trainer_state.json

{
  "best_metric": 0.6364051699638367,
  "best_model_checkpoint": "./vit-base-beans/checkpoint-120",
  "epoch": 3.5294117647058822,
  "global_step": 120,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.29,
      "learning_rate": 0.00018529411764705883,
      "loss": 3.1752,
      "step": 10
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00017058823529411766,
      "loss": 2.7524,
      "step": 20
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00015588235294117648,
      "loss": 2.3486,
      "step": 30
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0001411764705882353,
      "loss": 1.8512,
      "step": 40
    },
    {
      "epoch": 1.18,
      "eval_accuracy": 0.7372881355932204,
      "eval_loss": 1.7551679611206055,
      "eval_runtime": 80.0325,
      "eval_samples_per_second": 1.474,
      "eval_steps_per_second": 0.187,
      "step": 40
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0001264705882352941,
      "loss": 1.5475,
      "step": 50
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00011176470588235294,
      "loss": 1.2493,
      "step": 60
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.705882352941177e-05,
      "loss": 1.0852,
      "step": 70
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.23529411764706e-05,
      "loss": 0.8329,
      "step": 80
    },
    {
      "epoch": 2.35,
      "eval_accuracy": 0.9915254237288136,
      "eval_loss": 0.9235942363739014,
      "eval_runtime": 79.6393,
      "eval_samples_per_second": 1.482,
      "eval_steps_per_second": 0.188,
      "step": 80
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.764705882352942e-05,
      "loss": 0.7635,
      "step": 90
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.294117647058824e-05,
      "loss": 0.6581,
      "step": 100
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.8235294117647055e-05,
      "loss": 0.5925,
      "step": 110
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.5486,
      "step": 120
    },
    {
      "epoch": 3.53,
      "eval_accuracy": 0.9915254237288136,
      "eval_loss": 0.6364051699638367,
      "eval_runtime": 80.6234,
      "eval_samples_per_second": 1.464,
      "eval_steps_per_second": 0.186,
      "step": 120
    }
  ],
  "max_steps": 136,
  "num_train_epochs": 4,
  "total_flos": 1.467238905086976e+17,
  "trial_name": null,
  "trial_params": null
}