diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,27774 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9994323557237466,
+  "eval_steps": 500,
+  "global_step": 3963,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007568590350047304,
+      "grad_norm": 123.05032348632812,
+      "learning_rate": 0.0,
+      "loss": 2.0667,
+      "step": 1
+    },
+    {
+      "epoch": 0.0015137180700094607,
+      "grad_norm": 59.42024612426758,
+      "learning_rate": 3.5436764027111585e-06,
+      "loss": 1.9891,
+      "step": 2
+    },
+    {
+      "epoch": 0.002270577105014191,
+      "grad_norm": 74.41653442382812,
+      "learning_rate": 5.61659421298763e-06,
+      "loss": 1.9745,
+      "step": 3
+    },
+    {
+      "epoch": 0.0030274361400189215,
+      "grad_norm": 65.0828857421875,
+      "learning_rate": 7.087352805422317e-06,
+      "loss": 1.9442,
+      "step": 4
+    },
+    {
+      "epoch": 0.003784295175023652,
+      "grad_norm": 77.46288299560547,
+      "learning_rate": 8.228161798644422e-06,
+      "loss": 1.911,
+      "step": 5
+    },
+    {
+      "epoch": 0.004541154210028382,
+      "grad_norm": 140.7876434326172,
+      "learning_rate": 9.160270615698787e-06,
+      "loss": 1.9,
+      "step": 6
+    },
+    {
+      "epoch": 0.005298013245033113,
+      "grad_norm": 39.34813690185547,
+      "learning_rate": 9.948357391330555e-06,
+      "loss": 1.8421,
+      "step": 7
+    },
+    {
+      "epoch": 0.006054872280037843,
+      "grad_norm": 42.30311584472656,
+      "learning_rate": 1.0631029208133474e-05,
+      "loss": 1.8634,
+      "step": 8
+    },
+    {
+      "epoch": 0.006811731315042573,
+      "grad_norm": 56.799530029296875,
+      "learning_rate": 1.123318842597526e-05,
+      "loss": 1.8391,
+      "step": 9
+    },
+    {
+      "epoch": 0.007568590350047304,
+      "grad_norm": 74.55519104003906,
+      "learning_rate": 1.1771838201355582e-05,
+      "loss": 1.7531,
+      "step": 10
+    },
+    {
+      "epoch": 0.008325449385052034,
+      "grad_norm": 31.099952697753906,
+      "learning_rate": 1.2259106193757859e-05,
+      "loss": 1.7675,
+      "step": 11
+    },
+    {
+      "epoch": 0.009082308420056764,
+      "grad_norm": 49.344966888427734,
+      "learning_rate": 1.2703947018409945e-05,
+      "loss": 1.7896,
+      "step": 12
+    },
+    {
+      "epoch": 0.009839167455061495,
+      "grad_norm": 48.00835418701172,
+      "learning_rate": 1.311316090883172e-05,
+      "loss": 1.8585,
+      "step": 13
+    },
+    {
+      "epoch": 0.010596026490066225,
+      "grad_norm": 38.080078125,
+      "learning_rate": 1.3492033794041713e-05,
+      "loss": 1.7329,
+      "step": 14
+    },
+    {
+      "epoch": 0.011352885525070956,
+      "grad_norm": 45.796382904052734,
+      "learning_rate": 1.384475601163205e-05,
+      "loss": 1.8033,
+      "step": 15
+    },
+    {
+      "epoch": 0.012109744560075686,
+      "grad_norm": 35.87776184082031,
+      "learning_rate": 1.4174705610844634e-05,
+      "loss": 1.7784,
+      "step": 16
+    },
+    {
+      "epoch": 0.012866603595080416,
+      "grad_norm": 25.678325653076172,
+      "learning_rate": 1.4484645617497535e-05,
+      "loss": 1.7741,
+      "step": 17
+    },
+    {
+      "epoch": 0.013623462630085147,
+      "grad_norm": 28.66301918029785,
+      "learning_rate": 1.4776864828686414e-05,
+      "loss": 1.7957,
+      "step": 18
+    },
+    {
+      "epoch": 0.014380321665089877,
+      "grad_norm": 37.723976135253906,
+      "learning_rate": 1.505328048981752e-05,
+      "loss": 1.695,
+      "step": 19
+    },
+    {
+      "epoch": 0.015137180700094607,
+      "grad_norm": 36.247718811035156,
+      "learning_rate": 1.5315514604066738e-05,
+      "loss": 1.7189,
+      "step": 20
+    },
+    {
+      "epoch": 0.015894039735099338,
+      "grad_norm": 23.032033920288086,
+      "learning_rate": 1.5564951604318184e-05,
+      "loss": 1.6817,
+      "step": 21
+    },
+    {
+      "epoch": 0.016650898770104068,
+      "grad_norm": 28.1435489654541,
+      "learning_rate": 1.580278259646902e-05,
+      "loss": 1.7185,
+      "step": 22
+    },
+    {
+      "epoch": 0.0174077578051088,
+      "grad_norm": 75.195068359375,
+      "learning_rate": 1.603003975988117e-05,
+      "loss": 1.7109,
+      "step": 23
+    },
+    {
+      "epoch": 0.01816461684011353,
+      "grad_norm": 30.104032516479492,
+      "learning_rate": 1.6247623421121105e-05,
+      "loss": 1.6333,
+      "step": 24
+    },
+    {
+      "epoch": 0.01892147587511826,
+      "grad_norm": 24.25992774963379,
+      "learning_rate": 1.6456323597288844e-05,
+      "loss": 1.6416,
+      "step": 25
+    },
+    {
+      "epoch": 0.01967833491012299,
+      "grad_norm": 28.712949752807617,
+      "learning_rate": 1.6656837311542876e-05,
+      "loss": 1.6712,
+      "step": 26
+    },
+    {
+      "epoch": 0.02043519394512772,
+      "grad_norm": 26.72446060180664,
+      "learning_rate": 1.6849782638962885e-05,
+      "loss": 1.5939,
+      "step": 27
+    },
+    {
+      "epoch": 0.02119205298013245,
+      "grad_norm": 20.644784927368164,
+      "learning_rate": 1.7035710196752873e-05,
+      "loss": 1.6718,
+      "step": 28
+    },
+    {
+      "epoch": 0.02194891201513718,
+      "grad_norm": 24.643821716308594,
+      "learning_rate": 1.7215112617252848e-05,
+      "loss": 1.6778,
+      "step": 29
+    },
+    {
+      "epoch": 0.02270577105014191,
+      "grad_norm": 22.256315231323242,
+      "learning_rate": 1.738843241434321e-05,
+      "loss": 1.6233,
+      "step": 30
+    },
+    {
+      "epoch": 0.02346263008514664,
+      "grad_norm": 24.24241065979004,
+      "learning_rate": 1.7556068559516658e-05,
+      "loss": 1.6744,
+      "step": 31
+    },
+    {
+      "epoch": 0.024219489120151372,
+      "grad_norm": 37.310150146484375,
+      "learning_rate": 1.7718382013555794e-05,
+      "loss": 1.6556,
+      "step": 32
+    },
+    {
+      "epoch": 0.024976348155156102,
+      "grad_norm": 48.23684310913086,
+      "learning_rate": 1.7875700406745488e-05,
+      "loss": 1.6575,
+      "step": 33
+    },
+    {
+      "epoch": 0.025733207190160833,
+      "grad_norm": 24.166748046875,
+      "learning_rate": 1.8028322020208693e-05,
+      "loss": 1.6946,
+      "step": 34
+    },
+    {
+      "epoch": 0.026490066225165563,
+      "grad_norm": 29.902538299560547,
+      "learning_rate": 1.817651918997498e-05,
+      "loss": 1.6453,
+      "step": 35
+    },
+    {
+      "epoch": 0.027246925260170293,
+      "grad_norm": 27.081722259521484,
+      "learning_rate": 1.8320541231397574e-05,
+      "loss": 1.5933,
+      "step": 36
+    },
+    {
+      "epoch": 0.028003784295175024,
+      "grad_norm": 28.04783058166504,
+      "learning_rate": 1.8460616962803535e-05,
+      "loss": 1.676,
+      "step": 37
+    },
+    {
+      "epoch": 0.028760643330179754,
+      "grad_norm": 49.34148406982422,
+      "learning_rate": 1.859695689252868e-05,
+      "loss": 1.7431,
+      "step": 38
+    },
+    {
+      "epoch": 0.029517502365184484,
+      "grad_norm": 32.92803192138672,
+      "learning_rate": 1.872975512181935e-05,
+      "loss": 1.7004,
+      "step": 39
+    },
+    {
+      "epoch": 0.030274361400189215,
+      "grad_norm": 28.5103816986084,
+      "learning_rate": 1.8859191006777896e-05,
+      "loss": 1.646,
+      "step": 40
+    },
+    {
+      "epoch": 0.031031220435193945,
+      "grad_norm": 16.540956497192383,
+      "learning_rate": 1.8985430615062968e-05,
+      "loss": 1.6601,
+      "step": 41
+    },
+    {
+      "epoch": 0.031788079470198675,
+      "grad_norm": 26.588886260986328,
+      "learning_rate": 1.9108628007029345e-05,
+      "loss": 1.6273,
+      "step": 42
+    },
+    {
+      "epoch": 0.03254493850520341,
+      "grad_norm": 42.97163009643555,
+      "learning_rate": 1.9228926366101076e-05,
+      "loss": 1.5573,
+      "step": 43
+    },
+    {
+      "epoch": 0.033301797540208136,
+      "grad_norm": 40.343658447265625,
+      "learning_rate": 1.9346458999180177e-05,
+      "loss": 1.5821,
+      "step": 44
+    },
+    {
+      "epoch": 0.03405865657521287,
+      "grad_norm": 41.81525421142578,
+      "learning_rate": 1.946135022461968e-05,
+      "loss": 1.5927,
+      "step": 45
+    },
+    {
+      "epoch": 0.0348155156102176,
+      "grad_norm": 24.463436126708984,
+      "learning_rate": 1.9573716162592327e-05,
+      "loss": 1.6377,
+      "step": 46
+    },
+    {
+      "epoch": 0.03557237464522233,
+      "grad_norm": 21.16547203063965,
+      "learning_rate": 1.9683665440452515e-05,
+      "loss": 1.6151,
+      "step": 47
+    },
+    {
+      "epoch": 0.03632923368022706,
+      "grad_norm": 75.09718322753906,
+      "learning_rate": 1.9791299823832263e-05,
+      "loss": 1.6261,
+      "step": 48
+    },
+    {
+      "epoch": 0.03708609271523179,
+      "grad_norm": 30.04339027404785,
+      "learning_rate": 1.989671478266111e-05,
+      "loss": 1.6229,
+      "step": 49
+    },
+    {
+      "epoch": 0.03784295175023652,
+      "grad_norm": 100.00825500488281,
+      "learning_rate": 2e-05,
+      "loss": 1.6116,
+      "step": 50
+    },
+    {
+      "epoch": 0.03859981078524125,
+      "grad_norm": 28.68238639831543,
+      "learning_rate": 1.9999996777398546e-05,
+      "loss": 1.6644,
+      "step": 51
+    },
+    {
+      "epoch": 0.03935666982024598,
+      "grad_norm": 68.90675354003906,
+      "learning_rate": 1.9999987109596254e-05,
+      "loss": 1.6159,
+      "step": 52
+    },
+    {
+      "epoch": 0.04011352885525071,
+      "grad_norm": 81.32110595703125,
+      "learning_rate": 1.999997099659936e-05,
+      "loss": 1.7117,
+      "step": 53
+    },
+    {
+      "epoch": 0.04087038789025544,
+      "grad_norm": 171.38938903808594,
+      "learning_rate": 1.999994843841825e-05,
+      "loss": 1.5922,
+      "step": 54
+    },
+    {
+      "epoch": 0.041627246925260174,
+      "grad_norm": 43.32768249511719,
+      "learning_rate": 1.9999919435067465e-05,
+      "loss": 1.6245,
+      "step": 55
+    },
+    {
+      "epoch": 0.0423841059602649,
+      "grad_norm": 74.8743896484375,
+      "learning_rate": 1.9999883986565696e-05,
+      "loss": 1.6613,
+      "step": 56
+    },
+    {
+      "epoch": 0.043140964995269634,
+      "grad_norm": 18.4515438079834,
+      "learning_rate": 1.9999842092935797e-05,
+      "loss": 1.6048,
+      "step": 57
+    },
+    {
+      "epoch": 0.04389782403027436,
+      "grad_norm": 22.151155471801758,
+      "learning_rate": 1.999979375420477e-05,
+      "loss": 1.632,
+      "step": 58
+    },
+    {
+      "epoch": 0.044654683065279095,
+      "grad_norm": 19.156835556030273,
+      "learning_rate": 1.9999738970403775e-05,
+      "loss": 1.6189,
+      "step": 59
+    },
+    {
+      "epoch": 0.04541154210028382,
+      "grad_norm": 18.256269454956055,
+      "learning_rate": 1.999967774156812e-05,
+      "loss": 1.6306,
+      "step": 60
+    },
+    {
+      "epoch": 0.046168401135288556,
+      "grad_norm": 26.946420669555664,
+      "learning_rate": 1.999961006773728e-05,
+      "loss": 1.6411,
+      "step": 61
+    },
+    {
+      "epoch": 0.04692526017029328,
+      "grad_norm": 23.500892639160156,
+      "learning_rate": 1.999953594895487e-05,
+      "loss": 1.5876,
+      "step": 62
+    },
+    {
+      "epoch": 0.04768211920529802,
+      "grad_norm": 12.633379936218262,
+      "learning_rate": 1.999945538526867e-05,
+      "loss": 1.601,
+      "step": 63
+    },
+    {
+      "epoch": 0.048438978240302744,
+      "grad_norm": 15.04751205444336,
+      "learning_rate": 1.999936837673061e-05,
+      "loss": 1.6078,
+      "step": 64
+    },
+    {
+      "epoch": 0.04919583727530748,
+      "grad_norm": 11.406414985656738,
+      "learning_rate": 1.999927492339677e-05,
+      "loss": 1.5959,
+      "step": 65
+    },
+    {
+      "epoch": 0.049952696310312204,
+      "grad_norm": 11.040087699890137,
+      "learning_rate": 1.9999175025327395e-05,
+      "loss": 1.6025,
+      "step": 66
+    },
+    {
+      "epoch": 0.05070955534531694,
+      "grad_norm": 10.689179420471191,
+      "learning_rate": 1.999906868258687e-05,
+      "loss": 1.5797,
+      "step": 67
+    },
+    {
+      "epoch": 0.051466414380321665,
+      "grad_norm": 8.19336986541748,
+      "learning_rate": 1.9998955895243748e-05,
+      "loss": 1.5564,
+      "step": 68
+    },
+    {
+      "epoch": 0.0522232734153264,
+      "grad_norm": 13.38058853149414,
+      "learning_rate": 1.9998836663370726e-05,
+      "loss": 1.5584,
+      "step": 69
+    },
+    {
+      "epoch": 0.052980132450331126,
+      "grad_norm": 7.053563594818115,
+      "learning_rate": 1.9998710987044664e-05,
+      "loss": 1.5005,
+      "step": 70
+    },
+    {
+      "epoch": 0.05373699148533586,
+      "grad_norm": 6.7353105545043945,
+      "learning_rate": 1.9998578866346564e-05,
+      "loss": 1.5306,
+      "step": 71
+    },
+    {
+      "epoch": 0.054493850520340587,
+      "grad_norm": 5.975197792053223,
+      "learning_rate": 1.9998440301361598e-05,
+      "loss": 1.5885,
+      "step": 72
+    },
+    {
+      "epoch": 0.05525070955534532,
+      "grad_norm": 6.6494011878967285,
+      "learning_rate": 1.9998295292179073e-05,
+      "loss": 1.532,
+      "step": 73
+    },
+    {
+      "epoch": 0.05600756859035005,
+      "grad_norm": 5.434142589569092,
+      "learning_rate": 1.9998143838892468e-05,
+      "loss": 1.5764,
+      "step": 74
+    },
+    {
+      "epoch": 0.05676442762535478,
+      "grad_norm": 6.099053859710693,
+      "learning_rate": 1.99979859415994e-05,
+      "loss": 1.5187,
+      "step": 75
+    },
+    {
+      "epoch": 0.05752128666035951,
+      "grad_norm": 5.7470855712890625,
+      "learning_rate": 1.999782160040166e-05,
+      "loss": 1.5377,
+      "step": 76
+    },
+    {
+      "epoch": 0.05827814569536424,
+      "grad_norm": 5.577144145965576,
+      "learning_rate": 1.9997650815405167e-05,
+      "loss": 1.4817,
+      "step": 77
+    },
+    {
+      "epoch": 0.05903500473036897,
+      "grad_norm": 5.294313907623291,
+      "learning_rate": 1.999747358672001e-05,
+      "loss": 1.4812,
+      "step": 78
+    },
+    {
+      "epoch": 0.0597918637653737,
+      "grad_norm": 5.860252380371094,
+      "learning_rate": 1.9997289914460428e-05,
+      "loss": 1.524,
+      "step": 79
+    },
+    {
+      "epoch": 0.06054872280037843,
+      "grad_norm": 9.259795188903809,
+      "learning_rate": 1.9997099798744815e-05,
+      "loss": 1.4869,
+      "step": 80
+    },
+    {
+      "epoch": 0.06130558183538316,
+      "grad_norm": 5.9855852127075195,
+      "learning_rate": 1.999690323969571e-05,
+      "loss": 1.5187,
+      "step": 81
+    },
+    {
+      "epoch": 0.06206244087038789,
+      "grad_norm": 6.138685703277588,
+      "learning_rate": 1.9996700237439823e-05,
+      "loss": 1.5468,
+      "step": 82
+    },
+    {
+      "epoch": 0.06281929990539262,
+      "grad_norm": 6.596303939819336,
+      "learning_rate": 1.9996490792107997e-05,
+      "loss": 1.4899,
+      "step": 83
+    },
+    {
+      "epoch": 0.06357615894039735,
+      "grad_norm": 5.794029712677002,
+      "learning_rate": 1.9996274903835247e-05,
+      "loss": 1.4633,
+      "step": 84
+    },
+    {
+      "epoch": 0.06433301797540208,
+      "grad_norm": 5.815922260284424,
+      "learning_rate": 1.9996052572760723e-05,
+      "loss": 1.4968,
+      "step": 85
+    },
+    {
+      "epoch": 0.06508987701040682,
+      "grad_norm": 11.239161491394043,
+      "learning_rate": 1.9995823799027737e-05,
+      "loss": 1.4604,
+      "step": 86
+    },
+    {
+      "epoch": 0.06584673604541154,
+      "grad_norm": 5.367482662200928,
+      "learning_rate": 1.9995588582783753e-05,
+      "loss": 1.518,
+      "step": 87
+    },
+    {
+      "epoch": 0.06660359508041627,
+      "grad_norm": 19.351980209350586,
+      "learning_rate": 1.9995346924180394e-05,
+      "loss": 1.5267,
+      "step": 88
+    },
+    {
+      "epoch": 0.067360454115421,
+      "grad_norm": 6.7129316329956055,
+      "learning_rate": 1.999509882337342e-05,
+      "loss": 1.4639,
+      "step": 89
+    },
+    {
+      "epoch": 0.06811731315042574,
+      "grad_norm": 6.777989387512207,
+      "learning_rate": 1.999484428052276e-05,
+      "loss": 1.5127,
+      "step": 90
+    },
+    {
+      "epoch": 0.06887417218543046,
+      "grad_norm": 14.996123313903809,
+      "learning_rate": 1.9994583295792487e-05,
+      "loss": 1.5277,
+      "step": 91
+    },
+    {
+      "epoch": 0.0696310312204352,
+      "grad_norm": 6.6374311447143555,
+      "learning_rate": 1.9994315869350826e-05,
+      "loss": 1.4834,
+      "step": 92
+    },
+    {
+      "epoch": 0.07038789025543993,
+      "grad_norm": 11.19003963470459,
+      "learning_rate": 1.9994042001370154e-05,
+      "loss": 1.5084,
+      "step": 93
+    },
+    {
+      "epoch": 0.07114474929044466,
+      "grad_norm": 6.2547407150268555,
+      "learning_rate": 1.9993761692027007e-05,
+      "loss": 1.485,
+      "step": 94
+    },
+    {
+      "epoch": 0.07190160832544938,
+      "grad_norm": 6.645302772521973,
+      "learning_rate": 1.9993474941502067e-05,
+      "loss": 1.463,
+      "step": 95
+    },
+    {
+      "epoch": 0.07265846736045412,
+      "grad_norm": 7.073038578033447,
+      "learning_rate": 1.9993181749980168e-05,
+      "loss": 1.509,
+      "step": 96
+    },
+    {
+      "epoch": 0.07341532639545885,
+      "grad_norm": 6.401993274688721,
+      "learning_rate": 1.99928821176503e-05,
+      "loss": 1.4958,
+      "step": 97
+    },
+    {
+      "epoch": 0.07417218543046358,
+      "grad_norm": 6.126581192016602,
+      "learning_rate": 1.9992576044705596e-05,
+      "loss": 1.4449,
+      "step": 98
+    },
+    {
+      "epoch": 0.0749290444654683,
+      "grad_norm": 8.766273498535156,
+      "learning_rate": 1.9992263531343348e-05,
+      "loss": 1.5218,
+      "step": 99
+    },
+    {
+      "epoch": 0.07568590350047304,
+      "grad_norm": 5.65410852432251,
+      "learning_rate": 1.9991944577764996e-05,
+      "loss": 1.5205,
+      "step": 100
+    },
+    {
+      "epoch": 0.07644276253547777,
+      "grad_norm": 5.447603702545166,
+      "learning_rate": 1.9991619184176136e-05,
+      "loss": 1.4651,
+      "step": 101
+    },
+    {
+      "epoch": 0.0771996215704825,
+      "grad_norm": 5.317190647125244,
+      "learning_rate": 1.9991287350786512e-05,
+      "loss": 1.5059,
+      "step": 102
+    },
+    {
+      "epoch": 0.07795648060548722,
+      "grad_norm": 5.233520984649658,
+      "learning_rate": 1.9990949077810015e-05,
+      "loss": 1.4556,
+      "step": 103
+    },
+    {
+      "epoch": 0.07871333964049196,
+      "grad_norm": 4.955499649047852,
+      "learning_rate": 1.9990604365464693e-05,
+      "loss": 1.4236,
+      "step": 104
+    },
+    {
+      "epoch": 0.07947019867549669,
+      "grad_norm": 9.175353050231934,
+      "learning_rate": 1.9990253213972742e-05,
+      "loss": 1.4482,
+      "step": 105
+    },
+    {
+      "epoch": 0.08022705771050143,
+      "grad_norm": 5.2216057777404785,
+      "learning_rate": 1.998989562356051e-05,
+      "loss": 1.4902,
+      "step": 106
+    },
+    {
+      "epoch": 0.08098391674550615,
+      "grad_norm": 4.395474910736084,
+      "learning_rate": 1.9989531594458487e-05,
+      "loss": 1.4419,
+      "step": 107
+    },
+    {
+      "epoch": 0.08174077578051088,
+      "grad_norm": 4.641335487365723,
+      "learning_rate": 1.998916112690133e-05,
+      "loss": 1.4715,
+      "step": 108
+    },
+    {
+      "epoch": 0.08249763481551561,
+      "grad_norm": 5.315745830535889,
+      "learning_rate": 1.9988784221127834e-05,
+      "loss": 1.4742,
+      "step": 109
+    },
+    {
+      "epoch": 0.08325449385052035,
+      "grad_norm": 5.404274940490723,
+      "learning_rate": 1.998840087738095e-05,
+      "loss": 1.4579,
+      "step": 110
+    },
+    {
+      "epoch": 0.08401135288552507,
+      "grad_norm": 4.288702011108398,
+      "learning_rate": 1.9988011095907768e-05,
+      "loss": 1.49,
+      "step": 111
+    },
+    {
+      "epoch": 0.0847682119205298,
+      "grad_norm": 4.434887409210205,
+      "learning_rate": 1.9987614876959536e-05,
+      "loss": 1.4946,
+      "step": 112
+    },
+    {
+      "epoch": 0.08552507095553454,
+      "grad_norm": 5.428564071655273,
+      "learning_rate": 1.9987212220791657e-05,
+      "loss": 1.3817,
+      "step": 113
+    },
+    {
+      "epoch": 0.08628192999053927,
+      "grad_norm": 3.9893720149993896,
+      "learning_rate": 1.9986803127663672e-05,
+      "loss": 1.4428,
+      "step": 114
+    },
+    {
+      "epoch": 0.08703878902554399,
+      "grad_norm": 4.35543966293335,
+      "learning_rate": 1.998638759783928e-05,
+      "loss": 1.3801,
+      "step": 115
+    },
+    {
+      "epoch": 0.08779564806054872,
+      "grad_norm": 4.2772722244262695,
+      "learning_rate": 1.9985965631586318e-05,
+      "loss": 1.3975,
+      "step": 116
+    },
+    {
+      "epoch": 0.08855250709555346,
+      "grad_norm": 4.769036769866943,
+      "learning_rate": 1.9985537229176787e-05,
+      "loss": 1.4413,
+      "step": 117
+    },
+    {
+      "epoch": 0.08930936613055819,
+      "grad_norm": 4.7659759521484375,
+      "learning_rate": 1.9985102390886825e-05,
+      "loss": 1.4665,
+      "step": 118
+    },
+    {
+      "epoch": 0.09006622516556291,
+      "grad_norm": 5.218923091888428,
+      "learning_rate": 1.9984661116996723e-05,
+      "loss": 1.4544,
+      "step": 119
+    },
+    {
+      "epoch": 0.09082308420056764,
+      "grad_norm": 4.296699047088623,
+      "learning_rate": 1.9984213407790924e-05,
+      "loss": 1.3944,
+      "step": 120
+    },
+    {
+      "epoch": 0.09157994323557238,
+      "grad_norm": 3.866936683654785,
+      "learning_rate": 1.9983759263558003e-05,
+      "loss": 1.4273,
+      "step": 121
+    },
+    {
+      "epoch": 0.09233680227057711,
+      "grad_norm": 4.711172103881836,
+      "learning_rate": 1.99832986845907e-05,
+      "loss": 1.4208,
+      "step": 122
+    },
+    {
+      "epoch": 0.09309366130558183,
+      "grad_norm": 4.553902626037598,
+      "learning_rate": 1.9982831671185905e-05,
+      "loss": 1.525,
+      "step": 123
+    },
+    {
+      "epoch": 0.09385052034058657,
+      "grad_norm": 4.0878801345825195,
+      "learning_rate": 1.9982358223644635e-05,
+      "loss": 1.4621,
+      "step": 124
+    },
+    {
+      "epoch": 0.0946073793755913,
+      "grad_norm": 4.239192485809326,
+      "learning_rate": 1.9981878342272074e-05,
+      "loss": 1.4151,
+      "step": 125
+    },
+    {
+      "epoch": 0.09536423841059603,
+      "grad_norm": 3.9742391109466553,
+      "learning_rate": 1.9981392027377548e-05,
+      "loss": 1.4588,
+      "step": 126
+    },
+    {
+      "epoch": 0.09612109744560075,
+      "grad_norm": 4.459286212921143,
+      "learning_rate": 1.9980899279274523e-05,
+      "loss": 1.414,
+      "step": 127
+    },
+    {
+      "epoch": 0.09687795648060549,
+      "grad_norm": 4.164027214050293,
+      "learning_rate": 1.9980400098280622e-05,
+      "loss": 1.4572,
+      "step": 128
+    },
+    {
+      "epoch": 0.09763481551561022,
+      "grad_norm": 4.778876304626465,
+      "learning_rate": 1.9979894484717604e-05,
+      "loss": 1.4279,
+      "step": 129
+    },
+    {
+      "epoch": 0.09839167455061495,
+      "grad_norm": 4.639044761657715,
+      "learning_rate": 1.9979382438911383e-05,
+      "loss": 1.4343,
+      "step": 130
+    },
+    {
+      "epoch": 0.09914853358561967,
+      "grad_norm": 4.090446949005127,
+      "learning_rate": 1.9978863961192018e-05,
+      "loss": 1.4802,
+      "step": 131
+    },
+    {
+      "epoch": 0.09990539262062441,
+      "grad_norm": 4.360771656036377,
+      "learning_rate": 1.9978339051893702e-05,
+      "loss": 1.4552,
+      "step": 132
+    },
+    {
+      "epoch": 0.10066225165562914,
+      "grad_norm": 3.851464033126831,
+      "learning_rate": 1.9977807711354796e-05,
+      "loss": 1.3779,
+      "step": 133
+    },
+    {
+      "epoch": 0.10141911069063388,
+      "grad_norm": 4.016122341156006,
+      "learning_rate": 1.997726993991779e-05,
+      "loss": 1.4313,
+      "step": 134
+    },
+    {
+      "epoch": 0.1021759697256386,
+      "grad_norm": 4.009467124938965,
+      "learning_rate": 1.997672573792932e-05,
+      "loss": 1.491,
+      "step": 135
+    },
+    {
+      "epoch": 0.10293282876064333,
+      "grad_norm": 3.8596322536468506,
+      "learning_rate": 1.997617510574018e-05,
+      "loss": 1.4724,
+      "step": 136
+    },
+    {
+      "epoch": 0.10368968779564806,
+      "grad_norm": 3.6011574268341064,
+      "learning_rate": 1.9975618043705282e-05,
+      "loss": 1.3931,
+      "step": 137
+    },
+    {
+      "epoch": 0.1044465468306528,
+      "grad_norm": 4.025736331939697,
+      "learning_rate": 1.997505455218371e-05,
+      "loss": 1.4269,
+      "step": 138
+    },
+    {
+      "epoch": 0.10520340586565752,
+      "grad_norm": 3.760977268218994,
+      "learning_rate": 1.9974484631538685e-05,
+      "loss": 1.4311,
+      "step": 139
+    },
+    {
+      "epoch": 0.10596026490066225,
+      "grad_norm": 4.554644584655762,
+      "learning_rate": 1.9973908282137565e-05,
+      "loss": 1.4535,
+      "step": 140
+    },
+    {
+      "epoch": 0.10671712393566699,
+      "grad_norm": 4.12142276763916,
+      "learning_rate": 1.9973325504351856e-05,
+      "loss": 1.4111,
+      "step": 141
+    },
+    {
+      "epoch": 0.10747398297067172,
+      "grad_norm": 3.9459025859832764,
+      "learning_rate": 1.9972736298557207e-05,
+      "loss": 1.4424,
+      "step": 142
+    },
+    {
+      "epoch": 0.10823084200567644,
+      "grad_norm": 3.65413236618042,
+      "learning_rate": 1.9972140665133412e-05,
+      "loss": 1.3589,
+      "step": 143
+    },
+    {
+      "epoch": 0.10898770104068117,
+      "grad_norm": 3.935250997543335,
+      "learning_rate": 1.997153860446441e-05,
+      "loss": 1.3985,
+      "step": 144
+    },
+    {
+      "epoch": 0.1097445600756859,
+      "grad_norm": 4.394814968109131,
+      "learning_rate": 1.9970930116938273e-05,
+      "loss": 1.4304,
+      "step": 145
+    },
+    {
+      "epoch": 0.11050141911069064,
+      "grad_norm": 3.6491141319274902,
+      "learning_rate": 1.997031520294723e-05,
+      "loss": 1.3928,
+      "step": 146
+    },
+    {
+      "epoch": 0.11125827814569536,
+      "grad_norm": 4.235386848449707,
+      "learning_rate": 1.9969693862887643e-05,
+      "loss": 1.4712,
+      "step": 147
+    },
+    {
+      "epoch": 0.1120151371807001,
+      "grad_norm": 4.189138412475586,
+      "learning_rate": 1.996906609716002e-05,
+      "loss": 1.3609,
+      "step": 148
+    },
+    {
+      "epoch": 0.11277199621570483,
+      "grad_norm": 3.729450225830078,
+      "learning_rate": 1.9968431906169005e-05,
+      "loss": 1.4229,
+      "step": 149
+    },
+    {
+      "epoch": 0.11352885525070956,
+      "grad_norm": 3.915863513946533,
+      "learning_rate": 1.996779129032339e-05,
+      "loss": 1.3628,
+      "step": 150
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 4.461569786071777,
+      "learning_rate": 1.9967144250036104e-05,
+      "loss": 1.4087,
+      "step": 151
+    },
+    {
+      "epoch": 0.11504257332071902,
+      "grad_norm": 4.412698745727539,
+      "learning_rate": 1.9966490785724223e-05,
+      "loss": 1.4392,
+      "step": 152
+    },
+    {
+      "epoch": 0.11579943235572375,
+      "grad_norm": 4.236743450164795,
+      "learning_rate": 1.9965830897808954e-05,
+      "loss": 1.4391,
+      "step": 153
+    },
+    {
+      "epoch": 0.11655629139072848,
+      "grad_norm": 4.672597408294678,
+      "learning_rate": 1.996516458671566e-05,
+      "loss": 1.3995,
+      "step": 154
+    },
+    {
+      "epoch": 0.1173131504257332,
+      "grad_norm": 5.059709072113037,
+      "learning_rate": 1.9964491852873833e-05,
+      "loss": 1.3566,
+      "step": 155
+    },
+    {
+      "epoch": 0.11807000946073794,
+      "grad_norm": 4.973750114440918,
+      "learning_rate": 1.99638126967171e-05,
+      "loss": 1.3993,
+      "step": 156
+    },
+    {
+      "epoch": 0.11882686849574267,
+      "grad_norm": 4.362597942352295,
+      "learning_rate": 1.996312711868324e-05,
+      "loss": 1.4254,
+      "step": 157
+    },
+    {
+      "epoch": 0.1195837275307474,
+      "grad_norm": 4.407685279846191,
+      "learning_rate": 1.9962435119214164e-05,
+      "loss": 1.3983,
+      "step": 158
+    },
+    {
+      "epoch": 0.12034058656575213,
+      "grad_norm": 4.614277362823486,
+      "learning_rate": 1.9961736698755928e-05,
+      "loss": 1.412,
+      "step": 159
+    },
+    {
+      "epoch": 0.12109744560075686,
+      "grad_norm": 4.18186092376709,
+      "learning_rate": 1.9961031857758718e-05,
+      "loss": 1.3653,
+      "step": 160
+    },
+    {
+      "epoch": 0.12185430463576159,
+      "grad_norm": 4.011139392852783,
+      "learning_rate": 1.9960320596676866e-05,
+      "loss": 1.4234,
+      "step": 161
+    },
+    {
+      "epoch": 0.12261116367076633,
+      "grad_norm": 4.428970813751221,
+      "learning_rate": 1.9959602915968842e-05,
+      "loss": 1.3899,
+      "step": 162
+    },
+    {
+      "epoch": 0.12336802270577105,
+      "grad_norm": 4.968282222747803,
+      "learning_rate": 1.995887881609725e-05,
+      "loss": 1.4235,
+      "step": 163
+    },
+    {
+      "epoch": 0.12412488174077578,
+      "grad_norm": 4.600246906280518,
+      "learning_rate": 1.9958148297528833e-05,
+      "loss": 1.3656,
+      "step": 164
+    },
+    {
+      "epoch": 0.12488174077578051,
+      "grad_norm": 4.392306804656982,
+      "learning_rate": 1.9957411360734476e-05,
+      "loss": 1.3804,
+      "step": 165
+    },
+    {
+      "epoch": 0.12563859981078523,
+      "grad_norm": 4.066370964050293,
+      "learning_rate": 1.995666800618919e-05,
+      "loss": 1.4013,
+      "step": 166
+    },
+    {
+      "epoch": 0.12639545884578998,
+      "grad_norm": 3.9358901977539062,
+      "learning_rate": 1.995591823437214e-05,
+      "loss": 1.3887,
+      "step": 167
+    },
+    {
+      "epoch": 0.1271523178807947,
+      "grad_norm": 4.182436466217041,
+      "learning_rate": 1.9955162045766607e-05,
+      "loss": 1.4011,
+      "step": 168
+    },
+    {
+      "epoch": 0.12790917691579942,
+      "grad_norm": 3.8715391159057617,
+      "learning_rate": 1.9954399440860026e-05,
+      "loss": 1.3881,
+      "step": 169
+    },
+    {
+      "epoch": 0.12866603595080417,
+      "grad_norm": 4.34489107131958,
+      "learning_rate": 1.9953630420143958e-05,
+      "loss": 1.3679,
+      "step": 170
+    },
+    {
+      "epoch": 0.1294228949858089,
+      "grad_norm": 4.175931930541992,
+      "learning_rate": 1.9952854984114097e-05,
+      "loss": 1.3995,
+      "step": 171
+    },
+    {
+      "epoch": 0.13017975402081364,
+      "grad_norm": 6.231164932250977,
+      "learning_rate": 1.9952073133270288e-05,
+      "loss": 1.3826,
+      "step": 172
+    },
+    {
+      "epoch": 0.13093661305581836,
+      "grad_norm": 3.969299554824829,
+      "learning_rate": 1.9951284868116495e-05,
+      "loss": 1.427,
+      "step": 173
+    },
+    {
+      "epoch": 0.13169347209082308,
+      "grad_norm": 5.056988716125488,
+      "learning_rate": 1.9950490189160818e-05,
+      "loss": 1.4377,
+      "step": 174
+    },
+    {
+      "epoch": 0.13245033112582782,
+      "grad_norm": 3.9641916751861572,
+      "learning_rate": 1.99496890969155e-05,
+      "loss": 1.3729,
+      "step": 175
+    },
+    {
+      "epoch": 0.13320719016083254,
+      "grad_norm": 4.119785308837891,
+      "learning_rate": 1.9948881591896913e-05,
+      "loss": 1.4061,
+      "step": 176
+    },
+    {
+      "epoch": 0.13396404919583726,
+      "grad_norm": 4.154798984527588,
+      "learning_rate": 1.9948067674625557e-05,
+      "loss": 1.4383,
+      "step": 177
+    },
+    {
+      "epoch": 0.134720908230842,
+      "grad_norm": 4.396413326263428,
+      "learning_rate": 1.994724734562607e-05,
+      "loss": 1.3806,
+      "step": 178
+    },
+    {
+      "epoch": 0.13547776726584673,
+      "grad_norm": 10.802559852600098,
+      "learning_rate": 1.9946420605427235e-05,
+      "loss": 1.4279,
+      "step": 179
+    },
+    {
+      "epoch": 0.13623462630085148,
+      "grad_norm": 4.602297782897949,
+      "learning_rate": 1.9945587454561944e-05,
+      "loss": 1.3618,
+      "step": 180
+    },
+    {
+      "epoch": 0.1369914853358562,
+      "grad_norm": 4.874974727630615,
+      "learning_rate": 1.994474789356724e-05,
+      "loss": 1.3582,
+      "step": 181
+    },
+    {
+      "epoch": 0.13774834437086092,
+      "grad_norm": 5.023828983306885,
+      "learning_rate": 1.994390192298429e-05,
+      "loss": 1.3445,
+      "step": 182
+    },
+    {
+      "epoch": 0.13850520340586567,
+      "grad_norm": 4.938666343688965,
+      "learning_rate": 1.994304954335839e-05,
+      "loss": 1.4221,
+      "step": 183
+    },
+    {
+      "epoch": 0.1392620624408704,
+      "grad_norm": 5.975377559661865,
+      "learning_rate": 1.9942190755238973e-05,
+      "loss": 1.3947,
+      "step": 184
+    },
+    {
+      "epoch": 0.1400189214758751,
+      "grad_norm": 8.078311920166016,
+      "learning_rate": 1.9941325559179608e-05,
+      "loss": 1.3925,
+      "step": 185
+    },
+    {
+      "epoch": 0.14077578051087986,
+      "grad_norm": 5.0124897956848145,
+      "learning_rate": 1.9940453955737976e-05,
+      "loss": 1.3958,
+      "step": 186
+    },
+    {
+      "epoch": 0.14153263954588458,
+      "grad_norm": 4.94537353515625,
+      "learning_rate": 1.9939575945475905e-05,
+      "loss": 1.3855,
+      "step": 187
+    },
+    {
+      "epoch": 0.14228949858088932,
+      "grad_norm": 5.828818321228027,
+      "learning_rate": 1.9938691528959348e-05,
+      "loss": 1.4567,
+      "step": 188
+    },
+    {
+      "epoch": 0.14304635761589404,
+      "grad_norm": 4.672356605529785,
+      "learning_rate": 1.993780070675838e-05,
+      "loss": 1.3581,
+      "step": 189
+    },
+    {
+      "epoch": 0.14380321665089876,
+      "grad_norm": 5.052429676055908,
+      "learning_rate": 1.993690347944722e-05,
+      "loss": 1.3874,
+      "step": 190
+    },
+    {
+      "epoch": 0.1445600756859035,
+      "grad_norm": 4.454349040985107,
+      "learning_rate": 1.9935999847604204e-05,
+      "loss": 1.4282,
+      "step": 191
+    },
+    {
+      "epoch": 0.14531693472090823,
+      "grad_norm": 4.81812858581543,
+      "learning_rate": 1.9935089811811794e-05,
+      "loss": 1.4103,
+      "step": 192
+    },
+    {
+      "epoch": 0.14607379375591295,
+      "grad_norm": 3.8706412315368652,
+      "learning_rate": 1.993417337265659e-05,
+      "loss": 1.4024,
+      "step": 193
+    },
+    {
+      "epoch": 0.1468306527909177,
+      "grad_norm": 3.948594093322754,
+      "learning_rate": 1.9933250530729314e-05,
+      "loss": 1.387,
+      "step": 194
+    },
+    {
+      "epoch": 0.14758751182592242,
+      "grad_norm": 4.73719596862793,
+      "learning_rate": 1.993232128662482e-05,
+      "loss": 1.4528,
+      "step": 195
+    },
+    {
+      "epoch": 0.14834437086092717,
+      "grad_norm": 3.9017584323883057,
+      "learning_rate": 1.993138564094208e-05,
+      "loss": 1.4245,
+      "step": 196
+    },
+    {
+      "epoch": 0.14910122989593189,
+      "grad_norm": 6.6446309089660645,
+      "learning_rate": 1.9930443594284193e-05,
+      "loss": 1.4046,
+      "step": 197
+    },
+    {
+      "epoch": 0.1498580889309366,
+      "grad_norm": 4.191623210906982,
+      "learning_rate": 1.9929495147258395e-05,
+      "loss": 1.3987,
+      "step": 198
+    },
+    {
+      "epoch": 0.15061494796594135,
+      "grad_norm": 3.8362607955932617,
+      "learning_rate": 1.992854030047604e-05,
+      "loss": 1.3583,
+      "step": 199
+    },
+    {
+      "epoch": 0.15137180700094607,
+      "grad_norm": 4.051894187927246,
+      "learning_rate": 1.9927579054552603e-05,
+      "loss": 1.3856,
+      "step": 200
+    },
+    {
+      "epoch": 0.1521286660359508,
+      "grad_norm": 3.792412281036377,
+      "learning_rate": 1.992661141010769e-05,
+      "loss": 1.3961,
+      "step": 201
+    },
+    {
+      "epoch": 0.15288552507095554,
+      "grad_norm": 3.697641134262085,
+      "learning_rate": 1.992563736776503e-05,
+      "loss": 1.3808,
+      "step": 202
+    },
+    {
+      "epoch": 0.15364238410596026,
+      "grad_norm": 4.134721279144287,
+      "learning_rate": 1.992465692815248e-05,
+      "loss": 1.3594,
+      "step": 203
+    },
+    {
+      "epoch": 0.154399243140965,
+      "grad_norm": 4.171304225921631,
+      "learning_rate": 1.9923670091902013e-05,
+      "loss": 1.4217,
+      "step": 204
+    },
+    {
+      "epoch": 0.15515610217596973,
+      "grad_norm": 3.476039171218872,
+      "learning_rate": 1.992267685964973e-05,
+      "loss": 1.3967,
+      "step": 205
+    },
+    {
+      "epoch": 0.15591296121097445,
+      "grad_norm": 3.4347240924835205,
+      "learning_rate": 1.9921677232035846e-05,
+      "loss": 1.3422,
+      "step": 206
+    },
+    {
+      "epoch": 0.1566698202459792,
+      "grad_norm": 3.7200000286102295,
+      "learning_rate": 1.992067120970472e-05,
+      "loss": 1.3538,
+      "step": 207
+    },
+    {
+      "epoch": 0.15742667928098392,
+      "grad_norm": 3.8184263706207275,
+      "learning_rate": 1.9919658793304804e-05,
+      "loss": 1.3956,
+      "step": 208
+    },
+    {
+      "epoch": 0.15818353831598864,
+      "grad_norm": 3.761478900909424,
+      "learning_rate": 1.9918639983488694e-05,
+      "loss": 1.4233,
+      "step": 209
+    },
+    {
+      "epoch": 0.15894039735099338,
+      "grad_norm": 3.587502956390381,
+      "learning_rate": 1.99176147809131e-05,
+      "loss": 1.3514,
+      "step": 210
+    },
+    {
+      "epoch": 0.1596972563859981,
+      "grad_norm": 3.3828699588775635,
+      "learning_rate": 1.9916583186238847e-05,
+      "loss": 1.3766,
+      "step": 211
+    },
+    {
+      "epoch": 0.16045411542100285,
+      "grad_norm": 3.2444939613342285,
+      "learning_rate": 1.9915545200130893e-05,
+      "loss": 1.4051,
+      "step": 212
+    },
+    {
+      "epoch": 0.16121097445600757,
+      "grad_norm": 3.4360880851745605,
+      "learning_rate": 1.9914500823258298e-05,
+      "loss": 1.3364,
+      "step": 213
+    },
+    {
+      "epoch": 0.1619678334910123,
+      "grad_norm": 3.3002805709838867,
+      "learning_rate": 1.9913450056294255e-05,
+      "loss": 1.3807,
+      "step": 214
+    },
+    {
+      "epoch": 0.16272469252601704,
+      "grad_norm": 3.551203489303589,
+      "learning_rate": 1.991239289991608e-05,
+      "loss": 1.4077,
+      "step": 215
+    },
+    {
+      "epoch": 0.16348155156102176,
+      "grad_norm": 2.9857335090637207,
+      "learning_rate": 1.991132935480519e-05,
+      "loss": 1.3667,
+      "step": 216
+    },
+    {
+      "epoch": 0.16423841059602648,
+      "grad_norm": 3.935084342956543,
+      "learning_rate": 1.9910259421647136e-05,
+      "loss": 1.3973,
+      "step": 217
+    },
+    {
+      "epoch": 0.16499526963103123,
+      "grad_norm": 3.209479570388794,
+      "learning_rate": 1.9909183101131576e-05,
+      "loss": 1.3752,
+      "step": 218
+    },
+    {
+      "epoch": 0.16575212866603595,
+      "grad_norm": 3.311500072479248,
+      "learning_rate": 1.9908100393952293e-05,
+      "loss": 1.3566,
+      "step": 219
+    },
+    {
+      "epoch": 0.1665089877010407,
+      "grad_norm": 3.0751259326934814,
+      "learning_rate": 1.990701130080718e-05,
+      "loss": 1.411,
+      "step": 220
+    },
+    {
+      "epoch": 0.16726584673604541,
+      "grad_norm": 3.3133180141448975,
+      "learning_rate": 1.9905915822398257e-05,
+      "loss": 1.4006,
+      "step": 221
+    },
+    {
+      "epoch": 0.16802270577105013,
+      "grad_norm": 3.2017252445220947,
+      "learning_rate": 1.9904813959431646e-05,
+      "loss": 1.4028,
+      "step": 222
+    },
+    {
+      "epoch": 0.16877956480605488,
+      "grad_norm": 3.404691219329834,
+      "learning_rate": 1.9903705712617595e-05,
+      "loss": 1.355,
+      "step": 223
+    },
+    {
+      "epoch": 0.1695364238410596,
+      "grad_norm": 3.1049623489379883,
+      "learning_rate": 1.990259108267046e-05,
+      "loss": 1.3305,
+      "step": 224
+    },
+    {
+      "epoch": 0.17029328287606432,
+      "grad_norm": 3.3933444023132324,
+      "learning_rate": 1.990147007030871e-05,
+      "loss": 1.3718,
+      "step": 225
+    },
+    {
+      "epoch": 0.17105014191106907,
+      "grad_norm": 3.479591131210327,
+      "learning_rate": 1.9900342676254945e-05,
+      "loss": 1.393,
+      "step": 226
+    },
+    {
+      "epoch": 0.1718070009460738,
+      "grad_norm": 3.3810219764709473,
+      "learning_rate": 1.989920890123586e-05,
+      "loss": 1.3864,
+      "step": 227
+    },
+    {
+      "epoch": 0.17256385998107854,
+      "grad_norm": 3.4179928302764893,
+      "learning_rate": 1.9898068745982263e-05,
+      "loss": 1.3322,
+      "step": 228
+    },
+    {
+      "epoch": 0.17332071901608326,
+      "grad_norm": 3.288922071456909,
+      "learning_rate": 1.9896922211229088e-05,
+      "loss": 1.3738,
+      "step": 229
+    },
+    {
+      "epoch": 0.17407757805108798,
+      "grad_norm": 3.4045164585113525,
+      "learning_rate": 1.9895769297715373e-05,
+      "loss": 1.3509,
+      "step": 230
+    },
+    {
+      "epoch": 0.17483443708609273,
+      "grad_norm": 3.384779453277588,
+      "learning_rate": 1.9894610006184264e-05,
+      "loss": 1.3596,
+      "step": 231
+    },
+    {
+      "epoch": 0.17559129612109745,
+      "grad_norm": 3.6631815433502197,
+      "learning_rate": 1.989344433738303e-05,
+      "loss": 1.4126,
+      "step": 232
+    },
+    {
+      "epoch": 0.17634815515610217,
+      "grad_norm": 3.1958444118499756,
+      "learning_rate": 1.9892272292063034e-05,
+      "loss": 1.3711,
+      "step": 233
+    },
+    {
+      "epoch": 0.1771050141911069,
+      "grad_norm": 3.4087891578674316,
+      "learning_rate": 1.989109387097977e-05,
+      "loss": 1.3604,
+      "step": 234
+    },
+    {
+      "epoch": 0.17786187322611163,
+      "grad_norm": 3.5950968265533447,
+      "learning_rate": 1.988990907489282e-05,
+      "loss": 1.3464,
+      "step": 235
+    },
+    {
+      "epoch": 0.17861873226111638,
+      "grad_norm": 3.5223278999328613,
+      "learning_rate": 1.988871790456589e-05,
+      "loss": 1.3965,
+      "step": 236
+    },
+    {
+      "epoch": 0.1793755912961211,
+      "grad_norm": 3.259669780731201,
+      "learning_rate": 1.988752036076679e-05,
+      "loss": 1.3915,
+      "step": 237
+    },
+    {
+      "epoch": 0.18013245033112582,
+      "grad_norm": 3.1698622703552246,
+      "learning_rate": 1.9886316444267436e-05,
+      "loss": 1.3674,
+      "step": 238
+    },
+    {
+      "epoch": 0.18088930936613057,
+      "grad_norm": 3.336416482925415,
+      "learning_rate": 1.9885106155843857e-05,
+      "loss": 1.3476,
+      "step": 239
+    },
+    {
+      "epoch": 0.1816461684011353,
+      "grad_norm": 3.2986626625061035,
+      "learning_rate": 1.9883889496276188e-05,
+      "loss": 1.3139,
+      "step": 240
+    },
+    {
+      "epoch": 0.18240302743614,
+      "grad_norm": 3.2197721004486084,
+      "learning_rate": 1.9882666466348665e-05,
+      "loss": 1.3611,
+      "step": 241
+    },
+    {
+      "epoch": 0.18315988647114476,
+      "grad_norm": 3.120088815689087,
+      "learning_rate": 1.988143706684964e-05,
+      "loss": 1.364,
+      "step": 242
+    },
+    {
+      "epoch": 0.18391674550614948,
+      "grad_norm": 2.9464049339294434,
+      "learning_rate": 1.9880201298571558e-05,
+      "loss": 1.3295,
+      "step": 243
+    },
+    {
+      "epoch": 0.18467360454115422,
+      "grad_norm": 3.3369717597961426,
+      "learning_rate": 1.9878959162310983e-05,
+      "loss": 1.3669,
+      "step": 244
+    },
+    {
+      "epoch": 0.18543046357615894,
+      "grad_norm": 2.900787353515625,
+      "learning_rate": 1.987771065886857e-05,
+      "loss": 1.3565,
+      "step": 245
+    },
+    {
+      "epoch": 0.18618732261116366,
+      "grad_norm": 3.0211544036865234,
+      "learning_rate": 1.9876455789049096e-05,
+      "loss": 1.3882,
+      "step": 246
+    },
+    {
+      "epoch": 0.1869441816461684,
+      "grad_norm": 3.1576292514801025,
+      "learning_rate": 1.9875194553661415e-05,
+      "loss": 1.3075,
+      "step": 247
+    },
+    {
+      "epoch": 0.18770104068117313,
+      "grad_norm": 2.8850550651550293,
+      "learning_rate": 1.9873926953518515e-05,
+      "loss": 1.3665,
+      "step": 248
+    },
+    {
+      "epoch": 0.18845789971617785,
+      "grad_norm": 3.188582420349121,
+      "learning_rate": 1.9872652989437467e-05,
+      "loss": 1.3555,
+      "step": 249
+    },
+    {
+      "epoch": 0.1892147587511826,
+      "grad_norm": 4.313934803009033,
+      "learning_rate": 1.9871372662239446e-05,
+      "loss": 1.2937,
+      "step": 250
+    },
+    {
+      "epoch": 0.18997161778618732,
+      "grad_norm": 3.0744991302490234,
+      "learning_rate": 1.9870085972749733e-05,
+      "loss": 1.3289,
+      "step": 251
+    },
+    {
+      "epoch": 0.19072847682119207,
+      "grad_norm": 2.9217262268066406,
+      "learning_rate": 1.986879292179771e-05,
+      "loss": 1.33,
+      "step": 252
+    },
+    {
+      "epoch": 0.1914853358561968,
+      "grad_norm": 3.089919328689575,
+      "learning_rate": 1.986749351021686e-05,
+      "loss": 1.3532,
+      "step": 253
+    },
+    {
+      "epoch": 0.1922421948912015,
+      "grad_norm": 3.5609021186828613,
+      "learning_rate": 1.9866187738844753e-05,
+      "loss": 1.4002,
+      "step": 254
+    },
+    {
+      "epoch": 0.19299905392620625,
+      "grad_norm": 3.0856025218963623,
+      "learning_rate": 1.986487560852308e-05,
+      "loss": 1.3691,
+      "step": 255
+    },
+    {
+      "epoch": 0.19375591296121097,
+      "grad_norm": 2.679279327392578,
+      "learning_rate": 1.986355712009762e-05,
+      "loss": 1.3412,
+      "step": 256
+    },
+    {
+      "epoch": 0.1945127719962157,
+      "grad_norm": 3.1083905696868896,
+      "learning_rate": 1.9862232274418246e-05,
+      "loss": 1.3023,
+      "step": 257
+    },
+    {
+      "epoch": 0.19526963103122044,
+      "grad_norm": 2.726358413696289,
+      "learning_rate": 1.9860901072338936e-05,
+      "loss": 1.377,
+      "step": 258
+    },
+    {
+      "epoch": 0.19602649006622516,
+      "grad_norm": 2.966639995574951,
+      "learning_rate": 1.985956351471776e-05,
+      "loss": 1.304,
+      "step": 259
+    },
+    {
+      "epoch": 0.1967833491012299,
+      "grad_norm": 2.8776400089263916,
+      "learning_rate": 1.9858219602416887e-05,
+      "loss": 1.3481,
+      "step": 260
+    },
+    {
+      "epoch": 0.19754020813623463,
+      "grad_norm": 3.0099427700042725,
+      "learning_rate": 1.9856869336302588e-05,
+      "loss": 1.4332,
+      "step": 261
+    },
+    {
+      "epoch": 0.19829706717123935,
+      "grad_norm": 3.146959066390991,
+      "learning_rate": 1.985551271724522e-05,
+      "loss": 1.3372,
+      "step": 262
+    },
+    {
+      "epoch": 0.1990539262062441,
+      "grad_norm": 3.076327323913574,
+      "learning_rate": 1.9854149746119232e-05,
+      "loss": 1.3258,
+      "step": 263
+    },
+    {
+      "epoch": 0.19981078524124882,
+      "grad_norm": 2.8409347534179688,
+      "learning_rate": 1.9852780423803187e-05,
+      "loss": 1.2975,
+      "step": 264
+    },
+    {
+      "epoch": 0.20056764427625354,
+      "grad_norm": 3.1386849880218506,
+      "learning_rate": 1.9851404751179723e-05,
+      "loss": 1.3395,
+      "step": 265
+    },
+    {
+      "epoch": 0.20132450331125828,
+      "grad_norm": 3.104682445526123,
+      "learning_rate": 1.9850022729135578e-05,
+      "loss": 1.3667,
+      "step": 266
+    },
+    {
+      "epoch": 0.202081362346263,
+      "grad_norm": 3.337529182434082,
+      "learning_rate": 1.9848634358561584e-05,
+      "loss": 1.3145,
+      "step": 267
+    },
+    {
+      "epoch": 0.20283822138126775,
+      "grad_norm": 3.380446195602417,
+      "learning_rate": 1.984723964035266e-05,
+      "loss": 1.364,
+      "step": 268
+    },
+    {
+      "epoch": 0.20359508041627247,
+      "grad_norm": 3.161867141723633,
+      "learning_rate": 1.9845838575407824e-05,
+      "loss": 1.3333,
+      "step": 269
+    },
+    {
+      "epoch": 0.2043519394512772,
+      "grad_norm": 3.323434352874756,
+      "learning_rate": 1.9844431164630178e-05,
+      "loss": 1.3897,
+      "step": 270
+    },
+    {
+      "epoch": 0.20510879848628194,
+      "grad_norm": 3.4208099842071533,
+      "learning_rate": 1.984301740892692e-05,
+      "loss": 1.333,
+      "step": 271
+    },
+    {
+      "epoch": 0.20586565752128666,
+      "grad_norm": 3.178248643875122,
+      "learning_rate": 1.984159730920933e-05,
+      "loss": 1.3033,
+      "step": 272
+    },
+    {
+      "epoch": 0.20662251655629138,
+      "grad_norm": 3.0145297050476074,
+      "learning_rate": 1.9840170866392795e-05,
+      "loss": 1.3055,
+      "step": 273
+    },
+    {
+      "epoch": 0.20737937559129613,
+      "grad_norm": 3.6076059341430664,
+      "learning_rate": 1.9838738081396764e-05,
+      "loss": 1.3442,
+      "step": 274
+    },
+    {
+      "epoch": 0.20813623462630085,
+      "grad_norm": 3.3622937202453613,
+      "learning_rate": 1.9837298955144796e-05,
+      "loss": 1.3666,
+      "step": 275
+    },
+    {
+      "epoch": 0.2088930936613056,
+      "grad_norm": 3.782317876815796,
+      "learning_rate": 1.9835853488564527e-05,
+      "loss": 1.3791,
+      "step": 276
+    },
+    {
+      "epoch": 0.20964995269631032,
+      "grad_norm": 3.1874301433563232,
+      "learning_rate": 1.9834401682587688e-05,
+      "loss": 1.3703,
+      "step": 277
+    },
+    {
+      "epoch": 0.21040681173131504,
+      "grad_norm": 3.0065550804138184,
+      "learning_rate": 1.9832943538150083e-05,
+      "loss": 1.331,
+      "step": 278
+    },
+    {
+      "epoch": 0.21116367076631978,
+      "grad_norm": 3.953733444213867,
+      "learning_rate": 1.9831479056191618e-05,
+      "loss": 1.3855,
+      "step": 279
+    },
+    {
+      "epoch": 0.2119205298013245,
+      "grad_norm": 3.682438611984253,
+      "learning_rate": 1.983000823765627e-05,
+      "loss": 1.3605,
+      "step": 280
+    },
+    {
+      "epoch": 0.21267738883632922,
+      "grad_norm": 3.57037615776062,
+      "learning_rate": 1.9828531083492102e-05,
+      "loss": 1.3048,
+      "step": 281
+    },
+    {
+      "epoch": 0.21343424787133397,
+      "grad_norm": 3.4117233753204346,
+      "learning_rate": 1.9827047594651275e-05,
+      "loss": 1.3606,
+      "step": 282
+    },
+    {
+      "epoch": 0.2141911069063387,
+      "grad_norm": 2.914785623550415,
+      "learning_rate": 1.982555777209002e-05,
+      "loss": 1.3596,
+      "step": 283
+    },
+    {
+      "epoch": 0.21494796594134344,
+      "grad_norm": 3.271235942840576,
+      "learning_rate": 1.9824061616768652e-05,
+      "loss": 1.3208,
+      "step": 284
+    },
+    {
+      "epoch": 0.21570482497634816,
+      "grad_norm": 3.3142642974853516,
+      "learning_rate": 1.982255912965157e-05,
+      "loss": 1.3574,
+      "step": 285
+    },
+    {
+      "epoch": 0.21646168401135288,
+      "grad_norm": 3.752458095550537,
+      "learning_rate": 1.9821050311707253e-05,
+      "loss": 1.3818,
+      "step": 286
+    },
+    {
+      "epoch": 0.21721854304635763,
+      "grad_norm": 3.1010730266571045,
+      "learning_rate": 1.9819535163908266e-05,
+      "loss": 1.2799,
+      "step": 287
+    },
+    {
+      "epoch": 0.21797540208136235,
+      "grad_norm": 3.3089754581451416,
+      "learning_rate": 1.9818013687231252e-05,
+      "loss": 1.3719,
+      "step": 288
+    },
+    {
+      "epoch": 0.21873226111636707,
+      "grad_norm": 3.800584316253662,
+      "learning_rate": 1.9816485882656925e-05,
+      "loss": 1.3458,
+      "step": 289
+    },
+    {
+      "epoch": 0.2194891201513718,
+      "grad_norm": 3.5390021800994873,
+      "learning_rate": 1.9814951751170087e-05,
+      "loss": 1.3558,
+      "step": 290
+    },
+    {
+      "epoch": 0.22024597918637653,
+      "grad_norm": 3.37929630279541,
+      "learning_rate": 1.9813411293759618e-05,
+      "loss": 1.3236,
+      "step": 291
+    },
+    {
+      "epoch": 0.22100283822138128,
+      "grad_norm": 3.255699872970581,
+      "learning_rate": 1.9811864511418467e-05,
+      "loss": 1.3245,
+      "step": 292
+    },
+    {
+      "epoch": 0.221759697256386,
+      "grad_norm": 3.8194658756256104,
+      "learning_rate": 1.981031140514367e-05,
+      "loss": 1.3381,
+      "step": 293
+    },
+    {
+      "epoch": 0.22251655629139072,
+      "grad_norm": 3.8124804496765137,
+      "learning_rate": 1.9808751975936344e-05,
+      "loss": 1.3006,
+      "step": 294
+    },
+    {
+      "epoch": 0.22327341532639547,
+      "grad_norm": 3.637120246887207,
+      "learning_rate": 1.980718622480166e-05,
+      "loss": 1.3411,
+      "step": 295
+    },
+    {
+      "epoch": 0.2240302743614002,
+      "grad_norm": 3.8235883712768555,
+      "learning_rate": 1.9805614152748887e-05,
+      "loss": 1.3285,
+      "step": 296
+    },
+    {
+      "epoch": 0.2247871333964049,
+      "grad_norm": 3.892608642578125,
+      "learning_rate": 1.980403576079135e-05,
+      "loss": 1.3015,
+      "step": 297
+    },
+    {
+      "epoch": 0.22554399243140966,
+      "grad_norm": 3.9942359924316406,
+      "learning_rate": 1.9802451049946468e-05,
+      "loss": 1.3404,
+      "step": 298
+    },
+    {
+      "epoch": 0.22630085146641438,
+      "grad_norm": 3.8982861042022705,
+      "learning_rate": 1.9800860021235708e-05,
+      "loss": 1.3194,
+      "step": 299
+    },
+    {
+      "epoch": 0.22705771050141912,
+      "grad_norm": 4.402480125427246,
+      "learning_rate": 1.979926267568463e-05,
+      "loss": 1.3383,
+      "step": 300
+    },
+    {
+      "epoch": 0.22781456953642384,
+      "grad_norm": 3.9718708992004395,
+      "learning_rate": 1.979765901432286e-05,
+      "loss": 1.301,
+      "step": 301
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 3.8678481578826904,
+      "learning_rate": 1.979604903818409e-05,
+      "loss": 1.3269,
+      "step": 302
+    },
+    {
+      "epoch": 0.2293282876064333,
+      "grad_norm": 3.352957010269165,
+      "learning_rate": 1.979443274830609e-05,
+      "loss": 1.3332,
+      "step": 303
+    },
+    {
+      "epoch": 0.23008514664143803,
+      "grad_norm": 3.937535524368286,
+      "learning_rate": 1.9792810145730696e-05,
+      "loss": 1.3464,
+      "step": 304
+    },
+    {
+      "epoch": 0.23084200567644275,
+      "grad_norm": 4.081162929534912,
+      "learning_rate": 1.9791181231503804e-05,
+      "loss": 1.327,
+      "step": 305
+    },
+    {
+      "epoch": 0.2315988647114475,
+      "grad_norm": 3.5600180625915527,
+      "learning_rate": 1.97895460066754e-05,
+      "loss": 1.3536,
+      "step": 306
+    },
+    {
+      "epoch": 0.23235572374645222,
+      "grad_norm": 3.9321706295013428,
+      "learning_rate": 1.9787904472299512e-05,
+      "loss": 1.2999,
+      "step": 307
+    },
+    {
+      "epoch": 0.23311258278145697,
+      "grad_norm": 4.384609699249268,
+      "learning_rate": 1.978625662943426e-05,
+      "loss": 1.3461,
+      "step": 308
+    },
+    {
+      "epoch": 0.2338694418164617,
+      "grad_norm": 4.421790599822998,
+      "learning_rate": 1.978460247914181e-05,
+      "loss": 1.3324,
+      "step": 309
+    },
+    {
+      "epoch": 0.2346263008514664,
+      "grad_norm": 4.101651191711426,
+      "learning_rate": 1.9782942022488404e-05,
+      "loss": 1.2738,
+      "step": 310
+    },
+    {
+      "epoch": 0.23538315988647115,
+      "grad_norm": 4.219285488128662,
+      "learning_rate": 1.978127526054435e-05,
+      "loss": 1.3519,
+      "step": 311
+    },
+    {
+      "epoch": 0.23614001892147587,
+      "grad_norm": 3.5981838703155518,
+      "learning_rate": 1.9779602194384014e-05,
+      "loss": 1.3546,
+      "step": 312
+    },
+    {
+      "epoch": 0.2368968779564806,
+      "grad_norm": 3.758359432220459,
+      "learning_rate": 1.9777922825085835e-05,
+      "loss": 1.3264,
+      "step": 313
+    },
+    {
+      "epoch": 0.23765373699148534,
+      "grad_norm": 3.7645103931427,
+      "learning_rate": 1.97762371537323e-05,
+      "loss": 1.3135,
+      "step": 314
+    },
+    {
+      "epoch": 0.23841059602649006,
+      "grad_norm": 3.3905699253082275,
+      "learning_rate": 1.9774545181409973e-05,
+      "loss": 1.2848,
+      "step": 315
+    },
+    {
+      "epoch": 0.2391674550614948,
+      "grad_norm": 3.6732635498046875,
+      "learning_rate": 1.9772846909209473e-05,
+      "loss": 1.3487,
+      "step": 316
+    },
+    {
+      "epoch": 0.23992431409649953,
+      "grad_norm": 3.8122737407684326,
+      "learning_rate": 1.9771142338225476e-05,
+      "loss": 1.333,
+      "step": 317
+    },
+    {
+      "epoch": 0.24068117313150425,
+      "grad_norm": 4.025964260101318,
+      "learning_rate": 1.9769431469556728e-05,
+      "loss": 1.3431,
+      "step": 318
+    },
+    {
+      "epoch": 0.241438032166509,
+      "grad_norm": 3.054323196411133,
+      "learning_rate": 1.9767714304306024e-05,
+      "loss": 1.3279,
+      "step": 319
+    },
+    {
+      "epoch": 0.24219489120151372,
+      "grad_norm": 4.698709964752197,
+      "learning_rate": 1.9765990843580227e-05,
+      "loss": 1.3209,
+      "step": 320
+    },
+    {
+      "epoch": 0.24295175023651844,
+      "grad_norm": 3.341327428817749,
+      "learning_rate": 1.976426108849025e-05,
+      "loss": 1.3424,
+      "step": 321
+    },
+    {
+      "epoch": 0.24370860927152319,
+      "grad_norm": 3.9361190795898438,
+      "learning_rate": 1.9762525040151074e-05,
+      "loss": 1.3083,
+      "step": 322
+    },
+    {
+      "epoch": 0.2444654683065279,
+      "grad_norm": 3.340085506439209,
+      "learning_rate": 1.9760782699681716e-05,
+      "loss": 1.3358,
+      "step": 323
+    },
+    {
+      "epoch": 0.24522232734153265,
+      "grad_norm": 3.044618606567383,
+      "learning_rate": 1.9759034068205273e-05,
+      "loss": 1.3099,
+      "step": 324
+    },
+    {
+      "epoch": 0.24597918637653737,
+      "grad_norm": 3.619760274887085,
+      "learning_rate": 1.9757279146848883e-05,
+      "loss": 1.3455,
+      "step": 325
+    },
+    {
+      "epoch": 0.2467360454115421,
+      "grad_norm": 3.7121100425720215,
+      "learning_rate": 1.975551793674374e-05,
+      "loss": 1.3106,
+      "step": 326
+    },
+    {
+      "epoch": 0.24749290444654684,
+      "grad_norm": 3.5931692123413086,
+      "learning_rate": 1.9753750439025095e-05,
+      "loss": 1.2905,
+      "step": 327
+    },
+    {
+      "epoch": 0.24824976348155156,
+      "grad_norm": 3.603030204772949,
+      "learning_rate": 1.975197665483225e-05,
+      "loss": 1.3319,
+      "step": 328
+    },
+    {
+      "epoch": 0.24900662251655628,
+      "grad_norm": 3.6277918815612793,
+      "learning_rate": 1.9750196585308564e-05,
+      "loss": 1.3393,
+      "step": 329
+    },
+    {
+      "epoch": 0.24976348155156103,
+      "grad_norm": 3.5887362957000732,
+      "learning_rate": 1.974841023160143e-05,
+      "loss": 1.3866,
+      "step": 330
+    },
+    {
+      "epoch": 0.25052034058656575,
+      "grad_norm": 3.4283299446105957,
+      "learning_rate": 1.974661759486232e-05,
+      "loss": 1.329,
+      "step": 331
+    },
+    {
+      "epoch": 0.25127719962157047,
+      "grad_norm": 3.7355992794036865,
+      "learning_rate": 1.9744818676246724e-05,
+      "loss": 1.3129,
+      "step": 332
+    },
+    {
+      "epoch": 0.2520340586565752,
+      "grad_norm": 3.726663589477539,
+      "learning_rate": 1.974301347691421e-05,
+      "loss": 1.3665,
+      "step": 333
+    },
+    {
+      "epoch": 0.25279091769157996,
+      "grad_norm": 3.93129825592041,
+      "learning_rate": 1.9741201998028377e-05,
+      "loss": 1.3876,
+      "step": 334
+    },
+    {
+      "epoch": 0.2535477767265847,
+      "grad_norm": 3.588931083679199,
+      "learning_rate": 1.9739384240756873e-05,
+      "loss": 1.3715,
+      "step": 335
+    },
+    {
+      "epoch": 0.2543046357615894,
+      "grad_norm": 3.4406232833862305,
+      "learning_rate": 1.9737560206271404e-05,
+      "loss": 1.3013,
+      "step": 336
+    },
+    {
+      "epoch": 0.2550614947965941,
+      "grad_norm": 3.481201171875,
+      "learning_rate": 1.9735729895747714e-05,
+      "loss": 1.3625,
+      "step": 337
+    },
+    {
+      "epoch": 0.25581835383159884,
+      "grad_norm": 3.7452211380004883,
+      "learning_rate": 1.973389331036559e-05,
+      "loss": 1.3452,
+      "step": 338
+    },
+    {
+      "epoch": 0.2565752128666036,
+      "grad_norm": 3.8469581604003906,
+      "learning_rate": 1.973205045130887e-05,
+      "loss": 1.3824,
+      "step": 339
+    },
+    {
+      "epoch": 0.25733207190160834,
+      "grad_norm": 3.252890110015869,
+      "learning_rate": 1.9730201319765423e-05,
+      "loss": 1.311,
+      "step": 340
+    },
+    {
+      "epoch": 0.25808893093661306,
+      "grad_norm": 3.9583048820495605,
+      "learning_rate": 1.9728345916927187e-05,
+      "loss": 1.3244,
+      "step": 341
+    },
+    {
+      "epoch": 0.2588457899716178,
+      "grad_norm": 3.6613519191741943,
+      "learning_rate": 1.9726484243990115e-05,
+      "loss": 1.3539,
+      "step": 342
+    },
+    {
+      "epoch": 0.2596026490066225,
+      "grad_norm": 3.4180917739868164,
+      "learning_rate": 1.9724616302154218e-05,
+      "loss": 1.3353,
+      "step": 343
+    },
+    {
+      "epoch": 0.2603595080416273,
+      "grad_norm": 3.7470951080322266,
+      "learning_rate": 1.9722742092623536e-05,
+      "loss": 1.2864,
+      "step": 344
+    },
+    {
+      "epoch": 0.261116367076632,
+      "grad_norm": 4.141618251800537,
+      "learning_rate": 1.9720861616606165e-05,
+      "loss": 1.3486,
+      "step": 345
+    },
+    {
+      "epoch": 0.2618732261116367,
+      "grad_norm": 3.7161524295806885,
+      "learning_rate": 1.9718974875314226e-05,
+      "loss": 1.339,
+      "step": 346
+    },
+    {
+      "epoch": 0.26263008514664143,
+      "grad_norm": 4.011509895324707,
+      "learning_rate": 1.9717081869963887e-05,
+      "loss": 1.4027,
+      "step": 347
+    },
+    {
+      "epoch": 0.26338694418164615,
+      "grad_norm": 4.976902008056641,
+      "learning_rate": 1.9715182601775348e-05,
+      "loss": 1.3078,
+      "step": 348
+    },
+    {
+      "epoch": 0.2641438032166509,
+      "grad_norm": 3.8435733318328857,
+      "learning_rate": 1.9713277071972844e-05,
+      "loss": 1.3013,
+      "step": 349
+    },
+    {
+      "epoch": 0.26490066225165565,
+      "grad_norm": 3.3969762325286865,
+      "learning_rate": 1.971136528178466e-05,
+      "loss": 1.3078,
+      "step": 350
+    },
+    {
+      "epoch": 0.26565752128666037,
+      "grad_norm": 4.123608112335205,
+      "learning_rate": 1.9709447232443096e-05,
+      "loss": 1.3476,
+      "step": 351
+    },
+    {
+      "epoch": 0.2664143803216651,
+      "grad_norm": 3.974820137023926,
+      "learning_rate": 1.9707522925184507e-05,
+      "loss": 1.377,
+      "step": 352
+    },
+    {
+      "epoch": 0.2671712393566698,
+      "grad_norm": 4.08565616607666,
+      "learning_rate": 1.9705592361249267e-05,
+      "loss": 1.3559,
+      "step": 353
+    },
+    {
+      "epoch": 0.26792809839167453,
+      "grad_norm": 3.7338943481445312,
+      "learning_rate": 1.970365554188179e-05,
+      "loss": 1.2845,
+      "step": 354
+    },
+    {
+      "epoch": 0.2686849574266793,
+      "grad_norm": 3.806567430496216,
+      "learning_rate": 1.9701712468330518e-05,
+      "loss": 1.4283,
+      "step": 355
+    },
+    {
+      "epoch": 0.269441816461684,
+      "grad_norm": 3.4662294387817383,
+      "learning_rate": 1.9699763141847928e-05,
+      "loss": 1.3068,
+      "step": 356
+    },
+    {
+      "epoch": 0.27019867549668874,
+      "grad_norm": 3.5118749141693115,
+      "learning_rate": 1.9697807563690522e-05,
+      "loss": 1.266,
+      "step": 357
+    },
+    {
+      "epoch": 0.27095553453169346,
+      "grad_norm": 4.166219711303711,
+      "learning_rate": 1.969584573511885e-05,
+      "loss": 1.3355,
+      "step": 358
+    },
+    {
+      "epoch": 0.2717123935666982,
+      "grad_norm": 3.828523635864258,
+      "learning_rate": 1.969387765739746e-05,
+      "loss": 1.2712,
+      "step": 359
+    },
+    {
+      "epoch": 0.27246925260170296,
+      "grad_norm": 3.8785219192504883,
+      "learning_rate": 1.969190333179495e-05,
+      "loss": 1.2761,
+      "step": 360
+    },
+    {
+      "epoch": 0.2732261116367077,
+      "grad_norm": 3.772268056869507,
+      "learning_rate": 1.9689922759583947e-05,
+      "loss": 1.372,
+      "step": 361
+    },
+    {
+      "epoch": 0.2739829706717124,
+      "grad_norm": 3.7379493713378906,
+      "learning_rate": 1.968793594204109e-05,
+      "loss": 1.2843,
+      "step": 362
+    },
+    {
+      "epoch": 0.2747398297067171,
+      "grad_norm": 4.294455051422119,
+      "learning_rate": 1.9685942880447054e-05,
+      "loss": 1.3069,
+      "step": 363
+    },
+    {
+      "epoch": 0.27549668874172184,
+      "grad_norm": 4.1428728103637695,
+      "learning_rate": 1.9683943576086536e-05,
+      "loss": 1.366,
+      "step": 364
+    },
+    {
+      "epoch": 0.27625354777672656,
+      "grad_norm": 3.9030814170837402,
+      "learning_rate": 1.9681938030248257e-05,
+      "loss": 1.342,
+      "step": 365
+    },
+    {
+      "epoch": 0.27701040681173134,
+      "grad_norm": 4.4898681640625,
+      "learning_rate": 1.967992624422496e-05,
+      "loss": 1.2735,
+      "step": 366
+    },
+    {
+      "epoch": 0.27776726584673606,
+      "grad_norm": 4.548799514770508,
+      "learning_rate": 1.9677908219313414e-05,
+      "loss": 1.3589,
+      "step": 367
+    },
+    {
+      "epoch": 0.2785241248817408,
+      "grad_norm": 4.4808478355407715,
+      "learning_rate": 1.9675883956814403e-05,
+      "loss": 1.373,
+      "step": 368
+    },
+    {
+      "epoch": 0.2792809839167455,
+      "grad_norm": 4.146103858947754,
+      "learning_rate": 1.967385345803274e-05,
+      "loss": 1.2748,
+      "step": 369
+    },
+    {
+      "epoch": 0.2800378429517502,
+      "grad_norm": 5.006552696228027,
+      "learning_rate": 1.9671816724277254e-05,
+      "loss": 1.2852,
+      "step": 370
+    },
+    {
+      "epoch": 0.280794701986755,
+      "grad_norm": 4.279321670532227,
+      "learning_rate": 1.966977375686079e-05,
+      "loss": 1.3634,
+      "step": 371
+    },
+    {
+      "epoch": 0.2815515610217597,
+      "grad_norm": 5.318479537963867,
+      "learning_rate": 1.9667724557100214e-05,
+      "loss": 1.3184,
+      "step": 372
+    },
+    {
+      "epoch": 0.28230842005676443,
+      "grad_norm": 4.354931354522705,
+      "learning_rate": 1.966566912631641e-05,
+      "loss": 1.3018,
+      "step": 373
+    },
+    {
+      "epoch": 0.28306527909176915,
+      "grad_norm": 3.5126800537109375,
+      "learning_rate": 1.9663607465834275e-05,
+      "loss": 1.2811,
+      "step": 374
+    },
+    {
+      "epoch": 0.28382213812677387,
+      "grad_norm": 4.875300407409668,
+      "learning_rate": 1.9661539576982728e-05,
+      "loss": 1.3238,
+      "step": 375
+    },
+    {
+      "epoch": 0.28457899716177865,
+      "grad_norm": 4.699173450469971,
+      "learning_rate": 1.9659465461094692e-05,
+      "loss": 1.3223,
+      "step": 376
+    },
+    {
+      "epoch": 0.28533585619678337,
+      "grad_norm": 3.6528842449188232,
+      "learning_rate": 1.9657385119507118e-05,
+      "loss": 1.292,
+      "step": 377
+    },
+    {
+      "epoch": 0.2860927152317881,
+      "grad_norm": 3.849123239517212,
+      "learning_rate": 1.965529855356096e-05,
+      "loss": 1.3114,
+      "step": 378
+    },
+    {
+      "epoch": 0.2868495742667928,
+      "grad_norm": 3.7049927711486816,
+      "learning_rate": 1.9653205764601182e-05,
+      "loss": 1.3314,
+      "step": 379
+    },
+    {
+      "epoch": 0.2876064333017975,
+      "grad_norm": 4.335115909576416,
+      "learning_rate": 1.9651106753976768e-05,
+      "loss": 1.3719,
+      "step": 380
+    },
+    {
+      "epoch": 0.28836329233680225,
+      "grad_norm": 4.870954990386963,
+      "learning_rate": 1.964900152304071e-05,
+      "loss": 1.3264,
+      "step": 381
+    },
+    {
+      "epoch": 0.289120151371807,
+      "grad_norm": 4.583834648132324,
+      "learning_rate": 1.9646890073150005e-05,
+      "loss": 1.3743,
+      "step": 382
+    },
+    {
+      "epoch": 0.28987701040681174,
+      "grad_norm": 3.795956611633301,
+      "learning_rate": 1.964477240566566e-05,
+      "loss": 1.2997,
+      "step": 383
+    },
+    {
+      "epoch": 0.29063386944181646,
+      "grad_norm": 5.41873025894165,
+      "learning_rate": 1.9642648521952695e-05,
+      "loss": 1.3381,
+      "step": 384
+    },
+    {
+      "epoch": 0.2913907284768212,
+      "grad_norm": 4.2772393226623535,
+      "learning_rate": 1.9640518423380127e-05,
+      "loss": 1.3322,
+      "step": 385
+    },
+    {
+      "epoch": 0.2921475875118259,
+      "grad_norm": 10.241232872009277,
+      "learning_rate": 1.9638382111320996e-05,
+      "loss": 1.3249,
+      "step": 386
+    },
+    {
+      "epoch": 0.2929044465468307,
+      "grad_norm": 3.4204752445220947,
+      "learning_rate": 1.9636239587152323e-05,
+      "loss": 1.3295,
+      "step": 387
+    },
+    {
+      "epoch": 0.2936613055818354,
+      "grad_norm": 3.368516683578491,
+      "learning_rate": 1.9634090852255154e-05,
+      "loss": 1.3561,
+      "step": 388
+    },
+    {
+      "epoch": 0.2944181646168401,
+      "grad_norm": 3.5226809978485107,
+      "learning_rate": 1.9631935908014532e-05,
+      "loss": 1.3146,
+      "step": 389
+    },
+    {
+      "epoch": 0.29517502365184484,
+      "grad_norm": 3.446794271469116,
+      "learning_rate": 1.9629774755819495e-05,
+      "loss": 1.2973,
+      "step": 390
+    },
+    {
+      "epoch": 0.29593188268684956,
+      "grad_norm": 3.176982879638672,
+      "learning_rate": 1.9627607397063097e-05,
+      "loss": 1.3233,
+      "step": 391
+    },
+    {
+      "epoch": 0.29668874172185433,
+      "grad_norm": 3.0656180381774902,
+      "learning_rate": 1.9625433833142376e-05,
+      "loss": 1.3246,
+      "step": 392
+    },
+    {
+      "epoch": 0.29744560075685905,
+      "grad_norm": 3.4723055362701416,
+      "learning_rate": 1.9623254065458387e-05,
+      "loss": 1.3461,
+      "step": 393
+    },
+    {
+      "epoch": 0.29820245979186377,
+      "grad_norm": 3.2769827842712402,
+      "learning_rate": 1.962106809541616e-05,
+      "loss": 1.3242,
+      "step": 394
+    },
+    {
+      "epoch": 0.2989593188268685,
+      "grad_norm": 2.8769099712371826,
+      "learning_rate": 1.9618875924424756e-05,
+      "loss": 1.2548,
+      "step": 395
+    },
+    {
+      "epoch": 0.2997161778618732,
+      "grad_norm": 3.3671765327453613,
+      "learning_rate": 1.9616677553897204e-05,
+      "loss": 1.3241,
+      "step": 396
+    },
+    {
+      "epoch": 0.30047303689687793,
+      "grad_norm": 3.104637384414673,
+      "learning_rate": 1.9614472985250547e-05,
+      "loss": 1.3121,
+      "step": 397
+    },
+    {
+      "epoch": 0.3012298959318827,
+      "grad_norm": 3.6635613441467285,
+      "learning_rate": 1.9612262219905807e-05,
+      "loss": 1.3157,
+      "step": 398
+    },
+    {
+      "epoch": 0.3019867549668874,
+      "grad_norm": 3.4978229999542236,
+      "learning_rate": 1.9610045259288017e-05,
+      "loss": 1.3566,
+      "step": 399
+    },
+    {
+      "epoch": 0.30274361400189215,
+      "grad_norm": 3.084291458129883,
+      "learning_rate": 1.9607822104826198e-05,
+      "loss": 1.3157,
+      "step": 400
+    },
+    {
+      "epoch": 0.30350047303689687,
+      "grad_norm": 2.742034673690796,
+      "learning_rate": 1.9605592757953354e-05,
+      "loss": 1.2778,
+      "step": 401
+    },
+    {
+      "epoch": 0.3042573320719016,
+      "grad_norm": 2.89613938331604,
+      "learning_rate": 1.960335722010649e-05,
+      "loss": 1.3467,
+      "step": 402
+    },
+    {
+      "epoch": 0.30501419110690636,
+      "grad_norm": 3.3919119834899902,
+      "learning_rate": 1.9601115492726603e-05,
+      "loss": 1.3264,
+      "step": 403
+    },
+    {
+      "epoch": 0.3057710501419111,
+      "grad_norm": 2.5944290161132812,
+      "learning_rate": 1.9598867577258672e-05,
+      "loss": 1.3765,
+      "step": 404
+    },
+    {
+      "epoch": 0.3065279091769158,
+      "grad_norm": 2.6674866676330566,
+      "learning_rate": 1.9596613475151674e-05,
+      "loss": 1.3077,
+      "step": 405
+    },
+    {
+      "epoch": 0.3072847682119205,
+      "grad_norm": 2.9125916957855225,
+      "learning_rate": 1.9594353187858567e-05,
+      "loss": 1.3472,
+      "step": 406
+    },
+    {
+      "epoch": 0.30804162724692524,
+      "grad_norm": 2.746316909790039,
+      "learning_rate": 1.9592086716836292e-05,
+      "loss": 1.3137,
+      "step": 407
+    },
+    {
+      "epoch": 0.30879848628193,
+      "grad_norm": 3.5916221141815186,
+      "learning_rate": 1.958981406354579e-05,
+      "loss": 1.3181,
+      "step": 408
+    },
+    {
+      "epoch": 0.30955534531693474,
+      "grad_norm": 2.9677999019622803,
+      "learning_rate": 1.9587535229451973e-05,
+      "loss": 1.3094,
+      "step": 409
+    },
+    {
+      "epoch": 0.31031220435193946,
+      "grad_norm": 2.766179084777832,
+      "learning_rate": 1.9585250216023746e-05,
+      "loss": 1.3205,
+      "step": 410
+    },
+    {
+      "epoch": 0.3110690633869442,
+      "grad_norm": 3.019426107406616,
+      "learning_rate": 1.9582959024733992e-05,
+      "loss": 1.3053,
+      "step": 411
+    },
+    {
+      "epoch": 0.3118259224219489,
+      "grad_norm": 3.580401659011841,
+      "learning_rate": 1.9580661657059582e-05,
+      "loss": 1.3685,
+      "step": 412
+    },
+    {
+      "epoch": 0.3125827814569536,
+      "grad_norm": 3.2559759616851807,
+      "learning_rate": 1.957835811448136e-05,
+      "loss": 1.2975,
+      "step": 413
+    },
+    {
+      "epoch": 0.3133396404919584,
+      "grad_norm": 3.185425281524658,
+      "learning_rate": 1.957604839848415e-05,
+      "loss": 1.3391,
+      "step": 414
+    },
+    {
+      "epoch": 0.3140964995269631,
+      "grad_norm": 3.2222900390625,
+      "learning_rate": 1.9573732510556772e-05,
+      "loss": 1.2233,
+      "step": 415
+    },
+    {
+      "epoch": 0.31485335856196783,
+      "grad_norm": 3.3176467418670654,
+      "learning_rate": 1.9571410452192003e-05,
+      "loss": 1.32,
+      "step": 416
+    },
+    {
+      "epoch": 0.31561021759697255,
+      "grad_norm": 2.996213912963867,
+      "learning_rate": 1.9569082224886607e-05,
+      "loss": 1.3158,
+      "step": 417
+    },
+    {
+      "epoch": 0.3163670766319773,
+      "grad_norm": 2.757145881652832,
+      "learning_rate": 1.9566747830141327e-05,
+      "loss": 1.2747,
+      "step": 418
+    },
+    {
+      "epoch": 0.31712393566698205,
+      "grad_norm": 3.0630686283111572,
+      "learning_rate": 1.9564407269460873e-05,
+      "loss": 1.2863,
+      "step": 419
+    },
+    {
+      "epoch": 0.31788079470198677,
+      "grad_norm": 2.979710102081299,
+      "learning_rate": 1.956206054435394e-05,
+      "loss": 1.3017,
+      "step": 420
+    },
+    {
+      "epoch": 0.3186376537369915,
+      "grad_norm": 2.9305684566497803,
+      "learning_rate": 1.955970765633319e-05,
+      "loss": 1.2655,
+      "step": 421
+    },
+    {
+      "epoch": 0.3193945127719962,
+      "grad_norm": 3.2490427494049072,
+      "learning_rate": 1.955734860691526e-05,
+      "loss": 1.312,
+      "step": 422
+    },
+    {
+      "epoch": 0.32015137180700093,
+      "grad_norm": 2.647688388824463,
+      "learning_rate": 1.9554983397620754e-05,
+      "loss": 1.3009,
+      "step": 423
+    },
+    {
+      "epoch": 0.3209082308420057,
+      "grad_norm": 2.8015365600585938,
+      "learning_rate": 1.9552612029974246e-05,
+      "loss": 1.3069,
+      "step": 424
+    },
+    {
+      "epoch": 0.3216650898770104,
+      "grad_norm": 2.606043577194214,
+      "learning_rate": 1.9550234505504294e-05,
+      "loss": 1.2951,
+      "step": 425
+    },
+    {
+      "epoch": 0.32242194891201514,
+      "grad_norm": 2.9746274948120117,
+      "learning_rate": 1.9547850825743407e-05,
+      "loss": 1.2736,
+      "step": 426
+    },
+    {
+      "epoch": 0.32317880794701986,
+      "grad_norm": 3.0589208602905273,
+      "learning_rate": 1.9545460992228074e-05,
+      "loss": 1.3242,
+      "step": 427
+    },
+    {
+      "epoch": 0.3239356669820246,
+      "grad_norm": 3.041224956512451,
+      "learning_rate": 1.954306500649874e-05,
+      "loss": 1.3397,
+      "step": 428
+    },
+    {
+      "epoch": 0.3246925260170293,
+      "grad_norm": 2.700326681137085,
+      "learning_rate": 1.954066287009982e-05,
+      "loss": 1.2822,
+      "step": 429
+    },
+    {
+      "epoch": 0.3254493850520341,
+      "grad_norm": 2.7489256858825684,
+      "learning_rate": 1.95382545845797e-05,
+      "loss": 1.3056,
+      "step": 430
+    },
+    {
+      "epoch": 0.3262062440870388,
+      "grad_norm": 3.0966339111328125,
+      "learning_rate": 1.953584015149072e-05,
+      "loss": 1.3316,
+      "step": 431
+    },
+    {
+      "epoch": 0.3269631031220435,
+      "grad_norm": 2.661102533340454,
+      "learning_rate": 1.9533419572389186e-05,
+      "loss": 1.3017,
+      "step": 432
+    },
+    {
+      "epoch": 0.32771996215704824,
+      "grad_norm": 3.1965274810791016,
+      "learning_rate": 1.9530992848835367e-05,
+      "loss": 1.2975,
+      "step": 433
+    },
+    {
+      "epoch": 0.32847682119205296,
+      "grad_norm": 3.0282115936279297,
+      "learning_rate": 1.9528559982393497e-05,
+      "loss": 1.3261,
+      "step": 434
+    },
+    {
+      "epoch": 0.32923368022705773,
+      "grad_norm": 2.794201374053955,
+      "learning_rate": 1.9526120974631763e-05,
+      "loss": 1.3363,
+      "step": 435
+    },
+    {
+      "epoch": 0.32999053926206245,
+      "grad_norm": 2.8009607791900635,
+      "learning_rate": 1.9523675827122305e-05,
+      "loss": 1.2738,
+      "step": 436
+    },
+    {
+      "epoch": 0.3307473982970672,
+      "grad_norm": 3.1605050563812256,
+      "learning_rate": 1.952122454144123e-05,
+      "loss": 1.3242,
+      "step": 437
+    },
+    {
+      "epoch": 0.3315042573320719,
+      "grad_norm": 2.7758185863494873,
+      "learning_rate": 1.9518767119168608e-05,
+      "loss": 1.2546,
+      "step": 438
+    },
+    {
+      "epoch": 0.3322611163670766,
+      "grad_norm": 3.3435556888580322,
+      "learning_rate": 1.9516303561888446e-05,
+      "loss": 1.2966,
+      "step": 439
+    },
+    {
+      "epoch": 0.3330179754020814,
+      "grad_norm": 3.4312620162963867,
+      "learning_rate": 1.9513833871188724e-05,
+      "loss": 1.328,
+      "step": 440
+    },
+    {
+      "epoch": 0.3337748344370861,
+      "grad_norm": 3.4291491508483887,
+      "learning_rate": 1.951135804866136e-05,
+      "loss": 1.2927,
+      "step": 441
+    },
+    {
+      "epoch": 0.33453169347209083,
+      "grad_norm": 2.797574281692505,
+      "learning_rate": 1.9508876095902236e-05,
+      "loss": 1.3218,
+      "step": 442
+    },
+    {
+      "epoch": 0.33528855250709555,
+      "grad_norm": 3.1859307289123535,
+      "learning_rate": 1.9506388014511176e-05,
+      "loss": 1.2827,
+      "step": 443
+    },
+    {
+      "epoch": 0.33604541154210027,
+      "grad_norm": 3.4026360511779785,
+      "learning_rate": 1.950389380609196e-05,
+      "loss": 1.2879,
+      "step": 444
+    },
+    {
+      "epoch": 0.336802270577105,
+      "grad_norm": 3.2964580059051514,
+      "learning_rate": 1.9501393472252324e-05,
+      "loss": 1.2976,
+      "step": 445
+    },
+    {
+      "epoch": 0.33755912961210977,
+      "grad_norm": 3.227969169616699,
+      "learning_rate": 1.9498887014603937e-05,
+      "loss": 1.3191,
+      "step": 446
+    },
+    {
+      "epoch": 0.3383159886471145,
+      "grad_norm": 4.118795871734619,
+      "learning_rate": 1.949637443476243e-05,
+      "loss": 1.3112,
+      "step": 447
+    },
+    {
+      "epoch": 0.3390728476821192,
+      "grad_norm": 3.7260451316833496,
+      "learning_rate": 1.9493855734347367e-05,
+      "loss": 1.2836,
+      "step": 448
+    },
+    {
+      "epoch": 0.3398297067171239,
+      "grad_norm": 3.0048820972442627,
+      "learning_rate": 1.9491330914982265e-05,
+      "loss": 1.3106,
+      "step": 449
+    },
+    {
+      "epoch": 0.34058656575212864,
+      "grad_norm": 2.7483198642730713,
+      "learning_rate": 1.9488799978294586e-05,
+      "loss": 1.338,
+      "step": 450
+    },
+    {
+      "epoch": 0.3413434247871334,
+      "grad_norm": 3.021895170211792,
+      "learning_rate": 1.9486262925915736e-05,
+      "loss": 1.2931,
+      "step": 451
+    },
+    {
+      "epoch": 0.34210028382213814,
+      "grad_norm": 2.793663740158081,
+      "learning_rate": 1.948371975948106e-05,
+      "loss": 1.2895,
+      "step": 452
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 3.6046817302703857,
+      "learning_rate": 1.9481170480629835e-05,
+      "loss": 1.326,
+      "step": 453
+    },
+    {
+      "epoch": 0.3436140018921476,
+      "grad_norm": 2.8959131240844727,
+      "learning_rate": 1.9478615091005296e-05,
+      "loss": 1.3018,
+      "step": 454
+    },
+    {
+      "epoch": 0.3443708609271523,
+      "grad_norm": 2.869874954223633,
+      "learning_rate": 1.9476053592254608e-05,
+      "loss": 1.3181,
+      "step": 455
+    },
+    {
+      "epoch": 0.3451277199621571,
+      "grad_norm": 2.9448678493499756,
+      "learning_rate": 1.947348598602887e-05,
+      "loss": 1.2992,
+      "step": 456
+    },
+    {
+      "epoch": 0.3458845789971618,
+      "grad_norm": 2.8842031955718994,
+      "learning_rate": 1.9470912273983123e-05,
+      "loss": 1.3297,
+      "step": 457
+    },
+    {
+      "epoch": 0.3466414380321665,
+      "grad_norm": 3.329968214035034,
+      "learning_rate": 1.946833245777635e-05,
+      "loss": 1.3074,
+      "step": 458
+    },
+    {
+      "epoch": 0.34739829706717124,
+      "grad_norm": 2.8565642833709717,
+      "learning_rate": 1.9465746539071447e-05,
+      "loss": 1.3204,
+      "step": 459
+    },
+    {
+      "epoch": 0.34815515610217596,
+      "grad_norm": 3.0529487133026123,
+      "learning_rate": 1.946315451953527e-05,
+      "loss": 1.3249,
+      "step": 460
+    },
+    {
+      "epoch": 0.3489120151371807,
+      "grad_norm": 2.988011360168457,
+      "learning_rate": 1.946055640083859e-05,
+      "loss": 1.2612,
+      "step": 461
+    },
+    {
+      "epoch": 0.34966887417218545,
+      "grad_norm": 3.3266758918762207,
+      "learning_rate": 1.945795218465611e-05,
+      "loss": 1.3283,
+      "step": 462
+    },
+    {
+      "epoch": 0.35042573320719017,
+      "grad_norm": 3.2849862575531006,
+      "learning_rate": 1.945534187266648e-05,
+      "loss": 1.3476,
+      "step": 463
+    },
+    {
+      "epoch": 0.3511825922421949,
+      "grad_norm": 2.831113576889038,
+      "learning_rate": 1.945272546655226e-05,
+      "loss": 1.2726,
+      "step": 464
+    },
+    {
+      "epoch": 0.3519394512771996,
+      "grad_norm": 3.232224464416504,
+      "learning_rate": 1.9450102967999946e-05,
+      "loss": 1.3362,
+      "step": 465
+    },
+    {
+      "epoch": 0.35269631031220433,
+      "grad_norm": 3.704671621322632,
+      "learning_rate": 1.944747437869996e-05,
+      "loss": 1.3011,
+      "step": 466
+    },
+    {
+      "epoch": 0.3534531693472091,
+      "grad_norm": 2.6540513038635254,
+      "learning_rate": 1.944483970034665e-05,
+      "loss": 1.3268,
+      "step": 467
+    },
+    {
+      "epoch": 0.3542100283822138,
+      "grad_norm": 3.856849431991577,
+      "learning_rate": 1.944219893463829e-05,
+      "loss": 1.2762,
+      "step": 468
+    },
+    {
+      "epoch": 0.35496688741721855,
+      "grad_norm": 2.809225082397461,
+      "learning_rate": 1.943955208327708e-05,
+      "loss": 1.2515,
+      "step": 469
+    },
+    {
+      "epoch": 0.35572374645222327,
+      "grad_norm": 3.271754503250122,
+      "learning_rate": 1.943689914796914e-05,
+      "loss": 1.313,
+      "step": 470
+    },
+    {
+      "epoch": 0.356480605487228,
+      "grad_norm": 2.872096061706543,
+      "learning_rate": 1.9434240130424504e-05,
+      "loss": 1.2762,
+      "step": 471
+    },
+    {
+      "epoch": 0.35723746452223276,
+      "grad_norm": 2.9466817378997803,
+      "learning_rate": 1.9431575032357147e-05,
+      "loss": 1.3123,
+      "step": 472
+    },
+    {
+      "epoch": 0.3579943235572375,
+      "grad_norm": 3.358745813369751,
+      "learning_rate": 1.9428903855484938e-05,
+      "loss": 1.2684,
+      "step": 473
+    },
+    {
+      "epoch": 0.3587511825922422,
+      "grad_norm": 3.3290534019470215,
+      "learning_rate": 1.9426226601529685e-05,
+      "loss": 1.321,
+      "step": 474
+    },
+    {
+      "epoch": 0.3595080416272469,
+      "grad_norm": 3.1677582263946533,
+      "learning_rate": 1.9423543272217103e-05,
+      "loss": 1.2994,
+      "step": 475
+    },
+    {
+      "epoch": 0.36026490066225164,
+      "grad_norm": 3.891291618347168,
+      "learning_rate": 1.9420853869276822e-05,
+      "loss": 1.2783,
+      "step": 476
+    },
+    {
+      "epoch": 0.36102175969725636,
+      "grad_norm": 3.3545546531677246,
+      "learning_rate": 1.9418158394442395e-05,
+      "loss": 1.2985,
+      "step": 477
+    },
+    {
+      "epoch": 0.36177861873226114,
+      "grad_norm": 3.187551498413086,
+      "learning_rate": 1.941545684945128e-05,
+      "loss": 1.3401,
+      "step": 478
+    },
+    {
+      "epoch": 0.36253547776726586,
+      "grad_norm": 3.063565969467163,
+      "learning_rate": 1.9412749236044855e-05,
+      "loss": 1.2574,
+      "step": 479
+    },
+    {
+      "epoch": 0.3632923368022706,
+      "grad_norm": 3.0356266498565674,
+      "learning_rate": 1.9410035555968403e-05,
+      "loss": 1.2734,
+      "step": 480
+    },
+    {
+      "epoch": 0.3640491958372753,
+      "grad_norm": 4.256435871124268,
+      "learning_rate": 1.9407315810971123e-05,
+      "loss": 1.2623,
+      "step": 481
+    },
+    {
+      "epoch": 0.36480605487228,
+      "grad_norm": 3.298546075820923,
+      "learning_rate": 1.9404590002806122e-05,
+      "loss": 1.3079,
+      "step": 482
+    },
+    {
+      "epoch": 0.3655629139072848,
+      "grad_norm": 3.06703782081604,
+      "learning_rate": 1.9401858133230412e-05,
+      "loss": 1.347,
+      "step": 483
+    },
+    {
+      "epoch": 0.3663197729422895,
+      "grad_norm": 3.244100332260132,
+      "learning_rate": 1.9399120204004917e-05,
+      "loss": 1.298,
+      "step": 484
+    },
+    {
+      "epoch": 0.36707663197729423,
+      "grad_norm": 2.7238996028900146,
+      "learning_rate": 1.9396376216894462e-05,
+      "loss": 1.2434,
+      "step": 485
+    },
+    {
+      "epoch": 0.36783349101229895,
+      "grad_norm": 3.0345072746276855,
+      "learning_rate": 1.939362617366778e-05,
+      "loss": 1.3634,
+      "step": 486
+    },
+    {
+      "epoch": 0.36859035004730367,
+      "grad_norm": 3.2847676277160645,
+      "learning_rate": 1.9390870076097507e-05,
+      "loss": 1.3037,
+      "step": 487
+    },
+    {
+      "epoch": 0.36934720908230845,
+      "grad_norm": 2.8988196849823,
+      "learning_rate": 1.9388107925960183e-05,
+      "loss": 1.3137,
+      "step": 488
+    },
+    {
+      "epoch": 0.37010406811731317,
+      "grad_norm": 3.4510462284088135,
+      "learning_rate": 1.9385339725036244e-05,
+      "loss": 1.3042,
+      "step": 489
+    },
+    {
+      "epoch": 0.3708609271523179,
+      "grad_norm": 2.7349674701690674,
+      "learning_rate": 1.938256547511003e-05,
+      "loss": 1.3104,
+      "step": 490
+    },
+    {
+      "epoch": 0.3716177861873226,
+      "grad_norm": 2.6827545166015625,
+      "learning_rate": 1.9379785177969787e-05,
+      "loss": 1.3312,
+      "step": 491
+    },
+    {
+      "epoch": 0.3723746452223273,
+      "grad_norm": 2.871415138244629,
+      "learning_rate": 1.937699883540765e-05,
+      "loss": 1.2999,
+      "step": 492
+    },
+    {
+      "epoch": 0.37313150425733205,
+      "grad_norm": 3.261521339416504,
+      "learning_rate": 1.9374206449219646e-05,
+      "loss": 1.3027,
+      "step": 493
+    },
+    {
+      "epoch": 0.3738883632923368,
+      "grad_norm": 2.7188801765441895,
+      "learning_rate": 1.9371408021205708e-05,
+      "loss": 1.2688,
+      "step": 494
+    },
+    {
+      "epoch": 0.37464522232734154,
+      "grad_norm": 2.762587070465088,
+      "learning_rate": 1.936860355316967e-05,
