Method afrom_files

core/quivr_core/brain/brain.py:300–371

Create a brain from a list of file paths. Args: name (str): The name of the brain. file_paths (list[str | Path]): The list of file paths to add to the brain. vector_db (VectorStore | None): The vector store used to store the processed files.

(
        cls,
        *,
        name: str,
        file_paths: list[str | Path],
        vector_db: VectorStore | None = None,
        storage: StorageBase = TransparentStorage(),
        llm: LLMEndpoint | None = None,
        embedder: Embeddings | None = None,
        skip_file_error: bool = False,
        processor_kwargs: dict[str, Any] | None = None,
    )

    @classmethod
    async def afrom_files(
        cls,
        *,
        name: str,
        file_paths: list[str | Path],
        vector_db: VectorStore | None = None,
        storage: StorageBase = TransparentStorage(),
        llm: LLMEndpoint | None = None,
        embedder: Embeddings | None = None,
        skip_file_error: bool = False,
        processor_kwargs: dict[str, Any] | None = None,
    ):
        """
        Create a brain from a list of file paths.
        Args:
            name (str): The name of the brain.
            file_paths (list[str | Path]): The list of file paths to add to the brain.
            vector_db (VectorStore | None): The vector store used to store the processed files.
            storage (StorageBase): The storage used to store the files.
            llm (LLMEndpoint | None): The language model used to generate the answer.
            embedder (Embeddings | None): The embeddings used to create the index of the processed files.
            skip_file_error (bool): Whether to skip files that cannot be processed.
            processor_kwargs (dict[str, Any] | None): Additional arguments for the processor.
        Returns:
            Brain: The brain created from the file paths.
        Example:
        ```python
        brain = await Brain.afrom_files(name="My Brain", file_paths=["file1.pdf", "file2.pdf"])
        brain.print_info()
        ```
        """
        if llm is None:
            llm = default_llm()

        if embedder is None:
            embedder = default_embedder()

        processor_kwargs = processor_kwargs or {}

        brain_id = uuid4()

        # TODO: run in parallel using tasks

        for path in file_paths:
            file = await load_qfile(brain_id, path)
            await storage.upload_file(file)

        logger.debug(f"uploaded all files to {storage}")

        # Parse files
        docs = await process_files(
            storage=storage,
            skip_file_error=skip_file_error,
            **processor_kwargs,
        )

        # Building brain's vectordb
        if vector_db is None: