{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.3987563155849203,
  "eval_steps": 500,
  "global_step": 1800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00777302759424796,
      "grad_norm": 19.018803349985753,
      "learning_rate": 8.737864077669904e-07,
      "loss": 8.2473,
      "step": 10
    },
    {
      "epoch": 0.01554605518849592,
      "grad_norm": 17.967208179671328,
      "learning_rate": 1.8446601941747572e-06,
      "loss": 8.1304,
      "step": 20
    },
    {
      "epoch": 0.023319082782743878,
      "grad_norm": 9.21576906708124,
      "learning_rate": 2.8155339805825245e-06,
      "loss": 7.4913,
      "step": 30
    },
    {
      "epoch": 0.03109211037699184,
      "grad_norm": 4.8422984662740465,
      "learning_rate": 3.7864077669902915e-06,
      "loss": 6.7578,
      "step": 40
    },
    {
      "epoch": 0.0388651379712398,
      "grad_norm": 3.0883237528757603,
      "learning_rate": 4.7572815533980585e-06,
      "loss": 6.2789,
      "step": 50
    },
    {
      "epoch": 0.046638165565487756,
      "grad_norm": 3.0641013238554837,
      "learning_rate": 5.728155339805825e-06,
      "loss": 5.9381,
      "step": 60
    },
    {
      "epoch": 0.054411193159735714,
      "grad_norm": 3.6989178160715803,
      "learning_rate": 6.699029126213592e-06,
      "loss": 5.547,
      "step": 70
    },
    {
      "epoch": 0.06218422075398368,
      "grad_norm": 3.8926133061231023,
      "learning_rate": 7.66990291262136e-06,
      "loss": 4.9449,
      "step": 80
    },
    {
      "epoch": 0.06995724834823164,
      "grad_norm": 6.005297193380389,
      "learning_rate": 8.640776699029127e-06,
      "loss": 4.2741,
      "step": 90
    },
    {
      "epoch": 0.0777302759424796,
      "grad_norm": 4.634097942418493,
      "learning_rate": 9.611650485436892e-06,
      "loss": 3.692,
      "step": 100
    },
    {
      "epoch": 0.08550330353672755,
      "grad_norm": 3.9161281136802915,
      "learning_rate": 1.0582524271844662e-05,
      "loss": 3.3021,
      "step": 110
    },
    {
      "epoch": 0.09327633113097551,
      "grad_norm": 2.668188197386523,
      "learning_rate": 1.1553398058252427e-05,
      "loss": 3.0699,
      "step": 120
    },
    {
      "epoch": 0.10104935872522347,
      "grad_norm": 5.67542909747789,
      "learning_rate": 1.2524271844660197e-05,
      "loss": 2.9232,
      "step": 130
    },
    {
      "epoch": 0.10882238631947143,
      "grad_norm": 8.358891491079559,
      "learning_rate": 1.3495145631067962e-05,
      "loss": 2.8142,
      "step": 140
    },
    {
      "epoch": 0.11659541391371939,
      "grad_norm": 7.521330978238392,
      "learning_rate": 1.446601941747573e-05,
      "loss": 2.736,
      "step": 150
    },
    {
      "epoch": 0.12436844150796736,
      "grad_norm": 16.178722459983547,
      "learning_rate": 1.5436893203883496e-05,
      "loss": 2.6651,
      "step": 160
    },
    {
      "epoch": 0.1321414691022153,
      "grad_norm": 8.335965189770683,
      "learning_rate": 1.6407766990291263e-05,
      "loss": 2.6021,
      "step": 170
    },
    {
      "epoch": 0.13991449669646328,
      "grad_norm": 8.022474097021126,
      "learning_rate": 1.737864077669903e-05,
      "loss": 2.5458,
      "step": 180
    },
    {
      "epoch": 0.14768752429071122,
      "grad_norm": 11.647238033711355,
      "learning_rate": 1.8349514563106798e-05,
      "loss": 2.4889,
      "step": 190
    },
    {
      "epoch": 0.1554605518849592,
      "grad_norm": 11.910883080289608,
      "learning_rate": 1.9320388349514565e-05,
      "loss": 2.4326,
      "step": 200
    },
    {
      "epoch": 0.16323357947920716,
      "grad_norm": 11.410643498860912,
      "learning_rate": 2.0291262135922333e-05,
      "loss": 2.3697,
      "step": 210
    },
    {
      "epoch": 0.1710066070734551,
      "grad_norm": 11.108752154756692,
      "learning_rate": 2.1262135922330097e-05,
      "loss": 2.309,
