trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.931297709923664,
  "eval_steps": 500,
  "global_step": 96,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3053435114503817,
      "grad_norm": 6.722743498305272,
      "learning_rate": 5e-06,
      "loss": 0.733,
      "step": 10
    },
    {
      "epoch": 0.6106870229007634,
      "grad_norm": 1.6826556916714077,
      "learning_rate": 5e-06,
      "loss": 0.6358,
      "step": 20
    },
    {
      "epoch": 0.916030534351145,
      "grad_norm": 1.516348699189821,
      "learning_rate": 5e-06,
      "loss": 0.5846,
      "step": 30
    },
    {
      "epoch": 0.9770992366412213,
      "eval_loss": 0.5693601965904236,
      "eval_runtime": 23.4206,
      "eval_samples_per_second": 37.702,
      "eval_steps_per_second": 0.598,
      "step": 32
    },
    {
      "epoch": 1.2213740458015268,
      "grad_norm": 1.3385393305433098,
      "learning_rate": 5e-06,
      "loss": 0.5516,
      "step": 40
    },
    {
      "epoch": 1.5267175572519083,
      "grad_norm": 1.289866024739384,
      "learning_rate": 5e-06,
      "loss": 0.5098,
      "step": 50
    },
    {
      "epoch": 1.83206106870229,
      "grad_norm": 1.5587580916050303,
      "learning_rate": 5e-06,
      "loss": 0.4986,
      "step": 60
    },
    {
      "epoch": 1.984732824427481,
      "eval_loss": 0.5270054340362549,
      "eval_runtime": 22.667,
      "eval_samples_per_second": 38.955,
      "eval_steps_per_second": 0.618,
      "step": 65
    },
    {
      "epoch": 2.1374045801526718,
      "grad_norm": 1.5316434262726102,
      "learning_rate": 5e-06,
      "loss": 0.4818,
      "step": 70
    },
    {
      "epoch": 2.4427480916030535,
      "grad_norm": 2.2102135841908352,
      "learning_rate": 5e-06,
      "loss": 0.4381,
      "step": 80
    },
    {
      "epoch": 2.7480916030534353,
      "grad_norm": 2.3304153563587153,
      "learning_rate": 5e-06,
      "loss": 0.4286,
      "step": 90
    },
    {
      "epoch": 2.931297709923664,
      "eval_loss": 0.5195185542106628,
      "eval_runtime": 21.7824,
      "eval_samples_per_second": 40.537,
      "eval_steps_per_second": 0.643,
      "step": 96
    },
    {
      "epoch": 2.931297709923664,
      "step": 96,
      "total_flos": 160594195906560.0,
      "train_loss": 0.533353087802728,
      "train_runtime": 3467.0939,
      "train_samples_per_second": 14.508,
      "train_steps_per_second": 0.028
    }
  ],
  "logging_steps": 10,
  "max_steps": 96,
  "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": 160594195906560.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}