|
|
## Read This Before You Start
|
|
|
|
|
|
To start natural language processing and start speaking with pepper the following nodes should be running simultaneously:
|
|
|
- microphone
|
|
|
|
|
|
- microphone
|
|
|
- animated_speech
|
|
|
- dialogflow
|
|
|
- gpt
|
|
|
- dialogflow
|
|
|
- gpt
|
|
|
- ai_core
|
|
|
|
|
|
If you run the nodes one by one in a terminal, you can hit `Ctrl+Z` and then type `bg` to send the current node into the background. To bring it into the foreground again, type `fg`. You can have several nodes running in the background.
|
|
|
|
|
|
If you need to build, any package using python files should be built with the symlink flag. E.g. in lhw_intelligence you would run `colcon build --symlink-install`. Otherwise you'll get an error. Don't forget to source the install files after building by `source install/setup.zsh` for instance. You run these commands from the root directory in each container.
|
|
|
|
|
|
If you need to build, any package using python files should be built with the symlink flag. E.g. in lhw_intelligence you would run `colcon build --symlink-install`. Otherwise you'll get an error. Don't forget to source the install files after building by `source install/setup.zsh` for instance. You run these commands from the root directory in each container.
|
|
|
|
|
|
## Starting microphone and animated_speech
|
|
|
|
... | ... | @@ -27,7 +28,7 @@ After this run: \ |
|
|
`source ./install/setup.zsh` \
|
|
|
Then try the ros2 command again. \
|
|
|
Finally run: \
|
|
|
`ros2 run lhw_qi animated_speech -i $PEPPER_IP` \
|
|
|
`ros2 run lhw_qi animated_speech -i $PEPPER_IP` \\
|
|
|
|
|
|
## Starting Dialogflow and GPT-2
|
|
|
|
... | ... | @@ -47,7 +48,7 @@ In the same docker container, run: \ |
|
|
`ros2 run lhw_nlp gpt` \
|
|
|
And finally run the following (don't run dialogflow unnecessarily, there's a daily quota): \
|
|
|
`ros2 run lhw_nlp dialogflow` \
|
|
|
(If you get an error regarding the existence of DialogflowV2 then run `pip3 install google-cloud-dialogflow` and continue) \
|
|
|
(If you get an error regarding the existence of DialogflowV2 then run `pip3 install google-cloud-dialogflow` and continue) \\
|
|
|
|
|
|
## Starting ai_core
|
|
|
|
... | ... | @@ -56,6 +57,18 @@ Navigate to: \ |
|
|
Then run, \
|
|
|
`docker-compose run lhw_intelligence zsh` \
|
|
|
Once in the docker, run \
|
|
|
`ros2 run lhw_intelligence ai_core` \
|
|
|
`ros2 run lhw_intelligence ai_core` \\
|
|
|
|
|
|
Now pepper is listening and ready to respond!
|
|
|
|
|
|
# Starting GPT-3
|
|
|
|
|
|
Now pepper is listening and ready to respond! |
|
|
\ No newline at end of file |
|
|
1. Kör `git checkout implement-gpt3-again`
|
|
|
2. Kör `docker-compose run lhw_qi zsh`
|
|
|
3. I containern kör följande kommandon:
|
|
|
4. `export PEPPER_IP=<peppers ip>` är ni i labbet är den 10.133.5.240, annars tryck på knappen under plattan så ska den säga den om ni sitter på samma wifi.
|
|
|
5. Kör `ros2 run lhw_qi microphone -i $PEPPER_IP` och `ros2 run lhw_qi animated_speech -i $PEPPER_IP` i samma terminal (CTRL + z och `bg` för att lägga något i bakgrunden)
|
|
|
6. Starta ny terminal
|
|
|
7. Kör `docker-compose run lhw_nlp zsh`
|
|
|
8. I containern kör båda följande kommandon:
|
|
|
9. `ros2 run lhw_nlp gpt3` och `ros2 run lhw_nlp dialogflow` |
|
|
\ No newline at end of file |