{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.4163855421686746,
  "eval_steps": 500,
  "global_step": 368,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0038554216867469878,
      "grad_norm": 0.6997007131576538,
      "learning_rate": 0.0,
      "loss": 0.0405,
      "step": 1
    },
    {
      "epoch": 0.0077108433734939755,
      "grad_norm": 0.6004841327667236,
      "learning_rate": 5.128205128205128e-07,
      "loss": 0.0413,
      "step": 2
    },
    {
      "epoch": 0.011566265060240964,
      "grad_norm": 0.6796688437461853,
      "learning_rate": 1.0256410256410257e-06,
      "loss": 0.049,
      "step": 3
    },
    {
      "epoch": 0.015421686746987951,
      "grad_norm": 0.6812688708305359,
      "learning_rate": 1.5384615384615387e-06,
      "loss": 0.0486,
      "step": 4
    },
    {
      "epoch": 0.01927710843373494,
      "grad_norm": 0.6957109570503235,
      "learning_rate": 2.0512820512820513e-06,
      "loss": 0.033,
      "step": 5
    },
    {
      "epoch": 0.02313253012048193,
      "grad_norm": 0.6179328560829163,
      "learning_rate": 2.564102564102564e-06,
      "loss": 0.0505,
      "step": 6
    },
    {
      "epoch": 0.026987951807228915,
      "grad_norm": 0.5707671046257019,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 0.0409,
      "step": 7
    },
    {
      "epoch": 0.030843373493975902,
      "grad_norm": 0.41436561942100525,
      "learning_rate": 3.58974358974359e-06,
      "loss": 0.0433,
      "step": 8
    },
    {
      "epoch": 0.03469879518072289,
      "grad_norm": 0.3194092810153961,
      "learning_rate": 4.102564102564103e-06,
      "loss": 0.032,
      "step": 9
    },
    {
      "epoch": 0.03855421686746988,
      "grad_norm": 0.3646799325942993,
      "learning_rate": 4.615384615384616e-06,
      "loss": 0.0326,
      "step": 10
    },
    {
      "epoch": 0.042409638554216866,
      "grad_norm": 0.3623705804347992,
      "learning_rate": 5.128205128205128e-06,
      "loss": 0.0415,
      "step": 11
    },
    {
      "epoch": 0.04626506024096386,
      "grad_norm": 0.23459671437740326,
      "learning_rate": 5.641025641025641e-06,
      "loss": 0.0356,
      "step": 12
    },
    {
      "epoch": 0.05012048192771084,
      "grad_norm": 0.27805373072624207,
      "learning_rate": 6.153846153846155e-06,
      "loss": 0.0265,
      "step": 13
    },
    {
      "epoch": 0.05397590361445783,
      "grad_norm": 0.2236052304506302,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.0237,
      "step": 14
    },
    {
      "epoch": 0.05783132530120482,
      "grad_norm": 0.2813984453678131,
      "learning_rate": 7.17948717948718e-06,
      "loss": 0.0271,
      "step": 15
    },
    {
      "epoch": 0.061686746987951804,
      "grad_norm": 0.24619753658771515,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.0237,
      "step": 16
    },
    {
      "epoch": 0.0655421686746988,
      "grad_norm": 0.2730517089366913,
      "learning_rate": 8.205128205128205e-06,
      "loss": 0.0227,
      "step": 17
    },
    {
      "epoch": 0.06939759036144579,
      "grad_norm": 0.2350538820028305,
      "learning_rate": 8.717948717948719e-06,
      "loss": 0.0245,
      "step": 18
    },
    {
      "epoch": 0.07325301204819278,
      "grad_norm": 0.18193207681179047,
      "learning_rate": 9.230769230769232e-06,
      "loss": 0.0168,
      "step": 19
    },
    {
      "epoch": 0.07710843373493977,
      "grad_norm": 0.18165752291679382,
      "learning_rate": 9.743589743589744e-06,
      "loss": 0.0159,
      "step": 20
    },
    {
      "epoch": 0.08096385542168674,
      "grad_norm": 0.27797478437423706,
      "learning_rate": 1.0256410256410256e-05,
      "loss": 0.0301,
      "step": 21
    },
    {
      "epoch": 0.08481927710843373,
      "grad_norm": 0.26159757375717163,
      "learning_rate": 1.076923076923077e-05,
      "loss": 0.0279,
      "step": 22
    },
    {
      "epoch": 0.08867469879518072,
      "grad_norm": 0.20290671288967133,
      "learning_rate": 1.1282051282051283e-05,
      "loss": 0.0215,
      "step": 23
    },
    {
      "epoch": 0.09253012048192771,
      "grad_norm": 0.3016839921474457,
      "learning_rate": 1.1794871794871796e-05,
      "loss": 0.0239,
      "step": 24
    },
    {
      "epoch": 0.0963855421686747,
      "grad_norm": 0.23746410012245178,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.0287,
      "step": 25
    },
    {
      "epoch": 0.10024096385542168,
      "grad_norm": 0.20524649322032928,
      "learning_rate": 1.2820512820512823e-05,
      "loss": 0.0233,
      "step": 26
    },
    {
      "epoch": 0.10409638554216867,
      "grad_norm": 0.23074588179588318,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.0227,
      "step": 27
    },
    {
      "epoch": 0.10795180722891566,
      "grad_norm": 0.17793533205986023,
      "learning_rate": 1.3846153846153847e-05,
      "loss": 0.0206,
      "step": 28
    },
    {
      "epoch": 0.11180722891566265,
      "grad_norm": 0.19907526671886444,
      "learning_rate": 1.435897435897436e-05,
      "loss": 0.0291,
      "step": 29
    },
    {
      "epoch": 0.11566265060240964,
      "grad_norm": 0.22600896656513214,
      "learning_rate": 1.4871794871794874e-05,
      "loss": 0.0277,
      "step": 30
    },
    {
      "epoch": 0.11951807228915663,
      "grad_norm": 0.2135571539402008,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.0157,
      "step": 31
    },
    {
      "epoch": 0.12337349397590361,
      "grad_norm": 0.17844079434871674,
      "learning_rate": 1.5897435897435897e-05,
      "loss": 0.0207,
      "step": 32
    },
    {
      "epoch": 0.1272289156626506,
      "grad_norm": 0.21249860525131226,
      "learning_rate": 1.641025641025641e-05,
      "loss": 0.0258,
      "step": 33
    },
    {
      "epoch": 0.1310843373493976,
      "grad_norm": 0.18014496564865112,
      "learning_rate": 1.6923076923076924e-05,
      "loss": 0.0187,
      "step": 34
    },
    {
      "epoch": 0.13493975903614458,
      "grad_norm": 0.1790035516023636,
      "learning_rate": 1.7435897435897438e-05,
      "loss": 0.0141,
      "step": 35
    },
    {
      "epoch": 0.13879518072289157,
      "grad_norm": 0.18319612741470337,
      "learning_rate": 1.794871794871795e-05,
      "loss": 0.0153,
      "step": 36
    },
    {
      "epoch": 0.14265060240963856,
      "grad_norm": 0.16664963960647583,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.0146,
      "step": 37
    },
    {
      "epoch": 0.14650602409638555,
      "grad_norm": 0.17741602659225464,
      "learning_rate": 1.8974358974358975e-05,
      "loss": 0.0138,
      "step": 38
    },
    {
      "epoch": 0.15036144578313254,
      "grad_norm": 0.16826853156089783,
      "learning_rate": 1.9487179487179488e-05,
      "loss": 0.0149,
      "step": 39
    },
    {
      "epoch": 0.15421686746987953,
      "grad_norm": 0.17632468044757843,
      "learning_rate": 2e-05,
      "loss": 0.0315,
      "step": 40
    },
    {
      "epoch": 0.1580722891566265,
      "grad_norm": 0.21154020726680756,
      "learning_rate": 2.0512820512820512e-05,
      "loss": 0.0313,
      "step": 41
    },
    {
      "epoch": 0.16192771084337348,
      "grad_norm": 0.17007827758789062,
      "learning_rate": 2.102564102564103e-05,
      "loss": 0.0283,
      "step": 42
    },
    {
      "epoch": 0.16578313253012048,
      "grad_norm": 0.20888379216194153,
      "learning_rate": 2.153846153846154e-05,
      "loss": 0.0173,
      "step": 43
    },
    {
      "epoch": 0.16963855421686747,
      "grad_norm": 0.17302851378917694,
      "learning_rate": 2.2051282051282056e-05,
      "loss": 0.023,
      "step": 44
    },
    {
      "epoch": 0.17349397590361446,
      "grad_norm": 0.28571709990501404,
      "learning_rate": 2.2564102564102566e-05,
      "loss": 0.0528,
      "step": 45
    },
    {
      "epoch": 0.17734939759036145,
      "grad_norm": 0.2224675714969635,
      "learning_rate": 2.3076923076923076e-05,
      "loss": 0.0326,
