trainer_state.json

{
  "best_metric": 0.9410715683314885,
  "best_model_checkpoint": "model_saves/deberta-large_spell_5k_5_p3/checkpoint-268",
  "epoch": 4.0,
  "global_step": 1072,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9410715683314885,
      "eval_loss": 0.41616517305374146,
      "eval_runtime": 4.6708,
      "eval_samples_per_second": 938.167,
      "eval_steps_per_second": 7.493,
      "step": 268
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9403839488366372,
      "eval_loss": 0.44042402505874634,
      "eval_runtime": 4.6657,
      "eval_samples_per_second": 939.196,
      "eval_steps_per_second": 7.502,
      "step": 536
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9402777311910911,
      "eval_loss": 0.48104551434516907,
      "eval_runtime": 4.6927,
      "eval_samples_per_second": 933.79,
      "eval_steps_per_second": 7.458,
      "step": 804
    },
    {
      "epoch": 3.73,
      "learning_rate": 1e-05,
      "loss": 0.2516,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9393329531859703,
      "eval_loss": 0.5351813435554504,
      "eval_runtime": 4.6758,
      "eval_samples_per_second": 937.164,
      "eval_steps_per_second": 7.485,
      "step": 1072
    },
    {
      "epoch": 4.0,
      "step": 1072,
      "total_flos": 1.607398791839744e+16,
      "train_loss": 0.24524620397767025,
      "train_runtime": 459.3316,
      "train_samples_per_second": 1120.236,
      "train_steps_per_second": 8.752
    }
  ],
  "max_steps": 4020,
  "num_train_epochs": 15,
  "total_flos": 1.607398791839744e+16,
  "trial_name": null,
  "trial_params": null
}