diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21896 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.999839871897518,
+  "eval_steps": 500,
+  "global_step": 3122,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003202562049639712,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.597444089456869e-07,
+      "loss": 2.5505,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006405124099279424,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.194888178913738e-07,
+      "loss": 2.302,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009607686148919135,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.792332268370607e-07,
+      "loss": 2.0733,
+      "step": 3
+    },
+    {
+      "epoch": 0.0012810248198558848,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.389776357827476e-07,
+      "loss": 2.5665,
+      "step": 4
+    },
+    {
+      "epoch": 0.0016012810248198558,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.987220447284345e-07,
+      "loss": 2.2262,
+      "step": 5
+    },
+    {
+      "epoch": 0.001921537229783827,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.584664536741213e-07,
+      "loss": 2.7195,
+      "step": 6
+    },
+    {
+      "epoch": 0.0022417934347477983,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.1182108626198083e-06,
+      "loss": 2.3025,
+      "step": 7
+    },
+    {
+      "epoch": 0.0025620496397117695,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.2779552715654952e-06,
+      "loss": 1.7873,
+      "step": 8
+    },
+    {
+      "epoch": 0.0028823058446757407,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4376996805111822e-06,
+      "loss": 2.1716,
+      "step": 9
+    },
+    {
+      "epoch": 0.0032025620496397116,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.597444089456869e-06,
+      "loss": 2.1507,
+      "step": 10
+    },
+    {
+      "epoch": 0.003522818254603683,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.7571884984025559e-06,
+      "loss": 2.2224,
+      "step": 11
+    },
+    {
+      "epoch": 0.003843074459567654,
+      "grad_norm": 0.5,
+      "learning_rate": 1.9169329073482426e-06,
+      "loss": 2.3008,
+      "step": 12
+    },
+    {
+      "epoch": 0.004163330664531626,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.0766773162939296e-06,
+      "loss": 2.3001,
+      "step": 13
+    },
+    {
+      "epoch": 0.0044835868694955965,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.2364217252396165e-06,
+      "loss": 2.1583,
+      "step": 14
+    },
+    {
+      "epoch": 0.004803843074459567,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.3961661341853035e-06,
+      "loss": 2.2752,
+      "step": 15
+    },
+    {
+      "epoch": 0.005124099279423539,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.5559105431309904e-06,
+      "loss": 1.6753,
+      "step": 16
+    },
+    {
+      "epoch": 0.00544435548438751,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.7156549520766774e-06,
+      "loss": 2.4427,
+      "step": 17
+    },
+    {
+      "epoch": 0.0057646116893514815,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.8753993610223644e-06,
+      "loss": 2.3853,
+      "step": 18
+    },
+    {
+      "epoch": 0.006084867894315452,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.0351437699680513e-06,
+      "loss": 2.4479,
+      "step": 19
+    },
+    {
+      "epoch": 0.006405124099279423,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.194888178913738e-06,
+      "loss": 1.9668,
+      "step": 20
+    },
+    {
+      "epoch": 0.006725380304243395,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.354632587859425e-06,
+      "loss": 2.2987,
+      "step": 21
+    },
+    {
+      "epoch": 0.007045636509207366,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.5143769968051118e-06,
+      "loss": 1.7833,
+      "step": 22
+    },
+    {
+      "epoch": 0.007365892714171337,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.6741214057507987e-06,
+      "loss": 2.3454,
+      "step": 23
+    },
+    {
+      "epoch": 0.007686148919135308,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.833865814696485e-06,
+      "loss": 2.1669,
+      "step": 24
+    },
+    {
+      "epoch": 0.008006405124099279,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.993610223642173e-06,
+      "loss": 2.2589,
+      "step": 25
+    },
+    {
+      "epoch": 0.008326661329063251,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.153354632587859e-06,
+      "loss": 2.3922,
+      "step": 26
+    },
+    {
+      "epoch": 0.008646917534027222,
+      "grad_norm": 0.5,
+      "learning_rate": 4.3130990415335465e-06,
+      "loss": 2.8444,
+      "step": 27
+    },
+    {
+      "epoch": 0.008967173738991193,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.472843450479233e-06,
+      "loss": 1.6251,
+      "step": 28
+    },
+    {
+      "epoch": 0.009287429943955164,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.6325878594249205e-06,
+      "loss": 1.4672,
+      "step": 29
+    },
+    {
+      "epoch": 0.009607686148919135,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.792332268370607e-06,
+      "loss": 2.2324,
+      "step": 30
+    },
+    {
+      "epoch": 0.009927942353883107,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.952076677316294e-06,
+      "loss": 1.8752,
+      "step": 31
+    },
+    {
+      "epoch": 0.010248198558847078,
+      "grad_norm": 1.921875,
+      "learning_rate": 5.111821086261981e-06,
+      "loss": 2.4251,
+      "step": 32
+    },
+    {
+      "epoch": 0.010568454763811049,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.2715654952076674e-06,
+      "loss": 2.2702,
+      "step": 33
+    },
+    {
+      "epoch": 0.01088871096877502,
+      "grad_norm": 5.3125,
+      "learning_rate": 5.431309904153355e-06,
+      "loss": 2.2467,
+      "step": 34
+    },
+    {
+      "epoch": 0.01120896717373899,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.591054313099041e-06,
+      "loss": 1.8738,
+      "step": 35
+    },
+    {
+      "epoch": 0.011529223378702963,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.750798722044729e-06,
+      "loss": 2.0557,
+      "step": 36
+    },
+    {
+      "epoch": 0.011849479583666934,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.910543130990415e-06,
+      "loss": 2.236,
+      "step": 37
+    },
+    {
+      "epoch": 0.012169735788630905,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.070287539936103e-06,
+      "loss": 1.9037,
+      "step": 38
+    },
+    {
+      "epoch": 0.012489991993594875,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.230031948881789e-06,
+      "loss": 1.9012,
+      "step": 39
+    },
+    {
+      "epoch": 0.012810248198558846,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.389776357827476e-06,
+      "loss": 1.4725,
+      "step": 40
+    },
+    {
+      "epoch": 0.013130504403522819,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.549520766773164e-06,
+      "loss": 2.4334,
+      "step": 41
+    },
+    {
+      "epoch": 0.01345076060848679,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.70926517571885e-06,
+      "loss": 2.2842,
+      "step": 42
+    },
+    {
+      "epoch": 0.01377101681345076,
+      "grad_norm": 0.40625,
+      "learning_rate": 6.869009584664538e-06,
+      "loss": 1.993,
+      "step": 43
+    },
+    {
+      "epoch": 0.014091273018414731,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.0287539936102235e-06,
+      "loss": 1.7918,
+      "step": 44
+    },
+    {
+      "epoch": 0.014411529223378704,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.188498402555911e-06,
+      "loss": 2.573,
+      "step": 45
+    },
+    {
+      "epoch": 0.014731785428342675,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.3482428115015974e-06,
+      "loss": 2.0233,
+      "step": 46
+    },
+    {
+      "epoch": 0.015052041633306645,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.507987220447285e-06,
+      "loss": 2.1536,
+      "step": 47
+    },
+    {
+      "epoch": 0.015372297838270616,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.66773162939297e-06,
+      "loss": 2.4289,
+      "step": 48
+    },
+    {
+      "epoch": 0.015692554043234587,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.82747603833866e-06,
+      "loss": 1.7267,
+      "step": 49
+    },
+    {
+      "epoch": 0.016012810248198558,
+      "grad_norm": 0.384765625,
+      "learning_rate": 7.987220447284345e-06,
+      "loss": 1.8542,
+      "step": 50
+    },
+    {
+      "epoch": 0.01633306645316253,
+      "grad_norm": 0.494140625,
+      "learning_rate": 8.146964856230033e-06,
+      "loss": 1.6736,
+      "step": 51
+    },
+    {
+      "epoch": 0.016653322658126503,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.306709265175718e-06,
+      "loss": 2.1386,
+      "step": 52
+    },
+    {
+      "epoch": 0.016973578863090474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.466453674121406e-06,
+      "loss": 2.3592,
+      "step": 53
+    },
+    {
+      "epoch": 0.017293835068054444,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.626198083067093e-06,
+      "loss": 1.732,
+      "step": 54
+    },
+    {
+      "epoch": 0.017614091273018415,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.78594249201278e-06,
+      "loss": 1.5725,
+      "step": 55
+    },
+    {
+      "epoch": 0.017934347477982386,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.945686900958466e-06,
+      "loss": 2.4561,
+      "step": 56
+    },
+    {
+      "epoch": 0.018254603682946357,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.105431309904154e-06,
+      "loss": 1.8126,
+      "step": 57
+    },
+    {
+      "epoch": 0.018574859887910328,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.265175718849841e-06,
+      "loss": 2.1243,
+      "step": 58
+    },
+    {
+      "epoch": 0.0188951160928743,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.424920127795528e-06,
+      "loss": 2.3236,
+      "step": 59
+    },
+    {
+      "epoch": 0.01921537229783827,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.584664536741214e-06,
+      "loss": 2.3305,
+      "step": 60
+    },
+    {
+      "epoch": 0.01953562850280224,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.744408945686901e-06,
+      "loss": 2.1348,
+      "step": 61
+    },
+    {
+      "epoch": 0.019855884707766214,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.904153354632589e-06,
+      "loss": 1.91,
+      "step": 62
+    },
+    {
+      "epoch": 0.020176140912730185,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0063897763578276e-05,
+      "loss": 2.4788,
+      "step": 63
+    },
+    {
+      "epoch": 0.020496397117694156,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0223642172523962e-05,
+      "loss": 2.4574,
+      "step": 64
+    },
+    {
+      "epoch": 0.020816653322658127,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.038338658146965e-05,
+      "loss": 2.3022,
+      "step": 65
+    },
+    {
+      "epoch": 0.021136909527622098,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.0543130990415335e-05,
+      "loss": 1.4106,
+      "step": 66
+    },
+    {
+      "epoch": 0.02145716573258607,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0702875399361024e-05,
+      "loss": 2.4533,
+      "step": 67
+    },
+    {
+      "epoch": 0.02177742193755004,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.086261980830671e-05,
+      "loss": 2.6732,
+      "step": 68
+    },
+    {
+      "epoch": 0.02209767814251401,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1022364217252397e-05,
+      "loss": 2.09,
+      "step": 69
+    },
+    {
+      "epoch": 0.02241793434747798,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.1182108626198083e-05,
+      "loss": 2.4639,
+      "step": 70
+    },
+    {
+      "epoch": 0.022738190552441955,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.134185303514377e-05,
+      "loss": 2.195,
+      "step": 71
+    },
+    {
+      "epoch": 0.023058446757405926,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1501597444089457e-05,
+      "loss": 2.0553,
+      "step": 72
+    },
+    {
+      "epoch": 0.023378702962369897,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.1661341853035145e-05,
+      "loss": 1.6604,
+      "step": 73
+    },
+    {
+      "epoch": 0.023698959167333868,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.182108626198083e-05,
+      "loss": 2.1982,
+      "step": 74
+    },
+    {
+      "epoch": 0.02401921537229784,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.1980830670926518e-05,
+      "loss": 2.2353,
+      "step": 75
+    },
+    {
+      "epoch": 0.02433947157726181,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.2140575079872205e-05,
+      "loss": 2.0923,
+      "step": 76
+    },
+    {
+      "epoch": 0.02465972778222578,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.2300319488817893e-05,
+      "loss": 2.379,
+      "step": 77
+    },
+    {
+      "epoch": 0.02497998398718975,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.2460063897763578e-05,
+      "loss": 1.7649,
+      "step": 78
+    },
+    {
+      "epoch": 0.02530024019215372,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2619808306709266e-05,
+      "loss": 2.475,
+      "step": 79
+    },
+    {
+      "epoch": 0.025620496397117692,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.2779552715654951e-05,
+      "loss": 2.4901,
+      "step": 80
+    },
+    {
+      "epoch": 0.025940752602081667,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2939297124600639e-05,
+      "loss": 2.4741,
+      "step": 81
+    },
+    {
+      "epoch": 0.026261008807045638,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3099041533546328e-05,
+      "loss": 2.1992,
+      "step": 82
+    },
+    {
+      "epoch": 0.02658126501200961,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.3258785942492014e-05,
+      "loss": 2.6322,
+      "step": 83
+    },
+    {
+      "epoch": 0.02690152121697358,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.34185303514377e-05,
+      "loss": 2.2932,
+      "step": 84
+    },
+    {
+      "epoch": 0.02722177742193755,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.3578274760383387e-05,
+      "loss": 1.828,
+      "step": 85
+    },
+    {
+      "epoch": 0.02754203362690152,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.3738019169329076e-05,
+      "loss": 2.0083,
+      "step": 86
+    },
+    {
+      "epoch": 0.02786228983186549,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3897763578274761e-05,
+      "loss": 2.0055,
+      "step": 87
+    },
+    {
+      "epoch": 0.028182546036829462,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.4057507987220447e-05,
+      "loss": 2.3489,
+      "step": 88
+    },
+    {
+      "epoch": 0.028502802241793433,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.4217252396166134e-05,
+      "loss": 2.0319,
+      "step": 89
+    },
+    {
+      "epoch": 0.028823058446757407,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4376996805111822e-05,
+      "loss": 1.8418,
+      "step": 90
+    },
+    {
+      "epoch": 0.02914331465172138,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.453674121405751e-05,
+      "loss": 2.1933,
+      "step": 91
+    },
+    {
+      "epoch": 0.02946357085668535,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.4696485623003195e-05,
+      "loss": 2.238,
+      "step": 92
+    },
+    {
+      "epoch": 0.02978382706164932,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.485623003194888e-05,
+      "loss": 1.8154,
+      "step": 93
+    },
+    {
+      "epoch": 0.03010408326661329,
+      "grad_norm": 0.375,
+      "learning_rate": 1.501597444089457e-05,
+      "loss": 2.0664,
+      "step": 94
+    },
+    {
+      "epoch": 0.03042433947157726,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5175718849840257e-05,
+      "loss": 2.4674,
+      "step": 95
+    },
+    {
+      "epoch": 0.030744595676541232,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.533546325878594e-05,
+      "loss": 2.5225,
+      "step": 96
+    },
+    {
+      "epoch": 0.031064851881505203,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.549520766773163e-05,
+      "loss": 2.5732,
+      "step": 97
+    },
+    {
+      "epoch": 0.031385108086469174,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.565495207667732e-05,
+      "loss": 2.4263,
+      "step": 98
+    },
+    {
+      "epoch": 0.03170536429143315,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.5814696485623005e-05,
+      "loss": 1.8861,
+      "step": 99
+    },
+    {
+      "epoch": 0.032025620496397116,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.597444089456869e-05,
+      "loss": 2.187,
+      "step": 100
+    },
+    {
+      "epoch": 0.03234587670136109,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6134185303514376e-05,
+      "loss": 2.4841,
+      "step": 101
+    },
+    {
+      "epoch": 0.03266613290632506,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.6293929712460065e-05,
+      "loss": 2.1184,
+      "step": 102
+    },
+    {
+      "epoch": 0.03298638911128903,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.645367412140575e-05,
+      "loss": 2.5642,
+      "step": 103
+    },
+    {
+      "epoch": 0.033306645316253006,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6613418530351437e-05,
+      "loss": 2.2907,
+      "step": 104
+    },
+    {
+      "epoch": 0.03362690152121697,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6773162939297126e-05,
+      "loss": 2.4367,
+      "step": 105
+    },
+    {
+      "epoch": 0.03394715772618095,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.693290734824281e-05,
+      "loss": 2.5302,
+      "step": 106
+    },
+    {
+      "epoch": 0.034267413931144915,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.70926517571885e-05,
+      "loss": 2.1865,
+      "step": 107
+    },
+    {
+      "epoch": 0.03458767013610889,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.7252396166134186e-05,
+      "loss": 2.1695,
+      "step": 108
+    },
+    {
+      "epoch": 0.034907926341072856,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7412140575079875e-05,
+      "loss": 2.3653,
+      "step": 109
+    },
+    {
+      "epoch": 0.03522818254603683,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.757188498402556e-05,
+      "loss": 1.9492,
+      "step": 110
+    },
+    {
+      "epoch": 0.0355484387510008,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.7731629392971247e-05,
+      "loss": 2.7231,
+      "step": 111
+    },
+    {
+      "epoch": 0.03586869495596477,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7891373801916932e-05,
+      "loss": 2.2126,
+      "step": 112
+    },
+    {
+      "epoch": 0.03618895116092874,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.805111821086262e-05,
+      "loss": 2.4667,
+      "step": 113
+    },
+    {
+      "epoch": 0.036509207365892714,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8210862619808307e-05,
+      "loss": 2.3994,
+      "step": 114
+    },
+    {
+      "epoch": 0.03682946357085669,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8370607028753993e-05,
+      "loss": 2.2501,
+      "step": 115
+    },
+    {
+      "epoch": 0.037149719775820655,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.8530351437699682e-05,
+      "loss": 2.5277,
+      "step": 116
+    },
+    {
+      "epoch": 0.03746997598078463,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.869009584664537e-05,
+      "loss": 2.1283,
+      "step": 117
+    },
+    {
+      "epoch": 0.0377902321857486,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8849840255591057e-05,
+      "loss": 2.5671,
+      "step": 118
+    },
+    {
+      "epoch": 0.03811048839071257,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9009584664536742e-05,
+      "loss": 1.7323,
+      "step": 119
+    },
+    {
+      "epoch": 0.03843074459567654,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9169329073482428e-05,
+      "loss": 2.4496,
+      "step": 120
+    },
+    {
+      "epoch": 0.03875100080064051,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9329073482428117e-05,
+      "loss": 2.3247,
+      "step": 121
+    },
+    {
+      "epoch": 0.03907125700560448,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9488817891373803e-05,
+      "loss": 2.044,
+      "step": 122
+    },
+    {
+      "epoch": 0.039391513210568455,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.964856230031949e-05,
+      "loss": 2.2774,
+      "step": 123
+    },
+    {
+      "epoch": 0.03971176941553243,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9808306709265177e-05,
+      "loss": 2.6735,
+      "step": 124
+    },
+    {
+      "epoch": 0.040032025620496396,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.9968051118210863e-05,
+      "loss": 2.5676,
+      "step": 125
+    },
+    {
+      "epoch": 0.04035228182546037,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.0127795527156552e-05,
+      "loss": 2.132,
+      "step": 126
+    },
+    {
+      "epoch": 0.04067253803042434,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.0287539936102238e-05,
+      "loss": 2.6514,
+      "step": 127
+    },
+    {
+      "epoch": 0.04099279423538831,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.0447284345047924e-05,
+      "loss": 2.6641,
+      "step": 128
+    },
+    {
+      "epoch": 0.04131305044035228,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.0607028753993613e-05,
+      "loss": 2.442,
+      "step": 129
+    },
+    {
+      "epoch": 0.041633306645316254,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.07667731629393e-05,
+      "loss": 1.4356,
+      "step": 130
+    },
+    {
+      "epoch": 0.04195356285028022,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.0926517571884984e-05,
+      "loss": 1.8376,
+      "step": 131
+    },
+    {
+      "epoch": 0.042273819055244195,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.108626198083067e-05,
+      "loss": 2.4422,
+      "step": 132
+    },
+    {
+      "epoch": 0.04259407526020817,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.124600638977636e-05,
+      "loss": 2.1663,
+      "step": 133
+    },
+    {
+      "epoch": 0.04291433146517214,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.1405750798722048e-05,
+      "loss": 2.5161,
+      "step": 134
+    },
+    {
+      "epoch": 0.04323458767013611,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.1565495207667734e-05,
+      "loss": 2.295,
+      "step": 135
+    },
+    {
+      "epoch": 0.04355484387510008,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.172523961661342e-05,
+      "loss": 2.6238,
+      "step": 136
+    },
+    {
+      "epoch": 0.04387510008006405,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.188498402555911e-05,
+      "loss": 2.0364,
+      "step": 137
+    },
+    {
+      "epoch": 0.04419535628502802,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.2044728434504794e-05,
+      "loss": 2.7412,
+      "step": 138
+    },
+    {
+      "epoch": 0.044515612489991994,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.220447284345048e-05,
+      "loss": 1.8156,
+      "step": 139
+    },
+    {
+      "epoch": 0.04483586869495596,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.2364217252396165e-05,
+      "loss": 2.3815,
+      "step": 140
+    },
+    {
+      "epoch": 0.045156124899919936,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.2523961661341854e-05,
+      "loss": 2.6271,
+      "step": 141
+    },
+    {
+      "epoch": 0.04547638110488391,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.268370607028754e-05,
+      "loss": 1.6873,
+      "step": 142
+    },
+    {
+      "epoch": 0.04579663730984788,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.284345047923323e-05,
+      "loss": 2.3417,
+      "step": 143
+    },
+    {
+      "epoch": 0.04611689351481185,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.3003194888178915e-05,
+      "loss": 2.1776,
+      "step": 144
+    },
+    {
+      "epoch": 0.04643714971977582,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.3162939297124604e-05,
+      "loss": 2.2616,
+      "step": 145
+    },
+    {
+      "epoch": 0.046757405924739794,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.332268370607029e-05,
+      "loss": 2.5723,
+      "step": 146
+    },
+    {
+      "epoch": 0.04707766212970376,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.3482428115015975e-05,
+      "loss": 2.4548,
+      "step": 147
+    },
+    {
+      "epoch": 0.047397918334667735,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.364217252396166e-05,
+      "loss": 2.4736,
+      "step": 148
+    },
+    {
+      "epoch": 0.0477181745396317,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.380191693290735e-05,
+      "loss": 2.386,
+      "step": 149
+    },
+    {
+      "epoch": 0.04803843074459568,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.3961661341853036e-05,
+      "loss": 2.644,
+      "step": 150
+    },
+    {
+      "epoch": 0.04835868694955965,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.412140575079872e-05,
+      "loss": 2.1945,
+      "step": 151
+    },
+    {
+      "epoch": 0.04867894315452362,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.428115015974441e-05,
+      "loss": 2.2383,
+      "step": 152
+    },
+    {
+      "epoch": 0.04899919935948759,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.44408945686901e-05,
+      "loss": 2.5643,
+      "step": 153
+    },
+    {
+      "epoch": 0.04931945556445156,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.4600638977635785e-05,
+      "loss": 1.6623,
+      "step": 154
+    },
+    {
+      "epoch": 0.049639711769415534,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.476038338658147e-05,
+      "loss": 2.1212,
+      "step": 155
+    },
+    {
+      "epoch": 0.0499599679743795,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.4920127795527157e-05,
+      "loss": 2.0499,
+      "step": 156
+    },
+    {
+      "epoch": 0.050280224179343476,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.5079872204472842e-05,
+      "loss": 2.004,
+      "step": 157
+    },
+    {
+      "epoch": 0.05060048038430744,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.523961661341853e-05,
+      "loss": 2.4388,
+      "step": 158
+    },
+    {
+      "epoch": 0.05092073658927142,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.539936102236422e-05,
+      "loss": 2.2752,
+      "step": 159
+    },
+    {
+      "epoch": 0.051240992794235385,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.5559105431309903e-05,
+      "loss": 2.1717,
+      "step": 160
+    },
+    {
+      "epoch": 0.05156124899919936,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.5718849840255592e-05,
+      "loss": 2.3856,
+      "step": 161
+    },
+    {
+      "epoch": 0.05188150520416333,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.5878594249201278e-05,
+      "loss": 1.7901,
+      "step": 162
+    },
+    {
+      "epoch": 0.0522017614091273,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.6038338658146967e-05,
+      "loss": 2.1485,
+      "step": 163
+    },
+    {
+      "epoch": 0.052522017614091275,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.6198083067092656e-05,
+      "loss": 2.4288,
+      "step": 164
+    },
+    {
+      "epoch": 0.05284227381905524,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.6357827476038338e-05,
+      "loss": 2.4144,
+      "step": 165
+    },
+    {
+      "epoch": 0.05316253002401922,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.6517571884984027e-05,
+      "loss": 2.4021,
+      "step": 166
+    },
+    {
+      "epoch": 0.053482786228983184,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.6677316293929716e-05,
+      "loss": 2.2037,
+      "step": 167
+    },
+    {
+      "epoch": 0.05380304243394716,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.68370607028754e-05,
+      "loss": 2.33,
+      "step": 168
+    },
+    {
+      "epoch": 0.054123298638911126,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.6996805111821088e-05,
+      "loss": 2.3497,
+      "step": 169
+    },
+    {
+      "epoch": 0.0544435548438751,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.7156549520766773e-05,
+      "loss": 2.0505,
+      "step": 170
+    },
+    {
+      "epoch": 0.054763811048839074,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.7316293929712462e-05,
+      "loss": 2.5688,
+      "step": 171
+    },
+    {
+      "epoch": 0.05508406725380304,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.747603833865815e-05,
+      "loss": 2.5961,
+      "step": 172
+    },
+    {
+      "epoch": 0.055404323458767016,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.7635782747603834e-05,
+      "loss": 1.9069,
+      "step": 173
+    },
+    {
+      "epoch": 0.05572457966373098,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.7795527156549523e-05,
+      "loss": 2.3172,
+      "step": 174
+    },
+    {
+      "epoch": 0.05604483586869496,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.7955271565495212e-05,
+      "loss": 2.5545,
+      "step": 175
+    },
+    {
+      "epoch": 0.056365092073658925,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.8115015974440894e-05,
+      "loss": 2.248,
+      "step": 176
+    },
+    {
+      "epoch": 0.0566853482786229,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.8274760383386583e-05,
+      "loss": 2.4285,
+      "step": 177
+    },
+    {
+      "epoch": 0.057005604483586866,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.843450479233227e-05,
+      "loss": 1.4118,
+      "step": 178
+    },
+    {
+      "epoch": 0.05732586068855084,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.8594249201277955e-05,
+      "loss": 1.8937,
+      "step": 179
+    },
+    {
+      "epoch": 0.057646116893514815,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.8753993610223644e-05,
+      "loss": 1.7478,
+      "step": 180
+    },
+    {
+      "epoch": 0.05796637309847878,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.891373801916933e-05,
+      "loss": 1.8904,
+      "step": 181
+    },
+    {
+      "epoch": 0.05828662930344276,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.907348242811502e-05,
+      "loss": 2.6784,
+      "step": 182
+    },
+    {
+      "epoch": 0.058606885508406724,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.9233226837060707e-05,
+      "loss": 2.4389,
+      "step": 183
+    },
+    {
+      "epoch": 0.0589271417133707,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.939297124600639e-05,
+      "loss": 2.5659,
+      "step": 184
+    },
+    {
+      "epoch": 0.059247397918334666,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.955271565495208e-05,
+      "loss": 2.0976,
+      "step": 185
+    },
+    {
+      "epoch": 0.05956765412329864,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.971246006389776e-05,
+      "loss": 2.5468,
+      "step": 186
+    },
+    {
+      "epoch": 0.05988791032826261,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.987220447284345e-05,
+      "loss": 2.3513,
+      "step": 187
+    },
+    {
+      "epoch": 0.06020816653322658,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.003194888178914e-05,
+      "loss": 1.9096,
+      "step": 188
+    },
+    {
+      "epoch": 0.060528422738190556,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.0191693290734825e-05,
+      "loss": 1.768,
+      "step": 189
+    },
+    {
+      "epoch": 0.06084867894315452,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.0351437699680514e-05,
+      "loss": 1.7802,
+      "step": 190
+    },
+    {
+      "epoch": 0.0611689351481185,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.0511182108626203e-05,
+      "loss": 1.9335,
+      "step": 191
+    },
+    {
+      "epoch": 0.061489191353082465,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.067092651757188e-05,
+      "loss": 2.3403,
+      "step": 192
+    },
+    {
+      "epoch": 0.06180944755804644,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.083067092651757e-05,
+      "loss": 2.1596,
+      "step": 193
+    },
+    {
+      "epoch": 0.062129703763010406,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.099041533546326e-05,
+      "loss": 2.6126,
+      "step": 194
+    },
+    {
+      "epoch": 0.06244995996797438,
+      "grad_norm": 0.322265625,
+      "learning_rate": 3.115015974440895e-05,
+      "loss": 1.772,
+      "step": 195
+    },
+    {
+      "epoch": 0.06277021617293835,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.130990415335464e-05,
+      "loss": 2.694,
+      "step": 196
+    },
+    {
+      "epoch": 0.06309047237790232,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.146964856230032e-05,
+      "loss": 2.045,
+      "step": 197
+    },
+    {
+      "epoch": 0.0634107285828663,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.162939297124601e-05,
+      "loss": 1.7398,
+      "step": 198
+    },
+    {
+      "epoch": 0.06373098478783026,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.17891373801917e-05,
+      "loss": 2.344,
+      "step": 199
+    },
+    {
+      "epoch": 0.06405124099279423,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.194888178913738e-05,
+      "loss": 2.1837,
+      "step": 200
+    },
+    {
+      "epoch": 0.06437149719775821,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.210862619808307e-05,
+      "loss": 2.0782,
+      "step": 201
+    },
+    {
+      "epoch": 0.06469175340272218,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.226837060702875e-05,
+      "loss": 2.2655,
+      "step": 202
+    },
+    {
+      "epoch": 0.06501200960768615,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.242811501597444e-05,
+      "loss": 2.55,
+      "step": 203
+    },
+    {
+      "epoch": 0.06533226581265011,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.258785942492013e-05,
+      "loss": 2.2424,
+      "step": 204
+    },
+    {
+      "epoch": 0.0656525220176141,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.274760383386581e-05,
+      "loss": 2.5929,
+      "step": 205
+    },
+    {
+      "epoch": 0.06597277822257806,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.29073482428115e-05,
+      "loss": 2.2174,
+      "step": 206
+    },
+    {
+      "epoch": 0.06629303442754203,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.306709265175719e-05,
+      "loss": 2.685,
+      "step": 207
+    },
+    {
+      "epoch": 0.06661329063250601,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.322683706070287e-05,
+      "loss": 2.242,
+      "step": 208
+    },
+    {
+      "epoch": 0.06693354683746998,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.338658146964856e-05,
+      "loss": 1.9813,
+      "step": 209
+    },
+    {
+      "epoch": 0.06725380304243395,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.354632587859425e-05,
+      "loss": 2.0788,
+      "step": 210
+    },
+    {
+      "epoch": 0.06757405924739791,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.3706070287539934e-05,
+      "loss": 2.2187,
+      "step": 211
+    },
+    {
+      "epoch": 0.0678943154523619,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.386581469648562e-05,
+      "loss": 2.2772,
+      "step": 212
+    },
+    {
+      "epoch": 0.06821457165732586,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.402555910543131e-05,
+      "loss": 1.4853,
+      "step": 213
+    },
+    {
+      "epoch": 0.06853482786228983,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.4185303514377e-05,
+      "loss": 2.0633,
+      "step": 214
+    },
+    {
+      "epoch": 0.0688550840672538,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.434504792332269e-05,
+      "loss": 2.3552,
+      "step": 215
+    },
+    {
+      "epoch": 0.06917534027221778,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.450479233226837e-05,
+      "loss": 2.7011,
+      "step": 216
+    },
+    {
+      "epoch": 0.06949559647718175,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.466453674121406e-05,
+      "loss": 2.0144,
+      "step": 217
+    },
+    {
+      "epoch": 0.06981585268214571,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.482428115015975e-05,
+      "loss": 2.293,
+      "step": 218
+    },
+    {
+      "epoch": 0.0701361088871097,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.498402555910543e-05,
+      "loss": 2.2046,
+      "step": 219
+    },
+    {
+      "epoch": 0.07045636509207366,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.514376996805112e-05,
+      "loss": 2.0654,
+      "step": 220
+    },
+    {
+      "epoch": 0.07077662129703763,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.5303514376996804e-05,
+      "loss": 2.3501,
+      "step": 221
+    },
+    {
+      "epoch": 0.0710968775020016,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.546325878594249e-05,
+      "loss": 2.3258,
+      "step": 222
+    },
+    {
+      "epoch": 0.07141713370696558,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.562300319488818e-05,
+      "loss": 2.3782,
+      "step": 223
+    },
+    {
+      "epoch": 0.07173738991192954,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.5782747603833865e-05,
+      "loss": 2.2688,
+      "step": 224
+    },
+    {
+      "epoch": 0.07205764611689351,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.5942492012779554e-05,
+      "loss": 2.2505,
+      "step": 225
+    },
+    {
+      "epoch": 0.07237790232185748,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.610223642172524e-05,
+      "loss": 2.4121,
+      "step": 226
+    },
+    {
+      "epoch": 0.07269815852682146,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.6261980830670925e-05,
+      "loss": 2.4804,
+      "step": 227
+    },
+    {
+      "epoch": 0.07301841473178543,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.6421725239616614e-05,
+      "loss": 2.4248,
+      "step": 228
+    },
+    {
+      "epoch": 0.0733386709367494,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.65814696485623e-05,
+      "loss": 2.2826,
+      "step": 229
+    },
+    {
+      "epoch": 0.07365892714171338,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.6741214057507985e-05,
+      "loss": 1.8805,
+      "step": 230
+    },
+    {
+      "epoch": 0.07397918334667734,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.6900958466453675e-05,
+      "loss": 1.9896,
+      "step": 231
+    },
+    {
+      "epoch": 0.07429943955164131,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.7060702875399364e-05,
+      "loss": 2.6494,
+      "step": 232
+    },
+    {
+      "epoch": 0.07461969575660528,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.722044728434505e-05,
+      "loss": 1.5742,
+      "step": 233
+    },
+    {
+      "epoch": 0.07493995196156926,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.738019169329074e-05,
+      "loss": 1.9401,
+      "step": 234
+    },
+    {
+      "epoch": 0.07526020816653323,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.7539936102236424e-05,
+      "loss": 2.5195,
+      "step": 235
+    },
+    {
+      "epoch": 0.0755804643714972,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.769968051118211e-05,
+      "loss": 2.1442,
+      "step": 236
+    },
+    {
+      "epoch": 0.07590072057646118,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.7859424920127795e-05,
+      "loss": 2.2005,
+      "step": 237
+    },
+    {
+      "epoch": 0.07622097678142514,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.8019169329073485e-05,
+      "loss": 2.5538,
+      "step": 238
+    },
+    {
+      "epoch": 0.07654123298638911,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.8178913738019174e-05,
+      "loss": 2.22,
+      "step": 239
+    },
+    {
+      "epoch": 0.07686148919135308,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.8338658146964856e-05,
+      "loss": 2.4174,
+      "step": 240
+    },
+    {
+      "epoch": 0.07718174539631706,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.8498402555910545e-05,
+      "loss": 2.3724,
+      "step": 241
+    },
+    {
+      "epoch": 0.07750200160128103,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.8658146964856234e-05,
+      "loss": 2.1111,
+      "step": 242
+    },
+    {
+      "epoch": 0.077822257806245,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.8817891373801916e-05,
+      "loss": 2.2114,
+      "step": 243
+    },
+    {
+      "epoch": 0.07814251401120896,
+      "grad_norm": 0.5,
+      "learning_rate": 3.8977635782747605e-05,
+      "loss": 2.718,
+      "step": 244
+    },
+    {
+      "epoch": 0.07846277021617294,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.913738019169329e-05,
+      "loss": 2.1236,
+      "step": 245
+    },
+    {
+      "epoch": 0.07878302642113691,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.929712460063898e-05,
+      "loss": 2.3918,
+      "step": 246
+    },
+    {
+      "epoch": 0.07910328262610088,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.9456869009584666e-05,
+      "loss": 1.8762,
+      "step": 247
+    },
+    {
+      "epoch": 0.07942353883106486,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.9616613418530355e-05,
+      "loss": 2.2969,
+      "step": 248
+    },
+    {
+      "epoch": 0.07974379503602883,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.9776357827476044e-05,
+      "loss": 2.1578,
+      "step": 249
+    },
+    {
+      "epoch": 0.08006405124099279,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.9936102236421726e-05,
+      "loss": 2.5525,
+      "step": 250
+    },
+    {
+      "epoch": 0.08038430744595676,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.0095846645367415e-05,
+      "loss": 2.4644,
+      "step": 251
+    },
+    {
+      "epoch": 0.08070456365092074,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.0255591054313104e-05,
+      "loss": 1.2111,
+      "step": 252
+    },
+    {
+      "epoch": 0.08102481985588471,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.041533546325879e-05,
+      "loss": 2.522,
+      "step": 253
+    },
+    {
+      "epoch": 0.08134507606084868,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.0575079872204476e-05,
+      "loss": 1.5268,
+      "step": 254
+    },
+    {
+      "epoch": 0.08166533226581266,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.0734824281150165e-05,
+      "loss": 2.331,
+      "step": 255
+    },
+    {
+      "epoch": 0.08198558847077662,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.089456869009585e-05,
+      "loss": 2.2693,
+      "step": 256
+    },
+    {
+      "epoch": 0.08230584467574059,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.1054313099041536e-05,
+      "loss": 2.0041,
+      "step": 257
+    },
+    {
+      "epoch": 0.08262610088070456,
+      "grad_norm": 0.30078125,
+      "learning_rate": 4.1214057507987225e-05,
+      "loss": 1.814,
+      "step": 258
+    },
+    {
+      "epoch": 0.08294635708566854,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.137380191693291e-05,
+      "loss": 2.1807,
+      "step": 259
+    },
+    {
+      "epoch": 0.08326661329063251,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.15335463258786e-05,
+      "loss": 1.8303,
+      "step": 260
+    },
+    {
+      "epoch": 0.08358686949559647,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.169329073482428e-05,
+      "loss": 2.0817,
+      "step": 261
+    },
+    {
+      "epoch": 0.08390712570056044,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.185303514376997e-05,
+      "loss": 2.572,
+      "step": 262
+    },
+    {
+      "epoch": 0.08422738190552442,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.201277955271566e-05,
+      "loss": 2.4427,
+      "step": 263
+    },
+    {
+      "epoch": 0.08454763811048839,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.217252396166134e-05,
+      "loss": 1.6031,
+      "step": 264
+    },
+    {
+      "epoch": 0.08486789431545236,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.233226837060703e-05,
+      "loss": 2.3756,
+      "step": 265
+    },
+    {
+      "epoch": 0.08518815052041634,
+      "grad_norm": 0.75,
+      "learning_rate": 4.249201277955272e-05,
+      "loss": 2.5675,
+      "step": 266
+    },
+    {
+      "epoch": 0.0855084067253803,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.265175718849841e-05,
+      "loss": 2.3992,
+      "step": 267
+    },
+    {
+      "epoch": 0.08582866293034427,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.2811501597444096e-05,
+      "loss": 1.8396,
+      "step": 268
+    },
+    {
+      "epoch": 0.08614891913530824,
+      "grad_norm": 0.625,
+      "learning_rate": 4.297124600638978e-05,
+      "loss": 1.8471,
+      "step": 269
+    },
+    {
+      "epoch": 0.08646917534027222,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.313099041533547e-05,
+      "loss": 2.4466,
+      "step": 270
+    },
+    {
+      "epoch": 0.08678943154523619,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.3290734824281156e-05,
+      "loss": 2.4973,
+      "step": 271
+    },
+    {
+      "epoch": 0.08710968775020016,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.345047923322684e-05,
+      "loss": 2.3791,
+      "step": 272
+    },
+    {
+      "epoch": 0.08742994395516412,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.361022364217253e-05,
+      "loss": 2.2905,
+      "step": 273
+    },
+    {
+      "epoch": 0.0877502001601281,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.376996805111822e-05,
+      "loss": 2.1515,
+      "step": 274
+    },
+    {
+      "epoch": 0.08807045636509207,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.39297124600639e-05,
+      "loss": 2.2641,
+      "step": 275
+    },
+    {
+      "epoch": 0.08839071257005604,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.408945686900959e-05,
+      "loss": 2.4807,
+      "step": 276
+    },
+    {
+      "epoch": 0.08871096877502002,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.424920127795527e-05,
+      "loss": 2.3933,
+      "step": 277
+    },
+    {
+      "epoch": 0.08903122497998399,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.440894568690096e-05,
+      "loss": 1.7232,
+      "step": 278
+    },
+    {
+      "epoch": 0.08935148118494796,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.456869009584665e-05,
+      "loss": 2.3984,
+      "step": 279
+    },
+    {
+      "epoch": 0.08967173738991192,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.472843450479233e-05,
+      "loss": 2.2911,
+      "step": 280
+    },
+    {
+      "epoch": 0.0899919935948759,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.488817891373802e-05,
+      "loss": 2.3337,
+      "step": 281
+    },
+    {
+      "epoch": 0.09031224979983987,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.504792332268371e-05,
+      "loss": 2.0545,
+      "step": 282
+    },
+    {
+      "epoch": 0.09063250600480384,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.520766773162939e-05,
+      "loss": 2.1376,
+      "step": 283
+    },
+    {
+      "epoch": 0.09095276220976782,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.536741214057508e-05,
+      "loss": 2.1938,
+      "step": 284
+    },
+    {
+      "epoch": 0.09127301841473179,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.552715654952077e-05,
+      "loss": 2.4041,
+      "step": 285
+    },
+    {
+      "epoch": 0.09159327461969576,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.568690095846646e-05,
+      "loss": 2.2543,
+      "step": 286
+    },
+    {
+      "epoch": 0.09191353082465972,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.584664536741215e-05,
+      "loss": 2.0115,
+      "step": 287
+    },
+    {
+      "epoch": 0.0922337870296237,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.600638977635783e-05,
+      "loss": 2.3996,
+      "step": 288
+    },
+    {
+      "epoch": 0.09255404323458767,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.616613418530352e-05,
+      "loss": 2.7077,
+      "step": 289
+    },
+    {
+      "epoch": 0.09287429943955164,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.632587859424921e-05,
+      "loss": 2.0203,
+      "step": 290
+    },
+    {
+      "epoch": 0.0931945556445156,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.648562300319489e-05,
+      "loss": 2.2094,
+      "step": 291
+    },
+    {
+      "epoch": 0.09351481184947959,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.664536741214058e-05,
+      "loss": 2.2682,
+      "step": 292
+    },
+    {
+      "epoch": 0.09383506805444355,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.680511182108626e-05,
+      "loss": 2.1374,
+      "step": 293
+    },
+    {
+      "epoch": 0.09415532425940752,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.696485623003195e-05,
+      "loss": 2.272,
+      "step": 294
+    },
+    {
+      "epoch": 0.0944755804643715,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.712460063897764e-05,
+      "loss": 2.384,
+      "step": 295
+    },
+    {
+      "epoch": 0.09479583666933547,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.728434504792332e-05,
+      "loss": 2.4198,
+      "step": 296
+    },
+    {
+      "epoch": 0.09511609287429944,
+      "grad_norm": 0.3125,
+      "learning_rate": 4.744408945686901e-05,
+      "loss": 2.3007,
+      "step": 297
+    },
+    {
+      "epoch": 0.0954363490792634,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.76038338658147e-05,
+      "loss": 2.6156,
+      "step": 298
+    },
+    {
+      "epoch": 0.09575660528422739,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.776357827476038e-05,
+      "loss": 2.2731,
+      "step": 299
+    },
+    {
+      "epoch": 0.09607686148919135,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.792332268370607e-05,
+      "loss": 2.5316,
+      "step": 300
+    },
+    {
+      "epoch": 0.09639711769415532,
+      "grad_norm": 3.75,
+      "learning_rate": 4.8083067092651754e-05,
+      "loss": 1.9556,
+      "step": 301
+    },
+    {
+      "epoch": 0.0967173738991193,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.824281150159744e-05,
+      "loss": 2.4008,
+      "step": 302
+    },
+    {
+      "epoch": 0.09703763010408327,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.840255591054313e-05,
+      "loss": 2.2126,
+      "step": 303
+    },
+    {
+      "epoch": 0.09735788630904724,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.856230031948882e-05,
+      "loss": 1.9912,
+      "step": 304
+    },
+    {
+      "epoch": 0.0976781425140112,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.872204472843451e-05,
+      "loss": 2.3782,
+      "step": 305
+    },
+    {
+      "epoch": 0.09799839871897519,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.88817891373802e-05,
+      "loss": 2.2945,
+      "step": 306
+    },
+    {
+      "epoch": 0.09831865492393915,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.904153354632588e-05,
+      "loss": 2.7067,
+      "step": 307
+    },
+    {
+      "epoch": 0.09863891112890312,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.920127795527157e-05,
+      "loss": 2.0701,
+      "step": 308
+    },
+    {
+      "epoch": 0.09895916733386709,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.936102236421725e-05,
+      "loss": 2.5141,
+      "step": 309
+    },
+    {
+      "epoch": 0.09927942353883107,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.952076677316294e-05,
+      "loss": 2.1579,
+      "step": 310
+    },
+    {
+      "epoch": 0.09959967974379504,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.968051118210863e-05,
+      "loss": 2.4784,
+      "step": 311
+    },
+    {
+      "epoch": 0.099919935948759,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.984025559105431e-05,
+      "loss": 2.285,
+      "step": 312
+    },
+    {
+      "epoch": 0.10024019215372298,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5e-05,
+      "loss": 2.4196,
+      "step": 313
+    },
+    {
+      "epoch": 0.10056044835868695,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.9999984364699426e-05,
+      "loss": 1.4867,
+      "step": 314
+    },
+    {
+      "epoch": 0.10088070456365092,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.999993745881724e-05,
+      "loss": 2.2844,
+      "step": 315
+    },
+    {
+      "epoch": 0.10120096076861489,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.9999859282412125e-05,
+      "loss": 1.8481,
+      "step": 316
+    },
+    {
+      "epoch": 0.10152121697357887,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.9999749835581864e-05,
+      "loss": 2.4302,
+      "step": 317
+    },
+    {
+      "epoch": 0.10184147317854284,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.999960911846336e-05,
+      "loss": 2.3678,
+      "step": 318
+    },
+    {
+      "epoch": 0.1021617293835068,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.999943713123261e-05,
+      "loss": 2.1427,
+      "step": 319
+    },
+    {
+      "epoch": 0.10248198558847077,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.9999233874104755e-05,
+      "loss": 1.591,
+      "step": 320
+    },
+    {
+      "epoch": 0.10280224179343475,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.9998999347334036e-05,
+      "loss": 2.353,
+      "step": 321
+    },
+    {
+      "epoch": 0.10312249799839872,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.9998733551213795e-05,
+      "loss": 2.0951,
+      "step": 322
+    },
+    {
+      "epoch": 0.10344275420336269,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.9998436486076495e-05,
+      "loss": 2.4225,
+      "step": 323
+    },
+    {
+      "epoch": 0.10376301040832667,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.999810815229372e-05,
+      "loss": 1.8435,
+      "step": 324
+    },
+    {
+      "epoch": 0.10408326661329063,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.999774855027615e-05,
+      "loss": 2.0037,
+      "step": 325
+    },
+    {
+      "epoch": 0.1044035228182546,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.999735768047359e-05,
+      "loss": 2.6103,
+      "step": 326
+    },
+    {
+      "epoch": 0.10472377902321857,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.999693554337495e-05,
+      "loss": 2.0573,
+      "step": 327
+    },
+    {
+      "epoch": 0.10504403522818255,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.9996482139508236e-05,
+      "loss": 2.1116,
+      "step": 328
+    },
+    {
+      "epoch": 0.10536429143314652,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.999599746944059e-05,
+      "loss": 2.677,
+      "step": 329
+    },
+    {
+      "epoch": 0.10568454763811048,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.9995481533778256e-05,
+      "loss": 2.2031,
+      "step": 330
+    },
+    {
+      "epoch": 0.10600480384307447,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.999493433316656e-05,
+      "loss": 2.1728,
+      "step": 331
+    },
+    {
+      "epoch": 0.10632506004803843,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.9994355868289965e-05,
+      "loss": 2.239,
+      "step": 332
+    },
+    {
+      "epoch": 0.1066453162530024,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.999374613987202e-05,
+      "loss": 2.5327,
+      "step": 333
+    },
+    {
+      "epoch": 0.10696557245796637,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.9993105148675405e-05,
+      "loss": 2.2862,
+      "step": 334
+    },
+    {
+      "epoch": 0.10728582866293035,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.9992432895501874e-05,
+      "loss": 2.0678,
+      "step": 335
+    },
+    {
+      "epoch": 0.10760608486789432,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.99917293811923e-05,
+      "loss": 2.2222,
+      "step": 336
+    },
+    {
+      "epoch": 0.10792634107285828,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.9990994606626647e-05,
+      "loss": 2.195,
+      "step": 337
+    },
+    {
+      "epoch": 0.10824659727782225,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.9990228572724e-05,
+      "loss": 2.0068,
+      "step": 338
+    },
+    {
+      "epoch": 0.10856685348278623,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.998943128044254e-05,
+      "loss": 2.0167,
+      "step": 339
+    },
+    {
+      "epoch": 0.1088871096877502,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.9988602730779515e-05,
+      "loss": 2.3055,
+      "step": 340
+    },
+    {
+      "epoch": 0.10920736589271417,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.998774292477131e-05,
+      "loss": 2.2542,
+      "step": 341
+    },
+    {
+      "epoch": 0.10952762209767815,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.99868518634934e-05,
+      "loss": 2.564,
+      "step": 342
+    },
+    {
+      "epoch": 0.10984787830264212,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.998592954806033e-05,
+      "loss": 2.2191,
+      "step": 343
+    },
+    {
+      "epoch": 0.11016813450760608,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.998497597962576e-05,
+      "loss": 2.5834,
+      "step": 344
+    },
+    {
+      "epoch": 0.11048839071257005,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.998399115938244e-05,
+      "loss": 2.6079,
+      "step": 345
+    },
+    {
+      "epoch": 0.11080864691753403,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.99829750885622e-05,
+      "loss": 2.0724,
+      "step": 346
+    },
+    {
+      "epoch": 0.111128903122498,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.9981927768435964e-05,
+      "loss": 2.0157,
+      "step": 347
+    },
+    {
+      "epoch": 0.11144915932746197,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.998084920031376e-05,
+      "loss": 2.4644,
+      "step": 348
+    },
+    {
+      "epoch": 0.11176941553242595,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.997973938554466e-05,
+      "loss": 2.6107,
+      "step": 349
+    },
+    {
+      "epoch": 0.11208967173738991,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.997859832551689e-05,
+      "loss": 2.3877,
+      "step": 350
+    },
+    {
+      "epoch": 0.11240992794235388,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.997742602165767e-05,
+      "loss": 2.179,
+      "step": 351
+    },
+    {
+      "epoch": 0.11273018414731785,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.997622247543338e-05,
+      "loss": 2.5496,
+      "step": 352
+    },
+    {
+      "epoch": 0.11305044035228183,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.9974987688349415e-05,
+      "loss": 1.8469,
+      "step": 353
+    },
+    {
+      "epoch": 0.1133706965572458,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.9973721661950295e-05,
+      "loss": 2.3798,
+      "step": 354
+    },
+    {
+      "epoch": 0.11369095276220977,
+      "grad_norm": 0.875,
+      "learning_rate": 4.9972424397819596e-05,
+      "loss": 2.4496,
+      "step": 355
+    },
+    {
+      "epoch": 0.11401120896717373,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.997109589757997e-05,
+      "loss": 2.0667,
+      "step": 356
+    },
+    {
+      "epoch": 0.11433146517213771,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.996973616289312e-05,
+      "loss": 2.1528,
+      "step": 357
+    },
+    {
+      "epoch": 0.11465172137710168,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.996834519545985e-05,
+      "loss": 1.8263,
+      "step": 358
+    },
+    {
+      "epoch": 0.11497197758206565,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.996692299702e-05,
+      "loss": 2.629,
+      "step": 359
+    },
+    {
+      "epoch": 0.11529223378702963,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.996546956935252e-05,
+      "loss": 2.4788,
+      "step": 360
+    },
+    {
+      "epoch": 0.1156124899919936,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.9963984914275363e-05,
+      "loss": 2.2397,
+      "step": 361
+    },
+    {
+      "epoch": 0.11593274619695756,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.996246903364559e-05,
+      "loss": 2.3035,
+      "step": 362
+    },
+    {
+      "epoch": 0.11625300240192153,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.996092192935929e-05,
+      "loss": 2.6009,
+      "step": 363
+    },
+    {
+      "epoch": 0.11657325860688551,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.99593436033516e-05,
+      "loss": 2.3057,
+      "step": 364
+    },
+    {
+      "epoch": 0.11689351481184948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.9957734057596774e-05,
+      "loss": 2.5767,
+      "step": 365
+    },
+    {
+      "epoch": 0.11721377101681345,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.995609329410804e-05,
+      "loss": 1.9374,
+      "step": 366
+    },
+    {
+      "epoch": 0.11753402722177742,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.995442131493771e-05,
+      "loss": 1.9711,
+      "step": 367
+    },
+    {
+      "epoch": 0.1178542834267414,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.9952718122177126e-05,
+      "loss": 2.3285,
+      "step": 368
+    },
+    {
+      "epoch": 0.11817453963170536,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.9950983717956703e-05,
+      "loss": 2.0349,
+      "step": 369
+    },
+    {
+      "epoch": 0.11849479583666933,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.994921810444586e-05,
+      "loss": 2.1336,
+      "step": 370
+    },
+    {
+      "epoch": 0.11881505204163331,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.994742128385307e-05,
+      "loss": 2.319,
+      "step": 371
+    },
+    {
+      "epoch": 0.11913530824659728,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.994559325842585e-05,
+      "loss": 2.0951,
+      "step": 372
+    },
+    {
+      "epoch": 0.11945556445156125,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.994373403045073e-05,
+      "loss": 2.4311,
+      "step": 373
+    },
+    {
+      "epoch": 0.11977582065652521,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.994184360225328e-05,
+      "loss": 1.8299,
+      "step": 374
+    },
+    {
+      "epoch": 0.1200960768614892,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.993992197619809e-05,
+      "loss": 2.4937,
+      "step": 375
+    },
+    {
+      "epoch": 0.12041633306645316,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.9937969154688766e-05,
+      "loss": 2.3098,
+      "step": 376
+    },
+    {
+      "epoch": 0.12073658927141713,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.993598514016797e-05,
+      "loss": 2.2064,
+      "step": 377
+    },
+    {
+      "epoch": 0.12105684547638111,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.9933969935117335e-05,
+      "loss": 2.0565,
+      "step": 378
+    },
+    {
+      "epoch": 0.12137710168134508,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.9931923542057534e-05,
+      "loss": 2.6077,
+      "step": 379
+    },
+    {
+      "epoch": 0.12169735788630905,
+      "grad_norm": 0.375,
+      "learning_rate": 4.9929845963548246e-05,
+      "loss": 2.3129,
+      "step": 380
+    },
+    {
+      "epoch": 0.12201761409127301,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.9927737202188155e-05,
+      "loss": 2.5353,
+      "step": 381
+    },
+    {
+      "epoch": 0.122337870296237,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.992559726061494e-05,
+      "loss": 2.0995,
+      "step": 382
+    },
+    {
+      "epoch": 0.12265812650120096,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.9923426141505305e-05,
+      "loss": 1.9616,
+      "step": 383
+    },
+    {
+      "epoch": 0.12297838270616493,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.9921223847574935e-05,
+      "loss": 2.1245,
+      "step": 384
+    },
+    {
+      "epoch": 0.1232986389111289,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.991899038157851e-05,
+      "loss": 2.2955,
+      "step": 385
+    },
+    {
+      "epoch": 0.12361889511609288,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.991672574630971e-05,
+      "loss": 2.2364,
+      "step": 386
+    },
+    {
+      "epoch": 0.12393915132105685,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.9914429944601185e-05,
+      "loss": 2.2836,
+      "step": 387
+    },
+    {
+      "epoch": 0.12425940752602081,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.991210297932459e-05,
+      "loss": 1.9992,
+      "step": 388
+    },
+    {
+      "epoch": 0.1245796637309848,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.9909744853390534e-05,
+      "loss": 2.2879,
+      "step": 389
+    },
+    {
+      "epoch": 0.12489991993594876,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.990735556974863e-05,
+      "loss": 2.2375,
+      "step": 390
+    },
+    {
+      "epoch": 0.12522017614091274,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.990493513138744e-05,
+      "loss": 2.5697,
+      "step": 391
+    },
+    {
+      "epoch": 0.1255404323458767,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.990248354133452e-05,
+      "loss": 1.9927,
+      "step": 392
+    },
+    {
+      "epoch": 0.12586068855084068,
+      "grad_norm": 0.5,
+      "learning_rate": 4.9900000802656376e-05,
+      "loss": 2.2876,
+      "step": 393
+    },
+    {
+      "epoch": 0.12618094475580463,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.989748691845847e-05,
+      "loss": 2.3259,
+      "step": 394
+    },
+    {
+      "epoch": 0.1265012009607686,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.989494189188523e-05,
+      "loss": 2.6403,
+      "step": 395
+    },
+    {
+      "epoch": 0.1268214571657326,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.989236572612004e-05,
+      "loss": 2.4835,
+      "step": 396
+    },
+    {
+      "epoch": 0.12714171337069655,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.988975842438523e-05,
+      "loss": 2.2297,
+      "step": 397
+    },
+    {
+      "epoch": 0.12746196957566053,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.9887119989942073e-05,
+      "loss": 2.4837,
+      "step": 398
+    },
+    {
+      "epoch": 0.1277822257806245,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.9884450426090786e-05,
+      "loss": 2.9118,
+      "step": 399
+    },
+    {
+      "epoch": 0.12810248198558846,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.9881749736170525e-05,
+      "loss": 2.4227,
+      "step": 400
+    },
+    {
+      "epoch": 0.12842273819055244,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.987901792355938e-05,
+      "loss": 2.6651,
+      "step": 401
+    },
+    {
+      "epoch": 0.12874299439551642,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.987625499167436e-05,
+      "loss": 2.61,
+      "step": 402
+    },
+    {
+      "epoch": 0.12906325060048038,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.9873460943971416e-05,
+      "loss": 2.1197,
+      "step": 403
+    },
+    {
+      "epoch": 0.12938350680544436,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.987063578394541e-05,
+      "loss": 2.6976,
+      "step": 404
+    },
+    {
+      "epoch": 0.1297037630104083,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.986777951513011e-05,
+      "loss": 2.242,
+      "step": 405
+    },
+    {
+      "epoch": 0.1300240192153723,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.9864892141098215e-05,
+      "loss": 2.3003,
+      "step": 406
+    },
+    {
+      "epoch": 0.13034427542033628,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.986197366546133e-05,
+      "loss": 1.8946,
+      "step": 407
+    },
+    {
+      "epoch": 0.13066453162530023,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.985902409186993e-05,
+      "loss": 2.1781,
+      "step": 408
+    },
+    {
+      "epoch": 0.1309847878302642,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.985604342401343e-05,
+      "loss": 1.9627,
+      "step": 409
+    },
+    {
+      "epoch": 0.1313050440352282,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.9853031665620116e-05,
+      "loss": 2.161,
+      "step": 410
+    },
+    {
+      "epoch": 0.13162530024019214,
+      "grad_norm": 1.7734375,
+      "learning_rate": 4.984998882045717e-05,
+      "loss": 2.3552,
+      "step": 411
+    },
+    {
+      "epoch": 0.13194555644515613,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.9846914892330654e-05,
+      "loss": 2.2441,
+      "step": 412
+    },
+    {
+      "epoch": 0.1322658126501201,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.984380988508551e-05,
+      "loss": 1.9077,
+      "step": 413
+    },
+    {
+      "epoch": 0.13258606885508406,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.9840673802605566e-05,
+      "loss": 2.5129,
+      "step": 414
+    },
+    {
+      "epoch": 0.13290632506004804,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.98375066488135e-05,
+      "loss": 2.2193,
+      "step": 415
+    },
+    {
+      "epoch": 0.13322658126501202,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.983430842767086e-05,
+      "loss": 2.5459,
+      "step": 416
+    },
+    {
+      "epoch": 0.13354683746997598,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.9831079143178066e-05,
+      "loss": 2.6023,
+      "step": 417
+    },
+    {
+      "epoch": 0.13386709367493996,
+      "grad_norm": 9.5,
+      "learning_rate": 4.982781879937438e-05,
+      "loss": 2.1457,
+      "step": 418
+    },
+    {
+      "epoch": 0.1341873498799039,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.982452740033793e-05,
+      "loss": 2.0819,
+      "step": 419
+    },
+    {
+      "epoch": 0.1345076060848679,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.982120495018566e-05,
+      "loss": 2.6695,
+      "step": 420
+    },
+    {
+      "epoch": 0.13482786228983187,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.981785145307337e-05,
+      "loss": 2.4198,
+      "step": 421
+    },
+    {
+      "epoch": 0.13514811849479583,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.9814466913195717e-05,
+      "loss": 2.0445,
+      "step": 422
+    },
+    {
+      "epoch": 0.1354683746997598,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.981105133478614e-05,
+      "loss": 2.3541,
+      "step": 423
+    },
+    {
+      "epoch": 0.1357886309047238,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.9807604722116945e-05,
+      "loss": 1.7954,
+      "step": 424
+    },
+    {
+      "epoch": 0.13610888710968774,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.980412707949923e-05,
+      "loss": 2.3201,
+      "step": 425
+    },
+    {
+      "epoch": 0.13642914331465172,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.9800618411282914e-05,
+      "loss": 2.3104,
+      "step": 426
+    },
+    {
+      "epoch": 0.1367493995196157,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.979707872185672e-05,
+      "loss": 2.4099,
+      "step": 427
+    },
+    {
+      "epoch": 0.13706965572457966,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.979350801564818e-05,
+      "loss": 2.0,
+      "step": 428
+    },
+    {
+      "epoch": 0.13738991192954364,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.9789906297123626e-05,
+      "loss": 2.3145,
+      "step": 429
+    },
+    {
+      "epoch": 0.1377101681345076,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.978627357078817e-05,
+      "loss": 2.1489,
+      "step": 430
+    },
+    {
+      "epoch": 0.13803042433947157,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.97826098411857e-05,
+      "loss": 2.0548,
+      "step": 431
+    },
+    {
+      "epoch": 0.13835068054443556,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.9778915112898914e-05,
+      "loss": 2.3199,
+      "step": 432
+    },
+    {
+      "epoch": 0.1386709367493995,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.977518939054927e-05,
+      "loss": 2.1056,
+      "step": 433
+    },
+    {
+      "epoch": 0.1389911929543635,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.977143267879697e-05,
+      "loss": 2.2003,
+      "step": 434
+    },
+    {
+      "epoch": 0.13931144915932747,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.9767644982341014e-05,
+      "loss": 2.1945,
+      "step": 435
+    },
+    {
+      "epoch": 0.13963170536429143,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.9763826305919146e-05,
+      "loss": 2.4707,
+      "step": 436
+    },
+    {
+      "epoch": 0.1399519615692554,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.975997665430785e-05,
+      "loss": 2.5856,
+      "step": 437
+    },
+    {
+      "epoch": 0.1402722177742194,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.9756096032322376e-05,
+      "loss": 2.2698,
+      "step": 438
+    },
+    {
+      "epoch": 0.14059247397918334,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.975218444481668e-05,
+      "loss": 2.5266,
+      "step": 439
+    },
+    {
+      "epoch": 0.14091273018414732,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.974824189668348e-05,
+      "loss": 2.2826,
+      "step": 440
+    },
+    {
+      "epoch": 0.14123298638911128,
+      "grad_norm": 1.28125,
+      "learning_rate": 4.974426839285422e-05,
+      "loss": 1.8341,
+      "step": 441
+    },
+    {
+      "epoch": 0.14155324259407526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.9740263938299034e-05,
+      "loss": 2.6753,
+      "step": 442
+    },
+    {
+      "epoch": 0.14187349879903924,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.973622853802681e-05,
+      "loss": 2.436,
+      "step": 443
+    },
+    {
+      "epoch": 0.1421937550040032,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.97321621970851e-05,
+      "loss": 2.1716,
+      "step": 444
+    },
+    {
+      "epoch": 0.14251401120896717,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.97280649205602e-05,
+      "loss": 2.0777,
+      "step": 445
+    },
+    {
+      "epoch": 0.14283426741393115,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.9723936713577084e-05,
+      "loss": 1.7713,
+      "step": 446
+    },
+    {
+      "epoch": 0.1431545236188951,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.97197775812994e-05,
+      "loss": 2.645,
+      "step": 447
+    },
+    {
+      "epoch": 0.1434747798238591,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.97155875289295e-05,
+      "loss": 2.1758,
+      "step": 448
+    },
+    {
+      "epoch": 0.14379503602882307,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.9711366561708395e-05,
+      "loss": 2.2733,
+      "step": 449
+    },
+    {
+      "epoch": 0.14411529223378702,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.9707114684915776e-05,
+      "loss": 1.9709,
+      "step": 450
+    },
+    {
+      "epoch": 0.144435548438751,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.9702831903869996e-05,
+      "loss": 1.9222,
+      "step": 451
+    },
+    {
+      "epoch": 0.14475580464371496,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.969851822392805e-05,
+      "loss": 2.0273,
+      "step": 452
+    },
+    {
+      "epoch": 0.14507606084867894,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.9694173650485595e-05,
+      "loss": 2.286,
+      "step": 453
+    },
+    {
+      "epoch": 0.14539631705364292,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.968979818897694e-05,
+      "loss": 2.6001,
+      "step": 454
+    },
+    {
+      "epoch": 0.14571657325860687,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.968539184487502e-05,
+      "loss": 2.4443,
+      "step": 455
+    },
+    {
+      "epoch": 0.14603682946357086,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.9680954623691374e-05,
+      "loss": 2.2168,
+      "step": 456
+    },
+    {
+      "epoch": 0.14635708566853484,
+      "grad_norm": 0.5,
+      "learning_rate": 4.9676486530976196e-05,
+      "loss": 2.4374,
+      "step": 457
+    },
+    {
+      "epoch": 0.1466773418734988,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.967198757231829e-05,
+      "loss": 2.1691,
+      "step": 458
+    },
+    {
+      "epoch": 0.14699759807846277,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.966745775334505e-05,
+      "loss": 2.7222,
+      "step": 459
+    },
+    {
+      "epoch": 0.14731785428342675,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.966289707972249e-05,
+      "loss": 1.9953,
+      "step": 460
+    },
+    {
+      "epoch": 0.1476381104883907,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.965830555715522e-05,
+      "loss": 2.0458,
+      "step": 461
+    },
+    {
+      "epoch": 0.1479583666933547,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.9653683191386405e-05,
+      "loss": 2.2034,
+      "step": 462
+    },
+    {
+      "epoch": 0.14827862289831867,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.964902998819782e-05,
+      "loss": 2.0623,
+      "step": 463
+    },
+    {
+      "epoch": 0.14859887910328262,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.964434595340982e-05,
+      "loss": 2.0543,
+      "step": 464
+    },
+    {
+      "epoch": 0.1489191353082466,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.963963109288129e-05,
+      "loss": 1.7578,
+      "step": 465
+    },
+    {
+      "epoch": 0.14923939151321056,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.963488541250969e-05,
+      "loss": 1.9211,
+      "step": 466
+    },
+    {
+      "epoch": 0.14955964771817454,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.963010891823103e-05,
+      "loss": 2.1433,
+      "step": 467
+    },
+    {
+      "epoch": 0.14987990392313852,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.962530161601988e-05,
+      "loss": 2.7066,
+      "step": 468
+    },
+    {
+      "epoch": 0.15020016012810247,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.9620463511889314e-05,
+      "loss": 1.7484,
+      "step": 469
+    },
+    {
+      "epoch": 0.15052041633306645,
+      "grad_norm": 0.380859375,
+      "learning_rate": 4.961559461189096e-05,
+      "loss": 2.2189,
+      "step": 470
+    },
+    {
+      "epoch": 0.15084067253803043,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.961069492211494e-05,
+      "loss": 1.9058,
+      "step": 471
+    },
+    {
+      "epoch": 0.1511609287429944,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.960576444868992e-05,
+      "loss": 2.1691,
+      "step": 472
+    },
+    {
+      "epoch": 0.15148118494795837,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.960080319778305e-05,
+      "loss": 2.2587,
+      "step": 473
+    },
+    {
+      "epoch": 0.15180144115292235,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.959581117559998e-05,
+      "loss": 1.6996,
+      "step": 474
+    },
+    {
+      "epoch": 0.1521216973578863,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.959078838838485e-05,
+      "loss": 2.3875,
+      "step": 475
+    },
+    {
+      "epoch": 0.15244195356285029,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.958573484242027e-05,
+      "loss": 2.4589,
+      "step": 476
+    },
+    {
+      "epoch": 0.15276220976781424,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.9580650544027376e-05,
+      "loss": 2.1532,
+      "step": 477
+    },
+    {
+      "epoch": 0.15308246597277822,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.957553549956568e-05,
+      "loss": 2.2743,
+      "step": 478
+    },
+    {
+      "epoch": 0.1534027221777422,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.957038971543325e-05,
+      "loss": 1.9105,
+      "step": 479
+    },
+    {
+      "epoch": 0.15372297838270615,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.956521319806653e-05,
+      "loss": 2.2667,
+      "step": 480
+    },
+    {
+      "epoch": 0.15404323458767014,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.956000595394043e-05,
+      "loss": 1.9691,
+      "step": 481
+    },
+    {
+      "epoch": 0.15436349079263412,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.955476798956831e-05,
+      "loss": 1.549,
+      "step": 482
+    },
+    {
+      "epoch": 0.15468374699759807,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.954949931150193e-05,
+      "loss": 1.8682,
+      "step": 483
+    },
+    {
+      "epoch": 0.15500400320256205,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.9544199926331484e-05,
+      "loss": 1.846,
+      "step": 484
+    },
+    {
+      "epoch": 0.15532425940752603,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.953886984068558e-05,
+      "loss": 2.1578,
+      "step": 485
+    },
+    {
+      "epoch": 0.15564451561249,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.95335090612312e-05,
+      "loss": 2.2433,
+      "step": 486
+    },
+    {
+      "epoch": 0.15596477181745397,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.952811759467374e-05,
+      "loss": 2.2284,
+      "step": 487
+    },
+    {
+      "epoch": 0.15628502802241792,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.9522695447757e-05,
+      "loss": 2.5033,
+      "step": 488
+    },
+    {
+      "epoch": 0.1566052842273819,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.9517242627263096e-05,
+      "loss": 1.7995,
+      "step": 489
+    },
+    {
+      "epoch": 0.15692554043234588,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.951175914001256e-05,
+      "loss": 2.5606,
+      "step": 490
+    },
+    {
+      "epoch": 0.15724579663730984,
+      "grad_norm": 1.7265625,
+      "learning_rate": 4.950624499286428e-05,
+      "loss": 2.3475,
+      "step": 491
+    },
+    {
+      "epoch": 0.15756605284227382,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.950070019271547e-05,
+      "loss": 2.3772,
+      "step": 492
+    },
+    {
+      "epoch": 0.1578863090472378,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.949512474650171e-05,
+      "loss": 1.9745,
+      "step": 493
+    },
+    {
+      "epoch": 0.15820656525220175,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.94895186611969e-05,
+      "loss": 2.5581,
+      "step": 494
+    },
+    {
+      "epoch": 0.15852682145716573,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.948388194381326e-05,
+      "loss": 2.3336,
+      "step": 495
+    },
+    {
+      "epoch": 0.15884707766212972,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.947821460140134e-05,
+      "loss": 2.1051,
+      "step": 496
+    },
+    {
+      "epoch": 0.15916733386709367,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.947251664104998e-05,
+      "loss": 2.4866,
+      "step": 497
+    },
+    {
+      "epoch": 0.15948759007205765,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.946678806988633e-05,
+      "loss": 2.2679,
+      "step": 498
+    },
+    {
+      "epoch": 0.1598078462770216,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.9461028895075825e-05,
+      "loss": 2.4515,
+      "step": 499
+    },
+    {
+      "epoch": 0.16012810248198558,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.9455239123822184e-05,
+      "loss": 1.6854,
+      "step": 500
+    },
+    {
+      "epoch": 0.16044835868694957,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.944941876336738e-05,
+      "loss": 2.0233,
+      "step": 501
+    },
+    {
+      "epoch": 0.16076861489191352,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.944356782099167e-05,
+      "loss": 2.467,
+      "step": 502
+    },
+    {
+      "epoch": 0.1610888710968775,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.943768630401355e-05,
+      "loss": 2.4588,
+      "step": 503
+    },
+    {
+      "epoch": 0.16140912730184148,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.943177421978976e-05,
+      "loss": 2.4571,
+      "step": 504
+    },
+    {
+      "epoch": 0.16172938350680544,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.9425831575715286e-05,
+      "loss": 2.6707,
+      "step": 505
+    },
+    {
+      "epoch": 0.16204963971176942,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.941985837922332e-05,
+      "loss": 2.1903,
+      "step": 506
+    },
+    {
+      "epoch": 0.1623698959167334,
+      "grad_norm": 0.302734375,
+      "learning_rate": 4.941385463778528e-05,
+      "loss": 1.7641,
+      "step": 507
+    },
+    {
+      "epoch": 0.16269015212169735,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.9407820358910804e-05,
+      "loss": 2.4916,
+      "step": 508
+    },
+    {
+      "epoch": 0.16301040832666133,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.94017555501477e-05,
+      "loss": 2.0425,
+      "step": 509
+    },
+    {
+      "epoch": 0.1633306645316253,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.939566021908197e-05,
+      "loss": 2.0939,
+      "step": 510
+    },
+    {
+      "epoch": 0.16365092073658927,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.938953437333783e-05,
+      "loss": 2.3065,
+      "step": 511
+    },
+    {
+      "epoch": 0.16397117694155325,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.93833780205776e-05,
+      "loss": 2.7856,
+      "step": 512
+    },
+    {
+      "epoch": 0.1642914331465172,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.937719116850181e-05,
+      "loss": 2.3894,
+      "step": 513
+    },
+    {
+      "epoch": 0.16461168935148118,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.937097382484913e-05,
+      "loss": 2.5719,
+      "step": 514
+    },
+    {
+      "epoch": 0.16493194555644516,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.936472599739635e-05,
+      "loss": 1.8058,
+      "step": 515
+    },
+    {
+      "epoch": 0.16525220176140912,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.935844769395842e-05,
+      "loss": 2.2764,
+      "step": 516
+    },
+    {
+      "epoch": 0.1655724579663731,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.9352138922388375e-05,
+      "loss": 1.6205,
+      "step": 517
+    },
+    {
+      "epoch": 0.16589271417133708,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.9345799690577385e-05,
+      "loss": 1.8822,
+      "step": 518
+    },
+    {
+      "epoch": 0.16621297037630103,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.933943000645471e-05,
+      "loss": 2.3701,
+      "step": 519
+    },
+    {
+      "epoch": 0.16653322658126501,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.9333029877987715e-05,
+      "loss": 2.474,
+      "step": 520
+    },
+    {
+      "epoch": 0.166853482786229,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.932659931318182e-05,
+      "loss": 2.0294,
+      "step": 521
+    },
+    {
+      "epoch": 0.16717373899119295,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.932013832008054e-05,
+      "loss": 2.218,
+      "step": 522
+    },
+    {
+      "epoch": 0.16749399519615693,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.931364690676544e-05,
+      "loss": 2.5899,
+      "step": 523
+    },
+    {
+      "epoch": 0.16781425140112088,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.930712508135613e-05,
+      "loss": 1.828,
+      "step": 524
+    },
+    {
+      "epoch": 0.16813450760608487,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.930057285201027e-05,
+      "loss": 2.627,
+      "step": 525
+    },
+    {
+      "epoch": 0.16845476381104885,
+      "grad_norm": 1.25,
+      "learning_rate": 4.929399022692355e-05,
+      "loss": 2.1664,
+      "step": 526
+    },
+    {
+      "epoch": 0.1687750200160128,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.928737721432967e-05,
+      "loss": 1.8216,
+      "step": 527
+    },
+    {
+      "epoch": 0.16909527622097678,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.9280733822500346e-05,
+      "loss": 1.9037,
+      "step": 528
+    },
+    {
+      "epoch": 0.16941553242594076,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.927406005974529e-05,
+      "loss": 2.0505,
+      "step": 529
+    },
+    {
+      "epoch": 0.16973578863090472,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.9267355934412214e-05,
+      "loss": 2.3799,
+      "step": 530
+    },
+    {
+      "epoch": 0.1700560448358687,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.926062145488679e-05,
+      "loss": 2.2806,
+      "step": 531
+    },
+    {
+      "epoch": 0.17037630104083268,
+      "grad_norm": 0.314453125,
+      "learning_rate": 4.9253856629592663e-05,
+      "loss": 2.4759,
+      "step": 532
+    },
+    {
+      "epoch": 0.17069655724579663,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.924706146699146e-05,
+      "loss": 2.6853,
+      "step": 533
+    },
+    {
+      "epoch": 0.1710168134507606,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.924023597558271e-05,
+      "loss": 1.9226,
+      "step": 534
+    },
+    {
+      "epoch": 0.17133706965572457,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.923338016390392e-05,
+      "loss": 1.8459,
+      "step": 535
+    },
+    {
+      "epoch": 0.17165732586068855,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.922649404053048e-05,
+      "loss": 2.0748,
+      "step": 536
+    },
+    {
+      "epoch": 0.17197758206565253,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.921957761407575e-05,
+      "loss": 2.0169,
+      "step": 537
+    },
+    {
+      "epoch": 0.17229783827061648,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.921263089319095e-05,
+      "loss": 1.7651,
+      "step": 538
+    },
+    {
+      "epoch": 0.17261809447558046,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.920565388656519e-05,
+      "loss": 2.498,
+      "step": 539
+    },
+    {
+      "epoch": 0.17293835068054444,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.919864660292549e-05,
+      "loss": 2.1956,
+      "step": 540
+    },
+    {
+      "epoch": 0.1732586068855084,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.919160905103673e-05,
+      "loss": 2.4898,
+      "step": 541
+    },
+    {
+      "epoch": 0.17357886309047238,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.9184541239701666e-05,
+      "loss": 2.1794,
+      "step": 542
+    },
+    {
+      "epoch": 0.17389911929543636,
+      "grad_norm": 0.5,
+      "learning_rate": 4.9177443177760854e-05,
+      "loss": 2.1124,
+      "step": 543
+    },
+    {
+      "epoch": 0.17421937550040031,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.9170314874092734e-05,
+      "loss": 2.161,
+      "step": 544
+    },
+    {
+      "epoch": 0.1745396317053643,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.916315633761356e-05,
+      "loss": 2.2314,
+      "step": 545
+    },
+    {
+      "epoch": 0.17485988791032825,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.915596757727741e-05,
+      "loss": 1.9608,
+      "step": 546
+    },
+    {
+      "epoch": 0.17518014411529223,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.9148748602076146e-05,
+      "loss": 1.8623,
+      "step": 547
+    },
+    {
+      "epoch": 0.1755004003202562,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.914149942103944e-05,
+      "loss": 2.0954,
+      "step": 548
+    },
+    {
+      "epoch": 0.17582065652522016,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.913422004323474e-05,
+      "loss": 2.0595,
+      "step": 549
+    },
+    {
+      "epoch": 0.17614091273018415,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.912691047776727e-05,
+      "loss": 2.1955,
+      "step": 550
+    },
+    {
+      "epoch": 0.17646116893514813,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.9119570733780015e-05,
+      "loss": 2.0215,
+      "step": 551
+    },
+    {
+      "epoch": 0.17678142514011208,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.91122008204537e-05,
+      "loss": 2.29,
+      "step": 552
+    },
+    {
+      "epoch": 0.17710168134507606,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.910480074700678e-05,
+      "loss": 2.1564,
+      "step": 553
+    },
+    {
+      "epoch": 0.17742193755004004,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.909737052269545e-05,
+      "loss": 2.0404,
+      "step": 554
+    },
+    {
+      "epoch": 0.177742193755004,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.908991015681362e-05,
+      "loss": 2.3606,
+      "step": 555
+    },
+    {
+      "epoch": 0.17806244995996798,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.908241965869289e-05,
+      "loss": 2.4787,
+      "step": 556
+    },
+    {
+      "epoch": 0.17838270616493196,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.907489903770256e-05,
+      "loss": 2.3174,
+      "step": 557
+    },
+    {
+      "epoch": 0.1787029623698959,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.9067348303249594e-05,
+      "loss": 2.5595,
+      "step": 558
+    },
+    {
+      "epoch": 0.1790232185748599,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.905976746477864e-05,
+      "loss": 2.6765,
+      "step": 559
+    },
+    {
+      "epoch": 0.17934347477982385,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.905215653177199e-05,
+      "loss": 2.6524,
+      "step": 560
+    },
+    {
+      "epoch": 0.17966373098478783,
+      "grad_norm": 0.5,
+      "learning_rate": 4.904451551374959e-05,
+      "loss": 2.497,
+      "step": 561
+    },
+    {
+      "epoch": 0.1799839871897518,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.903684442026899e-05,
+      "loss": 2.0149,
+      "step": 562
+    },
+    {
+      "epoch": 0.18030424339471576,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.9029143260925395e-05,
+      "loss": 2.1776,
+      "step": 563
+    },
+    {
+      "epoch": 0.18062449959967974,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.902141204535159e-05,
+      "loss": 1.8415,
+      "step": 564
+    },
+    {
+      "epoch": 0.18094475580464373,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.901365078321798e-05,
+      "loss": 2.2762,
+      "step": 565
+    },
+    {
+      "epoch": 0.18126501200960768,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.9005859484232526e-05,
+      "loss": 2.6546,
+      "step": 566
+    },
+    {
+      "epoch": 0.18158526821457166,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.899803815814077e-05,
+      "loss": 2.1712,
+      "step": 567
+    },
+    {
+      "epoch": 0.18190552441953564,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.899018681472582e-05,
+      "loss": 2.5383,
+      "step": 568
+    },
+    {
+      "epoch": 0.1822257806244996,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.8982305463808334e-05,
+      "loss": 2.3668,
+      "step": 569
+    },
+    {
+      "epoch": 0.18254603682946358,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.897439411524648e-05,
+      "loss": 2.4888,
+      "step": 570
+    },
+    {
+      "epoch": 0.18286629303442753,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.896645277893597e-05,
+      "loss": 1.5993,
+      "step": 571
+    },
+    {
+      "epoch": 0.1831865492393915,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.895848146481002e-05,
+      "loss": 2.1428,
+      "step": 572
+    },
+    {
+      "epoch": 0.1835068054443555,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.895048018283933e-05,
+      "loss": 2.6575,
+      "step": 573
+    },
+    {
+      "epoch": 0.18382706164931945,
+      "grad_norm": 1.9296875,
+      "learning_rate": 4.8942448943032116e-05,
+      "loss": 2.0452,
+      "step": 574
+    },
+    {
+      "epoch": 0.18414731785428343,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.893438775543403e-05,
+      "loss": 1.9505,
+      "step": 575
+    },
+    {
+      "epoch": 0.1844675740592474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.89262966301282e-05,
+      "loss": 1.9657,
+      "step": 576
+    },
+    {
+      "epoch": 0.18478783026421136,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.8918175577235214e-05,
+      "loss": 2.5377,
+      "step": 577
+    },
+    {
+      "epoch": 0.18510808646917534,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.891002460691306e-05,
+      "loss": 2.1608,
+      "step": 578
+    },
+    {
+      "epoch": 0.18542834267413932,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.890184372935719e-05,
+      "loss": 2.0481,
+      "step": 579
+    },
+    {
+      "epoch": 0.18574859887910328,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.889363295480042e-05,
+      "loss": 1.9898,
+      "step": 580
+    },
+    {
+      "epoch": 0.18606885508406726,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.8885392293513e-05,
+      "loss": 2.3255,
+      "step": 581
+    },
+    {
+      "epoch": 0.1863891112890312,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.8877121755802536e-05,
+      "loss": 2.2675,
+      "step": 582
+    },
+    {
+      "epoch": 0.1867093674939952,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.8868821352014024e-05,
+      "loss": 2.2714,
+      "step": 583
+    },
+    {
+      "epoch": 0.18702962369895917,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.886049109252981e-05,
+      "loss": 2.0052,
+      "step": 584
+    },
+    {
+      "epoch": 0.18734987990392313,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.8852130987769574e-05,
+      "loss": 2.4493,
+      "step": 585
+    },
+    {
+      "epoch": 0.1876701361088871,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.884374104819035e-05,
+      "loss": 2.2796,
+      "step": 586
+    },
+    {
+      "epoch": 0.1879903923138511,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.883532128428645e-05,
+      "loss": 1.4809,
+      "step": 587
+    },
+    {
+      "epoch": 0.18831064851881504,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.882687170658955e-05,
+      "loss": 1.5241,
+      "step": 588
+    },
+    {
+      "epoch": 0.18863090472377902,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.881839232566856e-05,
+      "loss": 2.1891,
+      "step": 589
+    },
+    {
+      "epoch": 0.188951160928743,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.880988315212971e-05,
+      "loss": 2.2043,
+      "step": 590
+    },
+    {
+      "epoch": 0.18927141713370696,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.880134419661647e-05,
+      "loss": 2.3487,
+      "step": 591
+    },
+    {
+      "epoch": 0.18959167333867094,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.879277546980957e-05,
+      "loss": 2.6678,
+      "step": 592
+    },
+    {
+      "epoch": 0.1899119295436349,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.8784176982426984e-05,
+      "loss": 2.1424,
+      "step": 593
+    },
+    {
+      "epoch": 0.19023218574859888,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.87755487452239e-05,
+      "loss": 2.3376,
+      "step": 594
+    },
+    {
+      "epoch": 0.19055244195356286,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.8766890768992734e-05,
+      "loss": 2.4265,
+      "step": 595
+    },
+    {
+      "epoch": 0.1908726981585268,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.875820306456309e-05,
+      "loss": 2.366,
+      "step": 596
+    },
+    {
+      "epoch": 0.1911929543634908,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.8749485642801754e-05,
+      "loss": 2.4942,
+      "step": 597
+    },
+    {
+      "epoch": 0.19151321056845477,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.874073851461268e-05,
+      "loss": 1.9429,
+      "step": 598
+    },
+    {
+      "epoch": 0.19183346677341873,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.873196169093699e-05,
+      "loss": 2.1505,
+      "step": 599
+    },
+    {
+      "epoch": 0.1921537229783827,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.872315518275296e-05,
+      "loss": 2.373,
+      "step": 600
+    },
+    {
+      "epoch": 0.1924739791833467,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.871431900107597e-05,
+      "loss": 2.2588,
+      "step": 601
+    },
+    {
+      "epoch": 0.19279423538831064,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.870545315695853e-05,
+      "loss": 2.6125,
+      "step": 602
+    },
+    {
+      "epoch": 0.19311449159327462,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.869655766149024e-05,
+      "loss": 2.489,
+      "step": 603
+    },
+    {
+      "epoch": 0.1934347477982386,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.868763252579782e-05,
+      "loss": 2.29,
+      "step": 604
+    },
+    {
+      "epoch": 0.19375500400320256,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.867867776104503e-05,
+      "loss": 2.3659,
+      "step": 605
+    },
+    {
+      "epoch": 0.19407526020816654,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.866969337843271e-05,
+      "loss": 1.8565,
+      "step": 606
+    },
+    {
+      "epoch": 0.1943955164131305,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.8660679389198746e-05,
+      "loss": 2.5902,
+      "step": 607
+    },
+    {
+      "epoch": 0.19471577261809447,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.8651635804618046e-05,
+      "loss": 2.3034,
+      "step": 608
+    },
+    {
+      "epoch": 0.19503602882305845,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.8642562636002546e-05,
+      "loss": 2.4728,
+      "step": 609
+    },
+    {
+      "epoch": 0.1953562850280224,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.8633459894701186e-05,
+      "loss": 2.1875,
+      "step": 610
+    },
+    {
+      "epoch": 0.1956765412329864,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.8624327592099884e-05,
+      "loss": 2.2461,
+      "step": 611
+    },
+    {
+      "epoch": 0.19599679743795037,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.861516573962156e-05,
+      "loss": 1.8182,
+      "step": 612
+    },
+    {
+      "epoch": 0.19631705364291432,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.860597434872606e-05,
+      "loss": 1.7308,
+      "step": 613
+    },
+    {
+      "epoch": 0.1966373098478783,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.859675343091021e-05,
+      "loss": 2.5587,
+      "step": 614
+    },
+    {
+      "epoch": 0.1969575660528423,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.858750299770774e-05,
+      "loss": 2.2981,
+      "step": 615
+    },
+    {
+      "epoch": 0.19727782225780624,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.8578223060689346e-05,
+      "loss": 2.356,
+      "step": 616
+    },
+    {
+      "epoch": 0.19759807846277022,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.8568913631462565e-05,
+      "loss": 1.9542,
+      "step": 617
+    },
+    {
+      "epoch": 0.19791833466773417,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.855957472167187e-05,
+      "loss": 2.3769,
+      "step": 618
+    },
+    {
+      "epoch": 0.19823859087269816,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.85502063429986e-05,
+      "loss": 2.4044,
+      "step": 619
+    },
+    {
+      "epoch": 0.19855884707766214,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.8540808507160926e-05,
+      "loss": 2.4242,
+      "step": 620
+    },
+    {
+      "epoch": 0.1988791032826261,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.85313812259139e-05,
+      "loss": 2.3783,
+      "step": 621
+    },
+    {
+      "epoch": 0.19919935948759007,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.85219245110494e-05,
+      "loss": 2.5425,
+      "step": 622
+    },
+    {
+      "epoch": 0.19951961569255405,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.8512438374396095e-05,
+      "loss": 2.3095,
+      "step": 623
+    },
+    {
+      "epoch": 0.199839871897518,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.850292282781949e-05,
+      "loss": 2.5806,
+      "step": 624
+    },
+    {
+      "epoch": 0.200160128102482,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.849337788322185e-05,
+      "loss": 2.4287,
+      "step": 625
+    },
+    {
+      "epoch": 0.20048038430744597,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.848380355254222e-05,
+      "loss": 2.6023,
+      "step": 626
+    },
+    {
+      "epoch": 0.20080064051240992,
+      "grad_norm": 0.625,
+      "learning_rate": 4.847419984775641e-05,
+      "loss": 2.0865,
+      "step": 627
+    },
+    {
+      "epoch": 0.2011208967173739,
+      "grad_norm": 1.6484375,
+      "learning_rate": 4.8464566780876954e-05,
+      "loss": 1.8258,
+      "step": 628
+    },
+    {
+      "epoch": 0.20144115292233786,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.845490436395314e-05,
+      "loss": 2.2667,
+      "step": 629
+    },
+    {
+      "epoch": 0.20176140912730184,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.8445212609070934e-05,
+      "loss": 2.394,
+      "step": 630
+    },
+    {
+      "epoch": 0.20208166533226582,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.8435491528353026e-05,
+      "loss": 2.6123,
+      "step": 631
+    },
+    {
+      "epoch": 0.20240192153722977,
+      "grad_norm": 1.25,
+      "learning_rate": 4.8425741133958774e-05,
+      "loss": 2.2056,
+      "step": 632
+    },
+    {
+      "epoch": 0.20272217774219375,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.841596143808421e-05,
+      "loss": 2.6455,
+      "step": 633
+    },
+    {
+      "epoch": 0.20304243394715774,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.840615245296201e-05,
+      "loss": 2.1008,
+      "step": 634
+    },
+    {
+      "epoch": 0.2033626901521217,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.8396314190861495e-05,
+      "loss": 1.9474,
+      "step": 635
+    },
+    {
+      "epoch": 0.20368294635708567,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.838644666408858e-05,
+      "loss": 2.1314,
+      "step": 636
+    },
+    {
+      "epoch": 0.20400320256204965,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.8376549884985825e-05,
+      "loss": 2.4542,
+      "step": 637
+    },
+    {
+      "epoch": 0.2043234587670136,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.836662386593235e-05,
+      "loss": 2.2233,
+      "step": 638
+    },
+    {
+      "epoch": 0.20464371497197759,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.835666861934387e-05,
+      "loss": 1.8406,
+      "step": 639
+    },
+    {
+      "epoch": 0.20496397117694154,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.834668415767263e-05,
+      "loss": 2.3927,
+      "step": 640
+    },
+    {
+      "epoch": 0.20528422738190552,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.833667049340744e-05,
+      "loss": 2.4086,
+      "step": 641
+    },
+    {
+      "epoch": 0.2056044835868695,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.832662763907363e-05,
+      "loss": 1.7985,
+      "step": 642
+    },
+    {
+      "epoch": 0.20592473979183346,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.831655560723306e-05,
+      "loss": 2.2216,
+      "step": 643
+    },
+    {
+      "epoch": 0.20624499599679744,
+      "grad_norm": 1.3515625,
+      "learning_rate": 4.8306454410484055e-05,
+      "loss": 2.4505,
+      "step": 644
+    },
+    {
+      "epoch": 0.20656525220176142,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.829632406146143e-05,
+      "loss": 2.1029,
+      "step": 645
+    },
+    {
+      "epoch": 0.20688550840672537,
+      "grad_norm": 0.5,
+      "learning_rate": 4.828616457283648e-05,
+      "loss": 2.4966,
+      "step": 646
+    },
+    {
+      "epoch": 0.20720576461168935,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.8275975957316943e-05,
+      "loss": 1.7865,
+      "step": 647
+    },
+    {
+      "epoch": 0.20752602081665333,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.826575822764696e-05,
+      "loss": 1.674,
+      "step": 648
+    },
+    {
+      "epoch": 0.2078462770216173,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.825551139660715e-05,
+      "loss": 1.9932,
+      "step": 649
+    },
+    {
+      "epoch": 0.20816653322658127,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.824523547701445e-05,
+      "loss": 2.18,
+      "step": 650
+    },
+    {
+      "epoch": 0.20848678943154525,
+      "grad_norm": 1.7421875,
+      "learning_rate": 4.8234930481722264e-05,
+      "loss": 1.9729,
+      "step": 651
+    },
+    {
+      "epoch": 0.2088070456365092,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.8224596423620315e-05,
+      "loss": 2.0876,
+      "step": 652
+    },
+    {
+      "epoch": 0.20912730184147318,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.821423331563469e-05,
+      "loss": 2.0135,
+      "step": 653
+    },
+    {
+      "epoch": 0.20944755804643714,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.820384117072781e-05,
+      "loss": 2.3225,
+      "step": 654
+    },
+    {
+      "epoch": 0.20976781425140112,
+      "grad_norm": 1.65625,
+      "learning_rate": 4.819342000189844e-05,
+      "loss": 1.6923,
+      "step": 655
+    },
+    {
+      "epoch": 0.2100880704563651,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.818296982218161e-05,
+      "loss": 2.1741,
+      "step": 656
+    },
+    {
+      "epoch": 0.21040832666132905,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.817249064464865e-05,
+      "loss": 2.285,
+      "step": 657
+    },
+    {
+      "epoch": 0.21072858286629303,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.816198248240718e-05,
+      "loss": 2.1315,
+      "step": 658
+    },
+    {
+      "epoch": 0.21104883907125702,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.815144534860106e-05,
+      "loss": 1.9737,
+      "step": 659
+    },
+    {
+      "epoch": 0.21136909527622097,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.81408792564104e-05,
+      "loss": 1.6167,
+      "step": 660
+    },
+    {
+      "epoch": 0.21168935148118495,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.8130284219051494e-05,
+      "loss": 2.4866,
+      "step": 661
+    },
+    {
+      "epoch": 0.21200960768614893,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.81196602497769e-05,
+      "loss": 2.2202,
+      "step": 662
+    },
+    {
+      "epoch": 0.21232986389111289,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.810900736187531e-05,
+      "loss": 2.1335,
+      "step": 663
+    },
+    {
+      "epoch": 0.21265012009607687,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.809832556867164e-05,
+      "loss": 2.0997,
+      "step": 664
+    },
+    {
+      "epoch": 0.21297037630104082,
+      "grad_norm": 1.9765625,
+      "learning_rate": 4.80876148835269e-05,
+      "loss": 2.5337,
+      "step": 665
+    },
+    {
+      "epoch": 0.2132906325060048,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.80768753198383e-05,
+      "loss": 2.4361,
+      "step": 666
+    },
+    {
+      "epoch": 0.21361088871096878,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.8066106891039135e-05,
+      "loss": 2.7403,
+      "step": 667
+    },
+    {
+      "epoch": 0.21393114491593274,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.805530961059881e-05,
+      "loss": 1.8824,
+      "step": 668
+    },
+    {
+      "epoch": 0.21425140112089672,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.804448349202283e-05,
+      "loss": 1.9837,
+      "step": 669
+    },
+    {
+      "epoch": 0.2145716573258607,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.803362854885276e-05,
+      "loss": 2.0329,
+      "step": 670
+    },
+    {
+      "epoch": 0.21489191353082465,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.802274479466623e-05,
+      "loss": 2.314,
+      "step": 671
+    },
+    {
+      "epoch": 0.21521216973578863,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.8011832243076895e-05,
+      "loss": 2.324,
+      "step": 672
+    },
+    {
+      "epoch": 0.21553242594075261,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.8000890907734434e-05,
+      "loss": 1.7424,
+      "step": 673
+    },
+    {
+      "epoch": 0.21585268214571657,
+      "grad_norm": 0.37109375,
+      "learning_rate": 4.7989920802324537e-05,
+      "loss": 1.981,
+      "step": 674
+    },
+    {
+      "epoch": 0.21617293835068055,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.797892194056889e-05,
+      "loss": 1.6027,
+      "step": 675
+    },
+    {
+      "epoch": 0.2164931945556445,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.79678943362251e-05,
+      "loss": 2.5835,
+      "step": 676
+    },
+    {
+      "epoch": 0.21681345076060848,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.79568380030868e-05,
+      "loss": 2.4851,
+      "step": 677
+    },
+    {
+      "epoch": 0.21713370696557246,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.794575295498348e-05,
+      "loss": 2.6476,
+      "step": 678
+    },
+    {
+      "epoch": 0.21745396317053642,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.793463920578061e-05,
+      "loss": 2.6871,
+      "step": 679
+    },
+    {
+      "epoch": 0.2177742193755004,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.792349676937953e-05,
+      "loss": 2.124,
+      "step": 680
+    },
+    {
+      "epoch": 0.21809447558046438,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.791232565971745e-05,
+      "loss": 2.3615,
+      "step": 681
+    },
+    {
+      "epoch": 0.21841473178542833,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.7901125890767484e-05,
+      "loss": 1.8175,
+      "step": 682
+    },
+    {
+      "epoch": 0.21873498799039232,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.788989747653857e-05,
+      "loss": 2.2353,
+      "step": 683
+    },
+    {
+      "epoch": 0.2190552441953563,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.787864043107546e-05,
+      "loss": 1.9429,
+      "step": 684
+    },
+    {
+      "epoch": 0.21937550040032025,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.786735476845876e-05,
+      "loss": 2.1238,
+      "step": 685
+    },
+    {
+      "epoch": 0.21969575660528423,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.7856040502804836e-05,
+      "loss": 1.7298,
+      "step": 686
+    },
+    {
+      "epoch": 0.22001601281024818,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.784469764826585e-05,
+      "loss": 1.9109,
+      "step": 687
+    },
+    {
+      "epoch": 0.22033626901521217,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.783332621902971e-05,
+      "loss": 2.7153,
+      "step": 688
+    },
+    {
+      "epoch": 0.22065652522017615,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.782192622932008e-05,
+      "loss": 2.3001,
+      "step": 689
+    },
+    {
+      "epoch": 0.2209767814251401,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.7810497693396327e-05,
+      "loss": 2.0766,
+      "step": 690
+    },
+    {
+      "epoch": 0.22129703763010408,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.779904062555356e-05,
+      "loss": 2.2382,
+      "step": 691
+    },
+    {
+      "epoch": 0.22161729383506806,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.7787555040122534e-05,
+      "loss": 2.6912,
+      "step": 692
+    },
+    {
+      "epoch": 0.22193755004003202,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.777604095146971e-05,
+      "loss": 2.3497,
+      "step": 693
+    },
+    {
+      "epoch": 0.222257806244996,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.7764498373997194e-05,
+      "loss": 2.4798,
+      "step": 694
+    },
+    {
+      "epoch": 0.22257806244995998,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.775292732214269e-05,
+      "loss": 2.6314,
+      "step": 695
+    },
+    {
+      "epoch": 0.22289831865492393,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.774132781037958e-05,
+      "loss": 2.2279,
+      "step": 696
+    },
+    {
+      "epoch": 0.2232185748598879,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.7729699853216784e-05,
+      "loss": 2.6009,
+      "step": 697
+    },
+    {
+      "epoch": 0.2235388310648519,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.771804346519886e-05,
+      "loss": 2.1287,
+      "step": 698
+    },
+    {
+      "epoch": 0.22385908726981585,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.770635866090587e-05,
+      "loss": 2.0747,
+      "step": 699
+    },
+    {
+      "epoch": 0.22417934347477983,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.769464545495347e-05,
+      "loss": 2.1692,
+      "step": 700
+    },
+    {
+      "epoch": 0.22449959967974378,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.768290386199281e-05,
+      "loss": 2.1745,
+      "step": 701
+    },
+    {
+      "epoch": 0.22481985588470776,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.767113389671055e-05,
+      "loss": 2.4571,
+      "step": 702
+    },
+    {
+      "epoch": 0.22514011208967175,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.7659335573828855e-05,
+      "loss": 2.1443,
+      "step": 703
+    },
+    {
+      "epoch": 0.2254603682946357,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.7647508908105354e-05,
+      "loss": 2.645,
+      "step": 704
+    },
+    {
+      "epoch": 0.22578062449959968,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.7635653914333123e-05,
+      "loss": 2.5967,
+      "step": 705
+    },
+    {
+      "epoch": 0.22610088070456366,
+      "grad_norm": 0.3828125,
+      "learning_rate": 4.7623770607340656e-05,
+      "loss": 1.8238,
+      "step": 706
+    },
+    {
+      "epoch": 0.22642113690952761,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.761185900199191e-05,
+      "loss": 2.1985,
+      "step": 707
+    },
+    {
+      "epoch": 0.2267413931144916,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.7599919113186184e-05,
+      "loss": 1.9056,
+      "step": 708
+    },
+    {
+      "epoch": 0.22706164931945558,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.7587950955858184e-05,
+      "loss": 2.0725,
+      "step": 709
+    },
+    {
+      "epoch": 0.22738190552441953,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.757595454497798e-05,
+      "loss": 2.1442,
+      "step": 710
+    },
+    {
+      "epoch": 0.2277021617293835,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.756392989555094e-05,
+      "loss": 1.8137,
+      "step": 711
+    },
+    {
+      "epoch": 0.22802241793434747,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.755187702261782e-05,
+      "loss": 2.4148,
+      "step": 712
+    },
+    {
+      "epoch": 0.22834267413931145,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.753979594125463e-05,
+      "loss": 2.579,
+      "step": 713
+    },
+    {
+      "epoch": 0.22866293034427543,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.752768666657267e-05,
+      "loss": 2.4936,
+      "step": 714
+    },
+    {
+      "epoch": 0.22898318654923938,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.751554921371852e-05,
+      "loss": 2.104,
+      "step": 715
+    },
+    {
+      "epoch": 0.22930344275420336,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.7503383597873994e-05,
+      "loss": 2.0582,
+      "step": 716
+    },
+    {
+      "epoch": 0.22962369895916734,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.7491189834256147e-05,
+      "loss": 2.2863,
+      "step": 717
+    },
+    {
+      "epoch": 0.2299439551641313,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.747896793811721e-05,
+      "loss": 2.4744,
+      "step": 718
+    },
+    {
+      "epoch": 0.23026421136909528,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.746671792474465e-05,
+      "loss": 2.4798,
+      "step": 719
+    },
+    {
+      "epoch": 0.23058446757405926,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.7454439809461064e-05,
+      "loss": 2.4989,
+      "step": 720
+    },
+    {
+      "epoch": 0.2309047237790232,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.7442133607624216e-05,
+      "loss": 2.4011,
+      "step": 721
+    },
+    {
+      "epoch": 0.2312249799839872,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.7429799334627e-05,
+      "loss": 1.8769,
+      "step": 722
+    },
+    {
+      "epoch": 0.23154523618895115,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.7417437005897415e-05,
+      "loss": 2.1301,
+      "step": 723
+    },
+    {
+      "epoch": 0.23186549239391513,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.740504663689857e-05,
+      "loss": 1.8229,
+      "step": 724
+    },
+    {
+      "epoch": 0.2321857485988791,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.739262824312863e-05,
+      "loss": 2.2332,
+      "step": 725
+    },
+    {
+      "epoch": 0.23250600480384306,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.7380181840120825e-05,
+      "loss": 2.4415,
+      "step": 726
+    },
+    {
+      "epoch": 0.23282626100880705,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.736770744344341e-05,
+      "loss": 2.4059,
+      "step": 727
+    },
+    {
+      "epoch": 0.23314651721377103,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.735520506869966e-05,
+      "loss": 2.3278,
+      "step": 728
+    },
+    {
+      "epoch": 0.23346677341873498,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.734267473152785e-05,
+      "loss": 2.3155,
+      "step": 729
+    },
+    {
+      "epoch": 0.23378702962369896,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.7330116447601225e-05,
+      "loss": 1.8021,
+      "step": 730
+    },
+    {
+      "epoch": 0.23410728582866294,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.731753023262798e-05,
+      "loss": 2.0426,
+      "step": 731
+    },
+    {
+      "epoch": 0.2344275420336269,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.730491610235128e-05,
+      "loss": 1.5206,
+      "step": 732
+    },
+    {
+      "epoch": 0.23474779823859088,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.729227407254916e-05,
+      "loss": 2.1711,
+      "step": 733
+    },
+    {
+      "epoch": 0.23506805444355483,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.727960415903458e-05,
+      "loss": 2.246,
+      "step": 734
+    },
+    {
+      "epoch": 0.2353883106485188,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.7266906377655375e-05,
+      "loss": 1.8435,
+      "step": 735
+    },
+    {
+      "epoch": 0.2357085668534828,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.725418074429423e-05,
+      "loss": 1.9849,
+      "step": 736
+    },
+    {
+      "epoch": 0.23602882305844675,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.724142727486869e-05,
+      "loss": 2.6527,
+      "step": 737
+    },
+    {
+      "epoch": 0.23634907926341073,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.722864598533108e-05,
+      "loss": 2.5422,
+      "step": 738
+    },
+    {
+      "epoch": 0.2366693354683747,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.721583689166855e-05,
+      "loss": 2.0435,
+      "step": 739
+    },
+    {
+      "epoch": 0.23698959167333866,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.720300000990302e-05,
+      "loss": 1.571,
+      "step": 740
+    },
+    {
+      "epoch": 0.23730984787830264,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.719013535609119e-05,
+      "loss": 1.9472,
+      "step": 741
+    },
+    {
+      "epoch": 0.23763010408326662,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.717724294632446e-05,
+      "loss": 2.4293,
+      "step": 742
+    },
+    {
+      "epoch": 0.23795036028823058,
+      "grad_norm": 0.5,
+      "learning_rate": 4.716432279672897e-05,
+      "loss": 2.4161,
+      "step": 743
+    },
+    {
+      "epoch": 0.23827061649319456,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.7151374923465554e-05,
+      "loss": 2.4954,
+      "step": 744
+    },
+    {
+      "epoch": 0.23859087269815854,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.7138399342729723e-05,
+      "loss": 2.3457,
+      "step": 745
+    },
+    {
+      "epoch": 0.2389111289031225,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.712539607075164e-05,
+      "loss": 2.2736,
+      "step": 746
+    },
+    {
+      "epoch": 0.23923138510808648,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.711236512379612e-05,
+      "loss": 2.1532,
+      "step": 747
+    },
+    {
+      "epoch": 0.23955164131305043,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.709930651816258e-05,
+      "loss": 2.0565,
+      "step": 748
+    },
+    {
+      "epoch": 0.2398718975180144,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.708622027018504e-05,
+      "loss": 2.5243,
+      "step": 749
+    },
+    {
+      "epoch": 0.2401921537229784,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.70731063962321e-05,
+      "loss": 2.1374,
+      "step": 750
+    },
+    {
+      "epoch": 0.24051240992794234,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.705996491270689e-05,
+      "loss": 2.0636,
+      "step": 751
+    },
+    {
+      "epoch": 0.24083266613290633,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.704679583604711e-05,
+      "loss": 2.4744,
+      "step": 752
+    },
+    {
+      "epoch": 0.2411529223378703,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.703359918272495e-05,
+      "loss": 2.2096,
+      "step": 753
+    },
+    {
+      "epoch": 0.24147317854283426,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.7020374969247106e-05,
+      "loss": 2.576,
+      "step": 754
+    },
+    {
+      "epoch": 0.24179343474779824,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.700712321215474e-05,
+      "loss": 2.3694,
+      "step": 755
+    },
+    {
+      "epoch": 0.24211369095276222,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.699384392802346e-05,
+      "loss": 2.4635,
+      "step": 756
+    },
+    {
+      "epoch": 0.24243394715772618,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.698053713346334e-05,
+      "loss": 2.7555,
+      "step": 757
+    },
+    {
+      "epoch": 0.24275420336269016,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.696720284511881e-05,
+      "loss": 2.4657,
+      "step": 758
+    },
+    {
+      "epoch": 0.2430744595676541,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.6953841079668724e-05,
+      "loss": 1.6963,
+      "step": 759
+    },
+    {
+      "epoch": 0.2433947157726181,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.6940451853826316e-05,
+      "loss": 2.1481,
+      "step": 760
+    },
+    {
+      "epoch": 0.24371497197758207,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.692703518433913e-05,
+      "loss": 2.1694,
+      "step": 761
+    },
+    {
+      "epoch": 0.24403522818254603,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.6913591087989084e-05,
+      "loss": 2.0632,
+      "step": 762
+    },
+    {
+      "epoch": 0.24435548438751,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.690011958159235e-05,
+      "loss": 2.0139,
+      "step": 763
+    },
+    {
+      "epoch": 0.244675740592474,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.6886620681999434e-05,
+      "loss": 2.0248,
+      "step": 764
+    },
+    {
+      "epoch": 0.24499599679743794,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.6873094406095066e-05,
+      "loss": 2.3785,
+      "step": 765
+    },
+    {
+      "epoch": 0.24531625300240192,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.685954077079825e-05,
+      "loss": 2.6487,
+      "step": 766
+    },
+    {
+      "epoch": 0.2456365092073659,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.68459597930622e-05,
+      "loss": 2.0873,
+      "step": 767
+    },
+    {
+      "epoch": 0.24595676541232986,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.683235148987432e-05,
+      "loss": 2.0824,
+      "step": 768
+    },
+    {
+      "epoch": 0.24627702161729384,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.68187158782562e-05,
+      "loss": 2.6039,
+      "step": 769
+    },
+    {
+      "epoch": 0.2465972778222578,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.680505297526361e-05,
+      "loss": 2.1172,
+      "step": 770
+    },
+    {
+      "epoch": 0.24691753402722177,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.679136279798642e-05,
+      "loss": 2.3141,
+      "step": 771
+    },
+    {
+      "epoch": 0.24723779023218576,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.677764536354864e-05,
+      "loss": 2.0481,
+      "step": 772
+    },
+    {
+      "epoch": 0.2475580464371497,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.6763900689108365e-05,
+      "loss": 1.7297,
+      "step": 773
+    },
+    {
+      "epoch": 0.2478783026421137,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.675012879185777e-05,
+      "loss": 2.4405,
+      "step": 774
+    },
+    {
+      "epoch": 0.24819855884707767,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.673632968902306e-05,
+      "loss": 2.1231,
+      "step": 775
+    },
+    {
+      "epoch": 0.24851881505204163,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.67225033978645e-05,
+      "loss": 1.7979,
+      "step": 776
+    },
+    {
+      "epoch": 0.2488390712570056,
+      "grad_norm": 0.5,
+      "learning_rate": 4.670864993567635e-05,
+      "loss": 1.7725,
+      "step": 777
+    },
+    {
+      "epoch": 0.2491593274619696,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.6694769319786834e-05,
+      "loss": 2.0591,
+      "step": 778
+    },
+    {
+      "epoch": 0.24947958366693354,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.668086156755818e-05,
+      "loss": 2.1371,
+      "step": 779
+    },
+    {
+      "epoch": 0.24979983987189752,
+      "grad_norm": 0.37109375,
+      "learning_rate": 4.666692669638653e-05,
+      "loss": 2.1823,
+      "step": 780
+    },
+    {
+      "epoch": 0.2501200960768615,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.665296472370195e-05,
+      "loss": 1.8962,
+      "step": 781
+    },
+    {
+      "epoch": 0.2504403522818255,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.663897566696843e-05,
+      "loss": 2.6859,
+      "step": 782
+    },
+    {
+      "epoch": 0.25076060848678944,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.66249595436838e-05,
+      "loss": 2.4368,
+      "step": 783
+    },
+    {
+      "epoch": 0.2510808646917534,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.661091637137977e-05,
+      "loss": 1.895,
+      "step": 784
+    },
+    {
+      "epoch": 0.2514011208967174,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.659684616762188e-05,
+      "loss": 2.1725,
+      "step": 785
+    },
+    {
+      "epoch": 0.25172137710168135,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.6582748950009475e-05,
+      "loss": 2.7823,
+      "step": 786
+    },
+    {
+      "epoch": 0.2520416333066453,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.65686247361757e-05,
+      "loss": 2.3225,
+      "step": 787
+    },
+    {
+      "epoch": 0.25236188951160926,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.655447354378745e-05,
+      "loss": 2.1723,
+      "step": 788
+    },
+    {
+      "epoch": 0.25268214571657327,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.654029539054539e-05,
+      "loss": 1.9019,
+      "step": 789
+    },
+    {
+      "epoch": 0.2530024019215372,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.652609029418389e-05,
+      "loss": 2.178,
+      "step": 790
+    },
+    {
+      "epoch": 0.2533226581265012,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.651185827247103e-05,
+      "loss": 1.4949,
+      "step": 791
+    },
+    {
+      "epoch": 0.2536429143314652,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.649759934320855e-05,
+      "loss": 2.6352,
+      "step": 792
+    },
+    {
+      "epoch": 0.25396317053642914,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.6483313524231874e-05,
+      "loss": 2.1695,
+      "step": 793
+    },
+    {
+      "epoch": 0.2542834267413931,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.646900083341005e-05,
+      "loss": 2.4847,
+      "step": 794
+    },
+    {
+      "epoch": 0.2546036829463571,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.645466128864573e-05,
+      "loss": 2.0651,
+      "step": 795
+    },
+    {
+      "epoch": 0.25492393915132106,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.644029490787517e-05,
+      "loss": 2.7276,
+      "step": 796
+    },
+    {
+      "epoch": 0.255244195356285,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.642590170906816e-05,
+      "loss": 2.5108,
+      "step": 797
+    },
+    {
+      "epoch": 0.255564451561249,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.6411481710228096e-05,
+      "loss": 2.3693,
+      "step": 798
+    },
+    {
+      "epoch": 0.25588470776621297,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.6397034929391834e-05,
+      "loss": 2.333,
+      "step": 799
+    },
+    {
+      "epoch": 0.2562049639711769,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.638256138462975e-05,
+      "loss": 2.4582,
+      "step": 800
+    },
+    {
+      "epoch": 0.25652522017614093,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.636806109404573e-05,
+      "loss": 2.5472,
+      "step": 801
+    },
+    {
+      "epoch": 0.2568454763811049,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.635353407577706e-05,
+      "loss": 2.4438,
+      "step": 802
+    },
+    {
+      "epoch": 0.25716573258606884,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.633898034799449e-05,
+      "loss": 2.3358,
+      "step": 803
+    },
+    {
+      "epoch": 0.25748598879103285,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.632439992890218e-05,
+      "loss": 2.4705,
+      "step": 804
+    },
+    {
+      "epoch": 0.2578062449959968,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.630979283673766e-05,
+      "loss": 2.163,
+      "step": 805
+    },
+    {
+      "epoch": 0.25812650120096076,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.629515908977184e-05,
+      "loss": 1.7262,
+      "step": 806
+    },
+    {
+      "epoch": 0.25844675740592477,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.628049870630896e-05,
+      "loss": 2.2349,
+      "step": 807
+    },
+    {
+      "epoch": 0.2587670136108887,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.626581170468658e-05,
+      "loss": 2.3561,
+      "step": 808
+    },
+    {
+      "epoch": 0.25908726981585267,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.625109810327556e-05,
+      "loss": 2.5208,
+      "step": 809
+    },
+    {
+      "epoch": 0.2594075260208166,
+      "grad_norm": 1.953125,
+      "learning_rate": 4.6236357920480013e-05,
+      "loss": 2.0921,
+      "step": 810
+    },
+    {
+      "epoch": 0.25972778222578063,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.6221591174737314e-05,
+      "loss": 2.2984,
+      "step": 811
+    },
+    {
+      "epoch": 0.2600480384307446,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.620679788451808e-05,
+      "loss": 1.6466,
+      "step": 812
+    },
+    {
+      "epoch": 0.26036829463570854,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.61919780683261e-05,
+      "loss": 2.3331,
+      "step": 813
+    },
+    {
+      "epoch": 0.26068855084067255,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.6177131744698364e-05,
+      "loss": 2.313,
+      "step": 814
+    },
+    {
+      "epoch": 0.2610088070456365,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.616225893220502e-05,
+      "loss": 2.4781,
+      "step": 815
+    },
+    {
+      "epoch": 0.26132906325060046,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.6147359649449305e-05,
+      "loss": 2.2835,
+      "step": 816
+    },
+    {
+      "epoch": 0.26164931945556447,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.613243391506764e-05,
+      "loss": 2.2348,
+      "step": 817
+    },
+    {
+      "epoch": 0.2619695756605284,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.6117481747729476e-05,
+      "loss": 2.4842,
+      "step": 818
+    },
+    {
+      "epoch": 0.2622898318654924,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.6102503166137336e-05,
+      "loss": 2.0446,
+      "step": 819
+    },
+    {
+      "epoch": 0.2626100880704564,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.608749818902679e-05,
+      "loss": 1.8785,
+      "step": 820
+    },
+    {
+      "epoch": 0.26293034427542034,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.607246683516644e-05,
+      "loss": 2.5239,
+      "step": 821
+    },
+    {
+      "epoch": 0.2632506004803843,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.605740912335786e-05,
+      "loss": 2.61,
+      "step": 822
+    },
+    {
+      "epoch": 0.2635708566853483,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.604232507243559e-05,
+      "loss": 1.8585,
+      "step": 823
+    },
+    {
+      "epoch": 0.26389111289031225,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.6027214701267126e-05,
+      "loss": 2.3605,
+      "step": 824
+    },
+    {
+      "epoch": 0.2642113690952762,
+      "grad_norm": 0.3828125,
+      "learning_rate": 4.6012078028752885e-05,
+      "loss": 2.2571,
+      "step": 825
+    },
+    {
+      "epoch": 0.2645316253002402,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.5996915073826187e-05,
+      "loss": 1.8091,
+      "step": 826
+    },
+    {
+      "epoch": 0.26485188150520417,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.598172585545321e-05,
+      "loss": 2.3873,
+      "step": 827
+    },
+    {
+      "epoch": 0.2651721377101681,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.5966510392633e-05,
+      "loss": 2.171,
+      "step": 828
+    },
+    {
+      "epoch": 0.26549239391513213,
+      "grad_norm": 2.75,
+      "learning_rate": 4.595126870439742e-05,
+      "loss": 2.4189,
+      "step": 829
+    },
+    {
+      "epoch": 0.2658126501200961,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.593600080981114e-05,
+      "loss": 2.3643,
+      "step": 830
+    },
+    {
+      "epoch": 0.26613290632506004,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.592070672797161e-05,
+      "loss": 2.1314,
+      "step": 831
+    },
+    {
+      "epoch": 0.26645316253002405,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.590538647800904e-05,
+      "loss": 2.5095,
+      "step": 832
+    },
+    {
+      "epoch": 0.266773418734988,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.5890040079086374e-05,
+      "loss": 2.1051,
+      "step": 833
+    },
+    {
+      "epoch": 0.26709367493995195,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.587466755039924e-05,
+      "loss": 2.6451,
+      "step": 834
+    },
+    {
+      "epoch": 0.2674139311449159,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.585926891117597e-05,
+      "loss": 2.3557,
+      "step": 835
+    },
+    {
+      "epoch": 0.2677341873498799,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.584384418067756e-05,
+      "loss": 2.2055,
+      "step": 836
+    },
+    {
+      "epoch": 0.26805444355484387,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.582839337819763e-05,
+      "loss": 1.711,
+      "step": 837
+    },
+    {
+      "epoch": 0.2683746997598078,
+      "grad_norm": 1.6640625,
+      "learning_rate": 4.581291652306241e-05,
+      "loss": 2.7919,
+      "step": 838
+    },
+    {
+      "epoch": 0.26869495596477183,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.579741363463074e-05,
+      "loss": 2.3713,
+      "step": 839
+    },
+    {
+      "epoch": 0.2690152121697358,
+      "grad_norm": 0.625,
+      "learning_rate": 4.578188473229398e-05,
+      "loss": 2.4531,
+      "step": 840
+    },
+    {
+      "epoch": 0.26933546837469974,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.576632983547607e-05,
+      "loss": 2.2186,
+      "step": 841
+    },
+    {
+      "epoch": 0.26965572457966375,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.575074896363345e-05,
+      "loss": 2.3781,
+      "step": 842
+    },
+    {
+      "epoch": 0.2699759807846277,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.573514213625505e-05,
+      "loss": 2.1847,
+      "step": 843
+    },
+    {
+      "epoch": 0.27029623698959165,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.571950937286225e-05,
+      "loss": 2.1921,
+      "step": 844
+    },
+    {
+      "epoch": 0.27061649319455566,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.570385069300891e-05,
+      "loss": 2.6008,
+      "step": 845
+    },
+    {
+      "epoch": 0.2709367493995196,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.568816611628126e-05,
+      "loss": 2.5111,
+      "step": 846
+    },
+    {
+      "epoch": 0.27125700560448357,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.5672455662297966e-05,
+      "loss": 2.4546,
+      "step": 847
+    },
+    {
+      "epoch": 0.2715772618094476,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.565671935071002e-05,
+      "loss": 2.2381,
+      "step": 848
+    },
+    {
+      "epoch": 0.27189751801441153,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.56409572012008e-05,
+      "loss": 2.5955,
+      "step": 849
+    },
+    {
+      "epoch": 0.2722177742193755,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.562516923348597e-05,
+      "loss": 2.2343,
+      "step": 850
+    },
+    {
+      "epoch": 0.2725380304243395,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.56093554673135e-05,
+      "loss": 2.4699,
+      "step": 851
+    },
+    {
+      "epoch": 0.27285828662930345,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.559351592246363e-05,
+      "loss": 2.4249,
+      "step": 852
+    },
+    {
+      "epoch": 0.2731785428342674,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.5577650618748846e-05,
+      "loss": 1.7957,
+      "step": 853
+    },
+    {
+      "epoch": 0.2734987990392314,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.556175957601385e-05,
+      "loss": 2.3653,
+      "step": 854
+    },
+    {
+      "epoch": 0.27381905524419536,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.5545842814135545e-05,
+      "loss": 2.7702,
+      "step": 855
+    },
+    {
+      "epoch": 0.2741393114491593,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.552990035302299e-05,
+      "loss": 2.3475,
+      "step": 856
+    },
+    {
+      "epoch": 0.27445956765412327,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.55139322126174e-05,
+      "loss": 2.1684,
+      "step": 857
+    },
+    {
+      "epoch": 0.2747798238590873,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.549793841289211e-05,
+      "loss": 1.8475,
+      "step": 858
+    },
+    {
+      "epoch": 0.27510008006405123,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.548191897385257e-05,
+      "loss": 2.3642,
+      "step": 859
+    },
+    {
+      "epoch": 0.2754203362690152,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.546587391553624e-05,
+      "loss": 1.9096,
+      "step": 860
+    },
+    {
+      "epoch": 0.2757405924739792,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.54498032580127e-05,
+      "loss": 2.4123,
+      "step": 861
+    },
+    {
+      "epoch": 0.27606084867894315,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.54337070213835e-05,
+      "loss": 2.431,
+      "step": 862
+    },
+    {
+      "epoch": 0.2763811048839071,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.54175852257822e-05,
+      "loss": 2.5574,
+      "step": 863
+    },
+    {
+      "epoch": 0.2767013610888711,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.5401437891374326e-05,
+      "loss": 2.4494,
+      "step": 864
+    },
+    {
+      "epoch": 0.27702161729383507,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.5385265038357363e-05,
+      "loss": 2.2868,
+      "step": 865
+    },
+    {
+      "epoch": 0.277341873498799,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.5369066686960694e-05,
+      "loss": 2.2944,
+      "step": 866
+    },
+    {
+      "epoch": 0.277662129703763,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.535284285744561e-05,
+      "loss": 1.8372,
+      "step": 867
+    },
+    {
+      "epoch": 0.277982385908727,
+      "grad_norm": 0.5,
+      "learning_rate": 4.5336593570105266e-05,
+      "loss": 2.3709,
+      "step": 868
+    },
+    {
+      "epoch": 0.27830264211369093,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.5320318845264664e-05,
+      "loss": 2.1215,
+      "step": 869
+    },
+    {
+      "epoch": 0.27862289831865494,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.530401870328063e-05,
+      "loss": 2.8229,
+      "step": 870
+    },
+    {
+      "epoch": 0.2789431545236189,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.528769316454176e-05,
+      "loss": 2.272,
+      "step": 871
+    },
+    {
+      "epoch": 0.27926341072858285,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.5271342249468424e-05,
+      "loss": 2.1168,
+      "step": 872
+    },
+    {
+      "epoch": 0.27958366693354686,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.525496597851275e-05,
+      "loss": 1.5178,
+      "step": 873
+    },
+    {
+      "epoch": 0.2799039231385108,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.523856437215858e-05,
+      "loss": 1.6794,
+      "step": 874
+    },
+    {
+      "epoch": 0.28022417934347477,
+      "grad_norm": 0.375,
+      "learning_rate": 4.522213745092143e-05,
+      "loss": 2.3376,
+      "step": 875
+    },
+    {
+      "epoch": 0.2805444355484388,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.520568523534848e-05,
+      "loss": 2.4695,
+      "step": 876
+    },
+    {
+      "epoch": 0.28086469175340273,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.5189207746018566e-05,
+      "loss": 1.6355,
+      "step": 877
+    },
+    {
+      "epoch": 0.2811849479583667,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.5172705003542125e-05,
+      "loss": 2.0943,
+      "step": 878
+    },
+    {
+      "epoch": 0.2815052041633307,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.515617702856119e-05,
+      "loss": 1.9592,
+      "step": 879
+    },
+    {
+      "epoch": 0.28182546036829464,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.5139623841749346e-05,
+      "loss": 2.0776,
+      "step": 880
+    },
+    {
+      "epoch": 0.2821457165732586,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.5123045463811706e-05,
+      "loss": 2.2985,
+      "step": 881
+    },
+    {
+      "epoch": 0.28246597277822255,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.510644191548492e-05,
+      "loss": 1.9178,
+      "step": 882
+    },
+    {
+      "epoch": 0.28278622898318656,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.50898132175371e-05,
+      "loss": 2.3869,
+      "step": 883
+    },
+    {
+      "epoch": 0.2831064851881505,
+      "grad_norm": 0.625,
+      "learning_rate": 4.507315939076782e-05,
+      "loss": 1.5066,
+      "step": 884
+    },
+    {
+      "epoch": 0.28342674139311447,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.5056480456008085e-05,
+      "loss": 2.5622,
+      "step": 885
+    },
+    {
+      "epoch": 0.2837469975980785,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.5039776434120306e-05,
+      "loss": 2.3004,
+      "step": 886
+    },
+    {
+      "epoch": 0.28406725380304243,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.502304734599828e-05,
+      "loss": 2.3096,
+      "step": 887
+    },
+    {
+      "epoch": 0.2843875100080064,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.5006293212567164e-05,
+      "loss": 2.1645,
+      "step": 888
+    },
+    {
+      "epoch": 0.2847077662129704,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.4989514054783406e-05,
+      "loss": 2.2932,
+      "step": 889
+    },
+    {
+      "epoch": 0.28502802241793435,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.49727098936348e-05,
+      "loss": 2.1796,
+      "step": 890
+    },
+    {
+      "epoch": 0.2853482786228983,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.495588075014038e-05,
+      "loss": 1.7093,
+      "step": 891
+    },
+    {
+      "epoch": 0.2856685348278623,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.4939026645350454e-05,
+      "loss": 2.5258,
+      "step": 892
+    },
+    {
+      "epoch": 0.28598879103282626,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.4922147600346545e-05,
+      "loss": 1.979,
+      "step": 893
+    },
+    {
+      "epoch": 0.2863090472377902,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.4905243636241354e-05,
+      "loss": 2.0999,
+      "step": 894
+    },
+    {
+      "epoch": 0.2866293034427542,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.4888314774178776e-05,
+      "loss": 2.4738,
+      "step": 895
+    },
+    {
+      "epoch": 0.2869495596477182,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.4871361035333836e-05,
+      "loss": 2.3017,
+      "step": 896
+    },
+    {
+      "epoch": 0.28726981585268213,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.485438244091268e-05,
+      "loss": 2.2876,
+      "step": 897
+    },
+    {
+      "epoch": 0.28759007205764614,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.483737901215254e-05,
+      "loss": 2.1632,
+      "step": 898
+    },
+    {
+      "epoch": 0.2879103282626101,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.482035077032171e-05,
+      "loss": 2.4111,
+      "step": 899
+    },
+    {
+      "epoch": 0.28823058446757405,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.4803297736719534e-05,
+      "loss": 2.6698,
+      "step": 900
+    },
+    {
+      "epoch": 0.28855084067253806,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.478621993267635e-05,
+      "loss": 2.4385,
+      "step": 901
+    },
+    {
+      "epoch": 0.288871096877502,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.476911737955349e-05,
+      "loss": 1.9053,
+      "step": 902
+    },
+    {
+      "epoch": 0.28919135308246596,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.475199009874323e-05,
+      "loss": 1.8812,
+      "step": 903
+    },
+    {
+      "epoch": 0.2895116092874299,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.47348381116688e-05,
+      "loss": 1.812,
+      "step": 904
+    },
+    {
+      "epoch": 0.2898318654923939,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.47176614397843e-05,
+      "loss": 2.3822,
+      "step": 905
+    },
+    {
+      "epoch": 0.2901521216973579,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.470046010457473e-05,
+      "loss": 1.8144,
+      "step": 906
+    },
+    {
+      "epoch": 0.29047237790232183,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.4683234127555944e-05,
+      "loss": 2.3208,
+      "step": 907
+    },
+    {
+      "epoch": 0.29079263410728584,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.46659835302746e-05,
+      "loss": 2.0736,
+      "step": 908
+    },
+    {
+      "epoch": 0.2911128903122498,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.4648708334308156e-05,
+      "loss": 2.6272,
+      "step": 909
+    },
+    {
+      "epoch": 0.29143314651721375,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.463140856126485e-05,
+      "loss": 2.2224,
+      "step": 910
+    },
+    {
+      "epoch": 0.29175340272217776,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.461408423278365e-05,
+      "loss": 2.1467,
+      "step": 911
+    },
+    {
+      "epoch": 0.2920736589271417,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.4596735370534246e-05,
+      "loss": 2.5051,
+      "step": 912
+    },
+    {
+      "epoch": 0.29239391513210566,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.457936199621701e-05,
+      "loss": 2.0413,
+      "step": 913
+    },
+    {
+      "epoch": 0.2927141713370697,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.456196413156297e-05,
+      "loss": 2.1757,
+      "step": 914
+    },
+    {
+      "epoch": 0.2930344275420336,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.45445417983338e-05,
+      "loss": 2.51,
+      "step": 915
+    },
+    {
+      "epoch": 0.2933546837469976,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.4527095018321786e-05,
+      "loss": 2.3372,
+      "step": 916
+    },
+    {
+      "epoch": 0.2936749399519616,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.4509623813349755e-05,
+      "loss": 2.4337,
+      "step": 917
+    },
+    {
+      "epoch": 0.29399519615692554,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.4492128205271134e-05,
+      "loss": 2.0431,
+      "step": 918
+    },
+    {
+      "epoch": 0.2943154523618895,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.4474608215969825e-05,
+      "loss": 2.0454,
+      "step": 919
+    },
+    {
+      "epoch": 0.2946357085668535,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.445706386736027e-05,
+      "loss": 1.6327,
+      "step": 920
+    },
+    {
+      "epoch": 0.29495596477181746,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.4439495181387355e-05,
+      "loss": 2.1912,
+      "step": 921
+    },
+    {
+      "epoch": 0.2952762209767814,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.442190218002643e-05,
+      "loss": 2.5112,
+      "step": 922
+    },
+    {
+      "epoch": 0.2955964771817454,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.440428488528322e-05,
+      "loss": 2.2843,
+      "step": 923
+    },
+    {
+      "epoch": 0.2959167333867094,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.4386643319193874e-05,
+      "loss": 2.0363,
+      "step": 924
+    },
+    {
+      "epoch": 0.2962369895916733,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.436897750382489e-05,
+      "loss": 2.4265,
+      "step": 925
+    },
+    {
+      "epoch": 0.29655724579663734,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.435128746127309e-05,
+      "loss": 2.0594,
+      "step": 926
+    },
+    {
+      "epoch": 0.2968775020016013,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.4333573213665604e-05,
+      "loss": 2.3537,
+      "step": 927
+    },
+    {
+      "epoch": 0.29719775820656524,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.431583478315984e-05,
+      "loss": 2.7564,
+      "step": 928
+    },
+    {
+      "epoch": 0.2975180144115292,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.429807219194345e-05,
+      "loss": 2.0438,
+      "step": 929
+    },
+    {
+      "epoch": 0.2978382706164932,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.428028546223432e-05,
+      "loss": 2.0035,
+      "step": 930
+    },
+    {
+      "epoch": 0.29815852682145716,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.4262474616280506e-05,
+      "loss": 1.9675,
+      "step": 931
+    },
+    {
+      "epoch": 0.2984787830264211,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.4244639676360253e-05,
+      "loss": 2.2379,
+      "step": 932
+    },
+    {
+      "epoch": 0.2987990392313851,
+      "grad_norm": 1.8359375,
+      "learning_rate": 4.422678066478192e-05,
+      "loss": 1.8438,
+      "step": 933
+    },
+    {
+      "epoch": 0.2991192954363491,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.4208897603884016e-05,
+      "loss": 2.166,
+      "step": 934
+    },
+    {
+      "epoch": 0.29943955164131303,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.419099051603507e-05,
+      "loss": 2.4782,
+      "step": 935
+    },
+    {
+      "epoch": 0.29975980784627704,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.417305942363371e-05,
+      "loss": 1.9755,
+      "step": 936
+    },
+    {
+      "epoch": 0.300080064051241,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.415510434910859e-05,
+      "loss": 2.6677,
+      "step": 937
+    },
+    {
+      "epoch": 0.30040032025620494,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.413712531491833e-05,
+      "loss": 2.0538,
+      "step": 938
+    },
+    {
+      "epoch": 0.30072057646116895,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.411912234355156e-05,
+      "loss": 2.4744,
+      "step": 939
+    },
+    {
+      "epoch": 0.3010408326661329,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.41010954575268e-05,
+      "loss": 2.4799,
+      "step": 940
+    },
+    {
+      "epoch": 0.30136108887109686,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.408304467939254e-05,
+      "loss": 2.5785,
+      "step": 941
+    },
+    {
+      "epoch": 0.30168134507606087,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.4064970031727114e-05,
+      "loss": 2.2337,
+      "step": 942
+    },
+    {
+      "epoch": 0.3020016012810248,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.4046871537138734e-05,
+      "loss": 2.5195,
+      "step": 943
+    },
+    {
+      "epoch": 0.3023218574859888,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.402874921826542e-05,
+      "loss": 2.3741,
+      "step": 944
+    },
+    {
+      "epoch": 0.3026421136909528,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.4010603097775015e-05,
+      "loss": 2.4184,
+      "step": 945
+    },
+    {
+      "epoch": 0.30296236989591674,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.399243319836512e-05,
+      "loss": 2.1592,
+      "step": 946
+    },
+    {
+      "epoch": 0.3032826261008807,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.397423954276308e-05,
+      "loss": 1.8628,
+      "step": 947
+    },
+    {
+      "epoch": 0.3036028823058447,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.395602215372596e-05,
+      "loss": 1.8973,
+      "step": 948
+    },
+    {
+      "epoch": 0.30392313851080865,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.3937781054040505e-05,
+      "loss": 2.1797,
+      "step": 949
+    },
+    {
+      "epoch": 0.3042433947157726,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.391951626652312e-05,
+      "loss": 2.0715,
+      "step": 950
+    },
+    {
+      "epoch": 0.30456365092073656,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.3901227814019845e-05,
+      "loss": 1.859,
+      "step": 951
+    },
+    {
+      "epoch": 0.30488390712570057,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.388291571940631e-05,
+      "loss": 2.4701,
+      "step": 952
+    },
+    {
+      "epoch": 0.3052041633306645,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.386458000558773e-05,
+      "loss": 2.0972,
+      "step": 953
+    },
+    {
+      "epoch": 0.3055244195356285,
+      "grad_norm": 8.5,
+      "learning_rate": 4.384622069549886e-05,
+      "loss": 1.9298,
+      "step": 954
+    },
+    {
+      "epoch": 0.3058446757405925,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.382783781210395e-05,
+      "loss": 2.1858,
+      "step": 955
+    },
+    {
+      "epoch": 0.30616493194555644,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.3809431378396776e-05,
+      "loss": 2.1796,
+      "step": 956
+    },
+    {
+      "epoch": 0.3064851881505204,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.379100141740053e-05,
+      "loss": 2.071,
+      "step": 957
+    },
+    {
+      "epoch": 0.3068054443554844,
+      "grad_norm": 1.8359375,
+      "learning_rate": 4.377254795216785e-05,
+      "loss": 2.5639,
+      "step": 958
+    },
+    {
+      "epoch": 0.30712570056044836,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.3754071005780785e-05,
+      "loss": 2.4597,
+      "step": 959
+    },
+    {
+      "epoch": 0.3074459567654123,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.3735570601350736e-05,
+      "loss": 2.6462,
+      "step": 960
+    },
+    {
+      "epoch": 0.3077662129703763,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.3717046762018454e-05,
+      "loss": 2.2894,
+      "step": 961
+    },
+    {
+      "epoch": 0.30808646917534027,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.369849951095401e-05,
+      "loss": 1.6288,
+      "step": 962
+    },
+    {
+      "epoch": 0.3084067253803042,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.3679928871356743e-05,
+      "loss": 1.4223,
+      "step": 963
+    },
+    {
+      "epoch": 0.30872698158526823,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.366133486645526e-05,
+      "loss": 2.0063,
+      "step": 964
+    },
+    {
+      "epoch": 0.3090472377902322,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.364271751950739e-05,
+      "loss": 2.1687,
+      "step": 965
+    },
+    {
+      "epoch": 0.30936749399519614,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.362407685380015e-05,
+      "loss": 2.4734,
+      "step": 966
+    },
+    {
+      "epoch": 0.30968775020016015,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.3605412892649744e-05,
+      "loss": 2.215,
+      "step": 967
+    },
+    {
+      "epoch": 0.3100080064051241,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.3586725659401504e-05,
+      "loss": 2.1618,
+      "step": 968
+    },
+    {
+      "epoch": 0.31032826261008806,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.356801517742986e-05,
+      "loss": 2.6403,
+      "step": 969
+    },
+    {
+      "epoch": 0.31064851881505207,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.3549281470138334e-05,
+      "loss": 2.0227,
+      "step": 970
+    },
+    {
+      "epoch": 0.310968775020016,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.353052456095952e-05,
+      "loss": 2.4126,
+      "step": 971
+    },
+    {
+      "epoch": 0.31128903122498,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.351174447335498e-05,
+      "loss": 2.1611,
+      "step": 972
+    },
+    {
+      "epoch": 0.311609287429944,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.349294123081531e-05,
+      "loss": 2.2805,
+      "step": 973
+    },
+    {
+      "epoch": 0.31192954363490794,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.347411485686006e-05,
+      "loss": 2.2679,
+      "step": 974
+    },
+    {
+      "epoch": 0.3122497998398719,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.345526537503771e-05,
+      "loss": 2.2008,
+      "step": 975
+    },
+    {
+      "epoch": 0.31257005604483584,
+      "grad_norm": 2.0,
+      "learning_rate": 4.3436392808925654e-05,
+      "loss": 1.8056,
+      "step": 976
+    },
+    {
+      "epoch": 0.31289031224979985,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.341749718213014e-05,
+      "loss": 2.3376,
+      "step": 977
+    },
+    {
+      "epoch": 0.3132105684547638,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.339857851828628e-05,
+      "loss": 2.4653,
+      "step": 978
+    },
+    {
+      "epoch": 0.31353082465972776,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.337963684105799e-05,
+      "loss": 2.2227,
+      "step": 979
+    },
+    {
+      "epoch": 0.31385108086469177,
+      "grad_norm": 0.375,
+      "learning_rate": 4.336067217413797e-05,
+      "loss": 2.2355,
+      "step": 980
+    },
+    {
+      "epoch": 0.3141713370696557,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.334168454124769e-05,
+      "loss": 2.4895,
+      "step": 981
+    },
+    {
+      "epoch": 0.3144915932746197,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.332267396613734e-05,
+      "loss": 1.9417,
+      "step": 982
+    },
+    {
+      "epoch": 0.3148118494795837,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.330364047258579e-05,
+      "loss": 2.6751,
+      "step": 983
+    },
+    {
+      "epoch": 0.31513210568454764,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.3284584084400604e-05,
+      "loss": 1.9346,
+      "step": 984
+    },
+    {
+      "epoch": 0.3154523618895116,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.3265504825417966e-05,
+      "loss": 2.1739,
+      "step": 985
+    },
+    {
+      "epoch": 0.3157726180944756,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.3246402719502676e-05,
+      "loss": 1.7595,
+      "step": 986
+    },
+    {
+      "epoch": 0.31609287429943955,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.3227277790548104e-05,
+      "loss": 2.2438,
+      "step": 987
+    },
+    {
+      "epoch": 0.3164131305044035,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.3208130062476174e-05,
+      "loss": 2.4872,
+      "step": 988
+    },
+    {
+      "epoch": 0.3167333867093675,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.318895955923731e-05,
+      "loss": 2.4345,
+      "step": 989
+    },
+    {
+      "epoch": 0.31705364291433147,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.316976630481046e-05,
+      "loss": 2.2613,
+      "step": 990
+    },
+    {
+      "epoch": 0.3173738991192954,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.3150550323203e-05,
+      "loss": 2.1071,
+      "step": 991
+    },
+    {
+      "epoch": 0.31769415532425943,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.313131163845073e-05,
+      "loss": 2.1463,
+      "step": 992
+    },
+    {
+      "epoch": 0.3180144115292234,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.311205027461789e-05,
+      "loss": 1.4839,
+      "step": 993
+    },
+    {
+      "epoch": 0.31833466773418734,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.309276625579703e-05,
+      "loss": 2.4197,
+      "step": 994
+    },
+    {
+      "epoch": 0.31865492393915135,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.307345960610906e-05,
+      "loss": 2.0932,
+      "step": 995
+    },
+    {
+      "epoch": 0.3189751801441153,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.305413034970322e-05,
+      "loss": 2.1792,
+      "step": 996
+    },
+    {
+      "epoch": 0.31929543634907925,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.3034778510757e-05,
+      "loss": 2.5835,
+      "step": 997
+    },
+    {
+      "epoch": 0.3196156925540432,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.3015404113476156e-05,
+      "loss": 1.9542,
+      "step": 998
+    },
+    {
+      "epoch": 0.3199359487590072,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.2996007182094634e-05,
+      "loss": 1.7533,
+      "step": 999
+    },
+    {
+      "epoch": 0.32025620496397117,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.297658774087459e-05,
+      "loss": 2.4536,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3205764611689351,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.295714581410633e-05,
+      "loss": 2.0973,
+      "step": 1001
+    },
+    {
+      "epoch": 0.32089671737389913,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.293768142610828e-05,
+      "loss": 2.5078,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3212169735788631,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.291819460122697e-05,
+      "loss": 1.7711,
+      "step": 1003
+    },
+    {
+      "epoch": 0.32153722978382704,
+      "grad_norm": 0.3828125,
+      "learning_rate": 4.2898685363836974e-05,
+      "loss": 2.1171,
+      "step": 1004
+    },
+    {
+      "epoch": 0.32185748598879105,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.287915373834094e-05,
+      "loss": 2.2499,
+      "step": 1005
+    },
+    {
+      "epoch": 0.322177742193755,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.2859599749169474e-05,
+      "loss": 1.9514,
+      "step": 1006
+    },
+    {
+      "epoch": 0.32249799839871895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.284002342078119e-05,
+      "loss": 2.2806,
+      "step": 1007
+    },
+    {
+      "epoch": 0.32281825460368296,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.2820424777662616e-05,
+      "loss": 1.9288,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3231385108086469,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.280080384432821e-05,
+      "loss": 2.4497,
+      "step": 1009
+    },
+    {
+      "epoch": 0.32345876701361087,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.278116064532032e-05,
+      "loss": 2.273,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3237790232185749,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.2761495205209114e-05,
+      "loss": 2.2382,
+      "step": 1011
+    },
+    {
+      "epoch": 0.32409927942353883,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.274180754859261e-05,
+      "loss": 2.0357,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3244195356285028,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.27220977000966e-05,
+      "loss": 1.655,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3247397918334668,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.2702365684374626e-05,
+      "loss": 2.2382,
+      "step": 1014
+    },
+    {
+      "epoch": 0.32506004803843075,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.2682611526107986e-05,
+      "loss": 1.7932,
+      "step": 1015
+    },
+    {
+      "epoch": 0.3253803042433947,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.266283525000564e-05,
+      "loss": 2.5412,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3257005604483587,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.2643036880804236e-05,
+      "loss": 2.1839,
+      "step": 1017
+    },
+    {
+      "epoch": 0.32602081665332266,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.2623216443268055e-05,
+      "loss": 2.3921,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3263410728582866,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.2603373962188966e-05,
+      "loss": 2.1971,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3266613290632506,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.258350946238643e-05,
+      "loss": 2.4754,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3269815852682146,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.256362296870744e-05,
+      "loss": 2.1899,
+      "step": 1021
+    },
+    {
+      "epoch": 0.32730184147317853,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.25437145060265e-05,
+      "loss": 2.5574,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3276220976781425,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.252378409924559e-05,
+      "loss": 2.1373,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3279423538831065,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.2503831773294144e-05,
+      "loss": 2.4083,
+      "step": 1024
+    },
+    {
+      "epoch": 0.32826261008807045,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.248385755312901e-05,
+      "loss": 1.7979,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3285828662930344,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.246386146373444e-05,
+      "loss": 2.0333,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3289031224979984,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.244384353012199e-05,
+      "loss": 1.9149,
+      "step": 1027
+    },
+    {
+      "epoch": 0.32922337870296237,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.2423803777330606e-05,
+      "loss": 2.0417,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3295436349079263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.240374223042647e-05,
+      "loss": 1.9744,
+      "step": 1029
+    },
+    {
+      "epoch": 0.32986389111289033,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.2383658914503056e-05,
+      "loss": 2.2739,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3301841473178543,
+      "grad_norm": 0.30078125,
+      "learning_rate": 4.236355385468106e-05,
+      "loss": 1.5005,
+      "step": 1031
+    },
+    {
+      "epoch": 0.33050440352281824,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.234342707610837e-05,
+      "loss": 2.1598,
+      "step": 1032
+    },
+    {
+      "epoch": 0.33082465972778224,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.2323278603960046e-05,
+      "loss": 2.4382,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3311449159327462,
+      "grad_norm": 0.380859375,
+      "learning_rate": 4.230310846343828e-05,
+      "loss": 1.963,
+      "step": 1034
+    },
+    {
+      "epoch": 0.33146517213771015,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.228291667977238e-05,
+      "loss": 1.9068,
+      "step": 1035
+    },
+    {
+      "epoch": 0.33178542834267416,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.22627032782187e-05,
+      "loss": 1.4926,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3321056845476381,
+      "grad_norm": 0.5,
+      "learning_rate": 4.2242468284060644e-05,
+      "loss": 2.3178,
+      "step": 1037
+    },
+    {
+      "epoch": 0.33242594075260207,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.222221172260865e-05,
+      "loss": 2.6006,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3327461969575661,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.2201933619200095e-05,
+      "loss": 2.1655,
+      "step": 1039
+    },
+    {
+      "epoch": 0.33306645316253003,
+      "grad_norm": 0.875,
+      "learning_rate": 4.218163399919933e-05,
+      "loss": 2.2719,
+      "step": 1040
+    },
+    {
+      "epoch": 0.333386709367494,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.216131288799761e-05,
+      "loss": 2.1145,
+      "step": 1041
+    },
+    {
+      "epoch": 0.333706965572458,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.214097031101305e-05,
+      "loss": 1.9372,
+      "step": 1042
+    },
+    {
+      "epoch": 0.33402722177742195,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.212060629369065e-05,
+      "loss": 2.4681,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3343474779823859,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.210022086150221e-05,
+      "loss": 1.957,
+      "step": 1044
+    },
+    {
+      "epoch": 0.33466773418734985,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.207981403994632e-05,
+      "loss": 2.0795,
+      "step": 1045
+    },
+    {
+      "epoch": 0.33498799039231386,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.2059385854548316e-05,
+      "loss": 2.4652,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3353082465972778,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.203893633086027e-05,
+      "loss": 2.3613,
+      "step": 1047
+    },
+    {
+      "epoch": 0.33562850280224177,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.201846549446094e-05,
+      "loss": 2.4627,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3359487590072058,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.1997973370955734e-05,
+      "loss": 2.232,
+      "step": 1049
+    },
+    {
+      "epoch": 0.33626901521216973,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.19774599859767e-05,
+      "loss": 2.6032,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3365892714171337,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.1956925365182466e-05,
+      "loss": 2.4148,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3369095276220977,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.193636953425823e-05,
+      "loss": 1.6132,
+      "step": 1052
+    },
+    {
+      "epoch": 0.33722978382706165,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.191579251891572e-05,
+      "loss": 2.2119,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3375500400320256,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.1895194344893177e-05,
+      "loss": 2.0276,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3378702962369896,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.187457503795527e-05,
+      "loss": 2.0666,
+      "step": 1055
+    },
+    {
+      "epoch": 0.33819055244195356,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.185393462389313e-05,
+      "loss": 2.1673,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3385108086469175,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.1833273128524284e-05,
+      "loss": 2.0761,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3388310648518815,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.181259057769264e-05,
+      "loss": 2.4819,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3391513210568455,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.1791886997268414e-05,
+      "loss": 2.1933,
+      "step": 1059
+    },
+    {
+      "epoch": 0.33947157726180943,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.177116241314815e-05,
+      "loss": 2.2095,
+      "step": 1060
+    },
+    {
+      "epoch": 0.33979183346677344,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.175041685125465e-05,
+      "loss": 2.0342,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3401120896717374,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.172965033753697e-05,
+      "loss": 2.3994,
+      "step": 1062
+    },
+    {
+      "epoch": 0.34043234587670135,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.170886289797036e-05,
+      "loss": 2.0665,
+      "step": 1063
+    },
+    {
+      "epoch": 0.34075260208166536,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.1688054558556245e-05,
+      "loss": 1.9771,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3410728582866293,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.16672253453222e-05,
+      "loss": 2.5785,
+      "step": 1065
+    },
+    {
+      "epoch": 0.34139311449159326,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.1646375284321916e-05,
+      "loss": 2.0384,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3417133706965573,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.1625504401635126e-05,
+      "loss": 2.1372,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3420336269015212,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.1604612723367666e-05,
+      "loss": 1.9049,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3423538831064852,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.1583700275651314e-05,
+      "loss": 2.012,
+      "step": 1069
+    },
+    {
+      "epoch": 0.34267413931144913,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.1562767084643885e-05,
+      "loss": 2.1901,
+      "step": 1070
+    },
+    {
+      "epoch": 0.34299439551641314,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.154181317652911e-05,
+      "loss": 2.4813,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3433146517213771,
+      "grad_norm": 0.5,
+      "learning_rate": 4.152083857751665e-05,
+      "loss": 2.4138,
+      "step": 1072
+    },
+    {
+      "epoch": 0.34363490792634105,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.149984331384203e-05,
+      "loss": 2.4764,
+      "step": 1073
+    },
+    {
+      "epoch": 0.34395516413130506,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.147882741176662e-05,
+      "loss": 2.5071,
+      "step": 1074
+    },
+    {
+      "epoch": 0.344275420336269,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.145779089757764e-05,
+      "loss": 2.285,
+      "step": 1075
+    },
+    {
+      "epoch": 0.34459567654123296,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.1436733797588056e-05,
+      "loss": 2.2156,
+      "step": 1076
+    },
+    {
+      "epoch": 0.344915932746197,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.141565613813658e-05,
+      "loss": 2.0952,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3452361889511609,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.139455794558768e-05,
+      "loss": 2.3122,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3455564451561249,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.137343924633147e-05,
+      "loss": 1.9691,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3458767013610889,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.135230006678373e-05,
+      "loss": 2.2278,
+      "step": 1080
+    },
+    {
+      "epoch": 0.34619695756605284,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.133114043338585e-05,
+      "loss": 1.9843,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3465172137710168,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.130996037260482e-05,
+      "loss": 1.5984,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3468374699759808,
+      "grad_norm": 0.302734375,
+      "learning_rate": 4.128875991093315e-05,
+      "loss": 2.2697,
+      "step": 1083
+    },
+    {
+      "epoch": 0.34715772618094476,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.12675390748889e-05,
+      "loss": 2.314,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3474779823859087,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.12462978910156e-05,
+      "loss": 2.4867,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3477982385908727,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.1225036385882235e-05,
+      "loss": 1.9026,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3481184947958367,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.12037545860832e-05,
+      "loss": 2.6461,
+      "step": 1087
+    },
+    {
+      "epoch": 0.34843875100080063,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.118245251823829e-05,
+      "loss": 2.2835,
+      "step": 1088
+    },
+    {
+      "epoch": 0.34875900720576464,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.116113020899264e-05,
+      "loss": 2.3112,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3490792634107286,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.1139787685016707e-05,
+      "loss": 2.0939,
+      "step": 1090
+    },
+    {
+      "epoch": 0.34939951961569254,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.1118424973006237e-05,
+      "loss": 1.9811,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3497197758206565,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.1097042099682216e-05,
+      "loss": 2.5208,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3500400320256205,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.107563909179087e-05,
+      "loss": 1.6477,
+      "step": 1093
+    },
+    {
+      "epoch": 0.35036028823058446,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.105421597610358e-05,
+      "loss": 2.1296,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3506805444355484,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.1032772779416904e-05,
+      "loss": 2.5322,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3510008006405124,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.10113095285525e-05,
+      "loss": 1.8264,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3513210568454764,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.098982625035713e-05,
+      "loss": 2.2612,
+      "step": 1097
+    },
+    {
+      "epoch": 0.35164131305044033,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.096832297170259e-05,
+      "loss": 1.9801,
+      "step": 1098
+    },
+    {
+      "epoch": 0.35196156925540434,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.094679971948569e-05,
+      "loss": 2.2899,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3522818254603683,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.0925256520628244e-05,
+      "loss": 2.1428,
+      "step": 1100
+    },
+    {
+      "epoch": 0.35260208166533225,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.090369340207699e-05,
+      "loss": 1.8895,
+      "step": 1101
+    },
+    {
+      "epoch": 0.35292233787029625,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.088211039080361e-05,
+      "loss": 2.3364,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3532425940752602,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.086050751380464e-05,
+      "loss": 1.9986,
+      "step": 1103
+    },
+    {
+      "epoch": 0.35356285028022416,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.083888479810149e-05,
+      "loss": 2.2253,
+      "step": 1104
+    },
+    {
+      "epoch": 0.35388310648518817,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.081724227074036e-05,
+      "loss": 2.1126,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3542033626901521,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.079557995879225e-05,
+      "loss": 2.427,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3545236188951161,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.0773897889352913e-05,
+      "loss": 2.4259,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3548438751000801,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.075219608954278e-05,
+      "loss": 2.0609,
+      "step": 1108
+    },
+    {
+      "epoch": 0.35516413130504404,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.073047458650699e-05,
+      "loss": 2.5113,
+      "step": 1109
+    },
+    {
+      "epoch": 0.355484387510008,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.070873340741534e-05,
+      "loss": 2.1542,
+      "step": 1110
+    },
+    {
+      "epoch": 0.355804643714972,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.06869725794622e-05,
+      "loss": 2.0773,
+      "step": 1111
+    },
+    {
+      "epoch": 0.35612489991993596,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.066519212986654e-05,
+      "loss": 1.5918,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3564451561248999,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.064339208587187e-05,
+      "loss": 1.8691,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3567654123298639,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.0621572474746227e-05,
+      "loss": 2.0796,
+      "step": 1114
+    },
+    {
+      "epoch": 0.35708566853482787,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.059973332378207e-05,
+      "loss": 2.437,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3574059247397918,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.057787466029637e-05,
+      "loss": 2.6097,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3577261809447558,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.055599651163044e-05,
+      "loss": 2.567,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3580464371497198,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.053409890515002e-05,
+      "loss": 2.0911,
+      "step": 1118
+    },
+    {
+      "epoch": 0.35836669335468374,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.051218186824515e-05,
+      "loss": 2.3121,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3586869495596477,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.049024542833018e-05,
+      "loss": 2.4189,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3590072057646117,
+      "grad_norm": 0.5,
+      "learning_rate": 4.046828961284375e-05,
+      "loss": 2.6126,
+      "step": 1121
+    },
+    {
+      "epoch": 0.35932746196957566,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.0446314449248716e-05,
+      "loss": 2.208,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3596477181745396,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.042431996503213e-05,
+      "loss": 1.7206,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3599679743795036,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.040230618770524e-05,
+      "loss": 2.3154,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3602882305844676,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.03802731448034e-05,
+      "loss": 2.3005,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3606084867894315,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4.0358220863886074e-05,
+      "loss": 2.2069,
+      "step": 1126
+    },
+    {
+      "epoch": 0.36092874299439553,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.033614937253678e-05,
+      "loss": 2.4196,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3612489991993595,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.031405869836307e-05,
+      "loss": 2.4691,
+      "step": 1128
+    },
+    {
+      "epoch": 0.36156925540432344,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.0291948868996496e-05,
+      "loss": 2.7772,
+      "step": 1129
+    },
+    {
+      "epoch": 0.36188951160928745,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.026981991209256e-05,
+      "loss": 2.2456,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3622097678142514,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.024767185533069e-05,
+      "loss": 2.1406,
+      "step": 1131
+    },
+    {
+      "epoch": 0.36253002401921536,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.022550472641422e-05,
+      "loss": 2.3268,
+      "step": 1132
+    },
+    {
+      "epoch": 0.36285028022417937,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.020331855307031e-05,
+      "loss": 2.2335,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3631705364291433,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.018111336304997e-05,
+      "loss": 2.384,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3634907926341073,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.015888918412799e-05,
+      "loss": 2.0399,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3638110488390713,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.01366460441029e-05,
+      "loss": 2.2721,
+      "step": 1136
+    },
+    {
+      "epoch": 0.36413130504403524,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.011438397079696e-05,
+      "loss": 1.8505,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3644515612489992,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.0092102992056095e-05,
+      "loss": 1.8735,
+      "step": 1138
+    },
+    {
+      "epoch": 0.36477181745396314,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.006980313574991e-05,
+      "loss": 2.2612,
+      "step": 1139
+    },
+    {
+      "epoch": 0.36509207365892715,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.004748442977158e-05,
+      "loss": 2.2203,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3654123298638911,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.002514690203788e-05,
+      "loss": 2.1066,
+      "step": 1141
+    },
+    {
+      "epoch": 0.36573258606885506,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.000279058048915e-05,
+      "loss": 2.2341,
+      "step": 1142
+    },
+    {
+      "epoch": 0.36605284227381907,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.998041549308919e-05,
+      "loss": 2.4673,
+      "step": 1143
+    },
+    {
+      "epoch": 0.366373098478783,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.995802166782531e-05,
+      "loss": 2.1514,
+      "step": 1144
+    },
+    {
+      "epoch": 0.366693354683747,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.993560913270823e-05,
+      "loss": 1.8712,
+      "step": 1145
+    },
+    {
+      "epoch": 0.367013610888711,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.991317791577212e-05,
+      "loss": 2.6296,
+      "step": 1146
+    },
+    {
+      "epoch": 0.36733386709367494,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.989072804507444e-05,
+      "loss": 2.4734,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3676541232986389,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.986825954869607e-05,
+      "loss": 2.5021,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3679743795036029,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.9845772454741124e-05,
+      "loss": 2.0079,
+      "step": 1149
+    },
+    {
+      "epoch": 0.36829463570856685,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.982326679133701e-05,
+      "loss": 2.6182,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3686148919135308,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.9800742586634346e-05,
+      "loss": 1.8996,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3689351481184948,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.977819986880694e-05,
+      "loss": 2.4005,
+      "step": 1152
+    },
+    {
+      "epoch": 0.36925540432345877,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.9755638666051795e-05,
+      "loss": 2.2208,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3695756605284227,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.973305900658897e-05,
+      "loss": 1.9195,
+      "step": 1154
+    },
+    {
+      "epoch": 0.36989591673338673,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.971046091866166e-05,
+      "loss": 1.4788,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3702161729383507,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.968784443053612e-05,
+      "loss": 2.1755,
+      "step": 1156
+    },
+    {
+      "epoch": 0.37053642914331464,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.966520957050156e-05,
+      "loss": 2.3551,
+      "step": 1157
+    },
+    {
+      "epoch": 0.37085668534827865,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.964255636687023e-05,
+      "loss": 2.3383,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3711769415532426,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.961988484797729e-05,
+      "loss": 1.8462,
+      "step": 1159
+    },
+    {
+      "epoch": 0.37149719775820655,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.959719504218083e-05,
+      "loss": 2.2438,
+      "step": 1160
+    },
+    {
+      "epoch": 0.37181745396317056,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.95744869778618e-05,
+      "loss": 2.2278,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3721377101681345,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.9551760683424e-05,
+      "loss": 2.2491,
+      "step": 1162
+    },
+    {
+      "epoch": 0.37245796637309847,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.952901618729402e-05,
+      "loss": 1.8948,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3727782225780624,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.950625351792122e-05,
+      "loss": 1.8903,
+      "step": 1164
+    },
+    {
+      "epoch": 0.37309847878302643,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.948347270377769e-05,
+      "loss": 1.2149,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3734187349879904,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.946067377335824e-05,
+      "loss": 2.2527,
+      "step": 1166
+    },
+    {
+      "epoch": 0.37373899119295434,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.94378567551803e-05,
+      "loss": 1.5419,
+      "step": 1167
+    },
+    {
+      "epoch": 0.37405924739791835,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.941502167778395e-05,
+      "loss": 1.9236,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3743795036028823,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.9392168569731854e-05,
+      "loss": 2.3081,
+      "step": 1169
+    },
+    {
