{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 740,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.033783783783783786,
      "grad_norm": 8.29013495953546,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 0.5181,
      "step": 5
    },
    {
      "epoch": 0.06756756756756757,
      "grad_norm": 4.196123392105515,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 0.3567,
      "step": 10
    },
    {
      "epoch": 0.10135135135135136,
      "grad_norm": 3.2047846933733215,
      "learning_rate": 2.0270270270270273e-05,
      "loss": 0.3399,
      "step": 15
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 3.152224162446417,
      "learning_rate": 2.702702702702703e-05,
      "loss": 0.2552,
      "step": 20
    },
    {
      "epoch": 0.16891891891891891,
      "grad_norm": 3.649791800007489,
      "learning_rate": 3.3783783783783784e-05,
      "loss": 0.2705,
      "step": 25
    },
    {
      "epoch": 0.20270270270270271,
      "grad_norm": 3.4730019079608607,
      "learning_rate": 4.0540540540540545e-05,
      "loss": 0.2647,
      "step": 30
    },
    {
      "epoch": 0.23648648648648649,
      "grad_norm": 5.8436815551949195,
      "learning_rate": 4.72972972972973e-05,
      "loss": 0.3513,
      "step": 35
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 2.9019987565526995,
      "learning_rate": 4.9997978016429605e-05,
      "loss": 0.4888,
      "step": 40
    },
    {
      "epoch": 0.30405405405405406,
      "grad_norm": 2.9397780048282027,
      "learning_rate": 4.9985622766211935e-05,
      "loss": 0.2713,
      "step": 45
    },
    {
      "epoch": 0.33783783783783783,
      "grad_norm": 2.8421186668651917,
      "learning_rate": 4.996204175076325e-05,
      "loss": 0.2629,
      "step": 50
    },
    {
      "epoch": 0.3716216216216216,
      "grad_norm": 2.798031701878566,
      "learning_rate": 4.99272467427147e-05,
      "loss": 0.3088,
      "step": 55
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 2.70572378151596,
      "learning_rate": 4.9881255113192526e-05,
      "loss": 0.2351,
      "step": 60
    },
    {
      "epoch": 0.4391891891891892,
      "grad_norm": 2.6573100082008914,
      "learning_rate": 4.982408982314565e-05,
      "loss": 0.2722,
      "step": 65
    },
    {
      "epoch": 0.47297297297297297,
      "grad_norm": 2.613947983515025,
      "learning_rate": 4.975577941188258e-05,
      "loss": 0.2561,
      "step": 70
    },
    {
      "epoch": 0.5067567567567568,
      "grad_norm": 2.4731517146590503,
      "learning_rate": 4.967635798282344e-05,
      "loss": 0.2895,
      "step": 75
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 2.024939244762624,
      "learning_rate": 4.958586518647407e-05,
      "loss": 0.2432,
      "step": 80
    },
    {
      "epoch": 0.5743243243243243,
      "grad_norm": 2.075950925366443,
      "learning_rate": 4.9484346200630855e-05,
      "loss": 0.272,
      "step": 85
    },
    {
      "epoch": 0.6081081081081081,
      "grad_norm": 2.0824590269407257,
      "learning_rate": 4.937185170782607e-05,
      "loss": 0.2465,
      "step": 90
    },
    {
      "epoch": 0.6418918918918919,
      "grad_norm": 2.1834298841545765,
      "learning_rate": 4.9248437870025035e-05,
      "loss": 0.2561,
      "step": 95
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 3.0028526613842756,
      "learning_rate": 4.911416630058772e-05,
      "loss": 0.2112,
      "step": 100
    },
    {
      "epoch": 0.7094594594594594,
      "grad_norm": 6.923011878473891,
      "learning_rate": 4.896910403350873e-05,
      "loss": 0.2703,
      "step": 105
    },
    {
      "epoch": 0.7432432432432432,
      "grad_norm": 1.5341551012824088,
