{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.8315018315018317,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003663003663003663,
      "grad_norm": 33.24192428588867,
      "learning_rate": 0.0,
      "loss": 0.9555,
      "step": 1
    },
    {
      "epoch": 0.007326007326007326,
      "grad_norm": 23.005327224731445,
      "learning_rate": 2.1978021978021978e-07,
      "loss": 0.7557,
      "step": 2
    },
    {
      "epoch": 0.01098901098901099,
      "grad_norm": 12.516372680664062,
      "learning_rate": 4.3956043956043957e-07,
      "loss": 0.2322,
      "step": 3
    },
    {
      "epoch": 0.014652014652014652,
      "grad_norm": 22.350322723388672,
      "learning_rate": 6.593406593406594e-07,
      "loss": 0.5263,
      "step": 4
    },
    {
      "epoch": 0.018315018315018316,
      "grad_norm": 37.14425277709961,
      "learning_rate": 8.791208791208791e-07,
      "loss": 0.547,
      "step": 5
    },
    {
      "epoch": 0.02197802197802198,
      "grad_norm": 27.73367691040039,
      "learning_rate": 1.098901098901099e-06,
      "loss": 0.5922,
      "step": 6
    },
    {
      "epoch": 0.02564102564102564,
      "grad_norm": 28.463964462280273,
      "learning_rate": 1.3186813186813187e-06,
      "loss": 1.0195,
      "step": 7
    },
    {
      "epoch": 0.029304029304029304,
      "grad_norm": 12.688858032226562,
      "learning_rate": 1.5384615384615385e-06,
      "loss": 0.1519,
      "step": 8
    },
    {
      "epoch": 0.03296703296703297,
      "grad_norm": 24.222930908203125,
      "learning_rate": 1.7582417582417583e-06,
      "loss": 0.8008,
      "step": 9
    },
    {
      "epoch": 0.03663003663003663,
      "grad_norm": 22.45709800720215,
      "learning_rate": 1.9780219780219782e-06,
      "loss": 1.1024,
      "step": 10
    },
    {
      "epoch": 0.040293040293040296,
      "grad_norm": 23.01483917236328,
      "learning_rate": 2.197802197802198e-06,
      "loss": 0.3072,
      "step": 11
    },
    {
      "epoch": 0.04395604395604396,
      "grad_norm": 24.276216506958008,
      "learning_rate": 2.4175824175824177e-06,
      "loss": 0.8937,
      "step": 12
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 24.501638412475586,
      "learning_rate": 2.6373626373626375e-06,
      "loss": 0.3748,
      "step": 13
    },
    {
      "epoch": 0.05128205128205128,
      "grad_norm": 11.965837478637695,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.2221,
      "step": 14
    },
    {
      "epoch": 0.054945054945054944,
      "grad_norm": 8.884313583374023,
      "learning_rate": 3.076923076923077e-06,
      "loss": 0.1682,
      "step": 15
    },
    {
      "epoch": 0.05860805860805861,
      "grad_norm": 13.486218452453613,
      "learning_rate": 3.2967032967032968e-06,
      "loss": 0.3324,
      "step": 16
    },
    {
      "epoch": 0.06227106227106227,
      "grad_norm": 29.47451400756836,
      "learning_rate": 3.5164835164835165e-06,
      "loss": 0.9247,
      "step": 17
    },
    {
      "epoch": 0.06593406593406594,
      "grad_norm": 38.8739128112793,
      "learning_rate": 3.7362637362637363e-06,
      "loss": 1.3591,
      "step": 18
    },
    {
      "epoch": 0.0695970695970696,
      "grad_norm": 24.181066513061523,
      "learning_rate": 3.9560439560439565e-06,
      "loss": 0.4257,
      "step": 19
    },
    {
      "epoch": 0.07326007326007326,
      "grad_norm": 18.25806427001953,
      "learning_rate": 4.175824175824176e-06,
      "loss": 0.3534,
      "step": 20
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 4.121458053588867,
      "learning_rate": 4.395604395604396e-06,
      "loss": 0.0459,
      "step": 21
    },
    {
      "epoch": 0.08058608058608059,
      "grad_norm": 17.89643096923828,
      "learning_rate": 4.615384615384616e-06,
      "loss": 0.3707,
      "step": 22
    },
    {
      "epoch": 0.08424908424908426,
      "grad_norm": 43.25539016723633,
      "learning_rate": 4.8351648351648355e-06,
      "loss": 1.139,
      "step": 23
    },
    {
      "epoch": 0.08791208791208792,
      "grad_norm": 19.56612205505371,
      "learning_rate": 5.054945054945056e-06,
      "loss": 0.3819,
      "step": 24
    },
    {
      "epoch": 0.09157509157509157,
      "grad_norm": 18.20578956604004,
      "learning_rate": 5.274725274725275e-06,
      "loss": 0.516,
      "step": 25
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 23.16927146911621,
      "learning_rate": 5.494505494505494e-06,
      "loss": 0.7161,
      "step": 26
    },
    {
      "epoch": 0.0989010989010989,
      "grad_norm": 10.449734687805176,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.3049,
      "step": 27
    },
    {
      "epoch": 0.10256410256410256,
      "grad_norm": 33.13974380493164,
      "learning_rate": 5.934065934065934e-06,
      "loss": 1.0178,
      "step": 28
    },
    {
      "epoch": 0.10622710622710622,
      "grad_norm": 34.373470306396484,
      "learning_rate": 6.153846153846154e-06,
      "loss": 1.0162,
      "step": 29
    },
    {
      "epoch": 0.10989010989010989,
      "grad_norm": 22.710988998413086,
      "learning_rate": 6.373626373626373e-06,
      "loss": 0.5866,
      "step": 30
    },
    {
      "epoch": 0.11355311355311355,
      "grad_norm": 23.314502716064453,
      "learning_rate": 6.5934065934065935e-06,
      "loss": 0.6159,
      "step": 31
    },
    {
      "epoch": 0.11721611721611722,
      "grad_norm": 23.481319427490234,
      "learning_rate": 6.813186813186814e-06,
      "loss": 0.5441,
      "step": 32
    },
    {
      "epoch": 0.12087912087912088,
      "grad_norm": 35.16271209716797,
      "learning_rate": 7.032967032967033e-06,
      "loss": 0.9091,
      "step": 33
    },
    {
      "epoch": 0.12454212454212454,
      "grad_norm": 32.2298698425293,
      "learning_rate": 7.252747252747253e-06,
      "loss": 0.5156,
      "step": 34
    },
    {
      "epoch": 0.1282051282051282,
      "grad_norm": 36.708953857421875,
      "learning_rate": 7.4725274725274726e-06,
      "loss": 1.5839,
      "step": 35
    },
    {
      "epoch": 0.13186813186813187,
      "grad_norm": 34.64887619018555,
      "learning_rate": 7.692307692307692e-06,
      "loss": 1.2861,
      "step": 36
    },
    {
      "epoch": 0.13553113553113552,
      "grad_norm": 20.94220733642578,
      "learning_rate": 7.912087912087913e-06,
      "loss": 0.5027,
      "step": 37
    },
    {
      "epoch": 0.1391941391941392,
      "grad_norm": 30.93832015991211,
      "learning_rate": 8.131868131868132e-06,
      "loss": 0.3584,
      "step": 38
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 19.195362091064453,
      "learning_rate": 8.351648351648352e-06,
      "loss": 0.6912,
      "step": 39
    },
    {
      "epoch": 0.14652014652014653,
      "grad_norm": 21.054162979125977,
      "learning_rate": 8.571428571428571e-06,
      "loss": 0.8027,
      "step": 40
    },
    {
      "epoch": 0.15018315018315018,
      "grad_norm": 16.64535903930664,
      "learning_rate": 8.791208791208792e-06,
      "loss": 0.3004,
      "step": 41
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 12.1064453125,
      "learning_rate": 9.010989010989011e-06,
      "loss": 0.2158,
      "step": 42
    },
    {
      "epoch": 0.1575091575091575,
      "grad_norm": 16.20220947265625,
      "learning_rate": 9.230769230769232e-06,
      "loss": 0.4137,
      "step": 43
    },
    {
      "epoch": 0.16117216117216118,
      "grad_norm": 25.698654174804688,
      "learning_rate": 9.45054945054945e-06,
      "loss": 0.7716,
      "step": 44
    },
    {
      "epoch": 0.16483516483516483,
      "grad_norm": 7.480422019958496,
      "learning_rate": 9.670329670329671e-06,
      "loss": 0.1046,
      "step": 45
    },
    {
      "epoch": 0.1684981684981685,
      "grad_norm": 38.25539016723633,
      "learning_rate": 9.89010989010989e-06,
      "loss": 1.3913,
      "step": 46
    },
    {
      "epoch": 0.17216117216117216,
      "grad_norm": 24.113954544067383,
      "learning_rate": 1.0109890109890111e-05,
      "loss": 0.4632,
      "step": 47
    },
    {
      "epoch": 0.17582417582417584,
      "grad_norm": 22.136140823364258,
      "learning_rate": 1.032967032967033e-05,
      "loss": 0.6634,
      "step": 48
    },
    {
      "epoch": 0.1794871794871795,
      "grad_norm": 19.417444229125977,
      "learning_rate": 1.054945054945055e-05,
      "loss": 0.3991,
      "step": 49
    },
    {
      "epoch": 0.18315018315018314,
      "grad_norm": 13.265430450439453,
      "learning_rate": 1.076923076923077e-05,
      "loss": 0.2613,
      "step": 50
    },
    {
      "epoch": 0.18681318681318682,
      "grad_norm": 25.118703842163086,
      "learning_rate": 1.0989010989010989e-05,
      "loss": 0.9231,
      "step": 51
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 34.06997299194336,
      "learning_rate": 1.120879120879121e-05,
      "loss": 1.5809,
      "step": 52
    },
    {
      "epoch": 0.19413919413919414,
      "grad_norm": 40.32486343383789,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 1.4601,
      "step": 53
    },
    {
      "epoch": 0.1978021978021978,
      "grad_norm": 18.847017288208008,
      "learning_rate": 1.1648351648351648e-05,
      "loss": 0.2345,
      "step": 54
    },
    {
      "epoch": 0.20146520146520147,
      "grad_norm": 37.98270034790039,
      "learning_rate": 1.1868131868131868e-05,
      "loss": 0.9792,
      "step": 55
    },
    {
      "epoch": 0.20512820512820512,
      "grad_norm": 35.72782897949219,
      "learning_rate": 1.2087912087912089e-05,
      "loss": 1.1561,
      "step": 56
    },
    {
      "epoch": 0.2087912087912088,
      "grad_norm": 18.577186584472656,
      "learning_rate": 1.2307692307692308e-05,
      "loss": 0.5577,
      "step": 57
    },
    {
      "epoch": 0.21245421245421245,
      "grad_norm": 23.086456298828125,
      "learning_rate": 1.2527472527472529e-05,
      "loss": 0.5807,
      "step": 58
    },
    {
      "epoch": 0.21611721611721613,
      "grad_norm": 20.053525924682617,
      "learning_rate": 1.2747252747252747e-05,
      "loss": 0.7024,
      "step": 59
    },
    {
      "epoch": 0.21978021978021978,
      "grad_norm": 22.25934410095215,
      "learning_rate": 1.2967032967032968e-05,
      "loss": 1.1033,
      "step": 60
    },
    {
      "epoch": 0.22344322344322345,
      "grad_norm": 17.981454849243164,
      "learning_rate": 1.3186813186813187e-05,
      "loss": 0.2774,
      "step": 61
    },
    {
      "epoch": 0.2271062271062271,
      "grad_norm": 11.286524772644043,
      "learning_rate": 1.3406593406593408e-05,
      "loss": 0.1802,
      "step": 62
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 25.822996139526367,
