Release: v4.22.1

* Fix tokenizer load from one file

* Add a test

* Style

Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>

README.md

@@ -249,6 +249,8 @@ conda install -c huggingface transformers
 
 Follow the installation pages of Flax, PyTorch or TensorFlow to see how to install them with conda.
 
+> **_NOTE:_**  On Windows, you may be prompted to activate Developer Mode in order to benefit from caching. If this is not an option for you, please let us know in [this issue](https://github.com/huggingface/huggingface_hub/issues/1062).
+
 ## Model architectures
 
 **[All the model checkpoints](https://huggingface.co/models)** provided by 🤗 Transformers are seamlessly integrated from the huggingface.co [model hub](https://huggingface.co) where they are uploaded directly by [users](https://huggingface.co/users) and [organizations](https://huggingface.co/organizations).
@@ -290,12 +292,12 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (from Microsoft Research) released with the paper [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://arxiv.org/abs/1911.00536) by Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan.
 1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (from HuggingFace), released together with the paper [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108) by Victor Sanh, Lysandre Debut and Thomas Wolf. The same method has been applied to compress GPT2 into [DistilGPT2](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation), RoBERTa into [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation), Multilingual BERT into [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation) and a German version of DistilBERT.
 1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (from Microsoft Research) released with the paper [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) by Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei.
-1. **[Donut](https://huggingface.co/docs/transformers/main/model_doc/donut)** (from NAVER), released together with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
+1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (from NAVER), released together with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
 1. **[DPR](https://huggingface.co/docs/transformers/model_doc/dpr)** (from Facebook) released with the paper [Dense Passage Retrieval for Open-Domain Question Answering](https://arxiv.org/abs/2004.04906) by Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih.
 1. **[DPT](https://huggingface.co/docs/transformers/master/model_doc/dpt)** (from Intel Labs) released with the paper [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) by René Ranftl, Alexey Bochkovskiy, Vladlen Koltun.
 1. **[ELECTRA](https://huggingface.co/docs/transformers/model_doc/electra)** (from Google Research/Stanford University) released with the paper [ELECTRA: Pre-training text encoders as discriminators rather than generators](https://arxiv.org/abs/2003.10555) by Kevin Clark, Minh-Thang Luong, Quoc V. Le, Christopher D. Manning.
 1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (from Google Research) released with the paper [Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) by Sascha Rothe, Shashi Narayan, Aliaksei Severyn.
-1. **[ERNIE](https://huggingface.co/docs/transformers/main/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
 1. **[FlauBERT](https://huggingface.co/docs/transformers/model_doc/flaubert)** (from CNRS) released with the paper [FlauBERT: Unsupervised Language Model Pre-training for French](https://arxiv.org/abs/1912.05372) by Hang Le, Loïc Vial, Jibril Frej, Vincent Segonne, Maximin Coavoux, Benjamin Lecouteux, Alexandre Allauzen, Benoît Crabbé, Laurent Besacier, Didier Schwab.
 1. **[FLAVA](https://huggingface.co/docs/transformers/model_doc/flava)** (from Facebook AI) released with the paper [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) by Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach, and Douwe Kiela.
 1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (from Google Research) released with the paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
@@ -340,7 +342,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
-1. **[PEGASUS-X](https://huggingface.co/docs/transformers/main/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, and Peter J. Liu.
+1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, and Peter J. Liu.
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.
 1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
@@ -363,7 +365,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (from Tel Aviv University), released together with the paper [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) by Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy.
 1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (from Berkeley) released with the paper [SqueezeBERT: What can computer vision teach NLP about efficient neural networks?](https://arxiv.org/abs/2006.11316) by Forrest N. Iandola, Albert E. Shaw, Ravi Krishna, and Kurt W. Keutzer.
 1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (from Microsoft) released with the paper [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
-1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/main/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
 1. **[T5](https://huggingface.co/docs/transformers/model_doc/t5)** (from Google AI) released with the paper [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/abs/1910.10683) by Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu.
 1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (from Google AI) released in the repository [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) by Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu.
 1. **[TAPAS](https://huggingface.co/docs/transformers/model_doc/tapas)** (from Google AI) released with the paper [TAPAS: Weakly Supervised Table Parsing via Pre-training](https://arxiv.org/abs/2004.02349) by Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Müller, Francesco Piccinno and Julian Martin Eisenschlos.
@@ -375,7 +377,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (from Microsoft Research) released with the paper [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) by Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang.
 1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (from Microsoft Research) released with the paper [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) by Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu.
 1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (from Tsinghua University and Nankai University) released with the paper [Visual Attention Network](https://arxiv.org/abs/2202.09741) by Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
-1. **[VideoMAE](https://huggingface.co/docs/transformers/main/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
+1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
 1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (from NAVER AI Lab/Kakao Enterprise/Kakao Brain) released with the paper [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) by Wonjae Kim, Bokyung Son, Ildoo Kim.
 1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/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.
 1. **[VisualBERT](https://huggingface.co/docs/transformers/model_doc/visual_bert)** (from UCLA NLP) released with the paper [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) by Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang.
@@ -384,7 +386,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (from Facebook AI) released with the paper [FAIRSEQ S2T: Fast Speech-to-Text Modeling with FAIRSEQ](https://arxiv.org/abs/2010.05171) by Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Sravya Popuri, Dmytro Okhonko, Juan Pino.
 1. **[Wav2Vec2Phoneme](https://huggingface.co/docs/transformers/model_doc/wav2vec2_phoneme)** (from Facebook AI) released with the paper [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) by Qiantong Xu, Alexei Baevski, Michael Auli.
 1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
-1. **[X-CLIP](https://huggingface.co/docs/transformers/main/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
 1. **[XGLM](https://huggingface.co/docs/transformers/model_doc/xglm)** (From Facebook AI) released with the paper [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) by Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li.
 1. **[XLM](https://huggingface.co/docs/transformers/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.
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (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.

