{
  "best_metric": 0.6514522821576764,
  "best_model_checkpoint": "timesformer-base-finetuned-k400-finetuned-crema-d/checkpoint-2235",
  "epoch": 3.248991935483871,
  "eval_steps": 500,
  "global_step": 2976,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 8.966336250305176,
      "learning_rate": 1.6778523489932886e-06,
      "loss": 1.8029,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.845541954040527,
      "learning_rate": 3.3557046979865773e-06,
      "loss": 1.8707,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.349601745605469,
      "learning_rate": 5.033557046979865e-06,
      "loss": 1.8082,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.781116485595703,
      "learning_rate": 6.7114093959731546e-06,
      "loss": 1.8174,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.116007804870605,
      "learning_rate": 8.389261744966444e-06,
      "loss": 1.7626,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.84127426147461,
      "learning_rate": 1.006711409395973e-05,
      "loss": 1.7316,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.750720024108887,
      "learning_rate": 1.174496644295302e-05,
      "loss": 1.6928,
      "step": 70
    },
    {
      "epoch": 0.03,
      "grad_norm": 16.532550811767578,
      "learning_rate": 1.3422818791946309e-05,
      "loss": 1.6398,
      "step": 80
    },
    {
      "epoch": 0.03,
      "grad_norm": 14.644113540649414,
      "learning_rate": 1.51006711409396e-05,
      "loss": 1.6093,
      "step": 90
    },
    {
      "epoch": 0.03,
      "grad_norm": 13.7882661819458,
      "learning_rate": 1.6778523489932888e-05,
      "loss": 1.4837,
      "step": 100
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.735069274902344,
      "learning_rate": 1.8456375838926178e-05,
      "loss": 1.3557,
      "step": 110
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.975994110107422,
      "learning_rate": 2.013422818791946e-05,
      "loss": 1.4247,
      "step": 120
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.303879737854004,
      "learning_rate": 2.181208053691275e-05,
      "loss": 1.4661,
      "step": 130
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.194876670837402,
      "learning_rate": 2.348993288590604e-05,
      "loss": 1.3662,
      "step": 140
    },
    {
      "epoch": 0.05,
      "grad_norm": 15.312520980834961,
      "learning_rate": 2.516778523489933e-05,
      "loss": 1.3889,
      "step": 150
    },
    {
      "epoch": 0.05,
      "grad_norm": 12.085341453552246,
      "learning_rate": 2.6845637583892618e-05,
      "loss": 1.3331,
      "step": 160
    },
    {
      "epoch": 0.06,
      "grad_norm": 16.036479949951172,
      "learning_rate": 2.8523489932885905e-05,
      "loss": 1.3931,
      "step": 170
    },
    {
      "epoch": 0.06,
      "grad_norm": 13.334360122680664,
      "learning_rate": 3.02013422818792e-05,
      "loss": 1.5768,
      "step": 180
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.737225532531738,
      "learning_rate": 3.1879194630872485e-05,
      "loss": 1.6222,
      "step": 190
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.720209121704102,
      "learning_rate": 3.3557046979865775e-05,
      "loss": 1.4513,
      "step": 200
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.798286437988281,
      "learning_rate": 3.523489932885906e-05,
      "loss": 1.3589,
      "step": 210
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.686984062194824,
      "learning_rate": 3.6912751677852356e-05,
      "loss": 1.2702,
      "step": 220
    },
    {
      "epoch": 0.08,
      "grad_norm": 15.610692024230957,
      "learning_rate": 3.859060402684564e-05,
      "loss": 1.3324,
      "step": 230
    },
    {
      "epoch": 0.08,
      "grad_norm": 11.944442749023438,
      "learning_rate": 4.026845637583892e-05,
      "loss": 1.2537,
      "step": 240
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.310532569885254,
      "learning_rate": 4.194630872483222e-05,
      "loss": 1.1742,
      "step": 250
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.212057113647461,
      "learning_rate": 4.36241610738255e-05,
      "loss": 1.2357,
      "step": 260
    },
    {
      "epoch": 0.09,
      "grad_norm": 6.6078596115112305,
      "learning_rate": 4.530201342281879e-05,
      "loss": 1.3228,
      "step": 270
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.31572437286377,
      "learning_rate": 4.697986577181208e-05,
      "loss": 1.1611,
      "step": 280
    },
    {
      "epoch": 0.1,
      "grad_norm": 16.24120330810547,
      "learning_rate": 4.865771812080537e-05,
      "loss": 1.1575,
      "step": 290
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.7396087646484375,
      "learning_rate": 4.996265870052278e-05,
      "loss": 1.2793,
      "step": 300
    },
    {
      "epoch": 0.1,
      "grad_norm": 13.924488067626953,
      "learning_rate": 4.977595220313667e-05,
      "loss": 1.311,
      "step": 310
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.521751403808594,
      "learning_rate": 4.958924570575056e-05,
      "loss": 1.3315,
      "step": 320
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.977168083190918,
      "learning_rate": 4.9402539208364454e-05,
      "loss": 1.3456,
      "step": 330
    },
    {
      "epoch": 0.11,
      "grad_norm": 13.056800842285156,
      "learning_rate": 4.9215832710978346e-05,
      "loss": 1.1598,
      "step": 340
    },
    {
      "epoch": 0.12,
      "grad_norm": 15.543603897094727,
      "learning_rate": 4.902912621359224e-05,
      "loss": 1.3371,
      "step": 350
    },
    {
      "epoch": 0.12,
      "grad_norm": 14.090201377868652,
      "learning_rate": 4.884241971620613e-05,
      "loss": 1.7747,
      "step": 360
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.6073150634765625,
      "learning_rate": 4.8655713218820016e-05,
      "loss": 1.2209,
      "step": 370
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.058303833007812,
      "learning_rate": 4.846900672143391e-05,
      "loss": 1.1249,
      "step": 380
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.635152816772461,
