{
  "best_metric": 0.2772659659385681,
  "best_model_checkpoint": "./convnext-base-8e-5/checkpoint-8792",
  "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": 11.6831636428833,
      "learning_rate": 7.998365801595384e-05,
      "loss": 1.9321,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 23.91045379638672,
      "learning_rate": 7.993464541683746e-05,
      "loss": 0.9556,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 16.425535202026367,
      "learning_rate": 7.98530022508065e-05,
      "loss": 0.7791,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 18.70790672302246,
      "learning_rate": 7.973879522842682e-05,
      "loss": 0.6424,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 20.97853660583496,
      "learning_rate": 7.959211766816531e-05,
      "loss": 0.6717,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 19.02173614501953,
      "learning_rate": 7.941308942013942e-05,
      "loss": 0.5696,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 18.764028549194336,
      "learning_rate": 7.920185676818782e-05,
      "loss": 0.5981,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 18.06161880493164,
      "learning_rate": 7.895859231034193e-05,
      "loss": 0.5988,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 13.58993148803711,
      "learning_rate": 7.86834948177962e-05,
      "loss": 0.5275,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.574773788452148,
      "learning_rate": 7.837678907249235e-05,
      "loss": 0.5312,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.856858846918489,
      "eval_loss": 0.47219011187553406,
      "eval_runtime": 105.0222,
      "eval_samples_per_second": 23.947,
      "eval_steps_per_second": 1.504,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 15.577301979064941,
      "learning_rate": 7.80387256834502e-05,
      "loss": 0.5937,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.9820380210876465,
      "learning_rate": 7.766958088199526e-05,
      "loss": 0.4501,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.32746410369873,
      "learning_rate": 7.726965629605035e-05,
      "loss": 0.4311,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 17.756572723388672,
      "learning_rate": 7.683927870367564e-05,
      "loss": 0.4342,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 18.274375915527344,
      "learning_rate": 7.637879976605853e-05,
      "loss": 0.4238,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 20.55085563659668,
      "learning_rate": 7.588859574017165e-05,
      "loss": 0.3663,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.306856155395508,
      "learning_rate": 7.53690671713335e-05,
      "loss": 0.4696,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 17.713024139404297,
      "learning_rate": 7.482063856592323e-05,
      "loss": 0.4407,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.358209609985352,
      "learning_rate": 7.42437580445169e-05,
      "loss": 0.43,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 15.047048568725586,
      "learning_rate": 7.363889697572835e-05,
      "loss": 0.372,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 6.985563278198242,
      "learning_rate": 7.300654959105439e-05,
      "loss": 0.4064,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9025844930417495,
      "eval_loss": 0.36510175466537476,
      "eval_runtime": 106.7789,
      "eval_samples_per_second": 23.553,
      "eval_steps_per_second": 1.48,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 14.219414710998535,
      "learning_rate": 7.234723258103863e-05,
      "loss": 0.4234,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 8.225436210632324,
      "learning_rate": 7.1661484673084e-05,
      "loss": 0.3427,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 17.93203353881836,
      "learning_rate": 7.094986619125911e-05,
      "loss": 0.3326,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 6.593932151794434,
      "learning_rate": 7.02129585984578e-05,
      "loss": 0.3063,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 16.711748123168945,
      "learning_rate": 6.945136402128628e-05,
      "loss": 0.3645,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 4.706127643585205,
      "learning_rate": 6.86657047580661e-05,
      "loss": 0.3467,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.0358073711395264,
      "learning_rate": 6.785662277035447e-05,
      "loss": 0.3398,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 9.968092918395996,
      "learning_rate": 6.702477915839819e-05,
      "loss": 0.3555,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 8.269189834594727,
      "learning_rate": 6.617085362094902e-05,
      "loss": 0.3263,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 10.920413970947266,
      "learning_rate": 6.529554389988243e-05,
      "loss": 0.3033,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 11.657285690307617,
      "learning_rate": 6.43995652100733e-05,
      "loss": 0.318,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9145129224652088,
      "eval_loss": 0.32599496841430664,
      "eval_runtime": 105.5258,
      "eval_samples_per_second": 23.833,
      "eval_steps_per_second": 1.497,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 7.324591636657715,
      "learning_rate": 6.348364965499434e-05,
      "loss": 0.3295,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.3288770914077759,
      "learning_rate": 6.2548545628515e-05,
      "loss": 0.2353,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.16654238104820251,
      "learning_rate": 6.159501720338938e-05,
      "loss": 0.2515,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 11.06922721862793,
      "learning_rate": 6.062384350693302e-05,
      "loss": 0.2783,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 10.742453575134277,
      "learning_rate": 5.9635818084398626e-05,