      "learning_rate": 1.3626373626373627e-05,
      "loss": 0.651,
      "step": 63
    },
    {
      "epoch": 0.23443223443223443,
      "grad_norm": 16.457286834716797,
      "learning_rate": 1.3846153846153847e-05,
      "loss": 0.2946,
      "step": 64
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 26.712799072265625,
      "learning_rate": 1.4065934065934066e-05,
      "loss": 0.7763,
      "step": 65
    },
    {
      "epoch": 0.24175824175824176,
      "grad_norm": 21.4671630859375,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.4132,
      "step": 66
    },
    {
      "epoch": 0.2454212454212454,
      "grad_norm": 21.834922790527344,
      "learning_rate": 1.4505494505494506e-05,
      "loss": 0.6544,
      "step": 67
    },
    {
      "epoch": 0.2490842490842491,
      "grad_norm": 15.396453857421875,
      "learning_rate": 1.4725274725274726e-05,
      "loss": 0.2426,
      "step": 68
    },
    {
      "epoch": 0.25274725274725274,
      "grad_norm": 8.851480484008789,
      "learning_rate": 1.4945054945054945e-05,
      "loss": 0.125,
      "step": 69
    },
    {
      "epoch": 0.2564102564102564,
      "grad_norm": 22.21581268310547,
      "learning_rate": 1.5164835164835164e-05,
      "loss": 0.2585,
      "step": 70
    },
    {
      "epoch": 0.2600732600732601,
      "grad_norm": 23.589736938476562,
      "learning_rate": 1.5384615384615384e-05,
      "loss": 0.386,
      "step": 71
    },
    {
      "epoch": 0.26373626373626374,
      "grad_norm": 51.82280731201172,
      "learning_rate": 1.5604395604395605e-05,
      "loss": 1.1802,
      "step": 72
    },
    {
      "epoch": 0.2673992673992674,
      "grad_norm": 36.43033981323242,
      "learning_rate": 1.5824175824175826e-05,
      "loss": 0.5574,
      "step": 73
    },
    {
      "epoch": 0.27106227106227104,
      "grad_norm": 46.151885986328125,
      "learning_rate": 1.6043956043956043e-05,
      "loss": 0.9113,
      "step": 74
    },
    {
      "epoch": 0.27472527472527475,
      "grad_norm": 34.090213775634766,
      "learning_rate": 1.6263736263736265e-05,
      "loss": 1.2161,
      "step": 75
    },
    {
      "epoch": 0.2783882783882784,
      "grad_norm": 15.469125747680664,
      "learning_rate": 1.6483516483516486e-05,
      "loss": 0.1833,
      "step": 76
    },
    {
      "epoch": 0.28205128205128205,
      "grad_norm": 26.77261734008789,
      "learning_rate": 1.6703296703296703e-05,
      "loss": 0.4095,
      "step": 77
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 8.46114444732666,
      "learning_rate": 1.6923076923076924e-05,
      "loss": 0.0724,
      "step": 78
    },
    {
      "epoch": 0.2893772893772894,
      "grad_norm": 7.954617500305176,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 0.057,
      "step": 79
    },
    {
      "epoch": 0.29304029304029305,
      "grad_norm": 32.47618103027344,
      "learning_rate": 1.7362637362637366e-05,
      "loss": 0.8099,
      "step": 80
    },
    {
      "epoch": 0.2967032967032967,
      "grad_norm": 34.506927490234375,
      "learning_rate": 1.7582417582417584e-05,
      "loss": 0.5867,
      "step": 81
    },
    {
      "epoch": 0.30036630036630035,
      "grad_norm": 18.276355743408203,
      "learning_rate": 1.78021978021978e-05,
      "loss": 0.4387,
      "step": 82
    },
    {
      "epoch": 0.304029304029304,
      "grad_norm": 35.61729431152344,
      "learning_rate": 1.8021978021978023e-05,
      "loss": 0.9711,
      "step": 83
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 14.001388549804688,
      "learning_rate": 1.824175824175824e-05,
      "loss": 0.1431,
      "step": 84
    },
    {
      "epoch": 0.31135531135531136,
      "grad_norm": 27.521188735961914,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.3686,
      "step": 85
    },
    {
      "epoch": 0.315018315018315,
      "grad_norm": 38.0133171081543,
      "learning_rate": 1.8681318681318682e-05,
      "loss": 1.3866,
      "step": 86
    },
    {
      "epoch": 0.31868131868131866,
      "grad_norm": 30.895553588867188,
      "learning_rate": 1.89010989010989e-05,
      "loss": 0.6676,
      "step": 87
    },
    {
      "epoch": 0.32234432234432236,
      "grad_norm": 26.165082931518555,
      "learning_rate": 1.912087912087912e-05,
      "loss": 0.4763,
      "step": 88
    },
    {
      "epoch": 0.326007326007326,
      "grad_norm": 25.6451473236084,
      "learning_rate": 1.9340659340659342e-05,
      "loss": 0.6921,
      "step": 89
    },
    {
      "epoch": 0.32967032967032966,
      "grad_norm": 31.52683448791504,
      "learning_rate": 1.9560439560439563e-05,
      "loss": 0.8449,
      "step": 90
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 27.559072494506836,
      "learning_rate": 1.978021978021978e-05,
      "loss": 0.9726,
      "step": 91
    },
    {
      "epoch": 0.336996336996337,
      "grad_norm": 38.23103713989258,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 0.2568,
      "step": 92
    },
    {
      "epoch": 0.34065934065934067,
      "grad_norm": 28.575313568115234,
      "learning_rate": 2.0219780219780223e-05,
      "loss": 0.7039,
      "step": 93
    },
    {
      "epoch": 0.3443223443223443,
      "grad_norm": 31.54847526550293,
      "learning_rate": 2.043956043956044e-05,
      "loss": 0.835,
      "step": 94
    },
    {
      "epoch": 0.34798534798534797,
      "grad_norm": 34.27505111694336,
      "learning_rate": 2.065934065934066e-05,
      "loss": 1.0304,
      "step": 95
    },
    {
      "epoch": 0.3516483516483517,
      "grad_norm": 23.972553253173828,
      "learning_rate": 2.087912087912088e-05,
      "loss": 0.775,
      "step": 96
    },
    {
      "epoch": 0.3553113553113553,
      "grad_norm": 18.46526527404785,
      "learning_rate": 2.10989010989011e-05,
      "loss": 0.2856,
      "step": 97
    },
    {
      "epoch": 0.358974358974359,
      "grad_norm": 22.087251663208008,
      "learning_rate": 2.131868131868132e-05,
      "loss": 0.6849,
      "step": 98
    },
    {
      "epoch": 0.3626373626373626,
      "grad_norm": 13.144533157348633,
      "learning_rate": 2.153846153846154e-05,
      "loss": 0.2766,
      "step": 99
    },
    {
      "epoch": 0.3663003663003663,
      "grad_norm": 14.740280151367188,
      "learning_rate": 2.175824175824176e-05,
      "loss": 0.27,
      "step": 100
    },
    {
      "epoch": 0.36996336996337,
      "grad_norm": 17.15272331237793,
      "learning_rate": 2.1978021978021977e-05,
      "loss": 0.446,
      "step": 101
    },
    {
      "epoch": 0.37362637362637363,
      "grad_norm": 45.865509033203125,
      "learning_rate": 2.21978021978022e-05,
      "loss": 2.4265,
      "step": 102
    },
    {
      "epoch": 0.3772893772893773,
      "grad_norm": 22.298274993896484,
      "learning_rate": 2.241758241758242e-05,
      "loss": 1.5021,
      "step": 103
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 20.314172744750977,
      "learning_rate": 2.2637362637362637e-05,
      "loss": 0.508,
      "step": 104
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 11.217910766601562,
      "learning_rate": 2.2857142857142858e-05,
      "loss": 0.2282,
      "step": 105
    },
    {
      "epoch": 0.3882783882783883,
      "grad_norm": 21.36184310913086,
      "learning_rate": 2.307692307692308e-05,
      "loss": 0.4684,
      "step": 106
    },
    {
      "epoch": 0.39194139194139194,
      "grad_norm": 12.759861946105957,
      "learning_rate": 2.3296703296703297e-05,
      "loss": 0.3076,
      "step": 107
    },
    {
      "epoch": 0.3956043956043956,
      "grad_norm": 24.42287254333496,
      "learning_rate": 2.3516483516483518e-05,
      "loss": 1.3607,
      "step": 108
    },
    {
      "epoch": 0.3992673992673993,
      "grad_norm": 13.014902114868164,
      "learning_rate": 2.3736263736263735e-05,
      "loss": 0.4984,
      "step": 109
    },
    {
      "epoch": 0.40293040293040294,
      "grad_norm": 12.8681640625,
      "learning_rate": 2.395604395604396e-05,
      "loss": 0.4529,
      "step": 110
    },
    {
      "epoch": 0.4065934065934066,
      "grad_norm": 21.19939422607422,
      "learning_rate": 2.4175824175824177e-05,
      "loss": 1.0197,
      "step": 111
    },
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 20.60430145263672,
      "learning_rate": 2.4395604395604395e-05,
      "loss": 0.5367,
      "step": 112
    },
    {
      "epoch": 0.4139194139194139,
      "grad_norm": 34.49782943725586,
      "learning_rate": 2.4615384615384616e-05,
      "loss": 1.9045,
      "step": 113
    },
    {
      "epoch": 0.4175824175824176,
      "grad_norm": 28.380966186523438,
      "learning_rate": 2.4835164835164834e-05,
      "loss": 0.9019,
      "step": 114
    },
    {
      "epoch": 0.42124542124542125,
      "grad_norm": 18.234045028686523,
      "learning_rate": 2.5054945054945058e-05,
      "loss": 0.5529,
      "step": 115
    },
    {
      "epoch": 0.4249084249084249,
      "grad_norm": 18.759784698486328,
      "learning_rate": 2.5274725274725276e-05,
      "loss": 0.85,
      "step": 116
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 15.784387588500977,
      "learning_rate": 2.5494505494505493e-05,
      "loss": 0.429,
      "step": 117
    },
    {
      "epoch": 0.43223443223443225,
      "grad_norm": 23.149036407470703,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 0.8784,
      "step": 118
    },
    {
      "epoch": 0.4358974358974359,
      "grad_norm": 18.77080535888672,
      "learning_rate": 2.5934065934065935e-05,
      "loss": 0.537,
      "step": 119
    },
    {
      "epoch": 0.43956043956043955,
      "grad_norm": 24.311708450317383,
      "learning_rate": 2.6153846153846157e-05,
      "loss": 0.74,
      "step": 120
    },
    {
      "epoch": 0.4432234432234432,
      "grad_norm": 15.09874439239502,
      "learning_rate": 2.6373626373626374e-05,
      "loss": 0.2978,
      "step": 121
    },
    {
      "epoch": 0.4468864468864469,
      "grad_norm": 19.65829086303711,
      "learning_rate": 2.6593406593406592e-05,
      "loss": 0.8287,
      "step": 122
    },
    {
      "epoch": 0.45054945054945056,
      "grad_norm": 21.237165451049805,
      "learning_rate": 2.6813186813186816e-05,
      "loss": 1.1967,
      "step": 123
    },
    {
      "epoch": 0.4542124542124542,
      "grad_norm": 25.737913131713867,
      "learning_rate": 2.7032967032967034e-05,
      "loss": 0.9414,
      "step": 124
    },
    {
      "epoch": 0.45787545787545786,
      "grad_norm": 22.84954833984375,
      "learning_rate": 2.7252747252747255e-05,
      "loss": 0.398,
      "step": 125
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 35.505027770996094,
