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": 0.2418,
      "step": 50
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.8299319727891158e-05,
      "loss": 0.0112,
      "step": 100
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.7448979591836738e-05,
      "loss": 0.0161,
      "step": 150
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.6598639455782314e-05,
      "loss": 0.0154,
      "step": 200
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.5748299319727894e-05,
      "loss": 0.0234,
      "step": 250
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.4897959183673472e-05,
      "loss": 0.0225,
      "step": 300
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.4047619047619048e-05,
      "loss": 0.0092,
      "step": 350
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.3197278911564626e-05,
      "loss": 0.0103,
      "step": 400
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.2346938775510204e-05,
      "loss": 0.0004,
      "step": 450
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.1496598639455783e-05,
      "loss": 0.0077,
      "step": 500
    },
    {
      "epoch": 1.28,
      "eval_accuracy": 0.999681224099458,
      "eval_loss": 0.0006676316261291504,
      "eval_runtime": 12.272,
      "eval_samples_per_second": 255.622,
      "eval_steps_per_second": 16.053,
      "step": 500
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.064625850340136e-05,
      "loss": 0.0074,
      "step": 550
    },
    {
      "epoch": 1.53,
      "learning_rate": 9.795918367346939e-06,
      "loss": 0.0004,
      "step": 600
    },
    {
      "epoch": 1.66,
      "learning_rate": 8.945578231292518e-06,
      "loss": 0.0009,
      "step": 650
    },
    {
      "epoch": 1.79,
      "learning_rate": 8.095238095238097e-06,
      "loss": 0.0097,
      "step": 700
    },
    {
      "epoch": 1.91,
      "learning_rate": 7.244897959183675e-06,
      "loss": 0.0015,
      "step": 750
    },
    {
      "epoch": 2.04,
      "learning_rate": 6.394557823129253e-06,
      "loss": 0.006,
      "step": 800
    },
    {
      "epoch": 2.17,
      "learning_rate": 5.54421768707483e-06,
      "loss": 0.0003,
      "step": 850
    },
    {
      "epoch": 2.3,
      "learning_rate": 4.693877551020409e-06,
      "loss": 0.0005,
      "step": 900
    },
    {
      "epoch": 2.42,
      "learning_rate": 3.843537414965986e-06,
      "loss": 0.0002,
      "step": 950
    },
    {
      "epoch": 2.55,
      "learning_rate": 2.993197278911565e-06,
      "loss": 0.0043,
      "step": 1000
    },
    {
      "epoch": 2.55,
      "eval_accuracy": 0.999681224099458,
      "eval_loss": 0.0024934401735663414,
      "eval_runtime": 12.305,
      "eval_samples_per_second": 254.937,
      "eval_steps_per_second": 16.01,
      "step": 1000
    },
    {
      "epoch": 2.68,
      "learning_rate": 2.1428571428571427e-06,
      "loss": 0.0002,
      "step": 1050
    },
    {
      "epoch": 2.81,
      "learning_rate": 1.2925170068027212e-06,
      "loss": 0.0002,
      "step": 1100
    },
    {
      "epoch": 2.93,
      "learning_rate": 4.421768707482994e-07,
      "loss": 0.0002,
      "step": 1150
    },
    {
      "epoch": 3.0,
      "step": 1176,
      "total_flos": 3656104681030176.0,
      "train_loss": 0.016569241228019248,
      "train_runtime": 511.7207,
      "train_samples_per_second": 73.435,
      "train_steps_per_second": 2.298
    }
  ],
  "logging_steps": 50,
  "max_steps": 1176,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 3656104681030176.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}