{
  "best_metric": 0.30803602933883667,
  "best_model_checkpoint": "./opt_trained/checkpoint-258",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 430,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.023391812865497075,
      "grad_norm": 1.4262398481369019,
      "learning_rate": 9.995238095238095e-05,
      "loss": 2.4722,
      "step": 1
    },
    {
      "epoch": 0.04678362573099415,
      "grad_norm": 12.206507682800293,
      "learning_rate": 9.990476190476191e-05,
      "loss": 3.176,
      "step": 2
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 11.578840255737305,
      "learning_rate": 9.985714285714287e-05,
      "loss": 2.8396,
      "step": 3
    },
    {
      "epoch": 0.0935672514619883,
      "grad_norm": 9.648161888122559,
      "learning_rate": 9.980952380952382e-05,
      "loss": 2.7336,
      "step": 4
    },
    {
      "epoch": 0.11695906432748537,
      "grad_norm": 3.8782691955566406,
      "learning_rate": 9.976190476190477e-05,
      "loss": 2.6955,
      "step": 5
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 12.644487380981445,
      "learning_rate": 9.971428571428571e-05,
      "loss": 2.6405,
      "step": 6
    },
    {
      "epoch": 0.16374269005847952,
      "grad_norm": 11.057122230529785,
      "learning_rate": 9.966666666666667e-05,
      "loss": 2.7113,
      "step": 7
    },
    {
      "epoch": 0.1871345029239766,
      "grad_norm": 4.860190391540527,
      "learning_rate": 9.961904761904762e-05,
      "loss": 2.7076,
      "step": 8
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 4.317215442657471,
      "learning_rate": 9.957142857142858e-05,
      "loss": 2.6377,
      "step": 9
    },
    {
      "epoch": 0.23391812865497075,
      "grad_norm": 3.3068416118621826,
      "learning_rate": 9.952380952380953e-05,
      "loss": 2.5995,
      "step": 10
    },
    {
      "epoch": 0.2573099415204678,
      "grad_norm": 1.2752724885940552,
      "learning_rate": 9.947619047619048e-05,
      "loss": 2.6285,
      "step": 11
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 8.849737167358398,
      "learning_rate": 9.942857142857144e-05,
      "loss": 2.6217,
      "step": 12
    },
    {
      "epoch": 0.30409356725146197,
      "grad_norm": 5.594025611877441,
      "learning_rate": 9.938095238095238e-05,
      "loss": 2.6265,
      "step": 13
    },
    {
      "epoch": 0.32748538011695905,
      "grad_norm": 3.581617593765259,
      "learning_rate": 9.933333333333334e-05,
      "loss": 2.5984,
      "step": 14
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 5.33600378036499,
      "learning_rate": 9.92857142857143e-05,
      "loss": 2.6071,
      "step": 15
    },
    {
      "epoch": 0.3742690058479532,
      "grad_norm": 4.4274983406066895,
      "learning_rate": 9.923809523809524e-05,
      "loss": 2.5908,
      "step": 16
    },
    {
      "epoch": 0.39766081871345027,
      "grad_norm": 4.5507307052612305,
      "learning_rate": 9.91904761904762e-05,
      "loss": 2.5616,
      "step": 17
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 3.189161777496338,
      "learning_rate": 9.914285714285715e-05,
      "loss": 2.5784,
      "step": 18
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 2.706615924835205,
      "learning_rate": 9.909523809523809e-05,
      "loss": 2.5413,
      "step": 19
    },
    {
      "epoch": 0.4678362573099415,
      "grad_norm": 2.146662712097168,
      "learning_rate": 9.904761904761905e-05,
      "loss": 2.6019,
      "step": 20
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 3.2252964973449707,
      "learning_rate": 9.900000000000001e-05,
      "loss": 2.5914,
      "step": 21
    },
    {
      "epoch": 0.5146198830409356,
      "grad_norm": 2.8859879970550537,
      "learning_rate": 9.895238095238095e-05,
      "loss": 2.6133,
      "step": 22
    },
    {
      "epoch": 0.5380116959064327,
      "grad_norm": 2.5647897720336914,
      "learning_rate": 9.890476190476191e-05,
      "loss": 2.5425,
      "step": 23
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 3.0347073078155518,
      "learning_rate": 9.885714285714286e-05,
      "loss": 2.5732,
      "step": 24
    },
    {
      "epoch": 0.5847953216374269,
      "grad_norm": 1.8412858247756958,
      "learning_rate": 9.880952380952381e-05,
      "loss": 2.5776,
      "step": 25
    },
    {
      "epoch": 0.6081871345029239,
      "grad_norm": 3.1820366382598877,
      "learning_rate": 9.876190476190477e-05,
      "loss": 2.5566,
      "step": 26
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 2.4613687992095947,
      "learning_rate": 9.871428571428572e-05,
      "loss": 2.5113,
      "step": 27
    },
    {
      "epoch": 0.6549707602339181,
      "grad_norm": 5.942374229431152,
      "learning_rate": 9.866666666666668e-05,
      "loss": 2.575,
      "step": 28
    },
    {
      "epoch": 0.6783625730994152,
      "grad_norm": 7.427689075469971,
      "learning_rate": 9.861904761904762e-05,
      "loss": 2.5692,
      "step": 29
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 3.163085699081421,
      "learning_rate": 9.857142857142858e-05,
      "loss": 2.5192,
      "step": 30
    },
    {
      "epoch": 0.7251461988304093,
      "grad_norm": 3.421778678894043,
      "learning_rate": 9.852380952380952e-05,
      "loss": 2.5533,
      "step": 31
    },
    {
      "epoch": 0.7485380116959064,
      "grad_norm": 2.486320734024048,
      "learning_rate": 9.847619047619048e-05,
      "loss": 2.515,
      "step": 32
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 3.2825722694396973,
      "learning_rate": 9.842857142857144e-05,
      "loss": 2.4994,
      "step": 33
    },
    {
      "epoch": 0.7953216374269005,
      "grad_norm": 3.5643672943115234,
      "learning_rate": 9.838095238095238e-05,
      "loss": 2.5327,
      "step": 34
    },
    {
      "epoch": 0.8187134502923976,
      "grad_norm": 3.3127200603485107,
      "learning_rate": 9.833333333333333e-05,
      "loss": 2.6251,
      "step": 35
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 3.187095880508423,
      "learning_rate": 9.828571428571429e-05,
      "loss": 2.5233,
      "step": 36
    },
    {
      "epoch": 0.8654970760233918,
      "grad_norm": 3.3743860721588135,
      "learning_rate": 9.823809523809525e-05,
      "loss": 2.4999,
      "step": 37
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 2.3684120178222656,
      "learning_rate": 9.81904761904762e-05,
      "loss": 2.5302,
      "step": 38
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 3.7091619968414307,
      "learning_rate": 9.814285714285715e-05,
      "loss": 2.5003,
      "step": 39
    },
    {
      "epoch": 0.935672514619883,
      "grad_norm": 4.230418682098389,
      "learning_rate": 9.80952380952381e-05,
      "loss": 2.5379,
      "step": 40
    },
    {
      "epoch": 0.9590643274853801,
      "grad_norm": 3.5879616737365723,
      "learning_rate": 9.804761904761905e-05,
      "loss": 2.5652,
      "step": 41
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 2.621013879776001,
      "learning_rate": 9.8e-05,
      "loss": 2.5704,
      "step": 42
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.3536667823791504,
      "learning_rate": 9.795238095238097e-05,
      "loss": 1.9135,
      "step": 43
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.31598106026649475,
      "eval_runtime": 3.4639,
      "eval_samples_per_second": 62.358,
      "eval_steps_per_second": 15.59,
      "step": 43
    },
    {
      "epoch": 1.023391812865497,
      "grad_norm": 2.768101215362549,
      "learning_rate": 9.790476190476191e-05,
      "loss": 2.4537,
      "step": 44
    },
    {
      "epoch": 1.0467836257309941,
      "grad_norm": 2.276510000228882,
      "learning_rate": 9.785714285714286e-05,
      "loss": 2.4703,
      "step": 45
    },
    {
      "epoch": 1.0701754385964912,
      "grad_norm": 3.3676953315734863,
      "learning_rate": 9.780952380952382e-05,
      "loss": 2.5157,
      "step": 46
    },
    {
      "epoch": 1.0935672514619883,
      "grad_norm": 2.5821385383605957,
      "learning_rate": 9.776190476190476e-05,
      "loss": 2.5257,
      "step": 47
    },
    {
      "epoch": 1.1169590643274854,
      "grad_norm": 2.9606378078460693,
      "learning_rate": 9.771428571428572e-05,
      "loss": 2.4374,
      "step": 48
    },
    {
      "epoch": 1.1403508771929824,
      "grad_norm": 3.44348406791687,
      "learning_rate": 9.766666666666668e-05,
      "loss": 2.4795,
      "step": 49
    },
    {
      "epoch": 1.1637426900584795,
      "grad_norm": 2.2815728187561035,
      "learning_rate": 9.761904761904762e-05,
      "loss": 2.4883,
      "step": 50
    },
    {
      "epoch": 1.1871345029239766,
      "grad_norm": 2.4586877822875977,
      "learning_rate": 9.757142857142858e-05,
      "loss": 2.4754,
      "step": 51
    },
    {
      "epoch": 1.2105263157894737,
      "grad_norm": 2.3498520851135254,
      "learning_rate": 9.752380952380953e-05,
      "loss": 2.4607,
      "step": 52
    },
    {
      "epoch": 1.2339181286549707,
      "grad_norm": 2.0736541748046875,
      "learning_rate": 9.747619047619047e-05,
      "loss": 2.4729,
      "step": 53
    },
    {
      "epoch": 1.2573099415204678,
      "grad_norm": 3.3633615970611572,
      "learning_rate": 9.742857142857143e-05,
      "loss": 2.4651,
      "step": 54
    },
    {