      "step": 46
    },
    {
      "epoch": 0.18120481927710844,
      "grad_norm": 0.16989563405513763,
      "learning_rate": 2.3589743589743593e-05,
      "loss": 0.0106,
      "step": 47
    },
    {
      "epoch": 0.18506024096385543,
      "grad_norm": 0.17279551923274994,
      "learning_rate": 2.4102564102564103e-05,
      "loss": 0.0143,
      "step": 48
    },
    {
      "epoch": 0.18891566265060242,
      "grad_norm": 0.1571756899356842,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.0112,
      "step": 49
    },
    {
      "epoch": 0.1927710843373494,
      "grad_norm": 0.21857869625091553,
      "learning_rate": 2.512820512820513e-05,
      "loss": 0.0149,
      "step": 50
    },
    {
      "epoch": 0.1966265060240964,
      "grad_norm": 0.21509476006031036,
      "learning_rate": 2.5641025641025646e-05,
      "loss": 0.0248,
      "step": 51
    },
    {
      "epoch": 0.20048192771084336,
      "grad_norm": 0.15649166703224182,
      "learning_rate": 2.6153846153846157e-05,
      "loss": 0.0101,
      "step": 52
    },
    {
      "epoch": 0.20433734939759035,
      "grad_norm": 0.17392459511756897,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.0115,
      "step": 53
    },
    {
      "epoch": 0.20819277108433734,
      "grad_norm": 0.13562685251235962,
      "learning_rate": 2.7179487179487183e-05,
      "loss": 0.0102,
      "step": 54
    },
    {
      "epoch": 0.21204819277108433,
      "grad_norm": 0.22402575612068176,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.0164,
      "step": 55
    },
    {
      "epoch": 0.21590361445783132,
      "grad_norm": 0.1317652314901352,
      "learning_rate": 2.820512820512821e-05,
      "loss": 0.0091,
      "step": 56
    },
    {
      "epoch": 0.2197590361445783,
      "grad_norm": 0.11563073843717575,
      "learning_rate": 2.871794871794872e-05,
      "loss": 0.009,
      "step": 57
    },
    {
      "epoch": 0.2236144578313253,
      "grad_norm": 0.12244722992181778,
      "learning_rate": 2.923076923076923e-05,
      "loss": 0.0077,
      "step": 58
    },
    {
      "epoch": 0.2274698795180723,
      "grad_norm": 0.18595360219478607,
      "learning_rate": 2.9743589743589747e-05,
      "loss": 0.0118,
      "step": 59
    },
    {
      "epoch": 0.23132530120481928,
      "grad_norm": 0.16396519541740417,
      "learning_rate": 3.0256410256410257e-05,
      "loss": 0.0113,
      "step": 60
    },
    {
      "epoch": 0.23518072289156627,
      "grad_norm": 0.16600239276885986,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.0273,
      "step": 61
    },
    {
      "epoch": 0.23903614457831326,
      "grad_norm": 0.14368009567260742,
      "learning_rate": 3.128205128205129e-05,
      "loss": 0.0107,
      "step": 62
    },
    {
      "epoch": 0.24289156626506025,
      "grad_norm": 0.11151304095983505,
      "learning_rate": 3.1794871794871795e-05,
      "loss": 0.0114,
      "step": 63
    },
    {
      "epoch": 0.24674698795180722,
      "grad_norm": 0.16931672394275665,
      "learning_rate": 3.230769230769231e-05,
      "loss": 0.0101,
      "step": 64
    },
    {
      "epoch": 0.25060240963855424,
      "grad_norm": 0.13678501546382904,
      "learning_rate": 3.282051282051282e-05,
      "loss": 0.0074,
      "step": 65
    },
    {
      "epoch": 0.2544578313253012,
      "grad_norm": 0.12231199443340302,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0117,
      "step": 66
    },
    {
      "epoch": 0.2583132530120482,
      "grad_norm": 0.2577515244483948,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0304,
      "step": 67
    },
    {
      "epoch": 0.2621686746987952,
      "grad_norm": 0.10441906005144119,
      "learning_rate": 3.435897435897436e-05,
      "loss": 0.0077,
      "step": 68
    },
    {
      "epoch": 0.26602409638554214,
      "grad_norm": 0.18340735137462616,
      "learning_rate": 3.4871794871794875e-05,
      "loss": 0.0131,
      "step": 69
    },
    {
      "epoch": 0.26987951807228916,
      "grad_norm": 0.26318177580833435,
      "learning_rate": 3.538461538461539e-05,
      "loss": 0.0166,
      "step": 70
    },
    {
      "epoch": 0.2737349397590361,
      "grad_norm": 0.0868140459060669,
      "learning_rate": 3.58974358974359e-05,
      "loss": 0.0079,
      "step": 71
    },
    {
      "epoch": 0.27759036144578314,
      "grad_norm": 0.1882234513759613,
      "learning_rate": 3.6410256410256416e-05,
      "loss": 0.0113,
      "step": 72
    },
    {
      "epoch": 0.2814457831325301,
      "grad_norm": 0.1313207447528839,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.01,
      "step": 73
    },
    {
      "epoch": 0.2853012048192771,
      "grad_norm": 0.21643561124801636,
      "learning_rate": 3.7435897435897436e-05,
      "loss": 0.0126,
      "step": 74
    },
    {
      "epoch": 0.2891566265060241,
      "grad_norm": 0.35117974877357483,
      "learning_rate": 3.794871794871795e-05,
      "loss": 0.0322,
      "step": 75
    },
    {
      "epoch": 0.2930120481927711,
      "grad_norm": 0.14641538262367249,
      "learning_rate": 3.846153846153846e-05,
      "loss": 0.0086,
      "step": 76
    },
    {
      "epoch": 0.29686746987951806,
      "grad_norm": 0.18411016464233398,
      "learning_rate": 3.8974358974358976e-05,
      "loss": 0.0254,
      "step": 77
    },
    {
      "epoch": 0.3007228915662651,
      "grad_norm": 0.17950740456581116,
      "learning_rate": 3.948717948717949e-05,
      "loss": 0.0191,
      "step": 78
    },
    {
      "epoch": 0.30457831325301205,
      "grad_norm": 0.14514976739883423,
      "learning_rate": 4e-05,
      "loss": 0.0107,
      "step": 79
    },
    {
      "epoch": 0.30843373493975906,
      "grad_norm": 0.22098544239997864,
      "learning_rate": 3.999979972589652e-05,
      "loss": 0.0176,
      "step": 80
    },
    {
      "epoch": 0.312289156626506,
      "grad_norm": 0.11959940195083618,
      "learning_rate": 3.9999198907597046e-05,
      "loss": 0.0168,
      "step": 81
    },
    {
      "epoch": 0.316144578313253,
      "grad_norm": 0.13479092717170715,
      "learning_rate": 3.999819755713442e-05,
      "loss": 0.0246,
      "step": 82
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.15680080652236938,
      "learning_rate": 3.9996795694563096e-05,
      "loss": 0.0147,
      "step": 83
    },
    {
      "epoch": 0.32385542168674697,
      "grad_norm": 0.1106104925274849,
      "learning_rate": 3.999499334795875e-05,
      "loss": 0.017,
      "step": 84
    },
    {
      "epoch": 0.327710843373494,
      "grad_norm": 0.13254176080226898,
      "learning_rate": 3.999279055341771e-05,
      "loss": 0.0096,
      "step": 85
    },
    {
      "epoch": 0.33156626506024095,
      "grad_norm": 0.1391618400812149,
      "learning_rate": 3.999018735505626e-05,
      "loss": 0.0165,
      "step": 86
    },
    {
      "epoch": 0.33542168674698797,
      "grad_norm": 0.15080322325229645,
      "learning_rate": 3.998718380500971e-05,
      "loss": 0.0127,
      "step": 87
    },
    {
      "epoch": 0.33927710843373493,
      "grad_norm": 0.20322079956531525,
      "learning_rate": 3.998377996343139e-05,
      "loss": 0.053,
      "step": 88
    },
    {
      "epoch": 0.34313253012048195,
      "grad_norm": 0.10287915915250778,
      "learning_rate": 3.997997589849145e-05,
      "loss": 0.0112,
      "step": 89
    },
    {
      "epoch": 0.3469879518072289,
      "grad_norm": 0.14580729603767395,
      "learning_rate": 3.9975771686375434e-05,
      "loss": 0.009,
      "step": 90
    },
    {
      "epoch": 0.35084337349397593,
      "grad_norm": 0.16750818490982056,
      "learning_rate": 3.9971167411282835e-05,
      "loss": 0.0137,
      "step": 91
    },
    {
      "epoch": 0.3546987951807229,
      "grad_norm": 0.0841616690158844,
      "learning_rate": 3.996616316542537e-05,
      "loss": 0.0054,
      "step": 92
    },
    {
      "epoch": 0.35855421686746985,
      "grad_norm": 0.20032115280628204,