      "step": 220
    },
    {
      "epoch": 0.17877963466770308,
      "grad_norm": 19.998718279627422,
      "learning_rate": 2.2233009708737864e-05,
      "loss": 2.241,
      "step": 230
    },
    {
      "epoch": 0.18655266226195102,
      "grad_norm": 11.758179967806427,
      "learning_rate": 2.3203883495145632e-05,
      "loss": 2.1862,
      "step": 240
    },
    {
      "epoch": 0.194325689856199,
      "grad_norm": 13.686852090688575,
      "learning_rate": 2.41747572815534e-05,
      "loss": 2.1354,
      "step": 250
    },
    {
      "epoch": 0.20209871745044694,
      "grad_norm": 7.097958211758118,
      "learning_rate": 2.5145631067961167e-05,
      "loss": 2.0252,
      "step": 260
    },
    {
      "epoch": 0.2098717450446949,
      "grad_norm": 12.553623894402433,
      "learning_rate": 2.6116504854368934e-05,
      "loss": 1.92,
      "step": 270
    },
    {
      "epoch": 0.21764477263894286,
      "grad_norm": 7.033016298250838,
      "learning_rate": 2.7087378640776702e-05,
      "loss": 1.832,
      "step": 280
    },
    {
      "epoch": 0.22541780023319083,
      "grad_norm": 8.05454776441266,
      "learning_rate": 2.8058252427184466e-05,
      "loss": 1.7505,
      "step": 290
    },
    {
      "epoch": 0.23319082782743877,
      "grad_norm": 6.32342513670527,
      "learning_rate": 2.9029126213592233e-05,
      "loss": 1.6596,
      "step": 300
    },
    {
      "epoch": 0.24096385542168675,
      "grad_norm": 6.0607588085936435,
      "learning_rate": 3e-05,
      "loss": 1.5836,
      "step": 310
    },
    {
      "epoch": 0.24873688301593472,
      "grad_norm": 7.205879956336063,
      "learning_rate": 3.097087378640777e-05,
      "loss": 1.5148,
      "step": 320
    },
    {
      "epoch": 0.2565099106101827,
      "grad_norm": 5.233158820350977,
      "learning_rate": 3.194174757281553e-05,
      "loss": 1.4655,
      "step": 330
    },
    {
      "epoch": 0.2642829382044306,
      "grad_norm": 6.5668167184777495,
      "learning_rate": 3.29126213592233e-05,
      "loss": 1.3996,
      "step": 340
    },
    {
      "epoch": 0.2720559657986786,
      "grad_norm": 5.866172867997477,
      "learning_rate": 3.3883495145631074e-05,
      "loss": 1.3546,
      "step": 350
    },
    {
      "epoch": 0.27982899339292655,
      "grad_norm": 7.078627326366601,
      "learning_rate": 3.485436893203884e-05,
      "loss": 1.2957,
      "step": 360
    },
    {
      "epoch": 0.2876020209871745,
      "grad_norm": 7.013906915714377,
      "learning_rate": 3.58252427184466e-05,
      "loss": 1.265,
      "step": 370
    },
    {
      "epoch": 0.29537504858142244,
      "grad_norm": 5.721220928268079,
      "learning_rate": 3.679611650485437e-05,
      "loss": 1.2163,
      "step": 380
    },
    {
      "epoch": 0.3031480761756704,
      "grad_norm": 5.571852622479908,
      "learning_rate": 3.776699029126214e-05,
      "loss": 1.163,
      "step": 390
    },
    {
      "epoch": 0.3109211037699184,
      "grad_norm": 5.4843565963743295,
      "learning_rate": 3.873786407766991e-05,
      "loss": 1.1277,
      "step": 400
    },
    {
      "epoch": 0.31869413136416636,
      "grad_norm": 7.069970365890535,
      "learning_rate": 3.970873786407767e-05,
      "loss": 1.1094,
      "step": 410
    },
    {
      "epoch": 0.3264671589584143,
      "grad_norm": 5.425993491783754,
      "learning_rate": 4.0679611650485436e-05,
      "loss": 1.069,
      "step": 420
    },
    {
      "epoch": 0.33424018655266224,
      "grad_norm": 5.03054433074538,
      "learning_rate": 4.165048543689321e-05,
      "loss": 1.0382,
      "step": 430
    },
    {
      "epoch": 0.3420132141469102,
      "grad_norm": 7.103802226586121,
      "learning_rate": 4.262135922330097e-05,
      "loss": 0.9988,
      "step": 440
    },
    {
      "epoch": 0.3497862417411582,
      "grad_norm": 5.180871741408765,
      "learning_rate": 4.3592233009708735e-05,
      "loss": 0.979,
      "step": 450
    },
    {
      "epoch": 0.35755926933540616,
      "grad_norm": 5.168730423055745,
      "learning_rate": 4.4563106796116506e-05,
      "loss": 0.9442,
      "step": 460
    },
    {
      "epoch": 0.3653322969296541,
      "grad_norm": 4.509772926215358,
      "learning_rate": 4.553398058252427e-05,
      "loss": 0.9304,
      "step": 470
    },
    {
      "epoch": 0.37310532452390205,
      "grad_norm": 5.497064414551351,
      "learning_rate": 4.650485436893204e-05,
      "loss": 0.8971,
      "step": 480
    },
    {
      "epoch": 0.38087835211815,
      "grad_norm": 4.2878368849547375,
      "learning_rate": 4.747572815533981e-05,
      "loss": 0.8754,
      "step": 490
    },
    {
      "epoch": 0.388651379712398,
      "grad_norm": 5.023778493195039,
      "learning_rate": 4.8446601941747576e-05,
      "loss": 0.8535,
      "step": 500
    },
    {
      "epoch": 0.39642440730664597,
      "grad_norm": 4.85429288760384,
      "learning_rate": 4.941747572815535e-05,
      "loss": 0.8176,
      "step": 510
    },
    {
      "epoch": 0.4041974349008939,
      "grad_norm": 4.78285229843812,
      "learning_rate": 5.0388349514563104e-05,
      "loss": 0.8114,
      "step": 520
    },
    {
      "epoch": 0.41197046249514185,
      "grad_norm": 5.101751542601933,
      "learning_rate": 5.135922330097087e-05,
      "loss": 0.7752,
      "step": 530
    },
    {
      "epoch": 0.4197434900893898,
      "grad_norm": 5.448054773232898,
      "learning_rate": 5.233009708737864e-05,
      "loss": 0.755,
      "step": 540
    },
    {
      "epoch": 0.4275165176836378,
      "grad_norm": 4.141934961389859,
      "learning_rate": 5.3300970873786417e-05,
      "loss": 0.7273,
      "step": 550
    },
    {
      "epoch": 0.4352895452778857,
      "grad_norm": 4.514745713617805,
      "learning_rate": 5.427184466019418e-05,
      "loss": 0.7143,
      "step": 560
    },
    {
      "epoch": 0.4430625728721337,
      "grad_norm": 3.852074915498448,
      "learning_rate": 5.524271844660195e-05,
      "loss": 0.6954,
      "step": 570
    },
    {
      "epoch": 0.45083560046638166,
      "grad_norm": 4.549795432019635,
      "learning_rate": 5.6213592233009716e-05,
      "loss": 0.6695,
      "step": 580
    },
    {
      "epoch": 0.45860862806062963,
      "grad_norm": 3.920314510813715,
      "learning_rate": 5.718446601941748e-05,
      "loss": 0.6503,
      "step": 590
    },
    {
      "epoch": 0.46638165565487755,
      "grad_norm": 5.518859711149864,
      "learning_rate": 5.815533980582525e-05,
      "loss": 0.6249,
      "step": 600
    },
    {
      "epoch": 0.4741546832491255,
      "grad_norm": 4.876638883083892,
      "learning_rate": 5.9126213592233015e-05,
      "loss": 0.6064,
      "step": 610
    },
    {
      "epoch": 0.4819277108433735,
      "grad_norm": 5.2941697252063955,
      "learning_rate": 6.009708737864078e-05,
      "loss": 0.5911,
      "step": 620
    },
    {
      "epoch": 0.48970073843762146,
      "grad_norm": 4.245229081646686,
      "learning_rate": 6.106796116504854e-05,
      "loss": 0.5704,
      "step": 630
    },
    {
      "epoch": 0.49747376603186944,
      "grad_norm": 4.127680025681991,
      "learning_rate": 6.203883495145631e-05,
      "loss": 0.5589,
      "step": 640
    },
    {
      "epoch": 0.5052467936261174,
      "grad_norm": 3.5437231540377616,
      "learning_rate": 6.300970873786408e-05,
      "loss": 0.5378,
      "step": 650
    },
    {
      "epoch": 0.5130198212203654,
      "grad_norm": 4.180991087743002,
      "learning_rate": 6.398058252427184e-05,
      "loss": 0.5056,
      "step": 660
    },
    {
      "epoch": 0.5207928488146133,
      "grad_norm": 3.7905222943864034,
      "learning_rate": 6.495145631067961e-05,
      "loss": 0.4944,
      "step": 670
    },
    {
      "epoch": 0.5285658764088612,
      "grad_norm": 2.829726998778342,
      "learning_rate": 6.592233009708738e-05,
      "loss": 0.4706,
      "step": 680
    },
    {
      "epoch": 0.5363389040031092,
      "grad_norm": 2.9270824536877456,
      "learning_rate": 6.689320388349515e-05,
      "loss": 0.4539,
      "step": 690
    },
    {
      "epoch": 0.5441119315973572,
      "grad_norm": 2.742603316305724,
      "learning_rate": 6.786407766990291e-05,
      "loss": 0.4376,
      "step": 700
    },
    {
      "epoch": 0.5518849591916052,
      "grad_norm": 2.896538805797215,
      "learning_rate": 6.883495145631068e-05,
      "loss": 0.4113,
      "step": 710
    },
    {
      "epoch": 0.5596579867858531,
      "grad_norm": 3.4815145675652013,
      "learning_rate": 6.980582524271845e-05,
      "loss": 0.4005,
      "step": 720
    },
    {
      "epoch": 0.567431014380101,
      "grad_norm": 2.270621987351258,
      "learning_rate": 7.077669902912621e-05,
      "loss": 0.3886,
      "step": 730
    },
    {
      "epoch": 0.575204041974349,
      "grad_norm": 2.7187166457028344,
      "learning_rate": 7.174757281553398e-05,
      "loss": 0.3765,
      "step": 740
    },
    {
      "epoch": 0.582977069568597,
      "grad_norm": 3.184495607766821,
      "learning_rate": 7.271844660194175e-05,
      "loss": 0.3514,
      "step": 750
    },
    {
      "epoch": 0.5907500971628449,
      "grad_norm": 2.379059028542611,
      "learning_rate": 7.368932038834951e-05,
      "loss": 0.341,
      "step": 760
    },
    {
      "epoch": 0.5985231247570929,
      "grad_norm": 2.4310101684060705,
      "learning_rate": 7.466019417475728e-05,
      "loss": 0.3263,
      "step": 770
    },
    {
      "epoch": 0.6062961523513408,
      "grad_norm": 2.5710567898956005,
      "learning_rate": 7.563106796116505e-05,
      "loss": 0.3117,
      "step": 780
    },
    {
      "epoch": 0.6140691799455888,
      "grad_norm": 2.416082850923419,
      "learning_rate": 7.660194174757281e-05,
      "loss": 0.2944,
      "step": 790
    },
    {
      "epoch": 0.6218422075398368,
      "grad_norm": 2.109447352556029,
      "learning_rate": 7.757281553398059e-05,
      "loss": 0.2832,
      "step": 800
    },
    {
      "epoch": 0.6296152351340847,
      "grad_norm": 2.2994104123752988,
      "learning_rate": 7.854368932038836e-05,
      "loss": 0.2726,
      "step": 810
    },
    {
      "epoch": 0.6373882627283327,
      "grad_norm": 2.9267817958414524,
      "learning_rate": 7.951456310679612e-05,
      "loss": 0.2606,
      "step": 820
    },
    {
      "epoch": 0.6451612903225806,
      "grad_norm": 2.1699995560860197,
      "learning_rate": 8.048543689320389e-05,
      "loss": 0.2471,
      "step": 830
    },
    {
      "epoch": 0.6529343179168287,
      "grad_norm": 1.680681697729957,
      "learning_rate": 8.145631067961166e-05,
      "loss": 0.2354,
      "step": 840
    },
    {
      "epoch": 0.6607073455110766,
      "grad_norm": 1.88393966323043,
      "learning_rate": 8.242718446601942e-05,
      "loss": 0.2263,
      "step": 850
    },
    {
      "epoch": 0.6684803731053245,
      "grad_norm": 2.7385999995385957,
      "learning_rate": 8.339805825242719e-05,
      "loss": 0.2109,
      "step": 860
    },
    {
      "epoch": 0.6762534006995725,
      "grad_norm": 1.71125390353767,
      "learning_rate": 8.436893203883496e-05,
      "loss": 0.2034,
      "step": 870
    },
    {
      "epoch": 0.6840264282938204,
      "grad_norm": 1.8856456915119455,
      "learning_rate": 8.533980582524272e-05,
      "loss": 0.1928,
      "step": 880
    },
    {
      "epoch": 0.6917994558880683,
      "grad_norm": 1.54640226468132,
      "learning_rate": 8.631067961165049e-05,
      "loss": 0.1797,
      "step": 890
    },
    {
      "epoch": 0.6995724834823164,
      "grad_norm": 1.3309710354478357,
      "learning_rate": 8.728155339805826e-05,
      "loss": 0.1712,
      "step": 900
    },
    {
      "epoch": 0.7073455110765643,
      "grad_norm": 1.2897601826602498,
      "learning_rate": 8.825242718446602e-05,
      "loss": 0.1624,
      "step": 910
    },
    {
      "epoch": 0.7151185386708123,
      "grad_norm": 1.1022252566548159,
      "learning_rate": 8.922330097087379e-05,
      "loss": 0.1509,
      "step": 920
    },
    {
      "epoch": 0.7228915662650602,
      "grad_norm": 1.4797407171136665,
      "learning_rate": 9.019417475728156e-05,
      "loss": 0.1495,
      "step": 930
    },
    {
      "epoch": 0.7306645938593082,
      "grad_norm": 1.4708443604696282,
      "learning_rate": 9.116504854368933e-05,
      "loss": 0.1407,
      "step": 940
    },
    {
      "epoch": 0.7384376214535562,
      "grad_norm": 1.2062811965627402,
      "learning_rate": 9.213592233009709e-05,
      "loss": 0.1409,
      "step": 950
    },
    {
      "epoch": 0.7462106490478041,
      "grad_norm": 0.991340571431122,
      "learning_rate": 9.310679611650486e-05,
      "loss": 0.1267,
      "step": 960
    },
    {
      "epoch": 0.7539836766420521,
      "grad_norm": 1.0799976482859923,
      "learning_rate": 9.407766990291263e-05,
      "loss": 0.1294,
      "step": 970
    },
    {
      "epoch": 0.7617567042363,
      "grad_norm": 1.0055612260520772,
      "learning_rate": 9.504854368932039e-05,
      "loss": 0.1167,
      "step": 980
    },
    {
      "epoch": 0.769529731830548,
      "grad_norm": 0.99809772380525,
      "learning_rate": 9.601941747572816e-05,
      "loss": 0.1104,
      "step": 990
    },
    {
      "epoch": 0.777302759424796,
      "grad_norm": 0.8341966391057081,
      "learning_rate": 9.699029126213593e-05,
      "loss": 0.104,
      "step": 1000
    },
    {
      "epoch": 0.7850757870190439,
      "grad_norm": 0.7440675821604489,
      "learning_rate": 9.796116504854369e-05,
      "loss": 0.1,
      "step": 1010
    },
    {
      "epoch": 0.7928488146132919,
      "grad_norm": 1.0089217362175902,
      "learning_rate": 9.893203883495146e-05,
      "loss": 0.0926,
      "step": 1020
    },
    {
      "epoch": 0.8006218422075398,
      "grad_norm": 0.669475479205569,
      "learning_rate": 9.990291262135923e-05,
      "loss": 0.0855,
      "step": 1030
    },
    {
      "epoch": 0.8083948698017878,
      "grad_norm": 0.7508700947786369,
      "learning_rate": 9.999976722311928e-05,
      "loss": 0.0824,
      "step": 1040
    },
    {
      "epoch": 0.8161678973960358,
      "grad_norm": 0.5833904781662174,
      "learning_rate": 9.999896256507963e-05,
      "loss": 0.0794,
      "step": 1050
    },
    {
      "epoch": 0.8239409249902837,
      "grad_norm": 0.6285661168460489,
      "learning_rate": 9.99975831613399e-05,
      "loss": 0.0737,
      "step": 1060
    },
    {
      "epoch": 0.8317139525845316,
      "grad_norm": 0.574291945207083,
      "learning_rate": 9.99956290277565e-05,
      "loss": 0.0689,
      "step": 1070
    },
    {
      "epoch": 0.8394869801787797,
      "grad_norm": 0.4876511188851532,
      "learning_rate": 9.999310018679253e-05,
      "loss": 0.0659,
      "step": 1080
    },
    {
      "epoch": 0.8472600077730276,
      "grad_norm": 0.553792085705227,
      "learning_rate": 9.998999666751738e-05,
      "loss": 0.064,
      "step": 1090
    },
    {
      "epoch": 0.8550330353672756,
      "grad_norm": 0.4302541179456994,
      "learning_rate": 9.998631850560647e-05,
      "loss": 0.0612,
      "step": 1100
    },
    {
      "epoch": 0.8628060629615235,
      "grad_norm": 0.47710435973221127,
      "learning_rate": 9.998206574334089e-05,
      "loss": 0.0572,
      "step": 1110
    },
    {
      "epoch": 0.8705790905557714,
      "grad_norm": 0.6437234961571133,
      "learning_rate": 9.997723842960671e-05,
      "loss": 0.0556,
      "step": 1120
    },
    {
      "epoch": 0.8783521181500195,
      "grad_norm": 0.44297654610119586,
      "learning_rate": 9.997183661989468e-05,
      "loss": 0.0551,
      "step": 1130
    },
    {
      "epoch": 0.8861251457442674,
      "grad_norm": 0.40789308058272417,
