{
  "best_metric": 0.22431252896785736,
  "best_model_checkpoint": "./convnext-base-3e-5-wd-1e-8-raug/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": 21.20834732055664,
      "learning_rate": 2.999387175598269e-05,
      "loss": 2.5983,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 33.59589767456055,
      "learning_rate": 2.9975492031314045e-05,
      "loss": 1.4493,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 23.01809310913086,
      "learning_rate": 2.994487584405244e-05,
      "loss": 1.0707,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 22.695598602294922,
      "learning_rate": 2.990204821066006e-05,
      "loss": 0.875,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 19.60612678527832,
      "learning_rate": 2.984704412556199e-05,
      "loss": 0.8055,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 16.237865447998047,
      "learning_rate": 2.977990853255228e-05,
      "loss": 0.6791,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 20.869752883911133,
      "learning_rate": 2.970069628807043e-05,
      "loss": 0.7147,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 25.231393814086914,
      "learning_rate": 2.9609472116378222e-05,
      "loss": 0.6515,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 20.106616973876953,
      "learning_rate": 2.9506310556673573e-05,
      "loss": 0.5761,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 21.04306983947754,
      "learning_rate": 2.9391295902184625e-05,
      "loss": 0.6237,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.899403578528827,
      "eval_loss": 0.35866183042526245,
      "eval_runtime": 106.3012,
      "eval_samples_per_second": 23.659,
      "eval_steps_per_second": 1.486,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 28.916933059692383,
      "learning_rate": 2.9264522131293818e-05,
      "loss": 0.6392,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.14949369430542,
      "learning_rate": 2.9126092830748217e-05,
      "loss": 0.5424,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 16.16094398498535,
      "learning_rate": 2.897612111101888e-05,
      "loss": 0.4645,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 19.93354606628418,
      "learning_rate": 2.8814729513878365e-05,
      "loss": 0.4904,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 30.69086456298828,
      "learning_rate": 2.864204991227195e-05,
      "loss": 0.4554,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 26.814476013183594,
      "learning_rate": 2.8458223402564366e-05,
      "loss": 0.4355,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 36.27682876586914,
      "learning_rate": 2.826340018925006e-05,
      "loss": 0.4855,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 30.071430206298828,
      "learning_rate": 2.8057739462221215e-05,
      "loss": 0.4967,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 23.351360321044922,
      "learning_rate": 2.7841409266693838e-05,
      "loss": 0.4816,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 25.510053634643555,
      "learning_rate": 2.761458636589813e-05,
      "loss": 0.4129,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 17.746910095214844,
      "learning_rate": 2.7377456096645395e-05,
      "loss": 0.4599,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.921272365805169,
      "eval_loss": 0.27432167530059814,
      "eval_runtime": 106.7009,
      "eval_samples_per_second": 23.571,
      "eval_steps_per_second": 1.481,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.30936622619629,
      "learning_rate": 2.7130212217889484e-05,
      "loss": 0.4444,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 15.899140357971191,
      "learning_rate": 2.6873056752406504e-05,
      "loss": 0.3567,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 27.349393844604492,
      "learning_rate": 2.6606199821722166e-05,
      "loss": 0.3915,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 19.646141052246094,
      "learning_rate": 2.632985947442167e-05,
      "loss": 0.3738,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 16.584985733032227,
      "learning_rate": 2.6044261507982356e-05,
      "loss": 0.3886,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 14.146623611450195,
      "learning_rate": 2.5749639284274782e-05,
      "loss": 0.3752,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 15.094594955444336,
      "learning_rate": 2.5446233538882924e-05,
      "loss": 0.358,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 17.518064498901367,
      "learning_rate": 2.513429218439932e-05,
      "loss": 0.4038,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 12.49306583404541,
      "learning_rate": 2.4814070107855878e-05,
      "loss": 0.3582,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 19.405559539794922,
      "learning_rate": 2.448582896245591e-05,
      "loss": 0.3737,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 29.24690055847168,
      "learning_rate": 2.4149836953777488e-05,
      "loss": 0.359,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.925248508946322,
      "eval_loss": 0.25794780254364014,
      "eval_runtime": 106.6043,
      "eval_samples_per_second": 23.592,
      "eval_steps_per_second": 1.482,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.275596618652344,
      "learning_rate": 2.3806368620622876e-05,
      "loss": 0.373,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 16.224103927612305,
      "learning_rate": 2.345570461069312e-05,
      "loss": 0.2986,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 11.221643447875977,
      "learning_rate": 2.3098131451271016e-05,
      "loss": 0.3101,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 12.0270414352417,
      "learning_rate": 2.2733941315099883e-05,
      "loss": 0.3332,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 33.56585693359375,
      "learning_rate": 2.2363431781649483e-05,
