trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9993761696818466,
  "eval_steps": 500,
  "global_step": 10818,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.28,
      "learning_rate": 9.078217455621301e-06,
      "loss": 0.5163,
      "step": 1000
    },
    {
      "epoch": 0.55,
      "learning_rate": 8.153661242603551e-06,
      "loss": 0.4627,
      "step": 2000
    },
    {
      "epoch": 0.83,
      "learning_rate": 7.229105029585799e-06,
      "loss": 0.4393,
      "step": 3000
    },
    {
      "epoch": 1.11,
      "learning_rate": 6.304548816568048e-06,
      "loss": 0.3697,
      "step": 4000
    },
    {
      "epoch": 1.39,
      "learning_rate": 5.379992603550296e-06,
      "loss": 0.2788,
      "step": 5000
    },
    {
      "epoch": 1.66,
      "learning_rate": 4.455436390532545e-06,
      "loss": 0.2757,
      "step": 6000
    },
    {
      "epoch": 1.94,
      "learning_rate": 3.530880177514793e-06,
      "loss": 0.2708,
      "step": 7000
    },
    {
      "epoch": 2.22,
      "learning_rate": 2.6063239644970415e-06,
      "loss": 0.1397,
      "step": 8000
    },
    {
      "epoch": 2.5,
      "learning_rate": 1.6817677514792902e-06,
      "loss": 0.1026,
      "step": 9000
    },
    {
      "epoch": 2.77,
      "learning_rate": 7.572115384615385e-07,
      "loss": 0.0998,
      "step": 10000
    },
    {
      "epoch": 3.0,
      "step": 10818,
      "total_flos": 3075736122425344.0,
      "train_loss": 0.28055200490389104,
      "train_runtime": 252630.326,
      "train_samples_per_second": 5.482,
      "train_steps_per_second": 0.043
    }
  ],
  "logging_steps": 1000,
  "max_steps": 10818,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "total_flos": 3075736122425344.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}