      "epoch": 1.280701754385965,
      "grad_norm": 2.9030282497406006,
      "learning_rate": 9.738095238095239e-05,
      "loss": 2.4837,
      "step": 55
    },
    {
      "epoch": 1.304093567251462,
      "grad_norm": 3.896803855895996,
      "learning_rate": 9.733333333333335e-05,
      "loss": 2.53,
      "step": 56
    },
    {
      "epoch": 1.327485380116959,
      "grad_norm": 2.2422099113464355,
      "learning_rate": 9.728571428571429e-05,
      "loss": 2.4911,
      "step": 57
    },
    {
      "epoch": 1.3508771929824561,
      "grad_norm": 3.7257282733917236,
      "learning_rate": 9.723809523809524e-05,
      "loss": 2.5117,
      "step": 58
    },
    {
      "epoch": 1.3742690058479532,
      "grad_norm": 4.366884231567383,
      "learning_rate": 9.71904761904762e-05,
      "loss": 2.5469,
      "step": 59
    },
    {
      "epoch": 1.3976608187134503,
      "grad_norm": 2.365870714187622,
      "learning_rate": 9.714285714285715e-05,
      "loss": 2.522,
      "step": 60
    },
    {
      "epoch": 1.4210526315789473,
      "grad_norm": 2.6924397945404053,
      "learning_rate": 9.70952380952381e-05,
      "loss": 2.5042,
      "step": 61
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 2.8113691806793213,
      "learning_rate": 9.704761904761905e-05,
      "loss": 2.5424,
      "step": 62
    },
    {
      "epoch": 1.4678362573099415,
      "grad_norm": 2.295470952987671,
      "learning_rate": 9.7e-05,
      "loss": 2.5205,
      "step": 63
    },
    {
      "epoch": 1.4912280701754386,
      "grad_norm": 4.694986343383789,
      "learning_rate": 9.695238095238096e-05,
      "loss": 2.503,
      "step": 64
    },
    {
      "epoch": 1.5146198830409356,
      "grad_norm": 2.4416017532348633,
      "learning_rate": 9.69047619047619e-05,
      "loss": 2.5223,
      "step": 65
    },
    {
      "epoch": 1.5380116959064327,
      "grad_norm": 4.63900899887085,
      "learning_rate": 9.685714285714286e-05,
      "loss": 2.4851,
      "step": 66
    },
    {
      "epoch": 1.5614035087719298,
      "grad_norm": 2.8212132453918457,
      "learning_rate": 9.680952380952382e-05,
      "loss": 2.5191,
      "step": 67
    },
    {
      "epoch": 1.5847953216374269,
      "grad_norm": 2.3161513805389404,
      "learning_rate": 9.676190476190476e-05,
      "loss": 2.4891,
      "step": 68
    },
    {
      "epoch": 1.608187134502924,
      "grad_norm": 2.707838535308838,
      "learning_rate": 9.671428571428572e-05,
      "loss": 2.4927,
      "step": 69
    },
    {
      "epoch": 1.631578947368421,
      "grad_norm": 1.9635908603668213,
      "learning_rate": 9.666666666666667e-05,
      "loss": 2.4834,
      "step": 70
    },
    {
      "epoch": 1.654970760233918,
      "grad_norm": 2.5164906978607178,
      "learning_rate": 9.661904761904763e-05,
      "loss": 2.4375,
      "step": 71
    },
    {
      "epoch": 1.6783625730994152,
      "grad_norm": 2.799863576889038,
      "learning_rate": 9.657142857142858e-05,
      "loss": 2.4754,
      "step": 72
    },
    {
      "epoch": 1.7017543859649122,
      "grad_norm": 3.619720935821533,
      "learning_rate": 9.652380952380953e-05,
      "loss": 2.511,
      "step": 73
    },
    {
      "epoch": 1.7251461988304093,
      "grad_norm": 3.7119972705841064,
      "learning_rate": 9.647619047619047e-05,
      "loss": 2.4723,
      "step": 74
    },
    {
      "epoch": 1.7485380116959064,
      "grad_norm": 4.552167892456055,
      "learning_rate": 9.642857142857143e-05,
      "loss": 2.4928,
      "step": 75
    },
    {
      "epoch": 1.7719298245614035,
      "grad_norm": 2.3658089637756348,
      "learning_rate": 9.638095238095238e-05,
      "loss": 2.5233,
      "step": 76
    },
    {
      "epoch": 1.7953216374269005,
      "grad_norm": 2.2602102756500244,
      "learning_rate": 9.633333333333335e-05,
      "loss": 2.493,
      "step": 77
    },
    {
      "epoch": 1.8187134502923976,
      "grad_norm": 2.2393336296081543,
      "learning_rate": 9.628571428571429e-05,
      "loss": 2.4743,
      "step": 78
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 4.036160469055176,
      "learning_rate": 9.623809523809524e-05,
      "loss": 2.4973,
      "step": 79
    },
    {
      "epoch": 1.8654970760233918,
      "grad_norm": 2.1540491580963135,
      "learning_rate": 9.61904761904762e-05,
      "loss": 2.5163,
      "step": 80
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 3.3976409435272217,
      "learning_rate": 9.614285714285714e-05,
      "loss": 2.4244,
      "step": 81
    },
    {
      "epoch": 1.912280701754386,
      "grad_norm": 2.306898832321167,
      "learning_rate": 9.60952380952381e-05,
      "loss": 2.4487,
      "step": 82
    },
    {
      "epoch": 1.935672514619883,
      "grad_norm": 2.8536524772644043,
      "learning_rate": 9.604761904761906e-05,
      "loss": 2.4441,
      "step": 83
    },
    {
      "epoch": 1.95906432748538,
      "grad_norm": 2.7067620754241943,
      "learning_rate": 9.6e-05,
      "loss": 2.5898,
      "step": 84
    },
    {
      "epoch": 1.9824561403508771,
      "grad_norm": 2.6597952842712402,
      "learning_rate": 9.595238095238096e-05,
      "loss": 2.4564,
      "step": 85
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.054412603378296,
      "learning_rate": 9.59047619047619e-05,
      "loss": 1.8442,
      "step": 86
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.31219348311424255,
      "eval_runtime": 3.4985,
      "eval_samples_per_second": 61.74,
      "eval_steps_per_second": 15.435,
      "step": 86
    },
    {
      "epoch": 2.023391812865497,
      "grad_norm": 2.8484723567962646,
      "learning_rate": 9.585714285714285e-05,
      "loss": 2.4903,
      "step": 87
    },
    {
      "epoch": 2.046783625730994,
      "grad_norm": 3.8947722911834717,
      "learning_rate": 9.580952380952382e-05,
      "loss": 2.486,
      "step": 88
    },
    {
      "epoch": 2.0701754385964914,
      "grad_norm": 4.0316386222839355,
      "learning_rate": 9.576190476190477e-05,
      "loss": 2.4559,
      "step": 89
    },
    {
      "epoch": 2.0935672514619883,
      "grad_norm": 2.2441468238830566,
      "learning_rate": 9.571428571428573e-05,
      "loss": 2.4548,
      "step": 90
    },
    {
      "epoch": 2.116959064327485,
      "grad_norm": 4.1536359786987305,
      "learning_rate": 9.566666666666667e-05,
      "loss": 2.4341,
      "step": 91
    },
    {
      "epoch": 2.1403508771929824,
      "grad_norm": 5.011383533477783,
      "learning_rate": 9.561904761904761e-05,
      "loss": 2.4233,
      "step": 92
    },
    {
      "epoch": 2.1637426900584797,
      "grad_norm": 2.0493083000183105,
      "learning_rate": 9.557142857142857e-05,
      "loss": 2.4691,
      "step": 93
    },
    {
      "epoch": 2.1871345029239766,
      "grad_norm": 4.27736759185791,
      "learning_rate": 9.552380952380953e-05,
      "loss": 2.4666,
      "step": 94
    },
    {
      "epoch": 2.2105263157894735,
      "grad_norm": 3.795992136001587,
      "learning_rate": 9.547619047619049e-05,
      "loss": 2.4617,
      "step": 95
    },
    {
      "epoch": 2.2339181286549707,
      "grad_norm": 2.451568365097046,
      "learning_rate": 9.542857142857143e-05,
      "loss": 2.4485,
      "step": 96
    },
    {
      "epoch": 2.257309941520468,
      "grad_norm": 2.0920753479003906,
      "learning_rate": 9.538095238095238e-05,
      "loss": 2.4411,
      "step": 97
    },
    {
      "epoch": 2.280701754385965,
      "grad_norm": 3.969087600708008,
      "learning_rate": 9.533333333333334e-05,
      "loss": 2.47,
      "step": 98
    },
    {
      "epoch": 2.3040935672514617,
      "grad_norm": 3.598999500274658,
      "learning_rate": 9.52857142857143e-05,
      "loss": 2.4431,
      "step": 99
    },
    {
      "epoch": 2.327485380116959,
      "grad_norm": 1.720585584640503,
      "learning_rate": 9.523809523809524e-05,
      "loss": 2.4784,
      "step": 100
    },
    {
      "epoch": 2.3508771929824563,
      "grad_norm": 2.879593849182129,
      "learning_rate": 9.51904761904762e-05,
      "loss": 2.4321,
      "step": 101
    },
    {
      "epoch": 2.374269005847953,
      "grad_norm": 1.448943853378296,
      "learning_rate": 9.514285714285714e-05,
      "loss": 2.4054,
      "step": 102
    },
    {
      "epoch": 2.39766081871345,
      "grad_norm": 2.534806489944458,
      "learning_rate": 9.50952380952381e-05,
      "loss": 2.4682,
      "step": 103
    },
    {
      "epoch": 2.4210526315789473,
      "grad_norm": 2.2084572315216064,
      "learning_rate": 9.504761904761905e-05,
      "loss": 2.4459,
      "step": 104
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 2.6770689487457275,
      "learning_rate": 9.5e-05,
      "loss": 2.4461,
      "step": 105
    },
    {
      "epoch": 2.4678362573099415,
      "grad_norm": 2.4759695529937744,
      "learning_rate": 9.495238095238096e-05,
      "loss": 2.4658,
      "step": 106
    },
    {
      "epoch": 2.4912280701754383,
      "grad_norm": 2.4558472633361816,
      "learning_rate": 9.490476190476191e-05,
      "loss": 2.4309,
      "step": 107
    },
    {
      "epoch": 2.5146198830409356,
      "grad_norm": 2.584268808364868,
      "learning_rate": 9.485714285714287e-05,
      "loss": 2.4985,
      "step": 108
    },
    {
      "epoch": 2.538011695906433,
      "grad_norm": 3.1202220916748047,
      "learning_rate": 9.480952380952381e-05,
