{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9915254237288136,
  "global_step": 6001,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 9.974991663887962e-05,
      "loss": 9.7321,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 9.941647215738579e-05,
      "loss": 8.8682,
      "step": 40
    },
    {
      "epoch": 0.03,
      "learning_rate": 9.908302767589197e-05,
      "loss": 8.5062,
      "step": 60
    },
    {
      "epoch": 0.04,
      "learning_rate": 9.874958319439813e-05,
      "loss": 8.2194,
      "step": 80
    },
    {
      "epoch": 0.05,
      "learning_rate": 9.841613871290431e-05,
      "loss": 8.0283,
      "step": 100
    },
    {
      "epoch": 0.06,
      "learning_rate": 9.808269423141048e-05,
      "loss": 7.9013,
      "step": 120
    },
    {
      "epoch": 0.07,
      "learning_rate": 9.774924974991663e-05,
      "loss": 7.7487,
      "step": 140
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.741580526842281e-05,
      "loss": 7.6568,
      "step": 160
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.708236078692898e-05,
      "loss": 7.5793,
      "step": 180
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.674891630543515e-05,
      "loss": 7.4574,
      "step": 200
    },
    {
      "epoch": 0.1,
      "eval_accuracy": 0.16882355811759567,
      "eval_loss": 7.49609375,
      "eval_runtime": 975.6082,
      "eval_samples_per_second": 10.579,
      "eval_steps_per_second": 0.166,
      "step": 200
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.641547182394132e-05,
      "loss": 7.4258,
      "step": 220
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.608202734244749e-05,
      "loss": 7.3778,
      "step": 240
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.574858286095365e-05,
      "loss": 7.3281,
      "step": 260
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.541513837945982e-05,
      "loss": 7.2731,
      "step": 280
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.508169389796599e-05,
      "loss": 7.2491,
      "step": 300
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.474824941647216e-05,
      "loss": 7.1846,
      "step": 320
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.441480493497833e-05,
      "loss": 7.131,
      "step": 340
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.408136045348451e-05,
      "loss": 7.1157,
      "step": 360
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.374791597199066e-05,
      "loss": 7.0675,
      "step": 380
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.341447149049683e-05,
      "loss": 7.0445,
      "step": 400
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.19973070589428632,
      "eval_loss": 7.0546875,
      "eval_runtime": 892.7377,
      "eval_samples_per_second": 11.561,
      "eval_steps_per_second": 0.181,
      "step": 400
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.308102700900301e-05,
      "loss": 7.0021,
      "step": 420
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.274758252750917e-05,
      "loss": 6.9485,
      "step": 440
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.241413804601535e-05,
      "loss": 6.949,
      "step": 460
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.208069356452151e-05,
      "loss": 6.9056,
      "step": 480
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.174724908302767e-05,
      "loss": 6.8714,
      "step": 500
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.141380460153385e-05,
      "loss": 6.8231,
      "step": 520
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.108036012004002e-05,
      "loss": 6.8117,
      "step": 540
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.074691563854619e-05,
      "loss": 6.7348,
      "step": 560
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.041347115705235e-05,
      "loss": 6.7406,
      "step": 580
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.008002667555852e-05,
      "loss": 6.7483,
      "step": 600
    },
    {
      "epoch": 0.3,
      "eval_accuracy": 0.21896009610790168,
      "eval_loss": 6.79296875,
      "eval_runtime": 632.1671,
      "eval_samples_per_second": 16.326,
      "eval_steps_per_second": 0.256,
      "step": 600
    },
    {
      "epoch": 0.31,
      "learning_rate": 8.974658219406469e-05,
      "loss": 6.6812,
      "step": 620
    },
    {
      "epoch": 0.32,
      "learning_rate": 8.941313771257086e-05,
      "loss": 6.6846,
      "step": 640
    },
    {
      "epoch": 0.33,
      "learning_rate": 8.907969323107703e-05,
      "loss": 6.711,
      "step": 660
    },
    {
      "epoch": 0.34,
      "learning_rate": 8.87462487495832e-05,
      "loss": 6.6342,
      "step": 680
    },
    {
      "epoch": 0.35,
      "learning_rate": 8.841280426808936e-05,
      "loss": 6.6208,
      "step": 700
    },
    {
      "epoch": 0.36,
      "learning_rate": 8.807935978659554e-05,
      "loss": 6.6218,
      "step": 720
    },
    {
      "epoch": 0.37,
      "learning_rate": 8.77459153051017e-05,
      "loss": 6.5341,
      "step": 740
    },
    {
      "epoch": 0.38,
      "learning_rate": 8.741247082360787e-05,
      "loss": 6.5627,
      "step": 760
    },
    {
      "epoch": 0.39,
      "learning_rate": 8.707902634211405e-05,
      "loss": 6.4917,
      "step": 780
    },
    {
      "epoch": 0.4,
      "learning_rate": 8.67455818606202e-05,
      "loss": 6.4568,
      "step": 800
    },
