trainer_state.json

{
  "best_metric": 0.4515226483345032,
  "best_model_checkpoint": "./MRI_vit/checkpoint-100",
  "epoch": 10.0,
  "eval_steps": 100,
  "global_step": 180,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 7.522036552429199,
      "learning_rate": 0.00018888888888888888,
      "loss": 0.974,
      "step": 10
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 2.3464157581329346,
      "learning_rate": 0.00017777777777777779,
      "loss": 0.6625,
      "step": 20
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 9.9130220413208,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.4228,
      "step": 30
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 9.101266860961914,
      "learning_rate": 0.00015555555555555556,
      "loss": 0.4918,
      "step": 40
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 7.912335395812988,
      "learning_rate": 0.00014444444444444444,
      "loss": 0.1294,
      "step": 50
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 1.002270221710205,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.1837,
      "step": 60
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.04989054054021835,
      "learning_rate": 0.00012222222222222224,
      "loss": 0.0559,
      "step": 70
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 7.236814498901367,
      "learning_rate": 0.00011111111111111112,
      "loss": 0.0396,
      "step": 80
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.21504545211792,
      "learning_rate": 0.0001,
      "loss": 0.0791,
      "step": 90
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 0.02051878347992897,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.0236,
      "step": 100
    },
    {
      "epoch": 5.555555555555555,
      "eval_accuracy": 0.8941176470588236,
      "eval_loss": 0.4515226483345032,
      "eval_runtime": 1.7177,
      "eval_samples_per_second": 49.485,
      "eval_steps_per_second": 6.404,
      "step": 100
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 0.013602614402770996,
      "learning_rate": 7.777777777777778e-05,
      "loss": 0.0038,
      "step": 110
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.006600651424378157,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0013,
      "step": 120
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 0.004217671230435371,
      "learning_rate": 5.555555555555556e-05,
      "loss": 0.001,
      "step": 130
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 0.009464713744819164,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.0019,
      "step": 140
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.008202188648283482,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0003,
      "step": 150
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 0.019373826682567596,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.0004,
      "step": 160
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 0.004279292654246092,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.0004,
      "step": 170
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.004037588369101286,
      "learning_rate": 0.0,
      "loss": 0.0004,
      "step": 180
    },
    {
      "epoch": 10.0,
      "step": 180,
      "total_flos": 2.208541498297344e+17,
      "train_loss": 0.17066083699982199,
      "train_runtime": 133.3595,
      "train_samples_per_second": 21.371,
      "train_steps_per_second": 1.35
    }
  ],
  "logging_steps": 10,
  "max_steps": 180,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "total_flos": 2.208541498297344e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}