diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,27969 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9979836168872085,
+  "eval_steps": 199,
+  "global_step": 3968,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005040322580645161,
+      "grad_norm": 16.202816979237873,
+      "learning_rate": 1e-08,
+      "loss": 0.4122,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005040322580645161,
+      "eval_loss": 0.443972110748291,
+      "eval_runtime": 17.2668,
+      "eval_samples_per_second": 49.517,
+      "eval_steps_per_second": 1.042,
+      "step": 1
+    },
+    {
+      "epoch": 0.0010080645161290322,
+      "grad_norm": 16.205315003883026,
+      "learning_rate": 2e-08,
+      "loss": 0.4183,
+      "step": 2
+    },
+    {
+      "epoch": 0.0015120967741935483,
+      "grad_norm": 15.468979543825476,
+      "learning_rate": 3e-08,
+      "loss": 0.4182,
+      "step": 3
+    },
+    {
+      "epoch": 0.0020161290322580645,
+      "grad_norm": 15.468273739112808,
+      "learning_rate": 4e-08,
+      "loss": 0.402,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025201612903225806,
+      "grad_norm": 15.057464396772035,
+      "learning_rate": 5e-08,
+      "loss": 0.4211,
+      "step": 5
+    },
+    {
+      "epoch": 0.0030241935483870967,
+      "grad_norm": 15.204132593848971,
+      "learning_rate": 6e-08,
+      "loss": 0.4134,
+      "step": 6
+    },
+    {
+      "epoch": 0.003528225806451613,
+      "grad_norm": 17.615532773933918,
+      "learning_rate": 7e-08,
+      "loss": 0.3878,
+      "step": 7
+    },
+    {
+      "epoch": 0.004032258064516129,
+      "grad_norm": 16.340341169724105,
+      "learning_rate": 8e-08,
+      "loss": 0.3982,
+      "step": 8
+    },
+    {
+      "epoch": 0.0045362903225806455,
+      "grad_norm": 15.059609779441438,
+      "learning_rate": 9e-08,
+      "loss": 0.4178,
+      "step": 9
+    },
+    {
+      "epoch": 0.005040322580645161,
+      "grad_norm": 17.296721168903797,
+      "learning_rate": 1e-07,
+      "loss": 0.4115,
+      "step": 10
+    },
+    {
+      "epoch": 0.005544354838709678,
+      "grad_norm": 17.3187697691886,
+      "learning_rate": 1.0999999999999999e-07,
+      "loss": 0.4104,
+      "step": 11
+    },
+    {
+      "epoch": 0.006048387096774193,
+      "grad_norm": 18.030550149312244,
+      "learning_rate": 1.2e-07,
+      "loss": 0.4104,
+      "step": 12
+    },
+    {
+      "epoch": 0.00655241935483871,
+      "grad_norm": 14.002199418274493,
+      "learning_rate": 1.3e-07,
+      "loss": 0.4047,
+      "step": 13
+    },
+    {
+      "epoch": 0.007056451612903226,
+      "grad_norm": 153.88554829319122,
+      "learning_rate": 1.4e-07,
+      "loss": 0.3935,
+      "step": 14
+    },
+    {
+      "epoch": 0.007560483870967742,
+      "grad_norm": 6.155317751584709,
+      "learning_rate": 1.5e-07,
+      "loss": 0.3934,
+      "step": 15
+    },
+    {
+      "epoch": 0.008064516129032258,
+      "grad_norm": 8.516052807530963,
+      "learning_rate": 1.6e-07,
+      "loss": 0.3992,
+      "step": 16
+    },
+    {
+      "epoch": 0.008568548387096774,
+      "grad_norm": 12.572451947505744,
+      "learning_rate": 1.7000000000000001e-07,
+      "loss": 0.4091,
+      "step": 17
+    },
+    {
+      "epoch": 0.009072580645161291,
+      "grad_norm": 26.666526989945357,
+      "learning_rate": 1.8e-07,
+      "loss": 0.4194,
+      "step": 18
+    },
+    {
+      "epoch": 0.009576612903225807,
+      "grad_norm": 43.96981776742769,
+      "learning_rate": 1.8999999999999998e-07,
+      "loss": 0.4073,
+      "step": 19
+    },
+    {
+      "epoch": 0.010080645161290322,
+      "grad_norm": 58.38481172991244,
+      "learning_rate": 2e-07,
+      "loss": 0.4135,
+      "step": 20
+    },
+    {
+      "epoch": 0.010584677419354838,
+      "grad_norm": 8.92743387072961,
+      "learning_rate": 2.0999999999999997e-07,
+      "loss": 0.4233,
+      "step": 21
+    },
+    {
+      "epoch": 0.011088709677419355,
+      "grad_norm": 7.114049513845685,
+      "learning_rate": 2.1999999999999998e-07,
+      "loss": 0.3915,
+      "step": 22
+    },
+    {
+      "epoch": 0.011592741935483871,
+      "grad_norm": 7.5930944951978825,
+      "learning_rate": 2.3e-07,
+      "loss": 0.4119,
+      "step": 23
+    },
+    {
+      "epoch": 0.012096774193548387,
+      "grad_norm": 17.847773187914957,
+      "learning_rate": 2.4e-07,
+      "loss": 0.4035,
+      "step": 24
+    },
+    {
+      "epoch": 0.012600806451612902,
+      "grad_norm": 14.342646251100152,
+      "learning_rate": 2.5e-07,
+      "loss": 0.4023,
+      "step": 25
+    },
+    {
+      "epoch": 0.01310483870967742,
+      "grad_norm": 5.653916048149905,
+      "learning_rate": 2.6e-07,
+      "loss": 0.3999,
+      "step": 26
+    },
+    {
+      "epoch": 0.013608870967741936,
+      "grad_norm": 4.978130519093393,
+      "learning_rate": 2.7e-07,
+      "loss": 0.4012,
+      "step": 27
+    },
+    {
+      "epoch": 0.014112903225806451,
+      "grad_norm": 5.1928104536034345,
+      "learning_rate": 2.8e-07,
+      "loss": 0.3835,
+      "step": 28
+    },
+    {
+      "epoch": 0.014616935483870967,
+      "grad_norm": 5.713989015081132,
+      "learning_rate": 2.9e-07,
+      "loss": 0.4207,
+      "step": 29
+    },
+    {
+      "epoch": 0.015120967741935484,
+      "grad_norm": 10.819751935660008,
+      "learning_rate": 3e-07,
+      "loss": 0.3895,
+      "step": 30
+    },
+    {
+      "epoch": 0.015625,
+      "grad_norm": 9.415309461004313,
+      "learning_rate": 3.1e-07,
+      "loss": 0.4104,
+      "step": 31
+    },
+    {
+      "epoch": 0.016129032258064516,
+      "grad_norm": 3.283165538142218,
+      "learning_rate": 3.2e-07,
+      "loss": 0.4182,
+      "step": 32
+    },
+    {
+      "epoch": 0.01663306451612903,
+      "grad_norm": 6.31824176396841,
+      "learning_rate": 3.3e-07,
+      "loss": 0.3982,
+      "step": 33
+    },
+    {
+      "epoch": 0.017137096774193547,
+      "grad_norm": 2.471098407656219,
+      "learning_rate": 3.4000000000000003e-07,
+      "loss": 0.4051,
+      "step": 34
+    },
+    {
+      "epoch": 0.017641129032258066,
+      "grad_norm": 3.7875313205613943,
+      "learning_rate": 3.5e-07,
+      "loss": 0.3897,
+      "step": 35
+    },
+    {
+      "epoch": 0.018145161290322582,
+      "grad_norm": 21.98400784558835,
+      "learning_rate": 3.6e-07,
+      "loss": 0.398,
+      "step": 36
+    },
+    {
+      "epoch": 0.018649193548387098,
+      "grad_norm": 2.5064868793159714,
+      "learning_rate": 3.7e-07,
+      "loss": 0.3951,
+      "step": 37
+    },
+    {
+      "epoch": 0.019153225806451613,
+      "grad_norm": 24.059819417771145,
+      "learning_rate": 3.7999999999999996e-07,
+      "loss": 0.3989,
+      "step": 38
+    },
+    {
+      "epoch": 0.01965725806451613,
+      "grad_norm": 6.95393911781356,
+      "learning_rate": 3.8999999999999997e-07,
+      "loss": 0.4014,
+      "step": 39
+    },
+    {
+      "epoch": 0.020161290322580645,
+      "grad_norm": 2.8868850874278182,
+      "learning_rate": 4e-07,
+      "loss": 0.3864,
+      "step": 40
+    },
+    {
+      "epoch": 0.02066532258064516,
+      "grad_norm": 2.1387916921557624,
+      "learning_rate": 4.0999999999999994e-07,
+      "loss": 0.3765,
+      "step": 41
+    },
+    {
+      "epoch": 0.021169354838709676,
+      "grad_norm": 2.627363698866213,
+      "learning_rate": 4.1999999999999995e-07,
+      "loss": 0.3819,
+      "step": 42
+    },
+    {
+      "epoch": 0.021673387096774195,
+      "grad_norm": 4.399523991039246,
+      "learning_rate": 4.2999999999999996e-07,
+      "loss": 0.396,
+      "step": 43
+    },
+    {
+      "epoch": 0.02217741935483871,
+      "grad_norm": 6.2743195319461265,
+      "learning_rate": 4.3999999999999997e-07,
+      "loss": 0.3719,
+      "step": 44
+    },
+    {
+      "epoch": 0.022681451612903226,
+      "grad_norm": 3.5237571635789706,
+      "learning_rate": 4.5e-07,
+      "loss": 0.3787,
+      "step": 45
+    },
+    {
+      "epoch": 0.023185483870967742,
+      "grad_norm": 4.8083225470670845,
+      "learning_rate": 4.6e-07,
+      "loss": 0.3842,
+      "step": 46
+    },
+    {
+      "epoch": 0.023689516129032258,
+      "grad_norm": 4.2342259419992985,
+      "learning_rate": 4.6999999999999995e-07,
+      "loss": 0.3921,
+      "step": 47
+    },
+    {
+      "epoch": 0.024193548387096774,
+      "grad_norm": 5.233597040220851,
+      "learning_rate": 4.8e-07,
+      "loss": 0.3803,
+      "step": 48
+    },
+    {
+      "epoch": 0.02469758064516129,
+      "grad_norm": 3.3462375668312427,
+      "learning_rate": 4.9e-07,
+      "loss": 0.3762,
+      "step": 49
+    },
+    {
+      "epoch": 0.025201612903225805,
+      "grad_norm": 5.738883941273613,
+      "learning_rate": 5e-07,
+      "loss": 0.3691,
+      "step": 50
+    },
+    {
+      "epoch": 0.025705645161290324,
+      "grad_norm": 3.4906691092581608,
+      "learning_rate": 5.1e-07,
+      "loss": 0.3927,
+      "step": 51
+    },
+    {
+      "epoch": 0.02620967741935484,
+      "grad_norm": 3.1684109647850445,
+      "learning_rate": 5.2e-07,
+      "loss": 0.3649,
+      "step": 52
+    },
+    {
+      "epoch": 0.026713709677419355,
+      "grad_norm": 2.1354179898716037,
+      "learning_rate": 5.3e-07,
+      "loss": 0.3982,
+      "step": 53
+    },
+    {
+      "epoch": 0.02721774193548387,
+      "grad_norm": 9.732639588208007,
+      "learning_rate": 5.4e-07,
+      "loss": 0.3897,
+      "step": 54
+    },
+    {
+      "epoch": 0.027721774193548387,
+      "grad_norm": 1.114588852331627,
+      "learning_rate": 5.5e-07,
+      "loss": 0.3668,
+      "step": 55
+    },
+    {
+      "epoch": 0.028225806451612902,
+      "grad_norm": 6.114641493595002,
+      "learning_rate": 5.6e-07,
+      "loss": 0.377,
+      "step": 56
+    },
+    {
+      "epoch": 0.028729838709677418,
+      "grad_norm": 1.3354617010610528,
+      "learning_rate": 5.699999999999999e-07,
+      "loss": 0.38,
+      "step": 57
+    },
+    {
+      "epoch": 0.029233870967741934,
+      "grad_norm": 1.2850750187972262,
+      "learning_rate": 5.8e-07,
+      "loss": 0.3847,
+      "step": 58
+    },
+    {
+      "epoch": 0.029737903225806453,
+      "grad_norm": 1.3750785124101343,
+      "learning_rate": 5.9e-07,
+      "loss": 0.3592,
+      "step": 59
+    },
+    {
+      "epoch": 0.03024193548387097,
+      "grad_norm": 2.9929928043047975,
+      "learning_rate": 6e-07,
+      "loss": 0.3633,
+      "step": 60
+    },
+    {
+      "epoch": 0.030745967741935484,
+      "grad_norm": 6.301747495454331,
+      "learning_rate": 6.1e-07,
+      "loss": 0.3766,
+      "step": 61
+    },
+    {
+      "epoch": 0.03125,
+      "grad_norm": 2.1168564747094076,
+      "learning_rate": 6.2e-07,
+      "loss": 0.3713,
+      "step": 62
+    },
+    {
+      "epoch": 0.031754032258064516,
+      "grad_norm": 0.7578830680461082,
+      "learning_rate": 6.3e-07,
+      "loss": 0.369,
+      "step": 63
+    },
+    {
+      "epoch": 0.03225806451612903,
+      "grad_norm": 1.6734191696925034,
+      "learning_rate": 6.4e-07,
+      "loss": 0.366,
+      "step": 64
+    },
+    {
+      "epoch": 0.03276209677419355,
+      "grad_norm": 1.2789631524237748,
+      "learning_rate": 6.5e-07,
+      "loss": 0.3617,
+      "step": 65
+    },
+    {
+      "epoch": 0.03326612903225806,
+      "grad_norm": 0.9523361493720094,
+      "learning_rate": 6.6e-07,
+      "loss": 0.3383,
+      "step": 66
+    },
+    {
+      "epoch": 0.03377016129032258,
+      "grad_norm": 1.166610477006102,
+      "learning_rate": 6.7e-07,
+      "loss": 0.3626,
+      "step": 67
+    },
+    {
+      "epoch": 0.034274193548387094,
+      "grad_norm": 1.2998523358147898,
+      "learning_rate": 6.800000000000001e-07,
+      "loss": 0.3521,
+      "step": 68
+    },
+    {
+      "epoch": 0.03477822580645161,
+      "grad_norm": 1.4526130491588676,
+      "learning_rate": 6.9e-07,
+      "loss": 0.3464,
+      "step": 69
+    },
+    {
+      "epoch": 0.03528225806451613,
+      "grad_norm": 0.57853677880428,
+      "learning_rate": 7e-07,
+      "loss": 0.3459,
+      "step": 70
+    },
+    {
+      "epoch": 0.03578629032258065,
+      "grad_norm": 1.268886862963294,
+      "learning_rate": 7.1e-07,
+      "loss": 0.3658,
+      "step": 71
+    },
+    {
+      "epoch": 0.036290322580645164,
+      "grad_norm": 1.3395158890738286,
+      "learning_rate": 7.2e-07,
+      "loss": 0.3382,
+      "step": 72
+    },
+    {
+      "epoch": 0.03679435483870968,
+      "grad_norm": 0.7066691364777967,
+      "learning_rate": 7.3e-07,
+      "loss": 0.3484,
+      "step": 73
+    },
+    {
+      "epoch": 0.037298387096774195,
+      "grad_norm": 0.8562245745506359,
+      "learning_rate": 7.4e-07,
+      "loss": 0.3696,
+      "step": 74
+    },
+    {
+      "epoch": 0.03780241935483871,
+      "grad_norm": 1.583149325798458,
+      "learning_rate": 7.5e-07,
+      "loss": 0.352,
+      "step": 75
+    },
+    {
+      "epoch": 0.038306451612903226,
+      "grad_norm": 0.4718209553213981,
+      "learning_rate": 7.599999999999999e-07,
+      "loss": 0.3474,
+      "step": 76
+    },
+    {
+      "epoch": 0.03881048387096774,
+      "grad_norm": 16.741995431233445,
+      "learning_rate": 7.699999999999999e-07,
+      "loss": 0.3514,
+      "step": 77
+    },
+    {
+      "epoch": 0.03931451612903226,
+      "grad_norm": 0.5609500021734007,
+      "learning_rate": 7.799999999999999e-07,
+      "loss": 0.3562,
+      "step": 78
+    },
+    {
+      "epoch": 0.039818548387096774,
+      "grad_norm": 0.5482085713967506,
+      "learning_rate": 7.9e-07,
+      "loss": 0.3571,
+      "step": 79
+    },
+    {
+      "epoch": 0.04032258064516129,
+      "grad_norm": 0.5862216706734081,
+      "learning_rate": 8e-07,
+      "loss": 0.3775,
+      "step": 80
+    },
+    {
+      "epoch": 0.040826612903225805,
+      "grad_norm": 0.9433271815134637,
+      "learning_rate": 8.1e-07,
+      "loss": 0.3622,
+      "step": 81
+    },
+    {
+      "epoch": 0.04133064516129032,
+      "grad_norm": 2.312304687609339,
+      "learning_rate": 8.199999999999999e-07,
+      "loss": 0.3354,
+      "step": 82
+    },
+    {
+      "epoch": 0.041834677419354836,
+      "grad_norm": 0.7657086859357075,
+      "learning_rate": 8.299999999999999e-07,
+      "loss": 0.3386,
+      "step": 83
+    },
+    {
+      "epoch": 0.04233870967741935,
+      "grad_norm": 0.5693450749645984,
+      "learning_rate": 8.399999999999999e-07,
+      "loss": 0.336,
+      "step": 84
+    },
+    {
+      "epoch": 0.04284274193548387,
+      "grad_norm": 0.7237524467723284,
+      "learning_rate": 8.499999999999999e-07,
+      "loss": 0.3574,
+      "step": 85
+    },
+    {
+      "epoch": 0.04334677419354839,
+      "grad_norm": 1.6741553791841324,
+      "learning_rate": 8.599999999999999e-07,
+      "loss": 0.3678,
+      "step": 86
+    },
+    {
+      "epoch": 0.043850806451612906,
+      "grad_norm": 1.066623540734941,
+      "learning_rate": 8.699999999999999e-07,
+      "loss": 0.3528,
+      "step": 87
+    },
+    {
+      "epoch": 0.04435483870967742,
+      "grad_norm": 0.5236790478758251,
+      "learning_rate": 8.799999999999999e-07,
+      "loss": 0.3485,
+      "step": 88
+    },
+    {
+      "epoch": 0.04485887096774194,
+      "grad_norm": 0.37322566471667024,
+      "learning_rate": 8.9e-07,
+      "loss": 0.3403,
+      "step": 89
+    },
+    {
+      "epoch": 0.04536290322580645,
+      "grad_norm": 0.37334552504540236,
+      "learning_rate": 9e-07,
+      "loss": 0.334,
+      "step": 90
+    },
+    {
+      "epoch": 0.04586693548387097,
+      "grad_norm": 0.6233477368274808,
+      "learning_rate": 9.1e-07,
+      "loss": 0.3577,
+      "step": 91
+    },
+    {
+      "epoch": 0.046370967741935484,
+      "grad_norm": 0.380512676596753,
+      "learning_rate": 9.2e-07,
+      "loss": 0.3437,
+      "step": 92
+    },
+    {
+      "epoch": 0.046875,
+      "grad_norm": 0.7218988883748737,
+      "learning_rate": 9.3e-07,
+      "loss": 0.3622,
+      "step": 93
+    },
+    {
+      "epoch": 0.047379032258064516,
+      "grad_norm": 1.0217493525557952,
+      "learning_rate": 9.399999999999999e-07,
+      "loss": 0.3249,
+      "step": 94
+    },
+    {
+      "epoch": 0.04788306451612903,
+      "grad_norm": 0.8160482348861993,
+      "learning_rate": 9.499999999999999e-07,
+      "loss": 0.3569,
+      "step": 95
+    },
+    {
+      "epoch": 0.04838709677419355,
+      "grad_norm": 1.6620382396716078,
+      "learning_rate": 9.6e-07,
+      "loss": 0.3325,
+      "step": 96
+    },
+    {
+      "epoch": 0.04889112903225806,
+      "grad_norm": 0.9899117085666442,
+      "learning_rate": 9.7e-07,
+      "loss": 0.3565,
+      "step": 97
+    },
+    {
+      "epoch": 0.04939516129032258,
+      "grad_norm": 0.6672181715686732,
+      "learning_rate": 9.8e-07,
+      "loss": 0.3383,
+      "step": 98
+    },
+    {
+      "epoch": 0.049899193548387094,
+      "grad_norm": 0.8490767132728996,
+      "learning_rate": 9.9e-07,
+      "loss": 0.3628,
+      "step": 99
+    },
+    {
+      "epoch": 0.05040322580645161,
+      "grad_norm": 3.230091455346306,
+      "learning_rate": 1e-06,
+      "loss": 0.3525,
+      "step": 100
+    },
+    {
+      "epoch": 0.05090725806451613,
+      "grad_norm": 1.0048400815002736,
+      "learning_rate": 9.999999278765487e-07,
+      "loss": 0.3465,
+      "step": 101
+    },
+    {
+      "epoch": 0.05141129032258065,
+      "grad_norm": 0.46323136632477013,
+      "learning_rate": 9.999997115062153e-07,
+      "loss": 0.3613,
+      "step": 102
+    },
+    {
+      "epoch": 0.051915322580645164,
+      "grad_norm": 1.1303903715708508,
+      "learning_rate": 9.999993508890626e-07,
+      "loss": 0.3439,
+      "step": 103
+    },
+    {
+      "epoch": 0.05241935483870968,
+      "grad_norm": 0.8278300274368537,
+      "learning_rate": 9.999988460251948e-07,
+      "loss": 0.3364,
+      "step": 104
+    },
+    {
+      "epoch": 0.052923387096774195,
+      "grad_norm": 0.8385630360604959,
+      "learning_rate": 9.99998196914757e-07,
+      "loss": 0.3341,
+      "step": 105
+    },
+    {
+      "epoch": 0.05342741935483871,
+      "grad_norm": 0.5181207146190654,
+      "learning_rate": 9.999974035579367e-07,
+      "loss": 0.3504,
+      "step": 106
+    },
+    {
+      "epoch": 0.053931451612903226,
+      "grad_norm": 0.867225679439543,
+      "learning_rate": 9.999964659549629e-07,
+      "loss": 0.3487,
+      "step": 107
+    },
+    {
+      "epoch": 0.05443548387096774,
+      "grad_norm": 0.4359954195764285,
+      "learning_rate": 9.999953841061059e-07,
+      "loss": 0.3511,
+      "step": 108
+    },
+    {
+      "epoch": 0.05493951612903226,
+      "grad_norm": 0.49236276321466904,
+      "learning_rate": 9.99994158011678e-07,
+      "loss": 0.3301,
+      "step": 109
+    },
+    {
+      "epoch": 0.055443548387096774,
+      "grad_norm": 0.42334304009907897,
+      "learning_rate": 9.999927876720327e-07,
+      "loss": 0.3354,
+      "step": 110
+    },
+    {
+      "epoch": 0.05594758064516129,
+      "grad_norm": 0.3717776940463779,
+      "learning_rate": 9.999912730875654e-07,
+      "loss": 0.3536,
+      "step": 111
+    },
+    {
+      "epoch": 0.056451612903225805,
+      "grad_norm": 0.7134857290167043,
+      "learning_rate": 9.999896142587133e-07,
+      "loss": 0.3423,
+      "step": 112
+    },
+    {
+      "epoch": 0.05695564516129032,
+      "grad_norm": 0.6589031582715488,
+      "learning_rate": 9.999878111859545e-07,
+      "loss": 0.3528,
+      "step": 113
+    },
+    {
+      "epoch": 0.057459677419354836,
+      "grad_norm": 0.4715779115727926,
+      "learning_rate": 9.999858638698095e-07,
+      "loss": 0.331,
+      "step": 114
+    },
+    {
+      "epoch": 0.05796370967741935,
+      "grad_norm": 1.8324714803517166,
+      "learning_rate": 9.999837723108403e-07,
+      "loss": 0.3361,
+      "step": 115
+    },
+    {
+      "epoch": 0.05846774193548387,
+      "grad_norm": 3.4504051941803455,
+      "learning_rate": 9.999815365096497e-07,
+      "loss": 0.3381,
+      "step": 116
+    },
+    {
+      "epoch": 0.05897177419354839,
+      "grad_norm": 0.7082367647106664,
+      "learning_rate": 9.999791564668832e-07,
+      "loss": 0.3357,
+      "step": 117
+    },
+    {
+      "epoch": 0.059475806451612906,
+      "grad_norm": 0.8943478209896717,
+      "learning_rate": 9.99976632183227e-07,
+      "loss": 0.334,
+      "step": 118
+    },
+    {
+      "epoch": 0.05997983870967742,
+      "grad_norm": 2.639935863362195,
+      "learning_rate": 9.9997396365941e-07,
+      "loss": 0.3515,
+      "step": 119
+    },
+    {
+      "epoch": 0.06048387096774194,
+      "grad_norm": 0.5612525598880919,
+      "learning_rate": 9.999711508962014e-07,
+      "loss": 0.3421,
+      "step": 120
+    },
+    {
+      "epoch": 0.06098790322580645,
+      "grad_norm": 1.0108383298218186,
+      "learning_rate": 9.99968193894413e-07,
+      "loss": 0.3485,
+      "step": 121
+    },
+    {
+      "epoch": 0.06149193548387097,
+      "grad_norm": 1.5821951581009004,
+      "learning_rate": 9.999650926548979e-07,
+      "loss": 0.3293,
+      "step": 122
+    },
+    {
+      "epoch": 0.061995967741935484,
+      "grad_norm": 1.1703013492954635,
+      "learning_rate": 9.999618471785505e-07,
+      "loss": 0.3725,
+      "step": 123
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 0.3496060502874034,
+      "learning_rate": 9.999584574663074e-07,
+      "loss": 0.3401,
+      "step": 124
+    },
+    {
+      "epoch": 0.06300403225806452,
+      "grad_norm": 0.6132017993310452,
+      "learning_rate": 9.999549235191465e-07,
+      "loss": 0.3455,
+      "step": 125
+    },
+    {
+      "epoch": 0.06350806451612903,
+      "grad_norm": 1.5073872786250602,
+      "learning_rate": 9.999512453380869e-07,
+      "loss": 0.3389,
+      "step": 126
+    },
+    {
+      "epoch": 0.06401209677419355,
+      "grad_norm": 1.539376597083665,
+      "learning_rate": 9.999474229241903e-07,
+      "loss": 0.3332,
+      "step": 127
+    },
+    {
+      "epoch": 0.06451612903225806,
+      "grad_norm": 1.2824746138412246,
+      "learning_rate": 9.99943456278559e-07,
+      "loss": 0.3456,
+      "step": 128
+    },
+    {
+      "epoch": 0.06502016129032258,
+      "grad_norm": 1.4717370768722402,
+      "learning_rate": 9.99939345402338e-07,
+      "loss": 0.3361,
+      "step": 129
+    },
+    {
+      "epoch": 0.0655241935483871,
+      "grad_norm": 0.3607286264558088,
+      "learning_rate": 9.999350902967124e-07,
+      "loss": 0.3309,
+      "step": 130
+    },
+    {
+      "epoch": 0.06602822580645161,
+      "grad_norm": 0.3812650155511185,
+      "learning_rate": 9.999306909629103e-07,
+      "loss": 0.3336,
+      "step": 131
+    },
+    {
+      "epoch": 0.06653225806451613,
+      "grad_norm": 2.886726675491379,
+      "learning_rate": 9.999261474022007e-07,
+      "loss": 0.35,
+      "step": 132
+    },
+    {
+      "epoch": 0.06703629032258064,
+      "grad_norm": 2.482458443393656,
+      "learning_rate": 9.999214596158946e-07,
+      "loss": 0.333,
+      "step": 133
+    },
+    {
+      "epoch": 0.06754032258064516,
+      "grad_norm": 0.3364379109508271,
+      "learning_rate": 9.999166276053442e-07,
+      "loss": 0.3371,
+      "step": 134
+    },
+    {
+      "epoch": 0.06804435483870967,
+      "grad_norm": 0.7506818894597985,
+      "learning_rate": 9.999116513719434e-07,
+      "loss": 0.3339,
+      "step": 135
+    },
+    {
+      "epoch": 0.06854838709677419,
+      "grad_norm": 1.4202719086270377,
+      "learning_rate": 9.999065309171282e-07,
+      "loss": 0.3248,
+      "step": 136
+    },
+    {
+      "epoch": 0.0690524193548387,
+      "grad_norm": 0.7484158287932788,
+      "learning_rate": 9.999012662423754e-07,
+      "loss": 0.3322,
+      "step": 137
+    },
+    {
+      "epoch": 0.06955645161290322,
+      "grad_norm": 0.374995905774015,
+      "learning_rate": 9.998958573492042e-07,
+      "loss": 0.3388,
+      "step": 138
+    },
+    {
+      "epoch": 0.07006048387096774,
+      "grad_norm": 3.1451635192774416,
+      "learning_rate": 9.998903042391747e-07,
+      "loss": 0.3308,
+      "step": 139
+    },
+    {
+      "epoch": 0.07056451612903226,
+      "grad_norm": 4.100760115701697,
+      "learning_rate": 9.99884606913889e-07,
+      "loss": 0.3393,
+      "step": 140
+    },
+    {
+      "epoch": 0.07106854838709678,
+      "grad_norm": 1.0116256210926982,
+      "learning_rate": 9.99878765374991e-07,
+      "loss": 0.3365,
+      "step": 141
+    },
+    {
+      "epoch": 0.0715725806451613,
+      "grad_norm": 0.4898776317688679,
+      "learning_rate": 9.998727796241657e-07,
+      "loss": 0.3388,
+      "step": 142
+    },
+    {
+      "epoch": 0.07207661290322581,
+      "grad_norm": 1.2664512508752508,
+      "learning_rate": 9.9986664966314e-07,
+      "loss": 0.3245,
+      "step": 143
+    },
+    {
+      "epoch": 0.07258064516129033,
+      "grad_norm": 1.6834640389600903,
+      "learning_rate": 9.998603754936825e-07,
+      "loss": 0.3407,
+      "step": 144
+    },
+    {
+      "epoch": 0.07308467741935484,
+      "grad_norm": 1.3464610251528206,
+      "learning_rate": 9.99853957117603e-07,
+      "loss": 0.3225,
+      "step": 145
+    },
+    {
+      "epoch": 0.07358870967741936,
+      "grad_norm": 1.0984235521861139,
+      "learning_rate": 9.998473945367535e-07,
+      "loss": 0.3231,
+      "step": 146
+    },
+    {
+      "epoch": 0.07409274193548387,
+      "grad_norm": 2.8158986365995156,
+      "learning_rate": 9.998406877530267e-07,
+      "loss": 0.3331,
+      "step": 147
+    },
+    {
+      "epoch": 0.07459677419354839,
+      "grad_norm": 0.3788437430061732,
+      "learning_rate": 9.998338367683583e-07,
+      "loss": 0.3392,
+      "step": 148
+    },
+    {
+      "epoch": 0.0751008064516129,
+      "grad_norm": 0.6183083703029865,
+      "learning_rate": 9.99826841584724e-07,
+      "loss": 0.3305,
+      "step": 149
+    },
+    {
+      "epoch": 0.07560483870967742,
+      "grad_norm": 0.4095762625785864,
+      "learning_rate": 9.998197022041422e-07,
+      "loss": 0.3263,
+      "step": 150
+    },
+    {
+      "epoch": 0.07610887096774194,
+      "grad_norm": 1.2756291414954701,
+      "learning_rate": 9.998124186286724e-07,
+      "loss": 0.3394,
+      "step": 151
+    },
+    {
+      "epoch": 0.07661290322580645,
+      "grad_norm": 0.3597641387730504,
+      "learning_rate": 9.998049908604163e-07,
+      "loss": 0.3462,
+      "step": 152
+    },
+    {
+      "epoch": 0.07711693548387097,
+      "grad_norm": 0.3430879414583386,
+      "learning_rate": 9.997974189015163e-07,
+      "loss": 0.337,
+      "step": 153
+    },
+    {
+      "epoch": 0.07762096774193548,
+      "grad_norm": 1.1595831180201008,
+      "learning_rate": 9.997897027541571e-07,
+      "loss": 0.3372,
+      "step": 154
+    },
+    {
+      "epoch": 0.078125,
+      "grad_norm": 0.48618010029983355,
+      "learning_rate": 9.997818424205647e-07,
+      "loss": 0.3371,
+      "step": 155
+    },
+    {
+      "epoch": 0.07862903225806452,
+      "grad_norm": 0.39786757641752357,
+      "learning_rate": 9.997738379030068e-07,
+      "loss": 0.3314,
+      "step": 156
+    },
+    {
+      "epoch": 0.07913306451612903,
+      "grad_norm": 1.7056412415922666,
+      "learning_rate": 9.997656892037924e-07,
+      "loss": 0.319,
+      "step": 157
+    },
+    {
+      "epoch": 0.07963709677419355,
+      "grad_norm": 0.3653360435949616,
+      "learning_rate": 9.997573963252725e-07,
+      "loss": 0.3354,
+      "step": 158
+    },
+    {
+      "epoch": 0.08014112903225806,
+      "grad_norm": 2.829079281128038,
+      "learning_rate": 9.997489592698399e-07,
+      "loss": 0.3318,
+      "step": 159
+    },
+    {
+      "epoch": 0.08064516129032258,
+      "grad_norm": 1.2295812347072044,
+      "learning_rate": 9.997403780399282e-07,
+      "loss": 0.3494,
+      "step": 160
+    },
+    {
+      "epoch": 0.0811491935483871,
+      "grad_norm": 0.8812794930898853,
+      "learning_rate": 9.997316526380131e-07,
+      "loss": 0.3344,
+      "step": 161
+    },
+    {
+      "epoch": 0.08165322580645161,
+      "grad_norm": 0.8353193931880987,
+      "learning_rate": 9.997227830666118e-07,
+      "loss": 0.3305,
+      "step": 162
+    },
+    {
+      "epoch": 0.08215725806451613,
+      "grad_norm": 0.9923756258213962,
+      "learning_rate": 9.997137693282833e-07,
+      "loss": 0.3421,
+      "step": 163
+    },
+    {
+      "epoch": 0.08266129032258064,
+      "grad_norm": 2.476846910057538,
+      "learning_rate": 9.99704611425628e-07,
+      "loss": 0.3303,
+      "step": 164
+    },
+    {
+      "epoch": 0.08316532258064516,
+      "grad_norm": 3.0540963644253294,
+      "learning_rate": 9.996953093612877e-07,
+      "loss": 0.3244,
+      "step": 165
+    },
+    {
+      "epoch": 0.08366935483870967,
+      "grad_norm": 1.9232590922797959,
+      "learning_rate": 9.99685863137946e-07,
+      "loss": 0.3566,
+      "step": 166
+    },
+    {
+      "epoch": 0.08417338709677419,
+      "grad_norm": 0.8404131782552129,
+      "learning_rate": 9.996762727583285e-07,
+      "loss": 0.3248,
+      "step": 167
+    },
+    {
+      "epoch": 0.0846774193548387,
+      "grad_norm": 0.7858824684266251,
+      "learning_rate": 9.996665382252014e-07,
+      "loss": 0.3445,
+      "step": 168
+    },
+    {
+      "epoch": 0.08518145161290322,
+      "grad_norm": 0.7353820488132891,
+      "learning_rate": 9.996566595413734e-07,
+      "loss": 0.3284,
+      "step": 169
+    },
+    {
+      "epoch": 0.08568548387096774,
+      "grad_norm": 0.4062481131054341,
+      "learning_rate": 9.996466367096943e-07,
+      "loss": 0.3204,
+      "step": 170
+    },
+    {
+      "epoch": 0.08618951612903226,
+      "grad_norm": 0.7805121538010222,
+      "learning_rate": 9.996364697330555e-07,
+      "loss": 0.3222,
+      "step": 171
+    },
+    {
+      "epoch": 0.08669354838709678,
+      "grad_norm": 0.45723066112420185,
+      "learning_rate": 9.996261586143904e-07,
+      "loss": 0.3359,
+      "step": 172
+    },
+    {
+      "epoch": 0.0871975806451613,
+      "grad_norm": 0.9178689167005624,
+      "learning_rate": 9.996157033566737e-07,
+      "loss": 0.3446,
+      "step": 173
+    },
+    {
+      "epoch": 0.08770161290322581,
+      "grad_norm": 1.1988895413065521,
+      "learning_rate": 9.996051039629214e-07,
+      "loss": 0.3196,
+      "step": 174
+    },
+    {
+      "epoch": 0.08820564516129033,
+      "grad_norm": 1.5252246353478032,
+      "learning_rate": 9.995943604361915e-07,
+      "loss": 0.3296,
+      "step": 175
+    },
+    {
+      "epoch": 0.08870967741935484,
+      "grad_norm": 2.9977266959256217,
+      "learning_rate": 9.995834727795834e-07,
+      "loss": 0.332,
+      "step": 176
+    },
+    {
+      "epoch": 0.08921370967741936,
+      "grad_norm": 0.4475301770853643,
+      "learning_rate": 9.995724409962381e-07,
+      "loss": 0.3229,
+      "step": 177
+    },
+    {
+      "epoch": 0.08971774193548387,
+      "grad_norm": 0.5571776297010896,
+      "learning_rate": 9.995612650893384e-07,
+      "loss": 0.339,
+      "step": 178
+    },
+    {
+      "epoch": 0.09022177419354839,
+      "grad_norm": 0.3379507080955966,
+      "learning_rate": 9.995499450621084e-07,
+      "loss": 0.3363,
+      "step": 179
+    },
+    {
+      "epoch": 0.0907258064516129,
+      "grad_norm": 0.5115984237062103,
+      "learning_rate": 9.995384809178135e-07,
+      "loss": 0.3318,
+      "step": 180
+    },
+    {
+      "epoch": 0.09122983870967742,
+      "grad_norm": 0.597842695756601,
+      "learning_rate": 9.995268726597616e-07,
+      "loss": 0.3355,
+      "step": 181
+    },
+    {
+      "epoch": 0.09173387096774194,
+      "grad_norm": 1.6790645385073486,
+      "learning_rate": 9.995151202913013e-07,
+      "loss": 0.3379,
+      "step": 182
+    },
+    {
+      "epoch": 0.09223790322580645,
+      "grad_norm": 1.4108139487058173,
+      "learning_rate": 9.99503223815823e-07,
+      "loss": 0.3435,
+      "step": 183
+    },
+    {
+      "epoch": 0.09274193548387097,
+      "grad_norm": 1.1761800110707394,
+      "learning_rate": 9.99491183236759e-07,
+      "loss": 0.3304,
+      "step": 184
+    },
+    {
+      "epoch": 0.09324596774193548,
+      "grad_norm": 0.5429451949780051,
+      "learning_rate": 9.99478998557583e-07,
+      "loss": 0.3275,
+      "step": 185
+    },
+    {
+      "epoch": 0.09375,
+      "grad_norm": 1.0221880040659963,
+      "learning_rate": 9.994666697818097e-07,
+      "loss": 0.3248,
+      "step": 186
+    },
+    {
+      "epoch": 0.09425403225806452,
+      "grad_norm": 1.7971515949261194,
+      "learning_rate": 9.994541969129963e-07,
+      "loss": 0.3252,
+      "step": 187
+    },
+    {
+      "epoch": 0.09475806451612903,
+      "grad_norm": 0.9715025504301362,
+      "learning_rate": 9.99441579954741e-07,
+      "loss": 0.3275,
+      "step": 188
+    },
+    {
+      "epoch": 0.09526209677419355,
+      "grad_norm": 0.3672192941528604,
+      "learning_rate": 9.99428818910684e-07,
+      "loss": 0.3269,
+      "step": 189
+    },
+    {
+      "epoch": 0.09576612903225806,
+      "grad_norm": 0.520936669282708,
+      "learning_rate": 9.994159137845062e-07,
+      "loss": 0.3195,
+      "step": 190
+    },
+    {
+      "epoch": 0.09627016129032258,
+      "grad_norm": 0.7477175783403504,
+      "learning_rate": 9.994028645799312e-07,
+      "loss": 0.3136,
+      "step": 191
+    },
+    {
+      "epoch": 0.0967741935483871,
+      "grad_norm": 0.3729287311518564,
+      "learning_rate": 9.993896713007234e-07,
+      "loss": 0.3272,
+      "step": 192
+    },
+    {
+      "epoch": 0.09727822580645161,
+      "grad_norm": 0.34440602307523366,
+      "learning_rate": 9.99376333950689e-07,
+      "loss": 0.3233,
+      "step": 193
+    },
+    {
+      "epoch": 0.09778225806451613,
+      "grad_norm": 2.072604244189497,
+      "learning_rate": 9.993628525336757e-07,
+      "loss": 0.3436,
+      "step": 194
+    },
+    {
+      "epoch": 0.09828629032258064,
+      "grad_norm": 1.019190865070132,
+      "learning_rate": 9.993492270535728e-07,
+      "loss": 0.3313,
+      "step": 195
+    },
+    {
+      "epoch": 0.09879032258064516,
+      "grad_norm": 0.7904036059961682,
+      "learning_rate": 9.99335457514311e-07,
+      "loss": 0.3274,
+      "step": 196
+    },
+    {
+      "epoch": 0.09929435483870967,
+      "grad_norm": 0.6959515322641198,
+      "learning_rate": 9.993215439198632e-07,
+      "loss": 0.3208,
+      "step": 197
+    },
+    {
+      "epoch": 0.09979838709677419,
+      "grad_norm": 0.7837606339000123,
+      "learning_rate": 9.993074862742432e-07,
+      "loss": 0.3277,
+      "step": 198
+    },
+    {
+      "epoch": 0.1003024193548387,
+      "grad_norm": 0.3983563349530772,
+      "learning_rate": 9.992932845815062e-07,
+      "loss": 0.3259,
+      "step": 199
+    },
+    {
+      "epoch": 0.1003024193548387,
+      "eval_loss": 0.3568817675113678,
+      "eval_runtime": 17.0912,
+      "eval_samples_per_second": 50.026,
+      "eval_steps_per_second": 1.053,
+      "step": 199
+    },
+    {
+      "epoch": 0.10080645161290322,
+      "grad_norm": 2.1662289172188123,
+      "learning_rate": 9.992789388457496e-07,
+      "loss": 0.3374,
+      "step": 200
+    },
+    {
+      "epoch": 0.10131048387096774,
+      "grad_norm": 0.3792807469503325,
+      "learning_rate": 9.992644490711122e-07,
+      "loss": 0.3322,
+      "step": 201
+    },
+    {
+      "epoch": 0.10181451612903226,
+      "grad_norm": 0.8809934605286607,
+      "learning_rate": 9.99249815261774e-07,
+      "loss": 0.3234,
+      "step": 202
+    },
+    {
+      "epoch": 0.10231854838709678,
+      "grad_norm": 0.8130891805575898,
+      "learning_rate": 9.992350374219565e-07,
+      "loss": 0.3205,
+      "step": 203
+    },
+    {
+      "epoch": 0.1028225806451613,
+      "grad_norm": 0.6892675097386061,
+      "learning_rate": 9.992201155559235e-07,
+      "loss": 0.3265,
+      "step": 204
+    },
+    {
+      "epoch": 0.10332661290322581,
+      "grad_norm": 0.3620028450107508,
+      "learning_rate": 9.992050496679796e-07,
+      "loss": 0.3308,
+      "step": 205
+    },
+    {
+      "epoch": 0.10383064516129033,
+      "grad_norm": 1.1011333579285603,
+      "learning_rate": 9.991898397624713e-07,
+      "loss": 0.3297,
+      "step": 206
+    },
+    {
+      "epoch": 0.10433467741935484,
+      "grad_norm": 1.066929614971622,
+      "learning_rate": 9.991744858437867e-07,
+      "loss": 0.3343,
+      "step": 207
+    },
+    {
+      "epoch": 0.10483870967741936,
+      "grad_norm": 0.5043517533284387,
+      "learning_rate": 9.99158987916355e-07,
+      "loss": 0.3357,
+      "step": 208
+    },
+    {
+      "epoch": 0.10534274193548387,
+      "grad_norm": 2.028553721867032,
+      "learning_rate": 9.991433459846475e-07,
+      "loss": 0.3381,
+      "step": 209
+    },
+    {
+      "epoch": 0.10584677419354839,
+      "grad_norm": 1.0136025757003981,
+      "learning_rate": 9.991275600531766e-07,
+      "loss": 0.3256,
+      "step": 210
+    },
+    {
+      "epoch": 0.1063508064516129,
+      "grad_norm": 0.4260801721814953,
+      "learning_rate": 9.991116301264965e-07,
+      "loss": 0.3183,
+      "step": 211
+    },
+    {
+      "epoch": 0.10685483870967742,
+      "grad_norm": 0.8776821224341731,
+      "learning_rate": 9.990955562092032e-07,
+      "loss": 0.3086,
+      "step": 212
+    },
+    {
+      "epoch": 0.10735887096774194,
+      "grad_norm": 0.9670921019854246,
+      "learning_rate": 9.990793383059336e-07,
+      "loss": 0.3163,
+      "step": 213
+    },
+    {
+      "epoch": 0.10786290322580645,
+      "grad_norm": 0.4414053256258077,
+      "learning_rate": 9.990629764213663e-07,
+      "loss": 0.3274,
+      "step": 214
+    },
+    {
+      "epoch": 0.10836693548387097,
+      "grad_norm": 0.8635272396829062,
+      "learning_rate": 9.99046470560222e-07,
+      "loss": 0.3328,
+      "step": 215
+    },
+    {
+      "epoch": 0.10887096774193548,
+      "grad_norm": 1.2397064297417275,
+      "learning_rate": 9.990298207272625e-07,
+      "loss": 0.323,
+      "step": 216
+    },
+    {
+      "epoch": 0.109375,
+      "grad_norm": 0.4800605247066369,
+      "learning_rate": 9.99013026927291e-07,
+      "loss": 0.3248,
+      "step": 217
+    },
+    {
+      "epoch": 0.10987903225806452,
+      "grad_norm": 0.655524859523369,
+      "learning_rate": 9.989960891651521e-07,
+      "loss": 0.3237,
+      "step": 218
+    },
+    {
+      "epoch": 0.11038306451612903,
+      "grad_norm": 1.666081221879668,
+      "learning_rate": 9.98979007445733e-07,
+      "loss": 0.307,
+      "step": 219
+    },
+    {
+      "epoch": 0.11088709677419355,
+      "grad_norm": 0.6823811172697462,
+      "learning_rate": 9.989617817739612e-07,
+      "loss": 0.325,
+      "step": 220
+    },
+    {
+      "epoch": 0.11139112903225806,
+      "grad_norm": 0.38222379218675673,
+      "learning_rate": 9.989444121548061e-07,
+      "loss": 0.3284,
+      "step": 221
+    },
+    {
+      "epoch": 0.11189516129032258,
+      "grad_norm": 3.066211544417165,
+      "learning_rate": 9.989268985932789e-07,
+      "loss": 0.3192,
+      "step": 222
+    },
+    {
+      "epoch": 0.1123991935483871,
+      "grad_norm": 2.642378852763342,
+      "learning_rate": 9.989092410944321e-07,
+      "loss": 0.3381,
+      "step": 223
+    },
+    {
+      "epoch": 0.11290322580645161,
+      "grad_norm": 0.49898455707374983,
+      "learning_rate": 9.9889143966336e-07,
+      "loss": 0.3126,
+      "step": 224
+    },
+    {
+      "epoch": 0.11340725806451613,
+      "grad_norm": 0.8540861654265093,
+      "learning_rate": 9.988734943051981e-07,
+      "loss": 0.3272,
+      "step": 225
+    },
+    {
+      "epoch": 0.11391129032258064,
+      "grad_norm": 1.2446923755324546,
+      "learning_rate": 9.988554050251232e-07,
+      "loss": 0.329,
+      "step": 226
+    },
+    {
+      "epoch": 0.11441532258064516,
+      "grad_norm": 1.1002977014857587,
+      "learning_rate": 9.988371718283543e-07,
+      "loss": 0.3196,
+      "step": 227
+    },
+    {
+      "epoch": 0.11491935483870967,
+      "grad_norm": 0.42649460084589286,
+      "learning_rate": 9.988187947201517e-07,
+      "loss": 0.3226,
+      "step": 228
+    },
+    {
+      "epoch": 0.11542338709677419,
+      "grad_norm": 2.4011621649576824,
+      "learning_rate": 9.988002737058166e-07,
+      "loss": 0.3157,
+      "step": 229
+    },
+    {
+      "epoch": 0.1159274193548387,
+      "grad_norm": 1.9831170581698965,
+      "learning_rate": 9.987816087906924e-07,
+      "loss": 0.3128,
+      "step": 230
+    },
+    {
+      "epoch": 0.11643145161290322,
+      "grad_norm": 0.5007493489029178,
+      "learning_rate": 9.987627999801638e-07,
+      "loss": 0.3098,
+      "step": 231
+    },
+    {
+      "epoch": 0.11693548387096774,
+      "grad_norm": 1.3882086918200365,
+      "learning_rate": 9.987438472796572e-07,
+      "loss": 0.3205,
+      "step": 232
+    },
+    {
+      "epoch": 0.11743951612903226,
+      "grad_norm": 0.8267276195807572,
+      "learning_rate": 9.987247506946401e-07,
+      "loss": 0.3017,
+      "step": 233
+    },
+    {
+      "epoch": 0.11794354838709678,
+      "grad_norm": 0.5834541685362084,
+      "learning_rate": 9.98705510230622e-07,
+      "loss": 0.3211,
+      "step": 234
+    },
+    {
+      "epoch": 0.1184475806451613,
+      "grad_norm": 0.4068048722742785,
+      "learning_rate": 9.986861258931535e-07,
+      "loss": 0.317,
+      "step": 235
+    },
+    {
+      "epoch": 0.11895161290322581,
+      "grad_norm": 0.6447372842200539,
+      "learning_rate": 9.986665976878269e-07,
+      "loss": 0.3162,
+      "step": 236
+    },
+    {
+      "epoch": 0.11945564516129033,
+      "grad_norm": 0.3789031033771853,
+      "learning_rate": 9.986469256202758e-07,
+      "loss": 0.3434,
+      "step": 237
+    },
+    {
+      "epoch": 0.11995967741935484,
+      "grad_norm": 0.7092549539496891,
+      "learning_rate": 9.986271096961758e-07,
+      "loss": 0.3277,
+      "step": 238
+    },
+    {
+      "epoch": 0.12046370967741936,
+      "grad_norm": 0.5202838451750067,
+      "learning_rate": 9.986071499212435e-07,
+      "loss": 0.3138,
+      "step": 239
+    },
+    {
+      "epoch": 0.12096774193548387,
+      "grad_norm": 0.3561759374531645,
+      "learning_rate": 9.98587046301237e-07,
+      "loss": 0.3335,
+      "step": 240
+    },
+    {
+      "epoch": 0.12147177419354839,
+      "grad_norm": 0.3707646084919827,
+      "learning_rate": 9.985667988419562e-07,
+      "loss": 0.3162,
+      "step": 241
+    },
+    {
+      "epoch": 0.1219758064516129,
+      "grad_norm": 0.6547926315231538,
+      "learning_rate": 9.985464075492424e-07,
+      "loss": 0.3279,
+      "step": 242
+    },
+    {
+      "epoch": 0.12247983870967742,
+      "grad_norm": 0.6547474733966526,
+      "learning_rate": 9.985258724289784e-07,
+      "loss": 0.329,
+      "step": 243
+    },
+    {
+      "epoch": 0.12298387096774194,
+      "grad_norm": 1.4287976852804898,
+      "learning_rate": 9.985051934870886e-07,
+      "loss": 0.3171,
+      "step": 244
+    },
+    {
+      "epoch": 0.12348790322580645,
+      "grad_norm": 0.33812675544730314,
+      "learning_rate": 9.984843707295384e-07,
+      "loss": 0.3171,
+      "step": 245
+    },
+    {
+      "epoch": 0.12399193548387097,
+      "grad_norm": 0.5972872507124707,
+      "learning_rate": 9.98463404162335e-07,
+      "loss": 0.3195,
+      "step": 246
+    },
+    {
+      "epoch": 0.12449596774193548,
+      "grad_norm": 0.5523944045513064,
+      "learning_rate": 9.984422937915276e-07,
+      "loss": 0.3137,
+      "step": 247
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.3520728667966682,
+      "learning_rate": 9.98421039623206e-07,
+      "loss": 0.3181,
+      "step": 248
+    },
+    {
+      "epoch": 0.12550403225806453,
+      "grad_norm": 1.0978808931136894,
+      "learning_rate": 9.98399641663502e-07,
+      "loss": 0.3212,
+      "step": 249
+    },
+    {
+      "epoch": 0.12600806451612903,
+      "grad_norm": 0.4007468615552269,
+      "learning_rate": 9.98378099918589e-07,
+      "loss": 0.3235,
+      "step": 250
+    },
+    {
+      "epoch": 0.12651209677419356,
+      "grad_norm": 0.38517912392455744,
+      "learning_rate": 9.983564143946813e-07,
+      "loss": 0.3221,
+      "step": 251
+    },
+    {
+      "epoch": 0.12701612903225806,
+      "grad_norm": 0.8562512071253154,
+      "learning_rate": 9.98334585098035e-07,
+      "loss": 0.3168,
+      "step": 252
+    },
+    {
+      "epoch": 0.1275201612903226,
+      "grad_norm": 0.37641066199373835,
+      "learning_rate": 9.98312612034948e-07,
+      "loss": 0.3338,
+      "step": 253
+    },
+    {
+      "epoch": 0.1280241935483871,
+      "grad_norm": 0.3364345875156596,
+      "learning_rate": 9.982904952117597e-07,
+      "loss": 0.3149,
+      "step": 254
+    },
+    {
+      "epoch": 0.12852822580645162,
+      "grad_norm": 0.4087566783250614,
+      "learning_rate": 9.9826823463485e-07,
+      "loss": 0.3256,
+      "step": 255
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 0.7179205392340822,
+      "learning_rate": 9.982458303106411e-07,
+      "loss": 0.3375,
+      "step": 256
+    },
+    {
+      "epoch": 0.12953629032258066,
+      "grad_norm": 0.8800188682694119,
+      "learning_rate": 9.982232822455968e-07,
+      "loss": 0.3378,
+      "step": 257
+    },
+    {
+      "epoch": 0.13004032258064516,
+      "grad_norm": 0.4425193312322176,
+      "learning_rate": 9.982005904462219e-07,
+      "loss": 0.3303,
+      "step": 258
+    },
+    {
+      "epoch": 0.1305443548387097,
+      "grad_norm": 0.9161891570130994,
+      "learning_rate": 9.981777549190627e-07,
+      "loss": 0.3036,
+      "step": 259
+    },
+    {
+      "epoch": 0.1310483870967742,
+      "grad_norm": 0.3551274779066522,
+      "learning_rate": 9.981547756707074e-07,
+      "loss": 0.3195,
+      "step": 260
+    },
+    {
+      "epoch": 0.13155241935483872,
+      "grad_norm": 0.4319317585133586,
+      "learning_rate": 9.981316527077852e-07,
+      "loss": 0.3321,
+      "step": 261
+    },
+    {
+      "epoch": 0.13205645161290322,
+      "grad_norm": 1.1093458242995622,
+      "learning_rate": 9.981083860369668e-07,
+      "loss": 0.3341,
+      "step": 262
+    },
+    {
+      "epoch": 0.13256048387096775,
+      "grad_norm": 1.1354326631084213,
+      "learning_rate": 9.98084975664965e-07,
+      "loss": 0.3184,
+      "step": 263
+    },
+    {
+      "epoch": 0.13306451612903225,
+      "grad_norm": 0.3445717345214689,
+      "learning_rate": 9.980614215985327e-07,
+      "loss": 0.326,
+      "step": 264
+    },
+    {
+      "epoch": 0.13356854838709678,
+      "grad_norm": 1.795842384240905,
+      "learning_rate": 9.98037723844466e-07,
+      "loss": 0.3176,
+      "step": 265
+    },
+    {
+      "epoch": 0.13407258064516128,
+      "grad_norm": 1.1674146767927045,
+      "learning_rate": 9.98013882409601e-07,
+      "loss": 0.3153,
+      "step": 266
+    },
+    {
+      "epoch": 0.1345766129032258,
+      "grad_norm": 0.3532656598688029,
+      "learning_rate": 9.97989897300816e-07,
+      "loss": 0.3237,
+      "step": 267
+    },
+    {
+      "epoch": 0.1350806451612903,
+      "grad_norm": 1.0598457327232365,
+      "learning_rate": 9.979657685250305e-07,
+      "loss": 0.3316,
+      "step": 268
+    },
+    {
+      "epoch": 0.13558467741935484,
+      "grad_norm": 0.4426506579911205,
+      "learning_rate": 9.979414960892055e-07,
+      "loss": 0.3031,
+      "step": 269
+    },
+    {
+      "epoch": 0.13608870967741934,
+      "grad_norm": 0.4547436037084278,
+      "learning_rate": 9.979170800003436e-07,
+      "loss": 0.3122,
+      "step": 270
+    },
+    {
+      "epoch": 0.13659274193548387,
+      "grad_norm": 0.4462058135399398,
+      "learning_rate": 9.978925202654883e-07,
+      "loss": 0.3227,
+      "step": 271
+    },
+    {
+      "epoch": 0.13709677419354838,
+      "grad_norm": 1.6285597918474548,
+      "learning_rate": 9.978678168917253e-07,
+      "loss": 0.3103,
+      "step": 272
+    },
+    {
+      "epoch": 0.1376008064516129,
+      "grad_norm": 0.8755040490518289,
+      "learning_rate": 9.978429698861812e-07,
+      "loss": 0.3048,
+      "step": 273
+    },
+    {
+      "epoch": 0.1381048387096774,
+      "grad_norm": 0.90287853250279,
+      "learning_rate": 9.978179792560245e-07,
+      "loss": 0.3265,
+      "step": 274
+    },
+    {
+      "epoch": 0.13860887096774194,
+      "grad_norm": 0.3787373854311499,
+      "learning_rate": 9.977928450084642e-07,
+      "loss": 0.3241,
+      "step": 275
+    },
+    {
+      "epoch": 0.13911290322580644,
+      "grad_norm": 0.5847992322700967,
+      "learning_rate": 9.977675671507522e-07,
+      "loss": 0.3176,
+      "step": 276
+    },
+    {
+      "epoch": 0.13961693548387097,
+      "grad_norm": 0.3486681848522623,
+      "learning_rate": 9.977421456901803e-07,
+      "loss": 0.3158,
+      "step": 277
+    },
+    {
+      "epoch": 0.14012096774193547,
+      "grad_norm": 0.38818889942185175,
+      "learning_rate": 9.977165806340827e-07,
+      "loss": 0.3135,
+      "step": 278
+    },
+    {
+      "epoch": 0.140625,
+      "grad_norm": 0.4543982688708197,
+      "learning_rate": 9.97690871989835e-07,
+      "loss": 0.3191,
+      "step": 279
+    },
+    {
+      "epoch": 0.14112903225806453,
+      "grad_norm": 1.4162173420118023,
+      "learning_rate": 9.976650197648536e-07,
+      "loss": 0.314,
+      "step": 280
+    },
+    {
+      "epoch": 0.14163306451612903,
+      "grad_norm": 0.6252978926999877,
+      "learning_rate": 9.976390239665971e-07,
+      "loss": 0.3241,
+      "step": 281
+    },
+    {
+      "epoch": 0.14213709677419356,
+      "grad_norm": 0.3461689499768043,
+      "learning_rate": 9.976128846025646e-07,
+      "loss": 0.3219,
+      "step": 282
+    },
+    {
+      "epoch": 0.14264112903225806,
+      "grad_norm": 0.3699553094334184,
+      "learning_rate": 9.975866016802977e-07,
+      "loss": 0.3037,
+      "step": 283
+    },
+    {
+      "epoch": 0.1431451612903226,
+      "grad_norm": 0.39061647423480156,
+      "learning_rate": 9.975601752073783e-07,
+      "loss": 0.3063,
+      "step": 284
+    },
+    {
+      "epoch": 0.1436491935483871,
+      "grad_norm": 0.5258130756540553,
+      "learning_rate": 9.975336051914307e-07,
+      "loss": 0.3108,
+      "step": 285
+    },
+    {
+      "epoch": 0.14415322580645162,
+      "grad_norm": 0.8216411787070486,
+      "learning_rate": 9.9750689164012e-07,
+      "loss": 0.3056,
+      "step": 286
+    },
+    {
+      "epoch": 0.14465725806451613,
+      "grad_norm": 0.4881519867613522,
+      "learning_rate": 9.974800345611532e-07,
+      "loss": 0.3164,
+      "step": 287
+    },
+    {
+      "epoch": 0.14516129032258066,
+      "grad_norm": 0.43534512320301666,
+      "learning_rate": 9.974530339622779e-07,
+      "loss": 0.3211,
+      "step": 288
+    },
+    {
+      "epoch": 0.14566532258064516,
+      "grad_norm": 0.6993861205113443,
+      "learning_rate": 9.97425889851284e-07,
+      "loss": 0.3267,
+      "step": 289
+    },
+    {
+      "epoch": 0.1461693548387097,
+      "grad_norm": 1.0418090024102467,
+      "learning_rate": 9.973986022360022e-07,
+      "loss": 0.3239,
+      "step": 290
+    },
+    {
+      "epoch": 0.1466733870967742,
+      "grad_norm": 0.3319740811109165,
+      "learning_rate": 9.97371171124305e-07,
+      "loss": 0.3153,
+      "step": 291
+    },
+    {
+      "epoch": 0.14717741935483872,
+      "grad_norm": 0.7719912354248915,
+      "learning_rate": 9.973435965241058e-07,
+      "loss": 0.3101,
+      "step": 292
+    },
+    {
+      "epoch": 0.14768145161290322,
+      "grad_norm": 0.8608701948441182,
+      "learning_rate": 9.973158784433599e-07,
+      "loss": 0.3329,
+      "step": 293
+    },
+    {
+      "epoch": 0.14818548387096775,
+      "grad_norm": 0.5399943385601701,
+      "learning_rate": 9.972880168900638e-07,
+      "loss": 0.3136,
+      "step": 294
+    },
+    {
+      "epoch": 0.14868951612903225,
+      "grad_norm": 1.21808306827626,
+      "learning_rate": 9.972600118722555e-07,
+      "loss": 0.312,
+      "step": 295
+    },
+    {
+      "epoch": 0.14919354838709678,
+      "grad_norm": 2.232388742950908,
+      "learning_rate": 9.97231863398014e-07,
+      "loss": 0.3077,
+      "step": 296
+    },
+    {
+      "epoch": 0.14969758064516128,
+      "grad_norm": 0.40750824930135454,
+      "learning_rate": 9.972035714754602e-07,
+      "loss": 0.3423,
+      "step": 297
+    },
+    {
+      "epoch": 0.1502016129032258,
+      "grad_norm": 0.6444065733453856,
+      "learning_rate": 9.971751361127562e-07,
+      "loss": 0.3108,
+      "step": 298
+    },
+    {
+      "epoch": 0.1507056451612903,
+      "grad_norm": 0.7680402422746748,
+      "learning_rate": 9.971465573181049e-07,
+      "loss": 0.3089,
+      "step": 299
+    },
+    {
+      "epoch": 0.15120967741935484,
+      "grad_norm": 0.38712847346257356,
+      "learning_rate": 9.971178350997516e-07,
+      "loss": 0.2963,
+      "step": 300
+    },
+    {
+      "epoch": 0.15171370967741934,
+      "grad_norm": 0.921107522036033,
+      "learning_rate": 9.970889694659823e-07,
+      "loss": 0.3149,
+      "step": 301
+    },
+    {
+      "epoch": 0.15221774193548387,
+      "grad_norm": 1.009394455117804,
+      "learning_rate": 9.970599604251247e-07,
+      "loss": 0.3017,
+      "step": 302
+    },
+    {
+      "epoch": 0.15272177419354838,
+      "grad_norm": 0.4398399428836015,
+      "learning_rate": 9.970308079855476e-07,
+      "loss": 0.3036,
+      "step": 303
+    },
+    {
+      "epoch": 0.1532258064516129,
+      "grad_norm": 0.8373555944492586,
+      "learning_rate": 9.970015121556615e-07,
+      "loss": 0.3204,
+      "step": 304
+    },
+    {
+      "epoch": 0.1537298387096774,
+      "grad_norm": 1.226841071199288,
+      "learning_rate": 9.969720729439177e-07,
+      "loss": 0.31,
+      "step": 305
+    },
+    {
+      "epoch": 0.15423387096774194,
+      "grad_norm": 0.4736294159912266,
+      "learning_rate": 9.969424903588094e-07,
+      "loss": 0.3273,
+      "step": 306
+    },
+    {
+      "epoch": 0.15473790322580644,
+      "grad_norm": 1.3438422876161837,
+      "learning_rate": 9.969127644088713e-07,
+      "loss": 0.3229,
+      "step": 307
+    },
+    {
+      "epoch": 0.15524193548387097,
+      "grad_norm": 2.420491209892375,
+      "learning_rate": 9.968828951026786e-07,
+      "loss": 0.3016,
+      "step": 308
+    },
+    {
+      "epoch": 0.15574596774193547,
+      "grad_norm": 0.9363754119347487,
+      "learning_rate": 9.96852882448849e-07,
+      "loss": 0.3096,
+      "step": 309
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 0.7464871725596692,
+      "learning_rate": 9.968227264560404e-07,
+      "loss": 0.3114,
+      "step": 310
+    },
+    {
+      "epoch": 0.15675403225806453,
+      "grad_norm": 0.8045539981080978,
+      "learning_rate": 9.96792427132953e-07,
+      "loss": 0.3074,
+      "step": 311
+    },
+    {
+      "epoch": 0.15725806451612903,
+      "grad_norm": 0.7968575088428339,
+      "learning_rate": 9.967619844883277e-07,
+      "loss": 0.3202,
+      "step": 312
+    },
+    {
+      "epoch": 0.15776209677419356,
+      "grad_norm": 0.3462338506463475,
+      "learning_rate": 9.967313985309472e-07,
+      "loss": 0.3104,
+      "step": 313
+    },
+    {
+      "epoch": 0.15826612903225806,
+      "grad_norm": 0.3411937335850433,
+      "learning_rate": 9.967006692696353e-07,
+      "loss": 0.325,
+      "step": 314
+    },
+    {
+      "epoch": 0.1587701612903226,
+      "grad_norm": 1.6309339803104832,
+      "learning_rate": 9.966697967132573e-07,
+      "loss": 0.3048,
+      "step": 315
+    },
+    {
+      "epoch": 0.1592741935483871,
+      "grad_norm": 0.8520422369690264,
+      "learning_rate": 9.966387808707196e-07,
+      "loss": 0.3031,
+      "step": 316
+    },
+    {
+      "epoch": 0.15977822580645162,
+      "grad_norm": 0.550306745006942,
+      "learning_rate": 9.966076217509702e-07,
+      "loss": 0.3174,
+      "step": 317
+    },
+    {
+      "epoch": 0.16028225806451613,
+      "grad_norm": 0.3244041619511754,
+      "learning_rate": 9.965763193629982e-07,
+      "loss": 0.3051,
+      "step": 318
+    },
+    {
+      "epoch": 0.16078629032258066,
+      "grad_norm": 0.3289354707634873,
+      "learning_rate": 9.965448737158343e-07,
+      "loss": 0.3024,
+      "step": 319
+    },
+    {
+      "epoch": 0.16129032258064516,
+      "grad_norm": 0.5698828377828442,
+      "learning_rate": 9.9651328481855e-07,
+      "loss": 0.3197,
+      "step": 320
+    },
+    {
+      "epoch": 0.1617943548387097,
+      "grad_norm": 0.3603627753702205,
+      "learning_rate": 9.964815526802588e-07,
+      "loss": 0.3267,
+      "step": 321
+    },
+    {
+      "epoch": 0.1622983870967742,
+      "grad_norm": 0.41260756113885855,
+      "learning_rate": 9.964496773101155e-07,
+      "loss": 0.3128,
+      "step": 322
+    },
+    {
+      "epoch": 0.16280241935483872,
+      "grad_norm": 0.3123646419774433,
+      "learning_rate": 9.964176587173154e-07,
+      "loss": 0.2909,
+      "step": 323
+    },
+    {
+      "epoch": 0.16330645161290322,
+      "grad_norm": 0.32862366743648225,
+      "learning_rate": 9.963854969110958e-07,
+      "loss": 0.3249,
+      "step": 324
+    },
+    {
+      "epoch": 0.16381048387096775,
+      "grad_norm": 0.32317653597651286,
+      "learning_rate": 9.963531919007355e-07,
+      "loss": 0.3179,
+      "step": 325
+    },
+    {
+      "epoch": 0.16431451612903225,
+      "grad_norm": 0.3636694182909097,
+      "learning_rate": 9.963207436955539e-07,
+      "loss": 0.3094,
+      "step": 326
+    },
+    {
+      "epoch": 0.16481854838709678,
+      "grad_norm": 0.5601385575192833,
+      "learning_rate": 9.962881523049125e-07,
+      "loss": 0.313,
+      "step": 327
+    },
+    {
+      "epoch": 0.16532258064516128,
+      "grad_norm": 0.32805972924192683,
+      "learning_rate": 9.962554177382134e-07,
+      "loss": 0.304,
+      "step": 328
+    },
+    {
+      "epoch": 0.1658266129032258,
+      "grad_norm": 0.4699969421619995,
+      "learning_rate": 9.962225400049004e-07,
+      "loss": 0.309,
+      "step": 329
+    },
+    {
+      "epoch": 0.1663306451612903,
+      "grad_norm": 0.3767389273779974,
+      "learning_rate": 9.961895191144586e-07,
+      "loss": 0.2983,
+      "step": 330
+    },
+    {
+      "epoch": 0.16683467741935484,
+      "grad_norm": 0.503719665226859,
+      "learning_rate": 9.961563550764143e-07,
+      "loss": 0.3112,
+      "step": 331
+    },
+    {
+      "epoch": 0.16733870967741934,
+      "grad_norm": 0.3438328709283268,
+      "learning_rate": 9.961230479003348e-07,
+      "loss": 0.2985,
+      "step": 332
+    },
+    {
+      "epoch": 0.16784274193548387,
+      "grad_norm": 0.3083813504969578,
+      "learning_rate": 9.960895975958296e-07,
+      "loss": 0.3134,
+      "step": 333
+    },
+    {
+      "epoch": 0.16834677419354838,
+      "grad_norm": 0.4787512237902366,
+      "learning_rate": 9.960560041725486e-07,
+      "loss": 0.3213,
+      "step": 334
+    },
+    {
+      "epoch": 0.1688508064516129,
+      "grad_norm": 1.4378814525493548,
+      "learning_rate": 9.960222676401833e-07,
+      "loss": 0.3056,
+      "step": 335
+    },
+    {
+      "epoch": 0.1693548387096774,
+      "grad_norm": 0.3506129924302826,
+      "learning_rate": 9.959883880084664e-07,
+      "loss": 0.3174,
+      "step": 336
+    },
+    {
+      "epoch": 0.16985887096774194,
+      "grad_norm": 0.35990282931470485,
+      "learning_rate": 9.95954365287172e-07,
+      "loss": 0.3224,
+      "step": 337
+    },
+    {
+      "epoch": 0.17036290322580644,
+      "grad_norm": 1.093999056396477,
+      "learning_rate": 9.959201994861156e-07,
+      "loss": 0.2938,
+      "step": 338
+    },
+    {
+      "epoch": 0.17086693548387097,
+      "grad_norm": 0.7575441587383402,
+      "learning_rate": 9.958858906151536e-07,
+      "loss": 0.3092,
+      "step": 339
+    },
+    {
+      "epoch": 0.17137096774193547,
+      "grad_norm": 0.7138793464492631,
+      "learning_rate": 9.958514386841842e-07,
+      "loss": 0.3126,
+      "step": 340
+    },
+    {
+      "epoch": 0.171875,
+      "grad_norm": 0.6820850764538464,
+      "learning_rate": 9.95816843703146e-07,
+      "loss": 0.3051,
+      "step": 341
+    },
+    {
+      "epoch": 0.17237903225806453,
+      "grad_norm": 0.438185665282885,
+      "learning_rate": 9.957821056820202e-07,
+      "loss": 0.3107,
+      "step": 342
+    },
+    {
+      "epoch": 0.17288306451612903,
+      "grad_norm": 0.2993355268251931,
+      "learning_rate": 9.957472246308278e-07,
+      "loss": 0.3177,
+      "step": 343
+    },
+    {
+      "epoch": 0.17338709677419356,
+      "grad_norm": 0.3143562128856459,
+      "learning_rate": 9.957122005596324e-07,
+      "loss": 0.3028,
+      "step": 344
+    },
+    {
+      "epoch": 0.17389112903225806,
+      "grad_norm": 0.29963517583943,
+      "learning_rate": 9.956770334785377e-07,
+      "loss": 0.303,
+      "step": 345
+    },
+    {
+      "epoch": 0.1743951612903226,
+      "grad_norm": 0.41635137878897605,
+      "learning_rate": 9.956417233976895e-07,
+      "loss": 0.3007,
+      "step": 346
+    },
+    {
+      "epoch": 0.1748991935483871,
+      "grad_norm": 0.3795788244623765,
+      "learning_rate": 9.956062703272744e-07,
+      "loss": 0.3183,
+      "step": 347
+    },
+    {
+      "epoch": 0.17540322580645162,
+      "grad_norm": 0.5158539655619079,
+      "learning_rate": 9.955706742775204e-07,
+      "loss": 0.3034,
+      "step": 348
+    },
+    {
+      "epoch": 0.17590725806451613,
+      "grad_norm": 0.41143417528165055,
+      "learning_rate": 9.955349352586968e-07,
+      "loss": 0.3064,
+      "step": 349
+    },
+    {
+      "epoch": 0.17641129032258066,
+      "grad_norm": 0.3499570984862177,
+      "learning_rate": 9.95499053281114e-07,
+      "loss": 0.3165,
+      "step": 350
+    },
+    {
+      "epoch": 0.17691532258064516,
+      "grad_norm": 0.6117708541856488,
+      "learning_rate": 9.95463028355124e-07,
+      "loss": 0.3135,
+      "step": 351
+    },
+    {
+      "epoch": 0.1774193548387097,
+      "grad_norm": 0.3512173058497519,
+      "learning_rate": 9.954268604911193e-07,
+      "loss": 0.3086,
+      "step": 352
+    },
+    {
+      "epoch": 0.1779233870967742,
+      "grad_norm": 0.5459910083191652,
+      "learning_rate": 9.953905496995346e-07,
+      "loss": 0.3212,
+      "step": 353
+    },
+    {
+      "epoch": 0.17842741935483872,
+      "grad_norm": 0.48953620036372436,
+      "learning_rate": 9.953540959908448e-07,
+      "loss": 0.3082,
+      "step": 354
+    },
+    {
+      "epoch": 0.17893145161290322,
+      "grad_norm": 0.4806785690019039,
+      "learning_rate": 9.953174993755669e-07,
+      "loss": 0.3114,
+      "step": 355
+    },
+    {
+      "epoch": 0.17943548387096775,
+      "grad_norm": 0.27574681950143637,
+      "learning_rate": 9.95280759864259e-07,
+      "loss": 0.3088,
+      "step": 356
+    },
+    {
+      "epoch": 0.17993951612903225,
+      "grad_norm": 0.4046536057043104,
+      "learning_rate": 9.952438774675199e-07,
+      "loss": 0.3014,
+      "step": 357
+    },
+    {
+      "epoch": 0.18044354838709678,
+      "grad_norm": 0.526200938156327,
+      "learning_rate": 9.952068521959898e-07,
+      "loss": 0.3042,
+      "step": 358
+    },
+    {
+      "epoch": 0.18094758064516128,
+      "grad_norm": 0.5729896337994929,
+      "learning_rate": 9.951696840603508e-07,
+      "loss": 0.3096,
+      "step": 359
+    },
+    {
+      "epoch": 0.1814516129032258,
+      "grad_norm": 0.5261901087089867,
+      "learning_rate": 9.95132373071325e-07,
+      "loss": 0.3077,
+      "step": 360
+    },
+    {
+      "epoch": 0.1819556451612903,
+      "grad_norm": 0.6668832949382653,
+      "learning_rate": 9.950949192396772e-07,
+      "loss": 0.3025,
+      "step": 361
+    },
+    {
+      "epoch": 0.18245967741935484,
+      "grad_norm": 0.862055275986415,
+      "learning_rate": 9.950573225762117e-07,
+      "loss": 0.3015,
+      "step": 362
+    },
+    {
+      "epoch": 0.18296370967741934,
+      "grad_norm": 0.6293046065894792,
+      "learning_rate": 9.950195830917756e-07,
+      "loss": 0.302,
+      "step": 363
+    },
+    {
+      "epoch": 0.18346774193548387,
+      "grad_norm": 0.6877385211906398,
+      "learning_rate": 9.949817007972563e-07,
+      "loss": 0.3176,
+      "step": 364
+    },
+    {
+      "epoch": 0.18397177419354838,
+      "grad_norm": 0.29549386712677456,
+      "learning_rate": 9.949436757035825e-07,
+      "loss": 0.3089,
+      "step": 365
+    },
+    {
+      "epoch": 0.1844758064516129,
+      "grad_norm": 0.8557932164329561,
+      "learning_rate": 9.949055078217244e-07,
+      "loss": 0.3296,
+      "step": 366
+    },
+    {
+      "epoch": 0.1849798387096774,
+      "grad_norm": 0.5402766960899985,
+      "learning_rate": 9.948671971626927e-07,
+      "loss": 0.3175,
+      "step": 367
+    },
+    {
+      "epoch": 0.18548387096774194,
+      "grad_norm": 0.49998887489839794,
+      "learning_rate": 9.948287437375403e-07,
+      "loss": 0.314,
+      "step": 368
+    },
+    {
+      "epoch": 0.18598790322580644,
+      "grad_norm": 0.33956036638788806,
+      "learning_rate": 9.94790147557361e-07,
+      "loss": 0.3072,
+      "step": 369
+    },
+    {
+      "epoch": 0.18649193548387097,
+      "grad_norm": 0.4691366516694783,
+      "learning_rate": 9.94751408633289e-07,
+      "loss": 0.3007,
+      "step": 370
+    },
+    {
+      "epoch": 0.18699596774193547,
+      "grad_norm": 0.4294056345752334,
+      "learning_rate": 9.947125269765001e-07,
+      "loss": 0.3244,
+      "step": 371
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.5198726803714843,
+      "learning_rate": 9.946735025982121e-07,
+      "loss": 0.3309,
+      "step": 372
+    },
+    {
+      "epoch": 0.18800403225806453,
+      "grad_norm": 0.5083174614120537,
+      "learning_rate": 9.94634335509683e-07,
+      "loss": 0.3064,
+      "step": 373
+    },
+    {
+      "epoch": 0.18850806451612903,
+      "grad_norm": 0.4820500182649492,
+      "learning_rate": 9.94595025722212e-07,
+      "loss": 0.3042,
+      "step": 374
+    },
+    {
+      "epoch": 0.18901209677419356,
+      "grad_norm": 0.5290432017994986,
+      "learning_rate": 9.9455557324714e-07,
+      "loss": 0.2969,
+      "step": 375
+    },
+    {
+      "epoch": 0.18951612903225806,
+      "grad_norm": 0.9665951840791808,
+      "learning_rate": 9.945159780958487e-07,
+      "loss": 0.3053,
+      "step": 376
+    },
+    {
+      "epoch": 0.1900201612903226,
+      "grad_norm": 0.27376501652334334,
+      "learning_rate": 9.94476240279761e-07,
+      "loss": 0.2975,
+      "step": 377
+    },
+    {
+      "epoch": 0.1905241935483871,
+      "grad_norm": 0.6907529950403531,
+      "learning_rate": 9.944363598103412e-07,
+      "loss": 0.3028,
+      "step": 378
+    },
+    {
+      "epoch": 0.19102822580645162,
+      "grad_norm": 0.7206171183198804,
+      "learning_rate": 9.943963366990944e-07,
+      "loss": 0.3244,
+      "step": 379
+    },
+    {
+      "epoch": 0.19153225806451613,
+      "grad_norm": 0.4351702422546798,
+      "learning_rate": 9.943561709575671e-07,
+      "loss": 0.3019,
+      "step": 380
+    },
+    {
+      "epoch": 0.19203629032258066,
+      "grad_norm": 0.4993062351913905,
+      "learning_rate": 9.94315862597347e-07,
+      "loss": 0.3147,
+      "step": 381
+    },
+    {
+      "epoch": 0.19254032258064516,
+      "grad_norm": 1.900968501331666,
+      "learning_rate": 9.942754116300627e-07,
+      "loss": 0.2954,
+      "step": 382
+    },
+    {
+      "epoch": 0.1930443548387097,
+      "grad_norm": 1.6477482220874253,
+      "learning_rate": 9.942348180673838e-07,
+      "loss": 0.3121,
+      "step": 383
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 0.742153580574187,
+      "learning_rate": 9.941940819210215e-07,
+      "loss": 0.2992,
+      "step": 384
+    },
+    {
+      "epoch": 0.19405241935483872,
+      "grad_norm": 0.47705653322280234,
+      "learning_rate": 9.941532032027281e-07,
+      "loss": 0.3244,
+      "step": 385
+    },
+    {
+      "epoch": 0.19455645161290322,
+      "grad_norm": 0.911598765383663,
+      "learning_rate": 9.941121819242965e-07,
+      "loss": 0.319,
+      "step": 386
+    },
+    {
+      "epoch": 0.19506048387096775,
+      "grad_norm": 0.8928463372091003,
+      "learning_rate": 9.940710180975615e-07,
+      "loss": 0.3176,
+      "step": 387
+    },
+    {
+      "epoch": 0.19556451612903225,
+      "grad_norm": 0.5466103602367745,
+      "learning_rate": 9.940297117343983e-07,
+      "loss": 0.3117,
+      "step": 388
+    },
+    {
+      "epoch": 0.19606854838709678,
+      "grad_norm": 0.28786769682640595,
+      "learning_rate": 9.939882628467235e-07,
+      "loss": 0.319,
+      "step": 389
+    },
+    {
+      "epoch": 0.19657258064516128,
+      "grad_norm": 1.750270794203163,
+      "learning_rate": 9.939466714464953e-07,
+      "loss": 0.3097,
+      "step": 390
+    },
+    {
+      "epoch": 0.1970766129032258,
+      "grad_norm": 1.3109389597034193,
+      "learning_rate": 9.939049375457117e-07,
+      "loss": 0.3095,
+      "step": 391
+    },
+    {
+      "epoch": 0.1975806451612903,
+      "grad_norm": 0.3810245348808811,
+      "learning_rate": 9.938630611564136e-07,
+      "loss": 0.3015,
+      "step": 392
+    },
+    {
+      "epoch": 0.19808467741935484,
+      "grad_norm": 0.2822642693051148,
+      "learning_rate": 9.938210422906816e-07,
+      "loss": 0.3065,
+      "step": 393
+    },
+    {
+      "epoch": 0.19858870967741934,
+      "grad_norm": 0.4548013874567676,
+      "learning_rate": 9.93778880960638e-07,
+      "loss": 0.3059,
+      "step": 394
+    },
+    {
+      "epoch": 0.19909274193548387,
+      "grad_norm": 0.7403977526231464,
+      "learning_rate": 9.937365771784458e-07,
+      "loss": 0.3001,
+      "step": 395
+    },
+    {
+      "epoch": 0.19959677419354838,
+      "grad_norm": 0.40584105744807936,
+      "learning_rate": 9.936941309563097e-07,
+      "loss": 0.3018,
+      "step": 396
+    },
+    {
+      "epoch": 0.2001008064516129,
+      "grad_norm": 0.4027563712329421,
+      "learning_rate": 9.936515423064752e-07,
+      "loss": 0.2821,
+      "step": 397
+    },
+    {
+      "epoch": 0.2006048387096774,
+      "grad_norm": 0.3215902734384433,
+      "learning_rate": 9.936088112412288e-07,
+      "loss": 0.3093,
+      "step": 398
+    },
+    {
+      "epoch": 0.2006048387096774,
+      "eval_loss": 0.33719342947006226,
+      "eval_runtime": 17.0014,
+      "eval_samples_per_second": 50.29,
+      "eval_steps_per_second": 1.059,
+      "step": 398
+    },
+    {
+      "epoch": 0.20110887096774194,
+      "grad_norm": 0.9421611516647215,
+      "learning_rate": 9.935659377728982e-07,
+      "loss": 0.2979,
+      "step": 399
+    },
+    {
+      "epoch": 0.20161290322580644,
+      "grad_norm": 0.9245468490926179,
+      "learning_rate": 9.935229219138517e-07,
+      "loss": 0.2942,
+      "step": 400
+    },
+    {
+      "epoch": 0.20211693548387097,
+      "grad_norm": 0.6824205781396179,
+      "learning_rate": 9.934797636764999e-07,
+      "loss": 0.2971,
+      "step": 401
+    },
+    {
+      "epoch": 0.20262096774193547,
+      "grad_norm": 0.27185149783417195,
+      "learning_rate": 9.934364630732928e-07,
+      "loss": 0.3107,
+      "step": 402
+    },
+    {
+      "epoch": 0.203125,
+      "grad_norm": 0.6827808287541225,
+      "learning_rate": 9.933930201167228e-07,
+      "loss": 0.3184,
+      "step": 403
+    },
+    {
+      "epoch": 0.20362903225806453,
+      "grad_norm": 0.6173711604161257,
+      "learning_rate": 9.93349434819323e-07,
+      "loss": 0.3137,
+      "step": 404
+    },
+    {
+      "epoch": 0.20413306451612903,
+      "grad_norm": 0.8730361088623172,
+      "learning_rate": 9.933057071936674e-07,
+      "loss": 0.3045,
+      "step": 405
+    },
+    {
+      "epoch": 0.20463709677419356,
+      "grad_norm": 0.2347631790989028,
+      "learning_rate": 9.932618372523712e-07,
+      "loss": 0.3174,
+      "step": 406
+    },
+    {
+      "epoch": 0.20514112903225806,
+      "grad_norm": 0.2618825790896803,
+      "learning_rate": 9.932178250080905e-07,
+      "loss": 0.3053,
+      "step": 407
+    },
+    {
+      "epoch": 0.2056451612903226,
+      "grad_norm": 0.8173532124618281,
+      "learning_rate": 9.931736704735226e-07,
+      "loss": 0.2987,
+      "step": 408
+    },
+    {
+      "epoch": 0.2061491935483871,
+      "grad_norm": 0.6079169109005206,
+      "learning_rate": 9.931293736614059e-07,
+      "loss": 0.3131,
+      "step": 409
+    },
+    {
+      "epoch": 0.20665322580645162,
+      "grad_norm": 0.4779413914330351,
+      "learning_rate": 9.930849345845195e-07,
+      "loss": 0.3095,
+      "step": 410
+    },
+    {
+      "epoch": 0.20715725806451613,
+      "grad_norm": 0.5403444992322752,
+      "learning_rate": 9.930403532556841e-07,
+      "loss": 0.2985,
+      "step": 411
+    },
+    {
+      "epoch": 0.20766129032258066,
+      "grad_norm": 0.33724492734825506,
+      "learning_rate": 9.929956296877609e-07,
+      "loss": 0.3029,
+      "step": 412
+    },
+    {
+      "epoch": 0.20816532258064516,
+      "grad_norm": 0.803588317511041,
+      "learning_rate": 9.929507638936527e-07,
+      "loss": 0.304,
+      "step": 413
+    },
+    {
+      "epoch": 0.2086693548387097,
+      "grad_norm": 0.24653864233446288,
+      "learning_rate": 9.929057558863025e-07,
+      "loss": 0.2991,
+      "step": 414
+    },
+    {
+      "epoch": 0.2091733870967742,
+      "grad_norm": 0.3080563000251818,
+      "learning_rate": 9.928606056786953e-07,
+      "loss": 0.3045,
+      "step": 415
+    },
+    {
+      "epoch": 0.20967741935483872,
+      "grad_norm": 0.23276679829744804,
+      "learning_rate": 9.928153132838564e-07,
+      "loss": 0.2998,
+      "step": 416
+    },
+    {
+      "epoch": 0.21018145161290322,
+      "grad_norm": 0.3031284409615895,
+      "learning_rate": 9.927698787148524e-07,
+      "loss": 0.3129,
+      "step": 417
+    },
+    {
+      "epoch": 0.21068548387096775,
+      "grad_norm": 0.7190277552056444,
+      "learning_rate": 9.92724301984791e-07,
+      "loss": 0.318,
+      "step": 418
+    },
+    {
+      "epoch": 0.21118951612903225,
+      "grad_norm": 0.671405235151991,
+      "learning_rate": 9.92678583106821e-07,
+      "loss": 0.3152,
+      "step": 419
+    },
+    {
+      "epoch": 0.21169354838709678,
+      "grad_norm": 0.26427462576363475,
+      "learning_rate": 9.926327220941313e-07,
+      "loss": 0.3131,
+      "step": 420
+    },
+    {
+      "epoch": 0.21219758064516128,
+      "grad_norm": 0.21989852886239047,
+      "learning_rate": 9.925867189599534e-07,
+      "loss": 0.2913,
+      "step": 421
+    },
+    {
+      "epoch": 0.2127016129032258,
+      "grad_norm": 0.5902102443498939,
+      "learning_rate": 9.925405737175582e-07,
+      "loss": 0.3047,
+      "step": 422
+    },
+    {
+      "epoch": 0.2132056451612903,
+      "grad_norm": 0.6036703980264593,
+      "learning_rate": 9.924942863802586e-07,
+      "loss": 0.3023,
+      "step": 423
+    },
+    {
+      "epoch": 0.21370967741935484,
+      "grad_norm": 0.6785617848990195,
+      "learning_rate": 9.924478569614084e-07,
+      "loss": 0.3059,
+      "step": 424
+    },
+    {
+      "epoch": 0.21421370967741934,
+      "grad_norm": 0.41465435154148006,
+      "learning_rate": 9.924012854744019e-07,
+      "loss": 0.2937,
+      "step": 425
+    },
+    {
+      "epoch": 0.21471774193548387,
+      "grad_norm": 0.5756370196897739,
+      "learning_rate": 9.923545719326748e-07,
+      "loss": 0.3003,
+      "step": 426
+    },
+    {
+      "epoch": 0.21522177419354838,
+      "grad_norm": 1.1087103802236575,
+      "learning_rate": 9.923077163497037e-07,
+      "loss": 0.3115,
+      "step": 427
+    },
+    {
+      "epoch": 0.2157258064516129,
+      "grad_norm": 0.5075052144087366,
+      "learning_rate": 9.922607187390062e-07,
+      "loss": 0.2861,
+      "step": 428
+    },
+    {
+      "epoch": 0.2162298387096774,
+      "grad_norm": 0.21875262709516646,
+      "learning_rate": 9.922135791141407e-07,
+      "loss": 0.3111,
+      "step": 429
+    },
+    {
+      "epoch": 0.21673387096774194,
+      "grad_norm": 0.5081145594897964,
+      "learning_rate": 9.921662974887067e-07,
+      "loss": 0.3012,
+      "step": 430
+    },
+    {
+      "epoch": 0.21723790322580644,
+      "grad_norm": 0.3760477536962576,
+      "learning_rate": 9.921188738763447e-07,
+      "loss": 0.307,
+      "step": 431
+    },
+    {
+      "epoch": 0.21774193548387097,
+      "grad_norm": 0.3082054028318284,
+      "learning_rate": 9.92071308290736e-07,
+      "loss": 0.3171,
+      "step": 432
+    },
+    {
+      "epoch": 0.21824596774193547,
+      "grad_norm": 0.22029848103815816,
+      "learning_rate": 9.920236007456031e-07,
+      "loss": 0.2931,
+      "step": 433
+    },
+    {
+      "epoch": 0.21875,
+      "grad_norm": 0.41253335918523026,
+      "learning_rate": 9.919757512547094e-07,
+      "loss": 0.3011,
+      "step": 434
+    },
+    {
+      "epoch": 0.21925403225806453,
+      "grad_norm": 0.5610820660085883,
+      "learning_rate": 9.91927759831859e-07,
+      "loss": 0.3088,
+      "step": 435
+    },
+    {
+      "epoch": 0.21975806451612903,
+      "grad_norm": 0.35401767629764336,
+      "learning_rate": 9.918796264908973e-07,
+      "loss": 0.3003,
+      "step": 436
+    },
+    {
+      "epoch": 0.22026209677419356,
+      "grad_norm": 0.28203957531336254,
+      "learning_rate": 9.918313512457104e-07,
+      "loss": 0.3081,
+      "step": 437
+    },
+    {
+      "epoch": 0.22076612903225806,
+      "grad_norm": 0.2554173995348508,
+      "learning_rate": 9.917829341102254e-07,
+      "loss": 0.2942,
+      "step": 438
+    },
+    {
+      "epoch": 0.2212701612903226,
+      "grad_norm": 0.23639618149360714,
+      "learning_rate": 9.917343750984102e-07,
+      "loss": 0.3021,
+      "step": 439
+    },
+    {
+      "epoch": 0.2217741935483871,
+      "grad_norm": 0.4305124117168449,
+      "learning_rate": 9.91685674224274e-07,
+      "loss": 0.3085,
+      "step": 440
+    },
+    {
+      "epoch": 0.22227822580645162,
+      "grad_norm": 0.2220338714231415,
+      "learning_rate": 9.916368315018666e-07,
+      "loss": 0.3028,
+      "step": 441
+    },
+    {
+      "epoch": 0.22278225806451613,
+      "grad_norm": 0.2927983739524596,
+      "learning_rate": 9.91587846945279e-07,
+      "loss": 0.3017,
+      "step": 442
+    },
+    {
+      "epoch": 0.22328629032258066,
+      "grad_norm": 0.3420210184854715,
+      "learning_rate": 9.915387205686427e-07,
+      "loss": 0.3081,
+      "step": 443
+    },
+    {
+      "epoch": 0.22379032258064516,
+      "grad_norm": 0.39311494277912007,
+      "learning_rate": 9.914894523861304e-07,
+      "loss": 0.3028,
+      "step": 444
+    },
+    {
+      "epoch": 0.2242943548387097,
+      "grad_norm": 0.4151663339605208,
+      "learning_rate": 9.914400424119555e-07,
+      "loss": 0.2956,
+      "step": 445
+    },
+    {
+      "epoch": 0.2247983870967742,
+      "grad_norm": 0.2123585820142778,
+      "learning_rate": 9.91390490660373e-07,
+      "loss": 0.3075,
+      "step": 446
+    },
+    {
+      "epoch": 0.22530241935483872,
+      "grad_norm": 0.571261555860595,
+      "learning_rate": 9.913407971456778e-07,
+      "loss": 0.3013,
+      "step": 447
+    },
+    {
+      "epoch": 0.22580645161290322,
+      "grad_norm": 0.23955795641291014,
+      "learning_rate": 9.912909618822063e-07,
+      "loss": 0.3134,
+      "step": 448
+    },
+    {
+      "epoch": 0.22631048387096775,
+      "grad_norm": 0.570859883971083,
+      "learning_rate": 9.912409848843358e-07,
+      "loss": 0.3056,
+      "step": 449
+    },
+    {
+      "epoch": 0.22681451612903225,
+      "grad_norm": 0.48199321601840345,
+      "learning_rate": 9.911908661664842e-07,
+      "loss": 0.3154,
+      "step": 450
+    },
+    {
+      "epoch": 0.22731854838709678,
+      "grad_norm": 0.5422943126443341,
+      "learning_rate": 9.911406057431104e-07,
+      "loss": 0.2987,
+      "step": 451
+    },
+    {
+      "epoch": 0.22782258064516128,
+      "grad_norm": 0.3574993769315431,
+      "learning_rate": 9.910902036287143e-07,
+      "loss": 0.2941,
+      "step": 452
+    },
+    {
+      "epoch": 0.2283266129032258,
+      "grad_norm": 1.1833977307135548,
+      "learning_rate": 9.910396598378366e-07,
+      "loss": 0.3017,
+      "step": 453
+    },
+    {
+      "epoch": 0.2288306451612903,
+      "grad_norm": 0.25337165989867305,
+      "learning_rate": 9.90988974385059e-07,
+      "loss": 0.2925,
+      "step": 454
+    },
+    {
+      "epoch": 0.22933467741935484,
+      "grad_norm": 0.2480951752658535,
+      "learning_rate": 9.909381472850036e-07,
+      "loss": 0.3021,
+      "step": 455
+    },
+    {
+      "epoch": 0.22983870967741934,
+      "grad_norm": 0.409014941345192,
+      "learning_rate": 9.90887178552334e-07,
+      "loss": 0.3272,
+      "step": 456
+    },
+    {
+      "epoch": 0.23034274193548387,
+      "grad_norm": 0.7457987975927259,
+      "learning_rate": 9.908360682017544e-07,
+      "loss": 0.3048,
+      "step": 457
+    },
+    {
+      "epoch": 0.23084677419354838,
+      "grad_norm": 0.45956267234032566,
+      "learning_rate": 9.907848162480094e-07,
+      "loss": 0.3066,
+      "step": 458
+    },
+    {
+      "epoch": 0.2313508064516129,
+      "grad_norm": 0.3779765403542269,
+      "learning_rate": 9.907334227058855e-07,
+      "loss": 0.3009,
+      "step": 459
+    },
+    {
+      "epoch": 0.2318548387096774,
+      "grad_norm": 0.2666151159273384,
+      "learning_rate": 9.90681887590209e-07,
+      "loss": 0.3124,
+      "step": 460
+    },
+    {
+      "epoch": 0.23235887096774194,
+      "grad_norm": 0.2839428445664857,
+      "learning_rate": 9.906302109158474e-07,
+      "loss": 0.3034,
+      "step": 461
+    },
+    {
+      "epoch": 0.23286290322580644,
+      "grad_norm": 0.4380368120654733,
+      "learning_rate": 9.905783926977094e-07,
+      "loss": 0.3087,
+      "step": 462
+    },
+    {
+      "epoch": 0.23336693548387097,
+      "grad_norm": 0.46446987722443317,
+      "learning_rate": 9.90526432950744e-07,
+      "loss": 0.3066,
+      "step": 463
+    },
+    {
+      "epoch": 0.23387096774193547,
+      "grad_norm": 0.2400929003281859,
+      "learning_rate": 9.904743316899412e-07,
+      "loss": 0.2971,
+      "step": 464
+    },
+    {
+      "epoch": 0.234375,
+      "grad_norm": 0.22533888473808292,
+      "learning_rate": 9.904220889303322e-07,
+      "loss": 0.3034,
+      "step": 465
+    },
+    {
+      "epoch": 0.23487903225806453,
+      "grad_norm": 0.5169333077194566,
+      "learning_rate": 9.903697046869885e-07,
+      "loss": 0.3058,
+      "step": 466
+    },
+    {
+      "epoch": 0.23538306451612903,
+      "grad_norm": 0.37976744599257856,
+      "learning_rate": 9.903171789750227e-07,
+      "loss": 0.2943,
+      "step": 467
+    },
+    {
+      "epoch": 0.23588709677419356,
+      "grad_norm": 0.303653990179142,
+      "learning_rate": 9.90264511809588e-07,
+      "loss": 0.3037,
+      "step": 468
+    },
+    {
+      "epoch": 0.23639112903225806,
+      "grad_norm": 0.21581398194659623,
+      "learning_rate": 9.902117032058788e-07,
+      "loss": 0.3142,
+      "step": 469
+    },
+    {
+      "epoch": 0.2368951612903226,
+      "grad_norm": 0.4025038895307978,
+      "learning_rate": 9.9015875317913e-07,
+      "loss": 0.2886,
+      "step": 470
+    },
+    {
+      "epoch": 0.2373991935483871,
+      "grad_norm": 0.7490982140928762,
+      "learning_rate": 9.90105661744617e-07,
+      "loss": 0.3048,
+      "step": 471
+    },
+    {
+      "epoch": 0.23790322580645162,
+      "grad_norm": 0.2765613593877836,
+      "learning_rate": 9.900524289176571e-07,
+      "loss": 0.3113,
+      "step": 472
+    },
+    {
+      "epoch": 0.23840725806451613,
+      "grad_norm": 0.25206617170451434,
+      "learning_rate": 9.899990547136068e-07,
+      "loss": 0.3031,
+      "step": 473
+    },
+    {
+      "epoch": 0.23891129032258066,
+      "grad_norm": 0.5866005018512471,
+      "learning_rate": 9.899455391478646e-07,
+      "loss": 0.3011,
+      "step": 474
+    },
+    {
+      "epoch": 0.23941532258064516,
+      "grad_norm": 0.39806300430021535,
+      "learning_rate": 9.898918822358695e-07,
+      "loss": 0.2955,
+      "step": 475
+    },
+    {
+      "epoch": 0.2399193548387097,
+      "grad_norm": 0.2584605383681581,
+      "learning_rate": 9.898380839931012e-07,
+      "loss": 0.3044,
+      "step": 476
+    },
+    {
+      "epoch": 0.2404233870967742,
+      "grad_norm": 0.19700196645461712,
+      "learning_rate": 9.897841444350799e-07,
+      "loss": 0.2983,
+      "step": 477
+    },
+    {
+      "epoch": 0.24092741935483872,
+      "grad_norm": 0.306687320143844,
+      "learning_rate": 9.89730063577367e-07,
+      "loss": 0.2963,
+      "step": 478
+    },
+    {
+      "epoch": 0.24143145161290322,
+      "grad_norm": 0.19514269842432405,
+      "learning_rate": 9.896758414355646e-07,
+      "loss": 0.2852,
+      "step": 479
+    },
+    {
+      "epoch": 0.24193548387096775,
+      "grad_norm": 0.2903995442375573,
+      "learning_rate": 9.89621478025315e-07,
+      "loss": 0.3039,
+      "step": 480
+    },
+    {
+      "epoch": 0.24243951612903225,
+      "grad_norm": 0.24936460961507168,
+      "learning_rate": 9.895669733623024e-07,
+      "loss": 0.3025,
+      "step": 481
+    },
+    {
+      "epoch": 0.24294354838709678,
+      "grad_norm": 0.20112659720170964,
+      "learning_rate": 9.895123274622506e-07,
+      "loss": 0.308,
+      "step": 482
+    },
+    {
+      "epoch": 0.24344758064516128,
+      "grad_norm": 0.3732280770776686,
+      "learning_rate": 9.894575403409246e-07,
+      "loss": 0.3048,
+      "step": 483
+    },
+    {
+      "epoch": 0.2439516129032258,
+      "grad_norm": 0.3084274056795106,
+      "learning_rate": 9.894026120141304e-07,
+      "loss": 0.2999,
+      "step": 484
+    },
+    {
+      "epoch": 0.2444556451612903,
+      "grad_norm": 0.23552597211145404,
+      "learning_rate": 9.893475424977143e-07,
+      "loss": 0.3073,
+      "step": 485
+    },
+    {
+      "epoch": 0.24495967741935484,
+      "grad_norm": 1.0356894202770972,
+      "learning_rate": 9.892923318075634e-07,
+      "loss": 0.2958,
+      "step": 486
+    },
+    {
+      "epoch": 0.24546370967741934,
+      "grad_norm": 0.22009519521768645,
+      "learning_rate": 9.892369799596057e-07,
+      "loss": 0.3059,
+      "step": 487
+    },
+    {
+      "epoch": 0.24596774193548387,
+      "grad_norm": 0.2573429172383206,
+      "learning_rate": 9.891814869698101e-07,
+      "loss": 0.2992,
+      "step": 488
+    },
+    {
+      "epoch": 0.24647177419354838,
+      "grad_norm": 0.2350845142780554,
+      "learning_rate": 9.891258528541859e-07,
+      "loss": 0.3033,
+      "step": 489
+    },
+    {
+      "epoch": 0.2469758064516129,
+      "grad_norm": 0.2879929160489008,
+      "learning_rate": 9.89070077628783e-07,
+      "loss": 0.3117,
+      "step": 490
+    },
+    {
+      "epoch": 0.2474798387096774,
+      "grad_norm": 0.25961881540855053,
+      "learning_rate": 9.890141613096924e-07,
+      "loss": 0.3168,
+      "step": 491
+    },
+    {
+      "epoch": 0.24798387096774194,
+      "grad_norm": 0.19180476078696093,
+      "learning_rate": 9.889581039130455e-07,
+      "loss": 0.2868,
+      "step": 492
+    },
+    {
+      "epoch": 0.24848790322580644,
+      "grad_norm": 0.45241056266402024,
+      "learning_rate": 9.889019054550144e-07,
+      "loss": 0.304,
+      "step": 493
+    },
+    {
+      "epoch": 0.24899193548387097,
+      "grad_norm": 0.24790970349975638,
+      "learning_rate": 9.888455659518124e-07,
+      "loss": 0.3188,
+      "step": 494
+    },
+    {
+      "epoch": 0.24949596774193547,
+      "grad_norm": 0.22736377748906864,
+      "learning_rate": 9.887890854196928e-07,
+      "loss": 0.3086,
+      "step": 495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2157688569695134,
+      "learning_rate": 9.8873246387495e-07,
+      "loss": 0.3055,
+      "step": 496
+    },
+    {
+      "epoch": 0.2505040322580645,
+      "grad_norm": 0.2401397569367212,
+      "learning_rate": 9.886757013339188e-07,
+      "loss": 0.2903,
+      "step": 497
+    },
+    {
+      "epoch": 0.25100806451612906,
+      "grad_norm": 0.27724457476206976,
+      "learning_rate": 9.88618797812975e-07,
+      "loss": 0.2809,
+      "step": 498
+    },
+    {
+      "epoch": 0.25151209677419356,
+      "grad_norm": 0.22112121556855863,
+      "learning_rate": 9.885617533285349e-07,
+      "loss": 0.3083,
+      "step": 499
+    },
+    {
+      "epoch": 0.25201612903225806,
+      "grad_norm": 0.26659993041208113,
+      "learning_rate": 9.885045678970554e-07,
+      "loss": 0.2945,
+      "step": 500
+    },
+    {
+      "epoch": 0.25252016129032256,
+      "grad_norm": 0.474336083459808,
+      "learning_rate": 9.884472415350342e-07,
+      "loss": 0.3066,
+      "step": 501
+    },
+    {
+      "epoch": 0.2530241935483871,
+      "grad_norm": 0.3246752836417456,
+      "learning_rate": 9.883897742590094e-07,
