{
  "best_metric": 0.25579527020454407,
  "best_model_checkpoint": "./convnext-base-5e-5/checkpoint-4396",
  "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": 19.689619064331055,
      "learning_rate": 4.998978625997115e-05,
      "loss": 2.1359,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 24.598817825317383,
      "learning_rate": 4.995915338552341e-05,
      "loss": 1.0316,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 15.473122596740723,
      "learning_rate": 4.9908126406754066e-05,
      "loss": 0.783,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 34.29795837402344,
      "learning_rate": 4.9836747017766765e-05,
      "loss": 0.6182,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 24.46330451965332,
      "learning_rate": 4.9745073542603314e-05,
      "loss": 0.6141,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 20.168657302856445,
      "learning_rate": 4.963318088758713e-05,
      "loss": 0.542,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 15.94365119934082,
      "learning_rate": 4.950116048011739e-05,
      "loss": 0.5752,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 24.878143310546875,
      "learning_rate": 4.934912019396371e-05,
      "loss": 0.5303,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 12.92821216583252,
      "learning_rate": 4.917718426112262e-05,
      "loss": 0.4849,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.624749183654785,
      "learning_rate": 4.898549317030771e-05,
      "loss": 0.4987,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.88389662027833,
      "eval_loss": 0.38692283630371094,
      "eval_runtime": 106.9096,
      "eval_samples_per_second": 23.525,
      "eval_steps_per_second": 1.478,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 19.525230407714844,
      "learning_rate": 4.877420355215637e-05,
      "loss": 0.5385,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 5.739523410797119,
      "learning_rate": 4.854348805124703e-05,
      "loss": 0.3948,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 12.253121376037598,
      "learning_rate": 4.829353518503147e-05,
      "loss": 0.4011,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.450470924377441,
      "learning_rate": 4.8024549189797276e-05,
      "loss": 0.3934,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.626289367675781,
      "learning_rate": 4.7736749853786585e-05,
      "loss": 0.3618,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 18.44389533996582,
      "learning_rate": 4.743037233760728e-05,
      "loss": 0.3532,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 4.797609329223633,
      "learning_rate": 4.710566698208343e-05,
      "loss": 0.4287,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 16.017484664916992,
      "learning_rate": 4.676289910370202e-05,
      "loss": 0.3859,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.441079139709473,
      "learning_rate": 4.6402348777823063e-05,
      "loss": 0.3775,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 17.051143646240234,
      "learning_rate": 4.602431060983022e-05,
      "loss": 0.3309,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.007339477539062,
      "learning_rate": 4.562909349440899e-05,
      "loss": 0.3757,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9025844930417495,
      "eval_loss": 0.32946208119392395,
      "eval_runtime": 106.2955,
      "eval_samples_per_second": 23.66,
      "eval_steps_per_second": 1.486,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 15.104771614074707,
      "learning_rate": 4.5217020363149146e-05,
      "loss": 0.3766,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 14.0053071975708,
      "learning_rate": 4.478842792067751e-05,
      "loss": 0.3104,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 12.075423240661621,
      "learning_rate": 4.434366636953695e-05,
      "loss": 0.3068,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 8.154679298400879,
      "learning_rate": 4.388309912403612e-05,
      "loss": 0.2741,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 10.335779190063477,
      "learning_rate": 4.340710251330393e-05,
      "loss": 0.3255,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 10.061939239501953,
      "learning_rate": 4.2916065473791306e-05,
      "loss": 0.3065,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.7983992099761963,
      "learning_rate": 4.241038923147154e-05,
      "loss": 0.2864,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 12.939589500427246,
      "learning_rate": 4.1890486973998866e-05,
      "loss": 0.3275,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 2.2934823036193848,
      "learning_rate": 4.135678351309313e-05,
      "loss": 0.2704,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 14.341142654418945,
      "learning_rate": 4.0809714937426516e-05,
      "loss": 0.2824,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 18.908397674560547,
      "learning_rate": 4.024972825629582e-05,
      "loss": 0.2985,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9192842942345925,
      "eval_loss": 0.304299920797348,
      "eval_runtime": 106.6765,
      "eval_samples_per_second": 23.576,
      "eval_steps_per_second": 1.481,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 12.50020980834961,
      "learning_rate": 3.967728103437146e-05,
      "loss": 0.3084,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 5.194071292877197,
      "learning_rate": 3.909284101782187e-05,
      "loss": 0.2271,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.1949557065963745,
      "learning_rate": 3.849688575211836e-05,
      "loss": 0.2341,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 14.773088455200195,
      "learning_rate": 3.788990219183314e-05,
      "loss": 0.2542,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 8.208048820495605,
      "learning_rate": 3.727238630274914e-05,
