diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,66565 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.222222222222222,
+  "eval_steps": 2250,
+  "global_step": 9500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00044444444444444447,
+      "grad_norm": 0.7025980353355408,
+      "learning_rate": 2e-05,
+      "loss": 3.0289,
+      "step": 1
+    },
+    {
+      "epoch": 0.0008888888888888889,
+      "grad_norm": 0.8140734434127808,
+      "learning_rate": 4e-05,
+      "loss": 3.1428,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013333333333333333,
+      "grad_norm": 0.8431426286697388,
+      "learning_rate": 6e-05,
+      "loss": 2.759,
+      "step": 3
+    },
+    {
+      "epoch": 0.0017777777777777779,
+      "grad_norm": 0.8204770088195801,
+      "learning_rate": 8e-05,
+      "loss": 3.0823,
+      "step": 4
+    },
+    {
+      "epoch": 0.0022222222222222222,
+      "grad_norm": 0.7714145183563232,
+      "learning_rate": 0.0001,
+      "loss": 2.9085,
+      "step": 5
+    },
+    {
+      "epoch": 0.0026666666666666666,
+      "grad_norm": 0.8233251571655273,
+      "learning_rate": 0.00012,
+      "loss": 2.926,
+      "step": 6
+    },
+    {
+      "epoch": 0.003111111111111111,
+      "grad_norm": 0.919024646282196,
+      "learning_rate": 0.00014,
+      "loss": 2.9215,
+      "step": 7
+    },
+    {
+      "epoch": 0.0035555555555555557,
+      "grad_norm": 0.996406078338623,
+      "learning_rate": 0.00016,
+      "loss": 2.6709,
+      "step": 8
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 1.156698226928711,
+      "learning_rate": 0.00018,
+      "loss": 2.936,
+      "step": 9
+    },
+    {
+      "epoch": 0.0044444444444444444,
+      "grad_norm": 1.6368826627731323,
+      "learning_rate": 0.0002,
+      "loss": 3.0452,
+      "step": 10
+    },
+    {
+      "epoch": 0.004888888888888889,
+      "grad_norm": 1.3638646602630615,
+      "learning_rate": 0.00019998220640569397,
+      "loss": 2.7221,
+      "step": 11
+    },
+    {
+      "epoch": 0.005333333333333333,
+      "grad_norm": 1.2367130517959595,
+      "learning_rate": 0.00019996441281138792,
+      "loss": 3.1021,
+      "step": 12
+    },
+    {
+      "epoch": 0.0057777777777777775,
+      "grad_norm": 1.1429595947265625,
+      "learning_rate": 0.00019994661921708185,
+      "loss": 2.5428,
+      "step": 13
+    },
+    {
+      "epoch": 0.006222222222222222,
+      "grad_norm": 1.3077034950256348,
+      "learning_rate": 0.0001999288256227758,
+      "loss": 2.7311,
+      "step": 14
+    },
+    {
+      "epoch": 0.006666666666666667,
+      "grad_norm": 1.2416568994522095,
+      "learning_rate": 0.00019991103202846976,
+      "loss": 2.3485,
+      "step": 15
+    },
+    {
+      "epoch": 0.0071111111111111115,
+      "grad_norm": 1.2917921543121338,
+      "learning_rate": 0.00019989323843416372,
+      "loss": 2.6843,
+      "step": 16
+    },
+    {
+      "epoch": 0.007555555555555556,
+      "grad_norm": 1.388919711112976,
+      "learning_rate": 0.00019987544483985765,
+      "loss": 2.5847,
+      "step": 17
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 1.4038573503494263,
+      "learning_rate": 0.0001998576512455516,
+      "loss": 2.8317,
+      "step": 18
+    },
+    {
+      "epoch": 0.008444444444444444,
+      "grad_norm": 1.3480195999145508,
+      "learning_rate": 0.00019983985765124556,
+      "loss": 2.9627,
+      "step": 19
+    },
+    {
+      "epoch": 0.008888888888888889,
+      "grad_norm": 1.1358228921890259,
+      "learning_rate": 0.0001998220640569395,
+      "loss": 2.6335,
+      "step": 20
+    },
+    {
+      "epoch": 0.009333333333333334,
+      "grad_norm": 1.3050692081451416,
+      "learning_rate": 0.00019980427046263345,
+      "loss": 2.7682,
+      "step": 21
+    },
+    {
+      "epoch": 0.009777777777777778,
+      "grad_norm": 1.1249107122421265,
+      "learning_rate": 0.0001997864768683274,
+      "loss": 2.5659,
+      "step": 22
+    },
+    {
+      "epoch": 0.010222222222222223,
+      "grad_norm": 2.0670816898345947,
+      "learning_rate": 0.00019976868327402136,
+      "loss": 3.6833,
+      "step": 23
+    },
+    {
+      "epoch": 0.010666666666666666,
+      "grad_norm": 1.3983336687088013,
+      "learning_rate": 0.00019975088967971532,
+      "loss": 2.2004,
+      "step": 24
+    },
+    {
+      "epoch": 0.011111111111111112,
+      "grad_norm": 1.2295548915863037,
+      "learning_rate": 0.00019973309608540928,
+      "loss": 2.5752,
+      "step": 25
+    },
+    {
+      "epoch": 0.011555555555555555,
+      "grad_norm": 1.4657237529754639,
+      "learning_rate": 0.0001997153024911032,
+      "loss": 2.5437,
+      "step": 26
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 1.2808438539505005,
+      "learning_rate": 0.00019969750889679716,
+      "loss": 2.5708,
+      "step": 27
+    },
+    {
+      "epoch": 0.012444444444444444,
+      "grad_norm": 1.811769723892212,
+      "learning_rate": 0.00019967971530249112,
+      "loss": 2.7687,
+      "step": 28
+    },
+    {
+      "epoch": 0.012888888888888889,
+      "grad_norm": 1.4724963903427124,
+      "learning_rate": 0.00019966192170818507,
+      "loss": 2.4946,
+      "step": 29
+    },
+    {
+      "epoch": 0.013333333333333334,
+      "grad_norm": 1.31075119972229,
+      "learning_rate": 0.000199644128113879,
+      "loss": 2.2154,
+      "step": 30
+    },
+    {
+      "epoch": 0.013777777777777778,
+      "grad_norm": 1.5601913928985596,
+      "learning_rate": 0.00019962633451957296,
+      "loss": 2.9217,
+      "step": 31
+    },
+    {
+      "epoch": 0.014222222222222223,
+      "grad_norm": 1.3747111558914185,
+      "learning_rate": 0.00019960854092526692,
+      "loss": 2.921,
+      "step": 32
+    },
+    {
+      "epoch": 0.014666666666666666,
+      "grad_norm": 1.3256399631500244,
+      "learning_rate": 0.00019959074733096085,
+      "loss": 2.664,
+      "step": 33
+    },
+    {
+      "epoch": 0.015111111111111112,
+      "grad_norm": 1.421920657157898,
+      "learning_rate": 0.0001995729537366548,
+      "loss": 2.4761,
+      "step": 34
+    },
+    {
+      "epoch": 0.015555555555555555,
+      "grad_norm": 1.6060773134231567,
+      "learning_rate": 0.00019955516014234876,
+      "loss": 2.585,
+      "step": 35
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 1.2819411754608154,
+      "learning_rate": 0.00019953736654804272,
+      "loss": 2.4084,
+      "step": 36
+    },
+    {
+      "epoch": 0.016444444444444446,
+      "grad_norm": 1.6595838069915771,
+      "learning_rate": 0.00019951957295373667,
+      "loss": 2.9492,
+      "step": 37
+    },
+    {
+      "epoch": 0.016888888888888887,
+      "grad_norm": 1.7327772378921509,
+      "learning_rate": 0.00019950177935943063,
+      "loss": 2.4173,
+      "step": 38
+    },
+    {
+      "epoch": 0.017333333333333333,
+      "grad_norm": 1.7617017030715942,
+      "learning_rate": 0.00019948398576512456,
+      "loss": 2.0839,
+      "step": 39
+    },
+    {
+      "epoch": 0.017777777777777778,
+      "grad_norm": 1.677414894104004,
+      "learning_rate": 0.00019946619217081851,
+      "loss": 2.3126,
+      "step": 40
+    },
+    {
+      "epoch": 0.018222222222222223,
+      "grad_norm": 1.3741059303283691,
+      "learning_rate": 0.00019944839857651247,
+      "loss": 2.5466,
+      "step": 41
+    },
+    {
+      "epoch": 0.018666666666666668,
+      "grad_norm": 1.7205222845077515,
+      "learning_rate": 0.00019943060498220643,
+      "loss": 2.7476,
+      "step": 42
+    },
+    {
+      "epoch": 0.01911111111111111,
+      "grad_norm": 1.9464102983474731,
+      "learning_rate": 0.00019941281138790036,
+      "loss": 2.8767,
+      "step": 43
+    },
+    {
+      "epoch": 0.019555555555555555,
+      "grad_norm": 1.5541801452636719,
+      "learning_rate": 0.00019939501779359431,
+      "loss": 2.4521,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6664501428604126,
+      "learning_rate": 0.00019937722419928827,
+      "loss": 2.1682,
+      "step": 45
+    },
+    {
+      "epoch": 0.020444444444444446,
+      "grad_norm": 1.555294394493103,
+      "learning_rate": 0.0001993594306049822,
+      "loss": 2.6263,
+      "step": 46
+    },
+    {
+      "epoch": 0.020888888888888887,
+      "grad_norm": 1.8338146209716797,
+      "learning_rate": 0.00019934163701067616,
+      "loss": 2.6912,
+      "step": 47
+    },
+    {
+      "epoch": 0.021333333333333333,
+      "grad_norm": 1.5415942668914795,
+      "learning_rate": 0.0001993238434163701,
+      "loss": 2.5132,
+      "step": 48
+    },
+    {
+      "epoch": 0.021777777777777778,
+      "grad_norm": 1.7458958625793457,
+      "learning_rate": 0.00019930604982206407,
+      "loss": 2.6587,
+      "step": 49
+    },
+    {
+      "epoch": 0.022222222222222223,
+      "grad_norm": 5.483662128448486,
+      "learning_rate": 0.00019928825622775803,
+      "loss": 2.7683,
+      "step": 50
+    },
+    {
+      "epoch": 0.02266666666666667,
+      "grad_norm": 0.7813481688499451,
+      "learning_rate": 0.00019927046263345198,
+      "loss": 2.3085,
+      "step": 51
+    },
+    {
+      "epoch": 0.02311111111111111,
+      "grad_norm": 0.9960127472877502,
+      "learning_rate": 0.0001992526690391459,
+      "loss": 2.4272,
+      "step": 52
+    },
+    {
+      "epoch": 0.023555555555555555,
+      "grad_norm": 1.1653705835342407,
+      "learning_rate": 0.00019923487544483987,
+      "loss": 2.7689,
+      "step": 53
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 1.2431868314743042,
+      "learning_rate": 0.00019921708185053382,
+      "loss": 2.7339,
+      "step": 54
+    },
+    {
+      "epoch": 0.024444444444444446,
+      "grad_norm": 1.22008216381073,
+      "learning_rate": 0.00019919928825622778,
+      "loss": 2.4097,
+      "step": 55
+    },
+    {
+      "epoch": 0.024888888888888887,
+      "grad_norm": 1.1046444177627563,
+      "learning_rate": 0.0001991814946619217,
+      "loss": 2.5895,
+      "step": 56
+    },
+    {
+      "epoch": 0.025333333333333333,
+      "grad_norm": 1.812303066253662,
+      "learning_rate": 0.00019916370106761567,
+      "loss": 2.7829,
+      "step": 57
+    },
+    {
+      "epoch": 0.025777777777777778,
+      "grad_norm": 1.241114854812622,
+      "learning_rate": 0.00019914590747330962,
+      "loss": 2.4605,
+      "step": 58
+    },
+    {
+      "epoch": 0.026222222222222223,
+      "grad_norm": 1.2486529350280762,
+      "learning_rate": 0.00019912811387900355,
+      "loss": 2.5066,
+      "step": 59
+    },
+    {
+      "epoch": 0.02666666666666667,
+      "grad_norm": 1.2300881147384644,
+      "learning_rate": 0.0001991103202846975,
+      "loss": 2.676,
+      "step": 60
+    },
+    {
+      "epoch": 0.02711111111111111,
+      "grad_norm": 1.840987205505371,
+      "learning_rate": 0.00019909252669039147,
+      "loss": 2.5316,
+      "step": 61
+    },
+    {
+      "epoch": 0.027555555555555555,
+      "grad_norm": 1.1190531253814697,
+      "learning_rate": 0.00019907473309608542,
+      "loss": 2.3772,
+      "step": 62
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 1.255723476409912,
+      "learning_rate": 0.00019905693950177938,
+      "loss": 2.4208,
+      "step": 63
+    },
+    {
+      "epoch": 0.028444444444444446,
+      "grad_norm": 1.0897091627120972,
+      "learning_rate": 0.00019903914590747334,
+      "loss": 2.5596,
+      "step": 64
+    },
+    {
+      "epoch": 0.028888888888888888,
+      "grad_norm": 1.1145886182785034,
+      "learning_rate": 0.00019902135231316726,
+      "loss": 2.428,
+      "step": 65
+    },
+    {
+      "epoch": 0.029333333333333333,
+      "grad_norm": 1.608787178993225,
+      "learning_rate": 0.00019900355871886122,
+      "loss": 2.956,
+      "step": 66
+    },
+    {
+      "epoch": 0.029777777777777778,
+      "grad_norm": 1.2278952598571777,
+      "learning_rate": 0.00019898576512455518,
+      "loss": 2.9204,
+      "step": 67
+    },
+    {
+      "epoch": 0.030222222222222223,
+      "grad_norm": 1.2748639583587646,
+      "learning_rate": 0.00019896797153024913,
+      "loss": 2.8291,
+      "step": 68
+    },
+    {
+      "epoch": 0.030666666666666665,
+      "grad_norm": 1.2492533922195435,
+      "learning_rate": 0.00019895017793594306,
+      "loss": 2.8622,
+      "step": 69
+    },
+    {
+      "epoch": 0.03111111111111111,
+      "grad_norm": 1.2926253080368042,
+      "learning_rate": 0.00019893238434163702,
+      "loss": 2.7478,
+      "step": 70
+    },
+    {
+      "epoch": 0.03155555555555556,
+      "grad_norm": 1.654407024383545,
+      "learning_rate": 0.00019891459074733098,
+      "loss": 2.4137,
+      "step": 71
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 1.4059809446334839,
+      "learning_rate": 0.0001988967971530249,
+      "loss": 3.0662,
+      "step": 72
+    },
+    {
+      "epoch": 0.03244444444444444,
+      "grad_norm": 1.5277940034866333,
+      "learning_rate": 0.00019887900355871886,
+      "loss": 2.4363,
+      "step": 73
+    },
+    {
+      "epoch": 0.03288888888888889,
+      "grad_norm": 1.141005277633667,
+      "learning_rate": 0.00019886120996441282,
+      "loss": 2.4846,
+      "step": 74
+    },
+    {
+      "epoch": 0.03333333333333333,
+      "grad_norm": 1.2516539096832275,
+      "learning_rate": 0.00019884341637010678,
+      "loss": 2.3899,
+      "step": 75
+    },
+    {
+      "epoch": 0.033777777777777775,
+      "grad_norm": 1.2361774444580078,
+      "learning_rate": 0.00019882562277580073,
+      "loss": 2.4988,
+      "step": 76
+    },
+    {
+      "epoch": 0.03422222222222222,
+      "grad_norm": 1.2464451789855957,
+      "learning_rate": 0.0001988078291814947,
+      "loss": 2.4915,
+      "step": 77
+    },
+    {
+      "epoch": 0.034666666666666665,
+      "grad_norm": 1.488785982131958,
+      "learning_rate": 0.00019879003558718862,
+      "loss": 2.6796,
+      "step": 78
+    },
+    {
+      "epoch": 0.035111111111111114,
+      "grad_norm": 1.3093085289001465,
+      "learning_rate": 0.00019877224199288257,
+      "loss": 2.6892,
+      "step": 79
+    },
+    {
+      "epoch": 0.035555555555555556,
+      "grad_norm": 1.1957430839538574,
+      "learning_rate": 0.00019875444839857653,
+      "loss": 2.5761,
+      "step": 80
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 1.3030658960342407,
+      "learning_rate": 0.00019873665480427046,
+      "loss": 2.6818,
+      "step": 81
+    },
+    {
+      "epoch": 0.036444444444444446,
+      "grad_norm": 1.2497376203536987,
+      "learning_rate": 0.00019871886120996442,
+      "loss": 2.2396,
+      "step": 82
+    },
+    {
+      "epoch": 0.03688888888888889,
+      "grad_norm": 1.1968713998794556,
+      "learning_rate": 0.00019870106761565837,
+      "loss": 2.6936,
+      "step": 83
+    },
+    {
+      "epoch": 0.037333333333333336,
+      "grad_norm": 1.5414577722549438,
+      "learning_rate": 0.00019868327402135233,
+      "loss": 2.4236,
+      "step": 84
+    },
+    {
+      "epoch": 0.03777777777777778,
+      "grad_norm": 1.4061691761016846,
+      "learning_rate": 0.00019866548042704626,
+      "loss": 2.0842,
+      "step": 85
+    },
+    {
+      "epoch": 0.03822222222222222,
+      "grad_norm": 1.3697423934936523,
+      "learning_rate": 0.00019864768683274022,
+      "loss": 2.6033,
+      "step": 86
+    },
+    {
+      "epoch": 0.03866666666666667,
+      "grad_norm": 1.4249686002731323,
+      "learning_rate": 0.00019862989323843417,
+      "loss": 2.3001,
+      "step": 87
+    },
+    {
+      "epoch": 0.03911111111111111,
+      "grad_norm": 1.449453592300415,
+      "learning_rate": 0.00019861209964412813,
+      "loss": 2.7934,
+      "step": 88
+    },
+    {
+      "epoch": 0.03955555555555555,
+      "grad_norm": 1.6140450239181519,
+      "learning_rate": 0.00019859430604982209,
+      "loss": 2.9546,
+      "step": 89
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3802794218063354,
+      "learning_rate": 0.00019857651245551604,
+      "loss": 2.6017,
+      "step": 90
+    },
+    {
+      "epoch": 0.04044444444444444,
+      "grad_norm": 1.4572910070419312,
+      "learning_rate": 0.00019855871886120997,
+      "loss": 2.9772,
+      "step": 91
+    },
+    {
+      "epoch": 0.04088888888888889,
+      "grad_norm": 1.6315029859542847,
+      "learning_rate": 0.00019854092526690393,
+      "loss": 3.0303,
+      "step": 92
+    },
+    {
+      "epoch": 0.04133333333333333,
+      "grad_norm": 1.6254206895828247,
+      "learning_rate": 0.00019852313167259788,
+      "loss": 2.9715,
+      "step": 93
+    },
+    {
+      "epoch": 0.041777777777777775,
+      "grad_norm": 1.2871061563491821,
+      "learning_rate": 0.0001985053380782918,
+      "loss": 2.3368,
+      "step": 94
+    },
+    {
+      "epoch": 0.042222222222222223,
+      "grad_norm": 1.4380096197128296,
+      "learning_rate": 0.00019848754448398577,
+      "loss": 2.5723,
+      "step": 95
+    },
+    {
+      "epoch": 0.042666666666666665,
+      "grad_norm": 1.5760232210159302,
+      "learning_rate": 0.00019846975088967973,
+      "loss": 2.5501,
+      "step": 96
+    },
+    {
+      "epoch": 0.043111111111111114,
+      "grad_norm": 1.6527888774871826,
+      "learning_rate": 0.00019845195729537368,
+      "loss": 2.6406,
+      "step": 97
+    },
+    {
+      "epoch": 0.043555555555555556,
+      "grad_norm": 1.8810604810714722,
+      "learning_rate": 0.0001984341637010676,
+      "loss": 2.8451,
+      "step": 98
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 2.4290010929107666,
+      "learning_rate": 0.00019841637010676157,
+      "loss": 2.3485,
+      "step": 99
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 2.3675355911254883,
+      "learning_rate": 0.00019839857651245553,
+      "loss": 2.8247,
+      "step": 100
+    },
+    {
+      "epoch": 0.04488888888888889,
+      "grad_norm": 0.7726467847824097,
+      "learning_rate": 0.00019838078291814948,
+      "loss": 2.5495,
+      "step": 101
+    },
+    {
+      "epoch": 0.04533333333333334,
+      "grad_norm": 1.0265134572982788,
+      "learning_rate": 0.00019836298932384344,
+      "loss": 2.7777,
+      "step": 102
+    },
+    {
+      "epoch": 0.04577777777777778,
+      "grad_norm": 0.9280586838722229,
+      "learning_rate": 0.00019834519572953737,
+      "loss": 2.8585,
+      "step": 103
+    },
+    {
+      "epoch": 0.04622222222222222,
+      "grad_norm": 0.897459089756012,
+      "learning_rate": 0.00019832740213523132,
+      "loss": 2.5307,
+      "step": 104
+    },
+    {
+      "epoch": 0.04666666666666667,
+      "grad_norm": 1.299446702003479,
+      "learning_rate": 0.00019830960854092528,
+      "loss": 2.6633,
+      "step": 105
+    },
+    {
+      "epoch": 0.04711111111111111,
+      "grad_norm": 1.0051794052124023,
+      "learning_rate": 0.00019829181494661924,
+      "loss": 2.714,
+      "step": 106
+    },
+    {
+      "epoch": 0.04755555555555555,
+      "grad_norm": 1.096691370010376,
+      "learning_rate": 0.00019827402135231317,
+      "loss": 2.5904,
+      "step": 107
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 0.9809961318969727,
+      "learning_rate": 0.00019825622775800712,
+      "loss": 2.7201,
+      "step": 108
+    },
+    {
+      "epoch": 0.04844444444444444,
+      "grad_norm": 1.2203365564346313,
+      "learning_rate": 0.00019823843416370108,
+      "loss": 2.7113,
+      "step": 109
+    },
+    {
+      "epoch": 0.04888888888888889,
+      "grad_norm": 1.2605012655258179,
+      "learning_rate": 0.000198220640569395,
+      "loss": 2.7018,
+      "step": 110
+    },
+    {
+      "epoch": 0.04933333333333333,
+      "grad_norm": 1.0020304918289185,
+      "learning_rate": 0.00019820284697508897,
+      "loss": 2.4043,
+      "step": 111
+    },
+    {
+      "epoch": 0.049777777777777775,
+      "grad_norm": 0.9287395477294922,
+      "learning_rate": 0.00019818505338078292,
+      "loss": 2.1282,
+      "step": 112
+    },
+    {
+      "epoch": 0.050222222222222224,
+      "grad_norm": 1.0499564409255981,
+      "learning_rate": 0.00019816725978647688,
+      "loss": 2.724,
+      "step": 113
+    },
+    {
+      "epoch": 0.050666666666666665,
+      "grad_norm": 0.9938886165618896,
+      "learning_rate": 0.00019814946619217083,
+      "loss": 2.8035,
+      "step": 114
+    },
+    {
+      "epoch": 0.051111111111111114,
+      "grad_norm": 0.9068772196769714,
+      "learning_rate": 0.0001981316725978648,
+      "loss": 2.2173,
+      "step": 115
+    },
+    {
+      "epoch": 0.051555555555555556,
+      "grad_norm": 0.9217369556427002,
+      "learning_rate": 0.00019811387900355872,
+      "loss": 2.266,
+      "step": 116
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 0.9447048306465149,
+      "learning_rate": 0.00019809608540925268,
+      "loss": 2.5042,
+      "step": 117
+    },
+    {
+      "epoch": 0.052444444444444446,
+      "grad_norm": 1.1115142107009888,
+      "learning_rate": 0.00019807829181494663,
+      "loss": 2.9463,
+      "step": 118
+    },
+    {
+      "epoch": 0.05288888888888889,
+      "grad_norm": 1.0305101871490479,
+      "learning_rate": 0.0001980604982206406,
+      "loss": 2.5201,
+      "step": 119
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 1.2101026773452759,
+      "learning_rate": 0.00019804270462633452,
+      "loss": 2.6229,
+      "step": 120
+    },
+    {
+      "epoch": 0.05377777777777778,
+      "grad_norm": 1.163856029510498,
+      "learning_rate": 0.00019802491103202848,
+      "loss": 2.8074,
+      "step": 121
+    },
+    {
+      "epoch": 0.05422222222222222,
+      "grad_norm": 1.2083830833435059,
+      "learning_rate": 0.00019800711743772243,
+      "loss": 2.7174,
+      "step": 122
+    },
+    {
+      "epoch": 0.05466666666666667,
+      "grad_norm": 1.169143795967102,
+      "learning_rate": 0.00019798932384341636,
+      "loss": 2.3256,
+      "step": 123
+    },
+    {
+      "epoch": 0.05511111111111111,
+      "grad_norm": 1.1567578315734863,
+      "learning_rate": 0.00019797153024911032,
+      "loss": 2.4596,
+      "step": 124
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 1.2948107719421387,
+      "learning_rate": 0.00019795373665480427,
+      "loss": 2.5357,
+      "step": 125
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 1.2189359664916992,
+      "learning_rate": 0.00019793594306049823,
+      "loss": 2.823,
+      "step": 126
+    },
+    {
+      "epoch": 0.05644444444444444,
+      "grad_norm": 1.5137442350387573,
+      "learning_rate": 0.0001979181494661922,
+      "loss": 3.2187,
+      "step": 127
+    },
+    {
+      "epoch": 0.05688888888888889,
+      "grad_norm": 1.2950530052185059,
+      "learning_rate": 0.00019790035587188614,
+      "loss": 2.6476,
+      "step": 128
+    },
+    {
+      "epoch": 0.05733333333333333,
+      "grad_norm": 1.1926493644714355,
+      "learning_rate": 0.00019788256227758007,
+      "loss": 1.9832,
+      "step": 129
+    },
+    {
+      "epoch": 0.057777777777777775,
+      "grad_norm": 1.3409109115600586,
+      "learning_rate": 0.00019786476868327403,
+      "loss": 2.5699,
+      "step": 130
+    },
+    {
+      "epoch": 0.058222222222222224,
+      "grad_norm": 1.2740064859390259,
+      "learning_rate": 0.000197846975088968,
+      "loss": 2.6332,
+      "step": 131
+    },
+    {
+      "epoch": 0.058666666666666666,
+      "grad_norm": 1.7328652143478394,
+      "learning_rate": 0.00019782918149466194,
+      "loss": 2.946,
+      "step": 132
+    },
+    {
+      "epoch": 0.059111111111111114,
+      "grad_norm": 1.3805177211761475,
+      "learning_rate": 0.00019781138790035587,
+      "loss": 2.7012,
+      "step": 133
+    },
+    {
+      "epoch": 0.059555555555555556,
+      "grad_norm": 1.3198126554489136,
+      "learning_rate": 0.00019779359430604983,
+      "loss": 2.6473,
+      "step": 134
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4987982511520386,
+      "learning_rate": 0.00019777580071174379,
+      "loss": 2.8902,
+      "step": 135
+    },
+    {
+      "epoch": 0.060444444444444446,
+      "grad_norm": 1.4471278190612793,
+      "learning_rate": 0.00019775800711743772,
+      "loss": 2.8084,
+      "step": 136
+    },
+    {
+      "epoch": 0.06088888888888889,
+      "grad_norm": 1.1906874179840088,
+      "learning_rate": 0.00019774021352313167,
+      "loss": 2.2734,
+      "step": 137
+    },
+    {
+      "epoch": 0.06133333333333333,
+      "grad_norm": 1.349488615989685,
+      "learning_rate": 0.00019772241992882563,
+      "loss": 2.734,
+      "step": 138
+    },
+    {
+      "epoch": 0.06177777777777778,
+      "grad_norm": 1.1277025938034058,
+      "learning_rate": 0.00019770462633451958,
+      "loss": 1.7254,
+      "step": 139
+    },
+    {
+      "epoch": 0.06222222222222222,
+      "grad_norm": 1.277053713798523,
+      "learning_rate": 0.00019768683274021354,
+      "loss": 2.5841,
+      "step": 140
+    },
+    {
+      "epoch": 0.06266666666666666,
+      "grad_norm": 1.358282446861267,
+      "learning_rate": 0.0001976690391459075,
+      "loss": 2.4615,
+      "step": 141
+    },
+    {
+      "epoch": 0.06311111111111112,
+      "grad_norm": 1.463334560394287,
+      "learning_rate": 0.00019765124555160143,
+      "loss": 2.8497,
+      "step": 142
+    },
+    {
+      "epoch": 0.06355555555555556,
+      "grad_norm": 1.537904143333435,
+      "learning_rate": 0.00019763345195729538,
+      "loss": 2.4246,
+      "step": 143
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 1.3648548126220703,
+      "learning_rate": 0.00019761565836298934,
+      "loss": 2.6535,
+      "step": 144
+    },
+    {
+      "epoch": 0.06444444444444444,
+      "grad_norm": 1.3705697059631348,
+      "learning_rate": 0.0001975978647686833,
+      "loss": 2.6365,
+      "step": 145
+    },
+    {
+      "epoch": 0.06488888888888888,
+      "grad_norm": 1.5270709991455078,
+      "learning_rate": 0.00019758007117437723,
+      "loss": 2.7863,
+      "step": 146
+    },
+    {
+      "epoch": 0.06533333333333333,
+      "grad_norm": 1.413665771484375,
+      "learning_rate": 0.00019756227758007118,
+      "loss": 2.4596,
+      "step": 147
+    },
+    {
+      "epoch": 0.06577777777777778,
+      "grad_norm": 1.7925262451171875,
+      "learning_rate": 0.00019754448398576514,
+      "loss": 3.2978,
+      "step": 148
+    },
+    {
+      "epoch": 0.06622222222222222,
+      "grad_norm": 1.5601551532745361,
+      "learning_rate": 0.00019752669039145907,
+      "loss": 2.1796,
+      "step": 149
+    },
+    {
+      "epoch": 0.06666666666666667,
+      "grad_norm": 1.8019471168518066,
+      "learning_rate": 0.00019750889679715302,
+      "loss": 2.5945,
+      "step": 150
+    },
+    {
+      "epoch": 0.06711111111111111,
+      "grad_norm": 0.816262423992157,
+      "learning_rate": 0.00019749110320284698,
+      "loss": 2.478,
+      "step": 151
+    },
+    {
+      "epoch": 0.06755555555555555,
+      "grad_norm": 0.9055171608924866,
+      "learning_rate": 0.00019747330960854094,
+      "loss": 2.6226,
+      "step": 152
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 0.8419144153594971,
+      "learning_rate": 0.0001974555160142349,
+      "loss": 2.2478,
+      "step": 153
+    },
+    {
+      "epoch": 0.06844444444444445,
+      "grad_norm": 1.01926589012146,
+      "learning_rate": 0.00019743772241992885,
+      "loss": 2.8489,
+      "step": 154
+    },
+    {
+      "epoch": 0.06888888888888889,
+      "grad_norm": 1.3037279844284058,
+      "learning_rate": 0.00019741992882562278,
+      "loss": 3.1187,
+      "step": 155
+    },
+    {
+      "epoch": 0.06933333333333333,
+      "grad_norm": 1.0188137292861938,
+      "learning_rate": 0.00019740213523131674,
+      "loss": 2.6388,
+      "step": 156
+    },
+    {
+      "epoch": 0.06977777777777777,
+      "grad_norm": 1.1099557876586914,
+      "learning_rate": 0.0001973843416370107,
+      "loss": 2.7675,
+      "step": 157
+    },
+    {
+      "epoch": 0.07022222222222223,
+      "grad_norm": 0.9259713292121887,
+      "learning_rate": 0.00019736654804270465,
+      "loss": 2.7232,
+      "step": 158
+    },
+    {
+      "epoch": 0.07066666666666667,
+      "grad_norm": 1.0099951028823853,
+      "learning_rate": 0.00019734875444839858,
+      "loss": 2.3996,
+      "step": 159
+    },
+    {
+      "epoch": 0.07111111111111111,
+      "grad_norm": 1.045190691947937,
+      "learning_rate": 0.00019733096085409254,
+      "loss": 2.5773,
+      "step": 160
+    },
+    {
+      "epoch": 0.07155555555555555,
+      "grad_norm": 1.1050662994384766,
+      "learning_rate": 0.0001973131672597865,
+      "loss": 2.1176,
+      "step": 161
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 1.2864255905151367,
+      "learning_rate": 0.00019729537366548042,
+      "loss": 3.2324,
+      "step": 162
+    },
+    {
+      "epoch": 0.07244444444444445,
+      "grad_norm": 1.0812265872955322,
+      "learning_rate": 0.00019727758007117438,
+      "loss": 2.671,
+      "step": 163
+    },
+    {
+      "epoch": 0.07288888888888889,
+      "grad_norm": 1.0999687910079956,
+      "learning_rate": 0.00019725978647686833,
+      "loss": 2.7204,
+      "step": 164
+    },
+    {
+      "epoch": 0.07333333333333333,
+      "grad_norm": 1.0504425764083862,
+      "learning_rate": 0.0001972419928825623,
+      "loss": 2.5729,
+      "step": 165
+    },
+    {
+      "epoch": 0.07377777777777778,
+      "grad_norm": 1.059959053993225,
+      "learning_rate": 0.00019722419928825625,
+      "loss": 2.4222,
+      "step": 166
+    },
+    {
+      "epoch": 0.07422222222222222,
+      "grad_norm": 1.0592875480651855,
+      "learning_rate": 0.0001972064056939502,
+      "loss": 2.4263,
+      "step": 167
+    },
+    {
+      "epoch": 0.07466666666666667,
+      "grad_norm": 1.0814061164855957,
+      "learning_rate": 0.00019718861209964413,
+      "loss": 2.5768,
+      "step": 168
+    },
+    {
+      "epoch": 0.07511111111111111,
+      "grad_norm": 1.416944980621338,
+      "learning_rate": 0.0001971708185053381,
+      "loss": 3.0183,
+      "step": 169
+    },
+    {
+      "epoch": 0.07555555555555556,
+      "grad_norm": 1.0500316619873047,
+      "learning_rate": 0.00019715302491103205,
+      "loss": 2.6289,
+      "step": 170
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 1.2884352207183838,
+      "learning_rate": 0.000197135231316726,
+      "loss": 2.5171,
+      "step": 171
+    },
+    {
+      "epoch": 0.07644444444444444,
+      "grad_norm": 1.3116530179977417,
+      "learning_rate": 0.00019711743772241993,
+      "loss": 3.0034,
+      "step": 172
+    },
+    {
+      "epoch": 0.0768888888888889,
+      "grad_norm": 1.2504932880401611,
+      "learning_rate": 0.0001970996441281139,
+      "loss": 2.5675,
+      "step": 173
+    },
+    {
+      "epoch": 0.07733333333333334,
+      "grad_norm": 1.1547982692718506,
+      "learning_rate": 0.00019708185053380785,
+      "loss": 2.301,
+      "step": 174
+    },
+    {
+      "epoch": 0.07777777777777778,
+      "grad_norm": 1.1116724014282227,
+      "learning_rate": 0.00019706405693950177,
+      "loss": 2.4003,
+      "step": 175
+    },
+    {
+      "epoch": 0.07822222222222222,
+      "grad_norm": 1.109155297279358,
+      "learning_rate": 0.00019704626334519573,
+      "loss": 2.2188,
+      "step": 176
+    },
+    {
+      "epoch": 0.07866666666666666,
+      "grad_norm": 1.2276736497879028,
+      "learning_rate": 0.0001970284697508897,
+      "loss": 2.5917,
+      "step": 177
+    },
+    {
+      "epoch": 0.0791111111111111,
+      "grad_norm": 1.5495067834854126,
+      "learning_rate": 0.00019701067615658364,
+      "loss": 2.5696,
+      "step": 178
+    },
+    {
+      "epoch": 0.07955555555555556,
+      "grad_norm": 1.2796157598495483,
+      "learning_rate": 0.0001969928825622776,
+      "loss": 2.3558,
+      "step": 179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.151888132095337,
+      "learning_rate": 0.00019697508896797156,
+      "loss": 1.779,
+      "step": 180
+    },
+    {
+      "epoch": 0.08044444444444444,
+      "grad_norm": 1.2485597133636475,
+      "learning_rate": 0.0001969572953736655,
+      "loss": 2.1977,
+      "step": 181
+    },
+    {
+      "epoch": 0.08088888888888889,
+      "grad_norm": 1.392452597618103,
+      "learning_rate": 0.00019693950177935944,
+      "loss": 2.6898,
+      "step": 182
+    },
+    {
+      "epoch": 0.08133333333333333,
+      "grad_norm": 1.514426350593567,
+      "learning_rate": 0.0001969217081850534,
+      "loss": 2.8083,
+      "step": 183
+    },
+    {
+      "epoch": 0.08177777777777778,
+      "grad_norm": 1.472489356994629,
+      "learning_rate": 0.00019690391459074736,
+      "loss": 2.9328,
+      "step": 184
+    },
+    {
+      "epoch": 0.08222222222222222,
+      "grad_norm": 1.5749987363815308,
+      "learning_rate": 0.00019688612099644129,
+      "loss": 2.3176,
+      "step": 185
+    },
+    {
+      "epoch": 0.08266666666666667,
+      "grad_norm": 1.5175185203552246,
+      "learning_rate": 0.00019686832740213524,
+      "loss": 2.7677,
+      "step": 186
+    },
+    {
+      "epoch": 0.08311111111111111,
+      "grad_norm": 1.286679983139038,
+      "learning_rate": 0.0001968505338078292,
+      "loss": 2.08,
+      "step": 187
+    },
+    {
+      "epoch": 0.08355555555555555,
+      "grad_norm": 1.4742923974990845,
+      "learning_rate": 0.00019683274021352313,
+      "loss": 3.0767,
+      "step": 188
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 1.3634746074676514,
+      "learning_rate": 0.00019681494661921708,
+      "loss": 2.423,
+      "step": 189
+    },
+    {
+      "epoch": 0.08444444444444445,
+      "grad_norm": 1.4298174381256104,
+      "learning_rate": 0.00019679715302491104,
+      "loss": 2.1539,
+      "step": 190
+    },
+    {
+      "epoch": 0.08488888888888889,
+      "grad_norm": 1.6024513244628906,
+      "learning_rate": 0.000196779359430605,
+      "loss": 2.868,
+      "step": 191
+    },
+    {
+      "epoch": 0.08533333333333333,
+      "grad_norm": 1.4882041215896606,
+      "learning_rate": 0.00019676156583629895,
+      "loss": 2.4238,
+      "step": 192
+    },
+    {
+      "epoch": 0.08577777777777777,
+      "grad_norm": 1.4021437168121338,
+      "learning_rate": 0.00019674377224199288,
+      "loss": 2.6468,
+      "step": 193
+    },
+    {
+      "epoch": 0.08622222222222223,
+      "grad_norm": 1.5105438232421875,
+      "learning_rate": 0.00019672597864768684,
+      "loss": 2.3776,
+      "step": 194
+    },
+    {
+      "epoch": 0.08666666666666667,
+      "grad_norm": 1.753899097442627,
+      "learning_rate": 0.0001967081850533808,
+      "loss": 2.4496,
+      "step": 195
+    },
+    {
+      "epoch": 0.08711111111111111,
+      "grad_norm": 1.6667553186416626,
+      "learning_rate": 0.00019669039145907475,
+      "loss": 2.7937,
+      "step": 196
+    },
+    {
+      "epoch": 0.08755555555555555,
+      "grad_norm": 1.4890007972717285,
+      "learning_rate": 0.00019667259786476868,
+      "loss": 2.7493,
+      "step": 197
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 1.8673200607299805,
+      "learning_rate": 0.00019665480427046264,
+      "loss": 2.7379,
+      "step": 198
+    },
+    {
+      "epoch": 0.08844444444444445,
+      "grad_norm": 2.127183675765991,
+      "learning_rate": 0.0001966370106761566,
+      "loss": 3.239,
+      "step": 199
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 2.974283218383789,
+      "learning_rate": 0.00019661921708185052,
+      "loss": 2.9018,
+      "step": 200
+    },
+    {
+      "epoch": 0.08933333333333333,
+      "grad_norm": 0.7585410475730896,
+      "learning_rate": 0.00019660142348754448,
+      "loss": 2.1856,
+      "step": 201
+    },
+    {
+      "epoch": 0.08977777777777778,
+      "grad_norm": 0.8838407397270203,
+      "learning_rate": 0.00019658362989323844,
+      "loss": 2.353,
+      "step": 202
+    },
+    {
+      "epoch": 0.09022222222222222,
+      "grad_norm": 0.9938530325889587,
+      "learning_rate": 0.0001965658362989324,
+      "loss": 2.5122,
+      "step": 203
+    },
+    {
+      "epoch": 0.09066666666666667,
+      "grad_norm": 0.8707981109619141,
+      "learning_rate": 0.00019654804270462635,
+      "loss": 2.5396,
+      "step": 204
+    },
+    {
+      "epoch": 0.09111111111111111,
+      "grad_norm": 0.9297839999198914,
+      "learning_rate": 0.0001965302491103203,
+      "loss": 2.7153,
+      "step": 205
+    },
+    {
+      "epoch": 0.09155555555555556,
+      "grad_norm": 1.176153302192688,
+      "learning_rate": 0.00019651245551601424,
+      "loss": 2.5667,
+      "step": 206
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 1.1235204935073853,
+      "learning_rate": 0.0001964946619217082,
+      "loss": 2.5297,
+      "step": 207
+    },
+    {
+      "epoch": 0.09244444444444444,
+      "grad_norm": 1.123063564300537,
+      "learning_rate": 0.00019647686832740215,
+      "loss": 2.8088,
+      "step": 208
+    },
+    {
+      "epoch": 0.09288888888888888,
+      "grad_norm": 1.0872026681900024,
+      "learning_rate": 0.0001964590747330961,
+      "loss": 2.5135,
+      "step": 209
+    },
+    {
+      "epoch": 0.09333333333333334,
+      "grad_norm": 1.1103240251541138,
+      "learning_rate": 0.00019644128113879004,
+      "loss": 2.6343,
+      "step": 210
+    },
+    {
+      "epoch": 0.09377777777777778,
+      "grad_norm": 0.998205840587616,
+      "learning_rate": 0.000196423487544484,
+      "loss": 2.5504,
+      "step": 211
+    },
+    {
+      "epoch": 0.09422222222222222,
+      "grad_norm": 0.9748513698577881,
+      "learning_rate": 0.00019640569395017795,
+      "loss": 2.2131,
+      "step": 212
+    },
+    {
+      "epoch": 0.09466666666666666,
+      "grad_norm": 1.0071059465408325,
+      "learning_rate": 0.00019638790035587188,
+      "loss": 2.2381,
+      "step": 213
+    },
+    {
+      "epoch": 0.0951111111111111,
+      "grad_norm": 1.4692164659500122,
+      "learning_rate": 0.00019637010676156583,
+      "loss": 2.6001,
+      "step": 214
+    },
+    {
+      "epoch": 0.09555555555555556,
+      "grad_norm": 1.2764703035354614,
+      "learning_rate": 0.0001963523131672598,
+      "loss": 2.7826,
+      "step": 215
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 1.0466008186340332,
+      "learning_rate": 0.00019633451957295375,
+      "loss": 2.7655,
+      "step": 216
+    },
+    {
+      "epoch": 0.09644444444444444,
+      "grad_norm": 1.1789090633392334,
+      "learning_rate": 0.0001963167259786477,
+      "loss": 2.5712,
+      "step": 217
+    },
+    {
+      "epoch": 0.09688888888888889,
+      "grad_norm": 1.240662932395935,
+      "learning_rate": 0.00019629893238434166,
+      "loss": 2.7105,
+      "step": 218
+    },
+    {
+      "epoch": 0.09733333333333333,
+      "grad_norm": 1.3624532222747803,
+      "learning_rate": 0.0001962811387900356,
+      "loss": 2.7762,
+      "step": 219
+    },
+    {
+      "epoch": 0.09777777777777778,
+      "grad_norm": 1.0563603639602661,
+      "learning_rate": 0.00019626334519572955,
+      "loss": 2.3591,
+      "step": 220
+    },
+    {
+      "epoch": 0.09822222222222222,
+      "grad_norm": 1.0683754682540894,
+      "learning_rate": 0.0001962455516014235,
+      "loss": 1.9874,
+      "step": 221
+    },
+    {
+      "epoch": 0.09866666666666667,
+      "grad_norm": 1.7027884721755981,
+      "learning_rate": 0.00019622775800711746,
+      "loss": 2.5819,
+      "step": 222
+    },
+    {
+      "epoch": 0.09911111111111111,
+      "grad_norm": 1.2697211503982544,
+      "learning_rate": 0.0001962099644128114,
+      "loss": 2.5147,
+      "step": 223
+    },
+    {
+      "epoch": 0.09955555555555555,
+      "grad_norm": 1.1141654253005981,
+      "learning_rate": 0.00019619217081850534,
+      "loss": 2.196,
+      "step": 224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.337015986442566,
+      "learning_rate": 0.0001961743772241993,
+      "loss": 2.5916,
+      "step": 225
+    },
+    {
+      "epoch": 0.10044444444444445,
+      "grad_norm": 1.3354969024658203,
+      "learning_rate": 0.00019615658362989323,
+      "loss": 2.3433,
+      "step": 226
+    },
+    {
+      "epoch": 0.10088888888888889,
+      "grad_norm": 1.2302438020706177,
+      "learning_rate": 0.0001961387900355872,
+      "loss": 2.3541,
+      "step": 227
+    },
+    {
+      "epoch": 0.10133333333333333,
+      "grad_norm": 1.8002538681030273,
+      "learning_rate": 0.00019612099644128114,
+      "loss": 2.6135,
+      "step": 228
+    },
+    {
+      "epoch": 0.10177777777777777,
+      "grad_norm": 1.4334473609924316,
+      "learning_rate": 0.0001961032028469751,
+      "loss": 2.7333,
+      "step": 229
+    },
+    {
+      "epoch": 0.10222222222222223,
+      "grad_norm": 1.5954945087432861,
+      "learning_rate": 0.00019608540925266906,
+      "loss": 2.6602,
+      "step": 230
+    },
+    {
+      "epoch": 0.10266666666666667,
+      "grad_norm": 1.3982867002487183,
+      "learning_rate": 0.000196067615658363,
+      "loss": 2.7748,
+      "step": 231
+    },
+    {
+      "epoch": 0.10311111111111111,
+      "grad_norm": 1.322675108909607,
+      "learning_rate": 0.00019604982206405694,
+      "loss": 2.4225,
+      "step": 232
+    },
+    {
+      "epoch": 0.10355555555555555,
+      "grad_norm": 1.3092737197875977,
+      "learning_rate": 0.0001960320284697509,
+      "loss": 2.479,
+      "step": 233
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 1.3040847778320312,
+      "learning_rate": 0.00019601423487544486,
+      "loss": 2.6299,
+      "step": 234
+    },
+    {
+      "epoch": 0.10444444444444445,
+      "grad_norm": 1.3706668615341187,
+      "learning_rate": 0.0001959964412811388,
+      "loss": 2.2323,
+      "step": 235
+    },
+    {
+      "epoch": 0.10488888888888889,
+      "grad_norm": 1.5371273756027222,
+      "learning_rate": 0.00019597864768683274,
+      "loss": 2.8628,
+      "step": 236
+    },
+    {
+      "epoch": 0.10533333333333333,
+      "grad_norm": 1.5061933994293213,
+      "learning_rate": 0.0001959608540925267,
+      "loss": 2.962,
+      "step": 237
+    },
+    {
+      "epoch": 0.10577777777777778,
+      "grad_norm": 1.4326897859573364,
+      "learning_rate": 0.00019594306049822065,
+      "loss": 2.6172,
+      "step": 238
+    },
+    {
+      "epoch": 0.10622222222222222,
+      "grad_norm": 1.7283401489257812,
+      "learning_rate": 0.00019592526690391458,
+      "loss": 3.1021,
+      "step": 239
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 1.5328441858291626,
+      "learning_rate": 0.00019590747330960854,
+      "loss": 2.9195,
+      "step": 240
+    },
+    {
+      "epoch": 0.10711111111111112,
+      "grad_norm": 1.4138455390930176,
+      "learning_rate": 0.0001958896797153025,
+      "loss": 2.7317,
+      "step": 241
+    },
+    {
+      "epoch": 0.10755555555555556,
+      "grad_norm": 1.336175560951233,
+      "learning_rate": 0.00019587188612099645,
+      "loss": 2.6417,
+      "step": 242
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 1.5993636846542358,
+      "learning_rate": 0.0001958540925266904,
+      "loss": 2.5646,
+      "step": 243
+    },
+    {
+      "epoch": 0.10844444444444444,
+      "grad_norm": 1.464353084564209,
+      "learning_rate": 0.00019583629893238437,
+      "loss": 2.6526,
+      "step": 244
+    },
+    {
+      "epoch": 0.10888888888888888,
+      "grad_norm": 1.731520652770996,
+      "learning_rate": 0.0001958185053380783,
+      "loss": 2.5564,
+      "step": 245
+    },
+    {
+      "epoch": 0.10933333333333334,
+      "grad_norm": 1.7385467290878296,
+      "learning_rate": 0.00019580071174377225,
+      "loss": 2.8463,
+      "step": 246
+    },
+    {
+      "epoch": 0.10977777777777778,
+      "grad_norm": 1.7964988946914673,
+      "learning_rate": 0.0001957829181494662,
+      "loss": 2.7856,
+      "step": 247
+    },
+    {
+      "epoch": 0.11022222222222222,
+      "grad_norm": 1.8664138317108154,
+      "learning_rate": 0.00019576512455516017,
+      "loss": 2.6548,
+      "step": 248
+    },
+    {
+      "epoch": 0.11066666666666666,
+      "grad_norm": 1.7917791604995728,
+      "learning_rate": 0.0001957473309608541,
+      "loss": 2.5387,
+      "step": 249
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 1.9322112798690796,
+      "learning_rate": 0.00019572953736654805,
+      "loss": 1.9883,
+      "step": 250
+    },
+    {
+      "epoch": 0.11155555555555556,
+      "grad_norm": 0.8669712543487549,
+      "learning_rate": 0.000195711743772242,
+      "loss": 2.6931,
+      "step": 251
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 1.020733118057251,
+      "learning_rate": 0.00019569395017793594,
+      "loss": 3.1699,
+      "step": 252
+    },
+    {
+      "epoch": 0.11244444444444444,
+      "grad_norm": 0.8693115711212158,
+      "learning_rate": 0.0001956761565836299,
+      "loss": 2.4069,
+      "step": 253
+    },
+    {
+      "epoch": 0.11288888888888889,
+      "grad_norm": 1.043466567993164,
+      "learning_rate": 0.00019565836298932385,
+      "loss": 2.9878,
+      "step": 254
+    },
+    {
+      "epoch": 0.11333333333333333,
+      "grad_norm": 1.030266284942627,
+      "learning_rate": 0.0001956405693950178,
+      "loss": 2.6809,
+      "step": 255
+    },
+    {
+      "epoch": 0.11377777777777778,
+      "grad_norm": 0.9637128710746765,
+      "learning_rate": 0.00019562277580071176,
+      "loss": 2.2864,
+      "step": 256
+    },
+    {
+      "epoch": 0.11422222222222222,
+      "grad_norm": 1.2764060497283936,
+      "learning_rate": 0.00019560498220640572,
+      "loss": 2.5282,
+      "step": 257
+    },
+    {
+      "epoch": 0.11466666666666667,
+      "grad_norm": 1.0020864009857178,
+      "learning_rate": 0.00019558718861209965,
+      "loss": 2.6671,
+      "step": 258
+    },
+    {
+      "epoch": 0.11511111111111111,
+      "grad_norm": 0.9636043310165405,
+      "learning_rate": 0.0001955693950177936,
+      "loss": 2.6084,
+      "step": 259
+    },
+    {
+      "epoch": 0.11555555555555555,
+      "grad_norm": 0.9684137105941772,
+      "learning_rate": 0.00019555160142348756,
+      "loss": 2.6087,
+      "step": 260
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 0.9589288830757141,
+      "learning_rate": 0.00019553380782918152,
+      "loss": 2.4366,
+      "step": 261
+    },
+    {
+      "epoch": 0.11644444444444445,
+      "grad_norm": 0.9377467632293701,
+      "learning_rate": 0.00019551601423487545,
+      "loss": 2.1606,
+      "step": 262
+    },
+    {
+      "epoch": 0.11688888888888889,
+      "grad_norm": 1.1523168087005615,
+      "learning_rate": 0.0001954982206405694,
+      "loss": 2.8838,
+      "step": 263
+    },
+    {
+      "epoch": 0.11733333333333333,
+      "grad_norm": 1.0509337186813354,
+      "learning_rate": 0.00019548042704626336,
+      "loss": 2.41,
+      "step": 264
+    },
+    {
+      "epoch": 0.11777777777777777,
+      "grad_norm": 1.2317067384719849,
+      "learning_rate": 0.0001954626334519573,
+      "loss": 2.5933,
+      "step": 265
+    },
+    {
+      "epoch": 0.11822222222222223,
+      "grad_norm": 1.4088350534439087,
+      "learning_rate": 0.00019544483985765125,
+      "loss": 2.7486,
+      "step": 266
+    },
+    {
+      "epoch": 0.11866666666666667,
+      "grad_norm": 1.033850073814392,
+      "learning_rate": 0.0001954270462633452,
+      "loss": 2.407,
+      "step": 267
+    },
+    {
+      "epoch": 0.11911111111111111,
+      "grad_norm": 1.2386589050292969,
+      "learning_rate": 0.00019540925266903916,
+      "loss": 2.1679,
+      "step": 268
+    },
+    {
+      "epoch": 0.11955555555555555,
+      "grad_norm": 1.0948667526245117,
+      "learning_rate": 0.00019539145907473312,
+      "loss": 2.4034,
+      "step": 269
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3226970434188843,
+      "learning_rate": 0.00019537366548042707,
+      "loss": 2.7977,
+      "step": 270
+    },
+    {
+      "epoch": 0.12044444444444445,
+      "grad_norm": 1.3416509628295898,
+      "learning_rate": 0.000195355871886121,
+      "loss": 2.7933,
+      "step": 271
+    },
+    {
+      "epoch": 0.12088888888888889,
+      "grad_norm": 1.142828345298767,
+      "learning_rate": 0.00019533807829181496,
+      "loss": 2.1703,
+      "step": 272
+    },
+    {
+      "epoch": 0.12133333333333333,
+      "grad_norm": 1.4413820505142212,
+      "learning_rate": 0.00019532028469750892,
+      "loss": 3.1676,
+      "step": 273
+    },
+    {
+      "epoch": 0.12177777777777778,
+      "grad_norm": 1.3678597211837769,
+      "learning_rate": 0.00019530249110320287,
+      "loss": 2.5917,
+      "step": 274
+    },
+    {
+      "epoch": 0.12222222222222222,
+      "grad_norm": 1.165423035621643,
+      "learning_rate": 0.0001952846975088968,
+      "loss": 2.1922,
+      "step": 275
+    },
+    {
+      "epoch": 0.12266666666666666,
+      "grad_norm": 1.458446741104126,
+      "learning_rate": 0.00019526690391459076,
+      "loss": 2.5036,
+      "step": 276
+    },
+    {
+      "epoch": 0.12311111111111112,
+      "grad_norm": 1.5351003408432007,
+      "learning_rate": 0.00019524911032028471,
+      "loss": 2.849,
+      "step": 277
+    },
+    {
+      "epoch": 0.12355555555555556,
+      "grad_norm": 1.343141794204712,
+      "learning_rate": 0.00019523131672597864,
+      "loss": 2.3158,
+      "step": 278
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 1.5747337341308594,
+      "learning_rate": 0.0001952135231316726,
+      "loss": 3.3124,
+      "step": 279
+    },
+    {
+      "epoch": 0.12444444444444444,
+      "grad_norm": 1.538909912109375,
+      "learning_rate": 0.00019519572953736656,
+      "loss": 2.7005,
+      "step": 280
+    },
+    {
+      "epoch": 0.12488888888888888,
+      "grad_norm": 1.4149315357208252,
+      "learning_rate": 0.0001951779359430605,
+      "loss": 1.8815,
+      "step": 281
+    },
+    {
+      "epoch": 0.12533333333333332,
+      "grad_norm": 1.2315411567687988,
+      "learning_rate": 0.00019516014234875447,
+      "loss": 2.4296,
+      "step": 282
+    },
+    {
+      "epoch": 0.12577777777777777,
+      "grad_norm": 1.5550092458724976,
+      "learning_rate": 0.0001951423487544484,
+      "loss": 3.0271,
+      "step": 283
+    },
+    {
+      "epoch": 0.12622222222222224,
+      "grad_norm": 1.3692402839660645,
+      "learning_rate": 0.00019512455516014236,
+      "loss": 2.5704,
+      "step": 284
+    },
+    {
+      "epoch": 0.12666666666666668,
+      "grad_norm": 1.4310396909713745,
+      "learning_rate": 0.0001951067615658363,
+      "loss": 2.4914,
+      "step": 285
+    },
+    {
+      "epoch": 0.12711111111111112,
+      "grad_norm": 1.283097505569458,
+      "learning_rate": 0.00019508896797153027,
+      "loss": 2.31,
+      "step": 286
+    },
+    {
+      "epoch": 0.12755555555555556,
+      "grad_norm": 1.3749858140945435,
+      "learning_rate": 0.00019507117437722422,
+      "loss": 2.2563,
+      "step": 287
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 1.3623768091201782,
+      "learning_rate": 0.00019505338078291815,
+      "loss": 2.2465,
+      "step": 288
+    },
+    {
+      "epoch": 0.12844444444444444,
+      "grad_norm": 1.4308843612670898,
+      "learning_rate": 0.0001950355871886121,
+      "loss": 2.6771,
+      "step": 289
+    },
+    {
+      "epoch": 0.1288888888888889,
+      "grad_norm": 1.568965196609497,
+      "learning_rate": 0.00019501779359430604,
+      "loss": 2.3809,
+      "step": 290
+    },
+    {
+      "epoch": 0.12933333333333333,
+      "grad_norm": 1.429734468460083,
+      "learning_rate": 0.000195,
+      "loss": 2.6331,
+      "step": 291
+    },
+    {
+      "epoch": 0.12977777777777777,
+      "grad_norm": 1.3805055618286133,
+      "learning_rate": 0.00019498220640569395,
+      "loss": 2.2138,
+      "step": 292
+    },
+    {
+      "epoch": 0.1302222222222222,
+      "grad_norm": 1.5100244283676147,
+      "learning_rate": 0.0001949644128113879,
+      "loss": 2.1832,
+      "step": 293
+    },
+    {
+      "epoch": 0.13066666666666665,
+      "grad_norm": 1.2256643772125244,
+      "learning_rate": 0.00019494661921708187,
+      "loss": 2.1298,
+      "step": 294
+    },
+    {
+      "epoch": 0.13111111111111112,
+      "grad_norm": 1.6592442989349365,
+      "learning_rate": 0.00019492882562277582,
+      "loss": 2.5226,
+      "step": 295
+    },
+    {
+      "epoch": 0.13155555555555556,
+      "grad_norm": 1.3523834943771362,
+      "learning_rate": 0.00019491103202846975,
+      "loss": 2.3038,
+      "step": 296
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 1.4921329021453857,
+      "learning_rate": 0.0001948932384341637,
+      "loss": 2.2571,
+      "step": 297
+    },
+    {
+      "epoch": 0.13244444444444445,
+      "grad_norm": 1.7550227642059326,
+      "learning_rate": 0.00019487544483985766,
+      "loss": 3.1873,
+      "step": 298
+    },
+    {
+      "epoch": 0.1328888888888889,
+      "grad_norm": 1.6412465572357178,
+      "learning_rate": 0.00019485765124555162,
+      "loss": 2.63,
+      "step": 299
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 1.9257419109344482,
+      "learning_rate": 0.00019483985765124558,
+      "loss": 2.4374,
+      "step": 300
+    },
+    {
+      "epoch": 0.13377777777777777,
+      "grad_norm": 0.9691830277442932,
+      "learning_rate": 0.0001948220640569395,
+      "loss": 3.1218,
+      "step": 301
+    },
+    {
+      "epoch": 0.13422222222222221,
+      "grad_norm": 0.8326031565666199,
+      "learning_rate": 0.00019480427046263346,
+      "loss": 2.5508,
+      "step": 302
+    },
+    {
+      "epoch": 0.13466666666666666,
+      "grad_norm": 0.9384410381317139,
+      "learning_rate": 0.0001947864768683274,
+      "loss": 2.5186,
+      "step": 303
+    },
+    {
+      "epoch": 0.1351111111111111,
+      "grad_norm": 1.1212788820266724,
+      "learning_rate": 0.00019476868327402135,
+      "loss": 2.1688,
+      "step": 304
+    },
+    {
+      "epoch": 0.13555555555555557,
+      "grad_norm": 0.8847819566726685,
+      "learning_rate": 0.0001947508896797153,
+      "loss": 2.4118,
+      "step": 305
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 1.0403448343276978,
+      "learning_rate": 0.00019473309608540926,
+      "loss": 2.5872,
+      "step": 306
+    },
+    {
+      "epoch": 0.13644444444444445,
+      "grad_norm": 1.0033663511276245,
+      "learning_rate": 0.00019471530249110322,
+      "loss": 2.4532,
+      "step": 307
+    },
+    {
+      "epoch": 0.1368888888888889,
+      "grad_norm": 0.9713349342346191,
+      "learning_rate": 0.00019469750889679718,
+      "loss": 2.6743,
+      "step": 308
+    },
+    {
+      "epoch": 0.13733333333333334,
+      "grad_norm": 0.9954231381416321,
+      "learning_rate": 0.0001946797153024911,
+      "loss": 2.2478,
+      "step": 309
+    },
+    {
+      "epoch": 0.13777777777777778,
+      "grad_norm": 1.4259557723999023,
+      "learning_rate": 0.00019466192170818506,
+      "loss": 2.4252,
+      "step": 310
+    },
+    {
+      "epoch": 0.13822222222222222,
+      "grad_norm": 1.1478148698806763,
+      "learning_rate": 0.00019464412811387902,
+      "loss": 2.6959,
+      "step": 311
+    },
+    {
+      "epoch": 0.13866666666666666,
+      "grad_norm": 0.9570370316505432,
+      "learning_rate": 0.00019462633451957297,
+      "loss": 2.4102,
+      "step": 312
+    },
+    {
+      "epoch": 0.1391111111111111,
+      "grad_norm": 1.0587913990020752,
+      "learning_rate": 0.0001946085409252669,
+      "loss": 2.341,
+      "step": 313
+    },
+    {
+      "epoch": 0.13955555555555554,
+      "grad_norm": 0.9694374203681946,
+      "learning_rate": 0.00019459074733096086,
+      "loss": 2.3627,
+      "step": 314
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0712873935699463,
+      "learning_rate": 0.00019457295373665482,
+      "loss": 2.7581,
+      "step": 315
+    },
+    {
+      "epoch": 0.14044444444444446,
+      "grad_norm": 1.1833367347717285,
+      "learning_rate": 0.00019455516014234875,
+      "loss": 2.424,
+      "step": 316
+    },
+    {
+      "epoch": 0.1408888888888889,
+      "grad_norm": 1.1975206136703491,
+      "learning_rate": 0.0001945373665480427,
+      "loss": 2.992,
+      "step": 317
+    },
+    {
+      "epoch": 0.14133333333333334,
+      "grad_norm": 1.165632724761963,
+      "learning_rate": 0.00019451957295373666,
+      "loss": 2.3119,
+      "step": 318
+    },
+    {
+      "epoch": 0.14177777777777778,
+      "grad_norm": 1.2752189636230469,
+      "learning_rate": 0.00019450177935943062,
+      "loss": 2.755,
+      "step": 319
+    },
+    {
+      "epoch": 0.14222222222222222,
+      "grad_norm": 1.0351862907409668,
+      "learning_rate": 0.00019448398576512457,
+      "loss": 2.4693,
+      "step": 320
+    },
+    {
+      "epoch": 0.14266666666666666,
+      "grad_norm": 1.3102034330368042,
+      "learning_rate": 0.00019446619217081853,
+      "loss": 3.0192,
+      "step": 321
+    },
+    {
+      "epoch": 0.1431111111111111,
+      "grad_norm": 1.2251161336898804,
+      "learning_rate": 0.00019444839857651246,
+      "loss": 2.6388,
+      "step": 322
+    },
+    {
+      "epoch": 0.14355555555555555,
+      "grad_norm": 1.147139072418213,
+      "learning_rate": 0.00019443060498220641,
+      "loss": 2.1927,
+      "step": 323
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 1.6661100387573242,
+      "learning_rate": 0.00019441281138790037,
+      "loss": 2.3797,
+      "step": 324
+    },
+    {
+      "epoch": 0.14444444444444443,
+      "grad_norm": 1.5821012258529663,
+      "learning_rate": 0.00019439501779359433,
+      "loss": 2.6176,
+      "step": 325
+    },
+    {
+      "epoch": 0.1448888888888889,
+      "grad_norm": 1.4074416160583496,
+      "learning_rate": 0.00019437722419928826,
+      "loss": 2.887,
+      "step": 326
+    },
+    {
+      "epoch": 0.14533333333333334,
+      "grad_norm": 1.1706616878509521,
+      "learning_rate": 0.0001943594306049822,
+      "loss": 2.5354,
+      "step": 327
+    },
+    {
+      "epoch": 0.14577777777777778,
+      "grad_norm": 1.8788220882415771,
+      "learning_rate": 0.00019434163701067617,
+      "loss": 3.0704,
+      "step": 328
+    },
+    {
+      "epoch": 0.14622222222222223,
+      "grad_norm": 1.3559796810150146,
+      "learning_rate": 0.0001943238434163701,
+      "loss": 2.5688,
+      "step": 329
+    },
+    {
+      "epoch": 0.14666666666666667,
+      "grad_norm": 1.3595597743988037,
+      "learning_rate": 0.00019430604982206406,
+      "loss": 2.4277,
+      "step": 330
+    },
+    {
+      "epoch": 0.1471111111111111,
+      "grad_norm": 1.267451286315918,
+      "learning_rate": 0.000194288256227758,
+      "loss": 2.6135,
+      "step": 331
+    },
+    {
+      "epoch": 0.14755555555555555,
+      "grad_norm": 1.2744300365447998,
+      "learning_rate": 0.00019427046263345197,
+      "loss": 2.5527,
+      "step": 332
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 1.6571002006530762,
+      "learning_rate": 0.00019425266903914593,
+      "loss": 2.653,
+      "step": 333
+    },
+    {
+      "epoch": 0.14844444444444443,
+      "grad_norm": 1.295233130455017,
+      "learning_rate": 0.00019423487544483988,
+      "loss": 2.3443,
+      "step": 334
+    },
+    {
+      "epoch": 0.14888888888888888,
+      "grad_norm": 1.4817813634872437,
+      "learning_rate": 0.0001942170818505338,
+      "loss": 2.6859,
+      "step": 335
+    },
+    {
+      "epoch": 0.14933333333333335,
+      "grad_norm": 1.2812329530715942,
+      "learning_rate": 0.00019419928825622777,
+      "loss": 2.5331,
+      "step": 336
+    },
+    {
+      "epoch": 0.1497777777777778,
+      "grad_norm": 1.2227575778961182,
+      "learning_rate": 0.00019418149466192172,
+      "loss": 2.2328,
+      "step": 337
+    },
+    {
+      "epoch": 0.15022222222222223,
+      "grad_norm": 1.2625856399536133,
+      "learning_rate": 0.00019416370106761568,
+      "loss": 2.3834,
+      "step": 338
+    },
+    {
+      "epoch": 0.15066666666666667,
+      "grad_norm": 1.4927825927734375,
+      "learning_rate": 0.0001941459074733096,
+      "loss": 2.66,
+      "step": 339
+    },
+    {
+      "epoch": 0.1511111111111111,
+      "grad_norm": 1.5926779508590698,
+      "learning_rate": 0.00019412811387900357,
+      "loss": 2.8668,
+      "step": 340
+    },
+    {
+      "epoch": 0.15155555555555555,
+      "grad_norm": 1.4455138444900513,
+      "learning_rate": 0.00019411032028469752,
+      "loss": 2.313,
+      "step": 341
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 1.7588112354278564,
+      "learning_rate": 0.00019409252669039145,
+      "loss": 3.2341,
+      "step": 342
+    },
+    {
+      "epoch": 0.15244444444444444,
+      "grad_norm": 1.61067795753479,
+      "learning_rate": 0.0001940747330960854,
+      "loss": 3.0195,
+      "step": 343
+    },
+    {
+      "epoch": 0.15288888888888888,
+      "grad_norm": 1.5478051900863647,
+      "learning_rate": 0.00019405693950177937,
+      "loss": 2.5716,
+      "step": 344
+    },
+    {
+      "epoch": 0.15333333333333332,
+      "grad_norm": 1.436558485031128,
+      "learning_rate": 0.00019403914590747332,
+      "loss": 2.4509,
+      "step": 345
+    },
+    {
+      "epoch": 0.1537777777777778,
+      "grad_norm": 1.6531201601028442,
+      "learning_rate": 0.00019402135231316728,
+      "loss": 2.8739,
+      "step": 346
+    },
+    {
+      "epoch": 0.15422222222222223,
+      "grad_norm": 1.6915215253829956,
+      "learning_rate": 0.00019400355871886124,
+      "loss": 2.6791,
+      "step": 347
+    },
+    {
+      "epoch": 0.15466666666666667,
+      "grad_norm": 1.5415655374526978,
+      "learning_rate": 0.00019398576512455516,
+      "loss": 2.8461,
+      "step": 348
+    },
+    {
+      "epoch": 0.15511111111111112,
+      "grad_norm": 1.5193134546279907,
+      "learning_rate": 0.00019396797153024912,
+      "loss": 2.0756,
+      "step": 349
+    },
+    {
+      "epoch": 0.15555555555555556,
+      "grad_norm": 1.9580684900283813,
+      "learning_rate": 0.00019395017793594308,
+      "loss": 2.2343,
+      "step": 350
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 0.9276076555252075,
+      "learning_rate": 0.00019393238434163703,
+      "loss": 2.5678,
+      "step": 351
+    },
+    {
+      "epoch": 0.15644444444444444,
+      "grad_norm": 0.8778754472732544,
+      "learning_rate": 0.00019391459074733096,
+      "loss": 2.2734,
+      "step": 352
+    },
+    {
+      "epoch": 0.15688888888888888,
+      "grad_norm": 0.9423462152481079,
+      "learning_rate": 0.00019389679715302492,
+      "loss": 2.3079,
+      "step": 353
+    },
+    {
+      "epoch": 0.15733333333333333,
+      "grad_norm": 1.1275615692138672,
+      "learning_rate": 0.00019387900355871888,
+      "loss": 3.0597,
+      "step": 354
+    },
+    {
+      "epoch": 0.15777777777777777,
+      "grad_norm": 1.098436713218689,
+      "learning_rate": 0.0001938612099644128,
+      "loss": 2.5681,
+      "step": 355
+    },
+    {
+      "epoch": 0.1582222222222222,
+      "grad_norm": 1.2952444553375244,
+      "learning_rate": 0.00019384341637010676,
+      "loss": 3.2901,
+      "step": 356
+    },
+    {
+      "epoch": 0.15866666666666668,
+      "grad_norm": 0.9615929126739502,
+      "learning_rate": 0.00019382562277580072,
+      "loss": 2.1562,
+      "step": 357
+    },
+    {
+      "epoch": 0.15911111111111112,
+      "grad_norm": 1.0871940851211548,
+      "learning_rate": 0.00019380782918149468,
+      "loss": 2.483,
+      "step": 358
+    },
+    {
+      "epoch": 0.15955555555555556,
+      "grad_norm": 1.2492108345031738,
+      "learning_rate": 0.00019379003558718863,
+      "loss": 2.4314,
+      "step": 359
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.184037208557129,
+      "learning_rate": 0.0001937722419928826,
+      "loss": 2.5729,
+      "step": 360
+    },
+    {
+      "epoch": 0.16044444444444445,
+      "grad_norm": 1.07174813747406,
+      "learning_rate": 0.00019375444839857652,
+      "loss": 2.4092,
+      "step": 361
+    },
+    {
+      "epoch": 0.1608888888888889,
+      "grad_norm": 1.034098744392395,
+      "learning_rate": 0.00019373665480427047,
+      "loss": 2.3074,
+      "step": 362
+    },
+    {
+      "epoch": 0.16133333333333333,
+      "grad_norm": 1.224602460861206,
+      "learning_rate": 0.00019371886120996443,
+      "loss": 2.5208,
+      "step": 363
+    },
+    {
+      "epoch": 0.16177777777777777,
+      "grad_norm": 1.1033531427383423,
+      "learning_rate": 0.0001937010676156584,
+      "loss": 2.4352,
+      "step": 364
+    },
+    {
+      "epoch": 0.1622222222222222,
+      "grad_norm": 1.3021239042282104,
+      "learning_rate": 0.00019368327402135232,
+      "loss": 3.0264,
+      "step": 365
+    },
+    {
+      "epoch": 0.16266666666666665,
+      "grad_norm": 1.0758678913116455,
+      "learning_rate": 0.00019366548042704627,
+      "loss": 2.1807,
+      "step": 366
+    },
+    {
+      "epoch": 0.16311111111111112,
+      "grad_norm": 1.262274980545044,
+      "learning_rate": 0.00019364768683274023,
+      "loss": 2.0729,
+      "step": 367
+    },
+    {
+      "epoch": 0.16355555555555557,
+      "grad_norm": 1.1292612552642822,
+      "learning_rate": 0.00019362989323843416,
+      "loss": 2.3755,
+      "step": 368
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 1.226879596710205,
+      "learning_rate": 0.00019361209964412812,
+      "loss": 2.7226,
+      "step": 369
+    },
+    {
+      "epoch": 0.16444444444444445,
+      "grad_norm": 1.2748644351959229,
+      "learning_rate": 0.00019359430604982207,
+      "loss": 2.8053,
+      "step": 370
+    },
+    {
+      "epoch": 0.1648888888888889,
+      "grad_norm": 1.2830379009246826,
+      "learning_rate": 0.00019357651245551603,
+      "loss": 2.6742,
+      "step": 371
+    },
+    {
+      "epoch": 0.16533333333333333,
+      "grad_norm": 1.3083009719848633,
+      "learning_rate": 0.00019355871886120998,
+      "loss": 2.5176,
+      "step": 372
+    },
+    {
+      "epoch": 0.16577777777777777,
+      "grad_norm": 1.2429115772247314,
+      "learning_rate": 0.00019354092526690391,
+      "loss": 2.4358,
+      "step": 373
+    },
+    {
+      "epoch": 0.16622222222222222,
+      "grad_norm": 1.3192554712295532,
+      "learning_rate": 0.00019352313167259787,
+      "loss": 2.3644,
+      "step": 374
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 1.1091505289077759,
+      "learning_rate": 0.00019350533807829183,
+      "loss": 2.0567,
+      "step": 375
+    },
+    {
+      "epoch": 0.1671111111111111,
+      "grad_norm": 1.3943992853164673,
+      "learning_rate": 0.00019348754448398578,
+      "loss": 2.3812,
+      "step": 376
+    },
+    {
+      "epoch": 0.16755555555555557,
+      "grad_norm": 1.3048312664031982,
+      "learning_rate": 0.00019346975088967974,
+      "loss": 2.614,
+      "step": 377
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 1.2868915796279907,
+      "learning_rate": 0.00019345195729537367,
+      "loss": 2.8753,
+      "step": 378
+    },
+    {
+      "epoch": 0.16844444444444445,
+      "grad_norm": 1.7796956300735474,
+      "learning_rate": 0.00019343416370106763,
+      "loss": 2.8598,
+      "step": 379
+    },
+    {
+      "epoch": 0.1688888888888889,
+      "grad_norm": 1.2406312227249146,
+      "learning_rate": 0.00019341637010676156,
+      "loss": 2.3268,
+      "step": 380
+    },
+    {
+      "epoch": 0.16933333333333334,
+      "grad_norm": 1.1642827987670898,
+      "learning_rate": 0.0001933985765124555,
+      "loss": 1.8269,
+      "step": 381
+    },
+    {
+      "epoch": 0.16977777777777778,
+      "grad_norm": 1.325958251953125,
+      "learning_rate": 0.00019338078291814947,
+      "loss": 2.6527,
+      "step": 382
+    },
+    {
+      "epoch": 0.17022222222222222,
+      "grad_norm": 1.4351950883865356,
+      "learning_rate": 0.00019336298932384343,
+      "loss": 2.7466,
+      "step": 383
+    },
+    {
+      "epoch": 0.17066666666666666,
+      "grad_norm": 1.4853794574737549,
+      "learning_rate": 0.00019334519572953738,
+      "loss": 2.7099,
+      "step": 384
+    },
+    {
+      "epoch": 0.1711111111111111,
+      "grad_norm": 1.2756681442260742,
+      "learning_rate": 0.00019332740213523134,
+      "loss": 1.9808,
+      "step": 385
+    },
+    {
+      "epoch": 0.17155555555555554,
+      "grad_norm": 1.567001223564148,
+      "learning_rate": 0.00019330960854092527,
+      "loss": 2.6314,
+      "step": 386
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 1.4116157293319702,
+      "learning_rate": 0.00019329181494661922,
+      "loss": 2.3277,
+      "step": 387
+    },
+    {
+      "epoch": 0.17244444444444446,
+      "grad_norm": 1.7159870862960815,
+      "learning_rate": 0.00019327402135231318,
+      "loss": 3.2206,
+      "step": 388
+    },
+    {
+      "epoch": 0.1728888888888889,
+      "grad_norm": 1.3646256923675537,
+      "learning_rate": 0.00019325622775800714,
+      "loss": 2.4943,
+      "step": 389
+    },
+    {
+      "epoch": 0.17333333333333334,
+      "grad_norm": 1.47772216796875,
+      "learning_rate": 0.0001932384341637011,
+      "loss": 2.3482,
+      "step": 390
+    },
+    {
+      "epoch": 0.17377777777777778,
+      "grad_norm": 1.6584293842315674,
+      "learning_rate": 0.00019322064056939502,
+      "loss": 2.449,
+      "step": 391
+    },
+    {
+      "epoch": 0.17422222222222222,
+      "grad_norm": 1.6674505472183228,
+      "learning_rate": 0.00019320284697508898,
+      "loss": 2.319,
+      "step": 392
+    },
+    {
+      "epoch": 0.17466666666666666,
+      "grad_norm": 1.4847460985183716,
+      "learning_rate": 0.0001931850533807829,
+      "loss": 2.4977,
+      "step": 393
+    },
+    {
+      "epoch": 0.1751111111111111,
+      "grad_norm": 1.5762537717819214,
+      "learning_rate": 0.00019316725978647687,
+      "loss": 2.5036,
+      "step": 394
+    },
+    {
+      "epoch": 0.17555555555555555,
+      "grad_norm": 1.731825828552246,
+      "learning_rate": 0.00019314946619217082,
+      "loss": 2.9634,
+      "step": 395
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 1.5546553134918213,
+      "learning_rate": 0.00019313167259786478,
+      "loss": 2.9042,
+      "step": 396
+    },
+    {
+      "epoch": 0.17644444444444443,
+      "grad_norm": 1.7094860076904297,
+      "learning_rate": 0.00019311387900355873,
+      "loss": 2.6971,
+      "step": 397
+    },
+    {
+      "epoch": 0.1768888888888889,
+      "grad_norm": 1.4419265985488892,
+      "learning_rate": 0.0001930960854092527,
+      "loss": 2.0126,
+      "step": 398
+    },
+    {
+      "epoch": 0.17733333333333334,
+      "grad_norm": 1.516395926475525,
+      "learning_rate": 0.00019307829181494662,
+      "loss": 1.9048,
+      "step": 399
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.010030508041382,
+      "learning_rate": 0.00019306049822064058,
+      "loss": 2.6672,
+      "step": 400
+    },
+    {
+      "epoch": 0.17822222222222223,
+      "grad_norm": 0.8796574473381042,
+      "learning_rate": 0.00019304270462633453,
+      "loss": 2.3451,
+      "step": 401
+    },
+    {
+      "epoch": 0.17866666666666667,
+      "grad_norm": 1.060291051864624,
+      "learning_rate": 0.0001930249110320285,
+      "loss": 1.5039,
+      "step": 402
+    },
+    {
+      "epoch": 0.1791111111111111,
+      "grad_norm": 0.9270257949829102,
+      "learning_rate": 0.00019300711743772245,
+      "loss": 2.3051,
+      "step": 403
+    },
+    {
+      "epoch": 0.17955555555555555,
+      "grad_norm": 1.0624326467514038,
+      "learning_rate": 0.00019298932384341638,
+      "loss": 2.4763,
+      "step": 404
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.396337866783142,
+      "learning_rate": 0.00019297153024911033,
+      "loss": 3.0648,
+      "step": 405
+    },
+    {
+      "epoch": 0.18044444444444444,
+      "grad_norm": 1.185214877128601,
+      "learning_rate": 0.00019295373665480426,
+      "loss": 2.5809,
+      "step": 406
+    },
+    {
+      "epoch": 0.18088888888888888,
+      "grad_norm": 1.2690224647521973,
+      "learning_rate": 0.00019293594306049822,
+      "loss": 2.6203,
+      "step": 407
+    },
+    {
+      "epoch": 0.18133333333333335,
+      "grad_norm": 1.083329439163208,
+      "learning_rate": 0.00019291814946619217,
+      "loss": 2.6254,
+      "step": 408
+    },
+    {
+      "epoch": 0.1817777777777778,
+      "grad_norm": 1.1971805095672607,
+      "learning_rate": 0.00019290035587188613,
+      "loss": 2.6854,
+      "step": 409
+    },
+    {
+      "epoch": 0.18222222222222223,
+      "grad_norm": 1.253471851348877,
+      "learning_rate": 0.0001928825622775801,
+      "loss": 2.7219,
+      "step": 410
+    },
+    {
+      "epoch": 0.18266666666666667,
+      "grad_norm": 1.1367192268371582,
+      "learning_rate": 0.00019286476868327404,
+      "loss": 2.6906,
+      "step": 411
+    },
+    {
+      "epoch": 0.1831111111111111,
+      "grad_norm": 1.1325358152389526,
+      "learning_rate": 0.00019284697508896797,
+      "loss": 2.6536,
+      "step": 412
+    },
+    {
+      "epoch": 0.18355555555555556,
+      "grad_norm": 1.1050721406936646,
+      "learning_rate": 0.00019282918149466193,
+      "loss": 2.351,
+      "step": 413
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 1.1547539234161377,
+      "learning_rate": 0.0001928113879003559,
+      "loss": 2.3844,
+      "step": 414
+    },
+    {
+      "epoch": 0.18444444444444444,
+      "grad_norm": 1.0745432376861572,
+      "learning_rate": 0.00019279359430604984,
+      "loss": 1.9506,
+      "step": 415
+    },
+    {
+      "epoch": 0.18488888888888888,
+      "grad_norm": 1.3409316539764404,
+      "learning_rate": 0.0001927758007117438,
+      "loss": 3.0778,
+      "step": 416
+    },
+    {
+      "epoch": 0.18533333333333332,
+      "grad_norm": 1.029353380203247,
+      "learning_rate": 0.00019275800711743773,
+      "loss": 2.1147,
+      "step": 417
+    },
+    {
+      "epoch": 0.18577777777777776,
+      "grad_norm": 1.3205243349075317,
+      "learning_rate": 0.00019274021352313169,
+      "loss": 2.706,
+      "step": 418
+    },
+    {
+      "epoch": 0.18622222222222223,
+      "grad_norm": 1.2287330627441406,
+      "learning_rate": 0.00019272241992882562,
+      "loss": 2.9919,
+      "step": 419
+    },
+    {
+      "epoch": 0.18666666666666668,
+      "grad_norm": 1.2103333473205566,
+      "learning_rate": 0.00019270462633451957,
+      "loss": 2.5828,
+      "step": 420
+    },
+    {
+      "epoch": 0.18711111111111112,
+      "grad_norm": 1.2801356315612793,
+      "learning_rate": 0.00019268683274021353,
+      "loss": 2.7884,
+      "step": 421
+    },
+    {
+      "epoch": 0.18755555555555556,
+      "grad_norm": 1.2993440628051758,
+      "learning_rate": 0.00019266903914590748,
+      "loss": 2.3698,
+      "step": 422
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 1.173684000968933,
+      "learning_rate": 0.00019265124555160144,
+      "loss": 1.9735,
+      "step": 423
+    },
+    {
+      "epoch": 0.18844444444444444,
+      "grad_norm": 1.1559852361679077,
+      "learning_rate": 0.0001926334519572954,
+      "loss": 2.3973,
+      "step": 424
+    },
+    {
+      "epoch": 0.18888888888888888,
+      "grad_norm": 1.1459964513778687,
+      "learning_rate": 0.00019261565836298933,
+      "loss": 2.2315,
+      "step": 425
+    },
+    {
+      "epoch": 0.18933333333333333,
+      "grad_norm": 1.2700178623199463,
+      "learning_rate": 0.00019259786476868328,
+      "loss": 2.0161,
+      "step": 426
+    },
+    {
+      "epoch": 0.18977777777777777,
+      "grad_norm": 1.4809290170669556,
+      "learning_rate": 0.00019258007117437724,
+      "loss": 2.4066,
+      "step": 427
+    },
+    {
+      "epoch": 0.1902222222222222,
+      "grad_norm": 1.3454186916351318,
+      "learning_rate": 0.0001925622775800712,
+      "loss": 2.5743,
+      "step": 428
+    },
+    {
+      "epoch": 0.19066666666666668,
+      "grad_norm": 1.7205144166946411,
+      "learning_rate": 0.00019254448398576513,
+      "loss": 2.6614,
+      "step": 429
+    },
+    {
+      "epoch": 0.19111111111111112,
+      "grad_norm": 1.3243727684020996,
+      "learning_rate": 0.00019252669039145908,
+      "loss": 2.5333,
+      "step": 430
+    },
+    {
+      "epoch": 0.19155555555555556,
+      "grad_norm": 1.282810926437378,
+      "learning_rate": 0.00019250889679715304,
+      "loss": 2.4967,
+      "step": 431
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 1.4963980913162231,
+      "learning_rate": 0.00019249110320284697,
+      "loss": 2.5604,
+      "step": 432
+    },
+    {
+      "epoch": 0.19244444444444445,
+      "grad_norm": 1.4316112995147705,
+      "learning_rate": 0.00019247330960854092,
+      "loss": 2.5011,
+      "step": 433
+    },
+    {
+      "epoch": 0.1928888888888889,
+      "grad_norm": 1.550047755241394,
+      "learning_rate": 0.00019245551601423488,
+      "loss": 2.2658,
+      "step": 434
+    },
+    {
+      "epoch": 0.19333333333333333,
+      "grad_norm": 1.3296672105789185,
+      "learning_rate": 0.00019243772241992884,
+      "loss": 2.3954,
+      "step": 435
+    },
+    {
+      "epoch": 0.19377777777777777,
+      "grad_norm": 1.4095767736434937,
+      "learning_rate": 0.0001924199288256228,
+      "loss": 2.4617,
+      "step": 436
+    },
+    {
+      "epoch": 0.1942222222222222,
+      "grad_norm": 1.5941203832626343,
+      "learning_rate": 0.00019240213523131675,
+      "loss": 2.4245,
+      "step": 437
+    },
+    {
+      "epoch": 0.19466666666666665,
+      "grad_norm": 1.427892804145813,
+      "learning_rate": 0.00019238434163701068,
+      "loss": 2.6397,
+      "step": 438
+    },
+    {
+      "epoch": 0.19511111111111112,
+      "grad_norm": 1.512261152267456,
+      "learning_rate": 0.00019236654804270464,
+      "loss": 3.0727,
+      "step": 439
+    },
+    {
+      "epoch": 0.19555555555555557,
+      "grad_norm": 1.3837703466415405,
+      "learning_rate": 0.0001923487544483986,
+      "loss": 2.0297,
+      "step": 440
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 1.7951322793960571,
+      "learning_rate": 0.00019233096085409255,
+      "loss": 2.9765,
+      "step": 441
+    },
+    {
+      "epoch": 0.19644444444444445,
+      "grad_norm": 1.418602705001831,
+      "learning_rate": 0.00019231316725978648,
+      "loss": 2.4067,
+      "step": 442
+    },
+    {
+      "epoch": 0.1968888888888889,
+      "grad_norm": 1.8720040321350098,
+      "learning_rate": 0.00019229537366548044,
+      "loss": 2.911,
+      "step": 443
+    },
+    {
+      "epoch": 0.19733333333333333,
+      "grad_norm": 1.337851881980896,
+      "learning_rate": 0.0001922775800711744,
+      "loss": 2.2034,
+      "step": 444
+    },
+    {
+      "epoch": 0.19777777777777777,
+      "grad_norm": 1.770947813987732,
+      "learning_rate": 0.00019225978647686832,
+      "loss": 3.1644,
+      "step": 445
+    },
+    {
+      "epoch": 0.19822222222222222,
+      "grad_norm": 1.5249742269515991,
+      "learning_rate": 0.00019224199288256228,
+      "loss": 3.0308,
+      "step": 446
+    },
+    {
+      "epoch": 0.19866666666666666,
+      "grad_norm": 1.9548603296279907,
+      "learning_rate": 0.00019222419928825623,
+      "loss": 3.2289,
+      "step": 447
+    },
+    {
+      "epoch": 0.1991111111111111,
+      "grad_norm": 1.8034451007843018,
+      "learning_rate": 0.0001922064056939502,
+      "loss": 2.9985,
+      "step": 448
+    },
+    {
+      "epoch": 0.19955555555555557,
+      "grad_norm": 2.0380022525787354,
+      "learning_rate": 0.00019218861209964415,
+      "loss": 2.9506,
+      "step": 449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.272326946258545,
+      "learning_rate": 0.0001921708185053381,
+      "loss": 2.9212,
+      "step": 450
+    },
+    {
+      "epoch": 0.20044444444444445,
+      "grad_norm": 0.8263271450996399,
+      "learning_rate": 0.00019215302491103203,
+      "loss": 2.646,
+      "step": 451
+    },
+    {
+      "epoch": 0.2008888888888889,
+      "grad_norm": 0.7942408919334412,
+      "learning_rate": 0.000192135231316726,
+      "loss": 2.1999,
+      "step": 452
+    },
+    {
+      "epoch": 0.20133333333333334,
+      "grad_norm": 1.059103012084961,
+      "learning_rate": 0.00019211743772241995,
+      "loss": 1.499,
+      "step": 453
+    },
+    {
+      "epoch": 0.20177777777777778,
+      "grad_norm": 1.2836692333221436,
+      "learning_rate": 0.0001920996441281139,
+      "loss": 2.1925,
+      "step": 454
+    },
+    {
+      "epoch": 0.20222222222222222,
+      "grad_norm": 0.9823219776153564,
+      "learning_rate": 0.00019208185053380783,
+      "loss": 2.3254,
+      "step": 455
+    },
+    {
+      "epoch": 0.20266666666666666,
+      "grad_norm": 1.1425485610961914,
+      "learning_rate": 0.0001920640569395018,
+      "loss": 2.6993,
+      "step": 456
+    },
+    {
+      "epoch": 0.2031111111111111,
+      "grad_norm": 1.2006306648254395,
+      "learning_rate": 0.00019204626334519575,
+      "loss": 2.8467,
+      "step": 457
+    },
+    {
+      "epoch": 0.20355555555555555,
+      "grad_norm": 1.0690330266952515,
+      "learning_rate": 0.00019202846975088967,
+      "loss": 2.5,
+      "step": 458
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 1.0791610479354858,
+      "learning_rate": 0.00019201067615658363,
+      "loss": 2.6261,
+      "step": 459
+    },
+    {
+      "epoch": 0.20444444444444446,
+      "grad_norm": 1.281191349029541,
+      "learning_rate": 0.0001919928825622776,
+      "loss": 2.6089,
+      "step": 460
+    },
+    {
+      "epoch": 0.2048888888888889,
+      "grad_norm": 0.9611422419548035,
+      "learning_rate": 0.00019197508896797154,
+      "loss": 2.6225,
+      "step": 461
+    },
+    {
+      "epoch": 0.20533333333333334,
+      "grad_norm": 1.1073009967803955,
+      "learning_rate": 0.0001919572953736655,
+      "loss": 2.2157,
+      "step": 462
+    },
+    {
+      "epoch": 0.20577777777777778,
+      "grad_norm": 1.305037498474121,
+      "learning_rate": 0.00019193950177935943,
+      "loss": 3.2172,
+      "step": 463
+    },
+    {
+      "epoch": 0.20622222222222222,
+      "grad_norm": 1.1567325592041016,
+      "learning_rate": 0.00019192170818505339,
+      "loss": 2.4355,
+      "step": 464
+    },
+    {
+      "epoch": 0.20666666666666667,
+      "grad_norm": 1.2676507234573364,
+      "learning_rate": 0.00019190391459074734,
+      "loss": 2.405,
+      "step": 465
+    },
+    {
+      "epoch": 0.2071111111111111,
+      "grad_norm": 1.1165955066680908,
+      "learning_rate": 0.0001918861209964413,
+      "loss": 2.0347,
+      "step": 466
+    },
+    {
+      "epoch": 0.20755555555555555,
+      "grad_norm": 1.3841233253479004,
+      "learning_rate": 0.00019186832740213526,
+      "loss": 2.7864,
+      "step": 467
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 1.4399226903915405,
+      "learning_rate": 0.00019185053380782919,
+      "loss": 3.6052,
+      "step": 468
+    },
+    {
+      "epoch": 0.20844444444444443,
+      "grad_norm": 1.4364230632781982,
+      "learning_rate": 0.00019183274021352314,
+      "loss": 2.7003,
+      "step": 469
+    },
+    {
+      "epoch": 0.2088888888888889,
+      "grad_norm": 1.1740696430206299,
+      "learning_rate": 0.00019181494661921707,
+      "loss": 2.1421,
+      "step": 470
+    },
+    {
+      "epoch": 0.20933333333333334,
+      "grad_norm": 1.5531638860702515,
+      "learning_rate": 0.00019179715302491103,
+      "loss": 2.6817,
+      "step": 471
+    },
+    {
+      "epoch": 0.20977777777777779,
+      "grad_norm": 1.4464926719665527,
+      "learning_rate": 0.00019177935943060498,
+      "loss": 2.7101,
+      "step": 472
+    },
+    {
+      "epoch": 0.21022222222222223,
+      "grad_norm": 1.319682002067566,
+      "learning_rate": 0.00019176156583629894,
+      "loss": 2.3335,
+      "step": 473
+    },
+    {
+      "epoch": 0.21066666666666667,
+      "grad_norm": 1.347642183303833,
+      "learning_rate": 0.0001917437722419929,
+      "loss": 2.6819,
+      "step": 474
+    },
+    {
+      "epoch": 0.2111111111111111,
+      "grad_norm": 1.3280656337738037,
+      "learning_rate": 0.00019172597864768685,
+      "loss": 2.7256,
+      "step": 475
+    },
+    {
+      "epoch": 0.21155555555555555,
+      "grad_norm": 1.4412258863449097,
+      "learning_rate": 0.00019170818505338078,
+      "loss": 2.7524,
+      "step": 476
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 1.455552577972412,
+      "learning_rate": 0.00019169039145907474,
+      "loss": 2.5749,
+      "step": 477
+    },
+    {
+      "epoch": 0.21244444444444444,
+      "grad_norm": 1.2440650463104248,
+      "learning_rate": 0.0001916725978647687,
+      "loss": 2.3623,
+      "step": 478
+    },
+    {
+      "epoch": 0.21288888888888888,
+      "grad_norm": 1.2427901029586792,
+      "learning_rate": 0.00019165480427046265,
+      "loss": 2.3891,
+      "step": 479
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 1.2674572467803955,
+      "learning_rate": 0.0001916370106761566,
+      "loss": 2.2005,
+      "step": 480
+    },
+    {
+      "epoch": 0.2137777777777778,
+      "grad_norm": 1.4567019939422607,
+      "learning_rate": 0.00019161921708185054,
+      "loss": 2.7974,
+      "step": 481
+    },
+    {
+      "epoch": 0.21422222222222223,
+      "grad_norm": 1.25277578830719,
+      "learning_rate": 0.0001916014234875445,
+      "loss": 2.4577,
+      "step": 482
+    },
+    {
+      "epoch": 0.21466666666666667,
+      "grad_norm": 1.2980494499206543,
+      "learning_rate": 0.00019158362989323842,
+      "loss": 2.4017,
+      "step": 483
+    },
+    {
+      "epoch": 0.21511111111111111,
+      "grad_norm": 1.5980355739593506,
+      "learning_rate": 0.00019156583629893238,
+      "loss": 2.7121,
+      "step": 484
+    },
+    {
+      "epoch": 0.21555555555555556,
+      "grad_norm": 1.3960875272750854,
+      "learning_rate": 0.00019154804270462634,
+      "loss": 2.4198,
+      "step": 485
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 1.5180373191833496,
+      "learning_rate": 0.0001915302491103203,
+      "loss": 2.8813,
+      "step": 486
+    },
+    {
+      "epoch": 0.21644444444444444,
+      "grad_norm": 1.339158058166504,
+      "learning_rate": 0.00019151245551601425,
+      "loss": 2.4161,
+      "step": 487
+    },
+    {
+      "epoch": 0.21688888888888888,
+      "grad_norm": 1.708709955215454,
+      "learning_rate": 0.0001914946619217082,
+      "loss": 2.6625,
+      "step": 488
+    },
+    {
+      "epoch": 0.21733333333333332,
+      "grad_norm": 1.4037717580795288,
+      "learning_rate": 0.00019147686832740214,
+      "loss": 2.7222,
+      "step": 489
+    },
+    {
+      "epoch": 0.21777777777777776,
+      "grad_norm": 1.547869324684143,
+      "learning_rate": 0.0001914590747330961,
+      "loss": 2.4619,
+      "step": 490
+    },
+    {
+      "epoch": 0.21822222222222223,
+      "grad_norm": 2.0343785285949707,
+      "learning_rate": 0.00019144128113879005,
+      "loss": 2.8051,
+      "step": 491
+    },
+    {
+      "epoch": 0.21866666666666668,
+      "grad_norm": 1.5703917741775513,
+      "learning_rate": 0.000191423487544484,
+      "loss": 2.7364,
+      "step": 492
+    },
+    {
+      "epoch": 0.21911111111111112,
+      "grad_norm": 1.4888960123062134,
+      "learning_rate": 0.00019140569395017796,
+      "loss": 2.3952,
+      "step": 493
+    },
+    {
+      "epoch": 0.21955555555555556,
+      "grad_norm": 1.9154101610183716,
+      "learning_rate": 0.0001913879003558719,
+      "loss": 3.4184,
+      "step": 494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8733478784561157,
+      "learning_rate": 0.00019137010676156585,
+      "loss": 2.0588,
+      "step": 495
+    },
+    {
+      "epoch": 0.22044444444444444,
+      "grad_norm": 1.5168616771697998,
+      "learning_rate": 0.00019135231316725978,
+      "loss": 2.4865,
+      "step": 496
+    },
+    {
+      "epoch": 0.22088888888888888,
+      "grad_norm": 1.6080540418624878,
+      "learning_rate": 0.00019133451957295373,
+      "loss": 2.6185,
+      "step": 497
+    },
+    {
+      "epoch": 0.22133333333333333,
+      "grad_norm": 1.8169959783554077,
+      "learning_rate": 0.0001913167259786477,
+      "loss": 2.7514,
+      "step": 498
+    },
+    {
+      "epoch": 0.22177777777777777,
+      "grad_norm": 1.903592824935913,
+      "learning_rate": 0.00019129893238434165,
+      "loss": 2.926,
+      "step": 499
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 2.2263951301574707,
+      "learning_rate": 0.0001912811387900356,
+      "loss": 3.209,
+      "step": 500
+    },
+    {
+      "epoch": 0.22266666666666668,
+      "grad_norm": 0.9277405142784119,
+      "learning_rate": 0.00019126334519572956,
+      "loss": 2.8295,
+      "step": 501
+    },
+    {
+      "epoch": 0.22311111111111112,
+      "grad_norm": 0.7616639137268066,
+      "learning_rate": 0.0001912455516014235,
+      "loss": 2.3259,
+      "step": 502
+    },
+    {
+      "epoch": 0.22355555555555556,
+      "grad_norm": 0.863919734954834,
+      "learning_rate": 0.00019122775800711745,
+      "loss": 2.6629,
+      "step": 503
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.9117692112922668,
+      "learning_rate": 0.0001912099644128114,
+      "loss": 2.8012,
+      "step": 504
+    },
+    {
+      "epoch": 0.22444444444444445,
+      "grad_norm": 0.8689172863960266,
+      "learning_rate": 0.00019119217081850536,
+      "loss": 1.8062,
+      "step": 505
+    },
+    {
+      "epoch": 0.2248888888888889,
+      "grad_norm": 0.9966077208518982,
+      "learning_rate": 0.00019117437722419932,
+      "loss": 2.3426,
+      "step": 506
+    },
+    {
+      "epoch": 0.22533333333333333,
+      "grad_norm": 1.1402056217193604,
+      "learning_rate": 0.00019115658362989324,
+      "loss": 2.597,
+      "step": 507
+    },
+    {
+      "epoch": 0.22577777777777777,
+      "grad_norm": 1.0207332372665405,
+      "learning_rate": 0.0001911387900355872,
+      "loss": 2.7771,
+      "step": 508
+    },
+    {
+      "epoch": 0.2262222222222222,
+      "grad_norm": 1.0293519496917725,
+      "learning_rate": 0.00019112099644128113,
+      "loss": 2.2381,
+      "step": 509
+    },
+    {
+      "epoch": 0.22666666666666666,
+      "grad_norm": 1.0531278848648071,
+      "learning_rate": 0.0001911032028469751,
+      "loss": 2.42,
+      "step": 510
+    },
+    {
+      "epoch": 0.22711111111111112,
+      "grad_norm": 1.2546653747558594,
+      "learning_rate": 0.00019108540925266904,
+      "loss": 2.9447,
+      "step": 511
+    },
+    {
+      "epoch": 0.22755555555555557,
+      "grad_norm": 1.0765845775604248,
+      "learning_rate": 0.000191067615658363,
+      "loss": 2.3876,
+      "step": 512
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 1.0224113464355469,
+      "learning_rate": 0.00019104982206405696,
+      "loss": 2.5661,
+      "step": 513
+    },
+    {
+      "epoch": 0.22844444444444445,
+      "grad_norm": 1.1143425703048706,
+      "learning_rate": 0.0001910320284697509,
+      "loss": 2.4567,
+      "step": 514
+    },
+    {
+      "epoch": 0.2288888888888889,
+      "grad_norm": 1.2478740215301514,
+      "learning_rate": 0.00019101423487544484,
+      "loss": 2.2325,
+      "step": 515
+    },
+    {
+      "epoch": 0.22933333333333333,
+      "grad_norm": 1.5122989416122437,
+      "learning_rate": 0.0001909964412811388,
+      "loss": 2.7624,
+      "step": 516
+    },
+    {
+      "epoch": 0.22977777777777778,
+      "grad_norm": 1.2084643840789795,
+      "learning_rate": 0.00019097864768683276,
+      "loss": 2.1843,
+      "step": 517
+    },
+    {
+      "epoch": 0.23022222222222222,
+      "grad_norm": 2.0436813831329346,
+      "learning_rate": 0.0001909608540925267,
+      "loss": 2.7025,
+      "step": 518
+    },
+    {
+      "epoch": 0.23066666666666666,
+      "grad_norm": 1.6114445924758911,
+      "learning_rate": 0.00019094306049822067,
+      "loss": 2.5077,
+      "step": 519
+    },
+    {
+      "epoch": 0.2311111111111111,
+      "grad_norm": 1.163203239440918,
+      "learning_rate": 0.0001909252669039146,
+      "loss": 2.6062,
+      "step": 520
+    },
+    {
+      "epoch": 0.23155555555555554,
+      "grad_norm": 1.3827770948410034,
+      "learning_rate": 0.00019090747330960855,
+      "loss": 2.565,
+      "step": 521
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 1.443726658821106,
+      "learning_rate": 0.00019088967971530248,
+      "loss": 2.9211,
+      "step": 522
+    },
+    {
+      "epoch": 0.23244444444444445,
+      "grad_norm": 1.1512651443481445,
+      "learning_rate": 0.00019087188612099644,
+      "loss": 2.3893,
+      "step": 523
+    },
+    {
+      "epoch": 0.2328888888888889,
+      "grad_norm": 1.3335007429122925,
+      "learning_rate": 0.0001908540925266904,
+      "loss": 2.8188,
+      "step": 524
+    },
+    {
+      "epoch": 0.23333333333333334,
+      "grad_norm": 1.2235959768295288,
+      "learning_rate": 0.00019083629893238435,
+      "loss": 2.2223,
+      "step": 525
+    },
+    {
+      "epoch": 0.23377777777777778,
+      "grad_norm": 1.3788108825683594,
+      "learning_rate": 0.0001908185053380783,
+      "loss": 2.7354,
+      "step": 526
+    },
+    {
+      "epoch": 0.23422222222222222,
+      "grad_norm": 1.400914192199707,
+      "learning_rate": 0.00019080071174377227,
+      "loss": 2.3977,
+      "step": 527
+    },
+    {
+      "epoch": 0.23466666666666666,
+      "grad_norm": 1.4983909130096436,
+      "learning_rate": 0.0001907829181494662,
+      "loss": 2.7346,
+      "step": 528
+    },
+    {
+      "epoch": 0.2351111111111111,
+      "grad_norm": 1.453970193862915,
+      "learning_rate": 0.00019076512455516015,
+      "loss": 2.4243,
+      "step": 529
+    },
+    {
+      "epoch": 0.23555555555555555,
+      "grad_norm": 1.6744136810302734,
+      "learning_rate": 0.0001907473309608541,
+      "loss": 0.8518,
+      "step": 530
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 1.4733753204345703,
+      "learning_rate": 0.00019072953736654807,
+      "loss": 2.5426,
+      "step": 531
+    },
+    {
+      "epoch": 0.23644444444444446,
+      "grad_norm": 1.4669400453567505,
+      "learning_rate": 0.00019071174377224202,
+      "loss": 2.527,
+      "step": 532
+    },
+    {
+      "epoch": 0.2368888888888889,
+      "grad_norm": 1.413023829460144,
+      "learning_rate": 0.00019069395017793595,
+      "loss": 2.5775,
+      "step": 533
+    },
+    {
+      "epoch": 0.23733333333333334,
+      "grad_norm": 1.4842833280563354,
+      "learning_rate": 0.0001906761565836299,
+      "loss": 2.3223,
+      "step": 534
+    },
+    {
+      "epoch": 0.23777777777777778,
+      "grad_norm": 1.6651334762573242,
+      "learning_rate": 0.00019065836298932384,
+      "loss": 2.7261,
+      "step": 535
+    },
+    {
+      "epoch": 0.23822222222222222,
+      "grad_norm": 2.1817784309387207,
+      "learning_rate": 0.0001906405693950178,
+      "loss": 2.602,
+      "step": 536
+    },
+    {
+      "epoch": 0.23866666666666667,
+      "grad_norm": 1.4185001850128174,
+      "learning_rate": 0.00019062277580071175,
+      "loss": 2.5273,
+      "step": 537
+    },
+    {
+      "epoch": 0.2391111111111111,
+      "grad_norm": 1.6372658014297485,
+      "learning_rate": 0.0001906049822064057,
+      "loss": 2.7267,
+      "step": 538
+    },
+    {
+      "epoch": 0.23955555555555555,
+      "grad_norm": 2.114755868911743,
+      "learning_rate": 0.00019058718861209966,
+      "loss": 1.3715,
+      "step": 539
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5371288061141968,
+      "learning_rate": 0.00019056939501779362,
+      "loss": 2.5673,
+      "step": 540
+    },
+    {
+      "epoch": 0.24044444444444443,
+      "grad_norm": 1.3606349229812622,
+      "learning_rate": 0.00019055160142348755,
+      "loss": 2.5102,
+      "step": 541
+    },
+    {
+      "epoch": 0.2408888888888889,
+      "grad_norm": 1.7038285732269287,
+      "learning_rate": 0.0001905338078291815,
+      "loss": 2.8909,
+      "step": 542
+    },
+    {
+      "epoch": 0.24133333333333334,
+      "grad_norm": 1.6660969257354736,
+      "learning_rate": 0.00019051601423487546,
+      "loss": 2.3934,
+      "step": 543
+    },
+    {
+      "epoch": 0.24177777777777779,
+      "grad_norm": 1.4915132522583008,
+      "learning_rate": 0.00019049822064056942,
+      "loss": 2.6195,
+      "step": 544
+    },
+    {
+      "epoch": 0.24222222222222223,
+      "grad_norm": 1.606236219406128,
+      "learning_rate": 0.00019048042704626335,
+      "loss": 2.5366,
+      "step": 545
+    },
+    {
+      "epoch": 0.24266666666666667,
+      "grad_norm": 1.6464382410049438,
+      "learning_rate": 0.0001904626334519573,
+      "loss": 2.2821,
+      "step": 546
+    },
+    {
+      "epoch": 0.2431111111111111,
+      "grad_norm": 1.5627448558807373,
+      "learning_rate": 0.00019044483985765126,
+      "loss": 2.7557,
+      "step": 547
+    },
+    {
+      "epoch": 0.24355555555555555,
+      "grad_norm": 1.5537325143814087,
+      "learning_rate": 0.0001904270462633452,
+      "loss": 2.7917,
+      "step": 548
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 2.0328938961029053,
+      "learning_rate": 0.00019040925266903915,
+      "loss": 3.1031,
+      "step": 549
+    },
+    {
+      "epoch": 0.24444444444444444,
+      "grad_norm": 2.5331227779388428,
+      "learning_rate": 0.0001903914590747331,
+      "loss": 2.0921,
+      "step": 550
+    },
+    {
+      "epoch": 0.24488888888888888,
+      "grad_norm": 1.062377691268921,
+      "learning_rate": 0.00019037366548042706,
+      "loss": 2.7696,
+      "step": 551
+    },
+    {
+      "epoch": 0.24533333333333332,
+      "grad_norm": 0.9012869596481323,
+      "learning_rate": 0.00019035587188612102,
+      "loss": 1.8978,
+      "step": 552
+    },
+    {
+      "epoch": 0.2457777777777778,
+      "grad_norm": 0.9942989349365234,
+      "learning_rate": 0.00019033807829181495,
+      "loss": 1.9979,
+      "step": 553
+    },
+    {
+      "epoch": 0.24622222222222223,
+      "grad_norm": 1.0721116065979004,
+      "learning_rate": 0.0001903202846975089,
+      "loss": 2.6083,
+      "step": 554
+    },
+    {
+      "epoch": 0.24666666666666667,
+      "grad_norm": 1.0755621194839478,
+      "learning_rate": 0.00019030249110320286,
+      "loss": 2.7578,
+      "step": 555
+    },
+    {
+      "epoch": 0.24711111111111111,
+      "grad_norm": 1.080788254737854,
+      "learning_rate": 0.00019028469750889681,
+      "loss": 2.2836,
+      "step": 556
+    },
+    {
+      "epoch": 0.24755555555555556,
+      "grad_norm": 1.0383445024490356,
+      "learning_rate": 0.00019026690391459077,
+      "loss": 2.4542,
+      "step": 557
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 1.1483523845672607,
+      "learning_rate": 0.0001902491103202847,
+      "loss": 2.5689,
+      "step": 558
+    },
+    {
+      "epoch": 0.24844444444444444,
+      "grad_norm": 1.041678547859192,
+      "learning_rate": 0.00019023131672597866,
+      "loss": 2.2751,
+      "step": 559
+    },
+    {
+      "epoch": 0.24888888888888888,
+      "grad_norm": 1.1849489212036133,
+      "learning_rate": 0.0001902135231316726,
+      "loss": 2.6915,
+      "step": 560
+    },
+    {
+      "epoch": 0.24933333333333332,
+      "grad_norm": 1.1290448904037476,
+      "learning_rate": 0.00019019572953736654,
+      "loss": 2.184,
+      "step": 561
+    },
+    {
+      "epoch": 0.24977777777777777,
+      "grad_norm": 1.2467504739761353,
+      "learning_rate": 0.0001901779359430605,
+      "loss": 2.2676,
+      "step": 562
+    },
+    {
+      "epoch": 0.25022222222222223,
+      "grad_norm": 1.1683002710342407,
+      "learning_rate": 0.00019016014234875446,
+      "loss": 2.4178,
+      "step": 563
+    },
+    {
+      "epoch": 0.25066666666666665,
+      "grad_norm": 1.2386951446533203,
+      "learning_rate": 0.0001901423487544484,
+      "loss": 2.4034,
+      "step": 564
+    },
+    {
+      "epoch": 0.2511111111111111,
+      "grad_norm": 1.259753704071045,
+      "learning_rate": 0.00019012455516014237,
+      "loss": 2.5685,
+      "step": 565
+    },
+    {
+      "epoch": 0.25155555555555553,
+      "grad_norm": 1.5166339874267578,
+      "learning_rate": 0.0001901067615658363,
+      "loss": 2.6635,
+      "step": 566
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 1.23752760887146,
+      "learning_rate": 0.00019008896797153026,
+      "loss": 2.4881,
+      "step": 567
+    },
+    {
+      "epoch": 0.25244444444444447,
+      "grad_norm": 1.2667707204818726,
+      "learning_rate": 0.0001900711743772242,
+      "loss": 2.8058,
+      "step": 568
+    },
+    {
+      "epoch": 0.2528888888888889,
+      "grad_norm": 1.8489893674850464,
+      "learning_rate": 0.00019005338078291817,
+      "loss": 2.6723,
+      "step": 569
+    },
+    {
+      "epoch": 0.25333333333333335,
+      "grad_norm": 1.2785292863845825,
+      "learning_rate": 0.00019003558718861212,
+      "loss": 2.8666,
+      "step": 570
+    },
+    {
+      "epoch": 0.25377777777777777,
+      "grad_norm": 1.141205906867981,
+      "learning_rate": 0.00019001779359430605,
+      "loss": 2.1319,
+      "step": 571
+    },
+    {
+      "epoch": 0.25422222222222224,
+      "grad_norm": 1.4991300106048584,
+      "learning_rate": 0.00019,
+      "loss": 2.3779,
+      "step": 572
+    },
+    {
+      "epoch": 0.25466666666666665,
+      "grad_norm": 1.2517198324203491,
+      "learning_rate": 0.00018998220640569394,
+      "loss": 2.7036,
+      "step": 573
+    },
+    {
+      "epoch": 0.2551111111111111,
+      "grad_norm": 1.186219334602356,
+      "learning_rate": 0.0001899644128113879,
+      "loss": 2.6787,
+      "step": 574
+    },
+    {
+      "epoch": 0.25555555555555554,
+      "grad_norm": 1.2609152793884277,
+      "learning_rate": 0.00018994661921708185,
+      "loss": 2.5269,
+      "step": 575
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 1.4722431898117065,
+      "learning_rate": 0.0001899288256227758,
+      "loss": 2.6287,
+      "step": 576
+    },
+    {
+      "epoch": 0.2564444444444444,
+      "grad_norm": 1.3302136659622192,
+      "learning_rate": 0.00018991103202846977,
+      "loss": 2.5439,
+      "step": 577
+    },
+    {
+      "epoch": 0.2568888888888889,
+      "grad_norm": 1.270352840423584,
+      "learning_rate": 0.00018989323843416372,
+      "loss": 2.5176,
+      "step": 578
+    },
+    {
+      "epoch": 0.25733333333333336,
+      "grad_norm": 1.2411810159683228,
+      "learning_rate": 0.00018987544483985765,
+      "loss": 2.423,
+      "step": 579
+    },
+    {
+      "epoch": 0.2577777777777778,
+      "grad_norm": 1.3175048828125,
+      "learning_rate": 0.0001898576512455516,
+      "loss": 2.364,
+      "step": 580
+    },
+    {
+      "epoch": 0.25822222222222224,
+      "grad_norm": 1.4399092197418213,
+      "learning_rate": 0.00018983985765124556,
+      "loss": 2.7466,
+      "step": 581
+    },
+    {
+      "epoch": 0.25866666666666666,
+      "grad_norm": 1.234508752822876,
+      "learning_rate": 0.00018982206405693952,
+      "loss": 2.1863,
+      "step": 582
+    },
+    {
+      "epoch": 0.2591111111111111,
+      "grad_norm": 1.6190673112869263,
+      "learning_rate": 0.00018980427046263348,
+      "loss": 2.4877,
+      "step": 583
+    },
+    {
+      "epoch": 0.25955555555555554,
+      "grad_norm": 1.159323811531067,
+      "learning_rate": 0.0001897864768683274,
+      "loss": 2.2116,
+      "step": 584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.284498929977417,
+      "learning_rate": 0.00018976868327402136,
+      "loss": 2.2746,
+      "step": 585
+    },
+    {
+      "epoch": 0.2604444444444444,
+      "grad_norm": 1.366461992263794,
+      "learning_rate": 0.0001897508896797153,
+      "loss": 2.5624,
+      "step": 586
+    },
+    {
+      "epoch": 0.2608888888888889,
+      "grad_norm": 1.4767354726791382,
+      "learning_rate": 0.00018973309608540925,
+      "loss": 2.5998,
+      "step": 587
+    },
+    {
+      "epoch": 0.2613333333333333,
+      "grad_norm": 1.6025152206420898,
+      "learning_rate": 0.0001897153024911032,
+      "loss": 2.1852,
+      "step": 588
+    },
+    {
+      "epoch": 0.2617777777777778,
+      "grad_norm": 1.54243803024292,
+      "learning_rate": 0.00018969750889679716,
+      "loss": 2.9229,
+      "step": 589
+    },
+    {
+      "epoch": 0.26222222222222225,
+      "grad_norm": 1.440328598022461,
+      "learning_rate": 0.00018967971530249112,
+      "loss": 2.3776,
+      "step": 590
+    },
+    {
+      "epoch": 0.26266666666666666,
+      "grad_norm": 1.5140371322631836,
+      "learning_rate": 0.00018966192170818508,
+      "loss": 3.043,
+      "step": 591
+    },
+    {
+      "epoch": 0.26311111111111113,
+      "grad_norm": 1.7295174598693848,
+      "learning_rate": 0.000189644128113879,
+      "loss": 2.8572,
+      "step": 592
+    },
+    {
+      "epoch": 0.26355555555555554,
+      "grad_norm": 1.5222134590148926,
+      "learning_rate": 0.00018962633451957296,
+      "loss": 2.2202,
+      "step": 593
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 1.484958529472351,
+      "learning_rate": 0.00018960854092526692,
+      "loss": 2.1051,
+      "step": 594
+    },
+    {
+      "epoch": 0.2644444444444444,
+      "grad_norm": 1.4371466636657715,
+      "learning_rate": 0.00018959074733096087,
+      "loss": 2.301,
+      "step": 595
+    },
+    {
+      "epoch": 0.2648888888888889,
+      "grad_norm": 1.6050223112106323,
+      "learning_rate": 0.00018957295373665483,
+      "loss": 2.4561,
+      "step": 596
+    },
+    {
+      "epoch": 0.2653333333333333,
+      "grad_norm": 1.7809783220291138,
+      "learning_rate": 0.00018955516014234876,
+      "loss": 2.3243,
+      "step": 597
+    },
+    {
+      "epoch": 0.2657777777777778,
+      "grad_norm": 1.850594401359558,
+      "learning_rate": 0.00018953736654804272,
+      "loss": 2.6873,
+      "step": 598
+    },
+    {
+      "epoch": 0.26622222222222225,
+      "grad_norm": 1.9856559038162231,
+      "learning_rate": 0.00018951957295373665,
+      "loss": 2.3005,
+      "step": 599
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 2.3018789291381836,
+      "learning_rate": 0.0001895017793594306,
+      "loss": 1.6366,
+      "step": 600
+    },
+    {
+      "epoch": 0.26711111111111113,
+      "grad_norm": 0.9843171834945679,
+      "learning_rate": 0.00018948398576512456,
+      "loss": 2.7333,
+      "step": 601
+    },
+    {
+      "epoch": 0.26755555555555555,
+      "grad_norm": 0.9362220764160156,
+      "learning_rate": 0.00018946619217081852,
+      "loss": 2.8176,
+      "step": 602
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 0.9775174260139465,
+      "learning_rate": 0.00018944839857651247,
+      "loss": 2.3173,
+      "step": 603
+    },
+    {
+      "epoch": 0.26844444444444443,
+      "grad_norm": 1.0477993488311768,
+      "learning_rate": 0.00018943060498220643,
+      "loss": 2.0655,
+      "step": 604
+    },
+    {
+      "epoch": 0.2688888888888889,
+      "grad_norm": 1.0463943481445312,
+      "learning_rate": 0.00018941281138790036,
+      "loss": 1.998,
+      "step": 605
+    },
+    {
+      "epoch": 0.2693333333333333,
+      "grad_norm": 1.0541325807571411,
+      "learning_rate": 0.00018939501779359431,
+      "loss": 2.4205,
+      "step": 606
+    },
+    {
+      "epoch": 0.2697777777777778,
+      "grad_norm": 1.0537536144256592,
+      "learning_rate": 0.00018937722419928827,
+      "loss": 2.8228,
+      "step": 607
+    },
+    {
+      "epoch": 0.2702222222222222,
+      "grad_norm": 1.0244420766830444,
+      "learning_rate": 0.00018935943060498223,
+      "loss": 2.3281,
+      "step": 608
+    },
+    {
+      "epoch": 0.27066666666666667,
+      "grad_norm": 1.3767787218093872,
+      "learning_rate": 0.00018934163701067618,
+      "loss": 3.1796,
+      "step": 609
+    },
+    {
+      "epoch": 0.27111111111111114,
+      "grad_norm": 1.03878915309906,
+      "learning_rate": 0.0001893238434163701,
+      "loss": 1.968,
+      "step": 610
+    },
+    {
+      "epoch": 0.27155555555555555,
+      "grad_norm": 1.1602753400802612,
+      "learning_rate": 0.00018930604982206407,
+      "loss": 2.6853,
+      "step": 611
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 1.0449435710906982,
+      "learning_rate": 0.000189288256227758,
+      "loss": 2.7955,
+      "step": 612
+    },
+    {
+      "epoch": 0.27244444444444443,
+      "grad_norm": 1.095615029335022,
+      "learning_rate": 0.00018927046263345196,
+      "loss": 2.299,
+      "step": 613
+    },
+    {
+      "epoch": 0.2728888888888889,
+      "grad_norm": 1.4768877029418945,
+      "learning_rate": 0.0001892526690391459,
+      "loss": 3.057,
+      "step": 614
+    },
+    {
+      "epoch": 0.2733333333333333,
+      "grad_norm": 1.276252269744873,
+      "learning_rate": 0.00018923487544483987,
+      "loss": 2.585,
+      "step": 615
+    },
+    {
+      "epoch": 0.2737777777777778,
+      "grad_norm": 1.2952446937561035,
+      "learning_rate": 0.00018921708185053383,
+      "loss": 2.5297,
+      "step": 616
+    },
+    {
+      "epoch": 0.2742222222222222,
+      "grad_norm": 1.2312525510787964,
+      "learning_rate": 0.00018919928825622778,
+      "loss": 2.6543,
+      "step": 617
+    },
+    {
+      "epoch": 0.27466666666666667,
+      "grad_norm": 1.368359923362732,
+      "learning_rate": 0.0001891814946619217,
+      "loss": 2.6951,
+      "step": 618
+    },
+    {
+      "epoch": 0.2751111111111111,
+      "grad_norm": 1.50626802444458,
+      "learning_rate": 0.00018916370106761567,
+      "loss": 2.7336,
+      "step": 619
+    },
+    {
+      "epoch": 0.27555555555555555,
+      "grad_norm": 1.2678533792495728,
+      "learning_rate": 0.00018914590747330962,
+      "loss": 2.4208,
+      "step": 620
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 1.167494773864746,
+      "learning_rate": 0.00018912811387900358,
+      "loss": 2.1091,
+      "step": 621
+    },
+    {
+      "epoch": 0.27644444444444444,
+      "grad_norm": 1.12721586227417,
+      "learning_rate": 0.00018911032028469754,
+      "loss": 2.0721,
+      "step": 622
+    },
+    {
+      "epoch": 0.2768888888888889,
+      "grad_norm": 1.306931495666504,
+      "learning_rate": 0.00018909252669039147,
+      "loss": 2.4418,
+      "step": 623
+    },
+    {
+      "epoch": 0.2773333333333333,
+      "grad_norm": 1.8457114696502686,
+      "learning_rate": 0.00018907473309608542,
+      "loss": 3.0008,
+      "step": 624
+    },
+    {
+      "epoch": 0.2777777777777778,
+      "grad_norm": 1.4332703351974487,
+      "learning_rate": 0.00018905693950177935,
+      "loss": 2.3379,
+      "step": 625
+    },
+    {
+      "epoch": 0.2782222222222222,
+      "grad_norm": 1.4976214170455933,
+      "learning_rate": 0.0001890391459074733,
+      "loss": 2.8298,
+      "step": 626
+    },
+    {
+      "epoch": 0.2786666666666667,
+      "grad_norm": 1.3851099014282227,
+      "learning_rate": 0.00018902135231316727,
+      "loss": 2.608,
+      "step": 627
+    },
+    {
+      "epoch": 0.2791111111111111,
+      "grad_norm": 1.3901604413986206,
+      "learning_rate": 0.00018900355871886122,
+      "loss": 2.4981,
+      "step": 628
+    },
+    {
+      "epoch": 0.27955555555555556,
+      "grad_norm": 1.5062224864959717,
+      "learning_rate": 0.00018898576512455518,
+      "loss": 2.4534,
+      "step": 629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3895263671875,
+      "learning_rate": 0.00018896797153024913,
+      "loss": 2.5953,
+      "step": 630
+    },
+    {
+      "epoch": 0.28044444444444444,
+      "grad_norm": 1.5202879905700684,
+      "learning_rate": 0.00018895017793594306,
+      "loss": 2.2741,
+      "step": 631
+    },
+    {
+      "epoch": 0.2808888888888889,
+      "grad_norm": 1.2481839656829834,
+      "learning_rate": 0.00018893238434163702,
+      "loss": 2.3242,
+      "step": 632
+    },
+    {
+      "epoch": 0.2813333333333333,
+      "grad_norm": 1.2710275650024414,
+      "learning_rate": 0.00018891459074733098,
+      "loss": 2.1379,
+      "step": 633
+    },
+    {
+      "epoch": 0.2817777777777778,
+      "grad_norm": 1.3879283666610718,
+      "learning_rate": 0.00018889679715302493,
+      "loss": 1.353,
+      "step": 634
+    },
+    {
+      "epoch": 0.2822222222222222,
+      "grad_norm": 1.5457983016967773,
+      "learning_rate": 0.0001888790035587189,
+      "loss": 2.3432,
+      "step": 635
+    },
+    {
+      "epoch": 0.2826666666666667,
+      "grad_norm": 1.5545676946640015,
+      "learning_rate": 0.00018886120996441282,
+      "loss": 2.6554,
+      "step": 636
+    },
+    {
+      "epoch": 0.2831111111111111,
+      "grad_norm": 1.2374818325042725,
+      "learning_rate": 0.00018884341637010678,
+      "loss": 1.4633,
+      "step": 637
+    },
+    {
+      "epoch": 0.28355555555555556,
+      "grad_norm": 1.3478444814682007,
+      "learning_rate": 0.0001888256227758007,
+      "loss": 2.4976,
+      "step": 638
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 1.5220305919647217,
+      "learning_rate": 0.00018880782918149466,
+      "loss": 2.6628,
+      "step": 639
+    },
+    {
+      "epoch": 0.28444444444444444,
+      "grad_norm": 1.6763924360275269,
+      "learning_rate": 0.00018879003558718862,
+      "loss": 2.7547,
+      "step": 640
+    },
+    {
+      "epoch": 0.2848888888888889,
+      "grad_norm": 1.5462572574615479,
+      "learning_rate": 0.00018877224199288258,
+      "loss": 2.6934,
+      "step": 641
+    },
+    {
+      "epoch": 0.2853333333333333,
+      "grad_norm": 1.9419124126434326,
+      "learning_rate": 0.00018875444839857653,
+      "loss": 2.5849,
+      "step": 642
+    },
+    {
+      "epoch": 0.2857777777777778,
+      "grad_norm": 1.6451220512390137,
+      "learning_rate": 0.00018873665480427046,
+      "loss": 2.7719,
+      "step": 643
+    },
+    {
+      "epoch": 0.2862222222222222,
+      "grad_norm": 1.5274759531021118,
+      "learning_rate": 0.00018871886120996442,
+      "loss": 2.2116,
+      "step": 644
+    },
+    {
+      "epoch": 0.2866666666666667,
+      "grad_norm": 2.1637439727783203,
+      "learning_rate": 0.00018870106761565837,
+      "loss": 3.0492,
+      "step": 645
+    },
+    {
+      "epoch": 0.2871111111111111,
+      "grad_norm": 1.6823992729187012,
+      "learning_rate": 0.00018868327402135233,
+      "loss": 2.7498,
+      "step": 646
+    },
+    {
+      "epoch": 0.28755555555555556,
+      "grad_norm": 1.695084810256958,
+      "learning_rate": 0.0001886654804270463,
+      "loss": 2.6973,
+      "step": 647
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 1.8430315256118774,
+      "learning_rate": 0.00018864768683274024,
+      "loss": 1.9394,
+      "step": 648
+    },
+    {
+      "epoch": 0.28844444444444445,
+      "grad_norm": 2.2120563983917236,
+      "learning_rate": 0.00018862989323843417,
+      "loss": 1.9077,
+      "step": 649
+    },
+    {
+      "epoch": 0.28888888888888886,
+      "grad_norm": 2.309023141860962,
+      "learning_rate": 0.0001886120996441281,
+      "loss": 2.15,
+      "step": 650
+    },
+    {
+      "epoch": 0.28933333333333333,
+      "grad_norm": 1.1258161067962646,
+      "learning_rate": 0.00018859430604982206,
+      "loss": 2.9532,
+      "step": 651
+    },
+    {
+      "epoch": 0.2897777777777778,
+      "grad_norm": 0.9455536603927612,
+      "learning_rate": 0.00018857651245551602,
+      "loss": 2.695,
+      "step": 652
+    },
+    {
+      "epoch": 0.2902222222222222,
+      "grad_norm": 1.228993535041809,
+      "learning_rate": 0.00018855871886120997,
+      "loss": 1.3139,
+      "step": 653
+    },
+    {
+      "epoch": 0.2906666666666667,
+      "grad_norm": 1.0988825559616089,
+      "learning_rate": 0.00018854092526690393,
+      "loss": 2.3994,
+      "step": 654
+    },
+    {
+      "epoch": 0.2911111111111111,
+      "grad_norm": 1.0510218143463135,
+      "learning_rate": 0.00018852313167259788,
+      "loss": 2.9534,
+      "step": 655
+    },
+    {
+      "epoch": 0.29155555555555557,
+      "grad_norm": 1.1386710405349731,
+      "learning_rate": 0.00018850533807829181,
+      "loss": 2.2164,
+      "step": 656
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 1.1761900186538696,
+      "learning_rate": 0.00018848754448398577,
+      "loss": 2.6014,
+      "step": 657
+    },
+    {
+      "epoch": 0.29244444444444445,
+      "grad_norm": 1.0478448867797852,
+      "learning_rate": 0.00018846975088967973,
+      "loss": 2.497,
+      "step": 658
+    },
+    {
+      "epoch": 0.29288888888888887,
+      "grad_norm": 1.2289860248565674,
+      "learning_rate": 0.00018845195729537368,
+      "loss": 3.0106,
+      "step": 659
+    },
+    {
+      "epoch": 0.29333333333333333,
+      "grad_norm": 1.2881073951721191,
+      "learning_rate": 0.00018843416370106764,
+      "loss": 2.5109,
+      "step": 660
+    },
+    {
+      "epoch": 0.2937777777777778,
+      "grad_norm": 1.2944267988204956,
+      "learning_rate": 0.00018841637010676157,
+      "loss": 2.3959,
+      "step": 661
+    },
+    {
+      "epoch": 0.2942222222222222,
+      "grad_norm": 1.1457511186599731,
+      "learning_rate": 0.00018839857651245553,
+      "loss": 2.5717,
+      "step": 662
+    },
+    {
+      "epoch": 0.2946666666666667,
+      "grad_norm": 1.288615345954895,
+      "learning_rate": 0.00018838078291814946,
+      "loss": 2.6314,
+      "step": 663
+    },
+    {
+      "epoch": 0.2951111111111111,
+      "grad_norm": 1.153175711631775,
+      "learning_rate": 0.0001883629893238434,
+      "loss": 2.1529,
+      "step": 664
+    },
+    {
+      "epoch": 0.29555555555555557,
+      "grad_norm": 1.0861622095108032,
+      "learning_rate": 0.00018834519572953737,
+      "loss": 2.2134,
+      "step": 665
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 1.3027865886688232,
+      "learning_rate": 0.00018832740213523132,
+      "loss": 2.496,
+      "step": 666
+    },
+    {
+      "epoch": 0.29644444444444445,
+      "grad_norm": 1.2084095478057861,
+      "learning_rate": 0.00018830960854092528,
+      "loss": 2.6488,
+      "step": 667
+    },
+    {
+      "epoch": 0.29688888888888887,
+      "grad_norm": 1.2126069068908691,
+      "learning_rate": 0.00018829181494661924,
+      "loss": 2.2976,
+      "step": 668
+    },
+    {
+      "epoch": 0.29733333333333334,
+      "grad_norm": 1.392379879951477,
+      "learning_rate": 0.00018827402135231317,
+      "loss": 2.25,
+      "step": 669
+    },
+    {
+      "epoch": 0.29777777777777775,
+      "grad_norm": 1.228171944618225,
+      "learning_rate": 0.00018825622775800712,
+      "loss": 2.4179,
+      "step": 670
+    },
+    {
+      "epoch": 0.2982222222222222,
+      "grad_norm": 1.2194924354553223,
+      "learning_rate": 0.00018823843416370108,
+      "loss": 2.5822,
+      "step": 671
+    },
+    {
+      "epoch": 0.2986666666666667,
+      "grad_norm": 1.2600946426391602,
+      "learning_rate": 0.00018822064056939504,
+      "loss": 2.3714,
+      "step": 672
+    },
+    {
+      "epoch": 0.2991111111111111,
+      "grad_norm": 1.2756378650665283,
+      "learning_rate": 0.000188202846975089,
+      "loss": 2.3046,
+      "step": 673
+    },
+    {
+      "epoch": 0.2995555555555556,
+      "grad_norm": 1.4584718942642212,
+      "learning_rate": 0.00018818505338078292,
+      "loss": 2.6382,
+      "step": 674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.393619179725647,
+      "learning_rate": 0.00018816725978647688,
+      "loss": 2.7167,
+      "step": 675
+    },
+    {
+      "epoch": 0.30044444444444446,
+      "grad_norm": 1.6865702867507935,
+      "learning_rate": 0.0001881494661921708,
+      "loss": 2.446,
+      "step": 676
+    },
+    {
+      "epoch": 0.3008888888888889,
+      "grad_norm": 1.3696800470352173,
+      "learning_rate": 0.00018813167259786477,
+      "loss": 2.5134,
+      "step": 677
+    },
+    {
+      "epoch": 0.30133333333333334,
+      "grad_norm": 1.3241018056869507,
+      "learning_rate": 0.00018811387900355872,
+      "loss": 2.7836,
+      "step": 678
+    },
+    {
+      "epoch": 0.30177777777777776,
+      "grad_norm": 1.3688435554504395,
+      "learning_rate": 0.00018809608540925268,
+      "loss": 2.8028,
+      "step": 679
+    },
+    {
+      "epoch": 0.3022222222222222,
+      "grad_norm": 2.019115447998047,
+      "learning_rate": 0.00018807829181494663,
+      "loss": 2.4958,
+      "step": 680
+    },
+    {
+      "epoch": 0.30266666666666664,
+      "grad_norm": 1.3393666744232178,
+      "learning_rate": 0.0001880604982206406,
+      "loss": 2.4144,
+      "step": 681
+    },
+    {
+      "epoch": 0.3031111111111111,
+      "grad_norm": 1.5808879137039185,
+      "learning_rate": 0.00018804270462633452,
+      "loss": 1.5098,
+      "step": 682
+    },
+    {
+      "epoch": 0.3035555555555556,
+      "grad_norm": 1.5631835460662842,
+      "learning_rate": 0.00018802491103202848,
+      "loss": 2.6174,
+      "step": 683
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 1.338965892791748,
+      "learning_rate": 0.00018800711743772243,
+      "loss": 2.933,
+      "step": 684
+    },
+    {
+      "epoch": 0.30444444444444446,
+      "grad_norm": 1.4270402193069458,
+      "learning_rate": 0.0001879893238434164,
+      "loss": 2.4163,
+      "step": 685
+    },
+    {
+      "epoch": 0.3048888888888889,
+      "grad_norm": 1.6511561870574951,
+      "learning_rate": 0.00018797153024911035,
+      "loss": 2.8065,
+      "step": 686
+    },
+    {
+      "epoch": 0.30533333333333335,
+      "grad_norm": 1.3582799434661865,
+      "learning_rate": 0.00018795373665480428,
+      "loss": 2.4156,
+      "step": 687
+    },
+    {
+      "epoch": 0.30577777777777776,
+      "grad_norm": 1.3298442363739014,
+      "learning_rate": 0.00018793594306049823,
+      "loss": 2.7763,
+      "step": 688
+    },
+    {
+      "epoch": 0.30622222222222223,
+      "grad_norm": 1.5233420133590698,
+      "learning_rate": 0.00018791814946619216,
+      "loss": 2.605,
+      "step": 689
+    },
+    {
+      "epoch": 0.30666666666666664,
+      "grad_norm": 1.3514484167099,
+      "learning_rate": 0.00018790035587188612,
+      "loss": 2.6729,
+      "step": 690
+    },
+    {
+      "epoch": 0.3071111111111111,
+      "grad_norm": 1.5793657302856445,
+      "learning_rate": 0.00018788256227758007,
+      "loss": 2.8392,
+      "step": 691
+    },
+    {
+      "epoch": 0.3075555555555556,
+      "grad_norm": 1.590437889099121,
+      "learning_rate": 0.00018786476868327403,
+      "loss": 2.445,
+      "step": 692
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 1.5807512998580933,
+      "learning_rate": 0.000187846975088968,
+      "loss": 2.9341,
+      "step": 693
+    },
+    {
+      "epoch": 0.30844444444444447,
+      "grad_norm": 1.5815593004226685,
+      "learning_rate": 0.00018782918149466194,
+      "loss": 2.5062,
+      "step": 694
+    },
+    {
+      "epoch": 0.3088888888888889,
+      "grad_norm": 1.6342856884002686,
+      "learning_rate": 0.00018781138790035587,
+      "loss": 2.769,
+      "step": 695
+    },
+    {
+      "epoch": 0.30933333333333335,
+      "grad_norm": 1.8067598342895508,
+      "learning_rate": 0.00018779359430604983,
+      "loss": 2.3512,
+      "step": 696
+    },
+    {
+      "epoch": 0.30977777777777776,
+      "grad_norm": 1.7684136629104614,
+      "learning_rate": 0.0001877758007117438,
+      "loss": 3.2031,
+      "step": 697
+    },
+    {
+      "epoch": 0.31022222222222223,
+      "grad_norm": 1.7312310934066772,
+      "learning_rate": 0.00018775800711743774,
+      "loss": 2.7118,
+      "step": 698
+    },
+    {
+      "epoch": 0.31066666666666665,
+      "grad_norm": 1.7264273166656494,
+      "learning_rate": 0.0001877402135231317,
+      "loss": 2.7106,
+      "step": 699
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 1.8946504592895508,
+      "learning_rate": 0.00018772241992882563,
+      "loss": 1.8234,
+      "step": 700
+    },
+    {
+      "epoch": 0.31155555555555553,
+      "grad_norm": 0.897571325302124,
+      "learning_rate": 0.00018770462633451959,
+      "loss": 2.4317,
+      "step": 701
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 1.212544560432434,
+      "learning_rate": 0.00018768683274021351,
+      "loss": 1.817,
+      "step": 702
+    },
+    {
+      "epoch": 0.31244444444444447,
+      "grad_norm": 1.4370547533035278,
+      "learning_rate": 0.00018766903914590747,
+      "loss": 1.6912,
+      "step": 703
+    },
+    {
+      "epoch": 0.3128888888888889,
+      "grad_norm": 1.2379952669143677,
+      "learning_rate": 0.00018765124555160143,
+      "loss": 2.8551,
+      "step": 704
+    },
+    {
+      "epoch": 0.31333333333333335,
+      "grad_norm": 1.2043800354003906,
+      "learning_rate": 0.00018763345195729538,
+      "loss": 2.2021,
+      "step": 705
+    },
+    {
+      "epoch": 0.31377777777777777,
+      "grad_norm": 1.1653484106063843,
+      "learning_rate": 0.00018761565836298934,
+      "loss": 2.97,
+      "step": 706
+    },
+    {
+      "epoch": 0.31422222222222224,
+      "grad_norm": 1.2504090070724487,
+      "learning_rate": 0.0001875978647686833,
+      "loss": 2.8417,
+      "step": 707
+    },
+    {
+      "epoch": 0.31466666666666665,
+      "grad_norm": 1.186420202255249,
+      "learning_rate": 0.00018758007117437723,
+      "loss": 2.2231,
+      "step": 708
+    },
+    {
+      "epoch": 0.3151111111111111,
+      "grad_norm": 1.3034793138504028,
+      "learning_rate": 0.00018756227758007118,
+      "loss": 2.7367,
+      "step": 709
+    },
+    {
+      "epoch": 0.31555555555555553,
+      "grad_norm": 1.1502488851547241,
+      "learning_rate": 0.00018754448398576514,
+      "loss": 3.0465,
+      "step": 710
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 1.3152366876602173,
+      "learning_rate": 0.0001875266903914591,
+      "loss": 3.0754,
+      "step": 711
+    },
+    {
+      "epoch": 0.3164444444444444,
+      "grad_norm": 1.0930118560791016,
+      "learning_rate": 0.00018750889679715305,
+      "loss": 2.3183,
+      "step": 712
+    },
+    {
+      "epoch": 0.3168888888888889,
+      "grad_norm": 1.3411848545074463,
+      "learning_rate": 0.00018749110320284698,
+      "loss": 3.0055,
+      "step": 713
+    },
+    {
+      "epoch": 0.31733333333333336,
+      "grad_norm": 1.222649097442627,
+      "learning_rate": 0.00018747330960854094,
+      "loss": 2.541,
+      "step": 714
+    },
+    {
+      "epoch": 0.31777777777777777,
+      "grad_norm": 1.3231635093688965,
+      "learning_rate": 0.00018745551601423487,
+      "loss": 3.1424,
+      "step": 715
+    },
+    {
+      "epoch": 0.31822222222222224,
+      "grad_norm": 1.3029333353042603,
+      "learning_rate": 0.00018743772241992882,
+      "loss": 2.736,
+      "step": 716
+    },
+    {
+      "epoch": 0.31866666666666665,
+      "grad_norm": 1.1651556491851807,
+      "learning_rate": 0.00018741992882562278,
+      "loss": 2.4525,
+      "step": 717
+    },
+    {
+      "epoch": 0.3191111111111111,
+      "grad_norm": 1.1928997039794922,
+      "learning_rate": 0.00018740213523131674,
+      "loss": 2.338,
+      "step": 718
+    },
+    {
+      "epoch": 0.31955555555555554,
+      "grad_norm": 1.2094029188156128,
+      "learning_rate": 0.0001873843416370107,
+      "loss": 2.1728,
+      "step": 719
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3674081563949585,
+      "learning_rate": 0.00018736654804270465,
+      "loss": 2.7086,
+      "step": 720
+    },
+    {
+      "epoch": 0.3204444444444444,
+      "grad_norm": 1.2240111827850342,
+      "learning_rate": 0.00018734875444839858,
+      "loss": 2.4084,
+      "step": 721
+    },
+    {
+      "epoch": 0.3208888888888889,
+      "grad_norm": 1.625939965248108,
+      "learning_rate": 0.00018733096085409254,
+      "loss": 2.233,
+      "step": 722
+    },
+    {
+      "epoch": 0.32133333333333336,
+      "grad_norm": 1.572806477546692,
+      "learning_rate": 0.0001873131672597865,
+      "loss": 2.8494,
+      "step": 723
+    },
+    {
+      "epoch": 0.3217777777777778,
+      "grad_norm": 1.4796736240386963,
+      "learning_rate": 0.00018729537366548045,
+      "loss": 2.5345,
+      "step": 724
+    },
+    {
+      "epoch": 0.32222222222222224,
+      "grad_norm": 1.5258103609085083,
+      "learning_rate": 0.0001872775800711744,
+      "loss": 2.4471,
+      "step": 725
+    },
+    {
+      "epoch": 0.32266666666666666,
+      "grad_norm": 1.5818583965301514,
+      "learning_rate": 0.00018725978647686834,
+      "loss": 2.9746,
+      "step": 726
+    },
+    {
+      "epoch": 0.3231111111111111,
+      "grad_norm": 1.275765061378479,
+      "learning_rate": 0.00018724199288256226,
+      "loss": 2.4232,
+      "step": 727
+    },
+    {
+      "epoch": 0.32355555555555554,
+      "grad_norm": 1.3722361326217651,
+      "learning_rate": 0.00018722419928825622,
+      "loss": 1.4788,
+      "step": 728
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 1.369632363319397,
+      "learning_rate": 0.00018720640569395018,
+      "loss": 2.5712,
+      "step": 729
+    },
+    {
+      "epoch": 0.3244444444444444,
+      "grad_norm": 1.5968562364578247,
+      "learning_rate": 0.00018718861209964413,
+      "loss": 2.3008,
+      "step": 730
+    },
+    {
+      "epoch": 0.3248888888888889,
+      "grad_norm": 1.6501327753067017,
+      "learning_rate": 0.0001871708185053381,
+      "loss": 2.5323,
+      "step": 731
+    },
+    {
+      "epoch": 0.3253333333333333,
+      "grad_norm": 1.5736616849899292,
+      "learning_rate": 0.00018715302491103205,
+      "loss": 2.2086,
+      "step": 732
+    },
+    {
+      "epoch": 0.3257777777777778,
+      "grad_norm": 1.4434736967086792,
+      "learning_rate": 0.00018713523131672598,
+      "loss": 2.613,
+      "step": 733
+    },
+    {
+      "epoch": 0.32622222222222225,
+      "grad_norm": 1.5532594919204712,
+      "learning_rate": 0.00018711743772241993,
+      "loss": 2.4163,
+      "step": 734
+    },
+    {
+      "epoch": 0.32666666666666666,
+      "grad_norm": 1.4101078510284424,
+      "learning_rate": 0.0001870996441281139,
+      "loss": 2.6866,
+      "step": 735
+    },
+    {
+      "epoch": 0.32711111111111113,
+      "grad_norm": 1.3974218368530273,
+      "learning_rate": 0.00018708185053380785,
+      "loss": 2.0701,
+      "step": 736
+    },
+    {
+      "epoch": 0.32755555555555554,
+      "grad_norm": 1.8705499172210693,
+      "learning_rate": 0.0001870640569395018,
+      "loss": 2.5967,
+      "step": 737
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 1.7035057544708252,
+      "learning_rate": 0.00018704626334519576,
+      "loss": 2.589,
+      "step": 738
+    },
+    {
+      "epoch": 0.32844444444444443,
+      "grad_norm": 1.477556824684143,
+      "learning_rate": 0.0001870284697508897,
+      "loss": 2.576,
+      "step": 739
+    },
+    {
+      "epoch": 0.3288888888888889,
+      "grad_norm": 1.5129868984222412,
+      "learning_rate": 0.00018701067615658362,
+      "loss": 2.5122,
+      "step": 740
+    },
+    {
+      "epoch": 0.3293333333333333,
+      "grad_norm": 1.453865647315979,
+      "learning_rate": 0.00018699288256227757,
+      "loss": 2.1861,
+      "step": 741
+    },
+    {
+      "epoch": 0.3297777777777778,
+      "grad_norm": 1.3834642171859741,
+      "learning_rate": 0.00018697508896797153,
+      "loss": 2.3745,
+      "step": 742
+    },
+    {
+      "epoch": 0.3302222222222222,
+      "grad_norm": 1.610922932624817,
+      "learning_rate": 0.0001869572953736655,
+      "loss": 2.6936,
+      "step": 743
+    },
+    {
+      "epoch": 0.33066666666666666,
+      "grad_norm": 1.550073266029358,
+      "learning_rate": 0.00018693950177935944,
+      "loss": 2.3292,
+      "step": 744
+    },
+    {
+      "epoch": 0.33111111111111113,
+      "grad_norm": 1.5677452087402344,
+      "learning_rate": 0.0001869217081850534,
+      "loss": 2.8312,
+      "step": 745
+    },
+    {
+      "epoch": 0.33155555555555555,
+      "grad_norm": 1.6848689317703247,
+      "learning_rate": 0.00018690391459074733,
+      "loss": 2.9414,
+      "step": 746
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 1.5544129610061646,
+      "learning_rate": 0.00018688612099644129,
+      "loss": 2.3318,
+      "step": 747
+    },
+    {
+      "epoch": 0.33244444444444443,
+      "grad_norm": 1.705752968788147,
+      "learning_rate": 0.00018686832740213524,
+      "loss": 2.4389,
+      "step": 748
+    },
+    {
+      "epoch": 0.3328888888888889,
+      "grad_norm": 1.7628095149993896,
+      "learning_rate": 0.0001868505338078292,
+      "loss": 2.7991,
+      "step": 749
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 2.3342695236206055,
+      "learning_rate": 0.00018683274021352316,
+      "loss": 2.8554,
+      "step": 750
+    },
+    {
+      "epoch": 0.3337777777777778,
+      "grad_norm": 0.9579228758811951,
+      "learning_rate": 0.0001868149466192171,
+      "loss": 3.0001,
+      "step": 751
+    },
+    {
+      "epoch": 0.3342222222222222,
+      "grad_norm": 0.910830020904541,
+      "learning_rate": 0.00018679715302491104,
+      "loss": 2.7247,
+      "step": 752
+    },
+    {
+      "epoch": 0.33466666666666667,
+      "grad_norm": 0.9618215560913086,
+      "learning_rate": 0.00018677935943060497,
+      "loss": 2.1627,
+      "step": 753
+    },
+    {
+      "epoch": 0.33511111111111114,
+      "grad_norm": 1.398019790649414,
+      "learning_rate": 0.00018676156583629893,
+      "loss": 2.8774,
+      "step": 754
+    },
+    {
+      "epoch": 0.33555555555555555,
+      "grad_norm": 0.98659348487854,
+      "learning_rate": 0.00018674377224199288,
+      "loss": 2.207,
+      "step": 755
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 1.0544646978378296,
+      "learning_rate": 0.00018672597864768684,
+      "loss": 2.5833,
+      "step": 756
+    },
+    {
+      "epoch": 0.33644444444444443,
+      "grad_norm": 1.1433809995651245,
+      "learning_rate": 0.0001867081850533808,
+      "loss": 2.1706,
+      "step": 757
+    },
+    {
+      "epoch": 0.3368888888888889,
+      "grad_norm": 1.170425295829773,
+      "learning_rate": 0.00018669039145907475,
+      "loss": 2.3311,
+      "step": 758
+    },
+    {
+      "epoch": 0.3373333333333333,
+      "grad_norm": 1.2254000902175903,
+      "learning_rate": 0.00018667259786476868,
+      "loss": 2.671,
+      "step": 759
+    },
+    {
+      "epoch": 0.3377777777777778,
+      "grad_norm": 1.2437927722930908,
+      "learning_rate": 0.00018665480427046264,
+      "loss": 2.9828,
+      "step": 760
+    },
+    {
+      "epoch": 0.3382222222222222,
+      "grad_norm": 1.3588943481445312,
+      "learning_rate": 0.0001866370106761566,
+      "loss": 2.8734,
+      "step": 761
+    },
+    {
+      "epoch": 0.33866666666666667,
+      "grad_norm": 1.5109245777130127,
+      "learning_rate": 0.00018661921708185055,
+      "loss": 2.0159,
+      "step": 762
+    },
+    {
+      "epoch": 0.3391111111111111,
+      "grad_norm": 1.2116190195083618,
+      "learning_rate": 0.0001866014234875445,
+      "loss": 2.1226,
+      "step": 763
+    },
+    {
+      "epoch": 0.33955555555555555,
+      "grad_norm": 1.3050971031188965,
+      "learning_rate": 0.00018658362989323847,
+      "loss": 2.2373,
+      "step": 764
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2877930402755737,
+      "learning_rate": 0.0001865658362989324,
+      "loss": 2.5477,
+      "step": 765
+    },
+    {
+      "epoch": 0.34044444444444444,
+      "grad_norm": 1.306307077407837,
+      "learning_rate": 0.00018654804270462632,
+      "loss": 2.7654,
+      "step": 766
+    },
+    {
+      "epoch": 0.3408888888888889,
+      "grad_norm": 1.3450534343719482,
+      "learning_rate": 0.00018653024911032028,
+      "loss": 1.9589,
+      "step": 767
+    },
+    {
+      "epoch": 0.3413333333333333,
+      "grad_norm": 1.574854850769043,
+      "learning_rate": 0.00018651245551601424,
+      "loss": 2.778,
+      "step": 768
+    },
+    {
+      "epoch": 0.3417777777777778,
+      "grad_norm": 1.3935576677322388,
+      "learning_rate": 0.0001864946619217082,
+      "loss": 2.8835,
+      "step": 769
+    },
+    {
+      "epoch": 0.3422222222222222,
+      "grad_norm": 1.169109582901001,
+      "learning_rate": 0.00018647686832740215,
+      "loss": 1.8586,
+      "step": 770
+    },
+    {
+      "epoch": 0.3426666666666667,
+      "grad_norm": 1.6845237016677856,
+      "learning_rate": 0.0001864590747330961,
+      "loss": 2.3178,
+      "step": 771
+    },
+    {
+      "epoch": 0.3431111111111111,
+      "grad_norm": 2.100719690322876,
+      "learning_rate": 0.00018644128113879004,
+      "loss": 2.5948,
+      "step": 772
+    },
+    {
+      "epoch": 0.34355555555555556,
+      "grad_norm": 1.3049522638320923,
+      "learning_rate": 0.000186423487544484,
+      "loss": 2.2732,
+      "step": 773
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 1.610150933265686,
+      "learning_rate": 0.00018640569395017795,
+      "loss": 2.658,
+      "step": 774
+    },
+    {
+      "epoch": 0.34444444444444444,
+      "grad_norm": 2.3013813495635986,
+      "learning_rate": 0.0001863879003558719,
+      "loss": 1.0201,
+      "step": 775
+    },
+    {
+      "epoch": 0.3448888888888889,
+      "grad_norm": 1.3802649974822998,
+      "learning_rate": 0.00018637010676156586,
+      "loss": 2.4351,
+      "step": 776
+    },
+    {
+      "epoch": 0.3453333333333333,
+      "grad_norm": 1.4456000328063965,
+      "learning_rate": 0.0001863523131672598,
+      "loss": 2.9545,
+      "step": 777
+    },
+    {
+      "epoch": 0.3457777777777778,
+      "grad_norm": 1.7522518634796143,
+      "learning_rate": 0.00018633451957295375,
+      "loss": 2.3526,
+      "step": 778
+    },
+    {
+      "epoch": 0.3462222222222222,
+      "grad_norm": 1.460486650466919,
+      "learning_rate": 0.00018631672597864768,
+      "loss": 2.9393,
+      "step": 779
+    },
+    {
+      "epoch": 0.3466666666666667,
+      "grad_norm": 1.650462031364441,
+      "learning_rate": 0.00018629893238434163,
+      "loss": 2.8534,
+      "step": 780
+    },
+    {
+      "epoch": 0.3471111111111111,
+      "grad_norm": 1.3861716985702515,
+      "learning_rate": 0.0001862811387900356,
+      "loss": 2.7778,
+      "step": 781
+    },
+    {
+      "epoch": 0.34755555555555556,
+      "grad_norm": 1.4128412008285522,
+      "learning_rate": 0.00018626334519572955,
+      "loss": 2.5953,
+      "step": 782
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 1.5334755182266235,
+      "learning_rate": 0.0001862455516014235,
+      "loss": 2.4878,
+      "step": 783
+    },
+    {
+      "epoch": 0.34844444444444445,
+      "grad_norm": 1.4212511777877808,
+      "learning_rate": 0.00018622775800711746,
+      "loss": 2.4014,
+      "step": 784
+    },
+    {
+      "epoch": 0.3488888888888889,
+      "grad_norm": 1.7563406229019165,
+      "learning_rate": 0.0001862099644128114,
+      "loss": 2.8054,
+      "step": 785
+    },
+    {
+      "epoch": 0.34933333333333333,
+      "grad_norm": 1.8490791320800781,
+      "learning_rate": 0.00018619217081850535,
+      "loss": 2.5487,
+      "step": 786
+    },
+    {
+      "epoch": 0.3497777777777778,
+      "grad_norm": 1.55039644241333,
+      "learning_rate": 0.0001861743772241993,
+      "loss": 2.1236,
+      "step": 787
+    },
+    {
+      "epoch": 0.3502222222222222,
+      "grad_norm": 1.187299132347107,
+      "learning_rate": 0.00018615658362989326,
+      "loss": 1.773,
+      "step": 788
+    },
+    {
+      "epoch": 0.3506666666666667,
+      "grad_norm": 1.3866082429885864,
+      "learning_rate": 0.00018613879003558722,
+      "loss": 2.0484,
+      "step": 789
+    },
+    {
+      "epoch": 0.3511111111111111,
+      "grad_norm": 1.5214849710464478,
+      "learning_rate": 0.00018612099644128114,
+      "loss": 2.5582,
+      "step": 790
+    },
+    {
+      "epoch": 0.35155555555555557,
+      "grad_norm": 1.9252493381500244,
+      "learning_rate": 0.0001861032028469751,
+      "loss": 2.6299,
+      "step": 791
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 2.005993366241455,
+      "learning_rate": 0.00018608540925266903,
+      "loss": 2.9747,
+      "step": 792
+    },
+    {
+      "epoch": 0.35244444444444445,
+      "grad_norm": 1.4414646625518799,
+      "learning_rate": 0.000186067615658363,
+      "loss": 2.3433,
+      "step": 793
+    },
+    {
+      "epoch": 0.35288888888888886,
+      "grad_norm": 1.6313060522079468,
+      "learning_rate": 0.00018604982206405694,
+      "loss": 2.4237,
+      "step": 794
+    },
+    {
+      "epoch": 0.35333333333333333,
+      "grad_norm": 1.917683720588684,
+      "learning_rate": 0.0001860320284697509,
+      "loss": 2.7755,
+      "step": 795
+    },
+    {
+      "epoch": 0.3537777777777778,
+      "grad_norm": 1.681670069694519,
+      "learning_rate": 0.00018601423487544486,
+      "loss": 2.5065,
+      "step": 796
+    },
+    {
+      "epoch": 0.3542222222222222,
+      "grad_norm": 1.6937282085418701,
+      "learning_rate": 0.0001859964412811388,
+      "loss": 2.1956,
+      "step": 797
+    },
+    {
+      "epoch": 0.3546666666666667,
+      "grad_norm": 1.8412083387374878,
+      "learning_rate": 0.00018597864768683274,
+      "loss": 2.2921,
+      "step": 798
+    },
+    {
+      "epoch": 0.3551111111111111,
+      "grad_norm": 2.1753652095794678,
+      "learning_rate": 0.0001859608540925267,
+      "loss": 2.0335,
+      "step": 799
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 3.113302707672119,
+      "learning_rate": 0.00018594306049822066,
+      "loss": 2.3211,
+      "step": 800
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 1.0349335670471191,
+      "learning_rate": 0.0001859252669039146,
+      "loss": 2.5146,
+      "step": 801
+    },
+    {
+      "epoch": 0.35644444444444445,
+      "grad_norm": 1.299102544784546,
+      "learning_rate": 0.00018590747330960857,
+      "loss": 2.537,
+      "step": 802
+    },
+    {
+      "epoch": 0.35688888888888887,
+      "grad_norm": 1.0697323083877563,
+      "learning_rate": 0.0001858896797153025,
+      "loss": 2.2026,
+      "step": 803
+    },
+    {
+      "epoch": 0.35733333333333334,
+      "grad_norm": 1.0610216856002808,
+      "learning_rate": 0.00018587188612099645,
+      "loss": 2.6074,
+      "step": 804
+    },
+    {
+      "epoch": 0.35777777777777775,
+      "grad_norm": 1.3162498474121094,
+      "learning_rate": 0.00018585409252669038,
+      "loss": 3.1557,
+      "step": 805
+    },
+    {
+      "epoch": 0.3582222222222222,
+      "grad_norm": 1.2941645383834839,
+      "learning_rate": 0.00018583629893238434,
+      "loss": 2.8547,
+      "step": 806
+    },
+    {
+      "epoch": 0.3586666666666667,
+      "grad_norm": 1.1461007595062256,
+      "learning_rate": 0.0001858185053380783,
+      "loss": 2.7406,
+      "step": 807
+    },
+    {
+      "epoch": 0.3591111111111111,
+      "grad_norm": 1.1096692085266113,
+      "learning_rate": 0.00018580071174377225,
+      "loss": 2.3455,
+      "step": 808
+    },
+    {
+      "epoch": 0.3595555555555556,
+      "grad_norm": 1.158469796180725,
+      "learning_rate": 0.0001857829181494662,
+      "loss": 2.9105,
+      "step": 809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1534368991851807,
+      "learning_rate": 0.00018576512455516017,
+      "loss": 2.6555,
+      "step": 810
+    },
+    {
+      "epoch": 0.36044444444444446,
+      "grad_norm": 1.1266659498214722,
+      "learning_rate": 0.0001857473309608541,
+      "loss": 2.7283,
+      "step": 811
+    },
+    {
+      "epoch": 0.36088888888888887,
+      "grad_norm": 1.1437948942184448,
+      "learning_rate": 0.00018572953736654805,
+      "loss": 2.4803,
+      "step": 812
+    },
+    {
+      "epoch": 0.36133333333333334,
+      "grad_norm": 1.182286262512207,
+      "learning_rate": 0.000185711743772242,
+      "loss": 2.6214,
+      "step": 813
+    },
+    {
+      "epoch": 0.36177777777777775,
+      "grad_norm": 1.253722071647644,
+      "learning_rate": 0.00018569395017793596,
+      "loss": 2.6758,
+      "step": 814
+    },
+    {
+      "epoch": 0.3622222222222222,
+      "grad_norm": 1.2334574460983276,
+      "learning_rate": 0.00018567615658362992,
+      "loss": 2.6005,
+      "step": 815
+    },
+    {
+      "epoch": 0.3626666666666667,
+      "grad_norm": 1.1198886632919312,
+      "learning_rate": 0.00018565836298932385,
+      "loss": 2.4263,
+      "step": 816
+    },
+    {
+      "epoch": 0.3631111111111111,
+      "grad_norm": 1.501847743988037,
+      "learning_rate": 0.00018564056939501778,
+      "loss": 3.3744,
+      "step": 817
+    },
+    {
+      "epoch": 0.3635555555555556,
+      "grad_norm": 1.3934186697006226,
+      "learning_rate": 0.00018562277580071174,
+      "loss": 3.2117,
+      "step": 818
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 1.3710157871246338,
+      "learning_rate": 0.0001856049822064057,
+      "loss": 1.9331,
+      "step": 819
+    },
+    {
+      "epoch": 0.36444444444444446,
+      "grad_norm": 1.4316257238388062,
+      "learning_rate": 0.00018558718861209965,
+      "loss": 1.9532,
+      "step": 820
+    },
+    {
+      "epoch": 0.3648888888888889,
+      "grad_norm": 1.8586760759353638,
+      "learning_rate": 0.0001855693950177936,
+      "loss": 2.8284,
+      "step": 821
+    },
+    {
+      "epoch": 0.36533333333333334,
+      "grad_norm": 1.5303040742874146,
+      "learning_rate": 0.00018555160142348756,
+      "loss": 2.0609,
+      "step": 822
+    },
+    {
+      "epoch": 0.36577777777777776,
+      "grad_norm": 1.5688817501068115,
+      "learning_rate": 0.0001855338078291815,
+      "loss": 2.5499,
+      "step": 823
+    },
+    {
+      "epoch": 0.3662222222222222,
+      "grad_norm": 1.3707939386367798,
+      "learning_rate": 0.00018551601423487545,
+      "loss": 2.4933,
+      "step": 824
+    },
+    {
+      "epoch": 0.36666666666666664,
+      "grad_norm": 1.435909628868103,
+      "learning_rate": 0.0001854982206405694,
+      "loss": 2.7909,
+      "step": 825
+    },
+    {
+      "epoch": 0.3671111111111111,
+      "grad_norm": 1.5045204162597656,
+      "learning_rate": 0.00018548042704626336,
+      "loss": 2.4644,
+      "step": 826
+    },
+    {
+      "epoch": 0.3675555555555556,
+      "grad_norm": 1.3811103105545044,
+      "learning_rate": 0.00018546263345195732,
+      "loss": 2.1934,
+      "step": 827
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 1.7939069271087646,
+      "learning_rate": 0.00018544483985765127,
+      "loss": 2.2179,
+      "step": 828
+    },
+    {
+      "epoch": 0.36844444444444446,
+      "grad_norm": 1.7151756286621094,
+      "learning_rate": 0.0001854270462633452,
+      "loss": 2.9784,
+      "step": 829
+    },
+    {
+      "epoch": 0.3688888888888889,
+      "grad_norm": 1.6932202577590942,
+      "learning_rate": 0.00018540925266903913,
+      "loss": 2.5832,
+      "step": 830
+    },
+    {
+      "epoch": 0.36933333333333335,
+      "grad_norm": 1.4874944686889648,
+      "learning_rate": 0.0001853914590747331,
+      "loss": 2.2795,
+      "step": 831
+    },
+    {
+      "epoch": 0.36977777777777776,
+      "grad_norm": 1.5106111764907837,
+      "learning_rate": 0.00018537366548042705,
+      "loss": 1.6537,
+      "step": 832
+    },
+    {
+      "epoch": 0.37022222222222223,
+      "grad_norm": 1.5347083806991577,
+      "learning_rate": 0.000185355871886121,
+      "loss": 2.5899,
+      "step": 833
+    },
+    {
+      "epoch": 0.37066666666666664,
+      "grad_norm": 1.2724993228912354,
+      "learning_rate": 0.00018533807829181496,
+      "loss": 2.4979,
+      "step": 834
+    },
+    {
+      "epoch": 0.3711111111111111,
+      "grad_norm": 1.5324300527572632,
+      "learning_rate": 0.00018532028469750892,
+      "loss": 2.4469,
+      "step": 835
+    },
+    {
+      "epoch": 0.37155555555555553,
+      "grad_norm": 1.6249970197677612,
+      "learning_rate": 0.00018530249110320285,
+      "loss": 2.6245,
+      "step": 836
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 1.3731900453567505,
+      "learning_rate": 0.0001852846975088968,
+      "loss": 1.9859,
+      "step": 837
+    },
+    {
+      "epoch": 0.37244444444444447,
+      "grad_norm": 1.437991976737976,
+      "learning_rate": 0.00018526690391459076,
+      "loss": 2.4142,
+      "step": 838
+    },
+    {
+      "epoch": 0.3728888888888889,
+      "grad_norm": 2.2342700958251953,
+      "learning_rate": 0.00018524911032028471,
+      "loss": 1.0456,
+      "step": 839
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "grad_norm": 1.4727734327316284,
+      "learning_rate": 0.00018523131672597867,
+      "loss": 2.3314,
+      "step": 840
+    },
+    {
+      "epoch": 0.37377777777777776,
+      "grad_norm": 1.6986255645751953,
+      "learning_rate": 0.00018521352313167263,
+      "loss": 2.5555,
+      "step": 841
+    },
+    {
+      "epoch": 0.37422222222222223,
+      "grad_norm": 1.611127495765686,
+      "learning_rate": 0.00018519572953736656,
+      "loss": 2.5777,
+      "step": 842
+    },
+    {
+      "epoch": 0.37466666666666665,
+      "grad_norm": 1.5206453800201416,
+      "learning_rate": 0.0001851779359430605,
+      "loss": 2.8207,
+      "step": 843
+    },
+    {
+      "epoch": 0.3751111111111111,
+      "grad_norm": 1.5014015436172485,
+      "learning_rate": 0.00018516014234875444,
+      "loss": 2.2796,
+      "step": 844
+    },
+    {
+      "epoch": 0.37555555555555553,
+      "grad_norm": 1.9145801067352295,
+      "learning_rate": 0.0001851423487544484,
+      "loss": 2.432,
+      "step": 845
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 1.7269951105117798,
+      "learning_rate": 0.00018512455516014236,
+      "loss": 2.8725,
+      "step": 846
+    },
+    {
+      "epoch": 0.37644444444444447,
+      "grad_norm": 2.094856023788452,
+      "learning_rate": 0.0001851067615658363,
+      "loss": 2.5228,
+      "step": 847
+    },
+    {
+      "epoch": 0.3768888888888889,
+      "grad_norm": 4.175536155700684,
+      "learning_rate": 0.00018508896797153027,
+      "loss": 1.4835,
+      "step": 848
+    },
+    {
+      "epoch": 0.37733333333333335,
+      "grad_norm": 2.4652979373931885,
+      "learning_rate": 0.0001850711743772242,
+      "loss": 2.4617,
+      "step": 849
+    },
+    {
+      "epoch": 0.37777777777777777,
+      "grad_norm": 2.9979593753814697,
+      "learning_rate": 0.00018505338078291815,
+      "loss": 2.2239,
+      "step": 850
+    },
+    {
+      "epoch": 0.37822222222222224,
+      "grad_norm": 2.925299644470215,
+      "learning_rate": 0.0001850355871886121,
+      "loss": 1.4753,
+      "step": 851
+    },
+    {
+      "epoch": 0.37866666666666665,
+      "grad_norm": 1.156390905380249,
+      "learning_rate": 0.00018501779359430607,
+      "loss": 2.5781,
+      "step": 852
+    },
+    {
+      "epoch": 0.3791111111111111,
+      "grad_norm": 1.1416361331939697,
+      "learning_rate": 0.00018500000000000002,
+      "loss": 2.5683,
+      "step": 853
+    },
+    {
+      "epoch": 0.37955555555555553,
+      "grad_norm": 1.1092318296432495,
+      "learning_rate": 0.00018498220640569398,
+      "loss": 2.6212,
+      "step": 854
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1286892890930176,
+      "learning_rate": 0.0001849644128113879,
+      "loss": 2.3329,
+      "step": 855
+    },
+    {
+      "epoch": 0.3804444444444444,
+      "grad_norm": 1.3141859769821167,
+      "learning_rate": 0.00018494661921708184,
+      "loss": 2.6131,
+      "step": 856
+    },
+    {
+      "epoch": 0.3808888888888889,
+      "grad_norm": 1.1940083503723145,
+      "learning_rate": 0.0001849288256227758,
+      "loss": 2.6077,
+      "step": 857
+    },
+    {
+      "epoch": 0.38133333333333336,
+      "grad_norm": 1.2545088529586792,
+      "learning_rate": 0.00018491103202846975,
+      "loss": 2.6905,
+      "step": 858
+    },
+    {
+      "epoch": 0.38177777777777777,
+      "grad_norm": 1.0562766790390015,
+      "learning_rate": 0.0001848932384341637,
+      "loss": 2.1665,
+      "step": 859
+    },
+    {
+      "epoch": 0.38222222222222224,
+      "grad_norm": 1.2652605772018433,
+      "learning_rate": 0.00018487544483985767,
+      "loss": 2.6562,
+      "step": 860
+    },
+    {
+      "epoch": 0.38266666666666665,
+      "grad_norm": 1.213104486465454,
+      "learning_rate": 0.00018485765124555162,
+      "loss": 2.3212,
+      "step": 861
+    },
+    {
+      "epoch": 0.3831111111111111,
+      "grad_norm": 1.2592909336090088,
+      "learning_rate": 0.00018483985765124555,
+      "loss": 2.2561,
+      "step": 862
+    },
+    {
+      "epoch": 0.38355555555555554,
+      "grad_norm": 1.3437938690185547,
+      "learning_rate": 0.0001848220640569395,
+      "loss": 2.7051,
+      "step": 863
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 1.2356623411178589,
+      "learning_rate": 0.00018480427046263346,
+      "loss": 2.1958,
+      "step": 864
+    },
+    {
+      "epoch": 0.3844444444444444,
+      "grad_norm": 1.6262998580932617,
+      "learning_rate": 0.00018478647686832742,
+      "loss": 3.2701,
+      "step": 865
+    },
+    {
+      "epoch": 0.3848888888888889,
+      "grad_norm": 1.336805820465088,
+      "learning_rate": 0.00018476868327402138,
+      "loss": 2.8542,
+      "step": 866
+    },
+    {
+      "epoch": 0.38533333333333336,
+      "grad_norm": 1.4716001749038696,
+      "learning_rate": 0.00018475088967971533,
+      "loss": 2.5937,
+      "step": 867
+    },
+    {
+      "epoch": 0.3857777777777778,
+      "grad_norm": 1.3492522239685059,
+      "learning_rate": 0.00018473309608540926,
+      "loss": 2.7743,
+      "step": 868
+    },
+    {
+      "epoch": 0.38622222222222224,
+      "grad_norm": 1.2297523021697998,
+      "learning_rate": 0.0001847153024911032,
+      "loss": 2.1033,
+      "step": 869
+    },
+    {
+      "epoch": 0.38666666666666666,
+      "grad_norm": 1.3531607389450073,
+      "learning_rate": 0.00018469750889679715,
+      "loss": 2.3351,
+      "step": 870
+    },
+    {
+      "epoch": 0.38711111111111113,
+      "grad_norm": 1.213259220123291,
+      "learning_rate": 0.0001846797153024911,
+      "loss": 2.422,
+      "step": 871
+    },
+    {
+      "epoch": 0.38755555555555554,
+      "grad_norm": 1.6566977500915527,
+      "learning_rate": 0.00018466192170818506,
+      "loss": 3.0473,
+      "step": 872
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 1.389674425125122,
+      "learning_rate": 0.00018464412811387902,
+      "loss": 2.6186,
+      "step": 873
+    },
+    {
+      "epoch": 0.3884444444444444,
+      "grad_norm": 1.4744458198547363,
+      "learning_rate": 0.00018462633451957298,
+      "loss": 2.5226,
+      "step": 874
+    },
+    {
+      "epoch": 0.3888888888888889,
+      "grad_norm": 1.5133804082870483,
+      "learning_rate": 0.0001846085409252669,
+      "loss": 1.6426,
+      "step": 875
+    },
+    {
+      "epoch": 0.3893333333333333,
+      "grad_norm": 1.3070919513702393,
+      "learning_rate": 0.00018459074733096086,
+      "loss": 2.5463,
+      "step": 876
+    },
+    {
+      "epoch": 0.3897777777777778,
+      "grad_norm": 1.5536634922027588,
+      "learning_rate": 0.00018457295373665482,
+      "loss": 2.5429,
+      "step": 877
+    },
+    {
+      "epoch": 0.39022222222222225,
+      "grad_norm": 1.3954426050186157,
+      "learning_rate": 0.00018455516014234877,
+      "loss": 2.1144,
+      "step": 878
+    },
+    {
+      "epoch": 0.39066666666666666,
+      "grad_norm": 1.403937816619873,
+      "learning_rate": 0.00018453736654804273,
+      "loss": 2.4386,
+      "step": 879
+    },
+    {
+      "epoch": 0.39111111111111113,
+      "grad_norm": 1.692806601524353,
+      "learning_rate": 0.0001845195729537367,
+      "loss": 2.7535,
+      "step": 880
+    },
+    {
+      "epoch": 0.39155555555555555,
+      "grad_norm": 2.0360846519470215,
+      "learning_rate": 0.00018450177935943062,
+      "loss": 2.3359,
+      "step": 881
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 1.7397620677947998,
+      "learning_rate": 0.00018448398576512455,
+      "loss": 2.8038,
+      "step": 882
+    },
+    {
+      "epoch": 0.39244444444444443,
+      "grad_norm": 1.9279683828353882,
+      "learning_rate": 0.0001844661921708185,
+      "loss": 2.6394,
+      "step": 883
+    },
+    {
+      "epoch": 0.3928888888888889,
+      "grad_norm": 1.6761490106582642,
+      "learning_rate": 0.00018444839857651246,
+      "loss": 2.6922,
+      "step": 884
+    },
+    {
+      "epoch": 0.3933333333333333,
+      "grad_norm": 1.7097992897033691,
+      "learning_rate": 0.00018443060498220642,
+      "loss": 2.8203,
+      "step": 885
+    },
+    {
+      "epoch": 0.3937777777777778,
+      "grad_norm": 1.796673059463501,
+      "learning_rate": 0.00018441281138790037,
+      "loss": 3.291,
+      "step": 886
+    },
+    {
+      "epoch": 0.3942222222222222,
+      "grad_norm": 1.646627426147461,
+      "learning_rate": 0.00018439501779359433,
+      "loss": 2.6629,
+      "step": 887
+    },
+    {
+      "epoch": 0.39466666666666667,
+      "grad_norm": 1.7203751802444458,
+      "learning_rate": 0.00018437722419928826,
+      "loss": 2.3987,
+      "step": 888
+    },
+    {
+      "epoch": 0.39511111111111114,
+      "grad_norm": 2.0489413738250732,
+      "learning_rate": 0.00018435943060498221,
+      "loss": 2.5578,
+      "step": 889
+    },
+    {
+      "epoch": 0.39555555555555555,
+      "grad_norm": 1.5938488245010376,
+      "learning_rate": 0.00018434163701067617,
+      "loss": 2.1986,
+      "step": 890
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 1.604232907295227,
+      "learning_rate": 0.00018432384341637013,
+      "loss": 2.0916,
+      "step": 891
+    },
+    {
+      "epoch": 0.39644444444444443,
+      "grad_norm": 1.5554643869400024,
+      "learning_rate": 0.00018430604982206408,
+      "loss": 2.5466,
+      "step": 892
+    },
+    {
+      "epoch": 0.3968888888888889,
+      "grad_norm": 1.606425166130066,
+      "learning_rate": 0.000184288256227758,
+      "loss": 2.6946,
+      "step": 893
+    },
+    {
+      "epoch": 0.3973333333333333,
+      "grad_norm": 1.6642916202545166,
+      "learning_rate": 0.00018427046263345197,
+      "loss": 2.1499,
+      "step": 894
+    },
+    {
+      "epoch": 0.3977777777777778,
+      "grad_norm": 1.8029860258102417,
+      "learning_rate": 0.0001842526690391459,
+      "loss": 2.6755,
+      "step": 895
+    },
+    {
+      "epoch": 0.3982222222222222,
+      "grad_norm": 2.077056407928467,
+      "learning_rate": 0.00018423487544483986,
+      "loss": 2.7935,
+      "step": 896
+    },
+    {
+      "epoch": 0.39866666666666667,
+      "grad_norm": 1.7928773164749146,
+      "learning_rate": 0.0001842170818505338,
+      "loss": 2.8071,
+      "step": 897
+    },
+    {
+      "epoch": 0.39911111111111114,
+      "grad_norm": 2.6374940872192383,
+      "learning_rate": 0.00018419928825622777,
+      "loss": 3.2532,
+      "step": 898
+    },
+    {
+      "epoch": 0.39955555555555555,
+      "grad_norm": 2.1440799236297607,
+      "learning_rate": 0.00018418149466192173,
+      "loss": 1.447,
+      "step": 899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.737084984779358,
+      "learning_rate": 0.00018416370106761568,
+      "loss": 2.3376,
+      "step": 900
+    },
+    {
+      "epoch": 0.40044444444444444,
+      "grad_norm": 1.0363576412200928,
+      "learning_rate": 0.0001841459074733096,
+      "loss": 2.5708,
+      "step": 901
+    },
+    {
+      "epoch": 0.4008888888888889,
+      "grad_norm": 0.8888387084007263,
+      "learning_rate": 0.00018412811387900357,
+      "loss": 2.8373,
+      "step": 902
+    },
+    {
+      "epoch": 0.4013333333333333,
+      "grad_norm": 0.8826941847801208,
+      "learning_rate": 0.00018411032028469752,
+      "loss": 2.6704,
+      "step": 903
+    },
+    {
+      "epoch": 0.4017777777777778,
+      "grad_norm": 0.9509096741676331,
+      "learning_rate": 0.00018409252669039148,
+      "loss": 2.6441,
+      "step": 904
+    },
+    {
+      "epoch": 0.4022222222222222,
+      "grad_norm": 1.036983847618103,
+      "learning_rate": 0.00018407473309608544,
+      "loss": 2.3402,
+      "step": 905
+    },
+    {
+      "epoch": 0.4026666666666667,
+      "grad_norm": 0.9840334057807922,
+      "learning_rate": 0.00018405693950177937,
+      "loss": 2.2159,
+      "step": 906
+    },
+    {
+      "epoch": 0.4031111111111111,
+      "grad_norm": 1.118037223815918,
+      "learning_rate": 0.0001840391459074733,
+      "loss": 2.4926,
+      "step": 907
+    },
+    {
+      "epoch": 0.40355555555555556,
+      "grad_norm": 1.1225322484970093,
+      "learning_rate": 0.00018402135231316725,
+      "loss": 2.6811,
+      "step": 908
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 1.295507788658142,
+      "learning_rate": 0.0001840035587188612,
+      "loss": 2.8815,
+      "step": 909
+    },
+    {
+      "epoch": 0.40444444444444444,
+      "grad_norm": 1.1872442960739136,
+      "learning_rate": 0.00018398576512455517,
+      "loss": 3.0637,
+      "step": 910
+    },
+    {
+      "epoch": 0.4048888888888889,
+      "grad_norm": 1.2855168581008911,
+      "learning_rate": 0.00018396797153024912,
+      "loss": 2.4883,
+      "step": 911
+    },
+    {
+      "epoch": 0.4053333333333333,
+      "grad_norm": 1.232706904411316,
+      "learning_rate": 0.00018395017793594308,
+      "loss": 2.6065,
+      "step": 912
+    },
+    {
+      "epoch": 0.4057777777777778,
+      "grad_norm": 1.326191782951355,
+      "learning_rate": 0.000183932384341637,
+      "loss": 2.0947,
+      "step": 913
+    },
+    {
+      "epoch": 0.4062222222222222,
+      "grad_norm": 1.2210899591445923,
+      "learning_rate": 0.00018391459074733096,
+      "loss": 2.8879,
+      "step": 914
+    },
+    {
+      "epoch": 0.4066666666666667,
+      "grad_norm": 1.358302354812622,
+      "learning_rate": 0.00018389679715302492,
+      "loss": 2.6797,
+      "step": 915
+    },
+    {
+      "epoch": 0.4071111111111111,
+      "grad_norm": 1.2646130323410034,
+      "learning_rate": 0.00018387900355871888,
+      "loss": 2.2242,
+      "step": 916
+    },
+    {
+      "epoch": 0.40755555555555556,
+      "grad_norm": 1.2646642923355103,
+      "learning_rate": 0.00018386120996441283,
+      "loss": 2.5151,
+      "step": 917
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 1.4230983257293701,
+      "learning_rate": 0.0001838434163701068,
+      "loss": 1.5127,
+      "step": 918
+    },
+    {
+      "epoch": 0.40844444444444444,
+      "grad_norm": 1.5309816598892212,
+      "learning_rate": 0.00018382562277580072,
+      "loss": 2.9285,
+      "step": 919
+    },
+    {
+      "epoch": 0.4088888888888889,
+      "grad_norm": 1.2716691493988037,
+      "learning_rate": 0.00018380782918149465,
+      "loss": 2.5965,
+      "step": 920
+    },
+    {
+      "epoch": 0.4093333333333333,
+      "grad_norm": 1.433159589767456,
+      "learning_rate": 0.0001837900355871886,
+      "loss": 2.5312,
+      "step": 921
+    },
+    {
+      "epoch": 0.4097777777777778,
+      "grad_norm": 1.3177905082702637,
+      "learning_rate": 0.00018377224199288256,
+      "loss": 2.5805,
+      "step": 922
+    },
+    {
+      "epoch": 0.4102222222222222,
+      "grad_norm": 1.6881523132324219,
+      "learning_rate": 0.00018375444839857652,
+      "loss": 2.5188,
+      "step": 923
+    },
+    {
+      "epoch": 0.4106666666666667,
+      "grad_norm": 1.5823473930358887,
+      "learning_rate": 0.00018373665480427047,
+      "loss": 2.6616,
+      "step": 924
+    },
+    {
+      "epoch": 0.4111111111111111,
+      "grad_norm": 1.2907118797302246,
+      "learning_rate": 0.00018371886120996443,
+      "loss": 2.7342,
+      "step": 925
+    },
+    {
+      "epoch": 0.41155555555555556,
+      "grad_norm": 1.569952368736267,
+      "learning_rate": 0.00018370106761565836,
+      "loss": 2.6745,
+      "step": 926
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 1.2594976425170898,
+      "learning_rate": 0.00018368327402135232,
+      "loss": 2.1675,
+      "step": 927
+    },
+    {
+      "epoch": 0.41244444444444445,
+      "grad_norm": 1.449838638305664,
+      "learning_rate": 0.00018366548042704627,
+      "loss": 2.7803,
+      "step": 928
+    },
+    {
+      "epoch": 0.4128888888888889,
+      "grad_norm": 1.5406020879745483,
+      "learning_rate": 0.00018364768683274023,
+      "loss": 3.0986,
+      "step": 929
+    },
+    {
+      "epoch": 0.41333333333333333,
+      "grad_norm": 1.357981562614441,
+      "learning_rate": 0.0001836298932384342,
+      "loss": 2.2642,
+      "step": 930
+    },
+    {
+      "epoch": 0.4137777777777778,
+      "grad_norm": 1.4212137460708618,
+      "learning_rate": 0.00018361209964412814,
+      "loss": 2.4661,
+      "step": 931
+    },
+    {
+      "epoch": 0.4142222222222222,
+      "grad_norm": 1.3381963968276978,
+      "learning_rate": 0.00018359430604982207,
+      "loss": 2.1015,
+      "step": 932
+    },
+    {
+      "epoch": 0.4146666666666667,
+      "grad_norm": 1.223344326019287,
+      "learning_rate": 0.000183576512455516,
+      "loss": 1.0327,
+      "step": 933
+    },
+    {
+      "epoch": 0.4151111111111111,
+      "grad_norm": 1.5084744691848755,
+      "learning_rate": 0.00018355871886120996,
+      "loss": 3.0321,
+      "step": 934
+    },
+    {
+      "epoch": 0.41555555555555557,
+      "grad_norm": 1.3545995950698853,
+      "learning_rate": 0.00018354092526690392,
+      "loss": 1.6716,
+      "step": 935
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 1.775868535041809,
+      "learning_rate": 0.00018352313167259787,
+      "loss": 1.6868,
+      "step": 936
+    },
+    {
+      "epoch": 0.41644444444444445,
+      "grad_norm": 1.401804804801941,
+      "learning_rate": 0.00018350533807829183,
+      "loss": 2.2054,
+      "step": 937
+    },
+    {
+      "epoch": 0.41688888888888886,
+      "grad_norm": 1.4973540306091309,
+      "learning_rate": 0.00018348754448398578,
+      "loss": 2.396,
+      "step": 938
+    },
+    {
+      "epoch": 0.41733333333333333,
+      "grad_norm": 1.5535446405410767,
+      "learning_rate": 0.00018346975088967971,
+      "loss": 2.465,
+      "step": 939
+    },
+    {
+      "epoch": 0.4177777777777778,
+      "grad_norm": 1.7133632898330688,
+      "learning_rate": 0.00018345195729537367,
+      "loss": 2.8616,
+      "step": 940
+    },
+    {
+      "epoch": 0.4182222222222222,
+      "grad_norm": 1.5524804592132568,
+      "learning_rate": 0.00018343416370106763,
+      "loss": 2.1337,
+      "step": 941
+    },
+    {
+      "epoch": 0.4186666666666667,
+      "grad_norm": 1.4653924703598022,
+      "learning_rate": 0.00018341637010676158,
+      "loss": 2.3331,
+      "step": 942
+    },
+    {
+      "epoch": 0.4191111111111111,
+      "grad_norm": 1.5715734958648682,
+      "learning_rate": 0.00018339857651245554,
+      "loss": 2.1571,
+      "step": 943
+    },
+    {
+      "epoch": 0.41955555555555557,
+      "grad_norm": 1.7893381118774414,
+      "learning_rate": 0.0001833807829181495,
+      "loss": 2.7177,
+      "step": 944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0075345039367676,
+      "learning_rate": 0.00018336298932384343,
+      "loss": 2.3181,
+      "step": 945
+    },
+    {
+      "epoch": 0.42044444444444445,
+      "grad_norm": 2.4787044525146484,
+      "learning_rate": 0.00018334519572953736,
+      "loss": 2.8261,
+      "step": 946
+    },
+    {
+      "epoch": 0.42088888888888887,
+      "grad_norm": 1.778351068496704,
+      "learning_rate": 0.0001833274021352313,
+      "loss": 2.7114,
+      "step": 947
+    },
+    {
+      "epoch": 0.42133333333333334,
+      "grad_norm": 2.3666553497314453,
+      "learning_rate": 0.00018330960854092527,
+      "loss": 2.6281,
+      "step": 948
+    },
+    {
+      "epoch": 0.42177777777777775,
+      "grad_norm": 2.386976718902588,
+      "learning_rate": 0.00018329181494661922,
+      "loss": 3.5857,
+      "step": 949
+    },
+    {
+      "epoch": 0.4222222222222222,
+      "grad_norm": 1.791477918624878,
+      "learning_rate": 0.00018327402135231318,
+      "loss": 2.5117,
+      "step": 950
+    },
+    {
+      "epoch": 0.4226666666666667,
+      "grad_norm": 1.2231998443603516,
+      "learning_rate": 0.00018325622775800714,
+      "loss": 2.0187,
+      "step": 951
+    },
+    {
+      "epoch": 0.4231111111111111,
+      "grad_norm": 1.0212533473968506,
+      "learning_rate": 0.00018323843416370107,
+      "loss": 2.8211,
+      "step": 952
+    },
+    {
+      "epoch": 0.4235555555555556,
+      "grad_norm": 1.1677687168121338,
+      "learning_rate": 0.00018322064056939502,
+      "loss": 2.5803,
+      "step": 953
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 1.0932122468948364,
+      "learning_rate": 0.00018320284697508898,
+      "loss": 2.5059,
+      "step": 954
+    },
+    {
+      "epoch": 0.42444444444444446,
+      "grad_norm": 1.079312801361084,
+      "learning_rate": 0.00018318505338078294,
+      "loss": 2.4724,
+      "step": 955
+    },
+    {
+      "epoch": 0.42488888888888887,
+      "grad_norm": 1.324885368347168,
+      "learning_rate": 0.0001831672597864769,
+      "loss": 2.8136,
+      "step": 956
+    },
+    {
+      "epoch": 0.42533333333333334,
+      "grad_norm": 1.3072516918182373,
+      "learning_rate": 0.00018314946619217085,
+      "loss": 2.9349,
+      "step": 957
+    },
+    {
+      "epoch": 0.42577777777777776,
+      "grad_norm": 1.3189862966537476,
+      "learning_rate": 0.00018313167259786478,
+      "loss": 2.5912,
+      "step": 958
+    },
+    {
+      "epoch": 0.4262222222222222,
+      "grad_norm": 1.2304902076721191,
+      "learning_rate": 0.0001831138790035587,
+      "loss": 2.5022,
+      "step": 959
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "grad_norm": 1.3272204399108887,
+      "learning_rate": 0.00018309608540925266,
+      "loss": 3.0605,
+      "step": 960
+    },
+    {
+      "epoch": 0.4271111111111111,
+      "grad_norm": 1.2381232976913452,
+      "learning_rate": 0.00018307829181494662,
+      "loss": 2.0443,
+      "step": 961
+    },
+    {
+      "epoch": 0.4275555555555556,
+      "grad_norm": 1.268486499786377,
+      "learning_rate": 0.00018306049822064058,
+      "loss": 2.2911,
+      "step": 962
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 1.2793582677841187,
+      "learning_rate": 0.00018304270462633453,
+      "loss": 2.8313,
+      "step": 963
+    },
+    {
+      "epoch": 0.42844444444444446,
+      "grad_norm": 1.3932663202285767,
+      "learning_rate": 0.0001830249110320285,
+      "loss": 2.8657,
+      "step": 964
+    },
+    {
+      "epoch": 0.4288888888888889,
+      "grad_norm": 1.2110832929611206,
+      "learning_rate": 0.00018300711743772242,
+      "loss": 2.5214,
+      "step": 965
+    },
+    {
+      "epoch": 0.42933333333333334,
+      "grad_norm": 1.253836989402771,
+      "learning_rate": 0.00018298932384341638,
+      "loss": 2.5128,
+      "step": 966
+    },
+    {
+      "epoch": 0.42977777777777776,
+      "grad_norm": 1.33391273021698,
+      "learning_rate": 0.00018297153024911033,
+      "loss": 2.2437,
+      "step": 967
+    },
+    {
+      "epoch": 0.43022222222222223,
+      "grad_norm": 1.2081773281097412,
+      "learning_rate": 0.0001829537366548043,
+      "loss": 2.5682,
+      "step": 968
+    },
+    {
+      "epoch": 0.43066666666666664,
+      "grad_norm": 1.3490543365478516,
+      "learning_rate": 0.00018293594306049825,
+      "loss": 2.6874,
+      "step": 969
+    },
+    {
+      "epoch": 0.4311111111111111,
+      "grad_norm": 1.4848097562789917,
+      "learning_rate": 0.0001829181494661922,
+      "loss": 2.2042,
+      "step": 970
+    },
+    {
+      "epoch": 0.4315555555555556,
+      "grad_norm": 1.2465113401412964,
+      "learning_rate": 0.00018290035587188613,
+      "loss": 2.0922,
+      "step": 971
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 1.3515832424163818,
+      "learning_rate": 0.00018288256227758006,
+      "loss": 2.5685,
+      "step": 972
+    },
+    {
+      "epoch": 0.43244444444444446,
+      "grad_norm": 1.4319607019424438,
+      "learning_rate": 0.00018286476868327402,
+      "loss": 2.0091,
+      "step": 973
+    },
+    {
+      "epoch": 0.4328888888888889,
+      "grad_norm": 1.680587887763977,
+      "learning_rate": 0.00018284697508896797,
+      "loss": 2.6249,
+      "step": 974
+    },
+    {
+      "epoch": 0.43333333333333335,
+      "grad_norm": 1.3431737422943115,
+      "learning_rate": 0.00018282918149466193,
+      "loss": 2.5173,
+      "step": 975
+    },
+    {
+      "epoch": 0.43377777777777776,
+      "grad_norm": 1.2416516542434692,
+      "learning_rate": 0.0001828113879003559,
+      "loss": 2.644,
+      "step": 976
+    },
+    {
+      "epoch": 0.43422222222222223,
+      "grad_norm": 1.2653284072875977,
+      "learning_rate": 0.00018279359430604984,
+      "loss": 2.4739,
+      "step": 977
+    },
+    {
+      "epoch": 0.43466666666666665,
+      "grad_norm": 1.5131683349609375,
+      "learning_rate": 0.00018277580071174377,
+      "loss": 2.8553,
+      "step": 978
+    },
+    {
+      "epoch": 0.4351111111111111,
+      "grad_norm": 1.4946473836898804,
+      "learning_rate": 0.00018275800711743773,
+      "loss": 2.6073,
+      "step": 979
+    },
+    {
+      "epoch": 0.43555555555555553,
+      "grad_norm": 1.3179363012313843,
+      "learning_rate": 0.0001827402135231317,
+      "loss": 2.7802,
+      "step": 980
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 1.5546060800552368,
+      "learning_rate": 0.00018272241992882564,
+      "loss": 2.9142,
+      "step": 981
+    },
+    {
+      "epoch": 0.43644444444444447,
+      "grad_norm": 1.3515474796295166,
+      "learning_rate": 0.0001827046263345196,
+      "loss": 2.4532,
+      "step": 982
+    },
+    {
+      "epoch": 0.4368888888888889,
+      "grad_norm": 1.3547914028167725,
+      "learning_rate": 0.00018268683274021356,
+      "loss": 2.483,
+      "step": 983
+    },
+    {
+      "epoch": 0.43733333333333335,
+      "grad_norm": 1.2733529806137085,
+      "learning_rate": 0.00018266903914590749,
+      "loss": 2.0126,
+      "step": 984
+    },
+    {
+      "epoch": 0.43777777777777777,
+      "grad_norm": 1.6882998943328857,
+      "learning_rate": 0.00018265124555160141,
+      "loss": 3.432,
+      "step": 985
+    },
+    {
+      "epoch": 0.43822222222222224,
+      "grad_norm": 1.3751314878463745,
+      "learning_rate": 0.00018263345195729537,
+      "loss": 2.4604,
+      "step": 986
+    },
+    {
+      "epoch": 0.43866666666666665,
+      "grad_norm": 1.412575364112854,
+      "learning_rate": 0.00018261565836298933,
+      "loss": 2.4426,
+      "step": 987
+    },
+    {
+      "epoch": 0.4391111111111111,
+      "grad_norm": 1.7669273614883423,
+      "learning_rate": 0.00018259786476868328,
+      "loss": 2.5918,
+      "step": 988
+    },
+    {
+      "epoch": 0.43955555555555553,
+      "grad_norm": 1.605697751045227,
+      "learning_rate": 0.00018258007117437724,
+      "loss": 2.7675,
+      "step": 989
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.567189335823059,
+      "learning_rate": 0.0001825622775800712,
+      "loss": 2.9831,
+      "step": 990
+    },
+    {
+      "epoch": 0.44044444444444447,
+      "grad_norm": 1.3639848232269287,
+      "learning_rate": 0.00018254448398576513,
+      "loss": 1.9593,
+      "step": 991
+    },
+    {
+      "epoch": 0.4408888888888889,
+      "grad_norm": 1.586616039276123,
+      "learning_rate": 0.00018252669039145908,
+      "loss": 2.6094,
+      "step": 992
+    },
+    {
+      "epoch": 0.44133333333333336,
+      "grad_norm": 1.5296803712844849,
+      "learning_rate": 0.00018250889679715304,
+      "loss": 2.219,
+      "step": 993
+    },
+    {
+      "epoch": 0.44177777777777777,
+      "grad_norm": 1.6999601125717163,
+      "learning_rate": 0.000182491103202847,
+      "loss": 2.669,
+      "step": 994
+    },
+    {
+      "epoch": 0.44222222222222224,
+      "grad_norm": 2.0300962924957275,
+      "learning_rate": 0.00018247330960854095,
+      "loss": 2.6897,
+      "step": 995
+    },
+    {
+      "epoch": 0.44266666666666665,
+      "grad_norm": 1.7834362983703613,
+      "learning_rate": 0.00018245551601423488,
+      "loss": 2.3681,
+      "step": 996
+    },
+    {
+      "epoch": 0.4431111111111111,
+      "grad_norm": 1.7336286306381226,
+      "learning_rate": 0.0001824377224199288,
+      "loss": 2.5579,
+      "step": 997
+    },
+    {
+      "epoch": 0.44355555555555554,
+      "grad_norm": 1.8825653791427612,
+      "learning_rate": 0.00018241992882562277,
+      "loss": 2.8049,
+      "step": 998
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 1.9167088270187378,
+      "learning_rate": 0.00018240213523131672,
+      "loss": 2.9734,
+      "step": 999
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 2.2173099517822266,
+      "learning_rate": 0.00018238434163701068,
+      "loss": 2.6596,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4448888888888889,
+      "grad_norm": 0.9079247117042542,
+      "learning_rate": 0.00018236654804270464,
+      "loss": 2.6596,
+      "step": 1001
+    },
+    {
+      "epoch": 0.44533333333333336,
+      "grad_norm": 0.9305357336997986,
+      "learning_rate": 0.0001823487544483986,
+      "loss": 2.6803,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4457777777777778,
+      "grad_norm": 1.031278133392334,
+      "learning_rate": 0.00018233096085409252,
+      "loss": 2.5942,
+      "step": 1003
+    },
+    {
+      "epoch": 0.44622222222222224,
+      "grad_norm": 1.5068715810775757,
+      "learning_rate": 0.00018231316725978648,
+      "loss": 1.6077,
+      "step": 1004
+    },
+    {
+      "epoch": 0.44666666666666666,
+      "grad_norm": 1.17013680934906,
+      "learning_rate": 0.00018229537366548044,
+      "loss": 2.4984,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4471111111111111,
+      "grad_norm": 1.2330650091171265,
+      "learning_rate": 0.0001822775800711744,
+      "loss": 2.9625,
+      "step": 1006
+    },
+    {
+      "epoch": 0.44755555555555554,
+      "grad_norm": 1.1845786571502686,
+      "learning_rate": 0.00018225978647686835,
+      "loss": 2.8457,
+      "step": 1007
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 1.144061803817749,
+      "learning_rate": 0.0001822419928825623,
+      "loss": 2.0844,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4484444444444444,
+      "grad_norm": 1.383981466293335,
+      "learning_rate": 0.00018222419928825624,
+      "loss": 2.4893,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4488888888888889,
+      "grad_norm": 1.125385046005249,
+      "learning_rate": 0.00018220640569395016,
+      "loss": 2.2932,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4493333333333333,
+      "grad_norm": 1.2119320631027222,
+      "learning_rate": 0.00018218861209964412,
+      "loss": 2.4534,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4497777777777778,
+      "grad_norm": 1.108948826789856,
+      "learning_rate": 0.00018217081850533808,
+      "loss": 2.3211,
+      "step": 1012
+    },
+    {
+      "epoch": 0.45022222222222225,
+      "grad_norm": 1.3019822835922241,
+      "learning_rate": 0.00018215302491103203,
+      "loss": 2.3836,
+      "step": 1013
+    },
+    {
+      "epoch": 0.45066666666666666,
+      "grad_norm": 1.3383592367172241,
+      "learning_rate": 0.000182135231316726,
+      "loss": 2.7794,
+      "step": 1014
+    },
+    {
+      "epoch": 0.45111111111111113,
+      "grad_norm": 1.2700541019439697,
+      "learning_rate": 0.00018211743772241995,
+      "loss": 2.488,
+      "step": 1015
+    },
+    {
+      "epoch": 0.45155555555555554,
+      "grad_norm": 1.280701756477356,
+      "learning_rate": 0.00018209964412811388,
+      "loss": 2.5005,
+      "step": 1016
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 1.1454371213912964,
+      "learning_rate": 0.00018208185053380783,
+      "loss": 2.3253,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4524444444444444,
+      "grad_norm": 1.242236614227295,
+      "learning_rate": 0.0001820640569395018,
+      "loss": 2.4167,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4528888888888889,
+      "grad_norm": 1.2990704774856567,
+      "learning_rate": 0.00018204626334519575,
+      "loss": 2.1872,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4533333333333333,
+      "grad_norm": 1.283494472503662,
+      "learning_rate": 0.0001820284697508897,
+      "loss": 2.3171,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4537777777777778,
+      "grad_norm": 2.063596487045288,
+      "learning_rate": 0.00018201067615658366,
+      "loss": 1.681,
+      "step": 1021
+    },
+    {
+      "epoch": 0.45422222222222225,
+      "grad_norm": 1.3359391689300537,
+      "learning_rate": 0.0001819928825622776,
+      "loss": 2.0327,
+      "step": 1022
+    },
+    {
+      "epoch": 0.45466666666666666,
+      "grad_norm": 1.263917326927185,
+      "learning_rate": 0.00018197508896797152,
+      "loss": 2.239,
+      "step": 1023
+    },
+    {
+      "epoch": 0.45511111111111113,
+      "grad_norm": 1.8083239793777466,
+      "learning_rate": 0.00018195729537366547,
+      "loss": 2.7766,
+      "step": 1024
+    },
+    {
+      "epoch": 0.45555555555555555,
+      "grad_norm": 1.2244737148284912,
+      "learning_rate": 0.00018193950177935943,
+      "loss": 2.0234,
+      "step": 1025
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 1.4717793464660645,
+      "learning_rate": 0.0001819217081850534,
+      "loss": 2.655,
+      "step": 1026
+    },
+    {
+      "epoch": 0.45644444444444443,
+      "grad_norm": 1.5211389064788818,
+      "learning_rate": 0.00018190391459074734,
+      "loss": 2.335,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4568888888888889,
+      "grad_norm": 1.3322489261627197,
+      "learning_rate": 0.0001818861209964413,
+      "loss": 2.4195,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4573333333333333,
+      "grad_norm": 1.3629570007324219,
+      "learning_rate": 0.00018186832740213523,
+      "loss": 2.0856,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4577777777777778,
+      "grad_norm": 1.7404910326004028,
+      "learning_rate": 0.00018185053380782919,
+      "loss": 2.4039,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4582222222222222,
+      "grad_norm": 1.3935850858688354,
+      "learning_rate": 0.00018183274021352314,
+      "loss": 2.2005,
+      "step": 1031
+    },
+    {
+      "epoch": 0.45866666666666667,
+      "grad_norm": 1.3948017358779907,
+      "learning_rate": 0.0001818149466192171,
+      "loss": 2.7373,
+      "step": 1032
+    },
+    {
+      "epoch": 0.45911111111111114,
+      "grad_norm": 1.4649925231933594,
+      "learning_rate": 0.00018179715302491106,
+      "loss": 2.4669,
+      "step": 1033
+    },
+    {
+      "epoch": 0.45955555555555555,
+      "grad_norm": 1.5847502946853638,
+      "learning_rate": 0.000181779359430605,
+      "loss": 1.9865,
+      "step": 1034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.006155490875244,
+      "learning_rate": 0.00018176156583629894,
+      "loss": 0.4302,
+      "step": 1035
+    },
+    {
+      "epoch": 0.46044444444444443,
+      "grad_norm": 1.4123693704605103,
+      "learning_rate": 0.00018174377224199287,
+      "loss": 2.6673,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4608888888888889,
+      "grad_norm": 1.4881420135498047,
+      "learning_rate": 0.00018172597864768683,
+      "loss": 2.4905,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4613333333333333,
+      "grad_norm": 1.5692585706710815,
+      "learning_rate": 0.00018170818505338078,
+      "loss": 2.5459,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4617777777777778,
+      "grad_norm": 1.4341565370559692,
+      "learning_rate": 0.00018169039145907474,
+      "loss": 2.3733,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4622222222222222,
+      "grad_norm": 1.3901523351669312,
+      "learning_rate": 0.0001816725978647687,
+      "loss": 2.3915,
+      "step": 1040
+    },
+    {
+      "epoch": 0.46266666666666667,
+      "grad_norm": 1.5324454307556152,
+      "learning_rate": 0.00018165480427046265,
+      "loss": 2.3619,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4631111111111111,
+      "grad_norm": 1.567613959312439,
+      "learning_rate": 0.00018163701067615658,
+      "loss": 1.7382,
+      "step": 1042
+    },
+    {
+      "epoch": 0.46355555555555555,
+      "grad_norm": 1.8476368188858032,
+      "learning_rate": 0.00018161921708185054,
+      "loss": 2.8568,
+      "step": 1043
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 1.7850401401519775,
+      "learning_rate": 0.0001816014234875445,
+      "loss": 2.773,
+      "step": 1044
+    },
+    {
+      "epoch": 0.46444444444444444,
+      "grad_norm": 1.4900022745132446,
+      "learning_rate": 0.00018158362989323845,
+      "loss": 2.2433,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4648888888888889,
+      "grad_norm": 1.9751386642456055,
+      "learning_rate": 0.0001815658362989324,
+      "loss": 2.9478,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4653333333333333,
+      "grad_norm": 2.110889196395874,
+      "learning_rate": 0.00018154804270462637,
+      "loss": 3.1765,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4657777777777778,
+      "grad_norm": 1.8263683319091797,
+      "learning_rate": 0.0001815302491103203,
+      "loss": 2.7491,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4662222222222222,
+      "grad_norm": 1.875029444694519,
+      "learning_rate": 0.00018151245551601422,
+      "loss": 2.3277,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4666666666666667,
+      "grad_norm": 2.229357957839966,
+      "learning_rate": 0.00018149466192170818,
+      "loss": 2.8278,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4671111111111111,
+      "grad_norm": 1.0930174589157104,
+      "learning_rate": 0.00018147686832740214,
+      "loss": 2.5997,
+      "step": 1051
+    },
+    {
+      "epoch": 0.46755555555555556,
+      "grad_norm": 1.0300097465515137,
+      "learning_rate": 0.0001814590747330961,
+      "loss": 2.647,
+      "step": 1052
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 1.4961392879486084,
+      "learning_rate": 0.00018144128113879005,
+      "loss": 1.1492,
+      "step": 1053
+    },
+    {
+      "epoch": 0.46844444444444444,
+      "grad_norm": 1.1863456964492798,
+      "learning_rate": 0.000181423487544484,
+      "loss": 2.4534,
+      "step": 1054
+    },
+    {
+      "epoch": 0.4688888888888889,
+      "grad_norm": 1.156611442565918,
+      "learning_rate": 0.00018140569395017794,
+      "loss": 2.627,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4693333333333333,
+      "grad_norm": 1.1836708784103394,
+      "learning_rate": 0.0001813879003558719,
+      "loss": 2.3464,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4697777777777778,
+      "grad_norm": 1.2201026678085327,
+      "learning_rate": 0.00018137010676156585,
+      "loss": 2.8861,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4702222222222222,
+      "grad_norm": 2.332244396209717,
+      "learning_rate": 0.0001813523131672598,
+      "loss": 1.634,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4706666666666667,
+      "grad_norm": 1.337121605873108,
+      "learning_rate": 0.00018133451957295376,
+      "loss": 1.8932,
+      "step": 1059
+    },
+    {
+      "epoch": 0.4711111111111111,
+      "grad_norm": 1.2118984460830688,
+      "learning_rate": 0.00018131672597864772,
+      "loss": 2.2048,
+      "step": 1060
+    },
+    {
+      "epoch": 0.47155555555555556,
+      "grad_norm": 1.2091714143753052,
+      "learning_rate": 0.00018129893238434165,
+      "loss": 2.3127,
+      "step": 1061
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 1.2021245956420898,
+      "learning_rate": 0.00018128113879003558,
+      "loss": 2.6434,
+      "step": 1062
+    },
+    {
+      "epoch": 0.47244444444444444,
+      "grad_norm": 1.436784267425537,
+      "learning_rate": 0.00018126334519572953,
+      "loss": 2.3693,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4728888888888889,
+      "grad_norm": 2.271524429321289,
+      "learning_rate": 0.0001812455516014235,
+      "loss": 2.2956,
+      "step": 1064
+    },
+    {
+      "epoch": 0.47333333333333333,
+      "grad_norm": 1.162920355796814,
+      "learning_rate": 0.00018122775800711745,
+      "loss": 2.1083,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4737777777777778,
+      "grad_norm": 1.3496499061584473,
+      "learning_rate": 0.0001812099644128114,
+      "loss": 2.7196,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4742222222222222,
+      "grad_norm": 1.6458051204681396,
+      "learning_rate": 0.00018119217081850536,
+      "loss": 2.3444,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4746666666666667,
+      "grad_norm": 1.4568052291870117,
+      "learning_rate": 0.0001811743772241993,
+      "loss": 2.3294,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4751111111111111,
+      "grad_norm": 1.4898393154144287,
+      "learning_rate": 0.00018115658362989325,
+      "loss": 2.5576,
+      "step": 1069
+    },
+    {
+      "epoch": 0.47555555555555556,
+      "grad_norm": 1.3853182792663574,
+      "learning_rate": 0.0001811387900355872,
+      "loss": 2.4463,
+      "step": 1070
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 1.521707534790039,
+      "learning_rate": 0.00018112099644128116,
+      "loss": 2.7504,
+      "step": 1071
+    },
+    {
+      "epoch": 0.47644444444444445,
+      "grad_norm": 1.8744828701019287,
+      "learning_rate": 0.00018110320284697512,
+      "loss": 2.4272,
+      "step": 1072
+    },
+    {
+      "epoch": 0.47688888888888886,
+      "grad_norm": 1.4953957796096802,
+      "learning_rate": 0.00018108540925266907,
+      "loss": 2.2033,
+      "step": 1073
+    },
+    {
+      "epoch": 0.47733333333333333,
+      "grad_norm": 1.463110327720642,
+      "learning_rate": 0.000181067615658363,
+      "loss": 2.3522,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4777777777777778,
+      "grad_norm": 1.3929156064987183,
+      "learning_rate": 0.00018104982206405693,
+      "loss": 2.5893,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4782222222222222,
+      "grad_norm": 1.3469513654708862,
+      "learning_rate": 0.0001810320284697509,
+      "loss": 2.1896,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4786666666666667,
+      "grad_norm": 1.3536866903305054,
+      "learning_rate": 0.00018101423487544484,
+      "loss": 2.8212,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4791111111111111,
+      "grad_norm": 1.3798056840896606,
+      "learning_rate": 0.0001809964412811388,
+      "loss": 2.7518,
+      "step": 1078
+    },
+    {
+      "epoch": 0.47955555555555557,
+      "grad_norm": 1.553146243095398,
+      "learning_rate": 0.00018097864768683276,
+      "loss": 2.5848,
+      "step": 1079
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.532638669013977,
+      "learning_rate": 0.0001809608540925267,
+      "loss": 2.3551,
+      "step": 1080
+    },
+    {
+      "epoch": 0.48044444444444445,
+      "grad_norm": 1.4585469961166382,
+      "learning_rate": 0.00018094306049822064,
+      "loss": 2.6695,
+      "step": 1081
+    },
+    {
+      "epoch": 0.48088888888888887,
+      "grad_norm": 1.6650795936584473,
+      "learning_rate": 0.0001809252669039146,
+      "loss": 2.7979,
+      "step": 1082
+    },
+    {
+      "epoch": 0.48133333333333334,
+      "grad_norm": 1.3776137828826904,
+      "learning_rate": 0.00018090747330960856,
+      "loss": 2.1564,
+      "step": 1083
+    },
+    {
+      "epoch": 0.4817777777777778,
+      "grad_norm": 1.5997897386550903,
+      "learning_rate": 0.0001808896797153025,
+      "loss": 2.4603,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4822222222222222,
+      "grad_norm": 1.5599387884140015,
+      "learning_rate": 0.00018087188612099647,
+      "loss": 1.7888,
+      "step": 1085
+    },
+    {
+      "epoch": 0.4826666666666667,
+      "grad_norm": 1.5563712120056152,
+      "learning_rate": 0.0001808540925266904,
+      "loss": 2.438,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4831111111111111,
+      "grad_norm": 1.3926085233688354,
+      "learning_rate": 0.00018083629893238433,
+      "loss": 2.5696,
+      "step": 1087
+    },
+    {
+      "epoch": 0.48355555555555557,
+      "grad_norm": 1.7727118730545044,
+      "learning_rate": 0.00018081850533807828,
+      "loss": 2.2867,
+      "step": 1088
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 1.5934937000274658,
+      "learning_rate": 0.00018080071174377224,
+      "loss": 2.6703,
+      "step": 1089
+    },
+    {
+      "epoch": 0.48444444444444446,
+      "grad_norm": 1.789147973060608,
+      "learning_rate": 0.0001807829181494662,
+      "loss": 2.5851,
+      "step": 1090
+    },
+    {
+      "epoch": 0.48488888888888887,
+      "grad_norm": 1.4976032972335815,
+      "learning_rate": 0.00018076512455516015,
+      "loss": 2.1773,
+      "step": 1091
+    },
+    {
+      "epoch": 0.48533333333333334,
+      "grad_norm": 1.5114315748214722,
+      "learning_rate": 0.0001807473309608541,
+      "loss": 2.1908,
+      "step": 1092
+    },
+    {
+      "epoch": 0.48577777777777775,
+      "grad_norm": 1.5656942129135132,
+      "learning_rate": 0.00018072953736654804,
+      "loss": 2.5476,
+      "step": 1093
+    },
+    {
+      "epoch": 0.4862222222222222,
+      "grad_norm": 1.7908459901809692,
+      "learning_rate": 0.000180711743772242,
+      "loss": 2.6867,
+      "step": 1094
+    },
+    {
+      "epoch": 0.4866666666666667,
+      "grad_norm": 2.176135301589966,
+      "learning_rate": 0.00018069395017793595,
+      "loss": 2.0774,
+      "step": 1095
+    },
+    {
+      "epoch": 0.4871111111111111,
+      "grad_norm": 1.5971789360046387,
+      "learning_rate": 0.0001806761565836299,
+      "loss": 1.9446,
+      "step": 1096
+    },
+    {
+      "epoch": 0.4875555555555556,
+      "grad_norm": 1.709897518157959,
+      "learning_rate": 0.00018065836298932386,
+      "loss": 2.4818,
+      "step": 1097
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 1.9650827646255493,
+      "learning_rate": 0.00018064056939501782,
+      "loss": 2.5649,
+      "step": 1098
+    },
+    {
+      "epoch": 0.48844444444444446,
+      "grad_norm": 1.6556960344314575,
+      "learning_rate": 0.00018062277580071175,
+      "loss": 2.4576,
+      "step": 1099
+    },
+    {
+      "epoch": 0.4888888888888889,
+      "grad_norm": 2.5766797065734863,
+      "learning_rate": 0.00018060498220640568,
+      "loss": 2.1069,
+      "step": 1100
+    },
+    {
+      "epoch": 0.48933333333333334,
+      "grad_norm": 1.075654149055481,
+      "learning_rate": 0.00018058718861209964,
+      "loss": 2.5237,
+      "step": 1101
+    },
+    {
+      "epoch": 0.48977777777777776,
+      "grad_norm": 1.0496094226837158,
+      "learning_rate": 0.0001805693950177936,
+      "loss": 2.4905,
+      "step": 1102
+    },
+    {
+      "epoch": 0.4902222222222222,
+      "grad_norm": 0.9558612704277039,
+      "learning_rate": 0.00018055160142348755,
+      "loss": 2.9379,
+      "step": 1103
+    },
+    {
+      "epoch": 0.49066666666666664,
+      "grad_norm": 1.1676995754241943,
+      "learning_rate": 0.0001805338078291815,
+      "loss": 2.7487,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4911111111111111,
+      "grad_norm": 1.0585354566574097,
+      "learning_rate": 0.00018051601423487546,
+      "loss": 2.6956,
+      "step": 1105
+    },
+    {
+      "epoch": 0.4915555555555556,
+      "grad_norm": 1.1087442636489868,
+      "learning_rate": 0.0001804982206405694,
+      "loss": 2.6808,
+      "step": 1106
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 1.053286075592041,
+      "learning_rate": 0.00018048042704626335,
+      "loss": 2.2225,
+      "step": 1107
+    },
+    {
+      "epoch": 0.49244444444444446,
+      "grad_norm": 1.1056700944900513,
+      "learning_rate": 0.0001804626334519573,
+      "loss": 2.5699,
+      "step": 1108
+    },
+    {
+      "epoch": 0.4928888888888889,
+      "grad_norm": 1.198739767074585,
+      "learning_rate": 0.00018044483985765126,
+      "loss": 2.552,
+      "step": 1109
+    },
+    {
+      "epoch": 0.49333333333333335,
+      "grad_norm": 2.429079294204712,
+      "learning_rate": 0.00018042704626334522,
+      "loss": 1.7893,
+      "step": 1110
+    },
+    {
+      "epoch": 0.49377777777777776,
+      "grad_norm": 1.0832264423370361,
+      "learning_rate": 0.00018040925266903917,
+      "loss": 2.4136,
+      "step": 1111
+    },
+    {
+      "epoch": 0.49422222222222223,
+      "grad_norm": 1.0978045463562012,
+      "learning_rate": 0.0001803914590747331,
+      "loss": 2.1968,
+      "step": 1112
+    },
+    {
+      "epoch": 0.49466666666666664,
+      "grad_norm": 1.118681788444519,
+      "learning_rate": 0.00018037366548042703,
+      "loss": 2.3322,
+      "step": 1113
+    },
+    {
+      "epoch": 0.4951111111111111,
+      "grad_norm": 1.1858903169631958,
+      "learning_rate": 0.000180355871886121,
+      "loss": 2.5168,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4955555555555556,
+      "grad_norm": 1.3438916206359863,
+      "learning_rate": 0.00018033807829181495,
+      "loss": 2.7303,
+      "step": 1115
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 1.301822543144226,
+      "learning_rate": 0.0001803202846975089,
+      "loss": 2.824,
+      "step": 1116
+    },
+    {
+      "epoch": 0.49644444444444447,
+      "grad_norm": 1.2330950498580933,
+      "learning_rate": 0.00018030249110320286,
+      "loss": 3.0469,
+      "step": 1117
+    },
+    {
+      "epoch": 0.4968888888888889,
+      "grad_norm": 1.5200353860855103,
+      "learning_rate": 0.00018028469750889682,
+      "loss": 1.5884,
+      "step": 1118
+    },
+    {
+      "epoch": 0.49733333333333335,
+      "grad_norm": 1.2964918613433838,
+      "learning_rate": 0.00018026690391459075,
+      "loss": 2.4635,
+      "step": 1119
+    },
+    {
+      "epoch": 0.49777777777777776,
+      "grad_norm": 1.351252555847168,
+      "learning_rate": 0.0001802491103202847,
+      "loss": 2.1818,
+      "step": 1120
+    },
+    {
+      "epoch": 0.49822222222222223,
+      "grad_norm": 1.2766691446304321,
+      "learning_rate": 0.00018023131672597866,
+      "loss": 2.387,
+      "step": 1121
+    },
+    {
+      "epoch": 0.49866666666666665,
+      "grad_norm": 1.4819822311401367,
+      "learning_rate": 0.00018021352313167261,
+      "loss": 2.0414,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4991111111111111,
+      "grad_norm": 1.5996578931808472,
+      "learning_rate": 0.00018019572953736657,
+      "loss": 2.6798,
+      "step": 1123
+    },
+    {
+      "epoch": 0.49955555555555553,
+      "grad_norm": 1.4682111740112305,
+      "learning_rate": 0.00018017793594306053,
+      "loss": 2.6812,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.392949104309082,
+      "learning_rate": 0.00018016014234875446,
+      "loss": 2.4617,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5004444444444445,
+      "grad_norm": 1.4642528295516968,
+      "learning_rate": 0.0001801423487544484,
+      "loss": 2.6291,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5008888888888889,
+      "grad_norm": 1.2145447731018066,
+      "learning_rate": 0.00018012455516014234,
+      "loss": 2.0178,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5013333333333333,
+      "grad_norm": 1.6017488241195679,
+      "learning_rate": 0.0001801067615658363,
+      "loss": 2.5999,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5017777777777778,
+      "grad_norm": 1.3489327430725098,
+      "learning_rate": 0.00018008896797153026,
+      "loss": 1.565,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5022222222222222,
+      "grad_norm": 1.815772533416748,
+      "learning_rate": 0.0001800711743772242,
+      "loss": 2.4781,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5026666666666667,
+      "grad_norm": 1.6084818840026855,
+      "learning_rate": 0.00018005338078291817,
+      "loss": 2.2115,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5031111111111111,
+      "grad_norm": 1.483842372894287,
+      "learning_rate": 0.0001800355871886121,
+      "loss": 2.4087,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5035555555555555,
+      "grad_norm": 1.555029273033142,
+      "learning_rate": 0.00018001779359430605,
+      "loss": 2.6442,
+      "step": 1133
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 1.6016467809677124,
+      "learning_rate": 0.00018,
+      "loss": 2.3034,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5044444444444445,
+      "grad_norm": 1.4886064529418945,
+      "learning_rate": 0.00017998220640569397,
+      "loss": 2.3137,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5048888888888889,
+      "grad_norm": 1.590067744255066,
+      "learning_rate": 0.00017996441281138792,
+      "loss": 2.507,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5053333333333333,
+      "grad_norm": 1.2926700115203857,
+      "learning_rate": 0.00017994661921708188,
+      "loss": 1.8661,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5057777777777778,
+      "grad_norm": 1.8830050230026245,
+      "learning_rate": 0.0001799288256227758,
+      "loss": 2.7748,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5062222222222222,
+      "grad_norm": 1.6669584512710571,
+      "learning_rate": 0.00017991103202846974,
+      "loss": 2.7596,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5066666666666667,
+      "grad_norm": 1.7653512954711914,
+      "learning_rate": 0.0001798932384341637,
+      "loss": 3.2697,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5071111111111111,
+      "grad_norm": 1.8505072593688965,
+      "learning_rate": 0.00017987544483985765,
+      "loss": 2.9115,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5075555555555555,
+      "grad_norm": 1.5989995002746582,
+      "learning_rate": 0.0001798576512455516,
+      "loss": 2.0211,
+      "step": 1142
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 1.929032802581787,
+      "learning_rate": 0.00017983985765124557,
+      "loss": 2.6159,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5084444444444445,
+      "grad_norm": 1.9541597366333008,
+      "learning_rate": 0.00017982206405693952,
+      "loss": 2.6225,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5088888888888888,
+      "grad_norm": 2.0774333477020264,
+      "learning_rate": 0.00017980427046263345,
+      "loss": 2.6007,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5093333333333333,
+      "grad_norm": 1.6789966821670532,
+      "learning_rate": 0.0001797864768683274,
+      "loss": 2.505,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5097777777777778,
+      "grad_norm": 1.837697148323059,
+      "learning_rate": 0.00017976868327402136,
+      "loss": 3.0319,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5102222222222222,
+      "grad_norm": 1.7084914445877075,
+      "learning_rate": 0.00017975088967971532,
+      "loss": 2.4174,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5106666666666667,
+      "grad_norm": 2.1682441234588623,
+      "learning_rate": 0.00017973309608540928,
+      "loss": 1.1208,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5111111111111111,
+      "grad_norm": 1.9390794038772583,
+      "learning_rate": 0.00017971530249110323,
+      "loss": 2.4768,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5115555555555555,
+      "grad_norm": 0.9777438044548035,
+      "learning_rate": 0.00017969750889679716,
+      "loss": 3.1091,
+      "step": 1151
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.9038203358650208,
+      "learning_rate": 0.0001796797153024911,
+      "loss": 2.5902,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5124444444444445,
+      "grad_norm": 1.2871443033218384,
+      "learning_rate": 0.00017966192170818505,
+      "loss": 1.4081,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5128888888888888,
+      "grad_norm": 1.109168529510498,
+      "learning_rate": 0.000179644128113879,
+      "loss": 2.5515,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5133333333333333,
+      "grad_norm": 1.2267260551452637,
+      "learning_rate": 0.00017962633451957296,
+      "loss": 2.9597,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5137777777777778,
+      "grad_norm": 1.304792881011963,
+      "learning_rate": 0.00017960854092526692,
+      "loss": 2.4844,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5142222222222222,
+      "grad_norm": 1.1886632442474365,
+      "learning_rate": 0.00017959074733096088,
+      "loss": 2.5076,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5146666666666667,
+      "grad_norm": 1.3380016088485718,
+      "learning_rate": 0.0001795729537366548,
+      "loss": 2.7098,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5151111111111111,
+      "grad_norm": 2.1414008140563965,
+      "learning_rate": 0.00017955516014234876,
+      "loss": 1.6224,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5155555555555555,
+      "grad_norm": 1.3274937868118286,
+      "learning_rate": 0.00017953736654804272,
+      "loss": 2.9532,
+      "step": 1160
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 1.297349214553833,
+      "learning_rate": 0.00017951957295373667,
+      "loss": 2.4212,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5164444444444445,
+      "grad_norm": 1.2180557250976562,
+      "learning_rate": 0.00017950177935943063,
+      "loss": 2.6664,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5168888888888888,
+      "grad_norm": 1.4957740306854248,
+      "learning_rate": 0.0001794839857651246,
+      "loss": 2.5514,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5173333333333333,
+      "grad_norm": 1.2369511127471924,
+      "learning_rate": 0.00017946619217081852,
+      "loss": 2.3431,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5177777777777778,
+      "grad_norm": 1.2156001329421997,
+      "learning_rate": 0.00017944839857651245,
+      "loss": 2.5481,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5182222222222223,
+      "grad_norm": 1.2086858749389648,
+      "learning_rate": 0.0001794306049822064,
+      "loss": 2.0433,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5186666666666667,
+      "grad_norm": 1.3301823139190674,
+      "learning_rate": 0.00017941281138790036,
+      "loss": 2.3485,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5191111111111111,
+      "grad_norm": 1.3325672149658203,
+      "learning_rate": 0.00017939501779359432,
+      "loss": 2.6565,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5195555555555555,
+      "grad_norm": 1.1277254819869995,
+      "learning_rate": 0.00017937722419928827,
+      "loss": 2.1358,
+      "step": 1169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3675233125686646,
+      "learning_rate": 0.0001793594306049822,
+      "loss": 2.7481,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5204444444444445,
+      "grad_norm": 1.3147132396697998,
+      "learning_rate": 0.00017934163701067616,
+      "loss": 2.4171,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5208888888888888,
+      "grad_norm": 1.2295223474502563,
+      "learning_rate": 0.00017932384341637011,
+      "loss": 2.4101,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5213333333333333,
+      "grad_norm": 1.762349009513855,
+      "learning_rate": 0.00017930604982206407,
+      "loss": 3.3649,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5217777777777778,
+      "grad_norm": 1.4821921586990356,
+      "learning_rate": 0.00017928825622775803,
+      "loss": 2.7119,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5222222222222223,
+      "grad_norm": 1.4601001739501953,
+      "learning_rate": 0.00017927046263345198,
+      "loss": 2.1603,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5226666666666666,
+      "grad_norm": 1.397454857826233,
+      "learning_rate": 0.0001792526690391459,
+      "loss": 2.8709,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5231111111111111,
+      "grad_norm": 1.5645291805267334,
+      "learning_rate": 0.00017923487544483984,
+      "loss": 2.264,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5235555555555556,
+      "grad_norm": 1.3776110410690308,
+      "learning_rate": 0.0001792170818505338,
+      "loss": 2.5389,
+      "step": 1178
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 1.2663601636886597,
+      "learning_rate": 0.00017919928825622776,
+      "loss": 2.2745,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5244444444444445,
+      "grad_norm": 1.4239956140518188,
+      "learning_rate": 0.0001791814946619217,
+      "loss": 2.7701,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5248888888888888,
+      "grad_norm": 1.3543609380722046,
+      "learning_rate": 0.00017916370106761567,
+      "loss": 2.1195,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5253333333333333,
+      "grad_norm": 1.3027421236038208,
+      "learning_rate": 0.00017914590747330963,
+      "loss": 2.1416,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5257777777777778,
+      "grad_norm": 1.191349744796753,
+      "learning_rate": 0.00017912811387900355,
+      "loss": 1.8744,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5262222222222223,
+      "grad_norm": 2.0700368881225586,
+      "learning_rate": 0.0001791103202846975,
+      "loss": 1.3923,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5266666666666666,
+      "grad_norm": 1.5801739692687988,
+      "learning_rate": 0.00017909252669039147,
+      "loss": 2.7669,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5271111111111111,
+      "grad_norm": 1.549028992652893,
+      "learning_rate": 0.00017907473309608542,
+      "loss": 2.7438,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5275555555555556,
+      "grad_norm": 1.8961384296417236,
+      "learning_rate": 0.00017905693950177938,
+      "loss": 2.4417,
+      "step": 1187
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 1.741623044013977,
+      "learning_rate": 0.00017903914590747334,
+      "loss": 2.5803,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5284444444444445,
