Spaces:
Running
on
Zero
About the GPU Accessibility for the Project GeoWizard
Dear HuggingFace Community,
Thank you for providing a ZeroGPU environment for our GeoWizard space eight months ago through the community grant program. With your support, we successfully deployed our application, earning 104 likes and engaging researchers.
However, we’ve encountered a compiling error likely due to the CPU-only environment. Our local testing confirmed that the application runs properly in a GPU-enabled setup. We believe the GPU resources for our space may have been recycled, though we didn’t receive notification.
Could a GPU environment be re-assigned to GeoWizard? This would help us restore functionality and continue serving the community. Please let us know if further details are needed.
Best regards,
GeoWizard Authors.
This issue is probably not directly related to Zero GPU space, and it seems to be an error that occurs in both CPU space and Zero GPU space, and in pay-as-you-go spaces if the conditions are met.
Specifically, it is more likely to occur when you place files of several GB or more in Spaces. In this case, the model file is probably the trigger.
I think you can avoid this by moving these files to the model repo and rewriting the program to reference them from Spaces.
We're waiting for a solution, but it seems to be a difficult opponent.
https://discuss.huggingface.co/t/build-error-error-while-cloning-repository/113801
This issue is probably not directly related to Zero GPU space, and it seems to be an error that occurs in both CPU space and Zero GPU space, and in pay-as-you-go spaces if the conditions are met.
Specifically, it is more likely to occur when you place files of several GB or more in Spaces. In this case, the model file is probably the trigger.
I think you can avoid this by moving these files to the model repo and rewriting the program to reference them from Spaces.
We're waiting for a solution, but it seems to be a difficult opponent.
https://discuss.huggingface.co/t/build-error-error-while-cloning-repository/113801
Thank you for your experience John. After removing most of the weight files, however, we have now found that our space could indeed suffer from device issue:
