SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Add Vision Transformer and ViTFeatureExtractor (#10950)

Browse Source 
* Squash all commits into one

* Update ViTFeatureExtractor to use image_utils instead of torchvision

* Remove torchvision and add Pillow

* Small docs improvement

* Address most comments by @sgugger

* Fix tests

* Clean up conversion script

* Pooler first draft

* Fix quality

* Improve conversion script

* Make style and quality

* Make fix-copies

* Minor docs improvements

* Should use fix-copies instead of manual handling

* Revert "Should use fix-copies instead of manual handling"

This reverts commit fd4e591bce4496d41406425c82606a8fdaf8a50b.

* Place ViT in alphabetical order

Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
This commit is contained in:
NielsRogge
2021-04-01 17:16:05 +02:00
committed by GitHub
parent af6732225c
commit 30677dc743
25 changed files with 3072 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
docs/source/index.rst
View File

@@ -210,22 +210,26 @@ and conversion utilities for the following models:
43. :doc:`Transformer-XL <model_doc/transformerxl>` (from Google/CMU) released with the paper `Transformer-XL:
Attentive Language Models Beyond a Fixed-Length Context <https://arxiv.org/abs/1901.02860>`__ by Zihang Dai*,
Zhilin Yang*, Yiming Yang, Jaime Carbonell, Quoc V. Le, Ruslan Salakhutdinov.
44. :doc:`Wav2Vec2 <model_doc/wav2vec2>` (from Facebook AI) released with the paper `wav2vec 2.0: A Framework for
44. :doc:`Vision Transformer (ViT) <model_doc/vit>` (from Google AI) released with the paper `An Image is Worth 16x16
Words: Transformers for Image Recognition at Scale <https://arxiv.org/abs/2010.11929>`__ by Alexey Dosovitskiy,
Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias
Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby.
45. :doc:`Wav2Vec2 <model_doc/wav2vec2>` (from Facebook AI) released with the paper `wav2vec 2.0: A Framework for
Self-Supervised Learning of Speech Representations <https://arxiv.org/abs/2006.11477>`__ by Alexei Baevski, Henry
Zhou, Abdelrahman Mohamed, Michael Auli.
45. :doc:`XLM <model_doc/xlm>` (from Facebook) released together with the paper `Cross-lingual Language Model
46. :doc:`XLM <model_doc/xlm>` (from Facebook) released together with the paper `Cross-lingual Language Model
Pretraining <https://arxiv.org/abs/1901.07291>`__ by Guillaume Lample and Alexis Conneau.
46. :doc:`XLM-ProphetNet <model_doc/xlmprophetnet>` (from Microsoft Research) released with the paper `ProphetNet:
47. :doc:`XLM-ProphetNet <model_doc/xlmprophetnet>` (from Microsoft Research) released with the paper `ProphetNet:
Predicting Future N-gram for Sequence-to-Sequence Pre-training <https://arxiv.org/abs/2001.04063>`__ by Yu Yan,
Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
47. :doc:`XLM-RoBERTa <model_doc/xlmroberta>` (from Facebook AI), released together with the paper `Unsupervised
48. :doc:`XLM-RoBERTa <model_doc/xlmroberta>` (from Facebook AI), released together with the paper `Unsupervised
Cross-lingual Representation Learning at Scale <https://arxiv.org/abs/1911.02116>`__ by Alexis Conneau*, Kartikay
Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke
Zettlemoyer and Veselin Stoyanov.
48. :doc:`XLNet <model_doc/xlnet>` (from Google/CMU) released with the paper `​XLNet: Generalized Autoregressive
49. :doc:`XLNet <model_doc/xlnet>` (from Google/CMU) released with the paper `​XLNet: Generalized Autoregressive
Pretraining for Language Understanding <https://arxiv.org/abs/1906.08237>`__ by Zhilin Yang*, Zihang Dai*, Yiming
Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
49. :doc:`XLSR-Wav2Vec2 <model_doc/xlsr_wav2vec2>` (from Facebook AI) released with the paper `Unsupervised
50. :doc:`XLSR-Wav2Vec2 <model_doc/xlsr_wav2vec2>` (from Facebook AI) released with the paper `Unsupervised
Cross-Lingual Representation Learning For Speech Recognition <https://arxiv.org/abs/2006.13979>`__ by Alexis
Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
@@ -328,6 +332,8 @@ TensorFlow and/or Flax.
+-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
| Transformer-XL | ✅ | ❌ | ✅ | ✅ | ❌ |
+-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
| ViT | ❌ | ❌ | ✅ | ❌ | ❌ |
+-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
| Wav2Vec2 | ✅ | ❌ | ✅ | ❌ | ❌ |
+-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
| XLM | ✅ | ❌ | ✅ | ✅ | ❌ |
@@ -460,6 +466,7 @@ TensorFlow and/or Flax.
model_doc/t5
model_doc/tapas
model_doc/transformerxl
model_doc/vit
model_doc/wav2vec2
model_doc/xlm
model_doc/xlmprophetnet