+      "grad_norm": 1.3786072731018066,
+      "learning_rate": 0.00017902135231316727,
+      "loss": 2.0496,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5288888888888889,
+      "grad_norm": 1.7181576490402222,
+      "learning_rate": 0.0001790035587188612,
+      "loss": 2.5333,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5293333333333333,
+      "grad_norm": 1.6818015575408936,
+      "learning_rate": 0.00017898576512455515,
+      "loss": 1.9528,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5297777777777778,
+      "grad_norm": 1.7420971393585205,
+      "learning_rate": 0.0001789679715302491,
+      "loss": 1.8997,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5302222222222223,
+      "grad_norm": 1.5828181505203247,
+      "learning_rate": 0.00017895017793594307,
+      "loss": 2.5736,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5306666666666666,
+      "grad_norm": 1.89609956741333,
+      "learning_rate": 0.00017893238434163702,
+      "loss": 1.9595,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5311111111111111,
+      "grad_norm": 1.4787846803665161,
+      "learning_rate": 0.00017891459074733098,
+      "loss": 2.3573,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5315555555555556,
+      "grad_norm": 1.931437373161316,
+      "learning_rate": 0.0001788967971530249,
+      "loss": 3.1484,
+      "step": 1196
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 1.7234727144241333,
+      "learning_rate": 0.00017887900355871886,
+      "loss": 2.5473,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5324444444444445,
+      "grad_norm": 2.0949268341064453,
+      "learning_rate": 0.00017886120996441282,
+      "loss": 2.9174,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5328888888888889,
+      "grad_norm": 1.990982174873352,
+      "learning_rate": 0.00017884341637010678,
+      "loss": 2.9417,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 2.504199981689453,
+      "learning_rate": 0.00017882562277580073,
+      "loss": 2.3845,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5337777777777778,
+      "grad_norm": 1.2785906791687012,
+      "learning_rate": 0.0001788078291814947,
+      "loss": 1.7756,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5342222222222223,
+      "grad_norm": 1.2264808416366577,
+      "learning_rate": 0.00017879003558718862,
+      "loss": 2.578,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5346666666666666,
+      "grad_norm": 1.7778968811035156,
+      "learning_rate": 0.00017877224199288255,
+      "loss": 0.1643,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5351111111111111,
+      "grad_norm": 0.8537470102310181,
+      "learning_rate": 0.0001787544483985765,
+      "loss": 1.3101,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5355555555555556,
+      "grad_norm": 1.3113861083984375,
+      "learning_rate": 0.00017873665480427046,
+      "loss": 3.1657,
+      "step": 1205
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 1.2095390558242798,
+      "learning_rate": 0.00017871886120996442,
+      "loss": 2.4251,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5364444444444444,
+      "grad_norm": 1.25002121925354,
+      "learning_rate": 0.00017870106761565837,
+      "loss": 2.5725,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5368888888888889,
+      "grad_norm": 1.3035950660705566,
+      "learning_rate": 0.00017868327402135233,
+      "loss": 2.8178,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5373333333333333,
+      "grad_norm": 1.295156717300415,
+      "learning_rate": 0.00017866548042704626,
+      "loss": 2.3121,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5377777777777778,
+      "grad_norm": 1.2773317098617554,
+      "learning_rate": 0.00017864768683274022,
+      "loss": 2.637,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5382222222222223,
+      "grad_norm": 1.2895100116729736,
+      "learning_rate": 0.00017862989323843417,
+      "loss": 2.5377,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5386666666666666,
+      "grad_norm": 1.4635089635849,
+      "learning_rate": 0.00017861209964412813,
+      "loss": 2.134,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5391111111111111,
+      "grad_norm": 1.3764005899429321,
+      "learning_rate": 0.0001785943060498221,
+      "loss": 2.4902,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5395555555555556,
+      "grad_norm": 1.3150113821029663,
+      "learning_rate": 0.00017857651245551604,
+      "loss": 2.8578,
+      "step": 1214
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5147175788879395,
+      "learning_rate": 0.00017855871886120997,
+      "loss": 1.2944,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5404444444444444,
+      "grad_norm": 1.1644126176834106,
+      "learning_rate": 0.0001785409252669039,
+      "loss": 2.5019,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5408888888888889,
+      "grad_norm": 1.375962734222412,
+      "learning_rate": 0.00017852313167259786,
+      "loss": 2.5303,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5413333333333333,
+      "grad_norm": 1.2268140316009521,
+      "learning_rate": 0.00017850533807829182,
+      "loss": 2.1568,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5417777777777778,
+      "grad_norm": 1.3434302806854248,
+      "learning_rate": 0.00017848754448398577,
+      "loss": 2.1272,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5422222222222223,
+      "grad_norm": 1.3689292669296265,
+      "learning_rate": 0.00017846975088967973,
+      "loss": 2.4022,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5426666666666666,
+      "grad_norm": 1.6363227367401123,
+      "learning_rate": 0.00017845195729537368,
+      "loss": 2.8829,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5431111111111111,
+      "grad_norm": 1.4127588272094727,
+      "learning_rate": 0.00017843416370106761,
+      "loss": 2.699,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5435555555555556,
+      "grad_norm": 1.462015986442566,
+      "learning_rate": 0.00017841637010676157,
+      "loss": 2.8654,
+      "step": 1223
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 1.2841753959655762,
+      "learning_rate": 0.00017839857651245553,
+      "loss": 2.2034,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5444444444444444,
+      "grad_norm": 1.377759575843811,
+      "learning_rate": 0.00017838078291814948,
+      "loss": 2.5642,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5448888888888889,
+      "grad_norm": 1.3649755716323853,
+      "learning_rate": 0.00017836298932384344,
+      "loss": 2.7037,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5453333333333333,
+      "grad_norm": 1.399375319480896,
+      "learning_rate": 0.0001783451957295374,
+      "loss": 2.766,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5457777777777778,
+      "grad_norm": 1.3284432888031006,
+      "learning_rate": 0.00017832740213523133,
+      "loss": 2.6953,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5462222222222223,
+      "grad_norm": 1.5032292604446411,
+      "learning_rate": 0.00017830960854092526,
+      "loss": 2.4118,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5466666666666666,
+      "grad_norm": 1.4157973527908325,
+      "learning_rate": 0.0001782918149466192,
+      "loss": 2.6983,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5471111111111111,
+      "grad_norm": 1.2969857454299927,
+      "learning_rate": 0.00017827402135231317,
+      "loss": 2.1972,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5475555555555556,
+      "grad_norm": 1.4679317474365234,
+      "learning_rate": 0.00017825622775800712,
+      "loss": 2.6372,
+      "step": 1232
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 1.451851487159729,
+      "learning_rate": 0.00017823843416370108,
+      "loss": 2.5865,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5484444444444444,
+      "grad_norm": 1.4248473644256592,
+      "learning_rate": 0.00017822064056939504,
+      "loss": 2.3113,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5488888888888889,
+      "grad_norm": 1.5742985010147095,
+      "learning_rate": 0.00017820284697508897,
+      "loss": 2.6367,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5493333333333333,
+      "grad_norm": 1.569652795791626,
+      "learning_rate": 0.00017818505338078292,
+      "loss": 1.9242,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5497777777777778,
+      "grad_norm": 1.8335659503936768,
+      "learning_rate": 0.00017816725978647688,
+      "loss": 2.2898,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5502222222222222,
+      "grad_norm": 1.419884204864502,
+      "learning_rate": 0.00017814946619217084,
+      "loss": 2.3061,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5506666666666666,
+      "grad_norm": 1.5601950883865356,
+      "learning_rate": 0.0001781316725978648,
+      "loss": 2.5379,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5511111111111111,
+      "grad_norm": 1.4974377155303955,
+      "learning_rate": 0.00017811387900355875,
+      "loss": 2.5647,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5515555555555556,
+      "grad_norm": 1.5708105564117432,
+      "learning_rate": 0.00017809608540925268,
+      "loss": 2.8586,
+      "step": 1241
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 1.7998818159103394,
+      "learning_rate": 0.0001780782918149466,
+      "loss": 3.0485,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5524444444444444,
+      "grad_norm": 1.7063149213790894,
+      "learning_rate": 0.00017806049822064056,
+      "loss": 2.6647,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5528888888888889,
+      "grad_norm": 1.4614622592926025,
+      "learning_rate": 0.00017804270462633452,
+      "loss": 2.2354,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5533333333333333,
+      "grad_norm": 1.6693414449691772,
+      "learning_rate": 0.00017802491103202848,
+      "loss": 2.1268,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5537777777777778,
+      "grad_norm": 1.6453592777252197,
+      "learning_rate": 0.00017800711743772243,
+      "loss": 2.6987,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5542222222222222,
+      "grad_norm": 1.7275609970092773,
+      "learning_rate": 0.0001779893238434164,
+      "loss": 2.4686,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5546666666666666,
+      "grad_norm": 1.6857225894927979,
+      "learning_rate": 0.00017797153024911032,
+      "loss": 1.6319,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5551111111111111,
+      "grad_norm": 1.822630524635315,
+      "learning_rate": 0.00017795373665480428,
+      "loss": 2.215,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 1.8363982439041138,
+      "learning_rate": 0.00017793594306049823,
+      "loss": 1.699,
+      "step": 1250
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 1.0663748979568481,
+      "learning_rate": 0.0001779181494661922,
+      "loss": 2.6193,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5564444444444444,
+      "grad_norm": 1.552035927772522,
+      "learning_rate": 0.00017790035587188615,
+      "loss": 1.0698,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5568888888888889,
+      "grad_norm": 1.1600538492202759,
+      "learning_rate": 0.0001778825622775801,
+      "loss": 2.5013,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5573333333333333,
+      "grad_norm": 1.0800617933273315,
+      "learning_rate": 0.00017786476868327403,
+      "loss": 2.5511,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5577777777777778,
+      "grad_norm": 1.054673433303833,
+      "learning_rate": 0.00017784697508896796,
+      "loss": 2.3978,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5582222222222222,
+      "grad_norm": 0.9816464185714722,
+      "learning_rate": 0.00017782918149466192,
+      "loss": 2.0309,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5586666666666666,
+      "grad_norm": 1.2822006940841675,
+      "learning_rate": 0.00017781138790035587,
+      "loss": 2.4745,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5591111111111111,
+      "grad_norm": 1.2752128839492798,
+      "learning_rate": 0.00017779359430604983,
+      "loss": 2.6353,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5595555555555556,
+      "grad_norm": 1.1391791105270386,
+      "learning_rate": 0.0001777758007117438,
+      "loss": 2.6778,
+      "step": 1259
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1968642473220825,
+      "learning_rate": 0.00017775800711743772,
+      "loss": 2.9373,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5604444444444444,
+      "grad_norm": 1.2563437223434448,
+      "learning_rate": 0.00017774021352313167,
+      "loss": 2.3472,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5608888888888889,
+      "grad_norm": 1.2173583507537842,
+      "learning_rate": 0.00017772241992882563,
+      "loss": 1.8898,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5613333333333334,
+      "grad_norm": 1.0731583833694458,
+      "learning_rate": 0.00017770462633451959,
+      "loss": 2.4451,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5617777777777778,
+      "grad_norm": 1.0820194482803345,
+      "learning_rate": 0.00017768683274021354,
+      "loss": 2.7584,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5622222222222222,
+      "grad_norm": 1.2503118515014648,
+      "learning_rate": 0.0001776690391459075,
+      "loss": 2.5674,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5626666666666666,
+      "grad_norm": 1.265372633934021,
+      "learning_rate": 0.00017765124555160143,
+      "loss": 2.3988,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5631111111111111,
+      "grad_norm": 1.1392314434051514,
+      "learning_rate": 0.00017763345195729536,
+      "loss": 2.5185,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5635555555555556,
+      "grad_norm": 1.2027145624160767,
+      "learning_rate": 0.00017761565836298931,
+      "loss": 2.5246,
+      "step": 1268
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 2.07536244392395,
+      "learning_rate": 0.00017759786476868327,
+      "loss": 2.7272,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5644444444444444,
+      "grad_norm": 1.5870450735092163,
+      "learning_rate": 0.00017758007117437723,
+      "loss": 3.0011,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5648888888888889,
+      "grad_norm": 1.439990520477295,
+      "learning_rate": 0.00017756227758007118,
+      "loss": 2.0491,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5653333333333334,
+      "grad_norm": 2.632349967956543,
+      "learning_rate": 0.00017754448398576514,
+      "loss": 2.4149,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5657777777777778,
+      "grad_norm": 1.4456804990768433,
+      "learning_rate": 0.00017752669039145907,
+      "loss": 2.7718,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5662222222222222,
+      "grad_norm": 1.3915135860443115,
+      "learning_rate": 0.00017750889679715303,
+      "loss": 3.1489,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5666666666666667,
+      "grad_norm": 1.4057570695877075,
+      "learning_rate": 0.00017749110320284698,
+      "loss": 2.7257,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5671111111111111,
+      "grad_norm": 1.498278260231018,
+      "learning_rate": 0.00017747330960854094,
+      "loss": 2.1074,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5675555555555556,
+      "grad_norm": 1.3324486017227173,
+      "learning_rate": 0.0001774555160142349,
+      "loss": 2.5449,
+      "step": 1277
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 1.2378911972045898,
+      "learning_rate": 0.00017743772241992885,
+      "loss": 1.8185,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5684444444444444,
+      "grad_norm": 1.7115696668624878,
+      "learning_rate": 0.00017741992882562278,
+      "loss": 3.1832,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5688888888888889,
+      "grad_norm": 1.381099820137024,
+      "learning_rate": 0.0001774021352313167,
+      "loss": 2.3531,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5693333333333334,
+      "grad_norm": 1.3072692155838013,
+      "learning_rate": 0.00017738434163701067,
+      "loss": 2.4594,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5697777777777778,
+      "grad_norm": 1.5001025199890137,
+      "learning_rate": 0.00017736654804270462,
+      "loss": 2.822,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5702222222222222,
+      "grad_norm": 1.440004587173462,
+      "learning_rate": 0.00017734875444839858,
+      "loss": 2.6287,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5706666666666667,
+      "grad_norm": 1.5406244993209839,
+      "learning_rate": 0.00017733096085409254,
+      "loss": 2.4,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5711111111111111,
+      "grad_norm": 1.3207265138626099,
+      "learning_rate": 0.0001773131672597865,
+      "loss": 2.2423,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5715555555555556,
+      "grad_norm": 1.3449149131774902,
+      "learning_rate": 0.00017729537366548042,
+      "loss": 2.1636,
+      "step": 1286
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 1.5409855842590332,
+      "learning_rate": 0.00017727758007117438,
+      "loss": 2.5376,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5724444444444444,
+      "grad_norm": 1.6577012538909912,
+      "learning_rate": 0.00017725978647686834,
+      "loss": 2.1492,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5728888888888889,
+      "grad_norm": 1.4876697063446045,
+      "learning_rate": 0.0001772419928825623,
+      "loss": 2.4557,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5733333333333334,
+      "grad_norm": 1.600537657737732,
+      "learning_rate": 0.00017722419928825625,
+      "loss": 2.3974,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5737777777777778,
+      "grad_norm": 1.856227159500122,
+      "learning_rate": 0.0001772064056939502,
+      "loss": 2.8138,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5742222222222222,
+      "grad_norm": 1.80988609790802,
+      "learning_rate": 0.00017718861209964414,
+      "loss": 1.4204,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5746666666666667,
+      "grad_norm": 1.5674011707305908,
+      "learning_rate": 0.00017717081850533806,
+      "loss": 2.2016,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5751111111111111,
+      "grad_norm": 1.5148704051971436,
+      "learning_rate": 0.00017715302491103202,
+      "loss": 2.607,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5755555555555556,
+      "grad_norm": 1.924994945526123,
+      "learning_rate": 0.00017713523131672598,
+      "loss": 2.69,
+      "step": 1295
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 1.8337801694869995,
+      "learning_rate": 0.00017711743772241993,
+      "loss": 2.1814,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5764444444444444,
+      "grad_norm": 1.7834872007369995,
+      "learning_rate": 0.0001770996441281139,
+      "loss": 2.7517,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5768888888888889,
+      "grad_norm": 1.7494984865188599,
+      "learning_rate": 0.00017708185053380785,
+      "loss": 2.2146,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5773333333333334,
+      "grad_norm": 1.8861663341522217,
+      "learning_rate": 0.00017706405693950178,
+      "loss": 2.5008,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5777777777777777,
+      "grad_norm": 3.3163883686065674,
+      "learning_rate": 0.00017704626334519573,
+      "loss": 3.1244,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5782222222222222,
+      "grad_norm": 1.1315196752548218,
+      "learning_rate": 0.0001770284697508897,
+      "loss": 2.7789,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5786666666666667,
+      "grad_norm": 1.067335844039917,
+      "learning_rate": 0.00017701067615658365,
+      "loss": 3.2978,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5791111111111111,
+      "grad_norm": 1.1568325757980347,
+      "learning_rate": 0.0001769928825622776,
+      "loss": 2.8684,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5795555555555556,
+      "grad_norm": 1.2416324615478516,
+      "learning_rate": 0.00017697508896797156,
+      "loss": 3.2694,
+      "step": 1304
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.24941086769104,
+      "learning_rate": 0.0001769572953736655,
+      "loss": 2.8852,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5804444444444444,
+      "grad_norm": 1.0672186613082886,
+      "learning_rate": 0.00017693950177935942,
+      "loss": 2.4795,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5808888888888889,
+      "grad_norm": 1.0531431436538696,
+      "learning_rate": 0.00017692170818505337,
+      "loss": 2.5932,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5813333333333334,
+      "grad_norm": 1.2098814249038696,
+      "learning_rate": 0.00017690391459074733,
+      "loss": 2.3703,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5817777777777777,
+      "grad_norm": 1.1954690217971802,
+      "learning_rate": 0.0001768861209964413,
+      "loss": 2.5722,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5822222222222222,
+      "grad_norm": 1.2348884344100952,
+      "learning_rate": 0.00017686832740213524,
+      "loss": 2.2636,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5826666666666667,
+      "grad_norm": 1.145476222038269,
+      "learning_rate": 0.0001768505338078292,
+      "loss": 3.0043,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5831111111111111,
+      "grad_norm": 1.1092824935913086,
+      "learning_rate": 0.00017683274021352313,
+      "loss": 2.7379,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5835555555555556,
+      "grad_norm": 1.696060061454773,
+      "learning_rate": 0.00017681494661921709,
+      "loss": 1.8912,
+      "step": 1313
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 1.3610656261444092,
+      "learning_rate": 0.00017679715302491104,
+      "loss": 2.5365,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5844444444444444,
+      "grad_norm": 1.2558561563491821,
+      "learning_rate": 0.000176779359430605,
+      "loss": 2.5177,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5848888888888889,
+      "grad_norm": 1.0652177333831787,
+      "learning_rate": 0.00017676156583629896,
+      "loss": 2.0853,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5853333333333334,
+      "grad_norm": 1.3081934452056885,
+      "learning_rate": 0.0001767437722419929,
+      "loss": 2.6939,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5857777777777777,
+      "grad_norm": 1.3675099611282349,
+      "learning_rate": 0.00017672597864768684,
+      "loss": 2.4198,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5862222222222222,
+      "grad_norm": 1.2688225507736206,
+      "learning_rate": 0.00017670818505338077,
+      "loss": 2.472,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "grad_norm": 1.7709518671035767,
+      "learning_rate": 0.00017669039145907473,
+      "loss": 1.2899,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5871111111111111,
+      "grad_norm": 1.2735425233840942,
+      "learning_rate": 0.00017667259786476868,
+      "loss": 2.3649,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5875555555555556,
+      "grad_norm": 1.4274276494979858,
+      "learning_rate": 0.00017665480427046264,
+      "loss": 3.0392,
+      "step": 1322
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 1.1845803260803223,
+      "learning_rate": 0.0001766370106761566,
+      "loss": 2.2422,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5884444444444444,
+      "grad_norm": 1.505797028541565,
+      "learning_rate": 0.00017661921708185055,
+      "loss": 2.6997,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5888888888888889,
+      "grad_norm": 1.500378131866455,
+      "learning_rate": 0.00017660142348754448,
+      "loss": 2.4131,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5893333333333334,
+      "grad_norm": 1.666031837463379,
+      "learning_rate": 0.00017658362989323844,
+      "loss": 2.4642,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5897777777777777,
+      "grad_norm": 1.4224402904510498,
+      "learning_rate": 0.0001765658362989324,
+      "loss": 2.6931,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5902222222222222,
+      "grad_norm": 1.5196523666381836,
+      "learning_rate": 0.00017654804270462635,
+      "loss": 2.846,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5906666666666667,
+      "grad_norm": 1.455924391746521,
+      "learning_rate": 0.0001765302491103203,
+      "loss": 2.6315,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5911111111111111,
+      "grad_norm": 2.582533597946167,
+      "learning_rate": 0.00017651245551601427,
+      "loss": 1.3039,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5915555555555555,
+      "grad_norm": 1.7225983142852783,
+      "learning_rate": 0.0001764946619217082,
+      "loss": 2.5404,
+      "step": 1331
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 1.50846529006958,
+      "learning_rate": 0.00017647686832740212,
+      "loss": 2.2202,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5924444444444444,
+      "grad_norm": 1.578640103340149,
+      "learning_rate": 0.00017645907473309608,
+      "loss": 2.3533,
+      "step": 1333
+    },
+    {
+      "epoch": 0.5928888888888889,
+      "grad_norm": 1.3282861709594727,
+      "learning_rate": 0.00017644128113879004,
+      "loss": 2.2071,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5933333333333334,
+      "grad_norm": 1.6370536088943481,
+      "learning_rate": 0.000176423487544484,
+      "loss": 2.9182,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5937777777777777,
+      "grad_norm": 1.1675159931182861,
+      "learning_rate": 0.00017640569395017795,
+      "loss": 1.177,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5942222222222222,
+      "grad_norm": 1.2795166969299316,
+      "learning_rate": 0.0001763879003558719,
+      "loss": 1.9403,
+      "step": 1337
+    },
+    {
+      "epoch": 0.5946666666666667,
+      "grad_norm": 1.825806975364685,
+      "learning_rate": 0.00017637010676156584,
+      "loss": 2.655,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5951111111111111,
+      "grad_norm": 1.5679066181182861,
+      "learning_rate": 0.0001763523131672598,
+      "loss": 2.3021,
+      "step": 1339
+    },
+    {
+      "epoch": 0.5955555555555555,
+      "grad_norm": 1.534218430519104,
+      "learning_rate": 0.00017633451957295375,
+      "loss": 2.5934,
+      "step": 1340
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 1.8854663372039795,
+      "learning_rate": 0.0001763167259786477,
+      "loss": 2.6493,
+      "step": 1341
+    },
+    {
+      "epoch": 0.5964444444444444,
+      "grad_norm": 1.5538815259933472,
+      "learning_rate": 0.00017629893238434166,
+      "loss": 2.6666,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5968888888888889,
+      "grad_norm": 1.702937364578247,
+      "learning_rate": 0.00017628113879003562,
+      "loss": 2.6987,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5973333333333334,
+      "grad_norm": 1.9548336267471313,
+      "learning_rate": 0.00017626334519572955,
+      "loss": 2.4912,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5977777777777777,
+      "grad_norm": 1.5537859201431274,
+      "learning_rate": 0.00017624555160142348,
+      "loss": 2.2085,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5982222222222222,
+      "grad_norm": 1.6140497922897339,
+      "learning_rate": 0.00017622775800711743,
+      "loss": 2.6939,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5986666666666667,
+      "grad_norm": 2.120786666870117,
+      "learning_rate": 0.0001762099644128114,
+      "loss": 2.8833,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5991111111111111,
+      "grad_norm": 1.8668750524520874,
+      "learning_rate": 0.00017619217081850535,
+      "loss": 2.412,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5995555555555555,
+      "grad_norm": 1.6141252517700195,
+      "learning_rate": 0.0001761743772241993,
+      "loss": 2.3202,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0876715183258057,
+      "learning_rate": 0.00017615658362989323,
+      "loss": 1.9042,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6004444444444444,
+      "grad_norm": 0.9330457448959351,
+      "learning_rate": 0.0001761387900355872,
+      "loss": 2.1474,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6008888888888889,
+      "grad_norm": 0.9378175139427185,
+      "learning_rate": 0.00017612099644128115,
+      "loss": 2.6252,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6013333333333334,
+      "grad_norm": 1.2432550191879272,
+      "learning_rate": 0.0001761032028469751,
+      "loss": 3.2561,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6017777777777777,
+      "grad_norm": 1.2766064405441284,
+      "learning_rate": 0.00017608540925266906,
+      "loss": 2.0631,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6022222222222222,
+      "grad_norm": 1.170651912689209,
+      "learning_rate": 0.00017606761565836301,
+      "loss": 3.0557,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6026666666666667,
+      "grad_norm": 1.724684476852417,
+      "learning_rate": 0.00017604982206405694,
+      "loss": 1.6797,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6031111111111112,
+      "grad_norm": 0.9961637258529663,
+      "learning_rate": 0.00017603202846975087,
+      "loss": 2.8432,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6035555555555555,
+      "grad_norm": 1.1883872747421265,
+      "learning_rate": 0.00017601423487544483,
+      "loss": 3.071,
+      "step": 1358
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 1.0519766807556152,
+      "learning_rate": 0.0001759964412811388,
+      "loss": 2.1003,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6044444444444445,
+      "grad_norm": 1.0907334089279175,
+      "learning_rate": 0.00017597864768683274,
+      "loss": 2.6082,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6048888888888889,
+      "grad_norm": 1.1930400133132935,
+      "learning_rate": 0.0001759608540925267,
+      "loss": 2.4901,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6053333333333333,
+      "grad_norm": 1.688298225402832,
+      "learning_rate": 0.00017594306049822066,
+      "loss": 1.7889,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6057777777777777,
+      "grad_norm": 1.192671537399292,
+      "learning_rate": 0.00017592526690391459,
+      "loss": 1.313,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6062222222222222,
+      "grad_norm": 1.3298887014389038,
+      "learning_rate": 0.00017590747330960854,
+      "loss": 3.2103,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6066666666666667,
+      "grad_norm": 1.1152617931365967,
+      "learning_rate": 0.0001758896797153025,
+      "loss": 2.4196,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6071111111111112,
+      "grad_norm": 1.2884933948516846,
+      "learning_rate": 0.00017587188612099646,
+      "loss": 2.6931,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6075555555555555,
+      "grad_norm": 1.1782528162002563,
+      "learning_rate": 0.0001758540925266904,
+      "loss": 2.0241,
+      "step": 1367
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 1.228033423423767,
+      "learning_rate": 0.00017583629893238437,
+      "loss": 2.1323,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6084444444444445,
+      "grad_norm": 1.3437577486038208,
+      "learning_rate": 0.0001758185053380783,
+      "loss": 1.3953,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6088888888888889,
+      "grad_norm": 1.3058631420135498,
+      "learning_rate": 0.00017580071174377223,
+      "loss": 2.6964,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6093333333333333,
+      "grad_norm": 1.219984531402588,
+      "learning_rate": 0.00017578291814946618,
+      "loss": 2.2271,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6097777777777778,
+      "grad_norm": 1.363558292388916,
+      "learning_rate": 0.00017576512455516014,
+      "loss": 2.5532,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6102222222222222,
+      "grad_norm": 1.8681366443634033,
+      "learning_rate": 0.0001757473309608541,
+      "loss": 1.3838,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6106666666666667,
+      "grad_norm": 1.3558433055877686,
+      "learning_rate": 0.00017572953736654805,
+      "loss": 2.6416,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6111111111111112,
+      "grad_norm": 1.3465511798858643,
+      "learning_rate": 0.000175711743772242,
+      "loss": 2.3744,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6115555555555555,
+      "grad_norm": 1.6007107496261597,
+      "learning_rate": 0.00017569395017793594,
+      "loss": 2.3307,
+      "step": 1376
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 1.602541208267212,
+      "learning_rate": 0.0001756761565836299,
+      "loss": 2.5173,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6124444444444445,
+      "grad_norm": 1.6473790407180786,
+      "learning_rate": 0.00017565836298932385,
+      "loss": 2.6707,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6128888888888889,
+      "grad_norm": 1.376232385635376,
+      "learning_rate": 0.0001756405693950178,
+      "loss": 1.9534,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6133333333333333,
+      "grad_norm": 1.420652151107788,
+      "learning_rate": 0.00017562277580071176,
+      "loss": 2.5401,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6137777777777778,
+      "grad_norm": 1.1767398118972778,
+      "learning_rate": 0.00017560498220640572,
+      "loss": 1.1765,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6142222222222222,
+      "grad_norm": 1.3465992212295532,
+      "learning_rate": 0.00017558718861209965,
+      "loss": 2.0177,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6146666666666667,
+      "grad_norm": 1.6473318338394165,
+      "learning_rate": 0.00017556939501779358,
+      "loss": 2.2605,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6151111111111112,
+      "grad_norm": 1.4791382551193237,
+      "learning_rate": 0.00017555160142348754,
+      "loss": 2.494,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6155555555555555,
+      "grad_norm": 1.4145492315292358,
+      "learning_rate": 0.0001755338078291815,
+      "loss": 2.1339,
+      "step": 1385
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 1.6023871898651123,
+      "learning_rate": 0.00017551601423487545,
+      "loss": 2.8089,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6164444444444445,
+      "grad_norm": 1.548842191696167,
+      "learning_rate": 0.0001754982206405694,
+      "loss": 2.2959,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6168888888888889,
+      "grad_norm": 1.6211026906967163,
+      "learning_rate": 0.00017548042704626336,
+      "loss": 2.6256,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6173333333333333,
+      "grad_norm": 1.508934497833252,
+      "learning_rate": 0.0001754626334519573,
+      "loss": 2.4794,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6177777777777778,
+      "grad_norm": 1.704952359199524,
+      "learning_rate": 0.00017544483985765125,
+      "loss": 2.6322,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6182222222222222,
+      "grad_norm": 1.4746431112289429,
+      "learning_rate": 0.0001754270462633452,
+      "loss": 2.2247,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6186666666666667,
+      "grad_norm": 1.5690298080444336,
+      "learning_rate": 0.00017540925266903916,
+      "loss": 2.1532,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6191111111111111,
+      "grad_norm": 1.7075462341308594,
+      "learning_rate": 0.00017539145907473312,
+      "loss": 2.864,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6195555555555555,
+      "grad_norm": 1.909990906715393,
+      "learning_rate": 0.00017537366548042707,
+      "loss": 2.8461,
+      "step": 1394
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7880994081497192,
+      "learning_rate": 0.000175355871886121,
+      "loss": 2.6034,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6204444444444445,
+      "grad_norm": 1.8660807609558105,
+      "learning_rate": 0.00017533807829181493,
+      "loss": 2.372,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6208888888888889,
+      "grad_norm": 1.9713836908340454,
+      "learning_rate": 0.0001753202846975089,
+      "loss": 2.4238,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6213333333333333,
+      "grad_norm": 2.378556966781616,
+      "learning_rate": 0.00017530249110320285,
+      "loss": 2.9749,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6217777777777778,
+      "grad_norm": 3.1581037044525146,
+      "learning_rate": 0.0001752846975088968,
+      "loss": 0.1631,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 2.725257396697998,
+      "learning_rate": 0.00017526690391459076,
+      "loss": 2.3153,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6226666666666667,
+      "grad_norm": 1.687172770500183,
+      "learning_rate": 0.00017524911032028472,
+      "loss": 1.8111,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6231111111111111,
+      "grad_norm": 1.2486687898635864,
+      "learning_rate": 0.00017523131672597865,
+      "loss": 1.6803,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6235555555555555,
+      "grad_norm": 0.9351308345794678,
+      "learning_rate": 0.0001752135231316726,
+      "loss": 2.5876,
+      "step": 1403
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 1.1229432821273804,
+      "learning_rate": 0.00017519572953736656,
+      "loss": 2.5936,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6244444444444445,
+      "grad_norm": 0.9976657629013062,
+      "learning_rate": 0.00017517793594306051,
+      "loss": 2.7395,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6248888888888889,
+      "grad_norm": 0.9780849814414978,
+      "learning_rate": 0.00017516014234875447,
+      "loss": 2.1686,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6253333333333333,
+      "grad_norm": 1.0690516233444214,
+      "learning_rate": 0.00017514234875444843,
+      "loss": 2.2365,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6257777777777778,
+      "grad_norm": 1.1127312183380127,
+      "learning_rate": 0.00017512455516014236,
+      "loss": 2.5104,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6262222222222222,
+      "grad_norm": 1.1946388483047485,
+      "learning_rate": 0.00017510676156583629,
+      "loss": 2.7337,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6266666666666667,
+      "grad_norm": 1.0641489028930664,
+      "learning_rate": 0.00017508896797153024,
+      "loss": 2.5558,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6271111111111111,
+      "grad_norm": 1.127685546875,
+      "learning_rate": 0.0001750711743772242,
+      "loss": 2.517,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6275555555555555,
+      "grad_norm": 1.2236816883087158,
+      "learning_rate": 0.00017505338078291816,
+      "loss": 2.6978,
+      "step": 1412
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 1.4464499950408936,
+      "learning_rate": 0.0001750355871886121,
+      "loss": 2.6535,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6284444444444445,
+      "grad_norm": 1.4766771793365479,
+      "learning_rate": 0.00017501779359430607,
+      "loss": 3.3998,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6288888888888889,
+      "grad_norm": 1.2874871492385864,
+      "learning_rate": 0.000175,
+      "loss": 2.3482,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6293333333333333,
+      "grad_norm": 1.3195867538452148,
+      "learning_rate": 0.00017498220640569395,
+      "loss": 2.7828,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6297777777777778,
+      "grad_norm": 1.2177435159683228,
+      "learning_rate": 0.0001749644128113879,
+      "loss": 2.3438,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6302222222222222,
+      "grad_norm": 1.236802577972412,
+      "learning_rate": 0.00017494661921708187,
+      "loss": 2.6197,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6306666666666667,
+      "grad_norm": 1.363099217414856,
+      "learning_rate": 0.00017492882562277582,
+      "loss": 2.6425,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6311111111111111,
+      "grad_norm": 2.4340715408325195,
+      "learning_rate": 0.00017491103202846978,
+      "loss": 1.5395,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6315555555555555,
+      "grad_norm": 1.425446629524231,
+      "learning_rate": 0.0001748932384341637,
+      "loss": 2.6681,
+      "step": 1421
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 1.3142110109329224,
+      "learning_rate": 0.00017487544483985764,
+      "loss": 2.53,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6324444444444445,
+      "grad_norm": 1.947712779045105,
+      "learning_rate": 0.0001748576512455516,
+      "loss": 2.8098,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6328888888888888,
+      "grad_norm": 1.202399730682373,
+      "learning_rate": 0.00017483985765124555,
+      "loss": 1.8831,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6333333333333333,
+      "grad_norm": 1.3699499368667603,
+      "learning_rate": 0.0001748220640569395,
+      "loss": 2.9502,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6337777777777778,
+      "grad_norm": 1.4097647666931152,
+      "learning_rate": 0.00017480427046263347,
+      "loss": 2.8757,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6342222222222222,
+      "grad_norm": 1.3218767642974854,
+      "learning_rate": 0.00017478647686832742,
+      "loss": 2.299,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6346666666666667,
+      "grad_norm": 1.4610822200775146,
+      "learning_rate": 0.00017476868327402135,
+      "loss": 2.5177,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6351111111111111,
+      "grad_norm": 1.65794038772583,
+      "learning_rate": 0.0001747508896797153,
+      "loss": 2.8784,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6355555555555555,
+      "grad_norm": 1.5559744834899902,
+      "learning_rate": 0.00017473309608540926,
+      "loss": 2.3513,
+      "step": 1430
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 1.8620177507400513,
+      "learning_rate": 0.00017471530249110322,
+      "loss": 0.9563,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6364444444444445,
+      "grad_norm": 1.3389177322387695,
+      "learning_rate": 0.00017469750889679718,
+      "loss": 1.5891,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6368888888888888,
+      "grad_norm": 1.3747659921646118,
+      "learning_rate": 0.00017467971530249113,
+      "loss": 1.4177,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6373333333333333,
+      "grad_norm": 1.6347014904022217,
+      "learning_rate": 0.00017466192170818506,
+      "loss": 2.359,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6377777777777778,
+      "grad_norm": 1.6315451860427856,
+      "learning_rate": 0.000174644128113879,
+      "loss": 3.0319,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6382222222222222,
+      "grad_norm": 1.5996496677398682,
+      "learning_rate": 0.00017462633451957295,
+      "loss": 2.3993,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6386666666666667,
+      "grad_norm": 1.7360764741897583,
+      "learning_rate": 0.0001746085409252669,
+      "loss": 2.3704,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6391111111111111,
+      "grad_norm": 1.6707364320755005,
+      "learning_rate": 0.00017459074733096086,
+      "loss": 2.4405,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6395555555555555,
+      "grad_norm": 1.5770982503890991,
+      "learning_rate": 0.00017457295373665482,
+      "loss": 2.4132,
+      "step": 1439
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9009085893630981,
+      "learning_rate": 0.00017455516014234875,
+      "loss": 2.6834,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6404444444444445,
+      "grad_norm": 1.7474616765975952,
+      "learning_rate": 0.0001745373665480427,
+      "loss": 2.341,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6408888888888888,
+      "grad_norm": 2.131709337234497,
+      "learning_rate": 0.00017451957295373666,
+      "loss": 2.554,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6413333333333333,
+      "grad_norm": 2.146632194519043,
+      "learning_rate": 0.00017450177935943062,
+      "loss": 2.8733,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6417777777777778,
+      "grad_norm": 2.0064449310302734,
+      "learning_rate": 0.00017448398576512457,
+      "loss": 2.4239,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6422222222222222,
+      "grad_norm": 2.1821370124816895,
+      "learning_rate": 0.00017446619217081853,
+      "loss": 3.9109,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6426666666666667,
+      "grad_norm": 1.7780088186264038,
+      "learning_rate": 0.00017444839857651246,
+      "loss": 2.3761,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6431111111111111,
+      "grad_norm": 1.9316993951797485,
+      "learning_rate": 0.00017443060498220642,
+      "loss": 2.977,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6435555555555555,
+      "grad_norm": 2.113534927368164,
+      "learning_rate": 0.00017441281138790035,
+      "loss": 2.6501,
+      "step": 1448
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 2.1450841426849365,
+      "learning_rate": 0.0001743950177935943,
+      "loss": 1.5886,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6444444444444445,
+      "grad_norm": 2.311339855194092,
+      "learning_rate": 0.00017437722419928826,
+      "loss": 2.0959,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6448888888888888,
+      "grad_norm": 3.696924924850464,
+      "learning_rate": 0.00017435943060498222,
+      "loss": 1.73,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6453333333333333,
+      "grad_norm": 1.0169023275375366,
+      "learning_rate": 0.00017434163701067617,
+      "loss": 2.5696,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6457777777777778,
+      "grad_norm": 1.278465747833252,
+      "learning_rate": 0.0001743238434163701,
+      "loss": 2.876,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6462222222222223,
+      "grad_norm": 1.182702898979187,
+      "learning_rate": 0.00017430604982206406,
+      "loss": 2.4447,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6466666666666666,
+      "grad_norm": 1.3612101078033447,
+      "learning_rate": 0.00017428825622775801,
+      "loss": 2.5971,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6471111111111111,
+      "grad_norm": 1.2513208389282227,
+      "learning_rate": 0.00017427046263345197,
+      "loss": 1.7877,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6475555555555556,
+      "grad_norm": 1.108196496963501,
+      "learning_rate": 0.00017425266903914593,
+      "loss": 2.2443,
+      "step": 1457
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 1.3617606163024902,
+      "learning_rate": 0.00017423487544483988,
+      "loss": 2.4862,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6484444444444445,
+      "grad_norm": 1.2257615327835083,
+      "learning_rate": 0.0001742170818505338,
+      "loss": 2.6374,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6488888888888888,
+      "grad_norm": 1.6558245420455933,
+      "learning_rate": 0.00017419928825622777,
+      "loss": 1.4984,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6493333333333333,
+      "grad_norm": 1.308287262916565,
+      "learning_rate": 0.0001741814946619217,
+      "loss": 2.5151,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6497777777777778,
+      "grad_norm": 1.210597038269043,
+      "learning_rate": 0.00017416370106761566,
+      "loss": 2.0015,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6502222222222223,
+      "grad_norm": 1.6040871143341064,
+      "learning_rate": 0.0001741459074733096,
+      "loss": 2.5846,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6506666666666666,
+      "grad_norm": 1.5346297025680542,
+      "learning_rate": 0.00017412811387900357,
+      "loss": 2.6826,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6511111111111111,
+      "grad_norm": 1.5624873638153076,
+      "learning_rate": 0.00017411032028469752,
+      "loss": 2.7424,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6515555555555556,
+      "grad_norm": 1.3819774389266968,
+      "learning_rate": 0.00017409252669039145,
+      "loss": 2.7193,
+      "step": 1466
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 1.2688080072402954,
+      "learning_rate": 0.0001740747330960854,
+      "loss": 2.2891,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6524444444444445,
+      "grad_norm": 1.2661856412887573,
+      "learning_rate": 0.00017405693950177937,
+      "loss": 1.2741,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6528888888888889,
+      "grad_norm": 1.215834617614746,
+      "learning_rate": 0.00017403914590747332,
+      "loss": 2.3391,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6533333333333333,
+      "grad_norm": 1.450677752494812,
+      "learning_rate": 0.00017402135231316728,
+      "loss": 2.752,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6537777777777778,
+      "grad_norm": 1.2563883066177368,
+      "learning_rate": 0.00017400355871886124,
+      "loss": 2.3388,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6542222222222223,
+      "grad_norm": 1.5057649612426758,
+      "learning_rate": 0.00017398576512455517,
+      "loss": 2.4776,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6546666666666666,
+      "grad_norm": 1.5081661939620972,
+      "learning_rate": 0.00017396797153024912,
+      "loss": 2.4154,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6551111111111111,
+      "grad_norm": 1.3150933980941772,
+      "learning_rate": 0.00017395017793594305,
+      "loss": 2.4713,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6555555555555556,
+      "grad_norm": 1.830236554145813,
+      "learning_rate": 0.000173932384341637,
+      "loss": 2.4647,
+      "step": 1475
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 1.368726372718811,
+      "learning_rate": 0.00017391459074733097,
+      "loss": 2.1934,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6564444444444445,
+      "grad_norm": 1.3172972202301025,
+      "learning_rate": 0.00017389679715302492,
+      "loss": 2.269,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6568888888888889,
+      "grad_norm": 1.7238434553146362,
+      "learning_rate": 0.00017387900355871888,
+      "loss": 2.5588,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6573333333333333,
+      "grad_norm": 1.5723549127578735,
+      "learning_rate": 0.0001738612099644128,
+      "loss": 2.3316,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6577777777777778,
+      "grad_norm": 1.4798550605773926,
+      "learning_rate": 0.00017384341637010676,
+      "loss": 2.2944,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6582222222222223,
+      "grad_norm": 1.411055564880371,
+      "learning_rate": 0.00017382562277580072,
+      "loss": 2.5573,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6586666666666666,
+      "grad_norm": 2.178536891937256,
+      "learning_rate": 0.00017380782918149468,
+      "loss": 2.6412,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6591111111111111,
+      "grad_norm": 1.48294198513031,
+      "learning_rate": 0.00017379003558718863,
+      "loss": 2.2673,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6595555555555556,
+      "grad_norm": 1.597641944885254,
+      "learning_rate": 0.0001737722419928826,
+      "loss": 2.1974,
+      "step": 1484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7174814939498901,
+      "learning_rate": 0.00017375444839857652,
+      "loss": 2.7182,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6604444444444444,
+      "grad_norm": 2.032003879547119,
+      "learning_rate": 0.00017373665480427045,
+      "loss": 3.3137,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6608888888888889,
+      "grad_norm": 1.495666742324829,
+      "learning_rate": 0.0001737188612099644,
+      "loss": 1.8833,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6613333333333333,
+      "grad_norm": 1.6871966123580933,
+      "learning_rate": 0.00017370106761565836,
+      "loss": 3.0253,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6617777777777778,
+      "grad_norm": 1.5694763660430908,
+      "learning_rate": 0.00017368327402135232,
+      "loss": 1.9472,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6622222222222223,
+      "grad_norm": 1.8708395957946777,
+      "learning_rate": 0.00017366548042704627,
+      "loss": 2.8555,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6626666666666666,
+      "grad_norm": 1.8783323764801025,
+      "learning_rate": 0.00017364768683274023,
+      "loss": 2.4134,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6631111111111111,
+      "grad_norm": 2.046388626098633,
+      "learning_rate": 0.00017362989323843416,
+      "loss": 2.7402,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6635555555555556,
+      "grad_norm": 1.5785688161849976,
+      "learning_rate": 0.00017361209964412812,
+      "loss": 2.4242,
+      "step": 1493
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 1.7086628675460815,
+      "learning_rate": 0.00017359430604982207,
+      "loss": 2.3236,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6644444444444444,
+      "grad_norm": 1.5549596548080444,
+      "learning_rate": 0.00017357651245551603,
+      "loss": 2.23,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6648888888888889,
+      "grad_norm": 2.255401372909546,
+      "learning_rate": 0.00017355871886121,
+      "loss": 3.0353,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6653333333333333,
+      "grad_norm": 1.643557071685791,
+      "learning_rate": 0.00017354092526690394,
+      "loss": 2.4116,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6657777777777778,
+      "grad_norm": 2.1392343044281006,
+      "learning_rate": 0.00017352313167259787,
+      "loss": 3.4415,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6662222222222223,
+      "grad_norm": 2.6055009365081787,
+      "learning_rate": 0.0001735053380782918,
+      "loss": 0.1766,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 2.152284622192383,
+      "learning_rate": 0.00017348754448398576,
+      "loss": 2.9102,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6671111111111111,
+      "grad_norm": 0.9273963570594788,
+      "learning_rate": 0.00017346975088967971,
+      "loss": 2.428,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6675555555555556,
+      "grad_norm": 0.9428668022155762,
+      "learning_rate": 0.00017345195729537367,
+      "loss": 2.6534,
+      "step": 1502
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 0.93215012550354,
+      "learning_rate": 0.00017343416370106763,
+      "loss": 2.351,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6684444444444444,
+      "grad_norm": 1.035394549369812,
+      "learning_rate": 0.00017341637010676158,
+      "loss": 1.1251,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6688888888888889,
+      "grad_norm": 1.0628288984298706,
+      "learning_rate": 0.00017339857651245551,
+      "loss": 2.4944,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6693333333333333,
+      "grad_norm": 1.3216973543167114,
+      "learning_rate": 0.00017338078291814947,
+      "loss": 2.5529,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6697777777777778,
+      "grad_norm": 1.2949331998825073,
+      "learning_rate": 0.00017336298932384343,
+      "loss": 2.5878,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6702222222222223,
+      "grad_norm": 1.3365072011947632,
+      "learning_rate": 0.00017334519572953738,
+      "loss": 3.059,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6706666666666666,
+      "grad_norm": 1.181065320968628,
+      "learning_rate": 0.00017332740213523134,
+      "loss": 2.6428,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6711111111111111,
+      "grad_norm": 1.2061887979507446,
+      "learning_rate": 0.0001733096085409253,
+      "loss": 1.4412,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6715555555555556,
+      "grad_norm": 1.4257198572158813,
+      "learning_rate": 0.00017329181494661923,
+      "loss": 2.9952,
+      "step": 1511
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 1.2718660831451416,
+      "learning_rate": 0.00017327402135231316,
+      "loss": 2.3034,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6724444444444444,
+      "grad_norm": 1.2620964050292969,
+      "learning_rate": 0.0001732562277580071,
+      "loss": 1.7634,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6728888888888889,
+      "grad_norm": 1.4162237644195557,
+      "learning_rate": 0.00017323843416370107,
+      "loss": 2.2134,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6733333333333333,
+      "grad_norm": 1.3909696340560913,
+      "learning_rate": 0.00017322064056939502,
+      "loss": 2.7267,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6737777777777778,
+      "grad_norm": 1.563040852546692,
+      "learning_rate": 0.00017320284697508898,
+      "loss": 1.83,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6742222222222222,
+      "grad_norm": 1.417112112045288,
+      "learning_rate": 0.00017318505338078294,
+      "loss": 2.1559,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6746666666666666,
+      "grad_norm": 1.617037057876587,
+      "learning_rate": 0.00017316725978647687,
+      "loss": 2.7987,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6751111111111111,
+      "grad_norm": 1.5133682489395142,
+      "learning_rate": 0.00017314946619217082,
+      "loss": 2.5342,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6755555555555556,
+      "grad_norm": 1.3706659078598022,
+      "learning_rate": 0.00017313167259786478,
+      "loss": 2.583,
+      "step": 1520
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 1.600906252861023,
+      "learning_rate": 0.00017311387900355874,
+      "loss": 2.7443,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6764444444444444,
+      "grad_norm": 1.269012689590454,
+      "learning_rate": 0.0001730960854092527,
+      "loss": 2.3754,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6768888888888889,
+      "grad_norm": 1.382144808769226,
+      "learning_rate": 0.00017307829181494665,
+      "loss": 2.5199,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6773333333333333,
+      "grad_norm": 1.71562659740448,
+      "learning_rate": 0.00017306049822064058,
+      "loss": 3.278,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6777777777777778,
+      "grad_norm": 1.2466914653778076,
+      "learning_rate": 0.0001730427046263345,
+      "loss": 2.3992,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6782222222222222,
+      "grad_norm": 1.547672986984253,
+      "learning_rate": 0.00017302491103202846,
+      "loss": 2.4622,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6786666666666666,
+      "grad_norm": 1.5382349491119385,
+      "learning_rate": 0.00017300711743772242,
+      "loss": 2.1739,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6791111111111111,
+      "grad_norm": 1.4863885641098022,
+      "learning_rate": 0.00017298932384341638,
+      "loss": 2.6332,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6795555555555556,
+      "grad_norm": 1.6594899892807007,
+      "learning_rate": 0.00017297153024911033,
+      "loss": 3.0222,
+      "step": 1529
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.72464120388031,
+      "learning_rate": 0.00017295373665480426,
+      "loss": 2.6409,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6804444444444444,
+      "grad_norm": 1.5191189050674438,
+      "learning_rate": 0.00017293594306049822,
+      "loss": 2.6336,
+      "step": 1531
+    },
+    {
+      "epoch": 0.6808888888888889,
+      "grad_norm": 1.8320003747940063,
+      "learning_rate": 0.00017291814946619218,
+      "loss": 2.7605,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6813333333333333,
+      "grad_norm": 1.6135400533676147,
+      "learning_rate": 0.00017290035587188613,
+      "loss": 2.6308,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6817777777777778,
+      "grad_norm": 1.8295135498046875,
+      "learning_rate": 0.0001728825622775801,
+      "loss": 2.5421,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6822222222222222,
+      "grad_norm": 1.799838900566101,
+      "learning_rate": 0.00017286476868327405,
+      "loss": 3.0082,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6826666666666666,
+      "grad_norm": 1.427129864692688,
+      "learning_rate": 0.00017284697508896798,
+      "loss": 2.5288,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6831111111111111,
+      "grad_norm": 1.533745527267456,
+      "learning_rate": 0.00017282918149466193,
+      "loss": 2.7791,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6835555555555556,
+      "grad_norm": 1.6388285160064697,
+      "learning_rate": 0.00017281138790035586,
+      "loss": 3.325,
+      "step": 1538
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 1.5030821561813354,
+      "learning_rate": 0.00017279359430604982,
+      "loss": 2.0334,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6844444444444444,
+      "grad_norm": 1.2664015293121338,
+      "learning_rate": 0.00017277580071174377,
+      "loss": 2.0804,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6848888888888889,
+      "grad_norm": 1.496182918548584,
+      "learning_rate": 0.00017275800711743773,
+      "loss": 1.6106,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6853333333333333,
+      "grad_norm": 1.7371916770935059,
+      "learning_rate": 0.0001727402135231317,
+      "loss": 2.6694,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6857777777777778,
+      "grad_norm": 1.8223196268081665,
+      "learning_rate": 0.00017272241992882562,
+      "loss": 2.5409,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6862222222222222,
+      "grad_norm": 1.9429682493209839,
+      "learning_rate": 0.00017270462633451957,
+      "loss": 2.5704,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6866666666666666,
+      "grad_norm": 1.7982163429260254,
+      "learning_rate": 0.00017268683274021353,
+      "loss": 2.9834,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6871111111111111,
+      "grad_norm": 1.8245515823364258,
+      "learning_rate": 0.00017266903914590749,
+      "loss": 2.6373,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6875555555555556,
+      "grad_norm": 1.7472467422485352,
+      "learning_rate": 0.00017265124555160144,
+      "loss": 2.8218,
+      "step": 1547
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 1.9026468992233276,
+      "learning_rate": 0.0001726334519572954,
+      "loss": 3.0626,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6884444444444444,
+      "grad_norm": 2.187288522720337,
+      "learning_rate": 0.00017261565836298933,
+      "loss": 2.3243,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6888888888888889,
+      "grad_norm": 2.155287742614746,
+      "learning_rate": 0.00017259786476868329,
+      "loss": 3.0798,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6893333333333334,
+      "grad_norm": 2.2100751399993896,
+      "learning_rate": 0.00017258007117437721,
+      "loss": 2.1143,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6897777777777778,
+      "grad_norm": 1.0366290807724,
+      "learning_rate": 0.00017256227758007117,
+      "loss": 2.1292,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6902222222222222,
+      "grad_norm": 1.2031378746032715,
+      "learning_rate": 0.00017254448398576513,
+      "loss": 2.6824,
+      "step": 1553
+    },
+    {
+      "epoch": 0.6906666666666667,
+      "grad_norm": 1.154441475868225,
+      "learning_rate": 0.00017252669039145908,
+      "loss": 2.203,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6911111111111111,
+      "grad_norm": 1.2247297763824463,
+      "learning_rate": 0.00017250889679715304,
+      "loss": 2.8808,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6915555555555556,
+      "grad_norm": 1.3467925786972046,
+      "learning_rate": 0.00017249110320284697,
+      "loss": 2.9562,
+      "step": 1556
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 1.1302120685577393,
+      "learning_rate": 0.00017247330960854093,
+      "loss": 2.485,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6924444444444444,
+      "grad_norm": 1.2643386125564575,
+      "learning_rate": 0.00017245551601423488,
+      "loss": 2.572,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6928888888888889,
+      "grad_norm": 1.353574514389038,
+      "learning_rate": 0.00017243772241992884,
+      "loss": 2.671,
+      "step": 1559
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "grad_norm": 1.7643210887908936,
+      "learning_rate": 0.0001724199288256228,
+      "loss": 2.6473,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6937777777777778,
+      "grad_norm": 1.5903198719024658,
+      "learning_rate": 0.00017240213523131675,
+      "loss": 2.2954,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6942222222222222,
+      "grad_norm": 1.627631425857544,
+      "learning_rate": 0.00017238434163701068,
+      "loss": 2.233,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6946666666666667,
+      "grad_norm": 1.3801125288009644,
+      "learning_rate": 0.00017236654804270464,
+      "loss": 2.2972,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6951111111111111,
+      "grad_norm": 1.451066017150879,
+      "learning_rate": 0.00017234875444839857,
+      "loss": 2.013,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6955555555555556,
+      "grad_norm": 1.316686987876892,
+      "learning_rate": 0.00017233096085409252,
+      "loss": 2.3651,
+      "step": 1565
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 1.300595760345459,
+      "learning_rate": 0.00017231316725978648,
+      "loss": 2.0519,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6964444444444444,
+      "grad_norm": 1.4267830848693848,
+      "learning_rate": 0.00017229537366548044,
+      "loss": 2.2664,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6968888888888889,
+      "grad_norm": 1.5179320573806763,
+      "learning_rate": 0.0001722775800711744,
+      "loss": 2.3228,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6973333333333334,
+      "grad_norm": 1.7718604803085327,
+      "learning_rate": 0.00017225978647686832,
+      "loss": 2.6723,
+      "step": 1569
+    },
+    {
+      "epoch": 0.6977777777777778,
+      "grad_norm": 2.2856781482696533,
+      "learning_rate": 0.00017224199288256228,
+      "loss": 2.3014,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6982222222222222,
+      "grad_norm": 1.3740836381912231,
+      "learning_rate": 0.00017222419928825624,
+      "loss": 2.7384,
+      "step": 1571
+    },
+    {
+      "epoch": 0.6986666666666667,
+      "grad_norm": 1.3104565143585205,
+      "learning_rate": 0.0001722064056939502,
+      "loss": 2.4267,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6991111111111111,
+      "grad_norm": 1.5572513341903687,
+      "learning_rate": 0.00017218861209964415,
+      "loss": 2.7882,
+      "step": 1573
+    },
+    {
+      "epoch": 0.6995555555555556,
+      "grad_norm": 1.344378113746643,
+      "learning_rate": 0.0001721708185053381,
+      "loss": 2.6669,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6647869348526,
+      "learning_rate": 0.00017215302491103203,
+      "loss": 2.5764,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7004444444444444,
+      "grad_norm": 1.4547927379608154,
+      "learning_rate": 0.000172135231316726,
+      "loss": 2.4882,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7008888888888889,
+      "grad_norm": 1.69290030002594,
+      "learning_rate": 0.00017211743772241992,
+      "loss": 2.6278,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7013333333333334,
+      "grad_norm": 1.5832207202911377,
+      "learning_rate": 0.00017209964412811388,
+      "loss": 2.3207,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7017777777777777,
+      "grad_norm": 1.6772409677505493,
+      "learning_rate": 0.00017208185053380783,
+      "loss": 2.9327,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7022222222222222,
+      "grad_norm": 1.6155133247375488,
+      "learning_rate": 0.0001720640569395018,
+      "loss": 2.6978,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7026666666666667,
+      "grad_norm": 1.2734161615371704,
+      "learning_rate": 0.00017204626334519575,
+      "loss": 1.9025,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7031111111111111,
+      "grad_norm": 1.4908726215362549,
+      "learning_rate": 0.00017202846975088968,
+      "loss": 2.1544,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7035555555555556,
+      "grad_norm": 1.7874783277511597,
+      "learning_rate": 0.00017201067615658363,
+      "loss": 3.0701,
+      "step": 1583
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 2.0259952545166016,
+      "learning_rate": 0.0001719928825622776,
+      "loss": 2.926,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7044444444444444,
+      "grad_norm": 1.7444260120391846,
+      "learning_rate": 0.00017197508896797155,
+      "loss": 2.2407,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7048888888888889,
+      "grad_norm": 1.7600386142730713,
+      "learning_rate": 0.0001719572953736655,
+      "loss": 2.6499,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7053333333333334,
+      "grad_norm": 1.7595195770263672,
+      "learning_rate": 0.00017193950177935946,
+      "loss": 2.4446,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7057777777777777,
+      "grad_norm": 1.9496681690216064,
+      "learning_rate": 0.0001719217081850534,
+      "loss": 3.0214,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7062222222222222,
+      "grad_norm": 1.8185930252075195,
+      "learning_rate": 0.00017190391459074734,
+      "loss": 2.5126,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7066666666666667,
+      "grad_norm": 1.7105134725570679,
+      "learning_rate": 0.00017188612099644127,
+      "loss": 2.6629,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7071111111111111,
+      "grad_norm": 1.8199312686920166,
+      "learning_rate": 0.00017186832740213523,
+      "loss": 2.2038,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7075555555555556,
+      "grad_norm": 1.6544042825698853,
+      "learning_rate": 0.0001718505338078292,
+      "loss": 2.4603,
+      "step": 1592
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 1.5924146175384521,
+      "learning_rate": 0.00017183274021352314,
+      "loss": 2.5588,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7084444444444444,
+      "grad_norm": 1.6457511186599731,
+      "learning_rate": 0.0001718149466192171,
+      "loss": 2.7249,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7088888888888889,
+      "grad_norm": 1.7341830730438232,
+      "learning_rate": 0.00017179715302491103,
+      "loss": 2.7286,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7093333333333334,
+      "grad_norm": 1.8553625345230103,
+      "learning_rate": 0.00017177935943060499,
+      "loss": 3.2606,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7097777777777777,
+      "grad_norm": 2.046403408050537,
+      "learning_rate": 0.00017176156583629894,
+      "loss": 2.7661,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7102222222222222,
+      "grad_norm": 1.9650744199752808,
+      "learning_rate": 0.0001717437722419929,
+      "loss": 2.7783,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7106666666666667,
+      "grad_norm": 2.4216885566711426,
+      "learning_rate": 0.00017172597864768686,
+      "loss": 2.6446,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 2.498917818069458,
+      "learning_rate": 0.0001717081850533808,
+      "loss": 3.2051,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7115555555555556,
+      "grad_norm": 1.0230510234832764,
+      "learning_rate": 0.00017169039145907474,
+      "loss": 2.7331,
+      "step": 1601
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 1.6314057111740112,
+      "learning_rate": 0.00017167259786476867,
+      "loss": 1.3376,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7124444444444444,
+      "grad_norm": 1.033645510673523,
+      "learning_rate": 0.00017165480427046263,
+      "loss": 2.7086,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7128888888888889,
+      "grad_norm": 1.1229987144470215,
+      "learning_rate": 0.00017163701067615658,
+      "loss": 2.2418,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7133333333333334,
+      "grad_norm": 1.2182966470718384,
+      "learning_rate": 0.00017161921708185054,
+      "loss": 2.8067,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7137777777777777,
+      "grad_norm": 1.102390170097351,
+      "learning_rate": 0.0001716014234875445,
+      "loss": 3.0503,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7142222222222222,
+      "grad_norm": 1.0558178424835205,
+      "learning_rate": 0.00017158362989323845,
+      "loss": 3.2076,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7146666666666667,
+      "grad_norm": 1.236777424812317,
+      "learning_rate": 0.00017156583629893238,
+      "loss": 2.8685,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7151111111111111,
+      "grad_norm": 1.6383613348007202,
+      "learning_rate": 0.00017154804270462634,
+      "loss": 2.9057,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7155555555555555,
+      "grad_norm": 1.3844068050384521,
+      "learning_rate": 0.0001715302491103203,
+      "loss": 2.5375,
+      "step": 1610
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 1.1408178806304932,
+      "learning_rate": 0.00017151245551601425,
+      "loss": 2.5761,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7164444444444444,
+      "grad_norm": 1.1641733646392822,
+      "learning_rate": 0.0001714946619217082,
+      "loss": 2.6517,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7168888888888889,
+      "grad_norm": 1.1055686473846436,
+      "learning_rate": 0.00017147686832740214,
+      "loss": 2.1878,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7173333333333334,
+      "grad_norm": 1.1979750394821167,
+      "learning_rate": 0.0001714590747330961,
+      "loss": 2.4771,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7177777777777777,
+      "grad_norm": 1.271674633026123,
+      "learning_rate": 0.00017144128113879002,
+      "loss": 2.4917,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7182222222222222,
+      "grad_norm": 1.3329592943191528,
+      "learning_rate": 0.00017142348754448398,
+      "loss": 2.5694,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7186666666666667,
+      "grad_norm": 1.4913185834884644,
+      "learning_rate": 0.00017140569395017794,
+      "loss": 2.8739,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7191111111111111,
+      "grad_norm": 1.3278918266296387,
+      "learning_rate": 0.0001713879003558719,
+      "loss": 2.6695,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7195555555555555,
+      "grad_norm": 1.5261479616165161,
+      "learning_rate": 0.00017137010676156585,
+      "loss": 1.4149,
+      "step": 1619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4373037815093994,
+      "learning_rate": 0.00017135231316725978,
+      "loss": 2.8322,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7204444444444444,
+      "grad_norm": 1.2355538606643677,
+      "learning_rate": 0.00017133451957295374,
+      "loss": 2.1176,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7208888888888889,
+      "grad_norm": 1.29166579246521,
+      "learning_rate": 0.0001713167259786477,
+      "loss": 2.5198,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7213333333333334,
+      "grad_norm": 1.6880923509597778,
+      "learning_rate": 0.00017129893238434165,
+      "loss": 2.9581,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7217777777777777,
+      "grad_norm": 1.5091137886047363,
+      "learning_rate": 0.0001712811387900356,
+      "loss": 2.2347,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7222222222222222,
+      "grad_norm": 1.5347869396209717,
+      "learning_rate": 0.00017126334519572956,
+      "loss": 2.6514,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7226666666666667,
+      "grad_norm": 1.6462808847427368,
+      "learning_rate": 0.0001712455516014235,
+      "loss": 2.9572,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7231111111111111,
+      "grad_norm": 1.5206083059310913,
+      "learning_rate": 0.00017122775800711745,
+      "loss": 2.6225,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7235555555555555,
+      "grad_norm": 1.6654325723648071,
+      "learning_rate": 0.00017120996441281138,
+      "loss": 2.4817,
+      "step": 1628
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 1.6981536149978638,
+      "learning_rate": 0.00017119217081850533,
+      "loss": 2.5816,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7244444444444444,
+      "grad_norm": 1.5140714645385742,
+      "learning_rate": 0.0001711743772241993,
+      "loss": 2.6285,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7248888888888889,
+      "grad_norm": 1.4249933958053589,
+      "learning_rate": 0.00017115658362989325,
+      "loss": 1.8135,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7253333333333334,
+      "grad_norm": 1.72703218460083,
+      "learning_rate": 0.0001711387900355872,
+      "loss": 2.7962,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7257777777777777,
+      "grad_norm": 1.43552565574646,
+      "learning_rate": 0.00017112099644128113,
+      "loss": 2.0849,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7262222222222222,
+      "grad_norm": 1.728300929069519,
+      "learning_rate": 0.0001711032028469751,
+      "loss": 3.2766,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7266666666666667,
+      "grad_norm": 1.7892253398895264,
+      "learning_rate": 0.00017108540925266905,
+      "loss": 2.9754,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7271111111111112,
+      "grad_norm": 1.731090784072876,
+      "learning_rate": 0.000171067615658363,
+      "loss": 2.6769,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7275555555555555,
+      "grad_norm": 1.4827601909637451,
+      "learning_rate": 0.00017104982206405696,
+      "loss": 2.1877,
+      "step": 1637
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 1.4949201345443726,
+      "learning_rate": 0.00017103202846975091,
+      "loss": 2.7004,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7284444444444444,
+      "grad_norm": 1.5235779285430908,
+      "learning_rate": 0.00017101423487544484,
+      "loss": 2.451,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7288888888888889,
+      "grad_norm": 1.7077027559280396,
+      "learning_rate": 0.0001709964412811388,
+      "loss": 2.4066,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7293333333333333,
+      "grad_norm": 1.3838727474212646,
+      "learning_rate": 0.00017097864768683273,
+      "loss": 2.0036,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7297777777777777,
+      "grad_norm": 1.5762803554534912,
+      "learning_rate": 0.0001709608540925267,
+      "loss": 2.3423,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7302222222222222,
+      "grad_norm": 1.5735541582107544,
+      "learning_rate": 0.00017094306049822064,
+      "loss": 2.6597,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7306666666666667,
+      "grad_norm": 2.120513916015625,
+      "learning_rate": 0.0001709252669039146,
+      "loss": 3.0132,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7311111111111112,
+      "grad_norm": 1.8525890111923218,
+      "learning_rate": 0.00017090747330960856,
+      "loss": 2.3857,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7315555555555555,
+      "grad_norm": 1.7836278676986694,
+      "learning_rate": 0.00017088967971530249,
+      "loss": 2.9966,
+      "step": 1646
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 2.076381206512451,
+      "learning_rate": 0.00017087188612099644,
+      "loss": 3.142,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7324444444444445,
+      "grad_norm": 2.3997738361358643,
+      "learning_rate": 0.0001708540925266904,
+      "loss": 3.5624,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7328888888888889,
+      "grad_norm": 2.19384503364563,
+      "learning_rate": 0.00017083629893238435,
+      "loss": 2.0362,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7333333333333333,
+      "grad_norm": 2.8410866260528564,
+      "learning_rate": 0.0001708185053380783,
+      "loss": 2.3181,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7337777777777778,
+      "grad_norm": 1.1997778415679932,
+      "learning_rate": 0.00017080071174377227,
+      "loss": 1.5159,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7342222222222222,
+      "grad_norm": 1.0796202421188354,
+      "learning_rate": 0.0001707829181494662,
+      "loss": 3.136,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7346666666666667,
+      "grad_norm": 1.2189334630966187,
+      "learning_rate": 0.00017076512455516015,
+      "loss": 2.7014,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7351111111111112,
+      "grad_norm": 1.2097785472869873,
+      "learning_rate": 0.00017074733096085408,
+      "loss": 3.3386,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7355555555555555,
+      "grad_norm": 1.1241984367370605,
+      "learning_rate": 0.00017072953736654804,
+      "loss": 2.809,
+      "step": 1655
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 1.315993309020996,
+      "learning_rate": 0.000170711743772242,
+      "loss": 2.1582,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7364444444444445,
+      "grad_norm": 1.2245084047317505,
+      "learning_rate": 0.00017069395017793595,
+      "loss": 2.5975,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7368888888888889,
+      "grad_norm": 1.542858362197876,
+      "learning_rate": 0.0001706761565836299,
+      "loss": 1.5061,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7373333333333333,
+      "grad_norm": 1.0316481590270996,
+      "learning_rate": 0.00017065836298932384,
+      "loss": 1.3895,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7377777777777778,
+      "grad_norm": 1.2059721946716309,
+      "learning_rate": 0.0001706405693950178,
+      "loss": 2.5289,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7382222222222222,
+      "grad_norm": 1.303240418434143,
+      "learning_rate": 0.00017062277580071175,
+      "loss": 2.6186,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7386666666666667,
+      "grad_norm": 1.26139235496521,
+      "learning_rate": 0.0001706049822064057,
+      "loss": 2.7619,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7391111111111112,
+      "grad_norm": 1.4750614166259766,
+      "learning_rate": 0.00017058718861209966,
+      "loss": 2.4993,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7395555555555555,
+      "grad_norm": 1.3910586833953857,
+      "learning_rate": 0.00017056939501779362,
+      "loss": 2.2823,
+      "step": 1664
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4160467386245728,
+      "learning_rate": 0.00017055160142348755,
+      "loss": 2.102,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7404444444444445,
+      "grad_norm": 1.1895157098770142,
+      "learning_rate": 0.0001705338078291815,
+      "loss": 1.9158,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7408888888888889,
+      "grad_norm": 1.5959806442260742,
+      "learning_rate": 0.00017051601423487544,
+      "loss": 2.5473,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7413333333333333,
+      "grad_norm": 1.5768413543701172,
+      "learning_rate": 0.0001704982206405694,
+      "loss": 1.2562,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7417777777777778,
+      "grad_norm": 1.4953051805496216,
+      "learning_rate": 0.00017048042704626335,
+      "loss": 2.5303,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7422222222222222,
+      "grad_norm": 1.3557592630386353,
+      "learning_rate": 0.0001704626334519573,
+      "loss": 2.2733,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7426666666666667,
+      "grad_norm": 1.5126179456710815,
+      "learning_rate": 0.00017044483985765126,
+      "loss": 2.7248,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7431111111111111,
+      "grad_norm": 1.3499473333358765,
+      "learning_rate": 0.0001704270462633452,
+      "loss": 2.5348,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7435555555555555,
+      "grad_norm": 1.2511281967163086,
+      "learning_rate": 0.00017040925266903915,
+      "loss": 2.2185,
+      "step": 1673
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 1.4628006219863892,
+      "learning_rate": 0.0001703914590747331,
+      "loss": 2.3975,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7444444444444445,
+      "grad_norm": 1.4203425645828247,
+      "learning_rate": 0.00017037366548042706,
+      "loss": 2.1031,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7448888888888889,
+      "grad_norm": 1.3346225023269653,
+      "learning_rate": 0.00017035587188612102,
+      "loss": 1.9704,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7453333333333333,
+      "grad_norm": 1.3518871068954468,
+      "learning_rate": 0.00017033807829181497,
+      "loss": 2.6569,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7457777777777778,
+      "grad_norm": 1.5000810623168945,
+      "learning_rate": 0.0001703202846975089,
+      "loss": 2.1933,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7462222222222222,
+      "grad_norm": 1.6626880168914795,
+      "learning_rate": 0.00017030249110320286,
+      "loss": 2.166,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "grad_norm": 1.3020575046539307,
+      "learning_rate": 0.0001702846975088968,
+      "loss": 2.4969,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7471111111111111,
+      "grad_norm": 1.4085545539855957,
+      "learning_rate": 0.00017026690391459075,
+      "loss": 2.4218,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7475555555555555,
+      "grad_norm": 1.861728310585022,
+      "learning_rate": 0.0001702491103202847,
+      "loss": 2.2863,
+      "step": 1682
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 1.688712239265442,
+      "learning_rate": 0.00017023131672597866,
+      "loss": 2.773,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7484444444444445,
+      "grad_norm": 1.4581354856491089,
+      "learning_rate": 0.00017021352313167262,
+      "loss": 2.4764,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7488888888888889,
+      "grad_norm": 1.8125518560409546,
+      "learning_rate": 0.00017019572953736654,
+      "loss": 2.1238,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7493333333333333,
+      "grad_norm": 1.5851460695266724,
+      "learning_rate": 0.0001701779359430605,
+      "loss": 2.2135,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7497777777777778,
+      "grad_norm": 1.852522611618042,
+      "learning_rate": 0.00017016014234875446,
+      "loss": 2.3781,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7502222222222222,
+      "grad_norm": 1.500272512435913,
+      "learning_rate": 0.00017014234875444841,
+      "loss": 2.2147,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7506666666666667,
+      "grad_norm": 1.4343912601470947,
+      "learning_rate": 0.00017012455516014237,
+      "loss": 2.3782,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7511111111111111,
+      "grad_norm": 1.8600046634674072,
+      "learning_rate": 0.00017010676156583633,
+      "loss": 2.7655,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7515555555555555,
+      "grad_norm": 1.5437164306640625,
+      "learning_rate": 0.00017008896797153026,
+      "loss": 2.3394,
+      "step": 1691
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 1.9030184745788574,
+      "learning_rate": 0.0001700711743772242,
+      "loss": 2.7159,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7524444444444445,
+      "grad_norm": 2.081378698348999,
+      "learning_rate": 0.00017005338078291814,
+      "loss": 2.5658,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7528888888888889,
+      "grad_norm": 1.6532082557678223,
+      "learning_rate": 0.0001700355871886121,
+      "loss": 2.3354,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7533333333333333,
+      "grad_norm": 2.3726096153259277,
+      "learning_rate": 0.00017001779359430606,
+      "loss": 3.1188,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7537777777777778,
+      "grad_norm": 2.014913558959961,
+      "learning_rate": 0.00017,
+      "loss": 3.0022,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7542222222222222,
+      "grad_norm": 2.250953197479248,
+      "learning_rate": 0.00016998220640569397,
+      "loss": 2.8569,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7546666666666667,
+      "grad_norm": 1.7740085124969482,
+      "learning_rate": 0.0001699644128113879,
+      "loss": 2.3986,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7551111111111111,
+      "grad_norm": 1.9951434135437012,
+      "learning_rate": 0.00016994661921708185,
+      "loss": 1.5462,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7555555555555555,
+      "grad_norm": 2.3089957237243652,
+      "learning_rate": 0.0001699288256227758,
+      "loss": 2.038,
+      "step": 1700
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 0.9273460507392883,
+      "learning_rate": 0.00016991103202846977,
+      "loss": 2.2534,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7564444444444445,
+      "grad_norm": 1.0741509199142456,
+      "learning_rate": 0.00016989323843416372,
+      "loss": 2.6395,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7568888888888889,
+      "grad_norm": 1.2935476303100586,
+      "learning_rate": 0.00016987544483985765,
+      "loss": 2.7484,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7573333333333333,
+      "grad_norm": 1.2503875494003296,
+      "learning_rate": 0.0001698576512455516,
+      "loss": 2.4681,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7577777777777778,
+      "grad_norm": 1.1031012535095215,
+      "learning_rate": 0.00016983985765124557,
+      "loss": 2.5408,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7582222222222222,
+      "grad_norm": 1.1890628337860107,
+      "learning_rate": 0.0001698220640569395,
+      "loss": 2.9032,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7586666666666667,
+      "grad_norm": 1.1588826179504395,
+      "learning_rate": 0.00016980427046263345,
+      "loss": 2.3581,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7591111111111111,
+      "grad_norm": 1.2451859712600708,
+      "learning_rate": 0.0001697864768683274,
+      "loss": 2.4299,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7595555555555555,
+      "grad_norm": 1.4853380918502808,
+      "learning_rate": 0.00016976868327402137,
+      "loss": 3.0938,
+      "step": 1709
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.686055302619934,
+      "learning_rate": 0.0001697508896797153,
+      "loss": 2.3023,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7604444444444445,
+      "grad_norm": 1.3711706399917603,
+      "learning_rate": 0.00016973309608540925,
+      "loss": 2.6126,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7608888888888888,
+      "grad_norm": 1.3711973428726196,
+      "learning_rate": 0.0001697153024911032,
+      "loss": 2.634,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7613333333333333,
+      "grad_norm": 1.236276388168335,
+      "learning_rate": 0.00016969750889679716,
+      "loss": 2.6303,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7617777777777778,
+      "grad_norm": 1.2677000761032104,
+      "learning_rate": 0.00016967971530249112,
+      "loss": 2.646,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7622222222222222,
+      "grad_norm": 1.2618008852005005,
+      "learning_rate": 0.00016966192170818508,
+      "loss": 2.3963,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7626666666666667,
+      "grad_norm": 1.3823282718658447,
+      "learning_rate": 0.000169644128113879,
+      "loss": 2.6979,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7631111111111111,
+      "grad_norm": 1.3976502418518066,
+      "learning_rate": 0.00016962633451957296,
+      "loss": 1.525,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7635555555555555,
+      "grad_norm": 1.343619465827942,
+      "learning_rate": 0.0001696085409252669,
+      "loss": 2.9,
+      "step": 1718
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 1.1925876140594482,
+      "learning_rate": 0.00016959074733096085,
+      "loss": 1.9166,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7644444444444445,
+      "grad_norm": 1.782758355140686,
+      "learning_rate": 0.0001695729537366548,
+      "loss": 2.5998,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7648888888888888,
+      "grad_norm": 1.5515620708465576,
+      "learning_rate": 0.00016955516014234876,
+      "loss": 2.1778,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7653333333333333,
+      "grad_norm": 1.3295077085494995,
+      "learning_rate": 0.00016953736654804272,
+      "loss": 2.0455,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7657777777777778,
+      "grad_norm": 1.6880308389663696,
+      "learning_rate": 0.00016951957295373665,
+      "loss": 2.6615,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7662222222222222,
+      "grad_norm": 1.8657255172729492,
+      "learning_rate": 0.0001695017793594306,
+      "loss": 2.5311,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7666666666666667,
+      "grad_norm": 1.4304696321487427,
+      "learning_rate": 0.00016948398576512456,
+      "loss": 1.9907,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7671111111111111,
+      "grad_norm": 1.3921934366226196,
+      "learning_rate": 0.00016946619217081852,
+      "loss": 2.5942,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7675555555555555,
+      "grad_norm": 1.1713439226150513,
+      "learning_rate": 0.00016944839857651247,
+      "loss": 1.329,
+      "step": 1727
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 1.447230577468872,
+      "learning_rate": 0.00016943060498220643,
+      "loss": 2.2461,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7684444444444445,
+      "grad_norm": 1.5526431798934937,
+      "learning_rate": 0.00016941281138790036,
+      "loss": 2.5692,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7688888888888888,
+      "grad_norm": 1.864875078201294,
+      "learning_rate": 0.00016939501779359432,
+      "loss": 2.3305,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7693333333333333,
+      "grad_norm": 1.5413137674331665,
+      "learning_rate": 0.00016937722419928825,
+      "loss": 2.4763,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7697777777777778,
+      "grad_norm": 1.5602953433990479,
+      "learning_rate": 0.0001693594306049822,
+      "loss": 2.3397,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7702222222222223,
+      "grad_norm": 1.5029823780059814,
+      "learning_rate": 0.00016934163701067616,
+      "loss": 2.5413,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7706666666666667,
+      "grad_norm": 1.9721778631210327,
+      "learning_rate": 0.00016932384341637012,
+      "loss": 2.3882,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7711111111111111,
+      "grad_norm": 1.5762044191360474,
+      "learning_rate": 0.00016930604982206407,
+      "loss": 2.7452,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7715555555555556,
+      "grad_norm": 1.6823639869689941,
+      "learning_rate": 0.000169288256227758,
+      "loss": 2.706,
+      "step": 1736
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 2.066340446472168,
+      "learning_rate": 0.00016927046263345196,
+      "loss": 2.8768,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7724444444444445,
+      "grad_norm": 1.7574570178985596,
+      "learning_rate": 0.00016925266903914591,
+      "loss": 2.9213,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7728888888888888,
+      "grad_norm": 1.5815603733062744,
+      "learning_rate": 0.00016923487544483987,
+      "loss": 2.7557,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7733333333333333,
+      "grad_norm": 2.00929594039917,
+      "learning_rate": 0.00016921708185053383,
+      "loss": 2.828,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7737777777777778,
+      "grad_norm": 1.5500198602676392,
+      "learning_rate": 0.00016919928825622778,
+      "loss": 2.8645,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7742222222222223,
+      "grad_norm": 1.5847936868667603,
+      "learning_rate": 0.0001691814946619217,
+      "loss": 2.4433,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7746666666666666,
+      "grad_norm": 1.5422377586364746,
+      "learning_rate": 0.00016916370106761567,
+      "loss": 2.7777,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7751111111111111,
+      "grad_norm": 1.4462459087371826,
+      "learning_rate": 0.0001691459074733096,
+      "loss": 2.3321,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7755555555555556,
+      "grad_norm": 2.3052868843078613,
+      "learning_rate": 0.00016912811387900356,
+      "loss": 2.3839,
+      "step": 1745
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 1.589308261871338,
+      "learning_rate": 0.0001691103202846975,
+      "loss": 2.2261,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7764444444444445,
+      "grad_norm": 1.5655573606491089,
+      "learning_rate": 0.00016909252669039147,
+      "loss": 2.3206,
+      "step": 1747
+    },
+    {
+      "epoch": 0.7768888888888889,
+      "grad_norm": 1.4442392587661743,
+      "learning_rate": 0.00016907473309608542,
+      "loss": 1.8804,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7773333333333333,
+      "grad_norm": 2.0022120475769043,
+      "learning_rate": 0.00016905693950177935,
+      "loss": 3.0496,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 2.1214585304260254,
+      "learning_rate": 0.0001690391459074733,
+      "loss": 3.1286,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7782222222222223,
+      "grad_norm": 1.1207834482192993,
+      "learning_rate": 0.00016902135231316727,
+      "loss": 1.5612,
+      "step": 1751
+    },
+    {
+      "epoch": 0.7786666666666666,
+      "grad_norm": 1.0428500175476074,
+      "learning_rate": 0.00016900355871886122,
+      "loss": 2.3335,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7791111111111111,
+      "grad_norm": 1.0575370788574219,
+      "learning_rate": 0.00016898576512455518,
+      "loss": 1.4655,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7795555555555556,
+      "grad_norm": 1.2163389921188354,
+      "learning_rate": 0.00016896797153024914,
+      "loss": 2.3588,
+      "step": 1754
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.481124997138977,
+      "learning_rate": 0.00016895017793594307,
+      "loss": 2.8144,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7804444444444445,
+      "grad_norm": 1.3335241079330444,
+      "learning_rate": 0.00016893238434163702,
+      "loss": 3.0533,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7808888888888889,
+      "grad_norm": 1.2501187324523926,
+      "learning_rate": 0.00016891459074733095,
+      "loss": 2.7067,
+      "step": 1757
+    },
+    {
+      "epoch": 0.7813333333333333,
+      "grad_norm": 1.2316926717758179,
+      "learning_rate": 0.0001688967971530249,
+      "loss": 3.0781,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7817777777777778,
+      "grad_norm": 1.4136369228363037,
+      "learning_rate": 0.00016887900355871886,
+      "loss": 2.4057,
+      "step": 1759
+    },
+    {
+      "epoch": 0.7822222222222223,
+      "grad_norm": 1.2159130573272705,
+      "learning_rate": 0.00016886120996441282,
+      "loss": 2.3195,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7826666666666666,
+      "grad_norm": 1.3189160823822021,
+      "learning_rate": 0.00016884341637010678,
+      "loss": 2.6697,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7831111111111111,
+      "grad_norm": 1.2556674480438232,
+      "learning_rate": 0.0001688256227758007,
+      "loss": 2.4658,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7835555555555556,
+      "grad_norm": 1.4505153894424438,
+      "learning_rate": 0.00016880782918149466,
+      "loss": 2.4346,
+      "step": 1763
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 1.3776673078536987,
+      "learning_rate": 0.00016879003558718862,
+      "loss": 2.2668,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7844444444444445,
+      "grad_norm": 1.7608091831207275,
+      "learning_rate": 0.00016877224199288258,
+      "loss": 1.6074,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7848888888888889,
+      "grad_norm": 1.6191914081573486,
+      "learning_rate": 0.00016875444839857653,
+      "loss": 2.3601,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7853333333333333,
+      "grad_norm": 1.4360511302947998,
+      "learning_rate": 0.0001687366548042705,
+      "loss": 2.5914,
+      "step": 1767
+    },
+    {
+      "epoch": 0.7857777777777778,
+      "grad_norm": 1.3812462091445923,
+      "learning_rate": 0.00016871886120996442,
+      "loss": 2.5718,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7862222222222223,
+      "grad_norm": 1.522873878479004,
+      "learning_rate": 0.00016870106761565838,
+      "loss": 2.7388,
+      "step": 1769
+    },
+    {
+      "epoch": 0.7866666666666666,
+      "grad_norm": 1.7307029962539673,
+      "learning_rate": 0.0001686832740213523,
+      "loss": 2.0664,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7871111111111111,
+      "grad_norm": 1.383586049079895,
+      "learning_rate": 0.00016866548042704626,
+      "loss": 2.7238,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7875555555555556,
+      "grad_norm": 1.6344101428985596,
+      "learning_rate": 0.00016864768683274022,
+      "loss": 3.0245,
+      "step": 1772
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 1.4926892518997192,
+      "learning_rate": 0.00016862989323843417,
+      "loss": 2.4893,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7884444444444444,
+      "grad_norm": 1.7372292280197144,
+      "learning_rate": 0.00016861209964412813,
+      "loss": 2.9537,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7888888888888889,
+      "grad_norm": 1.3594554662704468,
+      "learning_rate": 0.00016859430604982206,
+      "loss": 1.9789,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7893333333333333,
+      "grad_norm": 1.3864822387695312,
+      "learning_rate": 0.00016857651245551602,
+      "loss": 2.585,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7897777777777778,
+      "grad_norm": 1.5721726417541504,
+      "learning_rate": 0.00016855871886120997,
+      "loss": 2.8356,
+      "step": 1777
+    },
+    {
+      "epoch": 0.7902222222222223,
+      "grad_norm": 1.3060152530670166,
+      "learning_rate": 0.00016854092526690393,
+      "loss": 2.2237,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7906666666666666,
+      "grad_norm": 1.349345326423645,
+      "learning_rate": 0.0001685231316725979,
+      "loss": 2.3712,
+      "step": 1779
+    },
+    {
+      "epoch": 0.7911111111111111,
+      "grad_norm": 1.4352922439575195,
+      "learning_rate": 0.00016850533807829184,
+      "loss": 1.6791,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7915555555555556,
+      "grad_norm": 1.5430026054382324,
+      "learning_rate": 0.00016848754448398577,
+      "loss": 2.7579,
+      "step": 1781
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 1.7820968627929688,
+      "learning_rate": 0.00016846975088967973,
+      "loss": 3.1443,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7924444444444444,
+      "grad_norm": 1.3988341093063354,
+      "learning_rate": 0.00016845195729537366,
+      "loss": 1.8977,
+      "step": 1783
+    },
+    {
+      "epoch": 0.7928888888888889,
+      "grad_norm": 1.705902338027954,
+      "learning_rate": 0.00016843416370106761,
+      "loss": 2.795,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7933333333333333,
+      "grad_norm": 1.6256381273269653,
+      "learning_rate": 0.00016841637010676157,
+      "loss": 2.6093,
+      "step": 1785
+    },
+    {
+      "epoch": 0.7937777777777778,
+      "grad_norm": 1.6664106845855713,
+      "learning_rate": 0.00016839857651245553,
+      "loss": 2.5954,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7942222222222223,
+      "grad_norm": 1.5220837593078613,
+      "learning_rate": 0.00016838078291814948,
+      "loss": 2.4056,
+      "step": 1787
+    },
+    {
+      "epoch": 0.7946666666666666,
+      "grad_norm": 2.4615044593811035,
+      "learning_rate": 0.00016836298932384341,
+      "loss": 3.1873,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7951111111111111,
+      "grad_norm": 1.54017174243927,
+      "learning_rate": 0.00016834519572953737,
+      "loss": 2.5285,
+      "step": 1789
+    },
+    {
+      "epoch": 0.7955555555555556,
+      "grad_norm": 1.5758004188537598,
+      "learning_rate": 0.00016832740213523133,
+      "loss": 2.4812,
+      "step": 1790
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 2.027078866958618,
+      "learning_rate": 0.00016830960854092528,
+      "loss": 2.5266,
+      "step": 1791
+    },
+    {
+      "epoch": 0.7964444444444444,
+      "grad_norm": 1.9708378314971924,
+      "learning_rate": 0.00016829181494661924,
+      "loss": 2.7944,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7968888888888889,
+      "grad_norm": 1.4897676706314087,
+      "learning_rate": 0.00016827402135231317,
+      "loss": 2.1396,
+      "step": 1793
+    },
+    {
+      "epoch": 0.7973333333333333,
+      "grad_norm": 1.5909985303878784,
+      "learning_rate": 0.00016825622775800713,
+      "loss": 2.1982,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7977777777777778,
+      "grad_norm": 1.780980110168457,
+      "learning_rate": 0.00016823843416370108,
+      "loss": 2.836,
+      "step": 1795
+    },
+    {
+      "epoch": 0.7982222222222223,
+      "grad_norm": 1.9581319093704224,
+      "learning_rate": 0.000168220640569395,
+      "loss": 3.0858,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7986666666666666,
+      "grad_norm": 2.094820737838745,
+      "learning_rate": 0.00016820284697508897,
+      "loss": 3.2028,
+      "step": 1797
+    },
+    {
+      "epoch": 0.7991111111111111,
+      "grad_norm": 2.108332872390747,
+      "learning_rate": 0.00016818505338078292,
+      "loss": 2.1878,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7995555555555556,
+      "grad_norm": 1.941762089729309,
+      "learning_rate": 0.00016816725978647688,
+      "loss": 2.6336,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6256611347198486,
+      "learning_rate": 0.0001681494661921708,
+      "loss": 1.605,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8004444444444444,
+      "grad_norm": 0.9606013894081116,
+      "learning_rate": 0.00016813167259786477,
+      "loss": 2.5109,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8008888888888889,
+      "grad_norm": 1.1120461225509644,
+      "learning_rate": 0.00016811387900355872,
+      "loss": 3.0039,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8013333333333333,
+      "grad_norm": 0.9408321976661682,
+      "learning_rate": 0.00016809608540925268,
+      "loss": 2.1882,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8017777777777778,
+      "grad_norm": 1.210302472114563,
+      "learning_rate": 0.00016807829181494664,
+      "loss": 2.4686,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8022222222222222,
+      "grad_norm": 1.1656526327133179,
+      "learning_rate": 0.0001680604982206406,
+      "loss": 2.546,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8026666666666666,
+      "grad_norm": 1.2500494718551636,
+      "learning_rate": 0.00016804270462633452,
+      "loss": 2.7287,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8031111111111111,
+      "grad_norm": 1.1803468465805054,
+      "learning_rate": 0.00016802491103202848,
+      "loss": 2.7445,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8035555555555556,
+      "grad_norm": 1.7277123928070068,
+      "learning_rate": 0.00016800711743772244,
+      "loss": 1.3246,
+      "step": 1808
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 1.6080886125564575,
+      "learning_rate": 0.00016798932384341636,
+      "loss": 2.639,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8044444444444444,
+      "grad_norm": 1.1558185815811157,
+      "learning_rate": 0.00016797153024911032,
+      "loss": 2.5563,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8048888888888889,
+      "grad_norm": 1.3278673887252808,
+      "learning_rate": 0.00016795373665480428,
+      "loss": 2.1776,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8053333333333333,
+      "grad_norm": 1.2279390096664429,
+      "learning_rate": 0.00016793594306049823,
+      "loss": 2.5615,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8057777777777778,
+      "grad_norm": 1.3550021648406982,
+      "learning_rate": 0.00016791814946619216,
+      "loss": 2.432,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8062222222222222,
+      "grad_norm": 1.2661508321762085,
+      "learning_rate": 0.00016790035587188612,
+      "loss": 2.4724,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8066666666666666,
+      "grad_norm": 1.5162663459777832,
+      "learning_rate": 0.00016788256227758008,
+      "loss": 2.7064,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8071111111111111,
+      "grad_norm": 1.526525616645813,
+      "learning_rate": 0.00016786476868327403,
+      "loss": 2.7493,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8075555555555556,
+      "grad_norm": 1.349353551864624,
+      "learning_rate": 0.000167846975088968,
+      "loss": 2.175,
+      "step": 1817
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 1.518172025680542,
+      "learning_rate": 0.00016782918149466195,
+      "loss": 2.9974,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8084444444444444,
+      "grad_norm": 1.3705731630325317,
+      "learning_rate": 0.00016781138790035588,
+      "loss": 2.691,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8088888888888889,
+      "grad_norm": 1.4127305746078491,
+      "learning_rate": 0.00016779359430604983,
+      "loss": 2.6794,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8093333333333333,
+      "grad_norm": 1.3548202514648438,
+      "learning_rate": 0.0001677758007117438,
+      "loss": 2.3797,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8097777777777778,
+      "grad_norm": 2.2642502784729004,
+      "learning_rate": 0.00016775800711743772,
+      "loss": 2.1909,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8102222222222222,
+      "grad_norm": 1.6806228160858154,
+      "learning_rate": 0.00016774021352313167,
+      "loss": 2.8928,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8106666666666666,
+      "grad_norm": 1.296937346458435,
+      "learning_rate": 0.00016772241992882563,
+      "loss": 2.0109,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8111111111111111,
+      "grad_norm": 1.387312650680542,
+      "learning_rate": 0.0001677046263345196,
+      "loss": 2.2319,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8115555555555556,
+      "grad_norm": 1.6626933813095093,
+      "learning_rate": 0.00016768683274021352,
+      "loss": 2.6542,
+      "step": 1826
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 1.658470869064331,
+      "learning_rate": 0.00016766903914590747,
+      "loss": 2.3784,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8124444444444444,
+      "grad_norm": 1.5517417192459106,
+      "learning_rate": 0.00016765124555160143,
+      "loss": 2.6614,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8128888888888889,
+      "grad_norm": 1.6006346940994263,
+      "learning_rate": 0.00016763345195729539,
+      "loss": 3.0233,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8133333333333334,
+      "grad_norm": 1.503570795059204,
+      "learning_rate": 0.00016761565836298934,
+      "loss": 2.2458,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8137777777777778,
+      "grad_norm": 1.9097468852996826,
+      "learning_rate": 0.0001675978647686833,
+      "loss": 2.6981,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8142222222222222,
+      "grad_norm": 1.504949927330017,
+      "learning_rate": 0.00016758007117437723,
+      "loss": 2.8705,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8146666666666667,
+      "grad_norm": 1.7333391904830933,
+      "learning_rate": 0.00016756227758007119,
+      "loss": 2.3296,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8151111111111111,
+      "grad_norm": 4.087153434753418,
+      "learning_rate": 0.00016754448398576511,
+      "loss": 0.2551,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8155555555555556,
+      "grad_norm": 1.6669485569000244,
+      "learning_rate": 0.00016752669039145907,
+      "loss": 2.2058,
+      "step": 1835
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 1.4292755126953125,
+      "learning_rate": 0.00016750889679715303,
+      "loss": 2.2523,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8164444444444444,
+      "grad_norm": 1.538756251335144,
+      "learning_rate": 0.00016749110320284698,
+      "loss": 2.3624,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8168888888888889,
+      "grad_norm": 1.5922455787658691,
+      "learning_rate": 0.00016747330960854094,
+      "loss": 2.6137,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8173333333333334,
+      "grad_norm": 1.5305728912353516,
+      "learning_rate": 0.00016745551601423487,
+      "loss": 2.1719,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8177777777777778,
+      "grad_norm": 2.139403820037842,
+      "learning_rate": 0.00016743772241992883,
+      "loss": 3.0687,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8182222222222222,
+      "grad_norm": 1.7138938903808594,
+      "learning_rate": 0.00016741992882562278,
+      "loss": 2.684,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8186666666666667,
+      "grad_norm": 2.1638147830963135,
+      "learning_rate": 0.00016740213523131674,
+      "loss": 2.8127,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8191111111111111,
+      "grad_norm": 2.317457914352417,
+      "learning_rate": 0.0001673843416370107,
+      "loss": 2.7219,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8195555555555556,
+      "grad_norm": 1.7413108348846436,
+      "learning_rate": 0.00016736654804270465,
+      "loss": 2.6101,
+      "step": 1844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0976204872131348,
+      "learning_rate": 0.00016734875444839858,
+      "loss": 2.5319,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8204444444444444,
+      "grad_norm": 2.517296075820923,
+      "learning_rate": 0.00016733096085409254,
+      "loss": 3.3618,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8208888888888889,
+      "grad_norm": 2.0704150199890137,
+      "learning_rate": 0.00016731316725978647,
+      "loss": 2.4757,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8213333333333334,
+      "grad_norm": 2.702256917953491,
+      "learning_rate": 0.00016729537366548042,
+      "loss": 0.1476,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8217777777777778,
+      "grad_norm": 1.8367481231689453,
+      "learning_rate": 0.00016727758007117438,
+      "loss": 1.8312,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8222222222222222,
+      "grad_norm": 1.943381428718567,
+      "learning_rate": 0.00016725978647686834,
+      "loss": 0.9415,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8226666666666667,
+      "grad_norm": 1.0285029411315918,
+      "learning_rate": 0.0001672419928825623,
+      "loss": 2.5791,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8231111111111111,
+      "grad_norm": 1.0174617767333984,
+      "learning_rate": 0.00016722419928825622,
+      "loss": 2.4917,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8235555555555556,
+      "grad_norm": 1.1286165714263916,
+      "learning_rate": 0.00016720640569395018,
+      "loss": 2.7286,
+      "step": 1853
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 1.2801088094711304,
+      "learning_rate": 0.00016718861209964414,
+      "loss": 2.2644,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8244444444444444,
+      "grad_norm": 1.3122279644012451,
+      "learning_rate": 0.0001671708185053381,
+      "loss": 1.8628,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8248888888888889,
+      "grad_norm": 1.3034883737564087,
+      "learning_rate": 0.00016715302491103205,
+      "loss": 2.8352,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8253333333333334,
+      "grad_norm": 1.3606019020080566,
+      "learning_rate": 0.000167135231316726,
+      "loss": 3.1646,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8257777777777778,
+      "grad_norm": 1.503804087638855,
+      "learning_rate": 0.00016711743772241993,
+      "loss": 2.3705,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8262222222222222,
+      "grad_norm": 1.2869994640350342,
+      "learning_rate": 0.0001670996441281139,
+      "loss": 2.587,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8266666666666667,
+      "grad_norm": 1.376165747642517,
+      "learning_rate": 0.00016708185053380782,
+      "loss": 2.2356,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8271111111111111,
+      "grad_norm": 1.1255593299865723,
+      "learning_rate": 0.00016706405693950178,
+      "loss": 2.1244,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8275555555555556,
+      "grad_norm": 1.390615701675415,
+      "learning_rate": 0.00016704626334519573,
+      "loss": 2.6707,
+      "step": 1862
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 1.195713996887207,
+      "learning_rate": 0.0001670284697508897,
+      "loss": 2.1111,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8284444444444444,
+      "grad_norm": 1.275231957435608,
+      "learning_rate": 0.00016701067615658365,
+      "loss": 2.8402,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8288888888888889,
+      "grad_norm": 1.294203758239746,
+      "learning_rate": 0.00016699288256227758,
+      "loss": 2.9545,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8293333333333334,
+      "grad_norm": 1.4004563093185425,
+      "learning_rate": 0.00016697508896797153,
+      "loss": 2.9775,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8297777777777777,
+      "grad_norm": 1.3818445205688477,
+      "learning_rate": 0.0001669572953736655,
+      "loss": 2.4705,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8302222222222222,
+      "grad_norm": 1.381339192390442,
+      "learning_rate": 0.00016693950177935945,
+      "loss": 2.0993,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8306666666666667,
+      "grad_norm": 1.2843163013458252,
+      "learning_rate": 0.0001669217081850534,
+      "loss": 2.7269,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8311111111111111,
+      "grad_norm": 1.5089333057403564,
+      "learning_rate": 0.00016690391459074736,
+      "loss": 2.442,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8315555555555556,
+      "grad_norm": 1.5830549001693726,
+      "learning_rate": 0.0001668861209964413,
+      "loss": 2.5014,
+      "step": 1871
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 1.250998616218567,
+      "learning_rate": 0.00016686832740213524,
+      "loss": 2.3065,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8324444444444444,
+      "grad_norm": 1.5048855543136597,
+      "learning_rate": 0.00016685053380782917,
+      "loss": 2.3963,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8328888888888889,
+      "grad_norm": 1.3290457725524902,
+      "learning_rate": 0.00016683274021352313,
+      "loss": 2.3719,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 1.4442025423049927,
+      "learning_rate": 0.0001668149466192171,
+      "loss": 2.5401,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8337777777777777,
+      "grad_norm": 1.5543490648269653,
+      "learning_rate": 0.00016679715302491104,
+      "loss": 2.6803,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8342222222222222,
+      "grad_norm": 1.3237760066986084,
+      "learning_rate": 0.000166779359430605,
+      "loss": 2.4261,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8346666666666667,
+      "grad_norm": 1.4572439193725586,
+      "learning_rate": 0.00016676156583629893,
+      "loss": 2.0907,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8351111111111111,
+      "grad_norm": 1.3316566944122314,
+      "learning_rate": 0.00016674377224199289,
+      "loss": 1.099,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8355555555555556,
+      "grad_norm": 1.640336513519287,
+      "learning_rate": 0.00016672597864768684,
+      "loss": 3.0017,
+      "step": 1880
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 1.6514651775360107,
+      "learning_rate": 0.0001667081850533808,
+      "loss": 2.6943,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8364444444444444,
+      "grad_norm": 1.7922955751419067,
+      "learning_rate": 0.00016669039145907476,
+      "loss": 3.0074,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8368888888888889,
+      "grad_norm": 1.293692708015442,
+      "learning_rate": 0.00016667259786476868,
+      "loss": 2.0004,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8373333333333334,
+      "grad_norm": 1.6649595499038696,
+      "learning_rate": 0.00016665480427046264,
+      "loss": 2.8518,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8377777777777777,
+      "grad_norm": 1.55930757522583,
+      "learning_rate": 0.0001666370106761566,
+      "loss": 2.5283,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8382222222222222,
+      "grad_norm": 1.808152198791504,
+      "learning_rate": 0.00016661921708185053,
+      "loss": 3.0004,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8386666666666667,
+      "grad_norm": 2.0772571563720703,
+      "learning_rate": 0.00016660142348754448,
+      "loss": 2.2274,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8391111111111111,
+      "grad_norm": 1.5994224548339844,
+      "learning_rate": 0.00016658362989323844,
+      "loss": 2.212,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8395555555555556,
+      "grad_norm": 1.7321178913116455,
+      "learning_rate": 0.0001665658362989324,
+      "loss": 2.4917,
+      "step": 1889
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8322235345840454,
+      "learning_rate": 0.00016654804270462633,
+      "loss": 3.1259,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8404444444444444,
+      "grad_norm": 1.6857374906539917,
+      "learning_rate": 0.00016653024911032028,
+      "loss": 2.81,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8408888888888889,
+      "grad_norm": 1.6723679304122925,
+      "learning_rate": 0.00016651245551601424,
+      "loss": 2.4271,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8413333333333334,
+      "grad_norm": 1.2392948865890503,
+      "learning_rate": 0.0001664946619217082,
+      "loss": 1.6478,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8417777777777777,
+      "grad_norm": 1.6006635427474976,
+      "learning_rate": 0.00016647686832740215,
+      "loss": 2.7513,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8422222222222222,
+      "grad_norm": 1.623317003250122,
+      "learning_rate": 0.0001664590747330961,
+      "loss": 2.3341,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8426666666666667,
+      "grad_norm": 1.8542609214782715,
+      "learning_rate": 0.00016644128113879004,
+      "loss": 2.7884,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8431111111111111,
+      "grad_norm": 1.8695321083068848,
+      "learning_rate": 0.000166423487544484,
+      "loss": 2.3178,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8435555555555555,
+      "grad_norm": 1.759669303894043,
+      "learning_rate": 0.00016640569395017795,
+      "loss": 2.8503,
+      "step": 1898
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 2.0336124897003174,
+      "learning_rate": 0.00016638790035587188,
+      "loss": 2.4729,
+      "step": 1899
+    },
+    {
+      "epoch": 0.8444444444444444,
+      "grad_norm": 2.0583572387695312,
+      "learning_rate": 0.00016637010676156584,
+      "loss": 2.7253,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8448888888888889,
+      "grad_norm": 0.9357160329818726,
+      "learning_rate": 0.0001663523131672598,
+      "loss": 1.1275,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8453333333333334,
+      "grad_norm": 1.1076958179473877,
+      "learning_rate": 0.00016633451957295375,
+      "loss": 2.4995,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8457777777777777,
+      "grad_norm": 1.0276774168014526,
+      "learning_rate": 0.00016631672597864768,
+      "loss": 2.427,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8462222222222222,
+      "grad_norm": 1.1579983234405518,
+      "learning_rate": 0.00016629893238434164,
+      "loss": 2.2964,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8466666666666667,
+      "grad_norm": 1.100016474723816,
+      "learning_rate": 0.0001662811387900356,
+      "loss": 2.4451,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8471111111111111,
+      "grad_norm": 1.2550407648086548,
+      "learning_rate": 0.00016626334519572955,
+      "loss": 2.985,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8475555555555555,
+      "grad_norm": 1.1398979425430298,
+      "learning_rate": 0.0001662455516014235,
+      "loss": 2.3818,
+      "step": 1907
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 1.3666727542877197,
+      "learning_rate": 0.00016622775800711746,
+      "loss": 2.3968,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8484444444444444,
+      "grad_norm": 1.4796351194381714,
+      "learning_rate": 0.0001662099644128114,
+      "loss": 1.2558,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8488888888888889,
+      "grad_norm": 1.3779754638671875,
+      "learning_rate": 0.00016619217081850535,
+      "loss": 3.1081,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8493333333333334,
+      "grad_norm": 1.4465447664260864,
+      "learning_rate": 0.0001661743772241993,
+      "loss": 2.4524,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8497777777777777,
+      "grad_norm": 1.3205504417419434,
+      "learning_rate": 0.00016615658362989323,
+      "loss": 2.7317,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8502222222222222,
+      "grad_norm": 1.286799669265747,
+      "learning_rate": 0.0001661387900355872,
+      "loss": 2.2615,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8506666666666667,
+      "grad_norm": 1.2320808172225952,
+      "learning_rate": 0.00016612099644128115,
+      "loss": 2.5005,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8511111111111112,
+      "grad_norm": 1.5686407089233398,
+      "learning_rate": 0.0001661032028469751,
+      "loss": 2.3068,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8515555555555555,
+      "grad_norm": 2.0578267574310303,
+      "learning_rate": 0.00016608540925266903,
+      "loss": 1.2472,
+      "step": 1916
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 1.3643602132797241,
+      "learning_rate": 0.000166067615658363,
+      "loss": 2.6433,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8524444444444444,
+      "grad_norm": 1.3753222227096558,
+      "learning_rate": 0.00016604982206405695,
+      "loss": 2.4551,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8528888888888889,
+      "grad_norm": 1.2894665002822876,
+      "learning_rate": 0.0001660320284697509,
+      "loss": 1.7715,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "grad_norm": 1.381399393081665,
+      "learning_rate": 0.00016601423487544486,
+      "loss": 2.5613,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8537777777777777,
+      "grad_norm": 1.5483062267303467,
+      "learning_rate": 0.00016599644128113881,
+      "loss": 2.9712,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8542222222222222,
+      "grad_norm": 1.481539011001587,
+      "learning_rate": 0.00016597864768683274,
+      "loss": 2.4221,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8546666666666667,
+      "grad_norm": 1.316893458366394,
+      "learning_rate": 0.0001659608540925267,
+      "loss": 2.0611,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8551111111111112,
+      "grad_norm": 1.398386836051941,
+      "learning_rate": 0.00016594306049822066,
+      "loss": 2.1331,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8555555555555555,
+      "grad_norm": 1.521558165550232,
+      "learning_rate": 0.0001659252669039146,
+      "loss": 2.6778,
+      "step": 1925
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 1.4601832628250122,
+      "learning_rate": 0.00016590747330960854,
+      "loss": 2.2369,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8564444444444445,
+      "grad_norm": 1.6191009283065796,
+      "learning_rate": 0.0001658896797153025,
+      "loss": 2.6541,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8568888888888889,
+      "grad_norm": 1.5223315954208374,
+      "learning_rate": 0.00016587188612099646,
+      "loss": 2.3937,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8573333333333333,
+      "grad_norm": 1.2509064674377441,
+      "learning_rate": 0.00016585409252669039,
+      "loss": 1.227,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8577777777777778,
+      "grad_norm": 1.5808355808258057,
+      "learning_rate": 0.00016583629893238434,
+      "loss": 2.3527,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8582222222222222,
+      "grad_norm": 1.5534776449203491,
+      "learning_rate": 0.0001658185053380783,
+      "loss": 2.8214,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8586666666666667,
+      "grad_norm": 1.675430178642273,
+      "learning_rate": 0.00016580071174377225,
+      "loss": 2.6934,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8591111111111112,
+      "grad_norm": 1.638614296913147,
+      "learning_rate": 0.0001657829181494662,
+      "loss": 2.6275,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8595555555555555,
+      "grad_norm": 1.4431530237197876,
+      "learning_rate": 0.00016576512455516017,
+      "loss": 2.4221,
+      "step": 1934
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.532965898513794,
+      "learning_rate": 0.0001657473309608541,
+      "loss": 2.1414,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8604444444444445,
+      "grad_norm": 1.8967840671539307,
+      "learning_rate": 0.00016572953736654805,
+      "loss": 2.2568,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8608888888888889,
+      "grad_norm": 1.6051034927368164,
+      "learning_rate": 0.000165711743772242,
+      "loss": 2.6386,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8613333333333333,
+      "grad_norm": 1.7966537475585938,
+      "learning_rate": 0.00016569395017793594,
+      "loss": 2.1053,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8617777777777778,
+      "grad_norm": 1.6629730463027954,
+      "learning_rate": 0.0001656761565836299,
+      "loss": 2.7147,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8622222222222222,
+      "grad_norm": 1.5632649660110474,
+      "learning_rate": 0.00016565836298932385,
+      "loss": 2.3997,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8626666666666667,
+      "grad_norm": 1.827580451965332,
+      "learning_rate": 0.0001656405693950178,
+      "loss": 2.118,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8631111111111112,
+      "grad_norm": 1.7074384689331055,
+      "learning_rate": 0.00016562277580071174,
+      "loss": 2.1276,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8635555555555555,
+      "grad_norm": 1.6580160856246948,
+      "learning_rate": 0.0001656049822064057,
+      "loss": 2.0692,
+      "step": 1943
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 2.2901294231414795,
+      "learning_rate": 0.00016558718861209965,
+      "loss": 2.8533,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8644444444444445,
+      "grad_norm": 1.9296009540557861,
+      "learning_rate": 0.0001655693950177936,
+      "loss": 2.5001,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8648888888888889,
+      "grad_norm": 1.7774969339370728,
+      "learning_rate": 0.00016555160142348756,
+      "loss": 3.0877,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8653333333333333,
+      "grad_norm": 1.7558300495147705,
+      "learning_rate": 0.00016553380782918152,
+      "loss": 2.1556,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8657777777777778,
+      "grad_norm": 1.870627760887146,
+      "learning_rate": 0.00016551601423487545,
+      "loss": 1.3417,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8662222222222222,
+      "grad_norm": 1.396712064743042,
+      "learning_rate": 0.0001654982206405694,
+      "loss": 1.0483,
+      "step": 1949
+    },
+    {
+      "epoch": 0.8666666666666667,
+      "grad_norm": 1.8367741107940674,
+      "learning_rate": 0.00016548042704626334,
+      "loss": 2.2391,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8671111111111112,
+      "grad_norm": 1.0035446882247925,
+      "learning_rate": 0.0001654626334519573,
+      "loss": 2.7096,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8675555555555555,
+      "grad_norm": 1.2445075511932373,
+      "learning_rate": 0.00016544483985765125,
+      "loss": 2.5628,
+      "step": 1952
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 1.1393455266952515,
+      "learning_rate": 0.0001654270462633452,
+      "loss": 2.8044,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8684444444444445,
+      "grad_norm": 1.042448878288269,
+      "learning_rate": 0.00016540925266903916,
+      "loss": 2.183,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8688888888888889,
+      "grad_norm": 1.0702394247055054,
+      "learning_rate": 0.0001653914590747331,
+      "loss": 2.544,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8693333333333333,
+      "grad_norm": 1.1012481451034546,
+      "learning_rate": 0.00016537366548042705,
+      "loss": 2.4327,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8697777777777778,
+      "grad_norm": 1.0068960189819336,
+      "learning_rate": 0.000165355871886121,
+      "loss": 2.1235,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8702222222222222,
+      "grad_norm": 1.271165370941162,
+      "learning_rate": 0.00016533807829181496,
+      "loss": 2.3766,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8706666666666667,
+      "grad_norm": 1.2693040370941162,
+      "learning_rate": 0.00016532028469750892,
+      "loss": 2.355,
+      "step": 1959
+    },
+    {
+      "epoch": 0.8711111111111111,
+      "grad_norm": 1.28933846950531,
+      "learning_rate": 0.00016530249110320287,
+      "loss": 2.1452,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8715555555555555,
+      "grad_norm": 1.3570892810821533,
+      "learning_rate": 0.0001652846975088968,
+      "loss": 2.8351,
+      "step": 1961
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 1.373550534248352,
+      "learning_rate": 0.00016526690391459076,
+      "loss": 2.1488,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8724444444444445,
+      "grad_norm": 1.1201701164245605,
+      "learning_rate": 0.0001652491103202847,
+      "loss": 2.1829,
+      "step": 1963
+    },
+    {
+      "epoch": 0.8728888888888889,
+      "grad_norm": 1.28224778175354,
+      "learning_rate": 0.00016523131672597865,
+      "loss": 2.6936,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8733333333333333,
+      "grad_norm": 1.540144443511963,
+      "learning_rate": 0.0001652135231316726,
+      "loss": 2.5752,
+      "step": 1965
+    },
+    {
+      "epoch": 0.8737777777777778,
+      "grad_norm": 1.3699182271957397,
+      "learning_rate": 0.00016519572953736656,
+      "loss": 2.5122,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8742222222222222,
+      "grad_norm": 1.5557972192764282,
+      "learning_rate": 0.00016517793594306052,
+      "loss": 2.2133,
+      "step": 1967
+    },
+    {
+      "epoch": 0.8746666666666667,
+      "grad_norm": 1.4491480588912964,
+      "learning_rate": 0.00016516014234875444,
+      "loss": 2.4586,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8751111111111111,
+      "grad_norm": 1.3533332347869873,
+      "learning_rate": 0.0001651423487544484,
+      "loss": 2.1343,
+      "step": 1969
+    },
+    {
+      "epoch": 0.8755555555555555,
+      "grad_norm": 1.8715101480484009,
+      "learning_rate": 0.00016512455516014236,
+      "loss": 2.5727,
+      "step": 1970
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 1.507683515548706,
+      "learning_rate": 0.00016510676156583631,
+      "loss": 1.8529,
+      "step": 1971
+    },
+    {
+      "epoch": 0.8764444444444445,
+      "grad_norm": 1.381305456161499,
+      "learning_rate": 0.00016508896797153027,
+      "loss": 2.2016,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8768888888888889,
+      "grad_norm": 1.4611423015594482,
+      "learning_rate": 0.0001650711743772242,
+      "loss": 2.4748,
+      "step": 1973
+    },
+    {
+      "epoch": 0.8773333333333333,
+      "grad_norm": 1.5445431470870972,
+      "learning_rate": 0.00016505338078291816,
+      "loss": 2.5277,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8777777777777778,
+      "grad_norm": 1.4540585279464722,
+      "learning_rate": 0.0001650355871886121,
+      "loss": 2.4206,
+      "step": 1975
+    },
+    {
+      "epoch": 0.8782222222222222,
+      "grad_norm": 1.4331034421920776,
+      "learning_rate": 0.00016501779359430604,
+      "loss": 2.6362,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8786666666666667,
+      "grad_norm": 1.668470859527588,
+      "learning_rate": 0.000165,
+      "loss": 2.5848,
+      "step": 1977
+    },
+    {
+      "epoch": 0.8791111111111111,
+      "grad_norm": 1.2265642881393433,
+      "learning_rate": 0.00016498220640569396,
+      "loss": 2.0659,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8795555555555555,
+      "grad_norm": 1.6252071857452393,
+      "learning_rate": 0.0001649644128113879,
+      "loss": 2.3045,
+      "step": 1979
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3835192918777466,
+      "learning_rate": 0.00016494661921708184,
+      "loss": 2.3903,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8804444444444445,
+      "grad_norm": 1.4605368375778198,
+      "learning_rate": 0.0001649288256227758,
+      "loss": 2.7493,
+      "step": 1981
+    },
+    {
+      "epoch": 0.8808888888888889,
+      "grad_norm": 1.863997459411621,
+      "learning_rate": 0.00016491103202846975,
+      "loss": 2.3149,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8813333333333333,
+      "grad_norm": 1.7310175895690918,
+      "learning_rate": 0.0001648932384341637,
+      "loss": 2.7611,
+      "step": 1983
+    },
+    {
+      "epoch": 0.8817777777777778,
+      "grad_norm": 1.4304709434509277,
+      "learning_rate": 0.00016487544483985767,
+      "loss": 1.2887,
+      "step": 1984
+    },
+    {
+      "epoch": 0.8822222222222222,
+      "grad_norm": 1.2733263969421387,
+      "learning_rate": 0.00016485765124555162,
+      "loss": 2.0624,
+      "step": 1985
+    },
+    {
+      "epoch": 0.8826666666666667,
+      "grad_norm": 1.5163559913635254,
+      "learning_rate": 0.00016483985765124555,
+      "loss": 2.7885,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8831111111111111,
+      "grad_norm": 1.3988975286483765,
+      "learning_rate": 0.0001648220640569395,
+      "loss": 2.1499,
+      "step": 1987
+    },
+    {
+      "epoch": 0.8835555555555555,
+      "grad_norm": 1.324731469154358,
+      "learning_rate": 0.00016480427046263347,
+      "loss": 2.1789,
+      "step": 1988
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 1.5569076538085938,
+      "learning_rate": 0.0001647864768683274,
+      "loss": 2.4129,
+      "step": 1989
+    },
+    {
+      "epoch": 0.8844444444444445,
+      "grad_norm": 2.2256200313568115,
+      "learning_rate": 0.00016476868327402135,
+      "loss": 2.9758,
+      "step": 1990
+    },
+    {
+      "epoch": 0.8848888888888888,
+      "grad_norm": 1.666374683380127,
+      "learning_rate": 0.0001647508896797153,
+      "loss": 2.7925,
+      "step": 1991
+    },
+    {
+      "epoch": 0.8853333333333333,
+      "grad_norm": 1.6541733741760254,
+      "learning_rate": 0.00016473309608540927,
+      "loss": 2.4019,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8857777777777778,
+      "grad_norm": 1.5037000179290771,
+      "learning_rate": 0.0001647153024911032,
+      "loss": 2.6672,
+      "step": 1993
+    },
+    {
+      "epoch": 0.8862222222222222,
+      "grad_norm": 1.7169650793075562,
+      "learning_rate": 0.00016469750889679715,
+      "loss": 2.7251,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8866666666666667,
+      "grad_norm": 1.9718469381332397,
+      "learning_rate": 0.0001646797153024911,
+      "loss": 2.1821,
+      "step": 1995
+    },
+    {
+      "epoch": 0.8871111111111111,
+      "grad_norm": 1.751865029335022,
+      "learning_rate": 0.00016466192170818506,
+      "loss": 2.6323,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8875555555555555,
+      "grad_norm": 1.602544903755188,
+      "learning_rate": 0.00016464412811387902,
+      "loss": 2.6532,
+      "step": 1997
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 1.865159273147583,
+      "learning_rate": 0.00016462633451957298,
+      "loss": 3.023,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8884444444444445,
+      "grad_norm": 1.7071006298065186,
+      "learning_rate": 0.0001646085409252669,
+      "loss": 2.6798,
+      "step": 1999
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 2.0052783489227295,
+      "learning_rate": 0.00016459074733096086,
+      "loss": 2.7293,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8893333333333333,
+      "grad_norm": 1.0259556770324707,
+      "learning_rate": 0.00016457295373665482,
+      "loss": 2.552,
+      "step": 2001
+    },
+    {
+      "epoch": 0.8897777777777778,
+      "grad_norm": 1.1333271265029907,
+      "learning_rate": 0.00016455516014234875,
+      "loss": 2.8014,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8902222222222222,
+      "grad_norm": 0.973343551158905,
+      "learning_rate": 0.0001645373665480427,
+      "loss": 2.5627,
+      "step": 2003
+    },
+    {
+      "epoch": 0.8906666666666667,
+      "grad_norm": 0.9888269901275635,
+      "learning_rate": 0.00016451957295373666,
+      "loss": 2.2681,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8911111111111111,
+      "grad_norm": 1.2307792901992798,
+      "learning_rate": 0.00016450177935943062,
+      "loss": 2.1229,
+      "step": 2005
+    },
+    {
+      "epoch": 0.8915555555555555,
+      "grad_norm": 1.0843390226364136,
+      "learning_rate": 0.00016448398576512455,
+      "loss": 2.3058,
+      "step": 2006
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 1.3246184587478638,
+      "learning_rate": 0.0001644661921708185,
+      "loss": 2.5498,
+      "step": 2007
+    },
+    {
+      "epoch": 0.8924444444444445,
+      "grad_norm": 1.156726360321045,
+      "learning_rate": 0.00016444839857651246,
+      "loss": 2.3375,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8928888888888888,
+      "grad_norm": 1.061392903327942,
+      "learning_rate": 0.00016443060498220642,
+      "loss": 1.2548,
+      "step": 2009
+    },
+    {
+      "epoch": 0.8933333333333333,
+      "grad_norm": 1.357146978378296,
+      "learning_rate": 0.00016441281138790037,
+      "loss": 2.8068,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8937777777777778,
+      "grad_norm": 1.3615721464157104,
+      "learning_rate": 0.00016439501779359433,
+      "loss": 2.4011,
+      "step": 2011
+    },
+    {
+      "epoch": 0.8942222222222223,
+      "grad_norm": 1.3508340120315552,
+      "learning_rate": 0.00016437722419928826,
+      "loss": 2.5085,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8946666666666667,
+      "grad_norm": 1.4771041870117188,
+      "learning_rate": 0.00016435943060498222,
+      "loss": 1.8489,
+      "step": 2013
+    },
+    {
+      "epoch": 0.8951111111111111,
+      "grad_norm": 1.2312934398651123,
+      "learning_rate": 0.00016434163701067617,
+      "loss": 1.9961,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8955555555555555,
+      "grad_norm": 1.308119535446167,
+      "learning_rate": 0.0001643238434163701,
+      "loss": 2.6029,
+      "step": 2015
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 1.3931859731674194,
+      "learning_rate": 0.00016430604982206406,
+      "loss": 2.5537,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8964444444444445,
+      "grad_norm": 1.5460842847824097,
+      "learning_rate": 0.00016428825622775802,
+      "loss": 2.5682,
+      "step": 2017
+    },
+    {
+      "epoch": 0.8968888888888888,
+      "grad_norm": 1.4253586530685425,
+      "learning_rate": 0.00016427046263345197,
+      "loss": 2.6531,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8973333333333333,
+      "grad_norm": 1.2461731433868408,
+      "learning_rate": 0.0001642526690391459,
+      "loss": 1.9027,
+      "step": 2019
+    },
+    {
+      "epoch": 0.8977777777777778,
+      "grad_norm": 1.418392539024353,
+      "learning_rate": 0.00016423487544483986,
+      "loss": 2.5344,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8982222222222223,
+      "grad_norm": 1.5666571855545044,
+      "learning_rate": 0.00016421708185053381,
+      "loss": 2.4009,
+      "step": 2021
+    },
+    {
+      "epoch": 0.8986666666666666,
+      "grad_norm": 1.2651710510253906,
+      "learning_rate": 0.00016419928825622777,
+      "loss": 2.1727,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8991111111111111,
+      "grad_norm": 1.4496339559555054,
+      "learning_rate": 0.00016418149466192173,
+      "loss": 2.3306,
+      "step": 2023
+    },
+    {
+      "epoch": 0.8995555555555556,
+      "grad_norm": 1.3731813430786133,
+      "learning_rate": 0.00016416370106761568,
+      "loss": 2.0778,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.383135199546814,
+      "learning_rate": 0.0001641459074733096,
+      "loss": 2.7001,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9004444444444445,
+      "grad_norm": 1.2729257345199585,
+      "learning_rate": 0.00016412811387900357,
+      "loss": 2.1518,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9008888888888889,
+      "grad_norm": 1.5172004699707031,
+      "learning_rate": 0.00016411032028469753,
+      "loss": 2.4128,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9013333333333333,
+      "grad_norm": 1.3917080163955688,
+      "learning_rate": 0.00016409252669039146,
+      "loss": 1.2132,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9017777777777778,
+      "grad_norm": 1.359440565109253,
+      "learning_rate": 0.0001640747330960854,
+      "loss": 2.6389,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9022222222222223,
+      "grad_norm": 1.6984691619873047,
+      "learning_rate": 0.00016405693950177937,
+      "loss": 2.5598,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9026666666666666,
+      "grad_norm": 1.3481029272079468,
+      "learning_rate": 0.00016403914590747332,
+      "loss": 2.2033,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9031111111111111,
+      "grad_norm": 1.4901320934295654,
+      "learning_rate": 0.00016402135231316725,
+      "loss": 2.227,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9035555555555556,
+      "grad_norm": 1.4756929874420166,
+      "learning_rate": 0.0001640035587188612,
+      "loss": 2.3098,
+      "step": 2033
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 2.4183623790740967,
+      "learning_rate": 0.00016398576512455517,
+      "loss": 2.2787,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9044444444444445,
+      "grad_norm": 1.7991214990615845,
+      "learning_rate": 0.00016396797153024912,
+      "loss": 2.3717,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9048888888888889,
+      "grad_norm": 1.7338757514953613,
+      "learning_rate": 0.00016395017793594308,
+      "loss": 2.9113,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9053333333333333,
+      "grad_norm": 1.5208733081817627,
+      "learning_rate": 0.00016393238434163704,
+      "loss": 2.0556,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9057777777777778,
+      "grad_norm": 1.5961337089538574,
+      "learning_rate": 0.00016391459074733097,
+      "loss": 2.093,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9062222222222223,
+      "grad_norm": 1.6487394571304321,
+      "learning_rate": 0.00016389679715302492,
+      "loss": 2.6833,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9066666666666666,
+      "grad_norm": 2.891885995864868,
+      "learning_rate": 0.00016387900355871888,
+      "loss": 1.6798,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9071111111111111,
+      "grad_norm": 1.8462321758270264,
+      "learning_rate": 0.0001638612099644128,
+      "loss": 2.5937,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9075555555555556,
+      "grad_norm": 1.7322661876678467,
+      "learning_rate": 0.00016384341637010676,
+      "loss": 2.7494,
+      "step": 2042
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 1.7559343576431274,
+      "learning_rate": 0.00016382562277580072,
+      "loss": 2.7197,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9084444444444445,
+      "grad_norm": 1.7972688674926758,
+      "learning_rate": 0.00016380782918149468,
+      "loss": 2.2508,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9088888888888889,
+      "grad_norm": 1.5934430360794067,
+      "learning_rate": 0.0001637900355871886,
+      "loss": 2.5825,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9093333333333333,
+      "grad_norm": 2.1732285022735596,
+      "learning_rate": 0.00016377224199288256,
+      "loss": 2.5181,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9097777777777778,
+      "grad_norm": 1.8256020545959473,
+      "learning_rate": 0.00016375444839857652,
+      "loss": 2.4343,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9102222222222223,
+      "grad_norm": 1.8094704151153564,
+      "learning_rate": 0.00016373665480427048,
+      "loss": 2.4485,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9106666666666666,
+      "grad_norm": 2.246121644973755,
+      "learning_rate": 0.00016371886120996443,
+      "loss": 3.1074,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9111111111111111,
+      "grad_norm": 3.0540249347686768,
+      "learning_rate": 0.0001637010676156584,
+      "loss": 1.8492,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9115555555555556,
+      "grad_norm": 0.9966058731079102,
+      "learning_rate": 0.00016368327402135232,
+      "loss": 2.6141,
+      "step": 2051
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 1.0565382242202759,
+      "learning_rate": 0.00016366548042704628,
+      "loss": 2.5051,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9124444444444444,
+      "grad_norm": 0.999159574508667,
+      "learning_rate": 0.00016364768683274023,
+      "loss": 1.2208,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9128888888888889,
+      "grad_norm": 1.0446531772613525,
+      "learning_rate": 0.00016362989323843416,
+      "loss": 2.403,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9133333333333333,
+      "grad_norm": 1.7847496271133423,
+      "learning_rate": 0.00016361209964412812,
+      "loss": 1.4985,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9137777777777778,
+      "grad_norm": 1.1442463397979736,
+      "learning_rate": 0.00016359430604982207,
+      "loss": 2.8388,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9142222222222223,
+      "grad_norm": 1.1940288543701172,
+      "learning_rate": 0.00016357651245551603,
+      "loss": 2.7245,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9146666666666666,
+      "grad_norm": 1.3934929370880127,
+      "learning_rate": 0.00016355871886120996,
+      "loss": 2.8598,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9151111111111111,
+      "grad_norm": 1.361688256263733,
+      "learning_rate": 0.00016354092526690392,
+      "loss": 2.7173,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9155555555555556,
+      "grad_norm": 1.3823915719985962,
+      "learning_rate": 0.00016352313167259787,
+      "loss": 1.5828,
+      "step": 2060
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 1.3563628196716309,
+      "learning_rate": 0.00016350533807829183,
+      "loss": 2.5457,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9164444444444444,
+      "grad_norm": 1.2287131547927856,
+      "learning_rate": 0.00016348754448398579,
+      "loss": 2.9229,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9168888888888889,
+      "grad_norm": 1.1974775791168213,
+      "learning_rate": 0.00016346975088967972,
+      "loss": 2.2515,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9173333333333333,
+      "grad_norm": 1.5387582778930664,
+      "learning_rate": 0.00016345195729537367,
+      "loss": 1.6172,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9177777777777778,
+      "grad_norm": 1.3966673612594604,
+      "learning_rate": 0.00016343416370106763,
+      "loss": 2.771,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9182222222222223,
+      "grad_norm": 1.7705440521240234,
+      "learning_rate": 0.00016341637010676156,
+      "loss": 1.446,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9186666666666666,
+      "grad_norm": 1.4579976797103882,
+      "learning_rate": 0.00016339857651245551,
+      "loss": 1.8905,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9191111111111111,
+      "grad_norm": 1.4872655868530273,
+      "learning_rate": 0.00016338078291814947,
+      "loss": 2.2954,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9195555555555556,
+      "grad_norm": 1.5094295740127563,
+      "learning_rate": 0.00016336298932384343,
+      "loss": 3.1148,
+      "step": 2069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5641443729400635,
+      "learning_rate": 0.00016334519572953736,
+      "loss": 2.7561,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9204444444444444,
+      "grad_norm": 1.4170724153518677,
+      "learning_rate": 0.00016332740213523131,
+      "loss": 2.2233,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9208888888888889,
+      "grad_norm": 1.633217215538025,
+      "learning_rate": 0.00016330960854092527,
+      "loss": 2.9578,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9213333333333333,
+      "grad_norm": 1.2382259368896484,
+      "learning_rate": 0.00016329181494661923,
+      "loss": 1.4224,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9217777777777778,
+      "grad_norm": 1.9555550813674927,
+      "learning_rate": 0.00016327402135231318,
+      "loss": 2.4531,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9222222222222223,
+      "grad_norm": 1.4946098327636719,
+      "learning_rate": 0.00016325622775800714,
+      "loss": 1.3729,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9226666666666666,
+      "grad_norm": 0.9218053817749023,
+      "learning_rate": 0.00016323843416370107,
+      "loss": 1.1931,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9231111111111111,
+      "grad_norm": 1.8889461755752563,
+      "learning_rate": 0.00016322064056939503,
+      "loss": 2.8331,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9235555555555556,
+      "grad_norm": 2.0126349925994873,
+      "learning_rate": 0.00016320284697508898,
+      "loss": 2.0097,
+      "step": 2078
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 1.5034337043762207,
+      "learning_rate": 0.0001631850533807829,
+      "loss": 2.4002,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9244444444444444,
+      "grad_norm": 1.582135558128357,
+      "learning_rate": 0.00016316725978647687,
+      "loss": 2.4081,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9248888888888889,
+      "grad_norm": 1.4720321893692017,
+      "learning_rate": 0.00016314946619217082,
+      "loss": 2.0486,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9253333333333333,
+      "grad_norm": 1.524876594543457,
+      "learning_rate": 0.00016313167259786478,
+      "loss": 2.4884,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9257777777777778,
+      "grad_norm": 1.3611582517623901,
+      "learning_rate": 0.0001631138790035587,
+      "loss": 2.2896,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9262222222222222,
+      "grad_norm": 1.6695072650909424,
+      "learning_rate": 0.00016309608540925267,
+      "loss": 2.5879,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9266666666666666,
+      "grad_norm": 1.5357855558395386,
+      "learning_rate": 0.00016307829181494662,
+      "loss": 2.4094,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9271111111111111,
+      "grad_norm": 1.6165261268615723,
+      "learning_rate": 0.00016306049822064058,
+      "loss": 2.3728,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9275555555555556,
+      "grad_norm": 1.7907167673110962,
+      "learning_rate": 0.00016304270462633454,
+      "loss": 2.4651,
+      "step": 2087
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 1.6630196571350098,
+      "learning_rate": 0.0001630249110320285,
+      "loss": 2.2993,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9284444444444444,
+      "grad_norm": 1.7729859352111816,
+      "learning_rate": 0.00016300711743772242,
+      "loss": 3.0367,
+      "step": 2089
+    },
+    {
+      "epoch": 0.9288888888888889,
+      "grad_norm": 1.6440625190734863,
+      "learning_rate": 0.00016298932384341638,
+      "loss": 2.4056,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9293333333333333,
+      "grad_norm": 1.8084213733673096,
+      "learning_rate": 0.00016297153024911034,
+      "loss": 2.5752,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9297777777777778,
+      "grad_norm": 1.6393537521362305,
+      "learning_rate": 0.00016295373665480426,
+      "loss": 2.2847,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9302222222222222,
+      "grad_norm": 1.849129319190979,
+      "learning_rate": 0.00016293594306049822,
+      "loss": 2.8678,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9306666666666666,
+      "grad_norm": 2.1448423862457275,
+      "learning_rate": 0.00016291814946619218,
+      "loss": 3.2427,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9311111111111111,
+      "grad_norm": 1.7885196208953857,
+      "learning_rate": 0.00016290035587188613,
+      "loss": 2.0247,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9315555555555556,
+      "grad_norm": 1.889359474182129,
+      "learning_rate": 0.00016288256227758006,
+      "loss": 2.492,
+      "step": 2096
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 1.7645171880722046,
+      "learning_rate": 0.00016286476868327402,
+      "loss": 2.6633,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9324444444444444,
+      "grad_norm": 1.768557071685791,
+      "learning_rate": 0.00016284697508896798,
+      "loss": 2.4075,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9328888888888889,
+      "grad_norm": 2.301161766052246,
+      "learning_rate": 0.00016282918149466193,
+      "loss": 2.7808,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 2.5840353965759277,
+      "learning_rate": 0.0001628113879003559,
+      "loss": 3.1963,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9337777777777778,
+      "grad_norm": 1.097058653831482,
+      "learning_rate": 0.00016279359430604985,
+      "loss": 2.4867,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9342222222222222,
+      "grad_norm": 0.9999614357948303,
+      "learning_rate": 0.00016277580071174378,
+      "loss": 1.277,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9346666666666666,
+      "grad_norm": 0.8796145915985107,
+      "learning_rate": 0.00016275800711743773,
+      "loss": 1.4156,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9351111111111111,
+      "grad_norm": 1.1534805297851562,
+      "learning_rate": 0.0001627402135231317,
+      "loss": 2.3798,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9355555555555556,
+      "grad_norm": 1.3361473083496094,
+      "learning_rate": 0.00016272241992882562,
+      "loss": 2.6389,
+      "step": 2105
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 1.3026865720748901,
+      "learning_rate": 0.00016270462633451957,
+      "loss": 2.9566,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9364444444444444,
+      "grad_norm": 1.2137173414230347,
+      "learning_rate": 0.00016268683274021353,
+      "loss": 2.7551,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9368888888888889,
+      "grad_norm": 1.1544345617294312,
+      "learning_rate": 0.0001626690391459075,
+      "loss": 1.6071,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9373333333333334,
+      "grad_norm": 1.3754730224609375,
+      "learning_rate": 0.00016265124555160142,
+      "loss": 2.681,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9377777777777778,
+      "grad_norm": 1.2891576290130615,
+      "learning_rate": 0.00016263345195729537,
+      "loss": 2.5061,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9382222222222222,
+      "grad_norm": 1.6436229944229126,
+      "learning_rate": 0.00016261565836298933,
+      "loss": 2.7442,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9386666666666666,
+      "grad_norm": 1.6582046747207642,
+      "learning_rate": 0.00016259786476868329,
+      "loss": 2.4712,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9391111111111111,
+      "grad_norm": 1.338775873184204,
+      "learning_rate": 0.00016258007117437724,
+      "loss": 2.2104,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9395555555555556,
+      "grad_norm": 1.4280105829238892,
+      "learning_rate": 0.0001625622775800712,
+      "loss": 1.706,
+      "step": 2114
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3031154870986938,
+      "learning_rate": 0.00016254448398576513,
+      "loss": 2.9918,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9404444444444444,
+      "grad_norm": 1.6146162748336792,
+      "learning_rate": 0.00016252669039145908,
+      "loss": 2.577,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9408888888888889,
+      "grad_norm": 1.5383062362670898,
+      "learning_rate": 0.00016250889679715304,
+      "loss": 2.38,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9413333333333334,
+      "grad_norm": 1.5242427587509155,
+      "learning_rate": 0.00016249110320284697,
+      "loss": 2.5093,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9417777777777778,
+      "grad_norm": 1.5081580877304077,
+      "learning_rate": 0.00016247330960854093,
+      "loss": 2.983,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9422222222222222,
+      "grad_norm": 1.362468957901001,
+      "learning_rate": 0.00016245551601423488,
+      "loss": 2.4969,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9426666666666667,
+      "grad_norm": 1.5495905876159668,
+      "learning_rate": 0.00016243772241992884,
+      "loss": 2.474,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9431111111111111,
+      "grad_norm": 1.6289684772491455,
+      "learning_rate": 0.00016241992882562277,
+      "loss": 2.7819,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9435555555555556,
+      "grad_norm": 1.3928167819976807,
+      "learning_rate": 0.00016240213523131673,
+      "loss": 2.3204,
+      "step": 2123
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 1.5994818210601807,
+      "learning_rate": 0.00016238434163701068,
+      "loss": 2.783,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9444444444444444,
+      "grad_norm": 1.4788901805877686,
+      "learning_rate": 0.00016236654804270464,
+      "loss": 2.4959,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9448888888888889,
+      "grad_norm": 1.7189639806747437,
+      "learning_rate": 0.0001623487544483986,
+      "loss": 2.7708,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9453333333333334,
+      "grad_norm": 1.6642398834228516,
+      "learning_rate": 0.00016233096085409255,
+      "loss": 1.4068,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9457777777777778,
+      "grad_norm": 1.9017895460128784,
+      "learning_rate": 0.00016231316725978648,
+      "loss": 2.0995,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9462222222222222,
+      "grad_norm": 1.4762321710586548,
+      "learning_rate": 0.00016229537366548044,
+      "loss": 2.8266,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9466666666666667,
+      "grad_norm": 1.1698795557022095,
+      "learning_rate": 0.0001622775800711744,
+      "loss": 1.2059,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9471111111111111,
+      "grad_norm": 1.7860333919525146,
+      "learning_rate": 0.00016225978647686832,
+      "loss": 3.0256,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9475555555555556,
+      "grad_norm": 1.6017791032791138,
+      "learning_rate": 0.00016224199288256228,
+      "loss": 3.0768,
+      "step": 2132
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 1.6588813066482544,
+      "learning_rate": 0.00016222419928825624,
+      "loss": 2.8438,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9484444444444444,
+      "grad_norm": 1.97148597240448,
+      "learning_rate": 0.0001622064056939502,
+      "loss": 2.4861,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9488888888888889,
+      "grad_norm": 1.5532220602035522,
+      "learning_rate": 0.00016218861209964412,
+      "loss": 2.999,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9493333333333334,
+      "grad_norm": 1.6381109952926636,
+      "learning_rate": 0.00016217081850533808,
+      "loss": 2.6297,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9497777777777778,
+      "grad_norm": 1.6247540712356567,
+      "learning_rate": 0.00016215302491103204,
+      "loss": 2.0808,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9502222222222222,
+      "grad_norm": 2.061701536178589,
+      "learning_rate": 0.000162135231316726,
+      "loss": 2.9111,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9506666666666667,
+      "grad_norm": 1.9049525260925293,
+      "learning_rate": 0.00016211743772241995,
+      "loss": 1.179,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9511111111111111,
+      "grad_norm": 1.4355841875076294,
+      "learning_rate": 0.0001620996441281139,
+      "loss": 2.2306,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9515555555555556,
+      "grad_norm": 2.1435563564300537,
+      "learning_rate": 0.00016208185053380783,
+      "loss": 3.1722,
+      "step": 2141
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 1.7206003665924072,
+      "learning_rate": 0.0001620640569395018,
+      "loss": 2.6664,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9524444444444444,
+      "grad_norm": 1.7475922107696533,
+      "learning_rate": 0.00016204626334519575,
+      "loss": 2.2984,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9528888888888889,
+      "grad_norm": 1.429494857788086,
+      "learning_rate": 0.00016202846975088968,
+      "loss": 2.0885,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9533333333333334,
+      "grad_norm": 1.6629289388656616,
+      "learning_rate": 0.00016201067615658363,
+      "loss": 2.312,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9537777777777777,
+      "grad_norm": 1.6623343229293823,
+      "learning_rate": 0.0001619928825622776,
+      "loss": 2.3646,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9542222222222222,
+      "grad_norm": 2.0395777225494385,
+      "learning_rate": 0.00016197508896797155,
+      "loss": 2.4601,
+      "step": 2147
+    },
+    {
+      "epoch": 0.9546666666666667,
+      "grad_norm": 2.142592191696167,
+      "learning_rate": 0.00016195729537366548,
+      "loss": 2.5194,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9551111111111111,
+      "grad_norm": 1.8677221536636353,
+      "learning_rate": 0.00016193950177935943,
+      "loss": 2.3944,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9555555555555556,
+      "grad_norm": 1.8945562839508057,
+      "learning_rate": 0.0001619217081850534,
+      "loss": 2.6568,
+      "step": 2150
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 1.6855690479278564,
+      "learning_rate": 0.00016190391459074735,
+      "loss": 3.2555,
+      "step": 2151
+    },
+    {
+      "epoch": 0.9564444444444444,
+      "grad_norm": 1.2461864948272705,
+      "learning_rate": 0.0001618861209964413,
+      "loss": 2.8541,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9568888888888889,
+      "grad_norm": 1.181875467300415,
+      "learning_rate": 0.00016186832740213523,
+      "loss": 2.6893,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9573333333333334,
+      "grad_norm": 1.1439327001571655,
+      "learning_rate": 0.0001618505338078292,
+      "loss": 2.8082,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9577777777777777,
+      "grad_norm": 1.2180849313735962,
+      "learning_rate": 0.00016183274021352314,
+      "loss": 2.4687,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9582222222222222,
+      "grad_norm": 1.2314586639404297,
+      "learning_rate": 0.0001618149466192171,
+      "loss": 2.1729,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9586666666666667,
+      "grad_norm": 1.3291667699813843,
+      "learning_rate": 0.00016179715302491103,
+      "loss": 2.5573,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9591111111111111,
+      "grad_norm": 1.359408974647522,
+      "learning_rate": 0.000161779359430605,
+      "loss": 2.2368,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9595555555555556,
+      "grad_norm": 1.2051069736480713,
+      "learning_rate": 0.00016176156583629894,
+      "loss": 2.2714,
+      "step": 2159
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4360554218292236,
+      "learning_rate": 0.00016174377224199287,
+      "loss": 2.4253,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9604444444444444,
+      "grad_norm": 1.3218653202056885,
+      "learning_rate": 0.00016172597864768683,
+      "loss": 2.38,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9608888888888889,
+      "grad_norm": 1.658354640007019,
+      "learning_rate": 0.00016170818505338079,
+      "loss": 3.0823,
+      "step": 2162
+    },
+    {
+      "epoch": 0.9613333333333334,
+      "grad_norm": 1.390121579170227,
+      "learning_rate": 0.00016169039145907474,
+      "loss": 2.6402,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9617777777777777,
+      "grad_norm": 1.279600977897644,
+      "learning_rate": 0.0001616725978647687,
+      "loss": 2.6816,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9622222222222222,
+      "grad_norm": 1.3409185409545898,
+      "learning_rate": 0.00016165480427046266,
+      "loss": 2.0361,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9626666666666667,
+      "grad_norm": 1.7012516260147095,
+      "learning_rate": 0.00016163701067615658,
+      "loss": 2.7103,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9631111111111111,
+      "grad_norm": 1.6114073991775513,
+      "learning_rate": 0.00016161921708185054,
+      "loss": 2.2603,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9635555555555556,
+      "grad_norm": 1.3906176090240479,
+      "learning_rate": 0.0001616014234875445,
+      "loss": 2.4579,
+      "step": 2168
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 1.3002898693084717,
+      "learning_rate": 0.00016158362989323845,
+      "loss": 2.3782,
+      "step": 2169
+    },
+    {
+      "epoch": 0.9644444444444444,
+      "grad_norm": 1.4082340002059937,
+      "learning_rate": 0.00016156583629893238,
+      "loss": 2.6653,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9648888888888889,
+      "grad_norm": 1.4760489463806152,
+      "learning_rate": 0.00016154804270462634,
+      "loss": 2.7057,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9653333333333334,
+      "grad_norm": 1.411620020866394,
+      "learning_rate": 0.0001615302491103203,
+      "loss": 2.3573,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9657777777777777,
+      "grad_norm": 1.4951653480529785,
+      "learning_rate": 0.00016151245551601423,
+      "loss": 2.9025,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9662222222222222,
+      "grad_norm": 1.6565364599227905,
+      "learning_rate": 0.00016149466192170818,
+      "loss": 3.272,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9666666666666667,
+      "grad_norm": 1.4833595752716064,
+      "learning_rate": 0.00016147686832740214,
+      "loss": 2.6249,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9671111111111111,
+      "grad_norm": 1.5216375589370728,
+      "learning_rate": 0.0001614590747330961,
+      "loss": 2.4307,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9675555555555555,
+      "grad_norm": 1.8078597784042358,
+      "learning_rate": 0.00016144128113879005,
+      "loss": 2.629,
+      "step": 2177
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 1.560192584991455,
+      "learning_rate": 0.000161423487544484,
+      "loss": 2.6558,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9684444444444444,
+      "grad_norm": 1.6416150331497192,
+      "learning_rate": 0.00016140569395017794,
+      "loss": 2.3076,
+      "step": 2179
+    },
+    {
+      "epoch": 0.9688888888888889,
+      "grad_norm": 1.6443932056427002,
+      "learning_rate": 0.0001613879003558719,
+      "loss": 2.9646,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9693333333333334,
+      "grad_norm": 2.160329580307007,
+      "learning_rate": 0.00016137010676156585,
+      "loss": 2.0787,
+      "step": 2181
+    },
+    {
+      "epoch": 0.9697777777777777,
+      "grad_norm": 1.5181187391281128,
+      "learning_rate": 0.00016135231316725978,
+      "loss": 2.5944,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9702222222222222,
+      "grad_norm": 1.5956158638000488,
+      "learning_rate": 0.00016133451957295374,
+      "loss": 2.5744,
+      "step": 2183
+    },
+    {
+      "epoch": 0.9706666666666667,
+      "grad_norm": 1.6073002815246582,
+      "learning_rate": 0.0001613167259786477,
+      "loss": 2.2124,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9711111111111111,
+      "grad_norm": 1.5423829555511475,
+      "learning_rate": 0.00016129893238434165,
+      "loss": 2.5337,
+      "step": 2185
+    },
+    {
+      "epoch": 0.9715555555555555,
+      "grad_norm": 1.5740853548049927,
+      "learning_rate": 0.00016128113879003558,
+      "loss": 2.7245,
+      "step": 2186
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 1.5205440521240234,
+      "learning_rate": 0.00016126334519572954,
+      "loss": 2.6604,
+      "step": 2187
+    },
+    {
+      "epoch": 0.9724444444444444,
+      "grad_norm": 1.425803303718567,
+      "learning_rate": 0.0001612455516014235,
+      "loss": 2.2333,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9728888888888889,
+      "grad_norm": 1.6136490106582642,
+      "learning_rate": 0.00016122775800711745,
+      "loss": 2.2972,
+      "step": 2189
+    },
+    {
+      "epoch": 0.9733333333333334,
+      "grad_norm": 2.0137991905212402,
+      "learning_rate": 0.0001612099644128114,
+      "loss": 1.3431,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9737777777777777,
+      "grad_norm": 1.7414988279342651,
+      "learning_rate": 0.00016119217081850536,
+      "loss": 2.8453,
+      "step": 2191
+    },
+    {
+      "epoch": 0.9742222222222222,
+      "grad_norm": 1.7436699867248535,
+      "learning_rate": 0.0001611743772241993,
+      "loss": 2.6342,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9746666666666667,
+      "grad_norm": 2.1652956008911133,
+      "learning_rate": 0.00016115658362989325,
+      "loss": 3.4064,
+      "step": 2193
+    },
+    {
+      "epoch": 0.9751111111111112,
+      "grad_norm": 1.4634653329849243,
+      "learning_rate": 0.0001611387900355872,
+      "loss": 2.2654,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9755555555555555,
+      "grad_norm": 1.380988359451294,
+      "learning_rate": 0.00016112099644128113,
+      "loss": 1.9905,
+      "step": 2195
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 1.78019380569458,
+      "learning_rate": 0.0001611032028469751,
+      "loss": 2.7685,
+      "step": 2196
+    },
+    {
+      "epoch": 0.9764444444444444,
+      "grad_norm": 2.219177722930908,
+      "learning_rate": 0.00016108540925266905,
+      "loss": 2.8879,
+      "step": 2197
+    },
+    {
+      "epoch": 0.9768888888888889,
+      "grad_norm": 2.295215129852295,
+      "learning_rate": 0.000161067615658363,
+      "loss": 2.7534,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9773333333333334,
+      "grad_norm": 2.250352144241333,
+      "learning_rate": 0.00016104982206405693,
+      "loss": 2.7764,
+      "step": 2199
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 2.36488938331604,
+      "learning_rate": 0.0001610320284697509,
+      "loss": 3.1503,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9782222222222222,
+      "grad_norm": 1.0454128980636597,
+      "learning_rate": 0.00016101423487544485,
+      "loss": 1.2486,
+      "step": 2201
+    },
+    {
+      "epoch": 0.9786666666666667,
+      "grad_norm": 1.087558388710022,
+      "learning_rate": 0.0001609964412811388,
+      "loss": 2.7034,
+      "step": 2202
+    },
+    {
+      "epoch": 0.9791111111111112,
+      "grad_norm": 1.1958833932876587,
+      "learning_rate": 0.00016097864768683276,
+      "loss": 2.5241,
+      "step": 2203
+    },
+    {
+      "epoch": 0.9795555555555555,
+      "grad_norm": 1.2436286211013794,
+      "learning_rate": 0.00016096085409252671,
+      "loss": 2.5663,
+      "step": 2204
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.29501211643219,
+      "learning_rate": 0.00016094306049822064,
+      "loss": 2.2681,
+      "step": 2205
+    },
+    {
+      "epoch": 0.9804444444444445,
+      "grad_norm": 1.4029202461242676,
+      "learning_rate": 0.0001609252669039146,
+      "loss": 2.6442,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9808888888888889,
+      "grad_norm": 1.2167294025421143,
+      "learning_rate": 0.00016090747330960856,
+      "loss": 2.3042,
+      "step": 2207
+    },
+    {
+      "epoch": 0.9813333333333333,
+      "grad_norm": 1.1777758598327637,
+      "learning_rate": 0.00016088967971530249,
+      "loss": 2.3734,
+      "step": 2208
+    },
+    {
+      "epoch": 0.9817777777777777,
+      "grad_norm": 1.352673053741455,
+      "learning_rate": 0.00016087188612099644,
+      "loss": 2.5139,
+      "step": 2209
+    },
+    {
+      "epoch": 0.9822222222222222,
+      "grad_norm": 1.5402523279190063,
+      "learning_rate": 0.0001608540925266904,
+      "loss": 2.4843,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9826666666666667,
+      "grad_norm": 1.278908371925354,
+      "learning_rate": 0.00016083629893238436,
+      "loss": 2.6108,
+      "step": 2211
+    },
+    {
+      "epoch": 0.9831111111111112,
+      "grad_norm": 1.1538258790969849,
+      "learning_rate": 0.00016081850533807829,
+      "loss": 2.221,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9835555555555555,
+      "grad_norm": 1.3440600633621216,
+      "learning_rate": 0.00016080071174377224,
+      "loss": 2.1903,
+      "step": 2213
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 1.4237117767333984,
+      "learning_rate": 0.0001607829181494662,
+      "loss": 2.6641,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9844444444444445,
+      "grad_norm": 1.4718806743621826,
+      "learning_rate": 0.00016076512455516015,
+      "loss": 1.0382,
+      "step": 2215
+    },
+    {
+      "epoch": 0.9848888888888889,
+      "grad_norm": 1.376482605934143,
+      "learning_rate": 0.0001607473309608541,
+      "loss": 2.2212,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9853333333333333,
+      "grad_norm": 1.2039564847946167,
+      "learning_rate": 0.00016072953736654807,
+      "loss": 1.969,
+      "step": 2217
+    },
+    {
+      "epoch": 0.9857777777777778,
+      "grad_norm": 1.4104335308074951,
+      "learning_rate": 0.000160711743772242,
+      "loss": 2.5837,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9862222222222222,
+      "grad_norm": 1.5590723752975464,
+      "learning_rate": 0.00016069395017793595,
+      "loss": 2.5401,
+      "step": 2219
+    },
+    {
+      "epoch": 0.9866666666666667,
+      "grad_norm": 1.4591701030731201,
+      "learning_rate": 0.0001606761565836299,
+      "loss": 2.1405,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9871111111111112,
+      "grad_norm": 1.6289016008377075,
+      "learning_rate": 0.00016065836298932384,
+      "loss": 2.6935,
+      "step": 2221
+    },
+    {
+      "epoch": 0.9875555555555555,
+      "grad_norm": 1.431257963180542,
+      "learning_rate": 0.0001606405693950178,
+      "loss": 2.7326,
+      "step": 2222
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 1.7325876951217651,
+      "learning_rate": 0.00016062277580071175,
+      "loss": 2.2347,
+      "step": 2223
+    },
+    {
+      "epoch": 0.9884444444444445,
+      "grad_norm": 1.4045909643173218,
+      "learning_rate": 0.0001606049822064057,
+      "loss": 2.4795,
+      "step": 2224
+    },
+    {
+      "epoch": 0.9888888888888889,
+      "grad_norm": 1.5909349918365479,
+      "learning_rate": 0.00016058718861209964,
+      "loss": 2.89,
+      "step": 2225
+    },
+    {
+      "epoch": 0.9893333333333333,
+      "grad_norm": 1.7079018354415894,
+      "learning_rate": 0.0001605693950177936,
+      "loss": 2.9235,
+      "step": 2226
+    },
+    {
+      "epoch": 0.9897777777777778,
+      "grad_norm": 1.401973843574524,
+      "learning_rate": 0.00016055160142348755,
+      "loss": 2.5211,
+      "step": 2227
+    },
+    {
+      "epoch": 0.9902222222222222,
+      "grad_norm": 1.7690025568008423,
+      "learning_rate": 0.0001605338078291815,
+      "loss": 2.8388,
+      "step": 2228
+    },
+    {
+      "epoch": 0.9906666666666667,
+      "grad_norm": 1.3721446990966797,
+      "learning_rate": 0.00016051601423487546,
+      "loss": 2.6306,
+      "step": 2229
+    },
+    {
+      "epoch": 0.9911111111111112,
+      "grad_norm": 1.9013787508010864,
+      "learning_rate": 0.00016049822064056942,
+      "loss": 3.1249,
+      "step": 2230
+    },
+    {
+      "epoch": 0.9915555555555555,
+      "grad_norm": 1.5793306827545166,
+      "learning_rate": 0.00016048042704626335,
+      "loss": 0.0805,
+      "step": 2231
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 1.6632827520370483,
+      "learning_rate": 0.0001604626334519573,
+      "loss": 2.1838,
+      "step": 2232
+    },
+    {
+      "epoch": 0.9924444444444445,
+      "grad_norm": 1.837686538696289,
+      "learning_rate": 0.00016044483985765126,
+      "loss": 2.245,
+      "step": 2233
+    },
+    {
+      "epoch": 0.9928888888888889,
+      "grad_norm": 2.039071559906006,
+      "learning_rate": 0.0001604270462633452,
+      "loss": 2.713,
+      "step": 2234
+    },
+    {
+      "epoch": 0.9933333333333333,
+      "grad_norm": 1.6307156085968018,
+      "learning_rate": 0.00016040925266903915,
+      "loss": 3.0165,
+      "step": 2235
+    },
+    {
+      "epoch": 0.9937777777777778,
+      "grad_norm": 1.4582569599151611,
+      "learning_rate": 0.0001603914590747331,
+      "loss": 2.3939,
+      "step": 2236
+    },
+    {
+      "epoch": 0.9942222222222222,
+      "grad_norm": 1.3615283966064453,
+      "learning_rate": 0.00016037366548042706,
+      "loss": 1.9555,
+      "step": 2237
+    },
+    {
+      "epoch": 0.9946666666666667,
+      "grad_norm": 1.5992584228515625,
+      "learning_rate": 0.000160355871886121,
+      "loss": 2.4956,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9951111111111111,
+      "grad_norm": 1.8983911275863647,
+      "learning_rate": 0.00016033807829181495,
+      "loss": 2.6738,
+      "step": 2239
+    },
+    {
+      "epoch": 0.9955555555555555,
+      "grad_norm": 1.622286081314087,
+      "learning_rate": 0.0001603202846975089,
+      "loss": 2.2483,
+      "step": 2240
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 1.2114402055740356,
+      "learning_rate": 0.00016030249110320286,
+      "loss": 1.3111,
+      "step": 2241
+    },
+    {
+      "epoch": 0.9964444444444445,
+      "grad_norm": 1.6320134401321411,
+      "learning_rate": 0.00016028469750889682,
+      "loss": 2.7794,
+      "step": 2242
+    },
+    {
+      "epoch": 0.9968888888888889,
+      "grad_norm": 1.804682970046997,
+      "learning_rate": 0.00016026690391459075,
+      "loss": 2.515,
+      "step": 2243
+    },
+    {
+      "epoch": 0.9973333333333333,
+      "grad_norm": 2.0232090950012207,
+      "learning_rate": 0.0001602491103202847,
+      "loss": 3.3673,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9977777777777778,
+      "grad_norm": 1.852048635482788,
+      "learning_rate": 0.00016023131672597866,
+      "loss": 2.7809,
+      "step": 2245
+    },
+    {
+      "epoch": 0.9982222222222222,
+      "grad_norm": 1.7070189714431763,
+      "learning_rate": 0.00016021352313167262,
+      "loss": 2.2775,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9986666666666667,
+      "grad_norm": 1.6251046657562256,
+      "learning_rate": 0.00016019572953736655,
+      "loss": 2.3802,
+      "step": 2247
+    },
+    {
+      "epoch": 0.9991111111111111,
+      "grad_norm": 1.7044910192489624,
+      "learning_rate": 0.0001601779359430605,
+      "loss": 2.5114,
+      "step": 2248
+    },
+    {
+      "epoch": 0.9995555555555555,
+      "grad_norm": 2.653667688369751,
+      "learning_rate": 0.00016016014234875446,
+      "loss": 1.5299,
+      "step": 2249
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5051262378692627,
+      "learning_rate": 0.0001601423487544484,
+      "loss": 1.7248,
+      "step": 2250
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 2.410618305206299,
+      "eval_runtime": 47.6193,
+      "eval_samples_per_second": 10.5,
+      "eval_steps_per_second": 10.5,
+      "step": 2250
+    },
+    {
+      "epoch": 1.0004444444444445,
+      "grad_norm": 0.9968487620353699,
+      "learning_rate": 0.00016012455516014234,
+      "loss": 2.3568,
+      "step": 2251
+    },
+    {
+      "epoch": 1.000888888888889,
+      "grad_norm": 1.1485272645950317,
+      "learning_rate": 0.0001601067615658363,
+      "loss": 2.6016,
+      "step": 2252
+    },
+    {
+      "epoch": 1.0013333333333334,
+      "grad_norm": 0.9918361902236938,
+      "learning_rate": 0.00016008896797153026,
+      "loss": 2.1317,
+      "step": 2253
+    },
+    {
+      "epoch": 1.0017777777777779,
+      "grad_norm": 1.443756103515625,
+      "learning_rate": 0.00016007117437722421,
+      "loss": 1.3992,
+      "step": 2254
+    },
+    {
+      "epoch": 1.0022222222222221,
+      "grad_norm": 1.2466870546340942,
+      "learning_rate": 0.00016005338078291817,
+      "loss": 2.7594,
+      "step": 2255
+    },
+    {
+      "epoch": 1.0026666666666666,
+      "grad_norm": 1.1465460062026978,
+      "learning_rate": 0.0001600355871886121,
+      "loss": 2.3712,
+      "step": 2256
+    },
+    {
+      "epoch": 1.003111111111111,
+      "grad_norm": 1.0530627965927124,
+      "learning_rate": 0.00016001779359430606,
+      "loss": 1.8851,
+      "step": 2257
+    },
+    {
+      "epoch": 1.0035555555555555,
+      "grad_norm": 1.195264458656311,
+      "learning_rate": 0.00016,
+      "loss": 2.2701,
+      "step": 2258
+    },
+    {
+      "epoch": 1.004,
+      "grad_norm": 1.2500135898590088,
+      "learning_rate": 0.00015998220640569397,
+      "loss": 2.167,
+      "step": 2259
+    },
+    {
+      "epoch": 1.0044444444444445,
+      "grad_norm": 1.24211585521698,
+      "learning_rate": 0.0001599644128113879,
+      "loss": 2.3523,
+      "step": 2260
+    },
+    {
+      "epoch": 1.004888888888889,
+      "grad_norm": 1.3692959547042847,
+      "learning_rate": 0.00015994661921708186,
+      "loss": 2.5483,
+      "step": 2261
+    },
+    {
+      "epoch": 1.0053333333333334,
+      "grad_norm": 1.1393674612045288,
+      "learning_rate": 0.0001599288256227758,
+      "loss": 2.3064,
+      "step": 2262
+    },
+    {
+      "epoch": 1.0057777777777779,
+      "grad_norm": 1.341461181640625,
+      "learning_rate": 0.00015991103202846974,
+      "loss": 2.0089,
+      "step": 2263
+    },
+    {
+      "epoch": 1.0062222222222221,
+      "grad_norm": 1.434979796409607,
+      "learning_rate": 0.0001598932384341637,
+      "loss": 1.9006,
+      "step": 2264
+    },
+    {
+      "epoch": 1.0066666666666666,
+      "grad_norm": 1.4398572444915771,
+      "learning_rate": 0.00015987544483985765,
+      "loss": 2.1331,
+      "step": 2265
+    },
+    {
+      "epoch": 1.007111111111111,
+      "grad_norm": 1.2505806684494019,
+      "learning_rate": 0.0001598576512455516,
+      "loss": 2.059,
+      "step": 2266
+    },
+    {
+      "epoch": 1.0075555555555555,
+      "grad_norm": 1.3361196517944336,
+      "learning_rate": 0.00015983985765124557,
+      "loss": 1.2448,
+      "step": 2267
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 1.3557100296020508,
+      "learning_rate": 0.00015982206405693952,
+      "loss": 2.2836,
+      "step": 2268
+    },
+    {
+      "epoch": 1.0084444444444445,
+      "grad_norm": 1.4562923908233643,
+      "learning_rate": 0.00015980427046263345,
+      "loss": 2.4052,
+      "step": 2269
+    },
+    {
+      "epoch": 1.008888888888889,
+      "grad_norm": 1.8324874639511108,
+      "learning_rate": 0.0001597864768683274,
+      "loss": 1.0363,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0093333333333334,
+      "grad_norm": 1.3576056957244873,
+      "learning_rate": 0.00015976868327402137,
+      "loss": 2.5,
+      "step": 2271
+    },
+    {
+      "epoch": 1.0097777777777779,
+      "grad_norm": 1.369577407836914,
+      "learning_rate": 0.00015975088967971532,
+      "loss": 2.4154,
+      "step": 2272
+    },
+    {
+      "epoch": 1.0102222222222221,
+      "grad_norm": 1.7282054424285889,
+      "learning_rate": 0.00015973309608540925,
+      "loss": 2.4238,
+      "step": 2273
+    },
+    {
+      "epoch": 1.0106666666666666,
+      "grad_norm": 1.4440052509307861,
+      "learning_rate": 0.0001597153024911032,
+      "loss": 1.8114,
+      "step": 2274
+    },
+    {
+      "epoch": 1.011111111111111,
+      "grad_norm": 1.8842155933380127,
+      "learning_rate": 0.00015969750889679717,
+      "loss": 2.5752,
+      "step": 2275
+    },
+    {
+      "epoch": 1.0115555555555555,
+      "grad_norm": 1.6671173572540283,
+      "learning_rate": 0.0001596797153024911,
+      "loss": 2.3995,
+      "step": 2276
+    },
+    {
+      "epoch": 1.012,
+      "grad_norm": 1.6246118545532227,
+      "learning_rate": 0.00015966192170818505,
+      "loss": 2.4257,
+      "step": 2277
+    },
+    {
+      "epoch": 1.0124444444444445,
+      "grad_norm": 1.416971206665039,
+      "learning_rate": 0.000159644128113879,
+      "loss": 1.9455,
+      "step": 2278
+    },
+    {
+      "epoch": 1.012888888888889,
+      "grad_norm": 1.754091739654541,
+      "learning_rate": 0.00015962633451957296,
+      "loss": 2.6238,
+      "step": 2279
+    },
+    {
+      "epoch": 1.0133333333333334,
+      "grad_norm": 1.3849114179611206,
+      "learning_rate": 0.00015960854092526692,
+      "loss": 1.5943,
+      "step": 2280
+    },
+    {
+      "epoch": 1.0137777777777779,
+      "grad_norm": 1.6501544713974,
+      "learning_rate": 0.00015959074733096088,
+      "loss": 1.9212,
+      "step": 2281
+    },
+    {
+      "epoch": 1.0142222222222221,
+      "grad_norm": 1.7187528610229492,
+      "learning_rate": 0.0001595729537366548,
+      "loss": 1.7647,
+      "step": 2282
+    },
+    {
+      "epoch": 1.0146666666666666,
+      "grad_norm": 1.8686177730560303,
+      "learning_rate": 0.00015955516014234876,
+      "loss": 2.4814,
+      "step": 2283
+    },
+    {
+      "epoch": 1.015111111111111,
+      "grad_norm": 1.5445845127105713,
+      "learning_rate": 0.00015953736654804272,
+      "loss": 1.9214,
+      "step": 2284
+    },
+    {
+      "epoch": 1.0155555555555555,
+      "grad_norm": 1.7692326307296753,
+      "learning_rate": 0.00015951957295373668,
+      "loss": 2.1594,
+      "step": 2285
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 1.6517410278320312,
+      "learning_rate": 0.0001595017793594306,
+      "loss": 2.1724,
+      "step": 2286
+    },
+    {
+      "epoch": 1.0164444444444445,
+      "grad_norm": 1.2496143579483032,
+      "learning_rate": 0.00015948398576512456,
+      "loss": 0.7354,
+      "step": 2287
+    },
+    {
+      "epoch": 1.016888888888889,
+      "grad_norm": 1.6711918115615845,
+      "learning_rate": 0.00015946619217081852,
+      "loss": 2.1376,
+      "step": 2288
+    },
+    {
+      "epoch": 1.0173333333333334,
+      "grad_norm": 1.8869587182998657,
+      "learning_rate": 0.00015944839857651245,
+      "loss": 2.2878,
+      "step": 2289
+    },
+    {
+      "epoch": 1.0177777777777777,
+      "grad_norm": 1.5782763957977295,
+      "learning_rate": 0.0001594306049822064,
+      "loss": 1.9263,
+      "step": 2290
+    },
+    {
+      "epoch": 1.0182222222222221,
+      "grad_norm": 1.8840655088424683,
+      "learning_rate": 0.00015941281138790036,
+      "loss": 2.4303,
+      "step": 2291
+    },
+    {
+      "epoch": 1.0186666666666666,
+      "grad_norm": 2.064854383468628,
+      "learning_rate": 0.00015939501779359432,
+      "loss": 2.246,
+      "step": 2292
+    },
+    {
+      "epoch": 1.019111111111111,
+      "grad_norm": 1.704014539718628,
+      "learning_rate": 0.00015937722419928827,
+      "loss": 1.6609,
+      "step": 2293
+    },
+    {
+      "epoch": 1.0195555555555555,
+      "grad_norm": 1.7326053380966187,
+      "learning_rate": 0.00015935943060498223,
+      "loss": 2.0684,
+      "step": 2294
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9503422975540161,
+      "learning_rate": 0.00015934163701067616,
+      "loss": 2.1345,
+      "step": 2295
+    },
+    {
+      "epoch": 1.0204444444444445,
+      "grad_norm": 2.1505517959594727,
+      "learning_rate": 0.00015932384341637012,
+      "loss": 2.2523,
+      "step": 2296
+    },
+    {
+      "epoch": 1.020888888888889,
+      "grad_norm": 2.059180736541748,
+      "learning_rate": 0.00015930604982206407,
+      "loss": 1.956,
+      "step": 2297
+    },
+    {
+      "epoch": 1.0213333333333334,
+      "grad_norm": 1.7710636854171753,
+      "learning_rate": 0.000159288256227758,
+      "loss": 1.8129,
+      "step": 2298
+    },
+    {
+      "epoch": 1.0217777777777777,
+      "grad_norm": 1.9192954301834106,
+      "learning_rate": 0.00015927046263345196,
+      "loss": 2.004,
+      "step": 2299
+    },
+    {
+      "epoch": 1.0222222222222221,
+      "grad_norm": 3.243239402770996,
+      "learning_rate": 0.00015925266903914591,
+      "loss": 1.0487,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0226666666666666,
+      "grad_norm": 1.2613039016723633,
+      "learning_rate": 0.00015923487544483987,
+      "loss": 2.2974,
+      "step": 2301
+    },
+    {
+      "epoch": 1.023111111111111,
+      "grad_norm": 1.9154945611953735,
+      "learning_rate": 0.0001592170818505338,
+      "loss": 1.0126,
+      "step": 2302
+    },
+    {
+      "epoch": 1.0235555555555556,
+      "grad_norm": 1.2044062614440918,
+      "learning_rate": 0.00015919928825622776,
+      "loss": 0.5583,
+      "step": 2303
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 1.3576868772506714,
+      "learning_rate": 0.00015918149466192171,
+      "loss": 2.2222,
+      "step": 2304
+    },
+    {
+      "epoch": 1.0244444444444445,
+      "grad_norm": 1.372114658355713,
+      "learning_rate": 0.00015916370106761567,
+      "loss": 2.2053,
+      "step": 2305
+    },
+    {
+      "epoch": 1.024888888888889,
+      "grad_norm": 1.9428937435150146,
+      "learning_rate": 0.00015914590747330963,
+      "loss": 1.9576,
+      "step": 2306
+    },
+    {
+      "epoch": 1.0253333333333334,
+      "grad_norm": 1.2017688751220703,
+      "learning_rate": 0.00015912811387900358,
+      "loss": 2.264,
+      "step": 2307
+    },
+    {
+      "epoch": 1.0257777777777777,
+      "grad_norm": 1.3299150466918945,
+      "learning_rate": 0.0001591103202846975,
+      "loss": 1.8914,
+      "step": 2308
+    },
+    {
+      "epoch": 1.0262222222222221,
+      "grad_norm": 1.3821178674697876,
+      "learning_rate": 0.00015909252669039147,
+      "loss": 2.7482,
+      "step": 2309
+    },
+    {
+      "epoch": 1.0266666666666666,
+      "grad_norm": 1.5535212755203247,
+      "learning_rate": 0.00015907473309608543,
+      "loss": 2.3124,
+      "step": 2310
+    },
+    {
+      "epoch": 1.027111111111111,
+      "grad_norm": 1.2688461542129517,
+      "learning_rate": 0.00015905693950177936,
+      "loss": 2.0699,
+      "step": 2311
+    },
+    {
+      "epoch": 1.0275555555555556,
+      "grad_norm": 1.5910311937332153,
+      "learning_rate": 0.0001590391459074733,
+      "loss": 2.258,
+      "step": 2312
+    },
+    {
+      "epoch": 1.028,
+      "grad_norm": 1.4131362438201904,
+      "learning_rate": 0.00015902135231316727,
+      "loss": 1.8134,
+      "step": 2313
+    },
+    {
+      "epoch": 1.0284444444444445,
+      "grad_norm": 1.4447015523910522,
+      "learning_rate": 0.00015900355871886122,
+      "loss": 1.4502,
+      "step": 2314
+    },
+    {
+      "epoch": 1.028888888888889,
+      "grad_norm": 1.6646491289138794,
+      "learning_rate": 0.00015898576512455515,
+      "loss": 2.3019,
+      "step": 2315
+    },
+    {
+      "epoch": 1.0293333333333334,
+      "grad_norm": 1.4541774988174438,
+      "learning_rate": 0.0001589679715302491,
+      "loss": 2.2436,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0297777777777777,
+      "grad_norm": 1.4342156648635864,
+      "learning_rate": 0.00015895017793594307,
+      "loss": 2.3697,
+      "step": 2317
+    },
+    {
+      "epoch": 1.0302222222222222,
+      "grad_norm": 1.6885074377059937,
+      "learning_rate": 0.00015893238434163702,
+      "loss": 2.3556,
+      "step": 2318
+    },
+    {
+      "epoch": 1.0306666666666666,
+      "grad_norm": 2.0182912349700928,
+      "learning_rate": 0.00015891459074733098,
+      "loss": 2.886,
+      "step": 2319
+    },
+    {
+      "epoch": 1.031111111111111,
+      "grad_norm": 1.5477981567382812,
+      "learning_rate": 0.00015889679715302494,
+      "loss": 2.3829,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0315555555555556,
+      "grad_norm": 1.4278366565704346,
+      "learning_rate": 0.00015887900355871887,
+      "loss": 1.8171,
+      "step": 2321
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 1.5529868602752686,
+      "learning_rate": 0.00015886120996441282,
+      "loss": 1.9937,
+      "step": 2322
+    },
+    {
+      "epoch": 1.0324444444444445,
+      "grad_norm": 1.7762391567230225,
+      "learning_rate": 0.00015884341637010678,
+      "loss": 1.9945,
+      "step": 2323
+    },
+    {
+      "epoch": 1.032888888888889,
+      "grad_norm": 1.378474473953247,
+      "learning_rate": 0.0001588256227758007,
+      "loss": 1.907,
+      "step": 2324
+    },
+    {
+      "epoch": 1.0333333333333334,
+      "grad_norm": 1.6256483793258667,
+      "learning_rate": 0.00015880782918149466,
+      "loss": 2.2908,
+      "step": 2325
+    },
+    {
+      "epoch": 1.0337777777777777,
+      "grad_norm": 1.5568405389785767,
+      "learning_rate": 0.00015879003558718862,
+      "loss": 2.2006,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0342222222222222,
+      "grad_norm": 1.5443711280822754,
+      "learning_rate": 0.00015877224199288258,
+      "loss": 1.892,
+      "step": 2327
+    },
+    {
+      "epoch": 1.0346666666666666,
+      "grad_norm": 1.584693193435669,
+      "learning_rate": 0.0001587544483985765,
+      "loss": 2.2243,
+      "step": 2328
+    },
+    {
+      "epoch": 1.035111111111111,
+      "grad_norm": 1.7238883972167969,
+      "learning_rate": 0.00015873665480427046,
+      "loss": 1.8239,
+      "step": 2329
+    },
+    {
+      "epoch": 1.0355555555555556,
+      "grad_norm": 1.7188372611999512,
+      "learning_rate": 0.00015871886120996442,
+      "loss": 2.1286,
+      "step": 2330
+    },
+    {
+      "epoch": 1.036,
+      "grad_norm": 1.8142226934432983,
+      "learning_rate": 0.00015870106761565838,
+      "loss": 2.342,
+      "step": 2331
+    },
+    {
+      "epoch": 1.0364444444444445,
+      "grad_norm": 1.9462339878082275,
+      "learning_rate": 0.00015868327402135233,
+      "loss": 2.3758,
+      "step": 2332
+    },
+    {
+      "epoch": 1.036888888888889,
+      "grad_norm": 1.4883403778076172,
+      "learning_rate": 0.00015866548042704626,
+      "loss": 1.6063,
+      "step": 2333
+    },
+    {
+      "epoch": 1.0373333333333334,
+      "grad_norm": 1.6989222764968872,
+      "learning_rate": 0.00015864768683274022,
+      "loss": 2.2398,
+      "step": 2334
+    },
+    {
+      "epoch": 1.0377777777777777,
+      "grad_norm": 1.6604561805725098,
+      "learning_rate": 0.00015862989323843418,
+      "loss": 2.2341,
+      "step": 2335
+    },
+    {
+      "epoch": 1.0382222222222222,
+      "grad_norm": 2.0556907653808594,
+      "learning_rate": 0.00015861209964412813,
+      "loss": 2.4467,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0386666666666666,
+      "grad_norm": 2.043485164642334,
+      "learning_rate": 0.00015859430604982206,
+      "loss": 1.7722,
+      "step": 2337
+    },
+    {
+      "epoch": 1.039111111111111,
+      "grad_norm": 1.7736142873764038,
+      "learning_rate": 0.00015857651245551602,
+      "loss": 1.9208,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0395555555555556,
+      "grad_norm": 2.0610883235931396,
+      "learning_rate": 0.00015855871886120997,
+      "loss": 2.8929,
+      "step": 2339
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8629289865493774,
+      "learning_rate": 0.0001585409252669039,
+      "loss": 1.9888,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0404444444444445,
+      "grad_norm": 2.201791286468506,
+      "learning_rate": 0.00015852313167259786,
+      "loss": 2.4039,
+      "step": 2341
+    },
+    {
+      "epoch": 1.040888888888889,
+      "grad_norm": 2.3649518489837646,
+      "learning_rate": 0.00015850533807829182,
+      "loss": 2.342,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0413333333333332,
+      "grad_norm": 1.763653039932251,
+      "learning_rate": 0.00015848754448398577,
+      "loss": 1.7666,
+      "step": 2343
+    },
+    {
+      "epoch": 1.0417777777777777,
+      "grad_norm": 1.721071481704712,
+      "learning_rate": 0.00015846975088967973,
+      "loss": 2.2088,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0422222222222222,
+      "grad_norm": 2.0251195430755615,
+      "learning_rate": 0.00015845195729537369,
+      "loss": 2.3982,
+      "step": 2345
+    },
+    {
+      "epoch": 1.0426666666666666,
+      "grad_norm": 1.797646403312683,
+      "learning_rate": 0.00015843416370106762,
+      "loss": 2.2807,
+      "step": 2346
+    },
+    {
+      "epoch": 1.043111111111111,
+      "grad_norm": 2.336357831954956,
+      "learning_rate": 0.00015841637010676157,
+      "loss": 2.7199,
+      "step": 2347
+    },
+    {
+      "epoch": 1.0435555555555556,
+      "grad_norm": 2.0868053436279297,
+      "learning_rate": 0.00015839857651245553,
+      "loss": 1.9292,
+      "step": 2348
+    },
+    {
+      "epoch": 1.044,
+      "grad_norm": 2.3412818908691406,
+      "learning_rate": 0.00015838078291814949,
+      "loss": 2.1445,
+      "step": 2349
+    },
+    {
+      "epoch": 1.0444444444444445,
+      "grad_norm": 2.016115188598633,
+      "learning_rate": 0.00015836298932384341,
+      "loss": 1.079,
+      "step": 2350
+    },
+    {
+      "epoch": 1.044888888888889,
+      "grad_norm": 1.2892199754714966,
+      "learning_rate": 0.00015834519572953737,
+      "loss": 1.1067,
+      "step": 2351
+    },
+    {
+      "epoch": 1.0453333333333332,
+      "grad_norm": 1.37278151512146,
+      "learning_rate": 0.00015832740213523133,
+      "loss": 2.208,
+      "step": 2352
+    },
+    {
+      "epoch": 1.0457777777777777,
+      "grad_norm": 1.2415770292282104,
+      "learning_rate": 0.00015830960854092526,
+      "loss": 2.2114,
+      "step": 2353
+    },
+    {
+      "epoch": 1.0462222222222222,
+      "grad_norm": 1.232061505317688,
+      "learning_rate": 0.0001582918149466192,
+      "loss": 1.4155,
+      "step": 2354
+    },
+    {
+      "epoch": 1.0466666666666666,
+      "grad_norm": 1.177029013633728,
+      "learning_rate": 0.00015827402135231317,
+      "loss": 2.5071,
+      "step": 2355
+    },
+    {
+      "epoch": 1.047111111111111,
+      "grad_norm": 1.4585161209106445,
+      "learning_rate": 0.00015825622775800713,
+      "loss": 2.3271,
+      "step": 2356
+    },
+    {
+      "epoch": 1.0475555555555556,
+      "grad_norm": 1.424669623374939,
+      "learning_rate": 0.00015823843416370108,
+      "loss": 2.1732,
+      "step": 2357
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 1.3325684070587158,
+      "learning_rate": 0.00015822064056939504,
+      "loss": 2.2668,
+      "step": 2358
+    },
+    {
+      "epoch": 1.0484444444444445,
+      "grad_norm": 1.5595697164535522,
+      "learning_rate": 0.00015820284697508897,
+      "loss": 2.3379,
+      "step": 2359
+    },
+    {
+      "epoch": 1.048888888888889,
+      "grad_norm": 1.6011629104614258,
+      "learning_rate": 0.00015818505338078293,
+      "loss": 2.8927,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0493333333333332,
+      "grad_norm": 1.100995421409607,
+      "learning_rate": 0.00015816725978647688,
+      "loss": 1.6169,
+      "step": 2361
+    },
+    {
+      "epoch": 1.0497777777777777,
+      "grad_norm": 1.3697389364242554,
+      "learning_rate": 0.00015814946619217084,
+      "loss": 2.3766,
+      "step": 2362
+    },
+    {
+      "epoch": 1.0502222222222222,
+      "grad_norm": 1.332924485206604,
+      "learning_rate": 0.00015813167259786477,
+      "loss": 1.7657,
+      "step": 2363
+    },
+    {
+      "epoch": 1.0506666666666666,
+      "grad_norm": 1.3072422742843628,
+      "learning_rate": 0.00015811387900355872,
+      "loss": 2.0262,
+      "step": 2364
+    },
+    {
+      "epoch": 1.051111111111111,
+      "grad_norm": 1.370421290397644,
+      "learning_rate": 0.00015809608540925268,
+      "loss": 1.8601,
+      "step": 2365
+    },
+    {
+      "epoch": 1.0515555555555556,
+      "grad_norm": 1.580460786819458,
+      "learning_rate": 0.0001580782918149466,
+      "loss": 1.3883,
+      "step": 2366
+    },
+    {
+      "epoch": 1.052,
+      "grad_norm": 1.478049635887146,
+      "learning_rate": 0.00015806049822064057,
+      "loss": 2.3788,
+      "step": 2367
+    },
+    {
+      "epoch": 1.0524444444444445,
+      "grad_norm": 1.421947717666626,
+      "learning_rate": 0.00015804270462633452,
+      "loss": 1.4549,
+      "step": 2368
+    },
+    {
+      "epoch": 1.052888888888889,
+      "grad_norm": 1.768334984779358,
+      "learning_rate": 0.00015802491103202848,
+      "loss": 1.0313,
+      "step": 2369
+    },
+    {
+      "epoch": 1.0533333333333332,
+      "grad_norm": 1.5838056802749634,
+      "learning_rate": 0.00015800711743772244,
+      "loss": 2.3386,
+      "step": 2370
+    },
+    {
+      "epoch": 1.0537777777777777,
+      "grad_norm": 1.6991932392120361,
+      "learning_rate": 0.0001579893238434164,
+      "loss": 2.5764,
+      "step": 2371
+    },
+    {
+      "epoch": 1.0542222222222222,
+      "grad_norm": 1.4423344135284424,
+      "learning_rate": 0.00015797153024911032,
+      "loss": 1.9598,
+      "step": 2372
+    },
+    {
+      "epoch": 1.0546666666666666,
+      "grad_norm": 1.508663535118103,
+      "learning_rate": 0.00015795373665480428,
+      "loss": 2.3974,
+      "step": 2373
+    },
+    {
+      "epoch": 1.055111111111111,
+      "grad_norm": 1.6960604190826416,
+      "learning_rate": 0.00015793594306049823,
+      "loss": 2.2544,
+      "step": 2374
+    },
+    {
+      "epoch": 1.0555555555555556,
+      "grad_norm": 1.786561131477356,
+      "learning_rate": 0.0001579181494661922,
+      "loss": 1.688,
+      "step": 2375
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 1.505338191986084,
+      "learning_rate": 0.00015790035587188612,
+      "loss": 2.2269,
+      "step": 2376
+    },
+    {
+      "epoch": 1.0564444444444445,
+      "grad_norm": 1.4534296989440918,
+      "learning_rate": 0.00015788256227758008,
+      "loss": 1.9174,
+      "step": 2377
+    },
+    {
+      "epoch": 1.056888888888889,
+      "grad_norm": 1.5155029296875,
+      "learning_rate": 0.00015786476868327403,
+      "loss": 2.2544,
+      "step": 2378
+    },
+    {
+      "epoch": 1.0573333333333332,
+      "grad_norm": 1.8670555353164673,
+      "learning_rate": 0.00015784697508896796,
+      "loss": 2.4167,
+      "step": 2379
+    },
+    {
+      "epoch": 1.0577777777777777,
+      "grad_norm": 1.633664846420288,
+      "learning_rate": 0.00015782918149466192,
+      "loss": 1.8265,
+      "step": 2380
+    },
+    {
+      "epoch": 1.0582222222222222,
+      "grad_norm": 1.5484849214553833,
+      "learning_rate": 0.00015781138790035588,
+      "loss": 2.1692,
+      "step": 2381
+    },
+    {
+      "epoch": 1.0586666666666666,
+      "grad_norm": 1.8765406608581543,
+      "learning_rate": 0.00015779359430604983,
+      "loss": 0.8998,
+      "step": 2382
+    },
+    {
+      "epoch": 1.0591111111111111,
+      "grad_norm": 1.6616917848587036,
+      "learning_rate": 0.0001577758007117438,
+      "loss": 2.3774,
+      "step": 2383
+    },
+    {
+      "epoch": 1.0595555555555556,
+      "grad_norm": 1.5309672355651855,
+      "learning_rate": 0.00015775800711743775,
+      "loss": 1.8152,
+      "step": 2384
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7901145219802856,
+      "learning_rate": 0.00015774021352313168,
+      "loss": 2.6988,
+      "step": 2385
+    },
+    {
+      "epoch": 1.0604444444444445,
+      "grad_norm": 1.5736534595489502,
+      "learning_rate": 0.00015772241992882563,
+      "loss": 1.9631,
+      "step": 2386
+    },
+    {
+      "epoch": 1.060888888888889,
+      "grad_norm": 1.8793672323226929,
+      "learning_rate": 0.0001577046263345196,
+      "loss": 2.7282,
+      "step": 2387
+    },
+    {
+      "epoch": 1.0613333333333332,
+      "grad_norm": 1.6914716958999634,
+      "learning_rate": 0.00015768683274021354,
+      "loss": 2.1709,
+      "step": 2388
+    },
+    {
+      "epoch": 1.0617777777777777,
+      "grad_norm": 1.847061038017273,
+      "learning_rate": 0.00015766903914590747,
+      "loss": 2.1721,
+      "step": 2389
+    },
+    {
+      "epoch": 1.0622222222222222,
+      "grad_norm": 1.698413610458374,
+      "learning_rate": 0.00015765124555160143,
+      "loss": 1.6681,
+      "step": 2390
+    },
+    {
+      "epoch": 1.0626666666666666,
+      "grad_norm": 1.9005299806594849,
+      "learning_rate": 0.0001576334519572954,
+      "loss": 1.9767,
+      "step": 2391
+    },
+    {
+      "epoch": 1.0631111111111111,
+      "grad_norm": 1.9315385818481445,
+      "learning_rate": 0.00015761565836298932,
+      "loss": 2.4512,
+      "step": 2392
+    },
+    {
+      "epoch": 1.0635555555555556,
+      "grad_norm": 1.114691972732544,
+      "learning_rate": 0.00015759786476868327,
+      "loss": 0.8507,
+      "step": 2393
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 1.4050822257995605,
+      "learning_rate": 0.00015758007117437723,
+      "loss": 1.3638,
+      "step": 2394
+    },
+    {
+      "epoch": 1.0644444444444445,
+      "grad_norm": 1.3316043615341187,
+      "learning_rate": 0.00015756227758007119,
+      "loss": 1.1736,
+      "step": 2395
+    },
+    {
+      "epoch": 1.064888888888889,
+      "grad_norm": 1.950830340385437,
+      "learning_rate": 0.00015754448398576514,
+      "loss": 2.0624,
+      "step": 2396
+    },
+    {
+      "epoch": 1.0653333333333332,
+      "grad_norm": 1.9094411134719849,
+      "learning_rate": 0.0001575266903914591,
+      "loss": 2.4114,
+      "step": 2397
+    },
+    {
+      "epoch": 1.0657777777777777,
+      "grad_norm": 2.046294927597046,
+      "learning_rate": 0.00015750889679715303,
+      "loss": 2.32,
+      "step": 2398
+    },
+    {
+      "epoch": 1.0662222222222222,
+      "grad_norm": 2.1060452461242676,
+      "learning_rate": 0.00015749110320284698,
+      "loss": 2.7453,
+      "step": 2399
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 3.819446325302124,
+      "learning_rate": 0.00015747330960854094,
+      "loss": 0.8665,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0671111111111111,
+      "grad_norm": 1.0843795537948608,
+      "learning_rate": 0.0001574555160142349,
+      "loss": 2.5427,
+      "step": 2401
+    },
+    {
+      "epoch": 1.0675555555555556,
+      "grad_norm": 1.219653606414795,
+      "learning_rate": 0.00015743772241992883,
+      "loss": 0.8994,
+      "step": 2402
+    },
+    {
+      "epoch": 1.068,
+      "grad_norm": 1.363571047782898,
+      "learning_rate": 0.00015741992882562278,
+      "loss": 2.6447,
+      "step": 2403
+    },
+    {
+      "epoch": 1.0684444444444445,
+      "grad_norm": 1.192108154296875,
+      "learning_rate": 0.00015740213523131674,
+      "loss": 2.1144,
+      "step": 2404
+    },
+    {
+      "epoch": 1.068888888888889,
+      "grad_norm": 1.539057731628418,
+      "learning_rate": 0.00015738434163701067,
+      "loss": 2.487,
+      "step": 2405
+    },
+    {
+      "epoch": 1.0693333333333332,
+      "grad_norm": 1.3476603031158447,
+      "learning_rate": 0.00015736654804270463,
+      "loss": 2.207,
+      "step": 2406
+    },
+    {
+      "epoch": 1.0697777777777777,
+      "grad_norm": 1.304561972618103,
+      "learning_rate": 0.00015734875444839858,
+      "loss": 2.0535,
+      "step": 2407
+    },
+    {
+      "epoch": 1.0702222222222222,
+      "grad_norm": 1.4313247203826904,
+      "learning_rate": 0.00015733096085409254,
+      "loss": 2.5509,
+      "step": 2408
+    },
+    {
+      "epoch": 1.0706666666666667,
+      "grad_norm": 1.3336185216903687,
+      "learning_rate": 0.0001573131672597865,
+      "loss": 1.8019,
+      "step": 2409
+    },
+    {
+      "epoch": 1.0711111111111111,
+      "grad_norm": 1.3517190217971802,
+      "learning_rate": 0.00015729537366548045,
+      "loss": 2.188,
+      "step": 2410
+    },
+    {
+      "epoch": 1.0715555555555556,
+      "grad_norm": 1.563821792602539,
+      "learning_rate": 0.00015727758007117438,
+      "loss": 2.2798,
+      "step": 2411
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 1.5215498208999634,
+      "learning_rate": 0.00015725978647686834,
+      "loss": 2.454,
+      "step": 2412
+    },
+    {
+      "epoch": 1.0724444444444445,
+      "grad_norm": 1.465469479560852,
+      "learning_rate": 0.0001572419928825623,
+      "loss": 1.8113,
+      "step": 2413
+    },
+    {
+      "epoch": 1.072888888888889,
+      "grad_norm": 1.5329351425170898,
+      "learning_rate": 0.00015722419928825622,
+      "loss": 2.3624,
+      "step": 2414
+    },
+    {
+      "epoch": 1.0733333333333333,
+      "grad_norm": 1.7038596868515015,
+      "learning_rate": 0.00015720640569395018,
+      "loss": 0.1039,
+      "step": 2415
+    },
+    {
+      "epoch": 1.0737777777777777,
+      "grad_norm": 1.4838898181915283,
+      "learning_rate": 0.00015718861209964414,
+      "loss": 1.8683,
+      "step": 2416
+    },
+    {
+      "epoch": 1.0742222222222222,
+      "grad_norm": 1.7675329446792603,
+      "learning_rate": 0.0001571708185053381,
+      "loss": 2.3259,
+      "step": 2417
+    },
+    {
+      "epoch": 1.0746666666666667,
+      "grad_norm": 1.687468409538269,
+      "learning_rate": 0.00015715302491103202,
+      "loss": 2.6304,
+      "step": 2418
+    },
+    {
+      "epoch": 1.0751111111111111,
+      "grad_norm": 1.5204507112503052,
+      "learning_rate": 0.00015713523131672598,
+      "loss": 2.5227,
+      "step": 2419
+    },
+    {
+      "epoch": 1.0755555555555556,
+      "grad_norm": 2.2087671756744385,
+      "learning_rate": 0.00015711743772241994,
+      "loss": 1.0911,
+      "step": 2420
+    },
+    {
+      "epoch": 1.076,
+      "grad_norm": 1.5000768899917603,
+      "learning_rate": 0.0001570996441281139,
+      "loss": 2.3532,
+      "step": 2421
+    },
+    {
+      "epoch": 1.0764444444444445,
+      "grad_norm": 1.5198241472244263,
+      "learning_rate": 0.00015708185053380785,
+      "loss": 2.0702,
+      "step": 2422
+    },
+    {
+      "epoch": 1.076888888888889,
+      "grad_norm": 1.6272002458572388,
+      "learning_rate": 0.00015706405693950178,
+      "loss": 2.0378,
+      "step": 2423
+    },
+    {
+      "epoch": 1.0773333333333333,
+      "grad_norm": 1.5990360975265503,
+      "learning_rate": 0.00015704626334519573,
+      "loss": 2.2515,
+      "step": 2424
+    },
+    {
+      "epoch": 1.0777777777777777,
+      "grad_norm": 1.4444339275360107,
+      "learning_rate": 0.0001570284697508897,
+      "loss": 2.2619,
+      "step": 2425
+    },
+    {
+      "epoch": 1.0782222222222222,
+      "grad_norm": 1.8956879377365112,
+      "learning_rate": 0.00015701067615658365,
+      "loss": 2.8021,
+      "step": 2426
+    },
+    {
+      "epoch": 1.0786666666666667,
+      "grad_norm": 1.6425714492797852,
+      "learning_rate": 0.00015699288256227758,
+      "loss": 2.0961,
+      "step": 2427
+    },
+    {
+      "epoch": 1.0791111111111111,
+      "grad_norm": 1.7819446325302124,
+      "learning_rate": 0.00015697508896797153,
+      "loss": 2.2295,
+      "step": 2428
+    },
+    {
+      "epoch": 1.0795555555555556,
+      "grad_norm": 1.8124161958694458,
+      "learning_rate": 0.0001569572953736655,
+      "loss": 2.3429,
+      "step": 2429
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7560713291168213,
+      "learning_rate": 0.00015693950177935942,
+      "loss": 2.0773,
+      "step": 2430
+    },
+    {
+      "epoch": 1.0804444444444445,
+      "grad_norm": 1.647606611251831,
+      "learning_rate": 0.00015692170818505338,
+      "loss": 1.9757,
+      "step": 2431
+    },
+    {
+      "epoch": 1.0808888888888888,
+      "grad_norm": 1.6525201797485352,
+      "learning_rate": 0.00015690391459074733,
+      "loss": 1.8019,
+      "step": 2432
+    },
+    {
+      "epoch": 1.0813333333333333,
+      "grad_norm": 1.7084051370620728,
+      "learning_rate": 0.0001568861209964413,
+      "loss": 2.6304,
+      "step": 2433
+    },
+    {
+      "epoch": 1.0817777777777777,
+      "grad_norm": 1.3071404695510864,
+      "learning_rate": 0.00015686832740213525,
+      "loss": 1.1213,
+      "step": 2434
+    },
+    {
+      "epoch": 1.0822222222222222,
+      "grad_norm": 1.7664408683776855,
+      "learning_rate": 0.0001568505338078292,
+      "loss": 2.335,
+      "step": 2435
+    },
+    {
+      "epoch": 1.0826666666666667,
+      "grad_norm": 1.7795616388320923,
+      "learning_rate": 0.00015683274021352313,
+      "loss": 1.8807,
+      "step": 2436
+    },
+    {
+      "epoch": 1.0831111111111111,
+      "grad_norm": 1.9509518146514893,
+      "learning_rate": 0.0001568149466192171,
+      "loss": 2.2775,
+      "step": 2437
+    },
+    {
+      "epoch": 1.0835555555555556,
+      "grad_norm": 1.7835257053375244,
+      "learning_rate": 0.00015679715302491104,
+      "loss": 1.8702,
+      "step": 2438
+    },
+    {
+      "epoch": 1.084,
+      "grad_norm": 1.7957788705825806,
+      "learning_rate": 0.000156779359430605,
+      "loss": 2.4921,
+      "step": 2439
+    },
+    {
+      "epoch": 1.0844444444444445,
+      "grad_norm": 1.2208243608474731,
+      "learning_rate": 0.00015676156583629893,
+      "loss": 1.2614,
+      "step": 2440
+    },
+    {
+      "epoch": 1.0848888888888888,
+      "grad_norm": 1.8217169046401978,
+      "learning_rate": 0.0001567437722419929,
+      "loss": 2.0923,
+      "step": 2441
+    },
+    {
+      "epoch": 1.0853333333333333,
+      "grad_norm": 2.509866952896118,
+      "learning_rate": 0.00015672597864768684,
+      "loss": 2.5114,
+      "step": 2442
+    },
+    {
+      "epoch": 1.0857777777777777,
+      "grad_norm": 2.0780751705169678,
+      "learning_rate": 0.00015670818505338077,
+      "loss": 2.5196,
+      "step": 2443
+    },
+    {
+      "epoch": 1.0862222222222222,
+      "grad_norm": 1.780432105064392,
+      "learning_rate": 0.00015669039145907473,
+      "loss": 1.6573,
+      "step": 2444
+    },
+    {
+      "epoch": 1.0866666666666667,
+      "grad_norm": 1.7413227558135986,
+      "learning_rate": 0.00015667259786476869,
+      "loss": 1.6994,
+      "step": 2445
+    },
+    {
+      "epoch": 1.0871111111111111,
+      "grad_norm": 2.0534093379974365,
+      "learning_rate": 0.00015665480427046264,
+      "loss": 2.5252,
+      "step": 2446
+    },
+    {
+      "epoch": 1.0875555555555556,
+      "grad_norm": 1.8891476392745972,
+      "learning_rate": 0.0001566370106761566,
+      "loss": 1.7454,
+      "step": 2447
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 2.5236616134643555,
+      "learning_rate": 0.00015661921708185056,
+      "loss": 2.7,
+      "step": 2448
+    },
+    {
+      "epoch": 1.0884444444444445,
+      "grad_norm": 2.130950689315796,
+      "learning_rate": 0.00015660142348754448,
+      "loss": 1.9737,
+      "step": 2449
+    },
+    {
+      "epoch": 1.0888888888888888,
+      "grad_norm": 3.0445713996887207,
+      "learning_rate": 0.00015658362989323844,
+      "loss": 1.5697,
+      "step": 2450
+    },
+    {
+      "epoch": 1.0893333333333333,
+      "grad_norm": 1.1887680292129517,
+      "learning_rate": 0.0001565658362989324,
+      "loss": 2.4287,
+      "step": 2451
+    },
+    {
+      "epoch": 1.0897777777777777,
+      "grad_norm": 1.3546632528305054,
+      "learning_rate": 0.00015654804270462635,
+      "loss": 1.377,
+      "step": 2452
+    },
+    {
+      "epoch": 1.0902222222222222,
+      "grad_norm": 1.6304256916046143,
+      "learning_rate": 0.00015653024911032028,
+      "loss": 2.4908,
+      "step": 2453
+    },
+    {
+      "epoch": 1.0906666666666667,
+      "grad_norm": 1.516430139541626,
+      "learning_rate": 0.00015651245551601424,
+      "loss": 2.0657,
+      "step": 2454
+    },
+    {
+      "epoch": 1.0911111111111111,
+      "grad_norm": 1.5727593898773193,
+      "learning_rate": 0.0001564946619217082,
+      "loss": 2.0812,
+      "step": 2455
+    },
+    {
+      "epoch": 1.0915555555555556,
+      "grad_norm": 1.2789214849472046,
+      "learning_rate": 0.00015647686832740213,
+      "loss": 2.01,
+      "step": 2456
+    },
+    {
+      "epoch": 1.092,
+      "grad_norm": 1.4972316026687622,
+      "learning_rate": 0.00015645907473309608,
+      "loss": 1.0983,
+      "step": 2457
+    },
+    {
+      "epoch": 1.0924444444444443,
+      "grad_norm": 1.3972692489624023,
+      "learning_rate": 0.00015644128113879004,
+      "loss": 2.2204,
+      "step": 2458
+    },
+    {
+      "epoch": 1.0928888888888888,
+      "grad_norm": 1.5354390144348145,
+      "learning_rate": 0.000156423487544484,
+      "loss": 2.3677,
+      "step": 2459
+    },
+    {
+      "epoch": 1.0933333333333333,
+      "grad_norm": 1.5079275369644165,
+      "learning_rate": 0.00015640569395017795,
+      "loss": 2.4312,
+      "step": 2460
+    },
+    {
+      "epoch": 1.0937777777777777,
+      "grad_norm": 1.4273076057434082,
+      "learning_rate": 0.0001563879003558719,
+      "loss": 2.1158,
+      "step": 2461
+    },
+    {
+      "epoch": 1.0942222222222222,
+      "grad_norm": 1.5340080261230469,
+      "learning_rate": 0.00015637010676156584,
+      "loss": 2.559,
+      "step": 2462
+    },
+    {
+      "epoch": 1.0946666666666667,
+      "grad_norm": 1.5617725849151611,
+      "learning_rate": 0.0001563523131672598,
+      "loss": 2.5294,
+      "step": 2463
+    },
+    {
+      "epoch": 1.0951111111111111,
+      "grad_norm": 1.6314741373062134,
+      "learning_rate": 0.00015633451957295375,
+      "loss": 2.1713,
+      "step": 2464
+    },
+    {
+      "epoch": 1.0955555555555556,
+      "grad_norm": 1.460752010345459,
+      "learning_rate": 0.0001563167259786477,
+      "loss": 2.3558,
+      "step": 2465
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 1.428756833076477,
+      "learning_rate": 0.00015629893238434164,
+      "loss": 2.2738,
+      "step": 2466
+    },
+    {
+      "epoch": 1.0964444444444443,
+      "grad_norm": 1.7158453464508057,
+      "learning_rate": 0.0001562811387900356,
+      "loss": 2.172,
+      "step": 2467
+    },
+    {
+      "epoch": 1.0968888888888888,
+      "grad_norm": 1.5051125288009644,
+      "learning_rate": 0.00015626334519572955,
+      "loss": 1.9741,
+      "step": 2468
+    },
+    {
+      "epoch": 1.0973333333333333,
+      "grad_norm": 1.4402563571929932,
+      "learning_rate": 0.00015624555160142348,
+      "loss": 2.2152,
+      "step": 2469
+    },
+    {
+      "epoch": 1.0977777777777777,
+      "grad_norm": 1.6247109174728394,
+      "learning_rate": 0.00015622775800711744,
+      "loss": 2.1543,
+      "step": 2470
+    },
+    {
+      "epoch": 1.0982222222222222,
+      "grad_norm": 1.6644169092178345,
+      "learning_rate": 0.0001562099644128114,
+      "loss": 2.2489,
+      "step": 2471
+    },
+    {
+      "epoch": 1.0986666666666667,
+      "grad_norm": 2.04425048828125,
+      "learning_rate": 0.00015619217081850535,
+      "loss": 2.2436,
+      "step": 2472
+    },
+    {
+      "epoch": 1.0991111111111111,
+      "grad_norm": 1.86391282081604,
+      "learning_rate": 0.0001561743772241993,
+      "loss": 1.8856,
+      "step": 2473
+    },
+    {
+      "epoch": 1.0995555555555556,
+      "grad_norm": 1.5092231035232544,
+      "learning_rate": 0.00015615658362989326,
+      "loss": 2.0027,
+      "step": 2474
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.5313433408737183,
+      "learning_rate": 0.0001561387900355872,
+      "loss": 1.4636,
+      "step": 2475
+    },
+    {
+      "epoch": 1.1004444444444443,
+      "grad_norm": 1.7613354921340942,
+      "learning_rate": 0.00015612099644128115,
+      "loss": 2.3526,
+      "step": 2476
+    },
+    {
+      "epoch": 1.1008888888888888,
+      "grad_norm": 1.567148208618164,
+      "learning_rate": 0.0001561032028469751,
+      "loss": 1.8585,
+      "step": 2477
+    },
+    {
+      "epoch": 1.1013333333333333,
+      "grad_norm": 2.1191651821136475,
+      "learning_rate": 0.00015608540925266906,
+      "loss": 2.188,
+      "step": 2478
+    },
+    {
+      "epoch": 1.1017777777777777,
+      "grad_norm": 1.6111822128295898,
+      "learning_rate": 0.000156067615658363,
+      "loss": 2.0623,
+      "step": 2479
+    },
+    {
+      "epoch": 1.1022222222222222,
+      "grad_norm": 1.5612345933914185,
+      "learning_rate": 0.00015604982206405695,
+      "loss": 1.881,
+      "step": 2480
+    },
+    {
+      "epoch": 1.1026666666666667,
+      "grad_norm": 1.9181139469146729,
+      "learning_rate": 0.0001560320284697509,
+      "loss": 2.4053,
+      "step": 2481
+    },
+    {
+      "epoch": 1.1031111111111112,
+      "grad_norm": 1.5462487936019897,
+      "learning_rate": 0.00015601423487544483,
+      "loss": 1.7245,
+      "step": 2482
+    },
+    {
+      "epoch": 1.1035555555555556,
+      "grad_norm": 1.9112005233764648,
+      "learning_rate": 0.0001559964412811388,
+      "loss": 2.0651,
+      "step": 2483
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 1.8536262512207031,
+      "learning_rate": 0.00015597864768683274,
+      "loss": 2.2609,
+      "step": 2484
+    },
+    {
+      "epoch": 1.1044444444444443,
+      "grad_norm": 1.6455966234207153,
+      "learning_rate": 0.0001559608540925267,
+      "loss": 1.8761,
+      "step": 2485
+    },
+    {
+      "epoch": 1.1048888888888888,
+      "grad_norm": 1.9066351652145386,
+      "learning_rate": 0.00015594306049822066,
+      "loss": 1.9272,
+      "step": 2486
+    },
+    {
+      "epoch": 1.1053333333333333,
+      "grad_norm": 1.5465588569641113,
+      "learning_rate": 0.00015592526690391461,
+      "loss": 1.6961,
+      "step": 2487
+    },
+    {
+      "epoch": 1.1057777777777777,
+      "grad_norm": 2.180607318878174,
+      "learning_rate": 0.00015590747330960854,
+      "loss": 2.4785,
+      "step": 2488
+    },
+    {
+      "epoch": 1.1062222222222222,
+      "grad_norm": 1.9382919073104858,
+      "learning_rate": 0.0001558896797153025,
+      "loss": 1.754,
+      "step": 2489
+    },
+    {
+      "epoch": 1.1066666666666667,
+      "grad_norm": 1.8899612426757812,
+      "learning_rate": 0.00015587188612099646,
+      "loss": 2.2687,
+      "step": 2490
+    },
+    {
+      "epoch": 1.1071111111111112,
+      "grad_norm": 1.633954644203186,
+      "learning_rate": 0.0001558540925266904,
+      "loss": 1.9051,
+      "step": 2491
+    },
+    {
+      "epoch": 1.1075555555555556,
+      "grad_norm": 3.747358560562134,
+      "learning_rate": 0.00015583629893238434,
+      "loss": 1.9931,
+      "step": 2492
+    },
+    {
+      "epoch": 1.108,
+      "grad_norm": 2.0736889839172363,
+      "learning_rate": 0.0001558185053380783,
+      "loss": 2.3253,
+      "step": 2493
+    },
+    {
+      "epoch": 1.1084444444444443,
+      "grad_norm": 1.9722511768341064,
+      "learning_rate": 0.00015580071174377226,
+      "loss": 2.2268,
+      "step": 2494
+    },
+    {
+      "epoch": 1.1088888888888888,
+      "grad_norm": 1.9055894613265991,
+      "learning_rate": 0.00015578291814946619,
+      "loss": 2.0168,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1093333333333333,
+      "grad_norm": 2.424161672592163,
+      "learning_rate": 0.00015576512455516014,
+      "loss": 2.4325,
+      "step": 2496
+    },
+    {
+      "epoch": 1.1097777777777778,
+      "grad_norm": 1.9778692722320557,
+      "learning_rate": 0.0001557473309608541,
+      "loss": 1.8312,
+      "step": 2497
+    },
+    {
+      "epoch": 1.1102222222222222,
+      "grad_norm": 2.590223550796509,
+      "learning_rate": 0.00015572953736654805,
+      "loss": 2.6732,
+      "step": 2498
+    },
+    {
+      "epoch": 1.1106666666666667,
+      "grad_norm": 1.8567228317260742,
+      "learning_rate": 0.000155711743772242,
+      "loss": 0.7758,
+      "step": 2499
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 2.436675786972046,
+      "learning_rate": 0.00015569395017793597,
+      "loss": 2.0896,
+      "step": 2500
+    },
+    {
+      "epoch": 1.1115555555555556,
+      "grad_norm": 1.4165303707122803,
+      "learning_rate": 0.0001556761565836299,
+      "loss": 2.3637,
+      "step": 2501
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 1.173225998878479,
+      "learning_rate": 0.00015565836298932385,
+      "loss": 2.1477,
+      "step": 2502
+    },
+    {
+      "epoch": 1.1124444444444443,
+      "grad_norm": 1.1812031269073486,
+      "learning_rate": 0.0001556405693950178,
+      "loss": 2.5036,
+      "step": 2503
+    },
+    {
+      "epoch": 1.1128888888888888,
+      "grad_norm": 1.4120737314224243,
+      "learning_rate": 0.00015562277580071177,
+      "loss": 2.1574,
+      "step": 2504
+    },
+    {
+      "epoch": 1.1133333333333333,
+      "grad_norm": 1.3986128568649292,
+      "learning_rate": 0.0001556049822064057,
+      "loss": 1.8443,
+      "step": 2505
+    },
+    {
+      "epoch": 1.1137777777777778,
+      "grad_norm": 1.6244175434112549,
+      "learning_rate": 0.00015558718861209965,
+      "loss": 2.7789,
+      "step": 2506
+    },
+    {
+      "epoch": 1.1142222222222222,
+      "grad_norm": 1.4272732734680176,
+      "learning_rate": 0.0001555693950177936,
+      "loss": 2.2718,
+      "step": 2507
+    },
+    {
+      "epoch": 1.1146666666666667,
+      "grad_norm": 1.3611066341400146,
+      "learning_rate": 0.00015555160142348754,
+      "loss": 2.3992,
+      "step": 2508
+    },
+    {
+      "epoch": 1.1151111111111112,
+      "grad_norm": 1.751434087753296,
+      "learning_rate": 0.0001555338078291815,
+      "loss": 2.1827,
+      "step": 2509
+    },
+    {
+      "epoch": 1.1155555555555556,
+      "grad_norm": 1.3341114521026611,
+      "learning_rate": 0.00015551601423487545,
+      "loss": 1.985,
+      "step": 2510
+    },
+    {
+      "epoch": 1.116,
+      "grad_norm": 1.3708674907684326,
+      "learning_rate": 0.0001554982206405694,
+      "loss": 2.0507,
+      "step": 2511
+    },
+    {
+      "epoch": 1.1164444444444444,
+      "grad_norm": 1.4740220308303833,
+      "learning_rate": 0.00015548042704626336,
+      "loss": 2.1808,
+      "step": 2512
+    },
+    {
+      "epoch": 1.1168888888888888,
+      "grad_norm": 1.5105361938476562,
+      "learning_rate": 0.0001554626334519573,
+      "loss": 2.055,
+      "step": 2513
+    },
+    {
+      "epoch": 1.1173333333333333,
+      "grad_norm": 1.4184283018112183,
+      "learning_rate": 0.00015544483985765125,
+      "loss": 2.0523,
+      "step": 2514
+    },
+    {
+      "epoch": 1.1177777777777778,
+      "grad_norm": 1.5358822345733643,
+      "learning_rate": 0.0001554270462633452,
+      "loss": 2.073,
+      "step": 2515
+    },
+    {
+      "epoch": 1.1182222222222222,
+      "grad_norm": 1.636608362197876,
+      "learning_rate": 0.00015540925266903916,
+      "loss": 2.0907,
+      "step": 2516
+    },
+    {
+      "epoch": 1.1186666666666667,
+      "grad_norm": 1.7694652080535889,
+      "learning_rate": 0.00015539145907473312,
+      "loss": 2.4836,
+      "step": 2517
+    },
+    {
+      "epoch": 1.1191111111111112,
+      "grad_norm": 1.6159368753433228,
+      "learning_rate": 0.00015537366548042705,
+      "loss": 1.9878,
+      "step": 2518
+    },
+    {
+      "epoch": 1.1195555555555556,
+      "grad_norm": 2.006478786468506,
+      "learning_rate": 0.000155355871886121,
+      "loss": 2.6112,
+      "step": 2519
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.530470371246338,
+      "learning_rate": 0.00015533807829181493,
+      "loss": 1.8775,
+      "step": 2520
+    },
+    {
+      "epoch": 1.1204444444444444,
+      "grad_norm": 1.8092018365859985,
+      "learning_rate": 0.0001553202846975089,
+      "loss": 2.0667,
+      "step": 2521
+    },
+    {
+      "epoch": 1.1208888888888888,
+      "grad_norm": 1.5595135688781738,
+      "learning_rate": 0.00015530249110320285,
+      "loss": 2.7584,
+      "step": 2522
+    },
+    {
+      "epoch": 1.1213333333333333,
+      "grad_norm": 1.5130575895309448,
+      "learning_rate": 0.0001552846975088968,
+      "loss": 2.3097,
+      "step": 2523
+    },
+    {
+      "epoch": 1.1217777777777778,
+      "grad_norm": 1.7195639610290527,
+      "learning_rate": 0.00015526690391459076,
+      "loss": 2.3219,
+      "step": 2524
+    },
+    {
+      "epoch": 1.1222222222222222,
+      "grad_norm": 1.7365368604660034,
+      "learning_rate": 0.00015524911032028472,
+      "loss": 1.8011,
+      "step": 2525
+    },
+    {
+      "epoch": 1.1226666666666667,
+      "grad_norm": 1.6279082298278809,
+      "learning_rate": 0.00015523131672597865,
+      "loss": 1.9524,
+      "step": 2526
+    },
+    {
+      "epoch": 1.1231111111111112,
+      "grad_norm": 1.9169297218322754,
+      "learning_rate": 0.0001552135231316726,
+      "loss": 2.3763,
+      "step": 2527
+    },
+    {
+      "epoch": 1.1235555555555556,
+      "grad_norm": 1.688704490661621,
+      "learning_rate": 0.00015519572953736656,
+      "loss": 2.1569,
+      "step": 2528
+    },
+    {
+      "epoch": 1.124,
+      "grad_norm": 1.2597405910491943,
+      "learning_rate": 0.00015517793594306052,
+      "loss": 1.3028,
+      "step": 2529
+    },
+    {
+      "epoch": 1.1244444444444444,
+      "grad_norm": 1.07369863986969,
+      "learning_rate": 0.00015516014234875445,
+      "loss": 0.8657,
+      "step": 2530
+    },
+    {
+      "epoch": 1.1248888888888888,
+      "grad_norm": 1.3311814069747925,
+      "learning_rate": 0.0001551423487544484,
+      "loss": 1.1018,
+      "step": 2531
+    },
+    {
+      "epoch": 1.1253333333333333,
+      "grad_norm": 1.7124340534210205,
+      "learning_rate": 0.00015512455516014236,
+      "loss": 2.2348,
+      "step": 2532
+    },
+    {
+      "epoch": 1.1257777777777778,
+      "grad_norm": 1.6865593194961548,
+      "learning_rate": 0.0001551067615658363,
+      "loss": 2.095,
+      "step": 2533
+    },
+    {
+      "epoch": 1.1262222222222222,
+      "grad_norm": 2.115900754928589,
+      "learning_rate": 0.00015508896797153024,
+      "loss": 2.2804,
+      "step": 2534
+    },
+    {
+      "epoch": 1.1266666666666667,
+      "grad_norm": 1.720361590385437,
+      "learning_rate": 0.0001550711743772242,
+      "loss": 2.125,
+      "step": 2535
+    },
+    {
+      "epoch": 1.1271111111111112,
+      "grad_norm": 1.8980624675750732,
+      "learning_rate": 0.00015505338078291816,
+      "loss": 2.2915,
+      "step": 2536
+    },
+    {
+      "epoch": 1.1275555555555556,
+      "grad_norm": 2.2021002769470215,
+      "learning_rate": 0.00015503558718861211,
+      "loss": 2.2995,
+      "step": 2537
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 1.804162621498108,
+      "learning_rate": 0.00015501779359430607,
+      "loss": 2.1668,
+      "step": 2538
+    },
+    {
+      "epoch": 1.1284444444444444,
+      "grad_norm": 1.6780054569244385,
+      "learning_rate": 0.000155,
+      "loss": 1.7612,
+      "step": 2539
+    },
+    {
+      "epoch": 1.1288888888888888,
+      "grad_norm": 2.008111000061035,
+      "learning_rate": 0.00015498220640569396,
+      "loss": 2.2902,
+      "step": 2540
+    },
+    {
+      "epoch": 1.1293333333333333,
+      "grad_norm": 2.181689500808716,
+      "learning_rate": 0.0001549644128113879,
+      "loss": 2.3495,
+      "step": 2541
+    },
+    {
+      "epoch": 1.1297777777777778,
+      "grad_norm": 2.064223289489746,
+      "learning_rate": 0.00015494661921708187,
+      "loss": 1.9178,
+      "step": 2542
+    },
+    {
+      "epoch": 1.1302222222222222,
+      "grad_norm": 2.1660356521606445,
+      "learning_rate": 0.0001549288256227758,
+      "loss": 2.6801,
+      "step": 2543
+    },
+    {
+      "epoch": 1.1306666666666667,
+      "grad_norm": 2.031355142593384,
+      "learning_rate": 0.00015491103202846976,
+      "loss": 2.2943,
+      "step": 2544
+    },
+    {
+      "epoch": 1.1311111111111112,
+      "grad_norm": 2.2684144973754883,
+      "learning_rate": 0.0001548932384341637,
+      "loss": 2.5049,
+      "step": 2545
+    },
+    {
+      "epoch": 1.1315555555555556,
+      "grad_norm": 2.3677666187286377,
+      "learning_rate": 0.00015487544483985764,
+      "loss": 2.6018,
+      "step": 2546
+    },
+    {
+      "epoch": 1.1320000000000001,
+      "grad_norm": 2.2659435272216797,
+      "learning_rate": 0.0001548576512455516,
+      "loss": 2.3025,
+      "step": 2547
+    },
+    {
+      "epoch": 1.1324444444444444,
+      "grad_norm": 2.0491015911102295,
+      "learning_rate": 0.00015483985765124555,
+      "loss": 2.3657,
+      "step": 2548
+    },
+    {
+      "epoch": 1.1328888888888888,
+      "grad_norm": 1.3263907432556152,
+      "learning_rate": 0.0001548220640569395,
+      "loss": 0.88,
+      "step": 2549
+    },
+    {
+      "epoch": 1.1333333333333333,
+      "grad_norm": 2.2537829875946045,
+      "learning_rate": 0.00015480427046263347,
+      "loss": 1.195,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1337777777777778,
+      "grad_norm": 1.338564395904541,
+      "learning_rate": 0.00015478647686832742,
+      "loss": 2.866,
+      "step": 2551
+    },
+    {
+      "epoch": 1.1342222222222222,
+      "grad_norm": 1.3049834966659546,
+      "learning_rate": 0.00015476868327402135,
+      "loss": 2.7732,
+      "step": 2552
+    },
+    {
+      "epoch": 1.1346666666666667,
+      "grad_norm": 0.8578532338142395,
+      "learning_rate": 0.0001547508896797153,
+      "loss": 1.0418,
+      "step": 2553
+    },
+    {
+      "epoch": 1.1351111111111112,
+      "grad_norm": 1.3947099447250366,
+      "learning_rate": 0.00015473309608540927,
+      "loss": 2.3232,
+      "step": 2554
+    },
+    {
+      "epoch": 1.1355555555555557,
+      "grad_norm": 1.4263209104537964,
+      "learning_rate": 0.00015471530249110322,
+      "loss": 2.4567,
+      "step": 2555
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 1.4634780883789062,
+      "learning_rate": 0.00015469750889679715,
+      "loss": 2.2001,
+      "step": 2556
+    },
+    {
+      "epoch": 1.1364444444444444,
+      "grad_norm": 1.4137287139892578,
+      "learning_rate": 0.0001546797153024911,
+      "loss": 2.3393,
+      "step": 2557
+    },
+    {
+      "epoch": 1.1368888888888888,
+      "grad_norm": 1.3310433626174927,
+      "learning_rate": 0.00015466192170818507,
+      "loss": 2.1137,
+      "step": 2558
+    },
+    {
+      "epoch": 1.1373333333333333,
+      "grad_norm": 1.6760700941085815,
+      "learning_rate": 0.000154644128113879,
+      "loss": 2.6479,
+      "step": 2559
+    },
+    {
+      "epoch": 1.1377777777777778,
+      "grad_norm": 1.6810277700424194,
+      "learning_rate": 0.00015462633451957295,
+      "loss": 1.7997,
+      "step": 2560
+    },
+    {
+      "epoch": 1.1382222222222222,
+      "grad_norm": 1.40380859375,
+      "learning_rate": 0.0001546085409252669,
+      "loss": 2.1069,
+      "step": 2561
+    },
+    {
+      "epoch": 1.1386666666666667,
+      "grad_norm": 1.451357126235962,
+      "learning_rate": 0.00015459074733096086,
+      "loss": 2.0879,
+      "step": 2562
+    },
+    {
+      "epoch": 1.1391111111111112,
+      "grad_norm": 1.7103229761123657,
+      "learning_rate": 0.00015457295373665482,
+      "loss": 2.1198,
+      "step": 2563
+    },
+    {
+      "epoch": 1.1395555555555554,
+      "grad_norm": 1.3014283180236816,
+      "learning_rate": 0.00015455516014234878,
+      "loss": 1.4409,
+      "step": 2564
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 1.7110216617584229,
+      "learning_rate": 0.0001545373665480427,
+      "loss": 2.064,
+      "step": 2565
+    },
+    {
+      "epoch": 1.1404444444444444,
+      "grad_norm": 1.2959778308868408,
+      "learning_rate": 0.00015451957295373666,
+      "loss": 1.9788,
+      "step": 2566
+    },
+    {
+      "epoch": 1.1408888888888888,
+      "grad_norm": 1.339138388633728,
+      "learning_rate": 0.00015450177935943062,
+      "loss": 1.525,
+      "step": 2567
+    },
+    {
+      "epoch": 1.1413333333333333,
+      "grad_norm": 2.068941116333008,
+      "learning_rate": 0.00015448398576512458,
+      "loss": 2.5298,
+      "step": 2568
+    },
+    {
+      "epoch": 1.1417777777777778,
+      "grad_norm": 1.6485881805419922,
+      "learning_rate": 0.0001544661921708185,
+      "loss": 2.0371,
+      "step": 2569
+    },
+    {
+      "epoch": 1.1422222222222222,
+      "grad_norm": 1.6411560773849487,
+      "learning_rate": 0.00015444839857651246,
+      "loss": 2.1563,
+      "step": 2570
+    },
+    {
+      "epoch": 1.1426666666666667,
+      "grad_norm": 1.8316292762756348,
+      "learning_rate": 0.00015443060498220642,
+      "loss": 1.9724,
+      "step": 2571
+    },
+    {
+      "epoch": 1.1431111111111112,
+      "grad_norm": 1.3947020769119263,
+      "learning_rate": 0.00015441281138790035,
+      "loss": 2.1488,
+      "step": 2572
+    },
+    {
+      "epoch": 1.1435555555555554,
+      "grad_norm": 1.722806692123413,
+      "learning_rate": 0.0001543950177935943,
+      "loss": 1.9657,
+      "step": 2573
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 1.944720983505249,
+      "learning_rate": 0.00015437722419928826,
+      "loss": 2.0344,
+      "step": 2574
+    },
+    {
+      "epoch": 1.1444444444444444,
+      "grad_norm": 1.69381582736969,
+      "learning_rate": 0.00015435943060498222,
+      "loss": 1.6078,
+      "step": 2575
+    },
+    {
+      "epoch": 1.1448888888888888,
+      "grad_norm": 1.678240180015564,
+      "learning_rate": 0.00015434163701067617,
+      "loss": 2.2514,
+      "step": 2576
+    },
+    {
+      "epoch": 1.1453333333333333,
+      "grad_norm": 1.8483023643493652,
+      "learning_rate": 0.00015432384341637013,
+      "loss": 2.3157,
+      "step": 2577
+    },
+    {
+      "epoch": 1.1457777777777778,
+      "grad_norm": 1.6956913471221924,
+      "learning_rate": 0.00015430604982206406,
+      "loss": 2.2523,
+      "step": 2578
+    },
+    {
+      "epoch": 1.1462222222222223,
+      "grad_norm": 1.861174464225769,
+      "learning_rate": 0.00015428825622775802,
+      "loss": 2.5922,
+      "step": 2579
+    },
+    {
+      "epoch": 1.1466666666666667,
+      "grad_norm": 1.6170905828475952,
+      "learning_rate": 0.00015427046263345197,
+      "loss": 1.7942,
+      "step": 2580
+    },
+    {
+      "epoch": 1.1471111111111112,
+      "grad_norm": 1.5658963918685913,
+      "learning_rate": 0.00015425266903914593,
+      "loss": 2.0127,
+      "step": 2581
+    },
+    {
+      "epoch": 1.1475555555555554,
+      "grad_norm": 1.7916998863220215,
+      "learning_rate": 0.00015423487544483986,
+      "loss": 2.2693,
+      "step": 2582
+    },
+    {
+      "epoch": 1.148,
+      "grad_norm": 1.8371453285217285,
+      "learning_rate": 0.00015421708185053381,
+      "loss": 1.8348,
+      "step": 2583
+    },
+    {
+      "epoch": 1.1484444444444444,
+      "grad_norm": 1.862358808517456,
+      "learning_rate": 0.00015419928825622777,
+      "loss": 2.536,
+      "step": 2584
+    },
+    {
+      "epoch": 1.1488888888888888,
+      "grad_norm": 2.036752700805664,
+      "learning_rate": 0.0001541814946619217,
+      "loss": 2.4652,
+      "step": 2585
+    },
+    {
+      "epoch": 1.1493333333333333,
+      "grad_norm": 2.268584966659546,
+      "learning_rate": 0.00015416370106761566,
+      "loss": 2.2486,
+      "step": 2586
+    },
+    {
+      "epoch": 1.1497777777777778,
+      "grad_norm": 1.7090932130813599,
+      "learning_rate": 0.00015414590747330961,
+      "loss": 1.95,
+      "step": 2587
+    },
+    {
+      "epoch": 1.1502222222222223,
+      "grad_norm": 1.6534048318862915,
+      "learning_rate": 0.00015412811387900357,
+      "loss": 1.301,
+      "step": 2588
+    },
+    {
+      "epoch": 1.1506666666666667,
+      "grad_norm": 1.7545114755630493,
+      "learning_rate": 0.00015411032028469753,
+      "loss": 1.753,
+      "step": 2589
+    },
+    {
+      "epoch": 1.1511111111111112,
+      "grad_norm": 2.4128382205963135,
+      "learning_rate": 0.00015409252669039148,
+      "loss": 2.5481,
+      "step": 2590
+    },
+    {
+      "epoch": 1.1515555555555554,
+      "grad_norm": 1.815370798110962,
+      "learning_rate": 0.0001540747330960854,
+      "loss": 2.091,
+      "step": 2591
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 2.445251941680908,
+      "learning_rate": 0.00015405693950177937,
+      "loss": 2.1765,
+      "step": 2592
+    },
+    {
+      "epoch": 1.1524444444444444,
+      "grad_norm": 1.9091752767562866,
+      "learning_rate": 0.00015403914590747333,
+      "loss": 1.5047,
+      "step": 2593
+    },
+    {
+      "epoch": 1.1528888888888889,
+      "grad_norm": 2.3893678188323975,
+      "learning_rate": 0.00015402135231316728,
+      "loss": 2.2924,
+      "step": 2594
+    },
+    {
+      "epoch": 1.1533333333333333,
+      "grad_norm": 2.1791248321533203,
+      "learning_rate": 0.0001540035587188612,
+      "loss": 2.1891,
+      "step": 2595
+    },
+    {
+      "epoch": 1.1537777777777778,
+      "grad_norm": 1.9411416053771973,
+      "learning_rate": 0.00015398576512455517,
+      "loss": 2.2953,
+      "step": 2596
+    },
+    {
+      "epoch": 1.1542222222222223,
+      "grad_norm": 2.478189468383789,
+      "learning_rate": 0.00015396797153024912,
+      "loss": 1.9645,
+      "step": 2597
+    },
+    {
+      "epoch": 1.1546666666666667,
+      "grad_norm": 2.280930280685425,
+      "learning_rate": 0.00015395017793594305,
+      "loss": 2.2849,
+      "step": 2598
+    },
+    {
+      "epoch": 1.1551111111111112,
+      "grad_norm": 2.5622832775115967,
+      "learning_rate": 0.000153932384341637,
+      "loss": 1.4624,
+      "step": 2599
+    },
+    {
+      "epoch": 1.1555555555555554,
+      "grad_norm": 2.4844248294830322,
+      "learning_rate": 0.00015391459074733097,
+      "loss": 1.5791,
+      "step": 2600
+    },
+    {
+      "epoch": 1.156,
+      "grad_norm": 1.2558754682540894,
+      "learning_rate": 0.00015389679715302492,
+      "loss": 2.402,
+      "step": 2601
+    },
+    {
+      "epoch": 1.1564444444444444,
+      "grad_norm": 1.8104875087738037,
+      "learning_rate": 0.00015387900355871888,
+      "loss": 1.0876,
+      "step": 2602
+    },
+    {
+      "epoch": 1.1568888888888889,
+      "grad_norm": 1.2649706602096558,
+      "learning_rate": 0.0001538612099644128,
+      "loss": 2.1157,
+      "step": 2603
+    },
+    {
+      "epoch": 1.1573333333333333,
+      "grad_norm": 1.4647830724716187,
+      "learning_rate": 0.00015384341637010677,
+      "loss": 2.2213,
+      "step": 2604
+    },
+    {
+      "epoch": 1.1577777777777778,
+      "grad_norm": 1.4536770582199097,
+      "learning_rate": 0.00015382562277580072,
+      "loss": 2.0763,
+      "step": 2605
+    },
+    {
+      "epoch": 1.1582222222222223,
+      "grad_norm": 1.4016244411468506,
+      "learning_rate": 0.00015380782918149468,
+      "loss": 2.3909,
+      "step": 2606
+    },
+    {
+      "epoch": 1.1586666666666667,
+      "grad_norm": 1.4496042728424072,
+      "learning_rate": 0.00015379003558718864,
+      "loss": 1.9667,
+      "step": 2607
+    },
+    {
+      "epoch": 1.1591111111111112,
+      "grad_norm": 1.384210228919983,
+      "learning_rate": 0.00015377224199288256,
+      "loss": 2.8226,
+      "step": 2608
+    },
+    {
+      "epoch": 1.1595555555555555,
+      "grad_norm": 1.5709418058395386,
+      "learning_rate": 0.00015375444839857652,
+      "loss": 2.7716,
+      "step": 2609
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6072601079940796,
+      "learning_rate": 0.00015373665480427045,
+      "loss": 2.2285,
+      "step": 2610
+    },
+    {
+      "epoch": 1.1604444444444444,
+      "grad_norm": 1.4224820137023926,
+      "learning_rate": 0.0001537188612099644,
+      "loss": 2.049,
+      "step": 2611
+    },
+    {
+      "epoch": 1.1608888888888889,
+      "grad_norm": 1.5850938558578491,
+      "learning_rate": 0.00015370106761565836,
+      "loss": 2.326,
+      "step": 2612
+    },
+    {
+      "epoch": 1.1613333333333333,
+      "grad_norm": 1.3699077367782593,
+      "learning_rate": 0.00015368327402135232,
+      "loss": 2.1054,
+      "step": 2613
+    },
+    {
+      "epoch": 1.1617777777777778,
+      "grad_norm": 1.5815706253051758,
+      "learning_rate": 0.00015366548042704628,
+      "loss": 2.4236,
+      "step": 2614
+    },
+    {
+      "epoch": 1.1622222222222223,
+      "grad_norm": 1.3918722867965698,
+      "learning_rate": 0.00015364768683274023,
+      "loss": 2.2995,
+      "step": 2615
+    },
+    {
+      "epoch": 1.1626666666666667,
+      "grad_norm": 1.4517208337783813,
+      "learning_rate": 0.00015362989323843416,
+      "loss": 1.8738,
+      "step": 2616
+    },
+    {
+      "epoch": 1.1631111111111112,
+      "grad_norm": 1.84287691116333,
+      "learning_rate": 0.00015361209964412812,
+      "loss": 1.3763,
+      "step": 2617
+    },
+    {
+      "epoch": 1.1635555555555555,
+      "grad_norm": 1.6634745597839355,
+      "learning_rate": 0.00015359430604982208,
+      "loss": 2.3855,
+      "step": 2618
+    },
+    {
+      "epoch": 1.164,
+      "grad_norm": 1.544952154159546,
+      "learning_rate": 0.00015357651245551603,
+      "loss": 1.9593,
+      "step": 2619
+    },
+    {
+      "epoch": 1.1644444444444444,
+      "grad_norm": 1.4931992292404175,
+      "learning_rate": 0.00015355871886121,
+      "loss": 2.1309,
+      "step": 2620
+    },
+    {
+      "epoch": 1.1648888888888889,
+      "grad_norm": 1.7971177101135254,
+      "learning_rate": 0.00015354092526690392,
+      "loss": 2.2084,
+      "step": 2621
+    },
+    {
+      "epoch": 1.1653333333333333,
+      "grad_norm": 1.4012914896011353,
+      "learning_rate": 0.00015352313167259787,
+      "loss": 1.5844,
+      "step": 2622
+    },
+    {
+      "epoch": 1.1657777777777778,
+      "grad_norm": 1.5000579357147217,
+      "learning_rate": 0.0001535053380782918,
+      "loss": 1.8046,
+      "step": 2623
+    },
+    {
+      "epoch": 1.1662222222222223,
+      "grad_norm": 1.830424427986145,
+      "learning_rate": 0.00015348754448398576,
+      "loss": 2.6892,
+      "step": 2624
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 1.5387390851974487,
+      "learning_rate": 0.00015346975088967972,
+      "loss": 1.9499,
+      "step": 2625
+    },
+    {
+      "epoch": 1.1671111111111112,
+      "grad_norm": 1.5739384889602661,
+      "learning_rate": 0.00015345195729537367,
+      "loss": 2.0548,
+      "step": 2626
+    },
+    {
+      "epoch": 1.1675555555555555,
+      "grad_norm": 1.7389553785324097,
+      "learning_rate": 0.00015343416370106763,
+      "loss": 2.3472,
+      "step": 2627
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 1.8690800666809082,
+      "learning_rate": 0.00015341637010676159,
+      "loss": 2.5385,
+      "step": 2628
+    },
+    {
+      "epoch": 1.1684444444444444,
+      "grad_norm": 1.6705001592636108,
+      "learning_rate": 0.00015339857651245552,
+      "loss": 2.2232,
+      "step": 2629
+    },
+    {
+      "epoch": 1.1688888888888889,
+      "grad_norm": 1.811651587486267,
+      "learning_rate": 0.00015338078291814947,
+      "loss": 2.3896,
+      "step": 2630
+    },
+    {
+      "epoch": 1.1693333333333333,
+      "grad_norm": 2.812932252883911,
+      "learning_rate": 0.00015336298932384343,
+      "loss": 2.3391,
+      "step": 2631
+    },
+    {
+      "epoch": 1.1697777777777778,
+      "grad_norm": 1.434435248374939,
+      "learning_rate": 0.00015334519572953739,
+      "loss": 0.9854,
+      "step": 2632
+    },
+    {
+      "epoch": 1.1702222222222223,
+      "grad_norm": 1.7616302967071533,
+      "learning_rate": 0.00015332740213523134,
+      "loss": 2.0482,
+      "step": 2633
+    },
+    {
+      "epoch": 1.1706666666666667,
+      "grad_norm": 2.4221534729003906,
+      "learning_rate": 0.00015330960854092527,
+      "loss": 2.338,
+      "step": 2634
+    },
+    {
+      "epoch": 1.1711111111111112,
+      "grad_norm": 1.9369844198226929,
+      "learning_rate": 0.00015329181494661923,
+      "loss": 1.8924,
+      "step": 2635
+    },
+    {
+      "epoch": 1.1715555555555555,
+      "grad_norm": 2.09136700630188,
+      "learning_rate": 0.00015327402135231316,
+      "loss": 2.5067,
+      "step": 2636
+    },
+    {
+      "epoch": 1.172,
+      "grad_norm": 1.7432854175567627,
+      "learning_rate": 0.0001532562277580071,
+      "loss": 1.6748,
+      "step": 2637
+    },
+    {
+      "epoch": 1.1724444444444444,
+      "grad_norm": 1.8707739114761353,
+      "learning_rate": 0.00015323843416370107,
+      "loss": 2.3218,
+      "step": 2638
+    },
+    {
+      "epoch": 1.1728888888888889,
+      "grad_norm": 1.9009658098220825,
+      "learning_rate": 0.00015322064056939503,
+      "loss": 2.1614,
+      "step": 2639
+    },
+    {
+      "epoch": 1.1733333333333333,
+      "grad_norm": 2.321162462234497,
+      "learning_rate": 0.00015320284697508898,
+      "loss": 2.0851,
+      "step": 2640
+    },
+    {
+      "epoch": 1.1737777777777778,
+      "grad_norm": 1.7875491380691528,
+      "learning_rate": 0.00015318505338078294,
+      "loss": 1.7379,
+      "step": 2641
+    },
+    {
+      "epoch": 1.1742222222222223,
+      "grad_norm": 1.9961577653884888,
+      "learning_rate": 0.00015316725978647687,
+      "loss": 2.4488,
+      "step": 2642
+    },
+    {
+      "epoch": 1.1746666666666667,
+      "grad_norm": 2.089043617248535,
+      "learning_rate": 0.00015314946619217083,
+      "loss": 2.2587,
+      "step": 2643
+    },
+    {
+      "epoch": 1.1751111111111112,
+      "grad_norm": 2.016988754272461,
+      "learning_rate": 0.00015313167259786478,
+      "loss": 2.3505,
+      "step": 2644
+    },
+    {
+      "epoch": 1.1755555555555555,
+      "grad_norm": 3.9953866004943848,
+      "learning_rate": 0.00015311387900355874,
+      "loss": 2.6823,
+      "step": 2645
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 2.324265956878662,
+      "learning_rate": 0.00015309608540925267,
+      "loss": 2.0942,
+      "step": 2646
+    },
+    {
+      "epoch": 1.1764444444444444,
+      "grad_norm": 1.8716621398925781,
+      "learning_rate": 0.00015307829181494662,
+      "loss": 1.8976,
+      "step": 2647
+    },
+    {
+      "epoch": 1.1768888888888889,
+      "grad_norm": 2.1721549034118652,
+      "learning_rate": 0.00015306049822064058,
+      "loss": 2.3372,
+      "step": 2648
+    },
+    {
+      "epoch": 1.1773333333333333,
+      "grad_norm": 2.4310812950134277,
+      "learning_rate": 0.0001530427046263345,
+      "loss": 2.3681,
+      "step": 2649
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 2.1429104804992676,
+      "learning_rate": 0.00015302491103202847,
+      "loss": 1.3194,
+      "step": 2650
+    },
+    {
+      "epoch": 1.1782222222222223,
+      "grad_norm": 1.4743183851242065,
+      "learning_rate": 0.00015300711743772242,
+      "loss": 2.7132,
+      "step": 2651
+    },
+    {
+      "epoch": 1.1786666666666668,
+      "grad_norm": 1.449602723121643,
+      "learning_rate": 0.00015298932384341638,
+      "loss": 2.0043,
+      "step": 2652
+    },
+    {
+      "epoch": 1.1791111111111112,
+      "grad_norm": 1.5418530702590942,
+      "learning_rate": 0.00015297153024911034,
+      "loss": 2.4505,
+      "step": 2653
+    },
+    {
+      "epoch": 1.1795555555555555,
+      "grad_norm": 1.408302664756775,
+      "learning_rate": 0.0001529537366548043,
+      "loss": 2.1342,
+      "step": 2654
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6608649492263794,
+      "learning_rate": 0.00015293594306049822,
+      "loss": 2.3212,
+      "step": 2655
+    },
+    {
+      "epoch": 1.1804444444444444,
+      "grad_norm": 1.560037612915039,
+      "learning_rate": 0.00015291814946619218,
+      "loss": 2.9029,
+      "step": 2656
+    },
+    {
+      "epoch": 1.1808888888888889,
+      "grad_norm": 1.5058655738830566,
+      "learning_rate": 0.00015290035587188613,
+      "loss": 2.5045,
+      "step": 2657
+    },
+    {
+      "epoch": 1.1813333333333333,
+      "grad_norm": 1.5224006175994873,
+      "learning_rate": 0.0001528825622775801,
+      "loss": 2.2639,
+      "step": 2658
+    },
+    {
+      "epoch": 1.1817777777777778,
+      "grad_norm": 1.611315131187439,
+      "learning_rate": 0.00015286476868327402,
+      "loss": 2.594,
+      "step": 2659
+    },
+    {
+      "epoch": 1.1822222222222223,
+      "grad_norm": 1.6001996994018555,
+      "learning_rate": 0.00015284697508896798,
+      "loss": 2.6139,
+      "step": 2660
+    },
+    {
+      "epoch": 1.1826666666666668,
+      "grad_norm": 1.5929144620895386,
+      "learning_rate": 0.00015282918149466193,
+      "loss": 2.6785,
+      "step": 2661
+    },
+    {
+      "epoch": 1.1831111111111112,
+      "grad_norm": 1.6971992254257202,
+      "learning_rate": 0.00015281138790035586,
+      "loss": 2.364,
+      "step": 2662
+    },
+    {
+      "epoch": 1.1835555555555555,
+      "grad_norm": 1.468849539756775,
+      "learning_rate": 0.00015279359430604982,
+      "loss": 1.773,
+      "step": 2663
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 1.352769374847412,
+      "learning_rate": 0.00015277580071174378,
+      "loss": 2.5955,
+      "step": 2664
+    },
+    {
+      "epoch": 1.1844444444444444,
+      "grad_norm": 1.4742112159729004,
+      "learning_rate": 0.00015275800711743773,
+      "loss": 1.9913,
+      "step": 2665
+    },
+    {
+      "epoch": 1.1848888888888889,
+      "grad_norm": 1.3926454782485962,
+      "learning_rate": 0.0001527402135231317,
+      "loss": 1.8751,
+      "step": 2666
+    },
+    {
+      "epoch": 1.1853333333333333,
+      "grad_norm": 1.6356984376907349,
+      "learning_rate": 0.00015272241992882565,
+      "loss": 2.353,
+      "step": 2667
+    },
+    {
+      "epoch": 1.1857777777777778,
+      "grad_norm": 1.5044867992401123,
+      "learning_rate": 0.00015270462633451958,
+      "loss": 1.988,
+      "step": 2668
+    },
+    {
+      "epoch": 1.1862222222222223,
+      "grad_norm": 1.8624123334884644,
+      "learning_rate": 0.00015268683274021353,
+      "loss": 2.5895,
+      "step": 2669
+    },
+    {
+      "epoch": 1.1866666666666668,
+      "grad_norm": 1.6830346584320068,
+      "learning_rate": 0.0001526690391459075,
+      "loss": 1.3593,
+      "step": 2670
+    },
+    {
+      "epoch": 1.1871111111111112,
+      "grad_norm": 1.6659824848175049,
+      "learning_rate": 0.00015265124555160144,
+      "loss": 1.8984,
+      "step": 2671
+    },
+    {
+      "epoch": 1.1875555555555555,
+      "grad_norm": 1.5918940305709839,
+      "learning_rate": 0.00015263345195729537,
+      "loss": 2.0415,
+      "step": 2672
+    },
+    {
+      "epoch": 1.188,
+      "grad_norm": 1.8598551750183105,
+      "learning_rate": 0.00015261565836298933,
+      "loss": 2.1718,
+      "step": 2673
+    },
+    {
+      "epoch": 1.1884444444444444,
+      "grad_norm": 1.764703631401062,
+      "learning_rate": 0.0001525978647686833,
+      "loss": 2.0413,
+      "step": 2674
+    },
+    {
+      "epoch": 1.1888888888888889,
+      "grad_norm": 1.7580000162124634,
+      "learning_rate": 0.00015258007117437722,
+      "loss": 2.1583,
+      "step": 2675
+    },
+    {
+      "epoch": 1.1893333333333334,
+      "grad_norm": 1.7123738527297974,
+      "learning_rate": 0.00015256227758007117,
+      "loss": 2.0192,
+      "step": 2676
+    },
+    {
+      "epoch": 1.1897777777777778,
+      "grad_norm": 2.88462233543396,
+      "learning_rate": 0.00015254448398576513,
+      "loss": 1.9526,
+      "step": 2677
+    },
+    {
+      "epoch": 1.1902222222222223,
+      "grad_norm": 1.6699835062026978,
+      "learning_rate": 0.00015252669039145909,
+      "loss": 2.0907,
+      "step": 2678
+    },
+    {
+      "epoch": 1.1906666666666668,
+      "grad_norm": 2.0208330154418945,
+      "learning_rate": 0.00015250889679715304,
+      "loss": 2.3712,
+      "step": 2679
+    },
+    {
+      "epoch": 1.1911111111111112,
+      "grad_norm": 1.9458266496658325,
+      "learning_rate": 0.000152491103202847,
+      "loss": 2.1126,
+      "step": 2680
+    },
+    {
+      "epoch": 1.1915555555555555,
+      "grad_norm": 1.6024980545043945,
+      "learning_rate": 0.00015247330960854093,
+      "loss": 1.8988,
+      "step": 2681
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 1.7899705171585083,
+      "learning_rate": 0.00015245551601423488,
+      "loss": 2.0417,
+      "step": 2682
+    },
+    {
+      "epoch": 1.1924444444444444,
+      "grad_norm": 1.6227293014526367,
+      "learning_rate": 0.00015243772241992884,
+      "loss": 2.1364,
+      "step": 2683
+    },
+    {
+      "epoch": 1.1928888888888889,
+      "grad_norm": 1.7193636894226074,
+      "learning_rate": 0.0001524199288256228,
+      "loss": 1.9889,
+      "step": 2684
+    },
+    {
+      "epoch": 1.1933333333333334,
+      "grad_norm": 1.7960073947906494,
+      "learning_rate": 0.00015240213523131673,
+      "loss": 2.0971,
+      "step": 2685
+    },
+    {
+      "epoch": 1.1937777777777778,
+      "grad_norm": 1.8026853799819946,
+      "learning_rate": 0.00015238434163701068,
+      "loss": 2.0781,
+      "step": 2686
+    },
+    {
+      "epoch": 1.1942222222222223,
+      "grad_norm": 1.0487536191940308,
+      "learning_rate": 0.00015236654804270464,
+      "loss": 0.076,
+      "step": 2687
+    },
+    {
+      "epoch": 1.1946666666666665,
+      "grad_norm": 2.0273492336273193,
+      "learning_rate": 0.00015234875444839857,
+      "loss": 2.2794,
+      "step": 2688
+    },
+    {
+      "epoch": 1.1951111111111112,
+      "grad_norm": 1.8268475532531738,
+      "learning_rate": 0.00015233096085409253,
+      "loss": 1.9992,
+      "step": 2689
+    },
+    {
+      "epoch": 1.1955555555555555,
+      "grad_norm": 1.6929086446762085,
+      "learning_rate": 0.00015231316725978648,
+      "loss": 2.0807,
+      "step": 2690
+    },
+    {
+      "epoch": 1.196,
+      "grad_norm": 2.158275842666626,
+      "learning_rate": 0.00015229537366548044,
+      "loss": 2.2911,
+      "step": 2691
+    },
+    {
+      "epoch": 1.1964444444444444,
+      "grad_norm": 1.98186457157135,
+      "learning_rate": 0.0001522775800711744,
+      "loss": 2.0243,
+      "step": 2692
+    },
+    {
+      "epoch": 1.196888888888889,
+      "grad_norm": 1.696062445640564,
+      "learning_rate": 0.00015225978647686832,
+      "loss": 1.8214,
+      "step": 2693
+    },
+    {
+      "epoch": 1.1973333333333334,
+      "grad_norm": 2.215367078781128,
+      "learning_rate": 0.00015224199288256228,
+      "loss": 2.1777,
+      "step": 2694
+    },
+    {
+      "epoch": 1.1977777777777778,
+      "grad_norm": 2.0742318630218506,
+      "learning_rate": 0.00015222419928825624,
+      "loss": 2.1899,
+      "step": 2695
+    },
+    {
+      "epoch": 1.1982222222222223,
+      "grad_norm": 2.0556631088256836,
+      "learning_rate": 0.0001522064056939502,
+      "loss": 2.1347,
+      "step": 2696
+    },
+    {
+      "epoch": 1.1986666666666665,
+      "grad_norm": 2.068554162979126,
+      "learning_rate": 0.00015218861209964415,
+      "loss": 1.8877,
+      "step": 2697
+    },
+    {
+      "epoch": 1.199111111111111,
+      "grad_norm": 2.118912696838379,
+      "learning_rate": 0.00015217081850533808,
+      "loss": 2.1662,
+      "step": 2698
+    },
+    {
+      "epoch": 1.1995555555555555,
+      "grad_norm": 1.5802162885665894,
+      "learning_rate": 0.00015215302491103204,
+      "loss": 0.963,
+      "step": 2699
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.575432062149048,
+      "learning_rate": 0.00015213523131672597,
+      "loss": 2.3078,
+      "step": 2700
+    },
+    {
+      "epoch": 1.2004444444444444,
+      "grad_norm": 1.573493480682373,
+      "learning_rate": 0.00015211743772241992,
+      "loss": 1.3132,
+      "step": 2701
+    },
+    {
+      "epoch": 1.200888888888889,
+      "grad_norm": 1.0483604669570923,
+      "learning_rate": 0.00015209964412811388,
+      "loss": 1.0147,
+      "step": 2702
+    },
+    {
+      "epoch": 1.2013333333333334,
+      "grad_norm": 0.8470864295959473,
+      "learning_rate": 0.00015208185053380784,
+      "loss": 0.9255,
+      "step": 2703
+    },
+    {
+      "epoch": 1.2017777777777778,
+      "grad_norm": 1.5103946924209595,
+      "learning_rate": 0.0001520640569395018,
+      "loss": 1.7368,
+      "step": 2704
+    },
+    {
+      "epoch": 1.2022222222222223,
+      "grad_norm": 1.2146368026733398,
+      "learning_rate": 0.00015204626334519575,
+      "loss": 1.5886,
+      "step": 2705
+    },
+    {
+      "epoch": 1.2026666666666666,
+      "grad_norm": 1.3831676244735718,
+      "learning_rate": 0.00015202846975088968,
+      "loss": 2.1525,
+      "step": 2706
+    },
+    {
+      "epoch": 1.203111111111111,
+      "grad_norm": 1.5528510808944702,
+      "learning_rate": 0.00015201067615658363,
+      "loss": 2.5657,
+      "step": 2707
+    },
+    {
+      "epoch": 1.2035555555555555,
+      "grad_norm": 1.551809549331665,
+      "learning_rate": 0.0001519928825622776,
+      "loss": 2.2547,
+      "step": 2708
+    },
+    {
+      "epoch": 1.204,
+      "grad_norm": 1.305998682975769,
+      "learning_rate": 0.00015197508896797155,
+      "loss": 1.3105,
+      "step": 2709
+    },
+    {
+      "epoch": 1.2044444444444444,
+      "grad_norm": 1.868577480316162,
+      "learning_rate": 0.0001519572953736655,
+      "loss": 2.6362,
+      "step": 2710
+    },
+    {
+      "epoch": 1.204888888888889,
+      "grad_norm": 1.6573137044906616,
+      "learning_rate": 0.00015193950177935943,
+      "loss": 2.3689,
+      "step": 2711
+    },
+    {
+      "epoch": 1.2053333333333334,
+      "grad_norm": 1.4828029870986938,
+      "learning_rate": 0.0001519217081850534,
+      "loss": 1.2541,
+      "step": 2712
+    },
+    {
+      "epoch": 1.2057777777777778,
+      "grad_norm": 1.4731237888336182,
+      "learning_rate": 0.00015190391459074732,
+      "loss": 2.0792,
+      "step": 2713
+    },
+    {
+      "epoch": 1.2062222222222223,
+      "grad_norm": 1.6513289213180542,
+      "learning_rate": 0.00015188612099644128,
+      "loss": 2.2811,
+      "step": 2714
+    },
+    {
+      "epoch": 1.2066666666666666,
+      "grad_norm": 1.5226035118103027,
+      "learning_rate": 0.00015186832740213523,
+      "loss": 2.117,
+      "step": 2715
+    },
+    {
+      "epoch": 1.207111111111111,
+      "grad_norm": 1.672688603401184,
+      "learning_rate": 0.0001518505338078292,
+      "loss": 2.0124,
+      "step": 2716
+    },
+    {
+      "epoch": 1.2075555555555555,
+      "grad_norm": 1.6700776815414429,
+      "learning_rate": 0.00015183274021352315,
+      "loss": 2.0039,
+      "step": 2717
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 1.613197922706604,
+      "learning_rate": 0.0001518149466192171,
+      "loss": 2.182,
+      "step": 2718
+    },
+    {
+      "epoch": 1.2084444444444444,
+      "grad_norm": 1.8841910362243652,
+      "learning_rate": 0.00015179715302491103,
+      "loss": 2.4719,
+      "step": 2719
+    },
+    {
+      "epoch": 1.208888888888889,
+      "grad_norm": 1.5783162117004395,
+      "learning_rate": 0.000151779359430605,
+      "loss": 2.2804,
+      "step": 2720
+    },
+    {
+      "epoch": 1.2093333333333334,
+      "grad_norm": 1.5242904424667358,
+      "learning_rate": 0.00015176156583629894,
+      "loss": 2.322,
+      "step": 2721
+    },
+    {
+      "epoch": 1.2097777777777778,
+      "grad_norm": 1.4535586833953857,
+      "learning_rate": 0.0001517437722419929,
+      "loss": 1.9133,
+      "step": 2722
+    },
+    {
+      "epoch": 1.2102222222222223,
+      "grad_norm": 2.0285212993621826,
+      "learning_rate": 0.00015172597864768686,
+      "loss": 2.0526,
+      "step": 2723
+    },
+    {
+      "epoch": 1.2106666666666666,
+      "grad_norm": 1.6039782762527466,
+      "learning_rate": 0.0001517081850533808,
+      "loss": 1.8729,
+      "step": 2724
+    },
+    {
+      "epoch": 1.211111111111111,
+      "grad_norm": 1.4324172735214233,
+      "learning_rate": 0.00015169039145907474,
+      "loss": 1.6432,
+      "step": 2725
+    },
+    {
+      "epoch": 1.2115555555555555,
+      "grad_norm": 1.6351962089538574,
+      "learning_rate": 0.00015167259786476867,
+      "loss": 2.4551,
+      "step": 2726
+    },
+    {
+      "epoch": 1.212,
+      "grad_norm": 1.7832766771316528,
+      "learning_rate": 0.00015165480427046263,
+      "loss": 2.1844,
+      "step": 2727
+    },
+    {
+      "epoch": 1.2124444444444444,
+      "grad_norm": 1.8114533424377441,
+      "learning_rate": 0.00015163701067615659,
+      "loss": 2.1506,
+      "step": 2728
+    },
+    {
+      "epoch": 1.212888888888889,
+      "grad_norm": 1.875593900680542,
+      "learning_rate": 0.00015161921708185054,
+      "loss": 2.0954,
+      "step": 2729
+    },
+    {
+      "epoch": 1.2133333333333334,
+      "grad_norm": 1.673852562904358,
+      "learning_rate": 0.0001516014234875445,
+      "loss": 2.1643,
+      "step": 2730
+    },
+    {
+      "epoch": 1.2137777777777778,
+      "grad_norm": 1.6588380336761475,
+      "learning_rate": 0.00015158362989323845,
+      "loss": 1.6013,
+      "step": 2731
+    },
+    {
+      "epoch": 1.2142222222222223,
+      "grad_norm": 1.8047765493392944,
+      "learning_rate": 0.00015156583629893238,
+      "loss": 2.441,
+      "step": 2732
+    },
+    {
+      "epoch": 1.2146666666666666,
+      "grad_norm": 1.8153445720672607,
+      "learning_rate": 0.00015154804270462634,
+      "loss": 2.2575,
+      "step": 2733
+    },
+    {
+      "epoch": 1.215111111111111,
+      "grad_norm": 1.9475229978561401,
+      "learning_rate": 0.0001515302491103203,
+      "loss": 2.0499,
+      "step": 2734
+    },
+    {
+      "epoch": 1.2155555555555555,
+      "grad_norm": 1.8662759065628052,
+      "learning_rate": 0.00015151245551601425,
+      "loss": 1.9091,
+      "step": 2735
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 1.9955482482910156,
+      "learning_rate": 0.0001514946619217082,
+      "loss": 2.5757,
+      "step": 2736
+    },
+    {
+      "epoch": 1.2164444444444444,
+      "grad_norm": 1.897420883178711,
+      "learning_rate": 0.00015147686832740214,
+      "loss": 1.8366,
+      "step": 2737
+    },
+    {
+      "epoch": 1.216888888888889,
+      "grad_norm": 1.6787134408950806,
+      "learning_rate": 0.0001514590747330961,
+      "loss": 0.962,
+      "step": 2738
+    },
+    {
+      "epoch": 1.2173333333333334,
+      "grad_norm": 1.790507197380066,
+      "learning_rate": 0.00015144128113879003,
+      "loss": 2.1025,
+      "step": 2739
+    },
+    {
+      "epoch": 1.2177777777777778,
+      "grad_norm": 2.025376558303833,
+      "learning_rate": 0.00015142348754448398,
+      "loss": 2.4571,
+      "step": 2740
+    },
+    {
+      "epoch": 1.2182222222222223,
+      "grad_norm": 2.094409942626953,
+      "learning_rate": 0.00015140569395017794,
+      "loss": 2.2626,
+      "step": 2741
+    },
+    {
+      "epoch": 1.2186666666666666,
+      "grad_norm": 2.1918323040008545,
+      "learning_rate": 0.0001513879003558719,
+      "loss": 2.4293,
+      "step": 2742
+    },
+    {
+      "epoch": 1.219111111111111,
+      "grad_norm": 2.2895400524139404,
+      "learning_rate": 0.00015137010676156585,
+      "loss": 2.4093,
+      "step": 2743
+    },
+    {
+      "epoch": 1.2195555555555555,
+      "grad_norm": 2.0072021484375,
+      "learning_rate": 0.0001513523131672598,
+      "loss": 2.3826,
+      "step": 2744
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0313949584960938,
+      "learning_rate": 0.00015133451957295374,
+      "loss": 1.6368,
+      "step": 2745
+    },
+    {
+      "epoch": 1.2204444444444444,
+      "grad_norm": 2.9698612689971924,
+      "learning_rate": 0.0001513167259786477,
+      "loss": 2.5953,
+      "step": 2746
+    },
+    {
+      "epoch": 1.220888888888889,
+      "grad_norm": 2.604888439178467,
+      "learning_rate": 0.00015129893238434165,
+      "loss": 2.9221,
+      "step": 2747
+    },
+    {
+      "epoch": 1.2213333333333334,
+      "grad_norm": 2.023358106613159,
+      "learning_rate": 0.0001512811387900356,
+      "loss": 1.8182,
+      "step": 2748
+    },
+    {
+      "epoch": 1.2217777777777779,
+      "grad_norm": 2.1045310497283936,
+      "learning_rate": 0.00015126334519572956,
+      "loss": 1.6869,
+      "step": 2749
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 2.9360511302948,
+      "learning_rate": 0.0001512455516014235,
+      "loss": 2.0189,
+      "step": 2750
+    },
+    {
+      "epoch": 1.2226666666666666,
+      "grad_norm": 1.2831294536590576,
+      "learning_rate": 0.00015122775800711745,
+      "loss": 2.4483,
+      "step": 2751
+    },
+    {
+      "epoch": 1.223111111111111,
+      "grad_norm": 1.1311028003692627,
+      "learning_rate": 0.00015120996441281138,
+      "loss": 1.2781,
+      "step": 2752
+    },
+    {
+      "epoch": 1.2235555555555555,
+      "grad_norm": 1.4160290956497192,
+      "learning_rate": 0.00015119217081850534,
+      "loss": 2.2221,
+      "step": 2753
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 1.3835599422454834,
+      "learning_rate": 0.0001511743772241993,
+      "loss": 2.2433,
+      "step": 2754
+    },
+    {
+      "epoch": 1.2244444444444444,
+      "grad_norm": 1.6253594160079956,
+      "learning_rate": 0.00015115658362989325,
+      "loss": 1.8161,
+      "step": 2755
+    },
+    {
+      "epoch": 1.224888888888889,
+      "grad_norm": 1.227462649345398,
+      "learning_rate": 0.0001511387900355872,
+      "loss": 1.9292,
+      "step": 2756
+    },
+    {
+      "epoch": 1.2253333333333334,
+      "grad_norm": 1.425794005393982,
+      "learning_rate": 0.00015112099644128116,
+      "loss": 1.6751,
+      "step": 2757
+    },
+    {
+      "epoch": 1.2257777777777779,
+      "grad_norm": 0.8352449536323547,
+      "learning_rate": 0.0001511032028469751,
+      "loss": 0.042,
+      "step": 2758
+    },
+    {
+      "epoch": 1.2262222222222223,
+      "grad_norm": 1.41720449924469,
+      "learning_rate": 0.00015108540925266905,
+      "loss": 2.2062,
+      "step": 2759
+    },
+    {
+      "epoch": 1.2266666666666666,
+      "grad_norm": 1.4105916023254395,
+      "learning_rate": 0.000151067615658363,
+      "loss": 1.8583,
+      "step": 2760
+    },
+    {
+      "epoch": 1.227111111111111,
+      "grad_norm": 1.653696060180664,
+      "learning_rate": 0.00015104982206405696,
+      "loss": 1.4183,
+      "step": 2761
+    },
+    {
+      "epoch": 1.2275555555555555,
+      "grad_norm": 1.5550695657730103,
+      "learning_rate": 0.0001510320284697509,
+      "loss": 2.0958,
+      "step": 2762
+    },
+    {
+      "epoch": 1.228,
+      "grad_norm": 1.5534552335739136,
+      "learning_rate": 0.00015101423487544485,
+      "loss": 1.1471,
+      "step": 2763
+    },
+    {
+      "epoch": 1.2284444444444444,
+      "grad_norm": 1.956020474433899,
+      "learning_rate": 0.0001509964412811388,
+      "loss": 2.408,
+      "step": 2764
+    },
+    {
+      "epoch": 1.228888888888889,
+      "grad_norm": 1.5008649826049805,
+      "learning_rate": 0.00015097864768683273,
+      "loss": 1.9145,
+      "step": 2765
+    },
+    {
+      "epoch": 1.2293333333333334,
+      "grad_norm": 1.4730578660964966,
+      "learning_rate": 0.0001509608540925267,
+      "loss": 1.8439,
+      "step": 2766
+    },
+    {
+      "epoch": 1.2297777777777779,
+      "grad_norm": 1.7233079671859741,
+      "learning_rate": 0.00015094306049822064,
+      "loss": 2.4389,
+      "step": 2767
+    },
+    {
+      "epoch": 1.2302222222222223,
+      "grad_norm": 1.4450547695159912,
+      "learning_rate": 0.0001509252669039146,
+      "loss": 2.1717,
+      "step": 2768
+    },
+    {
+      "epoch": 1.2306666666666666,
+      "grad_norm": 1.7250124216079712,
+      "learning_rate": 0.00015090747330960856,
+      "loss": 2.198,
+      "step": 2769
+    },
+    {
+      "epoch": 1.231111111111111,
+      "grad_norm": 2.009876012802124,
+      "learning_rate": 0.00015088967971530251,
+      "loss": 2.6704,
+      "step": 2770
+    },
+    {
+      "epoch": 1.2315555555555555,
+      "grad_norm": 1.9580446481704712,
+      "learning_rate": 0.00015087188612099644,
+      "loss": 1.803,
+      "step": 2771
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 1.7807945013046265,
+      "learning_rate": 0.0001508540925266904,
+      "loss": 2.401,
+      "step": 2772
+    },
+    {
+      "epoch": 1.2324444444444445,
+      "grad_norm": 1.9631510972976685,
+      "learning_rate": 0.00015083629893238436,
+      "loss": 2.762,
+      "step": 2773
+    },
+    {
+      "epoch": 1.232888888888889,
+      "grad_norm": 1.7688226699829102,
+      "learning_rate": 0.0001508185053380783,
+      "loss": 2.5009,
+      "step": 2774
+    },
+    {
+      "epoch": 1.2333333333333334,
+      "grad_norm": 1.898895263671875,
+      "learning_rate": 0.00015080071174377224,
+      "loss": 2.1815,
+      "step": 2775
+    },
+    {
+      "epoch": 1.2337777777777779,
+      "grad_norm": 1.748230218887329,
+      "learning_rate": 0.0001507829181494662,
+      "loss": 2.1702,
+      "step": 2776
+    },
+    {
+      "epoch": 1.2342222222222223,
+      "grad_norm": 1.829336404800415,
+      "learning_rate": 0.00015076512455516016,
+      "loss": 2.6297,
+      "step": 2777
+    },
+    {
+      "epoch": 1.2346666666666666,
+      "grad_norm": 1.7293047904968262,
+      "learning_rate": 0.00015074733096085409,
+      "loss": 2.0303,
+      "step": 2778
+    },
+    {
+      "epoch": 1.235111111111111,
+      "grad_norm": 1.6625522375106812,
+      "learning_rate": 0.00015072953736654804,
+      "loss": 1.9354,
+      "step": 2779
+    },
+    {
+      "epoch": 1.2355555555555555,
+      "grad_norm": 1.653939127922058,
+      "learning_rate": 0.000150711743772242,
+      "loss": 1.7279,
+      "step": 2780
+    },
+    {
+      "epoch": 1.236,
+      "grad_norm": 1.7402019500732422,
+      "learning_rate": 0.00015069395017793595,
+      "loss": 1.8008,
+      "step": 2781
+    },
+    {
+      "epoch": 1.2364444444444445,
+      "grad_norm": 1.5978055000305176,
+      "learning_rate": 0.0001506761565836299,
+      "loss": 2.0594,
+      "step": 2782
+    },
+    {
+      "epoch": 1.236888888888889,
+      "grad_norm": 1.8129159212112427,
+      "learning_rate": 0.00015065836298932384,
+      "loss": 1.8207,
+      "step": 2783
+    },
+    {
+      "epoch": 1.2373333333333334,
+      "grad_norm": 1.7434604167938232,
+      "learning_rate": 0.0001506405693950178,
+      "loss": 1.9957,
+      "step": 2784
+    },
+    {
+      "epoch": 1.2377777777777779,
+      "grad_norm": 1.829714059829712,
+      "learning_rate": 0.00015062277580071175,
+      "loss": 2.2293,
+      "step": 2785
+    },
+    {
+      "epoch": 1.2382222222222223,
+      "grad_norm": 2.231995105743408,
+      "learning_rate": 0.0001506049822064057,
+      "loss": 2.5919,
+      "step": 2786
+    },
+    {
+      "epoch": 1.2386666666666666,
+      "grad_norm": 2.2278225421905518,
+      "learning_rate": 0.00015058718861209967,
+      "loss": 2.299,
+      "step": 2787
+    },
+    {
+      "epoch": 1.239111111111111,
+      "grad_norm": 1.5640493631362915,
+      "learning_rate": 0.0001505693950177936,
+      "loss": 1.2429,
+      "step": 2788
+    },
+    {
+      "epoch": 1.2395555555555555,
+      "grad_norm": 1.987496256828308,
+      "learning_rate": 0.00015055160142348755,
+      "loss": 1.1852,
+      "step": 2789
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.910750389099121,
+      "learning_rate": 0.00015053380782918148,
+      "loss": 1.7426,
+      "step": 2790
+    },
+    {
+      "epoch": 1.2404444444444445,
+      "grad_norm": 1.7743321657180786,
+      "learning_rate": 0.00015051601423487544,
+      "loss": 1.7287,
+      "step": 2791
+    },
+    {
+      "epoch": 1.240888888888889,
+      "grad_norm": 1.9322333335876465,
+      "learning_rate": 0.0001504982206405694,
+      "loss": 2.3926,
+      "step": 2792
+    },
+    {
+      "epoch": 1.2413333333333334,
+      "grad_norm": 1.9463812112808228,
+      "learning_rate": 0.00015048042704626335,
+      "loss": 2.2058,
+      "step": 2793
+    },
+    {
+      "epoch": 1.2417777777777779,
+      "grad_norm": 1.9254072904586792,
+      "learning_rate": 0.0001504626334519573,
+      "loss": 2.167,
+      "step": 2794
+    },
+    {
+      "epoch": 1.2422222222222223,
+      "grad_norm": 2.315269947052002,
+      "learning_rate": 0.00015044483985765126,
+      "loss": 1.988,
+      "step": 2795
+    },
+    {
+      "epoch": 1.2426666666666666,
+      "grad_norm": 2.6116104125976562,
+      "learning_rate": 0.0001504270462633452,
+      "loss": 2.5243,
+      "step": 2796
+    },
+    {
+      "epoch": 1.243111111111111,
+      "grad_norm": 2.500777244567871,
+      "learning_rate": 0.00015040925266903915,
+      "loss": 2.4406,
+      "step": 2797
+    },
+    {
+      "epoch": 1.2435555555555555,
+      "grad_norm": 2.5886335372924805,
+      "learning_rate": 0.0001503914590747331,
+      "loss": 1.7912,
+      "step": 2798
+    },
+    {
+      "epoch": 1.244,
+      "grad_norm": 2.798053503036499,
+      "learning_rate": 0.00015037366548042706,
+      "loss": 1.7156,
+      "step": 2799
+    },
+    {
+      "epoch": 1.2444444444444445,
+      "grad_norm": 3.800767660140991,
+      "learning_rate": 0.00015035587188612102,
+      "loss": 1.2951,
+      "step": 2800
+    },
+    {
+      "epoch": 1.244888888888889,
+      "grad_norm": 1.2219860553741455,
+      "learning_rate": 0.00015033807829181495,
+      "loss": 2.3513,
+      "step": 2801
+    },
+    {
+      "epoch": 1.2453333333333334,
+      "grad_norm": 1.1778558492660522,
+      "learning_rate": 0.0001503202846975089,
+      "loss": 2.2511,
+      "step": 2802
+    },
+    {
+      "epoch": 1.2457777777777779,
+      "grad_norm": 1.369297981262207,
+      "learning_rate": 0.00015030249110320283,
+      "loss": 1.935,
+      "step": 2803
+    },
+    {
+      "epoch": 1.2462222222222223,
+      "grad_norm": 1.6750556230545044,
+      "learning_rate": 0.0001502846975088968,
+      "loss": 2.3057,
+      "step": 2804
+    },
+    {
+      "epoch": 1.2466666666666666,
+      "grad_norm": 1.4354381561279297,
+      "learning_rate": 0.00015026690391459075,
+      "loss": 2.4099,
+      "step": 2805
+    },
+    {
+      "epoch": 1.247111111111111,
+      "grad_norm": 1.3664571046829224,
+      "learning_rate": 0.0001502491103202847,
+      "loss": 2.0848,
+      "step": 2806
+    },
+    {
+      "epoch": 1.2475555555555555,
+      "grad_norm": 1.3316177129745483,
+      "learning_rate": 0.00015023131672597866,
+      "loss": 2.1252,
+      "step": 2807
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 1.3470760583877563,
+      "learning_rate": 0.00015021352313167262,
+      "loss": 2.3662,
+      "step": 2808
+    },
+    {
+      "epoch": 1.2484444444444445,
+      "grad_norm": 1.4858183860778809,
+      "learning_rate": 0.00015019572953736655,
+      "loss": 2.4822,
+      "step": 2809
+    },
+    {
+      "epoch": 1.248888888888889,
+      "grad_norm": 1.4876043796539307,
+      "learning_rate": 0.0001501779359430605,
+      "loss": 2.2044,
+      "step": 2810
+    },
+    {
+      "epoch": 1.2493333333333334,
+      "grad_norm": 1.807070016860962,
+      "learning_rate": 0.00015016014234875446,
+      "loss": 2.438,
+      "step": 2811
+    },
+    {
+      "epoch": 1.2497777777777777,
+      "grad_norm": 1.5529999732971191,
+      "learning_rate": 0.00015014234875444842,
+      "loss": 2.2415,
+      "step": 2812
+    },
+    {
+      "epoch": 1.2502222222222223,
+      "grad_norm": 1.5677090883255005,
+      "learning_rate": 0.00015012455516014237,
+      "loss": 2.1466,
+      "step": 2813
+    },
+    {
+      "epoch": 1.2506666666666666,
+      "grad_norm": 1.4437453746795654,
+      "learning_rate": 0.0001501067615658363,
+      "loss": 2.1075,
+      "step": 2814
+    },
+    {
+      "epoch": 1.251111111111111,
+      "grad_norm": 1.7834696769714355,
+      "learning_rate": 0.00015008896797153026,
+      "loss": 2.6351,
+      "step": 2815
+    },
+    {
+      "epoch": 1.2515555555555555,
+      "grad_norm": 1.593764305114746,
+      "learning_rate": 0.0001500711743772242,
+      "loss": 2.1957,
+      "step": 2816
+    },
+    {
+      "epoch": 1.252,
+      "grad_norm": 1.5799425840377808,
+      "learning_rate": 0.00015005338078291814,
+      "loss": 2.7452,
+      "step": 2817
+    },
+    {
+      "epoch": 1.2524444444444445,
+      "grad_norm": 1.6034505367279053,
+      "learning_rate": 0.0001500355871886121,
+      "loss": 1.7773,
+      "step": 2818
+    },
+    {
+      "epoch": 1.252888888888889,
+      "grad_norm": 1.1954715251922607,
+      "learning_rate": 0.00015001779359430606,
+      "loss": 1.204,
+      "step": 2819
+    },
+    {
+      "epoch": 1.2533333333333334,
+      "grad_norm": 1.638249158859253,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 2.394,
+      "step": 2820
+    },
+    {
+      "epoch": 1.2537777777777777,
+      "grad_norm": 1.640615463256836,
+      "learning_rate": 0.00014998220640569397,
+      "loss": 2.6767,
+      "step": 2821
+    },
+    {
+      "epoch": 1.2542222222222223,
+      "grad_norm": 1.6124838590621948,
+      "learning_rate": 0.0001499644128113879,
+      "loss": 2.2059,
+      "step": 2822
+    },
+    {
+      "epoch": 1.2546666666666666,
+      "grad_norm": 1.6705976724624634,
+      "learning_rate": 0.00014994661921708186,
+      "loss": 2.2517,
+      "step": 2823
+    },
+    {
+      "epoch": 1.255111111111111,
+      "grad_norm": 1.4649361371994019,
+      "learning_rate": 0.0001499288256227758,
+      "loss": 1.9224,
+      "step": 2824
+    },
+    {
+      "epoch": 1.2555555555555555,
+      "grad_norm": 1.701545238494873,
+      "learning_rate": 0.00014991103202846977,
+      "loss": 2.2677,
+      "step": 2825
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 1.726928949356079,
+      "learning_rate": 0.00014989323843416373,
+      "loss": 2.1658,
+      "step": 2826
+    },
+    {
+      "epoch": 1.2564444444444445,
+      "grad_norm": 1.6811003684997559,
+      "learning_rate": 0.00014987544483985766,
+      "loss": 2.0167,
+      "step": 2827
+    },
+    {
+      "epoch": 1.256888888888889,
+      "grad_norm": 1.8641170263290405,
+      "learning_rate": 0.0001498576512455516,
+      "loss": 1.7901,
+      "step": 2828
+    },
+    {
+      "epoch": 1.2573333333333334,
+      "grad_norm": 1.6056395769119263,
+      "learning_rate": 0.00014983985765124554,
+      "loss": 1.6597,
+      "step": 2829
+    },
+    {
+      "epoch": 1.2577777777777777,
+      "grad_norm": 1.687373399734497,
+      "learning_rate": 0.0001498220640569395,
+      "loss": 1.9804,
+      "step": 2830
+    },
+    {
+      "epoch": 1.2582222222222224,
+      "grad_norm": 1.6241012811660767,
+      "learning_rate": 0.00014980427046263345,
+      "loss": 2.0725,
+      "step": 2831
+    },
+    {
+      "epoch": 1.2586666666666666,
+      "grad_norm": 1.4386781454086304,
+      "learning_rate": 0.0001497864768683274,
+      "loss": 1.1918,
+      "step": 2832
+    },
+    {
+      "epoch": 1.259111111111111,
+      "grad_norm": 1.7372790575027466,
+      "learning_rate": 0.00014976868327402137,
+      "loss": 2.2247,
+      "step": 2833
+    },
+    {
+      "epoch": 1.2595555555555555,
+      "grad_norm": 1.8010145425796509,
+      "learning_rate": 0.00014975088967971532,
+      "loss": 2.1944,
+      "step": 2834
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9382820129394531,
+      "learning_rate": 0.00014973309608540925,
+      "loss": 2.3833,
+      "step": 2835
+    },
+    {
+      "epoch": 1.2604444444444445,
+      "grad_norm": 1.9000599384307861,
+      "learning_rate": 0.0001497153024911032,
+      "loss": 1.7084,
+      "step": 2836
+    },
+    {
+      "epoch": 1.260888888888889,
+      "grad_norm": 2.1467716693878174,
+      "learning_rate": 0.00014969750889679717,
+      "loss": 2.0741,
+      "step": 2837
+    },
+    {
+      "epoch": 1.2613333333333334,
+      "grad_norm": 2.0739872455596924,
+      "learning_rate": 0.00014967971530249112,
+      "loss": 2.2452,
+      "step": 2838
+    },
+    {
+      "epoch": 1.2617777777777777,
+      "grad_norm": 1.7084413766860962,
+      "learning_rate": 0.00014966192170818508,
+      "loss": 1.3884,
+      "step": 2839
+    },
+    {
+      "epoch": 1.2622222222222224,
+      "grad_norm": 1.855447769165039,
+      "learning_rate": 0.000149644128113879,
+      "loss": 1.9037,
+      "step": 2840
+    },
+    {
+      "epoch": 1.2626666666666666,
+      "grad_norm": 1.8656028509140015,
+      "learning_rate": 0.00014962633451957296,
+      "loss": 1.7658,
+      "step": 2841
+    },
+    {
+      "epoch": 1.263111111111111,
+      "grad_norm": 2.142399549484253,
+      "learning_rate": 0.0001496085409252669,
+      "loss": 1.9569,
+      "step": 2842
+    },
+    {
+      "epoch": 1.2635555555555555,
+      "grad_norm": 1.9603620767593384,
+      "learning_rate": 0.00014959074733096085,
+      "loss": 1.9807,
+      "step": 2843
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 2.183345317840576,
+      "learning_rate": 0.0001495729537366548,
+      "loss": 2.4544,
+      "step": 2844
+    },
+    {
+      "epoch": 1.2644444444444445,
+      "grad_norm": 2.205909490585327,
+      "learning_rate": 0.00014955516014234876,
+      "loss": 2.4811,
+      "step": 2845
+    },
+    {
+      "epoch": 1.264888888888889,
+      "grad_norm": 2.540581226348877,
+      "learning_rate": 0.00014953736654804272,
+      "loss": 2.9558,
+      "step": 2846
+    },
+    {
+      "epoch": 1.2653333333333334,
+      "grad_norm": 2.1151061058044434,
+      "learning_rate": 0.00014951957295373668,
+      "loss": 1.7681,
+      "step": 2847
+    },
+    {
+      "epoch": 1.2657777777777777,
+      "grad_norm": 2.5562145709991455,
+      "learning_rate": 0.0001495017793594306,
+      "loss": 2.3364,
+      "step": 2848
+    },
+    {
+      "epoch": 1.2662222222222224,
+      "grad_norm": 2.720233201980591,
+      "learning_rate": 0.00014948398576512456,
+      "loss": 1.7419,
+      "step": 2849
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 2.397717237472534,
+      "learning_rate": 0.00014946619217081852,
+      "loss": 1.9456,
+      "step": 2850
+    },
+    {
+      "epoch": 1.267111111111111,
+      "grad_norm": 1.2284705638885498,
+      "learning_rate": 0.00014944839857651248,
+      "loss": 2.54,
+      "step": 2851
+    },
+    {
+      "epoch": 1.2675555555555555,
+      "grad_norm": 1.3442673683166504,
+      "learning_rate": 0.00014943060498220643,
+      "loss": 2.1183,
+      "step": 2852
+    },
+    {
+      "epoch": 1.268,
+      "grad_norm": 1.3058741092681885,
+      "learning_rate": 0.00014941281138790036,
+      "loss": 2.0397,
+      "step": 2853
+    },
+    {
+      "epoch": 1.2684444444444445,
+      "grad_norm": 1.4303501844406128,
+      "learning_rate": 0.00014939501779359432,
+      "loss": 2.9471,
+      "step": 2854
+    },
+    {
+      "epoch": 1.268888888888889,
+      "grad_norm": 1.456242561340332,
+      "learning_rate": 0.00014937722419928825,
+      "loss": 2.1462,
+      "step": 2855
+    },
+    {
+      "epoch": 1.2693333333333334,
+      "grad_norm": 1.5883921384811401,
+      "learning_rate": 0.0001493594306049822,
+      "loss": 2.7315,
+      "step": 2856
+    },
+    {
+      "epoch": 1.2697777777777777,
+      "grad_norm": 1.4734134674072266,
+      "learning_rate": 0.00014934163701067616,
+      "loss": 1.95,
+      "step": 2857
+    },
+    {
+      "epoch": 1.2702222222222221,
+      "grad_norm": 1.4658904075622559,
+      "learning_rate": 0.00014932384341637012,
+      "loss": 2.5269,
+      "step": 2858
+    },
+    {
+      "epoch": 1.2706666666666666,
+      "grad_norm": 1.253811240196228,
+      "learning_rate": 0.00014930604982206407,
+      "loss": 1.3081,
+      "step": 2859
+    },
+    {
+      "epoch": 1.271111111111111,
+      "grad_norm": 1.178534984588623,
+      "learning_rate": 0.00014928825622775803,
+      "loss": 1.2465,
+      "step": 2860
+    },
+    {
+      "epoch": 1.2715555555555556,
+      "grad_norm": 1.6904065608978271,
+      "learning_rate": 0.00014927046263345196,
+      "loss": 2.314,
+      "step": 2861
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 1.5300809144973755,
+      "learning_rate": 0.00014925266903914592,
+      "loss": 2.0894,
+      "step": 2862
+    },
+    {
+      "epoch": 1.2724444444444445,
+      "grad_norm": 1.5079879760742188,
+      "learning_rate": 0.00014923487544483987,
+      "loss": 2.1044,
+      "step": 2863
+    },
+    {
+      "epoch": 1.272888888888889,
+      "grad_norm": 1.88065505027771,
+      "learning_rate": 0.00014921708185053383,
+      "loss": 2.1615,
+      "step": 2864
+    },
+    {
+      "epoch": 1.2733333333333334,
+      "grad_norm": 1.3186235427856445,
+      "learning_rate": 0.00014919928825622779,
+      "loss": 0.7738,
+      "step": 2865
+    },
+    {
+      "epoch": 1.2737777777777777,
+      "grad_norm": 2.2539889812469482,
+      "learning_rate": 0.00014918149466192171,
+      "loss": 2.2635,
+      "step": 2866
+    },
+    {
+      "epoch": 1.2742222222222221,
+      "grad_norm": 1.5537022352218628,
+      "learning_rate": 0.00014916370106761567,
+      "loss": 1.9448,
+      "step": 2867
+    },
+    {
+      "epoch": 1.2746666666666666,
+      "grad_norm": 1.676327109336853,
+      "learning_rate": 0.0001491459074733096,
+      "loss": 2.2653,
+      "step": 2868
+    },
+    {
+      "epoch": 1.275111111111111,
+      "grad_norm": 1.697751760482788,
+      "learning_rate": 0.00014912811387900356,
+      "loss": 2.2573,
+      "step": 2869
+    },
+    {
+      "epoch": 1.2755555555555556,
+      "grad_norm": 1.7857534885406494,
+      "learning_rate": 0.00014911032028469751,
+      "loss": 2.6627,
+      "step": 2870
+    },
+    {
+      "epoch": 1.276,
+      "grad_norm": 1.6772701740264893,
+      "learning_rate": 0.00014909252669039147,
+      "loss": 2.1949,
+      "step": 2871
+    },
+    {
+      "epoch": 1.2764444444444445,
+      "grad_norm": 1.546369194984436,
+      "learning_rate": 0.00014907473309608543,
+      "loss": 1.5786,
+      "step": 2872
+    },
+    {
+      "epoch": 1.276888888888889,
+      "grad_norm": 1.6441593170166016,
+      "learning_rate": 0.00014905693950177936,
+      "loss": 1.9288,
+      "step": 2873
+    },
+    {
+      "epoch": 1.2773333333333334,
+      "grad_norm": 1.613301396369934,
+      "learning_rate": 0.0001490391459074733,
+      "loss": 1.9617,
+      "step": 2874
+    },
+    {
+      "epoch": 1.2777777777777777,
+      "grad_norm": 2.057661771774292,
+      "learning_rate": 0.00014902135231316727,
+      "loss": 2.3796,
+      "step": 2875
+    },
+    {
+      "epoch": 1.2782222222222221,
+      "grad_norm": 2.0095505714416504,
+      "learning_rate": 0.00014900355871886123,
+      "loss": 2.1544,
+      "step": 2876
+    },
+    {
+      "epoch": 1.2786666666666666,
+      "grad_norm": 1.9702578783035278,
+      "learning_rate": 0.00014898576512455518,
+      "loss": 2.562,
+      "step": 2877
+    },
+    {
+      "epoch": 1.279111111111111,
+      "grad_norm": 1.7147190570831299,
+      "learning_rate": 0.00014896797153024914,
+      "loss": 2.1862,
+      "step": 2878
+    },
+    {
+      "epoch": 1.2795555555555556,
+      "grad_norm": 1.5411655902862549,
+      "learning_rate": 0.00014895017793594307,
+      "loss": 2.0377,
+      "step": 2879
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9299793243408203,
+      "learning_rate": 0.000148932384341637,
+      "loss": 2.4939,
+      "step": 2880
+    },
+    {
+      "epoch": 1.2804444444444445,
+      "grad_norm": 1.9820499420166016,
+      "learning_rate": 0.00014891459074733095,
+      "loss": 2.4094,
+      "step": 2881
+    },
+    {
+      "epoch": 1.280888888888889,
+      "grad_norm": 1.9093626737594604,
+      "learning_rate": 0.0001488967971530249,
+      "loss": 1.8737,
+      "step": 2882
+    },
+    {
+      "epoch": 1.2813333333333334,
+      "grad_norm": 1.7828611135482788,
+      "learning_rate": 0.00014887900355871887,
+      "loss": 2.3426,
+      "step": 2883
+    },
+    {
+      "epoch": 1.2817777777777777,
+      "grad_norm": 2.0726230144500732,
+      "learning_rate": 0.00014886120996441282,
+      "loss": 2.2123,
+      "step": 2884
+    },
+    {
+      "epoch": 1.2822222222222222,
+      "grad_norm": 1.8538103103637695,
+      "learning_rate": 0.00014884341637010678,
+      "loss": 1.9047,
+      "step": 2885
+    },
+    {
+      "epoch": 1.2826666666666666,
+      "grad_norm": 1.746737003326416,
+      "learning_rate": 0.0001488256227758007,
+      "loss": 1.693,
+      "step": 2886
+    },
+    {
+      "epoch": 1.283111111111111,
+      "grad_norm": 1.8844788074493408,
+      "learning_rate": 0.00014880782918149467,
+      "loss": 2.0716,
+      "step": 2887
+    },
+    {
+      "epoch": 1.2835555555555556,
+      "grad_norm": 1.7158288955688477,
+      "learning_rate": 0.00014879003558718862,
+      "loss": 1.9545,
+      "step": 2888
+    },
+    {
+      "epoch": 1.284,
+      "grad_norm": 1.926275610923767,
+      "learning_rate": 0.00014877224199288258,
+      "loss": 1.8723,
+      "step": 2889
+    },
+    {
+      "epoch": 1.2844444444444445,
+      "grad_norm": 2.2394421100616455,
+      "learning_rate": 0.00014875444839857654,
+      "loss": 2.3997,
+      "step": 2890
+    },
+    {
+      "epoch": 1.284888888888889,
+      "grad_norm": 1.957261323928833,
+      "learning_rate": 0.00014873665480427046,
+      "loss": 2.1043,
+      "step": 2891
+    },
+    {
+      "epoch": 1.2853333333333334,
+      "grad_norm": 2.291721820831299,
+      "learning_rate": 0.00014871886120996442,
+      "loss": 2.213,
+      "step": 2892
+    },
+    {
+      "epoch": 1.2857777777777777,
+      "grad_norm": 1.8697887659072876,
+      "learning_rate": 0.00014870106761565835,
+      "loss": 1.8325,
+      "step": 2893
+    },
+    {
+      "epoch": 1.2862222222222222,
+      "grad_norm": 1.8223001956939697,
+      "learning_rate": 0.0001486832740213523,
+      "loss": 1.8901,
+      "step": 2894
+    },
+    {
+      "epoch": 1.2866666666666666,
+      "grad_norm": 1.9460232257843018,
+      "learning_rate": 0.00014866548042704626,
+      "loss": 2.0251,
+      "step": 2895
+    },
+    {
+      "epoch": 1.287111111111111,
+      "grad_norm": 2.430386543273926,
+      "learning_rate": 0.00014864768683274022,
+      "loss": 2.591,
+      "step": 2896
+    },
+    {
+      "epoch": 1.2875555555555556,
+      "grad_norm": 2.0571846961975098,
+      "learning_rate": 0.00014862989323843418,
+      "loss": 2.0458,
+      "step": 2897
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 2.013607978820801,
+      "learning_rate": 0.00014861209964412813,
+      "loss": 1.9527,
+      "step": 2898
+    },
+    {
+      "epoch": 1.2884444444444445,
+      "grad_norm": 1.809848427772522,
+      "learning_rate": 0.00014859430604982206,
+      "loss": 1.2951,
+      "step": 2899
+    },
+    {
+      "epoch": 1.2888888888888888,
+      "grad_norm": 2.80146861076355,
+      "learning_rate": 0.00014857651245551602,
+      "loss": 1.8434,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2893333333333334,
+      "grad_norm": 0.43238645792007446,
+      "learning_rate": 0.00014855871886120998,
+      "loss": 0.0395,
+      "step": 2901
+    },
+    {
+      "epoch": 1.2897777777777777,
+      "grad_norm": 1.3477979898452759,
+      "learning_rate": 0.00014854092526690393,
+      "loss": 2.1978,
+      "step": 2902
+    },
+    {
+      "epoch": 1.2902222222222222,
+      "grad_norm": 1.4748413562774658,
+      "learning_rate": 0.0001485231316725979,
+      "loss": 2.2269,
+      "step": 2903
+    },
+    {
+      "epoch": 1.2906666666666666,
+      "grad_norm": 1.8670775890350342,
+      "learning_rate": 0.00014850533807829182,
+      "loss": 2.5789,
+      "step": 2904
+    },
+    {
+      "epoch": 1.291111111111111,
+      "grad_norm": 1.5996267795562744,
+      "learning_rate": 0.00014848754448398577,
+      "loss": 2.0898,
+      "step": 2905
+    },
+    {
+      "epoch": 1.2915555555555556,
+      "grad_norm": 1.5340416431427002,
+      "learning_rate": 0.0001484697508896797,
+      "loss": 2.0439,
+      "step": 2906
+    },
+    {
+      "epoch": 1.292,
+      "grad_norm": 1.7059005498886108,
+      "learning_rate": 0.00014845195729537366,
+      "loss": 2.6905,
+      "step": 2907
+    },
+    {
+      "epoch": 1.2924444444444445,
+      "grad_norm": 1.3028349876403809,
+      "learning_rate": 0.00014843416370106762,
+      "loss": 1.1318,
+      "step": 2908
+    },
+    {
+      "epoch": 1.2928888888888888,
+      "grad_norm": 1.6985855102539062,
+      "learning_rate": 0.00014841637010676157,
+      "loss": 2.2406,
+      "step": 2909
+    },
+    {
+      "epoch": 1.2933333333333334,
+      "grad_norm": 1.416417121887207,
+      "learning_rate": 0.00014839857651245553,
+      "loss": 2.1449,
+      "step": 2910
+    },
+    {
+      "epoch": 1.2937777777777777,
+      "grad_norm": 1.791305422782898,
+      "learning_rate": 0.00014838078291814949,
+      "loss": 2.4623,
+      "step": 2911
+    },
+    {
+      "epoch": 1.2942222222222222,
+      "grad_norm": 1.3889151811599731,
+      "learning_rate": 0.00014836298932384342,
+      "loss": 1.965,
+      "step": 2912
+    },
+    {
+      "epoch": 1.2946666666666666,
+      "grad_norm": 1.8636940717697144,
+      "learning_rate": 0.00014834519572953737,
+      "loss": 2.9771,
+      "step": 2913
+    },
+    {
+      "epoch": 1.295111111111111,
+      "grad_norm": 1.9207584857940674,
+      "learning_rate": 0.00014832740213523133,
+      "loss": 2.3587,
+      "step": 2914
+    },
+    {
+      "epoch": 1.2955555555555556,
+      "grad_norm": 1.6041591167449951,
+      "learning_rate": 0.00014830960854092528,
+      "loss": 1.986,
+      "step": 2915
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 1.9507296085357666,
+      "learning_rate": 0.00014829181494661924,
+      "loss": 2.7387,
+      "step": 2916
+    },
+    {
+      "epoch": 1.2964444444444445,
+      "grad_norm": 1.7080721855163574,
+      "learning_rate": 0.00014827402135231317,
+      "loss": 2.0368,
+      "step": 2917
+    },
+    {
+      "epoch": 1.2968888888888888,
+      "grad_norm": 1.7469477653503418,
+      "learning_rate": 0.00014825622775800713,
+      "loss": 2.0248,
+      "step": 2918
+    },
+    {
+      "epoch": 1.2973333333333334,
+      "grad_norm": 1.622348666191101,
+      "learning_rate": 0.00014823843416370106,
+      "loss": 2.3361,
+      "step": 2919
+    },
+    {
+      "epoch": 1.2977777777777777,
+      "grad_norm": 1.9793723821640015,
+      "learning_rate": 0.000148220640569395,
+      "loss": 2.7984,
+      "step": 2920
+    },
+    {
+      "epoch": 1.2982222222222222,
+      "grad_norm": 1.6501868963241577,
+      "learning_rate": 0.00014820284697508897,
+      "loss": 1.9731,
+      "step": 2921
+    },
+    {
+      "epoch": 1.2986666666666666,
+      "grad_norm": 1.7134915590286255,
+      "learning_rate": 0.00014818505338078293,
+      "loss": 2.3017,
+      "step": 2922
+    },
+    {
+      "epoch": 1.299111111111111,
+      "grad_norm": 1.6116629838943481,
+      "learning_rate": 0.00014816725978647688,
+      "loss": 2.0448,
+      "step": 2923
+    },
+    {
+      "epoch": 1.2995555555555556,
+      "grad_norm": 1.7651207447052002,
+      "learning_rate": 0.00014814946619217084,
+      "loss": 2.2993,
+      "step": 2924
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.5447845458984375,
+      "learning_rate": 0.00014813167259786477,
+      "loss": 1.6308,
+      "step": 2925
+    },
+    {
+      "epoch": 1.3004444444444445,
+      "grad_norm": 1.505743145942688,
+      "learning_rate": 0.00014811387900355873,
+      "loss": 1.54,
+      "step": 2926
+    },
+    {
+      "epoch": 1.3008888888888888,
+      "grad_norm": 1.8078949451446533,
+      "learning_rate": 0.00014809608540925268,
+      "loss": 2.3919,
+      "step": 2927
+    },
+    {
+      "epoch": 1.3013333333333335,
+      "grad_norm": 1.799320101737976,
+      "learning_rate": 0.00014807829181494664,
+      "loss": 2.7337,
+      "step": 2928
+    },
+    {
+      "epoch": 1.3017777777777777,
+      "grad_norm": 1.7614754438400269,
+      "learning_rate": 0.0001480604982206406,
+      "loss": 2.1453,
+      "step": 2929
+    },
+    {
+      "epoch": 1.3022222222222222,
+      "grad_norm": 1.8952438831329346,
+      "learning_rate": 0.00014804270462633452,
+      "loss": 2.2258,
+      "step": 2930
+    },
+    {
+      "epoch": 1.3026666666666666,
+      "grad_norm": 1.7075976133346558,
+      "learning_rate": 0.00014802491103202848,
+      "loss": 2.0048,
+      "step": 2931
+    },
+    {
+      "epoch": 1.303111111111111,
+      "grad_norm": 1.847962737083435,
+      "learning_rate": 0.0001480071174377224,
+      "loss": 1.7736,
+      "step": 2932
+    },
+    {
+      "epoch": 1.3035555555555556,
+      "grad_norm": 1.7101470232009888,
+      "learning_rate": 0.00014798932384341637,
+      "loss": 2.0625,
+      "step": 2933
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 1.760359764099121,
+      "learning_rate": 0.00014797153024911032,
+      "loss": 1.7883,
+      "step": 2934
+    },
+    {
+      "epoch": 1.3044444444444445,
+      "grad_norm": 2.244199275970459,
+      "learning_rate": 0.00014795373665480428,
+      "loss": 2.3975,
+      "step": 2935
+    },
+    {
+      "epoch": 1.3048888888888888,
+      "grad_norm": 2.010221242904663,
+      "learning_rate": 0.00014793594306049824,
+      "loss": 2.2006,
+      "step": 2936
+    },
+    {
+      "epoch": 1.3053333333333335,
+      "grad_norm": 1.8990859985351562,
+      "learning_rate": 0.0001479181494661922,
+      "loss": 2.0714,
+      "step": 2937
+    },
+    {
+      "epoch": 1.3057777777777777,
+      "grad_norm": 1.8823907375335693,
+      "learning_rate": 0.00014790035587188612,
+      "loss": 1.9876,
+      "step": 2938
+    },
+    {
+      "epoch": 1.3062222222222222,
+      "grad_norm": 1.7615153789520264,
+      "learning_rate": 0.00014788256227758008,
+      "loss": 2.2601,
+      "step": 2939
+    },
+    {
+      "epoch": 1.3066666666666666,
+      "grad_norm": 2.3797504901885986,
+      "learning_rate": 0.00014786476868327403,
+      "loss": 2.3143,
+      "step": 2940
+    },
+    {
+      "epoch": 1.3071111111111111,
+      "grad_norm": 2.3076117038726807,
+      "learning_rate": 0.000147846975088968,
+      "loss": 2.4184,
+      "step": 2941
+    },
+    {
+      "epoch": 1.3075555555555556,
+      "grad_norm": 2.3064980506896973,
+      "learning_rate": 0.00014782918149466195,
+      "loss": 2.0238,
+      "step": 2942
+    },
+    {
+      "epoch": 1.308,
+      "grad_norm": 2.125981330871582,
+      "learning_rate": 0.00014781138790035588,
+      "loss": 2.4634,
+      "step": 2943
+    },
+    {
+      "epoch": 1.3084444444444445,
+      "grad_norm": 2.4080536365509033,
+      "learning_rate": 0.00014779359430604983,
+      "loss": 2.204,
+      "step": 2944
+    },
+    {
+      "epoch": 1.3088888888888888,
+      "grad_norm": 2.146500587463379,
+      "learning_rate": 0.00014777580071174376,
+      "loss": 2.2429,
+      "step": 2945
+    },
+    {
+      "epoch": 1.3093333333333335,
+      "grad_norm": 2.469111680984497,
+      "learning_rate": 0.00014775800711743772,
+      "loss": 2.4771,
+      "step": 2946
+    },
+    {
+      "epoch": 1.3097777777777777,
+      "grad_norm": 2.244917392730713,
+      "learning_rate": 0.00014774021352313168,
+      "loss": 1.8265,
+      "step": 2947
+    },
+    {
+      "epoch": 1.3102222222222222,
+      "grad_norm": 2.1437273025512695,
+      "learning_rate": 0.00014772241992882563,
+      "loss": 2.496,
+      "step": 2948
+    },
+    {
+      "epoch": 1.3106666666666666,
+      "grad_norm": 2.3928580284118652,
+      "learning_rate": 0.0001477046263345196,
+      "loss": 1.9276,
+      "step": 2949
+    },
+    {
+      "epoch": 1.3111111111111111,
+      "grad_norm": 3.0038251876831055,
+      "learning_rate": 0.00014768683274021355,
+      "loss": 2.0514,
+      "step": 2950
+    },
+    {
+      "epoch": 1.3115555555555556,
+      "grad_norm": 1.4333771467208862,
+      "learning_rate": 0.00014766903914590747,
+      "loss": 2.8184,
+      "step": 2951
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 1.254792332649231,
+      "learning_rate": 0.00014765124555160143,
+      "loss": 2.1764,
+      "step": 2952
+    },
+    {
+      "epoch": 1.3124444444444445,
+      "grad_norm": 1.2323510646820068,
+      "learning_rate": 0.0001476334519572954,
+      "loss": 2.4008,
+      "step": 2953
+    },
+    {
+      "epoch": 1.3128888888888888,
+      "grad_norm": 1.026353120803833,
+      "learning_rate": 0.00014761565836298934,
+      "loss": 1.6053,
+      "step": 2954
+    },
+    {
+      "epoch": 1.3133333333333335,
+      "grad_norm": 1.8161970376968384,
+      "learning_rate": 0.0001475978647686833,
+      "loss": 2.4387,
+      "step": 2955
+    },
+    {
+      "epoch": 1.3137777777777777,
+      "grad_norm": 1.568258285522461,
+      "learning_rate": 0.00014758007117437723,
+      "loss": 2.1573,
+      "step": 2956
+    },
+    {
+      "epoch": 1.3142222222222222,
+      "grad_norm": 1.747540831565857,
+      "learning_rate": 0.00014756227758007116,
+      "loss": 2.3005,
+      "step": 2957
+    },
+    {
+      "epoch": 1.3146666666666667,
+      "grad_norm": 1.9220967292785645,
+      "learning_rate": 0.00014754448398576512,
+      "loss": 2.517,
+      "step": 2958
+    },
+    {
+      "epoch": 1.3151111111111111,
+      "grad_norm": 1.7257778644561768,
+      "learning_rate": 0.00014752669039145907,
+      "loss": 2.576,
+      "step": 2959
+    },
+    {
+      "epoch": 1.3155555555555556,
+      "grad_norm": 1.6477985382080078,
+      "learning_rate": 0.00014750889679715303,
+      "loss": 2.1995,
+      "step": 2960
+    },
+    {
+      "epoch": 1.316,
+      "grad_norm": 1.5505272150039673,
+      "learning_rate": 0.00014749110320284699,
+      "loss": 2.346,
+      "step": 2961
+    },
+    {
+      "epoch": 1.3164444444444445,
+      "grad_norm": 1.4658108949661255,
+      "learning_rate": 0.00014747330960854094,
+      "loss": 2.0652,
+      "step": 2962
+    },
+    {
+      "epoch": 1.3168888888888888,
+      "grad_norm": 1.694881558418274,
+      "learning_rate": 0.00014745551601423487,
+      "loss": 2.1266,
+      "step": 2963
+    },
+    {
+      "epoch": 1.3173333333333335,
+      "grad_norm": 1.606266736984253,
+      "learning_rate": 0.00014743772241992883,
+      "loss": 2.0245,
+      "step": 2964
+    },
+    {
+      "epoch": 1.3177777777777777,
+      "grad_norm": 1.483372688293457,
+      "learning_rate": 0.00014741992882562278,
+      "loss": 2.0158,
+      "step": 2965
+    },
+    {
+      "epoch": 1.3182222222222222,
+      "grad_norm": 1.510193943977356,
+      "learning_rate": 0.00014740213523131674,
+      "loss": 1.8094,
+      "step": 2966
+    },
+    {
+      "epoch": 1.3186666666666667,
+      "grad_norm": 1.2403696775436401,
+      "learning_rate": 0.0001473843416370107,
+      "loss": 1.1705,
+      "step": 2967
+    },
+    {
+      "epoch": 1.3191111111111111,
+      "grad_norm": 1.4912583827972412,
+      "learning_rate": 0.00014736654804270465,
+      "loss": 1.9255,
+      "step": 2968
+    },
+    {
+      "epoch": 1.3195555555555556,
+      "grad_norm": 1.7279419898986816,
+      "learning_rate": 0.00014734875444839858,
+      "loss": 1.2055,
+      "step": 2969
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8031781911849976,
+      "learning_rate": 0.0001473309608540925,
+      "loss": 2.2159,
+      "step": 2970
+    },
+    {
+      "epoch": 1.3204444444444445,
+      "grad_norm": 2.015742778778076,
+      "learning_rate": 0.00014731316725978647,
+      "loss": 2.5995,
+      "step": 2971
+    },
+    {
+      "epoch": 1.3208888888888888,
+      "grad_norm": 1.7225654125213623,
+      "learning_rate": 0.00014729537366548043,
+      "loss": 2.2011,
+      "step": 2972
+    },
+    {
+      "epoch": 1.3213333333333335,
+      "grad_norm": 1.881436824798584,
+      "learning_rate": 0.00014727758007117438,
+      "loss": 2.1361,
+      "step": 2973
+    },
+    {
+      "epoch": 1.3217777777777777,
+      "grad_norm": 1.8075528144836426,
+      "learning_rate": 0.00014725978647686834,
+      "loss": 2.4812,
+      "step": 2974
+    },
+    {
+      "epoch": 1.3222222222222222,
+      "grad_norm": 1.7048109769821167,
+      "learning_rate": 0.0001472419928825623,
+      "loss": 2.2206,
+      "step": 2975
+    },
+    {
+      "epoch": 1.3226666666666667,
+      "grad_norm": 1.8709800243377686,
+      "learning_rate": 0.00014722419928825622,
+      "loss": 2.0549,
+      "step": 2976
+    },
+    {
+      "epoch": 1.3231111111111111,
+      "grad_norm": 1.6973004341125488,
+      "learning_rate": 0.00014720640569395018,
+      "loss": 2.0749,
+      "step": 2977
+    },
+    {
+      "epoch": 1.3235555555555556,
+      "grad_norm": 2.0399484634399414,
+      "learning_rate": 0.00014718861209964414,
+      "loss": 2.59,
+      "step": 2978
+    },
+    {
+      "epoch": 1.324,
+      "grad_norm": 1.7553602457046509,
+      "learning_rate": 0.0001471708185053381,
+      "loss": 1.4675,
+      "step": 2979
+    },
+    {
+      "epoch": 1.3244444444444445,
+      "grad_norm": 1.8353267908096313,
+      "learning_rate": 0.00014715302491103205,
+      "loss": 2.1814,
+      "step": 2980
+    },
+    {
+      "epoch": 1.3248888888888888,
+      "grad_norm": 1.942821741104126,
+      "learning_rate": 0.000147135231316726,
+      "loss": 1.7173,
+      "step": 2981
+    },
+    {
+      "epoch": 1.3253333333333333,
+      "grad_norm": 2.025289535522461,
+      "learning_rate": 0.00014711743772241994,
+      "loss": 2.4813,
+      "step": 2982
+    },
+    {
+      "epoch": 1.3257777777777777,
+      "grad_norm": 1.890375018119812,
+      "learning_rate": 0.00014709964412811387,
+      "loss": 2.2062,
+      "step": 2983
+    },
+    {
+      "epoch": 1.3262222222222222,
+      "grad_norm": 1.8974454402923584,
+      "learning_rate": 0.00014708185053380782,
+      "loss": 2.3105,
+      "step": 2984
+    },
+    {
+      "epoch": 1.3266666666666667,
+      "grad_norm": 1.7386442422866821,
+      "learning_rate": 0.00014706405693950178,
+      "loss": 1.9486,
+      "step": 2985
+    },
+    {
+      "epoch": 1.3271111111111111,
+      "grad_norm": 1.974223017692566,
+      "learning_rate": 0.00014704626334519574,
+      "loss": 2.5025,
+      "step": 2986
+    },
+    {
+      "epoch": 1.3275555555555556,
+      "grad_norm": 2.720777988433838,
+      "learning_rate": 0.0001470284697508897,
+      "loss": 2.2163,
+      "step": 2987
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 2.455169200897217,
+      "learning_rate": 0.00014701067615658365,
+      "loss": 2.2537,
+      "step": 2988
+    },
+    {
+      "epoch": 1.3284444444444445,
+      "grad_norm": 1.8268187046051025,
+      "learning_rate": 0.00014699288256227758,
+      "loss": 1.3468,
+      "step": 2989
+    },
+    {
+      "epoch": 1.3288888888888888,
+      "grad_norm": 1.7359619140625,
+      "learning_rate": 0.00014697508896797153,
+      "loss": 2.2044,
+      "step": 2990
+    },
+    {
+      "epoch": 1.3293333333333333,
+      "grad_norm": 2.3060410022735596,
+      "learning_rate": 0.0001469572953736655,
+      "loss": 2.1335,
+      "step": 2991
+    },
+    {
+      "epoch": 1.3297777777777777,
+      "grad_norm": 2.0797529220581055,
+      "learning_rate": 0.00014693950177935945,
+      "loss": 2.4904,
+      "step": 2992
+    },
+    {
+      "epoch": 1.3302222222222222,
+      "grad_norm": 2.0389015674591064,
+      "learning_rate": 0.0001469217081850534,
+      "loss": 2.1793,
+      "step": 2993
+    },
+    {
+      "epoch": 1.3306666666666667,
+      "grad_norm": 1.99758780002594,
+      "learning_rate": 0.00014690391459074736,
+      "loss": 1.7031,
+      "step": 2994
+    },
+    {
+      "epoch": 1.3311111111111111,
+      "grad_norm": 2.201939582824707,
+      "learning_rate": 0.0001468861209964413,
+      "loss": 2.254,
+      "step": 2995
+    },
+    {
+      "epoch": 1.3315555555555556,
+      "grad_norm": 2.535717010498047,
+      "learning_rate": 0.00014686832740213522,
+      "loss": 2.2282,
+      "step": 2996
+    },
+    {
+      "epoch": 1.332,
+      "grad_norm": 2.139512062072754,
+      "learning_rate": 0.00014685053380782918,
+      "loss": 2.1161,
+      "step": 2997
+    },
+    {
+      "epoch": 1.3324444444444445,
+      "grad_norm": 2.589766502380371,
+      "learning_rate": 0.00014683274021352313,
+      "loss": 2.2188,
+      "step": 2998
+    },
+    {
+      "epoch": 1.3328888888888888,
+      "grad_norm": 2.3444082736968994,
+      "learning_rate": 0.0001468149466192171,
+      "loss": 2.3078,
+      "step": 2999
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 2.9983866214752197,
+      "learning_rate": 0.00014679715302491105,
+      "loss": 1.9401,
+      "step": 3000
+    },
+    {
+      "epoch": 1.3337777777777777,
+      "grad_norm": 1.0303854942321777,
+      "learning_rate": 0.000146779359430605,
+      "loss": 1.3335,
+      "step": 3001
+    },
+    {
+      "epoch": 1.3342222222222222,
+      "grad_norm": 1.3651078939437866,
+      "learning_rate": 0.00014676156583629893,
+      "loss": 2.7932,
+      "step": 3002
+    },
+    {
+      "epoch": 1.3346666666666667,
+      "grad_norm": 1.4741727113723755,
+      "learning_rate": 0.0001467437722419929,
+      "loss": 2.4145,
+      "step": 3003
+    },
+    {
+      "epoch": 1.3351111111111111,
+      "grad_norm": 1.308189868927002,
+      "learning_rate": 0.00014672597864768684,
+      "loss": 2.492,
+      "step": 3004
+    },
+    {
+      "epoch": 1.3355555555555556,
+      "grad_norm": 1.5847667455673218,
+      "learning_rate": 0.0001467081850533808,
+      "loss": 2.1946,
+      "step": 3005
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 1.329132318496704,
+      "learning_rate": 0.00014669039145907476,
+      "loss": 1.5467,
+      "step": 3006
+    },
+    {
+      "epoch": 1.3364444444444445,
+      "grad_norm": 1.514951229095459,
+      "learning_rate": 0.00014667259786476869,
+      "loss": 2.048,
+      "step": 3007
+    },
+    {
+      "epoch": 1.3368888888888888,
+      "grad_norm": 1.4534947872161865,
+      "learning_rate": 0.00014665480427046264,
+      "loss": 1.3235,
+      "step": 3008
+    },
+    {
+      "epoch": 1.3373333333333333,
+      "grad_norm": 1.4798189401626587,
+      "learning_rate": 0.00014663701067615657,
+      "loss": 2.0619,
+      "step": 3009
+    },
+    {
+      "epoch": 1.3377777777777777,
+      "grad_norm": 1.5956952571868896,
+      "learning_rate": 0.00014661921708185053,
+      "loss": 2.325,
+      "step": 3010
+    },
+    {
+      "epoch": 1.3382222222222222,
+      "grad_norm": 1.8076585531234741,
+      "learning_rate": 0.00014660142348754449,
+      "loss": 2.8001,
+      "step": 3011
+    },
+    {
+      "epoch": 1.3386666666666667,
+      "grad_norm": 1.6839090585708618,
+      "learning_rate": 0.00014658362989323844,
+      "loss": 2.4422,
+      "step": 3012
+    },
+    {
+      "epoch": 1.3391111111111111,
+      "grad_norm": 1.4808694124221802,
+      "learning_rate": 0.0001465658362989324,
+      "loss": 2.2948,
+      "step": 3013
+    },
+    {
+      "epoch": 1.3395555555555556,
+      "grad_norm": 1.645921230316162,
+      "learning_rate": 0.00014654804270462635,
+      "loss": 2.1147,
+      "step": 3014
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.490327000617981,
+      "learning_rate": 0.00014653024911032028,
+      "loss": 1.5932,
+      "step": 3015
+    },
+    {
+      "epoch": 1.3404444444444445,
+      "grad_norm": 1.7233227491378784,
+      "learning_rate": 0.00014651245551601424,
+      "loss": 2.5041,
+      "step": 3016
+    },
+    {
+      "epoch": 1.3408888888888888,
+      "grad_norm": 1.904835820198059,
+      "learning_rate": 0.0001464946619217082,
+      "loss": 2.4865,
+      "step": 3017
+    },
+    {
+      "epoch": 1.3413333333333333,
+      "grad_norm": 1.3651695251464844,
+      "learning_rate": 0.00014647686832740215,
+      "loss": 1.6397,
+      "step": 3018
+    },
+    {
+      "epoch": 1.3417777777777777,
+      "grad_norm": 1.8083473443984985,
+      "learning_rate": 0.0001464590747330961,
+      "loss": 2.0558,
+      "step": 3019
+    },
+    {
+      "epoch": 1.3422222222222222,
+      "grad_norm": 1.4256452322006226,
+      "learning_rate": 0.00014644128113879004,
+      "loss": 1.4262,
+      "step": 3020
+    },
+    {
+      "epoch": 1.3426666666666667,
+      "grad_norm": 1.686566948890686,
+      "learning_rate": 0.000146423487544484,
+      "loss": 2.235,
+      "step": 3021
+    },
+    {
+      "epoch": 1.3431111111111111,
+      "grad_norm": 1.6832934617996216,
+      "learning_rate": 0.00014640569395017793,
+      "loss": 2.3787,
+      "step": 3022
+    },
+    {
+      "epoch": 1.3435555555555556,
+      "grad_norm": 5.573652267456055,
+      "learning_rate": 0.00014638790035587188,
+      "loss": 2.2192,
+      "step": 3023
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 1.8312151432037354,
+      "learning_rate": 0.00014637010676156584,
+      "loss": 2.4134,
+      "step": 3024
+    },
+    {
+      "epoch": 1.3444444444444446,
+      "grad_norm": 2.171259641647339,
+      "learning_rate": 0.0001463523131672598,
+      "loss": 2.5597,
+      "step": 3025
+    },
+    {
+      "epoch": 1.3448888888888888,
+      "grad_norm": 1.7489248514175415,
+      "learning_rate": 0.00014633451957295375,
+      "loss": 2.0362,
+      "step": 3026
+    },
+    {
+      "epoch": 1.3453333333333333,
+      "grad_norm": 1.915249228477478,
+      "learning_rate": 0.0001463167259786477,
+      "loss": 2.0456,
+      "step": 3027
+    },
+    {
+      "epoch": 1.3457777777777777,
+      "grad_norm": 2.186251640319824,
+      "learning_rate": 0.00014629893238434164,
+      "loss": 2.448,
+      "step": 3028
+    },
+    {
+      "epoch": 1.3462222222222222,
+      "grad_norm": 2.2437589168548584,
+      "learning_rate": 0.0001462811387900356,
+      "loss": 0.093,
+      "step": 3029
+    },
+    {
+      "epoch": 1.3466666666666667,
+      "grad_norm": 1.4108742475509644,
+      "learning_rate": 0.00014626334519572955,
+      "loss": 1.0545,
+      "step": 3030
+    },
+    {
+      "epoch": 1.3471111111111111,
+      "grad_norm": 1.1790105104446411,
+      "learning_rate": 0.0001462455516014235,
+      "loss": 0.8722,
+      "step": 3031
+    },
+    {
+      "epoch": 1.3475555555555556,
+      "grad_norm": 1.8908590078353882,
+      "learning_rate": 0.00014622775800711746,
+      "loss": 2.3223,
+      "step": 3032
+    },
+    {
+      "epoch": 1.3479999999999999,
+      "grad_norm": 1.7565504312515259,
+      "learning_rate": 0.0001462099644128114,
+      "loss": 2.0448,
+      "step": 3033
+    },
+    {
+      "epoch": 1.3484444444444446,
+      "grad_norm": 2.0749552249908447,
+      "learning_rate": 0.00014619217081850535,
+      "loss": 2.4472,
+      "step": 3034
+    },
+    {
+      "epoch": 1.3488888888888888,
+      "grad_norm": 1.9328750371932983,
+      "learning_rate": 0.00014617437722419928,
+      "loss": 2.3066,
+      "step": 3035
+    },
+    {
+      "epoch": 1.3493333333333333,
+      "grad_norm": 1.9503514766693115,
+      "learning_rate": 0.00014615658362989324,
+      "loss": 2.1583,
+      "step": 3036
+    },
+    {
+      "epoch": 1.3497777777777777,
+      "grad_norm": 2.1976773738861084,
+      "learning_rate": 0.0001461387900355872,
+      "loss": 2.6388,
+      "step": 3037
+    },
+    {
+      "epoch": 1.3502222222222222,
+      "grad_norm": 1.8023751974105835,
+      "learning_rate": 0.00014612099644128115,
+      "loss": 2.0306,
+      "step": 3038
+    },
+    {
+      "epoch": 1.3506666666666667,
+      "grad_norm": 2.1533122062683105,
+      "learning_rate": 0.0001461032028469751,
+      "loss": 2.7645,
+      "step": 3039
+    },
+    {
+      "epoch": 1.3511111111111112,
+      "grad_norm": 1.889941930770874,
+      "learning_rate": 0.00014608540925266906,
+      "loss": 2.2246,
+      "step": 3040
+    },
+    {
+      "epoch": 1.3515555555555556,
+      "grad_norm": 2.390805244445801,
+      "learning_rate": 0.000146067615658363,
+      "loss": 2.8298,
+      "step": 3041
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 2.4162545204162598,
+      "learning_rate": 0.00014604982206405695,
+      "loss": 2.1917,
+      "step": 3042
+    },
+    {
+      "epoch": 1.3524444444444446,
+      "grad_norm": 2.0249195098876953,
+      "learning_rate": 0.0001460320284697509,
+      "loss": 2.6536,
+      "step": 3043
+    },
+    {
+      "epoch": 1.3528888888888888,
+      "grad_norm": 1.7109678983688354,
+      "learning_rate": 0.00014601423487544486,
+      "loss": 1.6031,
+      "step": 3044
+    },
+    {
+      "epoch": 1.3533333333333333,
+      "grad_norm": 1.919968843460083,
+      "learning_rate": 0.00014599644128113882,
+      "loss": 2.3276,
+      "step": 3045
+    },
+    {
+      "epoch": 1.3537777777777777,
+      "grad_norm": 2.2520694732666016,
+      "learning_rate": 0.00014597864768683275,
+      "loss": 2.5268,
+      "step": 3046
+    },
+    {
+      "epoch": 1.3542222222222222,
+      "grad_norm": 2.1359968185424805,
+      "learning_rate": 0.00014596085409252668,
+      "loss": 2.3626,
+      "step": 3047
+    },
+    {
+      "epoch": 1.3546666666666667,
+      "grad_norm": 2.3673102855682373,
+      "learning_rate": 0.00014594306049822063,
+      "loss": 2.5428,
+      "step": 3048
+    },
+    {
+      "epoch": 1.3551111111111112,
+      "grad_norm": 2.3437814712524414,
+      "learning_rate": 0.0001459252669039146,
+      "loss": 2.5333,
+      "step": 3049
+    },
+    {
+      "epoch": 1.3555555555555556,
+      "grad_norm": 1.5844188928604126,
+      "learning_rate": 0.00014590747330960854,
+      "loss": 0.678,
+      "step": 3050
+    },
+    {
+      "epoch": 1.3559999999999999,
+      "grad_norm": 0.8454328775405884,
+      "learning_rate": 0.0001458896797153025,
+      "loss": 0.0259,
+      "step": 3051
+    },
+    {
+      "epoch": 1.3564444444444446,
+      "grad_norm": 1.2323882579803467,
+      "learning_rate": 0.00014587188612099646,
+      "loss": 2.408,
+      "step": 3052
+    },
+    {
+      "epoch": 1.3568888888888888,
+      "grad_norm": 1.4226750135421753,
+      "learning_rate": 0.0001458540925266904,
+      "loss": 2.3492,
+      "step": 3053
+    },
+    {
+      "epoch": 1.3573333333333333,
+      "grad_norm": 1.3918040990829468,
+      "learning_rate": 0.00014583629893238434,
+      "loss": 2.2151,
+      "step": 3054
+    },
+    {
+      "epoch": 1.3577777777777778,
+      "grad_norm": 1.5795356035232544,
+      "learning_rate": 0.0001458185053380783,
+      "loss": 2.5257,
+      "step": 3055
+    },
+    {
+      "epoch": 1.3582222222222222,
+      "grad_norm": 1.3857512474060059,
+      "learning_rate": 0.00014580071174377226,
+      "loss": 2.3031,
+      "step": 3056
+    },
+    {
+      "epoch": 1.3586666666666667,
+      "grad_norm": 1.50767183303833,
+      "learning_rate": 0.0001457829181494662,
+      "loss": 1.606,
+      "step": 3057
+    },
+    {
+      "epoch": 1.3591111111111112,
+      "grad_norm": 1.5515854358673096,
+      "learning_rate": 0.00014576512455516017,
+      "loss": 2.4394,
+      "step": 3058
+    },
+    {
+      "epoch": 1.3595555555555556,
+      "grad_norm": 1.567515254020691,
+      "learning_rate": 0.0001457473309608541,
+      "loss": 2.7249,
+      "step": 3059
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 1.5846283435821533,
+      "learning_rate": 0.00014572953736654803,
+      "loss": 1.8917,
+      "step": 3060
+    },
+    {
+      "epoch": 1.3604444444444446,
+      "grad_norm": 1.4666316509246826,
+      "learning_rate": 0.00014571174377224198,
+      "loss": 2.1971,
+      "step": 3061
+    },
+    {
+      "epoch": 1.3608888888888888,
+      "grad_norm": 1.708336353302002,
+      "learning_rate": 0.00014569395017793594,
+      "loss": 2.177,
+      "step": 3062
+    },
+    {
+      "epoch": 1.3613333333333333,
+      "grad_norm": 1.7200583219528198,
+      "learning_rate": 0.0001456761565836299,
+      "loss": 2.3996,
+      "step": 3063
+    },
+    {
+      "epoch": 1.3617777777777778,
+      "grad_norm": 1.7250936031341553,
+      "learning_rate": 0.00014565836298932385,
+      "loss": 2.4772,
+      "step": 3064
+    },
+    {
+      "epoch": 1.3622222222222222,
+      "grad_norm": 1.6220717430114746,
+      "learning_rate": 0.0001456405693950178,
+      "loss": 1.7843,
+      "step": 3065
+    },
+    {
+      "epoch": 1.3626666666666667,
+      "grad_norm": 1.2669525146484375,
+      "learning_rate": 0.00014562277580071174,
+      "loss": 1.0277,
+      "step": 3066
+    },
+    {
+      "epoch": 1.3631111111111112,
+      "grad_norm": 1.7346656322479248,
+      "learning_rate": 0.0001456049822064057,
+      "loss": 1.7548,
+      "step": 3067
+    },
+    {
+      "epoch": 1.3635555555555556,
+      "grad_norm": 1.718342900276184,
+      "learning_rate": 0.00014558718861209965,
+      "loss": 1.7593,
+      "step": 3068
+    },
+    {
+      "epoch": 1.3639999999999999,
+      "grad_norm": 1.738747239112854,
+      "learning_rate": 0.0001455693950177936,
+      "loss": 2.1453,
+      "step": 3069
+    },
+    {
+      "epoch": 1.3644444444444446,
+      "grad_norm": 1.6312371492385864,
+      "learning_rate": 0.00014555160142348757,
+      "loss": 2.1307,
+      "step": 3070
+    },
+    {
+      "epoch": 1.3648888888888888,
+      "grad_norm": 1.9798495769500732,
+      "learning_rate": 0.00014553380782918152,
+      "loss": 1.9325,
+      "step": 3071
+    },
+    {
+      "epoch": 1.3653333333333333,
+      "grad_norm": 1.8481535911560059,
+      "learning_rate": 0.00014551601423487545,
+      "loss": 2.0689,
+      "step": 3072
+    },
+    {
+      "epoch": 1.3657777777777778,
+      "grad_norm": 1.7567691802978516,
+      "learning_rate": 0.00014549822064056938,
+      "loss": 1.8361,
+      "step": 3073
+    },
+    {
+      "epoch": 1.3662222222222222,
+      "grad_norm": 1.9224940538406372,
+      "learning_rate": 0.00014548042704626334,
+      "loss": 2.2146,
+      "step": 3074
+    },
+    {
+      "epoch": 1.3666666666666667,
+      "grad_norm": 2.118567705154419,
+      "learning_rate": 0.0001454626334519573,
+      "loss": 2.7325,
+      "step": 3075
+    },
+    {
+      "epoch": 1.3671111111111112,
+      "grad_norm": 1.6111234426498413,
+      "learning_rate": 0.00014544483985765125,
+      "loss": 1.5926,
+      "step": 3076
+    },
+    {
+      "epoch": 1.3675555555555556,
+      "grad_norm": 1.8418885469436646,
+      "learning_rate": 0.0001454270462633452,
+      "loss": 2.1703,
+      "step": 3077
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 3.9272093772888184,
+      "learning_rate": 0.00014540925266903916,
+      "loss": 0.0625,
+      "step": 3078
+    },
+    {
+      "epoch": 1.3684444444444446,
+      "grad_norm": 1.4438681602478027,
+      "learning_rate": 0.0001453914590747331,
+      "loss": 1.2624,
+      "step": 3079
+    },
+    {
+      "epoch": 1.3688888888888888,
+      "grad_norm": 2.313140392303467,
+      "learning_rate": 0.00014537366548042705,
+      "loss": 2.459,
+      "step": 3080
+    },
+    {
+      "epoch": 1.3693333333333333,
+      "grad_norm": 1.765426754951477,
+      "learning_rate": 0.000145355871886121,
+      "loss": 1.0009,
+      "step": 3081
+    },
+    {
+      "epoch": 1.3697777777777778,
+      "grad_norm": 1.7897223234176636,
+      "learning_rate": 0.00014533807829181496,
+      "loss": 1.6343,
+      "step": 3082
+    },
+    {
+      "epoch": 1.3702222222222222,
+      "grad_norm": 1.9620853662490845,
+      "learning_rate": 0.00014532028469750892,
+      "loss": 1.9386,
+      "step": 3083
+    },
+    {
+      "epoch": 1.3706666666666667,
+      "grad_norm": 2.0378286838531494,
+      "learning_rate": 0.00014530249110320288,
+      "loss": 1.7725,
+      "step": 3084
+    },
+    {
+      "epoch": 1.3711111111111112,
+      "grad_norm": 1.9394813776016235,
+      "learning_rate": 0.0001452846975088968,
+      "loss": 2.1801,
+      "step": 3085
+    },
+    {
+      "epoch": 1.3715555555555556,
+      "grad_norm": 2.0343053340911865,
+      "learning_rate": 0.00014526690391459073,
+      "loss": 2.0472,
+      "step": 3086
+    },
+    {
+      "epoch": 1.3719999999999999,
+      "grad_norm": 2.085235595703125,
+      "learning_rate": 0.0001452491103202847,
+      "loss": 1.969,
+      "step": 3087
+    },
+    {
+      "epoch": 1.3724444444444446,
+      "grad_norm": 1.9965012073516846,
+      "learning_rate": 0.00014523131672597865,
+      "loss": 2.3486,
+      "step": 3088
+    },
+    {
+      "epoch": 1.3728888888888888,
+      "grad_norm": 1.9986323118209839,
+      "learning_rate": 0.0001452135231316726,
+      "loss": 2.2368,
+      "step": 3089
+    },
+    {
+      "epoch": 1.3733333333333333,
+      "grad_norm": 2.003603935241699,
+      "learning_rate": 0.00014519572953736656,
+      "loss": 2.0619,
+      "step": 3090
+    },
+    {
+      "epoch": 1.3737777777777778,
+      "grad_norm": 2.0343897342681885,
+      "learning_rate": 0.00014517793594306052,
+      "loss": 2.2089,
+      "step": 3091
+    },
+    {
+      "epoch": 1.3742222222222222,
+      "grad_norm": 2.0502665042877197,
+      "learning_rate": 0.00014516014234875445,
+      "loss": 2.3439,
+      "step": 3092
+    },
+    {
+      "epoch": 1.3746666666666667,
+      "grad_norm": 2.019620895385742,
+      "learning_rate": 0.0001451423487544484,
+      "loss": 2.2632,
+      "step": 3093
+    },
+    {
+      "epoch": 1.3751111111111112,
+      "grad_norm": 1.9464764595031738,
+      "learning_rate": 0.00014512455516014236,
+      "loss": 1.7843,
+      "step": 3094
+    },
+    {
+      "epoch": 1.3755555555555556,
+      "grad_norm": 2.3599178791046143,
+      "learning_rate": 0.00014510676156583632,
+      "loss": 2.2186,
+      "step": 3095
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 2.2975494861602783,
+      "learning_rate": 0.00014508896797153027,
+      "loss": 2.5936,
+      "step": 3096
+    },
+    {
+      "epoch": 1.3764444444444446,
+      "grad_norm": 2.219733476638794,
+      "learning_rate": 0.00014507117437722423,
+      "loss": 2.0997,
+      "step": 3097
+    },
+    {
+      "epoch": 1.3768888888888888,
+      "grad_norm": 2.035273790359497,
+      "learning_rate": 0.00014505338078291816,
+      "loss": 2.0009,
+      "step": 3098
+    },
+    {
+      "epoch": 1.3773333333333333,
+      "grad_norm": 2.999622106552124,
+      "learning_rate": 0.0001450355871886121,
+      "loss": 2.7129,
+      "step": 3099
+    },
+    {
+      "epoch": 1.3777777777777778,
+      "grad_norm": 2.15091872215271,
+      "learning_rate": 0.00014501779359430604,
+      "loss": 2.1994,
+      "step": 3100
+    },
+    {
+      "epoch": 1.3782222222222222,
+      "grad_norm": 1.0955103635787964,
+      "learning_rate": 0.000145,
+      "loss": 1.416,
+      "step": 3101
+    },
+    {
+      "epoch": 1.3786666666666667,
+      "grad_norm": 1.3353731632232666,
+      "learning_rate": 0.00014498220640569396,
+      "loss": 2.4926,
+      "step": 3102
+    },
+    {
+      "epoch": 1.3791111111111112,
+      "grad_norm": 1.4232149124145508,
+      "learning_rate": 0.00014496441281138791,
+      "loss": 2.5539,
+      "step": 3103
+    },
+    {
+      "epoch": 1.3795555555555556,
+      "grad_norm": 1.4904459714889526,
+      "learning_rate": 0.00014494661921708187,
+      "loss": 2.5516,
+      "step": 3104
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.351136565208435,
+      "learning_rate": 0.0001449288256227758,
+      "loss": 1.9835,
+      "step": 3105
+    },
+    {
+      "epoch": 1.3804444444444444,
+      "grad_norm": 1.4860031604766846,
+      "learning_rate": 0.00014491103202846976,
+      "loss": 1.9876,
+      "step": 3106
+    },
+    {
+      "epoch": 1.3808888888888888,
+      "grad_norm": 1.582760214805603,
+      "learning_rate": 0.0001448932384341637,
+      "loss": 2.6782,
+      "step": 3107
+    },
+    {
+      "epoch": 1.3813333333333333,
+      "grad_norm": 1.3775660991668701,
+      "learning_rate": 0.00014487544483985767,
+      "loss": 1.6808,
+      "step": 3108
+    },
+    {
+      "epoch": 1.3817777777777778,
+      "grad_norm": 1.6874479055404663,
+      "learning_rate": 0.00014485765124555163,
+      "loss": 2.6314,
+      "step": 3109
+    },
+    {
+      "epoch": 1.3822222222222222,
+      "grad_norm": 1.7144551277160645,
+      "learning_rate": 0.00014483985765124558,
+      "loss": 2.7101,
+      "step": 3110
+    },
+    {
+      "epoch": 1.3826666666666667,
+      "grad_norm": 1.5098775625228882,
+      "learning_rate": 0.0001448220640569395,
+      "loss": 1.8929,
+      "step": 3111
+    },
+    {
+      "epoch": 1.3831111111111112,
+      "grad_norm": 1.653300404548645,
+      "learning_rate": 0.00014480427046263344,
+      "loss": 2.1654,
+      "step": 3112
+    },
+    {
+      "epoch": 1.3835555555555556,
+      "grad_norm": 1.4902502298355103,
+      "learning_rate": 0.0001447864768683274,
+      "loss": 2.3472,
+      "step": 3113
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 1.75850248336792,
+      "learning_rate": 0.00014476868327402135,
+      "loss": 1.99,
+      "step": 3114
+    },
+    {
+      "epoch": 1.3844444444444444,
+      "grad_norm": 1.5602999925613403,
+      "learning_rate": 0.0001447508896797153,
+      "loss": 1.8835,
+      "step": 3115
+    },
+    {
+      "epoch": 1.3848888888888888,
+      "grad_norm": 1.6873457431793213,
+      "learning_rate": 0.00014473309608540927,
+      "loss": 1.9495,
+      "step": 3116
+    },
+    {
+      "epoch": 1.3853333333333333,
+      "grad_norm": 1.1590341329574585,
+      "learning_rate": 0.00014471530249110322,
+      "loss": 1.2124,
+      "step": 3117
+    },
+    {
+      "epoch": 1.3857777777777778,
+      "grad_norm": 1.7708086967468262,
+      "learning_rate": 0.00014469750889679715,
+      "loss": 2.0342,
+      "step": 3118
+    },
+    {
+      "epoch": 1.3862222222222222,
+      "grad_norm": 1.969315767288208,
+      "learning_rate": 0.0001446797153024911,
+      "loss": 2.5779,
+      "step": 3119
+    },
+    {
+      "epoch": 1.3866666666666667,
+      "grad_norm": 1.6105482578277588,
+      "learning_rate": 0.00014466192170818507,
+      "loss": 2.2028,
+      "step": 3120
+    },
+    {
+      "epoch": 1.3871111111111112,
+      "grad_norm": 1.7451056241989136,
+      "learning_rate": 0.00014464412811387902,
+      "loss": 2.1619,
+      "step": 3121
+    },
+    {
+      "epoch": 1.3875555555555557,
+      "grad_norm": 1.5029910802841187,
+      "learning_rate": 0.00014462633451957298,
+      "loss": 2.2189,
+      "step": 3122
+    },
+    {
+      "epoch": 1.388,
+      "grad_norm": 1.8065159320831299,
+      "learning_rate": 0.0001446085409252669,
+      "loss": 2.52,
+      "step": 3123
+    },
+    {
+      "epoch": 1.3884444444444444,
+      "grad_norm": 1.7965675592422485,
+      "learning_rate": 0.00014459074733096086,
+      "loss": 2.6056,
+      "step": 3124
+    },
+    {
+      "epoch": 1.3888888888888888,
+      "grad_norm": 1.6334154605865479,
+      "learning_rate": 0.0001445729537366548,
+      "loss": 2.2306,
+      "step": 3125
+    },
+    {
+      "epoch": 1.3893333333333333,
+      "grad_norm": 2.191298484802246,
+      "learning_rate": 0.00014455516014234875,
+      "loss": 2.3345,
+      "step": 3126
+    },
+    {
+      "epoch": 1.3897777777777778,
+      "grad_norm": 2.035778045654297,
+      "learning_rate": 0.0001445373665480427,
+      "loss": 2.2855,
+      "step": 3127
+    },
+    {
+      "epoch": 1.3902222222222222,
+      "grad_norm": 1.8941333293914795,
+      "learning_rate": 0.00014451957295373666,
+      "loss": 2.0976,
+      "step": 3128
+    },
+    {
+      "epoch": 1.3906666666666667,
+      "grad_norm": 1.8983358144760132,
+      "learning_rate": 0.00014450177935943062,
+      "loss": 2.1321,
+      "step": 3129
+    },
+    {
+      "epoch": 1.3911111111111112,
+      "grad_norm": 1.8651962280273438,
+      "learning_rate": 0.00014448398576512458,
+      "loss": 2.2432,
+      "step": 3130
+    },
+    {
+      "epoch": 1.3915555555555557,
+      "grad_norm": 1.7466819286346436,
+      "learning_rate": 0.0001444661921708185,
+      "loss": 1.2187,
+      "step": 3131
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 1.9504824876785278,
+      "learning_rate": 0.00014444839857651246,
+      "loss": 1.0134,
+      "step": 3132
+    },
+    {
+      "epoch": 1.3924444444444444,
+      "grad_norm": 2.349276065826416,
+      "learning_rate": 0.00014443060498220642,
+      "loss": 2.3251,
+      "step": 3133
+    },
+    {
+      "epoch": 1.3928888888888888,
+      "grad_norm": 2.0128836631774902,
+      "learning_rate": 0.00014441281138790038,
+      "loss": 1.8844,
+      "step": 3134
+    },
+    {
+      "epoch": 1.3933333333333333,
+      "grad_norm": 1.8258366584777832,
+      "learning_rate": 0.00014439501779359433,
+      "loss": 2.1697,
+      "step": 3135
+    },
+    {
+      "epoch": 1.3937777777777778,
+      "grad_norm": 1.9964505434036255,
+      "learning_rate": 0.00014437722419928826,
+      "loss": 2.0272,
+      "step": 3136
+    },
+    {
+      "epoch": 1.3942222222222223,
+      "grad_norm": 2.2089779376983643,
+      "learning_rate": 0.0001443594306049822,
+      "loss": 2.4377,
+      "step": 3137
+    },
+    {
+      "epoch": 1.3946666666666667,
+      "grad_norm": 1.9052916765213013,
+      "learning_rate": 0.00014434163701067615,
+      "loss": 1.9119,
+      "step": 3138
+    },
+    {
+      "epoch": 1.3951111111111112,
+      "grad_norm": 2.0558083057403564,
+      "learning_rate": 0.0001443238434163701,
+      "loss": 2.0866,
+      "step": 3139
+    },
+    {
+      "epoch": 1.3955555555555557,
+      "grad_norm": 1.9824244976043701,
+      "learning_rate": 0.00014430604982206406,
+      "loss": 2.1271,
+      "step": 3140
+    },
+    {
+      "epoch": 1.396,
+      "grad_norm": 2.383279800415039,
+      "learning_rate": 0.00014428825622775802,
+      "loss": 2.5828,
+      "step": 3141
+    },
+    {
+      "epoch": 1.3964444444444444,
+      "grad_norm": 2.1160545349121094,
+      "learning_rate": 0.00014427046263345197,
+      "loss": 1.8748,
+      "step": 3142
+    },
+    {
+      "epoch": 1.3968888888888888,
+      "grad_norm": 1.8280696868896484,
+      "learning_rate": 0.0001442526690391459,
+      "loss": 2.0148,
+      "step": 3143
+    },
+    {
+      "epoch": 1.3973333333333333,
+      "grad_norm": 1.842757225036621,
+      "learning_rate": 0.00014423487544483986,
+      "loss": 1.9273,
+      "step": 3144
+    },
+    {
+      "epoch": 1.3977777777777778,
+      "grad_norm": 1.878212809562683,
+      "learning_rate": 0.00014421708185053382,
+      "loss": 1.8468,
+      "step": 3145
+    },
+    {
+      "epoch": 1.3982222222222223,
+      "grad_norm": 2.176372766494751,
+      "learning_rate": 0.00014419928825622777,
+      "loss": 1.7978,
+      "step": 3146
+    },
+    {
+      "epoch": 1.3986666666666667,
+      "grad_norm": 2.247149705886841,
+      "learning_rate": 0.00014418149466192173,
+      "loss": 2.298,
+      "step": 3147
+    },
+    {
+      "epoch": 1.3991111111111112,
+      "grad_norm": 2.320523977279663,
+      "learning_rate": 0.00014416370106761569,
+      "loss": 2.3023,
+      "step": 3148
+    },
+    {
+      "epoch": 1.3995555555555557,
+      "grad_norm": 2.5072226524353027,
+      "learning_rate": 0.00014414590747330961,
+      "loss": 2.0168,
+      "step": 3149
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.0211939811706543,
+      "learning_rate": 0.00014412811387900354,
+      "loss": 2.7225,
+      "step": 3150
+    },
+    {
+      "epoch": 1.4004444444444444,
+      "grad_norm": 1.3240656852722168,
+      "learning_rate": 0.0001441103202846975,
+      "loss": 2.5294,
+      "step": 3151
+    },
+    {
+      "epoch": 1.4008888888888889,
+      "grad_norm": 1.478697657585144,
+      "learning_rate": 0.00014409252669039146,
+      "loss": 2.2188,
+      "step": 3152
+    },
+    {
+      "epoch": 1.4013333333333333,
+      "grad_norm": 1.3129013776779175,
+      "learning_rate": 0.0001440747330960854,
+      "loss": 2.188,
+      "step": 3153
+    },
+    {
+      "epoch": 1.4017777777777778,
+      "grad_norm": 1.450279951095581,
+      "learning_rate": 0.00014405693950177937,
+      "loss": 2.3634,
+      "step": 3154
+    },
+    {
+      "epoch": 1.4022222222222223,
+      "grad_norm": 1.3494071960449219,
+      "learning_rate": 0.00014403914590747333,
+      "loss": 2.2367,
+      "step": 3155
+    },
+    {
+      "epoch": 1.4026666666666667,
+      "grad_norm": 1.4874467849731445,
+      "learning_rate": 0.00014402135231316726,
+      "loss": 2.548,
+      "step": 3156
+    },
+    {
+      "epoch": 1.403111111111111,
+      "grad_norm": 1.3934712409973145,
+      "learning_rate": 0.0001440035587188612,
+      "loss": 1.5867,
+      "step": 3157
+    },
+    {
+      "epoch": 1.4035555555555557,
+      "grad_norm": 1.5048962831497192,
+      "learning_rate": 0.00014398576512455517,
+      "loss": 2.5955,
+      "step": 3158
+    },
+    {
+      "epoch": 1.404,
+      "grad_norm": 1.5615451335906982,
+      "learning_rate": 0.00014396797153024913,
+      "loss": 2.1421,
+      "step": 3159
+    },
+    {
+      "epoch": 1.4044444444444444,
+      "grad_norm": 1.5293431282043457,
+      "learning_rate": 0.00014395017793594308,
+      "loss": 2.3808,
+      "step": 3160
+    },
+    {
+      "epoch": 1.4048888888888889,
+      "grad_norm": 1.3629491329193115,
+      "learning_rate": 0.00014393238434163704,
+      "loss": 1.8447,
+      "step": 3161
+    },
+    {
+      "epoch": 1.4053333333333333,
+      "grad_norm": 1.76398766040802,
+      "learning_rate": 0.00014391459074733097,
+      "loss": 1.8923,
+      "step": 3162
+    },
+    {
+      "epoch": 1.4057777777777778,
+      "grad_norm": 1.6600054502487183,
+      "learning_rate": 0.0001438967971530249,
+      "loss": 2.0265,
+      "step": 3163
+    },
+    {
+      "epoch": 1.4062222222222223,
+      "grad_norm": 1.6202727556228638,
+      "learning_rate": 0.00014387900355871885,
+      "loss": 2.2756,
+      "step": 3164
+    },
+    {
+      "epoch": 1.4066666666666667,
+      "grad_norm": 1.749403953552246,
+      "learning_rate": 0.0001438612099644128,
+      "loss": 2.5024,
+      "step": 3165
+    },
+    {
+      "epoch": 1.407111111111111,
+      "grad_norm": 1.7654697895050049,
+      "learning_rate": 0.00014384341637010677,
+      "loss": 2.4529,
+      "step": 3166
+    },
+    {
+      "epoch": 1.4075555555555557,
+      "grad_norm": 1.5884429216384888,
+      "learning_rate": 0.00014382562277580072,
+      "loss": 1.8367,
+      "step": 3167
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 1.5916013717651367,
+      "learning_rate": 0.00014380782918149468,
+      "loss": 1.8217,
+      "step": 3168
+    },
+    {
+      "epoch": 1.4084444444444444,
+      "grad_norm": 1.7130736112594604,
+      "learning_rate": 0.0001437900355871886,
+      "loss": 2.2377,
+      "step": 3169
+    },
+    {
+      "epoch": 1.4088888888888889,
+      "grad_norm": 1.05029296875,
+      "learning_rate": 0.00014377224199288257,
+      "loss": 0.5225,
+      "step": 3170
+    },
+    {
+      "epoch": 1.4093333333333333,
+      "grad_norm": 1.631998062133789,
+      "learning_rate": 0.00014375444839857652,
+      "loss": 2.1113,
+      "step": 3171
+    },
+    {
+      "epoch": 1.4097777777777778,
+      "grad_norm": 1.6177490949630737,
+      "learning_rate": 0.00014373665480427048,
+      "loss": 2.0905,
+      "step": 3172
+    },
+    {
+      "epoch": 1.4102222222222223,
+      "grad_norm": 1.727180004119873,
+      "learning_rate": 0.00014371886120996443,
+      "loss": 2.2642,
+      "step": 3173
+    },
+    {
+      "epoch": 1.4106666666666667,
+      "grad_norm": 1.755303144454956,
+      "learning_rate": 0.0001437010676156584,
+      "loss": 2.2187,
+      "step": 3174
+    },
+    {
+      "epoch": 1.411111111111111,
+      "grad_norm": 1.6000373363494873,
+      "learning_rate": 0.00014368327402135232,
+      "loss": 1.8188,
+      "step": 3175
+    },
+    {
+      "epoch": 1.4115555555555557,
+      "grad_norm": 2.0754306316375732,
+      "learning_rate": 0.00014366548042704625,
+      "loss": 2.0832,
+      "step": 3176
+    },
+    {
+      "epoch": 1.412,
+      "grad_norm": 1.7767425775527954,
+      "learning_rate": 0.0001436476868327402,
+      "loss": 2.1773,
+      "step": 3177
+    },
+    {
+      "epoch": 1.4124444444444444,
+      "grad_norm": 2.1510021686553955,
+      "learning_rate": 0.00014362989323843416,
+      "loss": 2.1337,
+      "step": 3178
+    },
+    {
+      "epoch": 1.4128888888888889,
+      "grad_norm": 1.9618239402770996,
+      "learning_rate": 0.00014361209964412812,
+      "loss": 2.108,
+      "step": 3179
+    },
+    {
+      "epoch": 1.4133333333333333,
+      "grad_norm": 1.8377591371536255,
+      "learning_rate": 0.00014359430604982208,
+      "loss": 1.7457,
+      "step": 3180
+    },
+    {
+      "epoch": 1.4137777777777778,
+      "grad_norm": 2.0039236545562744,
+      "learning_rate": 0.00014357651245551603,
+      "loss": 2.4462,
+      "step": 3181
+    },
+    {
+      "epoch": 1.4142222222222223,
+      "grad_norm": 1.9533127546310425,
+      "learning_rate": 0.00014355871886120996,
+      "loss": 2.0373,
+      "step": 3182
+    },
+    {
+      "epoch": 1.4146666666666667,
+      "grad_norm": 2.212468385696411,
+      "learning_rate": 0.00014354092526690392,
+      "loss": 2.2246,
+      "step": 3183
+    },
+    {
+      "epoch": 1.415111111111111,
+      "grad_norm": 1.4578838348388672,
+      "learning_rate": 0.00014352313167259788,
+      "loss": 1.1584,
+      "step": 3184
+    },
+    {
+      "epoch": 1.4155555555555557,
+      "grad_norm": 2.0764267444610596,
+      "learning_rate": 0.00014350533807829183,
+      "loss": 1.3613,
+      "step": 3185
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 2.053358316421509,
+      "learning_rate": 0.0001434875444839858,
+      "loss": 2.5823,
+      "step": 3186
+    },
+    {
+      "epoch": 1.4164444444444444,
+      "grad_norm": 0.93979811668396,
+      "learning_rate": 0.00014346975088967974,
+      "loss": 0.0561,
+      "step": 3187
+    },
+    {
+      "epoch": 1.4168888888888889,
+      "grad_norm": 1.6901674270629883,
+      "learning_rate": 0.00014345195729537367,
+      "loss": 1.9685,
+      "step": 3188
+    },
+    {
+      "epoch": 1.4173333333333333,
+      "grad_norm": 1.8892921209335327,
+      "learning_rate": 0.0001434341637010676,
+      "loss": 2.1895,
+      "step": 3189
+    },
+    {
+      "epoch": 1.4177777777777778,
+      "grad_norm": 2.0869650840759277,
+      "learning_rate": 0.00014341637010676156,
+      "loss": 2.1289,
+      "step": 3190
+    },
+    {
+      "epoch": 1.4182222222222223,
+      "grad_norm": 2.154538154602051,
+      "learning_rate": 0.00014339857651245552,
+      "loss": 2.005,
+      "step": 3191
+    },
+    {
+      "epoch": 1.4186666666666667,
+      "grad_norm": 2.2069180011749268,
+      "learning_rate": 0.00014338078291814947,
+      "loss": 2.2602,
+      "step": 3192
+    },
+    {
+      "epoch": 1.419111111111111,
+      "grad_norm": 2.003593921661377,
+      "learning_rate": 0.00014336298932384343,
+      "loss": 1.9024,
+      "step": 3193
+    },
+    {
+      "epoch": 1.4195555555555557,
+      "grad_norm": 2.1326658725738525,
+      "learning_rate": 0.00014334519572953739,
+      "loss": 2.3338,
+      "step": 3194
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.287719488143921,
+      "learning_rate": 0.00014332740213523132,
+      "loss": 2.1353,
+      "step": 3195
+    },
+    {
+      "epoch": 1.4204444444444444,
+      "grad_norm": 1.7974603176116943,
+      "learning_rate": 0.00014330960854092527,
+      "loss": 1.9235,
+      "step": 3196
+    },
+    {
+      "epoch": 1.4208888888888889,
+      "grad_norm": 2.7026476860046387,
+      "learning_rate": 0.00014329181494661923,
+      "loss": 1.8728,
+      "step": 3197
+    },
+    {
+      "epoch": 1.4213333333333333,
+      "grad_norm": 2.0487453937530518,
+      "learning_rate": 0.00014327402135231318,
+      "loss": 2.3309,
+      "step": 3198
+    },
+    {
+      "epoch": 1.4217777777777778,
+      "grad_norm": 2.1622159481048584,
+      "learning_rate": 0.00014325622775800714,
+      "loss": 2.104,
+      "step": 3199
+    },
+    {
+      "epoch": 1.4222222222222223,
+      "grad_norm": 2.4706900119781494,
+      "learning_rate": 0.0001432384341637011,
+      "loss": 2.2835,
+      "step": 3200
+    },
+    {
+      "epoch": 1.4226666666666667,
+      "grad_norm": 0.8910221457481384,
+      "learning_rate": 0.00014322064056939503,
+      "loss": 1.2837,
+      "step": 3201
+    },
+    {
+      "epoch": 1.423111111111111,
+      "grad_norm": 1.3645347356796265,
+      "learning_rate": 0.00014320284697508896,
+      "loss": 2.7213,
+      "step": 3202
+    },
+    {
+      "epoch": 1.4235555555555557,
+      "grad_norm": 1.3697624206542969,
+      "learning_rate": 0.0001431850533807829,
+      "loss": 2.6312,
+      "step": 3203
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 1.5852277278900146,
+      "learning_rate": 0.00014316725978647687,
+      "loss": 2.8356,
+      "step": 3204
+    },
+    {
+      "epoch": 1.4244444444444444,
+      "grad_norm": 1.2539130449295044,
+      "learning_rate": 0.00014314946619217083,
+      "loss": 1.7608,
+      "step": 3205
+    },
+    {
+      "epoch": 1.4248888888888889,
+      "grad_norm": 1.5661532878875732,
+      "learning_rate": 0.00014313167259786478,
+      "loss": 2.0956,
+      "step": 3206
+    },
+    {
+      "epoch": 1.4253333333333333,
+      "grad_norm": 1.6269443035125732,
+      "learning_rate": 0.00014311387900355874,
+      "loss": 2.2012,
+      "step": 3207
+    },
+    {
+      "epoch": 1.4257777777777778,
+      "grad_norm": 1.4714109897613525,
+      "learning_rate": 0.00014309608540925267,
+      "loss": 2.3546,
+      "step": 3208
+    },
+    {
+      "epoch": 1.4262222222222223,
+      "grad_norm": 1.7439886331558228,
+      "learning_rate": 0.00014307829181494662,
+      "loss": 2.7,
+      "step": 3209
+    },
+    {
+      "epoch": 1.4266666666666667,
+      "grad_norm": 1.4173275232315063,
+      "learning_rate": 0.00014306049822064058,
+      "loss": 2.0922,
+      "step": 3210
+    },
+    {
+      "epoch": 1.427111111111111,
+      "grad_norm": 1.5306942462921143,
+      "learning_rate": 0.00014304270462633454,
+      "loss": 2.0992,
+      "step": 3211
+    },
+    {
+      "epoch": 1.4275555555555557,
+      "grad_norm": 1.797987461090088,
+      "learning_rate": 0.0001430249110320285,
+      "loss": 2.9628,
+      "step": 3212
+    },
+    {
+      "epoch": 1.428,
+      "grad_norm": 2.1177406311035156,
+      "learning_rate": 0.00014300711743772245,
+      "loss": 2.1528,
+      "step": 3213
+    },
+    {
+      "epoch": 1.4284444444444444,
+      "grad_norm": 1.593675136566162,
+      "learning_rate": 0.00014298932384341638,
+      "loss": 2.5579,
+      "step": 3214
+    },
+    {
+      "epoch": 1.4288888888888889,
+      "grad_norm": 1.5105654001235962,
+      "learning_rate": 0.0001429715302491103,
+      "loss": 2.1294,
+      "step": 3215
+    },
+    {
+      "epoch": 1.4293333333333333,
+      "grad_norm": 1.520651936531067,
+      "learning_rate": 0.00014295373665480427,
+      "loss": 1.8752,
+      "step": 3216
+    },
+    {
+      "epoch": 1.4297777777777778,
+      "grad_norm": 1.612784504890442,
+      "learning_rate": 0.00014293594306049822,
+      "loss": 2.0609,
+      "step": 3217
+    },
+    {
+      "epoch": 1.4302222222222223,
+      "grad_norm": 1.5184054374694824,
+      "learning_rate": 0.00014291814946619218,
+      "loss": 1.4772,
+      "step": 3218
+    },
+    {
+      "epoch": 1.4306666666666668,
+      "grad_norm": 1.46523916721344,
+      "learning_rate": 0.00014290035587188614,
+      "loss": 1.9029,
+      "step": 3219
+    },
+    {
+      "epoch": 1.431111111111111,
+      "grad_norm": 1.4900418519973755,
+      "learning_rate": 0.0001428825622775801,
+      "loss": 1.5782,
+      "step": 3220
+    },
+    {
+      "epoch": 1.4315555555555557,
+      "grad_norm": 1.8551801443099976,
+      "learning_rate": 0.00014286476868327402,
+      "loss": 2.47,
+      "step": 3221
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 1.8610374927520752,
+      "learning_rate": 0.00014284697508896798,
+      "loss": 2.5916,
+      "step": 3222
+    },
+    {
+      "epoch": 1.4324444444444444,
+      "grad_norm": 1.73556649684906,
+      "learning_rate": 0.00014282918149466193,
+      "loss": 2.2463,
+      "step": 3223
+    },
+    {
+      "epoch": 1.4328888888888889,
+      "grad_norm": 1.574223279953003,
+      "learning_rate": 0.0001428113879003559,
+      "loss": 1.7704,
+      "step": 3224
+    },
+    {
+      "epoch": 1.4333333333333333,
+      "grad_norm": 2.1118242740631104,
+      "learning_rate": 0.00014279359430604985,
+      "loss": 1.9923,
+      "step": 3225
+    },
+    {
+      "epoch": 1.4337777777777778,
+      "grad_norm": 2.1607296466827393,
+      "learning_rate": 0.00014277580071174378,
+      "loss": 2.2045,
+      "step": 3226
+    },
+    {
+      "epoch": 1.4342222222222223,
+      "grad_norm": 1.532531499862671,
+      "learning_rate": 0.0001427580071174377,
+      "loss": 1.904,
+      "step": 3227
+    },
+    {
+      "epoch": 1.4346666666666668,
+      "grad_norm": 2.332968235015869,
+      "learning_rate": 0.00014274021352313166,
+      "loss": 2.3031,
+      "step": 3228
+    },
+    {
+      "epoch": 1.435111111111111,
+      "grad_norm": 1.7639070749282837,
+      "learning_rate": 0.00014272241992882562,
+      "loss": 2.3637,
+      "step": 3229
+    },
+    {
+      "epoch": 1.4355555555555555,
+      "grad_norm": 2.296191453933716,
+      "learning_rate": 0.00014270462633451958,
+      "loss": 1.664,
+      "step": 3230
+    },
+    {
+      "epoch": 1.436,
+      "grad_norm": 1.685143232345581,
+      "learning_rate": 0.00014268683274021353,
+      "loss": 1.9173,
+      "step": 3231
+    },
+    {
+      "epoch": 1.4364444444444444,
+      "grad_norm": 1.8601534366607666,
+      "learning_rate": 0.0001426690391459075,
+      "loss": 2.0097,
+      "step": 3232
+    },
+    {
+      "epoch": 1.4368888888888889,
+      "grad_norm": 1.818623423576355,
+      "learning_rate": 0.00014265124555160142,
+      "loss": 1.8948,
+      "step": 3233
+    },
+    {
+      "epoch": 1.4373333333333334,
+      "grad_norm": 2.0175423622131348,
+      "learning_rate": 0.00014263345195729537,
+      "loss": 2.5014,
+      "step": 3234
+    },
+    {
+      "epoch": 1.4377777777777778,
+      "grad_norm": 1.874712586402893,
+      "learning_rate": 0.00014261565836298933,
+      "loss": 2.4529,
+      "step": 3235
+    },
+    {
+      "epoch": 1.4382222222222223,
+      "grad_norm": 2.350339889526367,
+      "learning_rate": 0.0001425978647686833,
+      "loss": 2.5295,
+      "step": 3236
+    },
+    {
+      "epoch": 1.4386666666666668,
+      "grad_norm": 1.8386290073394775,
+      "learning_rate": 0.00014258007117437724,
+      "loss": 2.4049,
+      "step": 3237
+    },
+    {
+      "epoch": 1.439111111111111,
+      "grad_norm": 1.9711859226226807,
+      "learning_rate": 0.0001425622775800712,
+      "loss": 2.1427,
+      "step": 3238
+    },
+    {
+      "epoch": 1.4395555555555555,
+      "grad_norm": 1.9588954448699951,
+      "learning_rate": 0.00014254448398576513,
+      "loss": 2.0168,
+      "step": 3239
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.025226593017578,
+      "learning_rate": 0.00014252669039145906,
+      "loss": 1.9483,
+      "step": 3240
+    },
+    {
+      "epoch": 1.4404444444444444,
+      "grad_norm": 2.1897053718566895,
+      "learning_rate": 0.00014250889679715302,
+      "loss": 2.3079,
+      "step": 3241
+    },
+    {
+      "epoch": 1.4408888888888889,
+      "grad_norm": 1.9101537466049194,
+      "learning_rate": 0.00014249110320284697,
+      "loss": 1.5792,
+      "step": 3242
+    },
+    {
+      "epoch": 1.4413333333333334,
+      "grad_norm": 2.3152666091918945,
+      "learning_rate": 0.00014247330960854093,
+      "loss": 2.6493,
+      "step": 3243
+    },
+    {
+      "epoch": 1.4417777777777778,
+      "grad_norm": 3.443295478820801,
+      "learning_rate": 0.00014245551601423489,
+      "loss": 2.0788,
+      "step": 3244
+    },
+    {
+      "epoch": 1.4422222222222223,
+      "grad_norm": 2.00852370262146,
+      "learning_rate": 0.00014243772241992884,
+      "loss": 2.2241,
+      "step": 3245
+    },
+    {
+      "epoch": 1.4426666666666668,
+      "grad_norm": 2.1195225715637207,
+      "learning_rate": 0.00014241992882562277,
+      "loss": 2.2187,
+      "step": 3246
+    },
+    {
+      "epoch": 1.443111111111111,
+      "grad_norm": 2.060398817062378,
+      "learning_rate": 0.00014240213523131673,
+      "loss": 1.73,
+      "step": 3247
+    },
+    {
+      "epoch": 1.4435555555555555,
+      "grad_norm": 2.193606376647949,
+      "learning_rate": 0.00014238434163701068,
+      "loss": 2.5191,
+      "step": 3248
+    },
+    {
+      "epoch": 1.444,
+      "grad_norm": 2.3782386779785156,
+      "learning_rate": 0.00014236654804270464,
+      "loss": 2.3936,
+      "step": 3249
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 3.235896348953247,
+      "learning_rate": 0.0001423487544483986,
+      "loss": 2.299,
+      "step": 3250
+    },
+    {
+      "epoch": 1.444888888888889,
+      "grad_norm": 1.3519978523254395,
+      "learning_rate": 0.00014233096085409255,
+      "loss": 2.6515,
+      "step": 3251
+    },
+    {
+      "epoch": 1.4453333333333334,
+      "grad_norm": 1.7427107095718384,
+      "learning_rate": 0.00014231316725978648,
+      "loss": 2.758,
+      "step": 3252
+    },
+    {
+      "epoch": 1.4457777777777778,
+      "grad_norm": 1.2374142408370972,
+      "learning_rate": 0.0001422953736654804,
+      "loss": 2.4892,
+      "step": 3253
+    },
+    {
+      "epoch": 1.4462222222222223,
+      "grad_norm": 1.4171687364578247,
+      "learning_rate": 0.00014227758007117437,
+      "loss": 2.5246,
+      "step": 3254
+    },
+    {
+      "epoch": 1.4466666666666668,
+      "grad_norm": 1.9048703908920288,
+      "learning_rate": 0.00014225978647686833,
+      "loss": 2.5152,
+      "step": 3255
+    },
+    {
+      "epoch": 1.447111111111111,
+      "grad_norm": 1.6656217575073242,
+      "learning_rate": 0.00014224199288256228,
+      "loss": 2.3669,
+      "step": 3256
+    },
+    {
+      "epoch": 1.4475555555555555,
+      "grad_norm": 1.9828662872314453,
+      "learning_rate": 0.00014222419928825624,
+      "loss": 2.5827,
+      "step": 3257
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 1.593224287033081,
+      "learning_rate": 0.0001422064056939502,
+      "loss": 2.3798,
+      "step": 3258
+    },
+    {
+      "epoch": 1.4484444444444444,
+      "grad_norm": 1.7066659927368164,
+      "learning_rate": 0.00014218861209964412,
+      "loss": 2.5127,
+      "step": 3259
+    },
+    {
+      "epoch": 1.448888888888889,
+      "grad_norm": 1.5722315311431885,
+      "learning_rate": 0.00014217081850533808,
+      "loss": 1.6095,
+      "step": 3260
+    },
+    {
+      "epoch": 1.4493333333333334,
+      "grad_norm": 1.5399599075317383,
+      "learning_rate": 0.00014215302491103204,
+      "loss": 2.2268,
+      "step": 3261
+    },
+    {
+      "epoch": 1.4497777777777778,
+      "grad_norm": 2.0712404251098633,
+      "learning_rate": 0.000142135231316726,
+      "loss": 1.9943,
+      "step": 3262
+    },
+    {
+      "epoch": 1.4502222222222223,
+      "grad_norm": 1.9809646606445312,
+      "learning_rate": 0.00014211743772241995,
+      "loss": 2.8128,
+      "step": 3263
+    },
+    {
+      "epoch": 1.4506666666666668,
+      "grad_norm": 1.5712720155715942,
+      "learning_rate": 0.0001420996441281139,
+      "loss": 2.3628,
+      "step": 3264
+    },
+    {
+      "epoch": 1.451111111111111,
+      "grad_norm": 1.8636808395385742,
+      "learning_rate": 0.00014208185053380784,
+      "loss": 2.5732,
+      "step": 3265
+    },
+    {
+      "epoch": 1.4515555555555555,
+      "grad_norm": 1.9150162935256958,
+      "learning_rate": 0.00014206405693950177,
+      "loss": 2.2289,
+      "step": 3266
+    },
+    {
+      "epoch": 1.452,
+      "grad_norm": 1.132552146911621,
+      "learning_rate": 0.00014204626334519572,
+      "loss": 0.6418,
+      "step": 3267
+    },
+    {
+      "epoch": 1.4524444444444444,
+      "grad_norm": 1.7738691568374634,
+      "learning_rate": 0.00014202846975088968,
+      "loss": 2.2081,
+      "step": 3268
+    },
+    {
+      "epoch": 1.452888888888889,
+      "grad_norm": 1.6669455766677856,
+      "learning_rate": 0.00014201067615658364,
+      "loss": 1.912,
+      "step": 3269
+    },
+    {
+      "epoch": 1.4533333333333334,
+      "grad_norm": 1.720966100692749,
+      "learning_rate": 0.0001419928825622776,
+      "loss": 2.7566,
+      "step": 3270
+    },
+    {
+      "epoch": 1.4537777777777778,
+      "grad_norm": 1.593934178352356,
+      "learning_rate": 0.00014197508896797155,
+      "loss": 2.076,
+      "step": 3271
+    },
+    {
+      "epoch": 1.4542222222222223,
+      "grad_norm": 1.9219107627868652,
+      "learning_rate": 0.00014195729537366548,
+      "loss": 2.3327,
+      "step": 3272
+    },
+    {
+      "epoch": 1.4546666666666668,
+      "grad_norm": 1.694810390472412,
+      "learning_rate": 0.00014193950177935943,
+      "loss": 2.2732,
+      "step": 3273
+    },
+    {
+      "epoch": 1.455111111111111,
+      "grad_norm": 2.009329319000244,
+      "learning_rate": 0.0001419217081850534,
+      "loss": 1.7474,
+      "step": 3274
+    },
+    {
+      "epoch": 1.4555555555555555,
+      "grad_norm": 1.8578834533691406,
+      "learning_rate": 0.00014190391459074735,
+      "loss": 2.2934,
+      "step": 3275
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 1.8823623657226562,
+      "learning_rate": 0.0001418861209964413,
+      "loss": 2.2761,
+      "step": 3276
+    },
+    {
+      "epoch": 1.4564444444444444,
+      "grad_norm": 1.817597508430481,
+      "learning_rate": 0.00014186832740213526,
+      "loss": 1.9425,
+      "step": 3277
+    },
+    {
+      "epoch": 1.456888888888889,
+      "grad_norm": 2.0119707584381104,
+      "learning_rate": 0.0001418505338078292,
+      "loss": 2.2159,
+      "step": 3278
+    },
+    {
+      "epoch": 1.4573333333333334,
+      "grad_norm": 1.5498830080032349,
+      "learning_rate": 0.00014183274021352312,
+      "loss": 1.4838,
+      "step": 3279
+    },
+    {
+      "epoch": 1.4577777777777778,
+      "grad_norm": 1.7245508432388306,
+      "learning_rate": 0.00014181494661921708,
+      "loss": 2.362,
+      "step": 3280
+    },
+    {
+      "epoch": 1.458222222222222,
+      "grad_norm": 1.6693973541259766,
+      "learning_rate": 0.00014179715302491103,
+      "loss": 2.0165,
+      "step": 3281
+    },
+    {
+      "epoch": 1.4586666666666668,
+      "grad_norm": 1.8107409477233887,
+      "learning_rate": 0.000141779359430605,
+      "loss": 2.0808,
+      "step": 3282
+    },
+    {
+      "epoch": 1.459111111111111,
+      "grad_norm": 1.8532699346542358,
+      "learning_rate": 0.00014176156583629894,
+      "loss": 2.2752,
+      "step": 3283
+    },
+    {
+      "epoch": 1.4595555555555555,
+      "grad_norm": 1.8230167627334595,
+      "learning_rate": 0.0001417437722419929,
+      "loss": 2.1949,
+      "step": 3284
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.083483934402466,
+      "learning_rate": 0.00014172597864768683,
+      "loss": 2.4207,
+      "step": 3285
+    },
+    {
+      "epoch": 1.4604444444444444,
+      "grad_norm": 2.386479377746582,
+      "learning_rate": 0.0001417081850533808,
+      "loss": 2.4775,
+      "step": 3286
+    },
+    {
+      "epoch": 1.460888888888889,
+      "grad_norm": 1.8848226070404053,
+      "learning_rate": 0.00014169039145907474,
+      "loss": 2.1974,
+      "step": 3287
+    },
+    {
+      "epoch": 1.4613333333333334,
+      "grad_norm": 1.7829835414886475,
+      "learning_rate": 0.0001416725978647687,
+      "loss": 1.9449,
+      "step": 3288
+    },
+    {
+      "epoch": 1.4617777777777778,
+      "grad_norm": 1.645369529724121,
+      "learning_rate": 0.00014165480427046266,
+      "loss": 1.5355,
+      "step": 3289
+    },
+    {
+      "epoch": 1.462222222222222,
+      "grad_norm": 2.256089687347412,
+      "learning_rate": 0.0001416370106761566,
+      "loss": 2.5316,
+      "step": 3290
+    },
+    {
+      "epoch": 1.4626666666666668,
+      "grad_norm": 2.052887201309204,
+      "learning_rate": 0.00014161921708185054,
+      "loss": 2.3761,
+      "step": 3291
+    },
+    {
+      "epoch": 1.463111111111111,
+      "grad_norm": 2.1068687438964844,
+      "learning_rate": 0.00014160142348754447,
+      "loss": 0.9723,
+      "step": 3292
+    },
+    {
+      "epoch": 1.4635555555555555,
+      "grad_norm": 2.4331436157226562,
+      "learning_rate": 0.00014158362989323843,
+      "loss": 2.1379,
+      "step": 3293
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 1.9529526233673096,
+      "learning_rate": 0.00014156583629893239,
+      "loss": 2.0011,
+      "step": 3294
+    },
+    {
+      "epoch": 1.4644444444444444,
+      "grad_norm": 2.1232481002807617,
+      "learning_rate": 0.00014154804270462634,
+      "loss": 1.8544,
+      "step": 3295
+    },
+    {
+      "epoch": 1.464888888888889,
+      "grad_norm": 2.0463287830352783,
+      "learning_rate": 0.0001415302491103203,
+      "loss": 2.1042,
+      "step": 3296
+    },
+    {
+      "epoch": 1.4653333333333334,
+      "grad_norm": 2.2398715019226074,
+      "learning_rate": 0.00014151245551601425,
+      "loss": 2.389,
+      "step": 3297
+    },
+    {
+      "epoch": 1.4657777777777778,
+      "grad_norm": 2.3587806224823,
+      "learning_rate": 0.00014149466192170818,
+      "loss": 2.2351,
+      "step": 3298
+    },
+    {
+      "epoch": 1.466222222222222,
+      "grad_norm": 2.272650957107544,
+      "learning_rate": 0.00014147686832740214,
+      "loss": 2.1772,
+      "step": 3299
+    },
+    {
+      "epoch": 1.4666666666666668,
+      "grad_norm": 1.661880373954773,
+      "learning_rate": 0.0001414590747330961,
+      "loss": 1.2369,
+      "step": 3300
+    },
+    {
+      "epoch": 1.467111111111111,
+      "grad_norm": 1.3112317323684692,
+      "learning_rate": 0.00014144128113879005,
+      "loss": 2.3523,
+      "step": 3301
+    },
+    {
+      "epoch": 1.4675555555555555,
+      "grad_norm": 1.5207730531692505,
+      "learning_rate": 0.000141423487544484,
+      "loss": 2.3686,
+      "step": 3302
+    },
+    {
+      "epoch": 1.468,
+      "grad_norm": 1.6390271186828613,
+      "learning_rate": 0.00014140569395017797,
+      "loss": 0.0503,
+      "step": 3303
+    },
+    {
+      "epoch": 1.4684444444444444,
+      "grad_norm": 1.316325068473816,
+      "learning_rate": 0.0001413879003558719,
+      "loss": 2.2079,
+      "step": 3304
+    },
+    {
+      "epoch": 1.468888888888889,
+      "grad_norm": 1.6727405786514282,
+      "learning_rate": 0.00014137010676156583,
+      "loss": 2.8864,
+      "step": 3305
+    },
+    {
+      "epoch": 1.4693333333333334,
+      "grad_norm": 1.413974404335022,
+      "learning_rate": 0.00014135231316725978,
+      "loss": 1.4162,
+      "step": 3306
+    },
+    {
+      "epoch": 1.4697777777777778,
+      "grad_norm": 1.4699324369430542,
+      "learning_rate": 0.00014133451957295374,
+      "loss": 2.0043,
+      "step": 3307
+    },
+    {
+      "epoch": 1.470222222222222,
+      "grad_norm": 1.9841383695602417,
+      "learning_rate": 0.0001413167259786477,
+      "loss": 2.5486,
+      "step": 3308
+    },
+    {
+      "epoch": 1.4706666666666668,
+      "grad_norm": 1.5038107633590698,
+      "learning_rate": 0.00014129893238434165,
+      "loss": 2.3013,
+      "step": 3309
+    },
+    {
+      "epoch": 1.471111111111111,
+      "grad_norm": 1.5381580591201782,
+      "learning_rate": 0.0001412811387900356,
+      "loss": 2.1017,
+      "step": 3310
+    },
+    {
+      "epoch": 1.4715555555555555,
+      "grad_norm": 1.4826030731201172,
+      "learning_rate": 0.00014126334519572954,
+      "loss": 2.4797,
+      "step": 3311
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 1.4599792957305908,
+      "learning_rate": 0.0001412455516014235,
+      "loss": 1.9834,
+      "step": 3312
+    },
+    {
+      "epoch": 1.4724444444444444,
+      "grad_norm": 1.4552773237228394,
+      "learning_rate": 0.00014122775800711745,
+      "loss": 2.2435,
+      "step": 3313
+    },
+    {
+      "epoch": 1.472888888888889,
+      "grad_norm": 1.4118120670318604,
+      "learning_rate": 0.0001412099644128114,
+      "loss": 1.735,
+      "step": 3314
+    },
+    {
+      "epoch": 1.4733333333333334,
+      "grad_norm": 1.5949891805648804,
+      "learning_rate": 0.00014119217081850536,
+      "loss": 1.9896,
+      "step": 3315
+    },
+    {
+      "epoch": 1.4737777777777779,
+      "grad_norm": 1.7730779647827148,
+      "learning_rate": 0.0001411743772241993,
+      "loss": 2.0223,
+      "step": 3316
+    },
+    {
+      "epoch": 1.474222222222222,
+      "grad_norm": 1.7242622375488281,
+      "learning_rate": 0.00014115658362989322,
+      "loss": 2.3709,
+      "step": 3317
+    },
+    {
+      "epoch": 1.4746666666666668,
+      "grad_norm": 1.8402231931686401,
+      "learning_rate": 0.00014113879003558718,
+      "loss": 2.3489,
+      "step": 3318
+    },
+    {
+      "epoch": 1.475111111111111,
+      "grad_norm": 1.328906774520874,
+      "learning_rate": 0.00014112099644128113,
+      "loss": 1.0891,
+      "step": 3319
+    },
+    {
+      "epoch": 1.4755555555555555,
+      "grad_norm": 1.8179643154144287,
+      "learning_rate": 0.0001411032028469751,
+      "loss": 2.6636,
+      "step": 3320
+    },
+    {
+      "epoch": 1.476,
+      "grad_norm": 1.6425268650054932,
+      "learning_rate": 0.00014108540925266905,
+      "loss": 1.9906,
+      "step": 3321
+    },
+    {
+      "epoch": 1.4764444444444444,
+      "grad_norm": 1.7107009887695312,
+      "learning_rate": 0.000141067615658363,
+      "loss": 2.066,
+      "step": 3322
+    },
+    {
+      "epoch": 1.476888888888889,
+      "grad_norm": 1.7221518754959106,
+      "learning_rate": 0.00014104982206405693,
+      "loss": 2.4627,
+      "step": 3323
+    },
+    {
+      "epoch": 1.4773333333333334,
+      "grad_norm": 1.9194080829620361,
+      "learning_rate": 0.0001410320284697509,
+      "loss": 2.2223,
+      "step": 3324
+    },
+    {
+      "epoch": 1.4777777777777779,
+      "grad_norm": 1.7251501083374023,
+      "learning_rate": 0.00014101423487544485,
+      "loss": 1.902,
+      "step": 3325
+    },
+    {
+      "epoch": 1.478222222222222,
+      "grad_norm": 1.7051130533218384,
+      "learning_rate": 0.0001409964412811388,
+      "loss": 2.2801,
+      "step": 3326
+    },
+    {
+      "epoch": 1.4786666666666668,
+      "grad_norm": 2.0462424755096436,
+      "learning_rate": 0.00014097864768683276,
+      "loss": 2.2611,
+      "step": 3327
+    },
+    {
+      "epoch": 1.479111111111111,
+      "grad_norm": 1.6997382640838623,
+      "learning_rate": 0.00014096085409252672,
+      "loss": 2.0672,
+      "step": 3328
+    },
+    {
+      "epoch": 1.4795555555555555,
+      "grad_norm": 1.790878415107727,
+      "learning_rate": 0.00014094306049822065,
+      "loss": 1.866,
+      "step": 3329
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.1614255905151367,
+      "learning_rate": 0.00014092526690391458,
+      "loss": 2.691,
+      "step": 3330
+    },
+    {
+      "epoch": 1.4804444444444445,
+      "grad_norm": 1.456119418144226,
+      "learning_rate": 0.00014090747330960853,
+      "loss": 0.8112,
+      "step": 3331
+    },
+    {
+      "epoch": 1.480888888888889,
+      "grad_norm": 1.8652554750442505,
+      "learning_rate": 0.0001408896797153025,
+      "loss": 1.9741,
+      "step": 3332
+    },
+    {
+      "epoch": 1.4813333333333334,
+      "grad_norm": 1.9750827550888062,
+      "learning_rate": 0.00014087188612099644,
+      "loss": 2.1358,
+      "step": 3333
+    },
+    {
+      "epoch": 1.4817777777777779,
+      "grad_norm": 2.0224530696868896,
+      "learning_rate": 0.0001408540925266904,
+      "loss": 2.2299,
+      "step": 3334
+    },
+    {
+      "epoch": 1.482222222222222,
+      "grad_norm": 2.1135432720184326,
+      "learning_rate": 0.00014083629893238436,
+      "loss": 2.2098,
+      "step": 3335
+    },
+    {
+      "epoch": 1.4826666666666668,
+      "grad_norm": 1.8072072267532349,
+      "learning_rate": 0.0001408185053380783,
+      "loss": 1.9179,
+      "step": 3336
+    },
+    {
+      "epoch": 1.483111111111111,
+      "grad_norm": 2.2474746704101562,
+      "learning_rate": 0.00014080071174377224,
+      "loss": 2.6121,
+      "step": 3337
+    },
+    {
+      "epoch": 1.4835555555555555,
+      "grad_norm": 1.997774600982666,
+      "learning_rate": 0.0001407829181494662,
+      "loss": 1.8134,
+      "step": 3338
+    },
+    {
+      "epoch": 1.484,
+      "grad_norm": 1.8671613931655884,
+      "learning_rate": 0.00014076512455516016,
+      "loss": 1.8315,
+      "step": 3339
+    },
+    {
+      "epoch": 1.4844444444444445,
+      "grad_norm": 1.809183955192566,
+      "learning_rate": 0.0001407473309608541,
+      "loss": 1.9693,
+      "step": 3340
+    },
+    {
+      "epoch": 1.484888888888889,
+      "grad_norm": 2.1424343585968018,
+      "learning_rate": 0.00014072953736654807,
+      "loss": 1.7883,
+      "step": 3341
+    },
+    {
+      "epoch": 1.4853333333333334,
+      "grad_norm": 2.0938260555267334,
+      "learning_rate": 0.000140711743772242,
+      "loss": 2.2282,
+      "step": 3342
+    },
+    {
+      "epoch": 1.4857777777777779,
+      "grad_norm": 2.3834707736968994,
+      "learning_rate": 0.00014069395017793593,
+      "loss": 2.1236,
+      "step": 3343
+    },
+    {
+      "epoch": 1.4862222222222221,
+      "grad_norm": 2.215244770050049,
+      "learning_rate": 0.00014067615658362988,
+      "loss": 2.3009,
+      "step": 3344
+    },
+    {
+      "epoch": 1.4866666666666668,
+      "grad_norm": 1.9744479656219482,
+      "learning_rate": 0.00014065836298932384,
+      "loss": 2.0397,
+      "step": 3345
+    },
+    {
+      "epoch": 1.487111111111111,
+      "grad_norm": 2.0358409881591797,
+      "learning_rate": 0.0001406405693950178,
+      "loss": 1.7655,
+      "step": 3346
+    },
+    {
+      "epoch": 1.4875555555555555,
+      "grad_norm": 2.4060535430908203,
+      "learning_rate": 0.00014062277580071175,
+      "loss": 2.1293,
+      "step": 3347
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 2.8066680431365967,
+      "learning_rate": 0.0001406049822064057,
+      "loss": 2.1486,
+      "step": 3348
+    },
+    {
+      "epoch": 1.4884444444444445,
+      "grad_norm": 2.352820873260498,
+      "learning_rate": 0.00014058718861209964,
+      "loss": 1.4603,
+      "step": 3349
+    },
+    {
+      "epoch": 1.488888888888889,
+      "grad_norm": 2.9096052646636963,
+      "learning_rate": 0.0001405693950177936,
+      "loss": 2.8709,
+      "step": 3350
+    },
+    {
+      "epoch": 1.4893333333333334,
+      "grad_norm": 1.335901141166687,
+      "learning_rate": 0.00014055160142348755,
+      "loss": 2.7017,
+      "step": 3351
+    },
+    {
+      "epoch": 1.4897777777777779,
+      "grad_norm": 1.4894367456436157,
+      "learning_rate": 0.0001405338078291815,
+      "loss": 2.3371,
+      "step": 3352
+    },
+    {
+      "epoch": 1.4902222222222221,
+      "grad_norm": 1.6448569297790527,
+      "learning_rate": 0.00014051601423487547,
+      "loss": 2.3155,
+      "step": 3353
+    },
+    {
+      "epoch": 1.4906666666666666,
+      "grad_norm": 1.5843652486801147,
+      "learning_rate": 0.00014049822064056942,
+      "loss": 2.0601,
+      "step": 3354
+    },
+    {
+      "epoch": 1.491111111111111,
+      "grad_norm": 1.6974892616271973,
+      "learning_rate": 0.00014048042704626335,
+      "loss": 2.651,
+      "step": 3355
+    },
+    {
+      "epoch": 1.4915555555555555,
+      "grad_norm": 1.653384804725647,
+      "learning_rate": 0.00014046263345195728,
+      "loss": 2.0151,
+      "step": 3356
+    },
+    {
+      "epoch": 1.492,
+      "grad_norm": 1.6719948053359985,
+      "learning_rate": 0.00014044483985765124,
+      "loss": 2.2734,
+      "step": 3357
+    },
+    {
+      "epoch": 1.4924444444444445,
+      "grad_norm": 1.239280104637146,
+      "learning_rate": 0.0001404270462633452,
+      "loss": 0.9993,
+      "step": 3358
+    },
+    {
+      "epoch": 1.492888888888889,
+      "grad_norm": 1.8192747831344604,
+      "learning_rate": 0.00014040925266903915,
+      "loss": 2.4339,
+      "step": 3359
+    },
+    {
+      "epoch": 1.4933333333333334,
+      "grad_norm": 1.7065457105636597,
+      "learning_rate": 0.0001403914590747331,
+      "loss": 2.1031,
+      "step": 3360
+    },
+    {
+      "epoch": 1.4937777777777779,
+      "grad_norm": 1.8094873428344727,
+      "learning_rate": 0.00014037366548042706,
+      "loss": 2.3386,
+      "step": 3361
+    },
+    {
+      "epoch": 1.4942222222222221,
+      "grad_norm": 1.534125566482544,
+      "learning_rate": 0.000140355871886121,
+      "loss": 1.7805,
+      "step": 3362
+    },
+    {
+      "epoch": 1.4946666666666666,
+      "grad_norm": 1.619163990020752,
+      "learning_rate": 0.00014033807829181495,
+      "loss": 2.4424,
+      "step": 3363
+    },
+    {
+      "epoch": 1.495111111111111,
+      "grad_norm": 1.7250571250915527,
+      "learning_rate": 0.0001403202846975089,
+      "loss": 2.0668,
+      "step": 3364
+    },
+    {
+      "epoch": 1.4955555555555555,
+      "grad_norm": 1.4816625118255615,
+      "learning_rate": 0.00014030249110320286,
+      "loss": 2.0556,
+      "step": 3365
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 1.794979214668274,
+      "learning_rate": 0.00014028469750889682,
+      "loss": 1.9561,
+      "step": 3366
+    },
+    {
+      "epoch": 1.4964444444444445,
+      "grad_norm": 1.855759859085083,
+      "learning_rate": 0.00014026690391459078,
+      "loss": 2.3782,
+      "step": 3367
+    },
+    {
+      "epoch": 1.496888888888889,
+      "grad_norm": 1.706645131111145,
+      "learning_rate": 0.0001402491103202847,
+      "loss": 2.3613,
+      "step": 3368
+    },
+    {
+      "epoch": 1.4973333333333334,
+      "grad_norm": 1.6514323949813843,
+      "learning_rate": 0.00014023131672597863,
+      "loss": 2.3841,
+      "step": 3369
+    },
+    {
+      "epoch": 1.4977777777777779,
+      "grad_norm": 1.9299840927124023,
+      "learning_rate": 0.0001402135231316726,
+      "loss": 1.8054,
+      "step": 3370
+    },
+    {
+      "epoch": 1.4982222222222221,
+      "grad_norm": 1.72597336769104,
+      "learning_rate": 0.00014019572953736655,
+      "loss": 2.2748,
+      "step": 3371
+    },
+    {
+      "epoch": 1.4986666666666666,
+      "grad_norm": 1.791800618171692,
+      "learning_rate": 0.0001401779359430605,
+      "loss": 1.8032,
+      "step": 3372
+    },
+    {
+      "epoch": 1.499111111111111,
+      "grad_norm": 1.779707431793213,
+      "learning_rate": 0.00014016014234875446,
+      "loss": 2.1661,
+      "step": 3373
+    },
+    {
+      "epoch": 1.4995555555555555,
+      "grad_norm": 1.7183303833007812,
+      "learning_rate": 0.00014014234875444842,
+      "loss": 1.9356,
+      "step": 3374
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.637531042098999,
+      "learning_rate": 0.00014012455516014235,
+      "loss": 1.9201,
+      "step": 3375
+    },
+    {
+      "epoch": 1.5004444444444445,
+      "grad_norm": 1.4911421537399292,
+      "learning_rate": 0.0001401067615658363,
+      "loss": 1.2015,
+      "step": 3376
+    },
+    {
+      "epoch": 1.500888888888889,
+      "grad_norm": 1.6465941667556763,
+      "learning_rate": 0.00014008896797153026,
+      "loss": 1.8771,
+      "step": 3377
+    },
+    {
+      "epoch": 1.5013333333333332,
+      "grad_norm": 1.7987436056137085,
+      "learning_rate": 0.00014007117437722422,
+      "loss": 2.0635,
+      "step": 3378
+    },
+    {
+      "epoch": 1.5017777777777779,
+      "grad_norm": 1.6817963123321533,
+      "learning_rate": 0.00014005338078291817,
+      "loss": 1.8847,
+      "step": 3379
+    },
+    {
+      "epoch": 1.5022222222222221,
+      "grad_norm": 1.6715887784957886,
+      "learning_rate": 0.00014003558718861213,
+      "loss": 1.8009,
+      "step": 3380
+    },
+    {
+      "epoch": 1.5026666666666668,
+      "grad_norm": 2.189204216003418,
+      "learning_rate": 0.00014001779359430606,
+      "loss": 2.1705,
+      "step": 3381
+    },
+    {
+      "epoch": 1.503111111111111,
+      "grad_norm": 1.7621302604675293,
+      "learning_rate": 0.00014,
+      "loss": 2.3973,
+      "step": 3382
+    },
+    {
+      "epoch": 1.5035555555555555,
+      "grad_norm": 1.9194334745407104,
+      "learning_rate": 0.00013998220640569394,
+      "loss": 2.1278,
+      "step": 3383
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 2.001845121383667,
+      "learning_rate": 0.0001399644128113879,
+      "loss": 2.6398,
+      "step": 3384
+    },
+    {
+      "epoch": 1.5044444444444445,
+      "grad_norm": 1.9402610063552856,
+      "learning_rate": 0.00013994661921708186,
+      "loss": 1.8872,
+      "step": 3385
+    },
+    {
+      "epoch": 1.504888888888889,
+      "grad_norm": 1.9671640396118164,
+      "learning_rate": 0.00013992882562277581,
+      "loss": 2.1836,
+      "step": 3386
+    },
+    {
+      "epoch": 1.5053333333333332,
+      "grad_norm": 2.1770575046539307,
+      "learning_rate": 0.00013991103202846977,
+      "loss": 2.1509,
+      "step": 3387
+    },
+    {
+      "epoch": 1.5057777777777779,
+      "grad_norm": 2.059763193130493,
+      "learning_rate": 0.0001398932384341637,
+      "loss": 2.3197,
+      "step": 3388
+    },
+    {
+      "epoch": 1.5062222222222221,
+      "grad_norm": 1.899730920791626,
+      "learning_rate": 0.00013987544483985766,
+      "loss": 1.9101,
+      "step": 3389
+    },
+    {
+      "epoch": 1.5066666666666668,
+      "grad_norm": 1.6671397686004639,
+      "learning_rate": 0.0001398576512455516,
+      "loss": 1.8031,
+      "step": 3390
+    },
+    {
+      "epoch": 1.507111111111111,
+      "grad_norm": 2.2725398540496826,
+      "learning_rate": 0.00013983985765124557,
+      "loss": 2.2501,
+      "step": 3391
+    },
+    {
+      "epoch": 1.5075555555555555,
+      "grad_norm": 2.0739402770996094,
+      "learning_rate": 0.00013982206405693953,
+      "loss": 2.5371,
+      "step": 3392
+    },
+    {
+      "epoch": 1.508,
+      "grad_norm": 2.2705914974212646,
+      "learning_rate": 0.00013980427046263348,
+      "loss": 2.1084,
+      "step": 3393
+    },
+    {
+      "epoch": 1.5084444444444445,
+      "grad_norm": 1.981615662574768,
+      "learning_rate": 0.0001397864768683274,
+      "loss": 1.9832,
+      "step": 3394
+    },
+    {
+      "epoch": 1.508888888888889,
+      "grad_norm": 2.1815102100372314,
+      "learning_rate": 0.00013976868327402134,
+      "loss": 2.1987,
+      "step": 3395
+    },
+    {
+      "epoch": 1.5093333333333332,
+      "grad_norm": 1.9982526302337646,
+      "learning_rate": 0.0001397508896797153,
+      "loss": 2.1131,
+      "step": 3396
+    },
+    {
+      "epoch": 1.5097777777777779,
+      "grad_norm": 2.5864624977111816,
+      "learning_rate": 0.00013973309608540925,
+      "loss": 2.7484,
+      "step": 3397
+    },
+    {
+      "epoch": 1.5102222222222221,
+      "grad_norm": 2.7735939025878906,
+      "learning_rate": 0.0001397153024911032,
+      "loss": 2.8055,
+      "step": 3398
+    },
+    {
+      "epoch": 1.5106666666666668,
+      "grad_norm": 2.547844648361206,
+      "learning_rate": 0.00013969750889679717,
+      "loss": 2.8584,
+      "step": 3399
+    },
+    {
+      "epoch": 1.511111111111111,
+      "grad_norm": 3.193415880203247,
+      "learning_rate": 0.0001396797153024911,
+      "loss": 1.5981,
+      "step": 3400
+    },
+    {
+      "epoch": 1.5115555555555555,
+      "grad_norm": 1.2286021709442139,
+      "learning_rate": 0.00013966192170818505,
+      "loss": 2.0801,
+      "step": 3401
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 1.3546948432922363,
+      "learning_rate": 0.000139644128113879,
+      "loss": 2.2488,
+      "step": 3402
+    },
+    {
+      "epoch": 1.5124444444444445,
+      "grad_norm": 1.3589816093444824,
+      "learning_rate": 0.00013962633451957297,
+      "loss": 2.5642,
+      "step": 3403
+    },
+    {
+      "epoch": 1.512888888888889,
+      "grad_norm": 0.9513995051383972,
+      "learning_rate": 0.00013960854092526692,
+      "loss": 0.5388,
+      "step": 3404
+    },
+    {
+      "epoch": 1.5133333333333332,
+      "grad_norm": 1.2379084825515747,
+      "learning_rate": 0.00013959074733096088,
+      "loss": 1.2631,
+      "step": 3405
+    },
+    {
+      "epoch": 1.5137777777777779,
+      "grad_norm": 1.4364656209945679,
+      "learning_rate": 0.0001395729537366548,
+      "loss": 2.2967,
+      "step": 3406
+    },
+    {
+      "epoch": 1.5142222222222221,
+      "grad_norm": 1.6733119487762451,
+      "learning_rate": 0.00013955516014234874,
+      "loss": 2.4779,
+      "step": 3407
+    },
+    {
+      "epoch": 1.5146666666666668,
+      "grad_norm": 1.715740442276001,
+      "learning_rate": 0.0001395373665480427,
+      "loss": 2.4899,
+      "step": 3408
+    },
+    {
+      "epoch": 1.515111111111111,
+      "grad_norm": 1.6031004190444946,
+      "learning_rate": 0.00013951957295373665,
+      "loss": 2.3027,
+      "step": 3409
+    },
+    {
+      "epoch": 1.5155555555555555,
+      "grad_norm": 1.6325712203979492,
+      "learning_rate": 0.0001395017793594306,
+      "loss": 1.7408,
+      "step": 3410
+    },
+    {
+      "epoch": 1.516,
+      "grad_norm": 1.7524162530899048,
+      "learning_rate": 0.00013948398576512456,
+      "loss": 2.1623,
+      "step": 3411
+    },
+    {
+      "epoch": 1.5164444444444445,
+      "grad_norm": 1.493108868598938,
+      "learning_rate": 0.00013946619217081852,
+      "loss": 2.1662,
+      "step": 3412
+    },
+    {
+      "epoch": 1.516888888888889,
+      "grad_norm": 1.7417209148406982,
+      "learning_rate": 0.00013944839857651245,
+      "loss": 2.3913,
+      "step": 3413
+    },
+    {
+      "epoch": 1.5173333333333332,
+      "grad_norm": 1.671183705329895,
+      "learning_rate": 0.0001394306049822064,
+      "loss": 2.4024,
+      "step": 3414
+    },
+    {
+      "epoch": 1.517777777777778,
+      "grad_norm": 1.567742943763733,
+      "learning_rate": 0.00013941281138790036,
+      "loss": 1.9279,
+      "step": 3415
+    },
+    {
+      "epoch": 1.5182222222222221,
+      "grad_norm": 1.7501814365386963,
+      "learning_rate": 0.00013939501779359432,
+      "loss": 1.9253,
+      "step": 3416
+    },
+    {
+      "epoch": 1.5186666666666668,
+      "grad_norm": 1.5426925420761108,
+      "learning_rate": 0.00013937722419928828,
+      "loss": 2.021,
+      "step": 3417
+    },
+    {
+      "epoch": 1.519111111111111,
+      "grad_norm": 1.4181113243103027,
+      "learning_rate": 0.00013935943060498223,
+      "loss": 1.803,
+      "step": 3418
+    },
+    {
+      "epoch": 1.5195555555555555,
+      "grad_norm": 1.6199541091918945,
+      "learning_rate": 0.00013934163701067616,
+      "loss": 1.8904,
+      "step": 3419
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.6284871101379395,
+      "learning_rate": 0.0001393238434163701,
+      "loss": 2.1618,
+      "step": 3420
+    },
+    {
+      "epoch": 1.5204444444444445,
+      "grad_norm": 2.004983425140381,
+      "learning_rate": 0.00013930604982206405,
+      "loss": 2.4985,
+      "step": 3421
+    },
+    {
+      "epoch": 1.520888888888889,
+      "grad_norm": 1.890509843826294,
+      "learning_rate": 0.000139288256227758,
+      "loss": 2.0206,
+      "step": 3422
+    },
+    {
+      "epoch": 1.5213333333333332,
+      "grad_norm": 2.1789512634277344,
+      "learning_rate": 0.00013927046263345196,
+      "loss": 1.9493,
+      "step": 3423
+    },
+    {
+      "epoch": 1.521777777777778,
+      "grad_norm": 1.6540831327438354,
+      "learning_rate": 0.00013925266903914592,
+      "loss": 2.0598,
+      "step": 3424
+    },
+    {
+      "epoch": 1.5222222222222221,
+      "grad_norm": 2.1028473377227783,
+      "learning_rate": 0.00013923487544483987,
+      "loss": 2.4147,
+      "step": 3425
+    },
+    {
+      "epoch": 1.5226666666666666,
+      "grad_norm": 2.1696505546569824,
+      "learning_rate": 0.0001392170818505338,
+      "loss": 2.3895,
+      "step": 3426
+    },
+    {
+      "epoch": 1.523111111111111,
+      "grad_norm": 2.0171515941619873,
+      "learning_rate": 0.00013919928825622776,
+      "loss": 2.0155,
+      "step": 3427
+    },
+    {
+      "epoch": 1.5235555555555556,
+      "grad_norm": 2.0088951587677,
+      "learning_rate": 0.00013918149466192172,
+      "loss": 2.1746,
+      "step": 3428
+    },
+    {
+      "epoch": 1.524,
+      "grad_norm": 1.8271888494491577,
+      "learning_rate": 0.00013916370106761567,
+      "loss": 2.2367,
+      "step": 3429
+    },
+    {
+      "epoch": 1.5244444444444445,
+      "grad_norm": 2.0240986347198486,
+      "learning_rate": 0.00013914590747330963,
+      "loss": 2.5284,
+      "step": 3430
+    },
+    {
+      "epoch": 1.524888888888889,
+      "grad_norm": 1.7613776922225952,
+      "learning_rate": 0.00013912811387900359,
+      "loss": 1.8526,
+      "step": 3431
+    },
+    {
+      "epoch": 1.5253333333333332,
+      "grad_norm": 2.417299509048462,
+      "learning_rate": 0.00013911032028469751,
+      "loss": 2.6705,
+      "step": 3432
+    },
+    {
+      "epoch": 1.525777777777778,
+      "grad_norm": 2.0129454135894775,
+      "learning_rate": 0.00013909252669039144,
+      "loss": 1.8315,
+      "step": 3433
+    },
+    {
+      "epoch": 1.5262222222222221,
+      "grad_norm": 2.165886640548706,
+      "learning_rate": 0.0001390747330960854,
+      "loss": 2.5886,
+      "step": 3434
+    },
+    {
+      "epoch": 1.5266666666666666,
+      "grad_norm": 2.1423709392547607,
+      "learning_rate": 0.00013905693950177936,
+      "loss": 1.9441,
+      "step": 3435
+    },
+    {
+      "epoch": 1.527111111111111,
+      "grad_norm": 1.9875420331954956,
+      "learning_rate": 0.0001390391459074733,
+      "loss": 2.1317,
+      "step": 3436
+    },
+    {
+      "epoch": 1.5275555555555556,
+      "grad_norm": 2.0450797080993652,
+      "learning_rate": 0.00013902135231316727,
+      "loss": 2.0809,
+      "step": 3437
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 2.3254358768463135,
+      "learning_rate": 0.00013900355871886123,
+      "loss": 2.5318,
+      "step": 3438
+    },
+    {
+      "epoch": 1.5284444444444445,
+      "grad_norm": 2.041480779647827,
+      "learning_rate": 0.00013898576512455516,
+      "loss": 1.9015,
+      "step": 3439
+    },
+    {
+      "epoch": 1.528888888888889,
+      "grad_norm": 1.6092534065246582,
+      "learning_rate": 0.0001389679715302491,
+      "loss": 1.5966,
+      "step": 3440
+    },
+    {
+      "epoch": 1.5293333333333332,
+      "grad_norm": 2.0477304458618164,
+      "learning_rate": 0.00013895017793594307,
+      "loss": 2.1176,
+      "step": 3441
+    },
+    {
+      "epoch": 1.529777777777778,
+      "grad_norm": 2.8084466457366943,
+      "learning_rate": 0.00013893238434163703,
+      "loss": 2.5453,
+      "step": 3442
+    },
+    {
+      "epoch": 1.5302222222222222,
+      "grad_norm": 2.2902886867523193,
+      "learning_rate": 0.00013891459074733098,
+      "loss": 2.0898,
+      "step": 3443
+    },
+    {
+      "epoch": 1.5306666666666666,
+      "grad_norm": 2.420135259628296,
+      "learning_rate": 0.00013889679715302494,
+      "loss": 2.3878,
+      "step": 3444
+    },
+    {
+      "epoch": 1.531111111111111,
+      "grad_norm": 3.70841646194458,
+      "learning_rate": 0.00013887900355871887,
+      "loss": 2.4154,
+      "step": 3445
+    },
+    {
+      "epoch": 1.5315555555555556,
+      "grad_norm": 2.6717746257781982,
+      "learning_rate": 0.0001388612099644128,
+      "loss": 2.1492,
+      "step": 3446
+    },
+    {
+      "epoch": 1.532,
+      "grad_norm": 2.6898179054260254,
+      "learning_rate": 0.00013884341637010675,
+      "loss": 2.1948,
+      "step": 3447
+    },
+    {
+      "epoch": 1.5324444444444445,
+      "grad_norm": 2.706569194793701,
+      "learning_rate": 0.0001388256227758007,
+      "loss": 2.5001,
+      "step": 3448
+    },
+    {
+      "epoch": 1.532888888888889,
+      "grad_norm": 2.5174717903137207,
+      "learning_rate": 0.00013880782918149467,
+      "loss": 2.0004,
+      "step": 3449
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 2.5441904067993164,
+      "learning_rate": 0.00013879003558718862,
+      "loss": 2.0386,
+      "step": 3450
+    },
+    {
+      "epoch": 1.533777777777778,
+      "grad_norm": 1.2441266775131226,
+      "learning_rate": 0.00013877224199288258,
+      "loss": 2.3708,
+      "step": 3451
+    },
+    {
+      "epoch": 1.5342222222222222,
+      "grad_norm": 1.6138944625854492,
+      "learning_rate": 0.0001387544483985765,
+      "loss": 2.2968,
+      "step": 3452
+    },
+    {
+      "epoch": 1.5346666666666666,
+      "grad_norm": 1.7886635065078735,
+      "learning_rate": 0.00013873665480427047,
+      "loss": 2.3651,
+      "step": 3453
+    },
+    {
+      "epoch": 1.535111111111111,
+      "grad_norm": 1.5253329277038574,
+      "learning_rate": 0.00013871886120996442,
+      "loss": 2.409,
+      "step": 3454
+    },
+    {
+      "epoch": 1.5355555555555556,
+      "grad_norm": 1.6435463428497314,
+      "learning_rate": 0.00013870106761565838,
+      "loss": 2.1088,
+      "step": 3455
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 1.7751388549804688,
+      "learning_rate": 0.00013868327402135233,
+      "loss": 1.8769,
+      "step": 3456
+    },
+    {
+      "epoch": 1.5364444444444443,
+      "grad_norm": 1.4366142749786377,
+      "learning_rate": 0.0001386654804270463,
+      "loss": 1.3134,
+      "step": 3457
+    },
+    {
+      "epoch": 1.536888888888889,
+      "grad_norm": 1.9377721548080444,
+      "learning_rate": 0.00013864768683274022,
+      "loss": 2.3063,
+      "step": 3458
+    },
+    {
+      "epoch": 1.5373333333333332,
+      "grad_norm": 2.1702098846435547,
+      "learning_rate": 0.00013862989323843415,
+      "loss": 1.8274,
+      "step": 3459
+    },
+    {
+      "epoch": 1.537777777777778,
+      "grad_norm": 1.566851019859314,
+      "learning_rate": 0.0001386120996441281,
+      "loss": 2.2409,
+      "step": 3460
+    },
+    {
+      "epoch": 1.5382222222222222,
+      "grad_norm": 1.5840020179748535,
+      "learning_rate": 0.00013859430604982206,
+      "loss": 2.676,
+      "step": 3461
+    },
+    {
+      "epoch": 1.5386666666666666,
+      "grad_norm": 1.6341493129730225,
+      "learning_rate": 0.00013857651245551602,
+      "loss": 1.7889,
+      "step": 3462
+    },
+    {
+      "epoch": 1.539111111111111,
+      "grad_norm": 1.810591220855713,
+      "learning_rate": 0.00013855871886120998,
+      "loss": 2.312,
+      "step": 3463
+    },
+    {
+      "epoch": 1.5395555555555556,
+      "grad_norm": 1.6743979454040527,
+      "learning_rate": 0.00013854092526690393,
+      "loss": 2.1647,
+      "step": 3464
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6081904172897339,
+      "learning_rate": 0.00013852313167259786,
+      "loss": 2.3812,
+      "step": 3465
+    },
+    {
+      "epoch": 1.5404444444444443,
+      "grad_norm": 1.6036180257797241,
+      "learning_rate": 0.00013850533807829182,
+      "loss": 2.1979,
+      "step": 3466
+    },
+    {
+      "epoch": 1.540888888888889,
+      "grad_norm": 1.4913078546524048,
+      "learning_rate": 0.00013848754448398578,
+      "loss": 2.0216,
+      "step": 3467
+    },
+    {
+      "epoch": 1.5413333333333332,
+      "grad_norm": 1.6572915315628052,
+      "learning_rate": 0.00013846975088967973,
+      "loss": 2.0528,
+      "step": 3468
+    },
+    {
+      "epoch": 1.541777777777778,
+      "grad_norm": 2.2356021404266357,
+      "learning_rate": 0.0001384519572953737,
+      "loss": 2.1856,
+      "step": 3469
+    },
+    {
+      "epoch": 1.5422222222222222,
+      "grad_norm": 1.9612963199615479,
+      "learning_rate": 0.00013843416370106764,
+      "loss": 2.4308,
+      "step": 3470
+    },
+    {
+      "epoch": 1.5426666666666666,
+      "grad_norm": 1.8817795515060425,
+      "learning_rate": 0.00013841637010676157,
+      "loss": 2.1989,
+      "step": 3471
+    },
+    {
+      "epoch": 1.543111111111111,
+      "grad_norm": 1.5524616241455078,
+      "learning_rate": 0.0001383985765124555,
+      "loss": 1.9362,
+      "step": 3472
+    },
+    {
+      "epoch": 1.5435555555555556,
+      "grad_norm": 1.8420137166976929,
+      "learning_rate": 0.00013838078291814946,
+      "loss": 2.0204,
+      "step": 3473
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 1.6466306447982788,
+      "learning_rate": 0.00013836298932384342,
+      "loss": 2.0033,
+      "step": 3474
+    },
+    {
+      "epoch": 1.5444444444444443,
+      "grad_norm": 1.7015817165374756,
+      "learning_rate": 0.00013834519572953737,
+      "loss": 1.986,
+      "step": 3475
+    },
+    {
+      "epoch": 1.544888888888889,
+      "grad_norm": 2.1093199253082275,
+      "learning_rate": 0.00013832740213523133,
+      "loss": 2.3153,
+      "step": 3476
+    },
+    {
+      "epoch": 1.5453333333333332,
+      "grad_norm": 1.966652274131775,
+      "learning_rate": 0.00013830960854092529,
+      "loss": 2.2379,
+      "step": 3477
+    },
+    {
+      "epoch": 1.545777777777778,
+      "grad_norm": 1.2320265769958496,
+      "learning_rate": 0.00013829181494661922,
+      "loss": 0.9829,
+      "step": 3478
+    },
+    {
+      "epoch": 1.5462222222222222,
+      "grad_norm": 2.0316035747528076,
+      "learning_rate": 0.00013827402135231317,
+      "loss": 2.2463,
+      "step": 3479
+    },
+    {
+      "epoch": 1.5466666666666666,
+      "grad_norm": 1.9726709127426147,
+      "learning_rate": 0.00013825622775800713,
+      "loss": 1.8361,
+      "step": 3480
+    },
+    {
+      "epoch": 1.547111111111111,
+      "grad_norm": 1.76983642578125,
+      "learning_rate": 0.00013823843416370108,
+      "loss": 2.151,
+      "step": 3481
+    },
+    {
+      "epoch": 1.5475555555555556,
+      "grad_norm": 1.8415701389312744,
+      "learning_rate": 0.00013822064056939504,
+      "loss": 2.0344,
+      "step": 3482
+    },
+    {
+      "epoch": 1.548,
+      "grad_norm": 2.0059757232666016,
+      "learning_rate": 0.000138202846975089,
+      "loss": 2.1534,
+      "step": 3483
+    },
+    {
+      "epoch": 1.5484444444444443,
+      "grad_norm": 1.8996038436889648,
+      "learning_rate": 0.00013818505338078293,
+      "loss": 2.3525,
+      "step": 3484
+    },
+    {
+      "epoch": 1.548888888888889,
+      "grad_norm": 2.0750856399536133,
+      "learning_rate": 0.00013816725978647686,
+      "loss": 2.0834,
+      "step": 3485
+    },
+    {
+      "epoch": 1.5493333333333332,
+      "grad_norm": 2.5925469398498535,
+      "learning_rate": 0.0001381494661921708,
+      "loss": 2.6209,
+      "step": 3486
+    },
+    {
+      "epoch": 1.549777777777778,
+      "grad_norm": 2.2502434253692627,
+      "learning_rate": 0.00013813167259786477,
+      "loss": 2.4877,
+      "step": 3487
+    },
+    {
+      "epoch": 1.5502222222222222,
+      "grad_norm": 2.1007752418518066,
+      "learning_rate": 0.00013811387900355873,
+      "loss": 2.164,
+      "step": 3488
+    },
+    {
+      "epoch": 1.5506666666666666,
+      "grad_norm": 2.5511112213134766,
+      "learning_rate": 0.00013809608540925268,
+      "loss": 1.7419,
+      "step": 3489
+    },
+    {
+      "epoch": 1.551111111111111,
+      "grad_norm": 2.076032876968384,
+      "learning_rate": 0.0001380782918149466,
+      "loss": 2.2016,
+      "step": 3490
+    },
+    {
+      "epoch": 1.5515555555555556,
+      "grad_norm": 2.3464345932006836,
+      "learning_rate": 0.00013806049822064057,
+      "loss": 2.828,
+      "step": 3491
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 2.0291099548339844,
+      "learning_rate": 0.00013804270462633452,
+      "loss": 1.6438,
+      "step": 3492
+    },
+    {
+      "epoch": 1.5524444444444443,
+      "grad_norm": 2.3025505542755127,
+      "learning_rate": 0.00013802491103202848,
+      "loss": 1.8687,
+      "step": 3493
+    },
+    {
+      "epoch": 1.552888888888889,
+      "grad_norm": 2.7315847873687744,
+      "learning_rate": 0.00013800711743772244,
+      "loss": 2.6435,
+      "step": 3494
+    },
+    {
+      "epoch": 1.5533333333333332,
+      "grad_norm": 2.493013620376587,
+      "learning_rate": 0.0001379893238434164,
+      "loss": 2.3683,
+      "step": 3495
+    },
+    {
+      "epoch": 1.553777777777778,
+      "grad_norm": 2.1778078079223633,
+      "learning_rate": 0.00013797153024911032,
+      "loss": 2.2781,
+      "step": 3496
+    },
+    {
+      "epoch": 1.5542222222222222,
+      "grad_norm": 2.7922496795654297,
+      "learning_rate": 0.00013795373665480425,
+      "loss": 2.0727,
+      "step": 3497
+    },
+    {
+      "epoch": 1.5546666666666666,
+      "grad_norm": 2.76652193069458,
+      "learning_rate": 0.0001379359430604982,
+      "loss": 1.6055,
+      "step": 3498
+    },
+    {
+      "epoch": 1.555111111111111,
+      "grad_norm": 2.2120234966278076,
+      "learning_rate": 0.00013791814946619217,
+      "loss": 1.9275,
+      "step": 3499
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 3.5710928440093994,
+      "learning_rate": 0.00013790035587188612,
+      "loss": 2.7359,
+      "step": 3500
+    },
+    {
+      "epoch": 1.556,
+      "grad_norm": 1.3622146844863892,
+      "learning_rate": 0.00013788256227758008,
+      "loss": 2.8565,
+      "step": 3501
+    },
+    {
+      "epoch": 1.5564444444444443,
+      "grad_norm": 1.3293352127075195,
+      "learning_rate": 0.00013786476868327404,
+      "loss": 2.1254,
+      "step": 3502
+    },
+    {
+      "epoch": 1.556888888888889,
+      "grad_norm": 1.4484037160873413,
+      "learning_rate": 0.00013784697508896797,
+      "loss": 2.5264,
+      "step": 3503
+    },
+    {
+      "epoch": 1.5573333333333332,
+      "grad_norm": 1.5083986520767212,
+      "learning_rate": 0.00013782918149466192,
+      "loss": 2.212,
+      "step": 3504
+    },
+    {
+      "epoch": 1.557777777777778,
+      "grad_norm": 1.6028261184692383,
+      "learning_rate": 0.00013781138790035588,
+      "loss": 2.1341,
+      "step": 3505
+    },
+    {
+      "epoch": 1.5582222222222222,
+      "grad_norm": 1.4829325675964355,
+      "learning_rate": 0.00013779359430604983,
+      "loss": 1.8982,
+      "step": 3506
+    },
+    {
+      "epoch": 1.5586666666666666,
+      "grad_norm": 1.533338189125061,
+      "learning_rate": 0.0001377758007117438,
+      "loss": 2.2298,
+      "step": 3507
+    },
+    {
+      "epoch": 1.5591111111111111,
+      "grad_norm": 1.5170714855194092,
+      "learning_rate": 0.00013775800711743775,
+      "loss": 1.9902,
+      "step": 3508
+    },
+    {
+      "epoch": 1.5595555555555556,
+      "grad_norm": 1.5905641317367554,
+      "learning_rate": 0.00013774021352313168,
+      "loss": 2.1633,
+      "step": 3509
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.601283311843872,
+      "learning_rate": 0.0001377224199288256,
+      "loss": 2.1138,
+      "step": 3510
+    },
+    {
+      "epoch": 1.5604444444444443,
+      "grad_norm": 1.5424052476882935,
+      "learning_rate": 0.00013770462633451956,
+      "loss": 1.9056,
+      "step": 3511
+    },
+    {
+      "epoch": 1.560888888888889,
+      "grad_norm": 1.7833845615386963,
+      "learning_rate": 0.00013768683274021352,
+      "loss": 2.0569,
+      "step": 3512
+    },
+    {
+      "epoch": 1.5613333333333332,
+      "grad_norm": 1.729130506515503,
+      "learning_rate": 0.00013766903914590748,
+      "loss": 2.2351,
+      "step": 3513
+    },
+    {
+      "epoch": 1.561777777777778,
+      "grad_norm": 1.7116620540618896,
+      "learning_rate": 0.00013765124555160143,
+      "loss": 2.7517,
+      "step": 3514
+    },
+    {
+      "epoch": 1.5622222222222222,
+      "grad_norm": 1.9057334661483765,
+      "learning_rate": 0.0001376334519572954,
+      "loss": 2.7395,
+      "step": 3515
+    },
+    {
+      "epoch": 1.5626666666666666,
+      "grad_norm": 1.9144783020019531,
+      "learning_rate": 0.00013761565836298932,
+      "loss": 2.2624,
+      "step": 3516
+    },
+    {
+      "epoch": 1.5631111111111111,
+      "grad_norm": 1.756967306137085,
+      "learning_rate": 0.00013759786476868327,
+      "loss": 1.5217,
+      "step": 3517
+    },
+    {
+      "epoch": 1.5635555555555556,
+      "grad_norm": 2.028092861175537,
+      "learning_rate": 0.00013758007117437723,
+      "loss": 2.6578,
+      "step": 3518
+    },
+    {
+      "epoch": 1.564,
+      "grad_norm": 1.8625141382217407,
+      "learning_rate": 0.0001375622775800712,
+      "loss": 2.4186,
+      "step": 3519
+    },
+    {
+      "epoch": 1.5644444444444443,
+      "grad_norm": 2.1919872760772705,
+      "learning_rate": 0.00013754448398576514,
+      "loss": 2.2043,
+      "step": 3520
+    },
+    {
+      "epoch": 1.564888888888889,
+      "grad_norm": 1.709939956665039,
+      "learning_rate": 0.0001375266903914591,
+      "loss": 2.3172,
+      "step": 3521
+    },
+    {
+      "epoch": 1.5653333333333332,
+      "grad_norm": 2.1514196395874023,
+      "learning_rate": 0.00013750889679715303,
+      "loss": 2.4595,
+      "step": 3522
+    },
+    {
+      "epoch": 1.565777777777778,
+      "grad_norm": 1.8870500326156616,
+      "learning_rate": 0.00013749110320284696,
+      "loss": 2.6702,
+      "step": 3523
+    },
+    {
+      "epoch": 1.5662222222222222,
+      "grad_norm": 1.7913631200790405,
+      "learning_rate": 0.00013747330960854092,
+      "loss": 2.3106,
+      "step": 3524
+    },
+    {
+      "epoch": 1.5666666666666667,
+      "grad_norm": 1.8012170791625977,
+      "learning_rate": 0.00013745551601423487,
+      "loss": 1.7666,
+      "step": 3525
+    },
+    {
+      "epoch": 1.5671111111111111,
+      "grad_norm": 2.001098871231079,
+      "learning_rate": 0.00013743772241992883,
+      "loss": 2.1891,
+      "step": 3526
+    },
+    {
+      "epoch": 1.5675555555555556,
+      "grad_norm": 1.766248345375061,
+      "learning_rate": 0.00013741992882562279,
+      "loss": 2.1883,
+      "step": 3527
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 1.6869791746139526,
+      "learning_rate": 0.00013740213523131674,
+      "loss": 1.8203,
+      "step": 3528
+    },
+    {
+      "epoch": 1.5684444444444443,
+      "grad_norm": 1.724373459815979,
+      "learning_rate": 0.00013738434163701067,
+      "loss": 1.8375,
+      "step": 3529
+    },
+    {
+      "epoch": 1.568888888888889,
+      "grad_norm": 1.7977555990219116,
+      "learning_rate": 0.00013736654804270463,
+      "loss": 2.2153,
+      "step": 3530
+    },
+    {
+      "epoch": 1.5693333333333332,
+      "grad_norm": 1.736857295036316,
+      "learning_rate": 0.00013734875444839858,
+      "loss": 1.7471,
+      "step": 3531
+    },
+    {
+      "epoch": 1.569777777777778,
+      "grad_norm": 1.7593656778335571,
+      "learning_rate": 0.00013733096085409254,
+      "loss": 1.7479,
+      "step": 3532
+    },
+    {
+      "epoch": 1.5702222222222222,
+      "grad_norm": 1.9257147312164307,
+      "learning_rate": 0.0001373131672597865,
+      "loss": 1.9489,
+      "step": 3533
+    },
+    {
+      "epoch": 1.5706666666666667,
+      "grad_norm": 2.0456788539886475,
+      "learning_rate": 0.00013729537366548045,
+      "loss": 2.1165,
+      "step": 3534
+    },
+    {
+      "epoch": 1.5711111111111111,
+      "grad_norm": 2.0241599082946777,
+      "learning_rate": 0.00013727758007117438,
+      "loss": 2.3391,
+      "step": 3535
+    },
+    {
+      "epoch": 1.5715555555555556,
+      "grad_norm": 2.1118271350860596,
+      "learning_rate": 0.0001372597864768683,
+      "loss": 2.3182,
+      "step": 3536
+    },
+    {
+      "epoch": 1.572,
+      "grad_norm": 2.060579776763916,
+      "learning_rate": 0.00013724199288256227,
+      "loss": 2.3438,
+      "step": 3537
+    },
+    {
+      "epoch": 1.5724444444444443,
+      "grad_norm": 1.868186354637146,
+      "learning_rate": 0.00013722419928825623,
+      "loss": 1.67,
+      "step": 3538
+    },
+    {
+      "epoch": 1.572888888888889,
+      "grad_norm": 2.3484630584716797,
+      "learning_rate": 0.00013720640569395018,
+      "loss": 2.2858,
+      "step": 3539
+    },
+    {
+      "epoch": 1.5733333333333333,
+      "grad_norm": 2.5575010776519775,
+      "learning_rate": 0.00013718861209964414,
+      "loss": 2.4188,
+      "step": 3540
+    },
+    {
+      "epoch": 1.573777777777778,
+      "grad_norm": 1.9896149635314941,
+      "learning_rate": 0.0001371708185053381,
+      "loss": 1.6996,
+      "step": 3541
+    },
+    {
+      "epoch": 1.5742222222222222,
+      "grad_norm": 2.069122791290283,
+      "learning_rate": 0.00013715302491103202,
+      "loss": 1.705,
+      "step": 3542
+    },
+    {
+      "epoch": 1.5746666666666667,
+      "grad_norm": 2.1964266300201416,
+      "learning_rate": 0.00013713523131672598,
+      "loss": 1.8919,
+      "step": 3543
+    },
+    {
+      "epoch": 1.5751111111111111,
+      "grad_norm": 2.8693044185638428,
+      "learning_rate": 0.00013711743772241994,
+      "loss": 2.5328,
+      "step": 3544
+    },
+    {
+      "epoch": 1.5755555555555556,
+      "grad_norm": 2.714355707168579,
+      "learning_rate": 0.0001370996441281139,
+      "loss": 2.4895,
+      "step": 3545
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 2.324647903442383,
+      "learning_rate": 0.00013708185053380785,
+      "loss": 2.1343,
+      "step": 3546
+    },
+    {
+      "epoch": 1.5764444444444443,
+      "grad_norm": 2.0678861141204834,
+      "learning_rate": 0.0001370640569395018,
+      "loss": 2.1356,
+      "step": 3547
+    },
+    {
+      "epoch": 1.576888888888889,
+      "grad_norm": 1.9118576049804688,
+      "learning_rate": 0.00013704626334519574,
+      "loss": 1.1119,
+      "step": 3548
+    },
+    {
+      "epoch": 1.5773333333333333,
+      "grad_norm": 0.4530702829360962,
+      "learning_rate": 0.00013702846975088967,
+      "loss": 0.078,
+      "step": 3549
+    },
+    {
+      "epoch": 1.5777777777777777,
+      "grad_norm": 2.177344560623169,
+      "learning_rate": 0.00013701067615658362,
+      "loss": 1.7089,
+      "step": 3550
+    },
+    {
+      "epoch": 1.5782222222222222,
+      "grad_norm": 1.2385826110839844,
+      "learning_rate": 0.00013699288256227758,
+      "loss": 2.394,
+      "step": 3551
+    },
+    {
+      "epoch": 1.5786666666666667,
+      "grad_norm": 1.412482500076294,
+      "learning_rate": 0.00013697508896797154,
+      "loss": 2.4647,
+      "step": 3552
+    },
+    {
+      "epoch": 1.5791111111111111,
+      "grad_norm": 1.1037476062774658,
+      "learning_rate": 0.0001369572953736655,
+      "loss": 1.245,
+      "step": 3553
+    },
+    {
+      "epoch": 1.5795555555555556,
+      "grad_norm": 1.6340032815933228,
+      "learning_rate": 0.00013693950177935945,
+      "loss": 1.5744,
+      "step": 3554
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.600129246711731,
+      "learning_rate": 0.00013692170818505338,
+      "loss": 2.1649,
+      "step": 3555
+    },
+    {
+      "epoch": 1.5804444444444443,
+      "grad_norm": 1.5471206903457642,
+      "learning_rate": 0.00013690391459074733,
+      "loss": 1.8665,
+      "step": 3556
+    },
+    {
+      "epoch": 1.580888888888889,
+      "grad_norm": 1.4040687084197998,
+      "learning_rate": 0.0001368861209964413,
+      "loss": 2.1955,
+      "step": 3557
+    },
+    {
+      "epoch": 1.5813333333333333,
+      "grad_norm": 1.490673542022705,
+      "learning_rate": 0.00013686832740213525,
+      "loss": 2.0096,
+      "step": 3558
+    },
+    {
+      "epoch": 1.5817777777777777,
+      "grad_norm": 1.6409083604812622,
+      "learning_rate": 0.0001368505338078292,
+      "loss": 2.2744,
+      "step": 3559
+    },
+    {
+      "epoch": 1.5822222222222222,
+      "grad_norm": 1.574512004852295,
+      "learning_rate": 0.00013683274021352316,
+      "loss": 1.9902,
+      "step": 3560
+    },
+    {
+      "epoch": 1.5826666666666667,
+      "grad_norm": 1.7083110809326172,
+      "learning_rate": 0.0001368149466192171,
+      "loss": 1.6893,
+      "step": 3561
+    },
+    {
+      "epoch": 1.5831111111111111,
+      "grad_norm": 1.8733265399932861,
+      "learning_rate": 0.00013679715302491102,
+      "loss": 2.2119,
+      "step": 3562
+    },
+    {
+      "epoch": 1.5835555555555556,
+      "grad_norm": 1.8871500492095947,
+      "learning_rate": 0.00013677935943060498,
+      "loss": 2.3291,
+      "step": 3563
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 1.5511809587478638,
+      "learning_rate": 0.00013676156583629893,
+      "loss": 2.2039,
+      "step": 3564
