https://github.com/martcpp/build-with-ai-uniben
this is a chat bot that uses gemini google ai model to create a chat bot
https://github.com/martcpp/build-with-ai-uniben
Last synced: 10 months ago
JSON representation
this is a chat bot that uses gemini google ai model to create a chat bot
- Host: GitHub
- URL: https://github.com/martcpp/build-with-ai-uniben
- Owner: martcpp
- Created: 2024-04-25T11:41:03.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2024-04-25T12:49:04.000Z (about 2 years ago)
- Last Synced: 2024-04-26T13:54:46.700Z (about 2 years ago)
- Language: Python
- Size: 1.63 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# AI Chat UI (Proof of Concept)
This demonstrates a simple UI (using Streamlit's [Chat elements](https://docs.streamlit.io/library/api-reference/chat)) that uses interacts with an AI via an api (Google's [Gemini model](https://ai.google.dev/docs/gemini_api_overview) here).
## Demo
## Install
0. Ideally create a new Python environment (`venv`, `conda`, etc.).
- Tested on Python 3.11.5
1. Run `pip install -r requirements.txt` to install relevant packages.
2. Get an API key from Google's Google AI Studio formerly [MakerSuite](https://makersuite.google.com/app/apikey).
- See the Gemini API [docs](https://ai.google.dev/tutorials/setup) for more details.
3. Add your API key as an environment variable called `GOOGLE_API_KEY`.
- If you are on a Mac, this likely is editing `~/.zshrc` or `~/.bashrc` and adding `export GOOGLE_API_KEY=...` with your own key. Make sure you reload the shell.
- If you're having trouble and like to just get it to work, you can change the app script `app_chat.py` so the variable `GOOGLE_API_KEY` is set to the string value of your key. Note this is a bit more risky but probably fine if you're testing out without public access.
4. Run in your terminal: `streamlit run app_chat.py`.
- When using a Python environment, make sure `streamlit` is the correct one associated with your environment. Seen some weirdness before with `conda` and the shell not picking up the correct `streamlit`; had to manually call it with something like `MyUser/conda/envs/my_env/bin/streamlit`.
5. Browser should open where you can now interact with the UI! Otherwise, you might have to go to http://localhost:8501/ manually.