{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 84.50704225352112,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.28,
      "learning_rate": 9.974991663887962e-05,
      "loss": 7.941,
      "step": 20
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.941647215738579e-05,
      "loss": 5.5312,
      "step": 40
    },
    {
      "epoch": 0.85,
      "learning_rate": 9.908302767589197e-05,
      "loss": 5.3579,
      "step": 60
    },
    {
      "epoch": 1.13,
      "learning_rate": 9.874958319439813e-05,
      "loss": 5.1472,
      "step": 80
    },
    {
      "epoch": 1.41,
      "learning_rate": 9.841613871290431e-05,
      "loss": 4.9751,
      "step": 100
    },
    {
      "epoch": 1.69,
      "learning_rate": 9.808269423141048e-05,
      "loss": 4.8703,
      "step": 120
    },
    {
      "epoch": 1.97,
      "learning_rate": 9.774924974991663e-05,
      "loss": 4.7921,
      "step": 140
    },
    {
      "epoch": 2.25,
      "learning_rate": 9.741580526842281e-05,
      "loss": 4.7422,
      "step": 160
    },
    {
      "epoch": 2.54,
      "learning_rate": 9.708236078692898e-05,
      "loss": 4.6846,
      "step": 180
    },
    {
      "epoch": 2.82,
      "learning_rate": 9.674891630543515e-05,
      "loss": 4.6358,
      "step": 200
    },
    {
      "epoch": 2.82,
      "eval_validation_accuracy": 0.15669611355371005,
      "eval_validation_loss": 7.171875,
      "eval_validation_runtime": 104.6901,
      "eval_validation_samples_per_second": 6.877,
      "eval_validation_steps_per_second": 0.43,
      "step": 200
    },
    {
      "epoch": 3.1,
      "learning_rate": 9.641547182394132e-05,
      "loss": 4.5991,
      "step": 220
    },
    {
      "epoch": 3.38,
      "learning_rate": 9.608202734244749e-05,
      "loss": 4.5588,
      "step": 240
    },
    {
      "epoch": 3.66,
      "learning_rate": 9.574858286095365e-05,
      "loss": 4.5232,
      "step": 260
    },
    {
      "epoch": 3.94,
      "learning_rate": 9.541513837945982e-05,
      "loss": 4.4853,
      "step": 280
    },
    {
      "epoch": 4.23,
      "learning_rate": 9.508169389796599e-05,
      "loss": 4.4635,
      "step": 300
    },
    {
      "epoch": 4.51,
      "learning_rate": 9.474824941647216e-05,
      "loss": 4.4351,
      "step": 320
    },
    {
      "epoch": 4.79,
      "learning_rate": 9.441480493497833e-05,
      "loss": 4.4139,
      "step": 340
    },
    {
      "epoch": 5.07,
      "learning_rate": 9.408136045348451e-05,
      "loss": 4.3901,
      "step": 360
    },
    {
      "epoch": 5.35,
      "learning_rate": 9.374791597199066e-05,
      "loss": 4.3642,
      "step": 380
    },
    {
      "epoch": 5.63,
      "learning_rate": 9.341447149049683e-05,
      "loss": 4.3427,
      "step": 400
    },
    {
      "epoch": 5.63,
      "eval_validation_accuracy": 0.1884003691038376,
      "eval_validation_loss": 6.74609375,
      "eval_validation_runtime": 99.5037,
      "eval_validation_samples_per_second": 7.236,
      "eval_validation_steps_per_second": 0.452,
      "step": 400
    },
    {
      "epoch": 5.92,
      "learning_rate": 9.308102700900301e-05,
      "loss": 4.3131,
      "step": 420
    },
    {
      "epoch": 6.2,
      "learning_rate": 9.274758252750917e-05,
      "loss": 4.3033,
      "step": 440
    },
    {
      "epoch": 6.48,
      "learning_rate": 9.241413804601535e-05,
      "loss": 4.2887,
      "step": 460
    },
    {
      "epoch": 6.76,
      "learning_rate": 9.208069356452151e-05,
      "loss": 4.2642,
      "step": 480
    },
    {
      "epoch": 7.04,
      "learning_rate": 9.174724908302767e-05,
      "loss": 4.2506,
      "step": 500
    },
    {
      "epoch": 7.32,
      "learning_rate": 9.141380460153385e-05,
      "loss": 4.2259,
      "step": 520
    },
    {
      "epoch": 7.61,
      "learning_rate": 9.108036012004002e-05,
      "loss": 4.2121,
      "step": 540
    },
    {
      "epoch": 7.89,
      "learning_rate": 9.074691563854619e-05,
      "loss": 4.2003,
      "step": 560
    },
    {
      "epoch": 8.17,
      "learning_rate": 9.041347115705235e-05,
      "loss": 4.1863,
      "step": 580
    },
    {
      "epoch": 8.45,
      "learning_rate": 9.008002667555852e-05,
      "loss": 4.17,
      "step": 600
    },
    {
      "epoch": 8.45,
      "eval_validation_accuracy": 0.2065807686044618,
      "eval_validation_loss": 6.49609375,
      "eval_validation_runtime": 99.5839,
      "eval_validation_samples_per_second": 7.23,
      "eval_validation_steps_per_second": 0.452,
      "step": 600
    },
    {
      "epoch": 8.73,
      "learning_rate": 8.974658219406469e-05,
      "loss": 4.1538,
      "step": 620
    },
    {
      "epoch": 9.01,
      "learning_rate": 8.941313771257086e-05,
      "loss": 4.1417,
      "step": 640
    },
    {
      "epoch": 9.3,
      "learning_rate": 8.907969323107703e-05,
      "loss": 4.1313,
      "step": 660
    },
    {
      "epoch": 9.58,
      "learning_rate": 8.87462487495832e-05,
      "loss": 4.1168,
      "step": 680
    },
    {
      "epoch": 9.86,
      "learning_rate": 8.841280426808936e-05,
      "loss": 4.1002,
      "step": 700
    },
    {
      "epoch": 10.14,
      "learning_rate": 8.807935978659554e-05,
      "loss": 4.0983,
      "step": 720
    },
    {
      "epoch": 10.42,
      "learning_rate": 8.77459153051017e-05,
      "loss": 4.0787,
      "step": 740
    },
    {
      "epoch": 10.7,
      "learning_rate": 8.741247082360787e-05,
      "loss": 4.0627,
      "step": 760
    },
    {
      "epoch": 10.99,
      "learning_rate": 8.707902634211405e-05,
      "loss": 4.065,
      "step": 780
    },
    {
      "epoch": 11.27,
      "learning_rate": 8.67455818606202e-05,
