last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.039525691699604744,
  "eval_steps": 5,
  "global_step": 30,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0013175230566534915,
      "eval_loss": NaN,
      "eval_runtime": 44.835,
      "eval_samples_per_second": 7.137,
      "eval_steps_per_second": 3.569,
      "step": 1
    },
    {
      "epoch": 0.003952569169960474,
      "grad_norm": NaN,
      "learning_rate": 4e-05,
      "loss": 0.0,
      "step": 3
    },
    {
      "epoch": 0.006587615283267457,
      "eval_loss": NaN,
      "eval_runtime": 45.1723,
      "eval_samples_per_second": 7.084,
      "eval_steps_per_second": 3.542,
      "step": 5
    },
    {
      "epoch": 0.007905138339920948,
      "grad_norm": NaN,
      "learning_rate": 8e-05,
      "loss": 0.0,
      "step": 6
    },
    {
      "epoch": 0.011857707509881422,
      "grad_norm": NaN,
      "learning_rate": 0.00012,
      "loss": 0.0,
      "step": 9
    },
    {
      "epoch": 0.013175230566534914,
      "eval_loss": NaN,
      "eval_runtime": 45.3945,
      "eval_samples_per_second": 7.049,
      "eval_steps_per_second": 3.525,
      "step": 10
    },
    {
      "epoch": 0.015810276679841896,
      "grad_norm": NaN,
      "learning_rate": 0.00016,
      "loss": 0.0,
      "step": 12
    },
    {
      "epoch": 0.019762845849802372,
      "grad_norm": NaN,
      "learning_rate": 0.0002,
      "loss": 0.0,
      "step": 15
    },
    {
      "epoch": 0.019762845849802372,
      "eval_loss": NaN,
      "eval_runtime": 45.4855,
      "eval_samples_per_second": 7.035,
      "eval_steps_per_second": 3.518,
      "step": 15
    },
    {
      "epoch": 0.023715415019762844,
      "grad_norm": NaN,
      "learning_rate": 0.00018090169943749476,
      "loss": 0.0,
      "step": 18
    },
    {
      "epoch": 0.026350461133069828,
      "eval_loss": NaN,
      "eval_runtime": 45.5017,
      "eval_samples_per_second": 7.033,
      "eval_steps_per_second": 3.516,
      "step": 20
    },
    {
      "epoch": 0.02766798418972332,
      "grad_norm": NaN,
      "learning_rate": 0.00013090169943749476,
      "loss": 0.0,
      "step": 21
    },
    {
      "epoch": 0.03162055335968379,
      "grad_norm": NaN,
      "learning_rate": 6.909830056250527e-05,
      "loss": 0.0,
      "step": 24
    },
    {
      "epoch": 0.03293807641633729,
      "eval_loss": NaN,
      "eval_runtime": 45.5289,
      "eval_samples_per_second": 7.028,
      "eval_steps_per_second": 3.514,
      "step": 25
    },
    {
      "epoch": 0.03557312252964427,
      "grad_norm": NaN,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 0.0,
      "step": 27
    },
    {
      "epoch": 0.039525691699604744,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 0.0,
      "step": 30
    },
    {
      "epoch": 0.039525691699604744,
      "eval_loss": NaN,
      "eval_runtime": 45.5338,
      "eval_samples_per_second": 7.028,
      "eval_steps_per_second": 3.514,
      "step": 30
    }
  ],
  "logging_steps": 3,
  "max_steps": 30,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 15,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.971797071233024e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}