+      "loss": 1.316,
+      "step": 495
+    },
+    {
+      "epoch": 0.37540208136234626,
+      "grad_norm": 3.2157773971557617,
+      "learning_rate": 1.9365793046919233e-05,
+      "loss": 1.2818,
+      "step": 496
+    },
+    {
+      "epoch": 0.376158940397351,
+      "grad_norm": 2.720599889755249,
+      "learning_rate": 1.9362976504266017e-05,
+      "loss": 1.2767,
+      "step": 497
+    },
+    {
+      "epoch": 0.3769157994323557,
+      "grad_norm": 2.728111982345581,
+      "learning_rate": 1.936015392702552e-05,
+      "loss": 1.3292,
+      "step": 498
+    },
+    {
+      "epoch": 0.3776726584673605,
+      "grad_norm": 3.2255611419677734,
+      "learning_rate": 1.9357325317017127e-05,
+      "loss": 1.3165,
+      "step": 499
+    },
+    {
+      "epoch": 0.3784295175023652,
+      "grad_norm": 3.559377431869507,
+      "learning_rate": 1.935449067606413e-05,
+      "loss": 1.3284,
+      "step": 500
+    },
+    {
+      "epoch": 0.3791863765373699,
+      "grad_norm": 3.0584166049957275,
+      "learning_rate": 1.935165000599368e-05,
+      "loss": 1.2788,
+      "step": 501
+    },
+    {
+      "epoch": 0.37994323557237464,
+      "grad_norm": 3.422832727432251,
+      "learning_rate": 1.9348803308636836e-05,
+      "loss": 1.3315,
+      "step": 502
+    },
+    {
+      "epoch": 0.38070009460737936,
+      "grad_norm": 3.052250623703003,
+      "learning_rate": 1.9345950585828543e-05,
+      "loss": 1.2772,
+      "step": 503
+    },
+    {
+      "epoch": 0.38145695364238413,
+      "grad_norm": 3.164451837539673,
+      "learning_rate": 1.9343091839407608e-05,
+      "loss": 1.2796,
+      "step": 504
+    },
+    {
+      "epoch": 0.38221381267738885,
+      "grad_norm": 2.815291166305542,
+      "learning_rate": 1.9340227071216747e-05,
+      "loss": 1.2473,
+      "step": 505
+    },
+    {
+      "epoch": 0.3829706717123936,
+      "grad_norm": 2.953880548477173,
+      "learning_rate": 1.9337356283102543e-05,
+      "loss": 1.299,
+      "step": 506
+    },
+    {
+      "epoch": 0.3837275307473983,
+      "grad_norm": 2.8606905937194824,
+      "learning_rate": 1.9334479476915462e-05,
+      "loss": 1.3075,
+      "step": 507
+    },
+    {
+      "epoch": 0.384484389782403,
+      "grad_norm": 2.8410329818725586,
+      "learning_rate": 1.9331596654509848e-05,
+      "loss": 1.3377,
+      "step": 508
+    },
+    {
+      "epoch": 0.38524124881740773,
+      "grad_norm": 3.284508228302002,
+      "learning_rate": 1.9328707817743923e-05,
+      "loss": 1.2549,
+      "step": 509
+    },
+    {
+      "epoch": 0.3859981078524125,
+      "grad_norm": 2.892754554748535,
+      "learning_rate": 1.9325812968479793e-05,
+      "loss": 1.292,
+      "step": 510
+    },
+    {
+      "epoch": 0.38675496688741723,
+      "grad_norm": 3.1711387634277344,
+      "learning_rate": 1.932291210858343e-05,
+      "loss": 1.2258,
+      "step": 511
+    },
+    {
+      "epoch": 0.38751182592242195,
+      "grad_norm": 2.6686558723449707,
+      "learning_rate": 1.932000523992468e-05,
+      "loss": 1.246,
+      "step": 512
+    },
+    {
+      "epoch": 0.38826868495742667,
+      "grad_norm": 2.947592258453369,
+      "learning_rate": 1.9317092364377273e-05,
+      "loss": 1.2544,
+      "step": 513
+    },
+    {
+      "epoch": 0.3890255439924314,
+      "grad_norm": 3.5232532024383545,
+      "learning_rate": 1.93141734838188e-05,
+      "loss": 1.3041,
+      "step": 514
+    },
+    {
+      "epoch": 0.38978240302743616,
+      "grad_norm": 3.5411503314971924,
+      "learning_rate": 1.931124860013073e-05,
+      "loss": 1.2852,
+      "step": 515
+    },
+    {
+      "epoch": 0.3905392620624409,
+      "grad_norm": 3.0337717533111572,
+      "learning_rate": 1.93083177151984e-05,
+      "loss": 1.2867,
+      "step": 516
+    },
+    {
+      "epoch": 0.3912961210974456,
+      "grad_norm": 3.082791805267334,
+      "learning_rate": 1.9305380830911002e-05,
+      "loss": 1.2981,
+      "step": 517
+    },
+    {
+      "epoch": 0.3920529801324503,
+      "grad_norm": 3.1100258827209473,
+      "learning_rate": 1.9302437949161622e-05,
+      "loss": 1.2645,
+      "step": 518
+    },
+    {
+      "epoch": 0.39280983916745504,
+      "grad_norm": 3.3867480754852295,
+      "learning_rate": 1.9299489071847185e-05,
+      "loss": 1.3555,
+      "step": 519
+    },
+    {
+      "epoch": 0.3935666982024598,
+      "grad_norm": 3.1625607013702393,
+      "learning_rate": 1.9296534200868504e-05,
+      "loss": 1.3111,
+      "step": 520
+    },
+    {
+      "epoch": 0.39432355723746454,
+      "grad_norm": 4.334456443786621,
+      "learning_rate": 1.929357333813023e-05,
+      "loss": 1.3102,
+      "step": 521
+    },
+    {
+      "epoch": 0.39508041627246926,
+      "grad_norm": 3.650447130203247,
+      "learning_rate": 1.9290606485540903e-05,
+      "loss": 1.3129,
+      "step": 522
+    },
+    {
+      "epoch": 0.395837275307474,
+      "grad_norm": 3.8689398765563965,
+      "learning_rate": 1.9287633645012898e-05,
+      "loss": 1.2974,
+      "step": 523
+    },
+    {
+      "epoch": 0.3965941343424787,
+      "grad_norm": 3.520089864730835,
+      "learning_rate": 1.9284654818462474e-05,
+      "loss": 1.291,
+      "step": 524
+    },
+    {
+      "epoch": 0.3973509933774834,
+      "grad_norm": 4.220740795135498,
+      "learning_rate": 1.9281670007809735e-05,
+      "loss": 1.3039,
+      "step": 525
+    },
+    {
+      "epoch": 0.3981078524124882,
+      "grad_norm": 3.871176242828369,
+      "learning_rate": 1.9278679214978637e-05,
+      "loss": 1.2682,
+      "step": 526
+    },
+    {
+      "epoch": 0.3988647114474929,
+      "grad_norm": 3.2446093559265137,
+      "learning_rate": 1.9275682441897007e-05,
+      "loss": 1.2866,
+      "step": 527
+    },
+    {
+      "epoch": 0.39962157048249763,
+      "grad_norm": 3.475529432296753,
+      "learning_rate": 1.9272679690496517e-05,
+      "loss": 1.344,
+      "step": 528
+    },
+    {
+      "epoch": 0.40037842951750235,
+      "grad_norm": 3.29640531539917,
+      "learning_rate": 1.9269670962712695e-05,
+      "loss": 1.3257,
+      "step": 529
+    },
+    {
+      "epoch": 0.4011352885525071,
+      "grad_norm": 3.43729305267334,
+      "learning_rate": 1.9266656260484925e-05,
+      "loss": 1.3504,
+      "step": 530
+    },
+    {
+      "epoch": 0.40189214758751185,
+      "grad_norm": 3.6663601398468018,
+      "learning_rate": 1.9263635585756424e-05,
+      "loss": 1.2738,
+      "step": 531
+    },
+    {
+      "epoch": 0.40264900662251657,
+      "grad_norm": 3.4716086387634277,
+      "learning_rate": 1.9260608940474293e-05,
+      "loss": 1.2997,
+      "step": 532
+    },
+    {
+      "epoch": 0.4034058656575213,
+      "grad_norm": 3.0576701164245605,
+      "learning_rate": 1.9257576326589448e-05,
+      "loss": 1.2958,
+      "step": 533
+    },
+    {
+      "epoch": 0.404162724692526,
+      "grad_norm": 3.7031450271606445,
+      "learning_rate": 1.9254537746056664e-05,
+      "loss": 1.2537,
+      "step": 534
+    },
+    {
+      "epoch": 0.40491958372753073,
+      "grad_norm": 3.070580005645752,
+      "learning_rate": 1.925149320083457e-05,
+      "loss": 1.3362,
+      "step": 535
+    },
+    {
+      "epoch": 0.4056764427625355,
+      "grad_norm": 3.241197347640991,
+      "learning_rate": 1.9248442692885634e-05,
+      "loss": 1.2984,
+      "step": 536
+    },
+    {
+      "epoch": 0.4064333017975402,
+      "grad_norm": 2.7833101749420166,
+      "learning_rate": 1.9245386224176162e-05,
+      "loss": 1.2589,
+      "step": 537
+    },
+    {
+      "epoch": 0.40719016083254495,
+      "grad_norm": 2.8053226470947266,
+      "learning_rate": 1.9242323796676313e-05,
+      "loss": 1.277,
+      "step": 538
+    },
+    {
+      "epoch": 0.40794701986754967,
+      "grad_norm": 3.119124412536621,
+      "learning_rate": 1.9239255412360075e-05,
+      "loss": 1.2516,
+      "step": 539
+    },
+    {
+      "epoch": 0.4087038789025544,
+      "grad_norm": 3.013762950897217,
+      "learning_rate": 1.923618107320529e-05,
+      "loss": 1.2988,
+      "step": 540
+    },
+    {
+      "epoch": 0.4094607379375591,
+      "grad_norm": 2.8327529430389404,
+      "learning_rate": 1.923310078119362e-05,
+      "loss": 1.2596,
+      "step": 541
+    },
+    {
+      "epoch": 0.4102175969725639,
+      "grad_norm": 2.7732462882995605,
+      "learning_rate": 1.9230014538310575e-05,
+      "loss": 1.2525,
+      "step": 542
+    },
+    {
+      "epoch": 0.4109744560075686,
+      "grad_norm": 2.984377145767212,
+      "learning_rate": 1.9226922346545513e-05,
+      "loss": 1.2688,
+      "step": 543
+    },
+    {
+      "epoch": 0.4117313150425733,
+      "grad_norm": 3.146101474761963,
+      "learning_rate": 1.92238242078916e-05,
+      "loss": 1.3291,
+      "step": 544
+    },
+    {
+      "epoch": 0.41248817407757804,
+      "grad_norm": 2.911142587661743,
+      "learning_rate": 1.9220720124345855e-05,
+      "loss": 1.2372,
+      "step": 545
+    },
+    {
+      "epoch": 0.41324503311258276,
+      "grad_norm": 3.006364345550537,
+      "learning_rate": 1.921761009790912e-05,
+      "loss": 1.2157,
+      "step": 546
+    },
+    {
+      "epoch": 0.41400189214758754,
+      "grad_norm": 2.9054133892059326,
+      "learning_rate": 1.9214494130586074e-05,
+      "loss": 1.3591,
+      "step": 547
+    },
+    {
+      "epoch": 0.41475875118259226,
+      "grad_norm": 2.9922358989715576,
+      "learning_rate": 1.9211372224385222e-05,
+      "loss": 1.3093,
+      "step": 548
+    },
+    {
+      "epoch": 0.415515610217597,
+      "grad_norm": 2.6461005210876465,
+      "learning_rate": 1.9208244381318892e-05,
+      "loss": 1.2585,
+      "step": 549
+    },
+    {
+      "epoch": 0.4162724692526017,
+      "grad_norm": 2.7143542766571045,
+      "learning_rate": 1.9205110603403247e-05,
+      "loss": 1.2594,
+      "step": 550
+    },
+    {
+      "epoch": 0.4170293282876064,
+      "grad_norm": 2.9333744049072266,
+      "learning_rate": 1.9201970892658273e-05,
+      "loss": 1.3178,
+      "step": 551
+    },
+    {
+      "epoch": 0.4177861873226112,
+      "grad_norm": 2.956841230392456,
+      "learning_rate": 1.919882525110778e-05,
+      "loss": 1.2745,
+      "step": 552
+    },
+    {
+      "epoch": 0.4185430463576159,
+      "grad_norm": 3.0672903060913086,
+      "learning_rate": 1.91956736807794e-05,
+      "loss": 1.2648,
+      "step": 553
+    },
+    {
+      "epoch": 0.41929990539262063,
+      "grad_norm": 2.7969796657562256,
+      "learning_rate": 1.9192516183704587e-05,
+      "loss": 1.3154,
+      "step": 554
+    },
+    {
+      "epoch": 0.42005676442762535,
+      "grad_norm": 2.9009835720062256,
+      "learning_rate": 1.9189352761918616e-05,
+      "loss": 1.2412,
+      "step": 555
+    },
+    {
+      "epoch": 0.42081362346263007,
+      "grad_norm": 2.8731672763824463,
+      "learning_rate": 1.918618341746058e-05,
+      "loss": 1.2811,
+      "step": 556
+    },
+    {
+      "epoch": 0.4215704824976348,
+      "grad_norm": 2.7065563201904297,
+      "learning_rate": 1.918300815237339e-05,
+      "loss": 1.2895,
+      "step": 557
+    },
+    {
+      "epoch": 0.42232734153263957,
+      "grad_norm": 2.670109748840332,
+      "learning_rate": 1.9179826968703775e-05,
+      "loss": 1.2809,
+      "step": 558
+    },
+    {
+      "epoch": 0.4230842005676443,
+      "grad_norm": 2.9249067306518555,
+      "learning_rate": 1.9176639868502273e-05,
+      "loss": 1.3528,
+      "step": 559
+    },
+    {
+      "epoch": 0.423841059602649,
+      "grad_norm": 2.733651638031006,
+      "learning_rate": 1.917344685382325e-05,
+      "loss": 1.2516,
+      "step": 560
+    },
+    {
+      "epoch": 0.4245979186376537,
+      "grad_norm": 3.126077651977539,
+      "learning_rate": 1.9170247926724863e-05,
+      "loss": 1.3048,
+      "step": 561
+    },
+    {
+      "epoch": 0.42535477767265845,
+      "grad_norm": 3.024705648422241,
+      "learning_rate": 1.9167043089269096e-05,
+      "loss": 1.2871,
+      "step": 562
+    },
+    {
+      "epoch": 0.4261116367076632,
+      "grad_norm": 3.0809972286224365,
+      "learning_rate": 1.916383234352174e-05,
+      "loss": 1.2939,
+      "step": 563
+    },
+    {
+      "epoch": 0.42686849574266794,
+      "grad_norm": 2.8006155490875244,
+      "learning_rate": 1.9160615691552388e-05,
+      "loss": 1.2681,
+      "step": 564
+    },
+    {
+      "epoch": 0.42762535477767266,
+      "grad_norm": 3.146348714828491,
+      "learning_rate": 1.915739313543445e-05,
+      "loss": 1.299,
+      "step": 565
+    },
+    {
+      "epoch": 0.4283822138126774,
+      "grad_norm": 2.707672119140625,
+      "learning_rate": 1.915416467724514e-05,
+      "loss": 1.305,
+      "step": 566
+    },
+    {
+      "epoch": 0.4291390728476821,
+      "grad_norm": 3.0839362144470215,
+      "learning_rate": 1.9150930319065465e-05,
+      "loss": 1.2806,
+      "step": 567
+    },
+    {
+      "epoch": 0.4298959318826869,
+      "grad_norm": 2.6987831592559814,
+      "learning_rate": 1.9147690062980243e-05,
+      "loss": 1.2449,
+      "step": 568
+    },
+    {
+      "epoch": 0.4306527909176916,
+      "grad_norm": 3.5137927532196045,
+      "learning_rate": 1.9144443911078098e-05,
+      "loss": 1.2525,
+      "step": 569
+    },
+    {
+      "epoch": 0.4314096499526963,
+      "grad_norm": 2.656526803970337,
+      "learning_rate": 1.914119186545145e-05,
+      "loss": 1.2801,
+      "step": 570
+    },
+    {
+      "epoch": 0.43216650898770104,
+      "grad_norm": 2.7091798782348633,
+      "learning_rate": 1.9137933928196514e-05,
+      "loss": 1.2743,
+      "step": 571
+    },
+    {
+      "epoch": 0.43292336802270576,
+      "grad_norm": 2.6860084533691406,
+      "learning_rate": 1.913467010141331e-05,
+      "loss": 1.2569,
+      "step": 572
+    },
+    {
+      "epoch": 0.4336802270577105,
+      "grad_norm": 2.8987984657287598,
+      "learning_rate": 1.9131400387205653e-05,
+      "loss": 1.2411,
+      "step": 573
+    },
+    {
+      "epoch": 0.43443708609271525,
+      "grad_norm": 2.579749584197998,
+      "learning_rate": 1.9128124787681145e-05,
+      "loss": 1.2344,
+      "step": 574
+    },
+    {
+      "epoch": 0.43519394512771997,
+      "grad_norm": 2.835766553878784,
+      "learning_rate": 1.912484330495119e-05,
+      "loss": 1.2922,
+      "step": 575
+    },
+    {
+      "epoch": 0.4359508041627247,
+      "grad_norm": 3.549691915512085,
+      "learning_rate": 1.9121555941130986e-05,
+      "loss": 1.2908,
+      "step": 576
+    },
+    {
+      "epoch": 0.4367076631977294,
+      "grad_norm": 2.881730556488037,
+      "learning_rate": 1.911826269833951e-05,
+      "loss": 1.2787,
+      "step": 577
+    },
+    {
+      "epoch": 0.43746452223273413,
+      "grad_norm": 2.881334066390991,
+      "learning_rate": 1.9114963578699538e-05,
+      "loss": 1.3111,
+      "step": 578
+    },
+    {
+      "epoch": 0.4382213812677389,
+      "grad_norm": 2.941556453704834,
+      "learning_rate": 1.911165858433764e-05,
+      "loss": 1.2857,
+      "step": 579
+    },
+    {
+      "epoch": 0.4389782403027436,
+      "grad_norm": 2.6916472911834717,
+      "learning_rate": 1.9108347717384156e-05,
+      "loss": 1.2512,
+      "step": 580
+    },
+    {
+      "epoch": 0.43973509933774835,
+      "grad_norm": 3.0234310626983643,
+      "learning_rate": 1.9105030979973223e-05,
+      "loss": 1.2089,
+      "step": 581
+    },
+    {
+      "epoch": 0.44049195837275307,
+      "grad_norm": 2.7675161361694336,
+      "learning_rate": 1.9101708374242764e-05,
+      "loss": 1.3253,
+      "step": 582
+    },
+    {
+      "epoch": 0.4412488174077578,
+      "grad_norm": 2.746612310409546,
+      "learning_rate": 1.909837990233447e-05,
+      "loss": 1.2554,
+      "step": 583
+    },
+    {
+      "epoch": 0.44200567644276256,
+      "grad_norm": 2.629913091659546,
+      "learning_rate": 1.9095045566393834e-05,
+      "loss": 1.3158,
+      "step": 584
+    },
+    {
+      "epoch": 0.4427625354777673,
+      "grad_norm": 3.0382394790649414,
+      "learning_rate": 1.909170536857011e-05,
+      "loss": 1.3382,
+      "step": 585
+    },
+    {
+      "epoch": 0.443519394512772,
+      "grad_norm": 3.1332645416259766,
+      "learning_rate": 1.908835931101634e-05,
+      "loss": 1.2561,
+      "step": 586
+    },
+    {
+      "epoch": 0.4442762535477767,
+      "grad_norm": 2.91369891166687,
+      "learning_rate": 1.9085007395889342e-05,
+      "loss": 1.287,
+      "step": 587
+    },
+    {
+      "epoch": 0.44503311258278144,
+      "grad_norm": 2.6690382957458496,
+      "learning_rate": 1.9081649625349715e-05,
+      "loss": 1.275,
+      "step": 588
+    },
+    {
+      "epoch": 0.44578997161778616,
+      "grad_norm": 2.7576904296875,
+      "learning_rate": 1.9078286001561822e-05,
+      "loss": 1.2669,
+      "step": 589
+    },
+    {
+      "epoch": 0.44654683065279094,
+      "grad_norm": 2.731320381164551,
+      "learning_rate": 1.9074916526693804e-05,
+      "loss": 1.292,
+      "step": 590
+    },
+    {
+      "epoch": 0.44730368968779566,
+      "grad_norm": 2.6240909099578857,
+      "learning_rate": 1.9071541202917572e-05,
+      "loss": 1.2852,
+      "step": 591
+    },
+    {
+      "epoch": 0.4480605487228004,
+      "grad_norm": 2.8189620971679688,
+      "learning_rate": 1.906816003240881e-05,
+      "loss": 1.2655,
+      "step": 592
+    },
+    {
+      "epoch": 0.4488174077578051,
+      "grad_norm": 2.7323951721191406,
+      "learning_rate": 1.906477301734697e-05,
+      "loss": 1.2942,
+      "step": 593
+    },
+    {
+      "epoch": 0.4495742667928098,
+      "grad_norm": 2.8606555461883545,
+      "learning_rate": 1.9061380159915262e-05,
+      "loss": 1.3039,
+      "step": 594
+    },
+    {
+      "epoch": 0.4503311258278146,
+      "grad_norm": 2.7523887157440186,
+      "learning_rate": 1.9057981462300683e-05,
+      "loss": 1.2372,
+      "step": 595
+    },
+    {
+      "epoch": 0.4510879848628193,
+      "grad_norm": 3.1251001358032227,
+      "learning_rate": 1.9054576926693977e-05,
+      "loss": 1.2726,
+      "step": 596
+    },
+    {
+      "epoch": 0.45184484389782403,
+      "grad_norm": 3.1092488765716553,
+      "learning_rate": 1.9051166555289652e-05,
+      "loss": 1.3126,
+      "step": 597
+    },
+    {
+      "epoch": 0.45260170293282875,
+      "grad_norm": 2.722238302230835,
+      "learning_rate": 1.904775035028598e-05,
+      "loss": 1.2765,
+      "step": 598
+    },
+    {
+      "epoch": 0.4533585619678335,
+      "grad_norm": 3.9474592208862305,
+      "learning_rate": 1.9044328313885e-05,
+      "loss": 1.2389,
+      "step": 599
+    },
+    {
+      "epoch": 0.45411542100283825,
+      "grad_norm": 2.7783472537994385,
+      "learning_rate": 1.90409004482925e-05,
+      "loss": 1.2683,
+      "step": 600
+    },
+    {
+      "epoch": 0.45487228003784297,
+      "grad_norm": 2.7635014057159424,
+      "learning_rate": 1.9037466755718038e-05,
+      "loss": 1.3073,
+      "step": 601
+    },
+    {
+      "epoch": 0.4556291390728477,
+      "grad_norm": 2.899637222290039,
+      "learning_rate": 1.903402723837491e-05,
+      "loss": 1.2682,
+      "step": 602
+    },
+    {
+      "epoch": 0.4563859981078524,
+      "grad_norm": 2.5725064277648926,
+      "learning_rate": 1.9030581898480182e-05,
+      "loss": 1.2445,
+      "step": 603
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 2.767765760421753,
+      "learning_rate": 1.902713073825467e-05,
+      "loss": 1.3006,
+      "step": 604
+    },
+    {
+      "epoch": 0.45789971617786185,
+      "grad_norm": 2.7437305450439453,
+      "learning_rate": 1.902367375992293e-05,
+      "loss": 1.256,
+      "step": 605
+    },
+    {
+      "epoch": 0.4586565752128666,
+      "grad_norm": 2.764497756958008,
+      "learning_rate": 1.9020210965713287e-05,
+      "loss": 1.2316,
+      "step": 606
+    },
+    {
+      "epoch": 0.45941343424787134,
+      "grad_norm": 2.6510708332061768,
+      "learning_rate": 1.9016742357857802e-05,
+      "loss": 1.2413,
+      "step": 607
+    },
+    {
+      "epoch": 0.46017029328287606,
+      "grad_norm": 2.727973699569702,
+      "learning_rate": 1.9013267938592282e-05,
+      "loss": 1.2779,
+      "step": 608
+    },
+    {
+      "epoch": 0.4609271523178808,
+      "grad_norm": 2.7336103916168213,
+      "learning_rate": 1.900978771015629e-05,
+      "loss": 1.3133,
+      "step": 609
+    },
+    {
+      "epoch": 0.4616840113528855,
+      "grad_norm": 2.635427713394165,
+      "learning_rate": 1.9006301674793128e-05,
+      "loss": 1.233,
+      "step": 610
+    },
+    {
+      "epoch": 0.4624408703878903,
+      "grad_norm": 2.99351167678833,
+      "learning_rate": 1.900280983474984e-05,
+      "loss": 1.2353,
+      "step": 611
+    },
+    {
+      "epoch": 0.463197729422895,
+      "grad_norm": 3.155054807662964,
+      "learning_rate": 1.8999312192277217e-05,
+      "loss": 1.3258,
+      "step": 612
+    },
+    {
+      "epoch": 0.4639545884578997,
+      "grad_norm": 2.745626926422119,
+      "learning_rate": 1.8995808749629773e-05,
+      "loss": 1.2321,
+      "step": 613
+    },
+    {
+      "epoch": 0.46471144749290444,
+      "grad_norm": 2.662928819656372,
+      "learning_rate": 1.899229950906579e-05,
+      "loss": 1.2291,
+      "step": 614
+    },
+    {
+      "epoch": 0.46546830652790916,
+      "grad_norm": 2.684296131134033,
+      "learning_rate": 1.8988784472847262e-05,
+      "loss": 1.2575,
+      "step": 615
+    },
+    {
+      "epoch": 0.46622516556291393,
+      "grad_norm": 2.850404977798462,
+      "learning_rate": 1.8985263643239932e-05,
+      "loss": 1.2727,
+      "step": 616
+    },
+    {
+      "epoch": 0.46698202459791865,
+      "grad_norm": 2.8185768127441406,
+      "learning_rate": 1.8981737022513268e-05,
+      "loss": 1.2145,
+      "step": 617
+    },
+    {
+      "epoch": 0.4677388836329234,
+      "grad_norm": 2.865675449371338,
+      "learning_rate": 1.8978204612940476e-05,
+      "loss": 1.2602,
+      "step": 618
+    },
+    {
+      "epoch": 0.4684957426679281,
+      "grad_norm": 2.706779718399048,
+      "learning_rate": 1.8974666416798496e-05,
+      "loss": 1.2578,
+      "step": 619
+    },
+    {
+      "epoch": 0.4692526017029328,
+      "grad_norm": 2.7865641117095947,
+      "learning_rate": 1.8971122436368002e-05,
+      "loss": 1.2549,
+      "step": 620
+    },
+    {
+      "epoch": 0.47000946073793753,
+      "grad_norm": 3.0289227962493896,
+      "learning_rate": 1.8967572673933373e-05,
+      "loss": 1.2794,
+      "step": 621
+    },
+    {
+      "epoch": 0.4707663197729423,
+      "grad_norm": 2.986976146697998,
+      "learning_rate": 1.8964017131782748e-05,
+      "loss": 1.2666,
+      "step": 622
+    },
+    {
+      "epoch": 0.47152317880794703,
+      "grad_norm": 2.907590866088867,
+      "learning_rate": 1.896045581220797e-05,
+      "loss": 1.3149,
+      "step": 623
+    },
+    {
+      "epoch": 0.47228003784295175,
+      "grad_norm": 2.5124711990356445,
+      "learning_rate": 1.8956888717504607e-05,
+      "loss": 1.2692,
+      "step": 624
+    },
+    {
+      "epoch": 0.47303689687795647,
+      "grad_norm": 2.8450794219970703,
+      "learning_rate": 1.8953315849971956e-05,
+      "loss": 1.2385,
+      "step": 625
+    },
+    {
+      "epoch": 0.4737937559129612,
+      "grad_norm": 3.127713441848755,
+      "learning_rate": 1.8949737211913038e-05,
+      "loss": 1.2725,
+      "step": 626
+    },
+    {
+      "epoch": 0.47455061494796597,
+      "grad_norm": 3.0674550533294678,
+      "learning_rate": 1.894615280563458e-05,
+      "loss": 1.3016,
+      "step": 627
+    },
+    {
+      "epoch": 0.4753074739829707,
+      "grad_norm": 3.29008150100708,
+      "learning_rate": 1.894256263344704e-05,
+      "loss": 1.2382,
+      "step": 628
+    },
+    {
+      "epoch": 0.4760643330179754,
+      "grad_norm": 3.2081003189086914,
+      "learning_rate": 1.8938966697664592e-05,
+      "loss": 1.259,
+      "step": 629
+    },
+    {
+      "epoch": 0.4768211920529801,
+      "grad_norm": 2.922011613845825,
+      "learning_rate": 1.8935365000605116e-05,
+      "loss": 1.3017,
+      "step": 630
+    },
+    {
+      "epoch": 0.47757805108798485,
+      "grad_norm": 3.075958490371704,
+      "learning_rate": 1.893175754459021e-05,
+      "loss": 1.2595,
+      "step": 631
+    },
+    {
+      "epoch": 0.4783349101229896,
+      "grad_norm": 2.9022579193115234,
+      "learning_rate": 1.892814433194519e-05,
+      "loss": 1.3033,
+      "step": 632
+    },
+    {
+      "epoch": 0.47909176915799434,
+      "grad_norm": 2.9433717727661133,
+      "learning_rate": 1.8924525364999077e-05,
+      "loss": 1.2636,
+      "step": 633
+    },
+    {
+      "epoch": 0.47984862819299906,
+      "grad_norm": 2.9550983905792236,
+      "learning_rate": 1.89209006460846e-05,
+      "loss": 1.2936,
+      "step": 634
+    },
+    {
+      "epoch": 0.4806054872280038,
+      "grad_norm": 2.8603897094726562,
+      "learning_rate": 1.8917270177538198e-05,
+      "loss": 1.2497,
+      "step": 635
+    },
+    {
+      "epoch": 0.4813623462630085,
+      "grad_norm": 3.0159318447113037,
+      "learning_rate": 1.8913633961700014e-05,
+      "loss": 1.2627,
+      "step": 636
+    },
+    {
+      "epoch": 0.4821192052980132,
+      "grad_norm": 3.3943378925323486,
+      "learning_rate": 1.8909992000913896e-05,
+      "loss": 1.2977,
+      "step": 637
+    },
+    {
+      "epoch": 0.482876064333018,
+      "grad_norm": 2.8387339115142822,
+      "learning_rate": 1.8906344297527403e-05,
+      "loss": 1.2922,
+      "step": 638
+    },
+    {
+      "epoch": 0.4836329233680227,
+      "grad_norm": 2.8385610580444336,
+      "learning_rate": 1.8902690853891787e-05,
+      "loss": 1.2023,
+      "step": 639
+    },
+    {
+      "epoch": 0.48438978240302744,
+      "grad_norm": 3.155811309814453,
+      "learning_rate": 1.8899031672362e-05,
+      "loss": 1.3069,
+      "step": 640
+    },
+    {
+      "epoch": 0.48514664143803216,
+      "grad_norm": 3.442098617553711,
+      "learning_rate": 1.8895366755296693e-05,
+      "loss": 1.2361,
+      "step": 641
+    },
+    {
+      "epoch": 0.4859035004730369,
+      "grad_norm": 2.805680751800537,
+      "learning_rate": 1.8891696105058218e-05,
+      "loss": 1.2349,
+      "step": 642
+    },
+    {
+      "epoch": 0.48666035950804165,
+      "grad_norm": 2.7870709896087646,
+      "learning_rate": 1.8888019724012618e-05,
+      "loss": 1.3326,
+      "step": 643
+    },
+    {
+      "epoch": 0.48741721854304637,
+      "grad_norm": 2.8645455837249756,
+      "learning_rate": 1.8884337614529636e-05,
+      "loss": 1.2829,
+      "step": 644
+    },
+    {
+      "epoch": 0.4881740775780511,
+      "grad_norm": 2.8770759105682373,
+      "learning_rate": 1.88806497789827e-05,
+      "loss": 1.2268,
+      "step": 645
+    },
+    {
+      "epoch": 0.4889309366130558,
+      "grad_norm": 2.8018059730529785,
+      "learning_rate": 1.8876956219748934e-05,
+      "loss": 1.2566,
+      "step": 646
+    },
+    {
+      "epoch": 0.48968779564806053,
+      "grad_norm": 3.0624117851257324,
+      "learning_rate": 1.887325693920915e-05,
+      "loss": 1.2776,
+      "step": 647
+    },
+    {
+      "epoch": 0.4904446546830653,
+      "grad_norm": 2.7411904335021973,
+      "learning_rate": 1.886955193974785e-05,
+      "loss": 1.2941,
+      "step": 648
+    },
+    {
+      "epoch": 0.49120151371807,
+      "grad_norm": 2.4694104194641113,
+      "learning_rate": 1.8865841223753216e-05,
+      "loss": 1.245,
+      "step": 649
+    },
+    {
+      "epoch": 0.49195837275307475,
+      "grad_norm": 2.4889931678771973,
+      "learning_rate": 1.886212479361712e-05,
+      "loss": 1.2664,
+      "step": 650
+    },
+    {
+      "epoch": 0.49271523178807947,
+      "grad_norm": 2.699221134185791,
+      "learning_rate": 1.885840265173512e-05,
+      "loss": 1.245,
+      "step": 651
+    },
+    {
+      "epoch": 0.4934720908230842,
+      "grad_norm": 3.0901527404785156,
+      "learning_rate": 1.8854674800506447e-05,
+      "loss": 1.2683,
+      "step": 652
+    },
+    {
+      "epoch": 0.4942289498580889,
+      "grad_norm": 2.5710549354553223,
+      "learning_rate": 1.8850941242334024e-05,
+      "loss": 1.2677,
+      "step": 653
+    },
+    {
+      "epoch": 0.4949858088930937,
+      "grad_norm": 2.747673988342285,
+      "learning_rate": 1.8847201979624433e-05,
+      "loss": 1.2487,
+      "step": 654
+    },
+    {
+      "epoch": 0.4957426679280984,
+      "grad_norm": 2.6453075408935547,
+      "learning_rate": 1.8843457014787954e-05,
+      "loss": 1.2534,
+      "step": 655
+    },
+    {
+      "epoch": 0.4964995269631031,
+      "grad_norm": 2.3280134201049805,
+      "learning_rate": 1.8839706350238537e-05,
+      "loss": 1.2529,
+      "step": 656
+    },
+    {
+      "epoch": 0.49725638599810784,
+      "grad_norm": 2.353527307510376,
+      "learning_rate": 1.88359499883938e-05,
+      "loss": 1.2612,
+      "step": 657
+    },
+    {
+      "epoch": 0.49801324503311256,
+      "grad_norm": 2.827341318130493,
+      "learning_rate": 1.8832187931675036e-05,
+      "loss": 1.2883,
+      "step": 658
+    },
+    {
+      "epoch": 0.49877010406811734,
+      "grad_norm": 2.620957374572754,
+      "learning_rate": 1.882842018250721e-05,
+      "loss": 1.25,
+      "step": 659
+    },
+    {
+      "epoch": 0.49952696310312206,
+      "grad_norm": 2.600372076034546,
+      "learning_rate": 1.8824646743318955e-05,
+      "loss": 1.2497,
+      "step": 660
+    },
+    {
+      "epoch": 0.5002838221381267,
+      "grad_norm": 2.544832706451416,
+      "learning_rate": 1.882086761654257e-05,
+      "loss": 1.2656,
+      "step": 661
+    },
+    {
+      "epoch": 0.5010406811731315,
+      "grad_norm": 2.809065818786621,
+      "learning_rate": 1.881708280461403e-05,
+      "loss": 1.3098,
+      "step": 662
+    },
+    {
+      "epoch": 0.5017975402081363,
+      "grad_norm": 2.423124313354492,
+      "learning_rate": 1.881329230997296e-05,
+      "loss": 1.2676,
+      "step": 663
+    },
+    {
+      "epoch": 0.5025543992431409,
+      "grad_norm": 2.6886796951293945,
+      "learning_rate": 1.880949613506266e-05,
+      "loss": 1.2764,
+      "step": 664
+    },
+    {
+      "epoch": 0.5033112582781457,
+      "grad_norm": 2.9043877124786377,
+      "learning_rate": 1.8805694282330076e-05,
+      "loss": 1.2499,
+      "step": 665
+    },
+    {
+      "epoch": 0.5040681173131504,
+      "grad_norm": 2.5381906032562256,
+      "learning_rate": 1.880188675422584e-05,
+      "loss": 1.2429,
+      "step": 666
+    },
+    {
+      "epoch": 0.5048249763481552,
+      "grad_norm": 2.5368845462799072,
+      "learning_rate": 1.8798073553204216e-05,
+      "loss": 1.2992,
+      "step": 667
+    },
+    {
+      "epoch": 0.5055818353831599,
+      "grad_norm": 2.313969850540161,
+      "learning_rate": 1.879425468172314e-05,
+      "loss": 1.2602,
+      "step": 668
+    },
+    {
+      "epoch": 0.5063386944181646,
+      "grad_norm": 2.473052978515625,
+      "learning_rate": 1.8790430142244192e-05,
+      "loss": 1.2558,
+      "step": 669
+    },
+    {
+      "epoch": 0.5070955534531694,
+      "grad_norm": 2.5860140323638916,
+      "learning_rate": 1.878659993723262e-05,
+      "loss": 1.2489,
+      "step": 670
+    },
+    {
+      "epoch": 0.507852412488174,
+      "grad_norm": 2.7334864139556885,
+      "learning_rate": 1.8782764069157307e-05,
+      "loss": 1.2892,
+      "step": 671
+    },
+    {
+      "epoch": 0.5086092715231788,
+      "grad_norm": 2.7741503715515137,
+      "learning_rate": 1.8778922540490803e-05,
+      "loss": 1.214,
+      "step": 672
+    },
+    {
+      "epoch": 0.5093661305581836,
+      "grad_norm": 2.3246145248413086,
+      "learning_rate": 1.8775075353709294e-05,
+      "loss": 1.2301,
+      "step": 673
+    },
+    {
+      "epoch": 0.5101229895931882,
+      "grad_norm": 2.879974365234375,
+      "learning_rate": 1.8771222511292622e-05,
+      "loss": 1.2351,
+      "step": 674
+    },
+    {
+      "epoch": 0.510879848628193,
+      "grad_norm": 2.5754384994506836,
+      "learning_rate": 1.8767364015724266e-05,
+      "loss": 1.2701,
+      "step": 675
+    },
+    {
+      "epoch": 0.5116367076631977,
+      "grad_norm": 2.623716115951538,
+      "learning_rate": 1.8763499869491356e-05,
+      "loss": 1.2934,
+      "step": 676
+    },
+    {
+      "epoch": 0.5123935666982025,
+      "grad_norm": 2.6354804039001465,
+      "learning_rate": 1.8759630075084664e-05,
+      "loss": 1.2454,
+      "step": 677
+    },
+    {
+      "epoch": 0.5131504257332072,
+      "grad_norm": 2.550604820251465,
+      "learning_rate": 1.8755754634998593e-05,
+      "loss": 1.2555,
+      "step": 678
+    },
+    {
+      "epoch": 0.5139072847682119,
+      "grad_norm": 2.5519111156463623,
+      "learning_rate": 1.8751873551731196e-05,
+      "loss": 1.2384,
+      "step": 679
+    },
+    {
+      "epoch": 0.5146641438032167,
+      "grad_norm": 2.6348938941955566,
+      "learning_rate": 1.8747986827784167e-05,
+      "loss": 1.2453,
+      "step": 680
+    },
+    {
+      "epoch": 0.5154210028382213,
+      "grad_norm": 2.5110082626342773,
+      "learning_rate": 1.874409446566282e-05,
+      "loss": 1.3047,
+      "step": 681
+    },
+    {
+      "epoch": 0.5161778618732261,
+      "grad_norm": 2.5216503143310547,
+      "learning_rate": 1.8740196467876114e-05,
+      "loss": 1.2464,
+      "step": 682
+    },
+    {
+      "epoch": 0.5169347209082309,
+      "grad_norm": 2.737325668334961,
+      "learning_rate": 1.8736292836936643e-05,
+      "loss": 1.2666,
+      "step": 683
+    },
+    {
+      "epoch": 0.5176915799432356,
+      "grad_norm": 2.625519037246704,
+      "learning_rate": 1.8732383575360625e-05,
+      "loss": 1.2403,
+      "step": 684
+    },
+    {
+      "epoch": 0.5184484389782403,
+      "grad_norm": 2.784569263458252,
+      "learning_rate": 1.8728468685667914e-05,
+      "loss": 1.2627,
+      "step": 685
+    },
+    {
+      "epoch": 0.519205298013245,
+      "grad_norm": 2.7349774837493896,
+      "learning_rate": 1.8724548170381983e-05,
+      "loss": 1.2771,
+      "step": 686
+    },
+    {
+      "epoch": 0.5199621570482498,
+      "grad_norm": 2.681603193283081,
+      "learning_rate": 1.8720622032029936e-05,
+      "loss": 1.276,
+      "step": 687
+    },
+    {
+      "epoch": 0.5207190160832545,
+      "grad_norm": 2.767359972000122,
+      "learning_rate": 1.8716690273142504e-05,
+      "loss": 1.2279,
+      "step": 688
+    },
+    {
+      "epoch": 0.5214758751182592,
+      "grad_norm": 2.5928122997283936,
+      "learning_rate": 1.871275289625404e-05,
+      "loss": 1.2568,
+      "step": 689
+    },
+    {
+      "epoch": 0.522232734153264,
+      "grad_norm": 2.6970558166503906,
+      "learning_rate": 1.8708809903902517e-05,
+      "loss": 1.3101,
+      "step": 690
+    },
+    {
+      "epoch": 0.5229895931882687,
+      "grad_norm": 2.6737709045410156,
+      "learning_rate": 1.8704861298629524e-05,
+      "loss": 1.2575,
+      "step": 691
+    },
+    {
+      "epoch": 0.5237464522232734,
+      "grad_norm": 3.0363659858703613,
+      "learning_rate": 1.870090708298028e-05,
+      "loss": 1.3034,
+      "step": 692
+    },
+    {
+      "epoch": 0.5245033112582781,
+      "grad_norm": 2.817183017730713,
+      "learning_rate": 1.8696947259503603e-05,
+      "loss": 1.2962,
+      "step": 693
+    },
+    {
+      "epoch": 0.5252601702932829,
+      "grad_norm": 3.507577896118164,
+      "learning_rate": 1.8692981830751937e-05,
+      "loss": 1.2643,
+      "step": 694
+    },
+    {
+      "epoch": 0.5260170293282876,
+      "grad_norm": 2.9019994735717773,
+      "learning_rate": 1.868901079928134e-05,
+      "loss": 1.2968,
+      "step": 695
+    },
+    {
+      "epoch": 0.5267738883632923,
+      "grad_norm": 2.6820502281188965,
+      "learning_rate": 1.8685034167651477e-05,
+      "loss": 1.281,
+      "step": 696
+    },
+    {
+      "epoch": 0.5275307473982971,
+      "grad_norm": 2.5685501098632812,
+      "learning_rate": 1.8681051938425626e-05,
+      "loss": 1.2368,
+      "step": 697
+    },
+    {
+      "epoch": 0.5282876064333017,
+      "grad_norm": 2.943498134613037,
+      "learning_rate": 1.867706411417067e-05,
+      "loss": 1.2494,
+      "step": 698
+    },
+    {
+      "epoch": 0.5290444654683065,
+      "grad_norm": 2.9893808364868164,
+      "learning_rate": 1.8673070697457097e-05,
+      "loss": 1.3033,
+      "step": 699
+    },
+    {
+      "epoch": 0.5298013245033113,
+      "grad_norm": 3.192913293838501,
+      "learning_rate": 1.8669071690859002e-05,
+      "loss": 1.3122,
+      "step": 700
+    },
+    {
+      "epoch": 0.530558183538316,
+      "grad_norm": 2.6208715438842773,
+      "learning_rate": 1.866506709695409e-05,
+      "loss": 1.2335,
+      "step": 701
+    },
+    {
+      "epoch": 0.5313150425733207,
+      "grad_norm": 2.793226718902588,
+      "learning_rate": 1.8661056918323654e-05,
+      "loss": 1.2721,
+      "step": 702
+    },
+    {
+      "epoch": 0.5320719016083254,
+      "grad_norm": 2.809190034866333,
+      "learning_rate": 1.8657041157552597e-05,
+      "loss": 1.2318,
+      "step": 703
+    },
+    {
+      "epoch": 0.5328287606433302,
+      "grad_norm": 2.70646595954895,
+      "learning_rate": 1.865301981722942e-05,
+      "loss": 1.2471,
+      "step": 704
+    },
+    {
+      "epoch": 0.533585619678335,
+      "grad_norm": 2.691943407058716,
+      "learning_rate": 1.864899289994621e-05,
+      "loss": 1.2765,
+      "step": 705
+    },
+    {
+      "epoch": 0.5343424787133396,
+      "grad_norm": 2.6376893520355225,
+      "learning_rate": 1.864496040829867e-05,
+      "loss": 1.2932,
+      "step": 706
+    },
+    {
+      "epoch": 0.5350993377483444,
+      "grad_norm": 2.727936029434204,
+      "learning_rate": 1.8640922344886066e-05,
+      "loss": 1.2056,
+      "step": 707
+    },
+    {
+      "epoch": 0.5358561967833491,
+      "grad_norm": 2.599090337753296,
+      "learning_rate": 1.863687871231128e-05,
+      "loss": 1.2747,
+      "step": 708
+    },
+    {
+      "epoch": 0.5366130558183538,
+      "grad_norm": 2.9305431842803955,
+      "learning_rate": 1.863282951318078e-05,
+      "loss": 1.2593,
+      "step": 709
+    },
+    {
+      "epoch": 0.5373699148533586,
+      "grad_norm": 2.5242085456848145,
+      "learning_rate": 1.8628774750104615e-05,
+      "loss": 1.2669,
+      "step": 710
+    },
+    {
+      "epoch": 0.5381267738883633,
+      "grad_norm": 2.737729787826538,
+      "learning_rate": 1.862471442569642e-05,
+      "loss": 1.2515,
+      "step": 711
+    },
+    {
+      "epoch": 0.538883632923368,
+      "grad_norm": 2.8515143394470215,
+      "learning_rate": 1.8620648542573423e-05,
+      "loss": 1.2483,
+      "step": 712
+    },
+    {
+      "epoch": 0.5396404919583727,
+      "grad_norm": 2.8016417026519775,
+      "learning_rate": 1.8616577103356425e-05,
+      "loss": 1.2389,
+      "step": 713
+    },
+    {
+      "epoch": 0.5403973509933775,
+      "grad_norm": 2.9451699256896973,
+      "learning_rate": 1.861250011066982e-05,
+      "loss": 1.2345,
+      "step": 714
+    },
+    {
+      "epoch": 0.5411542100283823,
+      "grad_norm": 2.771279811859131,
+      "learning_rate": 1.8608417567141572e-05,
+      "loss": 1.2621,
+      "step": 715
+    },
+    {
+      "epoch": 0.5419110690633869,
+      "grad_norm": 2.9805190563201904,
+      "learning_rate": 1.860432947540322e-05,
+      "loss": 1.2348,
+      "step": 716
+    },
+    {
+      "epoch": 0.5426679280983917,
+      "grad_norm": 2.803847312927246,
+      "learning_rate": 1.8600235838089896e-05,
+      "loss": 1.241,
+      "step": 717
+    },
+    {
+      "epoch": 0.5434247871333964,
+      "grad_norm": 2.4871954917907715,
+      "learning_rate": 1.859613665784029e-05,
+      "loss": 1.2883,
+      "step": 718
+    },
+    {
+      "epoch": 0.5441816461684011,
+      "grad_norm": 3.067754030227661,
+      "learning_rate": 1.8592031937296673e-05,
+      "loss": 1.2833,
+      "step": 719
+    },
+    {
+      "epoch": 0.5449385052034059,
+      "grad_norm": 2.8348135948181152,
+      "learning_rate": 1.8587921679104887e-05,
+      "loss": 1.3083,
+      "step": 720
+    },
+    {
+      "epoch": 0.5456953642384106,
+      "grad_norm": 2.538663387298584,
+      "learning_rate": 1.8583805885914345e-05,
+      "loss": 1.2288,
+      "step": 721
+    },
+    {
+      "epoch": 0.5464522232734154,
+      "grad_norm": 2.7975425720214844,
+      "learning_rate": 1.857968456037801e-05,
+      "loss": 1.3166,
+      "step": 722
+    },
+    {
+      "epoch": 0.54720908230842,
+      "grad_norm": 2.60284423828125,
+      "learning_rate": 1.857555770515244e-05,
+      "loss": 1.251,
+      "step": 723
+    },
+    {
+      "epoch": 0.5479659413434248,
+      "grad_norm": 3.0047545433044434,
+      "learning_rate": 1.857142532289774e-05,
+      "loss": 1.2372,
+      "step": 724
+    },
+    {
+      "epoch": 0.5487228003784295,
+      "grad_norm": 2.7439827919006348,
+      "learning_rate": 1.8567287416277576e-05,
+      "loss": 1.2686,
+      "step": 725
+    },
+    {
+      "epoch": 0.5494796594134342,
+      "grad_norm": 2.7966012954711914,
+      "learning_rate": 1.856314398795918e-05,
+      "loss": 1.2997,
+      "step": 726
+    },
+    {
+      "epoch": 0.550236518448439,
+      "grad_norm": 2.4072394371032715,
+      "learning_rate": 1.855899504061335e-05,
+      "loss": 1.2371,
+      "step": 727
+    },
+    {
+      "epoch": 0.5509933774834437,
+      "grad_norm": 2.6710758209228516,
+      "learning_rate": 1.8554840576914425e-05,
+      "loss": 1.3084,
+      "step": 728
+    },
+    {
+      "epoch": 0.5517502365184485,
+      "grad_norm": 2.4834091663360596,
+      "learning_rate": 1.8550680599540315e-05,
+      "loss": 1.2335,
+      "step": 729
+    },
+    {
+      "epoch": 0.5525070955534531,
+      "grad_norm": 3.0747454166412354,
+      "learning_rate": 1.8546515111172475e-05,
+      "loss": 1.2691,
+      "step": 730
+    },
+    {
+      "epoch": 0.5532639545884579,
+      "grad_norm": 2.3881189823150635,
+      "learning_rate": 1.8542344114495918e-05,
+      "loss": 1.2852,
+      "step": 731
+    },
+    {
+      "epoch": 0.5540208136234627,
+      "grad_norm": 2.559795618057251,
+      "learning_rate": 1.85381676121992e-05,
+      "loss": 1.2266,
+      "step": 732
+    },
+    {
+      "epoch": 0.5547776726584673,
+      "grad_norm": 2.5426385402679443,
+      "learning_rate": 1.8533985606974436e-05,
+      "loss": 1.2136,
+      "step": 733
+    },
+    {
+      "epoch": 0.5555345316934721,
+      "grad_norm": 2.7627816200256348,
+      "learning_rate": 1.8529798101517283e-05,
+      "loss": 1.227,
+      "step": 734
+    },
+    {
+      "epoch": 0.5562913907284768,
+      "grad_norm": 3.559936285018921,
+      "learning_rate": 1.8525605098526935e-05,
+      "loss": 1.2823,
+      "step": 735
+    },
+    {
+      "epoch": 0.5570482497634816,
+      "grad_norm": 2.6380114555358887,
+      "learning_rate": 1.8521406600706146e-05,
+      "loss": 1.2077,
+      "step": 736
+    },
+    {
+      "epoch": 0.5578051087984863,
+      "grad_norm": 2.3080461025238037,
+      "learning_rate": 1.8517202610761203e-05,
+      "loss": 1.2146,
+      "step": 737
+    },
+    {
+      "epoch": 0.558561967833491,
+      "grad_norm": 2.245431423187256,
+      "learning_rate": 1.851299313140193e-05,
+      "loss": 1.2073,
+      "step": 738
+    },
+    {
+      "epoch": 0.5593188268684958,
+      "grad_norm": 2.4832706451416016,
+      "learning_rate": 1.8508778165341697e-05,
+      "loss": 1.2167,
+      "step": 739
+    },
+    {
+      "epoch": 0.5600756859035004,
+      "grad_norm": 2.646280288696289,
+      "learning_rate": 1.85045577152974e-05,
+      "loss": 1.2379,
+      "step": 740
+    },
+    {
+      "epoch": 0.5608325449385052,
+      "grad_norm": 2.449310302734375,
+      "learning_rate": 1.8500331783989486e-05,
+      "loss": 1.2085,
+      "step": 741
+    },
+    {
+      "epoch": 0.56158940397351,
+      "grad_norm": 2.7046239376068115,
+      "learning_rate": 1.8496100374141924e-05,
+      "loss": 1.2255,
+      "step": 742
+    },
+    {
+      "epoch": 0.5623462630085146,
+      "grad_norm": 2.5250003337860107,
+      "learning_rate": 1.849186348848221e-05,
+      "loss": 1.2028,
+      "step": 743
+    },
+    {
+      "epoch": 0.5631031220435194,
+      "grad_norm": 2.423783779144287,
+      "learning_rate": 1.848762112974138e-05,
+      "loss": 1.2485,
+      "step": 744
+    },
+    {
+      "epoch": 0.5638599810785241,
+      "grad_norm": 2.3143739700317383,
+      "learning_rate": 1.8483373300653995e-05,
+      "loss": 1.2238,
+      "step": 745
+    },
+    {
+      "epoch": 0.5646168401135289,
+      "grad_norm": 2.433070421218872,
+      "learning_rate": 1.8479120003958136e-05,
+      "loss": 1.2496,
+      "step": 746
+    },
+    {
+      "epoch": 0.5653736991485336,
+      "grad_norm": 2.5320703983306885,
+      "learning_rate": 1.8474861242395424e-05,
+      "loss": 1.2477,
+      "step": 747
+    },
+    {
+      "epoch": 0.5661305581835383,
+      "grad_norm": 5.566840171813965,
+      "learning_rate": 1.8470597018710976e-05,
+      "loss": 1.2086,
+      "step": 748
+    },
+    {
+      "epoch": 0.5668874172185431,
+      "grad_norm": 2.5624606609344482,
+      "learning_rate": 1.8466327335653458e-05,
+      "loss": 1.2458,
+      "step": 749
+    },
+    {
+      "epoch": 0.5676442762535477,
+      "grad_norm": 3.781528949737549,
+      "learning_rate": 1.846205219597504e-05,
+      "loss": 1.259,
+      "step": 750
+    },
+    {
+      "epoch": 0.5684011352885525,
+      "grad_norm": 2.4453654289245605,
+      "learning_rate": 1.8457771602431406e-05,
+      "loss": 1.2511,
+      "step": 751
+    },
+    {
+      "epoch": 0.5691579943235573,
+      "grad_norm": 2.4234702587127686,
+      "learning_rate": 1.8453485557781768e-05,
+      "loss": 1.2339,
+      "step": 752
+    },
+    {
+      "epoch": 0.569914853358562,
+      "grad_norm": 2.637007236480713,
+      "learning_rate": 1.8449194064788845e-05,
+      "loss": 1.2274,
+      "step": 753
+    },
+    {
+      "epoch": 0.5706717123935667,
+      "grad_norm": 2.557408332824707,
+      "learning_rate": 1.8444897126218865e-05,
+      "loss": 1.2718,
+      "step": 754
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 2.3460357189178467,
+      "learning_rate": 1.8440594744841564e-05,
+      "loss": 1.2522,
+      "step": 755
+    },
+    {
+      "epoch": 0.5721854304635762,
+      "grad_norm": 2.9702370166778564,
+      "learning_rate": 1.84362869234302e-05,
+      "loss": 1.2365,
+      "step": 756
+    },
+    {
+      "epoch": 0.5729422894985808,
+      "grad_norm": 2.4645347595214844,
+      "learning_rate": 1.843197366476153e-05,
+      "loss": 1.2497,
+      "step": 757
+    },
+    {
+      "epoch": 0.5736991485335856,
+      "grad_norm": 2.525984764099121,
+      "learning_rate": 1.8427654971615804e-05,
+      "loss": 1.2472,
+      "step": 758
+    },
+    {
+      "epoch": 0.5744560075685904,
+      "grad_norm": 2.598914861679077,
+      "learning_rate": 1.8423330846776797e-05,
+      "loss": 1.2783,
+      "step": 759
+    },
+    {
+      "epoch": 0.575212866603595,
+      "grad_norm": 2.464893341064453,
+      "learning_rate": 1.841900129303177e-05,
+      "loss": 1.2331,
+      "step": 760
+    },
+    {
+      "epoch": 0.5759697256385998,
+      "grad_norm": 2.517779588699341,
+      "learning_rate": 1.8414666313171488e-05,
+      "loss": 1.2087,
+      "step": 761
+    },
+    {
+      "epoch": 0.5767265846736045,
+      "grad_norm": 2.3364832401275635,
+      "learning_rate": 1.8410325909990207e-05,
+      "loss": 1.251,
+      "step": 762
+    },
+    {
+      "epoch": 0.5774834437086093,
+      "grad_norm": 2.348635673522949,
+      "learning_rate": 1.8405980086285693e-05,
+      "loss": 1.2424,
+      "step": 763
+    },
+    {
+      "epoch": 0.578240302743614,
+      "grad_norm": 2.472801446914673,
+      "learning_rate": 1.8401628844859193e-05,
+      "loss": 1.1972,
+      "step": 764
+    },
+    {
+      "epoch": 0.5789971617786187,
+      "grad_norm": 2.528832197189331,
+      "learning_rate": 1.839727218851545e-05,
+      "loss": 1.2904,
+      "step": 765
+    },
+    {
+      "epoch": 0.5797540208136235,
+      "grad_norm": 2.833585262298584,
+      "learning_rate": 1.83929101200627e-05,
+      "loss": 1.2284,
+      "step": 766
+    },
+    {
+      "epoch": 0.5805108798486281,
+      "grad_norm": 2.886864185333252,
+      "learning_rate": 1.838854264231267e-05,
+      "loss": 1.2529,
+      "step": 767
+    },
+    {
+      "epoch": 0.5812677388836329,
+      "grad_norm": 2.6184258460998535,
+      "learning_rate": 1.8384169758080564e-05,
+      "loss": 1.2422,
+      "step": 768
+    },
+    {
+      "epoch": 0.5820245979186377,
+      "grad_norm": 2.59594988822937,
+      "learning_rate": 1.8379791470185077e-05,
+      "loss": 1.2349,
+      "step": 769
+    },
+    {
+      "epoch": 0.5827814569536424,
+      "grad_norm": 2.359560489654541,
+      "learning_rate": 1.837540778144839e-05,
+      "loss": 1.2146,
+      "step": 770
+    },
+    {
+      "epoch": 0.5835383159886471,
+      "grad_norm": 3.088444709777832,
+      "learning_rate": 1.8371018694696155e-05,
+      "loss": 1.2667,
+      "step": 771
+    },
+    {
+      "epoch": 0.5842951750236518,
+      "grad_norm": 2.766091823577881,
+      "learning_rate": 1.836662421275752e-05,
+      "loss": 1.218,
+      "step": 772
+    },
+    {
+      "epoch": 0.5850520340586566,
+      "grad_norm": 2.739274263381958,
+      "learning_rate": 1.8362224338465093e-05,
+      "loss": 1.2618,
+      "step": 773
+    },
+    {
+      "epoch": 0.5858088930936614,
+      "grad_norm": 4.742860794067383,
+      "learning_rate": 1.835781907465497e-05,
+      "loss": 1.2989,
+      "step": 774
+    },
+    {
+      "epoch": 0.586565752128666,
+      "grad_norm": 3.0373001098632812,
+      "learning_rate": 1.8353408424166712e-05,
+      "loss": 1.283,
+      "step": 775
+    },
+    {
+      "epoch": 0.5873226111636708,
+      "grad_norm": 2.5657973289489746,
+      "learning_rate": 1.8348992389843365e-05,
+      "loss": 1.1942,
+      "step": 776
+    },
+    {
+      "epoch": 0.5880794701986755,
+      "grad_norm": 2.7128591537475586,
+      "learning_rate": 1.834457097453143e-05,
+      "loss": 1.262,
+      "step": 777
+    },
+    {
+      "epoch": 0.5888363292336802,
+      "grad_norm": 2.4917023181915283,
+      "learning_rate": 1.834014418108089e-05,
+      "loss": 1.2194,
+      "step": 778
+    },
+    {
+      "epoch": 0.589593188268685,
+      "grad_norm": 2.7309277057647705,
+      "learning_rate": 1.8335712012345188e-05,
+      "loss": 1.231,
+      "step": 779
+    },
+    {
+      "epoch": 0.5903500473036897,
+      "grad_norm": 2.894216537475586,
+      "learning_rate": 1.8331274471181224e-05,
+      "loss": 1.234,
+      "step": 780
+    },
+    {
+      "epoch": 0.5911069063386944,
+      "grad_norm": 2.491863250732422,
+      "learning_rate": 1.8326831560449375e-05,
+      "loss": 1.2335,
+      "step": 781
+    },
+    {
+      "epoch": 0.5918637653736991,
+      "grad_norm": 2.636247396469116,
+      "learning_rate": 1.832238328301348e-05,
+      "loss": 1.2371,
+      "step": 782
+    },
+    {
+      "epoch": 0.5926206244087039,
+      "grad_norm": 2.6232643127441406,
+      "learning_rate": 1.831792964174082e-05,
+      "loss": 1.2626,
+      "step": 783
+    },
+    {
+      "epoch": 0.5933774834437087,
+      "grad_norm": 2.667076826095581,
+      "learning_rate": 1.8313470639502148e-05,
+      "loss": 1.257,
+      "step": 784
+    },
+    {
+      "epoch": 0.5941343424787133,
+      "grad_norm": 2.936359405517578,
+      "learning_rate": 1.8309006279171675e-05,
+      "loss": 1.2509,
+      "step": 785
+    },
+    {
+      "epoch": 0.5948912015137181,
+      "grad_norm": 2.522406578063965,
+      "learning_rate": 1.8304536563627052e-05,
+      "loss": 1.2804,
+      "step": 786
+    },
+    {
+      "epoch": 0.5956480605487228,
+      "grad_norm": 2.542407512664795,
+      "learning_rate": 1.830006149574939e-05,
+      "loss": 1.2393,
+      "step": 787
+    },
+    {
+      "epoch": 0.5964049195837275,
+      "grad_norm": 2.5919876098632812,
+      "learning_rate": 1.8295581078423253e-05,
+      "loss": 1.2622,
+      "step": 788
+    },
+    {
+      "epoch": 0.5971617786187322,
+      "grad_norm": 2.7095932960510254,
+      "learning_rate": 1.8291095314536647e-05,
+      "loss": 1.2491,
+      "step": 789
+    },
+    {
+      "epoch": 0.597918637653737,
+      "grad_norm": 2.4110512733459473,
+      "learning_rate": 1.8286604206981028e-05,
+      "loss": 1.2622,
+      "step": 790
+    },
+    {
+      "epoch": 0.5986754966887418,
+      "grad_norm": 2.7041079998016357,
+      "learning_rate": 1.8282107758651295e-05,
+      "loss": 1.2563,
+      "step": 791
+    },
+    {
+      "epoch": 0.5994323557237464,
+      "grad_norm": 2.7525973320007324,
+      "learning_rate": 1.827760597244579e-05,
+      "loss": 1.2449,
+      "step": 792
+    },
+    {
+      "epoch": 0.6001892147587512,
+      "grad_norm": 2.612968921661377,
+      "learning_rate": 1.8273098851266297e-05,
+      "loss": 1.258,
+      "step": 793
+    },
+    {
+      "epoch": 0.6009460737937559,
+      "grad_norm": 2.6070921421051025,
+      "learning_rate": 1.826858639801804e-05,
+      "loss": 1.3045,
+      "step": 794
+    },
+    {
+      "epoch": 0.6017029328287606,
+      "grad_norm": 2.4890692234039307,
+      "learning_rate": 1.8264068615609668e-05,
+      "loss": 1.2253,
+      "step": 795
+    },
+    {
+      "epoch": 0.6024597918637654,
+      "grad_norm": 2.9760918617248535,
+      "learning_rate": 1.8259545506953285e-05,
+      "loss": 1.2673,
+      "step": 796
+    },
+    {
+      "epoch": 0.6032166508987701,
+      "grad_norm": 2.8577773571014404,
+      "learning_rate": 1.825501707496441e-05,
+      "loss": 1.264,
+      "step": 797
+    },
+    {
+      "epoch": 0.6039735099337749,
+      "grad_norm": 2.549546718597412,
+      "learning_rate": 1.825048332256201e-05,
+      "loss": 1.2228,
+      "step": 798
+    },
+    {
+      "epoch": 0.6047303689687795,
+      "grad_norm": 2.7687017917633057,
+      "learning_rate": 1.8245944252668462e-05,
+      "loss": 1.2522,
+      "step": 799
+    },
+    {
+      "epoch": 0.6054872280037843,
+      "grad_norm": 2.533287763595581,
+      "learning_rate": 1.824139986820959e-05,
+      "loss": 1.1939,
+      "step": 800
+    },
+    {
+      "epoch": 0.6062440870387891,
+      "grad_norm": 2.6402809619903564,
+      "learning_rate": 1.8236850172114633e-05,
+      "loss": 1.2417,
+      "step": 801
+    },
+    {
+      "epoch": 0.6070009460737937,
+      "grad_norm": 2.592946767807007,
+      "learning_rate": 1.8232295167316252e-05,
+      "loss": 1.2922,
+      "step": 802
+    },
+    {
+      "epoch": 0.6077578051087985,
+      "grad_norm": 2.6012048721313477,
+      "learning_rate": 1.8227734856750537e-05,
+      "loss": 1.2658,
+      "step": 803
+    },
+    {
+      "epoch": 0.6085146641438032,
+      "grad_norm": 2.737257242202759,
+      "learning_rate": 1.8223169243356995e-05,
+      "loss": 1.2955,
+      "step": 804
+    },
+    {
+      "epoch": 0.609271523178808,
+      "grad_norm": 2.8576440811157227,
+      "learning_rate": 1.8218598330078548e-05,
+      "loss": 1.261,
+      "step": 805
+    },
+    {
+      "epoch": 0.6100283822138127,
+      "grad_norm": 4.944385051727295,
+      "learning_rate": 1.8214022119861537e-05,
+      "loss": 1.2438,
+      "step": 806
+    },
+    {
+      "epoch": 0.6107852412488174,
+      "grad_norm": 2.8472225666046143,
+      "learning_rate": 1.820944061565572e-05,
+      "loss": 1.2305,
+      "step": 807
+    },
+    {
+      "epoch": 0.6115421002838222,
+      "grad_norm": 2.8943638801574707,
+      "learning_rate": 1.8204853820414267e-05,
+      "loss": 1.2608,
+      "step": 808
+    },
+    {
+      "epoch": 0.6122989593188268,
+      "grad_norm": 2.523142099380493,
+      "learning_rate": 1.820026173709375e-05,
+      "loss": 1.2721,
+      "step": 809
+    },
+    {
+      "epoch": 0.6130558183538316,
+      "grad_norm": 2.8089590072631836,
+      "learning_rate": 1.8195664368654157e-05,
+      "loss": 1.222,
+      "step": 810
+    },
+    {
+      "epoch": 0.6138126773888364,
+      "grad_norm": 2.9274590015411377,
+      "learning_rate": 1.8191061718058885e-05,
+      "loss": 1.2534,
+      "step": 811
+    },
+    {
+      "epoch": 0.614569536423841,
+      "grad_norm": 2.6819167137145996,
+      "learning_rate": 1.818645378827473e-05,
+      "loss": 1.2566,
+      "step": 812
+    },
+    {
+      "epoch": 0.6153263954588458,
+      "grad_norm": 2.5687010288238525,
+      "learning_rate": 1.8181840582271897e-05,
+      "loss": 1.2323,
+      "step": 813
+    },
+    {
+      "epoch": 0.6160832544938505,
+      "grad_norm": 2.636622428894043,
+      "learning_rate": 1.8177222103023983e-05,
+      "loss": 1.2007,
+      "step": 814
+    },
+    {
+      "epoch": 0.6168401135288553,
+      "grad_norm": 2.5585618019104004,
+      "learning_rate": 1.8172598353507988e-05,
+      "loss": 1.2169,
+      "step": 815
+    },
+    {
+      "epoch": 0.61759697256386,
+      "grad_norm": 2.880889415740967,
+      "learning_rate": 1.8167969336704322e-05,
+      "loss": 1.2211,
+      "step": 816
+    },
+    {
+      "epoch": 0.6183538315988647,
+      "grad_norm": 2.575530767440796,
+      "learning_rate": 1.8163335055596764e-05,
+      "loss": 1.2165,
+      "step": 817
+    },
+    {
+      "epoch": 0.6191106906338695,
+      "grad_norm": 2.65857195854187,
+      "learning_rate": 1.815869551317251e-05,
+      "loss": 1.2527,
+      "step": 818
+    },
+    {
+      "epoch": 0.6198675496688741,
+      "grad_norm": 2.7308692932128906,
+      "learning_rate": 1.8154050712422135e-05,
+      "loss": 1.245,
+      "step": 819
+    },
+    {
+      "epoch": 0.6206244087038789,
+      "grad_norm": 2.4128143787384033,
+      "learning_rate": 1.8149400656339606e-05,
+      "loss": 1.2274,
+      "step": 820
+    },
+    {
+      "epoch": 0.6213812677388836,
+      "grad_norm": 2.678269386291504,
+      "learning_rate": 1.8144745347922282e-05,
+      "loss": 1.2348,
+      "step": 821
+    },
+    {
+      "epoch": 0.6221381267738884,
+      "grad_norm": 2.4970011711120605,
+      "learning_rate": 1.81400847901709e-05,
+      "loss": 1.2525,
+      "step": 822
+    },
+    {
+      "epoch": 0.6228949858088931,
+      "grad_norm": 3.0284082889556885,
+      "learning_rate": 1.813541898608959e-05,
+      "loss": 1.2283,
+      "step": 823
+    },
+    {
+      "epoch": 0.6236518448438978,
+      "grad_norm": 2.5325472354888916,
+      "learning_rate": 1.813074793868585e-05,
+      "loss": 1.2177,
+      "step": 824
+    },
+    {
+      "epoch": 0.6244087038789026,
+      "grad_norm": 2.8422694206237793,
+      "learning_rate": 1.8126071650970566e-05,
+      "loss": 1.1957,
+      "step": 825
+    },
+    {
+      "epoch": 0.6251655629139072,
+      "grad_norm": 2.7805769443511963,
+      "learning_rate": 1.8121390125958012e-05,
+      "loss": 1.2406,
+      "step": 826
+    },
+    {
+      "epoch": 0.625922421948912,
+      "grad_norm": 3.035707473754883,
+      "learning_rate": 1.811670336666582e-05,
+      "loss": 1.2217,
+      "step": 827
+    },
+    {
+      "epoch": 0.6266792809839168,
+      "grad_norm": 2.6617417335510254,
+      "learning_rate": 1.8112011376115004e-05,
+      "loss": 1.2489,
+      "step": 828
+    },
+    {
+      "epoch": 0.6274361400189215,
+      "grad_norm": 2.593369722366333,
+      "learning_rate": 1.8107314157329953e-05,
+      "loss": 1.2582,
+      "step": 829
+    },
+    {
+      "epoch": 0.6281929990539262,
+      "grad_norm": 2.33566951751709,
+      "learning_rate": 1.810261171333842e-05,
+      "loss": 1.1726,
+      "step": 830
+    },
+    {
+      "epoch": 0.6289498580889309,
+      "grad_norm": 2.6399929523468018,
+      "learning_rate": 1.8097904047171525e-05,
+      "loss": 1.233,
+      "step": 831
+    },
+    {
+      "epoch": 0.6297067171239357,
+      "grad_norm": 2.833388328552246,
+      "learning_rate": 1.8093191161863765e-05,
+      "loss": 1.2465,
+      "step": 832
+    },
+    {
+      "epoch": 0.6304635761589404,
+      "grad_norm": 2.5618953704833984,
+      "learning_rate": 1.808847306045299e-05,
+      "loss": 1.2786,
+      "step": 833
+    },
+    {
+      "epoch": 0.6312204351939451,
+      "grad_norm": 2.449512004852295,
+      "learning_rate": 1.8083749745980417e-05,
+      "loss": 1.1866,
+      "step": 834
+    },
+    {
+      "epoch": 0.6319772942289499,
+      "grad_norm": 2.3261687755584717,
+      "learning_rate": 1.8079021221490623e-05,
+      "loss": 1.2293,
+      "step": 835
+    },
+    {
+      "epoch": 0.6327341532639545,
+      "grad_norm": 2.2670247554779053,
+      "learning_rate": 1.8074287490031544e-05,
+      "loss": 1.2307,
+      "step": 836
+    },
+    {
+      "epoch": 0.6334910122989593,
+      "grad_norm": 2.9090189933776855,
+      "learning_rate": 1.8069548554654465e-05,
+      "loss": 1.23,
+      "step": 837
+    },
+    {
+      "epoch": 0.6342478713339641,
+      "grad_norm": 2.2023513317108154,
+      "learning_rate": 1.8064804418414036e-05,
+      "loss": 1.2559,
+      "step": 838
+    },
+    {
+      "epoch": 0.6350047303689688,
+      "grad_norm": 2.3907856941223145,
+      "learning_rate": 1.8060055084368256e-05,
+      "loss": 1.1783,
+      "step": 839
+    },
+    {
+      "epoch": 0.6357615894039735,
+      "grad_norm": 2.7036445140838623,
+      "learning_rate": 1.805530055557847e-05,
+      "loss": 1.2268,
+      "step": 840
+    },
+    {
+      "epoch": 0.6365184484389782,
+      "grad_norm": 2.429286003112793,
+      "learning_rate": 1.805054083510938e-05,
+      "loss": 1.1904,
+      "step": 841
+    },
+    {
+      "epoch": 0.637275307473983,
+      "grad_norm": 2.644791603088379,
+      "learning_rate": 1.804577592602902e-05,
+      "loss": 1.1866,
+      "step": 842
+    },
+    {
+      "epoch": 0.6380321665089878,
+      "grad_norm": 2.7880802154541016,
+      "learning_rate": 1.804100583140879e-05,
+      "loss": 1.2817,
+      "step": 843
+    },
+    {
+      "epoch": 0.6387890255439924,
+      "grad_norm": 2.485358476638794,
+      "learning_rate": 1.8036230554323413e-05,
+      "loss": 1.281,
+      "step": 844
+    },
+    {
+      "epoch": 0.6395458845789972,
+      "grad_norm": 2.5849761962890625,
+      "learning_rate": 1.803145009785096e-05,
+      "loss": 1.248,
+      "step": 845
+    },
+    {
+      "epoch": 0.6403027436140019,
+      "grad_norm": 2.357409715652466,
+      "learning_rate": 1.8026664465072838e-05,
+      "loss": 1.2828,
+      "step": 846
+    },
+    {
+      "epoch": 0.6410596026490066,
+      "grad_norm": 2.4510414600372314,
+      "learning_rate": 1.80218736590738e-05,
+      "loss": 1.2275,
+      "step": 847
+    },
+    {
+      "epoch": 0.6418164616840114,
+      "grad_norm": 2.625035524368286,
+      "learning_rate": 1.8017077682941918e-05,
+      "loss": 1.2369,
+      "step": 848
+    },
+    {
+      "epoch": 0.6425733207190161,
+      "grad_norm": 2.4510104656219482,
+      "learning_rate": 1.8012276539768613e-05,
+      "loss": 1.2624,
+      "step": 849
+    },
+    {
+      "epoch": 0.6433301797540208,
+      "grad_norm": 2.6468582153320312,
+      "learning_rate": 1.800747023264862e-05,
+      "loss": 1.2964,
+      "step": 850
+    },
+    {
+      "epoch": 0.6440870387890255,
+      "grad_norm": 2.45991587638855,
+      "learning_rate": 1.800265876468002e-05,
+      "loss": 1.2359,
+      "step": 851
+    },
+    {
+      "epoch": 0.6448438978240303,
+      "grad_norm": 2.546734571456909,
+      "learning_rate": 1.799784213896421e-05,
+      "loss": 1.2124,
+      "step": 852
+    },
+    {
+      "epoch": 0.645600756859035,
+      "grad_norm": 2.265397787094116,
+      "learning_rate": 1.799302035860591e-05,
+      "loss": 1.1945,
+      "step": 853
+    },
+    {
+      "epoch": 0.6463576158940397,
+      "grad_norm": 2.4162395000457764,
+      "learning_rate": 1.7988193426713165e-05,
+      "loss": 1.2115,
+      "step": 854
+    },
+    {
+      "epoch": 0.6471144749290445,
+      "grad_norm": 2.2301483154296875,
+      "learning_rate": 1.7983361346397347e-05,
+      "loss": 1.2699,
+      "step": 855
+    },
+    {
+      "epoch": 0.6478713339640492,
+      "grad_norm": 2.2673699855804443,
+      "learning_rate": 1.797852412077314e-05,
+      "loss": 1.2525,
+      "step": 856
+    },
+    {
+      "epoch": 0.6486281929990539,
+      "grad_norm": 2.5041098594665527,
+      "learning_rate": 1.7973681752958543e-05,
+      "loss": 1.231,
+      "step": 857
+    },
+    {
+      "epoch": 0.6493850520340586,
+      "grad_norm": 2.5438284873962402,
+      "learning_rate": 1.7968834246074875e-05,
+      "loss": 1.2316,
+      "step": 858
+    },
+    {
+      "epoch": 0.6501419110690634,
+      "grad_norm": 2.4436419010162354,
+      "learning_rate": 1.7963981603246762e-05,
+      "loss": 1.2461,
+      "step": 859
+    },
+    {
+      "epoch": 0.6508987701040682,
+      "grad_norm": 2.3260018825531006,
+      "learning_rate": 1.795912382760215e-05,
+      "loss": 1.2575,
+      "step": 860
+    },
+    {
+      "epoch": 0.6516556291390728,
+      "grad_norm": 2.527569532394409,
+      "learning_rate": 1.7954260922272278e-05,
+      "loss": 1.2552,
+      "step": 861
+    },
+    {
+      "epoch": 0.6524124881740776,
+      "grad_norm": 2.5068411827087402,
+      "learning_rate": 1.7949392890391706e-05,
+      "loss": 1.2439,
+      "step": 862
+    },
+    {
+      "epoch": 0.6531693472090823,
+      "grad_norm": 2.8131117820739746,
+      "learning_rate": 1.7944519735098295e-05,
+      "loss": 1.2669,
+      "step": 863
+    },
+    {
+      "epoch": 0.653926206244087,
+      "grad_norm": 2.368083953857422,
+      "learning_rate": 1.79396414595332e-05,
+      "loss": 1.273,
+      "step": 864
+    },
+    {
+      "epoch": 0.6546830652790918,
+      "grad_norm": 2.4757819175720215,
+      "learning_rate": 1.7934758066840893e-05,
+      "loss": 1.2652,
+      "step": 865
+    },
+    {
+      "epoch": 0.6554399243140965,
+      "grad_norm": 2.7727437019348145,
+      "learning_rate": 1.7929869560169123e-05,
+      "loss": 1.2661,
+      "step": 866
+    },
+    {
+      "epoch": 0.6561967833491013,
+      "grad_norm": 2.5417017936706543,
+      "learning_rate": 1.7924975942668954e-05,
+      "loss": 1.2624,
+      "step": 867
+    },
+    {
+      "epoch": 0.6569536423841059,
+      "grad_norm": 3.0404696464538574,
+      "learning_rate": 1.792007721749474e-05,
+      "loss": 1.2149,
+      "step": 868
+    },
+    {
+      "epoch": 0.6577105014191107,
+      "grad_norm": 2.528648853302002,
+      "learning_rate": 1.7915173387804115e-05,
+      "loss": 1.2536,
+      "step": 869
+    },
+    {
+      "epoch": 0.6584673604541155,
+      "grad_norm": 2.5994584560394287,
+      "learning_rate": 1.791026445675802e-05,
+      "loss": 1.2146,
+      "step": 870
+    },
+    {
+      "epoch": 0.6592242194891201,
+      "grad_norm": 2.523890495300293,
+      "learning_rate": 1.7905350427520672e-05,
+      "loss": 1.2599,
+      "step": 871
+    },
+    {
+      "epoch": 0.6599810785241249,
+      "grad_norm": 3.055417537689209,
+      "learning_rate": 1.7900431303259585e-05,
+      "loss": 1.2447,
+      "step": 872
+    },
+    {
+      "epoch": 0.6607379375591296,
+      "grad_norm": 2.5144965648651123,
+      "learning_rate": 1.789550708714555e-05,
+      "loss": 1.2022,
+      "step": 873
+    },
+    {
+      "epoch": 0.6614947965941343,
+      "grad_norm": 2.5344860553741455,
+      "learning_rate": 1.789057778235264e-05,
+      "loss": 1.2578,
+      "step": 874
+    },
+    {
+      "epoch": 0.6622516556291391,
+      "grad_norm": 2.7370986938476562,
+      "learning_rate": 1.7885643392058207e-05,
+      "loss": 1.1964,
+      "step": 875
+    },
+    {
+      "epoch": 0.6630085146641438,
+      "grad_norm": 2.831005573272705,
+      "learning_rate": 1.7880703919442885e-05,
+      "loss": 1.2218,
+      "step": 876
+    },
+    {
+      "epoch": 0.6637653736991486,
+      "grad_norm": 2.2660505771636963,
+      "learning_rate": 1.787575936769059e-05,
+      "loss": 1.232,
+      "step": 877
+    },
+    {
+      "epoch": 0.6645222327341532,
+      "grad_norm": 2.5577943325042725,
+      "learning_rate": 1.78708097399885e-05,
+      "loss": 1.289,
+      "step": 878
+    },
+    {
+      "epoch": 0.665279091769158,
+      "grad_norm": 2.8066608905792236,
+      "learning_rate": 1.786585503952707e-05,
+      "loss": 1.2051,
+      "step": 879
+    },
+    {
+      "epoch": 0.6660359508041628,
+      "grad_norm": 2.683680295944214,
+      "learning_rate": 1.786089526950002e-05,
+      "loss": 1.2343,
+      "step": 880
+    },
+    {
+      "epoch": 0.6667928098391674,
+      "grad_norm": 2.571253538131714,
+      "learning_rate": 1.785593043310434e-05,
+      "loss": 1.2279,
+      "step": 881
+    },
+    {
+      "epoch": 0.6675496688741722,
+      "grad_norm": 2.2818214893341064,
+      "learning_rate": 1.78509605335403e-05,
+      "loss": 1.2346,
+      "step": 882
+    },
+    {
+      "epoch": 0.6683065279091769,
+      "grad_norm": 2.427520513534546,
+      "learning_rate": 1.7845985574011413e-05,
+      "loss": 1.251,
+      "step": 883
+    },
+    {
+      "epoch": 0.6690633869441817,
+      "grad_norm": 2.615901231765747,
+      "learning_rate": 1.784100555772446e-05,
+      "loss": 1.2697,
+      "step": 884
+    },
+    {
+      "epoch": 0.6698202459791863,
+      "grad_norm": 2.3778128623962402,
+      "learning_rate": 1.7836020487889495e-05,
+      "loss": 1.2291,
+      "step": 885
+    },
+    {
+      "epoch": 0.6705771050141911,
+      "grad_norm": 2.4669504165649414,
+      "learning_rate": 1.7831030367719802e-05,
+      "loss": 1.2365,
+      "step": 886
+    },
+    {
+      "epoch": 0.6713339640491959,
+      "grad_norm": 2.397721290588379,
+      "learning_rate": 1.782603520043195e-05,
+      "loss": 1.2718,
+      "step": 887
+    },
+    {
+      "epoch": 0.6720908230842005,
+      "grad_norm": 2.323598623275757,
+      "learning_rate": 1.782103498924574e-05,
+      "loss": 1.2706,
+      "step": 888
+    },
+    {
+      "epoch": 0.6728476821192053,
+      "grad_norm": 2.592615842819214,
+      "learning_rate": 1.7816029737384234e-05,
+      "loss": 1.2821,
+      "step": 889
+    },
+    {
+      "epoch": 0.67360454115421,
+      "grad_norm": 2.552388906478882,
+      "learning_rate": 1.7811019448073742e-05,
+      "loss": 1.2075,
+      "step": 890
+    },
+    {
+      "epoch": 0.6743614001892148,
+      "grad_norm": 2.659424304962158,
+      "learning_rate": 1.7806004124543818e-05,
+      "loss": 1.2365,
+      "step": 891
+    },
+    {
+      "epoch": 0.6751182592242195,
+      "grad_norm": 2.596625328063965,
+      "learning_rate": 1.7800983770027266e-05,
+      "loss": 1.2685,
+      "step": 892
+    },
+    {
+      "epoch": 0.6758751182592242,
+      "grad_norm": 2.485259771347046,
+      "learning_rate": 1.779595838776013e-05,
+      "loss": 1.2453,
+      "step": 893
+    },
+    {
+      "epoch": 0.676631977294229,
+      "grad_norm": 2.3858642578125,
+      "learning_rate": 1.7790927980981687e-05,
+      "loss": 1.1896,
+      "step": 894
+    },
+    {
+      "epoch": 0.6773888363292336,
+      "grad_norm": 2.53601336479187,
+      "learning_rate": 1.7785892552934468e-05,
+      "loss": 1.2533,
+      "step": 895
+    },
+    {
+      "epoch": 0.6781456953642384,
+      "grad_norm": 2.7505080699920654,
+      "learning_rate": 1.778085210686423e-05,
+      "loss": 1.2449,
+      "step": 896
+    },
+    {
+      "epoch": 0.6789025543992432,
+      "grad_norm": 2.4080655574798584,
+      "learning_rate": 1.7775806646019974e-05,
+      "loss": 1.1985,
+      "step": 897
+    },
+    {
+      "epoch": 0.6796594134342478,
+      "grad_norm": 2.742640972137451,
+      "learning_rate": 1.7770756173653923e-05,
+      "loss": 1.2434,
+      "step": 898
+    },
+    {
+      "epoch": 0.6804162724692526,
+      "grad_norm": 2.377990484237671,
+      "learning_rate": 1.776570069302153e-05,
+      "loss": 1.1726,
+      "step": 899
+    },
+    {
+      "epoch": 0.6811731315042573,
+      "grad_norm": 2.35687518119812,
+      "learning_rate": 1.7760640207381486e-05,
+      "loss": 1.2189,
+      "step": 900
+    },
+    {
+      "epoch": 0.6819299905392621,
+      "grad_norm": 2.576018810272217,
+      "learning_rate": 1.77555747199957e-05,
+      "loss": 1.2318,
+      "step": 901
+    },
+    {
+      "epoch": 0.6826868495742668,
+      "grad_norm": 2.3314318656921387,
+      "learning_rate": 1.7750504234129312e-05,
+      "loss": 1.1889,
+      "step": 902
+    },
+    {
+      "epoch": 0.6834437086092715,
+      "grad_norm": 2.3357717990875244,
+      "learning_rate": 1.7745428753050675e-05,
+      "loss": 1.2168,
+      "step": 903
+    },
+    {
+      "epoch": 0.6842005676442763,
+      "grad_norm": 2.2540555000305176,
+      "learning_rate": 1.774034828003137e-05,
+      "loss": 1.2017,
+      "step": 904
+    },
+    {
+      "epoch": 0.684957426679281,
+      "grad_norm": 2.325144052505493,
+      "learning_rate": 1.773526281834619e-05,
+      "loss": 1.2409,
+      "step": 905
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 2.731501340866089,
+      "learning_rate": 1.7730172371273147e-05,
+      "loss": 1.2765,
+      "step": 906
+    },
+    {
+      "epoch": 0.6864711447492905,
+      "grad_norm": 2.3535265922546387,
+      "learning_rate": 1.7725076942093468e-05,
+      "loss": 1.2353,
+      "step": 907
+    },
+    {
+      "epoch": 0.6872280037842952,
+      "grad_norm": 2.852663040161133,
+      "learning_rate": 1.7719976534091584e-05,
+      "loss": 1.2761,
+      "step": 908
+    },
+    {
+      "epoch": 0.6879848628192999,
+      "grad_norm": 2.5675928592681885,
+      "learning_rate": 1.7714871150555146e-05,
+      "loss": 1.1906,
+      "step": 909
+    },
+    {
+      "epoch": 0.6887417218543046,
+      "grad_norm": 2.1720049381256104,
+      "learning_rate": 1.7709760794775e-05,
+      "loss": 1.2057,
+      "step": 910
+    },
+    {
+      "epoch": 0.6894985808893094,
+      "grad_norm": 2.567373514175415,
+      "learning_rate": 1.7704645470045213e-05,
+      "loss": 1.2365,
+      "step": 911
+    },
+    {
+      "epoch": 0.6902554399243142,
+      "grad_norm": 2.38577938079834,
+      "learning_rate": 1.7699525179663034e-05,
+      "loss": 1.2047,
+      "step": 912
+    },
+    {
+      "epoch": 0.6910122989593188,
+      "grad_norm": 2.3595142364501953,
+      "learning_rate": 1.7694399926928932e-05,
+      "loss": 1.2329,
+      "step": 913
+    },
+    {
+      "epoch": 0.6917691579943236,
+      "grad_norm": 2.7524566650390625,
+      "learning_rate": 1.7689269715146562e-05,
+      "loss": 1.2461,
+      "step": 914
+    },
+    {
+      "epoch": 0.6925260170293283,
+      "grad_norm": 2.2120566368103027,
+      "learning_rate": 1.768413454762278e-05,
+      "loss": 1.2232,
+      "step": 915
+    },
+    {
+      "epoch": 0.693282876064333,
+      "grad_norm": 2.491506338119507,
+      "learning_rate": 1.767899442766764e-05,
+      "loss": 1.2185,
+      "step": 916
+    },
+    {
+      "epoch": 0.6940397350993377,
+      "grad_norm": 2.299386501312256,
+      "learning_rate": 1.7673849358594387e-05,
+      "loss": 1.2146,
+      "step": 917
+    },
+    {
+      "epoch": 0.6947965941343425,
+      "grad_norm": 2.367396831512451,
+      "learning_rate": 1.766869934371945e-05,
+      "loss": 1.2666,
+      "step": 918
+    },
+    {
+      "epoch": 0.6955534531693472,
+      "grad_norm": 2.379352331161499,
+      "learning_rate": 1.766354438636245e-05,
+      "loss": 1.2295,
+      "step": 919
+    },
+    {
+      "epoch": 0.6963103122043519,
+      "grad_norm": 2.91322660446167,
+      "learning_rate": 1.7658384489846197e-05,
+      "loss": 1.2211,
+      "step": 920
+    },
+    {
+      "epoch": 0.6970671712393567,
+      "grad_norm": 2.3727736473083496,
+      "learning_rate": 1.7653219657496675e-05,
+      "loss": 1.2478,
+      "step": 921
+    },
+    {
+      "epoch": 0.6978240302743614,
+      "grad_norm": 2.3029327392578125,
+      "learning_rate": 1.7648049892643064e-05,
+      "loss": 1.238,
+      "step": 922
+    },
+    {
+      "epoch": 0.6985808893093661,
+      "grad_norm": 2.3356475830078125,
+      "learning_rate": 1.7642875198617715e-05,
+      "loss": 1.1932,
+      "step": 923
+    },
+    {
+      "epoch": 0.6993377483443709,
+      "grad_norm": 2.5331709384918213,
+      "learning_rate": 1.7637695578756148e-05,
+      "loss": 1.1822,
+      "step": 924
+    },
+    {
+      "epoch": 0.7000946073793756,
+      "grad_norm": 2.714674949645996,
+      "learning_rate": 1.7632511036397078e-05,
+      "loss": 1.2454,
+      "step": 925
+    },
+    {
+      "epoch": 0.7008514664143803,
+      "grad_norm": 2.497758388519287,
+      "learning_rate": 1.7627321574882373e-05,
+      "loss": 1.2552,
+      "step": 926
+    },
+    {
+      "epoch": 0.701608325449385,
+      "grad_norm": 2.6237785816192627,
+      "learning_rate": 1.7622127197557085e-05,
+      "loss": 1.2334,
+      "step": 927
+    },
+    {
+      "epoch": 0.7023651844843898,
+      "grad_norm": 2.4308512210845947,
+      "learning_rate": 1.7616927907769436e-05,
+      "loss": 1.2516,
+      "step": 928
+    },
+    {
+      "epoch": 0.7031220435193946,
+      "grad_norm": 2.1913723945617676,
+      "learning_rate": 1.7611723708870797e-05,
+      "loss": 1.267,
+      "step": 929
+    },
+    {
+      "epoch": 0.7038789025543992,
+      "grad_norm": 2.6569485664367676,
+      "learning_rate": 1.7606514604215723e-05,
+      "loss": 1.2301,
+      "step": 930
+    },
+    {
+      "epoch": 0.704635761589404,
+      "grad_norm": 2.4195547103881836,
+      "learning_rate": 1.7601300597161918e-05,
+      "loss": 1.2464,
+      "step": 931
+    },
+    {
+      "epoch": 0.7053926206244087,
+      "grad_norm": 2.580186128616333,
+      "learning_rate": 1.7596081691070262e-05,
+      "loss": 1.2432,
+      "step": 932
+    },
+    {
+      "epoch": 0.7061494796594134,
+      "grad_norm": 2.4679551124572754,
+      "learning_rate": 1.759085788930477e-05,
+      "loss": 1.23,
+      "step": 933
+    },
+    {
+      "epoch": 0.7069063386944182,
+      "grad_norm": 2.256150722503662,
+      "learning_rate": 1.7585629195232633e-05,
+      "loss": 1.2362,
+      "step": 934
+    },
+    {
+      "epoch": 0.7076631977294229,
+      "grad_norm": 2.7825927734375,
+      "learning_rate": 1.7580395612224184e-05,
+      "loss": 1.2542,
+      "step": 935
+    },
+    {
+      "epoch": 0.7084200567644277,
+      "grad_norm": 2.568265676498413,
+      "learning_rate": 1.757515714365291e-05,
+      "loss": 1.3004,
+      "step": 936
+    },
+    {
+      "epoch": 0.7091769157994323,
+      "grad_norm": 2.422884702682495,
+      "learning_rate": 1.7569913792895455e-05,
+      "loss": 1.2382,
+      "step": 937
+    },
+    {
+      "epoch": 0.7099337748344371,
+      "grad_norm": 2.4318430423736572,
+      "learning_rate": 1.7564665563331597e-05,
+      "loss": 1.2421,
+      "step": 938
+    },
+    {
+      "epoch": 0.7106906338694419,
+      "grad_norm": 2.4188950061798096,
+      "learning_rate": 1.755941245834426e-05,
+      "loss": 1.244,
+      "step": 939
+    },
+    {
+      "epoch": 0.7114474929044465,
+      "grad_norm": 2.37963604927063,
+      "learning_rate": 1.7554154481319523e-05,
+      "loss": 1.2137,
+      "step": 940
+    },
+    {
+      "epoch": 0.7122043519394513,
+      "grad_norm": 2.5548665523529053,
+      "learning_rate": 1.7548891635646595e-05,
+      "loss": 1.241,
+      "step": 941
+    },
+    {
+      "epoch": 0.712961210974456,
+      "grad_norm": 2.384345054626465,
+      "learning_rate": 1.7543623924717827e-05,
+      "loss": 1.2299,
+      "step": 942
+    },
+    {
+      "epoch": 0.7137180700094607,
+      "grad_norm": 2.459399461746216,
+      "learning_rate": 1.7538351351928705e-05,
+      "loss": 1.2242,
+      "step": 943
+    },
+    {
+      "epoch": 0.7144749290444655,
+      "grad_norm": 2.2715373039245605,
+      "learning_rate": 1.7533073920677847e-05,
+      "loss": 1.2264,
+      "step": 944
+    },
+    {
+      "epoch": 0.7152317880794702,
+      "grad_norm": 2.423783302307129,
+      "learning_rate": 1.752779163436701e-05,
+      "loss": 1.2061,
+      "step": 945
+    },
+    {
+      "epoch": 0.715988647114475,
+      "grad_norm": 2.368046283721924,
+      "learning_rate": 1.7522504496401068e-05,
+      "loss": 1.2568,
+      "step": 946
+    },
+    {
+      "epoch": 0.7167455061494796,
+      "grad_norm": 2.7439255714416504,
+      "learning_rate": 1.7517212510188034e-05,
+      "loss": 1.2123,
+      "step": 947
+    },
+    {
+      "epoch": 0.7175023651844844,
+      "grad_norm": 2.3615167140960693,
+      "learning_rate": 1.751191567913904e-05,
+      "loss": 1.2188,
+      "step": 948
+    },
+    {
+      "epoch": 0.7182592242194891,
+      "grad_norm": 2.394190549850464,
+      "learning_rate": 1.7506614006668346e-05,
+      "loss": 1.21,
+      "step": 949
+    },
+    {
+      "epoch": 0.7190160832544938,
+      "grad_norm": 2.2254350185394287,
+      "learning_rate": 1.7501307496193324e-05,
+      "loss": 1.2306,
+      "step": 950
+    },
+    {
+      "epoch": 0.7197729422894986,
+      "grad_norm": 2.734381675720215,
+      "learning_rate": 1.749599615113447e-05,
+      "loss": 1.2367,
+      "step": 951
+    },
+    {
+      "epoch": 0.7205298013245033,
+      "grad_norm": 2.4236867427825928,
+      "learning_rate": 1.7490679974915404e-05,
+      "loss": 1.2013,
+      "step": 952
+    },
+    {
+      "epoch": 0.7212866603595081,
+      "grad_norm": 2.4105286598205566,
+      "learning_rate": 1.748535897096284e-05,
+      "loss": 1.1849,
+      "step": 953
+    },
+    {
+      "epoch": 0.7220435193945127,
+      "grad_norm": 3.396277666091919,
+      "learning_rate": 1.7480033142706626e-05,
+      "loss": 1.2018,
+      "step": 954
+    },
+    {
+      "epoch": 0.7228003784295175,
+      "grad_norm": 2.49308180809021,
+      "learning_rate": 1.7474702493579704e-05,
+      "loss": 1.2533,
+      "step": 955
+    },
+    {
+      "epoch": 0.7235572374645223,
+      "grad_norm": 2.2357521057128906,
+      "learning_rate": 1.7469367027018134e-05,
+      "loss": 1.253,
+      "step": 956
+    },
+    {
+      "epoch": 0.7243140964995269,
+      "grad_norm": 2.3083794116973877,
+      "learning_rate": 1.746402674646107e-05,
+      "loss": 1.2062,
+      "step": 957
+    },
+    {
+      "epoch": 0.7250709555345317,
+      "grad_norm": 2.087985038757324,
+      "learning_rate": 1.745868165535078e-05,
+      "loss": 1.2146,
+      "step": 958
+    },
+    {
+      "epoch": 0.7258278145695364,
+      "grad_norm": 2.1703999042510986,
+      "learning_rate": 1.7453331757132627e-05,
+      "loss": 1.2593,
+      "step": 959
+    },
+    {
+      "epoch": 0.7265846736045412,
+      "grad_norm": 2.644440174102783,
+      "learning_rate": 1.7447977055255076e-05,
+      "loss": 1.2744,
+      "step": 960
+    },
+    {
+      "epoch": 0.7273415326395459,
+      "grad_norm": 2.2902777194976807,
+      "learning_rate": 1.744261755316968e-05,
+      "loss": 1.2336,
+      "step": 961
+    },
+    {
+      "epoch": 0.7280983916745506,
+      "grad_norm": 2.1898083686828613,
+      "learning_rate": 1.7437253254331103e-05,
+      "loss": 1.1872,
+      "step": 962
+    },
+    {
+      "epoch": 0.7288552507095554,
+      "grad_norm": 2.192096710205078,
+      "learning_rate": 1.7431884162197076e-05,
+      "loss": 1.1904,
+      "step": 963
+    },
+    {
+      "epoch": 0.72961210974456,
+      "grad_norm": 2.344484806060791,
+      "learning_rate": 1.7426510280228447e-05,
+      "loss": 1.2086,
+      "step": 964
+    },
+    {
+      "epoch": 0.7303689687795648,
+      "grad_norm": 2.4366836547851562,
+      "learning_rate": 1.742113161188913e-05,
+      "loss": 1.2367,
+      "step": 965
+    },
+    {
+      "epoch": 0.7311258278145696,
+      "grad_norm": 2.3846473693847656,
+      "learning_rate": 1.7415748160646136e-05,
+      "loss": 1.2182,
+      "step": 966
+    },
+    {
+      "epoch": 0.7318826868495742,
+      "grad_norm": 2.432124614715576,
+      "learning_rate": 1.7410359929969555e-05,
+      "loss": 1.2345,
+      "step": 967
+    },
+    {
+      "epoch": 0.732639545884579,
+      "grad_norm": 2.427494525909424,
+      "learning_rate": 1.7404966923332558e-05,
+      "loss": 1.2284,
+      "step": 968
+    },
+    {
+      "epoch": 0.7333964049195837,
+      "grad_norm": 2.3191261291503906,
+      "learning_rate": 1.73995691442114e-05,
+      "loss": 1.2092,
+      "step": 969
+    },
+    {
+      "epoch": 0.7341532639545885,
+      "grad_norm": 2.3739922046661377,
+      "learning_rate": 1.7394166596085393e-05,
+      "loss": 1.2276,
+      "step": 970
+    },
+    {
+      "epoch": 0.7349101229895932,
+      "grad_norm": 2.29589581489563,
+      "learning_rate": 1.7388759282436953e-05,
+      "loss": 1.2422,
+      "step": 971
+    },
+    {
+      "epoch": 0.7356669820245979,
+      "grad_norm": 2.3834304809570312,
+      "learning_rate": 1.7383347206751542e-05,
+      "loss": 1.252,
+      "step": 972
+    },
+    {
+      "epoch": 0.7364238410596027,
+      "grad_norm": 2.2572319507598877,
+      "learning_rate": 1.7377930372517705e-05,
+      "loss": 1.2296,
+      "step": 973
+    },
+    {
+      "epoch": 0.7371807000946073,
+      "grad_norm": 2.6052353382110596,
+      "learning_rate": 1.7372508783227052e-05,
+      "loss": 1.2131,
+      "step": 974
+    },
+    {
+      "epoch": 0.7379375591296121,
+      "grad_norm": 2.4882845878601074,
+      "learning_rate": 1.7367082442374255e-05,
+      "loss": 1.2259,
+      "step": 975
+    },
+    {
+      "epoch": 0.7386944181646169,
+      "grad_norm": 2.301111936569214,
+      "learning_rate": 1.7361651353457053e-05,
+      "loss": 1.191,
+      "step": 976
+    },
+    {
+      "epoch": 0.7394512771996216,
+      "grad_norm": 2.496601104736328,
+      "learning_rate": 1.7356215519976236e-05,
+      "loss": 1.1749,
+      "step": 977
+    },
+    {
+      "epoch": 0.7402081362346263,
+      "grad_norm": 2.4782116413116455,
+      "learning_rate": 1.7350774945435667e-05,
+      "loss": 1.2282,
+      "step": 978
+    },
+    {
+      "epoch": 0.740964995269631,
+      "grad_norm": 2.3096814155578613,
+      "learning_rate": 1.7345329633342253e-05,
+      "loss": 1.1811,
+      "step": 979
+    },
+    {
+      "epoch": 0.7417218543046358,
+      "grad_norm": 2.5999755859375,
+      "learning_rate": 1.7339879587205966e-05,
+      "loss": 1.2081,
+      "step": 980
+    },
+    {
+      "epoch": 0.7424787133396404,
+      "grad_norm": 2.3727262020111084,
+      "learning_rate": 1.733442481053981e-05,
+      "loss": 1.2392,
+      "step": 981
+    },
+    {
+      "epoch": 0.7432355723746452,
+      "grad_norm": 2.621267318725586,
+      "learning_rate": 1.7328965306859864e-05,
+      "loss": 1.1715,
+      "step": 982
+    },
+    {
+      "epoch": 0.74399243140965,
+      "grad_norm": 2.786910057067871,
+      "learning_rate": 1.732350107968523e-05,
+      "loss": 1.2307,
+      "step": 983
+    },
+    {
+      "epoch": 0.7447492904446547,
+      "grad_norm": 2.5757007598876953,
+      "learning_rate": 1.7318032132538078e-05,
+      "loss": 1.204,
+      "step": 984
+    },
+    {
+      "epoch": 0.7455061494796594,
+      "grad_norm": 2.4591543674468994,
+      "learning_rate": 1.7312558468943595e-05,
+      "loss": 1.1665,
+      "step": 985
+    },
+    {
+      "epoch": 0.7462630085146641,
+      "grad_norm": 2.4593307971954346,
+      "learning_rate": 1.730708009243003e-05,
+      "loss": 1.2571,
+      "step": 986
+    },
+    {
+      "epoch": 0.7470198675496689,
+      "grad_norm": 2.507080554962158,
+      "learning_rate": 1.7301597006528654e-05,
+      "loss": 1.2222,
+      "step": 987
+    },
+    {
+      "epoch": 0.7477767265846736,
+      "grad_norm": 2.445662498474121,
+      "learning_rate": 1.7296109214773782e-05,
+      "loss": 1.2066,
+      "step": 988
+    },
+    {
+      "epoch": 0.7485335856196783,
+      "grad_norm": 2.341787099838257,
+      "learning_rate": 1.7290616720702768e-05,
+      "loss": 1.2395,
+      "step": 989
+    },
+    {
+      "epoch": 0.7492904446546831,
+      "grad_norm": 2.569960832595825,
+      "learning_rate": 1.728511952785598e-05,
+      "loss": 1.241,
+      "step": 990
+    },
+    {
+      "epoch": 0.7500473036896877,
+      "grad_norm": 2.5241215229034424,
+      "learning_rate": 1.7279617639776836e-05,
+      "loss": 1.2231,
+      "step": 991
+    },
+    {
+      "epoch": 0.7508041627246925,
+      "grad_norm": 2.4361581802368164,