      "learning_rate": 2.7472527472527473e-05,
      "loss": 1.0497,
      "step": 126
    },
    {
      "epoch": 0.4652014652014652,
      "grad_norm": 6.610748291015625,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.0491,
      "step": 127
    },
    {
      "epoch": 0.46886446886446886,
      "grad_norm": 33.34388732910156,
      "learning_rate": 2.7912087912087915e-05,
      "loss": 0.8991,
      "step": 128
    },
    {
      "epoch": 0.4725274725274725,
      "grad_norm": 17.098581314086914,
      "learning_rate": 2.8131868131868132e-05,
      "loss": 0.3217,
      "step": 129
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 11.438309669494629,
      "learning_rate": 2.8351648351648353e-05,
      "loss": 0.4301,
      "step": 130
    },
    {
      "epoch": 0.47985347985347987,
      "grad_norm": 25.803213119506836,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.8937,
      "step": 131
    },
    {
      "epoch": 0.4835164835164835,
      "grad_norm": 16.61037826538086,
      "learning_rate": 2.8791208791208792e-05,
      "loss": 0.3603,
      "step": 132
    },
    {
      "epoch": 0.48717948717948717,
      "grad_norm": 21.329975128173828,
      "learning_rate": 2.9010989010989013e-05,
      "loss": 0.4332,
      "step": 133
    },
    {
      "epoch": 0.4908424908424908,
      "grad_norm": 24.83706283569336,
      "learning_rate": 2.923076923076923e-05,
      "loss": 0.3967,
      "step": 134
    },
    {
      "epoch": 0.4945054945054945,
      "grad_norm": 8.3758544921875,
      "learning_rate": 2.945054945054945e-05,
      "loss": 0.1197,
      "step": 135
    },
    {
      "epoch": 0.4981684981684982,
      "grad_norm": 31.096702575683594,
      "learning_rate": 2.9670329670329673e-05,
      "loss": 2.2867,
      "step": 136
    },
    {
      "epoch": 0.5018315018315018,
      "grad_norm": 17.094390869140625,
      "learning_rate": 2.989010989010989e-05,
      "loss": 0.3064,
      "step": 137
    },
    {
      "epoch": 0.5054945054945055,
      "grad_norm": 23.401243209838867,
      "learning_rate": 3.010989010989011e-05,
      "loss": 0.9779,
      "step": 138
    },
    {
      "epoch": 0.5091575091575091,
      "grad_norm": 19.55811309814453,
      "learning_rate": 3.032967032967033e-05,
      "loss": 0.5665,
      "step": 139
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 18.668622970581055,
      "learning_rate": 3.0549450549450547e-05,
      "loss": 0.7068,
      "step": 140
    },
    {
      "epoch": 0.5164835164835165,
      "grad_norm": 9.49342155456543,
      "learning_rate": 3.076923076923077e-05,
      "loss": 0.2228,
      "step": 141
    },
    {
      "epoch": 0.5201465201465202,
      "grad_norm": 17.131006240844727,
      "learning_rate": 3.0989010989010995e-05,
      "loss": 0.8947,
      "step": 142
    },
    {
      "epoch": 0.5238095238095238,
      "grad_norm": 14.087484359741211,
      "learning_rate": 3.120879120879121e-05,
      "loss": 0.4394,
      "step": 143
    },
    {
      "epoch": 0.5274725274725275,
      "grad_norm": 14.246976852416992,
      "learning_rate": 3.142857142857143e-05,
      "loss": 0.7608,
      "step": 144
    },
    {
      "epoch": 0.5311355311355311,
      "grad_norm": 27.454071044921875,
      "learning_rate": 3.164835164835165e-05,
      "loss": 1.8982,
      "step": 145
    },
    {
      "epoch": 0.5347985347985348,
      "grad_norm": 8.580923080444336,
      "learning_rate": 3.1868131868131866e-05,
      "loss": 0.2199,
      "step": 146
    },
    {
      "epoch": 0.5384615384615384,
      "grad_norm": 12.200552940368652,
      "learning_rate": 3.208791208791209e-05,
      "loss": 0.4007,
      "step": 147
    },
    {
      "epoch": 0.5421245421245421,
      "grad_norm": 11.350752830505371,
      "learning_rate": 3.230769230769231e-05,
      "loss": 0.5359,
      "step": 148
    },
    {
      "epoch": 0.5457875457875457,
      "grad_norm": 21.45020866394043,
      "learning_rate": 3.252747252747253e-05,
      "loss": 1.4639,
      "step": 149
    },
    {
      "epoch": 0.5494505494505495,
      "grad_norm": 29.84933090209961,
      "learning_rate": 3.274725274725274e-05,
      "loss": 0.8764,
      "step": 150
    },
    {
      "epoch": 0.5531135531135531,
      "grad_norm": 14.899048805236816,
      "learning_rate": 3.296703296703297e-05,
      "loss": 0.3817,
      "step": 151
    },
    {
      "epoch": 0.5567765567765568,
      "grad_norm": 14.95295238494873,
      "learning_rate": 3.318681318681319e-05,
      "loss": 1.0153,
      "step": 152
    },
    {
      "epoch": 0.5604395604395604,
      "grad_norm": 13.904314994812012,
      "learning_rate": 3.3406593406593406e-05,
      "loss": 0.9891,
      "step": 153
    },
    {
      "epoch": 0.5641025641025641,
      "grad_norm": 14.465546607971191,
      "learning_rate": 3.362637362637363e-05,
      "loss": 0.4935,
      "step": 154
    },
    {
      "epoch": 0.5677655677655677,
      "grad_norm": 15.22211742401123,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.4973,
      "step": 155
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 19.977941513061523,
      "learning_rate": 3.406593406593406e-05,
      "loss": 0.5768,
      "step": 156
    },
    {
      "epoch": 0.575091575091575,
      "grad_norm": 21.778785705566406,
      "learning_rate": 3.4285714285714284e-05,
      "loss": 0.541,
      "step": 157
    },
    {
      "epoch": 0.5787545787545788,
      "grad_norm": 7.957052707672119,
      "learning_rate": 3.4505494505494505e-05,
      "loss": 0.1676,
      "step": 158
    },
    {
      "epoch": 0.5824175824175825,
      "grad_norm": 10.105476379394531,
      "learning_rate": 3.472527472527473e-05,
      "loss": 0.14,
      "step": 159
    },
    {
      "epoch": 0.5860805860805861,
      "grad_norm": 13.895249366760254,
      "learning_rate": 3.494505494505495e-05,
      "loss": 0.2135,
      "step": 160
    },
    {
      "epoch": 0.5897435897435898,
      "grad_norm": 15.14104175567627,
      "learning_rate": 3.516483516483517e-05,
      "loss": 0.2299,
      "step": 161
    },
    {
      "epoch": 0.5934065934065934,
      "grad_norm": 27.537504196166992,
      "learning_rate": 3.538461538461539e-05,
      "loss": 0.4517,
      "step": 162
    },
    {
      "epoch": 0.5970695970695971,
      "grad_norm": 22.290597915649414,
      "learning_rate": 3.56043956043956e-05,
      "loss": 0.2144,
      "step": 163
    },
    {
      "epoch": 0.6007326007326007,
      "grad_norm": 24.176603317260742,
      "learning_rate": 3.5824175824175824e-05,
      "loss": 0.4184,
      "step": 164
    },
    {
      "epoch": 0.6043956043956044,
      "grad_norm": 43.716552734375,
      "learning_rate": 3.6043956043956045e-05,
      "loss": 0.7672,
      "step": 165
    },
    {
      "epoch": 0.608058608058608,
      "grad_norm": 5.516793727874756,
      "learning_rate": 3.626373626373626e-05,
      "loss": 0.0332,
      "step": 166
    },
    {
      "epoch": 0.6117216117216118,
      "grad_norm": 13.202600479125977,
      "learning_rate": 3.648351648351648e-05,
      "loss": 0.1388,
      "step": 167
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 8.389626502990723,
      "learning_rate": 3.670329670329671e-05,
      "loss": 0.0284,
      "step": 168
    },
    {
      "epoch": 0.6190476190476191,
      "grad_norm": 11.500190734863281,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.1778,
      "step": 169
    },
    {
      "epoch": 0.6227106227106227,
      "grad_norm": 49.76407241821289,
      "learning_rate": 3.7142857142857143e-05,
      "loss": 0.8075,
      "step": 170
    },
    {
      "epoch": 0.6263736263736264,
      "grad_norm": 49.758705139160156,
      "learning_rate": 3.7362637362637365e-05,
      "loss": 1.3106,
      "step": 171
    },
    {
      "epoch": 0.63003663003663,
      "grad_norm": 7.655544281005859,
      "learning_rate": 3.7582417582417586e-05,
      "loss": 0.1362,
      "step": 172
    },
    {
      "epoch": 0.6336996336996337,
      "grad_norm": 29.778133392333984,
      "learning_rate": 3.78021978021978e-05,
      "loss": 0.2411,
      "step": 173
    },
    {
      "epoch": 0.6373626373626373,
      "grad_norm": 23.79543113708496,
      "learning_rate": 3.802197802197802e-05,
      "loss": 0.5665,
      "step": 174
    },
    {
      "epoch": 0.6410256410256411,
      "grad_norm": 25.333166122436523,
      "learning_rate": 3.824175824175824e-05,
      "loss": 0.5821,
      "step": 175
    },
    {
      "epoch": 0.6446886446886447,
      "grad_norm": 38.367759704589844,
      "learning_rate": 3.846153846153846e-05,
      "loss": 1.1098,
      "step": 176
    },
    {
      "epoch": 0.6483516483516484,
      "grad_norm": 31.53361701965332,
      "learning_rate": 3.8681318681318684e-05,
      "loss": 1.5399,
      "step": 177
    },
    {
      "epoch": 0.652014652014652,
      "grad_norm": 8.453901290893555,
      "learning_rate": 3.8901098901098905e-05,
      "loss": 0.1327,
      "step": 178
    },
    {
      "epoch": 0.6556776556776557,
      "grad_norm": 32.465980529785156,
      "learning_rate": 3.9120879120879126e-05,
      "loss": 0.8133,
      "step": 179
    },
    {
      "epoch": 0.6593406593406593,
      "grad_norm": 21.503114700317383,
      "learning_rate": 3.934065934065934e-05,
      "loss": 0.2472,
      "step": 180
    },
    {
      "epoch": 0.663003663003663,
      "grad_norm": 28.240659713745117,
      "learning_rate": 3.956043956043956e-05,
      "loss": 0.4718,
      "step": 181
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 6.919331073760986,
      "learning_rate": 3.978021978021978e-05,
      "loss": 0.0947,
      "step": 182
    },
    {
      "epoch": 0.6703296703296703,
      "grad_norm": 20.96783447265625,
      "learning_rate": 3.9999999999999996e-05,
      "loss": 1.1602,
      "step": 183
    },
    {
      "epoch": 0.673992673992674,
      "grad_norm": 17.967914581298828,
      "learning_rate": 4.021978021978022e-05,
      "loss": 0.3684,
      "step": 184
    },
    {
      "epoch": 0.6776556776556777,
      "grad_norm": 29.837678909301758,
      "learning_rate": 4.0439560439560445e-05,
      "loss": 0.5452,
      "step": 185
    },
    {
      "epoch": 0.6813186813186813,
      "grad_norm": 37.0803108215332,
      "learning_rate": 4.065934065934066e-05,
      "loss": 0.5983,
      "step": 186
    },
    {
      "epoch": 0.684981684981685,
      "grad_norm": 23.339448928833008,
      "learning_rate": 4.087912087912088e-05,
      "loss": 0.6255,
      "step": 187
    },
    {
      "epoch": 0.6886446886446886,
      "grad_norm": 13.779767036437988,
      "learning_rate": 4.10989010989011e-05,
      "loss": 0.3705,
      "step": 188
    },
    {
