{
  "best_metric": 0.22072996199131012,
  "best_model_checkpoint": "./convnext-base-wd1e-8-2e-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": 18.85420036315918,
      "learning_rate": 1.999591450398846e-05,
      "loss": 2.8171,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 28.546306610107422,
      "learning_rate": 1.9983661354209365e-05,
      "loss": 1.7712,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 28.488494873046875,
      "learning_rate": 1.9963250562701624e-05,
      "loss": 1.3111,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 27.985355377197266,
      "learning_rate": 1.9934698807106706e-05,
      "loss": 1.024,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 20.874189376831055,
      "learning_rate": 1.9898029417041328e-05,
      "loss": 0.9096,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 20.992918014526367,
      "learning_rate": 1.9853272355034854e-05,
      "loss": 0.7615,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 20.678836822509766,
      "learning_rate": 1.9800464192046956e-05,
      "loss": 0.7798,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 20.038610458374023,
      "learning_rate": 1.973964807758548e-05,
      "loss": 0.6927,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 23.2023868560791,
      "learning_rate": 1.967087370444905e-05,
      "loss": 0.6137,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 19.863248825073242,
      "learning_rate": 1.9594197268123087e-05,
      "loss": 0.6642,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8946322067594433,
      "eval_loss": 0.38488638401031494,
      "eval_runtime": 107.4706,
      "eval_samples_per_second": 23.402,
      "eval_steps_per_second": 1.47,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 62.42153549194336,
      "learning_rate": 1.950968142086255e-05,
      "loss": 0.6664,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.604940414428711,
      "learning_rate": 1.9417395220498815e-05,
      "loss": 0.5709,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 18.934337615966797,
      "learning_rate": 1.9317414074012588e-05,
      "loss": 0.4944,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 20.69709014892578,
      "learning_rate": 1.920981967591891e-05,
      "loss": 0.5209,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 33.886619567871094,
      "learning_rate": 1.9094699941514634e-05,
      "loss": 0.4758,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 24.116291046142578,
      "learning_rate": 1.8972148935042912e-05,
      "loss": 0.4422,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 27.108684539794922,
      "learning_rate": 1.8842266792833374e-05,
      "loss": 0.5052,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 37.22698211669922,
      "learning_rate": 1.870515964148081e-05,
      "loss": 0.5054,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 20.40926170349121,
      "learning_rate": 1.8560939511129225e-05,
      "loss": 0.4867,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 27.881574630737305,
      "learning_rate": 1.840972424393209e-05,
      "loss": 0.4333,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 14.44858455657959,
      "learning_rate": 1.8251637397763597e-05,
      "loss": 0.4658,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9165009940357853,
      "eval_loss": 0.2898237407207489,
      "eval_runtime": 108.8832,
      "eval_samples_per_second": 23.098,
      "eval_steps_per_second": 1.451,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 23.801347732543945,
      "learning_rate": 1.808680814525966e-05,
      "loss": 0.4546,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 20.20176887512207,
      "learning_rate": 1.7915371168271e-05,
      "loss": 0.3855,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 35.17919921875,
      "learning_rate": 1.773746654781478e-05,
      "loss": 0.4138,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 19.36662483215332,
      "learning_rate": 1.755323964961445e-05,
      "loss": 0.3929,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 21.20980453491211,
      "learning_rate": 1.736284100532157e-05,
      "loss": 0.3887,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 17.784828186035156,
      "learning_rate": 1.7166426189516524e-05,
      "loss": 0.3988,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 15.224771499633789,
      "learning_rate": 1.696415569258862e-05,
      "loss": 0.3724,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 32.63229751586914,
      "learning_rate": 1.6756194789599547e-05,
      "loss": 0.406,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 17.206344604492188,
      "learning_rate": 1.6542713405237254e-05,
      "loss": 0.3689,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 27.567245483398438,
      "learning_rate": 1.6323885974970606e-05,
      "loss": 0.3841,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 41.00786590576172,
      "learning_rate": 1.6099891302518326e-05,
      "loss": 0.3675,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9304174950298211,
      "eval_loss": 0.24960491061210632,
      "eval_runtime": 106.8163,
      "eval_samples_per_second": 23.545,
      "eval_steps_per_second": 1.479,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 23.188814163208008,
      "learning_rate": 1.5870912413748585e-05,
      "loss": 0.3788,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 30.5658016204834,
      "learning_rate": 1.563713640712875e-05,
      "loss": 0.3049,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 15.625422477722168,
      "learning_rate": 1.5398754300847346e-05,
      "loss": 0.3273,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 8.019862174987793,
      "learning_rate": 1.5155960876733255e-05,
      "loss": 0.3523,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 25.53868865966797,
      "learning_rate": 1.4908954521099656e-05,