      "learning_rate": 4.82823002240478e-05,
      "loss": 1.1293,
      "step": 390
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.893068313598633,
      "learning_rate": 4.809559372666169e-05,
      "loss": 1.2109,
      "step": 400
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.448178291320801,
      "learning_rate": 4.790888722927558e-05,
      "loss": 1.11,
      "step": 410
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.379781723022461,
      "learning_rate": 4.772218073188947e-05,
      "loss": 1.2579,
      "step": 420
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.443665504455566,
      "learning_rate": 4.753547423450336e-05,
      "loss": 1.2387,
      "step": 430
    },
    {
      "epoch": 0.15,
      "grad_norm": 12.107076644897461,
      "learning_rate": 4.7348767737117256e-05,
      "loss": 1.1839,
      "step": 440
    },
    {
      "epoch": 0.15,
      "grad_norm": 6.977869033813477,
      "learning_rate": 4.716206123973114e-05,
      "loss": 1.1067,
      "step": 450
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.275339603424072,
      "learning_rate": 4.697535474234503e-05,
      "loss": 1.137,
      "step": 460
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.69481086730957,
      "learning_rate": 4.6788648244958926e-05,
      "loss": 0.9908,
      "step": 470
    },
    {
      "epoch": 0.16,
      "grad_norm": 12.958833694458008,
      "learning_rate": 4.660194174757282e-05,
      "loss": 1.3072,
      "step": 480
    },
    {
      "epoch": 0.16,
      "grad_norm": 6.747406005859375,
      "learning_rate": 4.6415235250186703e-05,
      "loss": 1.1151,
      "step": 490
    },
    {
      "epoch": 0.17,
      "grad_norm": 6.857999324798584,
      "learning_rate": 4.6228528752800596e-05,
      "loss": 1.2205,
      "step": 500
    },
    {
      "epoch": 0.17,
      "grad_norm": 6.939675807952881,
      "learning_rate": 4.604182225541449e-05,
      "loss": 1.1204,
      "step": 510
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.679428100585938,
      "learning_rate": 4.585511575802838e-05,
      "loss": 1.4821,
      "step": 520
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.929995059967041,
      "learning_rate": 4.566840926064227e-05,
      "loss": 1.0272,
      "step": 530
    },
    {
      "epoch": 0.18,
      "grad_norm": 11.928690910339355,
      "learning_rate": 4.5481702763256165e-05,
      "loss": 1.0662,
      "step": 540
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.507076263427734,
      "learning_rate": 4.529499626587006e-05,
      "loss": 1.3273,
      "step": 550
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.756071090698242,
      "learning_rate": 4.510828976848395e-05,
      "loss": 1.1578,
      "step": 560
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.586543083190918,
      "learning_rate": 4.492158327109784e-05,
      "loss": 1.2698,
      "step": 570
    },
    {
      "epoch": 0.19,
      "grad_norm": 20.6762752532959,
      "learning_rate": 4.473487677371173e-05,
      "loss": 1.1172,
      "step": 580
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.282225608825684,
      "learning_rate": 4.454817027632562e-05,
      "loss": 1.2371,
      "step": 590
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.225829124450684,
      "learning_rate": 4.436146377893951e-05,
      "loss": 1.0255,
      "step": 600
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.627819061279297,
      "learning_rate": 4.4174757281553404e-05,
      "loss": 1.1431,
      "step": 610
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.365921974182129,
      "learning_rate": 4.398805078416729e-05,
      "loss": 1.2734,
      "step": 620
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.910810470581055,
      "learning_rate": 4.380134428678118e-05,
      "loss": 1.1357,
      "step": 630
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.954166412353516,
      "learning_rate": 4.3614637789395075e-05,
      "loss": 0.9257,
      "step": 640
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.132972717285156,
      "learning_rate": 4.342793129200897e-05,
      "loss": 0.8341,
      "step": 650
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.874831676483154,
      "learning_rate": 4.324122479462285e-05,
      "loss": 0.8737,
      "step": 660
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.769972801208496,
      "learning_rate": 4.3054518297236745e-05,
      "loss": 1.3776,
      "step": 670
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.323474884033203,
      "learning_rate": 4.286781179985064e-05,
      "loss": 1.0777,
      "step": 680
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.666136741638184,
      "learning_rate": 4.268110530246453e-05,
      "loss": 1.0246,
      "step": 690
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.188297271728516,
      "learning_rate": 4.2494398805078415e-05,
      "loss": 1.097,
      "step": 700
    },
    {
      "epoch": 0.24,
      "grad_norm": 6.143352031707764,
      "learning_rate": 4.230769230769231e-05,
      "loss": 0.9966,
      "step": 710
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.388320922851562,
      "learning_rate": 4.21209858103062e-05,
      "loss": 1.2716,
      "step": 720
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.024212837219238,
      "learning_rate": 4.193427931292009e-05,
      "loss": 1.0382,
      "step": 730
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.683081150054932,
      "learning_rate": 4.1747572815533984e-05,
      "loss": 1.0116,
      "step": 740
    },
    {
      "epoch": 0.25,
      "eval_accuracy": 0.5131396957123098,
      "eval_loss": 1.2478636503219604,
      "eval_runtime": 938.86,
      "eval_samples_per_second": 1.54,
      "eval_steps_per_second": 0.193,
      "step": 745
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.771040439605713,
      "learning_rate": 4.1560866318147876e-05,
      "loss": 0.7915,
      "step": 750
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.2587785720825195,
      "learning_rate": 4.137415982076177e-05,
      "loss": 0.7035,