      "epoch": 0.6923076923076923,
      "grad_norm": 15.792436599731445,
      "learning_rate": 4.131868131868132e-05,
      "loss": 0.4128,
      "step": 189
    },
    {
      "epoch": 0.6959706959706959,
      "grad_norm": 14.106623649597168,
      "learning_rate": 4.153846153846154e-05,
      "loss": 0.2914,
      "step": 190
    },
    {
      "epoch": 0.6996336996336996,
      "grad_norm": 34.428951263427734,
      "learning_rate": 4.175824175824176e-05,
      "loss": 1.2232,
      "step": 191
    },
    {
      "epoch": 0.7032967032967034,
      "grad_norm": 15.847033500671387,
      "learning_rate": 4.197802197802198e-05,
      "loss": 0.4129,
      "step": 192
    },
    {
      "epoch": 0.706959706959707,
      "grad_norm": 17.834794998168945,
      "learning_rate": 4.21978021978022e-05,
      "loss": 0.4158,
      "step": 193
    },
    {
      "epoch": 0.7106227106227107,
      "grad_norm": 29.807823181152344,
      "learning_rate": 4.241758241758242e-05,
      "loss": 0.9741,
      "step": 194
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 15.9482421875,
      "learning_rate": 4.263736263736264e-05,
      "loss": 0.1953,
      "step": 195
    },
    {
      "epoch": 0.717948717948718,
      "grad_norm": 37.89487075805664,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 1.1018,
      "step": 196
    },
    {
      "epoch": 0.7216117216117216,
      "grad_norm": 24.060779571533203,
      "learning_rate": 4.307692307692308e-05,
      "loss": 0.4774,
      "step": 197
    },
    {
      "epoch": 0.7252747252747253,
      "grad_norm": 18.701725006103516,
      "learning_rate": 4.32967032967033e-05,
      "loss": 0.2641,
      "step": 198
    },
    {
      "epoch": 0.7289377289377289,
      "grad_norm": 32.18348693847656,
      "learning_rate": 4.351648351648352e-05,
      "loss": 0.6958,
      "step": 199
    },
    {
      "epoch": 0.7326007326007326,
      "grad_norm": 16.504337310791016,
      "learning_rate": 4.3736263736263734e-05,
      "loss": 0.1933,
      "step": 200
    },
    {
      "epoch": 0.7362637362637363,
      "grad_norm": 34.5928840637207,
      "learning_rate": 4.3956043956043955e-05,
      "loss": 0.3712,
      "step": 201
    },
    {
      "epoch": 0.73992673992674,
      "grad_norm": 47.998512268066406,
      "learning_rate": 4.417582417582418e-05,
      "loss": 1.4578,
      "step": 202
    },
    {
      "epoch": 0.7435897435897436,
      "grad_norm": 29.871829986572266,
      "learning_rate": 4.43956043956044e-05,
      "loss": 0.7628,
      "step": 203
    },
    {
      "epoch": 0.7472527472527473,
      "grad_norm": 53.70481491088867,
      "learning_rate": 4.461538461538462e-05,
      "loss": 1.4017,
      "step": 204
    },
    {
      "epoch": 0.7509157509157509,
      "grad_norm": 58.087646484375,
      "learning_rate": 4.483516483516484e-05,
      "loss": 1.3168,
      "step": 205
    },
    {
      "epoch": 0.7545787545787546,
      "grad_norm": 44.62531280517578,
      "learning_rate": 4.505494505494505e-05,
      "loss": 0.8959,
      "step": 206
    },
    {
      "epoch": 0.7582417582417582,
      "grad_norm": 18.427953720092773,
      "learning_rate": 4.5274725274725274e-05,
      "loss": 0.4202,
      "step": 207
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 32.799434661865234,
      "learning_rate": 4.5494505494505495e-05,
      "loss": 0.5432,
      "step": 208
    },
    {
      "epoch": 0.7655677655677655,
      "grad_norm": 22.136354446411133,
      "learning_rate": 4.5714285714285716e-05,
      "loss": 1.0474,
      "step": 209
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 14.09807014465332,
      "learning_rate": 4.593406593406593e-05,
      "loss": 0.4048,
      "step": 210
    },
    {
      "epoch": 0.7728937728937729,
      "grad_norm": 16.818132400512695,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.4772,
      "step": 211
    },
    {
      "epoch": 0.7765567765567766,
      "grad_norm": 36.87644577026367,
      "learning_rate": 4.637362637362638e-05,
      "loss": 1.0203,
      "step": 212
    },
    {
      "epoch": 0.7802197802197802,
      "grad_norm": 23.279033660888672,
      "learning_rate": 4.6593406593406593e-05,
      "loss": 0.8223,
      "step": 213
    },
    {
      "epoch": 0.7838827838827839,
      "grad_norm": 21.23172378540039,
      "learning_rate": 4.6813186813186814e-05,
      "loss": 0.6838,
      "step": 214
    },
    {
      "epoch": 0.7875457875457875,
      "grad_norm": 15.129582405090332,
      "learning_rate": 4.7032967032967035e-05,
      "loss": 0.3939,
      "step": 215
    },
    {
      "epoch": 0.7912087912087912,
      "grad_norm": 38.20903778076172,
      "learning_rate": 4.725274725274725e-05,
      "loss": 0.4395,
      "step": 216
    },
    {
      "epoch": 0.7948717948717948,
      "grad_norm": 23.428571701049805,
      "learning_rate": 4.747252747252747e-05,
      "loss": 0.6657,
      "step": 217
    },
    {
      "epoch": 0.7985347985347986,
      "grad_norm": 15.892741203308105,
      "learning_rate": 4.769230769230769e-05,
      "loss": 0.3867,
      "step": 218
    },
    {
      "epoch": 0.8021978021978022,
      "grad_norm": 44.7977180480957,
      "learning_rate": 4.791208791208792e-05,
      "loss": 1.4335,
      "step": 219
    },
    {
      "epoch": 0.8058608058608059,
      "grad_norm": 18.13700294494629,
      "learning_rate": 4.8131868131868134e-05,
      "loss": 0.3965,
      "step": 220
    },
    {
      "epoch": 0.8095238095238095,
      "grad_norm": 23.00497817993164,
      "learning_rate": 4.8351648351648355e-05,
      "loss": 1.1319,
      "step": 221
    },
    {
      "epoch": 0.8131868131868132,
      "grad_norm": 27.63648796081543,
      "learning_rate": 4.8571428571428576e-05,
      "loss": 0.7782,
      "step": 222
    },
    {
      "epoch": 0.8168498168498168,
      "grad_norm": 23.91630744934082,
      "learning_rate": 4.879120879120879e-05,
      "loss": 0.7277,
      "step": 223
    },
    {
      "epoch": 0.8205128205128205,
      "grad_norm": 27.157682418823242,
      "learning_rate": 4.901098901098901e-05,
      "loss": 0.8309,
      "step": 224
    },
    {
      "epoch": 0.8241758241758241,
      "grad_norm": 20.686105728149414,
      "learning_rate": 4.923076923076923e-05,
      "loss": 0.4645,
      "step": 225
    },
    {
      "epoch": 0.8278388278388278,
      "grad_norm": 18.44706916809082,
      "learning_rate": 4.9450549450549446e-05,
      "loss": 0.6298,
      "step": 226
    },
    {
      "epoch": 0.8315018315018315,
      "grad_norm": 34.66194152832031,
      "learning_rate": 4.967032967032967e-05,
      "loss": 1.3282,
      "step": 227
    },
    {
      "epoch": 0.8351648351648352,
      "grad_norm": 26.68456268310547,
      "learning_rate": 4.9890109890109895e-05,
      "loss": 0.8652,
      "step": 228
    },
    {
      "epoch": 0.8388278388278388,
      "grad_norm": 18.36819839477539,
      "learning_rate": 5.0109890109890116e-05,
      "loss": 0.425,
      "step": 229
    },
    {
      "epoch": 0.8424908424908425,
      "grad_norm": 10.212838172912598,
      "learning_rate": 5.032967032967033e-05,
      "loss": 0.2183,
      "step": 230
    },
    {
      "epoch": 0.8461538461538461,
      "grad_norm": 28.40265464782715,
      "learning_rate": 5.054945054945055e-05,
      "loss": 1.6894,
      "step": 231
    },
    {
      "epoch": 0.8498168498168498,
      "grad_norm": 48.70882797241211,
      "learning_rate": 5.076923076923077e-05,
      "loss": 0.8564,
      "step": 232
    },
    {
      "epoch": 0.8534798534798534,
      "grad_norm": 38.576541900634766,
      "learning_rate": 5.098901098901099e-05,
      "loss": 0.8013,
      "step": 233
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 20.17264747619629,
      "learning_rate": 5.120879120879121e-05,
      "loss": 0.4553,
      "step": 234
    },
    {
      "epoch": 0.8608058608058609,
      "grad_norm": 33.383182525634766,
      "learning_rate": 5.142857142857143e-05,
      "loss": 0.9591,
      "step": 235
    },
    {
      "epoch": 0.8644688644688645,
      "grad_norm": 22.734106063842773,
      "learning_rate": 5.164835164835165e-05,
      "loss": 0.589,
      "step": 236
    },
    {
      "epoch": 0.8681318681318682,
      "grad_norm": 19.77442741394043,
      "learning_rate": 5.186813186813187e-05,
      "loss": 0.7066,
      "step": 237
    },
    {
      "epoch": 0.8717948717948718,
      "grad_norm": 32.36431884765625,
      "learning_rate": 5.208791208791209e-05,
      "loss": 0.8878,
      "step": 238
    },
    {
      "epoch": 0.8754578754578755,
      "grad_norm": 37.60574722290039,
      "learning_rate": 5.230769230769231e-05,
      "loss": 1.0034,
      "step": 239
    },
    {
      "epoch": 0.8791208791208791,
      "grad_norm": 28.051666259765625,
      "learning_rate": 5.252747252747253e-05,
      "loss": 0.9695,
      "step": 240
    },
    {
      "epoch": 0.8827838827838828,
      "grad_norm": 31.55886459350586,
      "learning_rate": 5.274725274725275e-05,
      "loss": 0.5416,
      "step": 241
    },
    {
      "epoch": 0.8864468864468864,
      "grad_norm": 17.856632232666016,
      "learning_rate": 5.296703296703297e-05,
      "loss": 0.3647,
      "step": 242
    },
    {
      "epoch": 0.8901098901098901,
      "grad_norm": 42.52962112426758,
      "learning_rate": 5.3186813186813184e-05,
      "loss": 1.3661,
      "step": 243
    },
    {
      "epoch": 0.8937728937728938,
      "grad_norm": 26.439769744873047,
      "learning_rate": 5.3406593406593405e-05,
      "loss": 0.6629,
      "step": 244
    },
    {
      "epoch": 0.8974358974358975,
      "grad_norm": 37.46576690673828,
      "learning_rate": 5.362637362637363e-05,
      "loss": 0.9631,
      "step": 245
    },
    {
      "epoch": 0.9010989010989011,
      "grad_norm": 29.706708908081055,
      "learning_rate": 5.384615384615385e-05,
      "loss": 1.0034,
      "step": 246
    },
    {
      "epoch": 0.9047619047619048,
      "grad_norm": 33.62871551513672,
      "learning_rate": 5.406593406593407e-05,
      "loss": 0.8036,
      "step": 247
    },
    {
      "epoch": 0.9084249084249084,
      "grad_norm": 41.97051239013672,
      "learning_rate": 5.428571428571429e-05,
      "loss": 1.309,
      "step": 248
    },
    {
      "epoch": 0.9120879120879121,
      "grad_norm": 37.57841110229492,
      "learning_rate": 5.450549450549451e-05,
      "loss": 1.2444,
      "step": 249
    },
    {
      "epoch": 0.9157509157509157,
      "grad_norm": 21.220727920532227,
      "learning_rate": 5.4725274725274724e-05,
      "loss": 0.6556,
      "step": 250
    },
    {
      "epoch": 0.9194139194139194,
      "grad_norm": 19.963764190673828,
      "learning_rate": 5.4945054945054945e-05,