      "learning_rate": 4.88133234899512e-05,
      "loss": 0.2445,
      "step": 110
    },
    {
      "epoch": 0.777027027027027,
      "grad_norm": 1.9243858701641752,
      "learning_rate": 4.864690244209105e-05,
      "loss": 0.208,
      "step": 115
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 1.6400221556474566,
      "learning_rate": 4.8469923974289874e-05,
      "loss": 0.2879,
      "step": 120
    },
    {
      "epoch": 0.8445945945945946,
      "grad_norm": 2.7494322371602946,
      "learning_rate": 4.828247644161577e-05,
      "loss": 0.2482,
      "step": 125
    },
    {
      "epoch": 0.8783783783783784,
      "grad_norm": 1.7599344699077832,
      "learning_rate": 4.808465342573274e-05,
      "loss": 0.2587,
      "step": 130
    },
    {
      "epoch": 0.9121621621621622,
      "grad_norm": 1.6037855553499427,
      "learning_rate": 4.787655368818087e-05,
      "loss": 0.2024,
      "step": 135
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 1.3955654859145117,
      "learning_rate": 4.765828112107034e-05,
      "loss": 0.2557,
      "step": 140
    },
    {
      "epoch": 0.9797297297297297,
      "grad_norm": 1.288730609348733,
      "learning_rate": 4.742994469521421e-05,
      "loss": 0.2232,
      "step": 145
    },
    {
      "epoch": 1.0135135135135136,
      "grad_norm": 1.7061202812270857,
      "learning_rate": 4.719165840572557e-05,
      "loss": 0.2171,
      "step": 150
    },
    {
      "epoch": 1.0472972972972974,
      "grad_norm": 2.1625112154042254,
      "learning_rate": 4.694354121510644e-05,
      "loss": 0.1566,
      "step": 155
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 2.211819043240105,
      "learning_rate": 4.668571699385668e-05,
      "loss": 0.1689,
      "step": 160
    },
    {
      "epoch": 1.114864864864865,
      "grad_norm": 1.4126203201911418,
      "learning_rate": 4.641831445863265e-05,
      "loss": 0.1716,
      "step": 165
    },
    {
      "epoch": 1.1486486486486487,
      "grad_norm": 2.0610795868124194,
      "learning_rate": 4.614146710798645e-05,
      "loss": 0.1605,
      "step": 170
    },
    {
      "epoch": 1.1824324324324325,
      "grad_norm": 1.7608279284054584,
      "learning_rate": 4.585531315571788e-05,
      "loss": 0.1594,
      "step": 175
    },
    {
      "epoch": 1.2162162162162162,
      "grad_norm": 1.5806659537041632,
      "learning_rate": 4.555999546187229e-05,
      "loss": 0.1484,
      "step": 180
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.5935232233887906,
      "learning_rate": 4.5255661461418854e-05,
      "loss": 0.1383,
      "step": 185
    },
    {
      "epoch": 1.2837837837837838,
      "grad_norm": 1.609634736806882,
      "learning_rate": 4.4942463090644896e-05,
      "loss": 0.1547,
      "step": 190
    },
    {
      "epoch": 1.3175675675675675,
      "grad_norm": 1.588101249847805,
      "learning_rate": 4.462055671130289e-05,
      "loss": 0.1482,
      "step": 195
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 1.5585026200892942,
      "learning_rate": 4.4290103032548094e-05,
      "loss": 0.146,
      "step": 200
    },
    {
      "epoch": 1.385135135135135,
      "grad_norm": 1.1467440704851377,
      "learning_rate": 4.395126703070589e-05,
      "loss": 0.1446,
      "step": 205
    },
    {
      "epoch": 1.4189189189189189,
      "grad_norm": 0.9207690008726266,
      "learning_rate": 4.360421786690862e-05,
      "loss": 0.151,
      "step": 210
    },
    {
      "epoch": 1.4527027027027026,
      "grad_norm": 1.1893240387302906,
      "learning_rate": 4.324912880264326e-05,
      "loss": 0.1451,
      "step": 215
    },
    {