+      "loss": 0.315,
+      "step": 502
+    },
+    {
+      "epoch": 0.2535282258064516,
+      "grad_norm": 0.19279895004344805,
+      "learning_rate": 9.883321660855604e-07,
+      "loss": 0.2902,
+      "step": 503
+    },
+    {
+      "epoch": 0.2540322580645161,
+      "grad_norm": 0.26362617716913594,
+      "learning_rate": 9.882744170313065e-07,
+      "loss": 0.3309,
+      "step": 504
+    },
+    {
+      "epoch": 0.2545362903225806,
+      "grad_norm": 0.5083521686121811,
+      "learning_rate": 9.882165271129078e-07,
+      "loss": 0.3039,
+      "step": 505
+    },
+    {
+      "epoch": 0.2550403225806452,
+      "grad_norm": 0.20136858022759457,
+      "learning_rate": 9.881584963470657e-07,
+      "loss": 0.3135,
+      "step": 506
+    },
+    {
+      "epoch": 0.2555443548387097,
+      "grad_norm": 0.24947755269293234,
+      "learning_rate": 9.88100324750521e-07,
+      "loss": 0.3,
+      "step": 507
+    },
+    {
+      "epoch": 0.2560483870967742,
+      "grad_norm": 0.1934113012969022,
+      "learning_rate": 9.880420123400567e-07,
+      "loss": 0.2998,
+      "step": 508
+    },
+    {
+      "epoch": 0.2565524193548387,
+      "grad_norm": 0.19024377903728643,
+      "learning_rate": 9.879835591324947e-07,
+      "loss": 0.3191,
+      "step": 509
+    },
+    {
+      "epoch": 0.25705645161290325,
+      "grad_norm": 0.19125874820918395,
+      "learning_rate": 9.87924965144699e-07,
+      "loss": 0.3075,
+      "step": 510
+    },
+    {
+      "epoch": 0.25756048387096775,
+      "grad_norm": 0.19244259184808277,
+      "learning_rate": 9.878662303935732e-07,
+      "loss": 0.2956,
+      "step": 511
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 0.22509070180606666,
+      "learning_rate": 9.878073548960623e-07,
+      "loss": 0.3218,
+      "step": 512
+    },
+    {
+      "epoch": 0.25856854838709675,
+      "grad_norm": 0.2717698227036736,
+      "learning_rate": 9.877483386691513e-07,
+      "loss": 0.3027,
+      "step": 513
+    },
+    {
+      "epoch": 0.2590725806451613,
+      "grad_norm": 0.1867442771638362,
+      "learning_rate": 9.876891817298658e-07,
+      "loss": 0.2846,
+      "step": 514
+    },
+    {
+      "epoch": 0.2595766129032258,
+      "grad_norm": 0.2377584026295176,
+      "learning_rate": 9.876298840952726e-07,
+      "loss": 0.2849,
+      "step": 515
+    },
+    {
+      "epoch": 0.2600806451612903,
+      "grad_norm": 0.2936917090291921,
+      "learning_rate": 9.875704457824786e-07,
+      "loss": 0.3013,
+      "step": 516
+    },
+    {
+      "epoch": 0.2605846774193548,
+      "grad_norm": 0.4263228296671513,
+      "learning_rate": 9.875108668086313e-07,
+      "loss": 0.2846,
+      "step": 517
+    },
+    {
+      "epoch": 0.2610887096774194,
+      "grad_norm": 0.2718679782241284,
+      "learning_rate": 9.874511471909189e-07,
+      "loss": 0.2842,
+      "step": 518
+    },
+    {
+      "epoch": 0.2615927419354839,
+      "grad_norm": 0.18725434376492076,
+      "learning_rate": 9.873912869465701e-07,
+      "loss": 0.3015,
+      "step": 519
+    },
+    {
+      "epoch": 0.2620967741935484,
+      "grad_norm": 0.19020214607186645,
+      "learning_rate": 9.873312860928541e-07,
+      "loss": 0.3149,
+      "step": 520
+    },
+    {
+      "epoch": 0.2626008064516129,
+      "grad_norm": 0.244741870819488,
+      "learning_rate": 9.87271144647081e-07,
+      "loss": 0.305,
+      "step": 521
+    },
+    {
+      "epoch": 0.26310483870967744,
+      "grad_norm": 0.22885289257455954,
+      "learning_rate": 9.872108626266014e-07,
+      "loss": 0.2933,
+      "step": 522
+    },
+    {
+      "epoch": 0.26360887096774194,
+      "grad_norm": 0.29727559200743137,
+      "learning_rate": 9.871504400488059e-07,
+      "loss": 0.2989,
+      "step": 523
+    },
+    {
+      "epoch": 0.26411290322580644,
+      "grad_norm": 0.3347319040673116,
+      "learning_rate": 9.870898769311261e-07,
+      "loss": 0.3003,
+      "step": 524
+    },
+    {
+      "epoch": 0.26461693548387094,
+      "grad_norm": 0.21285823806164653,
+      "learning_rate": 9.870291732910343e-07,
+      "loss": 0.2909,
+      "step": 525
+    },
+    {
+      "epoch": 0.2651209677419355,
+      "grad_norm": 0.19714692459161542,
+      "learning_rate": 9.86968329146043e-07,
+      "loss": 0.2863,
+      "step": 526
+    },
+    {
+      "epoch": 0.265625,
+      "grad_norm": 0.37762602100369747,
+      "learning_rate": 9.869073445137054e-07,
+      "loss": 0.3072,
+      "step": 527
+    },
+    {
+      "epoch": 0.2661290322580645,
+      "grad_norm": 0.2322570227181064,
+      "learning_rate": 9.868462194116149e-07,
+      "loss": 0.2952,
+      "step": 528
+    },
+    {
+      "epoch": 0.26663306451612906,
+      "grad_norm": 0.25028173693381045,
+      "learning_rate": 9.86784953857406e-07,
+      "loss": 0.2979,
+      "step": 529
+    },
+    {
+      "epoch": 0.26713709677419356,
+      "grad_norm": 0.23342075160977557,
+      "learning_rate": 9.867235478687534e-07,
+      "loss": 0.3029,
+      "step": 530
+    },
+    {
+      "epoch": 0.26764112903225806,
+      "grad_norm": 0.2202223368280641,
+      "learning_rate": 9.866620014633725e-07,
+      "loss": 0.2995,
+      "step": 531
+    },
+    {
+      "epoch": 0.26814516129032256,
+      "grad_norm": 0.40688579823270754,
+      "learning_rate": 9.866003146590186e-07,
+      "loss": 0.2998,
+      "step": 532
+    },
+    {
+      "epoch": 0.2686491935483871,
+      "grad_norm": 0.23558843514280095,
+      "learning_rate": 9.865384874734886e-07,
+      "loss": 0.2919,
+      "step": 533
+    },
+    {
+      "epoch": 0.2691532258064516,
+      "grad_norm": 0.350876653755429,
+      "learning_rate": 9.864765199246187e-07,
+      "loss": 0.2979,
+      "step": 534
+    },
+    {
+      "epoch": 0.2696572580645161,
+      "grad_norm": 0.5484769536987208,
+      "learning_rate": 9.864144120302865e-07,
+      "loss": 0.3098,
+      "step": 535
+    },
+    {
+      "epoch": 0.2701612903225806,
+      "grad_norm": 0.20073285824672496,
+      "learning_rate": 9.863521638084093e-07,
+      "loss": 0.299,
+      "step": 536
+    },
+    {
+      "epoch": 0.2706653225806452,
+      "grad_norm": 0.18356884936864395,
+      "learning_rate": 9.86289775276946e-07,
+      "loss": 0.3042,
+      "step": 537
+    },
+    {
+      "epoch": 0.2711693548387097,
+      "grad_norm": 0.2886255584956994,
+      "learning_rate": 9.862272464538946e-07,
+      "loss": 0.3132,
+      "step": 538
+    },
+    {
+      "epoch": 0.2716733870967742,
+      "grad_norm": 0.2658005187689356,
+      "learning_rate": 9.86164577357295e-07,
+      "loss": 0.308,
+      "step": 539
+    },
+    {
+      "epoch": 0.2721774193548387,
+      "grad_norm": 0.36659782067418106,
+      "learning_rate": 9.861017680052262e-07,
+      "loss": 0.2983,
+      "step": 540
+    },
+    {
+      "epoch": 0.27268145161290325,
+      "grad_norm": 0.26876552668494674,
+      "learning_rate": 9.860388184158086e-07,
+      "loss": 0.3026,
+      "step": 541
+    },
+    {
+      "epoch": 0.27318548387096775,
+      "grad_norm": 0.18123995001034987,
+      "learning_rate": 9.859757286072028e-07,
+      "loss": 0.2984,
+      "step": 542
+    },
+    {
+      "epoch": 0.27368951612903225,
+      "grad_norm": 0.3490662102838747,
+      "learning_rate": 9.859124985976097e-07,
+      "loss": 0.3006,
+      "step": 543
+    },
+    {
+      "epoch": 0.27419354838709675,
+      "grad_norm": 0.5077591494356637,
+      "learning_rate": 9.858491284052708e-07,
+      "loss": 0.2942,
+      "step": 544
+    },
+    {
+      "epoch": 0.2746975806451613,
+      "grad_norm": 0.1899610370247458,
+      "learning_rate": 9.857856180484682e-07,
+      "loss": 0.2952,
+      "step": 545
+    },
+    {
+      "epoch": 0.2752016129032258,
+      "grad_norm": 0.33906573112018934,
+      "learning_rate": 9.857219675455236e-07,
+      "loss": 0.2825,
+      "step": 546
+    },
+    {
+      "epoch": 0.2757056451612903,
+      "grad_norm": 0.24179860042616932,
+      "learning_rate": 9.856581769148007e-07,
+      "loss": 0.2963,
+      "step": 547
+    },
+    {
+      "epoch": 0.2762096774193548,
+      "grad_norm": 0.24173305167642747,
+      "learning_rate": 9.855942461747023e-07,
+      "loss": 0.3157,
+      "step": 548
+    },
+    {
+      "epoch": 0.2767137096774194,
+      "grad_norm": 0.28928103551655465,
+      "learning_rate": 9.855301753436718e-07,
+      "loss": 0.3087,
+      "step": 549
+    },
+    {
+      "epoch": 0.2772177419354839,
+      "grad_norm": 0.18922268926852628,
+      "learning_rate": 9.854659644401934e-07,
+      "loss": 0.3134,
+      "step": 550
+    },
+    {
+      "epoch": 0.2777217741935484,
+      "grad_norm": 0.20784139183584166,
+      "learning_rate": 9.854016134827916e-07,
+      "loss": 0.3051,
+      "step": 551
+    },
+    {
+      "epoch": 0.2782258064516129,
+      "grad_norm": 0.5667720250099785,
+      "learning_rate": 9.853371224900313e-07,
+      "loss": 0.2847,
+      "step": 552
+    },
+    {
+      "epoch": 0.27872983870967744,
+      "grad_norm": 0.34781718527992644,
+      "learning_rate": 9.852724914805175e-07,
+      "loss": 0.2873,
+      "step": 553
+    },
+    {
+      "epoch": 0.27923387096774194,
+      "grad_norm": 0.19716253139941464,
+      "learning_rate": 9.852077204728961e-07,
+      "loss": 0.3027,
+      "step": 554
+    },
+    {
+      "epoch": 0.27973790322580644,
+      "grad_norm": 0.19344190301481812,
+      "learning_rate": 9.85142809485853e-07,
+      "loss": 0.2897,
+      "step": 555
+    },
+    {
+      "epoch": 0.28024193548387094,
+      "grad_norm": 0.7127285715175351,
+      "learning_rate": 9.850777585381146e-07,
+      "loss": 0.2866,
+      "step": 556
+    },
+    {
+      "epoch": 0.2807459677419355,
+      "grad_norm": 0.40797443381529386,
+      "learning_rate": 9.85012567648448e-07,
+      "loss": 0.3056,
+      "step": 557
+    },
+    {
+      "epoch": 0.28125,
+      "grad_norm": 0.2136697521115876,
+      "learning_rate": 9.8494723683566e-07,
+      "loss": 0.3064,
+      "step": 558
+    },
+    {
+      "epoch": 0.2817540322580645,
+      "grad_norm": 0.26165293739971646,
+      "learning_rate": 9.848817661185984e-07,
+      "loss": 0.2921,
+      "step": 559
+    },
+    {
+      "epoch": 0.28225806451612906,
+      "grad_norm": 0.1846995068150634,
+      "learning_rate": 9.848161555161507e-07,
+      "loss": 0.3131,
+      "step": 560
+    },
+    {
+      "epoch": 0.28276209677419356,
+      "grad_norm": 1.0083499880979159,
+      "learning_rate": 9.847504050472454e-07,
+      "loss": 0.3053,
+      "step": 561
+    },
+    {
+      "epoch": 0.28326612903225806,
+      "grad_norm": 0.6105046373752532,
+      "learning_rate": 9.846845147308514e-07,
+      "loss": 0.3082,
+      "step": 562
+    },
+    {
+      "epoch": 0.28377016129032256,
+      "grad_norm": 0.32182959498303215,
+      "learning_rate": 9.846184845859772e-07,
+      "loss": 0.2886,
+      "step": 563
+    },
+    {
+      "epoch": 0.2842741935483871,
+      "grad_norm": 0.5616737715729443,
+      "learning_rate": 9.845523146316722e-07,
+      "loss": 0.3138,
+      "step": 564
+    },
+    {
+      "epoch": 0.2847782258064516,
+      "grad_norm": 0.34317343338743084,
+      "learning_rate": 9.844860048870261e-07,
+      "loss": 0.3058,
+      "step": 565
+    },
+    {
+      "epoch": 0.2852822580645161,
+      "grad_norm": 0.38672910543834893,
+      "learning_rate": 9.84419555371169e-07,
+      "loss": 0.2991,
+      "step": 566
+    },
+    {
+      "epoch": 0.2857862903225806,
+      "grad_norm": 0.27114300647024586,
+      "learning_rate": 9.843529661032706e-07,
+      "loss": 0.3083,
+      "step": 567
+    },
+    {
+      "epoch": 0.2862903225806452,
+      "grad_norm": 0.1896232339668482,
+      "learning_rate": 9.842862371025422e-07,
+      "loss": 0.3025,
+      "step": 568
+    },
+    {
+      "epoch": 0.2867943548387097,
+      "grad_norm": 0.3930098230344227,
+      "learning_rate": 9.842193683882344e-07,
+      "loss": 0.3079,
+      "step": 569
+    },
+    {
+      "epoch": 0.2872983870967742,
+      "grad_norm": 0.43124046850579134,
+      "learning_rate": 9.841523599796382e-07,
+      "loss": 0.2777,
+      "step": 570
+    },
+    {
+      "epoch": 0.2878024193548387,
+      "grad_norm": 0.7587117457845125,
+      "learning_rate": 9.840852118960853e-07,
+      "loss": 0.3064,
+      "step": 571
+    },
+    {
+      "epoch": 0.28830645161290325,
+      "grad_norm": 0.3117074250999561,
+      "learning_rate": 9.840179241569478e-07,
+      "loss": 0.299,
+      "step": 572
+    },
+    {
+      "epoch": 0.28881048387096775,
+      "grad_norm": 0.18731266500576274,
+      "learning_rate": 9.839504967816374e-07,
+      "loss": 0.3035,
+      "step": 573
+    },
+    {
+      "epoch": 0.28931451612903225,
+      "grad_norm": 0.43778747829645503,
+      "learning_rate": 9.838829297896065e-07,
+      "loss": 0.319,
+      "step": 574
+    },
+    {
+      "epoch": 0.28981854838709675,
+      "grad_norm": 0.40776414993659865,
+      "learning_rate": 9.83815223200348e-07,
+      "loss": 0.2955,
+      "step": 575
+    },
+    {
+      "epoch": 0.2903225806451613,
+      "grad_norm": 0.25990172322006877,
+      "learning_rate": 9.837473770333945e-07,
+      "loss": 0.3132,
+      "step": 576
+    },
+    {
+      "epoch": 0.2908266129032258,
+      "grad_norm": 0.2296880045553122,
+      "learning_rate": 9.836793913083195e-07,
+      "loss": 0.2936,
+      "step": 577
+    },
+    {
+      "epoch": 0.2913306451612903,
+      "grad_norm": 0.1953127561261216,
+      "learning_rate": 9.836112660447362e-07,
+      "loss": 0.3138,
+      "step": 578
+    },
+    {
+      "epoch": 0.2918346774193548,
+      "grad_norm": 0.27208394851949347,
+      "learning_rate": 9.835430012622988e-07,
+      "loss": 0.313,
+      "step": 579
+    },
+    {
+      "epoch": 0.2923387096774194,
+      "grad_norm": 0.2877633978499272,
+      "learning_rate": 9.834745969807006e-07,
+      "loss": 0.3013,
+      "step": 580
+    },
+    {
+      "epoch": 0.2928427419354839,
+      "grad_norm": 0.20402433175664483,
+      "learning_rate": 9.834060532196761e-07,
+      "loss": 0.2948,
+      "step": 581
+    },
+    {
+      "epoch": 0.2933467741935484,
+      "grad_norm": 0.3541481547851912,
+      "learning_rate": 9.833373699989999e-07,
+      "loss": 0.3057,
+      "step": 582
+    },
+    {
+      "epoch": 0.2938508064516129,
+      "grad_norm": 0.25513070455002546,
+      "learning_rate": 9.832685473384868e-07,
+      "loss": 0.2972,
+      "step": 583
+    },
+    {
+      "epoch": 0.29435483870967744,
+      "grad_norm": 0.6078105436530532,
+      "learning_rate": 9.83199585257991e-07,
+      "loss": 0.2974,
+      "step": 584
+    },
+    {
+      "epoch": 0.29485887096774194,
+      "grad_norm": 0.2701844685693586,
+      "learning_rate": 9.831304837774086e-07,
+      "loss": 0.314,
+      "step": 585
+    },
+    {
+      "epoch": 0.29536290322580644,
+      "grad_norm": 0.19399840828869006,
+      "learning_rate": 9.830612429166743e-07,
+      "loss": 0.3031,
+      "step": 586
+    },
+    {
+      "epoch": 0.29586693548387094,
+      "grad_norm": 0.184738711676414,
+      "learning_rate": 9.829918626957635e-07,
+      "loss": 0.3031,
+      "step": 587
+    },
+    {
+      "epoch": 0.2963709677419355,
+      "grad_norm": 0.3269160031771926,
+      "learning_rate": 9.829223431346926e-07,
+      "loss": 0.2911,
+      "step": 588
+    },
+    {
+      "epoch": 0.296875,
+      "grad_norm": 0.38332552633652756,
+      "learning_rate": 9.828526842535174e-07,
+      "loss": 0.2864,
+      "step": 589
+    },
+    {
+      "epoch": 0.2973790322580645,
+      "grad_norm": 0.21468223114928167,
+      "learning_rate": 9.827828860723338e-07,
+      "loss": 0.3109,
+      "step": 590
+    },
+    {
+      "epoch": 0.29788306451612906,
+      "grad_norm": 0.18987287925768598,
+      "learning_rate": 9.827129486112782e-07,
+      "loss": 0.3011,
+      "step": 591
+    },
+    {
+      "epoch": 0.29838709677419356,
+      "grad_norm": 0.3156090049134527,
+      "learning_rate": 9.82642871890527e-07,
+      "loss": 0.2906,
+      "step": 592
+    },
+    {
+      "epoch": 0.29889112903225806,
+      "grad_norm": 0.40698452999719154,
+      "learning_rate": 9.82572655930297e-07,
+      "loss": 0.306,
+      "step": 593
+    },
+    {
+      "epoch": 0.29939516129032256,
+      "grad_norm": 0.24367422580598258,
+      "learning_rate": 9.825023007508456e-07,
+      "loss": 0.3006,
+      "step": 594
+    },
+    {
+      "epoch": 0.2998991935483871,
+      "grad_norm": 0.27250412398540247,
+      "learning_rate": 9.82431806372469e-07,
+      "loss": 0.2989,
+      "step": 595
+    },
+    {
+      "epoch": 0.3004032258064516,
+      "grad_norm": 0.19246483798552555,
+      "learning_rate": 9.82361172815505e-07,
+      "loss": 0.2973,
+      "step": 596
+    },
+    {
+      "epoch": 0.3009072580645161,
+      "grad_norm": 0.18545484272367835,
+      "learning_rate": 9.822904001003306e-07,
+      "loss": 0.2996,
+      "step": 597
+    },
+    {
+      "epoch": 0.3009072580645161,
+      "eval_loss": 0.32989731431007385,
+      "eval_runtime": 18.7201,
+      "eval_samples_per_second": 45.673,
+      "eval_steps_per_second": 0.962,
+      "step": 597
+    },
+    {
+      "epoch": 0.3014112903225806,
+      "grad_norm": 0.5063679318757863,
+      "learning_rate": 9.822194882473635e-07,
+      "loss": 0.2864,
+      "step": 598
+    },
+    {
+      "epoch": 0.3019153225806452,
+      "grad_norm": 0.2855392122632754,
+      "learning_rate": 9.821484372770612e-07,
+      "loss": 0.295,
+      "step": 599
+    },
+    {
+      "epoch": 0.3024193548387097,
+      "grad_norm": 0.6735980580137383,
+      "learning_rate": 9.820772472099215e-07,
+      "loss": 0.2951,
+      "step": 600
+    },
+    {
+      "epoch": 0.3029233870967742,
+      "grad_norm": 0.35135382164679746,
+      "learning_rate": 9.82005918066482e-07,
+      "loss": 0.31,
+      "step": 601
+    },
+    {
+      "epoch": 0.3034274193548387,
+      "grad_norm": 0.19875119958210918,
+      "learning_rate": 9.819344498673215e-07,
+      "loss": 0.2956,
+      "step": 602
+    },
+    {
+      "epoch": 0.30393145161290325,
+      "grad_norm": 0.9315288935617498,
+      "learning_rate": 9.818628426330574e-07,
+      "loss": 0.3217,
+      "step": 603
+    },
+    {
+      "epoch": 0.30443548387096775,
+      "grad_norm": 1.0733717026504053,
+      "learning_rate": 9.817910963843481e-07,
+      "loss": 0.302,
+      "step": 604
+    },
+    {
+      "epoch": 0.30493951612903225,
+      "grad_norm": 0.1821439303858277,
+      "learning_rate": 9.81719211141892e-07,
+      "loss": 0.2951,
+      "step": 605
+    },
+    {
+      "epoch": 0.30544354838709675,
+      "grad_norm": 0.2227625770904509,
+      "learning_rate": 9.816471869264278e-07,
+      "loss": 0.2962,
+      "step": 606
+    },
+    {
+      "epoch": 0.3059475806451613,
+      "grad_norm": 0.6196036153090734,
+      "learning_rate": 9.815750237587338e-07,
+      "loss": 0.3029,
+      "step": 607
+    },
+    {
+      "epoch": 0.3064516129032258,
+      "grad_norm": 0.6010581227628021,
+      "learning_rate": 9.815027216596282e-07,
+      "loss": 0.3046,
+      "step": 608
+    },
+    {
+      "epoch": 0.3069556451612903,
+      "grad_norm": 0.3528611360569492,
+      "learning_rate": 9.814302806499707e-07,
+      "loss": 0.2932,
+      "step": 609
+    },
+    {
+      "epoch": 0.3074596774193548,
+      "grad_norm": 0.2979313249836626,
+      "learning_rate": 9.813577007506594e-07,
+      "loss": 0.2899,
+      "step": 610
+    },
+    {
+      "epoch": 0.3079637096774194,
+      "grad_norm": 0.489841762858414,
+      "learning_rate": 9.81284981982633e-07,
+      "loss": 0.2887,
+      "step": 611
+    },
+    {
+      "epoch": 0.3084677419354839,
+      "grad_norm": 0.32286398990319887,
+      "learning_rate": 9.81212124366871e-07,
+      "loss": 0.2972,
+      "step": 612
+    },
+    {
+      "epoch": 0.3089717741935484,
+      "grad_norm": 0.3654449282383879,
+      "learning_rate": 9.81139127924392e-07,
+      "loss": 0.2817,
+      "step": 613
+    },
+    {
+      "epoch": 0.3094758064516129,
+      "grad_norm": 0.23506329847966434,
+      "learning_rate": 9.810659926762551e-07,
+      "loss": 0.3041,
+      "step": 614
+    },
+    {
+      "epoch": 0.30997983870967744,
+      "grad_norm": 0.18739033718362563,
+      "learning_rate": 9.809927186435594e-07,
+      "loss": 0.3039,
+      "step": 615
+    },
+    {
+      "epoch": 0.31048387096774194,
+      "grad_norm": 0.4424195854136909,
+      "learning_rate": 9.809193058474438e-07,
+      "loss": 0.3027,
+      "step": 616
+    },
+    {
+      "epoch": 0.31098790322580644,
+      "grad_norm": 0.34612288423111537,
+      "learning_rate": 9.808457543090878e-07,
+      "loss": 0.2893,
+      "step": 617
+    },
+    {
+      "epoch": 0.31149193548387094,
+      "grad_norm": 0.6207470476564412,
+      "learning_rate": 9.807720640497103e-07,
+      "loss": 0.3089,
+      "step": 618
+    },
+    {
+      "epoch": 0.3119959677419355,
+      "grad_norm": 0.5606872023789801,
+      "learning_rate": 9.806982350905703e-07,
+      "loss": 0.3051,
+      "step": 619
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.6658134114343093,
+      "learning_rate": 9.806242674529676e-07,
+      "loss": 0.2879,
+      "step": 620
+    },
+    {
+      "epoch": 0.3130040322580645,
+      "grad_norm": 0.650400551133684,
+      "learning_rate": 9.80550161158241e-07,
+      "loss": 0.2998,
+      "step": 621
+    },
+    {
+      "epoch": 0.31350806451612906,
+      "grad_norm": 0.17947781258266207,
+      "learning_rate": 9.804759162277696e-07,
+      "loss": 0.3134,
+      "step": 622
+    },
+    {
+      "epoch": 0.31401209677419356,
+      "grad_norm": 0.3273834000987335,
+      "learning_rate": 9.804015326829728e-07,
+      "loss": 0.3108,
+      "step": 623
+    },
+    {
+      "epoch": 0.31451612903225806,
+      "grad_norm": 0.27388099966321255,
+      "learning_rate": 9.803270105453098e-07,
+      "loss": 0.2995,
+      "step": 624
+    },
+    {
+      "epoch": 0.31502016129032256,
+      "grad_norm": 0.4361569565465292,
+      "learning_rate": 9.802523498362797e-07,
+      "loss": 0.3037,
+      "step": 625
+    },
+    {
+      "epoch": 0.3155241935483871,
+      "grad_norm": 0.43147162718536736,
+      "learning_rate": 9.801775505774218e-07,
+      "loss": 0.3142,
+      "step": 626
+    },
+    {
+      "epoch": 0.3160282258064516,
+      "grad_norm": 0.23130882783317155,
+      "learning_rate": 9.801026127903149e-07,
+      "loss": 0.2981,
+      "step": 627
+    },
+    {
+      "epoch": 0.3165322580645161,
+      "grad_norm": 0.42494886746355115,
+      "learning_rate": 9.800275364965782e-07,
+      "loss": 0.3083,
+      "step": 628
+    },
+    {
+      "epoch": 0.3170362903225806,
+      "grad_norm": 0.23979029210918654,
+      "learning_rate": 9.79952321717871e-07,
+      "loss": 0.2927,
+      "step": 629
+    },
+    {
+      "epoch": 0.3175403225806452,
+      "grad_norm": 0.2649450830135195,
+      "learning_rate": 9.798769684758924e-07,
+      "loss": 0.3055,
+      "step": 630
+    },
+    {
+      "epoch": 0.3180443548387097,
+      "grad_norm": 0.20691288403035124,
+      "learning_rate": 9.798014767923807e-07,
+      "loss": 0.2939,
+      "step": 631
+    },
+    {
+      "epoch": 0.3185483870967742,
+      "grad_norm": 0.2818136169636899,
+      "learning_rate": 9.797258466891152e-07,
+      "loss": 0.3034,
+      "step": 632
+    },
+    {
+      "epoch": 0.3190524193548387,
+      "grad_norm": 0.1855253307389335,
+      "learning_rate": 9.796500781879148e-07,
+      "loss": 0.2845,
+      "step": 633
+    },
+    {
+      "epoch": 0.31955645161290325,
+      "grad_norm": 0.2383083001412052,
+      "learning_rate": 9.79574171310638e-07,
+      "loss": 0.2999,
+      "step": 634
+    },
+    {
+      "epoch": 0.32006048387096775,
+      "grad_norm": 0.2155300521952254,
+      "learning_rate": 9.794981260791837e-07,
+      "loss": 0.3115,
+      "step": 635
+    },
+    {
+      "epoch": 0.32056451612903225,
+      "grad_norm": 0.59372804366046,
+      "learning_rate": 9.794219425154904e-07,
+      "loss": 0.3057,
+      "step": 636
+    },
+    {
+      "epoch": 0.32106854838709675,
+      "grad_norm": 0.17584328105654679,
+      "learning_rate": 9.793456206415362e-07,
+      "loss": 0.3007,
+      "step": 637
+    },
+    {
+      "epoch": 0.3215725806451613,
+      "grad_norm": 0.2442097318205203,
+      "learning_rate": 9.792691604793402e-07,
+      "loss": 0.2897,
+      "step": 638
+    },
+    {
+      "epoch": 0.3220766129032258,
+      "grad_norm": 0.17991513417504787,
+      "learning_rate": 9.791925620509603e-07,
+      "loss": 0.289,
+      "step": 639
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 0.2022189687655052,
+      "learning_rate": 9.791158253784945e-07,
+      "loss": 0.3072,
+      "step": 640
+    },
+    {
+      "epoch": 0.3230846774193548,
+      "grad_norm": 0.3259212247774595,
+      "learning_rate": 9.79038950484081e-07,
+      "loss": 0.2989,
+      "step": 641
+    },
+    {
+      "epoch": 0.3235887096774194,
+      "grad_norm": 0.33045805468173267,
+      "learning_rate": 9.789619373898981e-07,
+      "loss": 0.3038,
+      "step": 642
+    },
+    {
+      "epoch": 0.3240927419354839,
+      "grad_norm": 0.24681922232019723,
+      "learning_rate": 9.788847861181631e-07,
+      "loss": 0.3138,
+      "step": 643
+    },
+    {
+      "epoch": 0.3245967741935484,
+      "grad_norm": 0.2518300810532734,
+      "learning_rate": 9.788074966911337e-07,
+      "loss": 0.2899,
+      "step": 644
+    },
+    {
+      "epoch": 0.3251008064516129,
+      "grad_norm": 0.29278633614058125,
+      "learning_rate": 9.787300691311077e-07,
+      "loss": 0.3139,
+      "step": 645
+    },
+    {
+      "epoch": 0.32560483870967744,
+      "grad_norm": 0.21645406918937576,
+      "learning_rate": 9.786525034604224e-07,
+      "loss": 0.309,
+      "step": 646
+    },
+    {
+      "epoch": 0.32610887096774194,
+      "grad_norm": 0.3666140583384514,
+      "learning_rate": 9.785747997014547e-07,
+      "loss": 0.2949,
+      "step": 647
+    },
+    {
+      "epoch": 0.32661290322580644,
+      "grad_norm": 0.1947942753801369,
+      "learning_rate": 9.78496957876622e-07,
+      "loss": 0.3075,
+      "step": 648
+    },
+    {
+      "epoch": 0.32711693548387094,
+      "grad_norm": 0.18162530785958844,
+      "learning_rate": 9.784189780083812e-07,
+      "loss": 0.294,
+      "step": 649
+    },
+    {
+      "epoch": 0.3276209677419355,
+      "grad_norm": 0.28107058253792394,
+      "learning_rate": 9.78340860119229e-07,
+      "loss": 0.2927,
+      "step": 650
+    },
+    {
+      "epoch": 0.328125,
+      "grad_norm": 0.1863145084715224,
+      "learning_rate": 9.782626042317015e-07,
+      "loss": 0.2977,
+      "step": 651
+    },
+    {
+      "epoch": 0.3286290322580645,
+      "grad_norm": 0.19255877376409836,
+      "learning_rate": 9.781842103683756e-07,
+      "loss": 0.3242,
+      "step": 652
+    },
+    {
+      "epoch": 0.32913306451612906,
+      "grad_norm": 0.22119159460683763,
+      "learning_rate": 9.78105678551867e-07,
+      "loss": 0.2949,
+      "step": 653
+    },
+    {
+      "epoch": 0.32963709677419356,
+      "grad_norm": 0.3370309906377243,
+      "learning_rate": 9.78027008804832e-07,
+      "loss": 0.3281,
+      "step": 654
+    },
+    {
+      "epoch": 0.33014112903225806,
+      "grad_norm": 0.21007597885166748,
+      "learning_rate": 9.779482011499662e-07,
+      "loss": 0.3011,
+      "step": 655
+    },
+    {
+      "epoch": 0.33064516129032256,
+      "grad_norm": 0.2114458853983953,
+      "learning_rate": 9.77869255610005e-07,
+      "loss": 0.3144,
+      "step": 656
+    },
+    {
+      "epoch": 0.3311491935483871,
+      "grad_norm": 0.4562451777517263,
+      "learning_rate": 9.77790172207724e-07,
+      "loss": 0.2804,
+      "step": 657
+    },
+    {
+      "epoch": 0.3316532258064516,
+      "grad_norm": 0.4033116791658588,
+      "learning_rate": 9.777109509659378e-07,
+      "loss": 0.2896,
+      "step": 658
+    },
+    {
+      "epoch": 0.3321572580645161,
+      "grad_norm": 0.19637371460146463,
+      "learning_rate": 9.776315919075015e-07,
+      "loss": 0.2851,
+      "step": 659
+    },
+    {
+      "epoch": 0.3326612903225806,
+      "grad_norm": 0.23427740658090485,
+      "learning_rate": 9.7755209505531e-07,
+      "loss": 0.2888,
+      "step": 660
+    },
+    {
+      "epoch": 0.3331653225806452,
+      "grad_norm": 0.2742494578051774,
+      "learning_rate": 9.77472460432297e-07,
+      "loss": 0.2811,
+      "step": 661
+    },
+    {
+      "epoch": 0.3336693548387097,
+      "grad_norm": 0.4976330279305446,
+      "learning_rate": 9.77392688061437e-07,
+      "loss": 0.2981,
+      "step": 662
+    },
+    {
+      "epoch": 0.3341733870967742,
+      "grad_norm": 0.20762430916034352,
+      "learning_rate": 9.773127779657442e-07,
+      "loss": 0.2928,
+      "step": 663
+    },
+    {
+      "epoch": 0.3346774193548387,
+      "grad_norm": 0.1974247496986688,
+      "learning_rate": 9.772327301682714e-07,
+      "loss": 0.2983,
+      "step": 664
+    },
+    {
+      "epoch": 0.33518145161290325,
+      "grad_norm": 0.2693856221779068,
+      "learning_rate": 9.771525446921123e-07,
+      "loss": 0.2966,
+      "step": 665
+    },
+    {
+      "epoch": 0.33568548387096775,
+      "grad_norm": 0.18606159238485928,
+      "learning_rate": 9.770722215604e-07,
+      "loss": 0.2989,
+      "step": 666
+    },
+    {
+      "epoch": 0.33618951612903225,
+      "grad_norm": 0.375528189461835,
+      "learning_rate": 9.769917607963068e-07,
+      "loss": 0.2908,
+      "step": 667
+    },
+    {
+      "epoch": 0.33669354838709675,
+      "grad_norm": 0.2859681728471663,
+      "learning_rate": 9.769111624230457e-07,
+      "loss": 0.2918,
+      "step": 668
+    },
+    {
+      "epoch": 0.3371975806451613,
+      "grad_norm": 0.20521406800135222,
+      "learning_rate": 9.768304264638684e-07,
+      "loss": 0.2987,
+      "step": 669
+    },
+    {
+      "epoch": 0.3377016129032258,
+      "grad_norm": 0.23397409956060417,
+      "learning_rate": 9.76749552942067e-07,
+      "loss": 0.3029,
+      "step": 670
+    },
+    {
+      "epoch": 0.3382056451612903,
+      "grad_norm": 0.4520922291898478,
+      "learning_rate": 9.766685418809727e-07,
+      "loss": 0.2963,
+      "step": 671
+    },
+    {
+      "epoch": 0.3387096774193548,
+      "grad_norm": 0.18452310561487084,
+      "learning_rate": 9.76587393303957e-07,
+      "loss": 0.2931,
+      "step": 672
+    },
+    {
+      "epoch": 0.3392137096774194,
+      "grad_norm": 0.22095628174169515,
+      "learning_rate": 9.765061072344305e-07,
+      "loss": 0.293,
+      "step": 673
+    },
+    {
+      "epoch": 0.3397177419354839,
+      "grad_norm": 0.45292152421659004,
+      "learning_rate": 9.764246836958439e-07,
+      "loss": 0.3098,
+      "step": 674
+    },
+    {
+      "epoch": 0.3402217741935484,
+      "grad_norm": 0.22491373287467453,
+      "learning_rate": 9.763431227116875e-07,
+      "loss": 0.297,
+      "step": 675
+    },
+    {
+      "epoch": 0.3407258064516129,
+      "grad_norm": 0.36919058274186534,
+      "learning_rate": 9.76261424305491e-07,
+      "loss": 0.3075,
+      "step": 676
+    },
+    {
+      "epoch": 0.34122983870967744,
+      "grad_norm": 0.3042293571669918,
+      "learning_rate": 9.761795885008236e-07,
+      "loss": 0.3018,
+      "step": 677
+    },
+    {
+      "epoch": 0.34173387096774194,
+      "grad_norm": 0.30087168043463913,
+      "learning_rate": 9.76097615321295e-07,
+      "loss": 0.2985,
+      "step": 678
+    },
+    {
+      "epoch": 0.34223790322580644,
+      "grad_norm": 0.24352689031383537,
+      "learning_rate": 9.760155047905534e-07,
+      "loss": 0.2938,
+      "step": 679
+    },
+    {
+      "epoch": 0.34274193548387094,
+      "grad_norm": 0.25871019913875265,
+      "learning_rate": 9.759332569322876e-07,
+      "loss": 0.2766,
+      "step": 680
+    },
+    {
+      "epoch": 0.3432459677419355,
+      "grad_norm": 0.3732275698199399,
+      "learning_rate": 9.758508717702253e-07,
+      "loss": 0.3068,
+      "step": 681
+    },
+    {
+      "epoch": 0.34375,
+      "grad_norm": 0.18175157181935991,
+      "learning_rate": 9.757683493281343e-07,
+      "loss": 0.3096,
+      "step": 682
+    },
+    {
+      "epoch": 0.3442540322580645,
+      "grad_norm": 0.3020686172952058,
+      "learning_rate": 9.756856896298218e-07,
+      "loss": 0.2874,
+      "step": 683
+    },
+    {
+      "epoch": 0.34475806451612906,
+      "grad_norm": 0.4287852398031048,
+      "learning_rate": 9.756028926991344e-07,
+      "loss": 0.3064,
+      "step": 684
+    },
+    {
+      "epoch": 0.34526209677419356,
+      "grad_norm": 0.3877586447077066,
+      "learning_rate": 9.755199585599587e-07,
+      "loss": 0.3034,
+      "step": 685
+    },
+    {
+      "epoch": 0.34576612903225806,
+      "grad_norm": 0.42768152833859835,
+      "learning_rate": 9.754368872362209e-07,
+      "loss": 0.3066,
+      "step": 686
+    },
+    {
+      "epoch": 0.34627016129032256,
+      "grad_norm": 0.2249355344351282,
+      "learning_rate": 9.75353678751886e-07,
+      "loss": 0.2882,
+      "step": 687
+    },
+    {
+      "epoch": 0.3467741935483871,
+      "grad_norm": 0.27912853498203316,
+      "learning_rate": 9.752703331309596e-07,
+      "loss": 0.3041,
+      "step": 688
+    },
+    {
+      "epoch": 0.3472782258064516,
+      "grad_norm": 0.43506449721846435,
+      "learning_rate": 9.751868503974862e-07,
+      "loss": 0.2846,
+      "step": 689
+    },
+    {
+      "epoch": 0.3477822580645161,
+      "grad_norm": 0.26206165096276,
+      "learning_rate": 9.7510323057555e-07,
+      "loss": 0.3068,
+      "step": 690
+    },
+    {
+      "epoch": 0.3482862903225806,
+      "grad_norm": 0.20904315219220723,
+      "learning_rate": 9.75019473689275e-07,
+      "loss": 0.2997,
+      "step": 691
+    },
+    {
+      "epoch": 0.3487903225806452,
+      "grad_norm": 0.19583961501100697,
+      "learning_rate": 9.749355797628243e-07,
+      "loss": 0.2819,
+      "step": 692
+    },
+    {
+      "epoch": 0.3492943548387097,
+      "grad_norm": 0.4477203330124689,
+      "learning_rate": 9.74851548820401e-07,
+      "loss": 0.2889,
+      "step": 693
+    },
+    {
+      "epoch": 0.3497983870967742,
+      "grad_norm": 0.465876564157471,
+      "learning_rate": 9.747673808862476e-07,
+      "loss": 0.2927,
+      "step": 694
+    },
+    {
+      "epoch": 0.3503024193548387,
+      "grad_norm": 0.19047436893148592,
+      "learning_rate": 9.746830759846456e-07,
+      "loss": 0.2972,
+      "step": 695
+    },
+    {
+      "epoch": 0.35080645161290325,
+      "grad_norm": 0.20298210286927948,
+      "learning_rate": 9.745986341399166e-07,
+      "loss": 0.2865,
+      "step": 696
+    },
+    {
+      "epoch": 0.35131048387096775,
+      "grad_norm": 0.21381675798513802,
+      "learning_rate": 9.745140553764219e-07,
+      "loss": 0.3057,
+      "step": 697
+    },
+    {
+      "epoch": 0.35181451612903225,
+      "grad_norm": 0.29038109612051655,
+      "learning_rate": 9.744293397185615e-07,
+      "loss": 0.2929,
+      "step": 698
+    },
+    {
+      "epoch": 0.35231854838709675,
+      "grad_norm": 0.3321256067582296,
+      "learning_rate": 9.743444871907756e-07,
+      "loss": 0.2896,
+      "step": 699
+    },
+    {
+      "epoch": 0.3528225806451613,
+      "grad_norm": 0.3056406476467989,
+      "learning_rate": 9.742594978175436e-07,
+      "loss": 0.2996,
+      "step": 700
+    },
+    {
+      "epoch": 0.3533266129032258,
+      "grad_norm": 0.22979049368338594,
+      "learning_rate": 9.741743716233843e-07,
+      "loss": 0.2879,
+      "step": 701
+    },
+    {
+      "epoch": 0.3538306451612903,
+      "grad_norm": 0.2167981736696398,
+      "learning_rate": 9.74089108632856e-07,
+      "loss": 0.3008,
+      "step": 702
+    },
+    {
+      "epoch": 0.3543346774193548,
+      "grad_norm": 0.6466745464300053,
+      "learning_rate": 9.74003708870557e-07,
+      "loss": 0.2883,
+      "step": 703
+    },
+    {
+      "epoch": 0.3548387096774194,
+      "grad_norm": 0.5205046458590996,
+      "learning_rate": 9.73918172361124e-07,
+      "loss": 0.2941,
+      "step": 704
+    },
+    {
+      "epoch": 0.3553427419354839,
+      "grad_norm": 0.4750444899997186,
+      "learning_rate": 9.73832499129234e-07,
+      "loss": 0.2978,
+      "step": 705
+    },
+    {
+      "epoch": 0.3558467741935484,
+      "grad_norm": 0.2755306914200271,
+      "learning_rate": 9.737466891996035e-07,
+      "loss": 0.2761,
+      "step": 706
+    },
+    {
+      "epoch": 0.3563508064516129,
+      "grad_norm": 0.23830584140125033,
+      "learning_rate": 9.736607425969878e-07,
+      "loss": 0.2817,
+      "step": 707
+    },
+    {
+      "epoch": 0.35685483870967744,
+      "grad_norm": 0.5272610922884131,
+      "learning_rate": 9.73574659346182e-07,
+      "loss": 0.3099,
+      "step": 708
+    },
+    {
+      "epoch": 0.35735887096774194,
+      "grad_norm": 0.4888851365749825,
+      "learning_rate": 9.734884394720207e-07,
+      "loss": 0.3035,
+      "step": 709
+    },
+    {
+      "epoch": 0.35786290322580644,
+      "grad_norm": 0.19529005617043382,
+      "learning_rate": 9.734020829993778e-07,
+      "loss": 0.288,
+      "step": 710
+    },
+    {
+      "epoch": 0.35836693548387094,
+      "grad_norm": 0.3168784896287719,
+      "learning_rate": 9.733155899531661e-07,
+      "loss": 0.2747,
+      "step": 711
+    },
+    {
+      "epoch": 0.3588709677419355,
+      "grad_norm": 0.20038440381297584,
+      "learning_rate": 9.73228960358339e-07,
+      "loss": 0.2855,
+      "step": 712
+    },
+    {
+      "epoch": 0.359375,
+      "grad_norm": 0.2637581920266508,
+      "learning_rate": 9.731421942398882e-07,
+      "loss": 0.3007,
+      "step": 713
+    },
+    {
+      "epoch": 0.3598790322580645,
+      "grad_norm": 0.677620290574485,
+      "learning_rate": 9.730552916228455e-07,
+      "loss": 0.31,
+      "step": 714
+    },
+    {
+      "epoch": 0.36038306451612906,
+      "grad_norm": 0.22952816515508798,
+      "learning_rate": 9.729682525322817e-07,
+      "loss": 0.3066,
+      "step": 715
+    },
+    {
+      "epoch": 0.36088709677419356,
+      "grad_norm": 0.1917187197394172,
+      "learning_rate": 9.728810769933066e-07,
+      "loss": 0.3017,
+      "step": 716
+    },
+    {
+      "epoch": 0.36139112903225806,
+      "grad_norm": 0.20951491712451265,
+      "learning_rate": 9.727937650310704e-07,
+      "loss": 0.2973,
+      "step": 717
+    },
+    {
+      "epoch": 0.36189516129032256,
+      "grad_norm": 0.3141631511750949,
+      "learning_rate": 9.727063166707619e-07,
+      "loss": 0.308,
+      "step": 718
+    },
+    {
+      "epoch": 0.3623991935483871,
+      "grad_norm": 0.18721524079592824,
+      "learning_rate": 9.726187319376088e-07,
+      "loss": 0.2927,
+      "step": 719
+    },
+    {
+      "epoch": 0.3629032258064516,
+      "grad_norm": 0.28391314645074966,
+      "learning_rate": 9.725310108568795e-07,
+      "loss": 0.2949,
+      "step": 720
+    },
+    {
+      "epoch": 0.3634072580645161,
+      "grad_norm": 0.17932969621501238,
+      "learning_rate": 9.724431534538809e-07,
+      "loss": 0.294,
+      "step": 721
+    },
+    {
+      "epoch": 0.3639112903225806,
+      "grad_norm": 0.21062063293701141,
+      "learning_rate": 9.723551597539591e-07,
+      "loss": 0.2836,
+      "step": 722
+    },
+    {
+      "epoch": 0.3644153225806452,
+      "grad_norm": 0.18043765044502827,
+      "learning_rate": 9.722670297824998e-07,
+      "loss": 0.3093,
+      "step": 723
+    },
+    {
+      "epoch": 0.3649193548387097,
+      "grad_norm": 0.29050763149878694,
+      "learning_rate": 9.72178763564928e-07,
+      "loss": 0.2907,
+      "step": 724
+    },
+    {
+      "epoch": 0.3654233870967742,
+      "grad_norm": 0.18845371146424408,
+      "learning_rate": 9.720903611267077e-07,
+      "loss": 0.2986,
+      "step": 725
+    },
+    {
+      "epoch": 0.3659274193548387,
+      "grad_norm": 0.18828740365731594,
+      "learning_rate": 9.720018224933427e-07,
+      "loss": 0.2872,
+      "step": 726
+    },
+    {
+      "epoch": 0.36643145161290325,
+      "grad_norm": 0.19190513878279883,
+      "learning_rate": 9.71913147690376e-07,
+      "loss": 0.3062,
+      "step": 727
+    },
+    {
+      "epoch": 0.36693548387096775,
+      "grad_norm": 0.19814485234368132,
+      "learning_rate": 9.718243367433893e-07,
+      "loss": 0.2929,
+      "step": 728
+    },
+    {
+      "epoch": 0.36743951612903225,
+      "grad_norm": 0.43417746968071463,
+      "learning_rate": 9.717353896780042e-07,
+      "loss": 0.2861,
+      "step": 729
+    },
+    {
+      "epoch": 0.36794354838709675,
+      "grad_norm": 0.2278557334184222,
+      "learning_rate": 9.716463065198817e-07,
+      "loss": 0.3085,
+      "step": 730
+    },
+    {
+      "epoch": 0.3684475806451613,
+      "grad_norm": 0.17793722781989707,
+      "learning_rate": 9.715570872947213e-07,
+      "loss": 0.2919,
+      "step": 731
+    },
+    {
+      "epoch": 0.3689516129032258,
+      "grad_norm": 0.20613998379871223,
+      "learning_rate": 9.714677320282623e-07,
+      "loss": 0.3027,
+      "step": 732
+    },
+    {
+      "epoch": 0.3694556451612903,
+      "grad_norm": 0.1966436661511563,
+      "learning_rate": 9.713782407462834e-07,
+      "loss": 0.2741,
+      "step": 733
+    },
+    {
+      "epoch": 0.3699596774193548,
+      "grad_norm": 0.28922865753235927,
+      "learning_rate": 9.712886134746019e-07,
+      "loss": 0.2858,
+      "step": 734
+    },
+    {
+      "epoch": 0.3704637096774194,
+      "grad_norm": 0.2702305708992764,
+      "learning_rate": 9.71198850239075e-07,
+      "loss": 0.297,
+      "step": 735
+    },
+    {
+      "epoch": 0.3709677419354839,
+      "grad_norm": 0.194378574505089,
+      "learning_rate": 9.711089510655985e-07,
+      "loss": 0.3142,
+      "step": 736
+    },
+    {
+      "epoch": 0.3714717741935484,
+      "grad_norm": 0.18465691225415032,
+      "learning_rate": 9.710189159801084e-07,
+      "loss": 0.2972,
+      "step": 737
+    },
+    {
+      "epoch": 0.3719758064516129,
+      "grad_norm": 0.19730530115872685,
+      "learning_rate": 9.709287450085786e-07,
+      "loss": 0.2954,
+      "step": 738
+    },
+    {
+      "epoch": 0.37247983870967744,
+      "grad_norm": 0.2189962450750751,
+      "learning_rate": 9.70838438177023e-07,
+      "loss": 0.2922,
+      "step": 739
+    },
+    {
+      "epoch": 0.37298387096774194,
+      "grad_norm": 0.4433211186238402,
+      "learning_rate": 9.707479955114948e-07,
+      "loss": 0.3027,
+      "step": 740
+    },
+    {
+      "epoch": 0.37348790322580644,
+      "grad_norm": 0.24228833774809522,
+      "learning_rate": 9.70657417038086e-07,
+      "loss": 0.2983,
+      "step": 741
+    },
+    {
+      "epoch": 0.37399193548387094,
+      "grad_norm": 0.18660389324918686,
+      "learning_rate": 9.70566702782928e-07,
+      "loss": 0.3081,
+      "step": 742
+    },
+    {
+      "epoch": 0.3744959677419355,
+      "grad_norm": 0.29156052417574047,
+      "learning_rate": 9.704758527721912e-07,
+      "loss": 0.2827,
+      "step": 743
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.30556626157560013,
+      "learning_rate": 9.703848670320855e-07,
+      "loss": 0.3043,
+      "step": 744
+    },
+    {
+      "epoch": 0.3755040322580645,
+      "grad_norm": 0.28242712763277444,
+      "learning_rate": 9.702937455888593e-07,
+      "loss": 0.2971,
+      "step": 745
+    },
+    {
+      "epoch": 0.37600806451612906,
+      "grad_norm": 0.18221657954549308,
+      "learning_rate": 9.70202488468801e-07,
+      "loss": 0.2947,
+      "step": 746
+    },
+    {
+      "epoch": 0.37651209677419356,
+      "grad_norm": 0.17989299690744823,
+      "learning_rate": 9.701110956982374e-07,
+      "loss": 0.2924,
+      "step": 747
+    },
+    {
+      "epoch": 0.37701612903225806,
+      "grad_norm": 0.18220343446730708,
+      "learning_rate": 9.700195673035349e-07,
+      "loss": 0.2885,
+      "step": 748
+    },
+    {
+      "epoch": 0.37752016129032256,
+      "grad_norm": 0.22346642161187105,
+      "learning_rate": 9.699279033110988e-07,
+      "loss": 0.3244,
+      "step": 749
+    },
+    {
+      "epoch": 0.3780241935483871,
+      "grad_norm": 0.19169547017843008,
+      "learning_rate": 9.698361037473738e-07,
+      "loss": 0.2961,
+      "step": 750
+    },
+    {
+      "epoch": 0.3785282258064516,
+      "grad_norm": 0.22065738901487775,
+      "learning_rate": 9.697441686388432e-07,
+      "loss": 0.3068,
+      "step": 751
+    },
+    {
+      "epoch": 0.3790322580645161,
+      "grad_norm": 0.19267369377139024,
+      "learning_rate": 9.696520980120299e-07,
+      "loss": 0.2898,
+      "step": 752
+    },
+    {
+      "epoch": 0.3795362903225806,
+      "grad_norm": 0.19447034655516351,
+      "learning_rate": 9.695598918934958e-07,
+      "loss": 0.2938,
+      "step": 753
+    },
+    {
+      "epoch": 0.3800403225806452,
+      "grad_norm": 0.17575932802696953,
+      "learning_rate": 9.694675503098415e-07,
+      "loss": 0.299,
+      "step": 754
+    },
+    {
+      "epoch": 0.3805443548387097,
+      "grad_norm": 0.27717817200263767,
+      "learning_rate": 9.693750732877071e-07,
+      "loss": 0.2902,
+      "step": 755
+    },
+    {
+      "epoch": 0.3810483870967742,
+      "grad_norm": 0.18017521043674514,
+      "learning_rate": 9.692824608537718e-07,
+      "loss": 0.2934,
+      "step": 756
+    },
+    {
+      "epoch": 0.3815524193548387,
+      "grad_norm": 0.18161011916708894,
+      "learning_rate": 9.691897130347536e-07,
+      "loss": 0.2958,
+      "step": 757
+    },
+    {
+      "epoch": 0.38205645161290325,
+      "grad_norm": 0.20730764541104563,
+      "learning_rate": 9.690968298574095e-07,
+      "loss": 0.2908,
+      "step": 758
+    },
+    {
+      "epoch": 0.38256048387096775,
+      "grad_norm": 0.2200352742107318,
+      "learning_rate": 9.69003811348536e-07,
+      "loss": 0.3019,
+      "step": 759
+    },
+    {
+      "epoch": 0.38306451612903225,
+      "grad_norm": 0.19253930185176787,
+      "learning_rate": 9.689106575349682e-07,
+      "loss": 0.2864,
+      "step": 760
+    },
+    {
+      "epoch": 0.38356854838709675,
+      "grad_norm": 0.21888290789299836,
+      "learning_rate": 9.688173684435806e-07,
+      "loss": 0.313,
+      "step": 761
+    },
+    {
+      "epoch": 0.3840725806451613,
+      "grad_norm": 0.2275047774015037,
+      "learning_rate": 9.687239441012863e-07,
+      "loss": 0.2927,
+      "step": 762
+    },
+    {
+      "epoch": 0.3845766129032258,
+      "grad_norm": 0.17433382207800896,
+      "learning_rate": 9.686303845350377e-07,
+      "loss": 0.2934,
+      "step": 763
+    },
+    {
+      "epoch": 0.3850806451612903,
+      "grad_norm": 0.1910468291040477,
+      "learning_rate": 9.68536689771826e-07,
+      "loss": 0.2892,
+      "step": 764
+    },
+    {
+      "epoch": 0.3855846774193548,
+      "grad_norm": 0.2073819671598411,
+      "learning_rate": 9.68442859838682e-07,
+      "loss": 0.2874,
+      "step": 765
+    },
+    {
+      "epoch": 0.3860887096774194,
+      "grad_norm": 0.18337565075052092,
+      "learning_rate": 9.683488947626746e-07,
+      "loss": 0.2986,
+      "step": 766
+    },
+    {
+      "epoch": 0.3865927419354839,
+      "grad_norm": 0.19262381834997364,
+      "learning_rate": 9.682547945709125e-07,
+      "loss": 0.2954,
+      "step": 767
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 0.19206670831950323,
+      "learning_rate": 9.681605592905425e-07,
+      "loss": 0.2796,
+      "step": 768
+    },
+    {
+      "epoch": 0.3876008064516129,
+      "grad_norm": 0.18021565531041417,
+      "learning_rate": 9.680661889487517e-07,
+      "loss": 0.2974,
+      "step": 769
+    },
+    {
+      "epoch": 0.38810483870967744,
+      "grad_norm": 0.20648881640932878,
+      "learning_rate": 9.679716835727647e-07,
+      "loss": 0.304,
+      "step": 770
+    },
+    {
+      "epoch": 0.38860887096774194,
+      "grad_norm": 0.21633577449914515,
+      "learning_rate": 9.67877043189846e-07,
+      "loss": 0.2815,
+      "step": 771
+    },
+    {
+      "epoch": 0.38911290322580644,
+      "grad_norm": 0.19027954862930285,
+      "learning_rate": 9.677822678272986e-07,
+      "loss": 0.2775,
+      "step": 772
+    },
+    {
+      "epoch": 0.38961693548387094,
+      "grad_norm": 0.1853233674997291,
+      "learning_rate": 9.676873575124647e-07,
+      "loss": 0.2736,
+      "step": 773
+    },
+    {
+      "epoch": 0.3901209677419355,
+      "grad_norm": 0.28029424195764874,
+      "learning_rate": 9.675923122727253e-07,
+      "loss": 0.2919,
+      "step": 774
+    },
+    {
+      "epoch": 0.390625,
+      "grad_norm": 0.2271137389969601,
+      "learning_rate": 9.674971321355003e-07,
+      "loss": 0.2884,
+      "step": 775
+    },
+    {
+      "epoch": 0.3911290322580645,
+      "grad_norm": 0.2482514592600029,
+      "learning_rate": 9.67401817128249e-07,
+      "loss": 0.2947,
+      "step": 776
+    },
+    {
+      "epoch": 0.39163306451612906,
+      "grad_norm": 0.29951392676665967,
+      "learning_rate": 9.673063672784684e-07,
+      "loss": 0.289,
+      "step": 777
+    },
+    {
+      "epoch": 0.39213709677419356,
+      "grad_norm": 0.17456103242486867,
+      "learning_rate": 9.67210782613696e-07,
+      "loss": 0.2995,
+      "step": 778
+    },
+    {
+      "epoch": 0.39264112903225806,
+      "grad_norm": 0.6967952430229244,
+      "learning_rate": 9.67115063161507e-07,
+      "loss": 0.3022,
+      "step": 779
+    },
+    {
+      "epoch": 0.39314516129032256,
+      "grad_norm": 0.29680448703729057,
+      "learning_rate": 9.67019208949516e-07,
+      "loss": 0.2862,
+      "step": 780
+    },
+    {
+      "epoch": 0.3936491935483871,
+      "grad_norm": 0.177999847832676,
+      "learning_rate": 9.669232200053759e-07,
+      "loss": 0.2962,
+      "step": 781
+    },
+    {
+      "epoch": 0.3941532258064516,
+      "grad_norm": 0.27026664375469966,
+      "learning_rate": 9.668270963567794e-07,
+      "loss": 0.2986,
+      "step": 782
+    },
+    {
+      "epoch": 0.3946572580645161,
+      "grad_norm": 0.19723594083135135,
+      "learning_rate": 9.667308380314576e-07,
+      "loss": 0.3063,
+      "step": 783
+    },
+    {
+      "epoch": 0.3951612903225806,
+      "grad_norm": 0.21862308310307982,
+      "learning_rate": 9.666344450571801e-07,
+      "loss": 0.2964,
+      "step": 784
+    },
+    {
+      "epoch": 0.3956653225806452,
+      "grad_norm": 0.1758663671801379,
+      "learning_rate": 9.665379174617558e-07,
+      "loss": 0.2914,
+      "step": 785
+    },
+    {
+      "epoch": 0.3961693548387097,
+      "grad_norm": 0.32096497636941274,
+      "learning_rate": 9.664412552730326e-07,
+      "loss": 0.2897,
+      "step": 786
+    },
+    {
+      "epoch": 0.3966733870967742,
+      "grad_norm": 0.17374107415020215,
+      "learning_rate": 9.663444585188965e-07,
+      "loss": 0.2867,
+      "step": 787
+    },
+    {
+      "epoch": 0.3971774193548387,
+      "grad_norm": 0.18444275375398292,
+      "learning_rate": 9.66247527227273e-07,
+      "loss": 0.2961,
+      "step": 788
+    },
+    {
+      "epoch": 0.39768145161290325,
+      "grad_norm": 0.19643289496174723,
+      "learning_rate": 9.661504614261261e-07,
+      "loss": 0.3125,
+      "step": 789
+    },
+    {
+      "epoch": 0.39818548387096775,
+      "grad_norm": 0.2167645010446501,
+      "learning_rate": 9.660532611434591e-07,
+      "loss": 0.2921,
+      "step": 790
+    },
+    {
+      "epoch": 0.39868951612903225,
+      "grad_norm": 0.1806328023049352,
+      "learning_rate": 9.659559264073129e-07,
+      "loss": 0.2846,
+      "step": 791
+    },
+    {
+      "epoch": 0.39919354838709675,
+      "grad_norm": 0.2612388559577726,
+      "learning_rate": 9.658584572457686e-07,
+      "loss": 0.2945,
+      "step": 792
+    },
+    {
+      "epoch": 0.3996975806451613,
+      "grad_norm": 0.2793761291173004,
+      "learning_rate": 9.657608536869451e-07,
+      "loss": 0.2837,
+      "step": 793
+    },
+    {
+      "epoch": 0.4002016129032258,
+      "grad_norm": 0.18310300206951588,
+      "learning_rate": 9.656631157590004e-07,
+      "loss": 0.2978,
+      "step": 794
+    },
+    {
+      "epoch": 0.4007056451612903,
+      "grad_norm": 0.18408805840331044,
+      "learning_rate": 9.655652434901317e-07,
+      "loss": 0.3079,
+      "step": 795
+    },
+    {
+      "epoch": 0.4012096774193548,
+      "grad_norm": 0.2342952244459477,
+      "learning_rate": 9.654672369085742e-07,
+      "loss": 0.2834,
+      "step": 796
+    },
+    {
+      "epoch": 0.4012096774193548,
+      "eval_loss": 0.3250449001789093,
+      "eval_runtime": 18.239,
+      "eval_samples_per_second": 46.878,
+      "eval_steps_per_second": 0.987,
+      "step": 796
+    },
+    {
+      "epoch": 0.4017137096774194,
+      "grad_norm": 0.6951664973155232,
+      "learning_rate": 9.653690960426024e-07,
+      "loss": 0.2949,
+      "step": 797
+    },
+    {
+      "epoch": 0.4022177419354839,
+      "grad_norm": 0.3007352567348458,
+      "learning_rate": 9.652708209205289e-07,
+      "loss": 0.308,
+      "step": 798
+    },
+    {
+      "epoch": 0.4027217741935484,
+      "grad_norm": 0.2111912313466037,
+      "learning_rate": 9.651724115707059e-07,
+      "loss": 0.2868,
+      "step": 799
+    },
+    {
+      "epoch": 0.4032258064516129,
+      "grad_norm": 0.25518959182252676,
+      "learning_rate": 9.650738680215237e-07,
+      "loss": 0.2926,
+      "step": 800
+    },
+    {
+      "epoch": 0.40372983870967744,
+      "grad_norm": 0.18835961947528512,
+      "learning_rate": 9.649751903014117e-07,
+      "loss": 0.2894,
+      "step": 801
+    },
+    {
+      "epoch": 0.40423387096774194,
+      "grad_norm": 0.17984205410768345,
+      "learning_rate": 9.648763784388375e-07,
+      "loss": 0.2857,
+      "step": 802
+    },
+    {
+      "epoch": 0.40473790322580644,
+      "grad_norm": 0.19376522082181055,
+      "learning_rate": 9.647774324623082e-07,
+      "loss": 0.2956,
+      "step": 803
+    },
+    {
+      "epoch": 0.40524193548387094,
+      "grad_norm": 0.2473519213828851,
+      "learning_rate": 9.646783524003684e-07,
+      "loss": 0.2816,
+      "step": 804
+    },
+    {
+      "epoch": 0.4057459677419355,
+      "grad_norm": 0.4135144683121957,
+      "learning_rate": 9.645791382816026e-07,
+      "loss": 0.2794,
+      "step": 805
+    },
+    {
+      "epoch": 0.40625,
+      "grad_norm": 0.22178842307704266,
+      "learning_rate": 9.644797901346333e-07,
+      "loss": 0.2859,
+      "step": 806
+    },
+    {
+      "epoch": 0.4067540322580645,
+      "grad_norm": 0.1865025598672438,
+      "learning_rate": 9.64380307988122e-07,
+      "loss": 0.2826,
+      "step": 807
+    },
+    {
+      "epoch": 0.40725806451612906,
+      "grad_norm": 0.21780748939537628,
+      "learning_rate": 9.642806918707685e-07,
+      "loss": 0.3003,
+      "step": 808
+    },
+    {
+      "epoch": 0.40776209677419356,
+      "grad_norm": 0.23524333096509437,
+      "learning_rate": 9.641809418113113e-07,
+      "loss": 0.2731,
+      "step": 809
+    },
+    {
+      "epoch": 0.40826612903225806,
+      "grad_norm": 0.3522546815652214,
+      "learning_rate": 9.64081057838528e-07,
+      "loss": 0.3006,
+      "step": 810
+    },
+    {
+      "epoch": 0.40877016129032256,
+      "grad_norm": 0.2060231832092746,
+      "learning_rate": 9.63981039981234e-07,
+      "loss": 0.2889,
+      "step": 811
+    },
+    {
+      "epoch": 0.4092741935483871,
+      "grad_norm": 0.17805134176779017,
+      "learning_rate": 9.638808882682845e-07,
+      "loss": 0.28,
+      "step": 812
+    },
+    {
+      "epoch": 0.4097782258064516,
+      "grad_norm": 0.21338586685342792,
+      "learning_rate": 9.637806027285721e-07,
+      "loss": 0.3266,
+      "step": 813
+    },
+    {
+      "epoch": 0.4102822580645161,
+      "grad_norm": 0.23761874762782823,
+      "learning_rate": 9.636801833910291e-07,
+      "loss": 0.2928,
+      "step": 814
+    },
+    {
+      "epoch": 0.4107862903225806,
+      "grad_norm": 0.349479056634183,
+      "learning_rate": 9.635796302846253e-07,
+      "loss": 0.2893,
+      "step": 815
+    },
+    {
+      "epoch": 0.4112903225806452,
+      "grad_norm": 0.1793918088824337,
+      "learning_rate": 9.6347894343837e-07,
+      "loss": 0.3022,
+      "step": 816
+    },
+    {
+      "epoch": 0.4117943548387097,
+      "grad_norm": 0.20651527805989847,
+      "learning_rate": 9.633781228813107e-07,
+      "loss": 0.289,
+      "step": 817
+    },
+    {
+      "epoch": 0.4122983870967742,
+      "grad_norm": 0.2120940285959409,
+      "learning_rate": 9.63277168642533e-07,
+      "loss": 0.2796,
+      "step": 818
+    },
+    {
+      "epoch": 0.4128024193548387,
+      "grad_norm": 0.18644947273341353,
+      "learning_rate": 9.631760807511624e-07,
+      "loss": 0.2823,
+      "step": 819
+    },
+    {
+      "epoch": 0.41330645161290325,
+      "grad_norm": 0.2559146778083527,
+      "learning_rate": 9.630748592363617e-07,
+      "loss": 0.3102,
+      "step": 820
+    },
+    {
+      "epoch": 0.41381048387096775,
+      "grad_norm": 0.20144278166717078,
+      "learning_rate": 9.629735041273325e-07,
+      "loss": 0.2949,
+      "step": 821
+    },
+    {
+      "epoch": 0.41431451612903225,
+      "grad_norm": 0.22358896888220894,
+      "learning_rate": 9.628720154533157e-07,
+      "loss": 0.2809,
+      "step": 822
+    },
+    {
+      "epoch": 0.41481854838709675,
+      "grad_norm": 0.2056875307546662,
+      "learning_rate": 9.627703932435895e-07,
+      "loss": 0.2838,
+      "step": 823
+    },
+    {
+      "epoch": 0.4153225806451613,
+      "grad_norm": 0.199611316920199,
+      "learning_rate": 9.626686375274715e-07,
+      "loss": 0.2731,
+      "step": 824
+    },
+    {
+      "epoch": 0.4158266129032258,
+      "grad_norm": 0.17872361611126808,
+      "learning_rate": 9.625667483343177e-07,
+      "loss": 0.2885,
+      "step": 825
+    },
+    {
+      "epoch": 0.4163306451612903,
+      "grad_norm": 0.1715080401060158,
+      "learning_rate": 9.624647256935226e-07,
+      "loss": 0.2918,
+      "step": 826
+    },
+    {
+      "epoch": 0.4168346774193548,
+      "grad_norm": 0.28942913404328185,
+      "learning_rate": 9.623625696345187e-07,
+      "loss": 0.2909,
+      "step": 827
+    },
+    {
+      "epoch": 0.4173387096774194,
+      "grad_norm": 0.5219653514726549,
+      "learning_rate": 9.62260280186778e-07,
+      "loss": 0.2917,
+      "step": 828
+    },
+    {
+      "epoch": 0.4178427419354839,
+      "grad_norm": 0.1803172885571653,
+      "learning_rate": 9.621578573798098e-07,
+      "loss": 0.29,
+      "step": 829
+    },
+    {
+      "epoch": 0.4183467741935484,
+      "grad_norm": 0.26437521269828795,
+      "learning_rate": 9.620553012431626e-07,
+      "loss": 0.2962,
+      "step": 830
+    },
+    {
+      "epoch": 0.4188508064516129,
+      "grad_norm": 0.22357063643965017,
+      "learning_rate": 9.619526118064234e-07,
+      "loss": 0.2841,
+      "step": 831
+    },
+    {
+      "epoch": 0.41935483870967744,
+      "grad_norm": 0.27227090245290375,
+      "learning_rate": 9.618497890992171e-07,
+      "loss": 0.2791,
+      "step": 832
+    },
+    {
+      "epoch": 0.41985887096774194,
+      "grad_norm": 0.18400884656891223,
+      "learning_rate": 9.61746833151208e-07,
+      "loss": 0.2922,
+      "step": 833
+    },
+    {
+      "epoch": 0.42036290322580644,
+      "grad_norm": 0.17737272960072972,
+      "learning_rate": 9.616437439920977e-07,
+      "loss": 0.2941,
+      "step": 834
+    },
+    {
+      "epoch": 0.42086693548387094,
+      "grad_norm": 0.1845994861654205,
+      "learning_rate": 9.61540521651627e-07,
+      "loss": 0.2993,
+      "step": 835
+    },
+    {
+      "epoch": 0.4213709677419355,
+      "grad_norm": 0.184243652976245,
+      "learning_rate": 9.61437166159575e-07,
+      "loss": 0.2954,
+      "step": 836
+    },
+    {
+      "epoch": 0.421875,
+      "grad_norm": 0.1962394295882692,
+      "learning_rate": 9.61333677545759e-07,
+      "loss": 0.2797,
+      "step": 837
+    },
+    {
+      "epoch": 0.4223790322580645,
+      "grad_norm": 0.19907302059194398,
+      "learning_rate": 9.612300558400348e-07,
+      "loss": 0.2913,
+      "step": 838
+    },
+    {
+      "epoch": 0.42288306451612906,
+      "grad_norm": 0.17791364001664575,
+      "learning_rate": 9.611263010722968e-07,
+      "loss": 0.308,
+      "step": 839
+    },
+    {
+      "epoch": 0.42338709677419356,
+      "grad_norm": 0.20244623183338803,
+      "learning_rate": 9.610224132724772e-07,
+      "loss": 0.2984,
+      "step": 840
+    },
+    {
+      "epoch": 0.42389112903225806,
+      "grad_norm": 0.1821553749592407,
+      "learning_rate": 9.609183924705473e-07,
+      "loss": 0.276,
+      "step": 841
+    },
+    {
+      "epoch": 0.42439516129032256,
+      "grad_norm": 0.18302758905749872,
+      "learning_rate": 9.608142386965166e-07,
+      "loss": 0.2886,
+      "step": 842
+    },
+    {
+      "epoch": 0.4248991935483871,
+      "grad_norm": 0.2814755691064198,
+      "learning_rate": 9.607099519804325e-07,
+      "loss": 0.2996,
+      "step": 843
+    },
+    {
+      "epoch": 0.4254032258064516,
+      "grad_norm": 0.2188163570039596,
+      "learning_rate": 9.60605532352381e-07,
+      "loss": 0.2849,
+      "step": 844
+    },
+    {
+      "epoch": 0.4259072580645161,
+      "grad_norm": 0.19063373332278225,
+      "learning_rate": 9.605009798424871e-07,
+      "loss": 0.2826,
+      "step": 845
+    },
+    {
+      "epoch": 0.4264112903225806,
+      "grad_norm": 0.22803495479430694,
+      "learning_rate": 9.60396294480913e-07,
+      "loss": 0.2881,
+      "step": 846
+    },
+    {
+      "epoch": 0.4269153225806452,
+      "grad_norm": 0.2532580997944776,
+      "learning_rate": 9.6029147629786e-07,
+      "loss": 0.2931,
+      "step": 847
+    },
+    {
+      "epoch": 0.4274193548387097,
+      "grad_norm": 0.29895228840340143,
+      "learning_rate": 9.601865253235673e-07,
+      "loss": 0.2851,
+      "step": 848
+    },
+    {
+      "epoch": 0.4279233870967742,
+      "grad_norm": 0.19966385962690142,
+      "learning_rate": 9.60081441588313e-07,
+      "loss": 0.2977,
+      "step": 849
+    },
+    {
+      "epoch": 0.4284274193548387,
+      "grad_norm": 0.3898444466529591,
+      "learning_rate": 9.599762251224125e-07,
+      "loss": 0.3104,
+      "step": 850
+    },
+    {
+      "epoch": 0.42893145161290325,
+      "grad_norm": 0.24259677387671844,
+      "learning_rate": 9.598708759562208e-07,
+      "loss": 0.2987,
+      "step": 851
+    },
+    {
+      "epoch": 0.42943548387096775,
+      "grad_norm": 0.20006735807640816,
+      "learning_rate": 9.5976539412013e-07,
+      "loss": 0.2864,
+      "step": 852
+    },
+    {
+      "epoch": 0.42993951612903225,
+      "grad_norm": 0.2882245398429841,
+      "learning_rate": 9.59659779644571e-07,
+      "loss": 0.2858,
+      "step": 853
+    },
+    {
+      "epoch": 0.43044354838709675,
+      "grad_norm": 0.22513281216087977,
+      "learning_rate": 9.59554032560013e-07,
+      "loss": 0.2824,
+      "step": 854
+    },
+    {
+      "epoch": 0.4309475806451613,
+      "grad_norm": 0.26571888341504696,
+      "learning_rate": 9.594481528969635e-07,
+      "loss": 0.3086,
+      "step": 855
+    },
+    {
+      "epoch": 0.4314516129032258,
+      "grad_norm": 0.19249797160463983,
+      "learning_rate": 9.59342140685968e-07,
+      "loss": 0.2889,
+      "step": 856
+    },
+    {
+      "epoch": 0.4319556451612903,
+      "grad_norm": 0.17883290237014954,
+      "learning_rate": 9.592359959576104e-07,
+      "loss": 0.2913,
+      "step": 857
+    },
+    {
+      "epoch": 0.4324596774193548,
+      "grad_norm": 0.38963242443484597,
+      "learning_rate": 9.591297187425128e-07,
+      "loss": 0.2905,
+      "step": 858
+    },
+    {
+      "epoch": 0.4329637096774194,
+      "grad_norm": 0.4499390873499669,
+      "learning_rate": 9.590233090713354e-07,
+      "loss": 0.3016,
+      "step": 859
+    },
+    {
+      "epoch": 0.4334677419354839,
+      "grad_norm": 0.23147026645054392,
+      "learning_rate": 9.58916766974777e-07,
+      "loss": 0.2753,
+      "step": 860
+    },
+    {
+      "epoch": 0.4339717741935484,
+      "grad_norm": 0.23004776559869194,
+      "learning_rate": 9.58810092483574e-07,
+      "loss": 0.2857,
+      "step": 861
+    },
+    {
+      "epoch": 0.4344758064516129,
+      "grad_norm": 0.38524648235237546,
+      "learning_rate": 9.587032856285016e-07,
+      "loss": 0.295,
+      "step": 862
+    },
+    {
+      "epoch": 0.43497983870967744,
+      "grad_norm": 0.26978296400638296,
+      "learning_rate": 9.585963464403727e-07,
+      "loss": 0.2924,
+      "step": 863
+    },
+    {
+      "epoch": 0.43548387096774194,
+      "grad_norm": 0.2304216233656417,
+      "learning_rate": 9.584892749500388e-07,
+      "loss": 0.2782,
+      "step": 864
+    },
+    {
+      "epoch": 0.43598790322580644,
+      "grad_norm": 0.20628041373927403,
+      "learning_rate": 9.58382071188389e-07,
+      "loss": 0.2938,
+      "step": 865
+    },
+    {
+      "epoch": 0.43649193548387094,
+      "grad_norm": 0.24806721102476476,
+      "learning_rate": 9.582747351863518e-07,
+      "loss": 0.3051,
+      "step": 866
+    },
+    {
+      "epoch": 0.4369959677419355,
+      "grad_norm": 0.18289677229578133,
+      "learning_rate": 9.58167266974892e-07,
+      "loss": 0.2928,
+      "step": 867
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 0.20539381909701737,
+      "learning_rate": 9.580596665850139e-07,
+      "loss": 0.2873,
+      "step": 868
+    },
+    {
+      "epoch": 0.4380040322580645,
+      "grad_norm": 0.18821302140885868,
+      "learning_rate": 9.579519340477592e-07,
+      "loss": 0.2903,
+      "step": 869
+    },
+    {
+      "epoch": 0.43850806451612906,
+      "grad_norm": 0.37407049401588166,
+      "learning_rate": 9.57844069394209e-07,
+      "loss": 0.2931,
+      "step": 870
+    },
+    {
+      "epoch": 0.43901209677419356,
+      "grad_norm": 0.2339579704921313,
+      "learning_rate": 9.577360726554804e-07,
+      "loss": 0.2847,
+      "step": 871
+    },
+    {
+      "epoch": 0.43951612903225806,
+      "grad_norm": 0.23524427674559492,
+      "learning_rate": 9.576279438627308e-07,
+      "loss": 0.2754,
+      "step": 872
+    },
+    {
+      "epoch": 0.44002016129032256,
+      "grad_norm": 0.18413711797969656,
+      "learning_rate": 9.57519683047154e-07,
+      "loss": 0.2893,
+      "step": 873
+    },
+    {
+      "epoch": 0.4405241935483871,
+      "grad_norm": 0.22031589128909412,
+      "learning_rate": 9.574112902399829e-07,
+      "loss": 0.2884,
+      "step": 874
+    },
+    {
+      "epoch": 0.4410282258064516,
+      "grad_norm": 0.22755059592326796,
+      "learning_rate": 9.573027654724882e-07,
+      "loss": 0.29,
+      "step": 875
+    },
+    {
+      "epoch": 0.4415322580645161,
+      "grad_norm": 0.18373793370970248,
+      "learning_rate": 9.571941087759782e-07,
+      "loss": 0.3064,
+      "step": 876
+    },
+    {
+      "epoch": 0.4420362903225806,
+      "grad_norm": 0.19409387364473213,
+      "learning_rate": 9.570853201818002e-07,
+      "loss": 0.2922,
+      "step": 877
+    },
+    {
+      "epoch": 0.4425403225806452,
+      "grad_norm": 0.21129036924339822,
+      "learning_rate": 9.569763997213387e-07,
+      "loss": 0.2714,
+      "step": 878
+    },
+    {
+      "epoch": 0.4430443548387097,
+      "grad_norm": 0.17440531697354897,
+      "learning_rate": 9.568673474260168e-07,
+      "loss": 0.2847,
+      "step": 879
+    },
+    {
+      "epoch": 0.4435483870967742,
+      "grad_norm": 0.18668182286896606,
+      "learning_rate": 9.567581633272955e-07,
+      "loss": 0.3118,
+      "step": 880
+    },
+    {
+      "epoch": 0.4440524193548387,
+      "grad_norm": 0.1797795667764904,
+      "learning_rate": 9.56648847456673e-07,
+      "loss": 0.3019,
+      "step": 881
+    },
+    {
+      "epoch": 0.44455645161290325,
+      "grad_norm": 0.3707822100666679,
+      "learning_rate": 9.565393998456874e-07,
+      "loss": 0.2912,
+      "step": 882
+    },
+    {
+      "epoch": 0.44506048387096775,
+      "grad_norm": 0.289665390914343,
+      "learning_rate": 9.564298205259126e-07,
+      "loss": 0.2895,
+      "step": 883
+    },
+    {
+      "epoch": 0.44556451612903225,
+      "grad_norm": 0.2056754789971548,
+      "learning_rate": 9.563201095289624e-07,
+      "loss": 0.298,
+      "step": 884
+    },
+    {
+      "epoch": 0.44606854838709675,
+      "grad_norm": 0.1814615855914537,
+      "learning_rate": 9.562102668864871e-07,
+      "loss": 0.2934,
+      "step": 885
+    },
+    {
+      "epoch": 0.4465725806451613,
+      "grad_norm": 0.37487748973707236,
+      "learning_rate": 9.56100292630176e-07,
+      "loss": 0.2973,
+      "step": 886
+    },
+    {
+      "epoch": 0.4470766129032258,
+      "grad_norm": 0.1759000528091719,
+      "learning_rate": 9.559901867917556e-07,
+      "loss": 0.3039,
+      "step": 887
+    },
+    {
+      "epoch": 0.4475806451612903,
+      "grad_norm": 0.32191288919848915,
+      "learning_rate": 9.558799494029914e-07,
+      "loss": 0.2893,
+      "step": 888
+    },
+    {
+      "epoch": 0.4480846774193548,
+      "grad_norm": 0.19810987185465256,
+      "learning_rate": 9.557695804956856e-07,
+      "loss": 0.3127,
+      "step": 889
+    },
+    {
+      "epoch": 0.4485887096774194,
+      "grad_norm": 0.24720381123354968,
+      "learning_rate": 9.556590801016793e-07,
+      "loss": 0.2918,
+      "step": 890
+    },
+    {
+      "epoch": 0.4490927419354839,
+      "grad_norm": 0.18168928920802488,
+      "learning_rate": 9.555484482528508e-07,
+      "loss": 0.2881,
+      "step": 891
+    },
+    {
+      "epoch": 0.4495967741935484,
+      "grad_norm": 0.18687548855169936,
+      "learning_rate": 9.554376849811173e-07,
+      "loss": 0.2905,
+      "step": 892
+    },
+    {
+      "epoch": 0.4501008064516129,
+      "grad_norm": 0.20251482097557522,
+      "learning_rate": 9.553267903184327e-07,
+      "loss": 0.2939,
+      "step": 893
+    },
+    {
+      "epoch": 0.45060483870967744,
+      "grad_norm": 0.2887402542876595,
+      "learning_rate": 9.552157642967897e-07,
+      "loss": 0.2878,
+      "step": 894
+    },
+    {
+      "epoch": 0.45110887096774194,
+      "grad_norm": 0.19945034327270725,
+      "learning_rate": 9.551046069482186e-07,
+      "loss": 0.2845,
+      "step": 895
+    },
+    {
+      "epoch": 0.45161290322580644,
+      "grad_norm": 0.23431424692187636,
+      "learning_rate": 9.549933183047877e-07,
+      "loss": 0.2917,
+      "step": 896
+    },
+    {
+      "epoch": 0.45211693548387094,
+      "grad_norm": 0.2190333169890677,
+      "learning_rate": 9.54881898398603e-07,
+      "loss": 0.2952,
+      "step": 897
+    },
+    {
+      "epoch": 0.4526209677419355,
+      "grad_norm": 0.17822949705219454,
+      "learning_rate": 9.547703472618086e-07,
+      "loss": 0.299,
+      "step": 898
+    },
+    {
+      "epoch": 0.453125,
+      "grad_norm": 0.2102549610365594,
+      "learning_rate": 9.54658664926586e-07,
+      "loss": 0.2935,
+      "step": 899
+    },
+    {
+      "epoch": 0.4536290322580645,
+      "grad_norm": 0.19289696781722984,
+      "learning_rate": 9.545468514251552e-07,
+      "loss": 0.2998,
+      "step": 900
+    },
+    {
+      "epoch": 0.45413306451612906,
+      "grad_norm": 0.3042982461642612,
+      "learning_rate": 9.544349067897734e-07,
+      "loss": 0.2937,
+      "step": 901
+    },
+    {
+      "epoch": 0.45463709677419356,
+      "grad_norm": 0.19107058828557277,
+      "learning_rate": 9.54322831052736e-07,
+      "loss": 0.2765,
+      "step": 902
+    },
+    {
+      "epoch": 0.45514112903225806,
+      "grad_norm": 0.28747582105179686,
+      "learning_rate": 9.542106242463764e-07,
+      "loss": 0.2833,
+      "step": 903
+    },
+    {
+      "epoch": 0.45564516129032256,
+      "grad_norm": 0.18458129956622982,
+      "learning_rate": 9.540982864030653e-07,
+      "loss": 0.2831,
+      "step": 904
+    },
+    {
+      "epoch": 0.4561491935483871,
+      "grad_norm": 0.2130309759800133,
+      "learning_rate": 9.539858175552115e-07,
+      "loss": 0.2836,
+      "step": 905
+    },
+    {
+      "epoch": 0.4566532258064516,
+      "grad_norm": 0.3057883689836788,
+      "learning_rate": 9.538732177352617e-07,
+      "loss": 0.3007,
+      "step": 906
+    },
+    {
+      "epoch": 0.4571572580645161,
+      "grad_norm": 0.31124910837676767,
+      "learning_rate": 9.537604869757001e-07,
+      "loss": 0.2764,
+      "step": 907
+    },
+    {
+      "epoch": 0.4576612903225806,
+      "grad_norm": 0.2441004343752962,
+      "learning_rate": 9.53647625309049e-07,
+      "loss": 0.2868,
+      "step": 908
+    },
+    {
+      "epoch": 0.4581653225806452,
+      "grad_norm": 0.2357500626232433,
+      "learning_rate": 9.535346327678682e-07,
+      "loss": 0.3016,
+      "step": 909
+    },
+    {
+      "epoch": 0.4586693548387097,
+      "grad_norm": 0.20066160584984583,
+      "learning_rate": 9.534215093847552e-07,
+      "loss": 0.3001,
+      "step": 910
+    },
+    {
+      "epoch": 0.4591733870967742,
+      "grad_norm": 0.22773495544084632,
+      "learning_rate": 9.533082551923458e-07,
+      "loss": 0.2756,
+      "step": 911
+    },
+    {
+      "epoch": 0.4596774193548387,
+      "grad_norm": 0.1906199840165581,
+      "learning_rate": 9.531948702233126e-07,
+      "loss": 0.2869,
+      "step": 912
+    },
+    {
+      "epoch": 0.46018145161290325,
+      "grad_norm": 0.3245639131414002,
+      "learning_rate": 9.530813545103667e-07,
+      "loss": 0.2708,
+      "step": 913
+    },
+    {
+      "epoch": 0.46068548387096775,
+      "grad_norm": 0.21271580471854526,
+      "learning_rate": 9.52967708086257e-07,
+      "loss": 0.2892,
+      "step": 914
+    },
+    {
+      "epoch": 0.46118951612903225,
+      "grad_norm": 0.2104148343959223,
+      "learning_rate": 9.528539309837693e-07,
+      "loss": 0.2877,
+      "step": 915
+    },
+    {
+      "epoch": 0.46169354838709675,
+      "grad_norm": 0.19824185563673147,
+      "learning_rate": 9.527400232357279e-07,
+      "loss": 0.276,
+      "step": 916
+    },
+    {
+      "epoch": 0.4621975806451613,
+      "grad_norm": 0.21951663687873915,
+      "learning_rate": 9.526259848749943e-07,
+      "loss": 0.2859,
+      "step": 917
+    },
+    {
+      "epoch": 0.4627016129032258,
+      "grad_norm": 0.37545971476160844,
+      "learning_rate": 9.52511815934468e-07,
+      "loss": 0.2774,
+      "step": 918
+    },
+    {
+      "epoch": 0.4632056451612903,
+      "grad_norm": 0.22438052196032507,
+      "learning_rate": 9.523975164470859e-07,
+      "loss": 0.2978,
+      "step": 919
+    },
+    {
+      "epoch": 0.4637096774193548,
+      "grad_norm": 0.1757823732376565,
+      "learning_rate": 9.522830864458227e-07,
+      "loss": 0.3022,
+      "step": 920
+    },
+    {
+      "epoch": 0.4642137096774194,
+      "grad_norm": 0.27657899716696727,
+      "learning_rate": 9.521685259636909e-07,
+      "loss": 0.2944,
+      "step": 921
+    },
+    {
+      "epoch": 0.4647177419354839,
+      "grad_norm": 0.18756079463012706,
+      "learning_rate": 9.520538350337404e-07,
+      "loss": 0.2884,
+      "step": 922
+    },
+    {
+      "epoch": 0.4652217741935484,
+      "grad_norm": 0.2062606714488354,
+      "learning_rate": 9.519390136890589e-07,
+      "loss": 0.2778,
+      "step": 923
+    },
+    {
+      "epoch": 0.4657258064516129,
+      "grad_norm": 0.366011125828711,
+      "learning_rate": 9.518240619627713e-07,
+      "loss": 0.3001,
+      "step": 924
+    },
+    {
+      "epoch": 0.46622983870967744,
+      "grad_norm": 0.2541955304250703,
+      "learning_rate": 9.51708979888041e-07,
+      "loss": 0.2905,
+      "step": 925
+    },
+    {
+      "epoch": 0.46673387096774194,
+      "grad_norm": 0.36412234331970916,
+      "learning_rate": 9.51593767498068e-07,
+      "loss": 0.2822,
+      "step": 926
+    },
+    {
+      "epoch": 0.46723790322580644,
+      "grad_norm": 0.21139676616548536,
+      "learning_rate": 9.514784248260908e-07,
+      "loss": 0.3045,
+      "step": 927
+    },
+    {
+      "epoch": 0.46774193548387094,
+      "grad_norm": 0.19079104145467554,
+      "learning_rate": 9.513629519053845e-07,
+      "loss": 0.2737,
+      "step": 928
+    },
+    {
+      "epoch": 0.4682459677419355,
+      "grad_norm": 0.23607892317649437,
+      "learning_rate": 9.512473487692628e-07,
+      "loss": 0.2999,
+      "step": 929
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 0.1803840363558066,
+      "learning_rate": 9.511316154510763e-07,
+      "loss": 0.287,
+      "step": 930
+    },
+    {
+      "epoch": 0.4692540322580645,
+      "grad_norm": 0.17453949054667584,
+      "learning_rate": 9.510157519842133e-07,
+      "loss": 0.2851,
+      "step": 931
+    },
+    {
+      "epoch": 0.46975806451612906,
+      "grad_norm": 0.24449967092054903,
+      "learning_rate": 9.508997584020997e-07,
+      "loss": 0.2872,
+      "step": 932
+    },
+    {
+      "epoch": 0.47026209677419356,
+      "grad_norm": 0.1745644546122727,
+      "learning_rate": 9.507836347381992e-07,
+      "loss": 0.3005,
+      "step": 933
+    },
+    {
+      "epoch": 0.47076612903225806,
+      "grad_norm": 0.18053174508073816,
+      "learning_rate": 9.506673810260123e-07,
+      "loss": 0.2929,
+      "step": 934
+    },
+    {
+      "epoch": 0.47127016129032256,
+      "grad_norm": 0.31029290882211796,
+      "learning_rate": 9.505509972990778e-07,
+      "loss": 0.2926,
+      "step": 935
+    },
+    {
+      "epoch": 0.4717741935483871,
+      "grad_norm": 0.22938531150844882,
+      "learning_rate": 9.504344835909716e-07,
+      "loss": 0.2885,
+      "step": 936
+    },
+    {
+      "epoch": 0.4722782258064516,
+      "grad_norm": 0.17882862009471923,
+      "learning_rate": 9.503178399353072e-07,
+      "loss": 0.2743,
+      "step": 937
+    },
+    {
+      "epoch": 0.4727822580645161,
+      "grad_norm": 0.1854810156795896,
+      "learning_rate": 9.502010663657354e-07,
+      "loss": 0.2955,
+      "step": 938
+    },
+    {
+      "epoch": 0.4732862903225806,
+      "grad_norm": 0.21043848944429472,
+      "learning_rate": 9.50084162915945e-07,
+      "loss": 0.2964,
+      "step": 939
+    },
+    {
+      "epoch": 0.4737903225806452,
+      "grad_norm": 0.2847988534984117,
+      "learning_rate": 9.499671296196617e-07,
+      "loss": 0.285,
+      "step": 940
+    },
+    {
+      "epoch": 0.4742943548387097,
+      "grad_norm": 0.25216723077802655,
+      "learning_rate": 9.498499665106487e-07,
+      "loss": 0.2836,
+      "step": 941
+    },
+    {
+      "epoch": 0.4747983870967742,
+      "grad_norm": 0.18397402571778532,
+      "learning_rate": 9.497326736227071e-07,
+      "loss": 0.2937,
+      "step": 942
+    },
+    {
+      "epoch": 0.4753024193548387,
+      "grad_norm": 0.28708542295201006,
+      "learning_rate": 9.496152509896753e-07,
+      "loss": 0.2851,
+      "step": 943
+    },
+    {
+      "epoch": 0.47580645161290325,
+      "grad_norm": 0.2460818429957608,
+      "learning_rate": 9.494976986454286e-07,
+      "loss": 0.29,
+      "step": 944
+    },
+    {
+      "epoch": 0.47631048387096775,
+      "grad_norm": 0.19878718222512176,
+      "learning_rate": 9.493800166238805e-07,
+      "loss": 0.2917,
+      "step": 945
+    },
+    {
+      "epoch": 0.47681451612903225,
+      "grad_norm": 0.22398506177876373,
+      "learning_rate": 9.492622049589812e-07,
+      "loss": 0.3093,
+      "step": 946
+    },
+    {
+      "epoch": 0.47731854838709675,
+      "grad_norm": 0.19821540893857262,
+      "learning_rate": 9.491442636847189e-07,
+      "loss": 0.2946,
+      "step": 947
+    },
+    {
+      "epoch": 0.4778225806451613,
+      "grad_norm": 0.19307685380255402,
+      "learning_rate": 9.490261928351189e-07,
+      "loss": 0.2969,
+      "step": 948
+    },
+    {
+      "epoch": 0.4783266129032258,
+      "grad_norm": 0.216166144258633,
+      "learning_rate": 9.489079924442438e-07,
+      "loss": 0.2981,
+      "step": 949
+    },
+    {
+      "epoch": 0.4788306451612903,
+      "grad_norm": 0.17612196294528637,
+      "learning_rate": 9.487896625461935e-07,
+      "loss": 0.2907,
+      "step": 950
+    },
+    {
+      "epoch": 0.4793346774193548,
+      "grad_norm": 0.2589705775054588,
+      "learning_rate": 9.486712031751058e-07,
+      "loss": 0.3049,
+      "step": 951
+    },
+    {
+      "epoch": 0.4798387096774194,
+      "grad_norm": 0.24483548456140278,
+      "learning_rate": 9.485526143651555e-07,
+      "loss": 0.3105,
+      "step": 952
+    },
+    {
+      "epoch": 0.4803427419354839,
+      "grad_norm": 0.20974832200044807,
+      "learning_rate": 9.484338961505544e-07,
+      "loss": 0.2935,
+      "step": 953
+    },
+    {
+      "epoch": 0.4808467741935484,
+      "grad_norm": 0.17639810431043276,
+      "learning_rate": 9.483150485655523e-07,
+      "loss": 0.2818,
+      "step": 954
+    },
+    {
+      "epoch": 0.4813508064516129,
+      "grad_norm": 0.17457844069908712,
+      "learning_rate": 9.481960716444358e-07,
+      "loss": 0.2954,
+      "step": 955
+    },
+    {
+      "epoch": 0.48185483870967744,
+      "grad_norm": 0.17714100105123767,
+      "learning_rate": 9.480769654215291e-07,
+      "loss": 0.2811,
+      "step": 956
+    },
+    {
+      "epoch": 0.48235887096774194,
+      "grad_norm": 0.1958920141207276,
+      "learning_rate": 9.479577299311934e-07,
+      "loss": 0.2898,
+      "step": 957
+    },
+    {
+      "epoch": 0.48286290322580644,
+      "grad_norm": 0.2882352875771043,
+      "learning_rate": 9.478383652078277e-07,
+      "loss": 0.296,
+      "step": 958
+    },
+    {
+      "epoch": 0.48336693548387094,
+      "grad_norm": 0.21959333859610755,
+      "learning_rate": 9.477188712858679e-07,
+      "loss": 0.2816,
+      "step": 959
+    },
+    {
+      "epoch": 0.4838709677419355,
+      "grad_norm": 0.33566206173811586,
+      "learning_rate": 9.475992481997872e-07,
+      "loss": 0.2957,
+      "step": 960
+    },
+    {
+      "epoch": 0.484375,
+      "grad_norm": 0.24126700688554154,
+      "learning_rate": 9.474794959840959e-07,
+      "loss": 0.2913,
+      "step": 961
+    },
+    {
+      "epoch": 0.4848790322580645,
+      "grad_norm": 0.25752819972368163,
+      "learning_rate": 9.473596146733422e-07,
+      "loss": 0.2815,
+      "step": 962
+    },
+    {
+      "epoch": 0.48538306451612906,
+      "grad_norm": 0.19963355802829852,
+      "learning_rate": 9.472396043021109e-07,
+      "loss": 0.2859,
+      "step": 963
+    },
+    {
+      "epoch": 0.48588709677419356,
+      "grad_norm": 0.1726400189184517,
+      "learning_rate": 9.471194649050243e-07,
+      "loss": 0.2932,
+      "step": 964
+    },
+    {
+      "epoch": 0.48639112903225806,
+      "grad_norm": 0.1769030343965752,
+      "learning_rate": 9.469991965167416e-07,
+      "loss": 0.3046,
+      "step": 965
+    },
+    {
+      "epoch": 0.48689516129032256,
+      "grad_norm": 0.19108815400810478,
+      "learning_rate": 9.4687879917196e-07,
+      "loss": 0.2912,
+      "step": 966
+    },
+    {
+      "epoch": 0.4873991935483871,
+      "grad_norm": 0.20802165343766263,
+      "learning_rate": 9.46758272905413e-07,
+      "loss": 0.2895,
+      "step": 967
+    },
+    {
+      "epoch": 0.4879032258064516,
+      "grad_norm": 0.21787519659072058,
+      "learning_rate": 9.466376177518718e-07,
+      "loss": 0.287,
+      "step": 968
+    },
+    {
+      "epoch": 0.4884072580645161,
+      "grad_norm": 0.1761978641368217,
+      "learning_rate": 9.465168337461447e-07,
+      "loss": 0.2778,
+      "step": 969
+    },
+    {
+      "epoch": 0.4889112903225806,
+      "grad_norm": 0.20346855806980782,
+      "learning_rate": 9.46395920923077e-07,
+      "loss": 0.2868,
+      "step": 970
+    },
+    {
+      "epoch": 0.4894153225806452,
+      "grad_norm": 0.1746303170409007,
+      "learning_rate": 9.462748793175514e-07,
+      "loss": 0.2853,
+      "step": 971
+    },
+    {
+      "epoch": 0.4899193548387097,
+      "grad_norm": 0.27860715367389927,
+      "learning_rate": 9.461537089644876e-07,
+      "loss": 0.2984,
+      "step": 972
+    },
+    {
+      "epoch": 0.4904233870967742,
+      "grad_norm": 0.17619075334859385,
+      "learning_rate": 9.460324098988426e-07,
+      "loss": 0.2902,
+      "step": 973
+    },
+    {
+      "epoch": 0.4909274193548387,
+      "grad_norm": 0.2092460695723593,
+      "learning_rate": 9.459109821556104e-07,
+      "loss": 0.2893,
+      "step": 974
+    },
+    {
+      "epoch": 0.49143145161290325,
+      "grad_norm": 0.19009628066171583,
+      "learning_rate": 9.457894257698221e-07,
+      "loss": 0.289,
+      "step": 975
+    },
+    {
+      "epoch": 0.49193548387096775,
+      "grad_norm": 0.19214552662936102,
+      "learning_rate": 9.456677407765459e-07,
+      "loss": 0.2914,
+      "step": 976
+    },
+    {
+      "epoch": 0.49243951612903225,
+      "grad_norm": 0.21669299426430066,
+      "learning_rate": 9.455459272108873e-07,
+      "loss": 0.2829,
+      "step": 977
+    },
+    {
+      "epoch": 0.49294354838709675,
+      "grad_norm": 0.22291135020577527,
+      "learning_rate": 9.454239851079888e-07,
+      "loss": 0.2927,
+      "step": 978
+    },
+    {
+      "epoch": 0.4934475806451613,
+      "grad_norm": 0.1874075382197213,
+      "learning_rate": 9.453019145030296e-07,
+      "loss": 0.2821,
+      "step": 979
+    },
+    {
+      "epoch": 0.4939516129032258,
+      "grad_norm": 0.17233245804235334,
+      "learning_rate": 9.451797154312269e-07,
+      "loss": 0.2764,
+      "step": 980
+    },
+    {
+      "epoch": 0.4944556451612903,
+      "grad_norm": 0.1834747384575909,
+      "learning_rate": 9.450573879278338e-07,
+      "loss": 0.2735,
+      "step": 981
+    },
+    {
+      "epoch": 0.4949596774193548,
+      "grad_norm": 0.17012166060463613,
+      "learning_rate": 9.449349320281414e-07,
+      "loss": 0.2676,
+      "step": 982
+    },
+    {
+      "epoch": 0.4954637096774194,
+      "grad_norm": 0.24518957923359821,
+      "learning_rate": 9.448123477674773e-07,
+      "loss": 0.2964,
+      "step": 983
+    },
+    {
+      "epoch": 0.4959677419354839,
+      "grad_norm": 0.2037237351416288,
+      "learning_rate": 9.446896351812064e-07,
+      "loss": 0.2989,
+      "step": 984
+    },
+    {
+      "epoch": 0.4964717741935484,
+      "grad_norm": 0.17254543605119085,
+      "learning_rate": 9.445667943047303e-07,
+      "loss": 0.2816,
+      "step": 985
+    },
+    {
+      "epoch": 0.4969758064516129,
+      "grad_norm": 0.17270938767860752,
+      "learning_rate": 9.444438251734881e-07,
+      "loss": 0.29,
+      "step": 986
+    },
+    {
+      "epoch": 0.49747983870967744,
+      "grad_norm": 0.18549793090433508,
+      "learning_rate": 9.443207278229554e-07,
+      "loss": 0.2938,
+      "step": 987
+    },
+    {
+      "epoch": 0.49798387096774194,
+      "grad_norm": 0.17378591850071334,
+      "learning_rate": 9.441975022886453e-07,
+      "loss": 0.2863,
+      "step": 988
+    },
+    {
+      "epoch": 0.49848790322580644,
+      "grad_norm": 0.2036952349756055,
+      "learning_rate": 9.440741486061075e-07,
+      "loss": 0.2979,
+      "step": 989
+    },
+    {
+      "epoch": 0.49899193548387094,
+      "grad_norm": 0.24105084066474844,
+      "learning_rate": 9.439506668109284e-07,
+      "loss": 0.2938,
+      "step": 990
+    },
+    {
+      "epoch": 0.4994959677419355,
+      "grad_norm": 0.18506171017034273,
+      "learning_rate": 9.438270569387323e-07,
+      "loss": 0.2985,
+      "step": 991
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.22764502062218736,
+      "learning_rate": 9.437033190251797e-07,
+      "loss": 0.2936,
+      "step": 992
+    },
+    {
+      "epoch": 0.5005040322580645,
+      "grad_norm": 0.1987156755550509,
+      "learning_rate": 9.43579453105968e-07,
+      "loss": 0.2941,
+      "step": 993
+    },
+    {
+      "epoch": 0.501008064516129,
+      "grad_norm": 0.18473840920584347,
+      "learning_rate": 9.43455459216832e-07,
+      "loss": 0.2951,
+      "step": 994
+    },
+    {
+      "epoch": 0.5015120967741935,
+      "grad_norm": 0.18615931801205335,
+      "learning_rate": 9.433313373935429e-07,
+      "loss": 0.2864,
+      "step": 995
+    },
+    {
+      "epoch": 0.5015120967741935,
+      "eval_loss": 0.3215525150299072,
+      "eval_runtime": 17.0311,
+      "eval_samples_per_second": 50.202,
+      "eval_steps_per_second": 1.057,
+      "step": 995
+    },
+    {
+      "epoch": 0.5020161290322581,
+      "grad_norm": 0.18646346465182403,
+      "learning_rate": 9.432070876719095e-07,
+      "loss": 0.2873,
+      "step": 996
+    },
+    {
+      "epoch": 0.5025201612903226,
+      "grad_norm": 0.3058678234050536,
+      "learning_rate": 9.430827100877767e-07,
+      "loss": 0.2898,
+      "step": 997
+    },
+    {
+      "epoch": 0.5030241935483871,
+      "grad_norm": 0.20468910289808356,
+      "learning_rate": 9.429582046770268e-07,
+      "loss": 0.2961,
+      "step": 998
+    },
+    {
+      "epoch": 0.5035282258064516,
+      "grad_norm": 0.18693796681842775,
+      "learning_rate": 9.428335714755788e-07,
+      "loss": 0.2917,
+      "step": 999
+    },
+    {
+      "epoch": 0.5040322580645161,
+      "grad_norm": 0.21831422621342852,
+      "learning_rate": 9.427088105193888e-07,
+      "loss": 0.2921,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5045362903225806,
+      "grad_norm": 0.2733584938342229,
+      "learning_rate": 9.425839218444492e-07,
+      "loss": 0.2848,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5050403225806451,
+      "grad_norm": 0.22194832047757593,
+      "learning_rate": 9.424589054867899e-07,
+      "loss": 0.2816,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5055443548387096,
+      "grad_norm": 0.2631541311468527,
+      "learning_rate": 9.423337614824772e-07,
+      "loss": 0.2985,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5060483870967742,
+      "grad_norm": 0.23862558481556265,
+      "learning_rate": 9.422084898676146e-07,
+      "loss": 0.2925,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5065524193548387,
+      "grad_norm": 0.18127086379393875,
+      "learning_rate": 9.420830906783418e-07,
+      "loss": 0.2893,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5070564516129032,
+      "grad_norm": 0.23817820439070797,
+      "learning_rate": 9.419575639508359e-07,
+      "loss": 0.2912,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5075604838709677,
+      "grad_norm": 0.19504523429748843,
+      "learning_rate": 9.418319097213108e-07,
+      "loss": 0.287,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5080645161290323,
+      "grad_norm": 0.20304831048415597,
+      "learning_rate": 9.417061280260165e-07,
+      "loss": 0.2804,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5085685483870968,
+      "grad_norm": 0.22706657240299058,
+      "learning_rate": 9.415802189012407e-07,
+      "loss": 0.3014,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5090725806451613,
+      "grad_norm": 0.1859609106588216,
+      "learning_rate": 9.414541823833071e-07,
+      "loss": 0.2984,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5095766129032258,
+      "grad_norm": 0.1734707367992668,
+      "learning_rate": 9.413280185085766e-07,
+      "loss": 0.2717,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5100806451612904,
+      "grad_norm": 0.1795148808426381,
+      "learning_rate": 9.412017273134464e-07,
+      "loss": 0.2963,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5105846774193549,
+      "grad_norm": 0.17831883735108947,
+      "learning_rate": 9.410753088343513e-07,
+      "loss": 0.2844,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5110887096774194,
+      "grad_norm": 0.17970109519050018,
+      "learning_rate": 9.409487631077617e-07,
+      "loss": 0.2961,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5115927419354839,
+      "grad_norm": 0.18545886299498057,
+      "learning_rate": 9.408220901701856e-07,
+      "loss": 0.2944,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5120967741935484,
+      "grad_norm": 0.2176487874487353,
+      "learning_rate": 9.40695290058167e-07,
+      "loss": 0.2805,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5126008064516129,
+      "grad_norm": 0.1848666221951678,
+      "learning_rate": 9.405683628082875e-07,
+      "loss": 0.2923,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5131048387096774,
+      "grad_norm": 0.179797631951874,
+      "learning_rate": 9.404413084571643e-07,
+      "loss": 0.2985,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5136088709677419,
+      "grad_norm": 0.18358769665622782,
+      "learning_rate": 9.403141270414521e-07,
+      "loss": 0.2935,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5141129032258065,
+      "grad_norm": 0.2395742446188053,
+      "learning_rate": 9.401868185978418e-07,
+      "loss": 0.2985,
+      "step": 1020
+    },
+    {
+      "epoch": 0.514616935483871,
+      "grad_norm": 0.21187575778017587,
+      "learning_rate": 9.400593831630612e-07,
+      "loss": 0.2934,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5151209677419355,
+      "grad_norm": 0.1735450576887722,
+      "learning_rate": 9.399318207738744e-07,
+      "loss": 0.2753,
+      "step": 1022
+    },
+    {
+      "epoch": 0.515625,
+      "grad_norm": 0.18638678577349388,
+      "learning_rate": 9.398041314670828e-07,
+      "loss": 0.2787,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 0.24434453479231824,
+      "learning_rate": 9.396763152795237e-07,
+      "loss": 0.2855,
+      "step": 1024
+    },
+    {
+      "epoch": 0.516633064516129,
+      "grad_norm": 0.18013654047589361,
+      "learning_rate": 9.395483722480711e-07,
+      "loss": 0.2946,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5171370967741935,
+      "grad_norm": 0.3202201260573074,
+      "learning_rate": 9.394203024096361e-07,
+      "loss": 0.2769,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5176411290322581,
+      "grad_norm": 0.17815074975274992,
+      "learning_rate": 9.392921058011658e-07,
+      "loss": 0.2979,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5181451612903226,
+      "grad_norm": 0.19221302362826442,
+      "learning_rate": 9.391637824596443e-07,
+      "loss": 0.2955,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5186491935483871,
+      "grad_norm": 0.2866346023755273,
+      "learning_rate": 9.390353324220921e-07,
+      "loss": 0.292,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5191532258064516,
+      "grad_norm": 0.17531514828702485,
+      "learning_rate": 9.389067557255661e-07,
+      "loss": 0.2925,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5196572580645161,
+      "grad_norm": 0.186495533039337,
+      "learning_rate": 9.3877805240716e-07,
+      "loss": 0.2951,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5201612903225806,
+      "grad_norm": 0.20577228144105744,
+      "learning_rate": 9.386492225040039e-07,
+      "loss": 0.2962,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5206653225806451,
+      "grad_norm": 0.17419768193385268,
+      "learning_rate": 9.385202660532643e-07,
+      "loss": 0.2855,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5211693548387096,
+      "grad_norm": 0.21933160804478288,
+      "learning_rate": 9.383911830921445e-07,
+      "loss": 0.3005,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5216733870967742,
+      "grad_norm": 0.18215272532714025,
+      "learning_rate": 9.38261973657884e-07,
+      "loss": 0.2836,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5221774193548387,
+      "grad_norm": 0.17976375005098702,
+      "learning_rate": 9.38132637787759e-07,
+      "loss": 0.287,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5226814516129032,
+      "grad_norm": 0.28128296670889563,
+      "learning_rate": 9.380031755190823e-07,
+      "loss": 0.2775,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5231854838709677,
+      "grad_norm": 0.23134564761372306,
+      "learning_rate": 9.378735868892024e-07,
+      "loss": 0.2968,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5236895161290323,
+      "grad_norm": 0.18342110673735257,
+      "learning_rate": 9.377438719355054e-07,
+      "loss": 0.2907,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5241935483870968,
+      "grad_norm": 0.18862222844193371,
+      "learning_rate": 9.37614030695413e-07,
+      "loss": 0.2911,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5246975806451613,
+      "grad_norm": 0.30077945487794844,
+      "learning_rate": 9.374840632063836e-07,
+      "loss": 0.2783,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5252016129032258,
+      "grad_norm": 0.18403419856623815,
+      "learning_rate": 9.37353969505912e-07,
+      "loss": 0.2873,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5257056451612904,
+      "grad_norm": 0.19055837973677883,
+      "learning_rate": 9.372237496315295e-07,
+      "loss": 0.3061,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5262096774193549,
+      "grad_norm": 0.1832527411516626,
+      "learning_rate": 9.370934036208037e-07,
+      "loss": 0.3085,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5267137096774194,
+      "grad_norm": 0.20346555869158944,
+      "learning_rate": 9.369629315113385e-07,
+      "loss": 0.3114,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5272177419354839,
+      "grad_norm": 0.1881315222709536,
+      "learning_rate": 9.368323333407746e-07,
+      "loss": 0.3027,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5277217741935484,
+      "grad_norm": 0.20889874047965842,
+      "learning_rate": 9.367016091467885e-07,
+      "loss": 0.2935,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5282258064516129,
+      "grad_norm": 0.22770873904210787,
+      "learning_rate": 9.365707589670933e-07,
+      "loss": 0.2808,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5287298387096774,
+      "grad_norm": 0.1992019569628758,
+      "learning_rate": 9.364397828394386e-07,
+      "loss": 0.2804,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5292338709677419,
+      "grad_norm": 0.1753127135289721,
+      "learning_rate": 9.3630868080161e-07,
+      "loss": 0.2811,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5297379032258065,
+      "grad_norm": 0.17661722427381488,
+      "learning_rate": 9.361774528914299e-07,
+      "loss": 0.2806,
+      "step": 1051
+    },
+    {
+      "epoch": 0.530241935483871,
+      "grad_norm": 0.1889290840413473,
+      "learning_rate": 9.360460991467567e-07,
+      "loss": 0.2913,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5307459677419355,
+      "grad_norm": 0.1829226305533146,
+      "learning_rate": 9.35914619605485e-07,
+      "loss": 0.2979,
+      "step": 1053
+    },
+    {
+      "epoch": 0.53125,
+      "grad_norm": 0.19271715211398896,
+      "learning_rate": 9.357830143055459e-07,
+      "loss": 0.282,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5317540322580645,
+      "grad_norm": 0.18190812734541403,
+      "learning_rate": 9.356512832849066e-07,
+      "loss": 0.2941,
+      "step": 1055
+    },
+    {
+      "epoch": 0.532258064516129,
+      "grad_norm": 0.18587084451978964,
+      "learning_rate": 9.35519426581571e-07,
+      "loss": 0.3082,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5327620967741935,
+      "grad_norm": 0.2198278002815469,
+      "learning_rate": 9.353874442335785e-07,
+      "loss": 0.302,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5332661290322581,
+      "grad_norm": 0.17574611035139306,
+      "learning_rate": 9.352553362790054e-07,
+      "loss": 0.2882,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5337701612903226,
+      "grad_norm": 0.1747327396768012,
+      "learning_rate": 9.351231027559643e-07,
+      "loss": 0.2867,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5342741935483871,
+      "grad_norm": 0.22698038959811936,
+      "learning_rate": 9.349907437026034e-07,
+      "loss": 0.2811,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5347782258064516,
+      "grad_norm": 0.18501161743729416,
+      "learning_rate": 9.348582591571075e-07,
+      "loss": 0.2917,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5352822580645161,
+      "grad_norm": 0.19282052250090315,
+      "learning_rate": 9.347256491576976e-07,
+      "loss": 0.2725,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5357862903225806,
+      "grad_norm": 0.18478580670133535,
+      "learning_rate": 9.345929137426311e-07,
+      "loss": 0.2774,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5362903225806451,
+      "grad_norm": 0.17685008655382356,
+      "learning_rate": 9.344600529502009e-07,
+      "loss": 0.2945,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5367943548387096,
+      "grad_norm": 0.20983177271923298,
+      "learning_rate": 9.34327066818737e-07,
+      "loss": 0.281,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5372983870967742,
+      "grad_norm": 0.2077770172618288,
+      "learning_rate": 9.341939553866048e-07,
+      "loss": 0.2922,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5378024193548387,
+      "grad_norm": 0.22247565527885046,
+      "learning_rate": 9.340607186922059e-07,
+      "loss": 0.2788,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5383064516129032,
+      "grad_norm": 0.2072454495569315,
+      "learning_rate": 9.339273567739787e-07,
+      "loss": 0.2919,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5388104838709677,
+      "grad_norm": 0.1745117991986638,
+      "learning_rate": 9.337938696703971e-07,
+      "loss": 0.2847,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5393145161290323,
+      "grad_norm": 0.2100754257810942,
+      "learning_rate": 9.336602574199713e-07,
+      "loss": 0.3018,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5398185483870968,
+      "grad_norm": 0.20378489641904865,
+      "learning_rate": 9.335265200612477e-07,
+      "loss": 0.2911,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5403225806451613,
+      "grad_norm": 0.20915414310743302,
+      "learning_rate": 9.333926576328087e-07,
+      "loss": 0.2934,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5408266129032258,
+      "grad_norm": 0.29515599957599536,
+      "learning_rate": 9.332586701732725e-07,
+      "loss": 0.2899,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5413306451612904,
+      "grad_norm": 0.1983091401677111,
+      "learning_rate": 9.331245577212938e-07,
+      "loss": 0.2741,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5418346774193549,
+      "grad_norm": 0.2538289487328285,
+      "learning_rate": 9.329903203155633e-07,
+      "loss": 0.2927,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5423387096774194,
+      "grad_norm": 0.17986202552902045,
+      "learning_rate": 9.328559579948078e-07,
+      "loss": 0.2922,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5428427419354839,
+      "grad_norm": 0.1825249327016375,
+      "learning_rate": 9.327214707977898e-07,
+      "loss": 0.3001,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5433467741935484,
+      "grad_norm": 0.2444225537955273,
+      "learning_rate": 9.325868587633079e-07,
+      "loss": 0.2947,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5438508064516129,
+      "grad_norm": 0.41133577581929226,
+      "learning_rate": 9.32452121930197e-07,
+      "loss": 0.2912,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5443548387096774,
+      "grad_norm": 0.26649460477657994,
+      "learning_rate": 9.323172603373278e-07,
+      "loss": 0.2992,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5448588709677419,
+      "grad_norm": 0.18547372683540594,
+      "learning_rate": 9.321822740236071e-07,
+      "loss": 0.3008,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5453629032258065,
+      "grad_norm": 0.2046461782645962,
+      "learning_rate": 9.320471630279776e-07,
+      "loss": 0.2998,
+      "step": 1082
+    },
+    {
+      "epoch": 0.545866935483871,
+      "grad_norm": 0.18110940549710675,
+      "learning_rate": 9.319119273894179e-07,
+      "loss": 0.2761,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5463709677419355,
+      "grad_norm": 0.20900035767008068,
+      "learning_rate": 9.317765671469428e-07,
+      "loss": 0.2808,
+      "step": 1084
+    },
+    {
+      "epoch": 0.546875,
+      "grad_norm": 0.36426577176365416,
+      "learning_rate": 9.316410823396026e-07,
+      "loss": 0.2904,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5473790322580645,
+      "grad_norm": 0.2316323693848249,
+      "learning_rate": 9.315054730064842e-07,
+      "loss": 0.284,
+      "step": 1086
+    },
+    {
+      "epoch": 0.547883064516129,
+      "grad_norm": 0.18001191011621256,
+      "learning_rate": 9.313697391867099e-07,
+      "loss": 0.2781,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5483870967741935,
+      "grad_norm": 0.18570361057953308,
+      "learning_rate": 9.31233880919438e-07,
+      "loss": 0.3016,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5488911290322581,
+      "grad_norm": 0.1965322122969676,
+      "learning_rate": 9.310978982438627e-07,
+      "loss": 0.2717,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5493951612903226,
+      "grad_norm": 0.18899479748188522,
+      "learning_rate": 9.309617911992143e-07,
+      "loss": 0.3011,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5498991935483871,
+      "grad_norm": 0.19454709669812478,
+      "learning_rate": 9.308255598247589e-07,
+      "loss": 0.2959,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5504032258064516,
+      "grad_norm": 0.25948798557199915,
+      "learning_rate": 9.306892041597982e-07,
+      "loss": 0.2894,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5509072580645161,
+      "grad_norm": 0.18772311499552344,
+      "learning_rate": 9.305527242436702e-07,
+      "loss": 0.2753,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5514112903225806,
+      "grad_norm": 0.1933597255454464,
+      "learning_rate": 9.304161201157484e-07,
+      "loss": 0.288,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5519153225806451,
+      "grad_norm": 0.1769738727867225,
+      "learning_rate": 9.30279391815442e-07,
+      "loss": 0.2872,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5524193548387096,
+      "grad_norm": 0.18100112677940808,
+      "learning_rate": 9.301425393821967e-07,
+      "loss": 0.2818,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5529233870967742,
+      "grad_norm": 0.23375376145862312,
+      "learning_rate": 9.300055628554933e-07,
+      "loss": 0.2949,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5534274193548387,
+      "grad_norm": 0.19417823391326344,
+      "learning_rate": 9.298684622748487e-07,
+      "loss": 0.2877,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5539314516129032,
+      "grad_norm": 0.1744889878512034,
+      "learning_rate": 9.297312376798158e-07,
+      "loss": 0.2842,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5544354838709677,
+      "grad_norm": 0.16932825168461108,
+      "learning_rate": 9.295938891099828e-07,
+      "loss": 0.2951,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5549395161290323,
+      "grad_norm": 0.17870713463323182,
+      "learning_rate": 9.294564166049739e-07,
+      "loss": 0.2962,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5554435483870968,
+      "grad_norm": 0.2132310949282931,
+      "learning_rate": 9.293188202044493e-07,
+      "loss": 0.289,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5559475806451613,
+      "grad_norm": 0.19629585998507623,
+      "learning_rate": 9.291810999481045e-07,
+      "loss": 0.2867,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5564516129032258,
+      "grad_norm": 0.173069340951293,
+      "learning_rate": 9.29043255875671e-07,
+      "loss": 0.2885,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5569556451612904,
+      "grad_norm": 0.18743625034354597,
+      "learning_rate": 9.289052880269159e-07,
+      "loss": 0.2889,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5574596774193549,
+      "grad_norm": 0.1747610249774233,
+      "learning_rate": 9.287671964416423e-07,
+      "loss": 0.2981,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5579637096774194,
+      "grad_norm": 0.1803315112726341,
+      "learning_rate": 9.286289811596883e-07,
+      "loss": 0.2958,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5584677419354839,
+      "grad_norm": 0.18669991895521104,
+      "learning_rate": 9.284906422209288e-07,
+      "loss": 0.2868,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5589717741935484,
+      "grad_norm": 0.1836448818997824,
+      "learning_rate": 9.283521796652732e-07,
+      "loss": 0.2896,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5594758064516129,
+      "grad_norm": 0.17629063140324913,
+      "learning_rate": 9.282135935326672e-07,
+      "loss": 0.2853,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5599798387096774,
+      "grad_norm": 0.17145118521395394,
+      "learning_rate": 9.280748838630923e-07,
+      "loss": 0.2845,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5604838709677419,
+      "grad_norm": 0.2113289196484219,
+      "learning_rate": 9.27936050696565e-07,
+      "loss": 0.2881,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5609879032258065,
+      "grad_norm": 0.21214281262489457,
+      "learning_rate": 9.277970940731381e-07,
+      "loss": 0.278,
+      "step": 1113
+    },
+    {
+      "epoch": 0.561491935483871,
+      "grad_norm": 0.27553648004742354,
+      "learning_rate": 9.276580140328996e-07,
+      "loss": 0.3008,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5619959677419355,
+      "grad_norm": 0.22325828203211032,
+      "learning_rate": 9.275188106159732e-07,
+      "loss": 0.2786,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.1847934093487764,
+      "learning_rate": 9.273794838625184e-07,
+      "loss": 0.2748,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5630040322580645,
+      "grad_norm": 0.1909450760618991,
+      "learning_rate": 9.272400338127299e-07,
+      "loss": 0.2772,
+      "step": 1117
+    },
+    {
+      "epoch": 0.563508064516129,
+      "grad_norm": 0.1793116127902729,
+      "learning_rate": 9.271004605068382e-07,
+      "loss": 0.3071,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5640120967741935,
+      "grad_norm": 0.1813871461319359,
+      "learning_rate": 9.269607639851095e-07,
+      "loss": 0.2968,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5645161290322581,
+      "grad_norm": 0.2411100780601475,
+      "learning_rate": 9.268209442878452e-07,
+      "loss": 0.287,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5650201612903226,
+      "grad_norm": 0.22883000547852828,
+      "learning_rate": 9.266810014553826e-07,
+      "loss": 0.2943,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5655241935483871,
+      "grad_norm": 0.2008124860780277,
+      "learning_rate": 9.265409355280941e-07,
+      "loss": 0.2911,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5660282258064516,
+      "grad_norm": 0.3472215822956789,
+      "learning_rate": 9.26400746546388e-07,
+      "loss": 0.2988,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5665322580645161,
+      "grad_norm": 0.26593221261781297,
+      "learning_rate": 9.262604345507079e-07,
+      "loss": 0.2965,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5670362903225806,
+      "grad_norm": 0.1797781151682231,
+      "learning_rate": 9.26119999581533e-07,
+      "loss": 0.2818,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5675403225806451,
+      "grad_norm": 0.18192713954001474,
+      "learning_rate": 9.25979441679378e-07,
+      "loss": 0.2847,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5680443548387096,
+      "grad_norm": 0.2591562379865001,
+      "learning_rate": 9.258387608847928e-07,
+      "loss": 0.2906,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5685483870967742,
+      "grad_norm": 0.23679205899885067,
+      "learning_rate": 9.256979572383631e-07,
+      "loss": 0.2826,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5690524193548387,
+      "grad_norm": 0.31203954672135736,
+      "learning_rate": 9.255570307807097e-07,
+      "loss": 0.2811,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5695564516129032,
+      "grad_norm": 0.24288407570831536,
+      "learning_rate": 9.254159815524891e-07,
+      "loss": 0.2914,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5700604838709677,
+      "grad_norm": 0.22988163476909568,
+      "learning_rate": 9.252748095943931e-07,
+      "loss": 0.2903,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5705645161290323,
+      "grad_norm": 0.2003178753492969,
+      "learning_rate": 9.251335149471491e-07,
+      "loss": 0.2816,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5710685483870968,
+      "grad_norm": 0.17031682685543056,
+      "learning_rate": 9.249920976515195e-07,
+      "loss": 0.275,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5715725806451613,
+      "grad_norm": 0.1827276740823613,
+      "learning_rate": 9.248505577483026e-07,
+      "loss": 0.2869,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5720766129032258,
+      "grad_norm": 0.2619379606957767,
+      "learning_rate": 9.247088952783313e-07,
+      "loss": 0.2812,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5725806451612904,
+      "grad_norm": 0.1802141216250462,
+      "learning_rate": 9.245671102824748e-07,
+      "loss": 0.2921,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5730846774193549,
+      "grad_norm": 0.19106913986785806,
+      "learning_rate": 9.244252028016371e-07,
+      "loss": 0.284,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5735887096774194,
+      "grad_norm": 0.1812826559618984,
+      "learning_rate": 9.242831728767575e-07,
+      "loss": 0.2803,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5740927419354839,
+      "grad_norm": 0.1734515314933056,
+      "learning_rate": 9.241410205488108e-07,
+      "loss": 0.2869,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5745967741935484,
+      "grad_norm": 0.19432579296048832,
+      "learning_rate": 9.23998745858807e-07,
+      "loss": 0.2856,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5751008064516129,
+      "grad_norm": 0.18377955104622667,
+      "learning_rate": 9.238563488477918e-07,
+      "loss": 0.2926,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5756048387096774,
+      "grad_norm": 0.18200221102162675,
+      "learning_rate": 9.237138295568454e-07,
+      "loss": 0.2881,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5761088709677419,
+      "grad_norm": 0.18619952457281,
+      "learning_rate": 9.23571188027084e-07,
+      "loss": 0.2842,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5766129032258065,
+      "grad_norm": 0.1962070569411852,
+      "learning_rate": 9.234284242996588e-07,
+      "loss": 0.2831,
+      "step": 1144
+    },
+    {
+      "epoch": 0.577116935483871,
+      "grad_norm": 0.17904916055767378,
+      "learning_rate": 9.23285538415756e-07,
+      "loss": 0.2927,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5776209677419355,
+      "grad_norm": 0.20485807453286825,
+      "learning_rate": 9.231425304165976e-07,
+      "loss": 0.283,
+      "step": 1146
+    },
+    {
+      "epoch": 0.578125,
+      "grad_norm": 0.17697640569451753,
+      "learning_rate": 9.229994003434405e-07,
+      "loss": 0.2987,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5786290322580645,
+      "grad_norm": 0.17424623251522328,
+      "learning_rate": 9.228561482375766e-07,
+      "loss": 0.2746,
+      "step": 1148
+    },
+    {
+      "epoch": 0.579133064516129,
+      "grad_norm": 0.19037141787800266,
+      "learning_rate": 9.227127741403336e-07,
+      "loss": 0.2873,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5796370967741935,
+      "grad_norm": 0.18343536870106394,
+      "learning_rate": 9.225692780930736e-07,
+      "loss": 0.2876,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5801411290322581,
+      "grad_norm": 0.1774843130008874,
+      "learning_rate": 9.224256601371947e-07,
+      "loss": 0.2934,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 0.1890382236967552,
+      "learning_rate": 9.222819203141295e-07,
+      "loss": 0.2716,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5811491935483871,
+      "grad_norm": 0.1813676084499092,
+      "learning_rate": 9.221380586653462e-07,
+      "loss": 0.2794,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5816532258064516,
+      "grad_norm": 0.170202685121712,
+      "learning_rate": 9.21994075232348e-07,
+      "loss": 0.2872,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5821572580645161,
+      "grad_norm": 0.1896602238301285,
+      "learning_rate": 9.218499700566733e-07,
+      "loss": 0.2904,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5826612903225806,
+      "grad_norm": 0.1816413232762944,
+      "learning_rate": 9.217057431798954e-07,
+      "loss": 0.2873,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5831653225806451,
+      "grad_norm": 0.19216255744528832,
+      "learning_rate": 9.215613946436229e-07,
+      "loss": 0.297,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5836693548387096,
+      "grad_norm": 0.17974789944825,
+      "learning_rate": 9.214169244894996e-07,
+      "loss": 0.293,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5841733870967742,
+      "grad_norm": 0.17992839360699955,
+      "learning_rate": 9.21272332759204e-07,
+      "loss": 0.287,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5846774193548387,
+      "grad_norm": 0.18513535414685894,
+      "learning_rate": 9.211276194944501e-07,
+      "loss": 0.2794,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5851814516129032,
+      "grad_norm": 0.1819833313378893,
+      "learning_rate": 9.209827847369866e-07,
+      "loss": 0.2744,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5856854838709677,
+      "grad_norm": 0.17281034086736458,
+      "learning_rate": 9.208378285285977e-07,
+      "loss": 0.2945,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5861895161290323,
+      "grad_norm": 0.23268827282788637,
+      "learning_rate": 9.206927509111022e-07,
+      "loss": 0.2931,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5866935483870968,
+      "grad_norm": 0.18019045409013396,
+      "learning_rate": 9.205475519263541e-07,
+      "loss": 0.2892,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5871975806451613,
+      "grad_norm": 0.19222136198482753,
+      "learning_rate": 9.204022316162424e-07,
+      "loss": 0.2977,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5877016129032258,
+      "grad_norm": 0.17442507458285844,
+      "learning_rate": 9.202567900226912e-07,
+      "loss": 0.286,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5882056451612904,
+      "grad_norm": 0.17138337080066757,
+      "learning_rate": 9.201112271876593e-07,
+      "loss": 0.284,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5887096774193549,
+      "grad_norm": 0.2425654731892946,
+      "learning_rate": 9.199655431531409e-07,
+      "loss": 0.2814,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5892137096774194,
+      "grad_norm": 0.1846246469710363,
+      "learning_rate": 9.198197379611647e-07,
+      "loss": 0.2854,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5897177419354839,
+      "grad_norm": 0.19156185603852782,
+      "learning_rate": 9.19673811653795e-07,
+      "loss": 0.2999,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5902217741935484,
+      "grad_norm": 0.17846888518866763,
+      "learning_rate": 9.195277642731303e-07,
+      "loss": 0.2944,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5907258064516129,
+      "grad_norm": 0.18016209804691175,
+      "learning_rate": 9.193815958613043e-07,
+      "loss": 0.2909,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5912298387096774,
+      "grad_norm": 0.1818568667655358,
+      "learning_rate": 9.192353064604858e-07,
+      "loss": 0.2977,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5917338709677419,
+      "grad_norm": 0.17227833790308136,
+      "learning_rate": 9.190888961128787e-07,
+      "loss": 0.2799,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5922379032258065,
+      "grad_norm": 0.1738364554972949,
+      "learning_rate": 9.18942364860721e-07,
+      "loss": 0.2764,
+      "step": 1175
+    },
+    {
+      "epoch": 0.592741935483871,
+      "grad_norm": 0.19302662529321246,
+      "learning_rate": 9.187957127462863e-07,
+      "loss": 0.2769,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5932459677419355,
+      "grad_norm": 0.17788471334305958,
+      "learning_rate": 9.186489398118827e-07,
+      "loss": 0.2824,
+      "step": 1177
+    },
+    {
+      "epoch": 0.59375,
+      "grad_norm": 0.17681503923586112,
+      "learning_rate": 9.185020460998534e-07,
+      "loss": 0.287,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5942540322580645,
+      "grad_norm": 0.176047828469218,
+      "learning_rate": 9.183550316525763e-07,
+      "loss": 0.2854,
+      "step": 1179
+    },
+    {
+      "epoch": 0.594758064516129,
+      "grad_norm": 0.17351698958327808,
+      "learning_rate": 9.182078965124643e-07,
+      "loss": 0.3057,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5952620967741935,
+      "grad_norm": 0.1821800882504011,
+      "learning_rate": 9.180606407219644e-07,
+      "loss": 0.2759,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5957661290322581,
+      "grad_norm": 0.17241468751173047,
+      "learning_rate": 9.179132643235598e-07,
+      "loss": 0.2909,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5962701612903226,
+      "grad_norm": 0.18984463320176223,
+      "learning_rate": 9.177657673597671e-07,
+      "loss": 0.2772,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5967741935483871,
+      "grad_norm": 0.17848369146299375,
+      "learning_rate": 9.176181498731385e-07,
+      "loss": 0.293,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5972782258064516,
+      "grad_norm": 0.17602029953890927,
+      "learning_rate": 9.174704119062606e-07,
+      "loss": 0.2843,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5977822580645161,
+      "grad_norm": 0.19474495339674502,
+      "learning_rate": 9.173225535017551e-07,
+      "loss": 0.2928,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5982862903225806,
+      "grad_norm": 0.17273250226175776,
+      "learning_rate": 9.171745747022778e-07,
+      "loss": 0.2914,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5987903225806451,
+      "grad_norm": 0.21145261381905367,
+      "learning_rate": 9.170264755505201e-07,
+      "loss": 0.2928,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5992943548387096,
+      "grad_norm": 0.2039002377170998,
+      "learning_rate": 9.168782560892077e-07,
+      "loss": 0.2701,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5997983870967742,
+      "grad_norm": 0.17988346160874077,
+      "learning_rate": 9.167299163611007e-07,
+      "loss": 0.2751,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6003024193548387,
+      "grad_norm": 0.2493981489367964,
+      "learning_rate": 9.165814564089944e-07,
+      "loss": 0.2657,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6008064516129032,
+      "grad_norm": 0.1797786958096351,
+      "learning_rate": 9.164328762757184e-07,
+      "loss": 0.2833,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6013104838709677,
+      "grad_norm": 0.19193093734013503,
+      "learning_rate": 9.162841760041373e-07,
+      "loss": 0.2858,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6018145161290323,
+      "grad_norm": 0.2769241015425666,
+      "learning_rate": 9.161353556371503e-07,
+      "loss": 0.3026,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6018145161290323,
+      "eval_loss": 0.31878647208213806,
+      "eval_runtime": 16.9743,
+      "eval_samples_per_second": 50.37,
+      "eval_steps_per_second": 1.06,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6023185483870968,
+      "grad_norm": 0.181263001899909,
+      "learning_rate": 9.159864152176908e-07,
+      "loss": 0.304,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6028225806451613,
+      "grad_norm": 0.17978051150161645,
+      "learning_rate": 9.158373547887275e-07,
+      "loss": 0.2873,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6033266129032258,
+      "grad_norm": 0.20272556727914262,
+      "learning_rate": 9.156881743932634e-07,
+      "loss": 0.2815,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6038306451612904,
+      "grad_norm": 0.21068658372368834,
+      "learning_rate": 9.155388740743361e-07,
+      "loss": 0.2853,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6043346774193549,
+      "grad_norm": 0.22437088927660387,
+      "learning_rate": 9.153894538750176e-07,
+      "loss": 0.283,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6048387096774194,
+      "grad_norm": 0.1931533084578361,
+      "learning_rate": 9.152399138384148e-07,
+      "loss": 0.282,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6053427419354839,
+      "grad_norm": 0.18479701434227452,
+      "learning_rate": 9.150902540076693e-07,
+      "loss": 0.2965,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6058467741935484,
+      "grad_norm": 0.2210412349427078,
+      "learning_rate": 9.149404744259568e-07,
+      "loss": 0.2888,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6063508064516129,
+      "grad_norm": 0.18226679427751902,
+      "learning_rate": 9.147905751364878e-07,
+      "loss": 0.2887,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6068548387096774,
+      "grad_norm": 0.1832755725258158,
+      "learning_rate": 9.146405561825075e-07,
+      "loss": 0.2948,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6073588709677419,
+      "grad_norm": 0.1720010372993489,
+      "learning_rate": 9.144904176072952e-07,
+      "loss": 0.2829,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6078629032258065,
+      "grad_norm": 0.17565878536855542,
+      "learning_rate": 9.14340159454165e-07,
+      "loss": 0.2815,
+      "step": 1206
+    },
+    {
+      "epoch": 0.608366935483871,
+      "grad_norm": 0.18227852681153148,
+      "learning_rate": 9.141897817664657e-07,
+      "loss": 0.2878,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6088709677419355,
+      "grad_norm": 0.1965323622835235,
+      "learning_rate": 9.140392845875799e-07,
+      "loss": 0.2845,
+      "step": 1208
+    },
+    {
+      "epoch": 0.609375,
+      "grad_norm": 0.17206440789039248,
+      "learning_rate": 9.138886679609254e-07,
+      "loss": 0.2852,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6098790322580645,
+      "grad_norm": 0.17264736183064056,
+      "learning_rate": 9.137379319299542e-07,
+      "loss": 0.274,
+      "step": 1210
+    },
+    {
+      "epoch": 0.610383064516129,
+      "grad_norm": 0.18598352903511223,
+      "learning_rate": 9.135870765381525e-07,
+      "loss": 0.2896,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6108870967741935,
+      "grad_norm": 0.19699736881940794,
+      "learning_rate": 9.134361018290413e-07,
+      "loss": 0.2763,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6113911290322581,
+      "grad_norm": 0.22313038636092158,
+      "learning_rate": 9.132850078461758e-07,
+      "loss": 0.2846,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6118951612903226,
+      "grad_norm": 0.30031307009566305,
+      "learning_rate": 9.131337946331458e-07,
+      "loss": 0.2989,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6123991935483871,
+      "grad_norm": 0.2329858913080578,
+      "learning_rate": 9.129824622335752e-07,
+      "loss": 0.3048,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6129032258064516,
+      "grad_norm": 0.17993489418040118,
+      "learning_rate": 9.128310106911226e-07,
+      "loss": 0.2932,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6134072580645161,
+      "grad_norm": 0.18148081451402603,
+      "learning_rate": 9.126794400494806e-07,
+      "loss": 0.3034,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6139112903225806,
+      "grad_norm": 0.20177903959056726,
+      "learning_rate": 9.125277503523766e-07,
+      "loss": 0.2919,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6144153225806451,
+      "grad_norm": 0.27324162227606025,
+      "learning_rate": 9.123759416435722e-07,
+      "loss": 0.2922,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6149193548387096,
+      "grad_norm": 0.34121798755615584,
+      "learning_rate": 9.122240139668631e-07,
+      "loss": 0.2832,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6154233870967742,
+      "grad_norm": 0.21697155231985218,
+      "learning_rate": 9.120719673660796e-07,
+      "loss": 0.282,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6159274193548387,
+      "grad_norm": 0.17674277405784744,
+      "learning_rate": 9.119198018850862e-07,
+      "loss": 0.2816,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6164314516129032,
+      "grad_norm": 0.16871547242020288,
+      "learning_rate": 9.117675175677815e-07,
+      "loss": 0.2756,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6169354838709677,
+      "grad_norm": 0.20057684589974267,
+      "learning_rate": 9.11615114458099e-07,
+      "loss": 0.275,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6174395161290323,
+      "grad_norm": 0.18893088728346058,
+      "learning_rate": 9.114625926000057e-07,
+      "loss": 0.2932,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6179435483870968,
+      "grad_norm": 0.24323661515306044,
+      "learning_rate": 9.113099520375032e-07,
+      "loss": 0.2884,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6184475806451613,
+      "grad_norm": 0.17657599055834897,
+      "learning_rate": 9.111571928146276e-07,
+      "loss": 0.2847,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6189516129032258,
+      "grad_norm": 0.1839549106035099,
+      "learning_rate": 9.110043149754487e-07,
+      "loss": 0.2976,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6194556451612904,
+      "grad_norm": 0.2056215870054611,
+      "learning_rate": 9.108513185640712e-07,
+      "loss": 0.2812,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6199596774193549,
+      "grad_norm": 0.17065753231349062,
+      "learning_rate": 9.106982036246331e-07,
+      "loss": 0.2857,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6204637096774194,
+      "grad_norm": 0.18476180000705647,
+      "learning_rate": 9.105449702013076e-07,
+      "loss": 0.2965,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6209677419354839,
+      "grad_norm": 0.2000432350203024,
+      "learning_rate": 9.103916183383013e-07,
+      "loss": 0.289,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6214717741935484,
+      "grad_norm": 0.17671157012984728,
+      "learning_rate": 9.102381480798553e-07,
+      "loss": 0.2887,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6219758064516129,
+      "grad_norm": 0.19186276950774964,
+      "learning_rate": 9.100845594702451e-07,
+      "loss": 0.2855,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6224798387096774,
+      "grad_norm": 0.17315544977260117,
+      "learning_rate": 9.099308525537796e-07,
+      "loss": 0.2787,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6229838709677419,
+      "grad_norm": 0.21400229908517301,
+      "learning_rate": 9.097770273748027e-07,
+      "loss": 0.2923,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6234879032258065,
+      "grad_norm": 0.19574643374456288,
+      "learning_rate": 9.096230839776917e-07,
+      "loss": 0.2849,
+      "step": 1237
+    },
+    {
+      "epoch": 0.623991935483871,
+      "grad_norm": 0.18182176245833126,
+      "learning_rate": 9.094690224068585e-07,
+      "loss": 0.282,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6244959677419355,
+      "grad_norm": 0.23124623736324354,
+      "learning_rate": 9.09314842706749e-07,
+      "loss": 0.2964,
+      "step": 1239
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.17853145875824106,
+      "learning_rate": 9.091605449218427e-07,
+      "loss": 0.3069,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6255040322580645,
+      "grad_norm": 0.2052782675999351,
+      "learning_rate": 9.09006129096654e-07,
+      "loss": 0.3013,
+      "step": 1241
+    },
+    {
+      "epoch": 0.626008064516129,
+      "grad_norm": 0.1823088396970403,
+      "learning_rate": 9.088515952757306e-07,
+      "loss": 0.2922,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6265120967741935,
+      "grad_norm": 0.20577771722463917,
+      "learning_rate": 9.086969435036547e-07,
+      "loss": 0.2748,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6270161290322581,
+      "grad_norm": 0.17813055873331182,
+      "learning_rate": 9.085421738250422e-07,
+      "loss": 0.2939,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6275201612903226,
+      "grad_norm": 0.1814603197270135,
+      "learning_rate": 9.083872862845436e-07,
+      "loss": 0.2884,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6280241935483871,
+      "grad_norm": 0.19148692203351,
+      "learning_rate": 9.082322809268425e-07,
+      "loss": 0.2726,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6285282258064516,
+      "grad_norm": 0.1933452685613287,
+      "learning_rate": 9.080771577966574e-07,
+      "loss": 0.2894,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6290322580645161,
+      "grad_norm": 0.18291707054663686,
+      "learning_rate": 9.0792191693874e-07,
+      "loss": 0.2999,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6295362903225806,
+      "grad_norm": 0.17702736035496525,
+      "learning_rate": 9.077665583978766e-07,
+      "loss": 0.3027,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6300403225806451,
+      "grad_norm": 0.1694200412404369,
+      "learning_rate": 9.076110822188872e-07,
+      "loss": 0.28,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6305443548387096,
+      "grad_norm": 0.18581080838795144,
+      "learning_rate": 9.074554884466254e-07,
+      "loss": 0.2938,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6310483870967742,
+      "grad_norm": 0.1997263136466941,
+      "learning_rate": 9.072997771259793e-07,
+      "loss": 0.2802,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6315524193548387,
+      "grad_norm": 0.21481822334565182,
+      "learning_rate": 9.071439483018708e-07,
+      "loss": 0.2932,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6320564516129032,
+      "grad_norm": 0.1791325184424046,
+      "learning_rate": 9.06988002019255e-07,
+      "loss": 0.2834,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6325604838709677,
+      "grad_norm": 0.1836618100696337,
+      "learning_rate": 9.06831938323122e-07,
+      "loss": 0.2938,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6330645161290323,
+      "grad_norm": 0.19292497762523103,
+      "learning_rate": 9.066757572584948e-07,
+      "loss": 0.2896,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6335685483870968,
+      "grad_norm": 0.18240313649056145,
+      "learning_rate": 9.065194588704311e-07,
+      "loss": 0.2814,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6340725806451613,
+      "grad_norm": 0.22052975724510962,
+      "learning_rate": 9.063630432040216e-07,
+      "loss": 0.2886,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6345766129032258,
+      "grad_norm": 0.20672385722341882,
+      "learning_rate": 9.062065103043915e-07,
+      "loss": 0.2954,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6350806451612904,
+      "grad_norm": 0.17995424113662914,
+      "learning_rate": 9.060498602166995e-07,
+      "loss": 0.2866,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6355846774193549,
+      "grad_norm": 0.17630973219376414,
+      "learning_rate": 9.058930929861381e-07,
+      "loss": 0.2803,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6360887096774194,
+      "grad_norm": 0.2239699462767389,
+      "learning_rate": 9.057362086579336e-07,
+      "loss": 0.2944,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6365927419354839,
+      "grad_norm": 0.30137928217223403,
+      "learning_rate": 9.055792072773466e-07,
+      "loss": 0.297,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6370967741935484,
+      "grad_norm": 0.22695762562998079,
+      "learning_rate": 9.054220888896706e-07,
+      "loss": 0.2815,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6376008064516129,
+      "grad_norm": 0.1751438430718199,
+      "learning_rate": 9.052648535402334e-07,
+      "loss": 0.2862,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6381048387096774,
+      "grad_norm": 0.17639322627147327,
+      "learning_rate": 9.051075012743965e-07,
+      "loss": 0.2648,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6386088709677419,
+      "grad_norm": 0.18539526527243927,
+      "learning_rate": 9.049500321375549e-07,
+      "loss": 0.2719,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6391129032258065,
+      "grad_norm": 0.2689595278098081,
+      "learning_rate": 9.047924461751376e-07,
+      "loss": 0.2875,
+      "step": 1268
+    },
+    {
+      "epoch": 0.639616935483871,
+      "grad_norm": 0.178579851083188,
+      "learning_rate": 9.046347434326072e-07,
+      "loss": 0.2915,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6401209677419355,
+      "grad_norm": 0.17904389890697903,
+      "learning_rate": 9.044769239554599e-07,
+      "loss": 0.2745,
+      "step": 1270
+    },
+    {
+      "epoch": 0.640625,
+      "grad_norm": 0.1896562559936013,
+      "learning_rate": 9.043189877892254e-07,
+      "loss": 0.2883,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6411290322580645,
+      "grad_norm": 0.1908189648372141,
+      "learning_rate": 9.041609349794678e-07,
+      "loss": 0.283,
+      "step": 1272
+    },
+    {
+      "epoch": 0.641633064516129,
+      "grad_norm": 0.21147318798440223,
+      "learning_rate": 9.040027655717841e-07,
+      "loss": 0.301,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6421370967741935,
+      "grad_norm": 0.18345975087456926,
+      "learning_rate": 9.03844479611805e-07,
+      "loss": 0.3139,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6426411290322581,
+      "grad_norm": 0.19734085403857743,
+      "learning_rate": 9.036860771451954e-07,
+      "loss": 0.2733,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6431451612903226,
+      "grad_norm": 0.17121744147827178,
+      "learning_rate": 9.035275582176533e-07,
+      "loss": 0.2846,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6436491935483871,
+      "grad_norm": 0.17622953005984843,
+      "learning_rate": 9.033689228749102e-07,
+      "loss": 0.2965,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6441532258064516,
+      "grad_norm": 0.18180991170706243,
+      "learning_rate": 9.032101711627316e-07,
+      "loss": 0.2863,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6446572580645161,
+      "grad_norm": 0.18383068016833068,
+      "learning_rate": 9.030513031269165e-07,
+      "loss": 0.2916,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 0.23069065361517252,
+      "learning_rate": 9.02892318813297e-07,
+      "loss": 0.285,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6456653225806451,
+      "grad_norm": 0.27998314668853796,
+      "learning_rate": 9.027332182677397e-07,
+      "loss": 0.2854,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6461693548387096,
+      "grad_norm": 0.17547605295017601,
+      "learning_rate": 9.025740015361433e-07,
+      "loss": 0.2839,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6466733870967742,
+      "grad_norm": 0.17884356887542877,
+      "learning_rate": 9.024146686644415e-07,
+      "loss": 0.2922,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6471774193548387,
+      "grad_norm": 0.18212333265861616,
+      "learning_rate": 9.022552196986006e-07,
+      "loss": 0.2735,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6476814516129032,
+      "grad_norm": 0.17605258249805855,
+      "learning_rate": 9.020956546846205e-07,
+      "loss": 0.2907,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6481854838709677,
+      "grad_norm": 0.23268702781185105,
+      "learning_rate": 9.01935973668535e-07,
+      "loss": 0.2899,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6486895161290323,
+      "grad_norm": 0.17715332276237264,
+      "learning_rate": 9.017761766964111e-07,
+      "loss": 0.2966,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6491935483870968,
+      "grad_norm": 0.18537544001467623,
+      "learning_rate": 9.01616263814349e-07,
+      "loss": 0.2851,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6496975806451613,
+      "grad_norm": 0.19742055543365888,
+      "learning_rate": 9.014562350684824e-07,
+      "loss": 0.2904,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6502016129032258,
+      "grad_norm": 0.17753937663898317,
+      "learning_rate": 9.012960905049791e-07,
+      "loss": 0.2909,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6507056451612904,
+      "grad_norm": 0.18127362373513947,
+      "learning_rate": 9.011358301700397e-07,
+      "loss": 0.2991,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6512096774193549,
+      "grad_norm": 0.1897788399464913,
+      "learning_rate": 9.009754541098982e-07,
+      "loss": 0.2995,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6517137096774194,
+      "grad_norm": 0.17813293972191205,
+      "learning_rate": 9.008149623708219e-07,
+      "loss": 0.2765,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6522177419354839,
+      "grad_norm": 0.18329773158412632,
+      "learning_rate": 9.006543549991119e-07,
+      "loss": 0.2936,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6527217741935484,
+      "grad_norm": 0.20529257571242965,
+      "learning_rate": 9.004936320411025e-07,
+      "loss": 0.2949,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6532258064516129,
+      "grad_norm": 0.21151253255173527,
+      "learning_rate": 9.003327935431612e-07,
+      "loss": 0.2771,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6537298387096774,
+      "grad_norm": 0.1796301906371901,
+      "learning_rate": 9.001718395516889e-07,
+      "loss": 0.2869,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6542338709677419,
+      "grad_norm": 0.1717234788709778,
+      "learning_rate": 9.000107701131196e-07,
+      "loss": 0.2823,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6547379032258065,
+      "grad_norm": 0.17845055491572026,
+      "learning_rate": 8.998495852739213e-07,
+      "loss": 0.3044,
+      "step": 1299
+    },
+    {
+      "epoch": 0.655241935483871,
+      "grad_norm": 0.17442341108486234,
+      "learning_rate": 8.996882850805944e-07,
+      "loss": 0.28,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6557459677419355,
+      "grad_norm": 0.20073841491430633,
+      "learning_rate": 8.995268695796734e-07,
+      "loss": 0.2811,
+      "step": 1301
+    },
+    {
+      "epoch": 0.65625,
+      "grad_norm": 0.21128884092445868,
+      "learning_rate": 8.993653388177254e-07,
+      "loss": 0.2929,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6567540322580645,
+      "grad_norm": 0.1850833947673501,
+      "learning_rate": 8.99203692841351e-07,
+      "loss": 0.2914,
+      "step": 1303
+    },
+    {
+      "epoch": 0.657258064516129,
+      "grad_norm": 0.19939616483650877,
+      "learning_rate": 8.990419316971842e-07,
+      "loss": 0.2859,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6577620967741935,
+      "grad_norm": 0.1750117264820056,
+      "learning_rate": 8.988800554318921e-07,
+      "loss": 0.2982,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6582661290322581,
+      "grad_norm": 0.2075447584954836,
+      "learning_rate": 8.987180640921747e-07,
+      "loss": 0.299,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6587701612903226,
+      "grad_norm": 0.1790266951422517,
+      "learning_rate": 8.985559577247661e-07,
+      "loss": 0.2776,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6592741935483871,
+      "grad_norm": 0.178773579825902,
+      "learning_rate": 8.983937363764324e-07,
+      "loss": 0.2936,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6597782258064516,
+      "grad_norm": 0.1738135103379992,
+      "learning_rate": 8.982314000939737e-07,
+      "loss": 0.2912,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6602822580645161,
+      "grad_norm": 0.18476555106324266,
+      "learning_rate": 8.98068948924223e-07,
+      "loss": 0.3009,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6607862903225806,
+      "grad_norm": 0.20244847551602765,
+      "learning_rate": 8.979063829140465e-07,
+      "loss": 0.2945,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6612903225806451,
+      "grad_norm": 0.19339358525691477,
+      "learning_rate": 8.977437021103433e-07,
+      "loss": 0.2788,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6617943548387096,
+      "grad_norm": 0.18712592000883932,
+      "learning_rate": 8.975809065600459e-07,
+      "loss": 0.3064,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6622983870967742,
+      "grad_norm": 0.22452691139273012,
+      "learning_rate": 8.974179963101201e-07,
+      "loss": 0.2898,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6628024193548387,
+      "grad_norm": 0.1796052556590874,
+      "learning_rate": 8.97254971407564e-07,
+      "loss": 0.2813,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6633064516129032,
+      "grad_norm": 0.17710526677908528,
+      "learning_rate": 8.970918318994096e-07,
+      "loss": 0.2894,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6638104838709677,
+      "grad_norm": 0.18446794030145566,
+      "learning_rate": 8.969285778327215e-07,
+      "loss": 0.2796,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6643145161290323,
+      "grad_norm": 0.22546463300214195,
+      "learning_rate": 8.967652092545976e-07,
+      "loss": 0.292,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6648185483870968,
+      "grad_norm": 0.24026811231174375,
+      "learning_rate": 8.966017262121687e-07,
+      "loss": 0.2846,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6653225806451613,
+      "grad_norm": 0.18394372591314476,
+      "learning_rate": 8.964381287525986e-07,
+      "loss": 0.2994,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6658266129032258,
+      "grad_norm": 0.1803179306900601,
+      "learning_rate": 8.962744169230841e-07,
+      "loss": 0.2797,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6663306451612904,
+      "grad_norm": 0.16839049819871527,
+      "learning_rate": 8.96110590770855e-07,
+      "loss": 0.2631,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6668346774193549,
+      "grad_norm": 0.22155662980251253,
+      "learning_rate": 8.959466503431744e-07,
+      "loss": 0.2816,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6673387096774194,
+      "grad_norm": 0.19139441389189604,
+      "learning_rate": 8.957825956873379e-07,
+      "loss": 0.2855,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6678427419354839,
+      "grad_norm": 0.19210293112712495,
+      "learning_rate": 8.956184268506742e-07,
+      "loss": 0.2817,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6683467741935484,
+      "grad_norm": 0.1857304337358492,
+      "learning_rate": 8.954541438805452e-07,
+      "loss": 0.2873,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6688508064516129,
+      "grad_norm": 0.17268224825580095,
+      "learning_rate": 8.952897468243454e-07,
+      "loss": 0.2713,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6693548387096774,
+      "grad_norm": 0.1957606146053459,
+      "learning_rate": 8.951252357295022e-07,
+      "loss": 0.2767,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6698588709677419,
+      "grad_norm": 0.1824912944669447,
+      "learning_rate": 8.949606106434763e-07,
+      "loss": 0.287,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6703629032258065,
+      "grad_norm": 0.1798044187166269,
+      "learning_rate": 8.947958716137608e-07,
+      "loss": 0.3125,
+      "step": 1330
+    },
+    {
+      "epoch": 0.670866935483871,
+      "grad_norm": 0.17749986390962377,
+      "learning_rate": 8.946310186878821e-07,
+      "loss": 0.2675,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6713709677419355,
+      "grad_norm": 0.17752734169122558,
+      "learning_rate": 8.94466051913399e-07,
+      "loss": 0.2772,
+      "step": 1332
+    },
+    {
+      "epoch": 0.671875,
+      "grad_norm": 0.1987933223640386,
+      "learning_rate": 8.943009713379034e-07,
+      "loss": 0.2838,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6723790322580645,
+      "grad_norm": 0.1792807492758479,
+      "learning_rate": 8.941357770090203e-07,
+      "loss": 0.2872,
+      "step": 1334
+    },
+    {
+      "epoch": 0.672883064516129,
+      "grad_norm": 0.17948534651864595,
+      "learning_rate": 8.939704689744071e-07,
+      "loss": 0.3002,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6733870967741935,
+      "grad_norm": 0.1812113705229493,
+      "learning_rate": 8.93805047281754e-07,
+      "loss": 0.2954,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6738911290322581,
+      "grad_norm": 0.18844957646714744,
+      "learning_rate": 8.936395119787842e-07,
+      "loss": 0.2697,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6743951612903226,
+      "grad_norm": 0.20929818984501958,
+      "learning_rate": 8.934738631132539e-07,
+      "loss": 0.2871,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6748991935483871,
+      "grad_norm": 0.188001165636237,
+      "learning_rate": 8.933081007329515e-07,
+      "loss": 0.2973,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6754032258064516,
+      "grad_norm": 0.16918518262975,
+      "learning_rate": 8.931422248856982e-07,
+      "loss": 0.2769,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6759072580645161,
+      "grad_norm": 0.17795218348215974,
+      "learning_rate": 8.929762356193486e-07,
+      "loss": 0.2957,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6764112903225806,
+      "grad_norm": 0.23569227174562507,
+      "learning_rate": 8.928101329817894e-07,
+      "loss": 0.296,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6769153225806451,
+      "grad_norm": 0.1810084239182294,
+      "learning_rate": 8.926439170209401e-07,
+      "loss": 0.3002,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6774193548387096,
+      "grad_norm": 0.17662865898790905,
+      "learning_rate": 8.92477587784753e-07,
+      "loss": 0.2877,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6779233870967742,
+      "grad_norm": 0.20068556141454189,
+      "learning_rate": 8.923111453212131e-07,
+      "loss": 0.2987,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6784274193548387,
+      "grad_norm": 0.1856885926434189,
+      "learning_rate": 8.921445896783381e-07,
+      "loss": 0.2762,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6789314516129032,
+      "grad_norm": 0.17738716689811637,
+      "learning_rate": 8.919779209041782e-07,
+      "loss": 0.2917,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6794354838709677,
+      "grad_norm": 0.17222917661740367,
+      "learning_rate": 8.918111390468162e-07,
+      "loss": 0.2801,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6799395161290323,
+      "grad_norm": 0.19151284132256455,
+      "learning_rate": 8.916442441543678e-07,
+      "loss": 0.2706,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6804435483870968,
+      "grad_norm": 0.17262129740464702,
+      "learning_rate": 8.91477236274981e-07,
+      "loss": 0.2769,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6809475806451613,
+      "grad_norm": 0.1738945976891568,
+      "learning_rate": 8.913101154568366e-07,
+      "loss": 0.2849,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6814516129032258,
+      "grad_norm": 0.18441311480041897,
+      "learning_rate": 8.91142881748148e-07,
+      "loss": 0.2684,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6819556451612904,
+      "grad_norm": 0.18119499884546747,
+      "learning_rate": 8.90975535197161e-07,
+      "loss": 0.2886,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6824596774193549,
+      "grad_norm": 0.1829298283270949,
+      "learning_rate": 8.90808075852154e-07,
+      "loss": 0.2929,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6829637096774194,
+      "grad_norm": 0.18164824873198268,
+      "learning_rate": 8.906405037614382e-07,
+      "loss": 0.2935,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6834677419354839,
+      "grad_norm": 0.20468371157832624,
+      "learning_rate": 8.904728189733568e-07,
+      "loss": 0.2748,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6839717741935484,
+      "grad_norm": 0.1825106724862103,
+      "learning_rate": 8.90305021536286e-07,
+      "loss": 0.3001,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6844758064516129,
+      "grad_norm": 0.17622106582753136,
+      "learning_rate": 8.901371114986343e-07,
+      "loss": 0.2819,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6849798387096774,
+      "grad_norm": 0.18643087038948788,
+      "learning_rate": 8.899690889088427e-07,
+      "loss": 0.2949,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6854838709677419,
+      "grad_norm": 0.1786561974876954,
+      "learning_rate": 8.898009538153847e-07,
+      "loss": 0.2798,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6859879032258065,
+      "grad_norm": 0.17854373951330638,
+      "learning_rate": 8.896327062667663e-07,
+      "loss": 0.2989,
+      "step": 1361
+    },
+    {
+      "epoch": 0.686491935483871,
+      "grad_norm": 0.1945546348671395,
+      "learning_rate": 8.894643463115257e-07,
+      "loss": 0.2961,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6869959677419355,
+      "grad_norm": 0.22049274658233767,
+      "learning_rate": 8.892958739982338e-07,
+      "loss": 0.275,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.2020492398451764,
+      "learning_rate": 8.891272893754937e-07,
+      "loss": 0.2949,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6880040322580645,
+      "grad_norm": 0.20796221258156572,
+      "learning_rate": 8.889585924919414e-07,
+      "loss": 0.2738,
+      "step": 1365
+    },
+    {
+      "epoch": 0.688508064516129,
+      "grad_norm": 0.17715818724452845,
+      "learning_rate": 8.887897833962444e-07,
+      "loss": 0.2779,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6890120967741935,
+      "grad_norm": 0.18291220238667136,
+      "learning_rate": 8.886208621371034e-07,
+      "loss": 0.2867,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6895161290322581,
+      "grad_norm": 0.1726979302547636,
+      "learning_rate": 8.88451828763251e-07,
+      "loss": 0.2777,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6900201612903226,
+      "grad_norm": 0.23127714111621359,
+      "learning_rate": 8.882826833234525e-07,
+      "loss": 0.3048,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6905241935483871,
+      "grad_norm": 0.20073601560327714,
+      "learning_rate": 8.881134258665051e-07,
+      "loss": 0.3165,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6910282258064516,
+      "grad_norm": 0.18006036389690872,
+      "learning_rate": 8.879440564412384e-07,
+      "loss": 0.2749,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6915322580645161,
+      "grad_norm": 0.2323105469609835,
+      "learning_rate": 8.877745750965146e-07,
+      "loss": 0.2909,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6920362903225806,
+      "grad_norm": 0.178854463220733,
+      "learning_rate": 8.876049818812281e-07,
+      "loss": 0.2951,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6925403225806451,
+      "grad_norm": 0.2431317742511991,
+      "learning_rate": 8.874352768443055e-07,
+      "loss": 0.2938,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6930443548387096,
+      "grad_norm": 0.17355381981942172,
+      "learning_rate": 8.872654600347055e-07,
+      "loss": 0.2904,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6935483870967742,
+      "grad_norm": 0.19800049499752437,
+      "learning_rate": 8.870955315014193e-07,
+      "loss": 0.284,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6940524193548387,
+      "grad_norm": 0.18380625874212533,
+      "learning_rate": 8.869254912934701e-07,
+      "loss": 0.2777,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6945564516129032,
+      "grad_norm": 0.2003918541066735,
+      "learning_rate": 8.867553394599137e-07,
+      "loss": 0.2759,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6950604838709677,
+      "grad_norm": 0.18695292816627587,
+      "learning_rate": 8.865850760498375e-07,
+      "loss": 0.2978,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6955645161290323,
+      "grad_norm": 0.18229058385029212,
+      "learning_rate": 8.864147011123617e-07,
+      "loss": 0.3109,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6960685483870968,
+      "grad_norm": 0.2363688988979662,
+      "learning_rate": 8.862442146966385e-07,
+      "loss": 0.3016,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6965725806451613,
+      "grad_norm": 0.1795567842038728,
+      "learning_rate": 8.860736168518517e-07,
+      "loss": 0.2957,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6970766129032258,
+      "grad_norm": 0.18655389810189696,
+      "learning_rate": 8.859029076272182e-07,
+      "loss": 0.2809,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6975806451612904,
+      "grad_norm": 0.17971498800102143,
+      "learning_rate": 8.857320870719864e-07,
+      "loss": 0.2838,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6980846774193549,
+      "grad_norm": 0.1915148111115413,
+      "learning_rate": 8.85561155235437e-07,
+      "loss": 0.2922,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6985887096774194,
+      "grad_norm": 0.1858146547352289,
+      "learning_rate": 8.853901121668828e-07,
+      "loss": 0.287,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6990927419354839,
+      "grad_norm": 0.19284974835593893,
+      "learning_rate": 8.852189579156684e-07,
+      "loss": 0.2994,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6995967741935484,
+      "grad_norm": 0.17664023987078498,
+      "learning_rate": 8.850476925311711e-07,
+      "loss": 0.2753,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7001008064516129,
+      "grad_norm": 0.17603041348701104,
+      "learning_rate": 8.848763160627997e-07,
+      "loss": 0.2862,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7006048387096774,
+      "grad_norm": 0.17114594668139757,
+      "learning_rate": 8.847048285599952e-07,
+      "loss": 0.2832,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7011088709677419,
+      "grad_norm": 0.18467144014813242,
+      "learning_rate": 8.84533230072231e-07,
+      "loss": 0.2761,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7016129032258065,
+      "grad_norm": 0.19349322968299676,
+      "learning_rate": 8.843615206490118e-07,
+      "loss": 0.2953,
+      "step": 1392
+    },
+    {
+      "epoch": 0.702116935483871,
+      "grad_norm": 0.18516765422030596,
+      "learning_rate": 8.841897003398749e-07,
+      "loss": 0.2892,
+      "step": 1393
+    },
+    {
+      "epoch": 0.702116935483871,
+      "eval_loss": 0.31658247113227844,
+      "eval_runtime": 16.8213,
+      "eval_samples_per_second": 50.829,
+      "eval_steps_per_second": 1.07,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7026209677419355,
+      "grad_norm": 0.17985038724483002,
+      "learning_rate": 8.840177691943895e-07,
+      "loss": 0.2785,
+      "step": 1394
+    },
+    {
+      "epoch": 0.703125,
+      "grad_norm": 0.19728174011424768,
+      "learning_rate": 8.838457272621565e-07,
+      "loss": 0.2942,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7036290322580645,
+      "grad_norm": 0.22391820163520273,
+      "learning_rate": 8.836735745928089e-07,
+      "loss": 0.285,
+      "step": 1396
+    },
+    {
+      "epoch": 0.704133064516129,
+      "grad_norm": 0.17792486538644323,
+      "learning_rate": 8.835013112360118e-07,
+      "loss": 0.2889,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7046370967741935,
+      "grad_norm": 0.17801651435359686,
+      "learning_rate": 8.833289372414621e-07,
+      "loss": 0.2652,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7051411290322581,
+      "grad_norm": 0.18753019276564675,
+      "learning_rate": 8.831564526588886e-07,
+      "loss": 0.28,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7056451612903226,
+      "grad_norm": 0.2040287956468218,
+      "learning_rate": 8.82983857538052e-07,
+      "loss": 0.2754,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7061491935483871,
+      "grad_norm": 0.2003540418757947,
+      "learning_rate": 8.828111519287451e-07,
+      "loss": 0.289,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7066532258064516,
+      "grad_norm": 0.20667446057464192,
+      "learning_rate": 8.82638335880792e-07,
+      "loss": 0.2983,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7071572580645161,
+      "grad_norm": 0.18640904253903684,
+      "learning_rate": 8.824654094440496e-07,
+      "loss": 0.28,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7076612903225806,
+      "grad_norm": 0.1841514649749984,
+      "learning_rate": 8.822923726684057e-07,
+      "loss": 0.2883,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7081653225806451,
+      "grad_norm": 0.18229858409313876,
+      "learning_rate": 8.821192256037804e-07,
+      "loss": 0.2909,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7086693548387096,
+      "grad_norm": 0.21389681862115453,
+      "learning_rate": 8.819459683001257e-07,
+      "loss": 0.2807,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7091733870967742,
+      "grad_norm": 0.17509601564433855,
+      "learning_rate": 8.817726008074252e-07,
+      "loss": 0.2866,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7096774193548387,
+      "grad_norm": 0.1766608417202575,
+      "learning_rate": 8.815991231756942e-07,
+      "loss": 0.2832,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7101814516129032,
+      "grad_norm": 0.17771797879063636,
+      "learning_rate": 8.814255354549801e-07,
+      "loss": 0.2859,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7106854838709677,
+      "grad_norm": 0.18708822232800043,
+      "learning_rate": 8.81251837695362e-07,
+      "loss": 0.2857,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7111895161290323,
+      "grad_norm": 0.17909240174834304,
+      "learning_rate": 8.810780299469502e-07,
+      "loss": 0.2839,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7116935483870968,
+      "grad_norm": 0.18475558009415183,
+      "learning_rate": 8.809041122598875e-07,
+      "loss": 0.2886,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7121975806451613,
+      "grad_norm": 0.18675745708053215,
+      "learning_rate": 8.80730084684348e-07,
+      "loss": 0.2709,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7127016129032258,
+      "grad_norm": 0.17219148431342304,
+      "learning_rate": 8.805559472705375e-07,
+      "loss": 0.2867,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7132056451612904,
+      "grad_norm": 0.19026935921175026,
+      "learning_rate": 8.803817000686937e-07,
+      "loss": 0.286,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7137096774193549,
+      "grad_norm": 0.19083134675660773,
+      "learning_rate": 8.802073431290857e-07,
+      "loss": 0.2782,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7142137096774194,
+      "grad_norm": 0.17182972306227576,
+      "learning_rate": 8.800328765020146e-07,
+      "loss": 0.2764,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7147177419354839,
+      "grad_norm": 0.1685745381686828,
+      "learning_rate": 8.798583002378128e-07,
+      "loss": 0.2793,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7152217741935484,
+      "grad_norm": 0.17135712002602105,
+      "learning_rate": 8.796836143868445e-07,
+      "loss": 0.283,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7157258064516129,
+      "grad_norm": 0.18005621021675017,
+      "learning_rate": 8.795088189995052e-07,
+      "loss": 0.2774,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7162298387096774,
+      "grad_norm": 0.1757591514793466,
+      "learning_rate": 8.793339141262228e-07,
+      "loss": 0.2938,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7167338709677419,
+      "grad_norm": 0.17230964184772374,
+      "learning_rate": 8.791588998174559e-07,
+      "loss": 0.2853,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7172379032258065,
+      "grad_norm": 0.19439553949354596,
+      "learning_rate": 8.789837761236954e-07,
+      "loss": 0.2706,
+      "step": 1423
+    },
+    {
+      "epoch": 0.717741935483871,
+      "grad_norm": 0.18372365279361863,
+      "learning_rate": 8.788085430954629e-07,
+      "loss": 0.2756,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7182459677419355,
+      "grad_norm": 0.17528943529169524,
+      "learning_rate": 8.786332007833123e-07,
+      "loss": 0.2999,
+      "step": 1425
+    },
+    {
+      "epoch": 0.71875,
+      "grad_norm": 0.1718666710733766,
+      "learning_rate": 8.78457749237829e-07,
+      "loss": 0.2828,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7192540322580645,
+      "grad_norm": 0.19363050866395,
+      "learning_rate": 8.782821885096294e-07,
+      "loss": 0.3002,
+      "step": 1427
+    },
+    {
+      "epoch": 0.719758064516129,
+      "grad_norm": 0.24520324213496356,
+      "learning_rate": 8.781065186493617e-07,
+      "loss": 0.2983,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7202620967741935,
+      "grad_norm": 0.17701465130657107,
+      "learning_rate": 8.779307397077056e-07,
+      "loss": 0.2941,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7207661290322581,
+      "grad_norm": 0.17040112419340506,
+      "learning_rate": 8.777548517353722e-07,
+      "loss": 0.2758,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7212701612903226,
+      "grad_norm": 0.1741219490089326,
+      "learning_rate": 8.775788547831042e-07,
+      "loss": 0.2726,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7217741935483871,
+      "grad_norm": 0.17863386512186544,
+      "learning_rate": 8.774027489016756e-07,
+      "loss": 0.2979,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7222782258064516,
+      "grad_norm": 0.1894540149341605,
+      "learning_rate": 8.772265341418918e-07,
+      "loss": 0.2775,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7227822580645161,
+      "grad_norm": 0.17905751081552182,
+      "learning_rate": 8.770502105545898e-07,
+      "loss": 0.2759,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7232862903225806,
+      "grad_norm": 0.2250191897070994,
+      "learning_rate": 8.768737781906377e-07,
+      "loss": 0.2792,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7237903225806451,
+      "grad_norm": 0.1873361676914799,
+      "learning_rate": 8.766972371009351e-07,
+      "loss": 0.2849,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7242943548387096,
+      "grad_norm": 0.18214326291659777,
+      "learning_rate": 8.765205873364132e-07,
+      "loss": 0.3074,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7247983870967742,
+      "grad_norm": 0.18444546366186285,
+      "learning_rate": 8.763438289480343e-07,
+      "loss": 0.2974,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7253024193548387,
+      "grad_norm": 0.19982363016080767,
+      "learning_rate": 8.76166961986792e-07,
+      "loss": 0.2661,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7258064516129032,
+      "grad_norm": 0.1781094109669711,
+      "learning_rate": 8.759899865037115e-07,
+      "loss": 0.2823,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7263104838709677,
+      "grad_norm": 0.1810221506084548,
+      "learning_rate": 8.758129025498488e-07,
+      "loss": 0.2994,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7268145161290323,
+      "grad_norm": 0.17107486611373632,
+      "learning_rate": 8.75635710176292e-07,
+      "loss": 0.2733,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7273185483870968,
+      "grad_norm": 0.1920323851595878,
+      "learning_rate": 8.754584094341597e-07,
+      "loss": 0.2949,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7278225806451613,
+      "grad_norm": 0.18240009322545153,
+      "learning_rate": 8.75281000374602e-07,
+      "loss": 0.2936,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7283266129032258,
+      "grad_norm": 0.17921579254401834,
+      "learning_rate": 8.751034830488006e-07,
+      "loss": 0.2885,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7288306451612904,
+      "grad_norm": 0.17928589014845386,
+      "learning_rate": 8.749258575079678e-07,
+      "loss": 0.303,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7293346774193549,
+      "grad_norm": 0.183832758370968,
+      "learning_rate": 8.747481238033478e-07,
+      "loss": 0.2843,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7298387096774194,
+      "grad_norm": 0.1725772769965713,
+      "learning_rate": 8.745702819862155e-07,
+      "loss": 0.2702,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7303427419354839,
+      "grad_norm": 0.18098089374754683,
+      "learning_rate": 8.743923321078772e-07,
+      "loss": 0.2931,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7308467741935484,
+      "grad_norm": 0.19180274311843135,
+      "learning_rate": 8.742142742196703e-07,
+      "loss": 0.2856,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7313508064516129,
+      "grad_norm": 0.17825200131791993,
+      "learning_rate": 8.740361083729634e-07,
+      "loss": 0.2902,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7318548387096774,
+      "grad_norm": 0.17705711042600694,
+      "learning_rate": 8.738578346191563e-07,
+      "loss": 0.2989,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7323588709677419,
+      "grad_norm": 0.18722947981092045,
+      "learning_rate": 8.736794530096797e-07,
+      "loss": 0.2837,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7328629032258065,
+      "grad_norm": 0.17077286964873561,
+      "learning_rate": 8.735009635959958e-07,
+      "loss": 0.2837,
+      "step": 1454
+    },
+    {
+      "epoch": 0.733366935483871,
+      "grad_norm": 0.17523945290796009,
+      "learning_rate": 8.733223664295976e-07,
+      "loss": 0.3038,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7338709677419355,
+      "grad_norm": 0.21527334289626718,
+      "learning_rate": 8.731436615620095e-07,
+      "loss": 0.2785,
+      "step": 1456
+    },
+    {
+      "epoch": 0.734375,
+      "grad_norm": 0.17398355824973538,
+      "learning_rate": 8.729648490447864e-07,
+      "loss": 0.2849,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7348790322580645,
+      "grad_norm": 0.19117546742540467,
+      "learning_rate": 8.727859289295147e-07,
+      "loss": 0.2922,
+      "step": 1458
+    },
+    {
+      "epoch": 0.735383064516129,
+      "grad_norm": 0.17261671346075716,
+      "learning_rate": 8.726069012678119e-07,
+      "loss": 0.2916,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7358870967741935,
+      "grad_norm": 0.17887054876224195,
+      "learning_rate": 8.724277661113262e-07,
+      "loss": 0.2873,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7363911290322581,
+      "grad_norm": 0.18810141505407652,
+      "learning_rate": 8.722485235117369e-07,
+      "loss": 0.2754,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7368951612903226,
+      "grad_norm": 0.1711988777610929,
+      "learning_rate": 8.720691735207549e-07,
+      "loss": 0.2795,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7373991935483871,
+      "grad_norm": 0.181403489411845,
+      "learning_rate": 8.718897161901208e-07,
+      "loss": 0.2812,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7379032258064516,
+      "grad_norm": 0.1741359968081314,
+      "learning_rate": 8.717101515716074e-07,
+      "loss": 0.2897,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7384072580645161,
+      "grad_norm": 0.18618138691906877,
+      "learning_rate": 8.71530479717018e-07,
+      "loss": 0.2862,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7389112903225806,
+      "grad_norm": 0.18489427612306297,
+      "learning_rate": 8.713507006781867e-07,
+      "loss": 0.2798,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7394153225806451,
+      "grad_norm": 0.2032269124376584,
+      "learning_rate": 8.711708145069787e-07,
+      "loss": 0.2826,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7399193548387096,
+      "grad_norm": 0.17163938575826002,
+      "learning_rate": 8.709908212552899e-07,
+      "loss": 0.284,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7404233870967742,
+      "grad_norm": 0.17551816826492267,
+      "learning_rate": 8.708107209750473e-07,
+      "loss": 0.2986,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7409274193548387,
+      "grad_norm": 0.1707997048312571,
+      "learning_rate": 8.706305137182089e-07,
+      "loss": 0.2795,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7414314516129032,
+      "grad_norm": 0.18060170508978393,
+      "learning_rate": 8.70450199536763e-07,
+      "loss": 0.2766,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7419354838709677,
+      "grad_norm": 0.17735325459231624,
+      "learning_rate": 8.702697784827295e-07,
+      "loss": 0.2748,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7424395161290323,
+      "grad_norm": 0.18812112690658891,
+      "learning_rate": 8.700892506081587e-07,
+      "loss": 0.2933,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7429435483870968,
+      "grad_norm": 0.17809974738272327,
+      "learning_rate": 8.699086159651314e-07,
+      "loss": 0.2793,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7434475806451613,
+      "grad_norm": 0.18817435016585066,
+      "learning_rate": 8.697278746057602e-07,
+      "loss": 0.29,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7439516129032258,
+      "grad_norm": 0.18251009568650395,
+      "learning_rate": 8.695470265821871e-07,
+      "loss": 0.2832,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7444556451612904,
+      "grad_norm": 0.17400386528720746,
+      "learning_rate": 8.693660719465865e-07,
+      "loss": 0.284,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7449596774193549,
+      "grad_norm": 0.22195293952140055,
+      "learning_rate": 8.69185010751162e-07,
+      "loss": 0.2793,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7454637096774194,
+      "grad_norm": 0.16879955516508652,
+      "learning_rate": 8.690038430481489e-07,
+      "loss": 0.2851,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7459677419354839,
+      "grad_norm": 0.17379174977002937,
+      "learning_rate": 8.688225688898129e-07,
+      "loss": 0.2997,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7464717741935484,
+      "grad_norm": 0.17053843236094848,
+      "learning_rate": 8.686411883284505e-07,
+      "loss": 0.2772,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7469758064516129,
+      "grad_norm": 0.1966482045417518,
+      "learning_rate": 8.684597014163891e-07,
+      "loss": 0.2921,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7474798387096774,
+      "grad_norm": 0.1842097594300451,
+      "learning_rate": 8.682781082059861e-07,
+      "loss": 0.2902,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7479838709677419,
+      "grad_norm": 0.17503229863085937,
+      "learning_rate": 8.680964087496303e-07,
+      "loss": 0.2974,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7484879032258065,
+      "grad_norm": 0.1848664095238554,
+      "learning_rate": 8.679146030997409e-07,
+      "loss": 0.2856,
+      "step": 1485
+    },
+    {
+      "epoch": 0.748991935483871,
+      "grad_norm": 0.1762096766799662,
+      "learning_rate": 8.677326913087675e-07,
+      "loss": 0.2715,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7494959677419355,
+      "grad_norm": 0.20658420010486583,
+      "learning_rate": 8.675506734291906e-07,
+      "loss": 0.2974,
+      "step": 1487
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17593053114884585,
+      "learning_rate": 8.673685495135214e-07,
+      "loss": 0.2883,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7505040322580645,
+      "grad_norm": 0.17964689526353583,
+      "learning_rate": 8.671863196143014e-07,
+      "loss": 0.2931,
+      "step": 1489
+    },
+    {
+      "epoch": 0.751008064516129,
+      "grad_norm": 0.1794940966748331,
+      "learning_rate": 8.670039837841028e-07,
+      "loss": 0.2693,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7515120967741935,
+      "grad_norm": 0.17496758427902515,
+      "learning_rate": 8.668215420755282e-07,
+      "loss": 0.2708,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7520161290322581,
+      "grad_norm": 0.173160052063297,
+      "learning_rate": 8.666389945412112e-07,
+      "loss": 0.2903,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7525201612903226,
+      "grad_norm": 0.19227463269140316,
+      "learning_rate": 8.664563412338154e-07,
+      "loss": 0.2928,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7530241935483871,
+      "grad_norm": 0.20236023713329304,
+      "learning_rate": 8.662735822060352e-07,
+      "loss": 0.2776,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7535282258064516,
+      "grad_norm": 0.20307254860433052,
+      "learning_rate": 8.660907175105956e-07,
+      "loss": 0.2789,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7540322580645161,
+      "grad_norm": 0.18535559574459684,
+      "learning_rate": 8.659077472002518e-07,
+      "loss": 0.2977,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7545362903225806,
+      "grad_norm": 0.19262270536952245,
+      "learning_rate": 8.657246713277895e-07,
+      "loss": 0.2884,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7550403225806451,