+      "learning_rate": 1.727411106001176e-05,
+      "loss": 1.2381,
+      "step": 992
+    },
+    {
+      "epoch": 0.7515610217596973,
+      "grad_norm": 2.3338370323181152,
+      "learning_rate": 1.7268599792110213e-05,
+      "loss": 1.2526,
+      "step": 993
+    },
+    {
+      "epoch": 0.752317880794702,
+      "grad_norm": 2.398029327392578,
+      "learning_rate": 1.726308383962467e-05,
+      "loss": 1.2405,
+      "step": 994
+    },
+    {
+      "epoch": 0.7530747398297067,
+      "grad_norm": 2.437852382659912,
+      "learning_rate": 1.7257563206110636e-05,
+      "loss": 1.2553,
+      "step": 995
+    },
+    {
+      "epoch": 0.7538315988647114,
+      "grad_norm": 2.763335704803467,
+      "learning_rate": 1.7252037895126622e-05,
+      "loss": 1.2342,
+      "step": 996
+    },
+    {
+      "epoch": 0.7545884578997162,
+      "grad_norm": 2.5191261768341064,
+      "learning_rate": 1.7246507910234162e-05,
+      "loss": 1.2188,
+      "step": 997
+    },
+    {
+      "epoch": 0.755345316934721,
+      "grad_norm": 2.5067646503448486,
+      "learning_rate": 1.72409732549978e-05,
+      "loss": 1.1707,
+      "step": 998
+    },
+    {
+      "epoch": 0.7561021759697256,
+      "grad_norm": 2.400637626647949,
+      "learning_rate": 1.7235433932985092e-05,
+      "loss": 1.1599,
+      "step": 999
+    },
+    {
+      "epoch": 0.7568590350047304,
+      "grad_norm": 2.410027027130127,
+      "learning_rate": 1.7229889947766597e-05,
+      "loss": 1.2442,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7576158940397351,
+      "grad_norm": 2.3706107139587402,
+      "learning_rate": 1.7224341302915885e-05,
+      "loss": 1.2264,
+      "step": 1001
+    },
+    {
+      "epoch": 0.7583727530747398,
+      "grad_norm": 2.4948031902313232,
+      "learning_rate": 1.7218788002009527e-05,
+      "loss": 1.2505,
+      "step": 1002
+    },
+    {
+      "epoch": 0.7591296121097446,
+      "grad_norm": 2.4337100982666016,
+      "learning_rate": 1.7213230048627093e-05,
+      "loss": 1.2416,
+      "step": 1003
+    },
+    {
+      "epoch": 0.7598864711447493,
+      "grad_norm": 2.2913546562194824,
+      "learning_rate": 1.7207667446351165e-05,
+      "loss": 1.22,
+      "step": 1004
+    },
+    {
+      "epoch": 0.760643330179754,
+      "grad_norm": 2.4365074634552,
+      "learning_rate": 1.72021001987673e-05,
+      "loss": 1.2398,
+      "step": 1005
+    },
+    {
+      "epoch": 0.7614001892147587,
+      "grad_norm": 2.4662392139434814,
+      "learning_rate": 1.7196528309464067e-05,
+      "loss": 1.2372,
+      "step": 1006
+    },
+    {
+      "epoch": 0.7621570482497635,
+      "grad_norm": 2.5085933208465576,
+      "learning_rate": 1.719095178203302e-05,
+      "loss": 1.2408,
+      "step": 1007
+    },
+    {
+      "epoch": 0.7629139072847683,
+      "grad_norm": 2.447695016860962,
+      "learning_rate": 1.7185370620068705e-05,
+      "loss": 1.2062,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7636707663197729,
+      "grad_norm": 2.8076727390289307,
+      "learning_rate": 1.717978482716865e-05,
+      "loss": 1.2051,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7644276253547777,
+      "grad_norm": 2.6247246265411377,
+      "learning_rate": 1.7174194406933377e-05,
+      "loss": 1.1861,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7651844843897824,
+      "grad_norm": 2.6273937225341797,
+      "learning_rate": 1.7168599362966382e-05,
+      "loss": 1.1919,
+      "step": 1011
+    },
+    {
+      "epoch": 0.7659413434247871,
+      "grad_norm": 2.363234281539917,
+      "learning_rate": 1.7162999698874144e-05,
+      "loss": 1.203,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7666982024597918,
+      "grad_norm": 2.4418020248413086,
+      "learning_rate": 1.7157395418266125e-05,
+      "loss": 1.2146,
+      "step": 1013
+    },
+    {
+      "epoch": 0.7674550614947966,
+      "grad_norm": 2.4737863540649414,
+      "learning_rate": 1.7151786524754755e-05,
+      "loss": 1.2149,
+      "step": 1014
+    },
+    {
+      "epoch": 0.7682119205298014,
+      "grad_norm": 2.2613844871520996,
+      "learning_rate": 1.7146173021955444e-05,
+      "loss": 1.2276,
+      "step": 1015
+    },
+    {
+      "epoch": 0.768968779564806,
+      "grad_norm": 2.626579523086548,
+      "learning_rate": 1.714055491348657e-05,
+      "loss": 1.2384,
+      "step": 1016
+    },
+    {
+      "epoch": 0.7697256385998108,
+      "grad_norm": 2.406792163848877,
+      "learning_rate": 1.7134932202969482e-05,
+      "loss": 1.2285,
+      "step": 1017
+    },
+    {
+      "epoch": 0.7704824976348155,
+      "grad_norm": 2.456866979598999,
+      "learning_rate": 1.7129304894028483e-05,
+      "loss": 1.1853,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7712393566698202,
+      "grad_norm": 2.5044846534729004,
+      "learning_rate": 1.7123672990290864e-05,
+      "loss": 1.212,
+      "step": 1019
+    },
+    {
+      "epoch": 0.771996215704825,
+      "grad_norm": 2.1986587047576904,
+      "learning_rate": 1.7118036495386856e-05,
+      "loss": 1.2106,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7727530747398297,
+      "grad_norm": 2.4531362056732178,
+      "learning_rate": 1.7112395412949662e-05,
+      "loss": 1.2466,
+      "step": 1021
+    },
+    {
+      "epoch": 0.7735099337748345,
+      "grad_norm": 2.2251899242401123,
+      "learning_rate": 1.7106749746615437e-05,
+      "loss": 1.1857,
+      "step": 1022
+    },
+    {
+      "epoch": 0.7742667928098391,
+      "grad_norm": 2.2850799560546875,
+      "learning_rate": 1.7101099500023287e-05,
+      "loss": 1.2499,
+      "step": 1023
+    },
+    {
+      "epoch": 0.7750236518448439,
+      "grad_norm": 2.3555169105529785,
+      "learning_rate": 1.709544467681528e-05,
+      "loss": 1.2139,
+      "step": 1024
+    },
+    {
+      "epoch": 0.7757805108798487,
+      "grad_norm": 2.54640531539917,
+      "learning_rate": 1.7089785280636428e-05,
+      "loss": 1.2121,
+      "step": 1025
+    },
+    {
+      "epoch": 0.7765373699148533,
+      "grad_norm": 2.3403546810150146,
+      "learning_rate": 1.708412131513469e-05,
+      "loss": 1.2294,
+      "step": 1026
+    },
+    {
+      "epoch": 0.7772942289498581,
+      "grad_norm": 2.450343370437622,
+      "learning_rate": 1.707845278396097e-05,
+      "loss": 1.2224,
+      "step": 1027
+    },
+    {
+      "epoch": 0.7780510879848628,
+      "grad_norm": 2.4089951515197754,
+      "learning_rate": 1.707277969076912e-05,
+      "loss": 1.2295,
+      "step": 1028
+    },
+    {
+      "epoch": 0.7788079470198676,
+      "grad_norm": 2.244898796081543,
+      "learning_rate": 1.7067102039215928e-05,
+      "loss": 1.2656,
+      "step": 1029
+    },
+    {
+      "epoch": 0.7795648060548723,
+      "grad_norm": 2.2754669189453125,
+      "learning_rate": 1.7061419832961122e-05,
+      "loss": 1.2106,
+      "step": 1030
+    },
+    {
+      "epoch": 0.780321665089877,
+      "grad_norm": 2.3827311992645264,
+      "learning_rate": 1.7055733075667368e-05,
+      "loss": 1.1916,
+      "step": 1031
+    },
+    {
+      "epoch": 0.7810785241248818,
+      "grad_norm": 3.2731504440307617,
+      "learning_rate": 1.7050041771000258e-05,
+      "loss": 1.2265,
+      "step": 1032
+    },
+    {
+      "epoch": 0.7818353831598864,
+      "grad_norm": 2.48207950592041,
+      "learning_rate": 1.7044345922628326e-05,
+      "loss": 1.2305,
+      "step": 1033
+    },
+    {
+      "epoch": 0.7825922421948912,
+      "grad_norm": 2.3561174869537354,
+      "learning_rate": 1.703864553422302e-05,
+      "loss": 1.2191,
+      "step": 1034
+    },
+    {
+      "epoch": 0.783349101229896,
+      "grad_norm": 2.4696364402770996,
+      "learning_rate": 1.703294060945873e-05,
+      "loss": 1.2354,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7841059602649006,
+      "grad_norm": 2.214374542236328,
+      "learning_rate": 1.7027231152012765e-05,
+      "loss": 1.1459,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7848628192999054,
+      "grad_norm": 2.8190994262695312,
+      "learning_rate": 1.7021517165565352e-05,
+      "loss": 1.2289,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7856196783349101,
+      "grad_norm": 2.3381307125091553,
+      "learning_rate": 1.701579865379964e-05,
+      "loss": 1.2142,
+      "step": 1038
+    },
+    {
+      "epoch": 0.7863765373699149,
+      "grad_norm": 2.4270827770233154,
+      "learning_rate": 1.7010075620401693e-05,
+      "loss": 1.175,
+      "step": 1039
+    },
+    {
+      "epoch": 0.7871333964049196,
+      "grad_norm": 2.5499768257141113,
+      "learning_rate": 1.7004348069060487e-05,
+      "loss": 1.1907,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7878902554399243,
+      "grad_norm": 2.8665435314178467,
+      "learning_rate": 1.6998616003467923e-05,
+      "loss": 1.232,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7886471144749291,
+      "grad_norm": 2.443026065826416,
+      "learning_rate": 1.6992879427318798e-05,
+      "loss": 1.206,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7894039735099337,
+      "grad_norm": 2.408712148666382,
+      "learning_rate": 1.6987138344310822e-05,
+      "loss": 1.1984,
+      "step": 1043
+    },
+    {
+      "epoch": 0.7901608325449385,
+      "grad_norm": 2.5489931106567383,
+      "learning_rate": 1.6981392758144616e-05,
+      "loss": 1.27,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7909176915799432,
+      "grad_norm": 2.2722368240356445,
+      "learning_rate": 1.6975642672523684e-05,
+      "loss": 1.2572,
+      "step": 1045
+    },
+    {
+      "epoch": 0.791674550614948,
+      "grad_norm": 2.921919822692871,
+      "learning_rate": 1.6969888091154452e-05,
+      "loss": 1.2433,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7924314096499527,
+      "grad_norm": 2.41582989692688,
+      "learning_rate": 1.6964129017746236e-05,
+      "loss": 1.2537,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7931882686849574,
+      "grad_norm": 2.557302474975586,
+      "learning_rate": 1.695836545601125e-05,
+      "loss": 1.2248,
+      "step": 1048
+    },
+    {
+      "epoch": 0.7939451277199622,
+      "grad_norm": 2.4108498096466064,
+      "learning_rate": 1.6952597409664587e-05,
+      "loss": 1.2337,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7947019867549668,
+      "grad_norm": 2.3067305088043213,
+      "learning_rate": 1.694682488242425e-05,
+      "loss": 1.1974,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7954588457899716,
+      "grad_norm": 2.3873379230499268,
+      "learning_rate": 1.6941047878011122e-05,
+      "loss": 1.204,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7962157048249764,
+      "grad_norm": 2.3166935443878174,
+      "learning_rate": 1.6935266400148963e-05,
+      "loss": 1.1585,
+      "step": 1052
+    },
+    {
+      "epoch": 0.796972563859981,
+      "grad_norm": 2.312579870223999,
+      "learning_rate": 1.6929480452564438e-05,
+      "loss": 1.2178,
+      "step": 1053
+    },
+    {
+      "epoch": 0.7977294228949858,
+      "grad_norm": 2.3897957801818848,
+      "learning_rate": 1.6923690038987075e-05,
+      "loss": 1.1867,
+      "step": 1054
+    },
+    {
+      "epoch": 0.7984862819299905,
+      "grad_norm": 2.5109200477600098,
+      "learning_rate": 1.6917895163149282e-05,
+      "loss": 1.2219,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7992431409649953,
+      "grad_norm": 2.4277844429016113,
+      "learning_rate": 1.6912095828786353e-05,
+      "loss": 1.2224,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.285210609436035,
+      "learning_rate": 1.6906292039636452e-05,
+      "loss": 1.233,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8007568590350047,
+      "grad_norm": 2.475517511367798,
+      "learning_rate": 1.690048379944061e-05,
+      "loss": 1.2606,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8015137180700095,
+      "grad_norm": 2.2740111351013184,
+      "learning_rate": 1.6894671111942733e-05,
+      "loss": 1.1516,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8022705771050141,
+      "grad_norm": 2.777266263961792,
+      "learning_rate": 1.6888853980889583e-05,
+      "loss": 1.2257,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8030274361400189,
+      "grad_norm": 2.4774162769317627,
+      "learning_rate": 1.6883032410030796e-05,
+      "loss": 1.2161,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8037842951750237,
+      "grad_norm": 2.4283878803253174,
+      "learning_rate": 1.6877206403118875e-05,
+      "loss": 1.2258,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8045411542100284,
+      "grad_norm": 2.2770519256591797,
+      "learning_rate": 1.687137596390917e-05,
+      "loss": 1.2209,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8052980132450331,
+      "grad_norm": 2.4034667015075684,
+      "learning_rate": 1.6865541096159895e-05,
+      "loss": 1.1773,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8060548722800378,
+      "grad_norm": 2.2633402347564697,
+      "learning_rate": 1.6859701803632117e-05,
+      "loss": 1.2552,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8068117313150426,
+      "grad_norm": 2.3959109783172607,
+      "learning_rate": 1.6853858090089753e-05,
+      "loss": 1.27,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8075685903500474,
+      "grad_norm": 2.4210898876190186,
+      "learning_rate": 1.6848009959299575e-05,
+      "loss": 1.2173,
+      "step": 1067
+    },
+    {
+      "epoch": 0.808325449385052,
+      "grad_norm": 2.3308327198028564,
+      "learning_rate": 1.6842157415031194e-05,
+      "loss": 1.2738,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8090823084200568,
+      "grad_norm": 3.167160987854004,
+      "learning_rate": 1.683630046105707e-05,
+      "loss": 1.2349,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8098391674550615,
+      "grad_norm": 2.1552276611328125,
+      "learning_rate": 1.6830439101152513e-05,
+      "loss": 1.2436,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8105960264900662,
+      "grad_norm": 2.5437731742858887,
+      "learning_rate": 1.682457333909566e-05,
+      "loss": 1.2039,
+      "step": 1071
+    },
+    {
+      "epoch": 0.811352885525071,
+      "grad_norm": 2.1334102153778076,
+      "learning_rate": 1.6818703178667496e-05,
+      "loss": 1.2173,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8121097445600757,
+      "grad_norm": 2.5743660926818848,
+      "learning_rate": 1.6812828623651832e-05,
+      "loss": 1.2132,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8128666035950805,
+      "grad_norm": 2.4903461933135986,
+      "learning_rate": 1.6806949677835328e-05,
+      "loss": 1.2428,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8136234626300851,
+      "grad_norm": 2.2703421115875244,
+      "learning_rate": 1.6801066345007447e-05,
+      "loss": 1.1828,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8143803216650899,
+      "grad_norm": 2.615246295928955,
+      "learning_rate": 1.6795178628960508e-05,
+      "loss": 1.2361,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8151371807000946,
+      "grad_norm": 2.6063549518585205,
+      "learning_rate": 1.6789286533489635e-05,
+      "loss": 1.2586,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8158940397350993,
+      "grad_norm": 2.725470542907715,
+      "learning_rate": 1.6783390062392788e-05,
+      "loss": 1.2166,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8166508987701041,
+      "grad_norm": 2.576597213745117,
+      "learning_rate": 1.6777489219470743e-05,
+      "loss": 1.231,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8174077578051088,
+      "grad_norm": 2.37703537940979,
+      "learning_rate": 1.677158400852708e-05,
+      "loss": 1.2663,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8181646168401135,
+      "grad_norm": 2.3021481037139893,
+      "learning_rate": 1.6765674433368232e-05,
+      "loss": 1.2091,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8189214758751182,
+      "grad_norm": 2.4437525272369385,
+      "learning_rate": 1.67597604978034e-05,
+      "loss": 1.2123,
+      "step": 1082
+    },
+    {
+      "epoch": 0.819678334910123,
+      "grad_norm": 2.470407724380493,
+      "learning_rate": 1.6753842205644628e-05,
+      "loss": 1.1948,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8204351939451278,
+      "grad_norm": 2.5628767013549805,
+      "learning_rate": 1.6747919560706752e-05,
+      "loss": 1.2347,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8211920529801324,
+      "grad_norm": 2.5520646572113037,
+      "learning_rate": 1.6741992566807416e-05,
+      "loss": 1.2319,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8219489120151372,
+      "grad_norm": 2.4275975227355957,
+      "learning_rate": 1.673606122776708e-05,
+      "loss": 1.1666,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8227057710501419,
+      "grad_norm": 2.803802728652954,
+      "learning_rate": 1.6730125547408984e-05,
+      "loss": 1.1861,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8234626300851466,
+      "grad_norm": 3.4586920738220215,
+      "learning_rate": 1.6724185529559185e-05,
+      "loss": 1.2278,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8242194891201514,
+      "grad_norm": 2.3933305740356445,
+      "learning_rate": 1.6718241178046526e-05,
+      "loss": 1.2148,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8249763481551561,
+      "grad_norm": 2.64758038520813,
+      "learning_rate": 1.671229249670264e-05,
+      "loss": 1.2138,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8257332071901609,
+      "grad_norm": 2.389108896255493,
+      "learning_rate": 1.6706339489361962e-05,
+      "loss": 1.2295,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8264900662251655,
+      "grad_norm": 2.5130155086517334,
+      "learning_rate": 1.6700382159861705e-05,
+      "loss": 1.196,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8272469252601703,
+      "grad_norm": 2.287849187850952,
+      "learning_rate": 1.6694420512041878e-05,
+      "loss": 1.2286,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8280037842951751,
+      "grad_norm": 2.378422737121582,
+      "learning_rate": 1.6688454549745263e-05,
+      "loss": 1.2518,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8287606433301797,
+      "grad_norm": 2.3797566890716553,
+      "learning_rate": 1.6682484276817433e-05,
+      "loss": 1.2228,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8295175023651845,
+      "grad_norm": 2.276672124862671,
+      "learning_rate": 1.667650969710673e-05,
+      "loss": 1.2281,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8302743614001892,
+      "grad_norm": 2.550900459289551,
+      "learning_rate": 1.6670530814464284e-05,
+      "loss": 1.2279,
+      "step": 1097
+    },
+    {
+      "epoch": 0.831031220435194,
+      "grad_norm": 2.5035128593444824,
+      "learning_rate": 1.6664547632743987e-05,
+      "loss": 1.1628,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8317880794701987,
+      "grad_norm": 2.664567708969116,
+      "learning_rate": 1.6658560155802506e-05,
+      "loss": 1.187,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8325449385052034,
+      "grad_norm": 2.5373306274414062,
+      "learning_rate": 1.665256838749928e-05,
+      "loss": 1.2422,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8333017975402082,
+      "grad_norm": 2.7911324501037598,
+      "learning_rate": 1.664657233169651e-05,
+      "loss": 1.1982,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8340586565752128,
+      "grad_norm": 2.663367509841919,
+      "learning_rate": 1.664057199225916e-05,
+      "loss": 1.2578,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8348155156102176,
+      "grad_norm": 2.486424684524536,
+      "learning_rate": 1.663456737305496e-05,
+      "loss": 1.2106,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8355723746452224,
+      "grad_norm": 2.503634214401245,
+      "learning_rate": 1.66285584779544e-05,
+      "loss": 1.2192,
+      "step": 1104
+    },
+    {
+      "epoch": 0.836329233680227,
+      "grad_norm": 2.679033041000366,
+      "learning_rate": 1.6622545310830712e-05,
+      "loss": 1.204,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8370860927152318,
+      "grad_norm": 2.7814950942993164,
+      "learning_rate": 1.66165278755599e-05,
+      "loss": 1.2133,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8378429517502365,
+      "grad_norm": 2.5719947814941406,
+      "learning_rate": 1.6610506176020707e-05,
+      "loss": 1.2457,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8385998107852413,
+      "grad_norm": 3.662503957748413,
+      "learning_rate": 1.660448021609463e-05,
+      "loss": 1.2288,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8393566698202459,
+      "grad_norm": 2.62904691696167,
+      "learning_rate": 1.659844999966591e-05,
+      "loss": 1.2202,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8401135288552507,
+      "grad_norm": 2.6756417751312256,
+      "learning_rate": 1.659241553062154e-05,
+      "loss": 1.2438,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8408703878902555,
+      "grad_norm": 2.762983798980713,
+      "learning_rate": 1.6586376812851233e-05,
+      "loss": 1.1499,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8416272469252601,
+      "grad_norm": 2.4654974937438965,
+      "learning_rate": 1.6580333850247462e-05,
+      "loss": 1.2398,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8423841059602649,
+      "grad_norm": 2.5800747871398926,
+      "learning_rate": 1.657428664670543e-05,
+      "loss": 1.2564,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8431409649952696,
+      "grad_norm": 2.4179458618164062,
+      "learning_rate": 1.6568235206123073e-05,
+      "loss": 1.1874,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8438978240302744,
+      "grad_norm": 2.4252541065216064,
+      "learning_rate": 1.6562179532401053e-05,
+      "loss": 1.2568,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8446546830652791,
+      "grad_norm": 2.5447540283203125,
+      "learning_rate": 1.6556119629442764e-05,
+      "loss": 1.1884,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8454115421002838,
+      "grad_norm": 2.5056309700012207,
+      "learning_rate": 1.655005550115433e-05,
+      "loss": 1.2665,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8461684011352886,
+      "grad_norm": 2.7429358959198,
+      "learning_rate": 1.65439871514446e-05,
+      "loss": 1.1835,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8469252601702932,
+      "grad_norm": 2.6835551261901855,
+      "learning_rate": 1.653791458422513e-05,
+      "loss": 1.2781,
+      "step": 1119
+    },
+    {
+      "epoch": 0.847682119205298,
+      "grad_norm": 2.8627474308013916,
+      "learning_rate": 1.653183780341021e-05,
+      "loss": 1.1931,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8484389782403028,
+      "grad_norm": 2.758310556411743,
+      "learning_rate": 1.652575681291684e-05,
+      "loss": 1.1944,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8491958372753075,
+      "grad_norm": 2.761715888977051,
+      "learning_rate": 1.6519671616664734e-05,
+      "loss": 1.2457,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8499526963103122,
+      "grad_norm": 2.8214142322540283,
+      "learning_rate": 1.6513582218576315e-05,
+      "loss": 1.2203,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8507095553453169,
+      "grad_norm": 2.6501047611236572,
+      "learning_rate": 1.6507488622576712e-05,
+      "loss": 1.2591,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8514664143803217,
+      "grad_norm": 2.4939935207366943,
+      "learning_rate": 1.6501390832593777e-05,
+      "loss": 1.2205,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8522232734153264,
+      "grad_norm": 2.5232126712799072,
+      "learning_rate": 1.6495288852558036e-05,
+      "loss": 1.2055,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8529801324503311,
+      "grad_norm": 2.805695056915283,
+      "learning_rate": 1.6489182686402753e-05,
+      "loss": 1.2069,
+      "step": 1127
+    },
+    {
+      "epoch": 0.8537369914853359,
+      "grad_norm": 2.588597059249878,
+      "learning_rate": 1.6483072338063844e-05,
+      "loss": 1.1991,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8544938505203405,
+      "grad_norm": 2.632336378097534,
+      "learning_rate": 1.6476957811479966e-05,
+      "loss": 1.2556,
+      "step": 1129
+    },
+    {
+      "epoch": 0.8552507095553453,
+      "grad_norm": 2.6084611415863037,
+      "learning_rate": 1.6470839110592445e-05,
+      "loss": 1.2257,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8560075685903501,
+      "grad_norm": 2.4824182987213135,
+      "learning_rate": 1.6464716239345296e-05,
+      "loss": 1.1898,
+      "step": 1131
+    },
+    {
+      "epoch": 0.8567644276253548,
+      "grad_norm": 2.5742006301879883,
+      "learning_rate": 1.6458589201685235e-05,
+      "loss": 1.2024,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8575212866603595,
+      "grad_norm": 2.2470591068267822,
+      "learning_rate": 1.6452458001561655e-05,
+      "loss": 1.2154,
+      "step": 1133
+    },
+    {
+      "epoch": 0.8582781456953642,
+      "grad_norm": 2.5146355628967285,
+      "learning_rate": 1.6446322642926636e-05,
+      "loss": 1.2202,
+      "step": 1134
+    },
+    {
+      "epoch": 0.859035004730369,
+      "grad_norm": 2.580735683441162,
+      "learning_rate": 1.644018312973493e-05,
+      "loss": 1.1595,
+      "step": 1135
+    },
+    {
+      "epoch": 0.8597918637653738,
+      "grad_norm": 2.558544635772705,
+      "learning_rate": 1.6434039465943984e-05,
+      "loss": 1.2048,
+      "step": 1136
+    },
+    {
+      "epoch": 0.8605487228003784,
+      "grad_norm": 2.5136754512786865,
+      "learning_rate": 1.64278916555139e-05,
+      "loss": 1.2003,
+      "step": 1137
+    },
+    {
+      "epoch": 0.8613055818353832,
+      "grad_norm": 2.7524209022521973,
+      "learning_rate": 1.6421739702407468e-05,
+      "loss": 1.1862,
+      "step": 1138
+    },
+    {
+      "epoch": 0.8620624408703879,
+      "grad_norm": 2.519251585006714,
+      "learning_rate": 1.6415583610590144e-05,
+      "loss": 1.2471,
+      "step": 1139
+    },
+    {
+      "epoch": 0.8628192999053926,
+      "grad_norm": 2.7237823009490967,
+      "learning_rate": 1.6409423384030046e-05,
+      "loss": 1.1878,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8635761589403973,
+      "grad_norm": 2.595668315887451,
+      "learning_rate": 1.6403259026697967e-05,
+      "loss": 1.2164,
+      "step": 1141
+    },
+    {
+      "epoch": 0.8643330179754021,
+      "grad_norm": 2.6703858375549316,
+      "learning_rate": 1.6397090542567356e-05,
+      "loss": 1.1944,
+      "step": 1142
+    },
+    {
+      "epoch": 0.8650898770104068,
+      "grad_norm": 2.558354139328003,
+      "learning_rate": 1.639091793561432e-05,
+      "loss": 1.2423,
+      "step": 1143
+    },
+    {
+      "epoch": 0.8658467360454115,
+      "grad_norm": 2.445343494415283,
+      "learning_rate": 1.6384741209817638e-05,
+      "loss": 1.1989,
+      "step": 1144
+    },
+    {
+      "epoch": 0.8666035950804163,
+      "grad_norm": 2.3659980297088623,
+      "learning_rate": 1.6378560369158724e-05,
+      "loss": 1.1969,
+      "step": 1145
+    },
+    {
+      "epoch": 0.867360454115421,
+      "grad_norm": 2.6195647716522217,
+      "learning_rate": 1.6372375417621654e-05,
+      "loss": 1.2012,
+      "step": 1146
+    },
+    {
+      "epoch": 0.8681173131504257,
+      "grad_norm": 2.528627395629883,
+      "learning_rate": 1.6366186359193155e-05,
+      "loss": 1.2365,
+      "step": 1147
+    },
+    {
+      "epoch": 0.8688741721854305,
+      "grad_norm": 2.360337734222412,
+      "learning_rate": 1.6359993197862604e-05,
+      "loss": 1.2192,
+      "step": 1148
+    },
+    {
+      "epoch": 0.8696310312204352,
+      "grad_norm": 2.3621413707733154,
+      "learning_rate": 1.635379593762201e-05,
+      "loss": 1.2015,
+      "step": 1149
+    },
+    {
+      "epoch": 0.8703878902554399,
+      "grad_norm": 2.5273406505584717,
+      "learning_rate": 1.6347594582466038e-05,
+      "loss": 1.187,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8711447492904446,
+      "grad_norm": 2.8172874450683594,
+      "learning_rate": 1.6341389136391985e-05,
+      "loss": 1.2271,
+      "step": 1151
+    },
+    {
+      "epoch": 0.8719016083254494,
+      "grad_norm": 2.3418102264404297,
+      "learning_rate": 1.6335179603399788e-05,
+      "loss": 1.2358,
+      "step": 1152
+    },
+    {
+      "epoch": 0.8726584673604542,
+      "grad_norm": 2.415493965148926,
+      "learning_rate": 1.632896598749202e-05,
+      "loss": 1.2717,
+      "step": 1153
+    },
+    {
+      "epoch": 0.8734153263954588,
+      "grad_norm": 2.3262200355529785,
+      "learning_rate": 1.6322748292673875e-05,
+      "loss": 1.2198,
+      "step": 1154
+    },
+    {
+      "epoch": 0.8741721854304636,
+      "grad_norm": 2.8730580806732178,
+      "learning_rate": 1.6316526522953195e-05,
+      "loss": 1.183,
+      "step": 1155
+    },
+    {
+      "epoch": 0.8749290444654683,
+      "grad_norm": 2.432713508605957,
+      "learning_rate": 1.631030068234043e-05,
+      "loss": 1.2093,
+      "step": 1156
+    },
+    {
+      "epoch": 0.875685903500473,
+      "grad_norm": 2.6007068157196045,
+      "learning_rate": 1.630407077484866e-05,
+      "loss": 1.2506,
+      "step": 1157
+    },
+    {
+      "epoch": 0.8764427625354778,
+      "grad_norm": 2.785717487335205,
+      "learning_rate": 1.6297836804493598e-05,
+      "loss": 1.2073,
+      "step": 1158
+    },
+    {
+      "epoch": 0.8771996215704825,
+      "grad_norm": 2.498161792755127,
+      "learning_rate": 1.629159877529356e-05,
+      "loss": 1.2297,
+      "step": 1159
+    },
+    {
+      "epoch": 0.8779564806054873,
+      "grad_norm": 2.6516387462615967,
+      "learning_rate": 1.628535669126948e-05,
+      "loss": 1.2242,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8787133396404919,
+      "grad_norm": 2.164231300354004,
+      "learning_rate": 1.627911055644492e-05,
+      "loss": 1.242,
+      "step": 1161
+    },
+    {
+      "epoch": 0.8794701986754967,
+      "grad_norm": 2.1503818035125732,
+      "learning_rate": 1.6272860374846037e-05,
+      "loss": 1.2187,
+      "step": 1162
+    },
+    {
+      "epoch": 0.8802270577105015,
+      "grad_norm": 1.9819633960723877,
+      "learning_rate": 1.6266606150501608e-05,
+      "loss": 1.2044,
+      "step": 1163
+    },
+    {
+      "epoch": 0.8809839167455061,
+      "grad_norm": 2.251472234725952,
+      "learning_rate": 1.6260347887443e-05,
+      "loss": 1.2262,
+      "step": 1164
+    },
+    {
+      "epoch": 0.8817407757805109,
+      "grad_norm": 2.2765519618988037,
+      "learning_rate": 1.625408558970421e-05,
+      "loss": 1.185,
+      "step": 1165
+    },
+    {
+      "epoch": 0.8824976348155156,
+      "grad_norm": 2.4048166275024414,
+      "learning_rate": 1.6247819261321803e-05,
+      "loss": 1.1973,
+      "step": 1166
+    },
+    {
+      "epoch": 0.8832544938505204,
+      "grad_norm": 2.234778881072998,
+      "learning_rate": 1.624154890633497e-05,
+      "loss": 1.1795,
+      "step": 1167
+    },
+    {
+      "epoch": 0.8840113528855251,
+      "grad_norm": 2.2173893451690674,
+      "learning_rate": 1.623527452878548e-05,
+      "loss": 1.1897,
+      "step": 1168
+    },
+    {
+      "epoch": 0.8847682119205298,
+      "grad_norm": 2.3917415142059326,
+      "learning_rate": 1.6228996132717702e-05,
+      "loss": 1.2425,
+      "step": 1169
+    },
+    {
+      "epoch": 0.8855250709555346,
+      "grad_norm": 2.254404306411743,
+      "learning_rate": 1.62227137221786e-05,
+      "loss": 1.2495,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8862819299905392,
+      "grad_norm": 2.1990530490875244,
+      "learning_rate": 1.6216427301217713e-05,
+      "loss": 1.1757,
+      "step": 1171
+    },
+    {
+      "epoch": 0.887038789025544,
+      "grad_norm": 2.3781630992889404,
+      "learning_rate": 1.6210136873887176e-05,
+      "loss": 1.2387,
+      "step": 1172
+    },
+    {
+      "epoch": 0.8877956480605487,
+      "grad_norm": 2.1774098873138428,
+      "learning_rate": 1.6203842444241703e-05,
+      "loss": 1.1937,
+      "step": 1173
+    },
+    {
+      "epoch": 0.8885525070955534,
+      "grad_norm": 2.3653695583343506,
+      "learning_rate": 1.619754401633858e-05,
+      "loss": 1.2115,
+      "step": 1174
+    },
+    {
+      "epoch": 0.8893093661305582,
+      "grad_norm": 2.3032443523406982,
+      "learning_rate": 1.619124159423769e-05,
+      "loss": 1.1802,
+      "step": 1175
+    },
+    {
+      "epoch": 0.8900662251655629,
+      "grad_norm": 2.2687666416168213,
+      "learning_rate": 1.618493518200147e-05,
+      "loss": 1.1868,
+      "step": 1176
+    },
+    {
+      "epoch": 0.8908230842005677,
+      "grad_norm": 2.3060355186462402,
+      "learning_rate": 1.6178624783694937e-05,
+      "loss": 1.1933,
+      "step": 1177
+    },
+    {
+      "epoch": 0.8915799432355723,
+      "grad_norm": 3.4064903259277344,
+      "learning_rate": 1.6172310403385677e-05,
+      "loss": 1.234,
+      "step": 1178
+    },
+    {
+      "epoch": 0.8923368022705771,
+      "grad_norm": 2.568434715270996,
+      "learning_rate": 1.616599204514385e-05,
+      "loss": 1.2115,
+      "step": 1179
+    },
+    {
+      "epoch": 0.8930936613055819,
+      "grad_norm": 2.2627182006835938,
+      "learning_rate": 1.6159669713042166e-05,
+      "loss": 1.2229,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8938505203405865,
+      "grad_norm": 2.2551677227020264,
+      "learning_rate": 1.615334341115591e-05,
+      "loss": 1.2481,
+      "step": 1181
+    },
+    {
+      "epoch": 0.8946073793755913,
+      "grad_norm": 2.4630017280578613,
+      "learning_rate": 1.6147013143562915e-05,
+      "loss": 1.1769,
+      "step": 1182
+    },
+    {
+      "epoch": 0.895364238410596,
+      "grad_norm": 2.2972466945648193,
+      "learning_rate": 1.6140678914343575e-05,
+      "loss": 1.2028,
+      "step": 1183
+    },
+    {
+      "epoch": 0.8961210974456008,
+      "grad_norm": 2.343468189239502,
+      "learning_rate": 1.6134340727580843e-05,
+      "loss": 1.2356,
+      "step": 1184
+    },
+    {
+      "epoch": 0.8968779564806055,
+      "grad_norm": 2.2180895805358887,
+      "learning_rate": 1.6127998587360208e-05,
+      "loss": 1.2082,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8976348155156102,
+      "grad_norm": 2.121718406677246,
+      "learning_rate": 1.6121652497769727e-05,
+      "loss": 1.2052,
+      "step": 1186
+    },
+    {
+      "epoch": 0.898391674550615,
+      "grad_norm": 2.2796201705932617,
+      "learning_rate": 1.6115302462899982e-05,
+      "loss": 1.2056,
+      "step": 1187
+    },
+    {
+      "epoch": 0.8991485335856196,
+      "grad_norm": 2.1909053325653076,
+      "learning_rate": 1.6108948486844118e-05,
+      "loss": 1.1556,
+      "step": 1188
+    },
+    {
+      "epoch": 0.8999053926206244,
+      "grad_norm": 2.4068331718444824,
+      "learning_rate": 1.610259057369781e-05,
+      "loss": 1.2258,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9006622516556292,
+      "grad_norm": 2.168159246444702,
+      "learning_rate": 1.6096228727559265e-05,
+      "loss": 1.1805,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9014191106906339,
+      "grad_norm": 2.129645586013794,
+      "learning_rate": 1.608986295252924e-05,
+      "loss": 1.1725,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9021759697256386,
+      "grad_norm": 2.3025479316711426,
+      "learning_rate": 1.608349325271101e-05,
+      "loss": 1.1873,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9029328287606433,
+      "grad_norm": 2.3402655124664307,
+      "learning_rate": 1.607711963221039e-05,
+      "loss": 1.2049,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9036896877956481,
+      "grad_norm": 2.844715118408203,
+      "learning_rate": 1.6070742095135722e-05,
+      "loss": 1.1654,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9044465468306528,
+      "grad_norm": 2.337291717529297,
+      "learning_rate": 1.6064360645597862e-05,
+      "loss": 1.2119,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9052034058656575,
+      "grad_norm": 2.2666563987731934,
+      "learning_rate": 1.60579752877102e-05,
+      "loss": 1.1662,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9059602649006623,
+      "grad_norm": 2.2546000480651855,
+      "learning_rate": 1.6051586025588634e-05,
+      "loss": 1.1612,
+      "step": 1197
+    },
+    {
+      "epoch": 0.906717123935667,
+      "grad_norm": 2.4789252281188965,
+      "learning_rate": 1.6045192863351594e-05,
+      "loss": 1.159,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9074739829706717,
+      "grad_norm": 2.5757458209991455,
+      "learning_rate": 1.6038795805120005e-05,
+      "loss": 1.1359,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9082308420056765,
+      "grad_norm": 2.5025620460510254,
+      "learning_rate": 1.603239485501732e-05,
+      "loss": 1.2513,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9089877010406812,
+      "grad_norm": 2.6904783248901367,
+      "learning_rate": 1.6025990017169495e-05,
+      "loss": 1.2275,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9097445600756859,
+      "grad_norm": 2.121021270751953,
+      "learning_rate": 1.6019581295704985e-05,
+      "loss": 1.2138,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9105014191106906,
+      "grad_norm": 2.2942512035369873,
+      "learning_rate": 1.601316869475476e-05,
+      "loss": 1.207,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9112582781456954,
+      "grad_norm": 2.4668707847595215,
+      "learning_rate": 1.6006752218452283e-05,
+      "loss": 1.2422,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9120151371807,
+      "grad_norm": 2.536863088607788,
+      "learning_rate": 1.600033187093351e-05,
+      "loss": 1.2256,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9127719962157048,
+      "grad_norm": 3.008856773376465,
+      "learning_rate": 1.599390765633691e-05,
+      "loss": 1.2129,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9135288552507096,
+      "grad_norm": 2.292177438735962,
+      "learning_rate": 1.5987479578803425e-05,
+      "loss": 1.2237,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 2.4802656173706055,
+      "learning_rate": 1.59810476424765e-05,
+      "loss": 1.2307,
+      "step": 1208
+    },
+    {
+      "epoch": 0.915042573320719,
+      "grad_norm": 2.225219964981079,
+      "learning_rate": 1.5974611851502064e-05,
+      "loss": 1.1845,
+      "step": 1209
+    },
+    {
+      "epoch": 0.9157994323557237,
+      "grad_norm": 2.584470272064209,
+      "learning_rate": 1.5968172210028525e-05,
+      "loss": 1.1756,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9165562913907285,
+      "grad_norm": 2.3518307209014893,
+      "learning_rate": 1.596172872220679e-05,
+      "loss": 1.1788,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9173131504257332,
+      "grad_norm": 2.3497278690338135,
+      "learning_rate": 1.595528139219021e-05,
+      "loss": 1.2084,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9180700094607379,
+      "grad_norm": 2.117664337158203,
+      "learning_rate": 1.594883022413466e-05,
+      "loss": 1.1765,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9188268684957427,
+      "grad_norm": 2.1322619915008545,
+      "learning_rate": 1.594237522219845e-05,
+      "loss": 1.1835,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9195837275307474,
+      "grad_norm": 2.0177836418151855,
+      "learning_rate": 1.5935916390542377e-05,
+      "loss": 1.178,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9203405865657521,
+      "grad_norm": 2.2327425479888916,
+      "learning_rate": 1.5929453733329713e-05,
+      "loss": 1.1916,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9210974456007569,
+      "grad_norm": 2.168905258178711,
+      "learning_rate": 1.592298725472618e-05,
+      "loss": 1.2139,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9218543046357616,
+      "grad_norm": 2.275158166885376,
+      "learning_rate": 1.591651695889998e-05,
+      "loss": 1.2014,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9226111636707663,
+      "grad_norm": 2.153704881668091,
+      "learning_rate": 1.5910042850021754e-05,
+      "loss": 1.2219,
+      "step": 1219
+    },
+    {
+      "epoch": 0.923368022705771,
+      "grad_norm": 2.161616802215576,
+      "learning_rate": 1.5903564932264624e-05,
+      "loss": 1.2452,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9241248817407758,
+      "grad_norm": 2.1606664657592773,
+      "learning_rate": 1.589708320980416e-05,
+      "loss": 1.1448,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9248817407757806,
+      "grad_norm": 2.040039300918579,
+      "learning_rate": 1.589059768681837e-05,
+      "loss": 1.235,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9256385998107852,
+      "grad_norm": 2.2927193641662598,
+      "learning_rate": 1.5884108367487732e-05,
+      "loss": 1.19,
+      "step": 1223
+    },
+    {
+      "epoch": 0.92639545884579,
+      "grad_norm": 2.2096221446990967,
+      "learning_rate": 1.587761525599516e-05,
+      "loss": 1.2349,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9271523178807947,
+      "grad_norm": 2.1982614994049072,
+      "learning_rate": 1.5871118356526017e-05,
+      "loss": 1.206,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9279091769157994,
+      "grad_norm": 2.2477710247039795,
+      "learning_rate": 1.5864617673268096e-05,
+      "loss": 1.2044,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9286660359508042,
+      "grad_norm": 2.126891851425171,
+      "learning_rate": 1.5858113210411646e-05,
+      "loss": 1.1685,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9294228949858089,
+      "grad_norm": 2.6382102966308594,
+      "learning_rate": 1.585160497214935e-05,
+      "loss": 1.2247,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9301797540208137,
+      "grad_norm": 2.1951191425323486,
+      "learning_rate": 1.5845092962676306e-05,
+      "loss": 1.1517,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9309366130558183,
+      "grad_norm": 2.299997091293335,
+      "learning_rate": 1.5838577186190064e-05,
+      "loss": 1.2327,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9316934720908231,
+      "grad_norm": 2.69441556930542,
+      "learning_rate": 1.5832057646890594e-05,
+      "loss": 1.1622,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9324503311258279,
+      "grad_norm": 2.3268439769744873,
+      "learning_rate": 1.582553434898029e-05,
+      "loss": 1.2181,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9332071901608325,
+      "grad_norm": 2.1350252628326416,
+      "learning_rate": 1.5819007296663974e-05,
+      "loss": 1.1818,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9339640491958373,
+      "grad_norm": 2.0754928588867188,
+      "learning_rate": 1.5812476494148876e-05,
+      "loss": 1.1847,
+      "step": 1234
+    },
+    {
+      "epoch": 0.934720908230842,
+      "grad_norm": 2.48238205909729,
+      "learning_rate": 1.5805941945644658e-05,
+      "loss": 1.222,
+      "step": 1235
+    },
+    {
+      "epoch": 0.9354777672658467,
+      "grad_norm": 2.202993154525757,
+      "learning_rate": 1.579940365536339e-05,
+      "loss": 1.2381,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9362346263008514,
+      "grad_norm": 2.424055337905884,
+      "learning_rate": 1.5792861627519554e-05,
+      "loss": 1.2035,
+      "step": 1237
+    },
+    {
+      "epoch": 0.9369914853358562,
+      "grad_norm": 2.270042896270752,
+      "learning_rate": 1.578631586633004e-05,
+      "loss": 1.2268,
+      "step": 1238
+    },
+    {
+      "epoch": 0.937748344370861,
+      "grad_norm": 2.38864803314209,
+      "learning_rate": 1.5779766376014146e-05,
+      "loss": 1.2202,
+      "step": 1239
+    },
+    {
+      "epoch": 0.9385052034058656,
+      "grad_norm": 2.137854814529419,
+      "learning_rate": 1.5773213160793574e-05,
+      "loss": 1.2246,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9392620624408704,
+      "grad_norm": 2.5035834312438965,
+      "learning_rate": 1.5766656224892424e-05,
+      "loss": 1.2246,
+      "step": 1241
+    },
+    {
+      "epoch": 0.9400189214758751,
+      "grad_norm": 2.498552083969116,
+      "learning_rate": 1.5760095572537207e-05,
+      "loss": 1.2037,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9407757805108798,
+      "grad_norm": 2.0278542041778564,
+      "learning_rate": 1.5753531207956806e-05,
+      "loss": 1.2197,
+      "step": 1243
+    },
+    {
+      "epoch": 0.9415326395458846,
+      "grad_norm": 2.1201868057250977,
+      "learning_rate": 1.5746963135382522e-05,
+      "loss": 1.1557,
+      "step": 1244
+    },
+    {
+      "epoch": 0.9422894985808893,
+      "grad_norm": 2.480867385864258,
+      "learning_rate": 1.574039135904802e-05,
+      "loss": 1.2006,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9430463576158941,
+      "grad_norm": 2.257807970046997,
+      "learning_rate": 1.573381588318938e-05,
+      "loss": 1.2235,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9438032166508987,
+      "grad_norm": 2.2047722339630127,
+      "learning_rate": 1.5727236712045053e-05,
+      "loss": 1.1904,
+      "step": 1247
+    },
+    {
+      "epoch": 0.9445600756859035,
+      "grad_norm": 2.2862167358398438,
+      "learning_rate": 1.5720653849855862e-05,
+      "loss": 1.2388,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9453169347209083,
+      "grad_norm": 2.2071452140808105,
+      "learning_rate": 1.571406730086503e-05,
+      "loss": 1.1967,
+      "step": 1249
+    },
+    {
+      "epoch": 0.9460737937559129,
+      "grad_norm": 2.1805355548858643,
+      "learning_rate": 1.5707477069318143e-05,
+      "loss": 1.1927,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9468306527909177,
+      "grad_norm": 2.156611204147339,
+      "learning_rate": 1.5700883159463162e-05,
+      "loss": 1.2216,
+      "step": 1251
+    },
+    {
+      "epoch": 0.9475875118259224,
+      "grad_norm": 2.2290961742401123,
+      "learning_rate": 1.5694285575550416e-05,
+      "loss": 1.2116,
+      "step": 1252
+    },
+    {
+      "epoch": 0.9483443708609272,
+      "grad_norm": 2.0691416263580322,
+      "learning_rate": 1.568768432183262e-05,
+      "loss": 1.2077,
+      "step": 1253
+    },
+    {
+      "epoch": 0.9491012298959319,
+      "grad_norm": 2.2860946655273438,
+      "learning_rate": 1.568107940256483e-05,
+      "loss": 1.2392,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9498580889309366,
+      "grad_norm": 2.3357367515563965,
+      "learning_rate": 1.567447082200448e-05,
+      "loss": 1.2469,
+      "step": 1255
+    },
+    {
+      "epoch": 0.9506149479659414,
+      "grad_norm": 2.224269151687622,
+      "learning_rate": 1.566785858441136e-05,
+      "loss": 1.2553,
+      "step": 1256
+    },
+    {
+      "epoch": 0.951371807000946,
+      "grad_norm": 2.274747133255005,
+      "learning_rate": 1.566124269404762e-05,
+      "loss": 1.1486,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9521286660359508,
+      "grad_norm": 2.205291271209717,
+      "learning_rate": 1.5654623155177758e-05,
+      "loss": 1.228,
+      "step": 1258
+    },
+    {
+      "epoch": 0.9528855250709556,
+      "grad_norm": 2.059138774871826,
+      "learning_rate": 1.564799997206863e-05,
+      "loss": 1.152,
+      "step": 1259
+    },
+    {
+      "epoch": 0.9536423841059603,
+      "grad_norm": 2.382854700088501,
+      "learning_rate": 1.564137314898944e-05,
+      "loss": 1.2262,
+      "step": 1260
+    },
+    {
+      "epoch": 0.954399243140965,
+      "grad_norm": 2.1926519870758057,
+      "learning_rate": 1.563474269021174e-05,
+      "loss": 1.1439,
+      "step": 1261
+    },
+    {
+      "epoch": 0.9551561021759697,
+      "grad_norm": 2.247835159301758,
+      "learning_rate": 1.5628108600009414e-05,
+      "loss": 1.2191,
+      "step": 1262
+    },
+    {
+      "epoch": 0.9559129612109745,
+      "grad_norm": 2.6202445030212402,
+      "learning_rate": 1.5621470882658696e-05,
+      "loss": 1.1955,
+      "step": 1263
+    },
+    {
+      "epoch": 0.9566698202459792,
+      "grad_norm": 2.1109254360198975,
+      "learning_rate": 1.5614829542438162e-05,
+      "loss": 1.2208,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9574266792809839,
+      "grad_norm": 2.0421035289764404,
+      "learning_rate": 1.5608184583628723e-05,
+      "loss": 1.2597,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9581835383159887,
+      "grad_norm": 2.3527796268463135,
+      "learning_rate": 1.5601536010513608e-05,
+      "loss": 1.2437,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9589403973509933,
+      "grad_norm": 2.39426851272583,
+      "learning_rate": 1.559488382737839e-05,
+      "loss": 1.2418,
+      "step": 1267
+    },
+    {
+      "epoch": 0.9596972563859981,
+      "grad_norm": 2.546283483505249,
+      "learning_rate": 1.558822803851097e-05,
+      "loss": 1.2295,
+      "step": 1268
+    },
+    {
+      "epoch": 0.9604541154210028,
+      "grad_norm": 2.275153160095215,
+      "learning_rate": 1.558156864820156e-05,
+      "loss": 1.1896,
+      "step": 1269
+    },
+    {
+      "epoch": 0.9612109744560076,
+      "grad_norm": 2.1879630088806152,
+      "learning_rate": 1.5574905660742707e-05,
+      "loss": 1.1766,
+      "step": 1270
+    },
+    {
+      "epoch": 0.9619678334910123,
+      "grad_norm": 2.3438518047332764,
+      "learning_rate": 1.556823908042927e-05,
+      "loss": 1.1828,
+      "step": 1271
+    },
+    {
+      "epoch": 0.962724692526017,
+      "grad_norm": 2.673069477081299,
+      "learning_rate": 1.5561568911558422e-05,
+      "loss": 1.1909,
+      "step": 1272
+    },
+    {
+      "epoch": 0.9634815515610218,
+      "grad_norm": 2.3552541732788086,
+      "learning_rate": 1.5554895158429654e-05,
+      "loss": 1.2246,
+      "step": 1273
+    },
+    {
+      "epoch": 0.9642384105960264,
+      "grad_norm": 2.1169943809509277,
+      "learning_rate": 1.5548217825344765e-05,
+      "loss": 1.2018,
+      "step": 1274
+    },
+    {
+      "epoch": 0.9649952696310312,
+      "grad_norm": 2.1841084957122803,
+      "learning_rate": 1.5541536916607863e-05,
+      "loss": 1.1496,
+      "step": 1275
+    },
+    {
+      "epoch": 0.965752128666036,
+      "grad_norm": 2.1429550647735596,
+      "learning_rate": 1.553485243652536e-05,
+      "loss": 1.1537,
+      "step": 1276
+    },
+    {
+      "epoch": 0.9665089877010407,
+      "grad_norm": 2.6029670238494873,
+      "learning_rate": 1.5528164389405972e-05,
+      "loss": 1.2348,
+      "step": 1277
+    },
+    {
+      "epoch": 0.9672658467360454,
+      "grad_norm": 2.105222463607788,
+      "learning_rate": 1.5521472779560705e-05,
+      "loss": 1.217,
+      "step": 1278
+    },
+    {
+      "epoch": 0.9680227057710501,
+      "grad_norm": 2.1541764736175537,
+      "learning_rate": 1.5514777611302875e-05,
+      "loss": 1.1844,
+      "step": 1279
+    },
+    {
+      "epoch": 0.9687795648060549,
+      "grad_norm": 2.0249156951904297,
+      "learning_rate": 1.5508078888948086e-05,
+      "loss": 1.2191,
+      "step": 1280
+    },
+    {
+      "epoch": 0.9695364238410596,
+      "grad_norm": 2.187110185623169,
+      "learning_rate": 1.550137661681423e-05,
+      "loss": 1.1811,
+      "step": 1281
+    },
+    {
+      "epoch": 0.9702932828760643,
+      "grad_norm": 2.27626371383667,
+      "learning_rate": 1.5494670799221485e-05,
+      "loss": 1.186,
+      "step": 1282
+    },
+    {
+      "epoch": 0.9710501419110691,
+      "grad_norm": 2.0355005264282227,
+      "learning_rate": 1.5487961440492327e-05,
+      "loss": 1.2338,
+      "step": 1283
+    },
+    {
+      "epoch": 0.9718070009460738,
+      "grad_norm": 2.126351833343506,
+      "learning_rate": 1.54812485449515e-05,
+      "loss": 1.2129,
+      "step": 1284
+    },
+    {
+      "epoch": 0.9725638599810785,
+      "grad_norm": 2.150451421737671,
+      "learning_rate": 1.5474532116926037e-05,
+      "loss": 1.1812,
+      "step": 1285
+    },
+    {
+      "epoch": 0.9733207190160833,
+      "grad_norm": 2.0796091556549072,
+      "learning_rate": 1.5467812160745245e-05,
+      "loss": 1.2273,
+      "step": 1286
+    },
+    {
+      "epoch": 0.974077578051088,
+      "grad_norm": 2.349214792251587,
+      "learning_rate": 1.5461088680740702e-05,
+      "loss": 1.2286,
+      "step": 1287
+    },
+    {
+      "epoch": 0.9748344370860927,
+      "grad_norm": 2.1848902702331543,
+      "learning_rate": 1.545436168124627e-05,
+      "loss": 1.2239,
+      "step": 1288
+    },
+    {
+      "epoch": 0.9755912961210974,
+      "grad_norm": 2.261702299118042,
+      "learning_rate": 1.544763116659806e-05,
+      "loss": 1.202,
+      "step": 1289
+    },
+    {
+      "epoch": 0.9763481551561022,
+      "grad_norm": 2.2427971363067627,
+      "learning_rate": 1.5440897141134464e-05,
+      "loss": 1.2133,
+      "step": 1290
+    },
+    {
+      "epoch": 0.977105014191107,
+      "grad_norm": 2.076875686645508,
+      "learning_rate": 1.5434159609196128e-05,
+      "loss": 1.2056,
+      "step": 1291
+    },
+    {
+      "epoch": 0.9778618732261116,
+      "grad_norm": 2.26599383354187,
+      "learning_rate": 1.542741857512597e-05,
+      "loss": 1.195,
+      "step": 1292
+    },
+    {
+      "epoch": 0.9786187322611164,
+      "grad_norm": 2.262747049331665,
+      "learning_rate": 1.5420674043269152e-05,
+      "loss": 1.2286,
+      "step": 1293
+    },
+    {
+      "epoch": 0.9793755912961211,
+      "grad_norm": 2.1384646892547607,
+      "learning_rate": 1.5413926017973097e-05,
+      "loss": 1.1843,
+      "step": 1294
+    },
+    {
+      "epoch": 0.9801324503311258,
+      "grad_norm": 2.3019633293151855,
+      "learning_rate": 1.540717450358748e-05,
+      "loss": 1.2474,
+      "step": 1295
+    },
+    {
+      "epoch": 0.9808893093661306,
+      "grad_norm": 2.6212801933288574,
+      "learning_rate": 1.5400419504464222e-05,
+      "loss": 1.2305,
+      "step": 1296
+    },
+    {
+      "epoch": 0.9816461684011353,
+      "grad_norm": 2.415092706680298,
+      "learning_rate": 1.5393661024957495e-05,
+      "loss": 1.2394,
+      "step": 1297
+    },
+    {
+      "epoch": 0.98240302743614,
+      "grad_norm": 2.392845392227173,
+      "learning_rate": 1.5386899069423712e-05,
+      "loss": 1.1971,
+      "step": 1298
+    },
+    {
+      "epoch": 0.9831598864711447,
+      "grad_norm": 2.291163206100464,
+      "learning_rate": 1.5380133642221525e-05,
+      "loss": 1.2004,
+      "step": 1299
+    },
+    {
+      "epoch": 0.9839167455061495,
+      "grad_norm": 2.191312313079834,
+      "learning_rate": 1.5373364747711825e-05,
+      "loss": 1.1677,
+      "step": 1300
+    },
+    {
+      "epoch": 0.9846736045411542,
+      "grad_norm": 2.2666783332824707,
+      "learning_rate": 1.536659239025774e-05,
+      "loss": 1.1656,
+      "step": 1301
+    },
+    {
+      "epoch": 0.9854304635761589,
+      "grad_norm": 2.578672409057617,
+      "learning_rate": 1.5359816574224626e-05,
+      "loss": 1.2021,
+      "step": 1302
+    },
+    {
+      "epoch": 0.9861873226111637,
+      "grad_norm": 2.1345741748809814,
+      "learning_rate": 1.5353037303980075e-05,
+      "loss": 1.2277,
+      "step": 1303
+    },
+    {
+      "epoch": 0.9869441816461684,
+      "grad_norm": 2.1685898303985596,
+      "learning_rate": 1.5346254583893895e-05,
+      "loss": 1.206,
+      "step": 1304
+    },
+    {
+      "epoch": 0.9877010406811731,
+      "grad_norm": 2.3150031566619873,
+      "learning_rate": 1.533946841833813e-05,
+      "loss": 1.1747,
+      "step": 1305
+    },
+    {
+      "epoch": 0.9884578997161778,
+      "grad_norm": 2.3677496910095215,
+      "learning_rate": 1.5332678811687034e-05,
+      "loss": 1.2502,
+      "step": 1306
+    },
+    {
+      "epoch": 0.9892147587511826,
+      "grad_norm": 2.0479371547698975,
+      "learning_rate": 1.5325885768317085e-05,
+      "loss": 1.129,
+      "step": 1307
+    },
+    {
+      "epoch": 0.9899716177861874,
+      "grad_norm": 2.272096633911133,
+      "learning_rate": 1.531908929260698e-05,
+      "loss": 1.1515,
+      "step": 1308
+    },
+    {
+      "epoch": 0.990728476821192,
+      "grad_norm": 2.233167886734009,
+      "learning_rate": 1.5312289388937613e-05,
+      "loss": 1.1576,
+      "step": 1309
+    },
+    {
+      "epoch": 0.9914853358561968,
+      "grad_norm": 2.329028606414795,
+      "learning_rate": 1.530548606169211e-05,
+      "loss": 1.2331,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9922421948912015,
+      "grad_norm": 2.3802735805511475,
+      "learning_rate": 1.5298679315255786e-05,
+      "loss": 1.1576,
+      "step": 1311
+    },
+    {
+      "epoch": 0.9929990539262062,
+      "grad_norm": 2.45041561126709,
+      "learning_rate": 1.5291869154016167e-05,
+      "loss": 1.2241,
+      "step": 1312
+    },
+    {
+      "epoch": 0.993755912961211,
+      "grad_norm": 2.528601884841919,
+      "learning_rate": 1.5285055582362975e-05,
+      "loss": 1.2257,
+      "step": 1313
+    },
+    {
+      "epoch": 0.9945127719962157,
+      "grad_norm": 2.2096829414367676,
+      "learning_rate": 1.5278238604688143e-05,
+      "loss": 1.1959,
+      "step": 1314
+    },
+    {
+      "epoch": 0.9952696310312205,
+      "grad_norm": 2.218921184539795,
+      "learning_rate": 1.5271418225385784e-05,
+      "loss": 1.1444,
+      "step": 1315
+    },
+    {
+      "epoch": 0.9960264900662251,
+      "grad_norm": 2.563999891281128,
+      "learning_rate": 1.526459444885221e-05,
+      "loss": 1.192,
+      "step": 1316
+    },
+    {
+      "epoch": 0.9967833491012299,
+      "grad_norm": 2.6427245140075684,
+      "learning_rate": 1.5257767279485934e-05,
+      "loss": 1.1575,
+      "step": 1317
+    },
+    {
+      "epoch": 0.9975402081362347,
+      "grad_norm": 2.0441269874572754,
+      "learning_rate": 1.5250936721687628e-05,
+      "loss": 1.1636,
+      "step": 1318
+    },
+    {
+      "epoch": 0.9982970671712393,
+      "grad_norm": 2.560488700866699,
+      "learning_rate": 1.5244102779860178e-05,
+      "loss": 1.2198,
+      "step": 1319
+    },
+    {
+      "epoch": 0.9990539262062441,
+      "grad_norm": 2.5017917156219482,
+      "learning_rate": 1.5237265458408637e-05,
+      "loss": 1.2215,
+      "step": 1320
+    },
+    {
+      "epoch": 0.9998107852412488,
+      "grad_norm": 2.3458917140960693,
+      "learning_rate": 1.5230424761740234e-05,
+      "loss": 1.1645,
+      "step": 1321
+    },
+    {
+      "epoch": 1.0005676442762534,
+      "grad_norm": 2.2212741374969482,
+      "learning_rate": 1.5223580694264382e-05,
+      "loss": 1.2054,
+      "step": 1322
+    },
+    {
+      "epoch": 1.0013245033112583,
+      "grad_norm": 2.394789695739746,
+      "learning_rate": 1.5216733260392658e-05,
+      "loss": 1.1085,
+      "step": 1323
+    },
+    {
+      "epoch": 1.002081362346263,
+      "grad_norm": 2.135875701904297,
+      "learning_rate": 1.5209882464538817e-05,
+      "loss": 1.1754,
+      "step": 1324
+    },
+    {
+      "epoch": 1.0028382213812677,
+      "grad_norm": 2.4602649211883545,
+      "learning_rate": 1.5203028311118778e-05,
+      "loss": 1.1202,
+      "step": 1325
+    },
+    {
+      "epoch": 1.0035950804162725,
+      "grad_norm": 2.5135326385498047,
+      "learning_rate": 1.5196170804550618e-05,
+      "loss": 1.147,
+      "step": 1326
+    },
+    {
+      "epoch": 1.0043519394512772,
+      "grad_norm": 2.3376166820526123,
+      "learning_rate": 1.518930994925459e-05,
+      "loss": 1.1715,
+      "step": 1327
+    },
+    {
+      "epoch": 1.0051087984862819,
+      "grad_norm": 2.2913684844970703,
+      "learning_rate": 1.518244574965309e-05,
+      "loss": 1.2037,
+      "step": 1328
+    },
+    {
+      "epoch": 1.0058656575212868,
+      "grad_norm": 2.34970760345459,
+      "learning_rate": 1.5175578210170678e-05,
+      "loss": 1.155,
+      "step": 1329
+    },
+    {
+      "epoch": 1.0066225165562914,
+      "grad_norm": 2.345945119857788,
+      "learning_rate": 1.5168707335234067e-05,
+      "loss": 1.1392,
+      "step": 1330
+    },
+    {
+      "epoch": 1.007379375591296,
+      "grad_norm": 2.2599105834960938,
+      "learning_rate": 1.5161833129272117e-05,
+      "loss": 1.1402,
+      "step": 1331
+    },
+    {
+      "epoch": 1.0081362346263008,
+      "grad_norm": 2.5355637073516846,
+      "learning_rate": 1.5154955596715836e-05,
+      "loss": 1.2212,
+      "step": 1332
+    },
+    {
+      "epoch": 1.0088930936613056,
+      "grad_norm": 2.4323315620422363,
+      "learning_rate": 1.5148074741998377e-05,
+      "loss": 1.1289,
+      "step": 1333
+    },
+    {
+      "epoch": 1.0096499526963103,
+      "grad_norm": 2.232952833175659,
+      "learning_rate": 1.5141190569555033e-05,
+      "loss": 1.1044,
+      "step": 1334
+    },
+    {
+      "epoch": 1.010406811731315,
+      "grad_norm": 2.1613996028900146,
+      "learning_rate": 1.513430308382324e-05,
+      "loss": 1.1678,
+      "step": 1335
+    },
+    {
+      "epoch": 1.0111636707663199,
+      "grad_norm": 2.369002342224121,
+      "learning_rate": 1.5127412289242562e-05,
+      "loss": 1.2219,
+      "step": 1336
+    },
+    {
+      "epoch": 1.0119205298013245,
+      "grad_norm": 2.2522876262664795,
+      "learning_rate": 1.51205181902547e-05,
+      "loss": 1.1728,
+      "step": 1337
+    },
+    {
+      "epoch": 1.0126773888363292,
+      "grad_norm": 2.172529935836792,
+      "learning_rate": 1.5113620791303489e-05,
+      "loss": 1.1599,
+      "step": 1338
+    },
+    {
+      "epoch": 1.013434247871334,
+      "grad_norm": 2.265456199645996,
+      "learning_rate": 1.5106720096834885e-05,
+      "loss": 1.1496,
+      "step": 1339
+    },
+    {
+      "epoch": 1.0141911069063387,
+      "grad_norm": 2.3640429973602295,
+      "learning_rate": 1.5099816111296968e-05,
+      "loss": 1.2001,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0149479659413434,
+      "grad_norm": 2.227107286453247,
+      "learning_rate": 1.5092908839139948e-05,
+      "loss": 1.1911,
+      "step": 1341
+    },
+    {
+      "epoch": 1.015704824976348,
+      "grad_norm": 2.3177998065948486,
+      "learning_rate": 1.5085998284816144e-05,
+      "loss": 1.1233,
+      "step": 1342
+    },
+    {
+      "epoch": 1.016461684011353,
+      "grad_norm": 2.167343854904175,
+      "learning_rate": 1.507908445277999e-05,
+      "loss": 1.2057,
+      "step": 1343
+    },
+    {
+      "epoch": 1.0172185430463576,
+      "grad_norm": 2.2151575088500977,
+      "learning_rate": 1.5072167347488042e-05,
+      "loss": 1.1828,
+      "step": 1344
+    },
+    {
+      "epoch": 1.0179754020813623,
+      "grad_norm": 2.031900405883789,
+      "learning_rate": 1.5065246973398959e-05,
+      "loss": 1.1408,
+      "step": 1345
+    },
+    {
+      "epoch": 1.0187322611163672,
+      "grad_norm": 2.3186428546905518,
+      "learning_rate": 1.5058323334973508e-05,
+      "loss": 1.1698,
+      "step": 1346
+    },
+    {
+      "epoch": 1.0194891201513718,
+      "grad_norm": 2.2243926525115967,
+      "learning_rate": 1.5051396436674562e-05,
+      "loss": 1.1983,
+      "step": 1347
+    },
+    {
+      "epoch": 1.0202459791863765,
+      "grad_norm": 2.0334129333496094,
+      "learning_rate": 1.5044466282967092e-05,
+      "loss": 1.1257,
+      "step": 1348
+    },
+    {
+      "epoch": 1.0210028382213812,
+      "grad_norm": 2.194042921066284,
+      "learning_rate": 1.503753287831817e-05,
+      "loss": 1.224,
+      "step": 1349
+    },
+    {
+      "epoch": 1.021759697256386,
+      "grad_norm": 2.2667534351348877,
+      "learning_rate": 1.5030596227196963e-05,
+      "loss": 1.2042,
+      "step": 1350
+    },
+    {
+      "epoch": 1.0225165562913907,
+      "grad_norm": 2.2039318084716797,
+      "learning_rate": 1.5023656334074732e-05,
+      "loss": 1.1965,
+      "step": 1351
+    },
+    {
+      "epoch": 1.0232734153263954,
+      "grad_norm": 2.3508946895599365,
+      "learning_rate": 1.5016713203424824e-05,
+      "loss": 1.1718,
+      "step": 1352
+    },
+    {
+      "epoch": 1.0240302743614003,
+      "grad_norm": 2.135310649871826,
+      "learning_rate": 1.5009766839722679e-05,
+      "loss": 1.1503,
+      "step": 1353
+    },
+    {
+      "epoch": 1.024787133396405,
+      "grad_norm": 2.2958900928497314,
+      "learning_rate": 1.5002817247445813e-05,
+      "loss": 1.2141,
+      "step": 1354
+    },
+    {
+      "epoch": 1.0255439924314096,
+      "grad_norm": 2.3174233436584473,
+      "learning_rate": 1.4995864431073828e-05,
+      "loss": 1.158,
+      "step": 1355
+    },
+    {
+      "epoch": 1.0263008514664145,
+      "grad_norm": 2.1523966789245605,
+      "learning_rate": 1.4988908395088405e-05,
+      "loss": 1.1757,
+      "step": 1356
+    },
+    {
+      "epoch": 1.0270577105014191,
+      "grad_norm": 2.2384963035583496,
+      "learning_rate": 1.4981949143973297e-05,
+      "loss": 1.1391,
+      "step": 1357
+    },
+    {
+      "epoch": 1.0278145695364238,
+      "grad_norm": 2.1168923377990723,
+      "learning_rate": 1.4974986682214332e-05,
+      "loss": 1.1306,
+      "step": 1358
+    },
+    {
+      "epoch": 1.0285714285714285,
+      "grad_norm": 2.392561435699463,
+      "learning_rate": 1.4968021014299409e-05,
+      "loss": 1.2224,
+      "step": 1359
+    },
+    {
+      "epoch": 1.0293282876064334,
+      "grad_norm": 2.2257487773895264,
+      "learning_rate": 1.4961052144718486e-05,
+      "loss": 1.1284,
+      "step": 1360
+    },
+    {
+      "epoch": 1.030085146641438,
+      "grad_norm": 2.0494847297668457,
+      "learning_rate": 1.4954080077963596e-05,
+      "loss": 1.1204,
+      "step": 1361
+    },
+    {
+      "epoch": 1.0308420056764427,
+      "grad_norm": 2.2043280601501465,
+      "learning_rate": 1.4947104818528822e-05,
+      "loss": 1.2135,
+      "step": 1362
+    },
+    {
+      "epoch": 1.0315988647114476,
+      "grad_norm": 2.1744041442871094,
+      "learning_rate": 1.494012637091031e-05,
+      "loss": 1.1533,
+      "step": 1363
+    },
+    {
+      "epoch": 1.0323557237464522,
+      "grad_norm": 2.1696369647979736,
+      "learning_rate": 1.4933144739606262e-05,
+      "loss": 1.171,
+      "step": 1364
+    },
+    {
+      "epoch": 1.033112582781457,
+      "grad_norm": 2.259871006011963,
+      "learning_rate": 1.4926159929116934e-05,
+      "loss": 1.1689,
+      "step": 1365
+    },
+    {
+      "epoch": 1.0338694418164618,
+      "grad_norm": 2.3883163928985596,
+      "learning_rate": 1.4919171943944628e-05,
+      "loss": 1.1808,
+      "step": 1366
+    },
+    {
+      "epoch": 1.0346263008514665,
+      "grad_norm": 2.4137635231018066,
+      "learning_rate": 1.4912180788593686e-05,
+      "loss": 1.2425,
+      "step": 1367
+    },
+    {
+      "epoch": 1.0353831598864711,
+      "grad_norm": 2.2681548595428467,
+      "learning_rate": 1.4905186467570509e-05,
+      "loss": 1.1808,
+      "step": 1368
+    },
+    {
+      "epoch": 1.0361400189214758,
+      "grad_norm": 2.2818410396575928,
+      "learning_rate": 1.4898188985383522e-05,
+      "loss": 1.1897,
+      "step": 1369
+    },
+    {
+      "epoch": 1.0368968779564807,
+      "grad_norm": 2.2151215076446533,
+      "learning_rate": 1.4891188346543201e-05,
+      "loss": 1.136,
+      "step": 1370
+    },
+    {
+      "epoch": 1.0376537369914853,
+      "grad_norm": 2.248666763305664,
+      "learning_rate": 1.488418455556205e-05,
+      "loss": 1.1758,
+      "step": 1371
+    },
+    {
+      "epoch": 1.03841059602649,
+      "grad_norm": 2.2258033752441406,
+      "learning_rate": 1.4877177616954602e-05,
+      "loss": 1.1628,
+      "step": 1372
+    },
+    {
+      "epoch": 1.0391674550614949,
+      "grad_norm": 2.117659091949463,
+      "learning_rate": 1.4870167535237428e-05,
+      "loss": 1.2191,
+      "step": 1373
+    },
+    {
+      "epoch": 1.0399243140964995,
+      "grad_norm": 2.367983102798462,
+      "learning_rate": 1.4863154314929114e-05,
+      "loss": 1.1415,
+      "step": 1374
+    },
+    {
+      "epoch": 1.0406811731315042,
+      "grad_norm": 2.134035587310791,
+      "learning_rate": 1.4856137960550278e-05,
+      "loss": 1.2032,
+      "step": 1375
+    },
+    {
+      "epoch": 1.0414380321665089,
+      "grad_norm": 2.392430543899536,
+      "learning_rate": 1.4849118476623556e-05,
+      "loss": 1.197,
+      "step": 1376
+    },
+    {
+      "epoch": 1.0421948912015138,
+      "grad_norm": 2.237036943435669,
+      "learning_rate": 1.4842095867673603e-05,
+      "loss": 1.168,
+      "step": 1377
+    },
+    {
+      "epoch": 1.0429517502365184,
+      "grad_norm": 2.338472604751587,
+      "learning_rate": 1.4835070138227077e-05,
+      "loss": 1.1629,
+      "step": 1378
+    },
+    {
+      "epoch": 1.043708609271523,
+      "grad_norm": 2.1505656242370605,
+      "learning_rate": 1.4828041292812662e-05,
+      "loss": 1.1559,
+      "step": 1379
+    },
+    {
+      "epoch": 1.044465468306528,
+      "grad_norm": 2.155229330062866,
+      "learning_rate": 1.4821009335961045e-05,
+      "loss": 1.1477,
+      "step": 1380
+    },
+    {
+      "epoch": 1.0452223273415326,
+      "grad_norm": 2.1831212043762207,
+      "learning_rate": 1.4813974272204918e-05,
+      "loss": 1.1486,
+      "step": 1381
+    },
+    {
+      "epoch": 1.0459791863765373,
+      "grad_norm": 2.2904438972473145,
+      "learning_rate": 1.4806936106078971e-05,
+      "loss": 1.1605,
+      "step": 1382
+    },
+    {
+      "epoch": 1.0467360454115422,
+      "grad_norm": 2.416222333908081,
+      "learning_rate": 1.4799894842119906e-05,
+      "loss": 1.1161,
+      "step": 1383
+    },
+    {
+      "epoch": 1.0474929044465469,
+      "grad_norm": 2.2631683349609375,
+      "learning_rate": 1.4792850484866408e-05,
+      "loss": 1.173,
+      "step": 1384
+    },
+    {
+      "epoch": 1.0482497634815515,
+      "grad_norm": 2.2983131408691406,
+      "learning_rate": 1.4785803038859166e-05,
+      "loss": 1.1584,
+      "step": 1385
+    },
+    {
+      "epoch": 1.0490066225165562,
+      "grad_norm": 2.1680402755737305,
+      "learning_rate": 1.4778752508640852e-05,
+      "loss": 1.1689,
+      "step": 1386
+    },
+    {
+      "epoch": 1.049763481551561,
+      "grad_norm": 2.161684036254883,
+      "learning_rate": 1.4771698898756137e-05,
+      "loss": 1.1772,
+      "step": 1387
+    },
+    {
+      "epoch": 1.0505203405865657,
+      "grad_norm": 2.048295021057129,
+      "learning_rate": 1.4764642213751664e-05,
+      "loss": 1.1598,
+      "step": 1388
+    },
+    {
+      "epoch": 1.0512771996215704,
+      "grad_norm": 2.0943684577941895,
+      "learning_rate": 1.4757582458176067e-05,
+      "loss": 1.1389,
+      "step": 1389
+    },
+    {
+      "epoch": 1.0520340586565753,
+      "grad_norm": 2.5327534675598145,
+      "learning_rate": 1.475051963657996e-05,
+      "loss": 1.1496,
+      "step": 1390
+    },
+    {
+      "epoch": 1.05279091769158,
+      "grad_norm": 2.1597166061401367,
+      "learning_rate": 1.4743453753515924e-05,
+      "loss": 1.1409,
+      "step": 1391
+    },
+    {
+      "epoch": 1.0535477767265846,
+      "grad_norm": 2.877094268798828,
+      "learning_rate": 1.4736384813538527e-05,
+      "loss": 1.1359,
+      "step": 1392
+    },
+    {
+      "epoch": 1.0543046357615895,
+      "grad_norm": 2.4092910289764404,
+      "learning_rate": 1.472931282120429e-05,
+      "loss": 1.1673,
+      "step": 1393
+    },
+    {
+      "epoch": 1.0550614947965942,
+      "grad_norm": 2.26458740234375,
+      "learning_rate": 1.4722237781071717e-05,
+      "loss": 1.167,
+      "step": 1394
+    },
+    {
+      "epoch": 1.0558183538315988,
+      "grad_norm": 2.2418289184570312,
+      "learning_rate": 1.4715159697701276e-05,
+      "loss": 1.1674,
+      "step": 1395
+    },
+    {
+      "epoch": 1.0565752128666035,
+      "grad_norm": 2.948460102081299,
+      "learning_rate": 1.470807857565538e-05,
+      "loss": 1.1459,
+      "step": 1396
+    },
+    {
+      "epoch": 1.0573320719016084,
+      "grad_norm": 2.1875085830688477,
+      "learning_rate": 1.4700994419498423e-05,
+      "loss": 1.1781,
+      "step": 1397
+    },
+    {
+      "epoch": 1.058088930936613,
+      "grad_norm": 2.3909361362457275,
+      "learning_rate": 1.4693907233796737e-05,
+      "loss": 1.1346,
+      "step": 1398
+    },
+    {
+      "epoch": 1.0588457899716177,
+      "grad_norm": 2.12752628326416,
+      "learning_rate": 1.4686817023118619e-05,
+      "loss": 1.1127,
+      "step": 1399
+    },
+    {
+      "epoch": 1.0596026490066226,
+      "grad_norm": 2.3758580684661865,
+      "learning_rate": 1.4679723792034304e-05,
+      "loss": 1.1667,
+      "step": 1400
+    },
+    {
+      "epoch": 1.0603595080416273,
+      "grad_norm": 2.23144793510437,
+      "learning_rate": 1.4672627545115991e-05,
+      "loss": 1.1693,
+      "step": 1401
+    },
+    {
+      "epoch": 1.061116367076632,
+      "grad_norm": 2.2588181495666504,
+      "learning_rate": 1.46655282869378e-05,
+      "loss": 1.1555,
+      "step": 1402
+    },
+    {
+      "epoch": 1.0618732261116368,
+      "grad_norm": 3.266263961791992,
+      "learning_rate": 1.4658426022075816e-05,
+      "loss": 1.1475,
+      "step": 1403
+    },
+    {
+      "epoch": 1.0626300851466415,
+      "grad_norm": 2.344022750854492,
+      "learning_rate": 1.4651320755108042e-05,
+      "loss": 1.1151,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0633869441816461,
+      "grad_norm": 2.701164722442627,
+      "learning_rate": 1.464421249061443e-05,
+      "loss": 1.1547,
+      "step": 1405
+    },
+    {
+      "epoch": 1.0641438032166508,
+      "grad_norm": 2.4534714221954346,
+      "learning_rate": 1.4637101233176856e-05,
+      "loss": 1.17,
+      "step": 1406
+    },
+    {
+      "epoch": 1.0649006622516557,
+      "grad_norm": 2.413388252258301,
+      "learning_rate": 1.462998698737913e-05,
+      "loss": 1.1852,
+      "step": 1407
+    },
+    {
+      "epoch": 1.0656575212866604,
+      "grad_norm": 2.2986247539520264,
+      "learning_rate": 1.4622869757806983e-05,
+      "loss": 1.1544,
+      "step": 1408
+    },
+    {
+      "epoch": 1.066414380321665,
+      "grad_norm": 2.308239221572876,
+      "learning_rate": 1.4615749549048076e-05,
+      "loss": 1.1572,
+      "step": 1409
+    },
+    {
+      "epoch": 1.06717123935667,
+      "grad_norm": 2.270495653152466,
+      "learning_rate": 1.4608626365691986e-05,
+      "loss": 1.1625,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0679280983916746,
+      "grad_norm": 2.0776920318603516,
+      "learning_rate": 1.4601500212330213e-05,
+      "loss": 1.1879,
+      "step": 1411
+    },
+    {
+      "epoch": 1.0686849574266792,
+      "grad_norm": 2.279533863067627,
+      "learning_rate": 1.4594371093556159e-05,
+      "loss": 1.1844,
+      "step": 1412
+    },
+    {
+      "epoch": 1.069441816461684,
+      "grad_norm": 2.263552188873291,
+      "learning_rate": 1.4587239013965149e-05,
+      "loss": 1.1192,
+      "step": 1413
+    },
+    {
+      "epoch": 1.0701986754966888,
+      "grad_norm": 2.1875579357147217,
+      "learning_rate": 1.4580103978154414e-05,
+      "loss": 1.1921,
+      "step": 1414
+    },
+    {
+      "epoch": 1.0709555345316935,
+      "grad_norm": 2.553298234939575,
+      "learning_rate": 1.4572965990723083e-05,
+      "loss": 1.1307,
+      "step": 1415
+    },
+    {
+      "epoch": 1.0717123935666981,
+      "grad_norm": 2.0610175132751465,
+      "learning_rate": 1.4565825056272199e-05,
+      "loss": 1.2057,
+      "step": 1416
+    },
+    {
+      "epoch": 1.072469252601703,
+      "grad_norm": 3.3085532188415527,
+      "learning_rate": 1.4558681179404704e-05,
+      "loss": 1.1566,
+      "step": 1417
+    },
+    {
+      "epoch": 1.0732261116367077,
+      "grad_norm": 2.1448001861572266,
+      "learning_rate": 1.4551534364725422e-05,
+      "loss": 1.1958,
+      "step": 1418
+    },
+    {
+      "epoch": 1.0739829706717123,
+      "grad_norm": 2.5602312088012695,
+      "learning_rate": 1.4544384616841084e-05,
+      "loss": 1.1513,
+      "step": 1419
+    },
+    {
+      "epoch": 1.0747398297067172,
+      "grad_norm": 2.34245228767395,
+      "learning_rate": 1.4537231940360315e-05,
+      "loss": 1.1331,
+      "step": 1420
+    },
+    {
+      "epoch": 1.0754966887417219,
+      "grad_norm": 2.5702669620513916,
+      "learning_rate": 1.4530076339893615e-05,
+      "loss": 1.1369,
+      "step": 1421
+    },
+    {
+      "epoch": 1.0762535477767265,
+      "grad_norm": 2.3624837398529053,
+      "learning_rate": 1.4522917820053375e-05,
+      "loss": 1.1618,
+      "step": 1422
+    },
+    {
+      "epoch": 1.0770104068117312,
+      "grad_norm": 2.3341963291168213,
+      "learning_rate": 1.4515756385453868e-05,
+      "loss": 1.1688,
+      "step": 1423
+    },
+    {
+      "epoch": 1.077767265846736,
+      "grad_norm": 2.32336688041687,
+      "learning_rate": 1.4508592040711246e-05,
+      "loss": 1.136,
+      "step": 1424
+    },
+    {
+      "epoch": 1.0785241248817408,
+      "grad_norm": 2.4956133365631104,
+      "learning_rate": 1.4501424790443544e-05,
+      "loss": 1.1893,
+      "step": 1425
+    },
+    {
+      "epoch": 1.0792809839167454,
+      "grad_norm": 2.5766842365264893,
+      "learning_rate": 1.4494254639270646e-05,
+      "loss": 1.104,
+      "step": 1426
+    },
+    {
+      "epoch": 1.0800378429517503,
+      "grad_norm": 2.3494839668273926,
+      "learning_rate": 1.4487081591814336e-05,
+      "loss": 1.1509,
+      "step": 1427
+    },
+    {
+      "epoch": 1.080794701986755,
+      "grad_norm": 2.25639009475708,
+      "learning_rate": 1.4479905652698248e-05,
+      "loss": 1.1518,
+      "step": 1428
+    },
+    {
+      "epoch": 1.0815515610217596,
+      "grad_norm": 2.454833984375,
+      "learning_rate": 1.4472726826547876e-05,
+      "loss": 1.2164,
+      "step": 1429
+    },
+    {
+      "epoch": 1.0823084200567645,
+      "grad_norm": 2.320312976837158,
+      "learning_rate": 1.4465545117990587e-05,
+      "loss": 1.1562,
+      "step": 1430
+    },
+    {
+      "epoch": 1.0830652790917692,
+      "grad_norm": 2.136070966720581,
+      "learning_rate": 1.4458360531655606e-05,
+      "loss": 1.1621,
+      "step": 1431
+    },
+    {
+      "epoch": 1.0838221381267739,
+      "grad_norm": 2.3895716667175293,
+      "learning_rate": 1.4451173072173996e-05,
+      "loss": 1.1442,
+      "step": 1432
+    },
+    {
+      "epoch": 1.0845789971617785,
+      "grad_norm": 2.185600757598877,
+      "learning_rate": 1.4443982744178694e-05,
+      "loss": 1.1352,
+      "step": 1433
+    },
+    {
+      "epoch": 1.0853358561967834,
+      "grad_norm": 2.2408831119537354,
+      "learning_rate": 1.4436789552304471e-05,
+      "loss": 1.1771,
+      "step": 1434
+    },
+    {
+      "epoch": 1.086092715231788,
+      "grad_norm": 2.2491331100463867,
+      "learning_rate": 1.4429593501187952e-05,
+      "loss": 1.1509,
+      "step": 1435
+    },
+    {
+      "epoch": 1.0868495742667927,
+      "grad_norm": 2.4954214096069336,
+      "learning_rate": 1.4422394595467597e-05,
+      "loss": 1.1333,
+      "step": 1436
+    },
+    {
+      "epoch": 1.0876064333017976,
+      "grad_norm": 2.427107572555542,
+      "learning_rate": 1.4415192839783716e-05,
+      "loss": 1.1806,
+      "step": 1437
+    },
+    {
+      "epoch": 1.0883632923368023,
+      "grad_norm": 5.063598155975342,
+      "learning_rate": 1.4407988238778448e-05,
+      "loss": 1.1364,
+      "step": 1438
+    },
+    {
+      "epoch": 1.089120151371807,
+      "grad_norm": 2.261101245880127,
+      "learning_rate": 1.4400780797095769e-05,
+      "loss": 1.171,
+      "step": 1439
+    },
+    {
+      "epoch": 1.0898770104068118,
+      "grad_norm": 2.3400819301605225,
+      "learning_rate": 1.4393570519381484e-05,
+      "loss": 1.1354,
+      "step": 1440
+    },
+    {
+      "epoch": 1.0906338694418165,
+      "grad_norm": 2.225931167602539,
+      "learning_rate": 1.438635741028323e-05,
+      "loss": 1.1674,
+      "step": 1441
+    },
+    {
+      "epoch": 1.0913907284768212,
+      "grad_norm": 2.3904130458831787,
+      "learning_rate": 1.437914147445047e-05,
+      "loss": 1.1513,
+      "step": 1442
+    },
+    {
+      "epoch": 1.0921475875118258,
+      "grad_norm": 2.4583778381347656,
+      "learning_rate": 1.4371922716534483e-05,
+      "loss": 1.1708,
+      "step": 1443
+    },
+    {
+      "epoch": 1.0929044465468307,
+      "grad_norm": 2.270364761352539,
+      "learning_rate": 1.436470114118837e-05,
+      "loss": 1.1708,
+      "step": 1444
+    },
+    {
+      "epoch": 1.0936613055818354,
+      "grad_norm": 2.190642833709717,
+      "learning_rate": 1.4357476753067053e-05,
+      "loss": 1.123,
+      "step": 1445
+    },
+    {
+      "epoch": 1.09441816461684,
+      "grad_norm": 2.3521080017089844,
+      "learning_rate": 1.4350249556827256e-05,
+      "loss": 1.1793,
+      "step": 1446
+    },
+    {
+      "epoch": 1.095175023651845,
+      "grad_norm": 2.4334235191345215,
+      "learning_rate": 1.4343019557127522e-05,
+      "loss": 1.175,
+      "step": 1447
+    },
+    {
+      "epoch": 1.0959318826868496,
+      "grad_norm": 2.2919211387634277,
+      "learning_rate": 1.4335786758628199e-05,
+      "loss": 1.1601,
+      "step": 1448
+    },
+    {
+      "epoch": 1.0966887417218543,
+      "grad_norm": 2.603358745574951,
+      "learning_rate": 1.4328551165991435e-05,
+      "loss": 1.1966,
+      "step": 1449
+    },
+    {
+      "epoch": 1.097445600756859,
+      "grad_norm": 2.47110915184021,
+      "learning_rate": 1.4321312783881192e-05,
+      "loss": 1.1623,
+      "step": 1450
+    },
+    {
+      "epoch": 1.0982024597918638,
+      "grad_norm": 2.54114031791687,
+      "learning_rate": 1.431407161696321e-05,
+      "loss": 1.1598,
+      "step": 1451
+    },
+    {
+      "epoch": 1.0989593188268685,
+      "grad_norm": 2.1958112716674805,
+      "learning_rate": 1.4306827669905041e-05,
+      "loss": 1.1317,
+      "step": 1452
+    },
+    {
+      "epoch": 1.0997161778618731,
+      "grad_norm": 2.2989349365234375,
+      "learning_rate": 1.4299580947376022e-05,
+      "loss": 1.1036,
+      "step": 1453
+    },
+    {
+      "epoch": 1.100473036896878,
+      "grad_norm": 2.331904411315918,
+      "learning_rate": 1.4292331454047278e-05,
+      "loss": 1.1331,
+      "step": 1454
+    },
+    {
+      "epoch": 1.1012298959318827,
+      "grad_norm": 2.376122236251831,
+      "learning_rate": 1.4285079194591722e-05,
+      "loss": 1.1649,
+      "step": 1455
+    },
+    {
+      "epoch": 1.1019867549668874,
+      "grad_norm": 2.453084945678711,
+      "learning_rate": 1.4277824173684056e-05,
+      "loss": 1.1636,
+      "step": 1456
+    },
+    {
+      "epoch": 1.1027436140018922,
+      "grad_norm": 2.4421586990356445,
+      "learning_rate": 1.4270566396000744e-05,
+      "loss": 1.1323,
+      "step": 1457
+    },
+    {
+      "epoch": 1.103500473036897,
+      "grad_norm": 2.308035135269165,
+      "learning_rate": 1.426330586622005e-05,
+      "loss": 1.13,
+      "step": 1458
+    },
+    {
+      "epoch": 1.1042573320719016,
+      "grad_norm": 2.403162956237793,
+      "learning_rate": 1.4256042589021994e-05,
+      "loss": 1.181,
+      "step": 1459
+    },
+    {
+      "epoch": 1.1050141911069062,
+      "grad_norm": 2.4109246730804443,
+      "learning_rate": 1.4248776569088377e-05,
+      "loss": 1.1597,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1057710501419111,
+      "grad_norm": 2.412398099899292,
+      "learning_rate": 1.4241507811102762e-05,
+      "loss": 1.118,
+      "step": 1461
+    },
+    {
+      "epoch": 1.1065279091769158,
+      "grad_norm": 2.8465075492858887,
+      "learning_rate": 1.4234236319750482e-05,
+      "loss": 1.1618,
+      "step": 1462
+    },
+    {
+      "epoch": 1.1072847682119205,
+      "grad_norm": 2.306621789932251,
+      "learning_rate": 1.4226962099718628e-05,
+      "loss": 1.2062,
+      "step": 1463
+    },
+    {
+      "epoch": 1.1080416272469253,
+      "grad_norm": 2.6074671745300293,
+      "learning_rate": 1.4219685155696053e-05,
+      "loss": 1.1676,
+      "step": 1464
+    },
+    {
+      "epoch": 1.10879848628193,
+      "grad_norm": 2.509995460510254,
+      "learning_rate": 1.421240549237336e-05,
+      "loss": 1.1771,
+      "step": 1465
+    },
+    {
+      "epoch": 1.1095553453169347,
+      "grad_norm": 2.535238742828369,
+      "learning_rate": 1.4205123114442916e-05,
+      "loss": 1.1682,
+      "step": 1466
+    },
+    {
+      "epoch": 1.1103122043519393,
+      "grad_norm": 2.4258975982666016,
+      "learning_rate": 1.4197838026598826e-05,
+      "loss": 1.0947,
+      "step": 1467
+    },
+    {
+      "epoch": 1.1110690633869442,
+      "grad_norm": 2.5997817516326904,
+      "learning_rate": 1.4190550233536946e-05,
+      "loss": 1.1471,
+      "step": 1468
+    },
+    {
+      "epoch": 1.1118259224219489,
+      "grad_norm": 2.358372449874878,
+      "learning_rate": 1.4183259739954877e-05,
+      "loss": 1.1564,
+      "step": 1469
+    },
+    {
+      "epoch": 1.1125827814569536,
+      "grad_norm": 2.323791027069092,
+      "learning_rate": 1.4175966550551963e-05,
+      "loss": 1.1936,
+      "step": 1470
+    },
+    {
+      "epoch": 1.1133396404919584,
+      "grad_norm": 2.334627151489258,
+      "learning_rate": 1.4168670670029277e-05,
+      "loss": 1.1514,
+      "step": 1471
+    },
+    {
+      "epoch": 1.114096499526963,
+      "grad_norm": 2.2344837188720703,
+      "learning_rate": 1.4161372103089637e-05,
+      "loss": 1.1378,
+      "step": 1472
+    },
+    {
+      "epoch": 1.1148533585619678,
+      "grad_norm": 2.200742483139038,
+      "learning_rate": 1.4154070854437587e-05,
+      "loss": 1.1783,
+      "step": 1473
+    },
+    {
+      "epoch": 1.1156102175969727,
+      "grad_norm": 2.2466723918914795,
+      "learning_rate": 1.4146766928779396e-05,
+      "loss": 1.1419,
+      "step": 1474
+    },
+    {
+      "epoch": 1.1163670766319773,
+      "grad_norm": 2.4173378944396973,
+      "learning_rate": 1.4139460330823071e-05,
+      "loss": 1.0991,
+      "step": 1475
+    },
+    {
+      "epoch": 1.117123935666982,
+      "grad_norm": 2.3149657249450684,
+      "learning_rate": 1.413215106527833e-05,
+      "loss": 1.1419,
+      "step": 1476
+    },
+    {
+      "epoch": 1.1178807947019869,
+      "grad_norm": 2.2564306259155273,
+      "learning_rate": 1.4124839136856612e-05,
+      "loss": 1.1693,
+      "step": 1477
+    },
+    {
+      "epoch": 1.1186376537369915,
+      "grad_norm": 2.131028652191162,
+      "learning_rate": 1.4117524550271077e-05,
+      "loss": 1.158,
+      "step": 1478
+    },
+    {
+      "epoch": 1.1193945127719962,
+      "grad_norm": 2.4710068702697754,
+      "learning_rate": 1.4110207310236595e-05,
+      "loss": 1.1934,
+      "step": 1479
+    },
+    {
+      "epoch": 1.1201513718070009,
+      "grad_norm": 2.138939380645752,
+      "learning_rate": 1.4102887421469747e-05,
+      "loss": 1.1196,
+      "step": 1480
+    },
+    {
+      "epoch": 1.1209082308420057,
+      "grad_norm": 2.542495012283325,
+      "learning_rate": 1.4095564888688822e-05,
+      "loss": 1.1693,
+      "step": 1481
+    },
+    {
+      "epoch": 1.1216650898770104,
+      "grad_norm": 2.4574832916259766,
+      "learning_rate": 1.4088239716613816e-05,
+      "loss": 1.1248,
+      "step": 1482
+    },
+    {
+      "epoch": 1.122421948912015,
+      "grad_norm": 2.133028268814087,
+      "learning_rate": 1.4080911909966419e-05,
+      "loss": 1.2361,
+      "step": 1483
+    },
+    {
+      "epoch": 1.12317880794702,
+      "grad_norm": 2.624393939971924,
+      "learning_rate": 1.4073581473470023e-05,
+      "loss": 1.1053,
+      "step": 1484
+    },
+    {
+      "epoch": 1.1239356669820246,
+      "grad_norm": 2.0480175018310547,
+      "learning_rate": 1.4066248411849717e-05,
+      "loss": 1.1364,
+      "step": 1485
+    },
+    {
+      "epoch": 1.1246925260170293,
+      "grad_norm": 2.2111339569091797,
+      "learning_rate": 1.4058912729832286e-05,
+      "loss": 1.1869,
+      "step": 1486
+    },
+    {
+      "epoch": 1.125449385052034,
+      "grad_norm": 2.4910013675689697,
+      "learning_rate": 1.4051574432146191e-05,
