Great tool

[GeorgePearse]

This looks really really nice! Could you guys start a slack or similar, or do you think discussion + Q and A can do the same? SLack seems to be the default approach to have more community engagement.

My company's core product is image classification (radiology), but we also have a lot of text that we could use to provide weak labels. Would love a chance to talk to the core contributors.

I'm using QDrant on the image side, and voxel51 are looking to work with them for scalable hybrid similarity search. Would you guys consider a similar integration?

[phurwicz]

This looks really really nice! Could you guys start a slack or similar, or do you think discussion + Q and A can do the same? SLack seems to be the default approach to have more community engagement.

Glad to see more community :) Discord is available if you don’t mind being the first few to join!

My company's core product is image classification (radiology), but we also have a lot of text that we could use to provide weak labels. Would love a chance to talk to the core contributors.

Calling up @haochuanwei

I'm using QDrant on the image side, and voxel51 are looking to work with them for scalable hybrid similarity search. Would you guys consider a similar integration?

QDrant sounds better than the naive vector search that we have right now. In my opinion it would be ideal to allow users to supply their own vector search functions, and provide a built-in or two. @haochuanwei for inputs.

[haochuanwei]

Would love a chance to talk to the core contributors.

Sure, @GeorgePearse send me an email so we can chat over Zoom or something.

it would be ideal to allow users to supply their own vector search functions, and provide a built-in or two

This involves an abstract vector search engine. Linking #11 (comment)