      "learning_rate": 3.99607590490251e-05,
      "loss": 0.0138,
      "step": 93
    },
    {
      "epoch": 0.3624096385542169,
      "grad_norm": 0.20202049612998962,
      "learning_rate": 3.9954955170312504e-05,
      "loss": 0.0128,
      "step": 94
    },
    {
      "epoch": 0.36626506024096384,
      "grad_norm": 0.10456931591033936,
      "learning_rate": 3.9948751645524235e-05,
      "loss": 0.0092,
      "step": 95
    },
    {
      "epoch": 0.37012048192771085,
      "grad_norm": 0.403626948595047,
      "learning_rate": 3.994214859890083e-05,
      "loss": 0.0206,
      "step": 96
    },
    {
      "epoch": 0.3739759036144578,
      "grad_norm": 0.11969339102506638,
      "learning_rate": 3.9935146162684206e-05,
      "loss": 0.0101,
      "step": 97
    },
    {
      "epoch": 0.37783132530120483,
      "grad_norm": 0.3118373453617096,
      "learning_rate": 3.992774447711503e-05,
      "loss": 0.0273,
      "step": 98
    },
    {
      "epoch": 0.3816867469879518,
      "grad_norm": 0.2649889588356018,
      "learning_rate": 3.9919943690429906e-05,
      "loss": 0.0386,
      "step": 99
    },
    {
      "epoch": 0.3855421686746988,
      "grad_norm": 0.12110533565282822,
      "learning_rate": 3.9911743958858376e-05,
      "loss": 0.0144,
      "step": 100
    },
    {
      "epoch": 0.3893975903614458,
      "grad_norm": 0.15890805423259735,
      "learning_rate": 3.9903145446619837e-05,
      "loss": 0.0145,
      "step": 101
    },
    {
      "epoch": 0.3932530120481928,
      "grad_norm": 0.18432742357254028,
      "learning_rate": 3.9894148325920226e-05,
      "loss": 0.0193,
      "step": 102
    },
    {
      "epoch": 0.39710843373493976,
      "grad_norm": 0.10168084502220154,
      "learning_rate": 3.9884752776948564e-05,
      "loss": 0.0159,
      "step": 103
    },
    {
      "epoch": 0.4009638554216867,
      "grad_norm": 0.1423511803150177,
      "learning_rate": 3.987495898787336e-05,
      "loss": 0.0106,
      "step": 104
    },
    {
      "epoch": 0.40481927710843374,
      "grad_norm": 0.12045685201883316,
      "learning_rate": 3.9864767154838864e-05,
      "loss": 0.0251,
      "step": 105
    },
    {
      "epoch": 0.4086746987951807,
      "grad_norm": 0.2883131802082062,
      "learning_rate": 3.985417748196108e-05,
      "loss": 0.0134,
      "step": 106
    },
    {
      "epoch": 0.4125301204819277,
      "grad_norm": 0.11933720856904984,
      "learning_rate": 3.9843190181323744e-05,
      "loss": 0.0128,
      "step": 107
    },
    {
      "epoch": 0.4163855421686747,
      "grad_norm": 0.13842257857322693,
      "learning_rate": 3.983180547297404e-05,
      "loss": 0.0177,
      "step": 108
    },
    {
      "epoch": 0.4202409638554217,
      "grad_norm": 0.10292483121156693,
      "learning_rate": 3.982002358491817e-05,
      "loss": 0.0145,
      "step": 109
    },
    {
      "epoch": 0.42409638554216866,
      "grad_norm": 0.13685393333435059,
      "learning_rate": 3.9807844753116864e-05,
      "loss": 0.0184,
      "step": 110
    },
    {
      "epoch": 0.4279518072289157,
      "grad_norm": 0.11288538575172424,
      "learning_rate": 3.979526922148058e-05,
      "loss": 0.0238,
      "step": 111
    },
    {
      "epoch": 0.43180722891566264,
      "grad_norm": 0.18025726079940796,
      "learning_rate": 3.978229724186464e-05,
      "loss": 0.0135,
      "step": 112
    },
    {
      "epoch": 0.43566265060240966,
      "grad_norm": 0.1295086145401001,
      "learning_rate": 3.9768929074064206e-05,
      "loss": 0.0109,
      "step": 113
    },
    {
      "epoch": 0.4395180722891566,
      "grad_norm": 0.16575755178928375,
      "learning_rate": 3.975516498580907e-05,
      "loss": 0.0163,
      "step": 114
    },
    {
      "epoch": 0.4433734939759036,
      "grad_norm": 0.08723966032266617,
      "learning_rate": 3.9741005252758255e-05,
      "loss": 0.0098,
      "step": 115
    },
    {
      "epoch": 0.4472289156626506,
      "grad_norm": 0.29995423555374146,
      "learning_rate": 3.972645015849457e-05,
      "loss": 0.0279,
      "step": 116
    },
    {
      "epoch": 0.45108433734939757,
      "grad_norm": 0.12310135364532471,
      "learning_rate": 3.971149999451886e-05,
      "loss": 0.0276,
      "step": 117
    },
    {
      "epoch": 0.4549397590361446,
      "grad_norm": 0.22947722673416138,
      "learning_rate": 3.9696155060244166e-05,
      "loss": 0.0143,
      "step": 118
    },
    {
      "epoch": 0.45879518072289155,
      "grad_norm": 0.1598968803882599,
      "learning_rate": 3.9680415662989806e-05,
      "loss": 0.0184,
      "step": 119
    },
    {
      "epoch": 0.46265060240963857,
      "grad_norm": 0.11719726771116257,
      "learning_rate": 3.966428211797514e-05,
      "loss": 0.0089,
      "step": 120
    },
    {
      "epoch": 0.46650602409638553,
      "grad_norm": 0.2086484581232071,
      "learning_rate": 3.9647754748313294e-05,
      "loss": 0.025,
      "step": 121
    },
    {
      "epoch": 0.47036144578313255,
      "grad_norm": 0.1487833857536316,
      "learning_rate": 3.963083388500469e-05,
      "loss": 0.0107,
      "step": 122
    },
    {
      "epoch": 0.4742168674698795,
      "grad_norm": 0.14869804680347443,
      "learning_rate": 3.96135198669304e-05,
      "loss": 0.0174,
      "step": 123
    },
    {
      "epoch": 0.47807228915662653,
      "grad_norm": 0.20308156311511993,
      "learning_rate": 3.959581304084536e-05,
      "loss": 0.0116,
      "step": 124
    },
    {
      "epoch": 0.4819277108433735,
      "grad_norm": 0.17671343684196472,
      "learning_rate": 3.957771376137144e-05,
      "loss": 0.0182,
      "step": 125
    },
    {
      "epoch": 0.4857831325301205,
      "grad_norm": 0.23302067816257477,
      "learning_rate": 3.955922239099035e-05,
      "loss": 0.013,
      "step": 126
    },
    {
      "epoch": 0.48963855421686747,
      "grad_norm": 0.11614499241113663,
      "learning_rate": 3.954033930003634e-05,
      "loss": 0.0134,
      "step": 127
    },
    {
      "epoch": 0.49349397590361443,
      "grad_norm": 0.1320069283246994,
      "learning_rate": 3.952106486668884e-05,
      "loss": 0.0097,
      "step": 128
    },
    {
      "epoch": 0.49734939759036145,
      "grad_norm": 0.11876890808343887,
      "learning_rate": 3.9501399476964806e-05,
      "loss": 0.0105,
      "step": 129
    },
    {
      "epoch": 0.5012048192771085,
      "grad_norm": 0.26870131492614746,
      "learning_rate": 3.9481343524711095e-05,
      "loss": 0.0157,
      "step": 130
    },
    {
      "epoch": 0.5050602409638554,
      "grad_norm": 0.12262927740812302,
      "learning_rate": 3.946089741159648e-05,
      "loss": 0.0089,
      "step": 131
    },
    {
      "epoch": 0.5089156626506024,
      "grad_norm": 0.16786855459213257,
      "learning_rate": 3.944006154710366e-05,
      "loss": 0.0149,
      "step": 132
    },
    {
      "epoch": 0.5127710843373494,
      "grad_norm": 0.17113229632377625,
      "learning_rate": 3.9418836348521045e-05,
      "loss": 0.0167,
      "step": 133
    },
    {
      "epoch": 0.5166265060240964,
      "grad_norm": 0.1865008920431137,
      "learning_rate": 3.939722224093439e-05,
      "loss": 0.0172,
      "step": 134
    },
    {
      "epoch": 0.5204819277108433,
      "grad_norm": 0.11489243060350418,
      "learning_rate": 3.937521965721831e-05,
      "loss": 0.0135,
      "step": 135
    },
    {
      "epoch": 0.5243373493975904,
      "grad_norm": 0.12064780294895172,
      "learning_rate": 3.9352829038027564e-05,
      "loss": 0.0224,
      "step": 136
    },
    {
      "epoch": 0.5281927710843374,
      "grad_norm": 0.21562877297401428,
      "learning_rate": 3.933005083178828e-05,
      "loss": 0.0197,
      "step": 137
    },
    {
      "epoch": 0.5320481927710843,
      "grad_norm": 0.15050241351127625,
      "learning_rate": 3.930688549468894e-05,
      "loss": 0.0316,
      "step": 138
    },
    {
      "epoch": 0.5359036144578313,
      "grad_norm": 0.17046259343624115,
      "learning_rate": 3.928333349067125e-05,
      "loss": 0.0233,
      "step": 139
    },
    {
      "epoch": 0.5397590361445783,
      "grad_norm": 0.15832003951072693,
      "learning_rate": 3.9259395291420864e-05,
      "loss": 0.0131,
      "step": 140
    },
    {
      "epoch": 0.5436144578313253,
      "grad_norm": 0.10384412109851837,
      "learning_rate": 3.923507137635792e-05,
      "loss": 0.0084,
      "step": 141
    },
    {
      "epoch": 0.5474698795180722,
      "grad_norm": 0.1093461811542511,
      "learning_rate": 3.921036223262745e-05,
      "loss": 0.0082,
      "step": 142
    },
    {
      "epoch": 0.5513253012048193,
      "grad_norm": 0.26693451404571533,
      "learning_rate": 3.9185268355089606e-05,
      "loss": 0.0198,
      "step": 143
    },
    {
      "epoch": 0.5551807228915663,
      "grad_norm": 0.10376594215631485,
      "learning_rate": 3.915979024630978e-05,
      "loss": 0.011,
      "step": 144
    },
    {
      "epoch": 0.5590361445783133,
      "grad_norm": 0.18130183219909668,
      "learning_rate": 3.913392841654851e-05,
      "loss": 0.0111,
      "step": 145
    },
    {
      "epoch": 0.5628915662650602,
      "grad_norm": 0.1697569340467453,
      "learning_rate": 3.9107683383751264e-05,
      "loss": 0.0188,
      "step": 146
    },
    {
      "epoch": 0.5667469879518072,
      "grad_norm": 0.1838744878768921,
      "learning_rate": 3.9081055673538093e-05,
      "loss": 0.0181,
      "step": 147
    },
    {
      "epoch": 0.5706024096385542,
      "grad_norm": 0.19388824701309204,
      "learning_rate": 3.9054045819193074e-05,
      "loss": 0.0136,
      "step": 148
    },
    {
      "epoch": 0.5744578313253013,
      "grad_norm": 0.14233097434043884,
      "learning_rate": 3.902665436165364e-05,
      "loss": 0.0285,
      "step": 149
    },
    {
      "epoch": 0.5783132530120482,
      "grad_norm": 0.14576232433319092,
      "learning_rate": 3.8998881849499756e-05,
      "loss": 0.0139,
      "step": 150
    },
    {
      "epoch": 0.5821686746987952,
      "grad_norm": 0.12378726154565811,
      "learning_rate": 3.897072883894291e-05,
      "loss": 0.0154,
      "step": 151
    },
    {
      "epoch": 0.5860240963855422,
      "grad_norm": 0.180088609457016,
      "learning_rate": 3.894219589381501e-05,
      "loss": 0.0284,
      "step": 152
    },
    {
      "epoch": 0.5898795180722891,
      "grad_norm": 0.1278950572013855,
      "learning_rate": 3.8913283585557054e-05,
      "loss": 0.006,
      "step": 153
    },
    {
      "epoch": 0.5937349397590361,
      "grad_norm": 0.09193156659603119,
      "learning_rate": 3.8883992493207696e-05,
      "loss": 0.0098,
      "step": 154
    },
    {
      "epoch": 0.5975903614457831,
      "grad_norm": 0.0820469856262207,
      "learning_rate": 3.885432320339167e-05,
      "loss": 0.008,
      "step": 155
    },
    {
      "epoch": 0.6014457831325302,
      "grad_norm": 0.1358262449502945,
      "learning_rate": 3.8824276310308e-05,
      "loss": 0.0134,
      "step": 156
    },
    {
      "epoch": 0.6053012048192771,
      "grad_norm": 0.1384797841310501,
      "learning_rate": 3.879385241571817e-05,
      "loss": 0.0211,
      "step": 157
    },
    {
      "epoch": 0.6091566265060241,
      "grad_norm": 0.1795671582221985,
      "learning_rate": 3.876305212893399e-05,
      "loss": 0.0189,
      "step": 158
    },
    {
      "epoch": 0.6130120481927711,
      "grad_norm": 0.09805607050657272,
      "learning_rate": 3.873187606680543e-05,
      "loss": 0.0074,
      "step": 159
    },
    {
      "epoch": 0.6168674698795181,
      "grad_norm": 0.10765130817890167,
      "learning_rate": 3.8700324853708304e-05,
      "loss": 0.0268,
      "step": 160
    },
    {
      "epoch": 0.620722891566265,
      "grad_norm": 0.15209679305553436,
      "learning_rate": 3.866839912153168e-05,
      "loss": 0.0148,
      "step": 161
    },
    {
      "epoch": 0.624578313253012,
      "grad_norm": 0.10442651808261871,
      "learning_rate": 3.863609950966531e-05,
      "loss": 0.0168,
      "step": 162
    },
    {
      "epoch": 0.6284337349397591,
      "grad_norm": 0.11361811310052872,
      "learning_rate": 3.860342666498677e-05,
      "loss": 0.0079,
      "step": 163
    },
    {
      "epoch": 0.632289156626506,
      "grad_norm": 0.09062547236680984,
      "learning_rate": 3.8570381241848534e-05,
      "loss": 0.0129,
      "step": 164
    },
    {
      "epoch": 0.636144578313253,
      "grad_norm": 0.10144251585006714,
      "learning_rate": 3.853696390206484e-05,
      "loss": 0.0179,
      "step": 165
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.10997909307479858,
      "learning_rate": 3.850317531489848e-05,
      "loss": 0.0218,
      "step": 166
    },
    {
      "epoch": 0.643855421686747,
      "grad_norm": 0.08607764542102814,
      "learning_rate": 3.846901615704734e-05,
      "loss": 0.0124,
      "step": 167
    },
    {
      "epoch": 0.6477108433734939,
      "grad_norm": 0.20264504849910736,
      "learning_rate": 3.843448711263089e-05,
      "loss": 0.0189,
      "step": 168
    },
    {
      "epoch": 0.651566265060241,
      "grad_norm": 0.10201306641101837,
      "learning_rate": 3.839958887317649e-05,
      "loss": 0.013,
      "step": 169
    },
    {
      "epoch": 0.655421686746988,
      "grad_norm": 0.15065032243728638,
      "learning_rate": 3.8364322137605484e-05,
      "loss": 0.0225,
      "step": 170
    },
    {
      "epoch": 0.659277108433735,
      "grad_norm": 0.1181226447224617,
      "learning_rate": 3.832868761221926e-05,
      "loss": 0.0117,
      "step": 171
    },
    {
      "epoch": 0.6631325301204819,
      "grad_norm": 0.277878075838089,
      "learning_rate": 3.829268601068509e-05,
      "loss": 0.0117,
      "step": 172
    },
    {
      "epoch": 0.6669879518072289,
      "grad_norm": 0.18172211945056915,
      "learning_rate": 3.825631805402182e-05,
      "loss": 0.0195,
      "step": 173
    },
    {
      "epoch": 0.6708433734939759,
      "grad_norm": 0.21377064287662506,
      "learning_rate": 3.821958447058543e-05,
      "loss": 0.0151,
      "step": 174
    },
    {
      "epoch": 0.6746987951807228,
      "grad_norm": 0.14039507508277893,
      "learning_rate": 3.818248599605448e-05,
      "loss": 0.0138,
      "step": 175
    },
    {
      "epoch": 0.6785542168674699,
      "grad_norm": 0.15213626623153687,
      "learning_rate": 3.8145023373415326e-05,
      "loss": 0.013,
      "step": 176
    },
    {
      "epoch": 0.6824096385542169,
      "grad_norm": 0.2724161446094513,
      "learning_rate": 3.810719735294731e-05,
      "loss": 0.0301,
      "step": 177
    },
    {
      "epoch": 0.6862650602409639,
      "grad_norm": 0.13877442479133606,
      "learning_rate": 3.806900869220765e-05,
      "loss": 0.0157,
      "step": 178
    },
    {
      "epoch": 0.6901204819277108,
      "grad_norm": 0.0968877524137497,
      "learning_rate": 3.8030458156016326e-05,
      "loss": 0.012,
      "step": 179
    },
    {
      "epoch": 0.6939759036144578,
      "grad_norm": 0.1643582135438919,
      "learning_rate": 3.799154651644075e-05,
      "loss": 0.0155,
      "step": 180
    },
    {
      "epoch": 0.6978313253012048,
      "grad_norm": 0.13681621849536896,
      "learning_rate": 3.795227455278029e-05,
      "loss": 0.0085,
      "step": 181
    },
    {
      "epoch": 0.7016867469879519,
      "grad_norm": 0.11992929130792618,
      "learning_rate": 3.7912643051550674e-05,
      "loss": 0.011,
      "step": 182
    },
    {
      "epoch": 0.7055421686746988,
      "grad_norm": 0.16314208507537842,
      "learning_rate": 3.787265280646825e-05,
      "loss": 0.0197,
      "step": 183
    },
    {
      "epoch": 0.7093975903614458,
      "grad_norm": 0.08850007504224777,
      "learning_rate": 3.783230461843406e-05,
      "loss": 0.009,
      "step": 184
    },
    {
      "epoch": 0.7132530120481928,
      "grad_norm": 0.12203098088502884,
      "learning_rate": 3.7791599295517825e-05,
      "loss": 0.0091,
      "step": 185
    },
    {
      "epoch": 0.7171084337349397,
      "grad_norm": 0.12872737646102905,
      "learning_rate": 3.775053765294175e-05,
      "loss": 0.0101,
      "step": 186
    },
    {
      "epoch": 0.7209638554216867,
      "grad_norm": 0.114129438996315,
      "learning_rate": 3.7709120513064196e-05,
      "loss": 0.0077,
      "step": 187
    },
    {
      "epoch": 0.7248192771084337,
      "grad_norm": 0.12316344678401947,
      "learning_rate": 3.7667348705363227e-05,
      "loss": 0.0103,
      "step": 188
    },
    {
      "epoch": 0.7286746987951808,
      "grad_norm": 0.07499723881483078,
      "learning_rate": 3.762522306641998e-05,
      "loss": 0.0059,
      "step": 189
    },
    {
      "epoch": 0.7325301204819277,
      "grad_norm": 0.10113950073719025,
      "learning_rate": 3.75827444399019e-05,
      "loss": 0.0137,
      "step": 190
    },
    {
      "epoch": 0.7363855421686747,
      "grad_norm": 0.10856342315673828,
      "learning_rate": 3.7539913676545874e-05,
      "loss": 0.0091,
      "step": 191
    },
    {
      "epoch": 0.7402409638554217,
      "grad_norm": 0.18475328385829926,
      "learning_rate": 3.749673163414117e-05,
      "loss": 0.0094,
      "step": 192
    },
    {
      "epoch": 0.7440963855421687,
      "grad_norm": 0.0707508996129036,
      "learning_rate": 3.745319917751229e-05,
      "loss": 0.0062,
      "step": 193
    },
    {
      "epoch": 0.7479518072289156,
      "grad_norm": 0.10805204510688782,
      "learning_rate": 3.740931717850159e-05,
      "loss": 0.0111,
      "step": 194
    },
    {
      "epoch": 0.7518072289156627,
      "grad_norm": 0.10874560475349426,
      "learning_rate": 3.736508651595188e-05,
      "loss": 0.0093,
      "step": 195
    },
    {
      "epoch": 0.7556626506024097,
      "grad_norm": 0.12469008564949036,
      "learning_rate": 3.732050807568878e-05,
      "loss": 0.012,
      "step": 196
    },
    {
      "epoch": 0.7595180722891566,
      "grad_norm": 0.0760217234492302,
      "learning_rate": 3.727558275050301e-05,
      "loss": 0.0068,
      "step": 197
    },
    {
      "epoch": 0.7633734939759036,
      "grad_norm": 0.1542130559682846,
      "learning_rate": 3.7230311440132494e-05,
      "loss": 0.0172,
      "step": 198
    },
    {
      "epoch": 0.7672289156626506,
      "grad_norm": 0.14483506977558136,
      "learning_rate": 3.718469505124434e-05,
      "loss": 0.0213,
      "step": 199
    },
    {
      "epoch": 0.7710843373493976,
      "grad_norm": 0.1366746425628662,
      "learning_rate": 3.713873449741669e-05,
      "loss": 0.0093,
      "step": 200
    },
    {
      "epoch": 0.7749397590361445,
      "grad_norm": 0.12670476734638214,
      "learning_rate": 3.709243069912041e-05,
      "loss": 0.0092,
      "step": 201
    },
    {
      "epoch": 0.7787951807228916,
      "grad_norm": 0.3366965353488922,
      "learning_rate": 3.704578458370067e-05,
      "loss": 0.017,
      "step": 202
    },
    {
      "epoch": 0.7826506024096386,
      "grad_norm": 0.1790863275527954,
      "learning_rate": 3.699879708535838e-05,
      "loss": 0.0247,
      "step": 203
    },
    {
      "epoch": 0.7865060240963856,
      "grad_norm": 0.20145761966705322,
      "learning_rate": 3.695146914513142e-05,
      "loss": 0.0075,
      "step": 204
    },
    {
      "epoch": 0.7903614457831325,
      "grad_norm": 0.09072735905647278,
      "learning_rate": 3.69038017108759e-05,
      "loss": 0.013,
      "step": 205
    },
    {
      "epoch": 0.7942168674698795,
      "grad_norm": 1.0004081726074219,
      "learning_rate": 3.685579573724707e-05,
      "loss": 0.0135,
      "step": 206
    },
    {
      "epoch": 0.7980722891566265,
      "grad_norm": 7.910810470581055,
      "learning_rate": 3.680745218568026e-05,
      "loss": 0.0986,
      "step": 207
    },
    {
      "epoch": 0.8019277108433734,
      "grad_norm": 0.19183991849422455,
      "learning_rate": 3.6758772024371626e-05,
      "loss": 0.0112,
      "step": 208
    },
    {
      "epoch": 0.8057831325301205,
      "grad_norm": 0.17618803679943085,
      "learning_rate": 3.6709756228258735e-05,
      "loss": 0.023,
      "step": 209
    },
    {
      "epoch": 0.8096385542168675,
      "grad_norm": 0.17773428559303284,
      "learning_rate": 3.6660405779001036e-05,
      "loss": 0.0166,
      "step": 210
    },
    {
      "epoch": 0.8134939759036145,
      "grad_norm": 0.11507993191480637,
      "learning_rate": 3.6610721664960236e-05,
      "loss": 0.008,
      "step": 211
    },
    {
      "epoch": 0.8173493975903614,
      "grad_norm": 0.21720387041568756,
      "learning_rate": 3.656070488118047e-05,
      "loss": 0.0364,
      "step": 212
    },
    {
      "epoch": 0.8212048192771084,
      "grad_norm": 0.1748373657464981,
      "learning_rate": 3.65103564293684e-05,
      "loss": 0.0121,
      "step": 213
    },
    {
      "epoch": 0.8250602409638554,
      "grad_norm": 0.11724656075239182,
      "learning_rate": 3.645967731787313e-05,
      "loss": 0.0178,
      "step": 214
    },
    {
      "epoch": 0.8289156626506025,
      "grad_norm": 0.13730894029140472,
      "learning_rate": 3.640866856166601e-05,
      "loss": 0.0078,
      "step": 215
    },
    {
      "epoch": 0.8327710843373494,
      "grad_norm": 0.14834603667259216,
      "learning_rate": 3.635733118232034e-05,
      "loss": 0.0148,
      "step": 216
    },
    {
      "epoch": 0.8366265060240964,
      "grad_norm": 0.14735879004001617,
      "learning_rate": 3.6305666207990886e-05,
      "loss": 0.0166,
      "step": 217
    },
    {
      "epoch": 0.8404819277108434,
      "grad_norm": 0.13057921826839447,
      "learning_rate": 3.625367467339329e-05,
      "loss": 0.0186,
      "step": 218
    },
    {
      "epoch": 0.8443373493975903,
      "grad_norm": 0.19306808710098267,
      "learning_rate": 3.6201357619783336e-05,
      "loss": 0.0152,
      "step": 219
    },
    {
      "epoch": 0.8481927710843373,
      "grad_norm": 0.1420055627822876,
      "learning_rate": 3.614871609493614e-05,
      "loss": 0.0131,
      "step": 220
    },
    {
      "epoch": 0.8520481927710843,
      "grad_norm": 0.11154947429895401,
      "learning_rate": 3.609575115312511e-05,
      "loss": 0.0103,
      "step": 221
    },
    {
      "epoch": 0.8559036144578314,
      "grad_norm": 0.07021575421094894,
      "learning_rate": 3.604246385510088e-05,
      "loss": 0.0054,
      "step": 222
    },
    {
      "epoch": 0.8597590361445783,
      "grad_norm": 0.13351315259933472,
      "learning_rate": 3.598885526807003e-05,
      "loss": 0.0141,
      "step": 223
    },
    {
      "epoch": 0.8636144578313253,
      "grad_norm": 0.43515488505363464,
      "learning_rate": 3.593492646567372e-05,
      "loss": 0.0174,
      "step": 224
    },
    {
      "epoch": 0.8674698795180723,
      "grad_norm": 0.14238739013671875,
      "learning_rate": 3.5880678527966224e-05,
      "loss": 0.013,
      "step": 225
    },
    {
      "epoch": 0.8713253012048193,
      "grad_norm": 0.11826923489570618,
      "learning_rate": 3.582611254139324e-05,
      "loss": 0.0149,
      "step": 226
    },
    {
      "epoch": 0.8751807228915662,
      "grad_norm": 0.22871695458889008,
      "learning_rate": 3.577122959877017e-05,
      "loss": 0.0394,
      "step": 227
    },
    {
      "epoch": 0.8790361445783132,
      "grad_norm": 0.20841334760189056,
      "learning_rate": 3.571603079926024e-05,
      "loss": 0.0163,
      "step": 228
    },
    {
      "epoch": 0.8828915662650603,
      "grad_norm": 0.1635965257883072,
      "learning_rate": 3.566051724835245e-05,
      "loss": 0.0103,
      "step": 229
    },
    {
      "epoch": 0.8867469879518072,
      "grad_norm": 0.11963551491498947,
      "learning_rate": 3.5604690057839456e-05,
      "loss": 0.0121,
      "step": 230
    },
    {
      "epoch": 0.8906024096385542,
      "grad_norm": 0.14931157231330872,
      "learning_rate": 3.554855034579532e-05,
      "loss": 0.0224,
      "step": 231
    },
    {
      "epoch": 0.8944578313253012,
      "grad_norm": 0.2835037410259247,
      "learning_rate": 3.549209923655309e-05,
      "loss": 0.0282,
      "step": 232
    },
    {
      "epoch": 0.8983132530120482,
      "grad_norm": 0.18211285769939423,
      "learning_rate": 3.5435337860682304e-05,
      "loss": 0.012,
      "step": 233
    },
    {
      "epoch": 0.9021686746987951,
      "grad_norm": 0.18591780960559845,
      "learning_rate": 3.537826735496632e-05,
      "loss": 0.015,
      "step": 234
    },
    {
      "epoch": 0.9060240963855422,
      "grad_norm": 0.09304528683423996,
      "learning_rate": 3.532088886237956e-05,
      "loss": 0.0087,
      "step": 235
    },
    {
      "epoch": 0.9098795180722892,
      "grad_norm": 0.1557615101337433,
      "learning_rate": 3.5263203532064674e-05,
      "loss": 0.018,
      "step": 236
    },
    {
      "epoch": 0.9137349397590362,
      "grad_norm": 0.14380118250846863,
      "learning_rate": 3.520521251930941e-05,
      "loss": 0.0106,
      "step": 237
    },
    {
      "epoch": 0.9175903614457831,
      "grad_norm": 0.31495094299316406,
      "learning_rate": 3.5146916985523604e-05,
      "loss": 0.0186,
      "step": 238
    },
    {
      "epoch": 0.9214457831325301,
      "grad_norm": 0.16599956154823303,
      "learning_rate": 3.5088318098215805e-05,
      "loss": 0.0092,
      "step": 239
    },
    {
      "epoch": 0.9253012048192771,
      "grad_norm": 0.13166308403015137,
      "learning_rate": 3.5029417030970005e-05,
      "loss": 0.0111,
      "step": 240
    },
    {
      "epoch": 0.929156626506024,
      "grad_norm": 0.1272827833890915,
      "learning_rate": 3.497021496342203e-05,
      "loss": 0.0192,
      "step": 241
    },
    {
      "epoch": 0.9330120481927711,
      "grad_norm": 0.11934683471918106,
      "learning_rate": 3.4910713081235975e-05,
      "loss": 0.0099,
      "step": 242
    },
    {
      "epoch": 0.9368674698795181,
      "grad_norm": 0.2631380259990692,
      "learning_rate": 3.485091257608047e-05,
      "loss": 0.016,
      "step": 243
    },
    {
      "epoch": 0.9407228915662651,
      "grad_norm": 0.3094984292984009,
      "learning_rate": 3.479081464560475e-05,
      "loss": 0.0309,
      "step": 244
    },
    {
      "epoch": 0.944578313253012,
      "grad_norm": 0.11957540363073349,
      "learning_rate": 3.473042049341474e-05,
      "loss": 0.0125,
      "step": 245
    },
    {
      "epoch": 0.948433734939759,
      "grad_norm": 0.08568521589040756,
      "learning_rate": 3.466973132904892e-05,
      "loss": 0.0086,
      "step": 246
    },
    {
      "epoch": 0.952289156626506,
      "grad_norm": 0.16930826008319855,
      "learning_rate": 3.4608748367954064e-05,
      "loss": 0.011,
      "step": 247
    },
    {
      "epoch": 0.9561445783132531,
      "grad_norm": 0.1169271171092987,
      "learning_rate": 3.4547472831460976e-05,
      "loss": 0.0085,
      "step": 248
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.16872183978557587,
      "learning_rate": 3.4485905946759965e-05,
      "loss": 0.0144,
      "step": 249
    },
    {
      "epoch": 0.963855421686747,
      "grad_norm": 0.09693744778633118,
      "learning_rate": 3.442404894687629e-05,
      "loss": 0.006,
      "step": 250
    },
    {
      "epoch": 0.967710843373494,
      "grad_norm": 0.17783288657665253,
      "learning_rate": 3.4361903070645484e-05,
      "loss": 0.0124,
      "step": 251
    },
    {
      "epoch": 0.971566265060241,
      "grad_norm": 0.2039792239665985,
      "learning_rate": 3.4299469562688494e-05,
      "loss": 0.0218,
      "step": 252
    },
    {
      "epoch": 0.9754216867469879,
      "grad_norm": 0.12866197526454926,
      "learning_rate": 3.423674967338681e-05,
      "loss": 0.0186,
      "step": 253
    },
    {
      "epoch": 0.9792771084337349,
      "grad_norm": 0.19284434616565704,
      "learning_rate": 3.4173744658857403e-05,
      "loss": 0.0112,
      "step": 254
    },
    {
      "epoch": 0.983132530120482,
      "grad_norm": 0.16508746147155762,
      "learning_rate": 3.411045578092754e-05,
      "loss": 0.0187,
      "step": 255
    },
    {
      "epoch": 0.9869879518072289,
      "grad_norm": 0.12111517041921616,
      "learning_rate": 3.404688430710955e-05,
      "loss": 0.0143,
      "step": 256
    },
    {
      "epoch": 0.9908433734939759,
      "grad_norm": 0.13150326907634735,
      "learning_rate": 3.398303151057543e-05,
      "loss": 0.0107,
      "step": 257
    },
    {
      "epoch": 0.9946987951807229,
      "grad_norm": 0.12251593172550201,
      "learning_rate": 3.391889867013134e-05,
      "loss": 0.014,
      "step": 258
    },
    {
      "epoch": 0.9985542168674699,
      "grad_norm": 0.09811683744192123,
      "learning_rate": 3.385448707019199e-05,
      "loss": 0.0109,
      "step": 259
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.229774609208107,
      "learning_rate": 3.3789798000754924e-05,
      "loss": 0.0096,
      "step": 260
    },
    {
      "epoch": 1.003855421686747,
      "grad_norm": 0.10416769236326218,
      "learning_rate": 3.372483275737468e-05,
      "loss": 0.0132,
      "step": 261
    },
    {
      "epoch": 1.007710843373494,
      "grad_norm": 0.1602071076631546,
      "learning_rate": 3.365959264113683e-05,
      "loss": 0.0085,
      "step": 262
    },
    {
      "epoch": 1.011566265060241,
      "grad_norm": 0.1573389619588852,
      "learning_rate": 3.359407895863199e-05,
      "loss": 0.0273,
      "step": 263
    },
    {
      "epoch": 1.0154216867469879,
      "grad_norm": 0.0807509571313858,
      "learning_rate": 3.3528293021929523e-05,
      "loss": 0.0176,
      "step": 264
    },
    {
      "epoch": 1.0192771084337349,
      "grad_norm": 0.12698227167129517,
      "learning_rate": 3.34622361485514e-05,
      "loss": 0.0275,
      "step": 265
    },
    {
      "epoch": 1.0231325301204819,
      "grad_norm": 0.07560165226459503,
      "learning_rate": 3.339590966144574e-05,
      "loss": 0.003,
      "step": 266
    },
    {
      "epoch": 1.026987951807229,
      "grad_norm": 0.11409149318933487,
      "learning_rate": 3.332931488896029e-05,
      "loss": 0.0076,
      "step": 267
    },
    {
      "epoch": 1.030843373493976,
      "grad_norm": 0.11506525427103043,
      "learning_rate": 3.326245316481591e-05,
      "loss": 0.0151,
      "step": 268
    },
    {
      "epoch": 1.034698795180723,
      "grad_norm": 0.09818965941667557,
      "learning_rate": 3.319532582807977e-05,
      "loss": 0.0135,
      "step": 269
    },
    {
      "epoch": 1.03855421686747,
      "grad_norm": 0.117642842233181,
      "learning_rate": 3.31279342231386e-05,
      "loss": 0.0075,
      "step": 270
    },
    {
      "epoch": 1.0424096385542168,
      "grad_norm": 0.12259496003389359,
      "learning_rate": 3.30602796996717e-05,
      "loss": 0.0108,
      "step": 271
    },
    {
      "epoch": 1.0462650602409638,
      "grad_norm": 0.18426232039928436,
      "learning_rate": 3.299236361262401e-05,
      "loss": 0.0062,
      "step": 272
    },
    {
      "epoch": 1.0501204819277108,
      "grad_norm": 0.10463278740644455,
      "learning_rate": 3.2924187322178865e-05,
      "loss": 0.0131,
      "step": 273
    },
    {
      "epoch": 1.0539759036144578,
      "grad_norm": 0.07446189224720001,
      "learning_rate": 3.285575219373079e-05,
      "loss": 0.0039,
      "step": 274
    },
    {
      "epoch": 1.0578313253012048,
      "grad_norm": 0.09327566623687744,
      "learning_rate": 3.278705959785821e-05,
      "loss": 0.0076,
      "step": 275
    },
    {
      "epoch": 1.0616867469879518,
      "grad_norm": 0.1121094599366188,
      "learning_rate": 3.27181109102959e-05,
      "loss": 0.0042,
      "step": 276
    },
    {
      "epoch": 1.0655421686746989,
      "grad_norm": 0.0797392874956131,
      "learning_rate": 3.2648907511907544e-05,
      "loss": 0.0055,
      "step": 277
    },
    {
      "epoch": 1.0693975903614459,
