{
  "best_metric": 0.22113871574401855,
  "best_model_checkpoint": "./convnext-base-wd-1e-8-3e-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": 23.46108627319336,
      "learning_rate": 2.999387175598269e-05,
      "loss": 2.6353,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 16.22706413269043,
      "learning_rate": 2.9975492031314045e-05,
      "loss": 1.4961,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 21.055408477783203,
      "learning_rate": 2.994487584405244e-05,
      "loss": 1.1313,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 33.7674446105957,
      "learning_rate": 2.990204821066006e-05,
      "loss": 0.8546,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 20.148395538330078,
      "learning_rate": 2.984704412556199e-05,
      "loss": 0.8015,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 19.732004165649414,
      "learning_rate": 2.977990853255228e-05,
      "loss": 0.691,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 25.433279037475586,
      "learning_rate": 2.970069628807043e-05,
      "loss": 0.7105,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 14.037881851196289,
      "learning_rate": 2.9609472116378222e-05,
      "loss": 0.64,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.451700210571289,
      "learning_rate": 2.9506310556673573e-05,
      "loss": 0.5995,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 12.546747207641602,
      "learning_rate": 2.9391295902184625e-05,
      "loss": 0.6062,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8930417495029821,
      "eval_loss": 0.36765438318252563,
      "eval_runtime": 105.6433,
      "eval_samples_per_second": 23.807,
      "eval_steps_per_second": 1.496,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 34.490543365478516,
      "learning_rate": 2.9264522131293818e-05,
      "loss": 0.6367,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 15.90696907043457,
      "learning_rate": 2.9126092830748217e-05,
      "loss": 0.5309,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 28.304794311523438,
      "learning_rate": 2.897612111101888e-05,
      "loss": 0.515,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 18.61859893798828,
      "learning_rate": 2.8814729513878365e-05,
      "loss": 0.484,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 19.281923294067383,
      "learning_rate": 2.864204991227195e-05,
      "loss": 0.5082,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 20.169174194335938,
      "learning_rate": 2.8458223402564366e-05,
      "loss": 0.4689,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 13.506836891174316,
      "learning_rate": 2.826340018925006e-05,
      "loss": 0.5311,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 22.1657657623291,
      "learning_rate": 2.8057739462221215e-05,
      "loss": 0.4909,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 20.638866424560547,
      "learning_rate": 2.7841409266693838e-05,
      "loss": 0.4316,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 15.684589385986328,
      "learning_rate": 2.761458636589813e-05,
      "loss": 0.4746,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 17.279956817626953,
      "learning_rate": 2.7377456096645395e-05,
      "loss": 0.4975,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9268389662027833,
      "eval_loss": 0.27913880348205566,
      "eval_runtime": 105.6176,
      "eval_samples_per_second": 23.812,
      "eval_steps_per_second": 1.496,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 29.95267105102539,
      "learning_rate": 2.7130212217889484e-05,
      "loss": 0.4716,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 14.320356369018555,
      "learning_rate": 2.6873056752406504e-05,
      "loss": 0.3472,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 24.206567764282227,
      "learning_rate": 2.6606199821722166e-05,
      "loss": 0.4262,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 16.749975204467773,
      "learning_rate": 2.632985947442167e-05,
      "loss": 0.3718,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 13.791542053222656,
      "learning_rate": 2.6044261507982356e-05,
      "loss": 0.4423,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 39.41408157348633,
      "learning_rate": 2.5749639284274782e-05,
      "loss": 0.3756,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 5.441478252410889,
      "learning_rate": 2.5446233538882924e-05,
      "loss": 0.3997,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 27.294628143310547,
      "learning_rate": 2.513429218439932e-05,
      "loss": 0.3874,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 18.142757415771484,
      "learning_rate": 2.4814070107855878e-05,
      "loss": 0.4291,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 18.32405662536621,
      "learning_rate": 2.448582896245591e-05,
      "loss": 0.4085,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 27.914020538330078,
      "learning_rate": 2.4149836953777488e-05,
      "loss": 0.3777,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9355864811133201,
      "eval_loss": 0.2496386617422104,
      "eval_runtime": 105.6667,
      "eval_samples_per_second": 23.801,
      "eval_steps_per_second": 1.495,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 11.372156143188477,
      "learning_rate": 2.3806368620622876e-05,
      "loss": 0.3952,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 6.642704963684082,
      "learning_rate": 2.345570461069312e-05,
      "loss": 0.3029,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 16.000930786132812,
      "learning_rate": 2.3098131451271016e-05,
      "loss": 0.3644,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 27.844514846801758,
      "learning_rate": 2.2733941315099883e-05,
      "loss": 0.3687,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 27.35566520690918,
      "learning_rate": 2.2363431781649483e-05,
      "loss": 0.3111,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 13.689902305603027,
      "learning_rate": 2.1986905593964048e-05,
      "loss": 0.3542,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 6.648231029510498,
      "learning_rate": 2.1604670411291174e-05,
      "loss": 0.3471,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 17.46345329284668,
      "learning_rate": 2.121703855769373e-05,
      "loss": 0.3704,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 13.02889347076416,
      "learning_rate": 2.0824326766850072e-05,
      "loss": 0.3491,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 3.550931453704834,
      "learning_rate": 2.042685592325123e-05,
      "loss": 0.33,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 52.093753814697266,
      "learning_rate": 2.0024950800006463e-05,
      "loss": 0.3135,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9395626242544731,
      "eval_loss": 0.22969551384449005,
      "eval_runtime": 105.7529,
      "eval_samples_per_second": 23.782,
      "eval_steps_per_second": 1.494,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.075785636901855,
      "learning_rate": 1.961893979347137e-05,
      "loss": 0.2703,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 16.759511947631836,
      "learning_rate": 1.9209154654915524e-05,
      "loss": 0.2964,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 34.864501953125,
      "learning_rate": 1.879593021944875e-05,
      "loss": 0.3281,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 31.949420928955078,
      "learning_rate": 1.837960413242765e-05,
      "loss": 0.2933,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 18.939651489257812,
      "learning_rate": 1.796051657356582e-05,
      "loss": 0.2837,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 10.814738273620605,
      "learning_rate": 1.7539009978973312e-05,
      "loss": 0.2615,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 19.958898544311523,
      "learning_rate": 1.711542876135233e-05,
      "loss": 0.2938,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 24.847991943359375,
      "learning_rate": 1.669011902857791e-05,
      "loss": 0.2847,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 31.440824508666992,
      "learning_rate": 1.6263428300893422e-05,
      "loss": 0.311,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 24.64202117919922,
      "learning_rate": 1.5835705226952112e-05,
      "loss": 0.319,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 20.617746353149414,
      "learning_rate": 1.540729929893649e-05,
      "loss": 0.3022,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9399602385685885,
      "eval_loss": 0.24197304248809814,
      "eval_runtime": 105.732,
      "eval_samples_per_second": 23.787,
      "eval_steps_per_second": 1.494,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 13.16674518585205,
      "learning_rate": 1.4978560566988603e-05,
      "loss": 0.3003,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 11.548135757446289,
      "learning_rate": 1.454983935318433e-05,
      "loss": 0.2383,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 15.562095642089844,
      "learning_rate": 1.4121485965285485e-05,
      "loss": 0.235,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 22.393648147583008,
      "learning_rate": 1.3693850410503614e-05,
      "loss": 0.2661,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 21.64914894104004,
      "learning_rate": 1.326728210950942e-05,
      "loss": 0.2697,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 7.229253768920898,
      "learning_rate": 1.2842129610921378e-05,
      "loss": 0.2556,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 17.345558166503906,
      "learning_rate": 1.2418740306506923e-05,
      "loss": 0.2717,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 32.90108871459961,
      "learning_rate": 1.1997460147328984e-05,
      "loss": 0.2346,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 26.533710479736328,
      "learning_rate": 1.1578633361069559e-05,
      "loss": 0.2724,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 16.73810577392578,
      "learning_rate": 1.1162602170761611e-05,
      "loss": 0.2518,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 31.530790328979492,
      "learning_rate": 1.0749706515158863e-05,
      "loss": 0.2481,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.945924453280318,
      "eval_loss": 0.2327275574207306,
      "eval_runtime": 105.9698,
      "eval_samples_per_second": 23.733,
      "eval_steps_per_second": 1.491,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 17.024036407470703,
      "learning_rate": 1.0340283770972167e-05,
      "loss": 0.2747,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 14.112195014953613,
      "learning_rate": 9.93466847719919e-06,
      "loss": 0.2236,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 7.770081043243408,
      "learning_rate": 9.533192061772919e-06,
      "loss": 0.2391,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 7.291871547698975,
      "learning_rate": 9.136182570752153e-06,
      "loss": 0.2501,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 9.997650146484375,
      "learning_rate": 8.743964400275304e-06,
      "loss": 0.2116,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.30319592356681824,
      "learning_rate": 8.356858031496596e-06,
      "loss": 0.2563,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 1.384082555770874,
      "learning_rate": 7.975179768721187e-06,
      "loss": 0.2203,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 17.061349868774414,
      "learning_rate": 7.599241480953112e-06,
      "loss": 0.2054,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 33.148921966552734,
      "learning_rate": 7.229350347067426e-06,
      "loss": 0.2328,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 8.242575645446777,