+      "epoch": 0.37469975980784626,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.9369297459609247e-05,
+      "loss": 2.1903,
+      "step": 1170
+    },
+    {
+      "epoch": 0.37502001601281026,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.934640837602383e-05,
+      "loss": 2.5142,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3753402722177742,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.932350134760585e-05,
+      "loss": 2.1751,
+      "step": 1172
+    },
+    {
+      "epoch": 0.37566052842273817,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.930057640300795e-05,
+      "loss": 2.3225,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3759807846277022,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.92776335709052e-05,
+      "loss": 2.1874,
+      "step": 1174
+    },
+    {
+      "epoch": 0.37630104083266613,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.925467287999507e-05,
+      "loss": 2.1818,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3766212970376301,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.9231694358997326e-05,
+      "loss": 2.558,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3769415532425941,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.920869803665405e-05,
+      "loss": 2.0283,
+      "step": 1177
+    },
+    {
+      "epoch": 0.37726180944755805,
+      "grad_norm": 0.5,
+      "learning_rate": 3.91856839417296e-05,
+      "loss": 2.6307,
+      "step": 1178
+    },
+    {
+      "epoch": 0.377582065652522,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.916265210301057e-05,
+      "loss": 1.9584,
+      "step": 1179
+    },
+    {
+      "epoch": 0.377902321857486,
+      "grad_norm": 0.5,
+      "learning_rate": 3.913960254930572e-05,
+      "loss": 1.7026,
+      "step": 1180
+    },
+    {
+      "epoch": 0.37822257806244997,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.9116535309446e-05,
+      "loss": 2.0822,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3785428342674139,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.909345041228447e-05,
+      "loss": 2.5519,
+      "step": 1182
+    },
+    {
+      "epoch": 0.37886309047237793,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.9070347886696254e-05,
+      "loss": 2.2184,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3791833466773419,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.9047227761578566e-05,
+      "loss": 1.765,
+      "step": 1184
+    },
+    {
+      "epoch": 0.37950360288230583,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.902409006585061e-05,
+      "loss": 2.3076,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3798238590872698,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.900093482845357e-05,
+      "loss": 2.0427,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3801441152922338,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.8977762078350576e-05,
+      "loss": 1.6712,
+      "step": 1187
+    },
+    {
+      "epoch": 0.38046437149719775,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.895457184452665e-05,
+      "loss": 2.4834,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3807846277021617,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.8931364155988726e-05,
+      "loss": 2.2172,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3811048839071257,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.89081390417655e-05,
+      "loss": 2.2178,
+      "step": 1190
+    },
+    {
+      "epoch": 0.38142514011208967,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.888489653090752e-05,
+      "loss": 2.2411,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3817453963170536,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.886163665248709e-05,
+      "loss": 2.7614,
+      "step": 1192
+    },
+    {
+      "epoch": 0.38206565252201763,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.8838359435598205e-05,
+      "loss": 2.2614,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3823859087269816,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.881506490935657e-05,
+      "loss": 2.0497,
+      "step": 1194
+    },
+    {
+      "epoch": 0.38270616493194554,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.879175310289955e-05,
+      "loss": 2.3673,
+      "step": 1195
+    },
+    {
+      "epoch": 0.38302642113690955,
+      "grad_norm": 4.6875,
+      "learning_rate": 3.876842404538611e-05,
+      "loss": 2.7132,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3833466773418735,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.874507776599678e-05,
+      "loss": 2.1813,
+      "step": 1197
+    },
+    {
+      "epoch": 0.38366693354683745,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.872171429393368e-05,
+      "loss": 2.3327,
+      "step": 1198
+    },
+    {
+      "epoch": 0.38398718975180146,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.8698333658420366e-05,
+      "loss": 1.5882,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3843074459567654,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.867493588870193e-05,
+      "loss": 2.0014,
+      "step": 1200
+    },
+    {
+      "epoch": 0.38462770216172937,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.865152101404485e-05,
+      "loss": 2.2738,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3849479583666934,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.862808906373701e-05,
+      "loss": 2.3236,
+      "step": 1202
+    },
+    {
+      "epoch": 0.38526821457165733,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.860464006708766e-05,
+      "loss": 2.6712,
+      "step": 1203
+    },
+    {
+      "epoch": 0.3855884707766213,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.8581174053427374e-05,
+      "loss": 1.9265,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3859087269815853,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.855769105210801e-05,
+      "loss": 2.5721,
+      "step": 1205
+    },
+    {
+      "epoch": 0.38622898318654925,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.8534191092502655e-05,
+      "loss": 2.0682,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3865492393915132,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.851067420400564e-05,
+      "loss": 2.3802,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3868694955964772,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.848714041603245e-05,
+      "loss": 2.3584,
+      "step": 1208
+    },
+    {
+      "epoch": 0.38718975180144116,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.846358975801971e-05,
+      "loss": 2.3271,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3875100080064051,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.844002225942514e-05,
+      "loss": 2.2613,
+      "step": 1210
+    },
+    {
+      "epoch": 0.38783026421136907,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.8416437949727555e-05,
+      "loss": 2.1979,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3881505204163331,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.839283685842676e-05,
+      "loss": 2.7155,
+      "step": 1212
+    },
+    {
+      "epoch": 0.38847077662129703,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.836921901504358e-05,
+      "loss": 1.9949,
+      "step": 1213
+    },
+    {
+      "epoch": 0.388791032826261,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.8345584449119776e-05,
+      "loss": 2.1843,
+      "step": 1214
+    },
+    {
+      "epoch": 0.389111289031225,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.832193319021802e-05,
+      "loss": 2.2712,
+      "step": 1215
+    },
+    {
+      "epoch": 0.38943154523618895,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.8298265267921884e-05,
+      "loss": 2.1854,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3897518014411529,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.827458071183578e-05,
+      "loss": 2.5162,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3900720576461169,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.825087955158492e-05,
+      "loss": 2.4592,
+      "step": 1218
+    },
+    {
+      "epoch": 0.39039231385108086,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.822716181681528e-05,
+      "loss": 1.9011,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3907125700560448,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.820342753719357e-05,
+      "loss": 2.2409,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3910328262610088,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.817967674240721e-05,
+      "loss": 1.9098,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3913530824659728,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.815590946216425e-05,
+      "loss": 2.0723,
+      "step": 1222
+    },
+    {
+      "epoch": 0.39167333867093673,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.813212572619338e-05,
+      "loss": 2.3035,
+      "step": 1223
+    },
+    {
+      "epoch": 0.39199359487590074,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.810832556424388e-05,
+      "loss": 1.9079,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3923138510808647,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.808450900608555e-05,
+      "loss": 1.7984,
+      "step": 1225
+    },
+    {
+      "epoch": 0.39263410728582865,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.806067608150872e-05,
+      "loss": 2.1234,
+      "step": 1226
+    },
+    {
+      "epoch": 0.39295436349079266,
+      "grad_norm": 0.5,
+      "learning_rate": 3.803682682032419e-05,
+      "loss": 2.2432,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3932746196957566,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.8012961252363184e-05,
+      "loss": 1.9167,
+      "step": 1228
+    },
+    {
+      "epoch": 0.39359487590072056,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.798907940747732e-05,
+      "loss": 2.6162,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3939151321056846,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.79651813155386e-05,
+      "loss": 2.239,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3942353883106485,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.794126700643933e-05,
+      "loss": 1.7196,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3945556445156125,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.791733651009209e-05,
+      "loss": 2.1263,
+      "step": 1232
+    },
+    {
+      "epoch": 0.39487590072057643,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.789338985642973e-05,
+      "loss": 2.0495,
+      "step": 1233
+    },
+    {
+      "epoch": 0.39519615692554044,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.786942707540529e-05,
+      "loss": 2.2755,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3955164131305044,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.784544819699202e-05,
+      "loss": 2.5356,
+      "step": 1235
+    },
+    {
+      "epoch": 0.39583666933546835,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.7821453251183245e-05,
+      "loss": 2.3301,
+      "step": 1236
+    },
+    {
+      "epoch": 0.39615692554043236,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.7797442267992436e-05,
+      "loss": 2.3236,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3964771817453963,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.7773415277453103e-05,
+      "loss": 2.3126,
+      "step": 1238
+    },
+    {
+      "epoch": 0.39679743795036027,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.77493723096188e-05,
+      "loss": 2.2927,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3971176941553243,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.772531339456302e-05,
+      "loss": 1.9049,
+      "step": 1240
+    },
+    {
+      "epoch": 0.39743795036028823,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.770123856237925e-05,
+      "loss": 2.5726,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3977582065652522,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.767714784318088e-05,
+      "loss": 2.2026,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3980784627702162,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.765304126710113e-05,
+      "loss": 2.08,
+      "step": 1243
+    },
+    {
+      "epoch": 0.39839871897518014,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.762891886429312e-05,
+      "loss": 1.8371,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3987189751801441,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.760478066492971e-05,
+      "loss": 2.2806,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3990392313851081,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.758062669920354e-05,
+      "loss": 2.312,
+      "step": 1246
+    },
+    {
+      "epoch": 0.39935948759007206,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.7556456997326975e-05,
+      "loss": 2.1427,
+      "step": 1247
+    },
+    {
+      "epoch": 0.399679743795036,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.7532271589532065e-05,
+      "loss": 2.2062,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.75080705060705e-05,
+      "loss": 2.4721,
+      "step": 1249
+    },
+    {
+      "epoch": 0.400320256204964,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.748385377721356e-05,
+      "loss": 2.5967,
+      "step": 1250
+    },
+    {
+      "epoch": 0.40064051240992793,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.7459621433252135e-05,
+      "loss": 2.1675,
+      "step": 1251
+    },
+    {
+      "epoch": 0.40096076861489194,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.74353735044966e-05,
+      "loss": 1.9072,
+      "step": 1252
+    },
+    {
+      "epoch": 0.4012810248198559,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.741111002127688e-05,
+      "loss": 1.6543,
+      "step": 1253
+    },
+    {
+      "epoch": 0.40160128102481985,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.738683101394228e-05,
+      "loss": 1.8961,
+      "step": 1254
+    },
+    {
+      "epoch": 0.40192153722978385,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.73625365128616e-05,
+      "loss": 2.2844,
+      "step": 1255
+    },
+    {
+      "epoch": 0.4022417934347478,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.733822654842298e-05,
+      "loss": 2.2592,
+      "step": 1256
+    },
+    {
+      "epoch": 0.40256204963971176,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.73139011510339e-05,
+      "loss": 2.8168,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4028823058446757,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.728956035112115e-05,
+      "loss": 1.9843,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4032025620496397,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.72652041791308e-05,
+      "loss": 2.2555,
+      "step": 1259
+    },
+    {
+      "epoch": 0.4035228182546037,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.7240832665528124e-05,
+      "loss": 2.3078,
+      "step": 1260
+    },
+    {
+      "epoch": 0.40384307445956763,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.721644584079761e-05,
+      "loss": 1.7157,
+      "step": 1261
+    },
+    {
+      "epoch": 0.40416333066453164,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.7192043735442865e-05,
+      "loss": 2.4214,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4044835868694956,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.7167626379986655e-05,
+      "loss": 2.5919,
+      "step": 1263
+    },
+    {
+      "epoch": 0.40480384307445955,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.714319380497077e-05,
+      "loss": 2.6869,
+      "step": 1264
+    },
+    {
+      "epoch": 0.40512409927942356,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.7118746040956076e-05,
+      "loss": 2.0477,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4054443554843875,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.709428311852242e-05,
+      "loss": 1.8267,
+      "step": 1266
+    },
+    {
+      "epoch": 0.40576461168935146,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.706980506826863e-05,
+      "loss": 2.1828,
+      "step": 1267
+    },
+    {
+      "epoch": 0.40608486789431547,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.704531192081241e-05,
+      "loss": 2.2251,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4064051240992794,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.70208037067904e-05,
+      "loss": 1.7606,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4067253803042434,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.699628045685805e-05,
+      "loss": 2.4729,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4070456365092074,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.697174220168965e-05,
+      "loss": 1.5516,
+      "step": 1271
+    },
+    {
+      "epoch": 0.40736589271417134,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.694718897197822e-05,
+      "loss": 2.0469,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4076861489191353,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.692262079843553e-05,
+      "loss": 1.997,
+      "step": 1273
+    },
+    {
+      "epoch": 0.4080064051240993,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.689803771179207e-05,
+      "loss": 2.3182,
+      "step": 1274
+    },
+    {
+      "epoch": 0.40832666132906326,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.687343974279691e-05,
+      "loss": 2.4617,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4086469175340272,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.684882692221782e-05,
+      "loss": 2.3608,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4089671737389912,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.682419928084109e-05,
+      "loss": 2.1872,
+      "step": 1277
+    },
+    {
+      "epoch": 0.40928742994395517,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.679955684947158e-05,
+      "loss": 2.0972,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4096076861489191,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.6774899658932615e-05,
+      "loss": 2.086,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4099279423538831,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.6750227740066015e-05,
+      "loss": 2.3214,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4102481985588471,
+      "grad_norm": 1.6640625,
+      "learning_rate": 3.6725541123732e-05,
+      "loss": 2.2449,
+      "step": 1281
+    },
+    {
+      "epoch": 0.41056845476381104,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.6700839840809206e-05,
+      "loss": 1.619,
+      "step": 1282
+    },
+    {
+      "epoch": 0.410888710968775,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.6676123922194574e-05,
+      "loss": 2.1188,
+      "step": 1283
+    },
+    {
+      "epoch": 0.411208967173739,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.6651393398803366e-05,
+      "loss": 2.6604,
+      "step": 1284
+    },
+    {
+      "epoch": 0.41152922337870296,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.6626648301569124e-05,
+      "loss": 1.9312,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4118494795836669,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.6601888661443603e-05,
+      "loss": 2.0071,
+      "step": 1286
+    },
+    {
+      "epoch": 0.4121697357886309,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.657711450939676e-05,
+      "loss": 2.1451,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4124899919935949,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.6552325876416704e-05,
+      "loss": 2.6817,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4128102481985588,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.652752279350965e-05,
+      "loss": 2.2703,
+      "step": 1289
+    },
+    {
+      "epoch": 0.41313050440352284,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.65027052916999e-05,
+      "loss": 1.6399,
+      "step": 1290
+    },
+    {
+      "epoch": 0.4134507606084868,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.647787340202975e-05,
+      "loss": 2.139,
+      "step": 1291
+    },
+    {
+      "epoch": 0.41377101681345074,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.6453027155559565e-05,
+      "loss": 2.0271,
+      "step": 1292
+    },
+    {
+      "epoch": 0.41409127301841475,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.6428166583367615e-05,
+      "loss": 2.1767,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4144115292233787,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.640329171655008e-05,
+      "loss": 1.9477,
+      "step": 1294
+    },
+    {
+      "epoch": 0.41473178542834266,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.637840258622106e-05,
+      "loss": 3.0508,
+      "step": 1295
+    },
+    {
+      "epoch": 0.41505204163330667,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.635349922351249e-05,
+      "loss": 2.6344,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4153722978382706,
+      "grad_norm": 4.125,
+      "learning_rate": 3.632858165957407e-05,
+      "loss": 2.1484,
+      "step": 1297
+    },
+    {
+      "epoch": 0.4156925540432346,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.63036499255733e-05,
+      "loss": 2.0567,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4160128102481986,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.627870405269539e-05,
+      "loss": 2.6683,
+      "step": 1299
+    },
+    {
+      "epoch": 0.41633306645316254,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.625374407214325e-05,
+      "loss": 2.0889,
+      "step": 1300
+    },
+    {
+      "epoch": 0.4166533226581265,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.62287700151374e-05,
+      "loss": 2.3849,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4169735788630905,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.6203781912916e-05,
+      "loss": 2.5107,
+      "step": 1302
+    },
+    {
+      "epoch": 0.41729383506805445,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.6178779796734785e-05,
+      "loss": 2.0597,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4176140912730184,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.615376369786699e-05,
+      "loss": 2.3948,
+      "step": 1304
+    },
+    {
+      "epoch": 0.41793434747798236,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.6128733647603345e-05,
+      "loss": 2.1924,
+      "step": 1305
+    },
+    {
+      "epoch": 0.41825460368294637,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.610368967725205e-05,
+      "loss": 2.1514,
+      "step": 1306
+    },
+    {
+      "epoch": 0.4185748598879103,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.607863181813871e-05,
+      "loss": 2.1708,
+      "step": 1307
+    },
+    {
+      "epoch": 0.4188951160928743,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.6053560101606285e-05,
+      "loss": 2.4148,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4192153722978383,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.602847455901509e-05,
+      "loss": 2.3323,
+      "step": 1309
+    },
+    {
+      "epoch": 0.41953562850280224,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.600337522174272e-05,
+      "loss": 2.2687,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4198558847077662,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.597826212118404e-05,
+      "loss": 2.5078,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4201761409127302,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.5953135288751094e-05,
+      "loss": 2.1704,
+      "step": 1312
+    },
+    {
+      "epoch": 0.42049639711769415,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.5927994755873157e-05,
+      "loss": 2.3606,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4208166533226581,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.5902840553996594e-05,
+      "loss": 1.9514,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4211369095276221,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.58776727145849e-05,
+      "loss": 2.6369,
+      "step": 1315
+    },
+    {
+      "epoch": 0.42145716573258607,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.58524912691186e-05,
+      "loss": 2.2219,
+      "step": 1316
+    },
+    {
+      "epoch": 0.42177742193755,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.5827296249095264e-05,
+      "loss": 1.7456,
+      "step": 1317
+    },
+    {
+      "epoch": 0.42209767814251403,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.580208768602941e-05,
+      "loss": 2.2512,
+      "step": 1318
+    },
+    {
+      "epoch": 0.422417934347478,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.577686561145254e-05,
+      "loss": 2.1483,
+      "step": 1319
+    },
+    {
+      "epoch": 0.42273819055244194,
+      "grad_norm": 0.3671875,
+      "learning_rate": 3.575163005691302e-05,
+      "loss": 1.9452,
+      "step": 1320
+    },
+    {
+      "epoch": 0.42305844675740595,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.572638105397607e-05,
+      "loss": 2.1301,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4233787029623699,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.5701118634223785e-05,
+      "loss": 2.099,
+      "step": 1322
+    },
+    {
+      "epoch": 0.42369895916733386,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.567584282925498e-05,
+      "loss": 2.2,
+      "step": 1323
+    },
+    {
+      "epoch": 0.42401921537229786,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.565055367068525e-05,
+      "loss": 2.4591,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4243394715772618,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.562525119014688e-05,
+      "loss": 2.1612,
+      "step": 1325
+    },
+    {
+      "epoch": 0.42465972778222577,
+      "grad_norm": 0.5,
+      "learning_rate": 3.5599935419288824e-05,
+      "loss": 1.9581,
+      "step": 1326
+    },
+    {
+      "epoch": 0.4249799839871897,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.557460638977666e-05,
+      "loss": 2.4488,
+      "step": 1327
+    },
+    {
+      "epoch": 0.42530024019215373,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.554926413329254e-05,
+      "loss": 2.6207,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4256204963971177,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.552390868153516e-05,
+      "loss": 2.0293,
+      "step": 1329
+    },
+    {
+      "epoch": 0.42594075260208164,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.549854006621976e-05,
+      "loss": 2.2195,
+      "step": 1330
+    },
+    {
+      "epoch": 0.42626100880704565,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.547315831907798e-05,
+      "loss": 2.9019,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4265812650120096,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.544776347185793e-05,
+      "loss": 2.5745,
+      "step": 1332
+    },
+    {
+      "epoch": 0.42690152121697356,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.54223555563241e-05,
+      "loss": 2.6656,
+      "step": 1333
+    },
+    {
+      "epoch": 0.42722177742193757,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.5396934604257326e-05,
+      "loss": 2.2501,
+      "step": 1334
+    },
+    {
+      "epoch": 0.4275420336269015,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.537150064745474e-05,
+      "loss": 2.2139,
+      "step": 1335
+    },
+    {
+      "epoch": 0.42786228983186547,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.534605371772974e-05,
+      "loss": 1.9124,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4281825460368295,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.5320593846911973e-05,
+      "loss": 2.6009,
+      "step": 1337
+    },
+    {
+      "epoch": 0.42850280224179343,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.529512106684724e-05,
+      "loss": 1.5908,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4288230584467574,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.526963540939752e-05,
+      "loss": 2.011,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4291433146517214,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.5244136906440886e-05,
+      "loss": 2.0686,
+      "step": 1340
+    },
+    {
+      "epoch": 0.42946357085668535,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.521862558987148e-05,
+      "loss": 1.9993,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4297838270616493,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.5193101491599453e-05,
+      "loss": 2.6751,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4301040832666133,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.5167564643550975e-05,
+      "loss": 2.1176,
+      "step": 1343
+    },
+    {
+      "epoch": 0.43042433947157727,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.5142015077668155e-05,
+      "loss": 2.3972,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4307445956765412,
+      "grad_norm": 0.291015625,
+      "learning_rate": 3.5116452825909e-05,
+      "loss": 1.8073,
+      "step": 1345
+    },
+    {
+      "epoch": 0.43106485188150523,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.5090877920247375e-05,
+      "loss": 2.6247,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4313851080864692,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.5065290392673e-05,
+      "loss": 1.8295,
+      "step": 1347
+    },
+    {
+      "epoch": 0.43170536429143314,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.503969027519137e-05,
+      "loss": 2.4511,
+      "step": 1348
+    },
+    {
+      "epoch": 0.43202562049639714,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.501407759982372e-05,
+      "loss": 2.578,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4323458767013611,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.498845239860702e-05,
+      "loss": 2.1768,
+      "step": 1350
+    },
+    {
+      "epoch": 0.43266613290632505,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.496281470359386e-05,
+      "loss": 2.084,
+      "step": 1351
+    },
+    {
+      "epoch": 0.432986389111289,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.493716454685251e-05,
+      "loss": 2.215,
+      "step": 1352
+    },
+    {
+      "epoch": 0.433306645316253,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.491150196046679e-05,
+      "loss": 2.1564,
+      "step": 1353
+    },
+    {
+      "epoch": 0.43362690152121697,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.4885826976536074e-05,
+      "loss": 2.0137,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4339471577261809,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.486013962717526e-05,
+      "loss": 1.9845,
+      "step": 1355
+    },
+    {
+      "epoch": 0.43426741393114493,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.483443994451471e-05,
+      "loss": 1.7616,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4345876701361089,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.480872796070018e-05,
+      "loss": 2.5005,
+      "step": 1357
+    },
+    {
+      "epoch": 0.43490792634107284,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.478300370789286e-05,
+      "loss": 2.4385,
+      "step": 1358
+    },
+    {
+      "epoch": 0.43522818254603685,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.475726721826926e-05,
+      "loss": 2.3023,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4355484387510008,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.473151852402119e-05,
+      "loss": 2.0413,
+      "step": 1360
+    },
+    {
+      "epoch": 0.43586869495596475,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.470575765735574e-05,
+      "loss": 2.4551,
+      "step": 1361
+    },
+    {
+      "epoch": 0.43618895116092876,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.4679984650495225e-05,
+      "loss": 2.5011,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4365092073658927,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.465419953567714e-05,
+      "loss": 1.651,
+      "step": 1363
+    },
+    {
+      "epoch": 0.43682946357085667,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.4628402345154136e-05,
+      "loss": 2.2163,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4371497197758207,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.4602593111193946e-05,
+      "loss": 2.1722,
+      "step": 1365
+    },
+    {
+      "epoch": 0.43746997598078463,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.457677186607938e-05,
+      "loss": 2.5205,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4377902321857486,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.455093864210828e-05,
+      "loss": 2.0087,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4381104883907126,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.452509347159346e-05,
+      "loss": 2.0194,
+      "step": 1368
+    },
+    {
+      "epoch": 0.43843074459567655,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.449923638686269e-05,
+      "loss": 2.6472,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4387510008006405,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.447336742025862e-05,
+      "loss": 2.1153,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4390712570056045,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.4447486604138776e-05,
+      "loss": 2.0821,
+      "step": 1371
+    },
+    {
+      "epoch": 0.43939151321056846,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.442159397087551e-05,
+      "loss": 2.3243,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4397117694155324,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.4395689552855955e-05,
+      "loss": 2.3805,
+      "step": 1373
+    },
+    {
+      "epoch": 0.44003202562049637,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.436977338248197e-05,
+      "loss": 2.2077,
+      "step": 1374
+    },
+    {
+      "epoch": 0.4403522818254604,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.434384549217013e-05,
+      "loss": 2.3821,
+      "step": 1375
+    },
+    {
+      "epoch": 0.44067253803042433,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.431790591435167e-05,
+      "loss": 2.0472,
+      "step": 1376
+    },
+    {
+      "epoch": 0.4409927942353883,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.4291954681472427e-05,
+      "loss": 1.6111,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4413130504403523,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.426599182599283e-05,
+      "loss": 2.5023,
+      "step": 1378
+    },
+    {
+      "epoch": 0.44163330664531625,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.424001738038784e-05,
+      "loss": 2.4245,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4419535628502802,
+      "grad_norm": 0.375,
+      "learning_rate": 3.4214031377146915e-05,
+      "loss": 2.2083,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4422738190552442,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.4188033848773995e-05,
+      "loss": 2.5568,
+      "step": 1381
+    },
+    {
+      "epoch": 0.44259407526020816,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.4162024827787384e-05,
+      "loss": 2.4897,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4429143314651721,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.4136004346719815e-05,
+      "loss": 2.2791,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4432345876701361,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.410997243811832e-05,
+      "loss": 1.5351,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4435548438751001,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.408392913454423e-05,
+      "loss": 2.3915,
+      "step": 1385
+    },
+    {
+      "epoch": 0.44387510008006403,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.4057874468573145e-05,
+      "loss": 1.9231,
+      "step": 1386
+    },
+    {
+      "epoch": 0.44419535628502804,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.4031808472794875e-05,
+      "loss": 2.1587,
+      "step": 1387
+    },
+    {
+      "epoch": 0.444515612489992,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.400573117981338e-05,
+      "loss": 1.8927,
+      "step": 1388
+    },
+    {
+      "epoch": 0.44483586869495595,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.3979642622246765e-05,
+      "loss": 1.9999,
+      "step": 1389
+    },
+    {
+      "epoch": 0.44515612489991996,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.395354283272724e-05,
+      "loss": 1.7148,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4454763811048839,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.3927431843901034e-05,
+      "loss": 2.6938,
+      "step": 1391
+    },
+    {
+      "epoch": 0.44579663730984787,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.39013096884284e-05,
+      "loss": 2.2425,
+      "step": 1392
+    },
+    {
+      "epoch": 0.4461168935148119,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.3875176398983565e-05,
+      "loss": 2.3499,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4464371497197758,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.3849032008254676e-05,
+      "loss": 2.2984,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4467574059247398,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.382287654894377e-05,
+      "loss": 1.9116,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4470776621297038,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.379671005376671e-05,
+      "loss": 2.2836,
+      "step": 1396
+    },
+    {
+      "epoch": 0.44739791833466774,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.377053255545319e-05,
+      "loss": 1.6366,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4477181745396317,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.374434408674665e-05,
+      "loss": 2.4968,
+      "step": 1398
+    },
+    {
+      "epoch": 0.44803843074459565,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.371814468040426e-05,
+      "loss": 2.4331,
+      "step": 1399
+    },
+    {
+      "epoch": 0.44835868694955966,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.3691934369196855e-05,
+      "loss": 2.381,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4486789431545236,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.366571318590894e-05,
+      "loss": 1.7545,
+      "step": 1401
+    },
+    {
+      "epoch": 0.44899919935948757,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.3639481163338584e-05,
+      "loss": 2.4128,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4493194555644516,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.3613238334297446e-05,
+      "loss": 1.8779,
+      "step": 1403
+    },
+    {
+      "epoch": 0.44963971176941553,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.358698473161068e-05,
+      "loss": 2.3488,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4499599679743795,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.3560720388116934e-05,
+      "loss": 1.6194,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4502802241793435,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.3534445336668266e-05,
+      "loss": 2.0731,
+      "step": 1406
+    },
+    {
+      "epoch": 0.45060048038430744,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.350815961013014e-05,
+      "loss": 2.0653,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4509207365892714,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.348186324138139e-05,
+      "loss": 2.5686,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4512409927942354,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.345555626331415e-05,
+      "loss": 2.0038,
+      "step": 1409
+    },
+    {
+      "epoch": 0.45156124899919936,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.34292387088338e-05,
+      "loss": 2.3573,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4518815052041633,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.340291061085898e-05,
+      "loss": 2.2184,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4522017614091273,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.3376572002321506e-05,
+      "loss": 2.8585,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4525220176140913,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.335022291616635e-05,
+      "loss": 2.5615,
+      "step": 1413
+    },
+    {
+      "epoch": 0.45284227381905523,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.332386338535158e-05,
+      "loss": 2.1242,
+      "step": 1414
+    },
+    {
+      "epoch": 0.45316253002401924,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.329749344284832e-05,
+      "loss": 2.2279,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4534827862289832,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.327111312164075e-05,
+      "loss": 2.3805,
+      "step": 1416
+    },
+    {
+      "epoch": 0.45380304243394715,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.324472245472598e-05,
+      "loss": 2.2084,
+      "step": 1417
+    },
+    {
+      "epoch": 0.45412329863891115,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.321832147511412e-05,
+      "loss": 2.6584,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4544435548438751,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.319191021582814e-05,
+      "loss": 1.9955,
+      "step": 1419
+    },
+    {
+      "epoch": 0.45476381104883906,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.3165488709903876e-05,
+      "loss": 2.1134,
+      "step": 1420
+    },
+    {
+      "epoch": 0.455084067253803,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.313905699038997e-05,
+      "loss": 2.4377,
+      "step": 1421
+    },
+    {
+      "epoch": 0.455404323458767,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.3112615090347875e-05,
+      "loss": 2.3482,
+      "step": 1422
+    },
+    {
+      "epoch": 0.455724579663731,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.3086163042851757e-05,
+      "loss": 2.2396,
+      "step": 1423
+    },
+    {
+      "epoch": 0.45604483586869493,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.305970088098844e-05,
+      "loss": 1.9499,
+      "step": 1424
+    },
+    {
+      "epoch": 0.45636509207365894,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.303322863785747e-05,
+      "loss": 2.4442,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4566853482786229,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.300674634657094e-05,
+      "loss": 2.0177,
+      "step": 1426
+    },
+    {
+      "epoch": 0.45700560448358685,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.2980254040253554e-05,
+      "loss": 2.3229,
+      "step": 1427
+    },
+    {
+      "epoch": 0.45732586068855086,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.2953751752042514e-05,
+      "loss": 1.5711,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4576461168935148,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.2927239515087524e-05,
+      "loss": 2.3031,
+      "step": 1429
+    },
+    {
+      "epoch": 0.45796637309847876,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.290071736255073e-05,
+      "loss": 2.4136,
+      "step": 1430
+    },
+    {
+      "epoch": 0.45828662930344277,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.287418532760667e-05,
+      "loss": 2.4615,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4586068855084067,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.284764344344226e-05,
+      "loss": 2.3343,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4589271417133707,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.282109174325672e-05,
+      "loss": 2.6001,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4592473979183347,
+      "grad_norm": 0.75,
+      "learning_rate": 3.279453026026157e-05,
+      "loss": 2.282,
+      "step": 1434
+    },
+    {
+      "epoch": 0.45956765412329864,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.2767959027680525e-05,
+      "loss": 2.6124,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4598879103282626,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.2741378078749545e-05,
+      "loss": 1.9807,
+      "step": 1436
+    },
+    {
+      "epoch": 0.4602081665332266,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.271478744671672e-05,
+      "loss": 2.7701,
+      "step": 1437
+    },
+    {
+      "epoch": 0.46052842273819056,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.2688187164842235e-05,
+      "loss": 2.2759,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4608486789431545,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.266157726639836e-05,
+      "loss": 2.1833,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4611689351481185,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.263495778466942e-05,
+      "loss": 2.2356,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4614891913530825,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.260832875295169e-05,
+      "loss": 2.2155,
+      "step": 1441
+    },
+    {
+      "epoch": 0.4618094475580464,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.25816902045534e-05,
+      "loss": 1.4778,
+      "step": 1442
+    },
+    {
+      "epoch": 0.46212970376301044,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.255504217279469e-05,
+      "loss": 1.7559,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4624499599679744,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.2528384691007555e-05,
+      "loss": 2.1804,
+      "step": 1444
+    },
+    {
+      "epoch": 0.46277021617293834,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.250171779253583e-05,
+      "loss": 2.2981,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4630904723779023,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.2475041510735086e-05,
+      "loss": 2.4689,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4634107285828663,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.244835587897268e-05,
+      "loss": 2.397,
+      "step": 1447
+    },
+    {
+      "epoch": 0.46373098478783026,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.242166093062764e-05,
+      "loss": 2.2169,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4640512409927942,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.239495669909064e-05,
+      "loss": 2.2407,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4643714971977582,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.236824321776397e-05,
+      "loss": 2.1264,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4646917534027222,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.2341520520061516e-05,
+      "loss": 2.2125,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4650120096076861,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.2314788639408656e-05,
+      "loss": 2.1495,
+      "step": 1452
+    },
+    {
+      "epoch": 0.46533226581265014,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.2288047609242266e-05,
+      "loss": 2.029,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4656525220176141,
+      "grad_norm": 0.625,
+      "learning_rate": 3.226129746301067e-05,
+      "loss": 1.9174,
+      "step": 1454
+    },
+    {
+      "epoch": 0.46597277822257804,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.223453823417362e-05,
+      "loss": 2.5737,
+      "step": 1455
+    },
+    {
+      "epoch": 0.46629303442754205,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.220776995620217e-05,
+      "loss": 2.2023,
+      "step": 1456
+    },
+    {
+      "epoch": 0.466613290632506,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.218099266257874e-05,
+      "loss": 2.6245,
+      "step": 1457
+    },
+    {
+      "epoch": 0.46693354683746996,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.2154206386797e-05,
+      "loss": 1.9389,
+      "step": 1458
+    },
+    {
+      "epoch": 0.46725380304243397,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.212741116236189e-05,
+      "loss": 2.0921,
+      "step": 1459
+    },
+    {
+      "epoch": 0.4675740592473979,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.210060702278951e-05,
+      "loss": 2.5264,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4678943154523619,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.207379400160711e-05,
+      "loss": 1.9612,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4682145716573259,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.204697213235309e-05,
+      "loss": 2.329,
+      "step": 1462
+    },
+    {
+      "epoch": 0.46853482786228984,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.2020141448576866e-05,
+      "loss": 2.4217,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4688550840672538,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.1993301983838906e-05,
+      "loss": 1.7458,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4691753402722178,
+      "grad_norm": 0.3828125,
+      "learning_rate": 3.1966453771710655e-05,
+      "loss": 2.112,
+      "step": 1465
+    },
+    {
+      "epoch": 0.46949559647718175,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.1939596845774525e-05,
+      "loss": 2.4703,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4698158526821457,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.191273123962378e-05,
+      "loss": 2.4275,
+      "step": 1467
+    },
+    {
+      "epoch": 0.47013610888710966,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.188585698686257e-05,
+      "loss": 2.4174,
+      "step": 1468
+    },
+    {
+      "epoch": 0.47045636509207367,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.185897412110587e-05,
+      "loss": 2.2741,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4707766212970376,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.1832082675979394e-05,
+      "loss": 2.2359,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4710968775020016,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.180518268511963e-05,
+      "loss": 2.174,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4714171337069656,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.1778274182173716e-05,
+      "loss": 2.2126,
+      "step": 1472
+    },
+    {
+      "epoch": 0.47173738991192954,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.175135720079947e-05,
+      "loss": 2.0206,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4720576461168935,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.172443177466529e-05,
+      "loss": 2.4025,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4723779023218575,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.169749793745014e-05,
+      "loss": 2.6801,
+      "step": 1475
+    },
+    {
+      "epoch": 0.47269815852682145,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.167055572284352e-05,
+      "loss": 2.2362,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4730184147317854,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.164360516454541e-05,
+      "loss": 2.0306,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4733386709367494,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.1616646296266196e-05,
+      "loss": 2.1496,
+      "step": 1478
+    },
+    {
+      "epoch": 0.47365892714171337,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.158967915172669e-05,
+      "loss": 2.6622,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4739791833466773,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.1562703764658055e-05,
+      "loss": 2.2467,
+      "step": 1480
+    },
+    {
+      "epoch": 0.47429943955164133,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.1535720168801725e-05,
+      "loss": 1.7394,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4746196957566053,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.150872839790946e-05,
+      "loss": 1.9964,
+      "step": 1482
+    },
+    {
+      "epoch": 0.47493995196156924,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.148172848574319e-05,
+      "loss": 2.3196,
+      "step": 1483
+    },
+    {
+      "epoch": 0.47526020816653325,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.145472046607508e-05,
+      "loss": 2.052,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4755804643714972,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.142770437268739e-05,
+      "loss": 1.8644,
+      "step": 1485
+    },
+    {
+      "epoch": 0.47590072057646116,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.140068023937251e-05,
+      "loss": 2.1707,
+      "step": 1486
+    },
+    {
+      "epoch": 0.47622097678142516,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.137364809993288e-05,
+      "loss": 1.9486,
+      "step": 1487
+    },
+    {
+      "epoch": 0.4765412329863891,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.1346607988180933e-05,
+      "loss": 1.5815,
+      "step": 1488
+    },
+    {
+      "epoch": 0.47686148919135307,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.131955993793911e-05,
+      "loss": 2.0043,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4771817453963171,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.129250398303975e-05,
+      "loss": 2.2674,
+      "step": 1490
+    },
+    {
+      "epoch": 0.47750200160128103,
+      "grad_norm": 0.3828125,
+      "learning_rate": 3.12654401573251e-05,
+      "loss": 2.037,
+      "step": 1491
+    },
+    {
+      "epoch": 0.477822257806245,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.123836849464723e-05,
+      "loss": 1.7428,
+      "step": 1492
+    },
+    {
+      "epoch": 0.47814251401120894,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.121128902886804e-05,
+      "loss": 2.6863,
+      "step": 1493
+    },
+    {
+      "epoch": 0.47846277021617295,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.118420179385919e-05,
+      "loss": 2.0104,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4787830264211369,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.115710682350201e-05,
+      "loss": 2.1841,
+      "step": 1495
+    },
+    {
+      "epoch": 0.47910328262610086,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.113000415168756e-05,
+      "loss": 2.503,
+      "step": 1496
+    },
+    {
+      "epoch": 0.47942353883106487,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.110289381231651e-05,
+      "loss": 1.9073,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4797437950360288,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.107577583929914e-05,
+      "loss": 1.9962,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4800640512409928,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.104865026655525e-05,
+      "loss": 2.0423,
+      "step": 1499
+    },
+    {
+      "epoch": 0.4803843074459568,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.102151712801415e-05,
+      "loss": 2.1058,
+      "step": 1500
+    },
+    {
+      "epoch": 0.48070456365092074,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.0994376457614636e-05,
+      "loss": 2.2794,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4810248198558847,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.09672282893049e-05,
+      "loss": 2.1654,
+      "step": 1502
+    },
+    {
+      "epoch": 0.4813450760608487,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.094007265704253e-05,
+      "loss": 1.8414,
+      "step": 1503
+    },
+    {
+      "epoch": 0.48166533226581265,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.091290959479444e-05,
+      "loss": 1.9365,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4819855884707766,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.0885739136536854e-05,
+      "loss": 2.2331,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4823058446757406,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.0858561316255224e-05,
+      "loss": 1.8893,
+      "step": 1506
+    },
+    {
+      "epoch": 0.48262610088070457,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.083137616794422e-05,
+      "loss": 1.8019,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4829463570856685,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.080418372560768e-05,
+      "loss": 1.971,
+      "step": 1508
+    },
+    {
+      "epoch": 0.48326661329063253,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.077698402325857e-05,
+      "loss": 2.0729,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4835868694955965,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.0749777094918924e-05,
+      "loss": 2.444,
+      "step": 1510
+    },
+    {
+      "epoch": 0.48390712570056044,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.072256297461983e-05,
+      "loss": 2.6036,
+      "step": 1511
+    },
+    {
+      "epoch": 0.48422738190552445,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.069534169640136e-05,
+      "loss": 2.5759,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4845476381104884,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.066811329431254e-05,
+      "loss": 2.5265,
+      "step": 1513
+    },
+    {
+      "epoch": 0.48486789431545235,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.0640877802411316e-05,
+      "loss": 2.5761,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4851881505204163,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.0613635254764495e-05,
+      "loss": 1.9635,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4855084067253803,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.058638568544772e-05,
+      "loss": 1.9067,
+      "step": 1516
+    },
+    {
+      "epoch": 0.48582866293034427,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.055912912854538e-05,
+      "loss": 2.7006,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4861489191353082,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.0531865618150654e-05,
+      "loss": 2.7706,
+      "step": 1518
+    },
+    {
+      "epoch": 0.48646917534027223,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.050459518836539e-05,
+      "loss": 2.2856,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4867894315452362,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.0477317873300103e-05,
+      "loss": 2.2043,
+      "step": 1520
+    },
+    {
+      "epoch": 0.48710968775020014,
+      "grad_norm": 1.3828125,
+      "learning_rate": 3.045003370707391e-05,
+      "loss": 1.7819,
+      "step": 1521
+    },
+    {
+      "epoch": 0.48742994395516415,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.0422742723814503e-05,
+      "loss": 2.2534,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4877502001601281,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.0395444957658097e-05,
+      "loss": 1.9324,
+      "step": 1523
+    },
+    {
+      "epoch": 0.48807045636509205,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.0368140442749398e-05,
+      "loss": 2.167,
+      "step": 1524
+    },
+    {
+      "epoch": 0.48839071257005606,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.034082921324155e-05,
+      "loss": 2.6054,
+      "step": 1525
+    },
+    {
+      "epoch": 0.48871096877502,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.03135113032961e-05,
+      "loss": 2.4174,
+      "step": 1526
+    },
+    {
+      "epoch": 0.48903122497998397,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.0286186747082934e-05,
+      "loss": 2.3323,
+      "step": 1527
+    },
+    {
+      "epoch": 0.489351481184948,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.0258855578780265e-05,
+      "loss": 2.1546,
+      "step": 1528
+    },
+    {
+      "epoch": 0.48967173738991193,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.023151783257459e-05,
+      "loss": 2.4691,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4899919935948759,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.0204173542660613e-05,
+      "loss": 2.0822,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4903122497998399,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.0176822743241222e-05,
+      "loss": 2.6899,
+      "step": 1531
+    },
+    {
+      "epoch": 0.49063250600480385,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.014946546852746e-05,
+      "loss": 2.5449,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4909527622097678,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.0122101752738468e-05,
+      "loss": 2.0894,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4912730184147318,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.0094731630101436e-05,
+      "loss": 2.2903,
+      "step": 1534
+    },
+    {
+      "epoch": 0.49159327461969576,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.0067355134851565e-05,
+      "loss": 2.2904,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4919135308246597,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.0039972301232045e-05,
+      "loss": 2.2971,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4922337870296237,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.001258316349398e-05,
+      "loss": 1.7656,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4925540432345877,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.998518775589635e-05,
+      "loss": 2.0205,
+      "step": 1538
+    },
+    {
+      "epoch": 0.49287429943955163,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.9957786112706006e-05,
+      "loss": 2.3737,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4931945556445156,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.9930378268197577e-05,
+      "loss": 2.1592,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4935148118494796,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.990296425665345e-05,
+      "loss": 2.3567,
+      "step": 1541
+    },
+    {
+      "epoch": 0.49383506805444355,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.987554411236373e-05,
+      "loss": 2.3767,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4941553242594075,
+      "grad_norm": 1.625,
+      "learning_rate": 2.9848117869626192e-05,
+      "loss": 2.5746,
+      "step": 1543
+    },
+    {
+      "epoch": 0.4944755804643715,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.9820685562746254e-05,
+      "loss": 1.4638,
+      "step": 1544
+    },
+    {
+      "epoch": 0.49479583666933546,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.9793247226036885e-05,
+      "loss": 1.9633,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4951160928742994,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.9765802893818627e-05,
+      "loss": 1.7408,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4954363490792634,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.9738352600419506e-05,
+      "loss": 2.0108,
+      "step": 1547
+    },
+    {
+      "epoch": 0.4957566052842274,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.9710896380175024e-05,
+      "loss": 2.3077,
+      "step": 1548
+    },
+    {
+      "epoch": 0.49607686148919133,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.9683434267428055e-05,
+      "loss": 2.0489,
+      "step": 1549
+    },
+    {
+      "epoch": 0.49639711769415534,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.9655966296528893e-05,
+      "loss": 2.8007,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4967173738991193,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.962849250183513e-05,
+      "loss": 2.0005,
+      "step": 1551
+    },
+    {
+      "epoch": 0.49703763010408325,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.9601012917711646e-05,
+      "loss": 1.5309,
+      "step": 1552
+    },
+    {
+      "epoch": 0.49735788630904726,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.957352757853056e-05,
+      "loss": 2.4253,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4976781425140112,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.9546036518671206e-05,
+      "loss": 1.8466,
+      "step": 1554
+    },
+    {
+      "epoch": 0.49799839871897517,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.9518539772520064e-05,
+      "loss": 2.2555,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4983186549239392,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.9491037374470715e-05,
+      "loss": 2.0387,
+      "step": 1556
+    },
+    {
+      "epoch": 0.49863891112890313,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.9463529358923824e-05,
+      "loss": 2.4217,
+      "step": 1557
+    },
+    {
+      "epoch": 0.4989591673338671,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.9436015760287082e-05,
+      "loss": 2.1835,
+      "step": 1558
+    },
+    {
+      "epoch": 0.4992794235388311,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.9408496612975156e-05,
+      "loss": 2.1866,
+      "step": 1559
+    },
+    {
+      "epoch": 0.49959967974379504,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.9380971951409652e-05,
