trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1176,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13,
      "learning_rate": 1.914965986394558e-05,
      "loss": 0.282,
      "step": 50
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.8299319727891158e-05,
      "loss": 0.0459,
      "step": 100
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.7448979591836738e-05,
      "loss": 0.0198,
      "step": 150
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.6598639455782314e-05,
      "loss": 0.0244,
      "step": 200
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.5748299319727894e-05,
      "loss": 0.0128,
      "step": 250
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.4897959183673472e-05,
      "loss": 0.0226,
      "step": 300
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.4047619047619048e-05,
      "loss": 0.018,
      "step": 350
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.3197278911564626e-05,
      "loss": 0.0089,
      "step": 400
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.2346938775510204e-05,
      "loss": 0.0164,
      "step": 450
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.1496598639455783e-05,
      "loss": 0.0043,
      "step": 500
    },
    {
      "epoch": 1.28,
      "eval_accuracy": 0.999681224099458,
      "eval_loss": 0.0025624127592891455,
      "eval_runtime": 16.1354,
      "eval_samples_per_second": 194.418,
      "eval_steps_per_second": 12.209,
      "step": 500
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.064625850340136e-05,
      "loss": 0.0267,
      "step": 550
    },
    {
      "epoch": 1.53,
      "learning_rate": 9.795918367346939e-06,
      "loss": 0.0096,
      "step": 600
    },
    {
      "epoch": 1.66,
      "learning_rate": 8.945578231292518e-06,
      "loss": 0.0046,
      "step": 650
    },
    {
      "epoch": 1.79,
      "learning_rate": 8.095238095238097e-06,
      "loss": 0.0189,
      "step": 700
    },
    {
      "epoch": 1.91,
      "learning_rate": 7.244897959183675e-06,
      "loss": 0.011,
      "step": 750
    },
    {
      "epoch": 2.04,
      "learning_rate": 6.394557823129253e-06,
      "loss": 0.0001,
      "step": 800
    },
    {
      "epoch": 2.17,
      "learning_rate": 5.54421768707483e-06,
      "loss": 0.0041,
      "step": 850
    },
    {
      "epoch": 2.3,
      "learning_rate": 4.693877551020409e-06,
      "loss": 0.0001,
      "step": 900
    },
    {
      "epoch": 2.42,
      "learning_rate": 3.843537414965986e-06,
      "loss": 0.0034,
      "step": 950
    },
    {
      "epoch": 2.55,
      "learning_rate": 2.993197278911565e-06,
      "loss": 0.0,
      "step": 1000
    },
    {
      "epoch": 2.55,
      "eval_accuracy": 0.999681224099458,
      "eval_loss": 0.0006749434396624565,
      "eval_runtime": 16.127,
      "eval_samples_per_second": 194.519,
      "eval_steps_per_second": 12.216,
      "step": 1000
    },
    {
      "epoch": 2.68,
      "learning_rate": 2.1428571428571427e-06,
      "loss": 0.0,
      "step": 1050
    },
    {
      "epoch": 2.81,
      "learning_rate": 1.2925170068027212e-06,
      "loss": 0.0,
      "step": 1100
    },
    {
      "epoch": 2.93,
      "learning_rate": 4.421768707482994e-07,
      "loss": 0.0,
      "step": 1150
    },
    {
      "epoch": 3.0,
      "step": 1176,
      "total_flos": 7498385275761072.0,
      "train_loss": 0.022714921345852703,
      "train_runtime": 755.4145,
      "train_samples_per_second": 49.745,
      "train_steps_per_second": 1.557
    }
  ],
  "logging_steps": 50,
  "max_steps": 1176,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 7498385275761072.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}