      "epoch": 1.4864864864864864,
      "grad_norm": 1.1348088575018935,
      "learning_rate": 4.288617711325207e-05,
      "loss": 0.133,
      "step": 220
    },
    {
      "epoch": 1.5202702702702702,
      "grad_norm": 1.2969069765004013,
      "learning_rate": 4.251554399942928e-05,
      "loss": 0.1482,
      "step": 225
    },
    {
      "epoch": 1.554054054054054,
      "grad_norm": 1.1549621334885833,
      "learning_rate": 4.21374144967581e-05,
      "loss": 0.1589,
      "step": 230
    },
    {
      "epoch": 1.5878378378378377,
      "grad_norm": 1.644137743103509,
      "learning_rate": 4.1751977383333224e-05,
      "loss": 0.1407,
      "step": 235
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 1.3525975163496182,
      "learning_rate": 4.1359425085514906e-05,
      "loss": 0.1363,
      "step": 240
    },
    {
      "epoch": 1.6554054054054053,
      "grad_norm": 2.0901857582974217,
      "learning_rate": 4.095995358186162e-05,
      "loss": 0.1772,
      "step": 245
    },
    {
      "epoch": 1.689189189189189,
      "grad_norm": 1.6270221012827155,
      "learning_rate": 4.055376230528936e-05,
      "loss": 0.1387,
      "step": 250
    },
    {
      "epoch": 1.722972972972973,
      "grad_norm": 1.430683434287028,
      "learning_rate": 4.0141054043506406e-05,
      "loss": 0.1655,
      "step": 255
    },
    {
      "epoch": 1.7567567567567568,
      "grad_norm": 0.8114050618763757,
      "learning_rate": 3.972203483777315e-05,
      "loss": 0.1366,
      "step": 260
    },
    {
      "epoch": 1.7905405405405406,
      "grad_norm": 1.269223528535312,
      "learning_rate": 3.929691388003772e-05,
      "loss": 0.1485,
      "step": 265
    },
    {
      "epoch": 1.8243243243243243,
      "grad_norm": 1.13503713932199,
      "learning_rate": 3.886590340849852e-05,
      "loss": 0.1197,
      "step": 270
    },
    {
      "epoch": 1.8581081081081081,
      "grad_norm": 0.9877113608753568,
      "learning_rate": 3.842921860164607e-05,
      "loss": 0.1334,
      "step": 275
    },
    {
      "epoch": 1.8918918918918919,
      "grad_norm": 0.8713265904231167,
      "learning_rate": 3.798707747083694e-05,
      "loss": 0.1334,
      "step": 280
    },
    {
      "epoch": 1.9256756756756757,
      "grad_norm": 0.8661557486157838,
      "learning_rate": 3.753970075145322e-05,
      "loss": 0.1429,
      "step": 285
    },
    {
      "epoch": 1.9594594594594594,
      "grad_norm": 1.219122887579692,
      "learning_rate": 3.7087311792702265e-05,
      "loss": 0.1356,
      "step": 290
    },
    {
      "epoch": 1.9932432432432432,
      "grad_norm": 1.1038755273323901,
      "learning_rate": 3.663013644611139e-05,
      "loss": 0.1349,
      "step": 295
    },
    {
      "epoch": 2.027027027027027,
      "grad_norm": 0.7179183269742409,
      "learning_rate": 3.616840295277328e-05,
      "loss": 0.0975,
      "step": 300
    },
    {
      "epoch": 2.060810810810811,
      "grad_norm": 0.7521542943209245,
      "learning_rate": 3.5702341829398525e-05,
      "loss": 0.096,
      "step": 305
    },
    {
      "epoch": 2.0945945945945947,
      "grad_norm": 0.7622848192686278,
      "learning_rate": 3.523218575323198e-05,
      "loss": 0.0947,
      "step": 310
    },
    {
      "epoch": 2.1283783783783785,
      "grad_norm": 1.25703408706443,
      "learning_rate": 3.475816944589058e-05,
      "loss": 0.0937,
      "step": 315
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 0.910003923674188,
      "learning_rate": 3.4280529556180404e-05,
      "loss": 0.0998,
      "step": 320
    },
    {
      "epoch": 2.195945945945946,
      "grad_norm": 0.6770052864732585,
      "learning_rate": 3.379950454195172e-05,