      "learning_rate": 9.996586037629944e-05,
      "loss": 0.052,
      "step": 1140
    },
    {
      "epoch": 0.8938981733385154,
      "grad_norm": 0.412683747214122,
      "learning_rate": 9.99593097675188e-05,
      "loss": 0.0503,
      "step": 1150
    },
    {
      "epoch": 0.9016712009327633,
      "grad_norm": 0.4042307583726218,
      "learning_rate": 9.995218486885302e-05,
      "loss": 0.0491,
      "step": 1160
    },
    {
      "epoch": 0.9094442285270112,
      "grad_norm": 0.3917357425568075,
      "learning_rate": 9.994448576220388e-05,
      "loss": 0.0472,
      "step": 1170
    },
    {
      "epoch": 0.9172172561212593,
      "grad_norm": 0.42847580686284464,
      "learning_rate": 9.99362125360738e-05,
      "loss": 0.0461,
      "step": 1180
    },
    {
      "epoch": 0.9249902837155072,
      "grad_norm": 0.36275912435050334,
      "learning_rate": 9.992736528556474e-05,
      "loss": 0.045,
      "step": 1190
    },
    {
      "epoch": 0.9327633113097551,
      "grad_norm": 0.29922320469821373,
      "learning_rate": 9.991794411237725e-05,
      "loss": 0.0429,
      "step": 1200
    },
    {
      "epoch": 0.9405363389040031,
      "grad_norm": 0.3914552553052167,
      "learning_rate": 9.990794912480913e-05,
      "loss": 0.0426,
      "step": 1210
    },
    {
      "epoch": 0.948309366498251,
      "grad_norm": 0.38509264588619113,
      "learning_rate": 9.989738043775428e-05,
      "loss": 0.0423,
      "step": 1220
    },
    {
      "epoch": 0.9560823940924991,
      "grad_norm": 0.3259823714473278,
      "learning_rate": 9.988623817270137e-05,
      "loss": 0.0406,
      "step": 1230
    },
    {
      "epoch": 0.963855421686747,
      "grad_norm": 0.31452752346688945,
      "learning_rate": 9.987452245773243e-05,
      "loss": 0.0398,
      "step": 1240
    },
    {
      "epoch": 0.9716284492809949,
      "grad_norm": 0.338353347661472,
      "learning_rate": 9.986223342752141e-05,
      "loss": 0.0386,
      "step": 1250
    },
    {
      "epoch": 0.9794014768752429,
      "grad_norm": 0.3002478510776732,
      "learning_rate": 9.984937122333256e-05,
      "loss": 0.0382,
      "step": 1260
    },
    {
      "epoch": 0.9871745044694908,
      "grad_norm": 0.40670865355319846,
      "learning_rate": 9.983593599301891e-05,
      "loss": 0.037,
      "step": 1270
    },
    {
      "epoch": 0.9949475320637389,
      "grad_norm": 0.6429283417769656,
      "learning_rate": 9.982192789102047e-05,
      "loss": 0.0373,
      "step": 1280
    },
    {
      "epoch": 1.0023319082782745,
      "grad_norm": 0.2801176127856459,
      "learning_rate": 9.980734707836248e-05,
      "loss": 0.0354,
      "step": 1290
    },
    {
      "epoch": 1.0101049358725223,
      "grad_norm": 0.3724656663427553,
      "learning_rate": 9.979219372265365e-05,
      "loss": 0.0344,
      "step": 1300
    },
    {
      "epoch": 1.0178779634667703,
      "grad_norm": 0.2573842211275922,
      "learning_rate": 9.97764679980841e-05,
      "loss": 0.0333,
      "step": 1310
    },
    {
      "epoch": 1.0256509910610183,
      "grad_norm": 0.24713329711260812,
      "learning_rate": 9.97601700854234e-05,
      "loss": 0.0336,
      "step": 1320
    },
    {
      "epoch": 1.0334240186552661,
      "grad_norm": 0.23958650404178872,
      "learning_rate": 9.974330017201859e-05,
      "loss": 0.0328,
      "step": 1330
    },
    {
      "epoch": 1.0411970462495141,
      "grad_norm": 0.22856953483452255,
      "learning_rate": 9.972585845179188e-05,
      "loss": 0.0316,
      "step": 1340
    },
    {
      "epoch": 1.0489700738437622,
      "grad_norm": 0.25194563553211635,
      "learning_rate": 9.970784512523848e-05,
      "loss": 0.0311,
      "step": 1350
    },
    {
      "epoch": 1.0567431014380102,
      "grad_norm": 0.23842502807724422,
      "learning_rate": 9.968926039942435e-05,
      "loss": 0.0306,
      "step": 1360
    },
    {
