trainer_state.json

{
  "best_metric": 0.4721340835094452,
  "best_model_checkpoint": "tiny_bert_km_50_v1_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 5.747946739196777,
      "learning_rate": 4.9e-05,
      "loss": 0.4666,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7855504587155964,
      "eval_loss": 0.4721340835094452,
      "eval_runtime": 0.3669,
      "eval_samples_per_second": 2376.728,
      "eval_steps_per_second": 10.902,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.643387794494629,
      "learning_rate": 4.8e-05,
      "loss": 0.2544,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8004587155963303,
      "eval_loss": 0.522304356098175,
      "eval_runtime": 0.3235,
      "eval_samples_per_second": 2695.928,
      "eval_steps_per_second": 12.367,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.789928913116455,
      "learning_rate": 4.7e-05,
      "loss": 0.1989,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8027522935779816,
      "eval_loss": 0.5052730441093445,
      "eval_runtime": 0.3266,
      "eval_samples_per_second": 2670.238,
      "eval_steps_per_second": 12.249,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.88606595993042,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1622,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.805045871559633,
      "eval_loss": 0.5874336361885071,
      "eval_runtime": 0.3218,
      "eval_samples_per_second": 2709.525,
      "eval_steps_per_second": 12.429,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 10.503990173339844,
      "learning_rate": 4.5e-05,
      "loss": 0.1356,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7912844036697247,
      "eval_loss": 0.6845178008079529,
      "eval_runtime": 0.3229,
      "eval_samples_per_second": 2700.711,
      "eval_steps_per_second": 12.389,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 6.96405553817749,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1131,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7912844036697247,
      "eval_loss": 0.7334715723991394,
      "eval_runtime": 0.3192,
      "eval_samples_per_second": 2732.097,
      "eval_steps_per_second": 12.533,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.22179452819053572,
      "train_runtime": 275.0501,
      "train_samples_per_second": 12243.041,
      "train_steps_per_second": 47.991
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 1.0596790573590528e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}