    {
      "epoch": 0.4,
      "eval_accuracy": 0.23763536355011933,
      "eval_loss": 6.5703125,
      "eval_runtime": 909.8761,
      "eval_samples_per_second": 11.343,
      "eval_steps_per_second": 0.178,
      "step": 800
    },
    {
      "epoch": 0.41,
      "learning_rate": 8.641213737912638e-05,
      "loss": 6.4541,
      "step": 820
    },
    {
      "epoch": 0.42,
      "learning_rate": 8.607869289763255e-05,
      "loss": 6.4661,
      "step": 840
    },
    {
      "epoch": 0.43,
      "learning_rate": 8.57452484161387e-05,
      "loss": 6.445,
      "step": 860
    },
    {
      "epoch": 0.44,
      "learning_rate": 8.541180393464489e-05,
      "loss": 6.4083,
      "step": 880
    },
    {
      "epoch": 0.45,
      "learning_rate": 8.507835945315106e-05,
      "loss": 6.4168,
      "step": 900
    },
    {
      "epoch": 0.46,
      "learning_rate": 8.474491497165722e-05,
      "loss": 6.3794,
      "step": 920
    },
    {
      "epoch": 0.47,
      "learning_rate": 8.441147049016339e-05,
      "loss": 6.3398,
      "step": 940
    },
    {
      "epoch": 0.48,
      "learning_rate": 8.407802600866956e-05,
      "loss": 6.3377,
      "step": 960
    },
    {
      "epoch": 0.49,
      "learning_rate": 8.374458152717573e-05,
      "loss": 6.3086,
      "step": 980
    },
    {
      "epoch": 0.5,
      "learning_rate": 8.34111370456819e-05,
      "loss": 6.2865,
      "step": 1000
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.25518331041094305,
      "eval_loss": 6.375,
      "eval_runtime": 974.2435,
      "eval_samples_per_second": 10.594,
      "eval_steps_per_second": 0.166,
      "step": 1000
    },
    {
      "epoch": 0.51,
      "learning_rate": 8.311103701233745e-05,
      "loss": 6.3177,
      "step": 1020
    },
    {
      "epoch": 0.52,
      "learning_rate": 8.277759253084362e-05,
      "loss": 6.2473,
      "step": 1040
    },
    {
      "epoch": 0.53,
      "learning_rate": 8.244414804934978e-05,
      "loss": 6.2194,
      "step": 1060
    },
    {
      "epoch": 0.54,
      "learning_rate": 8.211070356785595e-05,
      "loss": 6.2438,
      "step": 1080
    },
    {
      "epoch": 0.55,
      "learning_rate": 8.177725908636212e-05,
      "loss": 6.2338,
      "step": 1100
    },
    {
      "epoch": 0.56,
      "learning_rate": 8.14438146048683e-05,
      "loss": 6.1888,
      "step": 1120
    },
    {
      "epoch": 0.57,
      "learning_rate": 8.111037012337445e-05,
      "loss": 6.1476,
      "step": 1140
    },
    {
      "epoch": 0.58,
      "learning_rate": 8.077692564188064e-05,
      "loss": 6.1139,
      "step": 1160
    },
    {
      "epoch": 0.59,
      "learning_rate": 8.04434811603868e-05,
      "loss": 6.1182,
      "step": 1180
    },
    {
      "epoch": 0.6,
      "learning_rate": 8.011003667889296e-05,
      "loss": 6.1028,
      "step": 1200
    },
    {
      "epoch": 0.6,
      "eval_accuracy": 0.27931541153780454,
      "eval_loss": 6.1484375,
      "eval_runtime": 865.8054,
      "eval_samples_per_second": 11.921,
      "eval_steps_per_second": 0.187,
      "step": 1200
    },
    {
      "epoch": 0.61,
      "learning_rate": 7.977659219739914e-05,
      "loss": 6.0742,
      "step": 1220
    },
    {
      "epoch": 0.62,
      "learning_rate": 7.944314771590531e-05,
      "loss": 6.0706,
      "step": 1240
    },
    {
      "epoch": 0.63,
      "learning_rate": 7.910970323441148e-05,
      "loss": 6.0479,
      "step": 1260
    },
    {
      "epoch": 0.64,
      "learning_rate": 7.877625875291764e-05,
      "loss": 5.9765,
      "step": 1280
    },
    {
      "epoch": 0.65,
      "learning_rate": 7.844281427142381e-05,
      "loss": 5.9953,
      "step": 1300
    },
    {
      "epoch": 0.66,
      "learning_rate": 7.810936978992998e-05,
      "loss": 6.0057,
      "step": 1320
    },
    {
      "epoch": 0.67,
      "learning_rate": 7.777592530843615e-05,
      "loss": 6.0033,
      "step": 1340
    },
    {
      "epoch": 0.68,
      "learning_rate": 7.744248082694232e-05,
      "loss": 5.9667,
      "step": 1360
    },
    {
      "epoch": 0.69,
      "learning_rate": 7.710903634544848e-05,
      "loss": 5.9367,
      "step": 1380
    },
    {
      "epoch": 0.7,
      "learning_rate": 7.677559186395465e-05,
      "loss": 5.8888,
      "step": 1400
    },
    {
      "epoch": 0.7,
      "eval_accuracy": 0.29820689430587377,
      "eval_loss": 5.95703125,
      "eval_runtime": 778.5891,
      "eval_samples_per_second": 13.256,
      "eval_steps_per_second": 0.208,
      "step": 1400
    },
    {
      "epoch": 0.71,
      "learning_rate": 7.644214738246082e-05,
      "loss": 5.8829,
      "step": 1420
    },
    {
      "epoch": 0.72,
      "learning_rate": 7.610870290096699e-05,
      "loss": 5.8627,
      "step": 1440
    },
    {
      "epoch": 0.73,
      "learning_rate": 7.577525841947316e-05,
      "loss": 5.8251,
      "step": 1460
    },
    {
      "epoch": 0.74,
      "learning_rate": 7.544181393797934e-05,
      "loss": 5.8655,
      "step": 1480
    },
    {
      "epoch": 0.75,
      "learning_rate": 7.510836945648549e-05,
      "loss": 5.8127,
      "step": 1500
    },
    {
      "epoch": 0.76,
      "learning_rate": 7.477492497499166e-05,
      "loss": 5.8305,
      "step": 1520
    },
    {
      "epoch": 0.77,
      "learning_rate": 7.444148049349784e-05,
      "loss": 5.8159,
      "step": 1540
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.4108036012004e-05,
