FP16 about input and loss?

I have two questions about how to train the network correctly with fp16?

First, In main_fp16_optimizer.py, input will be .half() in data_prefetcher(), and model = network_to_half(model). Should input.half be necessary? #58

train_dataset = datasets.ImageFolder(
        traindir,
        transforms.Compose([
            transforms.RandomResizedCrop(crop_size),
            transforms.RandomHorizontalFlip(),
            # transforms.ToTensor(), Too slow
            # normalize,
        ]))

Second, should we concern about the operation in the criterion (loss function), which may be more complicated such as the loss function in object detection and sementation ?

if args.fp16:
            optimizer.backward(loss)

Start with main_fp16_optimizer.py, data_prefetcher(), network_to_half(model), and the optimizer.backward(loss) call. Determine from the existing training flow how input and criterion operations are expected to handle fp16, then document clear guidance for standard and more complex losses; done means both questions have an unambiguous usage answer.