+      "grad_norm": 0.17225775461167198,
+      "learning_rate": 8.655414899460251e-07,
+      "loss": 0.2798,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7555443548387096,
+      "grad_norm": 0.18058865422335652,
+      "learning_rate": 8.653582031078053e-07,
+      "loss": 0.2803,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7560483870967742,
+      "grad_norm": 0.193197708837322,
+      "learning_rate": 8.651748108660072e-07,
+      "loss": 0.285,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7565524193548387,
+      "grad_norm": 0.19695362977929617,
+      "learning_rate": 8.649913132735383e-07,
+      "loss": 0.2871,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7570564516129032,
+      "grad_norm": 0.17883642405940453,
+      "learning_rate": 8.648077103833365e-07,
+      "loss": 0.2799,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7575604838709677,
+      "grad_norm": 0.17172477648515241,
+      "learning_rate": 8.646240022483699e-07,
+      "loss": 0.292,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7580645161290323,
+      "grad_norm": 0.17575732800042274,
+      "learning_rate": 8.644401889216377e-07,
+      "loss": 0.2777,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7585685483870968,
+      "grad_norm": 0.17678434388091105,
+      "learning_rate": 8.642562704561684e-07,
+      "loss": 0.2956,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7590725806451613,
+      "grad_norm": 0.18058078973223446,
+      "learning_rate": 8.640722469050217e-07,
+      "loss": 0.2887,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7595766129032258,
+      "grad_norm": 0.1742859796225662,
+      "learning_rate": 8.638881183212869e-07,
+      "loss": 0.2787,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7600806451612904,
+      "grad_norm": 0.17057933697554573,
+      "learning_rate": 8.637038847580842e-07,
+      "loss": 0.2822,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7605846774193549,
+      "grad_norm": 0.17396796248874147,
+      "learning_rate": 8.635195462685637e-07,
+      "loss": 0.2898,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7610887096774194,
+      "grad_norm": 0.1754476685521095,
+      "learning_rate": 8.633351029059061e-07,
+      "loss": 0.2816,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7615927419354839,
+      "grad_norm": 0.1996562682521391,
+      "learning_rate": 8.63150554723322e-07,
+      "loss": 0.2802,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7620967741935484,
+      "grad_norm": 0.18654386311810392,
+      "learning_rate": 8.629659017740525e-07,
+      "loss": 0.2843,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7626008064516129,
+      "grad_norm": 0.1795688988462513,
+      "learning_rate": 8.627811441113688e-07,
+      "loss": 0.2972,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7631048387096774,
+      "grad_norm": 0.1700302651183351,
+      "learning_rate": 8.625962817885723e-07,
+      "loss": 0.276,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7636088709677419,
+      "grad_norm": 0.17276816293597855,
+      "learning_rate": 8.624113148589947e-07,
+      "loss": 0.2857,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7641129032258065,
+      "grad_norm": 0.17205240872531083,
+      "learning_rate": 8.622262433759976e-07,
+      "loss": 0.2898,
+      "step": 1516
+    },
+    {
+      "epoch": 0.764616935483871,
+      "grad_norm": 0.19308777537983587,
+      "learning_rate": 8.620410673929732e-07,
+      "loss": 0.2911,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7651209677419355,
+      "grad_norm": 0.18066352556552442,
+      "learning_rate": 8.618557869633438e-07,
+      "loss": 0.2851,
+      "step": 1518
+    },
+    {
+      "epoch": 0.765625,
+      "grad_norm": 0.1740551426558877,
+      "learning_rate": 8.616704021405613e-07,
+      "loss": 0.2754,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7661290322580645,
+      "grad_norm": 0.20220657101308243,
+      "learning_rate": 8.614849129781084e-07,
+      "loss": 0.2675,
+      "step": 1520
+    },
+    {
+      "epoch": 0.766633064516129,
+      "grad_norm": 0.17142601346124745,
+      "learning_rate": 8.612993195294971e-07,
+      "loss": 0.2697,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7671370967741935,
+      "grad_norm": 0.20383494070933988,
+      "learning_rate": 8.611136218482704e-07,
+      "loss": 0.2917,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7676411290322581,
+      "grad_norm": 0.17942668361212222,
+      "learning_rate": 8.609278199880007e-07,
+      "loss": 0.2996,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7681451612903226,
+      "grad_norm": 0.21582357596764432,
+      "learning_rate": 8.607419140022908e-07,
+      "loss": 0.2905,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7686491935483871,
+      "grad_norm": 0.19537946016922536,
+      "learning_rate": 8.605559039447734e-07,
+      "loss": 0.2897,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7691532258064516,
+      "grad_norm": 0.17446732973784299,
+      "learning_rate": 8.603697898691112e-07,
+      "loss": 0.2922,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7696572580645161,
+      "grad_norm": 0.17940436217661643,
+      "learning_rate": 8.601835718289971e-07,
+      "loss": 0.2897,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7701612903225806,
+      "grad_norm": 0.16967508800400594,
+      "learning_rate": 8.599972498781536e-07,
+      "loss": 0.2838,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7706653225806451,
+      "grad_norm": 0.17730457999766902,
+      "learning_rate": 8.598108240703337e-07,
+      "loss": 0.2863,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7711693548387096,
+      "grad_norm": 0.17432021927654012,
+      "learning_rate": 8.5962429445932e-07,
+      "loss": 0.2838,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7716733870967742,
+      "grad_norm": 0.18292364816715403,
+      "learning_rate": 8.594376610989249e-07,
+      "loss": 0.2889,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7721774193548387,
+      "grad_norm": 0.1692258045511517,
+      "learning_rate": 8.592509240429913e-07,
+      "loss": 0.2859,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7726814516129032,
+      "grad_norm": 0.17793048496272398,
+      "learning_rate": 8.590640833453916e-07,
+      "loss": 0.2962,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7731854838709677,
+      "grad_norm": 0.170590500112818,
+      "learning_rate": 8.58877139060028e-07,
+      "loss": 0.273,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7736895161290323,
+      "grad_norm": 0.1748601955404486,
+      "learning_rate": 8.58690091240833e-07,
+      "loss": 0.2844,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 0.18730945527731682,
+      "learning_rate": 8.585029399417687e-07,
+      "loss": 0.3091,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7746975806451613,
+      "grad_norm": 0.18332986307488439,
+      "learning_rate": 8.583156852168269e-07,
+      "loss": 0.2793,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7752016129032258,
+      "grad_norm": 0.18440491915084045,
+      "learning_rate": 8.581283271200297e-07,
+      "loss": 0.2862,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7757056451612904,
+      "grad_norm": 0.173582809480461,
+      "learning_rate": 8.579408657054286e-07,
+      "loss": 0.2792,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7762096774193549,
+      "grad_norm": 0.18014379318891854,
+      "learning_rate": 8.577533010271049e-07,
+      "loss": 0.2919,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7767137096774194,
+      "grad_norm": 0.1722503671402626,
+      "learning_rate": 8.575656331391702e-07,
+      "loss": 0.2697,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7772177419354839,
+      "grad_norm": 0.1727617532348802,
+      "learning_rate": 8.573778620957652e-07,
+      "loss": 0.3044,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7777217741935484,
+      "grad_norm": 0.17458552774869926,
+      "learning_rate": 8.571899879510609e-07,
+      "loss": 0.3022,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7782258064516129,
+      "grad_norm": 0.19575211026743364,
+      "learning_rate": 8.570020107592579e-07,
+      "loss": 0.2729,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7787298387096774,
+      "grad_norm": 0.18571171303798414,
+      "learning_rate": 8.568139305745861e-07,
+      "loss": 0.2797,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7792338709677419,
+      "grad_norm": 0.19155677402308682,
+      "learning_rate": 8.566257474513057e-07,
+      "loss": 0.2999,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7797379032258065,
+      "grad_norm": 0.19852085649960613,
+      "learning_rate": 8.564374614437065e-07,
+      "loss": 0.2948,
+      "step": 1547
+    },
+    {
+      "epoch": 0.780241935483871,
+      "grad_norm": 0.18180259804141408,
+      "learning_rate": 8.562490726061074e-07,
+      "loss": 0.272,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7807459677419355,
+      "grad_norm": 0.1757617028962345,
+      "learning_rate": 8.560605809928578e-07,
+      "loss": 0.2899,
+      "step": 1549
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 0.18002181651881002,
+      "learning_rate": 8.558719866583364e-07,
+      "loss": 0.2801,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7817540322580645,
+      "grad_norm": 0.18391600786228307,
+      "learning_rate": 8.556832896569512e-07,
+      "loss": 0.2953,
+      "step": 1551
+    },
+    {
+      "epoch": 0.782258064516129,
+      "grad_norm": 0.1696821891773392,
+      "learning_rate": 8.554944900431405e-07,
+      "loss": 0.2765,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7827620967741935,
+      "grad_norm": 0.17250755946675386,
+      "learning_rate": 8.553055878713714e-07,
+      "loss": 0.2742,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7832661290322581,
+      "grad_norm": 0.17460773929441345,
+      "learning_rate": 8.551165831961414e-07,
+      "loss": 0.2941,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7837701612903226,
+      "grad_norm": 0.17494634704080653,
+      "learning_rate": 8.549274760719767e-07,
+      "loss": 0.2838,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7842741935483871,
+      "grad_norm": 0.18400774964808703,
+      "learning_rate": 8.547382665534339e-07,
+      "loss": 0.2913,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7847782258064516,
+      "grad_norm": 0.17748115232473677,
+      "learning_rate": 8.545489546950988e-07,
+      "loss": 0.2859,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7852822580645161,
+      "grad_norm": 0.18218986931380948,
+      "learning_rate": 8.543595405515864e-07,
+      "loss": 0.2812,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7857862903225806,
+      "grad_norm": 0.1969916491736838,
+      "learning_rate": 8.541700241775419e-07,
+      "loss": 0.2888,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7862903225806451,
+      "grad_norm": 0.20960556550108175,
+      "learning_rate": 8.539804056276393e-07,
+      "loss": 0.2896,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7867943548387096,
+      "grad_norm": 0.18784903454787713,
+      "learning_rate": 8.537906849565824e-07,
+      "loss": 0.292,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7872983870967742,
+      "grad_norm": 0.17668988962371776,
+      "learning_rate": 8.536008622191047e-07,
+      "loss": 0.2832,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7878024193548387,
+      "grad_norm": 0.1938218281223095,
+      "learning_rate": 8.534109374699685e-07,
+      "loss": 0.2806,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7883064516129032,
+      "grad_norm": 0.1724503265255416,
+      "learning_rate": 8.532209107639661e-07,
+      "loss": 0.2836,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7888104838709677,
+      "grad_norm": 0.216684297634416,
+      "learning_rate": 8.530307821559192e-07,
+      "loss": 0.2834,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7893145161290323,
+      "grad_norm": 0.1952086001178248,
+      "learning_rate": 8.528405517006785e-07,
+      "loss": 0.2703,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7898185483870968,
+      "grad_norm": 0.17917540283085726,
+      "learning_rate": 8.526502194531242e-07,
+      "loss": 0.2795,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7903225806451613,
+      "grad_norm": 0.16957021131556896,
+      "learning_rate": 8.524597854681663e-07,
+      "loss": 0.2849,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7908266129032258,
+      "grad_norm": 0.17315335375074736,
+      "learning_rate": 8.522692498007436e-07,
+      "loss": 0.2706,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7913306451612904,
+      "grad_norm": 0.17415225412091026,
+      "learning_rate": 8.520786125058246e-07,
+      "loss": 0.2784,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7918346774193549,
+      "grad_norm": 0.19715463128787936,
+      "learning_rate": 8.518878736384067e-07,
+      "loss": 0.2791,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7923387096774194,
+      "grad_norm": 0.18485659414062103,
+      "learning_rate": 8.516970332535174e-07,
+      "loss": 0.2883,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7928427419354839,
+      "grad_norm": 0.17412481517409986,
+      "learning_rate": 8.515060914062124e-07,
+      "loss": 0.2891,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7933467741935484,
+      "grad_norm": 0.19186859393133207,
+      "learning_rate": 8.513150481515777e-07,
+      "loss": 0.2861,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7938508064516129,
+      "grad_norm": 0.17924622608650354,
+      "learning_rate": 8.511239035447277e-07,
+      "loss": 0.2769,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7943548387096774,
+      "grad_norm": 0.21636343101039252,
+      "learning_rate": 8.509326576408066e-07,
+      "loss": 0.2809,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7948588709677419,
+      "grad_norm": 0.1988798854713325,
+      "learning_rate": 8.507413104949878e-07,
+      "loss": 0.2817,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7953629032258065,
+      "grad_norm": 0.20133302174026105,
+      "learning_rate": 8.505498621624738e-07,
+      "loss": 0.273,
+      "step": 1578
+    },
+    {
+      "epoch": 0.795866935483871,
+      "grad_norm": 0.18247301896520096,
+      "learning_rate": 8.503583126984959e-07,
+      "loss": 0.2851,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7963709677419355,
+      "grad_norm": 0.179121678370444,
+      "learning_rate": 8.501666621583152e-07,
+      "loss": 0.2817,
+      "step": 1580
+    },
+    {
+      "epoch": 0.796875,
+      "grad_norm": 0.1794015027687013,
+      "learning_rate": 8.499749105972216e-07,
+      "loss": 0.277,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7973790322580645,
+      "grad_norm": 0.18799369028010127,
+      "learning_rate": 8.497830580705343e-07,
+      "loss": 0.2811,
+      "step": 1582
+    },
+    {
+      "epoch": 0.797883064516129,
+      "grad_norm": 0.19305091528117443,
+      "learning_rate": 8.495911046336015e-07,
+      "loss": 0.2995,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7983870967741935,
+      "grad_norm": 0.2066980461918033,
+      "learning_rate": 8.493990503418007e-07,
+      "loss": 0.2982,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7988911290322581,
+      "grad_norm": 0.17520616254907886,
+      "learning_rate": 8.492068952505382e-07,
+      "loss": 0.271,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7993951612903226,
+      "grad_norm": 0.17626401945753917,
+      "learning_rate": 8.490146394152497e-07,
+      "loss": 0.2858,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7998991935483871,
+      "grad_norm": 0.1752484381904052,
+      "learning_rate": 8.488222828913998e-07,
+      "loss": 0.2862,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8004032258064516,
+      "grad_norm": 0.18080248190940781,
+      "learning_rate": 8.486298257344821e-07,
+      "loss": 0.2815,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8009072580645161,
+      "grad_norm": 0.18782657218788468,
+      "learning_rate": 8.484372680000193e-07,
+      "loss": 0.2907,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8014112903225806,
+      "grad_norm": 0.17866164980608917,
+      "learning_rate": 8.482446097435631e-07,
+      "loss": 0.2863,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8019153225806451,
+      "grad_norm": 0.16998904102569148,
+      "learning_rate": 8.480518510206942e-07,
+      "loss": 0.2866,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8024193548387096,
+      "grad_norm": 0.17671260570065095,
+      "learning_rate": 8.478589918870225e-07,
+      "loss": 0.2965,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8024193548387096,
+      "eval_loss": 0.31474894285202026,
+      "eval_runtime": 17.2161,
+      "eval_samples_per_second": 49.663,
+      "eval_steps_per_second": 1.046,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8029233870967742,
+      "grad_norm": 0.17663891905474427,
+      "learning_rate": 8.476660323981863e-07,
+      "loss": 0.2894,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8034274193548387,
+      "grad_norm": 0.2011698318082641,
+      "learning_rate": 8.474729726098537e-07,
+      "loss": 0.2922,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8039314516129032,
+      "grad_norm": 0.20386721072025002,
+      "learning_rate": 8.472798125777208e-07,
+      "loss": 0.2797,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8044354838709677,
+      "grad_norm": 0.21744706928084864,
+      "learning_rate": 8.470865523575133e-07,
+      "loss": 0.2773,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8049395161290323,
+      "grad_norm": 0.1749022648400484,
+      "learning_rate": 8.468931920049855e-07,
+      "loss": 0.2893,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8054435483870968,
+      "grad_norm": 0.17378425941567904,
+      "learning_rate": 8.466997315759207e-07,
+      "loss": 0.2885,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8059475806451613,
+      "grad_norm": 0.2151441210122262,
+      "learning_rate": 8.465061711261312e-07,
+      "loss": 0.2812,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8064516129032258,
+      "grad_norm": 0.20189295912194238,
+      "learning_rate": 8.463125107114576e-07,
+      "loss": 0.2938,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8069556451612904,
+      "grad_norm": 0.1737385784520349,
+      "learning_rate": 8.461187503877701e-07,
+      "loss": 0.2913,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8074596774193549,
+      "grad_norm": 0.17511209441842182,
+      "learning_rate": 8.459248902109671e-07,
+      "loss": 0.2972,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8079637096774194,
+      "grad_norm": 0.17927894922934345,
+      "learning_rate": 8.457309302369762e-07,
+      "loss": 0.2751,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8084677419354839,
+      "grad_norm": 0.1711157379063894,
+      "learning_rate": 8.455368705217536e-07,
+      "loss": 0.293,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8089717741935484,
+      "grad_norm": 0.20561980849817957,
+      "learning_rate": 8.453427111212844e-07,
+      "loss": 0.2804,
+      "step": 1605
+    },
+    {
+      "epoch": 0.8094758064516129,
+      "grad_norm": 0.18016754167062388,
+      "learning_rate": 8.451484520915823e-07,
+      "loss": 0.2931,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8099798387096774,
+      "grad_norm": 0.19807530942971208,
+      "learning_rate": 8.449540934886898e-07,
+      "loss": 0.2947,
+      "step": 1607
+    },
+    {
+      "epoch": 0.8104838709677419,
+      "grad_norm": 0.18074849101785265,
+      "learning_rate": 8.447596353686783e-07,
+      "loss": 0.3026,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8109879032258065,
+      "grad_norm": 0.18227592034924972,
+      "learning_rate": 8.445650777876477e-07,
+      "loss": 0.2942,
+      "step": 1609
+    },
+    {
+      "epoch": 0.811491935483871,
+      "grad_norm": 0.1807387057488744,
+      "learning_rate": 8.443704208017265e-07,
+      "loss": 0.2905,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8119959677419355,
+      "grad_norm": 0.17752462791309626,
+      "learning_rate": 8.441756644670721e-07,
+      "loss": 0.3053,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.17960458304361665,
+      "learning_rate": 8.439808088398708e-07,
+      "loss": 0.2807,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8130040322580645,
+      "grad_norm": 0.17933572002728534,
+      "learning_rate": 8.437858539763368e-07,
+      "loss": 0.2874,
+      "step": 1613
+    },
+    {
+      "epoch": 0.813508064516129,
+      "grad_norm": 0.17388277996545867,
+      "learning_rate": 8.435907999327137e-07,
+      "loss": 0.2872,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8140120967741935,
+      "grad_norm": 0.20696380286831806,
+      "learning_rate": 8.433956467652731e-07,
+      "loss": 0.296,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8145161290322581,
+      "grad_norm": 0.19036323172621478,
+      "learning_rate": 8.432003945303157e-07,
+      "loss": 0.2919,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8150201612903226,
+      "grad_norm": 0.17417826300098163,
+      "learning_rate": 8.430050432841705e-07,
+      "loss": 0.2839,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8155241935483871,
+      "grad_norm": 0.1764029810149293,
+      "learning_rate": 8.428095930831951e-07,
+      "loss": 0.2801,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8160282258064516,
+      "grad_norm": 0.18297026792189108,
+      "learning_rate": 8.426140439837758e-07,
+      "loss": 0.2907,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8165322580645161,
+      "grad_norm": 0.1821471162865555,
+      "learning_rate": 8.42418396042327e-07,
+      "loss": 0.2871,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8170362903225806,
+      "grad_norm": 0.18365629188801313,
+      "learning_rate": 8.422226493152923e-07,
+      "loss": 0.2729,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8175403225806451,
+      "grad_norm": 0.17997053499497517,
+      "learning_rate": 8.420268038591432e-07,
+      "loss": 0.2928,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8180443548387096,
+      "grad_norm": 0.18199210579188108,
+      "learning_rate": 8.418308597303798e-07,
+      "loss": 0.2833,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8185483870967742,
+      "grad_norm": 0.17684377413059332,
+      "learning_rate": 8.41634816985531e-07,
+      "loss": 0.2821,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8190524193548387,
+      "grad_norm": 0.1729650092760548,
+      "learning_rate": 8.414386756811538e-07,
+      "loss": 0.2915,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8195564516129032,
+      "grad_norm": 0.16853170774887175,
+      "learning_rate": 8.412424358738337e-07,
+      "loss": 0.296,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8200604838709677,
+      "grad_norm": 0.17189289211370995,
+      "learning_rate": 8.410460976201847e-07,
+      "loss": 0.3022,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8205645161290323,
+      "grad_norm": 0.18303909561880122,
+      "learning_rate": 8.408496609768494e-07,
+      "loss": 0.27,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8210685483870968,
+      "grad_norm": 0.1689191180195712,
+      "learning_rate": 8.406531260004983e-07,
+      "loss": 0.283,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8215725806451613,
+      "grad_norm": 0.19059066020698312,
+      "learning_rate": 8.404564927478304e-07,
+      "loss": 0.2733,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8220766129032258,
+      "grad_norm": 0.18412586501703707,
+      "learning_rate": 8.402597612755736e-07,
+      "loss": 0.2958,
+      "step": 1631
+    },
+    {
+      "epoch": 0.8225806451612904,
+      "grad_norm": 0.1865008483611006,
+      "learning_rate": 8.400629316404833e-07,
+      "loss": 0.28,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8230846774193549,
+      "grad_norm": 0.1848533702786811,
+      "learning_rate": 8.398660038993439e-07,
+      "loss": 0.2806,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8235887096774194,
+      "grad_norm": 0.1737421957224456,
+      "learning_rate": 8.396689781089676e-07,
+      "loss": 0.285,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8240927419354839,
+      "grad_norm": 0.16653152620902592,
+      "learning_rate": 8.394718543261954e-07,
+      "loss": 0.2723,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8245967741935484,
+      "grad_norm": 0.1758054798983086,
+      "learning_rate": 8.392746326078961e-07,
+      "loss": 0.2889,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8251008064516129,
+      "grad_norm": 0.17075324603300618,
+      "learning_rate": 8.39077313010967e-07,
+      "loss": 0.2858,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8256048387096774,
+      "grad_norm": 0.18600083079394014,
+      "learning_rate": 8.388798955923335e-07,
+      "loss": 0.2735,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8261088709677419,
+      "grad_norm": 0.1961620789751493,
+      "learning_rate": 8.386823804089496e-07,
+      "loss": 0.2992,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8266129032258065,
+      "grad_norm": 0.2056030735817931,
+      "learning_rate": 8.384847675177968e-07,
+      "loss": 0.2966,
+      "step": 1640
+    },
+    {
+      "epoch": 0.827116935483871,
+      "grad_norm": 0.178659647296446,
+      "learning_rate": 8.382870569758853e-07,
+      "loss": 0.2943,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8276209677419355,
+      "grad_norm": 0.17144740006998577,
+      "learning_rate": 8.380892488402535e-07,
+      "loss": 0.2725,
+      "step": 1642
+    },
+    {
+      "epoch": 0.828125,
+      "grad_norm": 0.2048573646210055,
+      "learning_rate": 8.378913431679677e-07,
+      "loss": 0.2654,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8286290322580645,
+      "grad_norm": 0.19608901393965525,
+      "learning_rate": 8.376933400161226e-07,
+      "loss": 0.2657,
+      "step": 1644
+    },
+    {
+      "epoch": 0.829133064516129,
+      "grad_norm": 0.16996202799605425,
+      "learning_rate": 8.374952394418409e-07,
+      "loss": 0.2857,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8296370967741935,
+      "grad_norm": 0.1804403768350341,
+      "learning_rate": 8.37297041502273e-07,
+      "loss": 0.2889,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8301411290322581,
+      "grad_norm": 0.18021938104304921,
+      "learning_rate": 8.370987462545984e-07,
+      "loss": 0.2836,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8306451612903226,
+      "grad_norm": 0.18418845425920574,
+      "learning_rate": 8.369003537560237e-07,
+      "loss": 0.2812,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8311491935483871,
+      "grad_norm": 0.2151893686661586,
+      "learning_rate": 8.367018640637838e-07,
+      "loss": 0.2856,
+      "step": 1649
+    },
+    {
+      "epoch": 0.8316532258064516,
+      "grad_norm": 0.25324081072788207,
+      "learning_rate": 8.365032772351419e-07,
+      "loss": 0.2745,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8321572580645161,
+      "grad_norm": 0.17621766684272705,
+      "learning_rate": 8.363045933273889e-07,
+      "loss": 0.2813,
+      "step": 1651
+    },
+    {
+      "epoch": 0.8326612903225806,
+      "grad_norm": 0.1810792109703597,
+      "learning_rate": 8.361058123978442e-07,
+      "loss": 0.2958,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8331653225806451,
+      "grad_norm": 0.17290815775300664,
+      "learning_rate": 8.359069345038548e-07,
+      "loss": 0.2821,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8336693548387096,
+      "grad_norm": 0.17231901628427526,
+      "learning_rate": 8.357079597027954e-07,
+      "loss": 0.2793,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8341733870967742,
+      "grad_norm": 0.19031643192982242,
+      "learning_rate": 8.355088880520693e-07,
+      "loss": 0.2931,
+      "step": 1655
+    },
+    {
+      "epoch": 0.8346774193548387,
+      "grad_norm": 0.1773771991850869,
+      "learning_rate": 8.353097196091074e-07,
+      "loss": 0.2897,
+      "step": 1656
+    },
+    {
+      "epoch": 0.8351814516129032,
+      "grad_norm": 0.18048345032642082,
+      "learning_rate": 8.351104544313685e-07,
+      "loss": 0.2836,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8356854838709677,
+      "grad_norm": 0.18443696405092166,
+      "learning_rate": 8.349110925763393e-07,
+      "loss": 0.2846,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8361895161290323,
+      "grad_norm": 0.1772897016179952,
+      "learning_rate": 8.347116341015347e-07,
+      "loss": 0.2932,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8366935483870968,
+      "grad_norm": 0.1790343811345245,
+      "learning_rate": 8.34512079064497e-07,
+      "loss": 0.2915,
+      "step": 1660
+    },
+    {
+      "epoch": 0.8371975806451613,
+      "grad_norm": 0.18840621984787398,
+      "learning_rate": 8.343124275227968e-07,
+      "loss": 0.2816,
+      "step": 1661
+    },
+    {
+      "epoch": 0.8377016129032258,
+      "grad_norm": 0.17628267263776604,
+      "learning_rate": 8.341126795340321e-07,
+      "loss": 0.2805,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8382056451612904,
+      "grad_norm": 0.17600045722386132,
+      "learning_rate": 8.339128351558291e-07,
+      "loss": 0.29,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8387096774193549,
+      "grad_norm": 0.182325857327577,
+      "learning_rate": 8.337128944458415e-07,
+      "loss": 0.2903,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8392137096774194,
+      "grad_norm": 0.19535349793236487,
+      "learning_rate": 8.335128574617513e-07,
+      "loss": 0.2858,
+      "step": 1665
+    },
+    {
+      "epoch": 0.8397177419354839,
+      "grad_norm": 0.27705115942369163,
+      "learning_rate": 8.333127242612677e-07,
+      "loss": 0.2857,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8402217741935484,
+      "grad_norm": 0.185724216553934,
+      "learning_rate": 8.331124949021279e-07,
+      "loss": 0.2583,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8407258064516129,
+      "grad_norm": 0.18254445448952297,
+      "learning_rate": 8.329121694420969e-07,
+      "loss": 0.2915,
+      "step": 1668
+    },
+    {
+      "epoch": 0.8412298387096774,
+      "grad_norm": 0.18038012693100844,
+      "learning_rate": 8.327117479389672e-07,
+      "loss": 0.2792,
+      "step": 1669
+    },
+    {
+      "epoch": 0.8417338709677419,
+      "grad_norm": 0.19274233034208005,
+      "learning_rate": 8.325112304505592e-07,
+      "loss": 0.2805,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8422379032258065,
+      "grad_norm": 0.17681044961531925,
+      "learning_rate": 8.323106170347212e-07,
+      "loss": 0.2723,
+      "step": 1671
+    },
+    {
+      "epoch": 0.842741935483871,
+      "grad_norm": 0.17251043391808404,
+      "learning_rate": 8.321099077493285e-07,
+      "loss": 0.2696,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8432459677419355,
+      "grad_norm": 0.1752390712831652,
+      "learning_rate": 8.319091026522848e-07,
+      "loss": 0.2738,
+      "step": 1673
+    },
+    {
+      "epoch": 0.84375,
+      "grad_norm": 0.17935779324663959,
+      "learning_rate": 8.317082018015211e-07,
+      "loss": 0.2862,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8442540322580645,
+      "grad_norm": 0.19723380525297224,
+      "learning_rate": 8.315072052549961e-07,
+      "loss": 0.2961,
+      "step": 1675
+    },
+    {
+      "epoch": 0.844758064516129,
+      "grad_norm": 0.17594438243367191,
+      "learning_rate": 8.313061130706959e-07,
+      "loss": 0.2838,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8452620967741935,
+      "grad_norm": 0.18502761220597608,
+      "learning_rate": 8.311049253066344e-07,
+      "loss": 0.2904,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8457661290322581,
+      "grad_norm": 0.1735632363360221,
+      "learning_rate": 8.30903642020853e-07,
+      "loss": 0.286,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8462701612903226,
+      "grad_norm": 0.1685121845779261,
+      "learning_rate": 8.307022632714208e-07,
+      "loss": 0.266,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8467741935483871,
+      "grad_norm": 0.17378234473772378,
+      "learning_rate": 8.305007891164341e-07,
+      "loss": 0.2904,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8472782258064516,
+      "grad_norm": 0.2032632255195518,
+      "learning_rate": 8.302992196140173e-07,
+      "loss": 0.2895,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8477822580645161,
+      "grad_norm": 0.17748247042521456,
+      "learning_rate": 8.300975548223214e-07,
+      "loss": 0.3003,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8482862903225806,
+      "grad_norm": 0.195328665507149,
+      "learning_rate": 8.298957947995261e-07,
+      "loss": 0.2832,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8487903225806451,
+      "grad_norm": 0.17552608040636258,
+      "learning_rate": 8.296939396038375e-07,
+      "loss": 0.2866,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8492943548387096,
+      "grad_norm": 0.17824578366133345,
+      "learning_rate": 8.294919892934896e-07,
+      "loss": 0.2779,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8497983870967742,
+      "grad_norm": 0.17486942680363315,
+      "learning_rate": 8.29289943926744e-07,
+      "loss": 0.2617,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8503024193548387,
+      "grad_norm": 0.18914845225754362,
+      "learning_rate": 8.290878035618893e-07,
+      "loss": 0.2812,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8508064516129032,
+      "grad_norm": 0.18115901123985698,
+      "learning_rate": 8.28885568257242e-07,
+      "loss": 0.3004,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8513104838709677,
+      "grad_norm": 0.17449979315444422,
+      "learning_rate": 8.286832380711454e-07,
+      "loss": 0.273,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8518145161290323,
+      "grad_norm": 0.1782814890164981,
+      "learning_rate": 8.284808130619708e-07,
+      "loss": 0.2997,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8523185483870968,
+      "grad_norm": 0.1710342855678077,
+      "learning_rate": 8.282782932881165e-07,
+      "loss": 0.2864,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8528225806451613,
+      "grad_norm": 0.18551314368513988,
+      "learning_rate": 8.280756788080081e-07,
+      "loss": 0.2847,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8533266129032258,
+      "grad_norm": 0.17793439535374442,
+      "learning_rate": 8.278729696800988e-07,
+      "loss": 0.3006,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8538306451612904,
+      "grad_norm": 0.17754655120950547,
+      "learning_rate": 8.276701659628686e-07,
+      "loss": 0.2889,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8543346774193549,
+      "grad_norm": 0.18489194070233667,
+      "learning_rate": 8.274672677148256e-07,
+      "loss": 0.2725,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8548387096774194,
+      "grad_norm": 0.1967447154512761,
+      "learning_rate": 8.272642749945042e-07,
+      "loss": 0.2857,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8553427419354839,
+      "grad_norm": 0.17696581032696806,
+      "learning_rate": 8.270611878604669e-07,
+      "loss": 0.2793,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8558467741935484,
+      "grad_norm": 0.17041142623174294,
+      "learning_rate": 8.268580063713028e-07,
+      "loss": 0.2708,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8563508064516129,
+      "grad_norm": 0.17554540593528697,
+      "learning_rate": 8.266547305856288e-07,
+      "loss": 0.2708,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8568548387096774,
+      "grad_norm": 0.17308164296880016,
+      "learning_rate": 8.264513605620884e-07,
+      "loss": 0.2738,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8573588709677419,
+      "grad_norm": 0.2285975263899735,
+      "learning_rate": 8.262478963593529e-07,
+      "loss": 0.2732,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8578629032258065,
+      "grad_norm": 0.1765909866610657,
+      "learning_rate": 8.260443380361201e-07,
+      "loss": 0.2822,
+      "step": 1702
+    },
+    {
+      "epoch": 0.858366935483871,
+      "grad_norm": 0.18612762071005345,
+      "learning_rate": 8.258406856511157e-07,
+      "loss": 0.2897,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8588709677419355,
+      "grad_norm": 0.18274802390819692,
+      "learning_rate": 8.256369392630918e-07,
+      "loss": 0.2826,
+      "step": 1704
+    },
+    {
+      "epoch": 0.859375,
+      "grad_norm": 0.1702610425035792,
+      "learning_rate": 8.254330989308283e-07,
+      "loss": 0.2825,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8598790322580645,
+      "grad_norm": 0.18746425872965825,
+      "learning_rate": 8.252291647131315e-07,
+      "loss": 0.2831,
+      "step": 1706
+    },
+    {
+      "epoch": 0.860383064516129,
+      "grad_norm": 0.1729312918615105,
+      "learning_rate": 8.250251366688357e-07,
+      "loss": 0.2788,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8608870967741935,
+      "grad_norm": 0.17310685779871965,
+      "learning_rate": 8.248210148568011e-07,
+      "loss": 0.2807,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8613911290322581,
+      "grad_norm": 0.17388995344225858,
+      "learning_rate": 8.246167993359159e-07,
+      "loss": 0.278,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8618951612903226,
+      "grad_norm": 0.17561470000841797,
+      "learning_rate": 8.244124901650951e-07,
+      "loss": 0.2852,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8623991935483871,
+      "grad_norm": 0.19059405969768753,
+      "learning_rate": 8.242080874032804e-07,
+      "loss": 0.2811,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8629032258064516,
+      "grad_norm": 0.17640465354683116,
+      "learning_rate": 8.24003591109441e-07,
+      "loss": 0.2837,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8634072580645161,
+      "grad_norm": 0.1946696258918474,
+      "learning_rate": 8.237990013425725e-07,
+      "loss": 0.2943,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8639112903225806,
+      "grad_norm": 0.17143548711069195,
+      "learning_rate": 8.23594318161698e-07,
+      "loss": 0.2844,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8644153225806451,
+      "grad_norm": 0.170912525533573,
+      "learning_rate": 8.233895416258673e-07,
+      "loss": 0.2828,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8649193548387096,
+      "grad_norm": 0.17030957212565428,
+      "learning_rate": 8.231846717941572e-07,
+      "loss": 0.2841,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8654233870967742,
+      "grad_norm": 0.18030711489803564,
+      "learning_rate": 8.229797087256711e-07,
+      "loss": 0.2892,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8659274193548387,
+      "grad_norm": 0.1811276763324713,
+      "learning_rate": 8.2277465247954e-07,
+      "loss": 0.2833,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8664314516129032,
+      "grad_norm": 0.18181075140547173,
+      "learning_rate": 8.22569503114921e-07,
+      "loss": 0.2936,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8669354838709677,
+      "grad_norm": 0.17610807113952007,
+      "learning_rate": 8.223642606909986e-07,
+      "loss": 0.2918,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8674395161290323,
+      "grad_norm": 0.17498207358667517,
+      "learning_rate": 8.221589252669841e-07,
+      "loss": 0.2706,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8679435483870968,
+      "grad_norm": 0.18462559185235636,
+      "learning_rate": 8.219534969021151e-07,
+      "loss": 0.2862,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8684475806451613,
+      "grad_norm": 0.17613631479568562,
+      "learning_rate": 8.217479756556567e-07,
+      "loss": 0.2901,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8689516129032258,
+      "grad_norm": 0.17518479858622174,
+      "learning_rate": 8.215423615869005e-07,
+      "loss": 0.2611,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8694556451612904,
+      "grad_norm": 0.17109997839484933,
+      "learning_rate": 8.213366547551648e-07,
+      "loss": 0.2817,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8699596774193549,
+      "grad_norm": 0.1966008607228567,
+      "learning_rate": 8.211308552197948e-07,
+      "loss": 0.2792,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8704637096774194,
+      "grad_norm": 0.17565135968705492,
+      "learning_rate": 8.209249630401623e-07,
+      "loss": 0.2854,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8709677419354839,
+      "grad_norm": 0.17137189614479656,
+      "learning_rate": 8.207189782756661e-07,
+      "loss": 0.2662,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8714717741935484,
+      "grad_norm": 0.17297331900482513,
+      "learning_rate": 8.205129009857312e-07,
+      "loss": 0.2799,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8719758064516129,
+      "grad_norm": 0.1724815696106094,
+      "learning_rate": 8.203067312298101e-07,
+      "loss": 0.2815,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8724798387096774,
+      "grad_norm": 0.18096897924623875,
+      "learning_rate": 8.20100469067381e-07,
+      "loss": 0.2837,
+      "step": 1731
+    },
+    {
+      "epoch": 0.8729838709677419,
+      "grad_norm": 0.19781239041691703,
+      "learning_rate": 8.198941145579496e-07,
+      "loss": 0.2939,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8734879032258065,
+      "grad_norm": 0.17163633709375878,
+      "learning_rate": 8.196876677610479e-07,
+      "loss": 0.266,
+      "step": 1733
+    },
+    {
+      "epoch": 0.873991935483871,
+      "grad_norm": 0.18825663758858197,
+      "learning_rate": 8.194811287362344e-07,
+      "loss": 0.291,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8744959677419355,
+      "grad_norm": 0.19151890851189823,
+      "learning_rate": 8.192744975430941e-07,
+      "loss": 0.2887,
+      "step": 1735
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.17804405341187254,
+      "learning_rate": 8.190677742412393e-07,
+      "loss": 0.2848,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8755040322580645,
+      "grad_norm": 0.17441120813931166,
+      "learning_rate": 8.188609588903081e-07,
+      "loss": 0.297,
+      "step": 1737
+    },
+    {
+      "epoch": 0.876008064516129,
+      "grad_norm": 0.17132241278155927,
+      "learning_rate": 8.186540515499653e-07,
+      "loss": 0.2698,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8765120967741935,
+      "grad_norm": 0.17642561108901925,
+      "learning_rate": 8.184470522799029e-07,
+      "loss": 0.2908,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8770161290322581,
+      "grad_norm": 0.231128422621628,
+      "learning_rate": 8.182399611398385e-07,
+      "loss": 0.275,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8775201612903226,
+      "grad_norm": 0.17983640539436302,
+      "learning_rate": 8.180327781895166e-07,
+      "loss": 0.2746,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8780241935483871,
+      "grad_norm": 0.18298601240785914,
+      "learning_rate": 8.178255034887083e-07,
+      "loss": 0.2933,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8785282258064516,
+      "grad_norm": 0.17727125106526648,
+      "learning_rate": 8.176181370972112e-07,
+      "loss": 0.3026,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8790322580645161,
+      "grad_norm": 0.17900225615415002,
+      "learning_rate": 8.174106790748489e-07,
+      "loss": 0.2732,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8795362903225806,
+      "grad_norm": 0.16825501654368868,
+      "learning_rate": 8.172031294814721e-07,
+      "loss": 0.2648,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8800403225806451,
+      "grad_norm": 0.17838543584835861,
+      "learning_rate": 8.169954883769573e-07,
+      "loss": 0.2826,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8805443548387096,
+      "grad_norm": 0.20624261385412906,
+      "learning_rate": 8.16787755821208e-07,
+      "loss": 0.2931,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8810483870967742,
+      "grad_norm": 0.19622027217406832,
+      "learning_rate": 8.165799318741533e-07,
+      "loss": 0.2855,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8815524193548387,
+      "grad_norm": 0.1795090027517633,
+      "learning_rate": 8.163720165957494e-07,
+      "loss": 0.2855,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8820564516129032,
+      "grad_norm": 0.1727005576579451,
+      "learning_rate": 8.161640100459785e-07,
+      "loss": 0.2945,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8825604838709677,
+      "grad_norm": 0.17361486955355637,
+      "learning_rate": 8.159559122848494e-07,
+      "loss": 0.2845,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8830645161290323,
+      "grad_norm": 0.1844805830827498,
+      "learning_rate": 8.157477233723969e-07,
+      "loss": 0.298,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8835685483870968,
+      "grad_norm": 0.17263438233102688,
+      "learning_rate": 8.15539443368682e-07,
+      "loss": 0.2929,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8840725806451613,
+      "grad_norm": 0.18475595233458791,
+      "learning_rate": 8.153310723337923e-07,
+      "loss": 0.2844,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8845766129032258,
+      "grad_norm": 0.1730273189248541,
+      "learning_rate": 8.151226103278417e-07,
+      "loss": 0.284,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8850806451612904,
+      "grad_norm": 0.18104372418355474,
+      "learning_rate": 8.149140574109701e-07,
+      "loss": 0.2897,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8855846774193549,
+      "grad_norm": 0.17839123584579997,
+      "learning_rate": 8.147054136433437e-07,
+      "loss": 0.2908,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8860887096774194,
+      "grad_norm": 0.178362220369913,
+      "learning_rate": 8.144966790851551e-07,
+      "loss": 0.2724,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8865927419354839,
+      "grad_norm": 0.18349648424878212,
+      "learning_rate": 8.142878537966225e-07,
+      "loss": 0.2813,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8870967741935484,
+      "grad_norm": 0.1852606758136101,
+      "learning_rate": 8.140789378379911e-07,
+      "loss": 0.2705,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8876008064516129,
+      "grad_norm": 0.19163597573225657,
+      "learning_rate": 8.138699312695318e-07,
+      "loss": 0.2763,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8881048387096774,
+      "grad_norm": 0.16967793891115396,
+      "learning_rate": 8.136608341515417e-07,
+      "loss": 0.2878,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8886088709677419,
+      "grad_norm": 0.17409231137199876,
+      "learning_rate": 8.13451646544344e-07,
+      "loss": 0.2824,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8891129032258065,
+      "grad_norm": 0.1718892011569422,
+      "learning_rate": 8.132423685082879e-07,
+      "loss": 0.2819,
+      "step": 1764
+    },
+    {
+      "epoch": 0.889616935483871,
+      "grad_norm": 0.1768902921291496,
+      "learning_rate": 8.130330001037489e-07,
+      "loss": 0.2866,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8901209677419355,
+      "grad_norm": 0.17727162868649207,
+      "learning_rate": 8.128235413911286e-07,
+      "loss": 0.2904,
+      "step": 1766
+    },
+    {
+      "epoch": 0.890625,
+      "grad_norm": 0.18000850290405332,
+      "learning_rate": 8.126139924308544e-07,
+      "loss": 0.2821,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8911290322580645,
+      "grad_norm": 0.1950805957693296,
+      "learning_rate": 8.124043532833799e-07,
+      "loss": 0.2724,
+      "step": 1768
+    },
+    {
+      "epoch": 0.891633064516129,
+      "grad_norm": 0.16953136358361712,
+      "learning_rate": 8.121946240091847e-07,
+      "loss": 0.2897,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8921370967741935,
+      "grad_norm": 0.17382780411451854,
+      "learning_rate": 8.119848046687745e-07,
+      "loss": 0.2743,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8926411290322581,
+      "grad_norm": 0.17618015127985004,
+      "learning_rate": 8.117748953226807e-07,
+      "loss": 0.293,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8931451612903226,
+      "grad_norm": 0.19101986869600812,
+      "learning_rate": 8.115648960314609e-07,
+      "loss": 0.282,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8936491935483871,
+      "grad_norm": 0.18335641858591029,
+      "learning_rate": 8.113548068556989e-07,
+      "loss": 0.2782,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8941532258064516,
+      "grad_norm": 0.17305922736379498,
+      "learning_rate": 8.111446278560037e-07,
+      "loss": 0.2819,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8946572580645161,
+      "grad_norm": 0.17289190092892376,
+      "learning_rate": 8.109343590930107e-07,
+      "loss": 0.2722,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8951612903225806,
+      "grad_norm": 0.20419317136274495,
+      "learning_rate": 8.107240006273815e-07,
+      "loss": 0.2984,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8956653225806451,
+      "grad_norm": 0.17574416151651231,
+      "learning_rate": 8.105135525198026e-07,
+      "loss": 0.2785,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8961693548387096,
+      "grad_norm": 0.18197635365780418,
+      "learning_rate": 8.103030148309876e-07,
+      "loss": 0.2671,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8966733870967742,
+      "grad_norm": 0.182868231288254,
+      "learning_rate": 8.10092387621675e-07,
+      "loss": 0.2825,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8971774193548387,
+      "grad_norm": 0.1753018528864952,
+      "learning_rate": 8.098816709526293e-07,
+      "loss": 0.2742,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8976814516129032,
+      "grad_norm": 0.17996946202759295,
+      "learning_rate": 8.096708648846416e-07,
+      "loss": 0.2721,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8981854838709677,
+      "grad_norm": 0.17250723947592084,
+      "learning_rate": 8.094599694785272e-07,
+      "loss": 0.2857,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8986895161290323,
+      "grad_norm": 0.17857618093982208,
+      "learning_rate": 8.092489847951288e-07,
+      "loss": 0.289,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8991935483870968,
+      "grad_norm": 0.180980362554131,
+      "learning_rate": 8.09037910895314e-07,
+      "loss": 0.2815,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8996975806451613,
+      "grad_norm": 0.19044748994727145,
+      "learning_rate": 8.088267478399761e-07,
+      "loss": 0.2763,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9002016129032258,
+      "grad_norm": 0.18502618756219494,
+      "learning_rate": 8.086154956900348e-07,
+      "loss": 0.2777,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9007056451612904,
+      "grad_norm": 0.17795283297213882,
+      "learning_rate": 8.084041545064347e-07,
+      "loss": 0.277,
+      "step": 1787
+    },
+    {
+      "epoch": 0.9012096774193549,
+      "grad_norm": 0.18095952785719716,
+      "learning_rate": 8.081927243501465e-07,
+      "loss": 0.2849,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9017137096774194,
+      "grad_norm": 0.20352287553838194,
+      "learning_rate": 8.079812052821665e-07,
+      "loss": 0.268,
+      "step": 1789
+    },
+    {
+      "epoch": 0.9022177419354839,
+      "grad_norm": 0.1817677931102643,
+      "learning_rate": 8.077695973635165e-07,
+      "loss": 0.2821,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9027217741935484,
+      "grad_norm": 0.20806765911305017,
+      "learning_rate": 8.075579006552442e-07,
+      "loss": 0.2802,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9027217741935484,
+      "eval_loss": 0.31333088874816895,
+      "eval_runtime": 17.8076,
+      "eval_samples_per_second": 48.013,
+      "eval_steps_per_second": 1.011,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 0.18964086276352513,
+      "learning_rate": 8.073461152184229e-07,
+      "loss": 0.2955,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9037298387096774,
+      "grad_norm": 0.20100101509553,
+      "learning_rate": 8.071342411141511e-07,
+      "loss": 0.2833,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9042338709677419,
+      "grad_norm": 0.16556895445429864,
+      "learning_rate": 8.069222784035536e-07,
+      "loss": 0.2572,
+      "step": 1794
+    },
+    {
+      "epoch": 0.9047379032258065,
+      "grad_norm": 0.18585984965180008,
+      "learning_rate": 8.067102271477798e-07,
+      "loss": 0.292,
+      "step": 1795
+    },
+    {
+      "epoch": 0.905241935483871,
+      "grad_norm": 0.1714046882678454,
+      "learning_rate": 8.064980874080056e-07,
+      "loss": 0.2783,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9057459677419355,
+      "grad_norm": 0.1845849829174842,
+      "learning_rate": 8.062858592454318e-07,
+      "loss": 0.2719,
+      "step": 1797
+    },
+    {
+      "epoch": 0.90625,
+      "grad_norm": 0.18224225892643142,
+      "learning_rate": 8.060735427212848e-07,
+      "loss": 0.2812,
+      "step": 1798
+    },
+    {
+      "epoch": 0.9067540322580645,
+      "grad_norm": 0.1694582941399971,
+      "learning_rate": 8.05861137896817e-07,
+      "loss": 0.269,
+      "step": 1799
+    },
+    {
+      "epoch": 0.907258064516129,
+      "grad_norm": 0.1866565288452891,
+      "learning_rate": 8.056486448333053e-07,
+      "loss": 0.2867,
+      "step": 1800
+    },
+    {
+      "epoch": 0.9077620967741935,
+      "grad_norm": 0.18228705266437853,
+      "learning_rate": 8.054360635920532e-07,
+      "loss": 0.2919,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9082661290322581,
+      "grad_norm": 0.17847533993145076,
+      "learning_rate": 8.052233942343889e-07,
+      "loss": 0.2864,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9087701612903226,
+      "grad_norm": 0.16890284755735133,
+      "learning_rate": 8.050106368216661e-07,
+      "loss": 0.2748,
+      "step": 1803
+    },
+    {
+      "epoch": 0.9092741935483871,
+      "grad_norm": 0.17414946004467158,
+      "learning_rate": 8.047977914152639e-07,
+      "loss": 0.2886,
+      "step": 1804
+    },
+    {
+      "epoch": 0.9097782258064516,
+      "grad_norm": 0.17442792515844072,
+      "learning_rate": 8.045848580765869e-07,
+      "loss": 0.2926,
+      "step": 1805
+    },
+    {
+      "epoch": 0.9102822580645161,
+      "grad_norm": 0.17085485937836853,
+      "learning_rate": 8.043718368670654e-07,
+      "loss": 0.2847,
+      "step": 1806
+    },
+    {
+      "epoch": 0.9107862903225806,
+      "grad_norm": 0.17339005862232762,
+      "learning_rate": 8.041587278481541e-07,
+      "loss": 0.2729,
+      "step": 1807
+    },
+    {
+      "epoch": 0.9112903225806451,
+      "grad_norm": 0.17518880582664062,
+      "learning_rate": 8.039455310813343e-07,
+      "loss": 0.3047,
+      "step": 1808
+    },
+    {
+      "epoch": 0.9117943548387096,
+      "grad_norm": 0.17160540781556277,
+      "learning_rate": 8.037322466281116e-07,
+      "loss": 0.277,
+      "step": 1809
+    },
+    {
+      "epoch": 0.9122983870967742,
+      "grad_norm": 0.18984801005727484,
+      "learning_rate": 8.035188745500171e-07,
+      "loss": 0.2975,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9128024193548387,
+      "grad_norm": 0.18036724606771928,
+      "learning_rate": 8.033054149086076e-07,
+      "loss": 0.2738,
+      "step": 1811
+    },
+    {
+      "epoch": 0.9133064516129032,
+      "grad_norm": 0.1694811997928256,
+      "learning_rate": 8.030918677654648e-07,
+      "loss": 0.2802,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9138104838709677,
+      "grad_norm": 1.122599384666042,
+      "learning_rate": 8.028782331821956e-07,
+      "loss": 0.2729,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9143145161290323,
+      "grad_norm": 0.19587893697493347,
+      "learning_rate": 8.026645112204325e-07,
+      "loss": 0.2844,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9148185483870968,
+      "grad_norm": 0.18843188442999503,
+      "learning_rate": 8.024507019418327e-07,
+      "loss": 0.2622,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9153225806451613,
+      "grad_norm": 0.2412193144484893,
+      "learning_rate": 8.022368054080789e-07,
+      "loss": 0.2864,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9158266129032258,
+      "grad_norm": 0.2982039518983422,
+      "learning_rate": 8.020228216808792e-07,
+      "loss": 0.2827,
+      "step": 1817
+    },
+    {
+      "epoch": 0.9163306451612904,
+      "grad_norm": 0.24012806458676572,
+      "learning_rate": 8.018087508219664e-07,
+      "loss": 0.2749,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9168346774193549,
+      "grad_norm": 0.20105639329407363,
+      "learning_rate": 8.015945928930985e-07,
+      "loss": 0.2751,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9173387096774194,
+      "grad_norm": 0.17599257831809287,
+      "learning_rate": 8.013803479560588e-07,
+      "loss": 0.2726,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9178427419354839,
+      "grad_norm": 0.17545331694711577,
+      "learning_rate": 8.011660160726556e-07,
+      "loss": 0.2783,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9183467741935484,
+      "grad_norm": 0.1782601795128222,
+      "learning_rate": 8.009515973047225e-07,
+      "loss": 0.2721,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9188508064516129,
+      "grad_norm": 0.1927011349013866,
+      "learning_rate": 8.007370917141177e-07,
+      "loss": 0.2762,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9193548387096774,
+      "grad_norm": 0.1920937971165849,
+      "learning_rate": 8.005224993627251e-07,
+      "loss": 0.2704,
+      "step": 1824
+    },
+    {
+      "epoch": 0.9198588709677419,
+      "grad_norm": 0.1793904274021918,
+      "learning_rate": 8.003078203124532e-07,
+      "loss": 0.2906,
+      "step": 1825
+    },
+    {
+      "epoch": 0.9203629032258065,
+      "grad_norm": 0.1816291362076379,
+      "learning_rate": 8.000930546252351e-07,
+      "loss": 0.2691,
+      "step": 1826
+    },
+    {
+      "epoch": 0.920866935483871,
+      "grad_norm": 0.1783523495764457,
+      "learning_rate": 7.998782023630299e-07,
+      "loss": 0.2823,
+      "step": 1827
+    },
+    {
+      "epoch": 0.9213709677419355,
+      "grad_norm": 0.20855882950783844,
+      "learning_rate": 7.996632635878209e-07,
+      "loss": 0.2804,
+      "step": 1828
+    },
+    {
+      "epoch": 0.921875,
+      "grad_norm": 0.17492678074423923,
+      "learning_rate": 7.994482383616168e-07,
+      "loss": 0.2832,
+      "step": 1829
+    },
+    {
+      "epoch": 0.9223790322580645,
+      "grad_norm": 0.17234593489453404,
+      "learning_rate": 7.992331267464509e-07,
+      "loss": 0.2768,
+      "step": 1830
+    },
+    {
+      "epoch": 0.922883064516129,
+      "grad_norm": 0.17200804609809464,
+      "learning_rate": 7.990179288043815e-07,
+      "loss": 0.2677,
+      "step": 1831
+    },
+    {
+      "epoch": 0.9233870967741935,
+      "grad_norm": 0.17572805845486925,
+      "learning_rate": 7.98802644597492e-07,
+      "loss": 0.2821,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9238911290322581,
+      "grad_norm": 0.20644587156717698,
+      "learning_rate": 7.985872741878905e-07,
+      "loss": 0.2732,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9243951612903226,
+      "grad_norm": 0.17874642664132656,
+      "learning_rate": 7.983718176377101e-07,
+      "loss": 0.269,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9248991935483871,
+      "grad_norm": 0.17456990653485246,
+      "learning_rate": 7.981562750091085e-07,
+      "loss": 0.2745,
+      "step": 1835
+    },
+    {
+      "epoch": 0.9254032258064516,
+      "grad_norm": 0.16939889686969853,
+      "learning_rate": 7.979406463642686e-07,
+      "loss": 0.2916,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9259072580645161,
+      "grad_norm": 0.17330331884417968,
+      "learning_rate": 7.977249317653979e-07,
+      "loss": 0.2897,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9264112903225806,
+      "grad_norm": 0.17625952471758344,
+      "learning_rate": 7.975091312747286e-07,
+      "loss": 0.28,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9269153225806451,
+      "grad_norm": 0.18477999685889357,
+      "learning_rate": 7.97293244954518e-07,
+      "loss": 0.3018,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9274193548387096,
+      "grad_norm": 0.17443677971673763,
+      "learning_rate": 7.970772728670479e-07,
+      "loss": 0.2701,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9279233870967742,
+      "grad_norm": 0.17476215415812357,
+      "learning_rate": 7.968612150746247e-07,
+      "loss": 0.2917,
+      "step": 1841
+    },
+    {
+      "epoch": 0.9284274193548387,
+      "grad_norm": 0.172486146139852,
+      "learning_rate": 7.966450716395801e-07,
+      "loss": 0.2802,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9289314516129032,
+      "grad_norm": 0.17784609076169455,
+      "learning_rate": 7.9642884262427e-07,
+      "loss": 0.2795,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9294354838709677,
+      "grad_norm": 0.1837707194984302,
+      "learning_rate": 7.96212528091075e-07,
+      "loss": 0.2899,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9299395161290323,
+      "grad_norm": 0.18094587392424352,
+      "learning_rate": 7.959961281024004e-07,
+      "loss": 0.2939,
+      "step": 1845
+    },
+    {
+      "epoch": 0.9304435483870968,
+      "grad_norm": 0.18248542798816517,
+      "learning_rate": 7.957796427206766e-07,
+      "loss": 0.2926,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9309475806451613,
+      "grad_norm": 0.17614486195538318,
+      "learning_rate": 7.955630720083581e-07,
+      "loss": 0.2912,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9314516129032258,
+      "grad_norm": 0.1726471918494145,
+      "learning_rate": 7.953464160279244e-07,
+      "loss": 0.2891,
+      "step": 1848
+    },
+    {
+      "epoch": 0.9319556451612904,
+      "grad_norm": 0.1774055068005992,
+      "learning_rate": 7.951296748418789e-07,
+      "loss": 0.3031,
+      "step": 1849
+    },
+    {
+      "epoch": 0.9324596774193549,
+      "grad_norm": 0.18664971286954127,
+      "learning_rate": 7.949128485127508e-07,
+      "loss": 0.2798,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9329637096774194,
+      "grad_norm": 0.21309616857185437,
+      "learning_rate": 7.946959371030926e-07,
+      "loss": 0.2852,
+      "step": 1851
+    },
+    {
+      "epoch": 0.9334677419354839,
+      "grad_norm": 0.17453472429034075,
+      "learning_rate": 7.944789406754821e-07,
+      "loss": 0.2796,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9339717741935484,
+      "grad_norm": 0.17686343311528496,
+      "learning_rate": 7.942618592925214e-07,
+      "loss": 0.2821,
+      "step": 1853
+    },
+    {
+      "epoch": 0.9344758064516129,
+      "grad_norm": 0.2050361536049733,
+      "learning_rate": 7.940446930168372e-07,
+      "loss": 0.2855,
+      "step": 1854
+    },
+    {
+      "epoch": 0.9349798387096774,
+      "grad_norm": 0.19293005978824818,
+      "learning_rate": 7.938274419110806e-07,
+      "loss": 0.2853,
+      "step": 1855
+    },
+    {
+      "epoch": 0.9354838709677419,
+      "grad_norm": 0.18447151377144042,
+      "learning_rate": 7.936101060379272e-07,
+      "loss": 0.2909,
+      "step": 1856
+    },
+    {
+      "epoch": 0.9359879032258065,
+      "grad_norm": 0.17295084942405378,
+      "learning_rate": 7.93392685460077e-07,
+      "loss": 0.2828,
+      "step": 1857
+    },
+    {
+      "epoch": 0.936491935483871,
+      "grad_norm": 0.1695374241026228,
+      "learning_rate": 7.931751802402544e-07,
+      "loss": 0.2799,
+      "step": 1858
+    },
+    {
+      "epoch": 0.9369959677419355,
+      "grad_norm": 0.17220089424019935,
+      "learning_rate": 7.929575904412086e-07,
+      "loss": 0.2783,
+      "step": 1859
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.17583666760835148,
+      "learning_rate": 7.927399161257127e-07,
+      "loss": 0.268,
+      "step": 1860
+    },
+    {
+      "epoch": 0.9380040322580645,
+      "grad_norm": 0.181882166877518,
+      "learning_rate": 7.925221573565644e-07,
+      "loss": 0.2931,
+      "step": 1861
+    },
+    {
+      "epoch": 0.938508064516129,
+      "grad_norm": 0.17175675035279075,
+      "learning_rate": 7.923043141965857e-07,
+      "loss": 0.2763,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9390120967741935,
+      "grad_norm": 0.16867999230536798,
+      "learning_rate": 7.920863867086232e-07,
+      "loss": 0.2764,
+      "step": 1863
+    },
+    {
+      "epoch": 0.9395161290322581,
+      "grad_norm": 0.18179217734847186,
+      "learning_rate": 7.918683749555473e-07,
+      "loss": 0.2865,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9400201612903226,
+      "grad_norm": 0.17793466089999202,
+      "learning_rate": 7.916502790002535e-07,
+      "loss": 0.2933,
+      "step": 1865
+    },
+    {
+      "epoch": 0.9405241935483871,
+      "grad_norm": 0.1687244667253048,
+      "learning_rate": 7.914320989056608e-07,
+      "loss": 0.2646,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9410282258064516,
+      "grad_norm": 0.16924880345437254,
+      "learning_rate": 7.912138347347128e-07,
+      "loss": 0.2747,
+      "step": 1867
+    },
+    {
+      "epoch": 0.9415322580645161,
+      "grad_norm": 0.1998764610080855,
+      "learning_rate": 7.909954865503776e-07,
+      "loss": 0.2848,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9420362903225806,
+      "grad_norm": 0.17645688052721015,
+      "learning_rate": 7.907770544156471e-07,
+      "loss": 0.2744,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9425403225806451,
+      "grad_norm": 0.17444186260064237,
+      "learning_rate": 7.905585383935377e-07,
+      "loss": 0.2861,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9430443548387096,
+      "grad_norm": 0.18655327778687014,
+      "learning_rate": 7.903399385470898e-07,
+      "loss": 0.2723,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9435483870967742,
+      "grad_norm": 0.17432236510965454,
+      "learning_rate": 7.901212549393682e-07,
+      "loss": 0.2782,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9440524193548387,
+      "grad_norm": 0.1866368609884871,
+      "learning_rate": 7.899024876334619e-07,
+      "loss": 0.27,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9445564516129032,
+      "grad_norm": 0.1867440259653203,
+      "learning_rate": 7.896836366924836e-07,
+      "loss": 0.2861,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9450604838709677,
+      "grad_norm": 0.17504198342034763,
+      "learning_rate": 7.894647021795707e-07,
+      "loss": 0.288,
+      "step": 1875
+    },
+    {
+      "epoch": 0.9455645161290323,
+      "grad_norm": 0.17474732068264098,
+      "learning_rate": 7.892456841578843e-07,
+      "loss": 0.2839,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9460685483870968,
+      "grad_norm": 0.17433782582765533,
+      "learning_rate": 7.890265826906097e-07,
+      "loss": 0.2695,
+      "step": 1877
+    },
+    {
+      "epoch": 0.9465725806451613,
+      "grad_norm": 0.1732874776440476,
+      "learning_rate": 7.888073978409568e-07,
+      "loss": 0.2894,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9470766129032258,
+      "grad_norm": 0.1829899116719353,
+      "learning_rate": 7.885881296721584e-07,
+      "loss": 0.2943,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9475806451612904,
+      "grad_norm": 0.17001971535664087,
+      "learning_rate": 7.883687782474723e-07,
+      "loss": 0.2856,
+      "step": 1880
+    },
+    {
+      "epoch": 0.9480846774193549,
+      "grad_norm": 0.1780345466436045,
+      "learning_rate": 7.8814934363018e-07,
+      "loss": 0.2746,
+      "step": 1881
+    },
+    {
+      "epoch": 0.9485887096774194,
+      "grad_norm": 0.1811602910506968,
+      "learning_rate": 7.879298258835872e-07,
+      "loss": 0.2776,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9490927419354839,
+      "grad_norm": 0.18771497800701312,
+      "learning_rate": 7.877102250710231e-07,
+      "loss": 0.2829,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9495967741935484,
+      "grad_norm": 0.18472560622513845,
+      "learning_rate": 7.874905412558415e-07,
+      "loss": 0.2707,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9501008064516129,
+      "grad_norm": 0.1795065532871576,
+      "learning_rate": 7.872707745014195e-07,
+      "loss": 0.2812,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9506048387096774,
+      "grad_norm": 0.19450850100162909,
+      "learning_rate": 7.870509248711588e-07,
+      "loss": 0.262,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9511088709677419,
+      "grad_norm": 0.17189805933616106,
+      "learning_rate": 7.868309924284842e-07,
+      "loss": 0.2681,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9516129032258065,
+      "grad_norm": 0.17073814905474735,
+      "learning_rate": 7.866109772368453e-07,
+      "loss": 0.2774,
+      "step": 1888
+    },
+    {
+      "epoch": 0.952116935483871,
+      "grad_norm": 0.17066105523996877,
+      "learning_rate": 7.863908793597149e-07,
+      "loss": 0.2883,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9526209677419355,
+      "grad_norm": 0.16966954884907165,
+      "learning_rate": 7.861706988605898e-07,
+      "loss": 0.2832,
+      "step": 1890
+    },
+    {
+      "epoch": 0.953125,
+      "grad_norm": 0.17731653359539812,
+      "learning_rate": 7.859504358029909e-07,
+      "loss": 0.2856,
+      "step": 1891
+    },
+    {
+      "epoch": 0.9536290322580645,
+      "grad_norm": 0.18810146382045853,
+      "learning_rate": 7.857300902504628e-07,
+      "loss": 0.2853,
+      "step": 1892
+    },
+    {
+      "epoch": 0.954133064516129,
+      "grad_norm": 0.18445022791672494,
+      "learning_rate": 7.855096622665735e-07,
+      "loss": 0.2867,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9546370967741935,
+      "grad_norm": 0.1728990132935549,
+      "learning_rate": 7.852891519149152e-07,
+      "loss": 0.2874,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9551411290322581,
+      "grad_norm": 0.1761006592616753,
+      "learning_rate": 7.85068559259104e-07,
+      "loss": 0.2844,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9556451612903226,
+      "grad_norm": 0.183998556068685,
+      "learning_rate": 7.848478843627792e-07,
+      "loss": 0.2705,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9561491935483871,
+      "grad_norm": 0.17318304016376618,
+      "learning_rate": 7.846271272896044e-07,
+      "loss": 0.2778,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9566532258064516,
+      "grad_norm": 0.1824137043313419,
+      "learning_rate": 7.844062881032664e-07,
+      "loss": 0.2872,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9571572580645161,
+      "grad_norm": 0.17344844058434108,
+      "learning_rate": 7.841853668674763e-07,
+      "loss": 0.2787,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9576612903225806,
+      "grad_norm": 0.1758545032107461,
+      "learning_rate": 7.839643636459683e-07,
+      "loss": 0.2857,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9581653225806451,
+      "grad_norm": 0.1729782764700269,
+      "learning_rate": 7.837432785025004e-07,
+      "loss": 0.2765,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9586693548387096,
+      "grad_norm": 0.17898116556336083,
+      "learning_rate": 7.835221115008542e-07,
+      "loss": 0.2877,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9591733870967742,
+      "grad_norm": 0.1895613129053808,
+      "learning_rate": 7.833008627048352e-07,
+      "loss": 0.2754,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9596774193548387,
+      "grad_norm": 0.20979437224680517,
+      "learning_rate": 7.830795321782724e-07,
+      "loss": 0.2773,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9601814516129032,
+      "grad_norm": 0.17134279973430921,
+      "learning_rate": 7.828581199850182e-07,
+      "loss": 0.2785,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9606854838709677,
+      "grad_norm": 0.1703859949701216,
+      "learning_rate": 7.826366261889483e-07,
+      "loss": 0.2755,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9611895161290323,
+      "grad_norm": 0.1708391359023031,
+      "learning_rate": 7.824150508539628e-07,
+      "loss": 0.2784,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9616935483870968,
+      "grad_norm": 0.17468338877763465,
+      "learning_rate": 7.821933940439847e-07,
+      "loss": 0.2584,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9621975806451613,
+      "grad_norm": 0.17669530382160473,
+      "learning_rate": 7.819716558229604e-07,
+      "loss": 0.2594,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9627016129032258,
+      "grad_norm": 0.18182770484400235,
+      "learning_rate": 7.8174983625486e-07,
+      "loss": 0.2929,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9632056451612904,
+      "grad_norm": 0.1718420918323365,
+      "learning_rate": 7.815279354036772e-07,
+      "loss": 0.2741,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9637096774193549,
+      "grad_norm": 0.18047954008297915,
+      "learning_rate": 7.813059533334292e-07,
+      "loss": 0.282,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9642137096774194,
+      "grad_norm": 0.1708373327019462,
+      "learning_rate": 7.810838901081561e-07,
+      "loss": 0.2671,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9647177419354839,
+      "grad_norm": 0.17530947113755785,
+      "learning_rate": 7.80861745791922e-07,
+      "loss": 0.2863,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9652217741935484,
+      "grad_norm": 0.1856980391062388,
+      "learning_rate": 7.80639520448814e-07,
+      "loss": 0.2953,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9657258064516129,
+      "grad_norm": 0.17104956670285965,
+      "learning_rate": 7.80417214142943e-07,
+      "loss": 0.2929,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9662298387096774,
+      "grad_norm": 0.17968017354027993,
+      "learning_rate": 7.801948269384427e-07,
+      "loss": 0.2805,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9667338709677419,
+      "grad_norm": 0.16885090448355405,
+      "learning_rate": 7.799723588994706e-07,
+      "loss": 0.275,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9672379032258065,
+      "grad_norm": 0.16977668143159869,
+      "learning_rate": 7.797498100902071e-07,
+      "loss": 0.2955,
+      "step": 1919
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 0.17543252330569692,
+      "learning_rate": 7.795271805748565e-07,
+      "loss": 0.285,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9682459677419355,
+      "grad_norm": 0.1779214329877026,
+      "learning_rate": 7.793044704176459e-07,
+      "loss": 0.2751,
+      "step": 1921
+    },
+    {
+      "epoch": 0.96875,
+      "grad_norm": 0.1727984897805244,
+      "learning_rate": 7.790816796828259e-07,
+      "loss": 0.2917,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9692540322580645,
+      "grad_norm": 0.19093500891641543,
+      "learning_rate": 7.788588084346699e-07,
+      "loss": 0.2912,
+      "step": 1923
+    },
+    {
+      "epoch": 0.969758064516129,
+      "grad_norm": 0.1877331833500236,
+      "learning_rate": 7.786358567374752e-07,
+      "loss": 0.2828,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9702620967741935,
+      "grad_norm": 0.1964520978166042,
+      "learning_rate": 7.784128246555619e-07,
+      "loss": 0.2878,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9707661290322581,
+      "grad_norm": 0.1793857792176289,
+      "learning_rate": 7.781897122532732e-07,
+      "loss": 0.3009,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9712701612903226,
+      "grad_norm": 0.1778515329690811,
+      "learning_rate": 7.779665195949761e-07,
+      "loss": 0.2686,
+      "step": 1927
+    },
+    {
+      "epoch": 0.9717741935483871,
+      "grad_norm": 0.18899510212928142,
+      "learning_rate": 7.777432467450598e-07,
+      "loss": 0.2944,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9722782258064516,
+      "grad_norm": 0.18103789111426294,
+      "learning_rate": 7.775198937679375e-07,
+      "loss": 0.2853,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9727822580645161,
+      "grad_norm": 0.17187773087886613,
+      "learning_rate": 7.772964607280448e-07,
+      "loss": 0.2853,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9732862903225806,
+      "grad_norm": 0.17830119933021793,
+      "learning_rate": 7.77072947689841e-07,
+      "loss": 0.2883,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9737903225806451,
+      "grad_norm": 0.18468034488609253,
+      "learning_rate": 7.768493547178083e-07,
+      "loss": 0.2783,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9742943548387096,
+      "grad_norm": 0.1649149359989447,
+      "learning_rate": 7.766256818764517e-07,
+      "loss": 0.2712,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9747983870967742,
+      "grad_norm": 0.18006035200393466,
+      "learning_rate": 7.764019292302994e-07,
+      "loss": 0.2873,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9753024193548387,
+      "grad_norm": 0.18387324362971386,
+      "learning_rate": 7.761780968439027e-07,
+      "loss": 0.2825,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9758064516129032,
+      "grad_norm": 0.1756603042845968,
+      "learning_rate": 7.759541847818361e-07,
+      "loss": 0.2633,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9763104838709677,
+      "grad_norm": 0.17483974536625063,
+      "learning_rate": 7.757301931086963e-07,
+      "loss": 0.2982,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9768145161290323,
+      "grad_norm": 0.18502806181928816,
+      "learning_rate": 7.755061218891041e-07,
+      "loss": 0.2836,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9773185483870968,
+      "grad_norm": 0.17424427520526933,
+      "learning_rate": 7.752819711877024e-07,
+      "loss": 0.2909,
+      "step": 1939
+    },
+    {
+      "epoch": 0.9778225806451613,
+      "grad_norm": 0.17479328408301795,
+      "learning_rate": 7.750577410691572e-07,
+      "loss": 0.2892,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9783266129032258,
+      "grad_norm": 0.18236213050678932,
+      "learning_rate": 7.748334315981577e-07,
+      "loss": 0.2904,
+      "step": 1941
+    },
+    {
+      "epoch": 0.9788306451612904,
+      "grad_norm": 0.17005102876538494,
+      "learning_rate": 7.746090428394156e-07,
+      "loss": 0.2784,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9793346774193549,
+      "grad_norm": 0.17329730021771092,
+      "learning_rate": 7.743845748576659e-07,
+      "loss": 0.2836,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9798387096774194,
+      "grad_norm": 0.1714850364043798,
+      "learning_rate": 7.741600277176659e-07,
+      "loss": 0.2857,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9803427419354839,
+      "grad_norm": 0.19242374904544585,
+      "learning_rate": 7.739354014841963e-07,
+      "loss": 0.286,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9808467741935484,
+      "grad_norm": 0.17505647013555173,
+      "learning_rate": 7.737106962220603e-07,
+      "loss": 0.2961,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9813508064516129,
+      "grad_norm": 0.17558706687337872,
+      "learning_rate": 7.734859119960841e-07,
+      "loss": 0.2735,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9818548387096774,
+      "grad_norm": 0.17903519468817597,
+      "learning_rate": 7.732610488711162e-07,
+      "loss": 0.2901,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9823588709677419,
+      "grad_norm": 0.18371964890476014,
+      "learning_rate": 7.730361069120286e-07,
+      "loss": 0.2985,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9828629032258065,
+      "grad_norm": 0.1723762496803238,
+      "learning_rate": 7.728110861837156e-07,
+      "loss": 0.2837,
+      "step": 1950
+    },
+    {
+      "epoch": 0.983366935483871,
+      "grad_norm": 0.1790011706688029,
+      "learning_rate": 7.725859867510942e-07,
+      "loss": 0.2696,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9838709677419355,
+      "grad_norm": 0.18850089072217402,
+      "learning_rate": 7.72360808679104e-07,
+      "loss": 0.2811,
+      "step": 1952
+    },
+    {
+      "epoch": 0.984375,
+      "grad_norm": 0.17379961506239888,
+      "learning_rate": 7.72135552032708e-07,
+      "loss": 0.2743,
+      "step": 1953
+    },
+    {
+      "epoch": 0.9848790322580645,
+      "grad_norm": 0.18355475523994572,
+      "learning_rate": 7.719102168768907e-07,
+      "loss": 0.2918,
+      "step": 1954
+    },
+    {
+      "epoch": 0.985383064516129,
+      "grad_norm": 0.18238534494389397,
+      "learning_rate": 7.716848032766605e-07,
+      "loss": 0.2972,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9858870967741935,
+      "grad_norm": 0.1729414375179342,
+      "learning_rate": 7.714593112970473e-07,
+      "loss": 0.2761,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9863911290322581,
+      "grad_norm": 0.1768407372110884,
+      "learning_rate": 7.712337410031046e-07,
+      "loss": 0.2663,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9868951612903226,
+      "grad_norm": 0.1803839421907579,
+      "learning_rate": 7.710080924599077e-07,
+      "loss": 0.2913,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9873991935483871,
+      "grad_norm": 0.1722830670020056,
+      "learning_rate": 7.707823657325549e-07,
+      "loss": 0.2899,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9879032258064516,
+      "grad_norm": 0.17157007007009292,
+      "learning_rate": 7.705565608861673e-07,
+      "loss": 0.2832,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9884072580645161,
+      "grad_norm": 0.17179339985585704,
+      "learning_rate": 7.703306779858875e-07,
+      "loss": 0.275,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9889112903225806,
+      "grad_norm": 0.1750481355495183,
+      "learning_rate": 7.701047170968819e-07,
+      "loss": 0.2935,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9894153225806451,
+      "grad_norm": 0.1703164887136099,
+      "learning_rate": 7.698786782843386e-07,
+      "loss": 0.2578,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9899193548387096,
+      "grad_norm": 0.17963472103153666,
+      "learning_rate": 7.696525616134686e-07,
+      "loss": 0.2885,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9904233870967742,
+      "grad_norm": 0.17443507400988176,
+      "learning_rate": 7.694263671495047e-07,
+      "loss": 0.2992,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9909274193548387,
+      "grad_norm": 0.18131986198353908,
+      "learning_rate": 7.692000949577031e-07,
+      "loss": 0.2897,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9914314516129032,
+      "grad_norm": 0.19050009588750086,
+      "learning_rate": 7.689737451033415e-07,
+      "loss": 0.2975,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9919354838709677,
+      "grad_norm": 0.21986786250372692,
+      "learning_rate": 7.687473176517209e-07,
+      "loss": 0.2775,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9924395161290323,
+      "grad_norm": 0.1712446456152378,
+      "learning_rate": 7.685208126681637e-07,
+      "loss": 0.2653,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9929435483870968,
+      "grad_norm": 0.17416424744387057,
+      "learning_rate": 7.682942302180155e-07,
+      "loss": 0.3078,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9934475806451613,
+      "grad_norm": 0.17288742218707365,
+      "learning_rate": 7.680675703666439e-07,
+      "loss": 0.278,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9939516129032258,
+      "grad_norm": 0.1795604518756118,
+      "learning_rate": 7.67840833179439e-07,
+      "loss": 0.2849,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9944556451612904,
+      "grad_norm": 0.1715422047295615,
+      "learning_rate": 7.676140187218128e-07,
+      "loss": 0.2665,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9949596774193549,
+      "grad_norm": 0.17763829951279275,
+      "learning_rate": 7.673871270592e-07,
+      "loss": 0.2907,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9954637096774194,
+      "grad_norm": 0.1748484464046483,
+      "learning_rate": 7.671601582570573e-07,
+      "loss": 0.2722,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9959677419354839,
+      "grad_norm": 0.18807863525366347,
+      "learning_rate": 7.66933112380864e-07,
+      "loss": 0.2966,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9964717741935484,
+      "grad_norm": 0.19625503483008158,
+      "learning_rate": 7.667059894961214e-07,
+      "loss": 0.2688,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9969758064516129,
+      "grad_norm": 0.1714858168146041,
+      "learning_rate": 7.664787896683528e-07,
+      "loss": 0.2857,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9974798387096774,
+      "grad_norm": 0.17049746359373963,
+      "learning_rate": 7.662515129631045e-07,
+      "loss": 0.2732,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9979838709677419,
+      "grad_norm": 0.17253458678680952,
+      "learning_rate": 7.660241594459437e-07,
+      "loss": 0.2802,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9984879032258065,
+      "grad_norm": 0.17152385183060628,
+      "learning_rate": 7.65796729182461e-07,
+      "loss": 0.2775,
+      "step": 1981
+    },
+    {
+      "epoch": 0.998991935483871,
+      "grad_norm": 0.17962880378962143,
+      "learning_rate": 7.655692222382683e-07,
+      "loss": 0.282,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9994959677419355,
+      "grad_norm": 0.1815884076274797,
+      "learning_rate": 7.653416386790003e-07,
+      "loss": 0.2807,
+      "step": 1983
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.17036014949141842,
+      "learning_rate": 7.651139785703131e-07,
+      "loss": 0.2756,
+      "step": 1984
+    },
+    {
+      "epoch": 1.0005040322580645,
+      "grad_norm": 0.1858739335123262,
+      "learning_rate": 7.648862419778854e-07,
+      "loss": 0.2754,
+      "step": 1985
+    },
+    {
+      "epoch": 1.001008064516129,
+      "grad_norm": 0.17713241635811486,
+      "learning_rate": 7.646584289674178e-07,
+      "loss": 0.2913,
+      "step": 1986
+    },
+    {
+      "epoch": 1.0015120967741935,
+      "grad_norm": 0.17224329918547132,
+      "learning_rate": 7.644305396046328e-07,
+      "loss": 0.2838,
+      "step": 1987
+    },
+    {
+      "epoch": 1.002016129032258,
+      "grad_norm": 0.16815918206812863,
+      "learning_rate": 7.642025739552753e-07,
+      "loss": 0.2738,
+      "step": 1988
+    },
+    {
+      "epoch": 1.0003780718336484,
+      "grad_norm": 0.17810881094257833,
+      "learning_rate": 7.639745320851118e-07,
+      "loss": 0.2829,
+      "step": 1989
+    },
+    {
+      "epoch": 1.0008821676118462,
+      "grad_norm": 0.17225692284996372,
+      "learning_rate": 7.637464140599312e-07,
+      "loss": 0.2648,
+      "step": 1990
+    },
+    {
+      "epoch": 1.0008821676118462,
+      "eval_loss": 0.31192973256111145,
+      "eval_runtime": 16.7125,
+      "eval_samples_per_second": 51.159,
+      "eval_steps_per_second": 1.077,
+      "step": 1990
+    },
+    {
+      "epoch": 1.001386263390044,
+      "grad_norm": 0.18131745361437934,
+      "learning_rate": 7.635182199455437e-07,
+      "loss": 0.2904,
+      "step": 1991
+    },
+    {
+      "epoch": 1.001890359168242,
+      "grad_norm": 0.1674300111635892,
+      "learning_rate": 7.632899498077824e-07,
+      "loss": 0.2781,
+      "step": 1992
+    },
+    {
+      "epoch": 1.0023944549464399,
+      "grad_norm": 0.16930495343661658,
+      "learning_rate": 7.630616037125015e-07,
+      "loss": 0.2716,
+      "step": 1993
+    },
+    {
+      "epoch": 1.0028985507246377,
+      "grad_norm": 0.17149835744787098,
+      "learning_rate": 7.628331817255775e-07,
+      "loss": 0.3012,
+      "step": 1994
+    },
+    {