      "loss": 0.2884,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 10.652863502502441,
      "learning_rate": 2.1986905593964048e-05,
      "loss": 0.3069,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 18.348175048828125,
      "learning_rate": 2.1604670411291174e-05,
      "loss": 0.3166,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 30.576799392700195,
      "learning_rate": 2.121703855769373e-05,
      "loss": 0.321,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 18.137060165405273,
      "learning_rate": 2.0824326766850072e-05,
      "loss": 0.3415,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 6.155796051025391,
      "learning_rate": 2.042685592325123e-05,
      "loss": 0.302,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 9.89777660369873,
      "learning_rate": 2.0024950800006463e-05,
      "loss": 0.3047,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9387673956262426,
      "eval_loss": 0.24044357240200043,
      "eval_runtime": 105.8346,
      "eval_samples_per_second": 23.763,
      "eval_steps_per_second": 1.493,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 25.019359588623047,
      "learning_rate": 1.961893979347137e-05,
      "loss": 0.3006,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 13.57244873046875,
      "learning_rate": 1.9209154654915524e-05,
      "loss": 0.2743,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 30.206947326660156,
      "learning_rate": 1.879593021944875e-05,
      "loss": 0.2683,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 14.6887788772583,
      "learning_rate": 1.837960413242765e-05,
      "loss": 0.2623,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 11.224356651306152,
      "learning_rate": 1.796051657356582e-05,
      "loss": 0.2497,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 25.0120906829834,
      "learning_rate": 1.7539009978973312e-05,
      "loss": 0.256,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 19.190332412719727,
      "learning_rate": 1.711542876135233e-05,
      "loss": 0.2635,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 16.990314483642578,
      "learning_rate": 1.669011902857791e-05,
      "loss": 0.2395,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 32.22107696533203,
      "learning_rate": 1.6263428300893422e-05,
      "loss": 0.2839,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 21.929677963256836,
      "learning_rate": 1.5835705226952112e-05,
      "loss": 0.2759,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 2.2401912212371826,
      "learning_rate": 1.540729929893649e-05,
      "loss": 0.2869,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9407554671968191,
      "eval_loss": 0.23475004732608795,
      "eval_runtime": 106.6456,
      "eval_samples_per_second": 23.583,
      "eval_steps_per_second": 1.482,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 8.231451988220215,
      "learning_rate": 1.4978560566988603e-05,
      "loss": 0.2711,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 25.471065521240234,
      "learning_rate": 1.454983935318433e-05,
      "loss": 0.2116,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 3.9416098594665527,
      "learning_rate": 1.4121485965285485e-05,
      "loss": 0.2145,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 10.868898391723633,
      "learning_rate": 1.3693850410503614e-05,
      "loss": 0.2142,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.5696491599082947,
      "learning_rate": 1.326728210950942e-05,
      "loss": 0.1941,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 19.500898361206055,
      "learning_rate": 1.2842129610921378e-05,
      "loss": 0.225,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 14.424389839172363,
      "learning_rate": 1.2418740306506923e-05,
      "loss": 0.2361,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 12.626699447631836,
      "learning_rate": 1.1997460147328984e-05,
      "loss": 0.2245,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 19.012290954589844,
      "learning_rate": 1.1578633361069559e-05,
      "loss": 0.2374,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 19.391586303710938,
      "learning_rate": 1.1162602170761611e-05,
      "loss": 0.2108,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 15.383687973022461,
      "learning_rate": 1.0749706515158863e-05,
      "loss": 0.2468,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9455268389662028,
      "eval_loss": 0.2276468127965927,
      "eval_runtime": 106.8928,
      "eval_samples_per_second": 23.528,
      "eval_steps_per_second": 1.478,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.3950190842151642,
      "learning_rate": 1.0340283770972167e-05,
      "loss": 0.2195,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 20.661237716674805,
      "learning_rate": 9.93466847719919e-06,
      "loss": 0.2352,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 13.067805290222168,
      "learning_rate": 9.533192061772919e-06,
      "loss": 0.2083,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.3384680151939392,
      "learning_rate": 9.136182570752153e-06,
      "loss": 0.194,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 4.762894153594971,
      "learning_rate": 8.743964400275304e-06,
      "loss": 0.1912,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 10.7648286819458,
      "learning_rate": 8.356858031496596e-06,
      "loss": 0.25,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 3.155149459838867,
      "learning_rate": 7.975179768721187e-06,
      "loss": 0.1595,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 1.457185983657837,
      "learning_rate": 7.599241480953112e-06,
      "loss": 0.1812,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 17.26123809814453,
      "learning_rate": 7.229350347067426e-06,
      "loss": 0.1688,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 11.0090970993042,
