checkpoint-20/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 19.5,
  "eval_steps": 500,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "learning_rate": 5e-05,
      "loss": 1.7498,
      "step": 1
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0001,
      "loss": 1.7498,
      "step": 2
    },
    {
      "epoch": 2.5,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.7179,
      "step": 3
    },
    {
      "epoch": 3.5,
      "learning_rate": 0.0002,
      "loss": 1.6526,
      "step": 4
    },
    {
      "epoch": 4.5,
      "learning_rate": 0.00019444444444444446,
      "loss": 1.5549,
      "step": 5
    },
    {
      "epoch": 5.5,
      "learning_rate": 0.00018888888888888888,
      "loss": 1.4428,
      "step": 6
    },
    {
      "epoch": 6.5,
      "learning_rate": 0.00018333333333333334,
      "loss": 1.3413,
      "step": 7
    },
    {
      "epoch": 7.5,
      "learning_rate": 0.00017777777777777779,
      "loss": 1.2458,
      "step": 8
    },
    {
      "epoch": 8.5,
      "learning_rate": 0.00017222222222222224,
      "loss": 1.1408,
      "step": 9
    },
    {
      "epoch": 9.5,
      "learning_rate": 0.0001666666666666667,
      "loss": 1.0299,
      "step": 10
    },
    {
      "epoch": 10.5,
      "learning_rate": 0.0001611111111111111,
      "loss": 0.9223,
      "step": 11
    },
    {
      "epoch": 11.5,
      "learning_rate": 0.00015555555555555556,
      "loss": 0.8215,
      "step": 12
    },
    {
      "epoch": 12.5,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.7281,
      "step": 13
    },
    {
      "epoch": 13.5,
      "learning_rate": 0.00014444444444444444,
      "loss": 0.6339,
      "step": 14
    },
    {
      "epoch": 14.5,
      "learning_rate": 0.0001388888888888889,
      "loss": 0.5883,
      "step": 15
    },
    {
      "epoch": 15.5,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.4937,
      "step": 16
    },
    {
      "epoch": 16.5,
      "learning_rate": 0.00012777777777777776,
      "loss": 0.4356,
      "step": 17
    },
    {
      "epoch": 17.5,
      "learning_rate": 0.00012222222222222224,
      "loss": 0.3726,
      "step": 18
    },
    {
      "epoch": 18.5,
      "learning_rate": 0.00011666666666666668,
      "loss": 0.319,
      "step": 19
    },
    {
      "epoch": 19.5,
      "learning_rate": 0.00011111111111111112,
      "loss": 0.2746,
      "step": 20
    }
  ],
  "logging_steps": 1,
  "max_steps": 40,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 1749190862438400.0,
  "trial_name": null,
  "trial_params": null
}