trainer_state.json

{
  "best_metric": 0.08537869900465012,
  "best_model_checkpoint": "./vit-strawberries/checkpoint-40",
  "epoch": 22.0,
  "eval_steps": 40,
  "global_step": 110,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.0,
      "grad_norm": 0.6375295519828796,
      "learning_rate": 0.00018181818181818183,
      "loss": 0.3845,
      "step": 10
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.1383754014968872,
      "learning_rate": 0.00016363636363636366,
      "loss": 0.0393,
      "step": 20
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.08163678646087646,
      "learning_rate": 0.00014545454545454546,
      "loss": 0.0128,
      "step": 30
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.06157604977488518,
      "learning_rate": 0.00012727272727272728,
      "loss": 0.0087,
      "step": 40
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9230769230769231,
      "eval_loss": 0.08537869900465012,
      "eval_runtime": 0.4612,
      "eval_samples_per_second": 56.375,
      "eval_steps_per_second": 8.673,
      "step": 40
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.05511847883462906,
      "learning_rate": 0.00010909090909090909,
      "loss": 0.007,
      "step": 50
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.047300320118665695,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.0061,
      "step": 60
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.044113993644714355,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.0056,
      "step": 70
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.043533895164728165,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.0052,
      "step": 80
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.9230769230769231,
      "eval_loss": 0.14537294209003448,
      "eval_runtime": 0.4597,
      "eval_samples_per_second": 56.556,
      "eval_steps_per_second": 8.701,
      "step": 80
    },
    {
      "epoch": 18.0,
      "grad_norm": 0.040240347385406494,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.005,
      "step": 90
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.0400131419301033,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.0048,
      "step": 100
    },
    {
      "epoch": 22.0,
      "grad_norm": 0.039389099925756454,
      "learning_rate": 0.0,
      "loss": 0.0048,
      "step": 110
    },
    {
      "epoch": 22.0,
      "step": 110,
      "total_flos": 1.2615695909255578e+17,
      "train_loss": 0.04397649819200689,
      "train_runtime": 84.5517,
      "train_samples_per_second": 19.254,
      "train_steps_per_second": 1.301
    }
  ],
  "logging_steps": 10,
  "max_steps": 110,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 22,
  "save_steps": 40,
  "total_flos": 1.2615695909255578e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}