      "step": 760
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.185038566589355,
      "learning_rate": 4.118745332337566e-05,
      "loss": 0.9042,
      "step": 770
    },
    {
      "epoch": 1.01,
      "grad_norm": 18.68658447265625,
      "learning_rate": 4.1000746825989546e-05,
      "loss": 1.1033,
      "step": 780
    },
    {
      "epoch": 1.02,
      "grad_norm": 15.911717414855957,
      "learning_rate": 4.081404032860344e-05,
      "loss": 1.0388,
      "step": 790
    },
    {
      "epoch": 1.02,
      "grad_norm": 5.451622486114502,
      "learning_rate": 4.062733383121733e-05,
      "loss": 1.228,
      "step": 800
    },
    {
      "epoch": 1.02,
      "grad_norm": 6.26366662979126,
      "learning_rate": 4.0440627333831223e-05,
      "loss": 1.0408,
      "step": 810
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.665594100952148,
      "learning_rate": 4.025392083644511e-05,
      "loss": 0.9401,
      "step": 820
    },
    {
      "epoch": 1.03,
      "grad_norm": 5.607839107513428,
      "learning_rate": 4.0067214339059e-05,
      "loss": 0.8985,
      "step": 830
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.730631828308105,
      "learning_rate": 3.9880507841672894e-05,
      "loss": 0.9892,
      "step": 840
    },
    {
      "epoch": 1.04,
      "grad_norm": 7.875046730041504,
      "learning_rate": 3.9693801344286786e-05,
      "loss": 1.0674,
      "step": 850
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.48693561553955,
      "learning_rate": 3.950709484690067e-05,
      "loss": 1.0219,
      "step": 860
    },
    {
      "epoch": 1.04,
      "grad_norm": 7.560312747955322,
      "learning_rate": 3.9320388349514564e-05,
      "loss": 0.9777,
      "step": 870
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.460749626159668,
      "learning_rate": 3.9133681852128456e-05,
      "loss": 1.0962,
      "step": 880
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.8882598876953125,
      "learning_rate": 3.894697535474235e-05,
      "loss": 0.9755,
      "step": 890
    },
    {
      "epoch": 1.05,
      "grad_norm": 6.44077205657959,
      "learning_rate": 3.8760268857356234e-05,
      "loss": 1.0329,
      "step": 900
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.286242485046387,
      "learning_rate": 3.8573562359970126e-05,
      "loss": 1.012,
      "step": 910
    },
    {
      "epoch": 1.06,
      "grad_norm": 5.543011665344238,
      "learning_rate": 3.838685586258402e-05,
      "loss": 1.1146,
      "step": 920
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.843368530273438,
      "learning_rate": 3.820014936519791e-05,
      "loss": 0.9118,
      "step": 930
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.51323413848877,
      "learning_rate": 3.8013442867811796e-05,
      "loss": 0.8912,
      "step": 940
    },
    {
      "epoch": 1.07,
      "grad_norm": 13.254180908203125,
      "learning_rate": 3.782673637042569e-05,
      "loss": 1.165,
      "step": 950
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.183913230895996,
      "learning_rate": 3.764002987303958e-05,
      "loss": 0.9994,
      "step": 960
    },
    {
      "epoch": 1.08,
      "grad_norm": 12.324352264404297,
      "learning_rate": 3.745332337565347e-05,
      "loss": 0.9169,
      "step": 970
    },
    {
      "epoch": 1.08,
      "grad_norm": 5.288768768310547,
      "learning_rate": 3.7266616878267365e-05,
      "loss": 0.9028,
      "step": 980
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.353190422058105,
      "learning_rate": 3.707991038088126e-05,
      "loss": 0.9624,
      "step": 990
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.271513938903809,
      "learning_rate": 3.689320388349515e-05,
      "loss": 1.1194,
      "step": 1000
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.480072975158691,
      "learning_rate": 3.670649738610904e-05,
      "loss": 0.9825,
      "step": 1010
    },
    {
      "epoch": 1.09,
      "grad_norm": 6.073505401611328,
      "learning_rate": 3.651979088872293e-05,
      "loss": 1.0177,
      "step": 1020
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.153465270996094,
      "learning_rate": 3.633308439133682e-05,
      "loss": 0.9163,
      "step": 1030
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.76282787322998,
      "learning_rate": 3.614637789395071e-05,
      "loss": 0.8354,
      "step": 1040
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.979357719421387,
      "learning_rate": 3.5959671396564605e-05,
      "loss": 0.8525,
      "step": 1050
    },
    {
      "epoch": 1.11,
      "grad_norm": 3.172257900238037,
      "learning_rate": 3.577296489917849e-05,
      "loss": 0.7977,
      "step": 1060
    },
    {
      "epoch": 1.11,
      "grad_norm": 7.318731307983398,
      "learning_rate": 3.558625840179238e-05,
      "loss": 1.1231,
      "step": 1070
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.534405708312988,
      "learning_rate": 3.5399551904406275e-05,
      "loss": 0.8137,
      "step": 1080
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.472566604614258,
      "learning_rate": 3.521284540702017e-05,
      "loss": 0.8573,
      "step": 1090
    },
    {
      "epoch": 1.12,
      "grad_norm": 7.498010635375977,
      "learning_rate": 3.502613890963405e-05,
      "loss": 0.9724,
      "step": 1100
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.730813980102539,
      "learning_rate": 3.4839432412247945e-05,
      "loss": 0.8822,
      "step": 1110
    },
    {
      "epoch": 1.13,
      "grad_norm": 5.970120906829834,
      "learning_rate": 3.465272591486184e-05,
      "loss": 0.9178,
      "step": 1120
    },
    {
      "epoch": 1.13,
      "grad_norm": 7.76125431060791,
      "learning_rate": 3.446601941747573e-05,
      "loss": 0.7399,
      "step": 1130
    },
    {
      "epoch": 1.13,
      "grad_norm": 13.393411636352539,
      "learning_rate": 3.427931292008962e-05,
      "loss": 0.9639,
      "step": 1140
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.74706745147705,
      "learning_rate": 3.409260642270351e-05,