+    },
+    {
+      "epoch": 1.5844444444444443,
+      "grad_norm": 1.8675401210784912,
+      "learning_rate": 0.0001367437722419929,
+      "loss": 2.0337,
+      "step": 3565
+    },
+    {
+      "epoch": 1.584888888888889,
+      "grad_norm": 1.8697566986083984,
+      "learning_rate": 0.00013672597864768684,
+      "loss": 1.5566,
+      "step": 3566
+    },
+    {
+      "epoch": 1.5853333333333333,
+      "grad_norm": 1.7386986017227173,
+      "learning_rate": 0.0001367081850533808,
+      "loss": 2.0596,
+      "step": 3567
+    },
+    {
+      "epoch": 1.5857777777777777,
+      "grad_norm": 1.9721729755401611,
+      "learning_rate": 0.00013669039145907473,
+      "loss": 2.6456,
+      "step": 3568
+    },
+    {
+      "epoch": 1.5862222222222222,
+      "grad_norm": 1.8694285154342651,
+      "learning_rate": 0.0001366725978647687,
+      "loss": 2.3915,
+      "step": 3569
+    },
+    {
+      "epoch": 1.5866666666666667,
+      "grad_norm": 2.148411512374878,
+      "learning_rate": 0.00013665480427046264,
+      "loss": 2.6203,
+      "step": 3570
+    },
+    {
+      "epoch": 1.5871111111111111,
+      "grad_norm": 1.9500958919525146,
+      "learning_rate": 0.0001366370106761566,
+      "loss": 2.3018,
+      "step": 3571
+    },
+    {
+      "epoch": 1.5875555555555556,
+      "grad_norm": 1.888933777809143,
+      "learning_rate": 0.00013661921708185056,
+      "loss": 2.2713,
+      "step": 3572
+    },
+    {
+      "epoch": 1.588,
+      "grad_norm": 2.1509106159210205,
+      "learning_rate": 0.0001366014234875445,
+      "loss": 2.2562,
+      "step": 3573
+    },
+    {
+      "epoch": 1.5884444444444443,
+      "grad_norm": 1.9218084812164307,
+      "learning_rate": 0.00013658362989323844,
+      "loss": 1.9106,
+      "step": 3574
+    },
+    {
+      "epoch": 1.588888888888889,
+      "grad_norm": 2.161742687225342,
+      "learning_rate": 0.00013656583629893237,
+      "loss": 2.061,
+      "step": 3575
+    },
+    {
+      "epoch": 1.5893333333333333,
+      "grad_norm": 1.4747031927108765,
+      "learning_rate": 0.00013654804270462633,
+      "loss": 1.0917,
+      "step": 3576
+    },
+    {
+      "epoch": 1.5897777777777777,
+      "grad_norm": 2.0540664196014404,
+      "learning_rate": 0.00013653024911032029,
+      "loss": 2.2196,
+      "step": 3577
+    },
+    {
+      "epoch": 1.5902222222222222,
+      "grad_norm": 2.018321990966797,
+      "learning_rate": 0.00013651245551601424,
+      "loss": 2.4654,
+      "step": 3578
+    },
+    {
+      "epoch": 1.5906666666666667,
+      "grad_norm": 2.201575517654419,
+      "learning_rate": 0.0001364946619217082,
+      "loss": 2.1311,
+      "step": 3579
+    },
+    {
+      "epoch": 1.5911111111111111,
+      "grad_norm": 2.108130931854248,
+      "learning_rate": 0.00013647686832740213,
+      "loss": 2.4114,
+      "step": 3580
+    },
+    {
+      "epoch": 1.5915555555555554,
+      "grad_norm": 1.782263159751892,
+      "learning_rate": 0.00013645907473309608,
+      "loss": 1.9585,
+      "step": 3581
+    },
+    {
+      "epoch": 1.592,
+      "grad_norm": 2.01802134513855,
+      "learning_rate": 0.00013644128113879004,
+      "loss": 2.4353,
+      "step": 3582
+    },
+    {
+      "epoch": 1.5924444444444443,
+      "grad_norm": 2.023480176925659,
+      "learning_rate": 0.000136423487544484,
+      "loss": 1.6961,
+      "step": 3583
+    },
+    {
+      "epoch": 1.592888888888889,
+      "grad_norm": 2.0900518894195557,
+      "learning_rate": 0.00013640569395017795,
+      "loss": 2.1796,
+      "step": 3584
+    },
+    {
+      "epoch": 1.5933333333333333,
+      "grad_norm": 2.238590717315674,
+      "learning_rate": 0.0001363879003558719,
+      "loss": 2.2932,
+      "step": 3585
+    },
+    {
+      "epoch": 1.5937777777777777,
+      "grad_norm": 2.2192628383636475,
+      "learning_rate": 0.00013637010676156584,
+      "loss": 1.9587,
+      "step": 3586
+    },
+    {
+      "epoch": 1.5942222222222222,
+      "grad_norm": 2.1162943840026855,
+      "learning_rate": 0.00013635231316725977,
+      "loss": 2.5576,
+      "step": 3587
+    },
+    {
+      "epoch": 1.5946666666666667,
+      "grad_norm": 2.580493927001953,
+      "learning_rate": 0.00013633451957295373,
+      "loss": 2.4377,
+      "step": 3588
+    },
+    {
+      "epoch": 1.5951111111111111,
+      "grad_norm": 2.192513942718506,
+      "learning_rate": 0.00013631672597864768,
+      "loss": 2.2521,
+      "step": 3589
+    },
+    {
+      "epoch": 1.5955555555555554,
+      "grad_norm": 2.2190680503845215,
+      "learning_rate": 0.00013629893238434164,
+      "loss": 2.0167,
+      "step": 3590
+    },
+    {
+      "epoch": 1.596,
+      "grad_norm": 2.1189939975738525,
+      "learning_rate": 0.0001362811387900356,
+      "loss": 2.1514,
+      "step": 3591
+    },
+    {
+      "epoch": 1.5964444444444443,
+      "grad_norm": 1.9854724407196045,
+      "learning_rate": 0.00013626334519572955,
+      "loss": 2.4261,
+      "step": 3592
+    },
+    {
+      "epoch": 1.596888888888889,
+      "grad_norm": 2.232863426208496,
+      "learning_rate": 0.00013624555160142348,
+      "loss": 2.0627,
+      "step": 3593
+    },
+    {
+      "epoch": 1.5973333333333333,
+      "grad_norm": 2.1362996101379395,
+      "learning_rate": 0.00013622775800711744,
+      "loss": 2.0867,
+      "step": 3594
+    },
+    {
+      "epoch": 1.5977777777777777,
+      "grad_norm": 2.116586923599243,
+      "learning_rate": 0.0001362099644128114,
+      "loss": 2.0974,
+      "step": 3595
+    },
+    {
+      "epoch": 1.5982222222222222,
+      "grad_norm": 2.2818892002105713,
+      "learning_rate": 0.00013619217081850535,
+      "loss": 2.4957,
+      "step": 3596
+    },
+    {
+      "epoch": 1.5986666666666667,
+      "grad_norm": 2.1921818256378174,
+      "learning_rate": 0.0001361743772241993,
+      "loss": 2.0431,
+      "step": 3597
+    },
+    {
+      "epoch": 1.5991111111111111,
+      "grad_norm": 2.915421724319458,
+      "learning_rate": 0.00013615658362989326,
+      "loss": 2.8215,
+      "step": 3598
+    },
+    {
+      "epoch": 1.5995555555555554,
+      "grad_norm": 2.097731590270996,
+      "learning_rate": 0.0001361387900355872,
+      "loss": 1.9497,
+      "step": 3599
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.766552209854126,
+      "learning_rate": 0.00013612099644128112,
+      "loss": 2.1285,
+      "step": 3600
+    },
+    {
+      "epoch": 1.6004444444444443,
+      "grad_norm": 1.9238742589950562,
+      "learning_rate": 0.00013610320284697508,
+      "loss": 2.5526,
+      "step": 3601
+    },
+    {
+      "epoch": 1.600888888888889,
+      "grad_norm": 1.162853717803955,
+      "learning_rate": 0.00013608540925266903,
+      "loss": 1.0674,
+      "step": 3602
+    },
+    {
+      "epoch": 1.6013333333333333,
+      "grad_norm": 1.4226114749908447,
+      "learning_rate": 0.000136067615658363,
+      "loss": 2.2656,
+      "step": 3603
+    },
+    {
+      "epoch": 1.6017777777777777,
+      "grad_norm": 1.5394558906555176,
+      "learning_rate": 0.00013604982206405695,
+      "loss": 2.1054,
+      "step": 3604
+    },
+    {
+      "epoch": 1.6022222222222222,
+      "grad_norm": 1.7394784688949585,
+      "learning_rate": 0.0001360320284697509,
+      "loss": 2.3422,
+      "step": 3605
+    },
+    {
+      "epoch": 1.6026666666666667,
+      "grad_norm": 1.7149112224578857,
+      "learning_rate": 0.00013601423487544483,
+      "loss": 2.4442,
+      "step": 3606
+    },
+    {
+      "epoch": 1.6031111111111112,
+      "grad_norm": 1.8628418445587158,
+      "learning_rate": 0.0001359964412811388,
+      "loss": 2.5289,
+      "step": 3607
+    },
+    {
+      "epoch": 1.6035555555555554,
+      "grad_norm": 1.5874834060668945,
+      "learning_rate": 0.00013597864768683275,
+      "loss": 2.4502,
+      "step": 3608
+    },
+    {
+      "epoch": 1.604,
+      "grad_norm": 1.6472437381744385,
+      "learning_rate": 0.0001359608540925267,
+      "loss": 2.1657,
+      "step": 3609
+    },
+    {
+      "epoch": 1.6044444444444443,
+      "grad_norm": 1.950655221939087,
+      "learning_rate": 0.00013594306049822066,
+      "loss": 1.9845,
+      "step": 3610
+    },
+    {
+      "epoch": 1.604888888888889,
+      "grad_norm": 1.5699740648269653,
+      "learning_rate": 0.00013592526690391462,
+      "loss": 2.5056,
+      "step": 3611
+    },
+    {
+      "epoch": 1.6053333333333333,
+      "grad_norm": 1.7249257564544678,
+      "learning_rate": 0.00013590747330960855,
+      "loss": 2.4613,
+      "step": 3612
+    },
+    {
+      "epoch": 1.6057777777777777,
+      "grad_norm": 1.6163465976715088,
+      "learning_rate": 0.00013588967971530248,
+      "loss": 2.2756,
+      "step": 3613
+    },
+    {
+      "epoch": 1.6062222222222222,
+      "grad_norm": 1.6546217203140259,
+      "learning_rate": 0.00013587188612099643,
+      "loss": 2.0813,
+      "step": 3614
+    },
+    {
+      "epoch": 1.6066666666666667,
+      "grad_norm": 1.7059136629104614,
+      "learning_rate": 0.0001358540925266904,
+      "loss": 2.2557,
+      "step": 3615
+    },
+    {
+      "epoch": 1.6071111111111112,
+      "grad_norm": 1.9633029699325562,
+      "learning_rate": 0.00013583629893238434,
+      "loss": 2.3581,
+      "step": 3616
+    },
+    {
+      "epoch": 1.6075555555555554,
+      "grad_norm": 1.5104560852050781,
+      "learning_rate": 0.0001358185053380783,
+      "loss": 1.8789,
+      "step": 3617
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 1.7537540197372437,
+      "learning_rate": 0.00013580071174377226,
+      "loss": 2.1946,
+      "step": 3618
+    },
+    {
+      "epoch": 1.6084444444444443,
+      "grad_norm": 1.8772600889205933,
+      "learning_rate": 0.0001357829181494662,
+      "loss": 2.3868,
+      "step": 3619
+    },
+    {
+      "epoch": 1.608888888888889,
+      "grad_norm": 1.7722148895263672,
+      "learning_rate": 0.00013576512455516014,
+      "loss": 2.0009,
+      "step": 3620
+    },
+    {
+      "epoch": 1.6093333333333333,
+      "grad_norm": 1.662315011024475,
+      "learning_rate": 0.0001357473309608541,
+      "loss": 1.721,
+      "step": 3621
+    },
+    {
+      "epoch": 1.6097777777777778,
+      "grad_norm": 1.9207539558410645,
+      "learning_rate": 0.00013572953736654806,
+      "loss": 2.4827,
+      "step": 3622
+    },
+    {
+      "epoch": 1.6102222222222222,
+      "grad_norm": 1.8683892488479614,
+      "learning_rate": 0.000135711743772242,
+      "loss": 2.2517,
+      "step": 3623
+    },
+    {
+      "epoch": 1.6106666666666667,
+      "grad_norm": 1.8369477987289429,
+      "learning_rate": 0.00013569395017793597,
+      "loss": 2.0752,
+      "step": 3624
+    },
+    {
+      "epoch": 1.6111111111111112,
+      "grad_norm": 1.9077363014221191,
+      "learning_rate": 0.0001356761565836299,
+      "loss": 1.9962,
+      "step": 3625
+    },
+    {
+      "epoch": 1.6115555555555554,
+      "grad_norm": 1.9185972213745117,
+      "learning_rate": 0.00013565836298932383,
+      "loss": 1.8832,
+      "step": 3626
+    },
+    {
+      "epoch": 1.612,
+      "grad_norm": 1.9970186948776245,
+      "learning_rate": 0.00013564056939501778,
+      "loss": 2.2276,
+      "step": 3627
+    },
+    {
+      "epoch": 1.6124444444444443,
+      "grad_norm": 1.8423793315887451,
+      "learning_rate": 0.00013562277580071174,
+      "loss": 2.3652,
+      "step": 3628
+    },
+    {
+      "epoch": 1.612888888888889,
+      "grad_norm": 2.101058006286621,
+      "learning_rate": 0.0001356049822064057,
+      "loss": 2.3492,
+      "step": 3629
+    },
+    {
+      "epoch": 1.6133333333333333,
+      "grad_norm": 2.483633279800415,
+      "learning_rate": 0.00013558718861209965,
+      "loss": 2.4914,
+      "step": 3630
+    },
+    {
+      "epoch": 1.6137777777777778,
+      "grad_norm": 2.1490073204040527,
+      "learning_rate": 0.0001355693950177936,
+      "loss": 1.7586,
+      "step": 3631
+    },
+    {
+      "epoch": 1.6142222222222222,
+      "grad_norm": 2.103785276412964,
+      "learning_rate": 0.00013555160142348754,
+      "loss": 2.379,
+      "step": 3632
+    },
+    {
+      "epoch": 1.6146666666666667,
+      "grad_norm": 2.0692965984344482,
+      "learning_rate": 0.0001355338078291815,
+      "loss": 2.0689,
+      "step": 3633
+    },
+    {
+      "epoch": 1.6151111111111112,
+      "grad_norm": 2.025420665740967,
+      "learning_rate": 0.00013551601423487545,
+      "loss": 2.1845,
+      "step": 3634
+    },
+    {
+      "epoch": 1.6155555555555554,
+      "grad_norm": 1.7488460540771484,
+      "learning_rate": 0.0001354982206405694,
+      "loss": 1.6538,
+      "step": 3635
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 2.308826446533203,
+      "learning_rate": 0.00013548042704626337,
+      "loss": 2.2989,
+      "step": 3636
+    },
+    {
+      "epoch": 1.6164444444444444,
+      "grad_norm": 2.220733404159546,
+      "learning_rate": 0.00013546263345195732,
+      "loss": 2.6318,
+      "step": 3637
+    },
+    {
+      "epoch": 1.616888888888889,
+      "grad_norm": 2.038041830062866,
+      "learning_rate": 0.00013544483985765125,
+      "loss": 1.8199,
+      "step": 3638
+    },
+    {
+      "epoch": 1.6173333333333333,
+      "grad_norm": 2.3639976978302,
+      "learning_rate": 0.00013542704626334518,
+      "loss": 1.7709,
+      "step": 3639
+    },
+    {
+      "epoch": 1.6177777777777778,
+      "grad_norm": 2.1797988414764404,
+      "learning_rate": 0.00013540925266903914,
+      "loss": 1.9726,
+      "step": 3640
+    },
+    {
+      "epoch": 1.6182222222222222,
+      "grad_norm": 2.5964488983154297,
+      "learning_rate": 0.0001353914590747331,
+      "loss": 2.5678,
+      "step": 3641
+    },
+    {
+      "epoch": 1.6186666666666667,
+      "grad_norm": 2.0529944896698,
+      "learning_rate": 0.00013537366548042705,
+      "loss": 1.9994,
+      "step": 3642
+    },
+    {
+      "epoch": 1.6191111111111112,
+      "grad_norm": 1.9986488819122314,
+      "learning_rate": 0.000135355871886121,
+      "loss": 2.1701,
+      "step": 3643
+    },
+    {
+      "epoch": 1.6195555555555554,
+      "grad_norm": 2.469334840774536,
+      "learning_rate": 0.00013533807829181496,
+      "loss": 2.5633,
+      "step": 3644
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.4147121906280518,
+      "learning_rate": 0.0001353202846975089,
+      "loss": 2.2908,
+      "step": 3645
+    },
+    {
+      "epoch": 1.6204444444444444,
+      "grad_norm": 2.1340951919555664,
+      "learning_rate": 0.00013530249110320285,
+      "loss": 1.8365,
+      "step": 3646
+    },
+    {
+      "epoch": 1.620888888888889,
+      "grad_norm": 2.6548309326171875,
+      "learning_rate": 0.0001352846975088968,
+      "loss": 2.6132,
+      "step": 3647
+    },
+    {
+      "epoch": 1.6213333333333333,
+      "grad_norm": 2.923912286758423,
+      "learning_rate": 0.00013526690391459076,
+      "loss": 2.6588,
+      "step": 3648
+    },
+    {
+      "epoch": 1.6217777777777778,
+      "grad_norm": 2.134868621826172,
+      "learning_rate": 0.00013524911032028472,
+      "loss": 0.9781,
+      "step": 3649
+    },
+    {
+      "epoch": 1.6222222222222222,
+      "grad_norm": 3.1165969371795654,
+      "learning_rate": 0.00013523131672597868,
+      "loss": 1.6002,
+      "step": 3650
+    },
+    {
+      "epoch": 1.6226666666666667,
+      "grad_norm": 1.4649471044540405,
+      "learning_rate": 0.0001352135231316726,
+      "loss": 0.9932,
+      "step": 3651
+    },
+    {
+      "epoch": 1.6231111111111112,
+      "grad_norm": 1.4319626092910767,
+      "learning_rate": 0.00013519572953736653,
+      "loss": 2.754,
+      "step": 3652
+    },
+    {
+      "epoch": 1.6235555555555554,
+      "grad_norm": 1.6543512344360352,
+      "learning_rate": 0.0001351779359430605,
+      "loss": 2.3402,
+      "step": 3653
+    },
+    {
+      "epoch": 1.624,
+      "grad_norm": 1.540387511253357,
+      "learning_rate": 0.00013516014234875445,
+      "loss": 2.5623,
+      "step": 3654
+    },
+    {
+      "epoch": 1.6244444444444444,
+      "grad_norm": 1.4166914224624634,
+      "learning_rate": 0.0001351423487544484,
+      "loss": 1.6858,
+      "step": 3655
+    },
+    {
+      "epoch": 1.624888888888889,
+      "grad_norm": 2.0138063430786133,
+      "learning_rate": 0.00013512455516014236,
+      "loss": 2.6032,
+      "step": 3656
+    },
+    {
+      "epoch": 1.6253333333333333,
+      "grad_norm": 1.8225046396255493,
+      "learning_rate": 0.00013510676156583632,
+      "loss": 2.6059,
+      "step": 3657
+    },
+    {
+      "epoch": 1.6257777777777778,
+      "grad_norm": 1.8486915826797485,
+      "learning_rate": 0.00013508896797153025,
+      "loss": 2.212,
+      "step": 3658
+    },
+    {
+      "epoch": 1.6262222222222222,
+      "grad_norm": 1.6416893005371094,
+      "learning_rate": 0.0001350711743772242,
+      "loss": 2.4992,
+      "step": 3659
+    },
+    {
+      "epoch": 1.6266666666666667,
+      "grad_norm": 1.5142600536346436,
+      "learning_rate": 0.00013505338078291816,
+      "loss": 2.5075,
+      "step": 3660
+    },
+    {
+      "epoch": 1.6271111111111112,
+      "grad_norm": 1.2331926822662354,
+      "learning_rate": 0.00013503558718861212,
+      "loss": 0.9244,
+      "step": 3661
+    },
+    {
+      "epoch": 1.6275555555555554,
+      "grad_norm": 1.6997432708740234,
+      "learning_rate": 0.00013501779359430607,
+      "loss": 2.1417,
+      "step": 3662
+    },
+    {
+      "epoch": 1.6280000000000001,
+      "grad_norm": 1.8140586614608765,
+      "learning_rate": 0.00013500000000000003,
+      "loss": 2.3071,
+      "step": 3663
+    },
+    {
+      "epoch": 1.6284444444444444,
+      "grad_norm": 1.7522242069244385,
+      "learning_rate": 0.00013498220640569396,
+      "loss": 1.7328,
+      "step": 3664
+    },
+    {
+      "epoch": 1.628888888888889,
+      "grad_norm": 1.7086604833602905,
+      "learning_rate": 0.0001349644128113879,
+      "loss": 2.15,
+      "step": 3665
+    },
+    {
+      "epoch": 1.6293333333333333,
+      "grad_norm": 1.699811577796936,
+      "learning_rate": 0.00013494661921708184,
+      "loss": 2.1629,
+      "step": 3666
+    },
+    {
+      "epoch": 1.6297777777777778,
+      "grad_norm": 1.6811929941177368,
+      "learning_rate": 0.0001349288256227758,
+      "loss": 1.7282,
+      "step": 3667
+    },
+    {
+      "epoch": 1.6302222222222222,
+      "grad_norm": 1.8707060813903809,
+      "learning_rate": 0.00013491103202846976,
+      "loss": 2.4604,
+      "step": 3668
+    },
+    {
+      "epoch": 1.6306666666666667,
+      "grad_norm": 1.7901276350021362,
+      "learning_rate": 0.00013489323843416371,
+      "loss": 2.1453,
+      "step": 3669
+    },
+    {
+      "epoch": 1.6311111111111112,
+      "grad_norm": 1.7635587453842163,
+      "learning_rate": 0.00013487544483985764,
+      "loss": 2.688,
+      "step": 3670
+    },
+    {
+      "epoch": 1.6315555555555554,
+      "grad_norm": 2.1348886489868164,
+      "learning_rate": 0.0001348576512455516,
+      "loss": 2.6815,
+      "step": 3671
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 1.738974690437317,
+      "learning_rate": 0.00013483985765124556,
+      "loss": 2.1772,
+      "step": 3672
+    },
+    {
+      "epoch": 1.6324444444444444,
+      "grad_norm": 1.7756551504135132,
+      "learning_rate": 0.0001348220640569395,
+      "loss": 2.3101,
+      "step": 3673
+    },
+    {
+      "epoch": 1.6328888888888888,
+      "grad_norm": 1.9206149578094482,
+      "learning_rate": 0.00013480427046263347,
+      "loss": 2.318,
+      "step": 3674
+    },
+    {
+      "epoch": 1.6333333333333333,
+      "grad_norm": 1.740354299545288,
+      "learning_rate": 0.00013478647686832743,
+      "loss": 1.6665,
+      "step": 3675
+    },
+    {
+      "epoch": 1.6337777777777778,
+      "grad_norm": 1.8040590286254883,
+      "learning_rate": 0.00013476868327402135,
+      "loss": 1.9037,
+      "step": 3676
+    },
+    {
+      "epoch": 1.6342222222222222,
+      "grad_norm": 1.8580551147460938,
+      "learning_rate": 0.0001347508896797153,
+      "loss": 1.9794,
+      "step": 3677
+    },
+    {
+      "epoch": 1.6346666666666667,
+      "grad_norm": 1.9417858123779297,
+      "learning_rate": 0.00013473309608540924,
+      "loss": 2.0177,
+      "step": 3678
+    },
+    {
+      "epoch": 1.6351111111111112,
+      "grad_norm": 1.7505145072937012,
+      "learning_rate": 0.0001347153024911032,
+      "loss": 2.1137,
+      "step": 3679
+    },
+    {
+      "epoch": 1.6355555555555554,
+      "grad_norm": 2.044389486312866,
+      "learning_rate": 0.00013469750889679715,
+      "loss": 2.5816,
+      "step": 3680
+    },
+    {
+      "epoch": 1.6360000000000001,
+      "grad_norm": 1.7209917306900024,
+      "learning_rate": 0.0001346797153024911,
+      "loss": 1.2114,
+      "step": 3681
+    },
+    {
+      "epoch": 1.6364444444444444,
+      "grad_norm": 0.24984845519065857,
+      "learning_rate": 0.00013466192170818507,
+      "loss": 0.037,
+      "step": 3682
+    },
+    {
+      "epoch": 1.6368888888888888,
+      "grad_norm": 2.0459017753601074,
+      "learning_rate": 0.000134644128113879,
+      "loss": 2.2279,
+      "step": 3683
+    },
+    {
+      "epoch": 1.6373333333333333,
+      "grad_norm": 2.004143714904785,
+      "learning_rate": 0.00013462633451957295,
+      "loss": 2.1229,
+      "step": 3684
+    },
+    {
+      "epoch": 1.6377777777777778,
+      "grad_norm": 1.9758665561676025,
+      "learning_rate": 0.0001346085409252669,
+      "loss": 2.0448,
+      "step": 3685
+    },
+    {
+      "epoch": 1.6382222222222222,
+      "grad_norm": 2.130927324295044,
+      "learning_rate": 0.00013459074733096087,
+      "loss": 2.2224,
+      "step": 3686
+    },
+    {
+      "epoch": 1.6386666666666667,
+      "grad_norm": 2.007413148880005,
+      "learning_rate": 0.00013457295373665482,
+      "loss": 2.2341,
+      "step": 3687
+    },
+    {
+      "epoch": 1.6391111111111112,
+      "grad_norm": 2.422636032104492,
+      "learning_rate": 0.00013455516014234878,
+      "loss": 2.4104,
+      "step": 3688
+    },
+    {
+      "epoch": 1.6395555555555554,
+      "grad_norm": 2.4192779064178467,
+      "learning_rate": 0.0001345373665480427,
+      "loss": 2.3782,
+      "step": 3689
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 1.9421988725662231,
+      "learning_rate": 0.00013451957295373666,
+      "loss": 2.1848,
+      "step": 3690
+    },
+    {
+      "epoch": 1.6404444444444444,
+      "grad_norm": 2.0658280849456787,
+      "learning_rate": 0.0001345017793594306,
+      "loss": 1.9037,
+      "step": 3691
+    },
+    {
+      "epoch": 1.6408888888888888,
+      "grad_norm": 2.1715986728668213,
+      "learning_rate": 0.00013448398576512455,
+      "loss": 2.5529,
+      "step": 3692
+    },
+    {
+      "epoch": 1.6413333333333333,
+      "grad_norm": 2.3295042514801025,
+      "learning_rate": 0.0001344661921708185,
+      "loss": 2.2945,
+      "step": 3693
+    },
+    {
+      "epoch": 1.6417777777777778,
+      "grad_norm": 2.211021661758423,
+      "learning_rate": 0.00013444839857651246,
+      "loss": 2.2324,
+      "step": 3694
+    },
+    {
+      "epoch": 1.6422222222222222,
+      "grad_norm": 2.5660417079925537,
+      "learning_rate": 0.00013443060498220642,
+      "loss": 2.6198,
+      "step": 3695
+    },
+    {
+      "epoch": 1.6426666666666667,
+      "grad_norm": 2.0584888458251953,
+      "learning_rate": 0.00013441281138790035,
+      "loss": 1.9456,
+      "step": 3696
+    },
+    {
+      "epoch": 1.6431111111111112,
+      "grad_norm": 2.6596102714538574,
+      "learning_rate": 0.0001343950177935943,
+      "loss": 3.0825,
+      "step": 3697
+    },
+    {
+      "epoch": 1.6435555555555554,
+      "grad_norm": 2.368485689163208,
+      "learning_rate": 0.00013437722419928826,
+      "loss": 2.4511,
+      "step": 3698
+    },
+    {
+      "epoch": 1.6440000000000001,
+      "grad_norm": 2.7313435077667236,
+      "learning_rate": 0.00013435943060498222,
+      "loss": 1.0798,
+      "step": 3699
+    },
+    {
+      "epoch": 1.6444444444444444,
+      "grad_norm": 2.3031601905822754,
+      "learning_rate": 0.00013434163701067618,
+      "loss": 2.4437,
+      "step": 3700
+    },
+    {
+      "epoch": 1.6448888888888888,
+      "grad_norm": 1.3070570230484009,
+      "learning_rate": 0.00013432384341637013,
+      "loss": 2.4191,
+      "step": 3701
+    },
+    {
+      "epoch": 1.6453333333333333,
+      "grad_norm": 1.4076443910598755,
+      "learning_rate": 0.00013430604982206406,
+      "loss": 2.2877,
+      "step": 3702
+    },
+    {
+      "epoch": 1.6457777777777778,
+      "grad_norm": 1.5221251249313354,
+      "learning_rate": 0.000134288256227758,
+      "loss": 2.6039,
+      "step": 3703
+    },
+    {
+      "epoch": 1.6462222222222223,
+      "grad_norm": 1.4921314716339111,
+      "learning_rate": 0.00013427046263345195,
+      "loss": 2.7059,
+      "step": 3704
+    },
+    {
+      "epoch": 1.6466666666666665,
+      "grad_norm": 1.7720656394958496,
+      "learning_rate": 0.0001342526690391459,
+      "loss": 2.9481,
+      "step": 3705
+    },
+    {
+      "epoch": 1.6471111111111112,
+      "grad_norm": 1.6238889694213867,
+      "learning_rate": 0.00013423487544483986,
+      "loss": 1.6842,
+      "step": 3706
+    },
+    {
+      "epoch": 1.6475555555555554,
+      "grad_norm": 1.5820468664169312,
+      "learning_rate": 0.00013421708185053382,
+      "loss": 2.4521,
+      "step": 3707
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 1.584951639175415,
+      "learning_rate": 0.00013419928825622777,
+      "loss": 1.9391,
+      "step": 3708
+    },
+    {
+      "epoch": 1.6484444444444444,
+      "grad_norm": 1.8263479471206665,
+      "learning_rate": 0.0001341814946619217,
+      "loss": 2.4315,
+      "step": 3709
+    },
+    {
+      "epoch": 1.6488888888888888,
+      "grad_norm": 1.4678611755371094,
+      "learning_rate": 0.00013416370106761566,
+      "loss": 1.9397,
+      "step": 3710
+    },
+    {
+      "epoch": 1.6493333333333333,
+      "grad_norm": 1.4494836330413818,
+      "learning_rate": 0.00013414590747330962,
+      "loss": 1.6722,
+      "step": 3711
+    },
+    {
+      "epoch": 1.6497777777777778,
+      "grad_norm": 1.7007122039794922,
+      "learning_rate": 0.00013412811387900357,
+      "loss": 2.412,
+      "step": 3712
+    },
+    {
+      "epoch": 1.6502222222222223,
+      "grad_norm": 1.7339609861373901,
+      "learning_rate": 0.00013411032028469753,
+      "loss": 2.1592,
+      "step": 3713
+    },
+    {
+      "epoch": 1.6506666666666665,
+      "grad_norm": 1.8747652769088745,
+      "learning_rate": 0.00013409252669039148,
+      "loss": 2.8591,
+      "step": 3714
+    },
+    {
+      "epoch": 1.6511111111111112,
+      "grad_norm": 1.7554471492767334,
+      "learning_rate": 0.00013407473309608541,
+      "loss": 2.3202,
+      "step": 3715
+    },
+    {
+      "epoch": 1.6515555555555554,
+      "grad_norm": 1.5907261371612549,
+      "learning_rate": 0.00013405693950177934,
+      "loss": 1.7988,
+      "step": 3716
+    },
+    {
+      "epoch": 1.6520000000000001,
+      "grad_norm": 1.1176869869232178,
+      "learning_rate": 0.0001340391459074733,
+      "loss": 0.8156,
+      "step": 3717
+    },
+    {
+      "epoch": 1.6524444444444444,
+      "grad_norm": 1.6439223289489746,
+      "learning_rate": 0.00013402135231316726,
+      "loss": 2.4484,
+      "step": 3718
+    },
+    {
+      "epoch": 1.6528888888888889,
+      "grad_norm": 1.4052650928497314,
+      "learning_rate": 0.0001340035587188612,
+      "loss": 1.775,
+      "step": 3719
+    },
+    {
+      "epoch": 1.6533333333333333,
+      "grad_norm": 1.7691340446472168,
+      "learning_rate": 0.00013398576512455517,
+      "loss": 2.0681,
+      "step": 3720
+    },
+    {
+      "epoch": 1.6537777777777778,
+      "grad_norm": 1.88296377658844,
+      "learning_rate": 0.00013396797153024913,
+      "loss": 2.1395,
+      "step": 3721
+    },
+    {
+      "epoch": 1.6542222222222223,
+      "grad_norm": 1.7832571268081665,
+      "learning_rate": 0.00013395017793594306,
+      "loss": 1.7583,
+      "step": 3722
+    },
+    {
+      "epoch": 1.6546666666666665,
+      "grad_norm": 1.8587702512741089,
+      "learning_rate": 0.000133932384341637,
+      "loss": 1.9886,
+      "step": 3723
+    },
+    {
+      "epoch": 1.6551111111111112,
+      "grad_norm": 1.7117775678634644,
+      "learning_rate": 0.00013391459074733097,
+      "loss": 2.3072,
+      "step": 3724
+    },
+    {
+      "epoch": 1.6555555555555554,
+      "grad_norm": 1.7575953006744385,
+      "learning_rate": 0.00013389679715302493,
+      "loss": 2.2971,
+      "step": 3725
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 1.789642333984375,
+      "learning_rate": 0.00013387900355871888,
+      "loss": 1.8977,
+      "step": 3726
+    },
+    {
+      "epoch": 1.6564444444444444,
+      "grad_norm": 1.8398528099060059,
+      "learning_rate": 0.00013386120996441284,
+      "loss": 2.1343,
+      "step": 3727
+    },
+    {
+      "epoch": 1.6568888888888889,
+      "grad_norm": 1.2515662908554077,
+      "learning_rate": 0.00013384341637010677,
+      "loss": 0.9636,
+      "step": 3728
+    },
+    {
+      "epoch": 1.6573333333333333,
+      "grad_norm": 2.197495937347412,
+      "learning_rate": 0.0001338256227758007,
+      "loss": 2.1068,
+      "step": 3729
+    },
+    {
+      "epoch": 1.6577777777777778,
+      "grad_norm": 1.9065452814102173,
+      "learning_rate": 0.00013380782918149465,
+      "loss": 2.1467,
+      "step": 3730
+    },
+    {
+      "epoch": 1.6582222222222223,
+      "grad_norm": 2.0440008640289307,
+      "learning_rate": 0.0001337900355871886,
+      "loss": 2.2524,
+      "step": 3731
+    },
+    {
+      "epoch": 1.6586666666666665,
+      "grad_norm": 2.092015266418457,
+      "learning_rate": 0.00013377224199288257,
+      "loss": 1.9448,
+      "step": 3732
+    },
+    {
+      "epoch": 1.6591111111111112,
+      "grad_norm": 2.170003652572632,
+      "learning_rate": 0.00013375444839857652,
+      "loss": 2.7868,
+      "step": 3733
+    },
+    {
+      "epoch": 1.6595555555555555,
+      "grad_norm": 2.460923910140991,
+      "learning_rate": 0.00013373665480427048,
+      "loss": 2.7106,
+      "step": 3734
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 2.117558717727661,
+      "learning_rate": 0.0001337188612099644,
+      "loss": 1.8442,
+      "step": 3735
+    },
+    {
+      "epoch": 1.6604444444444444,
+      "grad_norm": 1.9140143394470215,
+      "learning_rate": 0.00013370106761565837,
+      "loss": 1.7407,
+      "step": 3736
+    },
+    {
+      "epoch": 1.6608888888888889,
+      "grad_norm": 2.178762912750244,
+      "learning_rate": 0.00013368327402135232,
+      "loss": 2.0877,
+      "step": 3737
+    },
+    {
+      "epoch": 1.6613333333333333,
+      "grad_norm": 1.4736453294754028,
+      "learning_rate": 0.00013366548042704628,
+      "loss": 1.2792,
+      "step": 3738
+    },
+    {
+      "epoch": 1.6617777777777778,
+      "grad_norm": 2.096111536026001,
+      "learning_rate": 0.00013364768683274023,
+      "loss": 2.0112,
+      "step": 3739
+    },
+    {
+      "epoch": 1.6622222222222223,
+      "grad_norm": 2.047461986541748,
+      "learning_rate": 0.0001336298932384342,
+      "loss": 2.578,
+      "step": 3740
+    },
+    {
+      "epoch": 1.6626666666666665,
+      "grad_norm": 2.3218650817871094,
+      "learning_rate": 0.00013361209964412812,
+      "loss": 2.2947,
+      "step": 3741
+    },
+    {
+      "epoch": 1.6631111111111112,
+      "grad_norm": 2.492847204208374,
+      "learning_rate": 0.00013359430604982205,
+      "loss": 2.4653,
+      "step": 3742
+    },
+    {
+      "epoch": 1.6635555555555555,
+      "grad_norm": 2.1130130290985107,
+      "learning_rate": 0.000133576512455516,
+      "loss": 1.9604,
+      "step": 3743
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 2.3125598430633545,
+      "learning_rate": 0.00013355871886120996,
+      "loss": 2.0361,
+      "step": 3744
+    },
+    {
+      "epoch": 1.6644444444444444,
+      "grad_norm": 2.435710906982422,
+      "learning_rate": 0.00013354092526690392,
+      "loss": 2.6202,
+      "step": 3745
+    },
+    {
+      "epoch": 1.6648888888888889,
+      "grad_norm": 2.7004244327545166,
+      "learning_rate": 0.00013352313167259788,
+      "loss": 2.2756,
+      "step": 3746
+    },
+    {
+      "epoch": 1.6653333333333333,
+      "grad_norm": 1.9285310506820679,
+      "learning_rate": 0.00013350533807829183,
+      "loss": 1.8988,
+      "step": 3747
+    },
+    {
+      "epoch": 1.6657777777777778,
+      "grad_norm": 2.2662885189056396,
+      "learning_rate": 0.00013348754448398576,
+      "loss": 2.2077,
+      "step": 3748
+    },
+    {
+      "epoch": 1.6662222222222223,
+      "grad_norm": 2.3850321769714355,
+      "learning_rate": 0.00013346975088967972,
+      "loss": 2.4601,
+      "step": 3749
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 2.6833248138427734,
+      "learning_rate": 0.00013345195729537367,
+      "loss": 1.1392,
+      "step": 3750
+    },
+    {
+      "epoch": 1.6671111111111112,
+      "grad_norm": 1.3608639240264893,
+      "learning_rate": 0.00013343416370106763,
+      "loss": 2.1485,
+      "step": 3751
+    },
+    {
+      "epoch": 1.6675555555555555,
+      "grad_norm": 1.7300606966018677,
+      "learning_rate": 0.0001334163701067616,
+      "loss": 2.5823,
+      "step": 3752
+    },
+    {
+      "epoch": 1.6680000000000001,
+      "grad_norm": 1.7243146896362305,
+      "learning_rate": 0.00013339857651245554,
+      "loss": 2.4296,
+      "step": 3753
+    },
+    {
+      "epoch": 1.6684444444444444,
+      "grad_norm": 1.6703439950942993,
+      "learning_rate": 0.00013338078291814947,
+      "loss": 2.1775,
+      "step": 3754
+    },
+    {
+      "epoch": 1.6688888888888889,
+      "grad_norm": 1.81732976436615,
+      "learning_rate": 0.0001333629893238434,
+      "loss": 2.548,
+      "step": 3755
+    },
+    {
+      "epoch": 1.6693333333333333,
+      "grad_norm": 1.9755940437316895,
+      "learning_rate": 0.00013334519572953736,
+      "loss": 2.2344,
+      "step": 3756
+    },
+    {
+      "epoch": 1.6697777777777778,
+      "grad_norm": 1.8577018976211548,
+      "learning_rate": 0.00013332740213523132,
+      "loss": 2.247,
+      "step": 3757
+    },
+    {
+      "epoch": 1.6702222222222223,
+      "grad_norm": 1.7116265296936035,
+      "learning_rate": 0.00013330960854092527,
+      "loss": 2.1821,
+      "step": 3758
+    },
+    {
+      "epoch": 1.6706666666666665,
+      "grad_norm": 1.8977397680282593,
+      "learning_rate": 0.00013329181494661923,
+      "loss": 2.1784,
+      "step": 3759
+    },
+    {
+      "epoch": 1.6711111111111112,
+      "grad_norm": 2.1667838096618652,
+      "learning_rate": 0.00013327402135231316,
+      "loss": 2.4484,
+      "step": 3760
+    },
+    {
+      "epoch": 1.6715555555555555,
+      "grad_norm": 1.9198535680770874,
+      "learning_rate": 0.00013325622775800712,
+      "loss": 2.5515,
+      "step": 3761
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 1.7330615520477295,
+      "learning_rate": 0.00013323843416370107,
+      "loss": 2.4671,
+      "step": 3762
+    },
+    {
+      "epoch": 1.6724444444444444,
+      "grad_norm": 1.9315346479415894,
+      "learning_rate": 0.00013322064056939503,
+      "loss": 1.9458,
+      "step": 3763
+    },
+    {
+      "epoch": 1.6728888888888889,
+      "grad_norm": 2.0418102741241455,
+      "learning_rate": 0.00013320284697508898,
+      "loss": 2.3197,
+      "step": 3764
+    },
+    {
+      "epoch": 1.6733333333333333,
+      "grad_norm": 1.8241764307022095,
+      "learning_rate": 0.00013318505338078294,
+      "loss": 2.5491,
+      "step": 3765
+    },
+    {
+      "epoch": 1.6737777777777778,
+      "grad_norm": 1.6973588466644287,
+      "learning_rate": 0.00013316725978647687,
+      "loss": 2.1914,
+      "step": 3766
+    },
+    {
+      "epoch": 1.6742222222222223,
+      "grad_norm": 2.0757534503936768,
+      "learning_rate": 0.00013314946619217083,
+      "loss": 2.9653,
+      "step": 3767
+    },
+    {
+      "epoch": 1.6746666666666665,
+      "grad_norm": 1.850299596786499,
+      "learning_rate": 0.00013313167259786476,
+      "loss": 2.241,
+      "step": 3768
+    },
+    {
+      "epoch": 1.6751111111111112,
+      "grad_norm": 1.8038710355758667,
+      "learning_rate": 0.0001331138790035587,
+      "loss": 2.2839,
+      "step": 3769
+    },
+    {
+      "epoch": 1.6755555555555555,
+      "grad_norm": 1.7989375591278076,
+      "learning_rate": 0.00013309608540925267,
+      "loss": 2.1376,
+      "step": 3770
+    },
+    {
+      "epoch": 1.6760000000000002,
+      "grad_norm": 1.8544578552246094,
+      "learning_rate": 0.00013307829181494663,
+      "loss": 2.1846,
+      "step": 3771
+    },
+    {
+      "epoch": 1.6764444444444444,
+      "grad_norm": 1.7627222537994385,
+      "learning_rate": 0.00013306049822064058,
+      "loss": 1.9189,
+      "step": 3772
+    },
+    {
+      "epoch": 1.6768888888888889,
+      "grad_norm": 1.8061233758926392,
+      "learning_rate": 0.0001330427046263345,
+      "loss": 1.9897,
+      "step": 3773
+    },
+    {
+      "epoch": 1.6773333333333333,
+      "grad_norm": 1.7060072422027588,
+      "learning_rate": 0.00013302491103202847,
+      "loss": 2.0127,
+      "step": 3774
+    },
+    {
+      "epoch": 1.6777777777777778,
+      "grad_norm": 2.054374933242798,
+      "learning_rate": 0.00013300711743772242,
+      "loss": 2.2547,
+      "step": 3775
+    },
+    {
+      "epoch": 1.6782222222222223,
+      "grad_norm": 1.944670557975769,
+      "learning_rate": 0.00013298932384341638,
+      "loss": 2.1802,
+      "step": 3776
+    },
+    {
+      "epoch": 1.6786666666666665,
+      "grad_norm": 2.326748847961426,
+      "learning_rate": 0.00013297153024911034,
+      "loss": 2.3484,
+      "step": 3777
+    },
+    {
+      "epoch": 1.6791111111111112,
+      "grad_norm": 1.707459807395935,
+      "learning_rate": 0.0001329537366548043,
+      "loss": 1.8628,
+      "step": 3778
+    },
+    {
+      "epoch": 1.6795555555555555,
+      "grad_norm": 2.1038479804992676,
+      "learning_rate": 0.00013293594306049822,
+      "loss": 1.6272,
+      "step": 3779
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 2.0666604042053223,
+      "learning_rate": 0.00013291814946619218,
+      "loss": 2.3157,
+      "step": 3780
+    },
+    {
+      "epoch": 1.6804444444444444,
+      "grad_norm": 1.8454957008361816,
+      "learning_rate": 0.0001329003558718861,
+      "loss": 2.0967,
+      "step": 3781
+    },
+    {
+      "epoch": 1.6808888888888889,
+      "grad_norm": 1.9296761751174927,
+      "learning_rate": 0.00013288256227758007,
+      "loss": 2.1812,
+      "step": 3782
+    },
+    {
+      "epoch": 1.6813333333333333,
+      "grad_norm": 1.9812657833099365,
+      "learning_rate": 0.00013286476868327402,
+      "loss": 2.0514,
+      "step": 3783
+    },
+    {
+      "epoch": 1.6817777777777778,
+      "grad_norm": 1.7936532497406006,
+      "learning_rate": 0.00013284697508896798,
+      "loss": 1.9351,
+      "step": 3784
+    },
+    {
+      "epoch": 1.6822222222222223,
+      "grad_norm": 2.2539501190185547,
+      "learning_rate": 0.00013282918149466194,
+      "loss": 2.297,
+      "step": 3785
+    },
+    {
+      "epoch": 1.6826666666666665,
+      "grad_norm": 2.0437049865722656,
+      "learning_rate": 0.00013281138790035586,
+      "loss": 1.9794,
+      "step": 3786
+    },
+    {
+      "epoch": 1.6831111111111112,
+      "grad_norm": 2.1592202186584473,
+      "learning_rate": 0.00013279359430604982,
+      "loss": 2.2224,
+      "step": 3787
+    },
+    {
+      "epoch": 1.6835555555555555,
+      "grad_norm": 2.0359416007995605,
+      "learning_rate": 0.00013277580071174378,
+      "loss": 2.0624,
+      "step": 3788
+    },
+    {
+      "epoch": 1.6840000000000002,
+      "grad_norm": 2.133168935775757,
+      "learning_rate": 0.00013275800711743773,
+      "loss": 2.0215,
+      "step": 3789
+    },
+    {
+      "epoch": 1.6844444444444444,
+      "grad_norm": 2.4152679443359375,
+      "learning_rate": 0.0001327402135231317,
+      "loss": 2.2511,
+      "step": 3790
+    },
+    {
+      "epoch": 1.6848888888888889,
+      "grad_norm": 2.4660964012145996,
+      "learning_rate": 0.00013272241992882565,
+      "loss": 2.6581,
+      "step": 3791
+    },
+    {
+      "epoch": 1.6853333333333333,
+      "grad_norm": 2.309948205947876,
+      "learning_rate": 0.00013270462633451958,
+      "loss": 2.3009,
+      "step": 3792
+    },
+    {
+      "epoch": 1.6857777777777778,
+      "grad_norm": 2.3204185962677,
+      "learning_rate": 0.00013268683274021353,
+      "loss": 2.0474,
+      "step": 3793
+    },
+    {
+      "epoch": 1.6862222222222223,
+      "grad_norm": 2.021592855453491,
+      "learning_rate": 0.00013266903914590746,
+      "loss": 1.6188,
+      "step": 3794
+    },
+    {
+      "epoch": 1.6866666666666665,
+      "grad_norm": 2.1225037574768066,
+      "learning_rate": 0.00013265124555160142,
+      "loss": 1.7655,
+      "step": 3795
+    },
+    {
+      "epoch": 1.6871111111111112,
+      "grad_norm": 2.4528112411499023,
+      "learning_rate": 0.00013263345195729538,
+      "loss": 2.146,
+      "step": 3796
+    },
+    {
+      "epoch": 1.6875555555555555,
+      "grad_norm": 2.4885880947113037,
+      "learning_rate": 0.00013261565836298933,
+      "loss": 2.6485,
+      "step": 3797
+    },
+    {
+      "epoch": 1.688,
+      "grad_norm": 2.7207860946655273,
+      "learning_rate": 0.0001325978647686833,
+      "loss": 2.7009,
+      "step": 3798
+    },
+    {
+      "epoch": 1.6884444444444444,
+      "grad_norm": 3.098876476287842,
+      "learning_rate": 0.00013258007117437722,
+      "loss": 1.9231,
+      "step": 3799
+    },
+    {
+      "epoch": 1.6888888888888889,
+      "grad_norm": 3.4644060134887695,
+      "learning_rate": 0.00013256227758007117,
+      "loss": 2.3836,
+      "step": 3800
+    },
+    {
+      "epoch": 1.6893333333333334,
+      "grad_norm": 1.1060537099838257,
+      "learning_rate": 0.00013254448398576513,
+      "loss": 1.2927,
+      "step": 3801
+    },
+    {
+      "epoch": 1.6897777777777778,
+      "grad_norm": 1.0716097354888916,
+      "learning_rate": 0.0001325266903914591,
+      "loss": 1.1274,
+      "step": 3802
+    },
+    {
+      "epoch": 1.6902222222222223,
+      "grad_norm": 1.659468412399292,
+      "learning_rate": 0.00013250889679715304,
+      "loss": 2.5655,
+      "step": 3803
+    },
+    {
+      "epoch": 1.6906666666666665,
+      "grad_norm": 1.416318416595459,
+      "learning_rate": 0.000132491103202847,
+      "loss": 2.0306,
+      "step": 3804
+    },
+    {
+      "epoch": 1.6911111111111112,
+      "grad_norm": 2.2710070610046387,
+      "learning_rate": 0.00013247330960854093,
+      "loss": 2.178,
+      "step": 3805
+    },
+    {
+      "epoch": 1.6915555555555555,
+      "grad_norm": 1.5939732789993286,
+      "learning_rate": 0.00013245551601423489,
+      "loss": 2.3397,
+      "step": 3806
+    },
+    {
+      "epoch": 1.692,
+      "grad_norm": 1.590168833732605,
+      "learning_rate": 0.00013243772241992882,
+      "loss": 2.2752,
+      "step": 3807
+    },
+    {
+      "epoch": 1.6924444444444444,
+      "grad_norm": 1.1703904867172241,
+      "learning_rate": 0.00013241992882562277,
+      "loss": 1.0365,
+      "step": 3808
+    },
+    {
+      "epoch": 1.6928888888888889,
+      "grad_norm": 1.6481446027755737,
+      "learning_rate": 0.00013240213523131673,
+      "loss": 2.2892,
+      "step": 3809
+    },
+    {
+      "epoch": 1.6933333333333334,
+      "grad_norm": 1.8637354373931885,
+      "learning_rate": 0.00013238434163701069,
+      "loss": 2.4122,
+      "step": 3810
+    },
+    {
+      "epoch": 1.6937777777777778,
+      "grad_norm": 1.6943954229354858,
+      "learning_rate": 0.00013236654804270464,
+      "loss": 2.4287,
+      "step": 3811
+    },
+    {
+      "epoch": 1.6942222222222223,
+      "grad_norm": 1.9139350652694702,
+      "learning_rate": 0.00013234875444839857,
+      "loss": 2.038,
+      "step": 3812
+    },
+    {
+      "epoch": 1.6946666666666665,
+      "grad_norm": 1.6153117418289185,
+      "learning_rate": 0.00013233096085409253,
+      "loss": 1.8512,
+      "step": 3813
+    },
+    {
+      "epoch": 1.6951111111111112,
+      "grad_norm": 1.473202109336853,
+      "learning_rate": 0.00013231316725978648,
+      "loss": 2.0833,
+      "step": 3814
+    },
+    {
+      "epoch": 1.6955555555555555,
+      "grad_norm": 1.6183151006698608,
+      "learning_rate": 0.00013229537366548044,
+      "loss": 1.797,
+      "step": 3815
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 1.648177981376648,
+      "learning_rate": 0.0001322775800711744,
+      "loss": 1.9791,
+      "step": 3816
+    },
+    {
+      "epoch": 1.6964444444444444,
+      "grad_norm": 1.758804202079773,
+      "learning_rate": 0.00013225978647686835,
+      "loss": 2.5341,
+      "step": 3817
+    },
+    {
+      "epoch": 1.696888888888889,
+      "grad_norm": 1.5626602172851562,
+      "learning_rate": 0.00013224199288256228,
+      "loss": 1.5439,
+      "step": 3818
+    },
+    {
+      "epoch": 1.6973333333333334,
+      "grad_norm": 1.9132189750671387,
+      "learning_rate": 0.0001322241992882562,
+      "loss": 2.4955,
+      "step": 3819
+    },
+    {
+      "epoch": 1.6977777777777778,
+      "grad_norm": 1.975059986114502,
+      "learning_rate": 0.00013220640569395017,
+      "loss": 2.4941,
+      "step": 3820
+    },
+    {
+      "epoch": 1.6982222222222223,
+      "grad_norm": 1.8108739852905273,
+      "learning_rate": 0.00013218861209964413,
+      "loss": 1.7566,
+      "step": 3821
+    },
+    {
+      "epoch": 1.6986666666666665,
+      "grad_norm": 1.5749622583389282,
+      "learning_rate": 0.00013217081850533808,
+      "loss": 2.0191,
+      "step": 3822
+    },
+    {
+      "epoch": 1.6991111111111112,
+      "grad_norm": 2.2732627391815186,
+      "learning_rate": 0.00013215302491103204,
+      "loss": 1.8025,
+      "step": 3823
+    },
+    {
+      "epoch": 1.6995555555555555,
+      "grad_norm": 1.6251347064971924,
+      "learning_rate": 0.000132135231316726,
+      "loss": 1.8303,
+      "step": 3824
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.9713494777679443,
+      "learning_rate": 0.00013211743772241992,
+      "loss": 2.1631,
+      "step": 3825
+    },
+    {
+      "epoch": 1.7004444444444444,
+      "grad_norm": 1.9133716821670532,
+      "learning_rate": 0.00013209964412811388,
+      "loss": 2.3184,
+      "step": 3826
+    },
+    {
+      "epoch": 1.700888888888889,
+      "grad_norm": 2.2946348190307617,
+      "learning_rate": 0.00013208185053380784,
+      "loss": 2.5479,
+      "step": 3827
+    },
+    {
+      "epoch": 1.7013333333333334,
+      "grad_norm": 1.764487862586975,
+      "learning_rate": 0.0001320640569395018,
+      "loss": 1.7169,
+      "step": 3828
+    },
+    {
+      "epoch": 1.7017777777777776,
+      "grad_norm": 1.9306504726409912,
+      "learning_rate": 0.00013204626334519575,
+      "loss": 1.8107,
+      "step": 3829
+    },
+    {
+      "epoch": 1.7022222222222223,
+      "grad_norm": 1.9167112112045288,
+      "learning_rate": 0.0001320284697508897,
+      "loss": 1.7252,
+      "step": 3830
+    },
+    {
+      "epoch": 1.7026666666666666,
+      "grad_norm": 2.143211603164673,
+      "learning_rate": 0.00013201067615658364,
+      "loss": 2.3653,
+      "step": 3831
+    },
+    {
+      "epoch": 1.7031111111111112,
+      "grad_norm": 2.1837661266326904,
+      "learning_rate": 0.00013199288256227757,
+      "loss": 2.3616,
+      "step": 3832
+    },
+    {
+      "epoch": 1.7035555555555555,
+      "grad_norm": 2.0470423698425293,
+      "learning_rate": 0.00013197508896797152,
+      "loss": 1.9548,
+      "step": 3833
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 2.4029338359832764,
+      "learning_rate": 0.00013195729537366548,
+      "loss": 2.0182,
+      "step": 3834
+    },
+    {
+      "epoch": 1.7044444444444444,
+      "grad_norm": 2.1079888343811035,
+      "learning_rate": 0.00013193950177935944,
+      "loss": 2.0754,
+      "step": 3835
+    },
+    {
+      "epoch": 1.704888888888889,
+      "grad_norm": 2.2155072689056396,
+      "learning_rate": 0.0001319217081850534,
+      "loss": 2.2782,
+      "step": 3836
+    },
+    {
+      "epoch": 1.7053333333333334,
+      "grad_norm": 2.1504364013671875,
+      "learning_rate": 0.00013190391459074735,
+      "loss": 2.1015,
+      "step": 3837
+    },
+    {
+      "epoch": 1.7057777777777776,
+      "grad_norm": 2.214022159576416,
+      "learning_rate": 0.00013188612099644128,
+      "loss": 2.2518,
+      "step": 3838
+    },
+    {
+      "epoch": 1.7062222222222223,
+      "grad_norm": 2.3961539268493652,
+      "learning_rate": 0.00013186832740213523,
+      "loss": 1.7802,
+      "step": 3839
+    },
+    {
+      "epoch": 1.7066666666666666,
+      "grad_norm": 2.034930467605591,
+      "learning_rate": 0.0001318505338078292,
+      "loss": 2.2164,
+      "step": 3840
+    },
+    {
+      "epoch": 1.7071111111111112,
+      "grad_norm": 2.0269289016723633,
+      "learning_rate": 0.00013183274021352315,
+      "loss": 1.7755,
+      "step": 3841
+    },
+    {
+      "epoch": 1.7075555555555555,
+      "grad_norm": 2.0814435482025146,
+      "learning_rate": 0.0001318149466192171,
+      "loss": 1.9628,
+      "step": 3842
+    },
+    {
+      "epoch": 1.708,
+      "grad_norm": 2.4242076873779297,
+      "learning_rate": 0.00013179715302491106,
+      "loss": 2.4269,
+      "step": 3843
+    },
+    {
+      "epoch": 1.7084444444444444,
+      "grad_norm": 2.6244444847106934,
+      "learning_rate": 0.000131779359430605,
+      "loss": 2.279,
+      "step": 3844
+    },
+    {
+      "epoch": 1.708888888888889,
+      "grad_norm": 2.184532403945923,
+      "learning_rate": 0.00013176156583629892,
+      "loss": 2.0163,
+      "step": 3845
+    },
+    {
+      "epoch": 1.7093333333333334,
+      "grad_norm": 2.2187845706939697,
+      "learning_rate": 0.00013174377224199288,
+      "loss": 2.226,
+      "step": 3846
+    },
+    {
+      "epoch": 1.7097777777777776,
+      "grad_norm": 2.7310564517974854,
+      "learning_rate": 0.00013172597864768683,
+      "loss": 2.1964,
+      "step": 3847
+    },
+    {
+      "epoch": 1.7102222222222223,
+      "grad_norm": 2.8237147331237793,
+      "learning_rate": 0.0001317081850533808,
+      "loss": 2.5273,
+      "step": 3848
+    },
+    {
+      "epoch": 1.7106666666666666,
+      "grad_norm": 1.9283852577209473,
+      "learning_rate": 0.00013169039145907474,
+      "loss": 1.3825,
+      "step": 3849
+    },
+    {
+      "epoch": 1.7111111111111112,
+      "grad_norm": 2.044787883758545,
+      "learning_rate": 0.00013167259786476867,
+      "loss": 0.2311,
+      "step": 3850
+    },
+    {
+      "epoch": 1.7115555555555555,
+      "grad_norm": 1.3595030307769775,
+      "learning_rate": 0.00013165480427046263,
+      "loss": 1.8332,
+      "step": 3851
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 1.7082335948944092,
+      "learning_rate": 0.0001316370106761566,
+      "loss": 1.3219,
+      "step": 3852
+    },
+    {
+      "epoch": 1.7124444444444444,
+      "grad_norm": 1.4628937244415283,
+      "learning_rate": 0.00013161921708185054,
+      "loss": 2.1888,
+      "step": 3853
+    },
+    {
+      "epoch": 1.712888888888889,
+      "grad_norm": 1.4949939250946045,
+      "learning_rate": 0.0001316014234875445,
+      "loss": 2.2021,
+      "step": 3854
+    },
+    {
+      "epoch": 1.7133333333333334,
+      "grad_norm": 1.4384099245071411,
+      "learning_rate": 0.00013158362989323846,
+      "loss": 1.9055,
+      "step": 3855
+    },
+    {
+      "epoch": 1.7137777777777776,
+      "grad_norm": 1.6663846969604492,
+      "learning_rate": 0.00013156583629893239,
+      "loss": 2.4469,
+      "step": 3856
+    },
+    {
+      "epoch": 1.7142222222222223,
+      "grad_norm": 1.5963435173034668,
+      "learning_rate": 0.00013154804270462634,
+      "loss": 2.1661,
+      "step": 3857
+    },
+    {
+      "epoch": 1.7146666666666666,
+      "grad_norm": 1.9488998651504517,
+      "learning_rate": 0.00013153024911032027,
+      "loss": 2.2052,
+      "step": 3858
+    },
+    {
+      "epoch": 1.7151111111111113,
+      "grad_norm": 1.6803126335144043,
+      "learning_rate": 0.00013151245551601423,
+      "loss": 2.437,
+      "step": 3859
+    },
+    {
+      "epoch": 1.7155555555555555,
+      "grad_norm": 1.6857129335403442,
+      "learning_rate": 0.00013149466192170818,
+      "loss": 1.8197,
+      "step": 3860
+    },
+    {
+      "epoch": 1.716,
+      "grad_norm": 1.7986105680465698,
+      "learning_rate": 0.00013147686832740214,
+      "loss": 2.2436,
+      "step": 3861
+    },
+    {
+      "epoch": 1.7164444444444444,
+      "grad_norm": 1.6643034219741821,
+      "learning_rate": 0.0001314590747330961,
+      "loss": 1.9625,
+      "step": 3862
+    },
+    {
+      "epoch": 1.716888888888889,
+      "grad_norm": 1.7694244384765625,
+      "learning_rate": 0.00013144128113879003,
+      "loss": 2.2189,
+      "step": 3863
+    },
+    {
+      "epoch": 1.7173333333333334,
+      "grad_norm": 1.951474905014038,
+      "learning_rate": 0.00013142348754448398,
+      "loss": 2.4388,
+      "step": 3864
+    },
+    {
+      "epoch": 1.7177777777777776,
+      "grad_norm": 1.966758370399475,
+      "learning_rate": 0.00013140569395017794,
+      "loss": 2.5255,
+      "step": 3865
+    },
+    {
+      "epoch": 1.7182222222222223,
+      "grad_norm": 1.5251381397247314,
+      "learning_rate": 0.0001313879003558719,
+      "loss": 1.4708,
+      "step": 3866
+    },
+    {
+      "epoch": 1.7186666666666666,
+      "grad_norm": 1.851199984550476,
+      "learning_rate": 0.00013137010676156585,
+      "loss": 2.4586,
+      "step": 3867
+    },
+    {
+      "epoch": 1.7191111111111113,
+      "grad_norm": 1.7658207416534424,
+      "learning_rate": 0.0001313523131672598,
+      "loss": 2.5013,
+      "step": 3868
+    },
+    {
+      "epoch": 1.7195555555555555,
+      "grad_norm": 2.4436843395233154,
+      "learning_rate": 0.00013133451957295374,
+      "loss": 2.5335,
+      "step": 3869
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8331066370010376,
+      "learning_rate": 0.0001313167259786477,
+      "loss": 2.3018,
+      "step": 3870
+    },
+    {
+      "epoch": 1.7204444444444444,
+      "grad_norm": 1.725740671157837,
+      "learning_rate": 0.00013129893238434163,
+      "loss": 2.4409,
+      "step": 3871
+    },
+    {
+      "epoch": 1.720888888888889,
+      "grad_norm": 2.027172088623047,
+      "learning_rate": 0.00013128113879003558,
+      "loss": 2.1246,
+      "step": 3872
+    },
+    {
+      "epoch": 1.7213333333333334,
+      "grad_norm": 1.6568158864974976,
+      "learning_rate": 0.00013126334519572954,
+      "loss": 1.8242,
+      "step": 3873
+    },
+    {
+      "epoch": 1.7217777777777776,
+      "grad_norm": 1.8668732643127441,
+      "learning_rate": 0.0001312455516014235,
+      "loss": 2.0447,
+      "step": 3874
+    },
+    {
+      "epoch": 1.7222222222222223,
+      "grad_norm": 1.9975069761276245,
+      "learning_rate": 0.00013122775800711745,
+      "loss": 2.1747,
+      "step": 3875
+    },
+    {
+      "epoch": 1.7226666666666666,
+      "grad_norm": 1.9104905128479004,
+      "learning_rate": 0.00013120996441281138,
+      "loss": 2.248,
+      "step": 3876
+    },
+    {
+      "epoch": 1.7231111111111113,
+      "grad_norm": 2.3680083751678467,
+      "learning_rate": 0.00013119217081850534,
+      "loss": 2.4361,
+      "step": 3877
+    },
+    {
+      "epoch": 1.7235555555555555,
+      "grad_norm": 2.0153682231903076,
+      "learning_rate": 0.0001311743772241993,
+      "loss": 1.9933,
+      "step": 3878
+    },
+    {
+      "epoch": 1.724,
+      "grad_norm": 2.112910032272339,
+      "learning_rate": 0.00013115658362989325,
+      "loss": 2.4397,
+      "step": 3879
+    },
+    {
+      "epoch": 1.7244444444444444,
+      "grad_norm": 1.9228010177612305,
+      "learning_rate": 0.0001311387900355872,
+      "loss": 2.7384,
+      "step": 3880
+    },
+    {
+      "epoch": 1.724888888888889,
+      "grad_norm": 2.054799795150757,
+      "learning_rate": 0.00013112099644128116,
+      "loss": 2.09,
+      "step": 3881
+    },
+    {
+      "epoch": 1.7253333333333334,
+      "grad_norm": 2.12286376953125,
+      "learning_rate": 0.0001311032028469751,
+      "loss": 2.0431,
+      "step": 3882
+    },
+    {
+      "epoch": 1.7257777777777776,
+      "grad_norm": 2.27078914642334,
+      "learning_rate": 0.00013108540925266905,
+      "loss": 2.5563,
+      "step": 3883
+    },
+    {
+      "epoch": 1.7262222222222223,
+      "grad_norm": 1.996488332748413,
+      "learning_rate": 0.00013106761565836298,
+      "loss": 2.4713,
+      "step": 3884
+    },
+    {
+      "epoch": 1.7266666666666666,
+      "grad_norm": 2.3658390045166016,
+      "learning_rate": 0.00013104982206405693,
+      "loss": 2.1061,
+      "step": 3885
+    },
+    {
+      "epoch": 1.7271111111111113,
+      "grad_norm": 1.9138381481170654,
+      "learning_rate": 0.0001310320284697509,
+      "loss": 2.6341,
+      "step": 3886
+    },
+    {
+      "epoch": 1.7275555555555555,
+      "grad_norm": 1.9364850521087646,
+      "learning_rate": 0.00013101423487544485,
+      "loss": 2.4424,
+      "step": 3887
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 2.730329751968384,
+      "learning_rate": 0.0001309964412811388,
+      "loss": 2.0531,
+      "step": 3888
+    },
+    {
+      "epoch": 1.7284444444444444,
+      "grad_norm": 2.1814632415771484,
+      "learning_rate": 0.00013097864768683273,
+      "loss": 2.3215,
+      "step": 3889
+    },
+    {
+      "epoch": 1.728888888888889,
+      "grad_norm": 2.137681484222412,
+      "learning_rate": 0.0001309608540925267,
+      "loss": 2.4898,
+      "step": 3890
+    },
+    {
+      "epoch": 1.7293333333333334,
+      "grad_norm": 2.490191698074341,
+      "learning_rate": 0.00013094306049822065,
+      "loss": 2.15,
+      "step": 3891
+    },
+    {
+      "epoch": 1.7297777777777776,
+      "grad_norm": 2.2890820503234863,
+      "learning_rate": 0.0001309252669039146,
+      "loss": 1.6876,
+      "step": 3892
+    },
+    {
+      "epoch": 1.7302222222222223,
+      "grad_norm": 1.9590189456939697,
+      "learning_rate": 0.00013090747330960856,
+      "loss": 1.9625,
+      "step": 3893
+    },
+    {
+      "epoch": 1.7306666666666666,
+      "grad_norm": 2.1057655811309814,
+      "learning_rate": 0.00013088967971530252,
+      "loss": 2.198,
+      "step": 3894
+    },
+    {
+      "epoch": 1.7311111111111113,
+      "grad_norm": 2.1267521381378174,
+      "learning_rate": 0.00013087188612099645,
+      "loss": 2.2824,
+      "step": 3895
+    },
+    {
+      "epoch": 1.7315555555555555,
+      "grad_norm": 1.9178341627120972,
+      "learning_rate": 0.0001308540925266904,
+      "loss": 1.9127,
+      "step": 3896
+    },
+    {
+      "epoch": 1.732,
+      "grad_norm": 2.151017189025879,
+      "learning_rate": 0.00013083629893238433,
+      "loss": 2.0591,
+      "step": 3897
+    },
+    {
+      "epoch": 1.7324444444444445,
+      "grad_norm": 2.1666769981384277,
+      "learning_rate": 0.0001308185053380783,
+      "loss": 2.1367,
+      "step": 3898
+    },
+    {
+      "epoch": 1.732888888888889,
+      "grad_norm": 2.2003917694091797,
+      "learning_rate": 0.00013080071174377224,
+      "loss": 0.9107,
+      "step": 3899
+    },
+    {
+      "epoch": 1.7333333333333334,
+      "grad_norm": 2.9451475143432617,
+      "learning_rate": 0.0001307829181494662,
+      "loss": 2.487,
+      "step": 3900
+    },
+    {
+      "epoch": 1.7337777777777776,
+      "grad_norm": 1.6602669954299927,
+      "learning_rate": 0.00013076512455516016,
+      "loss": 2.9994,
+      "step": 3901
+    },
+    {
+      "epoch": 1.7342222222222223,
+      "grad_norm": 1.3574813604354858,
+      "learning_rate": 0.0001307473309608541,
+      "loss": 2.2298,
+      "step": 3902
+    },
+    {
+      "epoch": 1.7346666666666666,
+      "grad_norm": 1.5164433717727661,
+      "learning_rate": 0.00013072953736654804,
+      "loss": 2.0965,
+      "step": 3903
+    },
+    {
+      "epoch": 1.7351111111111113,
+      "grad_norm": 1.5202592611312866,
+      "learning_rate": 0.000130711743772242,
+      "loss": 2.3827,
+      "step": 3904
+    },
+    {
+      "epoch": 1.7355555555555555,
+      "grad_norm": 1.3015217781066895,
+      "learning_rate": 0.00013069395017793596,
+      "loss": 1.6272,
+      "step": 3905
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 1.5526891946792603,
+      "learning_rate": 0.0001306761565836299,
+      "loss": 2.4231,
+      "step": 3906
+    },
+    {
+      "epoch": 1.7364444444444445,
+      "grad_norm": 1.6718441247940063,
+      "learning_rate": 0.00013065836298932387,
+      "loss": 2.8115,
+      "step": 3907
+    },
+    {
+      "epoch": 1.736888888888889,
+      "grad_norm": 1.7403420209884644,
+      "learning_rate": 0.0001306405693950178,
+      "loss": 2.7689,
+      "step": 3908
+    },
+    {
+      "epoch": 1.7373333333333334,
+      "grad_norm": 1.7474772930145264,
+      "learning_rate": 0.00013062277580071176,
+      "loss": 2.6111,
+      "step": 3909
+    },
+    {
+      "epoch": 1.7377777777777776,
+      "grad_norm": 1.6219923496246338,
+      "learning_rate": 0.00013060498220640568,
+      "loss": 2.5719,
+      "step": 3910
+    },
+    {
+      "epoch": 1.7382222222222223,
+      "grad_norm": 1.893513798713684,
+      "learning_rate": 0.00013058718861209964,
+      "loss": 2.7495,
+      "step": 3911
+    },
+    {
+      "epoch": 1.7386666666666666,
+      "grad_norm": 1.917389154434204,
+      "learning_rate": 0.0001305693950177936,
+      "loss": 2.4973,
+      "step": 3912
+    },
+    {
+      "epoch": 1.7391111111111113,
+      "grad_norm": 1.699096918106079,
+      "learning_rate": 0.00013055160142348755,
+      "loss": 1.7678,
+      "step": 3913
+    },
+    {
+      "epoch": 1.7395555555555555,
+      "grad_norm": 1.8528611660003662,
+      "learning_rate": 0.0001305338078291815,
+      "loss": 1.8641,
+      "step": 3914
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.741129994392395,
+      "learning_rate": 0.00013051601423487544,
+      "loss": 2.0575,
+      "step": 3915
+    },
+    {
+      "epoch": 1.7404444444444445,
+      "grad_norm": 1.7191723585128784,
+      "learning_rate": 0.0001304982206405694,
+      "loss": 2.1448,
+      "step": 3916
+    },
+    {
+      "epoch": 1.740888888888889,
+      "grad_norm": 1.8906670808792114,
+      "learning_rate": 0.00013048042704626335,
+      "loss": 2.3253,
+      "step": 3917
+    },
+    {
+      "epoch": 1.7413333333333334,
+      "grad_norm": 1.799043893814087,
+      "learning_rate": 0.0001304626334519573,
+      "loss": 2.2234,
+      "step": 3918
+    },
+    {
+      "epoch": 1.7417777777777776,
+      "grad_norm": 1.6781656742095947,
+      "learning_rate": 0.00013044483985765127,
+      "loss": 1.9896,
+      "step": 3919
+    },
+    {
+      "epoch": 1.7422222222222223,
+      "grad_norm": 1.7165361642837524,
+      "learning_rate": 0.00013042704626334522,
+      "loss": 2.0458,
+      "step": 3920
+    },
+    {
+      "epoch": 1.7426666666666666,
+      "grad_norm": 1.9061387777328491,
+      "learning_rate": 0.00013040925266903915,
+      "loss": 2.4378,
+      "step": 3921
+    },
+    {
+      "epoch": 1.743111111111111,
+      "grad_norm": 2.1114823818206787,
+      "learning_rate": 0.0001303914590747331,
+      "loss": 2.4924,
+      "step": 3922
+    },
+    {
+      "epoch": 1.7435555555555555,
+      "grad_norm": 1.9921493530273438,
+      "learning_rate": 0.00013037366548042704,
+      "loss": 2.5657,
+      "step": 3923
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 1.8910974264144897,
+      "learning_rate": 0.000130355871886121,
+      "loss": 2.0381,
+      "step": 3924
+    },
+    {
+      "epoch": 1.7444444444444445,
+      "grad_norm": 2.049863815307617,
+      "learning_rate": 0.00013033807829181495,
+      "loss": 1.9867,
+      "step": 3925
+    },
+    {
+      "epoch": 1.744888888888889,
+      "grad_norm": 1.9373703002929688,
+      "learning_rate": 0.0001303202846975089,
+      "loss": 2.3285,
+      "step": 3926
+    },
+    {
+      "epoch": 1.7453333333333334,
+      "grad_norm": 1.5811951160430908,
+      "learning_rate": 0.00013030249110320286,
+      "loss": 1.1489,
+      "step": 3927
+    },
+    {
+      "epoch": 1.7457777777777777,
+      "grad_norm": 2.237574577331543,
+      "learning_rate": 0.0001302846975088968,
+      "loss": 2.8071,
+      "step": 3928
+    },
+    {
+      "epoch": 1.7462222222222223,
+      "grad_norm": 1.7456693649291992,
+      "learning_rate": 0.00013026690391459075,
+      "loss": 1.5628,
+      "step": 3929
+    },
+    {
+      "epoch": 1.7466666666666666,
+      "grad_norm": 1.9937740564346313,
+      "learning_rate": 0.0001302491103202847,
+      "loss": 2.2982,
+      "step": 3930
+    },
+    {
+      "epoch": 1.747111111111111,
+      "grad_norm": 1.7053123712539673,
+      "learning_rate": 0.00013023131672597866,
+      "loss": 1.0467,
+      "step": 3931
+    },
+    {
+      "epoch": 1.7475555555555555,
+      "grad_norm": 2.3195912837982178,
+      "learning_rate": 0.00013021352313167262,
+      "loss": 2.5614,
+      "step": 3932
+    },
+    {
+      "epoch": 1.748,
+      "grad_norm": 2.1397838592529297,
+      "learning_rate": 0.00013019572953736655,
+      "loss": 1.9569,
+      "step": 3933
+    },
+    {
+      "epoch": 1.7484444444444445,
+      "grad_norm": 1.823757290840149,
+      "learning_rate": 0.0001301779359430605,
+      "loss": 1.8498,
+      "step": 3934
+    },
+    {
+      "epoch": 1.748888888888889,
+      "grad_norm": 1.9765830039978027,
+      "learning_rate": 0.00013016014234875443,
+      "loss": 2.2709,
+      "step": 3935
+    },
+    {
+      "epoch": 1.7493333333333334,
+      "grad_norm": 2.0341432094573975,
+      "learning_rate": 0.0001301423487544484,
+      "loss": 1.8853,
+      "step": 3936
+    },
+    {
+      "epoch": 1.7497777777777777,
+      "grad_norm": 1.8640036582946777,
+      "learning_rate": 0.00013012455516014235,
+      "loss": 1.9018,
+      "step": 3937
+    },
+    {
+      "epoch": 1.7502222222222223,
+      "grad_norm": 1.3533066511154175,
+      "learning_rate": 0.0001301067615658363,
+      "loss": 0.9755,
+      "step": 3938
+    },
+    {
+      "epoch": 1.7506666666666666,
+      "grad_norm": 2.0747861862182617,
+      "learning_rate": 0.00013008896797153026,
+      "loss": 2.1518,
+      "step": 3939
+    },
+    {
+      "epoch": 1.751111111111111,
+      "grad_norm": 2.1353771686553955,
+      "learning_rate": 0.0001300711743772242,
+      "loss": 2.3496,
+      "step": 3940
+    },
+    {
+      "epoch": 1.7515555555555555,
+      "grad_norm": 2.6912460327148438,
+      "learning_rate": 0.00013005338078291815,
+      "loss": 2.2062,
+      "step": 3941
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 1.905840277671814,
+      "learning_rate": 0.0001300355871886121,
+      "loss": 2.0854,
+      "step": 3942
+    },
+    {
+      "epoch": 1.7524444444444445,
+      "grad_norm": 2.2967617511749268,
+      "learning_rate": 0.00013001779359430606,
+      "loss": 1.8793,
+      "step": 3943
+    },
+    {
+      "epoch": 1.752888888888889,
+      "grad_norm": 2.267254590988159,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 1.8769,
+      "step": 3944
+    },
+    {
+      "epoch": 1.7533333333333334,
+      "grad_norm": 2.3356738090515137,
+      "learning_rate": 0.00012998220640569397,
+      "loss": 1.7641,
+      "step": 3945
+    },
+    {
+      "epoch": 1.7537777777777777,
+      "grad_norm": 2.722877264022827,
+      "learning_rate": 0.0001299644128113879,
+      "loss": 2.2982,
+      "step": 3946
+    },
+    {
+      "epoch": 1.7542222222222223,
+      "grad_norm": 2.4450271129608154,
+      "learning_rate": 0.00012994661921708186,
+      "loss": 2.1468,
+      "step": 3947
+    },
+    {
+      "epoch": 1.7546666666666666,
+      "grad_norm": 2.8750321865081787,
+      "learning_rate": 0.0001299288256227758,
+      "loss": 2.4269,
+      "step": 3948
+    },
+    {
+      "epoch": 1.755111111111111,
+      "grad_norm": 2.6799416542053223,
+      "learning_rate": 0.00012991103202846974,
+      "loss": 1.6945,
+      "step": 3949
+    },
+    {
+      "epoch": 1.7555555555555555,
+      "grad_norm": 3.447089433670044,
+      "learning_rate": 0.0001298932384341637,
+      "loss": 2.7562,
+      "step": 3950
+    },
+    {
+      "epoch": 1.756,
+      "grad_norm": 1.3063669204711914,
+      "learning_rate": 0.00012987544483985766,
+      "loss": 2.1046,
+      "step": 3951
+    },
+    {
+      "epoch": 1.7564444444444445,
+      "grad_norm": 1.4739446640014648,
+      "learning_rate": 0.0001298576512455516,
+      "loss": 2.3863,
+      "step": 3952
+    },
+    {
+      "epoch": 1.756888888888889,
+      "grad_norm": 1.543544888496399,
+      "learning_rate": 0.00012983985765124554,
+      "loss": 2.3732,
+      "step": 3953
+    },
+    {
+      "epoch": 1.7573333333333334,
+      "grad_norm": 1.6399084329605103,
+      "learning_rate": 0.0001298220640569395,
+      "loss": 2.5028,
+      "step": 3954
+    },
+    {
+      "epoch": 1.7577777777777777,
+      "grad_norm": 1.5914816856384277,
+      "learning_rate": 0.00012980427046263346,
+      "loss": 2.2064,
+      "step": 3955
+    },
+    {
+      "epoch": 1.7582222222222224,
+      "grad_norm": 1.6115366220474243,
+      "learning_rate": 0.0001297864768683274,
+      "loss": 2.2451,
+      "step": 3956
+    },
+    {
+      "epoch": 1.7586666666666666,
+      "grad_norm": 1.5162943601608276,
+      "learning_rate": 0.00012976868327402137,
+      "loss": 2.1785,
+      "step": 3957
+    },
+    {
+      "epoch": 1.759111111111111,
+      "grad_norm": 1.6043645143508911,
+      "learning_rate": 0.00012975088967971533,
+      "loss": 2.2091,
+      "step": 3958
+    },
+    {
+      "epoch": 1.7595555555555555,
+      "grad_norm": 1.4509108066558838,
+      "learning_rate": 0.00012973309608540925,
+      "loss": 2.3536,
+      "step": 3959
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6944358348846436,
+      "learning_rate": 0.0001297153024911032,
+      "loss": 2.5589,
+      "step": 3960
+    },
+    {
+      "epoch": 1.7604444444444445,
+      "grad_norm": 1.5533844232559204,
+      "learning_rate": 0.00012969750889679714,
+      "loss": 2.4115,
+      "step": 3961
+    },
+    {
+      "epoch": 1.7608888888888887,
+      "grad_norm": 1.4054555892944336,
+      "learning_rate": 0.0001296797153024911,
+      "loss": 2.1334,
+      "step": 3962
+    },
+    {
+      "epoch": 1.7613333333333334,
+      "grad_norm": 1.6660244464874268,
+      "learning_rate": 0.00012966192170818505,
+      "loss": 2.3195,
+      "step": 3963
+    },
+    {
+      "epoch": 1.7617777777777777,
+      "grad_norm": 1.5399365425109863,
+      "learning_rate": 0.000129644128113879,
+      "loss": 2.0814,
+      "step": 3964
+    },
+    {
+      "epoch": 1.7622222222222224,
+      "grad_norm": 1.5576121807098389,
+      "learning_rate": 0.00012962633451957297,
+      "loss": 1.8001,
+      "step": 3965
+    },
+    {
+      "epoch": 1.7626666666666666,
+      "grad_norm": 1.8270119428634644,
+      "learning_rate": 0.0001296085409252669,
+      "loss": 1.9568,
+      "step": 3966
+    },
+    {
+      "epoch": 1.763111111111111,
+      "grad_norm": 1.5569310188293457,
+      "learning_rate": 0.00012959074733096085,
+      "loss": 2.1702,
+      "step": 3967
+    },
+    {
+      "epoch": 1.7635555555555555,
+      "grad_norm": 1.7133228778839111,
+      "learning_rate": 0.0001295729537366548,
+      "loss": 2.4522,
+      "step": 3968
+    },
+    {
+      "epoch": 1.764,
+      "grad_norm": 1.6561294794082642,
+      "learning_rate": 0.00012955516014234877,
+      "loss": 2.0854,
+      "step": 3969
+    },
+    {
+      "epoch": 1.7644444444444445,
+      "grad_norm": 2.1431984901428223,
+      "learning_rate": 0.00012953736654804272,
+      "loss": 1.9774,
+      "step": 3970
+    },
+    {
+      "epoch": 1.7648888888888887,
+      "grad_norm": 1.6542658805847168,
+      "learning_rate": 0.00012951957295373668,
+      "loss": 1.9308,
+      "step": 3971
+    },
+    {
+      "epoch": 1.7653333333333334,
+      "grad_norm": 1.7912119626998901,
+      "learning_rate": 0.0001295017793594306,
+      "loss": 1.9613,
+      "step": 3972
+    },
+    {
+      "epoch": 1.7657777777777777,
+      "grad_norm": 1.2500418424606323,
+      "learning_rate": 0.00012948398576512456,
+      "loss": 0.6403,
+      "step": 3973
+    },
+    {
+      "epoch": 1.7662222222222224,
+      "grad_norm": 1.623761534690857,
+      "learning_rate": 0.0001294661921708185,
+      "loss": 2.2582,
+      "step": 3974
+    },
+    {
+      "epoch": 1.7666666666666666,
+      "grad_norm": 1.6567007303237915,
+      "learning_rate": 0.00012944839857651245,
+      "loss": 1.8474,
+      "step": 3975
+    },
+    {
+      "epoch": 1.767111111111111,
+      "grad_norm": 1.7294385433197021,
+      "learning_rate": 0.0001294306049822064,
+      "loss": 1.9536,
+      "step": 3976
+    },
+    {
+      "epoch": 1.7675555555555555,
+      "grad_norm": 1.9673961400985718,
+      "learning_rate": 0.00012941281138790036,
+      "loss": 2.0788,
+      "step": 3977
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 2.372072696685791,
+      "learning_rate": 0.00012939501779359432,
+      "loss": 2.1401,
+      "step": 3978
+    },
+    {
+      "epoch": 1.7684444444444445,
+      "grad_norm": 1.8470436334609985,
+      "learning_rate": 0.00012937722419928825,
+      "loss": 2.2509,
+      "step": 3979
+    },
+    {
+      "epoch": 1.7688888888888887,
+      "grad_norm": 1.7825738191604614,
+      "learning_rate": 0.0001293594306049822,
+      "loss": 2.1858,
+      "step": 3980
+    },
+    {
+      "epoch": 1.7693333333333334,
+      "grad_norm": 1.4077835083007812,
+      "learning_rate": 0.00012934163701067616,
+      "loss": 1.2364,
+      "step": 3981
+    },
+    {
+      "epoch": 1.7697777777777777,
+      "grad_norm": 2.0429136753082275,
+      "learning_rate": 0.00012932384341637012,
+      "loss": 2.2663,
+      "step": 3982
+    },
+    {
+      "epoch": 1.7702222222222224,
+      "grad_norm": 1.2380872964859009,
+      "learning_rate": 0.00012930604982206408,
+      "loss": 0.6942,
+      "step": 3983
+    },
+    {
+      "epoch": 1.7706666666666666,
+      "grad_norm": 2.0053813457489014,
+      "learning_rate": 0.00012928825622775803,
+      "loss": 2.1342,
+      "step": 3984
+    },
+    {
+      "epoch": 1.771111111111111,
+      "grad_norm": 1.8979686498641968,
+      "learning_rate": 0.00012927046263345196,
+      "loss": 2.4219,
+      "step": 3985
+    },
+    {
+      "epoch": 1.7715555555555556,
+      "grad_norm": 2.177061080932617,
+      "learning_rate": 0.00012925266903914592,
+      "loss": 2.2856,
+      "step": 3986
+    },
+    {
+      "epoch": 1.772,
+      "grad_norm": 2.1877963542938232,
+      "learning_rate": 0.00012923487544483985,
+      "loss": 2.0785,
+      "step": 3987
+    },
+    {
+      "epoch": 1.7724444444444445,
+      "grad_norm": 1.8126049041748047,
+      "learning_rate": 0.0001292170818505338,
+      "loss": 1.8393,
+      "step": 3988
+    },
+    {
+      "epoch": 1.7728888888888887,
+      "grad_norm": 2.09319806098938,
+      "learning_rate": 0.00012919928825622776,
+      "loss": 2.5437,
+      "step": 3989
+    },
+    {
+      "epoch": 1.7733333333333334,
+      "grad_norm": 1.2192652225494385,
+      "learning_rate": 0.00012918149466192172,
+      "loss": 1.0593,
+      "step": 3990
+    },
+    {
+      "epoch": 1.7737777777777777,
+      "grad_norm": 2.0721795558929443,
+      "learning_rate": 0.00012916370106761567,
+      "loss": 2.3111,
+      "step": 3991
+    },
+    {
+      "epoch": 1.7742222222222224,
+      "grad_norm": 1.9391847848892212,
+      "learning_rate": 0.0001291459074733096,
+      "loss": 1.9213,
+      "step": 3992
+    },
+    {
+      "epoch": 1.7746666666666666,
+      "grad_norm": 2.3300163745880127,
+      "learning_rate": 0.00012912811387900356,
+      "loss": 2.4885,
+      "step": 3993
+    },
+    {
+      "epoch": 1.775111111111111,
+      "grad_norm": 2.414308786392212,
+      "learning_rate": 0.00012911032028469752,
+      "loss": 2.4184,
+      "step": 3994
+    },
+    {
+      "epoch": 1.7755555555555556,
+      "grad_norm": 2.2488865852355957,
+      "learning_rate": 0.00012909252669039147,
+      "loss": 2.0251,
+      "step": 3995
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 3.426772117614746,
+      "learning_rate": 0.00012907473309608543,
+      "loss": 2.7753,
+      "step": 3996
+    },
+    {
+      "epoch": 1.7764444444444445,
+      "grad_norm": 2.1739494800567627,
+      "learning_rate": 0.00012905693950177938,
+      "loss": 2.0393,
+      "step": 3997
+    },
+    {
+      "epoch": 1.7768888888888887,
+      "grad_norm": 2.434339761734009,
+      "learning_rate": 0.00012903914590747331,
+      "loss": 2.2456,
+      "step": 3998
+    },
+    {
+      "epoch": 1.7773333333333334,
+      "grad_norm": 2.3951499462127686,
+      "learning_rate": 0.00012902135231316727,
+      "loss": 2.292,
+      "step": 3999
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 2.0735788345336914,
+      "learning_rate": 0.0001290035587188612,
+      "loss": 1.3612,
+      "step": 4000
+    },
+    {
+      "epoch": 1.7782222222222224,
+      "grad_norm": 0.45426124334335327,
+      "learning_rate": 0.00012898576512455516,
+      "loss": 0.0292,
+      "step": 4001
+    },
+    {
+      "epoch": 1.7786666666666666,
+      "grad_norm": 1.561895728111267,
+      "learning_rate": 0.0001289679715302491,
+      "loss": 2.864,
+      "step": 4002
+    },
+    {
+      "epoch": 1.779111111111111,
+      "grad_norm": 1.1529381275177002,
+      "learning_rate": 0.00012895017793594307,
+      "loss": 1.1453,
+      "step": 4003
+    },
+    {
+      "epoch": 1.7795555555555556,
+      "grad_norm": 1.7909576892852783,
+      "learning_rate": 0.00012893238434163703,
+      "loss": 2.1009,
+      "step": 4004
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.8325178623199463,
+      "learning_rate": 0.00012891459074733096,
+      "loss": 2.4447,
+      "step": 4005
+    },
+    {
+      "epoch": 1.7804444444444445,
+      "grad_norm": 1.695643663406372,
+      "learning_rate": 0.0001288967971530249,
+      "loss": 2.5286,
+      "step": 4006
+    },
+    {
+      "epoch": 1.7808888888888887,
+      "grad_norm": 1.6463677883148193,
+      "learning_rate": 0.00012887900355871887,
+      "loss": 2.2066,
+      "step": 4007
+    },
+    {
+      "epoch": 1.7813333333333334,
+      "grad_norm": 1.7361533641815186,
+      "learning_rate": 0.00012886120996441282,
+      "loss": 2.2354,
+      "step": 4008
+    },
+    {
+      "epoch": 1.7817777777777777,
+      "grad_norm": 2.212310791015625,
+      "learning_rate": 0.00012884341637010678,
+      "loss": 2.1203,
+      "step": 4009
+    },
+    {
+      "epoch": 1.7822222222222224,
+      "grad_norm": 1.4698344469070435,
+      "learning_rate": 0.00012882562277580074,
+      "loss": 1.8529,
+      "step": 4010
+    },
+    {
+      "epoch": 1.7826666666666666,
+      "grad_norm": 1.823663592338562,
+      "learning_rate": 0.00012880782918149467,
+      "loss": 2.2639,
+      "step": 4011
+    },
+    {
+      "epoch": 1.783111111111111,
+      "grad_norm": 1.7825374603271484,
+      "learning_rate": 0.00012879003558718862,
+      "loss": 1.9933,
+      "step": 4012
+    },
+    {
+      "epoch": 1.7835555555555556,
+      "grad_norm": 1.8066425323486328,
+      "learning_rate": 0.00012877224199288255,
+      "loss": 2.1136,
+      "step": 4013
+    },
+    {
+      "epoch": 1.784,
+      "grad_norm": 2.2551443576812744,
+      "learning_rate": 0.0001287544483985765,
+      "loss": 1.9374,
+      "step": 4014
+    },
+    {
+      "epoch": 1.7844444444444445,
+      "grad_norm": 1.7912770509719849,
+      "learning_rate": 0.00012873665480427047,
+      "loss": 2.2331,
+      "step": 4015
+    },
+    {
+      "epoch": 1.7848888888888887,
+      "grad_norm": 1.5528522729873657,
+      "learning_rate": 0.00012871886120996442,
+      "loss": 1.9012,
+      "step": 4016
+    },
+    {
+      "epoch": 1.7853333333333334,
+      "grad_norm": 1.6881825923919678,
+      "learning_rate": 0.00012870106761565838,
+      "loss": 2.061,
+      "step": 4017
+    },
+    {
+      "epoch": 1.7857777777777777,
+      "grad_norm": 1.2570315599441528,
+      "learning_rate": 0.0001286832740213523,
+      "loss": 1.2893,
+      "step": 4018
+    },
+    {
+      "epoch": 1.7862222222222224,
+      "grad_norm": 1.4894715547561646,
+      "learning_rate": 0.00012866548042704627,
+      "loss": 1.7158,
+      "step": 4019
+    },
+    {
+      "epoch": 1.7866666666666666,
+      "grad_norm": 1.969714879989624,
+      "learning_rate": 0.00012864768683274022,
+      "loss": 2.3733,
+      "step": 4020
+    },
+    {
+      "epoch": 1.787111111111111,
+      "grad_norm": 1.8792929649353027,
+      "learning_rate": 0.00012862989323843418,
+      "loss": 2.1233,
+      "step": 4021
+    },
+    {
+      "epoch": 1.7875555555555556,
+      "grad_norm": 1.8268121480941772,
+      "learning_rate": 0.00012861209964412813,
+      "loss": 2.136,
+      "step": 4022
+    },
+    {
+      "epoch": 1.788,
+      "grad_norm": 2.0907411575317383,
+      "learning_rate": 0.00012859430604982206,
+      "loss": 2.3892,
+      "step": 4023
+    },
+    {
+      "epoch": 1.7884444444444445,
+      "grad_norm": 1.901108741760254,
+      "learning_rate": 0.00012857651245551602,
+      "loss": 1.938,
+      "step": 4024
+    },
+    {
+      "epoch": 1.7888888888888888,
+      "grad_norm": 1.818852186203003,
+      "learning_rate": 0.00012855871886120998,
+      "loss": 2.5912,
+      "step": 4025
+    },
+    {
+      "epoch": 1.7893333333333334,
+      "grad_norm": 1.6948503255844116,
+      "learning_rate": 0.0001285409252669039,
+      "loss": 1.8355,
+      "step": 4026
+    },
+    {
+      "epoch": 1.7897777777777777,
+      "grad_norm": 2.053222417831421,
+      "learning_rate": 0.00012852313167259786,
+      "loss": 1.7958,
+      "step": 4027
+    },
+    {
+      "epoch": 1.7902222222222224,
+      "grad_norm": 1.777777910232544,
+      "learning_rate": 0.00012850533807829182,
+      "loss": 2.276,
+      "step": 4028
+    },
+    {
+      "epoch": 1.7906666666666666,
+      "grad_norm": 2.1247165203094482,
+      "learning_rate": 0.00012848754448398578,
+      "loss": 2.275,
+      "step": 4029
+    },
+    {
+      "epoch": 1.791111111111111,
+      "grad_norm": 2.1670467853546143,
+      "learning_rate": 0.0001284697508896797,
+      "loss": 2.1813,
+      "step": 4030
+    },
+    {
+      "epoch": 1.7915555555555556,
+      "grad_norm": 1.7745897769927979,
+      "learning_rate": 0.00012845195729537366,
+      "loss": 1.7079,
+      "step": 4031
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 1.98904287815094,
+      "learning_rate": 0.00012843416370106762,
+      "loss": 1.9803,
+      "step": 4032
+    },
+    {
+      "epoch": 1.7924444444444445,
+      "grad_norm": 1.8329356908798218,
+      "learning_rate": 0.00012841637010676157,
+      "loss": 2.3103,
+      "step": 4033
+    },
+    {
+      "epoch": 1.7928888888888888,
+      "grad_norm": 1.814437985420227,
+      "learning_rate": 0.00012839857651245553,
+      "loss": 2.0819,
+      "step": 4034
+    },
+    {
+      "epoch": 1.7933333333333334,
+      "grad_norm": 1.9947712421417236,
+      "learning_rate": 0.0001283807829181495,
+      "loss": 2.1793,
+      "step": 4035
+    },
+    {
+      "epoch": 1.7937777777777777,
+      "grad_norm": 1.7744113206863403,
+      "learning_rate": 0.00012836298932384342,
+      "loss": 1.4961,
+      "step": 4036
+    },
+    {
+      "epoch": 1.7942222222222224,
+      "grad_norm": 1.941453456878662,
+      "learning_rate": 0.00012834519572953737,
+      "loss": 1.5453,
+      "step": 4037
+    },
+    {
+      "epoch": 1.7946666666666666,
+      "grad_norm": 1.993354320526123,
+      "learning_rate": 0.00012832740213523133,
+      "loss": 2.5863,
+      "step": 4038
+    },
+    {
+      "epoch": 1.795111111111111,
+      "grad_norm": 1.9085299968719482,
+      "learning_rate": 0.00012830960854092526,
+      "loss": 2.0983,
+      "step": 4039
+    },
+    {
+      "epoch": 1.7955555555555556,
+      "grad_norm": 2.2730631828308105,
+      "learning_rate": 0.00012829181494661922,
+      "loss": 2.894,
+      "step": 4040
+    },
+    {
+      "epoch": 1.796,
+      "grad_norm": 2.3365797996520996,
+      "learning_rate": 0.00012827402135231317,
+      "loss": 2.123,
+      "step": 4041
+    },
+    {
+      "epoch": 1.7964444444444445,
+      "grad_norm": 1.8728188276290894,
+      "learning_rate": 0.00012825622775800713,
+      "loss": 1.4073,
+      "step": 4042
+    },
+    {
+      "epoch": 1.7968888888888888,
+      "grad_norm": 2.5472781658172607,
+      "learning_rate": 0.00012823843416370106,
+      "loss": 2.2789,
+      "step": 4043
+    },
+    {
+      "epoch": 1.7973333333333334,
+      "grad_norm": 2.269137382507324,
+      "learning_rate": 0.00012822064056939501,
+      "loss": 2.3703,
+      "step": 4044
+    },
+    {
+      "epoch": 1.7977777777777777,
+      "grad_norm": 2.4079058170318604,
+      "learning_rate": 0.00012820284697508897,
+      "loss": 1.6054,
+      "step": 4045
+    },
+    {
+      "epoch": 1.7982222222222224,
+      "grad_norm": 2.692018747329712,
+      "learning_rate": 0.00012818505338078293,
+      "loss": 2.6721,
+      "step": 4046
+    },
+    {
+      "epoch": 1.7986666666666666,
+      "grad_norm": 2.388993740081787,
+      "learning_rate": 0.00012816725978647688,
+      "loss": 2.0374,
+      "step": 4047
+    },
+    {
+      "epoch": 1.799111111111111,
+      "grad_norm": 1.563175082206726,
+      "learning_rate": 0.00012814946619217084,
+      "loss": 0.9111,
+      "step": 4048
+    },
+    {
+      "epoch": 1.7995555555555556,
+      "grad_norm": 2.669541835784912,
+      "learning_rate": 0.00012813167259786477,
+      "loss": 1.4868,
+      "step": 4049
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.8668127059936523,
+      "learning_rate": 0.00012811387900355873,
+      "loss": 1.8904,
+      "step": 4050
+    },
+    {
+      "epoch": 1.8004444444444445,
+      "grad_norm": 1.4953787326812744,
+      "learning_rate": 0.00012809608540925266,
+      "loss": 2.3084,
+      "step": 4051
+    },
+    {
+      "epoch": 1.8008888888888888,
+      "grad_norm": 1.6110552549362183,
+      "learning_rate": 0.0001280782918149466,
+      "loss": 2.9876,
+      "step": 4052
+    },
+    {
+      "epoch": 1.8013333333333335,
+      "grad_norm": 1.4815409183502197,
+      "learning_rate": 0.00012806049822064057,
+      "loss": 2.3245,
+      "step": 4053
+    },
+    {
+      "epoch": 1.8017777777777777,
+      "grad_norm": 1.6469542980194092,
+      "learning_rate": 0.00012804270462633453,
+      "loss": 2.4534,
+      "step": 4054
+    },
+    {
+      "epoch": 1.8022222222222222,
+      "grad_norm": 1.7720386981964111,
+      "learning_rate": 0.00012802491103202848,
+      "loss": 2.5591,
+      "step": 4055
+    },
+    {
+      "epoch": 1.8026666666666666,
+      "grad_norm": 1.607649564743042,
+      "learning_rate": 0.0001280071174377224,
+      "loss": 2.3919,
+      "step": 4056
+    },
+    {
+      "epoch": 1.803111111111111,
+      "grad_norm": 1.521120548248291,
+      "learning_rate": 0.00012798932384341637,
+      "loss": 2.292,
+      "step": 4057
+    },
+    {
+      "epoch": 1.8035555555555556,
+      "grad_norm": 1.7461004257202148,
+      "learning_rate": 0.00012797153024911032,
+      "loss": 2.1695,
+      "step": 4058
+    },
+    {
+      "epoch": 1.804,
+      "grad_norm": 1.6158878803253174,
+      "learning_rate": 0.00012795373665480428,
+      "loss": 1.4256,
+      "step": 4059
+    },
+    {
+      "epoch": 1.8044444444444445,
+      "grad_norm": 1.6743505001068115,
+      "learning_rate": 0.00012793594306049824,
+      "loss": 1.9986,
+      "step": 4060
+    },
+    {
+      "epoch": 1.8048888888888888,
+      "grad_norm": 1.6149520874023438,
+      "learning_rate": 0.0001279181494661922,
+      "loss": 2.5781,
+      "step": 4061
+    },
+    {
+      "epoch": 1.8053333333333335,
+      "grad_norm": 1.5467309951782227,
+      "learning_rate": 0.00012790035587188612,
+      "loss": 1.7904,
+      "step": 4062
+    },
+    {
+      "epoch": 1.8057777777777777,
+      "grad_norm": 1.7630541324615479,
+      "learning_rate": 0.00012788256227758008,
+      "loss": 2.2801,
+      "step": 4063
+    },
+    {
+      "epoch": 1.8062222222222222,
+      "grad_norm": 1.8055700063705444,
+      "learning_rate": 0.000127864768683274,
+      "loss": 1.9704,
+      "step": 4064
+    },
+    {
+      "epoch": 1.8066666666666666,
+      "grad_norm": 1.7958096265792847,
+      "learning_rate": 0.00012784697508896797,
+      "loss": 2.3531,
+      "step": 4065
+    },
+    {
+      "epoch": 1.8071111111111111,
+      "grad_norm": 1.426638126373291,
+      "learning_rate": 0.00012782918149466192,
+      "loss": 1.2682,
+      "step": 4066
+    },
+    {
+      "epoch": 1.8075555555555556,
+      "grad_norm": 1.6399476528167725,
+      "learning_rate": 0.00012781138790035588,
+      "loss": 2.1429,
+      "step": 4067
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 1.7020010948181152,
+      "learning_rate": 0.00012779359430604984,
+      "loss": 2.2423,
+      "step": 4068
+    },
+    {
+      "epoch": 1.8084444444444445,
+      "grad_norm": 1.5817952156066895,
+      "learning_rate": 0.00012777580071174376,
+      "loss": 1.7548,
+      "step": 4069
+    },
+    {
+      "epoch": 1.8088888888888888,
+      "grad_norm": 1.6301820278167725,
+      "learning_rate": 0.00012775800711743772,
+      "loss": 2.1788,
+      "step": 4070
+    },
+    {
+      "epoch": 1.8093333333333335,
+      "grad_norm": 1.7955340147018433,
+      "learning_rate": 0.00012774021352313168,
+      "loss": 1.756,
+      "step": 4071
+    },
+    {
+      "epoch": 1.8097777777777777,
+      "grad_norm": 1.5359266996383667,
+      "learning_rate": 0.00012772241992882563,
+      "loss": 1.714,
+      "step": 4072
+    },
+    {
+      "epoch": 1.8102222222222222,
+      "grad_norm": 1.6445212364196777,
+      "learning_rate": 0.0001277046263345196,
+      "loss": 1.9434,
+      "step": 4073
+    },
+    {
+      "epoch": 1.8106666666666666,
+      "grad_norm": 1.811740517616272,
+      "learning_rate": 0.00012768683274021355,
+      "loss": 1.7521,
+      "step": 4074
+    },
+    {
+      "epoch": 1.8111111111111111,
+      "grad_norm": 1.9442163705825806,
+      "learning_rate": 0.00012766903914590748,
+      "loss": 2.284,
+      "step": 4075
+    },
+    {
+      "epoch": 1.8115555555555556,
+      "grad_norm": 2.383162498474121,
+      "learning_rate": 0.00012765124555160143,
+      "loss": 2.1859,
+      "step": 4076
+    },
+    {
+      "epoch": 1.812,
+      "grad_norm": 1.7321439981460571,
+      "learning_rate": 0.00012763345195729536,
+      "loss": 1.6559,
+      "step": 4077
+    },
+    {
+      "epoch": 1.8124444444444445,
+      "grad_norm": 1.9843666553497314,
+      "learning_rate": 0.00012761565836298932,
+      "loss": 2.1682,
+      "step": 4078
+    },
+    {
+      "epoch": 1.8128888888888888,
+      "grad_norm": 1.7194164991378784,
+      "learning_rate": 0.00012759786476868328,
+      "loss": 1.4098,
+      "step": 4079
+    },
+    {
+      "epoch": 1.8133333333333335,
+      "grad_norm": 1.5130424499511719,
+      "learning_rate": 0.00012758007117437723,
+      "loss": 1.1193,
+      "step": 4080
+    },
+    {
+      "epoch": 1.8137777777777777,
+      "grad_norm": 2.0162298679351807,
+      "learning_rate": 0.0001275622775800712,
+      "loss": 2.1374,
+      "step": 4081
+    },
+    {
+      "epoch": 1.8142222222222222,
+      "grad_norm": 2.095752716064453,
+      "learning_rate": 0.00012754448398576512,
+      "loss": 1.7003,
+      "step": 4082
+    },
+    {
+      "epoch": 1.8146666666666667,
+      "grad_norm": 1.4571044445037842,
+      "learning_rate": 0.00012752669039145907,
+      "loss": 0.9885,
+      "step": 4083
+    },
+    {
+      "epoch": 1.8151111111111111,
+      "grad_norm": 2.0335910320281982,
+      "learning_rate": 0.00012750889679715303,
+      "loss": 1.7632,
+      "step": 4084
+    },
+    {
+      "epoch": 1.8155555555555556,
+      "grad_norm": 2.1240885257720947,
+      "learning_rate": 0.000127491103202847,
+      "loss": 1.9801,
+      "step": 4085
+    },
+    {
+      "epoch": 1.8159999999999998,
+      "grad_norm": 2.1505091190338135,
+      "learning_rate": 0.00012747330960854094,
+      "loss": 2.513,
+      "step": 4086
+    },
+    {
+      "epoch": 1.8164444444444445,
+      "grad_norm": 2.2950072288513184,
+      "learning_rate": 0.0001274555160142349,
+      "loss": 1.9917,
+      "step": 4087
+    },
+    {
+      "epoch": 1.8168888888888888,
+      "grad_norm": 2.2907001972198486,
+      "learning_rate": 0.00012743772241992883,
+      "loss": 2.3145,
+      "step": 4088
+    },
+    {
+      "epoch": 1.8173333333333335,
+      "grad_norm": 2.215815782546997,
+      "learning_rate": 0.00012741992882562279,
+      "loss": 1.8032,
+      "step": 4089
+    },
+    {
+      "epoch": 1.8177777777777777,
+      "grad_norm": 2.2718591690063477,
+      "learning_rate": 0.00012740213523131672,
+      "loss": 2.2278,
+      "step": 4090
+    },
+    {
+      "epoch": 1.8182222222222222,
+      "grad_norm": 2.494612455368042,
+      "learning_rate": 0.00012738434163701067,
+      "loss": 2.8879,
+      "step": 4091
+    },
+    {
+      "epoch": 1.8186666666666667,
+      "grad_norm": 2.505629539489746,
+      "learning_rate": 0.00012736654804270463,
+      "loss": 2.0987,
+      "step": 4092
+    },
+    {
+      "epoch": 1.8191111111111111,
+      "grad_norm": 2.47916841506958,
+      "learning_rate": 0.00012734875444839859,
+      "loss": 2.3081,
+      "step": 4093
+    },
+    {
+      "epoch": 1.8195555555555556,
+      "grad_norm": 2.0762839317321777,
+      "learning_rate": 0.00012733096085409254,
+      "loss": 1.929,
+      "step": 4094
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 2.3247344493865967,
+      "learning_rate": 0.00012731316725978647,
+      "loss": 2.4938,
+      "step": 4095
+    },
+    {
+      "epoch": 1.8204444444444445,
+      "grad_norm": 2.481536865234375,
+      "learning_rate": 0.00012729537366548043,
+      "loss": 2.1279,
+      "step": 4096
+    },
+    {
+      "epoch": 1.8208888888888888,
+      "grad_norm": 2.1468632221221924,
+      "learning_rate": 0.00012727758007117438,
+      "loss": 2.2491,
+      "step": 4097
+    },
+    {
+      "epoch": 1.8213333333333335,
+      "grad_norm": 2.509892702102661,
+      "learning_rate": 0.00012725978647686834,
+      "loss": 2.1366,
+      "step": 4098
+    },
+    {
+      "epoch": 1.8217777777777777,
+      "grad_norm": 2.757918119430542,
+      "learning_rate": 0.0001272419928825623,
+      "loss": 2.3787,
+      "step": 4099
+    },
+    {
+      "epoch": 1.8222222222222222,
+      "grad_norm": 2.95184326171875,
+      "learning_rate": 0.00012722419928825625,
+      "loss": 1.7402,
+      "step": 4100
+    },
+    {
+      "epoch": 1.8226666666666667,
+      "grad_norm": 1.111022710800171,
+      "learning_rate": 0.00012720640569395018,
+      "loss": 1.3282,
+      "step": 4101
+    },
+    {
+      "epoch": 1.8231111111111111,
+      "grad_norm": 1.570564866065979,
+      "learning_rate": 0.00012718861209964414,
+      "loss": 2.5706,
+      "step": 4102
+    },
+    {
+      "epoch": 1.8235555555555556,
+      "grad_norm": 1.5850750207901,
+      "learning_rate": 0.00012717081850533807,
+      "loss": 2.603,
+      "step": 4103
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 1.873236060142517,
+      "learning_rate": 0.00012715302491103203,
+      "loss": 2.306,
+      "step": 4104
+    },
+    {
+      "epoch": 1.8244444444444445,
+      "grad_norm": 1.5911369323730469,
+      "learning_rate": 0.00012713523131672598,
+      "loss": 2.377,
+      "step": 4105
+    },
+    {
+      "epoch": 1.8248888888888888,
+      "grad_norm": 1.6573972702026367,
+      "learning_rate": 0.00012711743772241994,
+      "loss": 2.1528,
+      "step": 4106
+    },
+    {
+      "epoch": 1.8253333333333335,
+      "grad_norm": 1.76816987991333,
+      "learning_rate": 0.0001270996441281139,
+      "loss": 1.9276,
+      "step": 4107
+    },
+    {
+      "epoch": 1.8257777777777777,
+      "grad_norm": 1.6484469175338745,
+      "learning_rate": 0.00012708185053380782,
+      "loss": 2.2868,
+      "step": 4108
+    },
+    {
+      "epoch": 1.8262222222222222,
+      "grad_norm": 1.8483304977416992,
+      "learning_rate": 0.00012706405693950178,
+      "loss": 1.8456,
+      "step": 4109
+    },
+    {
+      "epoch": 1.8266666666666667,
+      "grad_norm": 1.5337408781051636,
+      "learning_rate": 0.00012704626334519574,
+      "loss": 1.3085,
+      "step": 4110
+    },
+    {
+      "epoch": 1.8271111111111111,
+      "grad_norm": 1.832377314567566,
+      "learning_rate": 0.0001270284697508897,
+      "loss": 2.0178,
+      "step": 4111
+    },
+    {
+      "epoch": 1.8275555555555556,
+      "grad_norm": 1.631611943244934,
+      "learning_rate": 0.00012701067615658365,
+      "loss": 2.3294,
+      "step": 4112
+    },
+    {
+      "epoch": 1.8279999999999998,
+      "grad_norm": 1.6167796850204468,
+      "learning_rate": 0.00012699288256227758,
+      "loss": 2.2195,
+      "step": 4113
+    },
+    {
+      "epoch": 1.8284444444444445,
+      "grad_norm": 1.6683152914047241,
+      "learning_rate": 0.00012697508896797154,
+      "loss": 2.4356,
+      "step": 4114
+    },
+    {
+      "epoch": 1.8288888888888888,
+      "grad_norm": 1.8200979232788086,
+      "learning_rate": 0.0001269572953736655,
+      "loss": 2.0209,
+      "step": 4115
+    },
+    {
+      "epoch": 1.8293333333333335,
+      "grad_norm": 1.7240800857543945,
+      "learning_rate": 0.00012693950177935942,
+      "loss": 2.5672,
+      "step": 4116
+    },
+    {
+      "epoch": 1.8297777777777777,
+      "grad_norm": 1.9097305536270142,
+      "learning_rate": 0.00012692170818505338,
+      "loss": 2.4372,
+      "step": 4117
+    },
+    {
+      "epoch": 1.8302222222222222,
+      "grad_norm": 1.1662497520446777,
+      "learning_rate": 0.00012690391459074733,
+      "loss": 1.1567,
+      "step": 4118
+    },
+    {
+      "epoch": 1.8306666666666667,
+      "grad_norm": 1.9544744491577148,
+      "learning_rate": 0.0001268861209964413,
+      "loss": 2.2734,
+      "step": 4119
+    },
+    {
+      "epoch": 1.8311111111111111,
+      "grad_norm": 1.7287496328353882,
+      "learning_rate": 0.00012686832740213522,
+      "loss": 1.9544,
+      "step": 4120
+    },
+    {
+      "epoch": 1.8315555555555556,
+      "grad_norm": 1.6461305618286133,
+      "learning_rate": 0.00012685053380782918,
+      "loss": 2.2844,
+      "step": 4121
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 1.8667876720428467,
+      "learning_rate": 0.00012683274021352313,
+      "loss": 2.483,
+      "step": 4122
+    },
+    {
+      "epoch": 1.8324444444444445,
+      "grad_norm": 1.8884638547897339,
+      "learning_rate": 0.0001268149466192171,
+      "loss": 2.2412,
+      "step": 4123
+    },
+    {
+      "epoch": 1.8328888888888888,
+      "grad_norm": 1.9503480195999146,
+      "learning_rate": 0.00012679715302491105,
+      "loss": 2.3155,
+      "step": 4124
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 1.864048719406128,
+      "learning_rate": 0.000126779359430605,
+      "loss": 1.9597,
+      "step": 4125
+    },
+    {
+      "epoch": 1.8337777777777777,
+      "grad_norm": 1.8108760118484497,
+      "learning_rate": 0.00012676156583629893,
+      "loss": 2.4295,
+      "step": 4126
+    },
+    {
+      "epoch": 1.8342222222222222,
+      "grad_norm": 2.0015783309936523,
+      "learning_rate": 0.0001267437722419929,
+      "loss": 2.2783,
+      "step": 4127
+    },
+    {
+      "epoch": 1.8346666666666667,
+      "grad_norm": 1.8798251152038574,
+      "learning_rate": 0.00012672597864768685,
+      "loss": 1.6982,
+      "step": 4128
+    },
+    {
+      "epoch": 1.8351111111111111,
+      "grad_norm": 2.052774667739868,
+      "learning_rate": 0.00012670818505338078,
+      "loss": 2.3289,
+      "step": 4129
+    },
+    {
+      "epoch": 1.8355555555555556,
+      "grad_norm": 2.0394935607910156,
+      "learning_rate": 0.00012669039145907473,
+      "loss": 2.3255,
+      "step": 4130
+    },
+    {
+      "epoch": 1.8359999999999999,
+      "grad_norm": 3.176072120666504,
+      "learning_rate": 0.0001266725978647687,
+      "loss": 2.1502,
+      "step": 4131
+    },
+    {
+      "epoch": 1.8364444444444445,
+      "grad_norm": 2.176759719848633,
+      "learning_rate": 0.00012665480427046264,
+      "loss": 2.0181,
+      "step": 4132
+    },
+    {
+      "epoch": 1.8368888888888888,
+      "grad_norm": 1.9604226350784302,
+      "learning_rate": 0.00012663701067615657,
+      "loss": 2.3706,
+      "step": 4133
+    },
+    {
+      "epoch": 1.8373333333333335,
+      "grad_norm": 1.9100964069366455,
+      "learning_rate": 0.00012661921708185053,
+      "loss": 2.1983,
+      "step": 4134
+    },
+    {
+      "epoch": 1.8377777777777777,
+      "grad_norm": 1.7293579578399658,
+      "learning_rate": 0.0001266014234875445,
+      "loss": 2.0016,
+      "step": 4135
+    },
+    {
+      "epoch": 1.8382222222222222,
+      "grad_norm": 1.660583257675171,
+      "learning_rate": 0.00012658362989323844,
+      "loss": 1.5306,
+      "step": 4136
+    },
+    {
+      "epoch": 1.8386666666666667,
+      "grad_norm": 1.858089804649353,
+      "learning_rate": 0.0001265658362989324,
+      "loss": 2.1515,
+      "step": 4137
+    },
+    {
+      "epoch": 1.8391111111111111,
+      "grad_norm": 2.4940619468688965,
+      "learning_rate": 0.00012654804270462636,
+      "loss": 1.9427,
+      "step": 4138
+    },
+    {
+      "epoch": 1.8395555555555556,
+      "grad_norm": 2.1511716842651367,
+      "learning_rate": 0.00012653024911032029,
+      "loss": 2.0729,
+      "step": 4139
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 2.469057321548462,
+      "learning_rate": 0.00012651245551601424,
+      "loss": 1.8632,
+      "step": 4140
+    },
+    {
+      "epoch": 1.8404444444444445,
+      "grad_norm": 2.173260450363159,
+      "learning_rate": 0.0001264946619217082,
+      "loss": 2.422,
+      "step": 4141
+    },
+    {
+      "epoch": 1.8408888888888888,
+      "grad_norm": 2.3541669845581055,
+      "learning_rate": 0.00012647686832740213,
+      "loss": 2.3772,
+      "step": 4142
+    },
+    {
+      "epoch": 1.8413333333333335,
+      "grad_norm": 2.6083996295928955,
+      "learning_rate": 0.00012645907473309608,
+      "loss": 2.3829,
+      "step": 4143
+    },
+    {
+      "epoch": 1.8417777777777777,
+      "grad_norm": 2.0515520572662354,
+      "learning_rate": 0.00012644128113879004,
+      "loss": 2.0062,
+      "step": 4144
+    },
+    {
+      "epoch": 1.8422222222222222,
+      "grad_norm": 2.331029176712036,
+      "learning_rate": 0.000126423487544484,
+      "loss": 2.1876,
+      "step": 4145
+    },
+    {
+      "epoch": 1.8426666666666667,
+      "grad_norm": 2.6274282932281494,
+      "learning_rate": 0.00012640569395017793,
+      "loss": 2.344,
+      "step": 4146
+    },
+    {
+      "epoch": 1.8431111111111111,
+      "grad_norm": 2.3808679580688477,
+      "learning_rate": 0.00012638790035587188,
+      "loss": 1.9633,
+      "step": 4147
+    },
+    {
+      "epoch": 1.8435555555555556,
+      "grad_norm": 2.2616419792175293,
+      "learning_rate": 0.00012637010676156584,
+      "loss": 2.2526,
+      "step": 4148
+    },
+    {
+      "epoch": 1.8439999999999999,
+      "grad_norm": 1.2193570137023926,
+      "learning_rate": 0.0001263523131672598,
+      "loss": 0.1276,
+      "step": 4149
+    },
+    {
+      "epoch": 1.8444444444444446,
+      "grad_norm": 2.9776499271392822,
+      "learning_rate": 0.00012633451957295375,
+      "loss": 1.8211,
+      "step": 4150
+    },
+    {
+      "epoch": 1.8448888888888888,
+      "grad_norm": 1.406093955039978,
+      "learning_rate": 0.0001263167259786477,
+      "loss": 2.3213,
+      "step": 4151
+    },
+    {
+      "epoch": 1.8453333333333335,
+      "grad_norm": 1.0631603002548218,
+      "learning_rate": 0.00012629893238434164,
+      "loss": 1.0766,
+      "step": 4152
+    },
+    {
+      "epoch": 1.8457777777777777,
+      "grad_norm": 1.5336110591888428,
+      "learning_rate": 0.0001262811387900356,
+      "loss": 2.4137,
+      "step": 4153
+    },
+    {
+      "epoch": 1.8462222222222222,
+      "grad_norm": 1.60658597946167,
+      "learning_rate": 0.00012626334519572955,
+      "loss": 2.7207,
+      "step": 4154
+    },
+    {
+      "epoch": 1.8466666666666667,
+      "grad_norm": 1.6107380390167236,
+      "learning_rate": 0.00012624555160142348,
+      "loss": 2.6709,
+      "step": 4155
+    },
+    {
+      "epoch": 1.8471111111111111,
+      "grad_norm": 1.7542976140975952,
+      "learning_rate": 0.00012622775800711744,
+      "loss": 2.6682,
+      "step": 4156
+    },
+    {
+      "epoch": 1.8475555555555556,
+      "grad_norm": 1.7753915786743164,
+      "learning_rate": 0.0001262099644128114,
+      "loss": 2.0563,
+      "step": 4157
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 1.0910627841949463,
+      "learning_rate": 0.00012619217081850535,
+      "loss": 1.2284,
+      "step": 4158
+    },
+    {
+      "epoch": 1.8484444444444446,
+      "grad_norm": 1.4845707416534424,
+      "learning_rate": 0.00012617437722419928,
+      "loss": 1.7377,
+      "step": 4159
+    },
+    {
+      "epoch": 1.8488888888888888,
+      "grad_norm": 1.4500175714492798,
+      "learning_rate": 0.00012615658362989324,
+      "loss": 1.5395,
+      "step": 4160
+    },
+    {
+      "epoch": 1.8493333333333335,
+      "grad_norm": 1.6412715911865234,
+      "learning_rate": 0.0001261387900355872,
+      "loss": 1.5817,
+      "step": 4161
+    },
+    {
+      "epoch": 1.8497777777777777,
+      "grad_norm": 1.6677457094192505,
+      "learning_rate": 0.00012612099644128115,
+      "loss": 2.4675,
+      "step": 4162
+    },
+    {
+      "epoch": 1.8502222222222222,
+      "grad_norm": 1.9067307710647583,
+      "learning_rate": 0.0001261032028469751,
+      "loss": 2.383,
+      "step": 4163
+    },
+    {
+      "epoch": 1.8506666666666667,
+      "grad_norm": 1.6068315505981445,
+      "learning_rate": 0.00012608540925266906,
+      "loss": 1.9027,
+      "step": 4164
+    },
+    {
+      "epoch": 1.8511111111111112,
+      "grad_norm": 1.8001701831817627,
+      "learning_rate": 0.000126067615658363,
+      "loss": 2.42,
+      "step": 4165
+    },
+    {
+      "epoch": 1.8515555555555556,
+      "grad_norm": 2.3257319927215576,
+      "learning_rate": 0.00012604982206405695,
+      "loss": 2.0886,
+      "step": 4166
+    },
+    {
+      "epoch": 1.8519999999999999,
+      "grad_norm": 2.088073492050171,
+      "learning_rate": 0.00012603202846975088,
+      "loss": 2.7623,
+      "step": 4167
+    },
+    {
+      "epoch": 1.8524444444444446,
+      "grad_norm": 1.0101107358932495,
+      "learning_rate": 0.00012601423487544483,
+      "loss": 0.7091,
+      "step": 4168
+    },
+    {
+      "epoch": 1.8528888888888888,
+      "grad_norm": 1.636488676071167,
+      "learning_rate": 0.0001259964412811388,
+      "loss": 2.0673,
+      "step": 4169
+    },
+    {
+      "epoch": 1.8533333333333335,
+      "grad_norm": 1.791585922241211,
+      "learning_rate": 0.00012597864768683275,
+      "loss": 2.1496,
+      "step": 4170
+    },
+    {
+      "epoch": 1.8537777777777777,
+      "grad_norm": 1.8314164876937866,
+      "learning_rate": 0.0001259608540925267,
+      "loss": 2.2932,
+      "step": 4171
+    },
+    {
+      "epoch": 1.8542222222222222,
+      "grad_norm": 1.7913572788238525,
+      "learning_rate": 0.00012594306049822063,
+      "loss": 1.8432,
+      "step": 4172
+    },
+    {
+      "epoch": 1.8546666666666667,
+      "grad_norm": 1.7947697639465332,
+      "learning_rate": 0.0001259252669039146,
+      "loss": 1.7726,
+      "step": 4173
+    },
+    {
+      "epoch": 1.8551111111111112,
+      "grad_norm": 1.8670748472213745,
+      "learning_rate": 0.00012590747330960855,
+      "loss": 2.0514,
+      "step": 4174
+    },
+    {
+      "epoch": 1.8555555555555556,
+      "grad_norm": 1.9150646924972534,
+      "learning_rate": 0.0001258896797153025,
+      "loss": 2.2755,
+      "step": 4175
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 1.9982563257217407,
+      "learning_rate": 0.00012587188612099646,
+      "loss": 2.4126,
+      "step": 4176
+    },
+    {
+      "epoch": 1.8564444444444446,
+      "grad_norm": 1.914477825164795,
+      "learning_rate": 0.00012585409252669042,
+      "loss": 1.7552,
+      "step": 4177
+    },
+    {
+      "epoch": 1.8568888888888888,
+      "grad_norm": 2.1474571228027344,
+      "learning_rate": 0.00012583629893238435,
+      "loss": 2.5033,
+      "step": 4178
+    },
+    {
+      "epoch": 1.8573333333333333,
+      "grad_norm": 1.845629334449768,
+      "learning_rate": 0.0001258185053380783,
+      "loss": 2.109,
+      "step": 4179
+    },
+    {
+      "epoch": 1.8577777777777778,
+      "grad_norm": 2.0772738456726074,
+      "learning_rate": 0.00012580071174377223,
+      "loss": 1.5434,
+      "step": 4180
+    },
+    {
+      "epoch": 1.8582222222222222,
+      "grad_norm": 1.9606627225875854,
+      "learning_rate": 0.0001257829181494662,
+      "loss": 2.1156,
+      "step": 4181
+    },
+    {
+      "epoch": 1.8586666666666667,
+      "grad_norm": 2.0049219131469727,
+      "learning_rate": 0.00012576512455516014,
+      "loss": 1.9324,
+      "step": 4182
+    },
+    {
+      "epoch": 1.8591111111111112,
+      "grad_norm": 1.535262107849121,
+      "learning_rate": 0.0001257473309608541,
+      "loss": 1.2178,
+      "step": 4183
+    },
+    {
+      "epoch": 1.8595555555555556,
+      "grad_norm": 2.2102348804473877,
+      "learning_rate": 0.00012572953736654806,
+      "loss": 2.8349,
+      "step": 4184
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 1.7623968124389648,
+      "learning_rate": 0.000125711743772242,
+      "loss": 1.5631,
+      "step": 4185
+    },
+    {
+      "epoch": 1.8604444444444446,
+      "grad_norm": 2.135024070739746,
+      "learning_rate": 0.00012569395017793594,
+      "loss": 2.233,
+      "step": 4186
+    },
+    {
+      "epoch": 1.8608888888888888,
+      "grad_norm": 2.3472445011138916,
+      "learning_rate": 0.0001256761565836299,
+      "loss": 2.658,
+      "step": 4187
+    },
+    {
+      "epoch": 1.8613333333333333,
+      "grad_norm": 2.1595356464385986,
+      "learning_rate": 0.00012565836298932386,
+      "loss": 2.0647,
+      "step": 4188
+    },
+    {
+      "epoch": 1.8617777777777778,
+      "grad_norm": 2.6070942878723145,
+      "learning_rate": 0.0001256405693950178,
+      "loss": 2.0941,
+      "step": 4189
+    },
+    {
+      "epoch": 1.8622222222222222,
+      "grad_norm": 2.0253846645355225,
+      "learning_rate": 0.00012562277580071177,
+      "loss": 1.9656,
+      "step": 4190
+    },
+    {
+      "epoch": 1.8626666666666667,
+      "grad_norm": 2.2132487297058105,
+      "learning_rate": 0.0001256049822064057,
+      "loss": 2.3254,
+      "step": 4191
+    },
+    {
+      "epoch": 1.8631111111111112,
+      "grad_norm": 2.6254289150238037,
+      "learning_rate": 0.00012558718861209966,
+      "loss": 3.0945,
+      "step": 4192
+    },
+    {
+      "epoch": 1.8635555555555556,
+      "grad_norm": 2.6178314685821533,
+      "learning_rate": 0.00012556939501779358,
+      "loss": 2.216,
+      "step": 4193
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 2.047734498977661,
+      "learning_rate": 0.00012555160142348754,
+      "loss": 1.6887,
+      "step": 4194
+    },
+    {
+      "epoch": 1.8644444444444446,
+      "grad_norm": 3.436713695526123,
+      "learning_rate": 0.0001255338078291815,
+      "loss": 1.5871,
+      "step": 4195
+    },
+    {
+      "epoch": 1.8648888888888888,
+      "grad_norm": 2.280249834060669,
+      "learning_rate": 0.00012551601423487545,
+      "loss": 2.2639,
+      "step": 4196
+    },
+    {
+      "epoch": 1.8653333333333333,
+      "grad_norm": 2.708115816116333,
+      "learning_rate": 0.0001254982206405694,
+      "loss": 2.4468,
+      "step": 4197
+    },
+    {
+      "epoch": 1.8657777777777778,
+      "grad_norm": 2.6053898334503174,
+      "learning_rate": 0.00012548042704626334,
+      "loss": 2.3777,
+      "step": 4198
+    },
+    {
+      "epoch": 1.8662222222222222,
+      "grad_norm": 1.6615748405456543,
+      "learning_rate": 0.0001254626334519573,
+      "loss": 0.9062,
+      "step": 4199
+    },
+    {
+      "epoch": 1.8666666666666667,
+      "grad_norm": 2.4906649589538574,
+      "learning_rate": 0.00012544483985765125,
+      "loss": 0.8929,
+      "step": 4200
+    },
+    {
+      "epoch": 1.8671111111111112,
+      "grad_norm": 1.4890950918197632,
+      "learning_rate": 0.0001254270462633452,
+      "loss": 2.7543,
+      "step": 4201
+    },
+    {
+      "epoch": 1.8675555555555556,
+      "grad_norm": 1.5864791870117188,
+      "learning_rate": 0.00012540925266903917,
+      "loss": 2.6892,
+      "step": 4202
+    },
+    {
+      "epoch": 1.8679999999999999,
+      "grad_norm": 1.612914800643921,
+      "learning_rate": 0.0001253914590747331,
+      "loss": 1.7013,
+      "step": 4203
+    },
+    {
+      "epoch": 1.8684444444444446,
+      "grad_norm": 1.4477015733718872,
+      "learning_rate": 0.00012537366548042705,
+      "loss": 1.879,
+      "step": 4204
+    },
+    {
+      "epoch": 1.8688888888888888,
+      "grad_norm": 1.9278892278671265,
+      "learning_rate": 0.000125355871886121,
+      "loss": 2.1627,
+      "step": 4205
+    },
+    {
+      "epoch": 1.8693333333333333,
+      "grad_norm": 1.7229881286621094,
+      "learning_rate": 0.00012533807829181494,
+      "loss": 2.3671,
+      "step": 4206
+    },
+    {
+      "epoch": 1.8697777777777778,
+      "grad_norm": 1.8617095947265625,
+      "learning_rate": 0.0001253202846975089,
+      "loss": 2.6608,
+      "step": 4207
+    },
+    {
+      "epoch": 1.8702222222222222,
+      "grad_norm": 1.8363666534423828,
+      "learning_rate": 0.00012530249110320285,
+      "loss": 2.059,
+      "step": 4208
+    },
+    {
+      "epoch": 1.8706666666666667,
+      "grad_norm": 1.920870304107666,
+      "learning_rate": 0.0001252846975088968,
+      "loss": 1.7415,
+      "step": 4209
+    },
+    {
+      "epoch": 1.871111111111111,
+      "grad_norm": 1.610530138015747,
+      "learning_rate": 0.00012526690391459074,
+      "loss": 2.2202,
+      "step": 4210
+    },
+    {
+      "epoch": 1.8715555555555556,
+      "grad_norm": 1.8126658201217651,
+      "learning_rate": 0.0001252491103202847,
+      "loss": 2.2095,
+      "step": 4211
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 2.0825304985046387,
+      "learning_rate": 0.00012523131672597865,
+      "loss": 2.3952,
+      "step": 4212
+    },
+    {
+      "epoch": 1.8724444444444446,
+      "grad_norm": 1.455892562866211,
+      "learning_rate": 0.0001252135231316726,
+      "loss": 1.4388,
+      "step": 4213
+    },
+    {
+      "epoch": 1.8728888888888888,
+      "grad_norm": 1.9848651885986328,
+      "learning_rate": 0.00012519572953736656,
+      "loss": 2.325,
+      "step": 4214
+    },
+    {
+      "epoch": 1.8733333333333333,
+      "grad_norm": 1.972327709197998,
+      "learning_rate": 0.00012517793594306052,
+      "loss": 1.9154,
+      "step": 4215
+    },
+    {
+      "epoch": 1.8737777777777778,
+      "grad_norm": 1.8076159954071045,
+      "learning_rate": 0.00012516014234875445,
+      "loss": 2.4675,
+      "step": 4216
+    },
+    {
+      "epoch": 1.8742222222222222,
+      "grad_norm": 1.6413958072662354,
+      "learning_rate": 0.0001251423487544484,
+      "loss": 1.5596,
+      "step": 4217
+    },
+    {
+      "epoch": 1.8746666666666667,
+      "grad_norm": 1.9646254777908325,
+      "learning_rate": 0.00012512455516014236,
+      "loss": 2.1716,
+      "step": 4218
+    },
+    {
+      "epoch": 1.875111111111111,
+      "grad_norm": 1.7197006940841675,
+      "learning_rate": 0.0001251067615658363,
+      "loss": 2.0816,
+      "step": 4219
+    },
+    {
+      "epoch": 1.8755555555555556,
+      "grad_norm": 1.8920749425888062,
+      "learning_rate": 0.00012508896797153025,
+      "loss": 2.2979,
+      "step": 4220
+    },
+    {
+      "epoch": 1.876,
+      "grad_norm": 1.8623499870300293,
+      "learning_rate": 0.0001250711743772242,
+      "loss": 2.5679,
+      "step": 4221
+    },
+    {
+      "epoch": 1.8764444444444446,
+      "grad_norm": 2.2921416759490967,
+      "learning_rate": 0.00012505338078291816,
+      "loss": 2.4083,
+      "step": 4222
+    },
+    {
+      "epoch": 1.8768888888888888,
+      "grad_norm": 1.984156847000122,
+      "learning_rate": 0.0001250355871886121,
+      "loss": 2.6118,
+      "step": 4223
+    },
+    {
+      "epoch": 1.8773333333333333,
+      "grad_norm": 1.681645154953003,
+      "learning_rate": 0.00012501779359430605,
+      "loss": 1.4962,
+      "step": 4224
+    },
+    {
+      "epoch": 1.8777777777777778,
+      "grad_norm": 2.018657684326172,
+      "learning_rate": 0.000125,
+      "loss": 2.2015,
+      "step": 4225
+    },
+    {
+      "epoch": 1.8782222222222222,
+      "grad_norm": 1.8145815134048462,
+      "learning_rate": 0.00012498220640569396,
+      "loss": 1.4773,
+      "step": 4226
+    },
+    {
+      "epoch": 1.8786666666666667,
+      "grad_norm": 1.8681118488311768,
+      "learning_rate": 0.00012496441281138792,
+      "loss": 2.0218,
+      "step": 4227
+    },
+    {
+      "epoch": 1.879111111111111,
+      "grad_norm": 2.0949301719665527,
+      "learning_rate": 0.00012494661921708187,
+      "loss": 2.1287,
+      "step": 4228
+    },
+    {
+      "epoch": 1.8795555555555556,
+      "grad_norm": 1.3150235414505005,
+      "learning_rate": 0.0001249288256227758,
+      "loss": 1.1873,
+      "step": 4229
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.1858744621276855,
+      "learning_rate": 0.00012491103202846976,
+      "loss": 2.2247,
+      "step": 4230
+    },
+    {
+      "epoch": 1.8804444444444446,
+      "grad_norm": 2.0682497024536133,
+      "learning_rate": 0.00012489323843416371,
+      "loss": 2.2693,
+      "step": 4231
+    },
+    {
+      "epoch": 1.8808888888888888,
+      "grad_norm": 2.008765697479248,
+      "learning_rate": 0.00012487544483985764,
+      "loss": 1.8305,
+      "step": 4232
+    },
+    {
+      "epoch": 1.8813333333333333,
+      "grad_norm": 2.322118043899536,
+      "learning_rate": 0.0001248576512455516,
+      "loss": 2.6863,
+      "step": 4233
+    },
+    {
+      "epoch": 1.8817777777777778,
+      "grad_norm": 2.1211628913879395,
+      "learning_rate": 0.00012483985765124556,
+      "loss": 2.3278,
+      "step": 4234
+    },
+    {
+      "epoch": 1.8822222222222222,
+      "grad_norm": 2.126739501953125,
+      "learning_rate": 0.0001248220640569395,
+      "loss": 1.9615,
+      "step": 4235
+    },
+    {
+      "epoch": 1.8826666666666667,
+      "grad_norm": 1.4139108657836914,
+      "learning_rate": 0.00012480427046263344,
+      "loss": 1.3694,
+      "step": 4236
+    },
+    {
+      "epoch": 1.883111111111111,
+      "grad_norm": 2.1995489597320557,
+      "learning_rate": 0.0001247864768683274,
+      "loss": 2.1418,
+      "step": 4237
+    },
+    {
+      "epoch": 1.8835555555555556,
+      "grad_norm": 2.4655182361602783,
+      "learning_rate": 0.00012476868327402136,
+      "loss": 2.497,
+      "step": 4238
+    },
+    {
+      "epoch": 1.884,
+      "grad_norm": 2.2003283500671387,
+      "learning_rate": 0.0001247508896797153,
+      "loss": 2.6386,
+      "step": 4239
+    },
+    {
+      "epoch": 1.8844444444444446,
+      "grad_norm": 2.3272600173950195,
+      "learning_rate": 0.00012473309608540927,
+      "loss": 2.2253,
+      "step": 4240
+    },
+    {
+      "epoch": 1.8848888888888888,
+      "grad_norm": 1.7986118793487549,
+      "learning_rate": 0.00012471530249110323,
+      "loss": 2.112,
+      "step": 4241
+    },
+    {
+      "epoch": 1.8853333333333333,
+      "grad_norm": 2.1452407836914062,
+      "learning_rate": 0.00012469750889679715,
+      "loss": 1.9534,
+      "step": 4242
+    },
+    {
+      "epoch": 1.8857777777777778,
+      "grad_norm": 2.466782808303833,
+      "learning_rate": 0.0001246797153024911,
+      "loss": 2.5373,
+      "step": 4243
+    },
+    {
+      "epoch": 1.8862222222222222,
+      "grad_norm": 2.473870277404785,
+      "learning_rate": 0.00012466192170818507,
+      "loss": 2.505,
+      "step": 4244
+    },
+    {
+      "epoch": 1.8866666666666667,
+      "grad_norm": 2.5043163299560547,
+      "learning_rate": 0.000124644128113879,
+      "loss": 2.4174,
+      "step": 4245
+    },
+    {
+      "epoch": 1.887111111111111,
+      "grad_norm": 2.2881720066070557,
+      "learning_rate": 0.00012462633451957295,
+      "loss": 2.4042,
+      "step": 4246
+    },
+    {
+      "epoch": 1.8875555555555557,
+      "grad_norm": 2.357645034790039,
+      "learning_rate": 0.0001246085409252669,
+      "loss": 2.1379,
+      "step": 4247
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 2.677143096923828,
+      "learning_rate": 0.00012459074733096087,
+      "loss": 2.0372,
+      "step": 4248
+    },
+    {
+      "epoch": 1.8884444444444446,
+      "grad_norm": 2.4684627056121826,
+      "learning_rate": 0.0001245729537366548,
+      "loss": 1.7057,
+      "step": 4249
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 2.5674924850463867,
+      "learning_rate": 0.00012455516014234875,
+      "loss": 1.4432,
+      "step": 4250
+    },
+    {
+      "epoch": 1.8893333333333333,
+      "grad_norm": 1.4329222440719604,
+      "learning_rate": 0.0001245373665480427,
+      "loss": 2.3156,
+      "step": 4251
+    },
+    {
+      "epoch": 1.8897777777777778,
+      "grad_norm": 1.45597505569458,
+      "learning_rate": 0.00012451957295373667,
+      "loss": 2.4051,
+      "step": 4252
+    },
+    {
+      "epoch": 1.8902222222222222,
+      "grad_norm": 1.2270104885101318,
+      "learning_rate": 0.00012450177935943062,
+      "loss": 1.7157,
+      "step": 4253
+    },
+    {
+      "epoch": 1.8906666666666667,
+      "grad_norm": 1.1220571994781494,
+      "learning_rate": 0.00012448398576512458,
+      "loss": 1.2594,
+      "step": 4254
+    },
+    {
+      "epoch": 1.891111111111111,
+      "grad_norm": 1.5439119338989258,
+      "learning_rate": 0.0001244661921708185,
+      "loss": 2.2097,
+      "step": 4255
+    },
+    {
+      "epoch": 1.8915555555555557,
+      "grad_norm": 1.6956852674484253,
+      "learning_rate": 0.00012444839857651246,
+      "loss": 2.6823,
+      "step": 4256
+    },
+    {
+      "epoch": 1.892,
+      "grad_norm": 1.640121579170227,
+      "learning_rate": 0.00012443060498220642,
+      "loss": 2.2303,
+      "step": 4257
+    },
+    {
+      "epoch": 1.8924444444444446,
+      "grad_norm": 1.3527717590332031,
+      "learning_rate": 0.00012441281138790035,
+      "loss": 1.1705,
+      "step": 4258
+    },
+    {
+      "epoch": 1.8928888888888888,
+      "grad_norm": 1.7290159463882446,
+      "learning_rate": 0.0001243950177935943,
+      "loss": 2.4764,
+      "step": 4259
+    },
+    {
+      "epoch": 1.8933333333333333,
+      "grad_norm": 1.5144662857055664,
+      "learning_rate": 0.00012437722419928826,
+      "loss": 1.8495,
+      "step": 4260
+    },
+    {
+      "epoch": 1.8937777777777778,
+      "grad_norm": 1.8776041269302368,
+      "learning_rate": 0.00012435943060498222,
+      "loss": 2.7901,
+      "step": 4261
+    },
+    {
+      "epoch": 1.8942222222222223,
+      "grad_norm": 1.9808104038238525,
+      "learning_rate": 0.00012434163701067615,
+      "loss": 2.3308,
+      "step": 4262
+    },
+    {
+      "epoch": 1.8946666666666667,
+      "grad_norm": 1.6954816579818726,
+      "learning_rate": 0.0001243238434163701,
+      "loss": 2.1066,
+      "step": 4263
+    },
+    {
+      "epoch": 1.895111111111111,
+      "grad_norm": 2.016079902648926,
+      "learning_rate": 0.00012430604982206406,
+      "loss": 2.4023,
+      "step": 4264
+    },
+    {
+      "epoch": 1.8955555555555557,
+      "grad_norm": 1.6305691003799438,
+      "learning_rate": 0.00012428825622775802,
+      "loss": 1.8707,
+      "step": 4265
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 1.8553330898284912,
+      "learning_rate": 0.00012427046263345198,
+      "loss": 2.0848,
+      "step": 4266
+    },
+    {
+      "epoch": 1.8964444444444446,
+      "grad_norm": 1.7080594301223755,
+      "learning_rate": 0.00012425266903914593,
+      "loss": 2.0636,
+      "step": 4267
+    },
+    {
+      "epoch": 1.8968888888888888,
+      "grad_norm": 1.9433776140213013,
+      "learning_rate": 0.00012423487544483986,
+      "loss": 2.1638,
+      "step": 4268
+    },
+    {
+      "epoch": 1.8973333333333333,
+      "grad_norm": 1.7864351272583008,
+      "learning_rate": 0.00012421708185053382,
+      "loss": 2.1109,
+      "step": 4269
+    },
+    {
+      "epoch": 1.8977777777777778,
+      "grad_norm": 1.667856216430664,
+      "learning_rate": 0.00012419928825622777,
+      "loss": 1.4533,
+      "step": 4270
+    },
+    {
+      "epoch": 1.8982222222222223,
+      "grad_norm": 1.8229117393493652,
+      "learning_rate": 0.0001241814946619217,
+      "loss": 2.093,
+      "step": 4271
+    },
+    {
+      "epoch": 1.8986666666666667,
+      "grad_norm": 2.1975040435791016,
+      "learning_rate": 0.00012416370106761566,
+      "loss": 2.6783,
+      "step": 4272
+    },
+    {
+      "epoch": 1.899111111111111,
+      "grad_norm": 1.8947041034698486,
+      "learning_rate": 0.00012414590747330962,
+      "loss": 2.4428,
+      "step": 4273
+    },
+    {
+      "epoch": 1.8995555555555557,
+      "grad_norm": 1.981955647468567,
+      "learning_rate": 0.00012412811387900357,
+      "loss": 1.9542,
+      "step": 4274
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.057514190673828,
+      "learning_rate": 0.0001241103202846975,
+      "loss": 2.4705,
+      "step": 4275
+    },
+    {
+      "epoch": 1.9004444444444446,
+      "grad_norm": 1.9923932552337646,
+      "learning_rate": 0.00012409252669039146,
+      "loss": 2.1799,
+      "step": 4276
+    },
+    {
+      "epoch": 1.9008888888888889,
+      "grad_norm": 1.9740736484527588,
+      "learning_rate": 0.00012407473309608542,
+      "loss": 2.401,
+      "step": 4277
+    },
+    {
+      "epoch": 1.9013333333333333,
+      "grad_norm": 0.5957589149475098,
+      "learning_rate": 0.00012405693950177937,
+      "loss": 0.0373,
+      "step": 4278
+    },
+    {
+      "epoch": 1.9017777777777778,
+      "grad_norm": 1.835414171218872,
+      "learning_rate": 0.00012403914590747333,
+      "loss": 2.2666,
+      "step": 4279
+    },
+    {
+      "epoch": 1.9022222222222223,
+      "grad_norm": 1.9342777729034424,
+      "learning_rate": 0.00012402135231316728,
+      "loss": 2.1608,
+      "step": 4280
+    },
+    {
+      "epoch": 1.9026666666666667,
+      "grad_norm": 1.760406732559204,
+      "learning_rate": 0.00012400355871886121,
+      "loss": 1.7764,
+      "step": 4281
+    },
+    {
+      "epoch": 1.903111111111111,
+      "grad_norm": 2.0876519680023193,
+      "learning_rate": 0.00012398576512455517,
+      "loss": 2.6303,
+      "step": 4282
+    },
+    {
+      "epoch": 1.9035555555555557,
+      "grad_norm": 2.0983736515045166,
+      "learning_rate": 0.00012396797153024913,
+      "loss": 2.28,
+      "step": 4283
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 2.1319432258605957,
+      "learning_rate": 0.00012395017793594306,
+      "loss": 2.5054,
+      "step": 4284
+    },
+    {
+      "epoch": 1.9044444444444446,
+      "grad_norm": 1.9509506225585938,
+      "learning_rate": 0.000123932384341637,
+      "loss": 1.8634,
+      "step": 4285
+    },
+    {
+      "epoch": 1.9048888888888889,
+      "grad_norm": 2.6779415607452393,
+      "learning_rate": 0.00012391459074733097,
+      "loss": 2.8077,
+      "step": 4286
+    },
+    {
+      "epoch": 1.9053333333333333,
+      "grad_norm": 1.5328880548477173,
+      "learning_rate": 0.00012389679715302493,
+      "loss": 1.1721,
+      "step": 4287
+    },
+    {
+      "epoch": 1.9057777777777778,
+      "grad_norm": 2.5803184509277344,
+      "learning_rate": 0.00012387900355871886,
+      "loss": 2.3973,
+      "step": 4288
+    },
+    {
+      "epoch": 1.9062222222222223,
+      "grad_norm": 1.985290765762329,
+      "learning_rate": 0.0001238612099644128,
+      "loss": 2.2365,
+      "step": 4289
+    },
+    {
+      "epoch": 1.9066666666666667,
+      "grad_norm": 2.207040309906006,
+      "learning_rate": 0.00012384341637010677,
+      "loss": 1.5546,
+      "step": 4290
+    },
+    {
+      "epoch": 1.907111111111111,
+      "grad_norm": 2.6116459369659424,
+      "learning_rate": 0.00012382562277580072,
+      "loss": 2.4088,
+      "step": 4291
+    },
+    {
+      "epoch": 1.9075555555555557,
+      "grad_norm": 1.6348541975021362,
+      "learning_rate": 0.00012380782918149468,
+      "loss": 1.5725,
+      "step": 4292
+    },
+    {
+      "epoch": 1.908,
+      "grad_norm": 2.5747082233428955,
+      "learning_rate": 0.0001237900355871886,
+      "loss": 2.5069,
+      "step": 4293
+    },
+    {
+      "epoch": 1.9084444444444446,
+      "grad_norm": 2.236910104751587,
+      "learning_rate": 0.00012377224199288257,
+      "loss": 2.2303,
+      "step": 4294
+    },
+    {
+      "epoch": 1.9088888888888889,
+      "grad_norm": 2.4665608406066895,
+      "learning_rate": 0.00012375444839857652,
+      "loss": 2.0439,
+      "step": 4295
+    },
+    {
+      "epoch": 1.9093333333333333,
+      "grad_norm": 2.041987895965576,
+      "learning_rate": 0.00012373665480427045,
+      "loss": 2.1251,
+      "step": 4296
+    },
+    {
+      "epoch": 1.9097777777777778,
+      "grad_norm": 2.454289436340332,
+      "learning_rate": 0.0001237188612099644,
+      "loss": 2.5853,
+      "step": 4297
+    },
+    {
+      "epoch": 1.9102222222222223,
+      "grad_norm": 2.1545984745025635,
+      "learning_rate": 0.00012370106761565837,
+      "loss": 2.0791,
+      "step": 4298
+    },
+    {
+      "epoch": 1.9106666666666667,
+      "grad_norm": 2.7019107341766357,
+      "learning_rate": 0.00012368327402135232,
+      "loss": 2.4134,
+      "step": 4299
+    },
+    {
+      "epoch": 1.911111111111111,
+      "grad_norm": 3.2024686336517334,
+      "learning_rate": 0.00012366548042704625,
+      "loss": 1.6449,
+      "step": 4300
+    },
+    {
+      "epoch": 1.9115555555555557,
+      "grad_norm": 1.3786087036132812,
+      "learning_rate": 0.0001236476868327402,
+      "loss": 2.5061,
+      "step": 4301
+    },
+    {
+      "epoch": 1.912,
+      "grad_norm": 1.5719784498214722,
+      "learning_rate": 0.00012362989323843417,
+      "loss": 2.3814,
+      "step": 4302
+    },
+    {
+      "epoch": 1.9124444444444444,
+      "grad_norm": 1.4920408725738525,
+      "learning_rate": 0.00012361209964412812,
+      "loss": 2.0447,
+      "step": 4303
+    },
+    {
+      "epoch": 1.9128888888888889,
+      "grad_norm": 1.747694730758667,
+      "learning_rate": 0.00012359430604982208,
+      "loss": 2.8776,
+      "step": 4304
+    },
+    {
+      "epoch": 1.9133333333333333,
+      "grad_norm": 1.8610203266143799,
+      "learning_rate": 0.00012357651245551603,
+      "loss": 2.3367,
+      "step": 4305
+    },
+    {
+      "epoch": 1.9137777777777778,
+      "grad_norm": 1.5181238651275635,
+      "learning_rate": 0.00012355871886120996,
+      "loss": 1.9326,
+      "step": 4306
+    },
+    {
+      "epoch": 1.9142222222222223,
+      "grad_norm": 1.707848072052002,
+      "learning_rate": 0.00012354092526690392,
+      "loss": 2.3507,
+      "step": 4307
+    },
+    {
+      "epoch": 1.9146666666666667,
+      "grad_norm": 1.7067019939422607,
+      "learning_rate": 0.00012352313167259788,
+      "loss": 2.0366,
+      "step": 4308
+    },
+    {
+      "epoch": 1.915111111111111,
+      "grad_norm": 1.7387741804122925,
+      "learning_rate": 0.0001235053380782918,
+      "loss": 2.0162,
+      "step": 4309
+    },
+    {
+      "epoch": 1.9155555555555557,
+      "grad_norm": 1.7039339542388916,
+      "learning_rate": 0.00012348754448398576,
+      "loss": 2.1688,
+      "step": 4310
+    },
+    {
+      "epoch": 1.916,
+      "grad_norm": 1.6669498682022095,
+      "learning_rate": 0.00012346975088967972,
+      "loss": 2.2707,
+      "step": 4311
+    },
+    {
+      "epoch": 1.9164444444444444,
+      "grad_norm": 1.6338768005371094,
+      "learning_rate": 0.00012345195729537368,
+      "loss": 2.0841,
+      "step": 4312
+    },
+    {
+      "epoch": 1.9168888888888889,
+      "grad_norm": 1.6878036260604858,
+      "learning_rate": 0.0001234341637010676,
+      "loss": 2.2946,
+      "step": 4313
+    },
+    {
+      "epoch": 1.9173333333333333,
+      "grad_norm": 1.7214369773864746,
+      "learning_rate": 0.00012341637010676156,
+      "loss": 1.9289,
+      "step": 4314
+    },
+    {
+      "epoch": 1.9177777777777778,
+      "grad_norm": 1.6168575286865234,
+      "learning_rate": 0.00012339857651245552,
+      "loss": 1.9991,
+      "step": 4315
+    },
+    {
+      "epoch": 1.9182222222222223,
+      "grad_norm": 1.788225531578064,
+      "learning_rate": 0.00012338078291814947,
+      "loss": 2.3562,
+      "step": 4316
+    },
+    {
+      "epoch": 1.9186666666666667,
+      "grad_norm": 1.5336823463439941,
+      "learning_rate": 0.00012336298932384343,
+      "loss": 1.7089,
+      "step": 4317
+    },
+    {
+      "epoch": 1.919111111111111,
+      "grad_norm": 1.8361992835998535,
+      "learning_rate": 0.0001233451957295374,
+      "loss": 2.0708,
+      "step": 4318
+    },
+    {
+      "epoch": 1.9195555555555557,
+      "grad_norm": 1.857429027557373,
+      "learning_rate": 0.00012332740213523132,
+      "loss": 1.7238,
+      "step": 4319
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6928489208221436,
+      "learning_rate": 0.00012330960854092527,
+      "loss": 2.0781,
+      "step": 4320
+    },
+    {
+      "epoch": 1.9204444444444444,
+      "grad_norm": 1.6085439920425415,
+      "learning_rate": 0.00012329181494661923,
+      "loss": 1.37,
+      "step": 4321
+    },
+    {
+      "epoch": 1.9208888888888889,
+      "grad_norm": 1.9759423732757568,
+      "learning_rate": 0.00012327402135231316,
+      "loss": 2.561,
+      "step": 4322
+    },
+    {
+      "epoch": 1.9213333333333333,
+      "grad_norm": 1.8976612091064453,
+      "learning_rate": 0.00012325622775800712,
+      "loss": 2.1077,
+      "step": 4323
+    },
+    {
+      "epoch": 1.9217777777777778,
+      "grad_norm": 2.0411934852600098,
+      "learning_rate": 0.00012323843416370107,
+      "loss": 2.4446,
+      "step": 4324
+    },
+    {
+      "epoch": 1.9222222222222223,
+      "grad_norm": 2.429276704788208,
+      "learning_rate": 0.00012322064056939503,
+      "loss": 2.5635,
+      "step": 4325
+    },
+    {
+      "epoch": 1.9226666666666667,
+      "grad_norm": 2.421410322189331,
+      "learning_rate": 0.00012320284697508896,
+      "loss": 2.9828,
+      "step": 4326
+    },
+    {
+      "epoch": 1.923111111111111,
+      "grad_norm": 2.055605173110962,
+      "learning_rate": 0.00012318505338078291,
+      "loss": 2.5132,
+      "step": 4327
+    },
+    {
+      "epoch": 1.9235555555555557,
+      "grad_norm": 1.587828278541565,
+      "learning_rate": 0.00012316725978647687,
+      "loss": 1.3512,
+      "step": 4328
+    },
+    {
+      "epoch": 1.924,
+      "grad_norm": 2.128390312194824,
+      "learning_rate": 0.00012314946619217083,
+      "loss": 2.1455,
+      "step": 4329
+    },
+    {
+      "epoch": 1.9244444444444444,
+      "grad_norm": 1.7967475652694702,
+      "learning_rate": 0.00012313167259786478,
+      "loss": 0.8184,
+      "step": 4330
+    },
+    {
+      "epoch": 1.9248888888888889,
+      "grad_norm": 2.1603779792785645,
+      "learning_rate": 0.00012311387900355874,
+      "loss": 2.2261,
+      "step": 4331
+    },
+    {
+      "epoch": 1.9253333333333333,
+      "grad_norm": 2.0636863708496094,
+      "learning_rate": 0.00012309608540925267,
+      "loss": 2.0121,
+      "step": 4332
+    },
+    {
+      "epoch": 1.9257777777777778,
+      "grad_norm": 2.0725672245025635,
+      "learning_rate": 0.00012307829181494663,
+      "loss": 2.276,
+      "step": 4333
+    },
+    {
+      "epoch": 1.926222222222222,
+      "grad_norm": 2.1373484134674072,
+      "learning_rate": 0.00012306049822064058,
+      "loss": 2.3134,
+      "step": 4334
+    },
+    {
+      "epoch": 1.9266666666666667,
+      "grad_norm": 1.7675268650054932,
+      "learning_rate": 0.0001230427046263345,
+      "loss": 1.5304,
+      "step": 4335
+    },
+    {
+      "epoch": 1.927111111111111,
+      "grad_norm": 1.9493167400360107,
+      "learning_rate": 0.00012302491103202847,
+      "loss": 1.8928,
+      "step": 4336
+    },
+    {
+      "epoch": 1.9275555555555557,
+      "grad_norm": 2.15889835357666,
+      "learning_rate": 0.00012300711743772243,
+      "loss": 2.0092,
+      "step": 4337
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 2.0056865215301514,
+      "learning_rate": 0.00012298932384341638,
+      "loss": 2.3798,
+      "step": 4338
+    },
+    {
+      "epoch": 1.9284444444444444,
+      "grad_norm": 2.266223907470703,
+      "learning_rate": 0.0001229715302491103,
+      "loss": 2.3884,
+      "step": 4339
+    },
+    {
+      "epoch": 1.9288888888888889,
+      "grad_norm": 2.173867702484131,
+      "learning_rate": 0.00012295373665480427,
+      "loss": 2.5138,
+      "step": 4340
+    },
+    {
+      "epoch": 1.9293333333333333,
+      "grad_norm": 2.383648633956909,
+      "learning_rate": 0.00012293594306049822,
+      "loss": 2.4544,
+      "step": 4341
+    },
+    {
+      "epoch": 1.9297777777777778,
+      "grad_norm": 1.9824879169464111,
+      "learning_rate": 0.00012291814946619218,
+      "loss": 1.7544,
+      "step": 4342
+    },
+    {
+      "epoch": 1.930222222222222,
+      "grad_norm": 2.24977707862854,
+      "learning_rate": 0.00012290035587188614,
+      "loss": 2.9336,
+      "step": 4343
+    },
+    {
+      "epoch": 1.9306666666666668,
+      "grad_norm": 2.6044070720672607,
+      "learning_rate": 0.0001228825622775801,
+      "loss": 2.0662,
+      "step": 4344
+    },
+    {
+      "epoch": 1.931111111111111,
+      "grad_norm": 2.6351845264434814,
+      "learning_rate": 0.00012286476868327402,
+      "loss": 2.554,
+      "step": 4345
+    },
+    {
+      "epoch": 1.9315555555555557,
+      "grad_norm": 2.308326005935669,
+      "learning_rate": 0.00012284697508896798,
+      "loss": 2.0536,
+      "step": 4346
+    },
+    {
+      "epoch": 1.932,
+      "grad_norm": 2.2656610012054443,
+      "learning_rate": 0.00012282918149466194,
+      "loss": 2.3636,
+      "step": 4347
+    },
+    {
+      "epoch": 1.9324444444444444,
+      "grad_norm": 2.7390570640563965,
+      "learning_rate": 0.00012281138790035587,
+      "loss": 2.2897,
+      "step": 4348
+    },
+    {
+      "epoch": 1.9328888888888889,
+      "grad_norm": 2.2677807807922363,
+      "learning_rate": 0.00012279359430604982,
+      "loss": 2.1377,
+      "step": 4349
+    },
+    {
+      "epoch": 1.9333333333333333,
+      "grad_norm": 3.099202871322632,
+      "learning_rate": 0.00012277580071174378,
+      "loss": 2.5962,
+      "step": 4350
+    },
+    {
+      "epoch": 1.9337777777777778,
+      "grad_norm": 0.9823408126831055,
+      "learning_rate": 0.00012275800711743774,
+      "loss": 1.15,
+      "step": 4351
+    },
+    {
+      "epoch": 1.934222222222222,
+      "grad_norm": 1.5499428510665894,
+      "learning_rate": 0.00012274021352313166,
+      "loss": 1.9964,
+      "step": 4352
+    },
+    {
+      "epoch": 1.9346666666666668,
+      "grad_norm": 1.5423551797866821,
+      "learning_rate": 0.00012272241992882562,
+      "loss": 2.1596,
+      "step": 4353
+    },
+    {
+      "epoch": 1.935111111111111,
+      "grad_norm": 1.57822847366333,
+      "learning_rate": 0.00012270462633451958,
+      "loss": 2.2386,
+      "step": 4354
+    },
+    {
+      "epoch": 1.9355555555555557,
+      "grad_norm": 1.5377904176712036,
+      "learning_rate": 0.00012268683274021353,
+      "loss": 2.0222,
+      "step": 4355
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 1.53287935256958,
+      "learning_rate": 0.0001226690391459075,
+      "loss": 1.9897,
+      "step": 4356
+    },
+    {
+      "epoch": 1.9364444444444444,
+      "grad_norm": 1.51986563205719,
+      "learning_rate": 0.00012265124555160145,
+      "loss": 2.0826,
+      "step": 4357
+    },
+    {
+      "epoch": 1.9368888888888889,
+      "grad_norm": 1.7066446542739868,
+      "learning_rate": 0.00012263345195729538,
+      "loss": 1.8589,
+      "step": 4358
+    },
+    {
+      "epoch": 1.9373333333333334,
+      "grad_norm": 1.175493836402893,
+      "learning_rate": 0.00012261565836298933,
+      "loss": 1.0915,
+      "step": 4359
+    },
+    {
+      "epoch": 1.9377777777777778,
+      "grad_norm": 1.7773221731185913,
+      "learning_rate": 0.0001225978647686833,
+      "loss": 1.8631,
+      "step": 4360
+    },
+    {
+      "epoch": 1.938222222222222,
+      "grad_norm": 2.198300361633301,
+      "learning_rate": 0.00012258007117437722,
+      "loss": 2.6825,
+      "step": 4361
+    },
+    {
+      "epoch": 1.9386666666666668,
+      "grad_norm": 1.8990510702133179,
+      "learning_rate": 0.00012256227758007118,
+      "loss": 2.8493,
+      "step": 4362
+    },
+    {
+      "epoch": 1.939111111111111,
+      "grad_norm": 1.874263882637024,
+      "learning_rate": 0.00012254448398576513,
+      "loss": 2.588,
+      "step": 4363
+    },
+    {
+      "epoch": 1.9395555555555557,
+      "grad_norm": 1.9859611988067627,
+      "learning_rate": 0.0001225266903914591,
+      "loss": 2.3982,
+      "step": 4364
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6862070560455322,
+      "learning_rate": 0.00012250889679715302,
+      "loss": 2.0444,
+      "step": 4365
+    },
+    {
+      "epoch": 1.9404444444444444,
+      "grad_norm": 1.7006193399429321,
+      "learning_rate": 0.00012249110320284697,
+      "loss": 1.6788,
+      "step": 4366
+    },
+    {
+      "epoch": 1.9408888888888889,
+      "grad_norm": 1.6702203750610352,
+      "learning_rate": 0.00012247330960854093,
+      "loss": 2.0756,
+      "step": 4367
+    },
+    {
+      "epoch": 1.9413333333333334,
+      "grad_norm": 2.062119483947754,
+      "learning_rate": 0.0001224555160142349,
+      "loss": 3.1842,
+      "step": 4368
+    },
+    {
+      "epoch": 1.9417777777777778,
+      "grad_norm": 1.8628273010253906,
+      "learning_rate": 0.00012243772241992884,
+      "loss": 2.0465,
+      "step": 4369
+    },
+    {
+      "epoch": 1.942222222222222,
+      "grad_norm": 1.9983272552490234,
+      "learning_rate": 0.0001224199288256228,
+      "loss": 1.8614,
+      "step": 4370
+    },
+    {
+      "epoch": 1.9426666666666668,
+      "grad_norm": 1.8302124738693237,
+      "learning_rate": 0.00012240213523131673,
+      "loss": 2.2356,
+      "step": 4371
+    },
+    {
+      "epoch": 1.943111111111111,
+      "grad_norm": 2.0615060329437256,
+      "learning_rate": 0.00012238434163701069,
+      "loss": 2.6518,
+      "step": 4372
+    },
+    {
+      "epoch": 1.9435555555555557,
+      "grad_norm": 1.957165002822876,
+      "learning_rate": 0.00012236654804270464,
+      "loss": 2.5132,
+      "step": 4373
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 2.0981383323669434,
+      "learning_rate": 0.00012234875444839857,
+      "loss": 2.5369,
+      "step": 4374
+    },
+    {
+      "epoch": 1.9444444444444444,
+      "grad_norm": 2.0047876834869385,
+      "learning_rate": 0.00012233096085409253,
+      "loss": 1.8372,
+      "step": 4375
+    },
+    {
+      "epoch": 1.944888888888889,
+      "grad_norm": 1.8745818138122559,
+      "learning_rate": 0.00012231316725978649,
+      "loss": 2.0336,
+      "step": 4376
+    },
+    {
+      "epoch": 1.9453333333333334,
+      "grad_norm": 1.9470521211624146,
+      "learning_rate": 0.00012229537366548044,
+      "loss": 2.0855,
+      "step": 4377
+    },
+    {
+      "epoch": 1.9457777777777778,
+      "grad_norm": 1.992088794708252,
+      "learning_rate": 0.00012227758007117437,
+      "loss": 2.2125,
+      "step": 4378
+    },
+    {
+      "epoch": 1.946222222222222,
+      "grad_norm": 1.8258486986160278,
+      "learning_rate": 0.00012225978647686833,
+      "loss": 1.8891,
+      "step": 4379
+    },
+    {
+      "epoch": 1.9466666666666668,
+      "grad_norm": 1.850938081741333,
+      "learning_rate": 0.00012224199288256228,
+      "loss": 2.3489,
+      "step": 4380
+    },
+    {
+      "epoch": 1.947111111111111,
+      "grad_norm": 2.0930421352386475,
+      "learning_rate": 0.00012222419928825624,
+      "loss": 2.173,
+      "step": 4381
+    },
+    {
+      "epoch": 1.9475555555555557,
+      "grad_norm": 1.875168800354004,
+      "learning_rate": 0.0001222064056939502,
+      "loss": 1.7742,
+      "step": 4382
+    },
+    {
+      "epoch": 1.948,
+      "grad_norm": 2.2607669830322266,
+      "learning_rate": 0.00012218861209964413,
+      "loss": 2.6387,
+      "step": 4383
+    },
+    {
+      "epoch": 1.9484444444444444,
+      "grad_norm": 1.710083246231079,
+      "learning_rate": 0.00012217081850533808,
+      "loss": 2.132,
+      "step": 4384
+    },
+    {
+      "epoch": 1.948888888888889,
+      "grad_norm": 2.2312233448028564,
+      "learning_rate": 0.00012215302491103204,
+      "loss": 1.6657,
+      "step": 4385
+    },
+    {
+      "epoch": 1.9493333333333334,
+      "grad_norm": 2.3804073333740234,
+      "learning_rate": 0.000122135231316726,
+      "loss": 2.277,
+      "step": 4386
+    },
+    {
+      "epoch": 1.9497777777777778,
+      "grad_norm": 2.0216658115386963,
+      "learning_rate": 0.00012211743772241993,
+      "loss": 2.4888,
+      "step": 4387
+    },
+    {
+      "epoch": 1.950222222222222,
+      "grad_norm": 1.6631717681884766,
+      "learning_rate": 0.00012209964412811388,
+      "loss": 1.218,
+      "step": 4388
+    },
+    {
+      "epoch": 1.9506666666666668,
+      "grad_norm": 2.17562198638916,
+      "learning_rate": 0.00012208185053380784,
+      "loss": 1.9006,
+      "step": 4389
+    },
+    {
+      "epoch": 1.951111111111111,
+      "grad_norm": 1.9813545942306519,
+      "learning_rate": 0.00012206405693950178,
+      "loss": 1.9596,
+      "step": 4390
+    },
+    {
+      "epoch": 1.9515555555555557,
+      "grad_norm": 1.9598819017410278,
+      "learning_rate": 0.00012204626334519574,
+      "loss": 2.5526,
+      "step": 4391
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 2.731766700744629,
+      "learning_rate": 0.00012202846975088968,
+      "loss": 2.2122,
+      "step": 4392
+    },
+    {
+      "epoch": 1.9524444444444444,
+      "grad_norm": 2.2050271034240723,
+      "learning_rate": 0.00012201067615658364,
+      "loss": 2.17,
+      "step": 4393
+    },
+    {
+      "epoch": 1.952888888888889,
+      "grad_norm": 2.014303684234619,
+      "learning_rate": 0.0001219928825622776,
+      "loss": 2.0293,
+      "step": 4394
+    },
+    {
+      "epoch": 1.9533333333333334,
+      "grad_norm": 2.2613487243652344,
+      "learning_rate": 0.00012197508896797154,
+      "loss": 2.4478,
+      "step": 4395
+    },
+    {
+      "epoch": 1.9537777777777778,
+      "grad_norm": 2.023850917816162,
+      "learning_rate": 0.00012195729537366549,
+      "loss": 1.7262,
+      "step": 4396
+    },
+    {
+      "epoch": 1.954222222222222,
+      "grad_norm": 1.9727839231491089,
+      "learning_rate": 0.00012193950177935945,
+      "loss": 2.1561,
+      "step": 4397
+    },
+    {
+      "epoch": 1.9546666666666668,
+      "grad_norm": 2.2552168369293213,
+      "learning_rate": 0.00012192170818505339,
+      "loss": 2.3178,
+      "step": 4398
+    },
+    {
+      "epoch": 1.955111111111111,
+      "grad_norm": 2.353135108947754,
+      "learning_rate": 0.00012190391459074735,
+      "loss": 2.2246,
+      "step": 4399
+    },
+    {
+      "epoch": 1.9555555555555557,
+      "grad_norm": 2.8717939853668213,
+      "learning_rate": 0.00012188612099644128,
+      "loss": 2.3769,
+      "step": 4400
+    },
+    {
+      "epoch": 1.956,
+      "grad_norm": 1.117507815361023,
+      "learning_rate": 0.00012186832740213523,
+      "loss": 1.2481,
+      "step": 4401
+    },
+    {
+      "epoch": 1.9564444444444444,
+      "grad_norm": 1.0877759456634521,
+      "learning_rate": 0.00012185053380782918,
+      "loss": 0.9556,
+      "step": 4402
+    },
+    {
+      "epoch": 1.956888888888889,
+      "grad_norm": 1.1263483762741089,
+      "learning_rate": 0.00012183274021352313,
+      "loss": 1.1806,
+      "step": 4403
+    },
+    {
+      "epoch": 1.9573333333333334,
+      "grad_norm": 1.5997364521026611,
+      "learning_rate": 0.00012181494661921709,
+      "loss": 2.2267,
+      "step": 4404
+    },
+    {
+      "epoch": 1.9577777777777778,
+      "grad_norm": 1.4710240364074707,
+      "learning_rate": 0.00012179715302491103,
+      "loss": 2.4633,
+      "step": 4405
+    },
+    {
+      "epoch": 1.958222222222222,
+      "grad_norm": 1.5954550504684448,
+      "learning_rate": 0.00012177935943060499,
+      "loss": 2.3169,
+      "step": 4406
+    },
+    {
+      "epoch": 1.9586666666666668,
+      "grad_norm": 1.688092589378357,
+      "learning_rate": 0.00012176156583629895,
+      "loss": 2.6058,
+      "step": 4407
+    },
+    {
+      "epoch": 1.959111111111111,
+      "grad_norm": 3.413558006286621,
+      "learning_rate": 0.00012174377224199289,
+      "loss": 1.9166,
+      "step": 4408
+    },
+    {
+      "epoch": 1.9595555555555557,
+      "grad_norm": 1.7245512008666992,
+      "learning_rate": 0.00012172597864768685,
+      "loss": 2.4489,
+      "step": 4409
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.8946932554244995,
+      "learning_rate": 0.0001217081850533808,
+      "loss": 2.1216,
+      "step": 4410
+    },
+    {
+      "epoch": 1.9604444444444444,
+      "grad_norm": 1.766736626625061,
+      "learning_rate": 0.00012169039145907475,
+      "loss": 2.4328,
+      "step": 4411
+    },
+    {
+      "epoch": 1.960888888888889,
+      "grad_norm": 1.7106037139892578,
+      "learning_rate": 0.00012167259786476868,
+      "loss": 2.0669,
+      "step": 4412
+    },
+    {
+      "epoch": 1.9613333333333334,
+      "grad_norm": 1.758056879043579,
+      "learning_rate": 0.00012165480427046263,
+      "loss": 2.306,
+      "step": 4413
+    },
+    {
+      "epoch": 1.9617777777777778,
+      "grad_norm": 1.796230673789978,
+      "learning_rate": 0.00012163701067615659,
+      "loss": 2.3531,
+      "step": 4414
+    },
+    {
+      "epoch": 1.962222222222222,
+      "grad_norm": 1.641489863395691,
+      "learning_rate": 0.00012161921708185053,
+      "loss": 1.9206,
+      "step": 4415
+    },
+    {
+      "epoch": 1.9626666666666668,
+      "grad_norm": 1.558814287185669,
+      "learning_rate": 0.00012160142348754449,
+      "loss": 1.7989,
+      "step": 4416
+    },
+    {
+      "epoch": 1.963111111111111,
+      "grad_norm": 1.724256992340088,
+      "learning_rate": 0.00012158362989323844,
+      "loss": 2.0509,
+      "step": 4417
+    },
+    {
+      "epoch": 1.9635555555555557,
+      "grad_norm": 1.6848506927490234,
+      "learning_rate": 0.00012156583629893239,
+      "loss": 2.4329,
+      "step": 4418
+    },
+    {
+      "epoch": 1.964,
+      "grad_norm": 1.874718427658081,
+      "learning_rate": 0.00012154804270462634,
+      "loss": 2.2301,
+      "step": 4419
+    },
+    {
+      "epoch": 1.9644444444444444,
+      "grad_norm": 1.9206674098968506,
+      "learning_rate": 0.0001215302491103203,
+      "loss": 2.2543,
+      "step": 4420
+    },
+    {
+      "epoch": 1.964888888888889,
+      "grad_norm": 1.9313912391662598,
+      "learning_rate": 0.00012151245551601424,
+      "loss": 2.0722,
+      "step": 4421
+    },
+    {
+      "epoch": 1.9653333333333334,
+      "grad_norm": 2.0314974784851074,
+      "learning_rate": 0.0001214946619217082,
+      "loss": 2.2565,
+      "step": 4422
+    },
+    {
+      "epoch": 1.9657777777777778,
+      "grad_norm": 1.886389970779419,
+      "learning_rate": 0.00012147686832740214,
+      "loss": 2.3392,
+      "step": 4423
+    },
+    {
+      "epoch": 1.966222222222222,
+      "grad_norm": 2.023409843444824,
+      "learning_rate": 0.0001214590747330961,
+      "loss": 2.5448,
+      "step": 4424
+    },
+    {
+      "epoch": 1.9666666666666668,
+      "grad_norm": 1.7405117750167847,
+      "learning_rate": 0.00012144128113879003,
+      "loss": 1.7628,
+      "step": 4425
+    },
+    {
+      "epoch": 1.967111111111111,
+      "grad_norm": 1.9717170000076294,
+      "learning_rate": 0.00012142348754448398,
+      "loss": 2.027,
+      "step": 4426
+    },
+    {
+      "epoch": 1.9675555555555555,
+      "grad_norm": 1.9732880592346191,
+      "learning_rate": 0.00012140569395017794,
+      "loss": 2.3524,
+      "step": 4427
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 1.9365227222442627,
+      "learning_rate": 0.00012138790035587188,
+      "loss": 1.9627,
+      "step": 4428
+    },
+    {
+      "epoch": 1.9684444444444444,
+      "grad_norm": 2.359199285507202,
+      "learning_rate": 0.00012137010676156584,
+      "loss": 2.1723,
+      "step": 4429
+    },
+    {
+      "epoch": 1.968888888888889,
+      "grad_norm": 2.102796792984009,
+      "learning_rate": 0.00012135231316725978,
+      "loss": 2.388,
+      "step": 4430
+    },
+    {
+      "epoch": 1.9693333333333334,
+      "grad_norm": 2.1342241764068604,
+      "learning_rate": 0.00012133451957295374,
+      "loss": 2.6232,
+      "step": 4431
+    },
+    {
+      "epoch": 1.9697777777777778,
+      "grad_norm": 2.0955238342285156,
+      "learning_rate": 0.0001213167259786477,
+      "loss": 2.0068,
+      "step": 4432
+    },
+    {
+      "epoch": 1.970222222222222,
+      "grad_norm": 2.0554394721984863,
+      "learning_rate": 0.00012129893238434164,
+      "loss": 2.0677,
+      "step": 4433
+    },
+    {
+      "epoch": 1.9706666666666668,
+      "grad_norm": 2.020298957824707,
+      "learning_rate": 0.0001212811387900356,
+      "loss": 1.9219,
+      "step": 4434
+    },
+    {
+      "epoch": 1.971111111111111,
+      "grad_norm": 1.943292498588562,
+      "learning_rate": 0.00012126334519572955,
+      "loss": 2.1289,
+      "step": 4435
+    },
+    {
+      "epoch": 1.9715555555555555,
+      "grad_norm": 1.9721516370773315,
+      "learning_rate": 0.0001212455516014235,
+      "loss": 1.7805,
+      "step": 4436
+    },
+    {
+      "epoch": 1.972,
+      "grad_norm": 1.828384518623352,
+      "learning_rate": 0.00012122775800711745,
+      "loss": 1.8164,
+      "step": 4437
+    },
+    {
+      "epoch": 1.9724444444444444,
+      "grad_norm": 2.0130600929260254,
+      "learning_rate": 0.00012120996441281138,
+      "loss": 1.8864,
+      "step": 4438
+    },
+    {
+      "epoch": 1.972888888888889,
+      "grad_norm": 2.123643398284912,
+      "learning_rate": 0.00012119217081850534,
+      "loss": 1.9729,
+      "step": 4439
+    },
+    {
+      "epoch": 1.9733333333333334,
+      "grad_norm": 2.224832534790039,
+      "learning_rate": 0.00012117437722419928,
+      "loss": 2.3063,
+      "step": 4440
+    },
+    {
+      "epoch": 1.9737777777777779,
+      "grad_norm": 2.5760409832000732,
+      "learning_rate": 0.00012115658362989324,
+      "loss": 3.1176,
+      "step": 4441
+    },
+    {
+      "epoch": 1.974222222222222,
+      "grad_norm": 1.936394214630127,
+      "learning_rate": 0.0001211387900355872,
+      "loss": 1.9176,
+      "step": 4442
+    },
+    {
+      "epoch": 1.9746666666666668,
+      "grad_norm": 2.0854177474975586,
+      "learning_rate": 0.00012112099644128114,
+      "loss": 2.125,
+      "step": 4443
+    },
+    {
+      "epoch": 1.975111111111111,
+      "grad_norm": 2.2257535457611084,
+      "learning_rate": 0.00012110320284697509,
+      "loss": 2.1138,
+      "step": 4444
+    },
+    {
+      "epoch": 1.9755555555555555,
+      "grad_norm": 2.3724265098571777,
+      "learning_rate": 0.00012108540925266905,
+      "loss": 2.2497,
+      "step": 4445
+    },
+    {
+      "epoch": 1.976,
+      "grad_norm": 1.9573043584823608,
+      "learning_rate": 0.00012106761565836299,
+      "loss": 1.7342,
+      "step": 4446
+    },
+    {
+      "epoch": 1.9764444444444444,
+      "grad_norm": 2.2017226219177246,
+      "learning_rate": 0.00012104982206405695,
+      "loss": 2.292,
+      "step": 4447
+    },
+    {
+      "epoch": 1.976888888888889,
+      "grad_norm": 2.3729429244995117,
+      "learning_rate": 0.0001210320284697509,
+      "loss": 2.853,
+      "step": 4448
+    },
+    {
+      "epoch": 1.9773333333333334,
+      "grad_norm": 2.4285669326782227,
+      "learning_rate": 0.00012101423487544485,
+      "loss": 2.0607,
+      "step": 4449
+    },
+    {
+      "epoch": 1.9777777777777779,
+      "grad_norm": 2.3270487785339355,
+      "learning_rate": 0.0001209964412811388,
+      "loss": 1.2632,
+      "step": 4450
+    },
+    {
+      "epoch": 1.978222222222222,
+      "grad_norm": 1.3772460222244263,
+      "learning_rate": 0.00012097864768683273,
+      "loss": 2.7924,
+      "step": 4451
+    },
+    {
+      "epoch": 1.9786666666666668,
+      "grad_norm": 1.451092004776001,
+      "learning_rate": 0.00012096085409252669,
+      "loss": 2.6418,
+      "step": 4452
+    },
+    {
+      "epoch": 1.979111111111111,
+      "grad_norm": 1.4491418600082397,
+      "learning_rate": 0.00012094306049822063,
+      "loss": 2.5647,
+      "step": 4453
+    },
+    {
+      "epoch": 1.9795555555555555,
+      "grad_norm": 1.4295765161514282,
+      "learning_rate": 0.00012092526690391459,
+      "loss": 2.092,
+      "step": 4454
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.1193150281906128,
+      "learning_rate": 0.00012090747330960855,
+      "loss": 0.9929,
+      "step": 4455
+    },
+    {
+      "epoch": 1.9804444444444445,
+      "grad_norm": 1.6379578113555908,
+      "learning_rate": 0.00012088967971530249,
+      "loss": 2.621,
+      "step": 4456
+    },
+    {
+      "epoch": 1.980888888888889,
+      "grad_norm": 1.7126623392105103,
+      "learning_rate": 0.00012087188612099645,
+      "loss": 1.7564,
+      "step": 4457
+    },
+    {
+      "epoch": 1.9813333333333332,
+      "grad_norm": 1.670942783355713,
+      "learning_rate": 0.0001208540925266904,
+      "loss": 2.1591,
+      "step": 4458
+    },
+    {
+      "epoch": 1.9817777777777779,
+      "grad_norm": 1.9480079412460327,
+      "learning_rate": 0.00012083629893238435,
+      "loss": 2.5495,
+      "step": 4459
+    },
+    {
+      "epoch": 1.982222222222222,
+      "grad_norm": 1.5117992162704468,
+      "learning_rate": 0.0001208185053380783,
+      "loss": 2.1216,
+      "step": 4460
+    },
+    {
+      "epoch": 1.9826666666666668,
+      "grad_norm": 1.5676404237747192,
+      "learning_rate": 0.00012080071174377226,
+      "loss": 2.2749,
+      "step": 4461
+    },
+    {
+      "epoch": 1.983111111111111,
+      "grad_norm": 1.228750467300415,
+      "learning_rate": 0.0001207829181494662,
+      "loss": 0.9559,
+      "step": 4462
+    },
+    {
+      "epoch": 1.9835555555555555,
+      "grad_norm": 1.8625293970108032,
+      "learning_rate": 0.00012076512455516016,
+      "loss": 2.1203,
+      "step": 4463
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 1.5544917583465576,
+      "learning_rate": 0.00012074733096085409,
+      "loss": 2.1986,
+      "step": 4464
+    },
+    {
+      "epoch": 1.9844444444444445,
+      "grad_norm": 1.8329989910125732,
+      "learning_rate": 0.00012072953736654804,
+      "loss": 2.1676,
+      "step": 4465
+    },
+    {
+      "epoch": 1.984888888888889,
+      "grad_norm": 1.6483327150344849,
+      "learning_rate": 0.00012071174377224199,
+      "loss": 2.383,
+      "step": 4466
+    },
+    {
+      "epoch": 1.9853333333333332,
+      "grad_norm": 1.6952513456344604,
+      "learning_rate": 0.00012069395017793594,
+      "loss": 2.0415,
+      "step": 4467
+    },
+    {
+      "epoch": 1.9857777777777779,
+      "grad_norm": 1.7458889484405518,
+      "learning_rate": 0.0001206761565836299,
+      "loss": 2.0539,
+      "step": 4468
+    },
+    {
+      "epoch": 1.9862222222222221,
+      "grad_norm": 1.8751850128173828,
+      "learning_rate": 0.00012065836298932384,
+      "loss": 2.391,
+      "step": 4469
+    },
+    {
+      "epoch": 1.9866666666666668,
+      "grad_norm": 1.6705282926559448,
+      "learning_rate": 0.0001206405693950178,
+      "loss": 1.8026,
+      "step": 4470
+    },
+    {
+      "epoch": 1.987111111111111,
+      "grad_norm": 1.6303818225860596,
+      "learning_rate": 0.00012062277580071176,
+      "loss": 1.899,
+      "step": 4471
+    },
+    {
+      "epoch": 1.9875555555555555,
+      "grad_norm": 1.7543621063232422,
+      "learning_rate": 0.0001206049822064057,
+      "loss": 1.9756,
+      "step": 4472
+    },
+    {
+      "epoch": 1.988,
+      "grad_norm": 1.8945670127868652,
+      "learning_rate": 0.00012058718861209966,
+      "loss": 2.1054,
+      "step": 4473
+    },
+    {
+      "epoch": 1.9884444444444445,
+      "grad_norm": 1.76500403881073,
+      "learning_rate": 0.00012056939501779361,
+      "loss": 2.2563,
+      "step": 4474
+    },
+    {
+      "epoch": 1.988888888888889,
+      "grad_norm": 1.8483823537826538,
+      "learning_rate": 0.00012055160142348755,
+      "loss": 2.243,
+      "step": 4475
+    },
+    {
+      "epoch": 1.9893333333333332,
+      "grad_norm": 1.7314825057983398,
+      "learning_rate": 0.00012053380782918151,
+      "loss": 2.4037,
+      "step": 4476
+    },
+    {
+      "epoch": 1.9897777777777779,
+      "grad_norm": 1.545645833015442,
+      "learning_rate": 0.00012051601423487544,
+      "loss": 1.1406,
+      "step": 4477
+    },
+    {
+      "epoch": 1.9902222222222221,
+      "grad_norm": 1.576292634010315,
+      "learning_rate": 0.0001204982206405694,
+      "loss": 1.7195,
+      "step": 4478
+    },
+    {
+      "epoch": 1.9906666666666668,
+      "grad_norm": 1.6278116703033447,
+      "learning_rate": 0.00012048042704626334,
+      "loss": 1.8675,
+      "step": 4479
+    },
+    {
+      "epoch": 1.991111111111111,
+      "grad_norm": 1.6416962146759033,
+      "learning_rate": 0.0001204626334519573,
+      "loss": 1.9954,
+      "step": 4480
+    },
+    {
+      "epoch": 1.9915555555555555,
+      "grad_norm": 1.7250615358352661,
+      "learning_rate": 0.00012044483985765125,
+      "loss": 2.2059,
+      "step": 4481
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 1.973854899406433,
+      "learning_rate": 0.0001204270462633452,
+      "loss": 2.4063,
+      "step": 4482
+    },
+    {
+      "epoch": 1.9924444444444445,
+      "grad_norm": 2.003932476043701,
+      "learning_rate": 0.00012040925266903915,
+      "loss": 1.9116,
+      "step": 4483
+    },
+    {
+      "epoch": 1.992888888888889,
+      "grad_norm": 1.8206032514572144,
+      "learning_rate": 0.00012039145907473311,
+      "loss": 1.9458,
+      "step": 4484
+    },
+    {
+      "epoch": 1.9933333333333332,
+      "grad_norm": 2.3211958408355713,
+      "learning_rate": 0.00012037366548042705,
+      "loss": 2.565,
+      "step": 4485
+    },
+    {
+      "epoch": 1.9937777777777779,
+      "grad_norm": 2.1076741218566895,
+      "learning_rate": 0.00012035587188612101,
+      "loss": 2.1175,
+      "step": 4486
+    },
+    {
+      "epoch": 1.9942222222222221,
+      "grad_norm": 1.6778833866119385,
+      "learning_rate": 0.00012033807829181497,
+      "loss": 1.5964,
+      "step": 4487
+    },
+    {
+      "epoch": 1.9946666666666668,
+      "grad_norm": 2.298328399658203,
+      "learning_rate": 0.00012032028469750891,
+      "loss": 2.1779,
+      "step": 4488
+    },
+    {
+      "epoch": 1.995111111111111,
+      "grad_norm": 1.8983428478240967,
+      "learning_rate": 0.00012030249110320286,
+      "loss": 2.0109,
+      "step": 4489
+    },
+    {
+      "epoch": 1.9955555555555555,
+      "grad_norm": 1.9424333572387695,
+      "learning_rate": 0.0001202846975088968,
+      "loss": 1.6167,
+      "step": 4490
+    },
+    {
+      "epoch": 1.996,
+      "grad_norm": 1.9415746927261353,
+      "learning_rate": 0.00012026690391459075,
+      "loss": 1.7478,
+      "step": 4491
+    },
+    {
+      "epoch": 1.9964444444444445,
+      "grad_norm": 1.8782165050506592,
+      "learning_rate": 0.0001202491103202847,
+      "loss": 2.08,
+      "step": 4492
+    },
+    {
+      "epoch": 1.996888888888889,
+      "grad_norm": 2.077409267425537,
+      "learning_rate": 0.00012023131672597865,
+      "loss": 1.9588,
+      "step": 4493
+    },
+    {
+      "epoch": 1.9973333333333332,
+      "grad_norm": 2.3767199516296387,
+      "learning_rate": 0.0001202135231316726,
+      "loss": 2.4939,
+      "step": 4494
+    },
+    {
+      "epoch": 1.9977777777777779,
+      "grad_norm": 2.2348458766937256,
+      "learning_rate": 0.00012019572953736655,
+      "loss": 2.3313,
+      "step": 4495
+    },
+    {
+      "epoch": 1.9982222222222221,
+      "grad_norm": 2.509856700897217,
+      "learning_rate": 0.0001201779359430605,
+      "loss": 2.5201,
+      "step": 4496
+    },
+    {
+      "epoch": 1.9986666666666668,
+      "grad_norm": 2.5316574573516846,
+      "learning_rate": 0.00012016014234875446,
+      "loss": 2.3262,
+      "step": 4497
+    },
+    {
+      "epoch": 1.999111111111111,
+      "grad_norm": 2.4037058353424072,
+      "learning_rate": 0.0001201423487544484,
+      "loss": 1.8568,
+      "step": 4498
+    },
+    {
+      "epoch": 1.9995555555555555,
+      "grad_norm": 2.8537352085113525,
+      "learning_rate": 0.00012012455516014236,
+      "loss": 2.8726,
+      "step": 4499
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.3729751110076904,
+      "learning_rate": 0.00012010676156583632,
+      "loss": 2.0103,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 2.4265496730804443,
+      "eval_runtime": 47.2809,
+      "eval_samples_per_second": 10.575,
+      "eval_steps_per_second": 10.575,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0004444444444442,
+      "grad_norm": 0.8853614926338196,
+      "learning_rate": 0.00012008896797153026,
+      "loss": 0.8435,
+      "step": 4501
+    },
+    {
+      "epoch": 2.000888888888889,
+      "grad_norm": 1.3319120407104492,
+      "learning_rate": 0.00012007117437722422,
+      "loss": 2.0388,
+      "step": 4502
+    },
+    {
+      "epoch": 2.001333333333333,
+      "grad_norm": 1.6479343175888062,
+      "learning_rate": 0.00012005338078291815,
+      "loss": 1.9234,
+      "step": 4503
+    },
+    {
+      "epoch": 2.001777777777778,
+      "grad_norm": 1.6244651079177856,
+      "learning_rate": 0.0001200355871886121,
+      "loss": 2.0868,
+      "step": 4504
+    },
+    {
+      "epoch": 2.002222222222222,
+      "grad_norm": 1.5395774841308594,
+      "learning_rate": 0.00012001779359430605,
+      "loss": 2.1305,
+      "step": 4505
+    },
+    {
+      "epoch": 2.002666666666667,
+      "grad_norm": 1.7427664995193481,
+      "learning_rate": 0.00012,
+      "loss": 2.3715,
+      "step": 4506
+    },
+    {
+      "epoch": 2.003111111111111,
+      "grad_norm": 1.600896954536438,
+      "learning_rate": 0.00011998220640569396,
+      "loss": 1.9859,
+      "step": 4507
+    },
+    {
+      "epoch": 2.0035555555555558,
+      "grad_norm": 1.6297886371612549,
+      "learning_rate": 0.0001199644128113879,
+      "loss": 1.4048,
+      "step": 4508
+    },
+    {
+      "epoch": 2.004,
+      "grad_norm": 1.7642384767532349,
+      "learning_rate": 0.00011994661921708186,
+      "loss": 2.2935,
+      "step": 4509
+    },
+    {
+      "epoch": 2.0044444444444443,
+      "grad_norm": 1.5079247951507568,
+      "learning_rate": 0.00011992882562277582,
+      "loss": 1.7682,
+      "step": 4510
+    },
+    {
+      "epoch": 2.004888888888889,
+      "grad_norm": 1.8100050687789917,
+      "learning_rate": 0.00011991103202846976,
+      "loss": 1.9194,
+      "step": 4511
+    },
+    {
+      "epoch": 2.005333333333333,
+      "grad_norm": 1.9946025609970093,
+      "learning_rate": 0.00011989323843416371,
+      "loss": 1.9621,
+      "step": 4512
+    },
+    {
+      "epoch": 2.005777777777778,
+      "grad_norm": 1.6905927658081055,
+      "learning_rate": 0.00011987544483985766,
+      "loss": 2.0369,
+      "step": 4513
+    },
+    {
+      "epoch": 2.006222222222222,
+      "grad_norm": 1.6684443950653076,
+      "learning_rate": 0.00011985765124555161,
+      "loss": 1.8459,
+      "step": 4514
+    },
+    {
+      "epoch": 2.006666666666667,
+      "grad_norm": 2.0483903884887695,
+      "learning_rate": 0.00011983985765124557,
+      "loss": 1.9101,
+      "step": 4515
+    },
+    {
+      "epoch": 2.007111111111111,
+      "grad_norm": 2.2682251930236816,
+      "learning_rate": 0.0001198220640569395,
+      "loss": 1.5367,
+      "step": 4516
+    },
+    {
+      "epoch": 2.0075555555555558,
+      "grad_norm": 1.442750096321106,
+      "learning_rate": 0.00011980427046263346,
+      "loss": 0.6781,
+      "step": 4517
+    },
+    {
+      "epoch": 2.008,
+      "grad_norm": 2.0190680027008057,
+      "learning_rate": 0.0001197864768683274,
+      "loss": 1.9999,
+      "step": 4518
+    },
+    {
+      "epoch": 2.0084444444444443,
+      "grad_norm": 2.084582567214966,
+      "learning_rate": 0.00011976868327402136,
+      "loss": 2.1547,
+      "step": 4519
+    },
+    {
+      "epoch": 2.008888888888889,
+      "grad_norm": 2.291092872619629,
+      "learning_rate": 0.0001197508896797153,
+      "loss": 1.982,
+      "step": 4520
+    },
+    {
+      "epoch": 2.009333333333333,
+      "grad_norm": 2.4098877906799316,
+      "learning_rate": 0.00011973309608540926,
+      "loss": 1.9975,
+      "step": 4521
+    },
+    {
+      "epoch": 2.009777777777778,
+      "grad_norm": 2.174729585647583,
+      "learning_rate": 0.00011971530249110321,
+      "loss": 1.8608,
+      "step": 4522
+    },
+    {
+      "epoch": 2.010222222222222,
+      "grad_norm": 2.5682456493377686,
+      "learning_rate": 0.00011969750889679716,
+      "loss": 1.8649,
+      "step": 4523
+    },
+    {
+      "epoch": 2.010666666666667,
+      "grad_norm": 1.9961142539978027,
+      "learning_rate": 0.00011967971530249111,
+      "loss": 1.5078,
+      "step": 4524
+    },
+    {
+      "epoch": 2.011111111111111,
+      "grad_norm": 2.4178638458251953,
+      "learning_rate": 0.00011966192170818507,
+      "loss": 1.7468,
+      "step": 4525
+    },
+    {
+      "epoch": 2.0115555555555558,
+      "grad_norm": 2.30759334564209,
+      "learning_rate": 0.00011964412811387901,
+      "loss": 1.8143,
+      "step": 4526
+    },
+    {
+      "epoch": 2.012,
+      "grad_norm": 1.8575639724731445,
+      "learning_rate": 0.00011962633451957297,
+      "loss": 1.2505,
+      "step": 4527
+    },
+    {
+      "epoch": 2.0124444444444443,
+      "grad_norm": 2.4838311672210693,
+      "learning_rate": 0.0001196085409252669,
+      "loss": 1.9343,
+      "step": 4528
+    },
+    {
+      "epoch": 2.012888888888889,
+      "grad_norm": 2.2968297004699707,
+      "learning_rate": 0.00011959074733096085,
+      "loss": 1.8051,
+      "step": 4529
+    },
+    {
+      "epoch": 2.013333333333333,
+      "grad_norm": 2.502021074295044,
+      "learning_rate": 0.0001195729537366548,
+      "loss": 1.5853,
+      "step": 4530
+    },
+    {
+      "epoch": 2.013777777777778,
+      "grad_norm": 1.7160247564315796,
+      "learning_rate": 0.00011955516014234875,
+      "loss": 0.898,
+      "step": 4531
+    },
+    {
+      "epoch": 2.014222222222222,
+      "grad_norm": 2.109179973602295,
+      "learning_rate": 0.00011953736654804271,
+      "loss": 1.3351,
+      "step": 4532
+    },
+    {
+      "epoch": 2.014666666666667,
+      "grad_norm": 2.4689619541168213,
+      "learning_rate": 0.00011951957295373665,
+      "loss": 1.8757,
+      "step": 4533
+    },
+    {
+      "epoch": 2.015111111111111,
+      "grad_norm": 3.084038257598877,
+      "learning_rate": 0.00011950177935943061,
+      "loss": 2.2736,
+      "step": 4534
+    },
+    {
+      "epoch": 2.0155555555555558,
+      "grad_norm": 3.0082006454467773,
+      "learning_rate": 0.00011948398576512457,
+      "loss": 2.0318,
+      "step": 4535
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 2.722201108932495,
+      "learning_rate": 0.00011946619217081851,
+      "loss": 1.7501,
+      "step": 4536
+    },
+    {
+      "epoch": 2.0164444444444443,
+      "grad_norm": 2.107433795928955,
+      "learning_rate": 0.00011944839857651246,
+      "loss": 1.4507,
+      "step": 4537
+    },
+    {
+      "epoch": 2.016888888888889,
+      "grad_norm": 2.8475327491760254,
+      "learning_rate": 0.00011943060498220642,
+      "loss": 1.8524,
+      "step": 4538
+    },
+    {
+      "epoch": 2.017333333333333,
+      "grad_norm": 2.276287078857422,
+      "learning_rate": 0.00011941281138790036,
+      "loss": 1.5575,
+      "step": 4539
+    },
+    {
+      "epoch": 2.017777777777778,
+      "grad_norm": 2.9798543453216553,
+      "learning_rate": 0.00011939501779359432,
+      "loss": 1.8226,
+      "step": 4540
+    },
+    {
+      "epoch": 2.018222222222222,
+      "grad_norm": 2.7004785537719727,
+      "learning_rate": 0.00011937722419928825,
+      "loss": 1.6149,
+      "step": 4541
+    },
+    {
+      "epoch": 2.018666666666667,
+      "grad_norm": 2.5577023029327393,
+      "learning_rate": 0.0001193594306049822,
+      "loss": 1.6913,
+      "step": 4542
+    },
+    {
+      "epoch": 2.019111111111111,
+      "grad_norm": 2.958678960800171,
+      "learning_rate": 0.00011934163701067615,
+      "loss": 1.9823,
+      "step": 4543
+    },
+    {
+      "epoch": 2.0195555555555558,
+      "grad_norm": 2.862009048461914,
+      "learning_rate": 0.0001193238434163701,
+      "loss": 2.1328,
+      "step": 4544
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.533651113510132,
+      "learning_rate": 0.00011930604982206406,
+      "loss": 1.2886,
+      "step": 4545
+    },
+    {
+      "epoch": 2.0204444444444443,
+      "grad_norm": 2.7307307720184326,
+      "learning_rate": 0.000119288256227758,
+      "loss": 1.4595,
+      "step": 4546
+    },
+    {
+      "epoch": 2.020888888888889,
+      "grad_norm": 2.3221077919006348,
+      "learning_rate": 0.00011927046263345196,
+      "loss": 1.4627,
+      "step": 4547
+    },
+    {
+      "epoch": 2.021333333333333,
+      "grad_norm": 2.621258497238159,
+      "learning_rate": 0.00011925266903914592,
+      "loss": 1.9821,
+      "step": 4548
+    },
+    {
+      "epoch": 2.021777777777778,
+      "grad_norm": 3.0992743968963623,
+      "learning_rate": 0.00011923487544483986,
+      "loss": 1.9199,
+      "step": 4549
+    },
+    {
+      "epoch": 2.022222222222222,
+      "grad_norm": 2.962254524230957,
+      "learning_rate": 0.00011921708185053382,
+      "loss": 1.2242,
+      "step": 4550
+    },
+    {
+      "epoch": 2.022666666666667,
+      "grad_norm": 1.6041865348815918,
+      "learning_rate": 0.00011919928825622777,
+      "loss": 1.3209,
+      "step": 4551
+    },
+    {
+      "epoch": 2.023111111111111,
+      "grad_norm": 1.467262864112854,
+      "learning_rate": 0.00011918149466192172,
+      "loss": 1.234,
+      "step": 4552
+    },
+    {
+      "epoch": 2.0235555555555558,
+      "grad_norm": 1.7941569089889526,
+      "learning_rate": 0.00011916370106761567,
+      "loss": 2.2728,
+      "step": 4553
+    },
+    {
+      "epoch": 2.024,
+      "grad_norm": 1.5521703958511353,
+      "learning_rate": 0.0001191459074733096,
+      "loss": 1.8525,
+      "step": 4554
+    },
+    {
+      "epoch": 2.0244444444444443,
+      "grad_norm": 1.894801378250122,
+      "learning_rate": 0.00011912811387900356,
+      "loss": 2.1562,
+      "step": 4555
+    },
+    {
+      "epoch": 2.024888888888889,
+      "grad_norm": 1.414631962776184,
+      "learning_rate": 0.0001191103202846975,
+      "loss": 0.9182,
+      "step": 4556
+    },
+    {
+      "epoch": 2.025333333333333,
+      "grad_norm": 2.034543752670288,
+      "learning_rate": 0.00011909252669039146,
+      "loss": 2.1913,
+      "step": 4557
+    },
+    {
+      "epoch": 2.025777777777778,
+      "grad_norm": 1.8976682424545288,
+      "learning_rate": 0.00011907473309608542,
+      "loss": 2.0936,
+      "step": 4558
+    },
+    {
+      "epoch": 2.026222222222222,
+      "grad_norm": 1.6398429870605469,
+      "learning_rate": 0.00011905693950177936,
+      "loss": 1.7521,
+      "step": 4559
+    },
+    {
+      "epoch": 2.026666666666667,
+      "grad_norm": 1.8975718021392822,
+      "learning_rate": 0.00011903914590747332,
+      "loss": 1.6575,
+      "step": 4560
+    },
+    {
+      "epoch": 2.027111111111111,
+      "grad_norm": 1.5204815864562988,
+      "learning_rate": 0.00011902135231316727,
+      "loss": 0.9841,
+      "step": 4561
+    },
+    {
+      "epoch": 2.0275555555555558,
+      "grad_norm": 1.710460901260376,
+      "learning_rate": 0.00011900355871886121,
+      "loss": 1.4803,
+      "step": 4562
+    },
+    {
+      "epoch": 2.028,
+      "grad_norm": 1.8635836839675903,
+      "learning_rate": 0.00011898576512455517,
+      "loss": 2.0793,
+      "step": 4563
+    },
+    {
+      "epoch": 2.0284444444444443,
+      "grad_norm": 1.871050238609314,
+      "learning_rate": 0.00011896797153024913,
+      "loss": 1.0645,
+      "step": 4564
+    },
+    {
+      "epoch": 2.028888888888889,
+      "grad_norm": 2.071890115737915,
+      "learning_rate": 0.00011895017793594307,
+      "loss": 1.9001,
+      "step": 4565
+    },
+    {
+      "epoch": 2.029333333333333,
+      "grad_norm": 2.0963134765625,
+      "learning_rate": 0.00011893238434163703,
+      "loss": 2.1881,
+      "step": 4566
+    },
+    {
+      "epoch": 2.029777777777778,
+      "grad_norm": 1.9067516326904297,
+      "learning_rate": 0.00011891459074733096,
+      "loss": 1.7577,
+      "step": 4567
+    },
+    {
+      "epoch": 2.030222222222222,
+      "grad_norm": 2.036006212234497,
+      "learning_rate": 0.00011889679715302491,
+      "loss": 2.1429,
+      "step": 4568
+    },
+    {
+      "epoch": 2.030666666666667,
+      "grad_norm": 2.4452297687530518,
+      "learning_rate": 0.00011887900355871886,
+      "loss": 2.1539,
+      "step": 4569
+    },
+    {
+      "epoch": 2.031111111111111,
+      "grad_norm": 2.6038591861724854,
+      "learning_rate": 0.00011886120996441281,
+      "loss": 2.0575,
+      "step": 4570
+    },
+    {
+      "epoch": 2.0315555555555553,
+      "grad_norm": 1.8687160015106201,
+      "learning_rate": 0.00011884341637010677,
+      "loss": 1.633,
+      "step": 4571
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 2.6096248626708984,
+      "learning_rate": 0.00011882562277580071,
+      "loss": 1.9688,
+      "step": 4572
+    },
+    {
+      "epoch": 2.0324444444444443,
+      "grad_norm": 1.9733986854553223,
+      "learning_rate": 0.00011880782918149467,
+      "loss": 2.0018,
+      "step": 4573
+    },
+    {
+      "epoch": 2.032888888888889,
+      "grad_norm": 2.4421169757843018,
+      "learning_rate": 0.00011879003558718862,
+      "loss": 1.9457,
+      "step": 4574
+    },
+    {
+      "epoch": 2.033333333333333,
+      "grad_norm": 2.1608901023864746,
+      "learning_rate": 0.00011877224199288257,
+      "loss": 1.5918,
+      "step": 4575
+    },
+    {
+      "epoch": 2.033777777777778,
+      "grad_norm": 1.7584596872329712,
+      "learning_rate": 0.00011875444839857652,
+      "loss": 1.5401,
+      "step": 4576
+    },
+    {
+      "epoch": 2.034222222222222,
+      "grad_norm": 3.489712953567505,
+      "learning_rate": 0.00011873665480427048,
+      "loss": 0.9426,
+      "step": 4577
+    },
+    {
+      "epoch": 2.034666666666667,
+      "grad_norm": 2.114403486251831,
+      "learning_rate": 0.00011871886120996442,
+      "loss": 1.8606,
+      "step": 4578
+    },
+    {
+      "epoch": 2.035111111111111,
+      "grad_norm": 2.2650794982910156,
+      "learning_rate": 0.00011870106761565838,
+      "loss": 1.6157,
+      "step": 4579
+    },
+    {
+      "epoch": 2.0355555555555553,
+      "grad_norm": 2.404672145843506,
+      "learning_rate": 0.00011868327402135231,
+      "loss": 1.9152,
+      "step": 4580
+    },
+    {
+      "epoch": 2.036,
+      "grad_norm": 2.237191677093506,
+      "learning_rate": 0.00011866548042704627,
+      "loss": 1.8395,
+      "step": 4581
+    },
+    {
+      "epoch": 2.0364444444444443,
+      "grad_norm": 2.234955072402954,
+      "learning_rate": 0.00011864768683274021,
+      "loss": 1.7004,
+      "step": 4582
+    },
+    {
+      "epoch": 2.036888888888889,
+      "grad_norm": 1.999866247177124,
+      "learning_rate": 0.00011862989323843417,
+      "loss": 1.33,
+      "step": 4583
+    },
+    {
+      "epoch": 2.037333333333333,
+      "grad_norm": 2.5568530559539795,
+      "learning_rate": 0.00011861209964412812,
+      "loss": 1.7804,
+      "step": 4584
+    },
+    {
+      "epoch": 2.037777777777778,
+      "grad_norm": 2.2619681358337402,
+      "learning_rate": 0.00011859430604982206,
+      "loss": 1.7402,
+      "step": 4585
+    },
+    {
+      "epoch": 2.038222222222222,
+      "grad_norm": 2.6120476722717285,
+      "learning_rate": 0.00011857651245551602,
+      "loss": 1.6572,
+      "step": 4586
+    },
+    {
+      "epoch": 2.038666666666667,
+      "grad_norm": 2.367854595184326,
+      "learning_rate": 0.00011855871886120998,
+      "loss": 1.3876,
+      "step": 4587
+    },
+    {
+      "epoch": 2.039111111111111,
+      "grad_norm": 2.1282546520233154,
+      "learning_rate": 0.00011854092526690392,
+      "loss": 1.5421,
+      "step": 4588
+    },
+    {
+      "epoch": 2.0395555555555553,
+      "grad_norm": 2.2529513835906982,
+      "learning_rate": 0.00011852313167259788,
+      "loss": 1.2083,
+      "step": 4589
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.7248787879943848,
+      "learning_rate": 0.00011850533807829183,
+      "loss": 1.6816,
+      "step": 4590
+    },
+    {
+      "epoch": 2.0404444444444443,
+      "grad_norm": 3.0865466594696045,
+      "learning_rate": 0.00011848754448398578,
+      "loss": 1.5498,
+      "step": 4591
+    },
+    {
+      "epoch": 2.040888888888889,
+      "grad_norm": 3.441319465637207,
+      "learning_rate": 0.00011846975088967973,
+      "loss": 1.5423,
+      "step": 4592
+    },
+    {
+      "epoch": 2.041333333333333,
+      "grad_norm": 2.960602283477783,
+      "learning_rate": 0.00011845195729537366,
+      "loss": 1.7773,
+      "step": 4593
+    },
+    {
+      "epoch": 2.041777777777778,
+      "grad_norm": 3.269716501235962,
+      "learning_rate": 0.00011843416370106762,
+      "loss": 2.1755,
+      "step": 4594
+    },
+    {
+      "epoch": 2.042222222222222,
+      "grad_norm": 3.3861782550811768,
+      "learning_rate": 0.00011841637010676156,
+      "loss": 1.8699,
+      "step": 4595
+    },
+    {
+      "epoch": 2.042666666666667,
+      "grad_norm": 2.951418399810791,
+      "learning_rate": 0.00011839857651245552,
+      "loss": 1.9355,
+      "step": 4596
+    },
+    {
+      "epoch": 2.043111111111111,
+      "grad_norm": 2.9961771965026855,
+      "learning_rate": 0.00011838078291814948,
+      "loss": 2.1783,
+      "step": 4597
+    },
+    {
+      "epoch": 2.0435555555555553,
+      "grad_norm": 2.833582639694214,
+      "learning_rate": 0.00011836298932384342,
+      "loss": 1.8636,
+      "step": 4598
+    },
+    {
+      "epoch": 2.044,
+      "grad_norm": 3.0772078037261963,
+      "learning_rate": 0.00011834519572953737,
+      "loss": 1.4656,
+      "step": 4599
+    },
+    {
+      "epoch": 2.0444444444444443,
+      "grad_norm": 3.595390558242798,
+      "learning_rate": 0.00011832740213523133,
+      "loss": 1.5303,
+      "step": 4600
+    },
+    {
+      "epoch": 2.044888888888889,
+      "grad_norm": 1.3135249614715576,
+      "learning_rate": 0.00011830960854092527,
+      "loss": 1.0165,
+      "step": 4601
+    },
+    {
+      "epoch": 2.0453333333333332,
+      "grad_norm": 1.4123620986938477,
+      "learning_rate": 0.00011829181494661923,
+      "loss": 1.1516,
+      "step": 4602
+    },
+    {
+      "epoch": 2.045777777777778,
+      "grad_norm": 2.0700747966766357,
+      "learning_rate": 0.00011827402135231317,
+      "loss": 1.506,
+      "step": 4603
+    },
+    {
+      "epoch": 2.046222222222222,
+      "grad_norm": 1.950496792793274,
+      "learning_rate": 0.00011825622775800713,
+      "loss": 1.8448,
+      "step": 4604
+    },
+    {
+      "epoch": 2.046666666666667,
+      "grad_norm": 2.0276780128479004,
+      "learning_rate": 0.00011823843416370109,
+      "loss": 1.9754,
+      "step": 4605
+    },
+    {
+      "epoch": 2.047111111111111,
+      "grad_norm": 2.1765434741973877,
+      "learning_rate": 0.00011822064056939502,
+      "loss": 2.5343,
+      "step": 4606
+    },
+    {
+      "epoch": 2.0475555555555554,
+      "grad_norm": 1.9355638027191162,
+      "learning_rate": 0.00011820284697508896,
+      "loss": 2.0598,
+      "step": 4607
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 2.012378215789795,
+      "learning_rate": 0.00011818505338078292,
+      "loss": 1.7933,
+      "step": 4608
+    },
+    {
+      "epoch": 2.0484444444444443,
+      "grad_norm": 2.179774045944214,
+      "learning_rate": 0.00011816725978647687,
+      "loss": 2.1403,
+      "step": 4609
+    },
+    {
+      "epoch": 2.048888888888889,
+      "grad_norm": 2.4090864658355713,
+      "learning_rate": 0.00011814946619217081,
+      "loss": 1.648,
+      "step": 4610
+    },
+    {
+      "epoch": 2.0493333333333332,
+      "grad_norm": 2.0735795497894287,
+      "learning_rate": 0.00011813167259786477,
+      "loss": 1.7304,
+      "step": 4611
+    },
+    {
+      "epoch": 2.049777777777778,
+      "grad_norm": 2.167476177215576,
+      "learning_rate": 0.00011811387900355873,
+      "loss": 1.7997,
+      "step": 4612
+    },
+    {
+      "epoch": 2.050222222222222,
+      "grad_norm": 1.9912680387496948,
+      "learning_rate": 0.00011809608540925267,
+      "loss": 1.4525,
+      "step": 4613
+    },
+    {
+      "epoch": 2.050666666666667,
+      "grad_norm": 2.363731622695923,
+      "learning_rate": 0.00011807829181494663,
+      "loss": 2.1657,
+      "step": 4614
+    },
+    {
+      "epoch": 2.051111111111111,
+      "grad_norm": 1.7121641635894775,
+      "learning_rate": 0.00011806049822064058,
+      "loss": 1.3718,
+      "step": 4615
+    },
+    {
+      "epoch": 2.0515555555555554,
+      "grad_norm": 1.7621139287948608,
+      "learning_rate": 0.00011804270462633453,
+      "loss": 1.3004,
+      "step": 4616
+    },
+    {
+      "epoch": 2.052,
+      "grad_norm": 2.833408832550049,
+      "learning_rate": 0.00011802491103202848,
+      "loss": 1.7775,
+      "step": 4617
+    },
+    {
+      "epoch": 2.0524444444444443,
+      "grad_norm": 1.8854105472564697,
+      "learning_rate": 0.00011800711743772244,
+      "loss": 1.608,
+      "step": 4618
+    },
+    {
+      "epoch": 2.052888888888889,
+      "grad_norm": 2.169327974319458,
+      "learning_rate": 0.00011798932384341637,
+      "loss": 1.9171,
+      "step": 4619
+    },
+    {
+      "epoch": 2.0533333333333332,
+      "grad_norm": 2.554474115371704,
+      "learning_rate": 0.00011797153024911031,
+      "loss": 1.3938,
+      "step": 4620
+    },
+    {
+      "epoch": 2.053777777777778,
+      "grad_norm": 1.969563364982605,
+      "learning_rate": 0.00011795373665480427,
+      "loss": 1.7774,
+      "step": 4621
+    },
+    {
+      "epoch": 2.054222222222222,
+      "grad_norm": 1.854287028312683,
+      "learning_rate": 0.00011793594306049822,
+      "loss": 1.8198,
+      "step": 4622
+    },
+    {
+      "epoch": 2.054666666666667,
+      "grad_norm": 2.1573734283447266,
+      "learning_rate": 0.00011791814946619217,
+      "loss": 1.7923,
+      "step": 4623
+    },
+    {
+      "epoch": 2.055111111111111,
+      "grad_norm": 1.8452686071395874,
+      "learning_rate": 0.00011790035587188612,
+      "loss": 1.2323,
+      "step": 4624
+    },
+    {
+      "epoch": 2.0555555555555554,
+      "grad_norm": 2.0869951248168945,
+      "learning_rate": 0.00011788256227758008,
+      "loss": 1.9431,
+      "step": 4625
+    },
+    {
+      "epoch": 2.056,
+      "grad_norm": 1.9938613176345825,
+      "learning_rate": 0.00011786476868327402,
+      "loss": 1.5775,
+      "step": 4626
+    },
+    {
+      "epoch": 2.0564444444444443,
+      "grad_norm": 2.0701723098754883,
+      "learning_rate": 0.00011784697508896798,
+      "loss": 1.465,
+      "step": 4627
+    },
+    {
+      "epoch": 2.056888888888889,
+      "grad_norm": 2.5349276065826416,
+      "learning_rate": 0.00011782918149466194,
+      "loss": 2.0339,
+      "step": 4628
+    },
+    {
+      "epoch": 2.0573333333333332,
+      "grad_norm": 1.9578278064727783,
+      "learning_rate": 0.00011781138790035588,
+      "loss": 1.8306,
+      "step": 4629
+    },
+    {
+      "epoch": 2.057777777777778,
+      "grad_norm": 2.2757606506347656,
+      "learning_rate": 0.00011779359430604984,
+      "loss": 1.5594,
+      "step": 4630
+    },
+    {
+      "epoch": 2.058222222222222,
+      "grad_norm": 2.5828166007995605,
+      "learning_rate": 0.00011777580071174379,
+      "loss": 2.4208,
+      "step": 4631
+    },
+    {
+      "epoch": 2.058666666666667,
+      "grad_norm": 2.063826084136963,
+      "learning_rate": 0.00011775800711743772,
+      "loss": 1.3296,
+      "step": 4632
+    },
+    {
+      "epoch": 2.059111111111111,
+      "grad_norm": 2.9016664028167725,
+      "learning_rate": 0.00011774021352313167,
+      "loss": 2.1661,
+      "step": 4633
+    },
+    {
+      "epoch": 2.0595555555555554,
+      "grad_norm": 2.3844025135040283,
+      "learning_rate": 0.00011772241992882562,
+      "loss": 1.6781,
+      "step": 4634
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.9740389585494995,
+      "learning_rate": 0.00011770462633451958,
+      "loss": 1.4587,
+      "step": 4635
+    },
+    {
+      "epoch": 2.0604444444444443,
+      "grad_norm": 2.022944688796997,
+      "learning_rate": 0.00011768683274021352,
+      "loss": 1.5967,
+      "step": 4636
+    },
+    {
+      "epoch": 2.060888888888889,
+      "grad_norm": 2.8368794918060303,
+      "learning_rate": 0.00011766903914590748,
+      "loss": 1.9038,
+      "step": 4637
+    },
+    {
+      "epoch": 2.0613333333333332,
+      "grad_norm": 2.7894749641418457,
+      "learning_rate": 0.00011765124555160143,
+      "loss": 1.5072,
+      "step": 4638
+    },
+    {
+      "epoch": 2.061777777777778,
+      "grad_norm": 2.564565896987915,
+      "learning_rate": 0.00011763345195729538,
+      "loss": 1.4213,
+      "step": 4639
+    },
+    {
+      "epoch": 2.062222222222222,
+      "grad_norm": 2.5650315284729004,
+      "learning_rate": 0.00011761565836298933,
+      "loss": 1.6088,
+      "step": 4640
+    },
+    {
+      "epoch": 2.062666666666667,
+      "grad_norm": 1.4882378578186035,
+      "learning_rate": 0.00011759786476868329,
+      "loss": 0.7152,
+      "step": 4641
+    },
+    {
+      "epoch": 2.063111111111111,
+      "grad_norm": 3.052170991897583,
+      "learning_rate": 0.00011758007117437723,
+      "loss": 1.7892,
+      "step": 4642
+    },
+    {
+      "epoch": 2.0635555555555554,
+      "grad_norm": 3.016031265258789,
+      "learning_rate": 0.00011756227758007119,
+      "loss": 2.0574,
+      "step": 4643
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 2.477534532546997,
+      "learning_rate": 0.00011754448398576512,
+      "loss": 1.4391,
+      "step": 4644
+    },
+    {
+      "epoch": 2.0644444444444443,
+      "grad_norm": 2.703592300415039,
+      "learning_rate": 0.00011752669039145908,
+      "loss": 1.782,
+      "step": 4645
+    },
+    {
+      "epoch": 2.064888888888889,
+      "grad_norm": 2.8277475833892822,
+      "learning_rate": 0.00011750889679715302,
+      "loss": 2.0831,
+      "step": 4646
+    },
+    {
+      "epoch": 2.0653333333333332,
+      "grad_norm": 2.78928542137146,
+      "learning_rate": 0.00011749110320284697,
+      "loss": 1.3444,
+      "step": 4647
+    },
+    {
+      "epoch": 2.065777777777778,
+      "grad_norm": 2.8295578956604004,
+      "learning_rate": 0.00011747330960854093,
+      "loss": 1.4441,
+      "step": 4648
+    },
+    {
+      "epoch": 2.066222222222222,
+      "grad_norm": 1.9804134368896484,
+      "learning_rate": 0.00011745551601423487,
+      "loss": 0.8381,
+      "step": 4649
+    },
+    {
+      "epoch": 2.066666666666667,
+      "grad_norm": 3.729863166809082,
+      "learning_rate": 0.00011743772241992883,
+      "loss": 1.7128,
+      "step": 4650
+    },
+    {
+      "epoch": 2.067111111111111,
+      "grad_norm": 2.222588062286377,
+      "learning_rate": 0.00011741992882562279,
+      "loss": 0.9648,
+      "step": 4651
+    },
+    {
+      "epoch": 2.0675555555555554,
+      "grad_norm": 1.8874247074127197,
+      "learning_rate": 0.00011740213523131673,
+      "loss": 1.9391,
+      "step": 4652
+    },
+    {
+      "epoch": 2.068,
+      "grad_norm": 1.757788062095642,
+      "learning_rate": 0.00011738434163701069,
+      "loss": 1.8273,
+      "step": 4653
+    },
+    {
+      "epoch": 2.0684444444444443,
+      "grad_norm": 2.6937432289123535,
+      "learning_rate": 0.00011736654804270464,
+      "loss": 2.0317,
+      "step": 4654
+    },
+    {
+      "epoch": 2.068888888888889,
+      "grad_norm": 2.1685352325439453,
+      "learning_rate": 0.00011734875444839859,
+      "loss": 1.957,
+      "step": 4655
+    },
+    {
+      "epoch": 2.0693333333333332,
+      "grad_norm": 1.865443229675293,
+      "learning_rate": 0.00011733096085409254,
+      "loss": 1.8245,
+      "step": 4656
+    },
+    {
+      "epoch": 2.069777777777778,
+      "grad_norm": 1.8847582340240479,
+      "learning_rate": 0.00011731316725978647,
+      "loss": 2.0589,
+      "step": 4657
+    },
+    {
+      "epoch": 2.070222222222222,
+      "grad_norm": 2.063577890396118,
+      "learning_rate": 0.00011729537366548043,
+      "loss": 1.9999,
+      "step": 4658
+    },
+    {
+      "epoch": 2.070666666666667,
+      "grad_norm": 2.3595352172851562,
+      "learning_rate": 0.00011727758007117437,
+      "loss": 1.2466,
+      "step": 4659
+    },
+    {
+      "epoch": 2.071111111111111,
+      "grad_norm": 1.9096697568893433,
+      "learning_rate": 0.00011725978647686833,
+      "loss": 1.5026,
+      "step": 4660
+    },
+    {
+      "epoch": 2.0715555555555554,
+      "grad_norm": 2.064755439758301,
+      "learning_rate": 0.00011724199288256228,
+      "loss": 1.7534,
+      "step": 4661
+    },
+    {
+      "epoch": 2.072,
+      "grad_norm": 1.523712158203125,
+      "learning_rate": 0.00011722419928825623,
+      "loss": 1.2692,
+      "step": 4662
+    },
+    {
+      "epoch": 2.0724444444444443,
+      "grad_norm": 2.223828077316284,
+      "learning_rate": 0.00011720640569395018,
+      "loss": 1.8284,
+      "step": 4663
+    },
+    {
+      "epoch": 2.072888888888889,
+      "grad_norm": 2.1838390827178955,
+      "learning_rate": 0.00011718861209964414,
+      "loss": 1.6618,
+      "step": 4664
+    },
+    {
+      "epoch": 2.0733333333333333,
+      "grad_norm": 1.8496938943862915,
+      "learning_rate": 0.00011717081850533808,
+      "loss": 1.9343,
+      "step": 4665
+    },
+    {
+      "epoch": 2.073777777777778,
+      "grad_norm": 2.063234806060791,
+      "learning_rate": 0.00011715302491103204,
+      "loss": 1.4225,
+      "step": 4666
+    },
+    {
+      "epoch": 2.074222222222222,
+      "grad_norm": 2.388313055038452,
+      "learning_rate": 0.000117135231316726,
+      "loss": 1.5016,
+      "step": 4667
+    },
+    {
+      "epoch": 2.074666666666667,
+      "grad_norm": 2.2325823307037354,
+      "learning_rate": 0.00011711743772241994,
+      "loss": 1.7935,
+      "step": 4668
+    },
+    {
+      "epoch": 2.075111111111111,
+      "grad_norm": 1.600996732711792,
+      "learning_rate": 0.0001170996441281139,
+      "loss": 1.0136,
+      "step": 4669
+    },
+    {
+      "epoch": 2.0755555555555554,
+      "grad_norm": 2.17900013923645,
+      "learning_rate": 0.00011708185053380783,
+      "loss": 1.8612,
+      "step": 4670
+    },
+    {
+      "epoch": 2.076,
+      "grad_norm": 2.1239895820617676,
+      "learning_rate": 0.00011706405693950178,
+      "loss": 2.083,
+      "step": 4671
+    },
+    {
+      "epoch": 2.0764444444444443,
+      "grad_norm": 1.9274380207061768,
+      "learning_rate": 0.00011704626334519572,
+      "loss": 1.789,
+      "step": 4672
+    },
+    {
+      "epoch": 2.076888888888889,
+      "grad_norm": 2.4279444217681885,
+      "learning_rate": 0.00011702846975088968,
+      "loss": 2.0962,
+      "step": 4673
+    },
+    {
+      "epoch": 2.0773333333333333,
+      "grad_norm": 2.049731969833374,
+      "learning_rate": 0.00011701067615658364,
+      "loss": 1.529,
+      "step": 4674
+    },
+    {
+      "epoch": 2.077777777777778,
+      "grad_norm": 2.452834367752075,
+      "learning_rate": 0.00011699288256227758,
+      "loss": 1.5339,
+      "step": 4675
+    },
+    {
+      "epoch": 2.078222222222222,
+      "grad_norm": 2.3302857875823975,
+      "learning_rate": 0.00011697508896797154,
+      "loss": 1.8732,
+      "step": 4676
+    },
+    {
+      "epoch": 2.078666666666667,
+      "grad_norm": 2.1098620891571045,
+      "learning_rate": 0.0001169572953736655,
+      "loss": 1.4732,
+      "step": 4677
+    },
+    {
+      "epoch": 2.079111111111111,
+      "grad_norm": 2.499377489089966,
+      "learning_rate": 0.00011693950177935944,
+      "loss": 1.5989,
+      "step": 4678
+    },
+    {
+      "epoch": 2.0795555555555554,
+      "grad_norm": 2.321789503097534,
+      "learning_rate": 0.00011692170818505339,
+      "loss": 1.9255,
+      "step": 4679
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.7443134784698486,
+      "learning_rate": 0.00011690391459074735,
+      "loss": 0.6187,
+      "step": 4680
+    },
+    {
+      "epoch": 2.0804444444444443,
+      "grad_norm": 2.2512123584747314,
+      "learning_rate": 0.00011688612099644129,
+      "loss": 1.8004,
+      "step": 4681
+    },
+    {
+      "epoch": 2.080888888888889,
+      "grad_norm": 2.09832763671875,
+      "learning_rate": 0.00011686832740213525,
+      "loss": 1.3556,
+      "step": 4682
+    },
+    {
+      "epoch": 2.0813333333333333,
+      "grad_norm": 2.038501262664795,
+      "learning_rate": 0.00011685053380782918,
+      "loss": 1.1602,
+      "step": 4683
+    },
+    {
+      "epoch": 2.081777777777778,
+      "grad_norm": 2.386502504348755,
+      "learning_rate": 0.00011683274021352313,
+      "loss": 1.5816,
+      "step": 4684
+    },
+    {
+      "epoch": 2.082222222222222,
+      "grad_norm": 2.359687328338623,
+      "learning_rate": 0.00011681494661921708,
+      "loss": 1.7559,
+      "step": 4685
+    },
+    {
+      "epoch": 2.0826666666666664,
+      "grad_norm": 2.628209114074707,
+      "learning_rate": 0.00011679715302491103,
+      "loss": 1.7236,
+      "step": 4686
+    },
+    {
+      "epoch": 2.083111111111111,
+      "grad_norm": 2.598487615585327,
+      "learning_rate": 0.00011677935943060499,
+      "loss": 2.2871,
+      "step": 4687
+    },
+    {
+      "epoch": 2.0835555555555554,
+      "grad_norm": 2.430295944213867,
+      "learning_rate": 0.00011676156583629893,
+      "loss": 1.3882,
+      "step": 4688
+    },
+    {
+      "epoch": 2.084,
+      "grad_norm": 2.897634506225586,
+      "learning_rate": 0.00011674377224199289,
+      "loss": 1.9332,
+      "step": 4689
+    },
+    {
+      "epoch": 2.0844444444444443,
+      "grad_norm": 3.5412042140960693,
+      "learning_rate": 0.00011672597864768685,
+      "loss": 1.7709,
+      "step": 4690
+    },
+    {
+      "epoch": 2.084888888888889,
+      "grad_norm": 2.518477201461792,
+      "learning_rate": 0.00011670818505338079,
+      "loss": 1.8739,
+      "step": 4691
+    },
+    {
+      "epoch": 2.0853333333333333,
+      "grad_norm": 2.6702585220336914,
+      "learning_rate": 0.00011669039145907475,
+      "loss": 1.5968,
+      "step": 4692
+    },
+    {
+      "epoch": 2.085777777777778,
+      "grad_norm": 2.3972411155700684,
+      "learning_rate": 0.00011667259786476869,
+      "loss": 1.2457,
+      "step": 4693
+    },
+    {
+      "epoch": 2.086222222222222,
+      "grad_norm": 3.0265438556671143,
+      "learning_rate": 0.00011665480427046265,
+      "loss": 2.0656,
+      "step": 4694
+    },
+    {
+      "epoch": 2.086666666666667,
+      "grad_norm": 2.678575038909912,
+      "learning_rate": 0.0001166370106761566,
+      "loss": 1.3855,
+      "step": 4695
+    },
+    {
+      "epoch": 2.087111111111111,
+      "grad_norm": 2.815953016281128,
+      "learning_rate": 0.00011661921708185053,
+      "loss": 1.5628,
+      "step": 4696
+    },
+    {
+      "epoch": 2.0875555555555554,
+      "grad_norm": 3.07003116607666,
+      "learning_rate": 0.00011660142348754447,
+      "loss": 1.8748,
+      "step": 4697
+    },
+    {
+      "epoch": 2.088,
+      "grad_norm": 2.8302512168884277,
+      "learning_rate": 0.00011658362989323843,
+      "loss": 1.9202,
+      "step": 4698
+    },
+    {
+      "epoch": 2.0884444444444443,
+      "grad_norm": 2.839108467102051,
+      "learning_rate": 0.00011656583629893239,
+      "loss": 1.5236,
+      "step": 4699
+    },
+    {
+      "epoch": 2.088888888888889,
+      "grad_norm": 2.802971124649048,
+      "learning_rate": 0.00011654804270462633,
+      "loss": 1.2329,
+      "step": 4700
+    },
+    {
+      "epoch": 2.0893333333333333,
+      "grad_norm": 1.8777737617492676,
+      "learning_rate": 0.00011653024911032029,
+      "loss": 2.2772,
+      "step": 4701
+    },
+    {
+      "epoch": 2.089777777777778,
+      "grad_norm": 2.0970041751861572,
+      "learning_rate": 0.00011651245551601424,
+      "loss": 1.5844,
+      "step": 4702
+    },
+    {
+      "epoch": 2.090222222222222,
+      "grad_norm": 2.249812126159668,
+      "learning_rate": 0.00011649466192170819,
+      "loss": 1.9234,
+      "step": 4703
+    },
+    {
+      "epoch": 2.0906666666666665,
+      "grad_norm": 2.0751559734344482,
+      "learning_rate": 0.00011647686832740214,
+      "loss": 1.9922,
+      "step": 4704
+    },
+    {
+      "epoch": 2.091111111111111,
+      "grad_norm": 3.0285685062408447,
+      "learning_rate": 0.0001164590747330961,
+      "loss": 1.4631,
+      "step": 4705
+    },
+    {
+      "epoch": 2.0915555555555554,
+      "grad_norm": 1.9764230251312256,
+      "learning_rate": 0.00011644128113879004,
+      "loss": 1.3146,
+      "step": 4706
+    },
+    {
+      "epoch": 2.092,
+      "grad_norm": 2.0233638286590576,
+      "learning_rate": 0.000116423487544484,
+      "loss": 1.7718,
+      "step": 4707
+    },
+    {
+      "epoch": 2.0924444444444443,
+      "grad_norm": 2.1841344833374023,
+      "learning_rate": 0.00011640569395017796,
+      "loss": 2.2122,
+      "step": 4708
+    },
+    {
+      "epoch": 2.092888888888889,
+      "grad_norm": 2.00598406791687,
+      "learning_rate": 0.00011638790035587188,
+      "loss": 1.7922,
+      "step": 4709
+    },
+    {
+      "epoch": 2.0933333333333333,
+      "grad_norm": 2.4193332195281982,
+      "learning_rate": 0.00011637010676156583,
+      "loss": 1.916,
+      "step": 4710
+    },
+    {
+      "epoch": 2.093777777777778,
+      "grad_norm": 2.1774141788482666,
+      "learning_rate": 0.00011635231316725978,
+      "loss": 2.0158,
+      "step": 4711
+    },
+    {
+      "epoch": 2.094222222222222,
+      "grad_norm": 2.0492541790008545,
+      "learning_rate": 0.00011633451957295374,
+      "loss": 2.1842,
+      "step": 4712
+    },
+    {
+      "epoch": 2.0946666666666665,
+      "grad_norm": 2.008819818496704,
+      "learning_rate": 0.00011631672597864768,
+      "loss": 1.6566,
+      "step": 4713
+    },
+    {
+      "epoch": 2.095111111111111,
+      "grad_norm": 1.877214789390564,
+      "learning_rate": 0.00011629893238434164,
+      "loss": 1.9144,
+      "step": 4714
+    },
+    {
+      "epoch": 2.0955555555555554,
+      "grad_norm": 3.5000193119049072,
+      "learning_rate": 0.0001162811387900356,
+      "loss": 2.7184,
+      "step": 4715
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 2.240647554397583,
+      "learning_rate": 0.00011626334519572954,
+      "loss": 2.0603,
+      "step": 4716
+    },
+    {
+      "epoch": 2.0964444444444443,
+      "grad_norm": 1.966137409210205,
+      "learning_rate": 0.0001162455516014235,
+      "loss": 1.7733,
+      "step": 4717
+    },
+    {
+      "epoch": 2.096888888888889,
+      "grad_norm": 1.9724191427230835,
+      "learning_rate": 0.00011622775800711745,
+      "loss": 1.8949,
+      "step": 4718
+    },
+    {
+      "epoch": 2.0973333333333333,
+      "grad_norm": 2.057591199874878,
+      "learning_rate": 0.0001162099644128114,
+      "loss": 1.943,
+      "step": 4719
+    },
+    {
+      "epoch": 2.097777777777778,
+      "grad_norm": 2.395439624786377,
+      "learning_rate": 0.00011619217081850535,
+      "loss": 1.8616,
+      "step": 4720
+    },
+    {
+      "epoch": 2.098222222222222,
+      "grad_norm": 2.374725580215454,
+      "learning_rate": 0.00011617437722419931,
+      "loss": 1.5945,
+      "step": 4721
+    },
+    {
+      "epoch": 2.0986666666666665,
+      "grad_norm": 2.0969061851501465,
+      "learning_rate": 0.00011615658362989324,
+      "loss": 1.7964,
+      "step": 4722
+    },
+    {
+      "epoch": 2.099111111111111,
+      "grad_norm": 2.3275258541107178,
+      "learning_rate": 0.00011613879003558718,
+      "loss": 1.9125,
+      "step": 4723
+    },
+    {
+      "epoch": 2.0995555555555554,
+      "grad_norm": 2.1269102096557617,
+      "learning_rate": 0.00011612099644128114,
+      "loss": 1.2444,
+      "step": 4724
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.2072010040283203,
+      "learning_rate": 0.0001161032028469751,
+      "loss": 1.4687,
+      "step": 4725
+    },
+    {
+      "epoch": 2.1004444444444443,
+      "grad_norm": 2.1687400341033936,
+      "learning_rate": 0.00011608540925266904,
+      "loss": 1.8108,
+      "step": 4726
+    },
+    {
+      "epoch": 2.100888888888889,
+      "grad_norm": 2.358961820602417,
+      "learning_rate": 0.00011606761565836299,
+      "loss": 1.7005,
+      "step": 4727
+    },
+    {
+      "epoch": 2.1013333333333333,
+      "grad_norm": 2.413325786590576,
+      "learning_rate": 0.00011604982206405695,
+      "loss": 1.9044,
+      "step": 4728
+    },
+    {
+      "epoch": 2.101777777777778,
+      "grad_norm": 2.0373430252075195,
+      "learning_rate": 0.00011603202846975089,
+      "loss": 1.395,
+      "step": 4729
+    },
+    {
+      "epoch": 2.102222222222222,
+      "grad_norm": 2.424889326095581,
+      "learning_rate": 0.00011601423487544485,
+      "loss": 1.7084,
+      "step": 4730
+    },
+    {
+      "epoch": 2.1026666666666665,
+      "grad_norm": 2.3407599925994873,
+      "learning_rate": 0.0001159964412811388,
+      "loss": 1.3036,
+      "step": 4731
+    },
+    {
+      "epoch": 2.103111111111111,
+      "grad_norm": 2.332714080810547,
+      "learning_rate": 0.00011597864768683275,
+      "loss": 1.7091,
+      "step": 4732
+    },
+    {
+      "epoch": 2.1035555555555554,
+      "grad_norm": 2.412674903869629,
+      "learning_rate": 0.0001159608540925267,
+      "loss": 1.5195,
+      "step": 4733
+    },
+    {
+      "epoch": 2.104,
+      "grad_norm": 2.552485942840576,
+      "learning_rate": 0.00011594306049822066,
+      "loss": 1.2986,
+      "step": 4734
+    },
+    {
+      "epoch": 2.1044444444444443,
+      "grad_norm": 2.3344991207122803,
+      "learning_rate": 0.00011592526690391459,
+      "loss": 1.5909,
+      "step": 4735
+    },
+    {
+      "epoch": 2.104888888888889,
+      "grad_norm": 2.5788843631744385,
+      "learning_rate": 0.00011590747330960853,
+      "loss": 1.7849,
+      "step": 4736
+    },
+    {
+      "epoch": 2.1053333333333333,
+      "grad_norm": 2.2648143768310547,
+      "learning_rate": 0.00011588967971530249,
+      "loss": 1.688,
+      "step": 4737
+    },
+    {
+      "epoch": 2.105777777777778,
+      "grad_norm": 1.9284850358963013,
+      "learning_rate": 0.00011587188612099645,
+      "loss": 0.7604,
+      "step": 4738
+    },
+    {
+      "epoch": 2.106222222222222,
+      "grad_norm": 2.7157113552093506,
+      "learning_rate": 0.00011585409252669039,
+      "loss": 1.6809,
+      "step": 4739
+    },
+    {
+      "epoch": 2.1066666666666665,
+      "grad_norm": 2.454627752304077,
+      "learning_rate": 0.00011583629893238435,
+      "loss": 1.5076,
+      "step": 4740
+    },
+    {
+      "epoch": 2.107111111111111,
+      "grad_norm": 2.6363182067871094,
+      "learning_rate": 0.0001158185053380783,
+      "loss": 1.5113,
+      "step": 4741
+    },
+    {
+      "epoch": 2.1075555555555554,
+      "grad_norm": 2.9892001152038574,
+      "learning_rate": 0.00011580071174377225,
+      "loss": 1.453,
+      "step": 4742
+    },
+    {
+      "epoch": 2.108,
+      "grad_norm": 2.927412509918213,
+      "learning_rate": 0.0001157829181494662,
+      "loss": 1.7116,
+      "step": 4743
+    },
+    {
+      "epoch": 2.1084444444444443,
+      "grad_norm": 2.7504935264587402,
+      "learning_rate": 0.00011576512455516016,
+      "loss": 1.5148,
+      "step": 4744
+    },
+    {
+      "epoch": 2.108888888888889,
+      "grad_norm": 2.982175588607788,
+      "learning_rate": 0.0001157473309608541,
+      "loss": 1.7693,
+      "step": 4745
+    },
+    {
+      "epoch": 2.1093333333333333,
+      "grad_norm": 2.9303767681121826,
+      "learning_rate": 0.00011572953736654806,
+      "loss": 1.6931,
+      "step": 4746
+    },
+    {
+      "epoch": 2.109777777777778,
+      "grad_norm": 3.9879038333892822,