      "loss": 5.8354,
      "step": 1560
    },
    {
      "epoch": 0.79,
      "learning_rate": 7.377459153051018e-05,
      "loss": 5.7823,
      "step": 1580
    },
    {
      "epoch": 0.8,
      "learning_rate": 7.344114704901635e-05,
      "loss": 5.7362,
      "step": 1600
    },
    {
      "epoch": 0.8,
      "eval_accuracy": 0.3121272705508336,
      "eval_loss": 5.80078125,
      "eval_runtime": 525.7065,
      "eval_samples_per_second": 19.633,
      "eval_steps_per_second": 0.308,
      "step": 1600
    },
    {
      "epoch": 0.81,
      "learning_rate": 7.310770256752251e-05,
      "loss": 5.7201,
      "step": 1620
    },
    {
      "epoch": 0.82,
      "learning_rate": 7.277425808602868e-05,
      "loss": 5.7253,
      "step": 1640
    },
    {
      "epoch": 0.83,
      "learning_rate": 7.244081360453485e-05,
      "loss": 5.7258,
      "step": 1660
    },
    {
      "epoch": 0.84,
      "learning_rate": 7.210736912304102e-05,
      "loss": 5.6782,
      "step": 1680
    },
    {
      "epoch": 0.85,
      "learning_rate": 7.177392464154719e-05,
      "loss": 5.7047,
      "step": 1700
    },
    {
      "epoch": 0.86,
      "learning_rate": 7.144048016005335e-05,
      "loss": 5.6893,
      "step": 1720
    },
    {
      "epoch": 0.87,
      "learning_rate": 7.110703567855952e-05,
      "loss": 5.6119,
      "step": 1740
    },
    {
      "epoch": 0.88,
      "learning_rate": 7.077359119706569e-05,
      "loss": 5.6852,
      "step": 1760
    },
    {
      "epoch": 0.89,
      "learning_rate": 7.044014671557186e-05,
      "loss": 5.6021,
      "step": 1780
    },
    {
      "epoch": 0.9,
      "learning_rate": 7.010670223407803e-05,
      "loss": 5.6507,
      "step": 1800
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.32381624594057856,
      "eval_loss": 5.6796875,
      "eval_runtime": 634.7933,
      "eval_samples_per_second": 16.259,
      "eval_steps_per_second": 0.255,
      "step": 1800
    },
    {
      "epoch": 0.91,
      "learning_rate": 6.97732577525842e-05,
      "loss": 5.6308,
      "step": 1820
    },
    {
      "epoch": 0.92,
      "learning_rate": 6.943981327109038e-05,
      "loss": 5.6368,
      "step": 1840
    },
    {
      "epoch": 0.93,
      "learning_rate": 6.910636878959653e-05,
      "loss": 5.6018,
      "step": 1860
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.87729243081027e-05,
      "loss": 5.5617,
      "step": 1880
    },
    {
      "epoch": 0.95,
      "learning_rate": 6.843947982660888e-05,
      "loss": 5.5805,
      "step": 1900
    },
    {
      "epoch": 0.96,
      "learning_rate": 6.810603534511503e-05,
      "loss": 5.5503,
      "step": 1920
    },
    {
      "epoch": 0.97,
      "learning_rate": 6.777259086362121e-05,
      "loss": 5.555,
      "step": 1940
    },
    {
      "epoch": 0.98,
      "learning_rate": 6.743914638212738e-05,
      "loss": 5.5774,
      "step": 1960
    },
    {
      "epoch": 0.99,
      "learning_rate": 6.710570190063354e-05,
      "loss": 5.5667,
      "step": 1980
    },
    {
      "epoch": 1.0,
      "learning_rate": 6.677225741913972e-05,
      "loss": 5.565,
      "step": 2000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.33180064057103564,
      "eval_loss": 5.578125,
      "eval_runtime": 846.4215,
      "eval_samples_per_second": 12.194,
      "eval_steps_per_second": 0.191,
      "step": 2000
    },
    {
      "epoch": 1.01,
      "learning_rate": 6.647215738579527e-05,
      "loss": 5.5107,
      "step": 2020
    },
    {
      "epoch": 1.02,
      "learning_rate": 6.613871290430144e-05,
      "loss": 5.5002,
      "step": 2040
    },
    {
      "epoch": 1.03,
      "learning_rate": 6.58052684228076e-05,
      "loss": 5.4849,
      "step": 2060
    },
    {
      "epoch": 1.04,
      "learning_rate": 6.547182394131377e-05,
      "loss": 5.5153,
      "step": 2080
    },
    {
      "epoch": 1.05,
      "learning_rate": 6.513837945981994e-05,
      "loss": 5.4517,
      "step": 2100
    },
    {
      "epoch": 1.06,
      "learning_rate": 6.480493497832611e-05,
      "loss": 5.4576,
      "step": 2120
    },
    {
      "epoch": 1.07,
      "learning_rate": 6.447149049683228e-05,
      "loss": 5.454,
      "step": 2140
    },
    {
      "epoch": 1.08,
      "learning_rate": 6.413804601533845e-05,
      "loss": 5.432,
      "step": 2160
    },
    {
      "epoch": 1.09,
      "learning_rate": 6.380460153384463e-05,
      "loss": 5.4429,
      "step": 2180
    },
    {
      "epoch": 1.1,
      "learning_rate": 6.347115705235078e-05,
      "loss": 5.4688,
      "step": 2200
    },
    {
      "epoch": 1.1,
      "eval_accuracy": 0.33922944511943365,
      "eval_loss": 5.49609375,
      "eval_runtime": 969.4851,
      "eval_samples_per_second": 10.646,
      "eval_steps_per_second": 0.167,
      "step": 2200
    },
    {
      "epoch": 1.11,
      "learning_rate": 6.313771257085695e-05,
      "loss": 5.4398,
      "step": 2220
    },
    {
      "epoch": 1.12,
      "learning_rate": 6.280426808936313e-05,
      "loss": 5.3972,
      "step": 2240
    },
    {
      "epoch": 1.13,
      "learning_rate": 6.247082360786929e-05,
      "loss": 5.4121,
      "step": 2260
    },
    {
      "epoch": 1.14,
      "learning_rate": 6.213737912637547e-05,
      "loss": 5.4003,
      "step": 2280
    },
    {
      "epoch": 1.15,
      "learning_rate": 6.180393464488164e-05,
      "loss": 5.3916,
      "step": 2300
    },
