{
  "best_metric": 0.8339889850511408,
  "best_model_checkpoint": "./hatebert_finetuned_v5/checkpoint-3000",
  "epoch": 0.3759869657851861,
  "eval_steps": 1000,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06266449429753101,
      "grad_norm": 2.7858245372772217,
      "learning_rate": 4.160401002506266e-06,
      "loss": 0.5658,
      "step": 500
    },
    {
      "epoch": 0.12532898859506203,
      "grad_norm": 0.6836428642272949,
      "learning_rate": 8.320802005012531e-06,
      "loss": 0.3394,
      "step": 1000
    },
    {
      "epoch": 0.12532898859506203,
      "eval_accuracy": 0.9509008303305656,
      "eval_f1": 0.7696604439217992,
      "eval_loss": 0.3321492373943329,
      "eval_optimal_threshold": 0.7856788039207458,
      "eval_runtime": 24.5664,
      "eval_samples_per_second": 1299.13,
      "eval_steps_per_second": 40.625,
      "step": 1000
    },
    {
      "epoch": 0.18799348289259304,
      "grad_norm": 0.07630700618028641,
      "learning_rate": 1.2497911445279867e-05,
      "loss": 0.3098,
      "step": 1500
    },
    {
      "epoch": 0.25065797719012406,
      "grad_norm": 0.6884164214134216,
      "learning_rate": 1.66750208855472e-05,
      "loss": 0.3245,
      "step": 2000
    },
    {
      "epoch": 0.25065797719012406,
      "eval_accuracy": 0.9620241265862447,
      "eval_f1": 0.8181272509003601,
      "eval_loss": 0.24252600967884064,
      "eval_optimal_threshold": 0.6846858263015747,
      "eval_runtime": 24.5401,
      "eval_samples_per_second": 1300.526,
      "eval_steps_per_second": 40.668,
      "step": 2000
    },
    {
      "epoch": 0.3133224714876551,
      "grad_norm": 1.6269655227661133,
      "learning_rate": 1.9906234043540826e-05,
      "loss": 0.2939,
      "step": 2500
    },
    {
      "epoch": 0.3759869657851861,
      "grad_norm": 53.418663024902344,
      "learning_rate": 1.9442046140277588e-05,
      "loss": 0.2911,
      "step": 3000
    },
    {
      "epoch": 0.3759869657851861,
      "eval_accuracy": 0.9669434435218549,
      "eval_f1": 0.8339889850511408,
      "eval_loss": 0.32369741797447205,
      "eval_optimal_threshold": 0.639145016670227,
      "eval_runtime": 24.5157,
      "eval_samples_per_second": 1301.819,
      "eval_steps_per_second": 40.709,
      "step": 3000
    }
  ],
  "logging_steps": 500,
  "max_steps": 23937,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}