+      "learning_rate": 0.00011571174377224201,
+      "loss": 1.9336,
+      "step": 4747
+    },
+    {
+      "epoch": 2.110222222222222,
+      "grad_norm": 3.65157413482666,
+      "learning_rate": 0.00011569395017793594,
+      "loss": 1.7332,
+      "step": 4748
+    },
+    {
+      "epoch": 2.1106666666666665,
+      "grad_norm": 3.0835530757904053,
+      "learning_rate": 0.00011567615658362989,
+      "loss": 1.9205,
+      "step": 4749
+    },
+    {
+      "epoch": 2.111111111111111,
+      "grad_norm": 3.5538792610168457,
+      "learning_rate": 0.00011565836298932384,
+      "loss": 0.9,
+      "step": 4750
+    },
+    {
+      "epoch": 2.1115555555555554,
+      "grad_norm": 1.4265509843826294,
+      "learning_rate": 0.0001156405693950178,
+      "loss": 1.01,
+      "step": 4751
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 1.9871463775634766,
+      "learning_rate": 0.00011562277580071174,
+      "loss": 2.5466,
+      "step": 4752
+    },
+    {
+      "epoch": 2.1124444444444443,
+      "grad_norm": 2.88714337348938,
+      "learning_rate": 0.0001156049822064057,
+      "loss": 1.1221,
+      "step": 4753
+    },
+    {
+      "epoch": 2.112888888888889,
+      "grad_norm": 1.92192804813385,
+      "learning_rate": 0.00011558718861209966,
+      "loss": 1.7343,
+      "step": 4754
+    },
+    {
+      "epoch": 2.1133333333333333,
+      "grad_norm": 2.3003766536712646,
+      "learning_rate": 0.0001155693950177936,
+      "loss": 2.0344,
+      "step": 4755
+    },
+    {
+      "epoch": 2.113777777777778,
+      "grad_norm": 2.1846070289611816,
+      "learning_rate": 0.00011555160142348756,
+      "loss": 1.9369,
+      "step": 4756
+    },
+    {
+      "epoch": 2.1142222222222222,
+      "grad_norm": 2.2077274322509766,
+      "learning_rate": 0.00011553380782918151,
+      "loss": 1.8763,
+      "step": 4757
+    },
+    {
+      "epoch": 2.1146666666666665,
+      "grad_norm": 2.247318983078003,
+      "learning_rate": 0.00011551601423487545,
+      "loss": 1.6496,
+      "step": 4758
+    },
+    {
+      "epoch": 2.115111111111111,
+      "grad_norm": 2.483921527862549,
+      "learning_rate": 0.00011549822064056941,
+      "loss": 2.0657,
+      "step": 4759
+    },
+    {
+      "epoch": 2.1155555555555554,
+      "grad_norm": 2.4575605392456055,
+      "learning_rate": 0.00011548042704626334,
+      "loss": 1.9344,
+      "step": 4760
+    },
+    {
+      "epoch": 2.116,
+      "grad_norm": 2.4088830947875977,
+      "learning_rate": 0.0001154626334519573,
+      "loss": 1.8533,
+      "step": 4761
+    },
+    {
+      "epoch": 2.1164444444444444,
+      "grad_norm": 2.2537829875946045,
+      "learning_rate": 0.00011544483985765124,
+      "loss": 1.8034,
+      "step": 4762
+    },
+    {
+      "epoch": 2.116888888888889,
+      "grad_norm": 2.119475841522217,
+      "learning_rate": 0.0001154270462633452,
+      "loss": 1.9903,
+      "step": 4763
+    },
+    {
+      "epoch": 2.1173333333333333,
+      "grad_norm": 2.0730443000793457,
+      "learning_rate": 0.00011540925266903915,
+      "loss": 1.6459,
+      "step": 4764
+    },
+    {
+      "epoch": 2.117777777777778,
+      "grad_norm": 1.8100666999816895,
+      "learning_rate": 0.0001153914590747331,
+      "loss": 1.7129,
+      "step": 4765
+    },
+    {
+      "epoch": 2.1182222222222222,
+      "grad_norm": 1.7500770092010498,
+      "learning_rate": 0.00011537366548042705,
+      "loss": 1.5328,
+      "step": 4766
+    },
+    {
+      "epoch": 2.1186666666666665,
+      "grad_norm": 2.3155996799468994,
+      "learning_rate": 0.00011535587188612101,
+      "loss": 1.3889,
+      "step": 4767
+    },
+    {
+      "epoch": 2.119111111111111,
+      "grad_norm": 2.4421792030334473,
+      "learning_rate": 0.00011533807829181495,
+      "loss": 2.0543,
+      "step": 4768
+    },
+    {
+      "epoch": 2.1195555555555554,
+      "grad_norm": 2.4871792793273926,
+      "learning_rate": 0.00011532028469750891,
+      "loss": 2.2072,
+      "step": 4769
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.081446647644043,
+      "learning_rate": 0.00011530249110320286,
+      "loss": 1.4294,
+      "step": 4770
+    },
+    {
+      "epoch": 2.1204444444444444,
+      "grad_norm": 2.19559383392334,
+      "learning_rate": 0.00011528469750889681,
+      "loss": 1.9015,
+      "step": 4771
+    },
+    {
+      "epoch": 2.120888888888889,
+      "grad_norm": 2.453030586242676,
+      "learning_rate": 0.00011526690391459076,
+      "loss": 1.8428,
+      "step": 4772
+    },
+    {
+      "epoch": 2.1213333333333333,
+      "grad_norm": 2.7875795364379883,
+      "learning_rate": 0.0001152491103202847,
+      "loss": 1.5894,
+      "step": 4773
+    },
+    {
+      "epoch": 2.121777777777778,
+      "grad_norm": 2.395458459854126,
+      "learning_rate": 0.00011523131672597865,
+      "loss": 1.5671,
+      "step": 4774
+    },
+    {
+      "epoch": 2.1222222222222222,
+      "grad_norm": 2.4182310104370117,
+      "learning_rate": 0.0001152135231316726,
+      "loss": 1.6568,
+      "step": 4775
+    },
+    {
+      "epoch": 2.1226666666666665,
+      "grad_norm": 2.46071195602417,
+      "learning_rate": 0.00011519572953736655,
+      "loss": 1.8421,
+      "step": 4776
+    },
+    {
+      "epoch": 2.123111111111111,
+      "grad_norm": 2.5873522758483887,
+      "learning_rate": 0.0001151779359430605,
+      "loss": 1.5898,
+      "step": 4777
+    },
+    {
+      "epoch": 2.1235555555555554,
+      "grad_norm": 1.7765636444091797,
+      "learning_rate": 0.00011516014234875445,
+      "loss": 0.7924,
+      "step": 4778
+    },
+    {
+      "epoch": 2.124,
+      "grad_norm": 2.2877538204193115,
+      "learning_rate": 0.0001151423487544484,
+      "loss": 1.9232,
+      "step": 4779
+    },
+    {
+      "epoch": 2.1244444444444444,
+      "grad_norm": 2.445138692855835,
+      "learning_rate": 0.00011512455516014236,
+      "loss": 2.0002,
+      "step": 4780
+    },
+    {
+      "epoch": 2.124888888888889,
+      "grad_norm": 2.173436403274536,
+      "learning_rate": 0.0001151067615658363,
+      "loss": 1.6479,
+      "step": 4781
+    },
+    {
+      "epoch": 2.1253333333333333,
+      "grad_norm": 2.5610175132751465,
+      "learning_rate": 0.00011508896797153026,
+      "loss": 1.4669,
+      "step": 4782
+    },
+    {
+      "epoch": 2.1257777777777775,
+      "grad_norm": 2.4170966148376465,
+      "learning_rate": 0.0001150711743772242,
+      "loss": 1.7575,
+      "step": 4783
+    },
+    {
+      "epoch": 2.1262222222222222,
+      "grad_norm": 2.7330989837646484,
+      "learning_rate": 0.00011505338078291816,
+      "loss": 1.8737,
+      "step": 4784
+    },
+    {
+      "epoch": 2.1266666666666665,
+      "grad_norm": 2.4493215084075928,
+      "learning_rate": 0.00011503558718861212,
+      "loss": 1.2428,
+      "step": 4785
+    },
+    {
+      "epoch": 2.127111111111111,
+      "grad_norm": 2.54807710647583,
+      "learning_rate": 0.00011501779359430605,
+      "loss": 1.675,
+      "step": 4786
+    },
+    {
+      "epoch": 2.1275555555555554,
+      "grad_norm": 3.092026472091675,
+      "learning_rate": 0.00011499999999999999,
+      "loss": 1.8129,
+      "step": 4787
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 3.01973819732666,
+      "learning_rate": 0.00011498220640569395,
+      "loss": 1.9959,
+      "step": 4788
+    },
+    {
+      "epoch": 2.1284444444444444,
+      "grad_norm": 2.2665798664093018,
+      "learning_rate": 0.0001149644128113879,
+      "loss": 1.1747,
+      "step": 4789
+    },
+    {
+      "epoch": 2.128888888888889,
+      "grad_norm": 3.0983033180236816,
+      "learning_rate": 0.00011494661921708185,
+      "loss": 1.9808,
+      "step": 4790
+    },
+    {
+      "epoch": 2.1293333333333333,
+      "grad_norm": 2.7604308128356934,
+      "learning_rate": 0.0001149288256227758,
+      "loss": 1.7185,
+      "step": 4791
+    },
+    {
+      "epoch": 2.129777777777778,
+      "grad_norm": 2.712104558944702,
+      "learning_rate": 0.00011491103202846976,
+      "loss": 1.3634,
+      "step": 4792
+    },
+    {
+      "epoch": 2.1302222222222222,
+      "grad_norm": 2.7724192142486572,
+      "learning_rate": 0.0001148932384341637,
+      "loss": 1.2231,
+      "step": 4793
+    },
+    {
+      "epoch": 2.1306666666666665,
+      "grad_norm": 3.0472283363342285,
+      "learning_rate": 0.00011487544483985766,
+      "loss": 1.7844,
+      "step": 4794
+    },
+    {
+      "epoch": 2.131111111111111,
+      "grad_norm": 2.9061384201049805,
+      "learning_rate": 0.00011485765124555161,
+      "loss": 1.7998,
+      "step": 4795
+    },
+    {
+      "epoch": 2.1315555555555554,
+      "grad_norm": 3.207463502883911,
+      "learning_rate": 0.00011483985765124556,
+      "loss": 1.8822,
+      "step": 4796
+    },
+    {
+      "epoch": 2.132,
+      "grad_norm": 3.0004780292510986,
+      "learning_rate": 0.00011482206405693951,
+      "loss": 1.6778,
+      "step": 4797
+    },
+    {
+      "epoch": 2.1324444444444444,
+      "grad_norm": 3.055454730987549,
+      "learning_rate": 0.00011480427046263347,
+      "loss": 1.7763,
+      "step": 4798
+    },
+    {
+      "epoch": 2.132888888888889,
+      "grad_norm": 3.3208091259002686,
+      "learning_rate": 0.0001147864768683274,
+      "loss": 1.8971,
+      "step": 4799
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 2.958496570587158,
+      "learning_rate": 0.00011476868327402134,
+      "loss": 1.3253,
+      "step": 4800
+    },
+    {
+      "epoch": 2.1337777777777776,
+      "grad_norm": 1.970202922821045,
+      "learning_rate": 0.0001147508896797153,
+      "loss": 2.5294,
+      "step": 4801
+    },
+    {
+      "epoch": 2.1342222222222222,
+      "grad_norm": 1.807246446609497,
+      "learning_rate": 0.00011473309608540926,
+      "loss": 2.6724,
+      "step": 4802
+    },
+    {
+      "epoch": 2.1346666666666665,
+      "grad_norm": 1.7409089803695679,
+      "learning_rate": 0.0001147153024911032,
+      "loss": 1.8308,
+      "step": 4803
+    },
+    {
+      "epoch": 2.135111111111111,
+      "grad_norm": 1.3472347259521484,
+      "learning_rate": 0.00011469750889679716,
+      "loss": 1.1779,
+      "step": 4804
+    },
+    {
+      "epoch": 2.1355555555555554,
+      "grad_norm": 2.0794670581817627,
+      "learning_rate": 0.00011467971530249111,
+      "loss": 1.8861,
+      "step": 4805
+    },
+    {
+      "epoch": 2.136,
+      "grad_norm": 1.9827806949615479,
+      "learning_rate": 0.00011466192170818505,
+      "loss": 2.0878,
+      "step": 4806
+    },
+    {
+      "epoch": 2.1364444444444444,
+      "grad_norm": 1.8904576301574707,
+      "learning_rate": 0.00011464412811387901,
+      "loss": 1.7428,
+      "step": 4807
+    },
+    {
+      "epoch": 2.136888888888889,
+      "grad_norm": 1.7913202047348022,
+      "learning_rate": 0.00011462633451957297,
+      "loss": 1.2166,
+      "step": 4808
+    },
+    {
+      "epoch": 2.1373333333333333,
+      "grad_norm": 2.015347957611084,
+      "learning_rate": 0.00011460854092526691,
+      "loss": 1.8428,
+      "step": 4809
+    },
+    {
+      "epoch": 2.137777777777778,
+      "grad_norm": 2.0501961708068848,
+      "learning_rate": 0.00011459074733096087,
+      "loss": 2.1464,
+      "step": 4810
+    },
+    {
+      "epoch": 2.1382222222222222,
+      "grad_norm": 2.050212860107422,
+      "learning_rate": 0.00011457295373665482,
+      "loss": 2.0002,
+      "step": 4811
+    },
+    {
+      "epoch": 2.1386666666666665,
+      "grad_norm": 2.134122133255005,
+      "learning_rate": 0.00011455516014234875,
+      "loss": 2.1944,
+      "step": 4812
+    },
+    {
+      "epoch": 2.139111111111111,
+      "grad_norm": 1.8899743556976318,
+      "learning_rate": 0.0001145373665480427,
+      "loss": 1.6606,
+      "step": 4813
+    },
+    {
+      "epoch": 2.1395555555555554,
+      "grad_norm": 2.5020689964294434,
+      "learning_rate": 0.00011451957295373665,
+      "loss": 1.2598,
+      "step": 4814
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.9991487264633179,
+      "learning_rate": 0.00011450177935943061,
+      "loss": 1.9065,
+      "step": 4815
+    },
+    {
+      "epoch": 2.1404444444444444,
+      "grad_norm": 2.101409912109375,
+      "learning_rate": 0.00011448398576512455,
+      "loss": 2.0605,
+      "step": 4816
+    },
+    {
+      "epoch": 2.140888888888889,
+      "grad_norm": 1.888168454170227,
+      "learning_rate": 0.00011446619217081851,
+      "loss": 1.7381,
+      "step": 4817
+    },
+    {
+      "epoch": 2.1413333333333333,
+      "grad_norm": 1.9398518800735474,
+      "learning_rate": 0.00011444839857651247,
+      "loss": 1.6413,
+      "step": 4818
+    },
+    {
+      "epoch": 2.1417777777777776,
+      "grad_norm": 2.3134799003601074,
+      "learning_rate": 0.00011443060498220641,
+      "loss": 1.6505,
+      "step": 4819
+    },
+    {
+      "epoch": 2.1422222222222222,
+      "grad_norm": 2.276531934738159,
+      "learning_rate": 0.00011441281138790036,
+      "loss": 2.3464,
+      "step": 4820
+    },
+    {
+      "epoch": 2.1426666666666665,
+      "grad_norm": 1.9262102842330933,
+      "learning_rate": 0.00011439501779359432,
+      "loss": 1.7171,
+      "step": 4821
+    },
+    {
+      "epoch": 2.143111111111111,
+      "grad_norm": 1.9879403114318848,
+      "learning_rate": 0.00011437722419928826,
+      "loss": 1.3029,
+      "step": 4822
+    },
+    {
+      "epoch": 2.1435555555555554,
+      "grad_norm": 2.1766903400421143,
+      "learning_rate": 0.00011435943060498222,
+      "loss": 1.7259,
+      "step": 4823
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 2.2997825145721436,
+      "learning_rate": 0.00011434163701067618,
+      "loss": 2.3198,
+      "step": 4824
+    },
+    {
+      "epoch": 2.1444444444444444,
+      "grad_norm": 2.1982157230377197,
+      "learning_rate": 0.0001143238434163701,
+      "loss": 1.913,
+      "step": 4825
+    },
+    {
+      "epoch": 2.144888888888889,
+      "grad_norm": 2.041598320007324,
+      "learning_rate": 0.00011430604982206405,
+      "loss": 1.473,
+      "step": 4826
+    },
+    {
+      "epoch": 2.1453333333333333,
+      "grad_norm": 2.7166502475738525,
+      "learning_rate": 0.000114288256227758,
+      "loss": 1.6377,
+      "step": 4827
+    },
+    {
+      "epoch": 2.145777777777778,
+      "grad_norm": 2.4108223915100098,
+      "learning_rate": 0.00011427046263345196,
+      "loss": 1.5567,
+      "step": 4828
+    },
+    {
+      "epoch": 2.1462222222222223,
+      "grad_norm": 2.168083429336548,
+      "learning_rate": 0.0001142526690391459,
+      "loss": 1.4504,
+      "step": 4829
+    },
+    {
+      "epoch": 2.1466666666666665,
+      "grad_norm": 2.364166498184204,
+      "learning_rate": 0.00011423487544483986,
+      "loss": 2.1665,
+      "step": 4830
+    },
+    {
+      "epoch": 2.147111111111111,
+      "grad_norm": 2.303101062774658,
+      "learning_rate": 0.00011421708185053382,
+      "loss": 1.7004,
+      "step": 4831
+    },
+    {
+      "epoch": 2.1475555555555554,
+      "grad_norm": 2.227083921432495,
+      "learning_rate": 0.00011419928825622776,
+      "loss": 0.8074,
+      "step": 4832
+    },
+    {
+      "epoch": 2.148,
+      "grad_norm": 2.1052846908569336,
+      "learning_rate": 0.00011418149466192172,
+      "loss": 1.6949,
+      "step": 4833
+    },
+    {
+      "epoch": 2.1484444444444444,
+      "grad_norm": 2.4847774505615234,
+      "learning_rate": 0.00011416370106761567,
+      "loss": 1.601,
+      "step": 4834
+    },
+    {
+      "epoch": 2.148888888888889,
+      "grad_norm": 1.6619476079940796,
+      "learning_rate": 0.00011414590747330962,
+      "loss": 0.8843,
+      "step": 4835
+    },
+    {
+      "epoch": 2.1493333333333333,
+      "grad_norm": 2.540987491607666,
+      "learning_rate": 0.00011412811387900357,
+      "loss": 1.5859,
+      "step": 4836
+    },
+    {
+      "epoch": 2.1497777777777776,
+      "grad_norm": 2.535325765609741,
+      "learning_rate": 0.00011411032028469753,
+      "loss": 1.8662,
+      "step": 4837
+    },
+    {
+      "epoch": 2.1502222222222223,
+      "grad_norm": 2.593508243560791,
+      "learning_rate": 0.00011409252669039146,
+      "loss": 1.5427,
+      "step": 4838
+    },
+    {
+      "epoch": 2.1506666666666665,
+      "grad_norm": 2.8630247116088867,
+      "learning_rate": 0.0001140747330960854,
+      "loss": 1.5968,
+      "step": 4839
+    },
+    {
+      "epoch": 2.151111111111111,
+      "grad_norm": 2.9739978313446045,
+      "learning_rate": 0.00011405693950177936,
+      "loss": 1.7853,
+      "step": 4840
+    },
+    {
+      "epoch": 2.1515555555555554,
+      "grad_norm": 2.3142545223236084,
+      "learning_rate": 0.00011403914590747332,
+      "loss": 1.4719,
+      "step": 4841
+    },
+    {
+      "epoch": 2.152,
+      "grad_norm": 2.915494203567505,
+      "learning_rate": 0.00011402135231316726,
+      "loss": 1.4225,
+      "step": 4842
+    },
+    {
+      "epoch": 2.1524444444444444,
+      "grad_norm": 2.4613630771636963,
+      "learning_rate": 0.00011400355871886121,
+      "loss": 1.8753,
+      "step": 4843
+    },
+    {
+      "epoch": 2.152888888888889,
+      "grad_norm": 3.434368133544922,
+      "learning_rate": 0.00011398576512455517,
+      "loss": 1.4475,
+      "step": 4844
+    },
+    {
+      "epoch": 2.1533333333333333,
+      "grad_norm": 3.4849135875701904,
+      "learning_rate": 0.00011396797153024911,
+      "loss": 2.0161,
+      "step": 4845
+    },
+    {
+      "epoch": 2.153777777777778,
+      "grad_norm": 2.9939815998077393,
+      "learning_rate": 0.00011395017793594307,
+      "loss": 1.7339,
+      "step": 4846
+    },
+    {
+      "epoch": 2.1542222222222223,
+      "grad_norm": 2.8146097660064697,
+      "learning_rate": 0.00011393238434163703,
+      "loss": 1.5392,
+      "step": 4847
+    },
+    {
+      "epoch": 2.1546666666666665,
+      "grad_norm": 4.10907506942749,
+      "learning_rate": 0.00011391459074733097,
+      "loss": 1.2541,
+      "step": 4848
+    },
+    {
+      "epoch": 2.155111111111111,
+      "grad_norm": 2.995576858520508,
+      "learning_rate": 0.00011389679715302493,
+      "loss": 1.668,
+      "step": 4849
+    },
+    {
+      "epoch": 2.1555555555555554,
+      "grad_norm": 0.45589274168014526,
+      "learning_rate": 0.00011387900355871888,
+      "loss": 0.0619,
+      "step": 4850
+    },
+    {
+      "epoch": 2.156,
+      "grad_norm": 1.8959424495697021,
+      "learning_rate": 0.00011386120996441281,
+      "loss": 2.3939,
+      "step": 4851
+    },
+    {
+      "epoch": 2.1564444444444444,
+      "grad_norm": 1.9614112377166748,
+      "learning_rate": 0.00011384341637010676,
+      "loss": 1.8642,
+      "step": 4852
+    },
+    {
+      "epoch": 2.156888888888889,
+      "grad_norm": 2.2174315452575684,
+      "learning_rate": 0.00011382562277580071,
+      "loss": 1.6341,
+      "step": 4853
+    },
+    {
+      "epoch": 2.1573333333333333,
+      "grad_norm": 2.617152452468872,
+      "learning_rate": 0.00011380782918149467,
+      "loss": 1.9807,
+      "step": 4854
+    },
+    {
+      "epoch": 2.1577777777777776,
+      "grad_norm": 1.9597722291946411,
+      "learning_rate": 0.00011379003558718861,
+      "loss": 1.5573,
+      "step": 4855
+    },
+    {
+      "epoch": 2.1582222222222223,
+      "grad_norm": 2.3918566703796387,
+      "learning_rate": 0.00011377224199288257,
+      "loss": 2.0388,
+      "step": 4856
+    },
+    {
+      "epoch": 2.1586666666666665,
+      "grad_norm": 2.0133910179138184,
+      "learning_rate": 0.00011375444839857652,
+      "loss": 2.0006,
+      "step": 4857
+    },
+    {
+      "epoch": 2.159111111111111,
+      "grad_norm": 2.1372923851013184,
+      "learning_rate": 0.00011373665480427047,
+      "loss": 1.8971,
+      "step": 4858
+    },
+    {
+      "epoch": 2.1595555555555555,
+      "grad_norm": 2.200821876525879,
+      "learning_rate": 0.00011371886120996442,
+      "loss": 1.8821,
+      "step": 4859
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.4604291915893555,
+      "learning_rate": 0.00011370106761565838,
+      "loss": 1.9079,
+      "step": 4860
+    },
+    {
+      "epoch": 2.1604444444444444,
+      "grad_norm": 2.0270156860351562,
+      "learning_rate": 0.00011368327402135232,
+      "loss": 1.4849,
+      "step": 4861
+    },
+    {
+      "epoch": 2.160888888888889,
+      "grad_norm": 2.796276807785034,
+      "learning_rate": 0.00011366548042704628,
+      "loss": 2.0916,
+      "step": 4862
+    },
+    {
+      "epoch": 2.1613333333333333,
+      "grad_norm": 2.0537173748016357,
+      "learning_rate": 0.00011364768683274024,
+      "loss": 1.7261,
+      "step": 4863
+    },
+    {
+      "epoch": 2.1617777777777776,
+      "grad_norm": 1.9119226932525635,
+      "learning_rate": 0.00011362989323843417,
+      "loss": 1.5883,
+      "step": 4864
+    },
+    {
+      "epoch": 2.1622222222222223,
+      "grad_norm": 2.016460657119751,
+      "learning_rate": 0.00011361209964412811,
+      "loss": 2.0473,
+      "step": 4865
+    },
+    {
+      "epoch": 2.1626666666666665,
+      "grad_norm": 2.056744337081909,
+      "learning_rate": 0.00011359430604982207,
+      "loss": 1.5667,
+      "step": 4866
+    },
+    {
+      "epoch": 2.163111111111111,
+      "grad_norm": 2.0856573581695557,
+      "learning_rate": 0.00011357651245551602,
+      "loss": 2.1636,
+      "step": 4867
+    },
+    {
+      "epoch": 2.1635555555555555,
+      "grad_norm": 2.0556750297546387,
+      "learning_rate": 0.00011355871886120996,
+      "loss": 1.6923,
+      "step": 4868
+    },
+    {
+      "epoch": 2.164,
+      "grad_norm": 1.8435472249984741,
+      "learning_rate": 0.00011354092526690392,
+      "loss": 1.4836,
+      "step": 4869
+    },
+    {
+      "epoch": 2.1644444444444444,
+      "grad_norm": 1.7105246782302856,
+      "learning_rate": 0.00011352313167259788,
+      "loss": 1.2481,
+      "step": 4870
+    },
+    {
+      "epoch": 2.164888888888889,
+      "grad_norm": 1.367600679397583,
+      "learning_rate": 0.00011350533807829182,
+      "loss": 0.6025,
+      "step": 4871
+    },
+    {
+      "epoch": 2.1653333333333333,
+      "grad_norm": 2.135531425476074,
+      "learning_rate": 0.00011348754448398578,
+      "loss": 1.6163,
+      "step": 4872
+    },
+    {
+      "epoch": 2.1657777777777776,
+      "grad_norm": 1.9856308698654175,
+      "learning_rate": 0.00011346975088967972,
+      "loss": 1.6504,
+      "step": 4873
+    },
+    {
+      "epoch": 2.1662222222222223,
+      "grad_norm": 2.070834159851074,
+      "learning_rate": 0.00011345195729537368,
+      "loss": 1.7144,
+      "step": 4874
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 2.081918716430664,
+      "learning_rate": 0.00011343416370106763,
+      "loss": 1.6878,
+      "step": 4875
+    },
+    {
+      "epoch": 2.167111111111111,
+      "grad_norm": 2.117114305496216,
+      "learning_rate": 0.00011341637010676156,
+      "loss": 1.5885,
+      "step": 4876
+    },
+    {
+      "epoch": 2.1675555555555555,
+      "grad_norm": 2.200413227081299,
+      "learning_rate": 0.0001133985765124555,
+      "loss": 1.4892,
+      "step": 4877
+    },
+    {
+      "epoch": 2.168,
+      "grad_norm": 2.2609503269195557,
+      "learning_rate": 0.00011338078291814946,
+      "loss": 1.5828,
+      "step": 4878
+    },
+    {
+      "epoch": 2.1684444444444444,
+      "grad_norm": 2.2752559185028076,
+      "learning_rate": 0.00011336298932384342,
+      "loss": 1.6549,
+      "step": 4879
+    },
+    {
+      "epoch": 2.168888888888889,
+      "grad_norm": 2.138413667678833,
+      "learning_rate": 0.00011334519572953736,
+      "loss": 1.241,
+      "step": 4880
+    },
+    {
+      "epoch": 2.1693333333333333,
+      "grad_norm": 2.5910072326660156,
+      "learning_rate": 0.00011332740213523132,
+      "loss": 1.4842,
+      "step": 4881
+    },
+    {
+      "epoch": 2.1697777777777776,
+      "grad_norm": 2.1112563610076904,
+      "learning_rate": 0.00011330960854092527,
+      "loss": 1.5217,
+      "step": 4882
+    },
+    {
+      "epoch": 2.1702222222222223,
+      "grad_norm": 2.3650734424591064,
+      "learning_rate": 0.00011329181494661922,
+      "loss": 1.6714,
+      "step": 4883
+    },
+    {
+      "epoch": 2.1706666666666665,
+      "grad_norm": 2.035025119781494,
+      "learning_rate": 0.00011327402135231317,
+      "loss": 1.2853,
+      "step": 4884
+    },
+    {
+      "epoch": 2.171111111111111,
+      "grad_norm": 2.171036958694458,
+      "learning_rate": 0.00011325622775800713,
+      "loss": 1.4231,
+      "step": 4885
+    },
+    {
+      "epoch": 2.1715555555555555,
+      "grad_norm": 3.245302200317383,
+      "learning_rate": 0.00011323843416370107,
+      "loss": 1.7397,
+      "step": 4886
+    },
+    {
+      "epoch": 2.172,
+      "grad_norm": 3.072438955307007,
+      "learning_rate": 0.00011322064056939503,
+      "loss": 2.2672,
+      "step": 4887
+    },
+    {
+      "epoch": 2.1724444444444444,
+      "grad_norm": 2.8883869647979736,
+      "learning_rate": 0.00011320284697508899,
+      "loss": 1.7494,
+      "step": 4888
+    },
+    {
+      "epoch": 2.172888888888889,
+      "grad_norm": 2.5595476627349854,
+      "learning_rate": 0.00011318505338078292,
+      "loss": 1.3625,
+      "step": 4889
+    },
+    {
+      "epoch": 2.1733333333333333,
+      "grad_norm": 2.532860517501831,
+      "learning_rate": 0.00011316725978647686,
+      "loss": 1.0884,
+      "step": 4890
+    },
+    {
+      "epoch": 2.1737777777777776,
+      "grad_norm": 3.10447359085083,
+      "learning_rate": 0.00011314946619217082,
+      "loss": 2.1665,
+      "step": 4891
+    },
+    {
+      "epoch": 2.1742222222222223,
+      "grad_norm": 2.6869397163391113,
+      "learning_rate": 0.00011313167259786477,
+      "loss": 1.644,
+      "step": 4892
+    },
+    {
+      "epoch": 2.1746666666666665,
+      "grad_norm": 2.8844075202941895,
+      "learning_rate": 0.00011311387900355871,
+      "loss": 1.5886,
+      "step": 4893
+    },
+    {
+      "epoch": 2.175111111111111,
+      "grad_norm": 2.6155810356140137,
+      "learning_rate": 0.00011309608540925267,
+      "loss": 1.4173,
+      "step": 4894
+    },
+    {
+      "epoch": 2.1755555555555555,
+      "grad_norm": 3.163278341293335,
+      "learning_rate": 0.00011307829181494663,
+      "loss": 1.5141,
+      "step": 4895
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 2.6014788150787354,
+      "learning_rate": 0.00011306049822064057,
+      "loss": 1.3882,
+      "step": 4896
+    },
+    {
+      "epoch": 2.1764444444444444,
+      "grad_norm": 3.472792387008667,
+      "learning_rate": 0.00011304270462633453,
+      "loss": 1.8972,
+      "step": 4897
+    },
+    {
+      "epoch": 2.176888888888889,
+      "grad_norm": 2.989288568496704,
+      "learning_rate": 0.00011302491103202848,
+      "loss": 1.7598,
+      "step": 4898
+    },
+    {
+      "epoch": 2.1773333333333333,
+      "grad_norm": 3.674435615539551,
+      "learning_rate": 0.00011300711743772243,
+      "loss": 1.5909,
+      "step": 4899
+    },
+    {
+      "epoch": 2.1777777777777776,
+      "grad_norm": 2.7188007831573486,
+      "learning_rate": 0.00011298932384341638,
+      "loss": 0.8713,
+      "step": 4900
+    },
+    {
+      "epoch": 2.1782222222222223,
+      "grad_norm": 1.8986291885375977,
+      "learning_rate": 0.00011297153024911034,
+      "loss": 2.2709,
+      "step": 4901
+    },
+    {
+      "epoch": 2.1786666666666665,
+      "grad_norm": 1.9453548192977905,
+      "learning_rate": 0.00011295373665480427,
+      "loss": 2.0065,
+      "step": 4902
+    },
+    {
+      "epoch": 2.179111111111111,
+      "grad_norm": 1.927510380744934,
+      "learning_rate": 0.00011293594306049821,
+      "loss": 2.1494,
+      "step": 4903
+    },
+    {
+      "epoch": 2.1795555555555555,
+      "grad_norm": 2.2795588970184326,
+      "learning_rate": 0.00011291814946619217,
+      "loss": 1.9723,
+      "step": 4904
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.4390902519226074,
+      "learning_rate": 0.00011290035587188612,
+      "loss": 2.4457,
+      "step": 4905
+    },
+    {
+      "epoch": 2.1804444444444444,
+      "grad_norm": 2.1050994396209717,
+      "learning_rate": 0.00011288256227758007,
+      "loss": 1.3167,
+      "step": 4906
+    },
+    {
+      "epoch": 2.180888888888889,
+      "grad_norm": 2.186591386795044,
+      "learning_rate": 0.00011286476868327402,
+      "loss": 1.9273,
+      "step": 4907
+    },
+    {
+      "epoch": 2.1813333333333333,
+      "grad_norm": 2.417327404022217,
+      "learning_rate": 0.00011284697508896798,
+      "loss": 1.9586,
+      "step": 4908
+    },
+    {
+      "epoch": 2.1817777777777776,
+      "grad_norm": 2.4069387912750244,
+      "learning_rate": 0.00011282918149466192,
+      "loss": 1.3856,
+      "step": 4909
+    },
+    {
+      "epoch": 2.1822222222222223,
+      "grad_norm": 2.5273990631103516,
+      "learning_rate": 0.00011281138790035588,
+      "loss": 1.8896,
+      "step": 4910
+    },
+    {
+      "epoch": 2.1826666666666665,
+      "grad_norm": 2.1475119590759277,
+      "learning_rate": 0.00011279359430604984,
+      "loss": 1.9951,
+      "step": 4911
+    },
+    {
+      "epoch": 2.1831111111111112,
+      "grad_norm": 2.5976321697235107,
+      "learning_rate": 0.00011277580071174378,
+      "loss": 1.6981,
+      "step": 4912
+    },
+    {
+      "epoch": 2.1835555555555555,
+      "grad_norm": 1.4629095792770386,
+      "learning_rate": 0.00011275800711743774,
+      "loss": 0.9384,
+      "step": 4913
+    },
+    {
+      "epoch": 2.184,
+      "grad_norm": 2.086716413497925,
+      "learning_rate": 0.00011274021352313169,
+      "loss": 1.5494,
+      "step": 4914
+    },
+    {
+      "epoch": 2.1844444444444444,
+      "grad_norm": 1.9124023914337158,
+      "learning_rate": 0.00011272241992882562,
+      "loss": 1.5172,
+      "step": 4915
+    },
+    {
+      "epoch": 2.1848888888888887,
+      "grad_norm": 2.046513080596924,
+      "learning_rate": 0.00011270462633451956,
+      "loss": 1.8116,
+      "step": 4916
+    },
+    {
+      "epoch": 2.1853333333333333,
+      "grad_norm": 2.0074009895324707,
+      "learning_rate": 0.00011268683274021352,
+      "loss": 1.2871,
+      "step": 4917
+    },
+    {
+      "epoch": 2.1857777777777776,
+      "grad_norm": 2.108560800552368,
+      "learning_rate": 0.00011266903914590748,
+      "loss": 1.8694,
+      "step": 4918
+    },
+    {
+      "epoch": 2.1862222222222223,
+      "grad_norm": 2.6234912872314453,
+      "learning_rate": 0.00011265124555160142,
+      "loss": 1.577,
+      "step": 4919
+    },
+    {
+      "epoch": 2.1866666666666665,
+      "grad_norm": 2.239198684692383,
+      "learning_rate": 0.00011263345195729538,
+      "loss": 2.0347,
+      "step": 4920
+    },
+    {
+      "epoch": 2.1871111111111112,
+      "grad_norm": 2.2819833755493164,
+      "learning_rate": 0.00011261565836298933,
+      "loss": 2.3103,
+      "step": 4921
+    },
+    {
+      "epoch": 2.1875555555555555,
+      "grad_norm": 2.2331910133361816,
+      "learning_rate": 0.00011259786476868328,
+      "loss": 1.6046,
+      "step": 4922
+    },
+    {
+      "epoch": 2.188,
+      "grad_norm": 2.1115944385528564,
+      "learning_rate": 0.00011258007117437723,
+      "loss": 1.8757,
+      "step": 4923
+    },
+    {
+      "epoch": 2.1884444444444444,
+      "grad_norm": 2.104067087173462,
+      "learning_rate": 0.00011256227758007119,
+      "loss": 1.6774,
+      "step": 4924
+    },
+    {
+      "epoch": 2.188888888888889,
+      "grad_norm": 2.385721206665039,
+      "learning_rate": 0.00011254448398576513,
+      "loss": 2.1468,
+      "step": 4925
+    },
+    {
+      "epoch": 2.1893333333333334,
+      "grad_norm": 2.554507255554199,
+      "learning_rate": 0.00011252669039145909,
+      "loss": 2.0144,
+      "step": 4926
+    },
+    {
+      "epoch": 2.1897777777777776,
+      "grad_norm": 2.23824143409729,
+      "learning_rate": 0.00011250889679715305,
+      "loss": 1.7106,
+      "step": 4927
+    },
+    {
+      "epoch": 2.1902222222222223,
+      "grad_norm": 2.3488128185272217,
+      "learning_rate": 0.00011249110320284698,
+      "loss": 1.8532,
+      "step": 4928
+    },
+    {
+      "epoch": 2.1906666666666665,
+      "grad_norm": 2.960286855697632,
+      "learning_rate": 0.00011247330960854092,
+      "loss": 1.6043,
+      "step": 4929
+    },
+    {
+      "epoch": 2.1911111111111112,
+      "grad_norm": 2.2605531215667725,
+      "learning_rate": 0.00011245551601423487,
+      "loss": 1.4308,
+      "step": 4930
+    },
+    {
+      "epoch": 2.1915555555555555,
+      "grad_norm": 2.7191028594970703,
+      "learning_rate": 0.00011243772241992883,
+      "loss": 1.4713,
+      "step": 4931
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 2.499677896499634,
+      "learning_rate": 0.00011241992882562277,
+      "loss": 1.4607,
+      "step": 4932
+    },
+    {
+      "epoch": 2.1924444444444444,
+      "grad_norm": 2.556260347366333,
+      "learning_rate": 0.00011240213523131673,
+      "loss": 1.4645,
+      "step": 4933
+    },
+    {
+      "epoch": 2.1928888888888887,
+      "grad_norm": 2.443964958190918,
+      "learning_rate": 0.00011238434163701069,
+      "loss": 1.3057,
+      "step": 4934
+    },
+    {
+      "epoch": 2.1933333333333334,
+      "grad_norm": 2.543210744857788,
+      "learning_rate": 0.00011236654804270463,
+      "loss": 2.0736,
+      "step": 4935
+    },
+    {
+      "epoch": 2.1937777777777776,
+      "grad_norm": 2.3704352378845215,
+      "learning_rate": 0.00011234875444839859,
+      "loss": 1.3827,
+      "step": 4936
+    },
+    {
+      "epoch": 2.1942222222222223,
+      "grad_norm": 2.49611234664917,
+      "learning_rate": 0.00011233096085409254,
+      "loss": 1.7015,
+      "step": 4937
+    },
+    {
+      "epoch": 2.1946666666666665,
+      "grad_norm": 2.5088584423065186,
+      "learning_rate": 0.00011231316725978649,
+      "loss": 1.7995,
+      "step": 4938
+    },
+    {
+      "epoch": 2.1951111111111112,
+      "grad_norm": 2.3908095359802246,
+      "learning_rate": 0.00011229537366548044,
+      "loss": 1.5912,
+      "step": 4939
+    },
+    {
+      "epoch": 2.1955555555555555,
+      "grad_norm": 2.597078323364258,
+      "learning_rate": 0.0001122775800711744,
+      "loss": 1.5497,
+      "step": 4940
+    },
+    {
+      "epoch": 2.196,
+      "grad_norm": 2.943357467651367,
+      "learning_rate": 0.00011225978647686833,
+      "loss": 2.0209,
+      "step": 4941
+    },
+    {
+      "epoch": 2.1964444444444444,
+      "grad_norm": 3.4762535095214844,
+      "learning_rate": 0.00011224199288256227,
+      "loss": 2.0581,
+      "step": 4942
+    },
+    {
+      "epoch": 2.196888888888889,
+      "grad_norm": 2.8447139263153076,
+      "learning_rate": 0.00011222419928825623,
+      "loss": 1.42,
+      "step": 4943
+    },
+    {
+      "epoch": 2.1973333333333334,
+      "grad_norm": 2.7726380825042725,
+      "learning_rate": 0.00011220640569395018,
+      "loss": 1.608,
+      "step": 4944
+    },
+    {
+      "epoch": 2.1977777777777776,
+      "grad_norm": 2.942455768585205,
+      "learning_rate": 0.00011218861209964413,
+      "loss": 1.6582,
+      "step": 4945
+    },
+    {
+      "epoch": 2.1982222222222223,
+      "grad_norm": 3.3560426235198975,
+      "learning_rate": 0.00011217081850533808,
+      "loss": 1.897,
+      "step": 4946
+    },
+    {
+      "epoch": 2.1986666666666665,
+      "grad_norm": 3.4621267318725586,
+      "learning_rate": 0.00011215302491103204,
+      "loss": 2.2004,
+      "step": 4947
+    },
+    {
+      "epoch": 2.1991111111111112,
+      "grad_norm": 2.5123441219329834,
+      "learning_rate": 0.00011213523131672598,
+      "loss": 1.0752,
+      "step": 4948
+    },
+    {
+      "epoch": 2.1995555555555555,
+      "grad_norm": 1.6749694347381592,
+      "learning_rate": 0.00011211743772241994,
+      "loss": 0.6226,
+      "step": 4949
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.319334030151367,
+      "learning_rate": 0.0001120996441281139,
+      "loss": 1.4356,
+      "step": 4950
+    },
+    {
+      "epoch": 2.2004444444444444,
+      "grad_norm": 1.5259418487548828,
+      "learning_rate": 0.00011208185053380784,
+      "loss": 1.6631,
+      "step": 4951
+    },
+    {
+      "epoch": 2.2008888888888887,
+      "grad_norm": 1.3049883842468262,
+      "learning_rate": 0.0001120640569395018,
+      "loss": 0.9819,
+      "step": 4952
+    },
+    {
+      "epoch": 2.2013333333333334,
+      "grad_norm": 2.1342673301696777,
+      "learning_rate": 0.00011204626334519575,
+      "loss": 1.929,
+      "step": 4953
+    },
+    {
+      "epoch": 2.2017777777777776,
+      "grad_norm": 1.989028811454773,
+      "learning_rate": 0.00011202846975088968,
+      "loss": 2.0505,
+      "step": 4954
+    },
+    {
+      "epoch": 2.2022222222222223,
+      "grad_norm": 2.049135208129883,
+      "learning_rate": 0.00011201067615658362,
+      "loss": 1.8818,
+      "step": 4955
+    },
+    {
+      "epoch": 2.2026666666666666,
+      "grad_norm": 2.263387441635132,
+      "learning_rate": 0.00011199288256227758,
+      "loss": 2.0014,
+      "step": 4956
+    },
+    {
+      "epoch": 2.2031111111111112,
+      "grad_norm": 2.11501407623291,
+      "learning_rate": 0.00011197508896797154,
+      "loss": 1.9057,
+      "step": 4957
+    },
+    {
+      "epoch": 2.2035555555555555,
+      "grad_norm": 1.9304909706115723,
+      "learning_rate": 0.00011195729537366548,
+      "loss": 1.867,
+      "step": 4958
+    },
+    {
+      "epoch": 2.204,
+      "grad_norm": 2.2471048831939697,
+      "learning_rate": 0.00011193950177935944,
+      "loss": 1.7576,
+      "step": 4959
+    },
+    {
+      "epoch": 2.2044444444444444,
+      "grad_norm": 2.4755563735961914,
+      "learning_rate": 0.0001119217081850534,
+      "loss": 1.6393,
+      "step": 4960
+    },
+    {
+      "epoch": 2.204888888888889,
+      "grad_norm": 2.0645923614501953,
+      "learning_rate": 0.00011190391459074734,
+      "loss": 1.4799,
+      "step": 4961
+    },
+    {
+      "epoch": 2.2053333333333334,
+      "grad_norm": 2.2446157932281494,
+      "learning_rate": 0.00011188612099644129,
+      "loss": 2.2068,
+      "step": 4962
+    },
+    {
+      "epoch": 2.2057777777777776,
+      "grad_norm": 2.309330701828003,
+      "learning_rate": 0.00011186832740213524,
+      "loss": 2.1778,
+      "step": 4963
+    },
+    {
+      "epoch": 2.2062222222222223,
+      "grad_norm": 1.9780677556991577,
+      "learning_rate": 0.00011185053380782919,
+      "loss": 1.8793,
+      "step": 4964
+    },
+    {
+      "epoch": 2.2066666666666666,
+      "grad_norm": 2.061521530151367,
+      "learning_rate": 0.00011183274021352315,
+      "loss": 1.2263,
+      "step": 4965
+    },
+    {
+      "epoch": 2.2071111111111112,
+      "grad_norm": 2.245497703552246,
+      "learning_rate": 0.00011181494661921709,
+      "loss": 1.9673,
+      "step": 4966
+    },
+    {
+      "epoch": 2.2075555555555555,
+      "grad_norm": 2.020643472671509,
+      "learning_rate": 0.00011179715302491102,
+      "loss": 1.9154,
+      "step": 4967
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 2.15903902053833,
+      "learning_rate": 0.00011177935943060498,
+      "loss": 1.8257,
+      "step": 4968
+    },
+    {
+      "epoch": 2.2084444444444444,
+      "grad_norm": 2.490280866622925,
+      "learning_rate": 0.00011176156583629893,
+      "loss": 1.9819,
+      "step": 4969
+    },
+    {
+      "epoch": 2.2088888888888887,
+      "grad_norm": 2.4871490001678467,
+      "learning_rate": 0.00011174377224199288,
+      "loss": 1.946,
+      "step": 4970
+    },
+    {
+      "epoch": 2.2093333333333334,
+      "grad_norm": 2.053659200668335,
+      "learning_rate": 0.00011172597864768683,
+      "loss": 1.8871,
+      "step": 4971
+    },
+    {
+      "epoch": 2.2097777777777776,
+      "grad_norm": 2.2057316303253174,
+      "learning_rate": 0.00011170818505338079,
+      "loss": 1.5619,
+      "step": 4972
+    },
+    {
+      "epoch": 2.2102222222222223,
+      "grad_norm": 1.9269914627075195,
+      "learning_rate": 0.00011169039145907473,
+      "loss": 1.3407,
+      "step": 4973
+    },
+    {
+      "epoch": 2.2106666666666666,
+      "grad_norm": 2.1930150985717773,
+      "learning_rate": 0.00011167259786476869,
+      "loss": 1.7184,
+      "step": 4974
+    },
+    {
+      "epoch": 2.2111111111111112,
+      "grad_norm": 2.1484861373901367,
+      "learning_rate": 0.00011165480427046265,
+      "loss": 1.8792,
+      "step": 4975
+    },
+    {
+      "epoch": 2.2115555555555555,
+      "grad_norm": 2.3635623455047607,
+      "learning_rate": 0.00011163701067615659,
+      "loss": 1.8599,
+      "step": 4976
+    },
+    {
+      "epoch": 2.212,
+      "grad_norm": 2.525038003921509,
+      "learning_rate": 0.00011161921708185055,
+      "loss": 1.9892,
+      "step": 4977
+    },
+    {
+      "epoch": 2.2124444444444444,
+      "grad_norm": 2.2718634605407715,
+      "learning_rate": 0.0001116014234875445,
+      "loss": 1.6329,
+      "step": 4978
+    },
+    {
+      "epoch": 2.2128888888888887,
+      "grad_norm": 1.9843499660491943,
+      "learning_rate": 0.00011158362989323844,
+      "loss": 1.2032,
+      "step": 4979
+    },
+    {
+      "epoch": 2.2133333333333334,
+      "grad_norm": 2.85638165473938,
+      "learning_rate": 0.00011156583629893237,
+      "loss": 2.0707,
+      "step": 4980
+    },
+    {
+      "epoch": 2.2137777777777776,
+      "grad_norm": 2.3343589305877686,
+      "learning_rate": 0.00011154804270462633,
+      "loss": 1.3783,
+      "step": 4981
+    },
+    {
+      "epoch": 2.2142222222222223,
+      "grad_norm": 2.3680782318115234,
+      "learning_rate": 0.00011153024911032029,
+      "loss": 1.4898,
+      "step": 4982
+    },
+    {
+      "epoch": 2.2146666666666666,
+      "grad_norm": 1.55228853225708,
+      "learning_rate": 0.00011151245551601423,
+      "loss": 0.6674,
+      "step": 4983
+    },
+    {
+      "epoch": 2.2151111111111113,
+      "grad_norm": 2.7255640029907227,
+      "learning_rate": 0.00011149466192170819,
+      "loss": 1.7476,
+      "step": 4984
+    },
+    {
+      "epoch": 2.2155555555555555,
+      "grad_norm": 2.6943490505218506,
+      "learning_rate": 0.00011147686832740214,
+      "loss": 1.5847,
+      "step": 4985
+    },
+    {
+      "epoch": 2.216,
+      "grad_norm": 2.6895389556884766,
+      "learning_rate": 0.00011145907473309609,
+      "loss": 1.8233,
+      "step": 4986
+    },
+    {
+      "epoch": 2.2164444444444444,
+      "grad_norm": 2.645171642303467,
+      "learning_rate": 0.00011144128113879004,
+      "loss": 1.7922,
+      "step": 4987
+    },
+    {
+      "epoch": 2.2168888888888887,
+      "grad_norm": 3.0169036388397217,
+      "learning_rate": 0.000111423487544484,
+      "loss": 1.8521,
+      "step": 4988
+    },
+    {
+      "epoch": 2.2173333333333334,
+      "grad_norm": 2.563162088394165,
+      "learning_rate": 0.00011140569395017794,
+      "loss": 1.654,
+      "step": 4989
+    },
+    {
+      "epoch": 2.2177777777777776,
+      "grad_norm": 2.9162724018096924,
+      "learning_rate": 0.0001113879003558719,
+      "loss": 1.8161,
+      "step": 4990
+    },
+    {
+      "epoch": 2.2182222222222223,
+      "grad_norm": 2.8812615871429443,
+      "learning_rate": 0.00011137010676156586,
+      "loss": 1.8481,
+      "step": 4991
+    },
+    {
+      "epoch": 2.2186666666666666,
+      "grad_norm": 3.00016188621521,
+      "learning_rate": 0.00011135231316725978,
+      "loss": 1.7091,
+      "step": 4992
+    },
+    {
+      "epoch": 2.2191111111111113,
+      "grad_norm": 2.6278302669525146,
+      "learning_rate": 0.00011133451957295373,
+      "loss": 1.7687,
+      "step": 4993
+    },
+    {
+      "epoch": 2.2195555555555555,
+      "grad_norm": 2.530658006668091,
+      "learning_rate": 0.00011131672597864768,
+      "loss": 1.6654,
+      "step": 4994
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.8739798069000244,
+      "learning_rate": 0.00011129893238434164,
+      "loss": 1.3435,
+      "step": 4995
+    },
+    {
+      "epoch": 2.2204444444444444,
+      "grad_norm": 2.9799602031707764,
+      "learning_rate": 0.00011128113879003558,
+      "loss": 1.642,
+      "step": 4996
+    },
+    {
+      "epoch": 2.2208888888888887,
+      "grad_norm": 2.8478050231933594,
+      "learning_rate": 0.00011126334519572954,
+      "loss": 1.6135,
+      "step": 4997
+    },
+    {
+      "epoch": 2.2213333333333334,
+      "grad_norm": 3.293814182281494,
+      "learning_rate": 0.0001112455516014235,
+      "loss": 1.8149,
+      "step": 4998
+    },
+    {
+      "epoch": 2.2217777777777776,
+      "grad_norm": 2.8683934211730957,
+      "learning_rate": 0.00011122775800711744,
+      "loss": 1.3027,
+      "step": 4999
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 3.7593607902526855,
+      "learning_rate": 0.0001112099644128114,
+      "loss": 1.6055,
+      "step": 5000
+    },
+    {
+      "epoch": 2.2226666666666666,
+      "grad_norm": 1.727084755897522,
+      "learning_rate": 0.00011119217081850535,
+      "loss": 1.8026,
+      "step": 5001
+    },
+    {
+      "epoch": 2.2231111111111113,
+      "grad_norm": 1.792447805404663,
+      "learning_rate": 0.0001111743772241993,
+      "loss": 1.9238,
+      "step": 5002
+    },
+    {
+      "epoch": 2.2235555555555555,
+      "grad_norm": 1.3714830875396729,
+      "learning_rate": 0.00011115658362989325,
+      "loss": 0.1835,
+      "step": 5003
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 2.177612066268921,
+      "learning_rate": 0.00011113879003558721,
+      "loss": 2.0674,
+      "step": 5004
+    },
+    {
+      "epoch": 2.2244444444444444,
+      "grad_norm": 2.2636287212371826,
+      "learning_rate": 0.00011112099644128114,
+      "loss": 2.0529,
+      "step": 5005
+    },
+    {
+      "epoch": 2.2248888888888887,
+      "grad_norm": 2.174473285675049,
+      "learning_rate": 0.00011110320284697508,
+      "loss": 1.7144,
+      "step": 5006
+    },
+    {
+      "epoch": 2.2253333333333334,
+      "grad_norm": 2.079563617706299,
+      "learning_rate": 0.00011108540925266904,
+      "loss": 1.7053,
+      "step": 5007
+    },
+    {
+      "epoch": 2.2257777777777776,
+      "grad_norm": 2.0495669841766357,
+      "learning_rate": 0.000111067615658363,
+      "loss": 1.4467,
+      "step": 5008
+    },
+    {
+      "epoch": 2.2262222222222223,
+      "grad_norm": 2.1744115352630615,
+      "learning_rate": 0.00011104982206405694,
+      "loss": 1.6289,
+      "step": 5009
+    },
+    {
+      "epoch": 2.2266666666666666,
+      "grad_norm": 2.245788335800171,
+      "learning_rate": 0.00011103202846975089,
+      "loss": 2.0664,
+      "step": 5010
+    },
+    {
+      "epoch": 2.2271111111111113,
+      "grad_norm": 2.4176290035247803,
+      "learning_rate": 0.00011101423487544485,
+      "loss": 1.986,
+      "step": 5011
+    },
+    {
+      "epoch": 2.2275555555555555,
+      "grad_norm": 2.397927761077881,
+      "learning_rate": 0.00011099644128113879,
+      "loss": 1.8439,
+      "step": 5012
+    },
+    {
+      "epoch": 2.228,
+      "grad_norm": 2.479949951171875,
+      "learning_rate": 0.00011097864768683275,
+      "loss": 1.7942,
+      "step": 5013
+    },
+    {
+      "epoch": 2.2284444444444444,
+      "grad_norm": 2.213899850845337,
+      "learning_rate": 0.0001109608540925267,
+      "loss": 1.8292,
+      "step": 5014
+    },
+    {
+      "epoch": 2.2288888888888887,
+      "grad_norm": 2.117074728012085,
+      "learning_rate": 0.00011094306049822065,
+      "loss": 1.8464,
+      "step": 5015
+    },
+    {
+      "epoch": 2.2293333333333334,
+      "grad_norm": 2.4568095207214355,
+      "learning_rate": 0.0001109252669039146,
+      "loss": 1.7872,
+      "step": 5016
+    },
+    {
+      "epoch": 2.2297777777777776,
+      "grad_norm": 2.0903635025024414,
+      "learning_rate": 0.00011090747330960856,
+      "loss": 1.4369,
+      "step": 5017
+    },
+    {
+      "epoch": 2.2302222222222223,
+      "grad_norm": 2.1136090755462646,
+      "learning_rate": 0.00011088967971530249,
+      "loss": 1.8091,
+      "step": 5018
+    },
+    {
+      "epoch": 2.2306666666666666,
+      "grad_norm": 1.9191817045211792,
+      "learning_rate": 0.00011087188612099643,
+      "loss": 1.9579,
+      "step": 5019
+    },
+    {
+      "epoch": 2.2311111111111113,
+      "grad_norm": 2.3440165519714355,
+      "learning_rate": 0.00011085409252669039,
+      "loss": 1.9823,
+      "step": 5020
+    },
+    {
+      "epoch": 2.2315555555555555,
+      "grad_norm": 2.085297107696533,
+      "learning_rate": 0.00011083629893238435,
+      "loss": 1.8584,
+      "step": 5021
+    },
+    {
+      "epoch": 2.232,
+      "grad_norm": 2.513420820236206,
+      "learning_rate": 0.00011081850533807829,
+      "loss": 2.2256,
+      "step": 5022
+    },
+    {
+      "epoch": 2.2324444444444445,
+      "grad_norm": 2.3045380115509033,
+      "learning_rate": 0.00011080071174377225,
+      "loss": 1.4638,
+      "step": 5023
+    },
+    {
+      "epoch": 2.2328888888888887,
+      "grad_norm": 2.3060178756713867,
+      "learning_rate": 0.0001107829181494662,
+      "loss": 1.8818,
+      "step": 5024
+    },
+    {
+      "epoch": 2.2333333333333334,
+      "grad_norm": 2.0273396968841553,
+      "learning_rate": 0.00011076512455516015,
+      "loss": 1.5125,
+      "step": 5025
+    },
+    {
+      "epoch": 2.2337777777777776,
+      "grad_norm": 2.183797597885132,
+      "learning_rate": 0.0001107473309608541,
+      "loss": 1.4318,
+      "step": 5026
+    },
+    {
+      "epoch": 2.2342222222222223,
+      "grad_norm": 2.43717098236084,
+      "learning_rate": 0.00011072953736654806,
+      "loss": 1.7804,
+      "step": 5027
+    },
+    {
+      "epoch": 2.2346666666666666,
+      "grad_norm": 2.276435375213623,
+      "learning_rate": 0.000110711743772242,
+      "loss": 1.5594,
+      "step": 5028
+    },
+    {
+      "epoch": 2.2351111111111113,
+      "grad_norm": 2.2554962635040283,
+      "learning_rate": 0.00011069395017793596,
+      "loss": 1.4094,
+      "step": 5029
+    },
+    {
+      "epoch": 2.2355555555555555,
+      "grad_norm": 2.3153090476989746,
+      "learning_rate": 0.00011067615658362991,
+      "loss": 1.3073,
+      "step": 5030
+    },
+    {
+      "epoch": 2.2359999999999998,
+      "grad_norm": 2.353789806365967,
+      "learning_rate": 0.00011065836298932384,
+      "loss": 1.5179,
+      "step": 5031
+    },
+    {
+      "epoch": 2.2364444444444445,
+      "grad_norm": 2.6622314453125,
+      "learning_rate": 0.00011064056939501779,
+      "loss": 1.9487,
+      "step": 5032
+    },
+    {
+      "epoch": 2.2368888888888887,
+      "grad_norm": 2.1066958904266357,
+      "learning_rate": 0.00011062277580071174,
+      "loss": 1.3317,
+      "step": 5033
+    },
+    {
+      "epoch": 2.2373333333333334,
+      "grad_norm": 2.2060763835906982,
+      "learning_rate": 0.0001106049822064057,
+      "loss": 0.9029,
+      "step": 5034
+    },
+    {
+      "epoch": 2.2377777777777776,
+      "grad_norm": 0.20760299265384674,
+      "learning_rate": 0.00011058718861209964,
+      "loss": 0.037,
+      "step": 5035
+    },
+    {
+      "epoch": 2.2382222222222223,
+      "grad_norm": 2.2269294261932373,
+      "learning_rate": 0.0001105693950177936,
+      "loss": 1.4259,
+      "step": 5036
+    },
+    {
+      "epoch": 2.2386666666666666,
+      "grad_norm": 2.511350631713867,
+      "learning_rate": 0.00011055160142348756,
+      "loss": 2.0862,
+      "step": 5037
+    },
+    {
+      "epoch": 2.2391111111111113,
+      "grad_norm": 2.4969749450683594,
+      "learning_rate": 0.0001105338078291815,
+      "loss": 1.4302,
+      "step": 5038
+    },
+    {
+      "epoch": 2.2395555555555555,
+      "grad_norm": 2.8733372688293457,
+      "learning_rate": 0.00011051601423487546,
+      "loss": 1.4739,
+      "step": 5039
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.5586845874786377,
+      "learning_rate": 0.00011049822064056941,
+      "loss": 1.6102,
+      "step": 5040
+    },
+    {
+      "epoch": 2.2404444444444445,
+      "grad_norm": 3.037733554840088,
+      "learning_rate": 0.00011048042704626335,
+      "loss": 1.6267,
+      "step": 5041
+    },
+    {
+      "epoch": 2.2408888888888887,
+      "grad_norm": 2.6196389198303223,
+      "learning_rate": 0.00011046263345195731,
+      "loss": 1.3955,
+      "step": 5042
+    },
+    {
+      "epoch": 2.2413333333333334,
+      "grad_norm": 2.608210325241089,
+      "learning_rate": 0.00011044483985765127,
+      "loss": 1.8584,
+      "step": 5043
+    },
+    {
+      "epoch": 2.2417777777777776,
+      "grad_norm": 2.4374005794525146,
+      "learning_rate": 0.0001104270462633452,
+      "loss": 1.1662,
+      "step": 5044
+    },
+    {
+      "epoch": 2.2422222222222223,
+      "grad_norm": 2.6647913455963135,
+      "learning_rate": 0.00011040925266903914,
+      "loss": 1.7048,
+      "step": 5045
+    },
+    {
+      "epoch": 2.2426666666666666,
+      "grad_norm": 3.091082811355591,
+      "learning_rate": 0.0001103914590747331,
+      "loss": 1.6592,
+      "step": 5046
+    },
+    {
+      "epoch": 2.2431111111111113,
+      "grad_norm": 3.078878164291382,
+      "learning_rate": 0.00011037366548042705,
+      "loss": 1.6601,
+      "step": 5047
+    },
+    {
+      "epoch": 2.2435555555555555,
+      "grad_norm": 2.757037401199341,
+      "learning_rate": 0.000110355871886121,
+      "loss": 1.3657,
+      "step": 5048
+    },
+    {
+      "epoch": 2.2439999999999998,
+      "grad_norm": 4.261408805847168,
+      "learning_rate": 0.00011033807829181495,
+      "loss": 1.2643,
+      "step": 5049
+    },
+    {
+      "epoch": 2.2444444444444445,
+      "grad_norm": 4.232182025909424,
+      "learning_rate": 0.00011032028469750891,
+      "loss": 1.125,
+      "step": 5050
+    },
+    {
+      "epoch": 2.2448888888888887,
+      "grad_norm": 1.8265682458877563,
+      "learning_rate": 0.00011030249110320285,
+      "loss": 2.396,
+      "step": 5051
+    },
+    {
+      "epoch": 2.2453333333333334,
+      "grad_norm": 2.1252670288085938,
+      "learning_rate": 0.00011028469750889681,
+      "loss": 2.0516,
+      "step": 5052
+    },
+    {
+      "epoch": 2.2457777777777777,
+      "grad_norm": 2.073094129562378,
+      "learning_rate": 0.00011026690391459075,
+      "loss": 2.2738,
+      "step": 5053
+    },
+    {
+      "epoch": 2.2462222222222223,
+      "grad_norm": 2.2928173542022705,
+      "learning_rate": 0.00011024911032028471,
+      "loss": 1.8305,
+      "step": 5054
+    },
+    {
+      "epoch": 2.2466666666666666,
+      "grad_norm": 2.276385545730591,
+      "learning_rate": 0.00011023131672597866,
+      "loss": 2.1467,
+      "step": 5055
+    },
+    {
+      "epoch": 2.2471111111111113,
+      "grad_norm": 1.8952126502990723,
+      "learning_rate": 0.00011021352313167261,
+      "loss": 1.6379,
+      "step": 5056
+    },
+    {
+      "epoch": 2.2475555555555555,
+      "grad_norm": 2.017606735229492,
+      "learning_rate": 0.00011019572953736654,
+      "loss": 2.0341,
+      "step": 5057
+    },
+    {
+      "epoch": 2.248,
+      "grad_norm": 2.055772542953491,
+      "learning_rate": 0.00011017793594306049,
+      "loss": 2.0079,
+      "step": 5058
+    },
+    {
+      "epoch": 2.2484444444444445,
+      "grad_norm": 2.386974334716797,
+      "learning_rate": 0.00011016014234875445,
+      "loss": 2.0906,
+      "step": 5059
+    },
+    {
+      "epoch": 2.2488888888888887,
+      "grad_norm": 2.5190494060516357,
+      "learning_rate": 0.00011014234875444839,
+      "loss": 0.848,
+      "step": 5060
+    },
+    {
+      "epoch": 2.2493333333333334,
+      "grad_norm": 1.9555598497390747,
+      "learning_rate": 0.00011012455516014235,
+      "loss": 1.8228,
+      "step": 5061
+    },
+    {
+      "epoch": 2.2497777777777777,
+      "grad_norm": 2.0160269737243652,
+      "learning_rate": 0.0001101067615658363,
+      "loss": 1.8248,
+      "step": 5062
+    },
+    {
+      "epoch": 2.2502222222222223,
+      "grad_norm": 1.9360467195510864,
+      "learning_rate": 0.00011008896797153025,
+      "loss": 1.5145,
+      "step": 5063
+    },
+    {
+      "epoch": 2.2506666666666666,
+      "grad_norm": 2.0160787105560303,
+      "learning_rate": 0.0001100711743772242,
+      "loss": 1.72,
+      "step": 5064
+    },
+    {
+      "epoch": 2.2511111111111113,
+      "grad_norm": 2.07167649269104,
+      "learning_rate": 0.00011005338078291816,
+      "loss": 1.5055,
+      "step": 5065
+    },
+    {
+      "epoch": 2.2515555555555555,
+      "grad_norm": 2.21230149269104,
+      "learning_rate": 0.0001100355871886121,
+      "loss": 1.9066,
+      "step": 5066
+    },
+    {
+      "epoch": 2.252,
+      "grad_norm": 1.537184476852417,
+      "learning_rate": 0.00011001779359430606,
+      "loss": 0.8466,
+      "step": 5067
+    },
+    {
+      "epoch": 2.2524444444444445,
+      "grad_norm": 1.7883505821228027,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 1.1587,
+      "step": 5068
+    },
+    {
+      "epoch": 2.2528888888888887,
+      "grad_norm": 2.396481990814209,
+      "learning_rate": 0.00010998220640569396,
+      "loss": 2.3124,
+      "step": 5069
+    },
+    {
+      "epoch": 2.2533333333333334,
+      "grad_norm": 2.4314522743225098,
+      "learning_rate": 0.00010996441281138789,
+      "loss": 1.9213,
+      "step": 5070
+    },
+    {
+      "epoch": 2.2537777777777777,
+      "grad_norm": 2.3176066875457764,
+      "learning_rate": 0.00010994661921708185,
+      "loss": 1.7673,
+      "step": 5071
+    },
+    {
+      "epoch": 2.2542222222222223,
+      "grad_norm": 2.6162097454071045,
+      "learning_rate": 0.0001099288256227758,
+      "loss": 1.8331,
+      "step": 5072
+    },
+    {
+      "epoch": 2.2546666666666666,
+      "grad_norm": 2.079402446746826,
+      "learning_rate": 0.00010991103202846975,
+      "loss": 1.7033,
+      "step": 5073
+    },
+    {
+      "epoch": 2.2551111111111113,
+      "grad_norm": 2.6413557529449463,
+      "learning_rate": 0.0001098932384341637,
+      "loss": 2.1742,
+      "step": 5074
+    },
+    {
+      "epoch": 2.2555555555555555,
+      "grad_norm": 2.4262351989746094,
+      "learning_rate": 0.00010987544483985766,
+      "loss": 1.8914,
+      "step": 5075
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 2.247837543487549,
+      "learning_rate": 0.0001098576512455516,
+      "loss": 1.7544,
+      "step": 5076
+    },
+    {
+      "epoch": 2.2564444444444445,
+      "grad_norm": 2.4542994499206543,
+      "learning_rate": 0.00010983985765124556,
+      "loss": 1.5624,
+      "step": 5077
+    },
+    {
+      "epoch": 2.2568888888888887,
+      "grad_norm": 2.1061196327209473,
+      "learning_rate": 0.00010982206405693951,
+      "loss": 1.4963,
+      "step": 5078
+    },
+    {
+      "epoch": 2.2573333333333334,
+      "grad_norm": 2.599663257598877,
+      "learning_rate": 0.00010980427046263346,
+      "loss": 2.1344,
+      "step": 5079
+    },
+    {
+      "epoch": 2.2577777777777777,
+      "grad_norm": 2.2885448932647705,
+      "learning_rate": 0.00010978647686832741,
+      "loss": 1.6295,
+      "step": 5080
+    },
+    {
+      "epoch": 2.2582222222222224,
+      "grad_norm": 2.3721494674682617,
+      "learning_rate": 0.00010976868327402137,
+      "loss": 1.7371,
+      "step": 5081
+    },
+    {
+      "epoch": 2.2586666666666666,
+      "grad_norm": 2.5894291400909424,
+      "learning_rate": 0.00010975088967971531,
+      "loss": 1.7291,
+      "step": 5082
+    },
+    {
+      "epoch": 2.2591111111111113,
+      "grad_norm": 2.5275352001190186,
+      "learning_rate": 0.00010973309608540924,
+      "loss": 1.6438,
+      "step": 5083
+    },
+    {
+      "epoch": 2.2595555555555555,
+      "grad_norm": 2.664001941680908,
+      "learning_rate": 0.0001097153024911032,
+      "loss": 1.561,
+      "step": 5084
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.6688332557678223,
+      "learning_rate": 0.00010969750889679716,
+      "loss": 1.6883,
+      "step": 5085
+    },
+    {
+      "epoch": 2.2604444444444445,
+      "grad_norm": 3.350391387939453,
+      "learning_rate": 0.0001096797153024911,
+      "loss": 1.7163,
+      "step": 5086
+    },
+    {
+      "epoch": 2.2608888888888887,
+      "grad_norm": 2.204253911972046,
+      "learning_rate": 0.00010966192170818506,
+      "loss": 1.3111,
+      "step": 5087
+    },
+    {
+      "epoch": 2.2613333333333334,
+      "grad_norm": 2.883998394012451,
+      "learning_rate": 0.00010964412811387901,
+      "loss": 2.3049,
+      "step": 5088
+    },
+    {
+      "epoch": 2.2617777777777777,
+      "grad_norm": 3.0937411785125732,
+      "learning_rate": 0.00010962633451957295,
+      "loss": 2.2407,
+      "step": 5089
+    },
+    {
+      "epoch": 2.2622222222222224,
+      "grad_norm": 2.6656394004821777,
+      "learning_rate": 0.00010960854092526691,
+      "loss": 1.4929,
+      "step": 5090
+    },
+    {
+      "epoch": 2.2626666666666666,
+      "grad_norm": 2.6025304794311523,
+      "learning_rate": 0.00010959074733096087,
+      "loss": 1.6213,
+      "step": 5091
+    },
+    {
+      "epoch": 2.2631111111111113,
+      "grad_norm": 2.685457706451416,
+      "learning_rate": 0.00010957295373665481,
+      "loss": 1.5779,
+      "step": 5092
+    },
+    {
+      "epoch": 2.2635555555555555,
+      "grad_norm": 2.978306531906128,
+      "learning_rate": 0.00010955516014234877,
+      "loss": 1.9549,
+      "step": 5093
+    },
+    {
+      "epoch": 2.2640000000000002,
+      "grad_norm": 3.0156309604644775,
+      "learning_rate": 0.00010953736654804272,
+      "loss": 1.837,
+      "step": 5094
+    },
+    {
+      "epoch": 2.2644444444444445,
+      "grad_norm": 2.647887706756592,
+      "learning_rate": 0.00010951957295373667,
+      "loss": 1.496,
+      "step": 5095
+    },
+    {
+      "epoch": 2.2648888888888887,
+      "grad_norm": 3.1577885150909424,
+      "learning_rate": 0.0001095017793594306,
+      "loss": 1.7638,
+      "step": 5096
+    },
+    {
+      "epoch": 2.2653333333333334,
+      "grad_norm": 3.2039244174957275,
+      "learning_rate": 0.00010948398576512455,
+      "loss": 1.8427,
+      "step": 5097
+    },
+    {
+      "epoch": 2.2657777777777777,
+      "grad_norm": 3.6146154403686523,
+      "learning_rate": 0.00010946619217081851,
+      "loss": 1.961,
+      "step": 5098
+    },
+    {
+      "epoch": 2.2662222222222224,
+      "grad_norm": 2.9038968086242676,
+      "learning_rate": 0.00010944839857651245,
+      "loss": 1.731,
+      "step": 5099
+    },
+    {
+      "epoch": 2.2666666666666666,
+      "grad_norm": 1.7584317922592163,
+      "learning_rate": 0.00010943060498220641,
+      "loss": 0.5946,
+      "step": 5100
+    },
+    {
+      "epoch": 2.2671111111111113,
+      "grad_norm": 1.3620789051055908,
+      "learning_rate": 0.00010941281138790037,
+      "loss": 1.1835,
+      "step": 5101
+    },
+    {
+      "epoch": 2.2675555555555555,
+      "grad_norm": 1.787398338317871,
+      "learning_rate": 0.00010939501779359431,
+      "loss": 2.3021,
+      "step": 5102
+    },
+    {
+      "epoch": 2.268,
+      "grad_norm": 2.0760881900787354,
+      "learning_rate": 0.00010937722419928826,
+      "loss": 2.3895,
+      "step": 5103
+    },
+    {
+      "epoch": 2.2684444444444445,
+      "grad_norm": 1.6183782815933228,
+      "learning_rate": 0.00010935943060498222,
+      "loss": 1.7555,
+      "step": 5104
+    },
+    {
+      "epoch": 2.2688888888888887,
+      "grad_norm": 2.2935304641723633,
+      "learning_rate": 0.00010934163701067616,
+      "loss": 2.1392,
+      "step": 5105
+    },
+    {
+      "epoch": 2.2693333333333334,
+      "grad_norm": 2.3295974731445312,
+      "learning_rate": 0.00010932384341637012,
+      "loss": 1.9799,
+      "step": 5106
+    },
+    {
+      "epoch": 2.2697777777777777,
+      "grad_norm": 2.144185781478882,
+      "learning_rate": 0.00010930604982206408,
+      "loss": 1.6646,
+      "step": 5107
+    },
+    {
+      "epoch": 2.2702222222222224,
+      "grad_norm": 1.8681271076202393,
+      "learning_rate": 0.000109288256227758,
+      "loss": 1.7238,
+      "step": 5108
+    },
+    {
+      "epoch": 2.2706666666666666,
+      "grad_norm": 1.9623098373413086,
+      "learning_rate": 0.00010927046263345195,
+      "loss": 1.5286,
+      "step": 5109
+    },
+    {
+      "epoch": 2.2711111111111113,
+      "grad_norm": 2.0515475273132324,
+      "learning_rate": 0.0001092526690391459,
+      "loss": 1.853,
+      "step": 5110
+    },
+    {
+      "epoch": 2.2715555555555556,
+      "grad_norm": 2.088438034057617,
+      "learning_rate": 0.00010923487544483986,
+      "loss": 1.907,
+      "step": 5111
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 2.4442410469055176,
+      "learning_rate": 0.0001092170818505338,
+      "loss": 2.044,
+      "step": 5112
+    },
+    {
+      "epoch": 2.2724444444444445,
+      "grad_norm": 2.190702199935913,
+      "learning_rate": 0.00010919928825622776,
+      "loss": 2.2709,
+      "step": 5113
+    },
+    {
+      "epoch": 2.2728888888888887,
+      "grad_norm": 2.703242778778076,
+      "learning_rate": 0.00010918149466192172,
+      "loss": 2.2026,
+      "step": 5114
+    },
+    {
+      "epoch": 2.2733333333333334,
+      "grad_norm": 2.1000590324401855,
+      "learning_rate": 0.00010916370106761566,
+      "loss": 1.552,
+      "step": 5115
+    },
+    {
+      "epoch": 2.2737777777777777,
+      "grad_norm": 2.5693249702453613,
+      "learning_rate": 0.00010914590747330962,
+      "loss": 2.2501,
+      "step": 5116
+    },
+    {
+      "epoch": 2.2742222222222224,
+      "grad_norm": 2.3385515213012695,
+      "learning_rate": 0.00010912811387900357,
+      "loss": 2.283,
+      "step": 5117
+    },
+    {
+      "epoch": 2.2746666666666666,
+      "grad_norm": 1.8569934368133545,
+      "learning_rate": 0.00010911032028469752,
+      "loss": 1.5728,
+      "step": 5118
+    },
+    {
+      "epoch": 2.2751111111111113,
+      "grad_norm": 2.1497058868408203,
+      "learning_rate": 0.00010909252669039147,
+      "loss": 1.6939,
+      "step": 5119
+    },
+    {
+      "epoch": 2.2755555555555556,
+      "grad_norm": 2.399209499359131,
+      "learning_rate": 0.00010907473309608543,
+      "loss": 2.1534,
+      "step": 5120
+    },
+    {
+      "epoch": 2.276,
+      "grad_norm": 2.810572624206543,
+      "learning_rate": 0.00010905693950177936,
+      "loss": 1.8042,
+      "step": 5121
+    },
+    {
+      "epoch": 2.2764444444444445,
+      "grad_norm": 2.422611713409424,
+      "learning_rate": 0.0001090391459074733,
+      "loss": 1.7979,
+      "step": 5122
+    },
+    {
+      "epoch": 2.2768888888888887,
+      "grad_norm": 2.5435221195220947,
+      "learning_rate": 0.00010902135231316726,
+      "loss": 1.8007,
+      "step": 5123
+    },
+    {
+      "epoch": 2.2773333333333334,
+      "grad_norm": 2.7877004146575928,
+      "learning_rate": 0.00010900355871886122,
+      "loss": 1.9082,
+      "step": 5124
+    },
+    {
+      "epoch": 2.2777777777777777,
+      "grad_norm": 2.4548027515411377,
+      "learning_rate": 0.00010898576512455516,
+      "loss": 1.4372,
+      "step": 5125
+    },
+    {
+      "epoch": 2.2782222222222224,
+      "grad_norm": 2.1647870540618896,
+      "learning_rate": 0.00010896797153024911,
+      "loss": 1.5144,
+      "step": 5126
+    },
+    {
+      "epoch": 2.2786666666666666,
+      "grad_norm": 2.2049877643585205,
+      "learning_rate": 0.00010895017793594307,
+      "loss": 1.278,
+      "step": 5127
+    },
+    {
+      "epoch": 2.279111111111111,
+      "grad_norm": 2.6493771076202393,
+      "learning_rate": 0.00010893238434163701,
+      "loss": 2.1244,
+      "step": 5128
+    },
+    {
+      "epoch": 2.2795555555555556,
+      "grad_norm": 2.0453760623931885,
+      "learning_rate": 0.00010891459074733097,
+      "loss": 0.8898,
+      "step": 5129
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 2.6239845752716064,
+      "learning_rate": 0.00010889679715302493,
+      "loss": 1.7447,
+      "step": 5130
+    },
+    {
+      "epoch": 2.2804444444444445,
+      "grad_norm": 1.8818942308425903,
+      "learning_rate": 0.00010887900355871887,
+      "loss": 0.8104,
+      "step": 5131
+    },
+    {
+      "epoch": 2.2808888888888887,
+      "grad_norm": 2.51438045501709,
+      "learning_rate": 0.00010886120996441283,
+      "loss": 1.4221,
+      "step": 5132
+    },
+    {
+      "epoch": 2.2813333333333334,
+      "grad_norm": 2.23756742477417,
+      "learning_rate": 0.00010884341637010678,
+      "loss": 1.3229,
+      "step": 5133
+    },
+    {
+      "epoch": 2.2817777777777777,
+      "grad_norm": 2.578627109527588,
+      "learning_rate": 0.00010882562277580071,
+      "loss": 1.679,
+      "step": 5134
+    },
+    {
+      "epoch": 2.2822222222222224,
+      "grad_norm": 2.4972848892211914,
+      "learning_rate": 0.00010880782918149466,
+      "loss": 1.0042,
+      "step": 5135
+    },
+    {
+      "epoch": 2.2826666666666666,
+      "grad_norm": 2.4864046573638916,
+      "learning_rate": 0.00010879003558718861,
+      "loss": 1.6573,
+      "step": 5136
+    },
+    {
+      "epoch": 2.2831111111111113,
+      "grad_norm": 1.8569329977035522,
+      "learning_rate": 0.00010877224199288257,
+      "loss": 0.6083,
+      "step": 5137
+    },
+    {
+      "epoch": 2.2835555555555556,
+      "grad_norm": 2.6191885471343994,
+      "learning_rate": 0.00010875444839857651,
+      "loss": 1.7911,
+      "step": 5138
+    },
+    {
+      "epoch": 2.284,
+      "grad_norm": 2.874465227127075,
+      "learning_rate": 0.00010873665480427047,
+      "loss": 1.9264,
+      "step": 5139
+    },
+    {
+      "epoch": 2.2844444444444445,
+      "grad_norm": 2.813215494155884,
+      "learning_rate": 0.00010871886120996441,
+      "loss": 1.6019,
+      "step": 5140
+    },
+    {
+      "epoch": 2.2848888888888887,
+      "grad_norm": 2.852952003479004,
+      "learning_rate": 0.00010870106761565837,
+      "loss": 2.0002,
+      "step": 5141
+    },
+    {
+      "epoch": 2.2853333333333334,
+      "grad_norm": 3.3380825519561768,
+      "learning_rate": 0.00010868327402135232,
+      "loss": 1.6685,
+      "step": 5142
+    },
+    {
+      "epoch": 2.2857777777777777,
+      "grad_norm": 2.840141534805298,
+      "learning_rate": 0.00010866548042704627,
+      "loss": 1.6248,
+      "step": 5143
+    },
+    {
+      "epoch": 2.2862222222222224,
+      "grad_norm": 2.88301420211792,
+      "learning_rate": 0.00010864768683274022,
+      "loss": 1.3737,
+      "step": 5144
+    },
+    {
+      "epoch": 2.2866666666666666,
+      "grad_norm": 3.1075048446655273,
+      "learning_rate": 0.00010862989323843418,
+      "loss": 1.678,
+      "step": 5145
+    },
+    {
+      "epoch": 2.287111111111111,
+      "grad_norm": 3.333651542663574,
+      "learning_rate": 0.00010861209964412812,
+      "loss": 2.0464,
+      "step": 5146
+    },
+    {
+      "epoch": 2.2875555555555556,
+      "grad_norm": 4.473095417022705,
+      "learning_rate": 0.00010859430604982205,
+      "loss": 1.6364,
+      "step": 5147
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 2.844916343688965,
+      "learning_rate": 0.00010857651245551601,
+      "loss": 1.8926,
+      "step": 5148
+    },
+    {
+      "epoch": 2.2884444444444445,
+      "grad_norm": 5.156615257263184,
+      "learning_rate": 0.00010855871886120997,
+      "loss": 1.9084,
+      "step": 5149
+    },
+    {
+      "epoch": 2.2888888888888888,
+      "grad_norm": 3.648468494415283,
+      "learning_rate": 0.00010854092526690391,
+      "loss": 0.7005,
+      "step": 5150
+    },
+    {
+      "epoch": 2.2893333333333334,
+      "grad_norm": 2.095301628112793,
+      "learning_rate": 0.00010852313167259786,
+      "loss": 1.952,
+      "step": 5151
+    },
+    {
+      "epoch": 2.2897777777777777,
+      "grad_norm": 1.9257947206497192,
+      "learning_rate": 0.00010850533807829182,
+      "loss": 1.4345,
+      "step": 5152
+    },
+    {
+      "epoch": 2.2902222222222224,
+      "grad_norm": 2.2104406356811523,
+      "learning_rate": 0.00010848754448398576,
+      "loss": 1.9751,
+      "step": 5153
+    },
+    {
+      "epoch": 2.2906666666666666,
+      "grad_norm": 1.96810781955719,
+      "learning_rate": 0.00010846975088967972,
+      "loss": 2.024,
+      "step": 5154
+    },
+    {
+      "epoch": 2.2911111111111113,
+      "grad_norm": 2.201118230819702,
+      "learning_rate": 0.00010845195729537368,
+      "loss": 1.8283,
+      "step": 5155
+    },
+    {
+      "epoch": 2.2915555555555556,
+      "grad_norm": 2.238771438598633,
+      "learning_rate": 0.00010843416370106762,
+      "loss": 1.6468,
+      "step": 5156
+    },
+    {
+      "epoch": 2.292,
+      "grad_norm": 2.4652698040008545,
+      "learning_rate": 0.00010841637010676158,
+      "loss": 2.0724,
+      "step": 5157
+    },
+    {
+      "epoch": 2.2924444444444445,
+      "grad_norm": 1.8881862163543701,
+      "learning_rate": 0.00010839857651245553,
+      "loss": 1.2482,
+      "step": 5158
+    },
+    {
+      "epoch": 2.2928888888888888,
+      "grad_norm": 2.0697972774505615,
+      "learning_rate": 0.00010838078291814948,
+      "loss": 2.2032,
+      "step": 5159
+    },
+    {
+      "epoch": 2.2933333333333334,
+      "grad_norm": 2.31854510307312,
+      "learning_rate": 0.0001083629893238434,
+      "loss": 1.9085,
+      "step": 5160
+    },
+    {
+      "epoch": 2.2937777777777777,
+      "grad_norm": 2.013096332550049,
+      "learning_rate": 0.00010834519572953736,
+      "loss": 1.2551,
+      "step": 5161
+    },
+    {
+      "epoch": 2.2942222222222224,
+      "grad_norm": 2.2634682655334473,
+      "learning_rate": 0.00010832740213523132,
+      "loss": 1.5858,
+      "step": 5162
+    },
+    {
+      "epoch": 2.2946666666666666,
+      "grad_norm": 2.027539014816284,
+      "learning_rate": 0.00010830960854092526,
+      "loss": 1.3721,
+      "step": 5163
+    },
+    {
+      "epoch": 2.295111111111111,
+      "grad_norm": 2.623805284500122,
+      "learning_rate": 0.00010829181494661922,
+      "loss": 2.0648,
+      "step": 5164
+    },
+    {
+      "epoch": 2.2955555555555556,
+      "grad_norm": 1.4081056118011475,
+      "learning_rate": 0.00010827402135231317,
+      "loss": 0.8012,
+      "step": 5165
+    },
+    {
+      "epoch": 2.296,
+      "grad_norm": 1.6935027837753296,
+      "learning_rate": 0.00010825622775800712,
+      "loss": 1.0633,
+      "step": 5166
+    },
+    {
+      "epoch": 2.2964444444444445,
+      "grad_norm": 2.5808827877044678,
+      "learning_rate": 0.00010823843416370107,
+      "loss": 1.3108,
+      "step": 5167
+    },
+    {
+      "epoch": 2.2968888888888888,
+      "grad_norm": 2.5715770721435547,
+      "learning_rate": 0.00010822064056939503,
+      "loss": 1.5686,
+      "step": 5168
+    },
+    {
+      "epoch": 2.2973333333333334,
+      "grad_norm": 2.925513505935669,
+      "learning_rate": 0.00010820284697508897,
+      "loss": 1.9211,
+      "step": 5169
+    },
+    {
+      "epoch": 2.2977777777777777,
+      "grad_norm": 2.3410723209381104,
+      "learning_rate": 0.00010818505338078293,
+      "loss": 1.7527,
+      "step": 5170
+    },
+    {
+      "epoch": 2.2982222222222224,
+      "grad_norm": 2.2690529823303223,
+      "learning_rate": 0.00010816725978647689,
+      "loss": 1.5536,
+      "step": 5171
+    },
+    {
+      "epoch": 2.2986666666666666,
+      "grad_norm": 2.3402512073516846,
+      "learning_rate": 0.00010814946619217083,
+      "loss": 1.6756,
+      "step": 5172
+    },
+    {
+      "epoch": 2.2991111111111113,
+      "grad_norm": 2.567690849304199,
+      "learning_rate": 0.00010813167259786476,
+      "loss": 2.1322,
+      "step": 5173
+    },
+    {
+      "epoch": 2.2995555555555556,
+      "grad_norm": 2.5583088397979736,
+      "learning_rate": 0.00010811387900355872,
+      "loss": 1.7518,
+      "step": 5174
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.2459707260131836,
+      "learning_rate": 0.00010809608540925267,
+      "loss": 1.4129,
+      "step": 5175
+    },
+    {
+      "epoch": 2.3004444444444445,
+      "grad_norm": 3.016340970993042,
+      "learning_rate": 0.00010807829181494661,
+      "loss": 1.8386,
+      "step": 5176
+    },
+    {
+      "epoch": 2.3008888888888888,
+      "grad_norm": 2.6692094802856445,
+      "learning_rate": 0.00010806049822064057,
+      "loss": 1.5559,
+      "step": 5177
+    },
+    {
+      "epoch": 2.3013333333333335,
+      "grad_norm": 2.3022642135620117,
+      "learning_rate": 0.00010804270462633453,
+      "loss": 1.4982,
+      "step": 5178
+    },
+    {
+      "epoch": 2.3017777777777777,
+      "grad_norm": 2.4145843982696533,
+      "learning_rate": 0.00010802491103202847,
+      "loss": 1.8678,
+      "step": 5179
+    },
+    {
+      "epoch": 2.3022222222222224,
+      "grad_norm": 2.2984983921051025,
+      "learning_rate": 0.00010800711743772243,
+      "loss": 1.606,
+      "step": 5180
+    },
+    {
+      "epoch": 2.3026666666666666,
+      "grad_norm": 2.6932058334350586,
+      "learning_rate": 0.00010798932384341638,
+      "loss": 1.5845,
+      "step": 5181
+    },
+    {
+      "epoch": 2.303111111111111,
+      "grad_norm": 2.256645441055298,
+      "learning_rate": 0.00010797153024911033,
+      "loss": 1.7118,
+      "step": 5182
+    },
+    {
+      "epoch": 2.3035555555555556,
+      "grad_norm": 2.5675642490386963,
+      "learning_rate": 0.00010795373665480428,
+      "loss": 1.3647,
+      "step": 5183
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 2.930628776550293,
+      "learning_rate": 0.00010793594306049824,
+      "loss": 1.4687,
+      "step": 5184
+    },
+    {
+      "epoch": 2.3044444444444445,
+      "grad_norm": 3.3996012210845947,
+      "learning_rate": 0.00010791814946619218,
+      "loss": 1.5153,
+      "step": 5185
+    },
+    {
+      "epoch": 2.3048888888888888,
+      "grad_norm": 4.883398532867432,
+      "learning_rate": 0.00010790035587188611,
+      "loss": 1.8317,
+      "step": 5186
+    },
+    {
+      "epoch": 2.3053333333333335,
+      "grad_norm": 2.2228214740753174,
+      "learning_rate": 0.00010788256227758007,
+      "loss": 1.1171,
+      "step": 5187
+    },
+    {
+      "epoch": 2.3057777777777777,
+      "grad_norm": 2.6813435554504395,
+      "learning_rate": 0.00010786476868327402,
+      "loss": 1.5585,
+      "step": 5188
+    },
+    {
+      "epoch": 2.3062222222222224,
+      "grad_norm": 2.6681740283966064,
+      "learning_rate": 0.00010784697508896797,
+      "loss": 1.6806,
+      "step": 5189
+    },
+    {
+      "epoch": 2.3066666666666666,
+      "grad_norm": 2.8957736492156982,
+      "learning_rate": 0.00010782918149466192,
+      "loss": 1.8597,
+      "step": 5190
+    },
+    {
+      "epoch": 2.3071111111111113,
+      "grad_norm": 2.954521656036377,
+      "learning_rate": 0.00010781138790035588,
+      "loss": 1.6892,
+      "step": 5191
+    },
+    {
+      "epoch": 2.3075555555555556,
+      "grad_norm": 3.0856049060821533,
+      "learning_rate": 0.00010779359430604982,
+      "loss": 1.5892,
+      "step": 5192
+    },
+    {
+      "epoch": 2.308,
+      "grad_norm": 3.5431087017059326,
+      "learning_rate": 0.00010777580071174378,
+      "loss": 1.9527,
+      "step": 5193
+    },
+    {
+      "epoch": 2.3084444444444445,
+      "grad_norm": 2.8147850036621094,
+      "learning_rate": 0.00010775800711743774,
+      "loss": 1.6474,
+      "step": 5194
+    },
+    {
+      "epoch": 2.3088888888888888,
+      "grad_norm": 2.9207980632781982,
+      "learning_rate": 0.00010774021352313168,
+      "loss": 1.7115,
+      "step": 5195
+    },
+    {
+      "epoch": 2.3093333333333335,
+      "grad_norm": 1.972179651260376,
+      "learning_rate": 0.00010772241992882564,
+      "loss": 0.7474,
+      "step": 5196
+    },
+    {
+      "epoch": 2.3097777777777777,
+      "grad_norm": 3.6194729804992676,
+      "learning_rate": 0.00010770462633451959,
+      "loss": 1.577,
+      "step": 5197
+    },
+    {
+      "epoch": 2.3102222222222224,
+      "grad_norm": 3.8124332427978516,
+      "learning_rate": 0.00010768683274021354,
+      "loss": 1.6077,
+      "step": 5198
+    },
+    {
+      "epoch": 2.3106666666666666,
+      "grad_norm": 4.198216438293457,
+      "learning_rate": 0.00010766903914590746,
+      "loss": 2.2859,
+      "step": 5199
+    },
+    {
+      "epoch": 2.311111111111111,
+      "grad_norm": 3.1311981678009033,
+      "learning_rate": 0.00010765124555160142,
+      "loss": 1.2557,
+      "step": 5200
+    },
+    {
+      "epoch": 2.3115555555555556,
+      "grad_norm": 1.368379831314087,
+      "learning_rate": 0.00010763345195729538,
+      "loss": 1.233,
+      "step": 5201
+    },
+    {
+      "epoch": 2.312,
+      "grad_norm": 2.020923376083374,
+      "learning_rate": 0.00010761565836298932,
+      "loss": 1.9648,
+      "step": 5202
+    },
+    {
+      "epoch": 2.3124444444444445,
+      "grad_norm": 2.088132619857788,
+      "learning_rate": 0.00010759786476868328,
+      "loss": 2.2838,
+      "step": 5203
+    },
+    {
+      "epoch": 2.3128888888888888,
+      "grad_norm": 2.1785659790039062,
+      "learning_rate": 0.00010758007117437723,
+      "loss": 2.1716,
+      "step": 5204
+    },
+    {
+      "epoch": 2.3133333333333335,
+      "grad_norm": 2.086555004119873,
+      "learning_rate": 0.00010756227758007118,
+      "loss": 1.9242,
+      "step": 5205
+    },
+    {
+      "epoch": 2.3137777777777777,
+      "grad_norm": 2.2909762859344482,
+      "learning_rate": 0.00010754448398576513,
+      "loss": 2.3151,
+      "step": 5206
+    },
+    {
+      "epoch": 2.3142222222222224,
+      "grad_norm": 2.3706085681915283,
+      "learning_rate": 0.00010752669039145909,
+      "loss": 2.0996,
+      "step": 5207
+    },
+    {
+      "epoch": 2.3146666666666667,
+      "grad_norm": 2.0024471282958984,
+      "learning_rate": 0.00010750889679715303,
+      "loss": 1.8437,
+      "step": 5208
+    },
+    {
+      "epoch": 2.3151111111111113,
+      "grad_norm": 2.22165584564209,
+      "learning_rate": 0.00010749110320284699,
+      "loss": 1.6399,
+      "step": 5209
+    },
+    {
+      "epoch": 2.3155555555555556,
+      "grad_norm": 2.4060609340667725,
+      "learning_rate": 0.00010747330960854095,
+      "loss": 1.4666,
+      "step": 5210
+    },
+    {
+      "epoch": 2.316,
+      "grad_norm": 0.6937952041625977,
+      "learning_rate": 0.00010745551601423489,
+      "loss": 0.0337,
+      "step": 5211
+    },
+    {
+      "epoch": 2.3164444444444445,
+      "grad_norm": 2.10048508644104,
+      "learning_rate": 0.00010743772241992882,
+      "loss": 2.2067,
+      "step": 5212
+    },
+    {
+      "epoch": 2.3168888888888888,
+      "grad_norm": 2.3263375759124756,
+      "learning_rate": 0.00010741992882562277,
+      "loss": 2.0357,
+      "step": 5213
+    },
+    {
+      "epoch": 2.3173333333333335,
+      "grad_norm": 1.9152144193649292,
+      "learning_rate": 0.00010740213523131673,
+      "loss": 1.9073,
+      "step": 5214
+    },
+    {
+      "epoch": 2.3177777777777777,
+      "grad_norm": 2.3582661151885986,
+      "learning_rate": 0.00010738434163701067,
+      "loss": 1.9668,
+      "step": 5215
+    },
+    {
+      "epoch": 2.3182222222222224,
+      "grad_norm": 2.1141693592071533,
+      "learning_rate": 0.00010736654804270463,
+      "loss": 1.6874,
+      "step": 5216
+    },
+    {
+      "epoch": 2.3186666666666667,
+      "grad_norm": 2.0202839374542236,
+      "learning_rate": 0.00010734875444839859,
+      "loss": 1.7318,
+      "step": 5217
+    },
+    {
+      "epoch": 2.319111111111111,
+      "grad_norm": 2.4021458625793457,
+      "learning_rate": 0.00010733096085409253,
+      "loss": 1.9969,
+      "step": 5218
+    },
+    {
+      "epoch": 2.3195555555555556,
+      "grad_norm": 1.3853273391723633,
+      "learning_rate": 0.00010731316725978649,
+      "loss": 0.8318,
+      "step": 5219
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6327898502349854,
+      "learning_rate": 0.00010729537366548044,
+      "loss": 1.3863,
+      "step": 5220
+    },
+    {
+      "epoch": 2.3204444444444445,
+      "grad_norm": 2.009892463684082,
+      "learning_rate": 0.00010727758007117439,
+      "loss": 1.5846,
+      "step": 5221
+    },
+    {
+      "epoch": 2.320888888888889,
+      "grad_norm": 2.0592620372772217,
+      "learning_rate": 0.00010725978647686834,
+      "loss": 1.6894,
+      "step": 5222
+    },
+    {
+      "epoch": 2.3213333333333335,
+      "grad_norm": 2.1385364532470703,
+      "learning_rate": 0.0001072419928825623,
+      "loss": 1.6496,
+      "step": 5223
+    },
+    {
+      "epoch": 2.3217777777777777,
+      "grad_norm": 2.383230209350586,
+      "learning_rate": 0.00010722419928825623,
+      "loss": 1.707,
+      "step": 5224
+    },
+    {
+      "epoch": 2.3222222222222224,
+      "grad_norm": 2.0533084869384766,
+      "learning_rate": 0.00010720640569395017,
+      "loss": 1.4989,
+      "step": 5225
+    },
+    {
+      "epoch": 2.3226666666666667,
+      "grad_norm": 2.2495386600494385,
+      "learning_rate": 0.00010718861209964413,
+      "loss": 1.7552,
+      "step": 5226
+    },
+    {
+      "epoch": 2.3231111111111113,
+      "grad_norm": 2.7372584342956543,
+      "learning_rate": 0.00010717081850533808,
+      "loss": 1.5627,
+      "step": 5227
+    },
+    {
+      "epoch": 2.3235555555555556,
+      "grad_norm": 2.1765596866607666,
+      "learning_rate": 0.00010715302491103203,
+      "loss": 1.2697,
+      "step": 5228
+    },
+    {
+      "epoch": 2.324,
+      "grad_norm": 2.3925368785858154,
+      "learning_rate": 0.00010713523131672598,
+      "loss": 1.8882,
+      "step": 5229
+    },
+    {
+      "epoch": 2.3244444444444445,
+      "grad_norm": 3.3124077320098877,
+      "learning_rate": 0.00010711743772241993,
+      "loss": 1.5998,
+      "step": 5230
+    },
+    {
+      "epoch": 2.324888888888889,
+      "grad_norm": 2.3769872188568115,
+      "learning_rate": 0.00010709964412811388,
+      "loss": 1.8299,
+      "step": 5231
+    },
+    {
+      "epoch": 2.3253333333333335,
+      "grad_norm": 3.1028130054473877,
+      "learning_rate": 0.00010708185053380784,
+      "loss": 2.4035,
+      "step": 5232
+    },
+    {
+      "epoch": 2.3257777777777777,
+      "grad_norm": 2.543940544128418,
+      "learning_rate": 0.00010706405693950178,
+      "loss": 1.2689,
+      "step": 5233
+    },
+    {
+      "epoch": 2.3262222222222224,
+      "grad_norm": 2.7707502841949463,
+      "learning_rate": 0.00010704626334519574,
+      "loss": 1.9459,
+      "step": 5234
+    },
+    {
+      "epoch": 2.3266666666666667,
+      "grad_norm": 2.408712863922119,
+      "learning_rate": 0.0001070284697508897,
+      "loss": 1.7291,
+      "step": 5235
+    },
+    {
+      "epoch": 2.327111111111111,
+      "grad_norm": 3.0360755920410156,
+      "learning_rate": 0.00010701067615658364,
+      "loss": 2.274,
+      "step": 5236
+    },
+    {
+      "epoch": 2.3275555555555556,
+      "grad_norm": 2.5077056884765625,
+      "learning_rate": 0.00010699288256227757,
+      "loss": 1.5385,
+      "step": 5237
+    },
+    {
+      "epoch": 2.328,
+      "grad_norm": 3.110048770904541,
+      "learning_rate": 0.00010697508896797152,
+      "loss": 1.7831,
+      "step": 5238
+    },
+    {
+      "epoch": 2.3284444444444445,
+      "grad_norm": 2.6556756496429443,
+      "learning_rate": 0.00010695729537366548,
+      "loss": 1.9516,
+      "step": 5239
+    },
+    {
+      "epoch": 2.328888888888889,
+      "grad_norm": 3.066831350326538,
+      "learning_rate": 0.00010693950177935942,
+      "loss": 1.9794,
+      "step": 5240
+    },
+    {
+      "epoch": 2.3293333333333335,
+      "grad_norm": 2.9944040775299072,
+      "learning_rate": 0.00010692170818505338,
+      "loss": 1.2826,
+      "step": 5241
+    },
+    {
+      "epoch": 2.3297777777777777,
+      "grad_norm": 2.9339687824249268,
+      "learning_rate": 0.00010690391459074734,
+      "loss": 1.6169,
+      "step": 5242
+    },
+    {
+      "epoch": 2.330222222222222,
+      "grad_norm": 2.9092490673065186,
+      "learning_rate": 0.00010688612099644128,
+      "loss": 1.6369,
+      "step": 5243
+    },
+    {
+      "epoch": 2.3306666666666667,
+      "grad_norm": 3.148411512374878,
+      "learning_rate": 0.00010686832740213524,
+      "loss": 1.7855,
+      "step": 5244
+    },
+    {
+      "epoch": 2.3311111111111114,
+      "grad_norm": 2.889481544494629,
+      "learning_rate": 0.00010685053380782919,
+      "loss": 1.8247,
+      "step": 5245
+    },
+    {
+      "epoch": 2.3315555555555556,
+      "grad_norm": 3.1188857555389404,
+      "learning_rate": 0.00010683274021352314,
+      "loss": 1.301,
+      "step": 5246
+    },
+    {
+      "epoch": 2.332,
+      "grad_norm": 2.914860486984253,
+      "learning_rate": 0.00010681494661921709,
+      "loss": 1.5213,
+      "step": 5247
+    },
+    {
+      "epoch": 2.3324444444444445,
+      "grad_norm": 3.193268060684204,
+      "learning_rate": 0.00010679715302491105,
+      "loss": 1.5691,
+      "step": 5248
+    },
+    {
+      "epoch": 2.332888888888889,
+      "grad_norm": 2.0643980503082275,
+      "learning_rate": 0.00010677935943060499,
+      "loss": 0.7877,
+      "step": 5249
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 3.4626474380493164,
+      "learning_rate": 0.00010676156583629892,
+      "loss": 1.5895,
+      "step": 5250
+    },
+    {
+      "epoch": 2.3337777777777777,
+      "grad_norm": 1.721472144126892,
+      "learning_rate": 0.00010674377224199288,
+      "loss": 1.8599,
+      "step": 5251
+    },
+    {
+      "epoch": 2.3342222222222224,
+      "grad_norm": 2.4358739852905273,
+      "learning_rate": 0.00010672597864768683,
+      "loss": 2.4073,
+      "step": 5252
+    },
+    {
+      "epoch": 2.3346666666666667,
+      "grad_norm": 2.0674357414245605,
+      "learning_rate": 0.00010670818505338078,
+      "loss": 1.4055,
+      "step": 5253
+    },
+    {
+      "epoch": 2.335111111111111,
+      "grad_norm": 1.9118478298187256,
+      "learning_rate": 0.00010669039145907473,
+      "loss": 1.3337,
+      "step": 5254
+    },
+    {
+      "epoch": 2.3355555555555556,
+      "grad_norm": 2.8397645950317383,
+      "learning_rate": 0.00010667259786476869,
+      "loss": 2.0588,
+      "step": 5255
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 2.2740516662597656,
+      "learning_rate": 0.00010665480427046263,
+      "loss": 1.9389,
+      "step": 5256
+    },
+    {
+      "epoch": 2.3364444444444445,
+      "grad_norm": 1.9148417711257935,
+      "learning_rate": 0.00010663701067615659,
+      "loss": 1.1185,
+      "step": 5257
+    },
+    {
+      "epoch": 2.336888888888889,
+      "grad_norm": 1.7196369171142578,
+      "learning_rate": 0.00010661921708185055,
+      "loss": 0.9252,
+      "step": 5258
+    },
+    {
+      "epoch": 2.3373333333333335,
+      "grad_norm": 2.5018043518066406,
+      "learning_rate": 0.00010660142348754449,
+      "loss": 2.1507,
+      "step": 5259
+    },
+    {
+      "epoch": 2.3377777777777777,
+      "grad_norm": 2.2617380619049072,
+      "learning_rate": 0.00010658362989323845,
+      "loss": 1.6547,
+      "step": 5260
+    },
+    {
+      "epoch": 2.338222222222222,
+      "grad_norm": 2.1879079341888428,
+      "learning_rate": 0.0001065658362989324,
+      "loss": 1.8825,
+      "step": 5261
+    },
+    {
+      "epoch": 2.3386666666666667,
+      "grad_norm": 2.5784716606140137,
+      "learning_rate": 0.00010654804270462634,
+      "loss": 2.1602,
+      "step": 5262
+    },
+    {
+      "epoch": 2.339111111111111,
+      "grad_norm": 2.2693116664886475,
+      "learning_rate": 0.00010653024911032027,
+      "loss": 1.818,
+      "step": 5263
+    },
+    {
+      "epoch": 2.3395555555555556,
+      "grad_norm": 2.2588093280792236,
+      "learning_rate": 0.00010651245551601423,
+      "loss": 2.0916,
+      "step": 5264
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.266087055206299,
+      "learning_rate": 0.00010649466192170819,
+      "loss": 1.8388,
+      "step": 5265
+    },
+    {
+      "epoch": 2.3404444444444445,
+      "grad_norm": 2.1709249019622803,
+      "learning_rate": 0.00010647686832740213,
+      "loss": 1.9512,
+      "step": 5266
+    },
+    {
+      "epoch": 2.340888888888889,
+      "grad_norm": 1.934448003768921,
+      "learning_rate": 0.00010645907473309609,
+      "loss": 1.5891,
+      "step": 5267
+    },
+    {
+      "epoch": 2.3413333333333335,
+      "grad_norm": 1.595613956451416,
+      "learning_rate": 0.00010644128113879004,
+      "loss": 0.9631,
+      "step": 5268
+    },
+    {
+      "epoch": 2.3417777777777777,
+      "grad_norm": 2.1244924068450928,
+      "learning_rate": 0.00010642348754448399,
+      "loss": 1.49,
+      "step": 5269
+    },
+    {
+      "epoch": 2.3422222222222224,
+      "grad_norm": 2.107415199279785,
+      "learning_rate": 0.00010640569395017794,
+      "loss": 1.2837,
+      "step": 5270
+    },
+    {
+      "epoch": 2.3426666666666667,
+      "grad_norm": 2.0290896892547607,
+      "learning_rate": 0.0001063879003558719,
+      "loss": 1.426,
+      "step": 5271
+    },
+    {
+      "epoch": 2.343111111111111,
+      "grad_norm": 1.9935804605484009,
+      "learning_rate": 0.00010637010676156584,
+      "loss": 1.6032,
+      "step": 5272
+    },
+    {
+      "epoch": 2.3435555555555556,
+      "grad_norm": 2.401331663131714,
+      "learning_rate": 0.0001063523131672598,
+      "loss": 1.745,
+      "step": 5273
+    },
+    {
+      "epoch": 2.344,
+      "grad_norm": 2.32720685005188,
+      "learning_rate": 0.00010633451957295375,
+      "loss": 2.1566,
+      "step": 5274
+    },
+    {
+      "epoch": 2.3444444444444446,
+      "grad_norm": 2.5163567066192627,
+      "learning_rate": 0.0001063167259786477,
+      "loss": 1.8647,
+      "step": 5275
+    },
+    {
+      "epoch": 2.344888888888889,
+      "grad_norm": 2.6666061878204346,
+      "learning_rate": 0.00010629893238434163,
+      "loss": 2.0087,
+      "step": 5276
+    },
+    {
+      "epoch": 2.3453333333333335,
+      "grad_norm": 2.972123622894287,
+      "learning_rate": 0.00010628113879003558,
+      "loss": 2.0574,
+      "step": 5277
+    },
+    {
+      "epoch": 2.3457777777777777,
+      "grad_norm": 2.4128878116607666,
+      "learning_rate": 0.00010626334519572954,
+      "loss": 1.4681,
+      "step": 5278
+    },
+    {
+      "epoch": 2.346222222222222,
+      "grad_norm": 2.2778286933898926,
+      "learning_rate": 0.00010624555160142348,
+      "loss": 1.6174,
+      "step": 5279
+    },
+    {
+      "epoch": 2.3466666666666667,
+      "grad_norm": 2.547231435775757,
+      "learning_rate": 0.00010622775800711744,
+      "loss": 1.6653,
+      "step": 5280
+    },
+    {
+      "epoch": 2.347111111111111,
+      "grad_norm": 2.483854055404663,
+      "learning_rate": 0.0001062099644128114,
+      "loss": 1.5825,
+      "step": 5281
+    },
+    {
+      "epoch": 2.3475555555555556,
+      "grad_norm": 2.559788465499878,
+      "learning_rate": 0.00010619217081850534,
+      "loss": 2.0953,
+      "step": 5282
+    },
+    {
+      "epoch": 2.348,
+      "grad_norm": 3.0145204067230225,
+      "learning_rate": 0.0001061743772241993,
+      "loss": 1.9542,
+      "step": 5283
+    },
+    {
+      "epoch": 2.3484444444444446,
+      "grad_norm": 2.558265209197998,
+      "learning_rate": 0.00010615658362989325,
+      "loss": 1.8425,
+      "step": 5284
+    },
+    {
+      "epoch": 2.348888888888889,
+      "grad_norm": 2.9224250316619873,
+      "learning_rate": 0.0001061387900355872,
+      "loss": 1.947,
+      "step": 5285
+    },
+    {
+      "epoch": 2.3493333333333335,
+      "grad_norm": 2.7229421138763428,
+      "learning_rate": 0.00010612099644128115,
+      "loss": 1.305,
+      "step": 5286
+    },
+    {
+      "epoch": 2.3497777777777777,
+      "grad_norm": 2.2449328899383545,
+      "learning_rate": 0.00010610320284697511,
+      "loss": 1.4746,
+      "step": 5287
+    },
+    {
+      "epoch": 2.3502222222222224,
+      "grad_norm": 2.956235885620117,
+      "learning_rate": 0.00010608540925266905,
+      "loss": 1.9641,
+      "step": 5288
+    },
+    {
+      "epoch": 2.3506666666666667,
+      "grad_norm": 2.308121681213379,
+      "learning_rate": 0.00010606761565836298,
+      "loss": 1.4521,
+      "step": 5289
+    },
+    {
+      "epoch": 2.351111111111111,
+      "grad_norm": 2.414666175842285,
+      "learning_rate": 0.00010604982206405694,
+      "loss": 1.0793,
+      "step": 5290
+    },
+    {
+      "epoch": 2.3515555555555556,
+      "grad_norm": 3.1062090396881104,
+      "learning_rate": 0.0001060320284697509,
+      "loss": 1.5819,
+      "step": 5291
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 2.759904146194458,
+      "learning_rate": 0.00010601423487544484,
+      "loss": 2.2146,
+      "step": 5292
+    },
+    {
+      "epoch": 2.3524444444444446,
+      "grad_norm": 3.40423321723938,
+      "learning_rate": 0.00010599644128113879,
+      "loss": 1.8593,
+      "step": 5293
+    },
+    {
+      "epoch": 2.352888888888889,
+      "grad_norm": 3.157247543334961,
+      "learning_rate": 0.00010597864768683275,
+      "loss": 1.9475,
+      "step": 5294
+    },
+    {
+      "epoch": 2.3533333333333335,
+      "grad_norm": 2.931468963623047,
+      "learning_rate": 0.00010596085409252669,
+      "loss": 1.9819,
+      "step": 5295
+    },
+    {
+      "epoch": 2.3537777777777777,
+      "grad_norm": 2.865544557571411,
+      "learning_rate": 0.00010594306049822065,
+      "loss": 1.7654,
+      "step": 5296
+    },
+    {
+      "epoch": 2.354222222222222,
+      "grad_norm": 3.2829089164733887,
+      "learning_rate": 0.0001059252669039146,
+      "loss": 1.7053,
+      "step": 5297
+    },
+    {
+      "epoch": 2.3546666666666667,
+      "grad_norm": 2.915627956390381,
+      "learning_rate": 0.00010590747330960855,
+      "loss": 1.6627,
+      "step": 5298
+    },
+    {
+      "epoch": 2.355111111111111,
+      "grad_norm": 2.891852617263794,
+      "learning_rate": 0.0001058896797153025,
+      "loss": 1.4539,
+      "step": 5299
+    },
+    {
+      "epoch": 2.3555555555555556,
+      "grad_norm": 3.9856643676757812,
+      "learning_rate": 0.00010587188612099646,
+      "loss": 1.6694,
+      "step": 5300
+    },
+    {
+      "epoch": 2.356,
+      "grad_norm": 1.8876063823699951,
+      "learning_rate": 0.0001058540925266904,
+      "loss": 1.9182,
+      "step": 5301
+    },
+    {
+      "epoch": 2.3564444444444446,
+      "grad_norm": 1.3082079887390137,
+      "learning_rate": 0.00010583629893238433,
+      "loss": 0.945,
+      "step": 5302
+    },
+    {
+      "epoch": 2.356888888888889,
+      "grad_norm": 1.5244293212890625,
+      "learning_rate": 0.00010581850533807829,
+      "loss": 0.8277,
+      "step": 5303
+    },
+    {
+      "epoch": 2.3573333333333335,
+      "grad_norm": 2.210871458053589,
+      "learning_rate": 0.00010580071174377225,
+      "loss": 1.937,
+      "step": 5304
+    },
+    {
+      "epoch": 2.3577777777777778,
+      "grad_norm": 2.4062864780426025,
+      "learning_rate": 0.00010578291814946619,
+      "loss": 1.9807,
+      "step": 5305
+    },
+    {
+      "epoch": 2.3582222222222224,
+      "grad_norm": 2.0000367164611816,
+      "learning_rate": 0.00010576512455516015,
+      "loss": 1.4079,
+      "step": 5306
+    },
+    {
+      "epoch": 2.3586666666666667,
+      "grad_norm": 1.9247461557388306,
+      "learning_rate": 0.0001057473309608541,
+      "loss": 1.7219,
+      "step": 5307
+    },
+    {
+      "epoch": 2.359111111111111,
+      "grad_norm": 2.2574825286865234,
+      "learning_rate": 0.00010572953736654805,
+      "loss": 1.7616,
+      "step": 5308
+    },
+    {
+      "epoch": 2.3595555555555556,
+      "grad_norm": 2.254293203353882,
+      "learning_rate": 0.000105711743772242,
+      "loss": 2.0128,
+      "step": 5309
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.4639298915863037,
+      "learning_rate": 0.00010569395017793596,
+      "loss": 1.9739,
+      "step": 5310
+    },
+    {
+      "epoch": 2.3604444444444446,
+      "grad_norm": 2.103483200073242,
+      "learning_rate": 0.0001056761565836299,
+      "loss": 1.91,
+      "step": 5311
+    },
+    {
+      "epoch": 2.360888888888889,
+      "grad_norm": 0.13134007155895233,
+      "learning_rate": 0.00010565836298932386,
+      "loss": 0.0197,
+      "step": 5312
+    },
+    {
+      "epoch": 2.3613333333333335,
+      "grad_norm": 2.1380348205566406,
+      "learning_rate": 0.00010564056939501781,
+      "loss": 1.7965,
+      "step": 5313
+    },
+    {
+      "epoch": 2.3617777777777778,
+      "grad_norm": 1.9932422637939453,
+      "learning_rate": 0.00010562277580071176,
+      "loss": 1.6092,
+      "step": 5314
+    },
+    {
+      "epoch": 2.362222222222222,
+      "grad_norm": 2.254859209060669,
+      "learning_rate": 0.00010560498220640569,
+      "loss": 2.1432,
+      "step": 5315
+    },
+    {
+      "epoch": 2.3626666666666667,
+      "grad_norm": 2.0134313106536865,
+      "learning_rate": 0.00010558718861209964,
+      "loss": 1.8426,
+      "step": 5316
+    },
+    {
+      "epoch": 2.363111111111111,
+      "grad_norm": 2.157883644104004,
+      "learning_rate": 0.0001055693950177936,
+      "loss": 1.5952,
+      "step": 5317
+    },
+    {
+      "epoch": 2.3635555555555556,
+      "grad_norm": 2.2276217937469482,
+      "learning_rate": 0.00010555160142348754,
+      "loss": 1.8253,
+      "step": 5318
+    },
+    {
+      "epoch": 2.364,
+      "grad_norm": 2.406975269317627,
+      "learning_rate": 0.0001055338078291815,
+      "loss": 1.7489,
+      "step": 5319
+    },
+    {
+      "epoch": 2.3644444444444446,
+      "grad_norm": 2.4040451049804688,
+      "learning_rate": 0.00010551601423487544,
+      "loss": 1.7381,
+      "step": 5320
+    },
+    {
+      "epoch": 2.364888888888889,
+      "grad_norm": 2.3204426765441895,
+      "learning_rate": 0.0001054982206405694,
+      "loss": 2.0069,
+      "step": 5321
+    },
+    {
+      "epoch": 2.3653333333333335,
+      "grad_norm": 2.3594791889190674,
+      "learning_rate": 0.00010548042704626336,
+      "loss": 1.6934,
+      "step": 5322
+    },
+    {
+      "epoch": 2.3657777777777778,
+      "grad_norm": 2.970160961151123,
+      "learning_rate": 0.0001054626334519573,
+      "loss": 2.0672,
+      "step": 5323
+    },
+    {
+      "epoch": 2.3662222222222224,
+      "grad_norm": 2.4029760360717773,
+      "learning_rate": 0.00010544483985765125,
+      "loss": 1.6066,
+      "step": 5324
+    },
+    {
+      "epoch": 2.3666666666666667,
+      "grad_norm": 2.390634059906006,
+      "learning_rate": 0.00010542704626334521,
+      "loss": 1.8098,
+      "step": 5325
+    },
+    {
+      "epoch": 2.367111111111111,
+      "grad_norm": 2.158510208129883,
+      "learning_rate": 0.00010540925266903915,
+      "loss": 1.4622,
+      "step": 5326
+    },
+    {
+      "epoch": 2.3675555555555556,
+      "grad_norm": 2.2908811569213867,
+      "learning_rate": 0.00010539145907473311,
+      "loss": 1.4533,
+      "step": 5327
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 2.348654270172119,
+      "learning_rate": 0.00010537366548042704,
+      "loss": 1.7244,
+      "step": 5328
+    },
+    {
+      "epoch": 2.3684444444444446,
+      "grad_norm": 2.7863059043884277,
+      "learning_rate": 0.000105355871886121,
+      "loss": 1.9031,
+      "step": 5329
+    },
+    {
+      "epoch": 2.368888888888889,
+      "grad_norm": 1.919249415397644,
+      "learning_rate": 0.00010533807829181494,
+      "loss": 0.9351,
+      "step": 5330
+    },
+    {
+      "epoch": 2.3693333333333335,
+      "grad_norm": 2.1779160499572754,
+      "learning_rate": 0.0001053202846975089,
+      "loss": 1.311,
+      "step": 5331
+    },
+    {
+      "epoch": 2.3697777777777778,
+      "grad_norm": 2.9580140113830566,
+      "learning_rate": 0.00010530249110320285,
+      "loss": 1.8562,
+      "step": 5332
+    },
+    {
+      "epoch": 2.370222222222222,
+      "grad_norm": 2.4959969520568848,
+      "learning_rate": 0.0001052846975088968,
+      "loss": 1.3581,
+      "step": 5333
+    },
+    {
+      "epoch": 2.3706666666666667,
+      "grad_norm": 2.6963963508605957,
+      "learning_rate": 0.00010526690391459075,
+      "loss": 2.3081,
+      "step": 5334
+    },
+    {
+      "epoch": 2.371111111111111,
+      "grad_norm": 2.9075794219970703,
+      "learning_rate": 0.00010524911032028471,
+      "loss": 2.2261,
+      "step": 5335
+    },
+    {
+      "epoch": 2.3715555555555556,
+      "grad_norm": 2.346102714538574,
+      "learning_rate": 0.00010523131672597865,
+      "loss": 1.4546,
+      "step": 5336
+    },
+    {
+      "epoch": 2.372,
+      "grad_norm": 2.860933542251587,
+      "learning_rate": 0.00010521352313167261,
+      "loss": 1.6149,
+      "step": 5337
+    },
+    {
+      "epoch": 2.3724444444444446,
+      "grad_norm": 3.0181329250335693,
+      "learning_rate": 0.00010519572953736656,
+      "loss": 1.9698,
+      "step": 5338
+    },
+    {
+      "epoch": 2.372888888888889,
+      "grad_norm": 3.008528709411621,
+      "learning_rate": 0.00010517793594306051,
+      "loss": 1.943,
+      "step": 5339
+    },
+    {
+      "epoch": 2.3733333333333335,
+      "grad_norm": 2.5461809635162354,
+      "learning_rate": 0.00010516014234875444,
+      "loss": 1.473,
+      "step": 5340
+    },
+    {
+      "epoch": 2.3737777777777778,
+      "grad_norm": 2.90517258644104,
+      "learning_rate": 0.00010514234875444839,
+      "loss": 2.0846,
+      "step": 5341
+    },
+    {
+      "epoch": 2.3742222222222225,
+      "grad_norm": 2.84635329246521,
+      "learning_rate": 0.00010512455516014235,
+      "loss": 1.8785,
+      "step": 5342
+    },
+    {
+      "epoch": 2.3746666666666667,
+      "grad_norm": 2.7728958129882812,
+      "learning_rate": 0.00010510676156583629,
+      "loss": 1.4901,
+      "step": 5343
+    },
+    {
+      "epoch": 2.375111111111111,
+      "grad_norm": 2.4480459690093994,
+      "learning_rate": 0.00010508896797153025,
+      "loss": 1.4275,
+      "step": 5344
+    },
+    {
+      "epoch": 2.3755555555555556,
+      "grad_norm": 3.176147937774658,
+      "learning_rate": 0.0001050711743772242,
+      "loss": 1.3497,
+      "step": 5345
+    },
+    {
+      "epoch": 2.376,
+      "grad_norm": 2.7006847858428955,
+      "learning_rate": 0.00010505338078291815,
+      "loss": 1.2245,
+      "step": 5346
+    },
+    {
+      "epoch": 2.3764444444444446,
+      "grad_norm": 3.114924192428589,
+      "learning_rate": 0.0001050355871886121,
+      "loss": 1.6906,
+      "step": 5347
+    },
+    {
+      "epoch": 2.376888888888889,
+      "grad_norm": 3.2181596755981445,
+      "learning_rate": 0.00010501779359430606,
+      "loss": 1.4348,
+      "step": 5348
+    },
+    {
+      "epoch": 2.3773333333333335,
+      "grad_norm": 2.9679977893829346,
+      "learning_rate": 0.000105,
+      "loss": 1.8438,
+      "step": 5349
+    },
+    {
+      "epoch": 2.3777777777777778,
+      "grad_norm": 3.126516342163086,
+      "learning_rate": 0.00010498220640569396,
+      "loss": 1.0786,
+      "step": 5350
+    },
+    {
+      "epoch": 2.378222222222222,
+      "grad_norm": 1.8322924375534058,
+      "learning_rate": 0.00010496441281138792,
+      "loss": 2.0924,
+      "step": 5351
+    },
+    {
+      "epoch": 2.3786666666666667,
+      "grad_norm": 1.7690398693084717,
+      "learning_rate": 0.00010494661921708186,
+      "loss": 2.0291,
+      "step": 5352
+    },
+    {
+      "epoch": 2.379111111111111,
+      "grad_norm": 0.14878343045711517,
+      "learning_rate": 0.00010492882562277579,
+      "loss": 0.0181,
+      "step": 5353
+    },
+    {
+      "epoch": 2.3795555555555556,
+      "grad_norm": 2.2192952632904053,
+      "learning_rate": 0.00010491103202846975,
+      "loss": 1.5731,
+      "step": 5354
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.1451058387756348,
+      "learning_rate": 0.0001048932384341637,
+      "loss": 1.7856,
+      "step": 5355
+    },
+    {
+      "epoch": 2.3804444444444446,
+      "grad_norm": 2.5920441150665283,
+      "learning_rate": 0.00010487544483985765,
+      "loss": 2.0455,
+      "step": 5356
+    },
+    {
+      "epoch": 2.380888888888889,
+      "grad_norm": 2.153883218765259,
+      "learning_rate": 0.0001048576512455516,
+      "loss": 1.6193,
+      "step": 5357
+    },
+    {
+      "epoch": 2.3813333333333335,
+      "grad_norm": 2.2498092651367188,
+      "learning_rate": 0.00010483985765124556,
+      "loss": 2.1478,
+      "step": 5358
+    },
+    {
+      "epoch": 2.3817777777777778,
+      "grad_norm": 2.189810037612915,
+      "learning_rate": 0.0001048220640569395,
+      "loss": 1.8984,
+      "step": 5359
+    },
+    {
+      "epoch": 2.3822222222222225,
+      "grad_norm": 2.1763858795166016,
+      "learning_rate": 0.00010480427046263346,
+      "loss": 1.6781,
+      "step": 5360
+    },
+    {
+      "epoch": 2.3826666666666667,
+      "grad_norm": 2.4611992835998535,
+      "learning_rate": 0.00010478647686832741,
+      "loss": 1.9078,
+      "step": 5361
+    },
+    {
+      "epoch": 2.383111111111111,
+      "grad_norm": 2.345245122909546,
+      "learning_rate": 0.00010476868327402136,
+      "loss": 1.9821,
+      "step": 5362
+    },
+    {
+      "epoch": 2.3835555555555556,
+      "grad_norm": 2.14660906791687,
+      "learning_rate": 0.00010475088967971531,
+      "loss": 1.4488,
+      "step": 5363
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 2.06569504737854,
+      "learning_rate": 0.00010473309608540927,
+      "loss": 1.908,
+      "step": 5364
+    },
+    {
+      "epoch": 2.3844444444444446,
+      "grad_norm": 2.5986487865448,
+      "learning_rate": 0.00010471530249110321,
+      "loss": 2.0425,
+      "step": 5365
+    },
+    {
+      "epoch": 2.384888888888889,
+      "grad_norm": 2.309596061706543,
+      "learning_rate": 0.00010469750889679714,
+      "loss": 2.0211,
+      "step": 5366
+    },
+    {
+      "epoch": 2.3853333333333335,
+      "grad_norm": 2.2031354904174805,
+      "learning_rate": 0.0001046797153024911,
+      "loss": 1.793,
+      "step": 5367
+    },
+    {
+      "epoch": 2.3857777777777778,
+      "grad_norm": 2.5349233150482178,
+      "learning_rate": 0.00010466192170818506,
+      "loss": 2.126,
+      "step": 5368
+    },
+    {
+      "epoch": 2.386222222222222,
+      "grad_norm": 1.8240305185317993,
+      "learning_rate": 0.000104644128113879,
+      "loss": 1.3644,
+      "step": 5369
+    },
+    {
+      "epoch": 2.3866666666666667,
+      "grad_norm": 1.6334868669509888,
+      "learning_rate": 0.00010462633451957296,
+      "loss": 1.0386,
+      "step": 5370
+    },
+    {
+      "epoch": 2.387111111111111,
+      "grad_norm": 2.6212308406829834,
+      "learning_rate": 0.00010460854092526691,
+      "loss": 2.2779,
+      "step": 5371
+    },
+    {
+      "epoch": 2.3875555555555557,
+      "grad_norm": 2.4640159606933594,
+      "learning_rate": 0.00010459074733096085,
+      "loss": 2.1993,
+      "step": 5372
+    },
+    {
+      "epoch": 2.388,
+      "grad_norm": 2.5300076007843018,
+      "learning_rate": 0.00010457295373665481,
+      "loss": 2.1011,
+      "step": 5373
+    },
+    {
+      "epoch": 2.3884444444444446,
+      "grad_norm": 2.257607936859131,
+      "learning_rate": 0.00010455516014234877,
+      "loss": 1.6381,
+      "step": 5374
+    },
+    {
+      "epoch": 2.388888888888889,
+      "grad_norm": 2.7849531173706055,
+      "learning_rate": 0.00010453736654804271,
+      "loss": 1.4969,
+      "step": 5375
+    },
+    {
+      "epoch": 2.389333333333333,
+      "grad_norm": 2.2761592864990234,
+      "learning_rate": 0.00010451957295373667,
+      "loss": 1.4029,
+      "step": 5376
+    },
+    {
+      "epoch": 2.389777777777778,
+      "grad_norm": 2.228327512741089,
+      "learning_rate": 0.00010450177935943062,
+      "loss": 1.8044,
+      "step": 5377
+    },
+    {
+      "epoch": 2.3902222222222225,
+      "grad_norm": 1.8073506355285645,
+      "learning_rate": 0.00010448398576512457,
+      "loss": 0.9291,
+      "step": 5378
+    },
+    {
+      "epoch": 2.3906666666666667,
+      "grad_norm": 2.5977213382720947,
+      "learning_rate": 0.0001044661921708185,
+      "loss": 1.585,
+      "step": 5379
+    },
+    {
+      "epoch": 2.391111111111111,
+      "grad_norm": 2.726691246032715,
+      "learning_rate": 0.00010444839857651245,
+      "loss": 1.6747,
+      "step": 5380
+    },
+    {
+      "epoch": 2.3915555555555557,
+      "grad_norm": 2.13911509513855,
+      "learning_rate": 0.00010443060498220641,
+      "loss": 1.531,
+      "step": 5381
+    },
+    {
+      "epoch": 2.392,
+      "grad_norm": 2.832296133041382,
+      "learning_rate": 0.00010441281138790035,
+      "loss": 1.7403,
+      "step": 5382
+    },
+    {
+      "epoch": 2.3924444444444446,
+      "grad_norm": 2.713308811187744,
+      "learning_rate": 0.00010439501779359431,
+      "loss": 2.0669,
+      "step": 5383
+    },
+    {
+      "epoch": 2.392888888888889,
+      "grad_norm": 2.503474473953247,
+      "learning_rate": 0.00010437722419928826,
+      "loss": 1.8383,
+      "step": 5384
+    },
+    {
+      "epoch": 2.3933333333333335,
+      "grad_norm": 2.878114700317383,
+      "learning_rate": 0.00010435943060498221,
+      "loss": 1.7423,
+      "step": 5385
+    },
+    {
+      "epoch": 2.393777777777778,
+      "grad_norm": 3.174804449081421,
+      "learning_rate": 0.00010434163701067616,
+      "loss": 1.5972,
+      "step": 5386
+    },
+    {
+      "epoch": 2.394222222222222,
+      "grad_norm": 2.129746675491333,
+      "learning_rate": 0.00010432384341637012,
+      "loss": 1.0781,
+      "step": 5387
+    },
+    {
+      "epoch": 2.3946666666666667,
+      "grad_norm": 2.780766487121582,
+      "learning_rate": 0.00010430604982206406,
+      "loss": 1.649,
+      "step": 5388
+    },
+    {
+      "epoch": 2.395111111111111,
+      "grad_norm": 2.8491053581237793,
+      "learning_rate": 0.00010428825622775802,
+      "loss": 1.6903,
+      "step": 5389
+    },
+    {
+      "epoch": 2.3955555555555557,
+      "grad_norm": 2.460873603820801,
+      "learning_rate": 0.00010427046263345198,
+      "loss": 1.3277,
+      "step": 5390
+    },
+    {
+      "epoch": 2.396,
+      "grad_norm": 2.9226863384246826,
+      "learning_rate": 0.00010425266903914592,
+      "loss": 1.5445,
+      "step": 5391
+    },
+    {
+      "epoch": 2.3964444444444446,
+      "grad_norm": 3.39444899559021,
+      "learning_rate": 0.00010423487544483985,
+      "loss": 1.6876,
+      "step": 5392
+    },
+    {
+      "epoch": 2.396888888888889,
+      "grad_norm": 3.0782647132873535,
+      "learning_rate": 0.0001042170818505338,
+      "loss": 1.6426,
+      "step": 5393
+    },
+    {
+      "epoch": 2.397333333333333,
+      "grad_norm": 2.9284658432006836,
+      "learning_rate": 0.00010419928825622776,
+      "loss": 1.5015,
+      "step": 5394
+    },
+    {
+      "epoch": 2.397777777777778,
+      "grad_norm": 3.772498369216919,
+      "learning_rate": 0.0001041814946619217,
+      "loss": 1.734,
+      "step": 5395
+    },
+    {
+      "epoch": 2.398222222222222,
+      "grad_norm": 3.4858293533325195,
+      "learning_rate": 0.00010416370106761566,
+      "loss": 2.1559,
+      "step": 5396
+    },
+    {
+      "epoch": 2.3986666666666667,
+      "grad_norm": 2.8593242168426514,
+      "learning_rate": 0.00010414590747330962,
+      "loss": 1.2558,
+      "step": 5397
+    },
+    {
+      "epoch": 2.399111111111111,
+      "grad_norm": 3.39560866355896,
+      "learning_rate": 0.00010412811387900356,
+      "loss": 1.726,
+      "step": 5398
+    },
+    {
+      "epoch": 2.3995555555555557,
+      "grad_norm": 2.1574625968933105,
+      "learning_rate": 0.00010411032028469752,
+      "loss": 0.8812,
+      "step": 5399
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.058824062347412,
+      "learning_rate": 0.00010409252669039147,
+      "loss": 1.1959,
+      "step": 5400
+    },
+    {
+      "epoch": 2.4004444444444446,
+      "grad_norm": 1.906567931175232,
+      "learning_rate": 0.00010407473309608542,
+      "loss": 2.2807,
+      "step": 5401
+    },
+    {
+      "epoch": 2.400888888888889,
+      "grad_norm": 1.9457437992095947,
+      "learning_rate": 0.00010405693950177937,
+      "loss": 1.7017,
+      "step": 5402
+    },
+    {
+      "epoch": 2.4013333333333335,
+      "grad_norm": 2.0136795043945312,
+      "learning_rate": 0.00010403914590747333,
+      "loss": 2.0448,
+      "step": 5403
+    },
+    {
+      "epoch": 2.401777777777778,
+      "grad_norm": 2.323181390762329,
+      "learning_rate": 0.00010402135231316727,
+      "loss": 1.8618,
+      "step": 5404
+    },
+    {
+      "epoch": 2.402222222222222,
+      "grad_norm": 1.999695897102356,
+      "learning_rate": 0.0001040035587188612,
+      "loss": 1.8949,
+      "step": 5405
+    },
+    {
+      "epoch": 2.4026666666666667,
+      "grad_norm": 2.2009153366088867,
+      "learning_rate": 0.00010398576512455516,
+      "loss": 1.7722,
+      "step": 5406
+    },
+    {
+      "epoch": 2.403111111111111,
+      "grad_norm": 2.2078299522399902,
+      "learning_rate": 0.00010396797153024912,
+      "loss": 2.0193,
+      "step": 5407
+    },
+    {
+      "epoch": 2.4035555555555557,
+      "grad_norm": 2.5770654678344727,
+      "learning_rate": 0.00010395017793594306,
+      "loss": 2.0849,
+      "step": 5408
+    },
+    {
+      "epoch": 2.404,
+      "grad_norm": 2.2066149711608887,
+      "learning_rate": 0.00010393238434163701,
+      "loss": 1.5842,
+      "step": 5409
+    },
+    {
+      "epoch": 2.4044444444444446,
+      "grad_norm": 2.5766196250915527,
+      "learning_rate": 0.00010391459074733096,
+      "loss": 2.3273,
+      "step": 5410
+    },
+    {
+      "epoch": 2.404888888888889,
+      "grad_norm": 2.1553375720977783,
+      "learning_rate": 0.00010389679715302491,
+      "loss": 1.7654,
+      "step": 5411
+    },
+    {
+      "epoch": 2.405333333333333,
+      "grad_norm": 2.179262638092041,
+      "learning_rate": 0.00010387900355871887,
+      "loss": 1.8465,
+      "step": 5412
+    },
+    {
+      "epoch": 2.405777777777778,
+      "grad_norm": 2.3855044841766357,
+      "learning_rate": 0.00010386120996441281,
+      "loss": 1.5526,
+      "step": 5413
+    },
+    {
+      "epoch": 2.406222222222222,
+      "grad_norm": 1.9066109657287598,
+      "learning_rate": 0.00010384341637010677,
+      "loss": 1.1865,
+      "step": 5414
+    },
+    {
+      "epoch": 2.4066666666666667,
+      "grad_norm": 2.4043421745300293,
+      "learning_rate": 0.00010382562277580073,
+      "loss": 2.1115,
+      "step": 5415
+    },
+    {
+      "epoch": 2.407111111111111,
+      "grad_norm": 2.303219795227051,
+      "learning_rate": 0.00010380782918149467,
+      "loss": 1.6968,
+      "step": 5416
+    },
+    {
+      "epoch": 2.4075555555555557,
+      "grad_norm": 2.519510507583618,
+      "learning_rate": 0.00010379003558718863,
+      "loss": 1.4451,
+      "step": 5417
+    },
+    {
+      "epoch": 2.408,
+      "grad_norm": 2.0006396770477295,
+      "learning_rate": 0.00010377224199288256,
+      "loss": 1.6501,
+      "step": 5418
+    },
+    {
+      "epoch": 2.4084444444444446,
+      "grad_norm": 2.359020709991455,
+      "learning_rate": 0.00010375444839857651,
+      "loss": 2.1004,
+      "step": 5419
+    },
+    {
+      "epoch": 2.408888888888889,
+      "grad_norm": 2.581718683242798,
+      "learning_rate": 0.00010373665480427045,
+      "loss": 1.5921,
+      "step": 5420
+    },
+    {
+      "epoch": 2.4093333333333335,
+      "grad_norm": 2.2149415016174316,
+      "learning_rate": 0.00010371886120996441,
+      "loss": 1.4495,
+      "step": 5421
+    },
+    {
+      "epoch": 2.409777777777778,
+      "grad_norm": 2.1162991523742676,
+      "learning_rate": 0.00010370106761565837,
+      "loss": 1.7689,
+      "step": 5422
+    },
+    {
+      "epoch": 2.410222222222222,
+      "grad_norm": 2.3050718307495117,
+      "learning_rate": 0.00010368327402135231,
+      "loss": 1.7421,
+      "step": 5423
+    },
+    {
+      "epoch": 2.4106666666666667,
+      "grad_norm": 2.368021011352539,
+      "learning_rate": 0.00010366548042704627,
+      "loss": 1.8718,
+      "step": 5424
+    },
+    {
+      "epoch": 2.411111111111111,
+      "grad_norm": 2.574089765548706,
+      "learning_rate": 0.00010364768683274022,
+      "loss": 1.8538,
+      "step": 5425
+    },
+    {
+      "epoch": 2.4115555555555557,
+      "grad_norm": 2.1765310764312744,
+      "learning_rate": 0.00010362989323843417,
+      "loss": 1.6152,
+      "step": 5426
+    },
+    {
+      "epoch": 2.412,
+      "grad_norm": 2.298729419708252,
+      "learning_rate": 0.00010361209964412812,
+      "loss": 1.6085,
+      "step": 5427
+    },
+    {
+      "epoch": 2.4124444444444446,
+      "grad_norm": 2.547469139099121,
+      "learning_rate": 0.00010359430604982208,
+      "loss": 1.7991,
+      "step": 5428
+    },
+    {
+      "epoch": 2.412888888888889,
+      "grad_norm": 2.315479040145874,
+      "learning_rate": 0.00010357651245551602,
+      "loss": 1.5592,
+      "step": 5429
+    },
+    {
+      "epoch": 2.413333333333333,
+      "grad_norm": 2.59236478805542,
+      "learning_rate": 0.00010355871886120998,
+      "loss": 1.4449,
+      "step": 5430
+    },
+    {
+      "epoch": 2.413777777777778,
+      "grad_norm": 2.6196041107177734,
+      "learning_rate": 0.00010354092526690391,
+      "loss": 1.4048,
+      "step": 5431
+    },
+    {
+      "epoch": 2.414222222222222,
+      "grad_norm": 2.8616275787353516,
+      "learning_rate": 0.00010352313167259787,
+      "loss": 2.1669,
+      "step": 5432
+    },
+    {
+      "epoch": 2.4146666666666667,
+      "grad_norm": 2.7045931816101074,
+      "learning_rate": 0.00010350533807829181,
+      "loss": 1.8538,
+      "step": 5433
+    },
+    {
+      "epoch": 2.415111111111111,
+      "grad_norm": 2.812784194946289,
+      "learning_rate": 0.00010348754448398576,
+      "loss": 1.7349,
+      "step": 5434
+    },
+    {
+      "epoch": 2.4155555555555557,
+      "grad_norm": 2.665402889251709,
+      "learning_rate": 0.00010346975088967972,
+      "loss": 1.671,
+      "step": 5435
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 2.632174253463745,
+      "learning_rate": 0.00010345195729537366,
+      "loss": 1.297,
+      "step": 5436
+    },
+    {
+      "epoch": 2.4164444444444446,
+      "grad_norm": 2.9285874366760254,
+      "learning_rate": 0.00010343416370106762,
+      "loss": 1.6447,
+      "step": 5437
+    },
+    {
+      "epoch": 2.416888888888889,
+      "grad_norm": 2.873795747756958,
+      "learning_rate": 0.00010341637010676158,
+      "loss": 1.6297,
+      "step": 5438
+    },
+    {
+      "epoch": 2.4173333333333336,
+      "grad_norm": 1.693901538848877,
+      "learning_rate": 0.00010339857651245552,
+      "loss": 0.571,
+      "step": 5439
+    },
+    {
+      "epoch": 2.417777777777778,
+      "grad_norm": 2.4902453422546387,
+      "learning_rate": 0.00010338078291814948,
+      "loss": 1.7092,
+      "step": 5440
+    },
+    {
+      "epoch": 2.418222222222222,
+      "grad_norm": 2.657254457473755,
+      "learning_rate": 0.00010336298932384343,
+      "loss": 1.2449,
+      "step": 5441
+    },
+    {
+      "epoch": 2.4186666666666667,
+      "grad_norm": 2.572938919067383,
+      "learning_rate": 0.00010334519572953738,
+      "loss": 1.6842,
+      "step": 5442
+    },
+    {
+      "epoch": 2.419111111111111,
+      "grad_norm": 3.0407369136810303,
+      "learning_rate": 0.00010332740213523133,
+      "loss": 1.7859,
+      "step": 5443
+    },
+    {
+      "epoch": 2.4195555555555557,
+      "grad_norm": 2.9764583110809326,
+      "learning_rate": 0.00010330960854092526,
+      "loss": 1.6893,
+      "step": 5444
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.7912721633911133,
+      "learning_rate": 0.00010329181494661922,
+      "loss": 1.5315,
+      "step": 5445
+    },
+    {
+      "epoch": 2.4204444444444446,
+      "grad_norm": 3.436552047729492,
+      "learning_rate": 0.00010327402135231316,
+      "loss": 1.6444,
+      "step": 5446
+    },
+    {
+      "epoch": 2.420888888888889,
+      "grad_norm": 3.2824277877807617,
+      "learning_rate": 0.00010325622775800712,
+      "loss": 1.6908,
+      "step": 5447
+    },
+    {
+      "epoch": 2.421333333333333,
+      "grad_norm": 3.5586631298065186,
+      "learning_rate": 0.00010323843416370107,
+      "loss": 1.9766,
+      "step": 5448
+    },
+    {
+      "epoch": 2.421777777777778,
+      "grad_norm": 3.6402716636657715,
+      "learning_rate": 0.00010322064056939502,
+      "loss": 1.8899,
+      "step": 5449
+    },
+    {
+      "epoch": 2.422222222222222,
+      "grad_norm": 3.806692600250244,
+      "learning_rate": 0.00010320284697508897,
+      "loss": 1.7844,
+      "step": 5450
+    },
+    {
+      "epoch": 2.4226666666666667,
+      "grad_norm": 1.4669278860092163,
+      "learning_rate": 0.00010318505338078293,
+      "loss": 0.98,
+      "step": 5451
+    },
+    {
+      "epoch": 2.423111111111111,
+      "grad_norm": 1.7053353786468506,
+      "learning_rate": 0.00010316725978647687,
+      "loss": 2.1493,
+      "step": 5452
+    },
+    {
+      "epoch": 2.4235555555555557,
+      "grad_norm": 2.3028388023376465,
+      "learning_rate": 0.00010314946619217083,
+      "loss": 2.0852,
+      "step": 5453
+    },
+    {
+      "epoch": 2.424,
+      "grad_norm": 1.9444522857666016,
+      "learning_rate": 0.00010313167259786479,
+      "loss": 1.6929,
+      "step": 5454
+    },
+    {
+      "epoch": 2.4244444444444446,
+      "grad_norm": 2.3324930667877197,
+      "learning_rate": 0.00010311387900355873,
+      "loss": 1.9656,
+      "step": 5455
+    },
+    {
+      "epoch": 2.424888888888889,
+      "grad_norm": 2.4078588485717773,
+      "learning_rate": 0.00010309608540925266,
+      "loss": 1.8105,
+      "step": 5456
+    },
+    {
+      "epoch": 2.4253333333333336,
+      "grad_norm": 2.183835983276367,
+      "learning_rate": 0.00010307829181494661,
+      "loss": 1.921,
+      "step": 5457
+    },
+    {
+      "epoch": 2.425777777777778,
+      "grad_norm": 2.3134961128234863,
+      "learning_rate": 0.00010306049822064057,
+      "loss": 1.8601,
+      "step": 5458
+    },
+    {
+      "epoch": 2.426222222222222,
+      "grad_norm": 2.484114408493042,
+      "learning_rate": 0.00010304270462633451,
+      "loss": 1.701,
+      "step": 5459
+    },
+    {
+      "epoch": 2.4266666666666667,
+      "grad_norm": 2.430379867553711,
+      "learning_rate": 0.00010302491103202847,
+      "loss": 1.9341,
+      "step": 5460
+    },
+    {
+      "epoch": 2.427111111111111,
+      "grad_norm": 2.4869649410247803,
+      "learning_rate": 0.00010300711743772243,
+      "loss": 1.9392,
+      "step": 5461
+    },
+    {
+      "epoch": 2.4275555555555557,
+      "grad_norm": 2.402344226837158,
+      "learning_rate": 0.00010298932384341637,
+      "loss": 1.8691,
+      "step": 5462
+    },
+    {
+      "epoch": 2.428,
+      "grad_norm": 1.7501614093780518,
+      "learning_rate": 0.00010297153024911033,
+      "loss": 1.085,
+      "step": 5463
+    },
+    {
+      "epoch": 2.4284444444444446,
+      "grad_norm": 2.223562002182007,
+      "learning_rate": 0.00010295373665480428,
+      "loss": 1.3557,
+      "step": 5464
+    },
+    {
+      "epoch": 2.428888888888889,
+      "grad_norm": 2.1328225135803223,
+      "learning_rate": 0.00010293594306049823,
+      "loss": 1.7067,
+      "step": 5465
+    },
+    {
+      "epoch": 2.429333333333333,
+      "grad_norm": 2.1772732734680176,
+      "learning_rate": 0.00010291814946619218,
+      "loss": 1.5118,
+      "step": 5466
+    },
+    {
+      "epoch": 2.429777777777778,
+      "grad_norm": 2.320878505706787,
+      "learning_rate": 0.00010290035587188614,
+      "loss": 1.3153,
+      "step": 5467
+    },
+    {
+      "epoch": 2.430222222222222,
+      "grad_norm": 2.253594398498535,
+      "learning_rate": 0.00010288256227758008,
+      "loss": 1.8138,
+      "step": 5468
+    },
+    {
+      "epoch": 2.4306666666666668,
+      "grad_norm": 2.6910340785980225,
+      "learning_rate": 0.00010286476868327401,
+      "loss": 2.0876,
+      "step": 5469
+    },
+    {
+      "epoch": 2.431111111111111,
+      "grad_norm": 2.565636396408081,
+      "learning_rate": 0.00010284697508896797,
+      "loss": 1.9388,
+      "step": 5470
+    },
+    {
+      "epoch": 2.4315555555555557,
+      "grad_norm": 2.60259747505188,
+      "learning_rate": 0.00010282918149466192,
+      "loss": 1.5759,
+      "step": 5471
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 2.3194119930267334,
+      "learning_rate": 0.00010281138790035587,
+      "loss": 1.8334,
+      "step": 5472
+    },
+    {
+      "epoch": 2.4324444444444446,
+      "grad_norm": 2.5380783081054688,
+      "learning_rate": 0.00010279359430604982,
+      "loss": 1.7694,
+      "step": 5473
+    },
+    {
+      "epoch": 2.432888888888889,
+      "grad_norm": 2.5655555725097656,
+      "learning_rate": 0.00010277580071174378,
+      "loss": 2.0182,
+      "step": 5474
+    },
+    {
+      "epoch": 2.4333333333333336,
+      "grad_norm": 2.371737241744995,
+      "learning_rate": 0.00010275800711743772,
+      "loss": 1.805,
+      "step": 5475
+    },
+    {
+      "epoch": 2.433777777777778,
+      "grad_norm": 2.5527260303497314,
+      "learning_rate": 0.00010274021352313168,
+      "loss": 1.7959,
+      "step": 5476
+    },
+    {
+      "epoch": 2.434222222222222,
+      "grad_norm": 2.364063262939453,
+      "learning_rate": 0.00010272241992882564,
+      "loss": 1.4738,
+      "step": 5477
+    },
+    {
+      "epoch": 2.4346666666666668,
+      "grad_norm": 3.0714645385742188,
+      "learning_rate": 0.00010270462633451958,
+      "loss": 1.7449,
+      "step": 5478
+    },
+    {
+      "epoch": 2.435111111111111,
+      "grad_norm": 2.63435959815979,
+      "learning_rate": 0.00010268683274021354,
+      "loss": 1.9282,
+      "step": 5479
+    },
+    {
+      "epoch": 2.4355555555555557,
+      "grad_norm": 2.5400550365448,
+      "learning_rate": 0.00010266903914590749,
+      "loss": 1.199,
+      "step": 5480
+    },
+    {
+      "epoch": 2.436,
+      "grad_norm": 2.6152279376983643,
+      "learning_rate": 0.00010265124555160144,
+      "loss": 1.7645,
+      "step": 5481
+    },
+    {
+      "epoch": 2.4364444444444446,
+      "grad_norm": 3.2764761447906494,
+      "learning_rate": 0.00010263345195729536,
+      "loss": 1.8478,
+      "step": 5482
+    },
+    {
+      "epoch": 2.436888888888889,
+      "grad_norm": 2.65189790725708,
+      "learning_rate": 0.00010261565836298932,
+      "loss": 1.969,
+      "step": 5483
+    },
+    {
+      "epoch": 2.437333333333333,
+      "grad_norm": 2.8644859790802,
+      "learning_rate": 0.00010259786476868328,
+      "loss": 1.5599,
+      "step": 5484
+    },
+    {
+      "epoch": 2.437777777777778,
+      "grad_norm": 2.7728471755981445,
+      "learning_rate": 0.00010258007117437722,
+      "loss": 1.4802,
+      "step": 5485
+    },
+    {
+      "epoch": 2.438222222222222,
+      "grad_norm": 3.231962203979492,
+      "learning_rate": 0.00010256227758007118,
+      "loss": 1.8151,
+      "step": 5486
+    },
+    {
+      "epoch": 2.4386666666666668,
+      "grad_norm": 2.6833078861236572,
+      "learning_rate": 0.00010254448398576513,
+      "loss": 1.4914,
+      "step": 5487
+    },
+    {
+      "epoch": 2.439111111111111,
+      "grad_norm": 2.837379217147827,
+      "learning_rate": 0.00010252669039145908,
+      "loss": 2.1413,
+      "step": 5488
+    },
+    {
+      "epoch": 2.4395555555555557,
+      "grad_norm": 2.6518139839172363,
+      "learning_rate": 0.00010250889679715303,
+      "loss": 1.5023,
+      "step": 5489
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.0156285762786865,
+      "learning_rate": 0.00010249110320284699,
+      "loss": 0.9281,
+      "step": 5490
+    },
+    {
+      "epoch": 2.4404444444444446,
+      "grad_norm": 3.069772958755493,
+      "learning_rate": 0.00010247330960854093,
+      "loss": 2.1119,
+      "step": 5491
+    },
+    {
+      "epoch": 2.440888888888889,
+      "grad_norm": 3.10864520072937,
+      "learning_rate": 0.00010245551601423489,
+      "loss": 1.5862,
+      "step": 5492
+    },
+    {
+      "epoch": 2.4413333333333336,
+      "grad_norm": 2.859609365463257,
+      "learning_rate": 0.00010243772241992885,
+      "loss": 1.7583,
+      "step": 5493
+    },
+    {
+      "epoch": 2.441777777777778,
+      "grad_norm": 2.6531405448913574,
+      "learning_rate": 0.00010241992882562279,
+      "loss": 1.4576,
+      "step": 5494
+    },
+    {
+      "epoch": 2.442222222222222,
+      "grad_norm": 3.0856566429138184,
+      "learning_rate": 0.00010240213523131672,
+      "loss": 1.6381,
+      "step": 5495
+    },
+    {
+      "epoch": 2.4426666666666668,
+      "grad_norm": 3.7810099124908447,
+      "learning_rate": 0.00010238434163701067,
+      "loss": 2.2101,
+      "step": 5496
+    },
+    {
+      "epoch": 2.443111111111111,
+      "grad_norm": 3.085653066635132,
+      "learning_rate": 0.00010236654804270463,
+      "loss": 1.7354,
+      "step": 5497
+    },
+    {
+      "epoch": 2.4435555555555557,
+      "grad_norm": 2.9411280155181885,
+      "learning_rate": 0.00010234875444839857,
+      "loss": 1.1896,
+      "step": 5498
+    },
+    {
+      "epoch": 2.444,
+      "grad_norm": 2.1418721675872803,
+      "learning_rate": 0.00010233096085409253,
+      "loss": 1.0698,
+      "step": 5499
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 2.725583076477051,
+      "learning_rate": 0.00010231316725978647,
+      "loss": 0.6198,
+      "step": 5500
+    },
+    {
+      "epoch": 2.444888888888889,
+      "grad_norm": 1.92898690700531,
+      "learning_rate": 0.00010229537366548043,
+      "loss": 2.5948,
+      "step": 5501
+    },
+    {
+      "epoch": 2.445333333333333,
+      "grad_norm": 1.2224948406219482,
+      "learning_rate": 0.00010227758007117439,
+      "loss": 1.0915,
+      "step": 5502
+    },
+    {
+      "epoch": 2.445777777777778,
+      "grad_norm": 1.8230127096176147,
+      "learning_rate": 0.00010225978647686833,
+      "loss": 2.0617,
+      "step": 5503
+    },
+    {
+      "epoch": 2.446222222222222,
+      "grad_norm": 1.861149787902832,
+      "learning_rate": 0.00010224199288256229,
+      "loss": 1.6798,
+      "step": 5504
+    },
+    {
+      "epoch": 2.4466666666666668,
+      "grad_norm": 1.9252985715866089,
+      "learning_rate": 0.00010222419928825624,
+      "loss": 1.9089,
+      "step": 5505
+    },
+    {
+      "epoch": 2.447111111111111,
+      "grad_norm": 2.1263620853424072,
+      "learning_rate": 0.00010220640569395019,
+      "loss": 2.0751,
+      "step": 5506
+    },
+    {
+      "epoch": 2.4475555555555557,
+      "grad_norm": 2.2096424102783203,
+      "learning_rate": 0.00010218861209964414,
+      "loss": 2.0726,
+      "step": 5507
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 2.455075263977051,
+      "learning_rate": 0.00010217081850533807,
+      "loss": 2.2006,
+      "step": 5508
+    },
+    {
+      "epoch": 2.448444444444444,
+      "grad_norm": 2.1597304344177246,
+      "learning_rate": 0.00010215302491103203,
+      "loss": 1.5901,
+      "step": 5509
+    },
+    {
+      "epoch": 2.448888888888889,
+      "grad_norm": 2.4968745708465576,
+      "learning_rate": 0.00010213523131672597,
+      "loss": 1.8659,
+      "step": 5510
+    },
+    {
+      "epoch": 2.449333333333333,
+      "grad_norm": 2.074950695037842,
+      "learning_rate": 0.00010211743772241993,
+      "loss": 2.1795,
+      "step": 5511
+    },
+    {
+      "epoch": 2.449777777777778,
+      "grad_norm": 2.339791774749756,
+      "learning_rate": 0.00010209964412811388,
+      "loss": 1.5882,
+      "step": 5512
+    },
+    {
+      "epoch": 2.450222222222222,
+      "grad_norm": 2.3237624168395996,
+      "learning_rate": 0.00010208185053380783,
+      "loss": 1.4691,
+      "step": 5513
+    },
+    {
+      "epoch": 2.4506666666666668,
+      "grad_norm": 2.4327070713043213,
+      "learning_rate": 0.00010206405693950178,
+      "loss": 1.8784,
+      "step": 5514
+    },
+    {
+      "epoch": 2.451111111111111,
+      "grad_norm": 2.7746636867523193,
+      "learning_rate": 0.00010204626334519574,
+      "loss": 1.7476,
+      "step": 5515
+    },
+    {
+      "epoch": 2.4515555555555557,
+      "grad_norm": 1.4151685237884521,
+      "learning_rate": 0.00010202846975088968,
+      "loss": 0.7311,
+      "step": 5516
+    },
+    {
+      "epoch": 2.452,
+      "grad_norm": 2.369919776916504,
+      "learning_rate": 0.00010201067615658364,
+      "loss": 1.8284,
+      "step": 5517
+    },
+    {
+      "epoch": 2.4524444444444446,
+      "grad_norm": 2.3550519943237305,
+      "learning_rate": 0.0001019928825622776,
+      "loss": 1.4522,
+      "step": 5518
+    },
+    {
+      "epoch": 2.452888888888889,
+      "grad_norm": 2.1800880432128906,
+      "learning_rate": 0.00010197508896797154,
+      "loss": 1.7927,
+      "step": 5519
+    },
+    {
+      "epoch": 2.453333333333333,
+      "grad_norm": 1.9889439344406128,
+      "learning_rate": 0.0001019572953736655,
+      "loss": 1.4447,
+      "step": 5520
+    },
+    {
+      "epoch": 2.453777777777778,
+      "grad_norm": 2.0936267375946045,
+      "learning_rate": 0.00010193950177935942,
+      "loss": 1.6783,
+      "step": 5521
+    },
+    {
+      "epoch": 2.454222222222222,
+      "grad_norm": 2.577599287033081,
+      "learning_rate": 0.00010192170818505338,
+      "loss": 2.0707,
+      "step": 5522
+    },
+    {
+      "epoch": 2.4546666666666668,
+      "grad_norm": 2.550760507583618,
+      "learning_rate": 0.00010190391459074732,
+      "loss": 1.8725,
+      "step": 5523
+    },
+    {
+      "epoch": 2.455111111111111,
+      "grad_norm": 2.492544651031494,
+      "learning_rate": 0.00010188612099644128,
+      "loss": 1.8554,
+      "step": 5524
+    },
+    {
+      "epoch": 2.4555555555555557,
+      "grad_norm": 2.88720703125,
+      "learning_rate": 0.00010186832740213524,
+      "loss": 1.9373,
+      "step": 5525
+    },
+    {
+      "epoch": 2.456,
+      "grad_norm": 2.423215627670288,
+      "learning_rate": 0.00010185053380782918,
+      "loss": 1.8004,
+      "step": 5526
+    },
+    {
+      "epoch": 2.456444444444444,
+      "grad_norm": 2.4208219051361084,
+      "learning_rate": 0.00010183274021352314,
+      "loss": 2.0648,
+      "step": 5527
+    },
+    {
+      "epoch": 2.456888888888889,
+      "grad_norm": 2.6025261878967285,
+      "learning_rate": 0.00010181494661921709,
+      "loss": 1.7313,
+      "step": 5528
+    },
+    {
+      "epoch": 2.457333333333333,
+      "grad_norm": 1.9392775297164917,
+      "learning_rate": 0.00010179715302491104,
+      "loss": 1.209,
+      "step": 5529
+    },
+    {
+      "epoch": 2.457777777777778,
+      "grad_norm": 2.2096610069274902,
+      "learning_rate": 0.00010177935943060499,
+      "loss": 1.6588,
+      "step": 5530
+    },
+    {
+      "epoch": 2.458222222222222,
+      "grad_norm": 2.5169079303741455,
+      "learning_rate": 0.00010176156583629895,
+      "loss": 1.6311,
+      "step": 5531
+    },
+    {
+      "epoch": 2.458666666666667,
+      "grad_norm": 2.4798898696899414,
+      "learning_rate": 0.00010174377224199289,
+      "loss": 1.4915,
+      "step": 5532
+    },
+    {
+      "epoch": 2.459111111111111,
+      "grad_norm": 2.8230345249176025,
+      "learning_rate": 0.00010172597864768685,
+      "loss": 1.7837,
+      "step": 5533
+    },
+    {
+      "epoch": 2.4595555555555557,
+      "grad_norm": 2.252345561981201,
+      "learning_rate": 0.00010170818505338078,
+      "loss": 1.694,
+      "step": 5534
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.8912289142608643,
+      "learning_rate": 0.00010169039145907473,
+      "loss": 2.1645,
+      "step": 5535
+    },
+    {
+      "epoch": 2.4604444444444447,
+      "grad_norm": 2.01261305809021,
+      "learning_rate": 0.00010167259786476868,
+      "loss": 0.7449,
+      "step": 5536
+    },
+    {
+      "epoch": 2.460888888888889,
+      "grad_norm": 2.8200204372406006,
+      "learning_rate": 0.00010165480427046263,
+      "loss": 1.5755,
+      "step": 5537
+    },
+    {
+      "epoch": 2.461333333333333,
+      "grad_norm": 2.4388539791107178,
+      "learning_rate": 0.00010163701067615659,
+      "loss": 1.7643,
+      "step": 5538
+    },
+    {
+      "epoch": 2.461777777777778,
+      "grad_norm": 2.9990639686584473,
+      "learning_rate": 0.00010161921708185053,
+      "loss": 1.7626,
+      "step": 5539
+    },
+    {
+      "epoch": 2.462222222222222,
+      "grad_norm": 2.477353811264038,
+      "learning_rate": 0.00010160142348754449,
+      "loss": 1.3518,
+      "step": 5540
+    },
+    {
+      "epoch": 2.462666666666667,
+      "grad_norm": 2.8211402893066406,
+      "learning_rate": 0.00010158362989323845,
+      "loss": 1.5751,
+      "step": 5541
+    },
+    {
+      "epoch": 2.463111111111111,
+      "grad_norm": 2.7907443046569824,
+      "learning_rate": 0.00010156583629893239,
+      "loss": 2.068,
+      "step": 5542
+    },
+    {
+      "epoch": 2.4635555555555557,
+      "grad_norm": 2.866947650909424,
+      "learning_rate": 0.00010154804270462635,
+      "loss": 1.593,
+      "step": 5543
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 3.0840697288513184,
+      "learning_rate": 0.0001015302491103203,
+      "loss": 1.6982,
+      "step": 5544
+    },
+    {
+      "epoch": 2.464444444444444,
+      "grad_norm": 2.5934946537017822,
+      "learning_rate": 0.00010151245551601424,
+      "loss": 1.6589,
+      "step": 5545
+    },
+    {
+      "epoch": 2.464888888888889,
+      "grad_norm": 3.505406141281128,
+      "learning_rate": 0.0001014946619217082,
+      "loss": 1.4979,
+      "step": 5546
+    },
+    {
+      "epoch": 2.465333333333333,
+      "grad_norm": 3.8022685050964355,
+      "learning_rate": 0.00010147686832740213,
+      "loss": 2.0308,
+      "step": 5547
+    },
+    {
+      "epoch": 2.465777777777778,
+      "grad_norm": 3.6010377407073975,
+      "learning_rate": 0.00010145907473309609,
+      "loss": 1.9663,
+      "step": 5548
+    },
+    {
+      "epoch": 2.466222222222222,
+      "grad_norm": 3.2680211067199707,
+      "learning_rate": 0.00010144128113879003,
+      "loss": 2.0005,
+      "step": 5549
+    },
+    {
+      "epoch": 2.466666666666667,
+      "grad_norm": 4.470943927764893,
+      "learning_rate": 0.00010142348754448399,
+      "loss": 1.4878,
+      "step": 5550
+    },
+    {
+      "epoch": 2.467111111111111,
+      "grad_norm": 1.9788882732391357,
+      "learning_rate": 0.00010140569395017794,
+      "loss": 1.9776,
+      "step": 5551
+    },
+    {
+      "epoch": 2.4675555555555557,
+      "grad_norm": 1.8858304023742676,
+      "learning_rate": 0.00010138790035587189,
+      "loss": 2.3771,
+      "step": 5552
+    },
+    {
+      "epoch": 2.468,
+      "grad_norm": 2.0019304752349854,
+      "learning_rate": 0.00010137010676156584,
+      "loss": 1.513,
+      "step": 5553
+    },
+    {
+      "epoch": 2.4684444444444447,
+      "grad_norm": 2.7883894443511963,
+      "learning_rate": 0.0001013523131672598,
+      "loss": 2.1866,
+      "step": 5554
+    },
+    {
+      "epoch": 2.468888888888889,
+      "grad_norm": 2.3653924465179443,
+      "learning_rate": 0.00010133451957295374,
+      "loss": 1.6028,
+      "step": 5555
+    },
+    {
+      "epoch": 2.469333333333333,
+      "grad_norm": 2.258758544921875,
+      "learning_rate": 0.0001013167259786477,
+      "loss": 1.8427,
+      "step": 5556
+    },
+    {
+      "epoch": 2.469777777777778,
+      "grad_norm": 2.4392614364624023,
+      "learning_rate": 0.00010129893238434165,
+      "loss": 1.4794,
+      "step": 5557
+    },
+    {
+      "epoch": 2.470222222222222,
+      "grad_norm": 2.4021196365356445,
+      "learning_rate": 0.0001012811387900356,
+      "loss": 1.7046,
+      "step": 5558
+    },
+    {
+      "epoch": 2.470666666666667,
+      "grad_norm": 2.3524880409240723,
+      "learning_rate": 0.00010126334519572955,
+      "loss": 1.5135,
+      "step": 5559
+    },
+    {
+      "epoch": 2.471111111111111,
+      "grad_norm": 2.1956753730773926,
+      "learning_rate": 0.00010124555160142348,
+      "loss": 1.4995,
+      "step": 5560
+    },
+    {
+      "epoch": 2.4715555555555557,
+      "grad_norm": 2.2904982566833496,
+      "learning_rate": 0.00010122775800711744,
+      "loss": 1.7763,
+      "step": 5561
+    },
+    {
+      "epoch": 2.472,
+      "grad_norm": 2.2968485355377197,
+      "learning_rate": 0.00010120996441281138,
+      "loss": 1.882,
+      "step": 5562
+    },
+    {
+      "epoch": 2.4724444444444442,
+      "grad_norm": 2.505476951599121,
+      "learning_rate": 0.00010119217081850534,
+      "loss": 1.928,
+      "step": 5563
+    },
+    {
+      "epoch": 2.472888888888889,
+      "grad_norm": 2.5836219787597656,
+      "learning_rate": 0.0001011743772241993,
+      "loss": 2.0092,
+      "step": 5564
+    },
+    {
+      "epoch": 2.473333333333333,
+      "grad_norm": 2.1713929176330566,
+      "learning_rate": 0.00010115658362989324,
+      "loss": 1.6092,
+      "step": 5565
+    },
+    {
+      "epoch": 2.473777777777778,
+      "grad_norm": 2.1562206745147705,
+      "learning_rate": 0.0001011387900355872,
+      "loss": 1.1912,
+      "step": 5566
+    },
+    {
+      "epoch": 2.474222222222222,
+      "grad_norm": 2.7238199710845947,
+      "learning_rate": 0.00010112099644128115,
+      "loss": 1.8131,
+      "step": 5567
+    },
+    {
+      "epoch": 2.474666666666667,
+      "grad_norm": 2.7333614826202393,
+      "learning_rate": 0.0001011032028469751,
+      "loss": 2.0719,
+      "step": 5568
+    },
+    {
+      "epoch": 2.475111111111111,
+      "grad_norm": 1.9092731475830078,
+      "learning_rate": 0.00010108540925266905,
+      "loss": 1.5397,
+      "step": 5569
+    },
+    {
+      "epoch": 2.4755555555555557,
+      "grad_norm": 2.2644593715667725,
+      "learning_rate": 0.00010106761565836301,
+      "loss": 1.5289,
+      "step": 5570
+    },
+    {
+      "epoch": 2.476,
+      "grad_norm": 2.569396734237671,
+      "learning_rate": 0.00010104982206405695,
+      "loss": 1.8126,
+      "step": 5571
+    },
+    {
+      "epoch": 2.4764444444444447,
+      "grad_norm": 2.5156378746032715,
+      "learning_rate": 0.00010103202846975088,
+      "loss": 1.542,
+      "step": 5572
+    },
+    {
+      "epoch": 2.476888888888889,
+      "grad_norm": 2.716794013977051,
+      "learning_rate": 0.00010101423487544484,
+      "loss": 1.9589,
+      "step": 5573
+    },
+    {
+      "epoch": 2.477333333333333,
+      "grad_norm": 2.614689826965332,
+      "learning_rate": 0.0001009964412811388,
+      "loss": 2.0686,
+      "step": 5574
+    },
+    {
+      "epoch": 2.477777777777778,
+      "grad_norm": 3.7736783027648926,
+      "learning_rate": 0.00010097864768683274,
+      "loss": 1.7161,
+      "step": 5575
+    },
+    {
+      "epoch": 2.478222222222222,
+      "grad_norm": 2.7454824447631836,
+      "learning_rate": 0.00010096085409252669,
+      "loss": 2.2613,
+      "step": 5576
+    },
+    {
+      "epoch": 2.478666666666667,
+      "grad_norm": 2.346919298171997,
+      "learning_rate": 0.00010094306049822065,
+      "loss": 2.0465,
+      "step": 5577
+    },
+    {
+      "epoch": 2.479111111111111,
+      "grad_norm": 2.593102216720581,
+      "learning_rate": 0.00010092526690391459,
+      "loss": 1.9165,
+      "step": 5578
+    },
+    {
+      "epoch": 2.4795555555555557,
+      "grad_norm": 2.560601234436035,
+      "learning_rate": 0.00010090747330960855,
+      "loss": 1.8571,
+      "step": 5579
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.8583500385284424,
+      "learning_rate": 0.0001008896797153025,
+      "loss": 1.9994,
+      "step": 5580
+    },
+    {
+      "epoch": 2.4804444444444442,
+      "grad_norm": 2.5867176055908203,
+      "learning_rate": 0.00010087188612099645,
+      "loss": 1.6889,
+      "step": 5581
+    },
+    {
+      "epoch": 2.480888888888889,
+      "grad_norm": 6.32763147354126,
+      "learning_rate": 0.0001008540925266904,
+      "loss": 1.3415,
+      "step": 5582
+    },
+    {
+      "epoch": 2.481333333333333,
+      "grad_norm": 2.52264666557312,
+      "learning_rate": 0.00010083629893238435,
+      "loss": 1.8925,
+      "step": 5583
+    },
+    {
+      "epoch": 2.481777777777778,
+      "grad_norm": 2.677546977996826,
+      "learning_rate": 0.0001008185053380783,
+      "loss": 2.1266,
+      "step": 5584
+    },
+    {
+      "epoch": 2.482222222222222,
+      "grad_norm": 2.429607391357422,
+      "learning_rate": 0.00010080071174377223,
+      "loss": 1.6861,
+      "step": 5585
+    },
+    {
+      "epoch": 2.482666666666667,
+      "grad_norm": 2.5945723056793213,
+      "learning_rate": 0.00010078291814946619,
+      "loss": 1.3189,
+      "step": 5586
+    },
+    {
+      "epoch": 2.483111111111111,
+      "grad_norm": 2.5813865661621094,
+      "learning_rate": 0.00010076512455516015,
+      "loss": 1.6405,
+      "step": 5587
+    },
+    {
+      "epoch": 2.4835555555555557,
+      "grad_norm": 2.870296001434326,
+      "learning_rate": 0.00010074733096085409,
+      "loss": 1.6053,
+      "step": 5588
+    },
+    {
+      "epoch": 2.484,
+      "grad_norm": 3.1349499225616455,
+      "learning_rate": 0.00010072953736654805,
+      "loss": 1.7858,
+      "step": 5589
+    },
+    {
+      "epoch": 2.4844444444444447,
+      "grad_norm": 2.9594576358795166,
+      "learning_rate": 0.00010071174377224199,
+      "loss": 1.4855,
+      "step": 5590
+    },
+    {
+      "epoch": 2.484888888888889,
+      "grad_norm": 2.6506004333496094,
+      "learning_rate": 0.00010069395017793595,
+      "loss": 1.4352,
+      "step": 5591
+    },
+    {
+      "epoch": 2.485333333333333,
+      "grad_norm": 2.739713191986084,
+      "learning_rate": 0.0001006761565836299,
+      "loss": 1.5386,
+      "step": 5592
+    },
+    {
+      "epoch": 2.485777777777778,
+      "grad_norm": 2.450587034225464,
+      "learning_rate": 0.00010065836298932384,
+      "loss": 1.3184,
+      "step": 5593
+    },
+    {
+      "epoch": 2.486222222222222,
+      "grad_norm": 3.1459028720855713,
+      "learning_rate": 0.0001006405693950178,
+      "loss": 1.6811,
+      "step": 5594
+    },
+    {
+      "epoch": 2.486666666666667,
+      "grad_norm": 3.288677215576172,
+      "learning_rate": 0.00010062277580071176,
+      "loss": 1.7088,
+      "step": 5595
+    },
+    {
+      "epoch": 2.487111111111111,
+      "grad_norm": 3.5362205505371094,
+      "learning_rate": 0.0001006049822064057,
+      "loss": 1.4808,
+      "step": 5596
+    },
+    {
+      "epoch": 2.4875555555555557,
+      "grad_norm": 3.3700575828552246,
+      "learning_rate": 0.00010058718861209966,
+      "loss": 1.9626,
+      "step": 5597
+    },
+    {
+      "epoch": 2.488,
+      "grad_norm": 5.008380889892578,
+      "learning_rate": 0.00010056939501779359,
+      "loss": 1.8212,
+      "step": 5598
+    },
+    {
+      "epoch": 2.4884444444444442,
+      "grad_norm": 3.4378342628479004,
+      "learning_rate": 0.00010055160142348754,
+      "loss": 1.8416,
+      "step": 5599
+    },
+    {
+      "epoch": 2.488888888888889,
+      "grad_norm": 3.6014068126678467,
+      "learning_rate": 0.00010053380782918149,
+      "loss": 1.8809,
+      "step": 5600
+    },
+    {
+      "epoch": 2.489333333333333,
+      "grad_norm": 1.8102363348007202,
+      "learning_rate": 0.00010051601423487544,
+      "loss": 2.2235,
+      "step": 5601
+    },
+    {
+      "epoch": 2.489777777777778,
+      "grad_norm": 1.5259066820144653,
+      "learning_rate": 0.0001004982206405694,
+      "loss": 1.4081,
+      "step": 5602
+    },
+    {
+      "epoch": 2.490222222222222,
+      "grad_norm": 1.8424686193466187,
+      "learning_rate": 0.00010048042704626334,
+      "loss": 2.124,
+      "step": 5603
+    },
+    {
+      "epoch": 2.490666666666667,
+      "grad_norm": 2.1134793758392334,
+      "learning_rate": 0.0001004626334519573,
+      "loss": 2.0344,
+      "step": 5604
+    },
+    {
+      "epoch": 2.491111111111111,
+      "grad_norm": 1.8464834690093994,
+      "learning_rate": 0.00010044483985765125,
+      "loss": 1.4534,
+      "step": 5605
+    },
+    {
+      "epoch": 2.4915555555555557,
+      "grad_norm": 2.2974958419799805,
+      "learning_rate": 0.0001004270462633452,
+      "loss": 1.936,
+      "step": 5606
+    },
+    {
+      "epoch": 2.492,
+      "grad_norm": 2.2461228370666504,
+      "learning_rate": 0.00010040925266903915,
+      "loss": 1.9357,
+      "step": 5607
+    },
+    {
+      "epoch": 2.4924444444444447,
+      "grad_norm": 2.2530407905578613,
+      "learning_rate": 0.00010039145907473311,
+      "loss": 2.0635,
+      "step": 5608
+    },
+    {
+      "epoch": 2.492888888888889,
+      "grad_norm": 1.6925837993621826,
+      "learning_rate": 0.00010037366548042705,
+      "loss": 0.9727,
+      "step": 5609
+    },
+    {
+      "epoch": 2.493333333333333,
+      "grad_norm": 1.5781002044677734,
+      "learning_rate": 0.00010035587188612101,
+      "loss": 1.0898,
+      "step": 5610
+    },
+    {
+      "epoch": 2.493777777777778,
+      "grad_norm": 2.560060501098633,
+      "learning_rate": 0.00010033807829181494,
+      "loss": 1.6789,
+      "step": 5611
+    },
+    {
+      "epoch": 2.494222222222222,
+      "grad_norm": 2.309338092803955,
+      "learning_rate": 0.0001003202846975089,
+      "loss": 1.7006,
+      "step": 5612
+    },
+    {
+      "epoch": 2.494666666666667,
+      "grad_norm": 2.1079952716827393,
+      "learning_rate": 0.00010030249110320284,
+      "loss": 1.9404,
+      "step": 5613
+    },
+    {
+      "epoch": 2.495111111111111,
+      "grad_norm": 2.587090253829956,
+      "learning_rate": 0.0001002846975088968,
+      "loss": 2.0718,
+      "step": 5614
+    },
+    {
+      "epoch": 2.4955555555555557,
+      "grad_norm": 2.6044342517852783,
+      "learning_rate": 0.00010026690391459075,
+      "loss": 1.4223,
+      "step": 5615
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 2.266907215118408,
+      "learning_rate": 0.0001002491103202847,
+      "loss": 1.8536,
+      "step": 5616
+    },
+    {
+      "epoch": 2.4964444444444442,
+      "grad_norm": 2.3977086544036865,
+      "learning_rate": 0.00010023131672597865,
+      "loss": 1.5307,
+      "step": 5617
+    },
+    {
+      "epoch": 2.496888888888889,
+      "grad_norm": 2.1894779205322266,
+      "learning_rate": 0.00010021352313167261,
+      "loss": 1.6952,
+      "step": 5618
+    },
+    {
+      "epoch": 2.497333333333333,
+      "grad_norm": 2.3074259757995605,
+      "learning_rate": 0.00010019572953736655,
+      "loss": 1.7691,
+      "step": 5619
+    },
+    {
+      "epoch": 2.497777777777778,
+      "grad_norm": 2.2768285274505615,
+      "learning_rate": 0.00010017793594306051,
+      "loss": 1.696,
+      "step": 5620
+    },
+    {
+      "epoch": 2.498222222222222,
+      "grad_norm": 2.4698994159698486,
+      "learning_rate": 0.00010016014234875446,
+      "loss": 1.4205,
+      "step": 5621
+    },
+    {
+      "epoch": 2.498666666666667,
+      "grad_norm": 2.3182806968688965,
+      "learning_rate": 0.00010014234875444841,
+      "loss": 1.7386,
+      "step": 5622
+    },
+    {
+      "epoch": 2.499111111111111,
+      "grad_norm": 2.212176561355591,
+      "learning_rate": 0.00010012455516014236,
+      "loss": 1.4822,
+      "step": 5623
+    },
+    {
+      "epoch": 2.4995555555555553,
+      "grad_norm": 2.8632688522338867,
+      "learning_rate": 0.00010010676156583629,
+      "loss": 1.9683,
+      "step": 5624
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.298098087310791,
+      "learning_rate": 0.00010008896797153025,
+      "loss": 1.865,
+      "step": 5625
+    },
+    {
+      "epoch": 2.5004444444444447,
+      "grad_norm": 1.8092749118804932,
+      "learning_rate": 0.00010007117437722419,
+      "loss": 0.9924,
+      "step": 5626
+    },
+    {
+      "epoch": 2.500888888888889,
+      "grad_norm": 3.5151941776275635,
+      "learning_rate": 0.00010005338078291815,
+      "loss": 2.314,
+      "step": 5627
+    },
+    {
+      "epoch": 2.501333333333333,
+      "grad_norm": 1.888319730758667,
+      "learning_rate": 0.0001000355871886121,
+      "loss": 0.9312,
+      "step": 5628
+    },
+    {
+      "epoch": 2.501777777777778,
+      "grad_norm": 3.430162191390991,
+      "learning_rate": 0.00010001779359430605,
+      "loss": 1.7719,
+      "step": 5629
+    },
+    {
+      "epoch": 2.502222222222222,
+      "grad_norm": 2.4954025745391846,
+      "learning_rate": 0.0001,
+      "loss": 1.4474,
+      "step": 5630
+    },
+    {
+      "epoch": 2.502666666666667,
+      "grad_norm": 2.7440812587738037,
+      "learning_rate": 9.998220640569396e-05,
+      "loss": 2.1656,
+      "step": 5631
+    },
+    {
+      "epoch": 2.503111111111111,
+      "grad_norm": 2.9336841106414795,
+      "learning_rate": 9.99644128113879e-05,
+      "loss": 1.767,
+      "step": 5632
+    },
+    {
+      "epoch": 2.5035555555555558,
+      "grad_norm": 2.7770421504974365,
+      "learning_rate": 9.994661921708186e-05,
+      "loss": 1.8719,
+      "step": 5633
+    },
+    {
+      "epoch": 2.504,
+      "grad_norm": 3.4708054065704346,
+      "learning_rate": 9.99288256227758e-05,
+      "loss": 2.0225,
+      "step": 5634
+    },
+    {
+      "epoch": 2.5044444444444443,
+      "grad_norm": 3.865091323852539,
+      "learning_rate": 9.991103202846975e-05,
+      "loss": 1.9943,
+      "step": 5635
+    },
+    {
+      "epoch": 2.504888888888889,
+      "grad_norm": 2.653648853302002,
+      "learning_rate": 9.98932384341637e-05,
+      "loss": 1.7085,
+      "step": 5636
+    },
+    {
+      "epoch": 2.505333333333333,
+      "grad_norm": 3.2496232986450195,
+      "learning_rate": 9.987544483985766e-05,
+      "loss": 1.5987,
+      "step": 5637
+    },
+    {
+      "epoch": 2.505777777777778,
+      "grad_norm": 2.717987537384033,
+      "learning_rate": 9.98576512455516e-05,
+      "loss": 1.5691,
+      "step": 5638
+    },
+    {
+      "epoch": 2.506222222222222,
+      "grad_norm": 2.918710470199585,
+      "learning_rate": 9.983985765124556e-05,
+      "loss": 1.9691,
+      "step": 5639
+    },
+    {
+      "epoch": 2.506666666666667,
+      "grad_norm": 3.100358247756958,
+      "learning_rate": 9.98220640569395e-05,
+      "loss": 1.907,
+      "step": 5640
+    },
+    {
+      "epoch": 2.507111111111111,
+      "grad_norm": 2.9432833194732666,
+      "learning_rate": 9.980427046263346e-05,
+      "loss": 1.9512,
+      "step": 5641
+    },
+    {
+      "epoch": 2.5075555555555553,
+      "grad_norm": 3.8566741943359375,
+      "learning_rate": 9.97864768683274e-05,
+      "loss": 2.0891,
+      "step": 5642
+    },
+    {
+      "epoch": 2.508,
+      "grad_norm": 2.724581003189087,
+      "learning_rate": 9.976868327402136e-05,
+      "loss": 1.4336,
+      "step": 5643
+    },
+    {
+      "epoch": 2.5084444444444447,
+      "grad_norm": 2.8649446964263916,
+      "learning_rate": 9.975088967971531e-05,
+      "loss": 1.8689,
+      "step": 5644
+    },
+    {
+      "epoch": 2.508888888888889,
+      "grad_norm": 2.8804287910461426,
+      "learning_rate": 9.973309608540926e-05,
+      "loss": 1.902,
+      "step": 5645
+    },
+    {
+      "epoch": 2.509333333333333,
+      "grad_norm": 2.8557331562042236,
+      "learning_rate": 9.971530249110321e-05,
+      "loss": 1.7273,
+      "step": 5646
+    },
+    {
+      "epoch": 2.509777777777778,
+      "grad_norm": 3.825040340423584,
+      "learning_rate": 9.969750889679716e-05,
+      "loss": 2.0143,
+      "step": 5647
+    },
+    {
+      "epoch": 2.510222222222222,
+      "grad_norm": 2.0822556018829346,
+      "learning_rate": 9.96797153024911e-05,
+      "loss": 0.7991,
+      "step": 5648
+    },
+    {
+      "epoch": 2.510666666666667,
+      "grad_norm": 4.02184534072876,
+      "learning_rate": 9.966192170818506e-05,
+      "loss": 1.6318,
+      "step": 5649
+    },
+    {
+      "epoch": 2.511111111111111,
+      "grad_norm": 3.868868589401245,
+      "learning_rate": 9.964412811387901e-05,
+      "loss": 1.7922,
+      "step": 5650
+    },
+    {
+      "epoch": 2.5115555555555558,
+      "grad_norm": 1.9891340732574463,
+      "learning_rate": 9.962633451957296e-05,
+      "loss": 1.9876,
+      "step": 5651
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 2.9037976264953613,
+      "learning_rate": 9.960854092526691e-05,
+      "loss": 2.0403,
+      "step": 5652
+    },
+    {
+      "epoch": 2.5124444444444443,
+      "grad_norm": 2.0764451026916504,
+      "learning_rate": 9.959074733096086e-05,
+      "loss": 2.0316,
+      "step": 5653
+    },
+    {
+      "epoch": 2.512888888888889,
+      "grad_norm": 2.249202251434326,
+      "learning_rate": 9.957295373665481e-05,
+      "loss": 1.8978,
+      "step": 5654
+    },
+    {
+      "epoch": 2.513333333333333,
+      "grad_norm": 2.4839189052581787,
+      "learning_rate": 9.955516014234875e-05,
+      "loss": 2.0246,
+      "step": 5655
+    },
+    {
+      "epoch": 2.513777777777778,
+      "grad_norm": 2.4105618000030518,
+      "learning_rate": 9.953736654804271e-05,
+      "loss": 1.8856,
+      "step": 5656
+    },
+    {
+      "epoch": 2.514222222222222,
+      "grad_norm": 1.5155545473098755,
+      "learning_rate": 9.951957295373667e-05,
+      "loss": 1.0195,
+      "step": 5657
+    },
+    {
+      "epoch": 2.514666666666667,
+      "grad_norm": 2.674628973007202,
+      "learning_rate": 9.950177935943061e-05,
+      "loss": 2.1171,
+      "step": 5658
+    },
+    {
+      "epoch": 2.515111111111111,
+      "grad_norm": 2.4236528873443604,
+      "learning_rate": 9.948398576512457e-05,
+      "loss": 1.9027,
+      "step": 5659
+    },
+    {
+      "epoch": 2.5155555555555553,
+      "grad_norm": 2.4038853645324707,
+      "learning_rate": 9.946619217081851e-05,
+      "loss": 1.5235,
+      "step": 5660
+    },
+    {
+      "epoch": 2.516,
+      "grad_norm": 2.637871026992798,
+      "learning_rate": 9.944839857651245e-05,
+      "loss": 2.012,
+      "step": 5661
+    },
+    {
+      "epoch": 2.5164444444444447,
+      "grad_norm": 2.1370859146118164,
+      "learning_rate": 9.943060498220641e-05,
+      "loss": 1.6912,
+      "step": 5662
+    },
+    {
+      "epoch": 2.516888888888889,
+      "grad_norm": 1.37175452709198,
+      "learning_rate": 9.941281138790037e-05,
+      "loss": 0.8939,
+      "step": 5663
+    },
+    {
+      "epoch": 2.517333333333333,
+      "grad_norm": 2.60074520111084,
+      "learning_rate": 9.939501779359431e-05,
+      "loss": 2.1094,
+      "step": 5664
+    },
+    {
+      "epoch": 2.517777777777778,
+      "grad_norm": 2.421234607696533,
+      "learning_rate": 9.937722419928827e-05,
+      "loss": 2.2255,
+      "step": 5665
+    },
+    {
+      "epoch": 2.518222222222222,
+      "grad_norm": 1.9047367572784424,
+      "learning_rate": 9.935943060498221e-05,
+      "loss": 1.5189,
+      "step": 5666
+    },
+    {
+      "epoch": 2.518666666666667,
+      "grad_norm": 2.2271924018859863,
+      "learning_rate": 9.934163701067616e-05,
+      "loss": 1.6326,
+      "step": 5667
+    },
+    {
+      "epoch": 2.519111111111111,
+      "grad_norm": 2.0416018962860107,
+      "learning_rate": 9.932384341637011e-05,
+      "loss": 1.1365,
+      "step": 5668
+    },
+    {
+      "epoch": 2.5195555555555558,
+      "grad_norm": 2.345722198486328,
+      "learning_rate": 9.930604982206406e-05,
+      "loss": 1.771,
+      "step": 5669
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.515684127807617,
+      "learning_rate": 9.928825622775802e-05,
+      "loss": 1.5751,
+      "step": 5670
+    },
+    {
+      "epoch": 2.5204444444444443,
+      "grad_norm": 2.2543230056762695,
+      "learning_rate": 9.927046263345196e-05,
+      "loss": 1.505,
+      "step": 5671
+    },
+    {
+      "epoch": 2.520888888888889,
+      "grad_norm": 2.895059585571289,
+      "learning_rate": 9.92526690391459e-05,
+      "loss": 2.2744,
+      "step": 5672
+    },
+    {
+      "epoch": 2.521333333333333,
+      "grad_norm": 2.4043800830841064,
+      "learning_rate": 9.923487544483986e-05,
+      "loss": 1.5109,
+      "step": 5673
+    },
+    {
+      "epoch": 2.521777777777778,
+      "grad_norm": 2.604374885559082,
+      "learning_rate": 9.92170818505338e-05,
+      "loss": 1.5665,
+      "step": 5674
+    },
+    {
+      "epoch": 2.522222222222222,
+      "grad_norm": 2.587221145629883,
+      "learning_rate": 9.919928825622776e-05,
+      "loss": 1.7969,
+      "step": 5675
+    },
+    {
+      "epoch": 2.522666666666667,
+      "grad_norm": 2.516834259033203,
+      "learning_rate": 9.918149466192172e-05,
+      "loss": 1.5522,
+      "step": 5676
+    },
+    {
+      "epoch": 2.523111111111111,
+      "grad_norm": 2.5173187255859375,
+      "learning_rate": 9.916370106761566e-05,
+      "loss": 1.8983,
+      "step": 5677
+    },
+    {
+      "epoch": 2.5235555555555553,
+      "grad_norm": 3.197408676147461,
+      "learning_rate": 9.914590747330962e-05,
+      "loss": 2.0039,
+      "step": 5678
+    },
+    {
+      "epoch": 2.524,
+      "grad_norm": 2.5664446353912354,
+      "learning_rate": 9.912811387900356e-05,
+      "loss": 1.7915,
+      "step": 5679
+    },
+    {
+      "epoch": 2.5244444444444447,
+      "grad_norm": 2.8082244396209717,
+      "learning_rate": 9.91103202846975e-05,
+      "loss": 2.2351,
+      "step": 5680
+    },
+    {
+      "epoch": 2.524888888888889,
+      "grad_norm": 2.381884813308716,
+      "learning_rate": 9.909252669039146e-05,
+      "loss": 1.3298,
+      "step": 5681
+    },
+    {
+      "epoch": 2.525333333333333,
+      "grad_norm": 2.747676134109497,
+      "learning_rate": 9.907473309608542e-05,
+      "loss": 1.758,
+      "step": 5682
+    },
+    {
+      "epoch": 2.525777777777778,
+      "grad_norm": 1.99376380443573,
+      "learning_rate": 9.905693950177936e-05,
+      "loss": 0.815,
+      "step": 5683
+    },
+    {
+      "epoch": 2.526222222222222,
+      "grad_norm": 2.863327980041504,
+      "learning_rate": 9.903914590747332e-05,
+      "loss": 1.9873,
+      "step": 5684
+    },
+    {
+      "epoch": 2.5266666666666664,
+      "grad_norm": 2.060922861099243,
+      "learning_rate": 9.902135231316726e-05,
+      "loss": 1.206,
+      "step": 5685
+    },
+    {
+      "epoch": 2.527111111111111,
+      "grad_norm": 2.845263719558716,
+      "learning_rate": 9.900355871886122e-05,
+      "loss": 1.7081,
+      "step": 5686
+    },
+    {
+      "epoch": 2.5275555555555558,
+      "grad_norm": 2.831609010696411,
+      "learning_rate": 9.898576512455516e-05,
+      "loss": 1.9307,
+      "step": 5687
+    },
+    {
+      "epoch": 2.528,
+      "grad_norm": 2.4219014644622803,
+      "learning_rate": 9.896797153024912e-05,
+      "loss": 1.2403,
+      "step": 5688
+    },
+    {
+      "epoch": 2.5284444444444443,
+      "grad_norm": 3.4156317710876465,
+      "learning_rate": 9.895017793594307e-05,
+      "loss": 1.9255,
+      "step": 5689
+    },
+    {
+      "epoch": 2.528888888888889,
+      "grad_norm": 3.083696126937866,
+      "learning_rate": 9.893238434163702e-05,
+      "loss": 1.5899,
+      "step": 5690
+    },
+    {
+      "epoch": 2.529333333333333,
+      "grad_norm": 3.043922185897827,
+      "learning_rate": 9.891459074733097e-05,
+      "loss": 1.9652,
+      "step": 5691
+    },
+    {
+      "epoch": 2.529777777777778,
+      "grad_norm": 2.8423397541046143,
+      "learning_rate": 9.889679715302491e-05,
+      "loss": 1.7007,
+      "step": 5692
+    },
+    {
+      "epoch": 2.530222222222222,
+      "grad_norm": 3.143592596054077,
+      "learning_rate": 9.887900355871886e-05,
+      "loss": 1.613,
+      "step": 5693
+    },
+    {
+      "epoch": 2.530666666666667,
+      "grad_norm": 3.120520830154419,
+      "learning_rate": 9.886120996441281e-05,
+      "loss": 1.5702,
+      "step": 5694
+    },
+    {
+      "epoch": 2.531111111111111,
+      "grad_norm": 3.2891790866851807,
+      "learning_rate": 9.884341637010677e-05,
+      "loss": 2.051,
+      "step": 5695
+    },
+    {
+      "epoch": 2.5315555555555553,
+      "grad_norm": 2.3981990814208984,
+      "learning_rate": 9.882562277580071e-05,
+      "loss": 1.1692,
+      "step": 5696
+    },
+    {
+      "epoch": 2.532,
+      "grad_norm": 2.9288384914398193,
+      "learning_rate": 9.880782918149467e-05,
+      "loss": 1.5118,
+      "step": 5697
+    },
+    {
+      "epoch": 2.5324444444444447,
+      "grad_norm": 2.981546640396118,
+      "learning_rate": 9.879003558718861e-05,
+      "loss": 1.6747,
+      "step": 5698
+    },
+    {
+      "epoch": 2.532888888888889,
+      "grad_norm": 3.5307023525238037,
+      "learning_rate": 9.877224199288257e-05,
+      "loss": 1.7474,
+      "step": 5699
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 5.0346832275390625,
+      "learning_rate": 9.875444839857651e-05,
+      "loss": 1.8528,
+      "step": 5700
+    },
+    {
+      "epoch": 2.533777777777778,
+      "grad_norm": 1.5125519037246704,
+      "learning_rate": 9.873665480427047e-05,
+      "loss": 1.1769,
+      "step": 5701
+    },
+    {
+      "epoch": 2.534222222222222,
+      "grad_norm": 1.7165669202804565,
+      "learning_rate": 9.871886120996443e-05,
+      "loss": 1.086,
+      "step": 5702
+    },
+    {
+      "epoch": 2.5346666666666664,
+      "grad_norm": 1.4426929950714111,
+      "learning_rate": 9.870106761565837e-05,
+      "loss": 1.1691,
+      "step": 5703
+    },
+    {
+      "epoch": 2.535111111111111,
+      "grad_norm": 2.2641258239746094,
+      "learning_rate": 9.868327402135232e-05,
+      "loss": 2.1713,
+      "step": 5704
+    },
+    {
+      "epoch": 2.535555555555556,
+      "grad_norm": 2.1632139682769775,
+      "learning_rate": 9.866548042704627e-05,
+      "loss": 2.2283,
+      "step": 5705
+    },
+    {
+      "epoch": 2.536,
+      "grad_norm": 1.8564096689224243,
+      "learning_rate": 9.864768683274021e-05,
+      "loss": 1.0078,
+      "step": 5706
+    },
+    {
+      "epoch": 2.5364444444444443,
+      "grad_norm": 2.37397837638855,
+      "learning_rate": 9.862989323843417e-05,
+      "loss": 2.1491,
+      "step": 5707
+    },
+    {
+      "epoch": 2.536888888888889,
+      "grad_norm": 2.2556726932525635,
+      "learning_rate": 9.861209964412812e-05,
+      "loss": 2.0101,
+      "step": 5708
+    },
+    {
+      "epoch": 2.537333333333333,
+      "grad_norm": 2.226167678833008,
+      "learning_rate": 9.859430604982207e-05,
+      "loss": 1.8431,
+      "step": 5709
+    },
+    {
+      "epoch": 2.537777777777778,
+      "grad_norm": 2.111975908279419,
+      "learning_rate": 9.857651245551602e-05,
+      "loss": 1.9402,
+      "step": 5710
+    },
+    {
+      "epoch": 2.538222222222222,
+      "grad_norm": 2.208085060119629,
+      "learning_rate": 9.855871886120997e-05,
+      "loss": 1.8334,
+      "step": 5711
+    },
+    {
+      "epoch": 2.538666666666667,
+      "grad_norm": 2.404080867767334,
+      "learning_rate": 9.854092526690392e-05,
+      "loss": 2.0223,
+      "step": 5712
+    },
+    {
+      "epoch": 2.539111111111111,
+      "grad_norm": 2.602574586868286,
+      "learning_rate": 9.852313167259787e-05,
+      "loss": 2.0228,
+      "step": 5713
+    },
+    {
+      "epoch": 2.5395555555555553,
+      "grad_norm": 2.617043972015381,
+      "learning_rate": 9.850533807829182e-05,
+      "loss": 1.7762,
+      "step": 5714
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.011544704437256,
+      "learning_rate": 9.848754448398578e-05,
+      "loss": 1.5084,
+      "step": 5715
+    },
+    {
+      "epoch": 2.5404444444444443,
+      "grad_norm": 2.117644786834717,
+      "learning_rate": 9.846975088967972e-05,
+      "loss": 1.6064,
+      "step": 5716
+    },
+    {
+      "epoch": 2.540888888888889,
+      "grad_norm": 2.198139190673828,
+      "learning_rate": 9.845195729537368e-05,
+      "loss": 1.6679,
+      "step": 5717
+    },
+    {
+      "epoch": 2.541333333333333,
+      "grad_norm": 1.5015946626663208,
+      "learning_rate": 9.843416370106762e-05,
+      "loss": 0.9008,
+      "step": 5718
+    },
+    {
+      "epoch": 2.541777777777778,
+      "grad_norm": 2.3386454582214355,
+      "learning_rate": 9.841637010676156e-05,
+      "loss": 1.665,
+      "step": 5719
+    },
+    {
+      "epoch": 2.542222222222222,
+      "grad_norm": 2.063887357711792,
+      "learning_rate": 9.839857651245552e-05,
+      "loss": 1.4986,
+      "step": 5720
+    },
+    {
+      "epoch": 2.5426666666666664,
+      "grad_norm": 2.3388946056365967,
+      "learning_rate": 9.838078291814948e-05,
+      "loss": 2.0517,
+      "step": 5721
+    },
+    {
+      "epoch": 2.543111111111111,
+      "grad_norm": 2.838296890258789,
+      "learning_rate": 9.836298932384342e-05,
+      "loss": 1.8289,
+      "step": 5722
+    },
+    {
+      "epoch": 2.543555555555556,
+      "grad_norm": 2.528554677963257,
+      "learning_rate": 9.834519572953738e-05,
+      "loss": 2.0108,
+      "step": 5723
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 3.278167963027954,
+      "learning_rate": 9.832740213523132e-05,
+      "loss": 0.9733,
+      "step": 5724
+    },
+    {
+      "epoch": 2.5444444444444443,
+      "grad_norm": 2.756471633911133,
+      "learning_rate": 9.830960854092526e-05,
+      "loss": 1.7284,
+      "step": 5725
+    },
+    {
+      "epoch": 2.544888888888889,
+      "grad_norm": 2.6623575687408447,
+      "learning_rate": 9.829181494661922e-05,
+      "loss": 1.8741,
+      "step": 5726
+    },
+    {
+      "epoch": 2.5453333333333332,
+      "grad_norm": 2.228694438934326,
+      "learning_rate": 9.827402135231318e-05,
+      "loss": 1.5795,
+      "step": 5727
+    },
+    {
+      "epoch": 2.545777777777778,
+      "grad_norm": 1.9252150058746338,
+      "learning_rate": 9.825622775800712e-05,
+      "loss": 1.038,
+      "step": 5728
+    },
+    {
+      "epoch": 2.546222222222222,
+      "grad_norm": 2.6613807678222656,
+      "learning_rate": 9.823843416370107e-05,
+      "loss": 1.9298,
+      "step": 5729
+    },
+    {
+      "epoch": 2.546666666666667,
+      "grad_norm": 2.613647222518921,
+      "learning_rate": 9.822064056939502e-05,
+      "loss": 1.8275,
+      "step": 5730
+    },
+    {
+      "epoch": 2.547111111111111,
+      "grad_norm": 2.9346303939819336,
+      "learning_rate": 9.820284697508897e-05,
+      "loss": 1.9397,
+      "step": 5731
+    },
+    {
+      "epoch": 2.5475555555555554,
+      "grad_norm": 2.4454238414764404,
+      "learning_rate": 9.818505338078292e-05,
+      "loss": 1.6681,
+      "step": 5732
+    },
+    {
+      "epoch": 2.548,
+      "grad_norm": 2.5040907859802246,
+      "learning_rate": 9.816725978647687e-05,
+      "loss": 1.2844,
+      "step": 5733
+    },
+    {
+      "epoch": 2.5484444444444443,
+      "grad_norm": 2.631068706512451,
+      "learning_rate": 9.814946619217083e-05,
+      "loss": 1.482,
+      "step": 5734
+    },
+    {
+      "epoch": 2.548888888888889,
+      "grad_norm": 2.5751149654388428,
+      "learning_rate": 9.813167259786477e-05,
+      "loss": 1.7342,
+      "step": 5735
+    },
+    {
+      "epoch": 2.5493333333333332,
+      "grad_norm": 2.755535840988159,
+      "learning_rate": 9.811387900355873e-05,
+      "loss": 1.775,
+      "step": 5736
+    },
+    {
+      "epoch": 2.549777777777778,
+      "grad_norm": 3.007052421569824,
+      "learning_rate": 9.809608540925267e-05,
+      "loss": 1.8263,
+      "step": 5737
+    },
+    {
+      "epoch": 2.550222222222222,
+      "grad_norm": 2.7909440994262695,
+      "learning_rate": 9.807829181494662e-05,
+      "loss": 1.7401,
+      "step": 5738
+    },
+    {
+      "epoch": 2.5506666666666664,
+      "grad_norm": 3.113389492034912,
+      "learning_rate": 9.806049822064057e-05,
+      "loss": 1.4822,
+      "step": 5739
+    },
+    {
+      "epoch": 2.551111111111111,
+      "grad_norm": 3.607086420059204,
+      "learning_rate": 9.804270462633453e-05,
+      "loss": 2.3029,
+      "step": 5740
+    },
+    {
+      "epoch": 2.551555555555556,
+      "grad_norm": 2.5248825550079346,
+      "learning_rate": 9.802491103202847e-05,
+      "loss": 1.6365,
+      "step": 5741
+    },
+    {
+      "epoch": 2.552,
+      "grad_norm": 3.1107113361358643,
+      "learning_rate": 9.800711743772243e-05,
+      "loss": 1.6821,
+      "step": 5742
+    },
+    {
+      "epoch": 2.5524444444444443,
+      "grad_norm": 3.095695734024048,
+      "learning_rate": 9.798932384341637e-05,
+      "loss": 1.9611,
+      "step": 5743
+    },
+    {
+      "epoch": 2.552888888888889,
+      "grad_norm": 3.119293212890625,
+      "learning_rate": 9.797153024911033e-05,
+      "loss": 1.8032,
+      "step": 5744
+    },
+    {
+      "epoch": 2.5533333333333332,
+      "grad_norm": 3.1017251014709473,
+      "learning_rate": 9.795373665480427e-05,
+      "loss": 1.3616,
+      "step": 5745
+    },
+    {
+      "epoch": 2.553777777777778,
+      "grad_norm": 3.225269317626953,
+      "learning_rate": 9.793594306049823e-05,
+      "loss": 1.6475,
+      "step": 5746
+    },
+    {
+      "epoch": 2.554222222222222,
+      "grad_norm": 3.819324254989624,
+      "learning_rate": 9.791814946619218e-05,
+      "loss": 1.8154,
+      "step": 5747
+    },
+    {
+      "epoch": 2.554666666666667,
+      "grad_norm": 1.9647773504257202,
+      "learning_rate": 9.790035587188613e-05,
+      "loss": 0.7431,
+      "step": 5748
+    },
+    {
+      "epoch": 2.555111111111111,
+      "grad_norm": 3.4745988845825195,
+      "learning_rate": 9.788256227758008e-05,
+      "loss": 1.7861,
+      "step": 5749
+    },
+    {
+      "epoch": 2.5555555555555554,
+      "grad_norm": 4.805171489715576,
+      "learning_rate": 9.786476868327403e-05,
+      "loss": 1.0196,
+      "step": 5750
+    },
+    {
+      "epoch": 2.556,
+      "grad_norm": 1.983173131942749,
+      "learning_rate": 9.784697508896797e-05,
+      "loss": 2.2983,
+      "step": 5751
+    },
+    {
+      "epoch": 2.5564444444444443,
+      "grad_norm": 1.9651044607162476,
+      "learning_rate": 9.782918149466193e-05,
+      "loss": 1.9459,
+      "step": 5752
+    },
+    {
+      "epoch": 2.556888888888889,
+      "grad_norm": 2.155700445175171,
+      "learning_rate": 9.781138790035588e-05,
+      "loss": 2.1677,
+      "step": 5753
+    },
+    {
+      "epoch": 2.5573333333333332,
+      "grad_norm": 0.1379898637533188,
+      "learning_rate": 9.779359430604982e-05,
+      "loss": 0.0167,
+      "step": 5754
+    },
+    {
+      "epoch": 2.557777777777778,
+      "grad_norm": 2.319441080093384,
+      "learning_rate": 9.777580071174378e-05,
+      "loss": 2.1024,
+      "step": 5755
+    },
+    {
+      "epoch": 2.558222222222222,
+      "grad_norm": 2.3348388671875,
+      "learning_rate": 9.775800711743772e-05,
+      "loss": 2.4375,
+      "step": 5756
+    },
+    {
+      "epoch": 2.5586666666666664,
+      "grad_norm": 2.42160964012146,
+      "learning_rate": 9.774021352313168e-05,
+      "loss": 2.0312,
+      "step": 5757
+    },
+    {
+      "epoch": 2.559111111111111,
+      "grad_norm": 2.4215610027313232,
+      "learning_rate": 9.772241992882562e-05,
+      "loss": 2.1268,
+      "step": 5758
+    },
+    {
+      "epoch": 2.559555555555556,
+      "grad_norm": 1.9341044425964355,
+      "learning_rate": 9.770462633451958e-05,
+      "loss": 1.6431,
+      "step": 5759
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.569185972213745,
+      "learning_rate": 9.768683274021354e-05,
+      "loss": 1.9818,
+      "step": 5760
+    },
+    {
+      "epoch": 2.5604444444444443,
+      "grad_norm": 2.380908727645874,
+      "learning_rate": 9.766903914590748e-05,
+      "loss": 1.6281,
+      "step": 5761
+    },
+    {
+      "epoch": 2.560888888888889,
+      "grad_norm": 2.445352554321289,
+      "learning_rate": 9.765124555160144e-05,
+      "loss": 1.7746,
+      "step": 5762
+    },
+    {
+      "epoch": 2.5613333333333332,
+      "grad_norm": 1.9633126258850098,
+      "learning_rate": 9.763345195729538e-05,
+      "loss": 1.2796,
+      "step": 5763
+    },
+    {
+      "epoch": 2.561777777777778,
+      "grad_norm": 2.4921209812164307,
+      "learning_rate": 9.761565836298932e-05,
+      "loss": 2.1531,
+      "step": 5764
+    },
+    {
+      "epoch": 2.562222222222222,
+      "grad_norm": 1.6959223747253418,
+      "learning_rate": 9.759786476868328e-05,
+      "loss": 0.8272,
+      "step": 5765
+    },
+    {
+      "epoch": 2.562666666666667,
+      "grad_norm": 2.389455556869507,
+      "learning_rate": 9.758007117437723e-05,
+      "loss": 1.7277,
+      "step": 5766
+    },
+    {
+      "epoch": 2.563111111111111,
+      "grad_norm": 1.7974770069122314,
+      "learning_rate": 9.756227758007118e-05,
+      "loss": 1.1353,
+      "step": 5767
+    },
+    {
+      "epoch": 2.5635555555555554,
+      "grad_norm": 2.2047617435455322,
+      "learning_rate": 9.754448398576513e-05,
+      "loss": 1.7808,
+      "step": 5768
+    },
+    {
+      "epoch": 2.564,
+      "grad_norm": 3.0488312244415283,
+      "learning_rate": 9.752669039145908e-05,
+      "loss": 2.2577,
+      "step": 5769
+    },
+    {
+      "epoch": 2.5644444444444443,
+      "grad_norm": 2.719773054122925,
+      "learning_rate": 9.750889679715302e-05,
+      "loss": 2.1927,
+      "step": 5770
+    },
+    {
+      "epoch": 2.564888888888889,
+      "grad_norm": 2.654473304748535,
+      "learning_rate": 9.749110320284698e-05,
+      "loss": 1.8773,
+      "step": 5771
+    },
+    {
+      "epoch": 2.5653333333333332,
+      "grad_norm": 2.6189746856689453,
+      "learning_rate": 9.747330960854093e-05,
+      "loss": 1.5834,
+      "step": 5772
+    },
+    {
+      "epoch": 2.565777777777778,
+      "grad_norm": 2.9932594299316406,
+      "learning_rate": 9.745551601423488e-05,
+      "loss": 2.3835,
+      "step": 5773
+    },
+    {
+      "epoch": 2.566222222222222,
+      "grad_norm": 2.437887668609619,
+      "learning_rate": 9.743772241992883e-05,
+      "loss": 1.7243,
+      "step": 5774
+    },
+    {
+      "epoch": 2.5666666666666664,
+      "grad_norm": 2.175389051437378,
+      "learning_rate": 9.741992882562279e-05,
+      "loss": 1.4495,
+      "step": 5775
+    },
+    {
+      "epoch": 2.567111111111111,
+      "grad_norm": 2.9918575286865234,
+      "learning_rate": 9.740213523131673e-05,
+      "loss": 2.0336,
+      "step": 5776
+    },
+    {
+      "epoch": 2.567555555555556,
+      "grad_norm": 2.525195360183716,
+      "learning_rate": 9.738434163701067e-05,
+      "loss": 1.6138,
+      "step": 5777
+    },
+    {
+      "epoch": 2.568,
+      "grad_norm": 2.7021284103393555,
+      "learning_rate": 9.736654804270463e-05,
+      "loss": 1.7927,
+      "step": 5778
+    },
+    {
+      "epoch": 2.5684444444444443,
+      "grad_norm": 2.420149087905884,
+      "learning_rate": 9.734875444839859e-05,
+      "loss": 1.7339,
+      "step": 5779
+    },
+    {
+      "epoch": 2.568888888888889,
+      "grad_norm": 2.3807191848754883,
+      "learning_rate": 9.733096085409253e-05,
+      "loss": 0.6999,
+      "step": 5780
+    },
+    {
+      "epoch": 2.5693333333333332,
+      "grad_norm": 2.4973058700561523,
+      "learning_rate": 9.731316725978649e-05,
+      "loss": 1.6894,
+      "step": 5781
+    },
+    {
+      "epoch": 2.569777777777778,
+      "grad_norm": 2.848585844039917,
+      "learning_rate": 9.729537366548043e-05,
+      "loss": 1.9012,
+      "step": 5782
+    },
+    {
+      "epoch": 2.570222222222222,
+      "grad_norm": 2.616931676864624,
+      "learning_rate": 9.727758007117437e-05,
+      "loss": 1.5713,
+      "step": 5783
+    },
+    {
+      "epoch": 2.570666666666667,
+      "grad_norm": 2.4025752544403076,
+      "learning_rate": 9.725978647686833e-05,
+      "loss": 1.6347,
+      "step": 5784
+    },
+    {
+      "epoch": 2.571111111111111,
+      "grad_norm": 2.759446620941162,
+      "learning_rate": 9.724199288256229e-05,
+      "loss": 1.5939,
+      "step": 5785
+    },
+    {
+      "epoch": 2.5715555555555554,
+      "grad_norm": 2.421267032623291,
+      "learning_rate": 9.722419928825623e-05,
+      "loss": 1.3653,
+      "step": 5786
+    },
+    {
+      "epoch": 2.572,
+      "grad_norm": 3.161170482635498,
+      "learning_rate": 9.720640569395019e-05,
+      "loss": 1.5433,
+      "step": 5787
+    },
+    {
+      "epoch": 2.5724444444444443,
+      "grad_norm": 3.038508653640747,
+      "learning_rate": 9.718861209964413e-05,
+      "loss": 1.4943,
+      "step": 5788
+    },
+    {
+      "epoch": 2.572888888888889,
+      "grad_norm": 2.3798742294311523,
+      "learning_rate": 9.717081850533809e-05,
+      "loss": 1.1891,
+      "step": 5789
+    },
+    {
+      "epoch": 2.5733333333333333,
+      "grad_norm": 2.9611308574676514,
+      "learning_rate": 9.715302491103203e-05,
+      "loss": 1.6984,
+      "step": 5790
+    },
+    {
+      "epoch": 2.573777777777778,
+      "grad_norm": 3.015956401824951,
+      "learning_rate": 9.713523131672598e-05,
+      "loss": 1.6971,
+      "step": 5791
+    },
+    {
+      "epoch": 2.574222222222222,
+      "grad_norm": 2.819864273071289,
+      "learning_rate": 9.711743772241994e-05,
+      "loss": 1.5978,
+      "step": 5792
+    },
+    {
+      "epoch": 2.5746666666666664,
+      "grad_norm": 3.813222885131836,
+      "learning_rate": 9.709964412811388e-05,
+      "loss": 1.9956,
+      "step": 5793
+    },
+    {
+      "epoch": 2.575111111111111,
+      "grad_norm": 2.9637928009033203,
+      "learning_rate": 9.708185053380784e-05,
+      "loss": 1.7197,
+      "step": 5794
+    },
+    {
+      "epoch": 2.575555555555556,
+      "grad_norm": 3.4697177410125732,
+      "learning_rate": 9.706405693950178e-05,
+      "loss": 1.3374,
+      "step": 5795
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 3.505645751953125,
+      "learning_rate": 9.704626334519573e-05,
+      "loss": 1.9452,
+      "step": 5796
+    },
+    {
+      "epoch": 2.5764444444444443,
+      "grad_norm": 3.6183602809906006,
+      "learning_rate": 9.702846975088968e-05,
+      "loss": 1.6057,
+      "step": 5797
+    },
+    {
+      "epoch": 2.576888888888889,
+      "grad_norm": 3.725675106048584,
+      "learning_rate": 9.701067615658364e-05,
+      "loss": 1.6473,
+      "step": 5798
+    },
+    {
+      "epoch": 2.5773333333333333,
+      "grad_norm": 2.1364715099334717,
+      "learning_rate": 9.699288256227758e-05,
+      "loss": 0.8211,
+      "step": 5799
+    },
+    {
+      "epoch": 2.5777777777777775,
+      "grad_norm": 4.8051934242248535,
+      "learning_rate": 9.697508896797154e-05,
+      "loss": 1.2673,
+      "step": 5800
+    },
+    {
+      "epoch": 2.578222222222222,
+      "grad_norm": 1.737740397453308,
+      "learning_rate": 9.695729537366548e-05,
+      "loss": 2.0515,
+      "step": 5801
+    },
+    {
+      "epoch": 2.578666666666667,
+      "grad_norm": 1.431738018989563,
+      "learning_rate": 9.693950177935944e-05,
+      "loss": 1.0424,
+      "step": 5802
+    },
+    {
+      "epoch": 2.579111111111111,
+      "grad_norm": 1.8884087800979614,
+      "learning_rate": 9.692170818505338e-05,
+      "loss": 2.2214,
+      "step": 5803
+    },
+    {
+      "epoch": 2.5795555555555554,
+      "grad_norm": 2.2000203132629395,
+      "learning_rate": 9.690391459074734e-05,
+      "loss": 2.3403,
+      "step": 5804
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.952431559562683,
+      "learning_rate": 9.68861209964413e-05,
+      "loss": 2.2942,
+      "step": 5805
+    },
+    {
+      "epoch": 2.5804444444444443,
+      "grad_norm": 2.162529706954956,
+      "learning_rate": 9.686832740213524e-05,
+      "loss": 2.1921,
+      "step": 5806
+    },
+    {
+      "epoch": 2.580888888888889,
+      "grad_norm": 2.1220896244049072,
+      "learning_rate": 9.68505338078292e-05,
+      "loss": 1.914,
+      "step": 5807
+    },
+    {
+      "epoch": 2.5813333333333333,
+      "grad_norm": 1.9332010746002197,
+      "learning_rate": 9.683274021352314e-05,
+      "loss": 1.4451,
+      "step": 5808
+    },
+    {
+      "epoch": 2.581777777777778,
+      "grad_norm": 2.8427624702453613,
+      "learning_rate": 9.681494661921708e-05,
+      "loss": 2.0637,
+      "step": 5809
+    },
+    {
+      "epoch": 2.582222222222222,
+      "grad_norm": 1.7098627090454102,
+      "learning_rate": 9.679715302491104e-05,
+      "loss": 1.2616,
+      "step": 5810
+    },
+    {
+      "epoch": 2.5826666666666664,
+      "grad_norm": 2.2705390453338623,
+      "learning_rate": 9.677935943060499e-05,
+      "loss": 1.9668,
+      "step": 5811
+    },
+    {
+      "epoch": 2.583111111111111,
+      "grad_norm": 2.204101800918579,
+      "learning_rate": 9.676156583629894e-05,
+      "loss": 1.8447,
+      "step": 5812
+    },
+    {
+      "epoch": 2.583555555555556,
+      "grad_norm": 2.294118642807007,
+      "learning_rate": 9.674377224199289e-05,
+      "loss": 1.7681,
+      "step": 5813
+    },
+    {
+      "epoch": 2.584,
+      "grad_norm": 2.175595760345459,
+      "learning_rate": 9.672597864768683e-05,
+      "loss": 1.9735,
+      "step": 5814
+    },
+    {
+      "epoch": 2.5844444444444443,
+      "grad_norm": 2.1814606189727783,
+      "learning_rate": 9.670818505338078e-05,
+      "loss": 1.709,
+      "step": 5815
+    },
+    {
+      "epoch": 2.584888888888889,
+      "grad_norm": 2.43196702003479,
+      "learning_rate": 9.669039145907473e-05,
+      "loss": 2.0574,
+      "step": 5816
+    },
+    {
+      "epoch": 2.5853333333333333,
+      "grad_norm": 2.466120958328247,
+      "learning_rate": 9.667259786476869e-05,
+      "loss": 2.2044,
+      "step": 5817
+    },
+    {
+      "epoch": 2.5857777777777775,
+      "grad_norm": 1.9092581272125244,
+      "learning_rate": 9.665480427046263e-05,
+      "loss": 0.5758,
+      "step": 5818
+    },
+    {
+      "epoch": 2.586222222222222,
+      "grad_norm": 2.337799310684204,
+      "learning_rate": 9.663701067615659e-05,
+      "loss": 2.1761,
+      "step": 5819
+    },
+    {
+      "epoch": 2.586666666666667,
+      "grad_norm": 2.0676565170288086,
+      "learning_rate": 9.661921708185055e-05,
+      "loss": 1.7785,
+      "step": 5820
+    },
+    {
+      "epoch": 2.587111111111111,
+      "grad_norm": 2.228407382965088,
+      "learning_rate": 9.660142348754449e-05,
+      "loss": 1.3814,
+      "step": 5821
+    },
+    {
+      "epoch": 2.5875555555555554,
+      "grad_norm": 2.3954479694366455,
+      "learning_rate": 9.658362989323843e-05,
+      "loss": 1.3558,
+      "step": 5822
+    },
+    {
+      "epoch": 2.588,
+      "grad_norm": 2.1830084323883057,
+      "learning_rate": 9.656583629893239e-05,
+      "loss": 1.8961,
+      "step": 5823
+    },
+    {
+      "epoch": 2.5884444444444443,
+      "grad_norm": 2.57344913482666,
+      "learning_rate": 9.654804270462635e-05,
+      "loss": 1.8005,
+      "step": 5824
+    },
+    {
+      "epoch": 2.588888888888889,
+      "grad_norm": 2.3932812213897705,
+      "learning_rate": 9.653024911032029e-05,
+      "loss": 1.824,
+      "step": 5825
+    },
+    {
+      "epoch": 2.5893333333333333,
+      "grad_norm": 2.067878007888794,
+      "learning_rate": 9.651245551601425e-05,
+      "loss": 1.7157,
+      "step": 5826
+    },
+    {
+      "epoch": 2.589777777777778,
+      "grad_norm": 2.737030267715454,
+      "learning_rate": 9.649466192170819e-05,
+      "loss": 1.3283,
+      "step": 5827
+    },
+    {
+      "epoch": 2.590222222222222,
+      "grad_norm": 2.7902398109436035,
+      "learning_rate": 9.647686832740213e-05,
+      "loss": 1.8529,
+      "step": 5828
+    },
+    {
+      "epoch": 2.5906666666666665,
+      "grad_norm": 2.5697999000549316,
+      "learning_rate": 9.645907473309609e-05,
+      "loss": 1.7706,
+      "step": 5829
+    },
+    {
+      "epoch": 2.591111111111111,
+      "grad_norm": 2.486321449279785,
+      "learning_rate": 9.644128113879004e-05,
+      "loss": 1.4964,
+      "step": 5830
+    },
+    {
+      "epoch": 2.5915555555555554,
+      "grad_norm": 2.440593719482422,
+      "learning_rate": 9.642348754448399e-05,
+      "loss": 1.732,
+      "step": 5831
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 2.2928342819213867,
+      "learning_rate": 9.640569395017794e-05,
+      "loss": 1.0141,
+      "step": 5832
+    },
+    {
+      "epoch": 2.5924444444444443,
+      "grad_norm": 2.5680768489837646,
+      "learning_rate": 9.63879003558719e-05,
+      "loss": 1.3784,
+      "step": 5833
+    },
+    {
+      "epoch": 2.592888888888889,
+      "grad_norm": 2.402845621109009,
+      "learning_rate": 9.637010676156584e-05,
+      "loss": 1.7085,
+      "step": 5834
+    },
+    {
+      "epoch": 2.5933333333333333,
+      "grad_norm": 1.819101333618164,
+      "learning_rate": 9.635231316725979e-05,
+      "loss": 0.8828,
+      "step": 5835
+    },
+    {
+      "epoch": 2.5937777777777775,
+      "grad_norm": 2.9509596824645996,
+      "learning_rate": 9.633451957295374e-05,
+      "loss": 1.4475,
+      "step": 5836
+    },
+    {
+      "epoch": 2.594222222222222,
+      "grad_norm": 2.3825151920318604,
+      "learning_rate": 9.63167259786477e-05,
+      "loss": 1.2385,
+      "step": 5837
+    },
+    {
+      "epoch": 2.594666666666667,
+      "grad_norm": 3.0059218406677246,
+      "learning_rate": 9.629893238434164e-05,
+      "loss": 1.6829,
+      "step": 5838
+    },
+    {
+      "epoch": 2.595111111111111,
+      "grad_norm": 3.277036190032959,
+      "learning_rate": 9.62811387900356e-05,
+      "loss": 1.6309,
+      "step": 5839
+    },
+    {
+      "epoch": 2.5955555555555554,
+      "grad_norm": 2.9145870208740234,
+      "learning_rate": 9.626334519572954e-05,
+      "loss": 2.0603,
+      "step": 5840
+    },
+    {
+      "epoch": 2.596,
+      "grad_norm": 2.803736686706543,
+      "learning_rate": 9.624555160142348e-05,
+      "loss": 1.6505,
+      "step": 5841
+    },
+    {
+      "epoch": 2.5964444444444443,
+      "grad_norm": 3.3242056369781494,
+      "learning_rate": 9.622775800711744e-05,
+      "loss": 1.9847,
+      "step": 5842
+    },
+    {
+      "epoch": 2.596888888888889,
+      "grad_norm": 3.6770615577697754,
+      "learning_rate": 9.62099644128114e-05,
+      "loss": 1.4115,
+      "step": 5843
+    },
+    {
+      "epoch": 2.5973333333333333,
+      "grad_norm": 2.989746332168579,
+      "learning_rate": 9.619217081850534e-05,
+      "loss": 1.7387,
+      "step": 5844
+    },
+    {
+      "epoch": 2.597777777777778,
+      "grad_norm": 3.3722546100616455,
+      "learning_rate": 9.61743772241993e-05,
+      "loss": 1.6478,
+      "step": 5845
+    },
+    {
+      "epoch": 2.598222222222222,
+      "grad_norm": 3.2518842220306396,
+      "learning_rate": 9.615658362989324e-05,
+      "loss": 1.1471,
+      "step": 5846
+    },
+    {
+      "epoch": 2.5986666666666665,
+      "grad_norm": 3.7409679889678955,
+      "learning_rate": 9.61387900355872e-05,
+      "loss": 1.9251,
+      "step": 5847
+    },
+    {
+      "epoch": 2.599111111111111,
+      "grad_norm": 3.091366767883301,
+      "learning_rate": 9.612099644128114e-05,
+      "loss": 1.5747,
+      "step": 5848
+    },
+    {
+      "epoch": 2.5995555555555554,
+      "grad_norm": 3.8686630725860596,
+      "learning_rate": 9.61032028469751e-05,
+      "loss": 1.0119,
+      "step": 5849
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.7347426414489746,
+      "learning_rate": 9.608540925266905e-05,
+      "loss": 0.4375,
+      "step": 5850
+    },
+    {
+      "epoch": 2.6004444444444443,
+      "grad_norm": 1.894957184791565,
+      "learning_rate": 9.6067615658363e-05,
+      "loss": 2.3036,
+      "step": 5851
+    },
+    {
+      "epoch": 2.600888888888889,
+      "grad_norm": 1.9998407363891602,
+      "learning_rate": 9.604982206405695e-05,
+      "loss": 2.2295,
+      "step": 5852
+    },
+    {
+      "epoch": 2.6013333333333333,
+      "grad_norm": 2.1077568531036377,
+      "learning_rate": 9.60320284697509e-05,
+      "loss": 1.085,
+      "step": 5853
+    },
+    {
+      "epoch": 2.6017777777777775,
+      "grad_norm": 2.0301358699798584,
+      "learning_rate": 9.601423487544484e-05,
+      "loss": 1.1873,
+      "step": 5854
+    },
+    {
+      "epoch": 2.602222222222222,
+      "grad_norm": 2.020034074783325,
+      "learning_rate": 9.59964412811388e-05,
+      "loss": 1.7447,
+      "step": 5855
+    },
+    {
+      "epoch": 2.602666666666667,
+      "grad_norm": 2.428344964981079,
+      "learning_rate": 9.597864768683275e-05,
+      "loss": 2.155,
+      "step": 5856
+    },
+    {
+      "epoch": 2.603111111111111,
+      "grad_norm": 2.2994885444641113,
+      "learning_rate": 9.596085409252669e-05,
+      "loss": 2.0446,
+      "step": 5857
+    },
+    {
+      "epoch": 2.6035555555555554,
+      "grad_norm": 2.4764015674591064,
+      "learning_rate": 9.594306049822065e-05,
+      "loss": 1.8314,
+      "step": 5858
+    },
+    {
+      "epoch": 2.604,
+      "grad_norm": 2.4663918018341064,
+      "learning_rate": 9.592526690391459e-05,
+      "loss": 1.6086,
+      "step": 5859
+    },
+    {
+      "epoch": 2.6044444444444443,
+      "grad_norm": 2.3457717895507812,
+      "learning_rate": 9.590747330960854e-05,
+      "loss": 1.8214,
+      "step": 5860
+    },
+    {
+      "epoch": 2.604888888888889,
+      "grad_norm": 1.8627485036849976,
+      "learning_rate": 9.588967971530249e-05,
+      "loss": 1.225,
+      "step": 5861
+    },
+    {
+      "epoch": 2.6053333333333333,
+      "grad_norm": 2.610481023788452,
+      "learning_rate": 9.587188612099645e-05,
+      "loss": 1.8318,
+      "step": 5862
+    },
+    {
+      "epoch": 2.605777777777778,
+      "grad_norm": 2.533219337463379,
+      "learning_rate": 9.585409252669039e-05,
+      "loss": 1.8934,
+      "step": 5863
+    },
+    {
+      "epoch": 2.606222222222222,
+      "grad_norm": 2.506080389022827,
+      "learning_rate": 9.583629893238435e-05,
+      "loss": 2.0397,
+      "step": 5864
+    },
+    {
+      "epoch": 2.6066666666666665,
+      "grad_norm": 2.50927734375,
+      "learning_rate": 9.58185053380783e-05,
+      "loss": 1.8956,
+      "step": 5865
+    },
+    {
+      "epoch": 2.607111111111111,
+      "grad_norm": 2.1902568340301514,
+      "learning_rate": 9.580071174377225e-05,
+      "loss": 1.7044,
+      "step": 5866
+    },
+    {
+      "epoch": 2.6075555555555554,
+      "grad_norm": 2.695941925048828,
+      "learning_rate": 9.578291814946619e-05,
+      "loss": 1.936,
+      "step": 5867
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 2.4527158737182617,
+      "learning_rate": 9.576512455516015e-05,
+      "loss": 1.4728,
+      "step": 5868
+    },
+    {
+      "epoch": 2.6084444444444443,
+      "grad_norm": 2.4402294158935547,
+      "learning_rate": 9.57473309608541e-05,
+      "loss": 1.6587,
+      "step": 5869
+    },
+    {
+      "epoch": 2.608888888888889,
+      "grad_norm": 2.2047953605651855,
+      "learning_rate": 9.572953736654805e-05,
+      "loss": 1.5533,
+      "step": 5870
+    },
+    {
+      "epoch": 2.6093333333333333,
+      "grad_norm": 2.1342928409576416,
+      "learning_rate": 9.5711743772242e-05,
+      "loss": 1.6553,
+      "step": 5871
+    },
+    {
+      "epoch": 2.6097777777777775,
+      "grad_norm": 2.7137703895568848,
+      "learning_rate": 9.569395017793595e-05,
+      "loss": 2.0063,
+      "step": 5872
+    },
+    {
+      "epoch": 2.610222222222222,
+      "grad_norm": 2.3683669567108154,
+      "learning_rate": 9.567615658362989e-05,
+      "loss": 1.4502,
+      "step": 5873
+    },
+    {
+      "epoch": 2.610666666666667,
+      "grad_norm": 2.7055866718292236,
+      "learning_rate": 9.565836298932385e-05,
+      "loss": 2.0081,
+      "step": 5874
+    },
+    {
+      "epoch": 2.611111111111111,
+      "grad_norm": 2.435533046722412,
+      "learning_rate": 9.56405693950178e-05,
+      "loss": 1.4356,
+      "step": 5875
+    },
+    {
+      "epoch": 2.6115555555555554,
+      "grad_norm": 2.407223701477051,
+      "learning_rate": 9.562277580071174e-05,
+      "loss": 1.4788,
+      "step": 5876
+    },
+    {
+      "epoch": 2.612,
+      "grad_norm": 2.53997540473938,
+      "learning_rate": 9.56049822064057e-05,
+      "loss": 1.6948,
+      "step": 5877
+    },
+    {
+      "epoch": 2.6124444444444443,
+      "grad_norm": 2.5272786617279053,
+      "learning_rate": 9.558718861209966e-05,
+      "loss": 2.0246,
+      "step": 5878
+    },
+    {
+      "epoch": 2.612888888888889,
+      "grad_norm": 2.6360082626342773,
+      "learning_rate": 9.55693950177936e-05,
+      "loss": 1.6889,
+      "step": 5879
+    },
+    {
+      "epoch": 2.6133333333333333,
+      "grad_norm": 2.879133462905884,
+      "learning_rate": 9.555160142348754e-05,
+      "loss": 1.7153,
+      "step": 5880
+    },
+    {
+      "epoch": 2.613777777777778,
+      "grad_norm": 1.858364224433899,
+      "learning_rate": 9.55338078291815e-05,
+      "loss": 0.8528,
+      "step": 5881
+    },
+    {
+      "epoch": 2.6142222222222222,
+      "grad_norm": 2.2637181282043457,
+      "learning_rate": 9.551601423487546e-05,
+      "loss": 1.3998,
+      "step": 5882
+    },
+    {
+      "epoch": 2.6146666666666665,
+      "grad_norm": 2.4725327491760254,
+      "learning_rate": 9.54982206405694e-05,
+      "loss": 1.8587,
+      "step": 5883
+    },
+    {
+      "epoch": 2.615111111111111,
+      "grad_norm": 2.9442496299743652,
+      "learning_rate": 9.548042704626336e-05,
+      "loss": 2.0215,
+      "step": 5884
+    },
+    {
+      "epoch": 2.6155555555555554,
+      "grad_norm": 2.771085739135742,
+      "learning_rate": 9.54626334519573e-05,
+      "loss": 1.7668,
+      "step": 5885
+    },
+    {
+      "epoch": 2.616,
+      "grad_norm": 3.2240281105041504,
+      "learning_rate": 9.544483985765124e-05,
+      "loss": 1.9279,
+      "step": 5886
+    },
+    {
+      "epoch": 2.6164444444444444,
+      "grad_norm": 2.726102590560913,
+      "learning_rate": 9.54270462633452e-05,
+      "loss": 1.5733,
+      "step": 5887
+    },
+    {
+      "epoch": 2.616888888888889,
+      "grad_norm": 2.668149709701538,
+      "learning_rate": 9.540925266903915e-05,
+      "loss": 1.562,
+      "step": 5888
+    },
+    {
+      "epoch": 2.6173333333333333,
+      "grad_norm": 2.835635185241699,
+      "learning_rate": 9.53914590747331e-05,
+      "loss": 1.7907,
+      "step": 5889
+    },
+    {
+      "epoch": 2.6177777777777775,
+      "grad_norm": 3.4420056343078613,
+      "learning_rate": 9.537366548042705e-05,
+      "loss": 1.7993,
+      "step": 5890
+    },
+    {
+      "epoch": 2.6182222222222222,
+      "grad_norm": 3.626722574234009,
+      "learning_rate": 9.535587188612101e-05,
+      "loss": 2.1257,
+      "step": 5891
+    },
+    {
+      "epoch": 2.618666666666667,
+      "grad_norm": 2.6178202629089355,
+      "learning_rate": 9.533807829181495e-05,
+      "loss": 1.3251,
+      "step": 5892
+    },
+    {
+      "epoch": 2.619111111111111,
+      "grad_norm": 2.7256970405578613,
+      "learning_rate": 9.53202846975089e-05,
+      "loss": 1.5753,
+      "step": 5893
+    },
+    {
+      "epoch": 2.6195555555555554,
+      "grad_norm": 3.222346782684326,
+      "learning_rate": 9.530249110320285e-05,
+      "loss": 1.928,
+      "step": 5894
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.2846624851226807,
+      "learning_rate": 9.528469750889681e-05,
+      "loss": 1.8586,
+      "step": 5895
+    },
+    {
+      "epoch": 2.6204444444444444,
+      "grad_norm": 3.4086952209472656,
+      "learning_rate": 9.526690391459075e-05,
+      "loss": 1.5629,
+      "step": 5896
+    },
+    {
+      "epoch": 2.620888888888889,
+      "grad_norm": 3.3256282806396484,
+      "learning_rate": 9.524911032028471e-05,
+      "loss": 1.6927,
+      "step": 5897
+    },
+    {
+      "epoch": 2.6213333333333333,
+      "grad_norm": 3.0743277072906494,
+      "learning_rate": 9.523131672597865e-05,
+      "loss": 1.4407,
+      "step": 5898
+    },
+    {
+      "epoch": 2.621777777777778,
+      "grad_norm": 5.507002353668213,
+      "learning_rate": 9.52135231316726e-05,
+      "loss": 1.2145,
+      "step": 5899
+    },
+    {
+      "epoch": 2.6222222222222222,
+      "grad_norm": 4.734825134277344,
+      "learning_rate": 9.519572953736655e-05,
+      "loss": 1.7574,
+      "step": 5900
+    },
+    {
+      "epoch": 2.6226666666666665,
+      "grad_norm": 1.4950133562088013,
+      "learning_rate": 9.517793594306051e-05,
+      "loss": 1.0372,
+      "step": 5901
+    },
+    {
+      "epoch": 2.623111111111111,
+      "grad_norm": 2.024033784866333,
+      "learning_rate": 9.516014234875445e-05,
+      "loss": 1.9618,
+      "step": 5902
+    },
+    {
+      "epoch": 2.6235555555555554,
+      "grad_norm": 2.259506940841675,
+      "learning_rate": 9.514234875444841e-05,
+      "loss": 1.7976,
+      "step": 5903
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 2.1952743530273438,
+      "learning_rate": 9.512455516014235e-05,
+      "loss": 2.5925,
+      "step": 5904
+    },
+    {
+      "epoch": 2.6244444444444444,
+      "grad_norm": 2.353161096572876,
+      "learning_rate": 9.51067615658363e-05,
+      "loss": 1.9936,
+      "step": 5905
+    },
+    {
+      "epoch": 2.624888888888889,
+      "grad_norm": 1.892688274383545,
+      "learning_rate": 9.508896797153025e-05,
+      "loss": 1.7493,
+      "step": 5906
+    },
+    {
+      "epoch": 2.6253333333333333,
+      "grad_norm": 2.5873072147369385,
+      "learning_rate": 9.50711743772242e-05,
+      "loss": 1.8485,
+      "step": 5907
+    },
+    {
+      "epoch": 2.6257777777777775,
+      "grad_norm": 1.6135746240615845,
+      "learning_rate": 9.505338078291815e-05,
+      "loss": 0.9499,
+      "step": 5908
+    },
+    {
+      "epoch": 2.6262222222222222,
+      "grad_norm": 2.211524486541748,
+      "learning_rate": 9.50355871886121e-05,
+      "loss": 1.6318,
+      "step": 5909
+    },
+    {
+      "epoch": 2.626666666666667,
+      "grad_norm": 2.18351411819458,
+      "learning_rate": 9.501779359430606e-05,
+      "loss": 1.6961,
+      "step": 5910
+    },
+    {
+      "epoch": 2.627111111111111,
+      "grad_norm": 2.3403427600860596,
+      "learning_rate": 9.5e-05,
+      "loss": 1.2993,
+      "step": 5911
+    },
+    {
+      "epoch": 2.6275555555555554,
+      "grad_norm": 2.16619610786438,
+      "learning_rate": 9.498220640569395e-05,
+      "loss": 1.5,
+      "step": 5912
+    },
+    {
+      "epoch": 2.628,
+      "grad_norm": 2.4509599208831787,
+      "learning_rate": 9.49644128113879e-05,
+      "loss": 2.1853,
+      "step": 5913
+    },
+    {
+      "epoch": 2.6284444444444444,
+      "grad_norm": 2.2944180965423584,
+      "learning_rate": 9.494661921708186e-05,
+      "loss": 1.4259,
+      "step": 5914
+    },
+    {
+      "epoch": 2.628888888888889,
+      "grad_norm": 2.5902411937713623,
+      "learning_rate": 9.49288256227758e-05,
+      "loss": 1.5032,
+      "step": 5915
+    },
+    {
+      "epoch": 2.6293333333333333,
+      "grad_norm": 2.6031863689422607,
+      "learning_rate": 9.491103202846976e-05,
+      "loss": 1.9135,
+      "step": 5916
+    },
+    {
+      "epoch": 2.629777777777778,
+      "grad_norm": 2.3265883922576904,
+      "learning_rate": 9.48932384341637e-05,
+      "loss": 1.2576,
+      "step": 5917
+    },
+    {
+      "epoch": 2.6302222222222222,
+      "grad_norm": 2.358459949493408,
+      "learning_rate": 9.487544483985765e-05,
+      "loss": 1.8601,
+      "step": 5918
+    },
+    {
+      "epoch": 2.6306666666666665,
+      "grad_norm": 2.6963040828704834,
+      "learning_rate": 9.48576512455516e-05,
+      "loss": 2.0838,
+      "step": 5919
+    },
+    {
+      "epoch": 2.631111111111111,
+      "grad_norm": 2.430453300476074,
+      "learning_rate": 9.483985765124556e-05,
+      "loss": 1.6023,
+      "step": 5920
+    },
+    {
+      "epoch": 2.6315555555555554,
+      "grad_norm": 2.707322120666504,
+      "learning_rate": 9.48220640569395e-05,
+      "loss": 1.5959,
+      "step": 5921
+    },
+    {
+      "epoch": 2.632,
+      "grad_norm": 2.6505823135375977,
+      "learning_rate": 9.480427046263346e-05,
+      "loss": 1.9604,
+      "step": 5922
+    },
+    {
+      "epoch": 2.6324444444444444,
+      "grad_norm": 2.96512770652771,
+      "learning_rate": 9.478647686832742e-05,
+      "loss": 1.7178,
+      "step": 5923
+    },
+    {
+      "epoch": 2.632888888888889,
+      "grad_norm": 2.6181640625,
+      "learning_rate": 9.476868327402136e-05,
+      "loss": 2.2189,
+      "step": 5924
+    },
+    {
+      "epoch": 2.6333333333333333,
+      "grad_norm": 2.878074884414673,
+      "learning_rate": 9.47508896797153e-05,
+      "loss": 1.4863,
+      "step": 5925
+    },
+    {
+      "epoch": 2.6337777777777776,
+      "grad_norm": 2.7391178607940674,
+      "learning_rate": 9.473309608540926e-05,
+      "loss": 1.5675,
+      "step": 5926
+    },
+    {
+      "epoch": 2.6342222222222222,
+      "grad_norm": 2.5607688426971436,
+      "learning_rate": 9.471530249110321e-05,
+      "loss": 1.6423,
+      "step": 5927
+    },
+    {
+      "epoch": 2.634666666666667,
+      "grad_norm": 2.78110408782959,
+      "learning_rate": 9.469750889679716e-05,
+      "loss": 1.7438,
+      "step": 5928
+    },
+    {
+      "epoch": 2.635111111111111,
+      "grad_norm": 2.7865052223205566,
+      "learning_rate": 9.467971530249111e-05,
+      "loss": 1.8183,
+      "step": 5929
+    },
+    {
+      "epoch": 2.6355555555555554,
+      "grad_norm": 2.619302749633789,
+      "learning_rate": 9.466192170818506e-05,
+      "loss": 1.4989,
+      "step": 5930
+    },
+    {
+      "epoch": 2.636,
+      "grad_norm": 2.7835965156555176,
+      "learning_rate": 9.4644128113879e-05,
+      "loss": 1.3163,
+      "step": 5931
+    },
+    {
+      "epoch": 2.6364444444444444,
+      "grad_norm": 2.692561149597168,
+      "learning_rate": 9.462633451957296e-05,
+      "loss": 1.4886,
+      "step": 5932
+    },
+    {
+      "epoch": 2.6368888888888886,
+      "grad_norm": 2.836426019668579,
+      "learning_rate": 9.460854092526691e-05,
+      "loss": 1.9111,
+      "step": 5933
+    },
+    {
+      "epoch": 2.6373333333333333,
+      "grad_norm": 2.8386855125427246,
