KimWu

fail to download laion face image

8

hello, when I run ```bash download.sh ./laion_face_meta ./laion_face_data``` tha download fails ``` total - success: 0.005 - failed to download: 0.995 - failed to resize: 0.000 - images per sec:...

Pretrained model

12

Can trained models be provided, especially on the GQA dataset.

请问可以加载llava1.5，或者llava-next在huggingface上的模型做预训练吗？

I extracted a long video into frames with 1 fps. How can I modify the jepa code to support reading multiple frames of data input?