{
  "best_metric": 0.2340400665998459,
  "best_model_checkpoint": "./convnext-base-3e-5/checkpoint-4396",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 10990,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09,
      "grad_norm": 17.956300735473633,
      "learning_rate": 2.999387175598269e-05,
      "loss": 2.4621,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 29.822656631469727,
      "learning_rate": 2.9975492031314045e-05,
      "loss": 1.2937,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 17.655864715576172,
      "learning_rate": 2.994487584405244e-05,
      "loss": 0.9072,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 26.621862411499023,
      "learning_rate": 2.990204821066006e-05,
      "loss": 0.6825,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 30.1605224609375,
      "learning_rate": 2.984704412556199e-05,
      "loss": 0.6469,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 20.283042907714844,
      "learning_rate": 2.977990853255228e-05,
      "loss": 0.5517,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 23.60833740234375,
      "learning_rate": 2.970069628807043e-05,
      "loss": 0.5723,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 33.12726974487305,
      "learning_rate": 2.9609472116378222e-05,
      "loss": 0.5324,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.965418815612793,
      "learning_rate": 2.9506310556673573e-05,
      "loss": 0.4686,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.123353958129883,
      "learning_rate": 2.9391295902184625e-05,
      "loss": 0.4903,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8910536779324055,
      "eval_loss": 0.36899805068969727,
      "eval_runtime": 105.6572,
      "eval_samples_per_second": 23.803,
      "eval_steps_per_second": 1.495,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 21.59682273864746,
      "learning_rate": 2.9264522131293818e-05,
      "loss": 0.5236,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.269460678100586,
      "learning_rate": 2.9126092830748217e-05,
      "loss": 0.3906,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 15.037210464477539,
      "learning_rate": 2.897612111101888e-05,
      "loss": 0.3786,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.73908519744873,
      "learning_rate": 2.8814729513878365e-05,
      "loss": 0.3857,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 22.919809341430664,
      "learning_rate": 2.864204991227195e-05,
      "loss": 0.3413,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 16.75708770751953,
      "learning_rate": 2.8458223402564366e-05,
      "loss": 0.3258,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 13.096406936645508,
      "learning_rate": 2.826340018925006e-05,
      "loss": 0.4003,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 19.19637680053711,
      "learning_rate": 2.8057739462221215e-05,
      "loss": 0.385,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 17.63860321044922,
      "learning_rate": 2.7841409266693838e-05,
      "loss": 0.3486,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 15.726961135864258,
      "learning_rate": 2.761458636589813e-05,
      "loss": 0.3403,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 13.003602027893066,
      "learning_rate": 2.7377456096645395e-05,
      "loss": 0.3615,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9168986083499006,
      "eval_loss": 0.2985592782497406,
      "eval_runtime": 106.1115,
      "eval_samples_per_second": 23.701,
      "eval_steps_per_second": 1.489,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 31.20207405090332,
      "learning_rate": 2.7130212217889484e-05,
      "loss": 0.353,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 17.543489456176758,
      "learning_rate": 2.6873056752406504e-05,
      "loss": 0.307,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 16.423120498657227,
      "learning_rate": 2.6606199821722166e-05,
      "loss": 0.2946,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 10.412177085876465,
      "learning_rate": 2.632985947442167e-05,
      "loss": 0.2547,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 11.482078552246094,
      "learning_rate": 2.6044261507982356e-05,
      "loss": 0.3145,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 18.31868553161621,
      "learning_rate": 2.5749639284274782e-05,
      "loss": 0.2831,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.5178425312042236,
      "learning_rate": 2.5446233538882924e-05,
      "loss": 0.294,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 14.94119644165039,
      "learning_rate": 2.513429218439932e-05,
      "loss": 0.306,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 16.603517532348633,
      "learning_rate": 2.4814070107855878e-05,
      "loss": 0.2681,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 17.19023895263672,
      "learning_rate": 2.448582896245591e-05,
      "loss": 0.2738,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 16.443864822387695,
      "learning_rate": 2.4149836953777488e-05,
      "loss": 0.2686,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9304174950298211,
      "eval_loss": 0.26619353890419006,
      "eval_runtime": 105.8407,
      "eval_samples_per_second": 23.762,
      "eval_steps_per_second": 1.493,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.363679885864258,
      "learning_rate": 2.3806368620622876e-05,
      "loss": 0.2955,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 6.975516319274902,
      "learning_rate": 2.345570461069312e-05,
      "loss": 0.2085,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 3.4055590629577637,
      "learning_rate": 2.3098131451271016e-05,
      "loss": 0.2276,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 16.472017288208008,
      "learning_rate": 2.2733941315099883e-05,
      "loss": 0.2423,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 9.783247947692871,
      "learning_rate": 2.2363431781649483e-05,
