{
  "best_metric": 0.23055487871170044,
  "best_model_checkpoint": "./convnext-base-wd1e-8-4e-5-erasing/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": 21.871763229370117,
      "learning_rate": 3.999182900797692e-05,
      "loss": 2.4789,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 17.83153533935547,
      "learning_rate": 3.996732270841873e-05,
      "loss": 1.3163,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 19.49209213256836,
      "learning_rate": 3.992650112540325e-05,
      "loss": 1.0315,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 33.785457611083984,
      "learning_rate": 3.986939761421341e-05,
      "loss": 0.8006,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 22.69280242919922,
      "learning_rate": 3.9796058834082656e-05,
      "loss": 0.7755,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 17.21114158630371,
      "learning_rate": 3.970654471006971e-05,
      "loss": 0.6705,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 25.925142288208008,
      "learning_rate": 3.960092838409391e-05,
      "loss": 0.7039,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.736983299255371,
      "learning_rate": 3.947929615517096e-05,
      "loss": 0.6441,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 13.611920356750488,
      "learning_rate": 3.93417474088981e-05,
      "loss": 0.6039,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.253599166870117,
      "learning_rate": 3.9188394536246174e-05,
      "loss": 0.6036,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8930417495029821,
      "eval_loss": 0.3736543655395508,
      "eval_runtime": 106.8745,
      "eval_samples_per_second": 23.532,
      "eval_steps_per_second": 1.478,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 34.44277572631836,
      "learning_rate": 3.90193628417251e-05,
      "loss": 0.624,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 14.051901817321777,
      "learning_rate": 3.883479044099763e-05,
      "loss": 0.5383,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 23.078445434570312,
      "learning_rate": 3.8634828148025175e-05,
      "loss": 0.507,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 19.300426483154297,
      "learning_rate": 3.841963935183782e-05,
      "loss": 0.4953,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 22.54229736328125,
      "learning_rate": 3.818939988302927e-05,
      "loss": 0.4969,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 17.774635314941406,
      "learning_rate": 3.7944297870085823e-05,
      "loss": 0.4826,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 13.029126167297363,
      "learning_rate": 3.768453358566675e-05,
      "loss": 0.5326,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 22.84644317626953,
      "learning_rate": 3.741031928296162e-05,
      "loss": 0.4899,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 17.86623764038086,
      "learning_rate": 3.712187902225845e-05,
      "loss": 0.4475,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 15.729595184326172,
      "learning_rate": 3.681944848786418e-05,
      "loss": 0.4841,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 15.516929626464844,
      "learning_rate": 3.6503274795527193e-05,
      "loss": 0.4997,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9248508946322067,
      "eval_loss": 0.27299922704696655,
      "eval_runtime": 105.5397,
      "eval_samples_per_second": 23.83,
      "eval_steps_per_second": 1.497,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 32.8221321105957,
      "learning_rate": 3.617361629051932e-05,
      "loss": 0.4712,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 20.870311737060547,
      "learning_rate": 3.5830742336542e-05,
      "loss": 0.3474,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 22.007932662963867,
      "learning_rate": 3.547493309562956e-05,
      "loss": 0.434,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 11.817858695983887,
      "learning_rate": 3.51064792992289e-05,
      "loss": 0.3748,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 15.62509536743164,
      "learning_rate": 3.472568201064314e-05,
      "loss": 0.4473,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 29.18074607849121,
      "learning_rate": 3.433285237903305e-05,
      "loss": 0.384,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 3.7826242446899414,
      "learning_rate": 3.392831138517724e-05,
      "loss": 0.4136,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 37.68365478515625,
      "learning_rate": 3.3512389579199094e-05,
      "loss": 0.3916,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 16.420135498046875,
      "learning_rate": 3.308542681047451e-05,
      "loss": 0.4228,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 21.540740966796875,
      "learning_rate": 3.264777194994121e-05,
      "loss": 0.401,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 17.597993850708008,
      "learning_rate": 3.219978260503665e-05,
      "loss": 0.3663,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9260437375745527,
      "eval_loss": 0.26662740111351013,
      "eval_runtime": 105.4604,
      "eval_samples_per_second": 23.848,
      "eval_steps_per_second": 1.498,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 7.028054237365723,
      "learning_rate": 3.174182482749717e-05,
      "loss": 0.4076,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 6.525599479675293,
      "learning_rate": 3.12742728142575e-05,
      "loss": 0.3146,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 14.01247787475586,
      "learning_rate": 3.079750860169469e-05,
      "loss": 0.3465,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 23.31292152404785,
      "learning_rate": 3.031192175346651e-05,
      "loss": 0.3685,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 20.722320556640625,
      "learning_rate": 2.9817909042199313e-05,