      "loss": 2.4503,
      "step": 109
    },
    {
      "epoch": 2.56140350877193,
      "grad_norm": 3.697275161743164,
      "learning_rate": 9.476190476190476e-05,
      "loss": 2.4021,
      "step": 110
    },
    {
      "epoch": 2.5847953216374266,
      "grad_norm": 2.602041482925415,
      "learning_rate": 9.471428571428573e-05,
      "loss": 2.5068,
      "step": 111
    },
    {
      "epoch": 2.608187134502924,
      "grad_norm": 3.9455575942993164,
      "learning_rate": 9.466666666666667e-05,
      "loss": 2.4394,
      "step": 112
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 3.7310194969177246,
      "learning_rate": 9.461904761904762e-05,
      "loss": 2.4865,
      "step": 113
    },
    {
      "epoch": 2.654970760233918,
      "grad_norm": 4.220344066619873,
      "learning_rate": 9.457142857142858e-05,
      "loss": 2.4982,
      "step": 114
    },
    {
      "epoch": 2.678362573099415,
      "grad_norm": 1.7343074083328247,
      "learning_rate": 9.452380952380952e-05,
      "loss": 2.4484,
      "step": 115
    },
    {
      "epoch": 2.7017543859649122,
      "grad_norm": 3.8400301933288574,
      "learning_rate": 9.447619047619048e-05,
      "loss": 2.5013,
      "step": 116
    },
    {
      "epoch": 2.7251461988304095,
      "grad_norm": 5.945042133331299,
      "learning_rate": 9.442857142857144e-05,
      "loss": 2.5052,
      "step": 117
    },
    {
      "epoch": 2.7485380116959064,
      "grad_norm": 2.8387110233306885,
      "learning_rate": 9.438095238095238e-05,
      "loss": 2.5203,
      "step": 118
    },
    {
      "epoch": 2.7719298245614032,
      "grad_norm": 3.0027477741241455,
      "learning_rate": 9.433333333333334e-05,
      "loss": 2.4686,
      "step": 119
    },
    {
      "epoch": 2.7953216374269005,
      "grad_norm": 4.645102024078369,
      "learning_rate": 9.428571428571429e-05,
      "loss": 2.5442,
      "step": 120
    },
    {
      "epoch": 2.818713450292398,
      "grad_norm": 3.3703556060791016,
      "learning_rate": 9.423809523809524e-05,
      "loss": 2.4585,
      "step": 121
    },
    {
      "epoch": 2.8421052631578947,
      "grad_norm": 2.3330347537994385,
      "learning_rate": 9.41904761904762e-05,
      "loss": 2.4219,
      "step": 122
    },
    {
      "epoch": 2.8654970760233915,
      "grad_norm": 4.309964179992676,
      "learning_rate": 9.414285714285715e-05,
      "loss": 2.5121,
      "step": 123
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 4.484703540802002,
      "learning_rate": 9.40952380952381e-05,
      "loss": 2.5512,
      "step": 124
    },
    {
      "epoch": 2.912280701754386,
      "grad_norm": 3.8170957565307617,
      "learning_rate": 9.404761904761905e-05,
      "loss": 2.4961,
      "step": 125
    },
    {
      "epoch": 2.935672514619883,
      "grad_norm": 3.1967039108276367,
      "learning_rate": 9.4e-05,
      "loss": 2.4039,
      "step": 126
    },
    {
      "epoch": 2.95906432748538,
      "grad_norm": 3.305359363555908,
      "learning_rate": 9.395238095238095e-05,
      "loss": 2.4389,
      "step": 127
    },
    {
      "epoch": 2.982456140350877,
      "grad_norm": 3.0017144680023193,
      "learning_rate": 9.390476190476191e-05,
      "loss": 2.4839,
      "step": 128
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.907825231552124,
      "learning_rate": 9.385714285714287e-05,
      "loss": 1.7774,
      "step": 129
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.3095262944698334,
      "eval_runtime": 3.5049,
      "eval_samples_per_second": 61.627,
      "eval_steps_per_second": 15.407,
      "step": 129
    },
    {
      "epoch": 3.023391812865497,
      "grad_norm": 2.4059183597564697,
      "learning_rate": 9.380952380952381e-05,
      "loss": 2.4009,
      "step": 130
    },
    {
      "epoch": 3.046783625730994,
      "grad_norm": 2.391845226287842,
      "learning_rate": 9.376190476190476e-05,
      "loss": 2.4191,
      "step": 131
    },
    {
      "epoch": 3.0701754385964914,
      "grad_norm": 3.139408826828003,
      "learning_rate": 9.371428571428572e-05,
      "loss": 2.391,
      "step": 132
    },
    {
      "epoch": 3.0935672514619883,
      "grad_norm": 3.1654696464538574,
      "learning_rate": 9.366666666666668e-05,
      "loss": 2.3909,
      "step": 133
    },
    {
      "epoch": 3.116959064327485,
      "grad_norm": 1.5735834836959839,
      "learning_rate": 9.361904761904763e-05,
      "loss": 2.4318,
      "step": 134
    },
    {
      "epoch": 3.1403508771929824,
      "grad_norm": 2.855710029602051,
      "learning_rate": 9.357142857142858e-05,
      "loss": 2.3899,
      "step": 135
    },
    {
      "epoch": 3.1637426900584797,
      "grad_norm": 2.6927037239074707,
      "learning_rate": 9.352380952380952e-05,
      "loss": 2.4136,
      "step": 136
    },
    {
      "epoch": 3.1871345029239766,
      "grad_norm": 2.0338966846466064,
      "learning_rate": 9.347619047619048e-05,
      "loss": 2.3805,
      "step": 137
    },
    {
      "epoch": 3.2105263157894735,
      "grad_norm": 3.746018409729004,
      "learning_rate": 9.342857142857143e-05,
      "loss": 2.4159,
      "step": 138
    },
    {
      "epoch": 3.2339181286549707,
      "grad_norm": 2.1957383155822754,
      "learning_rate": 9.338095238095238e-05,
      "loss": 2.4393,
      "step": 139
    },
    {
      "epoch": 3.257309941520468,
      "grad_norm": 2.063310384750366,
      "learning_rate": 9.333333333333334e-05,
      "loss": 2.3983,
      "step": 140
    },
    {
      "epoch": 3.280701754385965,
      "grad_norm": 1.9041073322296143,
      "learning_rate": 9.328571428571429e-05,
      "loss": 2.4438,
      "step": 141
    },
    {
      "epoch": 3.3040935672514617,
      "grad_norm": 2.0360188484191895,
      "learning_rate": 9.323809523809525e-05,
      "loss": 2.4025,
      "step": 142
    },
    {
      "epoch": 3.327485380116959,
      "grad_norm": 2.820622444152832,
      "learning_rate": 9.319047619047619e-05,
      "loss": 2.4353,
      "step": 143
    },
    {
      "epoch": 3.3508771929824563,
      "grad_norm": 2.129838705062866,
      "learning_rate": 9.314285714285715e-05,
      "loss": 2.4528,
      "step": 144
    },
    {
      "epoch": 3.374269005847953,
      "grad_norm": 2.2732796669006348,
      "learning_rate": 9.309523809523811e-05,
      "loss": 2.4082,
      "step": 145
    },
    {
      "epoch": 3.39766081871345,
      "grad_norm": 2.280691623687744,
      "learning_rate": 9.304761904761905e-05,
      "loss": 2.4521,
      "step": 146
    },
    {
      "epoch": 3.4210526315789473,
      "grad_norm": 2.549703598022461,
      "learning_rate": 9.300000000000001e-05,
      "loss": 2.4777,
      "step": 147
    },
    {
      "epoch": 3.4444444444444446,
      "grad_norm": 1.628199815750122,
      "learning_rate": 9.295238095238096e-05,
      "loss": 2.3744,
      "step": 148
    },
    {
      "epoch": 3.4678362573099415,
      "grad_norm": 2.7309587001800537,
      "learning_rate": 9.29047619047619e-05,
      "loss": 2.4787,
      "step": 149
    },
    {
      "epoch": 3.4912280701754383,
      "grad_norm": 2.2030370235443115,
      "learning_rate": 9.285714285714286e-05,
      "loss": 2.4376,
      "step": 150
    },
    {
      "epoch": 3.5146198830409356,
      "grad_norm": 2.3776726722717285,
      "learning_rate": 9.280952380952382e-05,
      "loss": 2.4595,
      "step": 151
    },
    {
      "epoch": 3.538011695906433,
      "grad_norm": 3.5721945762634277,
      "learning_rate": 9.276190476190476e-05,
      "loss": 2.4642,
      "step": 152
    },
    {
      "epoch": 3.56140350877193,
      "grad_norm": 3.3094141483306885,
      "learning_rate": 9.271428571428572e-05,
      "loss": 2.4229,
      "step": 153
    },
    {
      "epoch": 3.5847953216374266,
      "grad_norm": 2.197993278503418,
      "learning_rate": 9.266666666666666e-05,
      "loss": 2.3844,
      "step": 154
    },
    {
      "epoch": 3.608187134502924,
      "grad_norm": 3.706082582473755,
      "learning_rate": 9.261904761904762e-05,
      "loss": 2.4332,
      "step": 155
    },
    {
      "epoch": 3.6315789473684212,
      "grad_norm": 2.791156053543091,
      "learning_rate": 9.257142857142858e-05,
      "loss": 2.448,
      "step": 156
    },
    {
      "epoch": 3.654970760233918,
      "grad_norm": 4.078625679016113,
      "learning_rate": 9.252380952380953e-05,
      "loss": 2.4481,
      "step": 157
    },
    {
      "epoch": 3.678362573099415,
      "grad_norm": 2.616171360015869,
      "learning_rate": 9.247619047619048e-05,
      "loss": 2.4338,
      "step": 158
    },
    {
      "epoch": 3.7017543859649122,
      "grad_norm": 3.2940022945404053,
      "learning_rate": 9.242857142857143e-05,
      "loss": 2.4327,
      "step": 159
    },
    {
      "epoch": 3.7251461988304095,
      "grad_norm": 5.047934532165527,
      "learning_rate": 9.238095238095239e-05,
      "loss": 2.4671,
      "step": 160
    },
    {
      "epoch": 3.7485380116959064,
      "grad_norm": 2.0488369464874268,
      "learning_rate": 9.233333333333333e-05,
      "loss": 2.4354,
      "step": 161
    },
    {
      "epoch": 3.7719298245614032,
      "grad_norm": 5.099597454071045,
      "learning_rate": 9.228571428571429e-05,
      "loss": 2.4336,
      "step": 162
    },
    {
      "epoch": 3.7953216374269005,
      "grad_norm": 3.812992572784424,
      "learning_rate": 9.223809523809525e-05,
      "loss": 2.4586,
      "step": 163
    },
    {
      "epoch": 3.818713450292398,
      "grad_norm": 3.237950086593628,
      "learning_rate": 9.21904761904762e-05,
      "loss": 2.4583,
      "step": 164
    },
    {
      "epoch": 3.8421052631578947,
      "grad_norm": 5.422150135040283,
      "learning_rate": 9.214285714285714e-05,
      "loss": 2.4427,
      "step": 165
    },
    {
      "epoch": 3.8654970760233915,
      "grad_norm": 3.7532801628112793,
      "learning_rate": 9.20952380952381e-05,
      "loss": 2.4103,
      "step": 166
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 3.172743320465088,
      "learning_rate": 9.204761904761906e-05,
      "loss": 2.4427,
      "step": 167
    },
    {
      "epoch": 3.912280701754386,
      "grad_norm": 5.759428977966309,
      "learning_rate": 9.200000000000001e-05,
      "loss": 2.4811,
      "step": 168
    },
    {
      "epoch": 3.935672514619883,
      "grad_norm": 2.9414196014404297,
      "learning_rate": 9.195238095238096e-05,
      "loss": 2.4602,
      "step": 169
    },
    {
      "epoch": 3.95906432748538,
      "grad_norm": 4.5333662033081055,
      "learning_rate": 9.19047619047619e-05,
      "loss": 2.4738,
      "step": 170
    },
    {
      "epoch": 3.982456140350877,
      "grad_norm": 3.230017900466919,
      "learning_rate": 9.185714285714286e-05,
      "loss": 2.4134,
      "step": 171
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.4366929531097412,
      "learning_rate": 9.18095238095238e-05,
      "loss": 1.8099,
      "step": 172
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.31109535694122314,
      "eval_runtime": 3.5128,
      "eval_samples_per_second": 61.489,
      "eval_steps_per_second": 15.372,
      "step": 172
    },
    {
      "epoch": 4.023391812865497,
      "grad_norm": 4.1390509605407715,
      "learning_rate": 9.176190476190476e-05,
      "loss": 2.4107,
      "step": 173
    },
    {
      "epoch": 4.046783625730994,
      "grad_norm": 4.0564985275268555,
      "learning_rate": 9.171428571428572e-05,
      "loss": 2.3704,
      "step": 174
    },
    {
      "epoch": 4.0701754385964914,
      "grad_norm": 3.3425748348236084,
      "learning_rate": 9.166666666666667e-05,
      "loss": 2.3986,
      "step": 175
    },
    {
      "epoch": 4.093567251461988,
      "grad_norm": 3.1439638137817383,
      "learning_rate": 9.161904761904763e-05,
      "loss": 2.3725,
      "step": 176
    },
    {
      "epoch": 4.116959064327485,
      "grad_norm": 2.531954765319824,
      "learning_rate": 9.157142857142857e-05,
      "loss": 2.4327,
      "step": 177
    },
    {
      "epoch": 4.140350877192983,
      "grad_norm": 3.1837306022644043,
      "learning_rate": 9.152380952380953e-05,
      "loss": 2.3846,
      "step": 178
    },
    {
      "epoch": 4.16374269005848,
      "grad_norm": 3.197889804840088,
      "learning_rate": 9.147619047619049e-05,
      "loss": 2.3887,
      "step": 179
    },
    {
      "epoch": 4.187134502923977,
      "grad_norm": 2.6710753440856934,
      "learning_rate": 9.142857142857143e-05,
      "loss": 2.4217,
      "step": 180
    },
    {
      "epoch": 4.2105263157894735,
      "grad_norm": 2.70224928855896,
      "learning_rate": 9.138095238095239e-05,
      "loss": 2.4039,
      "step": 181
    },
    {
      "epoch": 4.23391812865497,
      "grad_norm": 2.7430830001831055,
      "learning_rate": 9.133333333333334e-05,
      "loss": 2.4032,
      "step": 182
    },
    {
      "epoch": 4.257309941520468,
      "grad_norm": 2.428302764892578,
      "learning_rate": 9.128571428571428e-05,
      "loss": 2.394,
      "step": 183
    },
    {
      "epoch": 4.280701754385965,
      "grad_norm": 4.183929920196533,
      "learning_rate": 9.123809523809524e-05,
      "loss": 2.3949,
      "step": 184
    },
    {
      "epoch": 4.304093567251462,
      "grad_norm": 2.312898874282837,
      "learning_rate": 9.11904761904762e-05,
      "loss": 2.4049,
      "step": 185
    },
    {
      "epoch": 4.3274853801169595,
      "grad_norm": 2.6628363132476807,
      "learning_rate": 9.114285714285716e-05,
      "loss": 2.3866,
      "step": 186
    },
    {
      "epoch": 4.350877192982456,
      "grad_norm": 2.258638381958008,
      "learning_rate": 9.10952380952381e-05,
      "loss": 2.3608,
      "step": 187
    },
    {
      "epoch": 4.374269005847953,
      "grad_norm": 2.543140172958374,
      "learning_rate": 9.104761904761904e-05,
      "loss": 2.4382,
      "step": 188
    },
    {
      "epoch": 4.39766081871345,
      "grad_norm": 2.7995975017547607,
      "learning_rate": 9.1e-05,
      "loss": 2.3809,
      "step": 189
    },
    {
      "epoch": 4.421052631578947,
      "grad_norm": 2.617652654647827,
      "learning_rate": 9.095238095238096e-05,
      "loss": 2.3999,
      "step": 190
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 2.1277976036071777,
      "learning_rate": 9.09047619047619e-05,
      "loss": 2.3982,
      "step": 191
    },
    {
      "epoch": 4.4678362573099415,
      "grad_norm": 3.09194016456604,
      "learning_rate": 9.085714285714286e-05,
      "loss": 2.4261,
      "step": 192
    },
    {
      "epoch": 4.491228070175438,
      "grad_norm": 2.2097442150115967,
      "learning_rate": 9.080952380952381e-05,
      "loss": 2.4457,
      "step": 193
    },
    {
      "epoch": 4.514619883040936,
      "grad_norm": 2.3321805000305176,
      "learning_rate": 9.076190476190477e-05,
      "loss": 2.3977,
      "step": 194
    },
    {
      "epoch": 4.538011695906433,
      "grad_norm": 2.72891902923584,
      "learning_rate": 9.071428571428571e-05,
      "loss": 2.4663,
      "step": 195
    },
    {
      "epoch": 4.56140350877193,
      "grad_norm": 2.142327070236206,
      "learning_rate": 9.066666666666667e-05,
      "loss": 2.4171,
      "step": 196
    },
    {
      "epoch": 4.584795321637427,
      "grad_norm": 2.173485517501831,
      "learning_rate": 9.061904761904763e-05,
      "loss": 2.4219,
      "step": 197
    },
    {
      "epoch": 4.6081871345029235,
      "grad_norm": 2.439042568206787,
      "learning_rate": 9.057142857142857e-05,
      "loss": 2.3669,
      "step": 198
    },
    {
      "epoch": 4.631578947368421,
      "grad_norm": 3.0206170082092285,
      "learning_rate": 9.052380952380953e-05,
      "loss": 2.4399,
      "step": 199
    },
    {
      "epoch": 4.654970760233918,
      "grad_norm": 2.1517748832702637,
      "learning_rate": 9.047619047619048e-05,
      "loss": 2.4484,
      "step": 200
    },
    {
      "epoch": 4.678362573099415,
      "grad_norm": 3.037619113922119,
      "learning_rate": 9.042857142857143e-05,
      "loss": 2.44,
      "step": 201
    },
    {
      "epoch": 4.701754385964913,
      "grad_norm": 2.569814920425415,
      "learning_rate": 9.03809523809524e-05,
      "loss": 2.3586,
      "step": 202
    },
    {
      "epoch": 4.7251461988304095,
      "grad_norm": 2.983091115951538,
      "learning_rate": 9.033333333333334e-05,
      "loss": 2.3362,
      "step": 203
    },
    {
      "epoch": 4.748538011695906,
      "grad_norm": 2.2038419246673584,
      "learning_rate": 9.028571428571428e-05,
      "loss": 2.4397,
      "step": 204
    },
    {
      "epoch": 4.771929824561403,
      "grad_norm": 1.9491034746170044,
      "learning_rate": 9.023809523809524e-05,
      "loss": 2.3986,
      "step": 205
    },
    {
      "epoch": 4.7953216374269,
      "grad_norm": 1.9633510112762451,
      "learning_rate": 9.019047619047619e-05,
      "loss": 2.4017,
      "step": 206
    },
    {
      "epoch": 4.818713450292398,
      "grad_norm": 5.065985679626465,
      "learning_rate": 9.014285714285716e-05,
      "loss": 2.4338,
      "step": 207
    },
    {
      "epoch": 4.842105263157895,
      "grad_norm": 2.1132020950317383,
      "learning_rate": 9.00952380952381e-05,
      "loss": 2.3536,
      "step": 208
    },
    {
      "epoch": 4.8654970760233915,
      "grad_norm": 4.070071697235107,
      "learning_rate": 9.004761904761905e-05,
      "loss": 2.4411,
      "step": 209
    },
    {
      "epoch": 4.888888888888889,
      "grad_norm": 3.169827699661255,
      "learning_rate": 9e-05,
      "loss": 2.4099,
      "step": 210
    },
    {
      "epoch": 4.912280701754386,
      "grad_norm": 2.128169059753418,
      "learning_rate": 8.995238095238095e-05,
      "loss": 2.4253,
      "step": 211
    },
    {
      "epoch": 4.935672514619883,
      "grad_norm": 4.3199076652526855,
      "learning_rate": 8.990476190476191e-05,
      "loss": 2.4334,
      "step": 212
    },
    {
      "epoch": 4.95906432748538,
      "grad_norm": 2.538106679916382,
      "learning_rate": 8.985714285714287e-05,
      "loss": 2.4465,
      "step": 213
    },
    {
      "epoch": 4.982456140350877,
      "grad_norm": 4.168552398681641,
      "learning_rate": 8.980952380952381e-05,
      "loss": 2.3854,
      "step": 214
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.4958627223968506,
      "learning_rate": 8.976190476190477e-05,
      "loss": 1.8025,
      "step": 215
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.30952247977256775,
      "eval_runtime": 3.5036,
      "eval_samples_per_second": 61.651,
      "eval_steps_per_second": 15.413,
      "step": 215
    },
    {
      "epoch": 5.023391812865497,
      "grad_norm": 2.416876792907715,
      "learning_rate": 8.971428571428571e-05,
      "loss": 2.3852,
      "step": 216
    },
    {
      "epoch": 5.046783625730994,
      "grad_norm": 3.6754486560821533,
      "learning_rate": 8.966666666666666e-05,
      "loss": 2.4351,
      "step": 217
    },
    {
      "epoch": 5.0701754385964914,
      "grad_norm": 3.146263599395752,
      "learning_rate": 8.961904761904762e-05,
      "loss": 2.4231,
      "step": 218
    },
    {
      "epoch": 5.093567251461988,
      "grad_norm": 2.729116439819336,
      "learning_rate": 8.957142857142858e-05,
      "loss": 2.3589,
      "step": 219
    },
    {
      "epoch": 5.116959064327485,
      "grad_norm": 6.3045477867126465,
      "learning_rate": 8.952380952380953e-05,
      "loss": 2.3972,
      "step": 220
    },
    {
      "epoch": 5.140350877192983,
      "grad_norm": 4.70556116104126,
      "learning_rate": 8.947619047619048e-05,
      "loss": 2.3614,
      "step": 221
    },
    {
      "epoch": 5.16374269005848,
      "grad_norm": 4.070709705352783,
      "learning_rate": 8.942857142857142e-05,
      "loss": 2.3298,
      "step": 222
    },
    {
      "epoch": 5.187134502923977,
      "grad_norm": 5.384773254394531,
      "learning_rate": 8.938095238095238e-05,
      "loss": 2.4206,
      "step": 223
    },
    {
      "epoch": 5.2105263157894735,
      "grad_norm": 2.430166721343994,
      "learning_rate": 8.933333333333334e-05,
      "loss": 2.4027,
      "step": 224
    },
    {
      "epoch": 5.23391812865497,
      "grad_norm": 2.5975165367126465,
      "learning_rate": 8.92857142857143e-05,
      "loss": 2.4035,
      "step": 225
    },
    {
      "epoch": 5.257309941520468,
      "grad_norm": 2.0980706214904785,
      "learning_rate": 8.923809523809524e-05,
      "loss": 2.3119,
      "step": 226
    },
    {
      "epoch": 5.280701754385965,
      "grad_norm": 3.480513095855713,
      "learning_rate": 8.919047619047619e-05,
      "loss": 2.3753,
      "step": 227
    },
    {
      "epoch": 5.304093567251462,
      "grad_norm": 2.642237901687622,
      "learning_rate": 8.914285714285715e-05,
      "loss": 2.3885,
      "step": 228
    },
    {
      "epoch": 5.3274853801169595,
      "grad_norm": 3.1167006492614746,
      "learning_rate": 8.909523809523809e-05,
      "loss": 2.3668,
      "step": 229
    },
    {
      "epoch": 5.350877192982456,
      "grad_norm": 2.7304916381835938,
      "learning_rate": 8.904761904761905e-05,
      "loss": 2.3474,
      "step": 230
    },
    {
      "epoch": 5.374269005847953,
      "grad_norm": 2.4452908039093018,
      "learning_rate": 8.900000000000001e-05,
      "loss": 2.3501,
      "step": 231
    },
    {
      "epoch": 5.39766081871345,
      "grad_norm": 3.0729784965515137,
      "learning_rate": 8.895238095238095e-05,
      "loss": 2.4057,
      "step": 232
    },
    {
      "epoch": 5.421052631578947,
      "grad_norm": 2.315704345703125,
      "learning_rate": 8.890476190476191e-05,
      "loss": 2.3416,
      "step": 233
    },
    {
      "epoch": 5.444444444444445,
      "grad_norm": 2.8339195251464844,
      "learning_rate": 8.885714285714286e-05,
      "loss": 2.3419,
      "step": 234
    },
    {
      "epoch": 5.4678362573099415,
      "grad_norm": 4.628677845001221,
      "learning_rate": 8.880952380952381e-05,
      "loss": 2.3926,
      "step": 235
    },
    {
      "epoch": 5.491228070175438,
      "grad_norm": 3.533191204071045,
      "learning_rate": 8.876190476190477e-05,
      "loss": 2.4157,
      "step": 236
    },
    {
      "epoch": 5.514619883040936,
      "grad_norm": 2.3621981143951416,
      "learning_rate": 8.871428571428572e-05,
      "loss": 2.3415,
      "step": 237
    },
    {
      "epoch": 5.538011695906433,
      "grad_norm": 1.851884365081787,
      "learning_rate": 8.866666666666668e-05,
      "loss": 2.3789,
      "step": 238
    },
    {
      "epoch": 5.56140350877193,
      "grad_norm": 2.5211541652679443,
      "learning_rate": 8.861904761904762e-05,
      "loss": 2.3055,
      "step": 239
    },
    {
      "epoch": 5.584795321637427,
      "grad_norm": 3.753197431564331,
      "learning_rate": 8.857142857142857e-05,
      "loss": 2.3666,
      "step": 240
    },
    {
      "epoch": 5.6081871345029235,
      "grad_norm": 2.3129889965057373,
      "learning_rate": 8.852380952380954e-05,
      "loss": 2.3692,
      "step": 241
    },
    {
      "epoch": 5.631578947368421,
      "grad_norm": 2.3051323890686035,
      "learning_rate": 8.847619047619048e-05,
      "loss": 2.4062,
      "step": 242
    },
    {
      "epoch": 5.654970760233918,
      "grad_norm": 2.1116273403167725,
      "learning_rate": 8.842857142857143e-05,
      "loss": 2.3403,
      "step": 243
    },
    {
      "epoch": 5.678362573099415,
      "grad_norm": 2.010376214981079,
      "learning_rate": 8.838095238095239e-05,
      "loss": 2.2977,
      "step": 244
    },
    {
      "epoch": 5.701754385964913,
      "grad_norm": 1.641919732093811,
      "learning_rate": 8.833333333333333e-05,
      "loss": 2.3663,
      "step": 245
    },
    {
      "epoch": 5.7251461988304095,
      "grad_norm": 2.6189534664154053,
      "learning_rate": 8.828571428571429e-05,
      "loss": 2.3651,
      "step": 246
    },
    {
      "epoch": 5.748538011695906,
      "grad_norm": 2.2570645809173584,
      "learning_rate": 8.823809523809525e-05,
      "loss": 2.3802,
      "step": 247
    },
    {
      "epoch": 5.771929824561403,
      "grad_norm": 2.970245599746704,
      "learning_rate": 8.819047619047619e-05,
      "loss": 2.3843,
      "step": 248
    },
    {
      "epoch": 5.7953216374269,
      "grad_norm": 3.3681797981262207,
      "learning_rate": 8.814285714285715e-05,
      "loss": 2.3726,
      "step": 249
    },
    {
      "epoch": 5.818713450292398,
      "grad_norm": 2.2378933429718018,
      "learning_rate": 8.80952380952381e-05,
      "loss": 2.431,
      "step": 250
    },
    {
      "epoch": 5.842105263157895,
      "grad_norm": 3.3535726070404053,
      "learning_rate": 8.804761904761905e-05,
      "loss": 2.3884,
      "step": 251
    },
    {
      "epoch": 5.8654970760233915,
      "grad_norm": 2.3131027221679688,
      "learning_rate": 8.800000000000001e-05,
      "loss": 2.4059,
      "step": 252
    },
    {
      "epoch": 5.888888888888889,
      "grad_norm": 2.942883253097534,
      "learning_rate": 8.795238095238096e-05,
      "loss": 2.4084,
      "step": 253
    },
    {
      "epoch": 5.912280701754386,
      "grad_norm": 2.5141403675079346,
      "learning_rate": 8.790476190476191e-05,
      "loss": 2.408,
      "step": 254
    },
    {
      "epoch": 5.935672514619883,
      "grad_norm": 2.602572202682495,
      "learning_rate": 8.785714285714286e-05,
      "loss": 2.3673,
      "step": 255
    },
    {
      "epoch": 5.95906432748538,
      "grad_norm": 2.1632885932922363,
      "learning_rate": 8.78095238095238e-05,
      "loss": 2.3696,
      "step": 256
    },
    {
      "epoch": 5.982456140350877,
      "grad_norm": 4.0230302810668945,
      "learning_rate": 8.776190476190476e-05,
      "loss": 2.3337,
      "step": 257
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.374912977218628,
      "learning_rate": 8.771428571428572e-05,
      "loss": 1.7329,
      "step": 258
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.30803602933883667,
      "eval_runtime": 3.4815,
      "eval_samples_per_second": 62.042,
      "eval_steps_per_second": 15.511,
      "step": 258
    },
    {
      "epoch": 6.023391812865497,
      "grad_norm": 3.084089517593384,
      "learning_rate": 8.766666666666668e-05,
      "loss": 2.3514,
      "step": 259
    },
    {
      "epoch": 6.046783625730994,
      "grad_norm": 3.1659111976623535,
      "learning_rate": 8.761904761904762e-05,
      "loss": 2.3374,
      "step": 260
    },
    {
      "epoch": 6.0701754385964914,
      "grad_norm": 2.125979423522949,
      "learning_rate": 8.757142857142857e-05,
      "loss": 2.3415,
      "step": 261
    },
    {
      "epoch": 6.093567251461988,
      "grad_norm": 3.6721763610839844,
      "learning_rate": 8.752380952380953e-05,
      "loss": 2.3566,
      "step": 262
    },
    {
      "epoch": 6.116959064327485,
      "grad_norm": 3.2312450408935547,
      "learning_rate": 8.747619047619047e-05,
      "loss": 2.299,
      "step": 263
    },
    {
      "epoch": 6.140350877192983,
      "grad_norm": 5.158683776855469,
      "learning_rate": 8.742857142857144e-05,
      "loss": 2.3623,
      "step": 264
    },
    {
      "epoch": 6.16374269005848,
      "grad_norm": 3.5738871097564697,
      "learning_rate": 8.738095238095239e-05,
      "loss": 2.3442,
      "step": 265
    },
    {
      "epoch": 6.187134502923977,
      "grad_norm": 3.4018423557281494,
      "learning_rate": 8.733333333333333e-05,
      "loss": 2.3432,
      "step": 266
    },
    {
      "epoch": 6.2105263157894735,
      "grad_norm": 3.736330509185791,
      "learning_rate": 8.728571428571429e-05,
      "loss": 2.3612,
      "step": 267
    },
    {
      "epoch": 6.23391812865497,
      "grad_norm": 2.5863962173461914,
      "learning_rate": 8.723809523809524e-05,
      "loss": 2.339,
      "step": 268
    },
    {
      "epoch": 6.257309941520468,
      "grad_norm": 3.0965137481689453,
      "learning_rate": 8.71904761904762e-05,
      "loss": 2.3152,
      "step": 269
    },
    {
      "epoch": 6.280701754385965,
      "grad_norm": 3.284605026245117,
      "learning_rate": 8.714285714285715e-05,
      "loss": 2.3969,
      "step": 270
    },
    {
      "epoch": 6.304093567251462,
      "grad_norm": 2.4946165084838867,
      "learning_rate": 8.70952380952381e-05,
      "loss": 2.2957,
      "step": 271
    },
    {
      "epoch": 6.3274853801169595,
      "grad_norm": 3.2509989738464355,
      "learning_rate": 8.704761904761906e-05,
      "loss": 2.3151,
      "step": 272
    },
    {
      "epoch": 6.350877192982456,
      "grad_norm": 3.7346384525299072,
      "learning_rate": 8.7e-05,
      "loss": 2.313,
      "step": 273
    },
    {
      "epoch": 6.374269005847953,
      "grad_norm": 2.4033422470092773,
      "learning_rate": 8.695238095238095e-05,
      "loss": 2.3143,
      "step": 274
    },
    {
      "epoch": 6.39766081871345,
      "grad_norm": 3.5691659450531006,
      "learning_rate": 8.690476190476192e-05,
      "loss": 2.347,
      "step": 275
    },
    {
      "epoch": 6.421052631578947,
      "grad_norm": 3.408168315887451,
      "learning_rate": 8.685714285714286e-05,
      "loss": 2.3762,
      "step": 276
    },
    {
      "epoch": 6.444444444444445,
      "grad_norm": 2.701221227645874,
      "learning_rate": 8.680952380952382e-05,
      "loss": 2.3781,
      "step": 277
    },
    {
      "epoch": 6.4678362573099415,
      "grad_norm": 3.2413136959075928,
      "learning_rate": 8.676190476190477e-05,
      "loss": 2.3683,
      "step": 278
    },
    {
      "epoch": 6.491228070175438,
      "grad_norm": 3.9076733589172363,
      "learning_rate": 8.671428571428571e-05,
      "loss": 2.3921,
      "step": 279
    },
    {
      "epoch": 6.514619883040936,
      "grad_norm": 2.388068199157715,
      "learning_rate": 8.666666666666667e-05,
      "loss": 2.3121,
      "step": 280
    },
    {
      "epoch": 6.538011695906433,
      "grad_norm": 2.6373026371002197,
      "learning_rate": 8.661904761904763e-05,
      "loss": 2.3386,
      "step": 281
    },
    {
      "epoch": 6.56140350877193,
      "grad_norm": 3.8755855560302734,
      "learning_rate": 8.657142857142858e-05,
      "loss": 2.392,
      "step": 282
    },
    {
      "epoch": 6.584795321637427,
      "grad_norm": 2.8325467109680176,
      "learning_rate": 8.652380952380953e-05,
      "loss": 2.3382,
      "step": 283
    },
    {
      "epoch": 6.6081871345029235,
      "grad_norm": 3.6729838848114014,
      "learning_rate": 8.647619047619047e-05,
      "loss": 2.373,
      "step": 284
    },
    {
      "epoch": 6.631578947368421,
      "grad_norm": 3.537522077560425,
      "learning_rate": 8.642857142857143e-05,
      "loss": 2.3239,
      "step": 285
    },
    {
      "epoch": 6.654970760233918,
      "grad_norm": 3.385708808898926,
      "learning_rate": 8.638095238095239e-05,
      "loss": 2.3445,
      "step": 286
    },
    {
      "epoch": 6.678362573099415,
      "grad_norm": 2.634990692138672,
      "learning_rate": 8.633333333333334e-05,
      "loss": 2.3395,
      "step": 287
    },
    {
      "epoch": 6.701754385964913,
      "grad_norm": 3.4988627433776855,
      "learning_rate": 8.62857142857143e-05,
      "loss": 2.3774,
      "step": 288
    },
    {
      "epoch": 6.7251461988304095,
      "grad_norm": 2.7816081047058105,
      "learning_rate": 8.623809523809524e-05,
      "loss": 2.3252,
      "step": 289
    },
    {
      "epoch": 6.748538011695906,
      "grad_norm": 2.6087560653686523,
      "learning_rate": 8.61904761904762e-05,
      "loss": 2.3386,
      "step": 290
    },
    {
      "epoch": 6.771929824561403,
      "grad_norm": 2.98740553855896,
      "learning_rate": 8.614285714285714e-05,
      "loss": 2.3114,
      "step": 291
    },
    {
      "epoch": 6.7953216374269,
      "grad_norm": 2.4232819080352783,
      "learning_rate": 8.60952380952381e-05,
      "loss": 2.2925,
      "step": 292
    },
    {
      "epoch": 6.818713450292398,
      "grad_norm": 3.8474879264831543,
      "learning_rate": 8.604761904761906e-05,
      "loss": 2.3508,
      "step": 293
    },
    {
      "epoch": 6.842105263157895,
      "grad_norm": 3.4887146949768066,
      "learning_rate": 8.6e-05,
      "loss": 2.3654,
      "step": 294
    },
    {
      "epoch": 6.8654970760233915,
      "grad_norm": 2.830552101135254,
      "learning_rate": 8.595238095238096e-05,
      "loss": 2.3511,
      "step": 295
    },
    {
      "epoch": 6.888888888888889,
      "grad_norm": 3.660485029220581,
      "learning_rate": 8.59047619047619e-05,
      "loss": 2.3656,
      "step": 296
    },
    {
      "epoch": 6.912280701754386,
      "grad_norm": 3.242731809616089,
      "learning_rate": 8.585714285714286e-05,
      "loss": 2.368,
      "step": 297
    },
    {
      "epoch": 6.935672514619883,
      "grad_norm": 2.875051975250244,
      "learning_rate": 8.580952380952382e-05,
      "loss": 2.3471,
      "step": 298
    },
    {
      "epoch": 6.95906432748538,
      "grad_norm": 2.302536725997925,
      "learning_rate": 8.576190476190477e-05,
      "loss": 2.3601,
      "step": 299
    },
    {
      "epoch": 6.982456140350877,
      "grad_norm": 4.208285808563232,
      "learning_rate": 8.571428571428571e-05,
      "loss": 2.3197,
      "step": 300
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.7053803205490112,
      "learning_rate": 8.566666666666667e-05,
      "loss": 1.762,
      "step": 301
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.31038928031921387,
      "eval_runtime": 3.469,
      "eval_samples_per_second": 62.266,
      "eval_steps_per_second": 15.567,
      "step": 301
    },
    {
      "epoch": 7.023391812865497,
      "grad_norm": 3.534348964691162,
      "learning_rate": 8.561904761904762e-05,
      "loss": 2.3135,
      "step": 302
    },
    {
      "epoch": 7.046783625730994,
      "grad_norm": 2.444737434387207,
      "learning_rate": 8.557142857142857e-05,
      "loss": 2.3049,
      "step": 303
    },
    {
      "epoch": 7.0701754385964914,
      "grad_norm": 2.388345241546631,
      "learning_rate": 8.552380952380953e-05,
      "loss": 2.2772,
      "step": 304
    },
    {
      "epoch": 7.093567251461988,
      "grad_norm": 3.2836976051330566,
      "learning_rate": 8.547619047619048e-05,
      "loss": 2.2813,
      "step": 305
    },
    {
      "epoch": 7.116959064327485,
      "grad_norm": 2.2374956607818604,
      "learning_rate": 8.542857142857144e-05,
      "loss": 2.257,
      "step": 306
    },
    {
      "epoch": 7.140350877192983,
      "grad_norm": 2.89320969581604,
      "learning_rate": 8.538095238095238e-05,
      "loss": 2.3366,
      "step": 307
    },
    {
      "epoch": 7.16374269005848,
      "grad_norm": 2.2934420108795166,
      "learning_rate": 8.533333333333334e-05,
      "loss": 2.2922,
      "step": 308
    },
    {
      "epoch": 7.187134502923977,
      "grad_norm": 2.66508150100708,
      "learning_rate": 8.52857142857143e-05,
      "loss": 2.2824,
      "step": 309
    },
    {
      "epoch": 7.2105263157894735,
      "grad_norm": 2.668550729751587,
      "learning_rate": 8.523809523809524e-05,
      "loss": 2.3064,
      "step": 310
    },
    {
      "epoch": 7.23391812865497,
      "grad_norm": 2.6760103702545166,
      "learning_rate": 8.51904761904762e-05,
      "loss": 2.2928,
      "step": 311
    },
    {
      "epoch": 7.257309941520468,
      "grad_norm": 2.7909204959869385,
      "learning_rate": 8.514285714285714e-05,
      "loss": 2.2599,
      "step": 312
    },
    {
      "epoch": 7.280701754385965,
      "grad_norm": 2.3334522247314453,
      "learning_rate": 8.509523809523809e-05,
      "loss": 2.2526,
      "step": 313
    },
    {
      "epoch": 7.304093567251462,
      "grad_norm": 3.7709286212921143,
      "learning_rate": 8.504761904761905e-05,
      "loss": 2.3113,
      "step": 314
    },
    {
      "epoch": 7.3274853801169595,
      "grad_norm": 2.955559730529785,
      "learning_rate": 8.5e-05,
      "loss": 2.3101,
      "step": 315
    },
    {
      "epoch": 7.350877192982456,
      "grad_norm": 2.5498976707458496,
      "learning_rate": 8.495238095238096e-05,
      "loss": 2.2774,
      "step": 316
    },
    {
      "epoch": 7.374269005847953,
      "grad_norm": 2.2549405097961426,
      "learning_rate": 8.490476190476191e-05,
      "loss": 2.3074,
      "step": 317
    },
    {
      "epoch": 7.39766081871345,
      "grad_norm": 3.085324764251709,
      "learning_rate": 8.485714285714285e-05,
      "loss": 2.2579,
      "step": 318
    },
    {
      "epoch": 7.421052631578947,
      "grad_norm": 3.924325942993164,
      "learning_rate": 8.480952380952381e-05,
      "loss": 2.2452,
      "step": 319
    },
    {
      "epoch": 7.444444444444445,
      "grad_norm": 2.55413818359375,
      "learning_rate": 8.476190476190477e-05,
      "loss": 2.3174,
      "step": 320
    },
    {
      "epoch": 7.4678362573099415,
      "grad_norm": 3.082507371902466,
      "learning_rate": 8.471428571428573e-05,
      "loss": 2.3215,
      "step": 321
    },
    {
      "epoch": 7.491228070175438,
      "grad_norm": 3.0043070316314697,
      "learning_rate": 8.466666666666667e-05,
      "loss": 2.2896,
      "step": 322
    },
    {
      "epoch": 7.514619883040936,
      "grad_norm": 2.590393304824829,
      "learning_rate": 8.461904761904762e-05,
      "loss": 2.3028,
      "step": 323
    },
    {
      "epoch": 7.538011695906433,
      "grad_norm": 2.350078821182251,
      "learning_rate": 8.457142857142858e-05,
