ShiancoChat

ShiancoChat is an interactive chat application powered by a local Large Language Model (LLM). It features real-time chat, conversation history management, dark/light mode theming, and multi-language support (English and Chinese).

Features

• Interactive Chat: Engage in real-time conversations with a local LLM.
• Conversation Management: Create new chats, view recent conversations, rename, and delete existing ones.
• Theming: Switch between dark and light modes for a personalized experience.
• Multi-language Support: UI available in English and Chinese.
• LLM Integration: Connects with a local DeepSeek-R1 distilled model served via LM Studio.

Setup and Installation

To get ShiancoChat up and running on your local machine, follow these steps:

Prerequisites

• Node.js and Yarn: For the frontend.
• Python 3.8+ and pip: For the backend.
• MongoDB: A running MongoDB instance (local or cloud-hosted).
• LM Studio: To serve the local DeepSeek-R1 distilled model.

1. Backend Setup

1. Navigate to the backend directory:

   cd backend

2. Create a virtual environment (recommended):

   python -m venv venv
   source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

3. Install dependencies:

   pip install -r requirements.txt

4. Configure environment variables: Create a .env file in the backend/ directory with the following content:

   MONGO_URL=your_mongodb_connection_string
   DB_NAME=shiancochat_db
   LLM_BASE_URL=http://localhost:1234 # Or your LM Studio server address
   PORT=4100 # Or your desired backend port
   

   Replace your_mongodb_connection_string with your actual MongoDB connection string.
5. Run the backend server:

   uvicorn server:app --host 0.0.0.0 --port 4100 --reload

   The backend server should now be running on http://localhost:4100.

2. Frontend Setup

1. Navigate to the frontend directory:

   cd frontend

2. Install dependencies:

   yarn install

3. Configure environment variables: Create a .env file in the frontend/ directory with the following content:

   REACT_APP_BACKEND_URL=http://localhost:4100 # Or your backend server address
   

4. Run the frontend development server:

   yarn start

   The frontend application should now be accessible at http://localhost:4141 (or another port if 4141 is in use).

3. LM Studio Setup

1. Download and Install LM Studio: Get it from LM Studio website.
2. Download the DeepSeek-R1 distilled model:
   • Open LM Studio.
   • Go to the "Search" tab.
   • Search for deepseek/deepseek-r1-0528-qwen3-8b and download it.
3. Load and Serve the Model:
   • Go to the "My Models" tab.
   • Select the downloaded deepseek/deepseek-r1-0528-qwen3-8b model.
   • Go to the "Local Inference Server" tab (usually the > icon on the left sidebar).
   • Click "Start Server". Ensure the server is running on http://localhost:1234 (or the URL configured in your backend's .env file).

Usage

Once both the backend and frontend servers are running, open your web browser and navigate to http://localhost:4141. You can start a new chat, send messages, and interact with the local LLM.

Project Analysis

For a detailed analysis of the project's architecture, tech stack, completed features, and remaining tasks, please refer to the PROJECT_ANALYSIS.md file in the root directory of this repository.