    {
      "epoch": 1.16,
      "learning_rate": 6.147049016338779e-05,
      "loss": 5.3822,
      "step": 2320
    },
    {
      "epoch": 1.17,
      "learning_rate": 6.113704568189397e-05,
      "loss": 5.419,
      "step": 2340
    },
    {
      "epoch": 1.18,
      "learning_rate": 6.080360120040014e-05,
      "loss": 5.3861,
      "step": 2360
    },
    {
      "epoch": 1.19,
      "learning_rate": 6.04701567189063e-05,
      "loss": 5.4055,
      "step": 2380
    },
    {
      "epoch": 1.2,
      "learning_rate": 6.0136712237412475e-05,
      "loss": 5.4044,
      "step": 2400
    },
    {
      "epoch": 1.2,
      "eval_accuracy": 0.3456189203935214,
      "eval_loss": 5.421875,
      "eval_runtime": 894.657,
      "eval_samples_per_second": 11.536,
      "eval_steps_per_second": 0.181,
      "step": 2400
    },
    {
      "epoch": 1.21,
      "learning_rate": 5.980326775591864e-05,
      "loss": 5.3611,
      "step": 2420
    },
    {
      "epoch": 1.22,
      "learning_rate": 5.9469823274424805e-05,
      "loss": 5.2802,
      "step": 2440
    },
    {
      "epoch": 1.23,
      "learning_rate": 5.913637879293098e-05,
      "loss": 5.3635,
      "step": 2460
    },
    {
      "epoch": 1.24,
      "learning_rate": 5.8802934311437154e-05,
      "loss": 5.3047,
      "step": 2480
    },
    {
      "epoch": 1.25,
      "learning_rate": 5.8469489829943315e-05,
      "loss": 5.3242,
      "step": 2500
    },
    {
      "epoch": 1.26,
      "learning_rate": 5.813604534844948e-05,
      "loss": 5.3219,
      "step": 2520
    },
    {
      "epoch": 1.27,
      "learning_rate": 5.780260086695566e-05,
      "loss": 5.3164,
      "step": 2540
    },
    {
      "epoch": 1.28,
      "learning_rate": 5.746915638546182e-05,
      "loss": 5.3344,
      "step": 2560
    },
    {
      "epoch": 1.29,
      "learning_rate": 5.7135711903967994e-05,
      "loss": 5.2985,
      "step": 2580
    },
    {
      "epoch": 1.3,
      "learning_rate": 5.680226742247416e-05,
      "loss": 5.3323,
      "step": 2600
    },
    {
      "epoch": 1.3,
      "eval_accuracy": 0.3515776689895772,
      "eval_loss": 5.359375,
      "eval_runtime": 915.4215,
      "eval_samples_per_second": 11.275,
      "eval_steps_per_second": 0.177,
      "step": 2600
    },
    {
      "epoch": 1.31,
      "learning_rate": 5.646882294098032e-05,
      "loss": 5.3498,
      "step": 2620
    },
    {
      "epoch": 1.32,
      "learning_rate": 5.61353784594865e-05,
      "loss": 5.3374,
      "step": 2640
    },
    {
      "epoch": 1.33,
      "learning_rate": 5.580193397799267e-05,
      "loss": 5.258,
      "step": 2660
    },
    {
      "epoch": 1.34,
      "learning_rate": 5.5468489496498834e-05,
      "loss": 5.3057,
      "step": 2680
    },
    {
      "epoch": 1.35,
      "learning_rate": 5.5135045015005e-05,
      "loss": 5.3096,
      "step": 2700
    },
    {
      "epoch": 1.36,
      "learning_rate": 5.480160053351118e-05,
      "loss": 5.3103,
      "step": 2720
    },
    {
      "epoch": 1.37,
      "learning_rate": 5.446815605201734e-05,
      "loss": 5.294,
      "step": 2740
    },
    {
      "epoch": 1.38,
      "learning_rate": 5.413471157052351e-05,
      "loss": 5.2703,
      "step": 2760
    },
    {
      "epoch": 1.39,
      "learning_rate": 5.380126708902968e-05,
      "loss": 5.2838,
      "step": 2780
    },
    {
      "epoch": 1.4,
      "learning_rate": 5.346782260753584e-05,
      "loss": 5.2598,
      "step": 2800
    },
    {
      "epoch": 1.4,
      "eval_accuracy": 0.3561769305915198,
      "eval_loss": 5.3046875,
      "eval_runtime": 850.0259,
      "eval_samples_per_second": 12.142,
      "eval_steps_per_second": 0.191,
      "step": 2800
    },
    {
      "epoch": 1.41,
      "learning_rate": 5.3134378126042017e-05,
      "loss": 5.2614,
      "step": 2820
    },
    {
      "epoch": 1.42,
      "learning_rate": 5.280093364454819e-05,
      "loss": 5.2358,
      "step": 2840
    },
    {
      "epoch": 1.43,
      "learning_rate": 5.246748916305435e-05,
      "loss": 5.2465,
      "step": 2860
    },
    {
      "epoch": 1.44,
      "learning_rate": 5.213404468156052e-05,
      "loss": 5.2342,
      "step": 2880
    },
    {
      "epoch": 1.45,
      "learning_rate": 5.1800600200066695e-05,
      "loss": 5.2434,
      "step": 2900
    },
    {
      "epoch": 1.46,
      "learning_rate": 5.1467155718572856e-05,
      "loss": 5.2131,
      "step": 2920
    },
    {
      "epoch": 1.47,
      "learning_rate": 5.113371123707903e-05,
      "loss": 5.2167,
      "step": 2940
    },
    {
      "epoch": 1.48,
      "learning_rate": 5.08002667555852e-05,
      "loss": 5.1727,
      "step": 2960
    },
    {
      "epoch": 1.49,
      "learning_rate": 5.046682227409136e-05,
      "loss": 5.2107,
      "step": 2980
    },
    {
      "epoch": 1.5,
      "learning_rate": 5.0133377792597535e-05,
      "loss": 5.2159,
      "step": 3000
    },
    {
      "epoch": 1.5,
      "eval_accuracy": 0.3596428603716168,
      "eval_loss": 5.2578125,
      "eval_runtime": 533.7663,
      "eval_samples_per_second": 19.336,
      "eval_steps_per_second": 0.304,
      "step": 3000
    },
    {
      "epoch": 1.51,
      "learning_rate": 4.9833277759253086e-05,
      "loss": 5.2447,
      "step": 3020
    },
    {
      "epoch": 1.52,
      "learning_rate": 4.9499833277759254e-05,
      "loss": 5.2568,
      "step": 3040
    },
    {
      "epoch": 1.53,
