Docker Robot Server - DRS

A combination of Docker services to provide a conversational agent (e.g. one embodied by a Pepper humanoid social robot)with conversational abilities:

1. Speech Recognition → Whisper
2. Natural Language Understanding → Rasa
3. Speech Synthesis → Mimic 3
4. Web GUI → PHP

Clients

1. WatchOS (Swift)
2. Pepper Robot (Android Java) DRS Client

(Deprecated: Pepper Robot (NaoQi Python))

Usage

1. Make sure python 3.8 is installed, as it is needed for Rasa
   1. MacOS: brew install python@3.8
2. Install Rasa: python3.8 -m pip install rasa
3. Train Rasa Model: sudo rasa train
   1. Run Rasa Actions server: rasa run actions
   2. Run Rasa Model server in CLI: rm -rf models/;clear;rasa train --domain domain.yml --config config.yml;rasa run --enable-api --cors "*" --debug -p 5005
4. Make sure you have Docker installed and running
5. Start the containers with start-servers.sh
6. Validate that all containers are running with validate-containers.sh

Interacting with containers

Example: sudo docker exec -it conversational-agent-server /bin/bash

Trying Rasa locally without Docker export SKLEARN_ALLOW_DEPRECATED_SKLEARN_PACKAGE_INSTALL=True sudo python3.11 -m pip install scikit-learn dask --break-system-packages --ignore-requires-python

Curl Rasa Test curl --request POST http://localhost:5005/webhooks/rest/webhook --header 'content-type: application/json' --data '{ }"message": "hi" }'

Preparing Whisper API

1. Create user account

curl -X 'POST' \
  'http://localhost:9001/api/v1/users/' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "email": "user@email.com",
  "username": "whisper",
  "password": "whisper"
}'

2. Get api token

curl -X 'POST' \
  'http://localhost:9001/api/v1/users/get_token' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "email": "user@email.com",
  "password": "whisper"
}'