      "loss": 0.2667,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 6.648169994354248,
      "learning_rate": 5.86317482505708e-05,
      "loss": 0.2368,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 11.184340476989746,
      "learning_rate": 5.7612454430109804e-05,
      "loss": 0.2933,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 7.161631107330322,
      "learning_rate": 5.657876948718328e-05,
      "loss": 0.2561,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.8756735920906067,
      "learning_rate": 5.5531538044933525e-05,
      "loss": 0.2809,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.12508328258991241,
      "learning_rate": 5.447161579533662e-05,
      "loss": 0.2495,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 3.783909320831299,
      "learning_rate": 5.339986880001724e-05,
      "loss": 0.2725,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9196819085487078,
      "eval_loss": 0.3164690434932709,
      "eval_runtime": 105.5164,
      "eval_samples_per_second": 23.835,
      "eval_steps_per_second": 1.497,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 13.933430671691895,
      "learning_rate": 5.2317172782590326e-05,
      "loss": 0.2394,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 6.490962982177734,
      "learning_rate": 5.122441241310807e-05,
      "loss": 0.2062,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 4.284941673278809,
      "learning_rate": 5.012248058519667e-05,
      "loss": 0.2109,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 9.453885078430176,
      "learning_rate": 4.9012277686473734e-05,
      "loss": 0.2091,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.10080689936876297,
      "learning_rate": 4.789471086284219e-05,
      "loss": 0.2065,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 9.822230339050293,
      "learning_rate": 4.6770693277262165e-05,
      "loss": 0.2135,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 7.230046272277832,
      "learning_rate": 4.5641143363606216e-05,
      "loss": 0.1851,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 13.58958911895752,
      "learning_rate": 4.450698407620776e-05,
      "loss": 0.1805,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 7.531309127807617,
      "learning_rate": 4.33691421357158e-05,
      "loss": 0.2118,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 17.00364875793457,
      "learning_rate": 4.22285472718723e-05,
      "loss": 0.1952,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 10.512598037719727,
      "learning_rate": 4.108613146383063e-05,
      "loss": 0.212,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9256461232604374,
      "eval_loss": 0.3330000936985016,
      "eval_runtime": 104.2603,
      "eval_samples_per_second": 24.122,
      "eval_steps_per_second": 1.515,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.856630325317383,
      "learning_rate": 3.994282817863628e-05,
      "loss": 0.1993,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 5.001902103424072,
      "learning_rate": 3.879957160849155e-05,
      "loss": 0.1622,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 8.48959732055664,
      "learning_rate": 3.7657295907427964e-05,
      "loss": 0.1709,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 1.4750434160232544,
      "learning_rate": 3.651693442800964e-05,
      "loss": 0.1536,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.030608195811510086,
      "learning_rate": 3.537941895869179e-05,
      "loss": 0.1609,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.20574839413166046,
      "learning_rate": 3.4245678962457005e-05,
      "loss": 0.1318,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.6550854444503784,
      "learning_rate": 3.31166408173518e-05,
      "loss": 0.1667,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 14.868585586547852,
      "learning_rate": 3.199322705954396e-05,
      "loss": 0.1402,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 5.7412896156311035,
      "learning_rate": 3.087635562951882e-05,
      "loss": 0.1862,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 11.104312896728516,
      "learning_rate": 2.9766939122030964e-05,
      "loss": 0.1592,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 12.107190132141113,
      "learning_rate": 2.866588404042364e-05,
      "loss": 0.1711,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.927634194831014,
      "eval_loss": 0.3272738456726074,
      "eval_runtime": 105.6481,
      "eval_samples_per_second": 23.805,
      "eval_steps_per_second": 1.496,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 8.225330352783203,
      "learning_rate": 2.757409005592578e-05,
      "loss": 0.1678,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 20.111202239990234,
      "learning_rate": 2.6492449272531175e-05,
      "loss": 0.1286,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 23.527673721313477,
      "learning_rate": 2.5421845498061117e-05,
      "loss": 0.1245,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 15.480986595153809,
      "learning_rate": 2.4363153522005743e-05,
      "loss": 0.1185,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.05464939400553703,
      "learning_rate": 2.3317238400734145e-05,
      "loss": 0.125,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.09599953889846802,
      "learning_rate": 2.228495475065759e-05,
      "loss": 0.132,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 21.158004760742188,
      "learning_rate": 2.1267146049923167e-05,
      "loss": 0.1252,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 17.28738784790039,
      "learning_rate": 2.02646439492083e-05,
      "loss": 0.1289,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 13.451830863952637,
      "learning_rate": 1.9278267592179803e-05,
      "loss": 0.1473,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 12.016372680664062,
      "learning_rate": 1.8308822946172174e-05,
