Spaces:
Running
on
Zero
Running
on
Zero
Cannot export GLB
#3
by
DeathDaDev
- opened
With the free quota of ZERO GPU, it seems like I cannot export the glb as it requires more compute seconds than I have left.
I am genuinely curious, why have a list of images instead of the actual 3d model in the interface?
It saves compute if you don't like the staged results.
I completely understand the challenge with the limited compute quota on the ZERO GPU. Here are a few alternative approaches you can try:
- Local Docker Setup: If you have access to a machine with a compatible GPU, you can run the Docker container locally. This will give you full control over the compute resources and avoid the limitations of the free quota.
- Cloud Platforms: Consider using cloud services like AWS, Google Cloud, or Azure. These platforms offer GPU instances that can be used on a pay-as-you-go basis, providing more compute power and flexibility.
- Kaggle Notebooks: As you mentioned, Kaggle provides free GPU access. You can run the model within a Kaggle notebook environment, which will help you bypass some of the compute limitations. Just ensure that you enable GPU support in the settings.
- Optimizing the Model: Another approach could be optimizing the model to reduce compute requirements. For example, you can try lowering the resolution or simplifying certain steps in the process.