trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 6315,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.24,
      "learning_rate": 1.8416468725257326e-05,
      "loss": 0.2819,
      "step": 500
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.6832937450514647e-05,
      "loss": 0.1863,
      "step": 1000
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.5249406175771972e-05,
      "loss": 0.1603,
      "step": 1500
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.3665874901029297e-05,
      "loss": 0.1523,
      "step": 2000
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.208234362628662e-05,
      "loss": 0.1108,
      "step": 2500
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.0498812351543943e-05,
      "loss": 0.1091,
      "step": 3000
    },
    {
      "epoch": 1.66,
      "learning_rate": 8.915281076801267e-06,
      "loss": 0.1132,
      "step": 3500
    },
    {
      "epoch": 1.9,
      "learning_rate": 7.331749802058591e-06,
      "loss": 0.1071,
      "step": 4000
    },
    {
      "epoch": 2.14,
      "learning_rate": 5.748218527315916e-06,
      "loss": 0.0904,
      "step": 4500
    },
    {
      "epoch": 2.38,
      "learning_rate": 4.164687252573238e-06,
      "loss": 0.0743,
      "step": 5000
    },
    {
      "epoch": 2.61,
      "learning_rate": 2.581155977830562e-06,
      "loss": 0.0785,
      "step": 5500
    },
    {
      "epoch": 2.85,
      "learning_rate": 9.976247030878861e-07,
      "loss": 0.0701,
      "step": 6000
    },
    {
      "epoch": 3.0,
      "step": 6315,
      "total_flos": 1.6988827310258688e+16,
      "train_runtime": 1525.1643,
      "train_samples_per_second": 4.141
    }
  ],
  "max_steps": 6315,
  "num_train_epochs": 3,
  "total_flos": 1.6988827310258688e+16,
  "trial_name": null,
  "trial_params": null
}