trainer_state.json

{
  "best_metric": 0.8413172364234924,
  "best_model_checkpoint": "saves/BLOOM-7B/lora/train_1/checkpoint-200",
  "epoch": 1.3125512715340442,
  "eval_steps": 200,
  "global_step": 1200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.21875854525567404,
      "grad_norm": 0.41011762619018555,
      "learning_rate": 0.00029856911617379416,
      "loss": 0.817,
      "step": 200
    },
    {
      "epoch": 0.21875854525567404,
      "eval_loss": 0.8413172364234924,
      "eval_runtime": 2018.9188,
      "eval_samples_per_second": 23.339,
      "eval_steps_per_second": 0.73,
      "step": 200
    },
    {
      "epoch": 0.4375170905113481,
      "grad_norm": 0.5003934502601624,
      "learning_rate": 0.0002896017519370078,
      "loss": 0.381,
      "step": 400
    },
    {
      "epoch": 0.4375170905113481,
      "eval_loss": 0.8805686831474304,
      "eval_runtime": 2017.6263,
      "eval_samples_per_second": 23.354,
      "eval_steps_per_second": 0.73,
      "step": 400
    },
    {
      "epoch": 0.6562756357670222,
      "grad_norm": 0.671746551990509,
      "learning_rate": 0.0002728756302319302,
      "loss": 0.3025,
      "step": 600
    },
    {
      "epoch": 0.6562756357670222,
      "eval_loss": 1.017870545387268,
      "eval_runtime": 2018.3325,
      "eval_samples_per_second": 23.346,
      "eval_steps_per_second": 0.73,
      "step": 600
    },
    {
      "epoch": 0.8750341810226961,
      "grad_norm": 0.7479655146598816,
      "learning_rate": 0.00024932035201194605,
      "loss": 0.1879,
      "step": 800
    },
    {
      "epoch": 0.8750341810226961,
      "eval_loss": 1.1555241346359253,
      "eval_runtime": 2018.658,
      "eval_samples_per_second": 23.342,
      "eval_steps_per_second": 0.73,
      "step": 800
    },
    {
      "epoch": 1.0937927262783702,
      "grad_norm": 0.6166426539421082,
      "learning_rate": 0.00022024506768721243,
      "loss": 0.0744,
      "step": 1000
    },
    {
      "epoch": 1.0937927262783702,
      "eval_loss": 1.2785167694091797,
      "eval_runtime": 2018.1701,
      "eval_samples_per_second": 23.348,
      "eval_steps_per_second": 0.73,
      "step": 1000
    },
    {
      "epoch": 1.3125512715340442,
      "grad_norm": 0.3912193179130554,
      "learning_rate": 0.0001872657174323126,
      "loss": 0.0231,
      "step": 1200
    },
    {
      "epoch": 1.3125512715340442,
      "eval_loss": 1.336869478225708,
      "eval_runtime": 2018.0932,
      "eval_samples_per_second": 23.349,
      "eval_steps_per_second": 0.73,
      "step": 1200
    },
    {
      "epoch": 1.3125512715340442,
      "step": 1200,
      "total_flos": 1.7087206139898102e+18,
      "train_loss": 0.2976297422250112,
      "train_runtime": 65664.4367,
      "train_samples_per_second": 10.693,
      "train_steps_per_second": 0.042
    }
  ],
  "logging_steps": 200,
  "max_steps": 2742,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "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": 1.7087206139898102e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}