{
  "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": 12.621006186744474,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 1.3412,
      "step": 5
    },
    {
      "epoch": 0.06756756756756757,
      "grad_norm": 4.439018018077387,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 0.5039,
      "step": 10
    },
    {
      "epoch": 0.10135135135135136,
      "grad_norm": 3.552982041839376,
      "learning_rate": 2.0270270270270273e-05,
      "loss": 0.352,
      "step": 15
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 4.01963825097133,
      "learning_rate": 2.702702702702703e-05,
      "loss": 0.2372,
      "step": 20
    },
    {
      "epoch": 0.16891891891891891,
      "grad_norm": 3.660650757844536,
      "learning_rate": 3.3783783783783784e-05,
      "loss": 0.2532,
      "step": 25
    },
    {
      "epoch": 0.20270270270270271,
      "grad_norm": 3.387355775674936,
      "learning_rate": 4.0540540540540545e-05,
      "loss": 0.2316,
      "step": 30
    },
    {
      "epoch": 0.23648648648648649,
      "grad_norm": 2.971426754139791,
      "learning_rate": 4.72972972972973e-05,
      "loss": 0.2242,
      "step": 35
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 2.720388820178826,
      "learning_rate": 4.9997978016429605e-05,
      "loss": 0.2266,
      "step": 40
    },
    {
      "epoch": 0.30405405405405406,
      "grad_norm": 2.0750360180986642,
      "learning_rate": 4.9985622766211935e-05,
      "loss": 0.1979,
      "step": 45
    },
    {
      "epoch": 0.33783783783783783,
      "grad_norm": 2.1876189675786044,
      "learning_rate": 4.996204175076325e-05,
      "loss": 0.1982,
      "step": 50
    },
    {
      "epoch": 0.3716216216216216,
      "grad_norm": 1.4405929660585144,
      "learning_rate": 4.99272467427147e-05,
      "loss": 0.2389,
      "step": 55
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 1.847810228455433,
      "learning_rate": 4.9881255113192526e-05,
      "loss": 0.1929,
      "step": 60
    },
    {
      "epoch": 0.4391891891891892,
      "grad_norm": 1.906432406199186,
      "learning_rate": 4.982408982314565e-05,
      "loss": 0.2042,
      "step": 65
    },
    {
      "epoch": 0.47297297297297297,
      "grad_norm": 1.3624755819479784,
      "learning_rate": 4.975577941188258e-05,
      "loss": 0.1992,
      "step": 70
    },
    {
      "epoch": 0.5067567567567568,
      "grad_norm": 1.5125128194000783,
      "learning_rate": 4.967635798282344e-05,
      "loss": 0.2384,
      "step": 75
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 1.31202154484301,
      "learning_rate": 4.958586518647407e-05,
      "loss": 0.1914,
      "step": 80
    },
    {
      "epoch": 0.5743243243243243,
      "grad_norm": 1.6966838556264587,
      "learning_rate": 4.9484346200630855e-05,
      "loss": 0.2113,
      "step": 85
    },
    {
      "epoch": 0.6081081081081081,
      "grad_norm": 1.442350215495238,
      "learning_rate": 4.937185170782607e-05,
      "loss": 0.1978,
      "step": 90
    },
    {
      "epoch": 0.6418918918918919,
      "grad_norm": 1.499234738886479,
      "learning_rate": 4.9248437870025035e-05,
      "loss": 0.2069,
      "step": 95
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 1.8030875469109664,
      "learning_rate": 4.911416630058772e-05,
      "loss": 0.1689,
      "step": 100
    },
    {
      "epoch": 0.7094594594594594,
      "grad_norm": 8.10293651733264,
      "learning_rate": 4.896910403350873e-05,
      "loss": 0.3606,
      "step": 105
    },
    {
      "epoch": 0.7432432432432432,
      "grad_norm": 1.3802313535536794,
      "learning_rate": 4.88133234899512e-05,
      "loss": 0.2129,