      "loss": 0.7328,
      "step": 251
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 21.196062088012695,
      "learning_rate": 5.5164835164835166e-05,
      "loss": 0.5752,
      "step": 252
    },
    {
      "epoch": 0.9267399267399268,
      "grad_norm": 23.587268829345703,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.4801,
      "step": 253
    },
    {
      "epoch": 0.9304029304029304,
      "grad_norm": 16.09604263305664,
      "learning_rate": 5.560439560439561e-05,
      "loss": 0.4795,
      "step": 254
    },
    {
      "epoch": 0.9340659340659341,
      "grad_norm": 22.61296272277832,
      "learning_rate": 5.582417582417583e-05,
      "loss": 0.5807,
      "step": 255
    },
    {
      "epoch": 0.9377289377289377,
      "grad_norm": 28.715890884399414,
      "learning_rate": 5.604395604395604e-05,
      "loss": 1.3141,
      "step": 256
    },
    {
      "epoch": 0.9413919413919414,
      "grad_norm": 37.11213684082031,
      "learning_rate": 5.6263736263736264e-05,
      "loss": 1.7168,
      "step": 257
    },
    {
      "epoch": 0.945054945054945,
      "grad_norm": 13.693246841430664,
      "learning_rate": 5.6483516483516485e-05,
      "loss": 0.3207,
      "step": 258
    },
    {
      "epoch": 0.9487179487179487,
      "grad_norm": 18.186216354370117,
      "learning_rate": 5.6703296703296706e-05,
      "loss": 0.6265,
      "step": 259
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 23.68426513671875,
      "learning_rate": 5.692307692307692e-05,
      "loss": 0.5226,
      "step": 260
    },
    {
      "epoch": 0.9560439560439561,
      "grad_norm": 19.154836654663086,
      "learning_rate": 5.714285714285714e-05,
      "loss": 1.0116,
      "step": 261
    },
    {
      "epoch": 0.9597069597069597,
      "grad_norm": 17.64719009399414,
      "learning_rate": 5.736263736263737e-05,
      "loss": 0.5992,
      "step": 262
    },
    {
      "epoch": 0.9633699633699634,
      "grad_norm": 25.542757034301758,
      "learning_rate": 5.7582417582417584e-05,
      "loss": 0.8129,
      "step": 263
    },
    {
      "epoch": 0.967032967032967,
      "grad_norm": 25.94204330444336,
      "learning_rate": 5.7802197802197805e-05,
      "loss": 1.2194,
      "step": 264
    },
    {
      "epoch": 0.9706959706959707,
      "grad_norm": 13.693342208862305,
      "learning_rate": 5.8021978021978026e-05,
      "loss": 0.2565,
      "step": 265
    },
    {
      "epoch": 0.9743589743589743,
      "grad_norm": 20.760122299194336,
      "learning_rate": 5.824175824175824e-05,
      "loss": 0.4023,
      "step": 266
    },
    {
      "epoch": 0.978021978021978,
      "grad_norm": 20.00895118713379,
      "learning_rate": 5.846153846153846e-05,
      "loss": 0.2468,
      "step": 267
    },
    {
      "epoch": 0.9816849816849816,
      "grad_norm": 25.56069564819336,
      "learning_rate": 5.868131868131868e-05,
      "loss": 0.5648,
      "step": 268
    },
    {
      "epoch": 0.9853479853479854,
      "grad_norm": 38.19970703125,
      "learning_rate": 5.89010989010989e-05,
      "loss": 0.544,
      "step": 269
    },
    {
      "epoch": 0.989010989010989,
      "grad_norm": 37.63619613647461,
      "learning_rate": 5.9120879120879124e-05,
      "loss": 0.7556,
      "step": 270
    },
    {
      "epoch": 0.9926739926739927,
      "grad_norm": 10.586868286132812,
      "learning_rate": 5.9340659340659345e-05,
      "loss": 0.1003,
      "step": 271
    },
    {
      "epoch": 0.9963369963369964,
      "grad_norm": 17.579208374023438,
      "learning_rate": 5.9560439560439566e-05,
      "loss": 0.2931,
      "step": 272
    },
    {
      "epoch": 1.0,
      "grad_norm": 24.657121658325195,
      "learning_rate": 5.978021978021978e-05,
      "loss": 0.2372,
      "step": 273
    },
    {
      "epoch": 1.0036630036630036,
      "grad_norm": 29.52134895324707,
      "learning_rate": 6e-05,
      "loss": 0.5077,
      "step": 274
    },
    {
      "epoch": 1.0073260073260073,
      "grad_norm": 51.900062561035156,
      "learning_rate": 5.997557997557998e-05,
      "loss": 0.4404,
      "step": 275
    },
    {
      "epoch": 1.010989010989011,
      "grad_norm": 18.682769775390625,
      "learning_rate": 5.995115995115995e-05,
      "loss": 0.2405,
      "step": 276
    },
    {
      "epoch": 1.0146520146520146,
      "grad_norm": 87.95014953613281,
      "learning_rate": 5.992673992673993e-05,
      "loss": 2.8585,
      "step": 277
    },
    {
      "epoch": 1.0183150183150182,
      "grad_norm": 67.03990936279297,
      "learning_rate": 5.990231990231991e-05,
      "loss": 0.9746,
      "step": 278
    },
    {
      "epoch": 1.021978021978022,
      "grad_norm": 47.63545227050781,
      "learning_rate": 5.987789987789988e-05,
      "loss": 0.241,
      "step": 279
    },
    {
      "epoch": 1.0256410256410255,
      "grad_norm": 33.62876892089844,
      "learning_rate": 5.985347985347986e-05,
      "loss": 1.0003,
      "step": 280
    },
    {
      "epoch": 1.0293040293040292,
      "grad_norm": 30.26620864868164,
      "learning_rate": 5.982905982905983e-05,
      "loss": 0.7767,
      "step": 281
    },
    {
      "epoch": 1.032967032967033,
      "grad_norm": 33.785770416259766,
      "learning_rate": 5.98046398046398e-05,
      "loss": 0.899,
      "step": 282
    },
    {
      "epoch": 1.0366300366300367,
      "grad_norm": 33.753849029541016,
      "learning_rate": 5.978021978021978e-05,
      "loss": 1.8225,
      "step": 283
    },
    {
      "epoch": 1.0402930402930404,
      "grad_norm": 16.58989143371582,
      "learning_rate": 5.975579975579976e-05,
      "loss": 0.6211,
      "step": 284
    },
    {
      "epoch": 1.043956043956044,
      "grad_norm": 23.08768653869629,
      "learning_rate": 5.973137973137973e-05,
      "loss": 0.7541,
      "step": 285
    },
    {
      "epoch": 1.0476190476190477,
      "grad_norm": 24.57805824279785,
      "learning_rate": 5.970695970695971e-05,
      "loss": 0.8278,
      "step": 286
    },
    {
      "epoch": 1.0512820512820513,
      "grad_norm": 25.1593017578125,
      "learning_rate": 5.968253968253968e-05,
      "loss": 0.6932,
      "step": 287
    },
    {
      "epoch": 1.054945054945055,
      "grad_norm": 29.984054565429688,
      "learning_rate": 5.965811965811966e-05,
      "loss": 0.6987,
      "step": 288
    },
    {
      "epoch": 1.0586080586080586,
      "grad_norm": 28.183151245117188,
      "learning_rate": 5.963369963369964e-05,
      "loss": 0.8771,
      "step": 289
    },
    {
      "epoch": 1.0622710622710623,
      "grad_norm": 15.349969863891602,
      "learning_rate": 5.960927960927961e-05,
      "loss": 0.2906,
      "step": 290
    },
    {
      "epoch": 1.065934065934066,
      "grad_norm": 17.618196487426758,
      "learning_rate": 5.958485958485959e-05,
      "loss": 0.595,
      "step": 291
    },
    {
      "epoch": 1.0695970695970696,
      "grad_norm": 40.537925720214844,
      "learning_rate": 5.9560439560439566e-05,
      "loss": 1.3881,
      "step": 292
    },
    {
      "epoch": 1.0732600732600732,
      "grad_norm": 41.12261962890625,
      "learning_rate": 5.953601953601954e-05,
      "loss": 0.5402,
      "step": 293
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 38.4654655456543,
      "learning_rate": 5.951159951159951e-05,
      "loss": 0.3097,
      "step": 294
    },
    {
      "epoch": 1.0805860805860805,
      "grad_norm": 34.19886016845703,
      "learning_rate": 5.948717948717949e-05,
      "loss": 1.0228,
      "step": 295
    },
    {
      "epoch": 1.0842490842490842,
      "grad_norm": 19.727413177490234,
      "learning_rate": 5.946275946275946e-05,
      "loss": 0.1755,
      "step": 296
    },
    {
      "epoch": 1.0879120879120878,
      "grad_norm": 33.413352966308594,
      "learning_rate": 5.943833943833944e-05,
      "loss": 0.8087,
      "step": 297
    },
    {
      "epoch": 1.0915750915750915,
      "grad_norm": 29.848875045776367,
      "learning_rate": 5.941391941391942e-05,
      "loss": 0.673,
      "step": 298
    },
    {
      "epoch": 1.0952380952380953,
      "grad_norm": 18.643922805786133,
      "learning_rate": 5.938949938949939e-05,
      "loss": 0.4759,
      "step": 299
    },
    {
      "epoch": 1.098901098901099,
      "grad_norm": 28.923099517822266,
      "learning_rate": 5.936507936507937e-05,
      "loss": 0.6555,
      "step": 300
    },
    {
      "epoch": 1.1025641025641026,
      "grad_norm": 26.4990177154541,
      "learning_rate": 5.9340659340659345e-05,
      "loss": 0.4679,
      "step": 301
    },
    {
      "epoch": 1.1062271062271063,
      "grad_norm": 43.54881286621094,
      "learning_rate": 5.931623931623932e-05,
      "loss": 1.0861,
      "step": 302
    },
    {
      "epoch": 1.10989010989011,
      "grad_norm": 32.66098403930664,
      "learning_rate": 5.9291819291819295e-05,
      "loss": 0.677,
      "step": 303
    },
    {
      "epoch": 1.1135531135531136,
      "grad_norm": 43.79314422607422,
      "learning_rate": 5.9267399267399274e-05,
      "loss": 0.8883,
      "step": 304
    },
    {
      "epoch": 1.1172161172161172,
      "grad_norm": 44.49085235595703,
      "learning_rate": 5.9242979242979245e-05,
      "loss": 0.9553,
      "step": 305
    },
    {
      "epoch": 1.120879120879121,
      "grad_norm": 31.713787078857422,
      "learning_rate": 5.9218559218559224e-05,
      "loss": 0.6352,
      "step": 306
    },
    {
      "epoch": 1.1245421245421245,
      "grad_norm": 19.930402755737305,
      "learning_rate": 5.9194139194139196e-05,
      "loss": 0.7023,
      "step": 307
    },
    {
      "epoch": 1.1282051282051282,
      "grad_norm": 20.157196044921875,
      "learning_rate": 5.916971916971917e-05,
      "loss": 0.6241,
      "step": 308
    },
    {
      "epoch": 1.1318681318681318,
      "grad_norm": 26.819135665893555,
      "learning_rate": 5.9145299145299146e-05,
      "loss": 0.4788,
      "step": 309
    },
    {
      "epoch": 1.1355311355311355,
      "grad_norm": 24.948625564575195,
      "learning_rate": 5.9120879120879124e-05,
      "loss": 0.698,
      "step": 310
    },
    {
      "epoch": 1.1391941391941391,
      "grad_norm": 15.883389472961426,
      "learning_rate": 5.9096459096459096e-05,
      "loss": 0.3325,
      "step": 311
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 25.214584350585938,
      "learning_rate": 5.9072039072039074e-05,
      "loss": 0.4776,
      "step": 312
    },
    {
      "epoch": 1.1465201465201464,
      "grad_norm": 27.4523983001709,
      "learning_rate": 5.9047619047619046e-05,
      "loss": 0.6155,
      "step": 313
    },
    {
      "epoch": 1.15018315018315,