      "loss": 0.3136,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 17.478958129882812,
      "learning_rate": 2.93158741252854e-05,
      "loss": 0.3403,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 9.703248023986816,
      "learning_rate": 2.8806227215054902e-05,
      "loss": 0.3432,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 11.908836364746094,
      "learning_rate": 2.828938474359164e-05,
      "loss": 0.3695,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 22.338346481323242,
      "learning_rate": 2.7765769022466762e-05,
      "loss": 0.3569,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 3.3628461360931396,
      "learning_rate": 2.723580789766831e-05,
      "loss": 0.3324,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 17.76543617248535,
      "learning_rate": 2.669993440000862e-05,
      "loss": 0.3176,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9391650099403579,
      "eval_loss": 0.2379751056432724,
      "eval_runtime": 105.9425,
      "eval_samples_per_second": 23.739,
      "eval_steps_per_second": 1.491,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 13.366058349609375,
      "learning_rate": 2.6158586391295163e-05,
      "loss": 0.2818,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 10.28676700592041,
      "learning_rate": 2.5612206206554034e-05,
      "loss": 0.2891,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 15.261099815368652,
      "learning_rate": 2.5061240292598336e-05,
      "loss": 0.3196,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 23.628677368164062,
      "learning_rate": 2.4506138843236867e-05,
      "loss": 0.2921,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 28.432357788085938,
      "learning_rate": 2.3947355431421094e-05,
      "loss": 0.2641,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 13.766435623168945,
      "learning_rate": 2.3385346638631082e-05,
      "loss": 0.2599,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 17.775606155395508,
      "learning_rate": 2.2820571681803108e-05,
      "loss": 0.2917,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 34.89625549316406,
      "learning_rate": 2.225349203810388e-05,
      "loss": 0.2867,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 20.649389266967773,
      "learning_rate": 2.16845710678579e-05,
      "loss": 0.3066,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 11.82371711730957,
      "learning_rate": 2.111427363593615e-05,
      "loss": 0.302,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 18.50792694091797,
      "learning_rate": 2.0543065731915316e-05,
      "loss": 0.305,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9371769383697813,
      "eval_loss": 0.24580277502536774,
      "eval_runtime": 106.0764,
      "eval_samples_per_second": 23.709,
      "eval_steps_per_second": 1.489,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 17.39256477355957,
      "learning_rate": 1.997141408931814e-05,
      "loss": 0.2981,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 15.757665634155273,
      "learning_rate": 1.9399785804245775e-05,
      "loss": 0.237,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 13.927785873413086,
      "learning_rate": 1.8828647953713982e-05,
      "loss": 0.2281,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 27.117650985717773,
      "learning_rate": 1.825846721400482e-05,
      "loss": 0.2586,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 29.655193328857422,
      "learning_rate": 1.7689709479345895e-05,
      "loss": 0.2595,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 12.67216968536377,
      "learning_rate": 1.7122839481228503e-05,
      "loss": 0.2449,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 15.060705184936523,
      "learning_rate": 1.65583204086759e-05,
      "loss": 0.2558,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 35.11772918701172,
      "learning_rate": 1.599661352977198e-05,
      "loss": 0.2148,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 13.670873641967773,
      "learning_rate": 1.543817781475941e-05,
      "loss": 0.2593,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 16.197233200073242,
      "learning_rate": 1.4883469561015482e-05,
      "loss": 0.2482,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 26.83169937133789,
      "learning_rate": 1.433294202021182e-05,
      "loss": 0.2489,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.945924453280318,
      "eval_loss": 0.23894770443439484,
      "eval_runtime": 105.5878,
      "eval_samples_per_second": 23.819,
      "eval_steps_per_second": 1.496,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 15.461714744567871,
      "learning_rate": 1.378704502796289e-05,
      "loss": 0.2724,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 13.98656177520752,
      "learning_rate": 1.3246224636265588e-05,
      "loss": 0.2111,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 11.368219375610352,
      "learning_rate": 1.2710922749030558e-05,
      "loss": 0.2428,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 12.15278434753418,
      "learning_rate": 1.2181576761002872e-05,
      "loss": 0.2478,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 2.1350884437561035,
      "learning_rate": 1.1658619200367073e-05,
      "loss": 0.2005,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.2970704138278961,
      "learning_rate": 1.1142477375328796e-05,
      "loss": 0.2369,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 11.124874114990234,
      "learning_rate": 1.0633573024961584e-05,
      "loss": 0.2222,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 12.277088165283203,
      "learning_rate": 1.013232197460415e-05,
      "loss": 0.2092,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 26.41246223449707,
      "learning_rate": 9.639133796089901e-06,
      "loss": 0.2214,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 4.2534918785095215,
