TutorialsOfficial documentation5 min readTutorials

Clarify FastGPT Dataset Processing Model Roles

Core Dataset Processing Models Overview FastGPT utilizes two specialized, isolated models to prepare and organize dataset content for retrieval-augmented…

Core Dataset Processing Models Overview

FastGPT utilizes two specialized, isolated models to prepare and organize dataset content for retrieval-augmented generation workflows. Each model serves a distinct, non-interchangeable purpose during dataset management, and understanding their dedicated functions is critical for proper configuration of dataset ingestion and retrieval pipelines.

Model Comparison Reference Table

The following table outlines the core official specifications of each model, with no inferred or external details included:

Model TypePrimary Use CaseKey Functions
File Processing ModelEnhanced Processing and Q&A Splitting during data ingestionGenerate related questions and summaries; generate question-answer pairs
Embedding ModelVectorizationProcess and organize text data into a structure optimized for fast retrieval

Operational Workflow Alignment

The File Processing Model executes exclusively during the data ingestion phase, acting on raw uploaded dataset files to restructure content into more usable formats. The Embedding Model operates solely on the output generated by the File Processing Model, converting that processed text into a vectorized format designed to enable fast, accurate semantic retrieval. No overlapping configuration steps are required between the two models, as their operational scopes are strictly separated per official documentation.

Source: FastGPT official source

Applicability and version scope

Use this page for the documented Tutorials scenario. Confirm the FastGPT, dependency, API, and deployment versions in the official source before applying a change.

Safety guardrails

Use [REDACTED_CREDENTIAL] for credentials and private data. Confirm the documented environment and version before review.

Rollback guidance

Restore the prior technical-content authority snapshot. Restore saved configuration and data snapshots, then repeat the smallest verification scenario.