Extensions for the `fastai2` library in the form of data augmentations and inference utilities. Specific to computer vision models. Inference utilities are built with a focus on single-label (softmax) classifiers.
In recent versions of Pytorch the to_cuda function used here was not working as it should and there were some issues with device calculations. Hence, here it is proposed to change to_cuda function with directly setting model and x to cuda if available.
PD: Several formatting changes are done here automatically by black package but they are not changing the funtionality of the code.
In recent versions of Pytorch the
to_cudafunction used here was not working as it should and there were some issues with device calculations. Hence, here it is proposed to changeto_cudafunction with directly settingmodelandxto cuda if available.PD: Several formatting changes are done here automatically by
blackpackage but they are not changing the funtionality of the code.