      "step": 110
    },
    {
      "epoch": 0.777027027027027,
      "grad_norm": 1.6352039566876417,
      "learning_rate": 4.864690244209105e-05,
      "loss": 0.1701,
      "step": 115
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 1.447661676234873,
      "learning_rate": 4.8469923974289874e-05,
      "loss": 0.2402,
      "step": 120
    },
    {
      "epoch": 0.8445945945945946,
      "grad_norm": 1.334557195664052,
      "learning_rate": 4.828247644161577e-05,
      "loss": 0.2054,
      "step": 125
    },
    {
      "epoch": 0.8783783783783784,
      "grad_norm": 1.225300086570123,
      "learning_rate": 4.808465342573274e-05,
      "loss": 0.1909,
      "step": 130
    },
    {
      "epoch": 0.9121621621621622,
      "grad_norm": 1.492560775790108,
      "learning_rate": 4.787655368818087e-05,
      "loss": 0.174,
      "step": 135
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 1.0899068834797303,
      "learning_rate": 4.765828112107034e-05,
      "loss": 0.203,
      "step": 140
    },
    {
      "epoch": 0.9797297297297297,
      "grad_norm": 1.366371773428974,
      "learning_rate": 4.742994469521421e-05,
      "loss": 0.1957,
      "step": 145
    },
    {
      "epoch": 1.0135135135135136,
      "grad_norm": 0.7472054632033055,
      "learning_rate": 4.719165840572557e-05,
      "loss": 0.1764,
      "step": 150
    },
    {
      "epoch": 1.0472972972972974,
      "grad_norm": 1.096606205402809,
      "learning_rate": 4.694354121510644e-05,
      "loss": 0.1136,
      "step": 155
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 1.0525994837117485,
      "learning_rate": 4.668571699385668e-05,
      "loss": 0.1245,
      "step": 160
    },
    {
      "epoch": 1.114864864864865,
      "grad_norm": 1.0258352875007604,
      "learning_rate": 4.641831445863265e-05,
      "loss": 0.1224,
      "step": 165
    },
    {
      "epoch": 1.1486486486486487,
      "grad_norm": 1.0118533178220892,
      "learning_rate": 4.614146710798645e-05,
      "loss": 0.1233,
      "step": 170
    },
    {
      "epoch": 1.1824324324324325,
      "grad_norm": 1.3377148135130936,
      "learning_rate": 4.585531315571788e-05,
      "loss": 0.1245,
      "step": 175
    },
    {
      "epoch": 1.2162162162162162,
      "grad_norm": 1.2231503647700532,
      "learning_rate": 4.555999546187229e-05,
      "loss": 0.1153,
      "step": 180
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.2952625737397045,
      "learning_rate": 4.5255661461418854e-05,
      "loss": 0.1058,
      "step": 185
    },
    {
      "epoch": 1.2837837837837838,
      "grad_norm": 1.0339542892150495,
      "learning_rate": 4.4942463090644896e-05,
      "loss": 0.1153,
      "step": 190
    },
    {
      "epoch": 1.3175675675675675,
      "grad_norm": 1.0722232570490027,
      "learning_rate": 4.462055671130289e-05,
      "loss": 0.1204,
      "step": 195
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 1.0024563232848254,
      "learning_rate": 4.4290103032548094e-05,
      "loss": 0.115,
      "step": 200
    },
    {
      "epoch": 1.385135135135135,
      "grad_norm": 1.698722911031088,
      "learning_rate": 4.395126703070589e-05,
      "loss": 0.125,
      "step": 205
    },
    {
      "epoch": 1.4189189189189189,
      "grad_norm": 0.9849447980075114,
      "learning_rate": 4.360421786690862e-05,
      "loss": 0.1204,
      "step": 210
    },
    {
      "epoch": 1.4527027027027026,
      "grad_norm": 0.8335553215898444,
      "learning_rate": 4.324912880264326e-05,
      "loss": 0.1164,
      "step": 215
    },
    {
      "epoch": 1.4864864864864864,
      "grad_norm": 0.7791030356965772,
