trainer_state.json

{
  "best_metric": 0.7435600757598877,
  "best_model_checkpoint": "output/queen/checkpoint-69",
  "epoch": 1.0,
  "global_step": 69,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 0.00013537785052914348,
      "loss": 0.8946,
      "step": 5
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0001300082017869573,
      "loss": 1.1571,
      "step": 10
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00012137631040942562,
      "loss": 1.0891,
      "step": 15
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00010994073605561692,
      "loss": 1.0026,
      "step": 20
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.63089809342121e-05,
      "loss": 0.8956,
      "step": 25
    },
    {
      "epoch": 0.44,
      "learning_rate": 8.120521692221655e-05,
      "loss": 1.0037,
      "step": 30
    },
    {
      "epoch": 0.51,
      "learning_rate": 6.543181473690228e-05,
      "loss": 0.9259,
      "step": 35
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.982671888105533e-05,
      "loss": 0.9391,
      "step": 40
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.52189327703136e-05,
      "loss": 0.897,
      "step": 45
    },
    {
      "epoch": 0.74,
      "learning_rate": 2.2384478845846266e-05,
      "loss": 0.969,
      "step": 50
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.2005173242556645e-05,
      "loss": 0.9795,
      "step": 55
    },
    {
      "epoch": 0.88,
      "learning_rate": 4.632405062861206e-06,
      "loss": 1.0966,
      "step": 60
    },
    {
      "epoch": 0.96,
      "learning_rate": 6.578444419609618e-07,
      "loss": 0.9097,
      "step": 65
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.9767317771911621,
      "eval_runtime": 2.0747,
      "eval_samples_per_second": 46.272,
      "eval_steps_per_second": 5.784,
      "step": 68
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.7435600757598877,
      "eval_runtime": 1.811,
      "eval_samples_per_second": 48.039,
      "eval_steps_per_second": 6.074,
      "step": 69
    }
  ],
  "max_steps": 138,
  "num_train_epochs": 2,
  "total_flos": 70418202624000.0,
  "trial_name": null,
  "trial_params": null
}