README_ko.md

@@ -242,12 +242,12 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (from Microsoft Research) released with the paper [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://arxiv.org/abs/1911.00536) by Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan.
 1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (from HuggingFace), released together with the paper [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108) by Victor Sanh, Lysandre Debut and Thomas Wolf. The same method has been applied to compress GPT2 into [DistilGPT2](https://github.com/huggingface/transformers/tree/main/examples/distillation), RoBERTa into [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/distillation), Multilingual BERT into [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/distillation) and a German version of DistilBERT.
 1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (from Microsoft Research) released with the paper [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) by Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei.
-1. **[Donut](https://huggingface.co/docs/transformers/main/model_doc/donut)** (from NAVER) released with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
+1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (from NAVER) released with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
 1. **[DPR](https://huggingface.co/docs/transformers/model_doc/dpr)** (from Facebook) released with the paper [Dense Passage Retrieval for Open-Domain Question Answering](https://arxiv.org/abs/2004.04906) by Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih.
 1. **[DPT](https://huggingface.co/docs/transformers/master/model_doc/dpt)** (from Intel Labs) released with the paper [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) by René Ranftl, Alexey Bochkovskiy, Vladlen Koltun.
 1. **[ELECTRA](https://huggingface.co/docs/transformers/model_doc/electra)** (from Google Research/Stanford University) released with the paper [ELECTRA: Pre-training text encoders as discriminators rather than generators](https://arxiv.org/abs/2003.10555) by Kevin Clark, Minh-Thang Luong, Quoc V. Le, Christopher D. Manning.
 1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (from Google Research) released with the paper [Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) by Sascha Rothe, Shashi Narayan, Aliaksei Severyn.
-1. **[ERNIE](https://huggingface.co/docs/transformers/main/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
 1. **[FlauBERT](https://huggingface.co/docs/transformers/model_doc/flaubert)** (from CNRS) released with the paper [FlauBERT: Unsupervised Language Model Pre-training for French](https://arxiv.org/abs/1912.05372) by Hang Le, Loïc Vial, Jibril Frej, Vincent Segonne, Maximin Coavoux, Benjamin Lecouteux, Alexandre Allauzen, Benoît Crabbé, Laurent Besacier, Didier Schwab.
 1. **[FLAVA](https://huggingface.co/docs/transformers/model_doc/flava)** (from Facebook AI) released with the paper [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) by Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach, and Douwe Kiela.
 1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (from Google Research) released with the paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
@@ -292,7 +292,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
-1. **[PEGASUS-X](https://huggingface.co/docs/transformers/main/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, Peter J. Liu. 
+1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, Peter J. Liu. 
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.
 1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
@@ -315,7 +315,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (from Tel Aviv University), released together with the paper [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) by Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy.
 1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (from Berkeley) released with the paper [SqueezeBERT: What can computer vision teach NLP about efficient neural networks?](https://arxiv.org/abs/2006.11316) by Forrest N. Iandola, Albert E. Shaw, Ravi Krishna, and Kurt W. Keutzer.
 1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (from Microsoft) released with the paper [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
-1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/main/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
 1. **[T5](https://huggingface.co/docs/transformers/model_doc/t5)** (from Google AI) released with the paper [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/abs/1910.10683) by Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu.
 1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (from Google AI) released in the repository [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) by Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu.
 1. **[TAPAS](https://huggingface.co/docs/transformers/model_doc/tapas)** (from Google AI) released with the paper [TAPAS: Weakly Supervised Table Parsing via Pre-training](https://arxiv.org/abs/2004.02349) by Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Müller, Francesco Piccinno and Julian Martin Eisenschlos.
@@ -327,7 +327,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (from Microsoft Research) released with the paper [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) by Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang.
 1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (from Microsoft Research) released with the paper [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) by Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu.
 1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (from Tsinghua University and Nankai University) released with the paper [Visual Attention Network](https://arxiv.org/pdf/2202.09741.pdf) by Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
-1. **[VideoMAE](https://huggingface.co/docs/transformers/main/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
+1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
 1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (from NAVER AI Lab/Kakao Enterprise/Kakao Brain) released with the paper [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) by Wonjae Kim, Bokyung Son, Ildoo Kim.
 1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/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.
 1. **[VisualBERT](https://huggingface.co/docs/transformers/model_doc/visual_bert)** (from UCLA NLP) released with the paper [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) by Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang.
@@ -336,7 +336,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (from Facebook AI) released with the paper [FAIRSEQ S2T: Fast Speech-to-Text Modeling with FAIRSEQ](https://arxiv.org/abs/2010.05171) by Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Sravya Popuri, Dmytro Okhonko, Juan Pino.
 1. **[Wav2Vec2Phoneme](https://huggingface.co/docs/transformers/model_doc/wav2vec2_phoneme)** (from Facebook AI) released with the paper [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) by Qiantong Xu, Alexei Baevski, Michael Auli.
 1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
-1. **[X-CLIP](https://huggingface.co/docs/transformers/main/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
 1. **[XGLM](https://huggingface.co/docs/transformers/model_doc/xglm)** (From Facebook AI) released with the paper [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) by Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li. 
 1. **[XLM](https://huggingface.co/docs/transformers/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.
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (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.