      "learning_rate": 4.288617711325207e-05,
      "loss": 0.108,
      "step": 220
    },
    {
      "epoch": 1.5202702702702702,
      "grad_norm": 0.7851520577873489,
      "learning_rate": 4.251554399942928e-05,
      "loss": 0.1155,
      "step": 225
    },
    {
      "epoch": 1.554054054054054,
      "grad_norm": 0.903300450787851,
      "learning_rate": 4.21374144967581e-05,
      "loss": 0.1164,
      "step": 230
    },
    {
      "epoch": 1.5878378378378377,
      "grad_norm": 0.7456489319041527,
      "learning_rate": 4.1751977383333224e-05,
      "loss": 0.1069,
      "step": 235
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 1.0831836156816486,
      "learning_rate": 4.1359425085514906e-05,
      "loss": 0.113,
      "step": 240
    },
    {
      "epoch": 1.6554054054054053,
      "grad_norm": 0.9947103365276501,
      "learning_rate": 4.095995358186162e-05,
      "loss": 0.1223,
      "step": 245
    },
    {
      "epoch": 1.689189189189189,
      "grad_norm": 0.7930807717454621,
      "learning_rate": 4.055376230528936e-05,
      "loss": 0.0985,
      "step": 250
    },
    {
      "epoch": 1.722972972972973,
      "grad_norm": 0.8081662430780697,
      "learning_rate": 4.0141054043506406e-05,
      "loss": 0.1224,
      "step": 255
    },
    {
      "epoch": 1.7567567567567568,
      "grad_norm": 0.8136722738341788,
      "learning_rate": 3.972203483777315e-05,
      "loss": 0.1049,
      "step": 260
    },
    {
      "epoch": 1.7905405405405406,
      "grad_norm": 1.1697242547035394,
      "learning_rate": 3.929691388003772e-05,
      "loss": 0.1113,
      "step": 265
    },
    {
      "epoch": 1.8243243243243243,
      "grad_norm": 0.7908495493739465,
      "learning_rate": 3.886590340849852e-05,
      "loss": 0.092,
      "step": 270
    },
    {
      "epoch": 1.8581081081081081,
      "grad_norm": 1.0742819020203385,
      "learning_rate": 3.842921860164607e-05,
      "loss": 0.1019,
      "step": 275
    },
    {
      "epoch": 1.8918918918918919,
      "grad_norm": 0.8220992017657939,
      "learning_rate": 3.798707747083694e-05,
      "loss": 0.1047,
      "step": 280
    },
    {
      "epoch": 1.9256756756756757,
      "grad_norm": 0.734912679656405,
      "learning_rate": 3.753970075145322e-05,
      "loss": 0.1037,
      "step": 285
    },
    {
      "epoch": 1.9594594594594594,
      "grad_norm": 0.8174324747248585,
      "learning_rate": 3.7087311792702265e-05,
      "loss": 0.1059,
      "step": 290
    },
    {
      "epoch": 1.9932432432432432,
      "grad_norm": 0.9145873684161004,
      "learning_rate": 3.663013644611139e-05,
      "loss": 0.1005,
      "step": 295
    },
    {
      "epoch": 2.027027027027027,
      "grad_norm": 0.6551954661234566,
      "learning_rate": 3.616840295277328e-05,
      "loss": 0.065,
      "step": 300
    },
    {
      "epoch": 2.060810810810811,
      "grad_norm": 1.3033154952287334,
      "learning_rate": 3.5702341829398525e-05,
      "loss": 0.0642,
      "step": 305
    },
    {
      "epoch": 2.0945945945945947,
      "grad_norm": 0.6098962849414549,
      "learning_rate": 3.523218575323198e-05,
      "loss": 0.0668,
      "step": 310
    },
    {
      "epoch": 2.1283783783783785,
      "grad_norm": 0.8250176277136996,
      "learning_rate": 3.475816944589058e-05,
      "loss": 0.0645,
      "step": 315
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 0.7050979750969496,
      "learning_rate": 3.4280529556180404e-05,
      "loss": 0.0677,
      "step": 320
    },
    {
      "epoch": 2.195945945945946,
      "grad_norm": 0.81626878946061,
      "learning_rate": 3.379950454195172e-05,
      "loss": 0.0548,