      "loss": 0.0714,
      "step": 325
    },
    {
      "epoch": 2.22972972972973,
      "grad_norm": 0.9501804841017806,
      "learning_rate": 3.331533455105084e-05,
      "loss": 0.0856,
      "step": 330
    },
    {
      "epoch": 2.2635135135135136,
      "grad_norm": 0.9502759678893843,
      "learning_rate": 3.2828261301428206e-05,
      "loss": 0.0996,
      "step": 335
    },
    {
      "epoch": 2.2972972972972974,
      "grad_norm": 1.40596962540061,
      "learning_rate": 3.23385279604627e-05,
      "loss": 0.0691,
      "step": 340
    },
    {
      "epoch": 2.331081081081081,
      "grad_norm": 0.850391778909355,
      "learning_rate": 3.18463790235623e-05,
      "loss": 0.0828,
      "step": 345
    },
    {
      "epoch": 2.364864864864865,
      "grad_norm": 0.9046270921885601,
      "learning_rate": 3.135206019210167e-05,
      "loss": 0.0829,
      "step": 350
    },
    {
      "epoch": 2.3986486486486487,
      "grad_norm": 1.344251528726258,
      "learning_rate": 3.085581825075782e-05,
      "loss": 0.0761,
      "step": 355
    },
    {
      "epoch": 2.4324324324324325,
      "grad_norm": 1.3818381192158777,
      "learning_rate": 3.0357900944304774e-05,
      "loss": 0.0912,
      "step": 360
    },
    {
      "epoch": 2.4662162162162162,
      "grad_norm": 0.8834881506524517,
      "learning_rate": 2.9858556853929048e-05,
      "loss": 0.0843,
      "step": 365
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.9107496260787339,
      "learning_rate": 2.9358035273127483e-05,
      "loss": 0.0825,
      "step": 370
    },
    {
      "epoch": 2.5337837837837838,
      "grad_norm": 0.5784107872034662,
      "learning_rate": 2.8856586083249487e-05,
      "loss": 0.0721,
      "step": 375
    },
    {
      "epoch": 2.5675675675675675,
      "grad_norm": 0.7773606211954556,
      "learning_rate": 2.83544596287458e-05,
      "loss": 0.0874,
      "step": 380
    },
    {
      "epoch": 2.6013513513513513,
      "grad_norm": 0.6678593340764952,
      "learning_rate": 2.785190659218604e-05,
      "loss": 0.0859,
      "step": 385
    },
    {
      "epoch": 2.635135135135135,
      "grad_norm": 0.5939162753157207,
      "learning_rate": 2.7349177869107462e-05,
      "loss": 0.0809,
      "step": 390
    },
    {
      "epoch": 2.668918918918919,
      "grad_norm": 0.7283077606960382,
      "learning_rate": 2.684652444275741e-05,
      "loss": 0.0733,
      "step": 395
    },
    {
      "epoch": 2.7027027027027026,
      "grad_norm": 0.6502593242253658,
      "learning_rate": 2.634419725879193e-05,
      "loss": 0.0726,
      "step": 400
    },
    {
      "epoch": 2.7364864864864864,
      "grad_norm": 0.6122029804466594,
      "learning_rate": 2.58424470999932e-05,
      "loss": 0.088,
      "step": 405
    },
    {
      "epoch": 2.77027027027027,
      "grad_norm": 0.799053842544273,
      "learning_rate": 2.534152446106825e-05,
      "loss": 0.0729,
      "step": 410
    },
    {
      "epoch": 2.804054054054054,
      "grad_norm": 0.7095023200640361,
      "learning_rate": 2.4841679423591523e-05,
      "loss": 0.0668,
      "step": 415
    },
    {
      "epoch": 2.8378378378378377,
      "grad_norm": 0.5870222541144963,
      "learning_rate": 2.4343161531153647e-05,
      "loss": 0.0811,
      "step": 420
    },
    {
      "epoch": 2.8716216216216215,
      "grad_norm": 0.7486446531081316,
      "learning_rate": 2.3846219664778824e-05,
      "loss": 0.0743,
      "step": 425
    },
    {
      "epoch": 2.9054054054054053,
      "grad_norm": 0.626979823529596,
      "learning_rate": 2.3351101918672985e-05,
      "loss": 0.0827,
      "step": 430
    },
    {
