Description: Python wrapper and package for Bot Framework REST API and State REST API (not production ready, yet).
A snippet of a basic bot:
from pybotframework.botframework import BotFramework
from pybotframework.sklearn_connector import SklearnConnector
# Labels for ML model
target_names = ['neg', 'pos']
# Instatiate the connector to custom ML model
sklearn_lang_conn = SklearnConnector(model_file='sentiment.pkl',
target_names=target_names)
# Instatiate the bot
my_app = BotFramework(connectors=[sklearn_lang_conn])
if __name__ == '__main__':
"""Run the Flask app"""
my_app.run_server(host='0.0.0.0', port=3978, debug=True)See the README in the examples folder for running a test bot.
Install the pybotframework package from this repository with
pip install .
or
python setup.py
- Bot Framework Emulator https://github.com/Microsoft/BotFramework-Emulator
- Python installed (Anaconda 3.5 recommended) https://anaconda.org/
- Python experience https://docs.python.org/3/tutorial
- Pybotframework library installed from repo (python setup.py or pip install .)
- Docker (if macOS get https://docs.docker.com/docker-for-mac/install/ and if Windows get https://docs.docker.com/docker-for-windows/install/)
- Ngrok https://ngrok.com/download
- TensorFlow (tensorflow==1.3.0) and Scikit-learn (scikit-learn==0.18.2) installed
- Some knowledge of ML and neural nets (helpful)
You will also need the following models and data:
- Scikit-learn sentiment model (click to download)
- TensorFlow word2vec model (click to download)
To build and run the Docker image ensure you have Docker running. https://docs.docker.com/engine/installation/
Then set up the botframework emulator.
https://docs.microsoft.com/en-us/bot-framework/debug-bots-emulator
Make sure to go through the "Install and configure ngrok" settings (ensure you have ngrok installed - see Requirements above). Uncheck the "Bypass ngrok for local addresses" to enable the bot to talk with Docker.
Now build the Docker image on the command line.
cd <pybotframework base directory>
docker build .
docker images
Search for your Docker "Image ID" in the list
docker run -p 3978:3978 -id <image ID> python
docker ps
You should see your container ID running.
Now try to connect to the bot by entering the address in the Emulator: http://localhost:3978/api/messages
You should see the messages "User added!" and "Bot added!" This means that you are set up!
- Flask project