nat.plugins.langchain.dataset_loader.langsmith#

Functions#

load_langsmith_dataset(→ pandas.DataFrame)

Fetch a dataset from LangSmith and return as a pandas DataFrame.

Module Contents#

load_langsmith_dataset(
file_path,
*,
dataset_id: str | None = None,
dataset_name: str | None = None,
input_key: str = 'input',
output_key: str = 'output',
question_col: str = 'question',
answer_col: str = 'answer',
id_col: str = 'id',
split: str | None = None,
as_of: str | None = None,
limit: int | None = None,
) pandas.DataFrame#

Fetch a dataset from LangSmith and return as a pandas DataFrame.

Prefers dataset_id over dataset_name when both are provided. The file_path argument is ignored — data comes from the LangSmith API, not the filesystem.

Loads dataset of format https://docs.langchain.com/langsmith/example-data-format