+      "epoch": 1.0034026465028356,
+      "grad_norm": 0.17694237469391477,
+      "learning_rate": 7.626046839129087e-07,
+      "loss": 0.2826,
+      "step": 1995
+    },
+    {
+      "epoch": 1.0039067422810335,
+      "grad_norm": 0.18584553067561851,
+      "learning_rate": 7.623761103404154e-07,
+      "loss": 0.2834,
+      "step": 1996
+    },
+    {
+      "epoch": 1.0044108380592311,
+      "grad_norm": 0.17723779816231616,
+      "learning_rate": 7.621474610740396e-07,
+      "loss": 0.2794,
+      "step": 1997
+    },
+    {
+      "epoch": 1.004914933837429,
+      "grad_norm": 0.18721280669165485,
+      "learning_rate": 7.619187361797451e-07,
+      "loss": 0.2813,
+      "step": 1998
+    },
+    {
+      "epoch": 1.005419029615627,
+      "grad_norm": 0.1752600172196804,
+      "learning_rate": 7.616899357235178e-07,
+      "loss": 0.2636,
+      "step": 1999
+    },
+    {
+      "epoch": 1.0059231253938248,
+      "grad_norm": 0.1855969151006619,
+      "learning_rate": 7.614610597713651e-07,
+      "loss": 0.2923,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0064272211720227,
+      "grad_norm": 0.16729057622113636,
+      "learning_rate": 7.612321083893163e-07,
+      "loss": 0.2892,
+      "step": 2001
+    },
+    {
+      "epoch": 1.0069313169502205,
+      "grad_norm": 0.1693767296113182,
+      "learning_rate": 7.610030816434224e-07,
+      "loss": 0.2801,
+      "step": 2002
+    },
+    {
+      "epoch": 1.0074354127284184,
+      "grad_norm": 0.17521037625410535,
+      "learning_rate": 7.607739795997563e-07,
+      "loss": 0.2725,
+      "step": 2003
+    },
+    {
+      "epoch": 1.0079395085066163,
+      "grad_norm": 0.17979436463239115,
+      "learning_rate": 7.605448023244127e-07,
+      "loss": 0.2771,
+      "step": 2004
+    },
+    {
+      "epoch": 1.0084436042848142,
+      "grad_norm": 0.17931260879453892,
+      "learning_rate": 7.603155498835075e-07,
+      "loss": 0.2668,
+      "step": 2005
+    },
+    {
+      "epoch": 1.008947700063012,
+      "grad_norm": 0.17071051877009574,
+      "learning_rate": 7.600862223431787e-07,
+      "loss": 0.2828,
+      "step": 2006
+    },
+    {
+      "epoch": 1.00945179584121,
+      "grad_norm": 0.17196704916567862,
+      "learning_rate": 7.598568197695858e-07,
+      "loss": 0.2794,
+      "step": 2007
+    },
+    {
+      "epoch": 1.0099558916194078,
+      "grad_norm": 0.17562030132329703,
+      "learning_rate": 7.596273422289103e-07,
+      "loss": 0.2833,
+      "step": 2008
+    },
+    {
+      "epoch": 1.0104599873976055,
+      "grad_norm": 0.18089872134544313,
+      "learning_rate": 7.593977897873548e-07,
+      "loss": 0.2833,
+      "step": 2009
+    },
+    {
+      "epoch": 1.0109640831758033,
+      "grad_norm": 0.18222407962137802,
+      "learning_rate": 7.591681625111439e-07,
+      "loss": 0.2816,
+      "step": 2010
+    },
+    {
+      "epoch": 1.0114681789540012,
+      "grad_norm": 0.18109669021912,
+      "learning_rate": 7.589384604665235e-07,
+      "loss": 0.288,
+      "step": 2011
+    },
+    {
+      "epoch": 1.011972274732199,
+      "grad_norm": 0.17814912652297168,
+      "learning_rate": 7.587086837197614e-07,
+      "loss": 0.2753,
+      "step": 2012
+    },
+    {
+      "epoch": 1.012476370510397,
+      "grad_norm": 0.16991715478737968,
+      "learning_rate": 7.584788323371466e-07,
+      "loss": 0.2855,
+      "step": 2013
+    },
+    {
+      "epoch": 1.0129804662885948,
+      "grad_norm": 0.17035192594812326,
+      "learning_rate": 7.582489063849899e-07,
+      "loss": 0.2629,
+      "step": 2014
+    },
+    {
+      "epoch": 1.0134845620667927,
+      "grad_norm": 0.17513209286858666,
+      "learning_rate": 7.580189059296234e-07,
+      "loss": 0.2772,
+      "step": 2015
+    },
+    {
+      "epoch": 1.0139886578449906,
+      "grad_norm": 0.19229502375720714,
+      "learning_rate": 7.577888310374007e-07,
+      "loss": 0.2764,
+      "step": 2016
+    },
+    {
+      "epoch": 1.0144927536231885,
+      "grad_norm": 0.16703432210037336,
+      "learning_rate": 7.575586817746975e-07,
+      "loss": 0.2749,
+      "step": 2017
+    },
+    {
+      "epoch": 1.0149968494013863,
+      "grad_norm": 0.17186499930225924,
+      "learning_rate": 7.573284582079098e-07,
+      "loss": 0.2696,
+      "step": 2018
+    },
+    {
+      "epoch": 1.0155009451795842,
+      "grad_norm": 0.20803141423115293,
+      "learning_rate": 7.570981604034563e-07,
+      "loss": 0.274,
+      "step": 2019
+    },
+    {
+      "epoch": 1.0160050409577819,
+      "grad_norm": 0.17601420291337044,
+      "learning_rate": 7.56867788427776e-07,
+      "loss": 0.2908,
+      "step": 2020
+    },
+    {
+      "epoch": 1.0165091367359798,
+      "grad_norm": 0.18319232781478847,
+      "learning_rate": 7.566373423473299e-07,
+      "loss": 0.2923,
+      "step": 2021
+    },
+    {
+      "epoch": 1.0170132325141776,
+      "grad_norm": 0.17400976480499947,
+      "learning_rate": 7.564068222286004e-07,
+      "loss": 0.2808,
+      "step": 2022
+    },
+    {
+      "epoch": 1.0175173282923755,
+      "grad_norm": 0.17912905548016714,
+      "learning_rate": 7.56176228138091e-07,
+      "loss": 0.2865,
+      "step": 2023
+    },
+    {
+      "epoch": 1.0180214240705734,
+      "grad_norm": 0.1722722944490204,
+      "learning_rate": 7.559455601423266e-07,
+      "loss": 0.2741,
+      "step": 2024
+    },
+    {
+      "epoch": 1.0185255198487713,
+      "grad_norm": 0.16659736898121513,
+      "learning_rate": 7.557148183078539e-07,
+      "loss": 0.2821,
+      "step": 2025
+    },
+    {
+      "epoch": 1.0190296156269691,
+      "grad_norm": 0.17461443095442888,
+      "learning_rate": 7.554840027012399e-07,
+      "loss": 0.2764,
+      "step": 2026
+    },
+    {
+      "epoch": 1.019533711405167,
+      "grad_norm": 0.18391922179509013,
+      "learning_rate": 7.552531133890738e-07,
+      "loss": 0.295,
+      "step": 2027
+    },
+    {
+      "epoch": 1.020037807183365,
+      "grad_norm": 0.20322970192602707,
+      "learning_rate": 7.550221504379659e-07,
+      "loss": 0.2865,
+      "step": 2028
+    },
+    {
+      "epoch": 1.0205419029615628,
+      "grad_norm": 0.177400814926691,
+      "learning_rate": 7.547911139145472e-07,
+      "loss": 0.2911,
+      "step": 2029
+    },
+    {
+      "epoch": 1.0210459987397607,
+      "grad_norm": 0.1834139875378068,
+      "learning_rate": 7.545600038854705e-07,
+      "loss": 0.2795,
+      "step": 2030
+    },
+    {
+      "epoch": 1.0215500945179583,
+      "grad_norm": 0.17234090006004013,
+      "learning_rate": 7.543288204174096e-07,
+      "loss": 0.2812,
+      "step": 2031
+    },
+    {
+      "epoch": 1.0220541902961562,
+      "grad_norm": 0.19523445589007987,
+      "learning_rate": 7.540975635770595e-07,
+      "loss": 0.2634,
+      "step": 2032
+    },
+    {
+      "epoch": 1.022558286074354,
+      "grad_norm": 0.17425774393185609,
+      "learning_rate": 7.538662334311363e-07,
+      "loss": 0.2923,
+      "step": 2033
+    },
+    {
+      "epoch": 1.023062381852552,
+      "grad_norm": 0.18190325865887647,
+      "learning_rate": 7.536348300463775e-07,
+      "loss": 0.2763,
+      "step": 2034
+    },
+    {
+      "epoch": 1.0235664776307498,
+      "grad_norm": 0.17592758218306667,
+      "learning_rate": 7.534033534895415e-07,
+      "loss": 0.2869,
+      "step": 2035
+    },
+    {
+      "epoch": 1.0240705734089477,
+      "grad_norm": 0.17477622082253388,
+      "learning_rate": 7.531718038274076e-07,
+      "loss": 0.291,
+      "step": 2036
+    },
+    {
+      "epoch": 1.0245746691871456,
+      "grad_norm": 0.22007149638124626,
+      "learning_rate": 7.529401811267765e-07,
+      "loss": 0.2676,
+      "step": 2037
+    },
+    {
+      "epoch": 1.0250787649653434,
+      "grad_norm": 0.16688921868859855,
+      "learning_rate": 7.527084854544701e-07,
+      "loss": 0.267,
+      "step": 2038
+    },
+    {
+      "epoch": 1.0255828607435413,
+      "grad_norm": 0.1867065847855188,
+      "learning_rate": 7.524767168773311e-07,
+      "loss": 0.2694,
+      "step": 2039
+    },
+    {
+      "epoch": 1.0260869565217392,
+      "grad_norm": 0.205670590026773,
+      "learning_rate": 7.522448754622234e-07,
+      "loss": 0.2889,
+      "step": 2040
+    },
+    {
+      "epoch": 1.026591052299937,
+      "grad_norm": 0.17475031087338616,
+      "learning_rate": 7.520129612760318e-07,
+      "loss": 0.2697,
+      "step": 2041
+    },
+    {
+      "epoch": 1.0270951480781347,
+      "grad_norm": 0.18541787593421327,
+      "learning_rate": 7.517809743856618e-07,
+      "loss": 0.2774,
+      "step": 2042
+    },
+    {
+      "epoch": 1.0275992438563326,
+      "grad_norm": 0.17371102962725932,
+      "learning_rate": 7.515489148580405e-07,
+      "loss": 0.2594,
+      "step": 2043
+    },
+    {
+      "epoch": 1.0281033396345305,
+      "grad_norm": 0.17385943346589472,
+      "learning_rate": 7.513167827601154e-07,
+      "loss": 0.2675,
+      "step": 2044
+    },
+    {
+      "epoch": 1.0286074354127284,
+      "grad_norm": 0.17595128300347995,
+      "learning_rate": 7.510845781588554e-07,
+      "loss": 0.2793,
+      "step": 2045
+    },
+    {
+      "epoch": 1.0291115311909262,
+      "grad_norm": 0.179237071829832,
+      "learning_rate": 7.5085230112125e-07,
+      "loss": 0.2792,
+      "step": 2046
+    },
+    {
+      "epoch": 1.0296156269691241,
+      "grad_norm": 0.17451242012769375,
+      "learning_rate": 7.506199517143095e-07,
+      "loss": 0.276,
+      "step": 2047
+    },
+    {
+      "epoch": 1.030119722747322,
+      "grad_norm": 0.18158012272605775,
+      "learning_rate": 7.503875300050656e-07,
+      "loss": 0.2788,
+      "step": 2048
+    },
+    {
+      "epoch": 1.0306238185255199,
+      "grad_norm": 0.18636629727957202,
+      "learning_rate": 7.501550360605704e-07,
+      "loss": 0.293,
+      "step": 2049
+    },
+    {
+      "epoch": 1.0311279143037178,
+      "grad_norm": 0.17720408232848278,
+      "learning_rate": 7.499224699478969e-07,
+      "loss": 0.2813,
+      "step": 2050
+    },
+    {
+      "epoch": 1.0316320100819156,
+      "grad_norm": 0.1916262449849373,
+      "learning_rate": 7.496898317341389e-07,
+      "loss": 0.2677,
+      "step": 2051
+    },
+    {
+      "epoch": 1.0321361058601135,
+      "grad_norm": 0.17348752838152534,
+      "learning_rate": 7.494571214864113e-07,
+      "loss": 0.2751,
+      "step": 2052
+    },
+    {
+      "epoch": 1.0326402016383114,
+      "grad_norm": 0.19058244263360036,
+      "learning_rate": 7.492243392718493e-07,
+      "loss": 0.2722,
+      "step": 2053
+    },
+    {
+      "epoch": 1.033144297416509,
+      "grad_norm": 0.1890009158650643,
+      "learning_rate": 7.489914851576095e-07,
+      "loss": 0.2655,
+      "step": 2054
+    },
+    {
+      "epoch": 1.033648393194707,
+      "grad_norm": 0.172808831512677,
+      "learning_rate": 7.487585592108685e-07,
+      "loss": 0.2847,
+      "step": 2055
+    },
+    {
+      "epoch": 1.0341524889729048,
+      "grad_norm": 0.17546626349825928,
+      "learning_rate": 7.485255614988241e-07,
+      "loss": 0.2817,
+      "step": 2056
+    },
+    {
+      "epoch": 1.0346565847511027,
+      "grad_norm": 0.1775777311831324,
+      "learning_rate": 7.482924920886949e-07,
+      "loss": 0.2754,
+      "step": 2057
+    },
+    {
+      "epoch": 1.0351606805293005,
+      "grad_norm": 0.1749949082772431,
+      "learning_rate": 7.480593510477197e-07,
+      "loss": 0.2831,
+      "step": 2058
+    },
+    {
+      "epoch": 1.0356647763074984,
+      "grad_norm": 0.18632122267782072,
+      "learning_rate": 7.478261384431585e-07,
+      "loss": 0.2725,
+      "step": 2059
+    },
+    {
+      "epoch": 1.0361688720856963,
+      "grad_norm": 0.17774286945133272,
+      "learning_rate": 7.475928543422916e-07,
+      "loss": 0.2815,
+      "step": 2060
+    },
+    {
+      "epoch": 1.0366729678638942,
+      "grad_norm": 0.18490810909000924,
+      "learning_rate": 7.473594988124199e-07,
+      "loss": 0.2846,
+      "step": 2061
+    },
+    {
+      "epoch": 1.037177063642092,
+      "grad_norm": 0.1728261249927515,
+      "learning_rate": 7.471260719208649e-07,
+      "loss": 0.2689,
+      "step": 2062
+    },
+    {
+      "epoch": 1.03768115942029,
+      "grad_norm": 0.18211593138151266,
+      "learning_rate": 7.468925737349693e-07,
+      "loss": 0.2923,
+      "step": 2063
+    },
+    {
+      "epoch": 1.0381852551984878,
+      "grad_norm": 0.17343204712457802,
+      "learning_rate": 7.466590043220955e-07,
+      "loss": 0.2809,
+      "step": 2064
+    },
+    {
+      "epoch": 1.0386893509766855,
+      "grad_norm": 0.17100328567163497,
+      "learning_rate": 7.46425363749627e-07,
+      "loss": 0.2754,
+      "step": 2065
+    },
+    {
+      "epoch": 1.0391934467548833,
+      "grad_norm": 0.1818697379651732,
+      "learning_rate": 7.461916520849674e-07,
+      "loss": 0.2724,
+      "step": 2066
+    },
+    {
+      "epoch": 1.0396975425330812,
+      "grad_norm": 0.18227441948785947,
+      "learning_rate": 7.459578693955413e-07,
+      "loss": 0.2709,
+      "step": 2067
+    },
+    {
+      "epoch": 1.040201638311279,
+      "grad_norm": 0.1760139361145817,
+      "learning_rate": 7.457240157487935e-07,
+      "loss": 0.2806,
+      "step": 2068
+    },
+    {
+      "epoch": 1.040705734089477,
+      "grad_norm": 0.1795019228044322,
+      "learning_rate": 7.454900912121894e-07,
+      "loss": 0.2631,
+      "step": 2069
+    },
+    {
+      "epoch": 1.0412098298676749,
+      "grad_norm": 0.18888478288221203,
+      "learning_rate": 7.452560958532147e-07,
+      "loss": 0.2889,
+      "step": 2070
+    },
+    {
+      "epoch": 1.0417139256458727,
+      "grad_norm": 0.21616086911752216,
+      "learning_rate": 7.450220297393756e-07,
+      "loss": 0.2683,
+      "step": 2071
+    },
+    {
+      "epoch": 1.0422180214240706,
+      "grad_norm": 0.17939705286161006,
+      "learning_rate": 7.447878929381989e-07,
+      "loss": 0.293,
+      "step": 2072
+    },
+    {
+      "epoch": 1.0427221172022685,
+      "grad_norm": 0.1798424204037018,
+      "learning_rate": 7.445536855172313e-07,
+      "loss": 0.2706,
+      "step": 2073
+    },
+    {
+      "epoch": 1.0432262129804664,
+      "grad_norm": 0.17088142114120458,
+      "learning_rate": 7.443194075440405e-07,
+      "loss": 0.2819,
+      "step": 2074
+    },
+    {
+      "epoch": 1.0437303087586642,
+      "grad_norm": 0.17886842519012602,
+      "learning_rate": 7.44085059086214e-07,
+      "loss": 0.2893,
+      "step": 2075
+    },
+    {
+      "epoch": 1.0442344045368621,
+      "grad_norm": 0.1803195665228042,
+      "learning_rate": 7.4385064021136e-07,
+      "loss": 0.2581,
+      "step": 2076
+    },
+    {
+      "epoch": 1.0447385003150598,
+      "grad_norm": 0.21531931727822987,
+      "learning_rate": 7.436161509871069e-07,
+      "loss": 0.2852,
+      "step": 2077
+    },
+    {
+      "epoch": 1.0452425960932576,
+      "grad_norm": 0.16822216389718422,
+      "learning_rate": 7.433815914811033e-07,
+      "loss": 0.265,
+      "step": 2078
+    },
+    {
+      "epoch": 1.0457466918714555,
+      "grad_norm": 0.17669997866886272,
+      "learning_rate": 7.431469617610183e-07,
+      "loss": 0.272,
+      "step": 2079
+    },
+    {
+      "epoch": 1.0462507876496534,
+      "grad_norm": 0.17526090755685278,
+      "learning_rate": 7.429122618945409e-07,
+      "loss": 0.2741,
+      "step": 2080
+    },
+    {
+      "epoch": 1.0467548834278513,
+      "grad_norm": 0.17590646009524652,
+      "learning_rate": 7.426774919493808e-07,
+      "loss": 0.2825,
+      "step": 2081
+    },
+    {
+      "epoch": 1.0472589792060492,
+      "grad_norm": 0.172169218025432,
+      "learning_rate": 7.424426519932676e-07,
+      "loss": 0.2588,
+      "step": 2082
+    },
+    {
+      "epoch": 1.047763074984247,
+      "grad_norm": 0.1805765775734755,
+      "learning_rate": 7.422077420939511e-07,
+      "loss": 0.2686,
+      "step": 2083
+    },
+    {
+      "epoch": 1.048267170762445,
+      "grad_norm": 0.18419409997013503,
+      "learning_rate": 7.419727623192013e-07,
+      "loss": 0.273,
+      "step": 2084
+    },
+    {
+      "epoch": 1.0487712665406428,
+      "grad_norm": 0.18365937796388698,
+      "learning_rate": 7.417377127368087e-07,
+      "loss": 0.2884,
+      "step": 2085
+    },
+    {
+      "epoch": 1.0492753623188407,
+      "grad_norm": 0.1690690352871276,
+      "learning_rate": 7.415025934145836e-07,
+      "loss": 0.2711,
+      "step": 2086
+    },
+    {
+      "epoch": 1.0497794580970385,
+      "grad_norm": 0.17486780574416283,
+      "learning_rate": 7.412674044203561e-07,
+      "loss": 0.2664,
+      "step": 2087
+    },
+    {
+      "epoch": 1.0502835538752362,
+      "grad_norm": 0.1809124255952242,
+      "learning_rate": 7.410321458219771e-07,
+      "loss": 0.291,
+      "step": 2088
+    },
+    {
+      "epoch": 1.050787649653434,
+      "grad_norm": 0.17461075528139122,
+      "learning_rate": 7.407968176873169e-07,
+      "loss": 0.2829,
+      "step": 2089
+    },
+    {
+      "epoch": 1.051291745431632,
+      "grad_norm": 0.17457876173639864,
+      "learning_rate": 7.405614200842668e-07,
+      "loss": 0.2733,
+      "step": 2090
+    },
+    {
+      "epoch": 1.0517958412098298,
+      "grad_norm": 0.1729548254201044,
+      "learning_rate": 7.40325953080737e-07,
+      "loss": 0.2617,
+      "step": 2091
+    },
+    {
+      "epoch": 1.0522999369880277,
+      "grad_norm": 0.1731484045045451,
+      "learning_rate": 7.400904167446585e-07,
+      "loss": 0.2717,
+      "step": 2092
+    },
+    {
+      "epoch": 1.0528040327662256,
+      "grad_norm": 0.18214823178441017,
+      "learning_rate": 7.39854811143982e-07,
+      "loss": 0.2715,
+      "step": 2093
+    },
+    {
+      "epoch": 1.0533081285444235,
+      "grad_norm": 0.21466975211674327,
+      "learning_rate": 7.396191363466785e-07,
+      "loss": 0.2567,
+      "step": 2094
+    },
+    {
+      "epoch": 1.0538122243226213,
+      "grad_norm": 0.1830837150368865,
+      "learning_rate": 7.393833924207385e-07,
+      "loss": 0.2669,
+      "step": 2095
+    },
+    {
+      "epoch": 1.0543163201008192,
+      "grad_norm": 0.1772829383212441,
+      "learning_rate": 7.391475794341725e-07,
+      "loss": 0.2798,
+      "step": 2096
+    },
+    {
+      "epoch": 1.054820415879017,
+      "grad_norm": 0.17883144461072345,
+      "learning_rate": 7.389116974550114e-07,
+      "loss": 0.3041,
+      "step": 2097
+    },
+    {
+      "epoch": 1.055324511657215,
+      "grad_norm": 0.16977328934679697,
+      "learning_rate": 7.386757465513055e-07,
+      "loss": 0.2896,
+      "step": 2098
+    },
+    {
+      "epoch": 1.0558286074354126,
+      "grad_norm": 0.17447815375668077,
+      "learning_rate": 7.384397267911252e-07,
+      "loss": 0.2813,
+      "step": 2099
+    },
+    {
+      "epoch": 1.0563327032136105,
+      "grad_norm": 0.17549836356043094,
+      "learning_rate": 7.382036382425608e-07,
+      "loss": 0.2604,
+      "step": 2100
+    },
+    {
+      "epoch": 1.0568367989918084,
+      "grad_norm": 0.168631281997645,
+      "learning_rate": 7.379674809737226e-07,
+      "loss": 0.2595,
+      "step": 2101
+    },
+    {
+      "epoch": 1.0573408947700063,
+      "grad_norm": 0.1779153181273789,
+      "learning_rate": 7.377312550527399e-07,
+      "loss": 0.2769,
+      "step": 2102
+    },
+    {
+      "epoch": 1.0578449905482041,
+      "grad_norm": 0.17489911047054466,
+      "learning_rate": 7.37494960547763e-07,
+      "loss": 0.2638,
+      "step": 2103
+    },
+    {
+      "epoch": 1.058349086326402,
+      "grad_norm": 0.18710665444256072,
+      "learning_rate": 7.372585975269612e-07,
+      "loss": 0.2733,
+      "step": 2104
+    },
+    {
+      "epoch": 1.0588531821045999,
+      "grad_norm": 0.16631564764201226,
+      "learning_rate": 7.370221660585238e-07,
+      "loss": 0.2549,
+      "step": 2105
+    },
+    {
+      "epoch": 1.0593572778827978,
+      "grad_norm": 0.1763591739795769,
+      "learning_rate": 7.367856662106595e-07,
+      "loss": 0.2724,
+      "step": 2106
+    },
+    {
+      "epoch": 1.0598613736609956,
+      "grad_norm": 0.17005232527255398,
+      "learning_rate": 7.365490980515976e-07,
+      "loss": 0.294,
+      "step": 2107
+    },
+    {
+      "epoch": 1.0603654694391935,
+      "grad_norm": 0.18645746328153773,
+      "learning_rate": 7.36312461649586e-07,
+      "loss": 0.28,
+      "step": 2108
+    },
+    {
+      "epoch": 1.0608695652173914,
+      "grad_norm": 0.17883161921298946,
+      "learning_rate": 7.360757570728934e-07,
+      "loss": 0.2848,
+      "step": 2109
+    },
+    {
+      "epoch": 1.061373660995589,
+      "grad_norm": 0.16761850246628487,
+      "learning_rate": 7.358389843898071e-07,
+      "loss": 0.267,
+      "step": 2110
+    },
+    {
+      "epoch": 1.061877756773787,
+      "grad_norm": 0.16661816127263798,
+      "learning_rate": 7.356021436686347e-07,
+      "loss": 0.2669,
+      "step": 2111
+    },
+    {
+      "epoch": 1.0623818525519848,
+      "grad_norm": 0.17946012130007954,
+      "learning_rate": 7.353652349777033e-07,
+      "loss": 0.2973,
+      "step": 2112
+    },
+    {
+      "epoch": 1.0628859483301827,
+      "grad_norm": 0.18375956514983224,
+      "learning_rate": 7.351282583853597e-07,
+      "loss": 0.2679,
+      "step": 2113
+    },
+    {
+      "epoch": 1.0633900441083806,
+      "grad_norm": 0.17321969243454696,
+      "learning_rate": 7.348912139599701e-07,
+      "loss": 0.2699,
+      "step": 2114
+    },
+    {
+      "epoch": 1.0638941398865784,
+      "grad_norm": 0.17030523274157758,
+      "learning_rate": 7.346541017699204e-07,
+      "loss": 0.2752,
+      "step": 2115
+    },
+    {
+      "epoch": 1.0643982356647763,
+      "grad_norm": 0.17855378803648458,
+      "learning_rate": 7.344169218836161e-07,
+      "loss": 0.2634,
+      "step": 2116
+    },
+    {
+      "epoch": 1.0649023314429742,
+      "grad_norm": 0.17747877929878342,
+      "learning_rate": 7.341796743694817e-07,
+      "loss": 0.2796,
+      "step": 2117
+    },
+    {
+      "epoch": 1.065406427221172,
+      "grad_norm": 0.17074631177901406,
+      "learning_rate": 7.339423592959619e-07,
+      "loss": 0.2741,
+      "step": 2118
+    },
+    {
+      "epoch": 1.06591052299937,
+      "grad_norm": 0.16930821017014017,
+      "learning_rate": 7.337049767315207e-07,
+      "loss": 0.2669,
+      "step": 2119
+    },
+    {
+      "epoch": 1.0664146187775678,
+      "grad_norm": 0.17690264524984564,
+      "learning_rate": 7.334675267446415e-07,
+      "loss": 0.2725,
+      "step": 2120
+    },
+    {
+      "epoch": 1.0669187145557655,
+      "grad_norm": 0.17283246521488307,
+      "learning_rate": 7.33230009403827e-07,
+      "loss": 0.2883,
+      "step": 2121
+    },
+    {
+      "epoch": 1.0674228103339634,
+      "grad_norm": 0.1829853318315655,
+      "learning_rate": 7.329924247775997e-07,
+      "loss": 0.266,
+      "step": 2122
+    },
+    {
+      "epoch": 1.0679269061121612,
+      "grad_norm": 0.1685321993543116,
+      "learning_rate": 7.327547729345012e-07,
+      "loss": 0.2811,
+      "step": 2123
+    },
+    {
+      "epoch": 1.0684310018903591,
+      "grad_norm": 0.1849302770037306,
+      "learning_rate": 7.325170539430924e-07,
+      "loss": 0.2819,
+      "step": 2124
+    },
+    {
+      "epoch": 1.068935097668557,
+      "grad_norm": 0.1722656015767197,
+      "learning_rate": 7.32279267871954e-07,
+      "loss": 0.2689,
+      "step": 2125
+    },
+    {
+      "epoch": 1.0694391934467549,
+      "grad_norm": 0.1724271053593593,
+      "learning_rate": 7.320414147896857e-07,
+      "loss": 0.2667,
+      "step": 2126
+    },
+    {
+      "epoch": 1.0699432892249527,
+      "grad_norm": 0.17269435432280894,
+      "learning_rate": 7.318034947649064e-07,
+      "loss": 0.2799,
+      "step": 2127
+    },
+    {
+      "epoch": 1.0704473850031506,
+      "grad_norm": 0.17637659253320054,
+      "learning_rate": 7.31565507866255e-07,
+      "loss": 0.2916,
+      "step": 2128
+    },
+    {
+      "epoch": 1.0709514807813485,
+      "grad_norm": 0.1773899855628958,
+      "learning_rate": 7.313274541623891e-07,
+      "loss": 0.2817,
+      "step": 2129
+    },
+    {
+      "epoch": 1.0714555765595464,
+      "grad_norm": 0.18431910974801158,
+      "learning_rate": 7.310893337219857e-07,
+      "loss": 0.2909,
+      "step": 2130
+    },
+    {
+      "epoch": 1.0719596723377443,
+      "grad_norm": 0.16833157774674223,
+      "learning_rate": 7.30851146613741e-07,
+      "loss": 0.2765,
+      "step": 2131
+    },
+    {
+      "epoch": 1.0724637681159421,
+      "grad_norm": 0.17919301749031188,
+      "learning_rate": 7.306128929063705e-07,
+      "loss": 0.2882,
+      "step": 2132
+    },
+    {
+      "epoch": 1.0729678638941398,
+      "grad_norm": 0.17270532104058647,
+      "learning_rate": 7.303745726686091e-07,
+      "loss": 0.2781,
+      "step": 2133
+    },
+    {
+      "epoch": 1.0734719596723377,
+      "grad_norm": 0.1783614757197418,
+      "learning_rate": 7.301361859692103e-07,
+      "loss": 0.2754,
+      "step": 2134
+    },
+    {
+      "epoch": 1.0739760554505355,
+      "grad_norm": 0.17544697010356183,
+      "learning_rate": 7.298977328769476e-07,
+      "loss": 0.2854,
+      "step": 2135
+    },
+    {
+      "epoch": 1.0744801512287334,
+      "grad_norm": 0.18870456706038555,
+      "learning_rate": 7.296592134606133e-07,
+      "loss": 0.2766,
+      "step": 2136
+    },
+    {
+      "epoch": 1.0749842470069313,
+      "grad_norm": 0.18171100036742838,
+      "learning_rate": 7.294206277890185e-07,
+      "loss": 0.2744,
+      "step": 2137
+    },
+    {
+      "epoch": 1.0754883427851292,
+      "grad_norm": 0.1677025505094429,
+      "learning_rate": 7.291819759309936e-07,
+      "loss": 0.2647,
+      "step": 2138
+    },
+    {
+      "epoch": 1.075992438563327,
+      "grad_norm": 0.17288124268881677,
+      "learning_rate": 7.289432579553885e-07,
+      "loss": 0.2731,
+      "step": 2139
+    },
+    {
+      "epoch": 1.076496534341525,
+      "grad_norm": 0.1703425006119981,
+      "learning_rate": 7.287044739310717e-07,
+      "loss": 0.2795,
+      "step": 2140
+    },
+    {
+      "epoch": 1.0770006301197228,
+      "grad_norm": 0.20025547919726422,
+      "learning_rate": 7.284656239269308e-07,
+      "loss": 0.2877,
+      "step": 2141
+    },
+    {
+      "epoch": 1.0775047258979207,
+      "grad_norm": 0.17516136688479655,
+      "learning_rate": 7.282267080118727e-07,
+      "loss": 0.2929,
+      "step": 2142
+    },
+    {
+      "epoch": 1.0780088216761186,
+      "grad_norm": 0.17474524965964358,
+      "learning_rate": 7.279877262548232e-07,
+      "loss": 0.2768,
+      "step": 2143
+    },
+    {
+      "epoch": 1.0785129174543164,
+      "grad_norm": 0.17466518332420242,
+      "learning_rate": 7.277486787247269e-07,
+      "loss": 0.285,
+      "step": 2144
+    },
+    {
+      "epoch": 1.079017013232514,
+      "grad_norm": 0.1819043058750424,
+      "learning_rate": 7.275095654905477e-07,
+      "loss": 0.2952,
+      "step": 2145
+    },
+    {
+      "epoch": 1.079521109010712,
+      "grad_norm": 0.1737875425062245,
+      "learning_rate": 7.272703866212682e-07,
+      "loss": 0.2827,
+      "step": 2146
+    },
+    {
+      "epoch": 1.0800252047889098,
+      "grad_norm": 0.16958237624506534,
+      "learning_rate": 7.2703114218589e-07,
+      "loss": 0.268,
+      "step": 2147
+    },
+    {
+      "epoch": 1.0805293005671077,
+      "grad_norm": 0.17257636404154555,
+      "learning_rate": 7.267918322534336e-07,
+      "loss": 0.2971,
+      "step": 2148
+    },
+    {
+      "epoch": 1.0810333963453056,
+      "grad_norm": 0.16928675310322286,
+      "learning_rate": 7.265524568929386e-07,
+      "loss": 0.2736,
+      "step": 2149
+    },
+    {
+      "epoch": 1.0815374921235035,
+      "grad_norm": 0.1760838964224455,
+      "learning_rate": 7.263130161734632e-07,
+      "loss": 0.2849,
+      "step": 2150
+    },
+    {
+      "epoch": 1.0820415879017014,
+      "grad_norm": 0.18328166077952487,
+      "learning_rate": 7.260735101640845e-07,
+      "loss": 0.258,
+      "step": 2151
+    },
+    {
+      "epoch": 1.0825456836798992,
+      "grad_norm": 0.1739117009909244,
+      "learning_rate": 7.258339389338987e-07,
+      "loss": 0.2924,
+      "step": 2152
+    },
+    {
+      "epoch": 1.083049779458097,
+      "grad_norm": 0.1781115926870173,
+      "learning_rate": 7.255943025520203e-07,
+      "loss": 0.2596,
+      "step": 2153
+    },
+    {
+      "epoch": 1.083553875236295,
+      "grad_norm": 0.16844391904359102,
+      "learning_rate": 7.253546010875832e-07,
+      "loss": 0.2703,
+      "step": 2154
+    },
+    {
+      "epoch": 1.0840579710144929,
+      "grad_norm": 0.3278887015033608,
+      "learning_rate": 7.251148346097398e-07,
+      "loss": 0.2779,
+      "step": 2155
+    },
+    {
+      "epoch": 1.0845620667926905,
+      "grad_norm": 0.18990927841452848,
+      "learning_rate": 7.248750031876609e-07,
+      "loss": 0.2946,
+      "step": 2156
+    },
+    {
+      "epoch": 1.0850661625708884,
+      "grad_norm": 0.17850530711366283,
+      "learning_rate": 7.246351068905368e-07,
+      "loss": 0.2875,
+      "step": 2157
+    },
+    {
+      "epoch": 1.0855702583490863,
+      "grad_norm": 0.17728631017755883,
+      "learning_rate": 7.243951457875758e-07,
+      "loss": 0.2818,
+      "step": 2158
+    },
+    {
+      "epoch": 1.0860743541272841,
+      "grad_norm": 0.1774048928579683,
+      "learning_rate": 7.241551199480051e-07,
+      "loss": 0.3007,
+      "step": 2159
+    },
+    {
+      "epoch": 1.086578449905482,
+      "grad_norm": 0.17200567169289427,
+      "learning_rate": 7.239150294410712e-07,
+      "loss": 0.2717,
+      "step": 2160
+    },
+    {
+      "epoch": 1.08708254568368,
+      "grad_norm": 0.17241132084381322,
+      "learning_rate": 7.236748743360384e-07,
+      "loss": 0.2668,
+      "step": 2161
+    },
+    {
+      "epoch": 1.0875866414618778,
+      "grad_norm": 0.17166550486003698,
+      "learning_rate": 7.234346547021896e-07,
+      "loss": 0.2995,
+      "step": 2162
+    },
+    {
+      "epoch": 1.0880907372400757,
+      "grad_norm": 0.17146325451469707,
+      "learning_rate": 7.231943706088273e-07,
+      "loss": 0.2833,
+      "step": 2163
+    },
+    {
+      "epoch": 1.0885948330182735,
+      "grad_norm": 0.17437078445725077,
+      "learning_rate": 7.229540221252716e-07,
+      "loss": 0.2667,
+      "step": 2164
+    },
+    {
+      "epoch": 1.0890989287964714,
+      "grad_norm": 0.17975096501612886,
+      "learning_rate": 7.227136093208617e-07,
+      "loss": 0.2789,
+      "step": 2165
+    },
+    {
+      "epoch": 1.0896030245746693,
+      "grad_norm": 0.17177377044930098,
+      "learning_rate": 7.22473132264955e-07,
+      "loss": 0.2712,
+      "step": 2166
+    },
+    {
+      "epoch": 1.090107120352867,
+      "grad_norm": 0.16926098260953162,
+      "learning_rate": 7.22232591026928e-07,
+      "loss": 0.2655,
+      "step": 2167
+    },
+    {
+      "epoch": 1.0906112161310648,
+      "grad_norm": 0.1813817491695504,
+      "learning_rate": 7.21991985676175e-07,
+      "loss": 0.2782,
+      "step": 2168
+    },
+    {
+      "epoch": 1.0911153119092627,
+      "grad_norm": 0.1740637250383412,
+      "learning_rate": 7.217513162821094e-07,
+      "loss": 0.282,
+      "step": 2169
+    },
+    {
+      "epoch": 1.0916194076874606,
+      "grad_norm": 0.16909096484224836,
+      "learning_rate": 7.215105829141627e-07,
+      "loss": 0.2777,
+      "step": 2170
+    },
+    {
+      "epoch": 1.0921235034656585,
+      "grad_norm": 0.19169602165222183,
+      "learning_rate": 7.21269785641785e-07,
+      "loss": 0.2746,
+      "step": 2171
+    },
+    {
+      "epoch": 1.0926275992438563,
+      "grad_norm": 0.17606098979705256,
+      "learning_rate": 7.210289245344447e-07,
+      "loss": 0.2796,
+      "step": 2172
+    },
+    {
+      "epoch": 1.0931316950220542,
+      "grad_norm": 0.17081810769761613,
+      "learning_rate": 7.207879996616291e-07,
+      "loss": 0.272,
+      "step": 2173
+    },
+    {
+      "epoch": 1.093635790800252,
+      "grad_norm": 0.18390464264943393,
+      "learning_rate": 7.205470110928431e-07,
+      "loss": 0.2598,
+      "step": 2174
+    },
+    {
+      "epoch": 1.09413988657845,
+      "grad_norm": 0.1752617305053687,
+      "learning_rate": 7.203059588976107e-07,
+      "loss": 0.2702,
+      "step": 2175
+    },
+    {
+      "epoch": 1.0946439823566478,
+      "grad_norm": 0.17985065994012292,
+      "learning_rate": 7.20064843145474e-07,
+      "loss": 0.279,
+      "step": 2176
+    },
+    {
+      "epoch": 1.0951480781348457,
+      "grad_norm": 0.19788977536594046,
+      "learning_rate": 7.198236639059932e-07,
+      "loss": 0.2763,
+      "step": 2177
+    },
+    {
+      "epoch": 1.0956521739130434,
+      "grad_norm": 0.17078468603904942,
+      "learning_rate": 7.19582421248747e-07,
+      "loss": 0.2696,
+      "step": 2178
+    },
+    {
+      "epoch": 1.0961562696912412,
+      "grad_norm": 0.17485440247279033,
+      "learning_rate": 7.193411152433327e-07,
+      "loss": 0.2703,
+      "step": 2179
+    },
+    {
+      "epoch": 1.0966603654694391,
+      "grad_norm": 0.17440915689799694,
+      "learning_rate": 7.190997459593651e-07,
+      "loss": 0.272,
+      "step": 2180
+    },
+    {
+      "epoch": 1.097164461247637,
+      "grad_norm": 0.1798376956782012,
+      "learning_rate": 7.188583134664783e-07,
+      "loss": 0.2857,
+      "step": 2181
+    },
+    {
+      "epoch": 1.0976685570258349,
+      "grad_norm": 0.17843642267892243,
+      "learning_rate": 7.186168178343239e-07,
+      "loss": 0.3012,
+      "step": 2182
+    },
+    {
+      "epoch": 1.0981726528040328,
+      "grad_norm": 0.18526971863081118,
+      "learning_rate": 7.183752591325716e-07,
+      "loss": 0.2691,
+      "step": 2183
+    },
+    {
+      "epoch": 1.0986767485822306,
+      "grad_norm": 0.19144116070312203,
+      "learning_rate": 7.181336374309098e-07,
+      "loss": 0.2925,
+      "step": 2184
+    },
+    {
+      "epoch": 1.0991808443604285,
+      "grad_norm": 0.1727996053307209,
+      "learning_rate": 7.17891952799045e-07,
+      "loss": 0.2745,
+      "step": 2185
+    },
+    {
+      "epoch": 1.0996849401386264,
+      "grad_norm": 0.1706944820187512,
+      "learning_rate": 7.176502053067016e-07,
+      "loss": 0.2773,
+      "step": 2186
+    },
+    {
+      "epoch": 1.1001890359168243,
+      "grad_norm": 0.17180090645464358,
+      "learning_rate": 7.17408395023622e-07,
+      "loss": 0.2812,
+      "step": 2187
+    },
+    {
+      "epoch": 1.1006931316950221,
+      "grad_norm": 0.18042375948484118,
+      "learning_rate": 7.171665220195675e-07,
+      "loss": 0.2832,
+      "step": 2188
+    },
+    {
+      "epoch": 1.1011972274732198,
+      "grad_norm": 0.17833078926842466,
+      "learning_rate": 7.169245863643165e-07,
+      "loss": 0.2781,
+      "step": 2189
+    },
+    {
+      "epoch": 1.1011972274732198,
+      "eval_loss": 0.3109322786331177,
+      "eval_runtime": 18.3883,
+      "eval_samples_per_second": 46.497,
+      "eval_steps_per_second": 0.979,
+      "step": 2189
+    },
+    {
+      "epoch": 1.1017013232514177,
+      "grad_norm": 0.17668710642540988,
+      "learning_rate": 7.166825881276663e-07,
+      "loss": 0.2814,
+      "step": 2190
+    },
+    {
+      "epoch": 1.1022054190296156,
+      "grad_norm": 0.17892757631840098,
+      "learning_rate": 7.164405273794315e-07,
+      "loss": 0.287,
+      "step": 2191
+    },
+    {
+      "epoch": 1.1027095148078134,
+      "grad_norm": 0.19272913994143642,
+      "learning_rate": 7.161984041894453e-07,
+      "loss": 0.2671,
+      "step": 2192
+    },
+    {
+      "epoch": 1.1032136105860113,
+      "grad_norm": 0.1686523316395002,
+      "learning_rate": 7.159562186275589e-07,
+      "loss": 0.281,
+      "step": 2193
+    },
+    {
+      "epoch": 1.1037177063642092,
+      "grad_norm": 0.1802469776190519,
+      "learning_rate": 7.157139707636411e-07,
+      "loss": 0.2778,
+      "step": 2194
+    },
+    {
+      "epoch": 1.104221802142407,
+      "grad_norm": 0.17372788767408012,
+      "learning_rate": 7.15471660667579e-07,
+      "loss": 0.2766,
+      "step": 2195
+    },
+    {
+      "epoch": 1.104725897920605,
+      "grad_norm": 0.17515408548411832,
+      "learning_rate": 7.152292884092776e-07,
+      "loss": 0.2781,
+      "step": 2196
+    },
+    {
+      "epoch": 1.1052299936988028,
+      "grad_norm": 0.17120886057523532,
+      "learning_rate": 7.149868540586599e-07,
+      "loss": 0.2804,
+      "step": 2197
+    },
+    {
+      "epoch": 1.1057340894770007,
+      "grad_norm": 0.1967834646928095,
+      "learning_rate": 7.147443576856667e-07,
+      "loss": 0.2701,
+      "step": 2198
+    },
+    {
+      "epoch": 1.1062381852551986,
+      "grad_norm": 0.17730817500369223,
+      "learning_rate": 7.145017993602562e-07,
+      "loss": 0.2743,
+      "step": 2199
+    },
+    {
+      "epoch": 1.1067422810333964,
+      "grad_norm": 0.1963166032968741,
+      "learning_rate": 7.142591791524056e-07,
+      "loss": 0.2837,
+      "step": 2200
+    },
+    {
+      "epoch": 1.107246376811594,
+      "grad_norm": 0.17742060612270208,
+      "learning_rate": 7.14016497132109e-07,
+      "loss": 0.2878,
+      "step": 2201
+    },
+    {
+      "epoch": 1.107750472589792,
+      "grad_norm": 0.1746362136216461,
+      "learning_rate": 7.137737533693787e-07,
+      "loss": 0.2639,
+      "step": 2202
+    },
+    {
+      "epoch": 1.1082545683679899,
+      "grad_norm": 0.17804698672223912,
+      "learning_rate": 7.135309479342449e-07,
+      "loss": 0.2787,
+      "step": 2203
+    },
+    {
+      "epoch": 1.1087586641461877,
+      "grad_norm": 0.17579637022355174,
+      "learning_rate": 7.132880808967553e-07,
+      "loss": 0.2842,
+      "step": 2204
+    },
+    {
+      "epoch": 1.1092627599243856,
+      "grad_norm": 0.16989385004038,
+      "learning_rate": 7.130451523269757e-07,
+      "loss": 0.2677,
+      "step": 2205
+    },
+    {
+      "epoch": 1.1097668557025835,
+      "grad_norm": 0.17551046931385217,
+      "learning_rate": 7.128021622949894e-07,
+      "loss": 0.2797,
+      "step": 2206
+    },
+    {
+      "epoch": 1.1102709514807814,
+      "grad_norm": 0.17789389712157294,
+      "learning_rate": 7.125591108708973e-07,
+      "loss": 0.2638,
+      "step": 2207
+    },
+    {
+      "epoch": 1.1107750472589792,
+      "grad_norm": 0.1676683446730687,
+      "learning_rate": 7.123159981248187e-07,
+      "loss": 0.2743,
+      "step": 2208
+    },
+    {
+      "epoch": 1.1112791430371771,
+      "grad_norm": 0.17269955486897387,
+      "learning_rate": 7.120728241268897e-07,
+      "loss": 0.2753,
+      "step": 2209
+    },
+    {
+      "epoch": 1.111783238815375,
+      "grad_norm": 0.1772494652206068,
+      "learning_rate": 7.118295889472648e-07,
+      "loss": 0.2836,
+      "step": 2210
+    },
+    {
+      "epoch": 1.1122873345935729,
+      "grad_norm": 0.17063122243632267,
+      "learning_rate": 7.115862926561156e-07,
+      "loss": 0.2727,
+      "step": 2211
+    },
+    {
+      "epoch": 1.1127914303717708,
+      "grad_norm": 0.18443201636788845,
+      "learning_rate": 7.113429353236317e-07,
+      "loss": 0.2724,
+      "step": 2212
+    },
+    {
+      "epoch": 1.1132955261499684,
+      "grad_norm": 0.17187215404760892,
+      "learning_rate": 7.110995170200203e-07,
+      "loss": 0.2791,
+      "step": 2213
+    },
+    {
+      "epoch": 1.1137996219281663,
+      "grad_norm": 0.21059430107866162,
+      "learning_rate": 7.108560378155058e-07,
+      "loss": 0.2753,
+      "step": 2214
+    },
+    {
+      "epoch": 1.1143037177063642,
+      "grad_norm": 0.17008648209502905,
+      "learning_rate": 7.106124977803305e-07,
+      "loss": 0.2679,
+      "step": 2215
+    },
+    {
+      "epoch": 1.114807813484562,
+      "grad_norm": 0.17428666864442163,
+      "learning_rate": 7.103688969847544e-07,
+      "loss": 0.2788,
+      "step": 2216
+    },
+    {
+      "epoch": 1.11531190926276,
+      "grad_norm": 0.16978332708330493,
+      "learning_rate": 7.101252354990547e-07,
+      "loss": 0.2812,
+      "step": 2217
+    },
+    {
+      "epoch": 1.1158160050409578,
+      "grad_norm": 0.16763401695592228,
+      "learning_rate": 7.09881513393526e-07,
+      "loss": 0.2718,
+      "step": 2218
+    },
+    {
+      "epoch": 1.1163201008191557,
+      "grad_norm": 0.17695900270150158,
+      "learning_rate": 7.09637730738481e-07,
+      "loss": 0.2775,
+      "step": 2219
+    },
+    {
+      "epoch": 1.1168241965973535,
+      "grad_norm": 0.17577147289069614,
+      "learning_rate": 7.093938876042495e-07,
+      "loss": 0.2817,
+      "step": 2220
+    },
+    {
+      "epoch": 1.1173282923755514,
+      "grad_norm": 0.17875894680716742,
+      "learning_rate": 7.091499840611782e-07,
+      "loss": 0.2822,
+      "step": 2221
+    },
+    {
+      "epoch": 1.1178323881537493,
+      "grad_norm": 0.17471488732657428,
+      "learning_rate": 7.089060201796323e-07,
+      "loss": 0.2804,
+      "step": 2222
+    },
+    {
+      "epoch": 1.1183364839319472,
+      "grad_norm": 0.17146413279465827,
+      "learning_rate": 7.086619960299936e-07,
+      "loss": 0.2662,
+      "step": 2223
+    },
+    {
+      "epoch": 1.1188405797101448,
+      "grad_norm": 0.17497006413256116,
+      "learning_rate": 7.084179116826616e-07,
+      "loss": 0.273,
+      "step": 2224
+    },
+    {
+      "epoch": 1.1193446754883427,
+      "grad_norm": 0.17948306341455134,
+      "learning_rate": 7.081737672080533e-07,
+      "loss": 0.2772,
+      "step": 2225
+    },
+    {
+      "epoch": 1.1198487712665406,
+      "grad_norm": 0.18647108183544914,
+      "learning_rate": 7.079295626766026e-07,
+      "loss": 0.2788,
+      "step": 2226
+    },
+    {
+      "epoch": 1.1203528670447385,
+      "grad_norm": 0.1974823122970526,
+      "learning_rate": 7.076852981587613e-07,
+      "loss": 0.2993,
+      "step": 2227
+    },
+    {
+      "epoch": 1.1208569628229363,
+      "grad_norm": 0.17288627565030815,
+      "learning_rate": 7.07440973724998e-07,
+      "loss": 0.2775,
+      "step": 2228
+    },
+    {
+      "epoch": 1.1213610586011342,
+      "grad_norm": 0.18565848907871105,
+      "learning_rate": 7.071965894457987e-07,
+      "loss": 0.2763,
+      "step": 2229
+    },
+    {
+      "epoch": 1.121865154379332,
+      "grad_norm": 0.18033187048793917,
+      "learning_rate": 7.069521453916669e-07,
+      "loss": 0.2814,
+      "step": 2230
+    },
+    {
+      "epoch": 1.12236925015753,
+      "grad_norm": 0.17992401726222884,
+      "learning_rate": 7.067076416331233e-07,
+      "loss": 0.2729,
+      "step": 2231
+    },
+    {
+      "epoch": 1.1228733459357279,
+      "grad_norm": 0.17515567963879827,
+      "learning_rate": 7.064630782407053e-07,
+      "loss": 0.2971,
+      "step": 2232
+    },
+    {
+      "epoch": 1.1233774417139257,
+      "grad_norm": 0.16996972984196193,
+      "learning_rate": 7.062184552849683e-07,
+      "loss": 0.2752,
+      "step": 2233
+    },
+    {
+      "epoch": 1.1238815374921236,
+      "grad_norm": 0.1681151195763795,
+      "learning_rate": 7.059737728364844e-07,
+      "loss": 0.2764,
+      "step": 2234
+    },
+    {
+      "epoch": 1.1243856332703213,
+      "grad_norm": 0.17197455617986673,
+      "learning_rate": 7.05729030965843e-07,
+      "loss": 0.2732,
+      "step": 2235
+    },
+    {
+      "epoch": 1.1248897290485191,
+      "grad_norm": 0.17380812930247844,
+      "learning_rate": 7.054842297436506e-07,
+      "loss": 0.2713,
+      "step": 2236
+    },
+    {
+      "epoch": 1.125393824826717,
+      "grad_norm": 0.17146791708239711,
+      "learning_rate": 7.052393692405308e-07,
+      "loss": 0.274,
+      "step": 2237
+    },
+    {
+      "epoch": 1.125897920604915,
+      "grad_norm": 0.17522062714630593,
+      "learning_rate": 7.049944495271244e-07,
+      "loss": 0.291,
+      "step": 2238
+    },
+    {
+      "epoch": 1.1264020163831128,
+      "grad_norm": 0.18020842179105392,
+      "learning_rate": 7.047494706740891e-07,
+      "loss": 0.2776,
+      "step": 2239
+    },
+    {
+      "epoch": 1.1269061121613106,
+      "grad_norm": 0.17094614470132977,
+      "learning_rate": 7.045044327521e-07,
+      "loss": 0.2774,
+      "step": 2240
+    },
+    {
+      "epoch": 1.1274102079395085,
+      "grad_norm": 0.2013051302911893,
+      "learning_rate": 7.042593358318488e-07,
+      "loss": 0.274,
+      "step": 2241
+    },
+    {
+      "epoch": 1.1279143037177064,
+      "grad_norm": 0.17297868860952875,
+      "learning_rate": 7.040141799840446e-07,
+      "loss": 0.2651,
+      "step": 2242
+    },
+    {
+      "epoch": 1.1284183994959043,
+      "grad_norm": 0.17357556224692539,
+      "learning_rate": 7.037689652794132e-07,
+      "loss": 0.2742,
+      "step": 2243
+    },
+    {
+      "epoch": 1.1289224952741022,
+      "grad_norm": 0.1726295931526859,
+      "learning_rate": 7.035236917886977e-07,
+      "loss": 0.2749,
+      "step": 2244
+    },
+    {
+      "epoch": 1.1294265910523,
+      "grad_norm": 0.18408306057697696,
+      "learning_rate": 7.032783595826577e-07,
+      "loss": 0.2749,
+      "step": 2245
+    },
+    {
+      "epoch": 1.1299306868304977,
+      "grad_norm": 0.1760499835143201,
+      "learning_rate": 7.030329687320704e-07,
+      "loss": 0.2757,
+      "step": 2246
+    },
+    {
+      "epoch": 1.1304347826086956,
+      "grad_norm": 0.17625183802064207,
+      "learning_rate": 7.027875193077293e-07,
+      "loss": 0.2794,
+      "step": 2247
+    },
+    {
+      "epoch": 1.1309388783868934,
+      "grad_norm": 0.17473737339780746,
+      "learning_rate": 7.02542011380445e-07,
+      "loss": 0.2814,
+      "step": 2248
+    },
+    {
+      "epoch": 1.1314429741650913,
+      "grad_norm": 0.17699959608452862,
+      "learning_rate": 7.022964450210451e-07,
+      "loss": 0.2793,
+      "step": 2249
+    },
+    {
+      "epoch": 1.1319470699432892,
+      "grad_norm": 0.17541995578274833,
+      "learning_rate": 7.020508203003741e-07,
+      "loss": 0.2669,
+      "step": 2250
+    },
+    {
+      "epoch": 1.132451165721487,
+      "grad_norm": 0.1835401757832211,
+      "learning_rate": 7.01805137289293e-07,
+      "loss": 0.2743,
+      "step": 2251
+    },
+    {
+      "epoch": 1.132955261499685,
+      "grad_norm": 0.19020483202650845,
+      "learning_rate": 7.015593960586799e-07,
+      "loss": 0.2774,
+      "step": 2252
+    },
+    {
+      "epoch": 1.1334593572778828,
+      "grad_norm": 0.17000536309275166,
+      "learning_rate": 7.013135966794295e-07,
+      "loss": 0.2811,
+      "step": 2253
+    },
+    {
+      "epoch": 1.1339634530560807,
+      "grad_norm": 0.17801380513763565,
+      "learning_rate": 7.010677392224537e-07,
+      "loss": 0.2935,
+      "step": 2254
+    },
+    {
+      "epoch": 1.1344675488342786,
+      "grad_norm": 0.1787152664038223,
+      "learning_rate": 7.008218237586807e-07,
+      "loss": 0.2664,
+      "step": 2255
+    },
+    {
+      "epoch": 1.1349716446124765,
+      "grad_norm": 0.17254574236169345,
+      "learning_rate": 7.005758503590555e-07,
+      "loss": 0.2844,
+      "step": 2256
+    },
+    {
+      "epoch": 1.1354757403906741,
+      "grad_norm": 0.18663732747974232,
+      "learning_rate": 7.003298190945399e-07,
+      "loss": 0.2615,
+      "step": 2257
+    },
+    {
+      "epoch": 1.135979836168872,
+      "grad_norm": 0.18225915688322497,
+      "learning_rate": 7.000837300361127e-07,
+      "loss": 0.273,
+      "step": 2258
+    },
+    {
+      "epoch": 1.1364839319470699,
+      "grad_norm": 0.17525558203612415,
+      "learning_rate": 6.998375832547687e-07,
+      "loss": 0.2709,
+      "step": 2259
+    },
+    {
+      "epoch": 1.1369880277252677,
+      "grad_norm": 0.17721206051763022,
+      "learning_rate": 6.995913788215198e-07,
+      "loss": 0.2638,
+      "step": 2260
+    },
+    {
+      "epoch": 1.1374921235034656,
+      "grad_norm": 0.17478241116943685,
+      "learning_rate": 6.993451168073945e-07,
+      "loss": 0.2773,
+      "step": 2261
+    },
+    {
+      "epoch": 1.1379962192816635,
+      "grad_norm": 0.1774428076107305,
+      "learning_rate": 6.990987972834382e-07,
+      "loss": 0.2631,
+      "step": 2262
+    },
+    {
+      "epoch": 1.1385003150598614,
+      "grad_norm": 0.17410794951513323,
+      "learning_rate": 6.988524203207117e-07,
+      "loss": 0.2707,
+      "step": 2263
+    },
+    {
+      "epoch": 1.1390044108380593,
+      "grad_norm": 0.17441217480295762,
+      "learning_rate": 6.986059859902941e-07,
+      "loss": 0.2868,
+      "step": 2264
+    },
+    {
+      "epoch": 1.1395085066162571,
+      "grad_norm": 0.1747291161914758,
+      "learning_rate": 6.983594943632799e-07,
+      "loss": 0.2649,
+      "step": 2265
+    },
+    {
+      "epoch": 1.140012602394455,
+      "grad_norm": 0.17307228646736642,
+      "learning_rate": 6.981129455107802e-07,
+      "loss": 0.2861,
+      "step": 2266
+    },
+    {
+      "epoch": 1.1405166981726529,
+      "grad_norm": 0.18952720395194383,
+      "learning_rate": 6.978663395039231e-07,
+      "loss": 0.2865,
+      "step": 2267
+    },
+    {
+      "epoch": 1.1410207939508505,
+      "grad_norm": 0.17922645708493065,
+      "learning_rate": 6.976196764138526e-07,
+      "loss": 0.2808,
+      "step": 2268
+    },
+    {
+      "epoch": 1.1415248897290486,
+      "grad_norm": 0.17827854018956354,
+      "learning_rate": 6.973729563117297e-07,
+      "loss": 0.2667,
+      "step": 2269
+    },
+    {
+      "epoch": 1.1420289855072463,
+      "grad_norm": 0.18226122048319768,
+      "learning_rate": 6.971261792687315e-07,
+      "loss": 0.2784,
+      "step": 2270
+    },
+    {
+      "epoch": 1.1425330812854442,
+      "grad_norm": 0.1720926208778343,
+      "learning_rate": 6.968793453560518e-07,
+      "loss": 0.2696,
+      "step": 2271
+    },
+    {
+      "epoch": 1.143037177063642,
+      "grad_norm": 0.18305708913870564,
+      "learning_rate": 6.966324546449006e-07,
+      "loss": 0.281,
+      "step": 2272
+    },
+    {
+      "epoch": 1.14354127284184,
+      "grad_norm": 0.1731336224949858,
+      "learning_rate": 6.963855072065043e-07,
+      "loss": 0.2779,
+      "step": 2273
+    },
+    {
+      "epoch": 1.1440453686200378,
+      "grad_norm": 0.1748927423648997,
+      "learning_rate": 6.961385031121057e-07,
+      "loss": 0.2753,
+      "step": 2274
+    },
+    {
+      "epoch": 1.1445494643982357,
+      "grad_norm": 0.17333262216089212,
+      "learning_rate": 6.958914424329638e-07,
+      "loss": 0.2885,
+      "step": 2275
+    },
+    {
+      "epoch": 1.1450535601764336,
+      "grad_norm": 0.18183632440429925,
+      "learning_rate": 6.956443252403544e-07,
+      "loss": 0.2724,
+      "step": 2276
+    },
+    {
+      "epoch": 1.1455576559546314,
+      "grad_norm": 0.17479355568350904,
+      "learning_rate": 6.953971516055691e-07,
+      "loss": 0.2785,
+      "step": 2277
+    },
+    {
+      "epoch": 1.1460617517328293,
+      "grad_norm": 0.1785764248499925,
+      "learning_rate": 6.951499215999161e-07,
+      "loss": 0.2818,
+      "step": 2278
+    },
+    {
+      "epoch": 1.1465658475110272,
+      "grad_norm": 0.1707075256134137,
+      "learning_rate": 6.949026352947194e-07,
+      "loss": 0.2776,
+      "step": 2279
+    },
+    {
+      "epoch": 1.147069943289225,
+      "grad_norm": 0.17454520681763225,
+      "learning_rate": 6.946552927613201e-07,
+      "loss": 0.279,
+      "step": 2280
+    },
+    {
+      "epoch": 1.1475740390674227,
+      "grad_norm": 0.16725292243096776,
+      "learning_rate": 6.944078940710743e-07,
+      "loss": 0.2706,
+      "step": 2281
+    },
+    {
+      "epoch": 1.1480781348456206,
+      "grad_norm": 0.18158312079337943,
+      "learning_rate": 6.941604392953555e-07,
+      "loss": 0.2857,
+      "step": 2282
+    },
+    {
+      "epoch": 1.1485822306238185,
+      "grad_norm": 0.17686176298428924,
+      "learning_rate": 6.939129285055527e-07,
+      "loss": 0.2982,
+      "step": 2283
+    },
+    {
+      "epoch": 1.1490863264020164,
+      "grad_norm": 0.16814975626978187,
+      "learning_rate": 6.936653617730712e-07,
+      "loss": 0.2719,
+      "step": 2284
+    },
+    {
+      "epoch": 1.1495904221802142,
+      "grad_norm": 0.1759711441539806,
+      "learning_rate": 6.934177391693327e-07,
+      "loss": 0.2744,
+      "step": 2285
+    },
+    {
+      "epoch": 1.150094517958412,
+      "grad_norm": 0.17511838680958677,
+      "learning_rate": 6.931700607657744e-07,
+      "loss": 0.2739,
+      "step": 2286
+    },
+    {
+      "epoch": 1.15059861373661,
+      "grad_norm": 0.17740624835254679,
+      "learning_rate": 6.929223266338504e-07,
+      "loss": 0.269,
+      "step": 2287
+    },
+    {
+      "epoch": 1.1511027095148079,
+      "grad_norm": 0.1681514431068215,
+      "learning_rate": 6.926745368450301e-07,
+      "loss": 0.2749,
+      "step": 2288
+    },
+    {
+      "epoch": 1.1516068052930057,
+      "grad_norm": 0.1691426996495545,
+      "learning_rate": 6.924266914707995e-07,
+      "loss": 0.2644,
+      "step": 2289
+    },
+    {
+      "epoch": 1.1521109010712036,
+      "grad_norm": 0.1790330541665502,
+      "learning_rate": 6.921787905826605e-07,
+      "loss": 0.2823,
+      "step": 2290
+    },
+    {
+      "epoch": 1.1526149968494015,
+      "grad_norm": 0.1734370061342181,
+      "learning_rate": 6.919308342521308e-07,
+      "loss": 0.2764,
+      "step": 2291
+    },
+    {
+      "epoch": 1.1531190926275992,
+      "grad_norm": 0.18865399840145108,
+      "learning_rate": 6.916828225507443e-07,
+      "loss": 0.2811,
+      "step": 2292
+    },
+    {
+      "epoch": 1.153623188405797,
+      "grad_norm": 0.17616068936635845,
+      "learning_rate": 6.914347555500513e-07,
+      "loss": 0.2879,
+      "step": 2293
+    },
+    {
+      "epoch": 1.154127284183995,
+      "grad_norm": 0.17879090160590586,
+      "learning_rate": 6.911866333216169e-07,
+      "loss": 0.2717,
+      "step": 2294
+    },
+    {
+      "epoch": 1.1546313799621928,
+      "grad_norm": 0.18207500480662475,
+      "learning_rate": 6.909384559370233e-07,
+      "loss": 0.275,
+      "step": 2295
+    },
+    {
+      "epoch": 1.1551354757403907,
+      "grad_norm": 0.17522149059459446,
+      "learning_rate": 6.906902234678678e-07,
+      "loss": 0.2688,
+      "step": 2296
+    },
+    {
+      "epoch": 1.1556395715185885,
+      "grad_norm": 0.1938343793990561,
+      "learning_rate": 6.904419359857641e-07,
+      "loss": 0.2925,
+      "step": 2297
+    },
+    {
+      "epoch": 1.1561436672967864,
+      "grad_norm": 0.20218185185326207,
+      "learning_rate": 6.901935935623415e-07,
+      "loss": 0.2804,
+      "step": 2298
+    },
+    {
+      "epoch": 1.1566477630749843,
+      "grad_norm": 0.17255949907153817,
+      "learning_rate": 6.899451962692454e-07,
+      "loss": 0.2717,
+      "step": 2299
+    },
+    {
+      "epoch": 1.1571518588531822,
+      "grad_norm": 0.17432970101357595,
+      "learning_rate": 6.896967441781368e-07,
+      "loss": 0.2609,
+      "step": 2300
+    },
+    {
+      "epoch": 1.15765595463138,
+      "grad_norm": 0.17725867905667875,
+      "learning_rate": 6.894482373606927e-07,
+      "loss": 0.2877,
+      "step": 2301
+    },
+    {
+      "epoch": 1.158160050409578,
+      "grad_norm": 0.1817120717141377,
+      "learning_rate": 6.891996758886058e-07,
+      "loss": 0.2776,
+      "step": 2302
+    },
+    {
+      "epoch": 1.1586641461877756,
+      "grad_norm": 0.17327028373633524,
+      "learning_rate": 6.889510598335843e-07,
+      "loss": 0.2491,
+      "step": 2303
+    },
+    {
+      "epoch": 1.1591682419659735,
+      "grad_norm": 0.17396621817828842,
+      "learning_rate": 6.887023892673525e-07,
+      "loss": 0.2872,
+      "step": 2304
+    },
+    {
+      "epoch": 1.1596723377441713,
+      "grad_norm": 0.17083195902984658,
+      "learning_rate": 6.884536642616504e-07,
+      "loss": 0.2714,
+      "step": 2305
+    },
+    {
+      "epoch": 1.1601764335223692,
+      "grad_norm": 0.17991670145657968,
+      "learning_rate": 6.882048848882335e-07,
+      "loss": 0.2973,
+      "step": 2306
+    },
+    {
+      "epoch": 1.160680529300567,
+      "grad_norm": 0.17544125470825095,
+      "learning_rate": 6.879560512188733e-07,
+      "loss": 0.2767,
+      "step": 2307
+    },
+    {
+      "epoch": 1.161184625078765,
+      "grad_norm": 0.18423859716455426,
+      "learning_rate": 6.877071633253566e-07,
+      "loss": 0.2763,
+      "step": 2308
+    },
+    {
+      "epoch": 1.1616887208569628,
+      "grad_norm": 0.18681981813114995,
+      "learning_rate": 6.874582212794861e-07,
+      "loss": 0.2789,
+      "step": 2309
+    },
+    {
+      "epoch": 1.1621928166351607,
+      "grad_norm": 0.17640590031489775,
+      "learning_rate": 6.872092251530799e-07,
+      "loss": 0.288,
+      "step": 2310
+    },
+    {
+      "epoch": 1.1626969124133586,
+      "grad_norm": 0.17113861686922094,
+      "learning_rate": 6.869601750179721e-07,
+      "loss": 0.2721,
+      "step": 2311
+    },
+    {
+      "epoch": 1.1632010081915565,
+      "grad_norm": 0.1713692791805304,
+      "learning_rate": 6.867110709460118e-07,
+      "loss": 0.2767,
+      "step": 2312
+    },
+    {
+      "epoch": 1.1637051039697544,
+      "grad_norm": 0.1729727527955606,
+      "learning_rate": 6.864619130090642e-07,
+      "loss": 0.298,
+      "step": 2313
+    },
+    {
+      "epoch": 1.164209199747952,
+      "grad_norm": 0.18317780598435884,
+      "learning_rate": 6.862127012790098e-07,
+      "loss": 0.271,
+      "step": 2314
+    },
+    {
+      "epoch": 1.1647132955261499,
+      "grad_norm": 0.16947854182681354,
+      "learning_rate": 6.859634358277445e-07,
+      "loss": 0.2731,
+      "step": 2315
+    },
+    {
+      "epoch": 1.1652173913043478,
+      "grad_norm": 0.1764313134129964,
+      "learning_rate": 6.8571411672718e-07,
+      "loss": 0.2898,
+      "step": 2316
+    },
+    {
+      "epoch": 1.1657214870825456,
+      "grad_norm": 0.18128405805703027,
+      "learning_rate": 6.854647440492434e-07,
+      "loss": 0.2713,
+      "step": 2317
+    },
+    {
+      "epoch": 1.1662255828607435,
+      "grad_norm": 0.17302921586300837,
+      "learning_rate": 6.852153178658768e-07,
+      "loss": 0.2791,
+      "step": 2318
+    },
+    {
+      "epoch": 1.1667296786389414,
+      "grad_norm": 0.18238569158548426,
+      "learning_rate": 6.849658382490386e-07,
+      "loss": 0.2637,
+      "step": 2319
+    },
+    {
+      "epoch": 1.1672337744171393,
+      "grad_norm": 0.20560524437351874,
+      "learning_rate": 6.847163052707017e-07,
+      "loss": 0.2886,
+      "step": 2320
+    },
+    {
+      "epoch": 1.1677378701953371,
+      "grad_norm": 0.17279730398402016,
+      "learning_rate": 6.84466719002855e-07,
+      "loss": 0.276,
+      "step": 2321
+    },
+    {
+      "epoch": 1.168241965973535,
+      "grad_norm": 0.1748627881177702,
+      "learning_rate": 6.842170795175025e-07,
+      "loss": 0.2777,
+      "step": 2322
+    },
+    {
+      "epoch": 1.168746061751733,
+      "grad_norm": 0.20772510184287282,
+      "learning_rate": 6.839673868866639e-07,
+      "loss": 0.2745,
+      "step": 2323
+    },
+    {
+      "epoch": 1.1692501575299308,
+      "grad_norm": 0.1703532830956251,
+      "learning_rate": 6.837176411823738e-07,
+      "loss": 0.2631,
+      "step": 2324
+    },
+    {
+      "epoch": 1.1697542533081284,
+      "grad_norm": 0.1793111095271819,
+      "learning_rate": 6.834678424766822e-07,
+      "loss": 0.2818,
+      "step": 2325
+    },
+    {
+      "epoch": 1.1702583490863263,
+      "grad_norm": 0.17008601463744455,
+      "learning_rate": 6.832179908416546e-07,
+      "loss": 0.2637,
+      "step": 2326
+    },
+    {
+      "epoch": 1.1707624448645242,
+      "grad_norm": 0.16973624705065182,
+      "learning_rate": 6.829680863493717e-07,
+      "loss": 0.3012,
+      "step": 2327
+    },
+    {
+      "epoch": 1.171266540642722,
+      "grad_norm": 0.18216274445798897,
+      "learning_rate": 6.82718129071929e-07,
+      "loss": 0.2887,
+      "step": 2328
+    },
+    {
+      "epoch": 1.17177063642092,
+      "grad_norm": 0.18182861892057042,
+      "learning_rate": 6.824681190814383e-07,
+      "loss": 0.2825,
+      "step": 2329
+    },
+    {
+      "epoch": 1.1722747321991178,
+      "grad_norm": 0.18263373935726934,
+      "learning_rate": 6.822180564500254e-07,
+      "loss": 0.294,
+      "step": 2330
+    },
+    {
+      "epoch": 1.1727788279773157,
+      "grad_norm": 0.17351672150821956,
+      "learning_rate": 6.81967941249832e-07,
+      "loss": 0.2799,
+      "step": 2331
+    },
+    {
+      "epoch": 1.1732829237555136,
+      "grad_norm": 0.17364729038219348,
+      "learning_rate": 6.817177735530149e-07,
+      "loss": 0.2664,
+      "step": 2332
+    },
+    {
+      "epoch": 1.1737870195337115,
+      "grad_norm": 0.1808124777048507,
+      "learning_rate": 6.814675534317457e-07,
+      "loss": 0.2714,
+      "step": 2333
+    },
+    {
+      "epoch": 1.1742911153119093,
+      "grad_norm": 0.180877239519155,
+      "learning_rate": 6.812172809582114e-07,
+      "loss": 0.2828,
+      "step": 2334
+    },
+    {
+      "epoch": 1.1747952110901072,
+      "grad_norm": 0.17380368919234485,
+      "learning_rate": 6.809669562046142e-07,
+      "loss": 0.2737,
+      "step": 2335
+    },
+    {
+      "epoch": 1.1752993068683049,
+      "grad_norm": 0.19015792157284472,
+      "learning_rate": 6.807165792431712e-07,
+      "loss": 0.2909,
+      "step": 2336
+    },
+    {
+      "epoch": 1.175803402646503,
+      "grad_norm": 0.1739820937338746,
+      "learning_rate": 6.804661501461146e-07,
+      "loss": 0.2797,
+      "step": 2337
+    },
+    {
+      "epoch": 1.1763074984247006,
+      "grad_norm": 0.17829150310208366,
+      "learning_rate": 6.802156689856914e-07,
+      "loss": 0.272,
+      "step": 2338
+    },
+    {
+      "epoch": 1.1768115942028985,
+      "grad_norm": 0.17618058591368718,
+      "learning_rate": 6.799651358341644e-07,
+      "loss": 0.259,
+      "step": 2339
+    },
+    {
+      "epoch": 1.1773156899810964,
+      "grad_norm": 0.16972502373995604,
+      "learning_rate": 6.797145507638103e-07,
+      "loss": 0.2758,
+      "step": 2340
+    },
+    {
+      "epoch": 1.1778197857592942,
+      "grad_norm": 0.17376894673770923,
+      "learning_rate": 6.794639138469215e-07,
+      "loss": 0.2598,
+      "step": 2341
+    },
+    {
+      "epoch": 1.1783238815374921,
+      "grad_norm": 0.16890243175150166,
+      "learning_rate": 6.792132251558057e-07,
+      "loss": 0.2679,
+      "step": 2342
+    },
+    {
+      "epoch": 1.17882797731569,
+      "grad_norm": 0.16892397355410096,
+      "learning_rate": 6.789624847627842e-07,
+      "loss": 0.2713,
+      "step": 2343
+    },
+    {
+      "epoch": 1.1793320730938879,
+      "grad_norm": 0.18081668386093178,
+      "learning_rate": 6.787116927401947e-07,
+      "loss": 0.2839,
+      "step": 2344
+    },
+    {
+      "epoch": 1.1798361688720858,
+      "grad_norm": 0.1785462609641959,
+      "learning_rate": 6.784608491603887e-07,
+      "loss": 0.2767,
+      "step": 2345
+    },
+    {
+      "epoch": 1.1803402646502836,
+      "grad_norm": 0.17328874161653496,
+      "learning_rate": 6.782099540957334e-07,
+      "loss": 0.2792,
+      "step": 2346
+    },
+    {
+      "epoch": 1.1808443604284813,
+      "grad_norm": 0.17812431021686165,
+      "learning_rate": 6.779590076186103e-07,
+      "loss": 0.2756,
+      "step": 2347
+    },
+    {
+      "epoch": 1.1813484562066794,
+      "grad_norm": 0.18026369211177573,
+      "learning_rate": 6.777080098014157e-07,
+      "loss": 0.2829,
+      "step": 2348
+    },
+    {
+      "epoch": 1.181852551984877,
+      "grad_norm": 0.17224760523581167,
+      "learning_rate": 6.774569607165612e-07,
+      "loss": 0.2715,
+      "step": 2349
+    },
+    {
+      "epoch": 1.182356647763075,
+      "grad_norm": 0.17446195179033158,
+      "learning_rate": 6.772058604364728e-07,
+      "loss": 0.2816,
+      "step": 2350
+    },
+    {
+      "epoch": 1.1828607435412728,
+      "grad_norm": 0.18401389185741523,
+      "learning_rate": 6.769547090335915e-07,
+      "loss": 0.2756,
+      "step": 2351
+    },
+    {
+      "epoch": 1.1833648393194707,
+      "grad_norm": 0.17710337017025912,
+      "learning_rate": 6.767035065803728e-07,
+      "loss": 0.2662,
+      "step": 2352
+    },
+    {
+      "epoch": 1.1838689350976686,
+      "grad_norm": 0.17209284528655844,
+      "learning_rate": 6.76452253149287e-07,
+      "loss": 0.2643,
+      "step": 2353
+    },
+    {
+      "epoch": 1.1843730308758664,
+      "grad_norm": 0.19313801740333889,
+      "learning_rate": 6.762009488128193e-07,
+      "loss": 0.2859,
+      "step": 2354
+    },
+    {
+      "epoch": 1.1848771266540643,
+      "grad_norm": 0.185715764727419,
+      "learning_rate": 6.759495936434694e-07,
+      "loss": 0.2778,
+      "step": 2355
+    },
+    {
+      "epoch": 1.1853812224322622,
+      "grad_norm": 0.18586478407699625,
+      "learning_rate": 6.756981877137515e-07,
+      "loss": 0.2781,
+      "step": 2356
+    },
+    {
+      "epoch": 1.18588531821046,
+      "grad_norm": 0.18286265156559076,
+      "learning_rate": 6.754467310961951e-07,
+      "loss": 0.2941,
+      "step": 2357
+    },
+    {
+      "epoch": 1.186389413988658,
+      "grad_norm": 0.1827838570028727,
+      "learning_rate": 6.751952238633435e-07,
+      "loss": 0.2799,
+      "step": 2358
+    },
+    {
+      "epoch": 1.1868935097668558,
+      "grad_norm": 0.17564052156297236,
+      "learning_rate": 6.74943666087755e-07,
+      "loss": 0.2859,
+      "step": 2359
+    },
+    {
+      "epoch": 1.1873976055450535,
+      "grad_norm": 0.17276098802425446,
+      "learning_rate": 6.746920578420027e-07,
+      "loss": 0.273,
+      "step": 2360
+    },
+    {
+      "epoch": 1.1879017013232513,
+      "grad_norm": 0.16583786340074003,
+      "learning_rate": 6.744403991986737e-07,
+      "loss": 0.2751,
+      "step": 2361
+    },
+    {
+      "epoch": 1.1884057971014492,
+      "grad_norm": 0.1758298830227382,
+      "learning_rate": 6.741886902303703e-07,
+      "loss": 0.2817,
+      "step": 2362
+    },
+    {
+      "epoch": 1.188909892879647,
+      "grad_norm": 0.1786122710003466,
+      "learning_rate": 6.739369310097087e-07,
+      "loss": 0.2746,
+      "step": 2363
+    },
+    {
+      "epoch": 1.189413988657845,
+      "grad_norm": 0.1725746663086315,
+      "learning_rate": 6.7368512160932e-07,
+      "loss": 0.2697,
+      "step": 2364
+    },
+    {
+      "epoch": 1.1899180844360429,
+      "grad_norm": 0.17766797524511566,
+      "learning_rate": 6.734332621018497e-07,
+      "loss": 0.2731,
+      "step": 2365
+    },
+    {
+      "epoch": 1.1904221802142407,
+      "grad_norm": 0.1773919783826711,
+      "learning_rate": 6.731813525599576e-07,
+      "loss": 0.2811,
+      "step": 2366
+    },
+    {
+      "epoch": 1.1909262759924386,
+      "grad_norm": 0.18244881573424246,
+      "learning_rate": 6.72929393056318e-07,
+      "loss": 0.2733,
+      "step": 2367
+    },
+    {
+      "epoch": 1.1914303717706365,
+      "grad_norm": 0.17285562925317308,
+      "learning_rate": 6.7267738366362e-07,
+      "loss": 0.2691,
+      "step": 2368
+    },
+    {
+      "epoch": 1.1919344675488344,
+      "grad_norm": 0.18991372620629085,
+      "learning_rate": 6.724253244545663e-07,
+      "loss": 0.2787,
+      "step": 2369
+    },
+    {
+      "epoch": 1.1924385633270322,
+      "grad_norm": 0.17368277958423853,
+      "learning_rate": 6.721732155018747e-07,
+      "loss": 0.2746,
+      "step": 2370
+    },
+    {
+      "epoch": 1.19294265910523,
+      "grad_norm": 0.17143055764086812,
+      "learning_rate": 6.719210568782768e-07,
+      "loss": 0.2774,
+      "step": 2371
+    },
+    {
+      "epoch": 1.1934467548834278,
+      "grad_norm": 0.1723446011688051,
+      "learning_rate": 6.716688486565192e-07,
+      "loss": 0.2615,
+      "step": 2372
+    },
+    {
+      "epoch": 1.1939508506616257,
+      "grad_norm": 0.17415883235396237,
+      "learning_rate": 6.714165909093621e-07,
+      "loss": 0.2904,
+      "step": 2373
+    },
+    {
+      "epoch": 1.1944549464398235,
+      "grad_norm": 0.17281066848566462,
+      "learning_rate": 6.711642837095804e-07,
+      "loss": 0.2629,
+      "step": 2374
+    },
+    {
+      "epoch": 1.1949590422180214,
+      "grad_norm": 0.17706321838125264,
+      "learning_rate": 6.709119271299631e-07,
+      "loss": 0.2791,
+      "step": 2375
+    },
+    {
+      "epoch": 1.1954631379962193,
+      "grad_norm": 0.17409373969496145,
+      "learning_rate": 6.706595212433137e-07,
+      "loss": 0.286,
+      "step": 2376
+    },
+    {
+      "epoch": 1.1959672337744172,
+      "grad_norm": 0.17363748305455576,
+      "learning_rate": 6.704070661224496e-07,
+      "loss": 0.2809,
+      "step": 2377
+    },
+    {
+      "epoch": 1.196471329552615,
+      "grad_norm": 0.17224099285294484,
+      "learning_rate": 6.701545618402025e-07,
+      "loss": 0.2696,
+      "step": 2378
+    },
+    {
+      "epoch": 1.196975425330813,
+      "grad_norm": 0.17472439778273308,
+      "learning_rate": 6.699020084694183e-07,
+      "loss": 0.2719,
+      "step": 2379
+    },
+    {
+      "epoch": 1.1974795211090108,
+      "grad_norm": 0.17609087999551598,
+      "learning_rate": 6.69649406082957e-07,
+      "loss": 0.2687,
+      "step": 2380
+    },
+    {
+      "epoch": 1.1979836168872087,
+      "grad_norm": 0.17688131759879658,
+      "learning_rate": 6.693967547536932e-07,
+      "loss": 0.2839,
+      "step": 2381
+    },
+    {
+      "epoch": 1.1984877126654063,
+      "grad_norm": 0.1801110500802134,
+      "learning_rate": 6.69144054554515e-07,
+      "loss": 0.2898,
+      "step": 2382
+    },
+    {
+      "epoch": 1.1989918084436042,
+      "grad_norm": 0.17434074905342395,
+      "learning_rate": 6.688913055583247e-07,
+      "loss": 0.2673,
+      "step": 2383
+    },
+    {
+      "epoch": 1.199495904221802,
+      "grad_norm": 0.18206024523654654,
+      "learning_rate": 6.686385078380392e-07,
+      "loss": 0.2817,
+      "step": 2384
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.1875715751646641,
+      "learning_rate": 6.683856614665887e-07,
+      "loss": 0.2694,
+      "step": 2385
+    },
+    {
+      "epoch": 1.2005040957781978,
+      "grad_norm": 0.17720280812976338,
+      "learning_rate": 6.68132766516918e-07,
+      "loss": 0.293,
+      "step": 2386
+    },
+    {
+      "epoch": 1.2010081915563957,
+      "grad_norm": 0.1827446975631408,
+      "learning_rate": 6.678798230619856e-07,
+      "loss": 0.2607,
+      "step": 2387
+    },
+    {
+      "epoch": 1.2015122873345936,
+      "grad_norm": 0.18309536687811928,
+      "learning_rate": 6.676268311747644e-07,
+      "loss": 0.2776,
+      "step": 2388
+    },
+    {
+      "epoch": 1.2015122873345936,
+      "eval_loss": 0.30996033549308777,
+      "eval_runtime": 18.406,
+      "eval_samples_per_second": 46.452,
+      "eval_steps_per_second": 0.978,
+      "step": 2388
+    },
+    {
+      "epoch": 1.2020163831127915,
+      "grad_norm": 0.17691733319401884,
+      "learning_rate": 6.673737909282406e-07,
+      "loss": 0.2905,
+      "step": 2389
+    },
+    {
+      "epoch": 1.2025204788909893,
+      "grad_norm": 0.17048139269250298,
+      "learning_rate": 6.671207023954151e-07,
+      "loss": 0.2683,
+      "step": 2390
+    },
+    {
+      "epoch": 1.2030245746691872,
+      "grad_norm": 0.18100250219396277,
+      "learning_rate": 6.66867565649302e-07,
+      "loss": 0.2676,
+      "step": 2391
+    },
+    {
+      "epoch": 1.203528670447385,
+      "grad_norm": 0.1723576811925902,
+      "learning_rate": 6.666143807629302e-07,
+      "loss": 0.2808,
+      "step": 2392
+    },
+    {
+      "epoch": 1.2040327662255828,
+      "grad_norm": 0.17272371476581297,
+      "learning_rate": 6.663611478093415e-07,
+      "loss": 0.2668,
+      "step": 2393
+    },
+    {
+      "epoch": 1.2045368620037806,
+      "grad_norm": 0.1852834091042434,
+      "learning_rate": 6.661078668615922e-07,
+      "loss": 0.276,
+      "step": 2394
+    },
+    {
+      "epoch": 1.2050409577819785,
+      "grad_norm": 0.17253248810106345,
+      "learning_rate": 6.658545379927523e-07,
+      "loss": 0.2737,
+      "step": 2395
+    },
+    {
+      "epoch": 1.2055450535601764,
+      "grad_norm": 0.17189335242974707,
+      "learning_rate": 6.656011612759056e-07,
+      "loss": 0.2671,
+      "step": 2396
+    },
+    {
+      "epoch": 1.2060491493383743,
+      "grad_norm": 0.17237484748444734,
+      "learning_rate": 6.653477367841497e-07,
+      "loss": 0.2702,
+      "step": 2397
+    },
+    {
+      "epoch": 1.2065532451165721,
+      "grad_norm": 0.17030567087724324,
+      "learning_rate": 6.65094264590596e-07,
+      "loss": 0.2688,
+      "step": 2398
+    },
+    {
+      "epoch": 1.20705734089477,
+      "grad_norm": 0.17311851732438963,
+      "learning_rate": 6.648407447683698e-07,
+      "loss": 0.2945,
+      "step": 2399
+    },
+    {
+      "epoch": 1.207561436672968,
+      "grad_norm": 0.1868284113694255,
+      "learning_rate": 6.645871773906098e-07,
+      "loss": 0.2685,
+      "step": 2400
+    },
+    {
+      "epoch": 1.2080655324511658,
+      "grad_norm": 0.1797570381187558,
+      "learning_rate": 6.643335625304687e-07,
+      "loss": 0.2778,
+      "step": 2401
+    },
+    {
+      "epoch": 1.2085696282293636,
+      "grad_norm": 0.1736947065045101,
+      "learning_rate": 6.640799002611127e-07,
+      "loss": 0.2701,
+      "step": 2402
+    },
+    {
+      "epoch": 1.2090737240075615,
+      "grad_norm": 0.16988523916524562,
+      "learning_rate": 6.638261906557219e-07,
+      "loss": 0.2716,
+      "step": 2403
+    },
+    {
+      "epoch": 1.2095778197857592,
+      "grad_norm": 0.17637575120619103,
+      "learning_rate": 6.635724337874902e-07,
+      "loss": 0.2792,
+      "step": 2404
+    },
+    {
+      "epoch": 1.210081915563957,
+      "grad_norm": 0.18457061836282274,
+      "learning_rate": 6.633186297296244e-07,
+      "loss": 0.2726,
+      "step": 2405
+    },
+    {
+      "epoch": 1.210586011342155,
+      "grad_norm": 0.1835538319807229,
+      "learning_rate": 6.630647785553456e-07,
+      "loss": 0.2778,
+      "step": 2406
+    },
+    {
+      "epoch": 1.2110901071203528,
+      "grad_norm": 0.17108814635241743,
+      "learning_rate": 6.628108803378884e-07,
+      "loss": 0.2596,
+      "step": 2407
+    },
+    {
+      "epoch": 1.2115942028985507,
+      "grad_norm": 0.17109777831078982,
+      "learning_rate": 6.625569351505008e-07,
+      "loss": 0.2799,
+      "step": 2408
+    },
+    {
+      "epoch": 1.2120982986767486,
+      "grad_norm": 0.17949433198481227,
+      "learning_rate": 6.623029430664444e-07,
+      "loss": 0.2717,
+      "step": 2409
+    },
+    {
+      "epoch": 1.2126023944549464,
+      "grad_norm": 0.18703173955181984,
+      "learning_rate": 6.620489041589942e-07,
+      "loss": 0.2734,
+      "step": 2410
+    },
+    {
+      "epoch": 1.2131064902331443,
+      "grad_norm": 0.16778996136487268,
+      "learning_rate": 6.617948185014392e-07,
+      "loss": 0.2673,
+      "step": 2411
+    },
+    {
+      "epoch": 1.2136105860113422,
+      "grad_norm": 0.17151013082627964,
+      "learning_rate": 6.615406861670811e-07,
+      "loss": 0.2848,
+      "step": 2412
+    },
+    {
+      "epoch": 1.21411468178954,
+      "grad_norm": 0.17205458561571332,
+      "learning_rate": 6.612865072292359e-07,
+      "loss": 0.2788,
+      "step": 2413
+    },
+    {
+      "epoch": 1.214618777567738,
+      "grad_norm": 0.1697154843584016,
+      "learning_rate": 6.610322817612326e-07,
+      "loss": 0.2782,
+      "step": 2414
+    },
+    {
+      "epoch": 1.2151228733459356,
+      "grad_norm": 0.17155395420115768,
+      "learning_rate": 6.607780098364133e-07,
+      "loss": 0.2757,
+      "step": 2415
+    },
+    {
+      "epoch": 1.2156269691241337,
+      "grad_norm": 0.17914770936359822,
+      "learning_rate": 6.605236915281343e-07,
+      "loss": 0.2625,
+      "step": 2416
+    },
+    {
+      "epoch": 1.2161310649023314,
+      "grad_norm": 0.17886990627397134,
+      "learning_rate": 6.602693269097646e-07,
+      "loss": 0.2826,
+      "step": 2417
+    },
+    {
+      "epoch": 1.2166351606805292,
+      "grad_norm": 0.17137302898724377,
+      "learning_rate": 6.600149160546868e-07,
+      "loss": 0.2654,
+      "step": 2418
+    },
+    {
+      "epoch": 1.2171392564587271,
+      "grad_norm": 0.16812006244582586,
+      "learning_rate": 6.597604590362972e-07,
+      "loss": 0.2693,
+      "step": 2419
+    },
+    {
+      "epoch": 1.217643352236925,
+      "grad_norm": 0.1967968966854081,
+      "learning_rate": 6.595059559280047e-07,
+      "loss": 0.2851,
+      "step": 2420
+    },
+    {
+      "epoch": 1.2181474480151229,
+      "grad_norm": 0.18024230166349509,
+      "learning_rate": 6.592514068032321e-07,
+      "loss": 0.2802,
+      "step": 2421
+    },
+    {
+      "epoch": 1.2186515437933207,
+      "grad_norm": 0.17680540350469492,
+      "learning_rate": 6.58996811735415e-07,
+      "loss": 0.267,
+      "step": 2422
+    },
+    {
+      "epoch": 1.2191556395715186,
+      "grad_norm": 0.17123047011311054,
+      "learning_rate": 6.587421707980027e-07,
+      "loss": 0.2691,
+      "step": 2423
+    },
+    {
+      "epoch": 1.2196597353497165,
+      "grad_norm": 0.1817858622939275,
+      "learning_rate": 6.584874840644575e-07,
+      "loss": 0.2695,
+      "step": 2424
+    },
+    {
+      "epoch": 1.2201638311279144,
+      "grad_norm": 0.17367383253762286,
+      "learning_rate": 6.582327516082549e-07,
+      "loss": 0.2836,
+      "step": 2425
+    },
+    {
+      "epoch": 1.2206679269061123,
+      "grad_norm": 0.17880252652432166,
+      "learning_rate": 6.579779735028836e-07,
+      "loss": 0.2733,
+      "step": 2426
+    },
+    {
+      "epoch": 1.2211720226843101,
+      "grad_norm": 0.18216013842602857,
+      "learning_rate": 6.577231498218457e-07,
+      "loss": 0.2563,
+      "step": 2427
+    },
+    {
+      "epoch": 1.2216761184625078,
+      "grad_norm": 0.17923344984805104,
+      "learning_rate": 6.574682806386559e-07,
+      "loss": 0.2727,
+      "step": 2428
+    },
+    {
+      "epoch": 1.2221802142407057,
+      "grad_norm": 0.17305807088265074,
+      "learning_rate": 6.572133660268428e-07,
+      "loss": 0.2786,
+      "step": 2429
+    },
+    {
+      "epoch": 1.2226843100189035,
+      "grad_norm": 0.1826130969086575,
+      "learning_rate": 6.569584060599475e-07,
+      "loss": 0.2639,
+      "step": 2430
+    },
+    {
+      "epoch": 1.2231884057971014,
+      "grad_norm": 0.17180169078282023,
+      "learning_rate": 6.567034008115242e-07,
+      "loss": 0.2567,
+      "step": 2431
+    },
+    {
+      "epoch": 1.2236925015752993,
+      "grad_norm": 0.17999665867139894,
+      "learning_rate": 6.564483503551406e-07,
+      "loss": 0.2868,
+      "step": 2432
+    },
+    {
+      "epoch": 1.2241965973534972,
+      "grad_norm": 0.1734485518145704,
+      "learning_rate": 6.56193254764377e-07,
+      "loss": 0.2842,
+      "step": 2433
+    },
+    {
+      "epoch": 1.224700693131695,
+      "grad_norm": 0.19524933134868758,
+      "learning_rate": 6.55938114112827e-07,
+      "loss": 0.2805,
+      "step": 2434
+    },
+    {
+      "epoch": 1.225204788909893,
+      "grad_norm": 0.3859143312648577,
+      "learning_rate": 6.556829284740972e-07,
+      "loss": 0.2793,
+      "step": 2435
+    },
+    {
+      "epoch": 1.2257088846880908,
+      "grad_norm": 0.16949921412202032,
+      "learning_rate": 6.554276979218069e-07,
+      "loss": 0.2633,
+      "step": 2436
+    },
+    {
+      "epoch": 1.2262129804662887,
+      "grad_norm": 0.17509147431622993,
+      "learning_rate": 6.551724225295885e-07,
+      "loss": 0.2585,
+      "step": 2437
+    },
+    {
+      "epoch": 1.2267170762444866,
+      "grad_norm": 0.18102430967796426,
+      "learning_rate": 6.549171023710874e-07,
+      "loss": 0.2568,
+      "step": 2438
+    },
+    {
+      "epoch": 1.2272211720226842,
+      "grad_norm": 0.17595819751957628,
+      "learning_rate": 6.54661737519962e-07,
+      "loss": 0.2678,
+      "step": 2439
+    },
+    {
+      "epoch": 1.227725267800882,
+      "grad_norm": 0.18034785174616086,
+      "learning_rate": 6.544063280498834e-07,
+      "loss": 0.286,
+      "step": 2440
+    },
+    {
+      "epoch": 1.22822936357908,
+      "grad_norm": 0.1751845129218413,
+      "learning_rate": 6.541508740345357e-07,
+      "loss": 0.2704,
+      "step": 2441
+    },
+    {
+      "epoch": 1.2287334593572778,
+      "grad_norm": 0.1707508154159352,
+      "learning_rate": 6.538953755476155e-07,
+      "loss": 0.2657,
+      "step": 2442
+    },
+    {
+      "epoch": 1.2292375551354757,
+      "grad_norm": 0.17571528214389306,
+      "learning_rate": 6.53639832662833e-07,
+      "loss": 0.2737,
+      "step": 2443
+    },
+    {
+      "epoch": 1.2297416509136736,
+      "grad_norm": 0.1767996956782402,
+      "learning_rate": 6.533842454539105e-07,
+      "loss": 0.2692,
+      "step": 2444
+    },
+    {
+      "epoch": 1.2302457466918715,
+      "grad_norm": 0.17677555995052904,
+      "learning_rate": 6.531286139945834e-07,
+      "loss": 0.2747,
+      "step": 2445
+    },
+    {
+      "epoch": 1.2307498424700694,
+      "grad_norm": 0.17418823992379287,
+      "learning_rate": 6.528729383585997e-07,
+      "loss": 0.2687,
+      "step": 2446
+    },
+    {
+      "epoch": 1.2312539382482672,
+      "grad_norm": 0.1981211293612195,
+      "learning_rate": 6.526172186197203e-07,
+      "loss": 0.2699,
+      "step": 2447
+    },
+    {
+      "epoch": 1.231758034026465,
+      "grad_norm": 0.1770440172204685,
+      "learning_rate": 6.523614548517187e-07,
+      "loss": 0.2758,
+      "step": 2448
+    },
+    {
+      "epoch": 1.232262129804663,
+      "grad_norm": 0.1733214708512733,
+      "learning_rate": 6.521056471283811e-07,
+      "loss": 0.2761,
+      "step": 2449
+    },
+    {
+      "epoch": 1.2327662255828606,
+      "grad_norm": 0.2036211297547966,
+      "learning_rate": 6.518497955235068e-07,
+      "loss": 0.268,
+      "step": 2450
+    },
+    {
+      "epoch": 1.2332703213610585,
+      "grad_norm": 0.17325758117743048,
+      "learning_rate": 6.515939001109069e-07,
+      "loss": 0.2748,
+      "step": 2451
+    },
+    {
+      "epoch": 1.2337744171392564,
+      "grad_norm": 0.17462788382443756,
+      "learning_rate": 6.513379609644062e-07,
+      "loss": 0.2782,
+      "step": 2452
+    },
+    {
+      "epoch": 1.2342785129174543,
+      "grad_norm": 0.1711968410986884,
+      "learning_rate": 6.51081978157841e-07,
+      "loss": 0.2707,
+      "step": 2453
+    },
+    {
+      "epoch": 1.2347826086956522,
+      "grad_norm": 0.1683105761384845,
+      "learning_rate": 6.50825951765061e-07,
+      "loss": 0.2773,
+      "step": 2454
+    },
+    {
+      "epoch": 1.23528670447385,
+      "grad_norm": 0.17554589173116267,
+      "learning_rate": 6.505698818599284e-07,
+      "loss": 0.2676,
+      "step": 2455
+    },
+    {
+      "epoch": 1.235790800252048,
+      "grad_norm": 0.1759694214712948,
+      "learning_rate": 6.503137685163173e-07,
+      "loss": 0.2749,
+      "step": 2456
+    },
+    {
+      "epoch": 1.2362948960302458,
+      "grad_norm": 0.17957710933654297,
+      "learning_rate": 6.500576118081155e-07,
+      "loss": 0.2762,
+      "step": 2457
+    },
+    {
+      "epoch": 1.2367989918084437,
+      "grad_norm": 0.19053173662723633,
+      "learning_rate": 6.49801411809222e-07,
+      "loss": 0.2803,
+      "step": 2458
+    },
+    {
+      "epoch": 1.2373030875866415,
+      "grad_norm": 0.17718428947532663,
+      "learning_rate": 6.495451685935494e-07,
+      "loss": 0.2729,
+      "step": 2459
+    },
+    {
+      "epoch": 1.2378071833648394,
+      "grad_norm": 0.17785327429842798,
+      "learning_rate": 6.492888822350219e-07,
+      "loss": 0.2661,
+      "step": 2460
+    },
+    {
+      "epoch": 1.238311279143037,
+      "grad_norm": 0.17300591946217833,
+      "learning_rate": 6.490325528075766e-07,
+      "loss": 0.2867,
+      "step": 2461
+    },
+    {
+      "epoch": 1.238815374921235,
+      "grad_norm": 0.16921354432191585,
+      "learning_rate": 6.487761803851631e-07,
+      "loss": 0.2711,
+      "step": 2462
+    },
+    {
+      "epoch": 1.2393194706994328,
+      "grad_norm": 0.17810885372654386,
+      "learning_rate": 6.485197650417431e-07,
+      "loss": 0.258,
+      "step": 2463
+    },
+    {
+      "epoch": 1.2398235664776307,
+      "grad_norm": 0.18103805033195008,
+      "learning_rate": 6.482633068512911e-07,
+      "loss": 0.2856,
+      "step": 2464
+    },
+    {
+      "epoch": 1.2403276622558286,
+      "grad_norm": 0.18360526541296773,
+      "learning_rate": 6.480068058877934e-07,
+      "loss": 0.2743,
+      "step": 2465
+    },
+    {
+      "epoch": 1.2408317580340265,
+      "grad_norm": 0.18884668384982423,
+      "learning_rate": 6.47750262225249e-07,
+      "loss": 0.2769,
+      "step": 2466
+    },
+    {
+      "epoch": 1.2413358538122243,
+      "grad_norm": 0.2002321224410455,
+      "learning_rate": 6.474936759376693e-07,
+      "loss": 0.2707,
+      "step": 2467
+    },
+    {
+      "epoch": 1.2418399495904222,
+      "grad_norm": 0.18325425135295045,
+      "learning_rate": 6.472370470990778e-07,
+      "loss": 0.2694,
+      "step": 2468
+    },
+    {
+      "epoch": 1.24234404536862,
+      "grad_norm": 0.1696869877874564,
+      "learning_rate": 6.469803757835102e-07,
+      "loss": 0.2747,
+      "step": 2469
+    },
+    {
+      "epoch": 1.242848141146818,
+      "grad_norm": 0.3486889386014526,
+      "learning_rate": 6.467236620650147e-07,
+      "loss": 0.2826,
+      "step": 2470
+    },
+    {
+      "epoch": 1.2433522369250158,
+      "grad_norm": 0.16956469304659022,
+      "learning_rate": 6.464669060176516e-07,
+      "loss": 0.2687,
+      "step": 2471
+    },
+    {
+      "epoch": 1.2438563327032135,
+      "grad_norm": 0.20746929532665012,
+      "learning_rate": 6.462101077154935e-07,
+      "loss": 0.2721,
+      "step": 2472
+    },
+    {
+      "epoch": 1.2443604284814114,
+      "grad_norm": 0.18257378343955985,
+      "learning_rate": 6.459532672326249e-07,
+      "loss": 0.2725,
+      "step": 2473
+    },
+    {
+      "epoch": 1.2448645242596093,
+      "grad_norm": 0.17681868108144572,
+      "learning_rate": 6.45696384643143e-07,
+      "loss": 0.2798,
+      "step": 2474
+    },
+    {
+      "epoch": 1.2453686200378071,
+      "grad_norm": 0.19498569967155788,
+      "learning_rate": 6.454394600211565e-07,
+      "loss": 0.276,
+      "step": 2475
+    },
+    {
+      "epoch": 1.245872715816005,
+      "grad_norm": 0.19909897682075145,
+      "learning_rate": 6.45182493440787e-07,
+      "loss": 0.2883,
+      "step": 2476
+    },
+    {
+      "epoch": 1.2463768115942029,
+      "grad_norm": 0.17632241811925964,
+      "learning_rate": 6.449254849761672e-07,
+      "loss": 0.2688,
+      "step": 2477
+    },
+    {
+      "epoch": 1.2468809073724008,
+      "grad_norm": 0.1762000666442149,
+      "learning_rate": 6.446684347014428e-07,
+      "loss": 0.2741,
+      "step": 2478
+    },
+    {
+      "epoch": 1.2473850031505986,
+      "grad_norm": 0.1773686838716517,
+      "learning_rate": 6.444113426907713e-07,
+      "loss": 0.2899,
+      "step": 2479
+    },
+    {
+      "epoch": 1.2478890989287965,
+      "grad_norm": 0.23756861929225664,
+      "learning_rate": 6.44154209018322e-07,
+      "loss": 0.2687,
+      "step": 2480
+    },
+    {
+      "epoch": 1.2483931947069944,
+      "grad_norm": 0.16938250561272578,
+      "learning_rate": 6.438970337582764e-07,
+      "loss": 0.2817,
+      "step": 2481
+    },
+    {
+      "epoch": 1.2488972904851923,
+      "grad_norm": 0.18157388578326444,
+      "learning_rate": 6.436398169848278e-07,
+      "loss": 0.2718,
+      "step": 2482
+    },
+    {
+      "epoch": 1.24940138626339,
+      "grad_norm": 0.19502196013884435,
+      "learning_rate": 6.43382558772182e-07,
+      "loss": 0.2768,
+      "step": 2483
+    },
+    {
+      "epoch": 1.249905482041588,
+      "grad_norm": 0.1734733146967059,
+      "learning_rate": 6.431252591945561e-07,
+      "loss": 0.2769,
+      "step": 2484
+    },
+    {
+      "epoch": 1.2504095778197857,
+      "grad_norm": 0.188235466172801,
+      "learning_rate": 6.428679183261796e-07,
+      "loss": 0.2705,
+      "step": 2485
+    },
+    {
+      "epoch": 1.2509136735979836,
+      "grad_norm": 0.19152028195774132,
+      "learning_rate": 6.426105362412935e-07,
+      "loss": 0.2732,
+      "step": 2486
+    },
+    {
+      "epoch": 1.2514177693761814,
+      "grad_norm": 0.1695103765753016,
+      "learning_rate": 6.423531130141513e-07,
+      "loss": 0.2753,
+      "step": 2487
+    },
+    {
+      "epoch": 1.2519218651543793,
+      "grad_norm": 0.17465424303832264,
+      "learning_rate": 6.420956487190177e-07,
+      "loss": 0.2744,
+      "step": 2488
+    },
+    {
+      "epoch": 1.2524259609325772,
+      "grad_norm": 0.17601676662066001,
+      "learning_rate": 6.418381434301698e-07,
+      "loss": 0.2639,
+      "step": 2489
+    },
