I am wondering, if the notebooks have been tested at all using an Apple M1 (or possibly M2) as the platform instead of CUDA? I am continuously facing issues related to this, even when switching the device explicitly to MPS (which is the device type required on M1).
I don't have time right now to provide all the details, but it would be simple enough to replicate when trying to run the notebooks on a M1 Macbook and trying to set the device type to the MPS backend. Ideally, this should also be incorporated into the notebook code itself.
Information
The question or comment is about chapter:
Applies to several chapteers.
Question or comment
I am wondering, if the notebooks have been tested at all using an Apple M1 (or possibly M2) as the platform instead of CUDA? I am continuously facing issues related to this, even when switching the device explicitly to MPS (which is the device type required on M1). I don't have time right now to provide all the details, but it would be simple enough to replicate when trying to run the notebooks on a M1 Macbook and trying to set the device type to the MPS backend. Ideally, this should also be incorporated into the notebook code itself.