trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 7611,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "learning_rate": 4.6715280515044016e-05,
      "loss": 3.8283,
      "step": 500
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.343056103008804e-05,
      "loss": 2.9065,
      "step": 1000
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.0145841545132044e-05,
      "loss": 2.5591,
      "step": 1500
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.6861122060176065e-05,
      "loss": 2.4013,
      "step": 2000
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.357640257522008e-05,
      "loss": 2.2547,
      "step": 2500
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.0291683090264093e-05,
      "loss": 1.9816,
      "step": 3000
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.7006963605308104e-05,
      "loss": 1.8817,
      "step": 3500
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.372224412035212e-05,
      "loss": 1.8714,
      "step": 4000
    },
    {
      "epoch": 1.77,
      "learning_rate": 2.043752463539614e-05,
      "loss": 1.8331,
      "step": 4500
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.7152805150440153e-05,
      "loss": 1.7802,
      "step": 5000
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3868085665484169e-05,
      "loss": 1.553,
      "step": 5500
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.0583366180528183e-05,
      "loss": 1.5374,
      "step": 6000
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.298646695572199e-06,
      "loss": 1.5077,
      "step": 6500
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.013927210616213e-06,
      "loss": 1.5056,
      "step": 7000
    },
    {
      "epoch": 2.96,
      "learning_rate": 7.292077256602286e-07,
      "loss": 1.5086,
      "step": 7500
    },
    {
      "epoch": 3.0,
      "step": 7611,
      "total_flos": 1.2382261916860416e+16,
      "train_loss": 2.052058351100971,
      "train_runtime": 2409.5622,
      "train_samples_per_second": 31.577,
      "train_steps_per_second": 3.159
    }
  ],
  "max_steps": 7611,
  "num_train_epochs": 3,
  "total_flos": 1.2382261916860416e+16,
  "trial_name": null,
  "trial_params": null
}