Setup Rasa SDK on local machine

Prakash Dale · August 9, 2020

ChatBOT   Python   Rasa

About RASA

Building contextual assistants & chatbots that really help customers is hard. Rasa provides infrastructure & tools necessary for high-performing, resilient, proprietary contextual assistants that work. With Rasa, all developers can create better text- and voice-based assistants.

Actions

Actions are the things your bot runs in response to user input. There are four kinds of actions in Rasa:

  • Utterance Actions: start with utter_ and send a specific message to the user.
  • Retrieval Actions: start with respond_ and send a message selected by a retrieval model.
  • Custom Actions: run arbitrary code and send any number of messages (or none).
  • Default Actions: e.g. action_listen, action_restart, action_default_fallback.

Setup

  • Create new conda environment. Mine is ‘rasasdksrc’
  • Go do e:\git\rasahq folder on command prompt
  • Get latest source code from git clone https://github.com/RasaHQ/rasa-sdk.git

  • Install poetry using command curl -sSL https://raw.githubusercontent.com/python-poetry/poetry/master/get-poetry.py | python

    If you run into any issue with above command then work-around is

    • Get content of the file https://raw.githubusercontent.com/python-poetry/poetry/master/get-poetry.py and save it in a file get_poetry.py on your machine.
    • Run python get_poetry.py
  • Run poetry install to install dependencies needed for rasa-sdk

Debugging

  • Switch to conda environment for rasa. Mine is rasasrc.
  • On command prompt, go to the folder where you want to create new rasa project.
  • Run command rasa init and then follow instructions to create new project.
  • Switch back to conda environment corresponding to rasa-sdk.
  • Run commnad python -m rasa_sdk --actions actions
  • You are all set.

Twitter, Facebook