Rasa-Dx is envisioned as a web based platform that helps train Rasa contextual AI assistants at scale. It is designed to make the lives of those who work with ai-assistants easy, by giving them a no-coding web interface to adapt , train , test and maintain such assistants .
Adapting contextual AI assistants to specific domain is one of the bigger problems adopting chatbots within organizations. We aim to productionize cutting edge NLP solutions to help users complete the process of conveying domain knowledge to contextual assistants
At this juncture it layers on top of [Rasa Open Source] (https://rasa.com/)
Rasa DX only requires a recent version of Docker and Docker Compose.
Please do the below changes in docker/docker-compose.yml
-
set env variable server to public IP of the machine where trainer api docker container is running for example: http://localhost:8001/
-
set env variable SECRET_KEY to some random key.
-
use command openssl rand -hex 32 for generating random secret key
-
run the command.
cd docker/
docker-compose up -d- use username: test@demo.in and password: welcome@1 to try with demo user
git clone https://github.com/digiteinfotech/rasa-dx.git
cd rasa-dx/- For installing dependencies on windows run
setup.bat
- For installing dependencies on linux run
chmod 777 setup.sh
setup.sh
- For starting augment services run
uvicorn augment.server:app --host 0.0.0.0
- For starting trainer-api services run
uvicorn bot_trainer.api.app.main:app --host 0.0.0.0 --port 8080
We ❤️ contributions of all size and sorts. If you find a typo, if you want to improve a section of the documentation or if you want to help with a bug or a feature, here are the steps:
-
Fork the repo and create a new branch, say rasa-dx-issue1
-
Fix/improve the codebase
-
write test cases and documentation for code'
-
run test cases.
python -m pytest
- reformat code using black
python -m black bot_trainer
-
Commit the changes, with proper comments about the fix.
-
Make a pull request. It can simply be one of your commit messages.
-
Submit your pull request and wait for all checks passed.
-
Request reviews from one of the developers from our core team.
-
Get a 👍 and PR gets merged.
- Rasa - The bot framework used
- PiPy - Dependency Management
- Mongo - DB
- MongoEngine - ORM
- FastApi - Rest Api
- Uvicorn - ASGI Server
- Spacy - NLP
- Gensim - Embedding and Topic Modelling
- Sentence Transformer - Semantic Similarity
- Pytest - Testing
- MongoMock - Mocking DB
- Response - Mocking HTTP requests
- Black - Code Reformatting
- Digite, Inc. - Initial work - Digite
- sfahad1414
- deenaik
See also the list of contributors who participated in this project.
TODO