      "loss": 2.2711,
      "step": 324
    },
    {
      "epoch": 7.56140350877193,
      "grad_norm": 2.553441047668457,
      "learning_rate": 8.452380952380952e-05,
      "loss": 2.3428,
      "step": 325
    },
    {
      "epoch": 7.584795321637427,
      "grad_norm": 2.492304801940918,
      "learning_rate": 8.447619047619048e-05,
      "loss": 2.3009,
      "step": 326
    },
    {
      "epoch": 7.6081871345029235,
      "grad_norm": 3.329228162765503,
      "learning_rate": 8.442857142857144e-05,
      "loss": 2.3234,
      "step": 327
    },
    {
      "epoch": 7.631578947368421,
      "grad_norm": 2.6081371307373047,
      "learning_rate": 8.438095238095238e-05,
      "loss": 2.3028,
      "step": 328
    },
    {
      "epoch": 7.654970760233918,
      "grad_norm": 2.6931183338165283,
      "learning_rate": 8.433333333333334e-05,
      "loss": 2.2992,
      "step": 329
    },
    {
      "epoch": 7.678362573099415,
      "grad_norm": 2.0675008296966553,
      "learning_rate": 8.428571428571429e-05,
      "loss": 2.3682,
      "step": 330
    },
    {
      "epoch": 7.701754385964913,
      "grad_norm": 2.3411147594451904,
      "learning_rate": 8.423809523809524e-05,
      "loss": 2.2903,
      "step": 331
    },
    {
      "epoch": 7.7251461988304095,
      "grad_norm": 3.0642454624176025,
      "learning_rate": 8.41904761904762e-05,
      "loss": 2.3214,
      "step": 332
    },
    {
      "epoch": 7.748538011695906,
      "grad_norm": 3.137528896331787,
      "learning_rate": 8.414285714285715e-05,
      "loss": 2.3448,
      "step": 333
    },
    {
      "epoch": 7.771929824561403,
      "grad_norm": 1.907626748085022,
      "learning_rate": 8.40952380952381e-05,
      "loss": 2.3362,
      "step": 334
    },
    {
      "epoch": 7.7953216374269,
      "grad_norm": 2.599119186401367,
      "learning_rate": 8.404761904761905e-05,
      "loss": 2.2995,
      "step": 335
    },
    {
      "epoch": 7.818713450292398,
      "grad_norm": 2.1835620403289795,
      "learning_rate": 8.4e-05,
      "loss": 2.3506,
      "step": 336
    },
    {
      "epoch": 7.842105263157895,
      "grad_norm": 2.6803245544433594,
      "learning_rate": 8.395238095238095e-05,
      "loss": 2.2936,
      "step": 337
    },
    {
      "epoch": 7.8654970760233915,
      "grad_norm": 2.978414297103882,
      "learning_rate": 8.390476190476191e-05,
      "loss": 2.3327,
      "step": 338
    },
    {
      "epoch": 7.888888888888889,
      "grad_norm": 3.149514675140381,
      "learning_rate": 8.385714285714286e-05,
      "loss": 2.2973,
      "step": 339
    },
    {
      "epoch": 7.912280701754386,
      "grad_norm": 2.3366856575012207,
      "learning_rate": 8.380952380952382e-05,
      "loss": 2.3136,
      "step": 340
    },
    {
      "epoch": 7.935672514619883,
      "grad_norm": 2.9807534217834473,
      "learning_rate": 8.376190476190476e-05,
      "loss": 2.2854,
      "step": 341
    },
    {
      "epoch": 7.95906432748538,
      "grad_norm": 2.4020419120788574,
      "learning_rate": 8.371428571428572e-05,
      "loss": 2.3807,
      "step": 342
    },
    {
      "epoch": 7.982456140350877,
      "grad_norm": 3.0339105129241943,
      "learning_rate": 8.366666666666668e-05,
      "loss": 2.3336,
      "step": 343
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.9973078966140747,
      "learning_rate": 8.361904761904762e-05,
      "loss": 1.7577,
      "step": 344
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.3090885281562805,
      "eval_runtime": 3.4516,
      "eval_samples_per_second": 62.58,
      "eval_steps_per_second": 15.645,
      "step": 344
    },
    {
      "epoch": 8.023391812865498,
      "grad_norm": 2.9275968074798584,
      "learning_rate": 8.357142857142858e-05,
      "loss": 2.278,
      "step": 345
    },
    {
      "epoch": 8.046783625730994,
      "grad_norm": 2.7276556491851807,
      "learning_rate": 8.352380952380952e-05,
      "loss": 2.2362,
      "step": 346
    },
    {
      "epoch": 8.070175438596491,
      "grad_norm": 4.678555011749268,
      "learning_rate": 8.347619047619048e-05,
      "loss": 2.2201,
      "step": 347
    },
    {
      "epoch": 8.093567251461987,
      "grad_norm": 4.674102306365967,
      "learning_rate": 8.342857142857143e-05,
      "loss": 2.3193,
      "step": 348
    },
    {
      "epoch": 8.116959064327485,
      "grad_norm": 3.837069272994995,
      "learning_rate": 8.338095238095239e-05,
      "loss": 2.2565,
      "step": 349
    },
    {
      "epoch": 8.140350877192983,
      "grad_norm": 3.5084426403045654,
      "learning_rate": 8.333333333333334e-05,
      "loss": 2.2948,
      "step": 350
    },
    {
      "epoch": 8.163742690058479,
      "grad_norm": 2.2970736026763916,
      "learning_rate": 8.328571428571429e-05,
      "loss": 2.2341,
      "step": 351
    },
    {
      "epoch": 8.187134502923977,
      "grad_norm": 3.0149991512298584,
      "learning_rate": 8.323809523809523e-05,
      "loss": 2.2382,
      "step": 352
    },
    {
      "epoch": 8.210526315789474,
      "grad_norm": 3.103729724884033,
      "learning_rate": 8.319047619047619e-05,
      "loss": 2.2195,
      "step": 353
    },
    {
      "epoch": 8.23391812865497,
      "grad_norm": 3.5624403953552246,
      "learning_rate": 8.314285714285715e-05,
      "loss": 2.2907,
      "step": 354
    },
    {
      "epoch": 8.257309941520468,
      "grad_norm": 3.337334156036377,
      "learning_rate": 8.309523809523811e-05,
      "loss": 2.2537,
      "step": 355
    },
    {
      "epoch": 8.280701754385966,
      "grad_norm": 2.8476223945617676,
      "learning_rate": 8.304761904761905e-05,
      "loss": 2.2899,
      "step": 356
    },
    {
      "epoch": 8.304093567251462,
      "grad_norm": 3.415135622024536,
      "learning_rate": 8.3e-05,
      "loss": 2.2165,
      "step": 357
    },
    {
      "epoch": 8.32748538011696,
      "grad_norm": 3.373770236968994,
      "learning_rate": 8.295238095238096e-05,
      "loss": 2.2417,
      "step": 358
    },
    {
      "epoch": 8.350877192982455,
      "grad_norm": 2.926384925842285,
      "learning_rate": 8.29047619047619e-05,
      "loss": 2.2356,
      "step": 359
    },
    {
      "epoch": 8.374269005847953,
      "grad_norm": 3.749796152114868,
      "learning_rate": 8.285714285714287e-05,
      "loss": 2.2539,
      "step": 360
    },
    {
      "epoch": 8.397660818713451,
      "grad_norm": 2.3710129261016846,
      "learning_rate": 8.280952380952382e-05,
      "loss": 2.2066,
      "step": 361
    },
    {
      "epoch": 8.421052631578947,
      "grad_norm": 2.417635440826416,
      "learning_rate": 8.276190476190476e-05,
      "loss": 2.2212,
      "step": 362
    },
    {
      "epoch": 8.444444444444445,
      "grad_norm": 2.252788543701172,
      "learning_rate": 8.271428571428572e-05,
      "loss": 2.202,
      "step": 363
    },
    {
      "epoch": 8.46783625730994,
      "grad_norm": 2.487130641937256,
      "learning_rate": 8.266666666666667e-05,
      "loss": 2.2888,
      "step": 364
    },
    {
      "epoch": 8.491228070175438,
      "grad_norm": 3.3258721828460693,
      "learning_rate": 8.261904761904762e-05,
      "loss": 2.2453,
      "step": 365
    },
    {
      "epoch": 8.514619883040936,
      "grad_norm": 3.4949452877044678,
      "learning_rate": 8.257142857142858e-05,
      "loss": 2.2536,
      "step": 366
    },
    {
      "epoch": 8.538011695906432,
      "grad_norm": 2.923041343688965,
      "learning_rate": 8.252380952380953e-05,
      "loss": 2.2622,
      "step": 367
    },
    {
      "epoch": 8.56140350877193,
      "grad_norm": 4.235454559326172,
      "learning_rate": 8.247619047619049e-05,
      "loss": 2.2604,
      "step": 368
    },
    {
      "epoch": 8.584795321637428,
      "grad_norm": 3.192513942718506,
      "learning_rate": 8.242857142857143e-05,
      "loss": 2.26,
      "step": 369
    },
    {
      "epoch": 8.608187134502923,
      "grad_norm": 3.3745954036712646,
      "learning_rate": 8.238095238095238e-05,
      "loss": 2.3249,
      "step": 370
    },
    {
      "epoch": 8.631578947368421,
      "grad_norm": 2.696607828140259,
      "learning_rate": 8.233333333333333e-05,
      "loss": 2.2771,
      "step": 371
    },
    {
      "epoch": 8.654970760233919,
      "grad_norm": 3.3591575622558594,
      "learning_rate": 8.228571428571429e-05,
      "loss": 2.2722,
      "step": 372
    },
    {
      "epoch": 8.678362573099415,
      "grad_norm": 2.7265303134918213,
      "learning_rate": 8.223809523809525e-05,
      "loss": 2.3042,
      "step": 373
    },
    {
      "epoch": 8.701754385964913,
      "grad_norm": 3.1058547496795654,
      "learning_rate": 8.21904761904762e-05,
      "loss": 2.2676,
      "step": 374
    },
    {
      "epoch": 8.725146198830409,
      "grad_norm": 3.8888497352600098,
      "learning_rate": 8.214285714285714e-05,
      "loss": 2.2778,
      "step": 375
    },
    {
      "epoch": 8.748538011695906,
      "grad_norm": 2.25227427482605,
      "learning_rate": 8.20952380952381e-05,
      "loss": 2.2814,
      "step": 376
    },
    {
      "epoch": 8.771929824561404,
      "grad_norm": 4.398237228393555,
      "learning_rate": 8.204761904761906e-05,
      "loss": 2.317,
      "step": 377
    },
    {
      "epoch": 8.7953216374269,
      "grad_norm": 3.410182237625122,
      "learning_rate": 8.2e-05,
      "loss": 2.3045,
      "step": 378