      "loss": 0.2377,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 21.311338424682617,
      "learning_rate": 3.664484265660675e-05,
      "loss": 0.1974,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 24.28902816772461,
      "learning_rate": 3.6007784018818626e-05,
      "loss": 0.2736,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 6.190667629241943,
      "learning_rate": 3.536173092948955e-05,
      "loss": 0.2429,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 2.843653440475464,
      "learning_rate": 3.470721127808345e-05,
      "loss": 0.2423,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.1003132313489914,
      "learning_rate": 3.404475987208539e-05,
      "loss": 0.2132,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 6.872644424438477,
      "learning_rate": 3.3374918000010775e-05,
      "loss": 0.2535,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9320079522862823,
      "eval_loss": 0.25579527020454407,
      "eval_runtime": 106.3392,
      "eval_samples_per_second": 23.651,
      "eval_steps_per_second": 1.486,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 10.436262130737305,
      "learning_rate": 3.2698232989118956e-05,
      "loss": 0.2236,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 7.982916355133057,
      "learning_rate": 3.201525775819254e-05,
      "loss": 0.1939,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 4.680846214294434,
      "learning_rate": 3.132655036574792e-05,
      "loss": 0.1997,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 21.136667251586914,
      "learning_rate": 3.063267355404608e-05,
      "loss": 0.2021,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.10948444157838821,
      "learning_rate": 2.9934194289276367e-05,
      "loss": 0.1751,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 11.024763107299805,
      "learning_rate": 2.9231683298288853e-05,
      "loss": 0.1907,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 20.557159423828125,
      "learning_rate": 2.8525714602253885e-05,
      "loss": 0.1568,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.9859702587127686,
      "learning_rate": 2.7816865047629847e-05,
      "loss": 0.1751,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 5.471407890319824,
      "learning_rate": 2.710571383482237e-05,
      "loss": 0.1992,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 23.69923210144043,
      "learning_rate": 2.6392842044920184e-05,
      "loss": 0.2,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 25.619653701782227,
      "learning_rate": 2.5678832164894147e-05,
      "loss": 0.1951,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.931610337972167,
      "eval_loss": 0.2921680510044098,
      "eval_runtime": 106.4218,
      "eval_samples_per_second": 23.632,
      "eval_steps_per_second": 1.485,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.4416472911834717,
      "learning_rate": 2.4964267611647672e-05,
      "loss": 0.1827,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 2.3818318843841553,
      "learning_rate": 2.4249732255307216e-05,
      "loss": 0.1469,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 3.7632319927215576,
      "learning_rate": 2.3535809942142476e-05,
      "loss": 0.1482,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 7.223693370819092,
      "learning_rate": 2.2823084017506024e-05,
      "loss": 0.1487,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.026045968756079674,
      "learning_rate": 2.2112136849182368e-05,
      "loss": 0.1466,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.5752942562103271,
      "learning_rate": 2.1403549351535627e-05,
      "loss": 0.1331,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 10.485978126525879,
      "learning_rate": 2.0697900510844873e-05,
      "loss": 0.1633,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 1.757461667060852,
      "learning_rate": 1.9995766912214975e-05,
      "loss": 0.1351,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 11.111984252929688,
      "learning_rate": 1.9297722268449264e-05,
      "loss": 0.1856,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 2.5022082328796387,
      "learning_rate": 1.8604336951269353e-05,
      "loss": 0.1587,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 28.14973258972168,
      "learning_rate": 1.7916177525264775e-05,
      "loss": 0.1573,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9375745526838967,
      "eval_loss": 0.2673531472682953,
      "eval_runtime": 105.5648,
      "eval_samples_per_second": 23.824,
      "eval_steps_per_second": 1.497,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 12.279279708862305,
      "learning_rate": 1.7233806284953614e-05,
      "loss": 0.16,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 21.84239959716797,
      "learning_rate": 1.6557780795331985e-05,
      "loss": 0.1364,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 9.68105697631836,
      "learning_rate": 1.5888653436288196e-05,
      "loss": 0.1201,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 18.502185821533203,
      "learning_rate": 1.5226970951253588e-05,
      "loss": 0.1195,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.032358527183532715,
      "learning_rate": 1.457327400045884e-05,
      "loss": 0.1145,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.008106480352580547,
      "learning_rate": 1.3928096719160993e-05,
      "loss": 0.1555,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.5467287302017212,
      "learning_rate": 1.3291966281201978e-05,
      "loss": 0.1143,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.529351532459259,
      "learning_rate": 1.2665402468255186e-05,
      "loss": 0.1222,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 24.214012145996094,
      "learning_rate": 1.2048917245112376e-05,
      "loss": 0.1411,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 16.45935821533203,
