flow_inference.configure_dataset_card.HuggingFaceReadmeBuilder

class flow_inference.configure_dataset_card.HuggingFaceReadmeBuilder(repo_id, dataset, dataframes, parquet_paths, source_repos=None, description_text=None, evaluation_info=None, duplicate_info=None, tags=None, license_name='mit')

Build a Hugging Face dataset README with metadata and dataset statistics.

Parameters:

• repo_id (str)

• dataset (DatasetDict)

• dataframes (dict[str, DataFrame])

• parquet_paths (dict[str, list[str]])

• source_repos (list[str] | None)

• description_text (str | None)

• evaluation_info (dict[str, Any] | None)

• duplicate_info (dict[str, dict[str, int]] | None)

• tags (list[str] | None)

• license_name (str)

__init__(repo_id, dataset, dataframes, parquet_paths, source_repos=None, description_text=None, evaluation_info=None, duplicate_info=None, tags=None, license_name='mit')

Initialize the README builder.

Parameters:

• repo_id (str)

• dataset (DatasetDict)

• dataframes (dict[str, DataFrame])

• parquet_paths (dict[str, list[str]])

• source_repos (list[str] | None)

• description_text (str | None)

• evaluation_info (dict[str, Any] | None)

• duplicate_info (dict[str, dict[str, int]] | None)

• tags (list[str] | None)

• license_name (str)

Methods

__init__(repo_id, dataset, dataframes, ...)	Initialize the README builder.
from_handler(repo_id, dataset, dataframes, ...)	Create a README builder from dataset handler output.
render()	Render the complete README content with frontmatter and body sections.