https://github.com/yeukfei02/sanguozhi-llm

sanguozhi-llm
https://github.com/yeukfei02/sanguozhi-llm

chromadb embedding generative-ai llama3 llm ollama python sanguoshi sanguoyanyi streamlit vector-database

Last synced: about 1 month ago
JSON representation

sanguozhi-llm

• Host: GitHub
• URL: https://github.com/yeukfei02/sanguozhi-llm
• Owner: yeukfei02
• License: mit
• Created: 2024-08-23T14:52:32.000Z (5 months ago)
• Default Branch: main
• Last Pushed: 2024-08-28T09:54:16.000Z (5 months ago)
• Last Synced: 2024-11-03T08:42:06.210Z (3 months ago)
• Topics: chromadb, embedding, generative-ai, llama3, llm, ollama, python, sanguoshi, sanguoyanyi, streamlit, vector-database
• Language: Python
• Homepage:
• Size: 104 MB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# sanguozhi-llm

sanguozhi-llm

Data source: `三國志`, `三國演義`, `三國志十遊戲`

This project aims to generate text using the `Ollama Llama 3` model with integrated embeddings and a vector database.

Models and Tools:

LLM Model: `digimonster/llama3-chinese-response`

Purpose: Generates responses in Chinese characters.

Embedding Model: `shaw/dmeta-embedding-zh`

Purpose: Provides Chinese character embeddings.

Vector Database: `chromadb`

Description: An AI-native, open-source vector database for efficient data retrieval and storage.

documentation: 

api url: 

web url: 

## Requirement

- install python (v3.12)

- install ollama llama (v3)

## Testing and run

```zsh

// install dependencies

$ cd packages/api

$ pip install -r requirements.txt

// pull llm model

$ ollama pull digimonster/llama3-chinese-response

// pull embedding model

$ ollama pull shaw/dmeta-embedding-zh

// seed chromadb

$ cd packages/api

$ python seed_chromadb.py

// run ollama

$ cd packages/api

$ sh run-ollama.sh

// run in dev

$ cd packages/api

$ fastapi dev main.py

// run in prod

$ cd packages/api

$ fastapi run main.py

or

$ uvicorn main:app --host 0.0.0.0 --port 80

```

```zsh

// install dependencies

$ cd packages/web

$ pip install -r requirements.txt

// open sanguozhi_llm web

$ streamlit run sanguozhi_llm.py

```