[Large model knowledge base] (5): Running the BGE model of dity+fastchat in the local environment, you can use the embedding interface to vectorize the knowledge base, and the continuous adjustment is successful.

1. Video demonstration address

https://www.bilibili.com/video/BV1ei4y1e763/?vd_source=4b290247452adda4e56d84b659b0c8a2

[Large Model Knowledge Base] (5): Open source large model + knowledge base solution, the BGE model of dity + fastchat, you can use the embedding interface to vectorize the knowledge base and bind the chat application

2. About the dify project

https://github.com/langgenius/dify/blob/main/README_CN.md

Dify is an LLM application development platform, and more than 100,000 applications have been built based on Dify.AI. It combines the concepts of Backend as Service and LLMOps, covering the core technology stack required to build generative AI native applications, including a built-in RAG engine. With Dify, you can self-deploy capabilities like Assistants API and GPTs based on any model.

Project deployment script address:
https://gitee.com/fly-llm/dify-docker-compose

3. Configure the knowledge base bge-large-zh for embedding

Use the embedding interface.

Pay special attention to the fact that you need to add the /v1 path this time:

http://192.168.1.116:8000/v1

The local model has two more interfaces that can be used.

4. Then configure the knowledge base import

Then you can use high-quality indexing:

You can see the document paragraphs divided into:

5. Conduct recall testing

Then it can be used with the knowledge base.

6. Summary

The knowledge base is also very convenient to use. You can configure the embedding interface of bge.