trainer_state.json

{
  "best_metric": 0.9863948830227645,
  "best_model_checkpoint": "./output//29_roberta-large_nace_5__5e-6_0.01_0.06_07-21-22_10-40/checkpoint-46000",
  "epoch": 1.6596362630523478,
  "global_step": 48000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 1.1524720525527257e-06,
      "loss": 0.1493,
      "step": 2000
    },
    {
      "epoch": 0.07,
      "eval_accuracy": 0.9610440126806309,
      "eval_f1": 0.9764342504146886,
      "eval_loss": 0.05725020170211792,
      "eval_roc_auc": 0.9759211072420441,
      "eval_runtime": 1536.6369,
      "eval_samples_per_second": 33.461,
      "step": 2000
    },
    {
      "epoch": 0.14,
      "learning_rate": 2.3049441051054513e-06,
      "loss": 0.0528,
      "step": 4000
    },
    {
      "epoch": 0.14,
      "eval_accuracy": 0.9728883443219168,
      "eval_f1": 0.9829246052694726,
      "eval_loss": 0.05182106792926788,
      "eval_roc_auc": 0.9827894561625595,
      "eval_runtime": 1536.9258,
      "eval_samples_per_second": 33.454,
      "step": 4000
    },
    {
      "epoch": 0.21,
      "learning_rate": 3.457416157658177e-06,
      "loss": 0.0497,
      "step": 6000
    },
    {
      "epoch": 0.21,
      "eval_accuracy": 0.9731606278079235,
      "eval_f1": 0.9840875186474392,
      "eval_loss": 0.048033468425273895,
      "eval_roc_auc": 0.9838391681932317,
      "eval_runtime": 1538.305,
      "eval_samples_per_second": 33.424,
      "step": 6000
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.609888210210903e-06,
      "loss": 0.0462,
      "step": 8000
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 0.9722465332477586,
      "eval_f1": 0.9834792174869661,
      "eval_loss": 0.054564349353313446,
      "eval_roc_auc": 0.9832329187393456,
      "eval_runtime": 1538.1819,
      "eval_samples_per_second": 33.427,
      "step": 8000
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.951336320098873e-06,
      "loss": 0.0433,
      "step": 10000
    },
    {
      "epoch": 0.35,
      "eval_accuracy": 0.9736079506777914,
      "eval_f1": 0.9837862700755107,
      "eval_loss": 0.04221172630786896,
      "eval_roc_auc": 0.9836052885203017,
      "eval_runtime": 1535.859,
      "eval_samples_per_second": 33.478,
      "step": 10000
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.877770666431257e-06,
      "loss": 0.0442,
      "step": 12000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.9738413365229399,
      "eval_f1": 0.9840872143062064,
      "eval_loss": 0.047410748898983,
      "eval_roc_auc": 0.9838395246306029,
      "eval_runtime": 1536.4074,
      "eval_samples_per_second": 33.466,
      "step": 12000
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.804205012763641e-06,
      "loss": 0.0433,
      "step": 14000
    },
    {
      "epoch": 0.48,
      "eval_accuracy": 0.9747748799035338,
      "eval_f1": 0.9848328360492695,
      "eval_loss": 0.04587765783071518,
      "eval_roc_auc": 0.9845986397430451,
      "eval_runtime": 1536.4412,
      "eval_samples_per_second": 33.465,
      "step": 14000
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.730639359096026e-06,
      "loss": 0.0427,
      "step": 16000
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 0.9745609428788145,
      "eval_f1": 0.9846027155907081,
      "eval_loss": 0.04280721768736839,
      "eval_roc_auc": 0.9844085121898316,
