GitLab

* Adding tests to ImageFolder

* Adding DatasetFolder class

* Fix tests for pytest and code for lint checker

* Adding mock to requirements for ImageFolder tests

It was pointing to an old version where the validation pre-processing was the same as the training pre-processing, and thus had randomness in it (instead of CenterCrop).

* Add random affine transformation

* Rewrite __repr__

* Refactor affine transform and update tests

if on dtype -> elif on dtype

* Update transforms __repr__

* Add _pil_interpolation_to_str map for obtaining strings for PIL.Image interpolation options

* Fix typo

* Add RandomApply, RandomChoice, RandomOrder transformations

* Rename argument `proba` to `p`

* Set probability as configuration parameter in RandomHorizontalFlip and RandomHorizontalFlip (#414)

* Fix documentation

Tests should pass current release as well.

Fixes https://github.com/pytorch/vision/issues/399.

The only case of floating point supported by PIL seems to be `F`, so this should fix it.

Updated the doc of RandomResizedCrop.__call__() from 'Image to be flipped' to 'Image to be cropped and resized'

Fixes https://github.com/pytorch/vision/issues/397

* Add basic Omniglot dataset loader

* Remove unused import

* Add Omniglot random pair to sample pair of characters

* Precompute random set of pairs, deterministic after object instantiation

* Export OmniglotRandomPair via the datasets module interfact

* Fix naming convention, use sum instead of reduce

* Fix downloading to not download everything, fix Python2 syntax

* Fix end line lint

* Add random_seed, syntax fixes

* Remove randomized pair, take up as a separate generic wrapper

* Fix master conflict

* add __repr__ for datasets

* fix lint

Closes #309

* add __repr__ for all transform objects

* add tests for __repr__

* Update densenet.py

added missing weight init for densenet

* Update densenet.py

following official implementation [here](https://github.com/liuzhuang13/DenseNet/blob/master/models/densenet.lua#L160)

* Update densenet.py

changed to nn.init & fixed lint

* add scale and ratio params to RandomResizedCrop.get_params

* don't use Image as a context manager to avoid exceptions on close

* flake8

* EMNIST dataset + speedup *MNIST preprocessing

* param to cause FakeData to generate different images

docstring fix for Normalize: the std sequence should be (S1, ..., Sn).

* add tunable crop scale and aspect ratio in

* Added urls to Harris patches on Phototour

* fixed formatting

* fixed flake8 complains

* add docs for grayscale + random grayscale transforms

* add to_grayscale + randomGrayscale