      "loss": 4.0396,
      "step": 800
    },
    {
      "epoch": 11.27,
      "eval_validation_accuracy": 0.2191336915811757,
      "eval_validation_loss": 6.32421875,
      "eval_validation_runtime": 98.361,
      "eval_validation_samples_per_second": 7.32,
      "eval_validation_steps_per_second": 0.457,
      "step": 800
    },
    {
      "epoch": 11.55,
      "learning_rate": 8.641213737912638e-05,
      "loss": 4.0274,
      "step": 820
    },
    {
      "epoch": 11.83,
      "learning_rate": 8.607869289763255e-05,
      "loss": 4.0333,
      "step": 840
    },
    {
      "epoch": 12.11,
      "learning_rate": 8.57452484161387e-05,
      "loss": 4.0114,
      "step": 860
    },
    {
      "epoch": 12.39,
      "learning_rate": 8.541180393464489e-05,
      "loss": 4.0008,
      "step": 880
    },
    {
      "epoch": 12.68,
      "learning_rate": 8.507835945315106e-05,
      "loss": 3.9874,
      "step": 900
    },
    {
      "epoch": 12.96,
      "learning_rate": 8.474491497165722e-05,
      "loss": 3.9933,
      "step": 920
    },
    {
      "epoch": 13.24,
      "learning_rate": 8.441147049016339e-05,
      "loss": 3.9708,
      "step": 940
    },
    {
      "epoch": 13.52,
      "learning_rate": 8.407802600866956e-05,
      "loss": 3.9785,
      "step": 960
    },
    {
      "epoch": 13.8,
      "learning_rate": 8.374458152717573e-05,
      "loss": 3.9613,
      "step": 980
    },
    {
      "epoch": 14.08,
      "learning_rate": 8.34111370456819e-05,
      "loss": 3.9593,
      "step": 1000
    },
    {
      "epoch": 14.08,
      "eval_validation_accuracy": 0.2270911360799001,
      "eval_validation_loss": 6.203125,
      "eval_validation_runtime": 102.4606,
      "eval_validation_samples_per_second": 7.027,
      "eval_validation_steps_per_second": 0.439,
      "step": 1000
    },
    {
      "epoch": 14.37,
      "learning_rate": 8.311103701233745e-05,
      "loss": 3.962,
      "step": 1020
    },
    {
      "epoch": 14.65,
      "learning_rate": 8.277759253084362e-05,
      "loss": 3.9353,
      "step": 1040
    },
    {
      "epoch": 14.93,
      "learning_rate": 8.244414804934978e-05,
      "loss": 3.9261,
      "step": 1060
    },
    {
      "epoch": 15.21,
      "learning_rate": 8.211070356785595e-05,
      "loss": 3.9191,
      "step": 1080
    },
    {
      "epoch": 15.49,
      "learning_rate": 8.177725908636212e-05,
      "loss": 3.9148,
      "step": 1100
    },
    {
      "epoch": 15.77,
      "learning_rate": 8.14438146048683e-05,
      "loss": 3.9018,
      "step": 1120
    },
    {
      "epoch": 16.06,
      "learning_rate": 8.111037012337445e-05,
      "loss": 3.8976,
      "step": 1140
    },
    {
      "epoch": 16.34,
      "learning_rate": 8.077692564188064e-05,
      "loss": 3.8911,
      "step": 1160
    },
    {
      "epoch": 16.62,
      "learning_rate": 8.04434811603868e-05,
      "loss": 3.8881,
      "step": 1180
    },
    {
      "epoch": 16.9,
      "learning_rate": 8.011003667889296e-05,
      "loss": 3.8797,
      "step": 1200
    },
    {
      "epoch": 16.9,
      "eval_validation_accuracy": 0.23543125441024806,
      "eval_validation_loss": 6.0859375,
      "eval_validation_runtime": 104.2399,
      "eval_validation_samples_per_second": 6.907,
      "eval_validation_steps_per_second": 0.432,
      "step": 1200
    },
    {
      "epoch": 17.18,
      "learning_rate": 7.977659219739914e-05,
      "loss": 3.8746,
      "step": 1220
    },
    {
      "epoch": 17.46,
      "learning_rate": 7.944314771590531e-05,
      "loss": 3.8631,
      "step": 1240
    },
    {
      "epoch": 17.75,
      "learning_rate": 7.910970323441148e-05,
      "loss": 3.8594,
      "step": 1260
    },
    {
      "epoch": 18.03,
      "learning_rate": 7.877625875291764e-05,
      "loss": 3.8526,
      "step": 1280
    },
    {
      "epoch": 18.31,
      "learning_rate": 7.844281427142381e-05,
      "loss": 3.8454,
      "step": 1300
    },
    {
      "epoch": 18.59,
      "learning_rate": 7.810936978992998e-05,
      "loss": 3.8357,
      "step": 1320
    },
    {
      "epoch": 18.87,
      "learning_rate": 7.777592530843615e-05,
      "loss": 3.8338,
      "step": 1340
    },
    {
      "epoch": 19.15,
      "learning_rate": 7.744248082694232e-05,
      "loss": 3.8283,
      "step": 1360
    },
    {
      "epoch": 19.44,
      "learning_rate": 7.710903634544848e-05,
      "loss": 3.8205,
      "step": 1380
    },
    {
      "epoch": 19.72,
      "learning_rate": 7.677559186395465e-05,
      "loss": 3.8174,
      "step": 1400
    },
    {
      "epoch": 19.72,
      "eval_validation_accuracy": 0.24027981327688216,
      "eval_validation_loss": 6.0,
      "eval_validation_runtime": 101.9857,
      "eval_validation_samples_per_second": 7.06,
      "eval_validation_steps_per_second": 0.441,
      "step": 1400
    },
    {
      "epoch": 20.0,
      "learning_rate": 7.644214738246082e-05,
      "loss": 3.8131,
      "step": 1420
    },
    {
      "epoch": 20.28,
      "learning_rate": 7.610870290096699e-05,
      "loss": 3.7894,
      "step": 1440
    },
    {
      "epoch": 20.56,
      "learning_rate": 7.577525841947316e-05,
      "loss": 3.8069,
      "step": 1460
    },
    {
      "epoch": 20.85,
      "learning_rate": 7.544181393797934e-05,
      "loss": 3.799,
      "step": 1480
    },
    {
      "epoch": 21.13,
      "learning_rate": 7.510836945648549e-05,
      "loss": 3.7829,
      "step": 1500
    },
    {
      "epoch": 21.41,
      "learning_rate": 7.477492497499166e-05,
      "loss": 3.7852,
      "step": 1520
    },
    {
      "epoch": 21.69,
      "learning_rate": 7.444148049349784e-05,
      "loss": 3.7721,
      "step": 1540
    },
    {
      "epoch": 21.97,