      "eval_runtime": 1536.1052,
      "eval_samples_per_second": 33.472,
      "step": 16000
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.65707370542841e-06,
      "loss": 0.0401,
      "step": 18000
    },
    {
      "epoch": 0.62,
      "eval_accuracy": 0.9754555886185503,
      "eval_f1": 0.9849309815950921,
      "eval_loss": 0.0410287007689476,
      "eval_roc_auc": 0.9847790254315509,
      "eval_runtime": 1536.224,
      "eval_samples_per_second": 33.47,
      "step": 18000
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.583508051760795e-06,
      "loss": 0.0409,
      "step": 20000
    },
    {
      "epoch": 0.69,
      "eval_accuracy": 0.974891572826108,
      "eval_f1": 0.9855708668931557,
      "eval_loss": 0.040095292031764984,
      "eval_roc_auc": 0.9852792796060187,
      "eval_runtime": 1536.3646,
      "eval_samples_per_second": 33.467,
      "step": 20000
    },
    {
      "epoch": 0.76,
      "learning_rate": 4.509942398093178e-06,
      "loss": 0.0403,
      "step": 22000
    },
    {
      "epoch": 0.76,
      "eval_accuracy": 0.9755139350798374,
      "eval_f1": 0.9850088183421517,
      "eval_loss": 0.040499284863471985,
      "eval_roc_auc": 0.9848554210846383,
      "eval_runtime": 1536.2737,
      "eval_samples_per_second": 33.469,
      "step": 22000
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.436376744425563e-06,
      "loss": 0.0414,
      "step": 24000
    },
    {
      "epoch": 0.83,
      "eval_accuracy": 0.975319446875547,
      "eval_f1": 0.9852808881232654,
      "eval_loss": 0.041136886924505234,
      "eval_roc_auc": 0.985079469215943,
      "eval_runtime": 1536.5795,
      "eval_samples_per_second": 33.462,
      "step": 24000
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.362811090757947e-06,
      "loss": 0.0402,
      "step": 26000
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.9764474784604313,
      "eval_f1": 0.9861650624772766,
      "eval_loss": 0.04073059558868408,
      "eval_roc_auc": 0.9859694252223201,
      "eval_runtime": 1536.8522,
      "eval_samples_per_second": 33.456,
      "step": 26000
    },
    {
      "epoch": 0.97,
      "learning_rate": 4.289245437090332e-06,
      "loss": 0.0399,
      "step": 28000
    },
    {
      "epoch": 0.97,
      "eval_accuracy": 0.9755722815411245,
      "eval_f1": 0.9853538060958797,
      "eval_loss": 0.040422726422548294,
      "eval_roc_auc": 0.9851619243043377,
      "eval_runtime": 1537.2709,
      "eval_samples_per_second": 33.447,
      "step": 28000
    },
    {
      "epoch": 1.04,
      "learning_rate": 4.215679783422716e-06,
      "loss": 0.04,
      "step": 30000
    },
    {
      "epoch": 1.04,
      "eval_accuracy": 0.9761946437948539,
      "eval_f1": 0.9859556473987333,
      "eval_loss": 0.03985821455717087,
      "eval_roc_auc": 0.9857539906157645,
      "eval_runtime": 1537.4136,
      "eval_samples_per_second": 33.444,
      "step": 30000
    },
    {
      "epoch": 1.11,
      "learning_rate": 4.142114129755101e-06,
      "loss": 0.0373,
      "step": 32000
    },
    {
      "epoch": 1.11,
      "eval_accuracy": 0.9766419666647218,
      "eval_f1": 0.9855267070141279,
      "eval_loss": 0.04419500008225441,
      "eval_roc_auc": 0.9854023838792458,
      "eval_runtime": 1536.3227,
      "eval_samples_per_second": 33.468,
      "step": 32000
    },
    {
      "epoch": 1.18,