      "epoch": 1.064516129032258,
      "grad_norm": 0.2701262980707877,
      "learning_rate": 9.967010448798375e-05,
      "loss": 0.0303,
      "step": 1370
    },
    {
      "epoch": 1.072289156626506,
      "grad_norm": 0.25970806872919033,
      "learning_rate": 9.965037761111681e-05,
      "loss": 0.0295,
      "step": 1380
    },
    {
      "epoch": 1.080062184220754,
      "grad_norm": 0.24331306995976767,
      "learning_rate": 9.963007999558696e-05,
      "loss": 0.0291,
      "step": 1390
    },
    {
      "epoch": 1.0878352118150019,
      "grad_norm": 0.24100978412084842,
      "learning_rate": 9.96092118747184e-05,
      "loss": 0.0287,
      "step": 1400
    },
    {
      "epoch": 1.09560823940925,
      "grad_norm": 0.2184207676934967,
      "learning_rate": 9.958777348839336e-05,
      "loss": 0.0277,
      "step": 1410
    },
    {
      "epoch": 1.103381267003498,
      "grad_norm": 0.21996045383271537,
      "learning_rate": 9.956576508304938e-05,
      "loss": 0.0272,
      "step": 1420
    },
    {
      "epoch": 1.1111542945977457,
      "grad_norm": 0.21628224837569832,
      "learning_rate": 9.954318691167641e-05,
      "loss": 0.0273,
      "step": 1430
    },
    {
      "epoch": 1.1189273221919938,
      "grad_norm": 0.22298883167746006,
      "learning_rate": 9.952003923381399e-05,
      "loss": 0.0264,
      "step": 1440
    },
    {
      "epoch": 1.1267003497862418,
      "grad_norm": 0.2246150118171153,
      "learning_rate": 9.949632231554819e-05,
      "loss": 0.0265,
      "step": 1450
    },
    {
      "epoch": 1.1344733773804898,
      "grad_norm": 0.23186180017701422,
      "learning_rate": 9.947203642950862e-05,
      "loss": 0.0262,
      "step": 1460
    },
    {
      "epoch": 1.1422464049747376,
      "grad_norm": 0.24648849817769855,
      "learning_rate": 9.944718185486523e-05,
      "loss": 0.0256,
      "step": 1470
    },
    {
      "epoch": 1.1500194325689856,
      "grad_norm": 0.30413187225809407,
      "learning_rate": 9.942175887732517e-05,
      "loss": 0.025,
      "step": 1480
    },
    {
      "epoch": 1.1577924601632337,
      "grad_norm": 0.23138441255979614,
      "learning_rate": 9.93957677891294e-05,
      "loss": 0.0251,
      "step": 1490
    },
    {
      "epoch": 1.1655654877574815,
      "grad_norm": 0.24336635292704567,
      "learning_rate": 9.93692088890495e-05,
      "loss": 0.0244,
      "step": 1500
    },
    {
      "epoch": 1.1733385153517295,
      "grad_norm": 0.21185931343408212,
      "learning_rate": 9.934208248238403e-05,
      "loss": 0.0243,
      "step": 1510
    },
    {
      "epoch": 1.1811115429459775,
      "grad_norm": 0.19638440276738353,
      "learning_rate": 9.93143888809552e-05,
      "loss": 0.0239,
      "step": 1520
    },
    {
      "epoch": 1.1888845705402253,
      "grad_norm": 0.21756730427462398,
      "learning_rate": 9.928612840310514e-05,
      "loss": 0.0232,
      "step": 1530
    },
    {
      "epoch": 1.1966575981344734,
      "grad_norm": 0.21305245786411112,
      "learning_rate": 9.925730137369242e-05,
      "loss": 0.023,
      "step": 1540
    },
    {
      "epoch": 1.2044306257287214,
      "grad_norm": 0.22828714542453502,
      "learning_rate": 9.922790812408808e-05,
      "loss": 0.0226,
      "step": 1550
    },
    {
      "epoch": 1.2122036533229692,
      "grad_norm": 0.19930203197693303,
      "learning_rate": 9.919794899217203e-05,
      "loss": 0.0224,
      "step": 1560
    },
    {
      "epoch": 1.2199766809172172,
      "grad_norm": 0.20433748074735467,
      "learning_rate": 9.916742432232909e-05,
      "loss": 0.0222,
      "step": 1570
    },
    {
      "epoch": 1.2277497085114653,
      "grad_norm": 0.20184151603700423,
      "learning_rate": 9.913633446544493e-05,
      "loss": 0.0215,
      "step": 1580
    },
    {
      "epoch": 1.2355227361057133,
      "grad_norm": 0.2081529716929725,
      "learning_rate": 9.910467977890225e-05,
