" Dataset for training an NLU model to provide healthcare assistance. "
The dataset is in the format specified by the requirements of RASA framework as the model is tested and deployed using the same.
- Clone/ Download the repo
- Create an env for installing rasa
- We use conda to create the virtual environment, open your preferred command line/terminal/shell in the extracted directory and all the commands are to be performed here
- Execute
conda create --prefix ./env_name python=3.6
to create an env in the same directory - If conda is not installed
- Activate the env using
conda activate ./env_name
- Install the dependencies
-
pip install rasa
" Note: The env creation is optional and can also be performed directly in the base env but it is recommended to create a new env as the dependencies are large and might effect the storage/performance if directly installed"
Step 1: Speech to Text
------ pip install SpeechRecognition PyAudio
Step 2: Externally sending an input to rasa chatbot
------ rasa run -m models --endpoints endpoints.yml --port 5002 --credentials credentials.yml
Step 3: Text to speech conversion
------ pip install gtts and also install the media player for playing the audio after conversion,
set any media player path to the project Now to see your voice bot in action run the Voice.py file and also run the action server using
------ rasa run actions
At last run the voicebot python file
- Open the directory and activate the env from comman line.
- Train the model using
rasa train
this creates a model in the models directory. - Run the model :
rasa shell
. - The trained AI chatbot model will be now active