      "loss": 0.3003,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 13.111063003540039,
      "learning_rate": 1.46579370626427e-05,
      "loss": 0.3354,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 17.303651809692383,
      "learning_rate": 1.4403113607527451e-05,
      "loss": 0.3388,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 33.83345413208008,
      "learning_rate": 1.414469237179582e-05,
      "loss": 0.3382,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 28.170040130615234,
      "learning_rate": 1.3882884511233381e-05,
      "loss": 0.3522,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 2.074608564376831,
      "learning_rate": 1.3617903948834155e-05,
      "loss": 0.3237,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 15.362191200256348,
      "learning_rate": 1.334996720000431e-05,
      "loss": 0.3174,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9411530815109344,
      "eval_loss": 0.2325742095708847,
      "eval_runtime": 108.4122,
      "eval_samples_per_second": 23.198,
      "eval_steps_per_second": 1.457,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 19.770036697387695,
      "learning_rate": 1.3079293195647582e-05,
      "loss": 0.3204,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 16.560823440551758,
      "learning_rate": 1.2806103103277017e-05,
      "loss": 0.2899,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 25.254541397094727,
      "learning_rate": 1.2530620146299168e-05,
      "loss": 0.2782,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 20.078277587890625,
      "learning_rate": 1.2253069421618434e-05,
      "loss": 0.2908,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 9.352935791015625,
      "learning_rate": 1.1973677715710547e-05,
      "loss": 0.2804,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 23.662412643432617,
      "learning_rate": 1.1692673319315541e-05,
      "loss": 0.2707,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 27.34366798400879,
      "learning_rate": 1.1410285840901554e-05,
      "loss": 0.2717,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 19.72267723083496,
      "learning_rate": 1.112674601905194e-05,
      "loss": 0.2604,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 37.95018768310547,
      "learning_rate": 1.084228553392895e-05,
      "loss": 0.3012,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 16.562606811523438,
      "learning_rate": 1.0557136817968075e-05,
      "loss": 0.2882,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 2.053151845932007,
      "learning_rate": 1.0271532865957658e-05,
      "loss": 0.3106,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9435387673956263,
      "eval_loss": 0.2301262468099594,
      "eval_runtime": 107.4067,
      "eval_samples_per_second": 23.416,
      "eval_steps_per_second": 1.471,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.48727560043335,
      "learning_rate": 9.98570704465907e-06,
      "loss": 0.2853,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 19.252519607543945,
      "learning_rate": 9.699892902122887e-06,
      "loss": 0.2327,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 6.616147518157959,
      "learning_rate": 9.414323976856991e-06,
      "loss": 0.2412,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 24.550338745117188,
      "learning_rate": 9.12923360700241e-06,
      "loss": 0.2438,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.5674369931221008,
      "learning_rate": 8.844854739672947e-06,
      "loss": 0.2175,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 17.81463050842285,
      "learning_rate": 8.561419740614251e-06,
      "loss": 0.2442,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 20.672290802001953,
      "learning_rate": 8.27916020433795e-06,
      "loss": 0.2423,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 9.923518180847168,
      "learning_rate": 7.99830676488599e-06,
      "loss": 0.2537,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 41.107269287109375,
      "learning_rate": 7.719088907379705e-06,
      "loss": 0.2555,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 24.59263038635254,
      "learning_rate": 7.441734780507741e-06,
      "loss": 0.2303,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 15.823193550109863,
      "learning_rate": 7.16647101010591e-06,
      "loss": 0.2678,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9431411530815109,
      "eval_loss": 0.23032429814338684,
      "eval_runtime": 107.9379,
      "eval_samples_per_second": 23.3,
      "eval_steps_per_second": 1.464,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 2.0771257877349854,
      "learning_rate": 6.893522513981445e-06,
      "loss": 0.2379,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 28.53963279724121,
      "learning_rate": 6.623112318132794e-06,
      "loss": 0.2651,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 18.732324600219727,
      "learning_rate": 6.355461374515279e-06,
      "loss": 0.2332,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 1.8658305406570435,
      "learning_rate": 6.090788380501436e-06,
      "loss": 0.2196,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 10.409640312194824,
      "learning_rate": 5.829309600183536e-06,
      "loss": 0.2048,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 13.740955352783203,
      "learning_rate": 5.571238687664398e-06,
      "loss": 0.2776,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 11.248075485229492,
      "learning_rate": 5.316786512480792e-06,
      "loss": 0.1773,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 4.135138511657715,
      "learning_rate": 5.066160987302075e-06,
      "loss": 0.2077,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 27.468730926513672,
      "learning_rate": 4.819566898044951e-06,
      "loss": 0.1922,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 8.280457496643066,
