All rospackages are available in luxai-qtrobot, you can first clone both software and tutorials to your local device.
You should link or copy the following package to your src under catkin_ws and run the catkin_make
- headers
- motors_moveit
- qt_gspeech_app
- qtrobot_ikfast_left_arm_plugin
- qtrobot_ikfast_right_arm_plugin
- audio_common (in case something goes wrong with audio, can be found here)
- qt_gspeech_app running for speech recognition, check setup insturctions here
- motors_moveit running for arm control, see more instruction here
- openai credentials for accessing OpenAI GPT models
- Controlling QTrobot arms using MoveIt
- A complete guide to build a Conversational Social Robot with QTrobot and ChatGPT
- QTrobot Motion and Actuators
- QTrobot Audio processing and Microphone
- SMACH Tutorials
Note that you may also see qt_gspeech_interface for speech recognition. It could be another solution but is not tested.
Install the required python packages:
sudo pip3 install -r requirements.txt
Add OpenAI key to gpt_bot.yaml file:
"OPENAI_KEY":""To use OpenAI gpt-3.5-turbo model set engine: "chatgpt" or if you want to use text-davinci-003 set it to engine: "davinci3". Remember do not push you OPENAI_KEY to the repo.
There are 5 pre-configured characters for 'gpt-3.5-turbo' you can use (qtrobot, fisherman, astronaut, therapist and gollum). Simply change in the 'gpt_bot.yaml' character parameter on line 12 and have fun talking with them.
To use custom character prompt. Change the parameter 'use_prompt' to true and write your 'prompt' instead of the template that is in the 'gpt_bot.yaml'. QTrobot will then take your prompt to get reposnes from GPT.
make a link (or copy) to gpt_smach in catkin workspace source directory and build it.
$ cd ~/catkin_ws/src
$ ln -s ~/robot/code/tutorials/examples/gpt_smach ./
$ cd ~/catkin_ws/
$ catkin_make
Make sure you are connecting to QTrobot's Wi-Fi hot spot and have internet access
Launch the motors_moveit for control
$ roslaunch motors_moveit moveit_qtrobot.launch
Launch the visualization process
$ roslaunch gpt_smach visualize.launch
Launch the gpt_smach main process
$ roslaunch gpt_smach gpt_smach.launch