Transformers required by ImageNet training example

[Yancey1989]

To process the ImageNet dataset, we need

• ImageNetDataset to load ImageNet images with .tar.gz file.
• convert HWC to CHW format

and the following transform functions:

• Transofrms framework Add GoTorch Transformer framework proposal #196
• RandomResizedCrop: Add random crop transform #197
• RandomHorizontalFlip: Add random flip transform #200
• ToTensor Add ToTensor function #187
• Normalizer: Add normalize transform #213
• Resize: https://godoc.org/github.com/disintegration/imaging#Resize
• CenterCrop: Add center clip transform #203

[wangkuiyi]

The example training program that uses the above transformers

https://github.com/pytorch/examples/blob/4b119d735b802453479d739bf823f3f7d8d5d422/imagenet/main.py#L204-L214

[wangkuiyi]

The validation program that uses the above transformers

https://github.com/pytorch/examples/blob/4b119d735b802453479d739bf823f3f7d8d5d422/imagenet/main.py#L225-L230