{
  "best_metric": 0.23118199408054352,
  "best_model_checkpoint": "./convnext-base-wd1e-8-4e-5/checkpoint-7693",
  "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": 23.269968032836914,
      "learning_rate": 3.999182900797692e-05,
      "loss": 2.4371,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 30.385623931884766,
      "learning_rate": 3.996732270841873e-05,
      "loss": 1.2813,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 27.450511932373047,
      "learning_rate": 3.992650112540325e-05,
      "loss": 0.9817,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 23.956308364868164,
      "learning_rate": 3.986939761421341e-05,
      "loss": 0.8368,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 20.52164649963379,
      "learning_rate": 3.9796058834082656e-05,
      "loss": 0.77,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 12.296116828918457,
      "learning_rate": 3.970654471006971e-05,
      "loss": 0.661,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 17.72109031677246,
      "learning_rate": 3.960092838409391e-05,
      "loss": 0.7014,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 18.131221771240234,
      "learning_rate": 3.947929615517096e-05,
      "loss": 0.635,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 17.65522575378418,
      "learning_rate": 3.93417474088981e-05,
      "loss": 0.5691,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 12.841033935546875,
      "learning_rate": 3.9188394536246174e-05,
      "loss": 0.6271,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9053677932405566,
      "eval_loss": 0.34362277388572693,
      "eval_runtime": 106.9883,
      "eval_samples_per_second": 23.507,
      "eval_steps_per_second": 1.477,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 26.4544677734375,
      "learning_rate": 3.90193628417251e-05,
      "loss": 0.6271,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 13.971123695373535,
      "learning_rate": 3.883479044099763e-05,
      "loss": 0.5298,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 16.64786720275879,
      "learning_rate": 3.8634828148025175e-05,
      "loss": 0.478,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 22.18141746520996,
      "learning_rate": 3.841963935183782e-05,
      "loss": 0.4858,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 20.354249954223633,
      "learning_rate": 3.818939988302927e-05,
      "loss": 0.4578,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 18.752660751342773,
      "learning_rate": 3.7944297870085823e-05,
      "loss": 0.4527,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 36.6127815246582,
      "learning_rate": 3.768453358566675e-05,
      "loss": 0.4937,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 23.436214447021484,
      "learning_rate": 3.741031928296162e-05,
      "loss": 0.4916,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 27.608442306518555,
      "learning_rate": 3.712187902225845e-05,
      "loss": 0.4758,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 22.647695541381836,
      "learning_rate": 3.681944848786418e-05,
      "loss": 0.4312,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 14.084342002868652,
      "learning_rate": 3.6503274795527193e-05,
      "loss": 0.4528,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9256461232604374,
      "eval_loss": 0.27671584486961365,
      "eval_runtime": 108.0363,
      "eval_samples_per_second": 23.279,
      "eval_steps_per_second": 1.462,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 21.57204818725586,
      "learning_rate": 3.617361629051932e-05,
      "loss": 0.4516,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 21.228384017944336,
      "learning_rate": 3.5830742336542e-05,
      "loss": 0.3546,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 37.46253204345703,
      "learning_rate": 3.547493309562956e-05,
      "loss": 0.3906,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 17.27825355529785,
      "learning_rate": 3.51064792992289e-05,
      "loss": 0.3866,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 15.454079627990723,
      "learning_rate": 3.472568201064314e-05,
      "loss": 0.3813,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 14.477165222167969,
      "learning_rate": 3.433285237903305e-05,
      "loss": 0.3714,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 16.274826049804688,
      "learning_rate": 3.392831138517724e-05,
      "loss": 0.3561,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 11.944202423095703,
      "learning_rate": 3.3512389579199094e-05,
      "loss": 0.3941,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 10.119762420654297,
      "learning_rate": 3.308542681047451e-05,
      "loss": 0.3632,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 14.887975692749023,
      "learning_rate": 3.264777194994121e-05,
      "loss": 0.38,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 26.09409523010254,
      "learning_rate": 3.219978260503665e-05,
      "loss": 0.3762,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9268389662027833,
      "eval_loss": 0.25190117955207825,
      "eval_runtime": 107.1043,
      "eval_samples_per_second": 23.482,
      "eval_steps_per_second": 1.475,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.360994338989258,
      "learning_rate": 3.174182482749717e-05,
      "loss": 0.3766,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 20.89484977722168,
      "learning_rate": 3.12742728142575e-05,
      "loss": 0.3003,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 11.302007675170898,
      "learning_rate": 3.079750860169469e-05,
      "loss": 0.3166,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 11.336188316345215,
      "learning_rate": 3.031192175346651e-05,
      "loss": 0.3069,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 30.606346130371094,
      "learning_rate": 2.9817909042199313e-05,
