trainer_state.json

{
  "best_metric": 0.6237846612930298,
  "best_model_checkpoint": "tiny_bert_km_100_v1_cola/checkpoint-34",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 204,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.7257574200630188,
      "learning_rate": 4.9e-05,
      "loss": 0.6196,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6237846612930298,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3991,
      "eval_samples_per_second": 2613.323,
      "eval_steps_per_second": 12.528,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.121646761894226,
      "learning_rate": 4.8e-05,
      "loss": 0.6075,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6263473629951477,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.4095,
      "eval_samples_per_second": 2546.705,
      "eval_steps_per_second": 12.209,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.8357798457145691,
      "learning_rate": 4.7e-05,
      "loss": 0.5972,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6293219923973083,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3815,
      "eval_samples_per_second": 2733.875,
      "eval_steps_per_second": 13.106,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1285151243209839,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5804,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.668264627456665,
      "eval_loss": 0.635418713092804,
      "eval_matthews_correlation": -0.019583278456274576,
      "eval_runtime": 0.481,
      "eval_samples_per_second": 2168.225,
      "eval_steps_per_second": 10.394,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.3463857173919678,
      "learning_rate": 4.5e-05,
      "loss": 0.5561,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6673058271408081,
      "eval_loss": 0.6380937099456787,
      "eval_matthews_correlation": 0.036081175524456124,
      "eval_runtime": 0.3833,
      "eval_samples_per_second": 2721.024,
      "eval_steps_per_second": 13.044,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.237225294113159,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.5088,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6615532040596008,
      "eval_loss": 0.6592812538146973,
      "eval_matthews_correlation": 0.05373746752832175,
      "eval_runtime": 0.4768,
      "eval_samples_per_second": 2187.53,
      "eval_steps_per_second": 10.487,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 1345426898614272.0,
      "train_loss": 0.5782780647277832,
      "train_runtime": 41.8478,
      "train_samples_per_second": 10216.779,
      "train_steps_per_second": 40.623
    }
  ],
  "logging_steps": 1,
  "max_steps": 1700,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 1345426898614272.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}