{
  "best_metric": 0.24724383652210236,
  "best_model_checkpoint": "./convnext-tiny-new-8e-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": 29.175199508666992,
      "learning_rate": 7.998365801595384e-05,
      "loss": 2.5434,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 22.305095672607422,
      "learning_rate": 7.993464541683746e-05,
      "loss": 1.4809,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 20.301076889038086,
      "learning_rate": 7.98530022508065e-05,
      "loss": 1.2073,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 33.55647277832031,
      "learning_rate": 7.973879522842682e-05,
      "loss": 1.0004,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 17.641550064086914,
      "learning_rate": 7.959211766816531e-05,
      "loss": 0.9589,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 32.2988395690918,
      "learning_rate": 7.941308942013942e-05,
      "loss": 0.898,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 26.12941551208496,
      "learning_rate": 7.920185676818782e-05,
      "loss": 0.8587,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 24.61394500732422,
      "learning_rate": 7.895859231034193e-05,
      "loss": 0.8375,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 18.921947479248047,
      "learning_rate": 7.86834948177962e-05,
      "loss": 0.7619,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 25.84251594543457,
      "learning_rate": 7.837678907249235e-05,
      "loss": 0.7951,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8675944333996024,
      "eval_loss": 0.46665650606155396,
      "eval_runtime": 70.753,
      "eval_samples_per_second": 35.546,
      "eval_steps_per_second": 2.233,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 23.708938598632812,
      "learning_rate": 7.80387256834502e-05,
      "loss": 0.818,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 20.49905776977539,
      "learning_rate": 7.766958088199526e-05,
      "loss": 0.6782,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 17.50190544128418,
      "learning_rate": 7.726965629605035e-05,
      "loss": 0.6853,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 13.767265319824219,
      "learning_rate": 7.683927870367564e-05,
      "loss": 0.6206,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 16.55619239807129,
      "learning_rate": 7.637879976605853e-05,
      "loss": 0.626,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 17.5218448638916,
      "learning_rate": 7.588859574017165e-05,
      "loss": 0.6076,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 15.644243240356445,
      "learning_rate": 7.53690671713335e-05,
      "loss": 0.6628,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 27.80687141418457,
      "learning_rate": 7.482063856592323e-05,
      "loss": 0.6677,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 18.660202026367188,
      "learning_rate": 7.42437580445169e-05,
      "loss": 0.6035,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 14.59585189819336,
      "learning_rate": 7.363889697572835e-05,
      "loss": 0.5851,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 26.08392333984375,
      "learning_rate": 7.300654959105439e-05,
      "loss": 0.6475,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8946322067594433,
      "eval_loss": 0.3611379563808441,
      "eval_runtime": 70.7224,
      "eval_samples_per_second": 35.562,
      "eval_steps_per_second": 2.234,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 29.798437118530273,
      "learning_rate": 7.234723258103863e-05,
      "loss": 0.6025,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 12.655051231384277,
      "learning_rate": 7.1661484673084e-05,
      "loss": 0.5006,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 20.051542282104492,
      "learning_rate": 7.094986619125911e-05,
      "loss": 0.584,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 18.924829483032227,
      "learning_rate": 7.02129585984578e-05,
      "loss": 0.5216,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 10.966821670532227,
      "learning_rate": 6.945136402128628e-05,
      "loss": 0.5582,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 19.123348236083984,
      "learning_rate": 6.86657047580661e-05,
      "loss": 0.4905,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 18.773359298706055,
      "learning_rate": 6.785662277035447e-05,
      "loss": 0.5437,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 24.286714553833008,
      "learning_rate": 6.702477915839819e-05,
      "loss": 0.5196,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 10.000174522399902,
      "learning_rate": 6.617085362094902e-05,
      "loss": 0.5377,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 13.052635192871094,
      "learning_rate": 6.529554389988243e-05,
      "loss": 0.5193,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 39.736412048339844,
      "learning_rate": 6.43995652100733e-05,
      "loss": 0.4676,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8990059642147117,
      "eval_loss": 0.3682761788368225,
      "eval_runtime": 70.3285,
      "eval_samples_per_second": 35.761,
      "eval_steps_per_second": 2.247,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 11.428489685058594,
      "learning_rate": 6.348364965499434e-05,
      "loss": 0.5334,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 6.339554786682129,
      "learning_rate": 6.2548545628515e-05,
      "loss": 0.4044,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 11.18062973022461,
      "learning_rate": 6.159501720338938e-05,
      "loss": 0.4676,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 21.28518295288086,
      "learning_rate": 6.062384350693302e-05,
      "loss": 0.4708,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 20.110240936279297,
      "learning_rate": 5.9635818084398626e-05,
      "loss": 0.3822,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 19.29351806640625,
      "learning_rate": 5.86317482505708e-05,
      "loss": 0.4701,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 15.384368896484375,
      "learning_rate": 5.7612454430109804e-05,
      "loss": 0.4838,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 12.109505653381348,
      "learning_rate": 5.657876948718328e-05,
      "loss": 0.4809,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 14.05909538269043,
      "learning_rate": 5.5531538044933525e-05,
      "loss": 0.4506,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 4.24497652053833,
      "learning_rate": 5.447161579533662e-05,
      "loss": 0.4336,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 35.757415771484375,
      "learning_rate": 5.339986880001724e-05,
      "loss": 0.4181,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9240556660039762,
      "eval_loss": 0.28567156195640564,
      "eval_runtime": 70.8856,
      "eval_samples_per_second": 35.48,
      "eval_steps_per_second": 2.229,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.548322677612305,
      "learning_rate": 5.2317172782590326e-05,
      "loss": 0.3696,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 14.217641830444336,
      "learning_rate": 5.122441241310807e-05,
      "loss": 0.3903,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 19.054006576538086,
      "learning_rate": 5.012248058519667e-05,
      "loss": 0.3971,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 32.285396575927734,
      "learning_rate": 4.9012277686473734e-05,
      "loss": 0.3846,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 17.186689376831055,
      "learning_rate": 4.789471086284219e-05,
      "loss": 0.3451,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 17.26174545288086,
      "learning_rate": 4.6770693277262165e-05,
      "loss": 0.3571,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 9.664530754089355,
      "learning_rate": 4.5641143363606216e-05,
      "loss": 0.3848,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 22.393651962280273,
      "learning_rate": 4.450698407620776e-05,
      "loss": 0.3829,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 17.044584274291992,
      "learning_rate": 4.33691421357158e-05,
      "loss": 0.4061,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 18.844833374023438,
      "learning_rate": 4.22285472718723e-05,
      "loss": 0.3932,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 9.084076881408691,
      "learning_rate": 4.108613146383063e-05,
      "loss": 0.3648,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9149105367793241,
      "eval_loss": 0.29945191740989685,
      "eval_runtime": 70.985,
      "eval_samples_per_second": 35.43,
      "eval_steps_per_second": 2.226,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 15.977156639099121,
      "learning_rate": 3.994282817863628e-05,
      "loss": 0.3724,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 24.990800857543945,
      "learning_rate": 3.879957160849155e-05,
      "loss": 0.3132,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 18.232803344726562,
      "learning_rate": 3.7657295907427964e-05,
      "loss": 0.3289,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 11.785863876342773,
      "learning_rate": 3.651693442800964e-05,
      "loss": 0.3282,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 15.59240436553955,
      "learning_rate": 3.537941895869179e-05,
      "loss": 0.3132,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 17.803436279296875,
      "learning_rate": 3.4245678962457005e-05,
      "loss": 0.3315,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 18.81922149658203,
      "learning_rate": 3.31166408173518e-05,
      "loss": 0.3341,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 16.756589889526367,
      "learning_rate": 3.199322705954396e-05,
      "loss": 0.2879,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 24.542482376098633,
      "learning_rate": 3.087635562951882e-05,
      "loss": 0.3407,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 13.775747299194336,
      "learning_rate": 2.9766939122030964e-05,
      "loss": 0.3307,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 21.864892959594727,
      "learning_rate": 2.866588404042364e-05,
      "loss": 0.3241,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9312127236580517,
      "eval_loss": 0.27873358130455017,
      "eval_runtime": 70.6524,
      "eval_samples_per_second": 35.597,
      "eval_steps_per_second": 2.236,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 11.553291320800781,
      "learning_rate": 2.757409005592578e-05,
      "loss": 0.338,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 28.648635864257812,
      "learning_rate": 2.6492449272531175e-05,
      "loss": 0.2619,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 25.280593872070312,
      "learning_rate": 2.5421845498061117e-05,
      "loss": 0.3074,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 14.692756652832031,
      "learning_rate": 2.4363153522005743e-05,
      "loss": 0.317,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 6.455191135406494,
      "learning_rate": 2.3317238400734145e-05,
      "loss": 0.2545,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 10.77437973022461,
      "learning_rate": 2.228495475065759e-05,
      "loss": 0.3033,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 5.588379383087158,
      "learning_rate": 2.1267146049923167e-05,
      "loss": 0.2548,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 11.461979866027832,
      "learning_rate": 2.02646439492083e-05,
      "loss": 0.26,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 32.82159423828125,
      "learning_rate": 1.9278267592179803e-05,
      "loss": 0.2757,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 7.980381965637207,
      "learning_rate": 1.8308822946172174e-05,
      "loss": 0.2665,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 1.7043112516403198,
