trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 72,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 4.645751476287842,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 1.6038,
      "step": 5
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.9989657998085022,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.5001,
      "step": 10
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.7485830783843994,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.3532,
      "step": 15
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.6910512447357178,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.3014,
      "step": 20
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 0.622086226940155,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.2701,
      "step": 25
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.625781238079071,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.2061,
      "step": 30
    },
    {
      "epoch": 1.4583333333333333,
      "grad_norm": 0.6655622720718384,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.1961,
      "step": 35
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.5766973495483398,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.1949,
      "step": 40
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.48014524579048157,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.1787,
      "step": 45
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 0.4697082042694092,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.1609,
      "step": 50
    },
    {
      "epoch": 2.2916666666666665,
      "grad_norm": 0.45880624651908875,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.133,
      "step": 55
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.5088185667991638,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.1295,
      "step": 60
    },
    {
      "epoch": 2.7083333333333335,
      "grad_norm": 0.45895683765411377,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.1243,
      "step": 65
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 0.4871689975261688,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.1238,
      "step": 70
    },
    {
      "epoch": 3.0,
      "step": 72,
      "total_flos": 9.35320799412224e+16,
      "train_loss": 0.31486400547954774,
      "train_runtime": 306.1511,
      "train_samples_per_second": 14.532,
      "train_steps_per_second": 0.235
    }
  ],
  "logging_steps": 5,
  "max_steps": 72,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.35320799412224e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}