      "learning_rate": 7.4108036012004e-05,
      "loss": 3.7795,
      "step": 1560
    },
    {
      "epoch": 22.25,
      "learning_rate": 7.377459153051018e-05,
      "loss": 3.7653,
      "step": 1580
    },
    {
      "epoch": 22.54,
      "learning_rate": 7.344114704901635e-05,
      "loss": 3.7576,
      "step": 1600
    },
    {
      "epoch": 22.54,
      "eval_validation_accuracy": 0.24577837485751505,
      "eval_validation_loss": 5.9296875,
      "eval_validation_runtime": 103.1085,
      "eval_validation_samples_per_second": 6.983,
      "eval_validation_steps_per_second": 0.436,
      "step": 1600
    },
    {
      "epoch": 22.82,
      "learning_rate": 7.310770256752251e-05,
      "loss": 3.7596,
      "step": 1620
    },
    {
      "epoch": 23.1,
      "learning_rate": 7.277425808602868e-05,
      "loss": 3.7564,
      "step": 1640
    },
    {
      "epoch": 23.38,
      "learning_rate": 7.244081360453485e-05,
      "loss": 3.7419,
      "step": 1660
    },
    {
      "epoch": 23.66,
      "learning_rate": 7.210736912304102e-05,
      "loss": 3.7405,
      "step": 1680
    },
    {
      "epoch": 23.94,
      "learning_rate": 7.177392464154719e-05,
      "loss": 3.7394,
      "step": 1700
    },
    {
      "epoch": 24.23,
      "learning_rate": 7.144048016005335e-05,
      "loss": 3.7307,
      "step": 1720
    },
    {
      "epoch": 24.51,
      "learning_rate": 7.110703567855952e-05,
      "loss": 3.7282,
      "step": 1740
    },
    {
      "epoch": 24.79,
      "learning_rate": 7.077359119706569e-05,
      "loss": 3.7257,
      "step": 1760
    },
    {
      "epoch": 25.07,
      "learning_rate": 7.044014671557186e-05,
      "loss": 3.7163,
      "step": 1780
    },
    {
      "epoch": 25.35,
      "learning_rate": 7.010670223407803e-05,
      "loss": 3.7086,
      "step": 1800
    },
    {
      "epoch": 25.35,
      "eval_validation_accuracy": 0.2498147695815014,
      "eval_validation_loss": 5.87109375,
      "eval_validation_runtime": 101.5018,
      "eval_validation_samples_per_second": 7.093,
      "eval_validation_steps_per_second": 0.443,
      "step": 1800
    },
    {
      "epoch": 25.63,
      "learning_rate": 6.97732577525842e-05,
      "loss": 3.7055,
      "step": 1820
    },
    {
      "epoch": 25.92,
      "learning_rate": 6.943981327109038e-05,
      "loss": 3.7116,
      "step": 1840
    },
    {
      "epoch": 26.2,
      "learning_rate": 6.910636878959653e-05,
      "loss": 3.6953,
      "step": 1860
    },
    {
      "epoch": 26.48,
      "learning_rate": 6.87729243081027e-05,
      "loss": 3.7016,
      "step": 1880
    },
    {
      "epoch": 26.76,
      "learning_rate": 6.843947982660888e-05,
      "loss": 3.6928,
      "step": 1900
    },
    {
      "epoch": 27.04,
      "learning_rate": 6.810603534511503e-05,
      "loss": 3.6869,
      "step": 1920
    },
    {
      "epoch": 27.32,
      "learning_rate": 6.777259086362121e-05,
      "loss": 3.6707,
      "step": 1940
    },
    {
      "epoch": 27.61,
      "learning_rate": 6.743914638212738e-05,
      "loss": 3.6783,
      "step": 1960
    },
    {
      "epoch": 27.89,
      "learning_rate": 6.710570190063354e-05,
      "loss": 3.6793,
      "step": 1980
    },
    {
      "epoch": 28.17,
      "learning_rate": 6.677225741913972e-05,
      "loss": 3.6683,
      "step": 2000
    },
    {
      "epoch": 28.17,
      "eval_validation_accuracy": 0.25368289637952557,
      "eval_validation_loss": 5.80859375,
      "eval_validation_runtime": 101.9759,
      "eval_validation_samples_per_second": 7.06,
      "eval_validation_steps_per_second": 0.441,
      "step": 2000
    },
    {
      "epoch": 28.45,
      "learning_rate": 6.647215738579527e-05,
      "loss": 3.6639,
      "step": 2020
    },
    {
      "epoch": 28.73,
      "learning_rate": 6.613871290430144e-05,
      "loss": 3.6603,
      "step": 2040
    },
    {
      "epoch": 29.01,
      "learning_rate": 6.58052684228076e-05,
      "loss": 3.6586,
      "step": 2060
    },
    {
      "epoch": 29.3,
      "learning_rate": 6.547182394131377e-05,
      "loss": 3.6459,
      "step": 2080
    },
    {
      "epoch": 29.58,
      "learning_rate": 6.513837945981994e-05,
      "loss": 3.651,
      "step": 2100
    },
    {
      "epoch": 29.86,
      "learning_rate": 6.480493497832611e-05,
      "loss": 3.6535,
      "step": 2120
    },
    {
      "epoch": 30.14,
      "learning_rate": 6.447149049683228e-05,
      "loss": 3.6479,
      "step": 2140
    },
    {
      "epoch": 30.42,
      "learning_rate": 6.413804601533845e-05,
      "loss": 3.6367,
      "step": 2160
    },
    {
      "epoch": 30.7,
      "learning_rate": 6.380460153384463e-05,
      "loss": 3.6371,
      "step": 2180
    },
    {
      "epoch": 30.99,
      "learning_rate": 6.347115705235078e-05,
      "loss": 3.6359,
      "step": 2200
    },
    {
      "epoch": 30.99,
      "eval_validation_accuracy": 0.25744992672203226,
      "eval_validation_loss": 5.765625,
      "eval_validation_runtime": 101.6629,
      "eval_validation_samples_per_second": 7.082,
      "eval_validation_steps_per_second": 0.443,
      "step": 2200
    },
    {
      "epoch": 31.27,
      "learning_rate": 6.313771257085695e-05,
      "loss": 3.6198,
      "step": 2220
    },
    {
      "epoch": 31.55,
      "learning_rate": 6.280426808936313e-05,
      "loss": 3.6257,
      "step": 2240
    },
    {
      "epoch": 31.83,
      "learning_rate": 6.247082360786929e-05,
      "loss": 3.6197,
      "step": 2260
    },
    {
      "epoch": 32.11,
      "learning_rate": 6.213737912637547e-05,
      "loss": 3.6176,
      "step": 2280
    },
    {
      "epoch": 32.39,
      "learning_rate": 6.180393464488164e-05,