+      "learning_rate": 9.459074733096086e-05,
+      "loss": 1.8409,
+      "step": 5934
+    },
+    {
+      "epoch": 2.637777777777778,
+      "grad_norm": 2.6482274532318115,
+      "learning_rate": 9.457295373665481e-05,
+      "loss": 1.6835,
+      "step": 5935
+    },
+    {
+      "epoch": 2.6382222222222222,
+      "grad_norm": 2.3308165073394775,
+      "learning_rate": 9.455516014234877e-05,
+      "loss": 1.2158,
+      "step": 5936
+    },
+    {
+      "epoch": 2.6386666666666665,
+      "grad_norm": 2.602177619934082,
+      "learning_rate": 9.453736654804271e-05,
+      "loss": 1.8307,
+      "step": 5937
+    },
+    {
+      "epoch": 2.639111111111111,
+      "grad_norm": 3.2326931953430176,
+      "learning_rate": 9.451957295373665e-05,
+      "loss": 2.0871,
+      "step": 5938
+    },
+    {
+      "epoch": 2.6395555555555554,
+      "grad_norm": 2.9063456058502197,
+      "learning_rate": 9.450177935943061e-05,
+      "loss": 1.7372,
+      "step": 5939
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.6239306926727295,
+      "learning_rate": 9.448398576512457e-05,
+      "loss": 1.5927,
+      "step": 5940
+    },
+    {
+      "epoch": 2.6404444444444444,
+      "grad_norm": 3.01645827293396,
+      "learning_rate": 9.446619217081851e-05,
+      "loss": 1.5774,
+      "step": 5941
+    },
+    {
+      "epoch": 2.640888888888889,
+      "grad_norm": 3.1588776111602783,
+      "learning_rate": 9.444839857651247e-05,
+      "loss": 1.6249,
+      "step": 5942
+    },
+    {
+      "epoch": 2.6413333333333333,
+      "grad_norm": 3.4573142528533936,
+      "learning_rate": 9.443060498220641e-05,
+      "loss": 1.8496,
+      "step": 5943
+    },
+    {
+      "epoch": 2.6417777777777776,
+      "grad_norm": 2.9585986137390137,
+      "learning_rate": 9.441281138790035e-05,
+      "loss": 2.2489,
+      "step": 5944
+    },
+    {
+      "epoch": 2.6422222222222222,
+      "grad_norm": 3.1667559146881104,
+      "learning_rate": 9.439501779359431e-05,
+      "loss": 1.7358,
+      "step": 5945
+    },
+    {
+      "epoch": 2.642666666666667,
+      "grad_norm": 3.757124900817871,
+      "learning_rate": 9.437722419928827e-05,
+      "loss": 2.1035,
+      "step": 5946
+    },
+    {
+      "epoch": 2.643111111111111,
+      "grad_norm": 3.6398861408233643,
+      "learning_rate": 9.435943060498221e-05,
+      "loss": 2.1867,
+      "step": 5947
+    },
+    {
+      "epoch": 2.6435555555555554,
+      "grad_norm": 3.7026424407958984,
+      "learning_rate": 9.434163701067617e-05,
+      "loss": 1.5336,
+      "step": 5948
+    },
+    {
+      "epoch": 2.644,
+      "grad_norm": 3.198451519012451,
+      "learning_rate": 9.432384341637012e-05,
+      "loss": 0.8215,
+      "step": 5949
+    },
+    {
+      "epoch": 2.6444444444444444,
+      "grad_norm": 3.469071865081787,
+      "learning_rate": 9.430604982206405e-05,
+      "loss": 0.8383,
+      "step": 5950
+    },
+    {
+      "epoch": 2.6448888888888886,
+      "grad_norm": 1.5619441270828247,
+      "learning_rate": 9.428825622775801e-05,
+      "loss": 2.0473,
+      "step": 5951
+    },
+    {
+      "epoch": 2.6453333333333333,
+      "grad_norm": 1.3054723739624023,
+      "learning_rate": 9.427046263345196e-05,
+      "loss": 1.0846,
+      "step": 5952
+    },
+    {
+      "epoch": 2.645777777777778,
+      "grad_norm": 1.9514864683151245,
+      "learning_rate": 9.425266903914591e-05,
+      "loss": 1.7394,
+      "step": 5953
+    },
+    {
+      "epoch": 2.6462222222222223,
+      "grad_norm": 2.2087137699127197,
+      "learning_rate": 9.423487544483986e-05,
+      "loss": 1.6852,
+      "step": 5954
+    },
+    {
+      "epoch": 2.6466666666666665,
+      "grad_norm": 2.1699979305267334,
+      "learning_rate": 9.421708185053382e-05,
+      "loss": 1.6228,
+      "step": 5955
+    },
+    {
+      "epoch": 2.647111111111111,
+      "grad_norm": 2.410844564437866,
+      "learning_rate": 9.419928825622776e-05,
+      "loss": 2.018,
+      "step": 5956
+    },
+    {
+      "epoch": 2.6475555555555554,
+      "grad_norm": 2.351001024246216,
+      "learning_rate": 9.41814946619217e-05,
+      "loss": 1.5701,
+      "step": 5957
+    },
+    {
+      "epoch": 2.648,
+      "grad_norm": 2.365635633468628,
+      "learning_rate": 9.416370106761566e-05,
+      "loss": 1.8983,
+      "step": 5958
+    },
+    {
+      "epoch": 2.6484444444444444,
+      "grad_norm": 1.956677794456482,
+      "learning_rate": 9.414590747330962e-05,
+      "loss": 1.0431,
+      "step": 5959
+    },
+    {
+      "epoch": 2.648888888888889,
+      "grad_norm": 2.353614330291748,
+      "learning_rate": 9.412811387900356e-05,
+      "loss": 1.9778,
+      "step": 5960
+    },
+    {
+      "epoch": 2.6493333333333333,
+      "grad_norm": 2.23882794380188,
+      "learning_rate": 9.411032028469752e-05,
+      "loss": 1.4755,
+      "step": 5961
+    },
+    {
+      "epoch": 2.6497777777777776,
+      "grad_norm": 2.3019869327545166,
+      "learning_rate": 9.409252669039146e-05,
+      "loss": 1.1742,
+      "step": 5962
+    },
+    {
+      "epoch": 2.6502222222222223,
+      "grad_norm": 2.3776917457580566,
+      "learning_rate": 9.40747330960854e-05,
+      "loss": 1.6219,
+      "step": 5963
+    },
+    {
+      "epoch": 2.6506666666666665,
+      "grad_norm": 2.3942043781280518,
+      "learning_rate": 9.405693950177936e-05,
+      "loss": 1.7453,
+      "step": 5964
+    },
+    {
+      "epoch": 2.651111111111111,
+      "grad_norm": 2.4723472595214844,
+      "learning_rate": 9.403914590747332e-05,
+      "loss": 2.0211,
+      "step": 5965
+    },
+    {
+      "epoch": 2.6515555555555554,
+      "grad_norm": 2.4571118354797363,
+      "learning_rate": 9.402135231316726e-05,
+      "loss": 1.6631,
+      "step": 5966
+    },
+    {
+      "epoch": 2.652,
+      "grad_norm": 2.0202927589416504,
+      "learning_rate": 9.400355871886122e-05,
+      "loss": 1.291,
+      "step": 5967
+    },
+    {
+      "epoch": 2.6524444444444444,
+      "grad_norm": 2.624063014984131,
+      "learning_rate": 9.398576512455517e-05,
+      "loss": 1.9158,
+      "step": 5968
+    },
+    {
+      "epoch": 2.6528888888888886,
+      "grad_norm": 2.2154154777526855,
+      "learning_rate": 9.396797153024912e-05,
+      "loss": 1.46,
+      "step": 5969
+    },
+    {
+      "epoch": 2.6533333333333333,
+      "grad_norm": 2.480447769165039,
+      "learning_rate": 9.395017793594306e-05,
+      "loss": 2.152,
+      "step": 5970
+    },
+    {
+      "epoch": 2.653777777777778,
+      "grad_norm": 2.2933542728424072,
+      "learning_rate": 9.393238434163702e-05,
+      "loss": 1.5246,
+      "step": 5971
+    },
+    {
+      "epoch": 2.6542222222222223,
+      "grad_norm": 2.259690523147583,
+      "learning_rate": 9.391459074733097e-05,
+      "loss": 1.8234,
+      "step": 5972
+    },
+    {
+      "epoch": 2.6546666666666665,
+      "grad_norm": 2.219409942626953,
+      "learning_rate": 9.389679715302492e-05,
+      "loss": 1.8107,
+      "step": 5973
+    },
+    {
+      "epoch": 2.655111111111111,
+      "grad_norm": 2.7897937297821045,
+      "learning_rate": 9.387900355871887e-05,
+      "loss": 1.7732,
+      "step": 5974
+    },
+    {
+      "epoch": 2.6555555555555554,
+      "grad_norm": 2.3130133152008057,
+      "learning_rate": 9.386120996441281e-05,
+      "loss": 1.5716,
+      "step": 5975
+    },
+    {
+      "epoch": 2.656,
+      "grad_norm": 2.908433437347412,
+      "learning_rate": 9.384341637010676e-05,
+      "loss": 1.9881,
+      "step": 5976
+    },
+    {
+      "epoch": 2.6564444444444444,
+      "grad_norm": 2.9439632892608643,
+      "learning_rate": 9.382562277580071e-05,
+      "loss": 1.7524,
+      "step": 5977
+    },
+    {
+      "epoch": 2.656888888888889,
+      "grad_norm": 2.63321852684021,
+      "learning_rate": 9.380782918149467e-05,
+      "loss": 1.9221,
+      "step": 5978
+    },
+    {
+      "epoch": 2.6573333333333333,
+      "grad_norm": 2.665898323059082,
+      "learning_rate": 9.379003558718861e-05,
+      "loss": 1.9193,
+      "step": 5979
+    },
+    {
+      "epoch": 2.6577777777777776,
+      "grad_norm": 2.8060519695281982,
+      "learning_rate": 9.377224199288257e-05,
+      "loss": 1.7567,
+      "step": 5980
+    },
+    {
+      "epoch": 2.6582222222222223,
+      "grad_norm": 2.898313045501709,
+      "learning_rate": 9.375444839857653e-05,
+      "loss": 1.9132,
+      "step": 5981
+    },
+    {
+      "epoch": 2.6586666666666665,
+      "grad_norm": 3.1064798831939697,
+      "learning_rate": 9.373665480427047e-05,
+      "loss": 1.5257,
+      "step": 5982
+    },
+    {
+      "epoch": 2.659111111111111,
+      "grad_norm": 3.1904420852661133,
+      "learning_rate": 9.371886120996441e-05,
+      "loss": 2.2091,
+      "step": 5983
+    },
+    {
+      "epoch": 2.6595555555555555,
+      "grad_norm": 2.535499095916748,
+      "learning_rate": 9.370106761565837e-05,
+      "loss": 1.5115,
+      "step": 5984
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.571568727493286,
+      "learning_rate": 9.368327402135233e-05,
+      "loss": 1.7336,
+      "step": 5985
+    },
+    {
+      "epoch": 2.6604444444444444,
+      "grad_norm": 2.7381021976470947,
+      "learning_rate": 9.366548042704627e-05,
+      "loss": 1.7122,
+      "step": 5986
+    },
+    {
+      "epoch": 2.6608888888888886,
+      "grad_norm": 2.603264331817627,
+      "learning_rate": 9.364768683274022e-05,
+      "loss": 1.6372,
+      "step": 5987
+    },
+    {
+      "epoch": 2.6613333333333333,
+      "grad_norm": 2.825221300125122,
+      "learning_rate": 9.362989323843417e-05,
+      "loss": 1.8266,
+      "step": 5988
+    },
+    {
+      "epoch": 2.661777777777778,
+      "grad_norm": 3.076425313949585,
+      "learning_rate": 9.361209964412811e-05,
+      "loss": 1.8103,
+      "step": 5989
+    },
+    {
+      "epoch": 2.6622222222222223,
+      "grad_norm": 3.15580677986145,
+      "learning_rate": 9.359430604982207e-05,
+      "loss": 1.6719,
+      "step": 5990
+    },
+    {
+      "epoch": 2.6626666666666665,
+      "grad_norm": 3.181922674179077,
+      "learning_rate": 9.357651245551602e-05,
+      "loss": 1.8815,
+      "step": 5991
+    },
+    {
+      "epoch": 2.663111111111111,
+      "grad_norm": 2.974426746368408,
+      "learning_rate": 9.355871886120997e-05,
+      "loss": 1.6728,
+      "step": 5992
+    },
+    {
+      "epoch": 2.6635555555555555,
+      "grad_norm": 2.7631449699401855,
+      "learning_rate": 9.354092526690392e-05,
+      "loss": 1.5703,
+      "step": 5993
+    },
+    {
+      "epoch": 2.664,
+      "grad_norm": 3.3828341960906982,
+      "learning_rate": 9.352313167259788e-05,
+      "loss": 1.6526,
+      "step": 5994
+    },
+    {
+      "epoch": 2.6644444444444444,
+      "grad_norm": 3.690214157104492,
+      "learning_rate": 9.350533807829181e-05,
+      "loss": 2.2105,
+      "step": 5995
+    },
+    {
+      "epoch": 2.664888888888889,
+      "grad_norm": 4.263453960418701,
+      "learning_rate": 9.348754448398577e-05,
+      "loss": 1.8527,
+      "step": 5996
+    },
+    {
+      "epoch": 2.6653333333333333,
+      "grad_norm": 2.7068159580230713,
+      "learning_rate": 9.346975088967972e-05,
+      "loss": 0.7449,
+      "step": 5997
+    },
+    {
+      "epoch": 2.6657777777777776,
+      "grad_norm": 2.991706609725952,
+      "learning_rate": 9.345195729537366e-05,
+      "loss": 1.0355,
+      "step": 5998
+    },
+    {
+      "epoch": 2.6662222222222223,
+      "grad_norm": 4.530425071716309,
+      "learning_rate": 9.343416370106762e-05,
+      "loss": 1.912,
+      "step": 5999
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 4.280130386352539,
+      "learning_rate": 9.341637010676158e-05,
+      "loss": 1.333,
+      "step": 6000
+    },
+    {
+      "epoch": 2.667111111111111,
+      "grad_norm": 1.9667905569076538,
+      "learning_rate": 9.339857651245552e-05,
+      "loss": 2.5096,
+      "step": 6001
+    },
+    {
+      "epoch": 2.6675555555555555,
+      "grad_norm": 1.9078987836837769,
+      "learning_rate": 9.338078291814946e-05,
+      "loss": 2.1031,
+      "step": 6002
+    },
+    {
+      "epoch": 2.668,
+      "grad_norm": 2.1769826412200928,
+      "learning_rate": 9.336298932384342e-05,
+      "loss": 2.0237,
+      "step": 6003
+    },
+    {
+      "epoch": 2.6684444444444444,
+      "grad_norm": 0.29694122076034546,
+      "learning_rate": 9.334519572953738e-05,
+      "loss": 0.0224,
+      "step": 6004
+    },
+    {
+      "epoch": 2.6688888888888886,
+      "grad_norm": 1.7639079093933105,
+      "learning_rate": 9.332740213523132e-05,
+      "loss": 0.9714,
+      "step": 6005
+    },
+    {
+      "epoch": 2.6693333333333333,
+      "grad_norm": 2.3218326568603516,
+      "learning_rate": 9.330960854092528e-05,
+      "loss": 2.0642,
+      "step": 6006
+    },
+    {
+      "epoch": 2.669777777777778,
+      "grad_norm": 2.195533514022827,
+      "learning_rate": 9.329181494661923e-05,
+      "loss": 1.1775,
+      "step": 6007
+    },
+    {
+      "epoch": 2.6702222222222223,
+      "grad_norm": 1.75465989112854,
+      "learning_rate": 9.327402135231316e-05,
+      "loss": 0.9689,
+      "step": 6008
+    },
+    {
+      "epoch": 2.6706666666666665,
+      "grad_norm": 2.281090259552002,
+      "learning_rate": 9.325622775800712e-05,
+      "loss": 1.977,
+      "step": 6009
+    },
+    {
+      "epoch": 2.671111111111111,
+      "grad_norm": 1.779270052909851,
+      "learning_rate": 9.323843416370108e-05,
+      "loss": 0.882,
+      "step": 6010
+    },
+    {
+      "epoch": 2.6715555555555555,
+      "grad_norm": 2.8689680099487305,
+      "learning_rate": 9.322064056939502e-05,
+      "loss": 2.0544,
+      "step": 6011
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 2.2885069847106934,
+      "learning_rate": 9.320284697508897e-05,
+      "loss": 1.6721,
+      "step": 6012
+    },
+    {
+      "epoch": 2.6724444444444444,
+      "grad_norm": 2.3824915885925293,
+      "learning_rate": 9.318505338078293e-05,
+      "loss": 1.7404,
+      "step": 6013
+    },
+    {
+      "epoch": 2.672888888888889,
+      "grad_norm": 2.4072515964508057,
+      "learning_rate": 9.316725978647687e-05,
+      "loss": 1.7259,
+      "step": 6014
+    },
+    {
+      "epoch": 2.6733333333333333,
+      "grad_norm": 2.6987950801849365,
+      "learning_rate": 9.314946619217082e-05,
+      "loss": 1.9788,
+      "step": 6015
+    },
+    {
+      "epoch": 2.6737777777777776,
+      "grad_norm": 2.521949052810669,
+      "learning_rate": 9.313167259786477e-05,
+      "loss": 1.4704,
+      "step": 6016
+    },
+    {
+      "epoch": 2.6742222222222223,
+      "grad_norm": 3.2929980754852295,
+      "learning_rate": 9.311387900355873e-05,
+      "loss": 2.1755,
+      "step": 6017
+    },
+    {
+      "epoch": 2.6746666666666665,
+      "grad_norm": 1.4418671131134033,
+      "learning_rate": 9.309608540925267e-05,
+      "loss": 0.817,
+      "step": 6018
+    },
+    {
+      "epoch": 2.675111111111111,
+      "grad_norm": 1.5588750839233398,
+      "learning_rate": 9.307829181494663e-05,
+      "loss": 1.0768,
+      "step": 6019
+    },
+    {
+      "epoch": 2.6755555555555555,
+      "grad_norm": 2.279388189315796,
+      "learning_rate": 9.306049822064057e-05,
+      "loss": 1.2211,
+      "step": 6020
+    },
+    {
+      "epoch": 2.676,
+      "grad_norm": 2.7521657943725586,
+      "learning_rate": 9.304270462633452e-05,
+      "loss": 1.7047,
+      "step": 6021
+    },
+    {
+      "epoch": 2.6764444444444444,
+      "grad_norm": 2.4141414165496826,
+      "learning_rate": 9.302491103202847e-05,
+      "loss": 1.6488,
+      "step": 6022
+    },
+    {
+      "epoch": 2.6768888888888887,
+      "grad_norm": 2.966360330581665,
+      "learning_rate": 9.300711743772243e-05,
+      "loss": 1.8208,
+      "step": 6023
+    },
+    {
+      "epoch": 2.6773333333333333,
+      "grad_norm": 2.5755410194396973,
+      "learning_rate": 9.298932384341637e-05,
+      "loss": 2.2004,
+      "step": 6024
+    },
+    {
+      "epoch": 2.677777777777778,
+      "grad_norm": 2.3477094173431396,
+      "learning_rate": 9.297153024911033e-05,
+      "loss": 1.9106,
+      "step": 6025
+    },
+    {
+      "epoch": 2.6782222222222223,
+      "grad_norm": 3.050549030303955,
+      "learning_rate": 9.295373665480428e-05,
+      "loss": 2.0783,
+      "step": 6026
+    },
+    {
+      "epoch": 2.6786666666666665,
+      "grad_norm": 2.4758310317993164,
+      "learning_rate": 9.293594306049823e-05,
+      "loss": 1.7121,
+      "step": 6027
+    },
+    {
+      "epoch": 2.679111111111111,
+      "grad_norm": 2.427130699157715,
+      "learning_rate": 9.291814946619217e-05,
+      "loss": 1.9215,
+      "step": 6028
+    },
+    {
+      "epoch": 2.6795555555555555,
+      "grad_norm": 2.3932747840881348,
+      "learning_rate": 9.290035587188613e-05,
+      "loss": 1.7158,
+      "step": 6029
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.1773719787597656,
+      "learning_rate": 9.288256227758008e-05,
+      "loss": 1.6804,
+      "step": 6030
+    },
+    {
+      "epoch": 2.6804444444444444,
+      "grad_norm": 2.6314051151275635,
+      "learning_rate": 9.286476868327403e-05,
+      "loss": 1.5102,
+      "step": 6031
+    },
+    {
+      "epoch": 2.680888888888889,
+      "grad_norm": 2.660897970199585,
+      "learning_rate": 9.284697508896798e-05,
+      "loss": 1.9421,
+      "step": 6032
+    },
+    {
+      "epoch": 2.6813333333333333,
+      "grad_norm": 3.031437873840332,
+      "learning_rate": 9.282918149466193e-05,
+      "loss": 1.9283,
+      "step": 6033
+    },
+    {
+      "epoch": 2.6817777777777776,
+      "grad_norm": 2.821213722229004,
+      "learning_rate": 9.281138790035587e-05,
+      "loss": 1.7759,
+      "step": 6034
+    },
+    {
+      "epoch": 2.6822222222222223,
+      "grad_norm": 2.7068347930908203,
+      "learning_rate": 9.279359430604982e-05,
+      "loss": 1.7589,
+      "step": 6035
+    },
+    {
+      "epoch": 2.6826666666666665,
+      "grad_norm": 2.7685983180999756,
+      "learning_rate": 9.277580071174378e-05,
+      "loss": 2.0632,
+      "step": 6036
+    },
+    {
+      "epoch": 2.6831111111111112,
+      "grad_norm": 2.5510952472686768,
+      "learning_rate": 9.275800711743772e-05,
+      "loss": 1.6365,
+      "step": 6037
+    },
+    {
+      "epoch": 2.6835555555555555,
+      "grad_norm": 2.7311320304870605,
+      "learning_rate": 9.274021352313168e-05,
+      "loss": 1.5591,
+      "step": 6038
+    },
+    {
+      "epoch": 2.684,
+      "grad_norm": 3.251199722290039,
+      "learning_rate": 9.272241992882564e-05,
+      "loss": 2.0724,
+      "step": 6039
+    },
+    {
+      "epoch": 2.6844444444444444,
+      "grad_norm": 2.5799219608306885,
+      "learning_rate": 9.270462633451957e-05,
+      "loss": 1.643,
+      "step": 6040
+    },
+    {
+      "epoch": 2.6848888888888887,
+      "grad_norm": 3.4391684532165527,
+      "learning_rate": 9.268683274021352e-05,
+      "loss": 1.8439,
+      "step": 6041
+    },
+    {
+      "epoch": 2.6853333333333333,
+      "grad_norm": 3.380760908126831,
+      "learning_rate": 9.266903914590748e-05,
+      "loss": 2.3897,
+      "step": 6042
+    },
+    {
+      "epoch": 2.685777777777778,
+      "grad_norm": 2.8219289779663086,
+      "learning_rate": 9.265124555160142e-05,
+      "loss": 1.5193,
+      "step": 6043
+    },
+    {
+      "epoch": 2.6862222222222223,
+      "grad_norm": 2.5288350582122803,
+      "learning_rate": 9.263345195729538e-05,
+      "loss": 1.3944,
+      "step": 6044
+    },
+    {
+      "epoch": 2.6866666666666665,
+      "grad_norm": 3.494917154312134,
+      "learning_rate": 9.261565836298934e-05,
+      "loss": 1.5498,
+      "step": 6045
+    },
+    {
+      "epoch": 2.6871111111111112,
+      "grad_norm": 3.129040241241455,
+      "learning_rate": 9.259786476868328e-05,
+      "loss": 1.6695,
+      "step": 6046
+    },
+    {
+      "epoch": 2.6875555555555555,
+      "grad_norm": 3.235013484954834,
+      "learning_rate": 9.258007117437722e-05,
+      "loss": 1.9297,
+      "step": 6047
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 3.288318634033203,
+      "learning_rate": 9.256227758007118e-05,
+      "loss": 1.7766,
+      "step": 6048
+    },
+    {
+      "epoch": 2.6884444444444444,
+      "grad_norm": 3.6393327713012695,
+      "learning_rate": 9.254448398576513e-05,
+      "loss": 2.0743,
+      "step": 6049
+    },
+    {
+      "epoch": 2.688888888888889,
+      "grad_norm": 3.2492709159851074,
+      "learning_rate": 9.252669039145908e-05,
+      "loss": 0.7984,
+      "step": 6050
+    },
+    {
+      "epoch": 2.6893333333333334,
+      "grad_norm": 1.417654037475586,
+      "learning_rate": 9.250889679715303e-05,
+      "loss": 1.1905,
+      "step": 6051
+    },
+    {
+      "epoch": 2.6897777777777776,
+      "grad_norm": 2.2296252250671387,
+      "learning_rate": 9.249110320284699e-05,
+      "loss": 1.9266,
+      "step": 6052
+    },
+    {
+      "epoch": 2.6902222222222223,
+      "grad_norm": 2.10345721244812,
+      "learning_rate": 9.247330960854092e-05,
+      "loss": 2.199,
+      "step": 6053
+    },
+    {
+      "epoch": 2.6906666666666665,
+      "grad_norm": 2.360987901687622,
+      "learning_rate": 9.245551601423488e-05,
+      "loss": 2.2995,
+      "step": 6054
+    },
+    {
+      "epoch": 2.6911111111111112,
+      "grad_norm": 2.832897186279297,
+      "learning_rate": 9.243772241992883e-05,
+      "loss": 1.8118,
+      "step": 6055
+    },
+    {
+      "epoch": 2.6915555555555555,
+      "grad_norm": 2.3768863677978516,
+      "learning_rate": 9.241992882562278e-05,
+      "loss": 2.1399,
+      "step": 6056
+    },
+    {
+      "epoch": 2.692,
+      "grad_norm": 2.3208935260772705,
+      "learning_rate": 9.240213523131673e-05,
+      "loss": 2.2196,
+      "step": 6057
+    },
+    {
+      "epoch": 2.6924444444444444,
+      "grad_norm": 2.1701810359954834,
+      "learning_rate": 9.238434163701069e-05,
+      "loss": 2.0644,
+      "step": 6058
+    },
+    {
+      "epoch": 2.6928888888888887,
+      "grad_norm": 2.7813336849212646,
+      "learning_rate": 9.236654804270463e-05,
+      "loss": 2.1576,
+      "step": 6059
+    },
+    {
+      "epoch": 2.6933333333333334,
+      "grad_norm": 2.3689005374908447,
+      "learning_rate": 9.234875444839857e-05,
+      "loss": 2.148,
+      "step": 6060
+    },
+    {
+      "epoch": 2.693777777777778,
+      "grad_norm": 2.3087990283966064,
+      "learning_rate": 9.233096085409253e-05,
+      "loss": 2.139,
+      "step": 6061
+    },
+    {
+      "epoch": 2.6942222222222223,
+      "grad_norm": 2.087191581726074,
+      "learning_rate": 9.231316725978649e-05,
+      "loss": 1.7637,
+      "step": 6062
+    },
+    {
+      "epoch": 2.6946666666666665,
+      "grad_norm": 2.2079708576202393,
+      "learning_rate": 9.229537366548043e-05,
+      "loss": 2.0657,
+      "step": 6063
+    },
+    {
+      "epoch": 2.6951111111111112,
+      "grad_norm": 2.13736629486084,
+      "learning_rate": 9.227758007117439e-05,
+      "loss": 2.0342,
+      "step": 6064
+    },
+    {
+      "epoch": 2.6955555555555555,
+      "grad_norm": 2.382661819458008,
+      "learning_rate": 9.225978647686834e-05,
+      "loss": 1.5989,
+      "step": 6065
+    },
+    {
+      "epoch": 2.6959999999999997,
+      "grad_norm": 2.3359639644622803,
+      "learning_rate": 9.224199288256227e-05,
+      "loss": 2.0982,
+      "step": 6066
+    },
+    {
+      "epoch": 2.6964444444444444,
+      "grad_norm": 2.5619516372680664,
+      "learning_rate": 9.222419928825623e-05,
+      "loss": 1.9643,
+      "step": 6067
+    },
+    {
+      "epoch": 2.696888888888889,
+      "grad_norm": 2.3712518215179443,
+      "learning_rate": 9.220640569395019e-05,
+      "loss": 1.9981,
+      "step": 6068
+    },
+    {
+      "epoch": 2.6973333333333334,
+      "grad_norm": 2.2327232360839844,
+      "learning_rate": 9.218861209964413e-05,
+      "loss": 1.081,
+      "step": 6069
+    },
+    {
+      "epoch": 2.6977777777777776,
+      "grad_norm": 2.4085304737091064,
+      "learning_rate": 9.217081850533809e-05,
+      "loss": 1.8645,
+      "step": 6070
+    },
+    {
+      "epoch": 2.6982222222222223,
+      "grad_norm": 2.5223472118377686,
+      "learning_rate": 9.215302491103204e-05,
+      "loss": 1.9147,
+      "step": 6071
+    },
+    {
+      "epoch": 2.6986666666666665,
+      "grad_norm": 2.6552350521087646,
+      "learning_rate": 9.213523131672598e-05,
+      "loss": 2.2025,
+      "step": 6072
+    },
+    {
+      "epoch": 2.6991111111111112,
+      "grad_norm": 2.412116289138794,
+      "learning_rate": 9.211743772241993e-05,
+      "loss": 1.7127,
+      "step": 6073
+    },
+    {
+      "epoch": 2.6995555555555555,
+      "grad_norm": 2.261444091796875,
+      "learning_rate": 9.209964412811388e-05,
+      "loss": 1.4406,
+      "step": 6074
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.534367561340332,
+      "learning_rate": 9.208185053380784e-05,
+      "loss": 1.7541,
+      "step": 6075
+    },
+    {
+      "epoch": 2.7004444444444444,
+      "grad_norm": 2.575896978378296,
+      "learning_rate": 9.206405693950178e-05,
+      "loss": 1.9618,
+      "step": 6076
+    },
+    {
+      "epoch": 2.7008888888888887,
+      "grad_norm": 2.7344939708709717,
+      "learning_rate": 9.204626334519574e-05,
+      "loss": 1.9426,
+      "step": 6077
+    },
+    {
+      "epoch": 2.7013333333333334,
+      "grad_norm": 2.7714240550994873,
+      "learning_rate": 9.202846975088968e-05,
+      "loss": 1.8616,
+      "step": 6078
+    },
+    {
+      "epoch": 2.7017777777777776,
+      "grad_norm": 2.9522223472595215,
+      "learning_rate": 9.201067615658363e-05,
+      "loss": 2.3087,
+      "step": 6079
+    },
+    {
+      "epoch": 2.7022222222222223,
+      "grad_norm": 2.702958345413208,
+      "learning_rate": 9.199288256227758e-05,
+      "loss": 2.236,
+      "step": 6080
+    },
+    {
+      "epoch": 2.7026666666666666,
+      "grad_norm": 1.8355095386505127,
+      "learning_rate": 9.197508896797154e-05,
+      "loss": 0.6777,
+      "step": 6081
+    },
+    {
+      "epoch": 2.7031111111111112,
+      "grad_norm": 2.5094728469848633,
+      "learning_rate": 9.195729537366548e-05,
+      "loss": 1.8198,
+      "step": 6082
+    },
+    {
+      "epoch": 2.7035555555555555,
+      "grad_norm": 2.6528875827789307,
+      "learning_rate": 9.193950177935944e-05,
+      "loss": 1.7715,
+      "step": 6083
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 2.9194884300231934,
+      "learning_rate": 9.19217081850534e-05,
+      "loss": 1.7147,
+      "step": 6084
+    },
+    {
+      "epoch": 2.7044444444444444,
+      "grad_norm": 2.9441230297088623,
+      "learning_rate": 9.190391459074732e-05,
+      "loss": 1.854,
+      "step": 6085
+    },
+    {
+      "epoch": 2.704888888888889,
+      "grad_norm": 2.6675310134887695,
+      "learning_rate": 9.188612099644128e-05,
+      "loss": 2.0047,
+      "step": 6086
+    },
+    {
+      "epoch": 2.7053333333333334,
+      "grad_norm": 2.501265048980713,
+      "learning_rate": 9.186832740213524e-05,
+      "loss": 1.5176,
+      "step": 6087
+    },
+    {
+      "epoch": 2.7057777777777776,
+      "grad_norm": 2.112760066986084,
+      "learning_rate": 9.185053380782918e-05,
+      "loss": 1.0436,
+      "step": 6088
+    },
+    {
+      "epoch": 2.7062222222222223,
+      "grad_norm": 2.7593295574188232,
+      "learning_rate": 9.183274021352314e-05,
+      "loss": 1.6187,
+      "step": 6089
+    },
+    {
+      "epoch": 2.7066666666666666,
+      "grad_norm": 2.7610411643981934,
+      "learning_rate": 9.18149466192171e-05,
+      "loss": 1.7194,
+      "step": 6090
+    },
+    {
+      "epoch": 2.7071111111111112,
+      "grad_norm": 2.309563398361206,
+      "learning_rate": 9.179715302491104e-05,
+      "loss": 1.2155,
+      "step": 6091
+    },
+    {
+      "epoch": 2.7075555555555555,
+      "grad_norm": 3.307159900665283,
+      "learning_rate": 9.177935943060498e-05,
+      "loss": 1.9251,
+      "step": 6092
+    },
+    {
+      "epoch": 2.708,
+      "grad_norm": 2.4298038482666016,
+      "learning_rate": 9.176156583629894e-05,
+      "loss": 1.2146,
+      "step": 6093
+    },
+    {
+      "epoch": 2.7084444444444444,
+      "grad_norm": 2.78938627243042,
+      "learning_rate": 9.174377224199289e-05,
+      "loss": 1.7818,
+      "step": 6094
+    },
+    {
+      "epoch": 2.7088888888888887,
+      "grad_norm": 2.8825080394744873,
+      "learning_rate": 9.172597864768684e-05,
+      "loss": 1.6426,
+      "step": 6095
+    },
+    {
+      "epoch": 2.7093333333333334,
+      "grad_norm": 3.2043745517730713,
+      "learning_rate": 9.170818505338079e-05,
+      "loss": 1.5032,
+      "step": 6096
+    },
+    {
+      "epoch": 2.7097777777777776,
+      "grad_norm": 3.1680638790130615,
+      "learning_rate": 9.169039145907475e-05,
+      "loss": 1.3996,
+      "step": 6097
+    },
+    {
+      "epoch": 2.7102222222222223,
+      "grad_norm": 3.6439359188079834,
+      "learning_rate": 9.167259786476868e-05,
+      "loss": 2.1672,
+      "step": 6098
+    },
+    {
+      "epoch": 2.7106666666666666,
+      "grad_norm": 4.421741485595703,
+      "learning_rate": 9.165480427046263e-05,
+      "loss": 2.4516,
+      "step": 6099
+    },
+    {
+      "epoch": 2.7111111111111112,
+      "grad_norm": 3.2291154861450195,
+      "learning_rate": 9.163701067615659e-05,
+      "loss": 1.6698,
+      "step": 6100
+    },
+    {
+      "epoch": 2.7115555555555555,
+      "grad_norm": 1.8090177774429321,
+      "learning_rate": 9.161921708185053e-05,
+      "loss": 2.1445,
+      "step": 6101
+    },
+    {
+      "epoch": 2.7119999999999997,
+      "grad_norm": 1.2429306507110596,
+      "learning_rate": 9.160142348754449e-05,
+      "loss": 1.2912,
+      "step": 6102
+    },
+    {
+      "epoch": 2.7124444444444444,
+      "grad_norm": 2.093599319458008,
+      "learning_rate": 9.158362989323845e-05,
+      "loss": 2.4831,
+      "step": 6103
+    },
+    {
+      "epoch": 2.712888888888889,
+      "grad_norm": 1.8647911548614502,
+      "learning_rate": 9.156583629893239e-05,
+      "loss": 0.9684,
+      "step": 6104
+    },
+    {
+      "epoch": 2.7133333333333334,
+      "grad_norm": 2.1396780014038086,
+      "learning_rate": 9.154804270462633e-05,
+      "loss": 2.1317,
+      "step": 6105
+    },
+    {
+      "epoch": 2.7137777777777776,
+      "grad_norm": 2.1516218185424805,
+      "learning_rate": 9.153024911032029e-05,
+      "loss": 1.7869,
+      "step": 6106
+    },
+    {
+      "epoch": 2.7142222222222223,
+      "grad_norm": 2.357754945755005,
+      "learning_rate": 9.151245551601425e-05,
+      "loss": 2.3442,
+      "step": 6107
+    },
+    {
+      "epoch": 2.7146666666666666,
+      "grad_norm": 2.1616320610046387,
+      "learning_rate": 9.149466192170819e-05,
+      "loss": 2.2035,
+      "step": 6108
+    },
+    {
+      "epoch": 2.7151111111111113,
+      "grad_norm": 2.2545979022979736,
+      "learning_rate": 9.147686832740214e-05,
+      "loss": 2.3456,
+      "step": 6109
+    },
+    {
+      "epoch": 2.7155555555555555,
+      "grad_norm": 1.8658015727996826,
+      "learning_rate": 9.14590747330961e-05,
+      "loss": 1.5012,
+      "step": 6110
+    },
+    {
+      "epoch": 2.716,
+      "grad_norm": 2.5703420639038086,
+      "learning_rate": 9.144128113879003e-05,
+      "loss": 1.8163,
+      "step": 6111
+    },
+    {
+      "epoch": 2.7164444444444444,
+      "grad_norm": 2.278449296951294,
+      "learning_rate": 9.142348754448399e-05,
+      "loss": 1.7411,
+      "step": 6112
+    },
+    {
+      "epoch": 2.7168888888888887,
+      "grad_norm": 2.5653324127197266,
+      "learning_rate": 9.140569395017794e-05,
+      "loss": 1.8087,
+      "step": 6113
+    },
+    {
+      "epoch": 2.7173333333333334,
+      "grad_norm": 2.3541901111602783,
+      "learning_rate": 9.138790035587189e-05,
+      "loss": 2.0437,
+      "step": 6114
+    },
+    {
+      "epoch": 2.7177777777777776,
+      "grad_norm": 2.4183175563812256,
+      "learning_rate": 9.137010676156584e-05,
+      "loss": 1.6396,
+      "step": 6115
+    },
+    {
+      "epoch": 2.7182222222222223,
+      "grad_norm": 2.6298437118530273,
+      "learning_rate": 9.13523131672598e-05,
+      "loss": 2.3171,
+      "step": 6116
+    },
+    {
+      "epoch": 2.7186666666666666,
+      "grad_norm": 2.7562639713287354,
+      "learning_rate": 9.133451957295374e-05,
+      "loss": 2.0566,
+      "step": 6117
+    },
+    {
+      "epoch": 2.7191111111111113,
+      "grad_norm": 2.4366018772125244,
+      "learning_rate": 9.131672597864769e-05,
+      "loss": 1.9631,
+      "step": 6118
+    },
+    {
+      "epoch": 2.7195555555555555,
+      "grad_norm": 2.1214277744293213,
+      "learning_rate": 9.129893238434164e-05,
+      "loss": 1.3662,
+      "step": 6119
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 2.162672996520996,
+      "learning_rate": 9.12811387900356e-05,
+      "loss": 1.3789,
+      "step": 6120
+    },
+    {
+      "epoch": 2.7204444444444444,
+      "grad_norm": 2.6844053268432617,
+      "learning_rate": 9.126334519572954e-05,
+      "loss": 1.8149,
+      "step": 6121
+    },
+    {
+      "epoch": 2.720888888888889,
+      "grad_norm": 2.491919994354248,
+      "learning_rate": 9.12455516014235e-05,
+      "loss": 1.8213,
+      "step": 6122
+    },
+    {
+      "epoch": 2.7213333333333334,
+      "grad_norm": 2.4091458320617676,
+      "learning_rate": 9.122775800711744e-05,
+      "loss": 1.8815,
+      "step": 6123
+    },
+    {
+      "epoch": 2.7217777777777776,
+      "grad_norm": 2.2828023433685303,
+      "learning_rate": 9.120996441281138e-05,
+      "loss": 1.4552,
+      "step": 6124
+    },
+    {
+      "epoch": 2.7222222222222223,
+      "grad_norm": 2.2405014038085938,
+      "learning_rate": 9.119217081850534e-05,
+      "loss": 1.3707,
+      "step": 6125
+    },
+    {
+      "epoch": 2.7226666666666666,
+      "grad_norm": 2.6642508506774902,
+      "learning_rate": 9.11743772241993e-05,
+      "loss": 1.7702,
+      "step": 6126
+    },
+    {
+      "epoch": 2.7231111111111113,
+      "grad_norm": 2.5781748294830322,
+      "learning_rate": 9.115658362989324e-05,
+      "loss": 1.7319,
+      "step": 6127
+    },
+    {
+      "epoch": 2.7235555555555555,
+      "grad_norm": 2.8042945861816406,
+      "learning_rate": 9.11387900355872e-05,
+      "loss": 1.8082,
+      "step": 6128
+    },
+    {
+      "epoch": 2.724,
+      "grad_norm": 1.7119746208190918,
+      "learning_rate": 9.112099644128115e-05,
+      "loss": 0.7126,
+      "step": 6129
+    },
+    {
+      "epoch": 2.7244444444444444,
+      "grad_norm": 2.4615836143493652,
+      "learning_rate": 9.110320284697508e-05,
+      "loss": 1.6634,
+      "step": 6130
+    },
+    {
+      "epoch": 2.7248888888888887,
+      "grad_norm": 2.9589643478393555,
+      "learning_rate": 9.108540925266904e-05,
+      "loss": 1.829,
+      "step": 6131
+    },
+    {
+      "epoch": 2.7253333333333334,
+      "grad_norm": 2.783677577972412,
+      "learning_rate": 9.1067615658363e-05,
+      "loss": 1.98,
+      "step": 6132
+    },
+    {
+      "epoch": 2.7257777777777776,
+      "grad_norm": 3.0595808029174805,
+      "learning_rate": 9.104982206405694e-05,
+      "loss": 1.7831,
+      "step": 6133
+    },
+    {
+      "epoch": 2.7262222222222223,
+      "grad_norm": 2.033073902130127,
+      "learning_rate": 9.10320284697509e-05,
+      "loss": 0.923,
+      "step": 6134
+    },
+    {
+      "epoch": 2.7266666666666666,
+      "grad_norm": 3.1585965156555176,
+      "learning_rate": 9.101423487544485e-05,
+      "loss": 1.735,
+      "step": 6135
+    },
+    {
+      "epoch": 2.7271111111111113,
+      "grad_norm": 2.5209286212921143,
+      "learning_rate": 9.09964412811388e-05,
+      "loss": 1.5229,
+      "step": 6136
+    },
+    {
+      "epoch": 2.7275555555555555,
+      "grad_norm": 2.977914810180664,
+      "learning_rate": 9.097864768683274e-05,
+      "loss": 2.0039,
+      "step": 6137
+    },
+    {
+      "epoch": 2.7279999999999998,
+      "grad_norm": 2.497366189956665,
+      "learning_rate": 9.09608540925267e-05,
+      "loss": 1.4511,
+      "step": 6138
+    },
+    {
+      "epoch": 2.7284444444444444,
+      "grad_norm": 2.936041831970215,
+      "learning_rate": 9.094306049822065e-05,
+      "loss": 1.9052,
+      "step": 6139
+    },
+    {
+      "epoch": 2.728888888888889,
+      "grad_norm": 3.006460428237915,
+      "learning_rate": 9.092526690391459e-05,
+      "loss": 1.758,
+      "step": 6140
+    },
+    {
+      "epoch": 2.7293333333333334,
+      "grad_norm": 3.10183048248291,
+      "learning_rate": 9.090747330960855e-05,
+      "loss": 1.6804,
+      "step": 6141
+    },
+    {
+      "epoch": 2.7297777777777776,
+      "grad_norm": 3.5119240283966064,
+      "learning_rate": 9.08896797153025e-05,
+      "loss": 1.763,
+      "step": 6142
+    },
+    {
+      "epoch": 2.7302222222222223,
+      "grad_norm": 3.4593982696533203,
+      "learning_rate": 9.087188612099644e-05,
+      "loss": 2.3246,
+      "step": 6143
+    },
+    {
+      "epoch": 2.7306666666666666,
+      "grad_norm": 2.949693202972412,
+      "learning_rate": 9.085409252669039e-05,
+      "loss": 1.4612,
+      "step": 6144
+    },
+    {
+      "epoch": 2.7311111111111113,
+      "grad_norm": 3.316683769226074,
+      "learning_rate": 9.083629893238435e-05,
+      "loss": 2.0814,
+      "step": 6145
+    },
+    {
+      "epoch": 2.7315555555555555,
+      "grad_norm": 2.9927797317504883,
+      "learning_rate": 9.081850533807829e-05,
+      "loss": 1.5698,
+      "step": 6146
+    },
+    {
+      "epoch": 2.732,
+      "grad_norm": 2.82143497467041,
+      "learning_rate": 9.080071174377225e-05,
+      "loss": 1.6885,
+      "step": 6147
+    },
+    {
+      "epoch": 2.7324444444444445,
+      "grad_norm": 3.127837657928467,
+      "learning_rate": 9.07829181494662e-05,
+      "loss": 1.7723,
+      "step": 6148
+    },
+    {
+      "epoch": 2.7328888888888887,
+      "grad_norm": 3.9703283309936523,
+      "learning_rate": 9.076512455516015e-05,
+      "loss": 2.0333,
+      "step": 6149
+    },
+    {
+      "epoch": 2.7333333333333334,
+      "grad_norm": 2.2306530475616455,
+      "learning_rate": 9.074733096085409e-05,
+      "loss": 0.5828,
+      "step": 6150
+    },
+    {
+      "epoch": 2.7337777777777776,
+      "grad_norm": 2.2018139362335205,
+      "learning_rate": 9.072953736654805e-05,
+      "loss": 2.6287,
+      "step": 6151
+    },
+    {
+      "epoch": 2.7342222222222223,
+      "grad_norm": 1.7515716552734375,
+      "learning_rate": 9.0711743772242e-05,
+      "loss": 1.8294,
+      "step": 6152
+    },
+    {
+      "epoch": 2.7346666666666666,
+      "grad_norm": 1.612664818763733,
+      "learning_rate": 9.069395017793595e-05,
+      "loss": 1.3168,
+      "step": 6153
+    },
+    {
+      "epoch": 2.7351111111111113,
+      "grad_norm": 2.2013888359069824,
+      "learning_rate": 9.06761565836299e-05,
+      "loss": 2.3534,
+      "step": 6154
+    },
+    {
+      "epoch": 2.7355555555555555,
+      "grad_norm": 2.403697967529297,
+      "learning_rate": 9.065836298932386e-05,
+      "loss": 2.0122,
+      "step": 6155
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 2.392401695251465,
+      "learning_rate": 9.064056939501779e-05,
+      "loss": 2.1935,
+      "step": 6156
+    },
+    {
+      "epoch": 2.7364444444444445,
+      "grad_norm": 2.547696828842163,
+      "learning_rate": 9.062277580071175e-05,
+      "loss": 2.1032,
+      "step": 6157
+    },
+    {
+      "epoch": 2.736888888888889,
+      "grad_norm": 2.0454320907592773,
+      "learning_rate": 9.06049822064057e-05,
+      "loss": 1.6723,
+      "step": 6158
+    },
+    {
+      "epoch": 2.7373333333333334,
+      "grad_norm": 2.0025997161865234,
+      "learning_rate": 9.058718861209964e-05,
+      "loss": 1.5211,
+      "step": 6159
+    },
+    {
+      "epoch": 2.7377777777777776,
+      "grad_norm": 2.7177071571350098,
+      "learning_rate": 9.05693950177936e-05,
+      "loss": 2.2681,
+      "step": 6160
+    },
+    {
+      "epoch": 2.7382222222222223,
+      "grad_norm": 2.377166986465454,
+      "learning_rate": 9.055160142348756e-05,
+      "loss": 1.6563,
+      "step": 6161
+    },
+    {
+      "epoch": 2.7386666666666666,
+      "grad_norm": 2.444869041442871,
+      "learning_rate": 9.05338078291815e-05,
+      "loss": 1.7362,
+      "step": 6162
+    },
+    {
+      "epoch": 2.7391111111111113,
+      "grad_norm": 2.2301559448242188,
+      "learning_rate": 9.051601423487544e-05,
+      "loss": 2.0245,
+      "step": 6163
+    },
+    {
+      "epoch": 2.7395555555555555,
+      "grad_norm": 2.3340377807617188,
+      "learning_rate": 9.04982206405694e-05,
+      "loss": 1.6158,
+      "step": 6164
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.1984641551971436,
+      "learning_rate": 9.048042704626336e-05,
+      "loss": 1.6976,
+      "step": 6165
+    },
+    {
+      "epoch": 2.7404444444444445,
+      "grad_norm": 2.684303045272827,
+      "learning_rate": 9.04626334519573e-05,
+      "loss": 1.9596,
+      "step": 6166
+    },
+    {
+      "epoch": 2.7408888888888887,
+      "grad_norm": 2.037320852279663,
+      "learning_rate": 9.044483985765126e-05,
+      "loss": 1.5683,
+      "step": 6167
+    },
+    {
+      "epoch": 2.7413333333333334,
+      "grad_norm": 1.9725054502487183,
+      "learning_rate": 9.04270462633452e-05,
+      "loss": 1.5862,
+      "step": 6168
+    },
+    {
+      "epoch": 2.7417777777777776,
+      "grad_norm": 2.6273114681243896,
+      "learning_rate": 9.040925266903914e-05,
+      "loss": 2.0218,
+      "step": 6169
+    },
+    {
+      "epoch": 2.7422222222222223,
+      "grad_norm": 2.4630777835845947,
+      "learning_rate": 9.03914590747331e-05,
+      "loss": 1.7888,
+      "step": 6170
+    },
+    {
+      "epoch": 2.7426666666666666,
+      "grad_norm": 2.1881697177886963,
+      "learning_rate": 9.037366548042705e-05,
+      "loss": 1.811,
+      "step": 6171
+    },
+    {
+      "epoch": 2.7431111111111113,
+      "grad_norm": 2.804276943206787,
+      "learning_rate": 9.0355871886121e-05,
+      "loss": 2.1978,
+      "step": 6172
+    },
+    {
+      "epoch": 2.7435555555555555,
+      "grad_norm": 2.1326301097869873,
+      "learning_rate": 9.033807829181495e-05,
+      "loss": 1.7793,
+      "step": 6173
+    },
+    {
+      "epoch": 2.7439999999999998,
+      "grad_norm": 2.142714023590088,
+      "learning_rate": 9.032028469750891e-05,
+      "loss": 1.824,
+      "step": 6174
+    },
+    {
+      "epoch": 2.7444444444444445,
+      "grad_norm": 2.4953362941741943,
+      "learning_rate": 9.030249110320284e-05,
+      "loss": 1.922,
+      "step": 6175
+    },
+    {
+      "epoch": 2.744888888888889,
+      "grad_norm": 2.264841318130493,
+      "learning_rate": 9.02846975088968e-05,
+      "loss": 1.7474,
+      "step": 6176
+    },
+    {
+      "epoch": 2.7453333333333334,
+      "grad_norm": 2.464749813079834,
+      "learning_rate": 9.026690391459075e-05,
+      "loss": 1.7516,
+      "step": 6177
+    },
+    {
+      "epoch": 2.7457777777777777,
+      "grad_norm": 2.3408308029174805,
+      "learning_rate": 9.02491103202847e-05,
+      "loss": 1.6946,
+      "step": 6178
+    },
+    {
+      "epoch": 2.7462222222222223,
+      "grad_norm": 2.4677340984344482,
+      "learning_rate": 9.023131672597865e-05,
+      "loss": 1.6011,
+      "step": 6179
+    },
+    {
+      "epoch": 2.7466666666666666,
+      "grad_norm": 2.2355053424835205,
+      "learning_rate": 9.021352313167261e-05,
+      "loss": 1.5883,
+      "step": 6180
+    },
+    {
+      "epoch": 2.747111111111111,
+      "grad_norm": 0.23143476247787476,
+      "learning_rate": 9.019572953736655e-05,
+      "loss": 0.0327,
+      "step": 6181
+    },
+    {
+      "epoch": 2.7475555555555555,
+      "grad_norm": 2.9940497875213623,
+      "learning_rate": 9.01779359430605e-05,
+      "loss": 1.9261,
+      "step": 6182
+    },
+    {
+      "epoch": 2.748,
+      "grad_norm": 2.4658970832824707,
+      "learning_rate": 9.016014234875445e-05,
+      "loss": 1.8213,
+      "step": 6183
+    },
+    {
+      "epoch": 2.7484444444444445,
+      "grad_norm": 3.212437152862549,
+      "learning_rate": 9.014234875444841e-05,
+      "loss": 1.511,
+      "step": 6184
+    },
+    {
+      "epoch": 2.7488888888888887,
+      "grad_norm": 2.720136880874634,
+      "learning_rate": 9.012455516014235e-05,
+      "loss": 1.5774,
+      "step": 6185
+    },
+    {
+      "epoch": 2.7493333333333334,
+      "grad_norm": 2.8537580966949463,
+      "learning_rate": 9.010676156583631e-05,
+      "loss": 1.8448,
+      "step": 6186
+    },
+    {
+      "epoch": 2.7497777777777777,
+      "grad_norm": 2.702603578567505,
+      "learning_rate": 9.008896797153026e-05,
+      "loss": 1.3248,
+      "step": 6187
+    },
+    {
+      "epoch": 2.7502222222222223,
+      "grad_norm": 2.6414220333099365,
+      "learning_rate": 9.00711743772242e-05,
+      "loss": 1.7348,
+      "step": 6188
+    },
+    {
+      "epoch": 2.7506666666666666,
+      "grad_norm": 2.9192988872528076,
+      "learning_rate": 9.005338078291815e-05,
+      "loss": 1.4816,
+      "step": 6189
+    },
+    {
+      "epoch": 2.7511111111111113,
+      "grad_norm": 2.888636589050293,
+      "learning_rate": 9.00355871886121e-05,
+      "loss": 1.5585,
+      "step": 6190
+    },
+    {
+      "epoch": 2.7515555555555555,
+      "grad_norm": 3.021221399307251,
+      "learning_rate": 9.001779359430605e-05,
+      "loss": 1.8431,
+      "step": 6191
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 2.7811830043792725,
+      "learning_rate": 9e-05,
+      "loss": 1.7267,
+      "step": 6192
+    },
+    {
+      "epoch": 2.7524444444444445,
+      "grad_norm": 3.7655017375946045,
+      "learning_rate": 8.998220640569396e-05,
+      "loss": 1.9514,
+      "step": 6193
+    },
+    {
+      "epoch": 2.752888888888889,
+      "grad_norm": 2.969604015350342,
+      "learning_rate": 8.99644128113879e-05,
+      "loss": 1.8387,
+      "step": 6194
+    },
+    {
+      "epoch": 2.7533333333333334,
+      "grad_norm": 3.2840607166290283,
+      "learning_rate": 8.994661921708185e-05,
+      "loss": 1.4239,
+      "step": 6195
+    },
+    {
+      "epoch": 2.7537777777777777,
+      "grad_norm": 2.914557695388794,
+      "learning_rate": 8.99288256227758e-05,
+      "loss": 1.7825,
+      "step": 6196
+    },
+    {
+      "epoch": 2.7542222222222223,
+      "grad_norm": 3.4805359840393066,
+      "learning_rate": 8.991103202846976e-05,
+      "loss": 1.8883,
+      "step": 6197
+    },
+    {
+      "epoch": 2.7546666666666666,
+      "grad_norm": 3.3305397033691406,
+      "learning_rate": 8.98932384341637e-05,
+      "loss": 1.3508,
+      "step": 6198
+    },
+    {
+      "epoch": 2.755111111111111,
+      "grad_norm": 3.2841057777404785,
+      "learning_rate": 8.987544483985766e-05,
+      "loss": 1.4563,
+      "step": 6199
+    },
+    {
+      "epoch": 2.7555555555555555,
+      "grad_norm": 2.9176838397979736,
+      "learning_rate": 8.985765124555162e-05,
+      "loss": 1.0079,
+      "step": 6200
+    },
+    {
+      "epoch": 2.7560000000000002,
+      "grad_norm": 1.1757615804672241,
+      "learning_rate": 8.983985765124555e-05,
+      "loss": 1.0498,
+      "step": 6201
+    },
+    {
+      "epoch": 2.7564444444444445,
+      "grad_norm": 1.7046676874160767,
+      "learning_rate": 8.98220640569395e-05,
+      "loss": 1.9004,
+      "step": 6202
+    },
+    {
+      "epoch": 2.7568888888888887,
+      "grad_norm": 1.9083878993988037,
+      "learning_rate": 8.980427046263346e-05,
+      "loss": 2.1245,
+      "step": 6203
+    },
+    {
+      "epoch": 2.7573333333333334,
+      "grad_norm": 2.188979148864746,
+      "learning_rate": 8.97864768683274e-05,
+      "loss": 2.0422,
+      "step": 6204
+    },
+    {
+      "epoch": 2.7577777777777777,
+      "grad_norm": 2.305520534515381,
+      "learning_rate": 8.976868327402136e-05,
+      "loss": 1.8755,
+      "step": 6205
+    },
+    {
+      "epoch": 2.7582222222222224,
+      "grad_norm": 2.1911003589630127,
+      "learning_rate": 8.975088967971532e-05,
+      "loss": 1.9326,
+      "step": 6206
+    },
+    {
+      "epoch": 2.7586666666666666,
+      "grad_norm": 1.9062891006469727,
+      "learning_rate": 8.973309608540926e-05,
+      "loss": 1.7775,
+      "step": 6207
+    },
+    {
+      "epoch": 2.7591111111111113,
+      "grad_norm": 2.4695067405700684,
+      "learning_rate": 8.97153024911032e-05,
+      "loss": 2.0425,
+      "step": 6208
+    },
+    {
+      "epoch": 2.7595555555555555,
+      "grad_norm": 2.0994985103607178,
+      "learning_rate": 8.969750889679716e-05,
+      "loss": 1.6857,
+      "step": 6209
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.3211324214935303,
+      "learning_rate": 8.96797153024911e-05,
+      "loss": 2.2616,
+      "step": 6210
+    },
+    {
+      "epoch": 2.7604444444444445,
+      "grad_norm": 2.4278671741485596,
+      "learning_rate": 8.966192170818506e-05,
+      "loss": 1.8835,
+      "step": 6211
+    },
+    {
+      "epoch": 2.7608888888888887,
+      "grad_norm": 2.776087999343872,
+      "learning_rate": 8.964412811387901e-05,
+      "loss": 1.4898,
+      "step": 6212
+    },
+    {
+      "epoch": 2.7613333333333334,
+      "grad_norm": 2.4669578075408936,
+      "learning_rate": 8.962633451957296e-05,
+      "loss": 1.993,
+      "step": 6213
+    },
+    {
+      "epoch": 2.7617777777777777,
+      "grad_norm": 2.3764331340789795,
+      "learning_rate": 8.96085409252669e-05,
+      "loss": 1.6464,
+      "step": 6214
+    },
+    {
+      "epoch": 2.7622222222222224,
+      "grad_norm": 2.225770950317383,
+      "learning_rate": 8.959074733096086e-05,
+      "loss": 1.9112,
+      "step": 6215
+    },
+    {
+      "epoch": 2.7626666666666666,
+      "grad_norm": 2.581855535507202,
+      "learning_rate": 8.957295373665481e-05,
+      "loss": 1.7821,
+      "step": 6216
+    },
+    {
+      "epoch": 2.763111111111111,
+      "grad_norm": 2.1973445415496826,
+      "learning_rate": 8.955516014234876e-05,
+      "loss": 1.4587,
+      "step": 6217
+    },
+    {
+      "epoch": 2.7635555555555555,
+      "grad_norm": 2.4556639194488525,
+      "learning_rate": 8.953736654804271e-05,
+      "loss": 2.1099,
+      "step": 6218
+    },
+    {
+      "epoch": 2.7640000000000002,
+      "grad_norm": 1.6619460582733154,
+      "learning_rate": 8.951957295373667e-05,
+      "loss": 0.7801,
+      "step": 6219
+    },
+    {
+      "epoch": 2.7644444444444445,
+      "grad_norm": 2.5685040950775146,
+      "learning_rate": 8.95017793594306e-05,
+      "loss": 1.572,
+      "step": 6220
+    },
+    {
+      "epoch": 2.7648888888888887,
+      "grad_norm": 2.160173177719116,
+      "learning_rate": 8.948398576512455e-05,
+      "loss": 1.5836,
+      "step": 6221
+    },
+    {
+      "epoch": 2.7653333333333334,
+      "grad_norm": 2.594742774963379,
+      "learning_rate": 8.946619217081851e-05,
+      "loss": 1.307,
+      "step": 6222
+    },
+    {
+      "epoch": 2.7657777777777777,
+      "grad_norm": 2.604459285736084,
+      "learning_rate": 8.944839857651245e-05,
+      "loss": 1.8215,
+      "step": 6223
+    },
+    {
+      "epoch": 2.7662222222222224,
+      "grad_norm": 2.575556516647339,
+      "learning_rate": 8.943060498220641e-05,
+      "loss": 1.7588,
+      "step": 6224
+    },
+    {
+      "epoch": 2.7666666666666666,
+      "grad_norm": 2.51969051361084,
+      "learning_rate": 8.941281138790037e-05,
+      "loss": 1.8821,
+      "step": 6225
+    },
+    {
+      "epoch": 2.7671111111111113,
+      "grad_norm": 2.5474319458007812,
+      "learning_rate": 8.939501779359431e-05,
+      "loss": 2.1172,
+      "step": 6226
+    },
+    {
+      "epoch": 2.7675555555555555,
+      "grad_norm": 2.6223058700561523,
+      "learning_rate": 8.937722419928825e-05,
+      "loss": 1.7894,
+      "step": 6227
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 2.7866673469543457,
+      "learning_rate": 8.935943060498221e-05,
+      "loss": 1.7887,
+      "step": 6228
+    },
+    {
+      "epoch": 2.7684444444444445,
+      "grad_norm": 0.3962222933769226,
+      "learning_rate": 8.934163701067617e-05,
+      "loss": 0.0359,
+      "step": 6229
+    },
+    {
+      "epoch": 2.7688888888888887,
+      "grad_norm": 2.9397222995758057,
+      "learning_rate": 8.932384341637011e-05,
+      "loss": 1.8361,
+      "step": 6230
+    },
+    {
+      "epoch": 2.7693333333333334,
+      "grad_norm": 2.984018087387085,
+      "learning_rate": 8.930604982206407e-05,
+      "loss": 1.7147,
+      "step": 6231
+    },
+    {
+      "epoch": 2.7697777777777777,
+      "grad_norm": 3.337759017944336,
+      "learning_rate": 8.928825622775802e-05,
+      "loss": 2.0495,
+      "step": 6232
+    },
+    {
+      "epoch": 2.7702222222222224,
+      "grad_norm": 2.786367654800415,
+      "learning_rate": 8.927046263345195e-05,
+      "loss": 1.7017,
+      "step": 6233
+    },
+    {
+      "epoch": 2.7706666666666666,
+      "grad_norm": 3.031346321105957,
+      "learning_rate": 8.925266903914591e-05,
+      "loss": 1.8949,
+      "step": 6234
+    },
+    {
+      "epoch": 2.771111111111111,
+      "grad_norm": 2.8684723377227783,
+      "learning_rate": 8.923487544483986e-05,
+      "loss": 1.7939,
+      "step": 6235
+    },
+    {
+      "epoch": 2.7715555555555556,
+      "grad_norm": 2.4864673614501953,
+      "learning_rate": 8.921708185053381e-05,
+      "loss": 1.6988,
+      "step": 6236
+    },
+    {
+      "epoch": 2.7720000000000002,
+      "grad_norm": 2.5959644317626953,
+      "learning_rate": 8.919928825622776e-05,
+      "loss": 1.4515,
+      "step": 6237
+    },
+    {
+      "epoch": 2.7724444444444445,
+      "grad_norm": 2.3537802696228027,
+      "learning_rate": 8.918149466192172e-05,
+      "loss": 1.2161,
+      "step": 6238
+    },
+    {
+      "epoch": 2.7728888888888887,
+      "grad_norm": 2.6737425327301025,
+      "learning_rate": 8.916370106761566e-05,
+      "loss": 1.6259,
+      "step": 6239
+    },
+    {
+      "epoch": 2.7733333333333334,
+      "grad_norm": 2.907355785369873,
+      "learning_rate": 8.91459074733096e-05,
+      "loss": 1.846,
+      "step": 6240
+    },
+    {
+      "epoch": 2.7737777777777777,
+      "grad_norm": 3.1080589294433594,
+      "learning_rate": 8.912811387900356e-05,
+      "loss": 1.718,
+      "step": 6241
+    },
+    {
+      "epoch": 2.7742222222222224,
+      "grad_norm": 3.7171342372894287,
+      "learning_rate": 8.911032028469752e-05,
+      "loss": 1.7999,
+      "step": 6242
+    },
+    {
+      "epoch": 2.7746666666666666,
+      "grad_norm": 2.974966049194336,
+      "learning_rate": 8.909252669039146e-05,
+      "loss": 1.5202,
+      "step": 6243
+    },
+    {
+      "epoch": 2.7751111111111113,
+      "grad_norm": 2.403938055038452,
+      "learning_rate": 8.907473309608542e-05,
+      "loss": 1.3251,
+      "step": 6244
+    },
+    {
+      "epoch": 2.7755555555555556,
+      "grad_norm": 3.240013599395752,
+      "learning_rate": 8.905693950177937e-05,
+      "loss": 1.1542,
+      "step": 6245
+    },
+    {
+      "epoch": 2.776,
+      "grad_norm": 2.923494815826416,
+      "learning_rate": 8.90391459074733e-05,
+      "loss": 1.5198,
+      "step": 6246
+    },
+    {
+      "epoch": 2.7764444444444445,
+      "grad_norm": 3.5100812911987305,
+      "learning_rate": 8.902135231316726e-05,
+      "loss": 1.4171,
+      "step": 6247
+    },
+    {
+      "epoch": 2.7768888888888887,
+      "grad_norm": 2.885017156600952,
+      "learning_rate": 8.900355871886122e-05,
+      "loss": 1.4852,
+      "step": 6248
+    },
+    {
+      "epoch": 2.7773333333333334,
+      "grad_norm": 3.7107977867126465,
+      "learning_rate": 8.898576512455516e-05,
+      "loss": 1.3071,
+      "step": 6249
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 3.349717378616333,
+      "learning_rate": 8.896797153024912e-05,
+      "loss": 1.5356,
+      "step": 6250
+    },
+    {
+      "epoch": 2.7782222222222224,
+      "grad_norm": 2.166078805923462,
+      "learning_rate": 8.895017793594307e-05,
+      "loss": 1.9864,
+      "step": 6251
+    },
+    {
+      "epoch": 2.7786666666666666,
+      "grad_norm": 2.238597869873047,
+      "learning_rate": 8.893238434163702e-05,
+      "loss": 1.9695,
+      "step": 6252
+    },
+    {
+      "epoch": 2.779111111111111,
+      "grad_norm": 1.4127635955810547,
+      "learning_rate": 8.891459074733096e-05,
+      "loss": 1.1453,
+      "step": 6253
+    },
+    {
+      "epoch": 2.7795555555555556,
+      "grad_norm": 2.074450969696045,
+      "learning_rate": 8.889679715302492e-05,
+      "loss": 1.5515,
+      "step": 6254
+    },
+    {
+      "epoch": 2.7800000000000002,
+      "grad_norm": 2.3815135955810547,
+      "learning_rate": 8.887900355871886e-05,
+      "loss": 1.9383,
+      "step": 6255
+    },
+    {
+      "epoch": 2.7804444444444445,
+      "grad_norm": 2.3668477535247803,
+      "learning_rate": 8.886120996441281e-05,
+      "loss": 1.685,
+      "step": 6256
+    },
+    {
+      "epoch": 2.7808888888888887,
+      "grad_norm": 2.168121576309204,
+      "learning_rate": 8.884341637010677e-05,
+      "loss": 1.783,
+      "step": 6257
+    },
+    {
+      "epoch": 2.7813333333333334,
+      "grad_norm": 1.5242462158203125,
+      "learning_rate": 8.882562277580071e-05,
+      "loss": 1.0202,
+      "step": 6258
+    },
+    {
+      "epoch": 2.7817777777777777,
+      "grad_norm": 2.2029366493225098,
+      "learning_rate": 8.880782918149466e-05,
+      "loss": 1.4853,
+      "step": 6259
+    },
+    {
+      "epoch": 2.7822222222222224,
+      "grad_norm": 2.45417857170105,
+      "learning_rate": 8.879003558718861e-05,
+      "loss": 1.9181,
+      "step": 6260
+    },
+    {
+      "epoch": 2.7826666666666666,
+      "grad_norm": 2.351133346557617,
+      "learning_rate": 8.877224199288257e-05,
+      "loss": 1.9803,
+      "step": 6261
+    },
+    {
+      "epoch": 2.7831111111111113,
+      "grad_norm": 2.16433048248291,
+      "learning_rate": 8.875444839857651e-05,
+      "loss": 1.7061,
+      "step": 6262
+    },
+    {
+      "epoch": 2.7835555555555556,
+      "grad_norm": 2.3350791931152344,
+      "learning_rate": 8.873665480427047e-05,
+      "loss": 1.7437,
+      "step": 6263
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 2.5073392391204834,
+      "learning_rate": 8.871886120996443e-05,
+      "loss": 1.6962,
+      "step": 6264
+    },
+    {
+      "epoch": 2.7844444444444445,
+      "grad_norm": 2.518251657485962,
+      "learning_rate": 8.870106761565836e-05,
+      "loss": 1.6035,
+      "step": 6265
+    },
+    {
+      "epoch": 2.7848888888888887,
+      "grad_norm": 2.514759063720703,
+      "learning_rate": 8.868327402135231e-05,
+      "loss": 1.9147,
+      "step": 6266
+    },
+    {
+      "epoch": 2.7853333333333334,
+      "grad_norm": 2.542900562286377,
+      "learning_rate": 8.866548042704627e-05,
+      "loss": 1.8066,
+      "step": 6267
+    },
+    {
+      "epoch": 2.7857777777777777,
+      "grad_norm": 2.330296754837036,
+      "learning_rate": 8.864768683274021e-05,
+      "loss": 1.7774,
+      "step": 6268
+    },
+    {
+      "epoch": 2.7862222222222224,
+      "grad_norm": 2.444782018661499,
+      "learning_rate": 8.862989323843417e-05,
+      "loss": 2.0453,
+      "step": 6269
+    },
+    {
+      "epoch": 2.7866666666666666,
+      "grad_norm": 2.7158892154693604,
+      "learning_rate": 8.861209964412812e-05,
+      "loss": 2.2444,
+      "step": 6270
+    },
+    {
+      "epoch": 2.787111111111111,
+      "grad_norm": 2.286931037902832,
+      "learning_rate": 8.859430604982207e-05,
+      "loss": 1.1165,
+      "step": 6271
+    },
+    {
+      "epoch": 2.7875555555555556,
+      "grad_norm": 2.6693062782287598,
+      "learning_rate": 8.857651245551601e-05,
+      "loss": 1.8122,
+      "step": 6272
+    },
+    {
+      "epoch": 2.7880000000000003,
+      "grad_norm": 2.527357816696167,
+      "learning_rate": 8.855871886120997e-05,
+      "loss": 1.9374,
+      "step": 6273
+    },
+    {
+      "epoch": 2.7884444444444445,
+      "grad_norm": 2.357048273086548,
+      "learning_rate": 8.854092526690392e-05,
+      "loss": 1.9746,
+      "step": 6274
+    },
+    {
+      "epoch": 2.7888888888888888,
+      "grad_norm": 2.5850024223327637,
+      "learning_rate": 8.852313167259787e-05,
+      "loss": 1.3167,
+      "step": 6275
+    },
+    {
+      "epoch": 2.7893333333333334,
+      "grad_norm": 2.3246774673461914,
+      "learning_rate": 8.850533807829182e-05,
+      "loss": 1.5308,
+      "step": 6276
+    },
+    {
+      "epoch": 2.7897777777777777,
+      "grad_norm": 2.6649742126464844,
+      "learning_rate": 8.848754448398578e-05,
+      "loss": 1.9587,
+      "step": 6277
+    },
+    {
+      "epoch": 2.7902222222222224,
+      "grad_norm": 2.920469284057617,
+      "learning_rate": 8.846975088967971e-05,
+      "loss": 2.0161,
+      "step": 6278
+    },
+    {
+      "epoch": 2.7906666666666666,
+      "grad_norm": 2.4961137771606445,
+      "learning_rate": 8.845195729537367e-05,
+      "loss": 1.7213,
+      "step": 6279
+    },
+    {
+      "epoch": 2.7911111111111113,
+      "grad_norm": 2.586489677429199,
+      "learning_rate": 8.843416370106762e-05,
+      "loss": 1.6672,
+      "step": 6280
+    },
+    {
+      "epoch": 2.7915555555555556,
+      "grad_norm": 2.7179059982299805,
+      "learning_rate": 8.841637010676156e-05,
+      "loss": 1.6846,
+      "step": 6281
+    },
+    {
+      "epoch": 2.792,
+      "grad_norm": 2.3709304332733154,
+      "learning_rate": 8.839857651245552e-05,
+      "loss": 1.6412,
+      "step": 6282
+    },
+    {
+      "epoch": 2.7924444444444445,
+      "grad_norm": 2.750460624694824,
+      "learning_rate": 8.838078291814948e-05,
+      "loss": 1.928,
+      "step": 6283
+    },
+    {
+      "epoch": 2.7928888888888888,
+      "grad_norm": 2.755523681640625,
+      "learning_rate": 8.836298932384342e-05,
+      "loss": 1.3791,
+      "step": 6284
+    },
+    {
+      "epoch": 2.7933333333333334,
+      "grad_norm": 6.7336554527282715,
+      "learning_rate": 8.834519572953736e-05,
+      "loss": 1.7842,
+      "step": 6285
+    },
+    {
+      "epoch": 2.7937777777777777,
+      "grad_norm": 1.775144100189209,
+      "learning_rate": 8.832740213523132e-05,
+      "loss": 0.8977,
+      "step": 6286
+    },
+    {
+      "epoch": 2.7942222222222224,
+      "grad_norm": 1.6157419681549072,
+      "learning_rate": 8.830960854092528e-05,
+      "loss": 0.6727,
+      "step": 6287
+    },
+    {
+      "epoch": 2.7946666666666666,
+      "grad_norm": 3.1187615394592285,
+      "learning_rate": 8.829181494661922e-05,
+      "loss": 2.2613,
+      "step": 6288
+    },
+    {
+      "epoch": 2.795111111111111,
+      "grad_norm": 2.994696617126465,
+      "learning_rate": 8.827402135231318e-05,
+      "loss": 1.5852,
+      "step": 6289
+    },
+    {
+      "epoch": 2.7955555555555556,
+      "grad_norm": 2.502854585647583,
+      "learning_rate": 8.825622775800713e-05,
+      "loss": 1.5286,
+      "step": 6290
+    },
+    {
+      "epoch": 2.7960000000000003,
+      "grad_norm": 2.853635549545288,
+      "learning_rate": 8.823843416370106e-05,
+      "loss": 1.7994,
+      "step": 6291
+    },
+    {
+      "epoch": 2.7964444444444445,
+      "grad_norm": 4.293439865112305,
+      "learning_rate": 8.822064056939502e-05,
+      "loss": 2.0785,
+      "step": 6292
+    },
+    {
+      "epoch": 2.7968888888888888,
+      "grad_norm": 3.213901996612549,
+      "learning_rate": 8.820284697508897e-05,
+      "loss": 1.9483,
+      "step": 6293
+    },
+    {
+      "epoch": 2.7973333333333334,
+      "grad_norm": 3.6944658756256104,
+      "learning_rate": 8.818505338078292e-05,
+      "loss": 2.1967,
+      "step": 6294
+    },
+    {
+      "epoch": 2.7977777777777777,
+      "grad_norm": 2.9698970317840576,
+      "learning_rate": 8.816725978647687e-05,
+      "loss": 1.5886,
+      "step": 6295
+    },
+    {
+      "epoch": 2.7982222222222224,
+      "grad_norm": 3.9858527183532715,
+      "learning_rate": 8.814946619217083e-05,
+      "loss": 1.9551,
+      "step": 6296
+    },
+    {
+      "epoch": 2.7986666666666666,
+      "grad_norm": 4.04840612411499,
+      "learning_rate": 8.813167259786477e-05,
+      "loss": 2.0034,
+      "step": 6297
+    },
+    {
+      "epoch": 2.7991111111111113,
+      "grad_norm": 3.7091634273529053,
+      "learning_rate": 8.811387900355872e-05,
+      "loss": 2.1336,
+      "step": 6298
+    },
+    {
+      "epoch": 2.7995555555555556,
+      "grad_norm": 3.552135944366455,
+      "learning_rate": 8.809608540925267e-05,
+      "loss": 1.8169,
+      "step": 6299
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.049909830093384,
+      "learning_rate": 8.807829181494662e-05,
+      "loss": 0.6878,
+      "step": 6300
+    },
+    {
+      "epoch": 2.8004444444444445,
+      "grad_norm": 3.8162107467651367,
+      "learning_rate": 8.806049822064057e-05,
+      "loss": 2.3236,
+      "step": 6301
+    },
+    {
+      "epoch": 2.8008888888888888,
+      "grad_norm": 1.9239376783370972,
+      "learning_rate": 8.804270462633453e-05,
+      "loss": 1.5104,
+      "step": 6302
+    },
+    {
+      "epoch": 2.8013333333333335,
+      "grad_norm": 1.867793083190918,
+      "learning_rate": 8.802491103202847e-05,
+      "loss": 1.936,
+      "step": 6303
+    },
+    {
+      "epoch": 2.8017777777777777,
+      "grad_norm": 2.062577486038208,
+      "learning_rate": 8.800711743772242e-05,
+      "loss": 1.8487,
+      "step": 6304
+    },
+    {
+      "epoch": 2.8022222222222224,
+      "grad_norm": 2.304241895675659,
+      "learning_rate": 8.798932384341637e-05,
+      "loss": 1.9238,
+      "step": 6305
+    },
+    {
+      "epoch": 2.8026666666666666,
+      "grad_norm": 2.0936267375946045,
+      "learning_rate": 8.797153024911033e-05,
+      "loss": 1.6966,
+      "step": 6306
+    },
+    {
+      "epoch": 2.803111111111111,
+      "grad_norm": 2.325967788696289,
+      "learning_rate": 8.795373665480427e-05,
+      "loss": 2.0191,
+      "step": 6307
+    },
+    {
+      "epoch": 2.8035555555555556,
+      "grad_norm": 2.175997495651245,
+      "learning_rate": 8.793594306049823e-05,
+      "loss": 1.5685,
+      "step": 6308
+    },
+    {
+      "epoch": 2.8040000000000003,
+      "grad_norm": 1.519380807876587,
+      "learning_rate": 8.791814946619218e-05,
+      "loss": 0.9142,
+      "step": 6309
+    },
+    {
+      "epoch": 2.8044444444444445,
+      "grad_norm": 2.5699596405029297,
+      "learning_rate": 8.790035587188611e-05,
+      "loss": 2.3034,
+      "step": 6310
+    },
+    {
+      "epoch": 2.8048888888888888,
+      "grad_norm": 2.172449827194214,
+      "learning_rate": 8.788256227758007e-05,
+      "loss": 1.7681,
+      "step": 6311
+    },
+    {
+      "epoch": 2.8053333333333335,
+      "grad_norm": 2.67712140083313,
+      "learning_rate": 8.786476868327403e-05,
+      "loss": 2.1584,
+      "step": 6312
+    },
+    {
+      "epoch": 2.8057777777777777,
+      "grad_norm": 2.426130771636963,
+      "learning_rate": 8.784697508896797e-05,
+      "loss": 1.7722,
+      "step": 6313
+    },
+    {
+      "epoch": 2.806222222222222,
+      "grad_norm": 2.3680663108825684,
+      "learning_rate": 8.782918149466193e-05,
+      "loss": 1.886,
+      "step": 6314
+    },
+    {
+      "epoch": 2.8066666666666666,
+      "grad_norm": 2.257155179977417,
+      "learning_rate": 8.781138790035588e-05,
+      "loss": 1.7808,
+      "step": 6315
+    },
+    {
+      "epoch": 2.8071111111111113,
+      "grad_norm": 2.636523485183716,
+      "learning_rate": 8.779359430604983e-05,
+      "loss": 1.7881,
+      "step": 6316
+    },
+    {
+      "epoch": 2.8075555555555556,
+      "grad_norm": 2.501422882080078,
+      "learning_rate": 8.777580071174377e-05,
+      "loss": 1.814,
+      "step": 6317
+    },
+    {
+      "epoch": 2.808,
+      "grad_norm": 2.323500394821167,
+      "learning_rate": 8.775800711743772e-05,
+      "loss": 1.7802,
+      "step": 6318
+    },
+    {
+      "epoch": 2.8084444444444445,
+      "grad_norm": 1.7834057807922363,
+      "learning_rate": 8.774021352313168e-05,
+      "loss": 0.901,
+      "step": 6319
+    },
+    {
+      "epoch": 2.8088888888888888,
+      "grad_norm": 1.7829649448394775,
+      "learning_rate": 8.772241992882562e-05,
+      "loss": 0.9174,
+      "step": 6320
+    },
+    {
+      "epoch": 2.8093333333333335,
+      "grad_norm": 2.242635488510132,
+      "learning_rate": 8.770462633451958e-05,
+      "loss": 1.7682,
+      "step": 6321
+    },
+    {
+      "epoch": 2.8097777777777777,
+      "grad_norm": 2.7776167392730713,
+      "learning_rate": 8.768683274021354e-05,
+      "loss": 2.1089,
+      "step": 6322
+    },
+    {
+      "epoch": 2.8102222222222224,
+      "grad_norm": 2.151515483856201,
+      "learning_rate": 8.766903914590747e-05,
+      "loss": 1.4301,
+      "step": 6323
+    },
+    {
+      "epoch": 2.8106666666666666,
+      "grad_norm": 2.3415439128875732,
+      "learning_rate": 8.765124555160142e-05,
+      "loss": 1.8939,
+      "step": 6324
+    },
+    {
+      "epoch": 2.811111111111111,
+      "grad_norm": 2.3634235858917236,
+      "learning_rate": 8.763345195729538e-05,
+      "loss": 1.9442,
+      "step": 6325
+    },
+    {
+      "epoch": 2.8115555555555556,
+      "grad_norm": 2.3634817600250244,
+      "learning_rate": 8.761565836298932e-05,
+      "loss": 1.3034,
+      "step": 6326
+    },
+    {
+      "epoch": 2.8120000000000003,
+      "grad_norm": 2.2050936222076416,
+      "learning_rate": 8.759786476868328e-05,
+      "loss": 1.4394,
+      "step": 6327
+    },
+    {
+      "epoch": 2.8124444444444445,
+      "grad_norm": 2.491987466812134,
+      "learning_rate": 8.758007117437724e-05,
+      "loss": 1.4429,
+      "step": 6328
+    },
+    {
+      "epoch": 2.8128888888888888,
+      "grad_norm": 2.7539877891540527,
+      "learning_rate": 8.756227758007118e-05,
+      "loss": 1.8103,
+      "step": 6329
+    },
+    {
+      "epoch": 2.8133333333333335,
+      "grad_norm": 1.8611799478530884,
+      "learning_rate": 8.754448398576512e-05,
+      "loss": 0.8885,
+      "step": 6330
+    },
+    {
+      "epoch": 2.8137777777777777,
+      "grad_norm": 2.7642264366149902,
+      "learning_rate": 8.752669039145908e-05,
+      "loss": 1.6687,
+      "step": 6331
+    },
+    {
+      "epoch": 2.814222222222222,
+      "grad_norm": 2.5059311389923096,
+      "learning_rate": 8.750889679715303e-05,
+      "loss": 1.5068,
+      "step": 6332
+    },
+    {
+      "epoch": 2.8146666666666667,
+      "grad_norm": 2.761054515838623,
+      "learning_rate": 8.749110320284698e-05,
+      "loss": 1.7512,
+      "step": 6333
+    },
+    {
+      "epoch": 2.8151111111111113,
+      "grad_norm": 2.5050809383392334,
+      "learning_rate": 8.747330960854093e-05,
+      "loss": 1.4633,
+      "step": 6334
+    },
+    {
+      "epoch": 2.8155555555555556,
+      "grad_norm": 3.096000909805298,
+      "learning_rate": 8.745551601423489e-05,
+      "loss": 1.7558,
+      "step": 6335
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 3.1274073123931885,
+      "learning_rate": 8.743772241992882e-05,
+      "loss": 1.614,
+      "step": 6336
+    },
+    {
+      "epoch": 2.8164444444444445,
+      "grad_norm": 3.165975332260132,
+      "learning_rate": 8.741992882562278e-05,
+      "loss": 1.5489,
+      "step": 6337
+    },
+    {
+      "epoch": 2.8168888888888888,
+      "grad_norm": 3.016789674758911,
+      "learning_rate": 8.740213523131673e-05,
+      "loss": 1.8568,
+      "step": 6338
+    },
+    {
+      "epoch": 2.8173333333333335,
+      "grad_norm": 2.219804525375366,
+      "learning_rate": 8.738434163701068e-05,
+      "loss": 1.0983,
+      "step": 6339
+    },
+    {
+      "epoch": 2.8177777777777777,
+      "grad_norm": 3.128307819366455,
+      "learning_rate": 8.736654804270463e-05,
+      "loss": 1.4682,
+      "step": 6340
+    },
+    {
+      "epoch": 2.8182222222222224,
+      "grad_norm": 3.149846076965332,
+      "learning_rate": 8.734875444839859e-05,
+      "loss": 1.8692,
+      "step": 6341
+    },
+    {
+      "epoch": 2.8186666666666667,
+      "grad_norm": 2.897365093231201,
+      "learning_rate": 8.733096085409253e-05,
+      "loss": 1.5963,
+      "step": 6342
+    },
+    {
+      "epoch": 2.819111111111111,
+      "grad_norm": 3.354923963546753,
+      "learning_rate": 8.731316725978647e-05,
+      "loss": 1.8655,
+      "step": 6343
+    },
+    {
+      "epoch": 2.8195555555555556,
+      "grad_norm": 3.33359432220459,
+      "learning_rate": 8.729537366548043e-05,
+      "loss": 1.8005,
+      "step": 6344
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.5689620971679688,
+      "learning_rate": 8.727758007117437e-05,
+      "loss": 1.5202,
+      "step": 6345
+    },
+    {
+      "epoch": 2.8204444444444445,
+      "grad_norm": 4.082332134246826,
+      "learning_rate": 8.725978647686833e-05,
+      "loss": 2.2171,
+      "step": 6346
+    },
+    {
+      "epoch": 2.820888888888889,
+      "grad_norm": 3.0966172218322754,
+      "learning_rate": 8.724199288256229e-05,
+      "loss": 1.9997,
+      "step": 6347
+    },
+    {
+      "epoch": 2.8213333333333335,
+      "grad_norm": 3.0002849102020264,
+      "learning_rate": 8.722419928825623e-05,
+      "loss": 1.488,
+      "step": 6348
+    },
+    {
+      "epoch": 2.8217777777777777,
+      "grad_norm": 4.207763671875,
+      "learning_rate": 8.720640569395017e-05,
+      "loss": 1.2479,
+      "step": 6349
+    },
+    {
+      "epoch": 2.822222222222222,
+      "grad_norm": 2.4575085639953613,
+      "learning_rate": 8.718861209964413e-05,
+      "loss": 0.5755,
+      "step": 6350
+    },
+    {
+      "epoch": 2.8226666666666667,
+      "grad_norm": 1.834578275680542,
+      "learning_rate": 8.717081850533809e-05,
+      "loss": 2.0906,
+      "step": 6351
+    },
+    {
+      "epoch": 2.8231111111111113,
+      "grad_norm": 1.9886993169784546,
+      "learning_rate": 8.715302491103203e-05,
+      "loss": 1.6937,
+      "step": 6352
+    },
+    {
+      "epoch": 2.8235555555555556,
+      "grad_norm": 1.45979905128479,
+      "learning_rate": 8.713523131672599e-05,
+      "loss": 1.3237,
+      "step": 6353
+    },
+    {
+      "epoch": 2.824,
+      "grad_norm": 2.220140218734741,
+      "learning_rate": 8.711743772241994e-05,
+      "loss": 2.1938,
+      "step": 6354
+    },
+    {
+      "epoch": 2.8244444444444445,
+      "grad_norm": 2.36525821685791,
+      "learning_rate": 8.709964412811388e-05,
+      "loss": 1.7316,
+      "step": 6355
+    },
+    {
+      "epoch": 2.824888888888889,
+      "grad_norm": 2.672027587890625,
+      "learning_rate": 8.708185053380783e-05,
+      "loss": 2.1021,
+      "step": 6356
+    },
+    {
+      "epoch": 2.8253333333333335,
+      "grad_norm": 2.566497325897217,
+      "learning_rate": 8.706405693950178e-05,
+      "loss": 2.0366,
+      "step": 6357
+    },
+    {
+      "epoch": 2.8257777777777777,
+      "grad_norm": 2.3866355419158936,
+      "learning_rate": 8.704626334519573e-05,
+      "loss": 1.8904,
+      "step": 6358
+    },
+    {
+      "epoch": 2.8262222222222224,
+      "grad_norm": 2.158238172531128,
+      "learning_rate": 8.702846975088968e-05,
+      "loss": 1.9666,
+      "step": 6359
+    },
+    {
+      "epoch": 2.8266666666666667,
+      "grad_norm": 2.4062252044677734,
+      "learning_rate": 8.701067615658364e-05,
+      "loss": 1.7557,
+      "step": 6360
+    },
+    {
+      "epoch": 2.827111111111111,
+      "grad_norm": 2.215730905532837,
+      "learning_rate": 8.699288256227758e-05,
+      "loss": 1.0731,
+      "step": 6361
+    },
+    {
+      "epoch": 2.8275555555555556,
+      "grad_norm": 2.172459125518799,
+      "learning_rate": 8.697508896797153e-05,
+      "loss": 1.7543,
+      "step": 6362
+    },
+    {
+      "epoch": 2.828,
+      "grad_norm": 2.4174609184265137,
+      "learning_rate": 8.695729537366548e-05,
+      "loss": 1.5526,
+      "step": 6363
+    },
+    {
+      "epoch": 2.8284444444444445,
+      "grad_norm": 1.6709802150726318,
+      "learning_rate": 8.693950177935944e-05,
+      "loss": 0.8726,
+      "step": 6364
+    },
+    {
+      "epoch": 2.828888888888889,
+      "grad_norm": 2.830564022064209,
+      "learning_rate": 8.692170818505338e-05,
+      "loss": 1.9783,
+      "step": 6365
+    },
+    {
+      "epoch": 2.8293333333333335,
+      "grad_norm": 2.5049219131469727,
+      "learning_rate": 8.690391459074734e-05,
+      "loss": 1.7796,
+      "step": 6366
+    },
+    {
+      "epoch": 2.8297777777777777,
+      "grad_norm": 2.455636739730835,
+      "learning_rate": 8.68861209964413e-05,
+      "loss": 1.807,
+      "step": 6367
+    },
+    {
+      "epoch": 2.830222222222222,
+      "grad_norm": 2.934166431427002,
+      "learning_rate": 8.686832740213522e-05,
+      "loss": 1.4921,
+      "step": 6368
+    },
+    {
+      "epoch": 2.8306666666666667,
+      "grad_norm": 3.1317198276519775,
+      "learning_rate": 8.685053380782918e-05,
+      "loss": 1.7484,
+      "step": 6369
+    },
+    {
+      "epoch": 2.8311111111111114,
+      "grad_norm": 3.014157295227051,
+      "learning_rate": 8.683274021352314e-05,
+      "loss": 2.2238,
+      "step": 6370
+    },
+    {
+      "epoch": 2.8315555555555556,
+      "grad_norm": 2.56988787651062,
+      "learning_rate": 8.681494661921708e-05,
+      "loss": 1.6546,
+      "step": 6371
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 2.6320784091949463,
+      "learning_rate": 8.679715302491104e-05,
+      "loss": 2.1411,
+      "step": 6372
+    },
+    {
+      "epoch": 2.8324444444444445,
+      "grad_norm": 2.682128667831421,
+      "learning_rate": 8.6779359430605e-05,
+      "loss": 2.1698,
+      "step": 6373
+    },
+    {
+      "epoch": 2.832888888888889,
+      "grad_norm": 2.754392623901367,
+      "learning_rate": 8.676156583629894e-05,
+      "loss": 1.8696,
+      "step": 6374
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 2.4099462032318115,
+      "learning_rate": 8.674377224199288e-05,
+      "loss": 1.8406,
+      "step": 6375
+    },
+    {
+      "epoch": 2.8337777777777777,
+      "grad_norm": 3.0007071495056152,
+      "learning_rate": 8.672597864768684e-05,
+      "loss": 1.7611,
+      "step": 6376
+    },
+    {
+      "epoch": 2.8342222222222224,
+      "grad_norm": 3.2685530185699463,
+      "learning_rate": 8.670818505338079e-05,
+      "loss": 1.8421,
+      "step": 6377
+    },
+    {
+      "epoch": 2.8346666666666667,
+      "grad_norm": 2.7052693367004395,
+      "learning_rate": 8.669039145907474e-05,
+      "loss": 1.6633,
+      "step": 6378
+    },
+    {
+      "epoch": 2.835111111111111,
+      "grad_norm": 2.628025531768799,
+      "learning_rate": 8.667259786476869e-05,
+      "loss": 1.7274,
+      "step": 6379
+    },
+    {
+      "epoch": 2.8355555555555556,
+      "grad_norm": 1.5739624500274658,
+      "learning_rate": 8.665480427046265e-05,
+      "loss": 0.7554,
+      "step": 6380
+    },
+    {
+      "epoch": 2.836,
+      "grad_norm": 1.9430078268051147,
+      "learning_rate": 8.663701067615658e-05,
+      "loss": 0.8761,
+      "step": 6381
+    },
+    {
+      "epoch": 2.8364444444444445,
+      "grad_norm": 2.2438039779663086,
+      "learning_rate": 8.661921708185053e-05,
+      "loss": 1.1933,
+      "step": 6382
+    },
+    {
+      "epoch": 2.836888888888889,
+      "grad_norm": 2.6071414947509766,
+      "learning_rate": 8.660142348754449e-05,
+      "loss": 1.7854,
+      "step": 6383
+    },
+    {
+      "epoch": 2.8373333333333335,
+      "grad_norm": 1.6424260139465332,
+      "learning_rate": 8.658362989323843e-05,
+      "loss": 0.7366,
+      "step": 6384
+    },
+    {
+      "epoch": 2.8377777777777777,
+      "grad_norm": 2.981468439102173,
+      "learning_rate": 8.656583629893239e-05,
+      "loss": 1.7676,
+      "step": 6385
+    },
+    {
+      "epoch": 2.838222222222222,
+      "grad_norm": 2.440037965774536,
+      "learning_rate": 8.654804270462635e-05,
+      "loss": 1.5232,
+      "step": 6386
+    },
+    {
+      "epoch": 2.8386666666666667,
+      "grad_norm": 2.5505833625793457,
+      "learning_rate": 8.653024911032029e-05,
+      "loss": 1.4548,
+      "step": 6387
+    },
+    {
+      "epoch": 2.8391111111111114,
+      "grad_norm": 2.7578847408294678,
+      "learning_rate": 8.651245551601423e-05,
+      "loss": 1.5028,
+      "step": 6388
+    },
+    {
+      "epoch": 2.8395555555555556,
+      "grad_norm": 2.750519275665283,
+      "learning_rate": 8.649466192170819e-05,
+      "loss": 1.7589,
+      "step": 6389
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 2.7024030685424805,
+      "learning_rate": 8.647686832740213e-05,
+      "loss": 1.6456,
+      "step": 6390
+    },
+    {
+      "epoch": 2.8404444444444445,
+      "grad_norm": 2.8694217205047607,
+      "learning_rate": 8.645907473309609e-05,
+      "loss": 1.0521,
+      "step": 6391
+    },
+    {
+      "epoch": 2.840888888888889,
+      "grad_norm": 2.930608034133911,
+      "learning_rate": 8.644128113879004e-05,
+      "loss": 1.6561,
+      "step": 6392
+    },
+    {
+      "epoch": 2.8413333333333335,
+      "grad_norm": 2.9115495681762695,
+      "learning_rate": 8.642348754448399e-05,
+      "loss": 1.8796,
+      "step": 6393
+    },
+    {
+      "epoch": 2.8417777777777777,
+      "grad_norm": 3.2313528060913086,
+      "learning_rate": 8.640569395017793e-05,
+      "loss": 1.8615,
+      "step": 6394
+    },
+    {
+      "epoch": 2.8422222222222224,
+      "grad_norm": 2.823005437850952,
+      "learning_rate": 8.638790035587189e-05,
+      "loss": 1.7883,
+      "step": 6395
+    },
+    {
+      "epoch": 2.8426666666666667,
+      "grad_norm": 3.2536568641662598,
+      "learning_rate": 8.637010676156584e-05,
+      "loss": 1.734,
+      "step": 6396
+    },
+    {
+      "epoch": 2.843111111111111,
+      "grad_norm": 3.4909675121307373,
+      "learning_rate": 8.635231316725979e-05,
+      "loss": 1.5882,
+      "step": 6397
+    },
+    {
+      "epoch": 2.8435555555555556,
+      "grad_norm": 3.702479124069214,
+      "learning_rate": 8.633451957295374e-05,
+      "loss": 1.7106,
+      "step": 6398
+    },
+    {
+      "epoch": 2.844,
+      "grad_norm": 3.9514353275299072,
+      "learning_rate": 8.63167259786477e-05,
+      "loss": 2.0991,
+      "step": 6399
+    },
+    {
+      "epoch": 2.8444444444444446,
+      "grad_norm": 3.16878080368042,
+      "learning_rate": 8.629893238434164e-05,
+      "loss": 1.7365,
+      "step": 6400
+    },
+    {
+      "epoch": 2.844888888888889,
+      "grad_norm": 1.7129290103912354,
+      "learning_rate": 8.628113879003559e-05,
+      "loss": 2.2009,
+      "step": 6401
+    },
+    {
+      "epoch": 2.8453333333333335,
+      "grad_norm": 1.914031744003296,
+      "learning_rate": 8.626334519572954e-05,
+      "loss": 2.0489,
+      "step": 6402
+    },
+    {
+      "epoch": 2.8457777777777777,
+      "grad_norm": 2.1656267642974854,
+      "learning_rate": 8.624555160142348e-05,
+      "loss": 2.4192,
+      "step": 6403
+    },
+    {
+      "epoch": 2.846222222222222,
+      "grad_norm": 2.315261125564575,
+      "learning_rate": 8.622775800711744e-05,
+      "loss": 1.8947,
+      "step": 6404
+    },
+    {
+      "epoch": 2.8466666666666667,
+      "grad_norm": 2.6250321865081787,
+      "learning_rate": 8.62099644128114e-05,
+      "loss": 2.3402,
+      "step": 6405
+    },
+    {
+      "epoch": 2.8471111111111114,
+      "grad_norm": 2.3993537425994873,
+      "learning_rate": 8.619217081850534e-05,
+      "loss": 1.9262,
+      "step": 6406
+    },
+    {
+      "epoch": 2.8475555555555556,
+      "grad_norm": 2.0765573978424072,
+      "learning_rate": 8.617437722419928e-05,
+      "loss": 2.1101,
+      "step": 6407
+    },
+    {
+      "epoch": 2.848,
+      "grad_norm": 2.398111343383789,
+      "learning_rate": 8.615658362989324e-05,
+      "loss": 1.8188,
+      "step": 6408
+    },
+    {
+      "epoch": 2.8484444444444446,
+      "grad_norm": 2.267061471939087,
+      "learning_rate": 8.61387900355872e-05,
+      "loss": 2.3754,
+      "step": 6409
+    },
+    {
+      "epoch": 2.848888888888889,
+      "grad_norm": 2.5441832542419434,
+      "learning_rate": 8.612099644128114e-05,
+      "loss": 1.6375,
+      "step": 6410
+    },
+    {
+      "epoch": 2.8493333333333335,
+      "grad_norm": 2.2044379711151123,
+      "learning_rate": 8.61032028469751e-05,
+      "loss": 1.8951,
+      "step": 6411
+    },
+    {
+      "epoch": 2.8497777777777777,
+      "grad_norm": 2.2039835453033447,
+      "learning_rate": 8.608540925266905e-05,
+      "loss": 1.8844,
+      "step": 6412
+    },
+    {
+      "epoch": 2.8502222222222224,
+      "grad_norm": 2.50587797164917,
+      "learning_rate": 8.6067615658363e-05,
+      "loss": 1.9177,
+      "step": 6413
+    },
+    {
+      "epoch": 2.8506666666666667,
+      "grad_norm": 2.2638602256774902,
+      "learning_rate": 8.604982206405694e-05,
+      "loss": 1.5041,
+      "step": 6414
+    },
+    {
+      "epoch": 2.851111111111111,
+      "grad_norm": 2.9379706382751465,
+      "learning_rate": 8.60320284697509e-05,
+      "loss": 1.5956,
+      "step": 6415
+    },
+    {
+      "epoch": 2.8515555555555556,
+      "grad_norm": 2.415709972381592,
+      "learning_rate": 8.601423487544484e-05,
+      "loss": 1.5924,
+      "step": 6416
+    },
+    {
+      "epoch": 2.852,
+      "grad_norm": 2.4809000492095947,
+      "learning_rate": 8.59964412811388e-05,
+      "loss": 1.9965,
+      "step": 6417
+    },
+    {
+      "epoch": 2.8524444444444446,
+      "grad_norm": 2.643976926803589,
+      "learning_rate": 8.597864768683275e-05,
+      "loss": 2.1711,
+      "step": 6418
+    },
+    {
+      "epoch": 2.852888888888889,
+      "grad_norm": 2.432825803756714,
+      "learning_rate": 8.59608540925267e-05,
+      "loss": 1.4952,
+      "step": 6419
+    },
+    {
+      "epoch": 2.8533333333333335,
+      "grad_norm": 2.853224277496338,
+      "learning_rate": 8.594306049822064e-05,
+      "loss": 1.5422,
+      "step": 6420
+    },
+    {
+      "epoch": 2.8537777777777777,
+      "grad_norm": 2.624593496322632,
+      "learning_rate": 8.59252669039146e-05,
+      "loss": 1.5306,
+      "step": 6421
+    },
+    {
+      "epoch": 2.854222222222222,
+      "grad_norm": 2.337578296661377,
+      "learning_rate": 8.590747330960855e-05,
+      "loss": 1.6708,
+      "step": 6422
+    },
+    {
+      "epoch": 2.8546666666666667,
+      "grad_norm": 2.62070369720459,
+      "learning_rate": 8.588967971530249e-05,
+      "loss": 1.799,
+      "step": 6423
+    },
+    {
+      "epoch": 2.8551111111111114,
+      "grad_norm": 2.670642852783203,
+      "learning_rate": 8.587188612099645e-05,
+      "loss": 1.7654,
+      "step": 6424
+    },
+    {
+      "epoch": 2.8555555555555556,
+      "grad_norm": 2.330082416534424,
+      "learning_rate": 8.58540925266904e-05,
+      "loss": 1.4492,
+      "step": 6425
+    },
+    {
+      "epoch": 2.856,
+      "grad_norm": 2.2589471340179443,
+      "learning_rate": 8.583629893238434e-05,
+      "loss": 1.5373,
+      "step": 6426
+    },
+    {
+      "epoch": 2.8564444444444446,
+      "grad_norm": 2.6059587001800537,
+      "learning_rate": 8.581850533807829e-05,
+      "loss": 1.8362,
+      "step": 6427
+    },
+    {
+      "epoch": 2.856888888888889,
+      "grad_norm": 2.6167008876800537,
+      "learning_rate": 8.580071174377225e-05,
+      "loss": 1.7661,
+      "step": 6428
+    },
+    {
+      "epoch": 2.857333333333333,
+      "grad_norm": 2.616682291030884,
+      "learning_rate": 8.578291814946619e-05,
+      "loss": 1.8365,
+      "step": 6429
+    },
+    {
+      "epoch": 2.8577777777777778,
+      "grad_norm": 0.22476230561733246,
+      "learning_rate": 8.576512455516015e-05,
+      "loss": 0.0266,
+      "step": 6430
+    },
+    {
+      "epoch": 2.8582222222222224,
+      "grad_norm": 1.9581819772720337,
+      "learning_rate": 8.57473309608541e-05,
+      "loss": 1.0284,
+      "step": 6431
+    },
+    {
+      "epoch": 2.8586666666666667,
+      "grad_norm": 2.870471477508545,
+      "learning_rate": 8.572953736654805e-05,
+      "loss": 1.8517,
+      "step": 6432
+    },
+    {
+      "epoch": 2.859111111111111,
+      "grad_norm": 2.1929726600646973,
+      "learning_rate": 8.571174377224199e-05,
+      "loss": 1.1532,
+      "step": 6433
+    },
+    {
+      "epoch": 2.8595555555555556,
+      "grad_norm": 2.5478527545928955,
+      "learning_rate": 8.569395017793595e-05,
+      "loss": 1.6376,
+      "step": 6434
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.627499580383301,
+      "learning_rate": 8.567615658362989e-05,
+      "loss": 1.6505,
+      "step": 6435
+    },
+    {
+      "epoch": 2.8604444444444446,
+      "grad_norm": 2.877537727355957,
+      "learning_rate": 8.565836298932385e-05,
+      "loss": 2.0833,
+      "step": 6436
+    },
+    {
+      "epoch": 2.860888888888889,
+      "grad_norm": 2.8161423206329346,
+      "learning_rate": 8.56405693950178e-05,
+      "loss": 1.7967,
+      "step": 6437
+    },
+    {
+      "epoch": 2.8613333333333335,
+      "grad_norm": 3.0182766914367676,
+      "learning_rate": 8.562277580071175e-05,
+      "loss": 1.8528,
+      "step": 6438
+    },
+    {
+      "epoch": 2.8617777777777778,
+      "grad_norm": 2.9146299362182617,
+      "learning_rate": 8.560498220640569e-05,
+      "loss": 1.3118,
+      "step": 6439
+    },
+    {
+      "epoch": 2.862222222222222,
+      "grad_norm": 4.448184967041016,
+      "learning_rate": 8.558718861209964e-05,
+      "loss": 1.9648,
+      "step": 6440
+    },
+    {
+      "epoch": 2.8626666666666667,
+      "grad_norm": 3.721683979034424,
+      "learning_rate": 8.55693950177936e-05,
+      "loss": 2.0082,
+      "step": 6441
+    },
+    {
+      "epoch": 2.8631111111111114,
+      "grad_norm": 2.6923537254333496,
+      "learning_rate": 8.555160142348754e-05,
+      "loss": 1.5117,
+      "step": 6442
+    },
+    {
+      "epoch": 2.8635555555555556,
+      "grad_norm": 3.0055980682373047,
+      "learning_rate": 8.55338078291815e-05,
+      "loss": 1.688,
+      "step": 6443
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 2.95833158493042,
+      "learning_rate": 8.551601423487546e-05,
+      "loss": 1.8496,
+      "step": 6444
+    },
+    {
+      "epoch": 2.8644444444444446,
+      "grad_norm": 3.0208351612091064,
+      "learning_rate": 8.54982206405694e-05,
+      "loss": 1.6707,
+      "step": 6445
+    },
+    {
+      "epoch": 2.864888888888889,
+      "grad_norm": 3.0486018657684326,
+      "learning_rate": 8.548042704626334e-05,
+      "loss": 1.6162,
+      "step": 6446
+    },
+    {
+      "epoch": 2.865333333333333,
+      "grad_norm": 4.173706531524658,
+      "learning_rate": 8.54626334519573e-05,
+      "loss": 2.0012,
+      "step": 6447
+    },
+    {
+      "epoch": 2.8657777777777778,
+      "grad_norm": 3.727224349975586,
+      "learning_rate": 8.544483985765124e-05,
+      "loss": 2.3882,
+      "step": 6448
+    },
+    {
+      "epoch": 2.8662222222222224,
+      "grad_norm": 3.276524782180786,
+      "learning_rate": 8.54270462633452e-05,
+      "loss": 1.3258,
+      "step": 6449
+    },
+    {
+      "epoch": 2.8666666666666667,
+      "grad_norm": 3.5537030696868896,
+      "learning_rate": 8.540925266903916e-05,
+      "loss": 0.7531,
+      "step": 6450
+    },
+    {
+      "epoch": 2.867111111111111,
+      "grad_norm": 2.074697732925415,
+      "learning_rate": 8.53914590747331e-05,
+      "loss": 2.3748,
+      "step": 6451
+    },
+    {
+      "epoch": 2.8675555555555556,
+      "grad_norm": 1.607412576675415,
+      "learning_rate": 8.537366548042704e-05,
+      "loss": 1.1406,
+      "step": 6452
+    },
+    {
+      "epoch": 2.868,
+      "grad_norm": 1.7138793468475342,
+      "learning_rate": 8.5355871886121e-05,
+      "loss": 1.1559,
+      "step": 6453
+    },
+    {
+      "epoch": 2.8684444444444446,
+      "grad_norm": 2.1833138465881348,
+      "learning_rate": 8.533807829181495e-05,
+      "loss": 1.4804,
+      "step": 6454
+    },
+    {
+      "epoch": 2.868888888888889,
+      "grad_norm": 2.5808310508728027,
+      "learning_rate": 8.53202846975089e-05,
+      "loss": 2.0459,
+      "step": 6455
+    },
+    {
+      "epoch": 2.8693333333333335,
+      "grad_norm": 2.10170316696167,
+      "learning_rate": 8.530249110320285e-05,
+      "loss": 1.5515,
+      "step": 6456
+    },
+    {
+      "epoch": 2.8697777777777778,
+      "grad_norm": 2.5783615112304688,
+      "learning_rate": 8.528469750889681e-05,
+      "loss": 1.7325,
+      "step": 6457
+    },
+    {
+      "epoch": 2.870222222222222,
+      "grad_norm": 1.9656144380569458,
+      "learning_rate": 8.526690391459075e-05,
+      "loss": 1.6021,
+      "step": 6458
+    },
+    {
+      "epoch": 2.8706666666666667,
+      "grad_norm": 2.899205446243286,
+      "learning_rate": 8.52491103202847e-05,
+      "loss": 1.5858,
+      "step": 6459
+    },
+    {
+      "epoch": 2.871111111111111,
+      "grad_norm": 2.2554736137390137,
+      "learning_rate": 8.523131672597865e-05,
+      "loss": 1.8979,
+      "step": 6460
+    },
+    {
+      "epoch": 2.8715555555555556,
+      "grad_norm": 2.1474039554595947,
+      "learning_rate": 8.52135231316726e-05,
+      "loss": 1.6935,
+      "step": 6461
+    },
+    {
+      "epoch": 2.872,
+      "grad_norm": 2.697214365005493,
+      "learning_rate": 8.519572953736655e-05,
+      "loss": 1.746,
+      "step": 6462
+    },
+    {
+      "epoch": 2.8724444444444446,
+      "grad_norm": 2.4106836318969727,
+      "learning_rate": 8.517793594306051e-05,
+      "loss": 1.7322,
+      "step": 6463
+    },
+    {
+      "epoch": 2.872888888888889,
+      "grad_norm": 2.5296225547790527,
+      "learning_rate": 8.516014234875445e-05,
+      "loss": 1.6172,
+      "step": 6464
+    },
+    {
+      "epoch": 2.873333333333333,
+      "grad_norm": 2.625232696533203,
+      "learning_rate": 8.51423487544484e-05,
+      "loss": 1.9533,
+      "step": 6465
+    },
+    {
+      "epoch": 2.8737777777777778,
+      "grad_norm": 2.728658676147461,
+      "learning_rate": 8.512455516014235e-05,
+      "loss": 1.8563,
+      "step": 6466
+    },
+    {
+      "epoch": 2.8742222222222225,
+      "grad_norm": 2.2425432205200195,
+      "learning_rate": 8.510676156583631e-05,
+      "loss": 1.1704,
+      "step": 6467
+    },
+    {
+      "epoch": 2.8746666666666667,
+      "grad_norm": 1.9965929985046387,
+      "learning_rate": 8.508896797153025e-05,
+      "loss": 1.2467,
+      "step": 6468
+    },
+    {
+      "epoch": 2.875111111111111,
+      "grad_norm": 2.5652081966400146,
+      "learning_rate": 8.507117437722421e-05,
+      "loss": 1.9512,
+      "step": 6469
+    },
+    {
+      "epoch": 2.8755555555555556,
+      "grad_norm": 2.329965591430664,
+      "learning_rate": 8.505338078291816e-05,
+      "loss": 1.81,
+      "step": 6470
+    },
+    {
+      "epoch": 2.876,
+      "grad_norm": 3.135545015335083,
+      "learning_rate": 8.50355871886121e-05,
+      "loss": 1.9133,
+      "step": 6471
+    },
+    {
+      "epoch": 2.8764444444444446,
+      "grad_norm": 2.618896722793579,
+      "learning_rate": 8.501779359430605e-05,
+      "loss": 2.2505,
+      "step": 6472
+    },
+    {
+      "epoch": 2.876888888888889,
+      "grad_norm": 3.150949001312256,
+      "learning_rate": 8.5e-05,
+      "loss": 2.0441,
+      "step": 6473
+    },
+    {
+      "epoch": 2.8773333333333335,
+      "grad_norm": 2.217580556869507,
+      "learning_rate": 8.498220640569395e-05,
+      "loss": 1.4439,
+      "step": 6474
+    },
+    {
+      "epoch": 2.8777777777777778,
+      "grad_norm": 2.7368547916412354,
+      "learning_rate": 8.49644128113879e-05,
+      "loss": 2.0908,
+      "step": 6475
+    },
+    {
+      "epoch": 2.878222222222222,
+      "grad_norm": 2.5103628635406494,
+      "learning_rate": 8.494661921708186e-05,
+      "loss": 1.9196,
+      "step": 6476
+    },
+    {
+      "epoch": 2.8786666666666667,
+      "grad_norm": 2.70851469039917,
+      "learning_rate": 8.49288256227758e-05,
+      "loss": 1.9561,
+      "step": 6477
+    },
+    {
+      "epoch": 2.879111111111111,
+      "grad_norm": 2.451458692550659,
+      "learning_rate": 8.491103202846975e-05,
+      "loss": 1.6853,
+      "step": 6478
+    },
+    {
+      "epoch": 2.8795555555555556,
+      "grad_norm": 3.1393275260925293,
+      "learning_rate": 8.48932384341637e-05,
+      "loss": 1.797,
+      "step": 6479
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.9535908699035645,
+      "learning_rate": 8.487544483985765e-05,
+      "loss": 2.0018,
+      "step": 6480
+    },
+    {
+      "epoch": 2.8804444444444446,
+      "grad_norm": 2.7212975025177,
+      "learning_rate": 8.48576512455516e-05,
+      "loss": 1.7088,
+      "step": 6481
+    },
+    {
+      "epoch": 2.880888888888889,
+      "grad_norm": 2.8284823894500732,
+      "learning_rate": 8.483985765124556e-05,
+      "loss": 1.9613,
+      "step": 6482
+    },
+    {
+      "epoch": 2.881333333333333,
+      "grad_norm": 3.322567939758301,
+      "learning_rate": 8.48220640569395e-05,
+      "loss": 1.8317,
+      "step": 6483
+    },
+    {
+      "epoch": 2.8817777777777778,
+      "grad_norm": 2.870798349380493,
+      "learning_rate": 8.480427046263345e-05,
+      "loss": 1.6794,
+      "step": 6484
+    },
+    {
+      "epoch": 2.8822222222222225,
+      "grad_norm": 2.9603869915008545,
+      "learning_rate": 8.47864768683274e-05,
+      "loss": 1.9088,
+      "step": 6485
+    },
+    {
+      "epoch": 2.8826666666666667,
+      "grad_norm": 3.2760636806488037,
+      "learning_rate": 8.476868327402136e-05,
+      "loss": 1.3813,
+      "step": 6486
+    },
+    {
+      "epoch": 2.883111111111111,
+      "grad_norm": 2.696504831314087,
+      "learning_rate": 8.47508896797153e-05,
+      "loss": 1.692,
+      "step": 6487
+    },
+    {
+      "epoch": 2.8835555555555556,
+      "grad_norm": 3.375922679901123,
+      "learning_rate": 8.473309608540926e-05,
+      "loss": 1.5721,
+      "step": 6488
+    },
+    {
+      "epoch": 2.884,
+      "grad_norm": 2.901848316192627,
+      "learning_rate": 8.471530249110322e-05,
+      "loss": 1.589,
+      "step": 6489
+    },
+    {
+      "epoch": 2.8844444444444446,
+      "grad_norm": 2.677187919616699,
+      "learning_rate": 8.469750889679716e-05,
+      "loss": 2.0427,
+      "step": 6490
+    },
+    {
+      "epoch": 2.884888888888889,
+      "grad_norm": 3.2531144618988037,
+      "learning_rate": 8.46797153024911e-05,
+      "loss": 1.8249,
+      "step": 6491
+    },
+    {
+      "epoch": 2.8853333333333335,
+      "grad_norm": 3.3353047370910645,
+      "learning_rate": 8.466192170818506e-05,
+      "loss": 1.3361,
+      "step": 6492
+    },
+    {
+      "epoch": 2.8857777777777778,
+      "grad_norm": 2.4956469535827637,
+      "learning_rate": 8.4644128113879e-05,
+      "loss": 1.3096,
+      "step": 6493
+    },
+    {
+      "epoch": 2.886222222222222,
+      "grad_norm": 3.014362096786499,
+      "learning_rate": 8.462633451957296e-05,
+      "loss": 1.5515,
+      "step": 6494
+    },
+    {
+      "epoch": 2.8866666666666667,
+      "grad_norm": 2.6748619079589844,
+      "learning_rate": 8.460854092526691e-05,
+      "loss": 1.5532,
+      "step": 6495
+    },
+    {
+      "epoch": 2.887111111111111,
+      "grad_norm": 3.3072657585144043,
+      "learning_rate": 8.459074733096086e-05,
+      "loss": 1.744,
+      "step": 6496
+    },
+    {
+      "epoch": 2.8875555555555557,
+      "grad_norm": 3.1664316654205322,
+      "learning_rate": 8.45729537366548e-05,
+      "loss": 1.5507,
+      "step": 6497
+    },
+    {
+      "epoch": 2.888,
+      "grad_norm": 3.5255908966064453,
+      "learning_rate": 8.455516014234876e-05,
+      "loss": 2.1625,
+      "step": 6498
+    },
+    {
+      "epoch": 2.8884444444444446,
+      "grad_norm": 0.49165022373199463,
+      "learning_rate": 8.453736654804271e-05,
+      "loss": 0.0612,
+      "step": 6499
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 4.066702365875244,
+      "learning_rate": 8.451957295373666e-05,
+      "loss": 1.5139,
+      "step": 6500
+    },
+    {
+      "epoch": 2.889333333333333,
+      "grad_norm": 1.581194519996643,
+      "learning_rate": 8.450177935943061e-05,
+      "loss": 1.1063,
+      "step": 6501
+    },
+    {
+      "epoch": 2.889777777777778,
+      "grad_norm": 1.7735909223556519,
+      "learning_rate": 8.448398576512457e-05,
+      "loss": 1.9774,
+      "step": 6502
+    },
+    {
+      "epoch": 2.8902222222222225,
+      "grad_norm": 1.8580666780471802,
+      "learning_rate": 8.446619217081851e-05,
+      "loss": 2.2711,
+      "step": 6503
+    },
+    {
+      "epoch": 2.8906666666666667,
+      "grad_norm": 2.2903764247894287,
+      "learning_rate": 8.444839857651245e-05,
+      "loss": 2.2313,
+      "step": 6504
+    },
+    {
+      "epoch": 2.891111111111111,
+      "grad_norm": 2.7461414337158203,
+      "learning_rate": 8.443060498220641e-05,
+      "loss": 2.6372,
+      "step": 6505
+    },
+    {
+      "epoch": 2.8915555555555557,
+      "grad_norm": 2.5035669803619385,
+      "learning_rate": 8.441281138790035e-05,
+      "loss": 1.8447,
+      "step": 6506
+    },
+    {
+      "epoch": 2.892,
+      "grad_norm": 2.2642126083374023,
+      "learning_rate": 8.439501779359431e-05,
+      "loss": 2.3071,
+      "step": 6507
+    },
+    {
+      "epoch": 2.8924444444444446,
+      "grad_norm": 2.4640629291534424,
+      "learning_rate": 8.437722419928827e-05,
+      "loss": 2.0594,
+      "step": 6508
+    },
+    {
+      "epoch": 2.892888888888889,
+      "grad_norm": 2.5078279972076416,
+      "learning_rate": 8.435943060498221e-05,
+      "loss": 2.0226,
+      "step": 6509
+    },
+    {
+      "epoch": 2.8933333333333335,
+      "grad_norm": 2.1550204753875732,
+      "learning_rate": 8.434163701067615e-05,
+      "loss": 1.803,
+      "step": 6510
+    },
+    {
+      "epoch": 2.893777777777778,
+      "grad_norm": 2.478727340698242,
+      "learning_rate": 8.432384341637011e-05,
+      "loss": 1.9638,
+      "step": 6511
+    },
+    {
+      "epoch": 2.894222222222222,
+      "grad_norm": 2.432596206665039,
+      "learning_rate": 8.430604982206407e-05,
+      "loss": 1.8639,
+      "step": 6512
+    },
+    {
+      "epoch": 2.8946666666666667,
+      "grad_norm": 2.1648998260498047,
+      "learning_rate": 8.428825622775801e-05,
+      "loss": 1.8943,
+      "step": 6513
+    },
+    {
+      "epoch": 2.895111111111111,
+      "grad_norm": 2.2655625343322754,
+      "learning_rate": 8.427046263345197e-05,
+      "loss": 1.4152,
+      "step": 6514
+    },
+    {
+      "epoch": 2.8955555555555557,
+      "grad_norm": 2.6261086463928223,
+      "learning_rate": 8.425266903914592e-05,
+      "loss": 1.9381,
+      "step": 6515
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 2.0679445266723633,
+      "learning_rate": 8.423487544483986e-05,
+      "loss": 1.5273,
+      "step": 6516
+    },
+    {
+      "epoch": 2.8964444444444446,
+      "grad_norm": 2.227121591567993,
+      "learning_rate": 8.421708185053381e-05,
+      "loss": 1.5115,
+      "step": 6517
+    },
+    {
+      "epoch": 2.896888888888889,
+      "grad_norm": 2.8133530616760254,
+      "learning_rate": 8.419928825622776e-05,
+      "loss": 2.1005,
+      "step": 6518
+    },
+    {
+      "epoch": 2.897333333333333,
+      "grad_norm": 2.5659241676330566,
+      "learning_rate": 8.418149466192171e-05,
+      "loss": 1.6425,
+      "step": 6519
+    },
+    {
+      "epoch": 2.897777777777778,
+      "grad_norm": 3.140711784362793,
+      "learning_rate": 8.416370106761566e-05,
+      "loss": 1.7234,
+      "step": 6520
+    },
+    {
+      "epoch": 2.8982222222222225,
+      "grad_norm": 2.281583786010742,
+      "learning_rate": 8.414590747330962e-05,
+      "loss": 1.4164,
+      "step": 6521
+    },
+    {
+      "epoch": 2.8986666666666667,
+      "grad_norm": 2.2396039962768555,
+      "learning_rate": 8.412811387900356e-05,
+      "loss": 1.6061,
+      "step": 6522
+    },
+    {
+      "epoch": 2.899111111111111,
+      "grad_norm": 2.5794193744659424,
+      "learning_rate": 8.41103202846975e-05,
+      "loss": 1.8131,
+      "step": 6523
+    },
+    {
+      "epoch": 2.8995555555555557,
+      "grad_norm": 2.583009958267212,
+      "learning_rate": 8.409252669039146e-05,
+      "loss": 1.6304,
+      "step": 6524
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.1925716400146484,
+      "learning_rate": 8.40747330960854e-05,
+      "loss": 1.4056,
+      "step": 6525
+    },
+    {
+      "epoch": 2.9004444444444446,
+      "grad_norm": 2.4621171951293945,
+      "learning_rate": 8.405693950177936e-05,
+      "loss": 1.5926,
+      "step": 6526
+    },
+    {
+      "epoch": 2.900888888888889,
+      "grad_norm": 2.397620439529419,
+      "learning_rate": 8.403914590747332e-05,
+      "loss": 1.5116,
+      "step": 6527
+    },
+    {
+      "epoch": 2.9013333333333335,
+      "grad_norm": 2.841352939605713,
+      "learning_rate": 8.402135231316726e-05,
+      "loss": 1.5999,
+      "step": 6528
+    },
+    {
+      "epoch": 2.901777777777778,
+      "grad_norm": 3.072946548461914,
+      "learning_rate": 8.400355871886122e-05,
+      "loss": 2.002,
+      "step": 6529
+    },
+    {
+      "epoch": 2.902222222222222,
+      "grad_norm": 2.5697319507598877,
+      "learning_rate": 8.398576512455516e-05,
+      "loss": 1.8486,
+      "step": 6530
+    },
+    {
+      "epoch": 2.9026666666666667,
+      "grad_norm": 3.085151195526123,
+      "learning_rate": 8.396797153024912e-05,
+      "loss": 1.6757,
+      "step": 6531
+    },
+    {
+      "epoch": 2.903111111111111,
+      "grad_norm": 2.6995904445648193,
+      "learning_rate": 8.395017793594306e-05,
+      "loss": 1.4552,
+      "step": 6532
+    },
+    {
+      "epoch": 2.9035555555555557,
+      "grad_norm": 3.3730642795562744,
+      "learning_rate": 8.393238434163702e-05,
+      "loss": 1.8132,
+      "step": 6533
+    },
+    {
+      "epoch": 2.904,
+      "grad_norm": 2.572873115539551,
+      "learning_rate": 8.391459074733097e-05,
+      "loss": 1.8587,
+      "step": 6534
+    },
+    {
+      "epoch": 2.9044444444444446,
+      "grad_norm": 3.1449742317199707,
+      "learning_rate": 8.389679715302492e-05,
+      "loss": 1.3689,
+      "step": 6535
+    },
+    {
+      "epoch": 2.904888888888889,
+      "grad_norm": 2.7788970470428467,
+      "learning_rate": 8.387900355871886e-05,
+      "loss": 1.4992,
+      "step": 6536
+    },
+    {
+      "epoch": 2.905333333333333,
+      "grad_norm": 2.823796510696411,
+      "learning_rate": 8.386120996441282e-05,
+      "loss": 1.9288,
+      "step": 6537
+    },
+    {
+      "epoch": 2.905777777777778,
+      "grad_norm": 3.0071866512298584,
+      "learning_rate": 8.384341637010676e-05,
+      "loss": 2.0833,
+      "step": 6538
+    },
+    {
+      "epoch": 2.9062222222222225,
+      "grad_norm": 2.019472360610962,
+      "learning_rate": 8.382562277580071e-05,
+      "loss": 0.9878,
+      "step": 6539
+    },
+    {
+      "epoch": 2.9066666666666667,
+      "grad_norm": 2.4632411003112793,
+      "learning_rate": 8.380782918149467e-05,
+      "loss": 1.6176,
+      "step": 6540
+    },
+    {
+      "epoch": 2.907111111111111,
+      "grad_norm": 3.062143564224243,
+      "learning_rate": 8.379003558718861e-05,
+      "loss": 2.1748,
+      "step": 6541
+    },
+    {
+      "epoch": 2.9075555555555557,
+      "grad_norm": 4.060853958129883,
+      "learning_rate": 8.377224199288256e-05,
+      "loss": 1.9903,
+      "step": 6542
+    },
+    {
+      "epoch": 2.908,
+      "grad_norm": 3.1725289821624756,
+      "learning_rate": 8.375444839857651e-05,
+      "loss": 1.8678,
+      "step": 6543
+    },
+    {
+      "epoch": 2.9084444444444446,
+      "grad_norm": 3.047879934310913,
+      "learning_rate": 8.373665480427047e-05,
+      "loss": 1.6334,
+      "step": 6544
+    },
+    {
+      "epoch": 2.908888888888889,
+      "grad_norm": 3.0653295516967773,
+      "learning_rate": 8.371886120996441e-05,
+      "loss": 1.8163,
+      "step": 6545
+    },
+    {
+      "epoch": 2.9093333333333335,
+      "grad_norm": 3.4780588150024414,
+      "learning_rate": 8.370106761565837e-05,
+      "loss": 1.6996,
+      "step": 6546
+    },
+    {
+      "epoch": 2.909777777777778,
+      "grad_norm": 3.106790781021118,
+      "learning_rate": 8.368327402135233e-05,
+      "loss": 1.5692,
+      "step": 6547
+    },
+    {
+      "epoch": 2.910222222222222,
+      "grad_norm": 3.5817697048187256,
+      "learning_rate": 8.366548042704627e-05,
+      "loss": 2.3844,
+      "step": 6548
+    },
+    {
+      "epoch": 2.9106666666666667,
+      "grad_norm": 3.6129720211029053,
+      "learning_rate": 8.364768683274021e-05,
+      "loss": 1.8793,
+      "step": 6549
+    },
+    {
+      "epoch": 2.911111111111111,
+      "grad_norm": 3.0888235569000244,
+      "learning_rate": 8.362989323843417e-05,
+      "loss": 0.7461,
+      "step": 6550
+    },
+    {
+      "epoch": 2.9115555555555557,
+      "grad_norm": 1.7484018802642822,
+      "learning_rate": 8.361209964412811e-05,
+      "loss": 1.9615,
+      "step": 6551
+    },
+    {
+      "epoch": 2.912,
+      "grad_norm": 2.0905039310455322,
+      "learning_rate": 8.359430604982207e-05,
+      "loss": 2.2685,
+      "step": 6552
+    },
+    {
+      "epoch": 2.9124444444444446,
+      "grad_norm": 2.227344512939453,
+      "learning_rate": 8.357651245551602e-05,
+      "loss": 1.8803,
+      "step": 6553
+    },
+    {
+      "epoch": 2.912888888888889,
+      "grad_norm": 2.469000816345215,
+      "learning_rate": 8.355871886120997e-05,
+      "loss": 2.3178,
+      "step": 6554
+    },
+    {
+      "epoch": 2.913333333333333,
+      "grad_norm": 2.261646270751953,
+      "learning_rate": 8.354092526690391e-05,
+      "loss": 2.118,
+      "step": 6555
+    },
+    {
+      "epoch": 2.913777777777778,
+      "grad_norm": 2.2827672958374023,
+      "learning_rate": 8.352313167259787e-05,
+      "loss": 1.5105,
+      "step": 6556
+    },
+    {
+      "epoch": 2.9142222222222225,
+      "grad_norm": 2.4409070014953613,
+      "learning_rate": 8.350533807829182e-05,
+      "loss": 2.1229,
+      "step": 6557
+    },
+    {
+      "epoch": 2.9146666666666667,
+      "grad_norm": 1.7336750030517578,
+      "learning_rate": 8.348754448398577e-05,
+      "loss": 0.9735,
+      "step": 6558
+    },
+    {
+      "epoch": 2.915111111111111,
+      "grad_norm": 2.462920904159546,
+      "learning_rate": 8.346975088967972e-05,
+      "loss": 1.6474,
+      "step": 6559
+    },
+    {
+      "epoch": 2.9155555555555557,
+      "grad_norm": 2.5200788974761963,
+      "learning_rate": 8.345195729537368e-05,
+      "loss": 1.8454,
+      "step": 6560
+    },
+    {
+      "epoch": 2.916,
+      "grad_norm": 2.2653720378875732,
+      "learning_rate": 8.343416370106762e-05,
+      "loss": 1.7404,
+      "step": 6561
+    },
+    {
+      "epoch": 2.916444444444444,
+      "grad_norm": 2.323324680328369,
+      "learning_rate": 8.341637010676157e-05,
+      "loss": 1.7726,
+      "step": 6562
+    },
+    {
+      "epoch": 2.916888888888889,
+      "grad_norm": 2.575195074081421,
+      "learning_rate": 8.339857651245552e-05,
+      "loss": 2.0535,
+      "step": 6563
+    },
+    {
+      "epoch": 2.9173333333333336,
+      "grad_norm": 2.4426960945129395,
+      "learning_rate": 8.338078291814946e-05,
+      "loss": 1.8601,
+      "step": 6564
+    },
+    {
+      "epoch": 2.917777777777778,
+      "grad_norm": 2.3300108909606934,
+      "learning_rate": 8.336298932384342e-05,
+      "loss": 1.7842,
+      "step": 6565
+    },
+    {
+      "epoch": 2.918222222222222,
+      "grad_norm": 2.5958733558654785,
+      "learning_rate": 8.334519572953738e-05,
+      "loss": 1.9596,
+      "step": 6566
+    },
+    {
+      "epoch": 2.9186666666666667,
+      "grad_norm": 2.2606828212738037,
+      "learning_rate": 8.332740213523132e-05,
+      "loss": 1.2543,
+      "step": 6567
+    },
+    {
+      "epoch": 2.919111111111111,
+      "grad_norm": 2.573596954345703,
+      "learning_rate": 8.330960854092526e-05,
+      "loss": 1.6507,
+      "step": 6568
+    },
+    {
+      "epoch": 2.9195555555555557,
+      "grad_norm": 1.9967552423477173,
+      "learning_rate": 8.329181494661922e-05,
+      "loss": 1.3919,
+      "step": 6569
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.615650177001953,
+      "learning_rate": 8.327402135231316e-05,
+      "loss": 1.8423,
+      "step": 6570
+    },
+    {
+      "epoch": 2.9204444444444446,
+      "grad_norm": 2.7236530780792236,
+      "learning_rate": 8.325622775800712e-05,
+      "loss": 2.3231,
+      "step": 6571
+    },
+    {
+      "epoch": 2.920888888888889,
+      "grad_norm": 2.5713248252868652,
+      "learning_rate": 8.323843416370108e-05,
+      "loss": 1.869,
+      "step": 6572
+    },
+    {
+      "epoch": 2.921333333333333,
+      "grad_norm": 2.9785003662109375,
+      "learning_rate": 8.322064056939502e-05,
+      "loss": 2.2974,
+      "step": 6573
+    },
+    {
+      "epoch": 2.921777777777778,
+      "grad_norm": 2.4110934734344482,
+      "learning_rate": 8.320284697508898e-05,
+      "loss": 1.2225,
+      "step": 6574
+    },
+    {
+      "epoch": 2.9222222222222225,
+      "grad_norm": 2.372670888900757,
+      "learning_rate": 8.318505338078292e-05,
+      "loss": 1.2217,
+      "step": 6575
+    },
+    {
+      "epoch": 2.9226666666666667,
+      "grad_norm": 2.6958694458007812,
+      "learning_rate": 8.316725978647687e-05,
+      "loss": 1.913,
+      "step": 6576
+    },
+    {
+      "epoch": 2.923111111111111,
+      "grad_norm": 2.6148412227630615,
+      "learning_rate": 8.314946619217082e-05,
+      "loss": 1.4505,
+      "step": 6577
+    },
+    {
+      "epoch": 2.9235555555555557,
+      "grad_norm": 2.7632627487182617,
+      "learning_rate": 8.313167259786477e-05,
+      "loss": 2.0089,
+      "step": 6578
+    },
+    {
+      "epoch": 2.924,
+      "grad_norm": 2.448899984359741,
+      "learning_rate": 8.311387900355873e-05,
+      "loss": 1.7222,
+      "step": 6579
+    },
+    {
+      "epoch": 2.924444444444444,
+      "grad_norm": 1.7573479413986206,
+      "learning_rate": 8.309608540925267e-05,
+      "loss": 0.8537,
+      "step": 6580
+    },
+    {
+      "epoch": 2.924888888888889,
+      "grad_norm": 2.8462624549865723,
+      "learning_rate": 8.307829181494662e-05,
+      "loss": 2.0423,
+      "step": 6581
+    },
+    {
+      "epoch": 2.9253333333333336,
+      "grad_norm": 1.8303264379501343,
+      "learning_rate": 8.306049822064057e-05,
+      "loss": 0.9714,
+      "step": 6582
+    },
+    {
+      "epoch": 2.925777777777778,
+      "grad_norm": 2.0662248134613037,
+      "learning_rate": 8.304270462633452e-05,
+      "loss": 0.9561,
+      "step": 6583
+    },
+    {
+      "epoch": 2.926222222222222,
+      "grad_norm": 2.5108532905578613,
+      "learning_rate": 8.302491103202847e-05,
+      "loss": 1.4271,
+      "step": 6584
+    },
+    {
+      "epoch": 2.9266666666666667,
+      "grad_norm": 2.5698022842407227,
+      "learning_rate": 8.300711743772243e-05,
+      "loss": 1.4,
+      "step": 6585
+    },
+    {
+      "epoch": 2.927111111111111,
+      "grad_norm": 2.503103494644165,
+      "learning_rate": 8.298932384341637e-05,
+      "loss": 1.2983,
+      "step": 6586
+    },
+    {
+      "epoch": 2.9275555555555557,
+      "grad_norm": 2.902477264404297,
+      "learning_rate": 8.297153024911033e-05,
+      "loss": 1.7343,
+      "step": 6587
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 2.991421937942505,
+      "learning_rate": 8.295373665480427e-05,
+      "loss": 1.6752,
+      "step": 6588
+    },
+    {
+      "epoch": 2.9284444444444446,
+      "grad_norm": 2.4898111820220947,
+      "learning_rate": 8.293594306049823e-05,
+      "loss": 1.7385,
+      "step": 6589
+    },
+    {
+      "epoch": 2.928888888888889,
+      "grad_norm": 3.5942859649658203,
+      "learning_rate": 8.291814946619217e-05,
+      "loss": 2.0857,
+      "step": 6590
+    },
+    {
+      "epoch": 2.929333333333333,
+      "grad_norm": 1.9334756135940552,
+      "learning_rate": 8.290035587188613e-05,
+      "loss": 0.77,
+      "step": 6591
+    },
+    {
+      "epoch": 2.929777777777778,
+      "grad_norm": 3.102285385131836,
+      "learning_rate": 8.288256227758008e-05,
+      "loss": 1.8683,
+      "step": 6592
+    },
+    {
+      "epoch": 2.930222222222222,
+      "grad_norm": 2.8912463188171387,
+      "learning_rate": 8.286476868327403e-05,
+      "loss": 1.7117,
+      "step": 6593
+    },
+    {
+      "epoch": 2.9306666666666668,
+      "grad_norm": 3.238525152206421,
+      "learning_rate": 8.284697508896797e-05,
+      "loss": 2.01,
+      "step": 6594
+    },
+    {
+      "epoch": 2.931111111111111,
+      "grad_norm": 2.9973912239074707,
+      "learning_rate": 8.282918149466193e-05,
+      "loss": 1.7764,
+      "step": 6595
+    },
+    {
+      "epoch": 2.9315555555555557,
+      "grad_norm": 3.1456546783447266,
+      "learning_rate": 8.281138790035587e-05,
+      "loss": 2.0722,
+      "step": 6596
+    },
+    {
+      "epoch": 2.932,
+      "grad_norm": 2.999476671218872,
+      "learning_rate": 8.279359430604983e-05,
+      "loss": 1.7548,
+      "step": 6597
+    },
+    {
+      "epoch": 2.932444444444444,
+      "grad_norm": 3.8145954608917236,
+      "learning_rate": 8.277580071174378e-05,
+      "loss": 1.6423,
+      "step": 6598
+    },
+    {
+      "epoch": 2.932888888888889,
+      "grad_norm": 5.341122150421143,
+      "learning_rate": 8.275800711743773e-05,
+      "loss": 1.1883,
+      "step": 6599
+    },
+    {
+      "epoch": 2.9333333333333336,
+      "grad_norm": 3.1695616245269775,
+      "learning_rate": 8.274021352313167e-05,
+      "loss": 0.7776,
+      "step": 6600
+    },
+    {
+      "epoch": 2.933777777777778,
+      "grad_norm": 2.0281214714050293,
+      "learning_rate": 8.272241992882562e-05,
+      "loss": 1.7799,
+      "step": 6601
+    },
+    {
+      "epoch": 2.934222222222222,
+      "grad_norm": 2.2131547927856445,
+      "learning_rate": 8.270462633451958e-05,
+      "loss": 1.9286,
+      "step": 6602
+    },
+    {
+      "epoch": 2.9346666666666668,
+      "grad_norm": 2.3528523445129395,
+      "learning_rate": 8.268683274021352e-05,
+      "loss": 1.5521,
+      "step": 6603
+    },
+    {
+      "epoch": 2.935111111111111,
+      "grad_norm": 2.04451584815979,
+      "learning_rate": 8.266903914590748e-05,
+      "loss": 1.7594,
+      "step": 6604
+    },
+    {
+      "epoch": 2.9355555555555557,
+      "grad_norm": 0.19427204132080078,
+      "learning_rate": 8.265124555160144e-05,
+      "loss": 0.0238,
+      "step": 6605
+    },
+    {
+      "epoch": 2.936,
+      "grad_norm": 2.289008855819702,
+      "learning_rate": 8.263345195729538e-05,
+      "loss": 1.6694,
+      "step": 6606
+    },
+    {
+      "epoch": 2.9364444444444446,
+      "grad_norm": 3.0516107082366943,
+      "learning_rate": 8.261565836298932e-05,
+      "loss": 1.9453,
+      "step": 6607
+    },
+    {
+      "epoch": 2.936888888888889,
+      "grad_norm": 2.20947265625,
+      "learning_rate": 8.259786476868328e-05,
+      "loss": 1.7946,
+      "step": 6608
+    },
+    {
+      "epoch": 2.937333333333333,
+      "grad_norm": 2.571842670440674,
+      "learning_rate": 8.258007117437722e-05,
+      "loss": 1.6298,
+      "step": 6609
+    },
+    {
+      "epoch": 2.937777777777778,
+      "grad_norm": 2.839123249053955,
+      "learning_rate": 8.256227758007118e-05,
+      "loss": 1.9534,
+      "step": 6610
+    },
+    {
+      "epoch": 2.938222222222222,
+      "grad_norm": 2.8716540336608887,
+      "learning_rate": 8.254448398576514e-05,
+      "loss": 1.8971,
+      "step": 6611
+    },
+    {
+      "epoch": 2.9386666666666668,
+      "grad_norm": 2.287142276763916,
+      "learning_rate": 8.252669039145908e-05,
+      "loss": 1.73,
+      "step": 6612
+    },
+    {
+      "epoch": 2.939111111111111,
+      "grad_norm": 2.8752803802490234,
+      "learning_rate": 8.250889679715302e-05,
+      "loss": 1.7428,
+      "step": 6613
+    },
+    {
+      "epoch": 2.9395555555555557,
+      "grad_norm": 2.5958893299102783,
+      "learning_rate": 8.249110320284698e-05,
+      "loss": 1.9177,
+      "step": 6614
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 2.100194215774536,
+      "learning_rate": 8.247330960854092e-05,
+      "loss": 1.4296,
+      "step": 6615
+    },
+    {
+      "epoch": 2.940444444444444,
+      "grad_norm": 2.6018195152282715,
+      "learning_rate": 8.245551601423488e-05,
+      "loss": 1.7576,
+      "step": 6616
+    },
+    {
+      "epoch": 2.940888888888889,
+      "grad_norm": 2.3454983234405518,
+      "learning_rate": 8.243772241992883e-05,
+      "loss": 1.5927,
+      "step": 6617
+    },
+    {
+      "epoch": 2.9413333333333336,
+      "grad_norm": 2.5776448249816895,
+      "learning_rate": 8.241992882562278e-05,
+      "loss": 1.7602,
+      "step": 6618
+    },
+    {
+      "epoch": 2.941777777777778,
+      "grad_norm": 2.55011248588562,
+      "learning_rate": 8.240213523131673e-05,
+      "loss": 1.4901,
+      "step": 6619
+    },
+    {
+      "epoch": 2.942222222222222,
+      "grad_norm": 2.298241376876831,
+      "learning_rate": 8.238434163701068e-05,
+      "loss": 1.5664,
+      "step": 6620
+    },
+    {
+      "epoch": 2.9426666666666668,
+      "grad_norm": 2.403535842895508,
+      "learning_rate": 8.236654804270463e-05,
+      "loss": 1.7368,
+      "step": 6621
+    },
+    {
+      "epoch": 2.943111111111111,
+      "grad_norm": 2.3302366733551025,
+      "learning_rate": 8.234875444839858e-05,
+      "loss": 1.5256,
+      "step": 6622
+    },
+    {
+      "epoch": 2.9435555555555557,
+      "grad_norm": 2.4017152786254883,
+      "learning_rate": 8.233096085409253e-05,
+      "loss": 1.9532,
+      "step": 6623
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 1.9871573448181152,
+      "learning_rate": 8.231316725978649e-05,
+      "loss": 1.2034,
+      "step": 6624
+    },
+    {
+      "epoch": 2.9444444444444446,
+      "grad_norm": 2.877697229385376,
+      "learning_rate": 8.229537366548043e-05,
+      "loss": 1.7799,
+      "step": 6625
+    },
+    {
+      "epoch": 2.944888888888889,
+      "grad_norm": 2.7882204055786133,
+      "learning_rate": 8.227758007117437e-05,
+      "loss": 1.8585,
+      "step": 6626
+    },
+    {
+      "epoch": 2.945333333333333,
+      "grad_norm": 3.066232442855835,
+      "learning_rate": 8.225978647686833e-05,
+      "loss": 1.9722,
+      "step": 6627
+    },
+    {
+      "epoch": 2.945777777777778,
+      "grad_norm": 2.9001948833465576,
+      "learning_rate": 8.224199288256227e-05,
+      "loss": 2.0303,
+      "step": 6628
+    },
+    {
+      "epoch": 2.946222222222222,
+      "grad_norm": 3.1734731197357178,
+      "learning_rate": 8.222419928825623e-05,
+      "loss": 1.8917,
+      "step": 6629
+    },
+    {
+      "epoch": 2.9466666666666668,
+      "grad_norm": 1.936790108680725,
+      "learning_rate": 8.220640569395019e-05,
+      "loss": 0.9073,
+      "step": 6630
+    },
+    {
+      "epoch": 2.947111111111111,
+      "grad_norm": 2.223129987716675,
+      "learning_rate": 8.218861209964413e-05,
+      "loss": 0.9589,
+      "step": 6631
+    },
+    {
+      "epoch": 2.9475555555555557,
+      "grad_norm": 1.784337043762207,
+      "learning_rate": 8.217081850533809e-05,
+      "loss": 0.7341,
+      "step": 6632
+    },
+    {
+      "epoch": 2.948,
+      "grad_norm": 2.703338146209717,
+      "learning_rate": 8.215302491103203e-05,
+      "loss": 1.5117,
+      "step": 6633
+    },
+    {
+      "epoch": 2.948444444444444,
+      "grad_norm": 2.924023389816284,
+      "learning_rate": 8.213523131672599e-05,
+      "loss": 1.8323,
+      "step": 6634
+    },
+    {
+      "epoch": 2.948888888888889,
+      "grad_norm": 3.084257125854492,
+      "learning_rate": 8.211743772241993e-05,
+      "loss": 1.7585,
+      "step": 6635
+    },
+    {
+      "epoch": 2.9493333333333336,
+      "grad_norm": 2.6331820487976074,
+      "learning_rate": 8.209964412811389e-05,
+      "loss": 1.5509,
+      "step": 6636
+    },
+    {
+      "epoch": 2.949777777777778,
+      "grad_norm": 2.6916754245758057,
+      "learning_rate": 8.208185053380784e-05,
+      "loss": 1.6519,
+      "step": 6637
+    },
+    {
+      "epoch": 2.950222222222222,
+      "grad_norm": 3.28885555267334,
+      "learning_rate": 8.206405693950178e-05,
+      "loss": 1.9601,
+      "step": 6638
+    },
+    {
+      "epoch": 2.9506666666666668,
+      "grad_norm": 3.135986328125,
+      "learning_rate": 8.204626334519573e-05,
+      "loss": 1.9278,
+      "step": 6639
+    },
+    {
+      "epoch": 2.951111111111111,
+      "grad_norm": 2.845036745071411,
+      "learning_rate": 8.202846975088968e-05,
+      "loss": 1.611,
+      "step": 6640
+    },
+    {
+      "epoch": 2.9515555555555557,
+      "grad_norm": 2.581521987915039,
+      "learning_rate": 8.201067615658363e-05,
+      "loss": 1.4946,
+      "step": 6641
+    },
+    {
+      "epoch": 2.952,
+      "grad_norm": 2.706934928894043,
+      "learning_rate": 8.199288256227758e-05,
+      "loss": 1.6828,
+      "step": 6642
+    },
+    {
+      "epoch": 2.9524444444444446,
+      "grad_norm": 3.34147572517395,
+      "learning_rate": 8.197508896797154e-05,
+      "loss": 2.12,
+      "step": 6643
+    },
+    {
+      "epoch": 2.952888888888889,
+      "grad_norm": 3.1140594482421875,
+      "learning_rate": 8.195729537366548e-05,
+      "loss": 2.0541,
+      "step": 6644
+    },
+    {
+      "epoch": 2.953333333333333,
+      "grad_norm": 3.713144540786743,
+      "learning_rate": 8.193950177935944e-05,
+      "loss": 2.1426,
+      "step": 6645
+    },
+    {
+      "epoch": 2.953777777777778,
+      "grad_norm": 3.481236219406128,
+      "learning_rate": 8.192170818505338e-05,
+      "loss": 1.4833,
+      "step": 6646
+    },
+    {
+      "epoch": 2.954222222222222,
+      "grad_norm": 3.1434216499328613,
+      "learning_rate": 8.190391459074734e-05,
+      "loss": 1.8463,
+      "step": 6647
+    },
+    {
+      "epoch": 2.9546666666666668,
+      "grad_norm": 3.119110107421875,
+      "learning_rate": 8.188612099644128e-05,
+      "loss": 1.4728,
+      "step": 6648
+    },
+    {
+      "epoch": 2.955111111111111,
+      "grad_norm": 0.43881502747535706,
+      "learning_rate": 8.186832740213524e-05,
+      "loss": 0.0622,
+      "step": 6649
+    },
+    {
+      "epoch": 2.9555555555555557,
+      "grad_norm": 3.7765581607818604,
+      "learning_rate": 8.18505338078292e-05,
+      "loss": 1.5129,
+      "step": 6650
+    },
+    {
+      "epoch": 2.956,
+      "grad_norm": 2.1686058044433594,
+      "learning_rate": 8.183274021352314e-05,
+      "loss": 2.1518,
+      "step": 6651
+    },
+    {
+      "epoch": 2.956444444444444,
+      "grad_norm": 2.1492674350738525,
+      "learning_rate": 8.181494661921708e-05,
+      "loss": 1.6543,
+      "step": 6652
+    },
+    {
+      "epoch": 2.956888888888889,
+      "grad_norm": 2.0771119594573975,
+      "learning_rate": 8.179715302491104e-05,
+      "loss": 2.1437,
+      "step": 6653
+    },
+    {
+      "epoch": 2.9573333333333336,
+      "grad_norm": 2.154829978942871,
+      "learning_rate": 8.177935943060498e-05,
+      "loss": 2.0731,
+      "step": 6654
+    },
+    {
+      "epoch": 2.957777777777778,
+      "grad_norm": 2.0446035861968994,
+      "learning_rate": 8.176156583629894e-05,
+      "loss": 2.1788,
+      "step": 6655
+    },
+    {
+      "epoch": 2.958222222222222,
+      "grad_norm": 2.4353065490722656,
+      "learning_rate": 8.174377224199289e-05,
+      "loss": 1.9073,
+      "step": 6656
+    },
+    {
+      "epoch": 2.958666666666667,
+      "grad_norm": 2.2374634742736816,
+      "learning_rate": 8.172597864768684e-05,
+      "loss": 1.9468,
+      "step": 6657
+    },
+    {
+      "epoch": 2.959111111111111,
+      "grad_norm": 1.9991698265075684,
+      "learning_rate": 8.170818505338078e-05,
+      "loss": 2.0529,
+      "step": 6658
+    },
+    {
+      "epoch": 2.9595555555555557,
+      "grad_norm": 2.3634281158447266,
+      "learning_rate": 8.169039145907474e-05,
+      "loss": 2.2038,
+      "step": 6659
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.3970882892608643,
+      "learning_rate": 8.167259786476868e-05,
+      "loss": 2.1983,
+      "step": 6660
+    },
+    {
+      "epoch": 2.9604444444444447,
+      "grad_norm": 2.5298354625701904,
+      "learning_rate": 8.165480427046264e-05,
+      "loss": 2.1612,
+      "step": 6661
+    },
+    {
+      "epoch": 2.960888888888889,
+      "grad_norm": 2.4006104469299316,
+      "learning_rate": 8.163701067615659e-05,
+      "loss": 2.0669,
+      "step": 6662
+    },
+    {
+      "epoch": 2.961333333333333,
+      "grad_norm": 2.5116260051727295,
+      "learning_rate": 8.161921708185053e-05,
+      "loss": 1.9285,
+      "step": 6663
+    },
+    {
+      "epoch": 2.961777777777778,
+      "grad_norm": 2.3565287590026855,
+      "learning_rate": 8.160142348754449e-05,
+      "loss": 1.9953,
+      "step": 6664
+    },
+    {
+      "epoch": 2.962222222222222,
+      "grad_norm": 2.253330945968628,
+      "learning_rate": 8.158362989323843e-05,
+      "loss": 1.5129,
+      "step": 6665
+    },
+    {
+      "epoch": 2.962666666666667,
+      "grad_norm": 2.49210524559021,
+      "learning_rate": 8.156583629893239e-05,
+      "loss": 1.9875,
+      "step": 6666
+    },
+    {
+      "epoch": 2.963111111111111,
+      "grad_norm": 2.353780508041382,
+      "learning_rate": 8.154804270462633e-05,
+      "loss": 1.471,
+      "step": 6667
+    },
+    {
+      "epoch": 2.9635555555555557,
+      "grad_norm": 2.679733991622925,
+      "learning_rate": 8.153024911032029e-05,
+      "loss": 1.8062,
+      "step": 6668
+    },
+    {
+      "epoch": 2.964,
+      "grad_norm": 2.1088638305664062,
+      "learning_rate": 8.151245551601425e-05,
+      "loss": 1.6653,
+      "step": 6669
+    },
+    {
+      "epoch": 2.964444444444444,
+      "grad_norm": 2.271273612976074,
+      "learning_rate": 8.149466192170819e-05,
+      "loss": 1.8267,
+      "step": 6670
+    },
+    {
+      "epoch": 2.964888888888889,
+      "grad_norm": 2.468289613723755,
+      "learning_rate": 8.147686832740213e-05,
+      "loss": 1.5659,
+      "step": 6671
+    },
+    {
+      "epoch": 2.9653333333333336,
+      "grad_norm": 2.2520124912261963,
+      "learning_rate": 8.145907473309609e-05,
+      "loss": 1.6409,
+      "step": 6672
+    },
+    {
+      "epoch": 2.965777777777778,
+      "grad_norm": 2.9224207401275635,
+      "learning_rate": 8.144128113879003e-05,
+      "loss": 1.7884,
+      "step": 6673
+    },
+    {
+      "epoch": 2.966222222222222,
+      "grad_norm": 2.306745767593384,
+      "learning_rate": 8.142348754448399e-05,
+      "loss": 1.8612,
+      "step": 6674
+    },
+    {
+      "epoch": 2.966666666666667,
+      "grad_norm": 2.707267999649048,
+      "learning_rate": 8.140569395017794e-05,
+      "loss": 1.6955,
+      "step": 6675
+    },
+    {
+      "epoch": 2.967111111111111,
+      "grad_norm": 2.6562561988830566,
+      "learning_rate": 8.138790035587189e-05,
+      "loss": 1.8765,
+      "step": 6676
+    },
+    {
+      "epoch": 2.9675555555555553,
+      "grad_norm": 2.344191789627075,
+      "learning_rate": 8.137010676156584e-05,
+      "loss": 1.3519,
+      "step": 6677
+    },
+    {
+      "epoch": 2.968,
+      "grad_norm": 2.481653928756714,
+      "learning_rate": 8.135231316725979e-05,
+      "loss": 1.8528,
+      "step": 6678
+    },
+    {
+      "epoch": 2.9684444444444447,
+      "grad_norm": 2.491582155227661,
+      "learning_rate": 8.133451957295374e-05,
+      "loss": 1.8404,
+      "step": 6679
+    },
+    {
+      "epoch": 2.968888888888889,
+      "grad_norm": 2.41845440864563,
+      "learning_rate": 8.131672597864769e-05,
+      "loss": 1.413,
+      "step": 6680
+    },
+    {
+      "epoch": 2.969333333333333,
+      "grad_norm": 2.9708704948425293,
+      "learning_rate": 8.129893238434164e-05,
+      "loss": 2.0273,
+      "step": 6681
+    },
+    {
+      "epoch": 2.969777777777778,
+      "grad_norm": 3.0216286182403564,
+      "learning_rate": 8.12811387900356e-05,
+      "loss": 2.0485,
+      "step": 6682
+    },
+    {
+      "epoch": 2.970222222222222,
+      "grad_norm": 2.9172136783599854,
+      "learning_rate": 8.126334519572954e-05,
+      "loss": 1.5057,
+      "step": 6683
+    },
+    {
+      "epoch": 2.970666666666667,
+      "grad_norm": 3.159728765487671,
+      "learning_rate": 8.124555160142349e-05,
+      "loss": 1.9017,
+      "step": 6684
+    },
+    {
+      "epoch": 2.971111111111111,
+      "grad_norm": 2.4506280422210693,
+      "learning_rate": 8.122775800711744e-05,
+      "loss": 1.4581,
+      "step": 6685
+    },
+    {
+      "epoch": 2.9715555555555557,
+      "grad_norm": 2.9856271743774414,
+      "learning_rate": 8.120996441281138e-05,
+      "loss": 1.591,
+      "step": 6686
+    },
+    {
+      "epoch": 2.972,
+      "grad_norm": 3.172899007797241,
+      "learning_rate": 8.119217081850534e-05,
+      "loss": 1.7744,
+      "step": 6687
+    },
+    {
+      "epoch": 2.9724444444444442,
+      "grad_norm": 2.1155688762664795,
+      "learning_rate": 8.11743772241993e-05,
+      "loss": 1.1155,
+      "step": 6688
+    },
+    {
+      "epoch": 2.972888888888889,
+      "grad_norm": 2.4878904819488525,
+      "learning_rate": 8.115658362989324e-05,
+      "loss": 1.6012,
+      "step": 6689
+    },
+    {
+      "epoch": 2.9733333333333336,
+      "grad_norm": 2.819701671600342,
+      "learning_rate": 8.11387900355872e-05,
+      "loss": 1.6376,
+      "step": 6690
+    },
+    {
+      "epoch": 2.973777777777778,
+      "grad_norm": 3.1864094734191895,
+      "learning_rate": 8.112099644128114e-05,
+      "loss": 1.846,
+      "step": 6691
+    },
+    {
+      "epoch": 2.974222222222222,
+      "grad_norm": 3.0991835594177246,
+      "learning_rate": 8.11032028469751e-05,
+      "loss": 2.155,
+      "step": 6692
+    },
+    {
+      "epoch": 2.974666666666667,
+      "grad_norm": 2.6021008491516113,
+      "learning_rate": 8.108540925266904e-05,
+      "loss": 1.366,
+      "step": 6693
+    },
+    {
+      "epoch": 2.975111111111111,
+      "grad_norm": 3.0120420455932617,
+      "learning_rate": 8.1067615658363e-05,
+      "loss": 1.7095,
+      "step": 6694
+    },
+    {
+      "epoch": 2.9755555555555553,
+      "grad_norm": 3.54189395904541,
+      "learning_rate": 8.104982206405695e-05,
+      "loss": 1.5697,
+      "step": 6695
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 2.7960903644561768,
+      "learning_rate": 8.10320284697509e-05,
+      "loss": 1.5675,
+      "step": 6696
+    },
+    {
+      "epoch": 2.9764444444444447,
+      "grad_norm": 3.8344061374664307,
+      "learning_rate": 8.101423487544484e-05,
+      "loss": 1.8995,
+      "step": 6697
+    },
+    {
+      "epoch": 2.976888888888889,
+      "grad_norm": 3.2357256412506104,
+      "learning_rate": 8.09964412811388e-05,
+      "loss": 1.6873,
+      "step": 6698
+    },
+    {
+      "epoch": 2.977333333333333,
+      "grad_norm": 4.216027736663818,
+      "learning_rate": 8.097864768683274e-05,
+      "loss": 1.9569,
+      "step": 6699
+    },
+    {
+      "epoch": 2.977777777777778,
+      "grad_norm": 4.249828338623047,
+      "learning_rate": 8.09608540925267e-05,
+      "loss": 1.9055,
+      "step": 6700
+    },
+    {
+      "epoch": 2.978222222222222,
+      "grad_norm": 2.1459178924560547,
+      "learning_rate": 8.094306049822065e-05,
+      "loss": 2.1627,
+      "step": 6701
+    },
+    {
+      "epoch": 2.978666666666667,
+      "grad_norm": 1.4327117204666138,
+      "learning_rate": 8.09252669039146e-05,
+      "loss": 1.133,
+      "step": 6702
+    },
+    {
+      "epoch": 2.979111111111111,
+      "grad_norm": 1.9758896827697754,
+      "learning_rate": 8.090747330960855e-05,
+      "loss": 2.7843,
+      "step": 6703
+    },
+    {
+      "epoch": 2.9795555555555557,
+      "grad_norm": 2.312091112136841,
+      "learning_rate": 8.08896797153025e-05,
+      "loss": 2.1245,
+      "step": 6704
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.7319886684417725,
+      "learning_rate": 8.087188612099644e-05,
+      "loss": 0.4393,
+      "step": 6705
+    },
+    {
+      "epoch": 2.9804444444444442,
+      "grad_norm": 2.626613140106201,
+      "learning_rate": 8.085409252669039e-05,
+      "loss": 2.6569,
+      "step": 6706
+    },
+    {
+      "epoch": 2.980888888888889,
+      "grad_norm": 2.209878444671631,
+      "learning_rate": 8.083629893238435e-05,
+      "loss": 2.1272,
+      "step": 6707
+    },
+    {
+      "epoch": 2.981333333333333,
+      "grad_norm": 2.3179056644439697,
+      "learning_rate": 8.081850533807829e-05,
+      "loss": 1.7923,
+      "step": 6708
+    },
+    {
+      "epoch": 2.981777777777778,
+      "grad_norm": 2.408010959625244,
+      "learning_rate": 8.080071174377225e-05,
+      "loss": 1.7156,
+      "step": 6709
+    },
+    {
+      "epoch": 2.982222222222222,
+      "grad_norm": 3.3281619548797607,
+      "learning_rate": 8.078291814946619e-05,
+      "loss": 1.9335,
+      "step": 6710
+    },
+    {
+      "epoch": 2.982666666666667,
+      "grad_norm": 2.58201003074646,
+      "learning_rate": 8.076512455516015e-05,
+      "loss": 1.6612,
+      "step": 6711
+    },
+    {
+      "epoch": 2.983111111111111,
+      "grad_norm": 2.454719066619873,
+      "learning_rate": 8.074733096085409e-05,
+      "loss": 2.0895,
+      "step": 6712
+    },
+    {
+      "epoch": 2.9835555555555553,
+      "grad_norm": 2.9061319828033447,
+      "learning_rate": 8.072953736654805e-05,
+      "loss": 2.1351,
+      "step": 6713
+    },
+    {
+      "epoch": 2.984,
+      "grad_norm": 2.6297855377197266,
+      "learning_rate": 8.0711743772242e-05,
+      "loss": 1.849,
+      "step": 6714
+    },
+    {
+      "epoch": 2.9844444444444447,
+      "grad_norm": 2.431936264038086,
+      "learning_rate": 8.069395017793595e-05,
+      "loss": 1.809,
+      "step": 6715
+    },
+    {
+      "epoch": 2.984888888888889,
+      "grad_norm": 1.5601791143417358,
+      "learning_rate": 8.067615658362989e-05,
+      "loss": 0.7057,
+      "step": 6716
+    },
+    {
+      "epoch": 2.985333333333333,
+      "grad_norm": 2.3893916606903076,
+      "learning_rate": 8.065836298932385e-05,
+      "loss": 1.859,
+      "step": 6717
+    },
+    {
+      "epoch": 2.985777777777778,
+      "grad_norm": 2.8812737464904785,
+      "learning_rate": 8.064056939501779e-05,
+      "loss": 1.4955,
+      "step": 6718
+    },
+    {
+      "epoch": 2.986222222222222,
+      "grad_norm": 2.6631298065185547,
+      "learning_rate": 8.062277580071175e-05,
+      "loss": 1.8091,
+      "step": 6719
+    },
+    {
+      "epoch": 2.986666666666667,
+      "grad_norm": 2.2545714378356934,
+      "learning_rate": 8.06049822064057e-05,
+      "loss": 1.7845,
+      "step": 6720
+    },
+    {
+      "epoch": 2.987111111111111,
+      "grad_norm": 2.667125940322876,
+      "learning_rate": 8.058718861209965e-05,
+      "loss": 1.9743,
+      "step": 6721
+    },
+    {
+      "epoch": 2.9875555555555557,
+      "grad_norm": 2.8963820934295654,
+      "learning_rate": 8.05693950177936e-05,
+      "loss": 2.2487,
+      "step": 6722
+    },
+    {
+      "epoch": 2.988,
+      "grad_norm": 2.3376080989837646,
+      "learning_rate": 8.055160142348754e-05,
+      "loss": 1.7657,
+      "step": 6723
+    },
+    {
+      "epoch": 2.9884444444444442,
+      "grad_norm": 2.4739911556243896,
+      "learning_rate": 8.05338078291815e-05,
+      "loss": 1.4618,
+      "step": 6724
+    },
+    {
+      "epoch": 2.988888888888889,
+      "grad_norm": 2.836094856262207,
+      "learning_rate": 8.051601423487544e-05,
+      "loss": 1.8076,
+      "step": 6725
+    },
+    {
+      "epoch": 2.989333333333333,
+      "grad_norm": 2.611704111099243,
+      "learning_rate": 8.04982206405694e-05,
+      "loss": 1.8282,
+      "step": 6726
+    },
+    {
+      "epoch": 2.989777777777778,
+      "grad_norm": 2.016409158706665,
+      "learning_rate": 8.048042704626336e-05,
+      "loss": 1.579,
+      "step": 6727
+    },
+    {
+      "epoch": 2.990222222222222,
+      "grad_norm": 3.172990083694458,
+      "learning_rate": 8.04626334519573e-05,
+      "loss": 1.9213,
+      "step": 6728
+    },
+    {
+      "epoch": 2.990666666666667,
+      "grad_norm": 2.6530048847198486,
+      "learning_rate": 8.044483985765124e-05,
+      "loss": 1.3573,
+      "step": 6729
+    },
+    {
+      "epoch": 2.991111111111111,
+      "grad_norm": 2.725369453430176,
+      "learning_rate": 8.04270462633452e-05,
+      "loss": 1.7038,
+      "step": 6730
+    },
+    {
+      "epoch": 2.9915555555555553,
+      "grad_norm": 2.7324156761169434,
+      "learning_rate": 8.040925266903914e-05,
+      "loss": 1.9726,
+      "step": 6731
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 2.893840789794922,
+      "learning_rate": 8.03914590747331e-05,
+      "loss": 1.5817,
+      "step": 6732
+    },
+    {
+      "epoch": 2.9924444444444447,
+      "grad_norm": 3.1419413089752197,
+      "learning_rate": 8.037366548042706e-05,
+      "loss": 1.5595,
+      "step": 6733
+    },
+    {
+      "epoch": 2.992888888888889,
+      "grad_norm": 2.4272377490997314,
+      "learning_rate": 8.0355871886121e-05,
+      "loss": 1.338,
+      "step": 6734
+    },
+    {
+      "epoch": 2.993333333333333,
+      "grad_norm": 2.944011688232422,
+      "learning_rate": 8.033807829181496e-05,
+      "loss": 1.7456,
+      "step": 6735
+    },
+    {
+      "epoch": 2.993777777777778,
+      "grad_norm": 2.8351387977600098,
+      "learning_rate": 8.03202846975089e-05,
+      "loss": 1.6632,
+      "step": 6736
+    },
+    {
+      "epoch": 2.994222222222222,
+      "grad_norm": 3.1518852710723877,
+      "learning_rate": 8.030249110320285e-05,
+      "loss": 0.9622,
+      "step": 6737
+    },
+    {
+      "epoch": 2.994666666666667,
+      "grad_norm": 2.9322409629821777,
+      "learning_rate": 8.02846975088968e-05,
+      "loss": 1.1134,
+      "step": 6738
+    },
+    {
+      "epoch": 2.995111111111111,
+      "grad_norm": 2.9083549976348877,
+      "learning_rate": 8.026690391459075e-05,
+      "loss": 1.7185,
+      "step": 6739
+    },
+    {
+      "epoch": 2.9955555555555557,
+      "grad_norm": 2.974486827850342,
+      "learning_rate": 8.024911032028471e-05,
+      "loss": 1.9998,
+      "step": 6740
+    },
+    {
+      "epoch": 2.996,
+      "grad_norm": 2.9648959636688232,
+      "learning_rate": 8.023131672597865e-05,
+      "loss": 1.7564,
+      "step": 6741
+    },
+    {
+      "epoch": 2.9964444444444442,
+      "grad_norm": 3.1087772846221924,
+      "learning_rate": 8.02135231316726e-05,
+      "loss": 1.4747,
+      "step": 6742
+    },
+    {
+      "epoch": 2.996888888888889,
+      "grad_norm": 3.156559944152832,
+      "learning_rate": 8.019572953736655e-05,
+      "loss": 1.6082,
+      "step": 6743
+    },
+    {
+      "epoch": 2.997333333333333,
+      "grad_norm": 2.8267040252685547,
+      "learning_rate": 8.01779359430605e-05,
+      "loss": 1.7786,
+      "step": 6744
+    },
+    {
+      "epoch": 2.997777777777778,
+      "grad_norm": 3.1161437034606934,
+      "learning_rate": 8.016014234875445e-05,
+      "loss": 1.8029,
+      "step": 6745
+    },
+    {
+      "epoch": 2.998222222222222,
+      "grad_norm": 2.9387221336364746,
+      "learning_rate": 8.014234875444841e-05,
+      "loss": 1.4219,
+      "step": 6746
+    },
+    {
+      "epoch": 2.998666666666667,
+      "grad_norm": 3.949208974838257,
+      "learning_rate": 8.012455516014235e-05,
+      "loss": 2.0522,
+      "step": 6747
+    },
+    {
+      "epoch": 2.999111111111111,
+      "grad_norm": 3.7779977321624756,
+      "learning_rate": 8.010676156583631e-05,
+      "loss": 1.7612,
+      "step": 6748
+    },
+    {
+      "epoch": 2.9995555555555553,
+      "grad_norm": 3.2595958709716797,
+      "learning_rate": 8.008896797153025e-05,
+      "loss": 1.5605,
+      "step": 6749
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.8709285259246826,
+      "learning_rate": 8.00711743772242e-05,
+      "loss": 1.2562,
+      "step": 6750
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 2.560321807861328,
+      "eval_runtime": 47.1336,
+      "eval_samples_per_second": 10.608,
+      "eval_steps_per_second": 10.608,
+      "step": 6750
+    },
+    {
+      "epoch": 3.0004444444444442,
+      "grad_norm": 1.5255669355392456,
+      "learning_rate": 8.005338078291815e-05,
+      "loss": 1.4736,
+      "step": 6751
+    },
+    {
+      "epoch": 3.000888888888889,
+      "grad_norm": 1.8357324600219727,
+      "learning_rate": 8.003558718861211e-05,
+      "loss": 1.8316,
+      "step": 6752
+    },
+    {
+      "epoch": 3.001333333333333,
+      "grad_norm": 2.2091808319091797,
+      "learning_rate": 8.001779359430605e-05,
+      "loss": 1.994,
+      "step": 6753
+    },
+    {
+      "epoch": 3.001777777777778,
+      "grad_norm": 2.113086462020874,
+      "learning_rate": 8e-05,
+      "loss": 2.0161,
+      "step": 6754
+    },
+    {
+      "epoch": 3.002222222222222,
+      "grad_norm": 2.0146021842956543,
+      "learning_rate": 7.998220640569395e-05,
+      "loss": 1.3508,
+      "step": 6755
+    },
+    {
+      "epoch": 3.002666666666667,
+      "grad_norm": 2.153592348098755,
+      "learning_rate": 7.99644128113879e-05,
+      "loss": 1.2604,
+      "step": 6756
+    },
+    {
+      "epoch": 3.003111111111111,
+      "grad_norm": 1.5062990188598633,
+      "learning_rate": 7.994661921708185e-05,
+      "loss": 1.0908,
+      "step": 6757
+    },
+    {
+      "epoch": 3.0035555555555558,
+      "grad_norm": 2.1184024810791016,
+      "learning_rate": 7.99288256227758e-05,
+      "loss": 1.2938,
+      "step": 6758
+    },
+    {
+      "epoch": 3.004,
+      "grad_norm": 2.0012667179107666,
+      "learning_rate": 7.991103202846976e-05,
+      "loss": 1.3528,
+      "step": 6759
+    },
+    {
+      "epoch": 3.0044444444444443,
+      "grad_norm": 2.275223731994629,
+      "learning_rate": 7.98932384341637e-05,
+      "loss": 1.2119,
+      "step": 6760
+    },
+    {
+      "epoch": 3.004888888888889,
+      "grad_norm": 2.465026617050171,
+      "learning_rate": 7.987544483985766e-05,
+      "loss": 1.4336,
+      "step": 6761
+    },
+    {
+      "epoch": 3.005333333333333,
+      "grad_norm": 2.5878005027770996,
+      "learning_rate": 7.98576512455516e-05,
+      "loss": 1.4849,
+      "step": 6762
+    },
+    {
+      "epoch": 3.005777777777778,
+      "grad_norm": 2.4724271297454834,
+      "learning_rate": 7.983985765124555e-05,
+      "loss": 1.033,
+      "step": 6763
+    },
+    {
+      "epoch": 3.006222222222222,
+      "grad_norm": 2.9463562965393066,
+      "learning_rate": 7.98220640569395e-05,
+      "loss": 1.4481,
+      "step": 6764
+    },
+    {
+      "epoch": 3.006666666666667,
+      "grad_norm": 3.1810200214385986,
+      "learning_rate": 7.980427046263346e-05,
+      "loss": 0.9846,
+      "step": 6765
+    },
+    {
+      "epoch": 3.007111111111111,
+      "grad_norm": 2.737637758255005,
+      "learning_rate": 7.97864768683274e-05,
+      "loss": 1.0059,
+      "step": 6766
+    },
+    {
+      "epoch": 3.0075555555555558,
+      "grad_norm": 2.819664239883423,
+      "learning_rate": 7.976868327402136e-05,
+      "loss": 1.331,
+      "step": 6767
+    },
+    {
+      "epoch": 3.008,
+      "grad_norm": 2.880838632583618,
+      "learning_rate": 7.97508896797153e-05,
+      "loss": 1.6169,
+      "step": 6768
+    },
+    {
+      "epoch": 3.0084444444444443,
+      "grad_norm": 2.998161792755127,
+      "learning_rate": 7.973309608540926e-05,
+      "loss": 1.3927,
+      "step": 6769
+    },
+    {
+      "epoch": 3.008888888888889,
+      "grad_norm": 2.9646785259246826,
+      "learning_rate": 7.97153024911032e-05,
+      "loss": 1.1182,
+      "step": 6770
+    },
+    {
+      "epoch": 3.009333333333333,
+      "grad_norm": 3.8950388431549072,
+      "learning_rate": 7.969750889679716e-05,
+      "loss": 1.0995,
+      "step": 6771
+    },
+    {
+      "epoch": 3.009777777777778,
+      "grad_norm": 3.958792209625244,
+      "learning_rate": 7.967971530249112e-05,
+      "loss": 1.3026,
+      "step": 6772
+    },
+    {
+      "epoch": 3.010222222222222,
+      "grad_norm": 3.3509862422943115,
+      "learning_rate": 7.966192170818506e-05,
+      "loss": 1.511,
+      "step": 6773
+    },
+    {
+      "epoch": 3.010666666666667,
+      "grad_norm": 3.338815689086914,
+      "learning_rate": 7.9644128113879e-05,
+      "loss": 1.1785,
+      "step": 6774
+    },
+    {
+      "epoch": 3.011111111111111,
+      "grad_norm": 4.509990692138672,
+      "learning_rate": 7.962633451957296e-05,
+      "loss": 1.4605,
+      "step": 6775
+    },
+    {
+      "epoch": 3.0115555555555558,
+      "grad_norm": 3.72403883934021,
+      "learning_rate": 7.96085409252669e-05,
+      "loss": 1.3251,
+      "step": 6776
+    },
+    {
+      "epoch": 3.012,
+      "grad_norm": 3.0874853134155273,
+      "learning_rate": 7.959074733096086e-05,
+      "loss": 1.424,
+      "step": 6777
+    },
+    {
+      "epoch": 3.0124444444444443,
+      "grad_norm": 2.400561809539795,
+      "learning_rate": 7.957295373665481e-05,
+      "loss": 0.6489,
+      "step": 6778
+    },
+    {
+      "epoch": 3.012888888888889,
+      "grad_norm": 3.6095080375671387,
+      "learning_rate": 7.955516014234876e-05,
+      "loss": 0.9895,
+      "step": 6779
+    },
+    {
+      "epoch": 3.013333333333333,
+      "grad_norm": 4.059244632720947,
+      "learning_rate": 7.953736654804271e-05,
+      "loss": 1.0755,
+      "step": 6780
+    },
+    {
+      "epoch": 3.013777777777778,
+      "grad_norm": 3.3951244354248047,
+      "learning_rate": 7.951957295373666e-05,
+      "loss": 1.4593,
+      "step": 6781
+    },
+    {
+      "epoch": 3.014222222222222,
+      "grad_norm": 3.264514207839966,
+      "learning_rate": 7.950177935943061e-05,
+      "loss": 1.1751,
+      "step": 6782
+    },
+    {
+      "epoch": 3.014666666666667,
+      "grad_norm": 3.5838840007781982,
+      "learning_rate": 7.948398576512456e-05,
+      "loss": 1.2873,
+      "step": 6783
+    },
+    {
+      "epoch": 3.015111111111111,
+      "grad_norm": 3.029461622238159,
+      "learning_rate": 7.946619217081851e-05,
+      "loss": 0.5727,
+      "step": 6784
+    },
+    {
+      "epoch": 3.0155555555555558,
+      "grad_norm": 2.3278002738952637,
+      "learning_rate": 7.944839857651247e-05,
+      "loss": 0.5224,
+      "step": 6785
+    },
+    {
+      "epoch": 3.016,
+      "grad_norm": 3.3948307037353516,
+      "learning_rate": 7.943060498220641e-05,
+      "loss": 1.5969,
+      "step": 6786
+    },
+    {
+      "epoch": 3.0164444444444443,
+      "grad_norm": 3.4269049167633057,
+      "learning_rate": 7.941281138790035e-05,
+      "loss": 1.3439,
+      "step": 6787
+    },
+    {
+      "epoch": 3.016888888888889,
+      "grad_norm": 4.021193504333496,
+      "learning_rate": 7.939501779359431e-05,
+      "loss": 1.429,
+      "step": 6788
+    },
+    {
+      "epoch": 3.017333333333333,
+      "grad_norm": 3.9195425510406494,
+      "learning_rate": 7.937722419928825e-05,
+      "loss": 1.3733,
+      "step": 6789
+    },
+    {
+      "epoch": 3.017777777777778,
+      "grad_norm": 3.872466564178467,
+      "learning_rate": 7.935943060498221e-05,
+      "loss": 1.1731,
+      "step": 6790
+    },
+    {
+      "epoch": 3.018222222222222,
+      "grad_norm": 4.006219387054443,
+      "learning_rate": 7.934163701067617e-05,
+      "loss": 1.5755,
+      "step": 6791
+    },
+    {
+      "epoch": 3.018666666666667,
+      "grad_norm": 4.188656330108643,
+      "learning_rate": 7.932384341637011e-05,
+      "loss": 1.4132,
+      "step": 6792
+    },
+    {
+      "epoch": 3.019111111111111,
+      "grad_norm": 3.21732234954834,
+      "learning_rate": 7.930604982206407e-05,
+      "loss": 0.9631,
+      "step": 6793
+    },
+    {
+      "epoch": 3.0195555555555558,
+      "grad_norm": 2.5914382934570312,
+      "learning_rate": 7.928825622775801e-05,
+      "loss": 0.9383,
+      "step": 6794
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 3.1577656269073486,
+      "learning_rate": 7.927046263345195e-05,
+      "loss": 1.1981,
+      "step": 6795
+    },
+    {
+      "epoch": 3.0204444444444443,
+      "grad_norm": 3.7935643196105957,
+      "learning_rate": 7.925266903914591e-05,
+      "loss": 1.2181,
+      "step": 6796
+    },
+    {
+      "epoch": 3.020888888888889,
+      "grad_norm": 4.230493068695068,
+      "learning_rate": 7.923487544483986e-05,
+      "loss": 1.467,
+      "step": 6797
+    },
+    {
+      "epoch": 3.021333333333333,
+      "grad_norm": 4.8975982666015625,
+      "learning_rate": 7.921708185053381e-05,
+      "loss": 1.0094,
+      "step": 6798
+    },
+    {
+      "epoch": 3.021777777777778,
+      "grad_norm": 3.1656551361083984,
+      "learning_rate": 7.919928825622776e-05,
+      "loss": 0.5155,
+      "step": 6799
+    },
+    {
+      "epoch": 3.022222222222222,
+      "grad_norm": 4.159629821777344,
+      "learning_rate": 7.918149466192171e-05,
+      "loss": 0.1859,
+      "step": 6800
+    },
+    {
+      "epoch": 3.022666666666667,
+      "grad_norm": 3.048837900161743,
+      "learning_rate": 7.916370106761566e-05,
+      "loss": 0.8377,
+      "step": 6801
+    },
+    {
+      "epoch": 3.023111111111111,
+      "grad_norm": 2.2959177494049072,
+      "learning_rate": 7.91459074733096e-05,
+      "loss": 1.6241,
+      "step": 6802
+    },
+    {
+      "epoch": 3.0235555555555558,
+      "grad_norm": 1.9068602323532104,
+      "learning_rate": 7.912811387900356e-05,
+      "loss": 0.0712,
+      "step": 6803
+    },
+    {
+      "epoch": 3.024,
+      "grad_norm": 1.6793782711029053,
+      "learning_rate": 7.911032028469752e-05,
+      "loss": 0.5563,
+      "step": 6804
+    },
+    {
+      "epoch": 3.0244444444444443,
+      "grad_norm": 2.5858163833618164,
+      "learning_rate": 7.909252669039146e-05,
+      "loss": 1.3781,
+      "step": 6805
+    },
+    {
+      "epoch": 3.024888888888889,
+      "grad_norm": 2.612051010131836,
+      "learning_rate": 7.907473309608542e-05,
+      "loss": 1.5614,
+      "step": 6806
+    },
+    {
+      "epoch": 3.025333333333333,
+      "grad_norm": 2.8191487789154053,
+      "learning_rate": 7.905693950177936e-05,
+      "loss": 1.3573,
+      "step": 6807
+    },
+    {
+      "epoch": 3.025777777777778,
+      "grad_norm": 2.706998109817505,
+      "learning_rate": 7.90391459074733e-05,
+      "loss": 1.6064,
+      "step": 6808
+    },
+    {
+      "epoch": 3.026222222222222,
+      "grad_norm": 2.847705841064453,
+      "learning_rate": 7.902135231316726e-05,
+      "loss": 1.0591,
+      "step": 6809
+    },
+    {
+      "epoch": 3.026666666666667,
+      "grad_norm": 2.685861825942993,
+      "learning_rate": 7.900355871886122e-05,
+      "loss": 1.5385,
+      "step": 6810
+    },
+    {
+      "epoch": 3.027111111111111,
+      "grad_norm": 3.0858583450317383,
+      "learning_rate": 7.898576512455516e-05,
+      "loss": 1.4013,
+      "step": 6811
+    },
+    {
+      "epoch": 3.0275555555555558,
+      "grad_norm": 2.825329065322876,
+      "learning_rate": 7.896797153024912e-05,
+      "loss": 1.4406,
+      "step": 6812
+    },
+    {
+      "epoch": 3.028,
+      "grad_norm": 3.5686569213867188,
+      "learning_rate": 7.895017793594306e-05,
+      "loss": 1.3895,
+      "step": 6813
+    },
+    {
+      "epoch": 3.0284444444444443,
+      "grad_norm": 3.122117280960083,
+      "learning_rate": 7.893238434163702e-05,
+      "loss": 1.3254,
+      "step": 6814
+    },
+    {
+      "epoch": 3.028888888888889,
+      "grad_norm": 2.970949649810791,
+      "learning_rate": 7.891459074733096e-05,
+      "loss": 1.4918,
+      "step": 6815
+    },
+    {
+      "epoch": 3.029333333333333,
+      "grad_norm": 2.8904595375061035,
+      "learning_rate": 7.889679715302492e-05,
+      "loss": 1.2927,
+      "step": 6816
+    },
+    {
+      "epoch": 3.029777777777778,
+      "grad_norm": 3.0297486782073975,
+      "learning_rate": 7.887900355871887e-05,
+      "loss": 1.4865,
+      "step": 6817
+    },
+    {
+      "epoch": 3.030222222222222,
+      "grad_norm": 3.1951310634613037,
+      "learning_rate": 7.886120996441282e-05,
+      "loss": 1.1547,
+      "step": 6818
+    },
+    {
+      "epoch": 3.030666666666667,
+      "grad_norm": 1.9806190729141235,
+      "learning_rate": 7.884341637010677e-05,
+      "loss": 0.5659,
+      "step": 6819
+    },
+    {
+      "epoch": 3.031111111111111,
+      "grad_norm": 2.957610845565796,
+      "learning_rate": 7.882562277580072e-05,
+      "loss": 1.3614,
+      "step": 6820
+    },
+    {
+      "epoch": 3.0315555555555553,
+      "grad_norm": 3.6964633464813232,
+      "learning_rate": 7.880782918149466e-05,
+      "loss": 1.7276,
+      "step": 6821
+    },
+    {
+      "epoch": 3.032,
+      "grad_norm": 3.108067512512207,
+      "learning_rate": 7.879003558718861e-05,
+      "loss": 1.4918,
+      "step": 6822
+    },
+    {
+      "epoch": 3.0324444444444443,
+      "grad_norm": 2.6663601398468018,
+      "learning_rate": 7.877224199288257e-05,
+      "loss": 1.233,
+      "step": 6823
+    },
+    {
+      "epoch": 3.032888888888889,
+      "grad_norm": 3.010274648666382,
+      "learning_rate": 7.875444839857651e-05,
+      "loss": 1.3845,
+      "step": 6824
+    },
+    {
+      "epoch": 3.033333333333333,
+      "grad_norm": 3.353402614593506,
+      "learning_rate": 7.873665480427047e-05,
+      "loss": 1.3772,
+      "step": 6825
+    },
+    {
+      "epoch": 3.033777777777778,
+      "grad_norm": 2.9852728843688965,
+      "learning_rate": 7.871886120996441e-05,
+      "loss": 1.5947,
+      "step": 6826
+    },
+    {
+      "epoch": 3.034222222222222,
+      "grad_norm": 2.881147861480713,
+      "learning_rate": 7.870106761565837e-05,
+      "loss": 1.3783,
+      "step": 6827
+    },
+    {
+      "epoch": 3.034666666666667,
+      "grad_norm": 3.402742385864258,
+      "learning_rate": 7.868327402135231e-05,
+      "loss": 1.1179,
+      "step": 6828
+    },
+    {
+      "epoch": 3.035111111111111,
+      "grad_norm": 2.7964463233947754,
+      "learning_rate": 7.866548042704627e-05,
+      "loss": 0.6328,
+      "step": 6829
+    },
+    {
+      "epoch": 3.0355555555555553,
+      "grad_norm": 3.21201229095459,
+      "learning_rate": 7.864768683274023e-05,
+      "loss": 1.4121,
+      "step": 6830
+    },
+    {
+      "epoch": 3.036,
+      "grad_norm": 2.9257187843322754,
+      "learning_rate": 7.862989323843417e-05,
+      "loss": 0.7511,
+      "step": 6831
+    },
+    {
+      "epoch": 3.0364444444444443,
+      "grad_norm": 3.565912961959839,
+      "learning_rate": 7.861209964412811e-05,
+      "loss": 1.5339,
+      "step": 6832
+    },
+    {
+      "epoch": 3.036888888888889,
+      "grad_norm": 2.8367793560028076,
+      "learning_rate": 7.859430604982207e-05,
+      "loss": 1.2353,
+      "step": 6833
+    },
+    {
+      "epoch": 3.037333333333333,
+      "grad_norm": 3.7711472511291504,
+      "learning_rate": 7.857651245551601e-05,
+      "loss": 1.4139,
+      "step": 6834
+    },
+    {
+      "epoch": 3.037777777777778,
+      "grad_norm": 2.9545300006866455,
+      "learning_rate": 7.855871886120997e-05,
+      "loss": 1.3022,
+      "step": 6835
+    },
+    {
+      "epoch": 3.038222222222222,
+      "grad_norm": 3.401104688644409,
+      "learning_rate": 7.854092526690392e-05,
+      "loss": 1.3024,
+      "step": 6836
+    },
+    {
+      "epoch": 3.038666666666667,
+      "grad_norm": 3.2514753341674805,
+      "learning_rate": 7.852313167259787e-05,
+      "loss": 0.9559,
+      "step": 6837
+    },
+    {
+      "epoch": 3.039111111111111,
+      "grad_norm": 4.7731547355651855,
+      "learning_rate": 7.850533807829182e-05,
+      "loss": 1.3061,
+      "step": 6838
+    },
+    {
+      "epoch": 3.0395555555555553,
+      "grad_norm": 3.2436811923980713,
+      "learning_rate": 7.848754448398577e-05,
+      "loss": 1.0786,
+      "step": 6839
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.0958991050720215,
+      "learning_rate": 7.846975088967971e-05,
+      "loss": 1.2044,
+      "step": 6840
+    },
+    {
+      "epoch": 3.0404444444444443,
+      "grad_norm": 3.618880271911621,
+      "learning_rate": 7.845195729537367e-05,
+      "loss": 1.0556,
+      "step": 6841
+    },
+    {
+      "epoch": 3.040888888888889,
+      "grad_norm": 4.0584001541137695,
+      "learning_rate": 7.843416370106762e-05,
+      "loss": 1.1659,
+      "step": 6842
+    },
+    {
+      "epoch": 3.041333333333333,
+      "grad_norm": 4.1572418212890625,
+      "learning_rate": 7.841637010676157e-05,
+      "loss": 1.0593,
+      "step": 6843
+    },
+    {
+      "epoch": 3.041777777777778,
+      "grad_norm": 3.093219757080078,
+      "learning_rate": 7.839857651245552e-05,
+      "loss": 1.1444,
+      "step": 6844
+    },
+    {
+      "epoch": 3.042222222222222,
+      "grad_norm": 4.556105613708496,
+      "learning_rate": 7.838078291814947e-05,
+      "loss": 1.3215,
+      "step": 6845
+    },
+    {
+      "epoch": 3.042666666666667,
+      "grad_norm": 3.609612226486206,
+      "learning_rate": 7.836298932384342e-05,
+      "loss": 1.3856,
+      "step": 6846
+    },
+    {
+      "epoch": 3.043111111111111,
+      "grad_norm": 4.959066867828369,
+      "learning_rate": 7.834519572953736e-05,
+      "loss": 1.1098,
+      "step": 6847
+    },
+    {
+      "epoch": 3.0435555555555553,
+      "grad_norm": 3.45310640335083,
+      "learning_rate": 7.832740213523132e-05,
+      "loss": 0.7384,
+      "step": 6848
+    },
+    {
+      "epoch": 3.044,
+      "grad_norm": 5.438348770141602,
+      "learning_rate": 7.830960854092528e-05,
+      "loss": 1.2785,
+      "step": 6849
+    },
+    {
+      "epoch": 3.0444444444444443,
+      "grad_norm": 6.3002119064331055,
+      "learning_rate": 7.829181494661922e-05,
+      "loss": 0.9269,
+      "step": 6850
+    },
+    {
+      "epoch": 3.044888888888889,
+      "grad_norm": 2.523265838623047,
+      "learning_rate": 7.827402135231318e-05,
+      "loss": 1.9499,
+      "step": 6851
+    },
+    {
+      "epoch": 3.0453333333333332,
+      "grad_norm": 2.234962224960327,
+      "learning_rate": 7.825622775800712e-05,
+      "loss": 1.646,
+      "step": 6852
+    },
+    {
+      "epoch": 3.045777777777778,
+      "grad_norm": 2.712780475616455,
+      "learning_rate": 7.823843416370106e-05,
+      "loss": 1.793,
+      "step": 6853
+    },
+    {
+      "epoch": 3.046222222222222,
+      "grad_norm": 2.686671018600464,
+      "learning_rate": 7.822064056939502e-05,
+      "loss": 1.5653,
+      "step": 6854
+    },
+    {
+      "epoch": 3.046666666666667,
+      "grad_norm": 2.907186269760132,
+      "learning_rate": 7.820284697508898e-05,
+      "loss": 1.5827,
+      "step": 6855
+    },
+    {
+      "epoch": 3.047111111111111,
+      "grad_norm": 2.4521493911743164,
+      "learning_rate": 7.818505338078292e-05,
+      "loss": 1.5815,
+      "step": 6856
+    },
+    {
+      "epoch": 3.0475555555555554,
+      "grad_norm": 2.803241491317749,
+      "learning_rate": 7.816725978647688e-05,
+      "loss": 1.5903,
+      "step": 6857
+    },
+    {
+      "epoch": 3.048,
+      "grad_norm": 2.856114625930786,
+      "learning_rate": 7.814946619217082e-05,
+      "loss": 1.3927,
+      "step": 6858
+    },
+    {
+      "epoch": 3.0484444444444443,
+      "grad_norm": 3.376676082611084,
+      "learning_rate": 7.813167259786477e-05,
+      "loss": 1.561,
+      "step": 6859
+    },
+    {
+      "epoch": 3.048888888888889,
+      "grad_norm": 3.10386061668396,
+      "learning_rate": 7.811387900355872e-05,
+      "loss": 1.473,
+      "step": 6860
+    },
+    {
+      "epoch": 3.0493333333333332,
+      "grad_norm": 3.0056633949279785,
+      "learning_rate": 7.809608540925267e-05,
+      "loss": 1.3637,
+      "step": 6861
+    },
+    {
+      "epoch": 3.049777777777778,
+      "grad_norm": 3.1798784732818604,
+      "learning_rate": 7.807829181494663e-05,
+      "loss": 1.7298,
+      "step": 6862
+    },
+    {
+      "epoch": 3.050222222222222,
+      "grad_norm": 3.0518245697021484,
+      "learning_rate": 7.806049822064057e-05,
+      "loss": 1.2918,
+      "step": 6863
+    },
+    {
+      "epoch": 3.050666666666667,
+      "grad_norm": 2.996016502380371,
+      "learning_rate": 7.804270462633453e-05,
+      "loss": 1.5182,
+      "step": 6864
+    },
+    {
+      "epoch": 3.051111111111111,
+      "grad_norm": 3.0778775215148926,
+      "learning_rate": 7.802491103202847e-05,
+      "loss": 1.7565,
+      "step": 6865
+    },
+    {
+      "epoch": 3.0515555555555554,
+      "grad_norm": 2.45369553565979,
+      "learning_rate": 7.800711743772242e-05,
+      "loss": 0.9105,
+      "step": 6866
+    },
+    {
+      "epoch": 3.052,
+      "grad_norm": 3.1584999561309814,
+      "learning_rate": 7.798932384341637e-05,
+      "loss": 1.5315,
+      "step": 6867
+    },
+    {
+      "epoch": 3.0524444444444443,
+      "grad_norm": 2.7491531372070312,
+      "learning_rate": 7.797153024911033e-05,
+      "loss": 1.2398,
+      "step": 6868
+    },
+    {
+      "epoch": 3.052888888888889,
+      "grad_norm": 3.181427478790283,
+      "learning_rate": 7.795373665480427e-05,
+      "loss": 1.4826,
+      "step": 6869
+    },
+    {
+      "epoch": 3.0533333333333332,
+      "grad_norm": 2.462205648422241,
+      "learning_rate": 7.793594306049823e-05,
+      "loss": 0.7215,
+      "step": 6870
+    },
+    {
+      "epoch": 3.053777777777778,
+      "grad_norm": 2.8350601196289062,
+      "learning_rate": 7.791814946619217e-05,
+      "loss": 1.4129,
+      "step": 6871
+    },
+    {
+      "epoch": 3.054222222222222,
+      "grad_norm": 3.099984645843506,
+      "learning_rate": 7.790035587188613e-05,
+      "loss": 1.1478,
+      "step": 6872
+    },
+    {
+      "epoch": 3.054666666666667,
+      "grad_norm": 2.893434524536133,
+      "learning_rate": 7.788256227758007e-05,
+      "loss": 1.4698,
+      "step": 6873
+    },
+    {
+      "epoch": 3.055111111111111,
+      "grad_norm": 3.185155153274536,
+      "learning_rate": 7.786476868327403e-05,
+      "loss": 1.6558,
+      "step": 6874
+    },
+    {
+      "epoch": 3.0555555555555554,
+      "grad_norm": 2.3959739208221436,
+      "learning_rate": 7.784697508896798e-05,
+      "loss": 1.266,
+      "step": 6875
+    },
+    {
+      "epoch": 3.056,
+      "grad_norm": 3.058983564376831,
+      "learning_rate": 7.782918149466193e-05,
+      "loss": 1.3393,
+      "step": 6876
+    },
+    {
+      "epoch": 3.0564444444444443,
+      "grad_norm": 3.080317974090576,
+      "learning_rate": 7.781138790035588e-05,
+      "loss": 1.5509,
+      "step": 6877
+    },
+    {
+      "epoch": 3.056888888888889,
+      "grad_norm": 3.090117931365967,
+      "learning_rate": 7.779359430604983e-05,
+      "loss": 1.2201,
+      "step": 6878
+    },
+    {
+      "epoch": 3.0573333333333332,
+      "grad_norm": 2.1614432334899902,
+      "learning_rate": 7.777580071174377e-05,
+      "loss": 0.6346,
+      "step": 6879
+    },
+    {
+      "epoch": 3.057777777777778,
+      "grad_norm": 2.7697033882141113,
+      "learning_rate": 7.775800711743773e-05,
+      "loss": 1.1247,
+      "step": 6880
+    },
+    {
+      "epoch": 3.058222222222222,
+      "grad_norm": 3.1787707805633545,
+      "learning_rate": 7.774021352313168e-05,
+      "loss": 1.0795,
+      "step": 6881
+    },
+    {
+      "epoch": 3.058666666666667,
+      "grad_norm": 3.6657590866088867,
+      "learning_rate": 7.772241992882563e-05,
+      "loss": 1.6009,
+      "step": 6882
+    },
+    {
+      "epoch": 3.059111111111111,
+      "grad_norm": 3.2929611206054688,
+      "learning_rate": 7.770462633451958e-05,
+      "loss": 1.4993,
+      "step": 6883
+    },
+    {
+      "epoch": 3.0595555555555554,
+      "grad_norm": 3.753507137298584,
+      "learning_rate": 7.768683274021352e-05,
+      "loss": 1.6565,
+      "step": 6884
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 2.848525047302246,
+      "learning_rate": 7.766903914590747e-05,
+      "loss": 0.9654,
+      "step": 6885
+    },
+    {
+      "epoch": 3.0604444444444443,
+      "grad_norm": 3.355612277984619,
+      "learning_rate": 7.765124555160142e-05,
+      "loss": 1.0126,
+      "step": 6886
+    },
+    {
+      "epoch": 3.060888888888889,
+      "grad_norm": 3.229212522506714,
+      "learning_rate": 7.763345195729538e-05,
+      "loss": 0.7328,
+      "step": 6887
+    },
+    {
+      "epoch": 3.0613333333333332,
+      "grad_norm": 4.031661033630371,
+      "learning_rate": 7.761565836298932e-05,
+      "loss": 1.3381,
+      "step": 6888
+    },
+    {
+      "epoch": 3.061777777777778,
+      "grad_norm": 3.2130744457244873,
+      "learning_rate": 7.759786476868328e-05,
+      "loss": 1.1694,
+      "step": 6889
+    },
+    {
+      "epoch": 3.062222222222222,
+      "grad_norm": 4.325170040130615,
+      "learning_rate": 7.758007117437722e-05,
+      "loss": 0.9918,
+      "step": 6890
+    },
+    {
+      "epoch": 3.062666666666667,
+      "grad_norm": 3.6601080894470215,
+      "learning_rate": 7.756227758007118e-05,
+      "loss": 1.6125,
+      "step": 6891
+    },
+    {
+      "epoch": 3.063111111111111,
+      "grad_norm": 2.9753174781799316,
+      "learning_rate": 7.754448398576512e-05,
+      "loss": 0.8867,
+      "step": 6892
+    },
+    {
+      "epoch": 3.0635555555555554,
+      "grad_norm": 4.441684246063232,
+      "learning_rate": 7.752669039145908e-05,
+      "loss": 1.6672,
+      "step": 6893
+    },
+    {
+      "epoch": 3.064,
+      "grad_norm": 5.101708889007568,
+      "learning_rate": 7.750889679715304e-05,
+      "loss": 1.3812,
+      "step": 6894
+    },
+    {
+      "epoch": 3.0644444444444443,
+      "grad_norm": 4.84489631652832,
+      "learning_rate": 7.749110320284698e-05,
+      "loss": 1.4592,
+      "step": 6895
+    },
+    {
+      "epoch": 3.064888888888889,
+      "grad_norm": 4.635469913482666,
+      "learning_rate": 7.747330960854093e-05,
+      "loss": 1.2783,
+      "step": 6896
+    },
+    {
+      "epoch": 3.0653333333333332,
+      "grad_norm": 4.737509727478027,
+      "learning_rate": 7.745551601423488e-05,
+      "loss": 1.1961,
+      "step": 6897
+    },
+    {
+      "epoch": 3.065777777777778,
+      "grad_norm": 5.450290203094482,
+      "learning_rate": 7.743772241992882e-05,
+      "loss": 0.9652,
+      "step": 6898
+    },
+    {
+      "epoch": 3.066222222222222,
+      "grad_norm": 4.097311973571777,
+      "learning_rate": 7.741992882562278e-05,
+      "loss": 0.9152,
+      "step": 6899
+    },
+    {
+      "epoch": 3.066666666666667,
+      "grad_norm": 4.8908257484436035,
+      "learning_rate": 7.740213523131673e-05,
+      "loss": 0.9301,
+      "step": 6900
+    },
+    {
+      "epoch": 3.067111111111111,
+      "grad_norm": 1.8603249788284302,
+      "learning_rate": 7.738434163701068e-05,
+      "loss": 0.848,
+      "step": 6901
+    },
+    {
+      "epoch": 3.0675555555555554,
+      "grad_norm": 2.392153263092041,
+      "learning_rate": 7.736654804270463e-05,
+      "loss": 1.9415,
+      "step": 6902
+    },
+    {
+      "epoch": 3.068,
+      "grad_norm": 2.698664426803589,
+      "learning_rate": 7.734875444839858e-05,
+      "loss": 1.7879,
+      "step": 6903
+    },
+    {
+      "epoch": 3.0684444444444443,
+      "grad_norm": 2.744234561920166,
+      "learning_rate": 7.733096085409253e-05,
+      "loss": 1.7097,
+      "step": 6904
+    },
+    {
+      "epoch": 3.068888888888889,
+      "grad_norm": 3.156621217727661,
+      "learning_rate": 7.731316725978648e-05,
+      "loss": 2.1794,
+      "step": 6905
+    },
+    {
+      "epoch": 3.0693333333333332,
+      "grad_norm": 2.774522066116333,
+      "learning_rate": 7.729537366548043e-05,
+      "loss": 1.6727,
+      "step": 6906
+    },
+    {
+      "epoch": 3.069777777777778,
+      "grad_norm": 3.170665740966797,
+      "learning_rate": 7.727758007117439e-05,
+      "loss": 0.7644,
+      "step": 6907
+    },
+    {
+      "epoch": 3.070222222222222,
+      "grad_norm": 2.518483877182007,
+      "learning_rate": 7.725978647686833e-05,
+      "loss": 1.337,
+      "step": 6908
+    },
+    {
+      "epoch": 3.070666666666667,
+      "grad_norm": 2.7955567836761475,
+      "learning_rate": 7.724199288256229e-05,
+      "loss": 1.1453,
+      "step": 6909
+    },
+    {
+      "epoch": 3.071111111111111,
+      "grad_norm": 3.288102865219116,
+      "learning_rate": 7.722419928825623e-05,
+      "loss": 1.2716,
+      "step": 6910
+    },
+    {
+      "epoch": 3.0715555555555554,
+      "grad_norm": 3.349630355834961,
+      "learning_rate": 7.720640569395017e-05,
+      "loss": 1.716,
+      "step": 6911
+    },
+    {
+      "epoch": 3.072,
+      "grad_norm": 2.8400745391845703,
+      "learning_rate": 7.718861209964413e-05,
+      "loss": 1.2617,
+      "step": 6912
+    },
+    {
+      "epoch": 3.0724444444444443,
+      "grad_norm": 2.843729257583618,
+      "learning_rate": 7.717081850533809e-05,
+      "loss": 1.4572,
+      "step": 6913
+    },
+    {
+      "epoch": 3.072888888888889,
+      "grad_norm": 3.0647592544555664,
+      "learning_rate": 7.715302491103203e-05,
+      "loss": 1.5921,
+      "step": 6914
+    },
+    {
+      "epoch": 3.0733333333333333,
+      "grad_norm": 3.1365675926208496,
+      "learning_rate": 7.713523131672599e-05,
+      "loss": 1.5678,
+      "step": 6915
+    },
+    {
+      "epoch": 3.073777777777778,
+      "grad_norm": 2.7604968547821045,
+      "learning_rate": 7.711743772241993e-05,
+      "loss": 1.3843,
+      "step": 6916
+    },
+    {
+      "epoch": 3.074222222222222,
+      "grad_norm": 2.654712438583374,
+      "learning_rate": 7.709964412811389e-05,
+      "loss": 1.0891,
+      "step": 6917
+    },
+    {
+      "epoch": 3.074666666666667,
+      "grad_norm": 3.091094732284546,
+      "learning_rate": 7.708185053380783e-05,
+      "loss": 1.4574,
+      "step": 6918
+    },
+    {
+      "epoch": 3.075111111111111,
+      "grad_norm": 2.8697140216827393,
+      "learning_rate": 7.706405693950179e-05,
+      "loss": 1.2856,
+      "step": 6919
+    },
+    {
+      "epoch": 3.0755555555555554,
+      "grad_norm": 4.098965644836426,
+      "learning_rate": 7.704626334519574e-05,
+      "loss": 1.2447,
+      "step": 6920
+    },
+    {
+      "epoch": 3.076,
+      "grad_norm": 3.0717883110046387,
+      "learning_rate": 7.702846975088968e-05,
+      "loss": 1.2924,
+      "step": 6921
+    },
+    {
+      "epoch": 3.0764444444444443,
+      "grad_norm": 3.0707571506500244,
+      "learning_rate": 7.701067615658364e-05,
+      "loss": 1.3306,
+      "step": 6922
+    },
+    {
+      "epoch": 3.076888888888889,
+      "grad_norm": 3.0397446155548096,
+      "learning_rate": 7.699288256227758e-05,
+      "loss": 1.3013,
+      "step": 6923
+    },
+    {
+      "epoch": 3.0773333333333333,
+      "grad_norm": 3.693847894668579,
+      "learning_rate": 7.697508896797153e-05,
+      "loss": 1.4978,
+      "step": 6924
+    },
+    {
+      "epoch": 3.077777777777778,
+      "grad_norm": 3.112276077270508,
+      "learning_rate": 7.695729537366548e-05,
+      "loss": 1.4023,
+      "step": 6925
+    },
+    {
+      "epoch": 3.078222222222222,
+      "grad_norm": 3.29516863822937,
+      "learning_rate": 7.693950177935944e-05,
+      "loss": 1.2205,
+      "step": 6926
+    },
+    {
+      "epoch": 3.078666666666667,
+      "grad_norm": 2.247548818588257,
+      "learning_rate": 7.692170818505338e-05,
+      "loss": 0.6385,
+      "step": 6927
+    },
+    {
+      "epoch": 3.079111111111111,
+      "grad_norm": 2.3756346702575684,
+      "learning_rate": 7.690391459074734e-05,
+      "loss": 0.7201,
+      "step": 6928
+    },
+    {
+      "epoch": 3.0795555555555554,
+      "grad_norm": 3.00430965423584,
+      "learning_rate": 7.688612099644128e-05,
+      "loss": 1.0356,
+      "step": 6929
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.532029390335083,
+      "learning_rate": 7.686832740213523e-05,
+      "loss": 0.5134,
+      "step": 6930
+    },
+    {
+      "epoch": 3.0804444444444443,
+      "grad_norm": 0.30651673674583435,
+      "learning_rate": 7.685053380782918e-05,
+      "loss": 0.0335,
+      "step": 6931
+    },
+    {
+      "epoch": 3.080888888888889,
+      "grad_norm": 2.980158567428589,
+      "learning_rate": 7.683274021352314e-05,
+      "loss": 1.0148,
+      "step": 6932
+    },
+    {
+      "epoch": 3.0813333333333333,
+      "grad_norm": 3.9379446506500244,
+      "learning_rate": 7.681494661921708e-05,
+      "loss": 1.354,
+      "step": 6933
+    },
+    {
+      "epoch": 3.081777777777778,
+      "grad_norm": 3.313826560974121,
+      "learning_rate": 7.679715302491104e-05,
+      "loss": 1.5082,
+      "step": 6934
+    },
+    {
+      "epoch": 3.082222222222222,
+      "grad_norm": 2.866286039352417,
+      "learning_rate": 7.6779359430605e-05,
+      "loss": 0.7554,
+      "step": 6935
+    },
+    {
+      "epoch": 3.0826666666666664,
+      "grad_norm": 3.264652967453003,
+      "learning_rate": 7.676156583629894e-05,
+      "loss": 1.0634,
+      "step": 6936
+    },
+    {
+      "epoch": 3.083111111111111,
+      "grad_norm": 3.0582849979400635,
+      "learning_rate": 7.674377224199288e-05,
+      "loss": 0.9423,
+      "step": 6937
+    },
+    {
+      "epoch": 3.0835555555555554,
+      "grad_norm": 4.915375709533691,
+      "learning_rate": 7.672597864768684e-05,
+      "loss": 1.1836,
+      "step": 6938
+    },
+    {
+      "epoch": 3.084,
+      "grad_norm": 3.4640719890594482,
+      "learning_rate": 7.670818505338079e-05,
+      "loss": 1.2714,
+      "step": 6939
+    },
+    {
+      "epoch": 3.0844444444444443,
+      "grad_norm": 3.8669791221618652,
+      "learning_rate": 7.669039145907474e-05,
+      "loss": 1.0252,
+      "step": 6940
+    },
+    {
+      "epoch": 3.084888888888889,
+      "grad_norm": 4.150646686553955,
+      "learning_rate": 7.667259786476869e-05,
+      "loss": 1.432,
+      "step": 6941
+    },
+    {
+      "epoch": 3.0853333333333333,
+      "grad_norm": 3.675527334213257,
+      "learning_rate": 7.665480427046264e-05,
+      "loss": 1.2161,
+      "step": 6942
+    },
+    {
+      "epoch": 3.085777777777778,
+      "grad_norm": 4.122688293457031,
+      "learning_rate": 7.663701067615658e-05,
+      "loss": 1.2763,
+      "step": 6943
+    },
+    {
+      "epoch": 3.086222222222222,
+      "grad_norm": 3.758718967437744,
+      "learning_rate": 7.661921708185053e-05,
+      "loss": 1.2185,
+      "step": 6944
+    },
+    {
+      "epoch": 3.086666666666667,
+      "grad_norm": 4.356784343719482,
+      "learning_rate": 7.660142348754449e-05,
+      "loss": 1.4009,
+      "step": 6945
+    },
+    {
+      "epoch": 3.087111111111111,
+      "grad_norm": 4.016665458679199,
+      "learning_rate": 7.658362989323843e-05,
+      "loss": 0.9262,
+      "step": 6946
+    },
+    {
+      "epoch": 3.0875555555555554,
+      "grad_norm": 6.008724689483643,
+      "learning_rate": 7.656583629893239e-05,
+      "loss": 1.37,
+      "step": 6947
+    },
+    {
+      "epoch": 3.088,
+      "grad_norm": 5.219587326049805,
+      "learning_rate": 7.654804270462633e-05,
+      "loss": 1.1014,
+      "step": 6948
+    },
+    {
+      "epoch": 3.0884444444444443,
+      "grad_norm": 3.348543643951416,
+      "learning_rate": 7.653024911032029e-05,
+      "loss": 0.3617,
+      "step": 6949
+    },
+    {
+      "epoch": 3.088888888888889,
+      "grad_norm": 5.084573745727539,
+      "learning_rate": 7.651245551601423e-05,
+      "loss": 0.6737,
+      "step": 6950
+    },
+    {
+      "epoch": 3.0893333333333333,
+      "grad_norm": 2.369049072265625,
+      "learning_rate": 7.649466192170819e-05,
+      "loss": 1.7119,
+      "step": 6951
+    },
+    {
+      "epoch": 3.089777777777778,
+      "grad_norm": 2.740426540374756,
+      "learning_rate": 7.647686832740215e-05,
+      "loss": 1.6545,
+      "step": 6952
+    },
+    {
+      "epoch": 3.090222222222222,
+      "grad_norm": 2.695539712905884,
+      "learning_rate": 7.645907473309609e-05,
+      "loss": 1.7265,
+      "step": 6953
+    },
+    {
+      "epoch": 3.0906666666666665,
+      "grad_norm": 2.5847222805023193,
+      "learning_rate": 7.644128113879005e-05,
+      "loss": 1.1412,
+      "step": 6954
+    },
+    {
+      "epoch": 3.091111111111111,
+      "grad_norm": 2.5503435134887695,
+      "learning_rate": 7.642348754448399e-05,
+      "loss": 1.3825,
+      "step": 6955
+    },
+    {
+      "epoch": 3.0915555555555554,
+      "grad_norm": 3.02182674407959,
+      "learning_rate": 7.640569395017793e-05,
+      "loss": 1.3872,
+      "step": 6956
+    },
+    {
+      "epoch": 3.092,
+      "grad_norm": 2.3122494220733643,
+      "learning_rate": 7.638790035587189e-05,
+      "loss": 0.9554,
+      "step": 6957
+    },
+    {
+      "epoch": 3.0924444444444443,
+      "grad_norm": 2.8011865615844727,
+      "learning_rate": 7.637010676156584e-05,
+      "loss": 1.2296,
+      "step": 6958
+    },
+    {
+      "epoch": 3.092888888888889,
+      "grad_norm": 3.4765398502349854,
+      "learning_rate": 7.635231316725979e-05,
+      "loss": 2.2308,
+      "step": 6959
+    },
+    {
+      "epoch": 3.0933333333333333,
+      "grad_norm": 2.9357285499572754,
+      "learning_rate": 7.633451957295374e-05,
+      "loss": 1.394,
+      "step": 6960
+    },
+    {
+      "epoch": 3.093777777777778,
+      "grad_norm": 3.6758627891540527,
+      "learning_rate": 7.631672597864769e-05,
+      "loss": 1.7402,
+      "step": 6961
+    },
+    {
+      "epoch": 3.094222222222222,
+      "grad_norm": 3.1817829608917236,
+      "learning_rate": 7.629893238434164e-05,
+      "loss": 1.432,
+      "step": 6962
+    },
+    {
+      "epoch": 3.0946666666666665,
+      "grad_norm": 3.500157594680786,
+      "learning_rate": 7.628113879003559e-05,
+      "loss": 1.7906,
+      "step": 6963
+    },
+    {
+      "epoch": 3.095111111111111,
+      "grad_norm": 3.7627017498016357,
+      "learning_rate": 7.626334519572954e-05,
+      "loss": 1.7225,
+      "step": 6964
+    },
+    {
+      "epoch": 3.0955555555555554,
+      "grad_norm": 2.646817445755005,
+      "learning_rate": 7.62455516014235e-05,
+      "loss": 1.3691,
+      "step": 6965
+    },
+    {
+      "epoch": 3.096,
+      "grad_norm": 2.6752378940582275,
+      "learning_rate": 7.622775800711744e-05,
+      "loss": 1.4794,
+      "step": 6966
+    },
+    {
+      "epoch": 3.0964444444444443,
+      "grad_norm": 2.979374885559082,
+      "learning_rate": 7.62099644128114e-05,
+      "loss": 1.0324,
+      "step": 6967
+    },
+    {
+      "epoch": 3.096888888888889,
+      "grad_norm": 2.8177530765533447,
+      "learning_rate": 7.619217081850534e-05,
+      "loss": 1.5979,
+      "step": 6968
+    },
+    {
+      "epoch": 3.0973333333333333,
+      "grad_norm": 2.7737972736358643,
+      "learning_rate": 7.617437722419928e-05,
+      "loss": 1.138,
+      "step": 6969
+    },
+    {
+      "epoch": 3.097777777777778,
+      "grad_norm": 3.1239850521087646,
+      "learning_rate": 7.615658362989324e-05,
+      "loss": 1.4938,
+      "step": 6970
+    },
+    {
+      "epoch": 3.098222222222222,
+      "grad_norm": 3.2741284370422363,
+      "learning_rate": 7.61387900355872e-05,
+      "loss": 1.0675,
+      "step": 6971
+    },
+    {
+      "epoch": 3.0986666666666665,
+      "grad_norm": 2.790844678878784,
+      "learning_rate": 7.612099644128114e-05,
+      "loss": 1.2557,
+      "step": 6972
+    },
+    {
+      "epoch": 3.099111111111111,
+      "grad_norm": 2.8968873023986816,
+      "learning_rate": 7.61032028469751e-05,
+      "loss": 1.259,
+      "step": 6973
+    },
+    {
+      "epoch": 3.0995555555555554,
+      "grad_norm": 3.275770664215088,
+      "learning_rate": 7.608540925266904e-05,
+      "loss": 1.2327,
+      "step": 6974
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 3.203305721282959,
+      "learning_rate": 7.606761565836298e-05,
+      "loss": 0.9974,
+      "step": 6975
+    },
+    {
+      "epoch": 3.1004444444444443,
+      "grad_norm": 3.2165777683258057,
+      "learning_rate": 7.604982206405694e-05,
+      "loss": 1.6345,
+      "step": 6976
+    },
+    {
+      "epoch": 3.100888888888889,
+      "grad_norm": 3.0916032791137695,
+      "learning_rate": 7.60320284697509e-05,
+      "loss": 1.2841,
+      "step": 6977
+    },
+    {
+      "epoch": 3.1013333333333333,
+      "grad_norm": 3.5910959243774414,
+      "learning_rate": 7.601423487544484e-05,
+      "loss": 0.96,
+      "step": 6978
+    },
+    {
+      "epoch": 3.101777777777778,
+      "grad_norm": 3.80975341796875,
+      "learning_rate": 7.59964412811388e-05,
+      "loss": 1.5965,
+      "step": 6979
+    },
+    {
+      "epoch": 3.102222222222222,
+      "grad_norm": 2.8560242652893066,
+      "learning_rate": 7.597864768683275e-05,
+      "loss": 0.8661,
+      "step": 6980
+    },
+    {
+      "epoch": 3.1026666666666665,
+      "grad_norm": 4.064279079437256,
+      "learning_rate": 7.59608540925267e-05,
+      "loss": 1.3318,
+      "step": 6981
+    },
+    {
+      "epoch": 3.103111111111111,
+      "grad_norm": 3.475475311279297,
+      "learning_rate": 7.594306049822064e-05,
+      "loss": 0.9713,
+      "step": 6982
+    },
+    {
+      "epoch": 3.1035555555555554,
+      "grad_norm": 3.4135968685150146,
+      "learning_rate": 7.59252669039146e-05,
+      "loss": 1.1763,
+      "step": 6983
+    },
+    {
+      "epoch": 3.104,
+      "grad_norm": 2.9875054359436035,
+      "learning_rate": 7.590747330960855e-05,
+      "loss": 1.3583,
+      "step": 6984
+    },
+    {
+      "epoch": 3.1044444444444443,
+      "grad_norm": 3.5660154819488525,
+      "learning_rate": 7.58896797153025e-05,
+      "loss": 1.2142,
+      "step": 6985
+    },
+    {
+      "epoch": 3.104888888888889,
+      "grad_norm": 3.5056378841400146,
+      "learning_rate": 7.587188612099645e-05,
+      "loss": 1.3169,
+      "step": 6986
+    },
+    {
+      "epoch": 3.1053333333333333,
+      "grad_norm": 2.358790636062622,
+      "learning_rate": 7.58540925266904e-05,
+      "loss": 0.488,
+      "step": 6987
+    },
+    {
+      "epoch": 3.105777777777778,
+      "grad_norm": 3.0565037727355957,
+      "learning_rate": 7.583629893238434e-05,
+      "loss": 1.1014,
+      "step": 6988
+    },
+    {
+      "epoch": 3.106222222222222,
+      "grad_norm": 2.9908316135406494,
+      "learning_rate": 7.581850533807829e-05,
+      "loss": 1.0469,
+      "step": 6989
+    },
+    {
+      "epoch": 3.1066666666666665,
+      "grad_norm": 3.8466875553131104,
+      "learning_rate": 7.580071174377225e-05,
+      "loss": 0.9868,
+      "step": 6990
+    },
+    {
+      "epoch": 3.107111111111111,
+      "grad_norm": 4.6302032470703125,
+      "learning_rate": 7.578291814946619e-05,
+      "loss": 1.6431,
+      "step": 6991
+    },
+    {
+      "epoch": 3.1075555555555554,
+      "grad_norm": 3.230506420135498,
+      "learning_rate": 7.576512455516015e-05,
+      "loss": 0.8402,
+      "step": 6992
+    },
+    {
+      "epoch": 3.108,
+      "grad_norm": 3.8830463886260986,
+      "learning_rate": 7.57473309608541e-05,
+      "loss": 1.1773,
+      "step": 6993
+    },
+    {
+      "epoch": 3.1084444444444443,
+      "grad_norm": 4.235727310180664,
+      "learning_rate": 7.572953736654805e-05,
+      "loss": 1.2941,
+      "step": 6994
+    },
+    {
+      "epoch": 3.108888888888889,
+      "grad_norm": 4.56505012512207,
+      "learning_rate": 7.571174377224199e-05,
+      "loss": 1.1328,
+      "step": 6995
+    },
+    {
+      "epoch": 3.1093333333333333,
+      "grad_norm": 3.8152174949645996,
+      "learning_rate": 7.569395017793595e-05,
+      "loss": 0.8993,
+      "step": 6996
+    },
+    {
+      "epoch": 3.109777777777778,
+      "grad_norm": 5.836917877197266,
+      "learning_rate": 7.56761565836299e-05,
+      "loss": 1.6419,
+      "step": 6997
+    },
+    {
+      "epoch": 3.110222222222222,
+      "grad_norm": 3.6817240715026855,
+      "learning_rate": 7.565836298932385e-05,
+      "loss": 0.995,
+      "step": 6998
+    },
+    {
+      "epoch": 3.1106666666666665,
+      "grad_norm": 4.596709251403809,
+      "learning_rate": 7.56405693950178e-05,
+      "loss": 0.7939,
+      "step": 6999
+    },
+    {
+      "epoch": 3.111111111111111,
+      "grad_norm": 3.930978775024414,
+      "learning_rate": 7.562277580071175e-05,
+      "loss": 0.7911,
+      "step": 7000
+    },
+    {
+      "epoch": 3.1115555555555554,
+      "grad_norm": 2.6198768615722656,
+      "learning_rate": 7.560498220640569e-05,
+      "loss": 1.8534,
+      "step": 7001
+    },
+    {
+      "epoch": 3.112,
+      "grad_norm": 2.364030599594116,
+      "learning_rate": 7.558718861209965e-05,
+      "loss": 2.0188,
+      "step": 7002
+    },
+    {
+      "epoch": 3.1124444444444443,
+      "grad_norm": 2.803751230239868,
+      "learning_rate": 7.55693950177936e-05,
+      "loss": 1.405,
+      "step": 7003
+    },
+    {
+      "epoch": 3.112888888888889,
+      "grad_norm": 2.8711655139923096,
+      "learning_rate": 7.555160142348755e-05,
+      "loss": 1.3464,
+      "step": 7004
+    },
+    {
+      "epoch": 3.1133333333333333,
+      "grad_norm": 2.2702465057373047,
+      "learning_rate": 7.55338078291815e-05,
+      "loss": 1.1326,
+      "step": 7005
+    },
+    {
+      "epoch": 3.113777777777778,
+      "grad_norm": 3.279500961303711,
+      "learning_rate": 7.551601423487544e-05,
+      "loss": 1.7044,
+      "step": 7006
+    },
+    {
+      "epoch": 3.1142222222222222,
+      "grad_norm": 3.2445764541625977,
+      "learning_rate": 7.54982206405694e-05,
+      "loss": 1.5965,
+      "step": 7007
+    },
+    {
+      "epoch": 3.1146666666666665,
+      "grad_norm": 3.1852023601531982,
+      "learning_rate": 7.548042704626334e-05,
+      "loss": 1.7852,
+      "step": 7008
+    },
+    {
+      "epoch": 3.115111111111111,
+      "grad_norm": 2.400352716445923,
+      "learning_rate": 7.54626334519573e-05,
+      "loss": 0.6979,
+      "step": 7009
+    },
+    {
+      "epoch": 3.1155555555555554,
+      "grad_norm": 3.9273557662963867,
+      "learning_rate": 7.544483985765126e-05,
+      "loss": 1.7671,
+      "step": 7010
+    },
+    {
+      "epoch": 3.116,
+      "grad_norm": 3.297231674194336,
+      "learning_rate": 7.54270462633452e-05,
+      "loss": 1.6285,
+      "step": 7011
+    },
+    {
+      "epoch": 3.1164444444444444,
+      "grad_norm": 2.7438652515411377,
+      "learning_rate": 7.540925266903916e-05,
+      "loss": 1.2797,
+      "step": 7012
+    },
+    {
+      "epoch": 3.116888888888889,
+      "grad_norm": 2.7498419284820557,
+      "learning_rate": 7.53914590747331e-05,
+      "loss": 1.4858,
+      "step": 7013
+    },
+    {
+      "epoch": 3.1173333333333333,
+      "grad_norm": 2.6676368713378906,
+      "learning_rate": 7.537366548042704e-05,
+      "loss": 1.3337,
+      "step": 7014
+    },
+    {
+      "epoch": 3.117777777777778,
+      "grad_norm": 2.901026487350464,
+      "learning_rate": 7.5355871886121e-05,
+      "loss": 1.0484,
+      "step": 7015
+    },
+    {
+      "epoch": 3.1182222222222222,
+      "grad_norm": 2.0546693801879883,
+      "learning_rate": 7.533807829181496e-05,
+      "loss": 0.693,
+      "step": 7016
+    },
+    {
+      "epoch": 3.1186666666666665,
+      "grad_norm": 2.8888492584228516,
+      "learning_rate": 7.53202846975089e-05,
+      "loss": 1.1957,
+      "step": 7017
+    },
+    {
+      "epoch": 3.119111111111111,
+      "grad_norm": 3.6129257678985596,
+      "learning_rate": 7.530249110320285e-05,
+      "loss": 1.4137,
+      "step": 7018
+    },
+    {
+      "epoch": 3.1195555555555554,
+      "grad_norm": 2.8861212730407715,
+      "learning_rate": 7.52846975088968e-05,
+      "loss": 1.3716,
+      "step": 7019
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 2.638096809387207,
+      "learning_rate": 7.526690391459074e-05,
+      "loss": 1.1059,
+      "step": 7020
+    },
+    {
+      "epoch": 3.1204444444444444,
+      "grad_norm": 3.651658773422241,
+      "learning_rate": 7.52491103202847e-05,
+      "loss": 1.4312,
+      "step": 7021
+    },
+    {
+      "epoch": 3.120888888888889,
+      "grad_norm": 3.251763343811035,
+      "learning_rate": 7.523131672597865e-05,
+      "loss": 1.2436,
+      "step": 7022
+    },
+    {
+      "epoch": 3.1213333333333333,
+      "grad_norm": 3.0787482261657715,
+      "learning_rate": 7.52135231316726e-05,
+      "loss": 1.6314,
+      "step": 7023
+    },
+    {
+      "epoch": 3.121777777777778,
+      "grad_norm": 3.3005053997039795,
+      "learning_rate": 7.519572953736655e-05,
+      "loss": 1.3528,
+      "step": 7024
+    },
+    {
+      "epoch": 3.1222222222222222,
+      "grad_norm": 3.0210680961608887,
+      "learning_rate": 7.517793594306051e-05,
+      "loss": 1.028,
+      "step": 7025
+    },
+    {
+      "epoch": 3.1226666666666665,
+      "grad_norm": 3.53727126121521,
+      "learning_rate": 7.516014234875445e-05,
+      "loss": 1.2015,
+      "step": 7026
+    },
+    {
+      "epoch": 3.123111111111111,
+      "grad_norm": 3.3059957027435303,
+      "learning_rate": 7.51423487544484e-05,
+      "loss": 1.1474,
+      "step": 7027
+    },
+    {
+      "epoch": 3.1235555555555554,
+      "grad_norm": 3.240772008895874,
+      "learning_rate": 7.512455516014235e-05,
+      "loss": 1.1528,
+      "step": 7028
+    },
+    {
+      "epoch": 3.124,
+      "grad_norm": 2.7922959327697754,
+      "learning_rate": 7.510676156583631e-05,
+      "loss": 1.1426,
+      "step": 7029
+    },
+    {
+      "epoch": 3.1244444444444444,
+      "grad_norm": 3.379770278930664,
+      "learning_rate": 7.508896797153025e-05,
+      "loss": 1.3263,
+      "step": 7030
+    },
+    {
+      "epoch": 3.124888888888889,
+      "grad_norm": 2.729382038116455,
+      "learning_rate": 7.507117437722421e-05,
+      "loss": 1.1248,
+      "step": 7031
+    },
+    {
+      "epoch": 3.1253333333333333,
+      "grad_norm": 3.164997100830078,
+      "learning_rate": 7.505338078291815e-05,
+      "loss": 1.3626,
+      "step": 7032
+    },
+    {
+      "epoch": 3.1257777777777775,
+      "grad_norm": 2.980705499649048,
+      "learning_rate": 7.50355871886121e-05,
+      "loss": 1.0401,
+      "step": 7033
+    },
+    {
+      "epoch": 3.1262222222222222,
+      "grad_norm": 3.0273003578186035,
+      "learning_rate": 7.501779359430605e-05,
+      "loss": 1.0748,
+      "step": 7034
+    },
+    {
+      "epoch": 3.1266666666666665,
+      "grad_norm": 3.4305291175842285,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.38,
+      "step": 7035
+    },
+    {
+      "epoch": 3.127111111111111,
+      "grad_norm": 3.338956594467163,
+      "learning_rate": 7.498220640569395e-05,
+      "loss": 0.9866,
+      "step": 7036
+    },
+    {
+      "epoch": 3.1275555555555554,
+      "grad_norm": 3.5987491607666016,
+      "learning_rate": 7.49644128113879e-05,
+      "loss": 1.1284,
+      "step": 7037
+    },
+    {
+      "epoch": 3.128,
+      "grad_norm": 3.742375135421753,
+      "learning_rate": 7.494661921708186e-05,
+      "loss": 1.4609,
+      "step": 7038
+    },
+    {
+      "epoch": 3.1284444444444444,
+      "grad_norm": 3.7690203189849854,
+      "learning_rate": 7.49288256227758e-05,
+      "loss": 1.5164,
+      "step": 7039
+    },
+    {
+      "epoch": 3.128888888888889,
+      "grad_norm": 4.437491416931152,
+      "learning_rate": 7.491103202846975e-05,
+      "loss": 1.1423,
+      "step": 7040
+    },
+    {
+      "epoch": 3.1293333333333333,
+      "grad_norm": 3.674255609512329,
+      "learning_rate": 7.48932384341637e-05,
+      "loss": 1.3641,
+      "step": 7041
+    },
+    {
+      "epoch": 3.129777777777778,
+      "grad_norm": 3.8117153644561768,
+      "learning_rate": 7.487544483985766e-05,
+      "loss": 1.0165,
+      "step": 7042
+    },
+    {
+      "epoch": 3.1302222222222222,
+      "grad_norm": 3.8670129776000977,
+      "learning_rate": 7.48576512455516e-05,
+      "loss": 1.0165,
+      "step": 7043
+    },
+    {
+      "epoch": 3.1306666666666665,
+      "grad_norm": 3.9163060188293457,
+      "learning_rate": 7.483985765124556e-05,
+      "loss": 0.9119,
+      "step": 7044
+    },
+    {
+      "epoch": 3.131111111111111,
+      "grad_norm": 4.793590545654297,
+      "learning_rate": 7.48220640569395e-05,
+      "loss": 1.3223,
+      "step": 7045
+    },
+    {
+      "epoch": 3.1315555555555554,
+      "grad_norm": 5.119454383850098,
+      "learning_rate": 7.480427046263345e-05,
+      "loss": 1.1945,
+      "step": 7046
+    },
+    {
+      "epoch": 3.132,
+      "grad_norm": 4.798033714294434,
+      "learning_rate": 7.47864768683274e-05,
+      "loss": 1.407,
+      "step": 7047
+    },
+    {
+      "epoch": 3.1324444444444444,
+      "grad_norm": 2.8798067569732666,
+      "learning_rate": 7.476868327402136e-05,
+      "loss": 0.682,
+      "step": 7048
+    },
+    {
+      "epoch": 3.132888888888889,
+      "grad_norm": 2.7945594787597656,
+      "learning_rate": 7.47508896797153e-05,
+      "loss": 0.5815,
+      "step": 7049
+    },
+    {
+      "epoch": 3.1333333333333333,
+      "grad_norm": 4.523054122924805,
+      "learning_rate": 7.473309608540926e-05,
+      "loss": 0.7363,
+      "step": 7050
+    },
+    {
+      "epoch": 3.1337777777777776,
+      "grad_norm": 2.011254072189331,
+      "learning_rate": 7.471530249110322e-05,
+      "loss": 0.9446,
+      "step": 7051
+    },
+    {
+      "epoch": 3.1342222222222222,
+      "grad_norm": 1.6838473081588745,
+      "learning_rate": 7.469750889679716e-05,
+      "loss": 0.86,
+      "step": 7052
+    },
+    {
+      "epoch": 3.1346666666666665,
+      "grad_norm": 2.2713677883148193,
+      "learning_rate": 7.46797153024911e-05,
+      "loss": 1.7424,
+      "step": 7053
+    },
+    {
+      "epoch": 3.135111111111111,
+      "grad_norm": 2.4536499977111816,
+      "learning_rate": 7.466192170818506e-05,
+      "loss": 1.5837,
+      "step": 7054
+    },
+    {
+      "epoch": 3.1355555555555554,
+      "grad_norm": 1.8696529865264893,
+      "learning_rate": 7.464412811387901e-05,
+      "loss": 0.6377,
+      "step": 7055
+    },
+    {
+      "epoch": 3.136,
+      "grad_norm": 3.031843900680542,
+      "learning_rate": 7.462633451957296e-05,
+      "loss": 1.8678,
+      "step": 7056
+    },
+    {
+      "epoch": 3.1364444444444444,
+      "grad_norm": 2.593820333480835,
+      "learning_rate": 7.460854092526691e-05,
+      "loss": 1.4144,
+      "step": 7057
+    },
+    {
+      "epoch": 3.136888888888889,
+      "grad_norm": 2.432314872741699,
+      "learning_rate": 7.459074733096086e-05,
+      "loss": 1.5448,
+      "step": 7058
+    },
+    {
+      "epoch": 3.1373333333333333,
+      "grad_norm": 2.6551852226257324,
+      "learning_rate": 7.45729537366548e-05,
+      "loss": 1.3579,
+      "step": 7059
+    },
+    {
+      "epoch": 3.137777777777778,
+      "grad_norm": 3.1164588928222656,
+      "learning_rate": 7.455516014234876e-05,
+      "loss": 1.927,
+      "step": 7060
+    },
+    {
+      "epoch": 3.1382222222222222,
+      "grad_norm": 2.72015643119812,
+      "learning_rate": 7.453736654804271e-05,
+      "loss": 1.3657,
+      "step": 7061
+    },
+    {
+      "epoch": 3.1386666666666665,
+      "grad_norm": 3.068243980407715,
+      "learning_rate": 7.451957295373666e-05,
+      "loss": 1.4239,
+      "step": 7062
+    },
+    {
+      "epoch": 3.139111111111111,
+      "grad_norm": 2.426748514175415,
+      "learning_rate": 7.450177935943061e-05,
+      "loss": 1.1847,
+      "step": 7063
+    },
+    {
+      "epoch": 3.1395555555555554,
+      "grad_norm": 2.8654985427856445,
+      "learning_rate": 7.448398576512457e-05,
+      "loss": 1.5308,
+      "step": 7064
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 1.9712382555007935,
+      "learning_rate": 7.44661921708185e-05,
+      "loss": 0.7233,
+      "step": 7065
+    },
+    {
+      "epoch": 3.1404444444444444,
+      "grad_norm": 1.8847380876541138,
+      "learning_rate": 7.444839857651246e-05,
+      "loss": 0.7311,
+      "step": 7066
+    },
+    {
+      "epoch": 3.140888888888889,
+      "grad_norm": 2.814565658569336,
+      "learning_rate": 7.443060498220641e-05,
+      "loss": 1.1762,
+      "step": 7067
+    },
+    {
+      "epoch": 3.1413333333333333,
+      "grad_norm": 2.49889874458313,
+      "learning_rate": 7.441281138790035e-05,
+      "loss": 1.2439,
+      "step": 7068
+    },
+    {
+      "epoch": 3.1417777777777776,
+      "grad_norm": 3.1838126182556152,
+      "learning_rate": 7.439501779359431e-05,
+      "loss": 1.3337,
+      "step": 7069
+    },
+    {
+      "epoch": 3.1422222222222222,
+      "grad_norm": 2.6192221641540527,
+      "learning_rate": 7.437722419928827e-05,
+      "loss": 1.065,
+      "step": 7070
+    },
+    {
+      "epoch": 3.1426666666666665,
+      "grad_norm": 2.363414764404297,
+      "learning_rate": 7.435943060498221e-05,
+      "loss": 1.0256,
+      "step": 7071
+    },
+    {
+      "epoch": 3.143111111111111,
+      "grad_norm": 1.8646711111068726,
+      "learning_rate": 7.434163701067615e-05,
+      "loss": 0.6336,
+      "step": 7072
+    },
+    {
+      "epoch": 3.1435555555555554,
+      "grad_norm": 2.434340238571167,
+      "learning_rate": 7.432384341637011e-05,
+      "loss": 1.0433,
+      "step": 7073
+    },
+    {
+      "epoch": 3.144,
+      "grad_norm": 3.6076550483703613,
+      "learning_rate": 7.430604982206407e-05,
+      "loss": 1.1586,
+      "step": 7074
+    },
+    {
+      "epoch": 3.1444444444444444,
+      "grad_norm": 3.699741840362549,
+      "learning_rate": 7.428825622775801e-05,
+      "loss": 1.5315,
+      "step": 7075
+    },
+    {
+      "epoch": 3.144888888888889,
+      "grad_norm": 3.15507435798645,
+      "learning_rate": 7.427046263345197e-05,
+      "loss": 1.3789,
+      "step": 7076
+    },
+    {
+      "epoch": 3.1453333333333333,
+      "grad_norm": 2.9796347618103027,
+      "learning_rate": 7.425266903914591e-05,
+      "loss": 1.083,
+      "step": 7077
+    },
+    {
+      "epoch": 3.145777777777778,
+      "grad_norm": 2.9139842987060547,
+      "learning_rate": 7.423487544483985e-05,
+      "loss": 1.3829,
+      "step": 7078
+    },
+    {
+      "epoch": 3.1462222222222223,
+      "grad_norm": 3.6654200553894043,
+      "learning_rate": 7.421708185053381e-05,
+      "loss": 1.2362,
+      "step": 7079
+    },
+    {
+      "epoch": 3.1466666666666665,
+      "grad_norm": 3.540692090988159,
+      "learning_rate": 7.419928825622776e-05,
+      "loss": 1.4958,
+      "step": 7080
+    },
+    {
+      "epoch": 3.147111111111111,
+      "grad_norm": 3.061079740524292,
+      "learning_rate": 7.418149466192171e-05,
+      "loss": 1.317,
+      "step": 7081
+    },
+    {
+      "epoch": 3.1475555555555554,
+      "grad_norm": 3.567028522491455,
+      "learning_rate": 7.416370106761566e-05,
+      "loss": 1.3618,
+      "step": 7082
+    },
+    {
+      "epoch": 3.148,
+      "grad_norm": 3.034331798553467,
+      "learning_rate": 7.414590747330962e-05,
+      "loss": 0.9736,
+      "step": 7083
+    },
+    {
+      "epoch": 3.1484444444444444,
+      "grad_norm": 0.4216180741786957,
+      "learning_rate": 7.412811387900356e-05,
+      "loss": 0.0315,
+      "step": 7084
+    },
+    {
+      "epoch": 3.148888888888889,
+      "grad_norm": 2.6628007888793945,
+      "learning_rate": 7.41103202846975e-05,
+      "loss": 0.8623,
+      "step": 7085
+    },
+    {
+      "epoch": 3.1493333333333333,
+      "grad_norm": 3.0110785961151123,
+      "learning_rate": 7.409252669039146e-05,
+      "loss": 1.0985,
+      "step": 7086
+    },
+    {
+      "epoch": 3.1497777777777776,
+      "grad_norm": 3.624995231628418,
+      "learning_rate": 7.407473309608542e-05,
+      "loss": 1.1393,
+      "step": 7087
+    },
+    {
+      "epoch": 3.1502222222222223,
+      "grad_norm": 4.227591514587402,
+      "learning_rate": 7.405693950177936e-05,
+      "loss": 1.0868,
+      "step": 7088
+    },
+    {
+      "epoch": 3.1506666666666665,
+      "grad_norm": 3.4621124267578125,
+      "learning_rate": 7.403914590747332e-05,
+      "loss": 1.427,
+      "step": 7089
+    },
+    {
+      "epoch": 3.151111111111111,
+      "grad_norm": 3.3510584831237793,
+      "learning_rate": 7.402135231316726e-05,
+      "loss": 1.4212,
+      "step": 7090
+    },
+    {
+      "epoch": 3.1515555555555554,
+      "grad_norm": 3.500433921813965,
+      "learning_rate": 7.40035587188612e-05,
+      "loss": 1.1449,
+      "step": 7091
+    },
+    {
+      "epoch": 3.152,
+      "grad_norm": 2.601447582244873,
+      "learning_rate": 7.398576512455516e-05,
+      "loss": 0.8084,
+      "step": 7092
+    },
+    {
+      "epoch": 3.1524444444444444,
+      "grad_norm": 3.0802958011627197,
+      "learning_rate": 7.396797153024912e-05,
+      "loss": 1.1565,
+      "step": 7093
+    },
+    {
+      "epoch": 3.152888888888889,
+      "grad_norm": 3.7664129734039307,
+      "learning_rate": 7.395017793594306e-05,
+      "loss": 1.3433,
+      "step": 7094
+    },
+    {
+      "epoch": 3.1533333333333333,
+      "grad_norm": 3.885650157928467,
+      "learning_rate": 7.393238434163702e-05,
+      "loss": 1.2266,
+      "step": 7095
+    },
+    {
+      "epoch": 3.153777777777778,
+      "grad_norm": 5.502748489379883,
+      "learning_rate": 7.391459074733097e-05,
+      "loss": 1.8502,
+      "step": 7096
+    },
+    {
+      "epoch": 3.1542222222222223,
+      "grad_norm": 3.670057535171509,
+      "learning_rate": 7.389679715302492e-05,
+      "loss": 1.4684,
+      "step": 7097
+    },
+    {
+      "epoch": 3.1546666666666665,
+      "grad_norm": 4.47036075592041,
+      "learning_rate": 7.387900355871886e-05,
+      "loss": 1.183,
+      "step": 7098
+    },
+    {
+      "epoch": 3.155111111111111,
+      "grad_norm": 0.5041842460632324,
+      "learning_rate": 7.386120996441282e-05,
+      "loss": 0.0461,
+      "step": 7099
+    },
+    {
+      "epoch": 3.1555555555555554,
+      "grad_norm": 3.492530107498169,
+      "learning_rate": 7.384341637010677e-05,
+      "loss": 0.8281,
+      "step": 7100
+    },
+    {
+      "epoch": 3.156,
+      "grad_norm": 2.3296926021575928,
+      "learning_rate": 7.382562277580072e-05,
+      "loss": 1.9542,
+      "step": 7101
+    },
+    {
+      "epoch": 3.1564444444444444,
+      "grad_norm": 2.73773455619812,
+      "learning_rate": 7.380782918149467e-05,
+      "loss": 1.7659,
+      "step": 7102
+    },
+    {
+      "epoch": 3.156888888888889,
+      "grad_norm": 2.8689823150634766,
+      "learning_rate": 7.379003558718862e-05,
+      "loss": 1.7431,
+      "step": 7103
+    },
+    {
+      "epoch": 3.1573333333333333,
+      "grad_norm": 2.7786591053009033,
+      "learning_rate": 7.377224199288256e-05,
+      "loss": 1.6562,
+      "step": 7104
+    },
+    {
+      "epoch": 3.1577777777777776,
+      "grad_norm": 3.037637710571289,
+      "learning_rate": 7.375444839857651e-05,
+      "loss": 1.746,
+      "step": 7105
+    },
+    {
+      "epoch": 3.1582222222222223,
+      "grad_norm": 3.2228243350982666,
+      "learning_rate": 7.373665480427047e-05,
+      "loss": 1.8074,
+      "step": 7106
+    },
+    {
+      "epoch": 3.1586666666666665,
+      "grad_norm": 2.8026838302612305,
+      "learning_rate": 7.371886120996441e-05,
+      "loss": 1.786,
+      "step": 7107
+    },
+    {
+      "epoch": 3.159111111111111,
+      "grad_norm": 3.4075815677642822,
+      "learning_rate": 7.370106761565837e-05,
+      "loss": 1.6417,
+      "step": 7108
+    },
+    {
+      "epoch": 3.1595555555555555,
+      "grad_norm": 2.9315574169158936,
+      "learning_rate": 7.368327402135233e-05,
+      "loss": 1.3821,
+      "step": 7109
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.2183048725128174,
+      "learning_rate": 7.366548042704626e-05,
+      "loss": 1.3795,
+      "step": 7110
+    },
+    {
+      "epoch": 3.1604444444444444,
+      "grad_norm": 2.97540020942688,
+      "learning_rate": 7.364768683274021e-05,
+      "loss": 1.6864,
+      "step": 7111
+    },
+    {
+      "epoch": 3.160888888888889,
+      "grad_norm": 3.3829736709594727,
+      "learning_rate": 7.362989323843417e-05,
+      "loss": 1.5294,
+      "step": 7112
+    },
+    {
+      "epoch": 3.1613333333333333,
+      "grad_norm": 3.1063239574432373,
+      "learning_rate": 7.361209964412811e-05,
+      "loss": 1.451,
+      "step": 7113
+    },
+    {
+      "epoch": 3.1617777777777776,
+      "grad_norm": 3.2091407775878906,
+      "learning_rate": 7.359430604982207e-05,
+      "loss": 1.1878,
+      "step": 7114
+    },
+    {
+      "epoch": 3.1622222222222223,
+      "grad_norm": 3.398361921310425,
+      "learning_rate": 7.357651245551603e-05,
+      "loss": 1.5108,
+      "step": 7115
+    },
+    {
+      "epoch": 3.1626666666666665,
+      "grad_norm": 3.1675875186920166,
+      "learning_rate": 7.355871886120997e-05,
+      "loss": 1.294,
+      "step": 7116
+    },
+    {
+      "epoch": 3.163111111111111,
+      "grad_norm": 2.810655117034912,
+      "learning_rate": 7.354092526690391e-05,
+      "loss": 1.2984,
+      "step": 7117
+    },
+    {
+      "epoch": 3.1635555555555555,
+      "grad_norm": 3.2484841346740723,
+      "learning_rate": 7.352313167259787e-05,
+      "loss": 1.1205,
+      "step": 7118
+    },
+    {
+      "epoch": 3.164,
+      "grad_norm": 3.3352835178375244,
+      "learning_rate": 7.350533807829182e-05,
+      "loss": 1.5339,
+      "step": 7119
+    },
+    {
+      "epoch": 3.1644444444444444,
+      "grad_norm": 3.3871214389801025,
+      "learning_rate": 7.348754448398577e-05,
+      "loss": 1.3406,
+      "step": 7120
+    },
+    {
+      "epoch": 3.164888888888889,
+      "grad_norm": 3.104323148727417,
+      "learning_rate": 7.346975088967972e-05,
+      "loss": 1.636,
+      "step": 7121
+    },
+    {
+      "epoch": 3.1653333333333333,
+      "grad_norm": 2.902137041091919,
+      "learning_rate": 7.345195729537368e-05,
+      "loss": 1.4295,
+      "step": 7122
+    },
+    {
+      "epoch": 3.1657777777777776,
+      "grad_norm": 2.7861342430114746,
+      "learning_rate": 7.343416370106761e-05,
+      "loss": 0.9665,
+      "step": 7123
+    },
+    {
+      "epoch": 3.1662222222222223,