      "step": 325
    },
    {
      "epoch": 2.22972972972973,
      "grad_norm": 0.6774465388338856,
      "learning_rate": 3.331533455105084e-05,
      "loss": 0.0641,
      "step": 330
    },
    {
      "epoch": 2.2635135135135136,
      "grad_norm": 0.6717523431459212,
      "learning_rate": 3.2828261301428206e-05,
      "loss": 0.0698,
      "step": 335
    },
    {
      "epoch": 2.2972972972972974,
      "grad_norm": 0.6017205479686647,
      "learning_rate": 3.23385279604627e-05,
      "loss": 0.0535,
      "step": 340
    },
    {
      "epoch": 2.331081081081081,
      "grad_norm": 0.5836320359501072,
      "learning_rate": 3.18463790235623e-05,
      "loss": 0.063,
      "step": 345
    },
    {
      "epoch": 2.364864864864865,
      "grad_norm": 0.5556761669289445,
      "learning_rate": 3.135206019210167e-05,
      "loss": 0.0584,
      "step": 350
    },
    {
      "epoch": 2.3986486486486487,
      "grad_norm": 0.6489325331638421,
      "learning_rate": 3.085581825075782e-05,
      "loss": 0.057,
      "step": 355
    },
    {
      "epoch": 2.4324324324324325,
      "grad_norm": 1.0242942453940833,
      "learning_rate": 3.0357900944304774e-05,
      "loss": 0.0628,
      "step": 360
    },
    {
      "epoch": 2.4662162162162162,
      "grad_norm": 0.7315103373219937,
      "learning_rate": 2.9858556853929048e-05,
      "loss": 0.0671,
      "step": 365
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.5190642691663524,
      "learning_rate": 2.9358035273127483e-05,
      "loss": 0.0524,
      "step": 370
    },
    {
      "epoch": 2.5337837837837838,
      "grad_norm": 0.5763271776110046,
      "learning_rate": 2.8856586083249487e-05,
      "loss": 0.05,
      "step": 375
    },
    {
      "epoch": 2.5675675675675675,
      "grad_norm": 0.7672335789715156,
      "learning_rate": 2.83544596287458e-05,
      "loss": 0.0652,
      "step": 380
    },
    {
      "epoch": 2.6013513513513513,
      "grad_norm": 0.6370728790883294,
      "learning_rate": 2.785190659218604e-05,
      "loss": 0.0586,
      "step": 385
    },
    {
      "epoch": 2.635135135135135,
      "grad_norm": 0.6398265727137722,
      "learning_rate": 2.7349177869107462e-05,
      "loss": 0.061,
      "step": 390
    },
    {
      "epoch": 2.668918918918919,
      "grad_norm": 0.5524612383958989,
      "learning_rate": 2.684652444275741e-05,
      "loss": 0.0496,
      "step": 395
    },
    {
      "epoch": 2.7027027027027026,
      "grad_norm": 0.5670334737977824,
      "learning_rate": 2.634419725879193e-05,
      "loss": 0.0475,
      "step": 400
    },
    {
      "epoch": 2.7364864864864864,
      "grad_norm": 0.6466324477989005,
      "learning_rate": 2.58424470999932e-05,
      "loss": 0.0608,
      "step": 405
    },
    {
      "epoch": 2.77027027027027,
      "grad_norm": 0.6629874864808576,
      "learning_rate": 2.534152446106825e-05,
      "loss": 0.0587,
      "step": 410
    },
    {
      "epoch": 2.804054054054054,
      "grad_norm": 0.566115840572953,
      "learning_rate": 2.4841679423591523e-05,
      "loss": 0.0515,
      "step": 415
    },
    {
      "epoch": 2.8378378378378377,
      "grad_norm": 0.7183565847428235,
      "learning_rate": 2.4343161531153647e-05,
      "loss": 0.0603,
      "step": 420
    },
    {
      "epoch": 2.8716216216216215,
      "grad_norm": 0.5426694416307323,
      "learning_rate": 2.3846219664778824e-05,
      "loss": 0.0605,
      "step": 425
    },
    {
      "epoch": 2.9054054054054053,
      "grad_norm": 0.5593502911832727,
      "learning_rate": 2.3351101918672985e-05,
      "loss": 0.0591,
      "step": 430
    },
    {
      "epoch": 2.939189189189189,
