trainer_state.json

{
  "best_metric": 0.621593713760376,
  "best_model_checkpoint": "tiny_bert_km_50_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": 0.7628885507583618,
      "learning_rate": 4.9e-05,
      "loss": 0.6224,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.621593713760376,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.4018,
      "eval_samples_per_second": 2595.926,
      "eval_steps_per_second": 12.445,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0041512250900269,
      "learning_rate": 4.8e-05,
      "loss": 0.6083,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6220319271087646,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.4197,
      "eval_samples_per_second": 2485.372,
      "eval_steps_per_second": 11.915,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9402241110801697,
      "learning_rate": 4.7e-05,
      "loss": 0.5982,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6228198409080505,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.4066,
      "eval_samples_per_second": 2564.868,
      "eval_steps_per_second": 12.296,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.9954250454902649,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5783,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6883988380432129,
      "eval_loss": 0.6298257112503052,
      "eval_matthews_correlation": 0.08585482785878726,
      "eval_runtime": 0.4156,
      "eval_samples_per_second": 2509.564,
      "eval_steps_per_second": 12.031,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.6896628141403198,
      "learning_rate": 4.5e-05,
      "loss": 0.5458,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6788111329078674,
      "eval_loss": 0.6444757580757141,
      "eval_matthews_correlation": 0.058343519603323714,
      "eval_runtime": 0.4047,
      "eval_samples_per_second": 2576.964,
      "eval_steps_per_second": 12.354,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.9982967376708984,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.501,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6586769223213196,
      "eval_loss": 0.669975221157074,
      "eval_matthews_correlation": 0.11363806297455119,
      "eval_runtime": 0.413,
      "eval_samples_per_second": 2525.728,
      "eval_steps_per_second": 12.108,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 1345426898614272.0,
      "train_loss": 0.5756891848994237,
      "train_runtime": 42.7732,
      "train_samples_per_second": 9995.753,
      "train_steps_per_second": 39.745
    }
  ],
  "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
}