      "loss": 0.1245,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 3.5636794567108154,
      "learning_rate": 1.7357102143632386e-05,
      "loss": 0.1552,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9304174950298211,
      "eval_loss": 0.29928237199783325,
      "eval_runtime": 105.3329,
      "eval_samples_per_second": 23.877,
      "eval_steps_per_second": 1.5,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 9.268345832824707,
      "learning_rate": 1.6423882834868976e-05,
      "loss": 0.0984,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 8.35660457611084,
      "learning_rate": 1.550992755263471e-05,
      "loss": 0.0894,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 2.3893661499023438,
      "learning_rate": 1.4615983089061678e-05,
      "loss": 0.1349,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.8107450604438782,
      "learning_rate": 1.3742779885457966e-05,
      "loss": 0.0983,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 3.668052911758423,
      "learning_rate": 1.2891031435464663e-05,
      "loss": 0.1119,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 1.1919211149215698,
      "learning_rate": 1.206143370206086e-05,
      "loss": 0.0952,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 8.72845458984375,
      "learning_rate": 1.1254664548892755e-05,
      "loss": 0.0973,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.031977806240320206,
      "learning_rate": 1.0471383186391817e-05,
      "loss": 0.1219,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.5276290774345398,
      "learning_rate": 9.712229633134562e-06,
      "loss": 0.118,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.260431170463562,
      "learning_rate": 8.977824192883772e-06,
      "loss": 0.1148,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 6.11373233795166,
      "learning_rate": 8.26876694773886e-06,
      "loss": 0.1039,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.941948310139165,
      "eval_loss": 0.2772659659385681,
      "eval_runtime": 105.3277,
      "eval_samples_per_second": 23.878,
      "eval_steps_per_second": 1.5,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.6443412899971008,
      "learning_rate": 7.5856372678091964e-06,
      "loss": 0.0876,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.9504626989364624,
      "learning_rate": 6.928993337811354e-06,
      "loss": 0.1,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 1.0361099243164062,
      "learning_rate": 6.29937170097672e-06,
      "loss": 0.0938,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 10.742104530334473,
      "learning_rate": 5.697286820642514e-06,
      "loss": 0.087,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.1779290735721588,
      "learning_rate": 5.123230659884178e-06,
      "loss": 0.0974,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.006353577133268118,
      "learning_rate": 4.577672279532746e-06,
      "loss": 0.0764,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 7.467300891876221,
      "learning_rate": 4.061057454905615e-06,
      "loss": 0.0875,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 2.8993914127349854,
      "learning_rate": 3.573808311563891e-06,
      "loss": 0.0817,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.09099097549915314,
      "learning_rate": 3.116322980393922e-06,
      "loss": 0.1057,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 8.745915412902832,
      "learning_rate": 2.68897527229488e-06,
      "loss": 0.1134,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 20.60851287841797,
      "learning_rate": 2.2921143727381525e-06,
      "loss": 0.0996,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.945924453280318,
      "eval_loss": 0.27863016724586487,
      "eval_runtime": 105.7885,
      "eval_samples_per_second": 23.774,
      "eval_steps_per_second": 1.494,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 2.4663915634155273,
      "learning_rate": 1.926064556448202e-06,
      "loss": 0.0928,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 0.007756831590086222,
      "learning_rate": 1.5911249224379189e-06,
      "loss": 0.0681,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 0.06630658358335495,
      "learning_rate": 1.28756914961508e-06,
      "loss": 0.0806,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.05963192135095596,
      "learning_rate": 1.0156452731595112e-06,
      "loss": 0.0906,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 0.010499324649572372,
      "learning_rate": 7.755754818537542e-07,
      "loss": 0.0677,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 1.1624623537063599,
      "learning_rate": 5.675559365327798e-07,
      "loss": 0.0816,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 0.02799438126385212,
      "learning_rate": 3.917566098011438e-07,
      "loss": 0.1033,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 10.219303131103516,
      "learning_rate": 2.4832114714847367e-07,
      "loss": 0.0825,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 5.508051872253418,
      "learning_rate": 1.3736674957689045e-07,
      "loss": 0.117,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 0.7247039675712585,
      "learning_rate": 5.898407783614169e-08,
      "loss": 0.0994,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 0.1574389934539795,
      "learning_rate": 1.3237178344787106e-08,
      "loss": 0.0765,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9475149105367793,
      "eval_loss": 0.27801209688186646,
      "eval_runtime": 105.1016,
      "eval_samples_per_second": 23.929,
      "eval_steps_per_second": 1.503,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.2560113200065328,
      "train_runtime": 17058.8463,
      "train_samples_per_second": 10.306,
      "train_steps_per_second": 0.644
    }
  ],
  "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
}