      "loss": 0.2187,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 20.065593719482422,
      "learning_rate": 2.1986905593964048e-05,
      "loss": 0.1958,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 9.611791610717773,
      "learning_rate": 2.1604670411291174e-05,
      "loss": 0.2657,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 7.389467239379883,
      "learning_rate": 2.121703855769373e-05,
      "loss": 0.2479,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 7.945903301239014,
      "learning_rate": 2.0824326766850072e-05,
      "loss": 0.2279,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.05080638453364372,
      "learning_rate": 2.042685592325123e-05,
      "loss": 0.2171,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 8.823617935180664,
      "learning_rate": 2.0024950800006463e-05,
      "loss": 0.246,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9395626242544731,
      "eval_loss": 0.2340400665998459,
      "eval_runtime": 106.7505,
      "eval_samples_per_second": 23.56,
      "eval_steps_per_second": 1.48,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.573821067810059,
      "learning_rate": 1.961893979347137e-05,
      "loss": 0.2168,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 12.596579551696777,
      "learning_rate": 1.9209154654915524e-05,
      "loss": 0.1946,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 15.590298652648926,
      "learning_rate": 1.879593021944875e-05,
      "loss": 0.2028,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 20.42356300354004,
      "learning_rate": 1.837960413242765e-05,
      "loss": 0.1985,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 1.106755018234253,
      "learning_rate": 1.796051657356582e-05,
      "loss": 0.1907,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 4.9287495613098145,
      "learning_rate": 1.7539009978973312e-05,
      "loss": 0.1858,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 25.559234619140625,
      "learning_rate": 1.711542876135233e-05,
      "loss": 0.1694,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 22.039155960083008,
      "learning_rate": 1.669011902857791e-05,
      "loss": 0.1891,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 3.6209306716918945,
      "learning_rate": 1.6263428300893422e-05,
      "loss": 0.1981,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 30.3158016204834,
      "learning_rate": 1.5835705226952112e-05,
      "loss": 0.1867,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 32.05391311645508,
      "learning_rate": 1.540729929893649e-05,
      "loss": 0.201,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9411530815109344,
      "eval_loss": 0.25883305072784424,
      "eval_runtime": 105.5526,
      "eval_samples_per_second": 23.827,
      "eval_steps_per_second": 1.497,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.146540641784668,
      "learning_rate": 1.4978560566988603e-05,
      "loss": 0.1906,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 15.014126777648926,
      "learning_rate": 1.454983935318433e-05,
      "loss": 0.1563,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 7.286404609680176,
      "learning_rate": 1.4121485965285485e-05,
      "loss": 0.1526,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 19.885345458984375,
      "learning_rate": 1.3693850410503614e-05,
      "loss": 0.1482,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.03801773115992546,
      "learning_rate": 1.326728210950942e-05,
      "loss": 0.1619,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.256340354681015,
      "learning_rate": 1.2842129610921378e-05,
      "loss": 0.1431,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 6.246147632598877,
      "learning_rate": 1.2418740306506923e-05,
      "loss": 0.194,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.7258792519569397,
      "learning_rate": 1.1997460147328984e-05,
      "loss": 0.1388,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 5.175995826721191,
      "learning_rate": 1.1578633361069559e-05,
      "loss": 0.1943,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 5.505824089050293,
      "learning_rate": 1.1162602170761611e-05,
      "loss": 0.1617,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 22.753047943115234,
      "learning_rate": 1.0749706515158863e-05,
      "loss": 0.1757,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9439363817097416,
      "eval_loss": 0.2435537427663803,
      "eval_runtime": 105.5095,
      "eval_samples_per_second": 23.837,
      "eval_steps_per_second": 1.497,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 17.974355697631836,
      "learning_rate": 1.0340283770972167e-05,
      "loss": 0.1683,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 26.32249641418457,
      "learning_rate": 9.93466847719919e-06,
      "loss": 0.1313,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 8.388775825500488,
      "learning_rate": 9.533192061772919e-06,
      "loss": 0.1363,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 15.968542098999023,
      "learning_rate": 9.136182570752153e-06,
      "loss": 0.1241,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.21214519441127777,
      "learning_rate": 8.743964400275304e-06,
      "loss": 0.1201,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.1043318435549736,
      "learning_rate": 8.356858031496596e-06,
      "loss": 0.1438,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.6806092858314514,
      "learning_rate": 7.975179768721187e-06,
      "loss": 0.13,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.5344202518463135,
      "learning_rate": 7.599241480953112e-06,
      "loss": 0.1393,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 28.392440795898438,
      "learning_rate": 7.229350347067426e-06,
      "loss": 0.1441,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 21.353750228881836,
      "learning_rate": 6.865808604814564e-06,
