10. AI Large Model Offline Voice Assistant

1. Offline Voice Configuration

Before setting up auto-start, we must ensure that the program itself can work independently in an offline state. This needs to be done by modifying configuration files.

1. Locate the configuration file: In your project code, find and open the configuration file: config/yahboom.yaml

2. Modify configuration parameters: Please check the following parameters in the file and ensure their values match those shown below. If the parameters do not exist, please add them.

   ​x
   asr:                                    #Voice node parameters
     ros__parameters:
       VAD_MODE: 2                         #VAD sensitivity
       sample_rate: 16000                  #ASR recording audio sample rate
       frame_duration_ms:  30              #VAD frame size in ms
       use_oline_asr: False                #Whether to use online ASR recognition (True for online, False for offline)
       mic_serial_port: "/dev/ttyUSB0"     #Microphone serial port alias
       mic_index: 0                        #Microphone index
       language: 'zh'                      #ASR language
   ​
   model_service:                          #Model server node parameters
     ros__parameters:
       language: 'zh'                      #Large model interface language
       useolinetts: False                   #Whether to use online speech synthesis (True for online, False for offline)
   ​
        # Large model configuration
       # llm_platform: 'ollama'              # Optional platforms: 'ollama', 'tongyi', 'spark', 'qianfan', 'openrouter'
       llm_platform: 'ollama'              # Currently selected large model platform
   ​
   

   • use_oline_asr and useolinetts must be set to False.

   • llm_platform must be set to ollama.

     With this configuration, everything will be offline.

3. Save the file and rebuild the project to apply changes:

   xxxxxxxxxx
   cd ~/yahboom_ws
   colcon build
   source install/setup.bash
   

After completing this step, the program is already a purely offline voice service.

2. Creating a Boot Auto-start Service (Systemd)

Now, we will create a systemd service to make largemodel_control.launch.py run automatically at system startup.

2.1 Creating a Startup Script

To enable systemd to correctly load the ROS2 environment, the best practice is to create a simple bash script to encapsulate our startup command.

1. Create the script file: In the directory (~/yahboom_ws/src/largemodel/), create a file named start_largemodel.sh.

   xxxxxxxxxx
   vim ~/yahboom_ws/src/largemodel/start_largemodel.sh
   

2. Write the script content: Copy and paste the following content into the script file.

   xxxxxxxxxx
   #!/bin/bash
   ​
   # Source ROS2 Jazzy environment
   source /opt/ros/jazzy/setup.bash
   ​
   # Source Yahboom workspace environment
   source /home/sunrise/yahboom_ws/install/setup.bash
   ​
   # Start largemodel control script
   ros2 launch largemodel largemodel_control.launch.py
   

   Important: Please ensure to replace /home/sunrise/ in the script with your own user home directory path.

3. Save and exit

4. Grant execute permission to the script:

   xxxxxxxxxx
   chmod +x ~/yahboom_ws/src/largemodel/start_largemodel.sh
   

2.2 Creating the Systemd Service File

This is the most critical step. We will tell the system that we have a new service to manage.

1. Create the service file: You need to use sudo privileges to create this file.

   xxxxxxxxxx
   sudo vim /etc/systemd/system/largemodel.service
   

2. Write the service configuration: Copy and paste the following content completely into the service file.

   xxxxxxxxxx
   [Unit]
   Description=Robot Service
   After=network.target sound.target graphical.target multi-user.target
   Wants=network.target sound.target graphical.target multi-user.target
   ​
   [Service]
   Type=simple
   User=sunrise
   Group=sunrise
   Environment=DISPLAY=:0
   Environment=XDG_RUNTIME_DIR=/run/user/1000
   Environment=PULSE_SERVER=unix:/run/user/1000/pulse/native
   SupplementaryGroups=audio video
   ExecStartPre=/bin/sleep 10
   ExecStart=/home/sunrise/yahboom_ws/src/largemodel/start_largemodel.sh
   Restart=on-failure
   StandardOutput=journal
   StandardError=journal
   ​
   [Install]
   WantedBy=multi-user.target
   

   • The path in ExecStart must exactly match the startup script path.

3. Save and exit.

2.3 Managing and Debugging the Service

Now that your service is created, we need to let systemd load it and set it to auto-start on boot.

1. Reload the systemd daemon to read our newly created service file:

   xxxxxxxxxx
   sudo systemctl daemon-reload
   

2. Set the service to auto-start on boot:

   xxxxxxxxxx
   sudo systemctl enable largemodel.service
   

3. Start the service immediately:

   xxxxxxxxxx
   sudo systemctl start largemodel.service
   

4. Check service status: This is the most important command to verify whether the service is running successfully.

   xxxxxxxxxx
   sudo systemctl status largemodel.service
   

   • If you see Active: active (running), congratulations, the service has started successfully!
   • If the status is failed or other, please continue to the next step for debugging.

5. View service logs (essential for debugging): If the service fails to start, you can view all real-time logs generated by the ros2 launch command through the following command, which is crucial for locating problems.

   xxxxxxxxxx
   journalctl -u largemodel.service -f
   

After completing all the above steps, every time the system boots, it will automatically start the purely offline largemodel voice service.

To disable the online voice assistant service, run the following commands

xxxxxxxxxx
sudo systemctl disable largemodel.service

sudo systemctl stop largemodel.service