Skip to content

datarabbit-ai/chatbot-workshop

Repository files navigation

Harnessing GenAI power with your own custom chatbot @ PyConPL 2024

Generative AI is a very hot topic nowadays. And rightfully so, as with the recent developments in the area, quite a lot of new business applications have been enabled or made significantly easier in the implementation. AI is currently also getting more accessible than ever, with some models becoming small enough to run on one personal device while remaining quite capable, or with all 3rd party models APIs (such as OpenAI’s) available.

During this workshop, we will take advantage of that fact (as well as APIs making these models accessible) to start the adventure with generative AI, and build our own chatbot/assistant with conversational capabilities and the ability to interact with custom data.

Workshop brought to you by datarabbit.ai.

Preliminary requirements

  • Python is installed on your machine, preferably in version 3.10 (this is the oldest version that we tested on).
  • There will be some API keys required for the workshop – please refer to the notebook initial sections for details.

When you are running this workshop on windows, installing build tools might be required to install some pip packages

  • open this link
  • click Download Build Tools >
  • open downloaded file and install C++ build tools with the installer
  • you may need to reboot your device after installation

How to set up the workshop environment

Create virtual environment

$ python3 -m venv venv

Activate it

On Linux/MacOS

$ source venv/bin/activate

On Windows

$ venv\bin\activate

Install required packages

$ pip3 install -r requirements.txt

If you want to run additional content with huggingface - install required packages for it:

$ pip3 install -r requirements-hg.txt

Start a Jupyter server...

$ jupyter lab .

... go to the notebook workshop.ipynb!

workshop_solutions.ipynb is the copy of the notebook – but with reference solutions to the exercises.

Starting demo chatbot app

create file .streamlit/secrets.toml with content:

OPENAI_API_KEY = "YOUR_API_KEY"

run from terminal:

$ pip3 install -r requirements-demo.txt
$ streamlit run app.py

Contact

If you tried out the configuration or actual tutorials prior to the workshop, after it ended, or you strolled here from the Internet and when working with the materials you encountered some issues, or perhaps you would like to implement some more advanced chatbot project and would need help - feel free to contact us.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published