torch.accelerator.is_available#

torch.accelerator.is_available()[source]#

Check if the current accelerator is available at runtime: it was build, all therequired drivers are available and at least one device is visible.Seeaccelerator for details.

Returns

A boolean indicating if there is an availableaccelerator.

Return type

bool

Note

This API delegates to the device-specific version ofis_available.On CUDA, when the environment variablePYTORCH_NVML_BASED_CUDA_CHECK=1 is set,this function will NOT poison fork. Otherwise, it will. For more details, seePoison fork in multiprocessing.

Example:

>>>asserttorch.accelerator.is_available()"No available accelerators detected."