+      "loss": 1.16,
+      "step": 1487
+    },
+    {
+      "epoch": 1.1262062440870388,
+      "grad_norm": 2.419105052947998,
+      "learning_rate": 1.4044233523521587e-05,
+      "loss": 1.1637,
+      "step": 1488
+    },
+    {
+      "epoch": 1.1269631031220435,
+      "grad_norm": 2.4131598472595215,
+      "learning_rate": 1.4036890008690316e-05,
+      "loss": 1.1814,
+      "step": 1489
+    },
+    {
+      "epoch": 1.1277199621570482,
+      "grad_norm": 2.395854949951172,
+      "learning_rate": 1.4029543892385898e-05,
+      "loss": 1.1535,
+      "step": 1490
+    },
+    {
+      "epoch": 1.128476821192053,
+      "grad_norm": 2.0963070392608643,
+      "learning_rate": 1.4022195179343518e-05,
+      "loss": 1.1366,
+      "step": 1491
+    },
+    {
+      "epoch": 1.1292336802270577,
+      "grad_norm": 2.267829418182373,
+      "learning_rate": 1.4014843874300052e-05,
+      "loss": 1.1393,
+      "step": 1492
+    },
+    {
+      "epoch": 1.1299905392620624,
+      "grad_norm": 2.1519582271575928,
+      "learning_rate": 1.4007489981994038e-05,
+      "loss": 1.1728,
+      "step": 1493
+    },
+    {
+      "epoch": 1.1307473982970673,
+      "grad_norm": 2.194342613220215,
+      "learning_rate": 1.4000133507165684e-05,
+      "loss": 1.1586,
+      "step": 1494
+    },
+    {
+      "epoch": 1.131504257332072,
+      "grad_norm": 2.3476803302764893,
+      "learning_rate": 1.3992774454556855e-05,
+      "loss": 1.2297,
+      "step": 1495
+    },
+    {
+      "epoch": 1.1322611163670766,
+      "grad_norm": 2.1007235050201416,
+      "learning_rate": 1.3985412828911088e-05,
+      "loss": 1.1605,
+      "step": 1496
+    },
+    {
+      "epoch": 1.1330179754020813,
+      "grad_norm": 2.512786388397217,
+      "learning_rate": 1.397804863497358e-05,
+      "loss": 1.1765,
+      "step": 1497
+    },
+    {
+      "epoch": 1.1337748344370862,
+      "grad_norm": 2.1948659420013428,
+      "learning_rate": 1.397068187749117e-05,
+      "loss": 1.1912,
+      "step": 1498
+    },
+    {
+      "epoch": 1.1345316934720908,
+      "grad_norm": 2.348325729370117,
+      "learning_rate": 1.3963312561212359e-05,
+      "loss": 1.152,
+      "step": 1499
+    },
+    {
+      "epoch": 1.1352885525070955,
+      "grad_norm": 2.088045597076416,
+      "learning_rate": 1.3955940690887301e-05,
+      "loss": 1.0803,
+      "step": 1500
+    },
+    {
+      "epoch": 1.1360454115421004,
+      "grad_norm": 2.110816240310669,
+      "learning_rate": 1.3948566271267784e-05,
+      "loss": 1.1599,
+      "step": 1501
+    },
+    {
+      "epoch": 1.136802270577105,
+      "grad_norm": 2.306739330291748,
+      "learning_rate": 1.3941189307107255e-05,
+      "loss": 1.2,
+      "step": 1502
+    },
+    {
+      "epoch": 1.1375591296121097,
+      "grad_norm": 2.494978666305542,
+      "learning_rate": 1.3933809803160784e-05,
+      "loss": 1.1418,
+      "step": 1503
+    },
+    {
+      "epoch": 1.1383159886471144,
+      "grad_norm": 2.4510955810546875,
+      "learning_rate": 1.3926427764185093e-05,
+      "loss": 1.1455,
+      "step": 1504
+    },
+    {
+      "epoch": 1.1390728476821192,
+      "grad_norm": 2.415323495864868,
+      "learning_rate": 1.3919043194938528e-05,
+      "loss": 1.1361,
+      "step": 1505
+    },
+    {
+      "epoch": 1.139829706717124,
+      "grad_norm": 2.263831615447998,
+      "learning_rate": 1.391165610018107e-05,
+      "loss": 1.1087,
+      "step": 1506
+    },
+    {
+      "epoch": 1.1405865657521286,
+      "grad_norm": 2.5898752212524414,
+      "learning_rate": 1.3904266484674331e-05,
+      "loss": 1.1339,
+      "step": 1507
+    },
+    {
+      "epoch": 1.1413434247871335,
+      "grad_norm": 2.153635263442993,
+      "learning_rate": 1.3896874353181542e-05,
+      "loss": 1.1024,
+      "step": 1508
+    },
+    {
+      "epoch": 1.1421002838221381,
+      "grad_norm": 2.095327138900757,
+      "learning_rate": 1.3889479710467557e-05,
+      "loss": 1.2094,
+      "step": 1509
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 2.1070072650909424,
+      "learning_rate": 1.388208256129885e-05,
+      "loss": 1.143,
+      "step": 1510
+    },
+    {
+      "epoch": 1.1436140018921477,
+      "grad_norm": 2.2180447578430176,
+      "learning_rate": 1.3874682910443516e-05,
+      "loss": 1.1682,
+      "step": 1511
+    },
+    {
+      "epoch": 1.1443708609271523,
+      "grad_norm": 2.3871870040893555,
+      "learning_rate": 1.3867280762671246e-05,
+      "loss": 1.1671,
+      "step": 1512
+    },
+    {
+      "epoch": 1.145127719962157,
+      "grad_norm": 2.165802001953125,
+      "learning_rate": 1.3859876122753363e-05,
+      "loss": 1.1138,
+      "step": 1513
+    },
+    {
+      "epoch": 1.145884578997162,
+      "grad_norm": 2.162033796310425,
+      "learning_rate": 1.3852468995462785e-05,
+      "loss": 1.1719,
+      "step": 1514
+    },
+    {
+      "epoch": 1.1466414380321666,
+      "grad_norm": 2.163429021835327,
+      "learning_rate": 1.3845059385574023e-05,
+      "loss": 1.1483,
+      "step": 1515
+    },
+    {
+      "epoch": 1.1473982970671712,
+      "grad_norm": 2.177055597305298,
+      "learning_rate": 1.3837647297863203e-05,
+      "loss": 1.1372,
+      "step": 1516
+    },
+    {
+      "epoch": 1.1481551561021759,
+      "grad_norm": 2.3266491889953613,
+      "learning_rate": 1.383023273710805e-05,
+      "loss": 1.1825,
+      "step": 1517
+    },
+    {
+      "epoch": 1.1489120151371808,
+      "grad_norm": 2.4812657833099365,
+      "learning_rate": 1.3822815708087865e-05,
+      "loss": 1.1697,
+      "step": 1518
+    },
+    {
+      "epoch": 1.1496688741721854,
+      "grad_norm": 2.1462526321411133,
+      "learning_rate": 1.3815396215583564e-05,
+      "loss": 1.1203,
+      "step": 1519
+    },
+    {
+      "epoch": 1.15042573320719,
+      "grad_norm": 2.160487174987793,
+      "learning_rate": 1.3807974264377629e-05,
+      "loss": 1.1322,
+      "step": 1520
+    },
+    {
+      "epoch": 1.1511825922421948,
+      "grad_norm": 2.234320640563965,
+      "learning_rate": 1.3800549859254144e-05,
+      "loss": 1.1393,
+      "step": 1521
+    },
+    {
+      "epoch": 1.1519394512771997,
+      "grad_norm": 2.4396426677703857,
+      "learning_rate": 1.3793123004998765e-05,
+      "loss": 1.1671,
+      "step": 1522
+    },
+    {
+      "epoch": 1.1526963103122043,
+      "grad_norm": 2.3154118061065674,
+      "learning_rate": 1.3785693706398724e-05,
+      "loss": 1.1867,
+      "step": 1523
+    },
+    {
+      "epoch": 1.153453169347209,
+      "grad_norm": 2.2119319438934326,
+      "learning_rate": 1.377826196824284e-05,
+      "loss": 1.1582,
+      "step": 1524
+    },
+    {
+      "epoch": 1.1542100283822139,
+      "grad_norm": 2.2084405422210693,
+      "learning_rate": 1.3770827795321495e-05,
+      "loss": 1.1613,
+      "step": 1525
+    },
+    {
+      "epoch": 1.1549668874172185,
+      "grad_norm": 2.341912031173706,
+      "learning_rate": 1.3763391192426644e-05,
+      "loss": 1.1519,
+      "step": 1526
+    },
+    {
+      "epoch": 1.1557237464522232,
+      "grad_norm": 2.2736034393310547,
+      "learning_rate": 1.3755952164351814e-05,
+      "loss": 1.1465,
+      "step": 1527
+    },
+    {
+      "epoch": 1.156480605487228,
+      "grad_norm": 2.2359468936920166,
+      "learning_rate": 1.3748510715892075e-05,
+      "loss": 1.193,
+      "step": 1528
+    },
+    {
+      "epoch": 1.1572374645222328,
+      "grad_norm": 1.9430551528930664,
+      "learning_rate": 1.3741066851844082e-05,
+      "loss": 1.139,
+      "step": 1529
+    },
+    {
+      "epoch": 1.1579943235572374,
+      "grad_norm": 2.0962564945220947,
+      "learning_rate": 1.3733620577006035e-05,
+      "loss": 1.1442,
+      "step": 1530
+    },
+    {
+      "epoch": 1.1587511825922423,
+      "grad_norm": 2.0686581134796143,
+      "learning_rate": 1.3726171896177687e-05,
+      "loss": 1.1778,
+      "step": 1531
+    },
+    {
+      "epoch": 1.159508041627247,
+      "grad_norm": 2.120643138885498,
+      "learning_rate": 1.3718720814160342e-05,
+      "loss": 1.1789,
+      "step": 1532
+    },
+    {
+      "epoch": 1.1602649006622516,
+      "grad_norm": 2.4168310165405273,
+      "learning_rate": 1.3711267335756862e-05,
+      "loss": 1.1816,
+      "step": 1533
+    },
+    {
+      "epoch": 1.1610217596972563,
+      "grad_norm": 2.183661460876465,
+      "learning_rate": 1.3703811465771636e-05,
+      "loss": 1.1861,
+      "step": 1534
+    },
+    {
+      "epoch": 1.1617786187322612,
+      "grad_norm": 2.196077346801758,
+      "learning_rate": 1.3696353209010609e-05,
+      "loss": 1.1845,
+      "step": 1535
+    },
+    {
+      "epoch": 1.1625354777672658,
+      "grad_norm": 2.280958890914917,
+      "learning_rate": 1.3688892570281261e-05,
+      "loss": 1.1371,
+      "step": 1536
+    },
+    {
+      "epoch": 1.1632923368022705,
+      "grad_norm": 2.3048434257507324,
+      "learning_rate": 1.3681429554392602e-05,
+      "loss": 1.15,
+      "step": 1537
+    },
+    {
+      "epoch": 1.1640491958372754,
+      "grad_norm": 2.177098512649536,
+      "learning_rate": 1.367396416615518e-05,
+      "loss": 1.1537,
+      "step": 1538
+    },
+    {
+      "epoch": 1.16480605487228,
+      "grad_norm": 2.419185161590576,
+      "learning_rate": 1.3666496410381072e-05,
+      "loss": 1.1634,
+      "step": 1539
+    },
+    {
+      "epoch": 1.1655629139072847,
+      "grad_norm": 2.5214250087738037,
+      "learning_rate": 1.3659026291883874e-05,
+      "loss": 1.1669,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1663197729422894,
+      "grad_norm": 2.3618457317352295,
+      "learning_rate": 1.365155381547872e-05,
+      "loss": 1.2169,
+      "step": 1541
+    },
+    {
+      "epoch": 1.1670766319772943,
+      "grad_norm": 2.3737759590148926,
+      "learning_rate": 1.3644078985982243e-05,
+      "loss": 1.2004,
+      "step": 1542
+    },
+    {
+      "epoch": 1.167833491012299,
+      "grad_norm": 2.500761032104492,
+      "learning_rate": 1.3636601808212613e-05,
+      "loss": 1.1576,
+      "step": 1543
+    },
+    {
+      "epoch": 1.1685903500473036,
+      "grad_norm": 2.429725170135498,
+      "learning_rate": 1.36291222869895e-05,
+      "loss": 1.1412,
+      "step": 1544
+    },
+    {
+      "epoch": 1.1693472090823085,
+      "grad_norm": 2.4820287227630615,
+      "learning_rate": 1.3621640427134095e-05,
+      "loss": 1.1256,
+      "step": 1545
+    },
+    {
+      "epoch": 1.1701040681173132,
+      "grad_norm": 2.5075745582580566,
+      "learning_rate": 1.3614156233469081e-05,
+      "loss": 1.1426,
+      "step": 1546
+    },
+    {
+      "epoch": 1.1708609271523178,
+      "grad_norm": 2.5569803714752197,
+      "learning_rate": 1.3606669710818665e-05,
+      "loss": 1.1438,
+      "step": 1547
+    },
+    {
+      "epoch": 1.1716177861873227,
+      "grad_norm": 2.4151296615600586,
+      "learning_rate": 1.3599180864008538e-05,
+      "loss": 1.1325,
+      "step": 1548
+    },
+    {
+      "epoch": 1.1723746452223274,
+      "grad_norm": 2.1855361461639404,
+      "learning_rate": 1.3591689697865902e-05,
+      "loss": 1.1328,
+      "step": 1549
+    },
+    {
+      "epoch": 1.173131504257332,
+      "grad_norm": 2.130683422088623,
+      "learning_rate": 1.3584196217219443e-05,
+      "loss": 1.1411,
+      "step": 1550
+    },
+    {
+      "epoch": 1.173888363292337,
+      "grad_norm": 2.3407418727874756,
+      "learning_rate": 1.357670042689935e-05,
+      "loss": 1.1396,
+      "step": 1551
+    },
+    {
+      "epoch": 1.1746452223273416,
+      "grad_norm": 2.5837795734405518,
+      "learning_rate": 1.3569202331737292e-05,
+      "loss": 1.1858,
+      "step": 1552
+    },
+    {
+      "epoch": 1.1754020813623463,
+      "grad_norm": 2.2978811264038086,
+      "learning_rate": 1.3561701936566426e-05,
+      "loss": 1.1712,
+      "step": 1553
+    },
+    {
+      "epoch": 1.176158940397351,
+      "grad_norm": 2.5813682079315186,
+      "learning_rate": 1.355419924622139e-05,
+      "loss": 1.1282,
+      "step": 1554
+    },
+    {
+      "epoch": 1.1769157994323558,
+      "grad_norm": 2.0672824382781982,
+      "learning_rate": 1.3546694265538316e-05,
+      "loss": 1.1639,
+      "step": 1555
+    },
+    {
+      "epoch": 1.1776726584673605,
+      "grad_norm": 2.138291358947754,
+      "learning_rate": 1.3539186999354785e-05,
+      "loss": 1.1583,
+      "step": 1556
+    },
+    {
+      "epoch": 1.1784295175023651,
+      "grad_norm": 2.2805378437042236,
+      "learning_rate": 1.3531677452509873e-05,
+      "loss": 1.1315,
+      "step": 1557
+    },
+    {
+      "epoch": 1.1791863765373698,
+      "grad_norm": 2.4369373321533203,
+      "learning_rate": 1.3524165629844124e-05,
+      "loss": 1.1395,
+      "step": 1558
+    },
+    {
+      "epoch": 1.1799432355723747,
+      "grad_norm": 2.610330820083618,
+      "learning_rate": 1.3516651536199536e-05,
+      "loss": 1.1534,
+      "step": 1559
+    },
+    {
+      "epoch": 1.1807000946073793,
+      "grad_norm": 2.1532680988311768,
+      "learning_rate": 1.3509135176419583e-05,
+      "loss": 1.1266,
+      "step": 1560
+    },
+    {
+      "epoch": 1.181456953642384,
+      "grad_norm": 2.269569158554077,
+      "learning_rate": 1.3501616555349195e-05,
+      "loss": 1.1962,
+      "step": 1561
+    },
+    {
+      "epoch": 1.182213812677389,
+      "grad_norm": 2.3179900646209717,
+      "learning_rate": 1.3494095677834762e-05,
+      "loss": 1.1554,
+      "step": 1562
+    },
+    {
+      "epoch": 1.1829706717123936,
+      "grad_norm": 2.116596221923828,
+      "learning_rate": 1.3486572548724126e-05,
+      "loss": 1.124,
+      "step": 1563
+    },
+    {
+      "epoch": 1.1837275307473982,
+      "grad_norm": 2.4712612628936768,
+      "learning_rate": 1.347904717286658e-05,
+      "loss": 1.1336,
+      "step": 1564
+    },
+    {
+      "epoch": 1.1844843897824031,
+      "grad_norm": 2.1904261112213135,
+      "learning_rate": 1.3471519555112866e-05,
+      "loss": 1.1613,
+      "step": 1565
+    },
+    {
+      "epoch": 1.1852412488174078,
+      "grad_norm": 2.235826253890991,
+      "learning_rate": 1.3463989700315179e-05,
+      "loss": 1.1404,
+      "step": 1566
+    },
+    {
+      "epoch": 1.1859981078524124,
+      "grad_norm": 2.2662901878356934,
+      "learning_rate": 1.3456457613327136e-05,
+      "loss": 1.1985,
+      "step": 1567
+    },
+    {
+      "epoch": 1.1867549668874173,
+      "grad_norm": 2.3666486740112305,
+      "learning_rate": 1.3448923299003815e-05,
+      "loss": 1.125,
+      "step": 1568
+    },
+    {
+      "epoch": 1.187511825922422,
+      "grad_norm": 2.1387600898742676,
+      "learning_rate": 1.344138676220172e-05,
+      "loss": 1.157,
+      "step": 1569
+    },
+    {
+      "epoch": 1.1882686849574267,
+      "grad_norm": 2.4021949768066406,
+      "learning_rate": 1.3433848007778783e-05,
+      "loss": 1.1628,
+      "step": 1570
+    },
+    {
+      "epoch": 1.1890255439924313,
+      "grad_norm": 2.100867986679077,
+      "learning_rate": 1.3426307040594372e-05,
+      "loss": 1.1712,
+      "step": 1571
+    },
+    {
+      "epoch": 1.1897824030274362,
+      "grad_norm": 2.6818182468414307,
+      "learning_rate": 1.3418763865509283e-05,
+      "loss": 1.1505,
+      "step": 1572
+    },
+    {
+      "epoch": 1.1905392620624409,
+      "grad_norm": 2.2335827350616455,
+      "learning_rate": 1.3411218487385725e-05,
+      "loss": 1.1367,
+      "step": 1573
+    },
+    {
+      "epoch": 1.1912961210974455,
+      "grad_norm": 2.332047939300537,
+      "learning_rate": 1.3403670911087339e-05,
+      "loss": 1.1186,
+      "step": 1574
+    },
+    {
+      "epoch": 1.1920529801324504,
+      "grad_norm": 2.119150400161743,
+      "learning_rate": 1.339612114147917e-05,
+      "loss": 1.1748,
+      "step": 1575
+    },
+    {
+      "epoch": 1.192809839167455,
+      "grad_norm": 2.413939952850342,
+      "learning_rate": 1.3388569183427695e-05,
+      "loss": 1.178,
+      "step": 1576
+    },
+    {
+      "epoch": 1.1935666982024598,
+      "grad_norm": 2.265653371810913,
+      "learning_rate": 1.3381015041800787e-05,
+      "loss": 1.1532,
+      "step": 1577
+    },
+    {
+      "epoch": 1.1943235572374644,
+      "grad_norm": 2.1941394805908203,
+      "learning_rate": 1.3373458721467724e-05,
+      "loss": 1.1027,
+      "step": 1578
+    },
+    {
+      "epoch": 1.1950804162724693,
+      "grad_norm": 2.350780725479126,
+      "learning_rate": 1.3365900227299205e-05,
+      "loss": 1.1373,
+      "step": 1579
+    },
+    {
+      "epoch": 1.195837275307474,
+      "grad_norm": 2.401061773300171,
+      "learning_rate": 1.3358339564167313e-05,
+      "loss": 1.1602,
+      "step": 1580
+    },
+    {
+      "epoch": 1.1965941343424786,
+      "grad_norm": 2.3053834438323975,
+      "learning_rate": 1.3350776736945539e-05,
+      "loss": 1.0973,
+      "step": 1581
+    },
+    {
+      "epoch": 1.1973509933774835,
+      "grad_norm": 2.3348872661590576,
+      "learning_rate": 1.3343211750508769e-05,
+      "loss": 1.1439,
+      "step": 1582
+    },
+    {
+      "epoch": 1.1981078524124882,
+      "grad_norm": 2.255254030227661,
+      "learning_rate": 1.333564460973327e-05,
+      "loss": 1.1259,
+      "step": 1583
+    },
+    {
+      "epoch": 1.1988647114474928,
+      "grad_norm": 2.1117663383483887,
+      "learning_rate": 1.332807531949671e-05,
+      "loss": 1.1075,
+      "step": 1584
+    },
+    {
+      "epoch": 1.1996215704824977,
+      "grad_norm": 2.2909741401672363,
+      "learning_rate": 1.3320503884678141e-05,
+      "loss": 1.1518,
+      "step": 1585
+    },
+    {
+      "epoch": 1.2003784295175024,
+      "grad_norm": 2.2066426277160645,
+      "learning_rate": 1.331293031015799e-05,
+      "loss": 1.1617,
+      "step": 1586
+    },
+    {
+      "epoch": 1.201135288552507,
+      "grad_norm": 2.2523305416107178,
+      "learning_rate": 1.3305354600818068e-05,
+      "loss": 1.1961,
+      "step": 1587
+    },
+    {
+      "epoch": 1.201892147587512,
+      "grad_norm": 2.1978890895843506,
+      "learning_rate": 1.3297776761541566e-05,
+      "loss": 1.154,
+      "step": 1588
+    },
+    {
+      "epoch": 1.2026490066225166,
+      "grad_norm": 2.333961009979248,
+      "learning_rate": 1.3290196797213037e-05,
+      "loss": 1.1201,
+      "step": 1589
+    },
+    {
+      "epoch": 1.2034058656575213,
+      "grad_norm": 2.1890499591827393,
+      "learning_rate": 1.3282614712718412e-05,
+      "loss": 1.1166,
+      "step": 1590
+    },
+    {
+      "epoch": 1.204162724692526,
+      "grad_norm": 2.2715249061584473,
+      "learning_rate": 1.3275030512944995e-05,
+      "loss": 1.1702,
+      "step": 1591
+    },
+    {
+      "epoch": 1.2049195837275308,
+      "grad_norm": 2.378854513168335,
+      "learning_rate": 1.3267444202781434e-05,
+      "loss": 1.1674,
+      "step": 1592
+    },
+    {
+      "epoch": 1.2056764427625355,
+      "grad_norm": 2.085010290145874,
+      "learning_rate": 1.3259855787117758e-05,
+      "loss": 1.1709,
+      "step": 1593
+    },
+    {
+      "epoch": 1.2064333017975402,
+      "grad_norm": 2.3028149604797363,
+      "learning_rate": 1.3252265270845339e-05,
+      "loss": 1.1304,
+      "step": 1594
+    },
+    {
+      "epoch": 1.2071901608325448,
+      "grad_norm": 2.0950684547424316,
+      "learning_rate": 1.3244672658856908e-05,
+      "loss": 1.1585,
+      "step": 1595
+    },
+    {
+      "epoch": 1.2079470198675497,
+      "grad_norm": 2.2300803661346436,
+      "learning_rate": 1.3237077956046551e-05,
+      "loss": 1.1123,
+      "step": 1596
+    },
+    {
+      "epoch": 1.2087038789025544,
+      "grad_norm": 2.1364376544952393,
+      "learning_rate": 1.3229481167309692e-05,
+      "loss": 1.1112,
+      "step": 1597
+    },
+    {
+      "epoch": 1.209460737937559,
+      "grad_norm": 2.9876246452331543,
+      "learning_rate": 1.322188229754311e-05,
+      "loss": 1.1989,
+      "step": 1598
+    },
+    {
+      "epoch": 1.210217596972564,
+      "grad_norm": 2.1434969902038574,
+      "learning_rate": 1.3214281351644918e-05,
+      "loss": 1.1665,
+      "step": 1599
+    },
+    {
+      "epoch": 1.2109744560075686,
+      "grad_norm": 2.142533779144287,
+      "learning_rate": 1.3206678334514571e-05,
+      "loss": 1.1229,
+      "step": 1600
+    },
+    {
+      "epoch": 1.2117313150425733,
+      "grad_norm": 2.065274715423584,
+      "learning_rate": 1.3199073251052854e-05,
+      "loss": 1.1167,
+      "step": 1601
+    },
+    {
+      "epoch": 1.2124881740775781,
+      "grad_norm": 2.128526449203491,
+      "learning_rate": 1.3191466106161893e-05,
+      "loss": 1.1622,
+      "step": 1602
+    },
+    {
+      "epoch": 1.2132450331125828,
+      "grad_norm": 2.075362205505371,
+      "learning_rate": 1.3183856904745135e-05,
+      "loss": 1.1541,
+      "step": 1603
+    },
+    {
+      "epoch": 1.2140018921475875,
+      "grad_norm": 2.4913156032562256,
+      "learning_rate": 1.3176245651707357e-05,
+      "loss": 1.1635,
+      "step": 1604
+    },
+    {
+      "epoch": 1.2147587511825924,
+      "grad_norm": 2.1509463787078857,
+      "learning_rate": 1.3168632351954653e-05,
+      "loss": 1.1317,
+      "step": 1605
+    },
+    {
+      "epoch": 1.215515610217597,
+      "grad_norm": 2.2484796047210693,
+      "learning_rate": 1.3161017010394444e-05,
+      "loss": 1.1342,
+      "step": 1606
+    },
+    {
+      "epoch": 1.2162724692526017,
+      "grad_norm": 2.2622358798980713,
+      "learning_rate": 1.3153399631935463e-05,
+      "loss": 1.1416,
+      "step": 1607
+    },
+    {
+      "epoch": 1.2170293282876063,
+      "grad_norm": 2.4243550300598145,
+      "learning_rate": 1.3145780221487754e-05,
+      "loss": 1.1653,
+      "step": 1608
+    },
+    {
+      "epoch": 1.2177861873226112,
+      "grad_norm": 2.211627960205078,
+      "learning_rate": 1.3138158783962668e-05,
+      "loss": 1.177,
+      "step": 1609
+    },
+    {
+      "epoch": 1.218543046357616,
+      "grad_norm": 2.025865316390991,
+      "learning_rate": 1.3130535324272884e-05,
+      "loss": 1.1536,
+      "step": 1610
+    },
+    {
+      "epoch": 1.2192999053926206,
+      "grad_norm": 2.297100782394409,
+      "learning_rate": 1.3122909847332349e-05,
+      "loss": 1.2091,
+      "step": 1611
+    },
+    {
+      "epoch": 1.2200567644276255,
+      "grad_norm": 2.41648006439209,
+      "learning_rate": 1.3115282358056333e-05,
+      "loss": 1.183,
+      "step": 1612
+    },
+    {
+      "epoch": 1.2208136234626301,
+      "grad_norm": 2.1309831142425537,
+      "learning_rate": 1.3107652861361408e-05,
+      "loss": 1.1715,
+      "step": 1613
+    },
+    {
+      "epoch": 1.2215704824976348,
+      "grad_norm": 2.268522262573242,
+      "learning_rate": 1.3100021362165426e-05,
+      "loss": 1.1762,
+      "step": 1614
+    },
+    {
+      "epoch": 1.2223273415326394,
+      "grad_norm": 2.296552896499634,
+      "learning_rate": 1.3092387865387533e-05,
+      "loss": 1.104,
+      "step": 1615
+    },
+    {
+      "epoch": 1.2230842005676443,
+      "grad_norm": 2.1397440433502197,
+      "learning_rate": 1.3084752375948166e-05,
+      "loss": 1.1284,
+      "step": 1616
+    },
+    {
+      "epoch": 1.223841059602649,
+      "grad_norm": 2.097498893737793,
+      "learning_rate": 1.3077114898769048e-05,
+      "loss": 1.124,
+      "step": 1617
+    },
+    {
+      "epoch": 1.2245979186376537,
+      "grad_norm": 2.212064027786255,
+      "learning_rate": 1.3069475438773178e-05,
+      "loss": 1.1184,
+      "step": 1618
+    },
+    {
+      "epoch": 1.2253547776726585,
+      "grad_norm": 2.323784351348877,
+      "learning_rate": 1.3061834000884831e-05,
+      "loss": 1.1615,
+      "step": 1619
+    },
+    {
+      "epoch": 1.2261116367076632,
+      "grad_norm": 2.1432077884674072,
+      "learning_rate": 1.3054190590029572e-05,
+      "loss": 1.1564,
+      "step": 1620
+    },
+    {
+      "epoch": 1.2268684957426679,
+      "grad_norm": 2.3040294647216797,
+      "learning_rate": 1.3046545211134218e-05,
+      "loss": 1.1227,
+      "step": 1621
+    },
+    {
+      "epoch": 1.2276253547776728,
+      "grad_norm": 2.406848669052124,
+      "learning_rate": 1.3038897869126865e-05,
+      "loss": 1.1577,
+      "step": 1622
+    },
+    {
+      "epoch": 1.2283822138126774,
+      "grad_norm": 2.3050808906555176,
+      "learning_rate": 1.3031248568936877e-05,
+      "loss": 1.1924,
+      "step": 1623
+    },
+    {
+      "epoch": 1.229139072847682,
+      "grad_norm": 2.03425669670105,
+      "learning_rate": 1.3023597315494874e-05,
+      "loss": 1.1474,
+      "step": 1624
+    },
+    {
+      "epoch": 1.2298959318826868,
+      "grad_norm": 2.2921745777130127,
+      "learning_rate": 1.3015944113732734e-05,
+      "loss": 1.1815,
+      "step": 1625
+    },
+    {
+      "epoch": 1.2306527909176916,
+      "grad_norm": 2.248823881149292,
+      "learning_rate": 1.3008288968583603e-05,
+      "loss": 1.1482,
+      "step": 1626
+    },
+    {
+      "epoch": 1.2314096499526963,
+      "grad_norm": 2.3645107746124268,
+      "learning_rate": 1.3000631884981858e-05,
+      "loss": 1.1383,
+      "step": 1627
+    },
+    {
+      "epoch": 1.232166508987701,
+      "grad_norm": 2.0877134799957275,
+      "learning_rate": 1.2992972867863147e-05,
+      "loss": 1.2064,
+      "step": 1628
+    },
+    {
+      "epoch": 1.2329233680227059,
+      "grad_norm": 2.3611538410186768,
+      "learning_rate": 1.2985311922164359e-05,
+      "loss": 1.1582,
+      "step": 1629
+    },
+    {
+      "epoch": 1.2336802270577105,
+      "grad_norm": 2.087958335876465,
+      "learning_rate": 1.2977649052823616e-05,
+      "loss": 1.1553,
+      "step": 1630
+    },
+    {
+      "epoch": 1.2344370860927152,
+      "grad_norm": 2.2635905742645264,
+      "learning_rate": 1.2969984264780283e-05,
+      "loss": 1.1704,
+      "step": 1631
+    },
+    {
+      "epoch": 1.2351939451277199,
+      "grad_norm": 2.251215934753418,
+      "learning_rate": 1.2962317562974976e-05,
+      "loss": 1.1855,
+      "step": 1632
+    },
+    {
+      "epoch": 1.2359508041627247,
+      "grad_norm": 3.073653221130371,
+      "learning_rate": 1.2954648952349527e-05,
+      "loss": 1.1935,
+      "step": 1633
+    },
+    {
+      "epoch": 1.2367076631977294,
+      "grad_norm": 2.2269108295440674,
+      "learning_rate": 1.2946978437847004e-05,
+      "loss": 1.146,
+      "step": 1634
+    },
+    {
+      "epoch": 1.237464522232734,
+      "grad_norm": 2.4930577278137207,
+      "learning_rate": 1.2939306024411713e-05,
+      "loss": 1.1703,
+      "step": 1635
+    },
+    {
+      "epoch": 1.238221381267739,
+      "grad_norm": 2.6076838970184326,
+      "learning_rate": 1.2931631716989166e-05,
+      "loss": 1.1756,
+      "step": 1636
+    },
+    {
+      "epoch": 1.2389782403027436,
+      "grad_norm": 2.250298023223877,
+      "learning_rate": 1.292395552052611e-05,
+      "loss": 1.1607,
+      "step": 1637
+    },
+    {
+      "epoch": 1.2397350993377483,
+      "grad_norm": 2.2543821334838867,
+      "learning_rate": 1.29162774399705e-05,
+      "loss": 1.1653,
+      "step": 1638
+    },
+    {
+      "epoch": 1.2404919583727532,
+      "grad_norm": 2.1660284996032715,
+      "learning_rate": 1.290859748027151e-05,
+      "loss": 1.0827,
+      "step": 1639
+    },
+    {
+      "epoch": 1.2412488174077578,
+      "grad_norm": 2.392023801803589,
+      "learning_rate": 1.2900915646379524e-05,
+      "loss": 1.1651,
+      "step": 1640
+    },
+    {
+      "epoch": 1.2420056764427625,
+      "grad_norm": 2.147473096847534,
+      "learning_rate": 1.2893231943246143e-05,
+      "loss": 1.1012,
+      "step": 1641
+    },
+    {
+      "epoch": 1.2427625354777674,
+      "grad_norm": 2.2261102199554443,
+      "learning_rate": 1.2885546375824154e-05,
+      "loss": 1.1313,
+      "step": 1642
+    },
+    {
+      "epoch": 1.243519394512772,
+      "grad_norm": 2.1518611907958984,
+      "learning_rate": 1.2877858949067564e-05,
+      "loss": 1.1309,
+      "step": 1643
+    },
+    {
+      "epoch": 1.2442762535477767,
+      "grad_norm": 2.2095835208892822,
+      "learning_rate": 1.2870169667931567e-05,
+      "loss": 1.109,
+      "step": 1644
+    },
+    {
+      "epoch": 1.2450331125827814,
+      "grad_norm": 2.3556344509124756,
+      "learning_rate": 1.2862478537372556e-05,
+      "loss": 1.0937,
+      "step": 1645
+    },
+    {
+      "epoch": 1.2457899716177863,
+      "grad_norm": 2.178274154663086,
+      "learning_rate": 1.2854785562348121e-05,
+      "loss": 1.1857,
+      "step": 1646
+    },
+    {
+      "epoch": 1.246546830652791,
+      "grad_norm": 2.1965596675872803,
+      "learning_rate": 1.2847090747817033e-05,
+      "loss": 1.1292,
+      "step": 1647
+    },
+    {
+      "epoch": 1.2473036896877956,
+      "grad_norm": 2.359292984008789,
+      "learning_rate": 1.2839394098739257e-05,
+      "loss": 1.1753,
+      "step": 1648
+    },
+    {
+      "epoch": 1.2480605487228003,
+      "grad_norm": 2.189749002456665,
+      "learning_rate": 1.2831695620075926e-05,
+      "loss": 1.0852,
+      "step": 1649
+    },
+    {
+      "epoch": 1.2488174077578051,
+      "grad_norm": 2.1658499240875244,
+      "learning_rate": 1.2823995316789366e-05,
+      "loss": 1.1278,
+      "step": 1650
+    },
+    {
+      "epoch": 1.2495742667928098,
+      "grad_norm": 2.249729871749878,
+      "learning_rate": 1.281629319384308e-05,
+      "loss": 1.1829,
+      "step": 1651
+    },
+    {
+      "epoch": 1.2503311258278145,
+      "grad_norm": 2.2419471740722656,
+      "learning_rate": 1.2808589256201735e-05,
+      "loss": 1.1373,
+      "step": 1652
+    },
+    {
+      "epoch": 1.2510879848628194,
+      "grad_norm": 2.296644449234009,
+      "learning_rate": 1.280088350883117e-05,
+      "loss": 1.1724,
+      "step": 1653
+    },
+    {
+      "epoch": 1.251844843897824,
+      "grad_norm": 2.77717661857605,
+      "learning_rate": 1.2793175956698398e-05,
+      "loss": 1.1367,
+      "step": 1654
+    },
+    {
+      "epoch": 1.2526017029328287,
+      "grad_norm": 2.4535298347473145,
+      "learning_rate": 1.2785466604771584e-05,
+      "loss": 1.1874,
+      "step": 1655
+    },
+    {
+      "epoch": 1.2533585619678336,
+      "grad_norm": 2.1663715839385986,
+      "learning_rate": 1.2777755458020058e-05,
+      "loss": 1.1819,
+      "step": 1656
+    },
+    {
+      "epoch": 1.2541154210028382,
+      "grad_norm": 2.467954158782959,
+      "learning_rate": 1.2770042521414314e-05,
+      "loss": 1.1761,
+      "step": 1657
+    },
+    {
+      "epoch": 1.254872280037843,
+      "grad_norm": 2.3556721210479736,
+      "learning_rate": 1.2762327799925991e-05,
+      "loss": 1.1574,
+      "step": 1658
+    },
+    {
+      "epoch": 1.2556291390728478,
+      "grad_norm": 2.204136371612549,
+      "learning_rate": 1.2754611298527875e-05,
+      "loss": 1.0962,
+      "step": 1659
+    },
+    {
+      "epoch": 1.2563859981078525,
+      "grad_norm": 2.3233225345611572,
+      "learning_rate": 1.274689302219391e-05,
+      "loss": 1.1369,
+      "step": 1660
+    },
+    {
+      "epoch": 1.2571428571428571,
+      "grad_norm": 2.5201222896575928,
+      "learning_rate": 1.2739172975899181e-05,
+      "loss": 1.1593,
+      "step": 1661
+    },
+    {
+      "epoch": 1.257899716177862,
+      "grad_norm": 2.531087875366211,
+      "learning_rate": 1.273145116461991e-05,
+      "loss": 1.1411,
+      "step": 1662
+    },
+    {
+      "epoch": 1.2586565752128667,
+      "grad_norm": 2.510352373123169,
+      "learning_rate": 1.2723727593333454e-05,
+      "loss": 1.1426,
+      "step": 1663
+    },
+    {
+      "epoch": 1.2594134342478713,
+      "grad_norm": 2.217392921447754,
+      "learning_rate": 1.2716002267018314e-05,
+      "loss": 1.0712,
+      "step": 1664
+    },
+    {
+      "epoch": 1.260170293282876,
+      "grad_norm": 2.3988654613494873,
+      "learning_rate": 1.2708275190654126e-05,
+      "loss": 1.1884,
+      "step": 1665
+    },
+    {
+      "epoch": 1.2609271523178807,
+      "grad_norm": 2.151139259338379,
+      "learning_rate": 1.2700546369221628e-05,
+      "loss": 1.095,
+      "step": 1666
+    },
+    {
+      "epoch": 1.2616840113528855,
+      "grad_norm": 2.3287789821624756,
+      "learning_rate": 1.2692815807702711e-05,
+      "loss": 1.1726,
+      "step": 1667
+    },
+    {
+      "epoch": 1.2624408703878902,
+      "grad_norm": 2.1874783039093018,
+      "learning_rate": 1.268508351108038e-05,
+      "loss": 1.1389,
+      "step": 1668
+    },
+    {
+      "epoch": 1.2631977294228949,
+      "grad_norm": 2.501871347427368,
+      "learning_rate": 1.2677349484338747e-05,
+      "loss": 1.1717,
+      "step": 1669
+    },
+    {
+      "epoch": 1.2639545884578998,
+      "grad_norm": 2.2890784740448,
+      "learning_rate": 1.2669613732463053e-05,
+      "loss": 1.1069,
+      "step": 1670
+    },
+    {
+      "epoch": 1.2647114474929044,
+      "grad_norm": 2.41701340675354,
+      "learning_rate": 1.2661876260439642e-05,
+      "loss": 1.142,
+      "step": 1671
+    },
+    {
+      "epoch": 1.265468306527909,
+      "grad_norm": 2.5459794998168945,
+      "learning_rate": 1.2654137073255976e-05,
+      "loss": 1.1071,
+      "step": 1672
+    },
+    {
+      "epoch": 1.266225165562914,
+      "grad_norm": 2.1220319271087646,
+      "learning_rate": 1.2646396175900612e-05,
+      "loss": 1.1644,
+      "step": 1673
+    },
+    {
+      "epoch": 1.2669820245979186,
+      "grad_norm": 2.383187770843506,
+      "learning_rate": 1.2638653573363215e-05,
+      "loss": 1.1515,
+      "step": 1674
+    },
+    {
+      "epoch": 1.2677388836329233,
+      "grad_norm": 2.090808868408203,
+      "learning_rate": 1.2630909270634554e-05,
+      "loss": 1.1151,
+      "step": 1675
+    },
+    {
+      "epoch": 1.2684957426679282,
+      "grad_norm": 2.306619882583618,
+      "learning_rate": 1.2623163272706483e-05,
+      "loss": 1.177,
+      "step": 1676
+    },
+    {
+      "epoch": 1.2692526017029329,
+      "grad_norm": 2.4056408405303955,
+      "learning_rate": 1.261541558457195e-05,
+      "loss": 1.1811,
+      "step": 1677
+    },
+    {
+      "epoch": 1.2700094607379375,
+      "grad_norm": 2.4404773712158203,
+      "learning_rate": 1.2607666211225002e-05,
+      "loss": 1.1686,
+      "step": 1678
+    },
+    {
+      "epoch": 1.2707663197729424,
+      "grad_norm": 2.1623356342315674,
+      "learning_rate": 1.2599915157660776e-05,
+      "loss": 1.1472,
+      "step": 1679
+    },
+    {
+      "epoch": 1.271523178807947,
+      "grad_norm": 2.5244863033294678,
+      "learning_rate": 1.2592162428875465e-05,
+      "loss": 1.1093,
+      "step": 1680
+    },
+    {
+      "epoch": 1.2722800378429517,
+      "grad_norm": 2.649132251739502,
+      "learning_rate": 1.2584408029866373e-05,
+      "loss": 1.1644,
+      "step": 1681
+    },
+    {
+      "epoch": 1.2730368968779564,
+      "grad_norm": 2.245384931564331,
+      "learning_rate": 1.2576651965631862e-05,
+      "loss": 1.2137,
+      "step": 1682
+    },
+    {
+      "epoch": 1.2737937559129613,
+      "grad_norm": 3.0994908809661865,
+      "learning_rate": 1.256889424117137e-05,
+      "loss": 1.1189,
+      "step": 1683
+    },
+    {
+      "epoch": 1.274550614947966,
+      "grad_norm": 2.228210210800171,
+      "learning_rate": 1.2561134861485413e-05,
+      "loss": 1.1694,
+      "step": 1684
+    },
+    {
+      "epoch": 1.2753074739829706,
+      "grad_norm": 2.0974786281585693,
+      "learning_rate": 1.2553373831575572e-05,
+      "loss": 1.1661,
+      "step": 1685
+    },
+    {
+      "epoch": 1.2760643330179753,
+      "grad_norm": 2.1458041667938232,
+      "learning_rate": 1.2545611156444477e-05,
+      "loss": 1.0814,
+      "step": 1686
+    },
+    {
+      "epoch": 1.2768211920529802,
+      "grad_norm": 2.163115978240967,
+      "learning_rate": 1.253784684109584e-05,
+      "loss": 1.1643,
+      "step": 1687
+    },
+    {
+      "epoch": 1.2775780510879848,
+      "grad_norm": 2.122997999191284,
+      "learning_rate": 1.2530080890534416e-05,
+      "loss": 1.172,
+      "step": 1688
+    },
+    {
+      "epoch": 1.2783349101229895,
+      "grad_norm": 2.4748451709747314,
+      "learning_rate": 1.2522313309766021e-05,
+      "loss": 1.1489,
+      "step": 1689
+    },
+    {
+      "epoch": 1.2790917691579944,
+      "grad_norm": 2.201387882232666,
+      "learning_rate": 1.2514544103797517e-05,
+      "loss": 1.1509,
+      "step": 1690
+    },
+    {
+      "epoch": 1.279848628192999,
+      "grad_norm": 2.158069610595703,
+      "learning_rate": 1.2506773277636812e-05,
+      "loss": 1.1284,
+      "step": 1691
+    },
+    {
+      "epoch": 1.2806054872280037,
+      "grad_norm": 2.192920684814453,
+      "learning_rate": 1.2499000836292875e-05,
+      "loss": 1.156,
+      "step": 1692
+    },
+    {
+      "epoch": 1.2813623462630086,
+      "grad_norm": 2.266641855239868,
+      "learning_rate": 1.2491226784775685e-05,
+      "loss": 1.1298,
+      "step": 1693
+    },
+    {
+      "epoch": 1.2821192052980133,
+      "grad_norm": 2.677654981613159,
+      "learning_rate": 1.2483451128096289e-05,
+      "loss": 1.1472,
+      "step": 1694
+    },
+    {
+      "epoch": 1.282876064333018,
+      "grad_norm": 2.4137139320373535,
+      "learning_rate": 1.2475673871266756e-05,
+      "loss": 1.075,
+      "step": 1695
+    },
+    {
+      "epoch": 1.2836329233680228,
+      "grad_norm": 2.201813220977783,
+      "learning_rate": 1.2467895019300187e-05,
+      "loss": 1.1224,
+      "step": 1696
+    },
+    {
+      "epoch": 1.2843897824030275,
+      "grad_norm": 2.1659185886383057,
+      "learning_rate": 1.2460114577210703e-05,
+      "loss": 1.1606,
+      "step": 1697
+    },
+    {
+      "epoch": 1.2851466414380321,
+      "grad_norm": 2.3215322494506836,
+      "learning_rate": 1.245233255001347e-05,
+      "loss": 1.1408,
+      "step": 1698
+    },
+    {
+      "epoch": 1.285903500473037,
+      "grad_norm": 2.530764579772949,
+      "learning_rate": 1.2444548942724657e-05,
+      "loss": 1.1433,
+      "step": 1699
+    },
+    {
+      "epoch": 1.2866603595080417,
+      "grad_norm": 2.5110771656036377,
+      "learning_rate": 1.2436763760361461e-05,
+      "loss": 1.1644,
+      "step": 1700
+    },
+    {
+      "epoch": 1.2874172185430464,
+      "grad_norm": 2.0336203575134277,
+      "learning_rate": 1.2428977007942092e-05,
+      "loss": 1.1538,
+      "step": 1701
+    },
+    {
+      "epoch": 1.288174077578051,
+      "grad_norm": 2.3316948413848877,
+      "learning_rate": 1.2421188690485767e-05,
+      "loss": 1.1349,
+      "step": 1702
+    },
+    {
+      "epoch": 1.2889309366130557,
+      "grad_norm": 2.151745319366455,
+      "learning_rate": 1.241339881301273e-05,
+      "loss": 1.1464,
+      "step": 1703
+    },
+    {
+      "epoch": 1.2896877956480606,
+      "grad_norm": 2.1439285278320312,
+      "learning_rate": 1.2405607380544198e-05,
+      "loss": 1.1384,
+      "step": 1704
+    },
+    {
+      "epoch": 1.2904446546830652,
+      "grad_norm": 2.0676236152648926,
+      "learning_rate": 1.239781439810242e-05,
+      "loss": 1.1315,
+      "step": 1705
+    },
+    {
+      "epoch": 1.29120151371807,
+      "grad_norm": 2.1534860134124756,
+      "learning_rate": 1.239001987071064e-05,
+      "loss": 1.1232,
+      "step": 1706
+    },
+    {
+      "epoch": 1.2919583727530748,
+      "grad_norm": 2.4337046146392822,
+      "learning_rate": 1.238222380339308e-05,
+      "loss": 1.1637,
+      "step": 1707
+    },
+    {
+      "epoch": 1.2927152317880795,
+      "grad_norm": 3.3103768825531006,
+      "learning_rate": 1.2374426201174974e-05,
+      "loss": 1.1255,
+      "step": 1708
+    },
+    {
+      "epoch": 1.2934720908230841,
+      "grad_norm": 2.3964853286743164,
+      "learning_rate": 1.2366627069082533e-05,
+      "loss": 1.1474,
+      "step": 1709
+    },
+    {
+      "epoch": 1.294228949858089,
+      "grad_norm": 2.196171760559082,
+      "learning_rate": 1.235882641214296e-05,
+      "loss": 1.1152,
+      "step": 1710
+    },
+    {
+      "epoch": 1.2949858088930937,
+      "grad_norm": 2.2231311798095703,
+      "learning_rate": 1.2351024235384435e-05,
+      "loss": 1.0872,
+      "step": 1711
+    },
+    {
+      "epoch": 1.2957426679280983,
+      "grad_norm": 3.2890310287475586,
+      "learning_rate": 1.2343220543836132e-05,
+      "loss": 1.1376,
+      "step": 1712
+    },
+    {
+      "epoch": 1.2964995269631032,
+      "grad_norm": 2.37038516998291,
+      "learning_rate": 1.2335415342528186e-05,
+      "loss": 1.1309,
+      "step": 1713
+    },
+    {
+      "epoch": 1.2972563859981079,
+      "grad_norm": 2.0955164432525635,
+      "learning_rate": 1.2327608636491706e-05,
+      "loss": 1.1721,
+      "step": 1714
+    },
+    {
+      "epoch": 1.2980132450331126,
+      "grad_norm": 2.3492562770843506,
+      "learning_rate": 1.2319800430758787e-05,
+      "loss": 1.1855,
+      "step": 1715
+    },
+    {
+      "epoch": 1.2987701040681174,
+      "grad_norm": 2.366028308868408,
+      "learning_rate": 1.231199073036247e-05,
+      "loss": 1.1547,
+      "step": 1716
+    },
+    {
+      "epoch": 1.299526963103122,
+      "grad_norm": 2.163280725479126,
+      "learning_rate": 1.230417954033677e-05,
+      "loss": 1.1289,
+      "step": 1717
+    },
+    {
+      "epoch": 1.3002838221381268,
+      "grad_norm": 2.1231632232666016,
+      "learning_rate": 1.2296366865716663e-05,
+      "loss": 1.1386,
+      "step": 1718
+    },
+    {
+      "epoch": 1.3010406811731314,
+      "grad_norm": 2.1293210983276367,
+      "learning_rate": 1.2288552711538076e-05,
+      "loss": 1.185,
+      "step": 1719
+    },
+    {
+      "epoch": 1.3017975402081363,
+      "grad_norm": 2.0795953273773193,
+      "learning_rate": 1.2280737082837903e-05,
+      "loss": 1.0935,
+      "step": 1720
+    },
+    {
+      "epoch": 1.302554399243141,
+      "grad_norm": 2.1011762619018555,
+      "learning_rate": 1.2272919984653972e-05,
+      "loss": 1.1672,
+      "step": 1721
+    },
+    {
+      "epoch": 1.3033112582781456,
+      "grad_norm": 2.221156120300293,
+      "learning_rate": 1.2265101422025064e-05,
+      "loss": 1.1073,
+      "step": 1722
+    },
+    {
+      "epoch": 1.3040681173131503,
+      "grad_norm": 2.2249984741210938,
+      "learning_rate": 1.2257281399990913e-05,
+      "loss": 1.1659,
+      "step": 1723
+    },
+    {
+      "epoch": 1.3048249763481552,
+      "grad_norm": 4.573660850524902,
+      "learning_rate": 1.2249459923592176e-05,
+      "loss": 1.1835,
+      "step": 1724
+    },
+    {
+      "epoch": 1.3055818353831599,
+      "grad_norm": 2.1640846729278564,
+      "learning_rate": 1.2241636997870459e-05,
+      "loss": 1.1342,
+      "step": 1725
+    },
+    {
+      "epoch": 1.3063386944181645,
+      "grad_norm": 2.010333299636841,
+      "learning_rate": 1.223381262786831e-05,
+      "loss": 1.1,
+      "step": 1726
+    },
+    {
+      "epoch": 1.3070955534531694,
+      "grad_norm": 2.026108980178833,
+      "learning_rate": 1.2225986818629188e-05,
+      "loss": 1.1424,
+      "step": 1727
+    },
+    {
+      "epoch": 1.307852412488174,
+      "grad_norm": 2.0564801692962646,
+      "learning_rate": 1.22181595751975e-05,
+      "loss": 1.1336,
+      "step": 1728
+    },
+    {
+      "epoch": 1.3086092715231787,
+      "grad_norm": 1.9752734899520874,
+      "learning_rate": 1.2210330902618555e-05,
+      "loss": 1.1617,
+      "step": 1729
+    },
+    {
+      "epoch": 1.3093661305581836,
+      "grad_norm": 2.1132519245147705,
+      "learning_rate": 1.2202500805938606e-05,
+      "loss": 1.1841,
+      "step": 1730
+    },
+    {
+      "epoch": 1.3101229895931883,
+      "grad_norm": 2.167475461959839,
+      "learning_rate": 1.2194669290204813e-05,
+      "loss": 1.1255,
+      "step": 1731
+    },
+    {
+      "epoch": 1.310879848628193,
+      "grad_norm": 2.1125247478485107,
+      "learning_rate": 1.2186836360465249e-05,
+      "loss": 1.1434,
+      "step": 1732
+    },
+    {
+      "epoch": 1.3116367076631978,
+      "grad_norm": 2.126776933670044,
+      "learning_rate": 1.21790020217689e-05,
+      "loss": 1.1626,
+      "step": 1733
+    },
+    {
+      "epoch": 1.3123935666982025,
+      "grad_norm": 2.1454262733459473,
+      "learning_rate": 1.2171166279165668e-05,
+      "loss": 1.116,
+      "step": 1734
+    },
+    {
+      "epoch": 1.3131504257332072,
+      "grad_norm": 2.166027545928955,
+      "learning_rate": 1.216332913770634e-05,
+      "loss": 1.145,
+      "step": 1735
+    },
+    {
+      "epoch": 1.313907284768212,
+      "grad_norm": 2.0442612171173096,
+      "learning_rate": 1.2155490602442628e-05,
+      "loss": 1.1443,
+      "step": 1736
+    },
+    {
+      "epoch": 1.3146641438032167,
+      "grad_norm": 2.2408742904663086,
+      "learning_rate": 1.2147650678427136e-05,
+      "loss": 1.1297,
+      "step": 1737
+    },
+    {
+      "epoch": 1.3154210028382214,
+      "grad_norm": 2.039287805557251,
+      "learning_rate": 1.213980937071335e-05,
+      "loss": 1.1183,
+      "step": 1738
+    },
+    {
+      "epoch": 1.316177861873226,
+      "grad_norm": 2.4958298206329346,
+      "learning_rate": 1.213196668435566e-05,
+      "loss": 1.1127,
+      "step": 1739
+    },
+    {
+      "epoch": 1.3169347209082307,
+      "grad_norm": 2.0684995651245117,
+      "learning_rate": 1.212412262440935e-05,
+      "loss": 1.1092,
+      "step": 1740
+    },
+    {
+      "epoch": 1.3176915799432356,
+      "grad_norm": 2.2518489360809326,
+      "learning_rate": 1.2116277195930566e-05,
+      "loss": 1.1256,
+      "step": 1741
+    },
+    {
+      "epoch": 1.3184484389782403,
+      "grad_norm": 2.2096285820007324,
+      "learning_rate": 1.2108430403976363e-05,
+      "loss": 1.1785,
+      "step": 1742
+    },
+    {
+      "epoch": 1.319205298013245,
+      "grad_norm": 2.1161551475524902,
+      "learning_rate": 1.2100582253604663e-05,
+      "loss": 1.1816,
+      "step": 1743
+    },
+    {
+      "epoch": 1.3199621570482498,
+      "grad_norm": 2.2261106967926025,
+      "learning_rate": 1.2092732749874258e-05,
+      "loss": 1.1512,
+      "step": 1744
+    },
+    {
+      "epoch": 1.3207190160832545,
+      "grad_norm": 2.4035398960113525,
+      "learning_rate": 1.2084881897844827e-05,
+      "loss": 1.1505,
+      "step": 1745
+    },
+    {
+      "epoch": 1.3214758751182591,
+      "grad_norm": 2.2090861797332764,
+      "learning_rate": 1.2077029702576898e-05,
+      "loss": 1.1521,
+      "step": 1746
+    },
+    {
+      "epoch": 1.322232734153264,
+      "grad_norm": 2.054429769515991,
+      "learning_rate": 1.2069176169131889e-05,
+      "loss": 1.1863,
+      "step": 1747
+    },
+    {
+      "epoch": 1.3229895931882687,
+      "grad_norm": 2.0456814765930176,
+      "learning_rate": 1.2061321302572063e-05,
+      "loss": 1.1125,
+      "step": 1748
+    },
+    {
+      "epoch": 1.3237464522232734,
+      "grad_norm": 2.066861629486084,
+      "learning_rate": 1.2053465107960536e-05,
+      "loss": 1.1107,
+      "step": 1749
+    },
+    {
+      "epoch": 1.3245033112582782,
+      "grad_norm": 2.2116339206695557,
+      "learning_rate": 1.204560759036131e-05,
+      "loss": 1.1796,
+      "step": 1750
+    },
+    {
+      "epoch": 1.325260170293283,
+      "grad_norm": 2.355694532394409,
+      "learning_rate": 1.203774875483921e-05,
+      "loss": 1.1221,
+      "step": 1751
+    },
+    {
+      "epoch": 1.3260170293282876,
+      "grad_norm": 2.3318047523498535,
+      "learning_rate": 1.202988860645992e-05,
+      "loss": 1.1482,
+      "step": 1752
+    },
+    {
+      "epoch": 1.3267738883632925,
+      "grad_norm": 2.3973910808563232,
+      "learning_rate": 1.202202715028998e-05,
+      "loss": 1.09,
+      "step": 1753
+    },
+    {
+      "epoch": 1.3275307473982971,
+      "grad_norm": 2.3162357807159424,
+      "learning_rate": 1.2014164391396761e-05,
+      "loss": 1.1362,
+      "step": 1754
+    },
+    {
+      "epoch": 1.3282876064333018,
+      "grad_norm": 2.232311964035034,
+      "learning_rate": 1.2006300334848472e-05,
+      "loss": 1.1419,
+      "step": 1755
+    },
+    {
+      "epoch": 1.3290444654683065,
+      "grad_norm": 2.3239498138427734,
+      "learning_rate": 1.1998434985714172e-05,
+      "loss": 1.151,
+      "step": 1756
+    },
+    {
+      "epoch": 1.3298013245033111,
+      "grad_norm": 2.5459787845611572,
+      "learning_rate": 1.1990568349063742e-05,
+      "loss": 1.1649,
+      "step": 1757
+    },
+    {
+      "epoch": 1.330558183538316,
+      "grad_norm": 2.184105396270752,
+      "learning_rate": 1.1982700429967893e-05,
+      "loss": 1.1334,
+      "step": 1758
+    },
+    {
+      "epoch": 1.3313150425733207,
+      "grad_norm": 2.092010498046875,
+      "learning_rate": 1.1974831233498175e-05,
+      "loss": 1.0945,
+      "step": 1759
+    },
+    {
+      "epoch": 1.3320719016083253,
+      "grad_norm": 2.204160690307617,
+      "learning_rate": 1.1966960764726937e-05,
+      "loss": 1.1411,
+      "step": 1760
+    },
+    {
+      "epoch": 1.3328287606433302,
+      "grad_norm": 2.467329978942871,
+      "learning_rate": 1.195908902872738e-05,
+      "loss": 1.1259,
+      "step": 1761
+    },
+    {
+      "epoch": 1.3335856196783349,
+      "grad_norm": 2.2322754859924316,
+      "learning_rate": 1.1951216030573489e-05,
+      "loss": 1.1204,
+      "step": 1762
+    },
+    {
+      "epoch": 1.3343424787133396,
+      "grad_norm": 2.1422557830810547,
+      "learning_rate": 1.1943341775340087e-05,
+      "loss": 1.1306,
+      "step": 1763
+    },
+    {
+      "epoch": 1.3350993377483444,
+      "grad_norm": 2.393411159515381,
+      "learning_rate": 1.1935466268102802e-05,
+      "loss": 1.1409,
+      "step": 1764
+    },
+    {
+      "epoch": 1.335856196783349,
+      "grad_norm": 2.2010276317596436,
+      "learning_rate": 1.192758951393806e-05,
+      "loss": 1.0952,
+      "step": 1765
+    },
+    {
+      "epoch": 1.3366130558183538,
+      "grad_norm": 2.128002166748047,
+      "learning_rate": 1.1919711517923095e-05,
+      "loss": 1.1084,
+      "step": 1766
+    },
+    {
+      "epoch": 1.3373699148533587,
+      "grad_norm": 2.090876340866089,
+      "learning_rate": 1.1911832285135953e-05,
+      "loss": 1.1409,
+      "step": 1767
+    },
+    {
+      "epoch": 1.3381267738883633,
+      "grad_norm": 2.232081890106201,
+      "learning_rate": 1.1903951820655458e-05,
+      "loss": 1.176,
+      "step": 1768
+    },
+    {
+      "epoch": 1.338883632923368,
+      "grad_norm": 2.2187860012054443,
+      "learning_rate": 1.1896070129561237e-05,
+      "loss": 1.1094,
+      "step": 1769
+    },
+    {
+      "epoch": 1.3396404919583729,
+      "grad_norm": 2.2788565158843994,
+      "learning_rate": 1.1888187216933715e-05,
+      "loss": 1.1302,
+      "step": 1770
+    },
+    {
+      "epoch": 1.3403973509933775,
+      "grad_norm": 2.153656482696533,
+      "learning_rate": 1.1880303087854093e-05,
+      "loss": 1.1742,
+      "step": 1771
+    },
+    {
+      "epoch": 1.3411542100283822,
+      "grad_norm": 2.2120232582092285,
+      "learning_rate": 1.187241774740436e-05,
+      "loss": 1.1553,
+      "step": 1772
+    },
+    {
+      "epoch": 1.3419110690633869,
+      "grad_norm": 2.09063720703125,
+      "learning_rate": 1.1864531200667283e-05,
+      "loss": 1.1231,
+      "step": 1773
+    },
+    {
+      "epoch": 1.3426679280983917,
+      "grad_norm": 2.186126232147217,
+      "learning_rate": 1.1856643452726417e-05,
+      "loss": 1.1121,
+      "step": 1774
+    },
+    {
+      "epoch": 1.3434247871333964,
+      "grad_norm": 2.706040620803833,
+      "learning_rate": 1.1848754508666084e-05,
+      "loss": 1.1323,
+      "step": 1775
+    },
+    {
+      "epoch": 1.344181646168401,
+      "grad_norm": 2.1138103008270264,
+      "learning_rate": 1.1840864373571368e-05,
+      "loss": 1.1255,
+      "step": 1776
+    },
+    {
+      "epoch": 1.3449385052034057,
+      "grad_norm": 2.1181037425994873,
+      "learning_rate": 1.1832973052528136e-05,
+      "loss": 1.1406,
+      "step": 1777
+    },
+    {
+      "epoch": 1.3456953642384106,
+      "grad_norm": 2.0773799419403076,
+      "learning_rate": 1.1825080550623014e-05,
+      "loss": 1.1159,
+      "step": 1778
+    },
+    {
+      "epoch": 1.3464522232734153,
+      "grad_norm": 2.3848013877868652,
+      "learning_rate": 1.1817186872943385e-05,
+      "loss": 1.1687,
+      "step": 1779
+    },
+    {
+      "epoch": 1.34720908230842,
+      "grad_norm": 2.052957534790039,
+      "learning_rate": 1.1809292024577397e-05,
+      "loss": 1.176,
+      "step": 1780
+    },
+    {
+      "epoch": 1.3479659413434248,
+      "grad_norm": 2.1066739559173584,
+      "learning_rate": 1.1801396010613947e-05,
+      "loss": 1.1563,
+      "step": 1781
+    },
+    {
+      "epoch": 1.3487228003784295,
+      "grad_norm": 2.2263689041137695,
+      "learning_rate": 1.1793498836142685e-05,
+      "loss": 1.1763,
+      "step": 1782
+    },
+    {
+      "epoch": 1.3494796594134342,
+      "grad_norm": 2.263129711151123,
+      "learning_rate": 1.178560050625401e-05,
+      "loss": 1.1737,
+      "step": 1783
+    },
+    {
+      "epoch": 1.350236518448439,
+      "grad_norm": 2.3742623329162598,
+      "learning_rate": 1.1777701026039063e-05,
+      "loss": 1.123,
+      "step": 1784
+    },
+    {
+      "epoch": 1.3509933774834437,
+      "grad_norm": 2.1886773109436035,
+      "learning_rate": 1.1769800400589733e-05,
+      "loss": 1.1635,
+      "step": 1785
+    },
+    {
+      "epoch": 1.3517502365184484,
+      "grad_norm": 2.190129041671753,
+      "learning_rate": 1.1761898634998635e-05,
+      "loss": 1.1131,
+      "step": 1786
+    },
+    {
+      "epoch": 1.3525070955534533,
+      "grad_norm": 2.258070468902588,
+      "learning_rate": 1.1753995734359131e-05,
+      "loss": 1.1794,
+      "step": 1787
+    },
+    {
+      "epoch": 1.353263954588458,
+      "grad_norm": 2.0639896392822266,
+      "learning_rate": 1.1746091703765316e-05,
+      "loss": 1.1047,
+      "step": 1788
+    },
+    {
+      "epoch": 1.3540208136234626,
+      "grad_norm": 2.4623801708221436,
+      "learning_rate": 1.1738186548311998e-05,
+      "loss": 1.1642,
+      "step": 1789
+    },
+    {
+      "epoch": 1.3547776726584675,
+      "grad_norm": 2.082294225692749,
+      "learning_rate": 1.1730280273094724e-05,
+      "loss": 1.1374,
+      "step": 1790
+    },
+    {
+      "epoch": 1.3555345316934722,
+      "grad_norm": 1.9872076511383057,
+      "learning_rate": 1.1722372883209758e-05,
+      "loss": 1.1305,
+      "step": 1791
+    },
+    {
+      "epoch": 1.3562913907284768,
+      "grad_norm": 2.2372727394104004,
+      "learning_rate": 1.1714464383754085e-05,
+      "loss": 1.1261,
+      "step": 1792
+    },
+    {
+      "epoch": 1.3570482497634815,
+      "grad_norm": 2.16003680229187,
+      "learning_rate": 1.1706554779825399e-05,
+      "loss": 1.1289,
+      "step": 1793
+    },
+    {
+      "epoch": 1.3578051087984861,
+      "grad_norm": 2.2329182624816895,
+      "learning_rate": 1.1698644076522112e-05,
+      "loss": 1.1331,
+      "step": 1794
+    },
+    {
+      "epoch": 1.358561967833491,
+      "grad_norm": 2.2425284385681152,
+      "learning_rate": 1.1690732278943344e-05,
+      "loss": 1.2247,
+      "step": 1795
+    },
+    {
+      "epoch": 1.3593188268684957,
+      "grad_norm": 2.589672803878784,
+      "learning_rate": 1.1682819392188917e-05,
+      "loss": 1.144,
+      "step": 1796
+    },
+    {
+      "epoch": 1.3600756859035004,
+      "grad_norm": 2.2635231018066406,
+      "learning_rate": 1.1674905421359358e-05,
+      "loss": 1.1585,
+      "step": 1797
+    },
+    {
+      "epoch": 1.3608325449385053,
+      "grad_norm": 2.291184425354004,
+      "learning_rate": 1.1666990371555893e-05,
+      "loss": 1.1063,
+      "step": 1798
+    },
+    {
+      "epoch": 1.36158940397351,
+      "grad_norm": 2.289581298828125,
+      "learning_rate": 1.1659074247880442e-05,
+      "loss": 1.183,
+      "step": 1799
+    },
+    {
+      "epoch": 1.3623462630085146,
+      "grad_norm": 2.1125857830047607,
+      "learning_rate": 1.1651157055435616e-05,
+      "loss": 1.1226,
+      "step": 1800
+    },
+    {
+      "epoch": 1.3631031220435195,
+      "grad_norm": 2.1084022521972656,
+      "learning_rate": 1.1643238799324714e-05,
+      "loss": 1.1741,
+      "step": 1801
+    },
+    {
+      "epoch": 1.3638599810785241,
+      "grad_norm": 2.2463252544403076,
+      "learning_rate": 1.1635319484651733e-05,
+      "loss": 1.1459,
+      "step": 1802
+    },
+    {
+      "epoch": 1.3646168401135288,
+      "grad_norm": 2.2021613121032715,
+      "learning_rate": 1.1627399116521334e-05,
+      "loss": 1.1939,
+      "step": 1803
+    },
+    {
+      "epoch": 1.3653736991485337,
+      "grad_norm": 2.190654754638672,
+      "learning_rate": 1.1619477700038863e-05,
+      "loss": 1.0967,
+      "step": 1804
+    },
+    {
+      "epoch": 1.3661305581835383,
+      "grad_norm": 1.9912575483322144,
+      "learning_rate": 1.1611555240310356e-05,
+      "loss": 1.1268,
+      "step": 1805
+    },
+    {
+      "epoch": 1.366887417218543,
+      "grad_norm": 2.1702189445495605,
+      "learning_rate": 1.16036317424425e-05,
+      "loss": 1.1957,
+      "step": 1806
+    },
+    {
+      "epoch": 1.367644276253548,
+      "grad_norm": 2.0921695232391357,
+      "learning_rate": 1.1595707211542662e-05,
+      "loss": 1.1161,
+      "step": 1807
+    },
+    {
+      "epoch": 1.3684011352885526,
+      "grad_norm": 2.1319305896759033,
+      "learning_rate": 1.1587781652718877e-05,
+      "loss": 1.1411,
+      "step": 1808
+    },
+    {
+      "epoch": 1.3691579943235572,
+      "grad_norm": 2.2225658893585205,
+      "learning_rate": 1.1579855071079838e-05,
+      "loss": 1.1259,
+      "step": 1809
+    },
+    {
+      "epoch": 1.369914853358562,
+      "grad_norm": 1.943051815032959,
+      "learning_rate": 1.1571927471734894e-05,
+      "loss": 1.1088,
+      "step": 1810
+    },
+    {
+      "epoch": 1.3706717123935668,
+      "grad_norm": 2.3888943195343018,
+      "learning_rate": 1.156399885979406e-05,
+      "loss": 1.1416,
+      "step": 1811
+    },
+    {
+      "epoch": 1.3714285714285714,
+      "grad_norm": 2.145301580429077,
+      "learning_rate": 1.1556069240368002e-05,
+      "loss": 1.1412,
+      "step": 1812
+    },
+    {
+      "epoch": 1.372185430463576,
+      "grad_norm": 2.0973587036132812,
+      "learning_rate": 1.1548138618568023e-05,
+      "loss": 1.1358,
+      "step": 1813
+    },
+    {
+      "epoch": 1.3729422894985808,
+      "grad_norm": 2.2998507022857666,
+      "learning_rate": 1.1540206999506086e-05,
+      "loss": 1.152,
+      "step": 1814
+    },
+    {
+      "epoch": 1.3736991485335857,
+      "grad_norm": 2.1464428901672363,
+      "learning_rate": 1.1532274388294789e-05,
+      "loss": 1.105,
+      "step": 1815
+    },
+    {
+      "epoch": 1.3744560075685903,
+      "grad_norm": 2.0508806705474854,
+      "learning_rate": 1.152434079004738e-05,
+      "loss": 1.1425,
+      "step": 1816
+    },
+    {
+      "epoch": 1.375212866603595,
+      "grad_norm": 2.05656099319458,
+      "learning_rate": 1.151640620987772e-05,
+      "loss": 1.1042,
+      "step": 1817
+    },
+    {
+      "epoch": 1.3759697256385999,
+      "grad_norm": 2.153604030609131,
+      "learning_rate": 1.1508470652900332e-05,
+      "loss": 1.1361,
+      "step": 1818
+    },
+    {
+      "epoch": 1.3767265846736045,
+      "grad_norm": 2.9740560054779053,
+      "learning_rate": 1.1500534124230354e-05,
+      "loss": 1.1646,
+      "step": 1819
+    },
+    {
+      "epoch": 1.3774834437086092,
+      "grad_norm": 2.213672637939453,
+      "learning_rate": 1.149259662898354e-05,
+      "loss": 1.1348,
+      "step": 1820
+    },
+    {
+      "epoch": 1.378240302743614,
+      "grad_norm": 2.2703373432159424,
+      "learning_rate": 1.148465817227629e-05,
+      "loss": 1.1456,
+      "step": 1821
+    },
+    {
+      "epoch": 1.3789971617786188,
+      "grad_norm": 2.1815407276153564,
+      "learning_rate": 1.1476718759225611e-05,
+      "loss": 1.16,
+      "step": 1822
+    },
+    {
+      "epoch": 1.3797540208136234,
+      "grad_norm": 2.198272943496704,
+      "learning_rate": 1.1468778394949123e-05,
+      "loss": 1.1677,
+      "step": 1823
+    },
+    {
+      "epoch": 1.3805108798486283,
+      "grad_norm": 2.1629281044006348,
+      "learning_rate": 1.1460837084565064e-05,
+      "loss": 1.1269,
+      "step": 1824
+    },
+    {
+      "epoch": 1.381267738883633,
+      "grad_norm": 1.9979993104934692,
+      "learning_rate": 1.1452894833192287e-05,
+      "loss": 1.1243,
+      "step": 1825
+    },
+    {
+      "epoch": 1.3820245979186376,
+      "grad_norm": 2.1406540870666504,
+      "learning_rate": 1.144495164595024e-05,
+      "loss": 1.1819,
+      "step": 1826
+    },
+    {
+      "epoch": 1.3827814569536425,
+      "grad_norm": 2.2074644565582275,
+      "learning_rate": 1.1437007527958985e-05,
+      "loss": 1.1368,
+      "step": 1827
+    },
+    {
+      "epoch": 1.3835383159886472,
+      "grad_norm": 2.279019355773926,
+      "learning_rate": 1.1429062484339175e-05,
+      "loss": 1.1293,
+      "step": 1828
+    },
+    {
+      "epoch": 1.3842951750236518,
+      "grad_norm": 2.179516315460205,
+      "learning_rate": 1.1421116520212066e-05,
+      "loss": 1.1538,
+      "step": 1829
+    },
+    {
+      "epoch": 1.3850520340586565,
+      "grad_norm": 2.0977933406829834,
+      "learning_rate": 1.1413169640699505e-05,
+      "loss": 1.1259,
+      "step": 1830
+    },
+    {
+      "epoch": 1.3858088930936612,
+      "grad_norm": 2.1527068614959717,
+      "learning_rate": 1.1405221850923932e-05,
+      "loss": 1.0934,
+      "step": 1831
+    },
+    {
+      "epoch": 1.386565752128666,
+      "grad_norm": 2.2525691986083984,
+      "learning_rate": 1.1397273156008364e-05,
+      "loss": 1.2084,
+      "step": 1832
+    },
+    {
+      "epoch": 1.3873226111636707,
+      "grad_norm": 2.0335781574249268,
+      "learning_rate": 1.1389323561076419e-05,
+      "loss": 1.1224,
+      "step": 1833
+    },
+    {
+      "epoch": 1.3880794701986754,
+      "grad_norm": 2.142940044403076,
+      "learning_rate": 1.1381373071252273e-05,
+      "loss": 1.0934,
+      "step": 1834
+    },
+    {
+      "epoch": 1.3888363292336803,
+      "grad_norm": 2.2513163089752197,
+      "learning_rate": 1.1373421691660697e-05,
+      "loss": 1.161,
+      "step": 1835
+    },
+    {
+      "epoch": 1.389593188268685,
+      "grad_norm": 2.1784231662750244,
+      "learning_rate": 1.1365469427427037e-05,
+      "loss": 1.127,
+      "step": 1836
+    },
+    {
+      "epoch": 1.3903500473036896,
+      "grad_norm": 2.2235348224639893,
+      "learning_rate": 1.1357516283677185e-05,
+      "loss": 1.1595,
+      "step": 1837
+    },
+    {
+      "epoch": 1.3911069063386945,
+      "grad_norm": 2.08614182472229,
+      "learning_rate": 1.1349562265537626e-05,
+      "loss": 1.1083,
+      "step": 1838
+    },
+    {
+      "epoch": 1.3918637653736992,
+      "grad_norm": 1.9127520322799683,
+      "learning_rate": 1.1341607378135395e-05,
+      "loss": 1.1516,
+      "step": 1839
+    },
+    {
+      "epoch": 1.3926206244087038,
+      "grad_norm": 2.100748300552368,
+      "learning_rate": 1.1333651626598095e-05,
+      "loss": 1.1306,
+      "step": 1840
+    },
+    {
+      "epoch": 1.3933774834437087,
+      "grad_norm": 2.3174188137054443,
+      "learning_rate": 1.1325695016053878e-05,
+      "loss": 1.1893,
+      "step": 1841
+    },
+    {
+      "epoch": 1.3941343424787134,
+      "grad_norm": 2.4146411418914795,
+      "learning_rate": 1.1317737551631455e-05,
+      "loss": 1.1463,
+      "step": 1842
+    },
+    {
+      "epoch": 1.394891201513718,
+      "grad_norm": 2.604128837585449,
+      "learning_rate": 1.130977923846009e-05,
+      "loss": 1.131,
+      "step": 1843
+    },
+    {
+      "epoch": 1.395648060548723,
+      "grad_norm": 2.1692941188812256,
+      "learning_rate": 1.1301820081669586e-05,
+      "loss": 1.1504,
+      "step": 1844
+    },
+    {
+      "epoch": 1.3964049195837276,
+      "grad_norm": 1.9960031509399414,
+      "learning_rate": 1.1293860086390294e-05,
+      "loss": 1.133,
+      "step": 1845
+    },
+    {
+      "epoch": 1.3971617786187323,
+      "grad_norm": 2.2130203247070312,
+      "learning_rate": 1.1285899257753105e-05,
+      "loss": 1.1375,
+      "step": 1846
+    },
+    {
+      "epoch": 1.397918637653737,
+      "grad_norm": 2.1830358505249023,
+      "learning_rate": 1.1277937600889458e-05,
+      "loss": 1.1391,
+      "step": 1847
+    },
+    {
+      "epoch": 1.3986754966887418,
+      "grad_norm": 2.188948392868042,
+      "learning_rate": 1.1269975120931301e-05,
+      "loss": 1.1241,
+      "step": 1848
+    },
+    {
+      "epoch": 1.3994323557237465,
+      "grad_norm": 2.3037242889404297,
+      "learning_rate": 1.1262011823011132e-05,
+      "loss": 1.1221,
+      "step": 1849
+    },
+    {
+      "epoch": 1.4001892147587511,
+      "grad_norm": 2.0598981380462646,
+      "learning_rate": 1.1254047712261975e-05,
+      "loss": 1.108,
+      "step": 1850
+    },
+    {
+      "epoch": 1.4009460737937558,
+      "grad_norm": 2.116628885269165,
+      "learning_rate": 1.1246082793817372e-05,
+      "loss": 1.1166,
+      "step": 1851
+    },
+    {
+      "epoch": 1.4017029328287607,
+      "grad_norm": 2.09624981880188,
+      "learning_rate": 1.1238117072811389e-05,
+      "loss": 1.1216,
+      "step": 1852
+    },
+    {
+      "epoch": 1.4024597918637653,
+      "grad_norm": 2.092494487762451,
+      "learning_rate": 1.1230150554378606e-05,
+      "loss": 1.1447,
+      "step": 1853
+    },
+    {
+      "epoch": 1.40321665089877,
+      "grad_norm": 1.9815505743026733,
+      "learning_rate": 1.1222183243654119e-05,
+      "loss": 1.1939,
+      "step": 1854
+    },
+    {
+      "epoch": 1.403973509933775,
+      "grad_norm": 2.285538673400879,
+      "learning_rate": 1.121421514577354e-05,
+      "loss": 1.1981,
+      "step": 1855
+    },
+    {
+      "epoch": 1.4047303689687796,
+      "grad_norm": 2.4579432010650635,
+      "learning_rate": 1.1206246265872975e-05,
+      "loss": 1.1246,
+      "step": 1856
+    },
+    {
+      "epoch": 1.4054872280037842,
+      "grad_norm": 2.2195796966552734,
+      "learning_rate": 1.1198276609089051e-05,
+      "loss": 1.0943,
+      "step": 1857
+    },
+    {
+      "epoch": 1.4062440870387891,
+      "grad_norm": 2.3332061767578125,
+      "learning_rate": 1.1190306180558886e-05,
+      "loss": 1.1896,
+      "step": 1858
+    },
+    {
+      "epoch": 1.4070009460737938,
+      "grad_norm": 2.257955551147461,
+      "learning_rate": 1.1182334985420088e-05,
+      "loss": 1.1565,
+      "step": 1859
+    },
+    {
+      "epoch": 1.4077578051087984,
+      "grad_norm": 2.1527364253997803,
+      "learning_rate": 1.1174363028810782e-05,
+      "loss": 1.1269,
+      "step": 1860
+    },
+    {
+      "epoch": 1.4085146641438033,
+      "grad_norm": 2.168989896774292,
+      "learning_rate": 1.1166390315869555e-05,
+      "loss": 1.118,
+      "step": 1861
+    },
+    {
+      "epoch": 1.409271523178808,
+      "grad_norm": 2.1610758304595947,
+      "learning_rate": 1.1158416851735505e-05,
+      "loss": 1.1126,
+      "step": 1862
+    },
+    {
+      "epoch": 1.4100283822138127,
+      "grad_norm": 2.423572063446045,
+      "learning_rate": 1.1150442641548205e-05,
+      "loss": 1.1681,
+      "step": 1863
+    },
+    {
+      "epoch": 1.4107852412488175,
+      "grad_norm": 2.1142797470092773,
+      "learning_rate": 1.1142467690447708e-05,
+      "loss": 1.1159,
+      "step": 1864
+    },
+    {
+      "epoch": 1.4115421002838222,
+      "grad_norm": 2.206160068511963,
+      "learning_rate": 1.1134492003574541e-05,
+      "loss": 1.1007,
+      "step": 1865
+    },
+    {
+      "epoch": 1.4122989593188269,
+      "grad_norm": 2.223226547241211,
+      "learning_rate": 1.1126515586069716e-05,
+      "loss": 1.0648,
+      "step": 1866
+    },
+    {
+      "epoch": 1.4130558183538315,
+      "grad_norm": 2.488703966140747,
+      "learning_rate": 1.1118538443074713e-05,
+      "loss": 1.107,
+      "step": 1867
+    },
+    {
+      "epoch": 1.4138126773888362,
+      "grad_norm": 2.2958545684814453,
+      "learning_rate": 1.1110560579731469e-05,
+      "loss": 1.1553,
+      "step": 1868
+    },
+    {
+      "epoch": 1.414569536423841,
+      "grad_norm": 2.343440055847168,
+      "learning_rate": 1.1102582001182399e-05,
+      "loss": 1.1225,
+      "step": 1869
+    },
+    {
+      "epoch": 1.4153263954588458,
+      "grad_norm": 2.963460922241211,
+      "learning_rate": 1.1094602712570366e-05,
+      "loss": 1.1211,
+      "step": 1870
+    },
+    {
+      "epoch": 1.4160832544938504,
+      "grad_norm": 2.123777151107788,
+      "learning_rate": 1.1086622719038708e-05,
+      "loss": 1.0919,
+      "step": 1871
+    },
+    {
+      "epoch": 1.4168401135288553,
+      "grad_norm": 2.1496341228485107,
+      "learning_rate": 1.1078642025731197e-05,
+      "loss": 1.0807,
+      "step": 1872
+    },
+    {
+      "epoch": 1.41759697256386,
+      "grad_norm": 2.147340774536133,
+      "learning_rate": 1.107066063779207e-05,
+      "loss": 1.1372,
+      "step": 1873
+    },
+    {
+      "epoch": 1.4183538315988646,
+      "grad_norm": 2.05765438079834,
+      "learning_rate": 1.1062678560366013e-05,
+      "loss": 1.1531,
+      "step": 1874
+    },
+    {
+      "epoch": 1.4191106906338695,
+      "grad_norm": 2.409080982208252,
+      "learning_rate": 1.1054695798598142e-05,
+      "loss": 1.1531,
+      "step": 1875
+    },
+    {
+      "epoch": 1.4198675496688742,
+      "grad_norm": 1.9932847023010254,
+      "learning_rate": 1.104671235763403e-05,
+      "loss": 1.1766,
+      "step": 1876
+    },
+    {
+      "epoch": 1.4206244087038788,
+      "grad_norm": 2.2019896507263184,
+      "learning_rate": 1.1038728242619686e-05,
+      "loss": 1.1037,
+      "step": 1877
+    },
+    {
+      "epoch": 1.4213812677388837,
+      "grad_norm": 2.263040065765381,
+      "learning_rate": 1.1030743458701533e-05,
+      "loss": 1.1475,
+      "step": 1878
+    },
+    {
+      "epoch": 1.4221381267738884,
+      "grad_norm": 2.0611464977264404,
+      "learning_rate": 1.1022758011026455e-05,
+      "loss": 1.1652,
+      "step": 1879
+    },
+    {
+      "epoch": 1.422894985808893,
+      "grad_norm": 2.175058364868164,
+      "learning_rate": 1.1014771904741746e-05,
+      "loss": 1.1164,
+      "step": 1880
+    },
+    {
+      "epoch": 1.423651844843898,
+      "grad_norm": 2.152921676635742,
+      "learning_rate": 1.1006785144995127e-05,
+      "loss": 1.1327,
+      "step": 1881
+    },
+    {
+      "epoch": 1.4244087038789026,
+      "grad_norm": 2.4787025451660156,
+      "learning_rate": 1.0998797736934743e-05,
+      "loss": 1.1633,
+      "step": 1882
+    },
+    {
+      "epoch": 1.4251655629139073,
+      "grad_norm": 2.7934088706970215,
+      "learning_rate": 1.0990809685709149e-05,
+      "loss": 1.1831,
+      "step": 1883
+    },
+    {
+      "epoch": 1.425922421948912,
+      "grad_norm": 2.058727502822876,
+      "learning_rate": 1.0982820996467334e-05,
+      "loss": 1.1262,
+      "step": 1884
+    },
+    {
+      "epoch": 1.4266792809839166,
+      "grad_norm": 2.1673519611358643,
+      "learning_rate": 1.0974831674358674e-05,
+      "loss": 1.1478,
+      "step": 1885
+    },
+    {
+      "epoch": 1.4274361400189215,
+      "grad_norm": 2.481576442718506,
+      "learning_rate": 1.0966841724532966e-05,
+      "loss": 1.1166,
+      "step": 1886
+    },
+    {
+      "epoch": 1.4281929990539262,
+      "grad_norm": 2.131117820739746,
+      "learning_rate": 1.0958851152140413e-05,
+      "loss": 1.084,
+      "step": 1887
+    },
+    {
+      "epoch": 1.4289498580889308,
+      "grad_norm": 2.3017077445983887,
+      "learning_rate": 1.095085996233162e-05,
+      "loss": 1.1406,
+      "step": 1888
+    },
+    {
+      "epoch": 1.4297067171239357,
+      "grad_norm": 2.4619855880737305,
+      "learning_rate": 1.0942868160257574e-05,
+      "loss": 1.1287,
+      "step": 1889
+    },
+    {
+      "epoch": 1.4304635761589404,
+      "grad_norm": 2.3940885066986084,
+      "learning_rate": 1.0934875751069679e-05,
+      "loss": 1.1135,
+      "step": 1890
+    },
+    {
+      "epoch": 1.431220435193945,
+      "grad_norm": 2.3564260005950928,
+      "learning_rate": 1.0926882739919718e-05,
+      "loss": 1.1474,
+      "step": 1891
+    },
+    {
+      "epoch": 1.43197729422895,
+      "grad_norm": 2.1630711555480957,
+      "learning_rate": 1.091888913195986e-05,
+      "loss": 1.1622,
+      "step": 1892
+    },
+    {
+      "epoch": 1.4327341532639546,
+      "grad_norm": 2.173370838165283,
+      "learning_rate": 1.0910894932342666e-05,
+      "loss": 1.1384,
+      "step": 1893
+    },
+    {
+      "epoch": 1.4334910122989593,
+      "grad_norm": 2.09320068359375,
+      "learning_rate": 1.0902900146221075e-05,
+      "loss": 1.1625,
+      "step": 1894
+    },
+    {
+      "epoch": 1.4342478713339641,
+      "grad_norm": 2.508751630783081,
+      "learning_rate": 1.0894904778748406e-05,
+      "loss": 1.1457,
+      "step": 1895
+    },
+    {
+      "epoch": 1.4350047303689688,
+      "grad_norm": 2.234450578689575,
+      "learning_rate": 1.0886908835078349e-05,
+      "loss": 1.1785,
+      "step": 1896
+    },
+    {
+      "epoch": 1.4357615894039735,
+      "grad_norm": 2.431640148162842,
+      "learning_rate": 1.0878912320364962e-05,
+      "loss": 1.1465,
+      "step": 1897
+    },
+    {
+      "epoch": 1.4365184484389784,
+      "grad_norm": 2.068406581878662,
+      "learning_rate": 1.087091523976269e-05,
+      "loss": 1.1386,
+      "step": 1898
+    },
+    {
+      "epoch": 1.437275307473983,
+      "grad_norm": 2.1216137409210205,
+      "learning_rate": 1.0862917598426315e-05,
+      "loss": 1.1177,
+      "step": 1899
+    },
+    {
+      "epoch": 1.4380321665089877,
+      "grad_norm": 2.36860990524292,
+      "learning_rate": 1.0854919401511002e-05,
+      "loss": 1.1612,
+      "step": 1900
+    },
+    {
+      "epoch": 1.4387890255439924,
+      "grad_norm": 2.2473835945129395,
+      "learning_rate": 1.0846920654172264e-05,
+      "loss": 1.1123,
+      "step": 1901
+    },
+    {
+      "epoch": 1.4395458845789972,
+      "grad_norm": 2.1833431720733643,
+      "learning_rate": 1.0838921361565978e-05,
+      "loss": 1.133,
+      "step": 1902
+    },
+    {
+      "epoch": 1.440302743614002,
+      "grad_norm": 2.078742742538452,
+      "learning_rate": 1.0830921528848355e-05,
+      "loss": 1.1634,
+      "step": 1903
+    },
+    {
+      "epoch": 1.4410596026490066,
+      "grad_norm": 2.1691906452178955,
+      "learning_rate": 1.0822921161175974e-05,
+      "loss": 1.1557,
+      "step": 1904
+    },
+    {
+      "epoch": 1.4418164616840112,
+      "grad_norm": 2.200441598892212,
+      "learning_rate": 1.0814920263705746e-05,
+      "loss": 1.1438,
+      "step": 1905
+    },
+    {
+      "epoch": 1.4425733207190161,
+      "grad_norm": 2.1800084114074707,
+      "learning_rate": 1.0806918841594929e-05,
+      "loss": 1.1395,
+      "step": 1906
+    },
+    {
+      "epoch": 1.4433301797540208,
+      "grad_norm": 2.34407901763916,
+      "learning_rate": 1.0798916900001117e-05,
+      "loss": 1.1448,
+      "step": 1907
+    },
+    {
+      "epoch": 1.4440870387890254,
+      "grad_norm": 2.1149091720581055,
+      "learning_rate": 1.0790914444082244e-05,
+      "loss": 1.1664,
+      "step": 1908
+    },
+    {
+      "epoch": 1.4448438978240303,
+      "grad_norm": 2.3421874046325684,
+      "learning_rate": 1.0782911478996559e-05,
+      "loss": 1.1109,
+      "step": 1909
+    },
+    {
+      "epoch": 1.445600756859035,
+      "grad_norm": 2.139888286590576,
+      "learning_rate": 1.0774908009902663e-05,
+      "loss": 1.1852,
+      "step": 1910
+    },
+    {
+      "epoch": 1.4463576158940397,
+      "grad_norm": 2.1743266582489014,
+      "learning_rate": 1.0766904041959465e-05,
+      "loss": 1.0994,
+      "step": 1911
+    },
+    {
+      "epoch": 1.4471144749290445,
+      "grad_norm": 2.0239481925964355,
+      "learning_rate": 1.0758899580326203e-05,
+      "loss": 1.0712,
+      "step": 1912
+    },
+    {
+      "epoch": 1.4478713339640492,
+      "grad_norm": 2.5557572841644287,
+      "learning_rate": 1.0750894630162429e-05,
+      "loss": 1.0855,
+      "step": 1913
+    },
+    {
+      "epoch": 1.4486281929990539,
+      "grad_norm": 2.1770548820495605,
+      "learning_rate": 1.0742889196628014e-05,
+      "loss": 1.1541,
+      "step": 1914
+    },
+    {
+      "epoch": 1.4493850520340588,
+      "grad_norm": 2.065044641494751,
+      "learning_rate": 1.073488328488314e-05,
+      "loss": 1.0722,
+      "step": 1915
+    },
+    {
+      "epoch": 1.4501419110690634,
+      "grad_norm": 2.274731159210205,
+      "learning_rate": 1.0726876900088287e-05,
+      "loss": 1.1562,
+      "step": 1916
+    },
+    {
+      "epoch": 1.450898770104068,
+      "grad_norm": 2.2915658950805664,
+      "learning_rate": 1.0718870047404253e-05,
+      "loss": 1.1573,
+      "step": 1917
+    },
+    {
+      "epoch": 1.451655629139073,
+      "grad_norm": 2.391997814178467,
+      "learning_rate": 1.0710862731992138e-05,
+      "loss": 1.1434,
+      "step": 1918
+    },
+    {
+      "epoch": 1.4524124881740776,
+      "grad_norm": 2.3179776668548584,
+      "learning_rate": 1.0702854959013332e-05,
+      "loss": 1.1352,
+      "step": 1919
+    },
+    {
+      "epoch": 1.4531693472090823,
+      "grad_norm": 2.223360776901245,
+      "learning_rate": 1.0694846733629519e-05,
+      "loss": 1.1152,
+      "step": 1920
+    },
+    {
+      "epoch": 1.453926206244087,
+      "grad_norm": 2.222038984298706,
+      "learning_rate": 1.0686838061002684e-05,
+      "loss": 1.0796,
+      "step": 1921
+    },
+    {
+      "epoch": 1.4546830652790916,
+      "grad_norm": 2.1372921466827393,
+      "learning_rate": 1.0678828946295099e-05,
+      "loss": 1.1047,
+      "step": 1922
+    },
+    {
+      "epoch": 1.4554399243140965,
+      "grad_norm": 2.1707942485809326,
+      "learning_rate": 1.0670819394669308e-05,
+      "loss": 1.1509,
+      "step": 1923
+    },
+    {
+      "epoch": 1.4561967833491012,
+      "grad_norm": 2.112736463546753,
+      "learning_rate": 1.066280941128815e-05,
+      "loss": 1.1266,
+      "step": 1924
+    },
+    {
+      "epoch": 1.4569536423841059,
+      "grad_norm": 2.3490540981292725,
+      "learning_rate": 1.065479900131474e-05,
+      "loss": 1.1621,
+      "step": 1925
+    },
+    {
+      "epoch": 1.4577105014191107,
+      "grad_norm": 2.17901873588562,
+      "learning_rate": 1.0646788169912465e-05,
+      "loss": 1.0735,
+      "step": 1926
+    },
+    {
+      "epoch": 1.4584673604541154,
+      "grad_norm": 2.0860230922698975,
+      "learning_rate": 1.0638776922244982e-05,
+      "loss": 1.1362,
+      "step": 1927
+    },
+    {
+      "epoch": 1.45922421948912,
+      "grad_norm": 2.0391974449157715,
+      "learning_rate": 1.0630765263476221e-05,
+      "loss": 1.1316,
+      "step": 1928
+    },
+    {
+      "epoch": 1.459981078524125,
+      "grad_norm": 2.0687365531921387,
+      "learning_rate": 1.062275319877038e-05,
+      "loss": 1.128,
+      "step": 1929
+    },
+    {
+      "epoch": 1.4607379375591296,
+      "grad_norm": 2.0217580795288086,
+      "learning_rate": 1.0614740733291902e-05,
+      "loss": 1.1377,
+      "step": 1930
+    },
+    {
+      "epoch": 1.4614947965941343,
+      "grad_norm": 2.0296125411987305,
+      "learning_rate": 1.060672787220551e-05,
+      "loss": 1.1236,
+      "step": 1931
+    },
+    {
+      "epoch": 1.4622516556291392,
+      "grad_norm": 2.0273191928863525,
+      "learning_rate": 1.0598714620676171e-05,
+      "loss": 1.1271,
+      "step": 1932
+    },
+    {
+      "epoch": 1.4630085146641438,
+      "grad_norm": 2.011613130569458,
+      "learning_rate": 1.05907009838691e-05,
+      "loss": 1.1457,
+      "step": 1933
+    },
+    {
+      "epoch": 1.4637653736991485,
+      "grad_norm": 2.0048105716705322,
+      "learning_rate": 1.058268696694977e-05,
+      "loss": 1.1499,
+      "step": 1934
+    },
+    {
+      "epoch": 1.4645222327341534,
+      "grad_norm": 2.086610794067383,
+      "learning_rate": 1.0574672575083891e-05,
+      "loss": 1.1376,
+      "step": 1935
+    },
+    {
+      "epoch": 1.465279091769158,
+      "grad_norm": 2.2125232219696045,
+      "learning_rate": 1.0566657813437419e-05,
+      "loss": 1.1103,
+      "step": 1936
+    },
+    {
+      "epoch": 1.4660359508041627,
+      "grad_norm": 2.172860622406006,
+      "learning_rate": 1.0558642687176548e-05,
+      "loss": 1.1306,
+      "step": 1937
+    },
+    {
+      "epoch": 1.4667928098391674,
+      "grad_norm": 2.1361825466156006,
+      "learning_rate": 1.0550627201467702e-05,
+      "loss": 1.0978,
+      "step": 1938
+    },
+    {
+      "epoch": 1.4675496688741723,
+      "grad_norm": 2.0148260593414307,
+      "learning_rate": 1.0542611361477548e-05,
+      "loss": 1.0851,
+      "step": 1939
+    },
+    {
+      "epoch": 1.468306527909177,
+      "grad_norm": 2.103895664215088,
+      "learning_rate": 1.0534595172372967e-05,
+      "loss": 1.1197,
+      "step": 1940
+    },
+    {
+      "epoch": 1.4690633869441816,
+      "grad_norm": 2.1808462142944336,
+      "learning_rate": 1.0526578639321078e-05,
+      "loss": 1.1192,
+      "step": 1941
+    },
+    {
+      "epoch": 1.4698202459791863,
+      "grad_norm": 2.2360849380493164,
+      "learning_rate": 1.0518561767489211e-05,
+      "loss": 1.1142,
+      "step": 1942
+    },
+    {
+      "epoch": 1.4705771050141911,
+      "grad_norm": 2.243360996246338,
+      "learning_rate": 1.0510544562044925e-05,
+      "loss": 1.2133,
+      "step": 1943
+    },
+    {
+      "epoch": 1.4713339640491958,
+      "grad_norm": 2.07759690284729,
+      "learning_rate": 1.050252702815598e-05,
+      "loss": 1.1227,
+      "step": 1944
+    },
+    {
+      "epoch": 1.4720908230842005,
+      "grad_norm": 2.0380797386169434,
+      "learning_rate": 1.0494509170990362e-05,
+      "loss": 1.0894,
+      "step": 1945
+    },
+    {
+      "epoch": 1.4728476821192054,
+      "grad_norm": 2.184549570083618,
+      "learning_rate": 1.0486490995716264e-05,
+      "loss": 1.1314,
+      "step": 1946
+    },
+    {
+      "epoch": 1.47360454115421,
+      "grad_norm": 2.1510207653045654,
+      "learning_rate": 1.0478472507502069e-05,
+      "loss": 1.1688,
+      "step": 1947
+    },
+    {
+      "epoch": 1.4743614001892147,
+      "grad_norm": 2.1699905395507812,
+      "learning_rate": 1.0470453711516377e-05,
+      "loss": 1.1374,
+      "step": 1948
+    },
+    {
+      "epoch": 1.4751182592242196,
+      "grad_norm": 2.0163750648498535,
+      "learning_rate": 1.0462434612927984e-05,
+      "loss": 1.1469,
+      "step": 1949
+    },
+    {
+      "epoch": 1.4758751182592242,
+      "grad_norm": 2.176668882369995,
+      "learning_rate": 1.0454415216905875e-05,
+      "loss": 1.154,
+      "step": 1950
+    },
+    {
+      "epoch": 1.476631977294229,
+      "grad_norm": 2.18507981300354,
+      "learning_rate": 1.0446395528619236e-05,
+      "loss": 1.1175,
+      "step": 1951
+    },
+    {
+      "epoch": 1.4773888363292338,
+      "grad_norm": 2.033001661300659,
+      "learning_rate": 1.0438375553237428e-05,
+      "loss": 1.1129,
+      "step": 1952
+    },
+    {
+      "epoch": 1.4781456953642385,
+      "grad_norm": 2.0419886112213135,
+      "learning_rate": 1.0430355295930008e-05,
+      "loss": 1.1455,
+      "step": 1953
+    },
+    {
+      "epoch": 1.4789025543992431,
+      "grad_norm": 2.083308696746826,
+      "learning_rate": 1.0422334761866715e-05,
+      "loss": 1.1069,
+      "step": 1954
+    },
+    {
+      "epoch": 1.479659413434248,
+      "grad_norm": 2.0463309288024902,
+      "learning_rate": 1.0414313956217456e-05,
+      "loss": 1.1456,
+      "step": 1955
+    },
+    {
+      "epoch": 1.4804162724692527,
+      "grad_norm": 2.0065863132476807,
+      "learning_rate": 1.0406292884152327e-05,
+      "loss": 1.0829,
+      "step": 1956
+    },
+    {
+      "epoch": 1.4811731315042573,
+      "grad_norm": 1.8798035383224487,
+      "learning_rate": 1.0398271550841586e-05,
+      "loss": 1.1378,
+      "step": 1957
+    },
+    {
+      "epoch": 1.481929990539262,
+      "grad_norm": 2.483062267303467,
+      "learning_rate": 1.0390249961455658e-05,
+      "loss": 1.0775,
+      "step": 1958
+    },
+    {
+      "epoch": 1.4826868495742667,
+      "grad_norm": 1.995613694190979,
+      "learning_rate": 1.0382228121165146e-05,
+      "loss": 1.1936,
+      "step": 1959
+    },
+    {
+      "epoch": 1.4834437086092715,
+      "grad_norm": 2.1545281410217285,
+      "learning_rate": 1.03742060351408e-05,
+      "loss": 1.1802,
+      "step": 1960
+    },
+    {
+      "epoch": 1.4842005676442762,
+      "grad_norm": 2.1138501167297363,
+      "learning_rate": 1.0366183708553532e-05,
+      "loss": 1.102,
+      "step": 1961
+    },
+    {
+      "epoch": 1.4849574266792809,
+      "grad_norm": 2.1736159324645996,
+      "learning_rate": 1.0358161146574417e-05,
+      "loss": 1.1844,
+      "step": 1962
+    },
+    {
+      "epoch": 1.4857142857142858,
+      "grad_norm": 2.0476620197296143,
+      "learning_rate": 1.0350138354374675e-05,
+      "loss": 1.1117,
+      "step": 1963
+    },
+    {
+      "epoch": 1.4864711447492904,
+      "grad_norm": 2.070690631866455,
+      "learning_rate": 1.034211533712567e-05,
+      "loss": 1.0858,
+      "step": 1964
+    },
+    {
+      "epoch": 1.487228003784295,
+      "grad_norm": 2.256793975830078,
+      "learning_rate": 1.0334092099998926e-05,
+      "loss": 1.1564,
+      "step": 1965
+    },
+    {
+      "epoch": 1.4879848628193,
+      "grad_norm": 1.8769042491912842,
+      "learning_rate": 1.0326068648166088e-05,
+      "loss": 1.1211,
+      "step": 1966
+    },
+    {
+      "epoch": 1.4887417218543046,
+      "grad_norm": 2.271409749984741,
+      "learning_rate": 1.0318044986798961e-05,
+      "loss": 1.1329,
+      "step": 1967
+    },
+    {
+      "epoch": 1.4894985808893093,
+      "grad_norm": 2.035731315612793,
+      "learning_rate": 1.031002112106947e-05,
+      "loss": 1.0566,
+      "step": 1968
+    },
+    {
+      "epoch": 1.4902554399243142,
+      "grad_norm": 1.9846116304397583,
+      "learning_rate": 1.0301997056149678e-05,
+      "loss": 1.1373,
+      "step": 1969
+    },
+    {
+      "epoch": 1.4910122989593189,
+      "grad_norm": 2.304295301437378,
+      "learning_rate": 1.0293972797211774e-05,
+      "loss": 1.1098,
+      "step": 1970
+    },
+    {
+      "epoch": 1.4917691579943235,
+      "grad_norm": 2.189412832260132,
+      "learning_rate": 1.028594834942807e-05,
+      "loss": 1.1026,
+      "step": 1971
+    },
+    {
+      "epoch": 1.4925260170293284,
+      "grad_norm": 2.1527864933013916,
+      "learning_rate": 1.0277923717971006e-05,
+      "loss": 1.1262,
+      "step": 1972
+    },
+    {
+      "epoch": 1.493282876064333,
+      "grad_norm": 2.20159912109375,
+      "learning_rate": 1.026989890801314e-05,
+      "loss": 1.092,
+      "step": 1973
+    },
+    {
+      "epoch": 1.4940397350993377,
+      "grad_norm": 2.2014966011047363,
+      "learning_rate": 1.0261873924727138e-05,
+      "loss": 1.1267,
+      "step": 1974
+    },
+    {
+      "epoch": 1.4947965941343424,
+      "grad_norm": 2.194817304611206,
+      "learning_rate": 1.0253848773285778e-05,
+      "loss": 1.1565,
+      "step": 1975
+    },
+    {
+      "epoch": 1.4955534531693473,
+      "grad_norm": 2.061915636062622,
+      "learning_rate": 1.0245823458861958e-05,
+      "loss": 1.1291,
+      "step": 1976
+    },
+    {
+      "epoch": 1.496310312204352,
+      "grad_norm": 2.1642725467681885,
+      "learning_rate": 1.0237797986628672e-05,
+      "loss": 1.1161,
+      "step": 1977
+    },
+    {
+      "epoch": 1.4970671712393566,
+      "grad_norm": 2.0526773929595947,
+      "learning_rate": 1.022977236175901e-05,
+      "loss": 1.1583,
+      "step": 1978
+    },
+    {
+      "epoch": 1.4978240302743613,
+      "grad_norm": 2.0349247455596924,
+      "learning_rate": 1.0221746589426176e-05,
+      "loss": 1.1161,
+      "step": 1979
+    },
+    {
+      "epoch": 1.4985808893093662,
+      "grad_norm": 2.1763689517974854,
+      "learning_rate": 1.0213720674803458e-05,
+      "loss": 1.1344,
+      "step": 1980
+    },
+    {
+      "epoch": 1.4993377483443708,
+      "grad_norm": 2.139963150024414,
+      "learning_rate": 1.0205694623064236e-05,
+      "loss": 1.2086,
+      "step": 1981
+    },
+    {
+      "epoch": 1.5000946073793755,
+      "grad_norm": 2.633737564086914,
+      "learning_rate": 1.0197668439381978e-05,
+      "loss": 1.1523,
+      "step": 1982
+    },
+    {
+      "epoch": 1.5008514664143804,
+      "grad_norm": 2.0594277381896973,
+      "learning_rate": 1.0189642128930246e-05,
+      "loss": 1.1436,
+      "step": 1983
+    },
+    {
+      "epoch": 1.501608325449385,
+      "grad_norm": 2.1511809825897217,
+      "learning_rate": 1.0181615696882676e-05,
+      "loss": 1.1195,
+      "step": 1984
+    },
+    {
+      "epoch": 1.5023651844843897,
+      "grad_norm": 1.992146372795105,
+      "learning_rate": 1.0173589148412981e-05,
+      "loss": 1.1534,
+      "step": 1985
+    },
+    {
+      "epoch": 1.5031220435193946,
+      "grad_norm": 2.000650644302368,
+      "learning_rate": 1.0165562488694953e-05,
+      "loss": 1.1158,
+      "step": 1986
+    },
+    {
+      "epoch": 1.5038789025543993,
+      "grad_norm": 2.0944910049438477,
+      "learning_rate": 1.0157535722902456e-05,
+      "loss": 1.0991,
+      "step": 1987
+    },
+    {
+      "epoch": 1.504635761589404,
+      "grad_norm": 2.3380539417266846,
+      "learning_rate": 1.0149508856209416e-05,
+      "loss": 1.148,
+      "step": 1988
+    },
+    {
+      "epoch": 1.5053926206244088,
+      "grad_norm": 2.1337814331054688,
+      "learning_rate": 1.014148189378983e-05,
+      "loss": 1.1508,
+      "step": 1989
+    },
+    {
+      "epoch": 1.5061494796594135,
+      "grad_norm": 2.069946765899658,
+      "learning_rate": 1.0133454840817765e-05,
+      "loss": 1.1449,
+      "step": 1990
+    },
+    {
+      "epoch": 1.5069063386944181,
+      "grad_norm": 2.4194324016571045,
+      "learning_rate": 1.0125427702467327e-05,
+      "loss": 1.1833,
+      "step": 1991
+    },
+    {
+      "epoch": 1.507663197729423,
+      "grad_norm": 2.0037777423858643,
+      "learning_rate": 1.0117400483912687e-05,
+      "loss": 1.1053,
+      "step": 1992
+    },
+    {
+      "epoch": 1.5084200567644275,
+      "grad_norm": 1.9638372659683228,
+      "learning_rate": 1.010937319032807e-05,
+      "loss": 1.1249,
+      "step": 1993
+    },
+    {
+      "epoch": 1.5091769157994324,
+      "grad_norm": 2.185102939605713,
+      "learning_rate": 1.0101345826887752e-05,
+      "loss": 1.1369,
+      "step": 1994
+    },
+    {
+      "epoch": 1.5099337748344372,
+      "grad_norm": 2.193578004837036,
+      "learning_rate": 1.0093318398766042e-05,
+      "loss": 1.1268,
+      "step": 1995
+    },
+    {
+      "epoch": 1.5106906338694417,
+      "grad_norm": 2.1746068000793457,
+      "learning_rate": 1.0085290911137298e-05,
+      "loss": 1.1316,
+      "step": 1996
+    },
+    {
+      "epoch": 1.5114474929044466,
+      "grad_norm": 2.308969736099243,
+      "learning_rate": 1.0077263369175918e-05,
+      "loss": 1.11,
+      "step": 1997
+    },
+    {
+      "epoch": 1.5122043519394512,
+      "grad_norm": 2.2050511837005615,
+      "learning_rate": 1.0069235778056336e-05,
+      "loss": 1.1363,
+      "step": 1998
+    },
+    {
+      "epoch": 1.512961210974456,
+      "grad_norm": 2.351792812347412,
+      "learning_rate": 1.0061208142953012e-05,
+      "loss": 1.1222,
+      "step": 1999
+    },
+    {
+      "epoch": 1.5137180700094608,
+      "grad_norm": 2.144644021987915,
+      "learning_rate": 1.0053180469040433e-05,
+      "loss": 1.0997,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5144749290444655,
+      "grad_norm": 2.1637988090515137,
+      "learning_rate": 1.0045152761493127e-05,
+      "loss": 1.0968,
+      "step": 2001
+    },
+    {
+      "epoch": 1.5152317880794701,
+      "grad_norm": 2.200721502304077,
+      "learning_rate": 1.0037125025485616e-05,
+      "loss": 1.1016,
+      "step": 2002
+    },
+    {
+      "epoch": 1.515988647114475,
+      "grad_norm": 2.3035366535186768,
+      "learning_rate": 1.0029097266192467e-05,
+      "loss": 1.1659,
+      "step": 2003
+    },
+    {
+      "epoch": 1.5167455061494797,
+      "grad_norm": 2.34773588180542,
+      "learning_rate": 1.0021069488788253e-05,
+      "loss": 1.0888,
+      "step": 2004
+    },
+    {
+      "epoch": 1.5175023651844843,
+      "grad_norm": 2.268134117126465,
+      "learning_rate": 1.0013041698447547e-05,
+      "loss": 1.1519,
+      "step": 2005
+    },
+    {
+      "epoch": 1.5182592242194892,
+      "grad_norm": 2.331434726715088,
+      "learning_rate": 1.000501390034495e-05,
+      "loss": 1.1335,
+      "step": 2006
+    },
+    {
+      "epoch": 1.5190160832544939,
+      "grad_norm": 2.3400261402130127,
+      "learning_rate": 9.996986099655052e-06,
+      "loss": 1.1808,
+      "step": 2007
+    },
+    {
+      "epoch": 1.5197729422894986,
+      "grad_norm": 2.348576068878174,
+      "learning_rate": 9.988958301552454e-06,
+      "loss": 1.1358,
+      "step": 2008
+    },
+    {
+      "epoch": 1.5205298013245034,
+      "grad_norm": 2.131770610809326,
+      "learning_rate": 9.980930511211751e-06,
+      "loss": 1.0952,
+      "step": 2009
+    },
+    {
+      "epoch": 1.5212866603595079,
+      "grad_norm": 2.3337466716766357,
+      "learning_rate": 9.972902733807532e-06,
+      "loss": 1.1449,
+      "step": 2010
+    },
+    {
+      "epoch": 1.5220435193945128,
+      "grad_norm": 2.0936343669891357,
+      "learning_rate": 9.964874974514386e-06,
+      "loss": 1.1176,
+      "step": 2011
+    },
+    {
+      "epoch": 1.5228003784295177,
+      "grad_norm": 2.0963313579559326,
+      "learning_rate": 9.95684723850688e-06,
+      "loss": 1.1481,
+      "step": 2012
+    },
+    {
+      "epoch": 1.523557237464522,
+      "grad_norm": 2.055452823638916,
+      "learning_rate": 9.948819530959566e-06,
+      "loss": 1.1008,
+      "step": 2013
+    },
+    {
+      "epoch": 1.524314096499527,
+      "grad_norm": 2.257266044616699,
+      "learning_rate": 9.94079185704699e-06,
+      "loss": 1.1544,
+      "step": 2014
+    },
+    {
+      "epoch": 1.5250709555345316,
+      "grad_norm": 2.06075119972229,
+      "learning_rate": 9.932764221943666e-06,
+      "loss": 1.1153,
+      "step": 2015
+    },
+    {
+      "epoch": 1.5258278145695363,
+      "grad_norm": 2.2544174194335938,
+      "learning_rate": 9.924736630824083e-06,
+      "loss": 1.0718,
+      "step": 2016
+    },
+    {
+      "epoch": 1.5265846736045412,
+      "grad_norm": 2.1009559631347656,
+      "learning_rate": 9.916709088862707e-06,
+      "loss": 1.137,
+      "step": 2017
+    },
+    {
+      "epoch": 1.5273415326395459,
+      "grad_norm": 2.014848470687866,
+      "learning_rate": 9.908681601233964e-06,
+      "loss": 1.1268,
+      "step": 2018
+    },
+    {
+      "epoch": 1.5280983916745505,
+      "grad_norm": 2.2673892974853516,
+      "learning_rate": 9.900654173112251e-06,
+      "loss": 1.1528,
+      "step": 2019
+    },
+    {
+      "epoch": 1.5288552507095554,
+      "grad_norm": 2.206071138381958,
+      "learning_rate": 9.89262680967193e-06,
+      "loss": 1.1013,
+      "step": 2020
+    },
+    {
+      "epoch": 1.52961210974456,
+      "grad_norm": 2.102032423019409,
+      "learning_rate": 9.884599516087314e-06,
+      "loss": 1.105,
+      "step": 2021
+    },
+    {
+      "epoch": 1.5303689687795647,
+      "grad_norm": 2.1362051963806152,
+      "learning_rate": 9.876572297532677e-06,
+      "loss": 1.1479,
+      "step": 2022
+    },
+    {
+      "epoch": 1.5311258278145696,
+      "grad_norm": 2.0666024684906006,
+      "learning_rate": 9.868545159182238e-06,
+      "loss": 1.1257,
+      "step": 2023
+    },
+    {
+      "epoch": 1.5318826868495743,
+      "grad_norm": 1.9618515968322754,
+      "learning_rate": 9.860518106210167e-06,
+      "loss": 1.0649,
+      "step": 2024
+    },
+    {
+      "epoch": 1.532639545884579,
+      "grad_norm": 2.202753782272339,
+      "learning_rate": 9.852491143790587e-06,
+      "loss": 1.1016,
+      "step": 2025
+    },
+    {
+      "epoch": 1.5333964049195838,
+      "grad_norm": 1.9656624794006348,
+      "learning_rate": 9.844464277097549e-06,
+      "loss": 1.1435,
+      "step": 2026
+    },
+    {
+      "epoch": 1.5341532639545885,
+      "grad_norm": 2.06479811668396,
+      "learning_rate": 9.83643751130505e-06,
+      "loss": 1.1393,
+      "step": 2027
+    },
+    {
+      "epoch": 1.5349101229895932,
+      "grad_norm": 2.1722230911254883,
+      "learning_rate": 9.828410851587023e-06,
+      "loss": 1.1587,
+      "step": 2028
+    },
+    {
+      "epoch": 1.535666982024598,
+      "grad_norm": 2.090200901031494,
+      "learning_rate": 9.820384303117328e-06,
+      "loss": 1.109,
+      "step": 2029
+    },
+    {
+      "epoch": 1.5364238410596025,
+      "grad_norm": 1.9950278997421265,
+      "learning_rate": 9.812357871069754e-06,
+      "loss": 1.0648,
+      "step": 2030
+    },
+    {
+      "epoch": 1.5371807000946074,
+      "grad_norm": 2.141153573989868,
+      "learning_rate": 9.804331560618023e-06,
+      "loss": 1.1327,
+      "step": 2031
+    },
+    {
+      "epoch": 1.5379375591296123,
+      "grad_norm": 1.9659839868545532,
+      "learning_rate": 9.79630537693577e-06,
+      "loss": 1.1194,
+      "step": 2032
+    },
+    {
+      "epoch": 1.5386944181646167,
+      "grad_norm": 2.187727212905884,
+      "learning_rate": 9.788279325196547e-06,
+      "loss": 1.1225,
+      "step": 2033
+    },
+    {
+      "epoch": 1.5394512771996216,
+      "grad_norm": 2.4570298194885254,
+      "learning_rate": 9.780253410573827e-06,
+      "loss": 1.1252,
+      "step": 2034
+    },
+    {
+      "epoch": 1.5402081362346263,
+      "grad_norm": 2.2557146549224854,
+      "learning_rate": 9.772227638240993e-06,
+      "loss": 1.1698,
+      "step": 2035
+    },
+    {
+      "epoch": 1.540964995269631,
+      "grad_norm": 2.384152889251709,
+      "learning_rate": 9.764202013371333e-06,
+      "loss": 1.1447,
+      "step": 2036
+    },
+    {
+      "epoch": 1.5417218543046358,
+      "grad_norm": 2.0934481620788574,
+      "learning_rate": 9.756176541138045e-06,
+      "loss": 1.1429,
+      "step": 2037
+    },
+    {
+      "epoch": 1.5424787133396405,
+      "grad_norm": 2.418853282928467,
+      "learning_rate": 9.748151226714222e-06,
+      "loss": 1.1321,
+      "step": 2038
+    },
+    {
+      "epoch": 1.5432355723746451,
+      "grad_norm": 2.0321297645568848,
+      "learning_rate": 9.740126075272868e-06,
+      "loss": 1.0983,
+      "step": 2039
+    },
+    {
+      "epoch": 1.54399243140965,
+      "grad_norm": 2.088118076324463,
+      "learning_rate": 9.732101091986864e-06,
+      "loss": 1.1566,
+      "step": 2040
+    },
+    {
+      "epoch": 1.5447492904446547,
+      "grad_norm": 2.135477066040039,
+      "learning_rate": 9.724076282028993e-06,
+      "loss": 1.0886,
+      "step": 2041
+    },
+    {
+      "epoch": 1.5455061494796594,
+      "grad_norm": 2.1262335777282715,
+      "learning_rate": 9.716051650571933e-06,
+      "loss": 1.1461,
+      "step": 2042
+    },
+    {
+      "epoch": 1.5462630085146643,
+      "grad_norm": 2.2694787979125977,
+      "learning_rate": 9.708027202788229e-06,
+      "loss": 1.1725,
+      "step": 2043
+    },
+    {
+      "epoch": 1.547019867549669,
+      "grad_norm": 2.0884077548980713,
+      "learning_rate": 9.700002943850323e-06,
+      "loss": 1.1098,
+      "step": 2044
+    },
+    {
+      "epoch": 1.5477767265846736,
+      "grad_norm": 2.5522899627685547,
+      "learning_rate": 9.691978878930532e-06,
+      "loss": 1.0861,
+      "step": 2045
+    },
+    {
+      "epoch": 1.5485335856196785,
+      "grad_norm": 2.099339008331299,
+      "learning_rate": 9.68395501320104e-06,
+      "loss": 1.135,
+      "step": 2046
+    },
+    {
+      "epoch": 1.549290444654683,
+      "grad_norm": 2.0966038703918457,
+      "learning_rate": 9.675931351833911e-06,
+      "loss": 1.1468,
+      "step": 2047
+    },
+    {
+      "epoch": 1.5500473036896878,
+      "grad_norm": 1.972170352935791,
+      "learning_rate": 9.667907900001079e-06,
+      "loss": 1.0958,
+      "step": 2048
+    },
+    {
+      "epoch": 1.5508041627246927,
+      "grad_norm": 2.228671073913574,
+      "learning_rate": 9.659884662874332e-06,
+      "loss": 1.1338,
+      "step": 2049
+    },
+    {
+      "epoch": 1.5515610217596971,
+      "grad_norm": 1.9483565092086792,
+      "learning_rate": 9.65186164562533e-06,
+      "loss": 1.0757,
+      "step": 2050
+    },
+    {
+      "epoch": 1.552317880794702,
+      "grad_norm": 2.0136473178863525,
+      "learning_rate": 9.643838853425586e-06,
+      "loss": 1.1177,
+      "step": 2051
+    },
+    {
+      "epoch": 1.5530747398297067,
+      "grad_norm": 2.0868184566497803,
+      "learning_rate": 9.635816291446469e-06,
+      "loss": 1.1814,
+      "step": 2052
+    },
+    {
+      "epoch": 1.5538315988647113,
+      "grad_norm": 1.9814927577972412,
+      "learning_rate": 9.627793964859205e-06,
+      "loss": 1.1342,
+      "step": 2053
+    },
+    {
+      "epoch": 1.5545884578997162,
+      "grad_norm": 2.0822665691375732,
+      "learning_rate": 9.619771878834858e-06,
+      "loss": 1.0962,
+      "step": 2054
+    },
+    {
+      "epoch": 1.555345316934721,
+      "grad_norm": 2.0414929389953613,
+      "learning_rate": 9.611750038544343e-06,
+      "loss": 1.1178,
+      "step": 2055
+    },
+    {
+      "epoch": 1.5561021759697256,
+      "grad_norm": 2.134589672088623,
+      "learning_rate": 9.60372844915842e-06,
+      "loss": 1.133,
+      "step": 2056
+    },
+    {
+      "epoch": 1.5568590350047304,
+      "grad_norm": 2.3856427669525146,
+      "learning_rate": 9.595707115847676e-06,
+      "loss": 1.1252,
+      "step": 2057
+    },
+    {
+      "epoch": 1.557615894039735,
+      "grad_norm": 2.275172710418701,
+      "learning_rate": 9.587686043782545e-06,
+      "loss": 1.1195,
+      "step": 2058
+    },
+    {
+      "epoch": 1.5583727530747398,
+      "grad_norm": 2.2985713481903076,
+      "learning_rate": 9.579665238133291e-06,
+      "loss": 1.148,
+      "step": 2059
+    },
+    {
+      "epoch": 1.5591296121097447,
+      "grad_norm": 2.0514907836914062,
+      "learning_rate": 9.571644704069995e-06,
+      "loss": 1.1307,
+      "step": 2060
+    },
+    {
+      "epoch": 1.5598864711447493,
+      "grad_norm": 2.2182204723358154,
+      "learning_rate": 9.563624446762576e-06,
+      "loss": 1.0913,
+      "step": 2061
+    },
+    {
+      "epoch": 1.560643330179754,
+      "grad_norm": 2.628448486328125,
+      "learning_rate": 9.555604471380767e-06,
+      "loss": 1.1385,
+      "step": 2062
+    },
+    {
+      "epoch": 1.5614001892147589,
+      "grad_norm": 1.9690390825271606,
+      "learning_rate": 9.547584783094126e-06,
+      "loss": 1.1314,
+      "step": 2063
+    },
+    {
+      "epoch": 1.5621570482497635,
+      "grad_norm": 2.0425853729248047,
+      "learning_rate": 9.539565387072019e-06,
+      "loss": 1.1048,
+      "step": 2064
+    },
+    {
+      "epoch": 1.5629139072847682,
+      "grad_norm": 2.025308609008789,
+      "learning_rate": 9.531546288483624e-06,
+      "loss": 1.1012,
+      "step": 2065
+    },
+    {
+      "epoch": 1.563670766319773,
+      "grad_norm": 2.17830491065979,
+      "learning_rate": 9.523527492497934e-06,
+      "loss": 1.1926,
+      "step": 2066
+    },
+    {
+      "epoch": 1.5644276253547775,
+      "grad_norm": 3.244462251663208,
+      "learning_rate": 9.51550900428374e-06,
+      "loss": 1.0953,
+      "step": 2067
+    },
+    {
+      "epoch": 1.5651844843897824,
+      "grad_norm": 2.0593700408935547,
+      "learning_rate": 9.507490829009639e-06,
+      "loss": 1.1161,
+      "step": 2068
+    },
+    {
+      "epoch": 1.565941343424787,
+      "grad_norm": 2.0308477878570557,
+      "learning_rate": 9.49947297184402e-06,
+      "loss": 1.0959,
+      "step": 2069
+    },
+    {
+      "epoch": 1.5666982024597917,
+      "grad_norm": 2.1143085956573486,
+      "learning_rate": 9.491455437955081e-06,
+      "loss": 1.1541,
+      "step": 2070
+    },
+    {
+      "epoch": 1.5674550614947966,
+      "grad_norm": 2.3376524448394775,
+      "learning_rate": 9.483438232510792e-06,
+      "loss": 1.1283,
+      "step": 2071
+    },
+    {
+      "epoch": 1.5682119205298013,
+      "grad_norm": 2.194188117980957,
+      "learning_rate": 9.475421360678926e-06,
+      "loss": 1.1256,
+      "step": 2072
+    },
+    {
+      "epoch": 1.568968779564806,
+      "grad_norm": 2.12689208984375,
+      "learning_rate": 9.467404827627036e-06,
+      "loss": 1.1638,
+      "step": 2073
+    },
+    {
+      "epoch": 1.5697256385998108,
+      "grad_norm": 1.9852758646011353,
+      "learning_rate": 9.459388638522455e-06,
+      "loss": 1.1458,
+      "step": 2074
+    },
+    {
+      "epoch": 1.5704824976348155,
+      "grad_norm": 1.958489179611206,
+      "learning_rate": 9.4513727985323e-06,
+      "loss": 1.1222,
+      "step": 2075
+    },
+    {
+      "epoch": 1.5712393566698202,
+      "grad_norm": 2.1876025199890137,
+      "learning_rate": 9.443357312823454e-06,
+      "loss": 1.1322,
+      "step": 2076
+    },
+    {
+      "epoch": 1.571996215704825,
+      "grad_norm": 2.1041505336761475,
+      "learning_rate": 9.435342186562582e-06,
+      "loss": 1.1237,
+      "step": 2077
+    },
+    {
+      "epoch": 1.5727530747398297,
+      "grad_norm": 2.247180461883545,
+      "learning_rate": 9.427327424916113e-06,
+      "loss": 1.056,
+      "step": 2078
+    },
+    {
+      "epoch": 1.5735099337748344,
+      "grad_norm": 2.184521436691284,
+      "learning_rate": 9.419313033050232e-06,
+      "loss": 1.1022,
+      "step": 2079
+    },
+    {
+      "epoch": 1.5742667928098393,
+      "grad_norm": 2.078411340713501,
+      "learning_rate": 9.411299016130902e-06,
+      "loss": 1.1526,
+      "step": 2080
+    },
+    {
+      "epoch": 1.575023651844844,
+      "grad_norm": 2.154078245162964,
+      "learning_rate": 9.403285379323833e-06,
+      "loss": 1.2138,
+      "step": 2081
+    },
+    {
+      "epoch": 1.5757805108798486,
+      "grad_norm": 2.0813803672790527,
+      "learning_rate": 9.395272127794491e-06,
+      "loss": 1.0913,
+      "step": 2082
+    },
+    {
+      "epoch": 1.5765373699148535,
+      "grad_norm": 1.975311517715454,
+      "learning_rate": 9.387259266708104e-06,
+      "loss": 1.1674,
+      "step": 2083
+    },
+    {
+      "epoch": 1.577294228949858,
+      "grad_norm": 2.022935152053833,
+      "learning_rate": 9.379246801229626e-06,
+      "loss": 1.0664,
+      "step": 2084
+    },
+    {
+      "epoch": 1.5780510879848628,
+      "grad_norm": 2.3060450553894043,
+      "learning_rate": 9.371234736523781e-06,
+      "loss": 1.0884,
+      "step": 2085
+    },
+    {
+      "epoch": 1.5788079470198677,
+      "grad_norm": 2.224121570587158,
+      "learning_rate": 9.36322307775502e-06,
+      "loss": 1.1056,
+      "step": 2086
+    },
+    {
+      "epoch": 1.5795648060548722,
+      "grad_norm": 2.4794466495513916,
+      "learning_rate": 9.35521183008754e-06,
+      "loss": 1.1189,
+      "step": 2087
+    },
+    {
+      "epoch": 1.580321665089877,
+      "grad_norm": 2.0150938034057617,
+      "learning_rate": 9.347200998685261e-06,
+      "loss": 1.1063,
+      "step": 2088
+    },
+    {
+      "epoch": 1.5810785241248817,
+      "grad_norm": 2.3067104816436768,
+      "learning_rate": 9.339190588711852e-06,
+      "loss": 1.1081,
+      "step": 2089
+    },
+    {
+      "epoch": 1.5818353831598864,
+      "grad_norm": 2.071730613708496,
+      "learning_rate": 9.331180605330695e-06,
+      "loss": 1.1256,
+      "step": 2090
+    },
+    {
+      "epoch": 1.5825922421948913,
+      "grad_norm": 2.099440097808838,
+      "learning_rate": 9.323171053704904e-06,
+      "loss": 1.1306,
+      "step": 2091
+    },
+    {
+      "epoch": 1.583349101229896,
+      "grad_norm": 2.1519389152526855,
+      "learning_rate": 9.315161938997315e-06,
+      "loss": 1.1495,
+      "step": 2092
+    },
+    {
+      "epoch": 1.5841059602649006,
+      "grad_norm": 2.1621830463409424,
+      "learning_rate": 9.30715326637048e-06,
+      "loss": 1.1637,
+      "step": 2093
+    },
+    {
+      "epoch": 1.5848628192999055,
+      "grad_norm": 2.2661333084106445,
+      "learning_rate": 9.299145040986674e-06,
+      "loss": 1.1102,
+      "step": 2094
+    },
+    {
+      "epoch": 1.5856196783349101,
+      "grad_norm": 2.2131712436676025,
+      "learning_rate": 9.291137268007863e-06,
+      "loss": 1.1336,
+      "step": 2095
+    },
+    {
+      "epoch": 1.5863765373699148,
+      "grad_norm": 2.1026811599731445,
+      "learning_rate": 9.283129952595747e-06,
+      "loss": 1.0903,
+      "step": 2096
+    },
+    {
+      "epoch": 1.5871333964049197,
+      "grad_norm": 2.0890021324157715,
+      "learning_rate": 9.275123099911719e-06,
+      "loss": 1.1232,
+      "step": 2097
+    },
+    {
+      "epoch": 1.5878902554399243,
+      "grad_norm": 2.1274547576904297,
+      "learning_rate": 9.267116715116866e-06,
+      "loss": 1.1067,
+      "step": 2098
+    },
+    {
+      "epoch": 1.588647114474929,
+      "grad_norm": 2.193621873855591,
+      "learning_rate": 9.259110803371987e-06,
+      "loss": 1.1304,
+      "step": 2099
+    },
+    {
+      "epoch": 1.589403973509934,
+      "grad_norm": 2.1822807788848877,
+      "learning_rate": 9.251105369837574e-06,
+      "loss": 1.1568,
+      "step": 2100
+    },
+    {
+      "epoch": 1.5901608325449383,
+      "grad_norm": 1.883682131767273,
+      "learning_rate": 9.243100419673798e-06,
+      "loss": 1.129,
+      "step": 2101
+    },
+    {
+      "epoch": 1.5909176915799432,
+      "grad_norm": 2.0487449169158936,
+      "learning_rate": 9.235095958040535e-06,
+      "loss": 1.1436,
+      "step": 2102
+    },
+    {
+      "epoch": 1.5916745506149481,
+      "grad_norm": 2.079259157180786,
+      "learning_rate": 9.22709199009734e-06,
+      "loss": 1.1538,
+      "step": 2103
+    },
+    {
+      "epoch": 1.5924314096499526,
+      "grad_norm": 2.1335606575012207,
+      "learning_rate": 9.219088521003444e-06,
+      "loss": 1.1106,
+      "step": 2104
+    },
+    {
+      "epoch": 1.5931882686849574,
+      "grad_norm": 3.516350507736206,
+      "learning_rate": 9.211085555917764e-06,
+      "loss": 1.1,
+      "step": 2105
+    },
+    {
+      "epoch": 1.593945127719962,
+      "grad_norm": 2.226984739303589,
+      "learning_rate": 9.203083099998885e-06,
+      "loss": 1.149,
+      "step": 2106
+    },
+    {
+      "epoch": 1.5947019867549668,
+      "grad_norm": 2.139308452606201,
+      "learning_rate": 9.195081158405074e-06,
+      "loss": 1.1636,
+      "step": 2107
+    },
+    {
+      "epoch": 1.5954588457899717,
+      "grad_norm": 2.194244146347046,
+      "learning_rate": 9.187079736294258e-06,
+      "loss": 1.1441,
+      "step": 2108
+    },
+    {
+      "epoch": 1.5962157048249763,
+      "grad_norm": 2.349120855331421,
+      "learning_rate": 9.179078838824029e-06,
+      "loss": 1.1093,
+      "step": 2109
+    },
+    {
+      "epoch": 1.596972563859981,
+      "grad_norm": 2.0828843116760254,
+      "learning_rate": 9.171078471151646e-06,
+      "loss": 1.1357,
+      "step": 2110
+    },
+    {
+      "epoch": 1.5977294228949859,
+      "grad_norm": 2.10848331451416,
+      "learning_rate": 9.163078638434028e-06,
+      "loss": 1.1125,
+      "step": 2111
+    },
+    {
+      "epoch": 1.5984862819299905,
+      "grad_norm": 2.1101763248443604,
+      "learning_rate": 9.155079345827737e-06,
+      "loss": 1.0964,
+      "step": 2112
+    },
+    {
+      "epoch": 1.5992431409649952,
+      "grad_norm": 2.0837841033935547,
+      "learning_rate": 9.147080598488999e-06,
+      "loss": 1.1121,
+      "step": 2113
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9847743511199951,
+      "learning_rate": 9.13908240157369e-06,
+      "loss": 1.1153,
+      "step": 2114
+    },
+    {
+      "epoch": 1.6007568590350048,
+      "grad_norm": 2.0701873302459717,
+      "learning_rate": 9.131084760237314e-06,
+      "loss": 1.1046,
+      "step": 2115
+    },
+    {
+      "epoch": 1.6015137180700094,
+      "grad_norm": 2.1153228282928467,
+      "learning_rate": 9.123087679635039e-06,
+      "loss": 1.1126,
+      "step": 2116
+    },
+    {
+      "epoch": 1.6022705771050143,
+      "grad_norm": 2.0515284538269043,
+      "learning_rate": 9.115091164921654e-06,
+      "loss": 1.1266,
+      "step": 2117
+    },
+    {
+      "epoch": 1.603027436140019,
+      "grad_norm": 2.08406662940979,
+      "learning_rate": 9.107095221251597e-06,
+      "loss": 1.1355,
+      "step": 2118
+    },
+    {
+      "epoch": 1.6037842951750236,
+      "grad_norm": 2.6160190105438232,
+      "learning_rate": 9.099099853778927e-06,
+      "loss": 1.1226,
+      "step": 2119
+    },
+    {
+      "epoch": 1.6045411542100285,
+      "grad_norm": 2.024075984954834,
+      "learning_rate": 9.091105067657335e-06,
+      "loss": 1.0951,
+      "step": 2120
+    },
+    {
+      "epoch": 1.605298013245033,
+      "grad_norm": 1.9903373718261719,
+      "learning_rate": 9.083110868040142e-06,
+      "loss": 1.1485,
+      "step": 2121
+    },
+    {
+      "epoch": 1.6060548722800378,
+      "grad_norm": 2.307220458984375,
+      "learning_rate": 9.075117260080286e-06,
+      "loss": 1.1698,
+      "step": 2122
+    },
+    {
+      "epoch": 1.6068117313150427,
+      "grad_norm": 2.160867929458618,
+      "learning_rate": 9.067124248930324e-06,
+      "loss": 1.0852,
+      "step": 2123
+    },
+    {
+      "epoch": 1.6075685903500472,
+      "grad_norm": 2.255039691925049,
+      "learning_rate": 9.059131839742425e-06,
+      "loss": 1.1299,
+      "step": 2124
+    },
+    {
+      "epoch": 1.608325449385052,
+      "grad_norm": 2.122530698776245,
+      "learning_rate": 9.051140037668385e-06,
+      "loss": 1.1148,
+      "step": 2125
+    },
+    {
+      "epoch": 1.6090823084200567,
+      "grad_norm": 2.067059278488159,
+      "learning_rate": 9.043148847859588e-06,
+      "loss": 1.1335,
+      "step": 2126
+    },
+    {
+      "epoch": 1.6098391674550614,
+      "grad_norm": 3.3186850547790527,
+      "learning_rate": 9.035158275467037e-06,
+      "loss": 1.0892,
+      "step": 2127
+    },
+    {
+      "epoch": 1.6105960264900663,
+      "grad_norm": 2.4546923637390137,
+      "learning_rate": 9.02716832564133e-06,
+      "loss": 1.1939,
+      "step": 2128
+    },
+    {
+      "epoch": 1.611352885525071,
+      "grad_norm": 2.36734938621521,
+      "learning_rate": 9.01917900353267e-06,
+      "loss": 1.163,
+      "step": 2129
+    },
+    {
+      "epoch": 1.6121097445600756,
+      "grad_norm": 2.2100653648376465,
+      "learning_rate": 9.011190314290852e-06,
+      "loss": 1.0951,
+      "step": 2130
+    },
+    {
+      "epoch": 1.6128666035950805,
+      "grad_norm": 2.239097833633423,
+      "learning_rate": 9.003202263065263e-06,
+      "loss": 1.1554,
+      "step": 2131
+    },
+    {
+      "epoch": 1.6136234626300852,
+      "grad_norm": 2.2774319648742676,
+      "learning_rate": 8.995214855004877e-06,
+      "loss": 1.1237,
+      "step": 2132
+    },
+    {
+      "epoch": 1.6143803216650898,
+      "grad_norm": 2.1328752040863037,
+      "learning_rate": 8.987228095258256e-06,
+      "loss": 1.1154,
+      "step": 2133
+    },
+    {
+      "epoch": 1.6151371807000947,
+      "grad_norm": 2.3373916149139404,
+      "learning_rate": 8.979241988973546e-06,
+      "loss": 1.1058,
+      "step": 2134
+    },
+    {
+      "epoch": 1.6158940397350994,
+      "grad_norm": 2.126988172531128,
+      "learning_rate": 8.971256541298468e-06,
+      "loss": 1.1709,
+      "step": 2135
+    },
+    {
+      "epoch": 1.616650898770104,
+      "grad_norm": 2.3820157051086426,
+      "learning_rate": 8.963271757380319e-06,
+      "loss": 1.1332,
+      "step": 2136
+    },
+    {
+      "epoch": 1.617407757805109,
+      "grad_norm": 2.0995140075683594,
+      "learning_rate": 8.955287642365969e-06,
+      "loss": 1.1341,
+      "step": 2137
+    },
+    {
+      "epoch": 1.6181646168401134,
+      "grad_norm": 2.2463080883026123,
+      "learning_rate": 8.94730420140186e-06,
+      "loss": 1.1455,
+      "step": 2138
+    },
+    {
+      "epoch": 1.6189214758751183,
+      "grad_norm": 2.293729066848755,
+      "learning_rate": 8.939321439633991e-06,
+      "loss": 1.103,
+      "step": 2139
+    },
+    {
+      "epoch": 1.6196783349101231,
+      "grad_norm": 2.1475143432617188,
+      "learning_rate": 8.931339362207931e-06,
+      "loss": 1.124,
+      "step": 2140
+    },
+    {
+      "epoch": 1.6204351939451276,
+      "grad_norm": 2.087843179702759,
+      "learning_rate": 8.923357974268806e-06,
+      "loss": 1.1743,
+      "step": 2141
+    },
+    {
+      "epoch": 1.6211920529801325,
+      "grad_norm": 2.0908894538879395,
+      "learning_rate": 8.915377280961298e-06,
+      "loss": 1.0961,
+      "step": 2142
+    },
+    {
+      "epoch": 1.6219489120151371,
+      "grad_norm": 2.312263250350952,
+      "learning_rate": 8.907397287429635e-06,
+      "loss": 1.1523,
+      "step": 2143
+    },
+    {
+      "epoch": 1.6227057710501418,
+      "grad_norm": 2.30190110206604,
+      "learning_rate": 8.899417998817605e-06,
+      "loss": 1.1319,
+      "step": 2144
+    },
+    {
+      "epoch": 1.6234626300851467,
+      "grad_norm": 2.0427803993225098,
+      "learning_rate": 8.891439420268534e-06,
+      "loss": 1.1007,
+      "step": 2145
+    },
+    {
+      "epoch": 1.6242194891201513,
+      "grad_norm": 2.142066717147827,
+      "learning_rate": 8.88346155692529e-06,
+      "loss": 1.1465,
+      "step": 2146
+    },
+    {
+      "epoch": 1.624976348155156,
+      "grad_norm": 1.9997434616088867,
+      "learning_rate": 8.875484413930283e-06,
+      "loss": 1.0966,
+      "step": 2147
+    },
+    {
+      "epoch": 1.625733207190161,
+      "grad_norm": 2.0591166019439697,
+      "learning_rate": 8.86750799642546e-06,
+      "loss": 1.1409,
+      "step": 2148
+    },
+    {
+      "epoch": 1.6264900662251656,
+      "grad_norm": 2.044402599334717,
+      "learning_rate": 8.859532309552298e-06,
+      "loss": 1.1145,
+      "step": 2149
+    },
+    {
+      "epoch": 1.6272469252601702,
+      "grad_norm": 2.3767645359039307,
+      "learning_rate": 8.8515573584518e-06,
+      "loss": 1.1278,
+      "step": 2150
+    },
+    {
+      "epoch": 1.6280037842951751,
+      "grad_norm": 2.088170289993286,
+      "learning_rate": 8.843583148264496e-06,
+      "loss": 1.1607,
+      "step": 2151
+    },
+    {
+      "epoch": 1.6287606433301798,
+      "grad_norm": 2.026031255722046,
+      "learning_rate": 8.835609684130448e-06,
+      "loss": 1.1173,
+      "step": 2152
+    },
+    {
+      "epoch": 1.6295175023651844,
+      "grad_norm": 2.108065366744995,
+      "learning_rate": 8.827636971189222e-06,
+      "loss": 1.1735,
+      "step": 2153
+    },
+    {
+      "epoch": 1.6302743614001893,
+      "grad_norm": 2.12156081199646,
+      "learning_rate": 8.819665014579911e-06,
+      "loss": 1.0851,
+      "step": 2154
+    },
+    {
+      "epoch": 1.631031220435194,
+      "grad_norm": 2.0895984172821045,
+      "learning_rate": 8.81169381944112e-06,
+      "loss": 1.1261,
+      "step": 2155
+    },
+    {
+      "epoch": 1.6317880794701987,
+      "grad_norm": 2.119001865386963,
+      "learning_rate": 8.803723390910951e-06,
+      "loss": 1.1236,
+      "step": 2156
+    },
+    {
+      "epoch": 1.6325449385052035,
+      "grad_norm": 2.0929312705993652,
+      "learning_rate": 8.795753734127024e-06,
+      "loss": 1.157,
+      "step": 2157
+    },
+    {
+      "epoch": 1.633301797540208,
+      "grad_norm": 2.0275888442993164,
+      "learning_rate": 8.787784854226465e-06,
+      "loss": 1.1407,
+      "step": 2158
+    },
+    {
+      "epoch": 1.6340586565752129,
+      "grad_norm": 2.332402467727661,
+      "learning_rate": 8.779816756345884e-06,
+      "loss": 1.0916,
+      "step": 2159
+    },
+    {
+      "epoch": 1.6348155156102178,
+      "grad_norm": 2.0872373580932617,
+      "learning_rate": 8.7718494456214e-06,
+      "loss": 1.1492,
+      "step": 2160
+    },
+    {
+      "epoch": 1.6355723746452222,
+      "grad_norm": 2.1566085815429688,
+      "learning_rate": 8.763882927188615e-06,
+      "loss": 1.1397,
+      "step": 2161
+    },
+    {
+      "epoch": 1.636329233680227,
+      "grad_norm": 2.134572744369507,
+      "learning_rate": 8.75591720618263e-06,
+      "loss": 1.0967,
+      "step": 2162
+    },
+    {
+      "epoch": 1.6370860927152318,
+      "grad_norm": 2.061708450317383,
+      "learning_rate": 8.74795228773803e-06,
+      "loss": 1.0709,
+      "step": 2163
+    },
+    {
+      "epoch": 1.6378429517502364,
+      "grad_norm": 2.04203724861145,
+      "learning_rate": 8.739988176988869e-06,
+      "loss": 1.0671,
+      "step": 2164
+    },
+    {
+      "epoch": 1.6385998107852413,
+      "grad_norm": 2.0386204719543457,
+      "learning_rate": 8.732024879068702e-06,
+      "loss": 1.1021,
+      "step": 2165
+    },
+    {
+      "epoch": 1.639356669820246,
+      "grad_norm": 2.104109764099121,
+      "learning_rate": 8.724062399110547e-06,
+      "loss": 1.0964,
+      "step": 2166
+    },
+    {
+      "epoch": 1.6401135288552506,
+      "grad_norm": 2.078735113143921,
+      "learning_rate": 8.716100742246894e-06,
+      "loss": 1.1241,
+      "step": 2167
+    },
+    {
+      "epoch": 1.6408703878902555,
+      "grad_norm": 2.1530871391296387,
+      "learning_rate": 8.708139913609705e-06,
+      "loss": 1.118,
+      "step": 2168
+    },
+    {
+      "epoch": 1.6416272469252602,
+      "grad_norm": 2.1105563640594482,
+      "learning_rate": 8.700179918330419e-06,
+      "loss": 1.0883,
+      "step": 2169
+    },
+    {
+      "epoch": 1.6423841059602649,
+      "grad_norm": 2.056195020675659,
+      "learning_rate": 8.692220761539912e-06,
+      "loss": 1.1549,
+      "step": 2170
+    },
+    {
+      "epoch": 1.6431409649952697,
+      "grad_norm": 2.327533006668091,
+      "learning_rate": 8.684262448368546e-06,
+      "loss": 1.1097,
+      "step": 2171
+    },
+    {
+      "epoch": 1.6438978240302744,
+      "grad_norm": 2.111985206604004,
+      "learning_rate": 8.676304983946122e-06,
+      "loss": 1.1048,
+      "step": 2172
+    },
+    {
+      "epoch": 1.644654683065279,
+      "grad_norm": 2.1778697967529297,
+      "learning_rate": 8.668348373401908e-06,
+      "loss": 1.1644,
+      "step": 2173
+    },
+    {
+      "epoch": 1.645411542100284,
+      "grad_norm": 2.3045222759246826,
+      "learning_rate": 8.660392621864608e-06,
+      "loss": 1.1873,
+      "step": 2174
+    },
+    {
+      "epoch": 1.6461684011352884,
+      "grad_norm": 2.2667534351348877,
+      "learning_rate": 8.652437734462377e-06,
+      "loss": 1.0519,
+      "step": 2175
+    },
+    {
+      "epoch": 1.6469252601702933,
+      "grad_norm": 2.394404888153076,
+      "learning_rate": 8.644483716322818e-06,
+      "loss": 1.1324,
+      "step": 2176
+    },
+    {
+      "epoch": 1.6476821192052982,
+      "grad_norm": 2.009328603744507,
+      "learning_rate": 8.63653057257297e-06,
+      "loss": 1.147,
+      "step": 2177
+    },
+    {
+      "epoch": 1.6484389782403026,
+      "grad_norm": 2.072662591934204,
+      "learning_rate": 8.6285783083393e-06,
+      "loss": 1.1435,
+      "step": 2178
+    },
+    {
+      "epoch": 1.6491958372753075,
+      "grad_norm": 2.184267044067383,
+      "learning_rate": 8.620626928747725e-06,
+      "loss": 1.1896,
+      "step": 2179
+    },
+    {
+      "epoch": 1.6499526963103122,
+      "grad_norm": 2.0765388011932373,
+      "learning_rate": 8.612676438923587e-06,
+      "loss": 1.125,
+      "step": 2180
+    },
+    {
+      "epoch": 1.6507095553453168,
+      "grad_norm": 2.2719030380249023,
+      "learning_rate": 8.604726843991637e-06,
+      "loss": 1.074,
+      "step": 2181
+    },
+    {
+      "epoch": 1.6514664143803217,
+      "grad_norm": 2.0793111324310303,
+      "learning_rate": 8.596778149076073e-06,
+      "loss": 1.099,
+      "step": 2182
+    },
+    {
+      "epoch": 1.6522232734153264,
+      "grad_norm": 1.9967988729476929,
+      "learning_rate": 8.588830359300499e-06,
+      "loss": 1.1454,
+      "step": 2183
+    },
+    {
+      "epoch": 1.652980132450331,
+      "grad_norm": 2.3746399879455566,
+      "learning_rate": 8.580883479787936e-06,
+      "loss": 1.1424,
+      "step": 2184
+    },
+    {
+      "epoch": 1.653736991485336,
+      "grad_norm": 2.0627639293670654,
+      "learning_rate": 8.57293751566083e-06,
+      "loss": 1.1052,
+      "step": 2185
+    },
+    {
+      "epoch": 1.6544938505203406,
+      "grad_norm": 2.196162700653076,
+      "learning_rate": 8.564992472041021e-06,
+      "loss": 1.1002,
+      "step": 2186
+    },
+    {
+      "epoch": 1.6552507095553453,
+      "grad_norm": 2.3567469120025635,
+      "learning_rate": 8.557048354049763e-06,
+      "loss": 1.1426,
+      "step": 2187
+    },
+    {
+      "epoch": 1.6560075685903501,
+      "grad_norm": 2.0467660427093506,
+      "learning_rate": 8.549105166807716e-06,
+      "loss": 1.0916,
+      "step": 2188
+    },
+    {
+      "epoch": 1.6567644276253548,
+      "grad_norm": 2.107483386993408,
+      "learning_rate": 8.541162915434935e-06,
+      "loss": 1.1226,
+      "step": 2189
+    },
+    {
+      "epoch": 1.6575212866603595,
+      "grad_norm": 2.2864937782287598,
+      "learning_rate": 8.533221605050878e-06,
+      "loss": 1.1246,
+      "step": 2190
+    },
+    {
+      "epoch": 1.6582781456953644,
+      "grad_norm": 2.135864734649658,
+      "learning_rate": 8.525281240774391e-06,
+      "loss": 1.0364,
+      "step": 2191
+    },
+    {
+      "epoch": 1.659035004730369,
+      "grad_norm": 2.136951446533203,
+      "learning_rate": 8.517341827723709e-06,
+      "loss": 1.1753,
+      "step": 2192
+    },
+    {
+      "epoch": 1.6597918637653737,
+      "grad_norm": 2.1684775352478027,
+      "learning_rate": 8.509403371016462e-06,
+      "loss": 1.0812,
+      "step": 2193
+    },
+    {
+      "epoch": 1.6605487228003786,
+      "grad_norm": 2.195051431655884,
+      "learning_rate": 8.501465875769652e-06,
+      "loss": 1.0883,
+      "step": 2194
+    },
+    {
+      "epoch": 1.661305581835383,
+      "grad_norm": 2.2236487865448,
+      "learning_rate": 8.493529347099669e-06,
+      "loss": 1.0637,
+      "step": 2195
+    },
+    {
+      "epoch": 1.662062440870388,
+      "grad_norm": 2.1652839183807373,
+      "learning_rate": 8.48559379012228e-06,
+      "loss": 1.1285,
+      "step": 2196
+    },
+    {
+      "epoch": 1.6628192999053926,
+      "grad_norm": 2.3312926292419434,
+      "learning_rate": 8.477659209952627e-06,
+      "loss": 1.1303,
+      "step": 2197
+    },
+    {
+      "epoch": 1.6635761589403972,
+      "grad_norm": 2.1476340293884277,
+      "learning_rate": 8.46972561170521e-06,
+      "loss": 1.0911,
+      "step": 2198
+    },
+    {
+      "epoch": 1.6643330179754021,
+      "grad_norm": 2.205261707305908,
+      "learning_rate": 8.461793000493917e-06,
+      "loss": 1.1075,
+      "step": 2199
+    },
+    {
+      "epoch": 1.6650898770104068,
+      "grad_norm": 2.3584821224212646,
+      "learning_rate": 8.45386138143198e-06,
+      "loss": 1.1429,
+      "step": 2200
+    },
+    {
+      "epoch": 1.6658467360454114,
+      "grad_norm": 1.990213394165039,
+      "learning_rate": 8.445930759632e-06,
+      "loss": 1.0731,
+      "step": 2201
+    },
+    {
+      "epoch": 1.6666035950804163,
+      "grad_norm": 2.114382266998291,
+      "learning_rate": 8.43800114020594e-06,
+      "loss": 1.1304,
+      "step": 2202
+    },
+    {
+      "epoch": 1.667360454115421,
+      "grad_norm": 2.6425230503082275,
+      "learning_rate": 8.430072528265107e-06,
+      "loss": 1.1223,
+      "step": 2203
+    },
+    {
+      "epoch": 1.6681173131504257,
+      "grad_norm": 2.238675594329834,
+      "learning_rate": 8.422144928920168e-06,
+      "loss": 1.1187,
+      "step": 2204
+    },
+    {
+      "epoch": 1.6688741721854305,
+      "grad_norm": 2.0409348011016846,
+      "learning_rate": 8.414218347281127e-06,
+      "loss": 1.0912,
+      "step": 2205
+    },
+    {
+      "epoch": 1.6696310312204352,
+      "grad_norm": 2.5583693981170654,
+      "learning_rate": 8.406292788457338e-06,
+      "loss": 1.1433,
+      "step": 2206
+    },
+    {
+      "epoch": 1.6703878902554399,
+      "grad_norm": 2.24996018409729,
+      "learning_rate": 8.398368257557505e-06,
+      "loss": 1.1177,
+      "step": 2207
+    },
+    {
+      "epoch": 1.6711447492904448,
+      "grad_norm": 2.2110371589660645,
+      "learning_rate": 8.390444759689646e-06,
+      "loss": 1.1334,
+      "step": 2208
+    },
+    {
+      "epoch": 1.6719016083254494,
+      "grad_norm": 2.0102930068969727,
+      "learning_rate": 8.382522299961135e-06,
+      "loss": 1.0807,
+      "step": 2209
+    },
+    {
+      "epoch": 1.672658467360454,
+      "grad_norm": 2.319051504135132,
+      "learning_rate": 8.37460088347867e-06,
+      "loss": 1.0997,
+      "step": 2210
+    },
+    {
+      "epoch": 1.673415326395459,
+      "grad_norm": 2.3122832775115967,
+      "learning_rate": 8.36668051534827e-06,
+      "loss": 1.1591,
+      "step": 2211
+    },
+    {
+      "epoch": 1.6741721854304634,
+      "grad_norm": 2.39446759223938,
+      "learning_rate": 8.358761200675284e-06,
+      "loss": 1.1201,
+      "step": 2212
+    },
+    {
+      "epoch": 1.6749290444654683,
+      "grad_norm": 2.257894515991211,
+      "learning_rate": 8.350842944564386e-06,
+      "loss": 1.1094,
+      "step": 2213
+    },
+    {
+      "epoch": 1.6756859035004732,
+      "grad_norm": 2.2300925254821777,
+      "learning_rate": 8.34292575211956e-06,
+      "loss": 1.1613,
+      "step": 2214
+    },
+    {
+      "epoch": 1.6764427625354776,
+      "grad_norm": 2.2363715171813965,
+      "learning_rate": 8.33500962844411e-06,
+      "loss": 1.1485,
+      "step": 2215
+    },
+    {
+      "epoch": 1.6771996215704825,
+      "grad_norm": 2.0160231590270996,
+      "learning_rate": 8.327094578640643e-06,
+      "loss": 1.1136,
+      "step": 2216
+    },
+    {
+      "epoch": 1.6779564806054872,
+      "grad_norm": 2.279360294342041,
+      "learning_rate": 8.319180607811085e-06,
+      "loss": 1.1193,
+      "step": 2217
+    },
+    {
+      "epoch": 1.6787133396404919,
+      "grad_norm": 2.280641555786133,
+      "learning_rate": 8.31126772105666e-06,
+      "loss": 1.1655,
+      "step": 2218
+    },
+    {
+      "epoch": 1.6794701986754967,
+      "grad_norm": 2.077263832092285,
+      "learning_rate": 8.303355923477889e-06,
+      "loss": 1.1435,
+      "step": 2219
+    },
+    {
+      "epoch": 1.6802270577105014,
+      "grad_norm": 2.271101236343384,
+      "learning_rate": 8.295445220174604e-06,
+      "loss": 1.0986,
+      "step": 2220
+    },
+    {
+      "epoch": 1.680983916745506,
+      "grad_norm": 2.257680892944336,
+      "learning_rate": 8.28753561624592e-06,
+      "loss": 1.095,
+      "step": 2221
+    },
+    {
+      "epoch": 1.681740775780511,
+      "grad_norm": 2.18369722366333,
+      "learning_rate": 8.279627116790244e-06,
+      "loss": 1.1007,
+      "step": 2222
+    },
+    {
+      "epoch": 1.6824976348155156,
+      "grad_norm": 2.0165445804595947,
+      "learning_rate": 8.271719726905275e-06,
+      "loss": 1.1165,
+      "step": 2223
+    },
+    {
+      "epoch": 1.6832544938505203,
+      "grad_norm": 2.2388248443603516,
+      "learning_rate": 8.263813451688006e-06,
+      "loss": 1.1186,
+      "step": 2224
+    },
+    {
+      "epoch": 1.6840113528855252,
+      "grad_norm": 2.32745099067688,
+      "learning_rate": 8.255908296234688e-06,
+      "loss": 1.1812,
+      "step": 2225
+    },
+    {
+      "epoch": 1.6847682119205298,
+      "grad_norm": 2.361375093460083,
+      "learning_rate": 8.248004265640868e-06,
+      "loss": 1.1649,
+      "step": 2226
+    },
+    {
+      "epoch": 1.6855250709555345,
+      "grad_norm": 2.2414417266845703,
+      "learning_rate": 8.240101365001368e-06,
+      "loss": 1.1013,
+      "step": 2227
+    },
+    {
+      "epoch": 1.6862819299905394,
+      "grad_norm": 1.9859745502471924,
+      "learning_rate": 8.232199599410273e-06,
+      "loss": 1.1295,
+      "step": 2228
+    },
+    {
+      "epoch": 1.6870387890255438,
+      "grad_norm": 2.1757733821868896,
+      "learning_rate": 8.22429897396094e-06,
+      "loss": 1.1223,
+      "step": 2229
+    },
+    {
+      "epoch": 1.6877956480605487,
+      "grad_norm": 2.36989164352417,
+      "learning_rate": 8.216399493745992e-06,
+      "loss": 1.1337,
+      "step": 2230
+    },
+    {
+      "epoch": 1.6885525070955536,
+      "grad_norm": 1.9508718252182007,
+      "learning_rate": 8.208501163857318e-06,
+      "loss": 1.1351,
+      "step": 2231
+    },
+    {
+      "epoch": 1.689309366130558,
+      "grad_norm": 2.057548999786377,
+      "learning_rate": 8.200603989386055e-06,
+      "loss": 1.1382,
+      "step": 2232
+    },
+    {
+      "epoch": 1.690066225165563,
+      "grad_norm": 2.3355371952056885,
+      "learning_rate": 8.192707975422604e-06,
+      "loss": 1.1393,
+      "step": 2233
+    },
+    {
+      "epoch": 1.6908230842005676,
+      "grad_norm": 2.1525468826293945,
+      "learning_rate": 8.184813127056616e-06,
+      "loss": 1.1665,
+      "step": 2234
+    },
+    {
+      "epoch": 1.6915799432355723,
+      "grad_norm": 2.2342827320098877,
+      "learning_rate": 8.176919449376989e-06,
+      "loss": 1.1385,
+      "step": 2235
+    },
+    {
+      "epoch": 1.6923368022705771,
+      "grad_norm": 2.1949453353881836,
+      "learning_rate": 8.169026947471866e-06,
+      "loss": 1.1635,
+      "step": 2236
+    },
+    {
+      "epoch": 1.6930936613055818,
+      "grad_norm": 2.183218240737915,
+      "learning_rate": 8.161135626428633e-06,
+      "loss": 1.1433,
+      "step": 2237
+    },
+    {
+      "epoch": 1.6938505203405865,
+      "grad_norm": 2.1611173152923584,
+      "learning_rate": 8.153245491333922e-06,
+      "loss": 1.1083,
+      "step": 2238
+    },
+    {
+      "epoch": 1.6946073793755914,
+      "grad_norm": 2.0343189239501953,
+      "learning_rate": 8.145356547273584e-06,
+      "loss": 1.1334,
+      "step": 2239
+    },
+    {
+      "epoch": 1.695364238410596,
+      "grad_norm": 2.0303726196289062,
+      "learning_rate": 8.13746879933272e-06,
+      "loss": 1.1545,
+      "step": 2240
+    },
+    {
+      "epoch": 1.6961210974456007,
+      "grad_norm": 2.152381658554077,
+      "learning_rate": 8.129582252595645e-06,
+      "loss": 1.1316,
+      "step": 2241
+    },
+    {
+      "epoch": 1.6968779564806056,
+      "grad_norm": 2.0682196617126465,
+      "learning_rate": 8.12169691214591e-06,
+      "loss": 1.1396,
+      "step": 2242
+    },
+    {
+      "epoch": 1.6976348155156102,
+      "grad_norm": 2.07098388671875,
+      "learning_rate": 8.113812783066288e-06,
+      "loss": 1.0784,
+      "step": 2243
+    },
+    {
+      "epoch": 1.698391674550615,
+      "grad_norm": 2.1533405780792236,
+      "learning_rate": 8.105929870438762e-06,
+      "loss": 1.1151,
+      "step": 2244
+    },
+    {
+      "epoch": 1.6991485335856198,
+      "grad_norm": 2.348604679107666,
+      "learning_rate": 8.098048179344545e-06,
+      "loss": 1.0913,
+      "step": 2245
+    },
+    {
+      "epoch": 1.6999053926206245,
+      "grad_norm": 2.0196192264556885,
+      "learning_rate": 8.090167714864051e-06,
+      "loss": 1.1334,
+      "step": 2246
+    },
+    {
+      "epoch": 1.7006622516556291,
+      "grad_norm": 2.2407851219177246,
+      "learning_rate": 8.082288482076904e-06,
+      "loss": 1.1362,
+      "step": 2247
+    },
+    {
+      "epoch": 1.701419110690634,
+      "grad_norm": 2.0429224967956543,
+      "learning_rate": 8.074410486061943e-06,
+      "loss": 1.0377,
+      "step": 2248
+    },
+    {
+      "epoch": 1.7021759697256384,
+      "grad_norm": 2.149394989013672,
+      "learning_rate": 8.066533731897202e-06,
+      "loss": 1.1324,
+      "step": 2249
+    },
+    {
+      "epoch": 1.7029328287606433,
+      "grad_norm": 2.5104711055755615,
+      "learning_rate": 8.058658224659914e-06,
+      "loss": 1.1037,
+      "step": 2250
+    },
+    {
+      "epoch": 1.7036896877956482,
+      "grad_norm": 1.9572622776031494,
+      "learning_rate": 8.050783969426517e-06,
+      "loss": 1.1231,
+      "step": 2251
+    },
+    {
+      "epoch": 1.7044465468306527,
+      "grad_norm": 2.344362497329712,
+      "learning_rate": 8.042910971272627e-06,
+      "loss": 1.1054,
+      "step": 2252
+    },
+    {
+      "epoch": 1.7052034058656576,
+      "grad_norm": 2.160923719406128,
+      "learning_rate": 8.035039235273063e-06,
+      "loss": 1.1429,
+      "step": 2253
+    },
+    {
+      "epoch": 1.7059602649006622,
+      "grad_norm": 2.273373603820801,
+      "learning_rate": 8.027168766501831e-06,
+      "loss": 1.1073,
+      "step": 2254
+    },
+    {
+      "epoch": 1.7067171239356669,
+      "grad_norm": 2.043576955795288,
+      "learning_rate": 8.019299570032108e-06,
+      "loss": 1.1326,
+      "step": 2255
+    },
+    {
+      "epoch": 1.7074739829706718,
+      "grad_norm": 2.3227038383483887,
+      "learning_rate": 8.011431650936259e-06,
+      "loss": 1.0762,
+      "step": 2256
+    },
+    {
+      "epoch": 1.7082308420056764,
+      "grad_norm": 2.2618634700775146,
+      "learning_rate": 8.003565014285829e-06,
+      "loss": 1.1246,
+      "step": 2257
+    },
+    {
+      "epoch": 1.708987701040681,
+      "grad_norm": 2.072833776473999,
+      "learning_rate": 7.99569966515153e-06,
+      "loss": 1.128,
+      "step": 2258
+    },
+    {
+      "epoch": 1.709744560075686,
+      "grad_norm": 2.305095911026001,
+      "learning_rate": 7.987835608603241e-06,
+      "loss": 1.0533,
+      "step": 2259
+    },
+    {
+      "epoch": 1.7105014191106906,
+      "grad_norm": 2.25026535987854,
+      "learning_rate": 7.979972849710022e-06,
+      "loss": 1.1497,
+      "step": 2260
+    },
+    {
+      "epoch": 1.7112582781456953,
+      "grad_norm": 2.259713888168335,
+      "learning_rate": 7.972111393540079e-06,
+      "loss": 1.1364,
+      "step": 2261
+    },
+    {
+      "epoch": 1.7120151371807002,
+      "grad_norm": 2.3140814304351807,
+      "learning_rate": 7.964251245160795e-06,
+      "loss": 1.1363,
+      "step": 2262
+    },
+    {
+      "epoch": 1.7127719962157049,
+      "grad_norm": 2.25529408454895,
+      "learning_rate": 7.956392409638693e-06,
+      "loss": 1.1304,
+      "step": 2263
+    },
+    {
+      "epoch": 1.7135288552507095,
+      "grad_norm": 2.118211030960083,
+      "learning_rate": 7.948534892039462e-06,
+      "loss": 1.1227,
+      "step": 2264
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 2.260540008544922,
+      "learning_rate": 7.940678697427945e-06,
+      "loss": 1.0716,
+      "step": 2265
+    },
+    {
+      "epoch": 1.7150425733207189,
+      "grad_norm": 2.334322690963745,
+      "learning_rate": 7.932823830868114e-06,
+      "loss": 1.1458,
+      "step": 2266
+    },
+    {
+      "epoch": 1.7157994323557237,
+      "grad_norm": 2.1086597442626953,
+      "learning_rate": 7.9249702974231e-06,
+      "loss": 1.1264,
+      "step": 2267
+    },
+    {
+      "epoch": 1.7165562913907286,
+      "grad_norm": 2.286928176879883,
+      "learning_rate": 7.917118102155175e-06,
+      "loss": 1.1079,
+      "step": 2268
+    },
+    {
+      "epoch": 1.717313150425733,
+      "grad_norm": 2.132174491882324,
+      "learning_rate": 7.909267250125743e-06,
+      "loss": 1.1201,
+      "step": 2269
+    },
+    {
+      "epoch": 1.718070009460738,
+      "grad_norm": 2.0687386989593506,
+      "learning_rate": 7.901417746395338e-06,
+      "loss": 1.0981,
+      "step": 2270
+    },
+    {
+      "epoch": 1.7188268684957426,
+      "grad_norm": 2.2579548358917236,
+      "learning_rate": 7.893569596023638e-06,
+      "loss": 1.1193,
+      "step": 2271
+    },
+    {
+      "epoch": 1.7195837275307473,
+      "grad_norm": 2.1385035514831543,
+      "learning_rate": 7.885722804069435e-06,
+      "loss": 1.1826,
+      "step": 2272
+    },
+    {
+      "epoch": 1.7203405865657522,
+      "grad_norm": 1.9475092887878418,
+      "learning_rate": 7.877877375590657e-06,
+      "loss": 1.101,
+      "step": 2273
+    },
+    {
+      "epoch": 1.7210974456007568,
+      "grad_norm": 2.2321999073028564,
+      "learning_rate": 7.87003331564434e-06,
+      "loss": 1.121,
+      "step": 2274
+    },
+    {
+      "epoch": 1.7218543046357615,
+      "grad_norm": 2.091757297515869,
+      "learning_rate": 7.86219062928665e-06,
+      "loss": 1.0453,
+      "step": 2275
+    },
+    {
+      "epoch": 1.7226111636707664,
+      "grad_norm": 1.965161681175232,
+      "learning_rate": 7.854349321572868e-06,
+      "loss": 1.1628,
+      "step": 2276
+    },
+    {
+      "epoch": 1.723368022705771,
+      "grad_norm": 2.090461492538452,
+      "learning_rate": 7.846509397557372e-06,
+      "loss": 1.1603,
+      "step": 2277
+    },
+    {
+      "epoch": 1.7241248817407757,
+      "grad_norm": 2.0913045406341553,
+      "learning_rate": 7.83867086229366e-06,
+      "loss": 1.186,
+      "step": 2278
+    },
+    {
+      "epoch": 1.7248817407757806,
+      "grad_norm": 2.137932300567627,
+      "learning_rate": 7.83083372083434e-06,
+      "loss": 1.129,
+      "step": 2279
+    },
+    {
+      "epoch": 1.7256385998107853,
+      "grad_norm": 2.1641454696655273,
+      "learning_rate": 7.822997978231101e-06,
+      "loss": 1.1261,
+      "step": 2280
+    },
+    {
+      "epoch": 1.72639545884579,
+      "grad_norm": 1.977129340171814,
+      "learning_rate": 7.815163639534752e-06,
+      "loss": 1.1332,
+      "step": 2281
+    },
+    {
+      "epoch": 1.7271523178807948,
+      "grad_norm": 2.3945367336273193,
+      "learning_rate": 7.807330709795191e-06,
+      "loss": 1.0711,
+      "step": 2282
+    },
+    {
+      "epoch": 1.7279091769157995,
+      "grad_norm": 1.9358054399490356,
+      "learning_rate": 7.799499194061395e-06,
+      "loss": 1.0986,
+      "step": 2283
+    },
+    {
+      "epoch": 1.7286660359508041,
+      "grad_norm": 2.242386817932129,
+      "learning_rate": 7.791669097381447e-06,
+      "loss": 1.1113,
+      "step": 2284
+    },
+    {
+      "epoch": 1.729422894985809,
+      "grad_norm": 2.1633381843566895,
+      "learning_rate": 7.783840424802504e-06,
+      "loss": 1.1581,
+      "step": 2285
+    },
+    {
+      "epoch": 1.7301797540208135,
+      "grad_norm": 2.0929739475250244,
+      "learning_rate": 7.776013181370813e-06,
+      "loss": 1.1479,
+      "step": 2286
+    },
+    {
+      "epoch": 1.7309366130558184,
+      "grad_norm": 2.2305848598480225,
+      "learning_rate": 7.768187372131693e-06,
+      "loss": 1.1683,
+      "step": 2287
+    },
+    {
+      "epoch": 1.7316934720908232,
+      "grad_norm": 1.938439130783081,
+      "learning_rate": 7.76036300212954e-06,
+      "loss": 1.0774,
+      "step": 2288
+    },
+    {
+      "epoch": 1.7324503311258277,
+      "grad_norm": 2.031890869140625,
+      "learning_rate": 7.752540076407829e-06,
+      "loss": 1.0838,
+      "step": 2289
+    },
+    {
+      "epoch": 1.7332071901608326,
+      "grad_norm": 2.242338180541992,
+      "learning_rate": 7.744718600009093e-06,
+      "loss": 1.1161,
+      "step": 2290
+    },
+    {
+      "epoch": 1.7339640491958372,
+      "grad_norm": 2.1008989810943604,
+      "learning_rate": 7.736898577974936e-06,
+      "loss": 1.071,
+      "step": 2291
+    },
+    {
+      "epoch": 1.734720908230842,
+      "grad_norm": 1.9958288669586182,
+      "learning_rate": 7.72908001534603e-06,
+      "loss": 1.0919,
+      "step": 2292
+    },
+    {
+      "epoch": 1.7354777672658468,
+      "grad_norm": 2.2547950744628906,
+      "learning_rate": 7.7212629171621e-06,
+      "loss": 1.1024,
+      "step": 2293
+    },
+    {
+      "epoch": 1.7362346263008515,
+      "grad_norm": 2.198230743408203,
+      "learning_rate": 7.713447288461922e-06,
+      "loss": 1.1567,
+      "step": 2294
+    },
+    {
+      "epoch": 1.7369914853358561,
+      "grad_norm": 2.047135591506958,
+      "learning_rate": 7.705633134283342e-06,
+      "loss": 1.1345,
+      "step": 2295
+    },
+    {
+      "epoch": 1.737748344370861,
+      "grad_norm": 2.021092176437378,
+      "learning_rate": 7.697820459663234e-06,
+      "loss": 1.0968,
+      "step": 2296
+    },
+    {
+      "epoch": 1.7385052034058657,
+      "grad_norm": 2.04164719581604,
+      "learning_rate": 7.690009269637535e-06,
+      "loss": 1.1234,
+      "step": 2297
+    },
+    {
+      "epoch": 1.7392620624408703,
+      "grad_norm": 2.0042457580566406,
+      "learning_rate": 7.68219956924122e-06,
+      "loss": 1.1361,
+      "step": 2298
+    },
+    {
+      "epoch": 1.7400189214758752,
+      "grad_norm": 2.207336902618408,
+      "learning_rate": 7.674391363508293e-06,
+      "loss": 1.1241,
+      "step": 2299
+    },
+    {
+      "epoch": 1.7407757805108799,
+      "grad_norm": 2.4397289752960205,
+      "learning_rate": 7.666584657471819e-06,
+      "loss": 1.1499,
+      "step": 2300
+    },
+    {
+      "epoch": 1.7415326395458846,
+      "grad_norm": 1.9008210897445679,
+      "learning_rate": 7.65877945616387e-06,
+      "loss": 1.075,
+      "step": 2301
+    },
+    {
+      "epoch": 1.7422894985808894,
+      "grad_norm": 2.0731241703033447,
+      "learning_rate": 7.650975764615564e-06,
+      "loss": 1.1104,
+      "step": 2302
+    },
+    {
+      "epoch": 1.7430463576158939,
+      "grad_norm": 2.0302274227142334,
+      "learning_rate": 7.643173587857043e-06,
+      "loss": 1.1129,
+      "step": 2303
+    },
+    {
+      "epoch": 1.7438032166508988,
+      "grad_norm": 2.3482978343963623,
+      "learning_rate": 7.635372930917471e-06,
+      "loss": 1.1239,
+      "step": 2304
+    },
+    {
+      "epoch": 1.7445600756859037,
+      "grad_norm": 2.1537070274353027,
+      "learning_rate": 7.627573798825028e-06,
+      "loss": 1.1213,
+      "step": 2305
+    },
+    {
+      "epoch": 1.745316934720908,
+      "grad_norm": 3.2694427967071533,
+      "learning_rate": 7.619776196606923e-06,
+      "loss": 1.1201,
+      "step": 2306
+    },
+    {
+      "epoch": 1.746073793755913,
+      "grad_norm": 1.9938985109329224,
+      "learning_rate": 7.611980129289362e-06,
+      "loss": 1.1202,
+      "step": 2307
+    },
+    {
+      "epoch": 1.7468306527909176,
+      "grad_norm": 1.9866852760314941,
+      "learning_rate": 7.604185601897578e-06,
+      "loss": 1.1716,
+      "step": 2308
+    },
+    {
+      "epoch": 1.7475875118259223,
+      "grad_norm": 2.149052143096924,
+      "learning_rate": 7.596392619455805e-06,
+      "loss": 1.0911,
+      "step": 2309
+    },
+    {
+      "epoch": 1.7483443708609272,
+      "grad_norm": 2.1124277114868164,
+      "learning_rate": 7.588601186987277e-06,
+      "loss": 1.1686,
+      "step": 2310
+    },
+    {
+      "epoch": 1.7491012298959319,
+      "grad_norm": 2.1678647994995117,
+      "learning_rate": 7.5808113095142334e-06,
+      "loss": 1.1356,
+      "step": 2311
+    },
+    {
+      "epoch": 1.7498580889309365,
+      "grad_norm": 2.2444238662719727,
+      "learning_rate": 7.573022992057911e-06,
+      "loss": 1.1302,
+      "step": 2312
+    },
+    {
+      "epoch": 1.7506149479659414,
+      "grad_norm": 2.296766757965088,
+      "learning_rate": 7.565236239638542e-06,
+      "loss": 1.1317,
+      "step": 2313
+    },
+    {
+      "epoch": 1.751371807000946,
+      "grad_norm": 2.106170177459717,
+      "learning_rate": 7.557451057275346e-06,
+      "loss": 1.138,
+      "step": 2314
+    },
+    {
+      "epoch": 1.7521286660359507,
+      "grad_norm": 1.8964704275131226,
+      "learning_rate": 7.549667449986533e-06,
+      "loss": 1.1121,
+      "step": 2315
+    },
+    {
+      "epoch": 1.7528855250709556,
+      "grad_norm": 2.2497787475585938,
+      "learning_rate": 7.541885422789297e-06,
+      "loss": 1.1607,
+      "step": 2316
+    },
+    {
+      "epoch": 1.7536423841059603,
+      "grad_norm": 1.980980634689331,
+      "learning_rate": 7.53410498069982e-06,
+      "loss": 1.0806,
+      "step": 2317
+    },
+    {
+      "epoch": 1.754399243140965,
+      "grad_norm": 2.030378818511963,
+      "learning_rate": 7.526326128733247e-06,
+      "loss": 1.1,
+      "step": 2318
+    },
+    {
+      "epoch": 1.7551561021759698,
+      "grad_norm": 1.9196511507034302,
+      "learning_rate": 7.5185488719037105e-06,
+      "loss": 1.1771,
+      "step": 2319
+    },
+    {
+      "epoch": 1.7559129612109745,
+      "grad_norm": 1.9450955390930176,
+      "learning_rate": 7.510773215224318e-06,
+      "loss": 1.1347,
+      "step": 2320
+    },
+    {
+      "epoch": 1.7566698202459792,
+      "grad_norm": 1.9371559619903564,
+      "learning_rate": 7.502999163707131e-06,
+      "loss": 1.1026,
+      "step": 2321
+    },
+    {
+      "epoch": 1.757426679280984,
+      "grad_norm": 2.021090269088745,
+      "learning_rate": 7.4952267223631865e-06,
+      "loss": 1.1514,
+      "step": 2322
+    },
+    {
+      "epoch": 1.7581835383159885,
+      "grad_norm": 2.016483783721924,
+      "learning_rate": 7.487455896202487e-06,
+      "loss": 1.1315,
+      "step": 2323
+    },
+    {
+      "epoch": 1.7589403973509934,
+      "grad_norm": 2.051363945007324,
+      "learning_rate": 7.479686690233981e-06,
+      "loss": 1.1094,
+      "step": 2324
+    },
+    {
+      "epoch": 1.759697256385998,
+      "grad_norm": 2.3509905338287354,
+      "learning_rate": 7.471919109465584e-06,
+      "loss": 1.1104,
+      "step": 2325
+    },
+    {
+      "epoch": 1.7604541154210027,
+      "grad_norm": 2.0247390270233154,
+      "learning_rate": 7.46415315890416e-06,
+      "loss": 1.1122,
+      "step": 2326
+    },
+    {
+      "epoch": 1.7612109744560076,
+      "grad_norm": 2.1923465728759766,
+      "learning_rate": 7.456388843555525e-06,
+      "loss": 1.1308,
+      "step": 2327
+    },
+    {
+      "epoch": 1.7619678334910123,
+      "grad_norm": 2.132502555847168,
+      "learning_rate": 7.448626168424434e-06,
+      "loss": 1.1637,
+      "step": 2328
+    },
+    {
+      "epoch": 1.762724692526017,
+      "grad_norm": 1.9766474962234497,
+      "learning_rate": 7.440865138514587e-06,
+      "loss": 1.1019,
+      "step": 2329
+    },
+    {
+      "epoch": 1.7634815515610218,
+      "grad_norm": 2.1354434490203857,
+      "learning_rate": 7.433105758828631e-06,
+      "loss": 1.0869,
+      "step": 2330
+    },
+    {
+      "epoch": 1.7642384105960265,
+      "grad_norm": 2.135441303253174,
+      "learning_rate": 7.425348034368143e-06,
+      "loss": 1.1077,
+      "step": 2331
+    },
+    {
+      "epoch": 1.7649952696310311,
+      "grad_norm": 1.9634079933166504,
+      "learning_rate": 7.41759197013363e-06,
+      "loss": 1.0686,
+      "step": 2332
+    },
+    {
+      "epoch": 1.765752128666036,
+      "grad_norm": 2.202788829803467,
+      "learning_rate": 7.409837571124535e-06,
+      "loss": 1.0706,
+      "step": 2333
+    },
+    {
+      "epoch": 1.7665089877010407,
+      "grad_norm": 2.3422369956970215,
+      "learning_rate": 7.40208484233923e-06,
+      "loss": 1.1673,
+      "step": 2334
+    },
+    {
+      "epoch": 1.7672658467360454,
+      "grad_norm": 2.338772773742676,
+      "learning_rate": 7.394333788774995e-06,
+      "loss": 1.1037,
+      "step": 2335
+    },
+    {
+      "epoch": 1.7680227057710503,
+      "grad_norm": 2.2548608779907227,
+      "learning_rate": 7.386584415428051e-06,
+      "loss": 1.1583,
+      "step": 2336
+    },
+    {
+      "epoch": 1.768779564806055,
+      "grad_norm": 2.1475353240966797,
+      "learning_rate": 7.3788367272935235e-06,
+      "loss": 1.1232,
+      "step": 2337
+    },
+    {
+      "epoch": 1.7695364238410596,
+      "grad_norm": 2.1223628520965576,
+      "learning_rate": 7.37109072936545e-06,
+      "loss": 1.1164,
+      "step": 2338
+    },
+    {
+      "epoch": 1.7702932828760645,
+      "grad_norm": 2.0361294746398926,
+      "learning_rate": 7.363346426636786e-06,
+      "loss": 1.16,
+      "step": 2339
+    },
+    {
+      "epoch": 1.771050141911069,
+      "grad_norm": 2.1143364906311035,
+      "learning_rate": 7.355603824099389e-06,
+      "loss": 1.142,
+      "step": 2340
+    },
+    {
+      "epoch": 1.7718070009460738,
+      "grad_norm": 2.2214882373809814,
+      "learning_rate": 7.347862926744027e-06,
+      "loss": 1.1375,
+      "step": 2341
+    },
+    {
+      "epoch": 1.7725638599810787,
+      "grad_norm": 1.9182907342910767,
+      "learning_rate": 7.34012373956036e-06,
+      "loss": 1.1099,
+      "step": 2342
+    },
+    {
+      "epoch": 1.7733207190160831,
+      "grad_norm": 2.0895349979400635,
+      "learning_rate": 7.332386267536949e-06,
+      "loss": 1.1397,
+      "step": 2343
+    },
+    {
+      "epoch": 1.774077578051088,
+      "grad_norm": 2.078885793685913,
+      "learning_rate": 7.3246505156612554e-06,
+      "loss": 1.1296,
+      "step": 2344
+    },
+    {
+      "epoch": 1.7748344370860927,
+      "grad_norm": 2.180187702178955,
+      "learning_rate": 7.3169164889196235e-06,
+      "loss": 1.1458,
+      "step": 2345
+    },
+    {
+      "epoch": 1.7755912961210973,
+      "grad_norm": 2.3137030601501465,
+      "learning_rate": 7.309184192297289e-06,
+      "loss": 1.0713,
+      "step": 2346
+    },
+    {
+      "epoch": 1.7763481551561022,
+      "grad_norm": 2.0382871627807617,
+      "learning_rate": 7.3014536307783725e-06,
+      "loss": 1.1288,
+      "step": 2347
+    },
+    {
+      "epoch": 1.777105014191107,
+      "grad_norm": 2.004988670349121,
+      "learning_rate": 7.293724809345879e-06,
+      "loss": 1.1304,
+      "step": 2348
+    },
+    {
+      "epoch": 1.7778618732261116,
+      "grad_norm": 2.0899946689605713,
+      "learning_rate": 7.285997732981683e-06,
+      "loss": 1.1095,
+      "step": 2349
+    },
+    {
+      "epoch": 1.7786187322611164,
+      "grad_norm": 2.197770118713379,
+      "learning_rate": 7.2782724066665475e-06,
+      "loss": 1.1219,
+      "step": 2350
+    },
+    {
+      "epoch": 1.779375591296121,
+      "grad_norm": 1.9547758102416992,
+      "learning_rate": 7.270548835380095e-06,
+      "loss": 1.0707,
+      "step": 2351
+    },
+    {
+      "epoch": 1.7801324503311258,
+      "grad_norm": 2.0842347145080566,
+      "learning_rate": 7.262827024100821e-06,
+      "loss": 1.1485,
+      "step": 2352
+    },
+    {
+      "epoch": 1.7808893093661307,
+      "grad_norm": 2.189990520477295,
+      "learning_rate": 7.255106977806092e-06,
+      "loss": 1.1403,
+      "step": 2353
+    },
+    {
+      "epoch": 1.7816461684011353,
+      "grad_norm": 2.299306869506836,
+      "learning_rate": 7.247388701472129e-06,
+      "loss": 1.1001,
+      "step": 2354
+    },
+    {
+      "epoch": 1.78240302743614,
+      "grad_norm": 2.0084657669067383,
+      "learning_rate": 7.239672200074012e-06,
+      "loss": 1.0777,
+      "step": 2355
+    },
+    {
+      "epoch": 1.7831598864711449,
+      "grad_norm": 1.9171555042266846,
+      "learning_rate": 7.231957478585687e-06,
+      "loss": 1.1022,
+      "step": 2356
+    },
+    {
+      "epoch": 1.7839167455061493,
+      "grad_norm": 2.116420030593872,
+      "learning_rate": 7.224244541979941e-06,
+      "loss": 1.0945,
+      "step": 2357
+    },
+    {
+      "epoch": 1.7846736045411542,
+      "grad_norm": 2.250598669052124,
+      "learning_rate": 7.216533395228419e-06,
+      "loss": 1.1327,
+      "step": 2358
+    },
+    {
+      "epoch": 1.785430463576159,
+      "grad_norm": 2.0988495349884033,
+      "learning_rate": 7.208824043301604e-06,
+      "loss": 1.1452,
+      "step": 2359
+    },
+    {
+      "epoch": 1.7861873226111635,
+      "grad_norm": 1.9777265787124634,
+      "learning_rate": 7.201116491168829e-06,
+      "loss": 1.0838,
+      "step": 2360
+    },
+    {
+      "epoch": 1.7869441816461684,
+      "grad_norm": 2.1055500507354736,
+      "learning_rate": 7.19341074379827e-06,
+      "loss": 1.0996,
+      "step": 2361
+    },
+    {
+      "epoch": 1.787701040681173,
+      "grad_norm": 1.8813843727111816,
+      "learning_rate": 7.185706806156921e-06,
+      "loss": 1.1238,
+      "step": 2362
+    },
+    {
+      "epoch": 1.7884578997161777,
+      "grad_norm": 1.9652965068817139,
+      "learning_rate": 7.178004683210634e-06,
+      "loss": 1.1141,
+      "step": 2363
+    },
+    {
+      "epoch": 1.7892147587511826,
+      "grad_norm": 2.2869348526000977,
+      "learning_rate": 7.170304379924078e-06,
+      "loss": 1.1882,
+      "step": 2364
+    },
+    {
+      "epoch": 1.7899716177861873,
+      "grad_norm": 2.1046929359436035,
+      "learning_rate": 7.162605901260749e-06,
+      "loss": 1.0947,
+      "step": 2365
+    },
+    {
+      "epoch": 1.790728476821192,
+      "grad_norm": 2.0936052799224854,
+      "learning_rate": 7.1549092521829676e-06,
+      "loss": 1.1371,
+      "step": 2366
+    },
+    {
+      "epoch": 1.7914853358561968,
+      "grad_norm": 2.0121428966522217,
+      "learning_rate": 7.147214437651881e-06,
+      "loss": 1.102,
+      "step": 2367
+    },
+    {
+      "epoch": 1.7922421948912015,
+      "grad_norm": 2.144970178604126,
+      "learning_rate": 7.139521462627446e-06,
+      "loss": 1.1266,
+      "step": 2368
+    },
+    {
+      "epoch": 1.7929990539262062,
+      "grad_norm": 2.3722221851348877,
+      "learning_rate": 7.1318303320684356e-06,
+      "loss": 1.1499,
+      "step": 2369
+    },
+    {
+      "epoch": 1.793755912961211,
+      "grad_norm": 2.255847454071045,
+      "learning_rate": 7.124141050932441e-06,
+      "loss": 1.1243,
+      "step": 2370
+    },
+    {
+      "epoch": 1.7945127719962157,
+      "grad_norm": 2.1879565715789795,
+      "learning_rate": 7.116453624175847e-06,
+      "loss": 1.0995,
+      "step": 2371
+    },
+    {
+      "epoch": 1.7952696310312204,
+      "grad_norm": 2.267245292663574,
+      "learning_rate": 7.108768056753863e-06,
+      "loss": 1.156,
+      "step": 2372
+    },
+    {
+      "epoch": 1.7960264900662253,
+      "grad_norm": 2.1807005405426025,
+      "learning_rate": 7.101084353620476e-06,
+      "loss": 1.1588,
+      "step": 2373
+    },
+    {
+      "epoch": 1.79678334910123,
+      "grad_norm": 2.2159693241119385,
+      "learning_rate": 7.0934025197284924e-06,
+      "loss": 1.0647,
+      "step": 2374
+    },
+    {
+      "epoch": 1.7975402081362346,
+      "grad_norm": 2.1058151721954346,
+      "learning_rate": 7.085722560029507e-06,
+      "loss": 1.1166,
+      "step": 2375
+    },
+    {
+      "epoch": 1.7982970671712395,
+      "grad_norm": 2.202956438064575,
+      "learning_rate": 7.0780444794738945e-06,
+      "loss": 1.1524,
+      "step": 2376
+    },
+    {
+      "epoch": 1.799053926206244,
+      "grad_norm": 2.15413761138916,
+      "learning_rate": 7.070368283010836e-06,
+      "loss": 1.1331,
+      "step": 2377
+    },
+    {
+      "epoch": 1.7998107852412488,
+      "grad_norm": 1.968179702758789,
+      "learning_rate": 7.062693975588291e-06,
+      "loss": 1.0785,
+      "step": 2378
+    },
+    {
+      "epoch": 1.8005676442762537,
+      "grad_norm": 2.2872471809387207,
+      "learning_rate": 7.0550215621529965e-06,
+      "loss": 1.1364,
+      "step": 2379
+    },
+    {
+      "epoch": 1.8013245033112582,
+      "grad_norm": 2.0598936080932617,
+      "learning_rate": 7.047351047650476e-06,
+      "loss": 1.1238,
+      "step": 2380
+    },
+    {
+      "epoch": 1.802081362346263,
+      "grad_norm": 2.055774688720703,
+      "learning_rate": 7.039682437025028e-06,
+      "loss": 1.1336,
+      "step": 2381
+    },
+    {
+      "epoch": 1.8028382213812677,
+      "grad_norm": 2.1142072677612305,
+      "learning_rate": 7.032015735219719e-06,
+      "loss": 1.1216,
+      "step": 2382
+    },
+    {
+      "epoch": 1.8035950804162724,
+      "grad_norm": 2.067873001098633,
+      "learning_rate": 7.024350947176391e-06,
+      "loss": 1.1253,
+      "step": 2383
+    },
+    {
+      "epoch": 1.8043519394512773,
+      "grad_norm": 1.906582236289978,
+      "learning_rate": 7.016688077835645e-06,
+      "loss": 1.1002,
+      "step": 2384
+    },
+    {
+      "epoch": 1.805108798486282,
+      "grad_norm": 2.005889892578125,
+      "learning_rate": 7.009027132136853e-06,
+      "loss": 1.135,
+      "step": 2385
+    },
+    {
+      "epoch": 1.8058656575212866,
+      "grad_norm": 1.9194884300231934,
+      "learning_rate": 7.001368115018144e-06,
+      "loss": 1.0872,
+      "step": 2386
+    },
+    {
+      "epoch": 1.8066225165562915,
+      "grad_norm": 2.044262647628784,
+      "learning_rate": 6.993711031416402e-06,
+      "loss": 1.0973,
+      "step": 2387
+    },
+    {
+      "epoch": 1.8073793755912961,
+      "grad_norm": 2.0718541145324707,
+      "learning_rate": 6.986055886267265e-06,
+      "loss": 1.1224,
+      "step": 2388
+    },
+    {
+      "epoch": 1.8081362346263008,
+      "grad_norm": 2.132376194000244,
+      "learning_rate": 6.97840268450513e-06,
+      "loss": 1.083,
+      "step": 2389
+    },
+    {
+      "epoch": 1.8088930936613057,
+      "grad_norm": 2.1769330501556396,
+      "learning_rate": 6.970751431063124e-06,
+      "loss": 1.148,
+      "step": 2390
+    },
+    {
+      "epoch": 1.8096499526963103,
+      "grad_norm": 2.1449358463287354,
+      "learning_rate": 6.963102130873134e-06,
+      "loss": 1.0967,
+      "step": 2391
+    },
+    {
+      "epoch": 1.810406811731315,
+      "grad_norm": 1.9859085083007812,
+      "learning_rate": 6.955454788865785e-06,
+      "loss": 1.1101,
+      "step": 2392
+    },
+    {
+      "epoch": 1.81116367076632,
+      "grad_norm": 2.211151361465454,
+      "learning_rate": 6.947809409970431e-06,
+      "loss": 1.114,
+      "step": 2393
+    },
+    {
+      "epoch": 1.8119205298013243,
+      "grad_norm": 2.0564661026000977,
+      "learning_rate": 6.940165999115169e-06,
+      "loss": 1.0703,
+      "step": 2394
+    },
+    {
+      "epoch": 1.8126773888363292,
+      "grad_norm": 2.2020647525787354,
+      "learning_rate": 6.932524561226824e-06,
+      "loss": 1.0784,
+      "step": 2395
+    },
+    {
+      "epoch": 1.8134342478713341,
+      "grad_norm": 2.0232954025268555,
+      "learning_rate": 6.924885101230955e-06,
+      "loss": 1.1231,
+      "step": 2396
+    },
+    {
+      "epoch": 1.8141911069063386,
+      "grad_norm": 2.0655837059020996,
+      "learning_rate": 6.917247624051836e-06,
+      "loss": 1.102,
+      "step": 2397
+    },
+    {
+      "epoch": 1.8149479659413434,
+      "grad_norm": 2.0320346355438232,
+      "learning_rate": 6.90961213461247e-06,
+      "loss": 1.119,
+      "step": 2398
+    },
+    {
+      "epoch": 1.815704824976348,
+      "grad_norm": 2.415329694747925,
+      "learning_rate": 6.901978637834579e-06,
+      "loss": 1.1015,
+      "step": 2399
+    },
+    {
+      "epoch": 1.8164616840113528,
+      "grad_norm": 1.962516188621521,
+      "learning_rate": 6.894347138638595e-06,
+      "loss": 1.1063,
+      "step": 2400
+    },
+    {
+      "epoch": 1.8172185430463577,
+      "grad_norm": 2.263796329498291,
+      "learning_rate": 6.886717641943668e-06,
+      "loss": 1.0946,
+      "step": 2401
+    },
+    {
+      "epoch": 1.8179754020813623,
+      "grad_norm": 1.8907090425491333,
+      "learning_rate": 6.879090152667655e-06,
+      "loss": 1.0842,
+      "step": 2402
+    },
+    {
+      "epoch": 1.818732261116367,
+      "grad_norm": 2.3313028812408447,
+      "learning_rate": 6.871464675727122e-06,
+      "loss": 1.0972,
+      "step": 2403
+    },
+    {
+      "epoch": 1.8194891201513719,
+      "grad_norm": 2.123699903488159,
+      "learning_rate": 6.8638412160373294e-06,
+      "loss": 1.0953,
+      "step": 2404
+    },
+    {
+      "epoch": 1.8202459791863765,
+      "grad_norm": 2.1058695316314697,
+      "learning_rate": 6.856219778512248e-06,
+      "loss": 1.0983,
+      "step": 2405
+    },
+    {
+      "epoch": 1.8210028382213812,
+      "grad_norm": 2.0354390144348145,
+      "learning_rate": 6.8486003680645384e-06,
+      "loss": 1.1184,
+      "step": 2406
+    },
+    {
+      "epoch": 1.821759697256386,
+      "grad_norm": 2.448774814605713,
+      "learning_rate": 6.840982989605554e-06,
+      "loss": 1.1902,
+      "step": 2407
+    },
+    {
+      "epoch": 1.8225165562913908,
+      "grad_norm": 2.067413330078125,
+      "learning_rate": 6.833367648045347e-06,
+      "loss": 1.0844,
+      "step": 2408
+    },
+    {
+      "epoch": 1.8232734153263954,
+      "grad_norm": 1.8351151943206787,
+      "learning_rate": 6.825754348292647e-06,
+      "loss": 1.0751,
+      "step": 2409
+    },
+    {
+      "epoch": 1.8240302743614003,
+      "grad_norm": 2.036219835281372,
+      "learning_rate": 6.8181430952548664e-06,
+      "loss": 1.1118,
+      "step": 2410
+    },
+    {
+      "epoch": 1.824787133396405,
+      "grad_norm": 2.2903716564178467,
+      "learning_rate": 6.810533893838111e-06,
+      "loss": 1.1085,
+      "step": 2411
+    },
+    {
+      "epoch": 1.8255439924314096,
+      "grad_norm": 2.1487245559692383,
+      "learning_rate": 6.802926748947149e-06,
+      "loss": 1.0766,
+      "step": 2412
+    },
+    {
+      "epoch": 1.8263008514664145,
+      "grad_norm": 2.073429822921753,
+      "learning_rate": 6.795321665485434e-06,
+      "loss": 1.1056,
+      "step": 2413
+    },
+    {
+      "epoch": 1.827057710501419,
+      "grad_norm": 2.1071133613586426,
+      "learning_rate": 6.7877186483550865e-06,
+      "loss": 1.1688,
+      "step": 2414
+    },
+    {
+      "epoch": 1.8278145695364238,
+      "grad_norm": 2.153792381286621,
+      "learning_rate": 6.780117702456892e-06,
+      "loss": 1.1281,
+      "step": 2415
+    },
+    {
+      "epoch": 1.8285714285714287,
+      "grad_norm": 2.046393632888794,
+      "learning_rate": 6.772518832690312e-06,
+      "loss": 1.1413,
+      "step": 2416
+    },
+    {
+      "epoch": 1.8293282876064332,
+      "grad_norm": 2.0445821285247803,
+      "learning_rate": 6.764922043953452e-06,
+      "loss": 1.0761,
+      "step": 2417
+    },
+    {
+      "epoch": 1.830085146641438,
+      "grad_norm": 2.1296608448028564,
+      "learning_rate": 6.757327341143093e-06,
+      "loss": 1.1077,
+      "step": 2418
+    },
+    {
+      "epoch": 1.8308420056764427,
+      "grad_norm": 2.218290328979492,
+      "learning_rate": 6.749734729154663e-06,
+      "loss": 1.0869,
+      "step": 2419
+    },
+    {
+      "epoch": 1.8315988647114474,
+      "grad_norm": 2.161032199859619,
+      "learning_rate": 6.742144212882244e-06,
+      "loss": 1.1116,
+      "step": 2420
+    },
+    {
+      "epoch": 1.8323557237464523,
+      "grad_norm": 1.979115605354309,
+      "learning_rate": 6.734555797218567e-06,
+      "loss": 1.1321,
+      "step": 2421
+    },
+    {
+      "epoch": 1.833112582781457,
+      "grad_norm": 2.016322135925293,
+      "learning_rate": 6.726969487055008e-06,
+      "loss": 1.1265,
+      "step": 2422
+    },
+    {
+      "epoch": 1.8338694418164616,
+      "grad_norm": 1.943589448928833,
+      "learning_rate": 6.719385287281589e-06,
+      "loss": 1.079,
+      "step": 2423
+    },
+    {
+      "epoch": 1.8346263008514665,
+      "grad_norm": 2.0779478549957275,
+      "learning_rate": 6.711803202786965e-06,
+      "loss": 1.161,
+      "step": 2424
+    },
+    {
+      "epoch": 1.8353831598864712,
+      "grad_norm": 1.9313197135925293,
+      "learning_rate": 6.7042232384584396e-06,
+      "loss": 1.114,
+      "step": 2425
+    },
+    {
+      "epoch": 1.8361400189214758,
+      "grad_norm": 2.177368640899658,
+      "learning_rate": 6.6966453991819355e-06,
+      "loss": 1.1141,
+      "step": 2426
+    },
+    {
+      "epoch": 1.8368968779564807,
+      "grad_norm": 2.155545234680176,
+      "learning_rate": 6.689069689842015e-06,
+      "loss": 1.1058,
+      "step": 2427
+    },
+    {
+      "epoch": 1.8376537369914854,
+      "grad_norm": 2.1615564823150635,
+      "learning_rate": 6.681496115321863e-06,
+      "loss": 1.1445,
+      "step": 2428
+    },
+    {
+      "epoch": 1.83841059602649,
+      "grad_norm": 2.2454423904418945,
+      "learning_rate": 6.6739246805032895e-06,
+      "loss": 1.1575,
+      "step": 2429
+    },
+    {
+      "epoch": 1.839167455061495,
+      "grad_norm": 1.9341751337051392,
+      "learning_rate": 6.6663553902667345e-06,
+      "loss": 1.0993,
+      "step": 2430
+    },
+    {
+      "epoch": 1.8399243140964994,
+      "grad_norm": 2.307654857635498,
+      "learning_rate": 6.658788249491236e-06,
+      "loss": 1.0903,
+      "step": 2431
+    },
+    {
+      "epoch": 1.8406811731315043,
+      "grad_norm": 2.172126293182373,
+      "learning_rate": 6.651223263054462e-06,
+      "loss": 1.1384,
+      "step": 2432
+    },
+    {
+      "epoch": 1.8414380321665091,
+      "grad_norm": 2.061699151992798,
+      "learning_rate": 6.64366043583269e-06,
+      "loss": 1.1066,
+      "step": 2433
+    },
+    {
+      "epoch": 1.8421948912015136,
+      "grad_norm": 2.0565085411071777,
+      "learning_rate": 6.636099772700797e-06,
+      "loss": 1.1265,
+      "step": 2434
+    },
+    {
+      "epoch": 1.8429517502365185,
+      "grad_norm": 2.3713178634643555,
+      "learning_rate": 6.628541278532276e-06,
+      "loss": 1.1067,
+      "step": 2435
+    },
+    {
+      "epoch": 1.8437086092715231,
+      "grad_norm": 2.0300235748291016,
+      "learning_rate": 6.620984958199217e-06,
+      "loss": 1.1053,
+      "step": 2436
+    },
+    {
+      "epoch": 1.8444654683065278,
+      "grad_norm": 1.8853594064712524,
+      "learning_rate": 6.613430816572308e-06,
+      "loss": 1.1375,
+      "step": 2437
+    },
+    {
+      "epoch": 1.8452223273415327,
+      "grad_norm": 2.140911102294922,
+      "learning_rate": 6.605878858520832e-06,
+      "loss": 1.1372,
+      "step": 2438
+    },
+    {
+      "epoch": 1.8459791863765374,
+      "grad_norm": 2.0533270835876465,
+      "learning_rate": 6.598329088912666e-06,
+      "loss": 1.1054,
+      "step": 2439
+    },
+    {
+      "epoch": 1.846736045411542,
+      "grad_norm": 2.0813000202178955,
+      "learning_rate": 6.59078151261428e-06,
+      "loss": 1.0635,
+      "step": 2440
+    },
+    {
+      "epoch": 1.847492904446547,
+      "grad_norm": 2.0938546657562256,
+      "learning_rate": 6.5832361344907225e-06,
+      "loss": 1.1368,
+      "step": 2441
+    },
+    {
+      "epoch": 1.8482497634815516,
+      "grad_norm": 2.1274354457855225,
+      "learning_rate": 6.57569295940563e-06,
+      "loss": 1.1446,
+      "step": 2442
+    },
+    {
+      "epoch": 1.8490066225165562,
+      "grad_norm": 2.2737364768981934,
+      "learning_rate": 6.5681519922212175e-06,
+      "loss": 1.1007,
+      "step": 2443
+    },
+    {
+      "epoch": 1.8497634815515611,
+      "grad_norm": 2.0562212467193604,
+      "learning_rate": 6.560613237798282e-06,
+      "loss": 1.1033,
+      "step": 2444
+    },
+    {
+      "epoch": 1.8505203405865658,
+      "grad_norm": 2.1894006729125977,
+      "learning_rate": 6.553076700996186e-06,
+      "loss": 1.1733,
+      "step": 2445
+    },
+    {
+      "epoch": 1.8512771996215704,
+      "grad_norm": 2.1526927947998047,
+      "learning_rate": 6.545542386672864e-06,
+      "loss": 1.1254,
+      "step": 2446
+    },
+    {
+      "epoch": 1.8520340586565753,
+      "grad_norm": 2.335092306137085,
+      "learning_rate": 6.538010299684827e-06,
+      "loss": 1.089,
+      "step": 2447
+    },
+    {
+      "epoch": 1.85279091769158,
+      "grad_norm": 2.1147849559783936,
+      "learning_rate": 6.530480444887135e-06,
+      "loss": 1.1075,
+      "step": 2448
+    },
+    {
+      "epoch": 1.8535477767265847,
+      "grad_norm": 1.9990819692611694,
+      "learning_rate": 6.522952827133424e-06,
+      "loss": 1.1069,
+      "step": 2449
+    },
+    {
+      "epoch": 1.8543046357615895,
+      "grad_norm": 2.2554056644439697,
+      "learning_rate": 6.515427451275879e-06,
+      "loss": 1.1205,
+      "step": 2450
+    },
+    {
+      "epoch": 1.855061494796594,
+      "grad_norm": 2.143373489379883,
+      "learning_rate": 6.507904322165242e-06,
+      "loss": 1.1,
+      "step": 2451
+    },
+    {
+      "epoch": 1.8558183538315989,
+      "grad_norm": 2.145324468612671,
+      "learning_rate": 6.500383444650808e-06,
+      "loss": 1.124,
+      "step": 2452
+    },
+    {
+      "epoch": 1.8565752128666035,
+      "grad_norm": 2.0681822299957275,
+      "learning_rate": 6.492864823580418e-06,
+      "loss": 1.1404,
+      "step": 2453
+    },
+    {
+      "epoch": 1.8573320719016082,
+      "grad_norm": 2.0816290378570557,
+      "learning_rate": 6.485348463800467e-06,
+      "loss": 1.121,
+      "step": 2454
+    },
+    {
+      "epoch": 1.858088930936613,
+      "grad_norm": 2.210402488708496,
+      "learning_rate": 6.477834370155879e-06,
+      "loss": 1.081,
+      "step": 2455
+    },
+    {
+      "epoch": 1.8588457899716178,
+      "grad_norm": 2.258357286453247,
+      "learning_rate": 6.4703225474901266e-06,
+      "loss": 1.1221,
+      "step": 2456
+    },
+    {
+      "epoch": 1.8596026490066224,
+      "grad_norm": 2.336432456970215,
+      "learning_rate": 6.462813000645216e-06,
+      "loss": 1.1288,
+      "step": 2457
+    },
+    {
+      "epoch": 1.8603595080416273,
+      "grad_norm": 2.224451780319214,
+      "learning_rate": 6.4553057344616885e-06,
+      "loss": 1.1213,
+      "step": 2458
+    },
+    {
+      "epoch": 1.861116367076632,
+      "grad_norm": 2.012571096420288,
+      "learning_rate": 6.447800753778608e-06,
+      "loss": 1.1079,
+      "step": 2459
+    },
+    {
+      "epoch": 1.8618732261116366,
+      "grad_norm": 2.0077013969421387,
+      "learning_rate": 6.440298063433578e-06,
+      "loss": 1.1139,
+      "step": 2460
+    },
+    {
+      "epoch": 1.8626300851466415,
+      "grad_norm": 2.0572779178619385,
+      "learning_rate": 6.432797668262713e-06,
+      "loss": 1.1225,
+      "step": 2461
+    },
+    {
+      "epoch": 1.8633869441816462,
+      "grad_norm": 2.052415609359741,
+      "learning_rate": 6.425299573100653e-06,
+      "loss": 1.1232,
+      "step": 2462
+    },
+    {
+      "epoch": 1.8641438032166509,
+      "grad_norm": 2.1070804595947266,
+      "learning_rate": 6.41780378278056e-06,
+      "loss": 1.1425,
+      "step": 2463
+    },
+    {
+      "epoch": 1.8649006622516557,
+      "grad_norm": 2.1018309593200684,
+      "learning_rate": 6.410310302134102e-06,
+      "loss": 1.124,
+      "step": 2464
+    },
+    {
+      "epoch": 1.8656575212866604,
+      "grad_norm": 2.104137897491455,
+      "learning_rate": 6.4028191359914635e-06,
+      "loss": 1.1366,
+      "step": 2465
+    },
+    {
+      "epoch": 1.866414380321665,
+      "grad_norm": 2.196840763092041,
+      "learning_rate": 6.395330289181339e-06,
+      "loss": 1.1138,
+      "step": 2466
+    },
+    {
+      "epoch": 1.86717123935667,
+      "grad_norm": 2.0204899311065674,
+      "learning_rate": 6.38784376653092e-06,
+      "loss": 1.1571,
+      "step": 2467
+    },
+    {
+      "epoch": 1.8679280983916744,
+      "grad_norm": 2.1718480587005615,
+      "learning_rate": 6.380359572865909e-06,
+      "loss": 1.1265,
+      "step": 2468
+    },
+    {
+      "epoch": 1.8686849574266793,
+      "grad_norm": 2.2680718898773193,
+      "learning_rate": 6.372877713010501e-06,
+      "loss": 1.1218,
+      "step": 2469
+    },
+    {
+      "epoch": 1.8694418164616842,
+      "grad_norm": 1.9217084646224976,
+      "learning_rate": 6.365398191787388e-06,
+      "loss": 1.0846,
+      "step": 2470
+    },
+    {
+      "epoch": 1.8701986754966886,
+      "grad_norm": 2.0585711002349854,
+      "learning_rate": 6.35792101401776e-06,
+      "loss": 1.1274,
+      "step": 2471
+    },
+    {
+      "epoch": 1.8709555345316935,
+      "grad_norm": 1.989283800125122,
+      "learning_rate": 6.350446184521285e-06,
+      "loss": 1.095,
+      "step": 2472
+    },
+    {
+      "epoch": 1.8717123935666982,
+      "grad_norm": 1.886738657951355,
+      "learning_rate": 6.3429737081161265e-06,
+      "loss": 1.1196,
+      "step": 2473
+    },
+    {
+      "epoch": 1.8724692526017028,
+      "grad_norm": 1.9688234329223633,
+      "learning_rate": 6.335503589618933e-06,
+      "loss": 1.143,
+      "step": 2474
+    },
+    {
+      "epoch": 1.8732261116367077,
+      "grad_norm": 2.37060284614563,
+      "learning_rate": 6.328035833844823e-06,
+      "loss": 1.1088,
+      "step": 2475
+    },
+    {
+      "epoch": 1.8739829706717124,
+      "grad_norm": 2.3199589252471924,
+      "learning_rate": 6.320570445607399e-06,
+      "loss": 1.1072,
+      "step": 2476
+    },
+    {
+      "epoch": 1.874739829706717,
+      "grad_norm": 1.914215087890625,
+      "learning_rate": 6.313107429718741e-06,
+      "loss": 1.1222,
+      "step": 2477