      "learning_rate": 9.154411473086087e-06,
      "loss": 0.1903,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.8372716307640076,
      "learning_rate": 8.678551071816193e-06,
      "loss": 0.2356,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9451292246520875,
      "eval_loss": 0.24236075580120087,
      "eval_runtime": 105.4122,
      "eval_samples_per_second": 23.859,
      "eval_steps_per_second": 1.499,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 21.71571159362793,
      "learning_rate": 8.211941417434488e-06,
      "loss": 0.1892,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 35.81491470336914,
      "learning_rate": 7.754963776317355e-06,
      "loss": 0.2233,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 14.517807960510254,
      "learning_rate": 7.307991544530839e-06,
      "loss": 0.184,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 4.057854652404785,
      "learning_rate": 6.871389942728983e-06,
      "loss": 0.2041,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 9.720903396606445,
      "learning_rate": 6.445515717732332e-06,
      "loss": 0.1968,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 8.159006118774414,
      "learning_rate": 6.03071685103043e-06,
      "loss": 0.204,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 7.216451168060303,
      "learning_rate": 5.627332274446377e-06,
      "loss": 0.1778,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 16.071630477905273,
      "learning_rate": 5.2356915931959086e-06,
      "loss": 0.2121,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 2.6936397552490234,
      "learning_rate": 4.856114816567281e-06,
      "loss": 0.1652,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 13.142534255981445,
      "learning_rate": 4.488912096441886e-06,
      "loss": 0.1989,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 1.0733413696289062,
      "learning_rate": 4.13438347386943e-06,
      "loss": 0.1678,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9439363817097416,
      "eval_loss": 0.23690353333950043,
      "eval_runtime": 105.711,
      "eval_samples_per_second": 23.791,
      "eval_steps_per_second": 1.495,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 17.58012580871582,
      "learning_rate": 3.7928186339045982e-06,
      "loss": 0.1947,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 20.622407913208008,
      "learning_rate": 3.464496668905677e-06,
      "loss": 0.1857,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.7151262164115906,
      "learning_rate": 3.14968585048836e-06,
      "loss": 0.1732,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 8.048576354980469,
      "learning_rate": 2.848643410321257e-06,
      "loss": 0.1318,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.751136064529419,
      "learning_rate": 2.561615329942089e-06,
      "loss": 0.1989,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 3.61799693107605,
      "learning_rate": 2.288836139766373e-06,
      "loss": 0.1495,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 12.693122863769531,
      "learning_rate": 2.0305287274528073e-06,
      "loss": 0.1762,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 9.484522819519043,
      "learning_rate": 1.7869041557819456e-06,
      "loss": 0.1923,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 24.0538330078125,
      "learning_rate": 1.558161490196961e-06,
      "loss": 0.1779,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 1.2370848655700684,
      "learning_rate": 1.34448763614744e-06,
      "loss": 0.2347,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 18.4951229095459,
      "learning_rate": 1.1460571863690762e-06,
      "loss": 0.1565,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9499005964214712,
      "eval_loss": 0.23227240145206451,
      "eval_runtime": 105.2417,
      "eval_samples_per_second": 23.897,
      "eval_steps_per_second": 1.501,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 12.973793983459473,
      "learning_rate": 9.63032278224101e-07,
      "loss": 0.1495,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 7.522297382354736,
      "learning_rate": 7.955624612189594e-07,
      "loss": 0.1573,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 15.810741424560547,
      "learning_rate": 6.4378457480754e-07,
      "loss": 0.1692,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 15.419889450073242,
      "learning_rate": 5.078226365797556e-07,
      "loss": 0.1387,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 10.454876899719238,
      "learning_rate": 3.877877409268771e-07,
      "loss": 0.1696,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.05530254915356636,
      "learning_rate": 2.837779682663899e-07,
      "loss": 0.1417,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 14.749571800231934,
      "learning_rate": 1.958783049005719e-07,
      "loss": 0.1799,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 32.72957992553711,
      "learning_rate": 1.2416057357423684e-07,
      "loss": 0.1893,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 26.80739974975586,
      "learning_rate": 6.868337478844523e-08,
      "loss": 0.1796,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 9.608071327209473,
      "learning_rate": 2.9492038918070844e-08,
      "loss": 0.178,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 13.889327049255371,
      "learning_rate": 6.618589172393553e-09,
      "loss": 0.1643,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9502982107355865,
      "eval_loss": 0.23055487871170044,
      "eval_runtime": 105.2963,
      "eval_samples_per_second": 23.885,
      "eval_steps_per_second": 1.501,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.34581250545650966,
      "train_runtime": 18969.8729,
      "train_samples_per_second": 9.268,
      "train_steps_per_second": 0.579
    }
  ],
  "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
}