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

Problem in data generation #23

Open
iu110 opened this issue Apr 18, 2024 · 3 comments
Open

Problem in data generation #23

iu110 opened this issue Apr 18, 2024 · 3 comments

Comments

@iu110
Copy link

iu110 commented Apr 18, 2024

It's a amazing work. I am currently training the model with my own data using the dataset-generation code you provided.

微信图片_20240418175532

why always show this? please tell me what to do with it.

@iu110
Copy link
Author

iu110 commented Apr 18, 2024

I am training on 4 RTX4090 GPUs with 24Gb.

@iu110
Copy link
Author

iu110 commented Apr 18, 2024

微信图片_20240418180000

@frank-xwang
Copy link
Owner

Hi! It looks like you are currently using the CPU version of ms_deform_attn. For optimal performance, you might want to install the CUDA version. You can find detailed instructions on how to do this on the Mask2Former installation guide. Please ensure that the CUDA version used with your PyTorch installation (torch.__version__ in Python) matches the CUDA version used to install ms_deform_attn (nvcc --version in Linux). And please make sure CUDA_HOME is set. You can also check this instruction on some common CUDA bugs.

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

2 participants