      "grad_norm": 0.1773054301738739,
      "learning_rate": 3.2579450788657997e-05,
      "loss": 0.0196,
      "step": 278
    },
    {
      "epoch": 1.0732530120481927,
      "grad_norm": 0.8156651258468628,
      "learning_rate": 3.250974213158555e-05,
      "loss": 0.0096,
      "step": 279
    },
    {
      "epoch": 1.0771084337349397,
      "grad_norm": 0.10387098044157028,
      "learning_rate": 3.243978293677408e-05,
      "loss": 0.0071,
      "step": 280
    },
    {
      "epoch": 1.0809638554216867,
      "grad_norm": 0.29274797439575195,
      "learning_rate": 3.23695746053251e-05,
      "loss": 0.0147,
      "step": 281
    },
    {
      "epoch": 1.0848192771084337,
      "grad_norm": 0.0871700718998909,
      "learning_rate": 3.229911854332967e-05,
      "loss": 0.0105,
      "step": 282
    },
    {
      "epoch": 1.0886746987951808,
      "grad_norm": 0.13531728088855743,
      "learning_rate": 3.222841616184025e-05,
      "loss": 0.0188,
      "step": 283
    },
    {
      "epoch": 1.0925301204819278,
      "grad_norm": 0.11029268801212311,
      "learning_rate": 3.215746887684244e-05,
      "loss": 0.0067,
      "step": 284
    },
    {
      "epoch": 1.0963855421686748,
      "grad_norm": 0.10946464538574219,
      "learning_rate": 3.208627810922665e-05,
      "loss": 0.0142,
      "step": 285
    },
    {
      "epoch": 1.1002409638554216,
      "grad_norm": 0.05139852315187454,
      "learning_rate": 3.201484528475958e-05,
      "loss": 0.0053,
      "step": 286
    },
    {
      "epoch": 1.1040963855421686,
      "grad_norm": 0.10800574719905853,
      "learning_rate": 3.194317183405573e-05,
      "loss": 0.0082,
      "step": 287
    },
    {
      "epoch": 1.1079518072289156,
      "grad_norm": 0.08289000391960144,
      "learning_rate": 3.187125919254869e-05,
      "loss": 0.0058,
      "step": 288
    },
    {
      "epoch": 1.1118072289156626,
      "grad_norm": 0.2032063752412796,
      "learning_rate": 3.1799108800462466e-05,
      "loss": 0.0081,
      "step": 289
    },
    {
      "epoch": 1.1156626506024097,
      "grad_norm": 0.09905604273080826,
      "learning_rate": 3.172672210278254e-05,
      "loss": 0.0094,
      "step": 290
    },
    {
      "epoch": 1.1195180722891567,
      "grad_norm": 0.13566920161247253,
      "learning_rate": 3.1654100549227024e-05,
      "loss": 0.0065,
      "step": 291
    },
    {
      "epoch": 1.1233734939759037,
      "grad_norm": 0.13147567212581635,
      "learning_rate": 3.158124559421758e-05,
      "loss": 0.0056,
      "step": 292
    },
    {
      "epoch": 1.1272289156626507,
      "grad_norm": 0.0829431414604187,
      "learning_rate": 3.1508158696850275e-05,
      "loss": 0.0066,
      "step": 293
    },
    {
      "epoch": 1.1310843373493975,
      "grad_norm": 0.152206689119339,
      "learning_rate": 3.14348413208664e-05,
      "loss": 0.0072,
      "step": 294
    },
    {
      "epoch": 1.1349397590361445,
      "grad_norm": 0.10874905437231064,
      "learning_rate": 3.136129493462312e-05,
      "loss": 0.0071,
      "step": 295
    },
    {
      "epoch": 1.1387951807228915,
      "grad_norm": 0.08331774920225143,
      "learning_rate": 3.12875210110641e-05,
      "loss": 0.0079,
      "step": 296
    },
    {
      "epoch": 1.1426506024096386,
      "grad_norm": 0.08060551434755325,
      "learning_rate": 3.121352102768998e-05,
      "loss": 0.0059,
      "step": 297
    },
    {
      "epoch": 1.1465060240963856,
      "grad_norm": 0.08402010798454285,
      "learning_rate": 3.113929646652879e-05,
      "loss": 0.0047,
      "step": 298
    },
    {
      "epoch": 1.1503614457831326,
      "grad_norm": 0.07346951961517334,
      "learning_rate": 3.106484881410628e-05,
      "loss": 0.0071,
      "step": 299
    },
    {
      "epoch": 1.1542168674698796,
      "grad_norm": 0.11735919117927551,
      "learning_rate": 3.0990179561416124e-05,
      "loss": 0.0077,
      "step": 300
    },
    {
      "epoch": 1.1580722891566264,
      "grad_norm": 0.05967315658926964,
      "learning_rate": 3.091529020389009e-05,
      "loss": 0.0036,
      "step": 301
    },
    {
      "epoch": 1.1619277108433734,
      "grad_norm": 2.5192465782165527,
      "learning_rate": 3.0840182241368083e-05,
      "loss": 0.0089,
      "step": 302
    },
    {
      "epoch": 1.1657831325301204,
      "grad_norm": 0.12357142567634583,
      "learning_rate": 3.076485717806808e-05,
      "loss": 0.011,
      "step": 303
    },
    {
      "epoch": 1.1696385542168675,
      "grad_norm": 0.08216772973537445,
      "learning_rate": 3.0689316522556026e-05,
      "loss": 0.0147,
      "step": 304
    },
    {
      "epoch": 1.1734939759036145,
      "grad_norm": 0.10057099908590317,
      "learning_rate": 3.061356178771564e-05,
      "loss": 0.0143,
      "step": 305
    },
    {
      "epoch": 1.1773493975903615,
      "grad_norm": 0.14141112565994263,
      "learning_rate": 3.053759449071807e-05,
      "loss": 0.0104,
      "step": 306
    },
    {
      "epoch": 1.1812048192771085,
      "grad_norm": 0.08545457571744919,
      "learning_rate": 3.0461416152991555e-05,
      "loss": 0.0076,
      "step": 307
    },
    {
      "epoch": 1.1850602409638555,
      "grad_norm": 0.11981064826250076,
      "learning_rate": 3.038502830019092e-05,
      "loss": 0.0078,
      "step": 308
    },
    {
      "epoch": 1.1889156626506023,
      "grad_norm": 0.07742572575807571,
      "learning_rate": 3.0308432462167045e-05,
      "loss": 0.0054,
      "step": 309
    },
    {
      "epoch": 1.1927710843373494,
      "grad_norm": 0.08275239914655685,
      "learning_rate": 3.0231630172936205e-05,
      "loss": 0.0053,
      "step": 310
    },
    {
      "epoch": 1.1966265060240964,
      "grad_norm": 0.42046189308166504,
      "learning_rate": 3.015462297064936e-05,
      "loss": 0.0258,
      "step": 311
    },
    {
      "epoch": 1.2004819277108434,
      "grad_norm": 0.07977071404457092,
      "learning_rate": 3.0077412397561345e-05,
      "loss": 0.0153,
      "step": 312
    },
    {
      "epoch": 1.2043373493975904,
      "grad_norm": 0.09960279613733292,
      "learning_rate": 3.0000000000000004e-05,
      "loss": 0.0101,
      "step": 313
    },
    {
      "epoch": 1.2081927710843374,
      "grad_norm": 0.09214231371879578,
      "learning_rate": 2.992238732833518e-05,
      "loss": 0.0053,
      "step": 314
    },
    {
      "epoch": 1.2120481927710842,
      "grad_norm": 0.06532324850559235,
      "learning_rate": 2.98445759369477e-05,
      "loss": 0.0082,
      "step": 315
    },
    {
      "epoch": 1.2159036144578312,
      "grad_norm": 0.06730500608682632,
      "learning_rate": 2.9766567384198228e-05,
      "loss": 0.0063,
      "step": 316
    },
    {
      "epoch": 1.2197590361445783,
      "grad_norm": 0.1506088674068451,
      "learning_rate": 2.9688363232396056e-05,
      "loss": 0.0095,
      "step": 317
    },
    {
      "epoch": 1.2236144578313253,
      "grad_norm": 0.07647290825843811,
      "learning_rate": 2.960996504776783e-05,
      "loss": 0.0082,
      "step": 318
    },
    {
      "epoch": 1.2274698795180723,
      "grad_norm": 0.21594130992889404,
      "learning_rate": 2.9531374400426158e-05,
      "loss": 0.0251,
      "step": 319
    },
    {
      "epoch": 1.2313253012048193,
      "grad_norm": 0.1463146209716797,
      "learning_rate": 2.945259286433819e-05,
      "loss": 0.0087,
      "step": 320
    },
    {
      "epoch": 1.2351807228915663,
      "grad_norm": 0.08222866803407669,
      "learning_rate": 2.9373622017294075e-05,
      "loss": 0.0062,
      "step": 321
    },
    {
      "epoch": 1.2390361445783133,
      "grad_norm": 0.09863219410181046,
      "learning_rate": 2.9294463440875375e-05,
      "loss": 0.0067,
      "step": 322
    },
    {
      "epoch": 1.2428915662650604,
      "grad_norm": 0.13434955477714539,
      "learning_rate": 2.9215118720423375e-05,
      "loss": 0.0073,
      "step": 323
    },
    {
      "epoch": 1.2467469879518072,
