trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 10,
  "global_step": 36,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "learning_rate": 0.0,
      "loss": 1.796,
      "step": 1
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0001,
      "loss": 1.0996,
      "step": 2
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0001,
      "loss": 0.9834,
      "step": 3
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0001,
      "loss": 0.558,
      "step": 4
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0001,
      "loss": 0.3067,
      "step": 5
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0001,
      "loss": 0.4554,
      "step": 6
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0001,
      "loss": 0.4328,
      "step": 7
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0001,
      "loss": 0.2377,
      "step": 8
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0001,
      "loss": 0.3933,
      "step": 9
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0001,
      "loss": 0.3539,
      "step": 10
    },
    {
      "epoch": 0.83,
      "eval_accuracy": 0.01065855981246795,
      "eval_loss": 0.3053041696548462,
      "eval_runtime": 2.9764,
      "eval_samples_per_second": 7.391,
      "eval_steps_per_second": 0.336,
      "step": 10
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0001,
      "loss": 0.3407,
      "step": 11
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0001,
      "loss": 0.3791,
      "step": 12
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0001,
      "loss": 0.2567,
      "step": 13
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0001,
      "loss": 0.3081,
      "step": 14
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0001,
      "loss": 0.2464,
      "step": 15
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0001,
      "loss": 0.295,
      "step": 16
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0001,
      "loss": 0.243,
      "step": 17
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0001,
      "loss": 0.2889,
      "step": 18
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0001,
      "loss": 0.2491,
      "step": 19
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0001,
      "loss": 0.1398,
      "step": 20
    },
    {
      "epoch": 1.67,
      "eval_accuracy": 0.0107684418723903,
      "eval_loss": 0.2902178466320038,
      "eval_runtime": 2.52,
      "eval_samples_per_second": 8.73,
      "eval_steps_per_second": 0.397,
      "step": 20
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0001,
      "loss": 0.255,
      "step": 21
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0001,
      "loss": 0.233,
      "step": 22
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0001,
      "loss": 0.2733,
      "step": 23
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0001,
      "loss": 0.2045,
      "step": 24
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0001,
      "loss": 0.1885,
      "step": 25
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0001,
      "loss": 0.2345,
      "step": 26
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0001,
      "loss": 0.2209,
      "step": 27
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.0001,
      "loss": 0.2146,
      "step": 28
    },
    {
      "epoch": 2.42,
      "learning_rate": 0.0001,
      "loss": 0.1555,
      "step": 29
    },
    {
      "epoch": 2.5,
      "learning_rate": 0.0001,
      "loss": 0.2355,
      "step": 30
    },
    {
      "epoch": 2.5,
      "eval_accuracy": 0.0107684418723903,
      "eval_loss": 0.27231642603874207,
      "eval_runtime": 2.1494,
      "eval_samples_per_second": 10.235,
      "eval_steps_per_second": 0.465,
      "step": 30
    },
    {
      "epoch": 2.58,
      "learning_rate": 0.0001,
      "loss": 0.1539,
      "step": 31
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.0001,
      "loss": 0.119,
      "step": 32
    },
    {
      "epoch": 2.75,
      "learning_rate": 0.0001,
      "loss": 0.2046,
      "step": 33
    },
    {
      "epoch": 2.83,
      "learning_rate": 0.0001,
      "loss": 0.1672,
      "step": 34
    },
    {
      "epoch": 2.92,
      "learning_rate": 0.0001,
      "loss": 0.1556,
      "step": 35
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0001,
      "loss": 0.161,
      "step": 36
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 60,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 7706157613056.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}