      "loss": 3.6092,
      "step": 2300
    },
    {
      "epoch": 32.68,
      "learning_rate": 6.147049016338779e-05,
      "loss": 3.6158,
      "step": 2320
    },
    {
      "epoch": 32.96,
      "learning_rate": 6.113704568189397e-05,
      "loss": 3.6158,
      "step": 2340
    },
    {
      "epoch": 33.24,
      "learning_rate": 6.080360120040014e-05,
      "loss": 3.6058,
      "step": 2360
    },
    {
      "epoch": 33.52,
      "learning_rate": 6.04701567189063e-05,
      "loss": 3.5952,
      "step": 2380
    },
    {
      "epoch": 33.8,
      "learning_rate": 6.0136712237412475e-05,
      "loss": 3.5999,
      "step": 2400
    },
    {
      "epoch": 33.8,
      "eval_validation_accuracy": 0.2595444553004397,
      "eval_validation_loss": 5.72265625,
      "eval_validation_runtime": 102.324,
      "eval_validation_samples_per_second": 7.036,
      "eval_validation_steps_per_second": 0.44,
      "step": 2400
    },
    {
      "epoch": 34.08,
      "learning_rate": 5.980326775591864e-05,
      "loss": 3.5934,
      "step": 2420
    },
    {
      "epoch": 34.37,
      "learning_rate": 5.9469823274424805e-05,
      "loss": 3.5853,
      "step": 2440
    },
    {
      "epoch": 34.65,
      "learning_rate": 5.913637879293098e-05,
      "loss": 3.5871,
      "step": 2460
    },
    {
      "epoch": 34.93,
      "learning_rate": 5.8802934311437154e-05,
      "loss": 3.5904,
      "step": 2480
    },
    {
      "epoch": 35.21,
      "learning_rate": 5.8469489829943315e-05,
      "loss": 3.5769,
      "step": 2500
    },
    {
      "epoch": 35.49,
      "learning_rate": 5.813604534844948e-05,
      "loss": 3.5827,
      "step": 2520
    },
    {
      "epoch": 35.77,
      "learning_rate": 5.780260086695566e-05,
      "loss": 3.5765,
      "step": 2540
    },
    {
      "epoch": 36.06,
      "learning_rate": 5.746915638546182e-05,
      "loss": 3.5735,
      "step": 2560
    },
    {
      "epoch": 36.34,
      "learning_rate": 5.7135711903967994e-05,
      "loss": 3.5705,
      "step": 2580
    },
    {
      "epoch": 36.62,
      "learning_rate": 5.680226742247416e-05,
      "loss": 3.5674,
      "step": 2600
    },
    {
      "epoch": 36.62,
      "eval_validation_accuracy": 0.26218789013732835,
      "eval_validation_loss": 5.6875,
      "eval_validation_runtime": 103.4486,
      "eval_validation_samples_per_second": 6.96,
      "eval_validation_steps_per_second": 0.435,
      "step": 2600
    },
    {
      "epoch": 36.9,
      "learning_rate": 5.646882294098032e-05,
      "loss": 3.5564,
      "step": 2620
    },
    {
      "epoch": 37.18,
      "learning_rate": 5.61353784594865e-05,
      "loss": 3.558,
      "step": 2640
    },
    {
      "epoch": 37.46,
      "learning_rate": 5.580193397799267e-05,
      "loss": 3.5563,
      "step": 2660
    },
    {
      "epoch": 37.75,
      "learning_rate": 5.5468489496498834e-05,
      "loss": 3.55,
      "step": 2680
    },
    {
      "epoch": 38.03,
      "learning_rate": 5.5135045015005e-05,
      "loss": 3.5521,
      "step": 2700
    },
    {
      "epoch": 38.31,
      "learning_rate": 5.480160053351118e-05,
      "loss": 3.5432,
      "step": 2720
    },
    {
      "epoch": 38.59,
      "learning_rate": 5.446815605201734e-05,
      "loss": 3.5401,
      "step": 2740
    },
    {
      "epoch": 38.87,
      "learning_rate": 5.413471157052351e-05,
      "loss": 3.5478,
      "step": 2760
    },
    {
      "epoch": 39.15,
      "learning_rate": 5.380126708902968e-05,
      "loss": 3.535,
      "step": 2780
    },
    {
      "epoch": 39.44,
      "learning_rate": 5.346782260753584e-05,
      "loss": 3.539,
      "step": 2800
    },
    {
      "epoch": 39.44,
      "eval_validation_accuracy": 0.2642464582315584,
      "eval_validation_loss": 5.65625,
      "eval_validation_runtime": 106.3096,
      "eval_validation_samples_per_second": 6.773,
      "eval_validation_steps_per_second": 0.423,
      "step": 2800
    },
    {
      "epoch": 39.72,
      "learning_rate": 5.3134378126042017e-05,
      "loss": 3.5391,
      "step": 2820
    },
    {
      "epoch": 40.0,
      "learning_rate": 5.280093364454819e-05,
      "loss": 3.5356,
      "step": 2840
    },
    {
      "epoch": 40.28,
      "learning_rate": 5.246748916305435e-05,
      "loss": 3.5212,
      "step": 2860
    },
    {
      "epoch": 40.56,
      "learning_rate": 5.213404468156052e-05,
      "loss": 3.5203,
      "step": 2880
    },
    {
      "epoch": 40.85,
      "learning_rate": 5.1800600200066695e-05,
      "loss": 3.5268,
      "step": 2900
    },
    {
      "epoch": 41.13,
      "learning_rate": 5.1467155718572856e-05,
      "loss": 3.5222,
      "step": 2920
    },
    {
      "epoch": 41.41,
      "learning_rate": 5.113371123707903e-05,
      "loss": 3.5145,
      "step": 2940
    },
    {
      "epoch": 41.69,
      "learning_rate": 5.08002667555852e-05,
      "loss": 3.5183,
      "step": 2960
    },
    {
      "epoch": 41.97,
      "learning_rate": 5.046682227409136e-05,
      "loss": 3.5173,
      "step": 2980
    },
    {
      "epoch": 42.25,
      "learning_rate": 5.0133377792597535e-05,
      "loss": 3.5035,
      "step": 3000
    },
    {
      "epoch": 42.25,
      "eval_validation_accuracy": 0.2663450578081746,
      "eval_validation_loss": 5.62890625,
      "eval_validation_runtime": 102.2875,
      "eval_validation_samples_per_second": 7.039,
      "eval_validation_steps_per_second": 0.44,
      "step": 3000
    },
    {
      "epoch": 42.54,
      "learning_rate": 4.9833277759253086e-05,
      "loss": 3.5085,
      "step": 3020
    },
    {
      "epoch": 42.82,
      "learning_rate": 4.9499833277759254e-05,
      "loss": 3.5061,
