run-0/checkpoint-4216/trainer_state.json

{
  "best_metric": 0.7958715596330275,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-4216",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 4216,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 5.95961856842041,
      "learning_rate": 9.55389368279823e-06,
      "loss": 1.5369,
      "step": 527
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7339449541284404,
      "eval_loss": 1.2773902416229248,
      "eval_runtime": 2.5957,
      "eval_samples_per_second": 335.939,
      "eval_steps_per_second": 2.697,
      "step": 527
    },
    {
      "epoch": 2.0,
      "grad_norm": 15.252978324890137,
      "learning_rate": 8.492349940265094e-06,
      "loss": 1.2159,
      "step": 1054
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7786697247706422,
      "eval_loss": 1.022659182548523,
      "eval_runtime": 2.5741,
      "eval_samples_per_second": 338.753,
      "eval_steps_per_second": 2.719,
      "step": 1054
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.01114845275879,
      "learning_rate": 7.430806197731956e-06,
      "loss": 1.0132,
      "step": 1581
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7844036697247706,
      "eval_loss": 0.9622328281402588,
      "eval_runtime": 2.5865,
      "eval_samples_per_second": 337.137,
      "eval_steps_per_second": 2.706,
      "step": 1581
    },
    {
      "epoch": 4.0,
      "grad_norm": 15.912079811096191,
      "learning_rate": 6.371276769700781e-06,
      "loss": 0.9206,
      "step": 2108
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7901376146788991,
      "eval_loss": 0.9278557896614075,
      "eval_runtime": 2.5808,
      "eval_samples_per_second": 337.878,
      "eval_steps_per_second": 2.712,
      "step": 2108
    },
    {
      "epoch": 5.0,
      "grad_norm": 10.149016380310059,
      "learning_rate": 5.3097330271676446e-06,
      "loss": 0.8564,
      "step": 2635
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7901376146788991,
      "eval_loss": 0.9079629182815552,
      "eval_runtime": 2.5658,
      "eval_samples_per_second": 339.852,
      "eval_steps_per_second": 2.728,
      "step": 2635
    },
    {
      "epoch": 6.0,
      "grad_norm": 16.081409454345703,
      "learning_rate": 4.248189284634507e-06,
      "loss": 0.8139,
      "step": 3162
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7947247706422018,
      "eval_loss": 0.8956273794174194,
      "eval_runtime": 2.6131,
      "eval_samples_per_second": 333.701,
      "eval_steps_per_second": 2.679,
      "step": 3162
    },
    {
      "epoch": 7.0,
      "grad_norm": 11.411503791809082,
      "learning_rate": 3.1866455421013703e-06,
      "loss": 0.7798,
      "step": 3689
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7924311926605505,
      "eval_loss": 0.8861347436904907,
      "eval_runtime": 2.6019,
      "eval_samples_per_second": 335.138,
      "eval_steps_per_second": 2.69,
      "step": 3689
    },
    {
      "epoch": 8.0,
      "grad_norm": Infinity,
      "learning_rate": 2.127116114070194e-06,
      "loss": 0.7583,
      "step": 4216
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7958715596330275,
      "eval_loss": 0.8804019093513489,
      "eval_runtime": 2.543,
      "eval_samples_per_second": 342.909,
      "eval_steps_per_second": 2.753,
      "step": 4216
    }
  ],
  "logging_steps": 500,
  "max_steps": 5270,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 64687167744780.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.27608826195592573,
    "learning_rate": 1.0615437425331367e-05,
    "num_train_epochs": 10,
    "temperature": 2
  }
}