{
  "best_metric": 0.25531044602394104,
  "best_model_checkpoint": "./convnext-tiny-new-1e-4/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": 24.848655700683594,
      "learning_rate": 9.99795725199423e-05,
      "loss": 2.511,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 21.852506637573242,
      "learning_rate": 9.991830677104683e-05,
      "loss": 1.4511,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 14.456836700439453,
      "learning_rate": 9.981625281350813e-05,
      "loss": 1.1871,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 25.593276977539062,
      "learning_rate": 9.967349403553353e-05,
      "loss": 1.0097,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 21.799734115600586,
      "learning_rate": 9.949014708520663e-05,
      "loss": 0.9728,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 22.431013107299805,
      "learning_rate": 9.926636177517427e-05,
      "loss": 0.9024,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 28.28536033630371,
      "learning_rate": 9.900232096023477e-05,
      "loss": 0.8912,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 27.407716751098633,
      "learning_rate": 9.869824038792741e-05,
      "loss": 0.865,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 13.469520568847656,
      "learning_rate": 9.835436852224525e-05,
      "loss": 0.7768,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 17.95415496826172,
      "learning_rate": 9.797098634061542e-05,
      "loss": 0.8432,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.862027833001988,
      "eval_loss": 0.4732176661491394,
      "eval_runtime": 70.5844,
      "eval_samples_per_second": 35.631,
      "eval_steps_per_second": 2.238,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 28.058094024658203,
      "learning_rate": 9.754840710431274e-05,
      "loss": 0.827,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.800704002380371,
      "learning_rate": 9.708697610249406e-05,
      "loss": 0.733,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 17.161815643310547,
      "learning_rate": 9.658707037006294e-05,
      "loss": 0.7117,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 15.4861478805542,
      "learning_rate": 9.604909837959455e-05,
      "loss": 0.6609,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 20.579856872558594,
      "learning_rate": 9.547349970757317e-05,
      "loss": 0.6396,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 24.0612735748291,
      "learning_rate": 9.486074467521456e-05,
      "loss": 0.6321,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 23.076522827148438,
      "learning_rate": 9.421133396416686e-05,
      "loss": 0.7177,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 24.146862030029297,
      "learning_rate": 9.352579820740405e-05,
      "loss": 0.7033,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 13.88412857055664,
      "learning_rate": 9.280469755564613e-05,
      "loss": 0.637,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 18.419206619262695,
      "learning_rate": 9.204862121966044e-05,
      "loss": 0.6056,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 22.241830825805664,
      "learning_rate": 9.125818698881798e-05,
      "loss": 0.6603,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9009940357852882,
      "eval_loss": 0.35555583238601685,
      "eval_runtime": 70.2432,
      "eval_samples_per_second": 35.804,
      "eval_steps_per_second": 2.249,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 28.679054260253906,
      "learning_rate": 9.043404072629829e-05,
      "loss": 0.6342,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 25.620079040527344,
      "learning_rate": 8.957685584135502e-05,
      "loss": 0.5285,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 17.019142150878906,
      "learning_rate": 8.86873327390739e-05,
      "loss": 0.5882,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 15.56048583984375,
      "learning_rate": 8.776619824807224e-05,
      "loss": 0.545,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 10.068634033203125,
      "learning_rate": 8.681420502660786e-05,
      "loss": 0.5944,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 15.93220043182373,
      "learning_rate": 8.583213094758261e-05,
      "loss": 0.5378,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 16.840496063232422,
      "learning_rate": 8.482077846294308e-05,
      "loss": 0.5597,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 23.36423683166504,
      "learning_rate": 8.378097394799773e-05,
      "loss": 0.5476,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 14.018542289733887,
      "learning_rate": 8.271356702618626e-05,
      "loss": 0.5429,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 22.620033264160156,
      "learning_rate": 8.161942987485303e-05,
      "loss": 0.5361,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 29.02153778076172,
      "learning_rate": 8.049945651259163e-05,
      "loss": 0.487,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.882703777335984,
      "eval_loss": 0.4106706380844116,
      "eval_runtime": 69.7812,
      "eval_samples_per_second": 36.041,
      "eval_steps_per_second": 2.264,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 9.186837196350098,
      "learning_rate": 7.935456206874292e-05,
      "loss": 0.5448,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 10.694817543029785,
      "learning_rate": 7.818568203564374e-05,
      "loss": 0.4207,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 9.862343788146973,
      "learning_rate": 7.699377150423672e-05,
      "loss": 0.475,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 17.465476989746094,
      "learning_rate": 7.577980438366628e-05,
      "loss": 0.48,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 19.368242263793945,
      "learning_rate": 7.454477260549828e-05,
      "loss": 0.4126,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 24.577472686767578,