      "step": 3040
    },
    {
      "epoch": 43.1,
      "learning_rate": 4.916638879626542e-05,
      "loss": 3.506,
      "step": 3060
    },
    {
      "epoch": 43.38,
      "learning_rate": 4.88329443147716e-05,
      "loss": 3.5051,
      "step": 3080
    },
    {
      "epoch": 43.66,
      "learning_rate": 4.849949983327776e-05,
      "loss": 3.4926,
      "step": 3100
    },
    {
      "epoch": 43.94,
      "learning_rate": 4.8166055351783926e-05,
      "loss": 3.4964,
      "step": 3120
    },
    {
      "epoch": 44.23,
      "learning_rate": 4.78326108702901e-05,
      "loss": 3.4876,
      "step": 3140
    },
    {
      "epoch": 44.51,
      "learning_rate": 4.749916638879627e-05,
      "loss": 3.4919,
      "step": 3160
    },
    {
      "epoch": 44.79,
      "learning_rate": 4.716572190730244e-05,
      "loss": 3.495,
      "step": 3180
    },
    {
      "epoch": 45.07,
      "learning_rate": 4.6832277425808605e-05,
      "loss": 3.4821,
      "step": 3200
    },
    {
      "epoch": 45.07,
      "eval_validation_accuracy": 0.2684633338761331,
      "eval_validation_loss": 5.6015625,
      "eval_validation_runtime": 101.1239,
      "eval_validation_samples_per_second": 7.12,
      "eval_validation_steps_per_second": 0.445,
      "step": 3200
    },
    {
      "epoch": 45.35,
      "learning_rate": 4.649883294431477e-05,
      "loss": 3.4852,
      "step": 3220
    },
    {
      "epoch": 45.63,
      "learning_rate": 4.616538846282094e-05,
      "loss": 3.4793,
      "step": 3240
    },
    {
      "epoch": 45.92,
      "learning_rate": 4.5831943981327116e-05,
      "loss": 3.483,
      "step": 3260
    },
    {
      "epoch": 46.2,
      "learning_rate": 4.549849949983328e-05,
      "loss": 3.4746,
      "step": 3280
    },
    {
      "epoch": 46.48,
      "learning_rate": 4.5165055018339445e-05,
      "loss": 3.4752,
      "step": 3300
    },
    {
      "epoch": 46.76,
      "learning_rate": 4.483161053684562e-05,
      "loss": 3.4752,
      "step": 3320
    },
    {
      "epoch": 47.04,
      "learning_rate": 4.449816605535179e-05,
      "loss": 3.4763,
      "step": 3340
    },
    {
      "epoch": 47.32,
      "learning_rate": 4.4164721573857956e-05,
      "loss": 3.4687,
      "step": 3360
    },
    {
      "epoch": 47.61,
      "learning_rate": 4.3831277092364124e-05,
      "loss": 3.4674,
      "step": 3380
    },
    {
      "epoch": 47.89,
      "learning_rate": 4.349783261087029e-05,
      "loss": 3.4664,
      "step": 3400
    },
    {
      "epoch": 47.89,
      "eval_validation_accuracy": 0.2700048851978505,
      "eval_validation_loss": 5.58203125,
      "eval_validation_runtime": 99.6804,
      "eval_validation_samples_per_second": 7.223,
      "eval_validation_steps_per_second": 0.451,
      "step": 3400
    },
    {
      "epoch": 48.17,
      "learning_rate": 4.316438812937646e-05,
      "loss": 3.4543,
      "step": 3420
    },
    {
      "epoch": 48.45,
      "learning_rate": 4.2830943647882634e-05,
      "loss": 3.456,
      "step": 3440
    },
    {
      "epoch": 48.73,
      "learning_rate": 4.2497499166388796e-05,
      "loss": 3.4651,
      "step": 3460
    },
    {
      "epoch": 49.01,
      "learning_rate": 4.2164054684894964e-05,
      "loss": 3.4603,
      "step": 3480
    },
    {
      "epoch": 49.3,
      "learning_rate": 4.183061020340114e-05,
      "loss": 3.4466,
      "step": 3500
    },
    {
      "epoch": 49.58,
      "learning_rate": 4.1497165721907306e-05,
      "loss": 3.4556,
      "step": 3520
    },
    {
      "epoch": 49.86,
      "learning_rate": 4.1163721240413474e-05,
      "loss": 3.4501,
      "step": 3540
    },
    {
      "epoch": 50.14,
      "learning_rate": 4.083027675891964e-05,
      "loss": 3.4525,
      "step": 3560
    },
    {
      "epoch": 50.42,
      "learning_rate": 4.049683227742581e-05,
      "loss": 3.4449,
      "step": 3580
    },
    {
      "epoch": 50.7,
      "learning_rate": 4.016338779593198e-05,
      "loss": 3.4456,
      "step": 3600
    },
    {
      "epoch": 50.7,
      "eval_validation_accuracy": 0.2715294740270314,
      "eval_validation_loss": 5.55859375,
      "eval_validation_runtime": 99.8031,
      "eval_validation_samples_per_second": 7.214,
      "eval_validation_steps_per_second": 0.451,
      "step": 3600
    },
    {
      "epoch": 50.99,
      "learning_rate": 3.982994331443815e-05,
      "loss": 3.4467,
      "step": 3620
    },
    {
      "epoch": 51.27,
      "learning_rate": 3.9496498832944314e-05,
      "loss": 3.4418,
      "step": 3640
    },
    {
      "epoch": 51.55,
      "learning_rate": 3.916305435145048e-05,
      "loss": 3.4379,
      "step": 3660
    },
    {
      "epoch": 51.83,
      "learning_rate": 3.882960986995666e-05,
      "loss": 3.4415,
      "step": 3680
    },
    {
      "epoch": 52.11,
      "learning_rate": 3.8496165388462825e-05,
      "loss": 3.436,
      "step": 3700
    },
    {
      "epoch": 52.39,
      "learning_rate": 3.816272090696899e-05,
      "loss": 3.4276,
      "step": 3720
    },
    {
      "epoch": 52.68,
      "learning_rate": 3.7829276425475154e-05,
      "loss": 3.4394,
      "step": 3740
    },
    {
      "epoch": 52.96,
      "learning_rate": 3.749583194398133e-05,
      "loss": 3.4257,
      "step": 3760
    },
    {
      "epoch": 53.24,
      "learning_rate": 3.71623874624875e-05,
      "loss": 3.4201,
      "step": 3780
    },
    {
      "epoch": 53.52,
      "learning_rate": 3.6828942980993665e-05,
      "loss": 3.4257,
      "step": 3800
    },
    {
      "epoch": 53.52,
      "eval_validation_accuracy": 0.2729271834120393,
      "eval_validation_loss": 5.5390625,
      "eval_validation_runtime": 101.1833,