      "learning_rate": 4.5772057365430435e-06,
      "loss": 0.1982,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 15.988367080688477,
      "learning_rate": 4.339275535908096e-06,
      "loss": 0.2503,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9427435387673956,
      "eval_loss": 0.22976447641849518,
      "eval_runtime": 107.3436,
      "eval_samples_per_second": 23.429,
      "eval_steps_per_second": 1.472,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 22.717308044433594,
      "learning_rate": 4.105970708717244e-06,
      "loss": 0.2103,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.365151584148407,
      "learning_rate": 3.877481888158677e-06,
      "loss": 0.2023,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 11.08658504486084,
      "learning_rate": 3.6539957722654195e-06,
      "loss": 0.1878,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 2.89919114112854,
      "learning_rate": 3.4356949713644915e-06,
      "loss": 0.2046,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 1.526982069015503,
      "learning_rate": 3.222757858866166e-06,
      "loss": 0.2073,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 23.598388671875,
      "learning_rate": 3.015358425515215e-06,
      "loss": 0.1774,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 41.74139404296875,
      "learning_rate": 2.8136661372231887e-06,
      "loss": 0.201,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 8.005784034729004,
      "learning_rate": 2.6178457965979543e-06,
      "loss": 0.2074,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 14.387943267822266,
      "learning_rate": 2.4280574082836406e-06,
      "loss": 0.1943,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 20.383352279663086,
      "learning_rate": 2.244456048220943e-06,
      "loss": 0.1917,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.6621111631393433,
      "learning_rate": 2.067191736934715e-06,
      "loss": 0.2204,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.945924453280318,
      "eval_loss": 0.22159598767757416,
      "eval_runtime": 108.3116,
      "eval_samples_per_second": 23.22,
      "eval_steps_per_second": 1.459,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 18.94173240661621,
      "learning_rate": 1.8964093169522991e-06,
      "loss": 0.2453,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 26.928571701049805,
      "learning_rate": 1.7322483344528385e-06,
      "loss": 0.2176,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 1.4977689981460571,
      "learning_rate": 1.57484292524418e-06,
      "loss": 0.1887,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.4014904797077179,
      "learning_rate": 1.4243217051606285e-06,
      "loss": 0.1675,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 2.714075803756714,
      "learning_rate": 1.2808076649710444e-06,
      "loss": 0.1964,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 3.391688823699951,
      "learning_rate": 1.1444180698831864e-06,
      "loss": 0.1636,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 7.2056145668029785,
      "learning_rate": 1.0152643637264036e-06,
      "loss": 0.1679,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 6.323690414428711,
      "learning_rate": 8.934520778909728e-07,
      "loss": 0.215,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 10.003203392028809,
      "learning_rate": 7.790807450984805e-07,
      "loss": 0.169,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.6530151963233948,
      "learning_rate": 6.7224381807372e-07,
      "loss": 0.2152,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 1.6361758708953857,
      "learning_rate": 5.730285931845381e-07,
      "loss": 0.2013,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9463220675944334,
      "eval_loss": 0.22236517071723938,
      "eval_runtime": 107.9608,
      "eval_samples_per_second": 23.295,
      "eval_steps_per_second": 1.463,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 30.111351013183594,
      "learning_rate": 4.815161391120505e-07,
      "loss": 0.1982,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 2.49858021736145,
      "learning_rate": 3.977812306094797e-07,
      "loss": 0.2057,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 35.6863899230957,
      "learning_rate": 3.2189228740377e-07,
      "loss": 0.2417,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 20.32947540283203,
      "learning_rate": 2.539113182898778e-07,
      "loss": 0.1596,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 16.59613609313965,
      "learning_rate": 1.9389387046343855e-07,
      "loss": 0.1952,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 3.6704962253570557,
      "learning_rate": 1.4188898413319495e-07,
      "loss": 0.1823,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 7.785726070404053,
      "learning_rate": 9.793915245028595e-08,
      "loss": 0.2017,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 2.0824966430664062,
      "learning_rate": 6.208028678711842e-08,
      "loss": 0.2237,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 17.7388858795166,
      "learning_rate": 3.4341687394222614e-08,
      "loss": 0.1938,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 31.48530387878418,
      "learning_rate": 1.4746019459035422e-08,
      "loss": 0.1843,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 21.949007034301758,
      "learning_rate": 3.3092945861967764e-09,
      "loss": 0.1808,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9467196819085487,
      "eval_loss": 0.22072996199131012,
      "eval_runtime": 107.7576,
      "eval_samples_per_second": 23.339,
      "eval_steps_per_second": 1.466,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.36333368386432624,
      "train_runtime": 19367.8979,
      "train_samples_per_second": 9.077,
      "train_steps_per_second": 0.567
    }
  ],
  "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
}