+    },
+    {
+      "epoch": 1.875496688741722,
+      "grad_norm": 2.3843131065368652,
+      "learning_rate": 6.305646790989391e-06,
+      "loss": 1.1509,
+      "step": 2478
+    },
+    {
+      "epoch": 1.8762535477767266,
+      "grad_norm": 2.1501553058624268,
+      "learning_rate": 6.298188534228365e-06,
+      "loss": 1.0925,
+      "step": 2479
+    },
+    {
+      "epoch": 1.8770104068117313,
+      "grad_norm": 2.103590965270996,
+      "learning_rate": 6.290732664243141e-06,
+      "loss": 1.1068,
+      "step": 2480
+    },
+    {
+      "epoch": 1.8777672658467361,
+      "grad_norm": 2.1373817920684814,
+      "learning_rate": 6.283279185839658e-06,
+      "loss": 1.1228,
+      "step": 2481
+    },
+    {
+      "epoch": 1.8785241248817408,
+      "grad_norm": 1.9448984861373901,
+      "learning_rate": 6.275828103822317e-06,
+      "loss": 1.1138,
+      "step": 2482
+    },
+    {
+      "epoch": 1.8792809839167455,
+      "grad_norm": 2.133575916290283,
+      "learning_rate": 6.268379422993969e-06,
+      "loss": 1.137,
+      "step": 2483
+    },
+    {
+      "epoch": 1.8800378429517504,
+      "grad_norm": 2.0521798133850098,
+      "learning_rate": 6.26093314815592e-06,
+      "loss": 1.1077,
+      "step": 2484
+    },
+    {
+      "epoch": 1.8807947019867548,
+      "grad_norm": 2.154632091522217,
+      "learning_rate": 6.253489284107929e-06,
+      "loss": 1.0963,
+      "step": 2485
+    },
+    {
+      "epoch": 1.8815515610217597,
+      "grad_norm": 2.0606281757354736,
+      "learning_rate": 6.246047835648191e-06,
+      "loss": 1.1233,
+      "step": 2486
+    },
+    {
+      "epoch": 1.8823084200567646,
+      "grad_norm": 1.9377020597457886,
+      "learning_rate": 6.238608807573355e-06,
+      "loss": 1.128,
+      "step": 2487
+    },
+    {
+      "epoch": 1.883065279091769,
+      "grad_norm": 2.133552074432373,
+      "learning_rate": 6.231172204678507e-06,
+      "loss": 1.0872,
+      "step": 2488
+    },
+    {
+      "epoch": 1.883822138126774,
+      "grad_norm": 2.1637847423553467,
+      "learning_rate": 6.2237380317571626e-06,
+      "loss": 1.1051,
+      "step": 2489
+    },
+    {
+      "epoch": 1.8845789971617786,
+      "grad_norm": 2.0870983600616455,
+      "learning_rate": 6.216306293601277e-06,
+      "loss": 1.1296,
+      "step": 2490
+    },
+    {
+      "epoch": 1.8853358561967832,
+      "grad_norm": 2.129365921020508,
+      "learning_rate": 6.20887699500124e-06,
+      "loss": 1.1125,
+      "step": 2491
+    },
+    {
+      "epoch": 1.8860927152317881,
+      "grad_norm": 1.887802004814148,
+      "learning_rate": 6.20145014074586e-06,
+      "loss": 1.0531,
+      "step": 2492
+    },
+    {
+      "epoch": 1.8868495742667928,
+      "grad_norm": 2.163595199584961,
+      "learning_rate": 6.194025735622371e-06,
+      "loss": 1.0727,
+      "step": 2493
+    },
+    {
+      "epoch": 1.8876064333017974,
+      "grad_norm": 1.8616597652435303,
+      "learning_rate": 6.186603784416441e-06,
+      "loss": 1.0561,
+      "step": 2494
+    },
+    {
+      "epoch": 1.8883632923368023,
+      "grad_norm": 2.1504313945770264,
+      "learning_rate": 6.179184291912138e-06,
+      "loss": 1.1093,
+      "step": 2495
+    },
+    {
+      "epoch": 1.889120151371807,
+      "grad_norm": 1.997025728225708,
+      "learning_rate": 6.171767262891958e-06,
+      "loss": 1.0991,
+      "step": 2496
+    },
+    {
+      "epoch": 1.8898770104068117,
+      "grad_norm": 2.141857147216797,
+      "learning_rate": 6.164352702136799e-06,
+      "loss": 1.0897,
+      "step": 2497
+    },
+    {
+      "epoch": 1.8906338694418166,
+      "grad_norm": 1.9679754972457886,
+      "learning_rate": 6.15694061442598e-06,
+      "loss": 1.0971,
+      "step": 2498
+    },
+    {
+      "epoch": 1.8913907284768212,
+      "grad_norm": 2.048257350921631,
+      "learning_rate": 6.149531004537222e-06,
+      "loss": 1.1139,
+      "step": 2499
+    },
+    {
+      "epoch": 1.8921475875118259,
+      "grad_norm": 2.4383885860443115,
+      "learning_rate": 6.1421238772466375e-06,
+      "loss": 1.1028,
+      "step": 2500
+    },
+    {
+      "epoch": 1.8929044465468308,
+      "grad_norm": 2.0352723598480225,
+      "learning_rate": 6.134719237328751e-06,
+      "loss": 1.0957,
+      "step": 2501
+    },
+    {
+      "epoch": 1.8936613055818354,
+      "grad_norm": 2.1713624000549316,
+      "learning_rate": 6.127317089556489e-06,
+      "loss": 1.0726,
+      "step": 2502
+    },
+    {
+      "epoch": 1.89441816461684,
+      "grad_norm": 2.2224864959716797,
+      "learning_rate": 6.119917438701151e-06,
+      "loss": 1.0919,
+      "step": 2503
+    },
+    {
+      "epoch": 1.895175023651845,
+      "grad_norm": 2.1596179008483887,
+      "learning_rate": 6.112520289532445e-06,
+      "loss": 1.1273,
+      "step": 2504
+    },
+    {
+      "epoch": 1.8959318826868494,
+      "grad_norm": 2.018328905105591,
+      "learning_rate": 6.105125646818463e-06,
+      "loss": 1.1354,
+      "step": 2505
+    },
+    {
+      "epoch": 1.8966887417218543,
+      "grad_norm": 2.1755290031433105,
+      "learning_rate": 6.097733515325671e-06,
+      "loss": 1.1037,
+      "step": 2506
+    },
+    {
+      "epoch": 1.8974456007568592,
+      "grad_norm": 2.172973871231079,
+      "learning_rate": 6.090343899818931e-06,
+      "loss": 1.0656,
+      "step": 2507
+    },
+    {
+      "epoch": 1.8982024597918636,
+      "grad_norm": 2.193934679031372,
+      "learning_rate": 6.0829568050614725e-06,
+      "loss": 1.1252,
+      "step": 2508
+    },
+    {
+      "epoch": 1.8989593188268685,
+      "grad_norm": 2.3651788234710693,
+      "learning_rate": 6.075572235814909e-06,
+      "loss": 1.1242,
+      "step": 2509
+    },
+    {
+      "epoch": 1.8997161778618732,
+      "grad_norm": 2.107897996902466,
+      "learning_rate": 6.0681901968392184e-06,
+      "loss": 1.0937,
+      "step": 2510
+    },
+    {
+      "epoch": 1.9004730368968779,
+      "grad_norm": 2.57551908493042,
+      "learning_rate": 6.060810692892748e-06,
+      "loss": 1.129,
+      "step": 2511
+    },
+    {
+      "epoch": 1.9012298959318827,
+      "grad_norm": 2.0832760334014893,
+      "learning_rate": 6.053433728732217e-06,
+      "loss": 1.138,
+      "step": 2512
+    },
+    {
+      "epoch": 1.9019867549668874,
+      "grad_norm": 1.9609954357147217,
+      "learning_rate": 6.046059309112703e-06,
+      "loss": 1.1404,
+      "step": 2513
+    },
+    {
+      "epoch": 1.902743614001892,
+      "grad_norm": 2.1895411014556885,
+      "learning_rate": 6.038687438787642e-06,
+      "loss": 1.1378,
+      "step": 2514
+    },
+    {
+      "epoch": 1.903500473036897,
+      "grad_norm": 2.058955430984497,
+      "learning_rate": 6.031318122508833e-06,
+      "loss": 1.117,
+      "step": 2515
+    },
+    {
+      "epoch": 1.9042573320719016,
+      "grad_norm": 2.1496293544769287,
+      "learning_rate": 6.023951365026426e-06,
+      "loss": 1.1115,
+      "step": 2516
+    },
+    {
+      "epoch": 1.9050141911069063,
+      "grad_norm": 2.147587776184082,
+      "learning_rate": 6.016587171088913e-06,
+      "loss": 1.1419,
+      "step": 2517
+    },
+    {
+      "epoch": 1.9057710501419112,
+      "grad_norm": 2.470024824142456,
+      "learning_rate": 6.009225545443148e-06,
+      "loss": 1.1518,
+      "step": 2518
+    },
+    {
+      "epoch": 1.9065279091769158,
+      "grad_norm": 2.0301973819732666,
+      "learning_rate": 6.001866492834322e-06,
+      "loss": 1.0815,
+      "step": 2519
+    },
+    {
+      "epoch": 1.9072847682119205,
+      "grad_norm": 2.2255492210388184,
+      "learning_rate": 5.994510018005964e-06,
+      "loss": 1.1555,
+      "step": 2520
+    },
+    {
+      "epoch": 1.9080416272469254,
+      "grad_norm": 2.101928472518921,
+      "learning_rate": 5.987156125699951e-06,
+      "loss": 1.1251,
+      "step": 2521
+    },
+    {
+      "epoch": 1.9087984862819298,
+      "grad_norm": 2.0817983150482178,
+      "learning_rate": 5.979804820656483e-06,
+      "loss": 1.1233,
+      "step": 2522
+    },
+    {
+      "epoch": 1.9095553453169347,
+      "grad_norm": 2.082615375518799,
+      "learning_rate": 5.972456107614105e-06,
+      "loss": 1.1198,
+      "step": 2523
+    },
+    {
+      "epoch": 1.9103122043519396,
+      "grad_norm": 2.036180257797241,
+      "learning_rate": 5.965109991309686e-06,
+      "loss": 1.1056,
+      "step": 2524
+    },
+    {
+      "epoch": 1.911069063386944,
+      "grad_norm": 2.358384847640991,
+      "learning_rate": 5.9577664764784126e-06,
+      "loss": 1.1125,
+      "step": 2525
+    },
+    {
+      "epoch": 1.911825922421949,
+      "grad_norm": 2.097381830215454,
+      "learning_rate": 5.950425567853813e-06,
+      "loss": 1.1394,
+      "step": 2526
+    },
+    {
+      "epoch": 1.9125827814569536,
+      "grad_norm": 2.0445775985717773,
+      "learning_rate": 5.943087270167718e-06,
+      "loss": 1.1276,
+      "step": 2527
+    },
+    {
+      "epoch": 1.9133396404919583,
+      "grad_norm": 2.2490360736846924,
+      "learning_rate": 5.935751588150282e-06,
+      "loss": 1.0963,
+      "step": 2528
+    },
+    {
+      "epoch": 1.9140964995269631,
+      "grad_norm": 2.212881088256836,
+      "learning_rate": 5.928418526529981e-06,
+      "loss": 1.0829,
+      "step": 2529
+    },
+    {
+      "epoch": 1.9148533585619678,
+      "grad_norm": 1.9197094440460205,
+      "learning_rate": 5.921088090033585e-06,
+      "loss": 1.0947,
+      "step": 2530
+    },
+    {
+      "epoch": 1.9156102175969725,
+      "grad_norm": 2.0829176902770996,
+      "learning_rate": 5.913760283386186e-06,
+      "loss": 1.1466,
+      "step": 2531
+    },
+    {
+      "epoch": 1.9163670766319774,
+      "grad_norm": 2.326220750808716,
+      "learning_rate": 5.906435111311179e-06,
+      "loss": 1.131,
+      "step": 2532
+    },
+    {
+      "epoch": 1.917123935666982,
+      "grad_norm": 2.2894301414489746,
+      "learning_rate": 5.899112578530255e-06,
+      "loss": 1.1062,
+      "step": 2533
+    },
+    {
+      "epoch": 1.9178807947019867,
+      "grad_norm": 2.134059190750122,
+      "learning_rate": 5.891792689763407e-06,
+      "loss": 1.1116,
+      "step": 2534
+    },
+    {
+      "epoch": 1.9186376537369916,
+      "grad_norm": 2.1360747814178467,
+      "learning_rate": 5.884475449728925e-06,
+      "loss": 1.15,
+      "step": 2535
+    },
+    {
+      "epoch": 1.9193945127719962,
+      "grad_norm": 2.3759396076202393,
+      "learning_rate": 5.877160863143391e-06,
+      "loss": 1.0696,
+      "step": 2536
+    },
+    {
+      "epoch": 1.920151371807001,
+      "grad_norm": 2.216271162033081,
+      "learning_rate": 5.869848934721671e-06,
+      "loss": 1.166,
+      "step": 2537
+    },
+    {
+      "epoch": 1.9209082308420058,
+      "grad_norm": 2.0322463512420654,
+      "learning_rate": 5.86253966917693e-06,
+      "loss": 1.1031,
+      "step": 2538
+    },
+    {
+      "epoch": 1.9216650898770105,
+      "grad_norm": 1.9586721658706665,
+      "learning_rate": 5.855233071220603e-06,
+      "loss": 1.1062,
+      "step": 2539
+    },
+    {
+      "epoch": 1.9224219489120151,
+      "grad_norm": 2.202064037322998,
+      "learning_rate": 5.8479291455624186e-06,
+      "loss": 1.1295,
+      "step": 2540
+    },
+    {
+      "epoch": 1.92317880794702,
+      "grad_norm": 2.291038751602173,
+      "learning_rate": 5.840627896910365e-06,
+      "loss": 1.157,
+      "step": 2541
+    },
+    {
+      "epoch": 1.9239356669820245,
+      "grad_norm": 1.989047884941101,
+      "learning_rate": 5.833329329970726e-06,
+      "loss": 1.1506,
+      "step": 2542
+    },
+    {
+      "epoch": 1.9246925260170293,
+      "grad_norm": 1.9984663724899292,
+      "learning_rate": 5.82603344944804e-06,
+      "loss": 1.106,
+      "step": 2543
+    },
+    {
+      "epoch": 1.9254493850520342,
+      "grad_norm": 2.3392581939697266,
+      "learning_rate": 5.818740260045123e-06,
+      "loss": 1.1819,
+      "step": 2544
+    },
+    {
+      "epoch": 1.9262062440870387,
+      "grad_norm": 2.148768663406372,
+      "learning_rate": 5.811449766463058e-06,
+      "loss": 1.1439,
+      "step": 2545
+    },
+    {
+      "epoch": 1.9269631031220436,
+      "grad_norm": 2.033663511276245,
+      "learning_rate": 5.804161973401175e-06,
+      "loss": 1.1111,
+      "step": 2546
+    },
+    {
+      "epoch": 1.9277199621570482,
+      "grad_norm": 2.2173452377319336,
+      "learning_rate": 5.796876885557084e-06,
+      "loss": 1.0752,
+      "step": 2547
+    },
+    {
+      "epoch": 1.9284768211920529,
+      "grad_norm": 2.039340019226074,
+      "learning_rate": 5.78959450762664e-06,
+      "loss": 1.0878,
+      "step": 2548
+    },
+    {
+      "epoch": 1.9292336802270578,
+      "grad_norm": 2.22098445892334,
+      "learning_rate": 5.782314844303949e-06,
+      "loss": 1.1109,
+      "step": 2549
+    },
+    {
+      "epoch": 1.9299905392620624,
+      "grad_norm": 1.9632805585861206,
+      "learning_rate": 5.775037900281372e-06,
+      "loss": 1.0981,
+      "step": 2550
+    },
+    {
+      "epoch": 1.930747398297067,
+      "grad_norm": 2.195981025695801,
+      "learning_rate": 5.767763680249521e-06,
+      "loss": 1.0659,
+      "step": 2551
+    },
+    {
+      "epoch": 1.931504257332072,
+      "grad_norm": 2.3889381885528564,
+      "learning_rate": 5.760492188897241e-06,
+      "loss": 1.1027,
+      "step": 2552
+    },
+    {
+      "epoch": 1.9322611163670766,
+      "grad_norm": 2.212132692337036,
+      "learning_rate": 5.753223430911625e-06,
+      "loss": 1.1435,
+      "step": 2553
+    },
+    {
+      "epoch": 1.9330179754020813,
+      "grad_norm": 2.109678268432617,
+      "learning_rate": 5.7459574109780105e-06,
+      "loss": 1.1226,
+      "step": 2554
+    },
+    {
+      "epoch": 1.9337748344370862,
+      "grad_norm": 2.1192758083343506,
+      "learning_rate": 5.738694133779954e-06,
+      "loss": 1.158,
+      "step": 2555
+    },
+    {
+      "epoch": 1.9345316934720909,
+      "grad_norm": 2.249246597290039,
+      "learning_rate": 5.7314336039992565e-06,
+      "loss": 1.1342,
+      "step": 2556
+    },
+    {
+      "epoch": 1.9352885525070955,
+      "grad_norm": 2.1159563064575195,
+      "learning_rate": 5.7241758263159504e-06,
+      "loss": 1.0984,
+      "step": 2557
+    },
+    {
+      "epoch": 1.9360454115421004,
+      "grad_norm": 2.092686891555786,
+      "learning_rate": 5.7169208054082794e-06,
+      "loss": 1.1462,
+      "step": 2558
+    },
+    {
+      "epoch": 1.9368022705771049,
+      "grad_norm": 2.201853036880493,
+      "learning_rate": 5.7096685459527235e-06,
+      "loss": 1.1121,
+      "step": 2559
+    },
+    {
+      "epoch": 1.9375591296121097,
+      "grad_norm": 2.29725980758667,
+      "learning_rate": 5.702419052623982e-06,
+      "loss": 1.17,
+      "step": 2560
+    },
+    {
+      "epoch": 1.9383159886471146,
+      "grad_norm": 2.636181592941284,
+      "learning_rate": 5.695172330094961e-06,
+      "loss": 1.1398,
+      "step": 2561
+    },
+    {
+      "epoch": 1.939072847682119,
+      "grad_norm": 2.0273807048797607,
+      "learning_rate": 5.687928383036795e-06,
+      "loss": 1.0939,
+      "step": 2562
+    },
+    {
+      "epoch": 1.939829706717124,
+      "grad_norm": 2.2744505405426025,
+      "learning_rate": 5.680687216118814e-06,
+      "loss": 1.1233,
+      "step": 2563
+    },
+    {
+      "epoch": 1.9405865657521286,
+      "grad_norm": 1.976132869720459,
+      "learning_rate": 5.6734488340085665e-06,
+      "loss": 1.1573,
+      "step": 2564
+    },
+    {
+      "epoch": 1.9413434247871333,
+      "grad_norm": 5.837771415710449,
+      "learning_rate": 5.666213241371809e-06,
+      "loss": 1.1299,
+      "step": 2565
+    },
+    {
+      "epoch": 1.9421002838221382,
+      "grad_norm": 2.1729319095611572,
+      "learning_rate": 5.658980442872484e-06,
+      "loss": 1.1399,
+      "step": 2566
+    },
+    {
+      "epoch": 1.9428571428571428,
+      "grad_norm": 2.049954652786255,
+      "learning_rate": 5.651750443172749e-06,
+      "loss": 1.1012,
+      "step": 2567
+    },
+    {
+      "epoch": 1.9436140018921475,
+      "grad_norm": 2.170069694519043,
+      "learning_rate": 5.644523246932951e-06,
+      "loss": 1.1276,
+      "step": 2568
+    },
+    {
+      "epoch": 1.9443708609271524,
+      "grad_norm": 2.143918514251709,
+      "learning_rate": 5.637298858811633e-06,
+      "loss": 1.1127,
+      "step": 2569
+    },
+    {
+      "epoch": 1.945127719962157,
+      "grad_norm": 2.114530324935913,
+      "learning_rate": 5.6300772834655195e-06,
+      "loss": 1.0949,
+      "step": 2570
+    },
+    {
+      "epoch": 1.9458845789971617,
+      "grad_norm": 2.067688226699829,
+      "learning_rate": 5.6228585255495315e-06,
+      "loss": 1.1402,
+      "step": 2571
+    },
+    {
+      "epoch": 1.9466414380321666,
+      "grad_norm": 2.254387855529785,
+      "learning_rate": 5.615642589716773e-06,
+      "loss": 1.197,
+      "step": 2572
+    },
+    {
+      "epoch": 1.9473982970671713,
+      "grad_norm": 2.0530171394348145,
+      "learning_rate": 5.608429480618519e-06,
+      "loss": 1.0985,
+      "step": 2573
+    },
+    {
+      "epoch": 1.948155156102176,
+      "grad_norm": 1.9690866470336914,
+      "learning_rate": 5.6012192029042354e-06,
+      "loss": 1.1004,
+      "step": 2574
+    },
+    {
+      "epoch": 1.9489120151371808,
+      "grad_norm": 2.2864949703216553,
+      "learning_rate": 5.594011761221554e-06,
+      "loss": 1.1298,
+      "step": 2575
+    },
+    {
+      "epoch": 1.9496688741721855,
+      "grad_norm": 2.291849374771118,
+      "learning_rate": 5.5868071602162875e-06,
+      "loss": 1.1472,
+      "step": 2576
+    },
+    {
+      "epoch": 1.9504257332071901,
+      "grad_norm": 2.223792552947998,
+      "learning_rate": 5.579605404532403e-06,
+      "loss": 1.1307,
+      "step": 2577
+    },
+    {
+      "epoch": 1.951182592242195,
+      "grad_norm": 2.0151219367980957,
+      "learning_rate": 5.572406498812049e-06,
+      "loss": 1.1281,
+      "step": 2578
+    },
+    {
+      "epoch": 1.9519394512771995,
+      "grad_norm": 2.023299217224121,
+      "learning_rate": 5.565210447695529e-06,
+      "loss": 1.165,
+      "step": 2579
+    },
+    {
+      "epoch": 1.9526963103122044,
+      "grad_norm": 1.902061939239502,
+      "learning_rate": 5.5580172558213064e-06,
+      "loss": 1.061,
+      "step": 2580
+    },
+    {
+      "epoch": 1.953453169347209,
+      "grad_norm": 2.015148639678955,
+      "learning_rate": 5.550826927826003e-06,
+      "loss": 1.0866,
+      "step": 2581
+    },
+    {
+      "epoch": 1.9542100283822137,
+      "grad_norm": 1.9616479873657227,
+      "learning_rate": 5.5436394683443996e-06,
+      "loss": 1.0903,
+      "step": 2582
+    },
+    {
+      "epoch": 1.9549668874172186,
+      "grad_norm": 1.9738472700119019,
+      "learning_rate": 5.536454882009412e-06,
+      "loss": 1.1312,
+      "step": 2583
+    },
+    {
+      "epoch": 1.9557237464522232,
+      "grad_norm": 2.2209506034851074,
+      "learning_rate": 5.52927317345213e-06,
+      "loss": 1.0352,
+      "step": 2584
+    },
+    {
+      "epoch": 1.956480605487228,
+      "grad_norm": 2.2492170333862305,
+      "learning_rate": 5.522094347301757e-06,
+      "loss": 1.0982,
+      "step": 2585
+    },
+    {
+      "epoch": 1.9572374645222328,
+      "grad_norm": 1.9593442678451538,
+      "learning_rate": 5.514918408185666e-06,
+      "loss": 1.1162,
+      "step": 2586
+    },
+    {
+      "epoch": 1.9579943235572375,
+      "grad_norm": 2.0279417037963867,
+      "learning_rate": 5.507745360729356e-06,
+      "loss": 1.1148,
+      "step": 2587
+    },
+    {
+      "epoch": 1.9587511825922421,
+      "grad_norm": 2.1683051586151123,
+      "learning_rate": 5.500575209556462e-06,
+      "loss": 1.1078,
+      "step": 2588
+    },
+    {
+      "epoch": 1.959508041627247,
+      "grad_norm": 2.0510294437408447,
+      "learning_rate": 5.493407959288752e-06,
+      "loss": 1.1099,
+      "step": 2589
+    },
+    {
+      "epoch": 1.9602649006622517,
+      "grad_norm": 2.144102096557617,
+      "learning_rate": 5.486243614546135e-06,
+      "loss": 1.0938,
+      "step": 2590
+    },
+    {
+      "epoch": 1.9610217596972563,
+      "grad_norm": 1.9423801898956299,
+      "learning_rate": 5.479082179946628e-06,
+      "loss": 1.0941,
+      "step": 2591
+    },
+    {
+      "epoch": 1.9617786187322612,
+      "grad_norm": 1.9970104694366455,
+      "learning_rate": 5.471923660106387e-06,
+      "loss": 1.1106,
+      "step": 2592
+    },
+    {
+      "epoch": 1.962535477767266,
+      "grad_norm": 2.0680718421936035,
+      "learning_rate": 5.46476805963969e-06,
+      "loss": 1.1039,
+      "step": 2593
+    },
+    {
+      "epoch": 1.9632923368022706,
+      "grad_norm": 2.138693332672119,
+      "learning_rate": 5.457615383158917e-06,
+      "loss": 1.1283,
+      "step": 2594
+    },
+    {
+      "epoch": 1.9640491958372754,
+      "grad_norm": 2.0552265644073486,
+      "learning_rate": 5.450465635274581e-06,
+      "loss": 1.0889,
+      "step": 2595
+    },
+    {
+      "epoch": 1.9648060548722799,
+      "grad_norm": 1.974301815032959,
+      "learning_rate": 5.443318820595303e-06,
+      "loss": 1.1174,
+      "step": 2596
+    },
+    {
+      "epoch": 1.9655629139072848,
+      "grad_norm": 1.9770950078964233,
+      "learning_rate": 5.436174943727803e-06,
+      "loss": 1.1261,
+      "step": 2597
+    },
+    {
+      "epoch": 1.9663197729422897,
+      "grad_norm": 2.0605309009552,
+      "learning_rate": 5.42903400927692e-06,
+      "loss": 1.1453,
+      "step": 2598
+    },
+    {
+      "epoch": 1.967076631977294,
+      "grad_norm": 1.9402192831039429,
+      "learning_rate": 5.421896021845591e-06,
+      "loss": 1.1047,
+      "step": 2599
+    },
+    {
+      "epoch": 1.967833491012299,
+      "grad_norm": 2.0428860187530518,
+      "learning_rate": 5.4147609860348545e-06,
+      "loss": 1.0957,
+      "step": 2600
+    },
+    {
+      "epoch": 1.9685903500473036,
+      "grad_norm": 1.9550975561141968,
+      "learning_rate": 5.407628906443844e-06,
+      "loss": 1.1202,
+      "step": 2601
+    },
+    {
+      "epoch": 1.9693472090823083,
+      "grad_norm": 1.9513860940933228,
+      "learning_rate": 5.400499787669788e-06,
+      "loss": 1.0562,
+      "step": 2602
+    },
+    {
+      "epoch": 1.9701040681173132,
+      "grad_norm": 2.2109532356262207,
+      "learning_rate": 5.393373634308015e-06,
+      "loss": 1.1637,
+      "step": 2603
+    },
+    {
+      "epoch": 1.9708609271523179,
+      "grad_norm": 1.9870307445526123,
+      "learning_rate": 5.3862504509519245e-06,
+      "loss": 1.1243,
+      "step": 2604
+    },
+    {
+      "epoch": 1.9716177861873225,
+      "grad_norm": 2.027862310409546,
+      "learning_rate": 5.379130242193018e-06,
+      "loss": 1.0791,
+      "step": 2605
+    },
+    {
+      "epoch": 1.9723746452223274,
+      "grad_norm": 1.969875454902649,
+      "learning_rate": 5.372013012620875e-06,
+      "loss": 1.1593,
+      "step": 2606
+    },
+    {
+      "epoch": 1.973131504257332,
+      "grad_norm": 2.163132667541504,
+      "learning_rate": 5.3648987668231475e-06,
+      "loss": 1.1337,
+      "step": 2607
+    },
+    {
+      "epoch": 1.9738883632923367,
+      "grad_norm": 2.0548205375671387,
+      "learning_rate": 5.357787509385571e-06,
+      "loss": 1.1185,
+      "step": 2608
+    },
+    {
+      "epoch": 1.9746452223273416,
+      "grad_norm": 2.120103597640991,
+      "learning_rate": 5.350679244891962e-06,
+      "loss": 1.1419,
+      "step": 2609
+    },
+    {
+      "epoch": 1.9754020813623463,
+      "grad_norm": 2.1263537406921387,
+      "learning_rate": 5.343573977924188e-06,
+      "loss": 1.1043,
+      "step": 2610
+    },
+    {
+      "epoch": 1.976158940397351,
+      "grad_norm": 2.0377280712127686,
+      "learning_rate": 5.3364717130622e-06,
+      "loss": 1.0852,
+      "step": 2611
+    },
+    {
+      "epoch": 1.9769157994323558,
+      "grad_norm": 1.9558144807815552,
+      "learning_rate": 5.329372454884014e-06,
+      "loss": 1.1432,
+      "step": 2612
+    },
+    {
+      "epoch": 1.9776726584673603,
+      "grad_norm": 2.1129062175750732,
+      "learning_rate": 5.322276207965698e-06,
+      "loss": 1.1347,
+      "step": 2613
+    },
+    {
+      "epoch": 1.9784295175023652,
+      "grad_norm": 2.042936086654663,
+      "learning_rate": 5.315182976881382e-06,
+      "loss": 1.1201,
+      "step": 2614
+    },
+    {
+      "epoch": 1.97918637653737,
+      "grad_norm": 2.1350150108337402,
+      "learning_rate": 5.308092766203265e-06,
+      "loss": 1.0633,
+      "step": 2615
+    },
+    {
+      "epoch": 1.9799432355723745,
+      "grad_norm": 1.984386682510376,
+      "learning_rate": 5.301005580501579e-06,
+      "loss": 1.1045,
+      "step": 2616
+    },
+    {
+      "epoch": 1.9807000946073794,
+      "grad_norm": 1.9831804037094116,
+      "learning_rate": 5.293921424344624e-06,
+      "loss": 1.1376,
+      "step": 2617
+    },
+    {
+      "epoch": 1.981456953642384,
+      "grad_norm": 2.1934800148010254,
+      "learning_rate": 5.286840302298729e-06,
+      "loss": 1.1043,
+      "step": 2618
+    },
+    {
+      "epoch": 1.9822138126773887,
+      "grad_norm": 2.0572476387023926,
+      "learning_rate": 5.2797622189282835e-06,
+      "loss": 1.0807,
+      "step": 2619
+    },
+    {
+      "epoch": 1.9829706717123936,
+      "grad_norm": 2.0445570945739746,
+      "learning_rate": 5.272687178795715e-06,
+      "loss": 1.1075,
+      "step": 2620
+    },
+    {
+      "epoch": 1.9837275307473983,
+      "grad_norm": 2.0021796226501465,
+      "learning_rate": 5.265615186461479e-06,
+      "loss": 1.0971,
+      "step": 2621
+    },
+    {
+      "epoch": 1.984484389782403,
+      "grad_norm": 2.107959508895874,
+      "learning_rate": 5.258546246484077e-06,
+      "loss": 1.1329,
+      "step": 2622
+    },
+    {
+      "epoch": 1.9852412488174078,
+      "grad_norm": 1.8930269479751587,
+      "learning_rate": 5.251480363420041e-06,
+      "loss": 1.0927,
+      "step": 2623
+    },
+    {
+      "epoch": 1.9859981078524125,
+      "grad_norm": 2.0325968265533447,
+      "learning_rate": 5.244417541823935e-06,
+      "loss": 1.1009,
+      "step": 2624
+    },
+    {
+      "epoch": 1.9867549668874172,
+      "grad_norm": 2.3380300998687744,
+      "learning_rate": 5.237357786248337e-06,
+      "loss": 1.0821,
+      "step": 2625
+    },
+    {
+      "epoch": 1.987511825922422,
+      "grad_norm": 2.121659517288208,
+      "learning_rate": 5.230301101243864e-06,
+      "loss": 1.0595,
+      "step": 2626
+    },
+    {
+      "epoch": 1.9882686849574267,
+      "grad_norm": 2.329930067062378,
+      "learning_rate": 5.22324749135915e-06,
+      "loss": 1.1233,
+      "step": 2627
+    },
+    {
+      "epoch": 1.9890255439924314,
+      "grad_norm": 2.044088840484619,
+      "learning_rate": 5.216196961140837e-06,
+      "loss": 1.1064,
+      "step": 2628
+    },
+    {
+      "epoch": 1.9897824030274363,
+      "grad_norm": 2.201205015182495,
+      "learning_rate": 5.209149515133593e-06,
+      "loss": 1.1553,
+      "step": 2629
+    },
+    {
+      "epoch": 1.990539262062441,
+      "grad_norm": 2.029348850250244,
+      "learning_rate": 5.202105157880095e-06,
+      "loss": 1.119,
+      "step": 2630
+    },
+    {
+      "epoch": 1.9912961210974456,
+      "grad_norm": 2.332625150680542,
+      "learning_rate": 5.1950638939210296e-06,
+      "loss": 1.0767,
+      "step": 2631
+    },
+    {
+      "epoch": 1.9920529801324505,
+      "grad_norm": 2.2954776287078857,
+      "learning_rate": 5.188025727795084e-06,
+      "loss": 1.0942,
+      "step": 2632
+    },
+    {
+      "epoch": 1.992809839167455,
+      "grad_norm": 1.9560728073120117,
+      "learning_rate": 5.180990664038954e-06,
+      "loss": 1.0801,
+      "step": 2633
+    },
+    {
+      "epoch": 1.9935666982024598,
+      "grad_norm": 2.1281726360321045,
+      "learning_rate": 5.17395870718734e-06,
+      "loss": 1.1274,
+      "step": 2634
+    },
+    {
+      "epoch": 1.9943235572374647,
+      "grad_norm": 1.8601038455963135,
+      "learning_rate": 5.166929861772925e-06,
+      "loss": 1.0999,
+      "step": 2635
+    },
+    {
+      "epoch": 1.9950804162724691,
+      "grad_norm": 2.056415319442749,
+      "learning_rate": 5.159904132326399e-06,
+      "loss": 1.0721,
+      "step": 2636
+    },
+    {
+      "epoch": 1.995837275307474,
+      "grad_norm": 1.9574776887893677,
+      "learning_rate": 5.152881523376445e-06,
+      "loss": 1.1364,
+      "step": 2637
+    },
+    {
+      "epoch": 1.9965941343424787,
+      "grad_norm": 2.011434555053711,
+      "learning_rate": 5.145862039449723e-06,
+      "loss": 1.1044,
+      "step": 2638
+    },
+    {
+      "epoch": 1.9973509933774833,
+      "grad_norm": 1.9725828170776367,
+      "learning_rate": 5.138845685070891e-06,
+      "loss": 1.1425,
+      "step": 2639
+    },
+    {
+      "epoch": 1.9981078524124882,
+      "grad_norm": 2.337836742401123,
+      "learning_rate": 5.131832464762576e-06,
+      "loss": 1.1062,
+      "step": 2640
+    },
+    {
+      "epoch": 1.998864711447493,
+      "grad_norm": 2.0006585121154785,
+      "learning_rate": 5.1248223830454e-06,
+      "loss": 1.1117,
+      "step": 2641
+    },
+    {
+      "epoch": 1.9996215704824976,
+      "grad_norm": 2.063220977783203,
+      "learning_rate": 5.117815444437956e-06,
+      "loss": 1.0999,
+      "step": 2642
+    },
+    {
+      "epoch": 2.0003784295175024,
+      "grad_norm": 2.052854537963867,
+      "learning_rate": 5.110811653456801e-06,
+      "loss": 1.099,
+      "step": 2643
+    },
+    {
+      "epoch": 2.001135288552507,
+      "grad_norm": 2.046462059020996,
+      "learning_rate": 5.103811014616479e-06,
+      "loss": 1.0824,
+      "step": 2644
+    },
+    {
+      "epoch": 2.0018921475875118,
+      "grad_norm": 2.0290746688842773,
+      "learning_rate": 5.096813532429496e-06,
+      "loss": 1.0797,
+      "step": 2645
+    },
+    {
+      "epoch": 2.0026490066225167,
+      "grad_norm": 2.0245625972747803,
+      "learning_rate": 5.089819211406316e-06,
+      "loss": 1.078,
+      "step": 2646
+    },
+    {
+      "epoch": 2.003405865657521,
+      "grad_norm": 2.207991600036621,
+      "learning_rate": 5.082828056055375e-06,
+      "loss": 1.0669,
+      "step": 2647
+    },
+    {
+      "epoch": 2.004162724692526,
+      "grad_norm": 2.133127212524414,
+      "learning_rate": 5.075840070883069e-06,
+      "loss": 1.0906,
+      "step": 2648
+    },
+    {
+      "epoch": 2.004919583727531,
+      "grad_norm": 2.1217539310455322,
+      "learning_rate": 5.068855260393739e-06,
+      "loss": 1.1084,
+      "step": 2649
+    },
+    {
+      "epoch": 2.0056764427625353,
+      "grad_norm": 2.1101694107055664,
+      "learning_rate": 5.061873629089693e-06,
+      "loss": 1.0727,
+      "step": 2650
+    },
+    {
+      "epoch": 2.00643330179754,
+      "grad_norm": 2.0910747051239014,
+      "learning_rate": 5.054895181471185e-06,
+      "loss": 1.0533,
+      "step": 2651
+    },
+    {
+      "epoch": 2.007190160832545,
+      "grad_norm": 1.8526837825775146,
+      "learning_rate": 5.0479199220364085e-06,
+      "loss": 1.0245,
+      "step": 2652
+    },
+    {
+      "epoch": 2.0079470198675495,
+      "grad_norm": 2.0023531913757324,
+      "learning_rate": 5.040947855281515e-06,
+      "loss": 1.096,
+      "step": 2653
+    },
+    {
+      "epoch": 2.0087038789025544,
+      "grad_norm": 1.9686319828033447,
+      "learning_rate": 5.033978985700592e-06,
+      "loss": 1.0526,
+      "step": 2654
+    },
+    {
+      "epoch": 2.0094607379375593,
+      "grad_norm": 1.999506950378418,
+      "learning_rate": 5.02701331778567e-06,
+      "loss": 1.0875,
+      "step": 2655
+    },
+    {
+      "epoch": 2.0102175969725637,
+      "grad_norm": 2.096315860748291,
+      "learning_rate": 5.020050856026703e-06,
+      "loss": 1.0637,
+      "step": 2656
+    },
+    {
+      "epoch": 2.0109744560075686,
+      "grad_norm": 2.0473968982696533,
+      "learning_rate": 5.013091604911594e-06,
+      "loss": 1.1281,
+      "step": 2657
+    },
+    {
+      "epoch": 2.0117313150425735,
+      "grad_norm": 2.0134975910186768,
+      "learning_rate": 5.006135568926175e-06,
+      "loss": 1.0641,
+      "step": 2658
+    },
+    {
+      "epoch": 2.012488174077578,
+      "grad_norm": 1.9990431070327759,
+      "learning_rate": 4.999182752554189e-06,
+      "loss": 1.1031,
+      "step": 2659
+    },
+    {
+      "epoch": 2.013245033112583,
+      "grad_norm": 2.199671983718872,
+      "learning_rate": 4.992233160277321e-06,
+      "loss": 1.1043,
+      "step": 2660
+    },
+    {
+      "epoch": 2.0140018921475873,
+      "grad_norm": 2.0332555770874023,
+      "learning_rate": 4.985286796575174e-06,
+      "loss": 1.1057,
+      "step": 2661
+    },
+    {
+      "epoch": 2.014758751182592,
+      "grad_norm": 2.1199121475219727,
+      "learning_rate": 4.978343665925269e-06,
+      "loss": 1.036,
+      "step": 2662
+    },
+    {
+      "epoch": 2.015515610217597,
+      "grad_norm": 2.047947883605957,
+      "learning_rate": 4.9714037728030415e-06,
+      "loss": 1.0934,
+      "step": 2663
+    },
+    {
+      "epoch": 2.0162724692526015,
+      "grad_norm": 1.814427137374878,
+      "learning_rate": 4.964467121681834e-06,
+      "loss": 1.0809,
+      "step": 2664
+    },
+    {
+      "epoch": 2.0170293282876064,
+      "grad_norm": 2.189452648162842,
+      "learning_rate": 4.957533717032911e-06,
+      "loss": 1.0565,
+      "step": 2665
+    },
+    {
+      "epoch": 2.0177861873226113,
+      "grad_norm": 2.025991201400757,
+      "learning_rate": 4.95060356332544e-06,
+      "loss": 1.0633,
+      "step": 2666
+    },
+    {
+      "epoch": 2.0185430463576157,
+      "grad_norm": 2.3097431659698486,
+      "learning_rate": 4.943676665026492e-06,
+      "loss": 1.0527,
+      "step": 2667
+    },
+    {
+      "epoch": 2.0192999053926206,
+      "grad_norm": 2.235900402069092,
+      "learning_rate": 4.936753026601047e-06,
+      "loss": 1.0878,
+      "step": 2668
+    },
+    {
+      "epoch": 2.0200567644276255,
+      "grad_norm": 2.0237877368927,
+      "learning_rate": 4.929832652511963e-06,
+      "loss": 1.0243,
+      "step": 2669
+    },
+    {
+      "epoch": 2.02081362346263,
+      "grad_norm": 2.148148536682129,
+      "learning_rate": 4.922915547220014e-06,
+      "loss": 1.05,
+      "step": 2670
+    },
+    {
+      "epoch": 2.021570482497635,
+      "grad_norm": 2.204345464706421,
+      "learning_rate": 4.91600171518386e-06,
+      "loss": 1.0613,
+      "step": 2671
+    },
+    {
+      "epoch": 2.0223273415326397,
+      "grad_norm": 2.05426287651062,
+      "learning_rate": 4.909091160860053e-06,
+      "loss": 1.0683,
+      "step": 2672
+    },
+    {
+      "epoch": 2.023084200567644,
+      "grad_norm": 2.0507991313934326,
+      "learning_rate": 4.902183888703029e-06,
+      "loss": 1.1039,
+      "step": 2673
+    },
+    {
+      "epoch": 2.023841059602649,
+      "grad_norm": 2.111011505126953,
+      "learning_rate": 4.895279903165118e-06,
+      "loss": 1.0708,
+      "step": 2674
+    },
+    {
+      "epoch": 2.024597918637654,
+      "grad_norm": 2.152397871017456,
+      "learning_rate": 4.888379208696516e-06,
+      "loss": 1.135,
+      "step": 2675
+    },
+    {
+      "epoch": 2.0253547776726584,
+      "grad_norm": 2.062863826751709,
+      "learning_rate": 4.881481809745303e-06,
+      "loss": 1.0808,
+      "step": 2676
+    },
+    {
+      "epoch": 2.0261116367076633,
+      "grad_norm": 2.052548885345459,
+      "learning_rate": 4.874587710757442e-06,
+      "loss": 1.0729,
+      "step": 2677
+    },
+    {
+      "epoch": 2.026868495742668,
+      "grad_norm": 2.1231849193573,
+      "learning_rate": 4.8676969161767625e-06,
+      "loss": 1.13,
+      "step": 2678
+    },
+    {
+      "epoch": 2.0276253547776726,
+      "grad_norm": 1.9931443929672241,
+      "learning_rate": 4.860809430444969e-06,
+      "loss": 1.0863,
+      "step": 2679
+    },
+    {
+      "epoch": 2.0283822138126775,
+      "grad_norm": 1.980806589126587,
+      "learning_rate": 4.853925258001626e-06,
+      "loss": 1.1208,
+      "step": 2680
+    },
+    {
+      "epoch": 2.029139072847682,
+      "grad_norm": 2.0515875816345215,
+      "learning_rate": 4.847044403284166e-06,
+      "loss": 1.0301,
+      "step": 2681
+    },
+    {
+      "epoch": 2.029895931882687,
+      "grad_norm": 2.115715742111206,
+      "learning_rate": 4.840166870727887e-06,
+      "loss": 1.0677,
+      "step": 2682
+    },
+    {
+      "epoch": 2.0306527909176917,
+      "grad_norm": 1.9753094911575317,
+      "learning_rate": 4.833292664765935e-06,
+      "loss": 1.0814,
+      "step": 2683
+    },
+    {
+      "epoch": 2.031409649952696,
+      "grad_norm": 2.0974655151367188,
+      "learning_rate": 4.8264217898293226e-06,
+      "loss": 1.0553,
+      "step": 2684
+    },
+    {
+      "epoch": 2.032166508987701,
+      "grad_norm": 1.9045485258102417,
+      "learning_rate": 4.8195542503469145e-06,
+      "loss": 1.1018,
+      "step": 2685
+    },
+    {
+      "epoch": 2.032923368022706,
+      "grad_norm": 2.1343581676483154,
+      "learning_rate": 4.812690050745413e-06,
+      "loss": 1.1279,
+      "step": 2686
+    },
+    {
+      "epoch": 2.0336802270577103,
+      "grad_norm": 2.060368776321411,
+      "learning_rate": 4.805829195449382e-06,
+      "loss": 1.1021,
+      "step": 2687
+    },
+    {
+      "epoch": 2.0344370860927152,
+      "grad_norm": 2.2184059619903564,
+      "learning_rate": 4.798971688881224e-06,
+      "loss": 1.0819,
+      "step": 2688
+    },
+    {
+      "epoch": 2.03519394512772,
+      "grad_norm": 2.1077048778533936,
+      "learning_rate": 4.792117535461187e-06,
+      "loss": 1.0469,
+      "step": 2689
+    },
+    {
+      "epoch": 2.0359508041627246,
+      "grad_norm": 2.105867624282837,
+      "learning_rate": 4.7852667396073475e-06,
+      "loss": 1.1129,
+      "step": 2690
+    },
+    {
+      "epoch": 2.0367076631977294,
+      "grad_norm": 1.9376499652862549,
+      "learning_rate": 4.7784193057356234e-06,
+      "loss": 1.0889,
+      "step": 2691
+    },
+    {
+      "epoch": 2.0374645222327343,
+      "grad_norm": 2.1429734230041504,
+      "learning_rate": 4.771575238259769e-06,
+      "loss": 1.06,
+      "step": 2692
+    },
+    {
+      "epoch": 2.0382213812677388,
+      "grad_norm": 2.1066880226135254,
+      "learning_rate": 4.764734541591365e-06,
+      "loss": 1.0841,
+      "step": 2693
+    },
+    {
+      "epoch": 2.0389782403027437,
+      "grad_norm": 2.034998655319214,
+      "learning_rate": 4.757897220139822e-06,
+      "loss": 1.1155,
+      "step": 2694
+    },
+    {
+      "epoch": 2.0397350993377485,
+      "grad_norm": 2.0084969997406006,
+      "learning_rate": 4.751063278312371e-06,
+      "loss": 1.0365,
+      "step": 2695
+    },
+    {
+      "epoch": 2.040491958372753,
+      "grad_norm": 2.260364055633545,
+      "learning_rate": 4.744232720514074e-06,
+      "loss": 1.0722,
+      "step": 2696
+    },
+    {
+      "epoch": 2.041248817407758,
+      "grad_norm": 2.0515432357788086,
+      "learning_rate": 4.737405551147791e-06,
+      "loss": 1.1263,
+      "step": 2697
+    },
+    {
+      "epoch": 2.0420056764427623,
+      "grad_norm": 2.0565128326416016,
+      "learning_rate": 4.7305817746142186e-06,
+      "loss": 1.0697,
+      "step": 2698
+    },
+    {
+      "epoch": 2.042762535477767,
+      "grad_norm": 2.287740707397461,
+      "learning_rate": 4.723761395311858e-06,
+      "loss": 1.0616,
+      "step": 2699
+    },
+    {
+      "epoch": 2.043519394512772,
+      "grad_norm": 2.0816521644592285,
+      "learning_rate": 4.716944417637024e-06,
+      "loss": 1.0425,
+      "step": 2700
+    },
+    {
+      "epoch": 2.0442762535477765,
+      "grad_norm": 2.117865562438965,
+      "learning_rate": 4.710130845983837e-06,
+      "loss": 1.1141,
+      "step": 2701
+    },
+    {
+      "epoch": 2.0450331125827814,
+      "grad_norm": 1.9112534523010254,
+      "learning_rate": 4.703320684744216e-06,
+      "loss": 1.06,
+      "step": 2702
+    },
+    {
+      "epoch": 2.0457899716177863,
+      "grad_norm": 2.1456851959228516,
+      "learning_rate": 4.696513938307894e-06,
+      "loss": 1.0512,
+      "step": 2703
+    },
+    {
+      "epoch": 2.0465468306527907,
+      "grad_norm": 2.6872496604919434,
+      "learning_rate": 4.689710611062389e-06,
+      "loss": 1.0907,
+      "step": 2704
+    },
+    {
+      "epoch": 2.0473036896877956,
+      "grad_norm": 2.116586446762085,
+      "learning_rate": 4.682910707393024e-06,
+      "loss": 1.1179,
+      "step": 2705
+    },
+    {
+      "epoch": 2.0480605487228005,
+      "grad_norm": 2.1400527954101562,
+      "learning_rate": 4.676114231682915e-06,
+      "loss": 1.0673,
+      "step": 2706
+    },
+    {
+      "epoch": 2.048817407757805,
+      "grad_norm": 2.1281938552856445,
+      "learning_rate": 4.669321188312969e-06,
+      "loss": 1.0719,
+      "step": 2707
+    },
+    {
+      "epoch": 2.04957426679281,
+      "grad_norm": 2.1432082653045654,
+      "learning_rate": 4.662531581661873e-06,
+      "loss": 1.0844,
+      "step": 2708
+    },
+    {
+      "epoch": 2.0503311258278147,
+      "grad_norm": 1.9638357162475586,
+      "learning_rate": 4.655745416106105e-06,
+      "loss": 1.0379,
+      "step": 2709
+    },
+    {
+      "epoch": 2.051087984862819,
+      "grad_norm": 2.069023609161377,
+      "learning_rate": 4.648962696019928e-06,
+      "loss": 1.0808,
+      "step": 2710
+    },
+    {
+      "epoch": 2.051844843897824,
+      "grad_norm": 1.974176287651062,
+      "learning_rate": 4.6421834257753745e-06,
+      "loss": 1.1431,
+      "step": 2711
+    },
+    {
+      "epoch": 2.052601702932829,
+      "grad_norm": 2.0162038803100586,
+      "learning_rate": 4.635407609742265e-06,
+      "loss": 1.0715,
+      "step": 2712
+    },
+    {
+      "epoch": 2.0533585619678334,
+      "grad_norm": 2.0583693981170654,
+      "learning_rate": 4.628635252288178e-06,
+      "loss": 1.0583,
+      "step": 2713
+    },
+    {
+      "epoch": 2.0541154210028383,
+      "grad_norm": 1.934477686882019,
+      "learning_rate": 4.621866357778479e-06,
+      "loss": 1.071,
+      "step": 2714
+    },
+    {
+      "epoch": 2.054872280037843,
+      "grad_norm": 2.2252588272094727,
+      "learning_rate": 4.61510093057629e-06,
+      "loss": 1.1205,
+      "step": 2715
+    },
+    {
+      "epoch": 2.0556291390728476,
+      "grad_norm": 2.0399527549743652,
+      "learning_rate": 4.608338975042509e-06,
+      "loss": 1.1,
+      "step": 2716
+    },
+    {
+      "epoch": 2.0563859981078525,
+      "grad_norm": 2.1194961071014404,
+      "learning_rate": 4.601580495535781e-06,
+      "loss": 1.0566,
+      "step": 2717
+    },
+    {
+      "epoch": 2.057142857142857,
+      "grad_norm": 2.1798765659332275,
+      "learning_rate": 4.594825496412527e-06,
+      "loss": 1.07,
+      "step": 2718
+    },
+    {
+      "epoch": 2.057899716177862,
+      "grad_norm": 2.00516414642334,
+      "learning_rate": 4.588073982026908e-06,
+      "loss": 1.0404,
+      "step": 2719
+    },
+    {
+      "epoch": 2.0586565752128667,
+      "grad_norm": 2.1549298763275146,
+      "learning_rate": 4.581325956730851e-06,
+      "loss": 1.0873,
+      "step": 2720
+    },
+    {
+      "epoch": 2.059413434247871,
+      "grad_norm": 2.3754074573516846,
+      "learning_rate": 4.574581424874031e-06,
+      "loss": 1.0917,
+      "step": 2721
+    },
+    {
+      "epoch": 2.060170293282876,
+      "grad_norm": 2.118363857269287,
+      "learning_rate": 4.56784039080387e-06,
+      "loss": 1.0864,
+      "step": 2722
+    },
+    {
+      "epoch": 2.060927152317881,
+      "grad_norm": 1.9879770278930664,
+      "learning_rate": 4.561102858865542e-06,
+      "loss": 1.047,
+      "step": 2723
+    },
+    {
+      "epoch": 2.0616840113528854,
+      "grad_norm": 2.0962250232696533,
+      "learning_rate": 4.554368833401944e-06,
+      "loss": 1.0803,
+      "step": 2724
+    },
+    {
+      "epoch": 2.0624408703878903,
+      "grad_norm": 2.095574378967285,
+      "learning_rate": 4.547638318753733e-06,
+      "loss": 1.1101,
+      "step": 2725
+    },
+    {
+      "epoch": 2.063197729422895,
+      "grad_norm": 2.2542734146118164,
+      "learning_rate": 4.540911319259297e-06,
+      "loss": 1.0672,
+      "step": 2726
+    },
+    {
+      "epoch": 2.0639545884578996,
+      "grad_norm": 2.1071441173553467,
+      "learning_rate": 4.534187839254755e-06,
+      "loss": 1.0295,
+      "step": 2727
+    },
+    {
+      "epoch": 2.0647114474929045,
+      "grad_norm": 2.2289743423461914,
+      "learning_rate": 4.527467883073962e-06,
+      "loss": 1.0759,
+      "step": 2728
+    },
+    {
+      "epoch": 2.0654683065279094,
+      "grad_norm": 2.20210862159729,
+      "learning_rate": 4.520751455048502e-06,
+      "loss": 1.0952,
+      "step": 2729
+    },
+    {
+      "epoch": 2.066225165562914,
+      "grad_norm": 2.0890111923217773,
+      "learning_rate": 4.5140385595076795e-06,
+      "loss": 1.1055,
+      "step": 2730
+    },
+    {
+      "epoch": 2.0669820245979187,
+      "grad_norm": 2.0093884468078613,
+      "learning_rate": 4.507329200778518e-06,
+      "loss": 1.0836,
+      "step": 2731
+    },
+    {
+      "epoch": 2.0677388836329236,
+      "grad_norm": 2.171649932861328,
+      "learning_rate": 4.500623383185774e-06,
+      "loss": 1.0794,
+      "step": 2732
+    },
+    {
+      "epoch": 2.068495742667928,
+      "grad_norm": 1.8796758651733398,
+      "learning_rate": 4.493921111051916e-06,
+      "loss": 1.084,
+      "step": 2733
+    },
+    {
+      "epoch": 2.069252601702933,
+      "grad_norm": 2.1491153240203857,
+      "learning_rate": 4.487222388697128e-06,
+      "loss": 1.0629,
+      "step": 2734
+    },
+    {
+      "epoch": 2.0700094607379373,
+      "grad_norm": 2.3227274417877197,
+      "learning_rate": 4.4805272204392965e-06,
+      "loss": 1.0901,
+      "step": 2735
+    },
+    {
+      "epoch": 2.0707663197729422,
+      "grad_norm": 2.256610631942749,
+      "learning_rate": 4.47383561059403e-06,
+      "loss": 1.1046,
+      "step": 2736
+    },
+    {
+      "epoch": 2.071523178807947,
+      "grad_norm": 1.9754210710525513,
+      "learning_rate": 4.467147563474642e-06,
+      "loss": 1.0433,
+      "step": 2737
+    },
+    {
+      "epoch": 2.0722800378429516,
+      "grad_norm": 2.2307772636413574,
+      "learning_rate": 4.460463083392139e-06,
+      "loss": 1.079,
+      "step": 2738
+    },
+    {
+      "epoch": 2.0730368968779564,
+      "grad_norm": 1.8400083780288696,
+      "learning_rate": 4.453782174655236e-06,
+      "loss": 1.0615,
+      "step": 2739
+    },
+    {
+      "epoch": 2.0737937559129613,
+      "grad_norm": 1.9424253702163696,
+      "learning_rate": 4.447104841570351e-06,
+      "loss": 1.0318,
+      "step": 2740
+    },
+    {
+      "epoch": 2.0745506149479658,
+      "grad_norm": 2.008769989013672,
+      "learning_rate": 4.440431088441582e-06,
+      "loss": 1.0861,
+      "step": 2741
+    },
+    {
+      "epoch": 2.0753074739829707,
+      "grad_norm": 2.024463415145874,
+      "learning_rate": 4.4337609195707325e-06,
+      "loss": 1.0406,
+      "step": 2742
+    },
+    {
+      "epoch": 2.0760643330179755,
+      "grad_norm": 2.0557620525360107,
+      "learning_rate": 4.4270943392572924e-06,
+      "loss": 1.0983,
+      "step": 2743
+    },
+    {
+      "epoch": 2.07682119205298,
+      "grad_norm": 2.7445693016052246,
+      "learning_rate": 4.420431351798441e-06,
+      "loss": 1.1253,
+      "step": 2744
+    },
+    {
+      "epoch": 2.077578051087985,
+      "grad_norm": 2.1181790828704834,
+      "learning_rate": 4.413771961489035e-06,
+      "loss": 1.0808,
+      "step": 2745
+    },
+    {
+      "epoch": 2.0783349101229898,
+      "grad_norm": 2.035220146179199,
+      "learning_rate": 4.4071161726216116e-06,
+      "loss": 1.0242,
+      "step": 2746
+    },
+    {
+      "epoch": 2.079091769157994,
+      "grad_norm": 2.0690808296203613,
+      "learning_rate": 4.4004639894863945e-06,
+      "loss": 1.072,
+      "step": 2747
+    },
+    {
+      "epoch": 2.079848628192999,
+      "grad_norm": 2.145026683807373,
+      "learning_rate": 4.39381541637128e-06,
+      "loss": 1.0455,
+      "step": 2748
+    },
+    {
+      "epoch": 2.080605487228004,
+      "grad_norm": 1.9162312746047974,
+      "learning_rate": 4.387170457561837e-06,
+      "loss": 1.073,
+      "step": 2749
+    },
+    {
+      "epoch": 2.0813623462630084,
+      "grad_norm": 2.1280105113983154,
+      "learning_rate": 4.380529117341305e-06,
+      "loss": 1.151,
+      "step": 2750
+    },
+    {
+      "epoch": 2.0821192052980133,
+      "grad_norm": 2.1977860927581787,
+      "learning_rate": 4.373891399990595e-06,
+      "loss": 1.0732,
+      "step": 2751
+    },
+    {
+      "epoch": 2.0828760643330178,
+      "grad_norm": 2.101363182067871,
+      "learning_rate": 4.367257309788268e-06,
+      "loss": 1.0534,
+      "step": 2752
+    },
+    {
+      "epoch": 2.0836329233680226,
+      "grad_norm": 2.0512733459472656,
+      "learning_rate": 4.360626851010562e-06,
+      "loss": 1.0613,
+      "step": 2753
+    },
+    {
+      "epoch": 2.0843897824030275,
+      "grad_norm": 2.4588279724121094,
+      "learning_rate": 4.35400002793137e-06,
+      "loss": 1.058,
+      "step": 2754
+    },
+    {
+      "epoch": 2.085146641438032,
+      "grad_norm": 2.0685465335845947,
+      "learning_rate": 4.347376844822242e-06,
+      "loss": 1.0532,
+      "step": 2755
+    },
+    {
+      "epoch": 2.085903500473037,
+      "grad_norm": 2.1436235904693604,
+      "learning_rate": 4.340757305952384e-06,
+      "loss": 1.0618,
+      "step": 2756
+    },
+    {
+      "epoch": 2.0866603595080417,
+      "grad_norm": 2.1337039470672607,
+      "learning_rate": 4.334141415588644e-06,
+      "loss": 1.0852,
+      "step": 2757
+    },
+    {
+      "epoch": 2.087417218543046,
+      "grad_norm": 1.9831031560897827,
+      "learning_rate": 4.3275291779955245e-06,
+      "loss": 1.1146,
+      "step": 2758
+    },
+    {
+      "epoch": 2.088174077578051,
+      "grad_norm": 2.2047150135040283,
+      "learning_rate": 4.320920597435174e-06,
+      "loss": 1.0817,
+      "step": 2759
+    },
+    {
+      "epoch": 2.088930936613056,
+      "grad_norm": 2.065563201904297,
+      "learning_rate": 4.3143156781673846e-06,
+      "loss": 1.1424,
+      "step": 2760
+    },
+    {
+      "epoch": 2.0896877956480604,
+      "grad_norm": 2.1660046577453613,
+      "learning_rate": 4.307714424449583e-06,
+      "loss": 1.1019,
+      "step": 2761
+    },
+    {
+      "epoch": 2.0904446546830653,
+      "grad_norm": 2.0400032997131348,
+      "learning_rate": 4.301116840536844e-06,
+      "loss": 1.0345,
+      "step": 2762
+    },
+    {
+      "epoch": 2.09120151371807,
+      "grad_norm": 1.9771476984024048,
+      "learning_rate": 4.29452293068186e-06,
+      "loss": 1.0576,
+      "step": 2763
+    },
+    {
+      "epoch": 2.0919583727530746,
+      "grad_norm": 2.2237024307250977,
+      "learning_rate": 4.287932699134973e-06,
+      "loss": 1.0808,
+      "step": 2764
+    },
+    {
+      "epoch": 2.0927152317880795,
+      "grad_norm": 2.0638787746429443,
+      "learning_rate": 4.281346150144139e-06,
+      "loss": 1.0315,
+      "step": 2765
+    },
+    {
+      "epoch": 2.0934720908230844,
+      "grad_norm": 2.370335817337036,
+      "learning_rate": 4.27476328795495e-06,
+      "loss": 1.1243,
+      "step": 2766
+    },
+    {
+      "epoch": 2.094228949858089,
+      "grad_norm": 2.069380521774292,
+      "learning_rate": 4.268184116810623e-06,
+      "loss": 1.0801,
+      "step": 2767
+    },
+    {
+      "epoch": 2.0949858088930937,
+      "grad_norm": 2.310030221939087,
+      "learning_rate": 4.261608640951981e-06,
+      "loss": 1.0666,
+      "step": 2768
+    },
+    {
+      "epoch": 2.0957426679280986,
+      "grad_norm": 2.0185890197753906,
+      "learning_rate": 4.255036864617483e-06,
+      "loss": 1.0817,
+      "step": 2769
+    },
+    {
+      "epoch": 2.096499526963103,
+      "grad_norm": 2.0323379039764404,
+      "learning_rate": 4.248468792043194e-06,
+      "loss": 1.057,
+      "step": 2770
+    },
+    {
+      "epoch": 2.097256385998108,
+      "grad_norm": 2.362914562225342,
+      "learning_rate": 4.241904427462797e-06,
+      "loss": 1.0846,
+      "step": 2771
+    },
+    {
+      "epoch": 2.0980132450331124,
+      "grad_norm": 2.203740119934082,
+      "learning_rate": 4.235343775107575e-06,
+      "loss": 1.0565,
+      "step": 2772
+    },
+    {
+      "epoch": 2.0987701040681173,
+      "grad_norm": 2.006248712539673,
+      "learning_rate": 4.22878683920643e-06,
+      "loss": 1.1263,
+      "step": 2773
+    },
+    {
+      "epoch": 2.099526963103122,
+      "grad_norm": 2.120445489883423,
+      "learning_rate": 4.222233623985858e-06,
+      "loss": 1.0629,
+      "step": 2774
+    },
+    {
+      "epoch": 2.1002838221381266,
+      "grad_norm": 2.015179395675659,
+      "learning_rate": 4.2156841336699625e-06,
+      "loss": 1.0304,
+      "step": 2775
+    },
+    {
+      "epoch": 2.1010406811731315,
+      "grad_norm": 2.1381468772888184,
+      "learning_rate": 4.209138372480447e-06,
+      "loss": 1.0669,
+      "step": 2776
+    },
+    {
+      "epoch": 2.1017975402081364,
+      "grad_norm": 2.2807891368865967,
+      "learning_rate": 4.202596344636609e-06,
+      "loss": 1.0635,
+      "step": 2777
+    },
+    {
+      "epoch": 2.102554399243141,
+      "grad_norm": 2.1004843711853027,
+      "learning_rate": 4.196058054355347e-06,
+      "loss": 1.1306,
+      "step": 2778
+    },
+    {
+      "epoch": 2.1033112582781457,
+      "grad_norm": 2.092963695526123,
+      "learning_rate": 4.189523505851129e-06,
+      "loss": 1.0561,
+      "step": 2779
+    },
+    {
+      "epoch": 2.1040681173131506,
+      "grad_norm": 2.0627875328063965,
+      "learning_rate": 4.1829927033360314e-06,
+      "loss": 1.0671,
+      "step": 2780
+    },
+    {
+      "epoch": 2.104824976348155,
+      "grad_norm": 2.0852344036102295,
+      "learning_rate": 4.17646565101971e-06,
+      "loss": 1.056,
+      "step": 2781
+    },
+    {
+      "epoch": 2.10558183538316,
+      "grad_norm": 2.0920495986938477,
+      "learning_rate": 4.1699423531094065e-06,
+      "loss": 1.0415,
+      "step": 2782
+    },
+    {
+      "epoch": 2.106338694418165,
+      "grad_norm": 2.3179705142974854,
+      "learning_rate": 4.163422813809934e-06,
+      "loss": 1.0648,
+      "step": 2783
+    },
+    {
+      "epoch": 2.1070955534531692,
+      "grad_norm": 2.0878725051879883,
+      "learning_rate": 4.156907037323696e-06,
+      "loss": 1.0996,
+      "step": 2784
+    },
+    {
+      "epoch": 2.107852412488174,
+      "grad_norm": 2.1616759300231934,
+      "learning_rate": 4.1503950278506565e-06,
+      "loss": 1.0473,
+      "step": 2785
+    },
+    {
+      "epoch": 2.108609271523179,
+      "grad_norm": 2.314814805984497,
+      "learning_rate": 4.1438867895883555e-06,
+      "loss": 1.0717,
+      "step": 2786
+    },
+    {
+      "epoch": 2.1093661305581834,
+      "grad_norm": 2.105376958847046,
+      "learning_rate": 4.137382326731906e-06,
+      "loss": 1.1177,
+      "step": 2787
+    },
+    {
+      "epoch": 2.1101229895931883,
+      "grad_norm": 2.18996000289917,
+      "learning_rate": 4.130881643473987e-06,
+      "loss": 1.0923,
+      "step": 2788
+    },
+    {
+      "epoch": 2.1108798486281932,
+      "grad_norm": 2.0627288818359375,
+      "learning_rate": 4.124384744004844e-06,
+      "loss": 1.0307,
+      "step": 2789
+    },
+    {
+      "epoch": 2.1116367076631977,
+      "grad_norm": 2.2541861534118652,
+      "learning_rate": 4.117891632512271e-06,
+      "loss": 1.0543,
+      "step": 2790
+    },
+    {
+      "epoch": 2.1123935666982026,
+      "grad_norm": 2.0544228553771973,
+      "learning_rate": 4.111402313181631e-06,
+      "loss": 1.0987,
+      "step": 2791
+    },
+    {
+      "epoch": 2.113150425733207,
+      "grad_norm": 2.1496474742889404,
+      "learning_rate": 4.1049167901958454e-06,
+      "loss": 1.0422,
+      "step": 2792
+    },
+    {
+      "epoch": 2.113907284768212,
+      "grad_norm": 2.1363749504089355,
+      "learning_rate": 4.098435067735377e-06,
+      "loss": 1.0371,
+      "step": 2793
+    },
+    {
+      "epoch": 2.1146641438032168,
+      "grad_norm": 2.166128635406494,
+      "learning_rate": 4.091957149978247e-06,
+      "loss": 1.056,
+      "step": 2794
+    },
+    {
+      "epoch": 2.115421002838221,
+      "grad_norm": 2.3086111545562744,
+      "learning_rate": 4.085483041100028e-06,
+      "loss": 1.0582,
+      "step": 2795
+    },
+    {
+      "epoch": 2.116177861873226,
+      "grad_norm": 2.0368103981018066,
+      "learning_rate": 4.079012745273822e-06,
+      "loss": 0.9679,
+      "step": 2796
+    },
+    {
+      "epoch": 2.116934720908231,
+      "grad_norm": 2.146679639816284,
+      "learning_rate": 4.072546266670289e-06,
+      "loss": 1.0472,
+      "step": 2797
+    },
+    {
+      "epoch": 2.1176915799432354,
+      "grad_norm": 2.188101291656494,
+      "learning_rate": 4.0660836094576215e-06,
+      "loss": 1.1283,
+      "step": 2798
+    },
+    {
+      "epoch": 2.1184484389782403,
+      "grad_norm": 2.099888563156128,
+      "learning_rate": 4.059624777801554e-06,
+      "loss": 1.0708,
+      "step": 2799
+    },
+    {
+      "epoch": 2.119205298013245,
+      "grad_norm": 2.088252544403076,
+      "learning_rate": 4.053169775865346e-06,
+      "loss": 1.0619,
+      "step": 2800
+    },
+    {
+      "epoch": 2.1199621570482496,
+      "grad_norm": 2.0278518199920654,
+      "learning_rate": 4.046718607809791e-06,
+      "loss": 1.0549,
+      "step": 2801
+    },
+    {
+      "epoch": 2.1207190160832545,
+      "grad_norm": 1.9221056699752808,
+      "learning_rate": 4.040271277793217e-06,
+      "loss": 1.0776,
+      "step": 2802
+    },
+    {
+      "epoch": 2.1214758751182594,
+      "grad_norm": 2.296339511871338,
+      "learning_rate": 4.033827789971474e-06,
+      "loss": 1.0686,
+      "step": 2803
+    },
+    {
+      "epoch": 2.122232734153264,
+      "grad_norm": 2.1365742683410645,
+      "learning_rate": 4.027388148497936e-06,
+      "loss": 1.0812,
+      "step": 2804
+    },
+    {
+      "epoch": 2.1229895931882687,
+      "grad_norm": 1.9683605432510376,
+      "learning_rate": 4.020952357523498e-06,
+      "loss": 1.0168,
+      "step": 2805
+    },
+    {
+      "epoch": 2.1237464522232736,
+      "grad_norm": 2.0199337005615234,
+      "learning_rate": 4.014520421196579e-06,
+      "loss": 1.1035,
+      "step": 2806
+    },
+    {
+      "epoch": 2.124503311258278,
+      "grad_norm": 2.0269358158111572,
+      "learning_rate": 4.008092343663094e-06,
+      "loss": 1.0973,
+      "step": 2807
+    },
+    {
+      "epoch": 2.125260170293283,
+      "grad_norm": 2.286689519882202,
+      "learning_rate": 4.001668129066491e-06,
+      "loss": 1.0882,
+      "step": 2808
+    },
+    {
+      "epoch": 2.1260170293282874,
+      "grad_norm": 2.257807731628418,
+      "learning_rate": 3.995247781547721e-06,
+      "loss": 1.0877,
+      "step": 2809
+    },
+    {
+      "epoch": 2.1267738883632923,
+      "grad_norm": 2.049635171890259,
+      "learning_rate": 3.98883130524524e-06,
+      "loss": 1.0924,
+      "step": 2810
+    },
+    {
+      "epoch": 2.127530747398297,
+      "grad_norm": 2.112349033355713,
+      "learning_rate": 3.982418704295016e-06,
+      "loss": 1.0931,
+      "step": 2811
+    },
+    {
+      "epoch": 2.1282876064333016,
+      "grad_norm": 2.0468220710754395,
+      "learning_rate": 3.9760099828305104e-06,
+      "loss": 1.0842,
+      "step": 2812
+    },
+    {
+      "epoch": 2.1290444654683065,
+      "grad_norm": 2.0540926456451416,
+      "learning_rate": 3.969605144982682e-06,
+      "loss": 1.0924,
+      "step": 2813
+    },
+    {
+      "epoch": 2.1298013245033114,
+      "grad_norm": 2.1668741703033447,
+      "learning_rate": 3.963204194879998e-06,
+      "loss": 1.1271,
+      "step": 2814
+    },
+    {
+      "epoch": 2.130558183538316,
+      "grad_norm": 1.9331365823745728,
+      "learning_rate": 3.956807136648411e-06,
+      "loss": 1.1066,
+      "step": 2815
+    },
+    {
+      "epoch": 2.1313150425733207,
+      "grad_norm": 1.9183405637741089,
+      "learning_rate": 3.950413974411367e-06,
+      "loss": 1.1018,
+      "step": 2816
+    },
+    {
+      "epoch": 2.1320719016083256,
+      "grad_norm": 1.9769048690795898,
+      "learning_rate": 3.944024712289805e-06,
+      "loss": 1.0565,
+      "step": 2817
+    },
+    {
+      "epoch": 2.13282876064333,
+      "grad_norm": 1.961674690246582,
+      "learning_rate": 3.93763935440214e-06,
+      "loss": 1.0816,
+      "step": 2818
+    },
+    {
+      "epoch": 2.133585619678335,
+      "grad_norm": 2.0646157264709473,
+      "learning_rate": 3.931257904864283e-06,
+      "loss": 1.0373,
+      "step": 2819
+    },
+    {
+      "epoch": 2.13434247871334,
+      "grad_norm": 2.243910074234009,
+      "learning_rate": 3.92488036778961e-06,
+      "loss": 1.0423,
+      "step": 2820
+    },
+    {
+      "epoch": 2.1350993377483443,
+      "grad_norm": 2.192121744155884,
+      "learning_rate": 3.91850674728899e-06,
+      "loss": 1.0402,
+      "step": 2821
+    },
+    {
+      "epoch": 2.135856196783349,
+      "grad_norm": 2.0529327392578125,
+      "learning_rate": 3.912137047470764e-06,
+      "loss": 1.0676,
+      "step": 2822
+    },
+    {
+      "epoch": 2.136613055818354,
+      "grad_norm": 2.1558024883270264,
+      "learning_rate": 3.9057712724407366e-06,
+      "loss": 1.06,
+      "step": 2823
+    },
+    {
+      "epoch": 2.1373699148533585,
+      "grad_norm": 1.8848477602005005,
+      "learning_rate": 3.899409426302193e-06,
+      "loss": 1.0668,
+      "step": 2824
+    },
+    {
+      "epoch": 2.1381267738883634,
+      "grad_norm": 2.0971271991729736,
+      "learning_rate": 3.893051513155881e-06,
+      "loss": 1.1362,
+      "step": 2825
+    },
+    {
+      "epoch": 2.138883632923368,
+      "grad_norm": 2.3545618057250977,
+      "learning_rate": 3.88669753710002e-06,
+      "loss": 1.1014,
+      "step": 2826
+    },
+    {
+      "epoch": 2.1396404919583727,
+      "grad_norm": 2.0267715454101562,
+      "learning_rate": 3.880347502230277e-06,
+      "loss": 1.0955,
+      "step": 2827
+    },
+    {
+      "epoch": 2.1403973509933776,
+      "grad_norm": 2.065638780593872,
+      "learning_rate": 3.874001412639796e-06,
+      "loss": 1.0732,
+      "step": 2828
+    },
+    {
+      "epoch": 2.141154210028382,
+      "grad_norm": 2.221348285675049,
+      "learning_rate": 3.867659272419163e-06,
+      "loss": 1.0891,
+      "step": 2829
+    },
+    {
+      "epoch": 2.141911069063387,
+      "grad_norm": 2.0352323055267334,
+      "learning_rate": 3.861321085656425e-06,
+      "loss": 1.0615,
+      "step": 2830
+    },
+    {
+      "epoch": 2.142667928098392,
+      "grad_norm": 2.294567584991455,
+      "learning_rate": 3.854986856437086e-06,
+      "loss": 1.0886,
+      "step": 2831
+    },
+    {
+      "epoch": 2.1434247871333962,
+      "grad_norm": 2.132350444793701,
+      "learning_rate": 3.848656588844089e-06,
+      "loss": 1.0932,
+      "step": 2832
+    },
+    {
+      "epoch": 2.144181646168401,
+      "grad_norm": 2.0099170207977295,
+      "learning_rate": 3.842330286957837e-06,
+      "loss": 1.1081,
+      "step": 2833
+    },
+    {
+      "epoch": 2.144938505203406,
+      "grad_norm": 2.181610584259033,
+      "learning_rate": 3.836007954856154e-06,
+      "loss": 1.1125,
+      "step": 2834
+    },
+    {
+      "epoch": 2.1456953642384105,
+      "grad_norm": 2.204340934753418,
+      "learning_rate": 3.829689596614324e-06,
+      "loss": 1.1074,
+      "step": 2835
+    },
+    {
+      "epoch": 2.1464522232734153,
+      "grad_norm": 2.0872256755828857,
+      "learning_rate": 3.823375216305066e-06,
+      "loss": 1.0709,
+      "step": 2836
+    },
+    {
+      "epoch": 2.1472090823084202,
+      "grad_norm": 2.192131757736206,
+      "learning_rate": 3.8170648179985324e-06,
+      "loss": 1.0539,
+      "step": 2837
+    },
+    {
+      "epoch": 2.1479659413434247,
+      "grad_norm": 1.9136378765106201,
+      "learning_rate": 3.810758405762311e-06,
+      "loss": 1.0931,
+      "step": 2838
+    },
+    {
+      "epoch": 2.1487228003784296,
+      "grad_norm": 2.390619993209839,
+      "learning_rate": 3.8044559836614203e-06,
+      "loss": 1.0645,
+      "step": 2839
+    },
+    {
+      "epoch": 2.1494796594134344,
+      "grad_norm": 2.138697862625122,
+      "learning_rate": 3.798157555758304e-06,
+      "loss": 1.084,
+      "step": 2840
+    },
+    {
+      "epoch": 2.150236518448439,
+      "grad_norm": 1.9692342281341553,
+      "learning_rate": 3.791863126112828e-06,
+      "loss": 1.0735,
+      "step": 2841
+    },
+    {
+      "epoch": 2.1509933774834438,
+      "grad_norm": 2.02774715423584,
+      "learning_rate": 3.78557269878229e-06,
+      "loss": 1.0853,
+      "step": 2842
+    },
+    {
+      "epoch": 2.151750236518448,
+      "grad_norm": 1.9480324983596802,
+      "learning_rate": 3.779286277821402e-06,
+      "loss": 1.0504,
+      "step": 2843
+    },
+    {
+      "epoch": 2.152507095553453,
+      "grad_norm": 2.0497875213623047,
+      "learning_rate": 3.773003867282301e-06,
+      "loss": 1.0475,
+      "step": 2844
+    },
+    {
+      "epoch": 2.153263954588458,
+      "grad_norm": 2.0127995014190674,
+      "learning_rate": 3.766725471214524e-06,
+      "loss": 1.0773,
+      "step": 2845
+    },
+    {
+      "epoch": 2.1540208136234624,
+      "grad_norm": 1.920920729637146,
+      "learning_rate": 3.760451093665034e-06,
+      "loss": 1.0747,
+      "step": 2846
+    },
+    {
+      "epoch": 2.1547776726584673,
+      "grad_norm": 1.9610087871551514,
+      "learning_rate": 3.754180738678201e-06,
+      "loss": 1.0929,
+      "step": 2847
+    },
+    {
+      "epoch": 2.155534531693472,
+      "grad_norm": 2.1988742351531982,
+      "learning_rate": 3.7479144102957955e-06,
+      "loss": 1.0426,
+      "step": 2848
+    },
+    {
+      "epoch": 2.1562913907284766,
+      "grad_norm": 2.0719704627990723,
+      "learning_rate": 3.7416521125569987e-06,
+      "loss": 1.0965,
+      "step": 2849
+    },
+    {
+      "epoch": 2.1570482497634815,
+      "grad_norm": 2.073084592819214,
+      "learning_rate": 3.7353938494983966e-06,
+      "loss": 1.0428,
+      "step": 2850
+    },
+    {
+      "epoch": 2.1578051087984864,
+      "grad_norm": 1.9803792238235474,
+      "learning_rate": 3.729139625153964e-06,
+      "loss": 1.0724,
+      "step": 2851
+    },
+    {
+      "epoch": 2.158561967833491,
+      "grad_norm": 1.9936349391937256,
+      "learning_rate": 3.72288944355508e-06,
+      "loss": 1.0278,
+      "step": 2852
+    },
+    {
+      "epoch": 2.1593188268684957,
+      "grad_norm": 2.1690564155578613,
+      "learning_rate": 3.7166433087305177e-06,
+      "loss": 1.044,
+      "step": 2853
+    },
+    {
+      "epoch": 2.1600756859035006,
+      "grad_norm": 2.125483512878418,
+      "learning_rate": 3.7104012247064436e-06,
+      "loss": 1.0493,
+      "step": 2854
+    },
+    {
+      "epoch": 2.160832544938505,
+      "grad_norm": 2.031766653060913,
+      "learning_rate": 3.7041631955064067e-06,
+      "loss": 1.0746,
+      "step": 2855
+    },
+    {
+      "epoch": 2.16158940397351,
+      "grad_norm": 2.1385655403137207,
+      "learning_rate": 3.697929225151341e-06,
+      "loss": 1.0993,
+      "step": 2856
+    },
+    {
+      "epoch": 2.162346263008515,
+      "grad_norm": 2.363760471343994,
+      "learning_rate": 3.691699317659574e-06,
+      "loss": 1.0544,
+      "step": 2857
+    },
+    {
+      "epoch": 2.1631031220435193,
+      "grad_norm": 2.0311970710754395,
+      "learning_rate": 3.685473477046807e-06,
+      "loss": 1.0244,
+      "step": 2858
+    },
+    {
+      "epoch": 2.163859981078524,
+      "grad_norm": 2.2926740646362305,
+      "learning_rate": 3.679251707326123e-06,
+      "loss": 1.0813,
+      "step": 2859
+    },
+    {
+      "epoch": 2.164616840113529,
+      "grad_norm": 2.1094629764556885,
+      "learning_rate": 3.6730340125079804e-06,
+      "loss": 1.0729,
+      "step": 2860
+    },
+    {
+      "epoch": 2.1653736991485335,
+      "grad_norm": 2.2575571537017822,
+      "learning_rate": 3.6668203966002157e-06,
+      "loss": 1.1031,
+      "step": 2861
+    },
+    {
+      "epoch": 2.1661305581835384,
+      "grad_norm": 2.2293882369995117,
+      "learning_rate": 3.660610863608018e-06,
+      "loss": 1.0676,
+      "step": 2862
+    },
+    {
+      "epoch": 2.1668874172185433,
+      "grad_norm": 2.214388132095337,
+      "learning_rate": 3.6544054175339655e-06,
+      "loss": 1.1499,
+      "step": 2863
+    },
+    {
+      "epoch": 2.1676442762535477,
+      "grad_norm": 1.9468921422958374,
+      "learning_rate": 3.6482040623779925e-06,
+      "loss": 1.0726,
+      "step": 2864
+    },
+    {
+      "epoch": 2.1684011352885526,
+      "grad_norm": 2.0682532787323,
+      "learning_rate": 3.642006802137399e-06,
+      "loss": 1.0695,
+      "step": 2865
+    },
+    {
+      "epoch": 2.169157994323557,
+      "grad_norm": 2.1253714561462402,
+      "learning_rate": 3.6358136408068475e-06,
+      "loss": 1.116,
+      "step": 2866
+    },
+    {
+      "epoch": 2.169914853358562,
+      "grad_norm": 2.113579511642456,
+      "learning_rate": 3.6296245823783514e-06,
+      "loss": 1.0874,
+      "step": 2867
+    },
+    {
+      "epoch": 2.170671712393567,
+      "grad_norm": 1.9568238258361816,
+      "learning_rate": 3.623439630841282e-06,
+      "loss": 1.0363,
+      "step": 2868
+    },
+    {
+      "epoch": 2.1714285714285713,
+      "grad_norm": 2.3202977180480957,
+      "learning_rate": 3.6172587901823652e-06,
+      "loss": 1.0881,
+      "step": 2869
+    },
+    {
+      "epoch": 2.172185430463576,
+      "grad_norm": 2.232671022415161,
+      "learning_rate": 3.611082064385679e-06,
+      "loss": 1.0919,
+      "step": 2870
+    },
+    {
+      "epoch": 2.172942289498581,
+      "grad_norm": 1.9573837518692017,
+      "learning_rate": 3.6049094574326453e-06,
+      "loss": 1.017,
+      "step": 2871
+    },
+    {
+      "epoch": 2.1736991485335855,
+      "grad_norm": 2.110637664794922,
+      "learning_rate": 3.598740973302036e-06,
+      "loss": 1.1066,
+      "step": 2872
+    },
+    {
+      "epoch": 2.1744560075685904,
+      "grad_norm": 2.0418527126312256,
+      "learning_rate": 3.592576615969956e-06,
+      "loss": 1.0607,
+      "step": 2873
+    },
+    {
+      "epoch": 2.1752128666035953,
+      "grad_norm": 2.177208662033081,
+      "learning_rate": 3.5864163894098624e-06,
+      "loss": 1.0724,
+      "step": 2874
+    },
+    {
+      "epoch": 2.1759697256385997,
+      "grad_norm": 2.2924139499664307,
+      "learning_rate": 3.580260297592535e-06,
+      "loss": 1.0593,
+      "step": 2875
+    },
+    {
+      "epoch": 2.1767265846736046,
+      "grad_norm": 2.0772855281829834,
+      "learning_rate": 3.574108344486102e-06,
+      "loss": 1.0661,
+      "step": 2876
+    },
+    {
+      "epoch": 2.1774834437086095,
+      "grad_norm": 2.265467405319214,
+      "learning_rate": 3.5679605340560187e-06,
+      "loss": 1.0569,
+      "step": 2877
+    },
+    {
+      "epoch": 2.178240302743614,
+      "grad_norm": 2.154500722885132,
+      "learning_rate": 3.5618168702650713e-06,
+      "loss": 1.0332,
+      "step": 2878
+    },
+    {
+      "epoch": 2.178997161778619,
+      "grad_norm": 2.0559258460998535,
+      "learning_rate": 3.5556773570733666e-06,
+      "loss": 1.0697,
+      "step": 2879
+    },
+    {
+      "epoch": 2.1797540208136237,
+      "grad_norm": 2.4780728816986084,
+      "learning_rate": 3.5495419984383452e-06,
+      "loss": 1.0525,
+      "step": 2880
+    },
+    {
+      "epoch": 2.180510879848628,
+      "grad_norm": 2.0388307571411133,
+      "learning_rate": 3.543410798314767e-06,
+      "loss": 1.0224,
+      "step": 2881
+    },
+    {
+      "epoch": 2.181267738883633,
+      "grad_norm": 2.0700438022613525,
+      "learning_rate": 3.5372837606547056e-06,
+      "loss": 1.0795,
+      "step": 2882
+    },
+    {
+      "epoch": 2.1820245979186375,
+      "grad_norm": 2.08799409866333,
+      "learning_rate": 3.5311608894075606e-06,
+      "loss": 1.1147,
+      "step": 2883
+    },
+    {
+      "epoch": 2.1827814569536423,
+      "grad_norm": 2.04353928565979,
+      "learning_rate": 3.5250421885200357e-06,
+      "loss": 1.0693,
+      "step": 2884
+    },
+    {
+      "epoch": 2.1835383159886472,
+      "grad_norm": 2.1684114933013916,
+      "learning_rate": 3.5189276619361567e-06,
+      "loss": 1.0844,
+      "step": 2885
+    },
+    {
+      "epoch": 2.1842951750236517,
+      "grad_norm": 2.2811787128448486,
+      "learning_rate": 3.5128173135972515e-06,
+      "loss": 1.0921,
+      "step": 2886
+    },
+    {
+      "epoch": 2.1850520340586566,
+      "grad_norm": 2.293611764907837,
+      "learning_rate": 3.5067111474419603e-06,
+      "loss": 1.1276,
+      "step": 2887
+    },
+    {
+      "epoch": 2.1858088930936614,
+      "grad_norm": 1.9369990825653076,
+      "learning_rate": 3.5006091674062263e-06,
+      "loss": 1.0811,
+      "step": 2888
+    },
+    {
+      "epoch": 2.186565752128666,
+      "grad_norm": 2.1612861156463623,
+      "learning_rate": 3.494511377423291e-06,
+      "loss": 1.0987,
+      "step": 2889
+    },
+    {
+      "epoch": 2.1873226111636708,
+      "grad_norm": 2.301436424255371,
+      "learning_rate": 3.488417781423691e-06,
+      "loss": 1.1224,
+      "step": 2890
+    },
+    {
+      "epoch": 2.1880794701986757,
+      "grad_norm": 2.149083375930786,
+      "learning_rate": 3.482328383335271e-06,
+      "loss": 1.0906,
+      "step": 2891
+    },
+    {
+      "epoch": 2.18883632923368,
+      "grad_norm": 2.4687178134918213,
+      "learning_rate": 3.4762431870831625e-06,
+      "loss": 1.1381,
+      "step": 2892
+    },
+    {
+      "epoch": 2.189593188268685,
+      "grad_norm": 2.131269693374634,
+      "learning_rate": 3.4701621965897906e-06,
+      "loss": 1.0644,
+      "step": 2893
+    },
+    {
+      "epoch": 2.19035004730369,
+      "grad_norm": 2.0854032039642334,
+      "learning_rate": 3.464085415774874e-06,
+      "loss": 1.0703,
+      "step": 2894
+    },
+    {
+      "epoch": 2.1911069063386943,
+      "grad_norm": 1.988800287246704,
+      "learning_rate": 3.458012848555407e-06,
+      "loss": 1.0925,
+      "step": 2895
+    },
+    {
+      "epoch": 2.191863765373699,
+      "grad_norm": 2.0683155059814453,
+      "learning_rate": 3.451944498845673e-06,
+      "loss": 1.1212,
+      "step": 2896
+    },
+    {
+      "epoch": 2.192620624408704,
+      "grad_norm": 2.1208488941192627,
+      "learning_rate": 3.4458803705572385e-06,
+      "loss": 1.0917,
+      "step": 2897
+    },
+    {
+      "epoch": 2.1933774834437085,
+      "grad_norm": 1.9864528179168701,
+      "learning_rate": 3.4398204675989504e-06,
+      "loss": 1.1095,
+      "step": 2898
+    },
+    {
+      "epoch": 2.1941343424787134,
+      "grad_norm": 2.0708682537078857,
+      "learning_rate": 3.4337647938769283e-06,
+      "loss": 1.0989,
+      "step": 2899
+    },
+    {
+      "epoch": 2.194891201513718,
+      "grad_norm": 2.229597330093384,
+      "learning_rate": 3.4277133532945704e-06,
+      "loss": 1.1137,
+      "step": 2900
+    },
+    {
+      "epoch": 2.1956480605487227,
+      "grad_norm": 2.039870500564575,
+      "learning_rate": 3.4216661497525372e-06,
+      "loss": 1.0866,
+      "step": 2901
+    },
+    {
+      "epoch": 2.1964049195837276,
+      "grad_norm": 2.037367582321167,
+      "learning_rate": 3.4156231871487706e-06,
+      "loss": 1.0947,
+      "step": 2902
+    },
+    {
+      "epoch": 2.197161778618732,
+      "grad_norm": 2.3312087059020996,
+      "learning_rate": 3.4095844693784647e-06,
+      "loss": 1.0883,
+      "step": 2903
+    },
+    {
+      "epoch": 2.197918637653737,
+      "grad_norm": 2.1165080070495605,
+      "learning_rate": 3.4035500003340886e-06,
+      "loss": 1.07,
+      "step": 2904
+    },
+    {
+      "epoch": 2.198675496688742,
+      "grad_norm": 2.1637613773345947,
+      "learning_rate": 3.3975197839053727e-06,
+      "loss": 1.0012,
+      "step": 2905
+    },
+    {
+      "epoch": 2.1994323557237463,
+      "grad_norm": 2.1280291080474854,
+      "learning_rate": 3.3914938239792956e-06,
+      "loss": 1.0525,
+      "step": 2906
+    },
+    {
+      "epoch": 2.200189214758751,
+      "grad_norm": 2.1883440017700195,
+      "learning_rate": 3.385472124440102e-06,
+      "loss": 1.0466,
+      "step": 2907
+    },
+    {
+      "epoch": 2.200946073793756,
+      "grad_norm": 2.120882272720337,
+      "learning_rate": 3.3794546891692883e-06,
+      "loss": 1.0268,
+      "step": 2908
+    },
+    {
+      "epoch": 2.2017029328287605,
+      "grad_norm": 2.141380786895752,
+      "learning_rate": 3.3734415220456036e-06,
+      "loss": 1.0695,
+      "step": 2909
+    },
+    {
+      "epoch": 2.2024597918637654,
+      "grad_norm": 1.9636356830596924,
+      "learning_rate": 3.3674326269450386e-06,
+      "loss": 1.035,
+      "step": 2910
+    },
+    {
+      "epoch": 2.2032166508987703,
+      "grad_norm": 2.227339506149292,
+      "learning_rate": 3.361428007740842e-06,
+      "loss": 1.1143,
+      "step": 2911
+    },
+    {
+      "epoch": 2.2039735099337747,
+      "grad_norm": 2.206693172454834,
+      "learning_rate": 3.3554276683034933e-06,
+      "loss": 1.075,
+      "step": 2912
+    },
+    {
+      "epoch": 2.2047303689687796,
+      "grad_norm": 2.3205721378326416,
+      "learning_rate": 3.349431612500721e-06,
+      "loss": 1.0599,
+      "step": 2913
+    },
+    {
+      "epoch": 2.2054872280037845,
+      "grad_norm": 2.0222678184509277,
+      "learning_rate": 3.343439844197493e-06,
+      "loss": 1.083,
+      "step": 2914
+    },
+    {
+      "epoch": 2.206244087038789,
+      "grad_norm": 2.076840400695801,
+      "learning_rate": 3.337452367256012e-06,
+      "loss": 1.0306,
+      "step": 2915
+    },
+    {
+      "epoch": 2.207000946073794,
+      "grad_norm": 1.9618786573410034,
+      "learning_rate": 3.3314691855357197e-06,
+      "loss": 1.1021,
+      "step": 2916
+    },
+    {
+      "epoch": 2.2077578051087983,
+      "grad_norm": 2.168519973754883,
+      "learning_rate": 3.3254903028932716e-06,
+      "loss": 1.007,
+      "step": 2917
+    },
+    {
+      "epoch": 2.208514664143803,
+      "grad_norm": 2.0127992630004883,
+      "learning_rate": 3.3195157231825704e-06,
+      "loss": 1.0797,
+      "step": 2918
+    },
+    {
+      "epoch": 2.209271523178808,
+      "grad_norm": 2.0020880699157715,
+      "learning_rate": 3.3135454502547397e-06,
+      "loss": 1.1154,
+      "step": 2919
+    },
+    {
+      "epoch": 2.2100283822138125,
+      "grad_norm": 1.9836198091506958,
+      "learning_rate": 3.307579487958125e-06,
+      "loss": 1.0418,
+      "step": 2920
+    },
+    {
+      "epoch": 2.2107852412488174,
+      "grad_norm": 1.9691238403320312,
+      "learning_rate": 3.3016178401382957e-06,
+      "loss": 1.1094,
+      "step": 2921
+    },
+    {
+      "epoch": 2.2115421002838223,
+      "grad_norm": 2.1438305377960205,
+      "learning_rate": 3.2956605106380464e-06,
+      "loss": 1.0935,
+      "step": 2922
+    },
+    {
+      "epoch": 2.2122989593188267,
+      "grad_norm": 2.1357624530792236,
+      "learning_rate": 3.2897075032973656e-06,
+      "loss": 1.1033,
+      "step": 2923
+    },
+    {
+      "epoch": 2.2130558183538316,
+      "grad_norm": 2.027420997619629,
+      "learning_rate": 3.28375882195348e-06,
+      "loss": 1.0502,
+      "step": 2924
+    },
+    {
+      "epoch": 2.2138126773888365,
+      "grad_norm": 2.073096513748169,
+      "learning_rate": 3.2778144704408167e-06,
+      "loss": 1.0565,
+      "step": 2925
+    },
+    {
+      "epoch": 2.214569536423841,
+      "grad_norm": 2.12164306640625,
+      "learning_rate": 3.271874452591015e-06,
+      "loss": 1.103,
+      "step": 2926
+    },
+    {
+      "epoch": 2.215326395458846,
+      "grad_norm": 2.0933268070220947,
+      "learning_rate": 3.2659387722329226e-06,
+      "loss": 1.0776,
+      "step": 2927
+    },
+    {
+      "epoch": 2.2160832544938507,
+      "grad_norm": 2.033733367919922,
+      "learning_rate": 3.2600074331925834e-06,
+      "loss": 1.0642,
+      "step": 2928
+    },
+    {
+      "epoch": 2.216840113528855,
+      "grad_norm": 1.951857328414917,
+      "learning_rate": 3.2540804392932527e-06,
+      "loss": 1.0956,
+      "step": 2929
+    },
+    {
+      "epoch": 2.21759697256386,
+      "grad_norm": 2.0612125396728516,
+      "learning_rate": 3.2481577943553766e-06,
+      "loss": 1.0756,
+      "step": 2930
+    },
+    {
+      "epoch": 2.218353831598865,
+      "grad_norm": 1.9757081270217896,
+      "learning_rate": 3.2422395021966006e-06,
+      "loss": 1.0937,
+      "step": 2931
+    },
+    {
+      "epoch": 2.2191106906338693,
+      "grad_norm": 1.9480013847351074,
+      "learning_rate": 3.2363255666317706e-06,
+      "loss": 1.0986,
+      "step": 2932
+    },
+    {
+      "epoch": 2.2198675496688742,
+      "grad_norm": 2.087038040161133,
+      "learning_rate": 3.2304159914729194e-06,
+      "loss": 1.0518,
+      "step": 2933
+    },
+    {
+      "epoch": 2.2206244087038787,
+      "grad_norm": 2.119804859161377,
+      "learning_rate": 3.2245107805292625e-06,
+      "loss": 1.0356,
+      "step": 2934
+    },
+    {
+      "epoch": 2.2213812677388836,
+      "grad_norm": 2.2135863304138184,
+      "learning_rate": 3.2186099376072133e-06,
+      "loss": 1.0988,
+      "step": 2935
+    },
+    {
+      "epoch": 2.2221381267738884,
+      "grad_norm": 1.9379045963287354,
+      "learning_rate": 3.2127134665103684e-06,
+      "loss": 1.0762,
+      "step": 2936
+    },
+    {
+      "epoch": 2.222894985808893,
+      "grad_norm": 2.078213691711426,
+      "learning_rate": 3.206821371039495e-06,
+      "loss": 1.1031,
+      "step": 2937
+    },
+    {
+      "epoch": 2.2236518448438978,
+      "grad_norm": 1.7914735078811646,
+      "learning_rate": 3.2009336549925558e-06,
+      "loss": 1.0699,
+      "step": 2938
+    },
+    {
+      "epoch": 2.2244087038789027,
+      "grad_norm": 2.0808887481689453,
+      "learning_rate": 3.195050322164676e-06,
+      "loss": 1.0362,
+      "step": 2939
+    },
+    {
+      "epoch": 2.225165562913907,
+      "grad_norm": 2.2079849243164062,
+      "learning_rate": 3.1891713763481664e-06,
+      "loss": 1.0544,
+      "step": 2940
+    },
+    {
+      "epoch": 2.225922421948912,
+      "grad_norm": 2.137425184249878,
+      "learning_rate": 3.1832968213325056e-06,
+      "loss": 1.0465,
+      "step": 2941
+    },
+    {
+      "epoch": 2.226679280983917,
+      "grad_norm": 2.0021207332611084,
+      "learning_rate": 3.177426660904339e-06,
+      "loss": 1.0756,
+      "step": 2942
+    },
+    {
+      "epoch": 2.2274361400189213,
+      "grad_norm": 2.105543851852417,
+      "learning_rate": 3.1715608988474904e-06,
+      "loss": 1.0771,
+      "step": 2943
+    },
+    {
+      "epoch": 2.228192999053926,
+      "grad_norm": 1.925067663192749,
+      "learning_rate": 3.1656995389429347e-06,
+      "loss": 1.0919,
+      "step": 2944
+    },
+    {
+      "epoch": 2.228949858088931,
+      "grad_norm": 2.076474905014038,
+      "learning_rate": 3.159842584968813e-06,
+      "loss": 1.0684,
+      "step": 2945
+    },
+    {
+      "epoch": 2.2297067171239355,
+      "grad_norm": 2.0177693367004395,
+      "learning_rate": 3.15399004070043e-06,
+      "loss": 1.1303,
+      "step": 2946
+    },
+    {
+      "epoch": 2.2304635761589404,
+      "grad_norm": 1.9716448783874512,
+      "learning_rate": 3.1481419099102477e-06,
+      "loss": 1.0174,
+      "step": 2947
+    },
+    {
+      "epoch": 2.2312204351939453,
+      "grad_norm": 2.0462799072265625,
+      "learning_rate": 3.1422981963678823e-06,
+      "loss": 1.0466,
+      "step": 2948
+    },
+    {
+      "epoch": 2.2319772942289497,
+      "grad_norm": 2.242594003677368,
+      "learning_rate": 3.1364589038401055e-06,
+      "loss": 1.0786,
+      "step": 2949
+    },
+    {
+      "epoch": 2.2327341532639546,
+      "grad_norm": 2.011847972869873,
+      "learning_rate": 3.1306240360908325e-06,
+      "loss": 1.0735,
+      "step": 2950
+    },
+    {
+      "epoch": 2.2334910122989595,
+      "grad_norm": 1.8826643228530884,
+      "learning_rate": 3.124793596881128e-06,
+      "loss": 1.105,
+      "step": 2951
+    },
+    {
+      "epoch": 2.234247871333964,
+      "grad_norm": 2.173900604248047,
+      "learning_rate": 3.118967589969205e-06,
+      "loss": 1.0665,
+      "step": 2952
+    },
+    {
+      "epoch": 2.235004730368969,
+      "grad_norm": 2.065894365310669,
+      "learning_rate": 3.1131460191104214e-06,
+      "loss": 1.1092,
+      "step": 2953
+    },
+    {
+      "epoch": 2.2357615894039737,
+      "grad_norm": 2.1508965492248535,
+      "learning_rate": 3.107328888057271e-06,
+      "loss": 1.0692,
+      "step": 2954
+    },
+    {
+      "epoch": 2.236518448438978,
+      "grad_norm": 2.2163479328155518,
+      "learning_rate": 3.1015162005593918e-06,
+      "loss": 1.1466,
+      "step": 2955
+    },
+    {
+      "epoch": 2.237275307473983,
+      "grad_norm": 2.0771398544311523,
+      "learning_rate": 3.095707960363548e-06,
+      "loss": 1.1392,
+      "step": 2956
+    },
+    {
+      "epoch": 2.2380321665089875,
+      "grad_norm": 2.0793957710266113,
+      "learning_rate": 3.0899041712136474e-06,
+      "loss": 1.056,
+      "step": 2957
+    },
+    {
+      "epoch": 2.2387890255439924,
+      "grad_norm": 2.296447277069092,
+      "learning_rate": 3.084104836850719e-06,
+      "loss": 1.0637,
+      "step": 2958
+    },
+    {
+      "epoch": 2.2395458845789973,
+      "grad_norm": 2.2935092449188232,
+      "learning_rate": 3.0783099610129273e-06,
+      "loss": 1.0821,
+      "step": 2959
+    },
+    {
+      "epoch": 2.2403027436140017,
+      "grad_norm": 2.17787766456604,
+      "learning_rate": 3.0725195474355648e-06,
+      "loss": 1.0933,
+      "step": 2960
+    },
+    {
+      "epoch": 2.2410596026490066,
+      "grad_norm": 2.0003387928009033,
+      "learning_rate": 3.066733599851038e-06,
+      "loss": 1.0357,
+      "step": 2961
+    },
+    {
+      "epoch": 2.2418164616840115,
+      "grad_norm": 2.4547882080078125,
+      "learning_rate": 3.060952121988881e-06,
+      "loss": 1.1288,
+      "step": 2962
+    },
+    {
+      "epoch": 2.242573320719016,
+      "grad_norm": 2.319946765899658,
+      "learning_rate": 3.055175117575754e-06,
+      "loss": 1.09,
+      "step": 2963
+    },
+    {
+      "epoch": 2.243330179754021,
+      "grad_norm": 2.2627015113830566,
+      "learning_rate": 3.049402590335415e-06,
+      "loss": 1.1238,
+      "step": 2964
+    },
+    {
+      "epoch": 2.2440870387890257,
+      "grad_norm": 2.032540798187256,
+      "learning_rate": 3.043634543988752e-06,
+      "loss": 1.0622,
+      "step": 2965
+    },
+    {
+      "epoch": 2.24484389782403,
+      "grad_norm": 2.041095495223999,
+      "learning_rate": 3.037870982253763e-06,
+      "loss": 1.0729,
+      "step": 2966
+    },
+    {
+      "epoch": 2.245600756859035,
+      "grad_norm": 2.079834222793579,
+      "learning_rate": 3.032111908845547e-06,
+      "loss": 1.0541,
+      "step": 2967
+    },
+    {
+      "epoch": 2.24635761589404,
+      "grad_norm": 1.9280726909637451,
+      "learning_rate": 3.0263573274763165e-06,
+      "loss": 1.0715,
+      "step": 2968
+    },
+    {
+      "epoch": 2.2471144749290444,
+      "grad_norm": 1.9036996364593506,
+      "learning_rate": 3.0206072418553854e-06,
+      "loss": 1.0914,
+      "step": 2969
+    },
+    {
+      "epoch": 2.2478713339640493,
+      "grad_norm": 1.9973169565200806,
+      "learning_rate": 3.0148616556891774e-06,
+      "loss": 1.0612,
+      "step": 2970
+    },
+    {
+      "epoch": 2.248628192999054,
+      "grad_norm": 2.048168420791626,
+      "learning_rate": 3.009120572681206e-06,
+      "loss": 1.0573,
+      "step": 2971
+    },
+    {
+      "epoch": 2.2493850520340586,
+      "grad_norm": 1.9385312795639038,
+      "learning_rate": 3.0033839965320797e-06,
+      "loss": 1.0503,
+      "step": 2972
+    },
+    {
+      "epoch": 2.2501419110690635,
+      "grad_norm": 1.8759933710098267,
+      "learning_rate": 2.9976519309395154e-06,
+      "loss": 1.0739,