      "grad_norm": 48.60593795776367,
      "learning_rate": 5.9023199023199024e-05,
      "loss": 1.7225,
      "step": 314
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 27.19314193725586,
      "learning_rate": 5.8998778998779e-05,
      "loss": 0.6805,
      "step": 315
    },
    {
      "epoch": 1.1575091575091574,
      "grad_norm": 44.678768157958984,
      "learning_rate": 5.8974358974358975e-05,
      "loss": 0.5721,
      "step": 316
    },
    {
      "epoch": 1.1611721611721613,
      "grad_norm": 12.109644889831543,
      "learning_rate": 5.894993894993895e-05,
      "loss": 0.1079,
      "step": 317
    },
    {
      "epoch": 1.164835164835165,
      "grad_norm": 45.254730224609375,
      "learning_rate": 5.892551892551893e-05,
      "loss": 1.1492,
      "step": 318
    },
    {
      "epoch": 1.1684981684981686,
      "grad_norm": 65.83439636230469,
      "learning_rate": 5.89010989010989e-05,
      "loss": 0.7049,
      "step": 319
    },
    {
      "epoch": 1.1721611721611722,
      "grad_norm": 43.5418586730957,
      "learning_rate": 5.8876678876678875e-05,
      "loss": 0.4628,
      "step": 320
    },
    {
      "epoch": 1.1758241758241759,
      "grad_norm": 137.285400390625,
      "learning_rate": 5.885225885225885e-05,
      "loss": 1.4227,
      "step": 321
    },
    {
      "epoch": 1.1794871794871795,
      "grad_norm": 42.895565032958984,
      "learning_rate": 5.8827838827838825e-05,
      "loss": 0.4264,
      "step": 322
    },
    {
      "epoch": 1.1831501831501832,
      "grad_norm": 10.602986335754395,
      "learning_rate": 5.8803418803418803e-05,
      "loss": 0.0494,
      "step": 323
    },
    {
      "epoch": 1.1868131868131868,
      "grad_norm": 103.92290496826172,
      "learning_rate": 5.877899877899878e-05,
      "loss": 2.0111,
      "step": 324
    },
    {
      "epoch": 1.1904761904761905,
      "grad_norm": 36.497764587402344,
      "learning_rate": 5.8754578754578754e-05,
      "loss": 0.4768,
      "step": 325
    },
    {
      "epoch": 1.1941391941391941,
      "grad_norm": 45.52228546142578,
      "learning_rate": 5.873015873015873e-05,
      "loss": 0.994,
      "step": 326
    },
    {
      "epoch": 1.1978021978021978,
      "grad_norm": 24.81894302368164,
      "learning_rate": 5.870573870573871e-05,
      "loss": 0.5563,
      "step": 327
    },
    {
      "epoch": 1.2014652014652014,
      "grad_norm": 49.82950210571289,
      "learning_rate": 5.868131868131868e-05,
      "loss": 1.5448,
      "step": 328
    },
    {
      "epoch": 1.205128205128205,
      "grad_norm": 23.945913314819336,
      "learning_rate": 5.865689865689866e-05,
      "loss": 0.5256,
      "step": 329
    },
    {
      "epoch": 1.2087912087912087,
      "grad_norm": 20.63251304626465,
      "learning_rate": 5.863247863247864e-05,
      "loss": 0.3698,
      "step": 330
    },
    {
      "epoch": 1.2124542124542124,
      "grad_norm": 32.270328521728516,
      "learning_rate": 5.860805860805861e-05,
      "loss": 0.3518,
      "step": 331
    },
    {
      "epoch": 1.2161172161172162,
      "grad_norm": 32.445716857910156,
      "learning_rate": 5.858363858363858e-05,
      "loss": 0.857,
      "step": 332
    },
    {
      "epoch": 1.2197802197802199,
      "grad_norm": 59.69521713256836,
      "learning_rate": 5.855921855921856e-05,
      "loss": 1.3786,
      "step": 333
    },
    {
      "epoch": 1.2234432234432235,
      "grad_norm": 32.79878234863281,
      "learning_rate": 5.853479853479853e-05,
      "loss": 0.7648,
      "step": 334
    },
    {
      "epoch": 1.2271062271062272,
      "grad_norm": 26.749393463134766,
      "learning_rate": 5.851037851037851e-05,
      "loss": 0.4723,
      "step": 335
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 40.744102478027344,
      "learning_rate": 5.848595848595849e-05,
      "loss": 1.0543,
      "step": 336
    },
    {
      "epoch": 1.2344322344322345,
      "grad_norm": 34.2275505065918,
      "learning_rate": 5.846153846153846e-05,
      "loss": 0.4533,
      "step": 337
    },
    {
      "epoch": 1.2380952380952381,
      "grad_norm": 49.648136138916016,
      "learning_rate": 5.843711843711844e-05,
      "loss": 1.2112,
      "step": 338
    },
    {
      "epoch": 1.2417582417582418,
      "grad_norm": 64.69720458984375,
      "learning_rate": 5.841269841269841e-05,
      "loss": 1.2234,
      "step": 339
    },
    {
      "epoch": 1.2454212454212454,
      "grad_norm": 16.81964111328125,
      "learning_rate": 5.838827838827839e-05,
      "loss": 0.297,
      "step": 340
    },
    {
      "epoch": 1.249084249084249,
      "grad_norm": 17.393678665161133,
      "learning_rate": 5.836385836385837e-05,
      "loss": 0.2504,
      "step": 341
    },
    {
      "epoch": 1.2527472527472527,
      "grad_norm": 64.2254409790039,
      "learning_rate": 5.833943833943834e-05,
      "loss": 1.3656,
      "step": 342
    },
    {
      "epoch": 1.2564102564102564,
      "grad_norm": 48.991249084472656,
      "learning_rate": 5.831501831501832e-05,
      "loss": 1.0819,
      "step": 343
    },
    {
      "epoch": 1.26007326007326,
      "grad_norm": 22.78063201904297,
      "learning_rate": 5.82905982905983e-05,
      "loss": 0.1792,
      "step": 344
    },
    {
      "epoch": 1.2637362637362637,
      "grad_norm": 35.463233947753906,
      "learning_rate": 5.826617826617826e-05,
      "loss": 0.5663,
      "step": 345
    },
    {
      "epoch": 1.2673992673992673,
      "grad_norm": 54.528953552246094,
      "learning_rate": 5.824175824175824e-05,
      "loss": 1.5814,
      "step": 346
    },
    {
      "epoch": 1.271062271062271,
      "grad_norm": 44.60401916503906,
      "learning_rate": 5.821733821733822e-05,
      "loss": 0.6471,
      "step": 347
    },
    {
      "epoch": 1.2747252747252746,
      "grad_norm": 2.6468827724456787,
      "learning_rate": 5.819291819291819e-05,
      "loss": 0.0288,
      "step": 348
    },
    {
      "epoch": 1.2783882783882783,
      "grad_norm": 21.465364456176758,
      "learning_rate": 5.816849816849817e-05,
      "loss": 0.5259,
      "step": 349
    },
    {
      "epoch": 1.282051282051282,
      "grad_norm": 51.20866012573242,
      "learning_rate": 5.814407814407815e-05,
      "loss": 0.8054,
      "step": 350
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 33.52774429321289,
      "learning_rate": 5.811965811965812e-05,
      "loss": 0.494,
      "step": 351
    },
    {
      "epoch": 1.2893772893772895,
      "grad_norm": 39.15644836425781,
      "learning_rate": 5.80952380952381e-05,
      "loss": 1.6315,
      "step": 352
    },
    {
      "epoch": 1.293040293040293,
      "grad_norm": 24.35202407836914,
      "learning_rate": 5.8070818070818076e-05,
      "loss": 0.6189,
      "step": 353
    },
    {
      "epoch": 1.2967032967032968,
      "grad_norm": 39.99496841430664,
      "learning_rate": 5.804639804639805e-05,
      "loss": 1.2323,
      "step": 354
    },
    {
      "epoch": 1.3003663003663004,
      "grad_norm": 26.282432556152344,
      "learning_rate": 5.8021978021978026e-05,
      "loss": 0.5383,
      "step": 355
    },
    {
      "epoch": 1.304029304029304,
      "grad_norm": 36.909969329833984,
      "learning_rate": 5.7997557997558004e-05,
      "loss": 1.6886,
      "step": 356
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 18.90056037902832,
      "learning_rate": 5.7973137973137976e-05,
      "loss": 0.7226,
      "step": 357
    },
    {
      "epoch": 1.3113553113553114,
      "grad_norm": 21.10304832458496,
      "learning_rate": 5.794871794871795e-05,
      "loss": 0.8914,
      "step": 358
    },
    {
      "epoch": 1.315018315018315,
      "grad_norm": 18.380769729614258,
      "learning_rate": 5.7924297924297926e-05,
      "loss": 1.4304,
      "step": 359
    },
    {
      "epoch": 1.3186813186813187,
      "grad_norm": 17.992050170898438,
      "learning_rate": 5.78998778998779e-05,
      "loss": 1.0023,
      "step": 360
    },
    {
      "epoch": 1.3223443223443223,
      "grad_norm": 17.944400787353516,
      "learning_rate": 5.7875457875457876e-05,
      "loss": 0.7734,
      "step": 361
    },
    {
      "epoch": 1.326007326007326,
      "grad_norm": 19.117143630981445,
      "learning_rate": 5.7851037851037855e-05,
      "loss": 0.6923,
      "step": 362
    },
    {
      "epoch": 1.3296703296703296,
      "grad_norm": 21.4644718170166,
      "learning_rate": 5.7826617826617826e-05,
      "loss": 0.666,
      "step": 363
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 25.951030731201172,
      "learning_rate": 5.7802197802197805e-05,
      "loss": 1.522,
      "step": 364
    },
    {
      "epoch": 1.3369963369963371,
      "grad_norm": 32.20412063598633,
      "learning_rate": 5.7777777777777776e-05,
      "loss": 1.5771,
      "step": 365
    },
    {
      "epoch": 1.3406593406593408,
      "grad_norm": 26.847576141357422,
      "learning_rate": 5.7753357753357755e-05,
      "loss": 1.3427,
      "step": 366
    },
    {
      "epoch": 1.3443223443223444,
      "grad_norm": 18.596710205078125,
      "learning_rate": 5.772893772893773e-05,
      "loss": 0.5533,
      "step": 367
    },
    {
      "epoch": 1.347985347985348,
      "grad_norm": 23.6543025970459,
      "learning_rate": 5.7704517704517705e-05,
      "loss": 0.581,
      "step": 368
    },
    {
      "epoch": 1.3516483516483517,
      "grad_norm": 13.732353210449219,
      "learning_rate": 5.7680097680097684e-05,
      "loss": 0.1908,
      "step": 369
    },
    {
      "epoch": 1.3553113553113554,
      "grad_norm": 21.231159210205078,
      "learning_rate": 5.765567765567766e-05,
      "loss": 0.5858,
      "step": 370
    },
    {
      "epoch": 1.358974358974359,
      "grad_norm": 18.647363662719727,
      "learning_rate": 5.763125763125763e-05,
      "loss": 0.6205,
      "step": 371
    },
    {
      "epoch": 1.3626373626373627,
      "grad_norm": 20.302942276000977,
      "learning_rate": 5.7606837606837605e-05,
      "loss": 0.3637,
      "step": 372
    },
    {
      "epoch": 1.3663003663003663,
      "grad_norm": 18.72137451171875,
      "learning_rate": 5.7582417582417584e-05,
      "loss": 0.2262,
      "step": 373
    },
    {
      "epoch": 1.36996336996337,
      "grad_norm": 32.225738525390625,
      "learning_rate": 5.7557997557997555e-05,
      "loss": 0.5696,
      "step": 374
    },
    {
      "epoch": 1.3736263736263736,
      "grad_norm": 21.453779220581055,
      "learning_rate": 5.7533577533577534e-05,
      "loss": 0.3533,
      "step": 375
    },
    {
      "epoch": 1.3772893772893773,
      "grad_norm": 26.601511001586914,
      "learning_rate": 5.750915750915751e-05,
      "loss": 0.438,
