studio-lab-examples
studio-lab-examples copied to clipboard
aws
GPU unavailability issues
The objective of this issue is to summarize GPU unavailability issues to notify each issue when GPU is back.
If you encounter the "There is no runtime available right now. Please change the compute type or try again later." message, you do not need to raise an additional issue because we understand the problem and fighting to solve it 💪
The issues are ordered by time line.
2023
- March
- https://github.com/aws/studio-lab-examples/issues/202
- February
- https://github.com/aws/studio-lab-examples/issues/202
2022
- September
- https://github.com/aws/studio-lab-examples/issues/144
- August
- https://github.com/aws/studio-lab-examples/issues/141
- July
- https://github.com/aws/studio-lab-examples/issues/133
- June
- https://github.com/aws/studio-lab-examples/issues/115
- May
- https://github.com/aws/studio-lab-examples/issues/110
Requests about GPU
- https://github.com/aws/studio-lab-examples/issues/98
- https://github.com/aws/studio-lab-examples/issues/121
I cannot access to studio lab neither via CPU or GPU. No runtime available on both. I have tried to access for a week and the problem persists.
I now have access to a GPU instance but tensorflow is unable to list GPUs. I am able to see the GPU for my instance using !nvidia-smi but tf.config.list_physical_devices() doesn't list any GPUs, neither does adding the 'GPU' param. I've spent two days just trying to play with TensorFlow and this Studio Labs has made it impossible.
@RaistlinD2x If you have the TensorFlow specific problem, could you please raise an another issue? If nvidia-smi works, GPU itself is available.