      "learning_rate": 6.865808604814564e-06,
      "loss": 0.162,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 10.775118827819824,
      "learning_rate": 6.508913303862144e-06,
      "loss": 0.2098,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.947117296222664,
      "eval_loss": 0.23031435906887054,
      "eval_runtime": 106.5243,
      "eval_samples_per_second": 23.61,
      "eval_steps_per_second": 1.483,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 15.345827102661133,
      "learning_rate": 6.1589560630758656e-06,
      "loss": 0.179,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.20503534376621246,
      "learning_rate": 5.8162228322380155e-06,
      "loss": 0.1693,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 7.773012161254883,
      "learning_rate": 5.480993658398129e-06,
      "loss": 0.1712,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 1.7917510271072388,
      "learning_rate": 5.153542457046737e-06,
      "loss": 0.1761,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 1.3030850887298584,
      "learning_rate": 4.834136788299248e-06,
      "loss": 0.1733,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 22.458986282348633,
      "learning_rate": 4.523037638272822e-06,
      "loss": 0.1514,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 36.26962661743164,
      "learning_rate": 4.220499205834783e-06,
      "loss": 0.1778,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 10.59404468536377,
      "learning_rate": 3.926768694896931e-06,
      "loss": 0.1818,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 12.982277870178223,
      "learning_rate": 3.6420861124254607e-06,
      "loss": 0.1595,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 17.472332000732422,
      "learning_rate": 3.3666840723314145e-06,
      "loss": 0.1734,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.3642497658729553,
      "learning_rate": 3.1007876054020724e-06,
      "loss": 0.1944,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9495029821073558,
      "eval_loss": 0.22436106204986572,
      "eval_runtime": 106.33,
      "eval_samples_per_second": 23.653,
      "eval_steps_per_second": 1.486,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 17.4754695892334,
      "learning_rate": 2.8446139754284486e-06,
      "loss": 0.206,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 7.8490753173828125,
      "learning_rate": 2.5983725016792574e-06,
      "loss": 0.1825,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.25403663516044617,
      "learning_rate": 2.36226438786627e-06,
      "loss": 0.1642,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.2140350341796875,
      "learning_rate": 2.1364825577409424e-06,
      "loss": 0.134,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 2.0285985469818115,
      "learning_rate": 1.9212114974565664e-06,
      "loss": 0.1677,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 27.10205078125,
      "learning_rate": 1.7166271048247796e-06,
      "loss": 0.136,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 6.272194862365723,
      "learning_rate": 1.5228965455896054e-06,
      "loss": 0.1325,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 9.633159637451172,
      "learning_rate": 1.3401781168364591e-06,
      "loss": 0.186,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 14.02453327178955,
      "learning_rate": 1.1686211176477208e-06,
      "loss": 0.1499,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 12.952245712280273,
      "learning_rate": 1.00836572711058e-06,
      "loss": 0.1854,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 8.070415496826172,
      "learning_rate": 8.595428897768071e-07,
      "loss": 0.1739,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9506958250497017,
      "eval_loss": 0.2247273474931717,
      "eval_runtime": 106.3844,
      "eval_samples_per_second": 23.641,
      "eval_steps_per_second": 1.485,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 23.99713134765625,
      "learning_rate": 7.222742086680756e-07,
      "loss": 0.176,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 1.0274367332458496,
      "learning_rate": 5.966718459142196e-07,
      "loss": 0.1778,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 34.5405387878418,
      "learning_rate": 4.82838431105655e-07,
      "loss": 0.203,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 23.137969970703125,
      "learning_rate": 3.808669774348167e-07,
      "loss": 0.1419,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 24.021142959594727,
      "learning_rate": 2.908408056951578e-07,
      "loss": 0.1598,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.9892005920410156,
      "learning_rate": 2.1283347619979243e-07,
      "loss": 0.148,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 4.078668594360352,
      "learning_rate": 1.4690872867542892e-07,
      "loss": 0.181,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 0.3794693946838379,
      "learning_rate": 9.312043018067762e-08,
      "loss": 0.2031,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 16.30907440185547,
      "learning_rate": 5.1512531091333914e-08,
      "loss": 0.1606,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 22.63953971862793,
      "learning_rate": 2.211902918855313e-08,
      "loss": 0.1481,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 14.57466983795166,
      "learning_rate": 4.963941879295164e-09,
      "loss": 0.1508,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9487077534791253,
      "eval_loss": 0.22431252896785736,
      "eval_runtime": 106.5985,
      "eval_samples_per_second": 23.593,
      "eval_steps_per_second": 1.482,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.33142790373506276,
      "train_runtime": 19145.449,
      "train_samples_per_second": 9.183,
      "train_steps_per_second": 0.574
    }
  ],
  "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
}