+      "step": 2973
+    },
+    {
+      "epoch": 2.250898770104068,
+      "grad_norm": 2.2850966453552246,
+      "learning_rate": 2.9919243795983116e-06,
+      "loss": 1.0669,
+      "step": 2974
+    },
+    {
+      "epoch": 2.251655629139073,
+      "grad_norm": 2.017787218093872,
+      "learning_rate": 2.9862013462003634e-06,
+      "loss": 1.0993,
+      "step": 2975
+    },
+    {
+      "epoch": 2.2524124881740777,
+      "grad_norm": 1.9540081024169922,
+      "learning_rate": 2.980482834434648e-06,
+      "loss": 1.116,
+      "step": 2976
+    },
+    {
+      "epoch": 2.253169347209082,
+      "grad_norm": 2.056605577468872,
+      "learning_rate": 2.974768847987239e-06,
+      "loss": 1.0612,
+      "step": 2977
+    },
+    {
+      "epoch": 2.253926206244087,
+      "grad_norm": 2.3890209197998047,
+      "learning_rate": 2.969059390541273e-06,
+      "loss": 1.0817,
+      "step": 2978
+    },
+    {
+      "epoch": 2.254683065279092,
+      "grad_norm": 1.947478175163269,
+      "learning_rate": 2.963354465776983e-06,
+      "loss": 1.0236,
+      "step": 2979
+    },
+    {
+      "epoch": 2.2554399243140963,
+      "grad_norm": 2.1459438800811768,
+      "learning_rate": 2.9576540773716783e-06,
+      "loss": 1.0725,
+      "step": 2980
+    },
+    {
+      "epoch": 2.2561967833491012,
+      "grad_norm": 2.0048837661743164,
+      "learning_rate": 2.9519582289997423e-06,
+      "loss": 1.1,
+      "step": 2981
+    },
+    {
+      "epoch": 2.256953642384106,
+      "grad_norm": 2.1805686950683594,
+      "learning_rate": 2.9462669243326357e-06,
+      "loss": 1.0963,
+      "step": 2982
+    },
+    {
+      "epoch": 2.2577105014191106,
+      "grad_norm": 1.9871695041656494,
+      "learning_rate": 2.9405801670388784e-06,
+      "loss": 1.1184,
+      "step": 2983
+    },
+    {
+      "epoch": 2.2584673604541154,
+      "grad_norm": 2.142199754714966,
+      "learning_rate": 2.934897960784075e-06,
+      "loss": 1.0546,
+      "step": 2984
+    },
+    {
+      "epoch": 2.2592242194891203,
+      "grad_norm": 2.1177968978881836,
+      "learning_rate": 2.9292203092308823e-06,
+      "loss": 1.1136,
+      "step": 2985
+    },
+    {
+      "epoch": 2.2599810785241248,
+      "grad_norm": 2.1006743907928467,
+      "learning_rate": 2.9235472160390315e-06,
+      "loss": 1.0091,
+      "step": 2986
+    },
+    {
+      "epoch": 2.2607379375591297,
+      "grad_norm": 1.978402853012085,
+      "learning_rate": 2.917878684865312e-06,
+      "loss": 1.0677,
+      "step": 2987
+    },
+    {
+      "epoch": 2.2614947965941345,
+      "grad_norm": 2.2147839069366455,
+      "learning_rate": 2.9122147193635757e-06,
+      "loss": 1.0809,
+      "step": 2988
+    },
+    {
+      "epoch": 2.262251655629139,
+      "grad_norm": 2.2743515968322754,
+      "learning_rate": 2.9065553231847215e-06,
+      "loss": 1.0647,
+      "step": 2989
+    },
+    {
+      "epoch": 2.263008514664144,
+      "grad_norm": 2.038224935531616,
+      "learning_rate": 2.900900499976714e-06,
+      "loss": 1.0953,
+      "step": 2990
+    },
+    {
+      "epoch": 2.2637653736991483,
+      "grad_norm": 1.9364351034164429,
+      "learning_rate": 2.895250253384567e-06,
+      "loss": 1.0241,
+      "step": 2991
+    },
+    {
+      "epoch": 2.264522232734153,
+      "grad_norm": 2.0556015968322754,
+      "learning_rate": 2.8896045870503405e-06,
+      "loss": 1.0358,
+      "step": 2992
+    },
+    {
+      "epoch": 2.265279091769158,
+      "grad_norm": 2.6211061477661133,
+      "learning_rate": 2.8839635046131477e-06,
+      "loss": 1.058,
+      "step": 2993
+    },
+    {
+      "epoch": 2.2660359508041625,
+      "grad_norm": 2.0403685569763184,
+      "learning_rate": 2.87832700970914e-06,
+      "loss": 1.0957,
+      "step": 2994
+    },
+    {
+      "epoch": 2.2667928098391674,
+      "grad_norm": 2.106076955795288,
+      "learning_rate": 2.8726951059715184e-06,
+      "loss": 1.0622,
+      "step": 2995
+    },
+    {
+      "epoch": 2.2675496688741723,
+      "grad_norm": 1.878516674041748,
+      "learning_rate": 2.867067797030522e-06,
+      "loss": 1.0636,
+      "step": 2996
+    },
+    {
+      "epoch": 2.2683065279091768,
+      "grad_norm": 2.178928852081299,
+      "learning_rate": 2.861445086513431e-06,
+      "loss": 1.0347,
+      "step": 2997
+    },
+    {
+      "epoch": 2.2690633869441816,
+      "grad_norm": 2.5624477863311768,
+      "learning_rate": 2.855826978044558e-06,
+      "loss": 1.0171,
+      "step": 2998
+    },
+    {
+      "epoch": 2.2698202459791865,
+      "grad_norm": 1.9493463039398193,
+      "learning_rate": 2.8502134752452488e-06,
+      "loss": 1.0763,
+      "step": 2999
+    },
+    {
+      "epoch": 2.270577105014191,
+      "grad_norm": 1.9162508249282837,
+      "learning_rate": 2.844604581733879e-06,
+      "loss": 1.1071,
+      "step": 3000
+    },
+    {
+      "epoch": 2.271333964049196,
+      "grad_norm": 2.097134828567505,
+      "learning_rate": 2.8390003011258576e-06,
+      "loss": 1.0987,
+      "step": 3001
+    },
+    {
+      "epoch": 2.2720908230842007,
+      "grad_norm": 2.138456106185913,
+      "learning_rate": 2.83340063703362e-06,
+      "loss": 1.0962,
+      "step": 3002
+    },
+    {
+      "epoch": 2.272847682119205,
+      "grad_norm": 2.057185411453247,
+      "learning_rate": 2.8278055930666243e-06,
+      "loss": 1.0849,
+      "step": 3003
+    },
+    {
+      "epoch": 2.27360454115421,
+      "grad_norm": 2.094721555709839,
+      "learning_rate": 2.822215172831354e-06,
+      "loss": 1.1004,
+      "step": 3004
+    },
+    {
+      "epoch": 2.274361400189215,
+      "grad_norm": 2.1358296871185303,
+      "learning_rate": 2.8166293799312994e-06,
+      "loss": 1.0583,
+      "step": 3005
+    },
+    {
+      "epoch": 2.2751182592242194,
+      "grad_norm": 1.955474615097046,
+      "learning_rate": 2.8110482179669823e-06,
+      "loss": 1.0828,
+      "step": 3006
+    },
+    {
+      "epoch": 2.2758751182592243,
+      "grad_norm": 2.084411144256592,
+      "learning_rate": 2.805471690535935e-06,
+      "loss": 1.0635,
+      "step": 3007
+    },
+    {
+      "epoch": 2.2766319772942287,
+      "grad_norm": 2.111748218536377,
+      "learning_rate": 2.799899801232702e-06,
+      "loss": 1.0604,
+      "step": 3008
+    },
+    {
+      "epoch": 2.2773888363292336,
+      "grad_norm": 2.2352616786956787,
+      "learning_rate": 2.7943325536488373e-06,
+      "loss": 1.1397,
+      "step": 3009
+    },
+    {
+      "epoch": 2.2781456953642385,
+      "grad_norm": 2.1407878398895264,
+      "learning_rate": 2.788769951372908e-06,
+      "loss": 1.083,
+      "step": 3010
+    },
+    {
+      "epoch": 2.2789025543992434,
+      "grad_norm": 2.0809216499328613,
+      "learning_rate": 2.7832119979904798e-06,
+      "loss": 1.0496,
+      "step": 3011
+    },
+    {
+      "epoch": 2.279659413434248,
+      "grad_norm": 2.0093045234680176,
+      "learning_rate": 2.77765869708412e-06,
+      "loss": 1.0543,
+      "step": 3012
+    },
+    {
+      "epoch": 2.2804162724692527,
+      "grad_norm": 1.9954379796981812,
+      "learning_rate": 2.7721100522334056e-06,
+      "loss": 1.1095,
+      "step": 3013
+    },
+    {
+      "epoch": 2.281173131504257,
+      "grad_norm": 2.1180033683776855,
+      "learning_rate": 2.7665660670149092e-06,
+      "loss": 1.0982,
+      "step": 3014
+    },
+    {
+      "epoch": 2.281929990539262,
+      "grad_norm": 2.0234453678131104,
+      "learning_rate": 2.761026745002201e-06,
+      "loss": 1.0808,
+      "step": 3015
+    },
+    {
+      "epoch": 2.282686849574267,
+      "grad_norm": 2.1585819721221924,
+      "learning_rate": 2.7554920897658386e-06,
+      "loss": 1.0775,
+      "step": 3016
+    },
+    {
+      "epoch": 2.2834437086092714,
+      "grad_norm": 1.9864180088043213,
+      "learning_rate": 2.7499621048733775e-06,
+      "loss": 1.1177,
+      "step": 3017
+    },
+    {
+      "epoch": 2.2842005676442763,
+      "grad_norm": 2.359938621520996,
+      "learning_rate": 2.744436793889368e-06,
+      "loss": 1.0951,
+      "step": 3018
+    },
+    {
+      "epoch": 2.284957426679281,
+      "grad_norm": 2.0253729820251465,
+      "learning_rate": 2.7389161603753312e-06,
+      "loss": 1.0414,
+      "step": 3019
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 1.9887871742248535,
+      "learning_rate": 2.73340020788979e-06,
+      "loss": 1.0882,
+      "step": 3020
+    },
+    {
+      "epoch": 2.2864711447492905,
+      "grad_norm": 1.971255898475647,
+      "learning_rate": 2.7278889399882435e-06,
+      "loss": 1.0867,
+      "step": 3021
+    },
+    {
+      "epoch": 2.2872280037842954,
+      "grad_norm": 2.0510776042938232,
+      "learning_rate": 2.7223823602231664e-06,
+      "loss": 1.0738,
+      "step": 3022
+    },
+    {
+      "epoch": 2.2879848628193,
+      "grad_norm": 1.953727126121521,
+      "learning_rate": 2.7168804721440177e-06,
+      "loss": 1.0262,
+      "step": 3023
+    },
+    {
+      "epoch": 2.2887417218543047,
+      "grad_norm": 2.0659303665161133,
+      "learning_rate": 2.7113832792972323e-06,
+      "loss": 1.0748,
+      "step": 3024
+    },
+    {
+      "epoch": 2.289498580889309,
+      "grad_norm": 2.147465229034424,
+      "learning_rate": 2.705890785226219e-06,
+      "loss": 1.0691,
+      "step": 3025
+    },
+    {
+      "epoch": 2.290255439924314,
+      "grad_norm": 2.1003715991973877,
+      "learning_rate": 2.7004029934713516e-06,
+      "loss": 1.0623,
+      "step": 3026
+    },
+    {
+      "epoch": 2.291012298959319,
+      "grad_norm": 1.9521448612213135,
+      "learning_rate": 2.6949199075699754e-06,
+      "loss": 1.1025,
+      "step": 3027
+    },
+    {
+      "epoch": 2.291769157994324,
+      "grad_norm": 1.998204231262207,
+      "learning_rate": 2.689441531056408e-06,
+      "loss": 1.0875,
+      "step": 3028
+    },
+    {
+      "epoch": 2.2925260170293282,
+      "grad_norm": 1.98150634765625,
+      "learning_rate": 2.683967867461925e-06,
+      "loss": 1.1051,
+      "step": 3029
+    },
+    {
+      "epoch": 2.293282876064333,
+      "grad_norm": 2.0418317317962646,
+      "learning_rate": 2.678498920314767e-06,
+      "loss": 1.0871,
+      "step": 3030
+    },
+    {
+      "epoch": 2.2940397350993376,
+      "grad_norm": 2.075697898864746,
+      "learning_rate": 2.673034693140136e-06,
+      "loss": 1.1588,
+      "step": 3031
+    },
+    {
+      "epoch": 2.2947965941343424,
+      "grad_norm": 2.049619197845459,
+      "learning_rate": 2.6675751894601928e-06,
+      "loss": 1.1,
+      "step": 3032
+    },
+    {
+      "epoch": 2.2955534531693473,
+      "grad_norm": 1.9731786251068115,
+      "learning_rate": 2.6621204127940403e-06,
+      "loss": 1.0883,
+      "step": 3033
+    },
+    {
+      "epoch": 2.2963103122043518,
+      "grad_norm": 2.0121383666992188,
+      "learning_rate": 2.656670366657748e-06,
+      "loss": 1.0914,
+      "step": 3034
+    },
+    {
+      "epoch": 2.2970671712393567,
+      "grad_norm": 2.0904853343963623,
+      "learning_rate": 2.651225054564334e-06,
+      "loss": 1.0803,
+      "step": 3035
+    },
+    {
+      "epoch": 2.2978240302743616,
+      "grad_norm": 2.0923306941986084,
+      "learning_rate": 2.645784480023764e-06,
+      "loss": 1.0528,
+      "step": 3036
+    },
+    {
+      "epoch": 2.298580889309366,
+      "grad_norm": 2.1617391109466553,
+      "learning_rate": 2.6403486465429524e-06,
+      "loss": 1.1454,
+      "step": 3037
+    },
+    {
+      "epoch": 2.299337748344371,
+      "grad_norm": 1.9403904676437378,
+      "learning_rate": 2.634917557625747e-06,
+      "loss": 1.1051,
+      "step": 3038
+    },
+    {
+      "epoch": 2.3000946073793758,
+      "grad_norm": 2.3704395294189453,
+      "learning_rate": 2.629491216772951e-06,
+      "loss": 1.0736,
+      "step": 3039
+    },
+    {
+      "epoch": 2.30085146641438,
+      "grad_norm": 2.0632617473602295,
+      "learning_rate": 2.6240696274822976e-06,
+      "loss": 1.0948,
+      "step": 3040
+    },
+    {
+      "epoch": 2.301608325449385,
+      "grad_norm": 1.9306870698928833,
+      "learning_rate": 2.6186527932484595e-06,
+      "loss": 1.0978,
+      "step": 3041
+    },
+    {
+      "epoch": 2.3023651844843895,
+      "grad_norm": 2.1989099979400635,
+      "learning_rate": 2.61324071756305e-06,
+      "loss": 1.0513,
+      "step": 3042
+    },
+    {
+      "epoch": 2.3031220435193944,
+      "grad_norm": 2.053589105606079,
+      "learning_rate": 2.60783340391461e-06,
+      "loss": 1.0828,
+      "step": 3043
+    },
+    {
+      "epoch": 2.3038789025543993,
+      "grad_norm": 2.009385347366333,
+      "learning_rate": 2.602430855788607e-06,
+      "loss": 1.0859,
+      "step": 3044
+    },
+    {
+      "epoch": 2.304635761589404,
+      "grad_norm": 2.086993932723999,
+      "learning_rate": 2.597033076667443e-06,
+      "loss": 1.1311,
+      "step": 3045
+    },
+    {
+      "epoch": 2.3053926206244086,
+      "grad_norm": 2.1662371158599854,
+      "learning_rate": 2.5916400700304476e-06,
+      "loss": 1.0746,
+      "step": 3046
+    },
+    {
+      "epoch": 2.3061494796594135,
+      "grad_norm": 1.9955482482910156,
+      "learning_rate": 2.5862518393538662e-06,
+      "loss": 1.0841,
+      "step": 3047
+    },
+    {
+      "epoch": 2.306906338694418,
+      "grad_norm": 2.2083284854888916,
+      "learning_rate": 2.5808683881108743e-06,
+      "loss": 1.0738,
+      "step": 3048
+    },
+    {
+      "epoch": 2.307663197729423,
+      "grad_norm": 2.1207940578460693,
+      "learning_rate": 2.5754897197715566e-06,
+      "loss": 1.1198,
+      "step": 3049
+    },
+    {
+      "epoch": 2.3084200567644277,
+      "grad_norm": 2.125546455383301,
+      "learning_rate": 2.5701158378029245e-06,
+      "loss": 1.0487,
+      "step": 3050
+    },
+    {
+      "epoch": 2.309176915799432,
+      "grad_norm": 2.025674343109131,
+      "learning_rate": 2.564746745668899e-06,
+      "loss": 1.0782,
+      "step": 3051
+    },
+    {
+      "epoch": 2.309933774834437,
+      "grad_norm": 2.1108992099761963,
+      "learning_rate": 2.559382446830318e-06,
+      "loss": 1.1041,
+      "step": 3052
+    },
+    {
+      "epoch": 2.310690633869442,
+      "grad_norm": 2.149214267730713,
+      "learning_rate": 2.554022944744925e-06,
+      "loss": 1.0332,
+      "step": 3053
+    },
+    {
+      "epoch": 2.3114474929044464,
+      "grad_norm": 2.708857536315918,
+      "learning_rate": 2.5486682428673753e-06,
+      "loss": 1.0605,
+      "step": 3054
+    },
+    {
+      "epoch": 2.3122043519394513,
+      "grad_norm": 2.029184341430664,
+      "learning_rate": 2.5433183446492214e-06,
+      "loss": 1.1014,
+      "step": 3055
+    },
+    {
+      "epoch": 2.312961210974456,
+      "grad_norm": 1.9624137878417969,
+      "learning_rate": 2.537973253538931e-06,
+      "loss": 1.0924,
+      "step": 3056
+    },
+    {
+      "epoch": 2.3137180700094606,
+      "grad_norm": 2.015695571899414,
+      "learning_rate": 2.5326329729818673e-06,
+      "loss": 1.0847,
+      "step": 3057
+    },
+    {
+      "epoch": 2.3144749290444655,
+      "grad_norm": 2.0661072731018066,
+      "learning_rate": 2.5272975064202943e-06,
+      "loss": 1.0434,
+      "step": 3058
+    },
+    {
+      "epoch": 2.3152317880794704,
+      "grad_norm": 2.1128456592559814,
+      "learning_rate": 2.521966857293378e-06,
+      "loss": 1.0661,
+      "step": 3059
+    },
+    {
+      "epoch": 2.315988647114475,
+      "grad_norm": 2.1260507106781006,
+      "learning_rate": 2.5166410290371626e-06,
+      "loss": 1.084,
+      "step": 3060
+    },
+    {
+      "epoch": 2.3167455061494797,
+      "grad_norm": 2.091785430908203,
+      "learning_rate": 2.5113200250845996e-06,
+      "loss": 1.0772,
+      "step": 3061
+    },
+    {
+      "epoch": 2.3175023651844846,
+      "grad_norm": 2.1781039237976074,
+      "learning_rate": 2.5060038488655302e-06,
+      "loss": 1.0469,
+      "step": 3062
+    },
+    {
+      "epoch": 2.318259224219489,
+      "grad_norm": 2.150576114654541,
+      "learning_rate": 2.500692503806678e-06,
+      "loss": 1.1243,
+      "step": 3063
+    },
+    {
+      "epoch": 2.319016083254494,
+      "grad_norm": 2.021026372909546,
+      "learning_rate": 2.4953859933316555e-06,
+      "loss": 1.0894,
+      "step": 3064
+    },
+    {
+      "epoch": 2.3197729422894984,
+      "grad_norm": 2.0633111000061035,
+      "learning_rate": 2.490084320860961e-06,
+      "loss": 1.07,
+      "step": 3065
+    },
+    {
+      "epoch": 2.3205298013245033,
+      "grad_norm": 2.0326859951019287,
+      "learning_rate": 2.4847874898119706e-06,
+      "loss": 1.1148,
+      "step": 3066
+    },
+    {
+      "epoch": 2.321286660359508,
+      "grad_norm": 2.0705957412719727,
+      "learning_rate": 2.479495503598935e-06,
+      "loss": 1.056,
+      "step": 3067
+    },
+    {
+      "epoch": 2.3220435193945126,
+      "grad_norm": 2.0693790912628174,
+      "learning_rate": 2.474208365632993e-06,
+      "loss": 1.0678,
+      "step": 3068
+    },
+    {
+      "epoch": 2.3228003784295175,
+      "grad_norm": 2.152256965637207,
+      "learning_rate": 2.468926079322153e-06,
+      "loss": 1.0826,
+      "step": 3069
+    },
+    {
+      "epoch": 2.3235572374645224,
+      "grad_norm": 1.9246243238449097,
+      "learning_rate": 2.463648648071298e-06,
+      "loss": 1.0872,
+      "step": 3070
+    },
+    {
+      "epoch": 2.324314096499527,
+      "grad_norm": 2.0952563285827637,
+      "learning_rate": 2.4583760752821752e-06,
+      "loss": 1.0851,
+      "step": 3071
+    },
+    {
+      "epoch": 2.3250709555345317,
+      "grad_norm": 2.1515979766845703,
+      "learning_rate": 2.453108364353406e-06,
+      "loss": 1.0494,
+      "step": 3072
+    },
+    {
+      "epoch": 2.3258278145695366,
+      "grad_norm": 2.0528197288513184,
+      "learning_rate": 2.447845518680481e-06,
+      "loss": 1.0619,
+      "step": 3073
+    },
+    {
+      "epoch": 2.326584673604541,
+      "grad_norm": 2.0087759494781494,
+      "learning_rate": 2.4425875416557426e-06,
+      "loss": 1.0229,
+      "step": 3074
+    },
+    {
+      "epoch": 2.327341532639546,
+      "grad_norm": 2.0219340324401855,
+      "learning_rate": 2.437334436668407e-06,
+      "loss": 1.0621,
+      "step": 3075
+    },
+    {
+      "epoch": 2.328098391674551,
+      "grad_norm": 2.0388236045837402,
+      "learning_rate": 2.432086207104549e-06,
+      "loss": 1.111,
+      "step": 3076
+    },
+    {
+      "epoch": 2.3288552507095552,
+      "grad_norm": 2.142197370529175,
+      "learning_rate": 2.426842856347089e-06,
+      "loss": 1.0878,
+      "step": 3077
+    },
+    {
+      "epoch": 2.32961210974456,
+      "grad_norm": 1.8765891790390015,
+      "learning_rate": 2.4216043877758163e-06,
+      "loss": 1.1534,
+      "step": 3078
+    },
+    {
+      "epoch": 2.330368968779565,
+      "grad_norm": 2.0405936241149902,
+      "learning_rate": 2.416370804767367e-06,
+      "loss": 1.0497,
+      "step": 3079
+    },
+    {
+      "epoch": 2.3311258278145695,
+      "grad_norm": 1.989956259727478,
+      "learning_rate": 2.4111421106952317e-06,
+      "loss": 1.0716,
+      "step": 3080
+    },
+    {
+      "epoch": 2.3318826868495743,
+      "grad_norm": 2.097310781478882,
+      "learning_rate": 2.4059183089297432e-06,
+      "loss": 1.0932,
+      "step": 3081
+    },
+    {
+      "epoch": 2.332639545884579,
+      "grad_norm": 1.9613529443740845,
+      "learning_rate": 2.4006994028380835e-06,
+      "loss": 1.0853,
+      "step": 3082
+    },
+    {
+      "epoch": 2.3333964049195837,
+      "grad_norm": 2.1626081466674805,
+      "learning_rate": 2.3954853957842816e-06,
+      "loss": 1.1113,
+      "step": 3083
+    },
+    {
+      "epoch": 2.3341532639545886,
+      "grad_norm": 1.9847509860992432,
+      "learning_rate": 2.3902762911292063e-06,
+      "loss": 1.0632,
+      "step": 3084
+    },
+    {
+      "epoch": 2.334910122989593,
+      "grad_norm": 2.432072639465332,
+      "learning_rate": 2.385072092230568e-06,
+      "loss": 1.0981,
+      "step": 3085
+    },
+    {
+      "epoch": 2.335666982024598,
+      "grad_norm": 2.0696966648101807,
+      "learning_rate": 2.3798728024429136e-06,
+      "loss": 1.053,
+      "step": 3086
+    },
+    {
+      "epoch": 2.3364238410596028,
+      "grad_norm": 2.017995595932007,
+      "learning_rate": 2.374678425117631e-06,
+      "loss": 1.1,
+      "step": 3087
+    },
+    {
+      "epoch": 2.337180700094607,
+      "grad_norm": 2.056910753250122,
+      "learning_rate": 2.369488963602927e-06,
+      "loss": 1.1009,
+      "step": 3088
+    },
+    {
+      "epoch": 2.337937559129612,
+      "grad_norm": 2.1294288635253906,
+      "learning_rate": 2.3643044212438547e-06,
+      "loss": 1.1008,
+      "step": 3089
+    },
+    {
+      "epoch": 2.338694418164617,
+      "grad_norm": 1.9634032249450684,
+      "learning_rate": 2.3591248013822885e-06,
+      "loss": 1.0868,
+      "step": 3090
+    },
+    {
+      "epoch": 2.3394512771996214,
+      "grad_norm": 2.2118031978607178,
+      "learning_rate": 2.3539501073569357e-06,
+      "loss": 1.1081,
+      "step": 3091
+    },
+    {
+      "epoch": 2.3402081362346263,
+      "grad_norm": 2.3364391326904297,
+      "learning_rate": 2.348780342503326e-06,
+      "loss": 1.0373,
+      "step": 3092
+    },
+    {
+      "epoch": 2.340964995269631,
+      "grad_norm": 2.061373472213745,
+      "learning_rate": 2.343615510153806e-06,
+      "loss": 1.0809,
+      "step": 3093
+    },
+    {
+      "epoch": 2.3417218543046356,
+      "grad_norm": 1.9650219678878784,
+      "learning_rate": 2.338455613637553e-06,
+      "loss": 1.0659,
+      "step": 3094
+    },
+    {
+      "epoch": 2.3424787133396405,
+      "grad_norm": 2.1445631980895996,
+      "learning_rate": 2.333300656280552e-06,
+      "loss": 1.0708,
+      "step": 3095
+    },
+    {
+      "epoch": 2.3432355723746454,
+      "grad_norm": 2.209373950958252,
+      "learning_rate": 2.328150641405614e-06,
+      "loss": 1.0744,
+      "step": 3096
+    },
+    {
+      "epoch": 2.34399243140965,
+      "grad_norm": 2.036855936050415,
+      "learning_rate": 2.3230055723323587e-06,
+      "loss": 1.0878,
+      "step": 3097
+    },
+    {
+      "epoch": 2.3447492904446547,
+      "grad_norm": 2.2812464237213135,
+      "learning_rate": 2.317865452377222e-06,
+      "loss": 1.0321,
+      "step": 3098
+    },
+    {
+      "epoch": 2.345506149479659,
+      "grad_norm": 1.9373234510421753,
+      "learning_rate": 2.312730284853442e-06,
+      "loss": 1.1768,
+      "step": 3099
+    },
+    {
+      "epoch": 2.346263008514664,
+      "grad_norm": 1.9641289710998535,
+      "learning_rate": 2.3076000730710715e-06,
+      "loss": 1.0922,
+      "step": 3100
+    },
+    {
+      "epoch": 2.347019867549669,
+      "grad_norm": 2.2891197204589844,
+      "learning_rate": 2.3024748203369697e-06,
+      "loss": 1.0122,
+      "step": 3101
+    },
+    {
+      "epoch": 2.347776726584674,
+      "grad_norm": 2.042477607727051,
+      "learning_rate": 2.2973545299547907e-06,
+      "loss": 1.0209,
+      "step": 3102
+    },
+    {
+      "epoch": 2.3485335856196783,
+      "grad_norm": 2.082688331604004,
+      "learning_rate": 2.2922392052250012e-06,
+      "loss": 1.0879,
+      "step": 3103
+    },
+    {
+      "epoch": 2.349290444654683,
+      "grad_norm": 2.036217451095581,
+      "learning_rate": 2.287128849444857e-06,
+      "loss": 1.0736,
+      "step": 3104
+    },
+    {
+      "epoch": 2.3500473036896876,
+      "grad_norm": 2.099870443344116,
+      "learning_rate": 2.282023465908417e-06,
+      "loss": 1.0502,
+      "step": 3105
+    },
+    {
+      "epoch": 2.3508041627246925,
+      "grad_norm": 2.0198404788970947,
+      "learning_rate": 2.276923057906534e-06,
+      "loss": 1.1233,
+      "step": 3106
+    },
+    {
+      "epoch": 2.3515610217596974,
+      "grad_norm": 2.331169843673706,
+      "learning_rate": 2.271827628726853e-06,
+      "loss": 1.0535,
+      "step": 3107
+    },
+    {
+      "epoch": 2.352317880794702,
+      "grad_norm": 2.124520778656006,
+      "learning_rate": 2.2667371816538124e-06,
+      "loss": 1.0393,
+      "step": 3108
+    },
+    {
+      "epoch": 2.3530747398297067,
+      "grad_norm": 1.9916775226593018,
+      "learning_rate": 2.261651719968635e-06,
+      "loss": 1.0811,
+      "step": 3109
+    },
+    {
+      "epoch": 2.3538315988647116,
+      "grad_norm": 2.134824514389038,
+      "learning_rate": 2.2565712469493285e-06,
+      "loss": 1.0755,
+      "step": 3110
+    },
+    {
+      "epoch": 2.354588457899716,
+      "grad_norm": 2.083906412124634,
+      "learning_rate": 2.251495765870691e-06,
+      "loss": 1.0626,
+      "step": 3111
+    },
+    {
+      "epoch": 2.355345316934721,
+      "grad_norm": 2.06776762008667,
+      "learning_rate": 2.246425280004301e-06,
+      "loss": 1.0667,
+      "step": 3112
+    },
+    {
+      "epoch": 2.356102175969726,
+      "grad_norm": 2.042806386947632,
+      "learning_rate": 2.241359792618514e-06,
+      "loss": 1.0589,
+      "step": 3113
+    },
+    {
+      "epoch": 2.3568590350047303,
+      "grad_norm": 2.0756242275238037,
+      "learning_rate": 2.2362993069784754e-06,
+      "loss": 1.0757,
+      "step": 3114
+    },
+    {
+      "epoch": 2.357615894039735,
+      "grad_norm": 2.315819501876831,
+      "learning_rate": 2.231243826346082e-06,
+      "loss": 1.1302,
+      "step": 3115
+    },
+    {
+      "epoch": 2.3583727530747396,
+      "grad_norm": 2.038795232772827,
+      "learning_rate": 2.2261933539800276e-06,
+      "loss": 0.9891,
+      "step": 3116
+    },
+    {
+      "epoch": 2.3591296121097445,
+      "grad_norm": 1.8599226474761963,
+      "learning_rate": 2.2211478931357686e-06,
+      "loss": 1.0786,
+      "step": 3117
+    },
+    {
+      "epoch": 2.3598864711447494,
+      "grad_norm": 2.074420928955078,
+      "learning_rate": 2.2161074470655327e-06,
+      "loss": 1.0555,
+      "step": 3118
+    },
+    {
+      "epoch": 2.3606433301797543,
+      "grad_norm": 2.0066890716552734,
+      "learning_rate": 2.2110720190183143e-06,
+      "loss": 1.077,
+      "step": 3119
+    },
+    {
+      "epoch": 2.3614001892147587,
+      "grad_norm": 2.137488603591919,
+      "learning_rate": 2.2060416122398754e-06,
+      "loss": 1.0421,
+      "step": 3120
+    },
+    {
+      "epoch": 2.3621570482497636,
+      "grad_norm": 2.0553901195526123,
+      "learning_rate": 2.2010162299727382e-06,
+      "loss": 1.0978,
+      "step": 3121
+    },
+    {
+      "epoch": 2.362913907284768,
+      "grad_norm": 1.9921813011169434,
+      "learning_rate": 2.1959958754561846e-06,
+      "loss": 1.0452,
+      "step": 3122
+    },
+    {
+      "epoch": 2.363670766319773,
+      "grad_norm": 2.0219523906707764,
+      "learning_rate": 2.1909805519262607e-06,
+      "loss": 1.0474,
+      "step": 3123
+    },
+    {
+      "epoch": 2.364427625354778,
+      "grad_norm": 1.9772000312805176,
+      "learning_rate": 2.185970262615767e-06,
+      "loss": 1.0281,
+      "step": 3124
+    },
+    {
+      "epoch": 2.3651844843897822,
+      "grad_norm": 2.09308123588562,
+      "learning_rate": 2.1809650107542632e-06,
+      "loss": 1.082,
+      "step": 3125
+    },
+    {
+      "epoch": 2.365941343424787,
+      "grad_norm": 1.9955263137817383,
+      "learning_rate": 2.175964799568052e-06,
+      "loss": 1.0324,
+      "step": 3126
+    },
+    {
+      "epoch": 2.366698202459792,
+      "grad_norm": 2.1914641857147217,
+      "learning_rate": 2.1709696322801972e-06,
+      "loss": 1.093,
+      "step": 3127
+    },
+    {
+      "epoch": 2.3674550614947965,
+      "grad_norm": 1.9175313711166382,
+      "learning_rate": 2.1659795121105097e-06,
+      "loss": 1.0671,
+      "step": 3128
+    },
+    {
+      "epoch": 2.3682119205298013,
+      "grad_norm": 2.153555154800415,
+      "learning_rate": 2.16099444227554e-06,
+      "loss": 1.0857,
+      "step": 3129
+    },
+    {
+      "epoch": 2.3689687795648062,
+      "grad_norm": 2.1872153282165527,
+      "learning_rate": 2.1560144259885886e-06,
+      "loss": 1.0465,
+      "step": 3130
+    },
+    {
+      "epoch": 2.3697256385998107,
+      "grad_norm": 2.0605931282043457,
+      "learning_rate": 2.151039466459703e-06,
+      "loss": 1.0579,
+      "step": 3131
+    },
+    {
+      "epoch": 2.3704824976348156,
+      "grad_norm": 2.1121630668640137,
+      "learning_rate": 2.1460695668956603e-06,
+      "loss": 1.0797,
+      "step": 3132
+    },
+    {
+      "epoch": 2.37123935666982,
+      "grad_norm": 1.927918553352356,
+      "learning_rate": 2.1411047304999855e-06,
+      "loss": 1.065,
+      "step": 3133
+    },
+    {
+      "epoch": 2.371996215704825,
+      "grad_norm": 2.0213940143585205,
+      "learning_rate": 2.1361449604729334e-06,
+      "loss": 1.0243,
+      "step": 3134
+    },
+    {
+      "epoch": 2.3727530747398298,
+      "grad_norm": 2.0634241104125977,
+      "learning_rate": 2.1311902600115026e-06,
+      "loss": 1.0243,
+      "step": 3135
+    },
+    {
+      "epoch": 2.3735099337748347,
+      "grad_norm": 2.063898801803589,
+      "learning_rate": 2.126240632309412e-06,
+      "loss": 1.0804,
+      "step": 3136
+    },
+    {
+      "epoch": 2.374266792809839,
+      "grad_norm": 1.9965025186538696,
+      "learning_rate": 2.1212960805571153e-06,
+      "loss": 1.0819,
+      "step": 3137
+    },
+    {
+      "epoch": 2.375023651844844,
+      "grad_norm": 2.0790200233459473,
+      "learning_rate": 2.1163566079417965e-06,
+      "loss": 1.0549,
+      "step": 3138
+    },
+    {
+      "epoch": 2.3757805108798484,
+      "grad_norm": 2.2082910537719727,
+      "learning_rate": 2.1114222176473647e-06,
+      "loss": 1.041,
+      "step": 3139
+    },
+    {
+      "epoch": 2.3765373699148533,
+      "grad_norm": 2.1791975498199463,
+      "learning_rate": 2.1064929128544527e-06,
+      "loss": 1.0671,
+      "step": 3140
+    },
+    {
+      "epoch": 2.377294228949858,
+      "grad_norm": 2.011662006378174,
+      "learning_rate": 2.1015686967404155e-06,
+      "loss": 1.0451,
+      "step": 3141
+    },
+    {
+      "epoch": 2.3780510879848626,
+      "grad_norm": 2.092410087585449,
+      "learning_rate": 2.0966495724793328e-06,
+      "loss": 0.999,
+      "step": 3142
+    },
+    {
+      "epoch": 2.3788079470198675,
+      "grad_norm": 2.0836849212646484,
+      "learning_rate": 2.0917355432419856e-06,
+      "loss": 1.0693,
+      "step": 3143
+    },
+    {
+      "epoch": 2.3795648060548724,
+      "grad_norm": 2.1676831245422363,
+      "learning_rate": 2.0868266121958895e-06,
+      "loss": 1.0786,
+      "step": 3144
+    },
+    {
+      "epoch": 2.380321665089877,
+      "grad_norm": 1.94955313205719,
+      "learning_rate": 2.0819227825052655e-06,
+      "loss": 1.04,
+      "step": 3145
+    },
+    {
+      "epoch": 2.3810785241248817,
+      "grad_norm": 2.1713364124298096,
+      "learning_rate": 2.0770240573310464e-06,
+      "loss": 1.0997,
+      "step": 3146
+    },
+    {
+      "epoch": 2.3818353831598866,
+      "grad_norm": 1.9667205810546875,
+      "learning_rate": 2.07213043983088e-06,
+      "loss": 1.0526,
+      "step": 3147
+    },
+    {
+      "epoch": 2.382592242194891,
+      "grad_norm": 2.1327104568481445,
+      "learning_rate": 2.067241933159111e-06,
+      "loss": 1.0717,
+      "step": 3148
+    },
+    {
+      "epoch": 2.383349101229896,
+      "grad_norm": 2.0339972972869873,
+      "learning_rate": 2.0623585404668027e-06,
+      "loss": 1.0583,
+      "step": 3149
+    },
+    {
+      "epoch": 2.384105960264901,
+      "grad_norm": 2.172558069229126,
+      "learning_rate": 2.0574802649017087e-06,
+      "loss": 1.0814,
+      "step": 3150
+    },
+    {
+      "epoch": 2.3848628192999053,
+      "grad_norm": 2.1538596153259277,
+      "learning_rate": 2.0526071096082958e-06,
+      "loss": 1.0713,
+      "step": 3151
+    },
+    {
+      "epoch": 2.38561967833491,
+      "grad_norm": 2.009945869445801,
+      "learning_rate": 2.0477390777277238e-06,
+      "loss": 1.0783,
+      "step": 3152
+    },
+    {
+      "epoch": 2.386376537369915,
+      "grad_norm": 2.024836540222168,
+      "learning_rate": 2.042876172397855e-06,
+      "loss": 1.0648,
+      "step": 3153
+    },
+    {
+      "epoch": 2.3871333964049195,
+      "grad_norm": 1.9101126194000244,
+      "learning_rate": 2.0380183967532398e-06,
+      "loss": 1.0476,
+      "step": 3154
+    },
+    {
+      "epoch": 2.3878902554399244,
+      "grad_norm": 1.9727280139923096,
+      "learning_rate": 2.033165753925127e-06,
+      "loss": 1.0658,
+      "step": 3155
+    },
+    {
+      "epoch": 2.388647114474929,
+      "grad_norm": 2.0228683948516846,
+      "learning_rate": 2.0283182470414605e-06,
+      "loss": 1.0536,
+      "step": 3156
+    },
+    {
+      "epoch": 2.3894039735099337,
+      "grad_norm": 2.0721933841705322,
+      "learning_rate": 2.0234758792268626e-06,
+      "loss": 1.1227,
+      "step": 3157
+    },
+    {
+      "epoch": 2.3901608325449386,
+      "grad_norm": 2.15034818649292,
+      "learning_rate": 2.0186386536026563e-06,
+      "loss": 1.1274,
+      "step": 3158
+    },
+    {
+      "epoch": 2.390917691579943,
+      "grad_norm": 2.1228785514831543,
+      "learning_rate": 2.0138065732868377e-06,
+      "loss": 1.1003,
+      "step": 3159
+    },
+    {
+      "epoch": 2.391674550614948,
+      "grad_norm": 1.9539825916290283,
+      "learning_rate": 2.008979641394094e-06,
+      "loss": 1.0618,
+      "step": 3160
+    },
+    {
+      "epoch": 2.392431409649953,
+      "grad_norm": 2.054403066635132,
+      "learning_rate": 2.0041578610357924e-06,
+      "loss": 1.0761,
+      "step": 3161
+    },
+    {
+      "epoch": 2.3931882686849573,
+      "grad_norm": 2.2176952362060547,
+      "learning_rate": 1.9993412353199797e-06,
+      "loss": 1.0955,
+      "step": 3162
+    },
+    {
+      "epoch": 2.393945127719962,
+      "grad_norm": 1.9848977327346802,
+      "learning_rate": 1.9945297673513813e-06,
+      "loss": 1.0478,
+      "step": 3163
+    },
+    {
+      "epoch": 2.394701986754967,
+      "grad_norm": 2.0067944526672363,
+      "learning_rate": 1.9897234602313935e-06,
+      "loss": 1.0606,
+      "step": 3164
+    },
+    {
+      "epoch": 2.3954588457899715,
+      "grad_norm": 2.1978671550750732,
+      "learning_rate": 1.9849223170580863e-06,
+      "loss": 1.0673,
+      "step": 3165
+    },
+    {
+      "epoch": 2.3962157048249764,
+      "grad_norm": 2.103545665740967,
+      "learning_rate": 1.9801263409262044e-06,
+      "loss": 1.0697,
+      "step": 3166
+    },
+    {
+      "epoch": 2.3969725638599813,
+      "grad_norm": 2.0121638774871826,
+      "learning_rate": 1.975335534927164e-06,
+      "loss": 1.1051,
+      "step": 3167
+    },
+    {
+      "epoch": 2.3977294228949857,
+      "grad_norm": 2.227635383605957,
+      "learning_rate": 1.970549902149043e-06,
+      "loss": 1.1104,
+      "step": 3168
+    },
+    {
+      "epoch": 2.3984862819299906,
+      "grad_norm": 1.98334538936615,
+      "learning_rate": 1.965769445676593e-06,
+      "loss": 1.0538,
+      "step": 3169
+    },
+    {
+      "epoch": 2.3992431409649955,
+      "grad_norm": 2.092841863632202,
+      "learning_rate": 1.9609941685912137e-06,
+      "loss": 1.076,
+      "step": 3170
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.9714951515197754,
+      "learning_rate": 1.9562240739709797e-06,
+      "loss": 1.0408,
+      "step": 3171
+    },
+    {
+      "epoch": 2.400756859035005,
+      "grad_norm": 2.228931427001953,
+      "learning_rate": 1.951459164890623e-06,
+      "loss": 1.07,
+      "step": 3172
+    },
+    {
+      "epoch": 2.4015137180700092,
+      "grad_norm": 2.2674131393432617,
+      "learning_rate": 1.9466994444215306e-06,
+      "loss": 1.0301,
+      "step": 3173
+    },
+    {
+      "epoch": 2.402270577105014,
+      "grad_norm": 2.017943859100342,
+      "learning_rate": 1.941944915631745e-06,
+      "loss": 1.0159,
+      "step": 3174
+    },
+    {
+      "epoch": 2.403027436140019,
+      "grad_norm": 2.2940430641174316,
+      "learning_rate": 1.937195581585966e-06,
+      "loss": 1.0656,
+      "step": 3175
+    },
+    {
+      "epoch": 2.403784295175024,
+      "grad_norm": 2.035090684890747,
+      "learning_rate": 1.9324514453455404e-06,
+      "loss": 1.0572,
+      "step": 3176
+    },
+    {
+      "epoch": 2.4045411542100283,
+      "grad_norm": 2.0770013332366943,
+      "learning_rate": 1.927712509968461e-06,
+      "loss": 1.0763,
+      "step": 3177
+    },
+    {
+      "epoch": 2.4052980132450332,
+      "grad_norm": 2.1525466442108154,
+      "learning_rate": 1.9229787785093784e-06,
+      "loss": 1.069,
+      "step": 3178
+    },
+    {
+      "epoch": 2.4060548722800377,
+      "grad_norm": 1.9798002243041992,
+      "learning_rate": 1.9182502540195826e-06,
+      "loss": 1.04,
+      "step": 3179
+    },
+    {
+      "epoch": 2.4068117313150426,
+      "grad_norm": 2.0167088508605957,
+      "learning_rate": 1.9135269395470117e-06,
+      "loss": 1.0951,
+      "step": 3180
+    },
+    {
+      "epoch": 2.4075685903500474,
+      "grad_norm": 1.9377844333648682,
+      "learning_rate": 1.908808838136235e-06,
+      "loss": 1.0678,
+      "step": 3181
+    },
+    {
+      "epoch": 2.408325449385052,
+      "grad_norm": 1.8350154161453247,
+      "learning_rate": 1.904095952828474e-06,
+      "loss": 1.0734,
+      "step": 3182
+    },
+    {
+      "epoch": 2.4090823084200568,
+      "grad_norm": 2.102295160293579,
+      "learning_rate": 1.8993882866615832e-06,
+      "loss": 1.0446,
+      "step": 3183
+    },
+    {
+      "epoch": 2.4098391674550617,
+      "grad_norm": 2.003739833831787,
+      "learning_rate": 1.8946858426700479e-06,
+      "loss": 1.1126,
+      "step": 3184
+    },
+    {
+      "epoch": 2.410596026490066,
+      "grad_norm": 2.410670280456543,
+      "learning_rate": 1.8899886238849949e-06,
+      "loss": 1.0803,
+      "step": 3185
+    },
+    {
+      "epoch": 2.411352885525071,
+      "grad_norm": 1.9924821853637695,
+      "learning_rate": 1.8852966333341822e-06,
+      "loss": 1.0713,
+      "step": 3186
+    },
+    {
+      "epoch": 2.412109744560076,
+      "grad_norm": 2.1071012020111084,
+      "learning_rate": 1.880609874041989e-06,
+      "loss": 1.0772,
+      "step": 3187
+    },
+    {
+      "epoch": 2.4128666035950803,
+      "grad_norm": 2.1172494888305664,
+      "learning_rate": 1.8759283490294333e-06,
+      "loss": 1.0584,
+      "step": 3188
+    },
+    {
+      "epoch": 2.413623462630085,
+      "grad_norm": 2.2310361862182617,
+      "learning_rate": 1.8712520613141525e-06,
+      "loss": 1.039,
+      "step": 3189
+    },
+    {
+      "epoch": 2.4143803216650896,
+      "grad_norm": 1.90047287940979,
+      "learning_rate": 1.8665810139104157e-06,
+      "loss": 1.0776,
+      "step": 3190
+    },
+    {
+      "epoch": 2.4151371807000945,
+      "grad_norm": 2.110424041748047,
+      "learning_rate": 1.8619152098291044e-06,
+      "loss": 1.0546,
+      "step": 3191
+    },
+    {
+      "epoch": 2.4158940397350994,
+      "grad_norm": 1.9257051944732666,
+      "learning_rate": 1.8572546520777214e-06,
+      "loss": 1.0843,
+      "step": 3192
+    },
+    {
+      "epoch": 2.4166508987701043,
+      "grad_norm": 2.1789603233337402,
+      "learning_rate": 1.852599343660396e-06,
+      "loss": 1.0479,
+      "step": 3193
+    },
+    {
+      "epoch": 2.4174077578051087,
+      "grad_norm": 1.8343688249588013,
+      "learning_rate": 1.847949287577868e-06,
+      "loss": 0.9984,
+      "step": 3194
+    },
+    {
+      "epoch": 2.4181646168401136,
+      "grad_norm": 2.1044609546661377,
+      "learning_rate": 1.843304486827492e-06,
+      "loss": 1.0603,
+      "step": 3195
+    },
+    {
+      "epoch": 2.418921475875118,
+      "grad_norm": 2.0383095741271973,
+      "learning_rate": 1.838664944403236e-06,
+      "loss": 1.0733,
+      "step": 3196
+    },
+    {
+      "epoch": 2.419678334910123,
+      "grad_norm": 1.9661799669265747,
+      "learning_rate": 1.8340306632956847e-06,
+      "loss": 1.0257,
+      "step": 3197
+    },
+    {
+      "epoch": 2.420435193945128,
+      "grad_norm": 2.1200876235961914,
+      "learning_rate": 1.8294016464920133e-06,
+      "loss": 1.0597,
+      "step": 3198
+    },
+    {
+      "epoch": 2.4211920529801323,
+      "grad_norm": 2.138803005218506,
+      "learning_rate": 1.8247778969760206e-06,
+      "loss": 1.0518,
+      "step": 3199
+    },
+    {
+      "epoch": 2.421948912015137,
+      "grad_norm": 1.937528371810913,
+      "learning_rate": 1.8201594177281053e-06,
+      "loss": 1.1165,
+      "step": 3200
+    },
+    {
+      "epoch": 2.422705771050142,
+      "grad_norm": 1.99111008644104,
+      "learning_rate": 1.8155462117252693e-06,
+      "loss": 1.0687,
+      "step": 3201
+    },
+    {
+      "epoch": 2.4234626300851465,
+      "grad_norm": 2.2547271251678467,
+      "learning_rate": 1.8109382819411164e-06,
+      "loss": 1.0613,
+      "step": 3202
+    },
+    {
+      "epoch": 2.4242194891201514,
+      "grad_norm": 1.8853436708450317,
+      "learning_rate": 1.8063356313458443e-06,
+      "loss": 1.1003,
+      "step": 3203
+    },
+    {
+      "epoch": 2.4249763481551563,
+      "grad_norm": 1.9870060682296753,
+      "learning_rate": 1.801738262906254e-06,
+      "loss": 1.0924,
+      "step": 3204
+    },
+    {
+      "epoch": 2.4257332071901607,
+      "grad_norm": 1.9391242265701294,
+      "learning_rate": 1.7971461795857367e-06,
+      "loss": 1.0116,
+      "step": 3205
+    },
+    {
+      "epoch": 2.4264900662251656,
+      "grad_norm": 2.092609167098999,
+      "learning_rate": 1.7925593843442798e-06,
+      "loss": 1.1001,
+      "step": 3206
+    },
+    {
+      "epoch": 2.42724692526017,
+      "grad_norm": 2.029475212097168,
+      "learning_rate": 1.787977880138463e-06,
+      "loss": 1.0716,
+      "step": 3207
+    },
+    {
+      "epoch": 2.428003784295175,
+      "grad_norm": 2.054161787033081,
+      "learning_rate": 1.783401669921456e-06,
+      "loss": 1.0752,
+      "step": 3208
+    },
+    {
+      "epoch": 2.42876064333018,
+      "grad_norm": 1.762110710144043,
+      "learning_rate": 1.7788307566430083e-06,
+      "loss": 1.0639,
+      "step": 3209
+    },
+    {
+      "epoch": 2.4295175023651847,
+      "grad_norm": 2.044447898864746,
+      "learning_rate": 1.774265143249467e-06,
+      "loss": 1.0905,
+      "step": 3210
+    },
+    {
+      "epoch": 2.430274361400189,
+      "grad_norm": 2.024933338165283,
+      "learning_rate": 1.7697048326837516e-06,
+      "loss": 1.0152,
+      "step": 3211
+    },
+    {
+      "epoch": 2.431031220435194,
+      "grad_norm": 1.9569860696792603,
+      "learning_rate": 1.7651498278853708e-06,
+      "loss": 1.0603,
+      "step": 3212
+    },
+    {
+      "epoch": 2.4317880794701985,
+      "grad_norm": 2.0603296756744385,
+      "learning_rate": 1.760600131790414e-06,
+      "loss": 1.1086,
+      "step": 3213
+    },
+    {
+      "epoch": 2.4325449385052034,
+      "grad_norm": 1.8924018144607544,
+      "learning_rate": 1.7560557473315413e-06,
+      "loss": 1.0509,
+      "step": 3214
+    },
+    {
+      "epoch": 2.4333017975402083,
+      "grad_norm": 1.9490152597427368,
+      "learning_rate": 1.7515166774379947e-06,
+      "loss": 1.0518,
+      "step": 3215
+    },
+    {
+      "epoch": 2.4340586565752127,
+      "grad_norm": 2.046515703201294,
+      "learning_rate": 1.746982925035591e-06,
+      "loss": 1.0466,
+      "step": 3216
+    },
+    {
+      "epoch": 2.4348155156102176,
+      "grad_norm": 2.0436580181121826,
+      "learning_rate": 1.7424544930467205e-06,
+      "loss": 1.0642,
+      "step": 3217
+    },
+    {
+      "epoch": 2.4355723746452225,
+      "grad_norm": 2.1087872982025146,
+      "learning_rate": 1.7379313843903357e-06,
+      "loss": 1.0838,
+      "step": 3218
+    },
+    {
+      "epoch": 2.436329233680227,
+      "grad_norm": 1.896474838256836,
+      "learning_rate": 1.7334136019819681e-06,
+      "loss": 1.0678,
+      "step": 3219
+    },
+    {
+      "epoch": 2.437086092715232,
+      "grad_norm": 2.087778091430664,
+      "learning_rate": 1.7289011487337059e-06,
+      "loss": 1.0402,
+      "step": 3220
+    },
+    {
+      "epoch": 2.4378429517502367,
+      "grad_norm": 1.9922432899475098,
+      "learning_rate": 1.7243940275542126e-06,
+      "loss": 1.119,
+      "step": 3221
+    },
+    {
+      "epoch": 2.438599810785241,
+      "grad_norm": 2.1278886795043945,
+      "learning_rate": 1.7198922413487073e-06,
+      "loss": 1.1303,
+      "step": 3222
+    },
+    {
+      "epoch": 2.439356669820246,
+      "grad_norm": 2.059356689453125,
+      "learning_rate": 1.7153957930189735e-06,
+      "loss": 1.0732,
+      "step": 3223
+    },
+    {
+      "epoch": 2.440113528855251,
+      "grad_norm": 2.127638578414917,
+      "learning_rate": 1.7109046854633587e-06,
+      "loss": 1.0715,
+      "step": 3224
+    },
+    {
+      "epoch": 2.4408703878902553,
+      "grad_norm": 2.0281498432159424,
+      "learning_rate": 1.7064189215767526e-06,
+      "loss": 1.0378,
+      "step": 3225
+    },
+    {
+      "epoch": 2.4416272469252602,
+      "grad_norm": 2.8992788791656494,
+      "learning_rate": 1.7019385042506134e-06,
+      "loss": 1.0982,
+      "step": 3226
+    },
+    {
+      "epoch": 2.442384105960265,
+      "grad_norm": 2.046388626098633,
+      "learning_rate": 1.697463436372951e-06,
+      "loss": 1.0689,
+      "step": 3227
+    },
+    {
+      "epoch": 2.4431409649952696,
+      "grad_norm": 1.9682432413101196,
+      "learning_rate": 1.692993720828327e-06,
+      "loss": 1.0886,
+      "step": 3228
+    },
+    {
+      "epoch": 2.4438978240302744,
+      "grad_norm": 2.190717935562134,
+      "learning_rate": 1.6885293604978495e-06,
+      "loss": 1.0941,
+      "step": 3229
+    },
+    {
+      "epoch": 2.444654683065279,
+      "grad_norm": 2.263735294342041,
+      "learning_rate": 1.6840703582591808e-06,
+      "loss": 1.1485,
+      "step": 3230
+    },
+    {
+      "epoch": 2.4454115421002838,
+      "grad_norm": 1.917043924331665,
+      "learning_rate": 1.6796167169865243e-06,
+      "loss": 1.044,
+      "step": 3231
+    },
+    {
+      "epoch": 2.4461684011352887,
+      "grad_norm": 2.120823621749878,
+      "learning_rate": 1.6751684395506248e-06,
+      "loss": 1.0492,
+      "step": 3232
+    },
+    {
+      "epoch": 2.446925260170293,
+      "grad_norm": 1.9636114835739136,
+      "learning_rate": 1.6707255288187776e-06,
+      "loss": 1.0525,
+      "step": 3233
+    },
+    {
+      "epoch": 2.447682119205298,
+      "grad_norm": 2.068773031234741,
+      "learning_rate": 1.6662879876548164e-06,
+      "loss": 1.1185,
+      "step": 3234
+    },
+    {
+      "epoch": 2.448438978240303,
+      "grad_norm": 1.9608315229415894,
+      "learning_rate": 1.661855818919112e-06,
+      "loss": 1.0705,
+      "step": 3235
+    },
+    {
+      "epoch": 2.4491958372753073,
+      "grad_norm": 2.037750005722046,
+      "learning_rate": 1.65742902546857e-06,
+      "loss": 1.1109,
+      "step": 3236
+    },
+    {
+      "epoch": 2.449952696310312,
+      "grad_norm": 2.2666871547698975,
+      "learning_rate": 1.653007610156637e-06,
+      "loss": 1.0955,
+      "step": 3237
+    },
+    {
+      "epoch": 2.450709555345317,
+      "grad_norm": 2.0493760108947754,
+      "learning_rate": 1.6485915758332899e-06,
+      "loss": 1.0354,
+      "step": 3238
+    },
+    {
+      "epoch": 2.4514664143803215,
+      "grad_norm": 2.0443224906921387,
+      "learning_rate": 1.6441809253450347e-06,
+      "loss": 1.0605,
+      "step": 3239
+    },
+    {
+      "epoch": 2.4522232734153264,
+      "grad_norm": 1.7710448503494263,
+      "learning_rate": 1.6397756615349103e-06,
+      "loss": 1.064,
+      "step": 3240
+    },
+    {
+      "epoch": 2.4529801324503313,
+      "grad_norm": 2.0574166774749756,
+      "learning_rate": 1.6353757872424848e-06,
+      "loss": 1.0986,
+      "step": 3241
+    },
+    {
+      "epoch": 2.4537369914853357,
+      "grad_norm": 2.0350000858306885,
+      "learning_rate": 1.6309813053038476e-06,
+      "loss": 1.0722,
+      "step": 3242
+    },
+    {
+      "epoch": 2.4544938505203406,
+      "grad_norm": 2.178621530532837,
+      "learning_rate": 1.6265922185516136e-06,
+      "loss": 1.0777,
+      "step": 3243
+    },
+    {
+      "epoch": 2.4552507095553455,
+      "grad_norm": 1.874701976776123,
+      "learning_rate": 1.6222085298149237e-06,
+      "loss": 1.0678,
+      "step": 3244
+    },
+    {
+      "epoch": 2.45600756859035,
+      "grad_norm": 2.080073356628418,
+      "learning_rate": 1.617830241919439e-06,
+      "loss": 1.1109,
+      "step": 3245
+    },
+    {
+      "epoch": 2.456764427625355,
+      "grad_norm": 2.067389488220215,
+      "learning_rate": 1.6134573576873347e-06,
+      "loss": 1.0613,
+      "step": 3246
+    },
+    {
+      "epoch": 2.4575212866603593,
+      "grad_norm": 2.2184066772460938,
+      "learning_rate": 1.6090898799373013e-06,
+      "loss": 1.0445,
+      "step": 3247
+    },
+    {
+      "epoch": 2.458278145695364,
+      "grad_norm": 2.1395821571350098,
+      "learning_rate": 1.6047278114845524e-06,
+      "loss": 1.1291,
+      "step": 3248
+    },
+    {
+      "epoch": 2.459035004730369,
+      "grad_norm": 1.878059983253479,
+      "learning_rate": 1.6003711551408108e-06,
+      "loss": 1.1197,
+      "step": 3249
+    },
+    {
+      "epoch": 2.4597918637653735,
+      "grad_norm": 2.079202651977539,
+      "learning_rate": 1.5960199137143096e-06,
+      "loss": 1.0601,
+      "step": 3250
+    },
+    {
+      "epoch": 2.4605487228003784,
+      "grad_norm": 2.1114516258239746,
+      "learning_rate": 1.5916740900097936e-06,
+      "loss": 1.0981,
+      "step": 3251
+    },
+    {
+      "epoch": 2.4613055818353833,
+      "grad_norm": 2.0840392112731934,
+      "learning_rate": 1.5873336868285188e-06,
+      "loss": 1.0901,
+      "step": 3252
+    },
+    {
+      "epoch": 2.4620624408703877,
+      "grad_norm": 2.1868133544921875,
+      "learning_rate": 1.582998706968233e-06,
+      "loss": 1.1103,
+      "step": 3253
+    },
+    {
+      "epoch": 2.4628192999053926,
+      "grad_norm": 2.0153892040252686,
+      "learning_rate": 1.5786691532232047e-06,
+      "loss": 1.0825,
+      "step": 3254
+    },
+    {
+      "epoch": 2.4635761589403975,
+      "grad_norm": 2.147407054901123,
+      "learning_rate": 1.5743450283841957e-06,
+      "loss": 1.0529,
+      "step": 3255
+    },
+    {
+      "epoch": 2.464333017975402,
+      "grad_norm": 2.2476887702941895,
+      "learning_rate": 1.5700263352384732e-06,
+      "loss": 1.0551,
+      "step": 3256
+    },
+    {
+      "epoch": 2.465089877010407,
+      "grad_norm": 1.897383689880371,
+      "learning_rate": 1.5657130765698006e-06,
+      "loss": 1.0773,
+      "step": 3257
+    },
+    {
+      "epoch": 2.4658467360454117,
+      "grad_norm": 2.2343618869781494,
+      "learning_rate": 1.56140525515844e-06,
+      "loss": 1.0388,
+      "step": 3258
+    },
+    {
+      "epoch": 2.466603595080416,
+      "grad_norm": 2.182474136352539,
+      "learning_rate": 1.5571028737811414e-06,
+      "loss": 1.0837,
+      "step": 3259
+    },
+    {
+      "epoch": 2.467360454115421,
+      "grad_norm": 1.94349205493927,
+      "learning_rate": 1.5528059352111586e-06,
+      "loss": 1.0374,
+      "step": 3260
+    },
+    {
+      "epoch": 2.468117313150426,
+      "grad_norm": 2.3165524005889893,
+      "learning_rate": 1.5485144422182325e-06,
+      "loss": 1.049,
+      "step": 3261
+    },
+    {
+      "epoch": 2.4688741721854304,
+      "grad_norm": 2.2094292640686035,
+      "learning_rate": 1.5442283975685937e-06,
+      "loss": 1.0934,
+      "step": 3262
+    },
+    {
+      "epoch": 2.4696310312204353,
+      "grad_norm": 2.0244195461273193,
+      "learning_rate": 1.5399478040249638e-06,
+      "loss": 1.044,
+      "step": 3263
+    },
+    {
+      "epoch": 2.4703878902554397,
+      "grad_norm": 1.9300179481506348,
+      "learning_rate": 1.5356726643465427e-06,
+      "loss": 1.1156,
+      "step": 3264
+    },
+    {
+      "epoch": 2.4711447492904446,
+      "grad_norm": 2.06846022605896,
+      "learning_rate": 1.5314029812890258e-06,
+      "loss": 1.012,
+      "step": 3265
+    },
+    {
+      "epoch": 2.4719016083254495,
+      "grad_norm": 2.2604005336761475,
+      "learning_rate": 1.5271387576045804e-06,
+      "loss": 1.1042,
+      "step": 3266
+    },
+    {
+      "epoch": 2.4726584673604544,
+      "grad_norm": 2.3489127159118652,
+      "learning_rate": 1.5228799960418639e-06,
+      "loss": 1.05,
+      "step": 3267
+    },
+    {
+      "epoch": 2.473415326395459,
+      "grad_norm": 2.0200610160827637,
+      "learning_rate": 1.518626699346009e-06,
+      "loss": 1.1298,
+      "step": 3268
+    },
+    {
+      "epoch": 2.4741721854304637,
+      "grad_norm": 2.0748353004455566,
+      "learning_rate": 1.514378870258623e-06,
+      "loss": 1.0477,
+      "step": 3269
+    },
+    {
+      "epoch": 2.474929044465468,
+      "grad_norm": 2.0303914546966553,
+      "learning_rate": 1.510136511517792e-06,
+      "loss": 1.0319,
+      "step": 3270
+    },
+    {
+      "epoch": 2.475685903500473,
+      "grad_norm": 1.9617363214492798,
+      "learning_rate": 1.5058996258580788e-06,
+      "loss": 1.1149,
+      "step": 3271
+    },
+    {
+      "epoch": 2.476442762535478,
+      "grad_norm": 2.0544240474700928,
+      "learning_rate": 1.5016682160105153e-06,
+      "loss": 1.0733,
+      "step": 3272
+    },
+    {
+      "epoch": 2.4771996215704823,
+      "grad_norm": 2.3402349948883057,
+      "learning_rate": 1.4974422847026002e-06,
+      "loss": 1.0615,
+      "step": 3273
+    },
+    {
+      "epoch": 2.4779564806054872,
+      "grad_norm": 2.0029454231262207,
+      "learning_rate": 1.4932218346583082e-06,
+      "loss": 1.0688,
+      "step": 3274
+    },
+    {
+      "epoch": 2.478713339640492,
+      "grad_norm": 1.8487077951431274,
+      "learning_rate": 1.4890068685980732e-06,
+      "loss": 1.1071,
+      "step": 3275
+    },
+    {
+      "epoch": 2.4794701986754966,
+      "grad_norm": 2.282620906829834,
+      "learning_rate": 1.4847973892388003e-06,
+      "loss": 1.0802,
+      "step": 3276
+    },
+    {
+      "epoch": 2.4802270577105014,
+      "grad_norm": 1.9295916557312012,
+      "learning_rate": 1.4805933992938547e-06,
+      "loss": 1.0663,
+      "step": 3277
+    },
+    {
+      "epoch": 2.4809839167455063,
+      "grad_norm": 2.0331246852874756,
+      "learning_rate": 1.476394901473066e-06,
+      "loss": 1.0906,
+      "step": 3278
+    },
+    {
+      "epoch": 2.4817407757805108,
+      "grad_norm": 2.1244752407073975,
+      "learning_rate": 1.4722018984827247e-06,
+      "loss": 1.0655,
+      "step": 3279
+    },
+    {
+      "epoch": 2.4824976348155157,
+      "grad_norm": 2.140397787094116,
+      "learning_rate": 1.4680143930255675e-06,
+      "loss": 1.0969,
+      "step": 3280
+    },
+    {
+      "epoch": 2.48325449385052,
+      "grad_norm": 2.1430792808532715,
+      "learning_rate": 1.4638323878008022e-06,
+      "loss": 1.06,
+      "step": 3281
+    },
+    {
+      "epoch": 2.484011352885525,
+      "grad_norm": 2.28765606880188,
+      "learning_rate": 1.459655885504086e-06,
+      "loss": 1.1159,
+      "step": 3282
+    },
+    {
+      "epoch": 2.48476821192053,
+      "grad_norm": 1.9069238901138306,
+      "learning_rate": 1.455484888827526e-06,
+      "loss": 1.0083,
+      "step": 3283
+    },
+    {
+      "epoch": 2.4855250709555348,
+      "grad_norm": 1.848893642425537,
+      "learning_rate": 1.4513194004596865e-06,
+      "loss": 1.0527,
+      "step": 3284
+    },
+    {
+      "epoch": 2.486281929990539,
+      "grad_norm": 1.8594064712524414,
+      "learning_rate": 1.4471594230855774e-06,
+      "loss": 1.0815,
+      "step": 3285
+    },
+    {
+      "epoch": 2.487038789025544,
+      "grad_norm": 1.9376791715621948,
+      "learning_rate": 1.4430049593866543e-06,
+      "loss": 1.0403,
+      "step": 3286
+    },
+    {
+      "epoch": 2.4877956480605485,
+      "grad_norm": 2.031545639038086,
+      "learning_rate": 1.4388560120408215e-06,
+      "loss": 1.0378,
+      "step": 3287
+    },
+    {
+      "epoch": 2.4885525070955534,
+      "grad_norm": 2.0290255546569824,
+      "learning_rate": 1.4347125837224266e-06,
+      "loss": 1.05,
+      "step": 3288
+    },
+    {
+      "epoch": 2.4893093661305583,
+      "grad_norm": 2.070533037185669,
+      "learning_rate": 1.4305746771022623e-06,
+      "loss": 1.0854,
+      "step": 3289
+    },
+    {
+      "epoch": 2.4900662251655628,
+      "grad_norm": 3.2161612510681152,
+      "learning_rate": 1.4264422948475618e-06,
+      "loss": 1.0484,
+      "step": 3290
+    },
+    {
+      "epoch": 2.4908230842005676,
+      "grad_norm": 2.044058322906494,
+      "learning_rate": 1.4223154396219906e-06,
+      "loss": 1.0543,
+      "step": 3291
+    },
+    {
+      "epoch": 2.4915799432355725,
+      "grad_norm": 1.9972931146621704,
+      "learning_rate": 1.4181941140856595e-06,
+      "loss": 1.0482,
+      "step": 3292
+    },
+    {
+      "epoch": 2.492336802270577,
+      "grad_norm": 2.115438222885132,
+      "learning_rate": 1.4140783208951142e-06,
+      "loss": 1.1006,
+      "step": 3293
+    },
+    {
+      "epoch": 2.493093661305582,
+      "grad_norm": 1.9351952075958252,
+      "learning_rate": 1.4099680627033266e-06,
+      "loss": 1.0229,
+      "step": 3294
+    },
+    {
+      "epoch": 2.4938505203405867,
+      "grad_norm": 2.0593087673187256,
+      "learning_rate": 1.4058633421597104e-06,
+      "loss": 1.0653,
+      "step": 3295
+    },
+    {
+      "epoch": 2.494607379375591,
+      "grad_norm": 2.1452414989471436,
+      "learning_rate": 1.4017641619101074e-06,
+      "loss": 1.1039,
+      "step": 3296
+    },
+    {
+      "epoch": 2.495364238410596,
+      "grad_norm": 1.799978494644165,
+      "learning_rate": 1.3976705245967832e-06,
+      "loss": 1.0778,
+      "step": 3297
+    },
+    {
+      "epoch": 2.4961210974456005,
+      "grad_norm": 1.9537346363067627,
+      "learning_rate": 1.3935824328584335e-06,
+      "loss": 1.0688,
+      "step": 3298
+    },
+    {
+      "epoch": 2.4968779564806054,
+      "grad_norm": 2.1485798358917236,
+      "learning_rate": 1.3894998893301829e-06,
+      "loss": 1.0416,
+      "step": 3299
+    },
+    {
+      "epoch": 2.4976348155156103,
+      "grad_norm": 2.167556047439575,
+      "learning_rate": 1.3854228966435768e-06,
+      "loss": 1.0693,
+      "step": 3300
+    },
+    {
+      "epoch": 2.498391674550615,
+      "grad_norm": 2.1335930824279785,
+      "learning_rate": 1.3813514574265815e-06,
+      "loss": 1.0789,
+      "step": 3301
+    },
+    {
+      "epoch": 2.4991485335856196,
+      "grad_norm": 1.9826858043670654,
+      "learning_rate": 1.3772855743035818e-06,
+      "loss": 1.071,
+      "step": 3302
+    },
+    {
+      "epoch": 2.4999053926206245,
+      "grad_norm": 1.9714877605438232,
+      "learning_rate": 1.3732252498953874e-06,
+      "loss": 1.0915,
+      "step": 3303
+    },
+    {
+      "epoch": 2.500662251655629,
+      "grad_norm": 1.859167218208313,
+      "learning_rate": 1.3691704868192202e-06,
+      "loss": 1.0345,
+      "step": 3304
+    },
+    {
+      "epoch": 2.501419110690634,
+      "grad_norm": 1.9658515453338623,
+      "learning_rate": 1.3651212876887181e-06,
+      "loss": 1.0506,
+      "step": 3305
+    },
+    {
+      "epoch": 2.5021759697256387,
+      "grad_norm": 2.1254079341888428,
+      "learning_rate": 1.361077655113935e-06,
+      "loss": 1.0721,
+      "step": 3306
+    },
+    {
+      "epoch": 2.5029328287606436,
+      "grad_norm": 2.0176279544830322,
+      "learning_rate": 1.3570395917013365e-06,
+      "loss": 1.0711,
+      "step": 3307
+    },
+    {
+      "epoch": 2.503689687795648,
+      "grad_norm": 2.091989278793335,
+      "learning_rate": 1.353007100053791e-06,
+      "loss": 1.1103,
+      "step": 3308
+    },
+    {
+      "epoch": 2.504446546830653,
+      "grad_norm": 2.0587007999420166,
+      "learning_rate": 1.348980182770584e-06,
+      "loss": 1.0569,
+      "step": 3309
+    },
+    {
+      "epoch": 2.5052034058656574,
+      "grad_norm": 1.9162755012512207,
+      "learning_rate": 1.344958842447405e-06,
+      "loss": 1.0914,
+      "step": 3310
+    },
+    {
+      "epoch": 2.5059602649006623,
+      "grad_norm": 1.995250940322876,
+      "learning_rate": 1.3409430816763478e-06,
+      "loss": 1.0789,
+      "step": 3311
+    },
+    {
+      "epoch": 2.506717123935667,
+      "grad_norm": 2.189922571182251,
+      "learning_rate": 1.3369329030459152e-06,
+      "loss": 1.1112,
+      "step": 3312
+    },
+    {
+      "epoch": 2.5074739829706716,
+      "grad_norm": 2.006410837173462,
+      "learning_rate": 1.3329283091410014e-06,
+      "loss": 1.0911,
+      "step": 3313
+    },
+    {
+      "epoch": 2.5082308420056765,
+      "grad_norm": 1.9690699577331543,
+      "learning_rate": 1.3289293025429082e-06,
+      "loss": 1.0642,
+      "step": 3314
+    },
+    {
+      "epoch": 2.508987701040681,
+      "grad_norm": 1.9968689680099487,
+      "learning_rate": 1.324935885829334e-06,
+      "loss": 1.1037,
+      "step": 3315
+    },
+    {
+      "epoch": 2.509744560075686,
+      "grad_norm": 2.0873682498931885,
+      "learning_rate": 1.3209480615743746e-06,
+      "loss": 1.0804,
+      "step": 3316
+    },
+    {
+      "epoch": 2.5105014191106907,
+      "grad_norm": 2.100817918777466,
+      "learning_rate": 1.3169658323485212e-06,
+      "loss": 1.0313,
+      "step": 3317
+    },
+    {
+      "epoch": 2.5112582781456956,
+      "grad_norm": 2.1082022190093994,
+      "learning_rate": 1.3129892007186602e-06,
+      "loss": 1.0815,
+      "step": 3318
+    },
+    {
+      "epoch": 2.5120151371807,
+      "grad_norm": 2.102774143218994,
+      "learning_rate": 1.3090181692480642e-06,
+      "loss": 1.0529,
+      "step": 3319
+    },
+    {
+      "epoch": 2.512771996215705,
+      "grad_norm": 1.8931456804275513,
+      "learning_rate": 1.305052740496402e-06,
+      "loss": 1.029,
+      "step": 3320
+    },
+    {
+      "epoch": 2.5135288552507093,
+      "grad_norm": 1.9600942134857178,
+      "learning_rate": 1.301092917019724e-06,
+      "loss": 1.0499,
+      "step": 3321
+    },
+    {
+      "epoch": 2.5142857142857142,
+      "grad_norm": 1.9435330629348755,
+      "learning_rate": 1.2971387013704767e-06,
+      "loss": 1.0662,
+      "step": 3322
+    },
+    {
+      "epoch": 2.515042573320719,
+      "grad_norm": 2.6212613582611084,
+      "learning_rate": 1.2931900960974872e-06,
+      "loss": 1.0569,
+      "step": 3323
+    },
+    {
+      "epoch": 2.515799432355724,
+      "grad_norm": 2.1485402584075928,
+      "learning_rate": 1.2892471037459634e-06,
+      "loss": 1.0798,
+      "step": 3324
+    },
+    {
+      "epoch": 2.5165562913907285,
+      "grad_norm": 1.9852415323257446,
+      "learning_rate": 1.285309726857499e-06,
+      "loss": 1.0856,
+      "step": 3325
+    },
+    {
+      "epoch": 2.5173131504257333,
+      "grad_norm": 2.0448715686798096,
+      "learning_rate": 1.281377967970067e-06,
+      "loss": 1.0894,
+      "step": 3326
+    },
+    {
+      "epoch": 2.518070009460738,
+      "grad_norm": 1.966200590133667,
+      "learning_rate": 1.2774518296180222e-06,
+      "loss": 1.0583,
+      "step": 3327
+    },
+    {
+      "epoch": 2.5188268684957427,
+      "grad_norm": 2.0975255966186523,
+      "learning_rate": 1.2735313143320901e-06,
+      "loss": 1.1087,
+      "step": 3328
+    },
+    {
+      "epoch": 2.5195837275307476,
+      "grad_norm": 1.8325495719909668,
+      "learning_rate": 1.2696164246393766e-06,
+      "loss": 1.0857,
+      "step": 3329
+    },
+    {
+      "epoch": 2.520340586565752,
+      "grad_norm": 2.3163001537323,
+      "learning_rate": 1.265707163063358e-06,
+      "loss": 1.0535,
+      "step": 3330
+    },
+    {
+      "epoch": 2.521097445600757,
+      "grad_norm": 2.0799732208251953,
+      "learning_rate": 1.2618035321238856e-06,
+      "loss": 1.1036,
+      "step": 3331
+    },
+    {
+      "epoch": 2.5218543046357613,
+      "grad_norm": 1.9881048202514648,
+      "learning_rate": 1.257905534337181e-06,
+      "loss": 1.0693,
+      "step": 3332
+    },
+    {
+      "epoch": 2.522611163670766,
+      "grad_norm": 2.5260820388793945,
+      "learning_rate": 1.2540131722158336e-06,
+      "loss": 1.0339,
+      "step": 3333
+    },
+    {
+      "epoch": 2.523368022705771,
+      "grad_norm": 2.0032739639282227,
+      "learning_rate": 1.2501264482688052e-06,
+      "loss": 1.0888,
+      "step": 3334
+    },
+    {
+      "epoch": 2.524124881740776,
+      "grad_norm": 1.9609510898590088,
+      "learning_rate": 1.2462453650014107e-06,
+      "loss": 1.0654,
+      "step": 3335
+    },
+    {
+      "epoch": 2.5248817407757804,
+      "grad_norm": 2.018681049346924,
+      "learning_rate": 1.2423699249153408e-06,
+      "loss": 1.0763,
+      "step": 3336
+    },
+    {
+      "epoch": 2.5256385998107853,
+      "grad_norm": 1.9069342613220215,
+      "learning_rate": 1.2385001305086455e-06,
+      "loss": 1.0883,
+      "step": 3337
+    },
+    {
+      "epoch": 2.5263954588457898,
+      "grad_norm": 2.1848833560943604,
+      "learning_rate": 1.2346359842757345e-06,
+      "loss": 1.0894,
+      "step": 3338
+    },
+    {
+      "epoch": 2.5271523178807946,
+      "grad_norm": 2.210022449493408,
+      "learning_rate": 1.230777488707379e-06,
+      "loss": 1.0807,
+      "step": 3339
+    },
+    {
+      "epoch": 2.5279091769157995,
+      "grad_norm": 2.1791112422943115,
+      "learning_rate": 1.2269246462907065e-06,
+      "loss": 1.0707,
+      "step": 3340
+    },
+    {
+      "epoch": 2.5286660359508044,
+      "grad_norm": 1.9715594053268433,
+      "learning_rate": 1.2230774595092005e-06,
+      "loss": 1.0304,
+      "step": 3341
+    },
+    {
+      "epoch": 2.529422894985809,
+      "grad_norm": 2.082552433013916,
+      "learning_rate": 1.219235930842696e-06,
+      "loss": 1.088,
+      "step": 3342
+    },
+    {
+      "epoch": 2.5301797540208137,
+      "grad_norm": 2.0022144317626953,
+      "learning_rate": 1.215400062767385e-06,
+      "loss": 1.0153,
+      "step": 3343
+    },
+    {
+      "epoch": 2.530936613055818,
+      "grad_norm": 2.0545942783355713,
+      "learning_rate": 1.2115698577558096e-06,
+      "loss": 1.0895,
+      "step": 3344
+    },
+    {
+      "epoch": 2.531693472090823,
+      "grad_norm": 1.8354177474975586,
+      "learning_rate": 1.207745318276865e-06,
+      "loss": 1.082,
+      "step": 3345
+    },
+    {
+      "epoch": 2.532450331125828,
+      "grad_norm": 2.023404121398926,
+      "learning_rate": 1.203926446795787e-06,
+      "loss": 1.0675,
+      "step": 3346
+    },
+    {
+      "epoch": 2.5332071901608324,
+      "grad_norm": 1.8171032667160034,
+      "learning_rate": 1.2001132457741615e-06,
+      "loss": 1.0464,
+      "step": 3347
+    },
+    {
+      "epoch": 2.5339640491958373,
+      "grad_norm": 1.9232815504074097,
+      "learning_rate": 1.1963057176699249e-06,
+      "loss": 1.0647,
+      "step": 3348
+    },
+    {
+      "epoch": 2.5347209082308417,
+      "grad_norm": 1.9270484447479248,
+      "learning_rate": 1.1925038649373456e-06,
+      "loss": 1.1044,
+      "step": 3349
+    },
+    {
+      "epoch": 2.5354777672658466,
+      "grad_norm": 2.1430654525756836,
+      "learning_rate": 1.1887076900270418e-06,
+      "loss": 1.0809,
+      "step": 3350
+    },
+    {
+      "epoch": 2.5362346263008515,
+      "grad_norm": 2.0014898777008057,
+      "learning_rate": 1.1849171953859737e-06,
+      "loss": 1.0902,
+      "step": 3351
+    },
+    {
+      "epoch": 2.5369914853358564,
+      "grad_norm": 2.0257580280303955,
+      "learning_rate": 1.1811323834574302e-06,
+      "loss": 1.0911,
+      "step": 3352
+    },
+    {
+      "epoch": 2.537748344370861,
+      "grad_norm": 1.9925347566604614,
+      "learning_rate": 1.1773532566810477e-06,
+      "loss": 1.0666,
+      "step": 3353
+    },
+    {
+      "epoch": 2.5385052034058657,
+      "grad_norm": 2.072805404663086,
+      "learning_rate": 1.1735798174927917e-06,
+      "loss": 1.1,
+      "step": 3354
+    },
+    {
+      "epoch": 2.53926206244087,
+      "grad_norm": 2.0256335735321045,
+      "learning_rate": 1.1698120683249663e-06,
+      "loss": 1.0585,
+      "step": 3355
+    },
+    {
+      "epoch": 2.540018921475875,
+      "grad_norm": 2.217489004135132,
+      "learning_rate": 1.1660500116062037e-06,
+      "loss": 1.0761,
+      "step": 3356
+    },
+    {
+      "epoch": 2.54077578051088,
+      "grad_norm": 1.91505765914917,
+      "learning_rate": 1.1622936497614644e-06,
+      "loss": 1.0579,
+      "step": 3357
+    },
+    {
+      "epoch": 2.541532639545885,
+      "grad_norm": 1.9349195957183838,
+      "learning_rate": 1.1585429852120462e-06,
+      "loss": 1.074,
+      "step": 3358
+    },
+    {
+      "epoch": 2.5422894985808893,
+      "grad_norm": 1.915028691291809,
+      "learning_rate": 1.1547980203755697e-06,
+      "loss": 1.0717,
+      "step": 3359
+    },
+    {
+      "epoch": 2.543046357615894,
+      "grad_norm": 2.154155731201172,
+      "learning_rate": 1.1510587576659814e-06,
+      "loss": 1.057,
+      "step": 3360
+    },
+    {
+      "epoch": 2.5438032166508986,
+      "grad_norm": 2.6311988830566406,
+      "learning_rate": 1.1473251994935532e-06,
+      "loss": 1.1184,
+      "step": 3361
+    },
+    {
+      "epoch": 2.5445600756859035,
+      "grad_norm": 2.038876533508301,
+      "learning_rate": 1.1435973482648844e-06,
+      "loss": 1.0693,
+      "step": 3362
+    },
+    {
+      "epoch": 2.5453169347209084,
+      "grad_norm": 2.0545339584350586,
+      "learning_rate": 1.1398752063828815e-06,
+      "loss": 1.0887,
+      "step": 3363
+    },
+    {
+      "epoch": 2.546073793755913,
+      "grad_norm": 1.8213523626327515,
+      "learning_rate": 1.1361587762467873e-06,
+      "loss": 1.0866,
+      "step": 3364
+    },
+    {
+      "epoch": 2.5468306527909177,
+      "grad_norm": 2.055341958999634,
+      "learning_rate": 1.1324480602521524e-06,
+      "loss": 1.0923,
+      "step": 3365
+    },
+    {
+      "epoch": 2.5475875118259226,
+      "grad_norm": 2.2624013423919678,
+      "learning_rate": 1.1287430607908508e-06,
+      "loss": 1.0532,
+      "step": 3366
+    },
+    {
+      "epoch": 2.548344370860927,
+      "grad_norm": 2.7478582859039307,
+      "learning_rate": 1.1250437802510686e-06,
+      "loss": 1.0849,
+      "step": 3367
+    },
+    {
+      "epoch": 2.549101229895932,
+      "grad_norm": 1.8595587015151978,
+      "learning_rate": 1.1213502210173044e-06,
+      "loss": 1.0518,
+      "step": 3368
+    },
+    {
+      "epoch": 2.549858088930937,
+      "grad_norm": 2.0965359210968018,
+      "learning_rate": 1.1176623854703688e-06,
+      "loss": 1.0152,
+      "step": 3369
+    },
+    {
+      "epoch": 2.5506149479659412,
+      "grad_norm": 2.1459243297576904,
+      "learning_rate": 1.1139802759873852e-06,
+      "loss": 1.1309,
+      "step": 3370
+    },
+    {
+      "epoch": 2.551371807000946,
+      "grad_norm": 1.981099247932434,
+      "learning_rate": 1.110303894941786e-06,
+      "loss": 1.1094,
+      "step": 3371
+    },
+    {
+      "epoch": 2.5521286660359506,
+      "grad_norm": 2.0165510177612305,
+      "learning_rate": 1.10663324470331e-06,
+      "loss": 1.0812,
+      "step": 3372
+    },
+    {
+      "epoch": 2.5528855250709555,
+      "grad_norm": 1.839483380317688,
+      "learning_rate": 1.102968327638005e-06,
+      "loss": 1.0098,
+      "step": 3373
+    },
+    {
+      "epoch": 2.5536423841059603,
+      "grad_norm": 2.0099551677703857,
+      "learning_rate": 1.0993091461082154e-06,
+      "loss": 1.0111,
+      "step": 3374
+    },
+    {
+      "epoch": 2.5543992431409652,
+      "grad_norm": 2.1071789264678955,
+      "learning_rate": 1.0956557024725986e-06,
+      "loss": 1.018,
+      "step": 3375
+    },
+    {
+      "epoch": 2.5551561021759697,
+      "grad_norm": 2.2769486904144287,
+      "learning_rate": 1.0920079990861043e-06,
+      "loss": 1.0538,
+      "step": 3376
+    },
+    {
+      "epoch": 2.5559129612109746,
+      "grad_norm": 1.969045877456665,
+      "learning_rate": 1.088366038299989e-06,
+      "loss": 1.0186,
+      "step": 3377
+    },
+    {
+      "epoch": 2.556669820245979,
+      "grad_norm": 2.0804572105407715,
+      "learning_rate": 1.0847298224618053e-06,
+      "loss": 1.0581,
+      "step": 3378
+    },
+    {
+      "epoch": 2.557426679280984,
+      "grad_norm": 1.9232372045516968,
+      "learning_rate": 1.081099353915403e-06,
+      "loss": 1.0507,
+      "step": 3379
+    },
+    {
+      "epoch": 2.5581835383159888,
+      "grad_norm": 1.9603554010391235,
+      "learning_rate": 1.077474635000925e-06,
+      "loss": 1.0878,
+      "step": 3380
+    },
+    {
+      "epoch": 2.558940397350993,
+      "grad_norm": 2.1331593990325928,
+      "learning_rate": 1.07385566805481e-06,
+      "loss": 1.0751,
+      "step": 3381
+    },
+    {
+      "epoch": 2.559697256385998,
+      "grad_norm": 2.48708438873291,
+      "learning_rate": 1.070242455409791e-06,
+      "loss": 1.0929,
+      "step": 3382
+    },
+    {
+      "epoch": 2.560454115421003,
+      "grad_norm": 2.023110866546631,
+      "learning_rate": 1.066634999394886e-06,
+      "loss": 1.0477,
+      "step": 3383
+    },
+    {
+      "epoch": 2.5612109744560074,
+      "grad_norm": 2.201087236404419,
+      "learning_rate": 1.0630333023354118e-06,
+      "loss": 1.0736,
+      "step": 3384
+    },
+    {
+      "epoch": 2.5619678334910123,
+      "grad_norm": 2.0009500980377197,
+      "learning_rate": 1.0594373665529613e-06,
+      "loss": 1.0059,
+      "step": 3385
+    },
+    {
+      "epoch": 2.562724692526017,
+      "grad_norm": 1.9756640195846558,
+      "learning_rate": 1.0558471943654217e-06,
+      "loss": 1.0857,
+      "step": 3386
+    },
+    {
+      "epoch": 2.5634815515610216,
+      "grad_norm": 2.2636256217956543,
+      "learning_rate": 1.0522627880869646e-06,
+      "loss": 1.0848,
+      "step": 3387
+    },
+    {
+      "epoch": 2.5642384105960265,
+      "grad_norm": 2.0621442794799805,
+      "learning_rate": 1.0486841500280441e-06,
+      "loss": 1.0821,
+      "step": 3388
+    },
+    {
+      "epoch": 2.564995269631031,
+      "grad_norm": 2.0804009437561035,
+      "learning_rate": 1.0451112824953961e-06,
+      "loss": 1.072,
+      "step": 3389
+    },
+    {
+      "epoch": 2.565752128666036,
+      "grad_norm": 2.1010797023773193,
+      "learning_rate": 1.0415441877920349e-06,
+      "loss": 1.0384,
+      "step": 3390
+    },
+    {
+      "epoch": 2.5665089877010407,
+      "grad_norm": 2.0294950008392334,
+      "learning_rate": 1.037982868217254e-06,
+      "loss": 1.1007,
+      "step": 3391
+    },
+    {
+      "epoch": 2.5672658467360456,
+      "grad_norm": 2.2194080352783203,
+      "learning_rate": 1.0344273260666264e-06,
+      "loss": 1.0293,
+      "step": 3392
+    },
+    {
+      "epoch": 2.56802270577105,
+      "grad_norm": 2.0473580360412598,
+      "learning_rate": 1.0308775636320018e-06,
+      "loss": 1.0989,
+      "step": 3393
+    },
+    {
+      "epoch": 2.568779564806055,
+      "grad_norm": 2.1180901527404785,
+      "learning_rate": 1.027333583201503e-06,
+      "loss": 1.073,
+      "step": 3394
+    },
+    {
+      "epoch": 2.5695364238410594,
+      "grad_norm": 1.9611269235610962,
+      "learning_rate": 1.0237953870595262e-06,
+      "loss": 1.0638,
+      "step": 3395
+    },
+    {
+      "epoch": 2.5702932828760643,
+      "grad_norm": 2.031759023666382,
+      "learning_rate": 1.0202629774867378e-06,
+      "loss": 1.044,
+      "step": 3396
+    },
+    {
+      "epoch": 2.571050141911069,
+      "grad_norm": 2.155648708343506,
+      "learning_rate": 1.016736356760073e-06,
+      "loss": 1.0815,
+      "step": 3397
+    },
+    {
+      "epoch": 2.571807000946074,
+      "grad_norm": 2.0659499168395996,
+      "learning_rate": 1.0132155271527401e-06,
+      "loss": 1.0977,
+      "step": 3398
+    },
+    {
+      "epoch": 2.5725638599810785,
+      "grad_norm": 2.2170495986938477,
+      "learning_rate": 1.0097004909342112e-06,
+      "loss": 1.0449,
+      "step": 3399
+    },
+    {
+      "epoch": 2.5733207190160834,
+      "grad_norm": 1.9732736349105835,
+      "learning_rate": 1.0061912503702258e-06,
+      "loss": 1.0475,
+      "step": 3400
+    },
+    {
+      "epoch": 2.574077578051088,
+      "grad_norm": 1.9781739711761475,
+      "learning_rate": 1.0026878077227885e-06,
+      "loss": 1.0778,
+      "step": 3401
+    },
+    {
+      "epoch": 2.5748344370860927,
+      "grad_norm": 2.0298542976379395,
+      "learning_rate": 9.99190165250161e-07,
+      "loss": 1.0379,
+      "step": 3402
+    },
+    {
+      "epoch": 2.5755912961210976,
+      "grad_norm": 1.7894214391708374,
+      "learning_rate": 9.95698325206874e-07,
+      "loss": 1.069,
+      "step": 3403
+    },
+    {
+      "epoch": 2.576348155156102,
+      "grad_norm": 2.022477865219116,
+      "learning_rate": 9.922122898437122e-07,
+      "loss": 1.0623,
+      "step": 3404
+    },
+    {
+      "epoch": 2.577105014191107,
+      "grad_norm": 1.8968234062194824,
+      "learning_rate": 9.887320614077198e-07,
+      "loss": 1.04,
+      "step": 3405
+    },
+    {
+      "epoch": 2.5778618732261114,
+      "grad_norm": 2.217832326889038,
+      "learning_rate": 9.852576421422033e-07,
+      "loss": 1.0943,
+      "step": 3406
+    },
+    {
+      "epoch": 2.5786187322611163,
+      "grad_norm": 2.0771234035491943,
+      "learning_rate": 9.817890342867157e-07,
+      "loss": 1.0767,
+      "step": 3407
+    },
+    {
+      "epoch": 2.579375591296121,
+      "grad_norm": 1.9438964128494263,
+      "learning_rate": 9.783262400770708e-07,
+      "loss": 1.1243,
+      "step": 3408
+    },
+    {
+      "epoch": 2.580132450331126,
+      "grad_norm": 2.441040277481079,
+      "learning_rate": 9.748692617453326e-07,
+      "loss": 1.0624,
+      "step": 3409
+    },
+    {
+      "epoch": 2.5808893093661305,
+      "grad_norm": 2.0702598094940186,
+      "learning_rate": 9.714181015198182e-07,
+      "loss": 1.0696,
+      "step": 3410
+    },
+    {
+      "epoch": 2.5816461684011354,
+      "grad_norm": 2.0918853282928467,
+      "learning_rate": 9.67972761625091e-07,
+      "loss": 1.0814,
+      "step": 3411
+    },
+    {
+      "epoch": 2.58240302743614,
+      "grad_norm": 1.9000964164733887,
+      "learning_rate": 9.645332442819653e-07,
+      "loss": 1.0554,
+      "step": 3412
+    },
+    {
+      "epoch": 2.5831598864711447,
+      "grad_norm": 2.1118955612182617,
+      "learning_rate": 9.610995517075005e-07,
+      "loss": 1.0622,
+      "step": 3413
+    },
+    {
+      "epoch": 2.5839167455061496,
+      "grad_norm": 2.136005163192749,
+      "learning_rate": 9.57671686115003e-07,
+      "loss": 1.0871,
+      "step": 3414
+    },
+    {
+      "epoch": 2.5846736045411545,
+      "grad_norm": 2.0861973762512207,
+      "learning_rate": 9.542496497140228e-07,
+      "loss": 1.0348,
+      "step": 3415
+    },
+    {
+      "epoch": 2.585430463576159,
+      "grad_norm": 1.9754106998443604,
+      "learning_rate": 9.50833444710354e-07,
+      "loss": 1.0797,
+      "step": 3416
+    },
+    {
+      "epoch": 2.586187322611164,
+      "grad_norm": 2.138561964035034,
+      "learning_rate": 9.474230733060293e-07,
+      "loss": 1.1018,
+      "step": 3417
+    },
+    {
+      "epoch": 2.5869441816461682,
+      "grad_norm": 2.1578221321105957,
+      "learning_rate": 9.440185376993193e-07,
+      "loss": 1.1082,
+      "step": 3418
+    },
+    {
+      "epoch": 2.587701040681173,
+      "grad_norm": 1.928044080734253,
+      "learning_rate": 9.406198400847376e-07,
+      "loss": 1.0723,
+      "step": 3419
+    },
+    {
+      "epoch": 2.588457899716178,
+      "grad_norm": 2.0299084186553955,
+      "learning_rate": 9.372269826530338e-07,
+      "loss": 1.0557,
+      "step": 3420
+    },
+    {
+      "epoch": 2.5892147587511825,
+      "grad_norm": 2.100691556930542,
+      "learning_rate": 9.338399675911917e-07,
+      "loss": 1.0221,
+      "step": 3421
+    },
+    {
+      "epoch": 2.5899716177861873,
+      "grad_norm": 2.015913724899292,
+      "learning_rate": 9.304587970824288e-07,
+      "loss": 1.0651,
+      "step": 3422
+    },
+    {
+      "epoch": 2.590728476821192,
+      "grad_norm": 1.8734519481658936,
+      "learning_rate": 9.270834733061999e-07,
+      "loss": 1.0554,
+      "step": 3423
+    },
+    {
+      "epoch": 2.5914853358561967,
+      "grad_norm": 1.9088720083236694,
+      "learning_rate": 9.237139984381806e-07,
+      "loss": 1.0847,
+      "step": 3424
+    },
+    {
+      "epoch": 2.5922421948912016,
+      "grad_norm": 2.083169460296631,
+      "learning_rate": 9.203503746502859e-07,
+      "loss": 1.0714,
+      "step": 3425
+    },
+    {
+      "epoch": 2.5929990539262064,
+      "grad_norm": 1.9658424854278564,
+      "learning_rate": 9.169926041106579e-07,
+      "loss": 1.0231,
+      "step": 3426
+    },
+    {
+      "epoch": 2.593755912961211,
+      "grad_norm": 1.955154299736023,
+      "learning_rate": 9.13640688983662e-07,
+      "loss": 1.0539,
+      "step": 3427
+    },
+    {
+      "epoch": 2.5945127719962158,
+      "grad_norm": 2.0878820419311523,
+      "learning_rate": 9.102946314298959e-07,
+      "loss": 1.0911,
+      "step": 3428
+    },
+    {
+      "epoch": 2.59526963103122,
+      "grad_norm": 1.9909857511520386,
+      "learning_rate": 9.069544336061716e-07,
+      "loss": 1.083,
+      "step": 3429
+    },
+    {
+      "epoch": 2.596026490066225,
+      "grad_norm": 2.0630910396575928,
+      "learning_rate": 9.036200976655337e-07,
+      "loss": 1.0977,
+      "step": 3430
+    },
+    {
+      "epoch": 2.59678334910123,
+      "grad_norm": 1.982391595840454,
+      "learning_rate": 9.002916257572411e-07,
+      "loss": 1.0271,
+      "step": 3431
+    },
+    {
+      "epoch": 2.597540208136235,
+      "grad_norm": 1.9987069368362427,
+      "learning_rate": 8.969690200267786e-07,
+      "loss": 1.0813,
+      "step": 3432
+    },
+    {
+      "epoch": 2.5982970671712393,
+      "grad_norm": 1.983818531036377,
+      "learning_rate": 8.936522826158452e-07,
+      "loss": 1.0776,
+      "step": 3433
+    },
+    {
+      "epoch": 2.599053926206244,
+      "grad_norm": 1.9349209070205688,
+      "learning_rate": 8.903414156623622e-07,
+      "loss": 1.0509,
+      "step": 3434
+    },
+    {
+      "epoch": 2.5998107852412486,
+      "grad_norm": 2.01790714263916,
+      "learning_rate": 8.870364213004612e-07,
+      "loss": 1.0949,
+      "step": 3435
+    },
+    {
+      "epoch": 2.6005676442762535,
+      "grad_norm": 2.0256693363189697,
+      "learning_rate": 8.837373016604916e-07,
+      "loss": 1.0572,
+      "step": 3436
+    },
+    {
+      "epoch": 2.6013245033112584,
+      "grad_norm": 1.989288568496704,
+      "learning_rate": 8.804440588690183e-07,
+      "loss": 1.0321,
+      "step": 3437
+    },
+    {
+      "epoch": 2.602081362346263,
+      "grad_norm": 2.1254732608795166,
+      "learning_rate": 8.771566950488107e-07,
+      "loss": 1.0513,
+      "step": 3438
+    },
+    {
+      "epoch": 2.6028382213812677,
+      "grad_norm": 2.4187963008880615,
+      "learning_rate": 8.738752123188587e-07,
+      "loss": 1.0755,
+      "step": 3439
+    },
+    {
+      "epoch": 2.6035950804162726,
+      "grad_norm": 2.0207037925720215,
+      "learning_rate": 8.705996127943503e-07,
+      "loss": 1.0669,
+      "step": 3440
+    },
+    {
+      "epoch": 2.604351939451277,
+      "grad_norm": 2.1482834815979004,
+      "learning_rate": 8.6732989858669e-07,
+      "loss": 1.0675,
+      "step": 3441
+    },
+    {
+      "epoch": 2.605108798486282,
+      "grad_norm": 1.9809141159057617,
+      "learning_rate": 8.640660718034855e-07,
+      "loss": 1.1257,
+      "step": 3442
+    },
+    {
+      "epoch": 2.605865657521287,
+      "grad_norm": 2.1395535469055176,
+      "learning_rate": 8.608081345485507e-07,
+      "loss": 1.1311,
+      "step": 3443
+    },
+    {
+      "epoch": 2.6066225165562913,
+      "grad_norm": 2.1757044792175293,
+      "learning_rate": 8.575560889219027e-07,
+      "loss": 1.0888,
+      "step": 3444
+    },
+    {
+      "epoch": 2.607379375591296,
+      "grad_norm": 1.8578970432281494,
+      "learning_rate": 8.543099370197591e-07,
+      "loss": 1.0527,
+      "step": 3445
+    },
+    {
+      "epoch": 2.6081362346263006,
+      "grad_norm": 1.966065526008606,
+      "learning_rate": 8.51069680934539e-07,
+      "loss": 1.0301,
+      "step": 3446
+    },
+    {
+      "epoch": 2.6088930936613055,
+      "grad_norm": 1.9428819417953491,
+      "learning_rate": 8.478353227548625e-07,
+      "loss": 1.0255,
+      "step": 3447
+    },
+    {
+      "epoch": 2.6096499526963104,
+      "grad_norm": 1.873252272605896,
+      "learning_rate": 8.446068645655477e-07,
+      "loss": 1.092,
+      "step": 3448
+    },
+    {
+      "epoch": 2.6104068117313153,
+      "grad_norm": 2.0765583515167236,
+      "learning_rate": 8.413843084476109e-07,
+      "loss": 1.0985,
+      "step": 3449
+    },
+    {
+      "epoch": 2.6111636707663197,
+      "grad_norm": 2.0844457149505615,
+      "learning_rate": 8.381676564782655e-07,
+      "loss": 1.1138,
+      "step": 3450
+    },
+    {
+      "epoch": 2.6119205298013246,
+      "grad_norm": 1.9434021711349487,
+      "learning_rate": 8.349569107309078e-07,
+      "loss": 1.059,
+      "step": 3451
+    },
+    {
+      "epoch": 2.612677388836329,
+      "grad_norm": 2.0002236366271973,
+      "learning_rate": 8.317520732751409e-07,
+      "loss": 1.0701,
+      "step": 3452
+    },
+    {
+      "epoch": 2.613434247871334,
+      "grad_norm": 2.0566999912261963,
+      "learning_rate": 8.285531461767541e-07,
+      "loss": 1.066,
+      "step": 3453
+    },
+    {
+      "epoch": 2.614191106906339,
+      "grad_norm": 1.926048994064331,
+      "learning_rate": 8.253601314977264e-07,
+      "loss": 1.0811,
+      "step": 3454
+    },
+    {
+      "epoch": 2.6149479659413433,
+      "grad_norm": 2.086893320083618,
+      "learning_rate": 8.22173031296225e-07,
+      "loss": 1.0647,
+      "step": 3455
+    },
+    {
+      "epoch": 2.615704824976348,
+      "grad_norm": 2.1144938468933105,
+      "learning_rate": 8.189918476266104e-07,
+      "loss": 1.0737,
+      "step": 3456
+    },
+    {
+      "epoch": 2.616461684011353,
+      "grad_norm": 1.8913697004318237,
+      "learning_rate": 8.158165825394222e-07,
+      "loss": 1.0621,
+      "step": 3457
+    },
+    {
+      "epoch": 2.6172185430463575,
+      "grad_norm": 1.850129246711731,
+      "learning_rate": 8.126472380813851e-07,
+      "loss": 1.0475,
+      "step": 3458
+    },
+    {
+      "epoch": 2.6179754020813624,
+      "grad_norm": 1.9463554620742798,
+      "learning_rate": 8.094838162954142e-07,
+      "loss": 1.0652,
+      "step": 3459
+    },
+    {
+      "epoch": 2.6187322611163673,
+      "grad_norm": 1.9536323547363281,
+      "learning_rate": 8.063263192206013e-07,
+      "loss": 1.0567,
+      "step": 3460
+    },
+    {
+      "epoch": 2.6194891201513717,
+      "grad_norm": 1.9918063879013062,
+      "learning_rate": 8.031747488922231e-07,
+      "loss": 1.0604,
+      "step": 3461
+    },
+    {
+      "epoch": 2.6202459791863766,
+      "grad_norm": 2.074878215789795,
+      "learning_rate": 8.0002910734173e-07,
+      "loss": 1.0467,
+      "step": 3462
+    },
+    {
+      "epoch": 2.621002838221381,
+      "grad_norm": 2.0185697078704834,
+      "learning_rate": 7.968893965967558e-07,
+      "loss": 1.1014,
+      "step": 3463
+    },
+    {
+      "epoch": 2.621759697256386,
+      "grad_norm": 1.9658541679382324,
+      "learning_rate": 7.937556186811127e-07,
+      "loss": 1.0582,
+      "step": 3464
+    },
+    {
+      "epoch": 2.622516556291391,
+      "grad_norm": 2.0424704551696777,
+      "learning_rate": 7.906277756147835e-07,
+      "loss": 1.0843,
+      "step": 3465
+    },
+    {
+      "epoch": 2.6232734153263957,
+      "grad_norm": 2.1158058643341064,
+      "learning_rate": 7.875058694139282e-07,
+      "loss": 1.0359,
+      "step": 3466
+    },
+    {
+      "epoch": 2.6240302743614,
+      "grad_norm": 2.1120176315307617,
+      "learning_rate": 7.843899020908823e-07,
+      "loss": 1.0796,
+      "step": 3467
+    },
+    {
+      "epoch": 2.624787133396405,
+      "grad_norm": 1.910689353942871,
+      "learning_rate": 7.812798756541483e-07,
+      "loss": 1.0547,
+      "step": 3468
+    },
+    {
