ibl-data-manager (3.59.0-ai-plus)

Train a document through a worker process.

This endpoint queues larger documents for training through a worker process, which is more suitable for handling documents that would take too long to process directly.

Args: request: The HTTP request containing the document information. org: Organization key identifier.

Returns: Response: A confirmation that the document was queued for training, including a task ID for tracking the progress.

Raises: ValidationError: If the request data is invalid. BadRequest: If there was an error processing the document.

Train a document directly without using a worker.

This endpoint is designed for training smaller documents directly without queuing them through a worker process. For larger documents, use the TrainDocumentView endpoint instead.

Args: request: The HTTP request containing the document information. org: Organization key identifier.

Returns: Response: A confirmation that the document was trained successfully.

Raises: ValidationError: If the request data is invalid. BadRequest: If the document training failed.

Description: Submit a document file in a chat session. Checks file type, size, and ensures the session exists, then processes the file. Process document for use in a specific chat session. which will be processed and made available for the AI to reference when responding to queries in that session.

Methods:

• POST: Submits a document or media file for training within a specified chat session.

Parameters:

• org (str):
• session_id (str): UUID of the chat session for which the document is being submitted.
• *args, **kwargs: Additional arguments.
• file (file): Multipart file input.

Returns:

• POST: Returns a JSON message with the result of the upload process. Response code 200 on success. { "message": "File processed." }

Error Responses:

• 400 Bad Request:
• "File field is required" if the 'file' parameter is missing.
• "invalid file object" if the file object is not valid.
• Specific error message if an exception occurs while saving to Redis.
• 404 Not Found:
• "Session not found" if the provided session_id does not match any existing session.
• 413 Request Entity Too Large:
• "file is too large" including specific limits for document and media files if the file exceeds size limits.

Access Control: Requires authentication with a Bearer token in the Authorization header. The user associated with the session must exist and be authorized for the request.