Building from Source¶

Source Code Structure¶

.          
├── LICENSE                #  declaration        
├── README.md           
├── apps                   # Backend project root directory        
│ ├── application          # Applications
│ ├── common               # Project common resources directory
│ ├── dataset              # Knowledge base
│ ├── embedding            # Vectorization
│ ├── function_lib         # Function library
│ ├── ops                  # 
│ ├── setting             # Settings
│ ├── smartdoc             # Project main directory          
│ ├── users                # User related          
│ └── manage.py            # Django project entry point
├── installer              # Installation related  
├── pyproject.toml         # Backend dependency configuration file         
└── ui                     # Frontend project root directory          
├── config_example.yml     # Project configuration example             
├── main.py                # Project entry file

Setting Up the Development Environment¶

Environment Preparation¶

Frontend
Install node, v16 or above
Backend
Install python, version v3.11.x
Install pycharm
Database
Install PostgreSql, version v15.x

Initial Configuration¶

Create Database¶

Manually create a database named maxkb in PostgreSql for the MaxKB project.

# Create database
CREATE DATABASE "maxkb"; 
# Switch to database
\c "maxkb"; 
# Create vector extension
CREATE EXTENSION "vector";

Configuration File¶

Before starting the project, configuration files and directories need to be prepared.

Prepare configuration file

# Copy config_example.yml configuration file to /opt/maxkb/conf directory
cp config_example.yml /opt/maxkb/conf

Configure /opt/maxkb/conf/config_example.yml.

# Database configuration 
DB_NAME: maxkb
DB_HOST: localhost
DB_PORT: 5432
DB_USER: root
DB_PASSWORD: xxx
DB_ENGINE: django.db.backends.postgresql_psycopg2

# Model related configuration
EMBEDDING_MODEL_PATH: /opt/maxkb/model/
EMBEDDING_MODEL_NAME: /opt/maxkb/model/shibing624_text2vec-base-chinese

Development and Debugging¶

Running the Frontend¶

Note

All commands below should be executed in the ui directory of the MaxKB project.

# Install dependencies
npm install

# Start the project
npm run dev

Running the Backend¶

For the backend system, you need to start the web service, local model service, and asynchronous task queue service.

Note

All commands below should be executed in the root directory of the MaxKB project.

Web Service¶

# Install poetry package manager
pip install poetry

# Install backend dependencies
poetry install

# Start project
python main.py start

Local Model Service¶

# Start project
python main.py dev local_model

Asynchronous Task Service¶

# Start project
python main.py dev celery

Accessing MaxKB via Browser¶

Once both the frontend and backend are successfully started, you can access MaxKB through the frontend route (the access address is displayed when the frontend starts) in your browser.