last-checkpoint/trainer_state.json

{
  "best_metric": 3.539703845977783,
  "best_model_checkpoint": "/content/drive/MyDrive/brabant-from-scratch/bi-objective/checkpoint-5687",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 5687,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09,
      "grad_norm": 7.530444622039795,
      "learning_rate": 4.978020045718305e-05,
      "loss": 0.3375,
      "step": 500
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.392064094543457,
      "learning_rate": 4.9560400914366104e-05,
      "loss": 0.3686,
      "step": 1000
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.53122615814209,
      "learning_rate": 4.934060137154915e-05,
      "loss": 0.361,
      "step": 1500
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.576234340667725,
      "learning_rate": 4.91208018287322e-05,
      "loss": 0.3669,
      "step": 2000
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.397663593292236,
      "learning_rate": 4.890100228591525e-05,
      "loss": 0.3608,
      "step": 2500
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.71353530883789,
      "learning_rate": 4.86812027430983e-05,
      "loss": 0.3463,
      "step": 3000
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.535433769226074,
      "learning_rate": 4.8461403200281345e-05,
      "loss": 0.3457,
      "step": 3500
    },
    {
      "epoch": 0.7,
      "grad_norm": 12.396780014038086,
      "learning_rate": 4.8241603657464396e-05,
      "loss": 0.3392,
      "step": 4000
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.127644538879395,
      "learning_rate": 4.802180411464745e-05,
      "loss": 0.3473,
      "step": 4500
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.218430519104004,
      "learning_rate": 4.78020045718305e-05,
      "loss": 0.323,
      "step": 5000
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.705328941345215,
      "learning_rate": 4.758220502901354e-05,
      "loss": 0.3242,
      "step": 5500
    },
    {
      "epoch": 1.0,
      "eval_loss": 3.539703845977783,
      "eval_runtime": 41.371,
      "eval_samples_per_second": 1648.256,
      "eval_steps_per_second": 54.942,
      "step": 5687
    }
  ],
  "logging_steps": 500,
  "max_steps": 113740,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 2750382937430730.0,
  "train_batch_size": 30,
  "trial_name": null,
  "trial_params": null
}