+    {
+      "epoch": 1.252930056710775,
+      "grad_norm": 0.17576563315273216,
+      "learning_rate": 6.415805972218962e-07,
+      "loss": 0.2731,
+      "step": 2490
+    },
+    {
+      "epoch": 1.253434152488973,
+      "grad_norm": 0.1817582073444604,
+      "learning_rate": 6.413230101684972e-07,
+      "loss": 0.2915,
+      "step": 2491
+    },
+    {
+      "epoch": 1.2539382482671708,
+      "grad_norm": 0.1871857208126174,
+      "learning_rate": 6.410653823442853e-07,
+      "loss": 0.2669,
+      "step": 2492
+    },
+    {
+      "epoch": 1.2544423440453687,
+      "grad_norm": 0.17014233122771064,
+      "learning_rate": 6.408077138235843e-07,
+      "loss": 0.274,
+      "step": 2493
+    },
+    {
+      "epoch": 1.2549464398235664,
+      "grad_norm": 0.1714640417770811,
+      "learning_rate": 6.405500046807303e-07,
+      "loss": 0.2686,
+      "step": 2494
+    },
+    {
+      "epoch": 1.2554505356017645,
+      "grad_norm": 0.19243821454194698,
+      "learning_rate": 6.402922549900705e-07,
+      "loss": 0.2891,
+      "step": 2495
+    },
+    {
+      "epoch": 1.255954631379962,
+      "grad_norm": 0.17463123533718689,
+      "learning_rate": 6.400344648259644e-07,
+      "loss": 0.2699,
+      "step": 2496
+    },
+    {
+      "epoch": 1.25645872715816,
+      "grad_norm": 0.1910291136178679,
+      "learning_rate": 6.397766342627825e-07,
+      "loss": 0.2635,
+      "step": 2497
+    },
+    {
+      "epoch": 1.2569628229363579,
+      "grad_norm": 0.17138494155286757,
+      "learning_rate": 6.395187633749075e-07,
+      "loss": 0.2833,
+      "step": 2498
+    },
+    {
+      "epoch": 1.2574669187145557,
+      "grad_norm": 0.17443608504712982,
+      "learning_rate": 6.392608522367336e-07,
+      "loss": 0.2723,
+      "step": 2499
+    },
+    {
+      "epoch": 1.2579710144927536,
+      "grad_norm": 0.1699382749604669,
+      "learning_rate": 6.390029009226664e-07,
+      "loss": 0.2657,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2584751102709515,
+      "grad_norm": 0.17760734553417154,
+      "learning_rate": 6.387449095071234e-07,
+      "loss": 0.2738,
+      "step": 2501
+    },
+    {
+      "epoch": 1.2589792060491494,
+      "grad_norm": 0.17525575952040232,
+      "learning_rate": 6.384868780645335e-07,
+      "loss": 0.2703,
+      "step": 2502
+    },
+    {
+      "epoch": 1.2594833018273472,
+      "grad_norm": 0.18894681759150508,
+      "learning_rate": 6.382288066693372e-07,
+      "loss": 0.2816,
+      "step": 2503
+    },
+    {
+      "epoch": 1.2599873976055451,
+      "grad_norm": 0.18849104352320012,
+      "learning_rate": 6.379706953959865e-07,
+      "loss": 0.2766,
+      "step": 2504
+    },
+    {
+      "epoch": 1.2604914933837428,
+      "grad_norm": 0.17320664007502315,
+      "learning_rate": 6.377125443189446e-07,
+      "loss": 0.2558,
+      "step": 2505
+    },
+    {
+      "epoch": 1.2609955891619409,
+      "grad_norm": 0.1981593202287322,
+      "learning_rate": 6.37454353512687e-07,
+      "loss": 0.2854,
+      "step": 2506
+    },
+    {
+      "epoch": 1.2614996849401385,
+      "grad_norm": 0.1674126260844481,
+      "learning_rate": 6.371961230516997e-07,
+      "loss": 0.2687,
+      "step": 2507
+    },
+    {
+      "epoch": 1.2620037807183364,
+      "grad_norm": 0.17667010185551682,
+      "learning_rate": 6.36937853010481e-07,
+      "loss": 0.2699,
+      "step": 2508
+    },
+    {
+      "epoch": 1.2625078764965343,
+      "grad_norm": 0.17818670333223607,
+      "learning_rate": 6.366795434635398e-07,
+      "loss": 0.2687,
+      "step": 2509
+    },
+    {
+      "epoch": 1.2630119722747322,
+      "grad_norm": 0.17694309827192736,
+      "learning_rate": 6.364211944853971e-07,
+      "loss": 0.2635,
+      "step": 2510
+    },
+    {
+      "epoch": 1.26351606805293,
+      "grad_norm": 0.17785156941257393,
+      "learning_rate": 6.361628061505849e-07,
+      "loss": 0.2831,
+      "step": 2511
+    },
+    {
+      "epoch": 1.264020163831128,
+      "grad_norm": 0.17826472857233217,
+      "learning_rate": 6.359043785336467e-07,
+      "loss": 0.2731,
+      "step": 2512
+    },
+    {
+      "epoch": 1.2645242596093258,
+      "grad_norm": 0.18356140095576604,
+      "learning_rate": 6.356459117091369e-07,
+      "loss": 0.2601,
+      "step": 2513
+    },
+    {
+      "epoch": 1.2650283553875237,
+      "grad_norm": 0.2119595973428602,
+      "learning_rate": 6.353874057516222e-07,
+      "loss": 0.2809,
+      "step": 2514
+    },
+    {
+      "epoch": 1.2655324511657216,
+      "grad_norm": 0.16952440646698724,
+      "learning_rate": 6.351288607356793e-07,
+      "loss": 0.2731,
+      "step": 2515
+    },
+    {
+      "epoch": 1.2660365469439194,
+      "grad_norm": 0.17460234803900188,
+      "learning_rate": 6.348702767358974e-07,
+      "loss": 0.2797,
+      "step": 2516
+    },
+    {
+      "epoch": 1.2665406427221173,
+      "grad_norm": 0.17160497843474512,
+      "learning_rate": 6.34611653826876e-07,
+      "loss": 0.2698,
+      "step": 2517
+    },
+    {
+      "epoch": 1.267044738500315,
+      "grad_norm": 0.18778464548766757,
+      "learning_rate": 6.343529920832263e-07,
+      "loss": 0.2573,
+      "step": 2518
+    },
+    {
+      "epoch": 1.2675488342785128,
+      "grad_norm": 0.172443493887068,
+      "learning_rate": 6.340942915795708e-07,
+      "loss": 0.2748,
+      "step": 2519
+    },
+    {
+      "epoch": 1.2680529300567107,
+      "grad_norm": 0.17168130332053846,
+      "learning_rate": 6.338355523905427e-07,
+      "loss": 0.2625,
+      "step": 2520
+    },
+    {
+      "epoch": 1.2685570258349086,
+      "grad_norm": 0.17567263307465228,
+      "learning_rate": 6.335767745907869e-07,
+      "loss": 0.2859,
+      "step": 2521
+    },
+    {
+      "epoch": 1.2690611216131065,
+      "grad_norm": 0.20501744023706306,
+      "learning_rate": 6.33317958254959e-07,
+      "loss": 0.2773,
+      "step": 2522
+    },
+    {
+      "epoch": 1.2695652173913043,
+      "grad_norm": 0.1693368131106986,
+      "learning_rate": 6.33059103457726e-07,
+      "loss": 0.2653,
+      "step": 2523
+    },
+    {
+      "epoch": 1.2700693131695022,
+      "grad_norm": 0.18006664625555702,
+      "learning_rate": 6.32800210273766e-07,
+      "loss": 0.2645,
+      "step": 2524
+    },
+    {
+      "epoch": 1.2705734089477,
+      "grad_norm": 0.17456080347662142,
+      "learning_rate": 6.32541278777768e-07,
+      "loss": 0.2699,
+      "step": 2525
+    },
+    {
+      "epoch": 1.271077504725898,
+      "grad_norm": 0.16574198775311022,
+      "learning_rate": 6.322823090444322e-07,
+      "loss": 0.2662,
+      "step": 2526
+    },
+    {
+      "epoch": 1.2715816005040959,
+      "grad_norm": 0.18598160934075628,
+      "learning_rate": 6.320233011484696e-07,
+      "loss": 0.2828,
+      "step": 2527
+    },
+    {
+      "epoch": 1.2720856962822937,
+      "grad_norm": 0.1711774666753222,
+      "learning_rate": 6.317642551646024e-07,
+      "loss": 0.2803,
+      "step": 2528
+    },
+    {
+      "epoch": 1.2725897920604914,
+      "grad_norm": 0.17441523118583016,
+      "learning_rate": 6.315051711675639e-07,
+      "loss": 0.2864,
+      "step": 2529
+    },
+    {
+      "epoch": 1.2730938878386895,
+      "grad_norm": 0.18168000027383474,
+      "learning_rate": 6.312460492320981e-07,
+      "loss": 0.2786,
+      "step": 2530
+    },
+    {
+      "epoch": 1.2735979836168871,
+      "grad_norm": 0.18114157608848858,
+      "learning_rate": 6.309868894329602e-07,
+      "loss": 0.2805,
+      "step": 2531
+    },
+    {
+      "epoch": 1.274102079395085,
+      "grad_norm": 0.18530369106217595,
+      "learning_rate": 6.30727691844916e-07,
+      "loss": 0.2804,
+      "step": 2532
+    },
+    {
+      "epoch": 1.274606175173283,
+      "grad_norm": 0.1927697807229042,
+      "learning_rate": 6.304684565427427e-07,
+      "loss": 0.2667,
+      "step": 2533
+    },
+    {
+      "epoch": 1.2751102709514808,
+      "grad_norm": 0.17177026835602924,
+      "learning_rate": 6.302091836012278e-07,
+      "loss": 0.2633,
+      "step": 2534
+    },
+    {
+      "epoch": 1.2756143667296787,
+      "grad_norm": 0.20642706373970884,
+      "learning_rate": 6.299498730951699e-07,
+      "loss": 0.2878,
+      "step": 2535
+    },
+    {
+      "epoch": 1.2761184625078765,
+      "grad_norm": 0.17594445837282194,
+      "learning_rate": 6.296905250993787e-07,
+      "loss": 0.2946,
+      "step": 2536
+    },
+    {
+      "epoch": 1.2766225582860744,
+      "grad_norm": 0.1762144552346574,
+      "learning_rate": 6.294311396886745e-07,
+      "loss": 0.2915,
+      "step": 2537
+    },
+    {
+      "epoch": 1.2771266540642723,
+      "grad_norm": 0.17045816209341122,
+      "learning_rate": 6.291717169378881e-07,
+      "loss": 0.2782,
+      "step": 2538
+    },
+    {
+      "epoch": 1.2776307498424702,
+      "grad_norm": 0.18383362547518797,
+      "learning_rate": 6.289122569218615e-07,
+      "loss": 0.2748,
+      "step": 2539
+    },
+    {
+      "epoch": 1.2781348456206678,
+      "grad_norm": 0.1807772669969172,
+      "learning_rate": 6.286527597154475e-07,
+      "loss": 0.2885,
+      "step": 2540
+    },
+    {
+      "epoch": 1.278638941398866,
+      "grad_norm": 0.19302775133426964,
+      "learning_rate": 6.283932253935094e-07,
+      "loss": 0.2886,
+      "step": 2541
+    },
+    {
+      "epoch": 1.2791430371770636,
+      "grad_norm": 0.1888729384221993,
+      "learning_rate": 6.28133654030921e-07,
+      "loss": 0.2807,
+      "step": 2542
+    },
+    {
+      "epoch": 1.2796471329552614,
+      "grad_norm": 0.17228549927667094,
+      "learning_rate": 6.278740457025671e-07,
+      "loss": 0.2754,
+      "step": 2543
+    },
+    {
+      "epoch": 1.2801512287334593,
+      "grad_norm": 0.1729966859664312,
+      "learning_rate": 6.276144004833432e-07,
+      "loss": 0.2828,
+      "step": 2544
+    },
+    {
+      "epoch": 1.2806553245116572,
+      "grad_norm": 0.17500523408983926,
+      "learning_rate": 6.273547184481554e-07,
+      "loss": 0.2679,
+      "step": 2545
+    },
+    {
+      "epoch": 1.281159420289855,
+      "grad_norm": 0.17974221115337133,
+      "learning_rate": 6.270949996719202e-07,
+      "loss": 0.28,
+      "step": 2546
+    },
+    {
+      "epoch": 1.281663516068053,
+      "grad_norm": 0.18105222606013233,
+      "learning_rate": 6.268352442295648e-07,
+      "loss": 0.2794,
+      "step": 2547
+    },
+    {
+      "epoch": 1.2821676118462508,
+      "grad_norm": 0.18180911542908265,
+      "learning_rate": 6.265754521960272e-07,
+      "loss": 0.2815,
+      "step": 2548
+    },
+    {
+      "epoch": 1.2826717076244487,
+      "grad_norm": 0.19054982521679337,
+      "learning_rate": 6.263156236462557e-07,
+      "loss": 0.2603,
+      "step": 2549
+    },
+    {
+      "epoch": 1.2831758034026466,
+      "grad_norm": 0.17388626076521252,
+      "learning_rate": 6.260557586552094e-07,
+      "loss": 0.2787,
+      "step": 2550
+    },
+    {
+      "epoch": 1.2836798991808442,
+      "grad_norm": 0.18270654576376585,
+      "learning_rate": 6.257958572978573e-07,
+      "loss": 0.2664,
+      "step": 2551
+    },
+    {
+      "epoch": 1.2841839949590423,
+      "grad_norm": 0.1727969110053315,
+      "learning_rate": 6.255359196491799e-07,
+      "loss": 0.2669,
+      "step": 2552
+    },
+    {
+      "epoch": 1.28468809073724,
+      "grad_norm": 0.17719931615423382,
+      "learning_rate": 6.252759457841672e-07,
+      "loss": 0.2806,
+      "step": 2553
+    },
+    {
+      "epoch": 1.2851921865154379,
+      "grad_norm": 0.16515633828315085,
+      "learning_rate": 6.250159357778202e-07,
+      "loss": 0.2804,
+      "step": 2554
+    },
+    {
+      "epoch": 1.2856962822936358,
+      "grad_norm": 0.18820483320106177,
+      "learning_rate": 6.247558897051503e-07,
+      "loss": 0.2965,
+      "step": 2555
+    },
+    {
+      "epoch": 1.2862003780718336,
+      "grad_norm": 0.17604374414885252,
+      "learning_rate": 6.244958076411789e-07,
+      "loss": 0.2622,
+      "step": 2556
+    },
+    {
+      "epoch": 1.2867044738500315,
+      "grad_norm": 0.1696182132757248,
+      "learning_rate": 6.242356896609383e-07,
+      "loss": 0.2742,
+      "step": 2557
+    },
+    {
+      "epoch": 1.2872085696282294,
+      "grad_norm": 0.1737771138671047,
+      "learning_rate": 6.239755358394707e-07,
+      "loss": 0.2667,
+      "step": 2558
+    },
+    {
+      "epoch": 1.2877126654064273,
+      "grad_norm": 0.1801598415185936,
+      "learning_rate": 6.237153462518291e-07,
+      "loss": 0.2654,
+      "step": 2559
+    },
+    {
+      "epoch": 1.2882167611846251,
+      "grad_norm": 0.1756137146667004,
+      "learning_rate": 6.234551209730765e-07,
+      "loss": 0.2947,
+      "step": 2560
+    },
+    {
+      "epoch": 1.288720856962823,
+      "grad_norm": 0.17233304314756842,
+      "learning_rate": 6.231948600782863e-07,
+      "loss": 0.2776,
+      "step": 2561
+    },
+    {
+      "epoch": 1.2892249527410207,
+      "grad_norm": 0.18074095930111286,
+      "learning_rate": 6.229345636425421e-07,
+      "loss": 0.2778,
+      "step": 2562
+    },
+    {
+      "epoch": 1.2897290485192188,
+      "grad_norm": 0.17475122322002845,
+      "learning_rate": 6.226742317409378e-07,
+      "loss": 0.2907,
+      "step": 2563
+    },
+    {
+      "epoch": 1.2902331442974164,
+      "grad_norm": 0.17347744444476335,
+      "learning_rate": 6.224138644485775e-07,
+      "loss": 0.2737,
+      "step": 2564
+    },
+    {
+      "epoch": 1.2907372400756143,
+      "grad_norm": 0.17186378287100884,
+      "learning_rate": 6.221534618405757e-07,
+      "loss": 0.277,
+      "step": 2565
+    },
+    {
+      "epoch": 1.2912413358538122,
+      "grad_norm": 0.1705178171148712,
+      "learning_rate": 6.218930239920568e-07,
+      "loss": 0.2685,
+      "step": 2566
+    },
+    {
+      "epoch": 1.29174543163201,
+      "grad_norm": 0.1786888372127578,
+      "learning_rate": 6.216325509781556e-07,
+      "loss": 0.2785,
+      "step": 2567
+    },
+    {
+      "epoch": 1.292249527410208,
+      "grad_norm": 0.17449611257529518,
+      "learning_rate": 6.213720428740168e-07,
+      "loss": 0.2653,
+      "step": 2568
+    },
+    {
+      "epoch": 1.2927536231884058,
+      "grad_norm": 0.168813740260223,
+      "learning_rate": 6.211114997547956e-07,
+      "loss": 0.2629,
+      "step": 2569
+    },
+    {
+      "epoch": 1.2932577189666037,
+      "grad_norm": 0.16790027679595843,
+      "learning_rate": 6.208509216956572e-07,
+      "loss": 0.2706,
+      "step": 2570
+    },
+    {
+      "epoch": 1.2937618147448016,
+      "grad_norm": 0.17099748648486032,
+      "learning_rate": 6.205903087717761e-07,
+      "loss": 0.2687,
+      "step": 2571
+    },
+    {
+      "epoch": 1.2942659105229994,
+      "grad_norm": 0.18162517160448968,
+      "learning_rate": 6.203296610583382e-07,
+      "loss": 0.2747,
+      "step": 2572
+    },
+    {
+      "epoch": 1.294770006301197,
+      "grad_norm": 0.17063028798101706,
+      "learning_rate": 6.200689786305383e-07,
+      "loss": 0.2761,
+      "step": 2573
+    },
+    {
+      "epoch": 1.2952741020793952,
+      "grad_norm": 0.17398933181018333,
+      "learning_rate": 6.19808261563582e-07,
+      "loss": 0.283,
+      "step": 2574
+    },
+    {
+      "epoch": 1.2957781978575929,
+      "grad_norm": 0.16776550305812016,
+      "learning_rate": 6.195475099326843e-07,
+      "loss": 0.2617,
+      "step": 2575
+    },
+    {
+      "epoch": 1.2962822936357907,
+      "grad_norm": 0.1747053664007132,
+      "learning_rate": 6.192867238130708e-07,
+      "loss": 0.2774,
+      "step": 2576
+    },
+    {
+      "epoch": 1.2967863894139886,
+      "grad_norm": 0.18115021944031587,
+      "learning_rate": 6.190259032799761e-07,
+      "loss": 0.2751,
+      "step": 2577
+    },
+    {
+      "epoch": 1.2972904851921865,
+      "grad_norm": 0.1790395768010079,
+      "learning_rate": 6.187650484086459e-07,
+      "loss": 0.29,
+      "step": 2578
+    },
+    {
+      "epoch": 1.2977945809703844,
+      "grad_norm": 0.1719106072500581,
+      "learning_rate": 6.185041592743348e-07,
+      "loss": 0.2729,
+      "step": 2579
+    },
+    {
+      "epoch": 1.2982986767485822,
+      "grad_norm": 0.17444897223126585,
+      "learning_rate": 6.182432359523079e-07,
+      "loss": 0.2789,
+      "step": 2580
+    },
+    {
+      "epoch": 1.2988027725267801,
+      "grad_norm": 0.1766242163102816,
+      "learning_rate": 6.179822785178398e-07,
+      "loss": 0.2583,
+      "step": 2581
+    },
+    {
+      "epoch": 1.299306868304978,
+      "grad_norm": 0.18162411025003278,
+      "learning_rate": 6.177212870462152e-07,
+      "loss": 0.2686,
+      "step": 2582
+    },
+    {
+      "epoch": 1.2998109640831759,
+      "grad_norm": 0.16955702326217229,
+      "learning_rate": 6.174602616127287e-07,
+      "loss": 0.2736,
+      "step": 2583
+    },
+    {
+      "epoch": 1.3003150598613735,
+      "grad_norm": 0.173479162516625,
+      "learning_rate": 6.171992022926841e-07,
+      "loss": 0.2738,
+      "step": 2584
+    },
+    {
+      "epoch": 1.3008191556395716,
+      "grad_norm": 0.17139404620083853,
+      "learning_rate": 6.16938109161396e-07,
+      "loss": 0.2757,
+      "step": 2585
+    },
+    {
+      "epoch": 1.3013232514177693,
+      "grad_norm": 0.17095006127373458,
+      "learning_rate": 6.166769822941877e-07,
+      "loss": 0.2748,
+      "step": 2586
+    },
+    {
+      "epoch": 1.3018273471959672,
+      "grad_norm": 0.17350708973766807,
+      "learning_rate": 6.164158217663926e-07,
+      "loss": 0.2729,
+      "step": 2587
+    },
+    {
+      "epoch": 1.3018273471959672,
+      "eval_loss": 0.3091413378715515,
+      "eval_runtime": 17.1268,
+      "eval_samples_per_second": 49.922,
+      "eval_steps_per_second": 1.051,
+      "step": 2587
+    },
+    {
+      "epoch": 1.302331442974165,
+      "grad_norm": 0.19869258969365142,
+      "learning_rate": 6.161546276533542e-07,
+      "loss": 0.2881,
+      "step": 2588
+    },
+    {
+      "epoch": 1.302835538752363,
+      "grad_norm": 0.1703478811098866,
+      "learning_rate": 6.158934000304251e-07,
+      "loss": 0.2701,
+      "step": 2589
+    },
+    {
+      "epoch": 1.3033396345305608,
+      "grad_norm": 0.1696150699647383,
+      "learning_rate": 6.156321389729682e-07,
+      "loss": 0.2839,
+      "step": 2590
+    },
+    {
+      "epoch": 1.3038437303087587,
+      "grad_norm": 0.1738118040743564,
+      "learning_rate": 6.153708445563555e-07,
+      "loss": 0.2852,
+      "step": 2591
+    },
+    {
+      "epoch": 1.3043478260869565,
+      "grad_norm": 0.1767377470005941,
+      "learning_rate": 6.151095168559688e-07,
+      "loss": 0.2769,
+      "step": 2592
+    },
+    {
+      "epoch": 1.3048519218651544,
+      "grad_norm": 0.17056286239729945,
+      "learning_rate": 6.148481559471995e-07,
+      "loss": 0.2634,
+      "step": 2593
+    },
+    {
+      "epoch": 1.3053560176433523,
+      "grad_norm": 0.1964952722351761,
+      "learning_rate": 6.145867619054487e-07,
+      "loss": 0.2821,
+      "step": 2594
+    },
+    {
+      "epoch": 1.3058601134215502,
+      "grad_norm": 0.17002975410785548,
+      "learning_rate": 6.143253348061271e-07,
+      "loss": 0.2751,
+      "step": 2595
+    },
+    {
+      "epoch": 1.306364209199748,
+      "grad_norm": 0.17294236211348307,
+      "learning_rate": 6.140638747246543e-07,
+      "loss": 0.2777,
+      "step": 2596
+    },
+    {
+      "epoch": 1.3068683049779457,
+      "grad_norm": 0.18702273486347906,
+      "learning_rate": 6.138023817364603e-07,
+      "loss": 0.2817,
+      "step": 2597
+    },
+    {
+      "epoch": 1.3073724007561438,
+      "grad_norm": 0.18257740617721263,
+      "learning_rate": 6.135408559169842e-07,
+      "loss": 0.2869,
+      "step": 2598
+    },
+    {
+      "epoch": 1.3078764965343415,
+      "grad_norm": 0.1679551881506779,
+      "learning_rate": 6.132792973416744e-07,
+      "loss": 0.2671,
+      "step": 2599
+    },
+    {
+      "epoch": 1.3083805923125393,
+      "grad_norm": 0.1754134210625823,
+      "learning_rate": 6.130177060859894e-07,
+      "loss": 0.302,
+      "step": 2600
+    },
+    {
+      "epoch": 1.3088846880907372,
+      "grad_norm": 0.1772876906689951,
+      "learning_rate": 6.12756082225396e-07,
+      "loss": 0.2783,
+      "step": 2601
+    },
+    {
+      "epoch": 1.309388783868935,
+      "grad_norm": 0.20064464760978862,
+      "learning_rate": 6.124944258353714e-07,
+      "loss": 0.2839,
+      "step": 2602
+    },
+    {
+      "epoch": 1.309892879647133,
+      "grad_norm": 0.18138531276139028,
+      "learning_rate": 6.122327369914018e-07,
+      "loss": 0.2928,
+      "step": 2603
+    },
+    {
+      "epoch": 1.3103969754253308,
+      "grad_norm": 0.17693406432408068,
+      "learning_rate": 6.119710157689828e-07,
+      "loss": 0.264,
+      "step": 2604
+    },
+    {
+      "epoch": 1.3109010712035287,
+      "grad_norm": 0.17240536698687978,
+      "learning_rate": 6.117092622436194e-07,
+      "loss": 0.2803,
+      "step": 2605
+    },
+    {
+      "epoch": 1.3114051669817266,
+      "grad_norm": 0.17830285202833615,
+      "learning_rate": 6.11447476490826e-07,
+      "loss": 0.271,
+      "step": 2606
+    },
+    {
+      "epoch": 1.3119092627599245,
+      "grad_norm": 0.18109922157734024,
+      "learning_rate": 6.111856585861259e-07,
+      "loss": 0.2673,
+      "step": 2607
+    },
+    {
+      "epoch": 1.3124133585381221,
+      "grad_norm": 0.17133961085931276,
+      "learning_rate": 6.10923808605052e-07,
+      "loss": 0.2726,
+      "step": 2608
+    },
+    {
+      "epoch": 1.3129174543163202,
+      "grad_norm": 0.17649723679339357,
+      "learning_rate": 6.106619266231467e-07,
+      "loss": 0.2674,
+      "step": 2609
+    },
+    {
+      "epoch": 1.3134215500945179,
+      "grad_norm": 0.17184213041429233,
+      "learning_rate": 6.104000127159608e-07,
+      "loss": 0.279,
+      "step": 2610
+    },
+    {
+      "epoch": 1.3139256458727158,
+      "grad_norm": 0.17468087267169044,
+      "learning_rate": 6.101380669590551e-07,
+      "loss": 0.279,
+      "step": 2611
+    },
+    {
+      "epoch": 1.3144297416509136,
+      "grad_norm": 0.17337912043291634,
+      "learning_rate": 6.098760894279995e-07,
+      "loss": 0.2685,
+      "step": 2612
+    },
+    {
+      "epoch": 1.3149338374291115,
+      "grad_norm": 0.18366133473396995,
+      "learning_rate": 6.096140801983727e-07,
+      "loss": 0.2711,
+      "step": 2613
+    },
+    {
+      "epoch": 1.3154379332073094,
+      "grad_norm": 0.17222644312128002,
+      "learning_rate": 6.093520393457627e-07,
+      "loss": 0.2631,
+      "step": 2614
+    },
+    {
+      "epoch": 1.3159420289855073,
+      "grad_norm": 0.17738930991704355,
+      "learning_rate": 6.09089966945767e-07,
+      "loss": 0.2743,
+      "step": 2615
+    },
+    {
+      "epoch": 1.3164461247637051,
+      "grad_norm": 0.17185117857715593,
+      "learning_rate": 6.088278630739915e-07,
+      "loss": 0.2772,
+      "step": 2616
+    },
+    {
+      "epoch": 1.316950220541903,
+      "grad_norm": 0.16984493121172142,
+      "learning_rate": 6.085657278060515e-07,
+      "loss": 0.2656,
+      "step": 2617
+    },
+    {
+      "epoch": 1.317454316320101,
+      "grad_norm": 0.17192888657461453,
+      "learning_rate": 6.083035612175716e-07,
+      "loss": 0.277,
+      "step": 2618
+    },
+    {
+      "epoch": 1.3179584120982986,
+      "grad_norm": 0.18354467561304685,
+      "learning_rate": 6.080413633841853e-07,
+      "loss": 0.2701,
+      "step": 2619
+    },
+    {
+      "epoch": 1.3184625078764967,
+      "grad_norm": 0.17503373460612875,
+      "learning_rate": 6.077791343815349e-07,
+      "loss": 0.2825,
+      "step": 2620
+    },
+    {
+      "epoch": 1.3189666036546943,
+      "grad_norm": 0.1706123234531778,
+      "learning_rate": 6.075168742852718e-07,
+      "loss": 0.2662,
+      "step": 2621
+    },
+    {
+      "epoch": 1.3194706994328922,
+      "grad_norm": 0.17412767194593945,
+      "learning_rate": 6.072545831710567e-07,
+      "loss": 0.2722,
+      "step": 2622
+    },
+    {
+      "epoch": 1.31997479521109,
+      "grad_norm": 0.17240484877752552,
+      "learning_rate": 6.069922611145587e-07,
+      "loss": 0.2686,
+      "step": 2623
+    },
+    {
+      "epoch": 1.320478890989288,
+      "grad_norm": 0.16827994239810876,
+      "learning_rate": 6.06729908191456e-07,
+      "loss": 0.2641,
+      "step": 2624
+    },
+    {
+      "epoch": 1.3209829867674858,
+      "grad_norm": 0.17930969669376187,
+      "learning_rate": 6.064675244774362e-07,
+      "loss": 0.2852,
+      "step": 2625
+    },
+    {
+      "epoch": 1.3214870825456837,
+      "grad_norm": 0.18006157470320514,
+      "learning_rate": 6.062051100481949e-07,
+      "loss": 0.2758,
+      "step": 2626
+    },
+    {
+      "epoch": 1.3219911783238816,
+      "grad_norm": 0.1727448964101327,
+      "learning_rate": 6.059426649794374e-07,
+      "loss": 0.2714,
+      "step": 2627
+    },
+    {
+      "epoch": 1.3224952741020795,
+      "grad_norm": 0.17424563952604652,
+      "learning_rate": 6.056801893468773e-07,
+      "loss": 0.2787,
+      "step": 2628
+    },
+    {
+      "epoch": 1.3229993698802773,
+      "grad_norm": 0.1731354768377102,
+      "learning_rate": 6.054176832262371e-07,
+      "loss": 0.2776,
+      "step": 2629
+    },
+    {
+      "epoch": 1.323503465658475,
+      "grad_norm": 0.1804386379598718,
+      "learning_rate": 6.051551466932485e-07,
+      "loss": 0.2651,
+      "step": 2630
+    },
+    {
+      "epoch": 1.324007561436673,
+      "grad_norm": 0.17399836046155562,
+      "learning_rate": 6.048925798236512e-07,
+      "loss": 0.2702,
+      "step": 2631
+    },
+    {
+      "epoch": 1.3245116572148707,
+      "grad_norm": 0.43353712827340424,
+      "learning_rate": 6.046299826931946e-07,
+      "loss": 0.2753,
+      "step": 2632
+    },
+    {
+      "epoch": 1.3250157529930686,
+      "grad_norm": 0.17183489960437298,
+      "learning_rate": 6.043673553776361e-07,
+      "loss": 0.2857,
+      "step": 2633
+    },
+    {
+      "epoch": 1.3255198487712665,
+      "grad_norm": 0.17519525464739322,
+      "learning_rate": 6.041046979527422e-07,
+      "loss": 0.2724,
+      "step": 2634
+    },
+    {
+      "epoch": 1.3260239445494644,
+      "grad_norm": 0.18357488474151853,
+      "learning_rate": 6.038420104942877e-07,
+      "loss": 0.2662,
+      "step": 2635
+    },
+    {
+      "epoch": 1.3265280403276623,
+      "grad_norm": 0.1859694085856373,
+      "learning_rate": 6.035792930780565e-07,
+      "loss": 0.2615,
+      "step": 2636
+    },
+    {
+      "epoch": 1.3270321361058601,
+      "grad_norm": 0.17450754260587856,
+      "learning_rate": 6.033165457798408e-07,
+      "loss": 0.2754,
+      "step": 2637
+    },
+    {
+      "epoch": 1.327536231884058,
+      "grad_norm": 0.1668193821161752,
+      "learning_rate": 6.030537686754419e-07,
+      "loss": 0.2698,
+      "step": 2638
+    },
+    {
+      "epoch": 1.3280403276622559,
+      "grad_norm": 0.1728884240425731,
+      "learning_rate": 6.027909618406689e-07,
+      "loss": 0.275,
+      "step": 2639
+    },
+    {
+      "epoch": 1.3285444234404538,
+      "grad_norm": 0.17398093656070354,
+      "learning_rate": 6.025281253513404e-07,
+      "loss": 0.2627,
+      "step": 2640
+    },
+    {
+      "epoch": 1.3290485192186514,
+      "grad_norm": 0.203938933296385,
+      "learning_rate": 6.022652592832827e-07,
+      "loss": 0.2782,
+      "step": 2641
+    },
+    {
+      "epoch": 1.3295526149968495,
+      "grad_norm": 0.18386883362802814,
+      "learning_rate": 6.02002363712331e-07,
+      "loss": 0.2748,
+      "step": 2642
+    },
+    {
+      "epoch": 1.3300567107750472,
+      "grad_norm": 0.17575912101228022,
+      "learning_rate": 6.017394387143294e-07,
+      "loss": 0.2752,
+      "step": 2643
+    },
+    {
+      "epoch": 1.330560806553245,
+      "grad_norm": 0.18067748484917678,
+      "learning_rate": 6.0147648436513e-07,
+      "loss": 0.2815,
+      "step": 2644
+    },
+    {
+      "epoch": 1.331064902331443,
+      "grad_norm": 0.16917344843892523,
+      "learning_rate": 6.012135007405936e-07,
+      "loss": 0.2648,
+      "step": 2645
+    },
+    {
+      "epoch": 1.3315689981096408,
+      "grad_norm": 0.17412949415096388,
+      "learning_rate": 6.009504879165891e-07,
+      "loss": 0.2754,
+      "step": 2646
+    },
+    {
+      "epoch": 1.3320730938878387,
+      "grad_norm": 0.18047299736884104,
+      "learning_rate": 6.006874459689942e-07,
+      "loss": 0.2694,
+      "step": 2647
+    },
+    {
+      "epoch": 1.3325771896660366,
+      "grad_norm": 0.19127600843343956,
+      "learning_rate": 6.004243749736947e-07,
+      "loss": 0.2705,
+      "step": 2648
+    },
+    {
+      "epoch": 1.3330812854442344,
+      "grad_norm": 0.17608841608371475,
+      "learning_rate": 6.001612750065853e-07,
+      "loss": 0.2797,
+      "step": 2649
+    },
+    {
+      "epoch": 1.3335853812224323,
+      "grad_norm": 0.17931850245102035,
+      "learning_rate": 5.998981461435685e-07,
+      "loss": 0.2937,
+      "step": 2650
+    },
+    {
+      "epoch": 1.3340894770006302,
+      "grad_norm": 0.18450380693233284,
+      "learning_rate": 5.996349884605552e-07,
+      "loss": 0.2702,
+      "step": 2651
+    },
+    {
+      "epoch": 1.3345935727788278,
+      "grad_norm": 0.17087115966825997,
+      "learning_rate": 5.993718020334652e-07,
+      "loss": 0.2762,
+      "step": 2652
+    },
+    {
+      "epoch": 1.335097668557026,
+      "grad_norm": 0.1718079561805658,
+      "learning_rate": 5.991085869382258e-07,
+      "loss": 0.2715,
+      "step": 2653
+    },
+    {
+      "epoch": 1.3356017643352236,
+      "grad_norm": 0.18602810395230512,
+      "learning_rate": 5.988453432507729e-07,
+      "loss": 0.2744,
+      "step": 2654
+    },
+    {
+      "epoch": 1.3361058601134215,
+      "grad_norm": 0.18389684807636075,
+      "learning_rate": 5.985820710470509e-07,
+      "loss": 0.2808,
+      "step": 2655
+    },
+    {
+      "epoch": 1.3366099558916194,
+      "grad_norm": 0.19592914957135987,
+      "learning_rate": 5.983187704030123e-07,
+      "loss": 0.2676,
+      "step": 2656
+    },
+    {
+      "epoch": 1.3371140516698172,
+      "grad_norm": 0.18190367471809632,
+      "learning_rate": 5.980554413946172e-07,
+      "loss": 0.2938,
+      "step": 2657
+    },
+    {
+      "epoch": 1.337618147448015,
+      "grad_norm": 0.1717978516962743,
+      "learning_rate": 5.977920840978346e-07,
+      "loss": 0.2865,
+      "step": 2658
+    },
+    {
+      "epoch": 1.338122243226213,
+      "grad_norm": 0.1860456269002095,
+      "learning_rate": 5.975286985886418e-07,
+      "loss": 0.2763,
+      "step": 2659
+    },
+    {
+      "epoch": 1.3386263390044109,
+      "grad_norm": 0.17913391357600741,
+      "learning_rate": 5.972652849430235e-07,
+      "loss": 0.2702,
+      "step": 2660
+    },
+    {
+      "epoch": 1.3391304347826087,
+      "grad_norm": 0.17254123661223592,
+      "learning_rate": 5.97001843236973e-07,
+      "loss": 0.2711,
+      "step": 2661
+    },
+    {
+      "epoch": 1.3396345305608066,
+      "grad_norm": 0.17083369763214978,
+      "learning_rate": 5.967383735464916e-07,
+      "loss": 0.2582,
+      "step": 2662
+    },
+    {
+      "epoch": 1.3401386263390045,
+      "grad_norm": 0.17895529015768472,
+      "learning_rate": 5.964748759475887e-07,
+      "loss": 0.2818,
+      "step": 2663
+    },
+    {
+      "epoch": 1.3406427221172024,
+      "grad_norm": 0.17751494439480603,
+      "learning_rate": 5.962113505162818e-07,
+      "loss": 0.2741,
+      "step": 2664
+    },
+    {
+      "epoch": 1.3411468178954,
+      "grad_norm": 0.1703756144111887,
+      "learning_rate": 5.959477973285961e-07,
+      "loss": 0.2643,
+      "step": 2665
+    },
+    {
+      "epoch": 1.3416509136735981,
+      "grad_norm": 0.16962499032704564,
+      "learning_rate": 5.956842164605654e-07,
+      "loss": 0.2632,
+      "step": 2666
+    },
+    {
+      "epoch": 1.3421550094517958,
+      "grad_norm": 0.16795341629826513,
+      "learning_rate": 5.954206079882311e-07,
+      "loss": 0.274,
+      "step": 2667
+    },
+    {
+      "epoch": 1.3426591052299937,
+      "grad_norm": 0.17092016847119307,
+      "learning_rate": 5.951569719876421e-07,
+      "loss": 0.2778,
+      "step": 2668
+    },
+    {
+      "epoch": 1.3431632010081915,
+      "grad_norm": 0.1790697008564231,
+      "learning_rate": 5.948933085348564e-07,
+      "loss": 0.2677,
+      "step": 2669
+    },
+    {
+      "epoch": 1.3436672967863894,
+      "grad_norm": 0.16930642394721995,
+      "learning_rate": 5.94629617705939e-07,
+      "loss": 0.2744,
+      "step": 2670
+    },
+    {
+      "epoch": 1.3441713925645873,
+      "grad_norm": 0.17294047568014345,
+      "learning_rate": 5.943658995769631e-07,
+      "loss": 0.2542,
+      "step": 2671
+    },
+    {
+      "epoch": 1.3446754883427852,
+      "grad_norm": 0.1767866393054919,
+      "learning_rate": 5.941021542240098e-07,
+      "loss": 0.259,
+      "step": 2672
+    },
+    {
+      "epoch": 1.345179584120983,
+      "grad_norm": 0.1684270679379651,
+      "learning_rate": 5.938383817231678e-07,
+      "loss": 0.2668,
+      "step": 2673
+    },
+    {
+      "epoch": 1.345683679899181,
+      "grad_norm": 0.17591755928235953,
+      "learning_rate": 5.93574582150534e-07,
+      "loss": 0.271,
+      "step": 2674
+    },
+    {
+      "epoch": 1.3461877756773788,
+      "grad_norm": 0.17125486033048526,
+      "learning_rate": 5.933107555822131e-07,
+      "loss": 0.2793,
+      "step": 2675
+    },
+    {
+      "epoch": 1.3466918714555765,
+      "grad_norm": 0.17748188736278603,
+      "learning_rate": 5.930469020943171e-07,
+      "loss": 0.2777,
+      "step": 2676
+    },
+    {
+      "epoch": 1.3471959672337745,
+      "grad_norm": 0.16672523194129693,
+      "learning_rate": 5.927830217629661e-07,
+      "loss": 0.2536,
+      "step": 2677
+    },
+    {
+      "epoch": 1.3477000630119722,
+      "grad_norm": 0.1713361784166428,
+      "learning_rate": 5.925191146642883e-07,
+      "loss": 0.2622,
+      "step": 2678
+    },
+    {
+      "epoch": 1.34820415879017,
+      "grad_norm": 0.17829104370453977,
+      "learning_rate": 5.92255180874419e-07,
+      "loss": 0.2817,
+      "step": 2679
+    },
+    {
+      "epoch": 1.348708254568368,
+      "grad_norm": 0.18860577855505326,
+      "learning_rate": 5.919912204695014e-07,
+      "loss": 0.2799,
+      "step": 2680
+    },
+    {
+      "epoch": 1.3492123503465658,
+      "grad_norm": 0.1833287597973856,
+      "learning_rate": 5.917272335256865e-07,
+      "loss": 0.2858,
+      "step": 2681
+    },
+    {
+      "epoch": 1.3497164461247637,
+      "grad_norm": 0.19807321067808822,
+      "learning_rate": 5.914632201191332e-07,
+      "loss": 0.2787,
+      "step": 2682
+    },
+    {
+      "epoch": 1.3502205419029616,
+      "grad_norm": 0.16859634450252498,
+      "learning_rate": 5.911991803260074e-07,
+      "loss": 0.2764,
+      "step": 2683
+    },
+    {
+      "epoch": 1.3507246376811595,
+      "grad_norm": 0.190261438179237,
+      "learning_rate": 5.909351142224829e-07,
+      "loss": 0.272,
+      "step": 2684
+    },
+    {
+      "epoch": 1.3512287334593573,
+      "grad_norm": 0.17519831750598353,
+      "learning_rate": 5.906710218847413e-07,
+      "loss": 0.2603,
+      "step": 2685
+    },
+    {
+      "epoch": 1.3517328292375552,
+      "grad_norm": 0.18778966952607162,
+      "learning_rate": 5.904069033889716e-07,
+      "loss": 0.261,
+      "step": 2686
+    },
+    {
+      "epoch": 1.3522369250157529,
+      "grad_norm": 0.17347841958582969,
+      "learning_rate": 5.901427588113703e-07,
+      "loss": 0.2885,
+      "step": 2687
+    },
+    {
+      "epoch": 1.352741020793951,
+      "grad_norm": 0.17645693666916326,
+      "learning_rate": 5.898785882281415e-07,
+      "loss": 0.2831,
+      "step": 2688
+    },
+    {
+      "epoch": 1.3532451165721486,
+      "grad_norm": 0.17182233696335003,
+      "learning_rate": 5.896143917154967e-07,
+      "loss": 0.2712,
+      "step": 2689
+    },
+    {
+      "epoch": 1.3537492123503465,
+      "grad_norm": 0.17603565137881136,
+      "learning_rate": 5.893501693496553e-07,
+      "loss": 0.2634,
+      "step": 2690
+    },
+    {
+      "epoch": 1.3542533081285444,
+      "grad_norm": 0.18263857001668504,
+      "learning_rate": 5.890859212068433e-07,
+      "loss": 0.2738,
+      "step": 2691
+    },
+    {
+      "epoch": 1.3547574039067423,
+      "grad_norm": 0.1730156353984436,
+      "learning_rate": 5.888216473632949e-07,
+      "loss": 0.2716,
+      "step": 2692
+    },
+    {
+      "epoch": 1.3552614996849401,
+      "grad_norm": 0.17823231943674622,
+      "learning_rate": 5.885573478952515e-07,
+      "loss": 0.275,
+      "step": 2693
+    },
+    {
+      "epoch": 1.355765595463138,
+      "grad_norm": 0.17236615330455812,
+      "learning_rate": 5.882930228789617e-07,
+      "loss": 0.2674,
+      "step": 2694
+    },
+    {
+      "epoch": 1.356269691241336,
+      "grad_norm": 0.16824543267551886,
+      "learning_rate": 5.880286723906817e-07,
+      "loss": 0.2711,
+      "step": 2695
+    },
+    {
+      "epoch": 1.3567737870195338,
+      "grad_norm": 0.17367807155908455,
+      "learning_rate": 5.877642965066753e-07,
+      "loss": 0.2839,
+      "step": 2696
+    },
+    {
+      "epoch": 1.3572778827977316,
+      "grad_norm": 0.17538723204136947,
+      "learning_rate": 5.874998953032128e-07,
+      "loss": 0.2686,
+      "step": 2697
+    },
+    {
+      "epoch": 1.3577819785759293,
+      "grad_norm": 0.17766393535437983,
+      "learning_rate": 5.872354688565727e-07,
+      "loss": 0.2816,
+      "step": 2698
+    },
+    {
+      "epoch": 1.3582860743541274,
+      "grad_norm": 0.1699641810797796,
+      "learning_rate": 5.869710172430401e-07,
+      "loss": 0.2605,
+      "step": 2699
+    },
+    {
+      "epoch": 1.358790170132325,
+      "grad_norm": 0.17422760346977517,
+      "learning_rate": 5.867065405389078e-07,
+      "loss": 0.2846,
+      "step": 2700
+    },
+    {
+      "epoch": 1.359294265910523,
+      "grad_norm": 0.18149647107592987,
+      "learning_rate": 5.864420388204757e-07,
+      "loss": 0.2749,
+      "step": 2701
+    },
+    {
+      "epoch": 1.3597983616887208,
+      "grad_norm": 0.1847991842938757,
+      "learning_rate": 5.86177512164051e-07,
+      "loss": 0.2695,
+      "step": 2702
+    },
+    {
+      "epoch": 1.3603024574669187,
+      "grad_norm": 0.16847607424615257,
+      "learning_rate": 5.859129606459477e-07,
+      "loss": 0.2702,
+      "step": 2703
+    },
+    {
+      "epoch": 1.3608065532451166,
+      "grad_norm": 0.17245367094502653,
+      "learning_rate": 5.856483843424875e-07,
+      "loss": 0.2667,
+      "step": 2704
+    },
+    {
+      "epoch": 1.3613106490233144,
+      "grad_norm": 0.16958701012632785,
+      "learning_rate": 5.853837833299991e-07,
+      "loss": 0.2748,
+      "step": 2705
+    },
+    {
+      "epoch": 1.3618147448015123,
+      "grad_norm": 0.17294794292884255,
+      "learning_rate": 5.85119157684818e-07,
+      "loss": 0.2742,
+      "step": 2706
+    },
+    {
+      "epoch": 1.3623188405797102,
+      "grad_norm": 0.16738628688725057,
+      "learning_rate": 5.848545074832873e-07,
+      "loss": 0.2549,
+      "step": 2707
+    },
+    {
+      "epoch": 1.362822936357908,
+      "grad_norm": 0.17652623817244345,
+      "learning_rate": 5.845898328017566e-07,
+      "loss": 0.2758,
+      "step": 2708
+    },
+    {
+      "epoch": 1.3633270321361057,
+      "grad_norm": 0.16927838446248678,
+      "learning_rate": 5.843251337165834e-07,
+      "loss": 0.2714,
+      "step": 2709
+    },
+    {
+      "epoch": 1.3638311279143038,
+      "grad_norm": 0.17423646348688704,
+      "learning_rate": 5.840604103041313e-07,
+      "loss": 0.2652,
+      "step": 2710
+    },
+    {
+      "epoch": 1.3643352236925015,
+      "grad_norm": 0.17007075110833333,
+      "learning_rate": 5.837956626407717e-07,
+      "loss": 0.2815,
+      "step": 2711
+    },
+    {
+      "epoch": 1.3648393194706994,
+      "grad_norm": 0.17563065467297476,
+      "learning_rate": 5.835308908028826e-07,
+      "loss": 0.2837,
+      "step": 2712
+    },
+    {
+      "epoch": 1.3653434152488972,
+      "grad_norm": 0.17681431025674674,
+      "learning_rate": 5.832660948668488e-07,
+      "loss": 0.2872,
+      "step": 2713
+    },
+    {
+      "epoch": 1.3658475110270951,
+      "grad_norm": 0.17713189828891057,
+      "learning_rate": 5.830012749090624e-07,
+      "loss": 0.2704,
+      "step": 2714
+    },
+    {
+      "epoch": 1.366351606805293,
+      "grad_norm": 0.17125824414564597,
+      "learning_rate": 5.827364310059224e-07,
+      "loss": 0.2698,
+      "step": 2715
+    },
+    {
+      "epoch": 1.3668557025834909,
+      "grad_norm": 0.1721925153952408,
+      "learning_rate": 5.824715632338345e-07,
+      "loss": 0.2614,
+      "step": 2716
+    },
+    {
+      "epoch": 1.3673597983616887,
+      "grad_norm": 0.1738669650774558,
+      "learning_rate": 5.822066716692116e-07,
+      "loss": 0.2891,
+      "step": 2717
+    },
+    {
+      "epoch": 1.3678638941398866,
+      "grad_norm": 0.17612032698684543,
+      "learning_rate": 5.819417563884732e-07,
+      "loss": 0.2888,
+      "step": 2718
+    },
+    {
+      "epoch": 1.3683679899180845,
+      "grad_norm": 0.1710729077993677,
+      "learning_rate": 5.816768174680457e-07,
+      "loss": 0.2763,
+      "step": 2719
+    },
+    {
+      "epoch": 1.3688720856962822,
+      "grad_norm": 0.17794493706059122,
+      "learning_rate": 5.814118549843623e-07,
+      "loss": 0.2623,
+      "step": 2720
+    },
+    {
+      "epoch": 1.3693761814744803,
+      "grad_norm": 0.17143323856959825,
+      "learning_rate": 5.81146869013863e-07,
+      "loss": 0.2647,
+      "step": 2721
+    },
+    {
+      "epoch": 1.369880277252678,
+      "grad_norm": 0.19765533176326622,
+      "learning_rate": 5.808818596329947e-07,
+      "loss": 0.2896,
+      "step": 2722
+    },
+    {
+      "epoch": 1.3703843730308758,
+      "grad_norm": 0.17134806523272622,
+      "learning_rate": 5.806168269182108e-07,
+      "loss": 0.2844,
+      "step": 2723
+    },
+    {
+      "epoch": 1.3708884688090737,
+      "grad_norm": 0.17764670109993702,
+      "learning_rate": 5.803517709459718e-07,
+      "loss": 0.2857,
+      "step": 2724
+    },
+    {
+      "epoch": 1.3713925645872715,
+      "grad_norm": 0.17405509742325434,
+      "learning_rate": 5.800866917927448e-07,
+      "loss": 0.2808,
+      "step": 2725
+    },
+    {
+      "epoch": 1.3718966603654694,
+      "grad_norm": 0.17704791599385142,
+      "learning_rate": 5.798215895350032e-07,
+      "loss": 0.2569,
+      "step": 2726
+    },
+    {
+      "epoch": 1.3724007561436673,
+      "grad_norm": 0.19139286392663546,
+      "learning_rate": 5.795564642492274e-07,
+      "loss": 0.2782,
+      "step": 2727
+    },
+    {
+      "epoch": 1.3729048519218652,
+      "grad_norm": 0.18973954678563565,
+      "learning_rate": 5.792913160119046e-07,
+      "loss": 0.2646,
+      "step": 2728
+    },
+    {
+      "epoch": 1.373408947700063,
+      "grad_norm": 0.1717813204204622,
+      "learning_rate": 5.790261448995283e-07,
+      "loss": 0.2752,
+      "step": 2729
+    },
+    {
+      "epoch": 1.373913043478261,
+      "grad_norm": 0.1707475223327584,
+      "learning_rate": 5.787609509885987e-07,
+      "loss": 0.2551,
+      "step": 2730
+    },
+    {
+      "epoch": 1.3744171392564588,
+      "grad_norm": 0.1715006155883366,
+      "learning_rate": 5.784957343556227e-07,
+      "loss": 0.2837,
+      "step": 2731
+    },
+    {
+      "epoch": 1.3749212350346567,
+      "grad_norm": 0.16563216070091388,
+      "learning_rate": 5.782304950771136e-07,
+      "loss": 0.2482,
+      "step": 2732
+    },
+    {
+      "epoch": 1.3754253308128543,
+      "grad_norm": 0.1701892292799077,
+      "learning_rate": 5.779652332295912e-07,
+      "loss": 0.2713,
+      "step": 2733
+    },
+    {
+      "epoch": 1.3759294265910522,
+      "grad_norm": 0.16506563913482972,
+      "learning_rate": 5.776999488895821e-07,
+      "loss": 0.2603,
+      "step": 2734
+    },
+    {
+      "epoch": 1.37643352236925,
+      "grad_norm": 0.17768766370328332,
+      "learning_rate": 5.774346421336191e-07,
+      "loss": 0.2691,
+      "step": 2735
+    },
+    {
+      "epoch": 1.376937618147448,
+      "grad_norm": 0.1821534720498173,
+      "learning_rate": 5.771693130382413e-07,
+      "loss": 0.2755,
+      "step": 2736
+    },
+    {
+      "epoch": 1.3774417139256458,
+      "grad_norm": 0.1719459414964014,
+      "learning_rate": 5.769039616799949e-07,
+      "loss": 0.2834,
+      "step": 2737
+    },
+    {
+      "epoch": 1.3779458097038437,
+      "grad_norm": 0.17228563061652777,
+      "learning_rate": 5.766385881354319e-07,
+      "loss": 0.2977,
+      "step": 2738
+    },
+    {
+      "epoch": 1.3784499054820416,
+      "grad_norm": 0.17256562211361462,
+      "learning_rate": 5.76373192481111e-07,
+      "loss": 0.2736,
+      "step": 2739
+    },
+    {
+      "epoch": 1.3789540012602395,
+      "grad_norm": 0.17577017025513997,
+      "learning_rate": 5.76107774793597e-07,
+      "loss": 0.2693,
+      "step": 2740
+    },
+    {
+      "epoch": 1.3794580970384374,
+      "grad_norm": 0.17615100356954397,
+      "learning_rate": 5.758423351494617e-07,
+      "loss": 0.2755,
+      "step": 2741
+    },
+    {
+      "epoch": 1.3799621928166352,
+      "grad_norm": 0.1769331316897802,
+      "learning_rate": 5.755768736252824e-07,
+      "loss": 0.2651,
+      "step": 2742
+    },
+    {
+      "epoch": 1.3804662885948331,
+      "grad_norm": 0.18475639772963626,
+      "learning_rate": 5.753113902976433e-07,
+      "loss": 0.2818,
+      "step": 2743
+    },
+    {
+      "epoch": 1.3809703843730308,
+      "grad_norm": 0.17618416233072068,
+      "learning_rate": 5.750458852431346e-07,
+      "loss": 0.2993,
+      "step": 2744
+    },
+    {
+      "epoch": 1.3814744801512289,
+      "grad_norm": 0.1855289675606494,
+      "learning_rate": 5.747803585383529e-07,
+      "loss": 0.2645,
+      "step": 2745
+    },
+    {
+      "epoch": 1.3819785759294265,
+      "grad_norm": 0.17112395237932768,
+      "learning_rate": 5.74514810259901e-07,
+      "loss": 0.2796,
+      "step": 2746
+    },
+    {
+      "epoch": 1.3824826717076244,
+      "grad_norm": 0.17069954299670645,
+      "learning_rate": 5.742492404843877e-07,
+      "loss": 0.2725,
+      "step": 2747
+    },
+    {
+      "epoch": 1.3829867674858223,
+      "grad_norm": 0.17473478086312352,
+      "learning_rate": 5.739836492884287e-07,
+      "loss": 0.2567,
+      "step": 2748
+    },
+    {
+      "epoch": 1.3834908632640202,
+      "grad_norm": 0.17717349784959838,
+      "learning_rate": 5.737180367486453e-07,
+      "loss": 0.2764,
+      "step": 2749
+    },
+    {
+      "epoch": 1.383994959042218,
+      "grad_norm": 0.18599321000957772,
+      "learning_rate": 5.734524029416648e-07,
+      "loss": 0.2776,
+      "step": 2750
+    },
+    {
+      "epoch": 1.384499054820416,
+      "grad_norm": 0.17518820340484337,
+      "learning_rate": 5.731867479441211e-07,
+      "loss": 0.2585,
+      "step": 2751
+    },
+    {
+      "epoch": 1.3850031505986138,
+      "grad_norm": 0.1792914417737057,
+      "learning_rate": 5.72921071832654e-07,
+      "loss": 0.282,
+      "step": 2752
+    },
+    {
+      "epoch": 1.3855072463768117,
+      "grad_norm": 0.17099619039747274,
+      "learning_rate": 5.726553746839094e-07,
+      "loss": 0.2632,
+      "step": 2753
+    },
+    {
+      "epoch": 1.3860113421550095,
+      "grad_norm": 0.18225019166152914,
+      "learning_rate": 5.72389656574539e-07,
+      "loss": 0.2695,
+      "step": 2754
+    },
+    {
+      "epoch": 1.3865154379332072,
+      "grad_norm": 0.1707269538625628,
+      "learning_rate": 5.721239175812014e-07,
+      "loss": 0.272,
+      "step": 2755
+    },
+    {
+      "epoch": 1.3870195337114053,
+      "grad_norm": 0.1796542283862304,
+      "learning_rate": 5.718581577805604e-07,
+      "loss": 0.2721,
+      "step": 2756
+    },
+    {
+      "epoch": 1.387523629489603,
+      "grad_norm": 0.1770367282404336,
+      "learning_rate": 5.71592377249286e-07,
+      "loss": 0.2735,
+      "step": 2757
+    },
+    {
+      "epoch": 1.3880277252678008,
+      "grad_norm": 0.16921478099539097,
+      "learning_rate": 5.713265760640541e-07,
+      "loss": 0.2778,
+      "step": 2758
+    },
+    {
+      "epoch": 1.3885318210459987,
+      "grad_norm": 0.17268757073355442,
+      "learning_rate": 5.71060754301547e-07,
+      "loss": 0.2748,
+      "step": 2759
+    },
+    {
+      "epoch": 1.3890359168241966,
+      "grad_norm": 0.17598942075530893,
+      "learning_rate": 5.707949120384523e-07,
+      "loss": 0.2724,
+      "step": 2760
+    },
+    {
+      "epoch": 1.3895400126023945,
+      "grad_norm": 0.18011915470443154,
+      "learning_rate": 5.705290493514642e-07,
+      "loss": 0.2891,
+      "step": 2761
+    },
+    {
+      "epoch": 1.3900441083805923,
+      "grad_norm": 0.17876735405789806,
+      "learning_rate": 5.702631663172822e-07,
+      "loss": 0.2772,
+      "step": 2762
+    },
+    {
+      "epoch": 1.3905482041587902,
+      "grad_norm": 0.17718778143728184,
+      "learning_rate": 5.69997263012612e-07,
+      "loss": 0.2719,
+      "step": 2763
+    },
+    {
+      "epoch": 1.391052299936988,
+      "grad_norm": 0.17114075203463552,
+      "learning_rate": 5.697313395141651e-07,
+      "loss": 0.2689,
+      "step": 2764
+    },
+    {
+      "epoch": 1.391556395715186,
+      "grad_norm": 0.1880075108323104,
+      "learning_rate": 5.694653958986586e-07,
+      "loss": 0.2893,
+      "step": 2765
+    },
+    {
+      "epoch": 1.3920604914933836,
+      "grad_norm": 0.177316108146064,
+      "learning_rate": 5.691994322428159e-07,
+      "loss": 0.2774,
+      "step": 2766
+    },
+    {
+      "epoch": 1.3925645872715817,
+      "grad_norm": 0.16915555475740024,
+      "learning_rate": 5.689334486233655e-07,
+      "loss": 0.2605,
+      "step": 2767
+    },
+    {
+      "epoch": 1.3930686830497794,
+      "grad_norm": 0.1706580418948374,
+      "learning_rate": 5.686674451170421e-07,
+      "loss": 0.2669,
+      "step": 2768
+    },
+    {
+      "epoch": 1.3935727788279773,
+      "grad_norm": 0.18166657991393095,
+      "learning_rate": 5.684014218005861e-07,
+      "loss": 0.271,
+      "step": 2769
+    },
+    {
+      "epoch": 1.3940768746061751,
+      "grad_norm": 0.18074355543323026,
+      "learning_rate": 5.68135378750744e-07,
+      "loss": 0.2787,
+      "step": 2770
+    },
+    {
+      "epoch": 1.394580970384373,
+      "grad_norm": 0.18191764036464922,
+      "learning_rate": 5.678693160442668e-07,
+      "loss": 0.2694,
+      "step": 2771
+    },
+    {
+      "epoch": 1.3950850661625709,
+      "grad_norm": 0.1742981275140808,
+      "learning_rate": 5.676032337579126e-07,
+      "loss": 0.2668,
+      "step": 2772
+    },
+    {
+      "epoch": 1.3955891619407688,
+      "grad_norm": 0.182525466069335,
+      "learning_rate": 5.673371319684443e-07,
+      "loss": 0.2704,
+      "step": 2773
+    },
+    {
+      "epoch": 1.3960932577189666,
+      "grad_norm": 0.18422795498525135,
+      "learning_rate": 5.670710107526303e-07,
+      "loss": 0.2554,
+      "step": 2774
+    },
+    {
+      "epoch": 1.3965973534971645,
+      "grad_norm": 0.16935758864299294,
+      "learning_rate": 5.668048701872453e-07,
+      "loss": 0.2787,
+      "step": 2775
+    },
+    {
+      "epoch": 1.3971014492753624,
+      "grad_norm": 0.17250150189867655,
+      "learning_rate": 5.665387103490691e-07,
+      "loss": 0.269,
+      "step": 2776
+    },
+    {
+      "epoch": 1.39760554505356,
+      "grad_norm": 0.16845458026081864,
+      "learning_rate": 5.662725313148872e-07,
+      "loss": 0.2591,
+      "step": 2777
+    },
+    {
+      "epoch": 1.3981096408317581,
+      "grad_norm": 0.19087750959177846,
+      "learning_rate": 5.660063331614905e-07,
+      "loss": 0.2684,
+      "step": 2778
+    },
+    {
+      "epoch": 1.3986137366099558,
+      "grad_norm": 0.17287405774573983,
+      "learning_rate": 5.657401159656757e-07,
+      "loss": 0.2707,
+      "step": 2779
+    },
+    {
+      "epoch": 1.3991178323881537,
+      "grad_norm": 0.17266422585724345,
+      "learning_rate": 5.654738798042445e-07,
+      "loss": 0.2669,
+      "step": 2780
+    },
+    {
+      "epoch": 1.3996219281663516,
+      "grad_norm": 0.1687413663962204,
+      "learning_rate": 5.652076247540045e-07,
+      "loss": 0.269,
+      "step": 2781
+    },
+    {
+      "epoch": 1.4001260239445494,
+      "grad_norm": 0.17337825335332843,
+      "learning_rate": 5.649413508917689e-07,
+      "loss": 0.2687,
+      "step": 2782
+    },
+    {
+      "epoch": 1.4006301197227473,
+      "grad_norm": 0.17254045056559464,
+      "learning_rate": 5.646750582943558e-07,
+      "loss": 0.2806,
+      "step": 2783
+    },
+    {
+      "epoch": 1.4011342155009452,
+      "grad_norm": 0.17604086047481585,
+      "learning_rate": 5.644087470385889e-07,
+      "loss": 0.2794,
+      "step": 2784
+    },
+    {
+      "epoch": 1.401638311279143,
+      "grad_norm": 0.17593248783979848,
+      "learning_rate": 5.641424172012976e-07,
+      "loss": 0.2687,
+      "step": 2785
+    },
+    {
+      "epoch": 1.402142407057341,
+      "grad_norm": 0.17677274912632954,
+      "learning_rate": 5.638760688593162e-07,
+      "loss": 0.2667,
+      "step": 2786
+    },
+    {
+      "epoch": 1.402142407057341,
+      "eval_loss": 0.3082324266433716,
+      "eval_runtime": 17.3029,
+      "eval_samples_per_second": 49.414,
+      "eval_steps_per_second": 1.04,
+      "step": 2786
+    },
+    {
+      "epoch": 1.4026465028355388,
+      "grad_norm": 0.17615085346486575,
+      "learning_rate": 5.636097020894849e-07,
+      "loss": 0.2652,
+      "step": 2787
+    },
+    {
+      "epoch": 1.4031505986137365,
+      "grad_norm": 0.18280305120298956,
+      "learning_rate": 5.633433169686483e-07,
+      "loss": 0.2622,
+      "step": 2788
+    },
+    {
+      "epoch": 1.4036546943919346,
+      "grad_norm": 0.1766526757543098,
+      "learning_rate": 5.630769135736573e-07,
+      "loss": 0.2855,
+      "step": 2789
+    },
+    {
+      "epoch": 1.4041587901701322,
+      "grad_norm": 0.18013844905463533,
+      "learning_rate": 5.628104919813673e-07,
+      "loss": 0.2864,
+      "step": 2790
+    },
+    {
+      "epoch": 1.40466288594833,
+      "grad_norm": 0.17308521362796225,
+      "learning_rate": 5.625440522686395e-07,
+      "loss": 0.2822,
+      "step": 2791
+    },
+    {
+      "epoch": 1.405166981726528,
+      "grad_norm": 0.17384117426113055,
+      "learning_rate": 5.622775945123401e-07,
+      "loss": 0.2726,
+      "step": 2792
+    },
+    {
+      "epoch": 1.4056710775047259,
+      "grad_norm": 0.18309399117577832,
+      "learning_rate": 5.620111187893404e-07,
+      "loss": 0.2632,
+      "step": 2793
+    },
+    {
+      "epoch": 1.4061751732829237,
+      "grad_norm": 0.17415826316818256,
+      "learning_rate": 5.617446251765173e-07,
+      "loss": 0.2731,
+      "step": 2794
+    },
+    {
+      "epoch": 1.4066792690611216,
+      "grad_norm": 0.1739100478730036,
+      "learning_rate": 5.614781137507521e-07,
+      "loss": 0.2762,
+      "step": 2795
+    },
+    {
+      "epoch": 1.4071833648393195,
+      "grad_norm": 0.1764423898652386,
+      "learning_rate": 5.612115845889321e-07,
+      "loss": 0.2849,
+      "step": 2796
+    },
+    {
+      "epoch": 1.4076874606175174,
+      "grad_norm": 0.1760604570799819,
+      "learning_rate": 5.60945037767949e-07,
+      "loss": 0.2719,
+      "step": 2797
+    },
+    {
+      "epoch": 1.4081915563957152,
+      "grad_norm": 0.17431566775862142,
+      "learning_rate": 5.606784733647e-07,
+      "loss": 0.253,
+      "step": 2798
+    },
+    {
+      "epoch": 1.4086956521739131,
+      "grad_norm": 0.17776800316760477,
+      "learning_rate": 5.604118914560873e-07,
+      "loss": 0.2699,
+      "step": 2799
+    },
+    {
+      "epoch": 1.409199747952111,
+      "grad_norm": 0.18918069069185012,
+      "learning_rate": 5.601452921190183e-07,
+      "loss": 0.2717,
+      "step": 2800
+    },
+    {
+      "epoch": 1.4097038437303087,
+      "grad_norm": 0.1732601715342729,
+      "learning_rate": 5.598786754304051e-07,
+      "loss": 0.2714,
+      "step": 2801
+    },
+    {
+      "epoch": 1.4102079395085065,
+      "grad_norm": 0.1791884506349457,
+      "learning_rate": 5.596120414671649e-07,
+      "loss": 0.2693,
+      "step": 2802
+    },
+    {
+      "epoch": 1.4107120352867044,
+      "grad_norm": 0.17551762309525726,
+      "learning_rate": 5.5934539030622e-07,
+      "loss": 0.2931,
+      "step": 2803
+    },
+    {
+      "epoch": 1.4112161310649023,
+      "grad_norm": 0.17653447291530958,
+      "learning_rate": 5.590787220244975e-07,
+      "loss": 0.279,
+      "step": 2804
+    },
+    {
+      "epoch": 1.4117202268431002,
+      "grad_norm": 0.1715844081659434,
+      "learning_rate": 5.588120366989299e-07,
+      "loss": 0.2692,
+      "step": 2805
+    },
+    {
+      "epoch": 1.412224322621298,
+      "grad_norm": 0.17195024969559972,
+      "learning_rate": 5.585453344064538e-07,
+      "loss": 0.2755,
+      "step": 2806
+    },
+    {
+      "epoch": 1.412728418399496,
+      "grad_norm": 0.17614434935253434,
+      "learning_rate": 5.582786152240116e-07,
+      "loss": 0.2868,
+      "step": 2807
+    },
+    {
+      "epoch": 1.4132325141776938,
+      "grad_norm": 0.17488803322082694,
+      "learning_rate": 5.580118792285497e-07,
+      "loss": 0.2785,
+      "step": 2808
+    },
+    {
+      "epoch": 1.4137366099558917,
+      "grad_norm": 0.17403718634941737,
+      "learning_rate": 5.577451264970203e-07,
+      "loss": 0.2779,
+      "step": 2809
+    },
+    {
+      "epoch": 1.4142407057340896,
+      "grad_norm": 0.180098783645297,
+      "learning_rate": 5.574783571063795e-07,
+      "loss": 0.2805,
+      "step": 2810
+    },
+    {
+      "epoch": 1.4147448015122874,
+      "grad_norm": 0.1735655015170712,
+      "learning_rate": 5.572115711335886e-07,
+      "loss": 0.2695,
+      "step": 2811
+    },
+    {
+      "epoch": 1.415248897290485,
+      "grad_norm": 0.1741615072776797,
+      "learning_rate": 5.56944768655614e-07,
+      "loss": 0.2585,
+      "step": 2812
+    },
+    {
+      "epoch": 1.4157529930686832,
+      "grad_norm": 0.1743862849176332,
+      "learning_rate": 5.566779497494264e-07,
+      "loss": 0.2979,
+      "step": 2813
+    },
+    {
+      "epoch": 1.4162570888468808,
+      "grad_norm": 0.17758390356185494,
+      "learning_rate": 5.564111144920013e-07,
+      "loss": 0.2745,
+      "step": 2814
+    },
+    {
+      "epoch": 1.4167611846250787,
+      "grad_norm": 0.1834026948019301,
+      "learning_rate": 5.561442629603192e-07,
+      "loss": 0.2768,
+      "step": 2815
+    },
+    {
+      "epoch": 1.4172652804032766,
+      "grad_norm": 0.17218386664662416,
+      "learning_rate": 5.55877395231365e-07,
+      "loss": 0.278,
+      "step": 2816
+    },
+    {
+      "epoch": 1.4177693761814745,
+      "grad_norm": 0.17269835730848343,
+      "learning_rate": 5.556105113821285e-07,
+      "loss": 0.272,
+      "step": 2817
+    },
+    {
+      "epoch": 1.4182734719596723,
+      "grad_norm": 0.19670398507490758,
+      "learning_rate": 5.553436114896037e-07,
+      "loss": 0.2511,
+      "step": 2818
+    },
+    {
+      "epoch": 1.4187775677378702,
+      "grad_norm": 0.1715178941839397,
+      "learning_rate": 5.5507669563079e-07,
+      "loss": 0.2735,
+      "step": 2819
+    },
+    {
+      "epoch": 1.419281663516068,
+      "grad_norm": 0.1729826705555197,
+      "learning_rate": 5.548097638826907e-07,
+      "loss": 0.2707,
+      "step": 2820
+    },
+    {
+      "epoch": 1.419785759294266,
+      "grad_norm": 0.1725819935420221,
+      "learning_rate": 5.545428163223142e-07,
+      "loss": 0.2671,
+      "step": 2821
+    },
+    {
+      "epoch": 1.4202898550724639,
+      "grad_norm": 0.17665111931235344,
+      "learning_rate": 5.542758530266729e-07,
+      "loss": 0.2538,
+      "step": 2822
+    },
+    {
+      "epoch": 1.4207939508506615,
+      "grad_norm": 0.17897097101640105,
+      "learning_rate": 5.540088740727843e-07,
+      "loss": 0.2697,
+      "step": 2823
+    },
+    {
+      "epoch": 1.4212980466288596,
+      "grad_norm": 0.1687081289149051,
+      "learning_rate": 5.537418795376702e-07,
+      "loss": 0.2828,
+      "step": 2824
+    },
+    {
+      "epoch": 1.4218021424070573,
+      "grad_norm": 0.20222752341146646,
+      "learning_rate": 5.534748694983567e-07,
+      "loss": 0.2713,
+      "step": 2825
+    },
+    {
+      "epoch": 1.4223062381852551,
+      "grad_norm": 0.17189172904171757,
+      "learning_rate": 5.532078440318746e-07,
+      "loss": 0.2698,
+      "step": 2826
+    },
+    {
+      "epoch": 1.422810333963453,
+      "grad_norm": 0.17345223486754652,
+      "learning_rate": 5.52940803215259e-07,
+      "loss": 0.276,
+      "step": 2827
+    },
+    {
+      "epoch": 1.423314429741651,
+      "grad_norm": 0.2080580963062512,
+      "learning_rate": 5.526737471255498e-07,
+      "loss": 0.2878,
+      "step": 2828
+    },
+    {
+      "epoch": 1.4238185255198488,
+      "grad_norm": 0.1993251803507769,
+      "learning_rate": 5.524066758397907e-07,
+      "loss": 0.2741,
+      "step": 2829
+    },
+    {
+      "epoch": 1.4243226212980467,
+      "grad_norm": 0.19657833333195346,
+      "learning_rate": 5.521395894350303e-07,
+      "loss": 0.2743,
+      "step": 2830
+    },
+    {
+      "epoch": 1.4248267170762445,
+      "grad_norm": 0.18203360278973735,
+      "learning_rate": 5.518724879883215e-07,
+      "loss": 0.2769,
+      "step": 2831
+    },
+    {
+      "epoch": 1.4253308128544424,
+      "grad_norm": 0.18264070770705262,
+      "learning_rate": 5.516053715767211e-07,
+      "loss": 0.2853,
+      "step": 2832
+    },
+    {
+      "epoch": 1.4258349086326403,
+      "grad_norm": 0.18164978911251264,
+      "learning_rate": 5.513382402772906e-07,
+      "loss": 0.2755,
+      "step": 2833
+    },
+    {
+      "epoch": 1.426339004410838,
+      "grad_norm": 0.1662574674843768,
+      "learning_rate": 5.510710941670959e-07,
+      "loss": 0.2689,
+      "step": 2834
+    },
+    {
+      "epoch": 1.426843100189036,
+      "grad_norm": 0.17319135764139246,
+      "learning_rate": 5.508039333232069e-07,
+      "loss": 0.2856,
+      "step": 2835
+    },
+    {
+      "epoch": 1.4273471959672337,
+      "grad_norm": 0.1827176229024098,
+      "learning_rate": 5.505367578226978e-07,
+      "loss": 0.2842,
+      "step": 2836
+    },
+    {
+      "epoch": 1.4278512917454316,
+      "grad_norm": 0.1908930430427692,
+      "learning_rate": 5.502695677426471e-07,
+      "loss": 0.2634,
+      "step": 2837
+    },
+    {
+      "epoch": 1.4283553875236294,
+      "grad_norm": 0.18109239755144405,
+      "learning_rate": 5.500023631601376e-07,
+      "loss": 0.2786,
+      "step": 2838
+    },
+    {
+      "epoch": 1.4288594833018273,
+      "grad_norm": 0.18601457798351573,
+      "learning_rate": 5.497351441522561e-07,
+      "loss": 0.2773,
+      "step": 2839
+    },
+    {
+      "epoch": 1.4293635790800252,
+      "grad_norm": 0.1851254103241904,
+      "learning_rate": 5.494679107960936e-07,
+      "loss": 0.2918,
+      "step": 2840
+    },
+    {
+      "epoch": 1.429867674858223,
+      "grad_norm": 0.1740395548507666,
+      "learning_rate": 5.492006631687451e-07,
+      "loss": 0.2681,
+      "step": 2841
+    },
+    {
+      "epoch": 1.430371770636421,
+      "grad_norm": 0.17944562058468594,
+      "learning_rate": 5.489334013473103e-07,
+      "loss": 0.2643,
+      "step": 2842
+    },
+    {
+      "epoch": 1.4308758664146188,
+      "grad_norm": 0.1721575725843985,
+      "learning_rate": 5.486661254088921e-07,
+      "loss": 0.2678,
+      "step": 2843
+    },
+    {
+      "epoch": 1.4313799621928167,
+      "grad_norm": 0.18124114212980594,
+      "learning_rate": 5.483988354305984e-07,
+      "loss": 0.286,
+      "step": 2844
+    },
+    {
+      "epoch": 1.4318840579710144,
+      "grad_norm": 0.17177819687786705,
+      "learning_rate": 5.481315314895404e-07,
+      "loss": 0.2883,
+      "step": 2845
+    },
+    {
+      "epoch": 1.4323881537492125,
+      "grad_norm": 0.17999663813819824,
+      "learning_rate": 5.478642136628338e-07,
+      "loss": 0.2506,
+      "step": 2846
+    },
+    {
+      "epoch": 1.4328922495274101,
+      "grad_norm": 0.1759171770648982,
+      "learning_rate": 5.47596882027598e-07,
+      "loss": 0.2735,
+      "step": 2847
+    },
+    {
+      "epoch": 1.433396345305608,
+      "grad_norm": 0.1816944331335307,
+      "learning_rate": 5.473295366609566e-07,
+      "loss": 0.2849,
+      "step": 2848
+    },
+    {
+      "epoch": 1.4339004410838059,
+      "grad_norm": 0.17100685390259318,
+      "learning_rate": 5.470621776400371e-07,
+      "loss": 0.2612,
+      "step": 2849
+    },
+    {
+      "epoch": 1.4344045368620038,
+      "grad_norm": 0.1819142852011383,
+      "learning_rate": 5.467948050419707e-07,
+      "loss": 0.2783,
+      "step": 2850
+    },
+    {
+      "epoch": 1.4349086326402016,
+      "grad_norm": 0.18361474818908685,
+      "learning_rate": 5.465274189438931e-07,
+      "loss": 0.284,
+      "step": 2851
+    },
+    {
+      "epoch": 1.4354127284183995,
+      "grad_norm": 0.1752427494123583,
+      "learning_rate": 5.462600194229432e-07,
+      "loss": 0.2823,
+      "step": 2852
+    },
+    {
+      "epoch": 1.4359168241965974,
+      "grad_norm": 0.1773020482286541,
+      "learning_rate": 5.459926065562643e-07,
+      "loss": 0.2686,
+      "step": 2853
+    },
+    {
+      "epoch": 1.4364209199747953,
+      "grad_norm": 0.17566660311428434,
+      "learning_rate": 5.457251804210035e-07,
+      "loss": 0.2668,
+      "step": 2854
+    },
+    {
+      "epoch": 1.4369250157529931,
+      "grad_norm": 0.17963004958574516,
+      "learning_rate": 5.454577410943113e-07,
+      "loss": 0.2713,
+      "step": 2855
+    },
+    {
+      "epoch": 1.4374291115311908,
+      "grad_norm": 0.17390970931818708,
+      "learning_rate": 5.451902886533424e-07,
+      "loss": 0.2681,
+      "step": 2856
+    },
+    {
+      "epoch": 1.437933207309389,
+      "grad_norm": 0.17729799370957,
+      "learning_rate": 5.449228231752551e-07,
+      "loss": 0.272,
+      "step": 2857
+    },
+    {
+      "epoch": 1.4384373030875865,
+      "grad_norm": 0.17507123221756268,
+      "learning_rate": 5.446553447372117e-07,
+      "loss": 0.2775,
+      "step": 2858
+    },
+    {
+      "epoch": 1.4389413988657844,
+      "grad_norm": 0.17681228025938894,
+      "learning_rate": 5.443878534163779e-07,
+      "loss": 0.2831,
+      "step": 2859
+    },
+    {
+      "epoch": 1.4394454946439823,
+      "grad_norm": 0.1821736618960098,
+      "learning_rate": 5.441203492899233e-07,
+      "loss": 0.2841,
+      "step": 2860
+    },
+    {
+      "epoch": 1.4399495904221802,
+      "grad_norm": 0.17903890868161446,
+      "learning_rate": 5.438528324350214e-07,
+      "loss": 0.278,
+      "step": 2861
+    },
+    {
+      "epoch": 1.440453686200378,
+      "grad_norm": 0.180429015598911,
+      "learning_rate": 5.435853029288489e-07,
+      "loss": 0.2746,
+      "step": 2862
+    },
+    {
+      "epoch": 1.440957781978576,
+      "grad_norm": 0.17355901123245668,
+      "learning_rate": 5.433177608485865e-07,
+      "loss": 0.2679,
+      "step": 2863
+    },
+    {
+      "epoch": 1.4414618777567738,
+      "grad_norm": 0.180173248420686,
+      "learning_rate": 5.430502062714184e-07,
+      "loss": 0.2698,
+      "step": 2864
+    },
+    {
+      "epoch": 1.4419659735349717,
+      "grad_norm": 0.17214222893696038,
+      "learning_rate": 5.427826392745325e-07,
+      "loss": 0.2769,
+      "step": 2865
+    },
+    {
+      "epoch": 1.4424700693131696,
+      "grad_norm": 0.1726707684780785,
+      "learning_rate": 5.425150599351201e-07,
+      "loss": 0.2819,
+      "step": 2866
+    },
+    {
+      "epoch": 1.4429741650913672,
+      "grad_norm": 0.17661401865470128,
+      "learning_rate": 5.422474683303765e-07,
+      "loss": 0.2633,
+      "step": 2867
+    },
+    {
+      "epoch": 1.4434782608695653,
+      "grad_norm": 0.16966464132316653,
+      "learning_rate": 5.419798645374998e-07,
+      "loss": 0.2749,
+      "step": 2868
+    },
+    {
+      "epoch": 1.443982356647763,
+      "grad_norm": 0.17637511222129923,
+      "learning_rate": 5.417122486336923e-07,
+      "loss": 0.2747,
+      "step": 2869
+    },
+    {
+      "epoch": 1.4444864524259609,
+      "grad_norm": 0.1747096047216983,
+      "learning_rate": 5.414446206961596e-07,
+      "loss": 0.2701,
+      "step": 2870
+    },
+    {
+      "epoch": 1.4449905482041587,
+      "grad_norm": 0.18391353933023288,
+      "learning_rate": 5.411769808021104e-07,
+      "loss": 0.2786,
+      "step": 2871
+    },
+    {
+      "epoch": 1.4454946439823566,
+      "grad_norm": 0.1731778758855043,
+      "learning_rate": 5.409093290287573e-07,
+      "loss": 0.2568,
+      "step": 2872
+    },
+    {
+      "epoch": 1.4459987397605545,
+      "grad_norm": 0.16898295126793264,
+      "learning_rate": 5.406416654533163e-07,
+      "loss": 0.2657,
+      "step": 2873
+    },
+    {
+      "epoch": 1.4465028355387524,
+      "grad_norm": 0.18058265233021242,
+      "learning_rate": 5.403739901530065e-07,
+      "loss": 0.2765,
+      "step": 2874
+    },
+    {
+      "epoch": 1.4470069313169502,
+      "grad_norm": 0.17654601160313765,
+      "learning_rate": 5.401063032050507e-07,
+      "loss": 0.2789,
+      "step": 2875
+    },
+    {
+      "epoch": 1.4475110270951481,
+      "grad_norm": 0.17294349882056775,
+      "learning_rate": 5.398386046866747e-07,
+      "loss": 0.268,
+      "step": 2876
+    },
+    {
+      "epoch": 1.448015122873346,
+      "grad_norm": 0.17407010490339622,
+      "learning_rate": 5.395708946751083e-07,
+      "loss": 0.2752,
+      "step": 2877
+    },
+    {
+      "epoch": 1.4485192186515439,
+      "grad_norm": 0.1869029531049797,
+      "learning_rate": 5.393031732475837e-07,
+      "loss": 0.2839,
+      "step": 2878
+    },
+    {
+      "epoch": 1.4490233144297417,
+      "grad_norm": 0.17047646712682293,
+      "learning_rate": 5.390354404813373e-07,
+      "loss": 0.2783,
+      "step": 2879
+    },
+    {
+      "epoch": 1.4495274102079394,
+      "grad_norm": 0.1740578915583476,
+      "learning_rate": 5.38767696453608e-07,
+      "loss": 0.2903,
+      "step": 2880
+    },
+    {
+      "epoch": 1.4500315059861375,
+      "grad_norm": 0.18937601572967064,
+      "learning_rate": 5.384999412416383e-07,
+      "loss": 0.2649,
+      "step": 2881
+    },
+    {
+      "epoch": 1.4505356017643352,
+      "grad_norm": 0.17078602897509845,
+      "learning_rate": 5.382321749226743e-07,
+      "loss": 0.2753,
+      "step": 2882
+    },
+    {
+      "epoch": 1.451039697542533,
+      "grad_norm": 0.18237399698730128,
+      "learning_rate": 5.379643975739647e-07,
+      "loss": 0.2991,
+      "step": 2883
+    },
+    {
+      "epoch": 1.451543793320731,
+      "grad_norm": 0.1741266975993158,
+      "learning_rate": 5.376966092727613e-07,
+      "loss": 0.271,
+      "step": 2884
+    },
+    {
+      "epoch": 1.4520478890989288,
+      "grad_norm": 0.17610657049121992,
+      "learning_rate": 5.374288100963198e-07,
+      "loss": 0.2889,
+      "step": 2885
+    },
+    {
+      "epoch": 1.4525519848771267,
+      "grad_norm": 0.17138137935782394,
+      "learning_rate": 5.371610001218983e-07,
+      "loss": 0.2649,
+      "step": 2886
+    },
+    {
+      "epoch": 1.4530560806553245,
+      "grad_norm": 0.17838838912402527,
+      "learning_rate": 5.368931794267586e-07,
+      "loss": 0.2753,
+      "step": 2887
+    },
+    {
+      "epoch": 1.4535601764335224,
+      "grad_norm": 0.1746687660444875,
+      "learning_rate": 5.366253480881651e-07,
+      "loss": 0.2763,
+      "step": 2888
+    },
+    {
+      "epoch": 1.4540642722117203,
+      "grad_norm": 0.17490010732782957,
+      "learning_rate": 5.363575061833856e-07,
+      "loss": 0.273,
+      "step": 2889
+    },
+    {
+      "epoch": 1.4545683679899182,
+      "grad_norm": 0.1753044578433687,
+      "learning_rate": 5.360896537896909e-07,
+      "loss": 0.2818,
+      "step": 2890
+    },
+    {
+      "epoch": 1.4550724637681158,
+      "grad_norm": 0.17281297026171832,
+      "learning_rate": 5.358217909843545e-07,
+      "loss": 0.2616,
+      "step": 2891
+    },
+    {
+      "epoch": 1.455576559546314,
+      "grad_norm": 0.17311583025271762,
+      "learning_rate": 5.355539178446535e-07,
+      "loss": 0.2705,
+      "step": 2892
+    },
+    {
+      "epoch": 1.4560806553245116,
+      "grad_norm": 0.17027692820248555,
+      "learning_rate": 5.352860344478673e-07,
+      "loss": 0.2736,
+      "step": 2893
+    },
+    {
+      "epoch": 1.4565847511027095,
+      "grad_norm": 0.17311501562837145,
+      "learning_rate": 5.350181408712787e-07,
+      "loss": 0.2563,
+      "step": 2894
+    },
+    {
+      "epoch": 1.4570888468809073,
+      "grad_norm": 0.17543624536918084,
+      "learning_rate": 5.347502371921735e-07,
+      "loss": 0.2736,
+      "step": 2895
+    },
+    {
+      "epoch": 1.4575929426591052,
+      "grad_norm": 0.19910330074338614,
+      "learning_rate": 5.3448232348784e-07,
+      "loss": 0.273,
+      "step": 2896
+    },
+    {
+      "epoch": 1.458097038437303,
+      "grad_norm": 0.1738967237730725,
+      "learning_rate": 5.342143998355698e-07,
+      "loss": 0.274,
+      "step": 2897
+    },
+    {
+      "epoch": 1.458601134215501,
+      "grad_norm": 0.17570081461431064,
+      "learning_rate": 5.339464663126574e-07,
+      "loss": 0.2759,
+      "step": 2898
+    },
+    {
+      "epoch": 1.4591052299936988,
+      "grad_norm": 0.17791619225009586,
+      "learning_rate": 5.336785229963996e-07,
+      "loss": 0.2623,
+      "step": 2899
+    },
+    {
+      "epoch": 1.4596093257718967,
+      "grad_norm": 0.17002260231876098,
+      "learning_rate": 5.334105699640965e-07,
+      "loss": 0.266,
+      "step": 2900
+    },
+    {
+      "epoch": 1.4601134215500946,
+      "grad_norm": 0.18429176939879613,
+      "learning_rate": 5.33142607293051e-07,
+      "loss": 0.2819,
+      "step": 2901
+    },
+    {
+      "epoch": 1.4606175173282923,
+      "grad_norm": 0.17868955547716125,
+      "learning_rate": 5.328746350605685e-07,
+      "loss": 0.2682,
+      "step": 2902
+    },
+    {
+      "epoch": 1.4611216131064904,
+      "grad_norm": 0.16918396641956726,
+      "learning_rate": 5.326066533439575e-07,
+      "loss": 0.2755,
+      "step": 2903
+    },
+    {
+      "epoch": 1.461625708884688,
+      "grad_norm": 0.17069887650764975,
+      "learning_rate": 5.323386622205291e-07,
+      "loss": 0.2801,
+      "step": 2904
+    },
+    {
+      "epoch": 1.462129804662886,
+      "grad_norm": 0.1701037068935557,
+      "learning_rate": 5.320706617675968e-07,
+      "loss": 0.2765,
+      "step": 2905
+    },
+    {
+      "epoch": 1.4626339004410838,
+      "grad_norm": 0.17789523122182876,
+      "learning_rate": 5.318026520624774e-07,
+      "loss": 0.2912,
+      "step": 2906
+    },
+    {
+      "epoch": 1.4631379962192816,
+      "grad_norm": 0.19749764941108588,
+      "learning_rate": 5.315346331824898e-07,
+      "loss": 0.2853,
+      "step": 2907
+    },
+    {
+      "epoch": 1.4636420919974795,
+      "grad_norm": 0.1756855169152155,
+      "learning_rate": 5.31266605204956e-07,
+      "loss": 0.2832,
+      "step": 2908
+    },
+    {
+      "epoch": 1.4641461877756774,
+      "grad_norm": 0.1786872687703457,
+      "learning_rate": 5.309985682072001e-07,
+      "loss": 0.2665,
+      "step": 2909
+    },
+    {
+      "epoch": 1.4646502835538753,
+      "grad_norm": 0.1877786342320192,
+      "learning_rate": 5.307305222665494e-07,
+      "loss": 0.2774,
+      "step": 2910
+    },
+    {
+      "epoch": 1.4651543793320732,
+      "grad_norm": 0.17127707594522085,
+      "learning_rate": 5.304624674603335e-07,
+      "loss": 0.2646,
+      "step": 2911
+    },
+    {
+      "epoch": 1.465658475110271,
+      "grad_norm": 0.2560823194379178,
+      "learning_rate": 5.301944038658842e-07,
+      "loss": 0.2669,
+      "step": 2912
+    },
+    {
+      "epoch": 1.4661625708884687,
+      "grad_norm": 0.1766829426286369,
+      "learning_rate": 5.299263315605367e-07,
+      "loss": 0.2781,
+      "step": 2913
+    },
+    {
+      "epoch": 1.4666666666666668,
+      "grad_norm": 0.16776123278488433,
+      "learning_rate": 5.296582506216277e-07,
+      "loss": 0.2616,
+      "step": 2914
+    },
+    {
+      "epoch": 1.4671707624448644,
+      "grad_norm": 0.17648227246454712,
+      "learning_rate": 5.293901611264971e-07,
+      "loss": 0.2649,
+      "step": 2915
+    },
+    {
+      "epoch": 1.4676748582230623,
+      "grad_norm": 0.17432483953665048,
+      "learning_rate": 5.291220631524872e-07,
+      "loss": 0.2687,
+      "step": 2916
+    },
+    {
+      "epoch": 1.4681789540012602,
+      "grad_norm": 0.20520459105179642,
+      "learning_rate": 5.288539567769424e-07,
+      "loss": 0.2862,
+      "step": 2917
+    },
+    {
+      "epoch": 1.468683049779458,
+      "grad_norm": 0.18543376500369232,
+      "learning_rate": 5.285858420772099e-07,
+      "loss": 0.2541,
+      "step": 2918
+    },
+    {
+      "epoch": 1.469187145557656,
+      "grad_norm": 0.17322589885199907,
+      "learning_rate": 5.283177191306389e-07,
+      "loss": 0.2748,
+      "step": 2919
+    },
+    {
+      "epoch": 1.4696912413358538,
+      "grad_norm": 0.18616715044944854,
+      "learning_rate": 5.280495880145814e-07,
+      "loss": 0.2791,
+      "step": 2920
+    },
+    {
+      "epoch": 1.4701953371140517,
+      "grad_norm": 0.17972067487672042,
+      "learning_rate": 5.277814488063918e-07,
+      "loss": 0.283,
+      "step": 2921
+    },
+    {
+      "epoch": 1.4706994328922496,
+      "grad_norm": 0.19354601757670176,
+      "learning_rate": 5.27513301583426e-07,
+      "loss": 0.2733,
+      "step": 2922
+    },
+    {
+      "epoch": 1.4712035286704475,
+      "grad_norm": 0.17386234930511313,
+      "learning_rate": 5.272451464230433e-07,
+      "loss": 0.2667,
+      "step": 2923
+    },
+    {
+      "epoch": 1.4717076244486451,
+      "grad_norm": 0.1747812393098067,
+      "learning_rate": 5.269769834026045e-07,
+      "loss": 0.2538,
+      "step": 2924
+    },
+    {
+      "epoch": 1.4722117202268432,
+      "grad_norm": 0.18216928171454855,
+      "learning_rate": 5.267088125994732e-07,
+      "loss": 0.2689,
+      "step": 2925
+    },
+    {
+      "epoch": 1.4727158160050409,
+      "grad_norm": 0.18671286603946396,
+      "learning_rate": 5.264406340910148e-07,
+      "loss": 0.2682,
+      "step": 2926
+    },
+    {
+      "epoch": 1.4732199117832387,
+      "grad_norm": 0.17551742566377754,
+      "learning_rate": 5.261724479545974e-07,
+      "loss": 0.2913,
+      "step": 2927
+    },
+    {
+      "epoch": 1.4737240075614366,
+      "grad_norm": 0.24593696823308445,
+      "learning_rate": 5.259042542675907e-07,
+      "loss": 0.2717,
+      "step": 2928
+    },
+    {
+      "epoch": 1.4742281033396345,
+      "grad_norm": 0.1716232359748941,
+      "learning_rate": 5.256360531073674e-07,
+      "loss": 0.2791,
+      "step": 2929
+    },
+    {
+      "epoch": 1.4747321991178324,
+      "grad_norm": 0.17889186640749882,
+      "learning_rate": 5.253678445513014e-07,
+      "loss": 0.2559,
+      "step": 2930
+    },
+    {
+      "epoch": 1.4752362948960303,
+      "grad_norm": 0.16660287168913102,
+      "learning_rate": 5.250996286767693e-07,
+      "loss": 0.2796,
+      "step": 2931
+    },
+    {
+      "epoch": 1.4757403906742281,
+      "grad_norm": 0.17208791704520798,
+      "learning_rate": 5.248314055611499e-07,
+      "loss": 0.2789,
+      "step": 2932
+    },
+    {
+      "epoch": 1.476244486452426,
+      "grad_norm": 0.16785449767453983,
+      "learning_rate": 5.245631752818238e-07,
+      "loss": 0.2517,
+      "step": 2933
+    },
+    {
+      "epoch": 1.4767485822306239,
+      "grad_norm": 0.17104687297149335,
+      "learning_rate": 5.242949379161739e-07,
+      "loss": 0.2669,
+      "step": 2934
+    },
+    {
+      "epoch": 1.4772526780088215,
+      "grad_norm": 0.17816748061415208,
+      "learning_rate": 5.240266935415847e-07,
+      "loss": 0.2761,
+      "step": 2935
+    },
+    {
+      "epoch": 1.4777567737870196,
+      "grad_norm": 0.17497683850506593,
+      "learning_rate": 5.237584422354435e-07,
+      "loss": 0.2691,
+      "step": 2936
+    },
+    {
+      "epoch": 1.4782608695652173,
+      "grad_norm": 0.17896943432738507,
+      "learning_rate": 5.234901840751388e-07,
+      "loss": 0.2691,
+      "step": 2937
+    },
+    {
+      "epoch": 1.4787649653434152,
+      "grad_norm": 0.18082033748198661,
+      "learning_rate": 5.232219191380614e-07,
+      "loss": 0.2726,
+      "step": 2938
+    },
+    {
+      "epoch": 1.479269061121613,
+      "grad_norm": 0.17731545802828333,
+      "learning_rate": 5.229536475016044e-07,
+      "loss": 0.2817,
+      "step": 2939
+    },
+    {
+      "epoch": 1.479773156899811,
+      "grad_norm": 0.17098014110692908,
+      "learning_rate": 5.22685369243162e-07,
+      "loss": 0.2714,
+      "step": 2940
+    },
+    {
+      "epoch": 1.4802772526780088,
+      "grad_norm": 0.1813522016663038,
+      "learning_rate": 5.224170844401313e-07,
+      "loss": 0.2846,
+      "step": 2941
+    },
+    {
+      "epoch": 1.4807813484562067,
+      "grad_norm": 0.17197034879530912,
+      "learning_rate": 5.221487931699107e-07,
+      "loss": 0.2667,
+      "step": 2942
+    },
+    {
+      "epoch": 1.4812854442344046,
+      "grad_norm": 0.16697715148870823,
+      "learning_rate": 5.218804955099003e-07,
+      "loss": 0.2627,
+      "step": 2943
+    },
+    {
+      "epoch": 1.4817895400126024,
+      "grad_norm": 0.17040842764616276,
+      "learning_rate": 5.216121915375026e-07,
+      "loss": 0.2577,
+      "step": 2944
+    },
+    {
+      "epoch": 1.4822936357908003,
+      "grad_norm": 0.17579731593772854,
+      "learning_rate": 5.213438813301214e-07,
+      "loss": 0.2811,
+      "step": 2945
+    },
+    {
+      "epoch": 1.4827977315689982,
+      "grad_norm": 0.17045185683058092,
+      "learning_rate": 5.210755649651627e-07,
+      "loss": 0.255,
+      "step": 2946
+    },
+    {
+      "epoch": 1.483301827347196,
+      "grad_norm": 0.17068201983517348,
+      "learning_rate": 5.20807242520034e-07,
+      "loss": 0.2743,
+      "step": 2947
+    },
+    {
+      "epoch": 1.4838059231253937,
+      "grad_norm": 0.18181611693726946,
+      "learning_rate": 5.205389140721448e-07,
+      "loss": 0.2732,
+      "step": 2948
+    },
+    {
+      "epoch": 1.4843100189035916,
+      "grad_norm": 0.1852929761291196,
+      "learning_rate": 5.202705796989061e-07,
+      "loss": 0.2801,
+      "step": 2949
+    },
+    {
+      "epoch": 1.4848141146817895,
+      "grad_norm": 0.17600122731072515,
+      "learning_rate": 5.200022394777308e-07,
+      "loss": 0.2716,
+      "step": 2950
+    },
+    {
+      "epoch": 1.4853182104599874,
+      "grad_norm": 0.1718236542287459,
+      "learning_rate": 5.197338934860332e-07,
+      "loss": 0.2674,
+      "step": 2951
+    },
+    {
+      "epoch": 1.4858223062381852,
+      "grad_norm": 0.1775790487723521,
+      "learning_rate": 5.194655418012295e-07,
+      "loss": 0.2825,
+      "step": 2952
+    },
+    {
+      "epoch": 1.486326402016383,
+      "grad_norm": 0.1807315629873165,
+      "learning_rate": 5.191971845007378e-07,
+      "loss": 0.3003,
+      "step": 2953
+    },
+    {
+      "epoch": 1.486830497794581,
+      "grad_norm": 0.17385001875969464,
+      "learning_rate": 5.189288216619773e-07,
+      "loss": 0.2586,
+      "step": 2954
+    },
+    {
+      "epoch": 1.4873345935727789,
+      "grad_norm": 0.17077436408789423,
+      "learning_rate": 5.186604533623689e-07,
+      "loss": 0.2774,
+      "step": 2955
+    },
+    {
+      "epoch": 1.4878386893509767,
+      "grad_norm": 0.17354882163404403,
+      "learning_rate": 5.183920796793353e-07,
+      "loss": 0.2695,
+      "step": 2956
+    },
+    {
+      "epoch": 1.4883427851291746,
+      "grad_norm": 0.18056872611734137,
+      "learning_rate": 5.181237006903007e-07,
+      "loss": 0.2575,
+      "step": 2957
+    },
+    {
+      "epoch": 1.4888468809073725,
+      "grad_norm": 0.1747297098238214,
+      "learning_rate": 5.178553164726906e-07,
+      "loss": 0.2832,
+      "step": 2958
+    },
+    {
+      "epoch": 1.4893509766855701,
+      "grad_norm": 0.17268688557025336,
+      "learning_rate": 5.175869271039325e-07,
+      "loss": 0.2745,
+      "step": 2959
+    },
+    {
+      "epoch": 1.4898550724637682,
+      "grad_norm": 0.1723723042058958,
+      "learning_rate": 5.173185326614546e-07,
+      "loss": 0.2835,
+      "step": 2960
+    },
+    {
+      "epoch": 1.490359168241966,
+      "grad_norm": 0.16999642870279735,
+      "learning_rate": 5.170501332226875e-07,
+      "loss": 0.2734,
+      "step": 2961
+    },
+    {
+      "epoch": 1.4908632640201638,
+      "grad_norm": 0.16872935699465505,
+      "learning_rate": 5.167817288650625e-07,
+      "loss": 0.2756,
+      "step": 2962
+    },
+    {
+      "epoch": 1.4913673597983617,
+      "grad_norm": 0.18195730992832942,
+      "learning_rate": 5.165133196660128e-07,
+      "loss": 0.275,
+      "step": 2963
+    },
+    {
+      "epoch": 1.4918714555765595,
+      "grad_norm": 0.1694600503693703,
+      "learning_rate": 5.162449057029725e-07,
+      "loss": 0.2677,
+      "step": 2964
+    },
+    {
+      "epoch": 1.4923755513547574,
+      "grad_norm": 0.17420508681132724,
+      "learning_rate": 5.159764870533777e-07,
+      "loss": 0.2804,
+      "step": 2965
+    },
+    {
+      "epoch": 1.4928796471329553,
+      "grad_norm": 0.16859261160349226,
+      "learning_rate": 5.157080637946654e-07,
+      "loss": 0.2705,
+      "step": 2966
+    },
+    {
+      "epoch": 1.4933837429111532,
+      "grad_norm": 0.17551194260816222,
+      "learning_rate": 5.154396360042738e-07,
+      "loss": 0.2592,
+      "step": 2967
+    },
+    {
+      "epoch": 1.493887838689351,
+      "grad_norm": 0.17227527863330538,
+      "learning_rate": 5.15171203759643e-07,
+      "loss": 0.2674,
+      "step": 2968
+    },
+    {
+      "epoch": 1.494391934467549,
+      "grad_norm": 0.1728755992481674,
+      "learning_rate": 5.149027671382138e-07,
+      "loss": 0.2705,
+      "step": 2969
+    },
+    {
+      "epoch": 1.4948960302457466,
+      "grad_norm": 0.17382519055950477,
+      "learning_rate": 5.146343262174286e-07,
+      "loss": 0.2652,
+      "step": 2970
+    },
+    {
+      "epoch": 1.4954001260239447,
+      "grad_norm": 0.1769092204357377,
+      "learning_rate": 5.14365881074731e-07,
+      "loss": 0.2767,
+      "step": 2971
+    },
+    {
+      "epoch": 1.4959042218021423,
+      "grad_norm": 0.17476416087544372,
+      "learning_rate": 5.140974317875657e-07,
+      "loss": 0.2753,
+      "step": 2972
+    },
+    {
+      "epoch": 1.4964083175803402,
+      "grad_norm": 0.169687985399867,
+      "learning_rate": 5.138289784333787e-07,
+      "loss": 0.2765,
+      "step": 2973
+    },
+    {
+      "epoch": 1.496912413358538,
+      "grad_norm": 0.17261770629037,
+      "learning_rate": 5.13560521089617e-07,
+      "loss": 0.2732,
+      "step": 2974
+    },
+    {
+      "epoch": 1.497416509136736,
+      "grad_norm": 0.17477899451723963,
+      "learning_rate": 5.13292059833729e-07,
+      "loss": 0.2599,
+      "step": 2975
+    },
+    {
+      "epoch": 1.4979206049149338,
+      "grad_norm": 0.18102926267539213,
+      "learning_rate": 5.13023594743164e-07,
+      "loss": 0.2674,
+      "step": 2976
+    },
+    {
+      "epoch": 1.4984247006931317,
+      "grad_norm": 0.17738927553973277,
+      "learning_rate": 5.127551258953727e-07,
+      "loss": 0.2672,
+      "step": 2977
+    },
+    {
+      "epoch": 1.4989287964713296,
+      "grad_norm": 0.18069762077371987,
+      "learning_rate": 5.124866533678066e-07,
+      "loss": 0.2665,
+      "step": 2978
+    },
+    {
+      "epoch": 1.4994328922495275,
+      "grad_norm": 0.17624639732046102,
+      "learning_rate": 5.122181772379182e-07,
+      "loss": 0.2847,
+      "step": 2979
+    },
+    {
+      "epoch": 1.4999369880277253,
+      "grad_norm": 0.21149596045503335,
+      "learning_rate": 5.119496975831616e-07,
+      "loss": 0.2755,
+      "step": 2980
+    },
+    {
+      "epoch": 1.500441083805923,
+      "grad_norm": 0.17200655792505817,
+      "learning_rate": 5.116812144809911e-07,
+      "loss": 0.2776,
+      "step": 2981
+    },
+    {
+      "epoch": 1.500945179584121,
+      "grad_norm": 0.199818514038348,
+      "learning_rate": 5.114127280088627e-07,
+      "loss": 0.2721,
+      "step": 2982
+    },
+    {
+      "epoch": 1.5014492753623188,
+      "grad_norm": 0.1767000934409416,
+      "learning_rate": 5.111442382442328e-07,
+      "loss": 0.2897,
+      "step": 2983
+    },
+    {
+      "epoch": 1.5019533711405169,
+      "grad_norm": 0.17657617178530624,
+      "learning_rate": 5.108757452645594e-07,