      "grad_norm": 0.08647848665714264,
      "learning_rate": 2.913558944500736e-05,
      "loss": 0.006,
      "step": 324
    },
    {
      "epoch": 1.2506024096385542,
      "grad_norm": 0.07066208124160767,
      "learning_rate": 2.9055877207392752e-05,
      "loss": 0.0041,
      "step": 325
    },
    {
      "epoch": 1.2544578313253012,
      "grad_norm": 0.09136916697025299,
      "learning_rate": 2.897598360400925e-05,
      "loss": 0.0055,
      "step": 326
    },
    {
      "epoch": 1.2583132530120482,
      "grad_norm": 0.12554176151752472,
      "learning_rate": 2.8895910234918828e-05,
      "loss": 0.0117,
      "step": 327
    },
    {
      "epoch": 1.2621686746987952,
      "grad_norm": 0.1485210657119751,
      "learning_rate": 2.8815658703783715e-05,
      "loss": 0.0039,
      "step": 328
    },
    {
      "epoch": 1.266024096385542,
      "grad_norm": 0.09695762395858765,
      "learning_rate": 2.873523061783426e-05,
      "loss": 0.0091,
      "step": 329
    },
    {
      "epoch": 1.269879518072289,
      "grad_norm": 0.07787121087312698,
      "learning_rate": 2.8654627587836724e-05,
      "loss": 0.0041,
      "step": 330
    },
    {
      "epoch": 1.273734939759036,
      "grad_norm": 0.06576529145240784,
      "learning_rate": 2.8573851228061084e-05,
      "loss": 0.0041,
      "step": 331
    },
    {
      "epoch": 1.277590361445783,
      "grad_norm": 0.06353545188903809,
      "learning_rate": 2.8492903156248642e-05,
      "loss": 0.0071,
      "step": 332
    },
    {
      "epoch": 1.28144578313253,
      "grad_norm": 0.1047273725271225,
      "learning_rate": 2.8411784993579633e-05,
      "loss": 0.009,
      "step": 333
    },
    {
      "epoch": 1.2853012048192771,
      "grad_norm": 0.1017284020781517,
      "learning_rate": 2.8330498364640803e-05,
      "loss": 0.01,
      "step": 334
    },
    {
      "epoch": 1.2891566265060241,
      "grad_norm": 0.13284249603748322,
      "learning_rate": 2.8249044897392814e-05,
      "loss": 0.0124,
      "step": 335
    },
    {
      "epoch": 1.2930120481927712,
      "grad_norm": 0.08927687257528305,
      "learning_rate": 2.8167426223137682e-05,
      "loss": 0.0193,
      "step": 336
    },
    {
      "epoch": 1.2968674698795182,
      "grad_norm": 0.0922897607088089,
      "learning_rate": 2.80856439764861e-05,
      "loss": 0.0403,
      "step": 337
    },
    {
      "epoch": 1.3007228915662652,
      "grad_norm": 0.3542531430721283,
      "learning_rate": 2.8003699795324674e-05,
      "loss": 0.0053,
      "step": 338
    },
    {
      "epoch": 1.304578313253012,
      "grad_norm": 0.10456252843141556,
      "learning_rate": 2.792159532078314e-05,
      "loss": 0.0126,
      "step": 339
    },
    {
      "epoch": 1.308433734939759,
      "grad_norm": 0.06387709081172943,
      "learning_rate": 2.7839332197201505e-05,
      "loss": 0.0036,
      "step": 340
    },
    {
      "epoch": 1.312289156626506,
      "grad_norm": 0.06973070651292801,
      "learning_rate": 2.77569120720971e-05,
      "loss": 0.0094,
      "step": 341
    },
    {
      "epoch": 1.316144578313253,
      "grad_norm": 0.08025919646024704,
      "learning_rate": 2.7674336596131596e-05,
      "loss": 0.0081,
      "step": 342
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.15938566625118256,
      "learning_rate": 2.7591607423077932e-05,
      "loss": 0.0104,
      "step": 343
    },
    {
      "epoch": 1.3238554216867469,
      "grad_norm": 0.13838085532188416,
      "learning_rate": 2.75087262097872e-05,
      "loss": 0.004,
      "step": 344
    },
    {
      "epoch": 1.3277108433734939,
      "grad_norm": 0.06210877001285553,
      "learning_rate": 2.7425694616155474e-05,
      "loss": 0.0044,
      "step": 345
    },
    {
      "epoch": 1.331566265060241,
      "grad_norm": 0.094986192882061,
      "learning_rate": 2.7342514305090554e-05,
      "loss": 0.0206,
      "step": 346
    },
    {
      "epoch": 1.335421686746988,
      "grad_norm": 0.07161661237478256,
      "learning_rate": 2.7259186942478656e-05,
      "loss": 0.0033,
      "step": 347
    },
    {
      "epoch": 1.339277108433735,
      "grad_norm": 0.12837430834770203,
      "learning_rate": 2.717571419715107e-05,
      "loss": 0.0143,
      "step": 348
    },
    {
      "epoch": 1.343132530120482,
      "grad_norm": 0.07206165045499802,
      "learning_rate": 2.7092097740850712e-05,
      "loss": 0.0037,
      "step": 349
    },
    {
      "epoch": 1.346987951807229,
      "grad_norm": 0.06978686898946762,
      "learning_rate": 2.700833924819868e-05,
      "loss": 0.0059,
      "step": 350
    },
    {
      "epoch": 1.350843373493976,
      "grad_norm": 0.14167740941047668,
      "learning_rate": 2.692444039666066e-05,
      "loss": 0.0093,
      "step": 351
    },
    {
      "epoch": 1.354698795180723,
      "grad_norm": 0.08207441866397858,
      "learning_rate": 2.684040286651338e-05,
      "loss": 0.0087,
      "step": 352
    },
    {
      "epoch": 1.3585542168674698,
      "grad_norm": 0.08811763674020767,
      "learning_rate": 2.6756228340810946e-05,
      "loss": 0.0076,
      "step": 353
    },
    {
      "epoch": 1.3624096385542168,
      "grad_norm": 0.09116017818450928,
      "learning_rate": 2.6671918505351132e-05,
      "loss": 0.0043,
      "step": 354
    },
    {
      "epoch": 1.3662650602409638,
      "grad_norm": 0.06122549995779991,
      "learning_rate": 2.6587475048641596e-05,
      "loss": 0.0032,
      "step": 355
    },
    {
      "epoch": 1.3701204819277109,
      "grad_norm": 0.07415001839399338,
      "learning_rate": 2.6502899661866104e-05,
      "loss": 0.0072,
      "step": 356
    },
    {
      "epoch": 1.3739759036144579,
      "grad_norm": 0.08215410262346268,
      "learning_rate": 2.6418194038850634e-05,
      "loss": 0.009,
      "step": 357
    },
    {
      "epoch": 1.377831325301205,
      "grad_norm": 0.0975186750292778,
      "learning_rate": 2.6333359876029455e-05,
      "loss": 0.0045,
      "step": 358
    },
    {
      "epoch": 1.3816867469879517,
      "grad_norm": 0.20146572589874268,
      "learning_rate": 2.624839887241115e-05,
      "loss": 0.0094,
      "step": 359
    },
    {
      "epoch": 1.3855421686746987,
      "grad_norm": 0.08265519142150879,
      "learning_rate": 2.6163312729544616e-05,
      "loss": 0.0047,
      "step": 360
    },
    {
      "epoch": 1.3893975903614457,
      "grad_norm": 0.09653282910585403,
      "learning_rate": 2.607810315148494e-05,
      "loss": 0.0077,
      "step": 361
    },
    {
      "epoch": 1.3932530120481927,
      "grad_norm": 0.10554955899715424,
      "learning_rate": 2.5992771844759315e-05,
      "loss": 0.0091,
      "step": 362
    },
    {
      "epoch": 1.3971084337349398,
      "grad_norm": 0.06274604052305222,
      "learning_rate": 2.5907320518332827e-05,
      "loss": 0.0073,
      "step": 363
    },
    {
      "epoch": 1.4009638554216868,
      "grad_norm": 0.07398885488510132,
      "learning_rate": 2.582175088357426e-05,
      "loss": 0.0122,
      "step": 364
    },
    {
      "epoch": 1.4048192771084338,
      "grad_norm": 0.0733281746506691,
      "learning_rate": 2.5736064654221808e-05,
      "loss": 0.0107,
      "step": 365
    },
    {
      "epoch": 1.4086746987951808,
      "grad_norm": 0.08075340837240219,
      "learning_rate": 2.5650263546348743e-05,
      "loss": 0.0071,
      "step": 366
    },
    {
      "epoch": 1.4125301204819278,
      "grad_norm": 0.22409602999687195,
      "learning_rate": 2.5564349278329056e-05,
      "loss": 0.0078,
      "step": 367
    },
    {
      "epoch": 1.4163855421686746,
      "grad_norm": 0.07538998126983643,
      "learning_rate": 2.547832357080305e-05,
      "loss": 0.0115,
      "step": 368
    }
  ],
  "logging_steps": 1,
  "max_steps": 780,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 92,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.8819364940317983e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}