      "loss": 0.1211,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 3.2531442642211914,
      "learning_rate": 6.508913303862144e-06,
      "loss": 0.1699,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9443339960238568,
      "eval_loss": 0.2558170557022095,
      "eval_runtime": 106.0537,
      "eval_samples_per_second": 23.714,
      "eval_steps_per_second": 1.49,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 27.954816818237305,
      "learning_rate": 6.1589560630758656e-06,
      "loss": 0.1067,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 12.10332202911377,
      "learning_rate": 5.8162228322380155e-06,
      "loss": 0.1084,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 1.8917927742004395,
      "learning_rate": 5.480993658398129e-06,
      "loss": 0.1441,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.24689602851867676,
      "learning_rate": 5.153542457046737e-06,
      "loss": 0.1006,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 3.907996416091919,
      "learning_rate": 4.834136788299248e-06,
      "loss": 0.1243,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.24271999299526215,
      "learning_rate": 4.523037638272822e-06,
      "loss": 0.0981,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 20.29511833190918,
      "learning_rate": 4.220499205834783e-06,
      "loss": 0.1046,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 2.0506038665771484,
      "learning_rate": 3.926768694896931e-06,
      "loss": 0.1175,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.33996373414993286,
      "learning_rate": 3.6420861124254607e-06,
      "loss": 0.1491,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.5826025605201721,
      "learning_rate": 3.3666840723314145e-06,
      "loss": 0.1164,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 3.4220075607299805,
      "learning_rate": 3.1007876054020724e-06,
      "loss": 0.1178,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9479125248508946,
      "eval_loss": 0.2512185275554657,
      "eval_runtime": 105.3589,
      "eval_samples_per_second": 23.871,
      "eval_steps_per_second": 1.5,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 5.513002395629883,
      "learning_rate": 2.8446139754284486e-06,
      "loss": 0.1009,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 15.36313533782959,
      "learning_rate": 2.5983725016792574e-06,
      "loss": 0.1245,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 3.3562428951263428,
      "learning_rate": 2.36226438786627e-06,
      "loss": 0.1124,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 28.229764938354492,
      "learning_rate": 2.1364825577409424e-06,
      "loss": 0.0913,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 3.830212116241455,
      "learning_rate": 1.9212114974565664e-06,
      "loss": 0.1226,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.01726786606013775,
      "learning_rate": 1.7166271048247796e-06,
      "loss": 0.0825,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 10.166333198547363,
      "learning_rate": 1.5228965455896054e-06,
      "loss": 0.0989,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 18.327802658081055,
      "learning_rate": 1.3401781168364591e-06,
      "loss": 0.105,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.282446950674057,
      "learning_rate": 1.1686211176477208e-06,
      "loss": 0.1097,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 21.263484954833984,
      "learning_rate": 1.00836572711058e-06,
      "loss": 0.1297,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 4.42637825012207,
      "learning_rate": 8.595428897768071e-07,
      "loss": 0.1167,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9479125248508946,
      "eval_loss": 0.2536466717720032,
      "eval_runtime": 105.7662,
      "eval_samples_per_second": 23.779,
      "eval_steps_per_second": 1.494,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 9.55751895904541,
      "learning_rate": 7.222742086680756e-07,
      "loss": 0.1121,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 0.058909401297569275,
      "learning_rate": 5.966718459142196e-07,
      "loss": 0.0856,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 27.23402214050293,
      "learning_rate": 4.82838431105655e-07,
      "loss": 0.1041,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.20296531915664673,
      "learning_rate": 3.808669774348167e-07,
      "loss": 0.1079,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 0.6779181957244873,
      "learning_rate": 2.908408056951578e-07,
      "loss": 0.0855,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.20246420800685883,
      "learning_rate": 2.1283347619979243e-07,
      "loss": 0.0715,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 0.13543924689292908,
      "learning_rate": 1.4690872867542892e-07,
      "loss": 0.1181,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 12.729443550109863,
      "learning_rate": 9.312043018067762e-08,
      "loss": 0.1114,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 11.665053367614746,
      "learning_rate": 5.1512531091333914e-08,
      "loss": 0.131,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 5.462090969085693,
      "learning_rate": 2.211902918855313e-08,
      "loss": 0.1174,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 8.030549049377441,
      "learning_rate": 4.963941879295164e-09,
      "loss": 0.0889,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9495029821073558,
      "eval_loss": 0.25355264544487,
      "eval_runtime": 105.6261,
      "eval_samples_per_second": 23.81,
      "eval_steps_per_second": 1.496,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.25289429932751367,
      "train_runtime": 17079.3849,
      "train_samples_per_second": 10.294,
      "train_steps_per_second": 0.643
    }
  ],
  "logging_steps": 100,
  "max_steps": 10990,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 4.09349935387607e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}