      "loss": 0.0216,
      "step": 1590
    },
    {
      "epoch": 1.243295763699961,
      "grad_norm": 0.27882658958977674,
      "learning_rate": 9.907246062657648e-05,
      "loss": 0.0234,
      "step": 1600
    },
    {
      "epoch": 1.2510687912942091,
      "grad_norm": 0.30986570505669037,
      "learning_rate": 9.903967737883168e-05,
      "loss": 0.0248,
      "step": 1610
    },
    {
      "epoch": 1.258841818888457,
      "grad_norm": 0.2205524563203657,
      "learning_rate": 9.900633041251631e-05,
      "loss": 0.0222,
      "step": 1620
    },
    {
      "epoch": 1.266614846482705,
      "grad_norm": 0.19645809273796086,
      "learning_rate": 9.897242011095883e-05,
      "loss": 0.0208,
      "step": 1630
    },
    {
      "epoch": 1.274387874076953,
      "grad_norm": 0.20220041423279006,
      "learning_rate": 9.893794686396329e-05,
      "loss": 0.0203,
      "step": 1640
    },
    {
      "epoch": 1.282160901671201,
      "grad_norm": 0.19701852522226898,
      "learning_rate": 9.890291106780495e-05,
      "loss": 0.0199,
      "step": 1650
    },
    {
      "epoch": 1.289933929265449,
      "grad_norm": 0.19055264078653414,
      "learning_rate": 9.886731312522563e-05,
      "loss": 0.0196,
      "step": 1660
    },
    {
      "epoch": 1.2977069568596968,
      "grad_norm": 0.17498611562411595,
      "learning_rate": 9.883115344542911e-05,
      "loss": 0.0189,
      "step": 1670
    },
    {
      "epoch": 1.3054799844539449,
      "grad_norm": 0.18664040460319856,
      "learning_rate": 9.879443244407643e-05,
      "loss": 0.019,
      "step": 1680
    },
    {
      "epoch": 1.3132530120481927,
      "grad_norm": 0.186899833087349,
      "learning_rate": 9.875715054328108e-05,
      "loss": 0.0186,
      "step": 1690
    },
    {
      "epoch": 1.3210260396424407,
      "grad_norm": 0.18960588899249967,
      "learning_rate": 9.871930817160422e-05,
      "loss": 0.018,
      "step": 1700
    },
    {
      "epoch": 1.3287990672366887,
      "grad_norm": 0.2246998119833756,
      "learning_rate": 9.868090576404966e-05,
      "loss": 0.0185,
      "step": 1710
    },
    {
      "epoch": 1.3365720948309368,
      "grad_norm": 0.18612335330028418,
      "learning_rate": 9.864194376205895e-05,
      "loss": 0.0181,
      "step": 1720
    },
    {
      "epoch": 1.3443451224251846,
      "grad_norm": 0.19504144084562025,
      "learning_rate": 9.860242261350623e-05,
      "loss": 0.0174,
      "step": 1730
    },
    {
      "epoch": 1.3521181500194326,
      "grad_norm": 0.1940460381388096,
      "learning_rate": 9.856234277269313e-05,
      "loss": 0.0173,
      "step": 1740
    },
    {
      "epoch": 1.3598911776136804,
      "grad_norm": 0.1882305169943444,
      "learning_rate": 9.852170470034356e-05,
      "loss": 0.0173,
      "step": 1750
    },
    {
      "epoch": 1.3676642052079284,
      "grad_norm": 0.1844609631309131,
      "learning_rate": 9.848050886359835e-05,
      "loss": 0.0171,
      "step": 1760
    },
    {
      "epoch": 1.3754372328021764,
      "grad_norm": 0.16755276537647784,
      "learning_rate": 9.843875573600989e-05,
      "loss": 0.0169,
      "step": 1770
    },
    {
      "epoch": 1.3832102603964245,
      "grad_norm": 0.16696100318285242,
      "learning_rate": 9.839644579753677e-05,
      "loss": 0.0164,
      "step": 1780
    },
    {
      "epoch": 1.3909832879906725,
      "grad_norm": 0.1818918991872044,
      "learning_rate": 9.835357953453821e-05,
      "loss": 0.0163,
      "step": 1790
    },
    {
      "epoch": 1.3987563155849203,
      "grad_norm": 0.18963889299339048,
      "learning_rate": 9.831015743976843e-05,
      "loss": 0.0158,
      "step": 1800
    }
  ],
  "logging_steps": 10,
  "max_steps": 10296,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 1800,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 132633864437760.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}