trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 18.799348289259306,
  "global_step": 150000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.25,
      "learning_rate": 1.164473409366253e-05,
      "loss": 2.9668,
      "step": 10000
    },
    {
      "epoch": 2.51,
      "learning_rate": 3.2894681873250615e-06,
      "loss": 2.3684,
      "step": 20000
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.624013034214814e-05,
      "loss": 1.5028,
      "step": 30000
    },
    {
      "epoch": 5.01,
      "learning_rate": 1.498684045619752e-05,
      "loss": 1.3541,
      "step": 40000
    },
    {
      "epoch": 6.27,
      "learning_rate": 1.3733550570246898e-05,
      "loss": 1.2438,
      "step": 50000
    },
    {
      "epoch": 7.52,
      "learning_rate": 1.2480260684296278e-05,
      "loss": 1.1711,
      "step": 60000
    },
    {
      "epoch": 8.77,
      "learning_rate": 1.1226970798345658e-05,
      "loss": 1.1123,
      "step": 70000
    },
    {
      "epoch": 10.03,
      "learning_rate": 9.973680912395037e-06,
      "loss": 1.0557,
      "step": 80000
    },
    {
      "epoch": 11.28,
      "learning_rate": 8.720391026444417e-06,
      "loss": 1.0005,
      "step": 90000
    },
    {
      "epoch": 12.53,
      "learning_rate": 7.467101140493797e-06,
      "loss": 0.9576,
      "step": 100000
    },
    {
      "epoch": 13.79,
      "learning_rate": 6.213811254543176e-06,
      "loss": 0.9233,
      "step": 110000
    },
    {
      "epoch": 15.04,
      "learning_rate": 4.960521368592556e-06,
      "loss": 0.8876,
      "step": 120000
    },
    {
      "epoch": 16.29,
      "learning_rate": 3.7072314826419353e-06,
      "loss": 0.859,
      "step": 130000
    },
    {
      "epoch": 17.55,
      "learning_rate": 2.453941596691315e-06,
      "loss": 0.8366,
      "step": 140000
    },
    {
      "epoch": 18.8,
      "learning_rate": 1.2006517107406945e-06,
      "loss": 0.818,
      "step": 150000
    }
  ],
  "max_steps": 159580,
  "num_train_epochs": 20,
  "total_flos": 9.303396720850058e+19,
  "trial_name": null,
  "trial_params": null
}