last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.001315068493150685,
  "eval_steps": 5,
  "global_step": 15,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 8.767123287671233e-05,
      "grad_norm": 17.826961517333984,
      "learning_rate": 1e-05,
      "loss": 4.9034,
      "step": 1
    },
    {
      "epoch": 8.767123287671233e-05,
      "eval_loss": 5.864974498748779,
      "eval_runtime": 4180.0734,
      "eval_samples_per_second": 9.192,
      "eval_steps_per_second": 4.596,
      "step": 1
    },
    {
      "epoch": 0.00017534246575342467,
      "grad_norm": 15.42636489868164,
      "learning_rate": 2e-05,
      "loss": 4.5406,
      "step": 2
    },
    {
      "epoch": 0.00026301369863013696,
      "grad_norm": 15.912243843078613,
      "learning_rate": 3e-05,
      "loss": 4.8244,
      "step": 3
    },
    {
      "epoch": 0.00035068493150684934,
      "grad_norm": 15.158151626586914,
      "learning_rate": 4e-05,
      "loss": 4.8005,
      "step": 4
    },
    {
      "epoch": 0.00043835616438356166,
      "grad_norm": 14.64368724822998,
      "learning_rate": 5e-05,
      "loss": 4.4149,
      "step": 5
    },
    {
      "epoch": 0.00043835616438356166,
      "eval_loss": 4.748353958129883,
      "eval_runtime": 4179.9748,
      "eval_samples_per_second": 9.192,
      "eval_steps_per_second": 4.596,
      "step": 5
    },
    {
      "epoch": 0.0005260273972602739,
      "grad_norm": 12.26352310180664,
      "learning_rate": 6e-05,
      "loss": 4.1608,
      "step": 6
    },
    {
      "epoch": 0.0006136986301369863,
      "grad_norm": 8.713861465454102,
      "learning_rate": 7e-05,
      "loss": 3.7209,
      "step": 7
    },
    {
      "epoch": 0.0007013698630136987,
      "grad_norm": 8.520092964172363,
      "learning_rate": 8e-05,
      "loss": 3.3179,
      "step": 8
    },
    {
      "epoch": 0.0007890410958904109,
      "grad_norm": 8.435450553894043,
      "learning_rate": 9e-05,
      "loss": 2.7645,
      "step": 9
    },
    {
      "epoch": 0.0008767123287671233,
      "grad_norm": 7.701727390289307,
      "learning_rate": 0.0001,
      "loss": 2.4773,
      "step": 10
    },
    {
      "epoch": 0.0008767123287671233,
      "eval_loss": 2.0398104190826416,
      "eval_runtime": 4178.7123,
      "eval_samples_per_second": 9.195,
      "eval_steps_per_second": 4.597,
      "step": 10
    },
    {
      "epoch": 0.0009643835616438356,
      "grad_norm": 9.178277969360352,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.2672,
      "step": 11
    },
    {
      "epoch": 0.0010520547945205478,
      "grad_norm": 6.568050861358643,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.0439,
      "step": 12
    },
    {
      "epoch": 0.0011397260273972602,
      "grad_norm": 4.214171409606934,
      "learning_rate": 7.938926261462366e-05,
      "loss": 1.3993,
      "step": 13
    },
    {
      "epoch": 0.0012273972602739726,
      "grad_norm": 3.7734668254852295,
      "learning_rate": 6.545084971874738e-05,
      "loss": 1.5691,
      "step": 14
    },
    {
      "epoch": 0.001315068493150685,
      "grad_norm": 3.103121280670166,
      "learning_rate": 5e-05,
      "loss": 1.4274,
      "step": 15
    },
    {
      "epoch": 0.001315068493150685,
      "eval_loss": 1.1212583780288696,
      "eval_runtime": 4181.543,
      "eval_samples_per_second": 9.188,
      "eval_steps_per_second": 4.594,
      "step": 15
    }
  ],
  "logging_steps": 1,
  "max_steps": 20,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8651451609907200.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}