+      "loss": 0.2585,
+      "step": 2984
+    },
+    {
+      "epoch": 1.5024574669187145,
+      "grad_norm": 0.17907925642962816,
+      "learning_rate": 5.106072491473008e-07,
+      "loss": 0.2753,
+      "step": 2985
+    },
+    {
+      "epoch": 1.5024574669187145,
+      "eval_loss": 0.30766692757606506,
+      "eval_runtime": 17.4903,
+      "eval_samples_per_second": 48.884,
+      "eval_steps_per_second": 1.029,
+      "step": 2985
+    },
+    {
+      "epoch": 1.5029615626969124,
+      "grad_norm": 0.17603805841778422,
+      "learning_rate": 5.103387499699164e-07,
+      "loss": 0.2658,
+      "step": 2986
+    },
+    {
+      "epoch": 1.5034656584751103,
+      "grad_norm": 0.1772161820090902,
+      "learning_rate": 5.100702478098667e-07,
+      "loss": 0.2757,
+      "step": 2987
+    },
+    {
+      "epoch": 1.5039697542533081,
+      "grad_norm": 0.17226447203172757,
+      "learning_rate": 5.098017427446132e-07,
+      "loss": 0.2765,
+      "step": 2988
+    },
+    {
+      "epoch": 1.504473850031506,
+      "grad_norm": 0.17276877170127075,
+      "learning_rate": 5.095332348516172e-07,
+      "loss": 0.2752,
+      "step": 2989
+    },
+    {
+      "epoch": 1.504977945809704,
+      "grad_norm": 0.1847638673736269,
+      "learning_rate": 5.092647242083423e-07,
+      "loss": 0.2789,
+      "step": 2990
+    },
+    {
+      "epoch": 1.5054820415879018,
+      "grad_norm": 0.17901616149206137,
+      "learning_rate": 5.089962108922517e-07,
+      "loss": 0.2911,
+      "step": 2991
+    },
+    {
+      "epoch": 1.5059861373660994,
+      "grad_norm": 0.18718528180473848,
+      "learning_rate": 5.0872769498081e-07,
+      "loss": 0.2618,
+      "step": 2992
+    },
+    {
+      "epoch": 1.5064902331442975,
+      "grad_norm": 0.17561340990666174,
+      "learning_rate": 5.084591765514824e-07,
+      "loss": 0.2796,
+      "step": 2993
+    },
+    {
+      "epoch": 1.5069943289224952,
+      "grad_norm": 0.17573027328641988,
+      "learning_rate": 5.081906556817348e-07,
+      "loss": 0.2746,
+      "step": 2994
+    },
+    {
+      "epoch": 1.5074984247006933,
+      "grad_norm": 0.1946924830736341,
+      "learning_rate": 5.079221324490338e-07,
+      "loss": 0.2596,
+      "step": 2995
+    },
+    {
+      "epoch": 1.508002520478891,
+      "grad_norm": 0.19483610334363802,
+      "learning_rate": 5.076536069308466e-07,
+      "loss": 0.2661,
+      "step": 2996
+    },
+    {
+      "epoch": 1.5085066162570888,
+      "grad_norm": 0.17050635285106475,
+      "learning_rate": 5.073850792046411e-07,
+      "loss": 0.2845,
+      "step": 2997
+    },
+    {
+      "epoch": 1.5090107120352867,
+      "grad_norm": 0.17434918537651012,
+      "learning_rate": 5.071165493478862e-07,
+      "loss": 0.2763,
+      "step": 2998
+    },
+    {
+      "epoch": 1.5095148078134846,
+      "grad_norm": 0.17421166183775727,
+      "learning_rate": 5.068480174380507e-07,
+      "loss": 0.2849,
+      "step": 2999
+    },
+    {
+      "epoch": 1.5100189035916824,
+      "grad_norm": 0.17727506977234753,
+      "learning_rate": 5.065794835526047e-07,
+      "loss": 0.266,
+      "step": 3000
+    },
+    {
+      "epoch": 1.5105229993698803,
+      "grad_norm": 0.17988130416225281,
+      "learning_rate": 5.063109477690186e-07,
+      "loss": 0.2621,
+      "step": 3001
+    },
+    {
+      "epoch": 1.5110270951480782,
+      "grad_norm": 0.17698102799975615,
+      "learning_rate": 5.060424101647631e-07,
+      "loss": 0.2805,
+      "step": 3002
+    },
+    {
+      "epoch": 1.5115311909262759,
+      "grad_norm": 0.1800412666094284,
+      "learning_rate": 5.057738708173096e-07,
+      "loss": 0.2749,
+      "step": 3003
+    },
+    {
+      "epoch": 1.512035286704474,
+      "grad_norm": 0.17319416297445328,
+      "learning_rate": 5.055053298041302e-07,
+      "loss": 0.2782,
+      "step": 3004
+    },
+    {
+      "epoch": 1.5125393824826716,
+      "grad_norm": 0.1683959181018551,
+      "learning_rate": 5.052367872026971e-07,
+      "loss": 0.2595,
+      "step": 3005
+    },
+    {
+      "epoch": 1.5130434782608697,
+      "grad_norm": 0.19739859925036402,
+      "learning_rate": 5.049682430904835e-07,
+      "loss": 0.2671,
+      "step": 3006
+    },
+    {
+      "epoch": 1.5135475740390674,
+      "grad_norm": 0.17458057209940056,
+      "learning_rate": 5.046996975449624e-07,
+      "loss": 0.2812,
+      "step": 3007
+    },
+    {
+      "epoch": 1.5140516698172652,
+      "grad_norm": 0.17642668853441784,
+      "learning_rate": 5.044311506436077e-07,
+      "loss": 0.254,
+      "step": 3008
+    },
+    {
+      "epoch": 1.5145557655954631,
+      "grad_norm": 0.18216491050641825,
+      "learning_rate": 5.041626024638935e-07,
+      "loss": 0.2928,
+      "step": 3009
+    },
+    {
+      "epoch": 1.515059861373661,
+      "grad_norm": 0.16982395667266062,
+      "learning_rate": 5.038940530832944e-07,
+      "loss": 0.2792,
+      "step": 3010
+    },
+    {
+      "epoch": 1.5155639571518589,
+      "grad_norm": 0.17736924822640682,
+      "learning_rate": 5.03625502579285e-07,
+      "loss": 0.27,
+      "step": 3011
+    },
+    {
+      "epoch": 1.5160680529300568,
+      "grad_norm": 0.17990063325148373,
+      "learning_rate": 5.033569510293406e-07,
+      "loss": 0.2742,
+      "step": 3012
+    },
+    {
+      "epoch": 1.5165721487082546,
+      "grad_norm": 0.17269683446652181,
+      "learning_rate": 5.030883985109367e-07,
+      "loss": 0.2718,
+      "step": 3013
+    },
+    {
+      "epoch": 1.5170762444864523,
+      "grad_norm": 0.1713114478517227,
+      "learning_rate": 5.028198451015488e-07,
+      "loss": 0.292,
+      "step": 3014
+    },
+    {
+      "epoch": 1.5175803402646504,
+      "grad_norm": 0.17237373122427047,
+      "learning_rate": 5.025512908786531e-07,
+      "loss": 0.2708,
+      "step": 3015
+    },
+    {
+      "epoch": 1.518084436042848,
+      "grad_norm": 0.17979894721140827,
+      "learning_rate": 5.022827359197259e-07,
+      "loss": 0.2749,
+      "step": 3016
+    },
+    {
+      "epoch": 1.5185885318210461,
+      "grad_norm": 0.18029323092096203,
+      "learning_rate": 5.020141803022435e-07,
+      "loss": 0.2718,
+      "step": 3017
+    },
+    {
+      "epoch": 1.5190926275992438,
+      "grad_norm": 0.1709949058851969,
+      "learning_rate": 5.017456241036826e-07,
+      "loss": 0.2842,
+      "step": 3018
+    },
+    {
+      "epoch": 1.5195967233774417,
+      "grad_norm": 0.17615800092590866,
+      "learning_rate": 5.014770674015199e-07,
+      "loss": 0.2556,
+      "step": 3019
+    },
+    {
+      "epoch": 1.5201008191556395,
+      "grad_norm": 0.17565842258793046,
+      "learning_rate": 5.012085102732323e-07,
+      "loss": 0.2704,
+      "step": 3020
+    },
+    {
+      "epoch": 1.5206049149338374,
+      "grad_norm": 0.1706652936469358,
+      "learning_rate": 5.00939952796297e-07,
+      "loss": 0.2751,
+      "step": 3021
+    },
+    {
+      "epoch": 1.5211090107120353,
+      "grad_norm": 0.17904301457889313,
+      "learning_rate": 5.006713950481911e-07,
+      "loss": 0.2708,
+      "step": 3022
+    },
+    {
+      "epoch": 1.5216131064902332,
+      "grad_norm": 0.19348962154742488,
+      "learning_rate": 5.00402837106392e-07,
+      "loss": 0.2691,
+      "step": 3023
+    },
+    {
+      "epoch": 1.522117202268431,
+      "grad_norm": 0.17760169438436854,
+      "learning_rate": 5.001342790483769e-07,
+      "loss": 0.2812,
+      "step": 3024
+    },
+    {
+      "epoch": 1.5226212980466287,
+      "grad_norm": 0.1874622214559223,
+      "learning_rate": 4.998657209516231e-07,
+      "loss": 0.2703,
+      "step": 3025
+    },
+    {
+      "epoch": 1.5231253938248268,
+      "grad_norm": 0.19774473741326665,
+      "learning_rate": 4.995971628936078e-07,
+      "loss": 0.2642,
+      "step": 3026
+    },
+    {
+      "epoch": 1.5236294896030245,
+      "grad_norm": 0.19986370176179066,
+      "learning_rate": 4.993286049518088e-07,
+      "loss": 0.2949,
+      "step": 3027
+    },
+    {
+      "epoch": 1.5241335853812226,
+      "grad_norm": 0.16961474491016765,
+      "learning_rate": 4.990600472037029e-07,
+      "loss": 0.2767,
+      "step": 3028
+    },
+    {
+      "epoch": 1.5246376811594202,
+      "grad_norm": 0.18853446766118223,
+      "learning_rate": 4.987914897267678e-07,
+      "loss": 0.2781,
+      "step": 3029
+    },
+    {
+      "epoch": 1.525141776937618,
+      "grad_norm": 0.16987515007270199,
+      "learning_rate": 4.985229325984803e-07,
+      "loss": 0.2882,
+      "step": 3030
+    },
+    {
+      "epoch": 1.525645872715816,
+      "grad_norm": 0.1714486129335063,
+      "learning_rate": 4.982543758963174e-07,
+      "loss": 0.2704,
+      "step": 3031
+    },
+    {
+      "epoch": 1.5261499684940139,
+      "grad_norm": 0.16809069324008713,
+      "learning_rate": 4.979858196977566e-07,
+      "loss": 0.2591,
+      "step": 3032
+    },
+    {
+      "epoch": 1.5266540642722117,
+      "grad_norm": 0.1738515169060616,
+      "learning_rate": 4.977172640802741e-07,
+      "loss": 0.2582,
+      "step": 3033
+    },
+    {
+      "epoch": 1.5271581600504096,
+      "grad_norm": 0.17664113554519453,
+      "learning_rate": 4.974487091213469e-07,
+      "loss": 0.267,
+      "step": 3034
+    },
+    {
+      "epoch": 1.5276622558286075,
+      "grad_norm": 0.1916027084703846,
+      "learning_rate": 4.971801548984511e-07,
+      "loss": 0.2731,
+      "step": 3035
+    },
+    {
+      "epoch": 1.5281663516068051,
+      "grad_norm": 0.17313830459160134,
+      "learning_rate": 4.969116014890634e-07,
+      "loss": 0.2806,
+      "step": 3036
+    },
+    {
+      "epoch": 1.5286704473850032,
+      "grad_norm": 0.20413150350420245,
+      "learning_rate": 4.966430489706594e-07,
+      "loss": 0.2745,
+      "step": 3037
+    },
+    {
+      "epoch": 1.529174543163201,
+      "grad_norm": 0.1747943524307337,
+      "learning_rate": 4.96374497420715e-07,
+      "loss": 0.2634,
+      "step": 3038
+    },
+    {
+      "epoch": 1.529678638941399,
+      "grad_norm": 0.1728937587667476,
+      "learning_rate": 4.961059469167056e-07,
+      "loss": 0.2814,
+      "step": 3039
+    },
+    {
+      "epoch": 1.5301827347195966,
+      "grad_norm": 0.1743204044600501,
+      "learning_rate": 4.958373975361063e-07,
+      "loss": 0.2551,
+      "step": 3040
+    },
+    {
+      "epoch": 1.5306868304977945,
+      "grad_norm": 0.17502392690878435,
+      "learning_rate": 4.955688493563922e-07,
+      "loss": 0.267,
+      "step": 3041
+    },
+    {
+      "epoch": 1.5311909262759924,
+      "grad_norm": 0.19500065101179812,
+      "learning_rate": 4.953003024550375e-07,
+      "loss": 0.2819,
+      "step": 3042
+    },
+    {
+      "epoch": 1.5316950220541903,
+      "grad_norm": 0.17254980716871499,
+      "learning_rate": 4.950317569095166e-07,
+      "loss": 0.2649,
+      "step": 3043
+    },
+    {
+      "epoch": 1.5321991178323882,
+      "grad_norm": 0.1713124294788534,
+      "learning_rate": 4.94763212797303e-07,
+      "loss": 0.2563,
+      "step": 3044
+    },
+    {
+      "epoch": 1.532703213610586,
+      "grad_norm": 0.17577986512817073,
+      "learning_rate": 4.944946701958698e-07,
+      "loss": 0.273,
+      "step": 3045
+    },
+    {
+      "epoch": 1.533207309388784,
+      "grad_norm": 0.1701862456963333,
+      "learning_rate": 4.942261291826905e-07,
+      "loss": 0.2814,
+      "step": 3046
+    },
+    {
+      "epoch": 1.5337114051669816,
+      "grad_norm": 0.18127900976396943,
+      "learning_rate": 4.93957589835237e-07,
+      "loss": 0.2713,
+      "step": 3047
+    },
+    {
+      "epoch": 1.5342155009451797,
+      "grad_norm": 0.17254874589059224,
+      "learning_rate": 4.936890522309815e-07,
+      "loss": 0.2734,
+      "step": 3048
+    },
+    {
+      "epoch": 1.5347195967233773,
+      "grad_norm": 0.1717770510577077,
+      "learning_rate": 4.934205164473952e-07,
+      "loss": 0.2773,
+      "step": 3049
+    },
+    {
+      "epoch": 1.5352236925015754,
+      "grad_norm": 0.1908978528330254,
+      "learning_rate": 4.931519825619493e-07,
+      "loss": 0.2788,
+      "step": 3050
+    },
+    {
+      "epoch": 1.535727788279773,
+      "grad_norm": 0.171288059642186,
+      "learning_rate": 4.928834506521138e-07,
+      "loss": 0.2703,
+      "step": 3051
+    },
+    {
+      "epoch": 1.5362318840579712,
+      "grad_norm": 0.1675001073197896,
+      "learning_rate": 4.926149207953588e-07,
+      "loss": 0.2642,
+      "step": 3052
+    },
+    {
+      "epoch": 1.5367359798361688,
+      "grad_norm": 0.17021689322972322,
+      "learning_rate": 4.923463930691535e-07,
+      "loss": 0.2702,
+      "step": 3053
+    },
+    {
+      "epoch": 1.5372400756143667,
+      "grad_norm": 0.17371445641850253,
+      "learning_rate": 4.920778675509662e-07,
+      "loss": 0.2589,
+      "step": 3054
+    },
+    {
+      "epoch": 1.5377441713925646,
+      "grad_norm": 0.17478774431954305,
+      "learning_rate": 4.918093443182652e-07,
+      "loss": 0.2803,
+      "step": 3055
+    },
+    {
+      "epoch": 1.5382482671707625,
+      "grad_norm": 0.17591722779410152,
+      "learning_rate": 4.915408234485175e-07,
+      "loss": 0.2897,
+      "step": 3056
+    },
+    {
+      "epoch": 1.5387523629489603,
+      "grad_norm": 0.17280009330380483,
+      "learning_rate": 4.912723050191899e-07,
+      "loss": 0.2717,
+      "step": 3057
+    },
+    {
+      "epoch": 1.539256458727158,
+      "grad_norm": 0.16969841106908495,
+      "learning_rate": 4.910037891077482e-07,
+      "loss": 0.2602,
+      "step": 3058
+    },
+    {
+      "epoch": 1.539760554505356,
+      "grad_norm": 0.16865488414462979,
+      "learning_rate": 4.907352757916577e-07,
+      "loss": 0.2724,
+      "step": 3059
+    },
+    {
+      "epoch": 1.5402646502835537,
+      "grad_norm": 0.17584292304841348,
+      "learning_rate": 4.904667651483828e-07,
+      "loss": 0.2742,
+      "step": 3060
+    },
+    {
+      "epoch": 1.5407687460617518,
+      "grad_norm": 0.16673800323953514,
+      "learning_rate": 4.901982572553869e-07,
+      "loss": 0.2695,
+      "step": 3061
+    },
+    {
+      "epoch": 1.5412728418399495,
+      "grad_norm": 0.18326131122744185,
+      "learning_rate": 4.899297521901333e-07,
+      "loss": 0.2789,
+      "step": 3062
+    },
+    {
+      "epoch": 1.5417769376181476,
+      "grad_norm": 0.18805933934917177,
+      "learning_rate": 4.896612500300835e-07,
+      "loss": 0.272,
+      "step": 3063
+    },
+    {
+      "epoch": 1.5422810333963453,
+      "grad_norm": 0.1712806706085473,
+      "learning_rate": 4.893927508526993e-07,
+      "loss": 0.2669,
+      "step": 3064
+    },
+    {
+      "epoch": 1.5427851291745431,
+      "grad_norm": 0.1731870501531892,
+      "learning_rate": 4.891242547354406e-07,
+      "loss": 0.2812,
+      "step": 3065
+    },
+    {
+      "epoch": 1.543289224952741,
+      "grad_norm": 0.17506473097896597,
+      "learning_rate": 4.888557617557672e-07,
+      "loss": 0.2813,
+      "step": 3066
+    },
+    {
+      "epoch": 1.543793320730939,
+      "grad_norm": 0.17307157635246334,
+      "learning_rate": 4.885872719911375e-07,
+      "loss": 0.2695,
+      "step": 3067
+    },
+    {
+      "epoch": 1.5442974165091368,
+      "grad_norm": 0.17106830633462392,
+      "learning_rate": 4.883187855190089e-07,
+      "loss": 0.264,
+      "step": 3068
+    },
+    {
+      "epoch": 1.5448015122873346,
+      "grad_norm": 0.17296030737339352,
+      "learning_rate": 4.880503024168384e-07,
+      "loss": 0.2776,
+      "step": 3069
+    },
+    {
+      "epoch": 1.5453056080655325,
+      "grad_norm": 0.17398895118716073,
+      "learning_rate": 4.877818227620816e-07,
+      "loss": 0.2641,
+      "step": 3070
+    },
+    {
+      "epoch": 1.5458097038437302,
+      "grad_norm": 0.16983527604252388,
+      "learning_rate": 4.875133466321934e-07,
+      "loss": 0.2791,
+      "step": 3071
+    },
+    {
+      "epoch": 1.5463137996219283,
+      "grad_norm": 0.17446130630330534,
+      "learning_rate": 4.872448741046272e-07,
+      "loss": 0.2709,
+      "step": 3072
+    },
+    {
+      "epoch": 1.546817895400126,
+      "grad_norm": 0.17284082991627042,
+      "learning_rate": 4.86976405256836e-07,
+      "loss": 0.2671,
+      "step": 3073
+    },
+    {
+      "epoch": 1.547321991178324,
+      "grad_norm": 0.182632054641719,
+      "learning_rate": 4.867079401662711e-07,
+      "loss": 0.2776,
+      "step": 3074
+    },
+    {
+      "epoch": 1.5478260869565217,
+      "grad_norm": 0.17300501368762167,
+      "learning_rate": 4.864394789103829e-07,
+      "loss": 0.2794,
+      "step": 3075
+    },
+    {
+      "epoch": 1.5483301827347196,
+      "grad_norm": 0.17667332011696144,
+      "learning_rate": 4.861710215666213e-07,
+      "loss": 0.2853,
+      "step": 3076
+    },
+    {
+      "epoch": 1.5488342785129174,
+      "grad_norm": 0.16996586627278085,
+      "learning_rate": 4.859025682124341e-07,
+      "loss": 0.2786,
+      "step": 3077
+    },
+    {
+      "epoch": 1.5493383742911153,
+      "grad_norm": 0.17057715081373995,
+      "learning_rate": 4.85634118925269e-07,
+      "loss": 0.2691,
+      "step": 3078
+    },
+    {
+      "epoch": 1.5498424700693132,
+      "grad_norm": 0.17161445251025925,
+      "learning_rate": 4.853656737825713e-07,
+      "loss": 0.283,
+      "step": 3079
+    },
+    {
+      "epoch": 1.550346565847511,
+      "grad_norm": 0.17675037956109935,
+      "learning_rate": 4.850972328617863e-07,
+      "loss": 0.2721,
+      "step": 3080
+    },
+    {
+      "epoch": 1.550850661625709,
+      "grad_norm": 0.17208347939492113,
+      "learning_rate": 4.848287962403571e-07,
+      "loss": 0.2653,
+      "step": 3081
+    },
+    {
+      "epoch": 1.5513547574039066,
+      "grad_norm": 0.17004686976042202,
+      "learning_rate": 4.845603639957263e-07,
+      "loss": 0.2709,
+      "step": 3082
+    },
+    {
+      "epoch": 1.5518588531821047,
+      "grad_norm": 0.17995963542830856,
+      "learning_rate": 4.842919362053348e-07,
+      "loss": 0.265,
+      "step": 3083
+    },
+    {
+      "epoch": 1.5523629489603024,
+      "grad_norm": 0.18099808248082644,
+      "learning_rate": 4.840235129466222e-07,
+      "loss": 0.2651,
+      "step": 3084
+    },
+    {
+      "epoch": 1.5528670447385005,
+      "grad_norm": 0.17473760103531463,
+      "learning_rate": 4.837550942970275e-07,
+      "loss": 0.2841,
+      "step": 3085
+    },
+    {
+      "epoch": 1.5533711405166981,
+      "grad_norm": 0.19493429052326036,
+      "learning_rate": 4.834866803339872e-07,
+      "loss": 0.2789,
+      "step": 3086
+    },
+    {
+      "epoch": 1.553875236294896,
+      "grad_norm": 0.174886487259404,
+      "learning_rate": 4.832182711349374e-07,
+      "loss": 0.2897,
+      "step": 3087
+    },
+    {
+      "epoch": 1.5543793320730939,
+      "grad_norm": 0.1873744343078876,
+      "learning_rate": 4.829498667773126e-07,
+      "loss": 0.2596,
+      "step": 3088
+    },
+    {
+      "epoch": 1.5548834278512917,
+      "grad_norm": 0.1776819518952955,
+      "learning_rate": 4.826814673385454e-07,
+      "loss": 0.2592,
+      "step": 3089
+    },
+    {
+      "epoch": 1.5553875236294896,
+      "grad_norm": 0.17480879943963698,
+      "learning_rate": 4.824130728960677e-07,
+      "loss": 0.2854,
+      "step": 3090
+    },
+    {
+      "epoch": 1.5558916194076875,
+      "grad_norm": 0.17863519439054704,
+      "learning_rate": 4.821446835273093e-07,
+      "loss": 0.2848,
+      "step": 3091
+    },
+    {
+      "epoch": 1.5563957151858854,
+      "grad_norm": 0.20186594037979866,
+      "learning_rate": 4.818762993096994e-07,
+      "loss": 0.2742,
+      "step": 3092
+    },
+    {
+      "epoch": 1.556899810964083,
+      "grad_norm": 0.17253352705225278,
+      "learning_rate": 4.816079203206648e-07,
+      "loss": 0.2735,
+      "step": 3093
+    },
+    {
+      "epoch": 1.5574039067422811,
+      "grad_norm": 0.17021429772016514,
+      "learning_rate": 4.813395466376311e-07,
+      "loss": 0.2757,
+      "step": 3094
+    },
+    {
+      "epoch": 1.5579080025204788,
+      "grad_norm": 0.17831771845073935,
+      "learning_rate": 4.810711783380227e-07,
+      "loss": 0.2899,
+      "step": 3095
+    },
+    {
+      "epoch": 1.5584120982986769,
+      "grad_norm": 0.17539283313588858,
+      "learning_rate": 4.808028154992622e-07,
+      "loss": 0.2803,
+      "step": 3096
+    },
+    {
+      "epoch": 1.5589161940768745,
+      "grad_norm": 0.17570461231863696,
+      "learning_rate": 4.805344581987704e-07,
+      "loss": 0.2702,
+      "step": 3097
+    },
+    {
+      "epoch": 1.5594202898550724,
+      "grad_norm": 0.18256077525124403,
+      "learning_rate": 4.802661065139667e-07,
+      "loss": 0.279,
+      "step": 3098
+    },
+    {
+      "epoch": 1.5599243856332703,
+      "grad_norm": 0.1773908250652285,
+      "learning_rate": 4.799977605222693e-07,
+      "loss": 0.283,
+      "step": 3099
+    },
+    {
+      "epoch": 1.5604284814114682,
+      "grad_norm": 0.17701100191661834,
+      "learning_rate": 4.797294203010939e-07,
+      "loss": 0.2796,
+      "step": 3100
+    },
+    {
+      "epoch": 1.560932577189666,
+      "grad_norm": 0.17106773731405026,
+      "learning_rate": 4.794610859278552e-07,
+      "loss": 0.2669,
+      "step": 3101
+    },
+    {
+      "epoch": 1.561436672967864,
+      "grad_norm": 0.17235342667180142,
+      "learning_rate": 4.791927574799659e-07,
+      "loss": 0.2642,
+      "step": 3102
+    },
+    {
+      "epoch": 1.5619407687460618,
+      "grad_norm": 0.1791600844603549,
+      "learning_rate": 4.789244350348374e-07,
+      "loss": 0.2687,
+      "step": 3103
+    },
+    {
+      "epoch": 1.5624448645242595,
+      "grad_norm": 0.17661856365985878,
+      "learning_rate": 4.786561186698788e-07,
+      "loss": 0.2794,
+      "step": 3104
+    },
+    {
+      "epoch": 1.5629489603024576,
+      "grad_norm": 0.1755295092888704,
+      "learning_rate": 4.783878084624975e-07,
+      "loss": 0.2747,
+      "step": 3105
+    },
+    {
+      "epoch": 1.5634530560806552,
+      "grad_norm": 0.19305165095057605,
+      "learning_rate": 4.781195044900998e-07,
+      "loss": 0.2787,
+      "step": 3106
+    },
+    {
+      "epoch": 1.5639571518588533,
+      "grad_norm": 0.18063804583238724,
+      "learning_rate": 4.778512068300893e-07,
+      "loss": 0.2754,
+      "step": 3107
+    },
+    {
+      "epoch": 1.564461247637051,
+      "grad_norm": 0.18403206850653575,
+      "learning_rate": 4.775829155598686e-07,
+      "loss": 0.2778,
+      "step": 3108
+    },
+    {
+      "epoch": 1.5649653434152488,
+      "grad_norm": 0.1700956199275683,
+      "learning_rate": 4.773146307568379e-07,
+      "loss": 0.2727,
+      "step": 3109
+    },
+    {
+      "epoch": 1.5654694391934467,
+      "grad_norm": 0.18086574577750156,
+      "learning_rate": 4.770463524983956e-07,
+      "loss": 0.2912,
+      "step": 3110
+    },
+    {
+      "epoch": 1.5659735349716446,
+      "grad_norm": 0.17147115049908643,
+      "learning_rate": 4.7677808086193854e-07,
+      "loss": 0.27,
+      "step": 3111
+    },
+    {
+      "epoch": 1.5664776307498425,
+      "grad_norm": 0.1738895601221236,
+      "learning_rate": 4.7650981592486123e-07,
+      "loss": 0.2673,
+      "step": 3112
+    },
+    {
+      "epoch": 1.5669817265280404,
+      "grad_norm": 0.180723254015497,
+      "learning_rate": 4.7624155776455647e-07,
+      "loss": 0.2633,
+      "step": 3113
+    },
+    {
+      "epoch": 1.5674858223062382,
+      "grad_norm": 0.175819284778621,
+      "learning_rate": 4.7597330645841515e-07,
+      "loss": 0.2582,
+      "step": 3114
+    },
+    {
+      "epoch": 1.5679899180844359,
+      "grad_norm": 0.1753712910797449,
+      "learning_rate": 4.757050620838262e-07,
+      "loss": 0.2632,
+      "step": 3115
+    },
+    {
+      "epoch": 1.568494013862634,
+      "grad_norm": 0.17369096100282744,
+      "learning_rate": 4.754368247181761e-07,
+      "loss": 0.2798,
+      "step": 3116
+    },
+    {
+      "epoch": 1.5689981096408316,
+      "grad_norm": 0.17849770678240312,
+      "learning_rate": 4.751685944388501e-07,
+      "loss": 0.2842,
+      "step": 3117
+    },
+    {
+      "epoch": 1.5695022054190297,
+      "grad_norm": 0.19881328567873902,
+      "learning_rate": 4.749003713232308e-07,
+      "loss": 0.2733,
+      "step": 3118
+    },
+    {
+      "epoch": 1.5700063011972274,
+      "grad_norm": 0.1774218478893288,
+      "learning_rate": 4.7463215544869865e-07,
+      "loss": 0.2863,
+      "step": 3119
+    },
+    {
+      "epoch": 1.5705103969754255,
+      "grad_norm": 0.1733835174988626,
+      "learning_rate": 4.743639468926328e-07,
+      "loss": 0.2681,
+      "step": 3120
+    },
+    {
+      "epoch": 1.5710144927536231,
+      "grad_norm": 0.1736480466599747,
+      "learning_rate": 4.740957457324092e-07,
+      "loss": 0.2584,
+      "step": 3121
+    },
+    {
+      "epoch": 1.571518588531821,
+      "grad_norm": 0.17950796056364546,
+      "learning_rate": 4.738275520454027e-07,
+      "loss": 0.2508,
+      "step": 3122
+    },
+    {
+      "epoch": 1.572022684310019,
+      "grad_norm": 0.1778356015662689,
+      "learning_rate": 4.735593659089851e-07,
+      "loss": 0.2576,
+      "step": 3123
+    },
+    {
+      "epoch": 1.5725267800882168,
+      "grad_norm": 0.1687063104049516,
+      "learning_rate": 4.732911874005269e-07,
+      "loss": 0.2559,
+      "step": 3124
+    },
+    {
+      "epoch": 1.5730308758664147,
+      "grad_norm": 0.17493624040875075,
+      "learning_rate": 4.7302301659739547e-07,
+      "loss": 0.2648,
+      "step": 3125
+    },
+    {
+      "epoch": 1.5735349716446123,
+      "grad_norm": 0.17507468008329716,
+      "learning_rate": 4.7275485357695673e-07,
+      "loss": 0.2568,
+      "step": 3126
+    },
+    {
+      "epoch": 1.5740390674228104,
+      "grad_norm": 0.17975235409806142,
+      "learning_rate": 4.7248669841657404e-07,
+      "loss": 0.2776,
+      "step": 3127
+    },
+    {
+      "epoch": 1.574543163201008,
+      "grad_norm": 0.17454934150306223,
+      "learning_rate": 4.7221855119360824e-07,
+      "loss": 0.2681,
+      "step": 3128
+    },
+    {
+      "epoch": 1.5750472589792062,
+      "grad_norm": 0.18520223540546543,
+      "learning_rate": 4.7195041198541854e-07,
+      "loss": 0.2727,
+      "step": 3129
+    },
+    {
+      "epoch": 1.5755513547574038,
+      "grad_norm": 0.17055487749470938,
+      "learning_rate": 4.7168228086936096e-07,
+      "loss": 0.2692,
+      "step": 3130
+    },
+    {
+      "epoch": 1.576055450535602,
+      "grad_norm": 0.1751981137839451,
+      "learning_rate": 4.7141415792279015e-07,
+      "loss": 0.2754,
+      "step": 3131
+    },
+    {
+      "epoch": 1.5765595463137996,
+      "grad_norm": 0.180062722006383,
+      "learning_rate": 4.7114604322305747e-07,
+      "loss": 0.264,
+      "step": 3132
+    },
+    {
+      "epoch": 1.5770636420919975,
+      "grad_norm": 0.17530448742608076,
+      "learning_rate": 4.708779368475128e-07,
+      "loss": 0.2875,
+      "step": 3133
+    },
+    {
+      "epoch": 1.5775677378701953,
+      "grad_norm": 0.1716762643727378,
+      "learning_rate": 4.70609838873503e-07,
+      "loss": 0.27,
+      "step": 3134
+    },
+    {
+      "epoch": 1.5780718336483932,
+      "grad_norm": 0.17341353040640353,
+      "learning_rate": 4.703417493783723e-07,
+      "loss": 0.2623,
+      "step": 3135
+    },
+    {
+      "epoch": 1.578575929426591,
+      "grad_norm": 0.2080724128483008,
+      "learning_rate": 4.700736684394635e-07,
+      "loss": 0.2823,
+      "step": 3136
+    },
+    {
+      "epoch": 1.579080025204789,
+      "grad_norm": 0.17039241780253916,
+      "learning_rate": 4.6980559613411576e-07,
+      "loss": 0.2722,
+      "step": 3137
+    },
+    {
+      "epoch": 1.5795841209829868,
+      "grad_norm": 0.1742720965174898,
+      "learning_rate": 4.695375325396666e-07,
+      "loss": 0.2745,
+      "step": 3138
+    },
+    {
+      "epoch": 1.5800882167611845,
+      "grad_norm": 0.17895066474047672,
+      "learning_rate": 4.692694777334505e-07,
+      "loss": 0.2664,
+      "step": 3139
+    },
+    {
+      "epoch": 1.5805923125393826,
+      "grad_norm": 0.1779553651912266,
+      "learning_rate": 4.6900143179279984e-07,
+      "loss": 0.2729,
+      "step": 3140
+    },
+    {
+      "epoch": 1.5810964083175802,
+      "grad_norm": 0.1757573994664284,
+      "learning_rate": 4.687333947950441e-07,
+      "loss": 0.2628,
+      "step": 3141
+    },
+    {
+      "epoch": 1.5816005040957783,
+      "grad_norm": 0.17335221224324718,
+      "learning_rate": 4.684653668175102e-07,
+      "loss": 0.2564,
+      "step": 3142
+    },
+    {
+      "epoch": 1.582104599873976,
+      "grad_norm": 0.17526400475703355,
+      "learning_rate": 4.6819734793752257e-07,
+      "loss": 0.2724,
+      "step": 3143
+    },
+    {
+      "epoch": 1.5826086956521739,
+      "grad_norm": 0.17495007637044283,
+      "learning_rate": 4.679293382324031e-07,
+      "loss": 0.2844,
+      "step": 3144
+    },
+    {
+      "epoch": 1.5831127914303718,
+      "grad_norm": 0.17017524004684992,
+      "learning_rate": 4.67661337779471e-07,
+      "loss": 0.2772,
+      "step": 3145
+    },
+    {
+      "epoch": 1.5836168872085696,
+      "grad_norm": 0.17206013880701826,
+      "learning_rate": 4.6739334665604234e-07,
+      "loss": 0.2839,
+      "step": 3146
+    },
+    {
+      "epoch": 1.5841209829867675,
+      "grad_norm": 0.16939135952437226,
+      "learning_rate": 4.671253649394315e-07,
+      "loss": 0.2684,
+      "step": 3147
+    },
+    {
+      "epoch": 1.5846250787649654,
+      "grad_norm": 0.18799473978425724,
+      "learning_rate": 4.668573927069491e-07,
+      "loss": 0.2622,
+      "step": 3148
+    },
+    {
+      "epoch": 1.5851291745431633,
+      "grad_norm": 0.17747869099450092,
+      "learning_rate": 4.665894300359035e-07,
+      "loss": 0.2616,
+      "step": 3149
+    },
+    {
+      "epoch": 1.585633270321361,
+      "grad_norm": 0.17806445921975966,
+      "learning_rate": 4.6632147700360055e-07,
+      "loss": 0.2607,
+      "step": 3150
+    },
+    {
+      "epoch": 1.586137366099559,
+      "grad_norm": 0.17372413070642495,
+      "learning_rate": 4.6605353368734265e-07,
+      "loss": 0.2709,
+      "step": 3151
+    },
+    {
+      "epoch": 1.5866414618777567,
+      "grad_norm": 0.16458604740808605,
+      "learning_rate": 4.6578560016443013e-07,
+      "loss": 0.26,
+      "step": 3152
+    },
+    {
+      "epoch": 1.5871455576559548,
+      "grad_norm": 0.17882994895541682,
+      "learning_rate": 4.655176765121599e-07,
+      "loss": 0.27,
+      "step": 3153
+    },
+    {
+      "epoch": 1.5876496534341524,
+      "grad_norm": 0.17487944942600261,
+      "learning_rate": 4.652497628078266e-07,
+      "loss": 0.2579,
+      "step": 3154
+    },
+    {
+      "epoch": 1.5881537492123503,
+      "grad_norm": 0.17194013515573972,
+      "learning_rate": 4.6498185912872137e-07,
+      "loss": 0.2749,
+      "step": 3155
+    },
+    {
+      "epoch": 1.5886578449905482,
+      "grad_norm": 0.17344590975757027,
+      "learning_rate": 4.6471396555213273e-07,
+      "loss": 0.2625,
+      "step": 3156
+    },
+    {
+      "epoch": 1.589161940768746,
+      "grad_norm": 0.17496373280977637,
+      "learning_rate": 4.6444608215534657e-07,
+      "loss": 0.2715,
+      "step": 3157
+    },
+    {
+      "epoch": 1.589666036546944,
+      "grad_norm": 0.17877746835344072,
+      "learning_rate": 4.641782090156454e-07,
+      "loss": 0.2544,
+      "step": 3158
+    },
+    {
+      "epoch": 1.5901701323251418,
+      "grad_norm": 0.18160946077780746,
+      "learning_rate": 4.6391034621030903e-07,
+      "loss": 0.271,
+      "step": 3159
+    },
+    {
+      "epoch": 1.5906742281033397,
+      "grad_norm": 0.17284059925719977,
+      "learning_rate": 4.636424938166142e-07,
+      "loss": 0.2754,
+      "step": 3160
+    },
+    {
+      "epoch": 1.5911783238815373,
+      "grad_norm": 0.17028743922696857,
+      "learning_rate": 4.633746519118348e-07,
+      "loss": 0.263,
+      "step": 3161
+    },
+    {
+      "epoch": 1.5916824196597354,
+      "grad_norm": 0.1747381029040225,
+      "learning_rate": 4.631068205732413e-07,
+      "loss": 0.2684,
+      "step": 3162
+    },
+    {
+      "epoch": 1.592186515437933,
+      "grad_norm": 0.18722584514120708,
+      "learning_rate": 4.6283899987810164e-07,
+      "loss": 0.2854,
+      "step": 3163
+    },
+    {
+      "epoch": 1.5926906112161312,
+      "grad_norm": 0.16942535047104845,
+      "learning_rate": 4.6257118990368036e-07,
+      "loss": 0.2793,
+      "step": 3164
+    },
+    {
+      "epoch": 1.5931947069943289,
+      "grad_norm": 0.16883072945164623,
+      "learning_rate": 4.6230339072723874e-07,
+      "loss": 0.2707,
+      "step": 3165
+    },
+    {
+      "epoch": 1.5936988027725267,
+      "grad_norm": 0.17706329109752933,
+      "learning_rate": 4.6203560242603556e-07,
+      "loss": 0.2726,
+      "step": 3166
+    },
+    {
+      "epoch": 1.5942028985507246,
+      "grad_norm": 0.1686584353657218,
+      "learning_rate": 4.617678250773256e-07,
+      "loss": 0.2666,
+      "step": 3167
+    },
+    {
+      "epoch": 1.5947069943289225,
+      "grad_norm": 0.17166944845187698,
+      "learning_rate": 4.615000587583616e-07,
+      "loss": 0.2692,
+      "step": 3168
+    },
+    {
+      "epoch": 1.5952110901071204,
+      "grad_norm": 0.17764393811133042,
+      "learning_rate": 4.6123230354639194e-07,
+      "loss": 0.2578,
+      "step": 3169
+    },
+    {
+      "epoch": 1.5957151858853182,
+      "grad_norm": 0.18326712297594602,
+      "learning_rate": 4.6096455951866277e-07,
+      "loss": 0.2783,
+      "step": 3170
+    },
+    {
+      "epoch": 1.5962192816635161,
+      "grad_norm": 0.16700467021445017,
+      "learning_rate": 4.6069682675241626e-07,
+      "loss": 0.2729,
+      "step": 3171
+    },
+    {
+      "epoch": 1.5967233774417138,
+      "grad_norm": 0.17742240847400476,
+      "learning_rate": 4.6042910532489165e-07,
+      "loss": 0.2702,
+      "step": 3172
+    },
+    {
+      "epoch": 1.5972274732199119,
+      "grad_norm": 0.18272675092406107,
+      "learning_rate": 4.601613953133252e-07,
+      "loss": 0.2776,
+      "step": 3173
+    },
+    {
+      "epoch": 1.5977315689981095,
+      "grad_norm": 0.18849423602774695,
+      "learning_rate": 4.5989369679494935e-07,
+      "loss": 0.2831,
+      "step": 3174
+    },
+    {
+      "epoch": 1.5982356647763076,
+      "grad_norm": 0.19311234812321373,
+      "learning_rate": 4.5962600984699364e-07,
+      "loss": 0.271,
+      "step": 3175
+    },
+    {
+      "epoch": 1.5987397605545053,
+      "grad_norm": 0.17325148338292742,
+      "learning_rate": 4.593583345466837e-07,
+      "loss": 0.269,
+      "step": 3176
+    },
+    {
+      "epoch": 1.5992438563327032,
+      "grad_norm": 0.17501523218289874,
+      "learning_rate": 4.590906709712427e-07,
+      "loss": 0.2683,
+      "step": 3177
+    },
+    {
+      "epoch": 1.599747952110901,
+      "grad_norm": 0.18641330474821174,
+      "learning_rate": 4.588230191978898e-07,
+      "loss": 0.2681,
+      "step": 3178
+    },
+    {
+      "epoch": 1.600252047889099,
+      "grad_norm": 0.17265182176031832,
+      "learning_rate": 4.585553793038405e-07,
+      "loss": 0.2799,
+      "step": 3179
+    },
+    {
+      "epoch": 1.6007561436672968,
+      "grad_norm": 0.17101509452047053,
+      "learning_rate": 4.582877513663077e-07,
+      "loss": 0.2653,
+      "step": 3180
+    },
+    {
+      "epoch": 1.6012602394454947,
+      "grad_norm": 0.172204707085013,
+      "learning_rate": 4.5802013546250014e-07,
+      "loss": 0.2797,
+      "step": 3181
+    },
+    {
+      "epoch": 1.6017643352236925,
+      "grad_norm": 0.17967573980986776,
+      "learning_rate": 4.577525316696236e-07,
+      "loss": 0.262,
+      "step": 3182
+    },
+    {
+      "epoch": 1.6022684310018902,
+      "grad_norm": 0.17469624599365358,
+      "learning_rate": 4.574849400648797e-07,
+      "loss": 0.2636,
+      "step": 3183
+    },
+    {
+      "epoch": 1.6027725267800883,
+      "grad_norm": 0.1743708880339133,
+      "learning_rate": 4.5721736072546754e-07,
+      "loss": 0.2649,
+      "step": 3184
+    },
+    {
+      "epoch": 1.6027725267800883,
+      "eval_loss": 0.307079017162323,
+      "eval_runtime": 18.6674,
+      "eval_samples_per_second": 45.802,
+      "eval_steps_per_second": 0.964,
+      "step": 3184
+    },
+    {
+      "epoch": 1.603276622558286,
+      "grad_norm": 0.18603703776161118,
+      "learning_rate": 4.569497937285817e-07,
+      "loss": 0.2849,
+      "step": 3185
+    },
+    {
+      "epoch": 1.603780718336484,
+      "grad_norm": 0.1746753818130289,
+      "learning_rate": 4.566822391514135e-07,
+      "loss": 0.2644,
+      "step": 3186
+    },
+    {
+      "epoch": 1.6042848141146817,
+      "grad_norm": 0.19466418253252832,
+      "learning_rate": 4.5641469707115123e-07,
+      "loss": 0.2719,
+      "step": 3187
+    },
+    {
+      "epoch": 1.6047889098928798,
+      "grad_norm": 0.19722375083757057,
+      "learning_rate": 4.5614716756497856e-07,
+      "loss": 0.2765,
+      "step": 3188
+    },
+    {
+      "epoch": 1.6052930056710775,
+      "grad_norm": 0.17956583472244503,
+      "learning_rate": 4.5587965071007664e-07,
+      "loss": 0.2806,
+      "step": 3189
+    },
+    {
+      "epoch": 1.6057971014492753,
+      "grad_norm": 0.196228627957389,
+      "learning_rate": 4.556121465836221e-07,
+      "loss": 0.2753,
+      "step": 3190
+    },
+    {
+      "epoch": 1.6063011972274732,
+      "grad_norm": 0.1689632051276557,
+      "learning_rate": 4.553446552627884e-07,
+      "loss": 0.2714,
+      "step": 3191
+    },
+    {
+      "epoch": 1.606805293005671,
+      "grad_norm": 0.17456704042056104,
+      "learning_rate": 4.5507717682474475e-07,
+      "loss": 0.2732,
+      "step": 3192
+    },
+    {
+      "epoch": 1.607309388783869,
+      "grad_norm": 0.18246575730580003,
+      "learning_rate": 4.5480971134665765e-07,
+      "loss": 0.2684,
+      "step": 3193
+    },
+    {
+      "epoch": 1.6078134845620666,
+      "grad_norm": 0.16994370889715202,
+      "learning_rate": 4.545422589056888e-07,
+      "loss": 0.2581,
+      "step": 3194
+    },
+    {
+      "epoch": 1.6083175803402647,
+      "grad_norm": 0.1681073205628179,
+      "learning_rate": 4.5427481957899643e-07,
+      "loss": 0.2746,
+      "step": 3195
+    },
+    {
+      "epoch": 1.6088216761184624,
+      "grad_norm": 0.1935072839716343,
+      "learning_rate": 4.540073934437356e-07,
+      "loss": 0.2728,
+      "step": 3196
+    },
+    {
+      "epoch": 1.6093257718966605,
+      "grad_norm": 0.16743041342533738,
+      "learning_rate": 4.5373998057705667e-07,
+      "loss": 0.2709,
+      "step": 3197
+    },
+    {
+      "epoch": 1.6098298676748581,
+      "grad_norm": 0.18397921239376722,
+      "learning_rate": 4.53472581056107e-07,
+      "loss": 0.2801,
+      "step": 3198
+    },
+    {
+      "epoch": 1.6103339634530562,
+      "grad_norm": 0.17331537650260462,
+      "learning_rate": 4.5320519495802915e-07,
+      "loss": 0.2695,
+      "step": 3199
+    },
+    {
+      "epoch": 1.610838059231254,
+      "grad_norm": 0.17081500880990227,
+      "learning_rate": 4.5293782235996303e-07,
+      "loss": 0.271,
+      "step": 3200
+    },
+    {
+      "epoch": 1.6113421550094518,
+      "grad_norm": 0.17923914190076334,
+      "learning_rate": 4.526704633390435e-07,
+      "loss": 0.265,
+      "step": 3201
+    },
+    {
+      "epoch": 1.6118462507876496,
+      "grad_norm": 0.17772987495851134,
+      "learning_rate": 4.52403117972402e-07,
+      "loss": 0.2783,
+      "step": 3202
+    },
+    {
+      "epoch": 1.6123503465658475,
+      "grad_norm": 0.18238344872837062,
+      "learning_rate": 4.5213578633716627e-07,
+      "loss": 0.2756,
+      "step": 3203
+    },
+    {
+      "epoch": 1.6128544423440454,
+      "grad_norm": 0.17223510416546425,
+      "learning_rate": 4.5186846851045957e-07,
+      "loss": 0.2561,
+      "step": 3204
+    },
+    {
+      "epoch": 1.6133585381222433,
+      "grad_norm": 0.17827644377200683,
+      "learning_rate": 4.516011645694016e-07,
+      "loss": 0.2741,
+      "step": 3205
+    },
+    {
+      "epoch": 1.6138626339004412,
+      "grad_norm": 0.16814074106631602,
+      "learning_rate": 4.513338745911078e-07,
+      "loss": 0.2675,
+      "step": 3206
+    },
+    {
+      "epoch": 1.6143667296786388,
+      "grad_norm": 0.1774956008916992,
+      "learning_rate": 4.5106659865268973e-07,
+      "loss": 0.2697,
+      "step": 3207
+    },
+    {
+      "epoch": 1.614870825456837,
+      "grad_norm": 0.18296109118162968,
+      "learning_rate": 4.507993368312548e-07,
+      "loss": 0.2689,
+      "step": 3208
+    },
+    {
+      "epoch": 1.6153749212350346,
+      "grad_norm": 0.1858855435073671,
+      "learning_rate": 4.5053208920390646e-07,
+      "loss": 0.2669,
+      "step": 3209
+    },
+    {
+      "epoch": 1.6158790170132327,
+      "grad_norm": 0.17175203740808448,
+      "learning_rate": 4.5026485584774397e-07,
+      "loss": 0.2781,
+      "step": 3210
+    },
+    {
+      "epoch": 1.6163831127914303,
+      "grad_norm": 0.1792404257748635,
+      "learning_rate": 4.499976368398623e-07,
+      "loss": 0.2651,
+      "step": 3211
+    },
+    {
+      "epoch": 1.6168872085696282,
+      "grad_norm": 0.17147755827754907,
+      "learning_rate": 4.497304322573529e-07,
+      "loss": 0.2768,
+      "step": 3212
+    },
+    {
+      "epoch": 1.617391304347826,
+      "grad_norm": 0.1689049644593327,
+      "learning_rate": 4.494632421773021e-07,
+      "loss": 0.2644,
+      "step": 3213
+    },
+    {
+      "epoch": 1.617895400126024,
+      "grad_norm": 0.17249930136505048,
+      "learning_rate": 4.4919606667679314e-07,
+      "loss": 0.2695,
+      "step": 3214
+    },
+    {
+      "epoch": 1.6183994959042218,
+      "grad_norm": 0.17187403315174393,
+      "learning_rate": 4.489289058329042e-07,
+      "loss": 0.2753,
+      "step": 3215
+    },
+    {
+      "epoch": 1.6189035916824197,
+      "grad_norm": 0.17003988333355505,
+      "learning_rate": 4.4866175972270934e-07,
+      "loss": 0.2773,
+      "step": 3216
+    },
+    {
+      "epoch": 1.6194076874606176,
+      "grad_norm": 0.174941393420484,
+      "learning_rate": 4.4839462842327905e-07,
+      "loss": 0.2624,
+      "step": 3217
+    },
+    {
+      "epoch": 1.6199117832388152,
+      "grad_norm": 0.16731226660363765,
+      "learning_rate": 4.481275120116785e-07,
+      "loss": 0.2787,
+      "step": 3218
+    },
+    {
+      "epoch": 1.6204158790170133,
+      "grad_norm": 0.18079153758648603,
+      "learning_rate": 4.478604105649697e-07,
+      "loss": 0.2649,
+      "step": 3219
+    },
+    {
+      "epoch": 1.620919974795211,
+      "grad_norm": 0.1717217807780553,
+      "learning_rate": 4.475933241602093e-07,
+      "loss": 0.278,
+      "step": 3220
+    },
+    {
+      "epoch": 1.621424070573409,
+      "grad_norm": 0.1679242542537086,
+      "learning_rate": 4.473262528744502e-07,
+      "loss": 0.2614,
+      "step": 3221
+    },
+    {
+      "epoch": 1.6219281663516067,
+      "grad_norm": 0.17508722076028932,
+      "learning_rate": 4.47059196784741e-07,
+      "loss": 0.2681,
+      "step": 3222
+    },
+    {
+      "epoch": 1.6224322621298046,
+      "grad_norm": 0.17649494235855326,
+      "learning_rate": 4.467921559681255e-07,
+      "loss": 0.2636,
+      "step": 3223
+    },
+    {
+      "epoch": 1.6229363579080025,
+      "grad_norm": 0.171248906662282,
+      "learning_rate": 4.4652513050164344e-07,
+      "loss": 0.2675,
+      "step": 3224
+    },
+    {
+      "epoch": 1.6234404536862004,
+      "grad_norm": 0.1772203147323702,
+      "learning_rate": 4.462581204623298e-07,
+      "loss": 0.2676,
+      "step": 3225
+    },
+    {
+      "epoch": 1.6239445494643983,
+      "grad_norm": 0.17277840226247196,
+      "learning_rate": 4.4599112592721567e-07,
+      "loss": 0.2818,
+      "step": 3226
+    },
+    {
+      "epoch": 1.6244486452425961,
+      "grad_norm": 0.1888456571031224,
+      "learning_rate": 4.4572414697332694e-07,
+      "loss": 0.2582,
+      "step": 3227
+    },
+    {
+      "epoch": 1.624952741020794,
+      "grad_norm": 0.17481572073876417,
+      "learning_rate": 4.454571836776859e-07,
+      "loss": 0.2845,
+      "step": 3228
+    },
+    {
+      "epoch": 1.6254568367989917,
+      "grad_norm": 0.17113867643332575,
+      "learning_rate": 4.4519023611730913e-07,
+      "loss": 0.2628,
+      "step": 3229
+    },
+    {
+      "epoch": 1.6259609325771898,
+      "grad_norm": 0.1700358617278328,
+      "learning_rate": 4.4492330436920997e-07,
+      "loss": 0.272,
+      "step": 3230
+    },
+    {
+      "epoch": 1.6264650283553874,
+      "grad_norm": 0.17282908698853652,
+      "learning_rate": 4.4465638851039636e-07,
+      "loss": 0.2695,
+      "step": 3231
+    },
+    {
+      "epoch": 1.6269691241335855,
+      "grad_norm": 0.17626881997268046,
+      "learning_rate": 4.4438948861787164e-07,
+      "loss": 0.2744,
+      "step": 3232
+    },
+    {
+      "epoch": 1.6274732199117832,
+      "grad_norm": 0.1752662023406442,
+      "learning_rate": 4.4412260476863513e-07,
+      "loss": 0.2699,
+      "step": 3233
+    },
+    {
+      "epoch": 1.627977315689981,
+      "grad_norm": 0.18770480212311322,
+      "learning_rate": 4.4385573703968074e-07,
+      "loss": 0.2798,
+      "step": 3234
+    },
+    {
+      "epoch": 1.628481411468179,
+      "grad_norm": 0.17091794174949165,
+      "learning_rate": 4.435888855079987e-07,
+      "loss": 0.2554,
+      "step": 3235
+    },
+    {
+      "epoch": 1.6289855072463768,
+      "grad_norm": 0.18183277964556385,
+      "learning_rate": 4.4332205025057363e-07,
+      "loss": 0.2645,
+      "step": 3236
+    },
+    {
+      "epoch": 1.6294896030245747,
+      "grad_norm": 0.1721678222436179,
+      "learning_rate": 4.4305523134438603e-07,
+      "loss": 0.2714,
+      "step": 3237
+    },
+    {
+      "epoch": 1.6299936988027726,
+      "grad_norm": 0.17185890801401496,
+      "learning_rate": 4.427884288664114e-07,
+      "loss": 0.274,
+      "step": 3238
+    },
+    {
+      "epoch": 1.6304977945809704,
+      "grad_norm": 0.1705781529081766,
+      "learning_rate": 4.4252164289362055e-07,
+      "loss": 0.2689,
+      "step": 3239
+    },
+    {
+      "epoch": 1.631001890359168,
+      "grad_norm": 0.17045943428176855,
+      "learning_rate": 4.422548735029798e-07,
+      "loss": 0.2612,
+      "step": 3240
+    },
+    {
+      "epoch": 1.6315059861373662,
+      "grad_norm": 0.17057437607989734,
+      "learning_rate": 4.4198812077145014e-07,
+      "loss": 0.275,
+      "step": 3241
+    },
+    {
+      "epoch": 1.6320100819155638,
+      "grad_norm": 0.17813857592829013,
+      "learning_rate": 4.417213847759885e-07,
+      "loss": 0.2851,
+      "step": 3242
+    },
+    {
+      "epoch": 1.632514177693762,
+      "grad_norm": 0.1712522316623488,
+      "learning_rate": 4.4145466559354606e-07,
+      "loss": 0.2792,
+      "step": 3243
+    },
+    {
+      "epoch": 1.6330182734719596,
+      "grad_norm": 0.177089666182129,
+      "learning_rate": 4.4118796330107015e-07,
+      "loss": 0.2812,
+      "step": 3244
+    },
+    {
+      "epoch": 1.6335223692501575,
+      "grad_norm": 0.1804936269724506,
+      "learning_rate": 4.409212779755026e-07,
+      "loss": 0.2923,
+      "step": 3245
+    },
+    {
+      "epoch": 1.6340264650283554,
+      "grad_norm": 0.20349951530764857,
+      "learning_rate": 4.4065460969378e-07,
+      "loss": 0.2755,
+      "step": 3246
+    },
+    {
+      "epoch": 1.6345305608065532,
+      "grad_norm": 0.18026027565564015,
+      "learning_rate": 4.403879585328353e-07,
+      "loss": 0.27,
+      "step": 3247
+    },
+    {
+      "epoch": 1.6350346565847511,
+      "grad_norm": 0.17425845355042768,
+      "learning_rate": 4.4012132456959497e-07,
+      "loss": 0.2744,
+      "step": 3248
+    },
+    {
+      "epoch": 1.635538752362949,
+      "grad_norm": 0.17129421952442173,
+      "learning_rate": 4.3985470788098175e-07,
+      "loss": 0.2645,
+      "step": 3249
+    },
+    {
+      "epoch": 1.6360428481411469,
+      "grad_norm": 0.17611725539203973,
+      "learning_rate": 4.395881085439126e-07,
+      "loss": 0.2703,
+      "step": 3250
+    },
+    {
+      "epoch": 1.6365469439193445,
+      "grad_norm": 0.17504232119969135,
+      "learning_rate": 4.393215266353e-07,
+      "loss": 0.2543,
+      "step": 3251
+    },
+    {
+      "epoch": 1.6370510396975426,
+      "grad_norm": 0.17656439945011204,
+      "learning_rate": 4.3905496223205114e-07,
+      "loss": 0.2704,
+      "step": 3252
+    },
+    {
+      "epoch": 1.6375551354757403,
+      "grad_norm": 0.17063701784531435,
+      "learning_rate": 4.3878841541106805e-07,
+      "loss": 0.2889,
+      "step": 3253
+    },
+    {
+      "epoch": 1.6380592312539384,
+      "grad_norm": 0.16853391156671269,
+      "learning_rate": 4.385218862492479e-07,
+      "loss": 0.2796,
+      "step": 3254
+    },
+    {
+      "epoch": 1.638563327032136,
+      "grad_norm": 0.17227356978627587,
+      "learning_rate": 4.3825537482348274e-07,
+      "loss": 0.2848,
+      "step": 3255
+    },
+    {
+      "epoch": 1.6390674228103341,
+      "grad_norm": 0.17455061108226785,
+      "learning_rate": 4.3798888121065957e-07,
+      "loss": 0.2816,
+      "step": 3256
+    },
+    {
+      "epoch": 1.6395715185885318,
+      "grad_norm": 0.17637327346626572,
+      "learning_rate": 4.377224054876598e-07,
+      "loss": 0.2725,
+      "step": 3257
+    },
+    {
+      "epoch": 1.6400756143667297,
+      "grad_norm": 0.17175945902648324,
+      "learning_rate": 4.374559477313605e-07,
+      "loss": 0.2716,
+      "step": 3258
+    },
+    {
+      "epoch": 1.6405797101449275,
+      "grad_norm": 0.17464832215924994,
+      "learning_rate": 4.3718950801863263e-07,
+      "loss": 0.2816,
+      "step": 3259
+    },
+    {
+      "epoch": 1.6410838059231254,
+      "grad_norm": 0.17327368547696761,
+      "learning_rate": 4.369230864263428e-07,
+      "loss": 0.2708,
+      "step": 3260
+    },
+    {
+      "epoch": 1.6415879017013233,
+      "grad_norm": 0.17593115327303463,
+      "learning_rate": 4.3665668303135184e-07,
+      "loss": 0.271,
+      "step": 3261
+    },
+    {
+      "epoch": 1.642091997479521,
+      "grad_norm": 0.17355291392861003,
+      "learning_rate": 4.363902979105151e-07,
+      "loss": 0.2744,
+      "step": 3262
+    },
+    {
+      "epoch": 1.642596093257719,
+      "grad_norm": 0.17670933469610142,
+      "learning_rate": 4.361239311406837e-07,
+      "loss": 0.2767,
+      "step": 3263
+    },
+    {
+      "epoch": 1.6431001890359167,
+      "grad_norm": 0.1743548246208639,
+      "learning_rate": 4.358575827987022e-07,
+      "loss": 0.2877,
+      "step": 3264
+    },
+    {
+      "epoch": 1.6436042848141148,
+      "grad_norm": 0.1780920002901691,
+      "learning_rate": 4.3559125296141097e-07,
+      "loss": 0.2728,
+      "step": 3265
+    },
+    {
+      "epoch": 1.6441083805923125,
+      "grad_norm": 0.16978778461761998,
+      "learning_rate": 4.3532494170564413e-07,
+      "loss": 0.2638,
+      "step": 3266
+    },
+    {
+      "epoch": 1.6446124763705106,
+      "grad_norm": 0.19486305777071525,
+      "learning_rate": 4.35058649108231e-07,
+      "loss": 0.2745,
+      "step": 3267
+    },
+    {
+      "epoch": 1.6451165721487082,
+      "grad_norm": 0.18084492964621957,
+      "learning_rate": 4.3479237524599544e-07,
+      "loss": 0.2939,
+      "step": 3268
+    },
+    {
+      "epoch": 1.645620667926906,
+      "grad_norm": 0.17132944207036718,
+      "learning_rate": 4.345261201957556e-07,
+      "loss": 0.2664,
+      "step": 3269
+    },
+    {
+      "epoch": 1.646124763705104,
+      "grad_norm": 0.17437387694568648,
+      "learning_rate": 4.342598840343244e-07,
+      "loss": 0.2859,
+      "step": 3270
+    },
+    {
+      "epoch": 1.6466288594833018,
+      "grad_norm": 0.17301815118977484,
+      "learning_rate": 4.3399366683850946e-07,
+      "loss": 0.275,
+      "step": 3271
+    },
+    {
+      "epoch": 1.6471329552614997,
+      "grad_norm": 0.18122055882591992,
+      "learning_rate": 4.3372746868511284e-07,
+      "loss": 0.286,
+      "step": 3272
+    },
+    {
+      "epoch": 1.6476370510396976,
+      "grad_norm": 0.17723397432269952,
+      "learning_rate": 4.3346128965093077e-07,
+      "loss": 0.272,
+      "step": 3273
+    },
+    {
+      "epoch": 1.6481411468178955,
+      "grad_norm": 0.1801415778809958,
+      "learning_rate": 4.331951298127547e-07,
+      "loss": 0.2686,
+      "step": 3274
+    },
+    {
+      "epoch": 1.6486452425960931,
+      "grad_norm": 0.17591886578148874,
+      "learning_rate": 4.3292898924736976e-07,
+      "loss": 0.2549,
+      "step": 3275
+    },
+    {
+      "epoch": 1.6491493383742912,
+      "grad_norm": 0.1759321403355535,
+      "learning_rate": 4.326628680315558e-07,
+      "loss": 0.2836,
+      "step": 3276
+    },
+    {
+      "epoch": 1.6496534341524889,
+      "grad_norm": 0.17426684713872134,
+      "learning_rate": 4.323967662420874e-07,
+      "loss": 0.266,
+      "step": 3277
+    },
+    {
+      "epoch": 1.650157529930687,
+      "grad_norm": 0.1741569992164329,
+      "learning_rate": 4.3213068395573304e-07,
+      "loss": 0.2647,
+      "step": 3278
+    },
+    {
+      "epoch": 1.6506616257088846,
+      "grad_norm": 0.17222858934473642,
+      "learning_rate": 4.3186462124925616e-07,
+      "loss": 0.2735,
+      "step": 3279
+    },
+    {
+      "epoch": 1.6511657214870825,
+      "grad_norm": 0.1823409553601484,
+      "learning_rate": 4.315985781994137e-07,
+      "loss": 0.2756,
+      "step": 3280
+    },
+    {
+      "epoch": 1.6516698172652804,
+      "grad_norm": 0.18057706073850457,
+      "learning_rate": 4.3133255488295793e-07,
+      "loss": 0.2938,
+      "step": 3281
+    },
+    {
+      "epoch": 1.6521739130434783,
+      "grad_norm": 0.16850573295711344,
+      "learning_rate": 4.310665513766347e-07,
+      "loss": 0.258,
+      "step": 3282
+    },
+    {
+      "epoch": 1.6526780088216761,
+      "grad_norm": 0.17055663371284882,
+      "learning_rate": 4.308005677571842e-07,
+      "loss": 0.2773,
+      "step": 3283
+    },
+    {
+      "epoch": 1.653182104599874,
+      "grad_norm": 0.18266654108280814,
+      "learning_rate": 4.305346041013414e-07,
+      "loss": 0.2641,
+      "step": 3284
+    },
+    {
+      "epoch": 1.653686200378072,
+      "grad_norm": 0.18075413873305338,
+      "learning_rate": 4.302686604858349e-07,
+      "loss": 0.2729,
+      "step": 3285
+    },
+    {
+      "epoch": 1.6541902961562696,
+      "grad_norm": 0.16673471863682812,
+      "learning_rate": 4.3000273698738804e-07,
+      "loss": 0.2734,
+      "step": 3286
+    },
+    {
+      "epoch": 1.6546943919344677,
+      "grad_norm": 0.17371867860201198,
+      "learning_rate": 4.2973683368271775e-07,
+      "loss": 0.2755,
+      "step": 3287
+    },
+    {
+      "epoch": 1.6551984877126653,
+      "grad_norm": 0.17118499221022526,
+      "learning_rate": 4.2947095064853586e-07,
+      "loss": 0.2644,
+      "step": 3288
+    },
+    {
+      "epoch": 1.6557025834908634,
+      "grad_norm": 0.17084068049014356,
+      "learning_rate": 4.2920508796154755e-07,
+      "loss": 0.2749,
+      "step": 3289
+    },
+    {
+      "epoch": 1.656206679269061,
+      "grad_norm": 0.17288119445171216,
+      "learning_rate": 4.289392456984531e-07,
+      "loss": 0.2776,
+      "step": 3290
+    },
+    {
+      "epoch": 1.656710775047259,
+      "grad_norm": 0.17586923741312185,
+      "learning_rate": 4.2867342393594596e-07,
+      "loss": 0.2771,
+      "step": 3291
+    },
+    {
+      "epoch": 1.6572148708254568,
+      "grad_norm": 0.17101471939495422,
+      "learning_rate": 4.284076227507141e-07,
+      "loss": 0.263,
+      "step": 3292
+    },
+    {
+      "epoch": 1.6577189666036547,
+      "grad_norm": 0.16831960805801494,
+      "learning_rate": 4.2814184221943964e-07,
+      "loss": 0.2837,
+      "step": 3293
+    },
+    {
+      "epoch": 1.6582230623818526,
+      "grad_norm": 0.17622469939088511,
+      "learning_rate": 4.2787608241879847e-07,
+      "loss": 0.2639,
+      "step": 3294
+    },
+    {
+      "epoch": 1.6587271581600505,
+      "grad_norm": 0.16999028518865347,
+      "learning_rate": 4.2761034342546087e-07,
+      "loss": 0.2775,
+      "step": 3295
+    },
+    {
+      "epoch": 1.6592312539382483,
+      "grad_norm": 0.17456572812362198,
+      "learning_rate": 4.2734462531609063e-07,
+      "loss": 0.2736,
+      "step": 3296
+    },
+    {
+      "epoch": 1.659735349716446,
+      "grad_norm": 0.18460824603728268,
+      "learning_rate": 4.270789281673461e-07,
+      "loss": 0.2742,
+      "step": 3297
+    },
+    {
+      "epoch": 1.660239445494644,
+      "grad_norm": 0.17060321969936063,
+      "learning_rate": 4.26813252055879e-07,
+      "loss": 0.283,
+      "step": 3298
+    },
+    {
+      "epoch": 1.6607435412728417,
+      "grad_norm": 0.17365085031389696,
+      "learning_rate": 4.265475970583353e-07,
+      "loss": 0.2633,
+      "step": 3299
+    },
+    {
+      "epoch": 1.6612476370510398,
+      "grad_norm": 0.17714863665239125,
+      "learning_rate": 4.262819632513548e-07,
+      "loss": 0.268,
+      "step": 3300
+    },
+    {
+      "epoch": 1.6617517328292375,
+      "grad_norm": 0.1801000943195406,
+      "learning_rate": 4.260163507115712e-07,
+      "loss": 0.285,
+      "step": 3301
+    },
+    {
+      "epoch": 1.6622558286074354,
+      "grad_norm": 0.17866572538561554,
+      "learning_rate": 4.257507595156123e-07,
+      "loss": 0.2608,
+      "step": 3302
+    },
+    {
+      "epoch": 1.6627599243856332,
+      "grad_norm": 0.17544127376337787,
+      "learning_rate": 4.2548518974009906e-07,
+      "loss": 0.2845,
+      "step": 3303
+    },
+    {
+      "epoch": 1.6632640201638311,
+      "grad_norm": 0.17044959098671644,
+      "learning_rate": 4.2521964146164726e-07,
+      "loss": 0.267,
+      "step": 3304
+    },
+    {
+      "epoch": 1.663768115942029,
+      "grad_norm": 0.18277292620452268,
+      "learning_rate": 4.249541147568656e-07,
+      "loss": 0.2654,
+      "step": 3305
+    },
+    {
+      "epoch": 1.6642722117202269,
+      "grad_norm": 0.1700015852346814,
+      "learning_rate": 4.2468860970235676e-07,
+      "loss": 0.278,
+      "step": 3306
+    },
+    {
+      "epoch": 1.6647763074984248,
+      "grad_norm": 0.19816383195198753,
+      "learning_rate": 4.244231263747177e-07,
+      "loss": 0.2819,
+      "step": 3307
+    },
+    {
+      "epoch": 1.6652804032766224,
+      "grad_norm": 0.17293706700556136,
+      "learning_rate": 4.241576648505383e-07,
+      "loss": 0.2721,
+      "step": 3308
+    },
+    {
+      "epoch": 1.6657844990548205,
+      "grad_norm": 0.16833769270413493,
+      "learning_rate": 4.2389222520640297e-07,
+      "loss": 0.2629,
+      "step": 3309
+    },
+    {
+      "epoch": 1.6662885948330182,
+      "grad_norm": 0.1796549386579674,
+      "learning_rate": 4.2362680751888894e-07,
+      "loss": 0.281,
+      "step": 3310
+    },
+    {
+      "epoch": 1.6667926906112163,
+      "grad_norm": 0.17213164153650173,
+      "learning_rate": 4.2336141186456815e-07,
+      "loss": 0.278,
+      "step": 3311
+    },
+    {
+      "epoch": 1.667296786389414,
+      "grad_norm": 0.16665647034381148,
+      "learning_rate": 4.2309603832000523e-07,
+      "loss": 0.2677,
+      "step": 3312
+    },
+    {
+      "epoch": 1.6678008821676118,
+      "grad_norm": 0.19006288706354632,
+      "learning_rate": 4.2283068696175867e-07,
+      "loss": 0.2752,
+      "step": 3313
+    },
+    {
+      "epoch": 1.6683049779458097,
+      "grad_norm": 0.17508641308271217,
+      "learning_rate": 4.225653578663811e-07,
+      "loss": 0.2767,
+      "step": 3314
+    },
+    {
+      "epoch": 1.6688090737240076,
+      "grad_norm": 0.17313752219693374,
+      "learning_rate": 4.2230005111041793e-07,
+      "loss": 0.2752,
+      "step": 3315
+    },
+    {
+      "epoch": 1.6693131695022054,
+      "grad_norm": 0.17393250261360155,
+      "learning_rate": 4.2203476677040876e-07,
+      "loss": 0.2722,
+      "step": 3316
+    },
+    {
+      "epoch": 1.6698172652804033,
+      "grad_norm": 0.18422917201705832,
+      "learning_rate": 4.217695049228864e-07,
+      "loss": 0.271,
+      "step": 3317
+    },
+    {
+      "epoch": 1.6703213610586012,
+      "grad_norm": 0.1711040855484809,
+      "learning_rate": 4.2150426564437737e-07,
+      "loss": 0.2628,
+      "step": 3318
+    },
+    {
+      "epoch": 1.6708254568367988,
+      "grad_norm": 0.17294159684320054,
+      "learning_rate": 4.212390490114014e-07,
+      "loss": 0.2892,
+      "step": 3319
+    },
+    {
+      "epoch": 1.671329552614997,
+      "grad_norm": 0.18793079339537197,
+      "learning_rate": 4.2097385510047166e-07,
+      "loss": 0.2637,
+      "step": 3320
+    },
+    {
+      "epoch": 1.6718336483931946,
+      "grad_norm": 0.1773209876917145,
+      "learning_rate": 4.207086839880955e-07,
+      "loss": 0.27,
+      "step": 3321
+    },
+    {
+      "epoch": 1.6723377441713927,
+      "grad_norm": 0.17369452560686208,
+      "learning_rate": 4.204435357507725e-07,
+      "loss": 0.2695,
+      "step": 3322
+    },
+    {
+      "epoch": 1.6728418399495903,
+      "grad_norm": 0.17729418122505228,
+      "learning_rate": 4.201784104649969e-07,
+      "loss": 0.2631,
+      "step": 3323
+    },
+    {
+      "epoch": 1.6733459357277882,
+      "grad_norm": 0.17757333703014838,
+      "learning_rate": 4.199133082072552e-07,
+      "loss": 0.2761,
+      "step": 3324
+    },
+    {
+      "epoch": 1.673850031505986,
+      "grad_norm": 0.17481749922840548,
+      "learning_rate": 4.1964822905402817e-07,
+      "loss": 0.2948,
+      "step": 3325
+    },
+    {
+      "epoch": 1.674354127284184,
+      "grad_norm": 0.17607568877612143,
+      "learning_rate": 4.1938317308178903e-07,
+      "loss": 0.2789,
+      "step": 3326
+    },
+    {
+      "epoch": 1.6748582230623819,
+      "grad_norm": 0.17514273867736174,
+      "learning_rate": 4.191181403670054e-07,
+      "loss": 0.2693,
+      "step": 3327
+    },
+    {
+      "epoch": 1.6753623188405797,
+      "grad_norm": 0.16986181408491555,
+      "learning_rate": 4.1885313098613714e-07,
+      "loss": 0.2719,
+      "step": 3328
+    },
+    {
+      "epoch": 1.6758664146187776,
+      "grad_norm": 0.17536949245885125,
+      "learning_rate": 4.185881450156377e-07,
+      "loss": 0.294,
+      "step": 3329
+    },
+    {
+      "epoch": 1.6763705103969753,
+      "grad_norm": 0.1729391810469066,
+      "learning_rate": 4.183231825319544e-07,
+      "loss": 0.274,
+      "step": 3330
+    },
+    {
+      "epoch": 1.6768746061751734,
+      "grad_norm": 0.181209965516427,
+      "learning_rate": 4.1805824361152677e-07,
+      "loss": 0.2828,
+      "step": 3331
+    },
+    {
+      "epoch": 1.677378701953371,
+      "grad_norm": 0.17437527629077368,
+      "learning_rate": 4.177933283307884e-07,
+      "loss": 0.264,
+      "step": 3332
+    },
+    {
+      "epoch": 1.6778827977315691,
+      "grad_norm": 0.17749950810916038,
+      "learning_rate": 4.1752843676616533e-07,
+      "loss": 0.2774,
+      "step": 3333
+    },
+    {
+      "epoch": 1.6783868935097668,
+      "grad_norm": 0.17035105633216754,
+      "learning_rate": 4.1726356899407765e-07,
+      "loss": 0.2691,
+      "step": 3334
+    },
+    {
+      "epoch": 1.6788909892879649,
+      "grad_norm": 0.18282073284261757,
+      "learning_rate": 4.1699872509093774e-07,
+      "loss": 0.2774,
+      "step": 3335
+    },
+    {
+      "epoch": 1.6793950850661625,
+      "grad_norm": 0.1739086220354733,
+      "learning_rate": 4.167339051331513e-07,
+      "loss": 0.2761,
+      "step": 3336
+    },
+    {
+      "epoch": 1.6798991808443604,
+      "grad_norm": 0.178642966353365,
+      "learning_rate": 4.164691091971176e-07,
+      "loss": 0.2807,
+      "step": 3337
+    },
+    {
+      "epoch": 1.6804032766225583,
+      "grad_norm": 0.17158028839895453,
+      "learning_rate": 4.162043373592282e-07,
+      "loss": 0.2658,
+      "step": 3338
+    },
+    {
+      "epoch": 1.6809073724007562,
+      "grad_norm": 0.1761545331753215,
+      "learning_rate": 4.1593958969586864e-07,
+      "loss": 0.2683,
+      "step": 3339
+    },
+    {
+      "epoch": 1.681411468178954,
+      "grad_norm": 0.1790803594461759,
+      "learning_rate": 4.156748662834165e-07,
+      "loss": 0.2744,
+      "step": 3340
+    },
+    {
+      "epoch": 1.6819155639571517,
+      "grad_norm": 0.17622383520305496,
+      "learning_rate": 4.154101671982433e-07,
+      "loss": 0.2654,
+      "step": 3341
+    },
+    {
+      "epoch": 1.6824196597353498,
+      "grad_norm": 0.1749463588948772,
+      "learning_rate": 4.151454925167129e-07,
+      "loss": 0.292,
+      "step": 3342
+    },
+    {
+      "epoch": 1.6829237555135474,
+      "grad_norm": 0.17771616434870652,
+      "learning_rate": 4.14880842315182e-07,
+      "loss": 0.2689,
+      "step": 3343
+    },
+    {
+      "epoch": 1.6834278512917455,
+      "grad_norm": 0.171167947600746,
+      "learning_rate": 4.14616216670001e-07,
+      "loss": 0.2675,
+      "step": 3344
+    },
+    {
+      "epoch": 1.6839319470699432,
+      "grad_norm": 0.16943013435788834,
+      "learning_rate": 4.143516156575124e-07,
+      "loss": 0.2602,
+      "step": 3345
+    },
+    {
+      "epoch": 1.6844360428481413,
+      "grad_norm": 0.18285712491263306,
+      "learning_rate": 4.1408703935405234e-07,
+      "loss": 0.278,
+      "step": 3346
+    },
+    {
+      "epoch": 1.684940138626339,
+      "grad_norm": 0.17919325702734495,
+      "learning_rate": 4.1382248783594905e-07,
+      "loss": 0.2892,
+      "step": 3347
+    },
+    {
+      "epoch": 1.6854442344045368,
+      "grad_norm": 0.17248307230193116,
+      "learning_rate": 4.135579611795243e-07,
+      "loss": 0.2802,
+      "step": 3348
+    },
+    {
+      "epoch": 1.6859483301827347,
+      "grad_norm": 0.17017032309653435,
+      "learning_rate": 4.132934594610922e-07,
+      "loss": 0.2613,
+      "step": 3349
+    },
+    {
+      "epoch": 1.6864524259609326,
+      "grad_norm": 0.17232227547146478,
+      "learning_rate": 4.130289827569599e-07,
+      "loss": 0.28,
+      "step": 3350
+    },
+    {
+      "epoch": 1.6869565217391305,
+      "grad_norm": 0.19143726314461165,
+      "learning_rate": 4.127645311434275e-07,
+      "loss": 0.2882,
+      "step": 3351
+    },
+    {
+      "epoch": 1.6874606175173283,
+      "grad_norm": 0.17907586158542865,
+      "learning_rate": 4.125001046967871e-07,
+      "loss": 0.2795,
+      "step": 3352
+    },
+    {
+      "epoch": 1.6879647132955262,
+      "grad_norm": 0.1745434199070893,
+      "learning_rate": 4.122357034933248e-07,
+      "loss": 0.2827,
+      "step": 3353
+    },
+    {
+      "epoch": 1.6884688090737239,
+      "grad_norm": 0.18581346560412773,
+      "learning_rate": 4.119713276093181e-07,
+      "loss": 0.2744,
+      "step": 3354
+    },
+    {
+      "epoch": 1.688972904851922,
+      "grad_norm": 0.17050406426787684,
+      "learning_rate": 4.117069771210384e-07,
+      "loss": 0.2638,
+      "step": 3355
+    },
+    {
+      "epoch": 1.6894770006301196,
+      "grad_norm": 0.1678464885882017,
+      "learning_rate": 4.114426521047485e-07,
+      "loss": 0.274,
+      "step": 3356
+    },
+    {
+      "epoch": 1.6899810964083177,
+      "grad_norm": 0.18966378316531626,
+      "learning_rate": 4.1117835263670515e-07,
+      "loss": 0.2813,
+      "step": 3357
+    },
+    {
+      "epoch": 1.6904851921865154,
+      "grad_norm": 0.1700410651953894,
+      "learning_rate": 4.109140787931569e-07,
+      "loss": 0.261,
+      "step": 3358
+    },
+    {
+      "epoch": 1.6909892879647133,
+      "grad_norm": 0.165637909932199,
+      "learning_rate": 4.1064983065034476e-07,
+      "loss": 0.2654,
+      "step": 3359
+    },
+    {
+      "epoch": 1.6914933837429111,
+      "grad_norm": 0.1701023187284413,
+      "learning_rate": 4.103856082845032e-07,
+      "loss": 0.2779,
+      "step": 3360
+    },
+    {
+      "epoch": 1.691997479521109,
+      "grad_norm": 0.16764886438759627,
+      "learning_rate": 4.1012141177185846e-07,
+      "loss": 0.2495,
+      "step": 3361
+    },
+    {
+      "epoch": 1.692501575299307,
+      "grad_norm": 0.17964220409792864,
+      "learning_rate": 4.098572411886296e-07,
+      "loss": 0.2759,
+      "step": 3362
+    },
+    {
+      "epoch": 1.6930056710775048,
+      "grad_norm": 0.17356212055438042,
+      "learning_rate": 4.095930966110283e-07,
+      "loss": 0.2607,
+      "step": 3363
+    },
+    {
+      "epoch": 1.6935097668557026,
+      "grad_norm": 0.1748654301154168,
+      "learning_rate": 4.0932897811525866e-07,
+      "loss": 0.2647,
+      "step": 3364
+    },
+    {
+      "epoch": 1.6940138626339003,
+      "grad_norm": 0.17194284386041553,
+      "learning_rate": 4.0906488577751714e-07,
+      "loss": 0.2757,
+      "step": 3365
+    },
+    {
+      "epoch": 1.6945179584120984,
+      "grad_norm": 0.17475464121096163,
+      "learning_rate": 4.0880081967399265e-07,
+      "loss": 0.2766,
+      "step": 3366
+    },
+    {
+      "epoch": 1.695022054190296,
+      "grad_norm": 0.17506845809429164,
+      "learning_rate": 4.085367798808669e-07,
+      "loss": 0.2813,
+      "step": 3367
+    },
+    {
+      "epoch": 1.6955261499684942,
+      "grad_norm": 0.16997507190026173,
+      "learning_rate": 4.082727664743133e-07,
+      "loss": 0.2748,
+      "step": 3368
+    },
+    {
+      "epoch": 1.6960302457466918,
+      "grad_norm": 0.1823347794889387,
+      "learning_rate": 4.080087795304986e-07,
+      "loss": 0.2795,
+      "step": 3369
+    },
+    {
+      "epoch": 1.6965343415248897,
+      "grad_norm": 0.1714581235316788,
+      "learning_rate": 4.0774481912558103e-07,
+      "loss": 0.267,
+      "step": 3370
+    },
+    {
+      "epoch": 1.6970384373030876,
+      "grad_norm": 0.16918073676897602,
+      "learning_rate": 4.0748088533571174e-07,
+      "loss": 0.2659,
+      "step": 3371
+    },
+    {
+      "epoch": 1.6975425330812854,
+      "grad_norm": 0.1683780550275527,
+      "learning_rate": 4.072169782370339e-07,
+      "loss": 0.2792,
+      "step": 3372
+    },
+    {
+      "epoch": 1.6980466288594833,
+      "grad_norm": 0.1772465529025339,
+      "learning_rate": 4.069530979056829e-07,
+      "loss": 0.2675,
+      "step": 3373
+    },
+    {
+      "epoch": 1.6985507246376812,
+      "grad_norm": 0.20723872319464334,
+      "learning_rate": 4.06689244417787e-07,
+      "loss": 0.2761,
+      "step": 3374
+    },
+    {
+      "epoch": 1.699054820415879,
+      "grad_norm": 0.17653778927841088,
+      "learning_rate": 4.064254178494658e-07,
+      "loss": 0.278,
+      "step": 3375
+    },
+    {
+      "epoch": 1.6995589161940767,
+      "grad_norm": 0.1733117296454345,
+      "learning_rate": 4.0616161827683215e-07,
+      "loss": 0.2633,
+      "step": 3376
+    },
+    {
+      "epoch": 1.7000630119722748,
+      "grad_norm": 0.17322471461748393,
+      "learning_rate": 4.058978457759902e-07,
+      "loss": 0.2777,
+      "step": 3377
+    },
+    {
+      "epoch": 1.7005671077504725,
+      "grad_norm": 0.16784867847711188,
+      "learning_rate": 4.056341004230368e-07,
+      "loss": 0.2646,
+      "step": 3378
+    },
+    {
+      "epoch": 1.7010712035286706,
+      "grad_norm": 0.18157842039507768,
+      "learning_rate": 4.0537038229406093e-07,
+      "loss": 0.286,
+      "step": 3379
+    },
+    {
+      "epoch": 1.7015752993068682,
+      "grad_norm": 0.17284511881224912,
+      "learning_rate": 4.0510669146514356e-07,
+      "loss": 0.2667,
+      "step": 3380
+    },
+    {
+      "epoch": 1.7020793950850661,
+      "grad_norm": 0.1750808365681563,
+      "learning_rate": 4.0484302801235794e-07,
+      "loss": 0.2888,
+      "step": 3381
+    },
+    {
+      "epoch": 1.702583490863264,
+      "grad_norm": 0.17993571684627316,
+      "learning_rate": 4.0457939201176905e-07,
+      "loss": 0.2657,
+      "step": 3382
+    },
+    {
+      "epoch": 1.7030875866414619,
+      "grad_norm": 0.1860139337219952,
+      "learning_rate": 4.0431578353943464e-07,
+      "loss": 0.2712,
+      "step": 3383
+    },
+    {
+      "epoch": 1.7030875866414619,
+      "eval_loss": 0.3066566586494446,
+      "eval_runtime": 17.4494,
+      "eval_samples_per_second": 48.999,
+      "eval_steps_per_second": 1.032,
+      "step": 3383
+    },
+    {
+      "epoch": 1.7035916824196597,
+      "grad_norm": 0.1901404388554431,
+      "learning_rate": 4.0405220267140375e-07,
+      "loss": 0.2657,
+      "step": 3384
+    },
+    {
+      "epoch": 1.7040957781978576,
+      "grad_norm": 0.17602879795743603,
+      "learning_rate": 4.0378864948371824e-07,
+      "loss": 0.2814,
+      "step": 3385
+    },
+    {
+      "epoch": 1.7045998739760555,
+      "grad_norm": 0.17623442887630172,
+      "learning_rate": 4.035251240524114e-07,
+      "loss": 0.2811,
+      "step": 3386
+    },
+    {
+      "epoch": 1.7051039697542532,
+      "grad_norm": 0.18006049009972366,
+      "learning_rate": 4.0326162645350837e-07,
+      "loss": 0.2789,
+      "step": 3387
+    },
+    {
+      "epoch": 1.7056080655324513,
+      "grad_norm": 0.19068107643868426,
+      "learning_rate": 4.029981567630271e-07,
+      "loss": 0.2684,
+      "step": 3388
+    },
+    {
+      "epoch": 1.706112161310649,
+      "grad_norm": 0.17351592149290326,
+      "learning_rate": 4.027347150569765e-07,
+      "loss": 0.2635,
+      "step": 3389
+    },
+    {
+      "epoch": 1.706616257088847,
+      "grad_norm": 0.18358987651652428,
+      "learning_rate": 4.0247130141135825e-07,
+      "loss": 0.2971,
+      "step": 3390
+    },
+    {
+      "epoch": 1.7071203528670447,
+      "grad_norm": 0.1726651344266121,
+      "learning_rate": 4.0220791590216516e-07,
+      "loss": 0.2766,
+      "step": 3391
+    },
+    {
+      "epoch": 1.7076244486452425,
+      "grad_norm": 0.17121736158478892,
+      "learning_rate": 4.0194455860538284e-07,
+      "loss": 0.264,
+      "step": 3392
+    },
+    {
+      "epoch": 1.7081285444234404,
+      "grad_norm": 0.1723867701481724,
+      "learning_rate": 4.016812295969878e-07,
+      "loss": 0.2808,
+      "step": 3393
+    },
+    {
+      "epoch": 1.7086326402016383,
+      "grad_norm": 0.17772561012794783,
+      "learning_rate": 4.01417928952949e-07,
+      "loss": 0.2841,
+      "step": 3394
+    },
+    {
+      "epoch": 1.7091367359798362,
+      "grad_norm": 0.17268289420348543,
+      "learning_rate": 4.0115465674922697e-07,
+      "loss": 0.2715,
+      "step": 3395
+    },
+    {
+      "epoch": 1.709640831758034,
+      "grad_norm": 0.17538041239709554,
+      "learning_rate": 4.008914130617742e-07,
+      "loss": 0.2944,
+      "step": 3396
+    },
+    {
+      "epoch": 1.710144927536232,
+      "grad_norm": 0.1893702660671161,
+      "learning_rate": 4.006281979665349e-07,
+      "loss": 0.2851,
+      "step": 3397
+    },
+    {
+      "epoch": 1.7106490233144296,
+      "grad_norm": 0.17669490425000803,
+      "learning_rate": 4.003650115394446e-07,
+      "loss": 0.2619,
+      "step": 3398
+    },
+    {
+      "epoch": 1.7111531190926277,
+      "grad_norm": 0.17081475118990422,
+      "learning_rate": 4.0010185385643156e-07,
+      "loss": 0.2664,
+      "step": 3399
+    },
+    {
+      "epoch": 1.7116572148708253,
+      "grad_norm": 0.18139383971070436,
+      "learning_rate": 3.9983872499341463e-07,
+      "loss": 0.2722,
+      "step": 3400
+    },
+    {
+      "epoch": 1.7121613106490234,
+      "grad_norm": 0.16819965099097467,
+      "learning_rate": 3.9957562502630524e-07,
+      "loss": 0.2527,
+      "step": 3401
+    },
+    {
+      "epoch": 1.712665406427221,
+      "grad_norm": 0.17426197590337408,
+      "learning_rate": 3.99312554031006e-07,
+      "loss": 0.2675,
+      "step": 3402
+    },
+    {
+      "epoch": 1.7131695022054192,
+      "grad_norm": 0.1781326625588422,
+      "learning_rate": 3.99049512083411e-07,
+      "loss": 0.2769,
+      "step": 3403
+    },
+    {
+      "epoch": 1.7136735979836168,
+      "grad_norm": 0.18770153745691615,
+      "learning_rate": 3.9878649925940653e-07,
+      "loss": 0.266,
+      "step": 3404
+    },
+    {
+      "epoch": 1.7141776937618147,
+      "grad_norm": 0.17464611340669486,
+      "learning_rate": 3.9852351563486987e-07,
+      "loss": 0.2761,
+      "step": 3405
+    },
+    {
+      "epoch": 1.7146817895400126,
+      "grad_norm": 0.171635429988461,
+      "learning_rate": 3.9826056128567053e-07,
+      "loss": 0.2696,
+      "step": 3406
+    },
+    {
+      "epoch": 1.7151858853182105,
+      "grad_norm": 0.18495245117417844,
+      "learning_rate": 3.9799763628766895e-07,
+      "loss": 0.2975,
+      "step": 3407
+    },
+    {
+      "epoch": 1.7156899810964084,
+      "grad_norm": 0.169836479191478,
+      "learning_rate": 3.977347407167174e-07,
+      "loss": 0.2755,
+      "step": 3408
+    },
+    {
+      "epoch": 1.716194076874606,
+      "grad_norm": 0.17255997130578216,
+      "learning_rate": 3.9747187464865984e-07,
+      "loss": 0.2772,
+      "step": 3409
+    },
+    {
+      "epoch": 1.716698172652804,
+      "grad_norm": 0.16749153560856467,
+      "learning_rate": 3.972090381593311e-07,
+      "loss": 0.2736,
+      "step": 3410
+    },
+    {
+      "epoch": 1.7172022684310018,
+      "grad_norm": 0.17079123437345936,
+      "learning_rate": 3.9694623132455815e-07,
+      "loss": 0.2602,
+      "step": 3411
+    },
+    {
+      "epoch": 1.7177063642091999,
+      "grad_norm": 0.17321152114395624,
+      "learning_rate": 3.96683454220159e-07,
+      "loss": 0.275,
+      "step": 3412
+    },
+    {
+      "epoch": 1.7182104599873975,
+      "grad_norm": 0.16962109663638752,
+      "learning_rate": 3.964207069219435e-07,
+      "loss": 0.2732,
+      "step": 3413
+    },
+    {
+      "epoch": 1.7187145557655956,
+      "grad_norm": 0.1746549136635196,
+      "learning_rate": 3.9615798950571216e-07,
+      "loss": 0.2726,
+      "step": 3414
+    },
+    {
+      "epoch": 1.7192186515437933,
+      "grad_norm": 0.16750819214015739,
+      "learning_rate": 3.9589530204725787e-07,
+      "loss": 0.2659,
+      "step": 3415
+    },
+    {
+      "epoch": 1.7197227473219912,
+      "grad_norm": 0.1808474492424595,
+      "learning_rate": 3.956326446223639e-07,
+      "loss": 0.2684,
+      "step": 3416
+    },
+    {
+      "epoch": 1.720226843100189,
+      "grad_norm": 0.17539397294716638,
+      "learning_rate": 3.953700173068053e-07,
+      "loss": 0.2714,
+      "step": 3417
+    },
+    {
+      "epoch": 1.720730938878387,
+      "grad_norm": 0.17517500086130025,
+      "learning_rate": 3.9510742017634877e-07,
+      "loss": 0.2629,
+      "step": 3418
+    },
+    {
+      "epoch": 1.7212350346565848,
+      "grad_norm": 0.17325681094241782,
+      "learning_rate": 3.9484485330675153e-07,
+      "loss": 0.2714,
+      "step": 3419
+    },
+    {
+      "epoch": 1.7217391304347827,
+      "grad_norm": 0.17437360960471293,
+      "learning_rate": 3.9458231677376297e-07,
+      "loss": 0.2784,
+      "step": 3420
+    },
+    {
+      "epoch": 1.7222432262129805,
+      "grad_norm": 0.17254819902144547,
+      "learning_rate": 3.943198106531227e-07,
+      "loss": 0.2596,
+      "step": 3421
+    },
+    {
+      "epoch": 1.7227473219911782,
+      "grad_norm": 0.16965358248820828,
+      "learning_rate": 3.9405733502056267e-07,
+      "loss": 0.2705,
+      "step": 3422
+    },
+    {
+      "epoch": 1.7232514177693763,
+      "grad_norm": 0.17540405751695837,
+      "learning_rate": 3.9379488995180503e-07,
+      "loss": 0.275,
+      "step": 3423
+    },
+    {
+      "epoch": 1.723755513547574,
+      "grad_norm": 0.17044320687255016,
+      "learning_rate": 3.935324755225638e-07,
+      "loss": 0.261,
+      "step": 3424
+    },
+    {
+      "epoch": 1.724259609325772,
+      "grad_norm": 0.17079300009403095,
+      "learning_rate": 3.932700918085439e-07,
+      "loss": 0.2795,
+      "step": 3425
+    },
+    {
+      "epoch": 1.7247637051039697,
+      "grad_norm": 0.17492261380235571,
+      "learning_rate": 3.930077388854413e-07,
+      "loss": 0.2708,
+      "step": 3426
+    },
+    {
+      "epoch": 1.7252678008821676,
+      "grad_norm": 0.1697136489985747,
+      "learning_rate": 3.9274541682894325e-07,
+      "loss": 0.2646,
+      "step": 3427
+    },
+    {
+      "epoch": 1.7257718966603655,
+      "grad_norm": 0.17356605819677773,
+      "learning_rate": 3.9248312571472803e-07,
+      "loss": 0.2725,
+      "step": 3428
+    },
+    {
+      "epoch": 1.7262759924385633,
+      "grad_norm": 0.1798724083697512,
+      "learning_rate": 3.9222086561846515e-07,
+      "loss": 0.2689,
+      "step": 3429
+    },
+    {
+      "epoch": 1.7267800882167612,
+      "grad_norm": 0.172033557393297,
+      "learning_rate": 3.9195863661581466e-07,
+      "loss": 0.2675,
+      "step": 3430
+    },
+    {
+      "epoch": 1.727284183994959,
+      "grad_norm": 0.17813642880671784,
+      "learning_rate": 3.9169643878242836e-07,
+      "loss": 0.2538,
+      "step": 3431
+    },
+    {
+      "epoch": 1.727788279773157,
+      "grad_norm": 0.1744243470379689,
+      "learning_rate": 3.9143427219394857e-07,
+      "loss": 0.2719,
+      "step": 3432
+    },
+    {
+      "epoch": 1.7282923755513546,
+      "grad_norm": 0.19169988100579302,
+      "learning_rate": 3.911721369260086e-07,
+      "loss": 0.2729,
+      "step": 3433
+    },
+    {
+      "epoch": 1.7287964713295527,
+      "grad_norm": 0.1669226246028055,
+      "learning_rate": 3.9091003305423315e-07,
+      "loss": 0.2644,
+      "step": 3434
+    },
+    {
+      "epoch": 1.7293005671077504,
+      "grad_norm": 0.17848442108252346,
+      "learning_rate": 3.9064796065423716e-07,
+      "loss": 0.2765,
+      "step": 3435
+    },
+    {
+      "epoch": 1.7298046628859485,
+      "grad_norm": 0.17288599385442902,
+      "learning_rate": 3.9038591980162733e-07,
+      "loss": 0.2624,
+      "step": 3436
+    },
+    {
+      "epoch": 1.7303087586641461,
+      "grad_norm": 0.17418345316818656,
+      "learning_rate": 3.901239105720004e-07,
+      "loss": 0.2695,
+      "step": 3437
+    },
+    {
+      "epoch": 1.730812854442344,
+      "grad_norm": 0.16887161070123388,
+      "learning_rate": 3.8986193304094483e-07,
+      "loss": 0.269,
+      "step": 3438
+    },
+    {
+      "epoch": 1.7313169502205419,
+      "grad_norm": 0.1715448109776923,
+      "learning_rate": 3.8959998728403937e-07,
+      "loss": 0.2749,
+      "step": 3439
+    },
+    {
+      "epoch": 1.7318210459987398,
+      "grad_norm": 0.17162576830034787,
+      "learning_rate": 3.8933807337685344e-07,
+      "loss": 0.2866,
+      "step": 3440
+    },
+    {
+      "epoch": 1.7323251417769376,
+      "grad_norm": 0.17140705013497554,
+      "learning_rate": 3.89076191394948e-07,
+      "loss": 0.2515,
+      "step": 3441
+    },
+    {
+      "epoch": 1.7328292375551355,
+      "grad_norm": 0.17186684179031347,
+      "learning_rate": 3.8881434141387414e-07,
+      "loss": 0.2633,
+      "step": 3442
+    },
+    {
+      "epoch": 1.7333333333333334,
+      "grad_norm": 0.17247735874366607,
+      "learning_rate": 3.885525235091741e-07,
+      "loss": 0.2699,
+      "step": 3443
+    },
+    {
+      "epoch": 1.733837429111531,
+      "grad_norm": 0.19035869520485968,
+      "learning_rate": 3.8829073775638043e-07,
+      "loss": 0.2869,
+      "step": 3444
+    },
+    {
+      "epoch": 1.7343415248897291,
+      "grad_norm": 0.17126200809840406,
+      "learning_rate": 3.8802898423101715e-07,
+      "loss": 0.2659,
+      "step": 3445
+    },
+    {
+      "epoch": 1.7348456206679268,
+      "grad_norm": 0.17097316798308923,
+      "learning_rate": 3.877672630085983e-07,
+      "loss": 0.2747,
+      "step": 3446
+    },
+    {
+      "epoch": 1.735349716446125,
+      "grad_norm": 0.16663798950741054,
+      "learning_rate": 3.8750557416462863e-07,
+      "loss": 0.2666,
+      "step": 3447
+    },
+    {
+      "epoch": 1.7358538122243226,
+      "grad_norm": 0.17374809497432028,
+      "learning_rate": 3.8724391777460415e-07,
+      "loss": 0.2676,
+      "step": 3448
+    },
+    {
+      "epoch": 1.7363579080025204,
+      "grad_norm": 0.1705608920788751,
+      "learning_rate": 3.869822939140107e-07,
+      "loss": 0.273,
+      "step": 3449
+    },
+    {
+      "epoch": 1.7368620037807183,
+      "grad_norm": 0.1853493481955229,
+      "learning_rate": 3.867207026583255e-07,
+      "loss": 0.2858,
+      "step": 3450
+    },
+    {
+      "epoch": 1.7373660995589162,
+      "grad_norm": 0.1768536035338538,
+      "learning_rate": 3.8645914408301574e-07,
+      "loss": 0.2832,
+      "step": 3451
+    },
+    {
+      "epoch": 1.737870195337114,
+      "grad_norm": 0.16841766882214215,
+      "learning_rate": 3.861976182635397e-07,
+      "loss": 0.2623,
+      "step": 3452
+    },
+    {
+      "epoch": 1.738374291115312,
+      "grad_norm": 0.21467318021749093,
+      "learning_rate": 3.8593612527534585e-07,
+      "loss": 0.2769,
+      "step": 3453
+    },
+    {
+      "epoch": 1.7388783868935098,
+      "grad_norm": 0.17835594443596708,
+      "learning_rate": 3.8567466519387305e-07,
+      "loss": 0.2818,
+      "step": 3454
+    },
+    {
+      "epoch": 1.7393824826717075,
+      "grad_norm": 0.17944917802544677,
+      "learning_rate": 3.854132380945513e-07,
+      "loss": 0.285,
+      "step": 3455
+    },
+    {
+      "epoch": 1.7398865784499056,
+      "grad_norm": 0.1756271448121031,
+      "learning_rate": 3.8515184405280054e-07,
+      "loss": 0.2837,
+      "step": 3456
+    },
+    {
+      "epoch": 1.7403906742281032,
+      "grad_norm": 0.16873921221680957,
+      "learning_rate": 3.8489048314403124e-07,
+      "loss": 0.2668,
+      "step": 3457
+    },
+    {
+      "epoch": 1.7408947700063013,
+      "grad_norm": 0.1837468940499396,
+      "learning_rate": 3.846291554436445e-07,
+      "loss": 0.2771,
+      "step": 3458
+    },
+    {
+      "epoch": 1.741398865784499,
+      "grad_norm": 0.17119908182449262,
+      "learning_rate": 3.8436786102703186e-07,
+      "loss": 0.2642,
+      "step": 3459
+    },
+    {
+      "epoch": 1.7419029615626969,
+      "grad_norm": 0.17158536564497223,
+      "learning_rate": 3.841065999695748e-07,
+      "loss": 0.2716,
+      "step": 3460
+    },
+    {
+      "epoch": 1.7424070573408947,
+      "grad_norm": 0.18136191876998276,
+      "learning_rate": 3.838453723466459e-07,
+      "loss": 0.2701,
+      "step": 3461
+    },
+    {
+      "epoch": 1.7429111531190926,
+      "grad_norm": 0.17402251262948262,
+      "learning_rate": 3.8358417823360757e-07,
+      "loss": 0.2705,
+      "step": 3462
+    },
+    {
+      "epoch": 1.7434152488972905,
+      "grad_norm": 0.16714838859967632,
+      "learning_rate": 3.8332301770581244e-07,
+      "loss": 0.2705,
+      "step": 3463
+    },
+    {
+      "epoch": 1.7439193446754884,
+      "grad_norm": 0.16798057047478643,
+      "learning_rate": 3.8306189083860414e-07,
+      "loss": 0.265,
+      "step": 3464
+    },
+    {
+      "epoch": 1.7444234404536862,
+      "grad_norm": 0.1675965482058386,
+      "learning_rate": 3.8280079770731565e-07,
+      "loss": 0.2642,
+      "step": 3465
+    },
+    {
+      "epoch": 1.744927536231884,
+      "grad_norm": 0.1722718830074659,
+      "learning_rate": 3.8253973838727134e-07,
+      "loss": 0.2783,
+      "step": 3466
+    },
+    {
+      "epoch": 1.745431632010082,
+      "grad_norm": 0.17511740118469324,
+      "learning_rate": 3.822787129537846e-07,
+      "loss": 0.2548,
+      "step": 3467
+    },
+    {
+      "epoch": 1.7459357277882797,
+      "grad_norm": 0.1692592884492893,
+      "learning_rate": 3.820177214821602e-07,
+      "loss": 0.2756,
+      "step": 3468
+    },
+    {
+      "epoch": 1.7464398235664778,
+      "grad_norm": 0.1838015667126571,
+      "learning_rate": 3.8175676404769226e-07,
+      "loss": 0.2835,
+      "step": 3469
+    },
+    {
+      "epoch": 1.7469439193446754,
+      "grad_norm": 0.17221059457481896,
+      "learning_rate": 3.8149584072566516e-07,
+      "loss": 0.2679,
+      "step": 3470
+    },
+    {
+      "epoch": 1.7474480151228735,
+      "grad_norm": 0.17240520098361659,
+      "learning_rate": 3.8123495159135414e-07,
+      "loss": 0.2602,
+      "step": 3471
+    },
+    {
+      "epoch": 1.7479521109010712,
+      "grad_norm": 0.17645638344602915,
+      "learning_rate": 3.8097409672002384e-07,
+      "loss": 0.2664,
+      "step": 3472
+    },
+    {
+      "epoch": 1.748456206679269,
+      "grad_norm": 0.17119884887181155,
+      "learning_rate": 3.8071327618692926e-07,
+      "loss": 0.2771,
+      "step": 3473
+    },
+    {
+      "epoch": 1.748960302457467,
+      "grad_norm": 0.16737717354943357,
+      "learning_rate": 3.8045249006731554e-07,
+      "loss": 0.257,
+      "step": 3474
+    },
+    {
+      "epoch": 1.7494643982356648,
+      "grad_norm": 0.17011856009678725,
+      "learning_rate": 3.80191738436418e-07,
+      "loss": 0.2622,
+      "step": 3475
+    },
+    {
+      "epoch": 1.7499684940138627,
+      "grad_norm": 0.171796792732586,
+      "learning_rate": 3.799310213694618e-07,
+      "loss": 0.2818,
+      "step": 3476
+    },
+    {