102
docs/source/model_doc/vit.rst Normal file
View File

@@ -0,0 +1,102 @@
..
Copyright 2020 The HuggingFace Team. All rights reserved.
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
specific language governing permissions and limitations under the License.
Vision Transformer (ViT)
-----------------------------------------------------------------------------------------------------------------------
.. note::
This is a recently introduced model so the API hasn't been tested extensively. There may be some bugs or slight
breaking changes to fix it in the future. If you see something strange, file a `Github Issue
<https://github.com/huggingface/transformers/issues/new?assignees=&labels=&template=bug-report.md&title>`__.
Overview
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The Vision Transformer (ViT) model was proposed in `An Image is Worth 16x16 Words: Transformers for Image Recognition
at Scale <https://arxiv.org/abs/2010.11929>`__ by Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk
Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob
Uszkoreit, Neil Houlsby. It's the first paper that successfully trains a Transformer encoder on ImageNet, attaining
very good results compared to familiar convolutional architectures.
The abstract from the paper is the following:
*While the Transformer architecture has become the de-facto standard for natural language processing tasks, its
applications to computer vision remain limited. In vision, attention is either applied in conjunction with
convolutional networks, or used to replace certain components of convolutional networks while keeping their overall
structure in place. We show that this reliance on CNNs is not necessary and a pure transformer applied directly to
sequences of image patches can perform very well on image classification tasks. When pre-trained on large amounts of
data and transferred to multiple mid-sized or small image recognition benchmarks (ImageNet, CIFAR-100, VTAB, etc.),
Vision Transformer (ViT) attains excellent results compared to state-of-the-art convolutional networks while requiring
substantially fewer computational resources to train.*
Tips:
- To feed images to the Transformer encoder, each image is split into a sequence of fixed-size non-overlapping patches,
which are then linearly embedded. A [CLS] token is added to serve as representation of an entire image, which can be
used for classification. The authors also add absolute position embeddings, and feed the resulting sequence of
vectors to a standard Transformer encoder.
- The Vision Transformer was pre-trained using a resolution of 224x224. During fine-tuning, it is often beneficial to
use a higher resolution than pre-training `(Touvron et al., 2019) <https://arxiv.org/abs/1906.06423>`__, `(Kolesnikov
et al., 2020) <https://arxiv.org/abs/1912.11370>`__. The authors report the best results with a resolution of 384x384
during fine-tuning.
- As the Vision Transformer expects each image to be of the same size (resolution), one can use
:class:`~transformers.ViTFeatureExtractor` to resize (or rescale) and normalize images for the model.
- Both the patch resolution and image resolution used during pre-training or fine-tuning are reflected in the name of
each checkpoint. For example, :obj:`google/vit-base-patch16-224` refers to a base-sized architecture with patch
resolution of 16x16 and fine-tuning resolution of 224x224. All checkpoints can be found on the `hub
<https://huggingface.co/models?search=vit>`__.
- The available checkpoints are either (1) pre-trained on `ImageNet-21k <http://www.image-net.org/>`__ (a collection of
14 million images and 21k classes) only, or (2) also fine-tuned on `ImageNet
<http://www.image-net.org/challenges/LSVRC/2012/>`__ (also referred to as ILSVRC 2012, a collection of 1.3 million
images and 1,000 classes).
- The best results are obtained with supervised pre-training, which is not the case in NLP. The authors also performed
an experiment with a self-supervised pre-training objective, namely masked patched prediction (inspired by masked
language modeling). With this approach, the smaller ViT-B/16 model achieves 79.9% accuracy on ImageNet, a significant
improvement of 2% to training from scratch, but still 4% behind supervised pre-training.
The original code (written in JAX) can be found `here <https://github.com/google-research/vision_transformer>`__.
Note that we converted the weights from Ross Wightman's `timm library
<https://github.com/rwightman/pytorch-image-models>`__, who already converted the weights from JAX to PyTorch. Credits
go to him!
ViTConfig
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. autoclass:: transformers.ViTConfig
:members:
ViTFeatureExtractor
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. autoclass:: transformers.ViTFeatureExtractor
:members: __call__
ViTModel
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. autoclass:: transformers.ViTModel
:members: forward
ViTForImageClassification
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. autoclass:: transformers.ViTForImageClassification
:members: forward