README_zh-hans.md

@@ -266,12 +266,12 @@ conda install -c huggingface transformers
 1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (来自 Microsoft Research) 伴随论文 [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://arxiv.org/abs/1911.00536) 由 Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan 发布。
 1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (来自 HuggingFace), 伴随论文 [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108) 由 Victor Sanh, Lysandre Debut and Thomas Wolf 发布。 同样的方法也应用于压缩 GPT-2 到 [DistilGPT2](https://github.com/huggingface/transformers/tree/main/examples/distillation), RoBERTa 到 [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/distillation), Multilingual BERT 到 [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/distillation) 和德语版 DistilBERT。
 1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (来自 Microsoft Research) 伴随论文 [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) 由 Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei 发布。
-1. **[Donut](https://huggingface.co/docs/transformers/main/model_doc/donut)** (来自 NAVER) 伴随论文 [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) 由 Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park 发布。
+1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (来自 NAVER) 伴随论文 [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) 由 Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park 发布。
 1. **[DPR](https://huggingface.co/docs/transformers/model_doc/dpr)** (来自 Facebook) 伴随论文 [Dense Passage Retrieval for Open-Domain Question Answering](https://arxiv.org/abs/2004.04906) 由 Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih 发布。
 1. **[DPT](https://huggingface.co/docs/transformers/master/model_doc/dpt)** (来自 Intel Labs) 伴随论文 [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) 由 René Ranftl, Alexey Bochkovskiy, Vladlen Koltun 发布。
 1. **[ELECTRA](https://huggingface.co/docs/transformers/model_doc/electra)** (来自 Google Research/Stanford University) 伴随论文 [ELECTRA: Pre-training text encoders as discriminators rather than generators](https://arxiv.org/abs/2003.10555) 由 Kevin Clark, Minh-Thang Luong, Quoc V. Le, Christopher D. Manning 发布。
 1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (来自 Google Research) 伴随论文 [Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) 由 Sascha Rothe, Shashi Narayan, Aliaksei Severyn 发布。
-1. **[ERNIE](https://huggingface.co/docs/transformers/main/model_doc/ernie)** (来自 Baidu) 伴随论文 [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu 发布。
+1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (来自 Baidu) 伴随论文 [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu 发布。
 1. **[FlauBERT](https://huggingface.co/docs/transformers/model_doc/flaubert)** (来自 CNRS) 伴随论文 [FlauBERT: Unsupervised Language Model Pre-training for French](https://arxiv.org/abs/1912.05372) 由 Hang Le, Loïc Vial, Jibril Frej, Vincent Segonne, Maximin Coavoux, Benjamin Lecouteux, Alexandre Allauzen, Benoît Crabbé, Laurent Besacier, Didier Schwab 发布。
 1. **[FLAVA](https://huggingface.co/docs/transformers/model_doc/flava)** (来自 Facebook AI) 伴随论文 [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) 由 Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach, and Douwe Kiela 发布。
 1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (来自 Google Research) 伴随论文 [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) 由 James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon 发布。
@@ -316,7 +316,7 @@ conda install -c huggingface transformers
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (来自 Meta AI) 伴随论文 [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) 由 Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al 发布。
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (来自 Google AI) 伴随论文 [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) 由 Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby 发布。
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (来自 Google) 伴随论文 [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) 由 Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu 发布。
-1. **[PEGASUS-X](https://huggingface.co/docs/transformers/main/model_doc/pegasus_x)** (来自 Google) 伴随论文 [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) 由 Jason Phang, Yao Zhao, Peter J. Liu 发布。  
+1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (来自 Google) 伴随论文 [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) 由 Jason Phang, Yao Zhao, Peter J. Liu 发布。  
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (来自 Deepmind) 伴随论文 [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) 由 Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira 发布。
