简体中文|English

This project provides an intelligent agent for legal applications, offering knowledge-based Q&A functionality utilizing Faiss vector retrieval and LangChain.

1. Install Ollama locally as per the instructions on the official website (https://ollama.com/).
2. Pull the desired LLM model and Embedding model using the following commands in your terminal:
   ollama pull <llm_model_name> # e.g., ollama pull llama3
   ollama pull <embedding_model_name> # e.g., ollama pull nomic-embed-text
3. Navigate to the config directory of your project and edit the configuration file (e.g., config.yaml or config.json). Update the following settings:
   • Set the API base URL to the local Ollama server (default: http://localhost:11434)
   • Set the LLM model name to the one you pulled (e.g., "llama3")
   • Set the Embedding model name to the one you pulled (e.g., "nomic-embed-text")

1. Install uv

curl -LsSf https://astral.sh/uv/install.sh | sh

2. Install git lfs and pull large files

git lfs install
git lfs pull

3. Run the program

uv run run.py
# install nvm
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.3/install.sh | bash
# restart bash
nvm install
cd web/frontend
npm install
npm run dev

1. Install Python 3.12
2. Install dependencies

pip install -r requirements.txt

3. Run the program

python run.py
# install nvm
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.3/install.sh | bash
# restart bash
nvm install
cd web/frontend
npm install
npm run dev