+      "epoch": 2.6255439924314095,
+      "grad_norm": 2.1395435333251953,
+      "learning_rate": 7.781757921084019e-07,
+      "loss": 1.0968,
+      "step": 3469
+    },
+    {
+      "epoch": 2.6263008514664143,
+      "grad_norm": 1.9301244020462036,
+      "learning_rate": 7.750776534544889e-07,
+      "loss": 1.0543,
+      "step": 3470
+    },
+    {
+      "epoch": 2.6270577105014192,
+      "grad_norm": 1.9667290449142456,
+      "learning_rate": 7.719854616894243e-07,
+      "loss": 1.0575,
+      "step": 3471
+    },
+    {
+      "epoch": 2.627814569536424,
+      "grad_norm": 1.8904736042022705,
+      "learning_rate": 7.688992188063853e-07,
+      "loss": 1.0781,
+      "step": 3472
+    },
+    {
+      "epoch": 2.6285714285714286,
+      "grad_norm": 1.991716980934143,
+      "learning_rate": 7.658189267947159e-07,
+      "loss": 1.052,
+      "step": 3473
+    },
+    {
+      "epoch": 2.6293282876064334,
+      "grad_norm": 2.0166666507720947,
+      "learning_rate": 7.627445876399259e-07,
+      "loss": 1.0668,
+      "step": 3474
+    },
+    {
+      "epoch": 2.630085146641438,
+      "grad_norm": 2.020129919052124,
+      "learning_rate": 7.596762033236895e-07,
+      "loss": 1.0539,
+      "step": 3475
+    },
+    {
+      "epoch": 2.6308420056764428,
+      "grad_norm": 2.1295621395111084,
+      "learning_rate": 7.566137758238386e-07,
+      "loss": 1.0264,
+      "step": 3476
+    },
+    {
+      "epoch": 2.6315988647114477,
+      "grad_norm": 2.1414082050323486,
+      "learning_rate": 7.53557307114367e-07,
+      "loss": 1.0476,
+      "step": 3477
+    },
+    {
+      "epoch": 2.632355723746452,
+      "grad_norm": 1.8784303665161133,
+      "learning_rate": 7.505067991654335e-07,
+      "loss": 1.0252,
+      "step": 3478
+    },
+    {
+      "epoch": 2.633112582781457,
+      "grad_norm": 2.0736515522003174,
+      "learning_rate": 7.474622539433398e-07,
+      "loss": 1.0661,
+      "step": 3479
+    },
+    {
+      "epoch": 2.6338694418164614,
+      "grad_norm": 1.9620708227157593,
+      "learning_rate": 7.444236734105581e-07,
+      "loss": 1.0835,
+      "step": 3480
+    },
+    {
+      "epoch": 2.6346263008514663,
+      "grad_norm": 2.1406285762786865,
+      "learning_rate": 7.413910595257105e-07,
+      "loss": 1.0577,
+      "step": 3481
+    },
+    {
+      "epoch": 2.635383159886471,
+      "grad_norm": 1.9883054494857788,
+      "learning_rate": 7.383644142435741e-07,
+      "loss": 1.0546,
+      "step": 3482
+    },
+    {
+      "epoch": 2.636140018921476,
+      "grad_norm": 1.9472057819366455,
+      "learning_rate": 7.353437395150799e-07,
+      "loss": 1.0452,
+      "step": 3483
+    },
+    {
+      "epoch": 2.6368968779564805,
+      "grad_norm": 2.0464439392089844,
+      "learning_rate": 7.323290372873055e-07,
+      "loss": 1.1029,
+      "step": 3484
+    },
+    {
+      "epoch": 2.6376537369914854,
+      "grad_norm": 1.9990071058273315,
+      "learning_rate": 7.293203095034839e-07,
+      "loss": 1.1126,
+      "step": 3485
+    },
+    {
+      "epoch": 2.63841059602649,
+      "grad_norm": 2.022820234298706,
+      "learning_rate": 7.263175581029933e-07,
+      "loss": 1.0625,
+      "step": 3486
+    },
+    {
+      "epoch": 2.6391674550614947,
+      "grad_norm": 1.835789442062378,
+      "learning_rate": 7.233207850213639e-07,
+      "loss": 1.0732,
+      "step": 3487
+    },
+    {
+      "epoch": 2.6399243140964996,
+      "grad_norm": 1.9170242547988892,
+      "learning_rate": 7.20329992190268e-07,
+      "loss": 1.1162,
+      "step": 3488
+    },
+    {
+      "epoch": 2.6406811731315045,
+      "grad_norm": 1.9878673553466797,
+      "learning_rate": 7.173451815375276e-07,
+      "loss": 1.0664,
+      "step": 3489
+    },
+    {
+      "epoch": 2.641438032166509,
+      "grad_norm": 2.3943591117858887,
+      "learning_rate": 7.14366354987102e-07,
+      "loss": 1.1248,
+      "step": 3490
+    },
+    {
+      "epoch": 2.642194891201514,
+      "grad_norm": 1.926537275314331,
+      "learning_rate": 7.113935144591011e-07,
+      "loss": 1.0654,
+      "step": 3491
+    },
+    {
+      "epoch": 2.6429517502365183,
+      "grad_norm": 2.132347583770752,
+      "learning_rate": 7.084266618697722e-07,
+      "loss": 1.058,
+      "step": 3492
+    },
+    {
+      "epoch": 2.643708609271523,
+      "grad_norm": 2.0414459705352783,
+      "learning_rate": 7.054657991315009e-07,
+      "loss": 1.0699,
+      "step": 3493
+    },
+    {
+      "epoch": 2.644465468306528,
+      "grad_norm": 2.3885207176208496,
+      "learning_rate": 7.025109281528162e-07,
+      "loss": 1.0914,
+      "step": 3494
+    },
+    {
+      "epoch": 2.6452223273415325,
+      "grad_norm": 2.1995675563812256,
+      "learning_rate": 6.995620508383816e-07,
+      "loss": 1.0691,
+      "step": 3495
+    },
+    {
+      "epoch": 2.6459791863765374,
+      "grad_norm": 1.9995527267456055,
+      "learning_rate": 6.966191690889987e-07,
+      "loss": 1.0715,
+      "step": 3496
+    },
+    {
+      "epoch": 2.646736045411542,
+      "grad_norm": 2.06939435005188,
+      "learning_rate": 6.936822848016048e-07,
+      "loss": 1.0522,
+      "step": 3497
+    },
+    {
+      "epoch": 2.6474929044465467,
+      "grad_norm": 2.001063346862793,
+      "learning_rate": 6.907513998692701e-07,
+      "loss": 1.0726,
+      "step": 3498
+    },
+    {
+      "epoch": 2.6482497634815516,
+      "grad_norm": 2.1571309566497803,
+      "learning_rate": 6.878265161812005e-07,
+      "loss": 1.0609,
+      "step": 3499
+    },
+    {
+      "epoch": 2.6490066225165565,
+      "grad_norm": 2.0034592151641846,
+      "learning_rate": 6.849076356227285e-07,
+      "loss": 1.0636,
+      "step": 3500
+    },
+    {
+      "epoch": 2.649763481551561,
+      "grad_norm": 1.8944875001907349,
+      "learning_rate": 6.819947600753214e-07,
+      "loss": 1.061,
+      "step": 3501
+    },
+    {
+      "epoch": 2.650520340586566,
+      "grad_norm": 1.9522629976272583,
+      "learning_rate": 6.790878914165723e-07,
+      "loss": 1.1064,
+      "step": 3502
+    },
+    {
+      "epoch": 2.6512771996215703,
+      "grad_norm": 1.9700335264205933,
+      "learning_rate": 6.761870315202072e-07,
+      "loss": 1.0509,
+      "step": 3503
+    },
+    {
+      "epoch": 2.652034058656575,
+      "grad_norm": 1.9173399209976196,
+      "learning_rate": 6.732921822560753e-07,
+      "loss": 1.0467,
+      "step": 3504
+    },
+    {
+      "epoch": 2.65279091769158,
+      "grad_norm": 2.1325631141662598,
+      "learning_rate": 6.704033454901552e-07,
+      "loss": 1.0547,
+      "step": 3505
+    },
+    {
+      "epoch": 2.653547776726585,
+      "grad_norm": 1.9540364742279053,
+      "learning_rate": 6.67520523084541e-07,
+      "loss": 1.1084,
+      "step": 3506
+    },
+    {
+      "epoch": 2.6543046357615894,
+      "grad_norm": 2.026878595352173,
+      "learning_rate": 6.646437168974577e-07,
+      "loss": 1.0511,
+      "step": 3507
+    },
+    {
+      "epoch": 2.6550614947965943,
+      "grad_norm": 2.238311290740967,
+      "learning_rate": 6.617729287832535e-07,
+      "loss": 1.1053,
+      "step": 3508
+    },
+    {
+      "epoch": 2.6558183538315987,
+      "grad_norm": 2.0573630332946777,
+      "learning_rate": 6.589081605923916e-07,
+      "loss": 1.0377,
+      "step": 3509
+    },
+    {
+      "epoch": 2.6565752128666036,
+      "grad_norm": 2.0835254192352295,
+      "learning_rate": 6.56049414171461e-07,
+      "loss": 1.0403,
+      "step": 3510
+    },
+    {
+      "epoch": 2.6573320719016085,
+      "grad_norm": 1.888987421989441,
+      "learning_rate": 6.531966913631649e-07,
+      "loss": 1.0874,
+      "step": 3511
+    },
+    {
+      "epoch": 2.658088930936613,
+      "grad_norm": 1.9704266786575317,
+      "learning_rate": 6.503499940063245e-07,
+      "loss": 1.0968,
+      "step": 3512
+    },
+    {
+      "epoch": 2.658845789971618,
+      "grad_norm": 2.062167167663574,
+      "learning_rate": 6.475093239358764e-07,
+      "loss": 1.074,
+      "step": 3513
+    },
+    {
+      "epoch": 2.6596026490066222,
+      "grad_norm": 2.041229724884033,
+      "learning_rate": 6.446746829828747e-07,
+      "loss": 1.0881,
+      "step": 3514
+    },
+    {
+      "epoch": 2.660359508041627,
+      "grad_norm": 2.0432045459747314,
+      "learning_rate": 6.41846072974484e-07,
+      "loss": 1.0497,
+      "step": 3515
+    },
+    {
+      "epoch": 2.661116367076632,
+      "grad_norm": 2.037137746810913,
+      "learning_rate": 6.390234957339877e-07,
+      "loss": 1.1183,
+      "step": 3516
+    },
+    {
+      "epoch": 2.661873226111637,
+      "grad_norm": 2.0373597145080566,
+      "learning_rate": 6.362069530807692e-07,
+      "loss": 1.0721,
+      "step": 3517
+    },
+    {
+      "epoch": 2.6626300851466413,
+      "grad_norm": 2.2422561645507812,
+      "learning_rate": 6.333964468303339e-07,
+      "loss": 1.1072,
+      "step": 3518
+    },
+    {
+      "epoch": 2.6633869441816462,
+      "grad_norm": 2.035428285598755,
+      "learning_rate": 6.305919787942921e-07,
+      "loss": 1.0755,
+      "step": 3519
+    },
+    {
+      "epoch": 2.6641438032166507,
+      "grad_norm": 2.0684683322906494,
+      "learning_rate": 6.277935507803559e-07,
+      "loss": 1.0673,
+      "step": 3520
+    },
+    {
+      "epoch": 2.6649006622516556,
+      "grad_norm": 2.1325268745422363,
+      "learning_rate": 6.25001164592354e-07,
+      "loss": 1.0269,
+      "step": 3521
+    },
+    {
+      "epoch": 2.6656575212866604,
+      "grad_norm": 1.9109140634536743,
+      "learning_rate": 6.222148220302141e-07,
+      "loss": 1.0445,
+      "step": 3522
+    },
+    {
+      "epoch": 2.6664143803216653,
+      "grad_norm": 2.0545241832733154,
+      "learning_rate": 6.1943452488997e-07,
+      "loss": 1.03,
+      "step": 3523
+    },
+    {
+      "epoch": 2.6671712393566698,
+      "grad_norm": 1.9249234199523926,
+      "learning_rate": 6.166602749637587e-07,
+      "loss": 1.0247,
+      "step": 3524
+    },
+    {
+      "epoch": 2.6679280983916747,
+      "grad_norm": 2.097187042236328,
+      "learning_rate": 6.138920740398207e-07,
+      "loss": 1.0982,
+      "step": 3525
+    },
+    {
+      "epoch": 2.668684957426679,
+      "grad_norm": 2.3547725677490234,
+      "learning_rate": 6.111299239024957e-07,
+      "loss": 1.0561,
+      "step": 3526
+    },
+    {
+      "epoch": 2.669441816461684,
+      "grad_norm": 1.9423364400863647,
+      "learning_rate": 6.083738263322244e-07,
+      "loss": 1.0832,
+      "step": 3527
+    },
+    {
+      "epoch": 2.670198675496689,
+      "grad_norm": 1.9852162599563599,
+      "learning_rate": 6.056237831055416e-07,
+      "loss": 1.0746,
+      "step": 3528
+    },
+    {
+      "epoch": 2.6709555345316933,
+      "grad_norm": 2.095628261566162,
+      "learning_rate": 6.02879795995085e-07,
+      "loss": 1.0651,
+      "step": 3529
+    },
+    {
+      "epoch": 2.671712393566698,
+      "grad_norm": 2.109067916870117,
+      "learning_rate": 6.001418667695884e-07,
+      "loss": 1.0847,
+      "step": 3530
+    },
+    {
+      "epoch": 2.672469252601703,
+      "grad_norm": 1.9949407577514648,
+      "learning_rate": 5.97409997193879e-07,
+      "loss": 1.1257,
+      "step": 3531
+    },
+    {
+      "epoch": 2.6732261116367075,
+      "grad_norm": 2.216343641281128,
+      "learning_rate": 5.946841890288763e-07,
+      "loss": 1.0922,
+      "step": 3532
+    },
+    {
+      "epoch": 2.6739829706717124,
+      "grad_norm": 1.9918749332427979,
+      "learning_rate": 5.91964444031599e-07,
+      "loss": 1.0832,
+      "step": 3533
+    },
+    {
+      "epoch": 2.6747398297067173,
+      "grad_norm": 2.0035977363586426,
+      "learning_rate": 5.892507639551483e-07,
+      "loss": 1.1023,
+      "step": 3534
+    },
+    {
+      "epoch": 2.6754966887417218,
+      "grad_norm": 1.9594407081604004,
+      "learning_rate": 5.86543150548722e-07,
+      "loss": 1.0479,
+      "step": 3535
+    },
+    {
+      "epoch": 2.6762535477767266,
+      "grad_norm": 1.9996135234832764,
+      "learning_rate": 5.838416055576072e-07,
+      "loss": 1.0837,
+      "step": 3536
+    },
+    {
+      "epoch": 2.677010406811731,
+      "grad_norm": 2.032686710357666,
+      "learning_rate": 5.811461307231798e-07,
+      "loss": 1.0673,
+      "step": 3537
+    },
+    {
+      "epoch": 2.677767265846736,
+      "grad_norm": 1.7957963943481445,
+      "learning_rate": 5.784567277829007e-07,
+      "loss": 1.0672,
+      "step": 3538
+    },
+    {
+      "epoch": 2.678524124881741,
+      "grad_norm": 1.936874508857727,
+      "learning_rate": 5.757733984703174e-07,
+      "loss": 1.1329,
+      "step": 3539
+    },
+    {
+      "epoch": 2.6792809839167457,
+      "grad_norm": 2.072567939758301,
+      "learning_rate": 5.730961445150644e-07,
+      "loss": 1.1066,
+      "step": 3540
+    },
+    {
+      "epoch": 2.68003784295175,
+      "grad_norm": 1.8656892776489258,
+      "learning_rate": 5.704249676428575e-07,
+      "loss": 1.1158,
+      "step": 3541
+    },
+    {
+      "epoch": 2.680794701986755,
+      "grad_norm": 1.9235533475875854,
+      "learning_rate": 5.677598695754967e-07,
+      "loss": 1.0364,
+      "step": 3542
+    },
+    {
+      "epoch": 2.6815515610217595,
+      "grad_norm": 2.159919261932373,
+      "learning_rate": 5.651008520308641e-07,
+      "loss": 1.0813,
+      "step": 3543
+    },
+    {
+      "epoch": 2.6823084200567644,
+      "grad_norm": 2.093416929244995,
+      "learning_rate": 5.624479167229225e-07,
+      "loss": 1.082,
+      "step": 3544
+    },
+    {
+      "epoch": 2.6830652790917693,
+      "grad_norm": 1.9927400350570679,
+      "learning_rate": 5.598010653617116e-07,
+      "loss": 1.0486,
+      "step": 3545
+    },
+    {
+      "epoch": 2.6838221381267737,
+      "grad_norm": 1.986259937286377,
+      "learning_rate": 5.571602996533528e-07,
+      "loss": 1.0532,
+      "step": 3546
+    },
+    {
+      "epoch": 2.6845789971617786,
+      "grad_norm": 2.1325762271881104,
+      "learning_rate": 5.54525621300045e-07,
+      "loss": 1.1321,
+      "step": 3547
+    },
+    {
+      "epoch": 2.6853358561967835,
+      "grad_norm": 1.9752742052078247,
+      "learning_rate": 5.518970320000578e-07,
+      "loss": 1.0752,
+      "step": 3548
+    },
+    {
+      "epoch": 2.686092715231788,
+      "grad_norm": 1.9965808391571045,
+      "learning_rate": 5.492745334477438e-07,
+      "loss": 1.0721,
+      "step": 3549
+    },
+    {
+      "epoch": 2.686849574266793,
+      "grad_norm": 2.021066427230835,
+      "learning_rate": 5.466581273335216e-07,
+      "loss": 1.0819,
+      "step": 3550
+    },
+    {
+      "epoch": 2.6876064333017977,
+      "grad_norm": 2.0489556789398193,
+      "learning_rate": 5.440478153438891e-07,
+      "loss": 1.0542,
+      "step": 3551
+    },
+    {
+      "epoch": 2.688363292336802,
+      "grad_norm": 2.1207127571105957,
+      "learning_rate": 5.414435991614129e-07,
+      "loss": 1.0577,
+      "step": 3552
+    },
+    {
+      "epoch": 2.689120151371807,
+      "grad_norm": 2.099400281906128,
+      "learning_rate": 5.388454804647312e-07,
+      "loss": 1.0527,
+      "step": 3553
+    },
+    {
+      "epoch": 2.6898770104068115,
+      "grad_norm": 2.165239095687866,
+      "learning_rate": 5.362534609285534e-07,
+      "loss": 1.0728,
+      "step": 3554
+    },
+    {
+      "epoch": 2.6906338694418164,
+      "grad_norm": 1.964612364768982,
+      "learning_rate": 5.336675422236547e-07,
+      "loss": 1.1085,
+      "step": 3555
+    },
+    {
+      "epoch": 2.6913907284768213,
+      "grad_norm": 2.3169875144958496,
+      "learning_rate": 5.31087726016876e-07,
+      "loss": 1.0873,
+      "step": 3556
+    },
+    {
+      "epoch": 2.692147587511826,
+      "grad_norm": 2.005558490753174,
+      "learning_rate": 5.285140139711306e-07,
+      "loss": 1.0435,
+      "step": 3557
+    },
+    {
+      "epoch": 2.6929044465468306,
+      "grad_norm": 1.9185731410980225,
+      "learning_rate": 5.259464077453933e-07,
+      "loss": 1.1144,
+      "step": 3558
+    },
+    {
+      "epoch": 2.6936613055818355,
+      "grad_norm": 1.933445930480957,
+      "learning_rate": 5.233849089947034e-07,
+      "loss": 1.0526,
+      "step": 3559
+    },
+    {
+      "epoch": 2.69441816461684,
+      "grad_norm": 2.1504805088043213,
+      "learning_rate": 5.208295193701673e-07,
+      "loss": 1.0822,
+      "step": 3560
+    },
+    {
+      "epoch": 2.695175023651845,
+      "grad_norm": 2.1270816326141357,
+      "learning_rate": 5.182802405189443e-07,
+      "loss": 1.0848,
+      "step": 3561
+    },
+    {
+      "epoch": 2.6959318826868497,
+      "grad_norm": 2.112243890762329,
+      "learning_rate": 5.157370740842649e-07,
+      "loss": 1.0501,
+      "step": 3562
+    },
+    {
+      "epoch": 2.6966887417218546,
+      "grad_norm": 2.2307591438293457,
+      "learning_rate": 5.132000217054134e-07,
+      "loss": 1.1388,
+      "step": 3563
+    },
+    {
+      "epoch": 2.697445600756859,
+      "grad_norm": 2.053459405899048,
+      "learning_rate": 5.106690850177358e-07,
+      "loss": 1.0846,
+      "step": 3564
+    },
+    {
+      "epoch": 2.698202459791864,
+      "grad_norm": 2.0699667930603027,
+      "learning_rate": 5.08144265652635e-07,
+      "loss": 1.0567,
+      "step": 3565
+    },
+    {
+      "epoch": 2.6989593188268683,
+      "grad_norm": 2.0828826427459717,
+      "learning_rate": 5.056255652375729e-07,
+      "loss": 1.0729,
+      "step": 3566
+    },
+    {
+      "epoch": 2.6997161778618732,
+      "grad_norm": 1.9452773332595825,
+      "learning_rate": 5.031129853960639e-07,
+      "loss": 1.0788,
+      "step": 3567
+    },
+    {
+      "epoch": 2.700473036896878,
+      "grad_norm": 2.0344362258911133,
+      "learning_rate": 5.006065277476771e-07,
+      "loss": 1.054,
+      "step": 3568
+    },
+    {
+      "epoch": 2.7012298959318826,
+      "grad_norm": 1.900039792060852,
+      "learning_rate": 4.981061939080384e-07,
+      "loss": 1.0262,
+      "step": 3569
+    },
+    {
+      "epoch": 2.7019867549668874,
+      "grad_norm": 1.9985625743865967,
+      "learning_rate": 4.956119854888261e-07,
+      "loss": 1.0899,
+      "step": 3570
+    },
+    {
+      "epoch": 2.702743614001892,
+      "grad_norm": 2.088229179382324,
+      "learning_rate": 4.931239040977678e-07,
+      "loss": 1.1423,
+      "step": 3571
+    },
+    {
+      "epoch": 2.703500473036897,
+      "grad_norm": 1.9471749067306519,
+      "learning_rate": 4.90641951338641e-07,
+      "loss": 1.0762,
+      "step": 3572
+    },
+    {
+      "epoch": 2.7042573320719017,
+      "grad_norm": 1.8323947191238403,
+      "learning_rate": 4.88166128811277e-07,
+      "loss": 1.0531,
+      "step": 3573
+    },
+    {
+      "epoch": 2.7050141911069066,
+      "grad_norm": 1.9754245281219482,
+      "learning_rate": 4.856964381115542e-07,
+      "loss": 1.1185,
+      "step": 3574
+    },
+    {
+      "epoch": 2.705771050141911,
+      "grad_norm": 1.9460619688034058,
+      "learning_rate": 4.83232880831394e-07,
+      "loss": 1.1218,
+      "step": 3575
+    },
+    {
+      "epoch": 2.706527909176916,
+      "grad_norm": 2.2051377296447754,
+      "learning_rate": 4.807754585587696e-07,
+      "loss": 1.0773,
+      "step": 3576
+    },
+    {
+      "epoch": 2.7072847682119203,
+      "grad_norm": 1.9983853101730347,
+      "learning_rate": 4.783241728776997e-07,
+      "loss": 1.0325,
+      "step": 3577
+    },
+    {
+      "epoch": 2.708041627246925,
+      "grad_norm": 1.9599753618240356,
+      "learning_rate": 4.7587902536824234e-07,
+      "loss": 1.0701,
+      "step": 3578
+    },
+    {
+      "epoch": 2.70879848628193,
+      "grad_norm": 2.0052897930145264,
+      "learning_rate": 4.7344001760650454e-07,
+      "loss": 1.0672,
+      "step": 3579
+    },
+    {
+      "epoch": 2.709555345316935,
+      "grad_norm": 3.11828351020813,
+      "learning_rate": 4.710071511646324e-07,
+      "loss": 1.0932,
+      "step": 3580
+    },
+    {
+      "epoch": 2.7103122043519394,
+      "grad_norm": 2.1355981826782227,
+      "learning_rate": 4.685804276108169e-07,
+      "loss": 1.1196,
+      "step": 3581
+    },
+    {
+      "epoch": 2.7110690633869443,
+      "grad_norm": 2.2099850177764893,
+      "learning_rate": 4.6615984850928456e-07,
+      "loss": 1.1028,
+      "step": 3582
+    },
+    {
+      "epoch": 2.7118259224219488,
+      "grad_norm": 1.9474663734436035,
+      "learning_rate": 4.637454154203033e-07,
+      "loss": 1.07,
+      "step": 3583
+    },
+    {
+      "epoch": 2.7125827814569536,
+      "grad_norm": 2.1069188117980957,
+      "learning_rate": 4.613371299001815e-07,
+      "loss": 1.0899,
+      "step": 3584
+    },
+    {
+      "epoch": 2.7133396404919585,
+      "grad_norm": 2.008517265319824,
+      "learning_rate": 4.58934993501263e-07,
+      "loss": 1.0885,
+      "step": 3585
+    },
+    {
+      "epoch": 2.714096499526963,
+      "grad_norm": 1.88406503200531,
+      "learning_rate": 4.5653900777192763e-07,
+      "loss": 1.0659,
+      "step": 3586
+    },
+    {
+      "epoch": 2.714853358561968,
+      "grad_norm": 2.1920740604400635,
+      "learning_rate": 4.5414917425659094e-07,
+      "loss": 1.1038,
+      "step": 3587
+    },
+    {
+      "epoch": 2.7156102175969723,
+      "grad_norm": 2.047375440597534,
+      "learning_rate": 4.5176549449570765e-07,
+      "loss": 1.0542,
+      "step": 3588
+    },
+    {
+      "epoch": 2.716367076631977,
+      "grad_norm": 1.9768850803375244,
+      "learning_rate": 4.4938797002575485e-07,
+      "loss": 1.0511,
+      "step": 3589
+    },
+    {
+      "epoch": 2.717123935666982,
+      "grad_norm": 2.005725145339966,
+      "learning_rate": 4.4701660237925116e-07,
+      "loss": 1.0859,
+      "step": 3590
+    },
+    {
+      "epoch": 2.717880794701987,
+      "grad_norm": 2.0299482345581055,
+      "learning_rate": 4.446513930847431e-07,
+      "loss": 1.052,
+      "step": 3591
+    },
+    {
+      "epoch": 2.7186376537369914,
+      "grad_norm": 2.217197895050049,
+      "learning_rate": 4.4229234366681054e-07,
+      "loss": 1.083,
+      "step": 3592
+    },
+    {
+      "epoch": 2.7193945127719963,
+      "grad_norm": 2.225231170654297,
+      "learning_rate": 4.399394556460618e-07,
+      "loss": 1.1048,
+      "step": 3593
+    },
+    {
+      "epoch": 2.7201513718070007,
+      "grad_norm": 2.035879135131836,
+      "learning_rate": 4.375927305391286e-07,
+      "loss": 1.0064,
+      "step": 3594
+    },
+    {
+      "epoch": 2.7209082308420056,
+      "grad_norm": 2.046074628829956,
+      "learning_rate": 4.352521698586783e-07,
+      "loss": 1.0649,
+      "step": 3595
+    },
+    {
+      "epoch": 2.7216650898770105,
+      "grad_norm": 2.068490743637085,
+      "learning_rate": 4.329177751133964e-07,
+      "loss": 1.0509,
+      "step": 3596
+    },
+    {
+      "epoch": 2.7224219489120154,
+      "grad_norm": 2.191215991973877,
+      "learning_rate": 4.305895478079998e-07,
+      "loss": 1.0413,
+      "step": 3597
+    },
+    {
+      "epoch": 2.72317880794702,
+      "grad_norm": 2.310241937637329,
+      "learning_rate": 4.2826748944323e-07,
+      "loss": 1.0864,
+      "step": 3598
+    },
+    {
+      "epoch": 2.7239356669820247,
+      "grad_norm": 2.196274757385254,
+      "learning_rate": 4.2595160151584996e-07,
+      "loss": 1.0302,
+      "step": 3599
+    },
+    {
+      "epoch": 2.724692526017029,
+      "grad_norm": 2.0941972732543945,
+      "learning_rate": 4.2364188551864284e-07,
+      "loss": 1.0968,
+      "step": 3600
+    },
+    {
+      "epoch": 2.725449385052034,
+      "grad_norm": 2.1524224281311035,
+      "learning_rate": 4.213383429404197e-07,
+      "loss": 1.0739,
+      "step": 3601
+    },
+    {
+      "epoch": 2.726206244087039,
+      "grad_norm": 2.400557518005371,
+      "learning_rate": 4.190409752660077e-07,
+      "loss": 1.1176,
+      "step": 3602
+    },
+    {
+      "epoch": 2.7269631031220434,
+      "grad_norm": 2.0198590755462646,
+      "learning_rate": 4.16749783976255e-07,
+      "loss": 1.0351,
+      "step": 3603
+    },
+    {
+      "epoch": 2.7277199621570483,
+      "grad_norm": 2.151195526123047,
+      "learning_rate": 4.144647705480291e-07,
+      "loss": 1.0867,
+      "step": 3604
+    },
+    {
+      "epoch": 2.7284768211920527,
+      "grad_norm": 1.927239179611206,
+      "learning_rate": 4.1218593645421344e-07,
+      "loss": 1.0605,
+      "step": 3605
+    },
+    {
+      "epoch": 2.7292336802270576,
+      "grad_norm": 2.175260066986084,
+      "learning_rate": 4.099132831637103e-07,
+      "loss": 1.0312,
+      "step": 3606
+    },
+    {
+      "epoch": 2.7299905392620625,
+      "grad_norm": 2.2161762714385986,
+      "learning_rate": 4.0764681214143794e-07,
+      "loss": 1.0217,
+      "step": 3607
+    },
+    {
+      "epoch": 2.7307473982970674,
+      "grad_norm": 2.06466007232666,
+      "learning_rate": 4.053865248483281e-07,
+      "loss": 1.0851,
+      "step": 3608
+    },
+    {
+      "epoch": 2.731504257332072,
+      "grad_norm": 2.1965982913970947,
+      "learning_rate": 4.031324227413297e-07,
+      "loss": 1.0758,
+      "step": 3609
+    },
+    {
+      "epoch": 2.7322611163670767,
+      "grad_norm": 2.31892728805542,
+      "learning_rate": 4.008845072734016e-07,
+      "loss": 1.1159,
+      "step": 3610
+    },
+    {
+      "epoch": 2.733017975402081,
+      "grad_norm": 2.0228688716888428,
+      "learning_rate": 3.986427798935131e-07,
+      "loss": 1.0769,
+      "step": 3611
+    },
+    {
+      "epoch": 2.733774834437086,
+      "grad_norm": 2.0157992839813232,
+      "learning_rate": 3.964072420466503e-07,
+      "loss": 1.0597,
+      "step": 3612
+    },
+    {
+      "epoch": 2.734531693472091,
+      "grad_norm": 1.9818907976150513,
+      "learning_rate": 3.9417789517380527e-07,
+      "loss": 1.0732,
+      "step": 3613
+    },
+    {
+      "epoch": 2.735288552507096,
+      "grad_norm": 2.1533520221710205,
+      "learning_rate": 3.919547407119824e-07,
+      "loss": 1.063,
+      "step": 3614
+    },
+    {
+      "epoch": 2.7360454115421002,
+      "grad_norm": 2.073683738708496,
+      "learning_rate": 3.897377800941943e-07,
+      "loss": 1.0551,
+      "step": 3615
+    },
+    {
+      "epoch": 2.736802270577105,
+      "grad_norm": 2.021272897720337,
+      "learning_rate": 3.875270147494558e-07,
+      "loss": 1.1027,
+      "step": 3616
+    },
+    {
+      "epoch": 2.7375591296121096,
+      "grad_norm": 2.308957099914551,
+      "learning_rate": 3.853224461027956e-07,
+      "loss": 1.076,
+      "step": 3617
+    },
+    {
+      "epoch": 2.7383159886471145,
+      "grad_norm": 2.239806652069092,
+      "learning_rate": 3.8312407557524466e-07,
+      "loss": 1.0998,
+      "step": 3618
+    },
+    {
+      "epoch": 2.7390728476821193,
+      "grad_norm": 2.1331143379211426,
+      "learning_rate": 3.8093190458383777e-07,
+      "loss": 1.1259,
+      "step": 3619
+    },
+    {
+      "epoch": 2.739829706717124,
+      "grad_norm": 2.0615665912628174,
+      "learning_rate": 3.7874593454161647e-07,
+      "loss": 1.0191,
+      "step": 3620
+    },
+    {
+      "epoch": 2.7405865657521287,
+      "grad_norm": 1.9834305047988892,
+      "learning_rate": 3.7656616685762473e-07,
+      "loss": 1.0553,
+      "step": 3621
+    },
+    {
+      "epoch": 2.7413434247871336,
+      "grad_norm": 2.1964480876922607,
+      "learning_rate": 3.7439260293690597e-07,
+      "loss": 1.0388,
+      "step": 3622
+    },
+    {
+      "epoch": 2.742100283822138,
+      "grad_norm": 1.9631261825561523,
+      "learning_rate": 3.722252441805057e-07,
+      "loss": 1.0515,
+      "step": 3623
+    },
+    {
+      "epoch": 2.742857142857143,
+      "grad_norm": 2.0425281524658203,
+      "learning_rate": 3.7006409198547004e-07,
+      "loss": 1.0117,
+      "step": 3624
+    },
+    {
+      "epoch": 2.7436140018921478,
+      "grad_norm": 2.2030279636383057,
+      "learning_rate": 3.6790914774484625e-07,
+      "loss": 1.0627,
+      "step": 3625
+    },
+    {
+      "epoch": 2.744370860927152,
+      "grad_norm": 2.0677294731140137,
+      "learning_rate": 3.6576041284767873e-07,
+      "loss": 1.0504,
+      "step": 3626
+    },
+    {
+      "epoch": 2.745127719962157,
+      "grad_norm": 1.951145887374878,
+      "learning_rate": 3.6361788867900865e-07,
+      "loss": 1.0804,
+      "step": 3627
+    },
+    {
+      "epoch": 2.7458845789971615,
+      "grad_norm": 2.083099126815796,
+      "learning_rate": 3.614815766198731e-07,
+      "loss": 1.0902,
+      "step": 3628
+    },
+    {
+      "epoch": 2.7466414380321664,
+      "grad_norm": 2.0568675994873047,
+      "learning_rate": 3.593514780473093e-07,
+      "loss": 1.0659,
+      "step": 3629
+    },
+    {
+      "epoch": 2.7473982970671713,
+      "grad_norm": 2.0525779724121094,
+      "learning_rate": 3.572275943343428e-07,
+      "loss": 1.0885,
+      "step": 3630
+    },
+    {
+      "epoch": 2.748155156102176,
+      "grad_norm": 2.105832576751709,
+      "learning_rate": 3.55109926849998e-07,
+      "loss": 1.0754,
+      "step": 3631
+    },
+    {
+      "epoch": 2.7489120151371806,
+      "grad_norm": 2.1376736164093018,
+      "learning_rate": 3.5299847695929306e-07,
+      "loss": 1.1257,
+      "step": 3632
+    },
+    {
+      "epoch": 2.7496688741721855,
+      "grad_norm": 1.8944578170776367,
+      "learning_rate": 3.508932460232331e-07,
+      "loss": 1.0465,
+      "step": 3633
+    },
+    {
+      "epoch": 2.75042573320719,
+      "grad_norm": 2.0414884090423584,
+      "learning_rate": 3.4879423539882017e-07,
+      "loss": 1.0621,
+      "step": 3634
+    },
+    {
+      "epoch": 2.751182592242195,
+      "grad_norm": 2.1808700561523438,
+      "learning_rate": 3.467014464390431e-07,
+      "loss": 1.0513,
+      "step": 3635
+    },
+    {
+      "epoch": 2.7519394512771997,
+      "grad_norm": 2.060415506362915,
+      "learning_rate": 3.446148804928836e-07,
+      "loss": 1.0927,
+      "step": 3636
+    },
+    {
+      "epoch": 2.752696310312204,
+      "grad_norm": 1.9510079622268677,
+      "learning_rate": 3.425345389053098e-07,
+      "loss": 1.0684,
+      "step": 3637
+    },
+    {
+      "epoch": 2.753453169347209,
+      "grad_norm": 1.9349720478057861,
+      "learning_rate": 3.4046042301727504e-07,
+      "loss": 1.0437,
+      "step": 3638
+    },
+    {
+      "epoch": 2.754210028382214,
+      "grad_norm": 1.9200588464736938,
+      "learning_rate": 3.383925341657259e-07,
+      "loss": 1.0417,
+      "step": 3639
+    },
+    {
+      "epoch": 2.7549668874172184,
+      "grad_norm": 1.9135462045669556,
+      "learning_rate": 3.363308736835918e-07,
+      "loss": 1.0593,
+      "step": 3640
+    },
+    {
+      "epoch": 2.7557237464522233,
+      "grad_norm": 2.030207633972168,
+      "learning_rate": 3.342754428997865e-07,
+      "loss": 1.0311,
+      "step": 3641
+    },
+    {
+      "epoch": 2.756480605487228,
+      "grad_norm": 2.0563161373138428,
+      "learning_rate": 3.3222624313920995e-07,
+      "loss": 1.1101,
+      "step": 3642
+    },
+    {
+      "epoch": 2.7572374645222326,
+      "grad_norm": 2.1681125164031982,
+      "learning_rate": 3.301832757227478e-07,
+      "loss": 1.1007,
+      "step": 3643
+    },
+    {
+      "epoch": 2.7579943235572375,
+      "grad_norm": 1.9184566736221313,
+      "learning_rate": 3.281465419672603e-07,
+      "loss": 1.0738,
+      "step": 3644
+    },
+    {
+      "epoch": 2.758751182592242,
+      "grad_norm": 2.2246665954589844,
+      "learning_rate": 3.261160431855982e-07,
+      "loss": 1.0967,
+      "step": 3645
+    },
+    {
+      "epoch": 2.759508041627247,
+      "grad_norm": 2.213003396987915,
+      "learning_rate": 3.240917806865891e-07,
+      "loss": 1.1179,
+      "step": 3646
+    },
+    {
+      "epoch": 2.7602649006622517,
+      "grad_norm": 2.001859426498413,
+      "learning_rate": 3.2207375577504196e-07,
+      "loss": 1.0601,
+      "step": 3647
+    },
+    {
+      "epoch": 2.7610217596972566,
+      "grad_norm": 1.9976732730865479,
+      "learning_rate": 3.2006196975174716e-07,
+      "loss": 1.0809,
+      "step": 3648
+    },
+    {
+      "epoch": 2.761778618732261,
+      "grad_norm": 2.0903263092041016,
+      "learning_rate": 3.1805642391346757e-07,
+      "loss": 1.1196,
+      "step": 3649
+    },
+    {
+      "epoch": 2.762535477767266,
+      "grad_norm": 2.2454066276550293,
+      "learning_rate": 3.160571195529498e-07,
+      "loss": 1.1021,
+      "step": 3650
+    },
+    {
+      "epoch": 2.7632923368022704,
+      "grad_norm": 2.1280694007873535,
+      "learning_rate": 3.1406405795891286e-07,
+      "loss": 1.1027,
+      "step": 3651
+    },
+    {
+      "epoch": 2.7640491958372753,
+      "grad_norm": 2.1310126781463623,
+      "learning_rate": 3.1207724041605493e-07,
+      "loss": 0.9978,
+      "step": 3652
+    },
+    {
+      "epoch": 2.76480605487228,
+      "grad_norm": 2.2121293544769287,
+      "learning_rate": 3.1009666820505004e-07,
+      "loss": 1.0708,
+      "step": 3653
+    },
+    {
+      "epoch": 2.765562913907285,
+      "grad_norm": 2.256673812866211,
+      "learning_rate": 3.081223426025437e-07,
+      "loss": 1.094,
+      "step": 3654
+    },
+    {
+      "epoch": 2.7663197729422895,
+      "grad_norm": 2.2821056842803955,
+      "learning_rate": 3.0615426488115385e-07,
+      "loss": 1.0542,
+      "step": 3655
+    },
+    {
+      "epoch": 2.7670766319772944,
+      "grad_norm": 2.1040828227996826,
+      "learning_rate": 3.0419243630947764e-07,
+      "loss": 1.0439,
+      "step": 3656
+    },
+    {
+      "epoch": 2.767833491012299,
+      "grad_norm": 2.050218343734741,
+      "learning_rate": 3.022368581520758e-07,
+      "loss": 1.0747,
+      "step": 3657
+    },
+    {
+      "epoch": 2.7685903500473037,
+      "grad_norm": 1.962795376777649,
+      "learning_rate": 3.0028753166948504e-07,
+      "loss": 1.1227,
+      "step": 3658
+    },
+    {
+      "epoch": 2.7693472090823086,
+      "grad_norm": 2.256727933883667,
+      "learning_rate": 2.983444581182144e-07,
+      "loss": 1.123,
+      "step": 3659
+    },
+    {
+      "epoch": 2.770104068117313,
+      "grad_norm": 2.0236082077026367,
+      "learning_rate": 2.964076387507367e-07,
+      "loss": 1.0793,
+      "step": 3660
+    },
+    {
+      "epoch": 2.770860927152318,
+      "grad_norm": 2.2242016792297363,
+      "learning_rate": 2.944770748154961e-07,
+      "loss": 1.0849,
+      "step": 3661
+    },
+    {
+      "epoch": 2.7716177861873224,
+      "grad_norm": 1.8702110052108765,
+      "learning_rate": 2.9255276755690594e-07,
+      "loss": 1.0598,
+      "step": 3662
+    },
+    {
+      "epoch": 2.7723746452223272,
+      "grad_norm": 2.0809333324432373,
+      "learning_rate": 2.9063471821534544e-07,
+      "loss": 1.1008,
+      "step": 3663
+    },
+    {
+      "epoch": 2.773131504257332,
+      "grad_norm": 2.0802369117736816,
+      "learning_rate": 2.8872292802715856e-07,
+      "loss": 1.0757,
+      "step": 3664
+    },
+    {
+      "epoch": 2.773888363292337,
+      "grad_norm": 2.361516237258911,
+      "learning_rate": 2.868173982246573e-07,
+      "loss": 1.1079,
+      "step": 3665
+    },
+    {
+      "epoch": 2.7746452223273415,
+      "grad_norm": 2.069173574447632,
+      "learning_rate": 2.8491813003611614e-07,
+      "loss": 1.0559,
+      "step": 3666
+    },
+    {
+      "epoch": 2.7754020813623463,
+      "grad_norm": 1.9263156652450562,
+      "learning_rate": 2.830251246857745e-07,
+      "loss": 1.054,
+      "step": 3667
+    },
+    {
+      "epoch": 2.776158940397351,
+      "grad_norm": 1.9880831241607666,
+      "learning_rate": 2.811383833938352e-07,
+      "loss": 1.0915,
+      "step": 3668
+    },
+    {
+      "epoch": 2.7769157994323557,
+      "grad_norm": 1.977330207824707,
+      "learning_rate": 2.7925790737646375e-07,
+      "loss": 1.0582,
+      "step": 3669
+    },
+    {
+      "epoch": 2.7776726584673606,
+      "grad_norm": 2.2954440116882324,
+      "learning_rate": 2.7738369784578694e-07,
+      "loss": 1.093,
+      "step": 3670
+    },
+    {
+      "epoch": 2.7784295175023654,
+      "grad_norm": 1.9425572156906128,
+      "learning_rate": 2.755157560098875e-07,
+      "loss": 1.0558,
+      "step": 3671
+    },
+    {
+      "epoch": 2.77918637653737,
+      "grad_norm": 1.9914302825927734,
+      "learning_rate": 2.736540830728152e-07,
+      "loss": 1.1168,
+      "step": 3672
+    },
+    {
+      "epoch": 2.7799432355723748,
+      "grad_norm": 2.00032114982605,
+      "learning_rate": 2.717986802345765e-07,
+      "loss": 1.0692,
+      "step": 3673
+    },
+    {
+      "epoch": 2.780700094607379,
+      "grad_norm": 2.004713535308838,
+      "learning_rate": 2.6994954869113416e-07,
+      "loss": 1.0587,
+      "step": 3674
+    },
+    {
+      "epoch": 2.781456953642384,
+      "grad_norm": 2.0385048389434814,
+      "learning_rate": 2.6810668963441194e-07,
+      "loss": 1.064,
+      "step": 3675
+    },
+    {
+      "epoch": 2.782213812677389,
+      "grad_norm": 1.966386079788208,
+      "learning_rate": 2.662701042522874e-07,
+      "loss": 1.0086,
+      "step": 3676
+    },
+    {
+      "epoch": 2.7829706717123934,
+      "grad_norm": 2.2672669887542725,
+      "learning_rate": 2.644397937285963e-07,
+      "loss": 1.0867,
+      "step": 3677
+    },
+    {
+      "epoch": 2.7837275307473983,
+      "grad_norm": 1.9181667566299438,
+      "learning_rate": 2.626157592431278e-07,
+      "loss": 1.0969,
+      "step": 3678
+    },
+    {
+      "epoch": 2.7844843897824028,
+      "grad_norm": 2.0945870876312256,
+      "learning_rate": 2.607980019716272e-07,
+      "loss": 1.0493,
+      "step": 3679
+    },
+    {
+      "epoch": 2.7852412488174076,
+      "grad_norm": 1.9859826564788818,
+      "learning_rate": 2.589865230857932e-07,
+      "loss": 1.0695,
+      "step": 3680
+    },
+    {
+      "epoch": 2.7859981078524125,
+      "grad_norm": 1.9504257440567017,
+      "learning_rate": 2.5718132375327933e-07,
+      "loss": 1.0653,
+      "step": 3681
+    },
+    {
+      "epoch": 2.7867549668874174,
+      "grad_norm": 1.9905445575714111,
+      "learning_rate": 2.5538240513768625e-07,
+      "loss": 1.0907,
+      "step": 3682
+    },
+    {
+      "epoch": 2.787511825922422,
+      "grad_norm": 2.076355457305908,
+      "learning_rate": 2.535897683985702e-07,
+      "loss": 1.0448,
+      "step": 3683
+    },
+    {
+      "epoch": 2.7882686849574267,
+      "grad_norm": 1.986864447593689,
+      "learning_rate": 2.518034146914401e-07,
+      "loss": 1.0296,
+      "step": 3684
+    },
+    {
+      "epoch": 2.789025543992431,
+      "grad_norm": 1.9109041690826416,
+      "learning_rate": 2.5002334516774865e-07,
+      "loss": 1.0455,
+      "step": 3685
+    },
+    {
+      "epoch": 2.789782403027436,
+      "grad_norm": 2.183528423309326,
+      "learning_rate": 2.482495609749042e-07,
+      "loss": 1.0824,
+      "step": 3686
+    },
+    {
+      "epoch": 2.790539262062441,
+      "grad_norm": 2.090740203857422,
+      "learning_rate": 2.4648206325626e-07,
+      "loss": 1.0755,
+      "step": 3687
+    },
+    {
+      "epoch": 2.791296121097446,
+      "grad_norm": 2.1614151000976562,
+      "learning_rate": 2.447208531511184e-07,
+      "loss": 1.0617,
+      "step": 3688
+    },
+    {
+      "epoch": 2.7920529801324503,
+      "grad_norm": 1.9354277849197388,
+      "learning_rate": 2.429659317947277e-07,
+      "loss": 1.0702,
+      "step": 3689
+    },
+    {
+      "epoch": 2.792809839167455,
+      "grad_norm": 2.077448606491089,
+      "learning_rate": 2.412173003182842e-07,
+      "loss": 1.0656,
+      "step": 3690
+    },
+    {
+      "epoch": 2.7935666982024596,
+      "grad_norm": 1.9370477199554443,
+      "learning_rate": 2.394749598489302e-07,
+      "loss": 1.1324,
+      "step": 3691
+    },
+    {
+      "epoch": 2.7943235572374645,
+      "grad_norm": 2.0902650356292725,
+      "learning_rate": 2.3773891150975041e-07,
+      "loss": 1.1173,
+      "step": 3692
+    },
+    {
+      "epoch": 2.7950804162724694,
+      "grad_norm": 2.1968994140625,
+      "learning_rate": 2.3600915641977443e-07,
+      "loss": 1.1001,
+      "step": 3693
+    },
+    {
+      "epoch": 2.795837275307474,
+      "grad_norm": 1.9441262483596802,
+      "learning_rate": 2.342856956939765e-07,
+      "loss": 1.0932,
+      "step": 3694
+    },
+    {
+      "epoch": 2.7965941343424787,
+      "grad_norm": 2.0278730392456055,
+      "learning_rate": 2.3256853044327348e-07,
+      "loss": 1.1073,
+      "step": 3695
+    },
+    {
+      "epoch": 2.7973509933774836,
+      "grad_norm": 1.9890429973602295,
+      "learning_rate": 2.308576617745247e-07,
+      "loss": 1.0403,
+      "step": 3696
+    },
+    {
+      "epoch": 2.798107852412488,
+      "grad_norm": 1.9365586042404175,
+      "learning_rate": 2.2915309079052886e-07,
+      "loss": 1.0604,
+      "step": 3697
+    },
+    {
+      "epoch": 2.798864711447493,
+      "grad_norm": 2.051670789718628,
+      "learning_rate": 2.2745481859002917e-07,
+      "loss": 1.0739,
+      "step": 3698
+    },
+    {
+      "epoch": 2.799621570482498,
+      "grad_norm": 1.945073127746582,
+      "learning_rate": 2.2576284626770157e-07,
+      "loss": 1.0621,
+      "step": 3699
+    },
+    {
+      "epoch": 2.8003784295175023,
+      "grad_norm": 2.0430312156677246,
+      "learning_rate": 2.2407717491416676e-07,
+      "loss": 1.058,
+      "step": 3700
+    },
+    {
+      "epoch": 2.801135288552507,
+      "grad_norm": 2.074920177459717,
+      "learning_rate": 2.2239780561598455e-07,
+      "loss": 1.0765,
+      "step": 3701
+    },
+    {
+      "epoch": 2.8018921475875116,
+      "grad_norm": 2.323629140853882,
+      "learning_rate": 2.2072473945564961e-07,
+      "loss": 1.1025,
+      "step": 3702
+    },
+    {
+      "epoch": 2.8026490066225165,
+      "grad_norm": 1.8663524389266968,
+      "learning_rate": 2.1905797751159689e-07,
+      "loss": 1.0688,
+      "step": 3703
+    },
+    {
+      "epoch": 2.8034058656575214,
+      "grad_norm": 2.141047239303589,
+      "learning_rate": 2.1739752085819388e-07,
+      "loss": 1.0787,
+      "step": 3704
+    },
+    {
+      "epoch": 2.8041627246925263,
+      "grad_norm": 2.179725408554077,
+      "learning_rate": 2.15743370565744e-07,
+      "loss": 1.0848,
+      "step": 3705
+    },
+    {
+      "epoch": 2.8049195837275307,
+      "grad_norm": 2.0025246143341064,
+      "learning_rate": 2.1409552770048975e-07,
+      "loss": 1.0256,
+      "step": 3706
+    },
+    {
+      "epoch": 2.8056764427625356,
+      "grad_norm": 2.1321537494659424,
+      "learning_rate": 2.124539933246042e-07,
+      "loss": 1.1045,
+      "step": 3707
+    },
+    {
+      "epoch": 2.80643330179754,
+      "grad_norm": 2.0465590953826904,
+      "learning_rate": 2.108187684961972e-07,
+      "loss": 1.1277,
+      "step": 3708
+    },
+    {
+      "epoch": 2.807190160832545,
+      "grad_norm": 2.1039795875549316,
+      "learning_rate": 2.091898542693078e-07,
+      "loss": 1.0712,
+      "step": 3709
+    },
+    {
+      "epoch": 2.80794701986755,
+      "grad_norm": 2.3281686305999756,
+      "learning_rate": 2.0756725169391007e-07,
+      "loss": 1.1043,
+      "step": 3710
+    },
+    {
+      "epoch": 2.8087038789025542,
+      "grad_norm": 2.027113914489746,
+      "learning_rate": 2.0595096181591037e-07,
+      "loss": 1.1017,
+      "step": 3711
+    },
+    {
+      "epoch": 2.809460737937559,
+      "grad_norm": 2.1701509952545166,
+      "learning_rate": 2.04340985677141e-07,
+      "loss": 1.0812,
+      "step": 3712
+    },
+    {
+      "epoch": 2.810217596972564,
+      "grad_norm": 2.0328516960144043,
+      "learning_rate": 2.0273732431537025e-07,
+      "loss": 1.0951,
+      "step": 3713
+    },
+    {
+      "epoch": 2.8109744560075685,
+      "grad_norm": 2.0888283252716064,
+      "learning_rate": 2.0113997876429446e-07,
+      "loss": 1.1183,
+      "step": 3714
+    },
+    {
+      "epoch": 2.8117313150425733,
+      "grad_norm": 1.8546501398086548,
+      "learning_rate": 1.9954895005353692e-07,
+      "loss": 1.0742,
+      "step": 3715
+    },
+    {
+      "epoch": 2.8124881740775782,
+      "grad_norm": 1.985583782196045,
+      "learning_rate": 1.9796423920865021e-07,
+      "loss": 1.1027,
+      "step": 3716
+    },
+    {
+      "epoch": 2.8132450331125827,
+      "grad_norm": 2.0510141849517822,
+      "learning_rate": 1.9638584725111498e-07,
+      "loss": 1.0609,
+      "step": 3717
+    },
+    {
+      "epoch": 2.8140018921475876,
+      "grad_norm": 2.359945058822632,
+      "learning_rate": 1.9481377519834112e-07,
+      "loss": 1.0421,
+      "step": 3718
+    },
+    {
+      "epoch": 2.814758751182592,
+      "grad_norm": 2.107235908508301,
+      "learning_rate": 1.9324802406365883e-07,
+      "loss": 1.0114,
+      "step": 3719
+    },
+    {
+      "epoch": 2.815515610217597,
+      "grad_norm": 1.9575122594833374,
+      "learning_rate": 1.9168859485632866e-07,
+      "loss": 1.0763,
+      "step": 3720
+    },
+    {
+      "epoch": 2.8162724692526018,
+      "grad_norm": 2.15492582321167,
+      "learning_rate": 1.901354885815348e-07,
+      "loss": 1.0527,
+      "step": 3721
+    },
+    {
+      "epoch": 2.8170293282876067,
+      "grad_norm": 2.049591302871704,
+      "learning_rate": 1.8858870624038632e-07,
+      "loss": 1.0314,
+      "step": 3722
+    },
+    {
+      "epoch": 2.817786187322611,
+      "grad_norm": 2.623854875564575,
+      "learning_rate": 1.8704824882991584e-07,
+      "loss": 1.0589,
+      "step": 3723
+    },
+    {
+      "epoch": 2.818543046357616,
+      "grad_norm": 1.8997153043746948,
+      "learning_rate": 1.8551411734307744e-07,
+      "loss": 1.0561,
+      "step": 3724
+    },
+    {
+      "epoch": 2.8192999053926204,
+      "grad_norm": 2.2084269523620605,
+      "learning_rate": 1.8398631276875118e-07,
+      "loss": 1.0703,
+      "step": 3725
+    },
+    {
+      "epoch": 2.8200567644276253,
+      "grad_norm": 1.8947069644927979,
+      "learning_rate": 1.82464836091734e-07,
+      "loss": 1.0454,
+      "step": 3726
+    },
+    {
+      "epoch": 2.82081362346263,
+      "grad_norm": 1.8694313764572144,
+      "learning_rate": 1.8094968829274663e-07,
+      "loss": 1.0885,
+      "step": 3727
+    },
+    {
+      "epoch": 2.821570482497635,
+      "grad_norm": 2.0910801887512207,
+      "learning_rate": 1.7944087034843233e-07,
+      "loss": 1.0546,
+      "step": 3728
+    },
+    {
+      "epoch": 2.8223273415326395,
+      "grad_norm": 1.8177095651626587,
+      "learning_rate": 1.7793838323135016e-07,
+      "loss": 1.0405,
+      "step": 3729
+    },
+    {
+      "epoch": 2.8230842005676444,
+      "grad_norm": 2.050400733947754,
+      "learning_rate": 1.7644222790998186e-07,
+      "loss": 1.0905,
+      "step": 3730
+    },
+    {
+      "epoch": 2.823841059602649,
+      "grad_norm": 1.9098093509674072,
+      "learning_rate": 1.7495240534872614e-07,
+      "loss": 1.031,
+      "step": 3731
+    },
+    {
+      "epoch": 2.8245979186376537,
+      "grad_norm": 2.1355783939361572,
+      "learning_rate": 1.734689165078998e-07,
+      "loss": 1.0944,
+      "step": 3732
+    },
+    {
+      "epoch": 2.8253547776726586,
+      "grad_norm": 1.9840859174728394,
+      "learning_rate": 1.7199176234373553e-07,
+      "loss": 0.9905,
+      "step": 3733
+    },
+    {
+      "epoch": 2.826111636707663,
+      "grad_norm": 2.0721471309661865,
+      "learning_rate": 1.7052094380838532e-07,
+      "loss": 1.059,
+      "step": 3734
+    },
+    {
+      "epoch": 2.826868495742668,
+      "grad_norm": 2.345816135406494,
+      "learning_rate": 1.69056461849917e-07,
+      "loss": 1.0611,
+      "step": 3735
+    },
+    {
+      "epoch": 2.8276253547776724,
+      "grad_norm": 2.2599689960479736,
+      "learning_rate": 1.675983174123143e-07,
+      "loss": 1.0965,
+      "step": 3736
+    },
+    {
+      "epoch": 2.8283822138126773,
+      "grad_norm": 2.1439452171325684,
+      "learning_rate": 1.6614651143547243e-07,
+      "loss": 1.0312,
+      "step": 3737
+    },
+    {
+      "epoch": 2.829139072847682,
+      "grad_norm": 2.0652458667755127,
+      "learning_rate": 1.647010448552047e-07,
+      "loss": 1.0469,
+      "step": 3738
+    },
+    {
+      "epoch": 2.829895931882687,
+      "grad_norm": 2.0724799633026123,
+      "learning_rate": 1.63261918603237e-07,
+      "loss": 1.1211,
+      "step": 3739
+    },
+    {
+      "epoch": 2.8306527909176915,
+      "grad_norm": 1.8345634937286377,
+      "learning_rate": 1.618291336072078e-07,
+      "loss": 1.0573,
+      "step": 3740
+    },
+    {
+      "epoch": 2.8314096499526964,
+      "grad_norm": 1.9076229333877563,
+      "learning_rate": 1.6040269079066806e-07,
+      "loss": 1.0767,
+      "step": 3741
+    },
+    {
+      "epoch": 2.832166508987701,
+      "grad_norm": 1.9758639335632324,
+      "learning_rate": 1.5898259107308255e-07,
+      "loss": 1.0678,
+      "step": 3742
+    },
+    {
+      "epoch": 2.8329233680227057,
+      "grad_norm": 1.982330560684204,
+      "learning_rate": 1.5756883536982296e-07,
+      "loss": 1.1126,
+      "step": 3743
+    },
+    {
+      "epoch": 2.8336802270577106,
+      "grad_norm": 1.9688644409179688,
+      "learning_rate": 1.5616142459217799e-07,
+      "loss": 1.1118,
+      "step": 3744
+    },
+    {
+      "epoch": 2.8344370860927155,
+      "grad_norm": 2.031545639038086,
+      "learning_rate": 1.5476035964734117e-07,
+      "loss": 1.061,
+      "step": 3745
+    },
+    {
+      "epoch": 2.83519394512772,
+      "grad_norm": 1.9219672679901123,
+      "learning_rate": 1.5336564143841856e-07,
+      "loss": 1.031,
+      "step": 3746
+    },
+    {
+      "epoch": 2.835950804162725,
+      "grad_norm": 2.1881892681121826,
+      "learning_rate": 1.5197727086442445e-07,
+      "loss": 1.0689,
+      "step": 3747
+    },
+    {
+      "epoch": 2.8367076631977293,
+      "grad_norm": 1.885879397392273,
+      "learning_rate": 1.505952488202789e-07,
+      "loss": 1.0866,
+      "step": 3748
+    },
+    {
+      "epoch": 2.837464522232734,
+      "grad_norm": 2.17256760597229,
+      "learning_rate": 1.492195761968146e-07,
+      "loss": 1.0774,
+      "step": 3749
+    },
+    {
+      "epoch": 2.838221381267739,
+      "grad_norm": 2.0141475200653076,
+      "learning_rate": 1.4785025388076906e-07,
+      "loss": 1.1078,
+      "step": 3750
+    },
+    {
+      "epoch": 2.8389782403027435,
+      "grad_norm": 2.1120545864105225,
+      "learning_rate": 1.4648728275478566e-07,
+      "loss": 1.0698,
+      "step": 3751
+    },
+    {
+      "epoch": 2.8397350993377484,
+      "grad_norm": 1.9993555545806885,
+      "learning_rate": 1.451306636974159e-07,
+      "loss": 1.0529,
+      "step": 3752
+    },
+    {
+      "epoch": 2.840491958372753,
+      "grad_norm": 1.9042015075683594,
+      "learning_rate": 1.4378039758311616e-07,
+      "loss": 1.035,
+      "step": 3753
+    },
+    {
+      "epoch": 2.8412488174077577,
+      "grad_norm": 2.0726895332336426,
+      "learning_rate": 1.4243648528224414e-07,
+      "loss": 1.0772,
+      "step": 3754
+    },
+    {
+      "epoch": 2.8420056764427626,
+      "grad_norm": 2.004347085952759,
+      "learning_rate": 1.4109892766106804e-07,
+      "loss": 1.066,
+      "step": 3755
+    },
+    {
+      "epoch": 2.8427625354777675,
+      "grad_norm": 2.1998095512390137,
+      "learning_rate": 1.397677255817563e-07,
+      "loss": 1.1011,
+      "step": 3756
+    },
+    {
+      "epoch": 2.843519394512772,
+      "grad_norm": 1.9528348445892334,
+      "learning_rate": 1.3844287990238113e-07,
+      "loss": 1.105,
+      "step": 3757
+    },
+    {
+      "epoch": 2.844276253547777,
+      "grad_norm": 2.1867001056671143,
+      "learning_rate": 1.3712439147691946e-07,
+      "loss": 1.0787,
+      "step": 3758
+    },
+    {
+      "epoch": 2.8450331125827812,
+      "grad_norm": 2.0233795642852783,
+      "learning_rate": 1.3581226115524753e-07,
+      "loss": 1.0587,
+      "step": 3759
+    },
+    {
+      "epoch": 2.845789971617786,
+      "grad_norm": 1.9928818941116333,
+      "learning_rate": 1.345064897831441e-07,
+      "loss": 1.0421,
+      "step": 3760
+    },
+    {
+      "epoch": 2.846546830652791,
+      "grad_norm": 1.9571059942245483,
+      "learning_rate": 1.3320707820229063e-07,
+      "loss": 1.0569,
+      "step": 3761
+    },
+    {
+      "epoch": 2.847303689687796,
+      "grad_norm": 2.076955795288086,
+      "learning_rate": 1.3191402725026765e-07,
+      "loss": 1.0854,
+      "step": 3762
+    },
+    {
+      "epoch": 2.8480605487228003,
+      "grad_norm": 2.1233267784118652,
+      "learning_rate": 1.3062733776055504e-07,
+      "loss": 1.0457,
+      "step": 3763
+    },
+    {
+      "epoch": 2.8488174077578052,
+      "grad_norm": 1.9417656660079956,
+      "learning_rate": 1.2934701056253526e-07,
+      "loss": 1.0308,
+      "step": 3764
+    },
+    {
+      "epoch": 2.8495742667928097,
+      "grad_norm": 1.9117321968078613,
+      "learning_rate": 1.2807304648148552e-07,
+      "loss": 1.0519,
+      "step": 3765
+    },
+    {
+      "epoch": 2.8503311258278146,
+      "grad_norm": 1.987637996673584,
+      "learning_rate": 1.2680544633858457e-07,
+      "loss": 1.0602,
+      "step": 3766
+    },
+    {
+      "epoch": 2.8510879848628194,
+      "grad_norm": 2.072512626647949,
+      "learning_rate": 1.2554421095090923e-07,
+      "loss": 1.1063,
+      "step": 3767
+    },
+    {
+      "epoch": 2.851844843897824,
+      "grad_norm": 2.4176509380340576,
+      "learning_rate": 1.2428934113143005e-07,
+      "loss": 1.0889,
+      "step": 3768
+    },
+    {
+      "epoch": 2.8526017029328288,
+      "grad_norm": 2.25588059425354,
+      "learning_rate": 1.2304083768902016e-07,
+      "loss": 1.0894,
+      "step": 3769
+    },
+    {
+      "epoch": 2.853358561967833,
+      "grad_norm": 1.9961562156677246,
+      "learning_rate": 1.2179870142844305e-07,
+      "loss": 1.07,
+      "step": 3770
+    },
+    {
+      "epoch": 2.854115421002838,
+      "grad_norm": 2.0790538787841797,
+      "learning_rate": 1.2056293315036139e-07,
+      "loss": 1.1308,
+      "step": 3771
+    },
+    {
+      "epoch": 2.854872280037843,
+      "grad_norm": 2.107841968536377,
+      "learning_rate": 1.1933353365133393e-07,
+      "loss": 1.1053,
+      "step": 3772
+    },
+    {
+      "epoch": 2.855629139072848,
+      "grad_norm": 1.9669723510742188,
+      "learning_rate": 1.1811050372381292e-07,
+      "loss": 1.1049,
+      "step": 3773
+    },
+    {
+      "epoch": 2.8563859981078523,
+      "grad_norm": 1.9607486724853516,
+      "learning_rate": 1.1689384415614223e-07,
+      "loss": 1.0817,
+      "step": 3774
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 1.8808550834655762,
+      "learning_rate": 1.1568355573256491e-07,
+      "loss": 1.0328,
+      "step": 3775
+    },
+    {
+      "epoch": 2.8578997161778616,
+      "grad_norm": 2.0765459537506104,
+      "learning_rate": 1.1447963923321327e-07,
+      "loss": 1.0449,
+      "step": 3776
+    },
+    {
+      "epoch": 2.8586565752128665,
+      "grad_norm": 1.9006658792495728,
+      "learning_rate": 1.1328209543411224e-07,
+      "loss": 1.0046,
+      "step": 3777
+    },
+    {
+      "epoch": 2.8594134342478714,
+      "grad_norm": 2.1098666191101074,
+      "learning_rate": 1.1209092510718261e-07,
+      "loss": 1.0837,
+      "step": 3778
+    },
+    {
+      "epoch": 2.8601702932828763,
+      "grad_norm": 2.088935375213623,
+      "learning_rate": 1.1090612902023337e-07,
+      "loss": 1.0702,
+      "step": 3779
+    },
+    {
+      "epoch": 2.8609271523178808,
+      "grad_norm": 2.2443082332611084,
+      "learning_rate": 1.0972770793696717e-07,
+      "loss": 1.0864,
+      "step": 3780
+    },
+    {
+      "epoch": 2.8616840113528856,
+      "grad_norm": 2.380600929260254,
+      "learning_rate": 1.0855566261697372e-07,
+      "loss": 1.073,
+      "step": 3781
+    },
+    {
+      "epoch": 2.86244087038789,
+      "grad_norm": 2.0619399547576904,
+      "learning_rate": 1.073899938157375e-07,
+      "loss": 1.0486,
+      "step": 3782
+    },
+    {
+      "epoch": 2.863197729422895,
+      "grad_norm": 1.9430749416351318,
+      "learning_rate": 1.0623070228463008e-07,
+      "loss": 1.0664,
+      "step": 3783
+    },
+    {
+      "epoch": 2.8639545884579,
+      "grad_norm": 1.9037846326828003,
+      "learning_rate": 1.0507778877091445e-07,
+      "loss": 1.041,
+      "step": 3784
+    },
+    {
+      "epoch": 2.8647114474929043,
+      "grad_norm": 1.8566458225250244,
+      "learning_rate": 1.0393125401773843e-07,
+      "loss": 1.066,
+      "step": 3785
+    },
+    {
+      "epoch": 2.865468306527909,
+      "grad_norm": 1.8834096193313599,
+      "learning_rate": 1.027910987641447e-07,
+      "loss": 1.0902,
+      "step": 3786
+    },
+    {
+      "epoch": 2.866225165562914,
+      "grad_norm": 2.1945738792419434,
+      "learning_rate": 1.0165732374505733e-07,
+      "loss": 1.0731,
+      "step": 3787
+    },
+    {
+      "epoch": 2.8669820245979185,
+      "grad_norm": 2.057518720626831,
+      "learning_rate": 1.0052992969128971e-07,
+      "loss": 1.0807,
+      "step": 3788
+    },
+    {
+      "epoch": 2.8677388836329234,
+      "grad_norm": 1.897512435913086,
+      "learning_rate": 9.940891732954447e-08,
+      "loss": 1.0146,
+      "step": 3789
+    },
+    {
+      "epoch": 2.8684957426679283,
+      "grad_norm": 1.9884440898895264,
+      "learning_rate": 9.829428738240904e-08,
+      "loss": 1.0896,
+      "step": 3790
+    },
+    {
+      "epoch": 2.8692526017029327,
+      "grad_norm": 2.2165613174438477,
+      "learning_rate": 9.718604056835573e-08,
+      "loss": 1.014,
+      "step": 3791
+    },
+    {
+      "epoch": 2.8700094607379376,
+      "grad_norm": 2.517817258834839,
+      "learning_rate": 9.608417760174488e-08,
+      "loss": 1.0497,
+      "step": 3792
+    },
+    {
+      "epoch": 2.870766319772942,
+      "grad_norm": 1.8578647375106812,
+      "learning_rate": 9.498869919281952e-08,
+      "loss": 1.064,
+      "step": 3793
+    },
+    {
+      "epoch": 2.871523178807947,
+      "grad_norm": 2.168428897857666,
+      "learning_rate": 9.389960604770966e-08,
+      "loss": 1.1338,
+      "step": 3794
+    },
+    {
+      "epoch": 2.872280037842952,
+      "grad_norm": 1.902740716934204,
+      "learning_rate": 9.281689886842575e-08,
+      "loss": 1.0833,
+      "step": 3795
+    },
+    {
+      "epoch": 2.8730368968779567,
+      "grad_norm": 2.10799503326416,
+      "learning_rate": 9.174057835286632e-08,
+      "loss": 1.0498,
+      "step": 3796
+    },
+    {
+      "epoch": 2.873793755912961,
+      "grad_norm": 1.9497560262680054,
+      "learning_rate": 9.067064519481139e-08,
+      "loss": 1.0824,
+      "step": 3797
+    },
+    {
+      "epoch": 2.874550614947966,
+      "grad_norm": 2.052687406539917,
+      "learning_rate": 8.96071000839214e-08,
+      "loss": 1.0838,
+      "step": 3798
+    },
+    {
+      "epoch": 2.8753074739829705,
+      "grad_norm": 2.033168315887451,
+      "learning_rate": 8.854994370574378e-08,
+      "loss": 1.0696,
+      "step": 3799
+    },
+    {
+      "epoch": 2.8760643330179754,
+      "grad_norm": 1.928364872932434,
+      "learning_rate": 8.749917674170415e-08,
+      "loss": 1.0824,
+      "step": 3800
+    },
+    {
+      "epoch": 2.8768211920529803,
+      "grad_norm": 1.9337732791900635,
+      "learning_rate": 8.645479986911066e-08,
+      "loss": 1.117,
+      "step": 3801
+    },
+    {
+      "epoch": 2.8775780510879847,
+      "grad_norm": 1.929337501525879,
+      "learning_rate": 8.541681376115416e-08,
+      "loss": 1.0448,
+      "step": 3802
+    },
+    {
+      "epoch": 2.8783349101229896,
+      "grad_norm": 2.0957815647125244,
+      "learning_rate": 8.438521908690244e-08,
+      "loss": 1.0833,
+      "step": 3803
+    },
+    {
+      "epoch": 2.8790917691579945,
+      "grad_norm": 2.090304136276245,
+      "learning_rate": 8.336001651130706e-08,
+      "loss": 1.0567,
+      "step": 3804
+    },
+    {
+      "epoch": 2.879848628192999,
+      "grad_norm": 2.115295648574829,
+      "learning_rate": 8.234120669519771e-08,
+      "loss": 1.031,
+      "step": 3805
+    },
+    {
+      "epoch": 2.880605487228004,
+      "grad_norm": 1.9636808633804321,
+      "learning_rate": 8.132879029528445e-08,
+      "loss": 1.0494,
+      "step": 3806
+    },
+    {
+      "epoch": 2.8813623462630087,
+      "grad_norm": 1.9101999998092651,
+      "learning_rate": 8.03227679641533e-08,
+      "loss": 1.0571,
+      "step": 3807
+    },
+    {
+      "epoch": 2.882119205298013,
+      "grad_norm": 1.9551316499710083,
+      "learning_rate": 7.932314035027393e-08,
+      "loss": 1.0658,
+      "step": 3808
+    },
+    {
+      "epoch": 2.882876064333018,
+      "grad_norm": 2.1239876747131348,
+      "learning_rate": 7.832990809798869e-08,
+      "loss": 1.0788,
+      "step": 3809
+    },
+    {
+      "epoch": 2.8836329233680225,
+      "grad_norm": 1.9690558910369873,
+      "learning_rate": 7.734307184752134e-08,
+      "loss": 1.0772,
+      "step": 3810
+    },
+    {
+      "epoch": 2.8843897824030273,
+      "grad_norm": 2.071542263031006,
+      "learning_rate": 7.636263223496941e-08,
+      "loss": 1.0839,
+      "step": 3811
+    },
+    {
+      "epoch": 2.8851466414380322,
+      "grad_norm": 2.1702964305877686,
+      "learning_rate": 7.538858989231189e-08,
+      "loss": 1.0452,
+      "step": 3812
+    },
+    {
+      "epoch": 2.885903500473037,
+      "grad_norm": 2.1600115299224854,
+      "learning_rate": 7.442094544740037e-08,
+      "loss": 1.133,
+      "step": 3813
+    },
+    {
+      "epoch": 2.8866603595080416,
+      "grad_norm": 1.943969964981079,
+      "learning_rate": 7.34596995239646e-08,
+      "loss": 1.0342,
+      "step": 3814
+    },
+    {
+      "epoch": 2.8874172185430464,
+      "grad_norm": 2.029170513153076,
+      "learning_rate": 7.250485274160693e-08,
+      "loss": 1.0983,
+      "step": 3815
+    },
+    {
+      "epoch": 2.888174077578051,
+      "grad_norm": 2.1345629692077637,
+      "learning_rate": 7.1556405715809e-08,
+      "loss": 1.0854,
+      "step": 3816
+    },
+    {
+      "epoch": 2.8889309366130558,
+      "grad_norm": 1.8675469160079956,
+      "learning_rate": 7.061435905792389e-08,
+      "loss": 1.0661,
+      "step": 3817
+    },
+    {
+      "epoch": 2.8896877956480607,
+      "grad_norm": 2.0254111289978027,
+      "learning_rate": 6.967871337518176e-08,
+      "loss": 1.103,
+      "step": 3818
+    },
+    {
+      "epoch": 2.8904446546830656,
+      "grad_norm": 2.162344455718994,
+      "learning_rate": 6.874946927068538e-08,
+      "loss": 1.0878,
+      "step": 3819
+    },
+    {
+      "epoch": 2.89120151371807,
+      "grad_norm": 2.2124130725860596,
+      "learning_rate": 6.782662734341012e-08,
+      "loss": 1.0599,
+      "step": 3820
+    },
+    {
+      "epoch": 2.891958372753075,
+      "grad_norm": 2.1409800052642822,
+      "learning_rate": 6.691018818820837e-08,
+      "loss": 1.0525,
+      "step": 3821
+    },
+    {
+      "epoch": 2.8927152317880793,
+      "grad_norm": 2.1800687313079834,
+      "learning_rate": 6.600015239579959e-08,
+      "loss": 1.0602,
+      "step": 3822
+    },
+    {
+      "epoch": 2.893472090823084,
+      "grad_norm": 2.0903069972991943,
+      "learning_rate": 6.50965205527814e-08,
+      "loss": 1.0851,
+      "step": 3823
+    },
+    {
+      "epoch": 2.894228949858089,
+      "grad_norm": 1.9317938089370728,
+      "learning_rate": 6.419929324162068e-08,
+      "loss": 1.0319,
+      "step": 3824
+    },
+    {
+      "epoch": 2.8949858088930935,
+      "grad_norm": 2.0327014923095703,
+      "learning_rate": 6.330847104065472e-08,
+      "loss": 1.1128,
+      "step": 3825
+    },
+    {
+      "epoch": 2.8957426679280984,
+      "grad_norm": 2.1695809364318848,
+      "learning_rate": 6.242405452409559e-08,
+      "loss": 1.0591,
+      "step": 3826
+    },
+    {
+      "epoch": 2.896499526963103,
+      "grad_norm": 2.077954053878784,
+      "learning_rate": 6.154604426202468e-08,
+      "loss": 1.0295,
+      "step": 3827
+    },
+    {
+      "epoch": 2.8972563859981078,
+      "grad_norm": 2.0263519287109375,
+      "learning_rate": 6.067444082039482e-08,
+      "loss": 1.0147,
+      "step": 3828
+    },
+    {
+      "epoch": 2.8980132450331126,
+      "grad_norm": 2.1431772708892822,
+      "learning_rate": 5.980924476102595e-08,
+      "loss": 1.0512,
+      "step": 3829
+    },
+    {
+      "epoch": 2.8987701040681175,
+      "grad_norm": 1.9561032056808472,
+      "learning_rate": 5.895045664161168e-08,
+      "loss": 1.0426,
+      "step": 3830
+    },
+    {
+      "epoch": 2.899526963103122,
+      "grad_norm": 2.133995532989502,
+      "learning_rate": 5.8098077015713814e-08,
+      "loss": 1.0365,
+      "step": 3831
+    },
+    {
+      "epoch": 2.900283822138127,
+      "grad_norm": 2.5247886180877686,
+      "learning_rate": 5.7252106432762304e-08,
+      "loss": 1.1153,
+      "step": 3832
+    },
+    {
+      "epoch": 2.9010406811731313,
+      "grad_norm": 1.9548890590667725,
+      "learning_rate": 5.6412545438057476e-08,
+      "loss": 1.0739,
+      "step": 3833
+    },
+    {
+      "epoch": 2.901797540208136,
+      "grad_norm": 1.98203444480896,
+      "learning_rate": 5.557939457276783e-08,
+      "loss": 1.0844,
+      "step": 3834
+    },
+    {
+      "epoch": 2.902554399243141,
+      "grad_norm": 2.1283376216888428,
+      "learning_rate": 5.475265437393116e-08,
+      "loss": 1.0675,
+      "step": 3835
+    },
+    {
+      "epoch": 2.903311258278146,
+      "grad_norm": 1.917360782623291,
+      "learning_rate": 5.393232537444783e-08,
+      "loss": 1.0464,
+      "step": 3836
+    },
+    {
+      "epoch": 2.9040681173131504,
+      "grad_norm": 1.9345555305480957,
+      "learning_rate": 5.3118408103091954e-08,
+      "loss": 1.0937,
+      "step": 3837
+    },
+    {
+      "epoch": 2.9048249763481553,
+      "grad_norm": 2.1080758571624756,
+      "learning_rate": 5.2310903084502445e-08,
+      "loss": 1.0984,
+      "step": 3838
+    },
+    {
+      "epoch": 2.9055818353831597,
+      "grad_norm": 2.0316121578216553,
+      "learning_rate": 5.150981083918309e-08,
+      "loss": 1.105,
+      "step": 3839
+    },
+    {
+      "epoch": 2.9063386944181646,
+      "grad_norm": 2.1214966773986816,
+      "learning_rate": 5.0715131883506914e-08,
+      "loss": 1.0481,
+      "step": 3840
+    },
+    {
+      "epoch": 2.9070955534531695,
+      "grad_norm": 2.003058433532715,
+      "learning_rate": 4.99268667297129e-08,
+      "loss": 1.0848,
+      "step": 3841
+    },
+    {
+      "epoch": 2.907852412488174,
+      "grad_norm": 2.0405402183532715,
+      "learning_rate": 4.9145015885902656e-08,
+      "loss": 1.1065,
+      "step": 3842
+    },
+    {
+      "epoch": 2.908609271523179,
+      "grad_norm": 2.1864330768585205,
+      "learning_rate": 4.836957985604592e-08,
+      "loss": 1.1217,
+      "step": 3843
+    },
+    {
+      "epoch": 2.9093661305581833,
+      "grad_norm": 1.8275071382522583,
+      "learning_rate": 4.7600559139976164e-08,
+      "loss": 1.0634,
+      "step": 3844
+    },
+    {
+      "epoch": 2.910122989593188,
+      "grad_norm": 2.006591320037842,
+      "learning_rate": 4.683795423339395e-08,
+      "loss": 1.0702,
+      "step": 3845
+    },
+    {
+      "epoch": 2.910879848628193,
+      "grad_norm": 2.4923205375671387,
+      "learning_rate": 4.608176562786352e-08,
+      "loss": 1.0633,
+      "step": 3846
+    },
+    {
+      "epoch": 2.911636707663198,
+      "grad_norm": 1.8209044933319092,
+      "learning_rate": 4.533199381080951e-08,
+      "loss": 1.0856,
+      "step": 3847
+    },
+    {
+      "epoch": 2.9123935666982024,
+      "grad_norm": 2.0942399501800537,
+      "learning_rate": 4.458863926552586e-08,
+      "loss": 1.0936,
+      "step": 3848
+    },
+    {
+      "epoch": 2.9131504257332073,
+      "grad_norm": 2.149657964706421,
+      "learning_rate": 4.385170247116687e-08,
+      "loss": 1.0472,
+      "step": 3849
+    },
+    {
+      "epoch": 2.9139072847682117,
+      "grad_norm": 1.905176043510437,
+      "learning_rate": 4.3121183902750584e-08,
+      "loss": 1.0524,
+      "step": 3850
+    },
+    {
+      "epoch": 2.9146641438032166,
+      "grad_norm": 2.1164419651031494,
+      "learning_rate": 4.2397084031158755e-08,
+      "loss": 1.0406,
+      "step": 3851
+    },
+    {
+      "epoch": 2.9154210028382215,
+      "grad_norm": 1.955041527748108,
+      "learning_rate": 4.1679403323133525e-08,
+      "loss": 1.0788,
+      "step": 3852
+    },
+    {
+      "epoch": 2.9161778618732264,
+      "grad_norm": 2.0062403678894043,
+      "learning_rate": 4.096814224128301e-08,
+      "loss": 1.1013,
+      "step": 3853
+    },
+    {
+      "epoch": 2.916934720908231,
+      "grad_norm": 2.1404199600219727,
+      "learning_rate": 4.0263301244073465e-08,
+      "loss": 1.1179,
+      "step": 3854
+    },
+    {
+      "epoch": 2.9176915799432357,
+      "grad_norm": 1.968444585800171,
+      "learning_rate": 3.9564880785834875e-08,
+      "loss": 1.0376,
+      "step": 3855
+    },
+    {
+      "epoch": 2.91844843897824,
+      "grad_norm": 1.9066696166992188,
+      "learning_rate": 3.887288131676096e-08,
+      "loss": 1.0348,
+      "step": 3856
+    },
+    {
+      "epoch": 2.919205298013245,
+      "grad_norm": 2.04758620262146,
+      "learning_rate": 3.818730328290026e-08,
+      "loss": 1.0696,
+      "step": 3857
+    },
+    {
+      "epoch": 2.91996215704825,
+      "grad_norm": 1.8712373971939087,
+      "learning_rate": 3.750814712616839e-08,
+      "loss": 1.0559,
+      "step": 3858
+    },
+    {
+      "epoch": 2.9207190160832543,
+      "grad_norm": 1.9164494276046753,
+      "learning_rate": 3.6835413284338016e-08,
+      "loss": 1.0717,
+      "step": 3859
+    },
+    {
+      "epoch": 2.9214758751182592,
+      "grad_norm": 1.9720449447631836,
+      "learning_rate": 3.616910219104442e-08,
+      "loss": 1.1104,
+      "step": 3860
+    },
+    {
+      "epoch": 2.9222327341532637,
+      "grad_norm": 2.1905975341796875,
+      "learning_rate": 3.5509214275779944e-08,
+      "loss": 1.1058,
+      "step": 3861
+    },
+    {
+      "epoch": 2.9229895931882686,
+      "grad_norm": 1.912367582321167,
+      "learning_rate": 3.4855749963898434e-08,
+      "loss": 1.0694,
+      "step": 3862
+    },
+    {
+      "epoch": 2.9237464522232735,
+      "grad_norm": 2.054760456085205,
+      "learning_rate": 3.420870967661412e-08,
+      "loss": 1.0661,
+      "step": 3863
+    },
+    {
+      "epoch": 2.9245033112582783,
+      "grad_norm": 2.100724220275879,
+      "learning_rate": 3.3568093830998316e-08,
+      "loss": 1.0685,
+      "step": 3864
+    },
+    {
+      "epoch": 2.925260170293283,
+      "grad_norm": 1.9942377805709839,
+      "learning_rate": 3.2933902839982706e-08,
+      "loss": 1.0362,
+      "step": 3865
+    },
+    {
+      "epoch": 2.9260170293282877,
+      "grad_norm": 1.9845491647720337,
+      "learning_rate": 3.230613711235715e-08,
+      "loss": 1.0948,
+      "step": 3866
+    },
+    {
+      "epoch": 2.926773888363292,
+      "grad_norm": 1.9502067565917969,
+      "learning_rate": 3.168479705276969e-08,
+      "loss": 1.0292,
+      "step": 3867
+    },
+    {
+      "epoch": 2.927530747398297,
+      "grad_norm": 2.2125320434570312,
+      "learning_rate": 3.106988306172764e-08,
+      "loss": 1.0965,
+      "step": 3868
+    },
+    {
+      "epoch": 2.928287606433302,
+      "grad_norm": 2.0215206146240234,
+      "learning_rate": 3.046139553559317e-08,
+      "loss": 1.0451,
+      "step": 3869
+    },
+    {
+      "epoch": 2.9290444654683068,
+      "grad_norm": 2.161459445953369,
+      "learning_rate": 2.985933486658992e-08,
+      "loss": 1.0826,
+      "step": 3870
+    },
+    {
+      "epoch": 2.929801324503311,
+      "grad_norm": 2.112816572189331,
+      "learning_rate": 2.926370144279531e-08,
+      "loss": 1.1162,
+      "step": 3871
+    },
+    {
+      "epoch": 2.930558183538316,
+      "grad_norm": 2.043856382369995,
+      "learning_rate": 2.8674495648147115e-08,
+      "loss": 1.0377,
+      "step": 3872
+    },
+    {
+      "epoch": 2.9313150425733205,
+      "grad_norm": 2.230227470397949,
+      "learning_rate": 2.809171786243685e-08,
+      "loss": 1.0775,
+      "step": 3873
+    },
+    {
+      "epoch": 2.9320719016083254,
+      "grad_norm": 1.8375619649887085,
+      "learning_rate": 2.7515368461316434e-08,
+      "loss": 1.0217,
+      "step": 3874
+    },
+    {
+      "epoch": 2.9328287606433303,
+      "grad_norm": 1.9141755104064941,
+      "learning_rate": 2.694544781629039e-08,
+      "loss": 1.038,
+      "step": 3875
+    },
+    {
+      "epoch": 2.9335856196783348,
+      "grad_norm": 2.1602697372436523,
+      "learning_rate": 2.6381956294720323e-08,
+      "loss": 1.0714,
+      "step": 3876
+    },
+    {
+      "epoch": 2.9343424787133396,
+      "grad_norm": 1.923949956893921,
+      "learning_rate": 2.5824894259825987e-08,
+      "loss": 1.0561,
+      "step": 3877
+    },
+    {
+      "epoch": 2.9350993377483445,
+      "grad_norm": 2.1230146884918213,
+      "learning_rate": 2.5274262070678672e-08,
+      "loss": 1.0689,
+      "step": 3878
+    },
+    {
+      "epoch": 2.935856196783349,
+      "grad_norm": 2.0640814304351807,
+      "learning_rate": 2.4730060082210033e-08,
+      "loss": 1.0851,
+      "step": 3879
+    },
+    {
+      "epoch": 2.936613055818354,
+      "grad_norm": 1.8468024730682373,
+      "learning_rate": 2.4192288645203268e-08,
+      "loss": 1.0588,
+      "step": 3880
+    },
+    {
+      "epoch": 2.9373699148533587,
+      "grad_norm": 2.06715726852417,
+      "learning_rate": 2.3660948106297502e-08,
+      "loss": 1.04,
+      "step": 3881
+    },
+    {
+      "epoch": 2.938126773888363,
+      "grad_norm": 2.129422426223755,
+      "learning_rate": 2.313603880798671e-08,
+      "loss": 1.0935,
+      "step": 3882
+    },
+    {
+      "epoch": 2.938883632923368,
+      "grad_norm": 1.9994871616363525,
+      "learning_rate": 2.2617561088619707e-08,
+      "loss": 1.0823,
+      "step": 3883
+    },
+    {
+      "epoch": 2.9396404919583725,
+      "grad_norm": 1.9387072324752808,
+      "learning_rate": 2.2105515282399045e-08,
+      "loss": 1.0506,
+      "step": 3884
+    },
+    {
+      "epoch": 2.9403973509933774,
+      "grad_norm": 1.9773590564727783,
+      "learning_rate": 2.1599901719382117e-08,
+      "loss": 1.0713,
+      "step": 3885
+    },
+    {
+      "epoch": 2.9411542100283823,
+      "grad_norm": 1.9165699481964111,
+      "learning_rate": 2.110072072547893e-08,
+      "loss": 1.0889,
+      "step": 3886
+    },
+    {
+      "epoch": 2.941911069063387,
+      "grad_norm": 1.9767038822174072,
+      "learning_rate": 2.060797262245434e-08,
+      "loss": 1.1121,
+      "step": 3887
+    },
+    {
+      "epoch": 2.9426679280983916,
+      "grad_norm": 1.9442821741104126,
+      "learning_rate": 2.012165772792693e-08,
+      "loss": 1.0852,
+      "step": 3888
+    },
+    {
+      "epoch": 2.9434247871333965,
+      "grad_norm": 1.9988024234771729,
+      "learning_rate": 1.96417763553668e-08,
+      "loss": 1.0733,
+      "step": 3889
+    },
+    {
+      "epoch": 2.944181646168401,
+      "grad_norm": 2.0853540897369385,
+      "learning_rate": 1.91683288141e-08,
+      "loss": 1.0869,
+      "step": 3890
+    },
+    {
+      "epoch": 2.944938505203406,
+      "grad_norm": 1.8882020711898804,
+      "learning_rate": 1.8701315409300757e-08,
+      "loss": 1.0716,
+      "step": 3891
+    },
+    {
+      "epoch": 2.9456953642384107,
+      "grad_norm": 1.9533286094665527,
+      "learning_rate": 1.8240736442000363e-08,
+      "loss": 1.0977,
+      "step": 3892
+    },
+    {
+      "epoch": 2.946452223273415,
+      "grad_norm": 2.2869935035705566,
+      "learning_rate": 1.7786592209081624e-08,
+      "loss": 1.036,
+      "step": 3893
+    },
+    {
+      "epoch": 2.94720908230842,
+      "grad_norm": 1.8621643781661987,
+      "learning_rate": 1.733888300327774e-08,
+      "loss": 1.0655,
+      "step": 3894
+    },
+    {
+      "epoch": 2.947965941343425,
+      "grad_norm": 2.069187641143799,
+      "learning_rate": 1.689760911317565e-08,
+      "loss": 1.0701,
+      "step": 3895
+    },
+    {
+      "epoch": 2.9487228003784294,
+      "grad_norm": 2.112271547317505,
+      "learning_rate": 1.64627708232138e-08,
+      "loss": 1.0697,
+      "step": 3896
+    },
+    {
+      "epoch": 2.9494796594134343,
+      "grad_norm": 2.0788121223449707,
+      "learning_rate": 1.6034368413683266e-08,
+      "loss": 1.0746,
+      "step": 3897
+    },
+    {
+      "epoch": 2.950236518448439,
+      "grad_norm": 2.0578696727752686,
+      "learning_rate": 1.56124021607244e-08,
+      "loss": 1.0702,
+      "step": 3898
+    },
+    {
+      "epoch": 2.9509933774834436,
+      "grad_norm": 2.171917676925659,
+      "learning_rate": 1.519687233633019e-08,
+      "loss": 1.0755,
+      "step": 3899
+    },
+    {
+      "epoch": 2.9517502365184485,
+      "grad_norm": 1.9235490560531616,
+      "learning_rate": 1.4787779208345125e-08,
+      "loss": 1.0209,
+      "step": 3900
+    },
+    {
+      "epoch": 2.952507095553453,
+      "grad_norm": 2.046241283416748,
+      "learning_rate": 1.4385123040465213e-08,
+      "loss": 1.0691,
+      "step": 3901
+    },
+    {
+      "epoch": 2.953263954588458,
+      "grad_norm": 2.0865299701690674,
+      "learning_rate": 1.398890409223575e-08,
+      "loss": 1.0399,
+      "step": 3902
+    },
+    {
+      "epoch": 2.9540208136234627,
+      "grad_norm": 2.0731747150421143,
+      "learning_rate": 1.3599122619053542e-08,
+      "loss": 1.0499,
+      "step": 3903
+    },
+    {
+      "epoch": 2.9547776726584676,
+      "grad_norm": 2.198157548904419,
+      "learning_rate": 1.32157788721658e-08,
+      "loss": 1.0247,
+      "step": 3904
+    },
+    {
+      "epoch": 2.955534531693472,
+      "grad_norm": 2.0734500885009766,
+      "learning_rate": 1.2838873098669024e-08,
+      "loss": 1.0417,
+      "step": 3905
+    },
+    {
+      "epoch": 2.956291390728477,
+      "grad_norm": 1.8891007900238037,
+      "learning_rate": 1.2468405541513447e-08,
+      "loss": 1.0541,
+      "step": 3906
+    },
+    {
+      "epoch": 2.9570482497634814,
+      "grad_norm": 1.9352359771728516,
+      "learning_rate": 1.210437643949415e-08,
+      "loss": 1.0932,
+      "step": 3907
+    },
+    {
+      "epoch": 2.9578051087984862,
+      "grad_norm": 1.9413546323776245,
+      "learning_rate": 1.1746786027259944e-08,
+      "loss": 1.082,
+      "step": 3908
+    },
+    {
+      "epoch": 2.958561967833491,
+      "grad_norm": 2.091618299484253,
+      "learning_rate": 1.1395634535308943e-08,
+      "loss": 1.0216,
+      "step": 3909
+    },
+    {
+      "epoch": 2.959318826868496,
+      "grad_norm": 2.132253408432007,
+      "learning_rate": 1.1050922189986316e-08,
+      "loss": 1.0842,
+      "step": 3910
+    },
+    {
+      "epoch": 2.9600756859035005,
+      "grad_norm": 2.1053178310394287,
+      "learning_rate": 1.0712649213489865e-08,
+      "loss": 1.0316,
+      "step": 3911
+    },
+    {
+      "epoch": 2.9608325449385053,
+      "grad_norm": 2.1205570697784424,
+      "learning_rate": 1.0380815823864458e-08,
+      "loss": 1.065,
+      "step": 3912
+    },
+    {
+      "epoch": 2.96158940397351,
+      "grad_norm": 2.1178319454193115,
+      "learning_rate": 1.0055422235004254e-08,
+      "loss": 1.0666,
+      "step": 3913
+    },
+    {
+      "epoch": 2.9623462630085147,
+      "grad_norm": 2.021894693374634,
+      "learning_rate": 9.736468656653818e-09,
+      "loss": 1.0567,
+      "step": 3914
+    },
+    {
+      "epoch": 2.9631031220435196,
+      "grad_norm": 1.921276569366455,
+      "learning_rate": 9.423955294405891e-09,
+      "loss": 1.093,
+      "step": 3915
+    },
+    {
+      "epoch": 2.963859981078524,
+      "grad_norm": 2.062957286834717,
+      "learning_rate": 9.117882349702507e-09,
+      "loss": 1.0449,
+      "step": 3916
+    },
+    {
+      "epoch": 2.964616840113529,
+      "grad_norm": 2.258112668991089,
+      "learning_rate": 8.818250019831662e-09,
+      "loss": 1.0587,
+      "step": 3917
+    },
+    {
+      "epoch": 2.9653736991485333,
+      "grad_norm": 1.8993000984191895,
+      "learning_rate": 8.52505849793286e-09,
+      "loss": 1.0402,
+      "step": 3918
+    },
+    {
+      "epoch": 2.966130558183538,
+      "grad_norm": 1.8386201858520508,
+      "learning_rate": 8.23830797299268e-09,
+      "loss": 1.1062,
+      "step": 3919
+    },
+    {
+      "epoch": 2.966887417218543,
+      "grad_norm": 2.060410976409912,
+      "learning_rate": 7.957998629846991e-09,
+      "loss": 1.066,
+      "step": 3920
+    },
+    {
+      "epoch": 2.967644276253548,
+      "grad_norm": 2.098123073577881,
+      "learning_rate": 7.684130649177623e-09,
+      "loss": 1.0881,
+      "step": 3921
+    },
+    {
+      "epoch": 2.9684011352885524,
+      "grad_norm": 2.2169816493988037,
+      "learning_rate": 7.416704207515695e-09,
+      "loss": 1.0661,
+      "step": 3922
+    },
+    {
+      "epoch": 2.9691579943235573,
+      "grad_norm": 2.0316176414489746,
+      "learning_rate": 7.155719477241619e-09,
+      "loss": 1.0737,
+      "step": 3923
+    },
+    {
+      "epoch": 2.9699148533585618,
+      "grad_norm": 1.8836135864257812,
+      "learning_rate": 6.901176626581769e-09,
+      "loss": 1.0435,
+      "step": 3924
+    },
+    {
+      "epoch": 2.9706717123935666,
+      "grad_norm": 2.030869960784912,
+      "learning_rate": 6.653075819609588e-09,
+      "loss": 1.0358,
+      "step": 3925
+    },
+    {
+      "epoch": 2.9714285714285715,
+      "grad_norm": 2.1759679317474365,
+      "learning_rate": 6.411417216247812e-09,
+      "loss": 1.066,
+      "step": 3926
+    },
+    {
+      "epoch": 2.9721854304635764,
+      "grad_norm": 2.092773199081421,
+      "learning_rate": 6.176200972265136e-09,
+      "loss": 1.031,
+      "step": 3927
+    },
+    {
+      "epoch": 2.972942289498581,
+      "grad_norm": 2.3539814949035645,
+      "learning_rate": 5.947427239279547e-09,
+      "loss": 1.1136,
+      "step": 3928
+    },
+    {
+      "epoch": 2.9736991485335857,
+      "grad_norm": 2.3484017848968506,
+      "learning_rate": 5.725096164753884e-09,
+      "loss": 1.0145,
+      "step": 3929
+    },
+    {
+      "epoch": 2.97445600756859,
+      "grad_norm": 1.9310166835784912,
+      "learning_rate": 5.509207892001385e-09,
+      "loss": 1.0231,
+      "step": 3930
+    },
+    {
+      "epoch": 2.975212866603595,
+      "grad_norm": 1.9161075353622437,
+      "learning_rate": 5.299762560177922e-09,
+      "loss": 1.0041,
+      "step": 3931
+    },
+    {
+      "epoch": 2.9759697256386,
+      "grad_norm": 2.0112030506134033,
+      "learning_rate": 5.096760304289763e-09,
+      "loss": 1.1227,
+      "step": 3932
+    },
+    {
+      "epoch": 2.9767265846736044,
+      "grad_norm": 2.0244789123535156,
+      "learning_rate": 4.900201255189143e-09,
+      "loss": 1.0922,
+      "step": 3933
+    },
+    {
+      "epoch": 2.9774834437086093,
+      "grad_norm": 2.514658212661743,
+      "learning_rate": 4.710085539575363e-09,
+      "loss": 1.054,
+      "step": 3934
+    },
+    {
+      "epoch": 2.9782403027436137,
+      "grad_norm": 1.9655340909957886,
+      "learning_rate": 4.526413279993689e-09,
+      "loss": 1.0453,
+      "step": 3935
+    },
+    {
+      "epoch": 2.9789971617786186,
+      "grad_norm": 1.988950490951538,
+      "learning_rate": 4.349184594836453e-09,
+      "loss": 1.112,
+      "step": 3936
+    },
+    {
+      "epoch": 2.9797540208136235,
+      "grad_norm": 2.0003416538238525,
+      "learning_rate": 4.178399598341953e-09,
+      "loss": 1.0888,
+      "step": 3937
+    },
+    {
+      "epoch": 2.9805108798486284,
+      "grad_norm": 1.8892840147018433,
+      "learning_rate": 4.014058400597776e-09,
+      "loss": 1.0316,
+      "step": 3938
+    },
+    {
+      "epoch": 2.981267738883633,
+      "grad_norm": 1.8778574466705322,
+      "learning_rate": 3.856161107533029e-09,
+      "loss": 1.0264,
+      "step": 3939
+    },
+    {
+      "epoch": 2.9820245979186377,
+      "grad_norm": 1.9889436960220337,
+      "learning_rate": 3.70470782092722e-09,
+      "loss": 1.0775,
+      "step": 3940
+    },
+    {
+      "epoch": 2.982781456953642,
+      "grad_norm": 1.9453630447387695,
+      "learning_rate": 3.55969863840471e-09,
+      "loss": 1.0661,
+      "step": 3941
+    },
+    {
+      "epoch": 2.983538315988647,
+      "grad_norm": 2.1750118732452393,
+      "learning_rate": 3.421133653436929e-09,
+      "loss": 1.0582,
+      "step": 3942
+    },
+    {
+      "epoch": 2.984295175023652,
+      "grad_norm": 1.9517415761947632,
+      "learning_rate": 3.289012955339048e-09,
+      "loss": 1.061,
+      "step": 3943
+    },
+    {
+      "epoch": 2.985052034058657,
+      "grad_norm": 1.9353458881378174,
+      "learning_rate": 3.16333662927553e-09,
+      "loss": 1.0533,
+      "step": 3944
+    },
+    {
+      "epoch": 2.9858088930936613,
+      "grad_norm": 2.1572377681732178,
+      "learning_rate": 3.044104756254578e-09,
+      "loss": 1.0871,
+      "step": 3945
+    },
+    {
+      "epoch": 2.986565752128666,
+      "grad_norm": 1.9636902809143066,
+      "learning_rate": 2.9313174131325764e-09,
+      "loss": 1.0751,
+      "step": 3946
+    },
+    {
+      "epoch": 2.9873226111636706,
+      "grad_norm": 1.9330499172210693,
+      "learning_rate": 2.8249746726085392e-09,
+      "loss": 1.0858,
+      "step": 3947
+    },
+    {
+      "epoch": 2.9880794701986755,
+      "grad_norm": 1.8204699754714966,
+      "learning_rate": 2.7250766032307735e-09,
+      "loss": 1.034,
+      "step": 3948
+    },
+    {
+      "epoch": 2.9888363292336804,
+      "grad_norm": 1.9643014669418335,
+      "learning_rate": 2.6316232693913253e-09,
+      "loss": 1.062,
+      "step": 3949
+    },
+    {
+      "epoch": 2.989593188268685,
+      "grad_norm": 2.25469708442688,
+      "learning_rate": 2.544614731329312e-09,
+      "loss": 1.0581,
+      "step": 3950
+    },
+    {
+      "epoch": 2.9903500473036897,
+      "grad_norm": 2.0587730407714844,
+      "learning_rate": 2.464051045128703e-09,
+      "loss": 1.0528,
+      "step": 3951
+    },
+    {
+      "epoch": 2.9911069063386946,
+      "grad_norm": 1.9544748067855835,
+      "learning_rate": 2.389932262720538e-09,
+      "loss": 1.0655,
+      "step": 3952
+    },
+    {
+      "epoch": 2.991863765373699,
+      "grad_norm": 2.1084601879119873,
+      "learning_rate": 2.3222584318784854e-09,
+      "loss": 1.1346,
+      "step": 3953
+    },
+    {
+      "epoch": 2.992620624408704,
+      "grad_norm": 2.051255226135254,
+      "learning_rate": 2.261029596226618e-09,
+      "loss": 1.1014,
+      "step": 3954
+    },
+    {
+      "epoch": 2.993377483443709,
+      "grad_norm": 2.079298734664917,
+      "learning_rate": 2.206245795231637e-09,
+      "loss": 1.092,
+      "step": 3955
+    },
+    {
+      "epoch": 2.9941343424787132,
+      "grad_norm": 1.9083516597747803,
+      "learning_rate": 2.157907064203985e-09,
+      "loss": 1.0385,
+      "step": 3956
+    },
+    {
+      "epoch": 2.994891201513718,
+      "grad_norm": 2.315964460372925,
+      "learning_rate": 2.1160134343056167e-09,
+      "loss": 1.0794,
+      "step": 3957
+    },
+    {
+      "epoch": 2.9956480605487226,
+      "grad_norm": 2.072871685028076,
+      "learning_rate": 2.080564932537786e-09,
+      "loss": 1.0603,
+      "step": 3958
+    },
+    {
+      "epoch": 2.9964049195837275,
+      "grad_norm": 1.996877908706665,
+      "learning_rate": 2.0515615817510374e-09,
+      "loss": 1.066,
+      "step": 3959
+    },
+    {
+      "epoch": 2.9971617786187323,
+      "grad_norm": 2.1000583171844482,
+      "learning_rate": 2.0290034006407686e-09,
+      "loss": 1.0554,
+      "step": 3960
+    },
+    {
+      "epoch": 2.9979186376537372,
+      "grad_norm": 2.0836434364318848,
+      "learning_rate": 2.0128904037472256e-09,
+      "loss": 1.0769,
+      "step": 3961
+    },
+    {
+      "epoch": 2.9986754966887417,
+      "grad_norm": 1.999711275100708,
+      "learning_rate": 2.0032226014555062e-09,
+      "loss": 1.0731,
+      "step": 3962
+    },
+    {
+      "epoch": 2.9994323557237466,
+      "grad_norm": 2.6111867427825928,
+      "learning_rate": 2e-09,
+      "loss": 1.1002,
+      "step": 3963
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3963,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.729010317829918e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}