+      "grad_norm": 2.9004876613616943,
+      "learning_rate": 7.341637010676157e-05,
+      "loss": 1.2588,
+      "step": 7124
+    },
+    {
+      "epoch": 3.1666666666666665,
+      "grad_norm": 2.671241521835327,
+      "learning_rate": 7.339857651245552e-05,
+      "loss": 1.2189,
+      "step": 7125
+    },
+    {
+      "epoch": 3.167111111111111,
+      "grad_norm": 3.0016653537750244,
+      "learning_rate": 7.338078291814947e-05,
+      "loss": 0.9757,
+      "step": 7126
+    },
+    {
+      "epoch": 3.1675555555555555,
+      "grad_norm": 3.0266857147216797,
+      "learning_rate": 7.336298932384342e-05,
+      "loss": 1.2218,
+      "step": 7127
+    },
+    {
+      "epoch": 3.168,
+      "grad_norm": 2.868429183959961,
+      "learning_rate": 7.334519572953738e-05,
+      "loss": 0.8355,
+      "step": 7128
+    },
+    {
+      "epoch": 3.1684444444444444,
+      "grad_norm": 3.6476175785064697,
+      "learning_rate": 7.332740213523132e-05,
+      "loss": 0.864,
+      "step": 7129
+    },
+    {
+      "epoch": 3.168888888888889,
+      "grad_norm": 3.347186803817749,
+      "learning_rate": 7.330960854092526e-05,
+      "loss": 1.1723,
+      "step": 7130
+    },
+    {
+      "epoch": 3.1693333333333333,
+      "grad_norm": 3.4044735431671143,
+      "learning_rate": 7.329181494661922e-05,
+      "loss": 1.0311,
+      "step": 7131
+    },
+    {
+      "epoch": 3.1697777777777776,
+      "grad_norm": 2.955648422241211,
+      "learning_rate": 7.327402135231318e-05,
+      "loss": 1.1082,
+      "step": 7132
+    },
+    {
+      "epoch": 3.1702222222222223,
+      "grad_norm": 3.7140543460845947,
+      "learning_rate": 7.325622775800712e-05,
+      "loss": 0.9562,
+      "step": 7133
+    },
+    {
+      "epoch": 3.1706666666666665,
+      "grad_norm": 2.378147840499878,
+      "learning_rate": 7.323843416370108e-05,
+      "loss": 0.697,
+      "step": 7134
+    },
+    {
+      "epoch": 3.171111111111111,
+      "grad_norm": 3.5638556480407715,
+      "learning_rate": 7.322064056939502e-05,
+      "loss": 1.2156,
+      "step": 7135
+    },
+    {
+      "epoch": 3.1715555555555555,
+      "grad_norm": 3.8616840839385986,
+      "learning_rate": 7.320284697508896e-05,
+      "loss": 0.9394,
+      "step": 7136
+    },
+    {
+      "epoch": 3.172,
+      "grad_norm": 3.537442684173584,
+      "learning_rate": 7.318505338078292e-05,
+      "loss": 1.3647,
+      "step": 7137
+    },
+    {
+      "epoch": 3.1724444444444444,
+      "grad_norm": 3.341238260269165,
+      "learning_rate": 7.316725978647688e-05,
+      "loss": 1.0796,
+      "step": 7138
+    },
+    {
+      "epoch": 3.172888888888889,
+      "grad_norm": 3.2897703647613525,
+      "learning_rate": 7.314946619217082e-05,
+      "loss": 0.9163,
+      "step": 7139
+    },
+    {
+      "epoch": 3.1733333333333333,
+      "grad_norm": 3.147260904312134,
+      "learning_rate": 7.313167259786478e-05,
+      "loss": 0.9676,
+      "step": 7140
+    },
+    {
+      "epoch": 3.1737777777777776,
+      "grad_norm": 4.187511444091797,
+      "learning_rate": 7.311387900355873e-05,
+      "loss": 1.3926,
+      "step": 7141
+    },
+    {
+      "epoch": 3.1742222222222223,
+      "grad_norm": 4.7306809425354,
+      "learning_rate": 7.309608540925267e-05,
+      "loss": 1.3427,
+      "step": 7142
+    },
+    {
+      "epoch": 3.1746666666666665,
+      "grad_norm": 3.0592243671417236,
+      "learning_rate": 7.307829181494662e-05,
+      "loss": 0.7532,
+      "step": 7143
+    },
+    {
+      "epoch": 3.175111111111111,
+      "grad_norm": 4.022464275360107,
+      "learning_rate": 7.306049822064057e-05,
+      "loss": 1.3008,
+      "step": 7144
+    },
+    {
+      "epoch": 3.1755555555555555,
+      "grad_norm": 4.482253551483154,
+      "learning_rate": 7.304270462633453e-05,
+      "loss": 1.1332,
+      "step": 7145
+    },
+    {
+      "epoch": 3.176,
+      "grad_norm": 3.72058367729187,
+      "learning_rate": 7.302491103202847e-05,
+      "loss": 0.7426,
+      "step": 7146
+    },
+    {
+      "epoch": 3.1764444444444444,
+      "grad_norm": 4.921838760375977,
+      "learning_rate": 7.300711743772243e-05,
+      "loss": 1.0258,
+      "step": 7147
+    },
+    {
+      "epoch": 3.176888888888889,
+      "grad_norm": 4.758439540863037,
+      "learning_rate": 7.298932384341637e-05,
+      "loss": 1.3546,
+      "step": 7148
+    },
+    {
+      "epoch": 3.1773333333333333,
+      "grad_norm": 3.1380600929260254,
+      "learning_rate": 7.297153024911032e-05,
+      "loss": 1.0218,
+      "step": 7149
+    },
+    {
+      "epoch": 3.1777777777777776,
+      "grad_norm": 7.19141149520874,
+      "learning_rate": 7.295373665480427e-05,
+      "loss": 1.2103,
+      "step": 7150
+    },
+    {
+      "epoch": 3.1782222222222223,
+      "grad_norm": 1.9877086877822876,
+      "learning_rate": 7.293594306049823e-05,
+      "loss": 1.8951,
+      "step": 7151
+    },
+    {
+      "epoch": 3.1786666666666665,
+      "grad_norm": 2.509326934814453,
+      "learning_rate": 7.291814946619217e-05,
+      "loss": 1.9882,
+      "step": 7152
+    },
+    {
+      "epoch": 3.179111111111111,
+      "grad_norm": 2.56880521774292,
+      "learning_rate": 7.290035587188613e-05,
+      "loss": 1.8781,
+      "step": 7153
+    },
+    {
+      "epoch": 3.1795555555555555,
+      "grad_norm": 2.692734718322754,
+      "learning_rate": 7.288256227758008e-05,
+      "loss": 1.9204,
+      "step": 7154
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.635511875152588,
+      "learning_rate": 7.286476868327401e-05,
+      "loss": 1.1913,
+      "step": 7155
+    },
+    {
+      "epoch": 3.1804444444444444,
+      "grad_norm": 2.8718338012695312,
+      "learning_rate": 7.284697508896797e-05,
+      "loss": 1.432,
+      "step": 7156
+    },
+    {
+      "epoch": 3.180888888888889,
+      "grad_norm": 3.3866138458251953,
+      "learning_rate": 7.282918149466193e-05,
+      "loss": 1.7453,
+      "step": 7157
+    },
+    {
+      "epoch": 3.1813333333333333,
+      "grad_norm": 3.44844913482666,
+      "learning_rate": 7.281138790035587e-05,
+      "loss": 1.3587,
+      "step": 7158
+    },
+    {
+      "epoch": 3.1817777777777776,
+      "grad_norm": 2.8075125217437744,
+      "learning_rate": 7.279359430604983e-05,
+      "loss": 1.5954,
+      "step": 7159
+    },
+    {
+      "epoch": 3.1822222222222223,
+      "grad_norm": 2.6318461894989014,
+      "learning_rate": 7.277580071174378e-05,
+      "loss": 1.2991,
+      "step": 7160
+    },
+    {
+      "epoch": 3.1826666666666665,
+      "grad_norm": 2.9282736778259277,
+      "learning_rate": 7.275800711743773e-05,
+      "loss": 1.5801,
+      "step": 7161
+    },
+    {
+      "epoch": 3.1831111111111112,
+      "grad_norm": 2.870149850845337,
+      "learning_rate": 7.274021352313167e-05,
+      "loss": 0.8409,
+      "step": 7162
+    },
+    {
+      "epoch": 3.1835555555555555,
+      "grad_norm": 2.717829942703247,
+      "learning_rate": 7.272241992882563e-05,
+      "loss": 0.994,
+      "step": 7163
+    },
+    {
+      "epoch": 3.184,
+      "grad_norm": 2.9092955589294434,
+      "learning_rate": 7.270462633451958e-05,
+      "loss": 1.6491,
+      "step": 7164
+    },
+    {
+      "epoch": 3.1844444444444444,
+      "grad_norm": 2.3932883739471436,
+      "learning_rate": 7.268683274021352e-05,
+      "loss": 0.8707,
+      "step": 7165
+    },
+    {
+      "epoch": 3.1848888888888887,
+      "grad_norm": 3.078213691711426,
+      "learning_rate": 7.266903914590748e-05,
+      "loss": 1.0287,
+      "step": 7166
+    },
+    {
+      "epoch": 3.1853333333333333,
+      "grad_norm": 3.1520655155181885,
+      "learning_rate": 7.265124555160144e-05,
+      "loss": 1.1343,
+      "step": 7167
+    },
+    {
+      "epoch": 3.1857777777777776,
+      "grad_norm": 3.593202829360962,
+      "learning_rate": 7.263345195729537e-05,
+      "loss": 1.276,
+      "step": 7168
+    },
+    {
+      "epoch": 3.1862222222222223,
+      "grad_norm": 2.6163017749786377,
+      "learning_rate": 7.261565836298932e-05,
+      "loss": 1.226,
+      "step": 7169
+    },
+    {
+      "epoch": 3.1866666666666665,
+      "grad_norm": 2.836480140686035,
+      "learning_rate": 7.259786476868328e-05,
+      "loss": 1.2427,
+      "step": 7170
+    },
+    {
+      "epoch": 3.1871111111111112,
+      "grad_norm": 3.4099297523498535,
+      "learning_rate": 7.258007117437722e-05,
+      "loss": 1.4251,
+      "step": 7171
+    },
+    {
+      "epoch": 3.1875555555555555,
+      "grad_norm": 2.73268461227417,
+      "learning_rate": 7.256227758007118e-05,
+      "loss": 1.0855,
+      "step": 7172
+    },
+    {
+      "epoch": 3.188,
+      "grad_norm": 2.9737985134124756,
+      "learning_rate": 7.254448398576514e-05,
+      "loss": 1.451,
+      "step": 7173
+    },
+    {
+      "epoch": 3.1884444444444444,
+      "grad_norm": 3.1273388862609863,
+      "learning_rate": 7.252669039145908e-05,
+      "loss": 1.6291,
+      "step": 7174
+    },
+    {
+      "epoch": 3.188888888888889,
+      "grad_norm": 3.3678598403930664,
+      "learning_rate": 7.250889679715302e-05,
+      "loss": 1.6578,
+      "step": 7175
+    },
+    {
+      "epoch": 3.1893333333333334,
+      "grad_norm": 3.497072458267212,
+      "learning_rate": 7.249110320284698e-05,
+      "loss": 1.4176,
+      "step": 7176
+    },
+    {
+      "epoch": 3.1897777777777776,
+      "grad_norm": 3.735827922821045,
+      "learning_rate": 7.247330960854094e-05,
+      "loss": 1.5746,
+      "step": 7177
+    },
+    {
+      "epoch": 3.1902222222222223,
+      "grad_norm": 2.853653907775879,
+      "learning_rate": 7.245551601423488e-05,
+      "loss": 1.1089,
+      "step": 7178
+    },
+    {
+      "epoch": 3.1906666666666665,
+      "grad_norm": 3.3315930366516113,
+      "learning_rate": 7.243772241992883e-05,
+      "loss": 1.1311,
+      "step": 7179
+    },
+    {
+      "epoch": 3.1911111111111112,
+      "grad_norm": 2.9672279357910156,
+      "learning_rate": 7.241992882562279e-05,
+      "loss": 1.3171,
+      "step": 7180
+    },
+    {
+      "epoch": 3.1915555555555555,
+      "grad_norm": 3.4035391807556152,
+      "learning_rate": 7.240213523131672e-05,
+      "loss": 0.9667,
+      "step": 7181
+    },
+    {
+      "epoch": 3.192,
+      "grad_norm": 3.72255539894104,
+      "learning_rate": 7.238434163701068e-05,
+      "loss": 1.36,
+      "step": 7182
+    },
+    {
+      "epoch": 3.1924444444444444,
+      "grad_norm": 3.1913297176361084,
+      "learning_rate": 7.236654804270463e-05,
+      "loss": 1.0758,
+      "step": 7183
+    },
+    {
+      "epoch": 3.1928888888888887,
+      "grad_norm": 2.7058324813842773,
+      "learning_rate": 7.234875444839858e-05,
+      "loss": 0.6656,
+      "step": 7184
+    },
+    {
+      "epoch": 3.1933333333333334,
+      "grad_norm": 3.929481029510498,
+      "learning_rate": 7.233096085409253e-05,
+      "loss": 1.6081,
+      "step": 7185
+    },
+    {
+      "epoch": 3.1937777777777776,
+      "grad_norm": 3.4291040897369385,
+      "learning_rate": 7.231316725978649e-05,
+      "loss": 1.363,
+      "step": 7186
+    },
+    {
+      "epoch": 3.1942222222222223,
+      "grad_norm": 3.8611419200897217,
+      "learning_rate": 7.229537366548043e-05,
+      "loss": 1.6286,
+      "step": 7187
+    },
+    {
+      "epoch": 3.1946666666666665,
+      "grad_norm": 2.4565515518188477,
+      "learning_rate": 7.227758007117438e-05,
+      "loss": 0.4925,
+      "step": 7188
+    },
+    {
+      "epoch": 3.1951111111111112,
+      "grad_norm": 2.0368804931640625,
+      "learning_rate": 7.225978647686833e-05,
+      "loss": 0.5928,
+      "step": 7189
+    },
+    {
+      "epoch": 3.1955555555555555,
+      "grad_norm": 3.8311848640441895,
+      "learning_rate": 7.224199288256229e-05,
+      "loss": 1.3479,
+      "step": 7190
+    },
+    {
+      "epoch": 3.196,
+      "grad_norm": 4.187136173248291,
+      "learning_rate": 7.222419928825623e-05,
+      "loss": 0.9687,
+      "step": 7191
+    },
+    {
+      "epoch": 3.1964444444444444,
+      "grad_norm": 3.8171749114990234,
+      "learning_rate": 7.220640569395019e-05,
+      "loss": 1.265,
+      "step": 7192
+    },
+    {
+      "epoch": 3.196888888888889,
+      "grad_norm": 3.7234129905700684,
+      "learning_rate": 7.218861209964413e-05,
+      "loss": 1.1282,
+      "step": 7193
+    },
+    {
+      "epoch": 3.1973333333333334,
+      "grad_norm": 4.388239860534668,
+      "learning_rate": 7.217081850533807e-05,
+      "loss": 1.2207,
+      "step": 7194
+    },
+    {
+      "epoch": 3.1977777777777776,
+      "grad_norm": 3.327991247177124,
+      "learning_rate": 7.215302491103203e-05,
+      "loss": 1.1965,
+      "step": 7195
+    },
+    {
+      "epoch": 3.1982222222222223,
+      "grad_norm": 4.21957540512085,
+      "learning_rate": 7.213523131672599e-05,
+      "loss": 1.4889,
+      "step": 7196
+    },
+    {
+      "epoch": 3.1986666666666665,
+      "grad_norm": 4.019267559051514,
+      "learning_rate": 7.211743772241993e-05,
+      "loss": 1.3072,
+      "step": 7197
+    },
+    {
+      "epoch": 3.1991111111111112,
+      "grad_norm": 3.848752975463867,
+      "learning_rate": 7.209964412811389e-05,
+      "loss": 1.0561,
+      "step": 7198
+    },
+    {
+      "epoch": 3.1995555555555555,
+      "grad_norm": 3.2341089248657227,
+      "learning_rate": 7.208185053380784e-05,
+      "loss": 0.545,
+      "step": 7199
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.097386598587036,
+      "learning_rate": 7.206405693950177e-05,
+      "loss": 0.2964,
+      "step": 7200
+    },
+    {
+      "epoch": 3.2004444444444444,
+      "grad_norm": 2.824582099914551,
+      "learning_rate": 7.204626334519573e-05,
+      "loss": 1.753,
+      "step": 7201
+    },
+    {
+      "epoch": 3.2008888888888887,
+      "grad_norm": 2.1013436317443848,
+      "learning_rate": 7.202846975088968e-05,
+      "loss": 0.606,
+      "step": 7202
+    },
+    {
+      "epoch": 3.2013333333333334,
+      "grad_norm": 2.8657641410827637,
+      "learning_rate": 7.201067615658363e-05,
+      "loss": 1.8523,
+      "step": 7203
+    },
+    {
+      "epoch": 3.2017777777777776,
+      "grad_norm": 2.5109121799468994,
+      "learning_rate": 7.199288256227758e-05,
+      "loss": 1.1317,
+      "step": 7204
+    },
+    {
+      "epoch": 3.2022222222222223,
+      "grad_norm": 3.1287527084350586,
+      "learning_rate": 7.197508896797154e-05,
+      "loss": 1.8141,
+      "step": 7205
+    },
+    {
+      "epoch": 3.2026666666666666,
+      "grad_norm": 3.054079294204712,
+      "learning_rate": 7.195729537366548e-05,
+      "loss": 1.8931,
+      "step": 7206
+    },
+    {
+      "epoch": 3.2031111111111112,
+      "grad_norm": 3.2487199306488037,
+      "learning_rate": 7.193950177935943e-05,
+      "loss": 1.4648,
+      "step": 7207
+    },
+    {
+      "epoch": 3.2035555555555555,
+      "grad_norm": 2.855591058731079,
+      "learning_rate": 7.192170818505338e-05,
+      "loss": 1.2971,
+      "step": 7208
+    },
+    {
+      "epoch": 3.204,
+      "grad_norm": 2.5365679264068604,
+      "learning_rate": 7.190391459074734e-05,
+      "loss": 1.2072,
+      "step": 7209
+    },
+    {
+      "epoch": 3.2044444444444444,
+      "grad_norm": 3.0599007606506348,
+      "learning_rate": 7.188612099644128e-05,
+      "loss": 1.6128,
+      "step": 7210
+    },
+    {
+      "epoch": 3.204888888888889,
+      "grad_norm": 3.139268159866333,
+      "learning_rate": 7.186832740213524e-05,
+      "loss": 1.2958,
+      "step": 7211
+    },
+    {
+      "epoch": 3.2053333333333334,
+      "grad_norm": 3.227274179458618,
+      "learning_rate": 7.18505338078292e-05,
+      "loss": 1.4077,
+      "step": 7212
+    },
+    {
+      "epoch": 3.2057777777777776,
+      "grad_norm": 3.3886468410491943,
+      "learning_rate": 7.183274021352313e-05,
+      "loss": 1.5784,
+      "step": 7213
+    },
+    {
+      "epoch": 3.2062222222222223,
+      "grad_norm": 3.387749195098877,
+      "learning_rate": 7.181494661921708e-05,
+      "loss": 1.4535,
+      "step": 7214
+    },
+    {
+      "epoch": 3.2066666666666666,
+      "grad_norm": 2.8726046085357666,
+      "learning_rate": 7.179715302491104e-05,
+      "loss": 1.1744,
+      "step": 7215
+    },
+    {
+      "epoch": 3.2071111111111112,
+      "grad_norm": 2.2621724605560303,
+      "learning_rate": 7.177935943060498e-05,
+      "loss": 0.6728,
+      "step": 7216
+    },
+    {
+      "epoch": 3.2075555555555555,
+      "grad_norm": 3.217412233352661,
+      "learning_rate": 7.176156583629894e-05,
+      "loss": 1.6073,
+      "step": 7217
+    },
+    {
+      "epoch": 3.208,
+      "grad_norm": 2.7585723400115967,
+      "learning_rate": 7.17437722419929e-05,
+      "loss": 0.9378,
+      "step": 7218
+    },
+    {
+      "epoch": 3.2084444444444444,
+      "grad_norm": 3.4076976776123047,
+      "learning_rate": 7.172597864768684e-05,
+      "loss": 1.3989,
+      "step": 7219
+    },
+    {
+      "epoch": 3.2088888888888887,
+      "grad_norm": 2.9216151237487793,
+      "learning_rate": 7.170818505338078e-05,
+      "loss": 1.206,
+      "step": 7220
+    },
+    {
+      "epoch": 3.2093333333333334,
+      "grad_norm": 3.7456247806549072,
+      "learning_rate": 7.169039145907474e-05,
+      "loss": 1.8443,
+      "step": 7221
+    },
+    {
+      "epoch": 3.2097777777777776,
+      "grad_norm": 2.7865076065063477,
+      "learning_rate": 7.167259786476869e-05,
+      "loss": 0.9977,
+      "step": 7222
+    },
+    {
+      "epoch": 3.2102222222222223,
+      "grad_norm": 3.0865001678466797,
+      "learning_rate": 7.165480427046264e-05,
+      "loss": 1.3247,
+      "step": 7223
+    },
+    {
+      "epoch": 3.2106666666666666,
+      "grad_norm": 3.52168607711792,
+      "learning_rate": 7.163701067615659e-05,
+      "loss": 1.0324,
+      "step": 7224
+    },
+    {
+      "epoch": 3.2111111111111112,
+      "grad_norm": 4.151998519897461,
+      "learning_rate": 7.161921708185055e-05,
+      "loss": 1.5058,
+      "step": 7225
+    },
+    {
+      "epoch": 3.2115555555555555,
+      "grad_norm": 3.7610089778900146,
+      "learning_rate": 7.160142348754448e-05,
+      "loss": 1.8216,
+      "step": 7226
+    },
+    {
+      "epoch": 3.212,
+      "grad_norm": 2.9817423820495605,
+      "learning_rate": 7.158362989323843e-05,
+      "loss": 1.3767,
+      "step": 7227
+    },
+    {
+      "epoch": 3.2124444444444444,
+      "grad_norm": 2.8566582202911377,
+      "learning_rate": 7.156583629893239e-05,
+      "loss": 0.9962,
+      "step": 7228
+    },
+    {
+      "epoch": 3.2128888888888887,
+      "grad_norm": 2.9715373516082764,
+      "learning_rate": 7.154804270462633e-05,
+      "loss": 1.4952,
+      "step": 7229
+    },
+    {
+      "epoch": 3.2133333333333334,
+      "grad_norm": 3.730404853820801,
+      "learning_rate": 7.153024911032029e-05,
+      "loss": 1.5204,
+      "step": 7230
+    },
+    {
+      "epoch": 3.2137777777777776,
+      "grad_norm": 3.011878490447998,
+      "learning_rate": 7.151245551601425e-05,
+      "loss": 1.1504,
+      "step": 7231
+    },
+    {
+      "epoch": 3.2142222222222223,
+      "grad_norm": 3.1973698139190674,
+      "learning_rate": 7.149466192170819e-05,
+      "loss": 0.9828,
+      "step": 7232
+    },
+    {
+      "epoch": 3.2146666666666666,
+      "grad_norm": 3.3064098358154297,
+      "learning_rate": 7.147686832740213e-05,
+      "loss": 1.5142,
+      "step": 7233
+    },
+    {
+      "epoch": 3.2151111111111113,
+      "grad_norm": 2.9979686737060547,
+      "learning_rate": 7.145907473309609e-05,
+      "loss": 1.1844,
+      "step": 7234
+    },
+    {
+      "epoch": 3.2155555555555555,
+      "grad_norm": 3.717355489730835,
+      "learning_rate": 7.144128113879005e-05,
+      "loss": 1.7117,
+      "step": 7235
+    },
+    {
+      "epoch": 3.216,
+      "grad_norm": 4.831557273864746,
+      "learning_rate": 7.142348754448399e-05,
+      "loss": 1.553,
+      "step": 7236
+    },
+    {
+      "epoch": 3.2164444444444444,
+      "grad_norm": 2.6884803771972656,
+      "learning_rate": 7.140569395017795e-05,
+      "loss": 0.7039,
+      "step": 7237
+    },
+    {
+      "epoch": 3.2168888888888887,
+      "grad_norm": 3.3653738498687744,
+      "learning_rate": 7.138790035587189e-05,
+      "loss": 1.144,
+      "step": 7238
+    },
+    {
+      "epoch": 3.2173333333333334,
+      "grad_norm": 3.712017774581909,
+      "learning_rate": 7.137010676156583e-05,
+      "loss": 1.2988,
+      "step": 7239
+    },
+    {
+      "epoch": 3.2177777777777776,
+      "grad_norm": 4.0276360511779785,
+      "learning_rate": 7.135231316725979e-05,
+      "loss": 1.0672,
+      "step": 7240
+    },
+    {
+      "epoch": 3.2182222222222223,
+      "grad_norm": 3.8651421070098877,
+      "learning_rate": 7.133451957295374e-05,
+      "loss": 1.5012,
+      "step": 7241
+    },
+    {
+      "epoch": 3.2186666666666666,
+      "grad_norm": 3.2855637073516846,
+      "learning_rate": 7.131672597864769e-05,
+      "loss": 0.9562,
+      "step": 7242
+    },
+    {
+      "epoch": 3.2191111111111113,
+      "grad_norm": 3.601529598236084,
+      "learning_rate": 7.129893238434164e-05,
+      "loss": 1.2631,
+      "step": 7243
+    },
+    {
+      "epoch": 3.2195555555555555,
+      "grad_norm": 3.80122709274292,
+      "learning_rate": 7.12811387900356e-05,
+      "loss": 1.2177,
+      "step": 7244
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 4.019015789031982,
+      "learning_rate": 7.126334519572953e-05,
+      "loss": 1.2818,
+      "step": 7245
+    },
+    {
+      "epoch": 3.2204444444444444,
+      "grad_norm": 3.7049124240875244,
+      "learning_rate": 7.124555160142349e-05,
+      "loss": 0.9796,
+      "step": 7246
+    },
+    {
+      "epoch": 3.2208888888888887,
+      "grad_norm": 4.283298969268799,
+      "learning_rate": 7.122775800711744e-05,
+      "loss": 1.1318,
+      "step": 7247
+    },
+    {
+      "epoch": 3.2213333333333334,
+      "grad_norm": 5.531714916229248,
+      "learning_rate": 7.120996441281139e-05,
+      "loss": 1.4307,
+      "step": 7248
+    },
+    {
+      "epoch": 3.2217777777777776,
+      "grad_norm": 3.2561047077178955,
+      "learning_rate": 7.119217081850534e-05,
+      "loss": 1.09,
+      "step": 7249
+    },
+    {
+      "epoch": 3.2222222222222223,
+      "grad_norm": 4.089896202087402,
+      "learning_rate": 7.11743772241993e-05,
+      "loss": 0.5095,
+      "step": 7250
+    },
+    {
+      "epoch": 3.2226666666666666,
+      "grad_norm": 2.2933907508850098,
+      "learning_rate": 7.115658362989324e-05,
+      "loss": 2.2543,
+      "step": 7251
+    },
+    {
+      "epoch": 3.2231111111111113,
+      "grad_norm": 2.559863567352295,
+      "learning_rate": 7.113879003558718e-05,
+      "loss": 1.9739,
+      "step": 7252
+    },
+    {
+      "epoch": 3.2235555555555555,
+      "grad_norm": 1.9117332696914673,
+      "learning_rate": 7.112099644128114e-05,
+      "loss": 0.9344,
+      "step": 7253
+    },
+    {
+      "epoch": 3.224,
+      "grad_norm": 2.9356801509857178,
+      "learning_rate": 7.11032028469751e-05,
+      "loss": 1.5221,
+      "step": 7254
+    },
+    {
+      "epoch": 3.2244444444444444,
+      "grad_norm": 2.627389430999756,
+      "learning_rate": 7.108540925266904e-05,
+      "loss": 1.4828,
+      "step": 7255
+    },
+    {
+      "epoch": 3.2248888888888887,
+      "grad_norm": 3.1114463806152344,
+      "learning_rate": 7.1067615658363e-05,
+      "loss": 1.4934,
+      "step": 7256
+    },
+    {
+      "epoch": 3.2253333333333334,
+      "grad_norm": 2.5716917514801025,
+      "learning_rate": 7.104982206405695e-05,
+      "loss": 1.0995,
+      "step": 7257
+    },
+    {
+      "epoch": 3.2257777777777776,
+      "grad_norm": 2.8480474948883057,
+      "learning_rate": 7.103202846975088e-05,
+      "loss": 1.7053,
+      "step": 7258
+    },
+    {
+      "epoch": 3.2262222222222223,
+      "grad_norm": 2.725034475326538,
+      "learning_rate": 7.101423487544484e-05,
+      "loss": 1.1406,
+      "step": 7259
+    },
+    {
+      "epoch": 3.2266666666666666,
+      "grad_norm": 2.4350764751434326,
+      "learning_rate": 7.09964412811388e-05,
+      "loss": 1.2509,
+      "step": 7260
+    },
+    {
+      "epoch": 3.2271111111111113,
+      "grad_norm": 3.65389347076416,
+      "learning_rate": 7.097864768683274e-05,
+      "loss": 2.3713,
+      "step": 7261
+    },
+    {
+      "epoch": 3.2275555555555555,
+      "grad_norm": 3.542847156524658,
+      "learning_rate": 7.09608540925267e-05,
+      "loss": 1.5322,
+      "step": 7262
+    },
+    {
+      "epoch": 3.228,
+      "grad_norm": 2.748013496398926,
+      "learning_rate": 7.094306049822065e-05,
+      "loss": 1.3331,
+      "step": 7263
+    },
+    {
+      "epoch": 3.2284444444444444,
+      "grad_norm": 3.1858341693878174,
+      "learning_rate": 7.09252669039146e-05,
+      "loss": 1.7167,
+      "step": 7264
+    },
+    {
+      "epoch": 3.2288888888888887,
+      "grad_norm": 2.994159698486328,
+      "learning_rate": 7.090747330960854e-05,
+      "loss": 1.0373,
+      "step": 7265
+    },
+    {
+      "epoch": 3.2293333333333334,
+      "grad_norm": 2.560145616531372,
+      "learning_rate": 7.08896797153025e-05,
+      "loss": 0.9071,
+      "step": 7266
+    },
+    {
+      "epoch": 3.2297777777777776,
+      "grad_norm": 3.167236566543579,
+      "learning_rate": 7.087188612099645e-05,
+      "loss": 1.317,
+      "step": 7267
+    },
+    {
+      "epoch": 3.2302222222222223,
+      "grad_norm": 2.4366395473480225,
+      "learning_rate": 7.08540925266904e-05,
+      "loss": 0.8397,
+      "step": 7268
+    },
+    {
+      "epoch": 3.2306666666666666,
+      "grad_norm": 3.0970094203948975,
+      "learning_rate": 7.083629893238435e-05,
+      "loss": 1.2824,
+      "step": 7269
+    },
+    {
+      "epoch": 3.2311111111111113,
+      "grad_norm": 3.605058431625366,
+      "learning_rate": 7.08185053380783e-05,
+      "loss": 1.4352,
+      "step": 7270
+    },
+    {
+      "epoch": 3.2315555555555555,
+      "grad_norm": 3.437645673751831,
+      "learning_rate": 7.080071174377224e-05,
+      "loss": 1.4561,
+      "step": 7271
+    },
+    {
+      "epoch": 3.232,
+      "grad_norm": 2.9223389625549316,
+      "learning_rate": 7.078291814946619e-05,
+      "loss": 1.146,
+      "step": 7272
+    },
+    {
+      "epoch": 3.2324444444444445,
+      "grad_norm": 4.0127129554748535,
+      "learning_rate": 7.076512455516015e-05,
+      "loss": 1.5512,
+      "step": 7273
+    },
+    {
+      "epoch": 3.2328888888888887,
+      "grad_norm": 3.0993196964263916,
+      "learning_rate": 7.074733096085409e-05,
+      "loss": 1.178,
+      "step": 7274
+    },
+    {
+      "epoch": 3.2333333333333334,
+      "grad_norm": 3.2388551235198975,
+      "learning_rate": 7.072953736654805e-05,
+      "loss": 1.1,
+      "step": 7275
+    },
+    {
+      "epoch": 3.2337777777777776,
+      "grad_norm": 3.486222505569458,
+      "learning_rate": 7.0711743772242e-05,
+      "loss": 1.5717,
+      "step": 7276
+    },
+    {
+      "epoch": 3.2342222222222223,
+      "grad_norm": 3.0996615886688232,
+      "learning_rate": 7.069395017793595e-05,
+      "loss": 1.1912,
+      "step": 7277
+    },
+    {
+      "epoch": 3.2346666666666666,
+      "grad_norm": 3.7221884727478027,
+      "learning_rate": 7.067615658362989e-05,
+      "loss": 1.3382,
+      "step": 7278
+    },
+    {
+      "epoch": 3.2351111111111113,
+      "grad_norm": 2.7222108840942383,
+      "learning_rate": 7.065836298932385e-05,
+      "loss": 1.0546,
+      "step": 7279
+    },
+    {
+      "epoch": 3.2355555555555555,
+      "grad_norm": 3.1601808071136475,
+      "learning_rate": 7.06405693950178e-05,
+      "loss": 1.0409,
+      "step": 7280
+    },
+    {
+      "epoch": 3.2359999999999998,
+      "grad_norm": 3.1560659408569336,
+      "learning_rate": 7.062277580071175e-05,
+      "loss": 1.3843,
+      "step": 7281
+    },
+    {
+      "epoch": 3.2364444444444445,
+      "grad_norm": 2.998133897781372,
+      "learning_rate": 7.06049822064057e-05,
+      "loss": 1.227,
+      "step": 7282
+    },
+    {
+      "epoch": 3.2368888888888887,
+      "grad_norm": 2.981759548187256,
+      "learning_rate": 7.058718861209965e-05,
+      "loss": 1.0669,
+      "step": 7283
+    },
+    {
+      "epoch": 3.2373333333333334,
+      "grad_norm": 4.712173938751221,
+      "learning_rate": 7.056939501779359e-05,
+      "loss": 1.5214,
+      "step": 7284
+    },
+    {
+      "epoch": 3.2377777777777776,
+      "grad_norm": 3.807121515274048,
+      "learning_rate": 7.055160142348755e-05,
+      "loss": 1.1862,
+      "step": 7285
+    },
+    {
+      "epoch": 3.2382222222222223,
+      "grad_norm": 3.189521312713623,
+      "learning_rate": 7.05338078291815e-05,
+      "loss": 1.0769,
+      "step": 7286
+    },
+    {
+      "epoch": 3.2386666666666666,
+      "grad_norm": 3.0515081882476807,
+      "learning_rate": 7.051601423487545e-05,
+      "loss": 0.9761,
+      "step": 7287
+    },
+    {
+      "epoch": 3.2391111111111113,
+      "grad_norm": 3.679802894592285,
+      "learning_rate": 7.04982206405694e-05,
+      "loss": 1.1306,
+      "step": 7288
+    },
+    {
+      "epoch": 3.2395555555555555,
+      "grad_norm": 3.8492064476013184,
+      "learning_rate": 7.048042704626336e-05,
+      "loss": 1.0187,
+      "step": 7289
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 4.128783226013184,
+      "learning_rate": 7.046263345195729e-05,
+      "loss": 1.1335,
+      "step": 7290
+    },
+    {
+      "epoch": 3.2404444444444445,
+      "grad_norm": 3.4628500938415527,
+      "learning_rate": 7.044483985765124e-05,
+      "loss": 1.4144,
+      "step": 7291
+    },
+    {
+      "epoch": 3.2408888888888887,
+      "grad_norm": 3.7098937034606934,
+      "learning_rate": 7.04270462633452e-05,
+      "loss": 1.0089,
+      "step": 7292
+    },
+    {
+      "epoch": 3.2413333333333334,
+      "grad_norm": 3.4453952312469482,
+      "learning_rate": 7.040925266903914e-05,
+      "loss": 1.1287,
+      "step": 7293
+    },
+    {
+      "epoch": 3.2417777777777776,
+      "grad_norm": 4.36504602432251,
+      "learning_rate": 7.03914590747331e-05,
+      "loss": 1.435,
+      "step": 7294
+    },
+    {
+      "epoch": 3.2422222222222223,
+      "grad_norm": 4.6257476806640625,
+      "learning_rate": 7.037366548042706e-05,
+      "loss": 1.1206,
+      "step": 7295
+    },
+    {
+      "epoch": 3.2426666666666666,
+      "grad_norm": 5.652115821838379,
+      "learning_rate": 7.0355871886121e-05,
+      "loss": 1.1406,
+      "step": 7296
+    },
+    {
+      "epoch": 3.2431111111111113,
+      "grad_norm": 3.9133806228637695,
+      "learning_rate": 7.033807829181494e-05,
+      "loss": 0.8279,
+      "step": 7297
+    },
+    {
+      "epoch": 3.2435555555555555,
+      "grad_norm": 5.016626834869385,
+      "learning_rate": 7.03202846975089e-05,
+      "loss": 1.7871,
+      "step": 7298
+    },
+    {
+      "epoch": 3.2439999999999998,
+      "grad_norm": 2.0724196434020996,
+      "learning_rate": 7.030249110320286e-05,
+      "loss": 0.4856,
+      "step": 7299
+    },
+    {
+      "epoch": 3.2444444444444445,
+      "grad_norm": 5.151663303375244,
+      "learning_rate": 7.02846975088968e-05,
+      "loss": 1.195,
+      "step": 7300
+    },
+    {
+      "epoch": 3.2448888888888887,
+      "grad_norm": 1.3501733541488647,
+      "learning_rate": 7.026690391459075e-05,
+      "loss": 0.0231,
+      "step": 7301
+    },
+    {
+      "epoch": 3.2453333333333334,
+      "grad_norm": 2.257030963897705,
+      "learning_rate": 7.024911032028471e-05,
+      "loss": 1.4925,
+      "step": 7302
+    },
+    {
+      "epoch": 3.2457777777777777,
+      "grad_norm": 2.3157708644866943,
+      "learning_rate": 7.023131672597864e-05,
+      "loss": 1.3873,
+      "step": 7303
+    },
+    {
+      "epoch": 3.2462222222222223,
+      "grad_norm": 2.4795522689819336,
+      "learning_rate": 7.02135231316726e-05,
+      "loss": 1.7858,
+      "step": 7304
+    },
+    {
+      "epoch": 3.2466666666666666,
+      "grad_norm": 3.000107765197754,
+      "learning_rate": 7.019572953736655e-05,
+      "loss": 2.1163,
+      "step": 7305
+    },
+    {
+      "epoch": 3.2471111111111113,
+      "grad_norm": 3.0592353343963623,
+      "learning_rate": 7.01779359430605e-05,
+      "loss": 1.615,
+      "step": 7306
+    },
+    {
+      "epoch": 3.2475555555555555,
+      "grad_norm": 3.055565357208252,
+      "learning_rate": 7.016014234875445e-05,
+      "loss": 1.4699,
+      "step": 7307
+    },
+    {
+      "epoch": 3.248,
+      "grad_norm": 2.8485829830169678,
+      "learning_rate": 7.014234875444841e-05,
+      "loss": 1.61,
+      "step": 7308
+    },
+    {
+      "epoch": 3.2484444444444445,
+      "grad_norm": 0.6298081278800964,
+      "learning_rate": 7.012455516014235e-05,
+      "loss": 0.0241,
+      "step": 7309
+    },
+    {
+      "epoch": 3.2488888888888887,
+      "grad_norm": 3.0662970542907715,
+      "learning_rate": 7.01067615658363e-05,
+      "loss": 1.7241,
+      "step": 7310
+    },
+    {
+      "epoch": 3.2493333333333334,
+      "grad_norm": 4.107135772705078,
+      "learning_rate": 7.008896797153025e-05,
+      "loss": 2.0917,
+      "step": 7311
+    },
+    {
+      "epoch": 3.2497777777777777,
+      "grad_norm": 3.6882572174072266,
+      "learning_rate": 7.007117437722421e-05,
+      "loss": 1.5678,
+      "step": 7312
+    },
+    {
+      "epoch": 3.2502222222222223,
+      "grad_norm": 2.740238904953003,
+      "learning_rate": 7.005338078291815e-05,
+      "loss": 1.3232,
+      "step": 7313
+    },
+    {
+      "epoch": 3.2506666666666666,
+      "grad_norm": 3.1968138217926025,
+      "learning_rate": 7.003558718861211e-05,
+      "loss": 1.3232,
+      "step": 7314
+    },
+    {
+      "epoch": 3.2511111111111113,
+      "grad_norm": 3.4555680751800537,
+      "learning_rate": 7.001779359430606e-05,
+      "loss": 1.6254,
+      "step": 7315
+    },
+    {
+      "epoch": 3.2515555555555555,
+      "grad_norm": 3.412783145904541,
+      "learning_rate": 7e-05,
+      "loss": 1.671,
+      "step": 7316
+    },
+    {
+      "epoch": 3.252,
+      "grad_norm": 3.239433765411377,
+      "learning_rate": 6.998220640569395e-05,
+      "loss": 1.4749,
+      "step": 7317
+    },
+    {
+      "epoch": 3.2524444444444445,
+      "grad_norm": 2.9741363525390625,
+      "learning_rate": 6.996441281138791e-05,
+      "loss": 0.8289,
+      "step": 7318
+    },
+    {
+      "epoch": 3.2528888888888887,
+      "grad_norm": 3.045731782913208,
+      "learning_rate": 6.994661921708185e-05,
+      "loss": 1.6571,
+      "step": 7319
+    },
+    {
+      "epoch": 3.2533333333333334,
+      "grad_norm": 3.0729551315307617,
+      "learning_rate": 6.99288256227758e-05,
+      "loss": 1.4405,
+      "step": 7320
+    },
+    {
+      "epoch": 3.2537777777777777,
+      "grad_norm": 2.7762842178344727,
+      "learning_rate": 6.991103202846976e-05,
+      "loss": 1.1644,
+      "step": 7321
+    },
+    {
+      "epoch": 3.2542222222222223,
+      "grad_norm": 2.8989222049713135,
+      "learning_rate": 6.98932384341637e-05,
+      "loss": 1.2091,
+      "step": 7322
+    },
+    {
+      "epoch": 3.2546666666666666,
+      "grad_norm": 3.4842469692230225,
+      "learning_rate": 6.987544483985765e-05,
+      "loss": 1.489,
+      "step": 7323
+    },
+    {
+      "epoch": 3.2551111111111113,
+      "grad_norm": 3.5377140045166016,
+      "learning_rate": 6.98576512455516e-05,
+      "loss": 0.9816,
+      "step": 7324
+    },
+    {
+      "epoch": 3.2555555555555555,
+      "grad_norm": 3.346153974533081,
+      "learning_rate": 6.983985765124555e-05,
+      "loss": 1.6473,
+      "step": 7325
+    },
+    {
+      "epoch": 3.2560000000000002,
+      "grad_norm": 3.840273380279541,
+      "learning_rate": 6.98220640569395e-05,
+      "loss": 1.509,
+      "step": 7326
+    },
+    {
+      "epoch": 3.2564444444444445,
+      "grad_norm": 3.4088332653045654,
+      "learning_rate": 6.980427046263346e-05,
+      "loss": 1.1997,
+      "step": 7327
+    },
+    {
+      "epoch": 3.2568888888888887,
+      "grad_norm": 2.9059813022613525,
+      "learning_rate": 6.97864768683274e-05,
+      "loss": 1.133,
+      "step": 7328
+    },
+    {
+      "epoch": 3.2573333333333334,
+      "grad_norm": 3.1663708686828613,
+      "learning_rate": 6.976868327402135e-05,
+      "loss": 1.2817,
+      "step": 7329
+    },
+    {
+      "epoch": 3.2577777777777777,
+      "grad_norm": 3.088623523712158,
+      "learning_rate": 6.97508896797153e-05,
+      "loss": 1.142,
+      "step": 7330
+    },
+    {
+      "epoch": 3.2582222222222224,
+      "grad_norm": 2.9888076782226562,
+      "learning_rate": 6.973309608540926e-05,
+      "loss": 0.8191,
+      "step": 7331
+    },
+    {
+      "epoch": 3.2586666666666666,
+      "grad_norm": 3.0160036087036133,
+      "learning_rate": 6.97153024911032e-05,
+      "loss": 0.8145,
+      "step": 7332
+    },
+    {
+      "epoch": 3.2591111111111113,
+      "grad_norm": 4.1365885734558105,
+      "learning_rate": 6.969750889679716e-05,
+      "loss": 1.2396,
+      "step": 7333
+    },
+    {
+      "epoch": 3.2595555555555555,
+      "grad_norm": 4.00795316696167,
+      "learning_rate": 6.967971530249112e-05,
+      "loss": 1.7597,
+      "step": 7334
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 3.9610869884490967,
+      "learning_rate": 6.966192170818505e-05,
+      "loss": 1.4981,
+      "step": 7335
+    },
+    {
+      "epoch": 3.2604444444444445,
+      "grad_norm": 2.927777051925659,
+      "learning_rate": 6.9644128113879e-05,
+      "loss": 0.9722,
+      "step": 7336
+    },
+    {
+      "epoch": 3.2608888888888887,
+      "grad_norm": 3.2167749404907227,
+      "learning_rate": 6.962633451957296e-05,
+      "loss": 1.1545,
+      "step": 7337
+    },
+    {
+      "epoch": 3.2613333333333334,
+      "grad_norm": 3.5486435890197754,
+      "learning_rate": 6.96085409252669e-05,
+      "loss": 1.2762,
+      "step": 7338
+    },
+    {
+      "epoch": 3.2617777777777777,
+      "grad_norm": 2.870244026184082,
+      "learning_rate": 6.959074733096086e-05,
+      "loss": 0.8667,
+      "step": 7339
+    },
+    {
+      "epoch": 3.2622222222222224,
+      "grad_norm": 3.2309038639068604,
+      "learning_rate": 6.957295373665481e-05,
+      "loss": 1.0299,
+      "step": 7340
+    },
+    {
+      "epoch": 3.2626666666666666,
+      "grad_norm": 3.7159862518310547,
+      "learning_rate": 6.955516014234876e-05,
+      "loss": 1.2253,
+      "step": 7341
+    },
+    {
+      "epoch": 3.2631111111111113,
+      "grad_norm": 4.002220630645752,
+      "learning_rate": 6.95373665480427e-05,
+      "loss": 1.0,
+      "step": 7342
+    },
+    {
+      "epoch": 3.2635555555555555,
+      "grad_norm": 3.580116033554077,
+      "learning_rate": 6.951957295373666e-05,
+      "loss": 1.1418,
+      "step": 7343
+    },
+    {
+      "epoch": 3.2640000000000002,
+      "grad_norm": 3.863196849822998,
+      "learning_rate": 6.950177935943061e-05,
+      "loss": 1.2002,
+      "step": 7344
+    },
+    {
+      "epoch": 3.2644444444444445,
+      "grad_norm": 3.9584481716156006,
+      "learning_rate": 6.948398576512456e-05,
+      "loss": 1.1874,
+      "step": 7345
+    },
+    {
+      "epoch": 3.2648888888888887,
+      "grad_norm": 4.6558403968811035,
+      "learning_rate": 6.946619217081851e-05,
+      "loss": 1.1067,
+      "step": 7346
+    },
+    {
+      "epoch": 3.2653333333333334,
+      "grad_norm": 5.399629592895508,
+      "learning_rate": 6.944839857651247e-05,
+      "loss": 1.3862,
+      "step": 7347
+    },
+    {
+      "epoch": 3.2657777777777777,
+      "grad_norm": 4.66273307800293,
+      "learning_rate": 6.94306049822064e-05,
+      "loss": 1.7355,
+      "step": 7348
+    },
+    {
+      "epoch": 3.2662222222222224,
+      "grad_norm": 4.567107200622559,
+      "learning_rate": 6.941281138790035e-05,
+      "loss": 1.4238,
+      "step": 7349
+    },
+    {
+      "epoch": 3.2666666666666666,
+      "grad_norm": 3.7617597579956055,
+      "learning_rate": 6.939501779359431e-05,
+      "loss": 0.5293,
+      "step": 7350
+    },
+    {
+      "epoch": 3.2671111111111113,
+      "grad_norm": 2.0919718742370605,
+      "learning_rate": 6.937722419928825e-05,
+      "loss": 0.9802,
+      "step": 7351
+    },
+    {
+      "epoch": 3.2675555555555555,
+      "grad_norm": 2.003411293029785,
+      "learning_rate": 6.935943060498221e-05,
+      "loss": 0.7958,
+      "step": 7352
+    },
+    {
+      "epoch": 3.268,
+      "grad_norm": 2.802001714706421,
+      "learning_rate": 6.934163701067617e-05,
+      "loss": 1.5791,
+      "step": 7353
+    },
+    {
+      "epoch": 3.2684444444444445,
+      "grad_norm": 2.7242465019226074,
+      "learning_rate": 6.932384341637011e-05,
+      "loss": 1.2121,
+      "step": 7354
+    },
+    {
+      "epoch": 3.2688888888888887,
+      "grad_norm": 3.243849039077759,
+      "learning_rate": 6.930604982206405e-05,
+      "loss": 1.4581,
+      "step": 7355
+    },
+    {
+      "epoch": 3.2693333333333334,
+      "grad_norm": 2.92507004737854,
+      "learning_rate": 6.928825622775801e-05,
+      "loss": 0.9443,
+      "step": 7356
+    },
+    {
+      "epoch": 3.2697777777777777,
+      "grad_norm": 3.716792345046997,
+      "learning_rate": 6.927046263345197e-05,
+      "loss": 1.8721,
+      "step": 7357
+    },
+    {
+      "epoch": 3.2702222222222224,
+      "grad_norm": 3.27911639213562,
+      "learning_rate": 6.925266903914591e-05,
+      "loss": 1.8725,
+      "step": 7358
+    },
+    {
+      "epoch": 3.2706666666666666,
+      "grad_norm": 2.785362720489502,
+      "learning_rate": 6.923487544483987e-05,
+      "loss": 1.0283,
+      "step": 7359
+    },
+    {
+      "epoch": 3.2711111111111113,
+      "grad_norm": 3.401243209838867,
+      "learning_rate": 6.921708185053382e-05,
+      "loss": 1.4464,
+      "step": 7360
+    },
+    {
+      "epoch": 3.2715555555555556,
+      "grad_norm": 3.113215923309326,
+      "learning_rate": 6.919928825622775e-05,
+      "loss": 1.6867,
+      "step": 7361
+    },
+    {
+      "epoch": 3.2720000000000002,
+      "grad_norm": 3.3423285484313965,
+      "learning_rate": 6.918149466192171e-05,
+      "loss": 1.2452,
+      "step": 7362
+    },
+    {
+      "epoch": 3.2724444444444445,
+      "grad_norm": 2.914412021636963,
+      "learning_rate": 6.916370106761566e-05,
+      "loss": 1.0074,
+      "step": 7363
+    },
+    {
+      "epoch": 3.2728888888888887,
+      "grad_norm": 2.851064920425415,
+      "learning_rate": 6.914590747330961e-05,
+      "loss": 1.2674,
+      "step": 7364
+    },
+    {
+      "epoch": 3.2733333333333334,
+      "grad_norm": 3.0711374282836914,
+      "learning_rate": 6.912811387900356e-05,
+      "loss": 1.3246,
+      "step": 7365
+    },
+    {
+      "epoch": 3.2737777777777777,
+      "grad_norm": 3.009274959564209,
+      "learning_rate": 6.911032028469752e-05,
+      "loss": 1.3831,
+      "step": 7366
+    },
+    {
+      "epoch": 3.2742222222222224,
+      "grad_norm": 3.4454495906829834,
+      "learning_rate": 6.909252669039146e-05,
+      "loss": 1.5029,
+      "step": 7367
+    },
+    {
+      "epoch": 3.2746666666666666,
+      "grad_norm": 3.192911386489868,
+      "learning_rate": 6.90747330960854e-05,
+      "loss": 1.3722,
+      "step": 7368
+    },
+    {
+      "epoch": 3.2751111111111113,
+      "grad_norm": 3.413494110107422,
+      "learning_rate": 6.905693950177936e-05,
+      "loss": 1.3843,
+      "step": 7369
+    },
+    {
+      "epoch": 3.2755555555555556,
+      "grad_norm": 2.5494213104248047,
+      "learning_rate": 6.90391459074733e-05,
+      "loss": 0.9536,
+      "step": 7370
+    },
+    {
+      "epoch": 3.276,
+      "grad_norm": 3.192929267883301,
+      "learning_rate": 6.902135231316726e-05,
+      "loss": 1.687,
+      "step": 7371
+    },
+    {
+      "epoch": 3.2764444444444445,
+      "grad_norm": 3.70212459564209,
+      "learning_rate": 6.900355871886122e-05,
+      "loss": 1.5884,
+      "step": 7372
+    },
+    {
+      "epoch": 3.2768888888888887,
+      "grad_norm": 3.3133480548858643,
+      "learning_rate": 6.898576512455516e-05,
+      "loss": 1.5349,
+      "step": 7373
+    },
+    {
+      "epoch": 3.2773333333333334,
+      "grad_norm": 3.146245241165161,
+      "learning_rate": 6.89679715302491e-05,
+      "loss": 1.259,
+      "step": 7374
+    },
+    {
+      "epoch": 3.2777777777777777,
+      "grad_norm": 3.329066753387451,
+      "learning_rate": 6.895017793594306e-05,
+      "loss": 1.4778,
+      "step": 7375
+    },
+    {
+      "epoch": 3.2782222222222224,
+      "grad_norm": 3.4649317264556885,
+      "learning_rate": 6.893238434163702e-05,
+      "loss": 1.4338,
+      "step": 7376
+    },
+    {
+      "epoch": 3.2786666666666666,
+      "grad_norm": 3.8039920330047607,
+      "learning_rate": 6.891459074733096e-05,
+      "loss": 1.6371,
+      "step": 7377
+    },
+    {
+      "epoch": 3.279111111111111,
+      "grad_norm": 3.138587474822998,
+      "learning_rate": 6.889679715302492e-05,
+      "loss": 1.2055,
+      "step": 7378
+    },
+    {
+      "epoch": 3.2795555555555556,
+      "grad_norm": 3.8496856689453125,
+      "learning_rate": 6.887900355871887e-05,
+      "loss": 1.1513,
+      "step": 7379
+    },
+    {
+      "epoch": 3.2800000000000002,
+      "grad_norm": 1.128354787826538,
+      "learning_rate": 6.88612099644128e-05,
+      "loss": 0.039,
+      "step": 7380
+    },
+    {
+      "epoch": 3.2804444444444445,
+      "grad_norm": 3.4761149883270264,
+      "learning_rate": 6.884341637010676e-05,
+      "loss": 1.6024,
+      "step": 7381
+    },
+    {
+      "epoch": 3.2808888888888887,
+      "grad_norm": 3.339864492416382,
+      "learning_rate": 6.882562277580072e-05,
+      "loss": 1.3504,
+      "step": 7382
+    },
+    {
+      "epoch": 3.2813333333333334,
+      "grad_norm": 2.745142936706543,
+      "learning_rate": 6.880782918149466e-05,
+      "loss": 0.8555,
+      "step": 7383
+    },
+    {
+      "epoch": 3.2817777777777777,
+      "grad_norm": 4.043299198150635,
+      "learning_rate": 6.879003558718862e-05,
+      "loss": 1.5035,
+      "step": 7384
+    },
+    {
+      "epoch": 3.2822222222222224,
+      "grad_norm": 3.773738384246826,
+      "learning_rate": 6.877224199288257e-05,
+      "loss": 1.1627,
+      "step": 7385
+    },
+    {
+      "epoch": 3.2826666666666666,
+      "grad_norm": 3.9867827892303467,
+      "learning_rate": 6.875444839857652e-05,
+      "loss": 1.5756,
+      "step": 7386
+    },
+    {
+      "epoch": 3.2831111111111113,
+      "grad_norm": 3.7746942043304443,
+      "learning_rate": 6.873665480427046e-05,
+      "loss": 1.4665,
+      "step": 7387
+    },
+    {
+      "epoch": 3.2835555555555556,
+      "grad_norm": 3.536393642425537,
+      "learning_rate": 6.871886120996441e-05,
+      "loss": 1.0364,
+      "step": 7388
+    },
+    {
+      "epoch": 3.284,
+      "grad_norm": 3.190711498260498,
+      "learning_rate": 6.870106761565837e-05,
+      "loss": 0.9868,
+      "step": 7389
+    },
+    {
+      "epoch": 3.2844444444444445,
+      "grad_norm": 3.708233594894409,
+      "learning_rate": 6.868327402135231e-05,
+      "loss": 1.3567,
+      "step": 7390
+    },
+    {
+      "epoch": 3.2848888888888887,
+      "grad_norm": 3.1745657920837402,
+      "learning_rate": 6.866548042704627e-05,
+      "loss": 1.048,
+      "step": 7391
+    },
+    {
+      "epoch": 3.2853333333333334,
+      "grad_norm": 3.826011896133423,
+      "learning_rate": 6.864768683274023e-05,
+      "loss": 1.1933,
+      "step": 7392
+    },
+    {
+      "epoch": 3.2857777777777777,
+      "grad_norm": 3.8659939765930176,
+      "learning_rate": 6.862989323843416e-05,
+      "loss": 1.0345,
+      "step": 7393
+    },
+    {
+      "epoch": 3.2862222222222224,
+      "grad_norm": 4.651473522186279,
+      "learning_rate": 6.861209964412811e-05,
+      "loss": 1.0049,
+      "step": 7394
+    },
+    {
+      "epoch": 3.2866666666666666,
+      "grad_norm": 4.106167316436768,
+      "learning_rate": 6.859430604982207e-05,
+      "loss": 0.9622,
+      "step": 7395
+    },
+    {
+      "epoch": 3.287111111111111,
+      "grad_norm": 4.003452301025391,
+      "learning_rate": 6.857651245551601e-05,
+      "loss": 1.0904,
+      "step": 7396
+    },
+    {
+      "epoch": 3.2875555555555556,
+      "grad_norm": 0.4317832887172699,
+      "learning_rate": 6.855871886120997e-05,
+      "loss": 0.0452,
+      "step": 7397
+    },
+    {
+      "epoch": 3.288,
+      "grad_norm": 4.068361759185791,
+      "learning_rate": 6.854092526690393e-05,
+      "loss": 1.0543,
+      "step": 7398
+    },
+    {
+      "epoch": 3.2884444444444445,
+      "grad_norm": 3.885014772415161,
+      "learning_rate": 6.852313167259787e-05,
+      "loss": 0.5935,
+      "step": 7399
+    },
+    {
+      "epoch": 3.2888888888888888,
+      "grad_norm": 5.79363489151001,
+      "learning_rate": 6.850533807829181e-05,
+      "loss": 0.9945,
+      "step": 7400
+    },
+    {
+      "epoch": 3.2893333333333334,
+      "grad_norm": 2.343862295150757,
+      "learning_rate": 6.848754448398577e-05,
+      "loss": 1.873,
+      "step": 7401
+    },
+    {
+      "epoch": 3.2897777777777777,
+      "grad_norm": 2.8761746883392334,
+      "learning_rate": 6.846975088967972e-05,
+      "loss": 2.2376,
+      "step": 7402
+    },
+    {
+      "epoch": 3.2902222222222224,
+      "grad_norm": 2.3083155155181885,
+      "learning_rate": 6.845195729537367e-05,
+      "loss": 0.9427,
+      "step": 7403
+    },
+    {
+      "epoch": 3.2906666666666666,
+      "grad_norm": 2.915044069290161,
+      "learning_rate": 6.843416370106762e-05,
+      "loss": 1.3453,
+      "step": 7404
+    },
+    {
+      "epoch": 3.2911111111111113,
+      "grad_norm": 3.3559882640838623,
+      "learning_rate": 6.841637010676158e-05,
+      "loss": 2.2773,
+      "step": 7405
+    },
+    {
+      "epoch": 3.2915555555555556,
+      "grad_norm": 2.8695106506347656,
+      "learning_rate": 6.839857651245551e-05,
+      "loss": 1.099,
+      "step": 7406
+    },
+    {
+      "epoch": 3.292,
+      "grad_norm": 3.35927414894104,
+      "learning_rate": 6.838078291814947e-05,
+      "loss": 1.7923,
+      "step": 7407
+    },
+    {
+      "epoch": 3.2924444444444445,
+      "grad_norm": 3.227517604827881,
+      "learning_rate": 6.836298932384342e-05,
+      "loss": 1.651,
+      "step": 7408
+    },
+    {
+      "epoch": 3.2928888888888888,
+      "grad_norm": 3.0707974433898926,
+      "learning_rate": 6.834519572953737e-05,
+      "loss": 1.5187,
+      "step": 7409
+    },
+    {
+      "epoch": 3.2933333333333334,
+      "grad_norm": 2.92110538482666,
+      "learning_rate": 6.832740213523132e-05,
+      "loss": 1.5467,
+      "step": 7410
+    },
+    {
+      "epoch": 3.2937777777777777,
+      "grad_norm": 3.0438144207000732,
+      "learning_rate": 6.830960854092528e-05,
+      "loss": 1.6164,
+      "step": 7411
+    },
+    {
+      "epoch": 3.2942222222222224,
+      "grad_norm": 3.135681629180908,
+      "learning_rate": 6.829181494661922e-05,
+      "loss": 1.3172,
+      "step": 7412
+    },
+    {
+      "epoch": 3.2946666666666666,
+      "grad_norm": 3.0080184936523438,
+      "learning_rate": 6.827402135231316e-05,
+      "loss": 1.5289,
+      "step": 7413
+    },
+    {
+      "epoch": 3.295111111111111,
+      "grad_norm": 2.6062517166137695,
+      "learning_rate": 6.825622775800712e-05,
+      "loss": 1.47,
+      "step": 7414
+    },
+    {
+      "epoch": 3.2955555555555556,
+      "grad_norm": 3.1638424396514893,
+      "learning_rate": 6.823843416370106e-05,
+      "loss": 1.5606,
+      "step": 7415
+    },
+    {
+      "epoch": 3.296,
+      "grad_norm": 3.3744962215423584,
+      "learning_rate": 6.822064056939502e-05,
+      "loss": 1.6891,
+      "step": 7416
+    },
+    {
+      "epoch": 3.2964444444444445,
+      "grad_norm": 3.1398794651031494,
+      "learning_rate": 6.820284697508898e-05,
+      "loss": 1.4498,
+      "step": 7417
+    },
+    {
+      "epoch": 3.2968888888888888,
+      "grad_norm": 2.805931806564331,
+      "learning_rate": 6.818505338078292e-05,
+      "loss": 1.0785,
+      "step": 7418
+    },
+    {
+      "epoch": 3.2973333333333334,
+      "grad_norm": 2.9776318073272705,
+      "learning_rate": 6.816725978647686e-05,
+      "loss": 1.5049,
+      "step": 7419
+    },
+    {
+      "epoch": 3.2977777777777777,
+      "grad_norm": 2.0422849655151367,
+      "learning_rate": 6.814946619217082e-05,
+      "loss": 0.6512,
+      "step": 7420
+    },
+    {
+      "epoch": 3.2982222222222224,
+      "grad_norm": 3.6055731773376465,
+      "learning_rate": 6.813167259786478e-05,
+      "loss": 1.4551,
+      "step": 7421
+    },
+    {
+      "epoch": 3.2986666666666666,
+      "grad_norm": 3.5920650959014893,
+      "learning_rate": 6.811387900355872e-05,
+      "loss": 1.2965,
+      "step": 7422
+    },
+    {
+      "epoch": 3.2991111111111113,
+      "grad_norm": 2.8194003105163574,
+      "learning_rate": 6.809608540925268e-05,
+      "loss": 1.102,
+      "step": 7423
+    },
+    {
+      "epoch": 3.2995555555555556,
+      "grad_norm": 3.0865092277526855,
+      "learning_rate": 6.807829181494663e-05,
+      "loss": 1.0238,
+      "step": 7424
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 3.17631459236145,
+      "learning_rate": 6.806049822064056e-05,
+      "loss": 1.3226,
+      "step": 7425
+    },
+    {
+      "epoch": 3.3004444444444445,
+      "grad_norm": 3.16072678565979,
+      "learning_rate": 6.804270462633452e-05,
+      "loss": 1.5295,
+      "step": 7426
+    },
+    {
+      "epoch": 3.3008888888888888,
+      "grad_norm": 2.93965482711792,
+      "learning_rate": 6.802491103202847e-05,
+      "loss": 0.8391,
+      "step": 7427
+    },
+    {
+      "epoch": 3.3013333333333335,
+      "grad_norm": 3.356229543685913,
+      "learning_rate": 6.800711743772242e-05,
+      "loss": 1.6525,
+      "step": 7428
+    },
+    {
+      "epoch": 3.3017777777777777,
+      "grad_norm": 3.251253128051758,
+      "learning_rate": 6.798932384341637e-05,
+      "loss": 1.2885,
+      "step": 7429
+    },
+    {
+      "epoch": 3.3022222222222224,
+      "grad_norm": 3.1769371032714844,
+      "learning_rate": 6.797153024911033e-05,
+      "loss": 1.0414,
+      "step": 7430
+    },
+    {
+      "epoch": 3.3026666666666666,
+      "grad_norm": 3.4345128536224365,
+      "learning_rate": 6.795373665480427e-05,
+      "loss": 1.0469,
+      "step": 7431
+    },
+    {
+      "epoch": 3.303111111111111,
+      "grad_norm": 3.2680447101593018,
+      "learning_rate": 6.793594306049822e-05,
+      "loss": 1.3976,
+      "step": 7432
+    },
+    {
+      "epoch": 3.3035555555555556,
+      "grad_norm": 3.0672709941864014,
+      "learning_rate": 6.791814946619217e-05,
+      "loss": 1.2144,
+      "step": 7433
+    },
+    {
+      "epoch": 3.304,
+      "grad_norm": 3.2540392875671387,
+      "learning_rate": 6.790035587188613e-05,
+      "loss": 1.1354,
+      "step": 7434
+    },
+    {
+      "epoch": 3.3044444444444445,
+      "grad_norm": 3.3583109378814697,
+      "learning_rate": 6.788256227758007e-05,
+      "loss": 1.2996,
+      "step": 7435
+    },
+    {
+      "epoch": 3.3048888888888888,
+      "grad_norm": 4.012465476989746,
+      "learning_rate": 6.786476868327403e-05,
+      "loss": 1.5203,
+      "step": 7436
+    },
+    {
+      "epoch": 3.3053333333333335,
+      "grad_norm": 3.8637561798095703,
+      "learning_rate": 6.784697508896798e-05,
+      "loss": 1.3783,
+      "step": 7437
+    },
+    {
+      "epoch": 3.3057777777777777,
+      "grad_norm": 3.8466806411743164,
+      "learning_rate": 6.782918149466191e-05,
+      "loss": 1.3447,
+      "step": 7438
+    },
+    {
+      "epoch": 3.3062222222222224,
+      "grad_norm": 3.448899269104004,
+      "learning_rate": 6.781138790035587e-05,
+      "loss": 1.1285,
+      "step": 7439
+    },
+    {
+      "epoch": 3.3066666666666666,
+      "grad_norm": 3.3083012104034424,
+      "learning_rate": 6.779359430604983e-05,
+      "loss": 0.8823,
+      "step": 7440
+    },
+    {
+      "epoch": 3.3071111111111113,
+      "grad_norm": 4.381979465484619,
+      "learning_rate": 6.777580071174377e-05,
+      "loss": 1.1432,
+      "step": 7441
+    },
+    {
+      "epoch": 3.3075555555555556,
+      "grad_norm": 4.388772964477539,
+      "learning_rate": 6.775800711743773e-05,
+      "loss": 1.2684,
+      "step": 7442
+    },
+    {
+      "epoch": 3.308,
+      "grad_norm": 3.7288894653320312,
+      "learning_rate": 6.774021352313168e-05,
+      "loss": 1.035,
+      "step": 7443
+    },
+    {
+      "epoch": 3.3084444444444445,
+      "grad_norm": 4.472944736480713,
+      "learning_rate": 6.772241992882563e-05,
+      "loss": 1.217,
+      "step": 7444
+    },
+    {
+      "epoch": 3.3088888888888888,
+      "grad_norm": 4.284884929656982,
+      "learning_rate": 6.770462633451957e-05,
+      "loss": 1.237,
+      "step": 7445
+    },
+    {
+      "epoch": 3.3093333333333335,
+      "grad_norm": 6.807945251464844,
+      "learning_rate": 6.768683274021353e-05,
+      "loss": 1.2017,
+      "step": 7446
+    },
+    {
+      "epoch": 3.3097777777777777,
+      "grad_norm": 4.781992435455322,
+      "learning_rate": 6.766903914590748e-05,
+      "loss": 1.32,
+      "step": 7447
+    },
+    {
+      "epoch": 3.3102222222222224,
+      "grad_norm": 3.9543511867523193,
+      "learning_rate": 6.765124555160142e-05,
+      "loss": 1.0203,
+      "step": 7448
+    },
+    {
+      "epoch": 3.3106666666666666,
+      "grad_norm": 4.435871124267578,
+      "learning_rate": 6.763345195729538e-05,
+      "loss": 1.1069,
+      "step": 7449
+    },
+    {
+      "epoch": 3.311111111111111,
+      "grad_norm": 4.147932529449463,
+      "learning_rate": 6.761565836298934e-05,
+      "loss": 0.6211,
+      "step": 7450
+    },
+    {
+      "epoch": 3.3115555555555556,
+      "grad_norm": 2.8902626037597656,
+      "learning_rate": 6.759786476868327e-05,
+      "loss": 1.9548,
+      "step": 7451
+    },
+    {
+      "epoch": 3.312,
+      "grad_norm": 1.825205683708191,
+      "learning_rate": 6.758007117437722e-05,
+      "loss": 0.6829,
+      "step": 7452
+    },
+    {
+      "epoch": 3.3124444444444445,
+      "grad_norm": 3.156799077987671,
+      "learning_rate": 6.756227758007118e-05,
+      "loss": 1.5275,
+      "step": 7453
+    },
+    {
+      "epoch": 3.3128888888888888,
+      "grad_norm": 3.1396076679229736,
+      "learning_rate": 6.754448398576512e-05,
+      "loss": 1.4951,
+      "step": 7454
+    },
+    {
+      "epoch": 3.3133333333333335,
+      "grad_norm": 3.152575731277466,
+      "learning_rate": 6.752669039145908e-05,
+      "loss": 0.9246,
+      "step": 7455
+    },
+    {
+      "epoch": 3.3137777777777777,
+      "grad_norm": 3.4386932849884033,
+      "learning_rate": 6.750889679715304e-05,
+      "loss": 1.5932,
+      "step": 7456
+    },
+    {
+      "epoch": 3.3142222222222224,
+      "grad_norm": 3.287623882293701,
+      "learning_rate": 6.749110320284698e-05,
+      "loss": 1.3995,
+      "step": 7457
+    },
+    {
+      "epoch": 3.3146666666666667,
+      "grad_norm": 3.267625570297241,
+      "learning_rate": 6.747330960854092e-05,
+      "loss": 1.4844,
+      "step": 7458
+    },
+    {
+      "epoch": 3.3151111111111113,
+      "grad_norm": 2.988413095474243,
+      "learning_rate": 6.745551601423488e-05,
+      "loss": 1.1216,
+      "step": 7459
+    },
+    {
+      "epoch": 3.3155555555555556,
+      "grad_norm": 2.892350196838379,
+      "learning_rate": 6.743772241992882e-05,
+      "loss": 1.415,
+      "step": 7460
+    },
+    {
+      "epoch": 3.316,
+      "grad_norm": 3.1864426136016846,
+      "learning_rate": 6.741992882562278e-05,
+      "loss": 1.4968,
+      "step": 7461
+    },
+    {
+      "epoch": 3.3164444444444445,
+      "grad_norm": 2.9494309425354004,
+      "learning_rate": 6.740213523131673e-05,
+      "loss": 1.1238,
+      "step": 7462
+    },
+    {
+      "epoch": 3.3168888888888888,
+      "grad_norm": 2.9198689460754395,
+      "learning_rate": 6.738434163701068e-05,
+      "loss": 1.5899,
+      "step": 7463
+    },
+    {
+      "epoch": 3.3173333333333335,
+      "grad_norm": 3.371006965637207,
+      "learning_rate": 6.736654804270462e-05,
+      "loss": 1.564,
+      "step": 7464
+    },
+    {
+      "epoch": 3.3177777777777777,
+      "grad_norm": 2.8129475116729736,
+      "learning_rate": 6.734875444839858e-05,
+      "loss": 1.0162,
+      "step": 7465
+    },
+    {
+      "epoch": 3.3182222222222224,
+      "grad_norm": 3.1402010917663574,
+      "learning_rate": 6.733096085409253e-05,
+      "loss": 1.3211,
+      "step": 7466
+    },
+    {
+      "epoch": 3.3186666666666667,
+      "grad_norm": 2.9242656230926514,
+      "learning_rate": 6.731316725978648e-05,
+      "loss": 1.2605,
+      "step": 7467
+    },
+    {
+      "epoch": 3.319111111111111,
+      "grad_norm": 3.019500255584717,
+      "learning_rate": 6.729537366548043e-05,
+      "loss": 1.3583,
+      "step": 7468
+    },
+    {
+      "epoch": 3.3195555555555556,
+      "grad_norm": 3.6432340145111084,
+      "learning_rate": 6.727758007117439e-05,
+      "loss": 1.4444,
+      "step": 7469
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 3.2907845973968506,
+      "learning_rate": 6.725978647686833e-05,
+      "loss": 1.1577,
+      "step": 7470
+    },
+    {
+      "epoch": 3.3204444444444445,
+      "grad_norm": 2.75240421295166,
+      "learning_rate": 6.724199288256228e-05,
+      "loss": 1.5465,
+      "step": 7471
+    },
+    {
+      "epoch": 3.320888888888889,
+      "grad_norm": 3.2636733055114746,
+      "learning_rate": 6.722419928825623e-05,
+      "loss": 1.212,
+      "step": 7472
+    },
+    {
+      "epoch": 3.3213333333333335,
+      "grad_norm": 3.171222448348999,
+      "learning_rate": 6.720640569395017e-05,
+      "loss": 1.3743,
+      "step": 7473
+    },
+    {
+      "epoch": 3.3217777777777777,
+      "grad_norm": 2.8894879817962646,
+      "learning_rate": 6.718861209964413e-05,
+      "loss": 1.5747,
+      "step": 7474
+    },
+    {
+      "epoch": 3.3222222222222224,
+      "grad_norm": 3.2905118465423584,
+      "learning_rate": 6.717081850533809e-05,
+      "loss": 1.6193,
+      "step": 7475
+    },
+    {
+      "epoch": 3.3226666666666667,
+      "grad_norm": 3.048412799835205,
+      "learning_rate": 6.715302491103203e-05,
+      "loss": 1.1528,
+      "step": 7476
+    },
+    {
+      "epoch": 3.3231111111111113,
+      "grad_norm": 3.6255054473876953,
+      "learning_rate": 6.713523131672597e-05,
+      "loss": 1.3157,
+      "step": 7477
+    },
+    {
+      "epoch": 3.3235555555555556,
+      "grad_norm": 3.5292255878448486,
+      "learning_rate": 6.711743772241993e-05,
+      "loss": 1.2811,
+      "step": 7478
+    },
+    {
+      "epoch": 3.324,
+      "grad_norm": 2.83103346824646,
+      "learning_rate": 6.709964412811389e-05,
+      "loss": 0.9067,
+      "step": 7479
+    },
+    {
+      "epoch": 3.3244444444444445,
+      "grad_norm": 3.3375396728515625,
+      "learning_rate": 6.708185053380783e-05,
+      "loss": 1.307,
+      "step": 7480
+    },
+    {
+      "epoch": 3.324888888888889,
+      "grad_norm": 2.8192222118377686,
+      "learning_rate": 6.706405693950179e-05,
+      "loss": 1.0012,
+      "step": 7481
+    },
+    {
+      "epoch": 3.3253333333333335,
+      "grad_norm": 3.162932872772217,
+      "learning_rate": 6.704626334519574e-05,
+      "loss": 1.2155,
+      "step": 7482
+    },
+    {
+      "epoch": 3.3257777777777777,
+      "grad_norm": 4.290798187255859,
+      "learning_rate": 6.702846975088967e-05,
+      "loss": 1.3119,
+      "step": 7483
+    },
+    {
+      "epoch": 3.3262222222222224,
+      "grad_norm": 3.4641189575195312,
+      "learning_rate": 6.701067615658363e-05,
+      "loss": 1.0565,
+      "step": 7484
+    },
+    {
+      "epoch": 3.3266666666666667,
+      "grad_norm": 4.372963905334473,
+      "learning_rate": 6.699288256227758e-05,
+      "loss": 1.5057,
+      "step": 7485
+    },
+    {
+      "epoch": 3.327111111111111,
+      "grad_norm": 3.908487319946289,
+      "learning_rate": 6.697508896797153e-05,
+      "loss": 1.2567,
+      "step": 7486
+    },
+    {
+      "epoch": 3.3275555555555556,
+      "grad_norm": 3.5218594074249268,
+      "learning_rate": 6.695729537366548e-05,
+      "loss": 1.039,
+      "step": 7487
+    },
+    {
+      "epoch": 3.328,
+      "grad_norm": 4.635097980499268,
+      "learning_rate": 6.693950177935944e-05,
+      "loss": 1.1867,
+      "step": 7488
+    },
+    {
+      "epoch": 3.3284444444444445,
+      "grad_norm": 3.447298526763916,
+      "learning_rate": 6.692170818505338e-05,
+      "loss": 1.1312,
+      "step": 7489
+    },
+    {
+      "epoch": 3.328888888888889,
+      "grad_norm": 3.3791463375091553,
+      "learning_rate": 6.690391459074733e-05,
+      "loss": 0.8666,
+      "step": 7490
+    },
+    {
+      "epoch": 3.3293333333333335,
+      "grad_norm": 2.8901917934417725,
+      "learning_rate": 6.688612099644128e-05,
+      "loss": 1.1193,
+      "step": 7491
+    },
+    {
+      "epoch": 3.3297777777777777,
+      "grad_norm": 3.7830843925476074,
+      "learning_rate": 6.686832740213524e-05,
+      "loss": 1.152,
+      "step": 7492
+    },
+    {
+      "epoch": 3.330222222222222,
+      "grad_norm": 3.2505948543548584,
+      "learning_rate": 6.685053380782918e-05,
+      "loss": 0.9152,
+      "step": 7493
+    },
+    {
+      "epoch": 3.3306666666666667,
+      "grad_norm": 4.339881896972656,
+      "learning_rate": 6.683274021352314e-05,
+      "loss": 1.3848,
+      "step": 7494
+    },
+    {
+      "epoch": 3.3311111111111114,
+      "grad_norm": 4.330918312072754,
+      "learning_rate": 6.68149466192171e-05,
+      "loss": 1.5097,
+      "step": 7495
+    },
+    {
+      "epoch": 3.3315555555555556,
+      "grad_norm": 3.9478585720062256,
+      "learning_rate": 6.679715302491103e-05,
+      "loss": 1.2268,
+      "step": 7496
+    },
+    {
+      "epoch": 3.332,
+      "grad_norm": 4.484266757965088,
+      "learning_rate": 6.677935943060498e-05,
+      "loss": 1.1452,
+      "step": 7497
+    },
+    {
+      "epoch": 3.3324444444444445,
+      "grad_norm": 3.884369373321533,
+      "learning_rate": 6.676156583629894e-05,
+      "loss": 1.3011,
+      "step": 7498
+    },
+    {
+      "epoch": 3.332888888888889,
+      "grad_norm": 5.073292255401611,
+      "learning_rate": 6.674377224199288e-05,
+      "loss": 1.4814,
+      "step": 7499
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 4.949814796447754,
+      "learning_rate": 6.672597864768684e-05,
+      "loss": 0.5286,
+      "step": 7500
+    },
+    {
+      "epoch": 3.3337777777777777,
+      "grad_norm": 2.3786208629608154,
+      "learning_rate": 6.67081850533808e-05,
+      "loss": 1.8626,
+      "step": 7501
+    },
+    {
+      "epoch": 3.3342222222222224,
+      "grad_norm": 2.6522817611694336,
+      "learning_rate": 6.669039145907474e-05,
+      "loss": 2.0315,
+      "step": 7502
+    },
+    {
+      "epoch": 3.3346666666666667,
+      "grad_norm": 2.5052194595336914,
+      "learning_rate": 6.667259786476868e-05,
+      "loss": 1.5958,
+      "step": 7503
+    },
+    {
+      "epoch": 3.335111111111111,
+      "grad_norm": 2.0900769233703613,
+      "learning_rate": 6.665480427046264e-05,
+      "loss": 1.0634,
+      "step": 7504
+    },
+    {
+      "epoch": 3.3355555555555556,
+      "grad_norm": 3.4959940910339355,
+      "learning_rate": 6.663701067615658e-05,
+      "loss": 1.9428,
+      "step": 7505
+    },
+    {
+      "epoch": 3.336,
+      "grad_norm": 2.677666187286377,
+      "learning_rate": 6.661921708185054e-05,
+      "loss": 1.0751,
+      "step": 7506
+    },
+    {
+      "epoch": 3.3364444444444445,
+      "grad_norm": 3.592830181121826,
+      "learning_rate": 6.660142348754449e-05,
+      "loss": 2.0699,
+      "step": 7507
+    },
+    {
+      "epoch": 3.336888888888889,
+      "grad_norm": 2.6081550121307373,
+      "learning_rate": 6.658362989323844e-05,
+      "loss": 1.142,
+      "step": 7508
+    },
+    {
+      "epoch": 3.3373333333333335,
+      "grad_norm": 1.7961082458496094,
+      "learning_rate": 6.656583629893238e-05,
+      "loss": 0.5108,
+      "step": 7509
+    },
+    {
+      "epoch": 3.3377777777777777,
+      "grad_norm": 3.024773359298706,
+      "learning_rate": 6.654804270462633e-05,
+      "loss": 1.2942,
+      "step": 7510
+    },
+    {
+      "epoch": 3.338222222222222,
+      "grad_norm": 3.3666229248046875,
+      "learning_rate": 6.653024911032029e-05,
+      "loss": 1.7867,
+      "step": 7511
+    },
+    {
+      "epoch": 3.3386666666666667,
+      "grad_norm": 3.1930816173553467,
+      "learning_rate": 6.651245551601423e-05,
+      "loss": 1.3751,
+      "step": 7512
+    },
+    {
+      "epoch": 3.339111111111111,
+      "grad_norm": 2.655470609664917,
+      "learning_rate": 6.649466192170819e-05,
+      "loss": 1.6016,
+      "step": 7513
+    },
+    {
+      "epoch": 3.3395555555555556,
+      "grad_norm": 3.1630845069885254,
+      "learning_rate": 6.647686832740215e-05,
+      "loss": 1.3575,
+      "step": 7514
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 3.315333843231201,
+      "learning_rate": 6.645907473309609e-05,
+      "loss": 1.2997,
+      "step": 7515
+    },
+    {
+      "epoch": 3.3404444444444445,
+      "grad_norm": 2.40342378616333,
+      "learning_rate": 6.644128113879003e-05,
+      "loss": 0.9741,
+      "step": 7516
+    },
+    {
+      "epoch": 3.340888888888889,
+      "grad_norm": 0.24996036291122437,
+      "learning_rate": 6.642348754448399e-05,
+      "loss": 0.0318,
+      "step": 7517
+    },
+    {
+      "epoch": 3.3413333333333335,
+      "grad_norm": 2.008341073989868,
+      "learning_rate": 6.640569395017793e-05,
+      "loss": 0.5644,
+      "step": 7518
+    },
+    {
+      "epoch": 3.3417777777777777,
+      "grad_norm": 3.1700870990753174,
+      "learning_rate": 6.638790035587189e-05,
+      "loss": 1.164,
+      "step": 7519
+    },
+    {
+      "epoch": 3.3422222222222224,
+      "grad_norm": 3.5381534099578857,
+      "learning_rate": 6.637010676156585e-05,
+      "loss": 1.6558,
+      "step": 7520
+    },
+    {
+      "epoch": 3.3426666666666667,
+      "grad_norm": 3.9801483154296875,
+      "learning_rate": 6.635231316725979e-05,
+      "loss": 1.6166,
+      "step": 7521
+    },
+    {
+      "epoch": 3.343111111111111,
+      "grad_norm": 3.240372896194458,
+      "learning_rate": 6.633451957295373e-05,
+      "loss": 1.3277,
+      "step": 7522
+    },
+    {
+      "epoch": 3.3435555555555556,
+      "grad_norm": 3.508293390274048,
+      "learning_rate": 6.631672597864769e-05,
+      "loss": 1.3398,
+      "step": 7523
+    },
+    {
+      "epoch": 3.344,
+      "grad_norm": 3.4153871536254883,
+      "learning_rate": 6.629893238434164e-05,
+      "loss": 1.1003,
+      "step": 7524
+    },
+    {
+      "epoch": 3.3444444444444446,
+      "grad_norm": 3.2748870849609375,
+      "learning_rate": 6.628113879003559e-05,
+      "loss": 1.3643,
+      "step": 7525
+    },
+    {
+      "epoch": 3.344888888888889,
+      "grad_norm": 3.4981982707977295,
+      "learning_rate": 6.626334519572954e-05,
+      "loss": 1.3478,
+      "step": 7526
+    },
+    {
+      "epoch": 3.3453333333333335,
+      "grad_norm": 3.316903591156006,
+      "learning_rate": 6.62455516014235e-05,
+      "loss": 1.2191,
+      "step": 7527
+    },
+    {
+      "epoch": 3.3457777777777777,
+      "grad_norm": 3.2292816638946533,
+      "learning_rate": 6.622775800711744e-05,
+      "loss": 1.4199,
+      "step": 7528
+    },
+    {
+      "epoch": 3.346222222222222,
+      "grad_norm": 3.875464677810669,
+      "learning_rate": 6.620996441281139e-05,
+      "loss": 1.4645,
+      "step": 7529
+    },
+    {
+      "epoch": 3.3466666666666667,
+      "grad_norm": 3.4380528926849365,
+      "learning_rate": 6.619217081850534e-05,
+      "loss": 1.2479,
+      "step": 7530
+    },
+    {
+      "epoch": 3.347111111111111,
+      "grad_norm": 3.452080726623535,
+      "learning_rate": 6.617437722419929e-05,
+      "loss": 1.1636,
+      "step": 7531
+    },
+    {
+      "epoch": 3.3475555555555556,
+      "grad_norm": 3.524420738220215,
+      "learning_rate": 6.615658362989324e-05,
+      "loss": 1.2439,
+      "step": 7532
+    },
+    {
+      "epoch": 3.348,
+      "grad_norm": 3.042452096939087,
+      "learning_rate": 6.61387900355872e-05,
+      "loss": 1.1566,
+      "step": 7533
+    },
+    {
+      "epoch": 3.3484444444444446,
+      "grad_norm": 3.027488946914673,
+      "learning_rate": 6.612099644128114e-05,
+      "loss": 1.1191,
+      "step": 7534
+    },
+    {
+      "epoch": 3.348888888888889,
+      "grad_norm": 3.0509836673736572,
+      "learning_rate": 6.610320284697508e-05,
+      "loss": 1.0986,
+      "step": 7535
+    },
+    {
+      "epoch": 3.3493333333333335,
+      "grad_norm": 4.5463080406188965,
+      "learning_rate": 6.608540925266904e-05,
+      "loss": 1.5118,
+      "step": 7536
+    },
+    {
+      "epoch": 3.3497777777777777,
+      "grad_norm": 2.0029096603393555,
+      "learning_rate": 6.6067615658363e-05,
+      "loss": 0.5587,
+      "step": 7537
+    },
+    {
+      "epoch": 3.3502222222222224,
+      "grad_norm": 4.032919406890869,
+      "learning_rate": 6.604982206405694e-05,
+      "loss": 1.5668,
+      "step": 7538
+    },
+    {
+      "epoch": 3.3506666666666667,
+      "grad_norm": 3.9945425987243652,
+      "learning_rate": 6.60320284697509e-05,
+      "loss": 1.3243,
+      "step": 7539
+    },
+    {
+      "epoch": 3.351111111111111,
+      "grad_norm": 3.5870978832244873,
+      "learning_rate": 6.601423487544485e-05,
+      "loss": 1.1531,
+      "step": 7540
+    },
+    {
+      "epoch": 3.3515555555555556,
+      "grad_norm": 3.745569944381714,
+      "learning_rate": 6.599644128113878e-05,
+      "loss": 1.2786,
+      "step": 7541
+    },
+    {
+      "epoch": 3.352,
+      "grad_norm": 3.2785604000091553,
+      "learning_rate": 6.597864768683274e-05,
+      "loss": 1.0538,
+      "step": 7542
+    },
+    {
+      "epoch": 3.3524444444444446,
+      "grad_norm": 4.0194621086120605,
+      "learning_rate": 6.59608540925267e-05,
+      "loss": 1.2927,
+      "step": 7543
+    },
+    {
+      "epoch": 3.352888888888889,
+      "grad_norm": 4.010869026184082,
+      "learning_rate": 6.594306049822064e-05,
+      "loss": 1.2408,
+      "step": 7544
+    },
+    {
+      "epoch": 3.3533333333333335,
+      "grad_norm": 4.548504829406738,
+      "learning_rate": 6.59252669039146e-05,
+      "loss": 0.8367,
+      "step": 7545
+    },
+    {
+      "epoch": 3.3537777777777777,
+      "grad_norm": 3.896096706390381,
+      "learning_rate": 6.590747330960855e-05,
+      "loss": 0.8111,
+      "step": 7546
+    },
+    {
+      "epoch": 3.354222222222222,
+      "grad_norm": 4.7822957038879395,
+      "learning_rate": 6.58896797153025e-05,
+      "loss": 1.4794,
+      "step": 7547
+    },
+    {
+      "epoch": 3.3546666666666667,
+      "grad_norm": 4.28626823425293,
+      "learning_rate": 6.587188612099644e-05,
+      "loss": 1.0438,
+      "step": 7548
+    },
+    {
+      "epoch": 3.355111111111111,
+      "grad_norm": 4.136358737945557,
+      "learning_rate": 6.58540925266904e-05,
+      "loss": 0.9042,
+      "step": 7549
+    },
+    {
+      "epoch": 3.3555555555555556,
+      "grad_norm": 3.873260021209717,
+      "learning_rate": 6.583629893238434e-05,
+      "loss": 0.8692,
+      "step": 7550
+    },
+    {
+      "epoch": 3.356,
+      "grad_norm": 2.481637716293335,
+      "learning_rate": 6.58185053380783e-05,
+      "loss": 2.0427,
+      "step": 7551
+    },
+    {
+      "epoch": 3.3564444444444446,
+      "grad_norm": 2.9505603313446045,
+      "learning_rate": 6.580071174377225e-05,
+      "loss": 1.5954,
+      "step": 7552
+    },
+    {
+      "epoch": 3.356888888888889,
+      "grad_norm": 2.6145901679992676,
+      "learning_rate": 6.578291814946619e-05,
+      "loss": 1.3394,
+      "step": 7553
+    },
+    {
+      "epoch": 3.3573333333333335,
+      "grad_norm": 2.786313533782959,
+      "learning_rate": 6.576512455516014e-05,
+      "loss": 1.376,
+      "step": 7554
+    },
+    {
+      "epoch": 3.3577777777777778,
+      "grad_norm": 3.3209388256073,
+      "learning_rate": 6.574733096085409e-05,
+      "loss": 1.4978,
+      "step": 7555
+    },
+    {
+      "epoch": 3.3582222222222224,
+      "grad_norm": 3.2253782749176025,
+      "learning_rate": 6.572953736654805e-05,
+      "loss": 1.345,
+      "step": 7556
+    },
+    {
+      "epoch": 3.3586666666666667,
+      "grad_norm": 2.5213868618011475,
+      "learning_rate": 6.571174377224199e-05,
+      "loss": 1.0464,
+      "step": 7557
+    },
+    {
+      "epoch": 3.359111111111111,
+      "grad_norm": 3.1049644947052,
+      "learning_rate": 6.569395017793595e-05,
+      "loss": 1.2791,
+      "step": 7558
+    },
+    {
+      "epoch": 3.3595555555555556,
+      "grad_norm": 3.2726714611053467,
+      "learning_rate": 6.56761565836299e-05,
+      "loss": 1.1688,
+      "step": 7559
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 3.0971052646636963,
+      "learning_rate": 6.565836298932385e-05,
+      "loss": 1.1853,
+      "step": 7560
+    },
+    {
+      "epoch": 3.3604444444444446,
+      "grad_norm": 3.0031280517578125,
+      "learning_rate": 6.564056939501779e-05,
+      "loss": 1.5328,
+      "step": 7561
+    },
+    {
+      "epoch": 3.360888888888889,
+      "grad_norm": 4.010336399078369,
+      "learning_rate": 6.562277580071175e-05,
+      "loss": 1.3606,
+      "step": 7562
+    },
+    {
+      "epoch": 3.3613333333333335,
+      "grad_norm": 3.0197978019714355,
+      "learning_rate": 6.560498220640569e-05,
+      "loss": 1.5253,
+      "step": 7563
+    },
+    {
+      "epoch": 3.3617777777777778,
+      "grad_norm": 3.6670336723327637,
+      "learning_rate": 6.558718861209965e-05,
+      "loss": 2.0248,
+      "step": 7564
+    },
+    {
+      "epoch": 3.362222222222222,
+      "grad_norm": 3.820261001586914,
+      "learning_rate": 6.55693950177936e-05,
+      "loss": 1.6484,
+      "step": 7565
+    },
+    {
+      "epoch": 3.3626666666666667,
+      "grad_norm": 3.1288888454437256,
+      "learning_rate": 6.555160142348755e-05,
+      "loss": 1.1557,
+      "step": 7566
+    },
+    {
+      "epoch": 3.363111111111111,
+      "grad_norm": 3.1067354679107666,
+      "learning_rate": 6.553380782918149e-05,
+      "loss": 0.9204,
+      "step": 7567
+    },
+    {
+      "epoch": 3.3635555555555556,
+      "grad_norm": 3.1598074436187744,
+      "learning_rate": 6.551601423487545e-05,
+      "loss": 1.193,
+      "step": 7568
+    },
+    {
+      "epoch": 3.364,
+      "grad_norm": 2.919339656829834,
+      "learning_rate": 6.54982206405694e-05,
+      "loss": 1.3601,
+      "step": 7569
+    },
+    {
+      "epoch": 3.3644444444444446,
+      "grad_norm": 3.384775161743164,
+      "learning_rate": 6.548042704626335e-05,
+      "loss": 1.3367,
+      "step": 7570
+    },
+    {
+      "epoch": 3.364888888888889,
+      "grad_norm": 3.7233974933624268,
+      "learning_rate": 6.54626334519573e-05,
+      "loss": 1.748,
+      "step": 7571
+    },
+    {
+      "epoch": 3.3653333333333335,
+      "grad_norm": 3.6494290828704834,
+      "learning_rate": 6.544483985765126e-05,
+      "loss": 1.646,
+      "step": 7572
+    },
+    {
+      "epoch": 3.3657777777777778,
+      "grad_norm": 3.263012409210205,
+      "learning_rate": 6.54270462633452e-05,
+      "loss": 0.9094,
+      "step": 7573
+    },
+    {
+      "epoch": 3.3662222222222224,
+      "grad_norm": 3.7953481674194336,
+      "learning_rate": 6.540925266903914e-05,
+      "loss": 1.4884,
+      "step": 7574
+    },
+    {
+      "epoch": 3.3666666666666667,
+      "grad_norm": 2.8919782638549805,
+      "learning_rate": 6.53914590747331e-05,
+      "loss": 1.121,
+      "step": 7575
+    },
+    {
+      "epoch": 3.367111111111111,
+      "grad_norm": 2.639195680618286,
+      "learning_rate": 6.537366548042704e-05,
+      "loss": 0.5362,
+      "step": 7576
+    },
+    {
+      "epoch": 3.3675555555555556,
+      "grad_norm": 3.4763965606689453,
+      "learning_rate": 6.5355871886121e-05,
+      "loss": 1.3417,
+      "step": 7577
+    },
+    {
+      "epoch": 3.368,
+      "grad_norm": 2.803255558013916,
+      "learning_rate": 6.533807829181496e-05,
+      "loss": 1.0939,
+      "step": 7578
+    },
+    {
+      "epoch": 3.3684444444444446,
+      "grad_norm": 3.6555652618408203,
+      "learning_rate": 6.53202846975089e-05,
+      "loss": 1.3298,
+      "step": 7579
+    },
+    {
+      "epoch": 3.368888888888889,
+      "grad_norm": 3.3960537910461426,
+      "learning_rate": 6.530249110320284e-05,
+      "loss": 1.1507,
+      "step": 7580
+    },
+    {
+      "epoch": 3.3693333333333335,
+      "grad_norm": 2.8318889141082764,
+      "learning_rate": 6.52846975088968e-05,
+      "loss": 0.9001,
+      "step": 7581
+    },
+    {
+      "epoch": 3.3697777777777778,
+      "grad_norm": 3.948547601699829,
+      "learning_rate": 6.526690391459076e-05,
+      "loss": 0.7694,
+      "step": 7582
+    },
+    {
+      "epoch": 3.370222222222222,
+      "grad_norm": 3.6763882637023926,
+      "learning_rate": 6.52491103202847e-05,
+      "loss": 0.9653,
+      "step": 7583
+    },
+    {
+      "epoch": 3.3706666666666667,
+      "grad_norm": 4.624429225921631,
+      "learning_rate": 6.523131672597865e-05,
+      "loss": 1.2327,
+      "step": 7584
+    },
+    {
+      "epoch": 3.371111111111111,
+      "grad_norm": 3.7554173469543457,
+      "learning_rate": 6.521352313167261e-05,
+      "loss": 1.0042,
+      "step": 7585
+    },
+    {
+      "epoch": 3.3715555555555556,
+      "grad_norm": 2.8825485706329346,
+      "learning_rate": 6.519572953736655e-05,
+      "loss": 0.8331,
+      "step": 7586
+    },
+    {
+      "epoch": 3.372,
+      "grad_norm": 4.5201849937438965,
+      "learning_rate": 6.51779359430605e-05,
+      "loss": 1.8612,
+      "step": 7587
+    },
+    {
+      "epoch": 3.3724444444444446,
+      "grad_norm": 3.1740362644195557,
+      "learning_rate": 6.516014234875445e-05,
+      "loss": 1.1159,
+      "step": 7588
+    },
+    {
+      "epoch": 3.372888888888889,
+      "grad_norm": 3.3097000122070312,
+      "learning_rate": 6.51423487544484e-05,
+      "loss": 1.3252,
+      "step": 7589
+    },
+    {
+      "epoch": 3.3733333333333335,
+      "grad_norm": 3.417339563369751,
+      "learning_rate": 6.512455516014235e-05,
+      "loss": 0.9649,
+      "step": 7590
+    },
+    {
+      "epoch": 3.3737777777777778,
+      "grad_norm": 4.091862201690674,
+      "learning_rate": 6.510676156583631e-05,
+      "loss": 1.5291,
+      "step": 7591
+    },
+    {
+      "epoch": 3.3742222222222225,
+      "grad_norm": 5.507560729980469,
+      "learning_rate": 6.508896797153025e-05,
+      "loss": 0.6388,
+      "step": 7592
+    },
+    {
+      "epoch": 3.3746666666666667,
+      "grad_norm": 5.815156936645508,
+      "learning_rate": 6.50711743772242e-05,
+      "loss": 1.1697,
+      "step": 7593
+    },
+    {
+      "epoch": 3.375111111111111,
+      "grad_norm": 3.993685245513916,
+      "learning_rate": 6.505338078291815e-05,
+      "loss": 1.3822,
+      "step": 7594
+    },
+    {
+      "epoch": 3.3755555555555556,
+      "grad_norm": 3.4094078540802,
+      "learning_rate": 6.50355871886121e-05,
+      "loss": 0.8347,
+      "step": 7595
+    },
+    {
+      "epoch": 3.376,
+      "grad_norm": 3.818934202194214,
+      "learning_rate": 6.501779359430605e-05,
+      "loss": 1.0501,
+      "step": 7596
+    },
+    {
+      "epoch": 3.3764444444444446,
+      "grad_norm": 5.371906280517578,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 1.3774,
+      "step": 7597
+    },
+    {
+      "epoch": 3.376888888888889,
+      "grad_norm": 4.580707550048828,
+      "learning_rate": 6.498220640569395e-05,
+      "loss": 1.2469,
+      "step": 7598
+    },
+    {
+      "epoch": 3.3773333333333335,
+      "grad_norm": 5.118098258972168,
+      "learning_rate": 6.49644128113879e-05,
+      "loss": 1.7323,
+      "step": 7599
+    },
+    {
+      "epoch": 3.3777777777777778,
+      "grad_norm": 4.494351387023926,
+      "learning_rate": 6.494661921708185e-05,
+      "loss": 1.426,
+      "step": 7600
+    },
+    {
+      "epoch": 3.378222222222222,
+      "grad_norm": 1.999345302581787,
+      "learning_rate": 6.49288256227758e-05,
+      "loss": 1.0275,
+      "step": 7601
+    },
+    {
+      "epoch": 3.3786666666666667,
+      "grad_norm": 2.187124729156494,
+      "learning_rate": 6.491103202846975e-05,
+      "loss": 0.8653,
+      "step": 7602
+    },
+    {
+      "epoch": 3.379111111111111,
+      "grad_norm": 3.221345901489258,
+      "learning_rate": 6.48932384341637e-05,
+      "loss": 1.444,
+      "step": 7603
+    },
+    {
+      "epoch": 3.3795555555555556,
+      "grad_norm": 2.575716495513916,
+      "learning_rate": 6.487544483985766e-05,
+      "loss": 1.2003,
+      "step": 7604
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 3.7870845794677734,
+      "learning_rate": 6.48576512455516e-05,
+      "loss": 1.7092,
+      "step": 7605
+    },
+    {
+      "epoch": 3.3804444444444446,
+      "grad_norm": 2.4028003215789795,
+      "learning_rate": 6.483985765124555e-05,
+      "loss": 0.777,
+      "step": 7606
+    },
+    {
+      "epoch": 3.380888888888889,
+      "grad_norm": 2.2200560569763184,
+      "learning_rate": 6.48220640569395e-05,
+      "loss": 0.8364,
+      "step": 7607
+    },
+    {
+      "epoch": 3.3813333333333335,
+      "grad_norm": 3.2974655628204346,
+      "learning_rate": 6.480427046263345e-05,
+      "loss": 1.5625,
+      "step": 7608
+    },
+    {
+      "epoch": 3.3817777777777778,
+      "grad_norm": 3.614440441131592,
+      "learning_rate": 6.47864768683274e-05,
+      "loss": 1.6034,
+      "step": 7609
+    },
+    {
+      "epoch": 3.3822222222222225,
+      "grad_norm": 2.842564582824707,
+      "learning_rate": 6.476868327402136e-05,
+      "loss": 1.3101,
+      "step": 7610
+    },
+    {
+      "epoch": 3.3826666666666667,
+      "grad_norm": 3.6041154861450195,
+      "learning_rate": 6.47508896797153e-05,
+      "loss": 1.7356,
+      "step": 7611
+    },
+    {
+      "epoch": 3.383111111111111,
+      "grad_norm": 2.605125665664673,
+      "learning_rate": 6.473309608540925e-05,
+      "loss": 0.7838,
+      "step": 7612
+    },
+    {
+      "epoch": 3.3835555555555556,
+      "grad_norm": 3.500422477722168,
+      "learning_rate": 6.47153024911032e-05,
+      "loss": 1.4996,
+      "step": 7613
+    },
+    {
+      "epoch": 3.384,
+      "grad_norm": 2.843087673187256,
+      "learning_rate": 6.469750889679716e-05,
+      "loss": 1.1637,
+      "step": 7614
+    },
+    {
+      "epoch": 3.3844444444444446,
+      "grad_norm": 2.8948328495025635,
+      "learning_rate": 6.46797153024911e-05,
+      "loss": 1.3661,
+      "step": 7615
+    },
+    {
+      "epoch": 3.384888888888889,
+      "grad_norm": 3.8482465744018555,
+      "learning_rate": 6.466192170818506e-05,
+      "loss": 1.4079,
+      "step": 7616
+    },
+    {
+      "epoch": 3.3853333333333335,
+      "grad_norm": 3.166999101638794,
+      "learning_rate": 6.464412811387902e-05,
+      "loss": 1.2834,
+      "step": 7617
+    },
+    {
+      "epoch": 3.3857777777777778,
+      "grad_norm": 2.966557264328003,
+      "learning_rate": 6.462633451957296e-05,
+      "loss": 1.0525,
+      "step": 7618
+    },
+    {
+      "epoch": 3.386222222222222,
+      "grad_norm": 3.0287766456604004,
+      "learning_rate": 6.46085409252669e-05,
+      "loss": 1.0802,
+      "step": 7619
+    },
+    {
+      "epoch": 3.3866666666666667,
+      "grad_norm": 3.048327684402466,
+      "learning_rate": 6.459074733096086e-05,
+      "loss": 1.3006,
+      "step": 7620
+    },
+    {
+      "epoch": 3.387111111111111,
+      "grad_norm": 2.9743454456329346,
+      "learning_rate": 6.45729537366548e-05,
+      "loss": 1.1943,
+      "step": 7621
+    },
+    {
+      "epoch": 3.3875555555555557,
+      "grad_norm": 3.471954345703125,
+      "learning_rate": 6.455516014234876e-05,
+      "loss": 1.1794,
+      "step": 7622
+    },
+    {
+      "epoch": 3.388,
+      "grad_norm": 3.0449366569519043,
+      "learning_rate": 6.453736654804271e-05,
+      "loss": 1.1742,
+      "step": 7623
+    },
+    {
+      "epoch": 3.3884444444444446,
+      "grad_norm": 3.310814142227173,
+      "learning_rate": 6.451957295373666e-05,
+      "loss": 1.1978,
+      "step": 7624
+    },
+    {
+      "epoch": 3.388888888888889,
+      "grad_norm": 3.090825319290161,
+      "learning_rate": 6.45017793594306e-05,
+      "loss": 1.2543,
+      "step": 7625
+    },
+    {
+      "epoch": 3.389333333333333,
+      "grad_norm": 3.573758363723755,
+      "learning_rate": 6.448398576512456e-05,
+      "loss": 1.0293,
+      "step": 7626
+    },
+    {
+      "epoch": 3.389777777777778,
+      "grad_norm": 3.0098912715911865,
+      "learning_rate": 6.446619217081851e-05,
+      "loss": 1.0174,
+      "step": 7627
+    },
+    {
+      "epoch": 3.3902222222222225,
+      "grad_norm": 3.734511375427246,
+      "learning_rate": 6.444839857651246e-05,
+      "loss": 1.4176,
+      "step": 7628
+    },
+    {
+      "epoch": 3.3906666666666667,
+      "grad_norm": 3.394307851791382,
+      "learning_rate": 6.443060498220641e-05,
+      "loss": 1.4112,
+      "step": 7629
+    },
+    {
+      "epoch": 3.391111111111111,
+      "grad_norm": 3.520249605178833,
+      "learning_rate": 6.441281138790037e-05,
+      "loss": 1.3007,
+      "step": 7630
+    },
+    {
+      "epoch": 3.3915555555555557,
+      "grad_norm": 3.260222911834717,
+      "learning_rate": 6.439501779359431e-05,
+      "loss": 1.1919,
+      "step": 7631
+    },
+    {
+      "epoch": 3.392,
+      "grad_norm": 3.4656856060028076,
+      "learning_rate": 6.437722419928825e-05,
+      "loss": 1.1526,
+      "step": 7632
+    },
+    {
+      "epoch": 3.3924444444444446,
+      "grad_norm": 3.670583963394165,
+      "learning_rate": 6.435943060498221e-05,
+      "loss": 1.222,
+      "step": 7633
+    },
+    {
+      "epoch": 3.392888888888889,
+      "grad_norm": 3.5100245475769043,
+      "learning_rate": 6.434163701067615e-05,
+      "loss": 1.4112,
+      "step": 7634
+    },
+    {
+      "epoch": 3.3933333333333335,
+      "grad_norm": 3.7049787044525146,
+      "learning_rate": 6.432384341637011e-05,
+      "loss": 1.3744,
+      "step": 7635
+    },
+    {
+      "epoch": 3.393777777777778,
+      "grad_norm": 3.5180797576904297,
+      "learning_rate": 6.430604982206407e-05,
+      "loss": 1.0369,
+      "step": 7636
+    },
+    {
+      "epoch": 3.394222222222222,
+      "grad_norm": 3.888057231903076,
+      "learning_rate": 6.428825622775801e-05,
+      "loss": 1.3029,
+      "step": 7637
+    },
+    {
+      "epoch": 3.3946666666666667,
+      "grad_norm": 2.7446529865264893,
+      "learning_rate": 6.427046263345195e-05,
+      "loss": 1.0311,
+      "step": 7638
+    },
+    {
+      "epoch": 3.395111111111111,
+      "grad_norm": 4.01584005355835,
+      "learning_rate": 6.425266903914591e-05,
+      "loss": 1.0342,
+      "step": 7639
+    },
+    {
+      "epoch": 3.3955555555555557,
+      "grad_norm": 4.280939102172852,
+      "learning_rate": 6.423487544483985e-05,
+      "loss": 1.4491,
+      "step": 7640
+    },
+    {
+      "epoch": 3.396,
+      "grad_norm": 4.160184383392334,
+      "learning_rate": 6.421708185053381e-05,
+      "loss": 1.8072,
+      "step": 7641
+    },
+    {
+      "epoch": 3.3964444444444446,
+      "grad_norm": 3.48480486869812,
+      "learning_rate": 6.419928825622777e-05,
+      "loss": 1.0331,
+      "step": 7642
+    },
+    {
+      "epoch": 3.396888888888889,
+      "grad_norm": 3.321204423904419,
+      "learning_rate": 6.418149466192171e-05,
+      "loss": 0.9976,
+      "step": 7643
+    },
+    {
+      "epoch": 3.397333333333333,
+      "grad_norm": 3.8095834255218506,
+      "learning_rate": 6.416370106761567e-05,
+      "loss": 1.3911,
+      "step": 7644
+    },
+    {
+      "epoch": 3.397777777777778,
+      "grad_norm": 4.368819236755371,
+      "learning_rate": 6.414590747330961e-05,
+      "loss": 1.1837,
+      "step": 7645
+    },
+    {
+      "epoch": 3.398222222222222,
+      "grad_norm": 4.469260215759277,
+      "learning_rate": 6.412811387900356e-05,
+      "loss": 1.0485,
+      "step": 7646
+    },
+    {
+      "epoch": 3.3986666666666667,
+      "grad_norm": 4.630403518676758,
+      "learning_rate": 6.411032028469751e-05,
+      "loss": 1.3338,
+      "step": 7647
+    },
+    {
+      "epoch": 3.399111111111111,
+      "grad_norm": 4.27545166015625,
+      "learning_rate": 6.409252669039146e-05,
+      "loss": 1.071,
+      "step": 7648
+    },
+    {
+      "epoch": 3.3995555555555557,
+      "grad_norm": 5.530716896057129,
+      "learning_rate": 6.407473309608542e-05,
+      "loss": 1.4461,
+      "step": 7649
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 5.1273193359375,
+      "learning_rate": 6.405693950177936e-05,
+      "loss": 0.9624,
+      "step": 7650
+    },
+    {
+      "epoch": 3.4004444444444446,
+      "grad_norm": 2.5767862796783447,
+      "learning_rate": 6.40391459074733e-05,
+      "loss": 1.6648,
+      "step": 7651
+    },
+    {
+      "epoch": 3.400888888888889,
+      "grad_norm": 2.5170185565948486,
+      "learning_rate": 6.402135231316726e-05,
+      "loss": 1.7532,
+      "step": 7652
+    },
+    {
+      "epoch": 3.4013333333333335,
+      "grad_norm": 2.523386240005493,
+      "learning_rate": 6.40035587188612e-05,
+      "loss": 1.7704,
+      "step": 7653
+    },
+    {
+      "epoch": 3.401777777777778,
+      "grad_norm": 2.007690668106079,
+      "learning_rate": 6.398576512455516e-05,
+      "loss": 0.8182,
+      "step": 7654
+    },
+    {
+      "epoch": 3.402222222222222,
+      "grad_norm": 2.8901143074035645,
+      "learning_rate": 6.396797153024912e-05,
+      "loss": 1.8572,
+      "step": 7655
+    },
+    {
+      "epoch": 3.4026666666666667,
+      "grad_norm": 3.179241895675659,
+      "learning_rate": 6.395017793594306e-05,
+      "loss": 1.5891,
+      "step": 7656
+    },
+    {
+      "epoch": 3.403111111111111,
+      "grad_norm": 3.0536370277404785,
+      "learning_rate": 6.3932384341637e-05,
+      "loss": 1.4515,
+      "step": 7657
+    },
+    {
+      "epoch": 3.4035555555555557,
+      "grad_norm": 2.9210469722747803,
+      "learning_rate": 6.391459074733096e-05,
+      "loss": 1.8337,
+      "step": 7658
+    },
+    {
+      "epoch": 3.404,
+      "grad_norm": 3.83801531791687,
+      "learning_rate": 6.389679715302492e-05,
+      "loss": 1.8063,
+      "step": 7659
+    },
+    {
+      "epoch": 3.4044444444444446,
+      "grad_norm": 2.360499382019043,
+      "learning_rate": 6.387900355871886e-05,
+      "loss": 0.762,
+      "step": 7660
+    },
+    {
+      "epoch": 3.404888888888889,
+      "grad_norm": 3.5909371376037598,
+      "learning_rate": 6.386120996441282e-05,
+      "loss": 1.4551,
+      "step": 7661
+    },
+    {
+      "epoch": 3.405333333333333,
+      "grad_norm": 3.15690541267395,
+      "learning_rate": 6.384341637010677e-05,
+      "loss": 1.387,
+      "step": 7662
+    },
+    {
+      "epoch": 3.405777777777778,
+      "grad_norm": 3.711642026901245,
+      "learning_rate": 6.382562277580072e-05,
+      "loss": 1.6299,
+      "step": 7663
+    },
+    {
+      "epoch": 3.406222222222222,
+      "grad_norm": 3.540889263153076,
+      "learning_rate": 6.380782918149466e-05,
+      "loss": 1.5186,
+      "step": 7664
+    },
+    {
+      "epoch": 3.4066666666666667,
+      "grad_norm": 3.2987284660339355,
+      "learning_rate": 6.379003558718862e-05,
+      "loss": 1.0771,
+      "step": 7665
+    },
+    {
+      "epoch": 3.407111111111111,
+      "grad_norm": 3.3886878490448,
+      "learning_rate": 6.377224199288256e-05,
+      "loss": 1.8004,
+      "step": 7666
+    },
+    {
+      "epoch": 3.4075555555555557,
+      "grad_norm": 2.921910524368286,
+      "learning_rate": 6.375444839857652e-05,
+      "loss": 1.3376,
+      "step": 7667
+    },
+    {
+      "epoch": 3.408,
+      "grad_norm": 0.2479649782180786,
+      "learning_rate": 6.373665480427047e-05,
+      "loss": 0.0287,
+      "step": 7668
+    },
+    {
+      "epoch": 3.4084444444444446,
+      "grad_norm": 2.7667505741119385,
+      "learning_rate": 6.371886120996441e-05,
+      "loss": 1.1423,
+      "step": 7669
+    },
+    {
+      "epoch": 3.408888888888889,
+      "grad_norm": 2.919297933578491,
+      "learning_rate": 6.370106761565836e-05,
+      "loss": 1.5937,
+      "step": 7670
+    },
+    {
+      "epoch": 3.4093333333333335,
+      "grad_norm": 3.060208559036255,
+      "learning_rate": 6.368327402135231e-05,
+      "loss": 1.3546,
+      "step": 7671
+    },
+    {
+      "epoch": 3.409777777777778,
+      "grad_norm": 3.8419125080108643,
+      "learning_rate": 6.366548042704627e-05,
+      "loss": 1.6202,
+      "step": 7672
+    },
+    {
+      "epoch": 3.410222222222222,
+      "grad_norm": 3.074181318283081,
+      "learning_rate": 6.364768683274021e-05,
+      "loss": 1.1265,
+      "step": 7673
+    },
+    {
+      "epoch": 3.4106666666666667,
+      "grad_norm": 3.2371325492858887,
+      "learning_rate": 6.362989323843417e-05,
+      "loss": 0.9814,
+      "step": 7674
+    },
+    {
+      "epoch": 3.411111111111111,
+      "grad_norm": 3.5532615184783936,
+      "learning_rate": 6.361209964412813e-05,
+      "loss": 1.5752,
+      "step": 7675
+    },
+    {
+      "epoch": 3.4115555555555557,
+      "grad_norm": 2.725200653076172,
+      "learning_rate": 6.359430604982207e-05,
+      "loss": 0.9369,
+      "step": 7676
+    },
+    {
+      "epoch": 3.412,
+      "grad_norm": 2.41279673576355,
+      "learning_rate": 6.357651245551601e-05,
+      "loss": 1.2278,
+      "step": 7677
+    },
+    {
+      "epoch": 3.4124444444444446,
+      "grad_norm": 3.5397229194641113,
+      "learning_rate": 6.355871886120997e-05,
+      "loss": 1.1223,
+      "step": 7678
+    },
+    {
+      "epoch": 3.412888888888889,
+      "grad_norm": 3.2247607707977295,
+      "learning_rate": 6.354092526690391e-05,
+      "loss": 1.6677,
+      "step": 7679
+    },
+    {
+      "epoch": 3.413333333333333,
+      "grad_norm": 3.0843541622161865,
+      "learning_rate": 6.352313167259787e-05,
+      "loss": 1.3358,
+      "step": 7680
+    },
+    {
+      "epoch": 3.413777777777778,
+      "grad_norm": 3.328740358352661,
+      "learning_rate": 6.350533807829183e-05,
+      "loss": 1.5565,
+      "step": 7681
+    },
+    {
+      "epoch": 3.414222222222222,
+      "grad_norm": 2.958048105239868,
+      "learning_rate": 6.348754448398577e-05,
+      "loss": 1.2256,
+      "step": 7682
+    },
+    {
+      "epoch": 3.4146666666666667,
+      "grad_norm": 0.26679933071136475,
+      "learning_rate": 6.346975088967971e-05,
+      "loss": 0.0342,
+      "step": 7683
+    },
+    {
+      "epoch": 3.415111111111111,
+      "grad_norm": 4.010861396789551,
+      "learning_rate": 6.345195729537367e-05,
+      "loss": 1.9807,
+      "step": 7684
+    },
+    {
+      "epoch": 3.4155555555555557,
+      "grad_norm": 3.461297035217285,
+      "learning_rate": 6.343416370106761e-05,
+      "loss": 1.4831,
+      "step": 7685
+    },
+    {
+      "epoch": 3.416,
+      "grad_norm": 3.9611904621124268,
+      "learning_rate": 6.341637010676157e-05,
+      "loss": 0.9961,
+      "step": 7686
+    },
+    {
+      "epoch": 3.4164444444444446,
+      "grad_norm": 3.7136943340301514,
+      "learning_rate": 6.339857651245552e-05,
+      "loss": 1.3796,
+      "step": 7687
+    },
+    {
+      "epoch": 3.416888888888889,
+      "grad_norm": 3.2812464237213135,
+      "learning_rate": 6.338078291814947e-05,
+      "loss": 1.3365,
+      "step": 7688
+    },
+    {
+      "epoch": 3.4173333333333336,
+      "grad_norm": 3.156381130218506,
+      "learning_rate": 6.336298932384342e-05,
+      "loss": 1.0016,
+      "step": 7689
+    },
+    {
+      "epoch": 3.417777777777778,
+      "grad_norm": 1.95352041721344,
+      "learning_rate": 6.334519572953737e-05,
+      "loss": 0.5333,
+      "step": 7690
+    },
+    {
+      "epoch": 3.418222222222222,
+      "grad_norm": 3.408022880554199,
+      "learning_rate": 6.332740213523132e-05,
+      "loss": 1.2858,
+      "step": 7691
+    },
+    {
+      "epoch": 3.4186666666666667,
+      "grad_norm": 4.81541633605957,
+      "learning_rate": 6.330960854092527e-05,
+      "loss": 1.3646,
+      "step": 7692
+    },
+    {
+      "epoch": 3.419111111111111,
+      "grad_norm": 3.8997044563293457,
+      "learning_rate": 6.329181494661922e-05,
+      "loss": 1.4012,
+      "step": 7693
+    },
+    {
+      "epoch": 3.4195555555555557,
+      "grad_norm": 3.6768176555633545,
+      "learning_rate": 6.327402135231318e-05,
+      "loss": 1.3647,
+      "step": 7694
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 3.490180015563965,
+      "learning_rate": 6.325622775800712e-05,
+      "loss": 0.9646,
+      "step": 7695
+    },
+    {
+      "epoch": 3.4204444444444446,
+      "grad_norm": 3.8013601303100586,
+      "learning_rate": 6.323843416370106e-05,
+      "loss": 1.5347,
+      "step": 7696
+    },
+    {
+      "epoch": 3.420888888888889,
+      "grad_norm": 4.499445915222168,
+      "learning_rate": 6.322064056939502e-05,
+      "loss": 1.0899,
+      "step": 7697
+    },
+    {
+      "epoch": 3.421333333333333,
+      "grad_norm": 3.4003989696502686,
+      "learning_rate": 6.320284697508896e-05,
+      "loss": 0.9259,
+      "step": 7698
+    },
+    {
+      "epoch": 3.421777777777778,
+      "grad_norm": 4.400297164916992,
+      "learning_rate": 6.318505338078292e-05,
+      "loss": 1.6985,
+      "step": 7699
+    },
+    {
+      "epoch": 3.422222222222222,
+      "grad_norm": 4.983489513397217,
+      "learning_rate": 6.316725978647688e-05,
+      "loss": 1.1409,
+      "step": 7700
+    },
+    {
+      "epoch": 3.4226666666666667,
+      "grad_norm": 2.452336311340332,
+      "learning_rate": 6.314946619217082e-05,
+      "loss": 1.7351,
+      "step": 7701
+    },
+    {
+      "epoch": 3.423111111111111,
+      "grad_norm": 2.2893435955047607,
+      "learning_rate": 6.313167259786478e-05,
+      "loss": 2.1632,
+      "step": 7702
+    },
+    {
+      "epoch": 3.4235555555555557,
+      "grad_norm": 2.175917625427246,
+      "learning_rate": 6.311387900355872e-05,
+      "loss": 1.0005,
+      "step": 7703
+    },
+    {
+      "epoch": 3.424,
+      "grad_norm": 2.591916084289551,
+      "learning_rate": 6.309608540925268e-05,
+      "loss": 1.5803,
+      "step": 7704
+    },
+    {
+      "epoch": 3.4244444444444446,
+      "grad_norm": 2.7328264713287354,
+      "learning_rate": 6.307829181494662e-05,
+      "loss": 1.6264,
+      "step": 7705
+    },
+    {
+      "epoch": 3.424888888888889,
+      "grad_norm": 2.771141767501831,
+      "learning_rate": 6.306049822064057e-05,
+      "loss": 1.3346,
+      "step": 7706
+    },
+    {
+      "epoch": 3.4253333333333336,
+      "grad_norm": 3.2554972171783447,
+      "learning_rate": 6.304270462633453e-05,
+      "loss": 1.3667,
+      "step": 7707
+    },
+    {
+      "epoch": 3.425777777777778,
+      "grad_norm": 4.142671585083008,
+      "learning_rate": 6.302491103202847e-05,
+      "loss": 1.5744,
+      "step": 7708
+    },
+    {
+      "epoch": 3.426222222222222,
+      "grad_norm": 3.247718095779419,
+      "learning_rate": 6.300711743772242e-05,
+      "loss": 1.2485,
+      "step": 7709
+    },
+    {
+      "epoch": 3.4266666666666667,
+      "grad_norm": 3.285029649734497,
+      "learning_rate": 6.298932384341637e-05,
+      "loss": 1.3671,
+      "step": 7710
+    },
+    {
+      "epoch": 3.427111111111111,
+      "grad_norm": 2.8541810512542725,
+      "learning_rate": 6.297153024911032e-05,
+      "loss": 1.0453,
+      "step": 7711
+    },
+    {
+      "epoch": 3.4275555555555557,
+      "grad_norm": 3.167170286178589,
+      "learning_rate": 6.295373665480427e-05,
+      "loss": 1.1664,
+      "step": 7712
+    },
+    {
+      "epoch": 3.428,
+      "grad_norm": 3.222472667694092,
+      "learning_rate": 6.293594306049823e-05,
+      "loss": 1.0299,
+      "step": 7713
+    },
+    {
+      "epoch": 3.4284444444444446,
+      "grad_norm": 3.7111828327178955,
+      "learning_rate": 6.291814946619217e-05,
+      "loss": 1.3693,
+      "step": 7714
+    },
+    {
+      "epoch": 3.428888888888889,
+      "grad_norm": 3.3316597938537598,
+      "learning_rate": 6.290035587188612e-05,
+      "loss": 1.7048,
+      "step": 7715
+    },
+    {
+      "epoch": 3.429333333333333,
+      "grad_norm": 2.9972000122070312,
+      "learning_rate": 6.288256227758007e-05,
+      "loss": 1.1209,
+      "step": 7716
+    },
+    {
+      "epoch": 3.429777777777778,
+      "grad_norm": 3.5281057357788086,
+      "learning_rate": 6.286476868327403e-05,
+      "loss": 1.6927,
+      "step": 7717
+    },
+    {
+      "epoch": 3.430222222222222,
+      "grad_norm": 3.553957939147949,
+      "learning_rate": 6.284697508896797e-05,
+      "loss": 1.5864,
+      "step": 7718
+    },
+    {
+      "epoch": 3.4306666666666668,
+      "grad_norm": 4.420113563537598,
+      "learning_rate": 6.282918149466193e-05,
+      "loss": 1.5708,
+      "step": 7719
+    },
+    {
+      "epoch": 3.431111111111111,
+      "grad_norm": 4.024765968322754,
+      "learning_rate": 6.281138790035588e-05,
+      "loss": 1.3665,
+      "step": 7720
+    },
+    {
+      "epoch": 3.4315555555555557,
+      "grad_norm": 3.158511161804199,
+      "learning_rate": 6.279359430604983e-05,
+      "loss": 1.1667,
+      "step": 7721
+    },
+    {
+      "epoch": 3.432,
+      "grad_norm": 3.0785789489746094,
+      "learning_rate": 6.277580071174377e-05,
+      "loss": 1.1982,
+      "step": 7722
+    },
+    {
+      "epoch": 3.4324444444444446,
+      "grad_norm": 2.7140047550201416,
+      "learning_rate": 6.275800711743773e-05,
+      "loss": 1.1619,
+      "step": 7723
+    },
+    {
+      "epoch": 3.432888888888889,
+      "grad_norm": 3.3508801460266113,
+      "learning_rate": 6.274021352313167e-05,
+      "loss": 1.3535,
+      "step": 7724
+    },
+    {
+      "epoch": 3.4333333333333336,
+      "grad_norm": 2.8375468254089355,
+      "learning_rate": 6.272241992882563e-05,
+      "loss": 1.1456,
+      "step": 7725
+    },
+    {
+      "epoch": 3.433777777777778,
+      "grad_norm": 2.967170000076294,
+      "learning_rate": 6.270462633451958e-05,
+      "loss": 1.1639,
+      "step": 7726
+    },
+    {
+      "epoch": 3.434222222222222,
+      "grad_norm": 3.734205484390259,
+      "learning_rate": 6.268683274021353e-05,
+      "loss": 1.4107,
+      "step": 7727
+    },
+    {
+      "epoch": 3.4346666666666668,
+      "grad_norm": 3.417689085006714,
+      "learning_rate": 6.266903914590747e-05,
+      "loss": 1.2238,
+      "step": 7728
+    },
+    {
+      "epoch": 3.435111111111111,
+      "grad_norm": 3.4484238624572754,
+      "learning_rate": 6.265124555160143e-05,
+      "loss": 1.1604,
+      "step": 7729
+    },
+    {
+      "epoch": 3.4355555555555557,
+      "grad_norm": 1.4519230127334595,
+      "learning_rate": 6.263345195729537e-05,
+      "loss": 0.2928,
+      "step": 7730
+    },
+    {
+      "epoch": 3.436,
+      "grad_norm": 3.699568748474121,
+      "learning_rate": 6.261565836298932e-05,
+      "loss": 1.5317,
+      "step": 7731
+    },
+    {
+      "epoch": 3.4364444444444446,
+      "grad_norm": 3.111086845397949,
+      "learning_rate": 6.259786476868328e-05,
+      "loss": 1.3173,
+      "step": 7732
+    },
+    {
+      "epoch": 3.436888888888889,
+      "grad_norm": 2.9475185871124268,
+      "learning_rate": 6.258007117437722e-05,
+      "loss": 0.8653,
+      "step": 7733
+    },
+    {
+      "epoch": 3.437333333333333,
+      "grad_norm": 3.5937764644622803,
+      "learning_rate": 6.256227758007118e-05,
+      "loss": 1.4628,
+      "step": 7734
+    },
+    {
+      "epoch": 3.437777777777778,
+      "grad_norm": 3.6315152645111084,
+      "learning_rate": 6.254448398576512e-05,
+      "loss": 1.6076,
+      "step": 7735
+    },
+    {
+      "epoch": 3.438222222222222,
+      "grad_norm": 3.4667177200317383,
+      "learning_rate": 6.252669039145908e-05,
+      "loss": 1.4787,
+      "step": 7736
+    },
+    {
+      "epoch": 3.4386666666666668,
+      "grad_norm": 3.2693827152252197,
+      "learning_rate": 6.250889679715302e-05,
+      "loss": 1.1539,
+      "step": 7737
+    },
+    {
+      "epoch": 3.439111111111111,
+      "grad_norm": 3.5391995906829834,
+      "learning_rate": 6.249110320284698e-05,
+      "loss": 1.1326,
+      "step": 7738
+    },
+    {
+      "epoch": 3.4395555555555557,
+      "grad_norm": 4.569096088409424,
+      "learning_rate": 6.247330960854094e-05,
+      "loss": 1.5917,
+      "step": 7739
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 3.872166633605957,
+      "learning_rate": 6.245551601423488e-05,
+      "loss": 1.3355,
+      "step": 7740
+    },
+    {
+      "epoch": 3.4404444444444446,
+      "grad_norm": 3.8824992179870605,
+      "learning_rate": 6.243772241992882e-05,
+      "loss": 1.5446,
+      "step": 7741
+    },
+    {
+      "epoch": 3.440888888888889,
+      "grad_norm": 3.592456579208374,
+      "learning_rate": 6.241992882562278e-05,
+      "loss": 0.9791,
+      "step": 7742
+    },
+    {
+      "epoch": 3.4413333333333336,
+      "grad_norm": 3.4050142765045166,
+      "learning_rate": 6.240213523131672e-05,
+      "loss": 0.8251,
+      "step": 7743
+    },
+    {
+      "epoch": 3.441777777777778,
+      "grad_norm": 3.6496431827545166,
+      "learning_rate": 6.238434163701068e-05,
+      "loss": 1.2421,
+      "step": 7744
+    },
+    {
+      "epoch": 3.442222222222222,
+      "grad_norm": 4.084912300109863,
+      "learning_rate": 6.236654804270463e-05,
+      "loss": 1.1315,
+      "step": 7745
+    },
+    {
+      "epoch": 3.4426666666666668,
+      "grad_norm": 4.4305033683776855,
+      "learning_rate": 6.234875444839858e-05,
+      "loss": 0.8433,
+      "step": 7746
+    },
+    {
+      "epoch": 3.443111111111111,
+      "grad_norm": 4.542752742767334,
+      "learning_rate": 6.233096085409253e-05,
+      "loss": 1.3728,
+      "step": 7747
+    },
+    {
+      "epoch": 3.4435555555555557,
+      "grad_norm": 4.393490791320801,
+      "learning_rate": 6.231316725978648e-05,
+      "loss": 1.4556,
+      "step": 7748
+    },
+    {
+      "epoch": 3.444,
+      "grad_norm": 4.165998458862305,
+      "learning_rate": 6.229537366548043e-05,
+      "loss": 0.8499,
+      "step": 7749
+    },
+    {
+      "epoch": 3.4444444444444446,
+      "grad_norm": 3.2756288051605225,
+      "learning_rate": 6.227758007117438e-05,
+      "loss": 0.59,
+      "step": 7750
+    },
+    {
+      "epoch": 3.444888888888889,
+      "grad_norm": 1.687354564666748,
+      "learning_rate": 6.225978647686833e-05,
+      "loss": 0.7765,
+      "step": 7751
+    },
+    {
+      "epoch": 3.445333333333333,
+      "grad_norm": 2.0984833240509033,
+      "learning_rate": 6.224199288256229e-05,
+      "loss": 1.1706,
+      "step": 7752
+    },
+    {
+      "epoch": 3.445777777777778,
+      "grad_norm": 1.5732635259628296,
+      "learning_rate": 6.222419928825623e-05,
+      "loss": 0.4636,
+      "step": 7753
+    },
+    {
+      "epoch": 3.446222222222222,
+      "grad_norm": 2.6701159477233887,
+      "learning_rate": 6.220640569395018e-05,
+      "loss": 1.3432,
+      "step": 7754
+    },
+    {
+      "epoch": 3.4466666666666668,
+      "grad_norm": 2.630073308944702,
+      "learning_rate": 6.218861209964413e-05,
+      "loss": 1.229,
+      "step": 7755
+    },
+    {
+      "epoch": 3.447111111111111,
+      "grad_norm": 3.405998945236206,
+      "learning_rate": 6.217081850533807e-05,
+      "loss": 1.6136,
+      "step": 7756
+    },
+    {
+      "epoch": 3.4475555555555557,
+      "grad_norm": 3.1982181072235107,
+      "learning_rate": 6.215302491103203e-05,
+      "loss": 1.4239,
+      "step": 7757
+    },
+    {
+      "epoch": 3.448,
+      "grad_norm": 3.283080816268921,
+      "learning_rate": 6.213523131672599e-05,
+      "loss": 1.5826,
+      "step": 7758
+    },
+    {
+      "epoch": 3.448444444444444,
+      "grad_norm": 3.861956834793091,
+      "learning_rate": 6.211743772241993e-05,
+      "loss": 1.9406,
+      "step": 7759
+    },
+    {
+      "epoch": 3.448888888888889,
+      "grad_norm": 3.5665135383605957,
+      "learning_rate": 6.209964412811389e-05,
+      "loss": 1.5776,
+      "step": 7760
+    },
+    {
+      "epoch": 3.449333333333333,
+      "grad_norm": 3.1784329414367676,
+      "learning_rate": 6.208185053380783e-05,
+      "loss": 1.1607,
+      "step": 7761
+    },
+    {
+      "epoch": 3.449777777777778,
+      "grad_norm": 3.31244158744812,
+      "learning_rate": 6.206405693950179e-05,
+      "loss": 1.2746,
+      "step": 7762
+    },
+    {
+      "epoch": 3.450222222222222,
+      "grad_norm": 3.8500120639801025,
+      "learning_rate": 6.204626334519573e-05,
+      "loss": 1.4645,
+      "step": 7763
+    },
+    {
+      "epoch": 3.4506666666666668,
+      "grad_norm": 3.650425434112549,
+      "learning_rate": 6.202846975088969e-05,
+      "loss": 1.5679,
+      "step": 7764
+    },
+    {
+      "epoch": 3.451111111111111,
+      "grad_norm": 3.314197063446045,
+      "learning_rate": 6.201067615658364e-05,
+      "loss": 1.236,
+      "step": 7765
+    },
+    {
+      "epoch": 3.4515555555555557,
+      "grad_norm": 3.924046277999878,
+      "learning_rate": 6.199288256227759e-05,
+      "loss": 1.4479,
+      "step": 7766
+    },
+    {
+      "epoch": 3.452,
+      "grad_norm": 4.836942672729492,
+      "learning_rate": 6.197508896797153e-05,
+      "loss": 1.4685,
+      "step": 7767
+    },
+    {
+      "epoch": 3.4524444444444446,
+      "grad_norm": 3.9449543952941895,
+      "learning_rate": 6.195729537366548e-05,
+      "loss": 1.5383,
+      "step": 7768
+    },
+    {
+      "epoch": 3.452888888888889,
+      "grad_norm": 3.259953737258911,
+      "learning_rate": 6.193950177935943e-05,
+      "loss": 1.3374,
+      "step": 7769
+    },
+    {
+      "epoch": 3.453333333333333,
+      "grad_norm": 3.4236366748809814,
+      "learning_rate": 6.192170818505338e-05,
+      "loss": 1.4552,
+      "step": 7770
+    },
+    {
+      "epoch": 3.453777777777778,
+      "grad_norm": 3.391941547393799,
+      "learning_rate": 6.190391459074734e-05,
+      "loss": 1.3679,
+      "step": 7771
+    },
+    {
+      "epoch": 3.454222222222222,
+      "grad_norm": 2.923438787460327,
+      "learning_rate": 6.188612099644128e-05,
+      "loss": 1.0185,
+      "step": 7772
+    },
+    {
+      "epoch": 3.4546666666666668,
+      "grad_norm": 3.636385202407837,
+      "learning_rate": 6.186832740213523e-05,
+      "loss": 1.2137,
+      "step": 7773
+    },
+    {
+      "epoch": 3.455111111111111,
+      "grad_norm": 3.6334786415100098,
+      "learning_rate": 6.185053380782918e-05,
+      "loss": 1.2733,
+      "step": 7774
+    },
+    {
+      "epoch": 3.4555555555555557,
+      "grad_norm": 3.618476152420044,
+      "learning_rate": 6.183274021352313e-05,
+      "loss": 1.4823,
+      "step": 7775
+    },
+    {
+      "epoch": 3.456,
+      "grad_norm": 3.8086905479431152,
+      "learning_rate": 6.181494661921708e-05,
+      "loss": 1.2488,
+      "step": 7776
+    },
+    {
+      "epoch": 3.456444444444444,
+      "grad_norm": 4.449453353881836,
+      "learning_rate": 6.179715302491104e-05,
+      "loss": 1.2205,
+      "step": 7777
+    },
+    {
+      "epoch": 3.456888888888889,
+      "grad_norm": 2.2830700874328613,
+      "learning_rate": 6.177935943060498e-05,
+      "loss": 0.576,
+      "step": 7778
+    },
+    {
+      "epoch": 3.457333333333333,
+      "grad_norm": 3.2258355617523193,
+      "learning_rate": 6.176156583629894e-05,
+      "loss": 1.0031,
+      "step": 7779
+    },
+    {
+      "epoch": 3.457777777777778,
+      "grad_norm": 3.3974528312683105,
+      "learning_rate": 6.174377224199288e-05,
+      "loss": 1.2962,
+      "step": 7780
+    },
+    {
+      "epoch": 3.458222222222222,
+      "grad_norm": 3.8626692295074463,
+      "learning_rate": 6.172597864768684e-05,
+      "loss": 1.663,
+      "step": 7781
+    },
+    {
+      "epoch": 3.458666666666667,
+      "grad_norm": 2.1425042152404785,
+      "learning_rate": 6.170818505338078e-05,
+      "loss": 0.5912,
+      "step": 7782
+    },
+    {
+      "epoch": 3.459111111111111,
+      "grad_norm": 3.64699649810791,
+      "learning_rate": 6.169039145907474e-05,
+      "loss": 1.5683,
+      "step": 7783
+    },
+    {
+      "epoch": 3.4595555555555557,
+      "grad_norm": 3.0158698558807373,
+      "learning_rate": 6.16725978647687e-05,
+      "loss": 1.1897,
+      "step": 7784
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.8843777179718018,
+      "learning_rate": 6.165480427046264e-05,
+      "loss": 0.9849,
+      "step": 7785
+    },
+    {
+      "epoch": 3.4604444444444447,
+      "grad_norm": 4.387118339538574,
+      "learning_rate": 6.163701067615658e-05,
+      "loss": 1.2511,
+      "step": 7786
+    },
+    {
+      "epoch": 3.460888888888889,
+      "grad_norm": 3.0603246688842773,
+      "learning_rate": 6.161921708185054e-05,
+      "loss": 1.0083,
+      "step": 7787
+    },
+    {
+      "epoch": 3.461333333333333,
+      "grad_norm": 3.8462274074554443,
+      "learning_rate": 6.160142348754448e-05,
+      "loss": 1.4729,
+      "step": 7788
+    },
+    {
+      "epoch": 3.461777777777778,
+      "grad_norm": 3.917243719100952,
+      "learning_rate": 6.158362989323844e-05,
+      "loss": 1.104,
+      "step": 7789
+    },
+    {
+      "epoch": 3.462222222222222,
+      "grad_norm": 4.193759441375732,
+      "learning_rate": 6.156583629893239e-05,
+      "loss": 1.4846,
+      "step": 7790
+    },
+    {
+      "epoch": 3.462666666666667,
+      "grad_norm": 3.3156356811523438,
+      "learning_rate": 6.154804270462634e-05,
+      "loss": 0.9882,
+      "step": 7791
+    },
+    {
+      "epoch": 3.463111111111111,
+      "grad_norm": 4.5232062339782715,
+      "learning_rate": 6.153024911032029e-05,
+      "loss": 1.1761,
+      "step": 7792
+    },
+    {
+      "epoch": 3.4635555555555557,
+      "grad_norm": 3.9281444549560547,
+      "learning_rate": 6.151245551601423e-05,
+      "loss": 1.3836,
+      "step": 7793
+    },
+    {
+      "epoch": 3.464,
+      "grad_norm": 5.008172035217285,
+      "learning_rate": 6.149466192170819e-05,
+      "loss": 1.4874,
+      "step": 7794
+    },
+    {
+      "epoch": 3.464444444444444,
+      "grad_norm": 5.387003421783447,
+      "learning_rate": 6.147686832740213e-05,
+      "loss": 1.0767,
+      "step": 7795
+    },
+    {
+      "epoch": 3.464888888888889,
+      "grad_norm": 5.167206764221191,
+      "learning_rate": 6.145907473309609e-05,
+      "loss": 1.3537,
+      "step": 7796
+    },
+    {
+      "epoch": 3.465333333333333,
+      "grad_norm": 3.742565870285034,
+      "learning_rate": 6.144128113879005e-05,
+      "loss": 0.7384,
+      "step": 7797
+    },
+    {
+      "epoch": 3.465777777777778,
+      "grad_norm": 3.991476058959961,
+      "learning_rate": 6.142348754448399e-05,
+      "loss": 1.1434,
+      "step": 7798
+    },
+    {
+      "epoch": 3.466222222222222,
+      "grad_norm": 2.9769797325134277,
+      "learning_rate": 6.140569395017793e-05,
+      "loss": 0.496,
+      "step": 7799
+    },
+    {
+      "epoch": 3.466666666666667,
+      "grad_norm": 3.5614140033721924,
+      "learning_rate": 6.138790035587189e-05,
+      "loss": 0.4358,
+      "step": 7800
+    },
+    {
+      "epoch": 3.467111111111111,
+      "grad_norm": 2.5742027759552,
+      "learning_rate": 6.137010676156583e-05,
+      "loss": 2.0281,
+      "step": 7801
+    },
+    {
+      "epoch": 3.4675555555555557,
+      "grad_norm": 2.53623366355896,
+      "learning_rate": 6.135231316725979e-05,
+      "loss": 1.6929,
+      "step": 7802
+    },
+    {
+      "epoch": 3.468,
+      "grad_norm": 2.944135904312134,
+      "learning_rate": 6.133451957295375e-05,
+      "loss": 1.7063,
+      "step": 7803
+    },
+    {
+      "epoch": 3.4684444444444447,
+      "grad_norm": 2.7259559631347656,
+      "learning_rate": 6.131672597864769e-05,
+      "loss": 1.2452,
+      "step": 7804
+    },
+    {
+      "epoch": 3.468888888888889,
+      "grad_norm": 3.2671380043029785,
+      "learning_rate": 6.129893238434164e-05,
+      "loss": 1.7628,
+      "step": 7805
+    },
+    {
+      "epoch": 3.469333333333333,
+      "grad_norm": 2.8245341777801514,
+      "learning_rate": 6.128113879003559e-05,
+      "loss": 1.6624,
+      "step": 7806
+    },
+    {
+      "epoch": 3.469777777777778,
+      "grad_norm": 2.8411941528320312,
+      "learning_rate": 6.126334519572954e-05,
+      "loss": 1.4877,
+      "step": 7807
+    },
+    {
+      "epoch": 3.470222222222222,
+      "grad_norm": 3.039693593978882,
+      "learning_rate": 6.124555160142349e-05,
+      "loss": 1.8058,
+      "step": 7808
+    },
+    {
+      "epoch": 3.470666666666667,
+      "grad_norm": 3.047823667526245,
+      "learning_rate": 6.122775800711744e-05,
+      "loss": 1.0683,
+      "step": 7809
+    },
+    {
+      "epoch": 3.471111111111111,
+      "grad_norm": 3.6758055686950684,
+      "learning_rate": 6.12099644128114e-05,
+      "loss": 1.3514,
+      "step": 7810
+    },
+    {
+      "epoch": 3.4715555555555557,
+      "grad_norm": 2.9851505756378174,
+      "learning_rate": 6.119217081850534e-05,
+      "loss": 1.1167,
+      "step": 7811
+    },
+    {
+      "epoch": 3.472,
+      "grad_norm": 2.0843656063079834,
+      "learning_rate": 6.117437722419929e-05,
+      "loss": 0.533,
+      "step": 7812
+    },
+    {
+      "epoch": 3.4724444444444442,
+      "grad_norm": 3.570765733718872,
+      "learning_rate": 6.115658362989324e-05,
+      "loss": 1.5857,
+      "step": 7813
+    },
+    {
+      "epoch": 3.472888888888889,
+      "grad_norm": 2.951054573059082,
+      "learning_rate": 6.113879003558719e-05,
+      "loss": 0.9994,
+      "step": 7814
+    },
+    {
+      "epoch": 3.473333333333333,
+      "grad_norm": 3.410421133041382,
+      "learning_rate": 6.112099644128114e-05,
+      "loss": 1.492,
+      "step": 7815
+    },
+    {
+      "epoch": 3.473777777777778,
+      "grad_norm": 3.0904409885406494,
+      "learning_rate": 6.11032028469751e-05,
+      "loss": 1.6493,
+      "step": 7816
+    },
+    {
+      "epoch": 3.474222222222222,
+      "grad_norm": 3.2813360691070557,
+      "learning_rate": 6.108540925266904e-05,
+      "loss": 1.2129,
+      "step": 7817
+    },
+    {
+      "epoch": 3.474666666666667,
+      "grad_norm": 3.62602162361145,
+      "learning_rate": 6.1067615658363e-05,
+      "loss": 1.2179,
+      "step": 7818
+    },
+    {
+      "epoch": 3.475111111111111,
+      "grad_norm": 2.9935662746429443,
+      "learning_rate": 6.104982206405694e-05,
+      "loss": 1.2209,
+      "step": 7819
+    },
+    {
+      "epoch": 3.4755555555555557,
+      "grad_norm": 3.7781035900115967,
+      "learning_rate": 6.103202846975089e-05,
+      "loss": 1.8109,
+      "step": 7820
+    },
+    {
+      "epoch": 3.476,
+      "grad_norm": 3.0393505096435547,
+      "learning_rate": 6.101423487544484e-05,
+      "loss": 1.3712,
+      "step": 7821
+    },
+    {
+      "epoch": 3.4764444444444447,
+      "grad_norm": 3.436675786972046,
+      "learning_rate": 6.09964412811388e-05,
+      "loss": 1.1229,
+      "step": 7822
+    },
+    {
+      "epoch": 3.476888888888889,
+      "grad_norm": 3.587766170501709,
+      "learning_rate": 6.0978647686832747e-05,
+      "loss": 1.4697,
+      "step": 7823
+    },
+    {
+      "epoch": 3.477333333333333,
+      "grad_norm": 3.541997194290161,
+      "learning_rate": 6.0960854092526696e-05,
+      "loss": 0.978,
+      "step": 7824
+    },
+    {
+      "epoch": 3.477777777777778,
+      "grad_norm": 4.022648334503174,
+      "learning_rate": 6.094306049822064e-05,
+      "loss": 1.1844,
+      "step": 7825
+    },
+    {
+      "epoch": 3.478222222222222,
+      "grad_norm": 3.1931285858154297,
+      "learning_rate": 6.092526690391459e-05,
+      "loss": 1.3828,
+      "step": 7826
+    },
+    {
+      "epoch": 3.478666666666667,
+      "grad_norm": 2.8355612754821777,
+      "learning_rate": 6.0907473309608545e-05,
+      "loss": 0.8536,
+      "step": 7827
+    },
+    {
+      "epoch": 3.479111111111111,
+      "grad_norm": 3.4258205890655518,
+      "learning_rate": 6.0889679715302495e-05,
+      "loss": 1.2377,
+      "step": 7828
+    },
+    {
+      "epoch": 3.4795555555555557,
+      "grad_norm": 3.4359145164489746,
+      "learning_rate": 6.0871886120996445e-05,
+      "loss": 1.2237,
+      "step": 7829
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 3.928776502609253,
+      "learning_rate": 6.08540925266904e-05,
+      "loss": 1.4612,
+      "step": 7830
+    },
+    {
+      "epoch": 3.4804444444444442,
+      "grad_norm": 3.3484604358673096,
+      "learning_rate": 6.083629893238434e-05,
+      "loss": 1.3294,
+      "step": 7831
+    },
+    {
+      "epoch": 3.480888888888889,
+      "grad_norm": 4.023479461669922,
+      "learning_rate": 6.0818505338078294e-05,
+      "loss": 1.1446,
+      "step": 7832
+    },
+    {
+      "epoch": 3.481333333333333,
+      "grad_norm": 4.194918632507324,
+      "learning_rate": 6.0800711743772244e-05,
+      "loss": 1.5604,
+      "step": 7833
+    },
+    {
+      "epoch": 3.481777777777778,
+      "grad_norm": 3.498831033706665,
+      "learning_rate": 6.0782918149466193e-05,
+      "loss": 1.2972,
+      "step": 7834
+    },
+    {
+      "epoch": 3.482222222222222,
+      "grad_norm": 3.6610026359558105,
+      "learning_rate": 6.076512455516015e-05,
+      "loss": 1.0893,
+      "step": 7835
+    },
+    {
+      "epoch": 3.482666666666667,
+      "grad_norm": 3.50296950340271,
+      "learning_rate": 6.07473309608541e-05,
+      "loss": 1.252,
+      "step": 7836
+    },
+    {
+      "epoch": 3.483111111111111,
+      "grad_norm": 3.684080123901367,
+      "learning_rate": 6.072953736654805e-05,
+      "loss": 1.3696,
+      "step": 7837
+    },
+    {
+      "epoch": 3.4835555555555557,
+      "grad_norm": 3.986010789871216,
+      "learning_rate": 6.071174377224199e-05,
+      "loss": 1.3597,
+      "step": 7838
+    },
+    {
+      "epoch": 3.484,
+      "grad_norm": 3.7796928882598877,
+      "learning_rate": 6.069395017793594e-05,
+      "loss": 1.2141,
+      "step": 7839
+    },
+    {
+      "epoch": 3.4844444444444447,
+      "grad_norm": 3.9027299880981445,
+      "learning_rate": 6.067615658362989e-05,
+      "loss": 1.2926,
+      "step": 7840
+    },
+    {
+      "epoch": 3.484888888888889,
+      "grad_norm": 4.090729236602783,
+      "learning_rate": 6.065836298932385e-05,
+      "loss": 1.3292,
+      "step": 7841
+    },
+    {
+      "epoch": 3.485333333333333,
+      "grad_norm": 3.7317214012145996,
+      "learning_rate": 6.06405693950178e-05,
+      "loss": 1.2942,
+      "step": 7842
+    },
+    {
+      "epoch": 3.485777777777778,
+      "grad_norm": 5.53169584274292,
+      "learning_rate": 6.062277580071175e-05,
+      "loss": 1.1903,
+      "step": 7843
+    },
+    {
+      "epoch": 3.486222222222222,
+      "grad_norm": 4.516229629516602,
+      "learning_rate": 6.060498220640569e-05,
+      "loss": 1.246,
+      "step": 7844
+    },
+    {
+      "epoch": 3.486666666666667,
+      "grad_norm": 4.402373313903809,
+      "learning_rate": 6.058718861209964e-05,
+      "loss": 1.4393,
+      "step": 7845
+    },
+    {
+      "epoch": 3.487111111111111,
+      "grad_norm": 5.0335540771484375,
+      "learning_rate": 6.05693950177936e-05,
+      "loss": 0.9533,
+      "step": 7846
+    },
+    {
+      "epoch": 3.4875555555555557,
+      "grad_norm": 4.421757698059082,
+      "learning_rate": 6.0551601423487547e-05,
+      "loss": 1.1198,
+      "step": 7847
+    },
+    {
+      "epoch": 3.488,
+      "grad_norm": 3.6150312423706055,
+      "learning_rate": 6.0533807829181496e-05,
+      "loss": 1.3501,
+      "step": 7848
+    },
+    {
+      "epoch": 3.4884444444444442,
+      "grad_norm": 3.798158645629883,
+      "learning_rate": 6.051601423487545e-05,
+      "loss": 0.9775,
+      "step": 7849
+    },
+    {
+      "epoch": 3.488888888888889,
+      "grad_norm": 4.989626407623291,
+      "learning_rate": 6.04982206405694e-05,
+      "loss": 0.7092,
+      "step": 7850
+    },
+    {
+      "epoch": 3.489333333333333,
+      "grad_norm": 2.5157763957977295,
+      "learning_rate": 6.0480427046263345e-05,
+      "loss": 1.8447,
+      "step": 7851
+    },
+    {
+      "epoch": 3.489777777777778,
+      "grad_norm": 1.8457545042037964,
+      "learning_rate": 6.0462633451957295e-05,
+      "loss": 1.0517,
+      "step": 7852
+    },
+    {
+      "epoch": 3.490222222222222,
+      "grad_norm": 2.611934185028076,
+      "learning_rate": 6.0444839857651245e-05,
+      "loss": 1.78,
+      "step": 7853
+    },
+    {
+      "epoch": 3.490666666666667,
+      "grad_norm": 0.1728007197380066,
+      "learning_rate": 6.04270462633452e-05,
+      "loss": 0.0174,
+      "step": 7854
+    },
+    {
+      "epoch": 3.491111111111111,
+      "grad_norm": 2.6706738471984863,
+      "learning_rate": 6.040925266903915e-05,
+      "loss": 1.605,
+      "step": 7855
+    },
+    {
+      "epoch": 3.4915555555555557,
+      "grad_norm": 2.728020429611206,
+      "learning_rate": 6.03914590747331e-05,
+      "loss": 1.5846,
+      "step": 7856
+    },
+    {
+      "epoch": 3.492,
+      "grad_norm": 3.039168357849121,
+      "learning_rate": 6.0373665480427044e-05,
+      "loss": 1.7239,
+      "step": 7857
+    },
+    {
+      "epoch": 3.4924444444444447,
+      "grad_norm": 2.816784381866455,
+      "learning_rate": 6.0355871886120994e-05,
+      "loss": 1.4047,
+      "step": 7858
+    },
+    {
+      "epoch": 3.492888888888889,
+      "grad_norm": 2.839526653289795,
+      "learning_rate": 6.033807829181495e-05,
+      "loss": 1.4536,
+      "step": 7859
+    },
+    {
+      "epoch": 3.493333333333333,
+      "grad_norm": 3.2044901847839355,
+      "learning_rate": 6.03202846975089e-05,
+      "loss": 1.322,
+      "step": 7860
+    },
+    {
+      "epoch": 3.493777777777778,
+      "grad_norm": 3.872880458831787,
+      "learning_rate": 6.030249110320285e-05,
+      "loss": 1.7948,
+      "step": 7861
+    },
+    {
+      "epoch": 3.494222222222222,
+      "grad_norm": 2.9369702339172363,
+      "learning_rate": 6.0284697508896806e-05,
+      "loss": 1.1952,
+      "step": 7862
+    },
+    {
+      "epoch": 3.494666666666667,
+      "grad_norm": 2.457634687423706,
+      "learning_rate": 6.0266903914590756e-05,
+      "loss": 1.1545,
+      "step": 7863
+    },
+    {
+      "epoch": 3.495111111111111,
+      "grad_norm": 3.318244457244873,
+      "learning_rate": 6.02491103202847e-05,
+      "loss": 1.6481,
+      "step": 7864
+    },
+    {
+      "epoch": 3.4955555555555557,
+      "grad_norm": 3.1333770751953125,
+      "learning_rate": 6.023131672597865e-05,
+      "loss": 1.4185,
+      "step": 7865
+    },
+    {
+      "epoch": 3.496,
+      "grad_norm": 3.627277135848999,
+      "learning_rate": 6.02135231316726e-05,
+      "loss": 1.4683,
+      "step": 7866
+    },
+    {
+      "epoch": 3.4964444444444442,
+      "grad_norm": 2.975083827972412,
+      "learning_rate": 6.0195729537366555e-05,
+      "loss": 1.5541,
+      "step": 7867
+    },
+    {
+      "epoch": 3.496888888888889,
+      "grad_norm": 2.3885438442230225,
+      "learning_rate": 6.0177935943060504e-05,
+      "loss": 1.0314,
+      "step": 7868
+    },
+    {
+      "epoch": 3.497333333333333,
+      "grad_norm": 3.9095265865325928,
+      "learning_rate": 6.0160142348754454e-05,
+      "loss": 1.6982,
+      "step": 7869
+    },
+    {
+      "epoch": 3.497777777777778,
+      "grad_norm": 5.884151935577393,
+      "learning_rate": 6.01423487544484e-05,
+      "loss": 1.5449,
+      "step": 7870
+    },
+    {
+      "epoch": 3.498222222222222,
+      "grad_norm": 3.240220785140991,
+      "learning_rate": 6.012455516014235e-05,
+      "loss": 1.2482,
+      "step": 7871
+    },
+    {
+      "epoch": 3.498666666666667,
+      "grad_norm": 2.9373934268951416,
+      "learning_rate": 6.01067615658363e-05,
+      "loss": 1.3222,
+      "step": 7872
+    },
+    {
+      "epoch": 3.499111111111111,
+      "grad_norm": 3.125972270965576,
+      "learning_rate": 6.008896797153025e-05,
+      "loss": 1.4957,
+      "step": 7873
+    },
+    {
+      "epoch": 3.4995555555555553,
+      "grad_norm": 3.1987497806549072,
+      "learning_rate": 6.00711743772242e-05,
+      "loss": 1.5033,
+      "step": 7874
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 2.531674861907959,
+      "learning_rate": 6.005338078291816e-05,
+      "loss": 1.2043,
+      "step": 7875
+    },
+    {
+      "epoch": 3.5004444444444447,
+      "grad_norm": 3.2194294929504395,
+      "learning_rate": 6.003558718861211e-05,
+      "loss": 1.4242,
+      "step": 7876
+    },
+    {
+      "epoch": 3.500888888888889,
+      "grad_norm": 3.215784788131714,
+      "learning_rate": 6.001779359430605e-05,
+      "loss": 1.5275,
+      "step": 7877
+    },
+    {
+      "epoch": 3.501333333333333,
+      "grad_norm": 2.9531593322753906,
+      "learning_rate": 6e-05,
+      "loss": 1.3093,
+      "step": 7878
+    },
+    {
+      "epoch": 3.501777777777778,
+      "grad_norm": 3.470592737197876,
+      "learning_rate": 5.998220640569395e-05,
+      "loss": 1.5293,
+      "step": 7879
+    },
+    {
+      "epoch": 3.502222222222222,
+      "grad_norm": 3.2422022819519043,
+      "learning_rate": 5.996441281138791e-05,
+      "loss": 1.4052,
+      "step": 7880
+    },
+    {
+      "epoch": 3.502666666666667,
+      "grad_norm": 3.3211963176727295,
+      "learning_rate": 5.994661921708186e-05,
+      "loss": 1.304,
+      "step": 7881
+    },
+    {
+      "epoch": 3.503111111111111,
+      "grad_norm": 3.2284364700317383,
+      "learning_rate": 5.992882562277581e-05,
+      "loss": 1.1494,
+      "step": 7882
+    },
+    {
+      "epoch": 3.5035555555555558,
+      "grad_norm": 3.206890821456909,
+      "learning_rate": 5.991103202846975e-05,
+      "loss": 1.296,
+      "step": 7883
+    },
+    {
+      "epoch": 3.504,
+      "grad_norm": 2.572016716003418,
+      "learning_rate": 5.98932384341637e-05,
+      "loss": 0.7784,
+      "step": 7884
+    },
+    {
+      "epoch": 3.5044444444444443,
+      "grad_norm": 3.5026330947875977,
+      "learning_rate": 5.987544483985765e-05,
+      "loss": 1.3138,
+      "step": 7885
+    },
+    {
+      "epoch": 3.504888888888889,
+      "grad_norm": 3.5628318786621094,
+      "learning_rate": 5.9857651245551606e-05,
+      "loss": 1.1652,
+      "step": 7886
+    },
+    {
+      "epoch": 3.505333333333333,
+      "grad_norm": 3.418849468231201,
+      "learning_rate": 5.9839857651245556e-05,
+      "loss": 1.4964,
+      "step": 7887
+    },
+    {
+      "epoch": 3.505777777777778,
+      "grad_norm": 4.8035454750061035,
+      "learning_rate": 5.9822064056939506e-05,
+      "loss": 1.8758,
+      "step": 7888
+    },
+    {
+      "epoch": 3.506222222222222,
+      "grad_norm": 3.9143970012664795,
+      "learning_rate": 5.980427046263345e-05,
+      "loss": 1.4264,
+      "step": 7889
+    },
+    {
+      "epoch": 3.506666666666667,
+      "grad_norm": 3.6827564239501953,
+      "learning_rate": 5.97864768683274e-05,
+      "loss": 1.5077,
+      "step": 7890
+    },
+    {
+      "epoch": 3.507111111111111,
+      "grad_norm": 4.183266639709473,
+      "learning_rate": 5.9768683274021355e-05,
+      "loss": 1.2283,
+      "step": 7891
+    },
+    {
+      "epoch": 3.5075555555555553,
+      "grad_norm": 3.9071977138519287,
+      "learning_rate": 5.9750889679715304e-05,
+      "loss": 1.2634,
+      "step": 7892
+    },
+    {
+      "epoch": 3.508,
+      "grad_norm": 3.631640672683716,
+      "learning_rate": 5.9733096085409254e-05,
+      "loss": 1.5404,
+      "step": 7893
+    },
+    {
+      "epoch": 3.5084444444444447,
+      "grad_norm": 3.552340269088745,
+      "learning_rate": 5.971530249110321e-05,
+      "loss": 1.2201,
+      "step": 7894
+    },
+    {
+      "epoch": 3.508888888888889,
+      "grad_norm": 4.531553268432617,
+      "learning_rate": 5.969750889679716e-05,
+      "loss": 1.4324,
+      "step": 7895
+    },
+    {
+      "epoch": 3.509333333333333,
+      "grad_norm": 3.7588775157928467,
+      "learning_rate": 5.96797153024911e-05,
+      "loss": 1.2748,
+      "step": 7896
+    },
+    {
+      "epoch": 3.509777777777778,
+      "grad_norm": 4.838202476501465,
+      "learning_rate": 5.966192170818505e-05,
+      "loss": 1.1187,
+      "step": 7897
+    },
+    {
+      "epoch": 3.510222222222222,
+      "grad_norm": 4.4196367263793945,
+      "learning_rate": 5.9644128113879e-05,
+      "loss": 1.0293,
+      "step": 7898
+    },
+    {
+      "epoch": 3.510666666666667,
+      "grad_norm": 3.9595701694488525,
+      "learning_rate": 5.962633451957296e-05,
+      "loss": 1.232,
+      "step": 7899
+    },
+    {
+      "epoch": 3.511111111111111,
+      "grad_norm": 5.525210380554199,
+      "learning_rate": 5.960854092526691e-05,
+      "loss": 0.8993,
+      "step": 7900
+    },
+    {
+      "epoch": 3.5115555555555558,
+      "grad_norm": 2.474489212036133,
+      "learning_rate": 5.959074733096086e-05,
+      "loss": 1.8906,
+      "step": 7901
+    },
+    {
+      "epoch": 3.512,
+      "grad_norm": 1.5867375135421753,
+      "learning_rate": 5.95729537366548e-05,
+      "loss": 0.3189,
+      "step": 7902
+    },
+    {
+      "epoch": 3.5124444444444443,
+      "grad_norm": 2.7589077949523926,
+      "learning_rate": 5.955516014234875e-05,
+      "loss": 1.5188,
+      "step": 7903
+    },
+    {
+      "epoch": 3.512888888888889,
+      "grad_norm": 3.1739702224731445,
+      "learning_rate": 5.953736654804271e-05,
+      "loss": 1.4973,
+      "step": 7904
+    },
+    {
+      "epoch": 3.513333333333333,
+      "grad_norm": 3.220743179321289,
+      "learning_rate": 5.951957295373666e-05,
+      "loss": 1.7824,
+      "step": 7905
+    },
+    {
+      "epoch": 3.513777777777778,
+      "grad_norm": 3.1341934204101562,
+      "learning_rate": 5.950177935943061e-05,
+      "loss": 1.53,
+      "step": 7906
+    },
+    {
+      "epoch": 3.514222222222222,
+      "grad_norm": 2.954864501953125,
+      "learning_rate": 5.9483985765124564e-05,
+      "loss": 1.0464,
+      "step": 7907
+    },
+    {
+      "epoch": 3.514666666666667,
+      "grad_norm": 2.6394574642181396,
+      "learning_rate": 5.9466192170818513e-05,
+      "loss": 1.4441,
+      "step": 7908
+    },
+    {
+      "epoch": 3.515111111111111,
+      "grad_norm": 3.118455648422241,
+      "learning_rate": 5.9448398576512456e-05,
+      "loss": 1.438,
+      "step": 7909
+    },
+    {
+      "epoch": 3.5155555555555553,
+      "grad_norm": 2.812499523162842,
+      "learning_rate": 5.9430604982206406e-05,
+      "loss": 1.32,
+      "step": 7910
+    },
+    {
+      "epoch": 3.516,
+      "grad_norm": 3.96679425239563,
+      "learning_rate": 5.9412811387900356e-05,
+      "loss": 1.5107,
+      "step": 7911
+    },
+    {
+      "epoch": 3.5164444444444447,
+      "grad_norm": 3.8176608085632324,
+      "learning_rate": 5.939501779359431e-05,
+      "loss": 1.7569,
+      "step": 7912
+    },
+    {
+      "epoch": 3.516888888888889,
+      "grad_norm": 3.084055185317993,
+      "learning_rate": 5.937722419928826e-05,
+      "loss": 1.4477,
+      "step": 7913
+    },
+    {
+      "epoch": 3.517333333333333,
+      "grad_norm": 3.462454319000244,
+      "learning_rate": 5.935943060498221e-05,
+      "loss": 1.5585,
+      "step": 7914
+    },
+    {
+      "epoch": 3.517777777777778,
+      "grad_norm": 2.181887149810791,
+      "learning_rate": 5.9341637010676155e-05,
+      "loss": 0.452,
+      "step": 7915
+    },
+    {
+      "epoch": 3.518222222222222,
+      "grad_norm": 3.213517189025879,
+      "learning_rate": 5.9323843416370104e-05,
+      "loss": 1.5827,
+      "step": 7916
+    },
+    {
+      "epoch": 3.518666666666667,
+      "grad_norm": 3.1706907749176025,
+      "learning_rate": 5.930604982206406e-05,
+      "loss": 1.1167,
+      "step": 7917
+    },
+    {
+      "epoch": 3.519111111111111,
+      "grad_norm": 3.325066566467285,
+      "learning_rate": 5.928825622775801e-05,
+      "loss": 1.2163,
+      "step": 7918
+    },
+    {
+      "epoch": 3.5195555555555558,
+      "grad_norm": 3.789510488510132,
+      "learning_rate": 5.927046263345196e-05,
+      "loss": 1.5674,
+      "step": 7919
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 3.43265962600708,
+      "learning_rate": 5.925266903914592e-05,
+      "loss": 1.3059,
+      "step": 7920
+    },
+    {
+      "epoch": 3.5204444444444443,
+      "grad_norm": 3.1737518310546875,
+      "learning_rate": 5.923487544483987e-05,
+      "loss": 1.1704,
+      "step": 7921
+    },
+    {
+      "epoch": 3.520888888888889,
+      "grad_norm": 3.2082409858703613,
+      "learning_rate": 5.921708185053381e-05,
+      "loss": 1.5402,
+      "step": 7922
+    },
+    {
+      "epoch": 3.521333333333333,
+      "grad_norm": 3.5814785957336426,
+      "learning_rate": 5.919928825622776e-05,
+      "loss": 1.4676,
+      "step": 7923
+    },
+    {
+      "epoch": 3.521777777777778,
+      "grad_norm": 3.379570722579956,
+      "learning_rate": 5.918149466192171e-05,
+      "loss": 1.1777,
+      "step": 7924
+    },
+    {
+      "epoch": 3.522222222222222,
+      "grad_norm": 3.4962823390960693,
+      "learning_rate": 5.9163701067615666e-05,
+      "loss": 1.6759,
+      "step": 7925
+    },
+    {
+      "epoch": 3.522666666666667,
+      "grad_norm": 3.593132495880127,
+      "learning_rate": 5.9145907473309615e-05,
+      "loss": 1.5066,
+      "step": 7926
+    },
+    {
+      "epoch": 3.523111111111111,
+      "grad_norm": 3.335500478744507,
+      "learning_rate": 5.9128113879003565e-05,
+      "loss": 0.9421,
+      "step": 7927
+    },
+    {
+      "epoch": 3.5235555555555553,
+      "grad_norm": 3.0314064025878906,
+      "learning_rate": 5.911032028469751e-05,
+      "loss": 1.0284,
+      "step": 7928
+    },
+    {
+      "epoch": 3.524,
+      "grad_norm": 3.651460886001587,
+      "learning_rate": 5.909252669039146e-05,
+      "loss": 1.3157,
+      "step": 7929
+    },
+    {
+      "epoch": 3.5244444444444447,
+      "grad_norm": 3.4120965003967285,
+      "learning_rate": 5.907473309608541e-05,
+      "loss": 1.4894,
+      "step": 7930
+    },
+    {
+      "epoch": 3.524888888888889,
+      "grad_norm": 4.027737140655518,
+      "learning_rate": 5.9056939501779364e-05,
+      "loss": 1.6357,
+      "step": 7931
+    },
+    {
+      "epoch": 3.525333333333333,
+      "grad_norm": 2.242091178894043,
+      "learning_rate": 5.9039145907473314e-05,
+      "loss": 0.5504,
+      "step": 7932
+    },
+    {
+      "epoch": 3.525777777777778,
+      "grad_norm": 3.149824380874634,
+      "learning_rate": 5.902135231316726e-05,
+      "loss": 0.8101,
+      "step": 7933
+    },
+    {
+      "epoch": 3.526222222222222,
+      "grad_norm": 2.0692434310913086,
+      "learning_rate": 5.900355871886122e-05,
+      "loss": 0.4684,
+      "step": 7934
+    },
+    {
+      "epoch": 3.5266666666666664,
+      "grad_norm": 3.153052806854248,
+      "learning_rate": 5.8985765124555156e-05,
+      "loss": 1.1421,
+      "step": 7935
+    },
+    {
+      "epoch": 3.527111111111111,
+      "grad_norm": 3.0566887855529785,
+      "learning_rate": 5.896797153024911e-05,
+      "loss": 1.0727,
+      "step": 7936
+    },
+    {
+      "epoch": 3.5275555555555558,
+      "grad_norm": 3.488640546798706,
+      "learning_rate": 5.895017793594306e-05,
+      "loss": 1.2473,
+      "step": 7937
+    },
+    {
+      "epoch": 3.528,
+      "grad_norm": 3.4758243560791016,
+      "learning_rate": 5.893238434163701e-05,
+      "loss": 0.9908,
+      "step": 7938
+    },
+    {
+      "epoch": 3.5284444444444443,
+      "grad_norm": 5.093916416168213,
+      "learning_rate": 5.891459074733097e-05,
+      "loss": 1.1238,
+      "step": 7939
+    },
+    {
+      "epoch": 3.528888888888889,
+      "grad_norm": 3.6023635864257812,
+      "learning_rate": 5.889679715302492e-05,
+      "loss": 0.8337,
+      "step": 7940
+    },
+    {
+      "epoch": 3.529333333333333,
+      "grad_norm": 3.5062379837036133,
+      "learning_rate": 5.887900355871886e-05,
+      "loss": 1.244,
+      "step": 7941
+    },
+    {
+      "epoch": 3.529777777777778,
+      "grad_norm": 3.3245818614959717,
+      "learning_rate": 5.886120996441281e-05,
+      "loss": 1.0635,
+      "step": 7942
+    },
+    {
+      "epoch": 3.530222222222222,
+      "grad_norm": 3.718571662902832,
+      "learning_rate": 5.884341637010676e-05,
+      "loss": 1.2694,
+      "step": 7943
+    },
+    {
+      "epoch": 3.530666666666667,
+      "grad_norm": 4.398813247680664,
+      "learning_rate": 5.882562277580072e-05,
+      "loss": 1.2429,
+      "step": 7944
+    },
+    {
+      "epoch": 3.531111111111111,
+      "grad_norm": 3.6693432331085205,
+      "learning_rate": 5.880782918149467e-05,
+      "loss": 1.0828,
+      "step": 7945
+    },
+    {
+      "epoch": 3.5315555555555553,
+      "grad_norm": 4.250223159790039,
+      "learning_rate": 5.8790035587188616e-05,
+      "loss": 1.6145,
+      "step": 7946
+    },
+    {
+      "epoch": 3.532,
+      "grad_norm": 2.441047191619873,
+      "learning_rate": 5.877224199288256e-05,
+      "loss": 0.279,
+      "step": 7947
+    },
+    {
+      "epoch": 3.5324444444444447,
+      "grad_norm": 4.091504096984863,
+      "learning_rate": 5.875444839857651e-05,
+      "loss": 0.6715,
+      "step": 7948
+    },
+    {
+      "epoch": 3.532888888888889,
+      "grad_norm": 5.657823085784912,
+      "learning_rate": 5.8736654804270466e-05,
+      "loss": 1.2939,
+      "step": 7949
+    },
+    {
+      "epoch": 3.533333333333333,
+      "grad_norm": 4.15081787109375,
+      "learning_rate": 5.8718861209964415e-05,
+      "loss": 0.4457,
+      "step": 7950
+    },
+    {
+      "epoch": 3.533777777777778,
+      "grad_norm": 2.6434662342071533,
+      "learning_rate": 5.8701067615658365e-05,
+      "loss": 2.3563,
+      "step": 7951
+    },
+    {
+      "epoch": 3.534222222222222,
+      "grad_norm": 2.4858129024505615,
+      "learning_rate": 5.868327402135232e-05,
+      "loss": 1.8796,
+      "step": 7952
+    },
+    {
+      "epoch": 3.5346666666666664,
+      "grad_norm": 3.0322587490081787,
+      "learning_rate": 5.866548042704627e-05,
+      "loss": 1.9928,
+      "step": 7953
+    },
+    {
+      "epoch": 3.535111111111111,
+      "grad_norm": 2.7141330242156982,
+      "learning_rate": 5.8647686832740214e-05,
+      "loss": 1.6356,
+      "step": 7954
+    },
+    {
+      "epoch": 3.535555555555556,
+      "grad_norm": 3.070211172103882,
+      "learning_rate": 5.8629893238434164e-05,
+      "loss": 1.3506,
+      "step": 7955
+    },
+    {
+      "epoch": 3.536,
+      "grad_norm": 3.548081159591675,
+      "learning_rate": 5.8612099644128114e-05,
+      "loss": 1.8977,
+      "step": 7956
+    },
+    {
+      "epoch": 3.5364444444444443,
+      "grad_norm": 3.189389228820801,
+      "learning_rate": 5.859430604982207e-05,
+      "loss": 1.5622,
+      "step": 7957
+    },
+    {
+      "epoch": 3.536888888888889,
+      "grad_norm": 3.3881306648254395,
+      "learning_rate": 5.857651245551602e-05,
+      "loss": 1.7807,
+      "step": 7958
+    },
+    {
+      "epoch": 3.537333333333333,
+      "grad_norm": 2.211007595062256,
+      "learning_rate": 5.855871886120997e-05,
+      "loss": 0.688,
+      "step": 7959
+    },
+    {
+      "epoch": 3.537777777777778,
+      "grad_norm": 3.352668523788452,
+      "learning_rate": 5.854092526690391e-05,
+      "loss": 1.8142,
+      "step": 7960
+    },
+    {
+      "epoch": 3.538222222222222,
+      "grad_norm": 3.3707339763641357,
+      "learning_rate": 5.852313167259786e-05,
+      "loss": 1.3783,
+      "step": 7961
+    },
+    {
+      "epoch": 3.538666666666667,
+      "grad_norm": 3.5272951126098633,
+      "learning_rate": 5.850533807829182e-05,
+      "loss": 1.5301,
+      "step": 7962
+    },
+    {
+      "epoch": 3.539111111111111,
+      "grad_norm": 3.348961353302002,
+      "learning_rate": 5.848754448398577e-05,
+      "loss": 1.2704,
+      "step": 7963
+    },
+    {
+      "epoch": 3.5395555555555553,
+      "grad_norm": 3.5585920810699463,
+      "learning_rate": 5.846975088967972e-05,
+      "loss": 1.3461,
+      "step": 7964
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 3.355560064315796,
+      "learning_rate": 5.8451957295373675e-05,
+      "loss": 1.4307,
+      "step": 7965
+    },
+    {
+      "epoch": 3.5404444444444443,
+      "grad_norm": 3.4020283222198486,
+      "learning_rate": 5.8434163701067624e-05,
+      "loss": 1.4312,
+      "step": 7966
+    },
+    {
+      "epoch": 3.540888888888889,
+      "grad_norm": 3.8027524948120117,
+      "learning_rate": 5.841637010676157e-05,
+      "loss": 1.6558,
+      "step": 7967
+    },
+    {
+      "epoch": 3.541333333333333,
+      "grad_norm": 3.7693047523498535,
+      "learning_rate": 5.839857651245552e-05,
+      "loss": 1.2358,
+      "step": 7968
+    },
+    {
+      "epoch": 3.541777777777778,
+      "grad_norm": 0.21360933780670166,
+      "learning_rate": 5.838078291814947e-05,
+      "loss": 0.0254,
+      "step": 7969
+    },
+    {
+      "epoch": 3.542222222222222,
+      "grad_norm": 2.69741153717041,
+      "learning_rate": 5.836298932384342e-05,
+      "loss": 1.2167,
+      "step": 7970
+    },
+    {
+      "epoch": 3.5426666666666664,
+      "grad_norm": 2.8727669715881348,
+      "learning_rate": 5.834519572953737e-05,
+      "loss": 1.0769,
+      "step": 7971
+    },
+    {
+      "epoch": 3.543111111111111,
+      "grad_norm": 2.91815447807312,
+      "learning_rate": 5.832740213523132e-05,
+      "loss": 1.0958,
+      "step": 7972
+    },
+    {
+      "epoch": 3.543555555555556,
+      "grad_norm": 3.0180909633636475,
+      "learning_rate": 5.8309608540925266e-05,
+      "loss": 1.0416,
+      "step": 7973
+    },
+    {
+      "epoch": 3.544,
+      "grad_norm": 3.282245635986328,
+      "learning_rate": 5.8291814946619215e-05,
+      "loss": 1.1943,
+      "step": 7974
+    },
+    {
+      "epoch": 3.5444444444444443,
+      "grad_norm": 3.464338779449463,
+      "learning_rate": 5.8274021352313165e-05,
+      "loss": 1.088,
+      "step": 7975
+    },
+    {
+      "epoch": 3.544888888888889,
+      "grad_norm": 3.1577398777008057,
+      "learning_rate": 5.825622775800712e-05,
+      "loss": 1.3338,
+      "step": 7976
+    },
+    {
+      "epoch": 3.5453333333333332,
+      "grad_norm": 3.2722463607788086,
+      "learning_rate": 5.823843416370107e-05,
+      "loss": 1.2914,
+      "step": 7977
+    },
+    {
+      "epoch": 3.545777777777778,
+      "grad_norm": 3.671210527420044,
+      "learning_rate": 5.822064056939502e-05,
+      "loss": 1.3975,
+      "step": 7978
+    },
+    {
+      "epoch": 3.546222222222222,
+      "grad_norm": 3.583202600479126,
+      "learning_rate": 5.820284697508898e-05,
+      "loss": 1.2501,
+      "step": 7979
+    },
+    {
+      "epoch": 3.546666666666667,
+      "grad_norm": 2.0527262687683105,
+      "learning_rate": 5.8185053380782914e-05,
+      "loss": 0.5888,
+      "step": 7980
+    },
+    {
+      "epoch": 3.547111111111111,
+      "grad_norm": 3.583556890487671,
+      "learning_rate": 5.816725978647687e-05,
+      "loss": 1.0753,
+      "step": 7981
+    },
+    {
+      "epoch": 3.5475555555555554,
+      "grad_norm": 3.821152687072754,
+      "learning_rate": 5.814946619217082e-05,
+      "loss": 1.311,
+      "step": 7982
+    },
+    {
+      "epoch": 3.548,
+      "grad_norm": 3.369014263153076,
+      "learning_rate": 5.813167259786477e-05,
+      "loss": 1.2148,
+      "step": 7983
+    },
+    {
+      "epoch": 3.5484444444444443,
+      "grad_norm": 3.3493340015411377,
+      "learning_rate": 5.8113879003558726e-05,
+      "loss": 1.1789,
+      "step": 7984
+    },
+    {
+      "epoch": 3.548888888888889,
+      "grad_norm": 2.961345672607422,
+      "learning_rate": 5.8096085409252676e-05,
+      "loss": 1.0513,
+      "step": 7985
+    },
+    {
+      "epoch": 3.5493333333333332,
+      "grad_norm": 3.708880662918091,
+      "learning_rate": 5.807829181494662e-05,
+      "loss": 1.3949,
+      "step": 7986
+    },
+    {
+      "epoch": 3.549777777777778,
+      "grad_norm": 3.2943549156188965,
+      "learning_rate": 5.806049822064057e-05,
+      "loss": 1.1287,
+      "step": 7987
+    },
+    {
+      "epoch": 3.550222222222222,
+      "grad_norm": 3.0104446411132812,
+      "learning_rate": 5.804270462633452e-05,
+      "loss": 1.121,
+      "step": 7988
+    },
+    {
+      "epoch": 3.5506666666666664,
+      "grad_norm": 3.7163071632385254,
+      "learning_rate": 5.8024911032028475e-05,
+      "loss": 1.0807,
+      "step": 7989
+    },
+    {
+      "epoch": 3.551111111111111,
+      "grad_norm": 3.9415080547332764,
+      "learning_rate": 5.8007117437722425e-05,
+      "loss": 1.3248,
+      "step": 7990
+    },
+    {
+      "epoch": 3.551555555555556,
+      "grad_norm": 4.1947245597839355,
+      "learning_rate": 5.7989323843416374e-05,
+      "loss": 1.5827,
+      "step": 7991
+    },
+    {
+      "epoch": 3.552,
+      "grad_norm": 4.378093719482422,
+      "learning_rate": 5.797153024911033e-05,
+      "loss": 1.2251,
+      "step": 7992
+    },
+    {
+      "epoch": 3.5524444444444443,
+      "grad_norm": 3.898223638534546,
+      "learning_rate": 5.795373665480427e-05,
+      "loss": 1.4066,
+      "step": 7993
+    },
+    {
+      "epoch": 3.552888888888889,
+      "grad_norm": 4.775320529937744,
+      "learning_rate": 5.7935943060498223e-05,
+      "loss": 1.2938,
+      "step": 7994
+    },
+    {
+      "epoch": 3.5533333333333332,
+      "grad_norm": 4.527971267700195,
+      "learning_rate": 5.791814946619217e-05,
+      "loss": 1.3705,
+      "step": 7995
+    },
+    {
+      "epoch": 3.553777777777778,
+      "grad_norm": 4.258374214172363,
+      "learning_rate": 5.790035587188612e-05,
+      "loss": 1.2344,
+      "step": 7996
+    },
+    {
+      "epoch": 3.554222222222222,
+      "grad_norm": 3.4799654483795166,
+      "learning_rate": 5.788256227758008e-05,
+      "loss": 1.1858,
+      "step": 7997
+    },
+    {
+      "epoch": 3.554666666666667,
+      "grad_norm": 5.774474620819092,
+      "learning_rate": 5.786476868327403e-05,
+      "loss": 1.6127,
+      "step": 7998
+    },
+    {
+      "epoch": 3.555111111111111,
+      "grad_norm": 4.93737268447876,
+      "learning_rate": 5.784697508896797e-05,
+      "loss": 1.2544,
+      "step": 7999
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "grad_norm": 3.601134777069092,
+      "learning_rate": 5.782918149466192e-05,
+      "loss": 0.5138,
+      "step": 8000
+    },
+    {
+      "epoch": 3.556,
+      "grad_norm": 2.7162842750549316,
+      "learning_rate": 5.781138790035587e-05,
+      "loss": 2.1305,
+      "step": 8001
+    },
+    {
+      "epoch": 3.5564444444444443,
+      "grad_norm": 2.303110361099243,
+      "learning_rate": 5.779359430604983e-05,
+      "loss": 1.6483,
+      "step": 8002
+    },
+    {
+      "epoch": 3.556888888888889,
+      "grad_norm": 2.755901575088501,
+      "learning_rate": 5.777580071174378e-05,
+      "loss": 1.8632,
+      "step": 8003
+    },
+    {
+      "epoch": 3.5573333333333332,
+      "grad_norm": 1.7043906450271606,
+      "learning_rate": 5.775800711743773e-05,
+      "loss": 0.9941,
+      "step": 8004
+    },
+    {
+      "epoch": 3.557777777777778,
+      "grad_norm": 1.6068474054336548,
+      "learning_rate": 5.774021352313167e-05,
+      "loss": 0.6672,
+      "step": 8005
+    },
+    {
+      "epoch": 3.558222222222222,
+      "grad_norm": 2.530790328979492,
+      "learning_rate": 5.772241992882562e-05,
+      "loss": 1.2947,
+      "step": 8006
+    },
+    {
+      "epoch": 3.5586666666666664,
+      "grad_norm": 3.187232732772827,
+      "learning_rate": 5.7704626334519577e-05,
+      "loss": 1.9547,
+      "step": 8007
+    },
+    {
+      "epoch": 3.559111111111111,
+      "grad_norm": 2.78572416305542,
+      "learning_rate": 5.7686832740213526e-05,
+      "loss": 1.5949,
+      "step": 8008
+    },
+    {
+      "epoch": 3.559555555555556,
+      "grad_norm": 3.9255530834198,
+      "learning_rate": 5.7669039145907476e-05,
+      "loss": 1.6875,
+      "step": 8009
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 2.8568902015686035,
+      "learning_rate": 5.765124555160143e-05,
+      "loss": 1.7145,
+      "step": 8010
+    },
+    {
+      "epoch": 3.5604444444444443,
+      "grad_norm": 2.9281044006347656,
+      "learning_rate": 5.763345195729538e-05,
+      "loss": 1.518,
+      "step": 8011
+    },
+    {
+      "epoch": 3.560888888888889,
+      "grad_norm": 2.7695422172546387,
+      "learning_rate": 5.7615658362989325e-05,
+      "loss": 1.665,
+      "step": 8012
+    },
+    {
+      "epoch": 3.5613333333333332,
+      "grad_norm": 3.807720899581909,
+      "learning_rate": 5.7597864768683275e-05,
+      "loss": 1.395,
+      "step": 8013
+    },
+    {
+      "epoch": 3.561777777777778,
+      "grad_norm": 3.4705801010131836,
+      "learning_rate": 5.7580071174377225e-05,
+      "loss": 1.4524,
+      "step": 8014
+    },
+    {
+      "epoch": 3.562222222222222,
+      "grad_norm": 2.88387131690979,
+      "learning_rate": 5.756227758007118e-05,
+      "loss": 1.1103,
+      "step": 8015
+    },
+    {
+      "epoch": 3.562666666666667,
+      "grad_norm": 2.9393246173858643,
+      "learning_rate": 5.754448398576513e-05,
+      "loss": 1.0856,
+      "step": 8016
+    },
+    {
+      "epoch": 3.563111111111111,
+      "grad_norm": 3.0343334674835205,
+      "learning_rate": 5.752669039145908e-05,
+      "loss": 1.3709,
+      "step": 8017
+    },
+    {
+      "epoch": 3.5635555555555554,
+      "grad_norm": 2.985978126525879,
+      "learning_rate": 5.7508896797153023e-05,
+      "loss": 1.36,
+      "step": 8018
+    },
+    {
+      "epoch": 3.564,
+      "grad_norm": 3.5497658252716064,
+      "learning_rate": 5.749110320284697e-05,
+      "loss": 1.4632,
+      "step": 8019
+    },
+    {
+      "epoch": 3.5644444444444443,
+      "grad_norm": 2.939556837081909,
+      "learning_rate": 5.747330960854092e-05,
+      "loss": 1.417,
+      "step": 8020
+    },
+    {
+      "epoch": 3.564888888888889,
+      "grad_norm": 2.9384403228759766,
+      "learning_rate": 5.745551601423488e-05,
+      "loss": 0.9369,
+      "step": 8021
+    },
+    {
+      "epoch": 3.5653333333333332,
+      "grad_norm": 3.302537441253662,
+      "learning_rate": 5.743772241992883e-05,
+      "loss": 1.592,
+      "step": 8022
+    },
+    {
+      "epoch": 3.565777777777778,
+      "grad_norm": 3.0296859741210938,
+      "learning_rate": 5.741992882562278e-05,
+      "loss": 1.3002,
+      "step": 8023
+    },
+    {
+      "epoch": 3.566222222222222,
+      "grad_norm": 2.4445323944091797,
+      "learning_rate": 5.7402135231316735e-05,
+      "loss": 1.0328,
+      "step": 8024
+    },
+    {
+      "epoch": 3.5666666666666664,
+      "grad_norm": 2.847886562347412,
+      "learning_rate": 5.738434163701067e-05,
+      "loss": 1.151,
+      "step": 8025
+    },
+    {
+      "epoch": 3.567111111111111,
+      "grad_norm": 3.6882436275482178,
+      "learning_rate": 5.736654804270463e-05,
+      "loss": 1.717,
+      "step": 8026
+    },
+    {
+      "epoch": 3.567555555555556,
+      "grad_norm": 2.9185895919799805,
+      "learning_rate": 5.734875444839858e-05,
+      "loss": 1.4205,
+      "step": 8027
+    },
+    {
+      "epoch": 3.568,
+      "grad_norm": 3.1053783893585205,
+      "learning_rate": 5.733096085409253e-05,
+      "loss": 1.2341,
+      "step": 8028
+    },
+    {
+      "epoch": 3.5684444444444443,
+      "grad_norm": 3.3123791217803955,
+      "learning_rate": 5.7313167259786484e-05,
+      "loss": 1.1749,
+      "step": 8029
+    },
+    {
+      "epoch": 3.568888888888889,
+      "grad_norm": 3.7100796699523926,
+      "learning_rate": 5.7295373665480434e-05,
+      "loss": 1.2908,
+      "step": 8030
+    },
+    {
+      "epoch": 3.5693333333333332,
+      "grad_norm": 2.9911487102508545,
+      "learning_rate": 5.727758007117438e-05,
+      "loss": 0.8336,
+      "step": 8031
+    },
+    {
+      "epoch": 3.569777777777778,
+      "grad_norm": 4.015035629272461,
+      "learning_rate": 5.7259786476868326e-05,
+      "loss": 1.8666,
+      "step": 8032
+    },
+    {
+      "epoch": 3.570222222222222,
+      "grad_norm": 3.5821290016174316,
+      "learning_rate": 5.7241992882562276e-05,
+      "loss": 1.1474,
+      "step": 8033
+    },
+    {
+      "epoch": 3.570666666666667,
+      "grad_norm": 3.538398265838623,
+      "learning_rate": 5.722419928825623e-05,
+      "loss": 0.8516,
+      "step": 8034
+    },
+    {
+      "epoch": 3.571111111111111,
+      "grad_norm": 3.570612907409668,
+      "learning_rate": 5.720640569395018e-05,
+      "loss": 1.4082,
+      "step": 8035
+    },
+    {
+      "epoch": 3.5715555555555554,
+      "grad_norm": 2.5072994232177734,
+      "learning_rate": 5.718861209964413e-05,
+      "loss": 0.6298,
+      "step": 8036
+    },
+    {
+      "epoch": 3.572,
+      "grad_norm": 3.3981175422668457,
+      "learning_rate": 5.717081850533809e-05,
+      "loss": 1.1501,
+      "step": 8037
+    },
+    {
+      "epoch": 3.5724444444444443,
+      "grad_norm": 4.341430187225342,
+      "learning_rate": 5.7153024911032025e-05,
+      "loss": 1.4794,
+      "step": 8038
+    },
+    {
+      "epoch": 3.572888888888889,
+      "grad_norm": 4.046430587768555,
+      "learning_rate": 5.713523131672598e-05,
+      "loss": 1.3761,
+      "step": 8039
+    },
+    {
+      "epoch": 3.5733333333333333,
+      "grad_norm": 3.6594674587249756,
+      "learning_rate": 5.711743772241993e-05,
+      "loss": 1.5613,
+      "step": 8040
+    },
+    {
+      "epoch": 3.573777777777778,
+      "grad_norm": 4.007656574249268,
+      "learning_rate": 5.709964412811388e-05,
+      "loss": 1.3619,
+      "step": 8041
+    },
+    {
+      "epoch": 3.574222222222222,
+      "grad_norm": 4.279335021972656,
+      "learning_rate": 5.708185053380784e-05,
+      "loss": 1.6566,
+      "step": 8042
+    },
+    {
+      "epoch": 3.5746666666666664,
+      "grad_norm": 3.4057838916778564,
+      "learning_rate": 5.706405693950179e-05,
+      "loss": 1.2244,
+      "step": 8043
+    },
+    {
+      "epoch": 3.575111111111111,
+      "grad_norm": 3.479970932006836,
+      "learning_rate": 5.704626334519573e-05,
+      "loss": 1.4622,
+      "step": 8044
+    },
+    {
+      "epoch": 3.575555555555556,
+      "grad_norm": 3.1348230838775635,
+      "learning_rate": 5.702846975088968e-05,
+      "loss": 1.2252,
+      "step": 8045
+    },
+    {
+      "epoch": 3.576,
+      "grad_norm": 3.5639426708221436,
+      "learning_rate": 5.701067615658363e-05,
+      "loss": 1.0362,
+      "step": 8046
+    },
+    {
+      "epoch": 3.5764444444444443,
+      "grad_norm": 3.8750410079956055,
+      "learning_rate": 5.6992882562277586e-05,
+      "loss": 0.9873,
+      "step": 8047
+    },
+    {
+      "epoch": 3.576888888888889,
+      "grad_norm": 4.268735885620117,
+      "learning_rate": 5.6975088967971535e-05,
+      "loss": 1.7215,
+      "step": 8048
+    },
+    {
+      "epoch": 3.5773333333333333,
+      "grad_norm": 3.788827657699585,
+      "learning_rate": 5.6957295373665485e-05,
+      "loss": 1.0994,
+      "step": 8049
+    },
+    {
+      "epoch": 3.5777777777777775,
+      "grad_norm": 3.688361644744873,
+      "learning_rate": 5.693950177935944e-05,
+      "loss": 0.2493,
+      "step": 8050
+    },
+    {
+      "epoch": 3.578222222222222,
+      "grad_norm": 2.1738009452819824,
+      "learning_rate": 5.692170818505338e-05,
+      "loss": 1.7641,
+      "step": 8051
+    },
+    {
+      "epoch": 3.578666666666667,
+      "grad_norm": 2.439173698425293,
+      "learning_rate": 5.6903914590747334e-05,
+      "loss": 2.0032,
+      "step": 8052
+    },
+    {
+      "epoch": 3.579111111111111,
+      "grad_norm": 2.087841749191284,
+      "learning_rate": 5.6886120996441284e-05,
+      "loss": 1.0669,
+      "step": 8053
+    },
+    {
+      "epoch": 3.5795555555555554,
+      "grad_norm": 2.6237778663635254,
+      "learning_rate": 5.6868327402135234e-05,
+      "loss": 1.5711,
+      "step": 8054
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 2.9692187309265137,
+      "learning_rate": 5.685053380782919e-05,
+      "loss": 1.553,
+      "step": 8055
+    },
+    {
+      "epoch": 3.5804444444444443,
+      "grad_norm": 2.481410264968872,
+      "learning_rate": 5.683274021352314e-05,
+      "loss": 0.8767,
+      "step": 8056
+    },
+    {
+      "epoch": 3.580888888888889,
+      "grad_norm": 3.197645425796509,
+      "learning_rate": 5.681494661921708e-05,
+      "loss": 1.7859,
+      "step": 8057
+    },
+    {
+      "epoch": 3.5813333333333333,
+      "grad_norm": 3.208965301513672,
+      "learning_rate": 5.679715302491103e-05,
+      "loss": 1.5878,
+      "step": 8058
+    },
+    {
+      "epoch": 3.581777777777778,
+      "grad_norm": 3.1050381660461426,
+      "learning_rate": 5.677935943060498e-05,
+      "loss": 1.4308,
+      "step": 8059
+    },
+    {
+      "epoch": 3.582222222222222,
+      "grad_norm": 3.836418867111206,
+      "learning_rate": 5.676156583629894e-05,
+      "loss": 1.6935,
+      "step": 8060
+    },
+    {
+      "epoch": 3.5826666666666664,
+      "grad_norm": 3.386842727661133,
+      "learning_rate": 5.674377224199289e-05,
+      "loss": 1.5557,
+      "step": 8061
+    },
+    {
+      "epoch": 3.583111111111111,
+      "grad_norm": 2.919567108154297,
+      "learning_rate": 5.672597864768684e-05,
+      "loss": 1.1336,
+      "step": 8062
+    },
+    {
+      "epoch": 3.583555555555556,
+      "grad_norm": 3.271778106689453,
+      "learning_rate": 5.670818505338078e-05,
+      "loss": 1.3239,
+      "step": 8063
+    },
+    {
+      "epoch": 3.584,
+      "grad_norm": 4.067761421203613,
+      "learning_rate": 5.669039145907473e-05,
+      "loss": 1.423,
+      "step": 8064
+    },
+    {
+      "epoch": 3.5844444444444443,
+      "grad_norm": 3.229156970977783,
+      "learning_rate": 5.667259786476868e-05,
+      "loss": 1.3032,
+      "step": 8065
+    },
+    {
+      "epoch": 3.584888888888889,
+      "grad_norm": 3.7134037017822266,
+      "learning_rate": 5.665480427046264e-05,
+      "loss": 1.2842,
+      "step": 8066
+    },
+    {
+      "epoch": 3.5853333333333333,
+      "grad_norm": 3.65313982963562,
+      "learning_rate": 5.663701067615659e-05,
+      "loss": 1.4233,
+      "step": 8067
+    },
+    {
+      "epoch": 3.5857777777777775,
+      "grad_norm": 3.398033857345581,
+      "learning_rate": 5.661921708185054e-05,
+      "loss": 1.3633,
+      "step": 8068
+    },
+    {
+      "epoch": 3.586222222222222,
+      "grad_norm": 3.1917221546173096,
+      "learning_rate": 5.660142348754449e-05,
+      "loss": 1.0808,
+      "step": 8069
+    },
+    {
+      "epoch": 3.586666666666667,
+      "grad_norm": 3.6049952507019043,
+      "learning_rate": 5.658362989323843e-05,
+      "loss": 1.7095,
+      "step": 8070
+    },
+    {
+      "epoch": 3.587111111111111,
+      "grad_norm": 3.4335100650787354,
+      "learning_rate": 5.6565836298932386e-05,
+      "loss": 1.3466,
+      "step": 8071
+    },
+    {
+      "epoch": 3.5875555555555554,
+      "grad_norm": 3.463630437850952,
+      "learning_rate": 5.6548042704626336e-05,
+      "loss": 1.2838,
+      "step": 8072
+    },
+    {
+      "epoch": 3.588,
+      "grad_norm": 3.933293104171753,
+      "learning_rate": 5.6530249110320285e-05,
+      "loss": 1.4277,
+      "step": 8073
+    },
+    {
+      "epoch": 3.5884444444444443,
+      "grad_norm": 3.340559482574463,
+      "learning_rate": 5.651245551601424e-05,
+      "loss": 1.2219,
+      "step": 8074
+    },
+    {
+      "epoch": 3.588888888888889,
+      "grad_norm": 4.030251979827881,
+      "learning_rate": 5.649466192170819e-05,
+      "loss": 1.4176,
+      "step": 8075
+    },
+    {
+      "epoch": 3.5893333333333333,
+      "grad_norm": 3.1213886737823486,
+      "learning_rate": 5.6476868327402134e-05,
+      "loss": 1.2696,
+      "step": 8076
+    },
+    {
+      "epoch": 3.589777777777778,
+      "grad_norm": 2.9783213138580322,
+      "learning_rate": 5.6459074733096084e-05,
+      "loss": 1.2995,
+      "step": 8077
+    },
+    {
+      "epoch": 3.590222222222222,
+      "grad_norm": 2.2861247062683105,
+      "learning_rate": 5.6441281138790034e-05,
+      "loss": 0.642,
+      "step": 8078
+    },
+    {
+      "epoch": 3.5906666666666665,
+      "grad_norm": 3.1233749389648438,
+      "learning_rate": 5.642348754448399e-05,
+      "loss": 0.9259,
+      "step": 8079
+    },
+    {
+      "epoch": 3.591111111111111,
+      "grad_norm": 3.5224452018737793,
+      "learning_rate": 5.640569395017794e-05,
+      "loss": 1.4473,
+      "step": 8080
+    },
+    {
+      "epoch": 3.5915555555555554,
+      "grad_norm": 3.7169413566589355,
+      "learning_rate": 5.638790035587189e-05,
+      "loss": 1.1774,
+      "step": 8081
+    },
+    {
+      "epoch": 3.592,
+      "grad_norm": 3.676466226577759,
+      "learning_rate": 5.6370106761565846e-05,
+      "loss": 1.4331,
+      "step": 8082
+    },
+    {
+      "epoch": 3.5924444444444443,
+      "grad_norm": 3.7899556159973145,
+      "learning_rate": 5.635231316725978e-05,
+      "loss": 1.6973,
+      "step": 8083
+    },
+    {
+      "epoch": 3.592888888888889,
+      "grad_norm": 4.713149070739746,
+      "learning_rate": 5.633451957295374e-05,
+      "loss": 1.4462,
+      "step": 8084
+    },
+    {
+      "epoch": 3.5933333333333333,
+      "grad_norm": 3.4206809997558594,
+      "learning_rate": 5.631672597864769e-05,
+      "loss": 1.2807,
+      "step": 8085
+    },
+    {
+      "epoch": 3.5937777777777775,
+      "grad_norm": 3.8811004161834717,
+      "learning_rate": 5.629893238434164e-05,
+      "loss": 1.2656,
+      "step": 8086
+    },
+    {
+      "epoch": 3.594222222222222,
+      "grad_norm": 3.997771978378296,
+      "learning_rate": 5.6281138790035595e-05,
+      "loss": 1.5387,
+      "step": 8087
+    },
+    {
+      "epoch": 3.594666666666667,
+      "grad_norm": 3.7929420471191406,
+      "learning_rate": 5.6263345195729545e-05,
+      "loss": 1.2682,
+      "step": 8088
+    },
+    {
+      "epoch": 3.595111111111111,
+      "grad_norm": 3.0101442337036133,
+      "learning_rate": 5.624555160142349e-05,
+      "loss": 0.7153,
+      "step": 8089
+    },
+    {
+      "epoch": 3.5955555555555554,
+      "grad_norm": 3.637249708175659,
+      "learning_rate": 5.622775800711744e-05,
+      "loss": 1.3984,
+      "step": 8090
+    },
+    {
+      "epoch": 3.596,
+      "grad_norm": 4.056656360626221,
+      "learning_rate": 5.620996441281139e-05,
+      "loss": 1.3663,
+      "step": 8091
+    },
+    {
+      "epoch": 3.5964444444444443,
+      "grad_norm": 3.1671206951141357,
+      "learning_rate": 5.6192170818505344e-05,
+      "loss": 1.1746,
+      "step": 8092
+    },
+    {
+      "epoch": 3.596888888888889,
+      "grad_norm": 4.801858901977539,
+      "learning_rate": 5.617437722419929e-05,
+      "loss": 1.7342,
+      "step": 8093
+    },
+    {
+      "epoch": 3.5973333333333333,
+      "grad_norm": 3.729170322418213,
+      "learning_rate": 5.615658362989324e-05,
+      "loss": 0.959,
+      "step": 8094
+    },
+    {
+      "epoch": 3.597777777777778,
+      "grad_norm": 3.9978530406951904,
+      "learning_rate": 5.61387900355872e-05,
+      "loss": 1.2262,
+      "step": 8095
+    },
+    {
+      "epoch": 3.598222222222222,
+      "grad_norm": 3.4558112621307373,
+      "learning_rate": 5.6120996441281136e-05,
+      "loss": 1.0928,
+      "step": 8096
+    },
+    {
+      "epoch": 3.5986666666666665,
+      "grad_norm": 3.660371780395508,
+      "learning_rate": 5.610320284697509e-05,
+      "loss": 1.1768,
+      "step": 8097
+    },
+    {
+      "epoch": 3.599111111111111,
+      "grad_norm": 5.200618743896484,
+      "learning_rate": 5.608540925266904e-05,
+      "loss": 1.0061,
+      "step": 8098
+    },
+    {
+      "epoch": 3.5995555555555554,
+      "grad_norm": 4.098397731781006,
+      "learning_rate": 5.606761565836299e-05,
+      "loss": 1.1386,
+      "step": 8099
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.963202714920044,
+      "learning_rate": 5.604982206405695e-05,
+      "loss": 0.4808,
+      "step": 8100
+    },
+    {
+      "epoch": 3.6004444444444443,
+      "grad_norm": 1.6309106349945068,
+      "learning_rate": 5.60320284697509e-05,
+      "loss": 0.8084,
+      "step": 8101
+    },
+    {
+      "epoch": 3.600888888888889,
+      "grad_norm": 3.205033779144287,
+      "learning_rate": 5.601423487544484e-05,
+      "loss": 1.2626,
+      "step": 8102
+    },
+    {
+      "epoch": 3.6013333333333333,
+      "grad_norm": 3.2114498615264893,
+      "learning_rate": 5.599644128113879e-05,
+      "loss": 1.5981,
+      "step": 8103
+    },
+    {
+      "epoch": 3.6017777777777775,
+      "grad_norm": 3.0003700256347656,
+      "learning_rate": 5.597864768683274e-05,
+      "loss": 1.3144,
+      "step": 8104
+    },
+    {
+      "epoch": 3.602222222222222,
+      "grad_norm": 3.0486159324645996,
+      "learning_rate": 5.59608540925267e-05,
+      "loss": 1.3489,
+      "step": 8105
+    },
+    {
+      "epoch": 3.602666666666667,
+      "grad_norm": 2.8934648036956787,
+      "learning_rate": 5.5943060498220646e-05,
+      "loss": 1.316,
+      "step": 8106
+    },
+    {
+      "epoch": 3.603111111111111,
+      "grad_norm": 3.1762866973876953,
+      "learning_rate": 5.5925266903914596e-05,
+      "loss": 1.259,
+      "step": 8107
+    },
+    {
+      "epoch": 3.6035555555555554,
+      "grad_norm": 2.9977829456329346,
+      "learning_rate": 5.5907473309608546e-05,
+      "loss": 1.0371,
+      "step": 8108
+    },
+    {
+      "epoch": 3.604,
+      "grad_norm": 3.3716068267822266,
+      "learning_rate": 5.588967971530249e-05,
+      "loss": 1.4329,
+      "step": 8109
+    },
+    {
+      "epoch": 3.6044444444444443,
+      "grad_norm": 3.4693422317504883,
+      "learning_rate": 5.587188612099644e-05,
+      "loss": 1.3105,
+      "step": 8110
+    },
+    {
+      "epoch": 3.604888888888889,
+      "grad_norm": 3.0525338649749756,
+      "learning_rate": 5.5854092526690395e-05,
+      "loss": 1.3395,
+      "step": 8111
+    },
+    {
+      "epoch": 3.6053333333333333,
+      "grad_norm": 3.2203667163848877,
+      "learning_rate": 5.5836298932384345e-05,
+      "loss": 1.4343,
+      "step": 8112
+    },
+    {
+      "epoch": 3.605777777777778,
+      "grad_norm": 3.3371267318725586,
+      "learning_rate": 5.5818505338078294e-05,
+      "loss": 1.3877,
+      "step": 8113
+    },
+    {
+      "epoch": 3.606222222222222,
+      "grad_norm": 4.093338489532471,
+      "learning_rate": 5.580071174377225e-05,
+      "loss": 1.558,
+      "step": 8114
+    },
+    {
+      "epoch": 3.6066666666666665,
+      "grad_norm": 3.266066312789917,
+      "learning_rate": 5.578291814946619e-05,
+      "loss": 0.9591,
+      "step": 8115
+    },
+    {
+      "epoch": 3.607111111111111,
+      "grad_norm": 3.600327730178833,
+      "learning_rate": 5.5765124555160144e-05,
+      "loss": 1.4223,
+      "step": 8116
+    },
+    {
+      "epoch": 3.6075555555555554,
+      "grad_norm": 3.469787359237671,
+      "learning_rate": 5.574733096085409e-05,
+      "loss": 1.1383,
+      "step": 8117
+    },
+    {
+      "epoch": 3.608,
+      "grad_norm": 3.263157844543457,
+      "learning_rate": 5.572953736654804e-05,
+      "loss": 1.5695,
+      "step": 8118
+    },
+    {
+      "epoch": 3.6084444444444443,
+      "grad_norm": 2.616588830947876,
+      "learning_rate": 5.5711743772242e-05,
+      "loss": 0.9983,
+      "step": 8119
+    },
+    {
+      "epoch": 3.608888888888889,
+      "grad_norm": 3.373994827270508,
+      "learning_rate": 5.569395017793595e-05,
+      "loss": 1.3782,
+      "step": 8120
+    },
+    {
+      "epoch": 3.6093333333333333,
+      "grad_norm": 3.624722957611084,
+      "learning_rate": 5.567615658362989e-05,
+      "loss": 1.3283,
+      "step": 8121
+    },
+    {
+      "epoch": 3.6097777777777775,
+      "grad_norm": 3.547624349594116,
+      "learning_rate": 5.565836298932384e-05,
+      "loss": 0.9186,
+      "step": 8122
+    },
+    {
+      "epoch": 3.610222222222222,
+      "grad_norm": 3.4890434741973877,
+      "learning_rate": 5.564056939501779e-05,
+      "loss": 1.514,
+      "step": 8123
+    },
+    {
+      "epoch": 3.610666666666667,
+      "grad_norm": 3.2042863368988037,
+      "learning_rate": 5.562277580071175e-05,
+      "loss": 1.2029,
+      "step": 8124
+    },
+    {
+      "epoch": 3.611111111111111,
+      "grad_norm": 3.498734712600708,
+      "learning_rate": 5.56049822064057e-05,
+      "loss": 1.2113,
+      "step": 8125
+    },
+    {
+      "epoch": 3.6115555555555554,
+      "grad_norm": 3.9023139476776123,
+      "learning_rate": 5.558718861209965e-05,
+      "loss": 1.1575,
+      "step": 8126
+    },
+    {
+      "epoch": 3.612,
+      "grad_norm": 3.4858694076538086,
+      "learning_rate": 5.5569395017793604e-05,
+      "loss": 1.3176,
+      "step": 8127
+    },
+    {
+      "epoch": 3.6124444444444443,
+      "grad_norm": 3.3926501274108887,
+      "learning_rate": 5.555160142348754e-05,
+      "loss": 1.52,
+      "step": 8128
+    },
+    {
+      "epoch": 3.612888888888889,
+      "grad_norm": 3.7569804191589355,
+      "learning_rate": 5.55338078291815e-05,
+      "loss": 1.2774,
+      "step": 8129
+    },
+    {
+      "epoch": 3.6133333333333333,
+      "grad_norm": 3.865650177001953,
+      "learning_rate": 5.5516014234875446e-05,
+      "loss": 1.4652,
+      "step": 8130
+    },
+    {
+      "epoch": 3.613777777777778,
+      "grad_norm": 3.499619483947754,
+      "learning_rate": 5.5498220640569396e-05,
+      "loss": 1.3073,
+      "step": 8131
+    },
+    {
+      "epoch": 3.6142222222222222,
+      "grad_norm": 3.366881847381592,
+      "learning_rate": 5.548042704626335e-05,
+      "loss": 1.1339,
+      "step": 8132
+    },
+    {
+      "epoch": 3.6146666666666665,
+      "grad_norm": 3.8024709224700928,
+      "learning_rate": 5.54626334519573e-05,
+      "loss": 1.4588,
+      "step": 8133
+    },
+    {
+      "epoch": 3.615111111111111,
+      "grad_norm": 4.1196160316467285,
+      "learning_rate": 5.5444839857651245e-05,
+      "loss": 1.2886,
+      "step": 8134
+    },
+    {
+      "epoch": 3.6155555555555554,
+      "grad_norm": 3.782890796661377,
+      "learning_rate": 5.5427046263345195e-05,
+      "loss": 1.1498,
+      "step": 8135
+    },
+    {
+      "epoch": 3.616,
+      "grad_norm": 3.4527840614318848,
+      "learning_rate": 5.5409252669039145e-05,
+      "loss": 1.4733,
+      "step": 8136
+    },
+    {
+      "epoch": 3.6164444444444444,
+      "grad_norm": 3.2366058826446533,
+      "learning_rate": 5.53914590747331e-05,
+      "loss": 1.0925,
+      "step": 8137
+    },
+    {
+      "epoch": 3.616888888888889,
+      "grad_norm": 2.8542094230651855,
+      "learning_rate": 5.537366548042705e-05,
+      "loss": 0.7504,
+      "step": 8138
+    },
+    {
+      "epoch": 3.6173333333333333,
+      "grad_norm": 3.7491934299468994,
+      "learning_rate": 5.5355871886121e-05,
+      "loss": 1.4432,
+      "step": 8139
+    },
+    {
+      "epoch": 3.6177777777777775,
+      "grad_norm": 3.4385619163513184,
+      "learning_rate": 5.533807829181496e-05,
+      "loss": 0.8893,
+      "step": 8140
+    },
+    {
+      "epoch": 3.6182222222222222,
+      "grad_norm": 4.420047283172607,
+      "learning_rate": 5.5320284697508893e-05,
+      "loss": 1.8875,
+      "step": 8141
+    },
+    {
+      "epoch": 3.618666666666667,
+      "grad_norm": 4.307175636291504,
+      "learning_rate": 5.530249110320285e-05,
+      "loss": 1.1576,
+      "step": 8142
+    },
+    {
+      "epoch": 3.619111111111111,
+      "grad_norm": 3.8173797130584717,
+      "learning_rate": 5.52846975088968e-05,
+      "loss": 1.5327,
+      "step": 8143
+    },
+    {
+      "epoch": 3.6195555555555554,
+      "grad_norm": 4.208653926849365,
+      "learning_rate": 5.526690391459075e-05,
+      "loss": 0.8899,
+      "step": 8144
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 4.356147766113281,
+      "learning_rate": 5.5249110320284706e-05,
+      "loss": 1.4298,
+      "step": 8145
+    },
+    {
+      "epoch": 3.6204444444444444,
+      "grad_norm": 3.923570394515991,
+      "learning_rate": 5.5231316725978656e-05,
+      "loss": 1.0489,
+      "step": 8146
+    },
+    {
+      "epoch": 3.620888888888889,
+      "grad_norm": 5.018566608428955,
+      "learning_rate": 5.52135231316726e-05,
+      "loss": 1.1978,
+      "step": 8147
+    },
+    {
+      "epoch": 3.6213333333333333,
+      "grad_norm": 3.8071272373199463,
+      "learning_rate": 5.519572953736655e-05,
+      "loss": 0.5866,
+      "step": 8148
+    },
+    {
+      "epoch": 3.621777777777778,
+      "grad_norm": 5.471420764923096,
+      "learning_rate": 5.51779359430605e-05,
+      "loss": 1.0767,
+      "step": 8149
+    },
+    {
+      "epoch": 3.6222222222222222,
+      "grad_norm": 4.0738606452941895,
+      "learning_rate": 5.5160142348754454e-05,
+      "loss": 0.5901,
+      "step": 8150
+    },
+    {
+      "epoch": 3.6226666666666665,
+      "grad_norm": 2.3456101417541504,
+      "learning_rate": 5.5142348754448404e-05,
+      "loss": 1.7003,
+      "step": 8151
+    },
+    {
+      "epoch": 3.623111111111111,
+      "grad_norm": 2.777620315551758,
+      "learning_rate": 5.5124555160142354e-05,
+      "loss": 1.9537,
+      "step": 8152
+    },
+    {
+      "epoch": 3.6235555555555554,
+      "grad_norm": 2.4514338970184326,
+      "learning_rate": 5.5106761565836304e-05,
+      "loss": 1.5011,
+      "step": 8153
+    },
+    {
+      "epoch": 3.624,
+      "grad_norm": 2.7258999347686768,
+      "learning_rate": 5.5088967971530247e-05,
+      "loss": 1.7209,
+      "step": 8154
+    },
+    {
+      "epoch": 3.6244444444444444,
+      "grad_norm": 2.463881492614746,
+      "learning_rate": 5.5071174377224196e-05,
+      "loss": 1.3477,
+      "step": 8155
+    },
+    {
+      "epoch": 3.624888888888889,
+      "grad_norm": 2.6786627769470215,
+      "learning_rate": 5.505338078291815e-05,
+      "loss": 1.4005,
+      "step": 8156
+    },
+    {
+      "epoch": 3.6253333333333333,
+      "grad_norm": 3.139360189437866,
+      "learning_rate": 5.50355871886121e-05,
+      "loss": 1.9946,
+      "step": 8157
+    },
+    {
+      "epoch": 3.6257777777777775,
+      "grad_norm": 2.5752453804016113,
+      "learning_rate": 5.501779359430605e-05,
+      "loss": 1.3608,
+      "step": 8158
+    },
+    {
+      "epoch": 3.6262222222222222,
+      "grad_norm": 3.0296263694763184,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 1.5828,
+      "step": 8159
+    },
+    {
+      "epoch": 3.626666666666667,
+      "grad_norm": 3.2608470916748047,
+      "learning_rate": 5.4982206405693945e-05,
+      "loss": 1.9236,
+      "step": 8160
+    },
+    {
+      "epoch": 3.627111111111111,
+      "grad_norm": 2.6645302772521973,
+      "learning_rate": 5.49644128113879e-05,
+      "loss": 1.4359,
+      "step": 8161
+    },
+    {
+      "epoch": 3.6275555555555554,
+      "grad_norm": 3.320736885070801,
+      "learning_rate": 5.494661921708185e-05,
+      "loss": 1.5185,
+      "step": 8162
+    },
+    {
+      "epoch": 3.628,
+      "grad_norm": 2.807404041290283,
+      "learning_rate": 5.49288256227758e-05,
+      "loss": 1.4373,
+      "step": 8163
+    },
+    {
+      "epoch": 3.6284444444444444,
+      "grad_norm": 2.9995763301849365,
+      "learning_rate": 5.491103202846976e-05,
+      "loss": 1.4947,
+      "step": 8164
+    },
+    {
+      "epoch": 3.628888888888889,
+      "grad_norm": 2.971586227416992,
+      "learning_rate": 5.489323843416371e-05,
+      "loss": 1.1496,
+      "step": 8165
+    },
+    {
+      "epoch": 3.6293333333333333,
+      "grad_norm": 3.100977659225464,
+      "learning_rate": 5.487544483985766e-05,
+      "loss": 1.1148,
+      "step": 8166
+    },
+    {
+      "epoch": 3.629777777777778,
+      "grad_norm": 3.1417391300201416,
+      "learning_rate": 5.48576512455516e-05,
+      "loss": 1.6336,
+      "step": 8167
+    },
+    {
+      "epoch": 3.6302222222222222,
+      "grad_norm": 3.098752975463867,
+      "learning_rate": 5.483985765124555e-05,
+      "loss": 1.2825,
+      "step": 8168
+    },
+    {
+      "epoch": 3.6306666666666665,
+      "grad_norm": 3.1631276607513428,
+      "learning_rate": 5.4822064056939506e-05,
+      "loss": 1.8255,
+      "step": 8169
+    },
+    {
+      "epoch": 3.631111111111111,
+      "grad_norm": 3.218193769454956,
+      "learning_rate": 5.4804270462633456e-05,
+      "loss": 1.6269,
+      "step": 8170
+    },
+    {
+      "epoch": 3.6315555555555554,
+      "grad_norm": 3.31066632270813,
+      "learning_rate": 5.4786476868327405e-05,
+      "loss": 1.3839,
+      "step": 8171
+    },
+    {
+      "epoch": 3.632,
+      "grad_norm": 4.438895225524902,
+      "learning_rate": 5.476868327402136e-05,
+      "loss": 1.9307,
+      "step": 8172
+    },
+    {
+      "epoch": 3.6324444444444444,
+      "grad_norm": 3.1939444541931152,
+      "learning_rate": 5.47508896797153e-05,
+      "loss": 1.3621,
+      "step": 8173
+    },
+    {
+      "epoch": 3.632888888888889,
+      "grad_norm": 2.8579087257385254,
+      "learning_rate": 5.4733096085409255e-05,
+      "loss": 0.8286,
+      "step": 8174
+    },
+    {
+      "epoch": 3.6333333333333333,
+      "grad_norm": 3.163404941558838,
+      "learning_rate": 5.4715302491103204e-05,
+      "loss": 1.3634,
+      "step": 8175
+    },
+    {
+      "epoch": 3.6337777777777776,
+      "grad_norm": 3.0559070110321045,
+      "learning_rate": 5.4697508896797154e-05,
+      "loss": 1.2169,
+      "step": 8176
+    },
+    {
+      "epoch": 3.6342222222222222,
+      "grad_norm": 3.1324009895324707,
+      "learning_rate": 5.467971530249111e-05,
+      "loss": 1.1629,
+      "step": 8177
+    },
+    {
+      "epoch": 3.634666666666667,
+      "grad_norm": 2.9012513160705566,
+      "learning_rate": 5.466192170818506e-05,
+      "loss": 0.8936,
+      "step": 8178
+    },
+    {
+      "epoch": 3.635111111111111,
+      "grad_norm": 3.157163381576538,
+      "learning_rate": 5.4644128113879e-05,
+      "loss": 1.2233,
+      "step": 8179
+    },
+    {
+      "epoch": 3.6355555555555554,
+      "grad_norm": 3.3618364334106445,
+      "learning_rate": 5.462633451957295e-05,
+      "loss": 1.3142,
+      "step": 8180
+    },
+    {
+      "epoch": 3.636,
+      "grad_norm": 3.3610551357269287,
+      "learning_rate": 5.46085409252669e-05,
+      "loss": 0.9224,
+      "step": 8181
+    },
+    {
+      "epoch": 3.6364444444444444,
+      "grad_norm": 3.8400661945343018,
+      "learning_rate": 5.459074733096086e-05,
+      "loss": 1.2544,
+      "step": 8182
+    },
+    {
+      "epoch": 3.6368888888888886,
+      "grad_norm": 3.3344969749450684,
+      "learning_rate": 5.457295373665481e-05,
+      "loss": 1.2641,
+      "step": 8183
+    },
+    {
+      "epoch": 3.6373333333333333,
+      "grad_norm": 3.121795415878296,
+      "learning_rate": 5.455516014234876e-05,
+      "loss": 1.1398,
+      "step": 8184
+    },
+    {
+      "epoch": 3.637777777777778,
+      "grad_norm": 4.248197555541992,
+      "learning_rate": 5.4537366548042715e-05,
+      "loss": 1.0294,
+      "step": 8185
+    },
+    {
+      "epoch": 3.6382222222222222,
+      "grad_norm": 3.8400375843048096,
+      "learning_rate": 5.451957295373665e-05,
+      "loss": 1.2639,
+      "step": 8186
+    },
+    {
+      "epoch": 3.6386666666666665,
+      "grad_norm": 3.7360832691192627,
+      "learning_rate": 5.450177935943061e-05,
+      "loss": 0.9932,
+      "step": 8187
+    },
+    {
+      "epoch": 3.639111111111111,
+      "grad_norm": 3.7453458309173584,
+      "learning_rate": 5.448398576512456e-05,
+      "loss": 1.2704,
+      "step": 8188
+    },
+    {
+      "epoch": 3.6395555555555554,
+      "grad_norm": 3.8645150661468506,
+      "learning_rate": 5.446619217081851e-05,
+      "loss": 0.682,
+      "step": 8189
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 3.455167055130005,
+      "learning_rate": 5.4448398576512464e-05,
+      "loss": 1.3446,
+      "step": 8190
+    },
+    {
+      "epoch": 3.6404444444444444,
+      "grad_norm": 3.7151405811309814,
+      "learning_rate": 5.443060498220641e-05,
+      "loss": 1.0171,
+      "step": 8191
+    },
+    {
+      "epoch": 3.640888888888889,
+      "grad_norm": 4.202325820922852,
+      "learning_rate": 5.4412811387900356e-05,
+      "loss": 1.26,
+      "step": 8192
+    },
+    {
+      "epoch": 3.6413333333333333,
+      "grad_norm": 4.061198711395264,
+      "learning_rate": 5.4395017793594306e-05,
+      "loss": 0.9919,
+      "step": 8193
+    },
+    {
+      "epoch": 3.6417777777777776,
+      "grad_norm": 4.6272501945495605,
+      "learning_rate": 5.4377224199288256e-05,
+      "loss": 1.2987,
+      "step": 8194
+    },
+    {
+      "epoch": 3.6422222222222222,
+      "grad_norm": 4.520051956176758,
+      "learning_rate": 5.4359430604982205e-05,
+      "loss": 1.0514,
+      "step": 8195
+    },
+    {
+      "epoch": 3.642666666666667,
+      "grad_norm": 3.6569225788116455,
+      "learning_rate": 5.434163701067616e-05,
+      "loss": 1.2133,
+      "step": 8196
+    },
+    {
+      "epoch": 3.643111111111111,
+      "grad_norm": 4.312341213226318,
+      "learning_rate": 5.432384341637011e-05,
+      "loss": 1.1944,
+      "step": 8197
+    },
+    {
+      "epoch": 3.6435555555555554,
+      "grad_norm": 4.46543025970459,
+      "learning_rate": 5.430604982206406e-05,
+      "loss": 1.3214,
+      "step": 8198
+    },
+    {
+      "epoch": 3.644,
+      "grad_norm": 5.549198627471924,
+      "learning_rate": 5.4288256227758004e-05,
+      "loss": 0.8821,
+      "step": 8199
+    },
+    {
+      "epoch": 3.6444444444444444,
+      "grad_norm": 4.581550121307373,
+      "learning_rate": 5.4270462633451954e-05,
+      "loss": 1.044,
+      "step": 8200
+    },
+    {
+      "epoch": 3.6448888888888886,
+      "grad_norm": 2.513296127319336,
+      "learning_rate": 5.425266903914591e-05,
+      "loss": 0.9842,
+      "step": 8201
+    },
+    {
+      "epoch": 3.6453333333333333,
+      "grad_norm": 2.9860239028930664,
+      "learning_rate": 5.423487544483986e-05,
+      "loss": 2.2302,
+      "step": 8202
+    },
+    {
+      "epoch": 3.645777777777778,
+      "grad_norm": 2.648104429244995,
+      "learning_rate": 5.421708185053381e-05,
+      "loss": 1.7132,
+      "step": 8203
+    },
+    {
+      "epoch": 3.6462222222222223,
+      "grad_norm": 2.892163038253784,
+      "learning_rate": 5.4199288256227767e-05,
+      "loss": 1.3049,
+      "step": 8204
+    },
+    {
+      "epoch": 3.6466666666666665,
+      "grad_norm": 3.270510196685791,
+      "learning_rate": 5.41814946619217e-05,
+      "loss": 1.8626,
+      "step": 8205
+    },
+    {
+      "epoch": 3.647111111111111,
+      "grad_norm": 3.1879770755767822,
+      "learning_rate": 5.416370106761566e-05,
+      "loss": 1.513,
+      "step": 8206
+    },
+    {
+      "epoch": 3.6475555555555554,
+      "grad_norm": 2.4823553562164307,
+      "learning_rate": 5.414590747330961e-05,
+      "loss": 0.8951,
+      "step": 8207
+    },
+    {
+      "epoch": 3.648,
+      "grad_norm": 2.580904722213745,
+      "learning_rate": 5.412811387900356e-05,
+      "loss": 1.0414,
+      "step": 8208
+    },
+    {
+      "epoch": 3.6484444444444444,
+      "grad_norm": 3.6623318195343018,
+      "learning_rate": 5.4110320284697515e-05,
+      "loss": 1.6687,
+      "step": 8209
+    },
+    {
+      "epoch": 3.648888888888889,
+      "grad_norm": 3.58868408203125,
+      "learning_rate": 5.4092526690391465e-05,
+      "loss": 1.4268,
+      "step": 8210
+    },
+    {
+      "epoch": 3.6493333333333333,
+      "grad_norm": 3.2081754207611084,
+      "learning_rate": 5.4074733096085415e-05,
+      "loss": 1.6204,
+      "step": 8211
+    },
+    {
+      "epoch": 3.6497777777777776,
+      "grad_norm": 3.3098785877227783,
+      "learning_rate": 5.405693950177936e-05,
+      "loss": 1.1347,
+      "step": 8212
+    },
+    {
+      "epoch": 3.6502222222222223,
+      "grad_norm": 3.5921061038970947,
+      "learning_rate": 5.403914590747331e-05,
+      "loss": 1.5662,
+      "step": 8213
+    },
+    {
+      "epoch": 3.6506666666666665,
+      "grad_norm": 3.2691173553466797,
+      "learning_rate": 5.4021352313167264e-05,
+      "loss": 1.2848,
+      "step": 8214
+    },
+    {
+      "epoch": 3.651111111111111,
+      "grad_norm": 3.106048583984375,
+      "learning_rate": 5.4003558718861213e-05,
+      "loss": 0.8942,
+      "step": 8215
+    },
+    {
+      "epoch": 3.6515555555555554,
+      "grad_norm": 6.022765636444092,
+      "learning_rate": 5.398576512455516e-05,
+      "loss": 1.1837,
+      "step": 8216
+    },
+    {
+      "epoch": 3.652,
+      "grad_norm": 3.3533272743225098,
+      "learning_rate": 5.396797153024912e-05,
+      "loss": 1.3667,
+      "step": 8217
+    },
+    {
+      "epoch": 3.6524444444444444,
+      "grad_norm": 3.1057679653167725,
+      "learning_rate": 5.3950177935943056e-05,
+      "loss": 1.3119,
+      "step": 8218
+    },
+    {
+      "epoch": 3.6528888888888886,
+      "grad_norm": 2.8598716259002686,
+      "learning_rate": 5.393238434163701e-05,
+      "loss": 1.0251,
+      "step": 8219
+    },
+    {
+      "epoch": 3.6533333333333333,
+      "grad_norm": 3.3327109813690186,
+      "learning_rate": 5.391459074733096e-05,
+      "loss": 1.1523,
+      "step": 8220
+    },
+    {
+      "epoch": 3.653777777777778,
+      "grad_norm": 2.8015899658203125,
+      "learning_rate": 5.389679715302491e-05,
+      "loss": 1.087,
+      "step": 8221
+    },
+    {
+      "epoch": 3.6542222222222223,
+      "grad_norm": 3.4534077644348145,
+      "learning_rate": 5.387900355871887e-05,
+      "loss": 1.1916,
+      "step": 8222
+    },
+    {
+      "epoch": 3.6546666666666665,
+      "grad_norm": 2.8814008235931396,
+      "learning_rate": 5.386120996441282e-05,
+      "loss": 1.1617,
+      "step": 8223
+    },
+    {
+      "epoch": 3.655111111111111,
+      "grad_norm": 3.4058141708374023,
+      "learning_rate": 5.384341637010677e-05,
+      "loss": 1.2716,
+      "step": 8224
+    },
+    {
+      "epoch": 3.6555555555555554,
+      "grad_norm": 3.901968479156494,
+      "learning_rate": 5.382562277580071e-05,
+      "loss": 1.5236,
+      "step": 8225
+    },
+    {
+      "epoch": 3.656,
+      "grad_norm": 2.169243097305298,
+      "learning_rate": 5.380782918149466e-05,
+      "loss": 0.5998,
+      "step": 8226
+    },
+    {
+      "epoch": 3.6564444444444444,
+      "grad_norm": 3.659686803817749,
+      "learning_rate": 5.379003558718862e-05,
+      "loss": 1.4257,
+      "step": 8227
+    },
+    {
+      "epoch": 3.656888888888889,
+      "grad_norm": 3.8471503257751465,
+      "learning_rate": 5.3772241992882567e-05,
+      "loss": 1.3645,
+      "step": 8228
+    },
+    {
+      "epoch": 3.6573333333333333,
+      "grad_norm": 2.410559892654419,
+      "learning_rate": 5.3754448398576516e-05,
+      "loss": 0.6455,
+      "step": 8229
+    },
+    {
+      "epoch": 3.6577777777777776,
+      "grad_norm": 3.5442912578582764,
+      "learning_rate": 5.373665480427047e-05,
+      "loss": 1.3689,
+      "step": 8230
+    },
+    {
+      "epoch": 3.6582222222222223,
+      "grad_norm": 3.4644775390625,
+      "learning_rate": 5.371886120996441e-05,
+      "loss": 1.197,
+      "step": 8231
+    },
+    {
+      "epoch": 3.6586666666666665,
+      "grad_norm": 3.7939751148223877,
+      "learning_rate": 5.3701067615658365e-05,
+      "loss": 1.1387,
+      "step": 8232
+    },
+    {
+      "epoch": 3.659111111111111,
+      "grad_norm": 3.8991305828094482,
+      "learning_rate": 5.3683274021352315e-05,
+      "loss": 1.1945,
+      "step": 8233
+    },
+    {
+      "epoch": 3.6595555555555555,
+      "grad_norm": 2.8764808177948,
+      "learning_rate": 5.3665480427046265e-05,
+      "loss": 0.8908,
+      "step": 8234
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.422370433807373,
+      "learning_rate": 5.364768683274022e-05,
+      "loss": 1.0134,
+      "step": 8235
+    },
+    {
+      "epoch": 3.6604444444444444,
+      "grad_norm": 4.0114874839782715,
+      "learning_rate": 5.362989323843417e-05,
+      "loss": 1.0199,
+      "step": 8236
+    },
+    {
+      "epoch": 3.6608888888888886,
+      "grad_norm": 4.0204362869262695,
+      "learning_rate": 5.3612099644128114e-05,
+      "loss": 1.447,
+      "step": 8237
+    },
+    {
+      "epoch": 3.6613333333333333,
+      "grad_norm": 3.3613667488098145,
+      "learning_rate": 5.3594306049822064e-05,
+      "loss": 0.8356,
+      "step": 8238
+    },
+    {
+      "epoch": 3.661777777777778,
+      "grad_norm": 5.9143571853637695,
+      "learning_rate": 5.3576512455516014e-05,
+      "loss": 1.2582,
+      "step": 8239
+    },
+    {
+      "epoch": 3.6622222222222223,
+      "grad_norm": 3.864518404006958,
+      "learning_rate": 5.355871886120996e-05,
+      "loss": 1.2546,
+      "step": 8240
+    },
+    {
+      "epoch": 3.6626666666666665,
+      "grad_norm": 3.446545362472534,
+      "learning_rate": 5.354092526690392e-05,
+      "loss": 0.9876,
+      "step": 8241
+    },
+    {
+      "epoch": 3.663111111111111,
+      "grad_norm": 4.813685417175293,
+      "learning_rate": 5.352313167259787e-05,
+      "loss": 1.1929,
+      "step": 8242
+    },
+    {
+      "epoch": 3.6635555555555555,
+      "grad_norm": 4.656696319580078,
+      "learning_rate": 5.350533807829182e-05,
+      "loss": 1.0747,
+      "step": 8243
+    },
+    {
+      "epoch": 3.664,
+      "grad_norm": 4.323572158813477,
+      "learning_rate": 5.348754448398576e-05,
+      "loss": 1.2275,
+      "step": 8244
+    },
+    {
+      "epoch": 3.6644444444444444,
+      "grad_norm": 3.6625609397888184,
+      "learning_rate": 5.346975088967971e-05,
+      "loss": 1.0316,
+      "step": 8245
+    },
+    {
+      "epoch": 3.664888888888889,
+      "grad_norm": 4.840065002441406,
+      "learning_rate": 5.345195729537367e-05,
+      "loss": 1.3223,
+      "step": 8246
+    },
+    {
+      "epoch": 3.6653333333333333,
+      "grad_norm": 4.664702415466309,
+      "learning_rate": 5.343416370106762e-05,
+      "loss": 1.4522,
+      "step": 8247
+    },
+    {
+      "epoch": 3.6657777777777776,
+      "grad_norm": 4.857150077819824,
+      "learning_rate": 5.341637010676157e-05,
+      "loss": 1.0819,
+      "step": 8248
+    },
+    {
+      "epoch": 3.6662222222222223,
+      "grad_norm": 10.743447303771973,
+      "learning_rate": 5.3398576512455524e-05,
+      "loss": 1.3214,
+      "step": 8249
+    },
+    {
+      "epoch": 3.6666666666666665,
+      "grad_norm": 4.244997024536133,
+      "learning_rate": 5.338078291814946e-05,
+      "loss": 0.3447,
+      "step": 8250
+    },
+    {
+      "epoch": 3.667111111111111,
+      "grad_norm": 2.1878910064697266,
+      "learning_rate": 5.336298932384342e-05,
+      "loss": 1.7063,
+      "step": 8251
+    },
+    {
+      "epoch": 3.6675555555555555,
+      "grad_norm": 2.090615749359131,
+      "learning_rate": 5.334519572953737e-05,
+      "loss": 0.9182,
+      "step": 8252
+    },
+    {
+      "epoch": 3.668,
+      "grad_norm": 2.9112021923065186,
+      "learning_rate": 5.3327402135231316e-05,
+      "loss": 1.575,
+      "step": 8253
+    },
+    {
+      "epoch": 3.6684444444444444,
+      "grad_norm": 3.180088996887207,
+      "learning_rate": 5.330960854092527e-05,
+      "loss": 1.6393,
+      "step": 8254
+    },
+    {
+      "epoch": 3.6688888888888886,
+      "grad_norm": 3.059612512588501,
+      "learning_rate": 5.329181494661922e-05,
+      "loss": 1.631,
+      "step": 8255
+    },
+    {
+      "epoch": 3.6693333333333333,
+      "grad_norm": 3.003873825073242,
+      "learning_rate": 5.327402135231317e-05,
+      "loss": 1.4932,
+      "step": 8256
+    },
+    {
+      "epoch": 3.669777777777778,
+      "grad_norm": 3.1199257373809814,
+      "learning_rate": 5.3256227758007115e-05,
+      "loss": 1.5315,
+      "step": 8257
+    },
+    {
+      "epoch": 3.6702222222222223,
+      "grad_norm": 2.0866105556488037,
+      "learning_rate": 5.3238434163701065e-05,
+      "loss": 0.8337,
+      "step": 8258
+    },
+    {
+      "epoch": 3.6706666666666665,
+      "grad_norm": 3.190763473510742,
+      "learning_rate": 5.322064056939502e-05,
+      "loss": 1.4458,
+      "step": 8259
+    },
+    {
+      "epoch": 3.671111111111111,
+      "grad_norm": 3.141622304916382,
+      "learning_rate": 5.320284697508897e-05,
+      "loss": 1.4373,
+      "step": 8260
+    },
+    {
+      "epoch": 3.6715555555555555,
+      "grad_norm": 3.208282232284546,
+      "learning_rate": 5.318505338078292e-05,
+      "loss": 1.2228,
+      "step": 8261
+    },
+    {
+      "epoch": 3.672,
+      "grad_norm": 3.1854054927825928,
+      "learning_rate": 5.316725978647688e-05,
+      "loss": 1.1494,
+      "step": 8262
+    },
+    {
+      "epoch": 3.6724444444444444,
+      "grad_norm": 3.755186080932617,
+      "learning_rate": 5.3149466192170814e-05,
+      "loss": 1.5219,
+      "step": 8263
+    },
+    {
+      "epoch": 3.672888888888889,
+      "grad_norm": 3.6373298168182373,
+      "learning_rate": 5.313167259786477e-05,
+      "loss": 1.2791,
+      "step": 8264
+    },
+    {
+      "epoch": 3.6733333333333333,
+      "grad_norm": 3.869194984436035,
+      "learning_rate": 5.311387900355872e-05,
+      "loss": 1.854,
+      "step": 8265
+    },
+    {
+      "epoch": 3.6737777777777776,
+      "grad_norm": 3.1836750507354736,
+      "learning_rate": 5.309608540925267e-05,
+      "loss": 1.3653,
+      "step": 8266
+    },
+    {
+      "epoch": 3.6742222222222223,
+      "grad_norm": 3.2657806873321533,
+      "learning_rate": 5.3078291814946626e-05,
+      "loss": 1.5467,
+      "step": 8267
+    },
+    {
+      "epoch": 3.6746666666666665,
+      "grad_norm": 1.893934726715088,
+      "learning_rate": 5.3060498220640576e-05,
+      "loss": 0.7488,
+      "step": 8268
+    },
+    {
+      "epoch": 3.675111111111111,
+      "grad_norm": 3.510976552963257,
+      "learning_rate": 5.3042704626334526e-05,
+      "loss": 1.148,
+      "step": 8269
+    },
+    {
+      "epoch": 3.6755555555555555,
+      "grad_norm": 2.769212007522583,
+      "learning_rate": 5.302491103202847e-05,
+      "loss": 0.9423,
+      "step": 8270
+    },
+    {
+      "epoch": 3.676,
+      "grad_norm": 3.1665525436401367,
+      "learning_rate": 5.300711743772242e-05,
+      "loss": 1.9088,
+      "step": 8271
+    },
+    {
+      "epoch": 3.6764444444444444,
+      "grad_norm": 3.130796432495117,
+      "learning_rate": 5.2989323843416375e-05,
+      "loss": 1.2455,
+      "step": 8272
+    },
+    {
+      "epoch": 3.6768888888888887,
+      "grad_norm": 3.2460105419158936,
+      "learning_rate": 5.2971530249110324e-05,
+      "loss": 1.2274,
+      "step": 8273
+    },
+    {
+      "epoch": 3.6773333333333333,
+      "grad_norm": 3.6489906311035156,
+      "learning_rate": 5.2953736654804274e-05,
+      "loss": 1.3301,
+      "step": 8274
+    },
+    {
+      "epoch": 3.677777777777778,
+      "grad_norm": 3.5762670040130615,
+      "learning_rate": 5.293594306049823e-05,
+      "loss": 1.9313,
+      "step": 8275
+    },
+    {
+      "epoch": 3.6782222222222223,
+      "grad_norm": 3.458677053451538,
+      "learning_rate": 5.291814946619217e-05,
+      "loss": 1.0342,
+      "step": 8276
+    },
+    {
+      "epoch": 3.6786666666666665,
+      "grad_norm": 3.369166612625122,
+      "learning_rate": 5.290035587188612e-05,
+      "loss": 1.7108,
+      "step": 8277
+    },
+    {
+      "epoch": 3.679111111111111,
+      "grad_norm": 2.5164918899536133,
+      "learning_rate": 5.288256227758007e-05,
+      "loss": 0.6384,
+      "step": 8278
+    },
+    {
+      "epoch": 3.6795555555555555,
+      "grad_norm": 4.323822975158691,
+      "learning_rate": 5.286476868327402e-05,
+      "loss": 1.6491,
+      "step": 8279
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 3.677583932876587,
+      "learning_rate": 5.284697508896798e-05,
+      "loss": 0.9492,
+      "step": 8280
+    },
+    {
+      "epoch": 3.6804444444444444,
+      "grad_norm": 4.226683616638184,
+      "learning_rate": 5.282918149466193e-05,
+      "loss": 1.3614,
+      "step": 8281
+    },
+    {
+      "epoch": 3.680888888888889,
+      "grad_norm": 3.287234306335449,
+      "learning_rate": 5.281138790035588e-05,
+      "loss": 1.3625,
+      "step": 8282
+    },
+    {
+      "epoch": 3.6813333333333333,
+      "grad_norm": 3.5409836769104004,
+      "learning_rate": 5.279359430604982e-05,
+      "loss": 1.4074,
+      "step": 8283
+    },
+    {
+      "epoch": 3.6817777777777776,
+      "grad_norm": 3.3012094497680664,
+      "learning_rate": 5.277580071174377e-05,
+      "loss": 1.1894,
+      "step": 8284
+    },
+    {
+      "epoch": 3.6822222222222223,
+      "grad_norm": 3.455919027328491,
+      "learning_rate": 5.275800711743772e-05,
+      "loss": 1.3859,
+      "step": 8285
+    },
+    {
+      "epoch": 3.6826666666666665,
+      "grad_norm": 3.6948232650756836,
+      "learning_rate": 5.274021352313168e-05,
+      "loss": 1.1195,
+      "step": 8286
+    },
+    {
+      "epoch": 3.6831111111111112,
+      "grad_norm": 3.614387273788452,
+      "learning_rate": 5.272241992882563e-05,
+      "loss": 1.3584,
+      "step": 8287
+    },
+    {
+      "epoch": 3.6835555555555555,
+      "grad_norm": 3.2227156162261963,
+      "learning_rate": 5.270462633451958e-05,
+      "loss": 0.3771,
+      "step": 8288
+    },
+    {
+      "epoch": 3.684,
+      "grad_norm": 2.9594337940216064,
+      "learning_rate": 5.268683274021352e-05,
+      "loss": 1.0337,
+      "step": 8289
+    },
+    {
+      "epoch": 3.6844444444444444,
+      "grad_norm": 3.800976276397705,
+      "learning_rate": 5.266903914590747e-05,
+      "loss": 0.7237,
+      "step": 8290
+    },
+    {
+      "epoch": 3.6848888888888887,
+      "grad_norm": 4.719809055328369,
+      "learning_rate": 5.2651245551601426e-05,
+      "loss": 1.2393,
+      "step": 8291
+    },
+    {
+      "epoch": 3.6853333333333333,
+      "grad_norm": 4.721984386444092,
+      "learning_rate": 5.2633451957295376e-05,
+      "loss": 1.2577,
+      "step": 8292
+    },
+    {
+      "epoch": 3.685777777777778,
+      "grad_norm": 3.1835551261901855,
+      "learning_rate": 5.2615658362989326e-05,
+      "loss": 0.9904,
+      "step": 8293
+    },
+    {
+      "epoch": 3.6862222222222223,
+      "grad_norm": 4.421992301940918,
+      "learning_rate": 5.259786476868328e-05,
+      "loss": 1.4825,
+      "step": 8294
+    },
+    {
+      "epoch": 3.6866666666666665,
+      "grad_norm": 4.140617370605469,
+      "learning_rate": 5.258007117437722e-05,
+      "loss": 1.1782,
+      "step": 8295
+    },
+    {
+      "epoch": 3.6871111111111112,
+      "grad_norm": 3.6903188228607178,
+      "learning_rate": 5.2562277580071175e-05,
+      "loss": 1.5041,
+      "step": 8296
+    },
+    {
+      "epoch": 3.6875555555555555,
+      "grad_norm": 3.7069318294525146,
+      "learning_rate": 5.2544483985765124e-05,
+      "loss": 1.1797,
+      "step": 8297
+    },
+    {
+      "epoch": 3.6879999999999997,
+      "grad_norm": 4.413390636444092,
+      "learning_rate": 5.2526690391459074e-05,
+      "loss": 0.9219,
+      "step": 8298
+    },
+    {
+      "epoch": 3.6884444444444444,
+      "grad_norm": 5.5441508293151855,
+      "learning_rate": 5.250889679715303e-05,
+      "loss": 1.0509,
+      "step": 8299
+    },
+    {
+      "epoch": 3.688888888888889,
+      "grad_norm": 0.8163884282112122,
+      "learning_rate": 5.249110320284698e-05,
+      "loss": 0.0719,
+      "step": 8300
+    },
+    {
+      "epoch": 3.6893333333333334,
+      "grad_norm": 1.9359767436981201,
+      "learning_rate": 5.247330960854093e-05,
+      "loss": 0.8754,
+      "step": 8301
+    },
+    {
+      "epoch": 3.6897777777777776,
+      "grad_norm": 2.754122495651245,
+      "learning_rate": 5.245551601423487e-05,
+      "loss": 1.6614,
+      "step": 8302
+    },
+    {
+      "epoch": 3.6902222222222223,
+      "grad_norm": 2.6773335933685303,
+      "learning_rate": 5.243772241992882e-05,
+      "loss": 1.7607,
+      "step": 8303
+    },
+    {
+      "epoch": 3.6906666666666665,
+      "grad_norm": 2.679157257080078,
+      "learning_rate": 5.241992882562278e-05,
+      "loss": 1.3,
+      "step": 8304
+    },
+    {
+      "epoch": 3.6911111111111112,
+      "grad_norm": 2.7156100273132324,
+      "learning_rate": 5.240213523131673e-05,
+      "loss": 1.5878,
+      "step": 8305
+    },
+    {
+      "epoch": 3.6915555555555555,
+      "grad_norm": 3.0638999938964844,
+      "learning_rate": 5.238434163701068e-05,
+      "loss": 2.0461,
+      "step": 8306
+    },
+    {
+      "epoch": 3.692,
+      "grad_norm": 2.997682571411133,
+      "learning_rate": 5.2366548042704635e-05,
+      "loss": 1.6546,
+      "step": 8307
+    },
+    {
+      "epoch": 3.6924444444444444,
+      "grad_norm": 3.044121265411377,
+      "learning_rate": 5.234875444839857e-05,
+      "loss": 1.6228,
+      "step": 8308
+    },
+    {
+      "epoch": 3.6928888888888887,
+      "grad_norm": 2.8701446056365967,
+      "learning_rate": 5.233096085409253e-05,
+      "loss": 1.7426,
+      "step": 8309
+    },
+    {
+      "epoch": 3.6933333333333334,
+      "grad_norm": 3.174617052078247,
+      "learning_rate": 5.231316725978648e-05,
+      "loss": 1.5458,
+      "step": 8310
+    },
+    {
+      "epoch": 3.693777777777778,
+      "grad_norm": 3.201929807662964,
+      "learning_rate": 5.229537366548043e-05,
+      "loss": 1.6514,
+      "step": 8311
+    },
+    {
+      "epoch": 3.6942222222222223,
+      "grad_norm": 3.799694538116455,
+      "learning_rate": 5.2277580071174384e-05,
+      "loss": 1.6626,
+      "step": 8312
+    },
+    {
+      "epoch": 3.6946666666666665,
+      "grad_norm": 3.5478615760803223,
+      "learning_rate": 5.2259786476868334e-05,
+      "loss": 1.3328,
+      "step": 8313
+    },
+    {
