checkpoint-40/trainer_state.json

{
  "best_global_step": 40,
  "best_metric": 0.5844668745994568,
  "best_model_checkpoint": "//outputs/task7_microsoft/Phi-3.5-mini-instruct/checkpoint-40",
  "epoch": 4.470588235294118,
  "eval_steps": 500,
  "global_step": 40,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 0.3697907328605652,
      "learning_rate": 5e-05,
      "loss": 0.801,
      "step": 5
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.7140461802482605,
      "eval_runtime": 3.3729,
      "eval_samples_per_second": 4.447,
      "eval_steps_per_second": 0.593,
      "step": 9
    },
    {
      "epoch": 1.1176470588235294,
      "grad_norm": 0.23551630973815918,
      "learning_rate": 0.0001,
      "loss": 0.7259,
      "step": 10
    },
    {
      "epoch": 1.7058823529411766,
      "grad_norm": 0.197063609957695,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.6883,
      "step": 15
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.6461538076400757,
      "eval_runtime": 3.3729,
      "eval_samples_per_second": 4.447,
      "eval_steps_per_second": 0.593,
      "step": 18
    },
    {
      "epoch": 2.235294117647059,
      "grad_norm": 0.18328554928302765,
      "learning_rate": 0.0002,
      "loss": 0.6152,
      "step": 20
    },
    {
      "epoch": 2.8235294117647056,
      "grad_norm": 0.17514856159687042,
      "learning_rate": 0.00017071067811865476,
      "loss": 0.527,
      "step": 25
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.6003979444503784,
      "eval_runtime": 3.3785,
      "eval_samples_per_second": 4.44,
      "eval_steps_per_second": 0.592,
      "step": 27
    },
    {
      "epoch": 3.3529411764705883,
      "grad_norm": 0.24001598358154297,
      "learning_rate": 0.0001,
      "loss": 0.5294,
      "step": 30
    },
    {
      "epoch": 3.9411764705882355,
      "grad_norm": 0.2075182944536209,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 0.4269,
      "step": 35
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.5861222147941589,
      "eval_runtime": 3.3731,
      "eval_samples_per_second": 4.447,
      "eval_steps_per_second": 0.593,
      "step": 36
    },
    {
      "epoch": 4.470588235294118,
      "grad_norm": 0.21187624335289001,
      "learning_rate": 0.0,
      "loss": 0.4538,
      "step": 40
    },
    {
      "epoch": 4.470588235294118,
      "eval_loss": 0.5844668745994568,
      "eval_runtime": 3.3754,
      "eval_samples_per_second": 4.444,
      "eval_steps_per_second": 0.593,
      "step": 40
    }
  ],
  "logging_steps": 5,
  "max_steps": 40,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 2.1241123003551744e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}