{
  "best_metric": 0.2916193902492523,
  "best_model_checkpoint": "./convnext-base/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": 12.231302261352539,
      "learning_rate": 9.99795725199423e-05,
      "loss": 1.8816,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 15.359071731567383,
      "learning_rate": 9.991830677104683e-05,
      "loss": 0.9793,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 18.62959098815918,
      "learning_rate": 9.981625281350813e-05,
      "loss": 0.8021,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 15.989691734313965,
      "learning_rate": 9.967349403553353e-05,
      "loss": 0.6843,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 13.825401306152344,
      "learning_rate": 9.949014708520663e-05,
      "loss": 0.7125,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 17.203922271728516,
      "learning_rate": 9.926636177517427e-05,
      "loss": 0.5937,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 13.448897361755371,
      "learning_rate": 9.900232096023477e-05,
      "loss": 0.6399,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.717751502990723,
      "learning_rate": 9.869824038792741e-05,
      "loss": 0.6142,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.631546020507812,
      "learning_rate": 9.835436852224525e-05,
      "loss": 0.5774,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 14.26085376739502,
      "learning_rate": 9.797098634061542e-05,
      "loss": 0.5947,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.868389662027833,
      "eval_loss": 0.42419883608818054,
      "eval_runtime": 109.2833,
      "eval_samples_per_second": 23.014,
      "eval_steps_per_second": 1.446,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.6874361038208,
      "learning_rate": 9.754840710431274e-05,
      "loss": 0.6197,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.372115135192871,
      "learning_rate": 9.708697610249406e-05,
      "loss": 0.4679,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 12.95173168182373,
      "learning_rate": 9.658707037006294e-05,
      "loss": 0.4685,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.779945373535156,
      "learning_rate": 9.604909837959455e-05,
      "loss": 0.4719,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 16.67346954345703,
      "learning_rate": 9.547349970757317e-05,
      "loss": 0.4539,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 20.67113494873047,
      "learning_rate": 9.486074467521456e-05,
      "loss": 0.455,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 13.069211959838867,
      "learning_rate": 9.421133396416686e-05,
      "loss": 0.5008,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 16.976469039916992,
      "learning_rate": 9.352579820740405e-05,
      "loss": 0.4652,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.576595306396484,
      "learning_rate": 9.280469755564613e-05,
      "loss": 0.4612,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 12.198715209960938,
      "learning_rate": 9.204862121966044e-05,
      "loss": 0.3841,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 12.983030319213867,
      "learning_rate": 9.125818698881798e-05,
      "loss": 0.4798,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8727634194831014,
      "eval_loss": 0.42417293787002563,
      "eval_runtime": 109.7743,
      "eval_samples_per_second": 22.911,
      "eval_steps_per_second": 1.439,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.284225463867188,
      "learning_rate": 9.043404072629829e-05,
      "loss": 0.4405,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 10.147916793823242,
      "learning_rate": 8.957685584135502e-05,
      "loss": 0.3597,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 17.661441802978516,
      "learning_rate": 8.86873327390739e-05,
      "loss": 0.3931,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 11.3087797164917,
      "learning_rate": 8.776619824807224e-05,
      "loss": 0.3019,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 7.633837699890137,
      "learning_rate": 8.681420502660786e-05,
      "loss": 0.3771,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 13.283084869384766,
      "learning_rate": 8.583213094758261e-05,
      "loss": 0.3434,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.092094898223877,
      "learning_rate": 8.482077846294308e-05,
      "loss": 0.3502,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 7.117873191833496,
      "learning_rate": 8.378097394799773e-05,
      "loss": 0.4205,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 4.9123215675354,
      "learning_rate": 8.271356702618626e-05,
      "loss": 0.359,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 6.197904109954834,
      "learning_rate": 8.161942987485303e-05,
      "loss": 0.3543,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 15.387776374816895,
      "learning_rate": 8.049945651259163e-05,
      "loss": 0.3625,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9077534791252485,
      "eval_loss": 0.3553118109703064,
      "eval_runtime": 109.5619,
      "eval_samples_per_second": 22.955,
      "eval_steps_per_second": 1.442,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 12.840432167053223,
      "learning_rate": 7.935456206874292e-05,
      "loss": 0.3386,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 17.542098999023438,
      "learning_rate": 7.818568203564374e-05,
      "loss": 0.2708,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 2.256465196609497,
      "learning_rate": 7.699377150423672e-05,
      "loss": 0.2731,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 13.450770378112793,
      "learning_rate": 7.577980438366628e-05,
      "loss": 0.2903,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 12.049773216247559,
      "learning_rate": 7.454477260549828e-05,