+      "epoch": 3.6951111111111112,
+      "grad_norm": 3.048062562942505,
+      "learning_rate": 5.224199288256228e-05,
+      "loss": 1.4886,
+      "step": 8314
+    },
+    {
+      "epoch": 3.6955555555555555,
+      "grad_norm": 3.5230629444122314,
+      "learning_rate": 5.2224199288256226e-05,
+      "loss": 1.2701,
+      "step": 8315
+    },
+    {
+      "epoch": 3.6959999999999997,
+      "grad_norm": 3.156364917755127,
+      "learning_rate": 5.2206405693950176e-05,
+      "loss": 1.1824,
+      "step": 8316
+    },
+    {
+      "epoch": 3.6964444444444444,
+      "grad_norm": 3.9449214935302734,
+      "learning_rate": 5.218861209964413e-05,
+      "loss": 1.2558,
+      "step": 8317
+    },
+    {
+      "epoch": 3.696888888888889,
+      "grad_norm": 3.572063684463501,
+      "learning_rate": 5.217081850533808e-05,
+      "loss": 1.4913,
+      "step": 8318
+    },
+    {
+      "epoch": 3.6973333333333334,
+      "grad_norm": 3.8648767471313477,
+      "learning_rate": 5.215302491103203e-05,
+      "loss": 1.6746,
+      "step": 8319
+    },
+    {
+      "epoch": 3.6977777777777776,
+      "grad_norm": 3.326974391937256,
+      "learning_rate": 5.213523131672599e-05,
+      "loss": 1.213,
+      "step": 8320
+    },
+    {
+      "epoch": 3.6982222222222223,
+      "grad_norm": 3.392534017562866,
+      "learning_rate": 5.2117437722419925e-05,
+      "loss": 1.272,
+      "step": 8321
+    },
+    {
+      "epoch": 3.6986666666666665,
+      "grad_norm": 3.3309099674224854,
+      "learning_rate": 5.209964412811388e-05,
+      "loss": 1.559,
+      "step": 8322
+    },
+    {
+      "epoch": 3.6991111111111112,
+      "grad_norm": 3.49074387550354,
+      "learning_rate": 5.208185053380783e-05,
+      "loss": 1.2227,
+      "step": 8323
+    },
+    {
+      "epoch": 3.6995555555555555,
+      "grad_norm": 3.63338041305542,
+      "learning_rate": 5.206405693950178e-05,
+      "loss": 1.2891,
+      "step": 8324
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 3.5115768909454346,
+      "learning_rate": 5.204626334519574e-05,
+      "loss": 1.3273,
+      "step": 8325
+    },
+    {
+      "epoch": 3.7004444444444444,
+      "grad_norm": 3.8019955158233643,
+      "learning_rate": 5.202846975088969e-05,
+      "loss": 1.0889,
+      "step": 8326
+    },
+    {
+      "epoch": 3.7008888888888887,
+      "grad_norm": 3.4950296878814697,
+      "learning_rate": 5.2010676156583636e-05,
+      "loss": 1.379,
+      "step": 8327
+    },
+    {
+      "epoch": 3.7013333333333334,
+      "grad_norm": 3.5437874794006348,
+      "learning_rate": 5.199288256227758e-05,
+      "loss": 1.4179,
+      "step": 8328
+    },
+    {
+      "epoch": 3.7017777777777776,
+      "grad_norm": 3.734877109527588,
+      "learning_rate": 5.197508896797153e-05,
+      "loss": 1.0075,
+      "step": 8329
+    },
+    {
+      "epoch": 3.7022222222222223,
+      "grad_norm": 2.877346992492676,
+      "learning_rate": 5.195729537366548e-05,
+      "loss": 1.1345,
+      "step": 8330
+    },
+    {
+      "epoch": 3.7026666666666666,
+      "grad_norm": 3.68837571144104,
+      "learning_rate": 5.1939501779359435e-05,
+      "loss": 1.176,
+      "step": 8331
+    },
+    {
+      "epoch": 3.7031111111111112,
+      "grad_norm": 3.2419826984405518,
+      "learning_rate": 5.1921708185053385e-05,
+      "loss": 0.9544,
+      "step": 8332
+    },
+    {
+      "epoch": 3.7035555555555555,
+      "grad_norm": 3.9483325481414795,
+      "learning_rate": 5.1903914590747335e-05,
+      "loss": 1.0054,
+      "step": 8333
+    },
+    {
+      "epoch": 3.7039999999999997,
+      "grad_norm": 3.8043711185455322,
+      "learning_rate": 5.188612099644128e-05,
+      "loss": 1.1455,
+      "step": 8334
+    },
+    {
+      "epoch": 3.7044444444444444,
+      "grad_norm": 3.818793535232544,
+      "learning_rate": 5.186832740213523e-05,
+      "loss": 1.0459,
+      "step": 8335
+    },
+    {
+      "epoch": 3.704888888888889,
+      "grad_norm": 3.5451431274414062,
+      "learning_rate": 5.1850533807829184e-05,
+      "loss": 1.2172,
+      "step": 8336
+    },
+    {
+      "epoch": 3.7053333333333334,
+      "grad_norm": 3.9264607429504395,
+      "learning_rate": 5.1832740213523134e-05,
+      "loss": 1.5576,
+      "step": 8337
+    },
+    {
+      "epoch": 3.7057777777777776,
+      "grad_norm": 3.3462579250335693,
+      "learning_rate": 5.1814946619217083e-05,
+      "loss": 1.1177,
+      "step": 8338
+    },
+    {
+      "epoch": 3.7062222222222223,
+      "grad_norm": 3.5663697719573975,
+      "learning_rate": 5.179715302491104e-05,
+      "loss": 1.4554,
+      "step": 8339
+    },
+    {
+      "epoch": 3.7066666666666666,
+      "grad_norm": 3.879847288131714,
+      "learning_rate": 5.177935943060499e-05,
+      "loss": 1.3961,
+      "step": 8340
+    },
+    {
+      "epoch": 3.7071111111111112,
+      "grad_norm": 4.285848617553711,
+      "learning_rate": 5.176156583629893e-05,
+      "loss": 1.485,
+      "step": 8341
+    },
+    {
+      "epoch": 3.7075555555555555,
+      "grad_norm": 4.327765464782715,
+      "learning_rate": 5.174377224199288e-05,
+      "loss": 1.439,
+      "step": 8342
+    },
+    {
+      "epoch": 3.708,
+      "grad_norm": 3.3615176677703857,
+      "learning_rate": 5.172597864768683e-05,
+      "loss": 1.0735,
+      "step": 8343
+    },
+    {
+      "epoch": 3.7084444444444444,
+      "grad_norm": 5.214451789855957,
+      "learning_rate": 5.170818505338079e-05,
+      "loss": 1.2868,
+      "step": 8344
+    },
+    {
+      "epoch": 3.7088888888888887,
+      "grad_norm": 3.7910678386688232,
+      "learning_rate": 5.169039145907474e-05,
+      "loss": 1.2784,
+      "step": 8345
+    },
+    {
+      "epoch": 3.7093333333333334,
+      "grad_norm": 3.738447666168213,
+      "learning_rate": 5.167259786476869e-05,
+      "loss": 1.1879,
+      "step": 8346
+    },
+    {
+      "epoch": 3.7097777777777776,
+      "grad_norm": 4.8904643058776855,
+      "learning_rate": 5.165480427046263e-05,
+      "loss": 1.3139,
+      "step": 8347
+    },
+    {
+      "epoch": 3.7102222222222223,
+      "grad_norm": 5.133594036102295,
+      "learning_rate": 5.163701067615658e-05,
+      "loss": 1.5525,
+      "step": 8348
+    },
+    {
+      "epoch": 3.7106666666666666,
+      "grad_norm": 7.878839492797852,
+      "learning_rate": 5.161921708185054e-05,
+      "loss": 1.2252,
+      "step": 8349
+    },
+    {
+      "epoch": 3.7111111111111112,
+      "grad_norm": 3.2023541927337646,
+      "learning_rate": 5.160142348754449e-05,
+      "loss": 1.0264,
+      "step": 8350
+    },
+    {
+      "epoch": 3.7115555555555555,
+      "grad_norm": 2.548800230026245,
+      "learning_rate": 5.1583629893238437e-05,
+      "loss": 1.7163,
+      "step": 8351
+    },
+    {
+      "epoch": 3.7119999999999997,
+      "grad_norm": 2.4168503284454346,
+      "learning_rate": 5.156583629893239e-05,
+      "loss": 1.3757,
+      "step": 8352
+    },
+    {
+      "epoch": 3.7124444444444444,
+      "grad_norm": 2.8642630577087402,
+      "learning_rate": 5.154804270462633e-05,
+      "loss": 1.593,
+      "step": 8353
+    },
+    {
+      "epoch": 3.712888888888889,
+      "grad_norm": 3.071199893951416,
+      "learning_rate": 5.1530249110320286e-05,
+      "loss": 1.929,
+      "step": 8354
+    },
+    {
+      "epoch": 3.7133333333333334,
+      "grad_norm": 2.835141897201538,
+      "learning_rate": 5.1512455516014235e-05,
+      "loss": 1.5476,
+      "step": 8355
+    },
+    {
+      "epoch": 3.7137777777777776,
+      "grad_norm": 3.1856870651245117,
+      "learning_rate": 5.1494661921708185e-05,
+      "loss": 1.4111,
+      "step": 8356
+    },
+    {
+      "epoch": 3.7142222222222223,
+      "grad_norm": 3.279014825820923,
+      "learning_rate": 5.147686832740214e-05,
+      "loss": 1.6864,
+      "step": 8357
+    },
+    {
+      "epoch": 3.7146666666666666,
+      "grad_norm": 2.251955270767212,
+      "learning_rate": 5.145907473309609e-05,
+      "loss": 0.8535,
+      "step": 8358
+    },
+    {
+      "epoch": 3.7151111111111113,
+      "grad_norm": 3.4706077575683594,
+      "learning_rate": 5.144128113879004e-05,
+      "loss": 1.4023,
+      "step": 8359
+    },
+    {
+      "epoch": 3.7155555555555555,
+      "grad_norm": 3.1662933826446533,
+      "learning_rate": 5.1423487544483984e-05,
+      "loss": 1.1449,
+      "step": 8360
+    },
+    {
+      "epoch": 3.716,
+      "grad_norm": 3.3553576469421387,
+      "learning_rate": 5.1405693950177934e-05,
+      "loss": 1.1429,
+      "step": 8361
+    },
+    {
+      "epoch": 3.7164444444444444,
+      "grad_norm": 4.185884952545166,
+      "learning_rate": 5.138790035587189e-05,
+      "loss": 1.6231,
+      "step": 8362
+    },
+    {
+      "epoch": 3.7168888888888887,
+      "grad_norm": 3.4784839153289795,
+      "learning_rate": 5.137010676156584e-05,
+      "loss": 1.7953,
+      "step": 8363
+    },
+    {
+      "epoch": 3.7173333333333334,
+      "grad_norm": 3.0565781593322754,
+      "learning_rate": 5.135231316725979e-05,
+      "loss": 1.5965,
+      "step": 8364
+    },
+    {
+      "epoch": 3.7177777777777776,
+      "grad_norm": 3.630974054336548,
+      "learning_rate": 5.1334519572953746e-05,
+      "loss": 1.0787,
+      "step": 8365
+    },
+    {
+      "epoch": 3.7182222222222223,
+      "grad_norm": 3.7152397632598877,
+      "learning_rate": 5.131672597864768e-05,
+      "loss": 1.783,
+      "step": 8366
+    },
+    {
+      "epoch": 3.7186666666666666,
+      "grad_norm": 2.838580846786499,
+      "learning_rate": 5.129893238434164e-05,
+      "loss": 0.9311,
+      "step": 8367
+    },
+    {
+      "epoch": 3.7191111111111113,
+      "grad_norm": 3.2233829498291016,
+      "learning_rate": 5.128113879003559e-05,
+      "loss": 1.3966,
+      "step": 8368
+    },
+    {
+      "epoch": 3.7195555555555555,
+      "grad_norm": 3.2140471935272217,
+      "learning_rate": 5.126334519572954e-05,
+      "loss": 1.4011,
+      "step": 8369
+    },
+    {
+      "epoch": 3.7199999999999998,
+      "grad_norm": 3.5495688915252686,
+      "learning_rate": 5.1245551601423495e-05,
+      "loss": 1.5965,
+      "step": 8370
+    },
+    {
+      "epoch": 3.7204444444444444,
+      "grad_norm": 4.187902450561523,
+      "learning_rate": 5.1227758007117445e-05,
+      "loss": 1.4036,
+      "step": 8371
+    },
+    {
+      "epoch": 3.720888888888889,
+      "grad_norm": 3.8422319889068604,
+      "learning_rate": 5.1209964412811394e-05,
+      "loss": 1.5535,
+      "step": 8372
+    },
+    {
+      "epoch": 3.7213333333333334,
+      "grad_norm": 3.2021896839141846,
+      "learning_rate": 5.119217081850534e-05,
+      "loss": 1.4836,
+      "step": 8373
+    },
+    {
+      "epoch": 3.7217777777777776,
+      "grad_norm": 3.2228951454162598,
+      "learning_rate": 5.117437722419929e-05,
+      "loss": 1.2058,
+      "step": 8374
+    },
+    {
+      "epoch": 3.7222222222222223,
+      "grad_norm": 3.7099387645721436,
+      "learning_rate": 5.115658362989324e-05,
+      "loss": 1.2875,
+      "step": 8375
+    },
+    {
+      "epoch": 3.7226666666666666,
+      "grad_norm": 2.979400157928467,
+      "learning_rate": 5.113879003558719e-05,
+      "loss": 1.1885,
+      "step": 8376
+    },
+    {
+      "epoch": 3.7231111111111113,
+      "grad_norm": 3.7015509605407715,
+      "learning_rate": 5.112099644128114e-05,
+      "loss": 1.467,
+      "step": 8377
+    },
+    {
+      "epoch": 3.7235555555555555,
+      "grad_norm": 4.023457050323486,
+      "learning_rate": 5.110320284697509e-05,
+      "loss": 1.5333,
+      "step": 8378
+    },
+    {
+      "epoch": 3.724,
+      "grad_norm": 2.9902260303497314,
+      "learning_rate": 5.1085409252669036e-05,
+      "loss": 1.2314,
+      "step": 8379
+    },
+    {
+      "epoch": 3.7244444444444444,
+      "grad_norm": 4.326292991638184,
+      "learning_rate": 5.1067615658362985e-05,
+      "loss": 1.9829,
+      "step": 8380
+    },
+    {
+      "epoch": 3.7248888888888887,
+      "grad_norm": 3.4808270931243896,
+      "learning_rate": 5.104982206405694e-05,
+      "loss": 1.5185,
+      "step": 8381
+    },
+    {
+      "epoch": 3.7253333333333334,
+      "grad_norm": 3.8214287757873535,
+      "learning_rate": 5.103202846975089e-05,
+      "loss": 1.6433,
+      "step": 8382
+    },
+    {
+      "epoch": 3.7257777777777776,
+      "grad_norm": 3.6692192554473877,
+      "learning_rate": 5.101423487544484e-05,
+      "loss": 1.4233,
+      "step": 8383
+    },
+    {
+      "epoch": 3.7262222222222223,
+      "grad_norm": 2.361046552658081,
+      "learning_rate": 5.09964412811388e-05,
+      "loss": 0.7065,
+      "step": 8384
+    },
+    {
+      "epoch": 3.7266666666666666,
+      "grad_norm": 3.3511815071105957,
+      "learning_rate": 5.097864768683275e-05,
+      "loss": 1.0584,
+      "step": 8385
+    },
+    {
+      "epoch": 3.7271111111111113,
+      "grad_norm": 3.851069211959839,
+      "learning_rate": 5.096085409252669e-05,
+      "loss": 1.5869,
+      "step": 8386
+    },
+    {
+      "epoch": 3.7275555555555555,
+      "grad_norm": 3.2693023681640625,
+      "learning_rate": 5.094306049822064e-05,
+      "loss": 1.0885,
+      "step": 8387
+    },
+    {
+      "epoch": 3.7279999999999998,
+      "grad_norm": 3.3839292526245117,
+      "learning_rate": 5.092526690391459e-05,
+      "loss": 1.2459,
+      "step": 8388
+    },
+    {
+      "epoch": 3.7284444444444444,
+      "grad_norm": 4.055683135986328,
+      "learning_rate": 5.0907473309608546e-05,
+      "loss": 1.1619,
+      "step": 8389
+    },
+    {
+      "epoch": 3.728888888888889,
+      "grad_norm": 4.578985691070557,
+      "learning_rate": 5.0889679715302496e-05,
+      "loss": 1.3901,
+      "step": 8390
+    },
+    {
+      "epoch": 3.7293333333333334,
+      "grad_norm": 3.6490328311920166,
+      "learning_rate": 5.0871886120996446e-05,
+      "loss": 1.2829,
+      "step": 8391
+    },
+    {
+      "epoch": 3.7297777777777776,
+      "grad_norm": 4.452314853668213,
+      "learning_rate": 5.085409252669039e-05,
+      "loss": 1.9056,
+      "step": 8392
+    },
+    {
+      "epoch": 3.7302222222222223,
+      "grad_norm": 2.537201166152954,
+      "learning_rate": 5.083629893238434e-05,
+      "loss": 0.9342,
+      "step": 8393
+    },
+    {
+      "epoch": 3.7306666666666666,
+      "grad_norm": 3.178445339202881,
+      "learning_rate": 5.0818505338078295e-05,
+      "loss": 1.0333,
+      "step": 8394
+    },
+    {
+      "epoch": 3.7311111111111113,
+      "grad_norm": 3.890375852584839,
+      "learning_rate": 5.0800711743772245e-05,
+      "loss": 1.2319,
+      "step": 8395
+    },
+    {
+      "epoch": 3.7315555555555555,
+      "grad_norm": 3.9328713417053223,
+      "learning_rate": 5.0782918149466194e-05,
+      "loss": 0.9856,
+      "step": 8396
+    },
+    {
+      "epoch": 3.732,
+      "grad_norm": 4.24088191986084,
+      "learning_rate": 5.076512455516015e-05,
+      "loss": 1.2177,
+      "step": 8397
+    },
+    {
+      "epoch": 3.7324444444444445,
+      "grad_norm": 3.290562629699707,
+      "learning_rate": 5.07473309608541e-05,
+      "loss": 0.7268,
+      "step": 8398
+    },
+    {
+      "epoch": 3.7328888888888887,
+      "grad_norm": 3.8136353492736816,
+      "learning_rate": 5.0729537366548043e-05,
+      "loss": 0.3112,
+      "step": 8399
+    },
+    {
+      "epoch": 3.7333333333333334,
+      "grad_norm": 4.7547454833984375,
+      "learning_rate": 5.071174377224199e-05,
+      "loss": 0.8184,
+      "step": 8400
+    },
+    {
+      "epoch": 3.7337777777777776,
+      "grad_norm": 2.6292288303375244,
+      "learning_rate": 5.069395017793594e-05,
+      "loss": 1.9555,
+      "step": 8401
+    },
+    {
+      "epoch": 3.7342222222222223,
+      "grad_norm": 1.9523777961730957,
+      "learning_rate": 5.06761565836299e-05,
+      "loss": 0.8654,
+      "step": 8402
+    },
+    {
+      "epoch": 3.7346666666666666,
+      "grad_norm": 2.6126413345336914,
+      "learning_rate": 5.065836298932385e-05,
+      "loss": 1.7074,
+      "step": 8403
+    },
+    {
+      "epoch": 3.7351111111111113,
+      "grad_norm": 2.737042188644409,
+      "learning_rate": 5.06405693950178e-05,
+      "loss": 1.7787,
+      "step": 8404
+    },
+    {
+      "epoch": 3.7355555555555555,
+      "grad_norm": 3.067366600036621,
+      "learning_rate": 5.062277580071174e-05,
+      "loss": 1.4087,
+      "step": 8405
+    },
+    {
+      "epoch": 3.7359999999999998,
+      "grad_norm": 3.5647099018096924,
+      "learning_rate": 5.060498220640569e-05,
+      "loss": 1.8064,
+      "step": 8406
+    },
+    {
+      "epoch": 3.7364444444444445,
+      "grad_norm": 3.5738532543182373,
+      "learning_rate": 5.058718861209965e-05,
+      "loss": 2.0468,
+      "step": 8407
+    },
+    {
+      "epoch": 3.736888888888889,
+      "grad_norm": 3.4049813747406006,
+      "learning_rate": 5.05693950177936e-05,
+      "loss": 1.33,
+      "step": 8408
+    },
+    {
+      "epoch": 3.7373333333333334,
+      "grad_norm": 3.1586861610412598,
+      "learning_rate": 5.055160142348755e-05,
+      "loss": 1.7597,
+      "step": 8409
+    },
+    {
+      "epoch": 3.7377777777777776,
+      "grad_norm": 2.9543251991271973,
+      "learning_rate": 5.0533807829181504e-05,
+      "loss": 1.3858,
+      "step": 8410
+    },
+    {
+      "epoch": 3.7382222222222223,
+      "grad_norm": 3.1417078971862793,
+      "learning_rate": 5.051601423487544e-05,
+      "loss": 1.8369,
+      "step": 8411
+    },
+    {
+      "epoch": 3.7386666666666666,
+      "grad_norm": 3.0961074829101562,
+      "learning_rate": 5.04982206405694e-05,
+      "loss": 1.4527,
+      "step": 8412
+    },
+    {
+      "epoch": 3.7391111111111113,
+      "grad_norm": 3.145967483520508,
+      "learning_rate": 5.0480427046263346e-05,
+      "loss": 1.2984,
+      "step": 8413
+    },
+    {
+      "epoch": 3.7395555555555555,
+      "grad_norm": 2.759549856185913,
+      "learning_rate": 5.0462633451957296e-05,
+      "loss": 0.936,
+      "step": 8414
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 2.0562026500701904,
+      "learning_rate": 5.044483985765125e-05,
+      "loss": 0.7822,
+      "step": 8415
+    },
+    {
+      "epoch": 3.7404444444444445,
+      "grad_norm": 3.5271592140197754,
+      "learning_rate": 5.04270462633452e-05,
+      "loss": 1.9599,
+      "step": 8416
+    },
+    {
+      "epoch": 3.7408888888888887,
+      "grad_norm": 4.152881145477295,
+      "learning_rate": 5.040925266903915e-05,
+      "loss": 1.5676,
+      "step": 8417
+    },
+    {
+      "epoch": 3.7413333333333334,
+      "grad_norm": 3.7143032550811768,
+      "learning_rate": 5.0391459074733095e-05,
+      "loss": 1.4384,
+      "step": 8418
+    },
+    {
+      "epoch": 3.7417777777777776,
+      "grad_norm": 3.6692559719085693,
+      "learning_rate": 5.0373665480427045e-05,
+      "loss": 1.0879,
+      "step": 8419
+    },
+    {
+      "epoch": 3.7422222222222223,
+      "grad_norm": 2.8461532592773438,
+      "learning_rate": 5.0355871886120994e-05,
+      "loss": 1.2018,
+      "step": 8420
+    },
+    {
+      "epoch": 3.7426666666666666,
+      "grad_norm": 3.5240817070007324,
+      "learning_rate": 5.033807829181495e-05,
+      "loss": 1.7161,
+      "step": 8421
+    },
+    {
+      "epoch": 3.7431111111111113,
+      "grad_norm": 3.98225998878479,
+      "learning_rate": 5.03202846975089e-05,
+      "loss": 1.5325,
+      "step": 8422
+    },
+    {
+      "epoch": 3.7435555555555555,
+      "grad_norm": 3.3189501762390137,
+      "learning_rate": 5.030249110320285e-05,
+      "loss": 1.6055,
+      "step": 8423
+    },
+    {
+      "epoch": 3.7439999999999998,
+      "grad_norm": 3.6065752506256104,
+      "learning_rate": 5.028469750889679e-05,
+      "loss": 1.6128,
+      "step": 8424
+    },
+    {
+      "epoch": 3.7444444444444445,
+      "grad_norm": 3.7471060752868652,
+      "learning_rate": 5.026690391459074e-05,
+      "loss": 1.1217,
+      "step": 8425
+    },
+    {
+      "epoch": 3.744888888888889,
+      "grad_norm": 3.301255702972412,
+      "learning_rate": 5.02491103202847e-05,
+      "loss": 1.1631,
+      "step": 8426
+    },
+    {
+      "epoch": 3.7453333333333334,
+      "grad_norm": 3.3452343940734863,
+      "learning_rate": 5.023131672597865e-05,
+      "loss": 1.2224,
+      "step": 8427
+    },
+    {
+      "epoch": 3.7457777777777777,
+      "grad_norm": 3.3727502822875977,
+      "learning_rate": 5.02135231316726e-05,
+      "loss": 1.1533,
+      "step": 8428
+    },
+    {
+      "epoch": 3.7462222222222223,
+      "grad_norm": 3.6482954025268555,
+      "learning_rate": 5.0195729537366555e-05,
+      "loss": 1.4724,
+      "step": 8429
+    },
+    {
+      "epoch": 3.7466666666666666,
+      "grad_norm": 3.6111557483673096,
+      "learning_rate": 5.0177935943060505e-05,
+      "loss": 1.2438,
+      "step": 8430
+    },
+    {
+      "epoch": 3.747111111111111,
+      "grad_norm": 0.2538396120071411,
+      "learning_rate": 5.016014234875445e-05,
+      "loss": 0.0322,
+      "step": 8431
+    },
+    {
+      "epoch": 3.7475555555555555,
+      "grad_norm": 3.558535575866699,
+      "learning_rate": 5.01423487544484e-05,
+      "loss": 1.6036,
+      "step": 8432
+    },
+    {
+      "epoch": 3.748,
+      "grad_norm": 3.5707340240478516,
+      "learning_rate": 5.012455516014235e-05,
+      "loss": 1.0665,
+      "step": 8433
+    },
+    {
+      "epoch": 3.7484444444444445,
+      "grad_norm": 3.5678842067718506,
+      "learning_rate": 5.0106761565836304e-05,
+      "loss": 0.9684,
+      "step": 8434
+    },
+    {
+      "epoch": 3.7488888888888887,
+      "grad_norm": 3.770473003387451,
+      "learning_rate": 5.0088967971530254e-05,
+      "loss": 1.1987,
+      "step": 8435
+    },
+    {
+      "epoch": 3.7493333333333334,
+      "grad_norm": 3.830598831176758,
+      "learning_rate": 5.0071174377224204e-05,
+      "loss": 1.2741,
+      "step": 8436
+    },
+    {
+      "epoch": 3.7497777777777777,
+      "grad_norm": 3.5863218307495117,
+      "learning_rate": 5.0053380782918146e-05,
+      "loss": 1.7064,
+      "step": 8437
+    },
+    {
+      "epoch": 3.7502222222222223,
+      "grad_norm": 4.064740180969238,
+      "learning_rate": 5.0035587188612096e-05,
+      "loss": 1.5107,
+      "step": 8438
+    },
+    {
+      "epoch": 3.7506666666666666,
+      "grad_norm": 3.272308588027954,
+      "learning_rate": 5.001779359430605e-05,
+      "loss": 1.0924,
+      "step": 8439
+    },
+    {
+      "epoch": 3.7511111111111113,
+      "grad_norm": 3.53114914894104,
+      "learning_rate": 5e-05,
+      "loss": 1.1262,
+      "step": 8440
+    },
+    {
+      "epoch": 3.7515555555555555,
+      "grad_norm": 3.6443865299224854,
+      "learning_rate": 4.998220640569395e-05,
+      "loss": 1.2448,
+      "step": 8441
+    },
+    {
+      "epoch": 3.752,
+      "grad_norm": 4.559260845184326,
+      "learning_rate": 4.99644128113879e-05,
+      "loss": 1.0831,
+      "step": 8442
+    },
+    {
+      "epoch": 3.7524444444444445,
+      "grad_norm": 4.263270378112793,
+      "learning_rate": 4.994661921708185e-05,
+      "loss": 1.2406,
+      "step": 8443
+    },
+    {
+      "epoch": 3.752888888888889,
+      "grad_norm": 4.247504234313965,
+      "learning_rate": 4.99288256227758e-05,
+      "loss": 1.2629,
+      "step": 8444
+    },
+    {
+      "epoch": 3.7533333333333334,
+      "grad_norm": 3.752406358718872,
+      "learning_rate": 4.991103202846975e-05,
+      "loss": 1.1891,
+      "step": 8445
+    },
+    {
+      "epoch": 3.7537777777777777,
+      "grad_norm": 3.750194787979126,
+      "learning_rate": 4.98932384341637e-05,
+      "loss": 1.3531,
+      "step": 8446
+    },
+    {
+      "epoch": 3.7542222222222223,
+      "grad_norm": 5.040890216827393,
+      "learning_rate": 4.987544483985766e-05,
+      "loss": 1.2088,
+      "step": 8447
+    },
+    {
+      "epoch": 3.7546666666666666,
+      "grad_norm": 4.1202006340026855,
+      "learning_rate": 4.985765124555161e-05,
+      "loss": 1.2779,
+      "step": 8448
+    },
+    {
+      "epoch": 3.755111111111111,
+      "grad_norm": 4.093197345733643,
+      "learning_rate": 4.983985765124555e-05,
+      "loss": 0.832,
+      "step": 8449
+    },
+    {
+      "epoch": 3.7555555555555555,
+      "grad_norm": 4.263852119445801,
+      "learning_rate": 4.9822064056939506e-05,
+      "loss": 0.4735,
+      "step": 8450
+    },
+    {
+      "epoch": 3.7560000000000002,
+      "grad_norm": 1.3852638006210327,
+      "learning_rate": 4.9804270462633456e-05,
+      "loss": 0.6947,
+      "step": 8451
+    },
+    {
+      "epoch": 3.7564444444444445,
+      "grad_norm": 2.5553395748138428,
+      "learning_rate": 4.9786476868327406e-05,
+      "loss": 1.6021,
+      "step": 8452
+    },
+    {
+      "epoch": 3.7568888888888887,
+      "grad_norm": 2.2025933265686035,
+      "learning_rate": 4.9768683274021356e-05,
+      "loss": 1.1253,
+      "step": 8453
+    },
+    {
+      "epoch": 3.7573333333333334,
+      "grad_norm": 2.739720582962036,
+      "learning_rate": 4.9750889679715305e-05,
+      "loss": 1.8841,
+      "step": 8454
+    },
+    {
+      "epoch": 3.7577777777777777,
+      "grad_norm": 3.2897424697875977,
+      "learning_rate": 4.9733096085409255e-05,
+      "loss": 1.8541,
+      "step": 8455
+    },
+    {
+      "epoch": 3.7582222222222224,
+      "grad_norm": 3.067814350128174,
+      "learning_rate": 4.9715302491103205e-05,
+      "loss": 1.4044,
+      "step": 8456
+    },
+    {
+      "epoch": 3.7586666666666666,
+      "grad_norm": 2.3474533557891846,
+      "learning_rate": 4.9697508896797154e-05,
+      "loss": 1.1692,
+      "step": 8457
+    },
+    {
+      "epoch": 3.7591111111111113,
+      "grad_norm": 3.4697251319885254,
+      "learning_rate": 4.9679715302491104e-05,
+      "loss": 1.6912,
+      "step": 8458
+    },
+    {
+      "epoch": 3.7595555555555555,
+      "grad_norm": 3.053978443145752,
+      "learning_rate": 4.9661921708185054e-05,
+      "loss": 1.4072,
+      "step": 8459
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 3.327169179916382,
+      "learning_rate": 4.964412811387901e-05,
+      "loss": 1.6578,
+      "step": 8460
+    },
+    {
+      "epoch": 3.7604444444444445,
+      "grad_norm": 3.0925164222717285,
+      "learning_rate": 4.962633451957295e-05,
+      "loss": 1.6706,
+      "step": 8461
+    },
+    {
+      "epoch": 3.7608888888888887,
+      "grad_norm": 2.676941394805908,
+      "learning_rate": 4.96085409252669e-05,
+      "loss": 1.1866,
+      "step": 8462
+    },
+    {
+      "epoch": 3.7613333333333334,
+      "grad_norm": 3.976247787475586,
+      "learning_rate": 4.959074733096086e-05,
+      "loss": 1.4686,
+      "step": 8463
+    },
+    {
+      "epoch": 3.7617777777777777,
+      "grad_norm": 3.166390895843506,
+      "learning_rate": 4.957295373665481e-05,
+      "loss": 1.1084,
+      "step": 8464
+    },
+    {
+      "epoch": 3.7622222222222224,
+      "grad_norm": 3.5380375385284424,
+      "learning_rate": 4.955516014234875e-05,
+      "loss": 1.5746,
+      "step": 8465
+    },
+    {
+      "epoch": 3.7626666666666666,
+      "grad_norm": 2.127175807952881,
+      "learning_rate": 4.953736654804271e-05,
+      "loss": 0.4779,
+      "step": 8466
+    },
+    {
+      "epoch": 3.763111111111111,
+      "grad_norm": 3.756263494491577,
+      "learning_rate": 4.951957295373666e-05,
+      "loss": 1.2304,
+      "step": 8467
+    },
+    {
+      "epoch": 3.7635555555555555,
+      "grad_norm": 3.4877758026123047,
+      "learning_rate": 4.950177935943061e-05,
+      "loss": 1.7547,
+      "step": 8468
+    },
+    {
+      "epoch": 3.7640000000000002,
+      "grad_norm": 3.6990151405334473,
+      "learning_rate": 4.948398576512456e-05,
+      "loss": 1.4276,
+      "step": 8469
+    },
+    {
+      "epoch": 3.7644444444444445,
+      "grad_norm": 3.083824634552002,
+      "learning_rate": 4.946619217081851e-05,
+      "loss": 1.3949,
+      "step": 8470
+    },
+    {
+      "epoch": 3.7648888888888887,
+      "grad_norm": 3.808980703353882,
+      "learning_rate": 4.944839857651246e-05,
+      "loss": 1.247,
+      "step": 8471
+    },
+    {
+      "epoch": 3.7653333333333334,
+      "grad_norm": 3.5250051021575928,
+      "learning_rate": 4.943060498220641e-05,
+      "loss": 1.6016,
+      "step": 8472
+    },
+    {
+      "epoch": 3.7657777777777777,
+      "grad_norm": 3.2846179008483887,
+      "learning_rate": 4.941281138790036e-05,
+      "loss": 1.1513,
+      "step": 8473
+    },
+    {
+      "epoch": 3.7662222222222224,
+      "grad_norm": 3.1756033897399902,
+      "learning_rate": 4.9395017793594306e-05,
+      "loss": 1.2079,
+      "step": 8474
+    },
+    {
+      "epoch": 3.7666666666666666,
+      "grad_norm": 3.1101205348968506,
+      "learning_rate": 4.9377224199288256e-05,
+      "loss": 1.1411,
+      "step": 8475
+    },
+    {
+      "epoch": 3.7671111111111113,
+      "grad_norm": 3.436415433883667,
+      "learning_rate": 4.935943060498221e-05,
+      "loss": 1.2356,
+      "step": 8476
+    },
+    {
+      "epoch": 3.7675555555555555,
+      "grad_norm": 3.413377285003662,
+      "learning_rate": 4.934163701067616e-05,
+      "loss": 1.6142,
+      "step": 8477
+    },
+    {
+      "epoch": 3.768,
+      "grad_norm": 3.4654288291931152,
+      "learning_rate": 4.9323843416370105e-05,
+      "loss": 1.4022,
+      "step": 8478
+    },
+    {
+      "epoch": 3.7684444444444445,
+      "grad_norm": 1.9338456392288208,
+      "learning_rate": 4.930604982206406e-05,
+      "loss": 0.5082,
+      "step": 8479
+    },
+    {
+      "epoch": 3.7688888888888887,
+      "grad_norm": 3.8473315238952637,
+      "learning_rate": 4.928825622775801e-05,
+      "loss": 1.4336,
+      "step": 8480
+    },
+    {
+      "epoch": 3.7693333333333334,
+      "grad_norm": 2.3992502689361572,
+      "learning_rate": 4.927046263345196e-05,
+      "loss": 0.7091,
+      "step": 8481
+    },
+    {
+      "epoch": 3.7697777777777777,
+      "grad_norm": 3.7076449394226074,
+      "learning_rate": 4.925266903914591e-05,
+      "loss": 1.5299,
+      "step": 8482
+    },
+    {
+      "epoch": 3.7702222222222224,
+      "grad_norm": 3.1709039211273193,
+      "learning_rate": 4.923487544483986e-05,
+      "loss": 1.0703,
+      "step": 8483
+    },
+    {
+      "epoch": 3.7706666666666666,
+      "grad_norm": 3.58585262298584,
+      "learning_rate": 4.921708185053381e-05,
+      "loss": 1.3857,
+      "step": 8484
+    },
+    {
+      "epoch": 3.771111111111111,
+      "grad_norm": 3.486786365509033,
+      "learning_rate": 4.919928825622776e-05,
+      "loss": 1.5706,
+      "step": 8485
+    },
+    {
+      "epoch": 3.7715555555555556,
+      "grad_norm": 3.2322800159454346,
+      "learning_rate": 4.918149466192171e-05,
+      "loss": 0.9788,
+      "step": 8486
+    },
+    {
+      "epoch": 3.7720000000000002,
+      "grad_norm": 3.4824700355529785,
+      "learning_rate": 4.916370106761566e-05,
+      "loss": 1.3818,
+      "step": 8487
+    },
+    {
+      "epoch": 3.7724444444444445,
+      "grad_norm": 3.4638094902038574,
+      "learning_rate": 4.914590747330961e-05,
+      "loss": 1.3499,
+      "step": 8488
+    },
+    {
+      "epoch": 3.7728888888888887,
+      "grad_norm": 3.505972146987915,
+      "learning_rate": 4.912811387900356e-05,
+      "loss": 1.0351,
+      "step": 8489
+    },
+    {
+      "epoch": 3.7733333333333334,
+      "grad_norm": 3.177283525466919,
+      "learning_rate": 4.911032028469751e-05,
+      "loss": 1.4239,
+      "step": 8490
+    },
+    {
+      "epoch": 3.7737777777777777,
+      "grad_norm": 3.8296144008636475,
+      "learning_rate": 4.909252669039146e-05,
+      "loss": 1.449,
+      "step": 8491
+    },
+    {
+      "epoch": 3.7742222222222224,
+      "grad_norm": 3.507258415222168,
+      "learning_rate": 4.9074733096085415e-05,
+      "loss": 1.1,
+      "step": 8492
+    },
+    {
+      "epoch": 3.7746666666666666,
+      "grad_norm": 3.143571615219116,
+      "learning_rate": 4.9056939501779365e-05,
+      "loss": 0.668,
+      "step": 8493
+    },
+    {
+      "epoch": 3.7751111111111113,
+      "grad_norm": 4.362837314605713,
+      "learning_rate": 4.903914590747331e-05,
+      "loss": 1.2331,
+      "step": 8494
+    },
+    {
+      "epoch": 3.7755555555555556,
+      "grad_norm": 4.254435062408447,
+      "learning_rate": 4.9021352313167264e-05,
+      "loss": 1.0221,
+      "step": 8495
+    },
+    {
+      "epoch": 3.776,
+      "grad_norm": 3.589878797531128,
+      "learning_rate": 4.9003558718861214e-05,
+      "loss": 0.9752,
+      "step": 8496
+    },
+    {
+      "epoch": 3.7764444444444445,
+      "grad_norm": 4.932192325592041,
+      "learning_rate": 4.8985765124555164e-05,
+      "loss": 1.1548,
+      "step": 8497
+    },
+    {
+      "epoch": 3.7768888888888887,
+      "grad_norm": 4.631229400634766,
+      "learning_rate": 4.896797153024911e-05,
+      "loss": 1.2203,
+      "step": 8498
+    },
+    {
+      "epoch": 3.7773333333333334,
+      "grad_norm": 5.158580780029297,
+      "learning_rate": 4.895017793594306e-05,
+      "loss": 1.5583,
+      "step": 8499
+    },
+    {
+      "epoch": 3.7777777777777777,
+      "grad_norm": 4.34719181060791,
+      "learning_rate": 4.893238434163701e-05,
+      "loss": 0.8711,
+      "step": 8500
+    },
+    {
+      "epoch": 3.7782222222222224,
+      "grad_norm": 1.7901962995529175,
+      "learning_rate": 4.891459074733096e-05,
+      "loss": 0.969,
+      "step": 8501
+    },
+    {
+      "epoch": 3.7786666666666666,
+      "grad_norm": 2.875005006790161,
+      "learning_rate": 4.889679715302491e-05,
+      "loss": 1.8006,
+      "step": 8502
+    },
+    {
+      "epoch": 3.779111111111111,
+      "grad_norm": 0.1947576105594635,
+      "learning_rate": 4.887900355871886e-05,
+      "loss": 0.0163,
+      "step": 8503
+    },
+    {
+      "epoch": 3.7795555555555556,
+      "grad_norm": 2.6576716899871826,
+      "learning_rate": 4.886120996441281e-05,
+      "loss": 1.585,
+      "step": 8504
+    },
+    {
+      "epoch": 3.7800000000000002,
+      "grad_norm": 2.818967342376709,
+      "learning_rate": 4.884341637010677e-05,
+      "loss": 1.658,
+      "step": 8505
+    },
+    {
+      "epoch": 3.7804444444444445,
+      "grad_norm": 3.126274585723877,
+      "learning_rate": 4.882562277580072e-05,
+      "loss": 1.2724,
+      "step": 8506
+    },
+    {
+      "epoch": 3.7808888888888887,
+      "grad_norm": 3.2297914028167725,
+      "learning_rate": 4.880782918149466e-05,
+      "loss": 1.7334,
+      "step": 8507
+    },
+    {
+      "epoch": 3.7813333333333334,
+      "grad_norm": 3.0050487518310547,
+      "learning_rate": 4.879003558718862e-05,
+      "loss": 1.734,
+      "step": 8508
+    },
+    {
+      "epoch": 3.7817777777777777,
+      "grad_norm": 3.3488566875457764,
+      "learning_rate": 4.877224199288257e-05,
+      "loss": 1.7469,
+      "step": 8509
+    },
+    {
+      "epoch": 3.7822222222222224,
+      "grad_norm": 2.3252031803131104,
+      "learning_rate": 4.875444839857651e-05,
+      "loss": 0.6167,
+      "step": 8510
+    },
+    {
+      "epoch": 3.7826666666666666,
+      "grad_norm": 3.154351234436035,
+      "learning_rate": 4.8736654804270466e-05,
+      "loss": 1.5694,
+      "step": 8511
+    },
+    {
+      "epoch": 3.7831111111111113,
+      "grad_norm": 3.574852705001831,
+      "learning_rate": 4.8718861209964416e-05,
+      "loss": 1.4769,
+      "step": 8512
+    },
+    {
+      "epoch": 3.7835555555555556,
+      "grad_norm": 2.932189464569092,
+      "learning_rate": 4.8701067615658366e-05,
+      "loss": 1.2641,
+      "step": 8513
+    },
+    {
+      "epoch": 3.784,
+      "grad_norm": 3.2192270755767822,
+      "learning_rate": 4.8683274021352316e-05,
+      "loss": 1.5916,
+      "step": 8514
+    },
+    {
+      "epoch": 3.7844444444444445,
+      "grad_norm": 3.2948851585388184,
+      "learning_rate": 4.8665480427046265e-05,
+      "loss": 1.4189,
+      "step": 8515
+    },
+    {
+      "epoch": 3.7848888888888887,
+      "grad_norm": 3.3815078735351562,
+      "learning_rate": 4.8647686832740215e-05,
+      "loss": 1.5318,
+      "step": 8516
+    },
+    {
+      "epoch": 3.7853333333333334,
+      "grad_norm": 3.34417724609375,
+      "learning_rate": 4.8629893238434165e-05,
+      "loss": 1.5307,
+      "step": 8517
+    },
+    {
+      "epoch": 3.7857777777777777,
+      "grad_norm": 3.700117826461792,
+      "learning_rate": 4.8612099644128115e-05,
+      "loss": 0.8523,
+      "step": 8518
+    },
+    {
+      "epoch": 3.7862222222222224,
+      "grad_norm": 3.057548999786377,
+      "learning_rate": 4.8594306049822064e-05,
+      "loss": 1.1474,
+      "step": 8519
+    },
+    {
+      "epoch": 3.7866666666666666,
+      "grad_norm": 3.9406673908233643,
+      "learning_rate": 4.8576512455516014e-05,
+      "loss": 1.5789,
+      "step": 8520
+    },
+    {
+      "epoch": 3.787111111111111,
+      "grad_norm": 2.2049951553344727,
+      "learning_rate": 4.855871886120997e-05,
+      "loss": 0.9952,
+      "step": 8521
+    },
+    {
+      "epoch": 3.7875555555555556,
+      "grad_norm": 3.1826930046081543,
+      "learning_rate": 4.854092526690392e-05,
+      "loss": 1.1047,
+      "step": 8522
+    },
+    {
+      "epoch": 3.7880000000000003,
+      "grad_norm": 5.1929192543029785,
+      "learning_rate": 4.852313167259786e-05,
+      "loss": 1.3763,
+      "step": 8523
+    },
+    {
+      "epoch": 3.7884444444444445,
+      "grad_norm": 3.3125691413879395,
+      "learning_rate": 4.850533807829182e-05,
+      "loss": 1.5083,
+      "step": 8524
+    },
+    {
+      "epoch": 3.7888888888888888,
+      "grad_norm": 2.958933115005493,
+      "learning_rate": 4.848754448398577e-05,
+      "loss": 0.9042,
+      "step": 8525
+    },
+    {
+      "epoch": 3.7893333333333334,
+      "grad_norm": 3.5505728721618652,
+      "learning_rate": 4.846975088967972e-05,
+      "loss": 1.4218,
+      "step": 8526
+    },
+    {
+      "epoch": 3.7897777777777777,
+      "grad_norm": 3.6742985248565674,
+      "learning_rate": 4.845195729537367e-05,
+      "loss": 1.1994,
+      "step": 8527
+    },
+    {
+      "epoch": 3.7902222222222224,
+      "grad_norm": 3.1948156356811523,
+      "learning_rate": 4.843416370106762e-05,
+      "loss": 1.439,
+      "step": 8528
+    },
+    {
+      "epoch": 3.7906666666666666,
+      "grad_norm": 3.452651262283325,
+      "learning_rate": 4.841637010676157e-05,
+      "loss": 0.9603,
+      "step": 8529
+    },
+    {
+      "epoch": 3.7911111111111113,
+      "grad_norm": 3.6245038509368896,
+      "learning_rate": 4.839857651245552e-05,
+      "loss": 1.3753,
+      "step": 8530
+    },
+    {
+      "epoch": 3.7915555555555556,
+      "grad_norm": 3.7631113529205322,
+      "learning_rate": 4.838078291814947e-05,
+      "loss": 1.3403,
+      "step": 8531
+    },
+    {
+      "epoch": 3.792,
+      "grad_norm": 3.130178213119507,
+      "learning_rate": 4.836298932384342e-05,
+      "loss": 1.4048,
+      "step": 8532
+    },
+    {
+      "epoch": 3.7924444444444445,
+      "grad_norm": 4.133440017700195,
+      "learning_rate": 4.834519572953737e-05,
+      "loss": 1.1999,
+      "step": 8533
+    },
+    {
+      "epoch": 3.7928888888888888,
+      "grad_norm": 3.3631365299224854,
+      "learning_rate": 4.832740213523132e-05,
+      "loss": 1.1746,
+      "step": 8534
+    },
+    {
+      "epoch": 3.7933333333333334,
+      "grad_norm": 1.9450854063034058,
+      "learning_rate": 4.830960854092527e-05,
+      "loss": 0.3845,
+      "step": 8535
+    },
+    {
+      "epoch": 3.7937777777777777,
+      "grad_norm": 3.2177698612213135,
+      "learning_rate": 4.8291814946619216e-05,
+      "loss": 0.9918,
+      "step": 8536
+    },
+    {
+      "epoch": 3.7942222222222224,
+      "grad_norm": 3.5220441818237305,
+      "learning_rate": 4.827402135231317e-05,
+      "loss": 1.4969,
+      "step": 8537
+    },
+    {
+      "epoch": 3.7946666666666666,
+      "grad_norm": 4.0864996910095215,
+      "learning_rate": 4.825622775800712e-05,
+      "loss": 1.7111,
+      "step": 8538
+    },
+    {
+      "epoch": 3.795111111111111,
+      "grad_norm": 3.846219062805176,
+      "learning_rate": 4.8238434163701065e-05,
+      "loss": 1.1624,
+      "step": 8539
+    },
+    {
+      "epoch": 3.7955555555555556,
+      "grad_norm": 3.6705775260925293,
+      "learning_rate": 4.822064056939502e-05,
+      "loss": 1.132,
+      "step": 8540
+    },
+    {
+      "epoch": 3.7960000000000003,
+      "grad_norm": 3.94295334815979,
+      "learning_rate": 4.820284697508897e-05,
+      "loss": 1.6029,
+      "step": 8541
+    },
+    {
+      "epoch": 3.7964444444444445,
+      "grad_norm": 4.677298069000244,
+      "learning_rate": 4.818505338078292e-05,
+      "loss": 1.2734,
+      "step": 8542
+    },
+    {
+      "epoch": 3.7968888888888888,
+      "grad_norm": 3.895962715148926,
+      "learning_rate": 4.816725978647687e-05,
+      "loss": 0.8572,
+      "step": 8543
+    },
+    {
+      "epoch": 3.7973333333333334,
+      "grad_norm": 3.5614423751831055,
+      "learning_rate": 4.814946619217082e-05,
+      "loss": 1.3358,
+      "step": 8544
+    },
+    {
+      "epoch": 3.7977777777777777,
+      "grad_norm": 4.679959774017334,
+      "learning_rate": 4.813167259786477e-05,
+      "loss": 1.301,
+      "step": 8545
+    },
+    {
+      "epoch": 3.7982222222222224,
+      "grad_norm": 3.1577718257904053,
+      "learning_rate": 4.811387900355872e-05,
+      "loss": 0.7731,
+      "step": 8546
+    },
+    {
+      "epoch": 3.7986666666666666,
+      "grad_norm": 3.898137331008911,
+      "learning_rate": 4.809608540925267e-05,
+      "loss": 1.1303,
+      "step": 8547
+    },
+    {
+      "epoch": 3.7991111111111113,
+      "grad_norm": 4.067081928253174,
+      "learning_rate": 4.807829181494662e-05,
+      "loss": 1.231,
+      "step": 8548
+    },
+    {
+      "epoch": 3.7995555555555556,
+      "grad_norm": 3.7059855461120605,
+      "learning_rate": 4.806049822064057e-05,
+      "loss": 1.2194,
+      "step": 8549
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 4.398168563842773,
+      "learning_rate": 4.8042704626334526e-05,
+      "loss": 1.011,
+      "step": 8550
+    },
+    {
+      "epoch": 3.8004444444444445,
+      "grad_norm": 2.875924587249756,
+      "learning_rate": 4.8024911032028476e-05,
+      "loss": 1.9846,
+      "step": 8551
+    },
+    {
+      "epoch": 3.8008888888888888,
+      "grad_norm": 2.9650471210479736,
+      "learning_rate": 4.800711743772242e-05,
+      "loss": 2.1224,
+      "step": 8552
+    },
+    {
+      "epoch": 3.8013333333333335,
+      "grad_norm": 2.5322835445404053,
+      "learning_rate": 4.7989323843416375e-05,
+      "loss": 1.3393,
+      "step": 8553
+    },
+    {
+      "epoch": 3.8017777777777777,
+      "grad_norm": 3.003859281539917,
+      "learning_rate": 4.7971530249110325e-05,
+      "loss": 1.2544,
+      "step": 8554
+    },
+    {
+      "epoch": 3.8022222222222224,
+      "grad_norm": 3.0419626235961914,
+      "learning_rate": 4.795373665480427e-05,
+      "loss": 1.7736,
+      "step": 8555
+    },
+    {
+      "epoch": 3.8026666666666666,
+      "grad_norm": 3.1113429069519043,
+      "learning_rate": 4.7935943060498224e-05,
+      "loss": 1.4109,
+      "step": 8556
+    },
+    {
+      "epoch": 3.803111111111111,
+      "grad_norm": 3.09995698928833,
+      "learning_rate": 4.7918149466192174e-05,
+      "loss": 1.862,
+      "step": 8557
+    },
+    {
+      "epoch": 3.8035555555555556,
+      "grad_norm": 2.92777156829834,
+      "learning_rate": 4.7900355871886124e-05,
+      "loss": 1.3818,
+      "step": 8558
+    },
+    {
+      "epoch": 3.8040000000000003,
+      "grad_norm": 3.054277181625366,
+      "learning_rate": 4.7882562277580073e-05,
+      "loss": 1.6397,
+      "step": 8559
+    },
+    {
+      "epoch": 3.8044444444444445,
+      "grad_norm": 3.6034891605377197,
+      "learning_rate": 4.786476868327402e-05,
+      "loss": 1.8047,
+      "step": 8560
+    },
+    {
+      "epoch": 3.8048888888888888,
+      "grad_norm": 3.8269503116607666,
+      "learning_rate": 4.784697508896797e-05,
+      "loss": 1.6447,
+      "step": 8561
+    },
+    {
+      "epoch": 3.8053333333333335,
+      "grad_norm": 3.4392542839050293,
+      "learning_rate": 4.782918149466192e-05,
+      "loss": 1.5413,
+      "step": 8562
+    },
+    {
+      "epoch": 3.8057777777777777,
+      "grad_norm": 3.8955986499786377,
+      "learning_rate": 4.781138790035587e-05,
+      "loss": 1.7783,
+      "step": 8563
+    },
+    {
+      "epoch": 3.806222222222222,
+      "grad_norm": 3.504478693008423,
+      "learning_rate": 4.779359430604983e-05,
+      "loss": 1.0836,
+      "step": 8564
+    },
+    {
+      "epoch": 3.8066666666666666,
+      "grad_norm": 2.9569408893585205,
+      "learning_rate": 4.777580071174377e-05,
+      "loss": 1.1422,
+      "step": 8565
+    },
+    {
+      "epoch": 3.8071111111111113,
+      "grad_norm": 3.3336431980133057,
+      "learning_rate": 4.775800711743773e-05,
+      "loss": 1.6741,
+      "step": 8566
+    },
+    {
+      "epoch": 3.8075555555555556,
+      "grad_norm": 3.0466670989990234,
+      "learning_rate": 4.774021352313168e-05,
+      "loss": 1.207,
+      "step": 8567
+    },
+    {
+      "epoch": 3.808,
+      "grad_norm": 3.735114574432373,
+      "learning_rate": 4.772241992882562e-05,
+      "loss": 1.0772,
+      "step": 8568
+    },
+    {
+      "epoch": 3.8084444444444445,
+      "grad_norm": 3.699998617172241,
+      "learning_rate": 4.770462633451958e-05,
+      "loss": 1.3742,
+      "step": 8569
+    },
+    {
+      "epoch": 3.8088888888888888,
+      "grad_norm": 3.7992329597473145,
+      "learning_rate": 4.768683274021353e-05,
+      "loss": 1.3175,
+      "step": 8570
+    },
+    {
+      "epoch": 3.8093333333333335,
+      "grad_norm": 2.9106605052948,
+      "learning_rate": 4.766903914590748e-05,
+      "loss": 1.1899,
+      "step": 8571
+    },
+    {
+      "epoch": 3.8097777777777777,
+      "grad_norm": 3.2745654582977295,
+      "learning_rate": 4.7651245551601427e-05,
+      "loss": 1.0309,
+      "step": 8572
+    },
+    {
+      "epoch": 3.8102222222222224,
+      "grad_norm": 4.083169937133789,
+      "learning_rate": 4.7633451957295376e-05,
+      "loss": 1.5096,
+      "step": 8573
+    },
+    {
+      "epoch": 3.8106666666666666,
+      "grad_norm": 3.3155415058135986,
+      "learning_rate": 4.7615658362989326e-05,
+      "loss": 1.3212,
+      "step": 8574
+    },
+    {
+      "epoch": 3.811111111111111,
+      "grad_norm": 3.864699363708496,
+      "learning_rate": 4.7597864768683276e-05,
+      "loss": 1.3274,
+      "step": 8575
+    },
+    {
+      "epoch": 3.8115555555555556,
+      "grad_norm": 3.174386739730835,
+      "learning_rate": 4.7580071174377225e-05,
+      "loss": 1.1286,
+      "step": 8576
+    },
+    {
+      "epoch": 3.8120000000000003,
+      "grad_norm": 2.8707237243652344,
+      "learning_rate": 4.7562277580071175e-05,
+      "loss": 0.837,
+      "step": 8577
+    },
+    {
+      "epoch": 3.8124444444444445,
+      "grad_norm": 3.41292405128479,
+      "learning_rate": 4.7544483985765125e-05,
+      "loss": 1.3842,
+      "step": 8578
+    },
+    {
+      "epoch": 3.8128888888888888,
+      "grad_norm": 3.6212575435638428,
+      "learning_rate": 4.7526690391459075e-05,
+      "loss": 1.4985,
+      "step": 8579
+    },
+    {
+      "epoch": 3.8133333333333335,
+      "grad_norm": 3.4787771701812744,
+      "learning_rate": 4.750889679715303e-05,
+      "loss": 1.2398,
+      "step": 8580
+    },
+    {
+      "epoch": 3.8137777777777777,
+      "grad_norm": 3.5631048679351807,
+      "learning_rate": 4.7491103202846974e-05,
+      "loss": 1.2581,
+      "step": 8581
+    },
+    {
+      "epoch": 3.814222222222222,
+      "grad_norm": 3.253061056137085,
+      "learning_rate": 4.747330960854093e-05,
+      "loss": 0.8728,
+      "step": 8582
+    },
+    {
+      "epoch": 3.8146666666666667,
+      "grad_norm": 3.8624134063720703,
+      "learning_rate": 4.745551601423488e-05,
+      "loss": 1.3563,
+      "step": 8583
+    },
+    {
+      "epoch": 3.8151111111111113,
+      "grad_norm": 3.724658966064453,
+      "learning_rate": 4.743772241992882e-05,
+      "loss": 1.2391,
+      "step": 8584
+    },
+    {
+      "epoch": 3.8155555555555556,
+      "grad_norm": 4.286252975463867,
+      "learning_rate": 4.741992882562278e-05,
+      "loss": 1.6731,
+      "step": 8585
+    },
+    {
+      "epoch": 3.816,
+      "grad_norm": 4.29559326171875,
+      "learning_rate": 4.740213523131673e-05,
+      "loss": 1.3075,
+      "step": 8586
+    },
+    {
+      "epoch": 3.8164444444444445,
+      "grad_norm": 4.753545761108398,
+      "learning_rate": 4.738434163701068e-05,
+      "loss": 1.5687,
+      "step": 8587
+    },
+    {
+      "epoch": 3.8168888888888888,
+      "grad_norm": 2.7094476222991943,
+      "learning_rate": 4.736654804270463e-05,
+      "loss": 0.731,
+      "step": 8588
+    },
+    {
+      "epoch": 3.8173333333333335,
+      "grad_norm": 3.8449037075042725,
+      "learning_rate": 4.734875444839858e-05,
+      "loss": 1.4633,
+      "step": 8589
+    },
+    {
+      "epoch": 3.8177777777777777,
+      "grad_norm": 3.7161953449249268,
+      "learning_rate": 4.733096085409253e-05,
+      "loss": 0.9841,
+      "step": 8590
+    },
+    {
+      "epoch": 3.8182222222222224,
+      "grad_norm": 4.3916449546813965,
+      "learning_rate": 4.731316725978648e-05,
+      "loss": 1.2136,
+      "step": 8591
+    },
+    {
+      "epoch": 3.8186666666666667,
+      "grad_norm": 2.985351800918579,
+      "learning_rate": 4.729537366548043e-05,
+      "loss": 0.7911,
+      "step": 8592
+    },
+    {
+      "epoch": 3.819111111111111,
+      "grad_norm": 3.5371320247650146,
+      "learning_rate": 4.7277580071174384e-05,
+      "loss": 1.0655,
+      "step": 8593
+    },
+    {
+      "epoch": 3.8195555555555556,
+      "grad_norm": 9.645051956176758,
+      "learning_rate": 4.725978647686833e-05,
+      "loss": 1.1596,
+      "step": 8594
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 4.049137115478516,
+      "learning_rate": 4.7241992882562284e-05,
+      "loss": 1.3832,
+      "step": 8595
+    },
+    {
+      "epoch": 3.8204444444444445,
+      "grad_norm": 3.475752592086792,
+      "learning_rate": 4.7224199288256233e-05,
+      "loss": 1.171,
+      "step": 8596
+    },
+    {
+      "epoch": 3.820888888888889,
+      "grad_norm": 4.638784408569336,
+      "learning_rate": 4.7206405693950176e-05,
+      "loss": 1.3144,
+      "step": 8597
+    },
+    {
+      "epoch": 3.8213333333333335,
+      "grad_norm": 4.170307636260986,
+      "learning_rate": 4.718861209964413e-05,
+      "loss": 1.1174,
+      "step": 8598
+    },
+    {
+      "epoch": 3.8217777777777777,
+      "grad_norm": 4.556606292724609,
+      "learning_rate": 4.717081850533808e-05,
+      "loss": 1.1181,
+      "step": 8599
+    },
+    {
+      "epoch": 3.822222222222222,
+      "grad_norm": 4.029767990112305,
+      "learning_rate": 4.7153024911032026e-05,
+      "loss": 0.8181,
+      "step": 8600
+    },
+    {
+      "epoch": 3.8226666666666667,
+      "grad_norm": 2.57649827003479,
+      "learning_rate": 4.713523131672598e-05,
+      "loss": 2.2343,
+      "step": 8601
+    },
+    {
+      "epoch": 3.8231111111111113,
+      "grad_norm": 3.349898338317871,
+      "learning_rate": 4.711743772241993e-05,
+      "loss": 1.74,
+      "step": 8602
+    },
+    {
+      "epoch": 3.8235555555555556,
+      "grad_norm": 3.0327224731445312,
+      "learning_rate": 4.709964412811388e-05,
+      "loss": 1.4284,
+      "step": 8603
+    },
+    {
+      "epoch": 3.824,
+      "grad_norm": 3.228236436843872,
+      "learning_rate": 4.708185053380783e-05,
+      "loss": 1.6167,
+      "step": 8604
+    },
+    {
+      "epoch": 3.8244444444444445,
+      "grad_norm": 3.0474486351013184,
+      "learning_rate": 4.706405693950178e-05,
+      "loss": 1.2462,
+      "step": 8605
+    },
+    {
+      "epoch": 3.824888888888889,
+      "grad_norm": 3.509099006652832,
+      "learning_rate": 4.704626334519573e-05,
+      "loss": 1.7278,
+      "step": 8606
+    },
+    {
+      "epoch": 3.8253333333333335,
+      "grad_norm": 2.920429229736328,
+      "learning_rate": 4.702846975088968e-05,
+      "loss": 1.5507,
+      "step": 8607
+    },
+    {
+      "epoch": 3.8257777777777777,
+      "grad_norm": 1.919208288192749,
+      "learning_rate": 4.701067615658363e-05,
+      "loss": 0.6798,
+      "step": 8608
+    },
+    {
+      "epoch": 3.8262222222222224,
+      "grad_norm": 3.108910322189331,
+      "learning_rate": 4.699288256227759e-05,
+      "loss": 0.9496,
+      "step": 8609
+    },
+    {
+      "epoch": 3.8266666666666667,
+      "grad_norm": 3.4903645515441895,
+      "learning_rate": 4.697508896797153e-05,
+      "loss": 1.2665,
+      "step": 8610
+    },
+    {
+      "epoch": 3.827111111111111,
+      "grad_norm": 3.292814254760742,
+      "learning_rate": 4.6957295373665486e-05,
+      "loss": 1.7923,
+      "step": 8611
+    },
+    {
+      "epoch": 3.8275555555555556,
+      "grad_norm": 2.9696640968322754,
+      "learning_rate": 4.6939501779359436e-05,
+      "loss": 1.1515,
+      "step": 8612
+    },
+    {
+      "epoch": 3.828,
+      "grad_norm": 2.911243438720703,
+      "learning_rate": 4.692170818505338e-05,
+      "loss": 1.2357,
+      "step": 8613
+    },
+    {
+      "epoch": 3.8284444444444445,
+      "grad_norm": 3.220693588256836,
+      "learning_rate": 4.6903914590747335e-05,
+      "loss": 1.6873,
+      "step": 8614
+    },
+    {
+      "epoch": 3.828888888888889,
+      "grad_norm": 3.2728185653686523,
+      "learning_rate": 4.6886120996441285e-05,
+      "loss": 0.9055,
+      "step": 8615
+    },
+    {
+      "epoch": 3.8293333333333335,
+      "grad_norm": 3.27384090423584,
+      "learning_rate": 4.6868327402135235e-05,
+      "loss": 1.1947,
+      "step": 8616
+    },
+    {
+      "epoch": 3.8297777777777777,
+      "grad_norm": 3.1152114868164062,
+      "learning_rate": 4.6850533807829184e-05,
+      "loss": 1.2603,
+      "step": 8617
+    },
+    {
+      "epoch": 3.830222222222222,
+      "grad_norm": 2.450045108795166,
+      "learning_rate": 4.6832740213523134e-05,
+      "loss": 0.7963,
+      "step": 8618
+    },
+    {
+      "epoch": 3.8306666666666667,
+      "grad_norm": 3.4387762546539307,
+      "learning_rate": 4.6814946619217084e-05,
+      "loss": 1.3703,
+      "step": 8619
+    },
+    {
+      "epoch": 3.8311111111111114,
+      "grad_norm": 2.993537187576294,
+      "learning_rate": 4.6797153024911034e-05,
+      "loss": 1.0733,
+      "step": 8620
+    },
+    {
+      "epoch": 3.8315555555555556,
+      "grad_norm": 3.041719436645508,
+      "learning_rate": 4.677935943060498e-05,
+      "loss": 1.1477,
+      "step": 8621
+    },
+    {
+      "epoch": 3.832,
+      "grad_norm": 3.9076027870178223,
+      "learning_rate": 4.676156583629894e-05,
+      "loss": 1.3451,
+      "step": 8622
+    },
+    {
+      "epoch": 3.8324444444444445,
+      "grad_norm": 3.3174734115600586,
+      "learning_rate": 4.674377224199288e-05,
+      "loss": 0.8637,
+      "step": 8623
+    },
+    {
+      "epoch": 3.832888888888889,
+      "grad_norm": 3.7948451042175293,
+      "learning_rate": 4.672597864768683e-05,
+      "loss": 1.4609,
+      "step": 8624
+    },
+    {
+      "epoch": 3.8333333333333335,
+      "grad_norm": 3.765836477279663,
+      "learning_rate": 4.670818505338079e-05,
+      "loss": 1.1835,
+      "step": 8625
+    },
+    {
+      "epoch": 3.8337777777777777,
+      "grad_norm": 3.9058310985565186,
+      "learning_rate": 4.669039145907473e-05,
+      "loss": 1.5075,
+      "step": 8626
+    },
+    {
+      "epoch": 3.8342222222222224,
+      "grad_norm": 3.738953113555908,
+      "learning_rate": 4.667259786476869e-05,
+      "loss": 1.2718,
+      "step": 8627
+    },
+    {
+      "epoch": 3.8346666666666667,
+      "grad_norm": 2.5701253414154053,
+      "learning_rate": 4.665480427046264e-05,
+      "loss": 0.6246,
+      "step": 8628
+    },
+    {
+      "epoch": 3.835111111111111,
+      "grad_norm": 4.04362678527832,
+      "learning_rate": 4.663701067615658e-05,
+      "loss": 1.4515,
+      "step": 8629
+    },
+    {
+      "epoch": 3.8355555555555556,
+      "grad_norm": 1.7729761600494385,
+      "learning_rate": 4.661921708185054e-05,
+      "loss": 0.4256,
+      "step": 8630
+    },
+    {
+      "epoch": 3.836,
+      "grad_norm": 3.309948444366455,
+      "learning_rate": 4.660142348754449e-05,
+      "loss": 0.6741,
+      "step": 8631
+    },
+    {
+      "epoch": 3.8364444444444445,
+      "grad_norm": 3.3855819702148438,
+      "learning_rate": 4.658362989323844e-05,
+      "loss": 1.4831,
+      "step": 8632
+    },
+    {
+      "epoch": 3.836888888888889,
+      "grad_norm": 3.421257734298706,
+      "learning_rate": 4.656583629893239e-05,
+      "loss": 1.3999,
+      "step": 8633
+    },
+    {
+      "epoch": 3.8373333333333335,
+      "grad_norm": 3.1086058616638184,
+      "learning_rate": 4.6548042704626336e-05,
+      "loss": 0.8159,
+      "step": 8634
+    },
+    {
+      "epoch": 3.8377777777777777,
+      "grad_norm": 3.4166300296783447,
+      "learning_rate": 4.6530249110320286e-05,
+      "loss": 1.3853,
+      "step": 8635
+    },
+    {
+      "epoch": 3.838222222222222,
+      "grad_norm": 4.6446051597595215,
+      "learning_rate": 4.6512455516014236e-05,
+      "loss": 1.4196,
+      "step": 8636
+    },
+    {
+      "epoch": 3.8386666666666667,
+      "grad_norm": 3.4511923789978027,
+      "learning_rate": 4.6494661921708186e-05,
+      "loss": 1.4235,
+      "step": 8637
+    },
+    {
+      "epoch": 3.8391111111111114,
+      "grad_norm": 4.0160980224609375,
+      "learning_rate": 4.647686832740214e-05,
+      "loss": 1.1172,
+      "step": 8638
+    },
+    {
+      "epoch": 3.8395555555555556,
+      "grad_norm": 3.6638247966766357,
+      "learning_rate": 4.6459074733096085e-05,
+      "loss": 0.8709,
+      "step": 8639
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 3.9915387630462646,
+      "learning_rate": 4.644128113879004e-05,
+      "loss": 1.3121,
+      "step": 8640
+    },
+    {
+      "epoch": 3.8404444444444445,
+      "grad_norm": 3.331205368041992,
+      "learning_rate": 4.642348754448399e-05,
+      "loss": 0.9015,
+      "step": 8641
+    },
+    {
+      "epoch": 3.840888888888889,
+      "grad_norm": 3.3656539916992188,
+      "learning_rate": 4.6405693950177934e-05,
+      "loss": 1.0394,
+      "step": 8642
+    },
+    {
+      "epoch": 3.8413333333333335,
+      "grad_norm": 3.8786230087280273,
+      "learning_rate": 4.638790035587189e-05,
+      "loss": 1.1873,
+      "step": 8643
+    },
+    {
+      "epoch": 3.8417777777777777,
+      "grad_norm": 4.622972011566162,
+      "learning_rate": 4.637010676156584e-05,
+      "loss": 1.1365,
+      "step": 8644
+    },
+    {
+      "epoch": 3.8422222222222224,
+      "grad_norm": 3.3173351287841797,
+      "learning_rate": 4.635231316725978e-05,
+      "loss": 1.0542,
+      "step": 8645
+    },
+    {
+      "epoch": 3.8426666666666667,
+      "grad_norm": 4.794419765472412,
+      "learning_rate": 4.633451957295374e-05,
+      "loss": 1.4618,
+      "step": 8646
+    },
+    {
+      "epoch": 3.843111111111111,
+      "grad_norm": 4.264388084411621,
+      "learning_rate": 4.631672597864769e-05,
+      "loss": 0.9792,
+      "step": 8647
+    },
+    {
+      "epoch": 3.8435555555555556,
+      "grad_norm": 5.3533501625061035,
+      "learning_rate": 4.629893238434164e-05,
+      "loss": 1.031,
+      "step": 8648
+    },
+    {
+      "epoch": 3.844,
+      "grad_norm": 3.8206562995910645,
+      "learning_rate": 4.628113879003559e-05,
+      "loss": 0.9972,
+      "step": 8649
+    },
+    {
+      "epoch": 3.8444444444444446,
+      "grad_norm": 3.9211204051971436,
+      "learning_rate": 4.626334519572954e-05,
+      "loss": 0.4228,
+      "step": 8650
+    },
+    {
+      "epoch": 3.844888888888889,
+      "grad_norm": 2.347119092941284,
+      "learning_rate": 4.6245551601423495e-05,
+      "loss": 1.5638,
+      "step": 8651
+    },
+    {
+      "epoch": 3.8453333333333335,
+      "grad_norm": 2.3549845218658447,
+      "learning_rate": 4.622775800711744e-05,
+      "loss": 1.7397,
+      "step": 8652
+    },
+    {
+      "epoch": 3.8457777777777777,
+      "grad_norm": 2.7358062267303467,
+      "learning_rate": 4.620996441281139e-05,
+      "loss": 1.5028,
+      "step": 8653
+    },
+    {
+      "epoch": 3.846222222222222,
+      "grad_norm": 3.0936498641967773,
+      "learning_rate": 4.6192170818505344e-05,
+      "loss": 1.7485,
+      "step": 8654
+    },
+    {
+      "epoch": 3.8466666666666667,
+      "grad_norm": 1.712854266166687,
+      "learning_rate": 4.617437722419929e-05,
+      "loss": 0.8592,
+      "step": 8655
+    },
+    {
+      "epoch": 3.8471111111111114,
+      "grad_norm": 3.0746419429779053,
+      "learning_rate": 4.6156583629893244e-05,
+      "loss": 1.7988,
+      "step": 8656
+    },
+    {
+      "epoch": 3.8475555555555556,
+      "grad_norm": 2.9364185333251953,
+      "learning_rate": 4.6138790035587194e-05,
+      "loss": 1.5508,
+      "step": 8657
+    },
+    {
+      "epoch": 3.848,
+      "grad_norm": 3.4229776859283447,
+      "learning_rate": 4.6120996441281137e-05,
+      "loss": 1.9598,
+      "step": 8658
+    },
+    {
+      "epoch": 3.8484444444444446,
+      "grad_norm": 3.2707581520080566,
+      "learning_rate": 4.610320284697509e-05,
+      "loss": 2.055,
+      "step": 8659
+    },
+    {
+      "epoch": 3.848888888888889,
+      "grad_norm": 3.055335521697998,
+      "learning_rate": 4.608540925266904e-05,
+      "loss": 1.3324,
+      "step": 8660
+    },
+    {
+      "epoch": 3.8493333333333335,
+      "grad_norm": 3.291973114013672,
+      "learning_rate": 4.606761565836299e-05,
+      "loss": 2.014,
+      "step": 8661
+    },
+    {
+      "epoch": 3.8497777777777777,
+      "grad_norm": 2.739182233810425,
+      "learning_rate": 4.604982206405694e-05,
+      "loss": 0.6788,
+      "step": 8662
+    },
+    {
+      "epoch": 3.8502222222222224,
+      "grad_norm": 2.2576968669891357,
+      "learning_rate": 4.603202846975089e-05,
+      "loss": 0.7666,
+      "step": 8663
+    },
+    {
+      "epoch": 3.8506666666666667,
+      "grad_norm": 3.3199477195739746,
+      "learning_rate": 4.601423487544484e-05,
+      "loss": 1.4767,
+      "step": 8664
+    },
+    {
+      "epoch": 3.851111111111111,
+      "grad_norm": 3.386324405670166,
+      "learning_rate": 4.599644128113879e-05,
+      "loss": 1.273,
+      "step": 8665
+    },
+    {
+      "epoch": 3.8515555555555556,
+      "grad_norm": 3.1465635299682617,
+      "learning_rate": 4.597864768683274e-05,
+      "loss": 1.5099,
+      "step": 8666
+    },
+    {
+      "epoch": 3.852,
+      "grad_norm": 3.1457812786102295,
+      "learning_rate": 4.59608540925267e-05,
+      "loss": 1.2209,
+      "step": 8667
+    },
+    {
+      "epoch": 3.8524444444444446,
+      "grad_norm": 3.257748603820801,
+      "learning_rate": 4.594306049822064e-05,
+      "loss": 1.5008,
+      "step": 8668
+    },
+    {
+      "epoch": 3.852888888888889,
+      "grad_norm": 3.149172306060791,
+      "learning_rate": 4.592526690391459e-05,
+      "loss": 1.2685,
+      "step": 8669
+    },
+    {
+      "epoch": 3.8533333333333335,
+      "grad_norm": 2.9013049602508545,
+      "learning_rate": 4.590747330960855e-05,
+      "loss": 1.0847,
+      "step": 8670
+    },
+    {
+      "epoch": 3.8537777777777777,
+      "grad_norm": 2.962352752685547,
+      "learning_rate": 4.588967971530249e-05,
+      "loss": 0.8992,
+      "step": 8671
+    },
+    {
+      "epoch": 3.854222222222222,
+      "grad_norm": 3.3367416858673096,
+      "learning_rate": 4.5871886120996446e-05,
+      "loss": 1.6202,
+      "step": 8672
+    },
+    {
+      "epoch": 3.8546666666666667,
+      "grad_norm": 3.3775758743286133,
+      "learning_rate": 4.5854092526690396e-05,
+      "loss": 1.727,
+      "step": 8673
+    },
+    {
+      "epoch": 3.8551111111111114,
+      "grad_norm": 1.9687055349349976,
+      "learning_rate": 4.583629893238434e-05,
+      "loss": 0.7593,
+      "step": 8674
+    },
+    {
+      "epoch": 3.8555555555555556,
+      "grad_norm": 3.5084187984466553,
+      "learning_rate": 4.5818505338078295e-05,
+      "loss": 1.3926,
+      "step": 8675
+    },
+    {
+      "epoch": 3.856,
+      "grad_norm": 3.405186176300049,
+      "learning_rate": 4.5800711743772245e-05,
+      "loss": 1.5413,
+      "step": 8676
+    },
+    {
+      "epoch": 3.8564444444444446,
+      "grad_norm": 3.67044734954834,
+      "learning_rate": 4.5782918149466195e-05,
+      "loss": 1.5604,
+      "step": 8677
+    },
+    {
+      "epoch": 3.856888888888889,
+      "grad_norm": 3.343731641769409,
+      "learning_rate": 4.5765124555160144e-05,
+      "loss": 1.4601,
+      "step": 8678
+    },
+    {
+      "epoch": 3.857333333333333,
+      "grad_norm": 3.770385265350342,
+      "learning_rate": 4.5747330960854094e-05,
+      "loss": 1.3078,
+      "step": 8679
+    },
+    {
+      "epoch": 3.8577777777777778,
+      "grad_norm": 3.8169827461242676,
+      "learning_rate": 4.572953736654805e-05,
+      "loss": 1.362,
+      "step": 8680
+    },
+    {
+      "epoch": 3.8582222222222224,
+      "grad_norm": 3.3232243061065674,
+      "learning_rate": 4.5711743772241994e-05,
+      "loss": 1.4338,
+      "step": 8681
+    },
+    {
+      "epoch": 3.8586666666666667,
+      "grad_norm": 4.413866996765137,
+      "learning_rate": 4.569395017793594e-05,
+      "loss": 1.5802,
+      "step": 8682
+    },
+    {
+      "epoch": 3.859111111111111,
+      "grad_norm": 3.514523983001709,
+      "learning_rate": 4.56761565836299e-05,
+      "loss": 1.5358,
+      "step": 8683
+    },
+    {
+      "epoch": 3.8595555555555556,
+      "grad_norm": 3.5461461544036865,
+      "learning_rate": 4.565836298932384e-05,
+      "loss": 1.3908,
+      "step": 8684
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 3.1903622150421143,
+      "learning_rate": 4.56405693950178e-05,
+      "loss": 1.0759,
+      "step": 8685
+    },
+    {
+      "epoch": 3.8604444444444446,
+      "grad_norm": 3.541867971420288,
+      "learning_rate": 4.562277580071175e-05,
+      "loss": 1.4128,
+      "step": 8686
+    },
+    {
+      "epoch": 3.860888888888889,
+      "grad_norm": 3.281956434249878,
+      "learning_rate": 4.560498220640569e-05,
+      "loss": 1.2347,
+      "step": 8687
+    },
+    {
+      "epoch": 3.8613333333333335,
+      "grad_norm": 3.273998498916626,
+      "learning_rate": 4.558718861209965e-05,
+      "loss": 0.9553,
+      "step": 8688
+    },
+    {
+      "epoch": 3.8617777777777778,
+      "grad_norm": 3.5180327892303467,
+      "learning_rate": 4.55693950177936e-05,
+      "loss": 1.2825,
+      "step": 8689
+    },
+    {
+      "epoch": 3.862222222222222,
+      "grad_norm": 3.3110413551330566,
+      "learning_rate": 4.555160142348754e-05,
+      "loss": 1.1399,
+      "step": 8690
+    },
+    {
+      "epoch": 3.8626666666666667,
+      "grad_norm": 3.2750625610351562,
+      "learning_rate": 4.55338078291815e-05,
+      "loss": 1.181,
+      "step": 8691
+    },
+    {
+      "epoch": 3.8631111111111114,
+      "grad_norm": 3.3777711391448975,
+      "learning_rate": 4.551601423487545e-05,
+      "loss": 0.9839,
+      "step": 8692
+    },
+    {
+      "epoch": 3.8635555555555556,
+      "grad_norm": 4.099118232727051,
+      "learning_rate": 4.54982206405694e-05,
+      "loss": 1.184,
+      "step": 8693
+    },
+    {
+      "epoch": 3.864,
+      "grad_norm": 3.8149666786193848,
+      "learning_rate": 4.548042704626335e-05,
+      "loss": 1.387,
+      "step": 8694
+    },
+    {
+      "epoch": 3.8644444444444446,
+      "grad_norm": 3.8572885990142822,
+      "learning_rate": 4.5462633451957297e-05,
+      "loss": 1.1251,
+      "step": 8695
+    },
+    {
+      "epoch": 3.864888888888889,
+      "grad_norm": 3.872878313064575,
+      "learning_rate": 4.544483985765125e-05,
+      "loss": 1.2452,
+      "step": 8696
+    },
+    {
+      "epoch": 3.865333333333333,
+      "grad_norm": 4.542336940765381,
+      "learning_rate": 4.5427046263345196e-05,
+      "loss": 1.1995,
+      "step": 8697
+    },
+    {
+      "epoch": 3.8657777777777778,
+      "grad_norm": 4.309566974639893,
+      "learning_rate": 4.5409252669039146e-05,
+      "loss": 1.4818,
+      "step": 8698
+    },
+    {
+      "epoch": 3.8662222222222224,
+      "grad_norm": 4.477492332458496,
+      "learning_rate": 4.53914590747331e-05,
+      "loss": 1.4093,
+      "step": 8699
+    },
+    {
+      "epoch": 3.8666666666666667,
+      "grad_norm": 3.625142812728882,
+      "learning_rate": 4.5373665480427045e-05,
+      "loss": 0.8632,
+      "step": 8700
+    },
+    {
+      "epoch": 3.867111111111111,
+      "grad_norm": 1.706745982170105,
+      "learning_rate": 4.5355871886121e-05,
+      "loss": 1.0516,
+      "step": 8701
+    },
+    {
+      "epoch": 3.8675555555555556,
+      "grad_norm": 2.5270049571990967,
+      "learning_rate": 4.533807829181495e-05,
+      "loss": 1.6525,
+      "step": 8702
+    },
+    {
+      "epoch": 3.868,
+      "grad_norm": 1.587844729423523,
+      "learning_rate": 4.5320284697508894e-05,
+      "loss": 0.7329,
+      "step": 8703
+    },
+    {
+      "epoch": 3.8684444444444446,
+      "grad_norm": 3.1046266555786133,
+      "learning_rate": 4.530249110320285e-05,
+      "loss": 1.6955,
+      "step": 8704
+    },
+    {
+      "epoch": 3.868888888888889,
+      "grad_norm": 3.3074681758880615,
+      "learning_rate": 4.52846975088968e-05,
+      "loss": 1.6316,
+      "step": 8705
+    },
+    {
+      "epoch": 3.8693333333333335,
+      "grad_norm": 3.674323320388794,
+      "learning_rate": 4.526690391459075e-05,
+      "loss": 2.3296,
+      "step": 8706
+    },
+    {
+      "epoch": 3.8697777777777778,
+      "grad_norm": 3.3690025806427,
+      "learning_rate": 4.52491103202847e-05,
+      "loss": 1.4658,
+      "step": 8707
+    },
+    {
+      "epoch": 3.870222222222222,
+      "grad_norm": 3.0607759952545166,
+      "learning_rate": 4.523131672597865e-05,
+      "loss": 1.8122,
+      "step": 8708
+    },
+    {
+      "epoch": 3.8706666666666667,
+      "grad_norm": 3.5003387928009033,
+      "learning_rate": 4.52135231316726e-05,
+      "loss": 1.5672,
+      "step": 8709
+    },
+    {
+      "epoch": 3.871111111111111,
+      "grad_norm": 3.6459617614746094,
+      "learning_rate": 4.519572953736655e-05,
+      "loss": 1.6405,
+      "step": 8710
+    },
+    {
+      "epoch": 3.8715555555555556,
+      "grad_norm": 3.704782247543335,
+      "learning_rate": 4.51779359430605e-05,
+      "loss": 1.7152,
+      "step": 8711
+    },
+    {
+      "epoch": 3.872,
+      "grad_norm": 3.561216115951538,
+      "learning_rate": 4.5160142348754455e-05,
+      "loss": 1.2059,
+      "step": 8712
+    },
+    {
+      "epoch": 3.8724444444444446,
+      "grad_norm": 3.5210413932800293,
+      "learning_rate": 4.51423487544484e-05,
+      "loss": 1.4903,
+      "step": 8713
+    },
+    {
+      "epoch": 3.872888888888889,
+      "grad_norm": 3.023265838623047,
+      "learning_rate": 4.512455516014235e-05,
+      "loss": 1.3212,
+      "step": 8714
+    },
+    {
+      "epoch": 3.873333333333333,
+      "grad_norm": 3.663731098175049,
+      "learning_rate": 4.5106761565836305e-05,
+      "loss": 1.3391,
+      "step": 8715
+    },
+    {
+      "epoch": 3.8737777777777778,
+      "grad_norm": 3.64522385597229,
+      "learning_rate": 4.508896797153025e-05,
+      "loss": 0.937,
+      "step": 8716
+    },
+    {
+      "epoch": 3.8742222222222225,
+      "grad_norm": 3.8014652729034424,
+      "learning_rate": 4.5071174377224204e-05,
+      "loss": 1.5124,
+      "step": 8717
+    },
+    {
+      "epoch": 3.8746666666666667,
+      "grad_norm": 2.626232624053955,
+      "learning_rate": 4.5053380782918154e-05,
+      "loss": 0.9118,
+      "step": 8718
+    },
+    {
+      "epoch": 3.875111111111111,
+      "grad_norm": 3.8472952842712402,
+      "learning_rate": 4.50355871886121e-05,
+      "loss": 1.2868,
+      "step": 8719
+    },
+    {
+      "epoch": 3.8755555555555556,
+      "grad_norm": 3.5172226428985596,
+      "learning_rate": 4.501779359430605e-05,
+      "loss": 1.4457,
+      "step": 8720
+    },
+    {
+      "epoch": 3.876,
+      "grad_norm": 2.7889018058776855,
+      "learning_rate": 4.5e-05,
+      "loss": 1.2991,
+      "step": 8721
+    },
+    {
+      "epoch": 3.8764444444444446,
+      "grad_norm": 3.4194562435150146,
+      "learning_rate": 4.498220640569395e-05,
+      "loss": 1.5408,
+      "step": 8722
+    },
+    {
+      "epoch": 3.876888888888889,
+      "grad_norm": 3.4279887676239014,
+      "learning_rate": 4.49644128113879e-05,
+      "loss": 1.0715,
+      "step": 8723
+    },
+    {
+      "epoch": 3.8773333333333335,
+      "grad_norm": 3.7102408409118652,
+      "learning_rate": 4.494661921708185e-05,
+      "loss": 1.19,
+      "step": 8724
+    },
+    {
+      "epoch": 3.8777777777777778,
+      "grad_norm": 4.325709342956543,
+      "learning_rate": 4.492882562277581e-05,
+      "loss": 0.7462,
+      "step": 8725
+    },
+    {
+      "epoch": 3.878222222222222,
+      "grad_norm": 3.521773099899292,
+      "learning_rate": 4.491103202846975e-05,
+      "loss": 0.9747,
+      "step": 8726
+    },
+    {
+      "epoch": 3.8786666666666667,
+      "grad_norm": 3.7296512126922607,
+      "learning_rate": 4.48932384341637e-05,
+      "loss": 1.3466,
+      "step": 8727
+    },
+    {
+      "epoch": 3.879111111111111,
+      "grad_norm": 2.833895444869995,
+      "learning_rate": 4.487544483985766e-05,
+      "loss": 1.0873,
+      "step": 8728
+    },
+    {
+      "epoch": 3.8795555555555556,
+      "grad_norm": 3.184312105178833,
+      "learning_rate": 4.48576512455516e-05,
+      "loss": 1.184,
+      "step": 8729
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 2.265151023864746,
+      "learning_rate": 4.483985765124555e-05,
+      "loss": 0.5063,
+      "step": 8730
+    },
+    {
+      "epoch": 3.8804444444444446,
+      "grad_norm": 3.3480918407440186,
+      "learning_rate": 4.482206405693951e-05,
+      "loss": 1.0596,
+      "step": 8731
+    },
+    {
+      "epoch": 3.880888888888889,
+      "grad_norm": 4.041034698486328,
+      "learning_rate": 4.480427046263345e-05,
+      "loss": 1.4421,
+      "step": 8732
+    },
+    {
+      "epoch": 3.881333333333333,
+      "grad_norm": 2.776762008666992,
+      "learning_rate": 4.4786476868327406e-05,
+      "loss": 0.9752,
+      "step": 8733
+    },
+    {
+      "epoch": 3.8817777777777778,
+      "grad_norm": 3.5926032066345215,
+      "learning_rate": 4.4768683274021356e-05,
+      "loss": 1.1643,
+      "step": 8734
+    },
+    {
+      "epoch": 3.8822222222222225,
+      "grad_norm": 3.4924111366271973,
+      "learning_rate": 4.47508896797153e-05,
+      "loss": 1.1958,
+      "step": 8735
+    },
+    {
+      "epoch": 3.8826666666666667,
+      "grad_norm": 3.821939468383789,
+      "learning_rate": 4.4733096085409255e-05,
+      "loss": 1.3903,
+      "step": 8736
+    },
+    {
+      "epoch": 3.883111111111111,
+      "grad_norm": 2.8820078372955322,
+      "learning_rate": 4.4715302491103205e-05,
+      "loss": 0.5837,
+      "step": 8737
+    },
+    {
+      "epoch": 3.8835555555555556,
+      "grad_norm": 4.466034889221191,
+      "learning_rate": 4.4697508896797155e-05,
+      "loss": 1.4463,
+      "step": 8738
+    },
+    {
+      "epoch": 3.884,
+      "grad_norm": 3.6252734661102295,
+      "learning_rate": 4.4679715302491105e-05,
+      "loss": 1.1598,
+      "step": 8739
+    },
+    {
+      "epoch": 3.8844444444444446,
+      "grad_norm": 3.406493902206421,
+      "learning_rate": 4.4661921708185054e-05,
+      "loss": 0.9051,
+      "step": 8740
+    },
+    {
+      "epoch": 3.884888888888889,
+      "grad_norm": 3.9739794731140137,
+      "learning_rate": 4.464412811387901e-05,
+      "loss": 1.7437,
+      "step": 8741
+    },
+    {
+      "epoch": 3.8853333333333335,
+      "grad_norm": 4.158414840698242,
+      "learning_rate": 4.4626334519572954e-05,
+      "loss": 1.689,
+      "step": 8742
+    },
+    {
+      "epoch": 3.8857777777777778,
+      "grad_norm": 4.537271022796631,
+      "learning_rate": 4.4608540925266903e-05,
+      "loss": 1.1898,
+      "step": 8743
+    },
+    {
+      "epoch": 3.886222222222222,
+      "grad_norm": 2.962108612060547,
+      "learning_rate": 4.459074733096086e-05,
+      "loss": 1.0265,
+      "step": 8744
+    },
+    {
+      "epoch": 3.8866666666666667,
+      "grad_norm": 3.3965911865234375,
+      "learning_rate": 4.45729537366548e-05,
+      "loss": 1.1665,
+      "step": 8745
+    },
+    {
+      "epoch": 3.887111111111111,
+      "grad_norm": 4.494457244873047,
+      "learning_rate": 4.455516014234876e-05,
+      "loss": 1.4225,
+      "step": 8746
+    },
+    {
+      "epoch": 3.8875555555555557,
+      "grad_norm": 4.060354709625244,
+      "learning_rate": 4.453736654804271e-05,
+      "loss": 1.1413,
+      "step": 8747
+    },
+    {
+      "epoch": 3.888,
+      "grad_norm": 4.178462982177734,
+      "learning_rate": 4.451957295373665e-05,
+      "loss": 1.4544,
+      "step": 8748
+    },
+    {
+      "epoch": 3.8884444444444446,
+      "grad_norm": 3.7196671962738037,
+      "learning_rate": 4.450177935943061e-05,
+      "loss": 1.0486,
+      "step": 8749
+    },
+    {
+      "epoch": 3.888888888888889,
+      "grad_norm": 5.249838352203369,
+      "learning_rate": 4.448398576512456e-05,
+      "loss": 1.2359,
+      "step": 8750
+    },
+    {
+      "epoch": 3.889333333333333,
+      "grad_norm": 2.497725248336792,
+      "learning_rate": 4.446619217081851e-05,
+      "loss": 1.4867,
+      "step": 8751
+    },
+    {
+      "epoch": 3.889777777777778,
+      "grad_norm": 2.9607138633728027,
+      "learning_rate": 4.444839857651246e-05,
+      "loss": 1.9622,
+      "step": 8752
+    },
+    {
+      "epoch": 3.8902222222222225,
+      "grad_norm": 2.8709583282470703,
+      "learning_rate": 4.443060498220641e-05,
+      "loss": 1.7152,
+      "step": 8753
+    },
+    {
+      "epoch": 3.8906666666666667,
+      "grad_norm": 2.264155387878418,
+      "learning_rate": 4.441281138790036e-05,
+      "loss": 1.1801,
+      "step": 8754
+    },
+    {
+      "epoch": 3.891111111111111,
+      "grad_norm": 2.909691333770752,
+      "learning_rate": 4.439501779359431e-05,
+      "loss": 1.4905,
+      "step": 8755
+    },
+    {
+      "epoch": 3.8915555555555557,
+      "grad_norm": 2.7087597846984863,
+      "learning_rate": 4.437722419928826e-05,
+      "loss": 1.0053,
+      "step": 8756
+    },
+    {
+      "epoch": 3.892,
+      "grad_norm": 3.0195884704589844,
+      "learning_rate": 4.435943060498221e-05,
+      "loss": 1.4873,
+      "step": 8757
+    },
+    {
+      "epoch": 3.8924444444444446,
+      "grad_norm": 1.9923820495605469,
+      "learning_rate": 4.4341637010676156e-05,
+      "loss": 0.7845,
+      "step": 8758
+    },
+    {
+      "epoch": 3.892888888888889,
+      "grad_norm": 3.6413817405700684,
+      "learning_rate": 4.4323843416370106e-05,
+      "loss": 1.5665,
+      "step": 8759
+    },
+    {
+      "epoch": 3.8933333333333335,
+      "grad_norm": 3.3323585987091064,
+      "learning_rate": 4.430604982206406e-05,
+      "loss": 1.3828,
+      "step": 8760
+    },
+    {
+      "epoch": 3.893777777777778,
+      "grad_norm": 3.166923999786377,
+      "learning_rate": 4.4288256227758005e-05,
+      "loss": 1.5736,
+      "step": 8761
+    },
+    {
+      "epoch": 3.894222222222222,
+      "grad_norm": 3.236964464187622,
+      "learning_rate": 4.427046263345196e-05,
+      "loss": 1.5058,
+      "step": 8762
+    },
+    {
+      "epoch": 3.8946666666666667,
+      "grad_norm": 3.7442686557769775,
+      "learning_rate": 4.425266903914591e-05,
+      "loss": 1.1915,
+      "step": 8763
+    },
+    {
+      "epoch": 3.895111111111111,
+      "grad_norm": 3.836761236190796,
+      "learning_rate": 4.4234875444839854e-05,
+      "loss": 1.8305,
+      "step": 8764
+    },
+    {
+      "epoch": 3.8955555555555557,
+      "grad_norm": 3.648864984512329,
+      "learning_rate": 4.421708185053381e-05,
+      "loss": 1.4003,
+      "step": 8765
+    },
+    {
+      "epoch": 3.896,
+      "grad_norm": 3.3367178440093994,
+      "learning_rate": 4.419928825622776e-05,
+      "loss": 1.5771,
+      "step": 8766
+    },
+    {
+      "epoch": 3.8964444444444446,
+      "grad_norm": 4.111900806427002,
+      "learning_rate": 4.418149466192171e-05,
+      "loss": 1.5626,
+      "step": 8767
+    },
+    {
+      "epoch": 3.896888888888889,
+      "grad_norm": 3.5727264881134033,
+      "learning_rate": 4.416370106761566e-05,
+      "loss": 1.5124,
+      "step": 8768
+    },
+    {
+      "epoch": 3.897333333333333,
+      "grad_norm": 3.073716640472412,
+      "learning_rate": 4.414590747330961e-05,
+      "loss": 1.3135,
+      "step": 8769
+    },
+    {
+      "epoch": 3.897777777777778,
+      "grad_norm": 3.225069046020508,
+      "learning_rate": 4.4128113879003566e-05,
+      "loss": 1.4713,
+      "step": 8770
+    },
+    {
+      "epoch": 3.8982222222222225,
+      "grad_norm": 3.486926555633545,
+      "learning_rate": 4.411032028469751e-05,
+      "loss": 1.2325,
+      "step": 8771
+    },
+    {
+      "epoch": 3.8986666666666667,
+      "grad_norm": 3.7756948471069336,
+      "learning_rate": 4.409252669039146e-05,
+      "loss": 1.4494,
+      "step": 8772
+    },
+    {
+      "epoch": 3.899111111111111,
+      "grad_norm": 3.8202269077301025,
+      "learning_rate": 4.4074733096085415e-05,
+      "loss": 1.3107,
+      "step": 8773
+    },
+    {
+      "epoch": 3.8995555555555557,
+      "grad_norm": 3.3835747241973877,
+      "learning_rate": 4.405693950177936e-05,
+      "loss": 1.5086,
+      "step": 8774
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 3.0347068309783936,
+      "learning_rate": 4.403914590747331e-05,
+      "loss": 1.3952,
+      "step": 8775
+    },
+    {
+      "epoch": 3.9004444444444446,
+      "grad_norm": 3.6462948322296143,
+      "learning_rate": 4.4021352313167265e-05,
+      "loss": 1.0433,
+      "step": 8776
+    },
+    {
+      "epoch": 3.900888888888889,
+      "grad_norm": 4.498116970062256,
+      "learning_rate": 4.400355871886121e-05,
+      "loss": 1.5242,
+      "step": 8777
+    },
+    {
+      "epoch": 3.9013333333333335,
+      "grad_norm": 3.007981061935425,
+      "learning_rate": 4.3985765124555164e-05,
+      "loss": 1.2649,
+      "step": 8778
+    },
+    {
+      "epoch": 3.901777777777778,
+      "grad_norm": 4.1256208419799805,
+      "learning_rate": 4.3967971530249114e-05,
+      "loss": 0.9533,
+      "step": 8779
+    },
+    {
+      "epoch": 3.902222222222222,
+      "grad_norm": 3.121368169784546,
+      "learning_rate": 4.395017793594306e-05,
+      "loss": 1.1291,
+      "step": 8780
+    },
+    {
+      "epoch": 3.9026666666666667,
+      "grad_norm": 2.9976303577423096,
+      "learning_rate": 4.393238434163701e-05,
+      "loss": 1.0546,
+      "step": 8781
+    },
+    {
+      "epoch": 3.903111111111111,
+      "grad_norm": 3.8373613357543945,
+      "learning_rate": 4.391459074733096e-05,
+      "loss": 1.2271,
+      "step": 8782
+    },
+    {
+      "epoch": 3.9035555555555557,
+      "grad_norm": 3.4057092666625977,
+      "learning_rate": 4.389679715302491e-05,
+      "loss": 1.0357,
+      "step": 8783
+    },
+    {
+      "epoch": 3.904,
+      "grad_norm": 3.2696046829223633,
+      "learning_rate": 4.387900355871886e-05,
+      "loss": 1.0851,
+      "step": 8784
+    },
+    {
+      "epoch": 3.9044444444444446,
+      "grad_norm": 3.9882287979125977,
+      "learning_rate": 4.386120996441281e-05,
+      "loss": 0.895,
+      "step": 8785
+    },
+    {
+      "epoch": 3.904888888888889,
+      "grad_norm": 2.950498104095459,
+      "learning_rate": 4.384341637010677e-05,
+      "loss": 1.1437,
+      "step": 8786
+    },
+    {
+      "epoch": 3.905333333333333,
+      "grad_norm": 3.5203466415405273,
+      "learning_rate": 4.382562277580071e-05,
+      "loss": 1.3823,
+      "step": 8787
+    },
+    {
+      "epoch": 3.905777777777778,
+      "grad_norm": 2.380671262741089,
+      "learning_rate": 4.380782918149466e-05,
+      "loss": 0.7028,
+      "step": 8788
+    },
+    {
+      "epoch": 3.9062222222222225,
+      "grad_norm": 3.788663625717163,
+      "learning_rate": 4.379003558718862e-05,
+      "loss": 1.3234,
+      "step": 8789
+    },
+    {
+      "epoch": 3.9066666666666667,
+      "grad_norm": 2.707402467727661,
+      "learning_rate": 4.377224199288256e-05,
+      "loss": 0.5875,
+      "step": 8790
+    },
+    {
+      "epoch": 3.907111111111111,
+      "grad_norm": 3.8801896572113037,
+      "learning_rate": 4.375444839857652e-05,
+      "loss": 1.3117,
+      "step": 8791
+    },
+    {
+      "epoch": 3.9075555555555557,
+      "grad_norm": 3.4378976821899414,
+      "learning_rate": 4.373665480427047e-05,
+      "loss": 1.6461,
+      "step": 8792
+    },
+    {
+      "epoch": 3.908,
+      "grad_norm": 4.15399169921875,
+      "learning_rate": 4.371886120996441e-05,
+      "loss": 1.313,
+      "step": 8793
+    },
+    {
+      "epoch": 3.9084444444444446,
+      "grad_norm": 5.039565086364746,
+      "learning_rate": 4.3701067615658366e-05,
+      "loss": 1.3966,
+      "step": 8794
+    },
+    {
+      "epoch": 3.908888888888889,
+      "grad_norm": 4.0032854080200195,
+      "learning_rate": 4.3683274021352316e-05,
+      "loss": 1.4609,
+      "step": 8795
+    },
+    {
+      "epoch": 3.9093333333333335,
+      "grad_norm": 3.7062151432037354,
+      "learning_rate": 4.3665480427046266e-05,
+      "loss": 1.0876,
+      "step": 8796
+    },
+    {
+      "epoch": 3.909777777777778,
+      "grad_norm": 4.596202373504639,
+      "learning_rate": 4.3647686832740216e-05,
+      "loss": 1.5062,
+      "step": 8797
+    },
+    {
+      "epoch": 3.910222222222222,
+      "grad_norm": 4.020972728729248,
+      "learning_rate": 4.3629893238434165e-05,
+      "loss": 1.3268,
+      "step": 8798
+    },
+    {
+      "epoch": 3.9106666666666667,
+      "grad_norm": 4.46404504776001,
+      "learning_rate": 4.3612099644128115e-05,
+      "loss": 0.916,
+      "step": 8799
+    },
+    {
+      "epoch": 3.911111111111111,
+      "grad_norm": 3.5452182292938232,
+      "learning_rate": 4.3594306049822065e-05,
+      "loss": 0.8356,
+      "step": 8800
+    },
+    {
+      "epoch": 3.9115555555555557,
+      "grad_norm": 2.5292141437530518,
+      "learning_rate": 4.3576512455516014e-05,
+      "loss": 2.0216,
+      "step": 8801
+    },
+    {
+      "epoch": 3.912,
+      "grad_norm": 2.2973062992095947,
+      "learning_rate": 4.355871886120997e-05,
+      "loss": 0.8748,
+      "step": 8802
+    },
+    {
+      "epoch": 3.9124444444444446,
+      "grad_norm": 2.9960880279541016,
+      "learning_rate": 4.3540925266903914e-05,
+      "loss": 1.6939,
+      "step": 8803
+    },
+    {
+      "epoch": 3.912888888888889,
+      "grad_norm": 2.8162448406219482,
+      "learning_rate": 4.3523131672597864e-05,
+      "loss": 1.1804,
+      "step": 8804
+    },
+    {
+      "epoch": 3.913333333333333,
+      "grad_norm": 3.2202110290527344,
+      "learning_rate": 4.350533807829182e-05,
+      "loss": 1.2968,
+      "step": 8805
+    },
+    {
+      "epoch": 3.913777777777778,
+      "grad_norm": 3.30061936378479,
+      "learning_rate": 4.348754448398576e-05,
+      "loss": 1.4436,
+      "step": 8806
+    },
+    {
+      "epoch": 3.9142222222222225,
+      "grad_norm": 3.3244714736938477,
+      "learning_rate": 4.346975088967972e-05,
+      "loss": 1.6709,
+      "step": 8807
+    },
+    {
+      "epoch": 3.9146666666666667,
+      "grad_norm": 3.8059160709381104,
+      "learning_rate": 4.345195729537367e-05,
+      "loss": 1.5652,
+      "step": 8808
+    },
+    {
+      "epoch": 3.915111111111111,
+      "grad_norm": 4.244156360626221,
+      "learning_rate": 4.343416370106761e-05,
+      "loss": 1.6886,
+      "step": 8809
+    },
+    {
+      "epoch": 3.9155555555555557,
+      "grad_norm": 3.246370553970337,
+      "learning_rate": 4.341637010676157e-05,
+      "loss": 1.2196,
+      "step": 8810
+    },
+    {
+      "epoch": 3.916,
+      "grad_norm": 3.3758699893951416,
+      "learning_rate": 4.339857651245552e-05,
+      "loss": 1.3343,
+      "step": 8811
+    },
+    {
+      "epoch": 3.916444444444444,
+      "grad_norm": 3.9745421409606934,
+      "learning_rate": 4.338078291814947e-05,
+      "loss": 1.5243,
+      "step": 8812
+    },
+    {
+      "epoch": 3.916888888888889,
+      "grad_norm": 2.7803597450256348,
+      "learning_rate": 4.336298932384342e-05,
+      "loss": 0.7635,
+      "step": 8813
+    },
+    {
+      "epoch": 3.9173333333333336,
+      "grad_norm": 2.751984119415283,
+      "learning_rate": 4.334519572953737e-05,
+      "loss": 1.0825,
+      "step": 8814
+    },
+    {
+      "epoch": 3.917777777777778,
+      "grad_norm": 3.5463430881500244,
+      "learning_rate": 4.3327402135231324e-05,
+      "loss": 0.7509,
+      "step": 8815
+    },
+    {
+      "epoch": 3.918222222222222,
+      "grad_norm": 2.5409305095672607,
+      "learning_rate": 4.330960854092527e-05,
+      "loss": 0.7807,
+      "step": 8816
+    },
+    {
+      "epoch": 3.9186666666666667,
+      "grad_norm": 3.2667789459228516,
+      "learning_rate": 4.329181494661922e-05,
+      "loss": 1.3171,
+      "step": 8817
+    },
+    {
+      "epoch": 3.919111111111111,
+      "grad_norm": 2.9117629528045654,
+      "learning_rate": 4.327402135231317e-05,
+      "loss": 1.1148,
+      "step": 8818
+    },
+    {
+      "epoch": 3.9195555555555557,
+      "grad_norm": 3.8055434226989746,
+      "learning_rate": 4.3256227758007116e-05,
+      "loss": 1.6964,
+      "step": 8819
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 3.8291947841644287,
+      "learning_rate": 4.3238434163701066e-05,
+      "loss": 1.5032,
+      "step": 8820
+    },
+    {
+      "epoch": 3.9204444444444446,
+      "grad_norm": 3.757000207901001,
+      "learning_rate": 4.322064056939502e-05,
+      "loss": 1.4395,
+      "step": 8821
+    },
+    {
+      "epoch": 3.920888888888889,
+      "grad_norm": 3.474553108215332,
+      "learning_rate": 4.3202846975088965e-05,
+      "loss": 1.2536,
+      "step": 8822
+    },
+    {
+      "epoch": 3.921333333333333,
+      "grad_norm": 3.879840612411499,
+      "learning_rate": 4.318505338078292e-05,
+      "loss": 1.3524,
+      "step": 8823
+    },
+    {
+      "epoch": 3.921777777777778,
+      "grad_norm": 3.432889699935913,
+      "learning_rate": 4.316725978647687e-05,
+      "loss": 1.5825,
+      "step": 8824
+    },
+    {
+      "epoch": 3.9222222222222225,
+      "grad_norm": 3.799661159515381,
+      "learning_rate": 4.314946619217082e-05,
+      "loss": 1.4398,
+      "step": 8825
+    },
+    {
+      "epoch": 3.9226666666666667,
+      "grad_norm": 3.7226650714874268,
+      "learning_rate": 4.313167259786477e-05,
+      "loss": 1.1822,
+      "step": 8826
+    },
+    {
+      "epoch": 3.923111111111111,
+      "grad_norm": 3.7417545318603516,
+      "learning_rate": 4.311387900355872e-05,
+      "loss": 1.4244,
+      "step": 8827
+    },
+    {
+      "epoch": 3.9235555555555557,
+      "grad_norm": 3.5519144535064697,
+      "learning_rate": 4.309608540925267e-05,
+      "loss": 1.4199,
+      "step": 8828
+    },
+    {
+      "epoch": 3.924,
+      "grad_norm": 3.4887189865112305,
+      "learning_rate": 4.307829181494662e-05,
+      "loss": 1.4381,
+      "step": 8829
+    },
+    {
+      "epoch": 3.924444444444444,
+      "grad_norm": 3.223085880279541,
+      "learning_rate": 4.306049822064057e-05,
+      "loss": 1.0015,
+      "step": 8830
+    },
+    {
+      "epoch": 3.924888888888889,
+      "grad_norm": 3.3802876472473145,
+      "learning_rate": 4.3042704626334526e-05,
+      "loss": 1.0103,
+      "step": 8831
+    },
+    {
+      "epoch": 3.9253333333333336,
+      "grad_norm": 3.1447699069976807,
+      "learning_rate": 4.302491103202847e-05,
+      "loss": 1.2326,
+      "step": 8832
+    },
+    {
+      "epoch": 3.925777777777778,
+      "grad_norm": 3.206096887588501,
+      "learning_rate": 4.300711743772242e-05,
+      "loss": 0.9133,
+      "step": 8833
+    },
+    {
+      "epoch": 3.926222222222222,
+      "grad_norm": 3.67698335647583,
+      "learning_rate": 4.2989323843416376e-05,
+      "loss": 1.1838,
+      "step": 8834
+    },
+    {
+      "epoch": 3.9266666666666667,
+      "grad_norm": 3.849926233291626,
+      "learning_rate": 4.297153024911032e-05,
+      "loss": 1.1519,
+      "step": 8835
+    },
+    {
+      "epoch": 3.927111111111111,
+      "grad_norm": 3.2638113498687744,
+      "learning_rate": 4.2953736654804275e-05,
+      "loss": 1.0861,
+      "step": 8836
+    },
+    {
+      "epoch": 3.9275555555555557,
+      "grad_norm": 2.141021728515625,
+      "learning_rate": 4.2935943060498225e-05,
+      "loss": 0.3943,
+      "step": 8837
+    },
+    {
+      "epoch": 3.928,
+      "grad_norm": 2.9865190982818604,
+      "learning_rate": 4.291814946619217e-05,
+      "loss": 0.9882,
+      "step": 8838
+    },
+    {
+      "epoch": 3.9284444444444446,
+      "grad_norm": 3.7992167472839355,
+      "learning_rate": 4.2900355871886124e-05,
+      "loss": 1.3604,
+      "step": 8839
+    },
+    {
+      "epoch": 3.928888888888889,
+      "grad_norm": 5.145820617675781,
+      "learning_rate": 4.2882562277580074e-05,
+      "loss": 1.3215,
+      "step": 8840
+    },
+    {
+      "epoch": 3.929333333333333,
+      "grad_norm": 3.7761855125427246,
+      "learning_rate": 4.2864768683274024e-05,
+      "loss": 1.3284,
+      "step": 8841
+    },
+    {
+      "epoch": 3.929777777777778,
+      "grad_norm": 4.058622360229492,
+      "learning_rate": 4.284697508896797e-05,
+      "loss": 1.3971,
+      "step": 8842
+    },
+    {
+      "epoch": 3.930222222222222,
+      "grad_norm": 4.228296279907227,
+      "learning_rate": 4.282918149466192e-05,
+      "loss": 1.4811,
+      "step": 8843
+    },
+    {
+      "epoch": 3.9306666666666668,
+      "grad_norm": 4.4582085609436035,
+      "learning_rate": 4.281138790035587e-05,
+      "loss": 1.2761,
+      "step": 8844
+    },
+    {
+      "epoch": 3.931111111111111,
+      "grad_norm": 3.634289026260376,
+      "learning_rate": 4.279359430604982e-05,
+      "loss": 0.9966,
+      "step": 8845
+    },
+    {
+      "epoch": 3.9315555555555557,
+      "grad_norm": 5.011377334594727,
+      "learning_rate": 4.277580071174377e-05,
+      "loss": 1.1816,
+      "step": 8846
+    },
+    {
+      "epoch": 3.932,
+      "grad_norm": 4.198838233947754,
+      "learning_rate": 4.275800711743773e-05,
+      "loss": 1.4124,
+      "step": 8847
+    },
+    {
+      "epoch": 3.932444444444444,
+      "grad_norm": 4.471510887145996,
+      "learning_rate": 4.274021352313167e-05,
+      "loss": 1.0095,
+      "step": 8848
+    },
+    {
+      "epoch": 3.932888888888889,
+      "grad_norm": 3.8801698684692383,
+      "learning_rate": 4.272241992882562e-05,
+      "loss": 0.4851,
+      "step": 8849
+    },
+    {
+      "epoch": 3.9333333333333336,
+      "grad_norm": 3.203237533569336,
+      "learning_rate": 4.270462633451958e-05,
+      "loss": 0.411,
+      "step": 8850
+    },
+    {
+      "epoch": 3.933777777777778,
+      "grad_norm": 2.3682878017425537,
+      "learning_rate": 4.268683274021352e-05,
+      "loss": 2.0329,
+      "step": 8851
+    },
+    {
+      "epoch": 3.934222222222222,
+      "grad_norm": 2.3607728481292725,
+      "learning_rate": 4.266903914590748e-05,
+      "loss": 1.0491,
+      "step": 8852
+    },
+    {
+      "epoch": 3.9346666666666668,
+      "grad_norm": 3.4332642555236816,
+      "learning_rate": 4.265124555160143e-05,
+      "loss": 1.532,
+      "step": 8853
+    },
+    {
+      "epoch": 3.935111111111111,
+      "grad_norm": 3.4008309841156006,
+      "learning_rate": 4.263345195729538e-05,
+      "loss": 1.6618,
+      "step": 8854
+    },
+    {
+      "epoch": 3.9355555555555557,
+      "grad_norm": 3.161529779434204,
+      "learning_rate": 4.2615658362989326e-05,
+      "loss": 1.4769,
+      "step": 8855
+    },
+    {
+      "epoch": 3.936,
+      "grad_norm": 3.2563178539276123,
+      "learning_rate": 4.2597864768683276e-05,
+      "loss": 1.8546,
+      "step": 8856
+    },
+    {
+      "epoch": 3.9364444444444446,
+      "grad_norm": 3.0360910892486572,
+      "learning_rate": 4.2580071174377226e-05,
+      "loss": 1.0507,
+      "step": 8857
+    },
+    {
+      "epoch": 3.936888888888889,
+      "grad_norm": 3.976771116256714,
+      "learning_rate": 4.2562277580071176e-05,
+      "loss": 1.6975,
+      "step": 8858
+    },
+    {
+      "epoch": 3.937333333333333,
+      "grad_norm": 3.122298002243042,
+      "learning_rate": 4.2544483985765125e-05,
+      "loss": 1.4874,
+      "step": 8859
+    },
+    {
+      "epoch": 3.937777777777778,
+      "grad_norm": 3.1202025413513184,
+      "learning_rate": 4.252669039145908e-05,
+      "loss": 1.4433,
+      "step": 8860
+    },
+    {
+      "epoch": 3.938222222222222,
+      "grad_norm": 4.454637050628662,
+      "learning_rate": 4.2508896797153025e-05,
+      "loss": 1.6434,
+      "step": 8861
+    },
+    {
+      "epoch": 3.9386666666666668,
+      "grad_norm": 3.968618392944336,
+      "learning_rate": 4.2491103202846975e-05,
+      "loss": 1.365,
+      "step": 8862
+    },
+    {
+      "epoch": 3.939111111111111,
+      "grad_norm": 3.396158456802368,
+      "learning_rate": 4.247330960854093e-05,
+      "loss": 1.464,
+      "step": 8863
+    },
+    {
+      "epoch": 3.9395555555555557,
+      "grad_norm": 2.9130702018737793,
+      "learning_rate": 4.2455516014234874e-05,
+      "loss": 0.9835,
+      "step": 8864
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 3.9278974533081055,
+      "learning_rate": 4.2437722419928824e-05,
+      "loss": 1.585,
+      "step": 8865
+    },
+    {
+      "epoch": 3.940444444444444,
+      "grad_norm": 4.043193817138672,
+      "learning_rate": 4.241992882562278e-05,
+      "loss": 1.4772,
+      "step": 8866
+    },
+    {
+      "epoch": 3.940888888888889,
+      "grad_norm": 3.436844825744629,
+      "learning_rate": 4.240213523131672e-05,
+      "loss": 1.0931,
+      "step": 8867
+    },
+    {
+      "epoch": 3.9413333333333336,
+      "grad_norm": 3.061730146408081,
+      "learning_rate": 4.238434163701068e-05,
+      "loss": 1.3839,
+      "step": 8868
+    },
+    {
+      "epoch": 3.941777777777778,
+      "grad_norm": 3.5630033016204834,
+      "learning_rate": 4.236654804270463e-05,
+      "loss": 1.4972,
+      "step": 8869
+    },
+    {
+      "epoch": 3.942222222222222,
+      "grad_norm": 5.172977924346924,
+      "learning_rate": 4.234875444839858e-05,
+      "loss": 1.0713,
+      "step": 8870
+    },
+    {
+      "epoch": 3.9426666666666668,
+      "grad_norm": 3.7209513187408447,
+      "learning_rate": 4.233096085409253e-05,
+      "loss": 1.5315,
+      "step": 8871
+    },
+    {
+      "epoch": 3.943111111111111,
+      "grad_norm": 3.7008655071258545,
+      "learning_rate": 4.231316725978648e-05,
+      "loss": 1.2798,
+      "step": 8872
+    },
+    {
+      "epoch": 3.9435555555555557,
+      "grad_norm": 3.4888625144958496,
+      "learning_rate": 4.229537366548043e-05,
+      "loss": 1.139,
+      "step": 8873
+    },
+    {
+      "epoch": 3.944,
+      "grad_norm": 3.5707790851593018,
+      "learning_rate": 4.227758007117438e-05,
+      "loss": 1.2187,
+      "step": 8874
+    },
+    {
+      "epoch": 3.9444444444444446,
+      "grad_norm": 3.2470076084136963,
+      "learning_rate": 4.225978647686833e-05,
+      "loss": 1.5406,
+      "step": 8875
+    },
+    {
+      "epoch": 3.944888888888889,
+      "grad_norm": 3.3586134910583496,
+      "learning_rate": 4.2241992882562284e-05,
+      "loss": 1.3913,
+      "step": 8876
+    },
+    {
+      "epoch": 3.945333333333333,
+      "grad_norm": 4.539661407470703,
+      "learning_rate": 4.222419928825623e-05,
+      "loss": 1.8586,
+      "step": 8877
+    },
+    {
+      "epoch": 3.945777777777778,
+      "grad_norm": 3.6496472358703613,
+      "learning_rate": 4.220640569395018e-05,
+      "loss": 1.5811,
+      "step": 8878
+    },
+    {
+      "epoch": 3.946222222222222,
+      "grad_norm": 2.9896080493927,
+      "learning_rate": 4.218861209964413e-05,
+      "loss": 1.1142,
+      "step": 8879
+    },
+    {
+      "epoch": 3.9466666666666668,
+      "grad_norm": 4.495405673980713,
+      "learning_rate": 4.2170818505338076e-05,
+      "loss": 1.316,
+      "step": 8880
+    },
+    {
+      "epoch": 3.947111111111111,
+      "grad_norm": 2.01225209236145,
+      "learning_rate": 4.215302491103203e-05,
+      "loss": 0.4757,
+      "step": 8881
+    },
+    {
+      "epoch": 3.9475555555555557,
+      "grad_norm": 2.771531343460083,
+      "learning_rate": 4.213523131672598e-05,
+      "loss": 0.4966,
+      "step": 8882
+    },
+    {
+      "epoch": 3.948,
+      "grad_norm": 3.772594928741455,
+      "learning_rate": 4.211743772241993e-05,
+      "loss": 1.0786,
+      "step": 8883
+    },
+    {
+      "epoch": 3.948444444444444,
+      "grad_norm": 3.722163438796997,
+      "learning_rate": 4.209964412811388e-05,
+      "loss": 1.3533,
+      "step": 8884
+    },
+    {
+      "epoch": 3.948888888888889,
+      "grad_norm": 4.219447135925293,
+      "learning_rate": 4.208185053380783e-05,
+      "loss": 1.7263,
+      "step": 8885
+    },
+    {
+      "epoch": 3.9493333333333336,
+      "grad_norm": 3.999436855316162,
+      "learning_rate": 4.206405693950178e-05,
+      "loss": 1.3957,
+      "step": 8886
+    },
+    {
+      "epoch": 3.949777777777778,
+      "grad_norm": 3.6108932495117188,
+      "learning_rate": 4.204626334519573e-05,
+      "loss": 1.2958,
+      "step": 8887
+    },
+    {
+      "epoch": 3.950222222222222,
+      "grad_norm": 3.92063570022583,
+      "learning_rate": 4.202846975088968e-05,
+      "loss": 1.4854,
+      "step": 8888
+    },
+    {
+      "epoch": 3.9506666666666668,
+      "grad_norm": 3.5336453914642334,
+      "learning_rate": 4.201067615658363e-05,
+      "loss": 1.0695,
+      "step": 8889
+    },
+    {
+      "epoch": 3.951111111111111,
+      "grad_norm": 4.354097366333008,
+      "learning_rate": 4.199288256227758e-05,
+      "loss": 1.4669,
+      "step": 8890
+    },
+    {
+      "epoch": 3.9515555555555557,
+      "grad_norm": 4.140068054199219,
+      "learning_rate": 4.197508896797153e-05,
+      "loss": 1.292,
+      "step": 8891
+    },
+    {
+      "epoch": 3.952,
+      "grad_norm": 4.120720863342285,
+      "learning_rate": 4.1957295373665486e-05,
+      "loss": 1.2629,
+      "step": 8892
+    },
+    {
+      "epoch": 3.9524444444444446,
+      "grad_norm": 4.0067901611328125,
+      "learning_rate": 4.193950177935943e-05,
+      "loss": 1.318,
+      "step": 8893
+    },
+    {
+      "epoch": 3.952888888888889,
+      "grad_norm": 3.640927314758301,
+      "learning_rate": 4.192170818505338e-05,
+      "loss": 0.8191,
+      "step": 8894
+    },
+    {
+      "epoch": 3.953333333333333,
+      "grad_norm": 3.907336473464966,
+      "learning_rate": 4.1903914590747336e-05,
+      "loss": 1.0409,
+      "step": 8895
+    },
+    {
+      "epoch": 3.953777777777778,
+      "grad_norm": 4.239986419677734,
+      "learning_rate": 4.188612099644128e-05,
+      "loss": 1.4091,
+      "step": 8896
+    },
+    {
+      "epoch": 3.954222222222222,
+      "grad_norm": 4.236495494842529,
+      "learning_rate": 4.1868327402135235e-05,
+      "loss": 1.1437,
+      "step": 8897
+    },
+    {
+      "epoch": 3.9546666666666668,
+      "grad_norm": 3.6864097118377686,
+      "learning_rate": 4.1850533807829185e-05,
+      "loss": 1.0221,
+      "step": 8898
+    },
+    {
+      "epoch": 3.955111111111111,
+      "grad_norm": 5.639880180358887,
+      "learning_rate": 4.1832740213523135e-05,
+      "loss": 1.8105,
+      "step": 8899
+    },
+    {
+      "epoch": 3.9555555555555557,
+      "grad_norm": 3.3932342529296875,
+      "learning_rate": 4.1814946619217084e-05,
+      "loss": 0.5343,
+      "step": 8900
+    },
+    {
+      "epoch": 3.956,
+      "grad_norm": 2.1324033737182617,
+      "learning_rate": 4.1797153024911034e-05,
+      "loss": 2.0662,
+      "step": 8901
+    },
+    {
+      "epoch": 3.956444444444444,
+      "grad_norm": 2.3241565227508545,
+      "learning_rate": 4.1779359430604984e-05,
+      "loss": 1.8588,
+      "step": 8902
+    },
+    {
+      "epoch": 3.956888888888889,
+      "grad_norm": 2.4009382724761963,
+      "learning_rate": 4.1761565836298933e-05,
+      "loss": 2.0112,
+      "step": 8903
+    },
+    {
+      "epoch": 3.9573333333333336,
+      "grad_norm": 2.540668487548828,
+      "learning_rate": 4.174377224199288e-05,
+      "loss": 1.4803,
+      "step": 8904
+    },
+    {
+      "epoch": 3.957777777777778,
+      "grad_norm": 2.241854190826416,
+      "learning_rate": 4.172597864768684e-05,
+      "loss": 1.6004,
+      "step": 8905
+    },
+    {
+      "epoch": 3.958222222222222,
+      "grad_norm": 3.1838226318359375,
+      "learning_rate": 4.170818505338078e-05,
+      "loss": 1.6044,
+      "step": 8906
+    },
+    {
+      "epoch": 3.958666666666667,
+      "grad_norm": 2.7227938175201416,
+      "learning_rate": 4.169039145907473e-05,
+      "loss": 1.3226,
+      "step": 8907
+    },
+    {
+      "epoch": 3.959111111111111,
+      "grad_norm": 2.933244466781616,
+      "learning_rate": 4.167259786476869e-05,
+      "loss": 1.3356,
+      "step": 8908
+    },
+    {
+      "epoch": 3.9595555555555557,
+      "grad_norm": 2.794614315032959,
+      "learning_rate": 4.165480427046263e-05,
+      "loss": 1.3394,
+      "step": 8909
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.1257002353668213,
+      "learning_rate": 4.163701067615658e-05,
+      "loss": 1.6075,
+      "step": 8910
+    },
+    {
+      "epoch": 3.9604444444444447,
+      "grad_norm": 3.743485927581787,
+      "learning_rate": 4.161921708185054e-05,
+      "loss": 1.9586,
+      "step": 8911
+    },
+    {
+      "epoch": 3.960888888888889,
+      "grad_norm": 2.8176426887512207,
+      "learning_rate": 4.160142348754449e-05,
+      "loss": 1.7985,
+      "step": 8912
+    },
+    {
+      "epoch": 3.961333333333333,
+      "grad_norm": 3.1608290672302246,
+      "learning_rate": 4.158362989323844e-05,
+      "loss": 1.729,
+      "step": 8913
+    },
+    {
+      "epoch": 3.961777777777778,
+      "grad_norm": 3.7492446899414062,
+      "learning_rate": 4.156583629893239e-05,
+      "loss": 1.606,
+      "step": 8914
+    },
+    {
+      "epoch": 3.962222222222222,
+      "grad_norm": 3.0357353687286377,
+      "learning_rate": 4.154804270462634e-05,
+      "loss": 1.2235,
+      "step": 8915
+    },
+    {
+      "epoch": 3.962666666666667,
+      "grad_norm": 3.4608840942382812,
+      "learning_rate": 4.1530249110320287e-05,
+      "loss": 1.6648,
+      "step": 8916
+    },
+    {
+      "epoch": 3.963111111111111,
+      "grad_norm": 3.740095615386963,
+      "learning_rate": 4.1512455516014236e-05,
+      "loss": 1.4451,
+      "step": 8917
+    },
+    {
+      "epoch": 3.9635555555555557,
+      "grad_norm": 3.2377631664276123,
+      "learning_rate": 4.1494661921708186e-05,
+      "loss": 1.3791,
+      "step": 8918
+    },
+    {
+      "epoch": 3.964,
+      "grad_norm": 2.5673611164093018,
+      "learning_rate": 4.1476868327402136e-05,
+      "loss": 0.6982,
+      "step": 8919
+    },
+    {
+      "epoch": 3.964444444444444,
+      "grad_norm": 3.1579864025115967,
+      "learning_rate": 4.1459074733096085e-05,
+      "loss": 1.1644,
+      "step": 8920
+    },
+    {
+      "epoch": 3.964888888888889,
+      "grad_norm": 3.7532455921173096,
+      "learning_rate": 4.144128113879004e-05,
+      "loss": 1.3643,
+      "step": 8921
+    },
+    {
+      "epoch": 3.9653333333333336,
+      "grad_norm": 3.0965757369995117,
+      "learning_rate": 4.1423487544483985e-05,
+      "loss": 1.5047,
+      "step": 8922
+    },
+    {
+      "epoch": 3.965777777777778,
+      "grad_norm": 3.5118284225463867,
+      "learning_rate": 4.1405693950177935e-05,
+      "loss": 1.7549,
+      "step": 8923
+    },
+    {
+      "epoch": 3.966222222222222,
+      "grad_norm": 3.520890712738037,
+      "learning_rate": 4.138790035587189e-05,
+      "loss": 1.4312,
+      "step": 8924
+    },
+    {
+      "epoch": 3.966666666666667,
+      "grad_norm": 3.2968950271606445,
+      "learning_rate": 4.1370106761565834e-05,
+      "loss": 1.1051,
+      "step": 8925
+    },
+    {
+      "epoch": 3.967111111111111,
+      "grad_norm": 3.584757089614868,
+      "learning_rate": 4.135231316725979e-05,
+      "loss": 1.4383,
+      "step": 8926
+    },
+    {
+      "epoch": 3.9675555555555553,
+      "grad_norm": 2.9149937629699707,
+      "learning_rate": 4.133451957295374e-05,
+      "loss": 0.7798,
+      "step": 8927
+    },
+    {
+      "epoch": 3.968,
+      "grad_norm": 4.206515312194824,
+      "learning_rate": 4.131672597864769e-05,
+      "loss": 1.2928,
+      "step": 8928
+    },
+    {
+      "epoch": 3.9684444444444447,
+      "grad_norm": 3.031240224838257,
+      "learning_rate": 4.129893238434164e-05,
+      "loss": 1.0482,
+      "step": 8929
+    },
+    {
+      "epoch": 3.968888888888889,
+      "grad_norm": 3.6411449909210205,
+      "learning_rate": 4.128113879003559e-05,
+      "loss": 1.0632,
+      "step": 8930
+    },
+    {
+      "epoch": 3.969333333333333,
+      "grad_norm": 3.8553061485290527,
+      "learning_rate": 4.126334519572954e-05,
+      "loss": 1.5348,
+      "step": 8931
+    },
+    {
+      "epoch": 3.969777777777778,
+      "grad_norm": 3.2213852405548096,
+      "learning_rate": 4.124555160142349e-05,
+      "loss": 1.4563,
+      "step": 8932
+    },
+    {
+      "epoch": 3.970222222222222,
+      "grad_norm": 3.013375759124756,
+      "learning_rate": 4.122775800711744e-05,
+      "loss": 1.1047,
+      "step": 8933
+    },
+    {
+      "epoch": 3.970666666666667,
+      "grad_norm": 3.3184897899627686,
+      "learning_rate": 4.120996441281139e-05,
+      "loss": 1.2822,
+      "step": 8934
+    },
+    {
+      "epoch": 3.971111111111111,
+      "grad_norm": 3.349609375,
+      "learning_rate": 4.119217081850534e-05,
+      "loss": 1.4031,
+      "step": 8935
+    },
+    {
+      "epoch": 3.9715555555555557,
+      "grad_norm": 3.315196990966797,
+      "learning_rate": 4.117437722419929e-05,
+      "loss": 1.3917,
+      "step": 8936
+    },
+    {
+      "epoch": 3.972,
+      "grad_norm": 3.5079121589660645,
+      "learning_rate": 4.1156583629893244e-05,
+      "loss": 1.3935,
+      "step": 8937
+    },
+    {
+      "epoch": 3.9724444444444442,
+      "grad_norm": 3.8228821754455566,
+      "learning_rate": 4.113879003558719e-05,
+      "loss": 1.2339,
+      "step": 8938
+    },
+    {
+      "epoch": 3.972888888888889,
+      "grad_norm": 3.724152088165283,
+      "learning_rate": 4.112099644128114e-05,
+      "loss": 1.03,
+      "step": 8939
+    },
+    {
+      "epoch": 3.9733333333333336,
+      "grad_norm": 3.229318141937256,
+      "learning_rate": 4.1103202846975093e-05,
+      "loss": 0.7204,
+      "step": 8940
+    },
+    {
+      "epoch": 3.973777777777778,
+      "grad_norm": 4.946290493011475,
+      "learning_rate": 4.108540925266904e-05,
+      "loss": 1.647,
+      "step": 8941
+    },
+    {
+      "epoch": 3.974222222222222,
+      "grad_norm": 4.947329521179199,
+      "learning_rate": 4.106761565836299e-05,
+      "loss": 1.4321,
+      "step": 8942
+    },
+    {
+      "epoch": 3.974666666666667,
+      "grad_norm": 3.631192684173584,
+      "learning_rate": 4.104982206405694e-05,
+      "loss": 1.2327,
+      "step": 8943
+    },
+    {
+      "epoch": 3.975111111111111,
+      "grad_norm": 3.9020490646362305,
+      "learning_rate": 4.103202846975089e-05,
+      "loss": 1.3794,
+      "step": 8944
+    },
+    {
+      "epoch": 3.9755555555555553,
+      "grad_norm": 4.423495769500732,
+      "learning_rate": 4.101423487544484e-05,
+      "loss": 1.5862,
+      "step": 8945
+    },
+    {
+      "epoch": 3.976,
+      "grad_norm": 4.301177024841309,
+      "learning_rate": 4.099644128113879e-05,
+      "loss": 0.9748,
+      "step": 8946
+    },
+    {
+      "epoch": 3.9764444444444447,
+      "grad_norm": 3.9020419120788574,
+      "learning_rate": 4.097864768683274e-05,
+      "loss": 0.873,
+      "step": 8947
+    },
+    {
+      "epoch": 3.976888888888889,
+      "grad_norm": 4.565361022949219,
+      "learning_rate": 4.096085409252669e-05,
+      "loss": 0.9833,
+      "step": 8948
+    },
+    {
+      "epoch": 3.977333333333333,
+      "grad_norm": 3.912226676940918,
+      "learning_rate": 4.094306049822064e-05,
+      "loss": 1.3495,
+      "step": 8949
+    },
+    {
+      "epoch": 3.977777777777778,
+      "grad_norm": 5.293661594390869,
+      "learning_rate": 4.09252669039146e-05,
+      "loss": 1.3508,
+      "step": 8950
+    },
+    {
+      "epoch": 3.978222222222222,
+      "grad_norm": 0.17634160816669464,
+      "learning_rate": 4.090747330960854e-05,
+      "loss": 0.0163,
+      "step": 8951
+    },
+    {
+      "epoch": 3.978666666666667,
+      "grad_norm": 2.6012349128723145,
+      "learning_rate": 4.088967971530249e-05,
+      "loss": 1.7285,
+      "step": 8952
+    },
+    {
+      "epoch": 3.979111111111111,
+      "grad_norm": 2.6007161140441895,
+      "learning_rate": 4.0871886120996447e-05,
+      "loss": 0.9842,
+      "step": 8953
+    },
+    {
+      "epoch": 3.9795555555555557,
+      "grad_norm": 2.962765693664551,
+      "learning_rate": 4.085409252669039e-05,
+      "loss": 1.6334,
+      "step": 8954
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 3.1964375972747803,
+      "learning_rate": 4.083629893238434e-05,
+      "loss": 1.7721,
+      "step": 8955
+    },
+    {
+      "epoch": 3.9804444444444442,
+      "grad_norm": 3.102666139602661,
+      "learning_rate": 4.0818505338078296e-05,
+      "loss": 1.5754,
+      "step": 8956
+    },
+    {
+      "epoch": 3.980888888888889,
+      "grad_norm": 3.193918466567993,
+      "learning_rate": 4.0800711743772245e-05,
+      "loss": 2.0491,
+      "step": 8957
+    },
+    {
+      "epoch": 3.981333333333333,
+      "grad_norm": 2.9397614002227783,
+      "learning_rate": 4.0782918149466195e-05,
+      "loss": 1.044,
+      "step": 8958
+    },
+    {
+      "epoch": 3.981777777777778,
+      "grad_norm": 3.4079766273498535,
+      "learning_rate": 4.0765124555160145e-05,
+      "loss": 1.0677,
+      "step": 8959
+    },
+    {
+      "epoch": 3.982222222222222,
+      "grad_norm": 3.466888666152954,
+      "learning_rate": 4.0747330960854095e-05,
+      "loss": 1.5662,
+      "step": 8960
+    },
+    {
+      "epoch": 3.982666666666667,
+      "grad_norm": 2.9022140502929688,
+      "learning_rate": 4.0729537366548044e-05,
+      "loss": 1.1119,
+      "step": 8961
+    },
+    {
+      "epoch": 3.983111111111111,
+      "grad_norm": 3.6132586002349854,
+      "learning_rate": 4.0711743772241994e-05,
+      "loss": 1.4747,
+      "step": 8962
+    },
+    {
+      "epoch": 3.9835555555555553,
+      "grad_norm": 3.143056631088257,
+      "learning_rate": 4.0693950177935944e-05,
+      "loss": 1.6553,
+      "step": 8963
+    },
+    {
+      "epoch": 3.984,
+      "grad_norm": 3.3530349731445312,
+      "learning_rate": 4.0676156583629894e-05,
+      "loss": 1.1596,
+      "step": 8964
+    },
+    {
+      "epoch": 3.9844444444444447,
+      "grad_norm": 3.0167648792266846,
+      "learning_rate": 4.065836298932384e-05,
+      "loss": 1.0196,
+      "step": 8965
+    },
+    {
+      "epoch": 3.984888888888889,
+      "grad_norm": 2.661754846572876,
+      "learning_rate": 4.06405693950178e-05,
+      "loss": 0.8042,
+      "step": 8966
+    },
+    {
+      "epoch": 3.985333333333333,
+      "grad_norm": 4.018869400024414,
+      "learning_rate": 4.062277580071174e-05,
+      "loss": 1.6149,
+      "step": 8967
+    },
+    {
+      "epoch": 3.985777777777778,
+      "grad_norm": 3.3034558296203613,
+      "learning_rate": 4.060498220640569e-05,
+      "loss": 1.0005,
+      "step": 8968
+    },
+    {
+      "epoch": 3.986222222222222,
+      "grad_norm": 3.277484178543091,
+      "learning_rate": 4.058718861209965e-05,
+      "loss": 1.4676,
+      "step": 8969
+    },
+    {
+      "epoch": 3.986666666666667,
+      "grad_norm": 2.8795299530029297,
+      "learning_rate": 4.05693950177936e-05,
+      "loss": 1.0455,
+      "step": 8970
+    },
+    {
+      "epoch": 3.987111111111111,
+      "grad_norm": 3.6743695735931396,
+      "learning_rate": 4.055160142348755e-05,
+      "loss": 1.3309,
+      "step": 8971
+    },
+    {
+      "epoch": 3.9875555555555557,
+      "grad_norm": 2.998657464981079,
+      "learning_rate": 4.05338078291815e-05,
+      "loss": 1.3431,
+      "step": 8972
+    },
+    {
+      "epoch": 3.988,
+      "grad_norm": 3.6148931980133057,
+      "learning_rate": 4.051601423487545e-05,
+      "loss": 1.167,
+      "step": 8973
+    },
+    {
+      "epoch": 3.9884444444444442,
+      "grad_norm": 3.467222213745117,
+      "learning_rate": 4.04982206405694e-05,
+      "loss": 1.2744,
+      "step": 8974
+    },
+    {
+      "epoch": 3.988888888888889,
+      "grad_norm": 3.517871618270874,
+      "learning_rate": 4.048042704626335e-05,
+      "loss": 1.4644,
+      "step": 8975
+    },
+    {
+      "epoch": 3.989333333333333,
+      "grad_norm": 3.028439998626709,
+      "learning_rate": 4.04626334519573e-05,
+      "loss": 0.8676,
+      "step": 8976
+    },
+    {
+      "epoch": 3.989777777777778,
+      "grad_norm": 3.8672285079956055,
+      "learning_rate": 4.044483985765125e-05,
+      "loss": 1.3043,
+      "step": 8977
+    },
+    {
+      "epoch": 3.990222222222222,
+      "grad_norm": 3.13926362991333,
+      "learning_rate": 4.0427046263345196e-05,
+      "loss": 1.0397,
+      "step": 8978
+    },
+    {
+      "epoch": 3.990666666666667,
+      "grad_norm": 3.7574684619903564,
+      "learning_rate": 4.0409252669039146e-05,
+      "loss": 1.2111,
+      "step": 8979
+    },
+    {
+      "epoch": 3.991111111111111,
+      "grad_norm": 3.468540906906128,
+      "learning_rate": 4.0391459074733096e-05,
+      "loss": 1.4037,
+      "step": 8980
+    },
+    {
+      "epoch": 3.9915555555555553,
+      "grad_norm": 2.5756216049194336,
+      "learning_rate": 4.0373665480427046e-05,
+      "loss": 0.6104,
+      "step": 8981
+    },
+    {
+      "epoch": 3.992,
+      "grad_norm": 4.789612293243408,
+      "learning_rate": 4.0355871886121e-05,
+      "loss": 1.3617,
+      "step": 8982
+    },
+    {
+      "epoch": 3.9924444444444447,
+      "grad_norm": 4.018329620361328,
+      "learning_rate": 4.0338078291814945e-05,
+      "loss": 1.3874,
+      "step": 8983
+    },
+    {
+      "epoch": 3.992888888888889,
+      "grad_norm": 3.488252639770508,
+      "learning_rate": 4.0320284697508895e-05,
+      "loss": 1.2444,
+      "step": 8984
+    },
+    {
+      "epoch": 3.993333333333333,
+      "grad_norm": 3.8759312629699707,
+      "learning_rate": 4.030249110320285e-05,
+      "loss": 1.1931,
+      "step": 8985
+    },
+    {
+      "epoch": 3.993777777777778,
+      "grad_norm": 2.9700896739959717,
+      "learning_rate": 4.02846975088968e-05,
+      "loss": 0.9424,
+      "step": 8986
+    },
+    {
+      "epoch": 3.994222222222222,
+      "grad_norm": 3.435788154602051,
+      "learning_rate": 4.026690391459075e-05,
+      "loss": 1.1437,
+      "step": 8987
+    },
+    {
+      "epoch": 3.994666666666667,
+      "grad_norm": 3.1736974716186523,
+      "learning_rate": 4.02491103202847e-05,
+      "loss": 1.014,
+      "step": 8988
+    },
+    {
+      "epoch": 3.995111111111111,
+      "grad_norm": 3.615291118621826,
+      "learning_rate": 4.023131672597865e-05,
+      "loss": 1.4715,
+      "step": 8989
+    },
+    {
+      "epoch": 3.9955555555555557,
+      "grad_norm": 3.378284215927124,
+      "learning_rate": 4.02135231316726e-05,
+      "loss": 1.2936,
+      "step": 8990
+    },
+    {
+      "epoch": 3.996,
+      "grad_norm": 3.3950233459472656,
+      "learning_rate": 4.019572953736655e-05,
+      "loss": 1.1708,
+      "step": 8991
+    },
+    {
+      "epoch": 3.9964444444444442,
+      "grad_norm": 3.4889776706695557,
+      "learning_rate": 4.01779359430605e-05,
+      "loss": 1.192,
+      "step": 8992
+    },
+    {
+      "epoch": 3.996888888888889,
+      "grad_norm": 3.3020076751708984,
+      "learning_rate": 4.016014234875445e-05,
+      "loss": 0.991,
+      "step": 8993
+    },
+    {
+      "epoch": 3.997333333333333,
+      "grad_norm": 4.46552038192749,
+      "learning_rate": 4.01423487544484e-05,
+      "loss": 1.4608,
+      "step": 8994
+    },
+    {
+      "epoch": 3.997777777777778,
+      "grad_norm": 4.276278495788574,
+      "learning_rate": 4.0124555160142355e-05,
+      "loss": 1.2873,
+      "step": 8995
+    },
+    {
+      "epoch": 3.998222222222222,
+      "grad_norm": 3.8281071186065674,
+      "learning_rate": 4.01067615658363e-05,
+      "loss": 1.461,
+      "step": 8996
+    },
+    {
+      "epoch": 3.998666666666667,
+      "grad_norm": 5.66308069229126,
+      "learning_rate": 4.008896797153025e-05,
+      "loss": 1.2658,
+      "step": 8997
+    },
+    {
+      "epoch": 3.999111111111111,
+      "grad_norm": 4.838554382324219,
+      "learning_rate": 4.0071174377224204e-05,
+      "loss": 1.175,
+      "step": 8998
+    },
+    {
+      "epoch": 3.9995555555555553,
+      "grad_norm": 4.6421990394592285,
+      "learning_rate": 4.0053380782918154e-05,
+      "loss": 1.2439,
+      "step": 8999
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.736137628555298,
+      "learning_rate": 4.00355871886121e-05,
+      "loss": 0.6085,
+      "step": 9000
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 2.81941556930542,
+      "eval_runtime": 47.453,
+      "eval_samples_per_second": 10.537,
+      "eval_steps_per_second": 10.537,
+      "step": 9000
+    },
+    {
+      "epoch": 4.000444444444445,
+      "grad_norm": 2.5390360355377197,
+      "learning_rate": 4.0017793594306054e-05,
+      "loss": 1.3083,
+      "step": 9001
+    },
+    {
+      "epoch": 4.0008888888888885,
+      "grad_norm": 2.7254271507263184,
+      "learning_rate": 4e-05,
+      "loss": 0.5198,
+      "step": 9002
+    },
+    {
+      "epoch": 4.001333333333333,
+      "grad_norm": 2.502013683319092,
+      "learning_rate": 3.998220640569395e-05,
+      "loss": 1.3684,
+      "step": 9003
+    },
+    {
+      "epoch": 4.001777777777778,
+      "grad_norm": 2.1942808628082275,
+      "learning_rate": 3.99644128113879e-05,
+      "loss": 0.9158,
+      "step": 9004
+    },
+    {
+      "epoch": 4.002222222222223,
+      "grad_norm": 2.4862608909606934,
+      "learning_rate": 3.994661921708185e-05,
+      "loss": 1.1732,
+      "step": 9005
+    },
+    {
+      "epoch": 4.002666666666666,
+      "grad_norm": 2.521996259689331,
+      "learning_rate": 3.99288256227758e-05,
+      "loss": 0.9865,
+      "step": 9006
+    },
+    {
+      "epoch": 4.003111111111111,
+      "grad_norm": 2.7241384983062744,
+      "learning_rate": 3.991103202846975e-05,
+      "loss": 0.9629,
+      "step": 9007
+    },
+    {
+      "epoch": 4.003555555555556,
+      "grad_norm": 2.860523223876953,
+      "learning_rate": 3.98932384341637e-05,
+      "loss": 0.9462,
+      "step": 9008
+    },
+    {
+      "epoch": 4.004,
+      "grad_norm": 2.424485921859741,
+      "learning_rate": 3.987544483985765e-05,
+      "loss": 0.762,
+      "step": 9009
+    },
+    {
+      "epoch": 4.004444444444444,
+      "grad_norm": 3.2008659839630127,
+      "learning_rate": 3.98576512455516e-05,
+      "loss": 1.2424,
+      "step": 9010
+    },
+    {
+      "epoch": 4.004888888888889,
+      "grad_norm": 3.4407267570495605,
+      "learning_rate": 3.983985765124556e-05,
+      "loss": 0.9112,
+      "step": 9011
+    },
+    {
+      "epoch": 4.005333333333334,
+      "grad_norm": 2.80182147026062,
+      "learning_rate": 3.98220640569395e-05,
+      "loss": 1.006,
+      "step": 9012
+    },
+    {
+      "epoch": 4.005777777777777,
+      "grad_norm": 3.254213571548462,
+      "learning_rate": 3.980427046263345e-05,
+      "loss": 1.0918,
+      "step": 9013
+    },
+    {
+      "epoch": 4.006222222222222,
+      "grad_norm": 3.264662027359009,
+      "learning_rate": 3.978647686832741e-05,
+      "loss": 1.248,
+      "step": 9014
+    },
+    {
+      "epoch": 4.006666666666667,
+      "grad_norm": 2.7406811714172363,
+      "learning_rate": 3.9768683274021356e-05,
+      "loss": 0.7658,
+      "step": 9015
+    },
+    {
+      "epoch": 4.0071111111111115,
+      "grad_norm": 3.4597320556640625,
+      "learning_rate": 3.9750889679715306e-05,
+      "loss": 0.9957,
+      "step": 9016
+    },
+    {
+      "epoch": 4.007555555555555,
+      "grad_norm": 3.379284620285034,
+      "learning_rate": 3.9733096085409256e-05,
+      "loss": 1.1522,
+      "step": 9017
+    },
+    {
+      "epoch": 4.008,
+      "grad_norm": 2.9539339542388916,
+      "learning_rate": 3.9715302491103206e-05,
+      "loss": 0.8746,
+      "step": 9018
+    },
+    {
+      "epoch": 4.008444444444445,
+      "grad_norm": 3.017645835876465,
+      "learning_rate": 3.9697508896797155e-05,
+      "loss": 0.7478,
+      "step": 9019
+    },
+    {
+      "epoch": 4.0088888888888885,
+      "grad_norm": 3.781395196914673,
+      "learning_rate": 3.9679715302491105e-05,
+      "loss": 1.0097,
+      "step": 9020
+    },
+    {
+      "epoch": 4.009333333333333,
+      "grad_norm": 3.5405569076538086,
+      "learning_rate": 3.9661921708185055e-05,
+      "loss": 0.7791,
+      "step": 9021
+    },
+    {
+      "epoch": 4.009777777777778,
+      "grad_norm": 4.20705509185791,
+      "learning_rate": 3.9644128113879004e-05,
+      "loss": 0.98,
+      "step": 9022
+    },
+    {
+      "epoch": 4.010222222222223,
+      "grad_norm": 3.9584195613861084,
+      "learning_rate": 3.9626334519572954e-05,
+      "loss": 0.6952,
+      "step": 9023
+    },
+    {
+      "epoch": 4.010666666666666,
+      "grad_norm": 4.0206522941589355,
+      "learning_rate": 3.9608540925266904e-05,
+      "loss": 1.0842,
+      "step": 9024
+    },
+    {
+      "epoch": 4.011111111111111,
+      "grad_norm": 4.080977439880371,
+      "learning_rate": 3.9590747330960854e-05,
+      "loss": 0.7917,
+      "step": 9025
+    },
+    {
+      "epoch": 4.011555555555556,
+      "grad_norm": 4.305196285247803,
+      "learning_rate": 3.95729537366548e-05,
+      "loss": 1.0147,
+      "step": 9026
+    },
+    {
+      "epoch": 4.012,
+      "grad_norm": 3.7779810428619385,
+      "learning_rate": 3.955516014234876e-05,
+      "loss": 0.764,
+      "step": 9027
+    },
+    {
+      "epoch": 4.012444444444444,
+      "grad_norm": 3.0636918544769287,
+      "learning_rate": 3.953736654804271e-05,
+      "loss": 0.5695,
+      "step": 9028
+    },
+    {
+      "epoch": 4.012888888888889,
+      "grad_norm": 3.5221712589263916,
+      "learning_rate": 3.951957295373665e-05,
+      "loss": 0.7406,
+      "step": 9029
+    },
+    {
+      "epoch": 4.013333333333334,
+      "grad_norm": 4.276998996734619,
+      "learning_rate": 3.950177935943061e-05,
+      "loss": 0.8573,
+      "step": 9030
+    },
+    {
+      "epoch": 4.0137777777777774,
+      "grad_norm": 5.169942855834961,
+      "learning_rate": 3.948398576512456e-05,
+      "loss": 1.1206,
+      "step": 9031
+    },
+    {
+      "epoch": 4.014222222222222,
+      "grad_norm": 3.651606798171997,
+      "learning_rate": 3.946619217081851e-05,
+      "loss": 0.6189,
+      "step": 9032
+    },
+    {
+      "epoch": 4.014666666666667,
+      "grad_norm": 4.276918411254883,
+      "learning_rate": 3.944839857651246e-05,
+      "loss": 1.2343,
+      "step": 9033
+    },
+    {
+      "epoch": 4.0151111111111115,
+      "grad_norm": 3.116567850112915,
+      "learning_rate": 3.943060498220641e-05,
+      "loss": 0.5037,
+      "step": 9034
+    },
+    {
+      "epoch": 4.015555555555555,
+      "grad_norm": 4.222073078155518,
+      "learning_rate": 3.941281138790036e-05,
+      "loss": 0.9321,
+      "step": 9035
+    },
+    {
+      "epoch": 4.016,
+      "grad_norm": 3.9893438816070557,
+      "learning_rate": 3.939501779359431e-05,
+      "loss": 1.0099,
+      "step": 9036
+    },
+    {
+      "epoch": 4.016444444444445,
+      "grad_norm": 4.333362102508545,
+      "learning_rate": 3.937722419928826e-05,
+      "loss": 1.0413,
+      "step": 9037
+    },
+    {
+      "epoch": 4.0168888888888885,
+      "grad_norm": 3.694157361984253,
+      "learning_rate": 3.935943060498221e-05,
+      "loss": 0.6196,
+      "step": 9038
+    },
+    {
+      "epoch": 4.017333333333333,
+      "grad_norm": 4.171072006225586,
+      "learning_rate": 3.9341637010676157e-05,
+      "loss": 0.6015,
+      "step": 9039
+    },
+    {
+      "epoch": 4.017777777777778,
+      "grad_norm": 4.184826374053955,
+      "learning_rate": 3.932384341637011e-05,
+      "loss": 0.9599,
+      "step": 9040
+    },
+    {
+      "epoch": 4.018222222222223,
+      "grad_norm": 3.0356812477111816,
+      "learning_rate": 3.9306049822064056e-05,
+      "loss": 0.5598,
+      "step": 9041
+    },
+    {
+      "epoch": 4.018666666666666,
+      "grad_norm": 5.620950222015381,
+      "learning_rate": 3.9288256227758006e-05,
+      "loss": 0.7523,
+      "step": 9042
+    },
+    {
+      "epoch": 4.019111111111111,
+      "grad_norm": 3.5355186462402344,
+      "learning_rate": 3.927046263345196e-05,
+      "loss": 0.5461,
+      "step": 9043
+    },
+    {
+      "epoch": 4.019555555555556,
+      "grad_norm": 4.568739414215088,
+      "learning_rate": 3.925266903914591e-05,
+      "loss": 0.8424,
+      "step": 9044
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 3.9112837314605713,
+      "learning_rate": 3.9234875444839855e-05,
+      "loss": 0.6756,
+      "step": 9045
+    },
+    {
+      "epoch": 4.020444444444444,
+      "grad_norm": 4.544727325439453,
+      "learning_rate": 3.921708185053381e-05,
+      "loss": 0.7065,
+      "step": 9046
+    },
+    {
+      "epoch": 4.020888888888889,
+      "grad_norm": 4.089946746826172,
+      "learning_rate": 3.919928825622776e-05,
+      "loss": 1.122,
+      "step": 9047
+    },
+    {
+      "epoch": 4.021333333333334,
+      "grad_norm": 5.61737060546875,
+      "learning_rate": 3.918149466192171e-05,
+      "loss": 0.4246,
+      "step": 9048
+    },
+    {
+      "epoch": 4.0217777777777775,
+      "grad_norm": 4.636961460113525,
+      "learning_rate": 3.916370106761566e-05,
+      "loss": 0.4196,
+      "step": 9049
+    },
+    {
+      "epoch": 4.022222222222222,
+      "grad_norm": 7.075558662414551,
+      "learning_rate": 3.914590747330961e-05,
+      "loss": 0.5413,
+      "step": 9050
+    },
+    {
+      "epoch": 4.022666666666667,
+      "grad_norm": 2.6508331298828125,
+      "learning_rate": 3.912811387900356e-05,
+      "loss": 1.1433,
+      "step": 9051
+    },
+    {
+      "epoch": 4.0231111111111115,
+      "grad_norm": 2.6102499961853027,
+      "learning_rate": 3.911032028469751e-05,
+      "loss": 1.4944,
+      "step": 9052
+    },
+    {
+      "epoch": 4.023555555555555,
+      "grad_norm": 2.353135108947754,
+      "learning_rate": 3.909252669039146e-05,
+      "loss": 0.6206,
+      "step": 9053
+    },
+    {
+      "epoch": 4.024,
+      "grad_norm": 2.1263370513916016,
+      "learning_rate": 3.907473309608541e-05,
+      "loss": 0.9243,
+      "step": 9054
+    },
+    {
+      "epoch": 4.024444444444445,
+      "grad_norm": 3.224517583847046,
+      "learning_rate": 3.905693950177936e-05,
+      "loss": 1.5233,
+      "step": 9055
+    },
+    {
+      "epoch": 4.0248888888888885,
+      "grad_norm": 3.488454580307007,
+      "learning_rate": 3.9039145907473315e-05,
+      "loss": 1.5544,
+      "step": 9056
+    },
+    {
+      "epoch": 4.025333333333333,
+      "grad_norm": 3.6243624687194824,
+      "learning_rate": 3.9021352313167265e-05,
+      "loss": 1.1879,
+      "step": 9057
+    },
+    {
+      "epoch": 4.025777777777778,
+      "grad_norm": 3.8051021099090576,
+      "learning_rate": 3.900355871886121e-05,
+      "loss": 1.3218,
+      "step": 9058
+    },
+    {
+      "epoch": 4.026222222222223,
+      "grad_norm": 3.4981696605682373,
+      "learning_rate": 3.8985765124555164e-05,
+      "loss": 1.0695,
+      "step": 9059
+    },
+    {
+      "epoch": 4.026666666666666,
+      "grad_norm": 3.527284860610962,
+      "learning_rate": 3.8967971530249114e-05,
+      "loss": 1.0332,
+      "step": 9060
+    },
+    {
+      "epoch": 4.027111111111111,
+      "grad_norm": 3.228292942047119,
+      "learning_rate": 3.8950177935943064e-05,
+      "loss": 1.0043,
+      "step": 9061
+    },
+    {
+      "epoch": 4.027555555555556,
+      "grad_norm": 3.101989984512329,
+      "learning_rate": 3.8932384341637014e-05,
+      "loss": 0.7491,
+      "step": 9062
+    },
+    {
+      "epoch": 4.028,
+      "grad_norm": 3.7942066192626953,
+      "learning_rate": 3.891459074733096e-05,
+      "loss": 1.1937,
+      "step": 9063
+    },
+    {
+      "epoch": 4.028444444444444,
+      "grad_norm": 4.187127590179443,
+      "learning_rate": 3.889679715302491e-05,
+      "loss": 1.0301,
+      "step": 9064
+    },
+    {
+      "epoch": 4.028888888888889,
+      "grad_norm": 3.1858954429626465,
+      "learning_rate": 3.887900355871886e-05,
+      "loss": 0.8335,
+      "step": 9065
+    },
+    {
+      "epoch": 4.029333333333334,
+      "grad_norm": 3.0691826343536377,
+      "learning_rate": 3.886120996441281e-05,
+      "loss": 0.8928,
+      "step": 9066
+    },
+    {
+      "epoch": 4.0297777777777775,
+      "grad_norm": 2.3211822509765625,
+      "learning_rate": 3.884341637010676e-05,
+      "loss": 0.3684,
+      "step": 9067
+    },
+    {
+      "epoch": 4.030222222222222,
+      "grad_norm": 3.79524302482605,
+      "learning_rate": 3.882562277580071e-05,
+      "loss": 1.2104,
+      "step": 9068
+    },
+    {
+      "epoch": 4.030666666666667,
+      "grad_norm": 4.332894325256348,
+      "learning_rate": 3.880782918149466e-05,
+      "loss": 1.0843,
+      "step": 9069
+    },
+    {
+      "epoch": 4.0311111111111115,
+      "grad_norm": 4.316042423248291,
+      "learning_rate": 3.879003558718861e-05,
+      "loss": 0.5951,
+      "step": 9070
+    },
+    {
+      "epoch": 4.031555555555555,
+      "grad_norm": 3.6957225799560547,
+      "learning_rate": 3.877224199288256e-05,
+      "loss": 1.0805,
+      "step": 9071
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 4.04442834854126,
+      "learning_rate": 3.875444839857652e-05,
+      "loss": 1.1406,
+      "step": 9072
+    },
+    {
+      "epoch": 4.032444444444445,
+      "grad_norm": 3.6291229724884033,
+      "learning_rate": 3.873665480427047e-05,
+      "loss": 1.2136,
+      "step": 9073
+    },
+    {
+      "epoch": 4.0328888888888885,
+      "grad_norm": 3.7658095359802246,
+      "learning_rate": 3.871886120996441e-05,
+      "loss": 0.9391,
+      "step": 9074
+    },
+    {
+      "epoch": 4.033333333333333,
+      "grad_norm": 4.630577564239502,
+      "learning_rate": 3.870106761565837e-05,
+      "loss": 0.9134,
+      "step": 9075
+    },
+    {
+      "epoch": 4.033777777777778,
+      "grad_norm": 4.5560407638549805,
+      "learning_rate": 3.8683274021352317e-05,
+      "loss": 1.164,
+      "step": 9076
+    },
+    {
+      "epoch": 4.034222222222223,
+      "grad_norm": 3.639586925506592,
+      "learning_rate": 3.8665480427046266e-05,
+      "loss": 0.7082,
+      "step": 9077
+    },
+    {
+      "epoch": 4.034666666666666,
+      "grad_norm": 3.4507131576538086,
+      "learning_rate": 3.8647686832740216e-05,
+      "loss": 1.0112,
+      "step": 9078
+    },
+    {
+      "epoch": 4.035111111111111,
+      "grad_norm": 3.1812744140625,
+      "learning_rate": 3.8629893238434166e-05,
+      "loss": 0.5256,
+      "step": 9079
+    },
+    {
+      "epoch": 4.035555555555556,
+      "grad_norm": 3.704305410385132,
+      "learning_rate": 3.8612099644128115e-05,
+      "loss": 0.7616,
+      "step": 9080
+    },
+    {
+      "epoch": 4.036,
+      "grad_norm": 2.5341804027557373,
+      "learning_rate": 3.8594306049822065e-05,
+      "loss": 0.4988,
+      "step": 9081
+    },
+    {
+      "epoch": 4.036444444444444,
+      "grad_norm": 3.8586108684539795,
+      "learning_rate": 3.8576512455516015e-05,
+      "loss": 0.7027,
+      "step": 9082
+    },
+    {
+      "epoch": 4.036888888888889,
+      "grad_norm": 3.8743813037872314,
+      "learning_rate": 3.8558718861209965e-05,
+      "loss": 1.023,
+      "step": 9083
+    },
+    {
+      "epoch": 4.037333333333334,
+      "grad_norm": 3.394871711730957,
+      "learning_rate": 3.8540925266903914e-05,
+      "loss": 1.0268,
+      "step": 9084
+    },
+    {
+      "epoch": 4.0377777777777775,
+      "grad_norm": 4.589004993438721,
+      "learning_rate": 3.852313167259787e-05,
+      "loss": 1.0115,
+      "step": 9085
+    },
+    {
+      "epoch": 4.038222222222222,
+      "grad_norm": 3.1262948513031006,
+      "learning_rate": 3.850533807829182e-05,
+      "loss": 0.576,
+      "step": 9086
+    },
+    {
+      "epoch": 4.038666666666667,
+      "grad_norm": 4.138561725616455,
+      "learning_rate": 3.8487544483985763e-05,
+      "loss": 0.9221,
+      "step": 9087
+    },
+    {
+      "epoch": 4.0391111111111115,
+      "grad_norm": 3.105081796646118,
+      "learning_rate": 3.846975088967972e-05,
+      "loss": 0.4531,
+      "step": 9088
+    },
+    {
+      "epoch": 4.039555555555555,
+      "grad_norm": 4.928419589996338,
+      "learning_rate": 3.845195729537367e-05,
+      "loss": 0.906,
+      "step": 9089
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 4.220268249511719,
+      "learning_rate": 3.843416370106761e-05,
+      "loss": 0.8529,
+      "step": 9090
+    },
+    {
+      "epoch": 4.040444444444445,
+      "grad_norm": 5.022706031799316,
+      "learning_rate": 3.841637010676157e-05,
+      "loss": 0.9518,
+      "step": 9091
+    },
+    {
+      "epoch": 4.0408888888888885,
+      "grad_norm": 3.5744247436523438,
+      "learning_rate": 3.839857651245552e-05,
+      "loss": 0.8885,
+      "step": 9092
+    },
+    {
+      "epoch": 4.041333333333333,
+      "grad_norm": 3.773207902908325,
+      "learning_rate": 3.838078291814947e-05,
+      "loss": 0.7402,
+      "step": 9093
+    },
+    {
+      "epoch": 4.041777777777778,
+      "grad_norm": 3.8730592727661133,
+      "learning_rate": 3.836298932384342e-05,
+      "loss": 0.8627,
+      "step": 9094
+    },
+    {
+      "epoch": 4.042222222222223,
+      "grad_norm": 4.191495895385742,
+      "learning_rate": 3.834519572953737e-05,
+      "loss": 0.6384,
+      "step": 9095
+    },
+    {
+      "epoch": 4.042666666666666,
+      "grad_norm": 4.794586658477783,
+      "learning_rate": 3.832740213523132e-05,
+      "loss": 0.9203,
+      "step": 9096
+    },
+    {
+      "epoch": 4.043111111111111,
+      "grad_norm": 4.144529819488525,
+      "learning_rate": 3.830960854092527e-05,
+      "loss": 0.7881,
+      "step": 9097
+    },
+    {
+      "epoch": 4.043555555555556,
+      "grad_norm": 8.298238754272461,
+      "learning_rate": 3.829181494661922e-05,
+      "loss": 1.375,
+      "step": 9098
+    },
+    {
+      "epoch": 4.044,
+      "grad_norm": 3.5953052043914795,
+      "learning_rate": 3.827402135231317e-05,
+      "loss": 0.3941,
+      "step": 9099
+    },
+    {
+      "epoch": 4.044444444444444,
+      "grad_norm": 10.296087265014648,
+      "learning_rate": 3.825622775800712e-05,
+      "loss": 0.2957,
+      "step": 9100
+    },
+    {
+      "epoch": 4.044888888888889,
+      "grad_norm": 3.059195041656494,
+      "learning_rate": 3.823843416370107e-05,
+      "loss": 1.8857,
+      "step": 9101
+    },
+    {
+      "epoch": 4.045333333333334,
+      "grad_norm": 2.8062288761138916,
+      "learning_rate": 3.822064056939502e-05,
+      "loss": 1.7158,
+      "step": 9102
+    },
+    {
+      "epoch": 4.0457777777777775,
+      "grad_norm": 1.8304457664489746,
+      "learning_rate": 3.8202846975088966e-05,
+      "loss": 0.4758,
+      "step": 9103
+    },
+    {
+      "epoch": 4.046222222222222,
+      "grad_norm": 2.9134559631347656,
+      "learning_rate": 3.818505338078292e-05,
+      "loss": 1.2599,
+      "step": 9104
+    },
+    {
+      "epoch": 4.046666666666667,
+      "grad_norm": 2.96635365486145,
+      "learning_rate": 3.816725978647687e-05,
+      "loss": 1.1921,
+      "step": 9105
+    },
+    {
+      "epoch": 4.0471111111111115,
+      "grad_norm": 2.9679348468780518,
+      "learning_rate": 3.814946619217082e-05,
+      "loss": 1.0036,
+      "step": 9106
+    },
+    {
+      "epoch": 4.047555555555555,
+      "grad_norm": 3.5697379112243652,
+      "learning_rate": 3.813167259786477e-05,
+      "loss": 1.3876,
+      "step": 9107
+    },
+    {
+      "epoch": 4.048,
+      "grad_norm": 2.98809814453125,
+      "learning_rate": 3.811387900355872e-05,
+      "loss": 0.7762,
+      "step": 9108
+    },
+    {
+      "epoch": 4.048444444444445,
+      "grad_norm": 3.473585367202759,
+      "learning_rate": 3.809608540925267e-05,
+      "loss": 1.2346,
+      "step": 9109
+    },
+    {
+      "epoch": 4.0488888888888885,
+      "grad_norm": 3.487264394760132,
+      "learning_rate": 3.807829181494662e-05,
+      "loss": 1.0737,
+      "step": 9110
+    },
+    {
+      "epoch": 4.049333333333333,
+      "grad_norm": 3.552971839904785,
+      "learning_rate": 3.806049822064057e-05,
+      "loss": 1.1126,
+      "step": 9111
+    },
+    {
+      "epoch": 4.049777777777778,
+      "grad_norm": 4.056872844696045,
+      "learning_rate": 3.804270462633452e-05,
+      "loss": 1.2205,
+      "step": 9112
+    },
+    {
+      "epoch": 4.050222222222223,
+      "grad_norm": 3.7317447662353516,
+      "learning_rate": 3.802491103202847e-05,
+      "loss": 1.0883,
+      "step": 9113
+    },
+    {
+      "epoch": 4.050666666666666,
+      "grad_norm": 3.8739066123962402,
+      "learning_rate": 3.800711743772242e-05,
+      "loss": 0.8847,
+      "step": 9114
+    },
+    {
+      "epoch": 4.051111111111111,
+      "grad_norm": 3.574202060699463,
+      "learning_rate": 3.7989323843416376e-05,
+      "loss": 0.7981,
+      "step": 9115
+    },
+    {
+      "epoch": 4.051555555555556,
+      "grad_norm": 3.6287660598754883,
+      "learning_rate": 3.797153024911032e-05,
+      "loss": 1.0176,
+      "step": 9116
+    },
+    {
+      "epoch": 4.052,
+      "grad_norm": 3.372129201889038,
+      "learning_rate": 3.7953736654804275e-05,
+      "loss": 0.8463,
+      "step": 9117
+    },
+    {
+      "epoch": 4.052444444444444,
+      "grad_norm": 0.2473270297050476,
+      "learning_rate": 3.7935943060498225e-05,
+      "loss": 0.0263,
+      "step": 9118
+    },
+    {
+      "epoch": 4.052888888888889,
+      "grad_norm": 2.3387582302093506,
+      "learning_rate": 3.791814946619217e-05,
+      "loss": 0.4731,
+      "step": 9119
+    },
+    {
+      "epoch": 4.053333333333334,
+      "grad_norm": 3.4713640213012695,
+      "learning_rate": 3.7900355871886125e-05,
+      "loss": 1.0173,
+      "step": 9120
+    },
+    {
+      "epoch": 4.0537777777777775,
+      "grad_norm": 4.791953086853027,
+      "learning_rate": 3.7882562277580074e-05,
+      "loss": 1.3301,
+      "step": 9121
+    },
+    {
+      "epoch": 4.054222222222222,
+      "grad_norm": 3.886340618133545,
+      "learning_rate": 3.7864768683274024e-05,
+      "loss": 0.759,
+      "step": 9122
+    },
+    {
+      "epoch": 4.054666666666667,
+      "grad_norm": 3.3495802879333496,
+      "learning_rate": 3.7846975088967974e-05,
+      "loss": 0.8179,
+      "step": 9123
+    },
+    {
+      "epoch": 4.0551111111111116,
+      "grad_norm": 3.790015935897827,
+      "learning_rate": 3.7829181494661923e-05,
+      "loss": 1.1811,
+      "step": 9124
+    },
+    {
+      "epoch": 4.055555555555555,
+      "grad_norm": 4.013554096221924,
+      "learning_rate": 3.781138790035587e-05,
+      "loss": 0.7972,
+      "step": 9125
+    },
+    {
+      "epoch": 4.056,
+      "grad_norm": 4.106941223144531,
+      "learning_rate": 3.779359430604982e-05,
+      "loss": 1.1104,
+      "step": 9126
+    },
+    {
+      "epoch": 4.056444444444445,
+      "grad_norm": 4.0450334548950195,
+      "learning_rate": 3.777580071174377e-05,
+      "loss": 1.0491,
+      "step": 9127
+    },
+    {
+      "epoch": 4.0568888888888885,
+      "grad_norm": 4.481957912445068,
+      "learning_rate": 3.775800711743772e-05,
+      "loss": 1.1531,
+      "step": 9128
+    },
+    {
+      "epoch": 4.057333333333333,
+      "grad_norm": 3.5693199634552,
+      "learning_rate": 3.774021352313167e-05,
+      "loss": 0.8621,
+      "step": 9129
+    },
+    {
+      "epoch": 4.057777777777778,
+      "grad_norm": 2.899524688720703,
+      "learning_rate": 3.772241992882563e-05,
+      "loss": 0.5678,
+      "step": 9130
+    },
+    {
+      "epoch": 4.058222222222223,
+      "grad_norm": 4.416593551635742,
+      "learning_rate": 3.770462633451958e-05,
+      "loss": 1.0094,
+      "step": 9131
+    },
+    {
+      "epoch": 4.058666666666666,
+      "grad_norm": 4.252343654632568,
+      "learning_rate": 3.768683274021352e-05,
+      "loss": 0.7924,
+      "step": 9132
+    },
+    {
+      "epoch": 4.059111111111111,
+      "grad_norm": 4.362743377685547,
+      "learning_rate": 3.766903914590748e-05,
+      "loss": 0.5001,
+      "step": 9133
+    },
+    {
+      "epoch": 4.059555555555556,
+      "grad_norm": 4.057370662689209,
+      "learning_rate": 3.765124555160143e-05,
+      "loss": 0.8535,
+      "step": 9134
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 4.266189098358154,
+      "learning_rate": 3.763345195729537e-05,
+      "loss": 0.7122,
+      "step": 9135
+    },
+    {
+      "epoch": 4.060444444444444,
+      "grad_norm": 3.547168016433716,
+      "learning_rate": 3.761565836298933e-05,
+      "loss": 0.6885,
+      "step": 9136
+    },
+    {
+      "epoch": 4.060888888888889,
+      "grad_norm": 3.209439992904663,
+      "learning_rate": 3.759786476868328e-05,
+      "loss": 0.7069,
+      "step": 9137
+    },
+    {
+      "epoch": 4.061333333333334,
+      "grad_norm": 3.9220142364501953,
+      "learning_rate": 3.7580071174377226e-05,
+      "loss": 0.8905,
+      "step": 9138
+    },
+    {
+      "epoch": 4.0617777777777775,
+      "grad_norm": 3.8512017726898193,
+      "learning_rate": 3.7562277580071176e-05,
+      "loss": 0.9789,
+      "step": 9139
+    },
+    {
+      "epoch": 4.062222222222222,
+      "grad_norm": 4.788553714752197,
+      "learning_rate": 3.7544483985765126e-05,
+      "loss": 1.0029,
+      "step": 9140
+    },
+    {
+      "epoch": 4.062666666666667,
+      "grad_norm": 5.606046676635742,
+      "learning_rate": 3.7526690391459076e-05,
+      "loss": 0.781,
+      "step": 9141
+    },
+    {
+      "epoch": 4.063111111111111,
+      "grad_norm": 4.540804386138916,
+      "learning_rate": 3.7508896797153025e-05,
+      "loss": 1.0023,
+      "step": 9142
+    },
+    {
+      "epoch": 4.063555555555555,
+      "grad_norm": 4.766351699829102,
+      "learning_rate": 3.7491103202846975e-05,
+      "loss": 1.0042,
+      "step": 9143
+    },
+    {
+      "epoch": 4.064,
+      "grad_norm": 4.029290199279785,
+      "learning_rate": 3.747330960854093e-05,
+      "loss": 0.7966,
+      "step": 9144
+    },
+    {
+      "epoch": 4.064444444444445,
+      "grad_norm": 6.49293327331543,
+      "learning_rate": 3.7455516014234874e-05,
+      "loss": 0.6414,
+      "step": 9145
+    },
+    {
+      "epoch": 4.0648888888888886,
+      "grad_norm": 4.462486267089844,
+      "learning_rate": 3.743772241992883e-05,
+      "loss": 0.9538,
+      "step": 9146
+    },
+    {
+      "epoch": 4.065333333333333,
+      "grad_norm": 3.4527547359466553,
+      "learning_rate": 3.741992882562278e-05,
+      "loss": 0.4922,
+      "step": 9147
+    },
+    {
+      "epoch": 4.065777777777778,
+      "grad_norm": 3.8521132469177246,
+      "learning_rate": 3.7402135231316724e-05,
+      "loss": 0.7803,
+      "step": 9148
+    },
+    {
+      "epoch": 4.066222222222223,
+      "grad_norm": 3.7171268463134766,
+      "learning_rate": 3.738434163701068e-05,
+      "loss": 0.6302,
+      "step": 9149
+    },
+    {
+      "epoch": 4.066666666666666,
+      "grad_norm": 2.2403533458709717,
+      "learning_rate": 3.736654804270463e-05,
+      "loss": 0.2649,
+      "step": 9150
+    },
+    {
+      "epoch": 4.067111111111111,
+      "grad_norm": 2.9199023246765137,
+      "learning_rate": 3.734875444839858e-05,
+      "loss": 1.3195,
+      "step": 9151
+    },
+    {
+      "epoch": 4.067555555555556,
+      "grad_norm": 3.4877567291259766,
+      "learning_rate": 3.733096085409253e-05,
+      "loss": 1.3656,
+      "step": 9152
+    },
+    {
+      "epoch": 4.068,
+      "grad_norm": 2.9192590713500977,
+      "learning_rate": 3.731316725978648e-05,
+      "loss": 1.1056,
+      "step": 9153
+    },
+    {
+      "epoch": 4.068444444444444,
+      "grad_norm": 4.004071235656738,
+      "learning_rate": 3.729537366548043e-05,
+      "loss": 1.1344,
+      "step": 9154
+    },
+    {
+      "epoch": 4.068888888888889,
+      "grad_norm": 3.422654628753662,
+      "learning_rate": 3.727758007117438e-05,
+      "loss": 0.9968,
+      "step": 9155
+    },
+    {
+      "epoch": 4.069333333333334,
+      "grad_norm": 3.3041553497314453,
+      "learning_rate": 3.725978647686833e-05,
+      "loss": 1.4082,
+      "step": 9156
+    },
+    {
+      "epoch": 4.0697777777777775,
+      "grad_norm": 3.1231582164764404,
+      "learning_rate": 3.7241992882562285e-05,
+      "loss": 1.2489,
+      "step": 9157
+    },
+    {
+      "epoch": 4.070222222222222,
+      "grad_norm": 2.934847593307495,
+      "learning_rate": 3.722419928825623e-05,
+      "loss": 1.0434,
+      "step": 9158
+    },
+    {
+      "epoch": 4.070666666666667,
+      "grad_norm": 4.84507942199707,
+      "learning_rate": 3.720640569395018e-05,
+      "loss": 1.56,
+      "step": 9159
+    },
+    {
+      "epoch": 4.071111111111111,
+      "grad_norm": 4.400614261627197,
+      "learning_rate": 3.7188612099644134e-05,
+      "loss": 1.1331,
+      "step": 9160
+    },
+    {
+      "epoch": 4.071555555555555,
+      "grad_norm": 4.824821949005127,
+      "learning_rate": 3.717081850533808e-05,
+      "loss": 0.8231,
+      "step": 9161
+    },
+    {
+      "epoch": 4.072,
+      "grad_norm": 3.7322332859039307,
+      "learning_rate": 3.715302491103203e-05,
+      "loss": 1.0261,
+      "step": 9162
+    },
+    {
+      "epoch": 4.072444444444445,
+      "grad_norm": 3.8609957695007324,
+      "learning_rate": 3.713523131672598e-05,
+      "loss": 0.9616,
+      "step": 9163
+    },
+    {
+      "epoch": 4.072888888888889,
+      "grad_norm": 5.574636936187744,
+      "learning_rate": 3.7117437722419926e-05,
+      "loss": 1.2415,
+      "step": 9164
+    },
+    {
+      "epoch": 4.073333333333333,
+      "grad_norm": 4.391664028167725,
+      "learning_rate": 3.709964412811388e-05,
+      "loss": 1.3483,
+      "step": 9165
+    },
+    {
+      "epoch": 4.073777777777778,
+      "grad_norm": 3.788674831390381,
+      "learning_rate": 3.708185053380783e-05,
+      "loss": 1.0001,
+      "step": 9166
+    },
+    {
+      "epoch": 4.074222222222223,
+      "grad_norm": 4.3154072761535645,
+      "learning_rate": 3.706405693950178e-05,
+      "loss": 1.3682,
+      "step": 9167
+    },
+    {
+      "epoch": 4.074666666666666,
+      "grad_norm": 3.440797805786133,
+      "learning_rate": 3.704626334519573e-05,
+      "loss": 0.8641,
+      "step": 9168
+    },
+    {
+      "epoch": 4.075111111111111,
+      "grad_norm": 4.154467582702637,
+      "learning_rate": 3.702846975088968e-05,
+      "loss": 0.9788,
+      "step": 9169
+    },
+    {
+      "epoch": 4.075555555555556,
+      "grad_norm": 3.506791591644287,
+      "learning_rate": 3.701067615658363e-05,
+      "loss": 0.9756,
+      "step": 9170
+    },
+    {
+      "epoch": 4.076,
+      "grad_norm": 4.369688987731934,
+      "learning_rate": 3.699288256227758e-05,
+      "loss": 1.4468,
+      "step": 9171
+    },
+    {
+      "epoch": 4.076444444444444,
+      "grad_norm": 4.314336776733398,
+      "learning_rate": 3.697508896797153e-05,
+      "loss": 0.9896,
+      "step": 9172
+    },
+    {
+      "epoch": 4.076888888888889,
+      "grad_norm": 3.7983624935150146,
+      "learning_rate": 3.695729537366549e-05,
+      "loss": 1.0006,
+      "step": 9173
+    },
+    {
+      "epoch": 4.077333333333334,
+      "grad_norm": 3.259927749633789,
+      "learning_rate": 3.693950177935943e-05,
+      "loss": 0.7188,
+      "step": 9174
+    },
+    {
+      "epoch": 4.0777777777777775,
+      "grad_norm": 3.296593189239502,
+      "learning_rate": 3.6921708185053386e-05,
+      "loss": 0.5057,
+      "step": 9175
+    },
+    {
+      "epoch": 4.078222222222222,
+      "grad_norm": 3.704531669616699,
+      "learning_rate": 3.6903914590747336e-05,
+      "loss": 0.7359,
+      "step": 9176
+    },
+    {
+      "epoch": 4.078666666666667,
+      "grad_norm": 3.4545626640319824,
+      "learning_rate": 3.688612099644128e-05,
+      "loss": 0.784,
+      "step": 9177
+    },
+    {
+      "epoch": 4.079111111111111,
+      "grad_norm": 3.940635919570923,
+      "learning_rate": 3.6868327402135236e-05,
+      "loss": 0.9062,
+      "step": 9178
+    },
+    {
+      "epoch": 4.079555555555555,
+      "grad_norm": 3.7405149936676025,
+      "learning_rate": 3.6850533807829185e-05,
+      "loss": 1.1758,
+      "step": 9179
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 3.7189319133758545,
+      "learning_rate": 3.683274021352313e-05,
+      "loss": 1.0273,
+      "step": 9180
+    },
+    {
+      "epoch": 4.080444444444445,
+      "grad_norm": 3.3661651611328125,
+      "learning_rate": 3.6814946619217085e-05,
+      "loss": 0.5712,
+      "step": 9181
+    },
+    {
+      "epoch": 4.080888888888889,
+      "grad_norm": 4.403657913208008,
+      "learning_rate": 3.6797153024911034e-05,
+      "loss": 1.1332,
+      "step": 9182
+    },
+    {
+      "epoch": 4.081333333333333,
+      "grad_norm": 3.4443655014038086,
+      "learning_rate": 3.6779359430604984e-05,
+      "loss": 1.0131,
+      "step": 9183
+    },
+    {
+      "epoch": 4.081777777777778,
+      "grad_norm": 6.885471343994141,
+      "learning_rate": 3.6761565836298934e-05,
+      "loss": 0.9145,
+      "step": 9184
+    },
+    {
+      "epoch": 4.082222222222223,
+      "grad_norm": 4.67835807800293,
+      "learning_rate": 3.6743772241992884e-05,
+      "loss": 0.8626,
+      "step": 9185
+    },
+    {
+      "epoch": 4.082666666666666,
+      "grad_norm": 4.347279071807861,
+      "learning_rate": 3.672597864768684e-05,
+      "loss": 0.753,
+      "step": 9186
+    },
+    {
+      "epoch": 4.083111111111111,
+      "grad_norm": 4.374654293060303,
+      "learning_rate": 3.670818505338078e-05,
+      "loss": 0.8974,
+      "step": 9187
+    },
+    {
+      "epoch": 4.083555555555556,
+      "grad_norm": 3.8928260803222656,
+      "learning_rate": 3.669039145907473e-05,
+      "loss": 0.8108,
+      "step": 9188
+    },
+    {
+      "epoch": 4.084,
+      "grad_norm": 5.292436122894287,
+      "learning_rate": 3.667259786476869e-05,
+      "loss": 0.9941,
+      "step": 9189
+    },
+    {
+      "epoch": 4.084444444444444,
+      "grad_norm": 4.306451320648193,
+      "learning_rate": 3.665480427046263e-05,
+      "loss": 1.0679,
+      "step": 9190
+    },
+    {
+      "epoch": 4.084888888888889,
+      "grad_norm": 4.150672435760498,
+      "learning_rate": 3.663701067615659e-05,
+      "loss": 0.7973,
+      "step": 9191
+    },
+    {
+      "epoch": 4.085333333333334,
+      "grad_norm": 3.7112274169921875,
+      "learning_rate": 3.661921708185054e-05,
+      "loss": 0.7108,
+      "step": 9192
+    },
+    {
+      "epoch": 4.0857777777777775,
+      "grad_norm": 4.035175323486328,
+      "learning_rate": 3.660142348754448e-05,
+      "loss": 0.8418,
+      "step": 9193
+    },
+    {
+      "epoch": 4.086222222222222,
+      "grad_norm": 4.420337677001953,
+      "learning_rate": 3.658362989323844e-05,
+      "loss": 0.8937,
+      "step": 9194
+    },
+    {
+      "epoch": 4.086666666666667,
+      "grad_norm": 4.49367618560791,
+      "learning_rate": 3.656583629893239e-05,
+      "loss": 1.1035,
+      "step": 9195
+    },
+    {
+      "epoch": 4.087111111111111,
+      "grad_norm": 5.570310115814209,
+      "learning_rate": 3.654804270462634e-05,
+      "loss": 0.7943,
+      "step": 9196
+    },
+    {
+      "epoch": 4.087555555555555,
+      "grad_norm": 7.375543117523193,
+      "learning_rate": 3.653024911032029e-05,
+      "loss": 1.0006,
+      "step": 9197
+    },
+    {
+      "epoch": 4.088,
+      "grad_norm": 3.2930474281311035,
+      "learning_rate": 3.651245551601424e-05,
+      "loss": 0.3692,
+      "step": 9198
+    },
+    {
+      "epoch": 4.088444444444445,
+      "grad_norm": 2.375638961791992,
+      "learning_rate": 3.6494661921708186e-05,
+      "loss": 0.151,
+      "step": 9199
+    },
+    {
+      "epoch": 4.088888888888889,
+      "grad_norm": 3.8127799034118652,
+      "learning_rate": 3.6476868327402136e-05,
+      "loss": 0.318,
+      "step": 9200
+    },
+    {
+      "epoch": 4.089333333333333,
+      "grad_norm": 0.2315405160188675,
+      "learning_rate": 3.6459074733096086e-05,
+      "loss": 0.0145,
+      "step": 9201
+    },
+    {
+      "epoch": 4.089777777777778,
+      "grad_norm": 2.930769920349121,
+      "learning_rate": 3.644128113879004e-05,
+      "loss": 1.5272,
+      "step": 9202
+    },
+    {
+      "epoch": 4.090222222222223,
+      "grad_norm": 2.6000430583953857,
+      "learning_rate": 3.6423487544483985e-05,
+      "loss": 1.2693,
+      "step": 9203
+    },
+    {
+      "epoch": 4.0906666666666665,
+      "grad_norm": 3.497532367706299,
+      "learning_rate": 3.6405693950177935e-05,
+      "loss": 1.5588,
+      "step": 9204
+    },
+    {
+      "epoch": 4.091111111111111,
+      "grad_norm": 3.512805223464966,
+      "learning_rate": 3.638790035587189e-05,
+      "loss": 1.3456,
+      "step": 9205
+    },
+    {
+      "epoch": 4.091555555555556,
+      "grad_norm": 4.103816032409668,
+      "learning_rate": 3.6370106761565835e-05,
+      "loss": 0.9345,
+      "step": 9206
+    },
+    {
+      "epoch": 4.092,
+      "grad_norm": 3.1070504188537598,
+      "learning_rate": 3.635231316725979e-05,
+      "loss": 1.1946,
+      "step": 9207
+    },
+    {
+      "epoch": 4.092444444444444,
+      "grad_norm": 3.410989999771118,
+      "learning_rate": 3.633451957295374e-05,
+      "loss": 1.0064,
+      "step": 9208
+    },
+    {
+      "epoch": 4.092888888888889,
+      "grad_norm": 2.981509208679199,
+      "learning_rate": 3.6316725978647684e-05,
+      "loss": 0.8163,
+      "step": 9209
+    },
+    {
+      "epoch": 4.093333333333334,
+      "grad_norm": 2.2597262859344482,
+      "learning_rate": 3.629893238434164e-05,
+      "loss": 0.5533,
+      "step": 9210
+    },
+    {
+      "epoch": 4.0937777777777775,
+      "grad_norm": 2.4587838649749756,
+      "learning_rate": 3.628113879003559e-05,
+      "loss": 0.6522,
+      "step": 9211
+    },
+    {
+      "epoch": 4.094222222222222,
+      "grad_norm": 3.2743875980377197,
+      "learning_rate": 3.626334519572954e-05,
+      "loss": 0.9227,
+      "step": 9212
+    },
+    {
+      "epoch": 4.094666666666667,
+      "grad_norm": 3.8894238471984863,
+      "learning_rate": 3.624555160142349e-05,
+      "loss": 1.2398,
+      "step": 9213
+    },
+    {
+      "epoch": 4.095111111111111,
+      "grad_norm": 4.327610969543457,
+      "learning_rate": 3.622775800711744e-05,
+      "loss": 1.296,
+      "step": 9214
+    },
+    {
+      "epoch": 4.095555555555555,
+      "grad_norm": 3.491788864135742,
+      "learning_rate": 3.6209964412811396e-05,
+      "loss": 0.9961,
+      "step": 9215
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 4.091619968414307,
+      "learning_rate": 3.619217081850534e-05,
+      "loss": 1.0178,
+      "step": 9216
+    },
+    {
+      "epoch": 4.096444444444445,
+      "grad_norm": 5.12504243850708,
+      "learning_rate": 3.617437722419929e-05,
+      "loss": 0.9988,
+      "step": 9217
+    },
+    {
+      "epoch": 4.096888888888889,
+      "grad_norm": 5.116464614868164,
+      "learning_rate": 3.6156583629893245e-05,
+      "loss": 1.3894,
+      "step": 9218
+    },
+    {
+      "epoch": 4.097333333333333,
+      "grad_norm": 3.9038383960723877,
+      "learning_rate": 3.613879003558719e-05,
+      "loss": 1.2835,
+      "step": 9219
+    },
+    {
+      "epoch": 4.097777777777778,
+      "grad_norm": 4.344372272491455,
+      "learning_rate": 3.6120996441281144e-05,
+      "loss": 1.0891,
+      "step": 9220
+    },
+    {
+      "epoch": 4.098222222222223,
+      "grad_norm": 4.03400993347168,
+      "learning_rate": 3.6103202846975094e-05,
+      "loss": 1.1626,
+      "step": 9221
+    },
+    {
+      "epoch": 4.0986666666666665,
+      "grad_norm": 4.054080486297607,
+      "learning_rate": 3.608540925266904e-05,
+      "loss": 1.0427,
+      "step": 9222
+    },
+    {
+      "epoch": 4.099111111111111,
+      "grad_norm": 4.079352855682373,
+      "learning_rate": 3.606761565836299e-05,
+      "loss": 1.0579,
+      "step": 9223
+    },
+    {
+      "epoch": 4.099555555555556,
+      "grad_norm": 3.899838924407959,
+      "learning_rate": 3.604982206405694e-05,
+      "loss": 0.7473,
+      "step": 9224
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 3.900310754776001,
+      "learning_rate": 3.6032028469750886e-05,
+      "loss": 1.0974,
+      "step": 9225
+    },
+    {
+      "epoch": 4.100444444444444,
+      "grad_norm": 3.1842355728149414,
+      "learning_rate": 3.601423487544484e-05,
+      "loss": 0.7394,
+      "step": 9226
+    },
+    {
+      "epoch": 4.100888888888889,
+      "grad_norm": 3.721182346343994,
+      "learning_rate": 3.599644128113879e-05,
+      "loss": 0.7716,
+      "step": 9227
+    },
+    {
+      "epoch": 4.101333333333334,
+      "grad_norm": 5.3094353675842285,
+      "learning_rate": 3.597864768683274e-05,
+      "loss": 0.9151,
+      "step": 9228
+    },
+    {
+      "epoch": 4.1017777777777775,
+      "grad_norm": 3.914445638656616,
+      "learning_rate": 3.596085409252669e-05,
+      "loss": 1.0772,
+      "step": 9229
+    },
+    {
+      "epoch": 4.102222222222222,
+      "grad_norm": 4.377343654632568,
+      "learning_rate": 3.594306049822064e-05,
+      "loss": 0.9153,
+      "step": 9230
+    },
+    {
+      "epoch": 4.102666666666667,
+      "grad_norm": 3.879788637161255,
+      "learning_rate": 3.59252669039146e-05,
+      "loss": 0.6954,
+      "step": 9231
+    },
+    {
+      "epoch": 4.103111111111111,
+      "grad_norm": 4.046523571014404,
+      "learning_rate": 3.590747330960854e-05,
+      "loss": 0.8208,
+      "step": 9232
+    },
+    {
+      "epoch": 4.103555555555555,
+      "grad_norm": 4.044562339782715,
+      "learning_rate": 3.588967971530249e-05,
+      "loss": 0.9976,
+      "step": 9233
+    },
+    {
+      "epoch": 4.104,
+      "grad_norm": 3.7858481407165527,
+      "learning_rate": 3.587188612099645e-05,
+      "loss": 0.9062,
+      "step": 9234
+    },
+    {
+      "epoch": 4.104444444444445,
+      "grad_norm": 2.641073226928711,
+      "learning_rate": 3.585409252669039e-05,
+      "loss": 0.4488,
+      "step": 9235
+    },
+    {
+      "epoch": 4.104888888888889,
+      "grad_norm": 3.873842239379883,
+      "learning_rate": 3.5836298932384346e-05,
+      "loss": 0.8443,
+      "step": 9236
+    },
+    {
+      "epoch": 4.105333333333333,
+      "grad_norm": 3.6468522548675537,
+      "learning_rate": 3.5818505338078296e-05,
+      "loss": 0.8935,
+      "step": 9237
+    },
+    {
+      "epoch": 4.105777777777778,
+      "grad_norm": 4.058321952819824,
+      "learning_rate": 3.580071174377224e-05,
+      "loss": 0.7038,
+      "step": 9238
+    },
+    {
+      "epoch": 4.106222222222222,
+      "grad_norm": 4.622478008270264,
+      "learning_rate": 3.5782918149466196e-05,
+      "loss": 1.1684,
+      "step": 9239
+    },
+    {
+      "epoch": 4.1066666666666665,
+      "grad_norm": 3.3128762245178223,
+      "learning_rate": 3.5765124555160145e-05,
+      "loss": 0.6763,
+      "step": 9240
+    },
+    {
+      "epoch": 4.107111111111111,
+      "grad_norm": 3.4522128105163574,
+      "learning_rate": 3.5747330960854095e-05,
+      "loss": 0.7716,
+      "step": 9241
+    },
+    {
+      "epoch": 4.107555555555556,
+      "grad_norm": 4.44683837890625,
+      "learning_rate": 3.5729537366548045e-05,
+      "loss": 1.36,
+      "step": 9242
+    },
+    {
+      "epoch": 4.108,
+      "grad_norm": 4.440934181213379,
+      "learning_rate": 3.5711743772241995e-05,
+      "loss": 0.9069,
+      "step": 9243
+    },
+    {
+      "epoch": 4.108444444444444,
+      "grad_norm": 3.85774302482605,
+      "learning_rate": 3.5693950177935944e-05,
+      "loss": 0.7827,
+      "step": 9244
+    },
+    {
+      "epoch": 4.108888888888889,
+      "grad_norm": 5.362185955047607,
+      "learning_rate": 3.5676156583629894e-05,
+      "loss": 1.2746,
+      "step": 9245
+    },
+    {
+      "epoch": 4.109333333333334,
+      "grad_norm": 3.934819221496582,
+      "learning_rate": 3.5658362989323844e-05,
+      "loss": 0.7316,
+      "step": 9246
+    },
+    {
+      "epoch": 4.1097777777777775,
+      "grad_norm": 4.532813549041748,
+      "learning_rate": 3.56405693950178e-05,
+      "loss": 0.6302,
+      "step": 9247
+    },
+    {
+      "epoch": 4.110222222222222,
+      "grad_norm": 4.29311990737915,
+      "learning_rate": 3.562277580071174e-05,
+      "loss": 0.5404,
+      "step": 9248
+    },
+    {
+      "epoch": 4.110666666666667,
+      "grad_norm": 3.867619276046753,
+      "learning_rate": 3.560498220640569e-05,
+      "loss": 0.8902,
+      "step": 9249
+    },
+    {
+      "epoch": 4.111111111111111,
+      "grad_norm": 5.4382805824279785,
+      "learning_rate": 3.558718861209965e-05,
+      "loss": 0.8051,
+      "step": 9250
+    },
+    {
+      "epoch": 4.111555555555555,
+      "grad_norm": 1.7659415006637573,
+      "learning_rate": 3.556939501779359e-05,
+      "loss": 0.8157,
+      "step": 9251
+    },
+    {
+      "epoch": 4.112,
+      "grad_norm": 2.5079727172851562,
+      "learning_rate": 3.555160142348755e-05,
+      "loss": 1.3793,
+      "step": 9252
+    },
+    {
+      "epoch": 4.112444444444445,
+      "grad_norm": 3.136521100997925,
+      "learning_rate": 3.55338078291815e-05,
+      "loss": 1.223,
+      "step": 9253
+    },
+    {
+      "epoch": 4.112888888888889,
+      "grad_norm": 3.1797564029693604,
+      "learning_rate": 3.551601423487544e-05,
+      "loss": 1.2796,
+      "step": 9254
+    },
+    {
+      "epoch": 4.113333333333333,
+      "grad_norm": 3.613799810409546,
+      "learning_rate": 3.54982206405694e-05,
+      "loss": 1.1687,
+      "step": 9255
+    },
+    {
+      "epoch": 4.113777777777778,
+      "grad_norm": 3.8610999584198,
+      "learning_rate": 3.548042704626335e-05,
+      "loss": 0.957,
+      "step": 9256
+    },
+    {
+      "epoch": 4.114222222222223,
+      "grad_norm": 3.316948890686035,
+      "learning_rate": 3.54626334519573e-05,
+      "loss": 1.4552,
+      "step": 9257
+    },
+    {
+      "epoch": 4.1146666666666665,
+      "grad_norm": 3.1068124771118164,
+      "learning_rate": 3.544483985765125e-05,
+      "loss": 0.9491,
+      "step": 9258
+    },
+    {
+      "epoch": 4.115111111111111,
+      "grad_norm": 3.8049232959747314,
+      "learning_rate": 3.54270462633452e-05,
+      "loss": 1.0955,
+      "step": 9259
+    },
+    {
+      "epoch": 4.115555555555556,
+      "grad_norm": 3.263183355331421,
+      "learning_rate": 3.540925266903915e-05,
+      "loss": 0.8825,
+      "step": 9260
+    },
+    {
+      "epoch": 4.116,
+      "grad_norm": 3.227997303009033,
+      "learning_rate": 3.5391459074733096e-05,
+      "loss": 1.1426,
+      "step": 9261
+    },
+    {
+      "epoch": 4.116444444444444,
+      "grad_norm": 3.8824005126953125,
+      "learning_rate": 3.5373665480427046e-05,
+      "loss": 1.0386,
+      "step": 9262
+    },
+    {
+      "epoch": 4.116888888888889,
+      "grad_norm": 4.455204486846924,
+      "learning_rate": 3.5355871886121e-05,
+      "loss": 1.0883,
+      "step": 9263
+    },
+    {
+      "epoch": 4.117333333333334,
+      "grad_norm": 3.621575355529785,
+      "learning_rate": 3.5338078291814945e-05,
+      "loss": 1.4027,
+      "step": 9264
+    },
+    {
+      "epoch": 4.1177777777777775,
+      "grad_norm": 3.8370847702026367,
+      "learning_rate": 3.53202846975089e-05,
+      "loss": 1.212,
+      "step": 9265
+    },
+    {
+      "epoch": 4.118222222222222,
+      "grad_norm": 3.060487747192383,
+      "learning_rate": 3.530249110320285e-05,
+      "loss": 0.8656,
+      "step": 9266
+    },
+    {
+      "epoch": 4.118666666666667,
+      "grad_norm": 3.36088228225708,
+      "learning_rate": 3.5284697508896795e-05,
+      "loss": 0.7663,
+      "step": 9267
+    },
+    {
+      "epoch": 4.119111111111111,
+      "grad_norm": 3.684251308441162,
+      "learning_rate": 3.526690391459075e-05,
+      "loss": 0.9489,
+      "step": 9268
+    },
+    {
+      "epoch": 4.119555555555555,
+      "grad_norm": 4.51201868057251,
+      "learning_rate": 3.52491103202847e-05,
+      "loss": 1.1845,
+      "step": 9269
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 4.410719394683838,
+      "learning_rate": 3.5231316725978644e-05,
+      "loss": 1.702,
+      "step": 9270
+    },
+    {
+      "epoch": 4.120444444444445,
+      "grad_norm": 3.741725444793701,
+      "learning_rate": 3.52135231316726e-05,
+      "loss": 0.7647,
+      "step": 9271
+    },
+    {
+      "epoch": 4.120888888888889,
+      "grad_norm": 4.066876411437988,
+      "learning_rate": 3.519572953736655e-05,
+      "loss": 1.046,
+      "step": 9272
+    },
+    {
+      "epoch": 4.121333333333333,
+      "grad_norm": 3.6181998252868652,
+      "learning_rate": 3.51779359430605e-05,
+      "loss": 1.0439,
+      "step": 9273
+    },
+    {
+      "epoch": 4.121777777777778,
+      "grad_norm": 4.158766746520996,
+      "learning_rate": 3.516014234875445e-05,
+      "loss": 0.8574,
+      "step": 9274
+    },
+    {
+      "epoch": 4.122222222222222,
+      "grad_norm": 3.483020067214966,
+      "learning_rate": 3.51423487544484e-05,
+      "loss": 0.9312,
+      "step": 9275
+    },
+    {
+      "epoch": 4.1226666666666665,
+      "grad_norm": 3.8150248527526855,
+      "learning_rate": 3.5124555160142356e-05,
+      "loss": 0.7764,
+      "step": 9276
+    },
+    {
+      "epoch": 4.123111111111111,
+      "grad_norm": 4.526205062866211,
+      "learning_rate": 3.51067615658363e-05,
+      "loss": 0.9631,
+      "step": 9277
+    },
+    {
+      "epoch": 4.123555555555556,
+      "grad_norm": 3.8249216079711914,
+      "learning_rate": 3.508896797153025e-05,
+      "loss": 0.7817,
+      "step": 9278
+    },
+    {
+      "epoch": 4.124,
+      "grad_norm": 3.743093729019165,
+      "learning_rate": 3.5071174377224205e-05,
+      "loss": 0.9439,
+      "step": 9279
+    },
+    {
+      "epoch": 4.124444444444444,
+      "grad_norm": 4.130136966705322,
+      "learning_rate": 3.505338078291815e-05,
+      "loss": 0.8861,
+      "step": 9280
+    },
+    {
+      "epoch": 4.124888888888889,
+      "grad_norm": 4.320974826812744,
+      "learning_rate": 3.5035587188612104e-05,
+      "loss": 0.9908,
+      "step": 9281
+    },
+    {
+      "epoch": 4.125333333333334,
+      "grad_norm": 4.112794876098633,
+      "learning_rate": 3.5017793594306054e-05,
+      "loss": 0.8566,
+      "step": 9282
+    },
+    {
+      "epoch": 4.1257777777777775,
+      "grad_norm": 4.176024913787842,
+      "learning_rate": 3.5e-05,
+      "loss": 0.9719,
+      "step": 9283
+    },
+    {
+      "epoch": 4.126222222222222,
+      "grad_norm": 3.905268669128418,
+      "learning_rate": 3.4982206405693953e-05,
+      "loss": 0.8443,
+      "step": 9284
+    },
+    {
+      "epoch": 4.126666666666667,
+      "grad_norm": 4.007937431335449,
+      "learning_rate": 3.49644128113879e-05,
+      "loss": 0.6952,
+      "step": 9285
+    },
+    {
+      "epoch": 4.127111111111111,
+      "grad_norm": 4.117457866668701,
+      "learning_rate": 3.494661921708185e-05,
+      "loss": 1.1521,
+      "step": 9286
+    },
+    {
+      "epoch": 4.127555555555555,
+      "grad_norm": 4.604625225067139,
+      "learning_rate": 3.49288256227758e-05,
+      "loss": 0.7057,
+      "step": 9287
+    },
+    {
+      "epoch": 4.128,
+      "grad_norm": 4.086756706237793,
+      "learning_rate": 3.491103202846975e-05,
+      "loss": 0.7674,
+      "step": 9288
+    },
+    {
+      "epoch": 4.128444444444445,
+      "grad_norm": 3.6481192111968994,
+      "learning_rate": 3.48932384341637e-05,
+      "loss": 0.5105,
+      "step": 9289
+    },
+    {
+      "epoch": 4.128888888888889,
+      "grad_norm": 3.9914069175720215,
+      "learning_rate": 3.487544483985765e-05,
+      "loss": 0.8233,
+      "step": 9290
+    },
+    {
+      "epoch": 4.129333333333333,
+      "grad_norm": 4.540968418121338,
+      "learning_rate": 3.48576512455516e-05,
+      "loss": 1.2001,
+      "step": 9291
+    },
+    {
+      "epoch": 4.129777777777778,
+      "grad_norm": 4.1771769523620605,
+      "learning_rate": 3.483985765124556e-05,
+      "loss": 0.759,
+      "step": 9292
+    },
+    {
+      "epoch": 4.130222222222222,
+      "grad_norm": 4.386975288391113,
+      "learning_rate": 3.48220640569395e-05,
+      "loss": 0.8304,
+      "step": 9293
+    },
+    {
+      "epoch": 4.1306666666666665,
+      "grad_norm": 3.9730632305145264,
+      "learning_rate": 3.480427046263345e-05,
+      "loss": 0.7385,
+      "step": 9294
+    },
+    {
+      "epoch": 4.131111111111111,
+      "grad_norm": 3.7220194339752197,
+      "learning_rate": 3.478647686832741e-05,
+      "loss": 0.8544,
+      "step": 9295
+    },
+    {
+      "epoch": 4.131555555555556,
+      "grad_norm": 3.411055564880371,
+      "learning_rate": 3.476868327402135e-05,
+      "loss": 0.7001,
+      "step": 9296
+    },
+    {
+      "epoch": 4.132,
+      "grad_norm": 4.14982795715332,
+      "learning_rate": 3.4750889679715307e-05,
+      "loss": 0.9112,
+      "step": 9297
+    },
+    {
+      "epoch": 4.132444444444444,
+      "grad_norm": 4.247723579406738,
+      "learning_rate": 3.4733096085409256e-05,
+      "loss": 0.8633,
+      "step": 9298
+    },
+    {
+      "epoch": 4.132888888888889,
+      "grad_norm": 5.790535926818848,
+      "learning_rate": 3.47153024911032e-05,
+      "loss": 0.6881,
+      "step": 9299
+    },
+    {
+      "epoch": 4.133333333333334,
+      "grad_norm": 3.3628830909729004,
+      "learning_rate": 3.4697508896797156e-05,
+      "loss": 0.5777,
+      "step": 9300
+    },
+    {
+      "epoch": 4.1337777777777776,
+      "grad_norm": 2.9413001537323,
+      "learning_rate": 3.4679715302491105e-05,
+      "loss": 1.5624,
+      "step": 9301
+    },
+    {
+      "epoch": 4.134222222222222,
+      "grad_norm": 2.9262659549713135,
+      "learning_rate": 3.4661921708185055e-05,
+      "loss": 1.451,
+      "step": 9302
+    },
+    {
+      "epoch": 4.134666666666667,
+      "grad_norm": 2.8405566215515137,
+      "learning_rate": 3.4644128113879005e-05,
+      "loss": 1.6277,
+      "step": 9303
+    },
+    {
+      "epoch": 4.135111111111111,
+      "grad_norm": 1.9037202596664429,
+      "learning_rate": 3.4626334519572955e-05,
+      "loss": 0.3974,
+      "step": 9304
+    },
+    {
+      "epoch": 4.135555555555555,
+      "grad_norm": 3.3218352794647217,
+      "learning_rate": 3.460854092526691e-05,
+      "loss": 1.4065,
+      "step": 9305
+    },
+    {
+      "epoch": 4.136,
+      "grad_norm": 3.091264009475708,
+      "learning_rate": 3.4590747330960854e-05,
+      "loss": 1.1051,
+      "step": 9306
+    },
+    {
+      "epoch": 4.136444444444445,
+      "grad_norm": 3.341606616973877,
+      "learning_rate": 3.4572953736654804e-05,
+      "loss": 1.2214,
+      "step": 9307
+    },
+    {
+      "epoch": 4.136888888888889,
+      "grad_norm": 3.4468531608581543,
+      "learning_rate": 3.455516014234876e-05,
+      "loss": 1.1427,
+      "step": 9308
+    },
+    {
+      "epoch": 4.137333333333333,
+      "grad_norm": 3.611560583114624,
+      "learning_rate": 3.45373665480427e-05,
+      "loss": 1.0753,
+      "step": 9309
+    },
+    {
+      "epoch": 4.137777777777778,
+      "grad_norm": 4.02140474319458,
+      "learning_rate": 3.451957295373665e-05,
+      "loss": 1.0676,
+      "step": 9310
+    },
+    {
+      "epoch": 4.138222222222222,
+      "grad_norm": 3.795001268386841,
+      "learning_rate": 3.450177935943061e-05,
+      "loss": 1.0637,
+      "step": 9311
+    },
+    {
+      "epoch": 4.1386666666666665,
+      "grad_norm": 3.871553659439087,
+      "learning_rate": 3.448398576512455e-05,
+      "loss": 1.1937,
+      "step": 9312
+    },
+    {
+      "epoch": 4.139111111111111,
+      "grad_norm": 3.7541697025299072,
+      "learning_rate": 3.446619217081851e-05,
+      "loss": 0.9619,
+      "step": 9313
+    },
+    {
+      "epoch": 4.139555555555556,
+      "grad_norm": 3.9635727405548096,
+      "learning_rate": 3.444839857651246e-05,
+      "loss": 0.9555,
+      "step": 9314
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 4.11905574798584,
+      "learning_rate": 3.44306049822064e-05,
+      "loss": 1.1721,
+      "step": 9315
+    },
+    {
+      "epoch": 4.140444444444444,
+      "grad_norm": 4.109530448913574,
+      "learning_rate": 3.441281138790036e-05,
+      "loss": 1.0309,
+      "step": 9316
+    },
+    {
+      "epoch": 4.140888888888889,
+      "grad_norm": 4.221639633178711,
+      "learning_rate": 3.439501779359431e-05,
+      "loss": 0.8867,
+      "step": 9317
+    },
+    {
+      "epoch": 4.141333333333334,
+      "grad_norm": 4.1241068840026855,
+      "learning_rate": 3.437722419928826e-05,
+      "loss": 1.2,
+      "step": 9318
+    },
+    {
+      "epoch": 4.141777777777778,
+      "grad_norm": 3.115713596343994,
+      "learning_rate": 3.435943060498221e-05,
+      "loss": 0.8329,
+      "step": 9319
+    },
+    {
+      "epoch": 4.142222222222222,
+      "grad_norm": 4.011024475097656,
+      "learning_rate": 3.434163701067616e-05,
+      "loss": 0.9659,
+      "step": 9320
+    },
+    {
+      "epoch": 4.142666666666667,
+      "grad_norm": 4.23065185546875,
+      "learning_rate": 3.4323843416370113e-05,
+      "loss": 0.834,
+      "step": 9321
+    },
+    {
+      "epoch": 4.143111111111111,
+      "grad_norm": 4.004124641418457,
+      "learning_rate": 3.4306049822064056e-05,
+      "loss": 0.8905,
+      "step": 9322
+    },
+    {
+      "epoch": 4.143555555555555,
+      "grad_norm": 4.145106315612793,
+      "learning_rate": 3.4288256227758006e-05,
+      "loss": 0.8814,
+      "step": 9323
+    },
+    {
+      "epoch": 4.144,
+      "grad_norm": 3.7468135356903076,
+      "learning_rate": 3.427046263345196e-05,
+      "loss": 1.143,
+      "step": 9324
+    },
+    {
+      "epoch": 4.144444444444445,
+      "grad_norm": 4.9056596755981445,
+      "learning_rate": 3.4252669039145906e-05,
+      "loss": 0.8847,
+      "step": 9325
+    },
+    {
+      "epoch": 4.144888888888889,
+      "grad_norm": 3.6427161693573,
+      "learning_rate": 3.423487544483986e-05,
+      "loss": 0.7846,
+      "step": 9326
+    },
+    {
+      "epoch": 4.145333333333333,
+      "grad_norm": 3.6018316745758057,
+      "learning_rate": 3.421708185053381e-05,
+      "loss": 1.1534,
+      "step": 9327
+    },
+    {
+      "epoch": 4.145777777777778,
+      "grad_norm": 4.335727214813232,
+      "learning_rate": 3.4199288256227755e-05,
+      "loss": 0.7984,
+      "step": 9328
+    },
+    {
+      "epoch": 4.146222222222222,
+      "grad_norm": 2.9448020458221436,
+      "learning_rate": 3.418149466192171e-05,
+      "loss": 0.6594,
+      "step": 9329
+    },
+    {
+      "epoch": 4.1466666666666665,
+      "grad_norm": 5.207244396209717,
+      "learning_rate": 3.416370106761566e-05,
+      "loss": 0.9542,
+      "step": 9330
+    },
+    {
+      "epoch": 4.147111111111111,
+      "grad_norm": 4.200593948364258,
+      "learning_rate": 3.414590747330961e-05,
+      "loss": 0.7464,
+      "step": 9331
+    },
+    {
+      "epoch": 4.147555555555556,
+      "grad_norm": 3.8014614582061768,
+      "learning_rate": 3.412811387900356e-05,
+      "loss": 0.8468,
+      "step": 9332
+    },
+    {
+      "epoch": 4.148,
+      "grad_norm": 4.140382289886475,
+      "learning_rate": 3.411032028469751e-05,
+      "loss": 1.0457,
+      "step": 9333
+    },
+    {
+      "epoch": 4.148444444444444,
+      "grad_norm": 4.095797061920166,
+      "learning_rate": 3.409252669039146e-05,
+      "loss": 0.8541,
+      "step": 9334
+    },
+    {
+      "epoch": 4.148888888888889,
+      "grad_norm": 4.9063825607299805,
+      "learning_rate": 3.407473309608541e-05,
+      "loss": 1.012,
+      "step": 9335
+    },
+    {
+      "epoch": 4.149333333333334,
+      "grad_norm": 5.420862197875977,
+      "learning_rate": 3.405693950177936e-05,
+      "loss": 1.0987,
+      "step": 9336
+    },
+    {
+      "epoch": 4.149777777777778,
+      "grad_norm": 4.148375988006592,
+      "learning_rate": 3.4039145907473316e-05,
+      "loss": 0.7677,
+      "step": 9337
+    },
+    {
+      "epoch": 4.150222222222222,
+      "grad_norm": 4.134524345397949,
+      "learning_rate": 3.402135231316726e-05,
+      "loss": 0.7619,
+      "step": 9338
+    },
+    {
+      "epoch": 4.150666666666667,
+      "grad_norm": 3.2314364910125732,
+      "learning_rate": 3.400355871886121e-05,
+      "loss": 0.6115,
+      "step": 9339
+    },
+    {
+      "epoch": 4.151111111111111,
+      "grad_norm": 4.153497219085693,
+      "learning_rate": 3.3985765124555165e-05,
+      "loss": 1.0227,
+      "step": 9340
+    },
+    {
+      "epoch": 4.151555555555555,
+      "grad_norm": 5.603724002838135,
+      "learning_rate": 3.396797153024911e-05,
+      "loss": 1.2158,
+      "step": 9341
+    },
+    {
+      "epoch": 4.152,
+      "grad_norm": 4.0254716873168945,
+      "learning_rate": 3.3950177935943064e-05,
+      "loss": 0.9139,
+      "step": 9342
+    },
+    {
+      "epoch": 4.152444444444445,
+      "grad_norm": 3.7253804206848145,
+      "learning_rate": 3.3932384341637014e-05,
+      "loss": 0.6698,
+      "step": 9343
+    },
+    {
+      "epoch": 4.152888888888889,
+      "grad_norm": 3.8816728591918945,
+      "learning_rate": 3.391459074733096e-05,
+      "loss": 0.5557,
+      "step": 9344
+    },
+    {
+      "epoch": 4.153333333333333,
+      "grad_norm": 4.364738941192627,
+      "learning_rate": 3.3896797153024914e-05,
+      "loss": 0.7052,
+      "step": 9345
+    },
+    {
+      "epoch": 4.153777777777778,
+      "grad_norm": 4.066445827484131,
+      "learning_rate": 3.387900355871886e-05,
+      "loss": 0.6252,
+      "step": 9346
+    },
+    {
+      "epoch": 4.154222222222222,
+      "grad_norm": 5.738955497741699,
+      "learning_rate": 3.386120996441281e-05,
+      "loss": 1.3448,
+      "step": 9347
+    },
+    {
+      "epoch": 4.1546666666666665,
+      "grad_norm": 4.564332962036133,
+      "learning_rate": 3.384341637010676e-05,
+      "loss": 0.5821,
+      "step": 9348
+    },
+    {
+      "epoch": 4.155111111111111,
+      "grad_norm": 4.382009983062744,
+      "learning_rate": 3.382562277580071e-05,
+      "loss": 0.7055,
+      "step": 9349
+    },
+    {
+      "epoch": 4.155555555555556,
+      "grad_norm": 3.8401434421539307,
+      "learning_rate": 3.380782918149467e-05,
+      "loss": 0.4056,
+      "step": 9350
+    },
+    {
+      "epoch": 4.156,
+      "grad_norm": 2.941746711730957,
+      "learning_rate": 3.379003558718861e-05,
+      "loss": 1.3041,
+      "step": 9351
+    },
+    {
+      "epoch": 4.156444444444444,
+      "grad_norm": 3.0022342205047607,
+      "learning_rate": 3.377224199288256e-05,
+      "loss": 1.2223,
+      "step": 9352
+    },
+    {
+      "epoch": 4.156888888888889,
+      "grad_norm": 3.1641345024108887,
+      "learning_rate": 3.375444839857652e-05,
+      "loss": 1.1629,
+      "step": 9353
+    },
+    {
+      "epoch": 4.157333333333334,
+      "grad_norm": 3.8065378665924072,
+      "learning_rate": 3.373665480427046e-05,
+      "loss": 1.7449,
+      "step": 9354
+    },
+    {
+      "epoch": 4.157777777777778,
+      "grad_norm": 3.80554461479187,
+      "learning_rate": 3.371886120996441e-05,
+      "loss": 1.2545,
+      "step": 9355
+    },
+    {
+      "epoch": 4.158222222222222,
+      "grad_norm": 3.1130597591400146,
+      "learning_rate": 3.370106761565837e-05,
+      "loss": 0.8787,
+      "step": 9356
+    },
+    {
+      "epoch": 4.158666666666667,
+      "grad_norm": 3.7158026695251465,
+      "learning_rate": 3.368327402135231e-05,
+      "loss": 1.171,
+      "step": 9357
+    },
+    {
+      "epoch": 4.159111111111111,
+      "grad_norm": 3.599818468093872,
+      "learning_rate": 3.366548042704627e-05,
+      "loss": 1.0849,
+      "step": 9358
+    },
+    {
+      "epoch": 4.1595555555555555,
+      "grad_norm": 3.1732017993927,
+      "learning_rate": 3.3647686832740216e-05,
+      "loss": 0.9308,
+      "step": 9359
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 3.3031604290008545,
+      "learning_rate": 3.3629893238434166e-05,
+      "loss": 0.8818,
+      "step": 9360
+    },
+    {
+      "epoch": 4.160444444444445,
+      "grad_norm": 3.976465940475464,
+      "learning_rate": 3.3612099644128116e-05,
+      "loss": 1.2361,
+      "step": 9361
+    },
+    {
+      "epoch": 4.160888888888889,
+      "grad_norm": 3.4227919578552246,
+      "learning_rate": 3.3594306049822066e-05,
+      "loss": 0.9153,
+      "step": 9362
+    },
+    {
+      "epoch": 4.161333333333333,
+      "grad_norm": 4.071769714355469,
+      "learning_rate": 3.3576512455516015e-05,
+      "loss": 1.1655,
+      "step": 9363
+    },
+    {
+      "epoch": 4.161777777777778,
+      "grad_norm": 3.795288324356079,
+      "learning_rate": 3.3558718861209965e-05,
+      "loss": 0.9261,
+      "step": 9364
+    },
+    {
+      "epoch": 4.162222222222222,
+      "grad_norm": 4.52875280380249,
+      "learning_rate": 3.3540925266903915e-05,
+      "loss": 1.2094,
+      "step": 9365
+    },
+    {
+      "epoch": 4.1626666666666665,
+      "grad_norm": 3.7887825965881348,
+      "learning_rate": 3.352313167259787e-05,
+      "loss": 0.9441,
+      "step": 9366
+    },
+    {
+      "epoch": 4.163111111111111,
+      "grad_norm": 3.6934292316436768,
+      "learning_rate": 3.3505338078291814e-05,
+      "loss": 1.0685,
+      "step": 9367
+    },
+    {
+      "epoch": 4.163555555555556,
+      "grad_norm": 4.112295150756836,
+      "learning_rate": 3.3487544483985764e-05,
+      "loss": 0.8924,
+      "step": 9368
+    },
+    {
+      "epoch": 4.164,
+      "grad_norm": 4.231710910797119,
+      "learning_rate": 3.346975088967972e-05,
+      "loss": 1.2001,
+      "step": 9369
+    },
+    {
+      "epoch": 4.164444444444444,
+      "grad_norm": 4.923578262329102,
+      "learning_rate": 3.345195729537366e-05,
+      "loss": 1.2478,
+      "step": 9370
+    },
+    {
+      "epoch": 4.164888888888889,
+      "grad_norm": 4.2798895835876465,
+      "learning_rate": 3.343416370106762e-05,
+      "loss": 0.9576,
+      "step": 9371
+    },
+    {
+      "epoch": 4.165333333333333,
+      "grad_norm": 3.816706895828247,
+      "learning_rate": 3.341637010676157e-05,
+      "loss": 0.9874,
+      "step": 9372
+    },
+    {
+      "epoch": 4.165777777777778,
+      "grad_norm": 4.367424011230469,
+      "learning_rate": 3.339857651245551e-05,
+      "loss": 0.93,
+      "step": 9373
+    },
+    {
+      "epoch": 4.166222222222222,
+      "grad_norm": 3.3792166709899902,
+      "learning_rate": 3.338078291814947e-05,
+      "loss": 0.7842,
+      "step": 9374
+    },
+    {
+      "epoch": 4.166666666666667,
+      "grad_norm": 4.096624374389648,
+      "learning_rate": 3.336298932384342e-05,
+      "loss": 0.6063,
+      "step": 9375
+    },
+    {
+      "epoch": 4.167111111111111,
+      "grad_norm": 3.6580982208251953,
+      "learning_rate": 3.334519572953737e-05,
+      "loss": 1.2232,
+      "step": 9376
+    },
+    {
+      "epoch": 4.1675555555555555,
+      "grad_norm": 4.178760528564453,
+      "learning_rate": 3.332740213523132e-05,
+      "loss": 1.1404,
+      "step": 9377
+    },
+    {
+      "epoch": 4.168,
+      "grad_norm": 2.8333945274353027,
+      "learning_rate": 3.330960854092527e-05,
+      "loss": 0.3454,
+      "step": 9378
+    },
+    {
+      "epoch": 4.168444444444445,
+      "grad_norm": 4.03084135055542,
+      "learning_rate": 3.329181494661922e-05,
+      "loss": 0.9051,
+      "step": 9379
+    },
+    {
+      "epoch": 4.168888888888889,
+      "grad_norm": 4.452610015869141,
+      "learning_rate": 3.327402135231317e-05,
+      "loss": 0.8157,
+      "step": 9380
+    },
+    {
+      "epoch": 4.169333333333333,
+      "grad_norm": 3.308134078979492,
+      "learning_rate": 3.325622775800712e-05,
+      "loss": 0.5715,
+      "step": 9381
+    },
+    {
+      "epoch": 4.169777777777778,
+      "grad_norm": 5.042409896850586,
+      "learning_rate": 3.3238434163701074e-05,
+      "loss": 0.9263,
+      "step": 9382
+    },
+    {
+      "epoch": 4.170222222222222,
+      "grad_norm": 5.0167646408081055,
+      "learning_rate": 3.3220640569395016e-05,
+      "loss": 1.4425,
+      "step": 9383
+    },
+    {
+      "epoch": 4.1706666666666665,
+      "grad_norm": 3.9779274463653564,
+      "learning_rate": 3.3202846975088966e-05,
+      "loss": 0.8751,
+      "step": 9384
+    },
+    {
+      "epoch": 4.171111111111111,
+      "grad_norm": 3.9376795291900635,
+      "learning_rate": 3.318505338078292e-05,
+      "loss": 0.9257,
+      "step": 9385
+    },
+    {
+      "epoch": 4.171555555555556,
+      "grad_norm": 6.37144136428833,
+      "learning_rate": 3.3167259786476866e-05,
+      "loss": 0.7895,
+      "step": 9386
+    },
+    {
+      "epoch": 4.172,
+      "grad_norm": 3.9572556018829346,
+      "learning_rate": 3.314946619217082e-05,
+      "loss": 0.8676,
+      "step": 9387
+    },
+    {
+      "epoch": 4.172444444444444,
+      "grad_norm": 4.135096073150635,
+      "learning_rate": 3.313167259786477e-05,
+      "loss": 0.977,
+      "step": 9388
+    },
+    {
+      "epoch": 4.172888888888889,
+      "grad_norm": 4.981659412384033,
+      "learning_rate": 3.311387900355872e-05,
+      "loss": 0.8796,
+      "step": 9389
+    },
+    {
+      "epoch": 4.173333333333334,
+      "grad_norm": 1.5012166500091553,
+      "learning_rate": 3.309608540925267e-05,
+      "loss": 0.1896,
+      "step": 9390
+    },
+    {
+      "epoch": 4.173777777777778,
+      "grad_norm": 3.713898181915283,
+      "learning_rate": 3.307829181494662e-05,
+      "loss": 0.9846,
+      "step": 9391
+    },
+    {
+      "epoch": 4.174222222222222,
+      "grad_norm": 4.347773551940918,
+      "learning_rate": 3.306049822064057e-05,
+      "loss": 0.9879,
+      "step": 9392
+    },
+    {
+      "epoch": 4.174666666666667,
+      "grad_norm": 4.975940227508545,
+      "learning_rate": 3.304270462633452e-05,
+      "loss": 0.8384,
+      "step": 9393
+    },
+    {
+      "epoch": 4.175111111111111,
+      "grad_norm": 4.148822784423828,
+      "learning_rate": 3.302491103202847e-05,
+      "loss": 0.621,
+      "step": 9394
+    },
+    {
+      "epoch": 4.1755555555555555,
+      "grad_norm": 3.4947614669799805,
+      "learning_rate": 3.300711743772243e-05,
+      "loss": 0.6469,
+      "step": 9395
+    },
+    {
+      "epoch": 4.176,
+      "grad_norm": 3.335649251937866,
+      "learning_rate": 3.298932384341637e-05,
+      "loss": 0.5926,
+      "step": 9396
+    },
+    {
+      "epoch": 4.176444444444445,
+      "grad_norm": 5.286258220672607,
+      "learning_rate": 3.297153024911032e-05,
+      "loss": 1.1108,
+      "step": 9397
+    },
+    {
+      "epoch": 4.176888888888889,
+      "grad_norm": 4.690072059631348,
+      "learning_rate": 3.2953736654804276e-05,
+      "loss": 0.9287,
+      "step": 9398
+    },
+    {
+      "epoch": 4.177333333333333,
+      "grad_norm": 5.190115928649902,
+      "learning_rate": 3.293594306049822e-05,
+      "loss": 0.8765,
+      "step": 9399
+    },
+    {
+      "epoch": 4.177777777777778,
+      "grad_norm": 5.4480719566345215,
+      "learning_rate": 3.291814946619217e-05,
+      "loss": 0.8428,
+      "step": 9400
+    },
+    {
+      "epoch": 4.178222222222222,
+      "grad_norm": 1.7260178327560425,
+      "learning_rate": 3.2900355871886125e-05,
+      "loss": 0.6587,
+      "step": 9401
+    },
+    {
+      "epoch": 4.1786666666666665,
+      "grad_norm": 3.0524849891662598,
+      "learning_rate": 3.288256227758007e-05,
+      "loss": 1.6461,
+      "step": 9402
+    },
+    {
+      "epoch": 4.179111111111111,
+      "grad_norm": 3.80712890625,
+      "learning_rate": 3.2864768683274024e-05,
+      "loss": 1.5275,
+      "step": 9403
+    },
+    {
+      "epoch": 4.179555555555556,
+      "grad_norm": 3.202948808670044,
+      "learning_rate": 3.2846975088967974e-05,
+      "loss": 1.101,
+      "step": 9404
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 3.789433002471924,
+      "learning_rate": 3.2829181494661924e-05,
+      "loss": 1.4379,
+      "step": 9405
+    },
+    {
+      "epoch": 4.180444444444444,
+      "grad_norm": 3.376875162124634,
+      "learning_rate": 3.2811387900355874e-05,
+      "loss": 1.099,
+      "step": 9406
+    },
+    {
+      "epoch": 4.180888888888889,
+      "grad_norm": 3.4530889987945557,
+      "learning_rate": 3.279359430604982e-05,
+      "loss": 1.5196,
+      "step": 9407
+    },
+    {
+      "epoch": 4.181333333333333,
+      "grad_norm": 4.1841607093811035,
+      "learning_rate": 3.277580071174377e-05,
+      "loss": 1.3571,
+      "step": 9408
+    },
+    {
+      "epoch": 4.181777777777778,
+      "grad_norm": 4.057764053344727,
+      "learning_rate": 3.275800711743772e-05,
+      "loss": 1.0025,
+      "step": 9409
+    },
+    {
+      "epoch": 4.182222222222222,
+      "grad_norm": 3.6850550174713135,
+      "learning_rate": 3.274021352313167e-05,
+      "loss": 1.0223,
+      "step": 9410
+    },
+    {
+      "epoch": 4.182666666666667,
+      "grad_norm": 3.958878517150879,
+      "learning_rate": 3.272241992882563e-05,
+      "loss": 1.1901,
+      "step": 9411
+    },
+    {
+      "epoch": 4.183111111111111,
+      "grad_norm": 4.859387397766113,
+      "learning_rate": 3.270462633451957e-05,
+      "loss": 0.9504,
+      "step": 9412
+    },
+    {
+      "epoch": 4.1835555555555555,
+      "grad_norm": 2.9225330352783203,
+      "learning_rate": 3.268683274021352e-05,
+      "loss": 0.5792,
+      "step": 9413
+    },
+    {
+      "epoch": 4.184,
+      "grad_norm": 3.700796127319336,
+      "learning_rate": 3.266903914590748e-05,
+      "loss": 0.5557,
+      "step": 9414
+    },
+    {
+      "epoch": 4.184444444444445,
+      "grad_norm": 2.9793245792388916,
+      "learning_rate": 3.265124555160142e-05,
+      "loss": 0.5418,
+      "step": 9415
+    },
+    {
+      "epoch": 4.184888888888889,
+      "grad_norm": 4.47357702255249,
+      "learning_rate": 3.263345195729538e-05,
+      "loss": 1.2536,
+      "step": 9416
+    },
+    {
+      "epoch": 4.185333333333333,
+      "grad_norm": 3.908679485321045,
+      "learning_rate": 3.261565836298933e-05,
+      "loss": 1.2,
+      "step": 9417
+    },
+    {
+      "epoch": 4.185777777777778,
+      "grad_norm": 3.9256107807159424,
+      "learning_rate": 3.259786476868328e-05,
+      "loss": 0.7822,
+      "step": 9418
+    },
+    {
+      "epoch": 4.186222222222222,
+      "grad_norm": 4.371975421905518,
+      "learning_rate": 3.258007117437723e-05,
+      "loss": 0.8867,
+      "step": 9419
+    },
+    {
+      "epoch": 4.1866666666666665,
+      "grad_norm": 3.3875746726989746,
+      "learning_rate": 3.2562277580071177e-05,
+      "loss": 0.8726,
+      "step": 9420
+    },
+    {
+      "epoch": 4.187111111111111,
+      "grad_norm": 4.142739295959473,
+      "learning_rate": 3.2544483985765126e-05,
+      "loss": 0.6472,
+      "step": 9421
+    },
+    {
+      "epoch": 4.187555555555556,
+      "grad_norm": 3.9463632106781006,
+      "learning_rate": 3.2526690391459076e-05,
+      "loss": 1.1603,
+      "step": 9422
+    },
+    {
+      "epoch": 4.188,
+      "grad_norm": 5.386812210083008,
+      "learning_rate": 3.2508896797153026e-05,
+      "loss": 0.9674,
+      "step": 9423
+    },
+    {
+      "epoch": 4.188444444444444,
+      "grad_norm": 5.0619001388549805,
+      "learning_rate": 3.2491103202846975e-05,
+      "loss": 1.3021,
+      "step": 9424
+    },
+    {
+      "epoch": 4.188888888888889,
+      "grad_norm": 4.491568565368652,
+      "learning_rate": 3.2473309608540925e-05,
+      "loss": 1.1475,
+      "step": 9425
+    },
+    {
+      "epoch": 4.189333333333333,
+      "grad_norm": 4.122431755065918,
+      "learning_rate": 3.2455516014234875e-05,
+      "loss": 0.8021,
+      "step": 9426
+    },
+    {
+      "epoch": 4.189777777777778,
+      "grad_norm": 4.491034507751465,
+      "learning_rate": 3.243772241992883e-05,
+      "loss": 1.0464,
+      "step": 9427
+    },
+    {
+      "epoch": 4.190222222222222,
+      "grad_norm": 3.6117091178894043,
+      "learning_rate": 3.2419928825622774e-05,
+      "loss": 0.8858,
+      "step": 9428
+    },
+    {
+      "epoch": 4.190666666666667,
+      "grad_norm": 5.300388813018799,
+      "learning_rate": 3.2402135231316724e-05,
+      "loss": 1.0057,
+      "step": 9429
+    },
+    {
+      "epoch": 4.191111111111111,
+      "grad_norm": 3.5130858421325684,
+      "learning_rate": 3.238434163701068e-05,
+      "loss": 0.4761,
+      "step": 9430
+    },
+    {
+      "epoch": 4.1915555555555555,
+      "grad_norm": 4.297408103942871,
+      "learning_rate": 3.2366548042704623e-05,
+      "loss": 0.8196,
+      "step": 9431
+    },
+    {
+      "epoch": 4.192,
+      "grad_norm": 4.353086471557617,
+      "learning_rate": 3.234875444839858e-05,
+      "loss": 1.031,
+      "step": 9432
+    },
+    {
+      "epoch": 4.192444444444445,
+      "grad_norm": 3.7016918659210205,
+      "learning_rate": 3.233096085409253e-05,
+      "loss": 0.8314,
+      "step": 9433
+    },
+    {
+      "epoch": 4.192888888888889,
+      "grad_norm": 3.4741709232330322,
+      "learning_rate": 3.231316725978648e-05,
+      "loss": 0.894,
+      "step": 9434
+    },
+    {
+      "epoch": 4.193333333333333,
+      "grad_norm": 3.2105934619903564,
+      "learning_rate": 3.229537366548043e-05,
+      "loss": 0.5393,
+      "step": 9435
+    },
+    {
+      "epoch": 4.193777777777778,
+      "grad_norm": 4.7250447273254395,
+      "learning_rate": 3.227758007117438e-05,
+      "loss": 1.047,
+      "step": 9436
+    },
+    {
+      "epoch": 4.194222222222222,
+      "grad_norm": 4.391071796417236,
+      "learning_rate": 3.225978647686833e-05,
+      "loss": 0.7677,
+      "step": 9437
+    },
+    {
+      "epoch": 4.1946666666666665,
+      "grad_norm": 2.0137879848480225,
+      "learning_rate": 3.224199288256228e-05,
+      "loss": 0.2664,
+      "step": 9438
+    },
+    {
+      "epoch": 4.195111111111111,
+      "grad_norm": 4.647293567657471,
+      "learning_rate": 3.222419928825623e-05,
+      "loss": 0.6575,
+      "step": 9439
+    },
+    {
+      "epoch": 4.195555555555556,
+      "grad_norm": 3.5285840034484863,
+      "learning_rate": 3.2206405693950184e-05,
+      "loss": 0.8251,
+      "step": 9440
+    },
+    {
+      "epoch": 4.196,
+      "grad_norm": 4.494802474975586,
+      "learning_rate": 3.218861209964413e-05,
+      "loss": 0.7617,
+      "step": 9441
+    },
+    {
+      "epoch": 4.196444444444444,
+      "grad_norm": 3.300382614135742,
+      "learning_rate": 3.217081850533808e-05,
+      "loss": 0.8305,
+      "step": 9442
+    },
+    {
+      "epoch": 4.196888888888889,
+      "grad_norm": 5.507649898529053,
+      "learning_rate": 3.2153024911032034e-05,
+      "loss": 0.8494,
+      "step": 9443
+    },
+    {
+      "epoch": 4.197333333333333,
+      "grad_norm": 4.006106853485107,
+      "learning_rate": 3.2135231316725977e-05,
+      "loss": 0.6544,
+      "step": 9444
+    },
+    {
+      "epoch": 4.197777777777778,
+      "grad_norm": 4.67263126373291,
+      "learning_rate": 3.2117437722419926e-05,
+      "loss": 0.9304,
+      "step": 9445
+    },
+    {
+      "epoch": 4.198222222222222,
+      "grad_norm": 4.910307884216309,
+      "learning_rate": 3.209964412811388e-05,
+      "loss": 0.8954,
+      "step": 9446
+    },
+    {
+      "epoch": 4.198666666666667,
+      "grad_norm": 6.133388996124268,
+      "learning_rate": 3.208185053380783e-05,
+      "loss": 0.6887,
+      "step": 9447
+    },
+    {
+      "epoch": 4.199111111111111,
+      "grad_norm": 4.447690010070801,
+      "learning_rate": 3.206405693950178e-05,
+      "loss": 0.9454,
+      "step": 9448
+    },
+    {
+      "epoch": 4.1995555555555555,
+      "grad_norm": 4.388811111450195,
+      "learning_rate": 3.204626334519573e-05,
+      "loss": 0.8632,
+      "step": 9449
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 3.5809247493743896,
+      "learning_rate": 3.202846975088968e-05,
+      "loss": 0.1799,
+      "step": 9450
+    },
+    {
+      "epoch": 4.200444444444445,
+      "grad_norm": 2.599862575531006,
+      "learning_rate": 3.201067615658363e-05,
+      "loss": 1.5166,
+      "step": 9451
+    },
+    {
+      "epoch": 4.200888888888889,
+      "grad_norm": 3.278106451034546,
+      "learning_rate": 3.199288256227758e-05,
+      "loss": 1.6669,
+      "step": 9452
+    },
+    {
+      "epoch": 4.201333333333333,
+      "grad_norm": 2.277113676071167,
+      "learning_rate": 3.197508896797153e-05,
+      "loss": 0.5748,
+      "step": 9453
+    },
+    {
+      "epoch": 4.201777777777778,
+      "grad_norm": 3.8572428226470947,
+      "learning_rate": 3.195729537366548e-05,
+      "loss": 1.4989,
+      "step": 9454
+    },
+    {
+      "epoch": 4.202222222222222,
+      "grad_norm": 3.595043659210205,
+      "learning_rate": 3.193950177935943e-05,
+      "loss": 1.0113,
+      "step": 9455
+    },
+    {
+      "epoch": 4.2026666666666666,
+      "grad_norm": 3.511258840560913,
+      "learning_rate": 3.192170818505339e-05,
+      "loss": 1.4261,
+      "step": 9456
+    },
+    {
+      "epoch": 4.203111111111111,
+      "grad_norm": 3.4274392127990723,
+      "learning_rate": 3.190391459074733e-05,
+      "loss": 1.0837,
+      "step": 9457
+    },
+    {
+      "epoch": 4.203555555555556,
+      "grad_norm": 3.7889490127563477,
+      "learning_rate": 3.188612099644128e-05,
+      "loss": 1.3672,
+      "step": 9458
+    },
+    {
+      "epoch": 4.204,
+      "grad_norm": 3.413290500640869,
+      "learning_rate": 3.1868327402135236e-05,
+      "loss": 1.0793,
+      "step": 9459
+    },
+    {
+      "epoch": 4.204444444444444,
+      "grad_norm": 3.5661888122558594,
+      "learning_rate": 3.185053380782918e-05,
+      "loss": 1.704,
+      "step": 9460
+    },
+    {
+      "epoch": 4.204888888888889,
+      "grad_norm": 3.6765122413635254,
+      "learning_rate": 3.1832740213523135e-05,
+      "loss": 1.1704,
+      "step": 9461
+    },
+    {
+      "epoch": 4.205333333333333,
+      "grad_norm": 3.350315570831299,
+      "learning_rate": 3.1814946619217085e-05,
+      "loss": 1.5331,
+      "step": 9462
+    },
+    {
+      "epoch": 4.205777777777778,
+      "grad_norm": 3.8041248321533203,
+      "learning_rate": 3.1797153024911035e-05,
+      "loss": 1.1945,
+      "step": 9463
+    },
+    {
+      "epoch": 4.206222222222222,
+      "grad_norm": 3.120882987976074,
+      "learning_rate": 3.1779359430604985e-05,
+      "loss": 0.9308,
+      "step": 9464
+    },
+    {
+      "epoch": 4.206666666666667,
+      "grad_norm": 3.490065574645996,
+      "learning_rate": 3.1761565836298934e-05,
+      "loss": 1.0944,
+      "step": 9465
+    },
+    {
+      "epoch": 4.207111111111111,
+      "grad_norm": 3.804657459259033,
+      "learning_rate": 3.1743772241992884e-05,
+      "loss": 1.2981,
+      "step": 9466
+    },
+    {
+      "epoch": 4.2075555555555555,
+      "grad_norm": 4.033463954925537,
+      "learning_rate": 3.1725978647686834e-05,
+      "loss": 0.8921,
+      "step": 9467
+    },
+    {
+      "epoch": 4.208,
+      "grad_norm": 3.6122589111328125,
+      "learning_rate": 3.1708185053380783e-05,
+      "loss": 0.8588,
+      "step": 9468
+    },
+    {
+      "epoch": 4.208444444444444,
+      "grad_norm": 4.304235935211182,
+      "learning_rate": 3.169039145907473e-05,
+      "loss": 1.2107,
+      "step": 9469
+    },
+    {
+      "epoch": 4.208888888888889,
+      "grad_norm": 4.344990253448486,
+      "learning_rate": 3.167259786476868e-05,
+      "loss": 1.1185,
+      "step": 9470
+    },
+    {
+      "epoch": 4.209333333333333,
+      "grad_norm": 3.5502078533172607,
+      "learning_rate": 3.165480427046263e-05,
+      "loss": 1.0842,
+      "step": 9471
+    },
+    {
+      "epoch": 4.209777777777778,
+      "grad_norm": 4.145521640777588,
+      "learning_rate": 3.163701067615659e-05,
+      "loss": 1.2853,
+      "step": 9472
+    },
+    {
+      "epoch": 4.210222222222222,
+      "grad_norm": 3.878098964691162,
+      "learning_rate": 3.161921708185053e-05,
+      "loss": 0.8788,
+      "step": 9473
+    },
+    {
+      "epoch": 4.210666666666667,
+      "grad_norm": 3.4567415714263916,
+      "learning_rate": 3.160142348754448e-05,
+      "loss": 0.9139,
+      "step": 9474
+    },
+    {
+      "epoch": 4.211111111111111,
+      "grad_norm": 3.87986421585083,
+      "learning_rate": 3.158362989323844e-05,
+      "loss": 0.7227,
+      "step": 9475
+    },
+    {
+      "epoch": 4.211555555555556,
+      "grad_norm": 5.376968860626221,
+      "learning_rate": 3.156583629893239e-05,
+      "loss": 1.4257,
+      "step": 9476
+    },
+    {
+      "epoch": 4.212,
+      "grad_norm": 5.025868892669678,
+      "learning_rate": 3.154804270462634e-05,
+      "loss": 0.8064,
+      "step": 9477
+    },
+    {
+      "epoch": 4.212444444444444,
+      "grad_norm": 3.2380011081695557,
+      "learning_rate": 3.153024911032029e-05,
+      "loss": 0.6622,
+      "step": 9478
+    },
+    {
+      "epoch": 4.212888888888889,
+      "grad_norm": 4.465753555297852,
+      "learning_rate": 3.151245551601424e-05,
+      "loss": 0.9084,
+      "step": 9479
+    },
+    {
+      "epoch": 4.213333333333333,
+      "grad_norm": 4.312685489654541,
+      "learning_rate": 3.149466192170819e-05,
+      "loss": 1.0134,
+      "step": 9480
+    },
+    {
+      "epoch": 4.213777777777778,
+      "grad_norm": 3.1639740467071533,
+      "learning_rate": 3.147686832740214e-05,
+      "loss": 0.6682,
+      "step": 9481
+    },
+    {
+      "epoch": 4.214222222222222,
+      "grad_norm": 3.7357571125030518,
+      "learning_rate": 3.1459074733096086e-05,
+      "loss": 0.8452,
+      "step": 9482
+    },
+    {
+      "epoch": 4.214666666666667,
+      "grad_norm": 2.6543757915496826,
+      "learning_rate": 3.1441281138790036e-05,
+      "loss": 0.5144,
+      "step": 9483
+    },
+    {
+      "epoch": 4.215111111111111,
+      "grad_norm": 4.193578243255615,
+      "learning_rate": 3.1423487544483986e-05,
+      "loss": 0.6794,
+      "step": 9484
+    },
+    {
+      "epoch": 4.2155555555555555,
+      "grad_norm": 3.771239757537842,
+      "learning_rate": 3.140569395017794e-05,
+      "loss": 0.7003,
+      "step": 9485
+    },
+    {
+      "epoch": 4.216,
+      "grad_norm": 4.657698631286621,
+      "learning_rate": 3.1387900355871885e-05,
+      "loss": 1.0667,
+      "step": 9486
+    },
+    {
+      "epoch": 4.216444444444445,
+      "grad_norm": 3.7996761798858643,
+      "learning_rate": 3.1370106761565835e-05,
+      "loss": 0.7508,
+      "step": 9487
+    },
+    {
+      "epoch": 4.216888888888889,
+      "grad_norm": 3.673656940460205,
+      "learning_rate": 3.135231316725979e-05,
+      "loss": 0.9042,
+      "step": 9488
+    },
+    {
+      "epoch": 4.217333333333333,
+      "grad_norm": 4.215774059295654,
+      "learning_rate": 3.1334519572953734e-05,
+      "loss": 0.5879,
+      "step": 9489
+    },
+    {
+      "epoch": 4.217777777777778,
+      "grad_norm": 5.385339260101318,
+      "learning_rate": 3.1316725978647684e-05,
+      "loss": 1.1556,
+      "step": 9490
+    },
+    {
+      "epoch": 4.218222222222222,
+      "grad_norm": 5.48359489440918,
+      "learning_rate": 3.129893238434164e-05,
+      "loss": 0.8134,
+      "step": 9491
+    },
+    {
+      "epoch": 4.218666666666667,
+      "grad_norm": 4.134088516235352,
+      "learning_rate": 3.128113879003559e-05,
+      "loss": 0.9353,
+      "step": 9492
+    },
+    {
+      "epoch": 4.219111111111111,
+      "grad_norm": 4.422386646270752,
+      "learning_rate": 3.126334519572954e-05,
+      "loss": 0.7743,
+      "step": 9493
+    },
+    {
+      "epoch": 4.219555555555556,
+      "grad_norm": 4.430136680603027,
+      "learning_rate": 3.124555160142349e-05,
+      "loss": 0.8493,
+      "step": 9494
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 4.236592769622803,
+      "learning_rate": 3.122775800711744e-05,
+      "loss": 0.6579,
+      "step": 9495
+    },
+    {
+      "epoch": 4.220444444444444,
+      "grad_norm": 3.639747142791748,
+      "learning_rate": 3.120996441281139e-05,
+      "loss": 0.756,
+      "step": 9496
+    },
+    {
+      "epoch": 4.220888888888889,
+      "grad_norm": 4.818484783172607,
+      "learning_rate": 3.119217081850534e-05,
+      "loss": 1.1835,
+      "step": 9497
+    },
+    {
+      "epoch": 4.221333333333333,
+      "grad_norm": 4.322108745574951,
+      "learning_rate": 3.117437722419929e-05,
+      "loss": 0.6088,
+      "step": 9498
+    },
+    {
+      "epoch": 4.221777777777778,
+      "grad_norm": 4.885183334350586,
+      "learning_rate": 3.115658362989324e-05,
+      "loss": 0.8394,
+      "step": 9499
+    },
+    {
+      "epoch": 4.222222222222222,
+      "grad_norm": 0.4897964298725128,
+      "learning_rate": 3.113879003558719e-05,
+      "loss": 0.0723,
+      "step": 9500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 11250,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.299541558648013e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}