{
  "best_global_step": 1410,
  "best_metric": 0.7945503472501974,
  "best_model_checkpoint": "./roberta-toxic-classifier-en/checkpoint-1410",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 1410,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3546099290780142,
      "grad_norm": 9.611958503723145,
      "learning_rate": 9.884160756501183e-06,
      "loss": 0.653124008178711,
      "step": 50
    },
    {
      "epoch": 0.7092198581560284,
      "grad_norm": 7.586147308349609,
      "learning_rate": 9.765957446808511e-06,
      "loss": 0.5039330291748046,
      "step": 100
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.745,
      "eval_f1": 0.7445706232176288,
      "eval_loss": 0.5029547214508057,
      "eval_precision": 0.7506988625409678,
      "eval_recall": 0.7551099231212526,
      "eval_runtime": 0.3569,
      "eval_samples_per_second": 2801.728,
      "eval_steps_per_second": 44.828,
      "step": 141
    },
    {
      "epoch": 1.0638297872340425,
      "grad_norm": 9.09180736541748,
      "learning_rate": 9.64775413711584e-06,
      "loss": 0.4156942367553711,
      "step": 150
    },
    {
      "epoch": 1.4184397163120568,
      "grad_norm": 7.9193925857543945,
      "learning_rate": 9.529550827423168e-06,
      "loss": 0.38147411346435545,
      "step": 200
    },
    {
      "epoch": 1.773049645390071,
      "grad_norm": 11.232121467590332,
      "learning_rate": 9.411347517730498e-06,
      "loss": 0.3642482376098633,
      "step": 250
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.778,
      "eval_f1": 0.7759350171177573,
      "eval_loss": 0.45547381043434143,
      "eval_precision": 0.7752243747770282,
      "eval_recall": 0.7806237764181289,
      "eval_runtime": 0.3628,
      "eval_samples_per_second": 2756.355,
      "eval_steps_per_second": 44.102,
      "step": 282
    },
    {
      "epoch": 2.127659574468085,
      "grad_norm": 19.67840576171875,
      "learning_rate": 9.293144208037825e-06,
      "loss": 0.36101982116699216,
      "step": 300
    },
    {
      "epoch": 2.482269503546099,
      "grad_norm": 7.5036492347717285,
      "learning_rate": 9.174940898345155e-06,
      "loss": 0.3269504547119141,
      "step": 350
    },
    {
      "epoch": 2.8368794326241136,
      "grad_norm": 18.646142959594727,
      "learning_rate": 9.056737588652483e-06,
      "loss": 0.32073143005371096,
      "step": 400
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.775,
      "eval_f1": 0.7745832043448335,
      "eval_loss": 0.5032930374145508,
      "eval_precision": 0.780389401846648,
      "eval_recall": 0.7854649713288456,
      "eval_runtime": 0.3639,
      "eval_samples_per_second": 2748.102,
      "eval_steps_per_second": 43.97,
      "step": 423
    },
    {
      "epoch": 3.1914893617021276,
      "grad_norm": 6.803809642791748,
      "learning_rate": 8.938534278959812e-06,
      "loss": 0.29136837005615235,
      "step": 450
    },
    {
      "epoch": 3.546099290780142,
      "grad_norm": 13.090974807739258,
      "learning_rate": 8.82033096926714e-06,
      "loss": 0.27523290634155273,
      "step": 500
    },
    {
      "epoch": 3.900709219858156,
      "grad_norm": 10.417556762695312,
      "learning_rate": 8.70212765957447e-06,
      "loss": 0.2774989700317383,
      "step": 550
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.791,
      "eval_f1": 0.789619694817699,
      "eval_loss": 0.4875805974006653,
      "eval_precision": 0.7899062159912937,
      "eval_recall": 0.7961446186920396,
      "eval_runtime": 0.3651,
      "eval_samples_per_second": 2738.756,
      "eval_steps_per_second": 43.82,
      "step": 564
    },
    {
      "epoch": 4.25531914893617,
      "grad_norm": 13.110213279724121,
      "learning_rate": 8.583924349881797e-06,
      "loss": 0.240526180267334,
      "step": 600
    },
    {
      "epoch": 4.609929078014185,
      "grad_norm": 16.28187370300293,
      "learning_rate": 8.465721040189125e-06,
      "loss": 0.21295007705688476,
      "step": 650
    },
    {
      "epoch": 4.964539007092198,
      "grad_norm": 16.510469436645508,
      "learning_rate": 8.347517730496454e-06,
      "loss": 0.22765600204467773,
      "step": 700
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.788,
      "eval_f1": 0.7827120812082848,
      "eval_loss": 0.5206102132797241,
      "eval_precision": 0.7836996269193908,
      "eval_recall": 0.781890783950693,
      "eval_runtime": 0.3796,
      "eval_samples_per_second": 2634.502,
      "eval_steps_per_second": 42.152,