+      "loss": 2.0304,
+      "step": 1560
+    },
+    {
+      "epoch": 0.499919935948759,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.9353441810019093e-05,
+      "loss": 1.4819,
+      "step": 1561
+    },
+    {
+      "epoch": 0.500240192153723,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.9325906223238836e-05,
+      "loss": 1.8902,
+      "step": 1562
+    },
+    {
+      "epoch": 0.500560448358687,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.929836522551104e-05,
+      "loss": 2.3322,
+      "step": 1563
+    },
+    {
+      "epoch": 0.500880704563651,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.927081885128467e-05,
+      "loss": 1.9023,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5012009607686149,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.9243267135015383e-05,
+      "loss": 2.1684,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5015212169735789,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.921571011116552e-05,
+      "loss": 2.0673,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5018414731785429,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.9188147814204086e-05,
+      "loss": 2.5705,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5021617293835068,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.916058027860665e-05,
+      "loss": 2.6609,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5024819855884708,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.913300753885536e-05,
+      "loss": 2.2706,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5028022417934348,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.910542962943885e-05,
+      "loss": 2.5645,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5031224979983987,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.907784658485225e-05,
+      "loss": 2.7824,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5034427542033627,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.905025843959708e-05,
+      "loss": 1.5037,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5037630104083266,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.902266522818125e-05,
+      "loss": 2.2413,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5040832666132906,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.899506698511903e-05,
+      "loss": 2.5246,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5044035228182546,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.8967463744930958e-05,
+      "loss": 1.7726,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5047237790232185,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.8939855542143828e-05,
+      "loss": 1.8297,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5050440352281825,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.8912242411290646e-05,
+      "loss": 1.8675,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5053642914331465,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.8884624386910586e-05,
+      "loss": 2.3941,
+      "step": 1578
+    },
+    {
+      "epoch": 0.5056845476381104,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.8857001503548925e-05,
+      "loss": 2.5356,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5060048038430744,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.882937379575704e-05,
+      "loss": 2.5708,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5063250600480385,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.8801741298092332e-05,
+      "loss": 2.1658,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5066453162530024,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.8774104045118183e-05,
+      "loss": 1.7119,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5069655724579664,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.8746462071403952e-05,
+      "loss": 1.936,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5072858286629304,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.8718815411524863e-05,
+      "loss": 2.1219,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5076060848678943,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.8691164100062034e-05,
+      "loss": 2.5705,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5079263410728583,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.8663508171602386e-05,
+      "loss": 2.2458,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5082465972778223,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.8635847660738635e-05,
+      "loss": 1.8396,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5085668534827862,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.8608182602069195e-05,
+      "loss": 2.5934,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5088871096877502,
+      "grad_norm": 0.625,
+      "learning_rate": 2.8580513030198186e-05,
+      "loss": 2.4622,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5092073658927142,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.8552838979735393e-05,
+      "loss": 2.2395,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5095276220976781,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.8525160485296167e-05,
+      "loss": 2.172,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5098478783026421,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.849747758150143e-05,
+      "loss": 2.4499,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5101681345076061,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.846979030297764e-05,
+      "loss": 2.2869,
+      "step": 1593
+    },
+    {
+      "epoch": 0.51048839071257,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.8442098684356707e-05,
+      "loss": 1.8278,
+      "step": 1594
+    },
+    {
+      "epoch": 0.510808646917534,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.841440276027596e-05,
+      "loss": 2.2451,
+      "step": 1595
+    },
+    {
+      "epoch": 0.511128903122498,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.8386702565378137e-05,
+      "loss": 2.3676,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5114491593274619,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.8358998134311316e-05,
+      "loss": 2.4343,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5117694155324259,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.8331289501728854e-05,
+      "loss": 2.516,
+      "step": 1598
+    },
+    {
+      "epoch": 0.51208967173739,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.830357670228937e-05,
+      "loss": 2.5648,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5124099279423538,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.8275859770656714e-05,
+      "loss": 2.4266,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5127301841473179,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.8248138741499885e-05,
+      "loss": 2.4382,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5130504403522819,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.822041364949301e-05,
+      "loss": 2.4932,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5133706965572458,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.8192684529315305e-05,
+      "loss": 2.121,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5136909527622098,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.816495141565102e-05,
+      "loss": 2.0563,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5140112089671738,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.8137214343189395e-05,
+      "loss": 2.4438,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5143314651721377,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.8109473346624627e-05,
+      "loss": 2.3331,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5146517213771017,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.808172846065583e-05,
+      "loss": 2.2871,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5149719775820657,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.8053979719986972e-05,
+      "loss": 2.4588,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5152922337870296,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.802622715932684e-05,
+      "loss": 2.5462,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5156124899919936,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.799847081338901e-05,
+      "loss": 1.7209,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5159327461969576,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.7970710716891783e-05,
+      "loss": 2.7945,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5162530024019215,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.7942946904558155e-05,
+      "loss": 2.2602,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5165732586068855,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.791517941111577e-05,
+      "loss": 2.0592,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5168935148118495,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.7887408271296883e-05,
+      "loss": 2.1679,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5172137710168134,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7859633519838296e-05,
+      "loss": 2.5147,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5175340272217774,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.7831855191481336e-05,
+      "loss": 2.471,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5178542834267414,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.7804073320971812e-05,
+      "loss": 2.2137,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5181745396317053,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.777628794305995e-05,
+      "loss": 2.0848,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5184947958366694,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.774849909250037e-05,
+      "loss": 2.0605,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5188150520416333,
+      "grad_norm": 0.75,
+      "learning_rate": 2.7720706804052026e-05,
+      "loss": 2.6878,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5191353082465973,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.7692911112478193e-05,
+      "loss": 2.4631,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5194555644515613,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.7665112052546387e-05,
+      "loss": 2.5574,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5197758206565252,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.763730965902834e-05,
+      "loss": 2.0565,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5200960768614892,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.7609503966699945e-05,
+      "loss": 1.9217,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5204163330664532,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.7581695010341246e-05,
+      "loss": 2.049,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5207365892714171,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.7553882824736348e-05,
+      "loss": 2.1618,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5210568454763811,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.7526067444673402e-05,
+      "loss": 1.9671,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5213771016813451,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.749824890494455e-05,
+      "loss": 2.1159,
+      "step": 1628
+    },
+    {
+      "epoch": 0.521697357886309,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.74704272403459e-05,
+      "loss": 1.5528,
+      "step": 1629
+    },
+    {
+      "epoch": 0.522017614091273,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.744260248567745e-05,
+      "loss": 2.162,
+      "step": 1630
+    },
+    {
+      "epoch": 0.522337870296237,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.741477467574307e-05,
+      "loss": 1.8618,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5226581265012009,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.738694384535046e-05,
+      "loss": 1.8943,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5229783827061649,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.73591100293111e-05,
+      "loss": 2.0352,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5232986389111289,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.7331273262440175e-05,
+      "loss": 2.1791,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5236188951160928,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.7303433579556602e-05,
+      "loss": 2.4775,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5239391513210568,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.727559101548292e-05,
+      "loss": 2.1351,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5242594075260208,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.7247745605045277e-05,
+      "loss": 2.3471,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5245796637309847,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.7219897383073373e-05,
+      "loss": 1.6935,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5248999199359488,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.7192046384400444e-05,
+      "loss": 2.4096,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5252201761409128,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.7164192643863196e-05,
+      "loss": 2.2081,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5255404323458767,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.7136336196301737e-05,
+      "loss": 2.1331,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5258606885508407,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.7108477076559595e-05,
+      "loss": 2.1021,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5261809447558047,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.7080615319483634e-05,
+      "loss": 2.491,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5265012009607686,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.705275095992399e-05,
+      "loss": 2.7152,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5268214571657326,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.7024884032734082e-05,
+      "loss": 2.138,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5271417133706966,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.6997014572770534e-05,
+      "loss": 2.4889,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5274619695756605,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.696914261489315e-05,
+      "loss": 2.4738,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5277822257806245,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.694126819396482e-05,
+      "loss": 2.4388,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5281024819855885,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.6913391344851556e-05,
+      "loss": 2.0395,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5284227381905524,
+      "grad_norm": 1.125,
+      "learning_rate": 2.688551210242239e-05,
+      "loss": 2.4296,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5287429943955164,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.6857630501549347e-05,
+      "loss": 2.1873,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5290632506004804,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.68297465771074e-05,
+      "loss": 2.5596,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5293835068054443,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.6801860363974434e-05,
+      "loss": 1.793,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5297037630104083,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.6773971897031207e-05,
+      "loss": 2.0554,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5300240192153723,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.6746081211161268e-05,
+      "loss": 1.7286,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5303442754203362,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.6718188341250955e-05,
+      "loss": 2.0696,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5306645316253003,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.6690293322189353e-05,
+      "loss": 1.9679,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5309847878302643,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.6662396188868228e-05,
+      "loss": 2.3103,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5313050440352282,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.6634496976181968e-05,
+      "loss": 2.3372,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5316253002401922,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6606595719027583e-05,
+      "loss": 2.3772,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5319455564451562,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.657869245230464e-05,
+      "loss": 2.5822,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5322658126501201,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.6550787210915223e-05,
+      "loss": 1.8052,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5325860688550841,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.6522880029763862e-05,
+      "loss": 2.583,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5329063250600481,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.6494970943757548e-05,
+      "loss": 2.0871,
+      "step": 1664
+    },
+    {
+      "epoch": 0.533226581265012,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.6467059987805633e-05,
+      "loss": 2.2921,
+      "step": 1665
+    },
+    {
+      "epoch": 0.533546837469976,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.6439147196819795e-05,
+      "loss": 1.3349,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5338670936749399,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.6411232605714043e-05,
+      "loss": 2.4875,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5341873498799039,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.6383316249404615e-05,
+      "loss": 1.9192,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5345076060848679,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.6355398162809957e-05,
+      "loss": 2.4075,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5348278622898318,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.632747838085068e-05,
+      "loss": 2.5198,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5351481184947958,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.6299556938449528e-05,
+      "loss": 2.4702,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5354683746997598,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.627163387053131e-05,
+      "loss": 2.4223,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5357886309047237,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.624370921202286e-05,
+      "loss": 2.0182,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5361088871096877,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.6215782997853026e-05,
+      "loss": 2.2894,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5364291433146517,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.6187855262952587e-05,
+      "loss": 2.3239,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5367493995196156,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.615992604225422e-05,
+      "loss": 2.5619,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5370696557245797,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.6131995370692463e-05,
+      "loss": 1.886,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5373899119295437,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.6104063283203672e-05,
+      "loss": 2.2598,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5377101681345076,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.607612981472599e-05,
+      "loss": 1.7341,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5380304243394716,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.6048195000199248e-05,
+      "loss": 1.8696,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5383506805444356,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.6020258874565002e-05,
+      "loss": 2.6784,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5386709367493995,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.5992321472766418e-05,
+      "loss": 1.9824,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5389911929543635,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.596438282974828e-05,
+      "loss": 2.1182,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5393114491593275,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.59364429804569e-05,
+      "loss": 1.9854,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5396317053642914,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.5908501959840125e-05,
+      "loss": 2.27,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5399519615692554,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.588055980284725e-05,
+      "loss": 2.0739,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5402722177742194,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.5852616544429008e-05,
+      "loss": 2.2536,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5405924739791833,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.5824672219537483e-05,
+      "loss": 2.2221,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5409127301841473,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.5796726863126113e-05,
+      "loss": 2.7278,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5412329863891113,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.5768780510149633e-05,
+      "loss": 2.2353,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5415532425940752,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.5740833195563996e-05,
+      "loss": 1.9356,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5418734987990392,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.571288495432639e-05,
+      "loss": 1.7192,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5421937550040032,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.5684935821395135e-05,
+      "loss": 2.4505,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5425140112089671,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.5656985831729692e-05,
+      "loss": 1.8894,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5428342674139311,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.5629035020290566e-05,
+      "loss": 1.8297,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5431545236188952,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.5601083422039314e-05,
+      "loss": 2.2753,
+      "step": 1696
+    },
+    {
+      "epoch": 0.543474779823859,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.5573131071938456e-05,
+      "loss": 2.8383,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5437950360288231,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.554517800495148e-05,
+      "loss": 2.4237,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5441152922337871,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.5517224256042745e-05,
+      "loss": 2.0334,
+      "step": 1699
+    },
+    {
+      "epoch": 0.544435548438751,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.5489269860177474e-05,
+      "loss": 2.4893,
+      "step": 1700
+    },
+    {
+      "epoch": 0.544755804643715,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.5461314852321694e-05,
+      "loss": 2.0231,
+      "step": 1701
+    },
+    {
+      "epoch": 0.545076060848679,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.5433359267442204e-05,
+      "loss": 2.486,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5453963170536429,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.5405403140506517e-05,
+      "loss": 2.3233,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5457165732586069,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.5377446506482837e-05,
+      "loss": 2.0646,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5460368294635709,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.5349489400339997e-05,
+      "loss": 2.1051,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5463570856685348,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.5321531857047404e-05,
+      "loss": 2.5043,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5466773418734988,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.5293573911575036e-05,
+      "loss": 1.8195,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5469975980784628,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.5265615598893356e-05,
+      "loss": 2.6448,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5473178542834267,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.5237656953973303e-05,
+      "loss": 2.3232,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5476381104883907,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.5209698011786214e-05,
+      "loss": 1.9998,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5479583666933547,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.5181738807303806e-05,
+      "loss": 2.0763,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5482786228983186,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.5153779375498137e-05,
+      "loss": 2.3951,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5485988791032826,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.5125819751341522e-05,
+      "loss": 1.9252,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5489191353082465,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.509785996980653e-05,
+      "loss": 2.0553,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5492393915132106,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.506990006586594e-05,
+      "loss": 2.1069,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5495596477181746,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.5041940074492676e-05,
+      "loss": 2.3812,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5498799039231385,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.5013980030659743e-05,
+      "loss": 2.1561,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5502001601281025,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.498601996934026e-05,
+      "loss": 1.9662,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5505204163330665,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.4958059925507333e-05,
+      "loss": 1.932,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5508406725380304,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.4930099934134065e-05,
+      "loss": 2.3476,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5511609287429944,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.490214003019347e-05,
+      "loss": 2.2002,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5514811849479584,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.4874180248658484e-05,
+      "loss": 2.3532,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5518014411529223,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.4846220624501872e-05,
+      "loss": 2.258,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5521216973578863,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.48182611926962e-05,
+      "loss": 2.4362,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5524419535628503,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.479030198821379e-05,
+      "loss": 1.7804,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5527622097678142,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.4762343046026702e-05,
+      "loss": 2.7527,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5530824659727782,
+      "grad_norm": 0.625,
+      "learning_rate": 2.473438440110665e-05,
+      "loss": 2.4608,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5534027221777422,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.4706426088424973e-05,
+      "loss": 1.8141,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5537229783827061,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.46784681429526e-05,
+      "loss": 2.2575,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5540432345876701,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.4650510599660005e-05,
+      "loss": 2.6203,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5543634907926341,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.4622553493517165e-05,
+      "loss": 2.1001,
+      "step": 1731
+    },
+    {
+      "epoch": 0.554683746997598,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.4594596859493485e-05,
+      "loss": 2.3472,
+      "step": 1732
+    },
+    {
+      "epoch": 0.555004003202562,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.4566640732557802e-05,
+      "loss": 2.0663,
+      "step": 1733
+    },
+    {
+      "epoch": 0.555324259407526,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.4538685147678316e-05,
+      "loss": 2.3392,
+      "step": 1734
+    },
+    {
+      "epoch": 0.55564451561249,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.4510730139822535e-05,
+      "loss": 2.1949,
+      "step": 1735
+    },
+    {
+      "epoch": 0.555964771817454,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.4482775743957258e-05,
+      "loss": 2.517,
+      "step": 1736
+    },
+    {
+      "epoch": 0.556285028022418,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.4454821995048522e-05,
+      "loss": 2.4129,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5566052842273819,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.442686892806155e-05,
+      "loss": 2.5147,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5569255404323459,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.4398916577960695e-05,
+      "loss": 2.8204,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5572457966373099,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.437096497970944e-05,
+      "loss": 2.6784,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5575660528422738,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.4343014168270314e-05,
+      "loss": 2.1789,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5578863090472378,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.431506417860487e-05,
+      "loss": 2.8155,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5582065652522018,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.4287115045673617e-05,
+      "loss": 2.0444,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5585268214571657,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.4259166804436006e-05,
+      "loss": 2.646,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5588470776621297,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.4231219489850376e-05,
+      "loss": 2.2975,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5591673338670937,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.4203273136873892e-05,
+      "loss": 2.1118,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5594875900720576,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.4175327780462523e-05,
+      "loss": 2.1737,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5598078462770216,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.4147383455571004e-05,
+      "loss": 2.2012,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5601281024819856,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.4119440197152754e-05,
+      "loss": 1.4749,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5604483586869495,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.409149804015988e-05,
+      "loss": 2.1635,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5607686148919135,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.4063557019543107e-05,
+      "loss": 1.4912,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5610888710968776,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.403561717025173e-05,
+      "loss": 2.4878,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5614091273018414,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.400767852723359e-05,
+      "loss": 2.3552,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5617293835068055,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.3979741125435004e-05,
+      "loss": 1.6363,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5620496397117695,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.3951804999800754e-05,
+      "loss": 2.0071,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5623698959167334,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.3923870185274018e-05,
+      "loss": 2.3637,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5626901521216974,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.389593671679633e-05,
+      "loss": 2.4539,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5630104083266614,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.3868004629307543e-05,
+      "loss": 2.087,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5633306645316253,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.3840073957745793e-05,
+      "loss": 2.3264,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5636509207365893,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.381214473704742e-05,
+      "loss": 2.3831,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5639711769415532,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.3784217002146976e-05,
+      "loss": 2.5073,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5642914331465172,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.3756290787977147e-05,
+      "loss": 2.3088,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5646116893514812,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.3728366129468696e-05,
+      "loss": 2.3283,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5649319455564451,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.3700443061550478e-05,
+      "loss": 2.4001,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5652522017614091,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.3672521619149322e-05,
+      "loss": 2.3446,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5655724579663731,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.364460183719005e-05,
+      "loss": 2.1358,
+      "step": 1766
+    },
+    {
+      "epoch": 0.565892714171337,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.3616683750595394e-05,
+      "loss": 2.2924,
+      "step": 1767
+    },
+    {
+      "epoch": 0.566212970376301,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.3588767394285963e-05,
+      "loss": 1.6567,
+      "step": 1768
+    },
+    {
+      "epoch": 0.566533226581265,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.356085280318021e-05,
+      "loss": 2.1492,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5668534827862289,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.3532940012194383e-05,
+      "loss": 2.6153,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5671737389911929,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.3505029056242454e-05,
+      "loss": 2.1282,
+      "step": 1771
+    },
+    {
+      "epoch": 0.567493995196157,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.347711997023614e-05,
+      "loss": 2.281,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5678142514011209,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.344921278908479e-05,
+      "loss": 2.5034,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5681345076060849,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.342130754769536e-05,
+      "loss": 2.2516,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5684547638110489,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.3393404280972426e-05,
+      "loss": 2.1612,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5687750200160128,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.336550302381804e-05,
+      "loss": 2.4927,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5690952762209768,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.333760381113178e-05,
+      "loss": 2.2154,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5694155324259408,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.330970667781065e-05,
+      "loss": 1.8648,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5697357886309047,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.328181165874905e-05,
+      "loss": 2.3315,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5700560448358687,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.325391878883874e-05,
+      "loss": 2.1458,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5703763010408327,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.3226028102968806e-05,
+      "loss": 1.9153,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5706965572457966,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.3198139636025565e-05,
+      "loss": 1.8971,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5710168134507606,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.3170253422892606e-05,
+      "loss": 2.1762,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5713370696557246,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.3142369498450665e-05,
+      "loss": 1.8368,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5716573258606885,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.3114487897577614e-05,
+      "loss": 2.1639,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5719775820656525,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3086608655148447e-05,
+      "loss": 2.3979,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5722978382706165,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.305873180603519e-05,
+      "loss": 1.7269,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5726180944755804,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.303085738510686e-05,
+      "loss": 2.2162,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5729383506805444,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.3002985427229472e-05,
+      "loss": 2.1375,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5732586068855084,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.2975115967265924e-05,
+      "loss": 1.917,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5735788630904723,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.294724904007602e-05,
+      "loss": 2.2711,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5738991192954364,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.2919384680516382e-05,
+      "loss": 2.324,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5742193755004004,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.2891522923440408e-05,
+      "loss": 2.2216,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5745396317053643,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.286366380369827e-05,
+      "loss": 1.962,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5748598879103283,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.2835807356136817e-05,
+      "loss": 2.2255,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5751801441152923,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.2807953615599552e-05,
+      "loss": 1.8936,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5755004003202562,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.2780102616926633e-05,
+      "loss": 1.6788,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5758206565252202,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.2752254394954736e-05,
+      "loss": 2.1119,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5761409127301842,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.272440898451708e-05,
+      "loss": 2.2388,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5764611689351481,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.2696566420443403e-05,
+      "loss": 2.1596,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5767814251401121,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.266872673755983e-05,
+      "loss": 2.5837,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5771016813450761,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.2640889970688904e-05,
+      "loss": 1.9443,
+      "step": 1802
+    },
+    {
+      "epoch": 0.57742193755004,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.2613056154649542e-05,
+      "loss": 2.2111,
+      "step": 1803
+    },
+    {
+      "epoch": 0.577742193755004,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.2585225324256928e-05,
+      "loss": 2.4131,
+      "step": 1804
+    },
+    {
+      "epoch": 0.578062449959968,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.255739751432256e-05,
+      "loss": 2.4771,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5783827061649319,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.252957275965411e-05,
+      "loss": 2.0832,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5787029623698959,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.2501751095055448e-05,
+      "loss": 2.2442,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5790232185748598,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.2473932555326604e-05,
+      "loss": 2.3033,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5793434747798238,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.2446117175263658e-05,
+      "loss": 2.1901,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5796637309847879,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.2418304989658757e-05,
+      "loss": 2.1636,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5799839871897517,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.239049603330006e-05,
+      "loss": 2.5267,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5803042433947158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.2362690340971678e-05,
+      "loss": 1.9348,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5806244995996798,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.233488794745362e-05,
+      "loss": 2.3273,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5809447558046437,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.2307088887521816e-05,
+      "loss": 1.9374,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5812650120096077,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.2279293195947977e-05,
+      "loss": 2.5428,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5815852682145717,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.225150090749964e-05,
+      "loss": 1.9329,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5819055244195356,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.222371205694006e-05,
+      "loss": 2.0871,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5822257806244996,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.219592667902819e-05,
+      "loss": 2.1848,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5825460368294636,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.2168144808518667e-05,
+      "loss": 1.9535,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5828662930344275,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.2140366480161713e-05,
+      "loss": 2.2323,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5831865492393915,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.211259172870312e-05,
+      "loss": 2.3016,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5835068054443555,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.2084820588884234e-05,
+      "loss": 1.6852,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5838270616493194,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.2057053095441854e-05,
+      "loss": 2.3596,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5841473178542834,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.202928928310822e-05,
+      "loss": 1.8076,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5844675740592474,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.2001529186610996e-05,
+      "loss": 2.6958,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5847878302642113,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.197377284067317e-05,
+      "loss": 2.3731,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5851080864691753,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.194602028001303e-05,
+      "loss": 2.4798,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5854283426741393,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.1918271539344177e-05,
+      "loss": 2.7211,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5857485988791032,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.1890526653375372e-05,
+      "loss": 2.6734,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5860688550840673,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.186278565681061e-05,
+      "loss": 2.3266,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5863891112890313,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.183504858434899e-05,
+      "loss": 2.572,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5867093674939952,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.1807315470684697e-05,
+      "loss": 2.4207,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5870296236989592,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.1779586350506996e-05,
+      "loss": 2.1545,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5873498799039232,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.175186125850012e-05,
+      "loss": 2.287,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5876701361088871,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.1724140229343288e-05,
+      "loss": 2.3484,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5879903923138511,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.169642329771063e-05,
+      "loss": 2.0906,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5883106485188151,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.1668710498271162e-05,
+      "loss": 2.4591,
+      "step": 1837
+    },
+    {
+      "epoch": 0.588630904723779,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1641001865688686e-05,
+      "loss": 1.9441,
+      "step": 1838
+    },
+    {
+      "epoch": 0.588951160928743,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.1613297434621865e-05,
+      "loss": 2.4205,
+      "step": 1839
+    },
+    {
+      "epoch": 0.589271417133707,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.158559723972404e-05,
+      "loss": 2.1306,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5895916733386709,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.15579013156433e-05,
+      "loss": 1.831,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5899119295436349,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.1530209697022366e-05,
+      "loss": 1.9654,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5902321857485989,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.150252241849857e-05,
+      "loss": 2.4949,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5905524419535628,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.1474839514703843e-05,
+      "loss": 1.785,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5908726981585268,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.1447161020264616e-05,
+      "loss": 1.4549,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5911929543634908,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.141948696980181e-05,
+      "loss": 2.0587,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5915132105684547,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.1391817397930814e-05,
+      "loss": 1.8139,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5918334667734187,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.1364152339261374e-05,
+      "loss": 2.3478,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5921537229783828,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.133649182839761e-05,
+      "loss": 1.9641,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5924739791833467,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1308835899937972e-05,
+      "loss": 2.3559,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5927942353883107,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.128118458847515e-05,
+      "loss": 2.3271,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5931144915932747,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.1253537928596057e-05,
+      "loss": 2.0447,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5934347477982386,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.1225895954881823e-05,
+      "loss": 2.2835,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5937550040032026,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.119825870190767e-05,
+      "loss": 1.9083,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5940752602081665,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.1170626204242962e-05,
+      "loss": 2.0059,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5943955164131305,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.114299849645108e-05,
+      "loss": 2.408,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5947157726180945,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.1115375613089416e-05,
+      "loss": 2.1296,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5950360288230584,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1087757588709356e-05,
+      "loss": 2.2477,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5953562850280224,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.106014445785618e-05,
+      "loss": 2.5451,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5956765412329864,
+      "grad_norm": 0.5,
+      "learning_rate": 2.1032536255069045e-05,
+      "loss": 1.9078,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5959967974379503,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.1004933014880976e-05,
+      "loss": 2.3821,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5963170536429143,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.097733477181876e-05,
+      "loss": 1.7438,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5966373098478783,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.094974156040293e-05,
+      "loss": 1.9844,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5969575660528422,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.092215341514776e-05,
+      "loss": 2.3406,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5972778222578062,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.0894570370561156e-05,
+      "loss": 2.2888,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5975980784627702,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.0866992461144645e-05,
+      "loss": 2.4423,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5979183346677341,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.0839419721393354e-05,
+      "loss": 2.4976,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5982385908726982,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.0811852185795916e-05,
+      "loss": 2.1009,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5985588470776622,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.078428988883448e-05,
+      "loss": 2.1664,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5988791032826261,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.075673286498463e-05,
+      "loss": 2.3979,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5991993594875901,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.0729181148715332e-05,
+      "loss": 1.8152,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5995196156925541,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.070163477448896e-05,
+      "loss": 2.2508,
+      "step": 1872
+    },
+    {
+      "epoch": 0.599839871897518,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.0674093776761176e-05,
+      "loss": 2.1103,
+      "step": 1873
+    },
+    {
+      "epoch": 0.600160128102482,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.064655818998091e-05,
+      "loss": 2.4928,
+      "step": 1874
+    },
+    {
+      "epoch": 0.600480384307446,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.061902804859035e-05,
+      "loss": 2.1205,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6008006405124099,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.0591503387024853e-05,
+      "loss": 2.2769,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6011208967173739,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.056398423971292e-05,
+      "loss": 2.2979,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6014411529223379,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.053647064107618e-05,
+      "loss": 1.9518,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6017614091273018,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.0508962625529284e-05,
+      "loss": 2.1684,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6020816653322658,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.0481460227479942e-05,
+      "loss": 2.4742,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6024019215372298,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.0453963481328796e-05,
+      "loss": 2.1223,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6027221777421937,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.042647242146944e-05,
+      "loss": 1.8289,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6030424339471577,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.0398987082288363e-05,
+      "loss": 2.206,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6033626901521217,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.037150749816488e-05,
+      "loss": 1.9582,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6036829463570856,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.0344033703471106e-05,
+      "loss": 2.1088,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6040032025620496,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.0316565732571947e-05,
+      "loss": 1.6995,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6043234587670137,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.028910361982499e-05,
+      "loss": 2.3343,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6046437149719776,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.0261647399580493e-05,
+      "loss": 2.5429,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6049639711769416,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.023419710618138e-05,
+      "loss": 2.6612,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6052842273819056,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.0206752773963128e-05,
+      "loss": 2.0385,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6056044835868695,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.0179314437253756e-05,
+      "loss": 1.9074,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6059247397918335,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.015188213037381e-05,
+      "loss": 2.1805,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6062449959967975,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.012445588763627e-05,
+      "loss": 2.714,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6065652522017614,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.009703574334656e-05,
+      "loss": 2.5523,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6068855084067254,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.0069621731802435e-05,
+      "loss": 2.4568,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6072057646116894,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.0042213887293994e-05,
+      "loss": 2.021,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6075260208166533,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.0014812244103653e-05,
+      "loss": 2.636,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6078462770216173,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.9987416836506034e-05,
+      "loss": 1.8559,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6081665332265813,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9960027698767957e-05,
+      "loss": 1.9964,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6084867894315452,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.993264486514844e-05,
+      "loss": 2.045,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6088070456365092,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9905268369898576e-05,
+      "loss": 2.0631,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6091273018414731,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.9877898247261534e-05,
+      "loss": 1.8805,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6094475580464371,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9850534531472546e-05,
+      "loss": 2.0472,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6097678142514011,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.9823177256758787e-05,
+      "loss": 2.5672,
+      "step": 1904
+    },
+    {
+      "epoch": 0.610088070456365,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.9795826457339393e-05,
+      "loss": 1.8286,
+      "step": 1905
+    },
+    {
+      "epoch": 0.610408326661329,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9768482167425413e-05,
+      "loss": 1.6787,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6107285828662931,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.974114442121973e-05,
+      "loss": 2.3295,
+      "step": 1907
+    },
+    {
+      "epoch": 0.611048839071257,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9713813252917075e-05,
+      "loss": 2.6286,
+      "step": 1908
+    },
+    {
+      "epoch": 0.611369095276221,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9686488696703914e-05,
+      "loss": 2.2819,
+      "step": 1909
+    },
+    {
+      "epoch": 0.611689351481185,
+      "grad_norm": 0.375,
+      "learning_rate": 1.965917078675845e-05,
+      "loss": 2.2437,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6120096076861489,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9631859557250605e-05,
+      "loss": 2.1633,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6123298638911129,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.960455504234191e-05,
+      "loss": 2.1793,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6126501200960769,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.95772572761855e-05,
+      "loss": 2.4874,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6129703763010408,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9549966292926095e-05,
+      "loss": 2.1151,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6132906325060048,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.9522682126699903e-05,
+      "loss": 2.0204,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6136108887109688,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.949540481163461e-05,
+      "loss": 2.5082,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6139311449159327,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9468134381849355e-05,
+      "loss": 2.2812,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6142514011208967,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.944087087145462e-05,
+      "loss": 2.2605,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6145716573258607,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9413614314552293e-05,
+      "loss": 1.7432,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6148919135308246,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.938636474523551e-05,
+      "loss": 2.4423,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6152121697357886,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.935912219758868e-05,
+      "loss": 2.3166,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6155324259407526,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9331886705687464e-05,
+      "loss": 1.7659,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6158526821457165,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9304658303598648e-05,
+      "loss": 2.425,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6161729383506805,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.927743702538017e-05,
+      "loss": 1.8974,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6164931945556446,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.925022290508108e-05,
+      "loss": 1.9298,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6168134507606085,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.922301597674144e-05,
+      "loss": 2.4168,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6171337069655725,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.919581627439232e-05,
+      "loss": 2.331,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6174539631705365,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9168623832055786e-05,
+      "loss": 1.7454,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6177742193755004,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9141438683744785e-05,
+      "loss": 2.4372,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6180944755804644,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.9114260863463145e-05,
+      "loss": 2.159,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6184147317854284,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.908709040520556e-05,
+      "loss": 2.2052,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6187349879903923,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9059927342957468e-05,
+      "loss": 2.6305,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6190552441953563,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9032771710695106e-05,
+      "loss": 2.5403,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6193755004003203,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9005623542385377e-05,
+      "loss": 1.5766,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6196957566052842,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.8978482871985852e-05,
+      "loss": 2.1659,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6200160128102482,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.895134973344476e-05,
+      "loss": 1.9693,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6203362690152122,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8924224160700866e-05,
+      "loss": 2.5216,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6206565252201761,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8897106187683482e-05,
+      "loss": 1.6666,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6209767814251401,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8869995848312445e-05,
+      "loss": 2.3252,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6212970376301041,
+      "grad_norm": 0.5,
+      "learning_rate": 1.8842893176498003e-05,
+      "loss": 2.1655,
+      "step": 1940
+    },
+    {
+      "epoch": 0.621617293835068,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.881579820614082e-05,
+      "loss": 2.7821,
+      "step": 1941
+    },
+    {
+      "epoch": 0.621937550040032,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.878871097113196e-05,
+      "loss": 2.2319,
+      "step": 1942
+    },
+    {
+      "epoch": 0.622257806244996,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.876163150535277e-05,
+      "loss": 2.2277,
+      "step": 1943
+    },
+    {
+      "epoch": 0.62257806244996,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.8734559842674908e-05,
+      "loss": 1.8544,
+      "step": 1944
+    },
+    {
+      "epoch": 0.622898318654924,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.8707496016960262e-05,
+      "loss": 2.4545,
+      "step": 1945
+    },
+    {
+      "epoch": 0.623218574859888,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8680440062060895e-05,
+      "loss": 2.2949,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6235388310648519,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.8653392011819072e-05,
+      "loss": 2.4566,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6238590872698159,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8626351900067135e-05,
+      "loss": 1.9222,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6241793434747798,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.8599319760627494e-05,
+      "loss": 2.7594,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6244995996797438,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.8572295627312616e-05,
+      "loss": 2.2509,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6248198558847078,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.8545279533924934e-05,