      "epoch": 2.939189189189189,
      "grad_norm": 0.7519449237853378,
      "learning_rate": 2.2858055476364822e-05,
      "loss": 0.0626,
      "step": 435
    },
    {
      "epoch": 2.972972972972973,
      "grad_norm": 0.6067887360251734,
      "learning_rate": 2.2367326487301317e-05,
      "loss": 0.065,
      "step": 440
    },
    {
      "epoch": 3.0067567567567566,
      "grad_norm": 0.43015841504665747,
      "learning_rate": 2.1879159943959686e-05,
      "loss": 0.0588,
      "step": 445
    },
    {
      "epoch": 3.0405405405405403,
      "grad_norm": 0.5166996425697965,
      "learning_rate": 2.139379955953686e-05,
      "loss": 0.0403,
      "step": 450
    },
    {
      "epoch": 3.074324324324324,
      "grad_norm": 1.533221487720139,
      "learning_rate": 2.0911487646277623e-05,
      "loss": 0.045,
      "step": 455
    },
    {
      "epoch": 3.108108108108108,
      "grad_norm": 1.3591414304325815,
      "learning_rate": 2.0432464994502203e-05,
      "loss": 0.0428,
      "step": 460
    },
    {
      "epoch": 3.141891891891892,
      "grad_norm": 0.4695191476769462,
      "learning_rate": 1.995697075239365e-05,
      "loss": 0.0397,
      "step": 465
    },
    {
      "epoch": 3.175675675675676,
      "grad_norm": 0.7689436070838513,
      "learning_rate": 1.9485242306605028e-05,
      "loss": 0.0388,
      "step": 470
    },
    {
      "epoch": 3.2094594594594597,
      "grad_norm": 0.5707450348910151,
      "learning_rate": 1.9017515163746058e-05,
      "loss": 0.0398,
      "step": 475
    },
    {
      "epoch": 3.2432432432432434,
      "grad_norm": 0.6311825777201817,
      "learning_rate": 1.855402283280836e-05,
      "loss": 0.0397,
      "step": 480
    },
    {
      "epoch": 3.277027027027027,
      "grad_norm": 0.6341709550171472,
      "learning_rate": 1.8094996708587958e-05,
      "loss": 0.0366,
      "step": 485
    },
    {
      "epoch": 3.310810810810811,
      "grad_norm": 0.37977836752911515,
      "learning_rate": 1.7640665956163306e-05,
      "loss": 0.0364,
      "step": 490
    },
    {
      "epoch": 3.3445945945945947,
      "grad_norm": 0.3786127389464406,
      "learning_rate": 1.719125739648648e-05,
      "loss": 0.0316,
      "step": 495
    },
    {
      "epoch": 3.3783783783783785,
      "grad_norm": 0.4003900559035512,
      "learning_rate": 1.6746995393144668e-05,
      "loss": 0.0348,
      "step": 500
    },
    {
      "epoch": 3.4121621621621623,
      "grad_norm": 0.7612366325017373,
      "learning_rate": 1.6308101740348433e-05,
      "loss": 0.0373,
      "step": 505
    },
    {
      "epoch": 3.445945945945946,
      "grad_norm": 0.5347572825256244,
      "learning_rate": 1.5874795552202773e-05,
      "loss": 0.0409,
      "step": 510
    },
    {
      "epoch": 3.47972972972973,
      "grad_norm": 0.3804098138882223,
      "learning_rate": 1.5447293153316163e-05,
      "loss": 0.0288,
      "step": 515
    },
    {
      "epoch": 3.5135135135135136,
      "grad_norm": 0.7846805185343385,
      "learning_rate": 1.5025807970802252e-05,
      "loss": 0.0367,
      "step": 520
    },
    {
      "epoch": 3.5472972972972974,
      "grad_norm": 0.6955590006040003,
      "learning_rate": 1.4610550427728103e-05,
      "loss": 0.0355,
      "step": 525
    },
    {
      "epoch": 3.581081081081081,
      "grad_norm": 0.6081379754765148,
      "learning_rate": 1.4201727838062181e-05,
      "loss": 0.0384,
      "step": 530
    },
    {
      "epoch": 3.614864864864865,
      "grad_norm": 0.6589981768125784,
      "learning_rate": 1.3799544303174514e-05,
      "loss": 0.0454,
      "step": 535
    },
    {
      "epoch": 3.6486486486486487,
      "grad_norm": 0.6007971088374728,
