models/deberta-offensive/checkpoint-2500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.6778523489932886,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.34,
      "grad_norm": 1.5860745906829834,
      "learning_rate": 4.440715883668904e-05,
      "loss": 0.536,
      "step": 500
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.4774603843688965,
      "learning_rate": 3.881431767337807e-05,
      "loss": 0.5098,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8383720930232558,
      "eval_f1": 0.6971677559912854,
      "eval_loss": 0.3849833607673645,
      "eval_precision": 0.730593607305936,
      "eval_recall": 0.6666666666666666,
      "eval_runtime": 2.5975,
      "eval_samples_per_second": 331.086,
      "eval_steps_per_second": 41.578,
      "step": 1490
    },
    {
      "epoch": 1.01,
      "grad_norm": 4.723637104034424,
      "learning_rate": 3.3221476510067115e-05,
      "loss": 0.4939,
      "step": 1500
    },
    {
      "epoch": 1.34,
      "grad_norm": 7.389687538146973,
      "learning_rate": 2.7628635346756153e-05,
      "loss": 0.422,
      "step": 2000
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.1347562074661255,
      "learning_rate": 2.2035794183445192e-05,
      "loss": 0.4249,
      "step": 2500
    }
  ],
  "logging_steps": 500,
  "max_steps": 4470,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 1048312855592304.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}