      "learning_rate": 4.916638879626542e-05,
      "loss": 5.1794,
      "step": 3060
    },
    {
      "epoch": 1.54,
      "learning_rate": 4.88329443147716e-05,
      "loss": 5.1844,
      "step": 3080
    },
    {
      "epoch": 1.55,
      "learning_rate": 4.849949983327776e-05,
      "loss": 5.2218,
      "step": 3100
    },
    {
      "epoch": 1.56,
      "learning_rate": 4.8166055351783926e-05,
      "loss": 5.2347,
      "step": 3120
    },
    {
      "epoch": 1.57,
      "learning_rate": 4.78326108702901e-05,
      "loss": 5.1653,
      "step": 3140
    },
    {
      "epoch": 1.58,
      "learning_rate": 4.749916638879627e-05,
      "loss": 5.2375,
      "step": 3160
    },
    {
      "epoch": 1.59,
      "learning_rate": 4.716572190730244e-05,
      "loss": 5.19,
      "step": 3180
    },
    {
      "epoch": 1.6,
      "learning_rate": 4.6832277425808605e-05,
      "loss": 5.1992,
      "step": 3200
    },
    {
      "epoch": 1.6,
      "eval_accuracy": 0.36381432991685037,
      "eval_loss": 5.21484375,
      "eval_runtime": 1001.3018,
      "eval_samples_per_second": 10.308,
      "eval_steps_per_second": 0.162,
      "step": 3200
    },
    {
      "epoch": 1.61,
      "learning_rate": 4.649883294431477e-05,
      "loss": 5.2072,
      "step": 3220
    },
    {
      "epoch": 1.62,
      "learning_rate": 4.616538846282094e-05,
      "loss": 5.1528,
      "step": 3240
    },
    {
      "epoch": 1.63,
      "learning_rate": 4.5831943981327116e-05,
      "loss": 5.1562,
      "step": 3260
    },
    {
      "epoch": 1.64,
      "learning_rate": 4.549849949983328e-05,
      "loss": 5.1613,
      "step": 3280
    },
    {
      "epoch": 1.65,
      "learning_rate": 4.5165055018339445e-05,
      "loss": 5.1552,
      "step": 3300
    },
    {
      "epoch": 1.66,
      "learning_rate": 4.483161053684562e-05,
      "loss": 5.1224,
      "step": 3320
    },
    {
      "epoch": 1.67,
      "learning_rate": 4.449816605535179e-05,
      "loss": 5.1118,
      "step": 3340
    },
    {
      "epoch": 1.67,
      "learning_rate": 4.4164721573857956e-05,
      "loss": 5.138,
      "step": 3360
    },
    {
      "epoch": 1.68,
      "learning_rate": 4.3831277092364124e-05,
      "loss": 5.1441,
      "step": 3380
    },
    {
      "epoch": 1.69,
      "learning_rate": 4.349783261087029e-05,
      "loss": 5.1429,
      "step": 3400
    },
    {
      "epoch": 1.69,
      "eval_accuracy": 0.36717262536004136,
      "eval_loss": 5.1796875,
      "eval_runtime": 930.4199,
      "eval_samples_per_second": 11.093,
      "eval_steps_per_second": 0.174,
      "step": 3400
    },
    {
      "epoch": 1.7,
      "learning_rate": 4.316438812937646e-05,
      "loss": 5.1378,
      "step": 3420
    },
    {
      "epoch": 1.71,
      "learning_rate": 4.2830943647882634e-05,
      "loss": 5.0948,
      "step": 3440
    },
    {
      "epoch": 1.72,
      "learning_rate": 4.2497499166388796e-05,
      "loss": 5.108,
      "step": 3460
    },
    {
      "epoch": 1.73,
      "learning_rate": 4.2164054684894964e-05,
      "loss": 5.0677,
      "step": 3480
    },
    {
      "epoch": 1.74,
      "learning_rate": 4.183061020340114e-05,
      "loss": 5.1182,
      "step": 3500
    },
    {
      "epoch": 1.75,
      "learning_rate": 4.1497165721907306e-05,
      "loss": 5.1199,
      "step": 3520
    },
    {
      "epoch": 1.76,
      "learning_rate": 4.1163721240413474e-05,
      "loss": 5.1359,
      "step": 3540
    },
    {
      "epoch": 1.77,
      "learning_rate": 4.083027675891964e-05,
      "loss": 5.1278,
      "step": 3560
    },
    {
      "epoch": 1.78,
      "learning_rate": 4.049683227742581e-05,
      "loss": 5.1227,
      "step": 3580
    },
    {
      "epoch": 1.79,
      "learning_rate": 4.016338779593198e-05,
      "loss": 5.095,
      "step": 3600
    },
    {
      "epoch": 1.79,
      "eval_accuracy": 0.3696338259978766,
      "eval_loss": 5.14453125,
      "eval_runtime": 540.5754,
      "eval_samples_per_second": 19.093,
      "eval_steps_per_second": 0.3,
      "step": 3600
    },
    {
      "epoch": 1.8,
      "learning_rate": 3.982994331443815e-05,
      "loss": 5.1139,
      "step": 3620
    },
    {
      "epoch": 1.81,
      "learning_rate": 3.9496498832944314e-05,
      "loss": 5.0827,
      "step": 3640
    },
    {
      "epoch": 1.82,
      "learning_rate": 3.916305435145048e-05,
      "loss": 5.0918,
      "step": 3660
    },
    {
      "epoch": 1.83,
      "learning_rate": 3.882960986995666e-05,
      "loss": 5.0927,
      "step": 3680
    },
    {
      "epoch": 1.84,
      "learning_rate": 3.8496165388462825e-05,
      "loss": 5.0904,
      "step": 3700
    },
    {
      "epoch": 1.85,
      "learning_rate": 3.816272090696899e-05,
      "loss": 5.0245,
      "step": 3720
    },
    {
      "epoch": 1.86,
      "learning_rate": 3.7829276425475154e-05,
      "loss": 5.1174,
      "step": 3740
    },
    {
      "epoch": 1.87,
      "learning_rate": 3.749583194398133e-05,
      "loss": 5.1325,
      "step": 3760
    },
    {
      "epoch": 1.88,
      "learning_rate": 3.71623874624875e-05,
      "loss": 5.0408,
      "step": 3780
    },
    {
      "epoch": 1.89,
      "learning_rate": 3.6828942980993665e-05,
      "loss": 5.0646,
      "step": 3800
    },
    {
      "epoch": 1.89,
      "eval_accuracy": 0.37147085161338145,
      "eval_loss": 5.1171875,
      "eval_runtime": 1017.1553,
      "eval_samples_per_second": 10.147,
