Inspiration
Peter Danenberg runs a Bard workshop on Fridays. On the first meetup, we explored shopping recommendations via Bard and saw that it hallucinated Amazon links. He figured out a way to ground it and this is part of that result.
What it does
Uses the BLIP model to summarize an image
How we built it
Image link is sent to a vision LLM and the summary is used as a search term for Amazon. Shopping links follow
Challenges we ran into
No real challenges. Deployment on BentoCloud for the BLIP model was straightforward.
Accomplishments that we're proud of
This took about 20 minutes, which says a lot about how easy it is to integrate new models from BentoML. The code examples definitely help.
What we learned
BentoML model deployment is straightforward and intuitive.
What's next for Vision shopping
More integrations with other vision models
Built With
- python
- streamlit
Log in or sign up for Devpost to join the conversation.