      "loss": 1.112,
      "step": 1150
    },
    {
      "epoch": 1.14,
      "grad_norm": 7.2259602546691895,
      "learning_rate": 3.39058999253174e-05,
      "loss": 0.9841,
      "step": 1160
    },
    {
      "epoch": 1.14,
      "grad_norm": 5.164011001586914,
      "learning_rate": 3.371919342793129e-05,
      "loss": 0.8474,
      "step": 1170
    },
    {
      "epoch": 1.15,
      "grad_norm": 6.0777907371521,
      "learning_rate": 3.3532486930545184e-05,
      "loss": 1.005,
      "step": 1180
    },
    {
      "epoch": 1.15,
      "grad_norm": 6.80031156539917,
      "learning_rate": 3.334578043315908e-05,
      "loss": 0.8482,
      "step": 1190
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.064947128295898,
      "learning_rate": 3.315907393577297e-05,
      "loss": 0.7886,
      "step": 1200
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.459393501281738,
      "learning_rate": 3.297236743838686e-05,
      "loss": 0.7457,
      "step": 1210
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.584134101867676,
      "learning_rate": 3.2785660941000754e-05,
      "loss": 0.7781,
      "step": 1220
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.700101852416992,
      "learning_rate": 3.259895444361464e-05,
      "loss": 0.7307,
      "step": 1230
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.712447166442871,
      "learning_rate": 3.241224794622853e-05,
      "loss": 0.9998,
      "step": 1240
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.078315734863281,
      "learning_rate": 3.2225541448842424e-05,
      "loss": 0.9463,
      "step": 1250
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.657953262329102,
      "learning_rate": 3.2038834951456316e-05,
      "loss": 0.9286,
      "step": 1260
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.428277969360352,
      "learning_rate": 3.18521284540702e-05,
      "loss": 0.7539,
      "step": 1270
    },
    {
      "epoch": 1.18,
      "grad_norm": 4.560651779174805,
      "learning_rate": 3.1665421956684094e-05,
      "loss": 0.8404,
      "step": 1280
    },
    {
      "epoch": 1.18,
      "grad_norm": 5.551332473754883,
      "learning_rate": 3.1478715459297986e-05,
      "loss": 0.7146,
      "step": 1290
    },
    {
      "epoch": 1.19,
      "grad_norm": 8.496753692626953,
      "learning_rate": 3.129200896191188e-05,
      "loss": 0.8073,
      "step": 1300
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.748584747314453,
      "learning_rate": 3.1105302464525764e-05,
      "loss": 0.7927,
      "step": 1310
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.8315553665161133,
      "learning_rate": 3.0918595967139656e-05,
      "loss": 0.7411,
      "step": 1320
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.38379192352295,
      "learning_rate": 3.073188946975355e-05,
      "loss": 0.7508,
      "step": 1330
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.1679105758667,
      "learning_rate": 3.054518297236744e-05,
      "loss": 0.858,
      "step": 1340
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.699267864227295,
      "learning_rate": 3.035847647498133e-05,
      "loss": 0.9164,
      "step": 1350
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.339028358459473,
      "learning_rate": 3.0171769977595222e-05,
      "loss": 0.7847,
      "step": 1360
    },
    {
      "epoch": 1.21,
      "grad_norm": 5.491379261016846,
      "learning_rate": 2.9985063480209115e-05,
      "loss": 0.7382,
      "step": 1370
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.201659202575684,
      "learning_rate": 2.9798356982823007e-05,
      "loss": 0.7906,
      "step": 1380
    },
    {
      "epoch": 1.22,
      "grad_norm": 7.949913024902344,
      "learning_rate": 2.9611650485436892e-05,
      "loss": 0.6964,
      "step": 1390
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.574385643005371,
      "learning_rate": 2.9424943988050785e-05,
      "loss": 0.7372,
      "step": 1400
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.402859687805176,
      "learning_rate": 2.9238237490664677e-05,
      "loss": 0.8888,
      "step": 1410
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.103381156921387,
      "learning_rate": 2.905153099327857e-05,
      "loss": 0.8579,
      "step": 1420
    },
    {
      "epoch": 1.23,
      "grad_norm": 6.39555025100708,
      "learning_rate": 2.8864824495892455e-05,
      "loss": 0.8913,
      "step": 1430
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.563343048095703,
      "learning_rate": 2.8678117998506347e-05,
      "loss": 0.7085,
      "step": 1440
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.191370010375977,
      "learning_rate": 2.849141150112024e-05,
      "loss": 1.004,
      "step": 1450
    },
    {
      "epoch": 1.24,
      "grad_norm": 15.756904602050781,
      "learning_rate": 2.8304705003734132e-05,
      "loss": 0.9399,
      "step": 1460
    },
    {
      "epoch": 1.24,
      "grad_norm": 7.184589385986328,
      "learning_rate": 2.811799850634802e-05,
      "loss": 0.7033,
      "step": 1470
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.35378360748291,
      "learning_rate": 2.7931292008961913e-05,
      "loss": 0.7114,
      "step": 1480
    },
    {
      "epoch": 1.25,
      "grad_norm": 36.100059509277344,
      "learning_rate": 2.7744585511575805e-05,
      "loss": 0.9527,
      "step": 1490
    },
    {
      "epoch": 1.25,
      "eval_accuracy": 0.549792531120332,
      "eval_loss": 1.2327691316604614,
      "eval_runtime": 977.734,
      "eval_samples_per_second": 1.479,
      "eval_steps_per_second": 0.185,
      "step": 1490
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.35026741027832,
      "learning_rate": 2.7557879014189698e-05,
      "loss": 0.809,
      "step": 1500
    },
    {
      "epoch": 2.01,
      "grad_norm": 7.489639759063721,
      "learning_rate": 2.7371172516803583e-05,
      "loss": 0.7671,
      "step": 1510
    },
    {
      "epoch": 2.01,
      "grad_norm": 14.04826545715332,
      "learning_rate": 2.7184466019417475e-05,
      "loss": 0.6254,
      "step": 1520
