trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9901639344262296,
  "eval_steps": 500,
  "global_step": 114,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.26229508196721313,
      "grad_norm": 32.00107145638801,
      "learning_rate": 5e-06,
      "loss": 1.0567,
      "step": 10
    },
    {
      "epoch": 0.5245901639344263,
      "grad_norm": 1.971184652401555,
      "learning_rate": 5e-06,
      "loss": 0.9751,
      "step": 20
    },
    {
      "epoch": 0.7868852459016393,
      "grad_norm": 1.0825871925332298,
      "learning_rate": 5e-06,
      "loss": 0.9235,
      "step": 30
    },
    {
      "epoch": 0.9967213114754099,
      "eval_loss": 0.8894760012626648,
      "eval_runtime": 27.4292,
      "eval_samples_per_second": 37.478,
      "eval_steps_per_second": 0.62,
      "step": 38
    },
    {
      "epoch": 1.0491803278688525,
      "grad_norm": 1.115550400640531,
      "learning_rate": 5e-06,
      "loss": 0.9643,
      "step": 40
    },
    {
      "epoch": 1.3114754098360657,
      "grad_norm": 0.7517137885735146,
      "learning_rate": 5e-06,
      "loss": 0.8519,
      "step": 50
    },
    {
      "epoch": 1.5737704918032787,
      "grad_norm": 0.6663946347546653,
      "learning_rate": 5e-06,
      "loss": 0.8345,
      "step": 60
    },
    {
      "epoch": 1.8360655737704918,
      "grad_norm": 0.8125490139029613,
      "learning_rate": 5e-06,
      "loss": 0.8286,
      "step": 70
    },
    {
      "epoch": 1.9934426229508198,
      "eval_loss": 0.8515063524246216,
      "eval_runtime": 28.0142,
      "eval_samples_per_second": 36.696,
      "eval_steps_per_second": 0.607,
      "step": 76
    },
    {
      "epoch": 2.098360655737705,
      "grad_norm": 0.6569135220000142,
      "learning_rate": 5e-06,
      "loss": 0.8792,
      "step": 80
    },
    {
      "epoch": 2.360655737704918,
      "grad_norm": 0.7858189986120058,
      "learning_rate": 5e-06,
      "loss": 0.7768,
      "step": 90
    },
    {
      "epoch": 2.6229508196721314,
      "grad_norm": 0.6353257417534841,
      "learning_rate": 5e-06,
      "loss": 0.7699,
      "step": 100
    },
    {
      "epoch": 2.8852459016393444,
      "grad_norm": 0.748479252875262,
      "learning_rate": 5e-06,
      "loss": 0.771,
      "step": 110
    },
    {
      "epoch": 2.9901639344262296,
      "eval_loss": 0.8435602784156799,
      "eval_runtime": 26.4349,
      "eval_samples_per_second": 38.888,
      "eval_steps_per_second": 0.643,
      "step": 114
    },
    {
      "epoch": 2.9901639344262296,
      "step": 114,
      "total_flos": 190744866324480.0,
      "train_loss": 0.8725291072276601,
      "train_runtime": 4078.6063,
      "train_samples_per_second": 14.358,
      "train_steps_per_second": 0.028
    }
  ],
  "logging_steps": 10,
  "max_steps": 114,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 190744866324480.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}