      "eval_validation_samples_per_second": 7.116,
      "eval_validation_steps_per_second": 0.445,
      "step": 3800
    },
    {
      "epoch": 53.8,
      "learning_rate": 3.649549849949983e-05,
      "loss": 3.4246,
      "step": 3820
    },
    {
      "epoch": 54.08,
      "learning_rate": 3.6162054018006e-05,
      "loss": 3.426,
      "step": 3840
    },
    {
      "epoch": 54.37,
      "learning_rate": 3.582860953651217e-05,
      "loss": 3.4149,
      "step": 3860
    },
    {
      "epoch": 54.65,
      "learning_rate": 3.5495165055018344e-05,
      "loss": 3.419,
      "step": 3880
    },
    {
      "epoch": 54.93,
      "learning_rate": 3.516172057352451e-05,
      "loss": 3.4144,
      "step": 3900
    },
    {
      "epoch": 55.21,
      "learning_rate": 3.482827609203067e-05,
      "loss": 3.4105,
      "step": 3920
    },
    {
      "epoch": 55.49,
      "learning_rate": 3.449483161053685e-05,
      "loss": 3.4093,
      "step": 3940
    },
    {
      "epoch": 55.77,
      "learning_rate": 3.4161387129043016e-05,
      "loss": 3.4146,
      "step": 3960
    },
    {
      "epoch": 56.06,
      "learning_rate": 3.3827942647549183e-05,
      "loss": 3.412,
      "step": 3980
    },
    {
      "epoch": 56.34,
      "learning_rate": 3.349449816605535e-05,
      "loss": 3.4016,
      "step": 4000
    },
    {
      "epoch": 56.34,
      "eval_validation_accuracy": 0.27392186397437984,
      "eval_validation_loss": 5.5234375,
      "eval_validation_runtime": 101.7145,
      "eval_validation_samples_per_second": 7.079,
      "eval_validation_steps_per_second": 0.442,
      "step": 4000
    },
    {
      "epoch": 56.62,
      "learning_rate": 3.319439813271091e-05,
      "loss": 3.41,
      "step": 4020
    },
    {
      "epoch": 56.9,
      "learning_rate": 3.286095365121708e-05,
      "loss": 3.4075,
      "step": 4040
    },
    {
      "epoch": 57.18,
      "learning_rate": 3.252750916972324e-05,
      "loss": 3.409,
      "step": 4060
    },
    {
      "epoch": 57.46,
      "learning_rate": 3.2194064688229413e-05,
      "loss": 3.3959,
      "step": 4080
    },
    {
      "epoch": 57.75,
      "learning_rate": 3.186062020673558e-05,
      "loss": 3.3929,
      "step": 4100
    },
    {
      "epoch": 58.03,
      "learning_rate": 3.152717572524175e-05,
      "loss": 3.4073,
      "step": 4120
    },
    {
      "epoch": 58.31,
      "learning_rate": 3.119373124374792e-05,
      "loss": 3.3896,
      "step": 4140
    },
    {
      "epoch": 58.59,
      "learning_rate": 3.0860286762254085e-05,
      "loss": 3.3928,
      "step": 4160
    },
    {
      "epoch": 58.87,
      "learning_rate": 3.052684228076025e-05,
      "loss": 3.4,
      "step": 4180
    },
    {
      "epoch": 59.15,
      "learning_rate": 3.0193397799266425e-05,
      "loss": 3.3962,
      "step": 4200
    },
    {
      "epoch": 59.15,
      "eval_validation_accuracy": 0.2753813168322206,
      "eval_validation_loss": 5.51171875,
      "eval_validation_runtime": 100.6128,
      "eval_validation_samples_per_second": 7.156,
      "eval_validation_steps_per_second": 0.447,
      "step": 4200
    },
    {
      "epoch": 59.44,
      "learning_rate": 2.9859953317772593e-05,
      "loss": 3.3818,
      "step": 4220
    },
    {
      "epoch": 59.72,
      "learning_rate": 2.952650883627876e-05,
      "loss": 3.3882,
      "step": 4240
    },
    {
      "epoch": 60.0,
      "learning_rate": 2.919306435478493e-05,
      "loss": 3.387,
      "step": 4260
    },
    {
      "epoch": 60.28,
      "learning_rate": 2.88596198732911e-05,
      "loss": 3.384,
      "step": 4280
    },
    {
      "epoch": 60.56,
      "learning_rate": 2.8526175391797265e-05,
      "loss": 3.3829,
      "step": 4300
    },
    {
      "epoch": 60.85,
      "learning_rate": 2.8192730910303433e-05,
      "loss": 3.3849,
      "step": 4320
    },
    {
      "epoch": 61.13,
      "learning_rate": 2.7859286428809604e-05,
      "loss": 3.3766,
      "step": 4340
    },
    {
      "epoch": 61.41,
      "learning_rate": 2.7525841947315772e-05,
      "loss": 3.3792,
      "step": 4360
    },
    {
      "epoch": 61.69,
      "learning_rate": 2.719239746582194e-05,
      "loss": 3.3792,
      "step": 4380
    },
    {
      "epoch": 61.97,
      "learning_rate": 2.685895298432811e-05,
      "loss": 3.3873,
      "step": 4400
    },
    {
      "epoch": 61.97,
      "eval_validation_accuracy": 0.27612902350322965,
      "eval_validation_loss": 5.5,
      "eval_validation_runtime": 100.988,
      "eval_validation_samples_per_second": 7.13,
      "eval_validation_steps_per_second": 0.446,
      "step": 4400
    },
    {
      "epoch": 62.25,
      "learning_rate": 2.652550850283428e-05,
      "loss": 3.3768,
      "step": 4420
    },
    {
      "epoch": 62.54,
      "learning_rate": 2.6192064021340444e-05,
      "loss": 3.3768,
      "step": 4440
    },
    {
      "epoch": 62.82,
      "learning_rate": 2.585861953984662e-05,
      "loss": 3.3748,
      "step": 4460
    },
    {
      "epoch": 63.1,
      "learning_rate": 2.5525175058352783e-05,
      "loss": 3.3686,
      "step": 4480
    },
    {
      "epoch": 63.38,
      "learning_rate": 2.519173057685895e-05,
      "loss": 3.3712,
      "step": 4500
    },
    {
      "epoch": 63.66,
      "learning_rate": 2.4858286095365123e-05,
      "loss": 3.3711,
      "step": 4520
    },
    {
      "epoch": 63.94,
      "learning_rate": 2.452484161387129e-05,
      "loss": 3.364,
      "step": 4540
    },
    {
      "epoch": 64.23,
      "learning_rate": 2.4191397132377462e-05,
      "loss": 3.3646,
      "step": 4560
    },
    {
      "epoch": 64.51,
      "learning_rate": 2.3857952650883627e-05,
      "loss": 3.3617,