+      "loss": 2.6185,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6251401120896717,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8518271514256812e-05,
+      "loss": 2.1968,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6254603682946357,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8491271602090552e-05,
+      "loss": 2.1392,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6257806244995997,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8464279831198288e-05,
+      "loss": 2.3392,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6261008807045636,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.8437296235341954e-05,
+      "loss": 2.3781,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6264211369095276,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8410320848273315e-05,
+      "loss": 2.1818,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6267413931144916,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8383353703733803e-05,
+      "loss": 2.3265,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6270616493194555,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8356394835454596e-05,
+      "loss": 2.1143,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6273819055244195,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.8329444277156483e-05,
+      "loss": 1.686,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6277021617293835,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.8302502062549858e-05,
+      "loss": 2.2293,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6280224179343474,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.8275568225334722e-05,
+      "loss": 2.3718,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6283426741393114,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.824864279920054e-05,
+      "loss": 2.0935,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6286629303442755,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.8221725817826286e-05,
+      "loss": 2.1848,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6289831865492393,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.819481731488038e-05,
+      "loss": 2.5737,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6293034427542034,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8167917324020615e-05,
+      "loss": 2.2572,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6296236989591674,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.814102587889414e-05,
+      "loss": 2.2202,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6299439551641313,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.8114143013137434e-05,
+      "loss": 2.5762,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6302642113690953,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.8087268760376237e-05,
+      "loss": 2.2381,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6305844675740593,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.806040315422548e-05,
+      "loss": 2.179,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6309047237790232,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8033546228289347e-05,
+      "loss": 2.1213,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6312249799839872,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.8006698016161097e-05,
+      "loss": 1.9795,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6315452361889512,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.7979858551423147e-05,
+      "loss": 2.1043,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6318654923939151,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.795302786764692e-05,
+      "loss": 1.9355,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6321857485988791,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7926205998392888e-05,
+      "loss": 2.4791,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6325060048038431,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.7899392977210498e-05,
+      "loss": 2.5029,
+      "step": 1975
+    },
+    {
+      "epoch": 0.632826261008807,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.787258883763812e-05,
+      "loss": 2.1131,
+      "step": 1976
+    },
+    {
+      "epoch": 0.633146517213771,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.7845793613203e-05,
+      "loss": 2.3818,
+      "step": 1977
+    },
+    {
+      "epoch": 0.633466773418735,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.7819007337421272e-05,
+      "loss": 2.3837,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6337870296236989,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.7792230043797845e-05,
+      "loss": 2.3031,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6341072858286629,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.776546176582639e-05,
+      "loss": 2.1035,
+      "step": 1980
+    },
+    {
+      "epoch": 0.634427542033627,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.773870253698933e-05,
+      "loss": 2.4333,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6347477982385908,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.7711952390757736e-05,
+      "loss": 2.0962,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6350680544435549,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.7685211360591353e-05,
+      "loss": 2.0404,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6353883106485189,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.7658479479938493e-05,
+      "loss": 1.8208,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6357085668534828,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7631756782236024e-05,
+      "loss": 1.9733,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6360288230584468,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.760504330090937e-05,
+      "loss": 2.1959,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6363490792634108,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.757833906937237e-05,
+      "loss": 2.3309,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6366693354683747,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.7551644121027317e-05,
+      "loss": 2.45,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6369895916733387,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7524958489264913e-05,
+      "loss": 2.2078,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6373098478783027,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7498282207464185e-05,
+      "loss": 2.45,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6376301040832666,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.7471615308992444e-05,
+      "loss": 2.5175,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6379503602882306,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.7444957827205318e-05,
+      "loss": 1.9297,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6382706164931946,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.741830979544661e-05,
+      "loss": 2.3512,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6385908726981585,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7391671247048315e-05,
+      "loss": 2.4583,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6389111289031225,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.7365042215330584e-05,
+      "loss": 2.1487,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6392313851080864,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.7338422733601636e-05,
+      "loss": 1.959,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6395516413130504,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.7311812835157774e-05,
+      "loss": 2.6023,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6398718975180144,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7285212553283295e-05,
+      "loss": 2.2752,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6401921537229783,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.7258621921250454e-05,
+      "loss": 1.882,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6405124099279423,
+      "grad_norm": 2.875,
+      "learning_rate": 1.7232040972319474e-05,
+      "loss": 2.0141,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6408326661329063,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7205469739738442e-05,
+      "loss": 1.99,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6411529223378702,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7178908256743277e-05,
+      "loss": 1.6528,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6414731785428343,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.7152356556557745e-05,
+      "loss": 2.0187,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6417934347477983,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7125814672393337e-05,
+      "loss": 2.4054,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6421136909527622,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7099282637449273e-05,
+      "loss": 1.9533,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6424339471577262,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.7072760484912482e-05,
+      "loss": 2.1352,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6427542033626902,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.70462482479575e-05,
+      "loss": 2.5819,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6430744595676541,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.701974595974645e-05,
+      "loss": 2.4939,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6433947157726181,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6993253653429063e-05,
+      "loss": 2.0688,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6437149719775821,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.6966771362142532e-05,
+      "loss": 2.3022,
+      "step": 2010
+    },
+    {
+      "epoch": 0.644035228182546,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.694029911901156e-05,
+      "loss": 2.4128,
+      "step": 2011
+    },
+    {
+      "epoch": 0.64435548438751,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.691383695714826e-05,
+      "loss": 1.9259,
+      "step": 2012
+    },
+    {
+      "epoch": 0.644675740592474,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.688738490965212e-05,
+      "loss": 2.2305,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6449959967974379,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.686094300961003e-05,
+      "loss": 1.9657,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6453162530024019,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.683451129009614e-05,
+      "loss": 2.2175,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6456365092073659,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6808089784171863e-05,
+      "loss": 2.3302,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6459567654123298,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.6781678524885882e-05,
+      "loss": 1.7276,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6462770216172938,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6755277545274023e-05,
+      "loss": 1.9889,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6465972778222578,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.6728886878359258e-05,
+      "loss": 2.0543,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6469175340272217,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.670250655715168e-05,
+      "loss": 2.6313,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6472377902321858,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.667613661464842e-05,
+      "loss": 2.2489,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6475580464371498,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.664977708383365e-05,
+      "loss": 2.7663,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6478783026421137,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6623427997678497e-05,
+      "loss": 2.5957,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6481985588470777,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.659708938914102e-05,
+      "loss": 2.0438,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6485188150520417,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.6570761291166207e-05,
+      "loss": 2.1817,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6488390712570056,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.654444373668586e-05,
+      "loss": 2.3496,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6491593274619696,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6518136758618607e-05,
+      "loss": 2.2006,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6494795836669336,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.649184038986986e-05,
+      "loss": 1.8331,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6497998398718975,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.6465554663331746e-05,
+      "loss": 2.3697,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6501200960768615,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.6439279611883075e-05,
+      "loss": 2.2204,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6504403522818255,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.6413015268389323e-05,
+      "loss": 2.3094,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6507606084867894,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.6386761665702564e-05,
+      "loss": 2.6854,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6510808646917534,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.6360518836661418e-05,
+      "loss": 2.6873,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6514011208967174,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6334286814091072e-05,
+      "loss": 2.2687,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6517213771016813,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6308065630803144e-05,
+      "loss": 2.0965,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6520416333066453,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.6281855319595747e-05,
+      "loss": 2.6411,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6523618895116093,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6255655913253357e-05,
+      "loss": 2.0638,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6526821457165732,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.622946744454681e-05,
+      "loss": 2.3375,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6530024019215372,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.620328994623329e-05,
+      "loss": 2.0346,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6533226581265013,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.617712345105624e-05,
+      "loss": 2.1022,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6536429143314652,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6150967991745323e-05,
+      "loss": 2.2443,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6539631705364292,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.6124823601016437e-05,
+      "loss": 1.8261,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6542834267413931,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6098690311571608e-05,
+      "loss": 1.3938,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6546036829463571,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.6072568156098972e-05,
+      "loss": 2.2079,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6549239391513211,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.6046457167272773e-05,
+      "loss": 1.7936,
+      "step": 2045
+    },
+    {
+      "epoch": 0.655244195356285,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6020357377753237e-05,
+      "loss": 2.3265,
+      "step": 2046
+    },
+    {
+      "epoch": 0.655564451561249,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5994268820186624e-05,
+      "loss": 2.1049,
+      "step": 2047
+    },
+    {
+      "epoch": 0.655884707766213,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5968191527205134e-05,
+      "loss": 2.2532,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6562049639711769,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5942125531426848e-05,
+      "loss": 2.5682,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6565252201761409,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.591607086545577e-05,
+      "loss": 2.6043,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6568454763811049,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.589002756188169e-05,
+      "loss": 1.8284,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6571657325860688,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5863995653280187e-05,
+      "loss": 2.0559,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6574859887910328,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5837975172212615e-05,
+      "loss": 2.0543,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6578062449959968,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5811966151226014e-05,
+      "loss": 2.0767,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6581265012009607,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.578596862285308e-05,
+      "loss": 2.3316,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6584467574059247,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.575998261961217e-05,
+      "loss": 2.4229,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6587670136108887,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.5734008174007182e-05,
+      "loss": 2.3858,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6590872698158526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.5708045318527583e-05,
+      "loss": 2.5705,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6594075260208166,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.568209408564834e-05,
+      "loss": 2.8622,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6597277822257807,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5656154507829867e-05,
+      "loss": 2.4677,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6600480384307446,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.5630226617518033e-05,
+      "loss": 2.4751,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6603682946357086,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.560431044714405e-05,
+      "loss": 2.2698,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6606885508406726,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.5578406029124485e-05,
+      "loss": 2.0726,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6610088070456365,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.555251339586123e-05,
+      "loss": 1.9642,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6613290632506005,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.5526632579741386e-05,
+      "loss": 1.9934,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6616493194555645,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5500763613137316e-05,
+      "loss": 2.7078,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6619695756605284,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5474906528406542e-05,
+      "loss": 2.3785,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6622898318654924,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.544906135789172e-05,
+      "loss": 2.1679,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6626100880704564,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.542322813392062e-05,
+      "loss": 2.3335,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6629303442754203,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5397406888806063e-05,
+      "loss": 2.1106,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6632506004803843,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.537159765484587e-05,
+      "loss": 1.8888,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6635708566853483,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5345800464322856e-05,
+      "loss": 2.4113,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6638911128903122,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.5320015349504778e-05,
+      "loss": 1.6792,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6642113690952762,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.529424234264426e-05,
+      "loss": 2.4077,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6645316253002402,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.5268481475978817e-05,
+      "loss": 2.5355,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6648518815052041,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.5242732781730749e-05,
+      "loss": 2.0814,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6651721377101681,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5216996292107144e-05,
+      "loss": 2.5232,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6654923939151322,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5191272039299825e-05,
+      "loss": 2.3116,
+      "step": 2078
+    },
+    {
+      "epoch": 0.665812650120096,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.5165560055485301e-05,
+      "loss": 1.9626,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6661329063250601,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5139860372824741e-05,
+      "loss": 1.8281,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6664531625300241,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.5114173023463932e-05,
+      "loss": 2.2899,
+      "step": 2081
+    },
+    {
+      "epoch": 0.666773418734988,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.5088498039533219e-05,
+      "loss": 1.9186,
+      "step": 2082
+    },
+    {
+      "epoch": 0.667093674939952,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.5062835453147495e-05,
+      "loss": 2.5786,
+      "step": 2083
+    },
+    {
+      "epoch": 0.667413931144916,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5037185296406142e-05,
+      "loss": 2.3842,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6677341873498799,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5011547601392983e-05,
+      "loss": 2.313,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6680544435548439,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4985922400176278e-05,
+      "loss": 2.6968,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6683746997598079,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.4960309724808633e-05,
+      "loss": 1.9038,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6686949559647718,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4934709607327001e-05,
+      "loss": 1.8091,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6690152121697358,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4909122079752632e-05,
+      "loss": 2.426,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6693354683746997,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4883547174091011e-05,
+      "loss": 1.9204,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6696557245796637,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4857984922331847e-05,
+      "loss": 2.1271,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6699759807846277,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4832435356449026e-05,
+      "loss": 2.5165,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6702962369895916,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4806898508400552e-05,
+      "loss": 1.994,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6706164931945556,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.478137441012853e-05,
+      "loss": 2.0746,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6709367493995196,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4755863093559118e-05,
+      "loss": 1.7466,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6712570056044835,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.473036459060248e-05,
+      "loss": 2.247,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6715772618094475,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4704878933152761e-05,
+      "loss": 2.4396,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6718975180144116,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4679406153088032e-05,
+      "loss": 2.3092,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6722177742193755,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4653946282270258e-05,
+      "loss": 2.0661,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6725380304243395,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.4628499352545266e-05,
+      "loss": 2.3798,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6728582866293035,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.4603065395742677e-05,
+      "loss": 2.1861,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6731785428342674,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.4577644443675895e-05,
+      "loss": 1.6774,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6734987990392314,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.4552236528142078e-05,
+      "loss": 2.7617,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6738190552441954,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.452684168092203e-05,
+      "loss": 2.9759,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6741393114491593,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4501459933780249e-05,
+      "loss": 2.4151,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6744595676541233,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4476091318464836e-05,
+      "loss": 2.0188,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6747798238590873,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.4450735866707474e-05,
+      "loss": 2.5587,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6751000800640512,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.4425393610223342e-05,
+      "loss": 2.2319,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6754203362690152,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.4400064580711182e-05,
+      "loss": 2.3308,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6757405924739792,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.4374748809853125e-05,
+      "loss": 1.9952,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6760608486789431,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.4349446329314756e-05,
+      "loss": 2.4647,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6763811048839071,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.4324157170745025e-05,
+      "loss": 2.0062,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6767013610888711,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.429888136577622e-05,
+      "loss": 2.5416,
+      "step": 2113
+    },
+    {
+      "epoch": 0.677021617293835,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4273618946023925e-05,
+      "loss": 2.4491,
+      "step": 2114
+    },
+    {
+      "epoch": 0.677341873498799,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4248369943086998e-05,
+      "loss": 2.3076,
+      "step": 2115
+    },
+    {
+      "epoch": 0.677662129703763,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.422313438854746e-05,
+      "loss": 2.4278,
+      "step": 2116
+    },
+    {
+      "epoch": 0.677982385908727,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.419791231397059e-05,
+      "loss": 2.1202,
+      "step": 2117
+    },
+    {
+      "epoch": 0.678302642113691,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.4172703750904746e-05,
+      "loss": 1.8748,
+      "step": 2118
+    },
+    {
+      "epoch": 0.678622898318655,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.4147508730881403e-05,
+      "loss": 2.3358,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6789431545236189,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4122327285415105e-05,
+      "loss": 2.2464,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6792634107285829,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4097159446003417e-05,
+      "loss": 2.0882,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6795836669335469,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.4072005244126846e-05,
+      "loss": 1.9614,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6799039231385108,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.4046864711248913e-05,
+      "loss": 2.0258,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6802241793434748,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.4021737878815966e-05,
+      "loss": 1.9489,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6805444355484388,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3996624778257287e-05,
+      "loss": 2.5073,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6808646917534027,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.3971525440984914e-05,
+      "loss": 1.9678,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6811849479583667,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3946439898393718e-05,
+      "loss": 2.1684,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6815052041633307,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3921368181861293e-05,
+      "loss": 2.5053,
+      "step": 2128
+    },
+    {
+      "epoch": 0.6818254603682946,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3896310322747957e-05,
+      "loss": 2.5186,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6821457165732586,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3871266352396656e-05,
+      "loss": 2.1431,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6824659727782226,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3846236302133023e-05,
+      "loss": 2.2754,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6827862289831865,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3821220203265223e-05,
+      "loss": 2.122,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6831064851881505,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3796218087084e-05,
+      "loss": 2.2143,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6834267413931145,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.3771229984862608e-05,
+      "loss": 2.6696,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6837469975980784,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.374625592785677e-05,
+      "loss": 2.481,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6840672538030425,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3721295947304607e-05,
+      "loss": 2.4655,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6843875100080064,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.3696350074426708e-05,
+      "loss": 2.3388,
+      "step": 2137
+    },
+    {
+      "epoch": 0.6847077662129704,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3671418340425928e-05,
+      "loss": 2.2813,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6850280224179344,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.3646500776487519e-05,
+      "loss": 2.7145,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6853482786228983,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3621597413778939e-05,
+      "loss": 2.4922,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6856685348278623,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.3596708283449921e-05,
+      "loss": 2.0813,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6859887910328263,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3571833416632396e-05,
+      "loss": 1.8429,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6863090472377902,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.3546972844440441e-05,
+      "loss": 2.3635,
+      "step": 2143
+    },
+    {
+      "epoch": 0.6866293034427542,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.3522126597970244e-05,
+      "loss": 2.161,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6869495596477182,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3497294708300118e-05,
+      "loss": 2.2284,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6872698158526821,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3472477206490355e-05,
+      "loss": 2.4539,
+      "step": 2146
+    },
+    {
+      "epoch": 0.6875900720576461,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.34476741235833e-05,
+      "loss": 2.5645,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6879103282626101,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.3422885490603244e-05,
+      "loss": 2.0052,
+      "step": 2148
+    },
+    {
+      "epoch": 0.688230584467574,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3398111338556412e-05,
+      "loss": 1.8642,
+      "step": 2149
+    },
+    {
+      "epoch": 0.688550840672538,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3373351698430884e-05,
+      "loss": 2.543,
+      "step": 2150
+    },
+    {
+      "epoch": 0.688871096877502,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.3348606601196648e-05,
+      "loss": 2.08,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6891913530824659,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.332387607780543e-05,
+      "loss": 1.7781,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6895116092874299,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3299160159190801e-05,
+      "loss": 2.3629,
+      "step": 2153
+    },
+    {
+      "epoch": 0.689831865492394,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3274458876267998e-05,
+      "loss": 2.1964,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6901521216973578,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.324977225993399e-05,
+      "loss": 1.7751,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6904723779023219,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.3225100341067393e-05,
+      "loss": 1.352,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6907926341072859,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3200443150528438e-05,
+      "loss": 1.7647,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6911128903122498,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.317580071915891e-05,
+      "loss": 1.7487,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6914331465172138,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3151173077782192e-05,
+      "loss": 2.8391,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6917534027221778,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3126560257203097e-05,
+      "loss": 2.3178,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6920736589271417,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3101962288207948e-05,
+      "loss": 1.708,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6923939151321057,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.3077379201564472e-05,
+      "loss": 1.5599,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6927141713370697,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3052811028021788e-05,
+      "loss": 2.1572,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6930344275420336,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3028257798310356e-05,
+      "loss": 2.1006,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6933546837469976,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3003719543141957e-05,
+      "loss": 2.0096,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6936749399519616,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.29791962932096e-05,
+      "loss": 2.0032,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6939951961569255,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.29546880791876e-05,
+      "loss": 2.5821,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6943154523618895,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.2930194931731382e-05,
+      "loss": 2.3261,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6946357085668535,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.290571688147758e-05,
+      "loss": 2.4759,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6949559647718174,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2881253959043926e-05,
+      "loss": 2.61,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6952762209767814,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2856806195029241e-05,
+      "loss": 2.443,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6955964771817454,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.2832373620013352e-05,
+      "loss": 2.3406,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6959167333867093,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.280795626455714e-05,
+      "loss": 2.389,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6962369895916733,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2783554159202404e-05,
+      "loss": 2.4067,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6965572457966374,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2759167334471884e-05,
+      "loss": 2.178,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6968775020016013,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2734795820869208e-05,
+      "loss": 2.0938,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6971977582065653,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.2710439648878855e-05,
+      "loss": 1.9665,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6975180144115293,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2686098848966107e-05,
+      "loss": 2.3108,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6978382706164932,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2661773451577036e-05,
+      "loss": 1.9203,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6981585268214572,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.2637463487138398e-05,
+      "loss": 2.6032,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6984787830264212,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2613168986057728e-05,
+      "loss": 2.3923,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6987990392313851,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2588889978723134e-05,
+      "loss": 1.8835,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6991192954363491,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2564626495503401e-05,
+      "loss": 1.7873,
+      "step": 2183
+    },
+    {
+      "epoch": 0.699439551641313,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.2540378566747874e-05,
+      "loss": 1.8807,
+      "step": 2184
+    },
+    {
+      "epoch": 0.699759807846277,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2516146222786451e-05,
+      "loss": 2.3201,
+      "step": 2185
+    },
+    {
+      "epoch": 0.700080064051241,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2491929493929508e-05,
+      "loss": 1.9649,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7004003202562049,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2467728410467944e-05,
+      "loss": 1.8792,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7007205764611689,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2443543002673022e-05,
+      "loss": 2.3325,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7010408326661329,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.241937330079647e-05,
+      "loss": 2.6154,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7013610888710968,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.2395219335070301e-05,
+      "loss": 2.442,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7016813450760608,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.2371081135706888e-05,
+      "loss": 2.3294,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7020016012810248,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.234695873289887e-05,
+      "loss": 2.4601,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7023218574859887,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.2322852156819137e-05,
+      "loss": 2.0798,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7026421136909528,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2298761437620748e-05,
+      "loss": 2.283,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7029623698959168,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.2274686605436989e-05,
+      "loss": 2.3894,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7032826261008807,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2250627690381213e-05,
+      "loss": 2.4578,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7036028823058447,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2226584722546899e-05,
+      "loss": 2.2,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7039231385108087,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2202557732007569e-05,
+      "loss": 1.6341,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7042433947157726,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.217854674881677e-05,
+      "loss": 2.0122,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7045636509207366,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2154551803007986e-05,
+      "loss": 2.0305,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7048839071257006,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.2130572924594713e-05,
+      "loss": 1.7452,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7052041633306645,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2106610143570271e-05,
+      "loss": 1.9185,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7055244195356285,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.208266348990792e-05,
+      "loss": 2.1781,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7058446757405925,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2058732993560678e-05,
+      "loss": 2.6887,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7061649319455564,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.2034818684461402e-05,
+      "loss": 2.3117,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7064851881505204,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.2010920592522681e-05,
+      "loss": 2.0872,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7068054443554844,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.198703874763683e-05,
+      "loss": 2.1383,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7071257005604483,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1963173179675812e-05,
+      "loss": 2.1324,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7074459567654123,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.1939323918491288e-05,
+      "loss": 2.2598,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7077662129703763,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1915490993914458e-05,
+      "loss": 2.0589,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7080864691753402,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.1891674435756129e-05,
+      "loss": 2.522,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7084067253803042,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1867874273806623e-05,
+      "loss": 2.4149,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7087269815852683,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.1844090537835764e-05,
+      "loss": 2.5084,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7090472377902322,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1820323257592796e-05,
+      "loss": 2.5254,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7093674939951962,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.1796572462806437e-05,
+      "loss": 2.2132,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7096877502001602,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.177283818318472e-05,
+      "loss": 2.0243,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7100080064051241,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.1749120448415085e-05,
+      "loss": 2.1482,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7103282626100881,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.1725419288164222e-05,
+      "loss": 1.9561,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7106485188150521,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.1701734732078118e-05,
+      "loss": 2.4484,
+      "step": 2219
+    },
+    {
+      "epoch": 0.710968775020016,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.1678066809781988e-05,
+      "loss": 2.4673,
+      "step": 2220
+    },
+    {
+      "epoch": 0.71128903122498,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.1654415550880243e-05,
+      "loss": 2.1008,
+      "step": 2221
+    },
+    {
+      "epoch": 0.711609287429944,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.163078098495642e-05,
+      "loss": 2.556,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7119295436349079,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.1607163141573246e-05,
+      "loss": 2.6396,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7122497998398719,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.158356205027245e-05,
+      "loss": 2.5558,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7125700560448359,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.1559977740574861e-05,
+      "loss": 2.4041,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7128903122497998,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1536410241980297e-05,
+      "loss": 2.3616,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7132105684547638,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1512859583967552e-05,
+      "loss": 2.4075,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7135308246597278,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.148932579599436e-05,
+      "loss": 2.4303,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7138510808646917,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1465808907497352e-05,
+      "loss": 2.5852,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7141713370696557,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.1442308947891992e-05,
+      "loss": 2.5451,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7144915932746196,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.141882594657263e-05,
+      "loss": 1.6112,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7148118494795836,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1395359932912348e-05,
+      "loss": 2.3431,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7151321056845477,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1371910936263e-05,
+      "loss": 2.2054,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7154523618895116,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.1348478985955161e-05,
+      "loss": 1.7845,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7157726180944756,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.1325064111298087e-05,
+      "loss": 2.5106,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7160928742994396,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.1301666341579633e-05,
+      "loss": 2.3032,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7164131305044035,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.1278285706066335e-05,
+      "loss": 2.0377,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7167333867093675,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1254922234003224e-05,
+      "loss": 2.2911,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7170536429143315,
+      "grad_norm": 0.5,
+      "learning_rate": 1.1231575954613902e-05,
+      "loss": 1.9858,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7173738991192954,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1208246897100458e-05,
+      "loss": 2.2628,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7176941553242594,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.1184935090643433e-05,
+      "loss": 1.7735,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7180144115292234,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.1161640564401805e-05,
+      "loss": 1.6539,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7183346677341873,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.1138363347512926e-05,
+      "loss": 2.1595,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7186549239391513,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.1115103469092475e-05,
+      "loss": 1.9965,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7189751801441153,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.1091860958234508e-05,
+      "loss": 2.3327,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7192954363490792,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.1068635844011286e-05,
+      "loss": 2.1586,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7196156925540432,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1045428155473347e-05,
+      "loss": 1.9614,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7199359487590072,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.1022237921649429e-05,
+      "loss": 1.8978,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7202562049639711,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0999065171546432e-05,
+      "loss": 2.1097,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7205764611689351,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.0975909934149392e-05,
+      "loss": 2.1752,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7208967173738992,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.0952772238421444e-05,
+      "loss": 2.2443,
+      "step": 2251
+    },
+    {
+      "epoch": 0.721216973578863,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0929652113303754e-05,
+      "loss": 2.4343,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7215372297838271,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.0906549587715546e-05,
+      "loss": 2.1622,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7218574859887911,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0883464690554005e-05,
+      "loss": 2.4342,
+      "step": 2254
+    },
+    {
+      "epoch": 0.722177742193755,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.0860397450694285e-05,
+      "loss": 2.6058,
+      "step": 2255
+    },
+    {
+      "epoch": 0.722497998398719,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0837347896989437e-05,
+      "loss": 2.2438,
+      "step": 2256
+    },
+    {
+      "epoch": 0.722818254603683,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.081431605827041e-05,
+      "loss": 2.1936,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7231385108086469,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0791301963345954e-05,
+      "loss": 2.4717,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7234587670136109,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.0768305641002688e-05,
+      "loss": 2.1485,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7237790232185749,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.0745327120004935e-05,
+      "loss": 2.4316,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7240992794235388,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0722366429094798e-05,
+      "loss": 1.7809,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7244195356285028,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0699423596992056e-05,
+      "loss": 2.3482,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7247397918334668,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.0676498652394157e-05,
+      "loss": 2.037,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7250600480384307,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.065359162397617e-05,
+      "loss": 2.3973,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7253803042433947,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.0630702540390766e-05,
+      "loss": 2.2461,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7257005604483587,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.060783143026814e-05,
+      "loss": 2.4446,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7260208166533226,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0584978322216057e-05,
+      "loss": 1.8983,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7263410728582866,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.0562143244819707e-05,
+      "loss": 1.9454,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7266613290632506,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.0539326226641768e-05,
+      "loss": 2.2344,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7269815852682145,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.0516527296222308e-05,
+      "loss": 2.0809,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7273018414731786,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.0493746482078783e-05,
+      "loss": 1.7608,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7276220976781426,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.0470983812705984e-05,
+      "loss": 2.4922,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7279423538831065,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0448239316576009e-05,
+      "loss": 1.822,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7282626100880705,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.0425513022138203e-05,
+      "loss": 2.2417,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7285828662930345,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0402804957819174e-05,
+      "loss": 1.8278,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7289031224979984,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.0380115152022716e-05,
+      "loss": 2.0429,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7292233787029624,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.0357443633129777e-05,
+      "loss": 2.1444,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7295436349079263,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0334790429498445e-05,
+      "loss": 2.5259,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7298638911128903,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.0312155569463897e-05,
+      "loss": 2.4972,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7301841473178543,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0289539081338335e-05,
+      "loss": 1.7737,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7305044035228182,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.026694099341104e-05,
+      "loss": 2.4937,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7308246597277822,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.024436133394822e-05,
+      "loss": 1.7839,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7311449159327462,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.022180013119306e-05,
+      "loss": 2.5823,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7314651721377101,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0199257413365663e-05,
+      "loss": 2.5663,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7317854283426741,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.0176733208662993e-05,
+      "loss": 2.738,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7321056845476381,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.0154227545258873e-05,
+      "loss": 2.3974,
+      "step": 2286
+    },
+    {
+      "epoch": 0.732425940752602,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.0131740451303936e-05,
+      "loss": 2.6672,
+      "step": 2287
+    },
+    {
+      "epoch": 0.732746196957566,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.0109271954925562e-05,
+      "loss": 1.7195,
+      "step": 2288
+    },
+    {
+      "epoch": 0.73306645316253,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.0086822084227896e-05,
+      "loss": 2.2517,
+      "step": 2289
+    },
+    {
+      "epoch": 0.733386709367494,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.0064390867291774e-05,
+      "loss": 2.5776,
+      "step": 2290
+    },
+    {
+      "epoch": 0.733706965572458,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.00419783321747e-05,
+      "loss": 2.0832,
+      "step": 2291
+    },
+    {
+      "epoch": 0.734027221777422,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0019584506910817e-05,
+      "loss": 1.78,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7343474779823859,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.99720941951086e-06,
+      "loss": 2.3546,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7346677341873499,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.974853097962116e-06,
+      "loss": 2.3111,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7349879903923139,
+      "grad_norm": 0.625,
+      "learning_rate": 9.952515570228432e-06,
+      "loss": 1.9775,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7353082465972778,
+      "grad_norm": 0.494140625,
+      "learning_rate": 9.930196864250102e-06,
+      "loss": 2.029,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7356285028022418,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.90789700794391e-06,
+      "loss": 2.0413,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7359487590072058,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.885616029203046e-06,
+      "loss": 1.8798,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7362690152121697,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.8633539558971e-06,
+      "loss": 1.7159,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7365892714171337,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.84111081587201e-06,
+      "loss": 2.5267,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7369095276220977,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.818886636950037e-06,
+      "loss": 2.7178,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7372297838270616,
+      "grad_norm": 0.4765625,
+      "learning_rate": 9.796681446929698e-06,
+      "loss": 2.4225,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7375500400320256,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.77449527358579e-06,
+      "loss": 2.337,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7378702962369896,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.752328144669316e-06,
+      "loss": 2.5408,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7381905524419535,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.730180087907445e-06,
+      "loss": 1.9689,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7385108086469175,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.708051131003507e-06,
+      "loss": 2.0099,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7388310648518815,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.685941301636927e-06,
+      "loss": 2.1473,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7391513210568454,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.663850627463219e-06,
+      "loss": 1.9139,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7394715772618095,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.641779136113931e-06,
+      "loss": 2.5305,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7397918334667735,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.619726855196603e-06,
+      "loss": 2.4336,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7401120896717374,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.597693812294761e-06,
+      "loss": 2.3245,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7404323458767014,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.57568003496787e-06,
+      "loss": 2.3801,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7407526020816654,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.553685550751294e-06,
+      "loss": 2.4653,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7410728582866293,
+      "grad_norm": 0.494140625,
+      "learning_rate": 9.531710387156254e-06,
+      "loss": 2.639,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7413931144915933,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.509754571669821e-06,
+      "loss": 2.0894,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7417133706965573,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.48781813175486e-06,
+      "loss": 1.8972,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7420336269015212,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.465901094849985e-06,
+      "loss": 1.8963,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7423538831064852,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.44400348836956e-06,
+      "loss": 2.1828,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7426741393114492,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.422125339703635e-06,
+      "loss": 1.7612,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7429943955164131,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.400266676217927e-06,
+      "loss": 2.154,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7433146517213771,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.378427525253781e-06,
+      "loss": 2.6834,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7436349079263411,
+      "grad_norm": 0.5,
+      "learning_rate": 9.356607914128126e-06,
+      "loss": 1.8765,
+      "step": 2322
+    },
+    {
+      "epoch": 0.743955164131305,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.33480787013346e-06,
+      "loss": 2.3637,
+      "step": 2323
+    },
+    {
+      "epoch": 0.744275420336269,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.313027420537809e-06,
+      "loss": 2.5115,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7445956765412329,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.291266592584666e-06,
+      "loss": 1.9597,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7449159327461969,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.269525413493008e-06,
+      "loss": 2.3502,
+      "step": 2326
+    },
+    {
+      "epoch": 0.745236188951161,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.247803910457226e-06,
+      "loss": 2.5435,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7455564451561248,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.226102110647092e-06,
+      "loss": 2.3085,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7458767013610889,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.204420041207746e-06,
+      "loss": 2.0914,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7461969575660529,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.182757729259638e-06,
+      "loss": 1.7195,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7465172137710168,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.161115201898518e-06,
+      "loss": 2.1269,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7468374699759808,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.139492486195364e-06,
+      "loss": 2.4199,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7471577261809448,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.117889609196395e-06,
+      "loss": 2.0749,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7474779823859087,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.09630659792301e-06,
+      "loss": 2.6966,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7477982385908727,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.074743479371761e-06,
+      "loss": 1.7836,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7481184947958367,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.05320028051431e-06,
+      "loss": 2.27,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7484387510008006,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.031677028297413e-06,
+      "loss": 1.792,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7487590072057646,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.010173749642878e-06,
+      "loss": 2.2518,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7490792634107286,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.988690471447503e-06,
+      "loss": 1.1478,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7493995196156925,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.967227220583102e-06,
+      "loss": 2.3766,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7497197758206565,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.945784023896425e-06,
+      "loss": 1.8921,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7500400320256205,
+      "grad_norm": 0.486328125,
+      "learning_rate": 8.924360908209137e-06,
+      "loss": 1.6513,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7503602882305844,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.902957900317783e-06,
