search

baaivision / DIVA

Diffusion Feedback Helps CLIP See Better
https://rubics-xuan.github.io/DIVA/
MIT License
213 stars 11 forks source link

Looking forward to releasing CLIP model weights #3

Closed dulibubai closed 3 weeks ago

dulibubai commented 3 months ago

Nice works!!! Looking forward to releasing CLIP model weights !

NielsRogge commented 3 months ago

Indeed, if they pushed to HF, it would be great to link them to https://huggingface.co/papers/2407.20171 so that people can discover the models from the paper page.

See here on how to do that: https://huggingface.co/docs/hub/en/model-cards#linking-a-paper

Rubics-Xuan commented 3 months ago

Many thanks for your great interests on our work! We will release all the CLIP model weights very soon to serve the research community and facilitate everyone to use our DIVA-tuned CLIP models. Please stay tuned!

sipie800 commented 3 months ago

Great work! Will you guys release the trained MLLM demo trained on the CLIP? That would be exciting to get an instant insight of the excellect granularity in a MLLM task. Thanks.

Rubics-Xuan commented 3 months ago

Please refer to LLaVA(https://github.com/haotian-liu/LLaVA) for the MLLM training details.

sipie800 commented 3 months ago

Few people can afford training a MLLM while most people afford doing inference.