      "learning_rate": 6.865808604814564e-06,
      "loss": 0.2108,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.4348295032978058,
      "learning_rate": 6.508913303862144e-06,
      "loss": 0.2439,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9439363817097416,
      "eval_loss": 0.23284506797790527,
      "eval_runtime": 105.5247,
      "eval_samples_per_second": 23.833,
      "eval_steps_per_second": 1.497,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 30.08086585998535,
      "learning_rate": 6.1589560630758656e-06,
      "loss": 0.1879,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 41.7615852355957,
      "learning_rate": 5.8162228322380155e-06,
      "loss": 0.2135,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 8.607312202453613,
      "learning_rate": 5.480993658398129e-06,
      "loss": 0.1963,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 10.731322288513184,
      "learning_rate": 5.153542457046737e-06,
      "loss": 0.2423,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 10.43770980834961,
      "learning_rate": 4.834136788299248e-06,
      "loss": 0.2129,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 12.690656661987305,
      "learning_rate": 4.523037638272822e-06,
      "loss": 0.2149,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 18.027700424194336,
      "learning_rate": 4.220499205834783e-06,
      "loss": 0.185,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 15.369023323059082,
      "learning_rate": 3.926768694896931e-06,
      "loss": 0.22,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 44.863441467285156,
      "learning_rate": 3.6420861124254607e-06,
      "loss": 0.1803,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 9.732752799987793,
      "learning_rate": 3.3666840723314145e-06,
      "loss": 0.2207,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 1.730449914932251,
      "learning_rate": 3.1007876054020724e-06,
      "loss": 0.1849,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.94831013916501,
      "eval_loss": 0.2235245704650879,
      "eval_runtime": 105.4928,
      "eval_samples_per_second": 23.84,
      "eval_steps_per_second": 1.498,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 21.83309555053711,
      "learning_rate": 2.8446139754284486e-06,
      "loss": 0.2095,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 21.83153533935547,
      "learning_rate": 2.5983725016792574e-06,
      "loss": 0.196,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 1.2116672992706299,
      "learning_rate": 2.36226438786627e-06,
      "loss": 0.1769,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 12.81109619140625,
      "learning_rate": 2.1364825577409424e-06,
      "loss": 0.1455,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.36004820466041565,
      "learning_rate": 1.9212114974565664e-06,
      "loss": 0.2078,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 2.3525044918060303,
      "learning_rate": 1.7166271048247796e-06,
      "loss": 0.1607,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 15.3230619430542,
      "learning_rate": 1.5228965455896054e-06,
      "loss": 0.1831,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 14.115558624267578,
      "learning_rate": 1.3401781168364591e-06,
      "loss": 0.2013,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 2.3320865631103516,
      "learning_rate": 1.1686211176477208e-06,
      "loss": 0.1922,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 1.680296778678894,
      "learning_rate": 1.00836572711058e-06,
      "loss": 0.2407,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 12.502205848693848,
      "learning_rate": 8.595428897768071e-07,
      "loss": 0.1716,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9506958250497017,
      "eval_loss": 0.22239501774311066,
      "eval_runtime": 105.4214,
      "eval_samples_per_second": 23.857,
      "eval_steps_per_second": 1.499,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 19.13417625427246,
      "learning_rate": 7.222742086680756e-07,
      "loss": 0.1656,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 6.075860023498535,
      "learning_rate": 5.966718459142196e-07,
      "loss": 0.1705,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 17.40691375732422,
      "learning_rate": 4.82838431105655e-07,
      "loss": 0.1847,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 21.60211944580078,
      "learning_rate": 3.808669774348167e-07,
      "loss": 0.1602,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 11.399230003356934,
      "learning_rate": 2.908408056951578e-07,
      "loss": 0.1834,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.1577068567276001,
      "learning_rate": 2.1283347619979243e-07,
      "loss": 0.1565,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 5.944873332977295,
      "learning_rate": 1.4690872867542892e-07,
      "loss": 0.1948,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 35.5730094909668,
      "learning_rate": 9.312043018067762e-08,
      "loss": 0.2041,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 31.919750213623047,
      "learning_rate": 5.1512531091333914e-08,
      "loss": 0.1906,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 15.108580589294434,
      "learning_rate": 2.211902918855313e-08,
      "loss": 0.1973,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 29.040531158447266,
      "learning_rate": 4.963941879295164e-09,
      "loss": 0.1731,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9506958250497017,
      "eval_loss": 0.22113871574401855,
      "eval_runtime": 105.6004,
      "eval_samples_per_second": 23.816,
      "eval_steps_per_second": 1.496,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.35613980254224475,
      "train_runtime": 19061.1748,
      "train_samples_per_second": 9.223,
      "train_steps_per_second": 0.577
    }
  ],
  "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
}