      "step": 4580
    },
    {
      "epoch": 64.79,
      "learning_rate": 2.3524508169389798e-05,
      "loss": 3.3626,
      "step": 4600
    },
    {
      "epoch": 64.79,
      "eval_validation_accuracy": 0.2768007382076752,
      "eval_validation_loss": 5.48828125,
      "eval_validation_runtime": 99.65,
      "eval_validation_samples_per_second": 7.225,
      "eval_validation_steps_per_second": 0.452,
      "step": 4600
    },
    {
      "epoch": 65.07,
      "learning_rate": 2.3191063687895966e-05,
      "loss": 3.3644,
      "step": 4620
    },
    {
      "epoch": 65.35,
      "learning_rate": 2.2857619206402134e-05,
      "loss": 3.3611,
      "step": 4640
    },
    {
      "epoch": 65.63,
      "learning_rate": 2.2524174724908302e-05,
      "loss": 3.358,
      "step": 4660
    },
    {
      "epoch": 65.92,
      "learning_rate": 2.2190730243414473e-05,
      "loss": 3.3628,
      "step": 4680
    },
    {
      "epoch": 66.2,
      "learning_rate": 2.185728576192064e-05,
      "loss": 3.3557,
      "step": 4700
    },
    {
      "epoch": 66.48,
      "learning_rate": 2.152384128042681e-05,
      "loss": 3.3585,
      "step": 4720
    },
    {
      "epoch": 66.76,
      "learning_rate": 2.119039679893298e-05,
      "loss": 3.3613,
      "step": 4740
    },
    {
      "epoch": 67.04,
      "learning_rate": 2.0856952317439145e-05,
      "loss": 3.3578,
      "step": 4760
    },
    {
      "epoch": 67.32,
      "learning_rate": 2.0523507835945317e-05,
      "loss": 3.3487,
      "step": 4780
    },
    {
      "epoch": 67.61,
      "learning_rate": 2.0190063354451485e-05,
      "loss": 3.3489,
      "step": 4800
    },
    {
      "epoch": 67.61,
      "eval_validation_accuracy": 0.2774181729360039,
      "eval_validation_loss": 5.484375,
      "eval_validation_runtime": 103.1105,
      "eval_validation_samples_per_second": 6.983,
      "eval_validation_steps_per_second": 0.436,
      "step": 4800
    },
    {
      "epoch": 67.89,
      "learning_rate": 1.9856618872957653e-05,
      "loss": 3.3488,
      "step": 4820
    },
    {
      "epoch": 68.17,
      "learning_rate": 1.952317439146382e-05,
      "loss": 3.3569,
      "step": 4840
    },
    {
      "epoch": 68.45,
      "learning_rate": 1.9189729909969992e-05,
      "loss": 3.3489,
      "step": 4860
    },
    {
      "epoch": 68.73,
      "learning_rate": 1.885628542847616e-05,
      "loss": 3.3446,
      "step": 4880
    },
    {
      "epoch": 69.01,
      "learning_rate": 1.8522840946982328e-05,
      "loss": 3.3533,
      "step": 4900
    },
    {
      "epoch": 69.3,
      "learning_rate": 1.81893964654885e-05,
      "loss": 3.3408,
      "step": 4920
    },
    {
      "epoch": 69.58,
      "learning_rate": 1.7855951983994664e-05,
      "loss": 3.3448,
      "step": 4940
    },
    {
      "epoch": 69.86,
      "learning_rate": 1.7522507502500835e-05,
      "loss": 3.3509,
      "step": 4960
    },
    {
      "epoch": 70.14,
      "learning_rate": 1.7189063021007003e-05,
      "loss": 3.3399,
      "step": 4980
    },
    {
      "epoch": 70.42,
      "learning_rate": 1.685561853951317e-05,
      "loss": 3.3397,
      "step": 5000
    },
    {
      "epoch": 70.42,
      "eval_validation_accuracy": 0.2781387396189546,
      "eval_validation_loss": 5.47265625,
      "eval_validation_runtime": 101.8554,
      "eval_validation_samples_per_second": 7.069,
      "eval_validation_steps_per_second": 0.442,
      "step": 5000
    },
    {
      "epoch": 70.7,
      "learning_rate": 1.6555518506168722e-05,
      "loss": 3.3438,
      "step": 5020
    },
    {
      "epoch": 70.99,
      "learning_rate": 1.6222074024674894e-05,
      "loss": 3.3397,
      "step": 5040
    },
    {
      "epoch": 71.27,
      "learning_rate": 1.588862954318106e-05,
      "loss": 3.338,
      "step": 5060
    },
    {
      "epoch": 71.55,
      "learning_rate": 1.555518506168723e-05,
      "loss": 3.3334,
      "step": 5080
    },
    {
      "epoch": 71.83,
      "learning_rate": 1.52217405801934e-05,
      "loss": 3.3395,
      "step": 5100
    },
    {
      "epoch": 72.11,
      "learning_rate": 1.4888296098699567e-05,
      "loss": 3.34,
      "step": 5120
    },
    {
      "epoch": 72.39,
      "learning_rate": 1.4554851617205737e-05,
      "loss": 3.3338,
      "step": 5140
    },
    {
      "epoch": 72.68,
      "learning_rate": 1.4221407135711903e-05,
      "loss": 3.3357,
      "step": 5160
    },
    {
      "epoch": 72.96,
      "learning_rate": 1.3887962654218073e-05,
      "loss": 3.3378,
      "step": 5180
    },
    {
      "epoch": 73.24,
      "learning_rate": 1.3554518172724243e-05,
      "loss": 3.338,
      "step": 5200
    },
    {
      "epoch": 73.24,
      "eval_validation_accuracy": 0.27860486891385766,
      "eval_validation_loss": 5.46875,
      "eval_validation_runtime": 101.8164,
      "eval_validation_samples_per_second": 7.072,
      "eval_validation_steps_per_second": 0.442,
      "step": 5200
    },
    {
      "epoch": 73.52,
      "learning_rate": 1.3221073691230409e-05,
      "loss": 3.3351,
      "step": 5220
    },
    {
      "epoch": 73.8,
      "learning_rate": 1.2887629209736579e-05,
      "loss": 3.3352,
      "step": 5240
    },
    {
      "epoch": 74.08,
      "learning_rate": 1.2554184728242748e-05,
      "loss": 3.333,
      "step": 5260
    },
    {
      "epoch": 74.37,
      "learning_rate": 1.2220740246748916e-05,
      "loss": 3.3285,
      "step": 5280
    },
    {
      "epoch": 74.65,
      "learning_rate": 1.1887295765255086e-05,
      "loss": 3.3253,
      "step": 5300
    },
    {
      "epoch": 74.93,
      "learning_rate": 1.1553851283761254e-05,
      "loss": 3.3316,
      "step": 5320
    },
