Datasets:

tingtang2
/

the_stack_v2_python_repos_pretraining_dataset_repo_reconstruction-dataset

dataset_info:
  features:
    - name: repo_name
      dtype: string
    - name: combined_content
      dtype: string
    - name: file_paths
      list: string
  splits:
    - name: train
      num_bytes: 17855275374
      num_examples: 400000
  download_size: 5696432796
  dataset_size: 17855275374
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*