log_gpu_status#

muse.log.log_gpu_status()#

Log CUDA/GPU availability.

Optional accelerator backends are imported lazily and this is not called at import time, so importing this module (or muse) stays backend-free.