checkpoint-35/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 35,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.29,
      "learning_rate": 0.0001,
      "loss": 3.5616,
      "step": 2
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 3.4573,
      "step": 4
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0001870967741935484,
      "loss": 3.1267,
      "step": 6
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00017419354838709678,
      "loss": 2.564,
      "step": 8
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00016129032258064516,
      "loss": 2.1927,
      "step": 10
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00016129032258064516,
      "loss": 1.9417,
      "step": 12
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00015483870967741937,
      "loss": 1.9339,
      "step": 14
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.00014193548387096775,
      "loss": 1.4021,
      "step": 16
    },
    {
      "epoch": 2.57,
      "learning_rate": 0.00012903225806451613,
      "loss": 1.1358,
      "step": 18
    },
    {
      "epoch": 2.86,
      "learning_rate": 0.00011612903225806453,
      "loss": 1.0569,
      "step": 20
    },
    {
      "epoch": 3.14,
      "learning_rate": 0.0001032258064516129,
      "loss": 1.0232,
      "step": 22
    },
    {
      "epoch": 3.43,
      "learning_rate": 9.032258064516129e-05,
      "loss": 0.8228,
      "step": 24
    },
    {
      "epoch": 3.71,
      "learning_rate": 7.741935483870968e-05,
      "loss": 0.8717,
      "step": 26
    },
    {
      "epoch": 4.0,
      "learning_rate": 6.451612903225807e-05,
      "loss": 0.9224,
      "step": 28
    },
    {
      "epoch": 4.29,
      "learning_rate": 5.161290322580645e-05,
      "loss": 0.8255,
      "step": 30
    },
    {
      "epoch": 4.57,
      "learning_rate": 3.870967741935484e-05,
      "loss": 0.8517,
      "step": 32
    },
    {
      "epoch": 4.86,
      "learning_rate": 2.5806451612903226e-05,
      "loss": 0.7775,
      "step": 34
    }
  ],
  "logging_steps": 2,
  "max_steps": 35,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1.18070383214592e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}