      "learning_rate": 1.7357102143632386e-05,
      "loss": 0.2845,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9351888667992048,
      "eval_loss": 0.2659090757369995,
      "eval_runtime": 70.8922,
      "eval_samples_per_second": 35.476,
      "eval_steps_per_second": 2.229,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 27.45222282409668,
      "learning_rate": 1.6423882834868976e-05,
      "loss": 0.2111,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 24.539175033569336,
      "learning_rate": 1.550992755263471e-05,
      "loss": 0.2432,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 12.749982833862305,
      "learning_rate": 1.4615983089061678e-05,
      "loss": 0.2416,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 7.099791049957275,
      "learning_rate": 1.3742779885457966e-05,
      "loss": 0.2546,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 17.301631927490234,
      "learning_rate": 1.2891031435464663e-05,
      "loss": 0.2551,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 4.71995735168457,
      "learning_rate": 1.206143370206086e-05,
      "loss": 0.2446,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 26.869333267211914,
      "learning_rate": 1.1254664548892755e-05,
      "loss": 0.2178,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 10.9271821975708,
      "learning_rate": 1.0471383186391817e-05,
      "loss": 0.2505,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 2.89125657081604,
      "learning_rate": 9.712229633134562e-06,
      "loss": 0.2039,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 24.978134155273438,
      "learning_rate": 8.977824192883772e-06,
      "loss": 0.2622,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 4.031610488891602,
      "learning_rate": 8.26876694773886e-06,
      "loss": 0.2132,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9359840954274354,
      "eval_loss": 0.25515902042388916,
      "eval_runtime": 70.3386,
      "eval_samples_per_second": 35.756,
      "eval_steps_per_second": 2.246,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 12.905261993408203,
      "learning_rate": 7.5856372678091964e-06,
      "loss": 0.2099,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 14.154178619384766,
      "learning_rate": 6.928993337811354e-06,
      "loss": 0.2228,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 4.56654167175293,
      "learning_rate": 6.29937170097672e-06,
      "loss": 0.2123,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 7.129326820373535,
      "learning_rate": 5.697286820642514e-06,
      "loss": 0.182,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 4.556196212768555,
      "learning_rate": 5.123230659884178e-06,
      "loss": 0.2376,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 2.216639518737793,
      "learning_rate": 4.577672279532746e-06,
      "loss": 0.19,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 7.883602142333984,
      "learning_rate": 4.061057454905615e-06,
      "loss": 0.222,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 13.847503662109375,
      "learning_rate": 3.573808311563891e-06,
      "loss": 0.2279,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 8.522872924804688,
      "learning_rate": 3.116322980393922e-06,
      "loss": 0.2201,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 12.57330322265625,
      "learning_rate": 2.68897527229488e-06,
      "loss": 0.2548,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 34.16948318481445,
      "learning_rate": 2.2921143727381525e-06,
      "loss": 0.177,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9371769383697813,
      "eval_loss": 0.25127747654914856,
      "eval_runtime": 70.7449,
      "eval_samples_per_second": 35.55,
      "eval_steps_per_second": 2.233,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 13.760716438293457,
      "learning_rate": 1.926064556448202e-06,
      "loss": 0.1822,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 4.8457350730896,
      "learning_rate": 1.5911249224379189e-06,
      "loss": 0.2013,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 22.193262100219727,
      "learning_rate": 1.28756914961508e-06,
      "loss": 0.198,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 18.170265197753906,
      "learning_rate": 1.0156452731595112e-06,
      "loss": 0.1789,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 17.805940628051758,
      "learning_rate": 7.755754818537542e-07,
      "loss": 0.224,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 12.388289451599121,
      "learning_rate": 5.675559365327798e-07,
      "loss": 0.1946,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 6.379385948181152,
      "learning_rate": 3.917566098011438e-07,
      "loss": 0.2367,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 19.94425392150879,
      "learning_rate": 2.4832114714847367e-07,
      "loss": 0.2286,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 18.711742401123047,
      "learning_rate": 1.3736674957689045e-07,
      "loss": 0.2218,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 9.24134349822998,
      "learning_rate": 5.898407783614169e-08,
      "loss": 0.2121,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 14.733855247497559,
      "learning_rate": 1.3237178344787106e-08,
      "loss": 0.2131,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9391650099403579,
      "eval_loss": 0.24724383652210236,
      "eval_runtime": 71.3719,
      "eval_samples_per_second": 35.238,
      "eval_steps_per_second": 2.214,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 1.301428412334932e+19,
      "train_loss": 0.4344568561488005,
      "train_runtime": 10707.0666,
      "train_samples_per_second": 16.42,
      "train_steps_per_second": 1.026
    }
  ],
  "logging_steps": 100,
  "max_steps": 10990,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1.301428412334932e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}