metadata
			dataset_info:
  features:
    - name: markdown
      dtype: string
    - name: code
      dtype: string
    - name: output
      dtype: string
    - name: license
      dtype: string
    - name: path
      dtype: string
    - name: repo_name
      dtype: string
  splits:
    - name: train
      num_bytes: 14379582821
      num_examples: 9305991
  download_size: 6305278373
  dataset_size: 14379582821