    },
    {
      "epoch": 2.01,
      "grad_norm": 9.614288330078125,
      "learning_rate": 2.6997759522031368e-05,
      "loss": 0.724,
      "step": 1530
    },
    {
      "epoch": 2.02,
      "grad_norm": 10.15787410736084,
      "learning_rate": 2.681105302464526e-05,
      "loss": 0.6269,
      "step": 1540
    },
    {
      "epoch": 2.02,
      "grad_norm": 13.945058822631836,
      "learning_rate": 2.662434652725915e-05,
      "loss": 0.7043,
      "step": 1550
    },
    {
      "epoch": 2.02,
      "grad_norm": 8.97387981414795,
      "learning_rate": 2.643764002987304e-05,
      "loss": 1.0266,
      "step": 1560
    },
    {
      "epoch": 2.03,
      "grad_norm": 13.708014488220215,
      "learning_rate": 2.6250933532486934e-05,
      "loss": 0.7728,
      "step": 1570
    },
    {
      "epoch": 2.03,
      "grad_norm": 9.578474998474121,
      "learning_rate": 2.6064227035100826e-05,
      "loss": 0.8449,
      "step": 1580
    },
    {
      "epoch": 2.03,
      "grad_norm": 9.634424209594727,
      "learning_rate": 2.587752053771471e-05,
      "loss": 0.8558,
      "step": 1590
    },
    {
      "epoch": 2.04,
      "grad_norm": 12.515636444091797,
      "learning_rate": 2.5690814040328604e-05,
      "loss": 0.7223,
      "step": 1600
    },
    {
      "epoch": 2.04,
      "grad_norm": 10.763627052307129,
      "learning_rate": 2.5504107542942496e-05,
      "loss": 0.718,
      "step": 1610
    },
    {
      "epoch": 2.04,
      "grad_norm": 10.04530143737793,
      "learning_rate": 2.531740104555639e-05,
      "loss": 0.8171,
      "step": 1620
    },
    {
      "epoch": 2.05,
      "grad_norm": 6.750433921813965,
      "learning_rate": 2.5130694548170274e-05,
      "loss": 0.7821,
      "step": 1630
    },
    {
      "epoch": 2.05,
      "grad_norm": 8.174914360046387,
      "learning_rate": 2.4943988050784166e-05,
      "loss": 0.7136,
      "step": 1640
    },
    {
      "epoch": 2.05,
      "grad_norm": 4.98445463180542,
      "learning_rate": 2.475728155339806e-05,
      "loss": 0.6469,
      "step": 1650
    },
    {
      "epoch": 2.06,
      "grad_norm": 4.804347991943359,
      "learning_rate": 2.4570575056011947e-05,
      "loss": 0.5416,
      "step": 1660
    },
    {
      "epoch": 2.06,
      "grad_norm": 14.377196311950684,
      "learning_rate": 2.4383868558625843e-05,
      "loss": 0.7328,
      "step": 1670
    },
    {
      "epoch": 2.06,
      "grad_norm": 6.387362957000732,
      "learning_rate": 2.4197162061239732e-05,
      "loss": 0.7271,
      "step": 1680
    },
    {
      "epoch": 2.07,
      "grad_norm": 7.125580310821533,
      "learning_rate": 2.4010455563853624e-05,
      "loss": 0.8158,
      "step": 1690
    },
    {
      "epoch": 2.07,
      "grad_norm": 12.323750495910645,
      "learning_rate": 2.3823749066467517e-05,
      "loss": 0.8921,
      "step": 1700
    },
    {
      "epoch": 2.07,
      "grad_norm": 11.977945327758789,
      "learning_rate": 2.3637042569081406e-05,
      "loss": 0.6668,
      "step": 1710
    },
    {
      "epoch": 2.08,
      "grad_norm": 8.070152282714844,
      "learning_rate": 2.3450336071695298e-05,
      "loss": 0.7329,
      "step": 1720
    },
    {
      "epoch": 2.08,
      "grad_norm": 6.724440097808838,
      "learning_rate": 2.3263629574309187e-05,
      "loss": 0.7063,
      "step": 1730
    },
    {
      "epoch": 2.08,
      "grad_norm": 8.863656997680664,
      "learning_rate": 2.307692307692308e-05,
      "loss": 0.7954,
      "step": 1740
    },
    {
      "epoch": 2.09,
      "grad_norm": 9.189391136169434,
      "learning_rate": 2.2890216579536968e-05,
      "loss": 0.6772,
      "step": 1750
    },
    {
      "epoch": 2.09,
      "grad_norm": 4.208937168121338,
      "learning_rate": 2.270351008215086e-05,
      "loss": 0.6816,
      "step": 1760
    },
    {
      "epoch": 2.09,
      "grad_norm": 12.258294105529785,
      "learning_rate": 2.251680358476475e-05,
      "loss": 0.6676,
      "step": 1770
    },
    {
      "epoch": 2.1,
      "grad_norm": 8.332365036010742,
      "learning_rate": 2.233009708737864e-05,
      "loss": 0.7876,
      "step": 1780
    },
    {
      "epoch": 2.1,
      "grad_norm": 5.771688461303711,
      "learning_rate": 2.2143390589992534e-05,
      "loss": 0.9574,
      "step": 1790
    },
    {
      "epoch": 2.1,
      "grad_norm": 6.446303367614746,
      "learning_rate": 2.1956684092606426e-05,
      "loss": 0.7144,
      "step": 1800
    },
    {
      "epoch": 2.11,
      "grad_norm": 9.503754615783691,
      "learning_rate": 2.1769977595220315e-05,
      "loss": 0.6068,
      "step": 1810
    },
    {
      "epoch": 2.11,
      "grad_norm": 14.693286895751953,
      "learning_rate": 2.1583271097834207e-05,
      "loss": 0.6563,
      "step": 1820
    },
    {
      "epoch": 2.11,
      "grad_norm": 10.687201499938965,
      "learning_rate": 2.1396564600448096e-05,
      "loss": 0.7925,
      "step": 1830
    },
    {
      "epoch": 2.12,
      "grad_norm": 6.536046504974365,
      "learning_rate": 2.120985810306199e-05,
      "loss": 0.8295,
      "step": 1840
    },
    {
      "epoch": 2.12,
      "grad_norm": 7.004622459411621,
      "learning_rate": 2.1023151605675877e-05,
      "loss": 0.7893,
      "step": 1850
    },
    {
      "epoch": 2.12,
      "grad_norm": 9.96247673034668,
      "learning_rate": 2.083644510828977e-05,
      "loss": 0.8404,
      "step": 1860
    },
    {
      "epoch": 2.13,
      "grad_norm": 9.531902313232422,
      "learning_rate": 2.064973861090366e-05,
      "loss": 0.6834,
      "step": 1870
    },
    {
      "epoch": 2.13,
      "grad_norm": 11.442157745361328,
      "learning_rate": 2.046303211351755e-05,
      "loss": 0.7355,
      "step": 1880
    },
    {
      "epoch": 2.13,
      "grad_norm": 10.561468124389648,
      "learning_rate": 2.0276325616131443e-05,
      "loss": 0.7059,
      "step": 1890
    },
    {
      "epoch": 2.14,
      "grad_norm": 13.556699752807617,
      "learning_rate": 2.0089619118745336e-05,
      "loss": 0.8166,
      "step": 1900
    },
    {
      "epoch": 2.14,
      "grad_norm": 10.263440132141113,