      "eval_steps_per_second": 0.159,
      "step": 3800
    },
    {
      "epoch": 1.9,
      "learning_rate": 3.649549849949983e-05,
      "loss": 5.0465,
      "step": 3820
    },
    {
      "epoch": 1.91,
      "learning_rate": 3.6162054018006e-05,
      "loss": 5.1008,
      "step": 3840
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.582860953651217e-05,
      "loss": 5.0641,
      "step": 3860
    },
    {
      "epoch": 1.93,
      "learning_rate": 3.5495165055018344e-05,
      "loss": 5.085,
      "step": 3880
    },
    {
      "epoch": 1.94,
      "learning_rate": 3.516172057352451e-05,
      "loss": 5.0498,
      "step": 3900
    },
    {
      "epoch": 1.95,
      "learning_rate": 3.482827609203067e-05,
      "loss": 5.1401,
      "step": 3920
    },
    {
      "epoch": 1.96,
      "learning_rate": 3.449483161053685e-05,
      "loss": 5.0446,
      "step": 3940
    },
    {
      "epoch": 1.97,
      "learning_rate": 3.4161387129043016e-05,
      "loss": 5.0538,
      "step": 3960
    },
    {
      "epoch": 1.98,
      "learning_rate": 3.3827942647549183e-05,
      "loss": 5.0641,
      "step": 3980
    },
    {
      "epoch": 1.99,
      "learning_rate": 3.349449816605535e-05,
      "loss": 5.059,
      "step": 4000
    },
    {
      "epoch": 1.99,
      "eval_accuracy": 0.37421836716060286,
      "eval_loss": 5.0859375,
      "eval_runtime": 943.8518,
      "eval_samples_per_second": 10.935,
      "eval_steps_per_second": 0.172,
      "step": 4000
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.319439813271091e-05,
      "loss": 5.0309,
      "step": 4020
    },
    {
      "epoch": 2.01,
      "learning_rate": 3.286095365121708e-05,
      "loss": 5.0337,
      "step": 4040
    },
    {
      "epoch": 2.02,
      "learning_rate": 3.252750916972324e-05,
      "loss": 5.0333,
      "step": 4060
    },
    {
      "epoch": 2.03,
      "learning_rate": 3.2194064688229413e-05,
      "loss": 4.981,
      "step": 4080
    },
    {
      "epoch": 2.04,
      "learning_rate": 3.186062020673558e-05,
      "loss": 5.0187,
      "step": 4100
    },
    {
      "epoch": 2.05,
      "learning_rate": 3.152717572524175e-05,
      "loss": 4.9909,
      "step": 4120
    },
    {
      "epoch": 2.06,
      "learning_rate": 3.119373124374792e-05,
      "loss": 4.9913,
      "step": 4140
    },
    {
      "epoch": 2.07,
      "learning_rate": 3.0860286762254085e-05,
      "loss": 4.9999,
      "step": 4160
    },
    {
      "epoch": 2.08,
      "learning_rate": 3.052684228076025e-05,
      "loss": 5.0122,
      "step": 4180
    },
    {
      "epoch": 2.09,
      "learning_rate": 3.0193397799266425e-05,
      "loss": 5.0152,
      "step": 4200
    },
    {
      "epoch": 2.09,
      "eval_accuracy": 0.375609093672024,
      "eval_loss": 5.06640625,
      "eval_runtime": 893.7828,
      "eval_samples_per_second": 11.548,
      "eval_steps_per_second": 0.181,
      "step": 4200
    },
    {
      "epoch": 2.1,
      "learning_rate": 2.9859953317772593e-05,
      "loss": 4.9893,
      "step": 4220
    },
    {
      "epoch": 2.11,
      "learning_rate": 2.952650883627876e-05,
      "loss": 4.9974,
      "step": 4240
    },
    {
      "epoch": 2.12,
      "learning_rate": 2.919306435478493e-05,
      "loss": 5.0072,
      "step": 4260
    },
    {
      "epoch": 2.13,
      "learning_rate": 2.88596198732911e-05,
      "loss": 4.9678,
      "step": 4280
    },
    {
      "epoch": 2.14,
      "learning_rate": 2.8526175391797265e-05,
      "loss": 4.9532,
      "step": 4300
    },
    {
      "epoch": 2.15,
      "learning_rate": 2.8192730910303433e-05,
      "loss": 4.9994,
      "step": 4320
    },
    {
      "epoch": 2.16,
      "learning_rate": 2.7859286428809604e-05,
      "loss": 4.9984,
      "step": 4340
    },
    {
      "epoch": 2.17,
      "learning_rate": 2.7525841947315772e-05,
      "loss": 4.9659,
      "step": 4360
    },
    {
      "epoch": 2.18,
      "learning_rate": 2.719239746582194e-05,
      "loss": 4.9355,
      "step": 4380
    },
    {
      "epoch": 2.19,
      "learning_rate": 2.685895298432811e-05,
      "loss": 5.0251,
      "step": 4400
    },
    {
      "epoch": 2.19,
      "eval_accuracy": 0.376910266900496,
      "eval_loss": 5.046875,
      "eval_runtime": 856.9216,
      "eval_samples_per_second": 12.044,
      "eval_steps_per_second": 0.189,
      "step": 4400
    },
    {
      "epoch": 2.2,
      "learning_rate": 2.652550850283428e-05,
      "loss": 5.0034,
      "step": 4420
    },
    {
      "epoch": 2.21,
      "learning_rate": 2.6192064021340444e-05,
      "loss": 4.9666,
      "step": 4440
    },
    {
      "epoch": 2.22,
      "learning_rate": 2.585861953984662e-05,
      "loss": 4.9759,
      "step": 4460
    },
    {
      "epoch": 2.23,
      "learning_rate": 2.5525175058352783e-05,
      "loss": 5.0003,
      "step": 4480
    },
    {
      "epoch": 2.24,
      "learning_rate": 2.519173057685895e-05,
      "loss": 4.9566,
      "step": 4500
    },
    {
      "epoch": 2.25,
      "learning_rate": 2.4858286095365123e-05,
      "loss": 4.9922,
      "step": 4520
    },
    {
      "epoch": 2.26,
      "learning_rate": 2.452484161387129e-05,
      "loss": 4.9977,
      "step": 4540
    },
    {
      "epoch": 2.27,
      "learning_rate": 2.4191397132377462e-05,
      "loss": 4.978,
      "step": 4560
    },
    {
      "epoch": 2.28,
      "learning_rate": 2.3857952650883627e-05,
      "loss": 4.9841,