      "loss": 0.2849,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 5.758849143981934,
      "learning_rate": 7.32896853132135e-05,
      "loss": 0.2638,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 24.772600173950195,
      "learning_rate": 7.201556803763725e-05,
      "loss": 0.3027,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 4.821949481964111,
      "learning_rate": 7.07234618589791e-05,
      "loss": 0.2988,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 2.069714069366455,
      "learning_rate": 6.94144225561669e-05,
      "loss": 0.2806,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.6293389201164246,
      "learning_rate": 6.808951974417078e-05,
      "loss": 0.2627,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 8.25537109375,
      "learning_rate": 6.674983600002155e-05,
      "loss": 0.2777,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9184890656063618,
      "eval_loss": 0.32412779331207275,
      "eval_runtime": 109.2619,
      "eval_samples_per_second": 23.018,
      "eval_steps_per_second": 1.446,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.3740968704223633,
      "learning_rate": 6.539646597823791e-05,
      "loss": 0.2558,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 8.975493431091309,
      "learning_rate": 6.403051551638508e-05,
      "loss": 0.2225,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 7.33804988861084,
      "learning_rate": 6.265310073149584e-05,
      "loss": 0.2356,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 6.3810834884643555,
      "learning_rate": 6.126534710809216e-05,
      "loss": 0.2136,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.08626483380794525,
      "learning_rate": 5.9868388578552734e-05,
      "loss": 0.2346,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 8.195035934448242,
      "learning_rate": 5.8463366596577706e-05,
      "loss": 0.2214,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 2.5158026218414307,
      "learning_rate": 5.705142920450777e-05,
      "loss": 0.1991,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 1.2365189790725708,
      "learning_rate": 5.5633730095259695e-05,
      "loss": 0.2105,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 3.150585889816284,
      "learning_rate": 5.421142766964474e-05,
      "loss": 0.2264,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 21.01511001586914,
      "learning_rate": 5.278568408984037e-05,
      "loss": 0.213,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 16.65435791015625,
      "learning_rate": 5.135766432978829e-05,
      "loss": 0.2368,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9244532803180915,
      "eval_loss": 0.34125077724456787,
      "eval_runtime": 104.85,
      "eval_samples_per_second": 23.987,
      "eval_steps_per_second": 1.507,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.176883697509766,
      "learning_rate": 4.9928535223295344e-05,
      "loss": 0.2106,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.6758583188056946,
      "learning_rate": 4.849946451061443e-05,
      "loss": 0.1445,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 14.350722312927246,
      "learning_rate": 4.707161988428495e-05,
      "loss": 0.1812,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 1.7648365497589111,
      "learning_rate": 4.564616803501205e-05,
      "loss": 0.1644,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.08469072729349136,
      "learning_rate": 4.4224273698364735e-05,
      "loss": 0.1678,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 4.34414005279541,
      "learning_rate": 4.2807098703071255e-05,
      "loss": 0.1434,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 5.815812110900879,
      "learning_rate": 4.1395801021689746e-05,
      "loss": 0.188,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 1.176138997077942,
      "learning_rate": 3.999153382442995e-05,
      "loss": 0.1407,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 10.93614387512207,
      "learning_rate": 3.859544453689853e-05,
      "loss": 0.192,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 1.6799432039260864,
      "learning_rate": 3.7208673902538706e-05,
      "loss": 0.1552,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 15.9625883102417,
      "learning_rate": 3.583235505052955e-05,
      "loss": 0.1635,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9355864811133201,
      "eval_loss": 0.3116415739059448,
      "eval_runtime": 105.6961,
      "eval_samples_per_second": 23.795,
      "eval_steps_per_second": 1.495,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 8.467019081115723,
      "learning_rate": 3.446761256990723e-05,
      "loss": 0.1804,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 17.132722854614258,
      "learning_rate": 3.311556159066397e-05,
      "loss": 0.1311,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 16.02684783935547,
      "learning_rate": 3.177730687257639e-05,
      "loss": 0.139,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 13.188374519348145,
      "learning_rate": 3.0453941902507177e-05,
      "loss": 0.1163,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.007096346002072096,
      "learning_rate": 2.914654800091768e-05,
      "loss": 0.1489,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.02945764735341072,
      "learning_rate": 2.7856193438321986e-05,
      "loss": 0.1576,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 1.3139564990997314,
      "learning_rate": 2.6583932562403957e-05,
      "loss": 0.1117,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 4.071590900421143,
      "learning_rate": 2.5330804936510373e-05,
      "loss": 0.1233,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 11.481673240661621,
      "learning_rate": 2.409783449022475e-05,
      "loss": 0.1515,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 13.781807899475098,
      "learning_rate": 2.2886028682715217e-05,