      "grad_norm": 0.6027997142648732,
      "learning_rate": 2.2858055476364822e-05,
      "loss": 0.0505,
      "step": 435
    },
    {
      "epoch": 2.972972972972973,
      "grad_norm": 0.5030037170646953,
      "learning_rate": 2.2367326487301317e-05,
      "loss": 0.0466,
      "step": 440
    },
    {
      "epoch": 3.0067567567567566,
      "grad_norm": 0.49852590311868833,
      "learning_rate": 2.1879159943959686e-05,
      "loss": 0.0414,
      "step": 445
    },
    {
      "epoch": 3.0405405405405403,
      "grad_norm": 0.4323876114057528,
      "learning_rate": 2.139379955953686e-05,
      "loss": 0.0281,
      "step": 450
    },
    {
      "epoch": 3.074324324324324,
      "grad_norm": 0.4541697603850849,
      "learning_rate": 2.0911487646277623e-05,
      "loss": 0.0295,
      "step": 455
    },
    {
      "epoch": 3.108108108108108,
      "grad_norm": 0.4521805274146651,
      "learning_rate": 2.0432464994502203e-05,
      "loss": 0.0236,
      "step": 460
    },
    {
      "epoch": 3.141891891891892,
      "grad_norm": 0.48267823688054673,
      "learning_rate": 1.995697075239365e-05,
      "loss": 0.0298,
      "step": 465
    },
    {
      "epoch": 3.175675675675676,
      "grad_norm": 0.5203816613612195,
      "learning_rate": 1.9485242306605028e-05,
      "loss": 0.0285,
      "step": 470
    },
    {
      "epoch": 3.2094594594594597,
      "grad_norm": 0.3789443267017963,
      "learning_rate": 1.9017515163746058e-05,
      "loss": 0.0274,
      "step": 475
    },
    {
      "epoch": 3.2432432432432434,
      "grad_norm": 0.41224542055174457,
      "learning_rate": 1.855402283280836e-05,
      "loss": 0.028,
      "step": 480
    },
    {
      "epoch": 3.277027027027027,
      "grad_norm": 0.5925796703969819,
      "learning_rate": 1.8094996708587958e-05,
      "loss": 0.0264,
      "step": 485
    },
    {
      "epoch": 3.310810810810811,
      "grad_norm": 0.45808561787108665,
      "learning_rate": 1.7640665956163306e-05,
      "loss": 0.0277,
      "step": 490
    },
    {
      "epoch": 3.3445945945945947,
      "grad_norm": 0.4914906943906507,
      "learning_rate": 1.719125739648648e-05,
      "loss": 0.0245,
      "step": 495
    },
    {
      "epoch": 3.3783783783783785,
      "grad_norm": 0.5105302563432415,
      "learning_rate": 1.6746995393144668e-05,
      "loss": 0.0243,
      "step": 500
    },
    {
      "epoch": 3.4121621621621623,
      "grad_norm": 0.46664138904641295,
      "learning_rate": 1.6308101740348433e-05,
      "loss": 0.0253,
      "step": 505
    },
    {
      "epoch": 3.445945945945946,
      "grad_norm": 0.4930512078926301,
      "learning_rate": 1.5874795552202773e-05,
      "loss": 0.0287,
      "step": 510
    },
    {
      "epoch": 3.47972972972973,
      "grad_norm": 0.5750190708086297,
      "learning_rate": 1.5447293153316163e-05,
      "loss": 0.0247,
      "step": 515
    },
    {
      "epoch": 3.5135135135135136,
      "grad_norm": 0.4299195312738642,
      "learning_rate": 1.5025807970802252e-05,
      "loss": 0.025,
      "step": 520
    },
    {
      "epoch": 3.5472972972972974,
      "grad_norm": 0.4705848484203161,
      "learning_rate": 1.4610550427728103e-05,
      "loss": 0.0224,
      "step": 525
    },
    {
      "epoch": 3.581081081081081,
      "grad_norm": 0.4875384820191375,
      "learning_rate": 1.4201727838062181e-05,
      "loss": 0.0271,
      "step": 530
    },
    {
      "epoch": 3.614864864864865,
      "grad_norm": 0.4799225527601825,
      "learning_rate": 1.3799544303174514e-05,
      "loss": 0.0314,
      "step": 535
    },
    {
      "epoch": 3.6486486486486487,
      "grad_norm": 0.5268517933993498,