      "learning_rate": 7.32896853132135e-05,
      "loss": 0.4992,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 7.475000381469727,
      "learning_rate": 7.201556803763725e-05,
      "loss": 0.4931,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 13.887510299682617,
      "learning_rate": 7.07234618589791e-05,
      "loss": 0.4909,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 20.115652084350586,
      "learning_rate": 6.94144225561669e-05,
      "loss": 0.4814,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 14.20095443725586,
      "learning_rate": 6.808951974417078e-05,
      "loss": 0.457,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 29.256624221801758,
      "learning_rate": 6.674983600002155e-05,
      "loss": 0.4369,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9172962226640159,
      "eval_loss": 0.3118807375431061,
      "eval_runtime": 70.2783,
      "eval_samples_per_second": 35.786,
      "eval_steps_per_second": 2.248,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 10.440736770629883,
      "learning_rate": 6.539646597823791e-05,
      "loss": 0.4036,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 14.221583366394043,
      "learning_rate": 6.403051551638508e-05,
      "loss": 0.3902,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 18.483789443969727,
      "learning_rate": 6.265310073149584e-05,
      "loss": 0.4096,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 27.6778564453125,
      "learning_rate": 6.126534710809216e-05,
      "loss": 0.3778,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 18.118946075439453,
      "learning_rate": 5.9868388578552734e-05,
      "loss": 0.356,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 27.71598243713379,
      "learning_rate": 5.8463366596577706e-05,
      "loss": 0.3544,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 10.941313743591309,
      "learning_rate": 5.705142920450777e-05,
      "loss": 0.4005,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 20.2288818359375,
      "learning_rate": 5.5633730095259695e-05,
      "loss": 0.4007,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 13.51459789276123,
      "learning_rate": 5.421142766964474e-05,
      "loss": 0.411,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 10.38302993774414,
      "learning_rate": 5.278568408984037e-05,
      "loss": 0.4178,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 7.628846168518066,
      "learning_rate": 5.135766432978829e-05,
      "loss": 0.3791,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9196819085487078,
      "eval_loss": 0.3082761764526367,
      "eval_runtime": 71.5388,
      "eval_samples_per_second": 35.156,
      "eval_steps_per_second": 2.209,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 15.286186218261719,
      "learning_rate": 4.9928535223295344e-05,
      "loss": 0.3914,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 19.06636619567871,
      "learning_rate": 4.849946451061443e-05,
      "loss": 0.3268,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 16.880006790161133,
      "learning_rate": 4.707161988428495e-05,
      "loss": 0.3184,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 10.349203109741211,
      "learning_rate": 4.564616803501205e-05,
      "loss": 0.3417,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 17.491863250732422,
      "learning_rate": 4.4224273698364735e-05,
      "loss": 0.3218,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 9.212013244628906,
      "learning_rate": 4.2807098703071255e-05,
      "loss": 0.3143,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 19.431245803833008,
      "learning_rate": 4.1395801021689746e-05,
      "loss": 0.3262,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 16.24408531188965,
      "learning_rate": 3.999153382442995e-05,
      "loss": 0.2885,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 23.65041732788086,
      "learning_rate": 3.859544453689853e-05,
      "loss": 0.3555,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 10.8707275390625,
      "learning_rate": 3.7208673902538706e-05,
      "loss": 0.3404,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 18.250511169433594,
      "learning_rate": 3.583235505052955e-05,
      "loss": 0.3342,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9296222664015904,
      "eval_loss": 0.28032296895980835,
      "eval_runtime": 79.2042,
      "eval_samples_per_second": 31.753,
      "eval_steps_per_second": 1.995,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 6.232498645782471,
      "learning_rate": 3.446761256990723e-05,
      "loss": 0.3301,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 19.711963653564453,
      "learning_rate": 3.311556159066397e-05,
      "loss": 0.2638,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 11.230116844177246,
      "learning_rate": 3.177730687257639e-05,
      "loss": 0.312,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 9.200843811035156,
      "learning_rate": 3.0453941902507177e-05,
      "loss": 0.3046,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 18.336387634277344,
      "learning_rate": 2.914654800091768e-05,
      "loss": 0.2505,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 4.870161533355713,
      "learning_rate": 2.7856193438321986e-05,
      "loss": 0.3136,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 8.88219165802002,
      "learning_rate": 2.6583932562403957e-05,
      "loss": 0.2656,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 12.868867874145508,
      "learning_rate": 2.5330804936510373e-05,
      "loss": 0.2714,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 26.35186767578125,
      "learning_rate": 2.409783449022475e-05,
      "loss": 0.2675,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 13.712077140808105,
      "learning_rate": 2.2886028682715217e-05,
      "loss": 0.2664,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.8319320678710938,