      "learning_rate": 1.3404200609940754e-05,
      "loss": 0.0397,
      "step": 540
    },
    {
      "epoch": 3.6824324324324325,
      "grad_norm": 0.7325992049485736,
      "learning_rate": 1.3015894130500977e-05,
      "loss": 0.0374,
      "step": 545
    },
    {
      "epoch": 3.7162162162162162,
      "grad_norm": 0.37094326467204003,
      "learning_rate": 1.2634818723723174e-05,
      "loss": 0.0318,
      "step": 550
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.9146146847365886,
      "learning_rate": 1.2261164638420832e-05,
      "loss": 0.0328,
      "step": 555
    },
    {
      "epoch": 3.7837837837837838,
      "grad_norm": 0.5648038565491271,
      "learning_rate": 1.1895118418372734e-05,
      "loss": 0.0356,
      "step": 560
    },
    {
      "epoch": 3.8175675675675675,
      "grad_norm": 0.37538544226428106,
      "learning_rate": 1.1536862809192518e-05,
      "loss": 0.0398,
      "step": 565
    },
    {
      "epoch": 3.8513513513513513,
      "grad_norm": 0.4038884653715223,
      "learning_rate": 1.1186576667094342e-05,
      "loss": 0.0271,
      "step": 570
    },
    {
      "epoch": 3.885135135135135,
      "grad_norm": 0.6627132587861069,
      "learning_rate": 1.0844434869600428e-05,
      "loss": 0.0338,
      "step": 575
    },
    {
      "epoch": 3.918918918918919,
      "grad_norm": 0.7041617251510983,
      "learning_rate": 1.0510608228234848e-05,
      "loss": 0.0391,
      "step": 580
    },
    {
      "epoch": 3.9527027027027026,
      "grad_norm": 0.8131142697010945,
      "learning_rate": 1.0185263403247256e-05,
      "loss": 0.0342,
      "step": 585
    },
    {
      "epoch": 3.9864864864864864,
      "grad_norm": 0.5920567955107612,
      "learning_rate": 9.868562820409103e-06,
      "loss": 0.0337,
      "step": 590
    },
    {
      "epoch": 4.02027027027027,
      "grad_norm": 0.3120828177149261,
      "learning_rate": 9.560664589923895e-06,
      "loss": 0.0226,
      "step": 595
    },
    {
      "epoch": 4.054054054054054,
      "grad_norm": 0.2900775148851475,
      "learning_rate": 9.261722427491953e-06,
      "loss": 0.0147,
      "step": 600
    },
    {
      "epoch": 4.087837837837838,
      "grad_norm": 0.22784089723926296,
      "learning_rate": 8.971885577569058e-06,
      "loss": 0.0122,
      "step": 605
    },
    {
      "epoch": 4.121621621621622,
      "grad_norm": 0.2693260557173335,
      "learning_rate": 8.691298738857432e-06,
      "loss": 0.0119,
      "step": 610
    },
    {
      "epoch": 4.155405405405405,
      "grad_norm": 0.21618402703820352,
      "learning_rate": 8.420101992066028e-06,
      "loss": 0.0156,
      "step": 615
    },
    {
      "epoch": 4.1891891891891895,
      "grad_norm": 0.24421224725586618,
      "learning_rate": 8.158430729976372e-06,
      "loss": 0.0119,
      "step": 620
    },
    {
      "epoch": 4.222972972972973,
      "grad_norm": 0.33320132947770426,
      "learning_rate": 7.906415589848834e-06,
      "loss": 0.0146,
      "step": 625
    },
    {
      "epoch": 4.256756756756757,
      "grad_norm": 0.33015864806759715,
      "learning_rate": 7.664182388203037e-06,
      "loss": 0.0145,
      "step": 630
    },
    {
      "epoch": 4.29054054054054,
      "grad_norm": 0.36028113378267884,
      "learning_rate": 7.4318520580049444e-06,
      "loss": 0.015,
      "step": 635
    },
    {
      "epoch": 4.324324324324325,
      "grad_norm": 0.26126678042031565,
      "learning_rate": 7.209540588292083e-06,
      "loss": 0.0134,
      "step": 640
    },
    {
      "epoch": 4.358108108108108,
      "grad_norm": 0.549701638585684,
      "learning_rate": 6.9973589662669455e-06,
      "loss": 0.0136,
