trainer_state.json

{
  "best_metric": 0.5337400436401367,
  "best_model_checkpoint": "saves/Mistral-7B/lora/train_1/checkpoint-10",
  "epoch": 0.06562756357670221,
  "eval_steps": 10,
  "global_step": 60,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010937927262783703,
      "grad_norm": 0.9827508330345154,
      "learning_rate": 0.00029999015487222375,
      "loss": 0.8548,
      "step": 10
    },
    {
      "epoch": 0.010937927262783703,
      "eval_loss": 0.5337400436401367,
      "eval_runtime": 231.0066,
      "eval_samples_per_second": 20.398,
      "eval_steps_per_second": 0.641,
      "step": 10
    },
    {
      "epoch": 0.021875854525567406,
      "grad_norm": 0.6526839733123779,
      "learning_rate": 0.00029996062078124905,
      "loss": 0.4005,
      "step": 20
    },
    {
      "epoch": 0.021875854525567406,
      "eval_loss": 0.5784199833869934,
      "eval_runtime": 231.0005,
      "eval_samples_per_second": 20.398,
      "eval_steps_per_second": 0.641,
      "step": 20
    },
    {
      "epoch": 0.03281378178835111,
      "grad_norm": 0.5257601141929626,
      "learning_rate": 0.0002999114016039678,
      "loss": 0.3521,
      "step": 30
    },
    {
      "epoch": 0.03281378178835111,
      "eval_loss": 0.5616388320922852,
      "eval_runtime": 230.9639,
      "eval_samples_per_second": 20.401,
      "eval_steps_per_second": 0.641,
      "step": 30
    },
    {
      "epoch": 0.04375170905113481,
      "grad_norm": 0.553675651550293,
      "learning_rate": 0.00029984250380130117,
      "loss": 0.3565,
      "step": 40
    },
    {
      "epoch": 0.04375170905113481,
      "eval_loss": 0.5677462220191956,
      "eval_runtime": 231.0825,
      "eval_samples_per_second": 20.391,
      "eval_steps_per_second": 0.64,
      "step": 40
    },
    {
      "epoch": 0.05468963631391851,
      "grad_norm": 0.3802327811717987,
      "learning_rate": 0.0002997539364173515,
      "loss": 0.3418,
      "step": 50
    },
    {
      "epoch": 0.05468963631391851,
      "eval_loss": 0.5386738181114197,
      "eval_runtime": 231.0958,
      "eval_samples_per_second": 20.39,
      "eval_steps_per_second": 0.64,
      "step": 50
    },
    {
      "epoch": 0.06562756357670221,
      "grad_norm": 0.5235214829444885,
      "learning_rate": 0.00029964571107821494,
      "loss": 0.3298,
      "step": 60
    },
    {
      "epoch": 0.06562756357670221,
      "eval_loss": 0.5525479912757874,
      "eval_runtime": 230.9516,
      "eval_samples_per_second": 20.403,
      "eval_steps_per_second": 0.641,
      "step": 60
    },
    {
      "epoch": 0.06562756357670221,
      "step": 60,
      "total_flos": 1.143902298880082e+17,
      "train_loss": 0.43926427761713666,
      "train_runtime": 4213.5123,
      "train_samples_per_second": 166.635,
      "train_steps_per_second": 0.651
    }
  ],
  "logging_steps": 10,
  "max_steps": 2742,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 10,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.143902298880082e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}