      "loss": 0.1426,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 1.7268832921981812,
      "learning_rate": 2.169637767954048e-05,
      "loss": 0.1564,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9359840954274354,
      "eval_loss": 0.2996560335159302,
      "eval_runtime": 107.9382,
      "eval_samples_per_second": 23.3,
      "eval_steps_per_second": 1.464,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 7.562883377075195,
      "learning_rate": 2.052985354358622e-05,
      "loss": 0.1016,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 12.858866691589355,
      "learning_rate": 1.9387409440793386e-05,
      "loss": 0.1068,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.6010186076164246,
      "learning_rate": 1.82699788613271e-05,
      "loss": 0.1298,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 3.0228309631347656,
      "learning_rate": 1.7178474856822456e-05,
      "loss": 0.1028,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 1.0708427429199219,
      "learning_rate": 1.611378929433083e-05,
      "loss": 0.1333,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.3288826048374176,
      "learning_rate": 1.5076792127576073e-05,
      "loss": 0.0871,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 17.829463958740234,
      "learning_rate": 1.4068330686115943e-05,
      "loss": 0.1159,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.05485250800848007,
      "learning_rate": 1.308922898298977e-05,
      "loss": 0.1167,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.1920652985572815,
      "learning_rate": 1.2140287041418203e-05,
      "loss": 0.12,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.21363206207752228,
      "learning_rate": 1.1222280241104716e-05,
      "loss": 0.1011,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.3207741379737854,
      "learning_rate": 1.0335958684673574e-05,
      "loss": 0.1082,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9451292246520875,
      "eval_loss": 0.2916193902492523,
      "eval_runtime": 106.9736,
      "eval_samples_per_second": 23.51,
      "eval_steps_per_second": 1.477,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.5695350766181946,
      "learning_rate": 9.482046584761495e-06,
      "loss": 0.0866,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.629421591758728,
      "learning_rate": 8.661241672264192e-06,
      "loss": 0.1015,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.16660259664058685,
      "learning_rate": 7.874214626220899e-06,
      "loss": 0.0913,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 4.9127607345581055,
      "learning_rate": 7.1216085258031414e-06,
      "loss": 0.0762,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.2511623501777649,
      "learning_rate": 6.404038324855222e-06,
      "loss": 0.1046,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.004342870321124792,
      "learning_rate": 5.7220903494159316e-06,
      "loss": 0.0764,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 7.590545177459717,
      "learning_rate": 5.076321818632018e-06,
      "loss": 0.0868,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.1453438103199005,
      "learning_rate": 4.467260389454864e-06,
      "loss": 0.0799,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.7127562165260315,
      "learning_rate": 3.895403725492402e-06,
      "loss": 0.0994,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 8.983748435974121,
      "learning_rate": 3.3612190903686005e-06,
      "loss": 0.113,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 7.194179534912109,
      "learning_rate": 2.86514296592269e-06,
      "loss": 0.1146,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9431411530815109,
      "eval_loss": 0.2962559461593628,
      "eval_runtime": 107.8198,
      "eval_samples_per_second": 23.326,
      "eval_steps_per_second": 1.465,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 4.514409065246582,
      "learning_rate": 2.407580695560252e-06,
      "loss": 0.0801,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 0.0045217666774988174,
      "learning_rate": 1.9889061530473986e-06,
      "loss": 0.0669,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 0.0040066540241241455,
      "learning_rate": 1.6094614370188499e-06,
      "loss": 0.0772,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 3.6479694843292236,
      "learning_rate": 1.269556591449389e-06,
      "loss": 0.0743,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 0.03315883129835129,
      "learning_rate": 9.694693523171927e-07,
      "loss": 0.0791,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.5166642665863037,
      "learning_rate": 7.094449206659748e-07,
      "loss": 0.0764,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 1.8740558624267578,
      "learning_rate": 4.896957622514298e-07,
      "loss": 0.1069,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 9.233416557312012,
      "learning_rate": 3.104014339355921e-07,
      "loss": 0.0785,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 5.323319911956787,
      "learning_rate": 1.7170843697111304e-07,
      "loss": 0.1104,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 11.701455116271973,
      "learning_rate": 7.37300972951771e-08,
      "loss": 0.1072,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 0.15105102956295013,
      "learning_rate": 1.654647293098388e-08,
      "loss": 0.0801,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9439363817097416,
      "eval_loss": 0.2945658564567566,
      "eval_runtime": 107.8038,
      "eval_samples_per_second": 23.329,
      "eval_steps_per_second": 1.466,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.269900291632044,
      "train_runtime": 17210.4334,
      "train_samples_per_second": 10.215,
      "train_steps_per_second": 0.639
    }
  ],
  "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
}