    {
      "epoch": 75.21,
      "learning_rate": 1.1220406802267424e-05,
      "loss": 3.3236,
      "step": 5340
    },
    {
      "epoch": 75.49,
      "learning_rate": 1.0886962320773592e-05,
      "loss": 3.3257,
      "step": 5360
    },
    {
      "epoch": 75.77,
      "learning_rate": 1.055351783927976e-05,
      "loss": 3.3299,
      "step": 5380
    },
    {
      "epoch": 76.06,
      "learning_rate": 1.022007335778593e-05,
      "loss": 3.3305,
      "step": 5400
    },
    {
      "epoch": 76.06,
      "eval_validation_accuracy": 0.2790011127395104,
      "eval_validation_loss": 5.46484375,
      "eval_validation_runtime": 100.3292,
      "eval_validation_samples_per_second": 7.176,
      "eval_validation_steps_per_second": 0.449,
      "step": 5400
    },
    {
      "epoch": 76.34,
      "learning_rate": 9.886628876292097e-06,
      "loss": 3.321,
      "step": 5420
    },
    {
      "epoch": 76.62,
      "learning_rate": 9.553184394798265e-06,
      "loss": 3.3226,
      "step": 5440
    },
    {
      "epoch": 76.9,
      "learning_rate": 9.219739913304435e-06,
      "loss": 3.3276,
      "step": 5460
    },
    {
      "epoch": 77.18,
      "learning_rate": 8.886295431810605e-06,
      "loss": 3.3269,
      "step": 5480
    },
    {
      "epoch": 77.46,
      "learning_rate": 8.552850950316773e-06,
      "loss": 3.3205,
      "step": 5500
    },
    {
      "epoch": 77.75,
      "learning_rate": 8.219406468822942e-06,
      "loss": 3.3217,
      "step": 5520
    },
    {
      "epoch": 78.03,
      "learning_rate": 7.88596198732911e-06,
      "loss": 3.3265,
      "step": 5540
    },
    {
      "epoch": 78.31,
      "learning_rate": 7.552517505835278e-06,
      "loss": 3.3204,
      "step": 5560
    },
    {
      "epoch": 78.59,
      "learning_rate": 7.219073024341448e-06,
      "loss": 3.3213,
      "step": 5580
    },
    {
      "epoch": 78.87,
      "learning_rate": 6.885628542847617e-06,
      "loss": 3.3142,
      "step": 5600
    },
    {
      "epoch": 78.87,
      "eval_validation_accuracy": 0.27932543559680834,
      "eval_validation_loss": 5.4609375,
      "eval_validation_runtime": 99.2613,
      "eval_validation_samples_per_second": 7.254,
      "eval_validation_steps_per_second": 0.453,
      "step": 5600
    },
    {
      "epoch": 79.15,
      "learning_rate": 6.552184061353785e-06,
      "loss": 3.3212,
      "step": 5620
    },
    {
      "epoch": 79.44,
      "learning_rate": 6.2187395798599535e-06,
      "loss": 3.3174,
      "step": 5640
    },
    {
      "epoch": 79.72,
      "learning_rate": 5.885295098366122e-06,
      "loss": 3.3164,
      "step": 5660
    },
    {
      "epoch": 80.0,
      "learning_rate": 5.551850616872291e-06,
      "loss": 3.3221,
      "step": 5680
    },
    {
      "epoch": 80.28,
      "learning_rate": 5.21840613537846e-06,
      "loss": 3.3189,
      "step": 5700
    },
    {
      "epoch": 80.56,
      "learning_rate": 4.884961653884629e-06,
      "loss": 3.3212,
      "step": 5720
    },
    {
      "epoch": 80.85,
      "learning_rate": 4.551517172390797e-06,
      "loss": 3.3155,
      "step": 5740
    },
    {
      "epoch": 81.13,
      "learning_rate": 4.218072690896966e-06,
      "loss": 3.3167,
      "step": 5760
    },
    {
      "epoch": 81.41,
      "learning_rate": 3.8846282094031345e-06,
      "loss": 3.3181,
      "step": 5780
    },
    {
      "epoch": 81.69,
      "learning_rate": 3.5511837279093033e-06,
      "loss": 3.3108,
      "step": 5800
    },
    {
      "epoch": 81.69,
      "eval_validation_accuracy": 0.2794719915323237,
      "eval_validation_loss": 5.45703125,
      "eval_validation_runtime": 100.6515,
      "eval_validation_samples_per_second": 7.153,
      "eval_validation_steps_per_second": 0.447,
      "step": 5800
    },
    {
      "epoch": 81.97,
      "learning_rate": 3.217739246415472e-06,
      "loss": 3.3192,
      "step": 5820
    },
    {
      "epoch": 82.25,
      "learning_rate": 2.884294764921641e-06,
      "loss": 3.3208,
      "step": 5840
    },
    {
      "epoch": 82.54,
      "learning_rate": 2.5508502834278094e-06,
      "loss": 3.3063,
      "step": 5860
    },
    {
      "epoch": 82.82,
      "learning_rate": 2.2174058019339782e-06,
      "loss": 3.3141,
      "step": 5880
    },
    {
      "epoch": 83.1,
      "learning_rate": 1.8839613204401467e-06,
      "loss": 3.3149,
      "step": 5900
    },
    {
      "epoch": 83.38,
      "learning_rate": 1.5505168389463155e-06,
      "loss": 3.3159,
      "step": 5920
    },
    {
      "epoch": 83.66,
      "learning_rate": 1.2170723574524843e-06,
      "loss": 3.3104,
      "step": 5940
    },
    {
      "epoch": 83.94,
      "learning_rate": 8.836278759586529e-07,
      "loss": 3.3139,
      "step": 5960
    },
    {
      "epoch": 84.23,
      "learning_rate": 5.501833944648216e-07,
      "loss": 3.3131,
      "step": 5980
    },
    {
      "epoch": 84.51,
      "learning_rate": 2.1673891297099036e-07,
      "loss": 3.3126,
      "step": 6000
    },
    {
      "epoch": 84.51,
      "eval_validation_accuracy": 0.27968436193888074,
      "eval_validation_loss": 5.45703125,
      "eval_validation_runtime": 101.2584,
      "eval_validation_samples_per_second": 7.111,
      "eval_validation_steps_per_second": 0.444,
      "step": 6000
    },
    {
      "epoch": 84.51,
      "step": 6000,
      "total_flos": 1.2863463988238746e+19,
      "train_loss": 3.6562127278645833,
      "train_runtime": 21626.0725,
      "train_samples_per_second": 53.269,
      "train_steps_per_second": 0.277
    }
  ],
  "max_steps": 6000,
  "num_train_epochs": 85,
  "total_flos": 1.2863463988238746e+19,
  "trial_name": null,
  "trial_params": null
}