      "learning_rate": 2.169637767954048e-05,
      "loss": 0.2803,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9339960238568589,
      "eval_loss": 0.26732343435287476,
      "eval_runtime": 70.0882,
      "eval_samples_per_second": 35.883,
      "eval_steps_per_second": 2.254,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 15.552393913269043,
      "learning_rate": 2.052985354358622e-05,
      "loss": 0.2109,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 15.977888107299805,
      "learning_rate": 1.9387409440793386e-05,
      "loss": 0.2315,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 12.987915992736816,
      "learning_rate": 1.82699788613271e-05,
      "loss": 0.2421,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 7.147055625915527,
      "learning_rate": 1.7178474856822456e-05,
      "loss": 0.2454,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 25.91172981262207,
      "learning_rate": 1.611378929433083e-05,
      "loss": 0.2504,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 12.569241523742676,
      "learning_rate": 1.5076792127576073e-05,
      "loss": 0.2338,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 19.739892959594727,
      "learning_rate": 1.4068330686115943e-05,
      "loss": 0.2161,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 9.218408584594727,
      "learning_rate": 1.308922898298977e-05,
      "loss": 0.2329,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 2.0763704776763916,
      "learning_rate": 1.2140287041418203e-05,
      "loss": 0.2085,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 11.455533981323242,
      "learning_rate": 1.1222280241104716e-05,
      "loss": 0.2561,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 3.408818006515503,
      "learning_rate": 1.0335958684673574e-05,
      "loss": 0.2118,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9320079522862823,
      "eval_loss": 0.2637202739715576,
      "eval_runtime": 74.3067,
      "eval_samples_per_second": 33.846,
      "eval_steps_per_second": 2.126,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 10.710466384887695,
      "learning_rate": 9.482046584761495e-06,
      "loss": 0.2045,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 14.385254859924316,
      "learning_rate": 8.661241672264192e-06,
      "loss": 0.2176,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 10.243017196655273,
      "learning_rate": 7.874214626220899e-06,
      "loss": 0.2099,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 6.018984317779541,
      "learning_rate": 7.1216085258031414e-06,
      "loss": 0.1686,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 1.289069652557373,
      "learning_rate": 6.404038324855222e-06,
      "loss": 0.2214,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 2.6374850273132324,
      "learning_rate": 5.7220903494159316e-06,
      "loss": 0.1675,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 2.8265738487243652,
      "learning_rate": 5.076321818632018e-06,
      "loss": 0.2129,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 11.41331958770752,
      "learning_rate": 4.467260389454864e-06,
      "loss": 0.2097,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 22.140159606933594,
      "learning_rate": 3.895403725492402e-06,
      "loss": 0.2099,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 7.952567100524902,
      "learning_rate": 3.3612190903686005e-06,
      "loss": 0.2438,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 19.06670379638672,
      "learning_rate": 2.86514296592269e-06,
      "loss": 0.1706,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9347912524850894,
      "eval_loss": 0.25745028257369995,
      "eval_runtime": 74.1249,
      "eval_samples_per_second": 33.929,
      "eval_steps_per_second": 2.132,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 16.285367965698242,
      "learning_rate": 2.407580695560252e-06,
      "loss": 0.1872,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 5.482550621032715,
      "learning_rate": 1.9889061530473986e-06,
      "loss": 0.1863,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 8.077670097351074,
      "learning_rate": 1.6094614370188499e-06,
      "loss": 0.1759,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 14.164857864379883,
      "learning_rate": 1.269556591449389e-06,
      "loss": 0.1768,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 16.639026641845703,
      "learning_rate": 9.694693523171927e-07,
      "loss": 0.2211,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 13.088668823242188,
      "learning_rate": 7.094449206659748e-07,
      "loss": 0.1891,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 7.857825756072998,
      "learning_rate": 4.896957622514298e-07,
      "loss": 0.2362,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 16.95376968383789,
      "learning_rate": 3.104014339355921e-07,
      "loss": 0.2201,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 11.055394172668457,
      "learning_rate": 1.7170843697111304e-07,
      "loss": 0.2139,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 8.469569206237793,
      "learning_rate": 7.37300972951771e-08,
      "loss": 0.2026,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 18.598886489868164,
      "learning_rate": 1.654647293098388e-08,
      "loss": 0.1995,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9375745526838967,
      "eval_loss": 0.25531044602394104,
      "eval_runtime": 73.7798,
      "eval_samples_per_second": 34.088,
      "eval_steps_per_second": 2.142,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 1.301428412334932e+19,
      "train_loss": 0.4412916103637251,
      "train_runtime": 10918.1842,
      "train_samples_per_second": 16.102,
      "train_steps_per_second": 1.007
    }
  ],
  "logging_steps": 100,
  "max_steps": 10990,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1.301428412334932e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}