cilin yan

icon indicating a website link https://cilinyan.github.io/ icon indicating an email [email protected]

icon company BUAA icon location Beijing icon location I am a M.S. student at Beihang University, primarily focused on researching deep learning, and multimodal.

Results 5 comments of cilin yan

Make a wish for a docker image to run mamba.

Thank you so much for considering my request amidst your busy schedule with the ICML deadline. Wishing you all the best for the conference.

Shape cannot match the size during training

> You can try to use multiple gpus to run! And the error will go away! One simple approach is to ensure that only one video is trained on each...

What are the compute resources required for this?

Could you please provide the estimated training time for CLIPSelf on the COCO and CC3M datasets? Thanks!

What are the compute resources required for this?

Thank you for your reply! > The training of EVA-CLIP models is fast. With 8 A100-80Gs, it takes about 2 hours to train a ViT-B/16 on COCO for 6 epochs...

Text Prompt "person/human" Not Working for Segmentation (Works with "shoe")

After testing, the text prompt "a person" now works for human segmentation (where "person" or "human" alone failed earlier)!