      "loss": 0.2799,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 15.9932222366333,
      "learning_rate": 2.93158741252854e-05,
      "loss": 0.3054,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 12.633625984191895,
      "learning_rate": 2.8806227215054902e-05,
      "loss": 0.3178,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 33.243408203125,
      "learning_rate": 2.828938474359164e-05,
      "loss": 0.3152,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 18.574337005615234,
      "learning_rate": 2.7765769022466762e-05,
      "loss": 0.3448,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 4.619294166564941,
      "learning_rate": 2.723580789766831e-05,
      "loss": 0.3059,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 26.067737579345703,
      "learning_rate": 2.669993440000862e-05,
      "loss": 0.2979,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9371769383697813,
      "eval_loss": 0.2414184808731079,
      "eval_runtime": 107.4842,
      "eval_samples_per_second": 23.399,
      "eval_steps_per_second": 1.47,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 17.691198348999023,
      "learning_rate": 2.6158586391295163e-05,
      "loss": 0.3153,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 10.9254150390625,
      "learning_rate": 2.5612206206554034e-05,
      "loss": 0.2663,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 29.32948875427246,
      "learning_rate": 2.5061240292598336e-05,
      "loss": 0.2546,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 15.644346237182617,
      "learning_rate": 2.4506138843236867e-05,
      "loss": 0.2597,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 12.25552749633789,
      "learning_rate": 2.3947355431421094e-05,
      "loss": 0.2506,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 25.368473052978516,
      "learning_rate": 2.3385346638631082e-05,
      "loss": 0.2458,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 17.984514236450195,
      "learning_rate": 2.2820571681803108e-05,
      "loss": 0.2692,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 10.834085464477539,
      "learning_rate": 2.225349203810388e-05,
      "loss": 0.2347,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 23.141050338745117,
      "learning_rate": 2.16845710678579e-05,
      "loss": 0.2799,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 8.481077194213867,
      "learning_rate": 2.111427363593615e-05,
      "loss": 0.2713,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 1.8054014444351196,
      "learning_rate": 2.0543065731915316e-05,
      "loss": 0.2901,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9427435387673956,
      "eval_loss": 0.23892664909362793,
      "eval_runtime": 111.9885,
      "eval_samples_per_second": 22.458,
      "eval_steps_per_second": 1.411,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.089294910430908,
      "learning_rate": 1.997141408931814e-05,
      "loss": 0.2719,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 18.415925979614258,
      "learning_rate": 1.9399785804245775e-05,
      "loss": 0.2031,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 5.980634689331055,
      "learning_rate": 1.8828647953713982e-05,
      "loss": 0.2062,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 12.657254219055176,
      "learning_rate": 1.825846721400482e-05,
      "loss": 0.2194,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.39924386143684387,
      "learning_rate": 1.7689709479345895e-05,
      "loss": 0.1904,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 14.926392555236816,
      "learning_rate": 1.7122839481228503e-05,
      "loss": 0.2239,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 18.79764175415039,
      "learning_rate": 1.65583204086759e-05,
      "loss": 0.2309,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 8.083695411682129,
      "learning_rate": 1.599661352977198e-05,
      "loss": 0.2275,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 32.457725524902344,
      "learning_rate": 1.543817781475941e-05,
      "loss": 0.2326,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 15.148175239562988,
      "learning_rate": 1.4883469561015482e-05,
      "loss": 0.1995,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 18.512500762939453,
      "learning_rate": 1.433294202021182e-05,
      "loss": 0.2381,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.941948310139165,
      "eval_loss": 0.2407774031162262,
      "eval_runtime": 112.0327,
      "eval_samples_per_second": 22.449,
      "eval_steps_per_second": 1.41,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.42633116245269775,
      "learning_rate": 1.378704502796289e-05,
      "loss": 0.2012,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 22.5328426361084,
      "learning_rate": 1.3246224636265588e-05,
      "loss": 0.2241,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 12.483528137207031,
      "learning_rate": 1.2710922749030558e-05,
      "loss": 0.2037,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 1.0468084812164307,
      "learning_rate": 1.2181576761002872e-05,
      "loss": 0.1827,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 1.2360355854034424,
      "learning_rate": 1.1658619200367073e-05,
      "loss": 0.1708,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 17.188753128051758,
      "learning_rate": 1.1142477375328796e-05,
      "loss": 0.2357,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 1.7141854763031006,
      "learning_rate": 1.0633573024961584e-05,
      "loss": 0.1534,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 1.2022801637649536,
      "learning_rate": 1.013232197460415e-05,
      "loss": 0.1753,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 15.08975601196289,
      "learning_rate": 9.639133796089901e-06,
      "loss": 0.1601,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 29.775014877319336,