      "step": 4580
    },
    {
      "epoch": 2.29,
      "learning_rate": 2.3524508169389798e-05,
      "loss": 5.022,
      "step": 4600
    },
    {
      "epoch": 2.29,
      "eval_accuracy": 0.37901850832535505,
      "eval_loss": 5.02734375,
      "eval_runtime": 895.8024,
      "eval_samples_per_second": 11.522,
      "eval_steps_per_second": 0.181,
      "step": 4600
    },
    {
      "epoch": 2.3,
      "learning_rate": 2.3191063687895966e-05,
      "loss": 4.965,
      "step": 4620
    },
    {
      "epoch": 2.31,
      "learning_rate": 2.2857619206402134e-05,
      "loss": 4.9553,
      "step": 4640
    },
    {
      "epoch": 2.32,
      "learning_rate": 2.2524174724908302e-05,
      "loss": 4.9534,
      "step": 4660
    },
    {
      "epoch": 2.33,
      "learning_rate": 2.2190730243414473e-05,
      "loss": 4.9518,
      "step": 4680
    },
    {
      "epoch": 2.34,
      "learning_rate": 2.185728576192064e-05,
      "loss": 4.9339,
      "step": 4700
    },
    {
      "epoch": 2.35,
      "learning_rate": 2.152384128042681e-05,
      "loss": 4.9811,
      "step": 4720
    },
    {
      "epoch": 2.36,
      "learning_rate": 2.119039679893298e-05,
      "loss": 4.9615,
      "step": 4740
    },
    {
      "epoch": 2.37,
      "learning_rate": 2.0856952317439145e-05,
      "loss": 4.9562,
      "step": 4760
    },
    {
      "epoch": 2.38,
      "learning_rate": 2.0523507835945317e-05,
      "loss": 4.9349,
      "step": 4780
    },
    {
      "epoch": 2.39,
      "learning_rate": 2.0190063354451485e-05,
      "loss": 4.9939,
      "step": 4800
    },
    {
      "epoch": 2.39,
      "eval_accuracy": 0.37982254723521514,
      "eval_loss": 5.015625,
      "eval_runtime": 530.0593,
      "eval_samples_per_second": 19.471,
      "eval_steps_per_second": 0.306,
      "step": 4800
    },
    {
      "epoch": 2.4,
      "learning_rate": 1.9856618872957653e-05,
      "loss": 4.9611,
      "step": 4820
    },
    {
      "epoch": 2.41,
      "learning_rate": 1.952317439146382e-05,
      "loss": 4.914,
      "step": 4840
    },
    {
      "epoch": 2.42,
      "learning_rate": 1.9189729909969992e-05,
      "loss": 4.9286,
      "step": 4860
    },
    {
      "epoch": 2.43,
      "learning_rate": 1.885628542847616e-05,
      "loss": 4.9763,
      "step": 4880
    },
    {
      "epoch": 2.44,
      "learning_rate": 1.8522840946982328e-05,
      "loss": 4.947,
      "step": 4900
    },
    {
      "epoch": 2.45,
      "learning_rate": 1.81893964654885e-05,
      "loss": 4.8647,
      "step": 4920
    },
    {
      "epoch": 2.46,
      "learning_rate": 1.7855951983994664e-05,
      "loss": 4.9485,
      "step": 4940
    },
    {
      "epoch": 2.47,
      "learning_rate": 1.7522507502500835e-05,
      "loss": 4.914,
      "step": 4960
    },
    {
      "epoch": 2.48,
      "learning_rate": 1.7189063021007003e-05,
      "loss": 4.9502,
      "step": 4980
    },
    {
      "epoch": 2.49,
      "learning_rate": 1.685561853951317e-05,
      "loss": 4.924,
      "step": 5000
    },
    {
      "epoch": 2.49,
      "eval_accuracy": 0.381116573240788,
      "eval_loss": 5.0,
      "eval_runtime": 626.5478,
      "eval_samples_per_second": 16.473,
      "eval_steps_per_second": 0.259,
      "step": 5000
    },
    {
      "epoch": 2.5,
      "learning_rate": 1.6555518506168722e-05,
      "loss": 4.9324,
      "step": 5020
    },
    {
      "epoch": 2.51,
      "learning_rate": 1.6222074024674894e-05,
      "loss": 4.9441,
      "step": 5040
    },
    {
      "epoch": 2.52,
      "learning_rate": 1.588862954318106e-05,
      "loss": 4.9695,
      "step": 5060
    },
    {
      "epoch": 2.53,
      "learning_rate": 1.555518506168723e-05,
      "loss": 4.9103,
      "step": 5080
    },
    {
      "epoch": 2.54,
      "learning_rate": 1.52217405801934e-05,
      "loss": 4.9284,
      "step": 5100
    },
    {
      "epoch": 2.55,
      "learning_rate": 1.4888296098699567e-05,
      "loss": 4.9198,
      "step": 5120
    },
    {
      "epoch": 2.56,
      "learning_rate": 1.4554851617205737e-05,
      "loss": 4.9303,
      "step": 5140
    },
    {
      "epoch": 2.57,
      "learning_rate": 1.4221407135711903e-05,
      "loss": 4.8952,
      "step": 5160
    },
    {
      "epoch": 2.58,
      "learning_rate": 1.3887962654218073e-05,
      "loss": 4.9393,
      "step": 5180
    },
    {
      "epoch": 2.59,
      "learning_rate": 1.3554518172724243e-05,
      "loss": 4.9335,
      "step": 5200
    },
    {
      "epoch": 2.59,
      "eval_accuracy": 0.38211472315137435,
      "eval_loss": 4.98828125,
      "eval_runtime": 1028.0221,
      "eval_samples_per_second": 10.04,
      "eval_steps_per_second": 0.158,
      "step": 5200
    },
    {
      "epoch": 2.6,
      "learning_rate": 1.3221073691230409e-05,
      "loss": 4.919,
      "step": 5220
    },
    {
      "epoch": 2.61,
      "learning_rate": 1.2887629209736579e-05,
      "loss": 4.9249,
      "step": 5240
    },
    {
      "epoch": 2.62,
      "learning_rate": 1.2554184728242748e-05,
      "loss": 4.8953,
      "step": 5260
    },
    {
      "epoch": 2.63,
      "learning_rate": 1.2220740246748916e-05,
      "loss": 4.9371,
      "step": 5280
    },
    {
      "epoch": 2.64,
      "learning_rate": 1.1887295765255086e-05,
      "loss": 4.9752,
      "step": 5300
    },
    {
      "epoch": 2.65,
      "learning_rate": 1.1553851283761254e-05,
      "loss": 4.9155,
      "step": 5320