+      "loss": 1.5818,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7506805444355484,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.881575026993766e-06,
+      "loss": 2.2554,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7510008006405124,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.860212314983292e-06,
+      "loss": 2.2229,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7513210568454763,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.838869791007367e-06,
+      "loss": 2.2565,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7516413130504404,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.817547481761717e-06,
+      "loss": 2.5122,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7519615692554044,
+      "grad_norm": 0.47265625,
+      "learning_rate": 8.796245413916807e-06,
+      "loss": 2.6242,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7522818254603683,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.774963614117773e-06,
+      "loss": 2.2074,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7526020816653323,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.753702108984405e-06,
+      "loss": 2.5946,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7529223378702963,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.732460925111105e-06,
+      "loss": 2.4789,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7532425940752602,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.711240089066863e-06,
+      "loss": 2.1403,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7535628502802242,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.690039627395195e-06,
+      "loss": 2.0411,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7538831064851882,
+      "grad_norm": 0.494140625,
+      "learning_rate": 8.668859566614154e-06,
+      "loss": 2.4965,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7542033626901521,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.647699933216274e-06,
+      "loss": 2.275,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7545236188951161,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.626560753668531e-06,
+      "loss": 2.3405,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7548438751000801,
+      "grad_norm": 0.498046875,
+      "learning_rate": 8.605442054412318e-06,
+      "loss": 2.0917,
+      "step": 2357
+    },
+    {
+      "epoch": 0.755164131305044,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.584343861863418e-06,
+      "loss": 2.1474,
+      "step": 2358
+    },
+    {
+      "epoch": 0.755484387510008,
+      "grad_norm": 0.4765625,
+      "learning_rate": 8.563266202411948e-06,
+      "loss": 2.2758,
+      "step": 2359
+    },
+    {
+      "epoch": 0.755804643714972,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.542209102422366e-06,
+      "loss": 2.5798,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7561248999199359,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.521172588233378e-06,
+      "loss": 2.485,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7564451561248999,
+      "grad_norm": 0.5,
+      "learning_rate": 8.500156686157975e-06,
+      "loss": 2.354,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7567654123298639,
+      "grad_norm": 0.47265625,
+      "learning_rate": 8.479161422483353e-06,
+      "loss": 1.9641,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7570856685348278,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.458186823470888e-06,
+      "loss": 2.5159,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7574059247397918,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.437232915356114e-06,
+      "loss": 2.1078,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7577261809447559,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.416299724348695e-06,
+      "loss": 2.3243,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7580464371497198,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.395387276632338e-06,
+      "loss": 1.7567,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7583666933546838,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.374495598364876e-06,
+      "loss": 2.1981,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7586869495596478,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.353624715678093e-06,
+      "loss": 2.1722,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7590072057646117,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.332774654677802e-06,
+      "loss": 2.2055,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7593274619695757,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.311945441443759e-06,
+      "loss": 2.5394,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7596477181745396,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.291137102029644e-06,
+      "loss": 2.7369,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7599679743795036,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.270349662463032e-06,
+      "loss": 2.2094,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7602882305844676,
+      "grad_norm": 0.373046875,
+      "learning_rate": 8.249583148745357e-06,
+      "loss": 2.7189,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7606084867894315,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.228837586851859e-06,
+      "loss": 2.4148,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7609287429943955,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.20811300273159e-06,
+      "loss": 2.1258,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7612489991993595,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.187409422307365e-06,
+      "loss": 1.9111,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7615692554043234,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.166726871475713e-06,
+      "loss": 2.2352,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7618895116092874,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.146065376106873e-06,
+      "loss": 2.1225,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7622097678142514,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.125424962044742e-06,
+      "loss": 2.1848,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7625300240192153,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.10480565510683e-06,
+      "loss": 2.2108,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7628502802241793,
+      "grad_norm": 0.484375,
+      "learning_rate": 8.084207481084282e-06,
+      "loss": 2.359,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7631705364291433,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.063630465741775e-06,
+      "loss": 1.9673,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7634907926341072,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.043074634817541e-06,
+      "loss": 2.294,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7638110488390712,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.022540014023306e-06,
+      "loss": 2.1543,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7641313050440353,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.002026629044268e-06,
+      "loss": 2.2717,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7644515612489992,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.981534505539063e-06,
+      "loss": 2.7924,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7647718174539632,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.961063669139738e-06,
+      "loss": 2.4514,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7650920736589272,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.940614145451683e-06,
+      "loss": 2.3244,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7654123298638911,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.920185960053688e-06,
+      "loss": 1.8777,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7657325860688551,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.899779138497796e-06,
+      "loss": 2.4442,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7660528422738191,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.879393706309354e-06,
+      "loss": 2.174,
+      "step": 2392
+    },
+    {
+      "epoch": 0.766373098478783,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.859029688986952e-06,
+      "loss": 1.5029,
+      "step": 2393
+    },
+    {
+      "epoch": 0.766693354683747,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.838687112002397e-06,
+      "loss": 2.0569,
+      "step": 2394
+    },
+    {
+      "epoch": 0.767013610888711,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.818366000800666e-06,
+      "loss": 2.0224,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7673338670936749,
+      "grad_norm": 0.474609375,
+      "learning_rate": 7.798066380799907e-06,
+      "loss": 2.3784,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7676541232986389,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.777788277391348e-06,
+      "loss": 2.0502,
+      "step": 2397
+    },
+    {
+      "epoch": 0.7679743795036029,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.757531715939358e-06,
+      "loss": 1.6753,
+      "step": 2398
+    },
+    {
+      "epoch": 0.7682946357085668,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.73729672178131e-06,
+      "loss": 2.408,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7686148919135308,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.717083320227627e-06,
+      "loss": 2.0858,
+      "step": 2400
+    },
+    {
+      "epoch": 0.7689351481184948,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.69689153656172e-06,
+      "loss": 2.4706,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7692554043234587,
+      "grad_norm": 1.125,
+      "learning_rate": 7.676721396039963e-06,
+      "loss": 2.2135,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7695756605284227,
+      "grad_norm": 0.482421875,
+      "learning_rate": 7.65657292389163e-06,
+      "loss": 2.5483,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7698959167333868,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.636446145318945e-06,
+      "loss": 2.4565,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7702161729383507,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.6163410854969476e-06,
+      "loss": 2.0011,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7705364291433147,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.596257769573534e-06,
+      "loss": 2.4179,
+      "step": 2406
+    },
+    {
+      "epoch": 0.7708566853482787,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.576196222669396e-06,
+      "loss": 2.2249,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7711769415532426,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.5561564698780065e-06,
+      "loss": 2.3321,
+      "step": 2408
+    },
+    {
+      "epoch": 0.7714971977582066,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.536138536265563e-06,
+      "loss": 2.2056,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7718174539631706,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.516142446870989e-06,
+      "loss": 2.3361,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7721377101681345,
+      "grad_norm": 0.4765625,
+      "learning_rate": 7.496168226705852e-06,
+      "loss": 1.8415,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7724579663730985,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.476215900754413e-06,
+      "loss": 2.214,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7727782225780625,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.456285493973506e-06,
+      "loss": 2.1685,
+      "step": 2413
+    },
+    {
+      "epoch": 0.7730984787830264,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.4363770312925625e-06,
+      "loss": 1.6529,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7734187349879904,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.416490537613571e-06,
+      "loss": 2.5159,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7737389911929544,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.396626037811042e-06,
+      "loss": 1.9631,
+      "step": 2416
+    },
+    {
+      "epoch": 0.7740592473979183,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.376783556731948e-06,
+      "loss": 2.3239,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7743795036028823,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.35696311919577e-06,
+      "loss": 1.8956,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7746997598078462,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.337164749994368e-06,
+      "loss": 2.2789,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7750200160128102,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.317388473892023e-06,
+      "loss": 2.1517,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7753402722177742,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.297634315625377e-06,
+      "loss": 1.8434,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7756605284227381,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.277902299903405e-06,
+      "loss": 2.4362,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7759807846277021,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.2581924514073916e-06,
+      "loss": 2.2693,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7763010408326662,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.238504794790893e-06,
+      "loss": 2.4455,
+      "step": 2424
+    },
+    {
+      "epoch": 0.77662129703763,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.218839354679683e-06,
+      "loss": 2.5956,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7769415532425941,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.199196155671795e-06,
+      "loss": 2.187,
+      "step": 2426
+    },
+    {
+      "epoch": 0.7772618094475581,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.179575222337393e-06,
+      "loss": 2.5182,
+      "step": 2427
+    },
+    {
+      "epoch": 0.777582065652522,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.159976579218819e-06,
+      "loss": 1.9639,
+      "step": 2428
+    },
+    {
+      "epoch": 0.777902321857486,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.140400250830529e-06,
+      "loss": 2.2367,
+      "step": 2429
+    },
+    {
+      "epoch": 0.77822257806245,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.12084626165907e-06,
+      "loss": 2.4143,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7785428342674139,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.10131463616302e-06,
+      "loss": 1.9487,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7788630904723779,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.0818053987730395e-06,
+      "loss": 2.0972,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7791833466773419,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.062318573891716e-06,
+      "loss": 2.4615,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7795036028823058,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.042854185893674e-06,
+      "loss": 2.0897,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7798238590872698,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.023412259125414e-06,
+      "loss": 2.435,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7801441152922338,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.00399281790537e-06,
+      "loss": 2.3402,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7804643714971977,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.984595886523848e-06,
+      "loss": 2.1758,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7807846277021617,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.965221489243007e-06,
+      "loss": 2.3811,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7811048839071257,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.94586965029678e-06,
+      "loss": 1.8303,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7814251401120896,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.926540393890948e-06,
+      "loss": 2.0528,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7817453963170536,
+      "grad_norm": 0.3828125,
+      "learning_rate": 6.907233744202985e-06,
+      "loss": 2.1735,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7820656525220177,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.887949725382117e-06,
+      "loss": 2.369,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7823859087269815,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.8686883615492666e-06,
+      "loss": 2.4711,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7827061649319456,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.84944967679701e-06,
+      "loss": 2.5019,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7830264211369096,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.830233695189539e-06,
+      "loss": 2.4643,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7833466773418735,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.8110404407626946e-06,
+      "loss": 1.7722,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7836669335468375,
+      "grad_norm": 0.365234375,
+      "learning_rate": 6.791869937523831e-06,
+      "loss": 1.7393,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7839871897518015,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.772722209451904e-06,
+      "loss": 1.9059,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7843074459567654,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.753597280497329e-06,
+      "loss": 1.9944,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7846277021617294,
+      "grad_norm": 0.44921875,
+      "learning_rate": 6.734495174582037e-06,
+      "loss": 2.2997,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7849479583666934,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.715415915599399e-06,
+      "loss": 2.5738,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7852682145716573,
+      "grad_norm": 0.474609375,
+      "learning_rate": 6.69635952741422e-06,
+      "loss": 2.5298,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7855884707766213,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.677326033862666e-06,
+      "loss": 2.172,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7859087269815853,
+      "grad_norm": 0.37109375,
+      "learning_rate": 6.658315458752315e-06,
+      "loss": 1.9617,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7862289831865492,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.639327825862035e-06,
+      "loss": 2.0634,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7865492393915132,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.620363158942016e-06,
+      "loss": 2.2057,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7868694955964772,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.601421481713724e-06,
+      "loss": 2.4316,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7871897518014411,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.582502817869868e-06,
+      "loss": 2.2507,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7875100080064051,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.563607191074345e-06,
+      "loss": 2.264,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7878302642113691,
+      "grad_norm": 0.408203125,
+      "learning_rate": 6.5447346249622926e-06,
+      "loss": 2.4022,
+      "step": 2460
+    },
+    {
+      "epoch": 0.788150520416333,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.52588514313994e-06,
+      "loss": 2.6929,
+      "step": 2461
+    },
+    {
+      "epoch": 0.788470776621297,
+      "grad_norm": 0.5,
+      "learning_rate": 6.507058769184699e-06,
+      "loss": 2.1782,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7887910328262611,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.488255526645029e-06,
+      "loss": 1.8834,
+      "step": 2463
+    },
+    {
+      "epoch": 0.789111289031225,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.46947543904049e-06,
+      "loss": 2.4088,
+      "step": 2464
+    },
+    {
+      "epoch": 0.789431545236189,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.450718529861663e-06,
+      "loss": 2.7434,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7897518014411529,
+      "grad_norm": 0.365234375,
+      "learning_rate": 6.431984822570147e-06,
+      "loss": 2.0478,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7900720576461169,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.413274340598496e-06,
+      "loss": 2.5495,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7903923138510809,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.394587107350258e-06,
+      "loss": 1.9307,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7907125700560448,
+      "grad_norm": 0.37890625,
+      "learning_rate": 6.3759231461998525e-06,
+      "loss": 1.8534,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7910328262610088,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.357282480492616e-06,
+      "loss": 1.9793,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7913530824659728,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.338665133544744e-06,
+      "loss": 2.3047,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7916733386709367,
+      "grad_norm": 0.75,
+      "learning_rate": 6.32007112864326e-06,
+      "loss": 2.4341,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7919935948759007,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.3015004890459915e-06,
+      "loss": 1.6601,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7923138510808647,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.282953237981551e-06,
+      "loss": 2.1922,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7926341072858286,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.264429398649263e-06,
+      "loss": 1.6936,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7929543634907926,
+      "grad_norm": 0.48828125,
+      "learning_rate": 6.245928994219222e-06,
+      "loss": 1.8489,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7932746196957566,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.227452047832155e-06,
+      "loss": 2.7892,
+      "step": 2477
+    },
+    {
+      "epoch": 0.7935948759007205,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.208998582599476e-06,
+      "loss": 2.4221,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7939151321056845,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.1905686216032264e-06,
+      "loss": 2.2199,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7942353883106485,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.1721621878960526e-06,
+      "loss": 2.4266,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7945556445156124,
+      "grad_norm": 0.43359375,
+      "learning_rate": 6.153779304501139e-06,
+      "loss": 1.9723,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7948759007205765,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.135419994412272e-06,
+      "loss": 2.4696,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7951961569255405,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.1170842805936905e-06,
+      "loss": 2.081,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7955164131305044,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.098772185980159e-06,
+      "loss": 2.4716,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7958366693354684,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.080483733476883e-06,
+      "loss": 1.9128,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7961569255404324,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.062218945959497e-06,
+      "loss": 2.4175,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7964771817453963,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.04397784627404e-06,
+      "loss": 2.0246,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7967974379503603,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.025760457236926e-06,
+      "loss": 2.3537,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7971176941553243,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.007566801634876e-06,
+      "loss": 2.7204,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7974379503602882,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.989396902224986e-06,
+      "loss": 2.1869,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7977582065652522,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.97125078173458e-06,
+      "loss": 2.5756,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7980784627702162,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.953128462861268e-06,
+      "loss": 2.2436,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7983987189751801,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.935029968272885e-06,
+      "loss": 1.6643,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7987189751801441,
+      "grad_norm": 0.478515625,
+      "learning_rate": 5.916955320607465e-06,
+      "loss": 2.1245,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7990392313851081,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.898904542473197e-06,
+      "loss": 2.1199,
+      "step": 2495
+    },
+    {
+      "epoch": 0.799359487590072,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.8808776564484524e-06,
+      "loss": 2.4544,
+      "step": 2496
+    },
+    {
+      "epoch": 0.799679743795036,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.862874685081673e-06,
+      "loss": 2.386,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.844895650891416e-06,
+      "loss": 2.2664,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8003202562049639,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.826940576366291e-06,
+      "loss": 2.2293,
+      "step": 2499
+    },
+    {
+      "epoch": 0.800640512409928,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.809009483964936e-06,
+      "loss": 1.8299,
+      "step": 2500
+    },
+    {
+      "epoch": 0.800960768614892,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.791102396115991e-06,
+      "loss": 2.4019,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8012810248198559,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.773219335218083e-06,
+      "loss": 2.5534,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8016012810248199,
+      "grad_norm": 0.478515625,
+      "learning_rate": 5.75536032363975e-06,
+      "loss": 2.1044,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8019215372297839,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.737525383719499e-06,
+      "loss": 1.7572,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8022417934347478,
+      "grad_norm": 0.46484375,
+      "learning_rate": 5.719714537765688e-06,
+      "loss": 2.5762,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8025620496397118,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.7019278080565535e-06,
+      "loss": 2.3808,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8028823058446758,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.684165216840165e-06,
+      "loss": 2.4309,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8032025620496397,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.666426786334405e-06,
+      "loss": 2.4435,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8035228182546037,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.648712538726911e-06,
+      "loss": 2.1253,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8038430744595677,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.631022496175115e-06,
+      "loss": 2.5646,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8041633306645316,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.61335668080612e-06,
+      "loss": 2.106,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8044835868694956,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.595715114716782e-06,
+      "loss": 2.2236,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8048038430744595,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.578097819973577e-06,
+      "loss": 2.451,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8051240992794235,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.560504818612641e-06,
+      "loss": 2.1605,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8054443554843875,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.5429361326397285e-06,
+      "loss": 2.5384,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8057646116893514,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.525391784030182e-06,
+      "loss": 2.0079,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8060848678943154,
+      "grad_norm": 0.375,
+      "learning_rate": 5.507871794728872e-06,
+      "loss": 1.5529,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8064051240992794,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.4903761866502475e-06,
+      "loss": 2.1881,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8067253803042433,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.47290498167822e-06,
+      "loss": 2.1702,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8070456365092074,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.455458201666197e-06,
+      "loss": 2.2788,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8073658927141714,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.43803586843703e-06,
+      "loss": 2.2958,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8076861489191353,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.420638003783002e-06,
+      "loss": 2.4007,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8080064051240993,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.403264629465757e-06,
+      "loss": 2.1257,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8083266613290633,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.3859157672163564e-06,
+      "loss": 2.2356,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8086469175340272,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.368591438735149e-06,
+      "loss": 2.5254,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8089671737389912,
+      "grad_norm": 0.5,
+      "learning_rate": 5.3512916656918485e-06,
+      "loss": 2.1586,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8092874299439552,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.334016469725409e-06,
+      "loss": 2.3058,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8096076861489191,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.31676587244406e-06,
+      "loss": 2.061,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8099279423538831,
+      "grad_norm": 0.470703125,
+      "learning_rate": 5.299539895425271e-06,
+      "loss": 2.6985,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8102481985588471,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.282338560215711e-06,
+      "loss": 1.9756,
+      "step": 2530
+    },
+    {
+      "epoch": 0.810568454763811,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.265161888331205e-06,
+      "loss": 2.1155,
+      "step": 2531
+    },
+    {
+      "epoch": 0.810888710968775,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.248009901256776e-06,
+      "loss": 2.2378,
+      "step": 2532
+    },
+    {
+      "epoch": 0.811208967173739,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.230882620446517e-06,
+      "loss": 2.1956,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8115292233787029,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.213780067323654e-06,
+      "loss": 2.4922,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8118494795836669,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.19670226328047e-06,
+      "loss": 2.0935,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8121697357886309,
+      "grad_norm": 0.470703125,
+      "learning_rate": 5.1796492296782905e-06,
+      "loss": 1.7568,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8124899919935948,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.162620987847466e-06,
+      "loss": 2.527,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8128102481985588,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.145617559087332e-06,
+      "loss": 2.5302,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8131305044035229,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.128638964666166e-06,
+      "loss": 2.2782,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8134507606084868,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.111685225821233e-06,
+      "loss": 2.4849,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8137710168134508,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.0947563637586545e-06,
+      "loss": 2.1828,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8140912730184148,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.077852399653463e-06,
+      "loss": 2.3564,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8144115292233787,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.060973354649548e-06,
+      "loss": 1.8303,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8147317854283427,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.044119249859627e-06,
+      "loss": 2.3438,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8150520416333067,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.027290106365204e-06,
+      "loss": 1.9135,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8153722978382706,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.010485945216601e-06,
+      "loss": 2.1691,
+      "step": 2546
+    },
+    {
+      "epoch": 0.8156925540432346,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.993706787432844e-06,
+      "loss": 2.5787,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8160128102481986,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.976952654001718e-06,
+      "loss": 2.1237,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8163330664531625,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.960223565879696e-06,
+      "loss": 2.3624,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8166533226581265,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.94351954399192e-06,
+      "loss": 1.776,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8169735788630905,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.926840609232183e-06,
+      "loss": 2.3213,
+      "step": 2551
+    },
+    {
+      "epoch": 0.8172938350680544,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.910186782462908e-06,
+      "loss": 2.7406,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8176140912730184,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.8935580845150765e-06,
+      "loss": 2.2117,
+      "step": 2553
+    },
+    {
+      "epoch": 0.8179343474779824,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.876954536188297e-06,
+      "loss": 1.6675,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8182546036829463,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.860376158250659e-06,
+      "loss": 2.0331,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8185748598879103,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.84382297143881e-06,
+      "loss": 2.0956,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8188951160928744,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.827294996457873e-06,
+      "loss": 2.4513,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8192153722978383,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.810792253981439e-06,
+      "loss": 2.3092,
+      "step": 2558
+    },
+    {
+      "epoch": 0.8195356285028023,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.79431476465152e-06,
+      "loss": 1.9795,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8198558847077662,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.777862549078579e-06,
+      "loss": 2.2858,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8201761409127302,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.7614356278414235e-06,
+      "loss": 2.7803,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8204963971176942,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.745034021487252e-06,
+      "loss": 2.0644,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8208166533226581,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.728657750531581e-06,
+      "loss": 1.9754,
+      "step": 2563
+    },
+    {
+      "epoch": 0.8211369095276221,
+      "grad_norm": 1.390625,
+      "learning_rate": 4.712306835458249e-06,
+      "loss": 2.4305,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8214571657325861,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.695981296719376e-06,
+      "loss": 2.0684,
+      "step": 2565
+    },
+    {
+      "epoch": 0.82177742193755,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.679681154735341e-06,
+      "loss": 2.2563,
+      "step": 2566
+    },
+    {
+      "epoch": 0.822097678142514,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.6634064298947315e-06,
+      "loss": 2.162,
+      "step": 2567
+    },
+    {
+      "epoch": 0.822417934347478,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.647157142554396e-06,
+      "loss": 2.3221,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8227381905524419,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.6309333130393125e-06,
+      "loss": 2.307,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8230584467574059,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.614734961642644e-06,
+      "loss": 2.0686,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8233787029623699,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.598562108625676e-06,
+      "loss": 2.1714,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8236989591673338,
+      "grad_norm": 1.6796875,
+      "learning_rate": 4.582414774217811e-06,
+      "loss": 2.3787,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8240192153722978,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.5662929786165014e-06,
+      "loss": 2.2512,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8243394715772618,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.550196741987306e-06,
+      "loss": 2.0636,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8246597277822257,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.534126084463755e-06,
+      "loss": 1.9996,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8249799839871897,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.518081026147439e-06,
+      "loss": 1.8507,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8253002401921538,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.502061587107889e-06,
+      "loss": 2.1717,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8256204963971177,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.4860677873826025e-06,
+      "loss": 2.3655,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8259407526020817,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.4700996469770165e-06,
+      "loss": 2.014,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8262610088070457,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.454157185864469e-06,
+      "loss": 2.4824,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8265812650120096,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.438240423986154e-06,
+      "loss": 2.2044,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8269015212169736,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.422349381251162e-06,
+      "loss": 2.2686,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8272217774219376,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.406484077536377e-06,
+      "loss": 2.5848,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8275420336269015,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.3906445326865085e-06,
+      "loss": 1.95,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8278622898318655,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.374830766514037e-06,
+      "loss": 1.8195,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8281825460368295,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.3590427987992126e-06,
+      "loss": 1.9135,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8285028022417934,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.3432806492899826e-06,
+      "loss": 2.2409,
+      "step": 2587
+    },
+    {
+      "epoch": 0.8288230584467574,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.327544337702047e-06,
+      "loss": 1.7218,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8291433146517214,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.3118338837187415e-06,
+      "loss": 2.4292,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8294635708566853,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.296149306991098e-06,
+      "loss": 2.1419,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8297838270616493,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.280490627137751e-06,
+      "loss": 2.5364,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8301040832666133,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.264857863744956e-06,
+      "loss": 2.121,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8304243394715772,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.24925103636655e-06,
+      "loss": 2.5428,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8307445956765412,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.233670164523937e-06,
+      "loss": 2.0423,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8310648518815053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.218115267706021e-06,
+      "loss": 2.4526,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8313851080864691,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.2025863653692706e-06,
+      "loss": 2.3808,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8317053642914332,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.1870834769375925e-06,
+      "loss": 2.3936,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8320256204963972,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.171606621802377e-06,
+      "loss": 2.443,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8323458767013611,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.1561558193224435e-06,
+      "loss": 2.0169,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8326661329063251,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.14073108882404e-06,
+      "loss": 2.1607,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8329863891112891,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.125332449600766e-06,
+      "loss": 2.3544,
+      "step": 2601
+    },
+    {
+      "epoch": 0.833306645316253,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.109959920913637e-06,
+      "loss": 2.0641,
+      "step": 2602
+    },
+    {
+      "epoch": 0.833626901521217,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.094613521990956e-06,
+      "loss": 2.1128,
+      "step": 2603
+    },
+    {
+      "epoch": 0.833947157726181,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.079293272028395e-06,
+      "loss": 1.7678,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8342674139311449,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.063999190188869e-06,
+      "loss": 2.2154,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8345876701361089,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.04873129560259e-06,
+      "loss": 2.4884,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8349079263410728,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.033489607367011e-06,
+      "loss": 2.6735,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8352281825460368,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.018274144546802e-06,
+      "loss": 2.579,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8355484387510008,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.003084926173819e-06,
+      "loss": 2.5984,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8358686949559647,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.987921971247122e-06,
+      "loss": 2.7333,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8361889511609287,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.972785298732881e-06,
+      "loss": 2.0657,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8365092073658927,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.9576749275644175e-06,
+      "loss": 2.1598,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8368294635708566,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.942590876642146e-06,
+      "loss": 2.1292,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8371497197758206,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.927533164833561e-06,
+      "loss": 2.3755,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8374699759807847,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.91250181097321e-06,
+      "loss": 2.1559,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8377902321857486,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.897496833862677e-06,
+      "loss": 2.4086,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8381104883907126,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.882518252270531e-06,
+      "loss": 2.1332,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8384307445956766,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.867566084932367e-06,
+      "loss": 1.9732,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8387510008006405,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.852640350550696e-06,
+      "loss": 2.4946,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8390712570056045,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.837741067794992e-06,
+      "loss": 2.6402,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8393915132105685,
+      "grad_norm": 1.625,
+      "learning_rate": 3.822868255301637e-06,
+      "loss": 2.6491,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8397117694155324,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.808021931673905e-06,
+      "loss": 2.4719,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8400320256204964,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.7932021154819225e-06,
+      "loss": 1.9543,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8403522818254604,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.7784088252626914e-06,
+      "loss": 1.8275,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8406725380304243,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.7636420795199996e-06,
+      "loss": 1.8942,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8409927942353883,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.7489018967244494e-06,
+      "loss": 1.9354,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8413130504403523,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.7341882953134206e-06,
+      "loss": 1.8982,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8416333066453162,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.7195012936910383e-06,
+      "loss": 2.8696,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8419535628502802,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.7048409102281577e-06,
+      "loss": 2.3575,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8422738190552442,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.690207163262341e-06,
+      "loss": 2.4408,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8425940752602081,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.675600071097818e-06,
+      "loss": 2.1077,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8429143314651721,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.6610196520055124e-06,
+      "loss": 2.1918,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8432345876701361,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.646465924222947e-06,
+      "loss": 2.4826,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8435548438751,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.631938905954277e-06,
+      "loss": 2.3311,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8438751000800641,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.6174386153702473e-06,
+      "loss": 2.4787,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8441953562850281,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.6029650706081775e-06,
+      "loss": 1.9944,
+      "step": 2636
+    },
+    {
+      "epoch": 0.844515612489992,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.588518289771908e-06,
+      "loss": 2.4228,
+      "step": 2637
+    },
+    {
+      "epoch": 0.844835868694956,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.5740982909318406e-06,
+      "loss": 2.4106,
+      "step": 2638
+    },
+    {
+      "epoch": 0.84515612489992,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.5597050921248347e-06,
+      "loss": 2.4529,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8454763811048839,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.5453387113542737e-06,
+      "loss": 2.5682,
+      "step": 2640
+    },
+    {
+      "epoch": 0.8457966373098479,
+      "grad_norm": 1.609375,
+      "learning_rate": 3.5309991665899545e-06,
+      "loss": 2.0301,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8461168935148119,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.516686475768127e-06,
+      "loss": 2.1756,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8464371497197758,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.502400656791455e-06,
+      "loss": 1.9134,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8467574059247398,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.4881417275289847e-06,
+      "loss": 2.3704,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8470776621297038,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.4739097058161114e-06,
+      "loss": 2.5802,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8473979183346677,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.459704609454614e-06,
+      "loss": 1.5905,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8477181745396317,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.445526456212553e-06,
+      "loss": 2.2604,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8480384307445957,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.4313752638243075e-06,
+      "loss": 2.3032,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8483586869495596,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.41725104999053e-06,
+      "loss": 2.215,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8486789431545236,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.4031538323781254e-06,
+      "loss": 2.3237,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8489991993594876,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.389083628620235e-06,
+      "loss": 1.5601,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8493194555644515,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.3750404563162085e-06,
+      "loss": 2.3487,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8496397117694156,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.3610243330315733e-06,
+      "loss": 1.9077,
+      "step": 2653
+    },
+    {
+      "epoch": 0.8499599679743794,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.347035276298052e-06,
+      "loss": 1.879,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8502802241793435,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.3330733036134765e-06,
+      "loss": 2.0285,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8506004803843075,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.3191384324418236e-06,
+      "loss": 2.5377,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8509207365892714,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.3052306802131644e-06,
+      "loss": 1.8052,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8512409927942354,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.291350064323659e-06,
+      "loss": 1.8514,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8515612489991994,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.277496602135496e-06,
+      "loss": 2.2399,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8518815052041633,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.2636703109769406e-06,
+      "loss": 2.274,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8522017614091273,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.249871208142238e-06,
+      "loss": 2.2057,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8525220176140913,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.236099310891638e-06,
+      "loss": 2.5783,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8528422738190552,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.2223546364513636e-06,
+      "loss": 2.2508,
+      "step": 2663
+    },
+    {
+      "epoch": 0.8531625300240192,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.2086372020135815e-06,
+      "loss": 2.5538,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8534827862289832,
+      "grad_norm": 0.625,
+      "learning_rate": 3.194947024736392e-06,
+      "loss": 2.0804,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8538030424339471,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.1812841217438023e-06,
+      "loss": 2.5366,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8541232986389111,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.1676485101256844e-06,
+      "loss": 2.1118,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8544435548438751,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.1540402069378084e-06,
+      "loss": 2.3653,
+      "step": 2668
+    },
+    {
+      "epoch": 0.854763811048839,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.140459229201753e-06,
+      "loss": 2.4334,
+      "step": 2669
+    },
+    {
+      "epoch": 0.855084067253803,
+      "grad_norm": 0.5,
+      "learning_rate": 3.1269055939049356e-06,
+      "loss": 2.5745,
+      "step": 2670
+    },
+    {
+      "epoch": 0.855404323458767,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.11337931800057e-06,
+      "loss": 1.7917,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8557245796637309,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.0998804184076495e-06,
+      "loss": 2.3825,
+      "step": 2672
+    },
+    {
+      "epoch": 0.856044835868695,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.0864089120109186e-06,
+      "loss": 2.1879,
+      "step": 2673
+    },
+    {
+      "epoch": 0.856365092073659,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.07296481566087e-06,
+      "loss": 2.2279,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8566853482786229,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.059548146173691e-06,
+      "loss": 2.1929,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8570056044835869,
+      "grad_norm": 0.625,
+      "learning_rate": 3.046158920331277e-06,
+      "loss": 2.2886,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8573258606885509,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.032797154881198e-06,
+      "loss": 2.8405,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8576461168935148,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.019462866536668e-06,
+      "loss": 1.712,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8579663730984788,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.0061560719765376e-06,
+      "loss": 2.4549,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8582866293034428,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.9928767878452698e-06,
+      "loss": 2.2048,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8586068855084067,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.9796250307528973e-06,
+      "loss": 2.1138,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8589271417133707,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9664008172750557e-06,
+      "loss": 2.0606,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8592473979183347,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.9532041639528967e-06,
+      "loss": 2.1709,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8595676541232986,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.9400350872931138e-06,
+      "loss": 2.5132,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8598879103282626,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.926893603767908e-06,
+      "loss": 2.2391,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8602081665332266,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.9137797298149584e-06,
+      "loss": 2.4567,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8605284227381905,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.9006934818374187e-06,
+      "loss": 2.511,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8608486789431545,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.8876348762038846e-06,
+      "loss": 2.4784,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8611689351481185,
+      "grad_norm": 0.625,
+      "learning_rate": 2.874603929248365e-06,
+      "loss": 1.7693,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8614891913530824,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.8616006572702857e-06,
+      "loss": 2.4669,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8618094475580464,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.848625076534453e-06,
+      "loss": 1.9782,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8621297037630105,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.835677203271037e-06,
+      "loss": 2.2473,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8624499599679744,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.8227570536755425e-06,
+      "loss": 2.3449,
+      "step": 2693
+    },
+    {
+      "epoch": 0.8627702161729384,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.8098646439088106e-06,
+      "loss": 2.292,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8630904723779024,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.796999990096974e-06,
+      "loss": 2.395,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8634107285828663,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.7841631083314584e-06,
+      "loss": 2.19,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8637309847878303,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.771354014668931e-06,
+      "loss": 1.5913,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8640512409927943,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.75857272513132e-06,
+      "loss": 2.1674,
+      "step": 2698
+    },
+    {
+      "epoch": 0.8643714971977582,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.74581925570577e-06,
+      "loss": 2.3245,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8646917534027222,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.733093622344629e-06,
+      "loss": 1.8666,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8650120096076861,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.7203958409654227e-06,
+      "loss": 2.2844,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8653322658126501,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.707725927450844e-06,
+      "loss": 1.9818,
+      "step": 2702
+    },
+    {
+      "epoch": 0.8656525220176141,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6950838976487287e-06,
+      "loss": 1.9651,
+      "step": 2703
+    },
+    {
+      "epoch": 0.865972778222578,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.68246976737202e-06,
+      "loss": 2.1429,
+      "step": 2704
+    },
+    {
+      "epoch": 0.866293034427542,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.6698835523987832e-06,
+      "loss": 2.1287,
+      "step": 2705
+    },
+    {
+      "epoch": 0.866613290632506,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.657325268472155e-06,
+      "loss": 2.2263,
+      "step": 2706
+    },
+    {
+      "epoch": 0.8669335468374699,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.6447949313003456e-06,
+      "loss": 2.1328,
+      "step": 2707
+    },
+    {
+      "epoch": 0.8672538030424339,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.6322925565565954e-06,
+      "loss": 2.1588,
+      "step": 2708
+    },
+    {
+      "epoch": 0.8675740592473979,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.619818159879178e-06,
+      "loss": 2.1747,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8678943154523618,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.607371756871374e-06,
+      "loss": 1.9288,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8682145716573259,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.594953363101435e-06,
+      "loss": 2.3231,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8685348278622899,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.5825629941025887e-06,
+      "loss": 2.2347,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8688550840672538,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.5702006653730072e-06,
+      "loss": 2.5465,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8691753402722178,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.557866392375788e-06,
+      "loss": 2.163,
+      "step": 2714
+    },
+    {
+      "epoch": 0.8694955964771818,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.545560190538937e-06,
+      "loss": 2.5304,
+      "step": 2715
+    },
+    {
+      "epoch": 0.8698158526821457,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.5332820752553498e-06,
+      "loss": 2.4132,
+      "step": 2716
+    },
+    {
+      "epoch": 0.8701361088871097,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.521032061882786e-06,
+      "loss": 2.2646,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8704563650920737,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.5088101657438605e-06,
+      "loss": 2.1072,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8707766212970376,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.4966164021260092e-06,
+      "loss": 2.3744,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8710968775020016,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.4844507862814835e-06,
+      "loss": 1.145,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8714171337069656,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.472313333427334e-06,
+      "loss": 1.9044,
+      "step": 2721
+    },
+    {
+      "epoch": 0.8717373899119295,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.4602040587453774e-06,
+      "loss": 1.9285,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8720576461168935,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.4481229773821835e-06,
+      "loss": 2.2034,
+      "step": 2723
+    },
+    {
+      "epoch": 0.8723779023218575,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.436070104449059e-06,
+      "loss": 2.475,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8726981585268214,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.4240454550220367e-06,
+      "loss": 2.528,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8730184147317854,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.412049044141823e-06,
+      "loss": 2.067,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8733386709367494,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.4000808868138212e-06,
+      "loss": 1.8833,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8736589271417133,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.3881409980080954e-06,
+      "loss": 2.4556,
+      "step": 2728
+    },
+    {
+      "epoch": 0.8739791833466773,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.3762293926593433e-06,
+      "loss": 1.9098,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8742994395516414,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.3643460856668823e-06,
+      "loss": 2.4019,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8746196957566053,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.3524910918946484e-06,
+      "loss": 2.2306,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8749399519615693,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.340664426171149e-06,
+      "loss": 2.2344,
+      "step": 2732
+    },
+    {
+      "epoch": 0.8752602081665333,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.3288661032894564e-06,
+      "loss": 2.1087,
+      "step": 2733
+    },
+    {
+      "epoch": 0.8755804643714972,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.3170961380071993e-06,
