Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Llava Data Points Missing #21

Open
Ruining0916 opened this issue Feb 13, 2025 · 5 comments
Open

Llava Data Points Missing #21

Ruining0916 opened this issue Feb 13, 2025 · 5 comments

Comments

@Ruining0916
Copy link

Hi Authors,

Thanks a lot for your amazing work! I wish to confirm with you that is it normal for some of the llava images are missing from zipped datasets from huggingface but required from dataloader in your code?

for example,

  • data/llava_data/llava_images/vg/VG_100K_2/214.jpg
  • data/llava_data/llava_images/vg/VG_100K/497987.jpg
  • data/llava_data/llava_images/vg/VG_100K/285815.jpg
  • ...

A lot of images are missing from the llava_images from huggingface, but your script is trying to find them?

Thanks,
Ruining

@lxtGH
Copy link
Collaborator

lxtGH commented Feb 15, 2025

@zhang-tao-whu Check this part.

@bill-hx-liu
Copy link

@Ruining0916
Hi Ruining0916 ,

I have encountered the same issue. Could you please share if there is a solution available?

Best regards

@Ruining0916
Copy link
Author

No, I haven't found the solution. I wonder if these images are supposed to be missing.

@bill-hx-liu
Copy link

ok,thanks!

@bill-hx-liu
Copy link

@Ruining0916
Hi Ruining0916 ,

could you tell me, what GPU model and VRAM size do you use for training for the sa2va_4b?

Best regards

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants