checkpoint-800/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 26.016666666666666,
  "global_step": 800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 11.1049,
      "step": 1
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00025,
      "loss": 8.1845,
      "step": 50
    },
    {
      "epoch": 3.01,
      "learning_rate": 0.0005,
      "loss": 5.9529,
      "step": 100
    },
    {
      "epoch": 4.03,
      "learning_rate": 0.0005833333333333333,
      "loss": 5.1965,
      "step": 150
    },
    {
      "epoch": 6.02,
      "learning_rate": 0.0005555555555555556,
      "loss": 4.8925,
      "step": 200
    },
    {
      "epoch": 8.01,
      "learning_rate": 0.0005277777777777777,
      "loss": 4.7182,
      "step": 250
    },
    {
      "epoch": 9.03,
      "learning_rate": 0.0005,
      "loss": 4.561,
      "step": 300
    },
    {
      "epoch": 11.02,
      "learning_rate": 0.00047222222222222224,
      "loss": 4.4282,
      "step": 350
    },
    {
      "epoch": 13.01,
      "learning_rate": 0.00044444444444444436,
      "loss": 4.2476,
      "step": 400
    },
    {
      "epoch": 14.03,
      "learning_rate": 0.00041666666666666664,
      "loss": 4.0259,
      "step": 450
    },
    {
      "epoch": 16.02,
      "learning_rate": 0.00038888888888888887,
      "loss": 3.8942,
      "step": 500
    },
    {
      "epoch": 18.01,
      "learning_rate": 0.0003611111111111111,
      "loss": 3.7583,
      "step": 550
    },
    {
      "epoch": 19.02,
      "learning_rate": 0.0003333333333333333,
      "loss": 3.6224,
      "step": 600
    },
    {
      "epoch": 21.02,
      "learning_rate": 0.00030555555555555555,
      "loss": 3.55,
      "step": 650
    },
    {
      "epoch": 23.01,
      "learning_rate": 0.0002777777777777778,
      "loss": 3.4646,
      "step": 700
    },
    {
      "epoch": 24.02,
      "learning_rate": 0.00025,
      "loss": 3.3628,
      "step": 750
    },
    {
      "epoch": 26.02,
      "learning_rate": 0.00022222222222222218,
      "loss": 3.3208,
      "step": 800
    }
  ],
  "max_steps": 1200,
  "num_train_epochs": 9223372036854775807,
  "total_flos": 2.16523300405248e+17,
  "trial_name": null,
  "trial_params": null
}