      "learning_rate": 9.154411473086087e-06,
      "loss": 0.1623,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 6.941558361053467,
      "learning_rate": 8.678551071816193e-06,
      "loss": 0.2084,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9463220675944334,
      "eval_loss": 0.23118199408054352,
      "eval_runtime": 111.9458,
      "eval_samples_per_second": 22.466,
      "eval_steps_per_second": 1.411,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 13.732728958129883,
      "learning_rate": 8.211941417434488e-06,
      "loss": 0.169,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.291477233171463,
      "learning_rate": 7.754963776317355e-06,
      "loss": 0.1618,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 4.71694278717041,
      "learning_rate": 7.307991544530839e-06,
      "loss": 0.1636,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.7449569702148438,
      "learning_rate": 6.871389942728983e-06,
      "loss": 0.1638,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 1.087550401687622,
      "learning_rate": 6.445515717732332e-06,
      "loss": 0.1695,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 14.644157409667969,
      "learning_rate": 6.03071685103043e-06,
      "loss": 0.1437,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 35.039981842041016,
      "learning_rate": 5.627332274446377e-06,
      "loss": 0.174,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 7.3837080001831055,
      "learning_rate": 5.2356915931959086e-06,
      "loss": 0.1604,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 10.73990249633789,
      "learning_rate": 4.856114816567281e-06,
      "loss": 0.1443,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 14.905359268188477,
      "learning_rate": 4.488912096441886e-06,
      "loss": 0.162,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.12942950427532196,
      "learning_rate": 4.13438347386943e-06,
      "loss": 0.1742,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9451292246520875,
      "eval_loss": 0.23588231205940247,
      "eval_runtime": 112.1261,
      "eval_samples_per_second": 22.43,
      "eval_steps_per_second": 1.409,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 17.91644859313965,
      "learning_rate": 3.7928186339045982e-06,
      "loss": 0.1865,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 2.5071592330932617,
      "learning_rate": 3.464496668905677e-06,
      "loss": 0.1669,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.10696269571781158,
      "learning_rate": 3.14968585048836e-06,
      "loss": 0.1535,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.128840371966362,
      "learning_rate": 2.848643410321257e-06,
      "loss": 0.1154,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 3.628100633621216,
      "learning_rate": 2.561615329942089e-06,
      "loss": 0.1564,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 11.36262321472168,
      "learning_rate": 2.288836139766373e-06,
      "loss": 0.1267,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 6.693118095397949,
      "learning_rate": 2.0305287274528073e-06,
      "loss": 0.119,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 10.11449146270752,
      "learning_rate": 1.7869041557819456e-06,
      "loss": 0.1749,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 20.39699363708496,
      "learning_rate": 1.558161490196961e-06,
      "loss": 0.1405,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 23.157207489013672,
      "learning_rate": 1.34448763614744e-06,
      "loss": 0.1765,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 13.510420799255371,
      "learning_rate": 1.1460571863690762e-06,
      "loss": 0.1582,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9479125248508946,
      "eval_loss": 0.23640932142734528,
      "eval_runtime": 111.9275,
      "eval_samples_per_second": 22.47,
      "eval_steps_per_second": 1.412,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 16.904611587524414,
      "learning_rate": 9.63032278224101e-07,
      "loss": 0.1588,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 0.36576932668685913,
      "learning_rate": 7.955624612189594e-07,
      "loss": 0.1708,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 26.957815170288086,
      "learning_rate": 6.4378457480754e-07,
      "loss": 0.1847,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 17.24859619140625,
      "learning_rate": 5.078226365797556e-07,
      "loss": 0.1401,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 10.3253755569458,
      "learning_rate": 3.877877409268771e-07,
      "loss": 0.1398,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.26469576358795166,
      "learning_rate": 2.837779682663899e-07,
      "loss": 0.1309,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 7.175375461578369,
      "learning_rate": 1.958783049005719e-07,
      "loss": 0.1637,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 0.38347867131233215,
      "learning_rate": 1.2416057357423684e-07,
      "loss": 0.1888,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 11.241958618164062,
      "learning_rate": 6.868337478844523e-08,
      "loss": 0.156,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 19.296527862548828,
      "learning_rate": 2.9492038918070844e-08,
      "loss": 0.135,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 14.173076629638672,
      "learning_rate": 6.618589172393553e-09,
      "loss": 0.1451,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9495029821073558,
      "eval_loss": 0.2357349991798401,
      "eval_runtime": 111.9428,
      "eval_samples_per_second": 22.467,
      "eval_steps_per_second": 1.411,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.3212610531113601,
      "train_runtime": 19604.4461,
      "train_samples_per_second": 8.968,
      "train_steps_per_second": 0.561
    }
  ],
  "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
}