      "step": 705
    },
    {
      "epoch": 5.319148936170213,
      "grad_norm": 24.18474006652832,
      "learning_rate": 8.229314420803784e-06,
      "loss": 0.18396900177001954,
      "step": 750
    },
    {
      "epoch": 5.673758865248227,
      "grad_norm": 21.326946258544922,
      "learning_rate": 8.111111111111112e-06,
      "loss": 0.19418764114379883,
      "step": 800
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.796,
      "eval_f1": 0.7937694099378882,
      "eval_loss": 0.6245253086090088,
      "eval_precision": 0.7925987495934372,
      "eval_recall": 0.7978223818924188,
      "eval_runtime": 0.3664,
      "eval_samples_per_second": 2729.117,
      "eval_steps_per_second": 43.666,
      "step": 846
    },
    {
      "epoch": 6.028368794326241,
      "grad_norm": 7.92404842376709,
      "learning_rate": 7.992907801418441e-06,
      "loss": 0.1742597770690918,
      "step": 850
    },
    {
      "epoch": 6.382978723404255,
      "grad_norm": 11.012433052062988,
      "learning_rate": 7.874704491725769e-06,
      "loss": 0.14716711044311523,
      "step": 900
    },
    {
      "epoch": 6.73758865248227,
      "grad_norm": 9.80257511138916,
      "learning_rate": 7.756501182033097e-06,
      "loss": 0.13163516998291017,
      "step": 950
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.796,
      "eval_f1": 0.7943341062607118,
      "eval_loss": 0.6543664932250977,
      "eval_precision": 0.7938576995006228,
      "eval_recall": 0.7999109007606132,
      "eval_runtime": 0.3667,
      "eval_samples_per_second": 2727.034,
      "eval_steps_per_second": 43.633,
      "step": 987
    },
    {
      "epoch": 7.092198581560283,
      "grad_norm": 2.5416817665100098,
      "learning_rate": 7.638297872340426e-06,
      "loss": 0.1342453956604004,
      "step": 1000
    },
    {
      "epoch": 7.446808510638298,
      "grad_norm": 9.653868675231934,
      "learning_rate": 7.520094562647755e-06,
      "loss": 0.11633362770080566,
      "step": 1050
    },
    {
      "epoch": 7.801418439716312,
      "grad_norm": 16.958995819091797,
      "learning_rate": 7.401891252955083e-06,
      "loss": 0.11314620018005371,
      "step": 1100
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.777,
      "eval_f1": 0.77662450579424,
      "eval_loss": 0.788261353969574,
      "eval_precision": 0.7828304736199473,
      "eval_recall": 0.7878068917035529,
      "eval_runtime": 0.3673,
      "eval_samples_per_second": 2722.508,
      "eval_steps_per_second": 43.56,
      "step": 1128
    },
    {
      "epoch": 8.156028368794326,
      "grad_norm": 9.911176681518555,
      "learning_rate": 7.283687943262412e-06,
      "loss": 0.10539370536804199,
      "step": 1150
    },
    {
      "epoch": 8.51063829787234,
      "grad_norm": 14.041365623474121,
      "learning_rate": 7.1654846335697405e-06,
      "loss": 0.10781905174255371,
      "step": 1200
    },
    {
      "epoch": 8.865248226950355,
      "grad_norm": 10.613764762878418,
      "learning_rate": 7.047281323877069e-06,
      "loss": 0.10644716262817383,
      "step": 1250
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.788,
      "eval_f1": 0.7864203650196856,
      "eval_loss": 0.7684714198112488,
      "eval_precision": 0.7862955357821888,
      "eval_recall": 0.7923333782916652,
      "eval_runtime": 0.3673,
      "eval_samples_per_second": 2722.633,
      "eval_steps_per_second": 43.562,
      "step": 1269
    },
    {
      "epoch": 9.21985815602837,
      "grad_norm": 15.78783130645752,
      "learning_rate": 6.929078014184397e-06,
      "loss": 0.09560779571533203,
      "step": 1300
    },
    {
      "epoch": 9.574468085106384,
      "grad_norm": 10.481488227844238,
      "learning_rate": 6.810874704491726e-06,
      "loss": 0.0669502067565918,
      "step": 1350
    },
    {
      "epoch": 9.929078014184396,
      "grad_norm": 52.73137664794922,
      "learning_rate": 6.692671394799055e-06,
      "loss": 0.07912938117980957,
      "step": 1400
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.796,
      "eval_f1": 0.7945503472501974,
      "eval_loss": 0.9117002487182617,
      "eval_precision": 0.7945365556929553,
      "eval_recall": 0.8008059802755537,
      "eval_runtime": 0.3717,
      "eval_samples_per_second": 2690.34,
      "eval_steps_per_second": 43.045,
      "step": 1410
    }
  ],
  "logging_steps": 50,
  "max_steps": 4230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5919998745600000.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}