{
  "best_metric": 0.22793245315551758,
  "best_model_checkpoint": "./convnext-base-1e-5/checkpoint-10990",
  "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": 16.468570709228516,
      "learning_rate": 9.99795725199423e-06,
      "loss": 3.073,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 25.60658836364746,
      "learning_rate": 9.991830677104682e-06,
      "loss": 2.2679,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 20.34626579284668,
      "learning_rate": 9.981625281350812e-06,
      "loss": 1.7379,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 24.378923416137695,
      "learning_rate": 9.967349403553353e-06,
      "loss": 1.2829,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 22.901121139526367,
      "learning_rate": 9.949014708520664e-06,
      "loss": 1.0767,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 21.893035888671875,
      "learning_rate": 9.926636177517427e-06,
      "loss": 0.8904,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 27.2147216796875,
      "learning_rate": 9.900232096023478e-06,
      "loss": 0.8418,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 35.865386962890625,
      "learning_rate": 9.86982403879274e-06,
      "loss": 0.7409,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.933706283569336,
      "learning_rate": 9.835436852224525e-06,
      "loss": 0.6466,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.289974212646484,
      "learning_rate": 9.797098634061543e-06,
      "loss": 0.6315,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.882703777335984,
      "eval_loss": 0.46150895953178406,
      "eval_runtime": 104.4689,
      "eval_samples_per_second": 24.074,
      "eval_steps_per_second": 1.512,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 21.11526870727539,
      "learning_rate": 9.754840710431274e-06,
      "loss": 0.6502,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 15.997396469116211,
      "learning_rate": 9.708697610249407e-06,
      "loss": 0.5333,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 23.112411499023438,
      "learning_rate": 9.658707037006294e-06,
      "loss": 0.519,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 20.990036010742188,
      "learning_rate": 9.604909837959456e-06,
      "loss": 0.4819,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 42.656829833984375,
      "learning_rate": 9.547349970757317e-06,
      "loss": 0.4743,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 23.052263259887695,
      "learning_rate": 9.486074467521456e-06,
      "loss": 0.4237,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 14.877523422241211,
      "learning_rate": 9.421133396416687e-06,
      "loss": 0.4838,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 22.42397689819336,
      "learning_rate": 9.352579820740404e-06,
      "loss": 0.4973,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 18.889188766479492,
      "learning_rate": 9.280469755564613e-06,
      "loss": 0.4411,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 24.70013999938965,
      "learning_rate": 9.204862121966044e-06,
      "loss": 0.3991,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 17.89081573486328,
      "learning_rate": 9.125818698881798e-06,
      "loss": 0.4379,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9041749502982107,
      "eval_loss": 0.3370091915130615,
      "eval_runtime": 104.295,
      "eval_samples_per_second": 24.114,
      "eval_steps_per_second": 1.515,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 41.93815612792969,
      "learning_rate": 9.04340407262983e-06,
      "loss": 0.4006,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 20.993558883666992,
      "learning_rate": 8.9576855841355e-06,
      "loss": 0.3839,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 27.155363082885742,
      "learning_rate": 8.86873327390739e-06,
      "loss": 0.3816,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 17.326332092285156,
      "learning_rate": 8.776619824807225e-06,
      "loss": 0.3352,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 12.79124927520752,
      "learning_rate": 8.681420502660785e-06,
      "loss": 0.4064,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 37.5992546081543,
      "learning_rate": 8.583213094758262e-06,
      "loss": 0.3464,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 20.579957962036133,
      "learning_rate": 8.48207784629431e-06,
      "loss": 0.3477,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 20.70344352722168,
      "learning_rate": 8.378097394799774e-06,
      "loss": 0.3709,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 7.414881229400635,
      "learning_rate": 8.271356702618627e-06,
      "loss": 0.3247,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 19.997455596923828,
      "learning_rate": 8.161942987485303e-06,
      "loss": 0.3426,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 19.284587860107422,
      "learning_rate": 8.049945651259163e-06,
      "loss": 0.334,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9264413518886679,
      "eval_loss": 0.27322468161582947,
      "eval_runtime": 103.8233,
      "eval_samples_per_second": 24.224,
      "eval_steps_per_second": 1.522,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 19.56566047668457,
      "learning_rate": 7.935456206874292e-06,
      "loss": 0.3583,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 7.5106120109558105,
      "learning_rate": 7.818568203564375e-06,
      "loss": 0.2861,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 16.039175033569336,
      "learning_rate": 7.699377150423673e-06,
      "loss": 0.3014,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 20.261781692504883,
      "learning_rate": 7.577980438366628e-06,
      "loss": 0.3144,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 10.892475128173828,
      "learning_rate": 7.454477260549828e-06,
