fabricatio_rag.models.document

Base class for document models.

Classes

StoredDocumentModel	A base class for document models.
SearchedDocumentModel	A base class for document models retrieved from a vector database.

Module Contents

class fabricatio_rag.models.document.StoredDocumentModel[ST](/, **data: Any)

Bases: fabricatio_core.models.generic.Base, fabricatio_core.models.generic.Vectorizable

A base class for document models.

abstractmethod prepare_insertion(vector: Sequence[float]) → ST

Prepares the data for insertion into a vector database.

classmethod from_txt_files[S: StoredDocumentModel[ST]](files: Sequence[pathlib.Path], chunk_size: int = 512, overlap: float = 0.2) → List[S]

Create documents by splitting text files into chunks.

Parameters:

• files – Sequence of text file paths to read.

• chunk_size – Maximum word count per chunk.

• overlap – Overlap ratio between consecutive chunks (0.0-1.0).

Returns:

List of text chunk document model instances, one per chunk.

classmethod with_text_chunk(chunk: str) → Self

Abstractmethod:

Create with a text chunk.

class fabricatio_rag.models.document.SearchedDocumentModel[SD](/, **data: Any)

Bases: fabricatio_capabilities.models.generic.AsPrompt, fabricatio_core.models.generic.Base

A base class for document models retrieved from a vector database.

classmethod from_raw(raw: SD) → Self

Abstractmethod:

Create the searched model from the rawdata searched from the db.