      "learning_rate": 4.068548476087484e-06,
      "loss": 0.0357,
      "step": 34000
    },
    {
      "epoch": 1.18,
      "eval_accuracy": 0.9768948013302993,
      "eval_f1": 0.9859001025621362,
      "eval_loss": 0.04071442782878876,
      "eval_roc_auc": 0.9857602731082165,
      "eval_runtime": 1536.5165,
      "eval_samples_per_second": 33.463,
      "step": 34000
    },
    {
      "epoch": 1.24,
      "learning_rate": 3.994982822419869e-06,
      "loss": 0.0365,
      "step": 36000
    },
    {
      "epoch": 1.24,
      "eval_accuracy": 0.976252990256141,
      "eval_f1": 0.9857402622260503,
      "eval_loss": 0.04006611183285713,
      "eval_roc_auc": 0.9855463976313715,
      "eval_runtime": 1536.8448,
      "eval_samples_per_second": 33.456,
      "step": 36000
    },
    {
      "epoch": 1.31,
      "learning_rate": 3.921417168752253e-06,
      "loss": 0.0362,
      "step": 38000
    },
    {
      "epoch": 1.31,
      "eval_accuracy": 0.9752805492346889,
      "eval_f1": 0.9857657021203471,
      "eval_loss": 0.03810262680053711,
      "eval_roc_auc": 0.9854908677538746,
      "eval_runtime": 1536.4894,
      "eval_samples_per_second": 33.464,
      "step": 38000
    },
    {
      "epoch": 1.38,
      "learning_rate": 3.847851515084638e-06,
      "loss": 0.0367,
      "step": 40000
    },
    {
      "epoch": 1.38,
      "eval_accuracy": 0.9741719664702336,
      "eval_f1": 0.9848820340536,
      "eval_loss": 0.03949220851063728,
      "eval_roc_auc": 0.984581025607102,
      "eval_runtime": 1536.5636,
      "eval_samples_per_second": 33.462,
      "step": 40000
    },
    {
      "epoch": 1.45,
      "learning_rate": 3.7742858614170217e-06,
      "loss": 0.0366,
      "step": 42000
    },
    {
      "epoch": 1.45,
      "eval_accuracy": 0.9754166909776922,
      "eval_f1": 0.9858603603173542,
      "eval_loss": 0.03979261964559555,
      "eval_roc_auc": 0.9855892138192017,
      "eval_runtime": 1537.1787,
      "eval_samples_per_second": 33.449,
      "step": 42000
    },
    {
      "epoch": 1.52,
      "learning_rate": 3.7007202077494066e-06,
      "loss": 0.0354,
      "step": 44000
    },
    {
      "epoch": 1.52,
      "eval_accuracy": 0.9768559036894412,
      "eval_f1": 0.9861624145916668,
      "eval_loss": 0.039536893367767334,
      "eval_roc_auc": 0.9859729895960303,
      "eval_runtime": 1536.4855,
      "eval_samples_per_second": 33.464,
      "step": 44000
    },
    {
      "epoch": 1.59,
      "learning_rate": 3.6271545540817902e-06,
      "loss": 0.0365,
      "step": 46000
    },
    {
      "epoch": 1.59,
      "eval_accuracy": 0.9768364548690122,
      "eval_f1": 0.9863948830227645,
      "eval_loss": 0.04265940561890602,
      "eval_roc_auc": 0.9861797953532997,
      "eval_runtime": 1536.6035,
      "eval_samples_per_second": 33.461,
      "step": 46000
    },
    {
      "epoch": 1.66,
      "learning_rate": 3.553588900414175e-06,
      "loss": 0.0358,
      "step": 48000
    },
    {
      "epoch": 1.66,
      "eval_accuracy": 0.9745609428788145,
      "eval_f1": 0.9850252764160049,
      "eval_loss": 0.039769496768713,
      "eval_roc_auc": 0.9847700991727389,
      "eval_runtime": 1536.7946,
      "eval_samples_per_second": 33.457,
      "step": 48000
    }
  ],
  "max_steps": 144610,
  "num_train_epochs": 5,
  "total_flos": 6.9866971594752e+16,
  "trial_name": null,
  "trial_params": null
}