trainer_state.json

{
  "best_metric": 0.940546070506155,
  "best_model_checkpoint": "model_saves/roberta-large_lemon_5k_1_p3/checkpoint-536",
  "epoch": 5.0,
  "global_step": 1340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9403671776294458,
      "eval_loss": 0.42676737904548645,
      "eval_runtime": 3.3616,
      "eval_samples_per_second": 1303.542,
      "eval_steps_per_second": 10.412,
      "step": 268
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.940546070506155,
      "eval_loss": 0.43358537554740906,
      "eval_runtime": 3.365,
      "eval_samples_per_second": 1302.228,
      "eval_steps_per_second": 10.401,
      "step": 536
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9396739677321974,
      "eval_loss": 0.46013274788856506,
      "eval_runtime": 3.3631,
      "eval_samples_per_second": 1302.97,
      "eval_steps_per_second": 10.407,
      "step": 804
    },
    {
      "epoch": 3.73,
      "learning_rate": 1e-05,
      "loss": 0.3185,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9386956473126936,
      "eval_loss": 0.49785834550857544,
      "eval_runtime": 3.3607,
      "eval_samples_per_second": 1303.893,
      "eval_steps_per_second": 10.414,
      "step": 1072
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9384608504120127,
      "eval_loss": 0.5383838415145874,
      "eval_runtime": 3.3614,
      "eval_samples_per_second": 1303.635,
      "eval_steps_per_second": 10.412,
      "step": 1340
    },
    {
      "epoch": 5.0,
      "step": 1340,
      "total_flos": 1.7336177338089472e+16,
      "train_loss": 0.289523668431524,
      "train_runtime": 403.2725,
      "train_samples_per_second": 1275.961,
      "train_steps_per_second": 9.968
    }
  ],
  "max_steps": 4020,
  "num_train_epochs": 15,
  "total_flos": 1.7336177338089472e+16,
  "trial_name": null,
  "trial_params": null
}