+      "loss": 2.0231,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8759007205764612,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.305354545046537e-06,
+      "loss": 2.5689,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8762209767814252,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.2936413390941325e-06,
+      "loss": 2.1854,
+      "step": 2736
+    },
+    {
+      "epoch": 0.8765412329863891,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.2819565348011466e-06,
+      "loss": 2.2984,
+      "step": 2737
+    },
+    {
+      "epoch": 0.8768614891913531,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.2703001467832137e-06,
+      "loss": 2.3775,
+      "step": 2738
+    },
+    {
+      "epoch": 0.8771817453963171,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.258672189620431e-06,
+      "loss": 2.5423,
+      "step": 2739
+    },
+    {
+      "epoch": 0.877502001601281,
+      "grad_norm": 1.5,
+      "learning_rate": 2.2470726778573146e-06,
+      "loss": 1.8138,
+      "step": 2740
+    },
+    {
+      "epoch": 0.877822257806245,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.235501626002817e-06,
+      "loss": 1.9844,
+      "step": 2741
+    },
+    {
+      "epoch": 0.878142514011209,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.2239590485302887e-06,
+      "loss": 1.8902,
+      "step": 2742
+    },
+    {
+      "epoch": 0.8784627702161729,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.212444959877466e-06,
+      "loss": 1.7303,
+      "step": 2743
+    },
+    {
+      "epoch": 0.8787830264211369,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.2009593744464456e-06,
+      "loss": 2.1603,
+      "step": 2744
+    },
+    {
+      "epoch": 0.8791032826261009,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.1895023066036745e-06,
+      "loss": 1.5858,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8794235388310648,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.1780737706799255e-06,
+      "loss": 2.3251,
+      "step": 2746
+    },
+    {
+      "epoch": 0.8797437950360288,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.1666737809702963e-06,
+      "loss": 2.1973,
+      "step": 2747
+    },
+    {
+      "epoch": 0.8800640512409927,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.1553023517341566e-06,
+      "loss": 2.2133,
+      "step": 2748
+    },
+    {
+      "epoch": 0.8803843074459567,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.1439594971951666e-06,
+      "loss": 2.7235,
+      "step": 2749
+    },
+    {
+      "epoch": 0.8807045636509208,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.1326452315412413e-06,
+      "loss": 1.9469,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8810248198558847,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.1213595689245386e-06,
+      "loss": 2.0951,
+      "step": 2751
+    },
+    {
+      "epoch": 0.8813450760608487,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.110102523461435e-06,
+      "loss": 2.1507,
+      "step": 2752
+    },
+    {
+      "epoch": 0.8816653322658127,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.098874109232521e-06,
+      "loss": 2.0099,
+      "step": 2753
+    },
+    {
+      "epoch": 0.8819855884707766,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.08767434028255e-06,
+      "loss": 1.7282,
+      "step": 2754
+    },
+    {
+      "epoch": 0.8823058446757406,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.07650323062048e-06,
+      "loss": 2.3165,
+      "step": 2755
+    },
+    {
+      "epoch": 0.8826261008807046,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.065360794219395e-06,
+      "loss": 2.093,
+      "step": 2756
+    },
+    {
+      "epoch": 0.8829463570856685,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.054247045016522e-06,
+      "loss": 1.8592,
+      "step": 2757
+    },
+    {
+      "epoch": 0.8832666132906325,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.0431619969132065e-06,
+      "loss": 2.0496,
+      "step": 2758
+    },
+    {
+      "epoch": 0.8835868694955965,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.032105663774897e-06,
+      "loss": 2.394,
+      "step": 2759
+    },
+    {
+      "epoch": 0.8839071257005604,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.021078059431114e-06,
+      "loss": 1.8352,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8842273819055244,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.010079197675463e-06,
+      "loss": 1.8796,
+      "step": 2761
+    },
+    {
+      "epoch": 0.8845476381104884,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9991090922655694e-06,
+      "loss": 1.7734,
+      "step": 2762
+    },
+    {
+      "epoch": 0.8848678943154523,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.9881677569231094e-06,
+      "loss": 2.5961,
+      "step": 2763
+    },
+    {
+      "epoch": 0.8851881505204163,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.977255205333775e-06,
+      "loss": 1.9614,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8855084067253803,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.966371451147239e-06,
+      "loss": 2.529,
+      "step": 2765
+    },
+    {
+      "epoch": 0.8858286629303442,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9555165079771703e-06,
+      "loss": 2.4828,
+      "step": 2766
+    },
+    {
+      "epoch": 0.8861489191353082,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9446903894011937e-06,
+      "loss": 2.0968,
+      "step": 2767
+    },
+    {
+      "epoch": 0.8864691753402723,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.933893108960866e-06,
+      "loss": 1.7931,
+      "step": 2768
+    },
+    {
+      "epoch": 0.8867894315452362,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9231246801617033e-06,
+      "loss": 2.7387,
+      "step": 2769
+    },
+    {
+      "epoch": 0.8871096877502002,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9123851164731004e-06,
+      "loss": 2.1077,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8874299439551642,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.901674431328368e-06,
+      "loss": 2.2398,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8877502001601281,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.8909926381246862e-06,
+      "loss": 2.5055,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8880704563650921,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8803397502231034e-06,
+      "loss": 2.0305,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8883907125700561,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.8697157809485055e-06,
+      "loss": 2.3018,
+      "step": 2774
+    },
+    {
+      "epoch": 0.88871096877502,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8591207435896096e-06,
+      "loss": 2.3575,
+      "step": 2775
+    },
+    {
+      "epoch": 0.889031224979984,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.8485546513989372e-06,
+      "loss": 2.1141,
+      "step": 2776
+    },
+    {
+      "epoch": 0.889351481184948,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8380175175928216e-06,
+      "loss": 1.9234,
+      "step": 2777
+    },
+    {
+      "epoch": 0.8896717373899119,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.8275093553513533e-06,
+      "loss": 2.3495,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8899919935948759,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.817030177818399e-06,
+      "loss": 2.2586,
+      "step": 2779
+    },
+    {
+      "epoch": 0.8903122497998399,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.8065799981015625e-06,
+      "loss": 2.1218,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8906325060048038,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.7961588292721848e-06,
+      "loss": 2.3084,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8909527622097678,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7857666843653088e-06,
+      "loss": 1.9461,
+      "step": 2782
+    },
+    {
+      "epoch": 0.8912730184147318,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.775403576379689e-06,
+      "loss": 2.3538,
+      "step": 2783
+    },
+    {
+      "epoch": 0.8915932746196957,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.76506951827774e-06,
+      "loss": 2.2233,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8919135308246597,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.7547645229855497e-06,
+      "loss": 2.5818,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8922337870296237,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.7444886033928603e-06,
+      "loss": 2.1235,
+      "step": 2786
+    },
+    {
+      "epoch": 0.8925540432345876,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.7342417723530374e-06,
+      "loss": 2.3278,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8928742994395517,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.7240240426830617e-06,
+      "loss": 2.1486,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8931945556445157,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.713835427163521e-06,
+      "loss": 2.7315,
+      "step": 2789
+    },
+    {
+      "epoch": 0.8935148118494796,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7036759385385682e-06,
+      "loss": 1.8048,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8938350680544436,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.693545589515952e-06,
+      "loss": 2.6576,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8941553242594076,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.6834443927669446e-06,
+      "loss": 1.56,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8944755804643715,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.6733723609263674e-06,
+      "loss": 2.1219,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8947958366693355,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.6633295065925647e-06,
+      "loss": 1.3158,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8951160928742994,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.653315842327377e-06,
+      "loss": 2.4756,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8954363490792634,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6433313806561379e-06,
+      "loss": 2.2544,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8957566052842274,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6333761340676518e-06,
+      "loss": 2.0936,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8960768614891913,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.6234501150141745e-06,
+      "loss": 2.3256,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8963971176941553,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6135533359114247e-06,
+      "loss": 2.2462,
+      "step": 2799
+    },
+    {
+      "epoch": 0.8967173738991193,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.6036858091385137e-06,
+      "loss": 2.377,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8970376301040832,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.593847547037991e-06,
+      "loss": 2.2872,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8973578863090472,
+      "grad_norm": 0.5,
+      "learning_rate": 1.5840385619157904e-06,
+      "loss": 1.8975,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8976781425140112,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.574258866041231e-06,
+      "loss": 2.7625,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8979983987189751,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.5645084716469777e-06,
+      "loss": 2.2334,
+      "step": 2804
+    },
+    {
+      "epoch": 0.8983186549239391,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.5547873909290716e-06,
+      "loss": 2.3039,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8986389111289032,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.5450956360468698e-06,
+      "loss": 1.9124,
+      "step": 2806
+    },
+    {
+      "epoch": 0.898959167333867,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.5354332191230502e-06,
+      "loss": 1.9301,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8992794235388311,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.5258001522435978e-06,
+      "loss": 2.5829,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8995996797437951,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.516196447457785e-06,
+      "loss": 2.4198,
+      "step": 2809
+    },
+    {
+      "epoch": 0.899919935948759,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5066221167781557e-06,
+      "loss": 2.4602,
+      "step": 2810
+    },
+    {
+      "epoch": 0.900240192153723,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.4970771721805165e-06,
+      "loss": 2.4408,
+      "step": 2811
+    },
+    {
+      "epoch": 0.900560448358687,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.4875616256039031e-06,
+      "loss": 2.3519,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9008807045636509,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.478075488950606e-06,
+      "loss": 1.9272,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9012009607686149,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.468618774086103e-06,
+      "loss": 2.1546,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9015212169735789,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.4591914928390793e-06,
+      "loss": 2.2757,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9018414731785428,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.449793657001408e-06,
+      "loss": 2.3683,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9021617293835068,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.4404252783281308e-06,
+      "loss": 1.7778,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9024819855884708,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.431086368537432e-06,
+      "loss": 2.2228,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9028022417934347,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.4217769393106567e-06,
+      "loss": 2.0747,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9031224979983987,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.4124970022922513e-06,
+      "loss": 1.7582,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9034427542033627,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4032465690897944e-06,
+      "loss": 2.561,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9037630104083266,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3940256512739447e-06,
+      "loss": 1.9619,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9040832666132906,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.3848342603784453e-06,
+      "loss": 2.1804,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9044035228182546,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3756724079001165e-06,
+      "loss": 2.6895,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9047237790232185,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.366540105298819e-06,
+      "loss": 1.9138,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9050440352281826,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3574373639974518e-06,
+      "loss": 2.2131,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9053642914331466,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3483641953819542e-06,
+      "loss": 2.1942,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9056845476381105,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3393206108012562e-06,
+      "loss": 1.9268,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9060048038430745,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.33030662156729e-06,
+      "loss": 2.5872,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9063250600480385,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3213222389549724e-06,
+      "loss": 2.0611,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9066453162530024,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3123674742021862e-06,
+      "loss": 2.4414,
+      "step": 2831
+    },
+    {
+      "epoch": 0.9069655724579664,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.303442338509761e-06,
+      "loss": 2.0825,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9072858286629304,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2945468430414803e-06,
+      "loss": 1.8121,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9076060848678943,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2856809989240326e-06,
+      "loss": 2.3212,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9079263410728583,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.2768448172470444e-06,
+      "loss": 2.4806,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9082465972778223,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2680383090630077e-06,
+      "loss": 2.4582,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9085668534827862,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2592614853873224e-06,
+      "loss": 2.4116,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9088871096877502,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.2505143571982515e-06,
+      "loss": 2.1586,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9092073658927142,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2417969354369147e-06,
+      "loss": 2.3296,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9095276220976781,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.2331092310072623e-06,
+      "loss": 1.7858,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9098478783026421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.2244512547761016e-06,
+      "loss": 2.1908,
+      "step": 2841
+    },
+    {
+      "epoch": 0.910168134507606,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.2158230175730201e-06,
+      "loss": 2.1441,
+      "step": 2842
+    },
+    {
+      "epoch": 0.91048839071257,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2072245301904345e-06,
+      "loss": 2.2306,
+      "step": 2843
+    },
+    {
+      "epoch": 0.910808646917534,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.1986558033835333e-06,
+      "loss": 2.5195,
+      "step": 2844
+    },
+    {
+      "epoch": 0.911128903122498,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.190116847870293e-06,
+      "loss": 2.6438,
+      "step": 2845
+    },
+    {
+      "epoch": 0.911449159327462,
+      "grad_norm": 1.0,
+      "learning_rate": 1.181607674331439e-06,
+      "loss": 2.0309,
+      "step": 2846
+    },
+    {
+      "epoch": 0.911769415532426,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1731282934104553e-06,
+      "loss": 2.1805,
+      "step": 2847
+    },
+    {
+      "epoch": 0.9120896717373899,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.1646787157135463e-06,
+      "loss": 2.2949,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9124099279423539,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.1562589518096607e-06,
+      "loss": 2.3779,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9127301841473179,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.1478690122304293e-06,
+      "loss": 1.8918,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9130504403522818,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.1395089074701936e-06,
+      "loss": 2.4792,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9133706965572458,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.1311786479859781e-06,
+      "loss": 2.6007,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9136909527622098,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.12287824419747e-06,
+      "loss": 2.0799,
+      "step": 2853
+    },
+    {
+      "epoch": 0.9140112089671737,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.114607706487006e-06,
+      "loss": 2.4274,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9143314651721377,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1063670451995856e-06,
+      "loss": 2.5174,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9146517213771017,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0981562706428166e-06,
+      "loss": 2.1216,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9149719775820656,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0899753930869394e-06,
+      "loss": 2.0923,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9152922337870296,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.0818244227647905e-06,
+      "loss": 2.3559,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9156124899919936,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.0737033698717979e-06,
+      "loss": 2.3995,
+      "step": 2859
+    },
+    {
+      "epoch": 0.9159327461969575,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.0656122445659721e-06,
+      "loss": 2.1872,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9162530024019215,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.057551056967887e-06,
+      "loss": 1.9036,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9165732586068855,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.0495198171606657e-06,
+      "loss": 2.4604,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9168935148118494,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0415185351899837e-06,
+      "loss": 2.326,
+      "step": 2863
+    },
+    {
+      "epoch": 0.9172137710168135,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.0335472210640323e-06,
+      "loss": 2.0918,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9175340272217775,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0256058847535221e-06,
+      "loss": 2.2654,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9178542834267414,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.0176945361916657e-06,
+      "loss": 2.1514,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9181745396317054,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.0098131852741777e-06,
+      "loss": 2.1784,
+      "step": 2867
+    },
+    {
+      "epoch": 0.9184947958366694,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.001961841859228e-06,
+      "loss": 2.1977,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9188150520416333,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.941405157674778e-07,
+      "loss": 2.4212,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9191353082465973,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.863492167820232e-07,
+      "loss": 1.8568,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9194555644515613,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.785879546484077e-07,
+      "loss": 2.2068,
+      "step": 2871
+    },
+    {
+      "epoch": 0.9197758206565252,
+      "grad_norm": 0.48046875,
+      "learning_rate": 9.708567390746077e-07,
+      "loss": 2.2889,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9200960768614892,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.631555797310122e-07,
+      "loss": 2.3926,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9204163330664532,
+      "grad_norm": 0.490234375,
+      "learning_rate": 9.554844862504158e-07,
+      "loss": 2.2207,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9207365892714171,
+      "grad_norm": 0.478515625,
+      "learning_rate": 9.47843468228013e-07,
+      "loss": 2.3968,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9210568454763811,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.402325352213582e-07,
+      "loss": 2.2347,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9213771016813451,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.326516967504056e-07,
+      "loss": 2.4828,
+      "step": 2877
+    },
+    {
+      "epoch": 0.921697357886309,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.251009622974426e-07,
+      "loss": 2.2806,
+      "step": 2878
+    },
+    {
+      "epoch": 0.922017614091273,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.175803413071088e-07,
+      "loss": 2.2707,
+      "step": 2879
+    },
+    {
+      "epoch": 0.922337870296237,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.100898431863791e-07,
+      "loss": 1.9029,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9226581265012009,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.026294773045535e-07,
+      "loss": 1.0364,
+      "step": 2881
+    },
+    {
+      "epoch": 0.922978382706165,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.95199252993223e-07,
+      "loss": 2.0725,
+      "step": 2882
+    },
+    {
+      "epoch": 0.923298638911129,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.877991795463086e-07,
+      "loss": 2.2058,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9236188951160929,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.804292662199842e-07,
+      "loss": 2.1161,
+      "step": 2884
+    },
+    {
+      "epoch": 0.9239391513210569,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.730895222327284e-07,
+      "loss": 2.715,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9242594075260209,
+      "grad_norm": 0.29296875,
+      "learning_rate": 8.657799567652614e-07,
+      "loss": 1.7732,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9245796637309848,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.585005789605638e-07,
+      "loss": 2.3266,
+      "step": 2887
+    },
+    {
+      "epoch": 0.9248999199359488,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.512513979238579e-07,
+      "loss": 2.277,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9252201761409127,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.440324227225959e-07,
+      "loss": 2.2216,
+      "step": 2889
+    },
+    {
+      "epoch": 0.9255404323458767,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.368436623864384e-07,
+      "loss": 2.1062,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9258606885508407,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.296851259072702e-07,
+      "loss": 2.6251,
+      "step": 2891
+    },
+    {
+      "epoch": 0.9261809447558046,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.225568222391539e-07,
+      "loss": 3.0058,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9265012009607686,
+      "grad_norm": 0.46875,
+      "learning_rate": 8.154587602983432e-07,
+      "loss": 2.0105,
+      "step": 2893
+    },
+    {
+      "epoch": 0.9268214571657326,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.08390948963264e-07,
+      "loss": 2.0824,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9271417133706965,
+      "grad_norm": 0.4921875,
+      "learning_rate": 8.013533970745113e-07,
+      "loss": 2.2465,
+      "step": 2895
+    },
+    {
+      "epoch": 0.9274619695756605,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.943461134348129e-07,
+      "loss": 2.2219,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9277822257806245,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.873691068090605e-07,
+      "loss": 2.332,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9281024819855884,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.804223859242482e-07,
+      "loss": 2.7178,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9284227381905524,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.735059594695171e-07,
+      "loss": 1.4854,
+      "step": 2899
+    },
+    {
+      "epoch": 0.9287429943955164,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.666198360960858e-07,
+      "loss": 2.2354,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9290632506004803,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.597640244172921e-07,
+      "loss": 1.8512,
+      "step": 2901
+    },
+    {
+      "epoch": 0.9293835068054443,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.529385330085431e-07,
+      "loss": 1.8892,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9297037630104084,
+      "grad_norm": 0.38671875,
+      "learning_rate": 7.461433704073373e-07,
+      "loss": 2.5126,
+      "step": 2903
+    },
+    {
+      "epoch": 0.9300240192153723,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.393785451132146e-07,
+      "loss": 2.1277,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9303442754203363,
+      "grad_norm": 0.453125,
+      "learning_rate": 7.326440655877925e-07,
+      "loss": 2.4802,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9306645316253003,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.259399402547134e-07,
+      "loss": 2.1335,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9309847878302642,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.192661774996584e-07,
+      "loss": 2.5345,
+      "step": 2907
+    },
+    {
+      "epoch": 0.9313050440352282,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.126227856703332e-07,
+      "loss": 2.3705,
+      "step": 2908
+    },
+    {
+      "epoch": 0.9316253002401922,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.060097730764548e-07,
+      "loss": 2.6065,
+      "step": 2909
+    },
+    {
+      "epoch": 0.9319455564451561,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.994271479897314e-07,
+      "loss": 2.1943,
+      "step": 2910
+    },
+    {
+      "epoch": 0.9322658126501201,
+      "grad_norm": 0.47265625,
+      "learning_rate": 6.928749186438738e-07,
+      "loss": 1.9554,
+      "step": 2911
+    },
+    {
+      "epoch": 0.9325860688550841,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.863530932345625e-07,
+      "loss": 2.3002,
+      "step": 2912
+    },
+    {
+      "epoch": 0.932906325060048,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.798616799194634e-07,
+      "loss": 2.4129,
+      "step": 2913
+    },
+    {
+      "epoch": 0.933226581265012,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.734006868181847e-07,
+      "loss": 2.1435,
+      "step": 2914
+    },
+    {
+      "epoch": 0.933546837469976,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.66970122012292e-07,
+      "loss": 1.9209,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9338670936749399,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.605699935452902e-07,
+      "loss": 2.4338,
+      "step": 2916
+    },
+    {
+      "epoch": 0.9341873498799039,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.542003094226201e-07,
+      "loss": 2.2675,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9345076060848679,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.478610776116278e-07,
+      "loss": 1.4878,
+      "step": 2918
+    },
+    {
+      "epoch": 0.9348278622898318,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.415523060415846e-07,
+      "loss": 2.2072,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9351481184947958,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.352740026036475e-07,
+      "loss": 2.3663,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9354683746997599,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.290261751508708e-07,
+      "loss": 2.3013,
+      "step": 2921
+    },
+    {
+      "epoch": 0.9357886309047238,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.228088314981867e-07,
+      "loss": 2.0383,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9361088871096878,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.16621979422402e-07,
+      "loss": 1.8135,
+      "step": 2923
+    },
+    {
+      "epoch": 0.9364291433146518,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.104656266621766e-07,
+      "loss": 2.2271,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9367493995196157,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.043397809180257e-07,
+      "loss": 2.2725,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9370696557245797,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.982444498523005e-07,
+      "loss": 2.1763,
+      "step": 2926
+    },
+    {
+      "epoch": 0.9373899119295437,
+      "grad_norm": 0.5,
+      "learning_rate": 5.921796410891994e-07,
+      "loss": 2.3399,
+      "step": 2927
+    },
+    {
+      "epoch": 0.9377101681345076,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.861453622147157e-07,
+      "loss": 1.8565,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9380304243394716,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.801416207766808e-07,
+      "loss": 2.3172,
+      "step": 2929
+    },
+    {
+      "epoch": 0.9383506805444356,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.741684242847156e-07,
+      "loss": 1.864,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9386709367493995,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.682257802102408e-07,
+      "loss": 2.4591,
+      "step": 2931
+    },
+    {
+      "epoch": 0.9389911929543635,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.623136959864494e-07,
+      "loss": 2.759,
+      "step": 2932
+    },
+    {
+      "epoch": 0.9393114491593275,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.564321790083316e-07,
+      "loss": 1.5265,
+      "step": 2933
+    },
+    {
+      "epoch": 0.9396317053642914,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.505812366326219e-07,
+      "loss": 2.2601,
+      "step": 2934
+    },
+    {
+      "epoch": 0.9399519615692554,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.447608761778189e-07,
+      "loss": 1.9696,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9402722177742193,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.389711049241741e-07,
+      "loss": 2.2893,
+      "step": 2936
+    },
+    {
+      "epoch": 0.9405924739791833,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.332119301136695e-07,
+      "loss": 2.2355,
+      "step": 2937
+    },
+    {
+      "epoch": 0.9409127301841473,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.274833589500205e-07,
+      "loss": 2.0032,
+      "step": 2938
+    },
+    {
+      "epoch": 0.9412329863891112,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.217853985986648e-07,
+      "loss": 2.6002,
+      "step": 2939
+    },
+    {
+      "epoch": 0.9415532425940752,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.161180561867401e-07,
+      "loss": 2.2874,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9418734987990393,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.104813388031038e-07,
+      "loss": 2.5509,
+      "step": 2941
+    },
+    {
+      "epoch": 0.9421937550040032,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.048752534982909e-07,
+      "loss": 2.1587,
+      "step": 2942
+    },
+    {
+      "epoch": 0.9425140112089672,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.992998072845311e-07,
+      "loss": 2.4545,
+      "step": 2943
+    },
+    {
+      "epoch": 0.9428342674139312,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.937550071357239e-07,
+      "loss": 1.585,
+      "step": 2944
+    },
+    {
+      "epoch": 0.9431545236188951,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.882408599874433e-07,
+      "loss": 2.0142,
+      "step": 2945
+    },
+    {
+      "epoch": 0.9434747798238591,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.827573727369112e-07,
+      "loss": 2.179,
+      "step": 2946
+    },
+    {
+      "epoch": 0.9437950360288231,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.773045522430103e-07,
+      "loss": 2.4145,
+      "step": 2947
+    },
+    {
+      "epoch": 0.944115292233787,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.7188240532625714e-07,
+      "loss": 2.4136,
+      "step": 2948
+    },
+    {
+      "epoch": 0.944435548438751,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.664909387688043e-07,
+      "loss": 1.7031,
+      "step": 2949
+    },
+    {
+      "epoch": 0.944755804643715,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.6113015931442683e-07,
+      "loss": 2.3827,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9450760608486789,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.558000736685164e-07,
+      "loss": 2.4576,
+      "step": 2951
+    },
+    {
+      "epoch": 0.9453963170536429,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.5050068849807334e-07,
+      "loss": 2.6722,
+      "step": 2952
+    },
+    {
+      "epoch": 0.9457165732586069,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.4523201043169813e-07,
+      "loss": 2.5232,
+      "step": 2953
+    },
+    {
+      "epoch": 0.9460368294635708,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.3999404605957196e-07,
+      "loss": 2.327,
+      "step": 2954
+    },
+    {
+      "epoch": 0.9463570856685348,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.347868019334789e-07,
+      "loss": 2.152,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9466773418734988,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.296102845667532e-07,
+      "loss": 2.3246,
+      "step": 2956
+    },
+    {
+      "epoch": 0.9469975980784627,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.244645004343156e-07,
+      "loss": 2.4348,
+      "step": 2957
+    },
+    {
+      "epoch": 0.9473178542834267,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.193494559726313e-07,
+      "loss": 2.3624,
+      "step": 2958
+    },
+    {
+      "epoch": 0.9476381104883908,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.1426515757972684e-07,
+      "loss": 2.3969,
+      "step": 2959
+    },
+    {
+      "epoch": 0.9479583666933546,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.0921161161515674e-07,
+      "loss": 2.1616,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9482786228983187,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.041888244000286e-07,
+      "loss": 1.9173,
+      "step": 2961
+    },
+    {
+      "epoch": 0.9485988791032827,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.991968022169529e-07,
+      "loss": 2.3724,
+      "step": 2962
+    },
+    {
+      "epoch": 0.9489191353082466,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.9423555131007925e-07,
+      "loss": 2.1529,
+      "step": 2963
+    },
+    {
+      "epoch": 0.9492393915132106,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.8930507788506046e-07,
+      "loss": 2.1868,
+      "step": 2964
+    },
+    {
+      "epoch": 0.9495596477181746,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.8440538810904367e-07,
+      "loss": 1.8231,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9498799039231385,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.7953648811068777e-07,
+      "loss": 2.235,
+      "step": 2966
+    },
+    {
+      "epoch": 0.9502001601281025,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.746983839801238e-07,
+      "loss": 2.3122,
+      "step": 2967
+    },
+    {
+      "epoch": 0.9505204163330665,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.698910817689694e-07,
+      "loss": 2.4373,
+      "step": 2968
+    },
+    {
+      "epoch": 0.9508406725380304,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.6511458749031723e-07,
+      "loss": 1.7601,
+      "step": 2969
+    },
+    {
+      "epoch": 0.9511609287429944,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.6036890711871886e-07,
+      "loss": 2.7486,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9514811849479584,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.556540465901842e-07,
+      "loss": 2.1883,
+      "step": 2971
+    },
+    {
+      "epoch": 0.9518014411529223,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.509700118021736e-07,
+      "loss": 2.4816,
+      "step": 2972
+    },
+    {
+      "epoch": 0.9521216973578863,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.463168086135976e-07,
+      "loss": 2.3101,
+      "step": 2973
+    },
+    {
+      "epoch": 0.9524419535628503,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.416944428447838e-07,
+      "loss": 2.7864,
+      "step": 2974
+    },
+    {
+      "epoch": 0.9527622097678142,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.3710292027750724e-07,
+      "loss": 2.1871,
+      "step": 2975
+    },
+    {
+      "epoch": 0.9530824659727782,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.3254224665494883e-07,
+      "loss": 2.508,
+      "step": 2976
+    },
+    {
+      "epoch": 0.9534027221777422,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.2801242768171483e-07,
+      "loss": 1.6672,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9537229783827061,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.235134690238062e-07,
+      "loss": 2.5607,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9540432345876702,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.1904537630862984e-07,
+      "loss": 2.4434,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9543634907926342,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.146081551249874e-07,
+      "loss": 2.1326,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9546837469975981,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.102018110230587e-07,
+      "loss": 2.3844,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9550040032025621,
+      "grad_norm": 0.3828125,
+      "learning_rate": 3.058263495144015e-07,
+      "loss": 2.5421,
+      "step": 2982
+    },
+    {
+      "epoch": 0.955324259407526,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.014817760719546e-07,
+      "loss": 1.6864,
+      "step": 2983
+    },
+    {
+      "epoch": 0.95564451561249,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.971680961300127e-07,
+      "loss": 2.5311,
+      "step": 2984
+    },
+    {
+      "epoch": 0.955964771817454,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.9288531508422644e-07,
+      "loss": 2.2674,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9562850280224179,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.886334382916078e-07,
+      "loss": 2.4979,
+      "step": 2986
+    },
+    {
+      "epoch": 0.9566052842273819,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.8441247107050264e-07,
+      "loss": 2.3117,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9569255404323459,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.802224187005986e-07,
+      "loss": 2.069,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9572457966373098,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.7606328642291736e-07,
+      "loss": 2.3921,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9575660528422738,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7193507943979457e-07,
+      "loss": 2.7724,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9578863090472378,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.678378029148998e-07,
+      "loss": 2.1656,
+      "step": 2991
+    },
+    {
+      "epoch": 0.9582065652522017,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.6377146197319733e-07,
+      "loss": 2.1326,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9585268214571657,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.5973606170096585e-07,
+      "loss": 2.2594,
+      "step": 2993
+    },
+    {
+      "epoch": 0.9588470776621297,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.5573160714578436e-07,
+      "loss": 2.2226,
+      "step": 2994
+    },
+    {
+      "epoch": 0.9591673338670936,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.517581033165184e-07,
+      "loss": 2.1275,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9594875900720576,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.4781555518332e-07,
+      "loss": 2.3456,
+      "step": 2996
+    },
+    {
+      "epoch": 0.9598078462770216,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.439039676776306e-07,
+      "loss": 2.5249,
+      "step": 2997
+    },
+    {
+      "epoch": 0.9601281024819855,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.4002334569215023e-07,
+      "loss": 1.8815,
+      "step": 2998
+    },
+    {
+      "epoch": 0.9604483586869496,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.3617369408085732e-07,
+      "loss": 2.5048,
+      "step": 2999
+    },
+    {
+      "epoch": 0.9607686148919136,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.3235501765898615e-07,
+      "loss": 2.205,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9610888710968775,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.285673212030326e-07,
+      "loss": 2.0211,
+      "step": 3001
+    },
+    {
+      "epoch": 0.9614091273018415,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.2481060945073738e-07,
+      "loss": 2.1957,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9617293835068055,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.21084887101089e-07,
+      "loss": 2.2572,
+      "step": 3003
+    },
+    {
+      "epoch": 0.9620496397117694,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.1739015881430125e-07,
+      "loss": 2.4002,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9623698959167334,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.137264292118385e-07,
+      "loss": 1.8485,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9626901521216974,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.1009370287637664e-07,
+      "loss": 2.4418,
+      "step": 3006
+    },
+    {
+      "epoch": 0.9630104083266613,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.0649198435181983e-07,
+      "loss": 2.4171,
+      "step": 3007
+    },
+    {
+      "epoch": 0.9633306645316253,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.02921278143281e-07,
+      "loss": 2.5708,
+      "step": 3008
+    },
+    {
+      "epoch": 0.9636509207365893,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.99381588717093e-07,
+      "loss": 2.2956,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9639711769415532,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9587292050077255e-07,
+      "loss": 2.6983,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9642914331465172,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9239527788305622e-07,
+      "loss": 2.1624,
+      "step": 3011
+    },
+    {
+      "epoch": 0.9646116893514812,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.889486652138589e-07,
+      "loss": 1.8433,
+      "step": 3012
+    },
+    {
+      "epoch": 0.9649319455564451,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.855330868042876e-07,
+      "loss": 2.5753,
+      "step": 3013
+    },
+    {
+      "epoch": 0.9652522017614091,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.8214854692662765e-07,
+      "loss": 1.9499,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9655724579663731,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7879504981434536e-07,
+      "loss": 1.9731,
+      "step": 3015
+    },
+    {
+      "epoch": 0.965892714171337,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.7547259966207708e-07,
+      "loss": 2.0553,
+      "step": 3016
+    },
+    {
+      "epoch": 0.966212970376301,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.7218120062562071e-07,
+      "loss": 2.3178,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9665332265812651,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.6892085682193583e-07,
+      "loss": 1.948,
+      "step": 3018
+    },
+    {
+      "epoch": 0.966853482786229,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6569157232914079e-07,
+      "loss": 2.2639,
+      "step": 3019
+    },
+    {
+      "epoch": 0.967173738991193,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6249335118650456e-07,
+      "loss": 2.1195,
+      "step": 3020
+    },
+    {
+      "epoch": 0.967493995196157,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5932619739443543e-07,
+      "loss": 2.0017,
+      "step": 3021
+    },
+    {
+      "epoch": 0.9678142514011209,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.5619011491448677e-07,
+      "loss": 1.485,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9681345076060849,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.5308510766934848e-07,
+      "loss": 2.2819,
+      "step": 3023
+    },
+    {
+      "epoch": 0.9684547638110489,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.5001117954283329e-07,
+      "loss": 1.7135,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9687750200160128,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.4696833437988778e-07,
+      "loss": 2.3383,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9690952762209768,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.4395657598657296e-07,
+      "loss": 1.7648,
+      "step": 3026
+    },
+    {
+      "epoch": 0.9694155324259408,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.4097590813007545e-07,
+      "loss": 2.0177,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9697357886309047,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.380263345386795e-07,
+      "loss": 1.8389,
+      "step": 3028
+    },
+    {
+      "epoch": 0.9700560448358687,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.35107858901784e-07,
+      "loss": 2.0451,
+      "step": 3029
+    },
+    {
+      "epoch": 0.9703763010408326,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3222048486989102e-07,
+      "loss": 2.4008,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9706965572457966,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.2936421605459494e-07,
+      "loss": 2.2855,
+      "step": 3031
+    },
+    {
+      "epoch": 0.9710168134507606,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2653905602858507e-07,
+      "loss": 2.2895,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9713370696557245,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2374500832564306e-07,
+      "loss": 2.4505,
+      "step": 3033
+    },
+    {
+      "epoch": 0.9716573258606885,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.209820764406261e-07,
+      "loss": 2.6746,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9719775820656525,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.1825026382947801e-07,
+      "loss": 2.5954,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9722978382706164,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.1554957390921828e-07,
+      "loss": 2.7764,
+      "step": 3036
+    },
+    {
+      "epoch": 0.9726180944755805,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1288001005793358e-07,
+      "loss": 2.1283,
+      "step": 3037
+    },
+    {
+      "epoch": 0.9729383506805445,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.1024157561477233e-07,
+      "loss": 2.1614,
+      "step": 3038
+    },
+    {
+      "epoch": 0.9732586068855084,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0763427387996128e-07,
+      "loss": 2.4447,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9735788630904724,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0505810811476947e-07,
+      "loss": 2.3156,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9738991192954364,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0251308154153317e-07,
+      "loss": 2.5195,
+      "step": 3041
+    },
+    {
+      "epoch": 0.9742193755004003,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.999919734362539e-08,
+      "loss": 2.5644,
+      "step": 3042
+    },
+    {
+      "epoch": 0.9745396317053643,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.75164586654781e-08,
+      "loss": 2.3122,
+      "step": 3043
+    },
+    {
+      "epoch": 0.9748598879103283,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.506486861255715e-08,
+      "loss": 2.1659,
+      "step": 3044
+    },
+    {
+      "epoch": 0.9751801441152922,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.264443025137626e-08,
+      "loss": 2.1065,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9755004003202562,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.025514660946921e-08,
+      "loss": 2.4461,
+      "step": 3046
+    },
+    {
+      "epoch": 0.9758206565252202,
+      "grad_norm": 0.35546875,
+      "learning_rate": 8.78970206754176e-08,
+      "loss": 2.4571,
+      "step": 3047
+    },
+    {
+      "epoch": 0.9761409127301841,
+      "grad_norm": 0.4765625,
+      "learning_rate": 8.55700553988148e-08,
+      "loss": 1.889,
+      "step": 3048
+    },
+    {
+      "epoch": 0.9764611689351481,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.327425369029085e-08,
+      "loss": 2.25,
+      "step": 3049
+    },
+    {
+      "epoch": 0.9767814251401121,
+      "grad_norm": 0.380859375,
+      "learning_rate": 8.100961842148758e-08,
+      "loss": 2.4113,
+      "step": 3050
+    },
+    {
+      "epoch": 0.977101681345076,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.877615242506408e-08,
+      "loss": 2.4718,
+      "step": 3051
+    },
+    {
+      "epoch": 0.97742193755004,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.657385849469678e-08,
+      "loss": 2.5222,
+      "step": 3052
+    },
+    {
+      "epoch": 0.977742193755004,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.44027393850627e-08,
+      "loss": 2.1825,
+      "step": 3053
+    },
+    {
+      "epoch": 0.9780624499599679,
+      "grad_norm": 0.484375,
+      "learning_rate": 7.226279781185341e-08,
+      "loss": 1.9089,
+      "step": 3054
+    },
+    {
+      "epoch": 0.978382706164932,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.015403645176111e-08,
+      "loss": 2.3184,
+      "step": 3055
+    },
+    {
+      "epoch": 0.978702962369896,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.80764579424703e-08,
+      "loss": 1.9132,
+      "step": 3056
+    },
+    {
+      "epoch": 0.9790232185748599,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.603006488266616e-08,
+      "loss": 1.9538,
+      "step": 3057
+    },
+    {
+      "epoch": 0.9793434747798239,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.40148598320317e-08,
+      "loss": 2.1982,
+      "step": 3058
+    },
+    {
+      "epoch": 0.9796637309847879,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.203084531123115e-08,
+      "loss": 1.9992,
+      "step": 3059
+    },
+    {
+      "epoch": 0.9799839871897518,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.007802380191552e-08,
+      "loss": 2.0144,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9803042433947158,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.815639774672532e-08,
+      "loss": 1.9409,
+      "step": 3061
+    },
+    {
+      "epoch": 0.9806244995996798,
+      "grad_norm": 0.365234375,
+      "learning_rate": 5.6265969549271216e-08,
+      "loss": 2.4769,
+      "step": 3062
+    },
+    {
+      "epoch": 0.9809447558046437,
+      "grad_norm": 0.37890625,
+      "learning_rate": 5.44067415741506e-08,
+      "loss": 2.4554,
+      "step": 3063
+    },
+    {
+      "epoch": 0.9812650120096077,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.257871614692822e-08,
+      "loss": 2.1183,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9815852682145717,
+      "grad_norm": 0.75,
+      "learning_rate": 5.07818955541417e-08,
+      "loss": 2.0154,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9819055244195356,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.901628204330155e-08,
+      "loss": 2.1623,
+      "step": 3066
+    },
+    {
+      "epoch": 0.9822257806244996,
+      "grad_norm": 0.625,
+      "learning_rate": 4.728187782287452e-08,
+      "loss": 1.8426,
+      "step": 3067
+    },
+    {
+      "epoch": 0.9825460368294636,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.5578685062297456e-08,
+      "loss": 2.0737,
+      "step": 3068
+    },
+    {
+      "epoch": 0.9828662930344275,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.390670589196622e-08,
+      "loss": 2.1993,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9831865492393915,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.2265942403227345e-08,
+      "loss": 2.5824,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9835068054443555,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.065639664839471e-08,
+      "loss": 2.3295,
+      "step": 3071
+    },
+    {
+      "epoch": 0.9838270616493194,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.907807064072178e-08,
+      "loss": 1.9268,
+      "step": 3072
+    },
+    {
+      "epoch": 0.9841473178542834,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.7530966354418215e-08,
+      "loss": 2.4965,
+      "step": 3073
+    },
+    {
+      "epoch": 0.9844675740592475,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.6015085724638856e-08,
+      "loss": 2.1137,
+      "step": 3074
+    },
+    {
+      "epoch": 0.9847878302642114,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.453043064748362e-08,
+      "loss": 1.6508,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9851080864691754,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.307700297999483e-08,
+      "loss": 2.1851,
+      "step": 3076
+    },
+    {
+      "epoch": 0.9854283426741393,
+      "grad_norm": 0.314453125,
+      "learning_rate": 3.1654804540157124e-08,
+      "loss": 1.754,
+      "step": 3077
+    },
+    {
+      "epoch": 0.9857485988791033,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.02638371068864e-08,
+      "loss": 2.1616,
+      "step": 3078
+    },
+    {
+      "epoch": 0.9860688550840673,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.890410242003816e-08,
+      "loss": 2.733,
+      "step": 3079
+    },
+    {
+      "epoch": 0.9863891112890312,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.757560218040467e-08,
+      "loss": 2.0874,
+      "step": 3080
+    },
+    {
+      "epoch": 0.9867093674939952,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.6278338049706696e-08,
+      "loss": 2.343,
+      "step": 3081
+    },
+    {
+      "epoch": 0.9870296236989592,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.5012311650587927e-08,
+      "loss": 2.3688,
+      "step": 3082
+    },
+    {
+      "epoch": 0.9873498799039231,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.377752456662885e-08,
+      "loss": 1.8563,
+      "step": 3083
+    },
+    {
+      "epoch": 0.9876701361088871,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.257397834233288e-08,
+      "loss": 2.2285,
+      "step": 3084
+    },
+    {
+      "epoch": 0.9879903923138511,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.1401674483118027e-08,
+      "loss": 2.2832,
+      "step": 3085
+    },
+    {
+      "epoch": 0.988310648518815,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.026061445533356e-08,
+      "loss": 2.1044,
+      "step": 3086
+    },
+    {
+      "epoch": 0.988630904723779,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9150799686246112e-08,
+      "loss": 2.234,
+      "step": 3087
+    },
+    {
+      "epoch": 0.988951160928743,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.8072231564036922e-08,
+      "loss": 1.7073,
+      "step": 3088
+    },
+    {
+      "epoch": 0.9892714171337069,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.702491143780738e-08,
+      "loss": 2.6741,
+      "step": 3089
+    },
+    {
+      "epoch": 0.9895916733386709,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.6008840617565134e-08,
+      "loss": 2.2903,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9899119295436349,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.5024020374243554e-08,
+      "loss": 2.1201,
+      "step": 3091
+    },
+    {
+      "epoch": 0.9902321857485988,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4070451939673934e-08,
+      "loss": 2.3257,
+      "step": 3092
+    },
+    {
+      "epoch": 0.9905524419535628,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3148136506604958e-08,
+      "loss": 2.3063,
+      "step": 3093
+    },
+    {
+      "epoch": 0.9908726981585269,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2257075228688797e-08,
+      "loss": 2.2024,
+      "step": 3094
+    },
+    {
+      "epoch": 0.9911929543634908,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.1397269220486672e-08,
+      "loss": 2.2727,
+      "step": 3095
+    },
+    {
+      "epoch": 0.9915132105684548,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.0568719557468853e-08,
+      "loss": 1.8462,
+      "step": 3096
+    },
+    {
+      "epoch": 0.9918334667734188,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.771427276000778e-09,
+      "loss": 1.2467,
+      "step": 3097
+    },
+    {
+      "epoch": 0.9921537229783827,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.00539337335693e-09,
+      "loss": 1.7548,
+      "step": 3098
+    },
+    {
+      "epoch": 0.9924739791833467,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.270618807706965e-09,
+      "loss": 1.8665,
+      "step": 3099
+    },
+    {
+      "epoch": 0.9927942353883107,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.56710449813236e-09,
+      "loss": 1.8483,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9931144915932746,
+      "grad_norm": 0.75,
+      "learning_rate": 6.894851324595886e-09,
+      "loss": 2.3824,
+      "step": 3101
+    },
+    {
+      "epoch": 0.9934347477982386,
+      "grad_norm": 0.48828125,
+      "learning_rate": 6.2538601279776846e-09,
+      "loss": 2.3081,
+      "step": 3102
+    },
+    {
+      "epoch": 0.9937550040032026,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.644131710039191e-09,
+      "loss": 2.6313,
+      "step": 3103
+    },
+    {
+      "epoch": 0.9940752602081665,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.065666833442562e-09,
+      "loss": 2.2497,
+      "step": 3104
+    },
+    {
+      "epoch": 0.9943955164131305,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.518466221750672e-09,
+      "loss": 2.2095,
+      "step": 3105
+    },
+    {
+      "epoch": 0.9947157726180945,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.002530559410467e-09,
+      "loss": 2.0491,
+      "step": 3106
+    },
+    {
+      "epoch": 0.9950360288230584,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.5178604917668334e-09,
+      "loss": 1.9228,
+      "step": 3107
+    },
+    {
+      "epoch": 0.9953562850280224,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.0644566250598305e-09,
+      "loss": 2.6178,
+      "step": 3108
+    },
+    {
+      "epoch": 0.9956765412329864,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6423195264163593e-09,
+      "loss": 2.0714,
+      "step": 3109
+    },
+    {
+      "epoch": 0.9959967974379503,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.2514497238557144e-09,
+      "loss": 2.1848,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9963170536429143,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.891847706286809e-09,
+      "loss": 2.4682,
+      "step": 3111
+    },
+    {
+      "epoch": 0.9966373098478783,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.5635139235109507e-09,
+      "loss": 2.3354,
+      "step": 3112
+    },
+    {
+      "epoch": 0.9969575660528422,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2664487862107389e-09,
+      "loss": 1.8637,
+      "step": 3113
+    },
+    {
+      "epoch": 0.9972778222578063,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.0006526659667171e-09,
+      "loss": 2.0926,
+      "step": 3114
+    },
+    {
+      "epoch": 0.9975980784627703,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.661258952434969e-10,
+      "loss": 2.0889,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9979183346677342,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.628687673897571e-10,
+      "loss": 1.7024,
+      "step": 3116
+    },
+    {
+      "epoch": 0.9982385908726982,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.9088153664379457e-10,
+      "loss": 2.477,
+      "step": 3117
+    },
+    {
+      "epoch": 0.9985588470776622,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.5016441813630054e-10,
+      "loss": 2.363,
+      "step": 3118
+    },
+    {
+      "epoch": 0.9988791032826261,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.4071758787648214e-10,
+      "loss": 2.3924,
+      "step": 3119
+    },
+    {
+      "epoch": 0.9991993594875901,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.254118276316501e-11,
+      "loss": 2.2253,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9995196156925541,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5635300579242008e-11,
+      "loss": 2.4826,
+      "step": 3121
+    },
+    {
+      "epoch": 0.999839871897518,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0,
+      "loss": 2.4466,
+      "step": 3122
+    },
+    {
+      "epoch": 0.999839871897518,
+      "step": 3122,
+      "total_flos": 1.7832492586531553e+18,
+      "train_loss": 2.216206028322468,
+      "train_runtime": 33247.722,
+      "train_samples_per_second": 0.751,
+      "train_steps_per_second": 0.094
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3122,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.7832492586531553e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}