      "step": 376
    },
    {
      "epoch": 1.380952380952381,
      "grad_norm": 49.10448455810547,
      "learning_rate": 5.7484737484737484e-05,
      "loss": 0.6742,
      "step": 377
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 51.251136779785156,
      "learning_rate": 5.746031746031746e-05,
      "loss": 0.7096,
      "step": 378
    },
    {
      "epoch": 1.3882783882783882,
      "grad_norm": 35.14614486694336,
      "learning_rate": 5.743589743589744e-05,
      "loss": 1.5348,
      "step": 379
    },
    {
      "epoch": 1.3919413919413919,
      "grad_norm": 58.83134078979492,
      "learning_rate": 5.741147741147741e-05,
      "loss": 1.303,
      "step": 380
    },
    {
      "epoch": 1.3956043956043955,
      "grad_norm": 34.27029800415039,
      "learning_rate": 5.738705738705739e-05,
      "loss": 0.3682,
      "step": 381
    },
    {
      "epoch": 1.3992673992673992,
      "grad_norm": 59.508628845214844,
      "learning_rate": 5.736263736263737e-05,
      "loss": 0.6489,
      "step": 382
    },
    {
      "epoch": 1.4029304029304028,
      "grad_norm": 24.804059982299805,
      "learning_rate": 5.733821733821734e-05,
      "loss": 0.325,
      "step": 383
    },
    {
      "epoch": 1.4065934065934065,
      "grad_norm": 20.69612693786621,
      "learning_rate": 5.731379731379731e-05,
      "loss": 0.1529,
      "step": 384
    },
    {
      "epoch": 1.4102564102564101,
      "grad_norm": 29.134044647216797,
      "learning_rate": 5.728937728937729e-05,
      "loss": 0.8694,
      "step": 385
    },
    {
      "epoch": 1.4139194139194138,
      "grad_norm": 37.44430923461914,
      "learning_rate": 5.726495726495726e-05,
      "loss": 0.9174,
      "step": 386
    },
    {
      "epoch": 1.4175824175824177,
      "grad_norm": 36.84721755981445,
      "learning_rate": 5.724053724053724e-05,
      "loss": 0.3522,
      "step": 387
    },
    {
      "epoch": 1.4212454212454213,
      "grad_norm": 44.15989685058594,
      "learning_rate": 5.721611721611722e-05,
      "loss": 1.4677,
      "step": 388
    },
    {
      "epoch": 1.424908424908425,
      "grad_norm": 16.73012351989746,
      "learning_rate": 5.719169719169719e-05,
      "loss": 0.1621,
      "step": 389
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 35.41815185546875,
      "learning_rate": 5.716727716727717e-05,
      "loss": 0.6702,
      "step": 390
    },
    {
      "epoch": 1.4322344322344323,
      "grad_norm": 19.04936408996582,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.1845,
      "step": 391
    },
    {
      "epoch": 1.435897435897436,
      "grad_norm": 22.89434242248535,
      "learning_rate": 5.711843711843712e-05,
      "loss": 0.5694,
      "step": 392
    },
    {
      "epoch": 1.4395604395604396,
      "grad_norm": 22.125951766967773,
      "learning_rate": 5.70940170940171e-05,
      "loss": 0.821,
      "step": 393
    },
    {
      "epoch": 1.4432234432234432,
      "grad_norm": 37.83376693725586,
      "learning_rate": 5.706959706959707e-05,
      "loss": 0.4658,
      "step": 394
    },
    {
      "epoch": 1.4468864468864469,
      "grad_norm": 38.37764358520508,
      "learning_rate": 5.704517704517705e-05,
      "loss": 0.4146,
      "step": 395
    },
    {
      "epoch": 1.4505494505494505,
      "grad_norm": 21.50092315673828,
      "learning_rate": 5.702075702075703e-05,
      "loss": 0.5044,
      "step": 396
    },
    {
      "epoch": 1.4542124542124542,
      "grad_norm": 20.02173614501953,
      "learning_rate": 5.699633699633699e-05,
      "loss": 0.4955,
      "step": 397
    },
    {
      "epoch": 1.4578754578754578,
      "grad_norm": 21.474336624145508,
      "learning_rate": 5.697191697191697e-05,
      "loss": 0.3818,
      "step": 398
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 22.903839111328125,
      "learning_rate": 5.694749694749695e-05,
      "loss": 0.7603,
      "step": 399
    },
    {
      "epoch": 1.4652014652014653,
      "grad_norm": 20.22893524169922,
      "learning_rate": 5.692307692307692e-05,
      "loss": 0.5612,
      "step": 400
    },
    {
      "epoch": 1.468864468864469,
      "grad_norm": 32.34550857543945,
      "learning_rate": 5.68986568986569e-05,
      "loss": 0.4659,
      "step": 401
    },
    {
      "epoch": 1.4725274725274726,
      "grad_norm": 49.979034423828125,
      "learning_rate": 5.687423687423688e-05,
      "loss": 0.6784,
      "step": 402
    },
    {
      "epoch": 1.4761904761904763,
      "grad_norm": 79.79581451416016,
      "learning_rate": 5.684981684981685e-05,
      "loss": 0.9404,
      "step": 403
    },
    {
      "epoch": 1.47985347985348,
      "grad_norm": 17.678560256958008,
      "learning_rate": 5.682539682539683e-05,
      "loss": 0.1675,
      "step": 404
    },
    {
      "epoch": 1.4835164835164836,
      "grad_norm": 21.246519088745117,
      "learning_rate": 5.6800976800976806e-05,
      "loss": 0.2428,
      "step": 405
    },
    {
      "epoch": 1.4871794871794872,
      "grad_norm": 34.815452575683594,
      "learning_rate": 5.677655677655678e-05,
      "loss": 0.3925,
      "step": 406
    },
    {
      "epoch": 1.4908424908424909,
      "grad_norm": 73.8591079711914,
      "learning_rate": 5.6752136752136756e-05,
      "loss": 1.3163,
      "step": 407
    },
    {
      "epoch": 1.4945054945054945,
      "grad_norm": 66.63922882080078,
      "learning_rate": 5.6727716727716735e-05,
      "loss": 0.9653,
      "step": 408
    },
    {
      "epoch": 1.4981684981684982,
      "grad_norm": 52.39488220214844,
      "learning_rate": 5.6703296703296706e-05,
      "loss": 0.9322,
      "step": 409
    },
    {
      "epoch": 1.5018315018315018,
      "grad_norm": 13.078998565673828,
      "learning_rate": 5.667887667887668e-05,
      "loss": 0.1168,
      "step": 410
    },
    {
      "epoch": 1.5054945054945055,
      "grad_norm": 41.32448959350586,
      "learning_rate": 5.6654456654456657e-05,
      "loss": 0.9296,
      "step": 411
    },
    {
      "epoch": 1.5091575091575091,
      "grad_norm": 26.448543548583984,
      "learning_rate": 5.663003663003663e-05,
      "loss": 0.5474,
      "step": 412
    },
    {
      "epoch": 1.5128205128205128,
      "grad_norm": 29.58432960510254,
      "learning_rate": 5.660561660561661e-05,
      "loss": 0.6573,
      "step": 413
    },
    {
      "epoch": 1.5164835164835164,
      "grad_norm": 28.568214416503906,
      "learning_rate": 5.6581196581196585e-05,
      "loss": 0.9223,
      "step": 414
    },
    {
      "epoch": 1.52014652014652,
      "grad_norm": 31.92661476135254,
      "learning_rate": 5.655677655677656e-05,
      "loss": 1.0601,
      "step": 415
    },
    {
      "epoch": 1.5238095238095237,
      "grad_norm": 31.934263229370117,
      "learning_rate": 5.6532356532356535e-05,
      "loss": 0.6288,
      "step": 416
    },
    {
      "epoch": 1.5274725274725274,
      "grad_norm": 21.51350975036621,
      "learning_rate": 5.650793650793651e-05,
      "loss": 0.7378,
      "step": 417
    },
    {
      "epoch": 1.531135531135531,
      "grad_norm": 19.010095596313477,
      "learning_rate": 5.6483516483516485e-05,
      "loss": 0.7792,
      "step": 418
    },
    {
      "epoch": 1.5347985347985347,
      "grad_norm": 21.7001895904541,
      "learning_rate": 5.6459096459096464e-05,
      "loss": 0.7885,
      "step": 419
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 21.400882720947266,
      "learning_rate": 5.6434676434676436e-05,
      "loss": 0.942,
      "step": 420
    },
    {
      "epoch": 1.542124542124542,
      "grad_norm": 30.14664649963379,
      "learning_rate": 5.6410256410256414e-05,
      "loss": 0.7675,
      "step": 421
    },
    {
      "epoch": 1.5457875457875456,
      "grad_norm": 33.25088882446289,
      "learning_rate": 5.6385836385836386e-05,
      "loss": 1.1349,
      "step": 422
    },
    {
      "epoch": 1.5494505494505495,
      "grad_norm": 22.923208236694336,
      "learning_rate": 5.636141636141636e-05,
      "loss": 0.7145,
      "step": 423
    },
    {
      "epoch": 1.5531135531135531,
      "grad_norm": 20.00519371032715,
      "learning_rate": 5.6336996336996336e-05,
      "loss": 0.5107,
      "step": 424
    },
    {
      "epoch": 1.5567765567765568,
      "grad_norm": 21.95383071899414,
      "learning_rate": 5.6312576312576314e-05,
      "loss": 0.7836,
      "step": 425
    },
    {
      "epoch": 1.5604395604395604,
      "grad_norm": 27.24031639099121,
      "learning_rate": 5.6288156288156286e-05,
      "loss": 0.4955,
      "step": 426
    },
    {
      "epoch": 1.564102564102564,
      "grad_norm": 45.48428726196289,
      "learning_rate": 5.6263736263736264e-05,
      "loss": 1.016,
      "step": 427
    },
    {
      "epoch": 1.5677655677655677,
      "grad_norm": 20.055965423583984,
      "learning_rate": 5.623931623931624e-05,
      "loss": 0.325,
      "step": 428
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 22.020767211914062,
      "learning_rate": 5.6214896214896215e-05,
      "loss": 0.45,
      "step": 429
    },
    {
      "epoch": 1.575091575091575,
      "grad_norm": 32.608741760253906,
      "learning_rate": 5.619047619047619e-05,
      "loss": 0.6561,
      "step": 430
    },
    {
      "epoch": 1.578754578754579,
      "grad_norm": 38.14396667480469,
      "learning_rate": 5.616605616605617e-05,
      "loss": 0.6387,
      "step": 431
    },
    {
      "epoch": 1.5824175824175826,
      "grad_norm": 26.266948699951172,
      "learning_rate": 5.614163614163614e-05,
      "loss": 0.5593,
      "step": 432
    },
    {
      "epoch": 1.5860805860805862,
      "grad_norm": 16.37360954284668,
      "learning_rate": 5.611721611721612e-05,
      "loss": 0.1591,
      "step": 433
    },
    {
      "epoch": 1.5897435897435899,
      "grad_norm": 21.9448299407959,
      "learning_rate": 5.60927960927961e-05,
      "loss": 0.2129,
      "step": 434
    },
    {
      "epoch": 1.5934065934065935,
      "grad_norm": 30.096052169799805,
      "learning_rate": 5.6068376068376065e-05,
      "loss": 0.3384,
      "step": 435
    },
    {
      "epoch": 1.5970695970695972,
      "grad_norm": 40.15864181518555,
      "learning_rate": 5.604395604395604e-05,
      "loss": 0.5181,
      "step": 436
    },
    {
      "epoch": 1.6007326007326008,
      "grad_norm": 63.40933609008789,
      "learning_rate": 5.601953601953602e-05,
      "loss": 0.8834,
      "step": 437
    },
    {
      "epoch": 1.6043956043956045,
      "grad_norm": 40.0787353515625,
      "learning_rate": 5.5995115995115993e-05,
      "loss": 0.437,
      "step": 438
    },
    {
      "epoch": 1.6080586080586081,
      "grad_norm": 40.136863708496094,
