finetune/cola/trainer_state.json

{
  "best_metric": 0.7931937172774869,
  "best_model_checkpoint": "finetune_results/omarmomen/structformer_s1_final_with_pos/cola/checkpoint-400",
  "epoch": 10.0,
  "global_step": 690,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 5.8,
      "eval_accuracy": 0.689892053604126,
      "eval_f1": 0.7931937172774869,
      "eval_loss": 0.762876570224762,
      "eval_mcc": 0.19495355846277174,
      "eval_runtime": 1.9776,
      "eval_samples_per_second": 515.265,
      "eval_steps_per_second": 64.724,
      "step": 400
    },
    {
      "epoch": 7.25,
      "learning_rate": 1.3768115942028985e-05,
      "loss": 0.4623,
      "step": 500
    },
    {
      "epoch": 10.0,
      "step": 690,
      "total_flos": 6814792144343040.0,
      "train_loss": 0.3857684093972911,
      "train_runtime": 391.5771,
      "train_samples_per_second": 208.49,
      "train_steps_per_second": 1.762
    }
  ],
  "max_steps": 690,
  "num_train_epochs": 10,
  "total_flos": 6814792144343040.0,
  "trial_name": null,
  "trial_params": null
}