      "learning_rate": 1.9902912621359225e-05,
      "loss": 0.976,
      "step": 1910
    },
    {
      "epoch": 2.14,
      "grad_norm": 11.936306953430176,
      "learning_rate": 1.9716206123973117e-05,
      "loss": 0.7265,
      "step": 1920
    },
    {
      "epoch": 2.15,
      "grad_norm": 12.076094627380371,
      "learning_rate": 1.9529499626587006e-05,
      "loss": 0.6839,
      "step": 1930
    },
    {
      "epoch": 2.15,
      "grad_norm": 8.33880615234375,
      "learning_rate": 1.9342793129200898e-05,
      "loss": 0.7764,
      "step": 1940
    },
    {
      "epoch": 2.15,
      "grad_norm": 8.386175155639648,
      "learning_rate": 1.9156086631814787e-05,
      "loss": 0.8813,
      "step": 1950
    },
    {
      "epoch": 2.16,
      "grad_norm": 4.594736576080322,
      "learning_rate": 1.896938013442868e-05,
      "loss": 0.6866,
      "step": 1960
    },
    {
      "epoch": 2.16,
      "grad_norm": 15.1980619430542,
      "learning_rate": 1.8782673637042568e-05,
      "loss": 0.6357,
      "step": 1970
    },
    {
      "epoch": 2.16,
      "grad_norm": 3.2236995697021484,
      "learning_rate": 1.859596713965646e-05,
      "loss": 0.4694,
      "step": 1980
    },
    {
      "epoch": 2.17,
      "grad_norm": 7.324807167053223,
      "learning_rate": 1.8409260642270353e-05,
      "loss": 0.8348,
      "step": 1990
    },
    {
      "epoch": 2.17,
      "grad_norm": 8.34453010559082,
      "learning_rate": 1.8222554144884245e-05,
      "loss": 0.6198,
      "step": 2000
    },
    {
      "epoch": 2.17,
      "grad_norm": 6.823315620422363,
      "learning_rate": 1.8035847647498134e-05,
      "loss": 0.5476,
      "step": 2010
    },
    {
      "epoch": 2.18,
      "grad_norm": 6.084024906158447,
      "learning_rate": 1.7849141150112026e-05,
      "loss": 0.8824,
      "step": 2020
    },
    {
      "epoch": 2.18,
      "grad_norm": 8.106437683105469,
      "learning_rate": 1.7662434652725915e-05,
      "loss": 0.6111,
      "step": 2030
    },
    {
      "epoch": 2.18,
      "grad_norm": 7.96425724029541,
      "learning_rate": 1.7475728155339808e-05,
      "loss": 0.6311,
      "step": 2040
    },
    {
      "epoch": 2.19,
      "grad_norm": 8.175187110900879,
      "learning_rate": 1.7289021657953697e-05,
      "loss": 0.5308,
      "step": 2050
    },
    {
      "epoch": 2.19,
      "grad_norm": 6.085997581481934,
      "learning_rate": 1.710231516056759e-05,
      "loss": 0.5263,
      "step": 2060
    },
    {
      "epoch": 2.19,
      "grad_norm": 7.611990928649902,
      "learning_rate": 1.6915608663181478e-05,
      "loss": 0.7262,
      "step": 2070
    },
    {
      "epoch": 2.2,
      "grad_norm": 7.725258827209473,
      "learning_rate": 1.672890216579537e-05,
      "loss": 0.7899,
      "step": 2080
    },
    {
      "epoch": 2.2,
      "grad_norm": 6.377608299255371,
      "learning_rate": 1.654219566840926e-05,
      "loss": 0.6879,
      "step": 2090
    },
    {
      "epoch": 2.2,
      "grad_norm": 12.69424057006836,
      "learning_rate": 1.635548917102315e-05,
      "loss": 0.8104,
      "step": 2100
    },
    {
      "epoch": 2.21,
      "grad_norm": 2.8327956199645996,
      "learning_rate": 1.6168782673637044e-05,
      "loss": 0.5905,
      "step": 2110
    },
    {
      "epoch": 2.21,
      "grad_norm": 8.960968017578125,
      "learning_rate": 1.5982076176250936e-05,
      "loss": 0.5654,
      "step": 2120
    },
    {
      "epoch": 2.22,
      "grad_norm": 12.242358207702637,
      "learning_rate": 1.5795369678864825e-05,
      "loss": 0.6423,
      "step": 2130
    },
    {
      "epoch": 2.22,
      "grad_norm": 5.338672161102295,
      "learning_rate": 1.5608663181478717e-05,
      "loss": 0.4424,
      "step": 2140
    },
    {
      "epoch": 2.22,
      "grad_norm": 11.80203628540039,
      "learning_rate": 1.5421956684092606e-05,
      "loss": 0.7185,
      "step": 2150
    },
    {
      "epoch": 2.23,
      "grad_norm": 10.600192070007324,
      "learning_rate": 1.5235250186706498e-05,
      "loss": 0.7381,
      "step": 2160
    },
    {
      "epoch": 2.23,
      "grad_norm": 11.627554893493652,
      "learning_rate": 1.5048543689320387e-05,
      "loss": 0.4774,
      "step": 2170
    },
    {
      "epoch": 2.23,
      "grad_norm": 5.562707424163818,
      "learning_rate": 1.486183719193428e-05,
      "loss": 0.7739,
      "step": 2180
    },
    {
      "epoch": 2.24,
      "grad_norm": 12.66571044921875,
      "learning_rate": 1.467513069454817e-05,
      "loss": 0.702,
      "step": 2190
    },
    {
      "epoch": 2.24,
      "grad_norm": 8.979747772216797,
      "learning_rate": 1.4488424197162062e-05,
      "loss": 0.7674,
      "step": 2200
    },
    {
      "epoch": 2.24,
      "grad_norm": 8.47741413116455,
      "learning_rate": 1.4301717699775951e-05,
      "loss": 0.7117,
      "step": 2210
    },
    {
      "epoch": 2.25,
      "grad_norm": 12.884946823120117,
      "learning_rate": 1.4115011202389844e-05,
      "loss": 0.5667,
      "step": 2220
    },
    {
      "epoch": 2.25,
      "grad_norm": 6.181754112243652,
      "learning_rate": 1.3928304705003734e-05,
      "loss": 0.5409,
      "step": 2230
    },
    {
      "epoch": 2.25,
      "eval_accuracy": 0.6514522821576764,
      "eval_loss": 0.9363940358161926,
      "eval_runtime": 1062.414,
      "eval_samples_per_second": 1.361,
      "eval_steps_per_second": 0.17,
      "step": 2235
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.820518493652344,
      "learning_rate": 1.3741598207617627e-05,
      "loss": 0.6216,
      "step": 2240
    },
    {
      "epoch": 3.01,
      "grad_norm": 11.719588279724121,
      "learning_rate": 1.3554891710231516e-05,
      "loss": 0.7012,
      "step": 2250
    },
    {
      "epoch": 3.01,
      "grad_norm": 8.87963581085205,
      "learning_rate": 1.3368185212845408e-05,
      "loss": 0.4752,
      "step": 2260
    },
    {
      "epoch": 3.01,
      "grad_norm": 7.130947113037109,
      "learning_rate": 1.3181478715459297e-05,
      "loss": 0.5228,
      "step": 2270
    },
    {
      "epoch": 3.02,
      "grad_norm": 5.969545841217041,
