labenz
labenz
@LiJunnan1992 – appreciate all your engagement in the discussion! I'm also interested in fine-tuning BLIP on a proprietary dataset in an effort to generate "contextual captions" and have a few...
thanks very much for the feedback – appreciate it!!
Thank you! It seems this only works for nucleus sampling method – is that right? Thanks again. :)
thanks for the pointer! Changing the runtime to Python 3.8 worked, as suggested [here](https://github.com/keithrozario/Klayers/issues/154#issuecomment-1115739823) Neither 3.7 nor 3.9 worked, though, so there does still seem to be an issue out...
thanks for the clarification. at the risk of asking another newb question... is there a reason many packages available in 3.8 are not available in 3.9? I notice that Pillow,...
and one more possibly dumb question: is there a reason that common ML packages are not included in this project? I'm thinking about things like... - transformers - timm -...
understood – thank you!
We ended up using a different approach, which used BLIP image-text matching instead of captioning. (For context, our problem was “image selection”, so we found that generating “ideal captions” and...
Looks like this made it about 1/3rd of the way through training and was stopped? Any plans to continue? > I've opened up the wandb for this model if anyone...
Love it - thanks for the update. Believe this would be a differentiating feature for sure. Do you have a commercial partner in mind? I’ve tried many but of course...