checkpoint-18/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.285714285714286,
  "eval_steps": 500,
  "global_step": 18,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.57,
      "grad_norm": 0.2423473596572876,
      "learning_rate": 0.0001,
      "loss": 4.0857,
      "step": 1
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.2876349985599518,
      "learning_rate": 0.0002,
      "loss": 4.396,
      "step": 2
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.25892937183380127,
      "learning_rate": 0.0001875,
      "loss": 4.257,
      "step": 3
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.24695058166980743,
      "learning_rate": 0.000175,
      "loss": 4.2171,
      "step": 4
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.439170777797699,
      "learning_rate": 0.00016250000000000002,
      "loss": 4.0037,
      "step": 5
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.3510642647743225,
      "learning_rate": 0.00015000000000000001,
      "loss": 4.1377,
      "step": 6
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3439651131629944,
      "learning_rate": 0.0001375,
      "loss": 4.0279,
      "step": 7
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.3093964159488678,
      "learning_rate": 0.000125,
      "loss": 3.8389,
      "step": 8
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.3536764979362488,
      "learning_rate": 0.00011250000000000001,
      "loss": 4.1566,
      "step": 9
    },
    {
      "epoch": 5.71,
      "grad_norm": 0.3965250849723816,
      "learning_rate": 0.0001,
      "loss": 3.969,
      "step": 10
    },
    {
      "epoch": 6.29,
      "grad_norm": 0.4037889838218689,
      "learning_rate": 8.75e-05,
      "loss": 3.9695,
      "step": 11
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.3280973434448242,
      "learning_rate": 7.500000000000001e-05,
      "loss": 3.7841,
      "step": 12
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.3607095777988434,
      "learning_rate": 6.25e-05,
      "loss": 3.8726,
      "step": 13
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.3536747992038727,
      "learning_rate": 5e-05,
      "loss": 3.8848,
      "step": 14
    },
    {
      "epoch": 8.57,
      "grad_norm": 0.42837923765182495,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 3.8312,
      "step": 15
    },
    {
      "epoch": 9.14,
      "grad_norm": 0.3861906826496124,
      "learning_rate": 2.5e-05,
      "loss": 3.9284,
      "step": 16
    },
    {
      "epoch": 9.71,
      "grad_norm": 0.3450121581554413,
      "learning_rate": 1.25e-05,
      "loss": 3.8218,
      "step": 17
    },
    {
      "epoch": 10.29,
      "grad_norm": 0.3705593943595886,
      "learning_rate": 0.0,
      "loss": 3.726,
      "step": 18
    }
  ],
  "logging_steps": 1,
  "max_steps": 18,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 18,
  "save_steps": 500,
  "total_flos": 5730050919936000.0,
  "train_batch_size": 5,
  "trial_name": null,
  "trial_params": null
}