      "learning_rate": 1.2994772218073189e-05,
      "loss": 0.5243,
      "step": 2280
    },
    {
      "epoch": 3.02,
      "grad_norm": 9.089691162109375,
      "learning_rate": 1.2808065720687081e-05,
      "loss": 0.5782,
      "step": 2290
    },
    {
      "epoch": 3.02,
      "grad_norm": 22.645851135253906,
      "learning_rate": 1.2621359223300972e-05,
      "loss": 0.5316,
      "step": 2300
    },
    {
      "epoch": 3.03,
      "grad_norm": 13.614704132080078,
      "learning_rate": 1.2434652725914863e-05,
      "loss": 0.6667,
      "step": 2310
    },
    {
      "epoch": 3.03,
      "grad_norm": 4.408351421356201,
      "learning_rate": 1.2247946228528753e-05,
      "loss": 0.5386,
      "step": 2320
    },
    {
      "epoch": 3.03,
      "grad_norm": 11.060683250427246,
      "learning_rate": 1.2061239731142644e-05,
      "loss": 0.6878,
      "step": 2330
    },
    {
      "epoch": 3.04,
      "grad_norm": 7.305804252624512,
      "learning_rate": 1.1874533233756534e-05,
      "loss": 0.653,
      "step": 2340
    },
    {
      "epoch": 3.04,
      "grad_norm": 12.793059349060059,
      "learning_rate": 1.1687826736370427e-05,
      "loss": 0.5231,
      "step": 2350
    },
    {
      "epoch": 3.04,
      "grad_norm": 10.4277982711792,
      "learning_rate": 1.1501120238984317e-05,
      "loss": 0.6953,
      "step": 2360
    },
    {
      "epoch": 3.05,
      "grad_norm": 8.336786270141602,
      "learning_rate": 1.1314413741598208e-05,
      "loss": 0.6445,
      "step": 2370
    },
    {
      "epoch": 3.05,
      "grad_norm": 9.598668098449707,
      "learning_rate": 1.1127707244212099e-05,
      "loss": 0.4292,
      "step": 2380
    },
    {
      "epoch": 3.05,
      "grad_norm": 9.169483184814453,
      "learning_rate": 1.094100074682599e-05,
      "loss": 0.8099,
      "step": 2390
    },
    {
      "epoch": 3.06,
      "grad_norm": 6.843069553375244,
      "learning_rate": 1.0754294249439881e-05,
      "loss": 0.6355,
      "step": 2400
    },
    {
      "epoch": 3.06,
      "grad_norm": 11.256698608398438,
      "learning_rate": 1.0567587752053772e-05,
      "loss": 0.6483,
      "step": 2410
    },
    {
      "epoch": 3.06,
      "grad_norm": 11.205333709716797,
      "learning_rate": 1.0380881254667663e-05,
      "loss": 0.4663,
      "step": 2420
    },
    {
      "epoch": 3.07,
      "grad_norm": 9.922324180603027,
      "learning_rate": 1.0194174757281553e-05,
      "loss": 0.5095,
      "step": 2430
    },
    {
      "epoch": 3.07,
      "grad_norm": 4.500748157501221,
      "learning_rate": 1.0007468259895444e-05,
      "loss": 0.4991,
      "step": 2440
    },
    {
      "epoch": 3.07,
      "grad_norm": 9.476024627685547,
      "learning_rate": 9.820761762509336e-06,
      "loss": 0.4489,
      "step": 2450
    },
    {
      "epoch": 3.08,
      "grad_norm": 8.485849380493164,
      "learning_rate": 9.634055265123227e-06,
      "loss": 0.567,
      "step": 2460
    },
    {
      "epoch": 3.08,
      "grad_norm": 9.129056930541992,
      "learning_rate": 9.447348767737117e-06,
      "loss": 0.5535,
      "step": 2470
    },
    {
      "epoch": 3.08,
      "grad_norm": 7.4503631591796875,
      "learning_rate": 9.260642270351008e-06,
      "loss": 0.4628,
      "step": 2480
    },
    {
      "epoch": 3.09,
      "grad_norm": 11.07274341583252,
      "learning_rate": 9.073935772964899e-06,
      "loss": 0.4511,
      "step": 2490
    },
    {
      "epoch": 3.09,
      "grad_norm": 7.829531192779541,
      "learning_rate": 8.88722927557879e-06,
      "loss": 0.5419,
      "step": 2500
    },
    {
      "epoch": 3.09,
      "grad_norm": 3.9611074924468994,
      "learning_rate": 8.700522778192682e-06,
      "loss": 0.6373,
      "step": 2510
    },
    {
      "epoch": 3.1,
      "grad_norm": 14.42736530303955,
      "learning_rate": 8.513816280806572e-06,
      "loss": 0.6642,
      "step": 2520
    },
    {
      "epoch": 3.1,
      "grad_norm": 10.837944030761719,
      "learning_rate": 8.327109783420463e-06,
      "loss": 0.5636,
      "step": 2530
    },
    {
      "epoch": 3.1,
      "grad_norm": 12.138989448547363,
      "learning_rate": 8.140403286034353e-06,
      "loss": 0.5138,
      "step": 2540
    },
    {
      "epoch": 3.11,
      "grad_norm": 2.928316831588745,
      "learning_rate": 7.953696788648244e-06,
      "loss": 0.4422,
      "step": 2550
    },
    {
      "epoch": 3.11,
      "grad_norm": 11.172917366027832,
      "learning_rate": 7.766990291262136e-06,
      "loss": 0.5037,
      "step": 2560
    },
    {
      "epoch": 3.11,
      "grad_norm": 13.547741889953613,
      "learning_rate": 7.580283793876028e-06,
      "loss": 0.3689,
      "step": 2570
    },
    {
      "epoch": 3.12,
      "grad_norm": 10.06125545501709,
      "learning_rate": 7.393577296489919e-06,
      "loss": 0.5504,
      "step": 2580
    },
    {
      "epoch": 3.12,
      "grad_norm": 12.435505867004395,
      "learning_rate": 7.20687079910381e-06,
      "loss": 0.5295,
      "step": 2590
    },
    {
      "epoch": 3.12,
      "grad_norm": 9.098969459533691,
      "learning_rate": 7.0201643017177005e-06,
      "loss": 0.4936,
      "step": 2600
    },
    {
      "epoch": 3.13,
      "grad_norm": 8.082307815551758,
      "learning_rate": 6.833457804331592e-06,
      "loss": 0.5826,
      "step": 2610
    },
    {
      "epoch": 3.13,
      "grad_norm": 8.9844331741333,
      "learning_rate": 6.6467513069454825e-06,
      "loss": 0.477,
      "step": 2620
    },
    {
      "epoch": 3.13,
      "grad_norm": 2.6742870807647705,
      "learning_rate": 6.460044809559373e-06,
      "loss": 0.4234,
      "step": 2630
    },
    {
      "epoch": 3.14,
      "grad_norm": 12.013636589050293,
      "learning_rate": 6.273338312173265e-06,
      "loss": 0.5719,
      "step": 2640
    },
    {
      "epoch": 3.14,
      "grad_norm": 9.816898345947266,
      "learning_rate": 6.086631814787154e-06,
      "loss": 0.5904,
      "step": 2650
    },
    {
      "epoch": 3.14,
      "grad_norm": 10.359482765197754,
      "learning_rate": 5.899925317401046e-06,
      "loss": 0.5888,