      "loss": 0.2924,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 26.032447814941406,
      "learning_rate": 7.32896853132135e-06,
      "loss": 0.2658,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 12.755492210388184,
      "learning_rate": 7.2015568037637255e-06,
      "loss": 0.3108,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 17.241613388061523,
      "learning_rate": 7.07234618589791e-06,
      "loss": 0.3001,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 30.844722747802734,
      "learning_rate": 6.941442255616691e-06,
      "loss": 0.3036,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.245629906654358,
      "learning_rate": 6.808951974417077e-06,
      "loss": 0.2921,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 32.295589447021484,
      "learning_rate": 6.674983600002155e-06,
      "loss": 0.2928,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9300198807157057,
      "eval_loss": 0.24411818385124207,
      "eval_runtime": 103.9147,
      "eval_samples_per_second": 24.203,
      "eval_steps_per_second": 1.52,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 25.101146697998047,
      "learning_rate": 6.539646597823791e-06,
      "loss": 0.2702,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 15.633359909057617,
      "learning_rate": 6.4030515516385085e-06,
      "loss": 0.2647,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 20.75804328918457,
      "learning_rate": 6.265310073149584e-06,
      "loss": 0.2643,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 10.434358596801758,
      "learning_rate": 6.126534710809217e-06,
      "loss": 0.2728,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 2.034339189529419,
      "learning_rate": 5.9868388578552736e-06,
      "loss": 0.2526,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 15.276336669921875,
      "learning_rate": 5.8463366596577706e-06,
      "loss": 0.2443,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 31.355663299560547,
      "learning_rate": 5.705142920450777e-06,
      "loss": 0.2183,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 12.438169479370117,
      "learning_rate": 5.56337300952597e-06,
      "loss": 0.2489,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 13.544615745544434,
      "learning_rate": 5.421142766964475e-06,
      "loss": 0.2771,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 21.137969970703125,
      "learning_rate": 5.2785684089840375e-06,
      "loss": 0.2631,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 33.35549545288086,
      "learning_rate": 5.135766432978829e-06,
      "loss": 0.2626,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9308151093439364,
      "eval_loss": 0.24758844077587128,
      "eval_runtime": 103.2721,
      "eval_samples_per_second": 24.353,
      "eval_steps_per_second": 1.53,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 21.884260177612305,
      "learning_rate": 4.992853522329535e-06,
      "loss": 0.2438,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 11.009835243225098,
      "learning_rate": 4.849946451061444e-06,
      "loss": 0.2043,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 32.59590530395508,
      "learning_rate": 4.7071619884284955e-06,
      "loss": 0.2166,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 2.92531418800354,
      "learning_rate": 4.564616803501205e-06,
      "loss": 0.2368,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.7258028984069824,
      "learning_rate": 4.422427369836474e-06,
      "loss": 0.2164,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 2.2120249271392822,
      "learning_rate": 4.280709870307126e-06,
      "loss": 0.2097,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 22.97938346862793,
      "learning_rate": 4.139580102168975e-06,
      "loss": 0.252,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 10.621854782104492,
      "learning_rate": 3.999153382442995e-06,
      "loss": 0.2068,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 16.208162307739258,
      "learning_rate": 3.859544453689853e-06,
      "loss": 0.2531,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 11.100607872009277,
      "learning_rate": 3.7208673902538705e-06,
      "loss": 0.216,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 30.43475914001465,
      "learning_rate": 3.583235505052955e-06,
      "loss": 0.2611,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9379721669980119,
      "eval_loss": 0.2386939376592636,
      "eval_runtime": 103.8465,
      "eval_samples_per_second": 24.218,
      "eval_steps_per_second": 1.521,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 36.211509704589844,
      "learning_rate": 3.4467612569907226e-06,
      "loss": 0.2483,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 35.222206115722656,
      "learning_rate": 3.311556159066397e-06,
      "loss": 0.2143,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 42.25660705566406,
      "learning_rate": 3.1777306872576396e-06,
      "loss": 0.202,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 18.040760040283203,
      "learning_rate": 3.045394190250718e-06,
      "loss": 0.201,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.627560019493103,
      "learning_rate": 2.914654800091768e-06,
      "loss": 0.1875,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.3061462640762329,
      "learning_rate": 2.785619343832199e-06,
      "loss": 0.2351,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 3.600099563598633,
      "learning_rate": 2.658393256240396e-06,
      "loss": 0.1924,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 3.9403347969055176,
      "learning_rate": 2.5330804936510374e-06,
      "loss": 0.2057,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 22.147470474243164,
      "learning_rate": 2.4097834490224754e-06,
      "loss": 0.214,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 32.04662322998047,
      "learning_rate": 2.2886028682715217e-06,
      "loss": 0.1874,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 16.505617141723633,
      "learning_rate": 2.169637767954048e-06,
      "loss": 0.2312,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9395626242544731,
      "eval_loss": 0.232571080327034,
      "eval_runtime": 103.7864,
      "eval_samples_per_second": 24.232,
      "eval_steps_per_second": 1.522,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 31.77814483642578,
      "learning_rate": 2.052985354358622e-06,
      "loss": 0.1744,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 27.672454833984375,
      "learning_rate": 1.9387409440793387e-06,
      "loss": 0.1897,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 5.418594837188721,
      "learning_rate": 1.8269978861327097e-06,
      "loss": 0.2036,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 3.6510262489318848,
      "learning_rate": 1.7178474856822457e-06,
      "loss": 0.1938,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 42.624839782714844,
      "learning_rate": 1.611378929433083e-06,
      "loss": 0.2046,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 5.9900383949279785,
      "learning_rate": 1.5076792127576074e-06,
      "loss": 0.1706,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 40.733543395996094,
      "learning_rate": 1.4068330686115943e-06,
      "loss": 0.1955,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 21.81285858154297,
      "learning_rate": 1.3089228982989771e-06,
      "loss": 0.1917,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 6.0959672927856445,
      "learning_rate": 1.2140287041418203e-06,
      "loss": 0.2048,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 2.601531982421875,
      "learning_rate": 1.1222280241104716e-06,
      "loss": 0.1836,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.49629172682762146,
      "learning_rate": 1.0335958684673574e-06,
      "loss": 0.18,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9379721669980119,
      "eval_loss": 0.2283337563276291,
      "eval_runtime": 103.326,
      "eval_samples_per_second": 24.34,
      "eval_steps_per_second": 1.529,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 34.25825119018555,
      "learning_rate": 9.482046584761496e-07,
      "loss": 0.1761,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 21.655656814575195,
      "learning_rate": 8.661241672264193e-07,
      "loss": 0.2155,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 4.431504249572754,
      "learning_rate": 7.8742146262209e-07,
      "loss": 0.2004,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 17.799575805664062,
      "learning_rate": 7.121608525803142e-07,
      "loss": 0.171,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 20.39786720275879,
      "learning_rate": 6.404038324855222e-07,
      "loss": 0.1945,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 1.1519782543182373,
      "learning_rate": 5.722090349415932e-07,
      "loss": 0.1503,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 17.413345336914062,
      "learning_rate": 5.076321818632018e-07,
      "loss": 0.1892,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 5.572515964508057,
      "learning_rate": 4.467260389454864e-07,
      "loss": 0.1988,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 3.1129953861236572,
      "learning_rate": 3.8954037254924026e-07,
      "loss": 0.1971,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 14.042587280273438,
      "learning_rate": 3.3612190903686e-07,
      "loss": 0.1976,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 16.600317001342773,
      "learning_rate": 2.8651429659226906e-07,
      "loss": 0.198,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9379721669980119,
      "eval_loss": 0.228021040558815,
      "eval_runtime": 103.859,
      "eval_samples_per_second": 24.216,
      "eval_steps_per_second": 1.521,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 20.190162658691406,
      "learning_rate": 2.407580695560252e-07,
      "loss": 0.1804,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 0.5975027680397034,
      "learning_rate": 1.9889061530473986e-07,
      "loss": 0.1574,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 33.140628814697266,
      "learning_rate": 1.60946143701885e-07,
      "loss": 0.1855,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 20.75016975402832,
      "learning_rate": 1.269556591449389e-07,
      "loss": 0.184,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 12.881192207336426,
      "learning_rate": 9.694693523171927e-08,
      "loss": 0.1763,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 10.315765380859375,
      "learning_rate": 7.094449206659748e-08,
      "loss": 0.1565,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 3.9750170707702637,
      "learning_rate": 4.896957622514298e-08,
      "loss": 0.206,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 17.893590927124023,
      "learning_rate": 3.104014339355921e-08,
      "loss": 0.196,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 12.213136672973633,
      "learning_rate": 1.7170843697111307e-08,
      "loss": 0.1999,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 35.23320007324219,
      "learning_rate": 7.373009729517711e-09,
      "loss": 0.1873,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 27.422964096069336,
      "learning_rate": 1.6546472930983882e-09,
      "loss": 0.164,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9375745526838967,
      "eval_loss": 0.22793245315551758,
      "eval_runtime": 103.2002,
      "eval_samples_per_second": 24.37,
      "eval_steps_per_second": 1.531,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.3625432834503757,
      "train_runtime": 16413.7517,
      "train_samples_per_second": 10.711,
      "train_steps_per_second": 0.67
    }
  ],
  "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
}