    },
    {
      "epoch": 2.66,
      "learning_rate": 1.1220406802267424e-05,
      "loss": 4.9085,
      "step": 5340
    },
    {
      "epoch": 2.67,
      "learning_rate": 1.0886962320773592e-05,
      "loss": 4.9508,
      "step": 5360
    },
    {
      "epoch": 2.68,
      "learning_rate": 1.055351783927976e-05,
      "loss": 4.8926,
      "step": 5380
    },
    {
      "epoch": 2.69,
      "learning_rate": 1.022007335778593e-05,
      "loss": 4.9231,
      "step": 5400
    },
    {
      "epoch": 2.69,
      "eval_accuracy": 0.38292519929510394,
      "eval_loss": 4.98046875,
      "eval_runtime": 786.1183,
      "eval_samples_per_second": 13.129,
      "eval_steps_per_second": 0.206,
      "step": 5400
    },
    {
      "epoch": 2.7,
      "learning_rate": 9.886628876292097e-06,
      "loss": 4.9256,
      "step": 5420
    },
    {
      "epoch": 2.71,
      "learning_rate": 9.553184394798265e-06,
      "loss": 4.9355,
      "step": 5440
    },
    {
      "epoch": 2.72,
      "learning_rate": 9.219739913304435e-06,
      "loss": 4.9676,
      "step": 5460
    },
    {
      "epoch": 2.73,
      "learning_rate": 8.886295431810605e-06,
      "loss": 4.8958,
      "step": 5480
    },
    {
      "epoch": 2.74,
      "learning_rate": 8.552850950316773e-06,
      "loss": 4.8705,
      "step": 5500
    },
    {
      "epoch": 2.75,
      "learning_rate": 8.219406468822942e-06,
      "loss": 4.9561,
      "step": 5520
    },
    {
      "epoch": 2.76,
      "learning_rate": 7.88596198732911e-06,
      "loss": 4.9005,
      "step": 5540
    },
    {
      "epoch": 2.77,
      "learning_rate": 7.552517505835278e-06,
      "loss": 4.8861,
      "step": 5560
    },
    {
      "epoch": 2.78,
      "learning_rate": 7.219073024341448e-06,
      "loss": 4.8598,
      "step": 5580
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.885628542847617e-06,
      "loss": 4.8886,
      "step": 5600
    },
    {
      "epoch": 2.79,
      "eval_accuracy": 0.3835265126706772,
      "eval_loss": 4.97265625,
      "eval_runtime": 539.9358,
      "eval_samples_per_second": 19.115,
      "eval_steps_per_second": 0.3,
      "step": 5600
    },
    {
      "epoch": 2.8,
      "learning_rate": 6.552184061353785e-06,
      "loss": 4.9213,
      "step": 5620
    },
    {
      "epoch": 2.81,
      "learning_rate": 6.2187395798599535e-06,
      "loss": 4.8934,
      "step": 5640
    },
    {
      "epoch": 2.82,
      "learning_rate": 5.885295098366122e-06,
      "loss": 4.8822,
      "step": 5660
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.551850616872291e-06,
      "loss": 4.9389,
      "step": 5680
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.21840613537846e-06,
      "loss": 4.9103,
      "step": 5700
    },
    {
      "epoch": 2.85,
      "learning_rate": 4.884961653884629e-06,
      "loss": 4.8878,
      "step": 5720
    },
    {
      "epoch": 2.86,
      "learning_rate": 4.551517172390797e-06,
      "loss": 4.9294,
      "step": 5740
    },
    {
      "epoch": 2.87,
      "learning_rate": 4.218072690896966e-06,
      "loss": 4.9309,
      "step": 5760
    },
    {
      "epoch": 2.88,
      "learning_rate": 3.8846282094031345e-06,
      "loss": 4.8913,
      "step": 5780
    },
    {
      "epoch": 2.89,
      "learning_rate": 3.5511837279093033e-06,
      "loss": 4.9419,
      "step": 5800
    },
    {
      "epoch": 2.89,
      "eval_accuracy": 0.38428875689298764,
      "eval_loss": 4.96484375,
      "eval_runtime": 548.4975,
      "eval_samples_per_second": 18.817,
      "eval_steps_per_second": 0.295,
      "step": 5800
    },
    {
      "epoch": 2.9,
      "learning_rate": 3.217739246415472e-06,
      "loss": 4.9406,
      "step": 5820
    },
    {
      "epoch": 2.91,
      "learning_rate": 2.884294764921641e-06,
      "loss": 4.8887,
      "step": 5840
    },
    {
      "epoch": 2.92,
      "learning_rate": 2.5508502834278094e-06,
      "loss": 4.9541,
      "step": 5860
    },
    {
      "epoch": 2.93,
      "learning_rate": 2.2174058019339782e-06,
      "loss": 4.8878,
      "step": 5880
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.8839613204401467e-06,
      "loss": 4.9403,
      "step": 5900
    },
    {
      "epoch": 2.95,
      "learning_rate": 1.5505168389463155e-06,
      "loss": 4.8785,
      "step": 5920
    },
    {
      "epoch": 2.96,
      "learning_rate": 1.2170723574524843e-06,
      "loss": 4.9111,
      "step": 5940
    },
    {
      "epoch": 2.97,
      "learning_rate": 8.836278759586529e-07,
      "loss": 4.8861,
      "step": 5960
    },
    {
      "epoch": 2.98,
      "learning_rate": 5.501833944648216e-07,
      "loss": 4.8789,
      "step": 5980
    },
    {
      "epoch": 2.99,
      "learning_rate": 2.1673891297099036e-07,
      "loss": 4.9227,
      "step": 6000
    },
    {
      "epoch": 2.99,
      "eval_accuracy": 0.3842425129408517,
      "eval_loss": 4.96484375,
      "eval_runtime": 533.8356,
      "eval_samples_per_second": 19.334,
      "eval_steps_per_second": 0.303,
      "step": 6000
    },
    {
      "epoch": 2.99,
      "step": 6001,
      "total_flos": 6.431553336633197e+18,
      "train_loss": 0.0007667992834527579,
      "train_runtime": 4.5998,
      "train_samples_per_second": 125223.897,
      "train_steps_per_second": 1304.416
    }
  ],
  "max_steps": 6000,
  "num_train_epochs": 3,
  "total_flos": 6.431553336633197e+18,
  "trial_name": null,
  "trial_params": null
}