      "step": 2660
    },
    {
      "epoch": 3.15,
      "grad_norm": 6.7240824699401855,
      "learning_rate": 5.7132188200149364e-06,
      "loss": 0.4304,
      "step": 2670
    },
    {
      "epoch": 3.15,
      "grad_norm": 8.453876495361328,
      "learning_rate": 5.526512322628828e-06,
      "loss": 0.5359,
      "step": 2680
    },
    {
      "epoch": 3.15,
      "grad_norm": 14.156214714050293,
      "learning_rate": 5.3398058252427185e-06,
      "loss": 0.6238,
      "step": 2690
    },
    {
      "epoch": 3.16,
      "grad_norm": 7.631002426147461,
      "learning_rate": 5.153099327856609e-06,
      "loss": 0.3718,
      "step": 2700
    },
    {
      "epoch": 3.16,
      "grad_norm": 13.941666603088379,
      "learning_rate": 4.966392830470501e-06,
      "loss": 0.5593,
      "step": 2710
    },
    {
      "epoch": 3.16,
      "grad_norm": 7.249992847442627,
      "learning_rate": 4.779686333084392e-06,
      "loss": 0.4882,
      "step": 2720
    },
    {
      "epoch": 3.17,
      "grad_norm": 15.234835624694824,
      "learning_rate": 4.592979835698283e-06,
      "loss": 0.4251,
      "step": 2730
    },
    {
      "epoch": 3.17,
      "grad_norm": 11.385998725891113,
      "learning_rate": 4.406273338312174e-06,
      "loss": 0.3966,
      "step": 2740
    },
    {
      "epoch": 3.17,
      "grad_norm": 11.18847370147705,
      "learning_rate": 4.219566840926065e-06,
      "loss": 0.4952,
      "step": 2750
    },
    {
      "epoch": 3.18,
      "grad_norm": 5.429126739501953,
      "learning_rate": 4.032860343539955e-06,
      "loss": 0.4902,
      "step": 2760
    },
    {
      "epoch": 3.18,
      "grad_norm": 10.847857475280762,
      "learning_rate": 3.846153846153847e-06,
      "loss": 0.6141,
      "step": 2770
    },
    {
      "epoch": 3.18,
      "grad_norm": 8.353704452514648,
      "learning_rate": 3.6594473487677374e-06,
      "loss": 0.5792,
      "step": 2780
    },
    {
      "epoch": 3.19,
      "grad_norm": 6.862905979156494,
      "learning_rate": 3.4727408513816284e-06,
      "loss": 0.6625,
      "step": 2790
    },
    {
      "epoch": 3.19,
      "grad_norm": 10.818923950195312,
      "learning_rate": 3.2860343539955195e-06,
      "loss": 0.5551,
      "step": 2800
    },
    {
      "epoch": 3.19,
      "grad_norm": 12.606778144836426,
      "learning_rate": 3.09932785660941e-06,
      "loss": 0.7523,
      "step": 2810
    },
    {
      "epoch": 3.2,
      "grad_norm": 14.034588813781738,
      "learning_rate": 2.912621359223301e-06,
      "loss": 0.3656,
      "step": 2820
    },
    {
      "epoch": 3.2,
      "grad_norm": 11.530349731445312,
      "learning_rate": 2.725914861837192e-06,
      "loss": 0.5173,
      "step": 2830
    },
    {
      "epoch": 3.2,
      "grad_norm": 4.992223262786865,
      "learning_rate": 2.539208364451083e-06,
      "loss": 0.5454,
      "step": 2840
    },
    {
      "epoch": 3.21,
      "grad_norm": 18.497516632080078,
      "learning_rate": 2.352501867064974e-06,
      "loss": 0.4571,
      "step": 2850
    },
    {
      "epoch": 3.21,
      "grad_norm": 11.636422157287598,
      "learning_rate": 2.165795369678865e-06,
      "loss": 0.5505,
      "step": 2860
    },
    {
      "epoch": 3.21,
      "grad_norm": 15.542102813720703,
      "learning_rate": 1.979088872292756e-06,
      "loss": 0.6034,
      "step": 2870
    },
    {
      "epoch": 3.22,
      "grad_norm": 5.758667945861816,
      "learning_rate": 1.7923823749066467e-06,
      "loss": 0.4881,
      "step": 2880
    },
    {
      "epoch": 3.22,
      "grad_norm": 8.1626615524292,
      "learning_rate": 1.6056758775205377e-06,
      "loss": 0.52,
      "step": 2890
    },
    {
      "epoch": 3.22,
      "grad_norm": 3.6557774543762207,
      "learning_rate": 1.4189693801344288e-06,
      "loss": 0.6039,
      "step": 2900
    },
    {
      "epoch": 3.23,
      "grad_norm": 11.118461608886719,
      "learning_rate": 1.2322628827483198e-06,
      "loss": 0.3346,
      "step": 2910
    },
    {
      "epoch": 3.23,
      "grad_norm": 7.280307292938232,
      "learning_rate": 1.0455563853622106e-06,
      "loss": 0.4897,
      "step": 2920
    },
    {
      "epoch": 3.23,
      "grad_norm": 8.602575302124023,
      "learning_rate": 8.588498879761017e-07,
      "loss": 0.4007,
      "step": 2930
    },
    {
      "epoch": 3.24,
      "grad_norm": 11.347135543823242,
      "learning_rate": 6.721433905899926e-07,
      "loss": 0.3186,
      "step": 2940
    },
    {
      "epoch": 3.24,
      "grad_norm": 4.977802753448486,
      "learning_rate": 4.854368932038835e-07,
      "loss": 0.3876,
      "step": 2950
    },
    {
      "epoch": 3.24,
      "grad_norm": 5.7857184410095215,
      "learning_rate": 2.987303958177745e-07,
      "loss": 0.3962,
      "step": 2960
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.2075411081314087,
      "learning_rate": 1.1202389843166542e-07,
      "loss": 0.5551,
      "step": 2970
    },
    {
      "epoch": 3.25,
      "eval_accuracy": 0.6438450899031812,
      "eval_loss": 0.9391621351242065,
      "eval_runtime": 1083.7107,
      "eval_samples_per_second": 1.334,
      "eval_steps_per_second": 0.167,
      "step": 2976
    },
    {
      "epoch": 3.25,
      "step": 2976,
      "total_flos": 2.0841444704018792e+19,
      "train_loss": 0.8609641519605472,
      "train_runtime": 58939.9677,
      "train_samples_per_second": 0.404,
      "train_steps_per_second": 0.05
    },
    {
      "epoch": 3.25,
      "eval_accuracy": 0.6119610570236439,
      "eval_loss": 0.9484543204307556,
      "eval_runtime": 1115.5395,
      "eval_samples_per_second": 1.289,
      "eval_steps_per_second": 0.161,
      "step": 2976
    },
    {
      "epoch": 3.25,
      "eval_accuracy": 0.6119610570236439,
      "eval_loss": 0.9484543204307556,
      "eval_runtime": 1071.6262,
      "eval_samples_per_second": 1.342,
      "eval_steps_per_second": 0.168,
      "step": 2976
    }
  ],
  "logging_steps": 10,
  "max_steps": 2976,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "total_flos": 2.0841444704018792e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}