+      "epoch": 1.7504725897920603,
+      "grad_norm": 0.1928931218051459,
+      "learning_rate": 3.796703389416619e-07,
+      "loss": 0.2763,
+      "step": 3477
+    },
+    {
+      "epoch": 1.7509766855702584,
+      "grad_norm": 0.20147336279423894,
+      "learning_rate": 3.7940969122822395e-07,
+      "loss": 0.2824,
+      "step": 3478
+    },
+    {
+      "epoch": 1.751480781348456,
+      "grad_norm": 0.175565231620901,
+      "learning_rate": 3.7914907830434296e-07,
+      "loss": 0.2864,
+      "step": 3479
+    },
+    {
+      "epoch": 1.7519848771266542,
+      "grad_norm": 0.17087917079117643,
+      "learning_rate": 3.7888850024520443e-07,
+      "loss": 0.2664,
+      "step": 3480
+    },
+    {
+      "epoch": 1.7524889729048518,
+      "grad_norm": 0.17664931365772008,
+      "learning_rate": 3.7862795712598307e-07,
+      "loss": 0.2758,
+      "step": 3481
+    },
+    {
+      "epoch": 1.75299306868305,
+      "grad_norm": 0.17456510661593722,
+      "learning_rate": 3.783674490218445e-07,
+      "loss": 0.2734,
+      "step": 3482
+    },
+    {
+      "epoch": 1.7534971644612476,
+      "grad_norm": 0.18609071641176797,
+      "learning_rate": 3.7810697600794335e-07,
+      "loss": 0.267,
+      "step": 3483
+    },
+    {
+      "epoch": 1.7540012602394455,
+      "grad_norm": 0.17283492786683485,
+      "learning_rate": 3.7784653815942436e-07,
+      "loss": 0.2837,
+      "step": 3484
+    },
+    {
+      "epoch": 1.7545053560176433,
+      "grad_norm": 0.17490689341326401,
+      "learning_rate": 3.775861355514226e-07,
+      "loss": 0.2799,
+      "step": 3485
+    },
+    {
+      "epoch": 1.7550094517958412,
+      "grad_norm": 0.17672370129936685,
+      "learning_rate": 3.7732576825906223e-07,
+      "loss": 0.2539,
+      "step": 3486
+    },
+    {
+      "epoch": 1.755513547574039,
+      "grad_norm": 0.17912383987865965,
+      "learning_rate": 3.7706543635745795e-07,
+      "loss": 0.2949,
+      "step": 3487
+    },
+    {
+      "epoch": 1.756017643352237,
+      "grad_norm": 0.1744911022180228,
+      "learning_rate": 3.768051399217137e-07,
+      "loss": 0.2803,
+      "step": 3488
+    },
+    {
+      "epoch": 1.7565217391304349,
+      "grad_norm": 0.17364431498626157,
+      "learning_rate": 3.7654487902692355e-07,
+      "loss": 0.2761,
+      "step": 3489
+    },
+    {
+      "epoch": 1.7570258349086325,
+      "grad_norm": 0.1746170171542789,
+      "learning_rate": 3.762846537481708e-07,
+      "loss": 0.2696,
+      "step": 3490
+    },
+    {
+      "epoch": 1.7575299306868306,
+      "grad_norm": 0.17817739668123456,
+      "learning_rate": 3.7602446416052926e-07,
+      "loss": 0.2801,
+      "step": 3491
+    },
+    {
+      "epoch": 1.7580340264650283,
+      "grad_norm": 0.17133707420290878,
+      "learning_rate": 3.757643103390618e-07,
+      "loss": 0.278,
+      "step": 3492
+    },
+    {
+      "epoch": 1.7585381222432264,
+      "grad_norm": 0.17227766863166188,
+      "learning_rate": 3.75504192358821e-07,
+      "loss": 0.2655,
+      "step": 3493
+    },
+    {
+      "epoch": 1.759042218021424,
+      "grad_norm": 0.1841860670831539,
+      "learning_rate": 3.7524411029484984e-07,
+      "loss": 0.284,
+      "step": 3494
+    },
+    {
+      "epoch": 1.759546313799622,
+      "grad_norm": 0.18128560254995008,
+      "learning_rate": 3.7498406422217967e-07,
+      "loss": 0.2688,
+      "step": 3495
+    },
+    {
+      "epoch": 1.7600504095778198,
+      "grad_norm": 0.1777006939689815,
+      "learning_rate": 3.7472405421583283e-07,
+      "loss": 0.2665,
+      "step": 3496
+    },
+    {
+      "epoch": 1.7605545053560177,
+      "grad_norm": 0.17651881222532234,
+      "learning_rate": 3.7446408035082e-07,
+      "loss": 0.2659,
+      "step": 3497
+    },
+    {
+      "epoch": 1.7610586011342155,
+      "grad_norm": 0.17121563039114812,
+      "learning_rate": 3.742041427021426e-07,
+      "loss": 0.2661,
+      "step": 3498
+    },
+    {
+      "epoch": 1.7615626969124134,
+      "grad_norm": 0.17086616322817438,
+      "learning_rate": 3.7394424134479085e-07,
+      "loss": 0.2748,
+      "step": 3499
+    },
+    {
+      "epoch": 1.7620667926906113,
+      "grad_norm": 0.1845306405915281,
+      "learning_rate": 3.736843763537443e-07,
+      "loss": 0.2666,
+      "step": 3500
+    },
+    {
+      "epoch": 1.762570888468809,
+      "grad_norm": 0.18385553877215485,
+      "learning_rate": 3.734245478039729e-07,
+      "loss": 0.2876,
+      "step": 3501
+    },
+    {
+      "epoch": 1.763074984247007,
+      "grad_norm": 0.1766369636170071,
+      "learning_rate": 3.7316475577043515e-07,
+      "loss": 0.2662,
+      "step": 3502
+    },
+    {
+      "epoch": 1.7635790800252047,
+      "grad_norm": 0.17153706729666343,
+      "learning_rate": 3.7290500032807994e-07,
+      "loss": 0.2629,
+      "step": 3503
+    },
+    {
+      "epoch": 1.7640831758034028,
+      "grad_norm": 0.17299192192087665,
+      "learning_rate": 3.726452815518446e-07,
+      "loss": 0.2619,
+      "step": 3504
+    },
+    {
+      "epoch": 1.7645872715816004,
+      "grad_norm": 0.19384856527161845,
+      "learning_rate": 3.7238559951665684e-07,
+      "loss": 0.264,
+      "step": 3505
+    },
+    {
+      "epoch": 1.7650913673597983,
+      "grad_norm": 0.17125569407787852,
+      "learning_rate": 3.721259542974329e-07,
+      "loss": 0.2736,
+      "step": 3506
+    },
+    {
+      "epoch": 1.7655954631379962,
+      "grad_norm": 0.17240058028350044,
+      "learning_rate": 3.7186634596907905e-07,
+      "loss": 0.2747,
+      "step": 3507
+    },
+    {
+      "epoch": 1.766099558916194,
+      "grad_norm": 0.1788859679300591,
+      "learning_rate": 3.716067746064907e-07,
+      "loss": 0.275,
+      "step": 3508
+    },
+    {
+      "epoch": 1.766603654694392,
+      "grad_norm": 0.17068346062555592,
+      "learning_rate": 3.7134724028455235e-07,
+      "loss": 0.2596,
+      "step": 3509
+    },
+    {
+      "epoch": 1.7671077504725898,
+      "grad_norm": 0.17273509001374454,
+      "learning_rate": 3.710877430781384e-07,
+      "loss": 0.2734,
+      "step": 3510
+    },
+    {
+      "epoch": 1.7676118462507877,
+      "grad_norm": 0.19441499623673344,
+      "learning_rate": 3.708282830621118e-07,
+      "loss": 0.2865,
+      "step": 3511
+    },
+    {
+      "epoch": 1.7681159420289854,
+      "grad_norm": 0.18157876524943378,
+      "learning_rate": 3.705688603113256e-07,
+      "loss": 0.2659,
+      "step": 3512
+    },
+    {
+      "epoch": 1.7686200378071835,
+      "grad_norm": 0.1755434232731158,
+      "learning_rate": 3.703094749006214e-07,
+      "loss": 0.2657,
+      "step": 3513
+    },
+    {
+      "epoch": 1.7691241335853811,
+      "grad_norm": 0.17616250911630144,
+      "learning_rate": 3.700501269048301e-07,
+      "loss": 0.2696,
+      "step": 3514
+    },
+    {
+      "epoch": 1.7696282293635792,
+      "grad_norm": 0.1717712747533756,
+      "learning_rate": 3.697908163987724e-07,
+      "loss": 0.263,
+      "step": 3515
+    },
+    {
+      "epoch": 1.7701323251417769,
+      "grad_norm": 0.1746899392847966,
+      "learning_rate": 3.6953154345725733e-07,
+      "loss": 0.2833,
+      "step": 3516
+    },
+    {
+      "epoch": 1.7706364209199748,
+      "grad_norm": 0.1705652748493287,
+      "learning_rate": 3.6927230815508403e-07,
+      "loss": 0.2637,
+      "step": 3517
+    },
+    {
+      "epoch": 1.7711405166981726,
+      "grad_norm": 0.18596072083429435,
+      "learning_rate": 3.6901311056703985e-07,
+      "loss": 0.2953,
+      "step": 3518
+    },
+    {
+      "epoch": 1.7716446124763705,
+      "grad_norm": 0.17652711756186662,
+      "learning_rate": 3.6875395076790186e-07,
+      "loss": 0.2861,
+      "step": 3519
+    },
+    {
+      "epoch": 1.7721487082545684,
+      "grad_norm": 0.17706846127443582,
+      "learning_rate": 3.684948288324362e-07,
+      "loss": 0.2646,
+      "step": 3520
+    },
+    {
+      "epoch": 1.7726528040327663,
+      "grad_norm": 0.17967101443910952,
+      "learning_rate": 3.6823574483539767e-07,
+      "loss": 0.2862,
+      "step": 3521
+    },
+    {
+      "epoch": 1.7731568998109641,
+      "grad_norm": 0.1747251079117009,
+      "learning_rate": 3.6797669885153046e-07,
+      "loss": 0.2683,
+      "step": 3522
+    },
+    {
+      "epoch": 1.7736609955891618,
+      "grad_norm": 0.1687110695892719,
+      "learning_rate": 3.6771769095556783e-07,
+      "loss": 0.2653,
+      "step": 3523
+    },
+    {
+      "epoch": 1.77416509136736,
+      "grad_norm": 0.17110856789881085,
+      "learning_rate": 3.6745872122223196e-07,
+      "loss": 0.2758,
+      "step": 3524
+    },
+    {
+      "epoch": 1.7746691871455575,
+      "grad_norm": 0.17460922566080275,
+      "learning_rate": 3.6719978972623385e-07,
+      "loss": 0.2753,
+      "step": 3525
+    },
+    {
+      "epoch": 1.7751732829237556,
+      "grad_norm": 0.1847111261299794,
+      "learning_rate": 3.669408965422739e-07,
+      "loss": 0.2731,
+      "step": 3526
+    },
+    {
+      "epoch": 1.7756773787019533,
+      "grad_norm": 0.17449754787956734,
+      "learning_rate": 3.6668204174504085e-07,
+      "loss": 0.2726,
+      "step": 3527
+    },
+    {
+      "epoch": 1.7761814744801512,
+      "grad_norm": 0.17338254071796924,
+      "learning_rate": 3.664232254092131e-07,
+      "loss": 0.2793,
+      "step": 3528
+    },
+    {
+      "epoch": 1.776685570258349,
+      "grad_norm": 0.17628576320824763,
+      "learning_rate": 3.6616444760945744e-07,
+      "loss": 0.2793,
+      "step": 3529
+    },
+    {
+      "epoch": 1.777189666036547,
+      "grad_norm": 0.1761083816514799,
+      "learning_rate": 3.659057084204292e-07,
+      "loss": 0.2762,
+      "step": 3530
+    },
+    {
+      "epoch": 1.7776937618147448,
+      "grad_norm": 0.17618815238730148,
+      "learning_rate": 3.656470079167737e-07,
+      "loss": 0.2723,
+      "step": 3531
+    },
+    {
+      "epoch": 1.7781978575929427,
+      "grad_norm": 0.2082525337835813,
+      "learning_rate": 3.6538834617312395e-07,
+      "loss": 0.2656,
+      "step": 3532
+    },
+    {
+      "epoch": 1.7787019533711406,
+      "grad_norm": 0.1740382361506214,
+      "learning_rate": 3.6512972326410274e-07,
+      "loss": 0.2874,
+      "step": 3533
+    },
+    {
+      "epoch": 1.7792060491493382,
+      "grad_norm": 0.18497511420153842,
+      "learning_rate": 3.6487113926432067e-07,
+      "loss": 0.2831,
+      "step": 3534
+    },
+    {
+      "epoch": 1.7797101449275363,
+      "grad_norm": 0.1868818820255656,
+      "learning_rate": 3.6461259424837787e-07,
+      "loss": 0.2704,
+      "step": 3535
+    },
+    {
+      "epoch": 1.780214240705734,
+      "grad_norm": 0.1747555047792214,
+      "learning_rate": 3.643540882908631e-07,
+      "loss": 0.2742,
+      "step": 3536
+    },
+    {
+      "epoch": 1.780718336483932,
+      "grad_norm": 0.16955294868925422,
+      "learning_rate": 3.640956214663534e-07,
+      "loss": 0.2805,
+      "step": 3537
+    },
+    {
+      "epoch": 1.7812224322621297,
+      "grad_norm": 0.17195016907202454,
+      "learning_rate": 3.638371938494151e-07,
+      "loss": 0.2726,
+      "step": 3538
+    },
+    {
+      "epoch": 1.7817265280403278,
+      "grad_norm": 0.17077392343569042,
+      "learning_rate": 3.635788055146028e-07,
+      "loss": 0.2695,
+      "step": 3539
+    },
+    {
+      "epoch": 1.7822306238185255,
+      "grad_norm": 0.16922021837013979,
+      "learning_rate": 3.633204565364602e-07,
+      "loss": 0.2607,
+      "step": 3540
+    },
+    {
+      "epoch": 1.7827347195967234,
+      "grad_norm": 0.17020929306602342,
+      "learning_rate": 3.63062146989519e-07,
+      "loss": 0.2774,
+      "step": 3541
+    },
+    {
+      "epoch": 1.7832388153749212,
+      "grad_norm": 0.18083426533544908,
+      "learning_rate": 3.628038769483002e-07,
+      "loss": 0.2841,
+      "step": 3542
+    },
+    {
+      "epoch": 1.7837429111531191,
+      "grad_norm": 0.1790034539370043,
+      "learning_rate": 3.625456464873131e-07,
+      "loss": 0.2713,
+      "step": 3543
+    },
+    {
+      "epoch": 1.784247006931317,
+      "grad_norm": 0.1729253600603901,
+      "learning_rate": 3.6228745568105534e-07,
+      "loss": 0.2714,
+      "step": 3544
+    },
+    {
+      "epoch": 1.7847511027095146,
+      "grad_norm": 0.20077594866662438,
+      "learning_rate": 3.620293046040137e-07,
+      "loss": 0.2694,
+      "step": 3545
+    },
+    {
+      "epoch": 1.7852551984877127,
+      "grad_norm": 0.17484350438222895,
+      "learning_rate": 3.6177119333066274e-07,
+      "loss": 0.2577,
+      "step": 3546
+    },
+    {
+      "epoch": 1.7857592942659104,
+      "grad_norm": 0.17015014074186688,
+      "learning_rate": 3.6151312193546653e-07,
+      "loss": 0.2756,
+      "step": 3547
+    },
+    {
+      "epoch": 1.7862633900441085,
+      "grad_norm": 0.18310485144032457,
+      "learning_rate": 3.612550904928765e-07,
+      "loss": 0.2605,
+      "step": 3548
+    },
+    {
+      "epoch": 1.7867674858223062,
+      "grad_norm": 0.1697602759063394,
+      "learning_rate": 3.6099709907733355e-07,
+      "loss": 0.277,
+      "step": 3549
+    },
+    {
+      "epoch": 1.7872715816005043,
+      "grad_norm": 0.17942080037574837,
+      "learning_rate": 3.6073914776326654e-07,
+      "loss": 0.2766,
+      "step": 3550
+    },
+    {
+      "epoch": 1.787775677378702,
+      "grad_norm": 0.1749196729191675,
+      "learning_rate": 3.6048123662509257e-07,
+      "loss": 0.2702,
+      "step": 3551
+    },
+    {
+      "epoch": 1.7882797731568998,
+      "grad_norm": 0.17142762886817356,
+      "learning_rate": 3.6022336573721757e-07,
+      "loss": 0.2646,
+      "step": 3552
+    },
+    {
+      "epoch": 1.7887838689350977,
+      "grad_norm": 0.16935198240355007,
+      "learning_rate": 3.599655351740356e-07,
+      "loss": 0.2799,
+      "step": 3553
+    },
+    {
+      "epoch": 1.7892879647132955,
+      "grad_norm": 0.17626517942064382,
+      "learning_rate": 3.597077450099295e-07,
+      "loss": 0.2608,
+      "step": 3554
+    },
+    {
+      "epoch": 1.7897920604914934,
+      "grad_norm": 0.1756679530760868,
+      "learning_rate": 3.5944999531926963e-07,
+      "loss": 0.2762,
+      "step": 3555
+    },
+    {
+      "epoch": 1.7902961562696913,
+      "grad_norm": 0.17715691284816204,
+      "learning_rate": 3.5919228617641564e-07,
+      "loss": 0.2753,
+      "step": 3556
+    },
+    {
+      "epoch": 1.7908002520478892,
+      "grad_norm": 0.1778053393171107,
+      "learning_rate": 3.5893461765571464e-07,
+      "loss": 0.2705,
+      "step": 3557
+    },
+    {
+      "epoch": 1.7913043478260868,
+      "grad_norm": 0.17959544599325605,
+      "learning_rate": 3.5867698983150277e-07,
+      "loss": 0.2629,
+      "step": 3558
+    },
+    {
+      "epoch": 1.791808443604285,
+      "grad_norm": 0.1662083229493581,
+      "learning_rate": 3.5841940277810395e-07,
+      "loss": 0.2628,
+      "step": 3559
+    },
+    {
+      "epoch": 1.7923125393824826,
+      "grad_norm": 0.16874268935126943,
+      "learning_rate": 3.581618565698301e-07,
+      "loss": 0.27,
+      "step": 3560
+    },
+    {
+      "epoch": 1.7928166351606807,
+      "grad_norm": 0.17322137849836952,
+      "learning_rate": 3.5790435128098227e-07,
+      "loss": 0.2669,
+      "step": 3561
+    },
+    {
+      "epoch": 1.7933207309388783,
+      "grad_norm": 0.1742075041952224,
+      "learning_rate": 3.576468869858486e-07,
+      "loss": 0.2726,
+      "step": 3562
+    },
+    {
+      "epoch": 1.7938248267170762,
+      "grad_norm": 0.168819659068297,
+      "learning_rate": 3.573894637587064e-07,
+      "loss": 0.2707,
+      "step": 3563
+    },
+    {
+      "epoch": 1.794328922495274,
+      "grad_norm": 0.19374260839838295,
+      "learning_rate": 3.571320816738205e-07,
+      "loss": 0.2779,
+      "step": 3564
+    },
+    {
+      "epoch": 1.794833018273472,
+      "grad_norm": 0.16947665412300375,
+      "learning_rate": 3.568747408054439e-07,
+      "loss": 0.2748,
+      "step": 3565
+    },
+    {
+      "epoch": 1.7953371140516698,
+      "grad_norm": 0.17469189198050658,
+      "learning_rate": 3.5661744122781804e-07,
+      "loss": 0.2618,
+      "step": 3566
+    },
+    {
+      "epoch": 1.7958412098298677,
+      "grad_norm": 0.1702481414901748,
+      "learning_rate": 3.5636018301517213e-07,
+      "loss": 0.2597,
+      "step": 3567
+    },
+    {
+      "epoch": 1.7963453056080656,
+      "grad_norm": 0.17222409493077534,
+      "learning_rate": 3.5610296624172365e-07,
+      "loss": 0.2649,
+      "step": 3568
+    },
+    {
+      "epoch": 1.7968494013862633,
+      "grad_norm": 0.16931592244630336,
+      "learning_rate": 3.55845790981678e-07,
+      "loss": 0.2731,
+      "step": 3569
+    },
+    {
+      "epoch": 1.7973534971644614,
+      "grad_norm": 0.17620807131039093,
+      "learning_rate": 3.555886573092287e-07,
+      "loss": 0.2749,
+      "step": 3570
+    },
+    {
+      "epoch": 1.797857592942659,
+      "grad_norm": 0.18052515979122832,
+      "learning_rate": 3.55331565298557e-07,
+      "loss": 0.2878,
+      "step": 3571
+    },
+    {
+      "epoch": 1.798361688720857,
+      "grad_norm": 0.17191185207152138,
+      "learning_rate": 3.550745150238328e-07,
+      "loss": 0.2862,
+      "step": 3572
+    },
+    {
+      "epoch": 1.7988657844990548,
+      "grad_norm": 0.1757024787598414,
+      "learning_rate": 3.548175065592132e-07,
+      "loss": 0.2649,
+      "step": 3573
+    },
+    {
+      "epoch": 1.7993698802772526,
+      "grad_norm": 0.1729303037797562,
+      "learning_rate": 3.545605399788434e-07,
+      "loss": 0.2562,
+      "step": 3574
+    },
+    {
+      "epoch": 1.7998739760554505,
+      "grad_norm": 0.17438697638244505,
+      "learning_rate": 3.543036153568571e-07,
+      "loss": 0.2601,
+      "step": 3575
+    },
+    {
+      "epoch": 1.8003780718336484,
+      "grad_norm": 0.1785781990776129,
+      "learning_rate": 3.5404673276737503e-07,
+      "loss": 0.2732,
+      "step": 3576
+    },
+    {
+      "epoch": 1.8008821676118463,
+      "grad_norm": 0.17317098668224498,
+      "learning_rate": 3.5378989228450664e-07,
+      "loss": 0.2797,
+      "step": 3577
+    },
+    {
+      "epoch": 1.8013862633900442,
+      "grad_norm": 0.1690137234088919,
+      "learning_rate": 3.5353309398234833e-07,
+      "loss": 0.2721,
+      "step": 3578
+    },
+    {
+      "epoch": 1.801890359168242,
+      "grad_norm": 0.17537848542244042,
+      "learning_rate": 3.5327633793498535e-07,
+      "loss": 0.2714,
+      "step": 3579
+    },
+    {
+      "epoch": 1.8023944549464397,
+      "grad_norm": 0.16986985399350213,
+      "learning_rate": 3.530196242164899e-07,
+      "loss": 0.2637,
+      "step": 3580
+    },
+    {
+      "epoch": 1.8028985507246378,
+      "grad_norm": 0.16970500175613376,
+      "learning_rate": 3.5276295290092225e-07,
+      "loss": 0.2919,
+      "step": 3581
+    },
+    {
+      "epoch": 1.8034026465028354,
+      "grad_norm": 0.1798669561885536,
+      "learning_rate": 3.525063240623307e-07,
+      "loss": 0.2743,
+      "step": 3582
+    },
+    {
+      "epoch": 1.8034026465028354,
+      "eval_loss": 0.3062366843223572,
+      "eval_runtime": 17.926,
+      "eval_samples_per_second": 47.696,
+      "eval_steps_per_second": 1.004,
+      "step": 3582
+    },
+    {
+      "epoch": 1.8039067422810335,
+      "grad_norm": 0.17610026348303112,
+      "learning_rate": 3.5224973777475096e-07,
+      "loss": 0.2622,
+      "step": 3583
+    },
+    {
+      "epoch": 1.8044108380592312,
+      "grad_norm": 0.1819103774919584,
+      "learning_rate": 3.519931941122066e-07,
+      "loss": 0.2749,
+      "step": 3584
+    },
+    {
+      "epoch": 1.804914933837429,
+      "grad_norm": 0.1717127969736599,
+      "learning_rate": 3.517366931487089e-07,
+      "loss": 0.2703,
+      "step": 3585
+    },
+    {
+      "epoch": 1.805419029615627,
+      "grad_norm": 0.17885330880659442,
+      "learning_rate": 3.514802349582569e-07,
+      "loss": 0.2907,
+      "step": 3586
+    },
+    {
+      "epoch": 1.8059231253938248,
+      "grad_norm": 0.18707259558766182,
+      "learning_rate": 3.5122381961483684e-07,
+      "loss": 0.276,
+      "step": 3587
+    },
+    {
+      "epoch": 1.8064272211720227,
+      "grad_norm": 0.1782099276673349,
+      "learning_rate": 3.509674471924234e-07,
+      "loss": 0.2716,
+      "step": 3588
+    },
+    {
+      "epoch": 1.8069313169502206,
+      "grad_norm": 0.17525854196465432,
+      "learning_rate": 3.507111177649783e-07,
+      "loss": 0.2575,
+      "step": 3589
+    },
+    {
+      "epoch": 1.8074354127284185,
+      "grad_norm": 0.18183285195462048,
+      "learning_rate": 3.5045483140645063e-07,
+      "loss": 0.2871,
+      "step": 3590
+    },
+    {
+      "epoch": 1.807939508506616,
+      "grad_norm": 0.16829292657965922,
+      "learning_rate": 3.50198588190778e-07,
+      "loss": 0.277,
+      "step": 3591
+    },
+    {
+      "epoch": 1.8084436042848142,
+      "grad_norm": 0.16750119184558296,
+      "learning_rate": 3.4994238819188446e-07,
+      "loss": 0.2688,
+      "step": 3592
+    },
+    {
+      "epoch": 1.8089477000630119,
+      "grad_norm": 0.19435552393670724,
+      "learning_rate": 3.496862314836826e-07,
+      "loss": 0.2752,
+      "step": 3593
+    },
+    {
+      "epoch": 1.80945179584121,
+      "grad_norm": 0.17151739988556397,
+      "learning_rate": 3.4943011814007154e-07,
+      "loss": 0.2752,
+      "step": 3594
+    },
+    {
+      "epoch": 1.8099558916194076,
+      "grad_norm": 0.17044681265940018,
+      "learning_rate": 3.4917404823493895e-07,
+      "loss": 0.2764,
+      "step": 3595
+    },
+    {
+      "epoch": 1.8104599873976055,
+      "grad_norm": 0.1706463322105864,
+      "learning_rate": 3.489180218421591e-07,
+      "loss": 0.2795,
+      "step": 3596
+    },
+    {
+      "epoch": 1.8109640831758034,
+      "grad_norm": 0.1743724675299324,
+      "learning_rate": 3.486620390355939e-07,
+      "loss": 0.2778,
+      "step": 3597
+    },
+    {
+      "epoch": 1.8114681789540013,
+      "grad_norm": 0.1672255541484769,
+      "learning_rate": 3.48406099889093e-07,
+      "loss": 0.2602,
+      "step": 3598
+    },
+    {
+      "epoch": 1.8119722747321991,
+      "grad_norm": 0.17213091184076965,
+      "learning_rate": 3.4815020447649325e-07,
+      "loss": 0.2662,
+      "step": 3599
+    },
+    {
+      "epoch": 1.812476370510397,
+      "grad_norm": 0.17382490282350782,
+      "learning_rate": 3.478943528716189e-07,
+      "loss": 0.2702,
+      "step": 3600
+    },
+    {
+      "epoch": 1.8129804662885949,
+      "grad_norm": 0.1707010922925502,
+      "learning_rate": 3.476385451482813e-07,
+      "loss": 0.2694,
+      "step": 3601
+    },
+    {
+      "epoch": 1.8134845620667925,
+      "grad_norm": 0.16861692088845995,
+      "learning_rate": 3.473827813802798e-07,
+      "loss": 0.2682,
+      "step": 3602
+    },
+    {
+      "epoch": 1.8139886578449906,
+      "grad_norm": 0.17226530366785384,
+      "learning_rate": 3.4712706164140044e-07,
+      "loss": 0.2707,
+      "step": 3603
+    },
+    {
+      "epoch": 1.8144927536231883,
+      "grad_norm": 0.17374229482438772,
+      "learning_rate": 3.468713860054166e-07,
+      "loss": 0.2665,
+      "step": 3604
+    },
+    {
+      "epoch": 1.8149968494013864,
+      "grad_norm": 0.17166711837546353,
+      "learning_rate": 3.466157545460895e-07,
+      "loss": 0.2614,
+      "step": 3605
+    },
+    {
+      "epoch": 1.815500945179584,
+      "grad_norm": 0.18055980554179216,
+      "learning_rate": 3.463601673371669e-07,
+      "loss": 0.286,
+      "step": 3606
+    },
+    {
+      "epoch": 1.816005040957782,
+      "grad_norm": 0.16875286919833732,
+      "learning_rate": 3.4610462445238447e-07,
+      "loss": 0.2754,
+      "step": 3607
+    },
+    {
+      "epoch": 1.8165091367359798,
+      "grad_norm": 0.17716433583796581,
+      "learning_rate": 3.4584912596546435e-07,
+      "loss": 0.2741,
+      "step": 3608
+    },
+    {
+      "epoch": 1.8170132325141777,
+      "grad_norm": 0.17475847439692482,
+      "learning_rate": 3.4559367195011663e-07,
+      "loss": 0.2709,
+      "step": 3609
+    },
+    {
+      "epoch": 1.8175173282923756,
+      "grad_norm": 0.17444775099348855,
+      "learning_rate": 3.4533826248003807e-07,
+      "loss": 0.2818,
+      "step": 3610
+    },
+    {
+      "epoch": 1.8180214240705734,
+      "grad_norm": 0.16496716406158368,
+      "learning_rate": 3.4508289762891253e-07,
+      "loss": 0.2597,
+      "step": 3611
+    },
+    {
+      "epoch": 1.8185255198487713,
+      "grad_norm": 0.17214388173996828,
+      "learning_rate": 3.448275774704116e-07,
+      "loss": 0.2858,
+      "step": 3612
+    },
+    {
+      "epoch": 1.819029615626969,
+      "grad_norm": 0.17019231591147854,
+      "learning_rate": 3.4457230207819317e-07,
+      "loss": 0.2678,
+      "step": 3613
+    },
+    {
+      "epoch": 1.819533711405167,
+      "grad_norm": 0.17699657912936148,
+      "learning_rate": 3.443170715259027e-07,
+      "loss": 0.2577,
+      "step": 3614
+    },
+    {
+      "epoch": 1.8200378071833647,
+      "grad_norm": 0.1752044669776778,
+      "learning_rate": 3.440618858871728e-07,
+      "loss": 0.2641,
+      "step": 3615
+    },
+    {
+      "epoch": 1.8205419029615628,
+      "grad_norm": 0.1796499070557965,
+      "learning_rate": 3.438067452356229e-07,
+      "loss": 0.2818,
+      "step": 3616
+    },
+    {
+      "epoch": 1.8210459987397605,
+      "grad_norm": 0.17743300718849356,
+      "learning_rate": 3.435516496448594e-07,
+      "loss": 0.2632,
+      "step": 3617
+    },
+    {
+      "epoch": 1.8215500945179586,
+      "grad_norm": 0.17048364525336646,
+      "learning_rate": 3.4329659918847574e-07,
+      "loss": 0.2797,
+      "step": 3618
+    },
+    {
+      "epoch": 1.8220541902961562,
+      "grad_norm": 0.17700606194697294,
+      "learning_rate": 3.430415939400526e-07,
+      "loss": 0.2738,
+      "step": 3619
+    },
+    {
+      "epoch": 1.822558286074354,
+      "grad_norm": 0.1732841514749814,
+      "learning_rate": 3.4278663397315714e-07,
+      "loss": 0.2767,
+      "step": 3620
+    },
+    {
+      "epoch": 1.823062381852552,
+      "grad_norm": 0.1746313075508639,
+      "learning_rate": 3.4253171936134407e-07,
+      "loss": 0.2737,
+      "step": 3621
+    },
+    {
+      "epoch": 1.8235664776307499,
+      "grad_norm": 0.16856675923392617,
+      "learning_rate": 3.4227685017815435e-07,
+      "loss": 0.2704,
+      "step": 3622
+    },
+    {
+      "epoch": 1.8240705734089477,
+      "grad_norm": 0.17817947743263224,
+      "learning_rate": 3.420220264971164e-07,
+      "loss": 0.2633,
+      "step": 3623
+    },
+    {
+      "epoch": 1.8245746691871454,
+      "grad_norm": 0.17390789109467564,
+      "learning_rate": 3.417672483917451e-07,
+      "loss": 0.2631,
+      "step": 3624
+    },
+    {
+      "epoch": 1.8250787649653435,
+      "grad_norm": 0.170222280271219,
+      "learning_rate": 3.4151251593554255e-07,
+      "loss": 0.2666,
+      "step": 3625
+    },
+    {
+      "epoch": 1.8255828607435411,
+      "grad_norm": 0.18503069700309455,
+      "learning_rate": 3.4125782920199744e-07,
+      "loss": 0.2727,
+      "step": 3626
+    },
+    {
+      "epoch": 1.8260869565217392,
+      "grad_norm": 0.17199894183725506,
+      "learning_rate": 3.41003188264585e-07,
+      "loss": 0.268,
+      "step": 3627
+    },
+    {
+      "epoch": 1.826591052299937,
+      "grad_norm": 0.17933515312179926,
+      "learning_rate": 3.407485931967681e-07,
+      "loss": 0.2793,
+      "step": 3628
+    },
+    {
+      "epoch": 1.827095148078135,
+      "grad_norm": 0.17059652406791323,
+      "learning_rate": 3.404940440719953e-07,
+      "loss": 0.2635,
+      "step": 3629
+    },
+    {
+      "epoch": 1.8275992438563327,
+      "grad_norm": 0.17365046482826368,
+      "learning_rate": 3.4023954096370285e-07,
+      "loss": 0.2762,
+      "step": 3630
+    },
+    {
+      "epoch": 1.8281033396345305,
+      "grad_norm": 0.17048013344794072,
+      "learning_rate": 3.399850839453131e-07,
+      "loss": 0.2672,
+      "step": 3631
+    },
+    {
+      "epoch": 1.8286074354127284,
+      "grad_norm": 0.18254834974101053,
+      "learning_rate": 3.397306730902355e-07,
+      "loss": 0.2855,
+      "step": 3632
+    },
+    {
+      "epoch": 1.8291115311909263,
+      "grad_norm": 0.1782057659837957,
+      "learning_rate": 3.394763084718659e-07,
+      "loss": 0.2723,
+      "step": 3633
+    },
+    {
+      "epoch": 1.8296156269691242,
+      "grad_norm": 0.1869729195366187,
+      "learning_rate": 3.3922199016358663e-07,
+      "loss": 0.2815,
+      "step": 3634
+    },
+    {
+      "epoch": 1.830119722747322,
+      "grad_norm": 0.17300425848362272,
+      "learning_rate": 3.389677182387676e-07,
+      "loss": 0.2766,
+      "step": 3635
+    },
+    {
+      "epoch": 1.83062381852552,
+      "grad_norm": 0.17409263469313568,
+      "learning_rate": 3.38713492770764e-07,
+      "loss": 0.2725,
+      "step": 3636
+    },
+    {
+      "epoch": 1.8311279143037176,
+      "grad_norm": 0.17168330174108185,
+      "learning_rate": 3.384593138329188e-07,
+      "loss": 0.2636,
+      "step": 3637
+    },
+    {
+      "epoch": 1.8316320100819157,
+      "grad_norm": 0.17321147308428417,
+      "learning_rate": 3.382051814985607e-07,
+      "loss": 0.2643,
+      "step": 3638
+    },
+    {
+      "epoch": 1.8321361058601133,
+      "grad_norm": 0.17382144255134466,
+      "learning_rate": 3.3795109584100565e-07,
+      "loss": 0.2774,
+      "step": 3639
+    },
+    {
+      "epoch": 1.8326402016383114,
+      "grad_norm": 0.17237782646607822,
+      "learning_rate": 3.376970569335557e-07,
+      "loss": 0.2912,
+      "step": 3640
+    },
+    {
+      "epoch": 1.833144297416509,
+      "grad_norm": 0.18252171874104944,
+      "learning_rate": 3.374430648494991e-07,
+      "loss": 0.2874,
+      "step": 3641
+    },
+    {
+      "epoch": 1.833648393194707,
+      "grad_norm": 0.17047186028552003,
+      "learning_rate": 3.371891196621116e-07,
+      "loss": 0.2752,
+      "step": 3642
+    },
+    {
+      "epoch": 1.8341524889729048,
+      "grad_norm": 0.1670392087043219,
+      "learning_rate": 3.3693522144465426e-07,
+      "loss": 0.2659,
+      "step": 3643
+    },
+    {
+      "epoch": 1.8346565847511027,
+      "grad_norm": 0.173528591535757,
+      "learning_rate": 3.366813702703757e-07,
+      "loss": 0.2817,
+      "step": 3644
+    },
+    {
+      "epoch": 1.8351606805293006,
+      "grad_norm": 0.17390840294297605,
+      "learning_rate": 3.364275662125099e-07,
+      "loss": 0.2736,
+      "step": 3645
+    },
+    {
+      "epoch": 1.8356647763074985,
+      "grad_norm": 0.16690990450267892,
+      "learning_rate": 3.361738093442781e-07,
+      "loss": 0.2603,
+      "step": 3646
+    },
+    {
+      "epoch": 1.8361688720856963,
+      "grad_norm": 0.17814929644569416,
+      "learning_rate": 3.359200997388874e-07,
+      "loss": 0.2645,
+      "step": 3647
+    },
+    {
+      "epoch": 1.836672967863894,
+      "grad_norm": 0.18228489937252948,
+      "learning_rate": 3.3566643746953136e-07,
+      "loss": 0.2847,
+      "step": 3648
+    },
+    {
+      "epoch": 1.837177063642092,
+      "grad_norm": 0.17393875067169687,
+      "learning_rate": 3.354128226093903e-07,
+      "loss": 0.2589,
+      "step": 3649
+    },
+    {
+      "epoch": 1.8376811594202898,
+      "grad_norm": 0.17147582807039788,
+      "learning_rate": 3.351592552316302e-07,
+      "loss": 0.2617,
+      "step": 3650
+    },
+    {
+      "epoch": 1.8381852551984879,
+      "grad_norm": 0.17359008786164207,
+      "learning_rate": 3.34905735409404e-07,
+      "loss": 0.2672,
+      "step": 3651
+    },
+    {
+      "epoch": 1.8386893509766855,
+      "grad_norm": 0.17091069458322225,
+      "learning_rate": 3.346522632158502e-07,
+      "loss": 0.2737,
+      "step": 3652
+    },
+    {
+      "epoch": 1.8391934467548834,
+      "grad_norm": 0.17036599489519516,
+      "learning_rate": 3.343988387240945e-07,
+      "loss": 0.2655,
+      "step": 3653
+    },
+    {
+      "epoch": 1.8396975425330813,
+      "grad_norm": 0.1751804328426936,
+      "learning_rate": 3.3414546200724765e-07,
+      "loss": 0.2778,
+      "step": 3654
+    },
+    {
+      "epoch": 1.8402016383112791,
+      "grad_norm": 0.1705462977459914,
+      "learning_rate": 3.338921331384078e-07,
+      "loss": 0.2775,
+      "step": 3655
+    },
+    {
+      "epoch": 1.840705734089477,
+      "grad_norm": 0.17625220305705913,
+      "learning_rate": 3.336388521906587e-07,
+      "loss": 0.271,
+      "step": 3656
+    },
+    {
+      "epoch": 1.841209829867675,
+      "grad_norm": 0.1711600986441844,
+      "learning_rate": 3.333856192370699e-07,
+      "loss": 0.2773,
+      "step": 3657
+    },
+    {
+      "epoch": 1.8417139256458728,
+      "grad_norm": 0.1861777517647817,
+      "learning_rate": 3.3313243435069796e-07,
+      "loss": 0.278,
+      "step": 3658
+    },
+    {
+      "epoch": 1.8422180214240704,
+      "grad_norm": 0.17302149508680212,
+      "learning_rate": 3.328792976045849e-07,
+      "loss": 0.2668,
+      "step": 3659
+    },
+    {
+      "epoch": 1.8427221172022685,
+      "grad_norm": 0.17199119227595985,
+      "learning_rate": 3.3262620907175935e-07,
+      "loss": 0.2591,
+      "step": 3660
+    },
+    {
+      "epoch": 1.8432262129804662,
+      "grad_norm": 0.17344070020135455,
+      "learning_rate": 3.323731688252356e-07,
+      "loss": 0.2716,
+      "step": 3661
+    },
+    {
+      "epoch": 1.8437303087586643,
+      "grad_norm": 0.1724615950455862,
+      "learning_rate": 3.3212017693801444e-07,
+      "loss": 0.2701,
+      "step": 3662
+    },
+    {
+      "epoch": 1.844234404536862,
+      "grad_norm": 0.1751562140997247,
+      "learning_rate": 3.3186723348308204e-07,
+      "loss": 0.2808,
+      "step": 3663
+    },
+    {
+      "epoch": 1.8447385003150598,
+      "grad_norm": 0.18510250956865856,
+      "learning_rate": 3.316143385334113e-07,
+      "loss": 0.2698,
+      "step": 3664
+    },
+    {
+      "epoch": 1.8452425960932577,
+      "grad_norm": 0.19779933795030186,
+      "learning_rate": 3.3136149216196094e-07,
+      "loss": 0.2812,
+      "step": 3665
+    },
+    {
+      "epoch": 1.8457466918714556,
+      "grad_norm": 0.17539119190444977,
+      "learning_rate": 3.311086944416752e-07,
+      "loss": 0.267,
+      "step": 3666
+    },
+    {
+      "epoch": 1.8462507876496534,
+      "grad_norm": 0.1688698268367981,
+      "learning_rate": 3.3085594544548506e-07,
+      "loss": 0.2706,
+      "step": 3667
+    },
+    {
+      "epoch": 1.8467548834278513,
+      "grad_norm": 0.16370546456235416,
+      "learning_rate": 3.306032452463067e-07,
+      "loss": 0.2593,
+      "step": 3668
+    },
+    {
+      "epoch": 1.8472589792060492,
+      "grad_norm": 0.16664793193380006,
+      "learning_rate": 3.303505939170429e-07,
+      "loss": 0.2747,
+      "step": 3669
+    },
+    {
+      "epoch": 1.8477630749842469,
+      "grad_norm": 0.1920173379347818,
+      "learning_rate": 3.3009799153058185e-07,
+      "loss": 0.2808,
+      "step": 3670
+    },
+    {
+      "epoch": 1.848267170762445,
+      "grad_norm": 0.17169232742511578,
+      "learning_rate": 3.298454381597976e-07,
+      "loss": 0.2737,
+      "step": 3671
+    },
+    {
+      "epoch": 1.8487712665406426,
+      "grad_norm": 0.1768506079726659,
+      "learning_rate": 3.295929338775505e-07,
+      "loss": 0.2821,
+      "step": 3672
+    },
+    {
+      "epoch": 1.8492753623188407,
+      "grad_norm": 0.17791535437923783,
+      "learning_rate": 3.2934047875668624e-07,
+      "loss": 0.3001,
+      "step": 3673
+    },
+    {
+      "epoch": 1.8497794580970384,
+      "grad_norm": 0.1739360428422386,
+      "learning_rate": 3.290880728700368e-07,
+      "loss": 0.2636,
+      "step": 3674
+    },
+    {
+      "epoch": 1.8502835538752362,
+      "grad_norm": 0.17787282666122797,
+      "learning_rate": 3.2883571629041964e-07,
+      "loss": 0.2658,
+      "step": 3675
+    },
+    {
+      "epoch": 1.8507876496534341,
+      "grad_norm": 0.17402992624760774,
+      "learning_rate": 3.2858340909063785e-07,
+      "loss": 0.2821,
+      "step": 3676
+    },
+    {
+      "epoch": 1.851291745431632,
+      "grad_norm": 0.16863049632313487,
+      "learning_rate": 3.283311513434809e-07,
+      "loss": 0.2762,
+      "step": 3677
+    },
+    {
+      "epoch": 1.8517958412098299,
+      "grad_norm": 0.17175916416523554,
+      "learning_rate": 3.280789431217231e-07,
+      "loss": 0.2646,
+      "step": 3678
+    },
+    {
+      "epoch": 1.8522999369880278,
+      "grad_norm": 0.1816439994226865,
+      "learning_rate": 3.278267844981254e-07,
+      "loss": 0.2751,
+      "step": 3679
+    },
+    {
+      "epoch": 1.8528040327662256,
+      "grad_norm": 0.17768054704900385,
+      "learning_rate": 3.275746755454337e-07,
+      "loss": 0.2934,
+      "step": 3680
+    },
+    {
+      "epoch": 1.8533081285444233,
+      "grad_norm": 0.17526569011593024,
+      "learning_rate": 3.2732261633638014e-07,
+      "loss": 0.2783,
+      "step": 3681
+    },
+    {
+      "epoch": 1.8538122243226214,
+      "grad_norm": 0.1683831095084952,
+      "learning_rate": 3.2707060694368185e-07,
+      "loss": 0.271,
+      "step": 3682
+    },
+    {
+      "epoch": 1.854316320100819,
+      "grad_norm": 0.184280262405193,
+      "learning_rate": 3.268186474400424e-07,
+      "loss": 0.2726,
+      "step": 3683
+    },
+    {
+      "epoch": 1.8548204158790171,
+      "grad_norm": 0.17522567840778597,
+      "learning_rate": 3.2656673789815045e-07,
+      "loss": 0.2847,
+      "step": 3684
+    },
+    {
+      "epoch": 1.8553245116572148,
+      "grad_norm": 0.17558489027707613,
+      "learning_rate": 3.2631487839067995e-07,
+      "loss": 0.2865,
+      "step": 3685
+    },
+    {
+      "epoch": 1.855828607435413,
+      "grad_norm": 0.17002382304061303,
+      "learning_rate": 3.260630689902913e-07,
+      "loss": 0.2491,
+      "step": 3686
+    },
+    {
+      "epoch": 1.8563327032136105,
+      "grad_norm": 0.17988503644984635,
+      "learning_rate": 3.2581130976962966e-07,
+      "loss": 0.2584,
+      "step": 3687
+    },
+    {
+      "epoch": 1.8568367989918084,
+      "grad_norm": 0.17012660875417476,
+      "learning_rate": 3.255596008013263e-07,
+      "loss": 0.2755,
+      "step": 3688
+    },
+    {
+      "epoch": 1.8573408947700063,
+      "grad_norm": 0.17339868792540347,
+      "learning_rate": 3.2530794215799726e-07,
+      "loss": 0.2957,
+      "step": 3689
+    },
+    {
+      "epoch": 1.8578449905482042,
+      "grad_norm": 0.17081604800735575,
+      "learning_rate": 3.2505633391224497e-07,
+      "loss": 0.2619,
+      "step": 3690
+    },
+    {
+      "epoch": 1.858349086326402,
+      "grad_norm": 0.17873255506104102,
+      "learning_rate": 3.248047761366566e-07,
+      "loss": 0.2494,
+      "step": 3691
+    },
+    {
+      "epoch": 1.8588531821045997,
+      "grad_norm": 0.18817722225475367,
+      "learning_rate": 3.2455326890380493e-07,
+      "loss": 0.2767,
+      "step": 3692
+    },
+    {
+      "epoch": 1.8593572778827978,
+      "grad_norm": 0.17152115532488543,
+      "learning_rate": 3.243018122862484e-07,
+      "loss": 0.2821,
+      "step": 3693
+    },
+    {
+      "epoch": 1.8598613736609955,
+      "grad_norm": 0.16839244253657323,
+      "learning_rate": 3.240504063565307e-07,
+      "loss": 0.2529,
+      "step": 3694
+    },
+    {
+      "epoch": 1.8603654694391936,
+      "grad_norm": 0.1730644160338181,
+      "learning_rate": 3.2379905118718075e-07,
+      "loss": 0.2669,
+      "step": 3695
+    },
+    {
+      "epoch": 1.8608695652173912,
+      "grad_norm": 0.1760485174615298,
+      "learning_rate": 3.2354774685071297e-07,
+      "loss": 0.2525,
+      "step": 3696
+    },
+    {
+      "epoch": 1.8613736609955893,
+      "grad_norm": 0.1724016942572103,
+      "learning_rate": 3.232964934196273e-07,
+      "loss": 0.2646,
+      "step": 3697
+    },
+    {
+      "epoch": 1.861877756773787,
+      "grad_norm": 0.1695018921199434,
+      "learning_rate": 3.230452909664084e-07,
+      "loss": 0.2597,
+      "step": 3698
+    },
+    {
+      "epoch": 1.8623818525519849,
+      "grad_norm": 0.17585413997700072,
+      "learning_rate": 3.2279413956352713e-07,
+      "loss": 0.2672,
+      "step": 3699
+    },
+    {
+      "epoch": 1.8628859483301827,
+      "grad_norm": 0.17301030936638456,
+      "learning_rate": 3.2254303928343886e-07,
+      "loss": 0.2656,
+      "step": 3700
+    },
+    {
+      "epoch": 1.8633900441083806,
+      "grad_norm": 0.17993787406188083,
+      "learning_rate": 3.2229199019858426e-07,
+      "loss": 0.2731,
+      "step": 3701
+    },
+    {
+      "epoch": 1.8638941398865785,
+      "grad_norm": 0.17406377463523626,
+      "learning_rate": 3.2204099238138986e-07,
+      "loss": 0.2663,
+      "step": 3702
+    },
+    {
+      "epoch": 1.8643982356647764,
+      "grad_norm": 0.17543670483670815,
+      "learning_rate": 3.217900459042666e-07,
+      "loss": 0.2744,
+      "step": 3703
+    },
+    {
+      "epoch": 1.8649023314429742,
+      "grad_norm": 0.18680717781511091,
+      "learning_rate": 3.2153915083961124e-07,
+      "loss": 0.2768,
+      "step": 3704
+    },
+    {
+      "epoch": 1.865406427221172,
+      "grad_norm": 0.1799452415248967,
+      "learning_rate": 3.2128830725980527e-07,
+      "loss": 0.2829,
+      "step": 3705
+    },
+    {
+      "epoch": 1.86591052299937,
+      "grad_norm": 0.17633674870398874,
+      "learning_rate": 3.210375152372157e-07,
+      "loss": 0.2732,
+      "step": 3706
+    },
+    {
+      "epoch": 1.8664146187775676,
+      "grad_norm": 0.17843417862899802,
+      "learning_rate": 3.207867748441945e-07,
+      "loss": 0.2672,
+      "step": 3707
+    },
+    {
+      "epoch": 1.8669187145557657,
+      "grad_norm": 0.1716848365869712,
+      "learning_rate": 3.2053608615307836e-07,
+      "loss": 0.2626,
+      "step": 3708
+    },
+    {
+      "epoch": 1.8674228103339634,
+      "grad_norm": 0.17365023761079248,
+      "learning_rate": 3.202854492361897e-07,
+      "loss": 0.2682,
+      "step": 3709
+    },
+    {
+      "epoch": 1.8679269061121613,
+      "grad_norm": 0.17180044980161524,
+      "learning_rate": 3.2003486416583566e-07,
+      "loss": 0.2616,
+      "step": 3710
+    },
+    {
+      "epoch": 1.8684310018903592,
+      "grad_norm": 0.1722207017048821,
+      "learning_rate": 3.1978433101430857e-07,
+      "loss": 0.2749,
+      "step": 3711
+    },
+    {
+      "epoch": 1.868935097668557,
+      "grad_norm": 0.16714750604744708,
+      "learning_rate": 3.1953384985388543e-07,
+      "loss": 0.2548,
+      "step": 3712
+    },
+    {
+      "epoch": 1.869439193446755,
+      "grad_norm": 0.17097699024786078,
+      "learning_rate": 3.192834207568288e-07,
+      "loss": 0.2771,
+      "step": 3713
+    },
+    {
+      "epoch": 1.8699432892249528,
+      "grad_norm": 0.18094945147962893,
+      "learning_rate": 3.1903304379538585e-07,
+      "loss": 0.2816,
+      "step": 3714
+    },
+    {
+      "epoch": 1.8704473850031507,
+      "grad_norm": 0.188619079337791,
+      "learning_rate": 3.1878271904178855e-07,
+      "loss": 0.2558,
+      "step": 3715
+    },
+    {
+      "epoch": 1.8709514807813483,
+      "grad_norm": 0.17733938393348464,
+      "learning_rate": 3.1853244656825446e-07,
+      "loss": 0.2689,
+      "step": 3716
+    },
+    {
+      "epoch": 1.8714555765595464,
+      "grad_norm": 0.17881276774598476,
+      "learning_rate": 3.1828222644698515e-07,
+      "loss": 0.2676,
+      "step": 3717
+    },
+    {
+      "epoch": 1.871959672337744,
+      "grad_norm": 0.18577840959577496,
+      "learning_rate": 3.1803205875016806e-07,
+      "loss": 0.2862,
+      "step": 3718
+    },
+    {
+      "epoch": 1.8724637681159422,
+      "grad_norm": 0.1708409232652426,
+      "learning_rate": 3.1778194354997456e-07,
+      "loss": 0.2637,
+      "step": 3719
+    },
+    {
+      "epoch": 1.8729678638941398,
+      "grad_norm": 0.17019804041538505,
+      "learning_rate": 3.1753188091856176e-07,
+      "loss": 0.2607,
+      "step": 3720
+    },
+    {
+      "epoch": 1.8734719596723377,
+      "grad_norm": 0.17419845830941813,
+      "learning_rate": 3.172818709280709e-07,
+      "loss": 0.2902,
+      "step": 3721
+    },
+    {
+      "epoch": 1.8739760554505356,
+      "grad_norm": 0.16739241329355717,
+      "learning_rate": 3.1703191365062843e-07,
+      "loss": 0.2526,
+      "step": 3722
+    },
+    {
+      "epoch": 1.8744801512287335,
+      "grad_norm": 0.1736290579522879,
+      "learning_rate": 3.167820091583455e-07,
+      "loss": 0.2879,
+      "step": 3723
+    },
+    {
+      "epoch": 1.8749842470069313,
+      "grad_norm": 0.1829860060599394,
+      "learning_rate": 3.1653215752331784e-07,
+      "loss": 0.2577,
+      "step": 3724
+    },
+    {
+      "epoch": 1.8754883427851292,
+      "grad_norm": 0.17208808352446284,
+      "learning_rate": 3.1628235881762624e-07,
+      "loss": 0.2624,
+      "step": 3725
+    },
+    {
+      "epoch": 1.875992438563327,
+      "grad_norm": 0.17226579687443258,
+      "learning_rate": 3.160326131133361e-07,
+      "loss": 0.2724,
+      "step": 3726
+    },
+    {
+      "epoch": 1.8764965343415247,
+      "grad_norm": 0.1758528503018195,
+      "learning_rate": 3.1578292048249743e-07,
+      "loss": 0.2635,
+      "step": 3727
+    },
+    {
+      "epoch": 1.8770006301197228,
+      "grad_norm": 0.17602460233894715,
+      "learning_rate": 3.1553328099714493e-07,
+      "loss": 0.2698,
+      "step": 3728
+    },
+    {
+      "epoch": 1.8775047258979205,
+      "grad_norm": 0.17916344278800375,
+      "learning_rate": 3.152836947292984e-07,
+      "loss": 0.2717,
+      "step": 3729
+    },
+    {
+      "epoch": 1.8780088216761186,
+      "grad_norm": 0.17570265601183882,
+      "learning_rate": 3.1503416175096156e-07,
+      "loss": 0.2794,
+      "step": 3730
+    },
+    {
+      "epoch": 1.8785129174543163,
+      "grad_norm": 0.1730642589307263,
+      "learning_rate": 3.147846821341231e-07,
+      "loss": 0.2521,
+      "step": 3731
+    },
+    {
+      "epoch": 1.8790170132325141,
+      "grad_norm": 0.1773966697187365,
+      "learning_rate": 3.145352559507567e-07,
+      "loss": 0.2722,
+      "step": 3732
+    },
+    {
+      "epoch": 1.879521109010712,
+      "grad_norm": 0.17094044431659802,
+      "learning_rate": 3.1428588327281993e-07,
+      "loss": 0.2758,
+      "step": 3733
+    },
+    {
+      "epoch": 1.8800252047889099,
+      "grad_norm": 0.1851002728085005,
+      "learning_rate": 3.140365641722555e-07,
+      "loss": 0.2734,
+      "step": 3734
+    },
+    {
+      "epoch": 1.8805293005671078,
+      "grad_norm": 0.18808599356233432,
+      "learning_rate": 3.137872987209902e-07,
+      "loss": 0.2826,
+      "step": 3735
+    },
+    {
+      "epoch": 1.8810333963453056,
+      "grad_norm": 0.16938800737251228,
+      "learning_rate": 3.1353808699093583e-07,
+      "loss": 0.2684,
+      "step": 3736
+    },
+    {
+      "epoch": 1.8815374921235035,
+      "grad_norm": 0.17470028516437633,
+      "learning_rate": 3.132889290539883e-07,
+      "loss": 0.2722,
+      "step": 3737
+    },
+    {
+      "epoch": 1.8820415879017012,
+      "grad_norm": 0.17323545442596444,
+      "learning_rate": 3.13039824982028e-07,
+      "loss": 0.2674,
+      "step": 3738
+    },
+    {
+      "epoch": 1.8825456836798993,
+      "grad_norm": 0.168222975337303,
+      "learning_rate": 3.127907748469201e-07,
+      "loss": 0.2596,
+      "step": 3739
+    },
+    {
+      "epoch": 1.883049779458097,
+      "grad_norm": 0.18129268141555105,
+      "learning_rate": 3.12541778720514e-07,
+      "loss": 0.27,
+      "step": 3740
+    },
+    {
+      "epoch": 1.883553875236295,
+      "grad_norm": 0.1747582775376678,
+      "learning_rate": 3.122928366746434e-07,
+      "loss": 0.263,
+      "step": 3741
+    },
+    {
+      "epoch": 1.8840579710144927,
+      "grad_norm": 0.17513777293951543,
+      "learning_rate": 3.1204394878112665e-07,
+      "loss": 0.2868,
+      "step": 3742
+    },
+    {
+      "epoch": 1.8845620667926906,
+      "grad_norm": 0.16905872732868277,
+      "learning_rate": 3.1179511511176646e-07,
+      "loss": 0.2661,
+      "step": 3743
+    },
+    {
+      "epoch": 1.8850661625708884,
+      "grad_norm": 0.1711316990404876,
+      "learning_rate": 3.1154633573834973e-07,
+      "loss": 0.2722,
+      "step": 3744
+    },
+    {
+      "epoch": 1.8855702583490863,
+      "grad_norm": 0.1739717737065482,
+      "learning_rate": 3.112976107326475e-07,
+      "loss": 0.268,
+      "step": 3745
+    },
+    {
+      "epoch": 1.8860743541272842,
+      "grad_norm": 0.17944928774305824,
+      "learning_rate": 3.110489401664158e-07,
+      "loss": 0.2812,
+      "step": 3746
+    },
+    {
+      "epoch": 1.886578449905482,
+      "grad_norm": 0.1762596235261297,
+      "learning_rate": 3.108003241113942e-07,
+      "loss": 0.2604,
+      "step": 3747
+    },
+    {
+      "epoch": 1.88708254568368,
+      "grad_norm": 0.17582649188250596,
+      "learning_rate": 3.1055176263930725e-07,
+      "loss": 0.27,
+      "step": 3748
+    },
+    {
+      "epoch": 1.8875866414618776,
+      "grad_norm": 0.18349586560030343,
+      "learning_rate": 3.10303255821863e-07,
+      "loss": 0.2731,
+      "step": 3749
+    },
+    {
+      "epoch": 1.8880907372400757,
+      "grad_norm": 0.18096878339273567,
+      "learning_rate": 3.100548037307546e-07,
+      "loss": 0.2817,
+      "step": 3750
+    },
+    {
+      "epoch": 1.8885948330182734,
+      "grad_norm": 0.17078057550904882,
+      "learning_rate": 3.0980640643765867e-07,
+      "loss": 0.2677,
+      "step": 3751
+    },
+    {
+      "epoch": 1.8890989287964715,
+      "grad_norm": 0.16893384397010494,
+      "learning_rate": 3.09558064014236e-07,
+      "loss": 0.2664,
+      "step": 3752
+    },
+    {
+      "epoch": 1.889603024574669,
+      "grad_norm": 0.18217985296544154,
+      "learning_rate": 3.093097765321324e-07,
+      "loss": 0.2688,
+      "step": 3753
+    },
+    {
+      "epoch": 1.8901071203528672,
+      "grad_norm": 0.17031724602389556,
+      "learning_rate": 3.0906154406297677e-07,
+      "loss": 0.2675,
+      "step": 3754
+    },
+    {
+      "epoch": 1.8906112161310649,
+      "grad_norm": 0.17612084822756136,
+      "learning_rate": 3.0881336667838313e-07,
+      "loss": 0.259,
+      "step": 3755
+    },
+    {
+      "epoch": 1.8911153119092627,
+      "grad_norm": 0.1661742309383176,
+      "learning_rate": 3.085652444499488e-07,
+      "loss": 0.2555,
+      "step": 3756
+    },
+    {
+      "epoch": 1.8916194076874606,
+      "grad_norm": 0.17479975552258276,
+      "learning_rate": 3.0831717744925556e-07,
+      "loss": 0.2719,
+      "step": 3757
+    },
+    {
+      "epoch": 1.8921235034656585,
+      "grad_norm": 0.1723634999690176,
+      "learning_rate": 3.080691657478691e-07,
+      "loss": 0.2728,
+      "step": 3758
+    },
+    {
+      "epoch": 1.8926275992438564,
+      "grad_norm": 0.17468461669614443,
+      "learning_rate": 3.0782120941733954e-07,
+      "loss": 0.2765,
+      "step": 3759
+    },
+    {
+      "epoch": 1.893131695022054,
+      "grad_norm": 0.1710752190489546,
+      "learning_rate": 3.075733085292006e-07,
+      "loss": 0.2687,
+      "step": 3760
+    },
+    {
+      "epoch": 1.8936357908002521,
+      "grad_norm": 0.16637775831353452,
+      "learning_rate": 3.0732546315496986e-07,
+      "loss": 0.2613,
+      "step": 3761
+    },
+    {
+      "epoch": 1.8941398865784498,
+      "grad_norm": 0.18290536724761589,
+      "learning_rate": 3.070776733661497e-07,
+      "loss": 0.2847,
+      "step": 3762
+    },
+    {
+      "epoch": 1.8946439823566479,
+      "grad_norm": 0.17059102780201477,
+      "learning_rate": 3.068299392342255e-07,
+      "loss": 0.2651,
+      "step": 3763
+    },
+    {
+      "epoch": 1.8951480781348455,
+      "grad_norm": 0.17464195983436098,
+      "learning_rate": 3.065822608306674e-07,
+      "loss": 0.2901,
+      "step": 3764
+    },
+    {
+      "epoch": 1.8956521739130436,
+      "grad_norm": 0.16589700319021694,
+      "learning_rate": 3.063346382269286e-07,
+      "loss": 0.2665,
+      "step": 3765
+    },
+    {
+      "epoch": 1.8961562696912413,
+      "grad_norm": 0.17656288621526112,
+      "learning_rate": 3.060870714944473e-07,
+      "loss": 0.2642,
+      "step": 3766
+    },
+    {
+      "epoch": 1.8966603654694392,
+      "grad_norm": 0.18513022931135426,
+      "learning_rate": 3.058395607046446e-07,
+      "loss": 0.2714,
+      "step": 3767
+    },
+    {
+      "epoch": 1.897164461247637,
+      "grad_norm": 0.17250542864882773,
+      "learning_rate": 3.0559210592892567e-07,
+      "loss": 0.2758,
+      "step": 3768
+    },
+    {
+      "epoch": 1.897668557025835,
+      "grad_norm": 0.17243208247681027,
+      "learning_rate": 3.053447072386801e-07,
+      "loss": 0.2769,
+      "step": 3769
+    },
+    {
+      "epoch": 1.8981726528040328,
+      "grad_norm": 0.17667298757324432,
+      "learning_rate": 3.050973647052805e-07,
+      "loss": 0.2703,
+      "step": 3770
+    },
+    {
+      "epoch": 1.8986767485822307,
+      "grad_norm": 0.16775068473994517,
+      "learning_rate": 3.0485007840008394e-07,
+      "loss": 0.2497,
+      "step": 3771
+    },
+    {
+      "epoch": 1.8991808443604286,
+      "grad_norm": 0.17502025512014624,
+      "learning_rate": 3.046028483944308e-07,
+      "loss": 0.2672,
+      "step": 3772
+    },
+    {
+      "epoch": 1.8996849401386262,
+      "grad_norm": 0.17430596589676,
+      "learning_rate": 3.043556747596456e-07,
+      "loss": 0.2753,
+      "step": 3773
+    },
+    {
+      "epoch": 1.9001890359168243,
+      "grad_norm": 0.17212088868648046,
+      "learning_rate": 3.0410855756703614e-07,
+      "loss": 0.2712,
+      "step": 3774
+    },
+    {
+      "epoch": 1.900693131695022,
+      "grad_norm": 0.17745212599865196,
+      "learning_rate": 3.0386149688789434e-07,
+      "loss": 0.2847,
+      "step": 3775
+    },
+    {
+      "epoch": 1.90119722747322,
+      "grad_norm": 0.17153573115828954,
+      "learning_rate": 3.036144927934958e-07,
+      "loss": 0.276,
+      "step": 3776
+    },
+    {
+      "epoch": 1.9017013232514177,
+      "grad_norm": 0.1780316189262351,
+      "learning_rate": 3.033675453550994e-07,
+      "loss": 0.2849,
+      "step": 3777
+    },
+    {
+      "epoch": 1.9022054190296156,
+      "grad_norm": 0.1832700985266359,
+      "learning_rate": 3.031206546439482e-07,
+      "loss": 0.2594,
+      "step": 3778
+    },
+    {
+      "epoch": 1.9027095148078135,
+      "grad_norm": 0.1699394230622764,
+      "learning_rate": 3.0287382073126837e-07,
+      "loss": 0.2766,
+      "step": 3779
+    },
+    {
+      "epoch": 1.9032136105860114,
+      "grad_norm": 0.17706015901322072,
+      "learning_rate": 3.0262704368827036e-07,
+      "loss": 0.2752,
+      "step": 3780
+    },
+    {
+      "epoch": 1.9037177063642092,
+      "grad_norm": 0.1775906710352642,
+      "learning_rate": 3.0238032358614753e-07,
+      "loss": 0.2721,
+      "step": 3781
+    },
+    {
+      "epoch": 1.9037177063642092,
+      "eval_loss": 0.3058184087276459,
+      "eval_runtime": 17.9595,
+      "eval_samples_per_second": 47.607,
+      "eval_steps_per_second": 1.002,
+      "step": 3781
+    },
+    {
+      "epoch": 1.904221802142407,
+      "grad_norm": 0.18004522794254016,
+      "learning_rate": 3.0213366049607703e-07,
+      "loss": 0.2564,
+      "step": 3782
+    },
+    {
+      "epoch": 1.904725897920605,
+      "grad_norm": 0.17630616450078446,
+      "learning_rate": 3.0188705448921994e-07,
+      "loss": 0.2788,
+      "step": 3783
+    },
+    {
+      "epoch": 1.9052299936988026,
+      "grad_norm": 0.17486538442100466,
+      "learning_rate": 3.0164050563672004e-07,
+      "loss": 0.2699,
+      "step": 3784
+    },
+    {
+      "epoch": 1.9057340894770007,
+      "grad_norm": 0.17611096571524854,
+      "learning_rate": 3.0139401400970586e-07,
+      "loss": 0.2754,
+      "step": 3785
+    },
+    {
+      "epoch": 1.9062381852551984,
+      "grad_norm": 0.1737782779159714,
+      "learning_rate": 3.0114757967928816e-07,
+      "loss": 0.2657,
+      "step": 3786
+    },
+    {
+      "epoch": 1.9067422810333965,
+      "grad_norm": 0.1739385161722817,
+      "learning_rate": 3.0090120271656194e-07,
+      "loss": 0.2653,
+      "step": 3787
+    },
+    {
+      "epoch": 1.9072463768115941,
+      "grad_norm": 0.17747427811726257,
+      "learning_rate": 3.0065488319260535e-07,
+      "loss": 0.2728,
+      "step": 3788
+    },
+    {
+      "epoch": 1.907750472589792,
+      "grad_norm": 0.17131014859330937,
+      "learning_rate": 3.004086211784802e-07,
+      "loss": 0.2675,
+      "step": 3789
+    },
+    {
+      "epoch": 1.90825456836799,
+      "grad_norm": 0.17148523515796457,
+      "learning_rate": 3.001624167452315e-07,
+      "loss": 0.2672,
+      "step": 3790
+    },
+    {
+      "epoch": 1.9087586641461878,
+      "grad_norm": 0.16882516072988113,
+      "learning_rate": 2.999162699638873e-07,
+      "loss": 0.2637,
+      "step": 3791
+    },
+    {
+      "epoch": 1.9092627599243857,
+      "grad_norm": 0.1786046049841935,
+      "learning_rate": 2.996701809054601e-07,
+      "loss": 0.2848,
+      "step": 3792
+    },
+    {
+      "epoch": 1.9097668557025835,
+      "grad_norm": 0.17145262177042117,
+      "learning_rate": 2.994241496409444e-07,
+      "loss": 0.2712,
+      "step": 3793
+    },
+    {
+      "epoch": 1.9102709514807814,
+      "grad_norm": 0.171516448246734,
+      "learning_rate": 2.991781762413194e-07,
+      "loss": 0.2847,
+      "step": 3794
+    },
+    {
+      "epoch": 1.910775047258979,
+      "grad_norm": 0.17763020432437332,
+      "learning_rate": 2.989322607775462e-07,
+      "loss": 0.2837,
+      "step": 3795
+    },
+    {
+      "epoch": 1.9112791430371772,
+      "grad_norm": 0.1746370724053624,
+      "learning_rate": 2.986864033205704e-07,
+      "loss": 0.2727,
+      "step": 3796
+    },
+    {
+      "epoch": 1.9117832388153748,
+      "grad_norm": 0.185899284676683,
+      "learning_rate": 2.984406039413202e-07,
+      "loss": 0.2887,
+      "step": 3797
+    },
+    {
+      "epoch": 1.912287334593573,
+      "grad_norm": 0.16862868846992204,
+      "learning_rate": 2.98194862710707e-07,
+      "loss": 0.2645,
+      "step": 3798
+    },
+    {
+      "epoch": 1.9127914303717706,
+      "grad_norm": 0.17734208433370674,
+      "learning_rate": 2.9794917969962595e-07,
+      "loss": 0.2769,
+      "step": 3799
+    },
+    {
+      "epoch": 1.9132955261499685,
+      "grad_norm": 0.1646715304601581,
+      "learning_rate": 2.977035549789548e-07,
+      "loss": 0.2596,
+      "step": 3800
+    },
+    {
+      "epoch": 1.9137996219281663,
+      "grad_norm": 0.1718225932856344,
+      "learning_rate": 2.9745798861955497e-07,
+      "loss": 0.281,
+      "step": 3801
+    },
+    {
+      "epoch": 1.9143037177063642,
+      "grad_norm": 0.17618806208047236,
+      "learning_rate": 2.972124806922707e-07,
+      "loss": 0.2662,
+      "step": 3802
+    },
+    {
+      "epoch": 1.914807813484562,
+      "grad_norm": 0.1719512615295124,
+      "learning_rate": 2.9696703126792967e-07,
+      "loss": 0.2669,
+      "step": 3803
+    },
+    {
+      "epoch": 1.91531190926276,
+      "grad_norm": 0.18136678946666576,
+      "learning_rate": 2.967216404173423e-07,
+      "loss": 0.2602,
+      "step": 3804
+    },
+    {
+      "epoch": 1.9158160050409578,
+      "grad_norm": 0.1746419414840373,
+      "learning_rate": 2.9647630821130234e-07,
+      "loss": 0.2723,
+      "step": 3805
+    },
+    {
+      "epoch": 1.9163201008191555,
+      "grad_norm": 0.17198386237681593,
+      "learning_rate": 2.9623103472058685e-07,
+      "loss": 0.276,
+      "step": 3806
+    },
+    {
+      "epoch": 1.9168241965973536,
+      "grad_norm": 0.17409294867594544,
+      "learning_rate": 2.959858200159554e-07,
+      "loss": 0.2668,
+      "step": 3807
+    },
+    {
+      "epoch": 1.9173282923755512,
+      "grad_norm": 0.16517572018109172,
+      "learning_rate": 2.9574066416815123e-07,
+      "loss": 0.2668,
+      "step": 3808
+    },
+    {
+      "epoch": 1.9178323881537493,
+      "grad_norm": 0.17327174886971222,
+      "learning_rate": 2.9549556724789995e-07,
+      "loss": 0.2654,
+      "step": 3809
+    },
+    {
+      "epoch": 1.918336483931947,
+      "grad_norm": 0.17123477242224328,
+      "learning_rate": 2.952505293259108e-07,
+      "loss": 0.2666,
+      "step": 3810
+    },
+    {
+      "epoch": 1.9188405797101449,
+      "grad_norm": 0.1711720330959814,
+      "learning_rate": 2.950055504728757e-07,
+      "loss": 0.275,
+      "step": 3811
+    },
+    {
+      "epoch": 1.9193446754883428,
+      "grad_norm": 0.17427388711209385,
+      "learning_rate": 2.9476063075946915e-07,
+      "loss": 0.2762,
+      "step": 3812
+    },
+    {
+      "epoch": 1.9198487712665406,
+      "grad_norm": 0.17268175591003512,
+      "learning_rate": 2.945157702563494e-07,
+      "loss": 0.2744,
+      "step": 3813
+    },
+    {
+      "epoch": 1.9203528670447385,
+      "grad_norm": 0.16809166747283374,
+      "learning_rate": 2.9427096903415694e-07,
+      "loss": 0.2778,
+      "step": 3814
+    },
+    {
+      "epoch": 1.9208569628229364,
+      "grad_norm": 0.17330157587108844,
+      "learning_rate": 2.940262271635156e-07,
+      "loss": 0.2768,
+      "step": 3815
+    },
+    {
+      "epoch": 1.9213610586011343,
+      "grad_norm": 0.1738154543130853,
+      "learning_rate": 2.9378154471503156e-07,
+      "loss": 0.2878,
+      "step": 3816
+    },
+    {
+      "epoch": 1.921865154379332,
+      "grad_norm": 0.18288271259832445,
+      "learning_rate": 2.9353692175929475e-07,
+      "loss": 0.2671,
+      "step": 3817
+    },
+    {
+      "epoch": 1.92236925015753,
+      "grad_norm": 0.18141220460600882,
+      "learning_rate": 2.9329235836687684e-07,
+      "loss": 0.2806,
+      "step": 3818
+    },
+    {
+      "epoch": 1.9228733459357277,
+      "grad_norm": 0.18790688920159956,
+      "learning_rate": 2.930478546083331e-07,
+      "loss": 0.2681,
+      "step": 3819
+    },
+    {
+      "epoch": 1.9233774417139258,
+      "grad_norm": 0.16964152403213711,
+      "learning_rate": 2.9280341055420133e-07,
+      "loss": 0.2664,
+      "step": 3820
+    },
+    {
+      "epoch": 1.9238815374921234,
+      "grad_norm": 0.1671671212800262,
+      "learning_rate": 2.9255902627500204e-07,
+      "loss": 0.2641,
+      "step": 3821
+    },
+    {
+      "epoch": 1.9243856332703215,
+      "grad_norm": 0.17197335612344677,
+      "learning_rate": 2.923147018412387e-07,
+      "loss": 0.2696,
+      "step": 3822
+    },
+    {
+      "epoch": 1.9248897290485192,
+      "grad_norm": 0.17243819294137702,
+      "learning_rate": 2.920704373233972e-07,
+      "loss": 0.2754,
+      "step": 3823
+    },
+    {
+      "epoch": 1.925393824826717,
+      "grad_norm": 0.17043519865324416,
+      "learning_rate": 2.918262327919466e-07,
+      "loss": 0.2715,
+      "step": 3824
+    },
+    {
+      "epoch": 1.925897920604915,
+      "grad_norm": 0.1722409742574077,
+      "learning_rate": 2.915820883173383e-07,
+      "loss": 0.2578,
+      "step": 3825
+    },
+    {
+      "epoch": 1.9264020163831128,
+      "grad_norm": 0.1740557480804691,
+      "learning_rate": 2.9133800397000627e-07,
+      "loss": 0.2765,
+      "step": 3826
+    },
+    {
+      "epoch": 1.9269061121613107,
+      "grad_norm": 0.1680406260361292,
+      "learning_rate": 2.910939798203677e-07,
+      "loss": 0.2795,
+      "step": 3827
+    },
+    {
+      "epoch": 1.9274102079395083,
+      "grad_norm": 0.17149022418835733,
+      "learning_rate": 2.9085001593882187e-07,
+      "loss": 0.2604,
+      "step": 3828
+    },
+    {
+      "epoch": 1.9279143037177064,
+      "grad_norm": 0.17010503345411235,
+      "learning_rate": 2.9060611239575085e-07,
+      "loss": 0.2827,
+      "step": 3829
+    },
+    {
+      "epoch": 1.928418399495904,
+      "grad_norm": 0.1749493958223586,
+      "learning_rate": 2.9036226926151897e-07,
+      "loss": 0.2758,
+      "step": 3830
+    },
+    {
+      "epoch": 1.9289224952741022,
+      "grad_norm": 0.17008445038688108,
+      "learning_rate": 2.90118486606474e-07,
+      "loss": 0.2645,
+      "step": 3831
+    },
+    {
+      "epoch": 1.9294265910522999,
+      "grad_norm": 0.17652489240339891,
+      "learning_rate": 2.898747645009454e-07,
+      "loss": 0.2746,
+      "step": 3832
+    },
+    {
+      "epoch": 1.929930686830498,
+      "grad_norm": 0.22700479625148187,
+      "learning_rate": 2.896311030152457e-07,
+      "loss": 0.2753,
+      "step": 3833
+    },
+    {
+      "epoch": 1.9304347826086956,
+      "grad_norm": 0.17427450545863746,
+      "learning_rate": 2.8938750221966965e-07,
+      "loss": 0.2703,
+      "step": 3834
+    },
+    {
+      "epoch": 1.9309388783868935,
+      "grad_norm": 0.17151048823577203,
+      "learning_rate": 2.891439621844943e-07,
+      "loss": 0.2595,
+      "step": 3835
+    },
+    {
+      "epoch": 1.9314429741650914,
+      "grad_norm": 0.17149156103858243,
+      "learning_rate": 2.8890048297997985e-07,
+      "loss": 0.252,
+      "step": 3836
+    },
+    {
+      "epoch": 1.9319470699432892,
+      "grad_norm": 0.16817818345200922,
+      "learning_rate": 2.886570646763682e-07,
+      "loss": 0.2643,
+      "step": 3837
+    },
+    {
+      "epoch": 1.9324511657214871,
+      "grad_norm": 0.1694609499296146,
+      "learning_rate": 2.8841370734388444e-07,
+      "loss": 0.2774,
+      "step": 3838
+    },
+    {
+      "epoch": 1.9329552614996848,
+      "grad_norm": 0.18663748392100837,
+      "learning_rate": 2.8817041105273513e-07,
+      "loss": 0.2794,
+      "step": 3839
+    },
+    {
+      "epoch": 1.9334593572778829,
+      "grad_norm": 0.17290683655437977,
+      "learning_rate": 2.8792717587311027e-07,
+      "loss": 0.2751,
+      "step": 3840
+    },
+    {
+      "epoch": 1.9339634530560805,
+      "grad_norm": 0.17362487357823306,
+      "learning_rate": 2.876840018751814e-07,
+      "loss": 0.263,
+      "step": 3841
+    },
+    {
+      "epoch": 1.9344675488342786,
+      "grad_norm": 0.18573004563502468,
+      "learning_rate": 2.8744088912910257e-07,
+      "loss": 0.2749,
+      "step": 3842
+    },
+    {
+      "epoch": 1.9349716446124763,
+      "grad_norm": 0.16915731492502192,
+      "learning_rate": 2.8719783770501074e-07,
+      "loss": 0.2589,
+      "step": 3843
+    },
+    {
+      "epoch": 1.9354757403906744,
+      "grad_norm": 0.17885497820565566,
+      "learning_rate": 2.8695484767302423e-07,
+      "loss": 0.2829,
+      "step": 3844
+    },
+    {
+      "epoch": 1.935979836168872,
+      "grad_norm": 0.16662065368985032,
+      "learning_rate": 2.8671191910324466e-07,
+      "loss": 0.2675,
+      "step": 3845
+    },
+    {
+      "epoch": 1.93648393194707,
+      "grad_norm": 0.18153383756126074,
+      "learning_rate": 2.86469052065755e-07,
+      "loss": 0.2636,
+      "step": 3846
+    },
+    {
+      "epoch": 1.9369880277252678,
+      "grad_norm": 0.17318438090821503,
+      "learning_rate": 2.8622624663062125e-07,
+      "loss": 0.2942,
+      "step": 3847
+    },
+    {
+      "epoch": 1.9374921235034657,
+      "grad_norm": 0.1733848907134762,
+      "learning_rate": 2.859835028678911e-07,
+      "loss": 0.2782,
+      "step": 3848
+    },
+    {
+      "epoch": 1.9379962192816635,
+      "grad_norm": 0.16974026130832987,
+      "learning_rate": 2.8574082084759434e-07,
+      "loss": 0.2549,
+      "step": 3849
+    },
+    {
+      "epoch": 1.9385003150598614,
+      "grad_norm": 0.17091600219366612,
+      "learning_rate": 2.854982006397438e-07,
+      "loss": 0.2838,
+      "step": 3850
+    },
+    {
+      "epoch": 1.9390044108380593,
+      "grad_norm": 0.17588856131135513,
+      "learning_rate": 2.852556423143333e-07,
+      "loss": 0.2667,
+      "step": 3851
+    },
+    {
+      "epoch": 1.939508506616257,
+      "grad_norm": 0.16743774165838593,
+      "learning_rate": 2.8501314594133996e-07,
+      "loss": 0.2672,
+      "step": 3852
+    },
+    {
+      "epoch": 1.940012602394455,
+      "grad_norm": 0.17244186575284368,
+      "learning_rate": 2.8477071159072206e-07,
+      "loss": 0.2764,
+      "step": 3853
+    },
+    {
+      "epoch": 1.9405166981726527,
+      "grad_norm": 0.17790819827202037,
+      "learning_rate": 2.845283393324208e-07,
+      "loss": 0.2717,
+      "step": 3854
+    },
+    {
+      "epoch": 1.9410207939508508,
+      "grad_norm": 0.17254105972174483,
+      "learning_rate": 2.8428602923635894e-07,
+      "loss": 0.28,
+      "step": 3855
+    },
+    {
+      "epoch": 1.9415248897290485,
+      "grad_norm": 0.16993890956105945,
+      "learning_rate": 2.84043781372441e-07,
+      "loss": 0.2677,
+      "step": 3856
+    },
+    {
+      "epoch": 1.9420289855072463,
+      "grad_norm": 0.17692085437275623,
+      "learning_rate": 2.838015958105547e-07,
+      "loss": 0.2878,
+      "step": 3857
+    },
+    {
+      "epoch": 1.9425330812854442,
+      "grad_norm": 0.175016522473398,
+      "learning_rate": 2.8355947262056865e-07,
+      "loss": 0.2736,
+      "step": 3858
+    },
+    {
+      "epoch": 1.943037177063642,
+      "grad_norm": 0.1845066084934637,
+      "learning_rate": 2.833174118723338e-07,
+      "loss": 0.2712,
+      "step": 3859
+    },
+    {
+      "epoch": 1.94354127284184,
+      "grad_norm": 0.17292158533010546,
+      "learning_rate": 2.8307541363568356e-07,
+      "loss": 0.2648,
+      "step": 3860
+    },
+    {
+      "epoch": 1.9440453686200379,
+      "grad_norm": 0.17403302200728327,
+      "learning_rate": 2.8283347798043265e-07,
+      "loss": 0.2601,
+      "step": 3861
+    },
+    {
+      "epoch": 1.9445494643982357,
+      "grad_norm": 0.17209410998540525,
+      "learning_rate": 2.825916049763779e-07,
+      "loss": 0.2885,
+      "step": 3862
+    },
+    {
+      "epoch": 1.9450535601764334,
+      "grad_norm": 0.1730295963728701,
+      "learning_rate": 2.8234979469329856e-07,
+      "loss": 0.2773,
+      "step": 3863
+    },
+    {
+      "epoch": 1.9455576559546315,
+      "grad_norm": 0.17409432528563223,
+      "learning_rate": 2.8210804720095516e-07,
+      "loss": 0.2762,
+      "step": 3864
+    },
+    {
+      "epoch": 1.9460617517328291,
+      "grad_norm": 0.18150338712543956,
+      "learning_rate": 2.818663625690902e-07,
+      "loss": 0.2729,
+      "step": 3865
+    },
+    {
+      "epoch": 1.9465658475110272,
+      "grad_norm": 0.17569994960621013,
+      "learning_rate": 2.8162474086742854e-07,
+      "loss": 0.2756,
+      "step": 3866
+    },
+    {
+      "epoch": 1.947069943289225,
+      "grad_norm": 0.17879985777229393,
+      "learning_rate": 2.813831821656762e-07,
+      "loss": 0.2919,
+      "step": 3867
+    },
+    {
+      "epoch": 1.9475740390674228,
+      "grad_norm": 0.17230743106972518,
+      "learning_rate": 2.811416865335217e-07,
+      "loss": 0.276,
+      "step": 3868
+    },
+    {
+      "epoch": 1.9480781348456206,
+      "grad_norm": 0.17984069998116,
+      "learning_rate": 2.8090025404063477e-07,
+      "loss": 0.2785,
+      "step": 3869
+    },
+    {
+      "epoch": 1.9485822306238185,
+      "grad_norm": 0.1744539834333857,
+      "learning_rate": 2.8065888475666745e-07,
+      "loss": 0.2884,
+      "step": 3870
+    },
+    {
+      "epoch": 1.9490863264020164,
+      "grad_norm": 0.18027869943306707,
+      "learning_rate": 2.80417578751253e-07,
+      "loss": 0.278,
+      "step": 3871
+    },
+    {
+      "epoch": 1.9495904221802143,
+      "grad_norm": 0.17865984324678172,
+      "learning_rate": 2.801763360940068e-07,
+      "loss": 0.2602,
+      "step": 3872
+    },
+    {
+      "epoch": 1.9500945179584122,
+      "grad_norm": 0.17671158591958067,
+      "learning_rate": 2.7993515685452613e-07,
+      "loss": 0.2779,
+      "step": 3873
+    },
+    {
+      "epoch": 1.9505986137366098,
+      "grad_norm": 0.1762952439624626,
+      "learning_rate": 2.796940411023892e-07,
+      "loss": 0.2593,
+      "step": 3874
+    },
+    {
+      "epoch": 1.951102709514808,
+      "grad_norm": 0.1823791079352628,
+      "learning_rate": 2.794529889071569e-07,
+      "loss": 0.2758,
+      "step": 3875
+    },
+    {
+      "epoch": 1.9516068052930056,
+      "grad_norm": 0.178230741517296,
+      "learning_rate": 2.792120003383709e-07,
+      "loss": 0.2628,
+      "step": 3876
+    },
+    {
+      "epoch": 1.9521109010712037,
+      "grad_norm": 0.17483875950866737,
+      "learning_rate": 2.7897107546555525e-07,
+      "loss": 0.2742,
+      "step": 3877
+    },
+    {
+      "epoch": 1.9526149968494013,
+      "grad_norm": 0.17465758090677505,
+      "learning_rate": 2.787302143582152e-07,
+      "loss": 0.276,
+      "step": 3878
+    },
+    {
+      "epoch": 1.9531190926275992,
+      "grad_norm": 0.17129898524450557,
+      "learning_rate": 2.784894170858373e-07,
+      "loss": 0.2652,
+      "step": 3879
+    },
+    {
+      "epoch": 1.953623188405797,
+      "grad_norm": 0.1658101810588016,
+      "learning_rate": 2.782486837178907e-07,
+      "loss": 0.263,
+      "step": 3880
+    },
+    {
+      "epoch": 1.954127284183995,
+      "grad_norm": 0.1755462609743805,
+      "learning_rate": 2.780080143238249e-07,
+      "loss": 0.2587,
+      "step": 3881
+    },
+    {
+      "epoch": 1.9546313799621928,
+      "grad_norm": 0.17509315570998626,
+      "learning_rate": 2.7776740897307203e-07,
+      "loss": 0.2666,
+      "step": 3882
+    },
+    {
+      "epoch": 1.9551354757403907,
+      "grad_norm": 0.16889327157699743,
+      "learning_rate": 2.7752686773504486e-07,
+      "loss": 0.2704,
+      "step": 3883
+    },
+    {
+      "epoch": 1.9556395715185886,
+      "grad_norm": 0.17329963412035843,
+      "learning_rate": 2.7728639067913826e-07,
+      "loss": 0.2765,
+      "step": 3884
+    },
+    {
+      "epoch": 1.9561436672967862,
+      "grad_norm": 0.17191012346840492,
+      "learning_rate": 2.7704597787472825e-07,
+      "loss": 0.2689,
+      "step": 3885
+    },
+    {
+      "epoch": 1.9566477630749843,
+      "grad_norm": 0.16839979801264043,
+      "learning_rate": 2.7680562939117265e-07,
+      "loss": 0.2665,
+      "step": 3886
+    },
+    {
+      "epoch": 1.957151858853182,
+      "grad_norm": 0.17446755444392845,
+      "learning_rate": 2.765653452978103e-07,
+      "loss": 0.2837,
+      "step": 3887
+    },
+    {
+      "epoch": 1.95765595463138,
+      "grad_norm": 0.1726644996987495,
+      "learning_rate": 2.7632512566396185e-07,
+      "loss": 0.2825,
+      "step": 3888
+    },
+    {
+      "epoch": 1.9581600504095777,
+      "grad_norm": 0.18240867556012677,
+      "learning_rate": 2.7608497055892877e-07,
+      "loss": 0.2799,
+      "step": 3889
+    },
+    {
+      "epoch": 1.9586641461877756,
+      "grad_norm": 0.1852368217976518,
+      "learning_rate": 2.758448800519948e-07,
+      "loss": 0.2774,
+      "step": 3890
+    },
+    {
+      "epoch": 1.9591682419659735,
+      "grad_norm": 0.17519851532629838,
+      "learning_rate": 2.756048542124244e-07,
+      "loss": 0.2664,
+      "step": 3891
+    },
+    {
+      "epoch": 1.9596723377441714,
+      "grad_norm": 0.17550478596249003,
+      "learning_rate": 2.7536489310946325e-07,
+      "loss": 0.2765,
+      "step": 3892
+    },
+    {
+      "epoch": 1.9601764335223693,
+      "grad_norm": 0.17986773014699392,
+      "learning_rate": 2.751249968123391e-07,
+      "loss": 0.2741,
+      "step": 3893
+    },
+    {
+      "epoch": 1.9606805293005671,
+      "grad_norm": 0.18286596209791772,
+      "learning_rate": 2.748851653902604e-07,
+      "loss": 0.274,
+      "step": 3894
+    },
+    {
+      "epoch": 1.961184625078765,
+      "grad_norm": 0.16845294749892417,
+      "learning_rate": 2.7464539891241677e-07,
+      "loss": 0.2706,
+      "step": 3895
+    },
+    {
+      "epoch": 1.9616887208569627,
+      "grad_norm": 0.1667625234228712,
+      "learning_rate": 2.744056974479798e-07,
+      "loss": 0.2689,
+      "step": 3896
+    },
+    {
+      "epoch": 1.9621928166351608,
+      "grad_norm": 0.17330081052295093,
+      "learning_rate": 2.741660610661013e-07,
+      "loss": 0.271,
+      "step": 3897
+    },
+    {
+      "epoch": 1.9626969124133584,
+      "grad_norm": 0.16872657786163298,
+      "learning_rate": 2.7392648983591547e-07,
+      "loss": 0.2622,
+      "step": 3898
+    },
+    {
+      "epoch": 1.9632010081915565,
+      "grad_norm": 0.17255810013786174,
+      "learning_rate": 2.736869838265368e-07,
+      "loss": 0.2772,
+      "step": 3899
+    },
+    {
+      "epoch": 1.9637051039697542,
+      "grad_norm": 0.18979660370274565,
+      "learning_rate": 2.7344754310706135e-07,
+      "loss": 0.2774,
+      "step": 3900
+    },
+    {
+      "epoch": 1.9642091997479523,
+      "grad_norm": 0.2005994736740754,
+      "learning_rate": 2.732081677465664e-07,
+      "loss": 0.2558,
+      "step": 3901
+    },
+    {
+      "epoch": 1.96471329552615,
+      "grad_norm": 0.16769942763481963,
+      "learning_rate": 2.7296885781410997e-07,
+      "loss": 0.2729,
+      "step": 3902
+    },
+    {
+      "epoch": 1.9652173913043478,
+      "grad_norm": 0.17196075848567396,
+      "learning_rate": 2.7272961337873184e-07,
+      "loss": 0.2597,
+      "step": 3903
+    },
+    {
+      "epoch": 1.9657214870825457,
+      "grad_norm": 0.17574737438377164,
+      "learning_rate": 2.724904345094522e-07,
+      "loss": 0.2787,
+      "step": 3904
+    },
+    {
+      "epoch": 1.9662255828607436,
+      "grad_norm": 0.1801545153423894,
+      "learning_rate": 2.7225132127527305e-07,
+      "loss": 0.2665,
+      "step": 3905
+    },
+    {
+      "epoch": 1.9667296786389414,
+      "grad_norm": 0.18767570216174043,
+      "learning_rate": 2.720122737451767e-07,
+      "loss": 0.2963,
+      "step": 3906
+    },
+    {
+      "epoch": 1.967233774417139,
+      "grad_norm": 0.17392592966729542,
+      "learning_rate": 2.717732919881273e-07,
+      "loss": 0.2664,
+      "step": 3907
+    },
+    {
+      "epoch": 1.9677378701953372,
+      "grad_norm": 0.18336201230753177,
+      "learning_rate": 2.715343760730693e-07,
+      "loss": 0.2722,
+      "step": 3908
+    },
+    {
+      "epoch": 1.9682419659735348,
+      "grad_norm": 0.17088548349889449,
+      "learning_rate": 2.7129552606892834e-07,
+      "loss": 0.2686,
+      "step": 3909
+    },
+    {
+      "epoch": 1.968746061751733,
+      "grad_norm": 0.17989017966945003,
+      "learning_rate": 2.710567420446116e-07,
+      "loss": 0.2661,
+      "step": 3910
+    },
+    {
+      "epoch": 1.9692501575299306,
+      "grad_norm": 0.17148613832259865,
+      "learning_rate": 2.708180240690063e-07,
+      "loss": 0.2682,
+      "step": 3911
+    },
+    {
+      "epoch": 1.9697542533081287,
+      "grad_norm": 0.16861185373135207,
+      "learning_rate": 2.705793722109816e-07,
+      "loss": 0.2704,
+      "step": 3912
+    },
+    {
+      "epoch": 1.9702583490863264,
+      "grad_norm": 0.19696241693330502,
+      "learning_rate": 2.7034078653938663e-07,
+      "loss": 0.2718,
+      "step": 3913
+    },
+    {
+      "epoch": 1.9707624448645242,
+      "grad_norm": 0.18050920428069256,
+      "learning_rate": 2.7010226712305227e-07,
+      "loss": 0.2849,
+      "step": 3914
+    },
+    {
+      "epoch": 1.971266540642722,
+      "grad_norm": 0.17461718966022174,
+      "learning_rate": 2.698638140307897e-07,
+      "loss": 0.2709,
+      "step": 3915
+    },
+    {
+      "epoch": 1.97177063642092,
+      "grad_norm": 0.16987029931279596,
+      "learning_rate": 2.6962542733139094e-07,
+      "loss": 0.2749,
+      "step": 3916
+    },
+    {
+      "epoch": 1.9722747321991179,
+      "grad_norm": 0.1780473567269595,
+      "learning_rate": 2.6938710709362953e-07,
+      "loss": 0.2801,
+      "step": 3917
+    },
+    {
+      "epoch": 1.9727788279773157,
+      "grad_norm": 0.18378204778035598,
+      "learning_rate": 2.691488533862589e-07,
+      "loss": 0.2737,
+      "step": 3918
+    },
+    {
+      "epoch": 1.9732829237555136,
+      "grad_norm": 0.17887945423278645,
+      "learning_rate": 2.689106662780143e-07,
+      "loss": 0.2733,
+      "step": 3919
+    },
+    {
+      "epoch": 1.9737870195337113,
+      "grad_norm": 0.16923234306379617,
+      "learning_rate": 2.686725458376109e-07,
+      "loss": 0.2718,
+      "step": 3920
+    },
+    {
+      "epoch": 1.9742911153119094,
+      "grad_norm": 0.17182101067457767,
+      "learning_rate": 2.684344921337449e-07,
+      "loss": 0.272,
+      "step": 3921
+    },
+    {
+      "epoch": 1.974795211090107,
+      "grad_norm": 0.1811448482903096,
+      "learning_rate": 2.681965052350935e-07,
+      "loss": 0.2757,
+      "step": 3922
+    },
+    {
+      "epoch": 1.9752993068683051,
+      "grad_norm": 0.17076536993145938,
+      "learning_rate": 2.6795858521031455e-07,
+      "loss": 0.272,
+      "step": 3923
+    },
+    {
+      "epoch": 1.9758034026465028,
+      "grad_norm": 0.17273807428844049,
+      "learning_rate": 2.6772073212804626e-07,
+      "loss": 0.2623,
+      "step": 3924
+    },
+    {
+      "epoch": 1.9763074984247007,
+      "grad_norm": 0.170028734836891,
+      "learning_rate": 2.674829460569077e-07,
+      "loss": 0.2783,
+      "step": 3925
+    },
+    {
+      "epoch": 1.9768115942028985,
+      "grad_norm": 0.16806227375301971,
+      "learning_rate": 2.67245227065499e-07,
+      "loss": 0.2556,
+      "step": 3926
+    },
+    {
+      "epoch": 1.9773156899810964,
+      "grad_norm": 0.17164028145315433,
+      "learning_rate": 2.6700757522240025e-07,
+      "loss": 0.2745,
+      "step": 3927
+    },
+    {
+      "epoch": 1.9778197857592943,
+      "grad_norm": 0.16995123735916823,
+      "learning_rate": 2.66769990596173e-07,
+      "loss": 0.2746,
+      "step": 3928
+    },
+    {
+      "epoch": 1.9783238815374922,
+      "grad_norm": 0.17598543808774827,
+      "learning_rate": 2.6653247325535843e-07,
+      "loss": 0.2622,
+      "step": 3929
+    },
+    {
+      "epoch": 1.97882797731569,
+      "grad_norm": 0.17359313560540118,
+      "learning_rate": 2.662950232684793e-07,
+      "loss": 0.2776,
+      "step": 3930
+    },
+    {
+      "epoch": 1.9793320730938877,
+      "grad_norm": 0.167656438622552,
+      "learning_rate": 2.6605764070403817e-07,
+      "loss": 0.262,
+      "step": 3931
+    },
+    {
+      "epoch": 1.9798361688720858,
+      "grad_norm": 0.17411531156307122,
+      "learning_rate": 2.6582032563051835e-07,
+      "loss": 0.2746,
+      "step": 3932
+    },
+    {
+      "epoch": 1.9803402646502835,
+      "grad_norm": 0.17074273937666753,
+      "learning_rate": 2.6558307811638415e-07,
+      "loss": 0.2745,
+      "step": 3933
+    },
+    {
+      "epoch": 1.9808443604284816,
+      "grad_norm": 0.17097062535567104,
+      "learning_rate": 2.653458982300795e-07,
+      "loss": 0.2762,
+      "step": 3934
+    },
+    {
+      "epoch": 1.9813484562066792,
+      "grad_norm": 0.1683002481107667,
+      "learning_rate": 2.6510878604002984e-07,
+      "loss": 0.28,
+      "step": 3935
+    },
+    {
+      "epoch": 1.981852551984877,
+      "grad_norm": 0.17106853535957925,
+      "learning_rate": 2.648717416146401e-07,
+      "loss": 0.2673,
+      "step": 3936
+    },
+    {
+      "epoch": 1.982356647763075,
+      "grad_norm": 0.18084388076626143,
+      "learning_rate": 2.6463476502229664e-07,
+      "loss": 0.2863,
+      "step": 3937
+    },
+    {
+      "epoch": 1.9828607435412728,
+      "grad_norm": 0.16736153224440417,
+      "learning_rate": 2.643978563313654e-07,
+      "loss": 0.2522,
+      "step": 3938
+    },
+    {
+      "epoch": 1.9833648393194707,
+      "grad_norm": 0.1708328607685619,
+      "learning_rate": 2.64161015610193e-07,
+      "loss": 0.2713,
+      "step": 3939
+    },
+    {
+      "epoch": 1.9838689350976686,
+      "grad_norm": 0.16701316011846237,
+      "learning_rate": 2.639242429271068e-07,
+      "loss": 0.2684,
+      "step": 3940
+    },
+    {
+      "epoch": 1.9843730308758665,
+      "grad_norm": 0.19205395390104565,
+      "learning_rate": 2.6368753835041384e-07,
+      "loss": 0.2652,
+      "step": 3941
+    },
+    {
+      "epoch": 1.9848771266540641,
+      "grad_norm": 0.1717816839539935,
+      "learning_rate": 2.634509019484025e-07,
+      "loss": 0.277,
+      "step": 3942
+    },
+    {
+      "epoch": 1.9853812224322622,
+      "grad_norm": 0.17022274259442283,
+      "learning_rate": 2.6321433378934035e-07,
+      "loss": 0.2609,
+      "step": 3943
+    },
+    {
+      "epoch": 1.9858853182104599,
+      "grad_norm": 0.17756035746010013,
+      "learning_rate": 2.629778339414763e-07,
+      "loss": 0.2712,
+      "step": 3944
+    },
+    {
+      "epoch": 1.986389413988658,
+      "grad_norm": 0.17227310821132044,
+      "learning_rate": 2.627414024730389e-07,
+      "loss": 0.2636,
+      "step": 3945
+    },
+    {
+      "epoch": 1.9868935097668556,
+      "grad_norm": 0.1759199181751579,
+      "learning_rate": 2.625050394522369e-07,
+      "loss": 0.2692,
+      "step": 3946
+    },
+    {
+      "epoch": 1.9873976055450535,
+      "grad_norm": 0.17477365439151,
+      "learning_rate": 2.6226874494726005e-07,
+      "loss": 0.2719,
+      "step": 3947
+    },
+    {
+      "epoch": 1.9879017013232514,
+      "grad_norm": 0.17546888078854525,
+      "learning_rate": 2.620325190262774e-07,
+      "loss": 0.2693,
+      "step": 3948
+    },
+    {
+      "epoch": 1.9884057971014493,
+      "grad_norm": 0.17339448401055704,
+      "learning_rate": 2.6179636175743905e-07,
+      "loss": 0.2735,
+      "step": 3949
+    },
+    {
+      "epoch": 1.9889098928796471,
+      "grad_norm": 0.1692734111192285,
+      "learning_rate": 2.615602732088748e-07,
+      "loss": 0.2712,
+      "step": 3950
+    },
+    {
+      "epoch": 1.989413988657845,
+      "grad_norm": 0.1779563761961173,
+      "learning_rate": 2.6132425344869446e-07,
+      "loss": 0.2596,
+      "step": 3951
+    },
+    {
+      "epoch": 1.989918084436043,
+      "grad_norm": 0.17689388931184377,
+      "learning_rate": 2.610883025449887e-07,
+      "loss": 0.2756,
+      "step": 3952
+    },
+    {
+      "epoch": 1.9904221802142406,
+      "grad_norm": 0.18739670056285473,
+      "learning_rate": 2.6085242056582764e-07,
+      "loss": 0.2713,
+      "step": 3953
+    },
+    {
+      "epoch": 1.9909262759924387,
+      "grad_norm": 0.17988497242704043,
+      "learning_rate": 2.6061660757926163e-07,
+      "loss": 0.272,
+      "step": 3954
+    },
+    {
+      "epoch": 1.9914303717706363,
+      "grad_norm": 0.1864336481558806,
+      "learning_rate": 2.6038086365332155e-07,
+      "loss": 0.2731,
+      "step": 3955
+    },
+    {
+      "epoch": 1.9919344675488344,
+      "grad_norm": 0.17348195377594694,
+      "learning_rate": 2.6014518885601803e-07,
+      "loss": 0.2747,
+      "step": 3956
+    },
+    {
+      "epoch": 1.992438563327032,
+      "grad_norm": 0.17263796785386343,
+      "learning_rate": 2.599095832553415e-07,
+      "loss": 0.2837,
+      "step": 3957
+    },
+    {
+      "epoch": 1.99294265910523,
+      "grad_norm": 0.17274927381019026,
+      "learning_rate": 2.596740469192631e-07,
+      "loss": 0.2724,
+      "step": 3958
+    },
+    {
+      "epoch": 1.9934467548834278,
+      "grad_norm": 0.18942508993924725,
+      "learning_rate": 2.594385799157333e-07,
+      "loss": 0.2704,
+      "step": 3959
+    },
+    {
+      "epoch": 1.9939508506616257,
+      "grad_norm": 0.16909085704359703,
+      "learning_rate": 2.59203182312683e-07,
+      "loss": 0.2658,
+      "step": 3960
+    },
+    {
+      "epoch": 1.9944549464398236,
+      "grad_norm": 0.17982713814067064,
+      "learning_rate": 2.5896785417802313e-07,
+      "loss": 0.3008,
+      "step": 3961
+    },
+    {
+      "epoch": 1.9949590422180214,
+      "grad_norm": 0.17137213119070738,
+      "learning_rate": 2.5873259557964395e-07,
+      "loss": 0.2798,
+      "step": 3962
+    },
+    {
+      "epoch": 1.9954631379962193,
+      "grad_norm": 0.1794592545298311,
+      "learning_rate": 2.584974065854165e-07,
+      "loss": 0.2686,
+      "step": 3963
+    },
+    {
+      "epoch": 1.995967233774417,
+      "grad_norm": 0.17623111078697679,
+      "learning_rate": 2.5826228726319116e-07,
+      "loss": 0.2792,
+      "step": 3964
+    },
+    {
+      "epoch": 1.996471329552615,
+      "grad_norm": 0.1704272750332749,
+      "learning_rate": 2.5802723768079857e-07,
+      "loss": 0.2706,
+      "step": 3965
+    },
+    {
+      "epoch": 1.9969754253308127,
+      "grad_norm": 0.17123275269483276,
+      "learning_rate": 2.5779225790604887e-07,
+      "loss": 0.28,
+      "step": 3966
+    },
+    {
+      "epoch": 1.9974795211090108,
+      "grad_norm": 0.170434187962785,
+      "learning_rate": 2.5755734800673243e-07,
+      "loss": 0.2765,
+      "step": 3967
+    },
+    {
+      "epoch": 1.9979836168872085,
+      "grad_norm": 0.1680976681438808,
+      "learning_rate": 2.573225080506193e-07,
+      "loss": 0.2617,
+      "step": 3968
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5949,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 992,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6240331729207296.0,
+  "train_batch_size": 6,
+  "trial_name": null,
+  "trial_params": null
+}