      "step": 645
    },
    {
      "epoch": 4.391891891891892,
      "grad_norm": 0.30168954386975444,
      "learning_rate": 6.7954131218875404e-06,
      "loss": 0.0151,
      "step": 650
    },
    {
      "epoch": 4.425675675675675,
      "grad_norm": 0.43179940752419366,
      "learning_rate": 6.603803874982687e-06,
      "loss": 0.017,
      "step": 655
    },
    {
      "epoch": 4.45945945945946,
      "grad_norm": 0.26869880474008145,
      "learning_rate": 6.422626884918559e-06,
      "loss": 0.0131,
      "step": 660
    },
    {
      "epoch": 4.493243243243243,
      "grad_norm": 0.4055684732783103,
      "learning_rate": 6.2519726028415145e-06,
      "loss": 0.017,
      "step": 665
    },
    {
      "epoch": 4.527027027027027,
      "grad_norm": 0.334734546863203,
      "learning_rate": 6.091926226521089e-06,
      "loss": 0.0135,
      "step": 670
    },
    {
      "epoch": 4.5608108108108105,
      "grad_norm": 0.5546791873819881,
      "learning_rate": 5.942567657815696e-06,
      "loss": 0.0138,
      "step": 675
    },
    {
      "epoch": 4.594594594594595,
      "grad_norm": 0.4128562858606657,
      "learning_rate": 5.8039714627822754e-06,
      "loss": 0.0123,
      "step": 680
    },
    {
      "epoch": 4.628378378378378,
      "grad_norm": 0.33985067404728464,
      "learning_rate": 5.676206834449797e-06,
      "loss": 0.014,
      "step": 685
    },
    {
      "epoch": 4.662162162162162,
      "grad_norm": 0.3646278429623203,
      "learning_rate": 5.55933755827518e-06,
      "loss": 0.0105,
      "step": 690
    },
    {
      "epoch": 4.695945945945946,
      "grad_norm": 0.2663115053288746,
      "learning_rate": 5.453421980298957e-06,
      "loss": 0.012,
      "step": 695
    },
    {
      "epoch": 4.72972972972973,
      "grad_norm": 0.30374384905412244,
      "learning_rate": 5.358512978016445e-06,
      "loss": 0.0108,
      "step": 700
    },
    {
      "epoch": 4.763513513513513,
      "grad_norm": 0.47429438189169476,
      "learning_rate": 5.27465793397911e-06,
      "loss": 0.0143,
      "step": 705
    },
    {
      "epoch": 4.797297297297297,
      "grad_norm": 0.2683488363965782,
      "learning_rate": 5.201898712139201e-06,
      "loss": 0.0121,
      "step": 710
    },
    {
      "epoch": 4.831081081081081,
      "grad_norm": 0.4245927469089233,
      "learning_rate": 5.1402716369495194e-06,
      "loss": 0.0125,
      "step": 715
    },
    {
      "epoch": 4.864864864864865,
      "grad_norm": 0.40748023019876795,
      "learning_rate": 5.089807475228711e-06,
      "loss": 0.0151,
      "step": 720
    },
    {
      "epoch": 4.898648648648649,
      "grad_norm": 0.2454774836937338,
      "learning_rate": 5.050531420801205e-06,
      "loss": 0.0112,
      "step": 725
    },
    {
      "epoch": 4.9324324324324325,
      "grad_norm": 0.343369387002044,
      "learning_rate": 5.022463081919386e-06,
      "loss": 0.0126,
      "step": 730
    },
    {
      "epoch": 4.966216216216216,
      "grad_norm": 0.28684110351791725,
      "learning_rate": 5.005616471474332e-06,
      "loss": 0.0116,
      "step": 735
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.3365561300076367,
      "learning_rate": 5e-06,
      "loss": 0.013,
      "step": 740
    },
    {
      "epoch": 5.0,
      "step": 740,
      "total_flos": 519543052959744.0,
      "train_loss": 0.11201934994676628,
      "train_runtime": 13636.3107,
      "train_samples_per_second": 3.469,
      "train_steps_per_second": 0.054
    }
  ],
  "logging_steps": 5,
  "max_steps": 740,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 519543052959744.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}