    },
    {
      "epoch": 8.818713450292398,
      "grad_norm": 2.4230539798736572,
      "learning_rate": 8.195238095238096e-05,
      "loss": 2.3019,
      "step": 379
    },
    {
      "epoch": 8.842105263157894,
      "grad_norm": 3.115366220474243,
      "learning_rate": 8.19047619047619e-05,
      "loss": 2.2107,
      "step": 380
    },
    {
      "epoch": 8.865497076023392,
      "grad_norm": 4.855501174926758,
      "learning_rate": 8.185714285714286e-05,
      "loss": 2.2712,
      "step": 381
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 2.77266001701355,
      "learning_rate": 8.180952380952381e-05,
      "loss": 2.2569,
      "step": 382
    },
    {
      "epoch": 8.912280701754385,
      "grad_norm": 2.3498106002807617,
      "learning_rate": 8.176190476190477e-05,
      "loss": 2.2951,
      "step": 383
    },
    {
      "epoch": 8.935672514619883,
      "grad_norm": 3.1771063804626465,
      "learning_rate": 8.171428571428572e-05,
      "loss": 2.2536,
      "step": 384
    },
    {
      "epoch": 8.95906432748538,
      "grad_norm": 2.889338970184326,
      "learning_rate": 8.166666666666667e-05,
      "loss": 2.2749,
      "step": 385
    },
    {
      "epoch": 8.982456140350877,
      "grad_norm": 5.354519367218018,
      "learning_rate": 8.161904761904763e-05,
      "loss": 2.322,
      "step": 386
    },
    {
      "epoch": 9.0,
      "grad_norm": 3.0712740421295166,
      "learning_rate": 8.157142857142857e-05,
      "loss": 1.7048,
      "step": 387
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.3110826313495636,
      "eval_runtime": 3.4732,
      "eval_samples_per_second": 62.191,
      "eval_steps_per_second": 15.548,
      "step": 387
    },
    {
      "epoch": 9.023391812865498,
      "grad_norm": 4.209856033325195,
      "learning_rate": 8.152380952380953e-05,
      "loss": 2.2796,
      "step": 388
    },
    {
      "epoch": 9.046783625730994,
      "grad_norm": 4.553707122802734,
      "learning_rate": 8.147619047619049e-05,
      "loss": 2.1917,
      "step": 389
    },
    {
      "epoch": 9.070175438596491,
      "grad_norm": 2.235180377960205,
      "learning_rate": 8.142857142857143e-05,
      "loss": 2.2586,
      "step": 390
    },
    {
      "epoch": 9.093567251461987,
      "grad_norm": 3.6155264377593994,
      "learning_rate": 8.138095238095238e-05,
      "loss": 2.2361,
      "step": 391
    },
    {
      "epoch": 9.116959064327485,
      "grad_norm": 3.4730417728424072,
      "learning_rate": 8.133333333333334e-05,
      "loss": 2.2257,
      "step": 392
    },
    {
      "epoch": 9.140350877192983,
      "grad_norm": 3.765535831451416,
      "learning_rate": 8.128571428571428e-05,
      "loss": 2.1642,
      "step": 393
    },
    {
      "epoch": 9.163742690058479,
      "grad_norm": 3.1897642612457275,
      "learning_rate": 8.123809523809525e-05,
      "loss": 2.2229,
      "step": 394
    },
    {
      "epoch": 9.187134502923977,
      "grad_norm": 3.8044841289520264,
      "learning_rate": 8.11904761904762e-05,
      "loss": 2.1882,
      "step": 395
    },
    {
      "epoch": 9.210526315789474,
      "grad_norm": 3.561450719833374,
      "learning_rate": 8.114285714285714e-05,
      "loss": 2.1809,
      "step": 396
    },
    {
      "epoch": 9.23391812865497,
      "grad_norm": 3.2488150596618652,
      "learning_rate": 8.10952380952381e-05,
      "loss": 2.2111,
      "step": 397
    },
    {
      "epoch": 9.257309941520468,
      "grad_norm": 3.5442261695861816,
      "learning_rate": 8.104761904761905e-05,
      "loss": 2.1749,
      "step": 398
    },
    {
      "epoch": 9.280701754385966,
      "grad_norm": 2.66875958442688,
      "learning_rate": 8.1e-05,
      "loss": 2.1776,
      "step": 399
    },
    {
      "epoch": 9.304093567251462,
      "grad_norm": 2.672934055328369,
      "learning_rate": 8.095238095238096e-05,
      "loss": 2.1938,
      "step": 400
    },
    {
      "epoch": 9.32748538011696,
      "grad_norm": 2.8488197326660156,
      "learning_rate": 8.090476190476191e-05,
      "loss": 2.1875,
      "step": 401
    },
    {
      "epoch": 9.350877192982455,
      "grad_norm": 3.4410758018493652,
      "learning_rate": 8.085714285714287e-05,
      "loss": 2.2253,
      "step": 402
    },
    {
      "epoch": 9.374269005847953,
      "grad_norm": 3.20196270942688,
      "learning_rate": 8.080952380952381e-05,
      "loss": 2.2043,
      "step": 403
    },
    {
      "epoch": 9.397660818713451,
      "grad_norm": 3.485410213470459,
      "learning_rate": 8.076190476190475e-05,
      "loss": 2.2155,
      "step": 404
    },
    {
      "epoch": 9.421052631578947,
      "grad_norm": 3.2198355197906494,
      "learning_rate": 8.071428571428573e-05,
      "loss": 2.2339,
      "step": 405
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 4.449166774749756,
      "learning_rate": 8.066666666666667e-05,
      "loss": 2.2157,
      "step": 406
    },
    {
      "epoch": 9.46783625730994,
      "grad_norm": 3.3024957180023193,
      "learning_rate": 8.061904761904763e-05,
      "loss": 2.2373,
      "step": 407
    },
    {
      "epoch": 9.491228070175438,
      "grad_norm": 4.262597560882568,
      "learning_rate": 8.057142857142857e-05,
      "loss": 2.2118,
      "step": 408
    },
    {
      "epoch": 9.514619883040936,
      "grad_norm": 3.014378070831299,
      "learning_rate": 8.052380952380952e-05,
      "loss": 2.1513,
      "step": 409
    },
    {
      "epoch": 9.538011695906432,
      "grad_norm": 2.3644843101501465,
      "learning_rate": 8.047619047619048e-05,
      "loss": 2.2105,
      "step": 410
    },
    {
      "epoch": 9.56140350877193,
      "grad_norm": 3.573030471801758,
      "learning_rate": 8.042857142857144e-05,
      "loss": 2.2014,
      "step": 411
    },
    {
      "epoch": 9.584795321637428,
      "grad_norm": 3.49285626411438,
      "learning_rate": 8.03809523809524e-05,
      "loss": 2.2258,
      "step": 412
    },
    {
      "epoch": 9.608187134502923,
      "grad_norm": 2.701261281967163,
      "learning_rate": 8.033333333333334e-05,
      "loss": 2.1926,
      "step": 413
    },
    {
      "epoch": 9.631578947368421,
      "grad_norm": 3.1829402446746826,
      "learning_rate": 8.028571428571428e-05,
      "loss": 2.1833,
      "step": 414
    },
    {
      "epoch": 9.654970760233919,
      "grad_norm": 3.5617990493774414,
      "learning_rate": 8.023809523809524e-05,
      "loss": 2.2629,
      "step": 415
    },
    {
      "epoch": 9.678362573099415,
      "grad_norm": 3.1133735179901123,
      "learning_rate": 8.01904761904762e-05,
      "loss": 2.2344,
      "step": 416
    },
    {
      "epoch": 9.701754385964913,
      "grad_norm": 3.0228543281555176,
      "learning_rate": 8.014285714285715e-05,
      "loss": 2.2076,
      "step": 417
    },
    {
      "epoch": 9.725146198830409,
      "grad_norm": 4.193742752075195,
      "learning_rate": 8.00952380952381e-05,
      "loss": 2.2325,
      "step": 418
    },
    {
      "epoch": 9.748538011695906,
      "grad_norm": 4.473887920379639,
      "learning_rate": 8.004761904761905e-05,
      "loss": 2.2241,
      "step": 419
    },
    {
      "epoch": 9.771929824561404,
      "grad_norm": 3.292799234390259,
      "learning_rate": 8e-05,
      "loss": 2.2584,
      "step": 420
    },
    {
      "epoch": 9.7953216374269,
      "grad_norm": 6.8649516105651855,
      "learning_rate": 7.995238095238095e-05,
      "loss": 2.2649,
      "step": 421
    },
    {
      "epoch": 9.818713450292398,
      "grad_norm": 3.9372193813323975,
      "learning_rate": 7.990476190476191e-05,
      "loss": 2.3003,
      "step": 422
    },
    {
      "epoch": 9.842105263157894,
      "grad_norm": 3.112377405166626,
      "learning_rate": 7.985714285714287e-05,
      "loss": 2.2106,
      "step": 423
    },
    {
      "epoch": 9.865497076023392,
      "grad_norm": 2.902355909347534,
      "learning_rate": 7.980952380952381e-05,
      "loss": 2.2946,
      "step": 424
    },
    {
      "epoch": 9.88888888888889,
      "grad_norm": 2.473977565765381,
      "learning_rate": 7.976190476190477e-05,
      "loss": 2.2175,
      "step": 425
    },
    {
      "epoch": 9.912280701754385,
      "grad_norm": 4.093216419219971,
      "learning_rate": 7.971428571428572e-05,
      "loss": 2.2092,
      "step": 426
    },
    {
      "epoch": 9.935672514619883,
      "grad_norm": 2.5776782035827637,
      "learning_rate": 7.966666666666666e-05,
      "loss": 2.219,
      "step": 427
    },
    {
      "epoch": 9.95906432748538,
      "grad_norm": 3.246060371398926,
      "learning_rate": 7.961904761904763e-05,
      "loss": 2.239,
      "step": 428
    },
    {
      "epoch": 9.982456140350877,
      "grad_norm": 2.8515846729278564,
      "learning_rate": 7.957142857142858e-05,
      "loss": 2.2154,
      "step": 429
    },
    {
      "epoch": 10.0,
      "grad_norm": 2.685945510864258,
      "learning_rate": 7.952380952380952e-05,
      "loss": 1.6512,
      "step": 430
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.3121136426925659,
      "eval_runtime": 3.5135,
      "eval_samples_per_second": 61.478,
      "eval_steps_per_second": 15.369,
      "step": 430
    }
  ],
  "logging_steps": 1,
  "max_steps": 2100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.151327228493824e+16,
  "train_batch_size": 12,
  "trial_name": null,
  "trial_params": null
}