      "learning_rate": 5.597069597069597e-05,
      "loss": 0.4834,
      "step": 439
    },
    {
      "epoch": 1.6117216117216118,
      "grad_norm": 27.898317337036133,
      "learning_rate": 5.594627594627595e-05,
      "loss": 0.4862,
      "step": 440
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 31.5762882232666,
      "learning_rate": 5.592185592185592e-05,
      "loss": 0.1878,
      "step": 441
    },
    {
      "epoch": 1.619047619047619,
      "grad_norm": 88.90093994140625,
      "learning_rate": 5.58974358974359e-05,
      "loss": 1.3343,
      "step": 442
    },
    {
      "epoch": 1.6227106227106227,
      "grad_norm": 57.7340202331543,
      "learning_rate": 5.587301587301587e-05,
      "loss": 0.3032,
      "step": 443
    },
    {
      "epoch": 1.6263736263736264,
      "grad_norm": 57.28425979614258,
      "learning_rate": 5.584859584859585e-05,
      "loss": 1.3972,
      "step": 444
    },
    {
      "epoch": 1.63003663003663,
      "grad_norm": 39.866302490234375,
      "learning_rate": 5.582417582417583e-05,
      "loss": 0.4026,
      "step": 445
    },
    {
      "epoch": 1.6336996336996337,
      "grad_norm": 41.72932815551758,
      "learning_rate": 5.57997557997558e-05,
      "loss": 0.5407,
      "step": 446
    },
    {
      "epoch": 1.6373626373626373,
      "grad_norm": 60.77634811401367,
      "learning_rate": 5.577533577533578e-05,
      "loss": 0.8581,
      "step": 447
    },
    {
      "epoch": 1.641025641025641,
      "grad_norm": 28.382030487060547,
      "learning_rate": 5.575091575091575e-05,
      "loss": 0.3759,
      "step": 448
    },
    {
      "epoch": 1.6446886446886446,
      "grad_norm": 62.1085205078125,
      "learning_rate": 5.572649572649572e-05,
      "loss": 1.0749,
      "step": 449
    },
    {
      "epoch": 1.6483516483516483,
      "grad_norm": 41.8302001953125,
      "learning_rate": 5.57020757020757e-05,
      "loss": 0.5884,
      "step": 450
    },
    {
      "epoch": 1.652014652014652,
      "grad_norm": 24.128931045532227,
      "learning_rate": 5.567765567765568e-05,
      "loss": 0.6113,
      "step": 451
    },
    {
      "epoch": 1.6556776556776556,
      "grad_norm": 19.634384155273438,
      "learning_rate": 5.565323565323565e-05,
      "loss": 0.3902,
      "step": 452
    },
    {
      "epoch": 1.6593406593406592,
      "grad_norm": 18.17875099182129,
      "learning_rate": 5.562881562881563e-05,
      "loss": 0.3137,
      "step": 453
    },
    {
      "epoch": 1.6630036630036629,
      "grad_norm": 39.68446731567383,
      "learning_rate": 5.560439560439561e-05,
      "loss": 0.7587,
      "step": 454
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 29.387836456298828,
      "learning_rate": 5.557997557997558e-05,
      "loss": 0.6397,
      "step": 455
    },
    {
      "epoch": 1.6703296703296702,
      "grad_norm": 19.08424949645996,
      "learning_rate": 5.555555555555556e-05,
      "loss": 0.2484,
      "step": 456
    },
    {
      "epoch": 1.673992673992674,
      "grad_norm": 36.07701873779297,
      "learning_rate": 5.553113553113554e-05,
      "loss": 0.8587,
      "step": 457
    },
    {
      "epoch": 1.6776556776556777,
      "grad_norm": 52.062339782714844,
      "learning_rate": 5.550671550671551e-05,
      "loss": 1.6675,
      "step": 458
    },
    {
      "epoch": 1.6813186813186813,
      "grad_norm": 45.415687561035156,
      "learning_rate": 5.548229548229549e-05,
      "loss": 1.653,
      "step": 459
    },
    {
      "epoch": 1.684981684981685,
      "grad_norm": 31.457420349121094,
      "learning_rate": 5.5457875457875465e-05,
      "loss": 0.4578,
      "step": 460
    },
    {
      "epoch": 1.6886446886446886,
      "grad_norm": 33.14665603637695,
      "learning_rate": 5.543345543345543e-05,
      "loss": 1.3327,
      "step": 461
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 25.720529556274414,
      "learning_rate": 5.540903540903541e-05,
      "loss": 0.5,
      "step": 462
    },
    {
      "epoch": 1.695970695970696,
      "grad_norm": 23.71514129638672,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.434,
      "step": 463
    },
    {
      "epoch": 1.6996336996336996,
      "grad_norm": 45.231746673583984,
      "learning_rate": 5.536019536019536e-05,
      "loss": 0.9448,
      "step": 464
    },
    {
      "epoch": 1.7032967032967035,
      "grad_norm": 17.44647789001465,
      "learning_rate": 5.533577533577534e-05,
      "loss": 0.3183,
      "step": 465
    },
    {
      "epoch": 1.7069597069597071,
      "grad_norm": 18.627901077270508,
      "learning_rate": 5.531135531135531e-05,
      "loss": 0.4137,
      "step": 466
    },
    {
      "epoch": 1.7106227106227108,
      "grad_norm": 45.57220458984375,
      "learning_rate": 5.528693528693529e-05,
      "loss": 1.0096,
      "step": 467
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 27.329822540283203,
      "learning_rate": 5.5262515262515266e-05,
      "loss": 0.5416,
      "step": 468
    },
    {
      "epoch": 1.717948717948718,
      "grad_norm": 46.70027160644531,
      "learning_rate": 5.523809523809524e-05,
      "loss": 0.983,
      "step": 469
    },
    {
      "epoch": 1.7216117216117217,
      "grad_norm": 32.47868728637695,
      "learning_rate": 5.5213675213675216e-05,
      "loss": 1.5687,
      "step": 470
    },
    {
      "epoch": 1.7252747252747254,
      "grad_norm": 16.49342155456543,
      "learning_rate": 5.5189255189255194e-05,
      "loss": 0.3101,
      "step": 471
    },
    {
      "epoch": 1.728937728937729,
      "grad_norm": 26.58381462097168,
      "learning_rate": 5.5164835164835166e-05,
      "loss": 0.7027,
      "step": 472
    },
    {
      "epoch": 1.7326007326007327,
      "grad_norm": 17.435213088989258,
      "learning_rate": 5.5140415140415144e-05,
      "loss": 0.3958,
      "step": 473
    },
    {
      "epoch": 1.7362637362637363,
      "grad_norm": 19.37874412536621,
      "learning_rate": 5.5115995115995116e-05,
      "loss": 0.3979,
      "step": 474
    },
    {
      "epoch": 1.73992673992674,
      "grad_norm": 16.509248733520508,
      "learning_rate": 5.509157509157509e-05,
      "loss": 0.5121,
      "step": 475
    },
    {
      "epoch": 1.7435897435897436,
      "grad_norm": 9.653852462768555,
      "learning_rate": 5.5067155067155066e-05,
      "loss": 0.1386,
      "step": 476
    },
    {
      "epoch": 1.7472527472527473,
      "grad_norm": 26.486963272094727,
      "learning_rate": 5.5042735042735045e-05,
      "loss": 1.0307,
      "step": 477
    },
    {
      "epoch": 1.750915750915751,
      "grad_norm": 17.766828536987305,
      "learning_rate": 5.5018315018315016e-05,
      "loss": 0.278,
      "step": 478
    },
    {
      "epoch": 1.7545787545787546,
      "grad_norm": 12.930633544921875,
      "learning_rate": 5.4993894993894995e-05,
      "loss": 0.1487,
      "step": 479
    },
    {
      "epoch": 1.7582417582417582,
      "grad_norm": 44.64267349243164,
      "learning_rate": 5.496947496947497e-05,
      "loss": 0.7036,
      "step": 480
    },
    {
      "epoch": 1.7619047619047619,
      "grad_norm": 17.474651336669922,
      "learning_rate": 5.4945054945054945e-05,
      "loss": 0.1666,
      "step": 481
    },
    {
      "epoch": 1.7655677655677655,
      "grad_norm": 48.3519401550293,
      "learning_rate": 5.4920634920634923e-05,
      "loss": 0.6157,
      "step": 482
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 18.429521560668945,
      "learning_rate": 5.48962148962149e-05,
      "loss": 0.2588,
      "step": 483
    },
    {
      "epoch": 1.7728937728937728,
      "grad_norm": 66.73760986328125,
      "learning_rate": 5.4871794871794874e-05,
      "loss": 0.654,
      "step": 484
    },
    {
      "epoch": 1.7765567765567765,
      "grad_norm": 53.831539154052734,
      "learning_rate": 5.484737484737485e-05,
      "loss": 0.7538,
      "step": 485
    },
    {
      "epoch": 1.7802197802197801,
      "grad_norm": 52.023895263671875,
      "learning_rate": 5.482295482295483e-05,
      "loss": 1.6623,
      "step": 486
    },
    {
      "epoch": 1.7838827838827838,
      "grad_norm": 38.4475212097168,
      "learning_rate": 5.4798534798534795e-05,
      "loss": 0.5079,
      "step": 487
    },
    {
      "epoch": 1.7875457875457874,
      "grad_norm": 25.642650604248047,
      "learning_rate": 5.4774114774114774e-05,
      "loss": 0.3825,
      "step": 488
    },
    {
      "epoch": 1.791208791208791,
      "grad_norm": 57.916900634765625,
      "learning_rate": 5.474969474969475e-05,
      "loss": 0.9583,
      "step": 489
    },
    {
      "epoch": 1.7948717948717947,
      "grad_norm": 39.23340606689453,
      "learning_rate": 5.4725274725274724e-05,
      "loss": 0.4724,
      "step": 490
    },
    {
      "epoch": 1.7985347985347986,
      "grad_norm": 24.188661575317383,
      "learning_rate": 5.47008547008547e-05,
      "loss": 0.4471,
      "step": 491
    },
    {
      "epoch": 1.8021978021978022,
      "grad_norm": 68.73822021484375,
      "learning_rate": 5.4676434676434674e-05,
      "loss": 0.6618,
      "step": 492
    },
    {
      "epoch": 1.8058608058608059,
      "grad_norm": 26.382184982299805,
      "learning_rate": 5.465201465201465e-05,
      "loss": 0.5835,
      "step": 493
    },
    {
      "epoch": 1.8095238095238095,
      "grad_norm": 31.758886337280273,
      "learning_rate": 5.462759462759463e-05,
      "loss": 0.622,
      "step": 494
    },
    {
      "epoch": 1.8131868131868132,
      "grad_norm": 26.657405853271484,
      "learning_rate": 5.46031746031746e-05,
      "loss": 0.6003,
      "step": 495
    },
    {
      "epoch": 1.8168498168498168,
      "grad_norm": 31.248491287231445,
      "learning_rate": 5.457875457875458e-05,
      "loss": 0.4929,
      "step": 496
    },
    {
      "epoch": 1.8205128205128205,
      "grad_norm": 53.82766342163086,
      "learning_rate": 5.455433455433456e-05,
      "loss": 2.0716,
      "step": 497
    },
    {
      "epoch": 1.8241758241758241,
      "grad_norm": 46.39777374267578,
      "learning_rate": 5.452991452991453e-05,
      "loss": 1.6767,
      "step": 498
    },
    {
      "epoch": 1.8278388278388278,
      "grad_norm": 39.58620071411133,
      "learning_rate": 5.450549450549451e-05,
      "loss": 0.8274,
      "step": 499
    },
    {
      "epoch": 1.8315018315018317,
      "grad_norm": 29.395286560058594,
      "learning_rate": 5.448107448107448e-05,
      "loss": 1.1441,
      "step": 500
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 2730,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "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": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}