      "learning_rate": 1.3404200609940754e-05,
      "loss": 0.0295,
      "step": 540
    },
    {
      "epoch": 3.6824324324324325,
      "grad_norm": 0.48358322486680966,
      "learning_rate": 1.3015894130500977e-05,
      "loss": 0.0258,
      "step": 545
    },
    {
      "epoch": 3.7162162162162162,
      "grad_norm": 0.45660895428611165,
      "learning_rate": 1.2634818723723174e-05,
      "loss": 0.0251,
      "step": 550
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.6337079251512344,
      "learning_rate": 1.2261164638420832e-05,
      "loss": 0.0258,
      "step": 555
    },
    {
      "epoch": 3.7837837837837838,
      "grad_norm": 0.5808617819017178,
      "learning_rate": 1.1895118418372734e-05,
      "loss": 0.0253,
      "step": 560
    },
    {
      "epoch": 3.8175675675675675,
      "grad_norm": 0.4794592149017628,
      "learning_rate": 1.1536862809192518e-05,
      "loss": 0.0278,
      "step": 565
    },
    {
      "epoch": 3.8513513513513513,
      "grad_norm": 0.38444321464200576,
      "learning_rate": 1.1186576667094342e-05,
      "loss": 0.0219,
      "step": 570
    },
    {
      "epoch": 3.885135135135135,
      "grad_norm": 0.43815837663453977,
      "learning_rate": 1.0844434869600428e-05,
      "loss": 0.0258,
      "step": 575
    },
    {
      "epoch": 3.918918918918919,
      "grad_norm": 0.5514451957555104,
      "learning_rate": 1.0510608228234848e-05,
      "loss": 0.026,
      "step": 580
    },
    {
      "epoch": 3.9527027027027026,
      "grad_norm": 0.5614013829897044,
      "learning_rate": 1.0185263403247256e-05,
      "loss": 0.0279,
      "step": 585
    },
    {
      "epoch": 3.9864864864864864,
      "grad_norm": 0.4309993163791908,
      "learning_rate": 9.868562820409103e-06,
      "loss": 0.0236,
      "step": 590
    },
    {
      "epoch": 4.02027027027027,
      "grad_norm": 0.3694546301591644,
      "learning_rate": 9.560664589923895e-06,
      "loss": 0.018,
      "step": 595
    },
    {
      "epoch": 4.054054054054054,
      "grad_norm": 0.32081562696250526,
      "learning_rate": 9.261722427491953e-06,
      "loss": 0.0118,
      "step": 600
    },
    {
      "epoch": 4.087837837837838,
      "grad_norm": 0.25798129529445024,
      "learning_rate": 8.971885577569058e-06,
      "loss": 0.0103,
      "step": 605
    },
    {
      "epoch": 4.121621621621622,
      "grad_norm": 0.27127431922061396,
      "learning_rate": 8.691298738857432e-06,
      "loss": 0.011,
      "step": 610
    },
    {
      "epoch": 4.155405405405405,
      "grad_norm": 0.2639376342400401,
      "learning_rate": 8.420101992066028e-06,
      "loss": 0.0124,
      "step": 615
    },
    {
      "epoch": 4.1891891891891895,
      "grad_norm": 0.32468635218976477,
      "learning_rate": 8.158430729976372e-06,
      "loss": 0.0102,
      "step": 620
    },
    {
      "epoch": 4.222972972972973,
      "grad_norm": 0.29954422733054015,
      "learning_rate": 7.906415589848834e-06,
      "loss": 0.0089,
      "step": 625
    },
    {
      "epoch": 4.256756756756757,
      "grad_norm": 0.27369651618976626,
      "learning_rate": 7.664182388203037e-06,
      "loss": 0.0109,
      "step": 630
    },
    {
      "epoch": 4.29054054054054,
      "grad_norm": 0.4357364688045452,
      "learning_rate": 7.4318520580049444e-06,
      "loss": 0.0129,
      "step": 635
    },
    {
      "epoch": 4.324324324324325,
      "grad_norm": 0.2817491467718038,
      "learning_rate": 7.209540588292083e-06,
      "loss": 0.0107,
      "step": 640
    },
    {
      "epoch": 4.358108108108108,
      "grad_norm": 0.1994039245459065,
      "learning_rate": 6.9973589662669455e-06,
      "loss": 0.0085,
      "step": 645
    },
    {
      "epoch": 4.391891891891892,
      "grad_norm": 0.2969666151220339,
      "learning_rate": 6.7954131218875404e-06,
      "loss": 0.012,
      "step": 650
    },
    {
      "epoch": 4.425675675675675,
      "grad_norm": 0.45430115974806207,
      "learning_rate": 6.603803874982687e-06,
      "loss": 0.0134,
      "step": 655
    },
    {
      "epoch": 4.45945945945946,
      "grad_norm": 0.37183133433063675,
      "learning_rate": 6.422626884918559e-06,
      "loss": 0.0101,
      "step": 660
    },
    {
      "epoch": 4.493243243243243,
      "grad_norm": 0.4467174694705207,
      "learning_rate": 6.2519726028415145e-06,
      "loss": 0.0091,
      "step": 665
    },
    {
      "epoch": 4.527027027027027,
      "grad_norm": 0.36401832143798657,
      "learning_rate": 6.091926226521089e-06,
      "loss": 0.011,
      "step": 670
    },
    {
      "epoch": 4.5608108108108105,
      "grad_norm": 0.36310471427170216,
      "learning_rate": 5.942567657815696e-06,
      "loss": 0.01,
      "step": 675
    },
    {
      "epoch": 4.594594594594595,
      "grad_norm": 0.35651788551881625,
      "learning_rate": 5.8039714627822754e-06,
      "loss": 0.0086,
      "step": 680
    },
    {
      "epoch": 4.628378378378378,
      "grad_norm": 0.3845984404770356,
      "learning_rate": 5.676206834449797e-06,
      "loss": 0.0101,
      "step": 685
    },
    {
      "epoch": 4.662162162162162,
      "grad_norm": 0.27019219415223567,
      "learning_rate": 5.55933755827518e-06,
      "loss": 0.008,
      "step": 690
    },
    {
      "epoch": 4.695945945945946,
      "grad_norm": 0.34211781250715345,
      "learning_rate": 5.453421980298957e-06,
      "loss": 0.0092,
      "step": 695
    },
    {
      "epoch": 4.72972972972973,
      "grad_norm": 0.30161807459303475,
      "learning_rate": 5.358512978016445e-06,
      "loss": 0.0105,
      "step": 700
    },
    {
      "epoch": 4.763513513513513,
      "grad_norm": 0.4873844480794417,
      "learning_rate": 5.27465793397911e-06,
      "loss": 0.0107,
      "step": 705
    },
    {
      "epoch": 4.797297297297297,
      "grad_norm": 0.26520217431622456,
      "learning_rate": 5.201898712139201e-06,
      "loss": 0.0091,
      "step": 710
    },
    {
      "epoch": 4.831081081081081,
      "grad_norm": 0.4915175457694254,
      "learning_rate": 5.1402716369495194e-06,
      "loss": 0.0081,
      "step": 715
    },
    {
      "epoch": 4.864864864864865,
      "grad_norm": 0.35656355334133155,
      "learning_rate": 5.089807475228711e-06,
      "loss": 0.0124,
      "step": 720
    },
    {
      "epoch": 4.898648648648649,
      "grad_norm": 0.34007651253392635,
      "learning_rate": 5.050531420801205e-06,
      "loss": 0.009,
      "step": 725
    },
    {
      "epoch": 4.9324324324324325,
      "grad_norm": 0.29734787018395237,
      "learning_rate": 5.022463081919386e-06,
      "loss": 0.0107,
      "step": 730
    },
    {
      "epoch": 4.966216216216216,
      "grad_norm": 0.2534028775081779,
      "learning_rate": 5.005616471474332e-06,
      "loss": 0.008,
      "step": 735
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2546674518371035,
      "learning_rate": 5e-06,
      "loss": 0.0076,
      "step": 740
    },
    {
      "epoch": 5.0,
      "step": 740,
      "total_flos": 697349512364032.0,
      "train_loss": 0.09491675717605128,
      "train_runtime": 14744.8121,
      "train_samples_per_second": 3.208,
      "train_steps_per_second": 0.05
    }
  ],
  "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": 697349512364032.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}