checkpoint-1953/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1953,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "learning_rate": 0.00013061224489795917,
      "loss": 1.1362,
      "step": 130
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.000192942515651679,
      "loss": 0.6999,
      "step": 260
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00017814456459874786,
      "loss": 0.6471,
      "step": 390
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00016334661354581675,
      "loss": 0.6069,
      "step": 520
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0001485486624928856,
      "loss": 0.6069,
      "step": 650
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00013375071143995447,
      "loss": 0.5496,
      "step": 780
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00011895276038702335,
      "loss": 0.5527,
      "step": 910
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00010415480933409221,
      "loss": 0.5454,
      "step": 1040
    },
    {
      "epoch": 1.8,
      "learning_rate": 8.935685828116107e-05,
      "loss": 0.5216,
      "step": 1170
    },
    {
      "epoch": 2.0,
      "learning_rate": 7.455890722822995e-05,
      "loss": 0.4949,
      "step": 1300
    },
    {
      "epoch": 2.2,
      "learning_rate": 5.976095617529881e-05,
      "loss": 0.4732,
      "step": 1430
    },
    {
      "epoch": 2.4,
      "learning_rate": 4.4963005122367675e-05,
      "loss": 0.4562,
      "step": 1560
    },
    {
      "epoch": 2.6,
      "learning_rate": 3.0165054069436543e-05,
      "loss": 0.4539,
      "step": 1690
    },
    {
      "epoch": 2.8,
      "learning_rate": 1.5367103016505408e-05,
      "loss": 0.4469,
      "step": 1820
    },
    {
      "epoch": 3.0,
      "learning_rate": 5.691519635742744e-07,
      "loss": 0.4726,
      "step": 1950
    }
  ],
  "logging_steps": 130,
  "max_steps": 1953,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 1.5876741161130394e+17,
  "trial_name": null,
  "trial_params": null
}