      "learning_rate": 1.1443014341357609e-05,
      "loss": 0.1166,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 13.261152267456055,
      "learning_rate": 1.084818883977024e-05,
      "loss": 0.1432,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9415506958250497,
      "eval_loss": 0.28571948409080505,
      "eval_runtime": 105.6389,
      "eval_samples_per_second": 23.808,
      "eval_steps_per_second": 1.496,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 26.173646926879883,
      "learning_rate": 1.026492677179311e-05,
      "loss": 0.1028,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 10.5139799118042,
      "learning_rate": 9.693704720396693e-06,
      "loss": 0.0937,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.7591624855995178,
      "learning_rate": 9.13498943066355e-06,
      "loss": 0.1398,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.3186585009098053,
      "learning_rate": 8.589237428411228e-06,
      "loss": 0.099,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 1.547601342201233,
      "learning_rate": 8.056894647165414e-06,
      "loss": 0.1252,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.27220067381858826,
      "learning_rate": 7.538396063788037e-06,
      "loss": 0.0982,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 10.86794662475586,
      "learning_rate": 7.0341653430579715e-06,
      "loss": 0.0997,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.1966288536787033,
      "learning_rate": 6.544614491494885e-06,
      "loss": 0.1128,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.6555065512657166,
      "learning_rate": 6.070143520709101e-06,
      "loss": 0.1243,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 1.6866192817687988,
      "learning_rate": 5.611140120552358e-06,
      "loss": 0.0909,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 15.261884689331055,
      "learning_rate": 5.167979342336787e-06,
      "loss": 0.1087,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9447316103379721,
      "eval_loss": 0.2808384597301483,
      "eval_runtime": 106.5038,
      "eval_samples_per_second": 23.614,
      "eval_steps_per_second": 1.484,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.6104148030281067,
      "learning_rate": 4.7410232923807475e-06,
      "loss": 0.0853,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 6.473132133483887,
      "learning_rate": 4.330620836132096e-06,
      "loss": 0.1124,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 1.1223357915878296,
      "learning_rate": 3.937107313110449e-06,
      "loss": 0.0953,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 17.79993438720703,
      "learning_rate": 3.5608042629015707e-06,
      "loss": 0.0812,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.5151718854904175,
      "learning_rate": 3.202019162427611e-06,
      "loss": 0.11,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.01641092263162136,
      "learning_rate": 2.8610451747079658e-06,
      "loss": 0.0731,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 9.250214576721191,
      "learning_rate": 2.538160909316009e-06,
      "loss": 0.0917,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 8.201703071594238,
      "learning_rate": 2.233630194727432e-06,
      "loss": 0.0934,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.14482106268405914,
      "learning_rate": 1.947701862746201e-06,
      "loss": 0.1005,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 9.274855613708496,
      "learning_rate": 1.6806095451843002e-06,
      "loss": 0.1125,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 3.39308500289917,
      "learning_rate": 1.432571482961345e-06,
      "loss": 0.1043,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9475149105367793,
      "eval_loss": 0.27623292803764343,
      "eval_runtime": 106.365,
      "eval_samples_per_second": 23.645,
      "eval_steps_per_second": 1.485,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 10.197725296020508,
      "learning_rate": 1.203790347780126e-06,
      "loss": 0.0978,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 0.014358041808009148,
      "learning_rate": 9.944530765236993e-07,
      "loss": 0.074,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 0.04214104637503624,
      "learning_rate": 8.047307185094249e-07,
      "loss": 0.0886,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.1321563571691513,
      "learning_rate": 6.347782957246945e-07,
      "loss": 0.0955,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 0.13576732575893402,
      "learning_rate": 4.847346761585963e-07,
      "loss": 0.0866,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.12991665303707123,
      "learning_rate": 3.547224603329874e-07,
      "loss": 0.0682,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 0.03759434446692467,
      "learning_rate": 2.448478811257149e-07,
      "loss": 0.1133,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 13.521113395690918,
      "learning_rate": 1.5520071696779604e-07,
      "loss": 0.094,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 10.216411590576172,
      "learning_rate": 8.585421848555652e-08,
      "loss": 0.1004,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 1.9031656980514526,
      "learning_rate": 3.686504864758855e-08,
      "loss": 0.0947,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 7.584500789642334,
      "learning_rate": 8.27323646549194e-09,
      "loss": 0.0757,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.94831013916501,
      "eval_loss": 0.2754751145839691,
      "eval_runtime": 105.8006,
      "eval_samples_per_second": 23.771,
      "eval_steps_per_second": 1.493,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.24376181263181707,
      "train_runtime": 17060.3388,
      "train_samples_per_second": 10.305,
      "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
}