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (来自 VinAI Research) 伴随论文 [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) 由 Dat Quoc Nguyen and Anh Tuan Nguyen 发布。
 1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (来自 UCLA NLP) 伴随论文 [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) 由 Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang 发布。
@@ -339,7 +339,7 @@ conda install -c huggingface transformers
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (来自 Tel Aviv University) 伴随论文 [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) 由 Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy 发布。
 1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (来自 Berkeley) 伴随论文 [SqueezeBERT: What can computer vision teach NLP about efficient neural networks?](https://arxiv.org/abs/2006.11316) 由 Forrest N. Iandola, Albert E. Shaw, Ravi Krishna, and Kurt W. Keutzer 发布。
 1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (来自 Microsoft) 伴随论文 [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) 由 Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo 发布。
-1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/main/model_doc/swinv2)** (来自 Microsoft) 伴随论文 [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) 由 Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo 发布。
+1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (来自 Microsoft) 伴随论文 [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) 由 Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo 发布。
 1. **[T5](https://huggingface.co/docs/transformers/model_doc/t5)** (来自 Google AI) 伴随论文 [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/abs/1910.10683) 由 Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu 发布。
 1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (来自 Google AI) 伴随论文 [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) 由 Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu 发布。
 1. **[TAPAS](https://huggingface.co/docs/transformers/model_doc/tapas)** (来自 Google AI) 伴随论文 [TAPAS: Weakly Supervised Table Parsing via Pre-training](https://arxiv.org/abs/2004.02349) 由 Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Müller, Francesco Piccinno and Julian Martin Eisenschlos 发布。
@@ -351,7 +351,7 @@ conda install -c huggingface transformers
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (来自 Microsoft Research) 伴随论文 [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) 由 Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang 发布。
 1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (来自 Microsoft Research) 伴随论文 [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) 由 Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu 发布。
 1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (来自 Tsinghua University and Nankai University) 伴随论文 [Visual Attention Network](https://arxiv.org/pdf/2202.09741.pdf) 由 Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu 发布。
-1. **[VideoMAE](https://huggingface.co/docs/transformers/main/model_doc/videomae)** (来自 Multimedia Computing Group, Nanjing University) 伴随论文 [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) 由 Zhan Tong, Yibing Song, Jue Wang, Limin Wang 发布。
+1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (来自 Multimedia Computing Group, Nanjing University) 伴随论文 [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) 由 Zhan Tong, Yibing Song, Jue Wang, Limin Wang 发布。
 1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (来自 NAVER AI Lab/Kakao Enterprise/Kakao Brain) 伴随论文 [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) 由 Wonjae Kim, Bokyung Son, Ildoo Kim 发布。
 1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/model_doc/vit)** (来自 Google AI) 伴随论文 [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://arxiv.org/abs/2010.11929) 由 Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby 发布。
 1. **[VisualBERT](https://huggingface.co/docs/transformers/model_doc/visual_bert)** (来自 UCLA NLP) 伴随论文 [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) 由 Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang 发布。
@@ -360,7 +360,7 @@ conda install -c huggingface transformers
 1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (来自 Facebook AI) 伴随论文 [FAIRSEQ S2T: Fast Speech-to-Text Modeling with FAIRSEQ](https://arxiv.org/abs/2010.05171) 由 Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Sravya Popuri, Dmytro Okhonko, Juan Pino 发布。
 1. **[Wav2Vec2Phoneme](https://huggingface.co/docs/transformers/model_doc/wav2vec2_phoneme)** (来自 Facebook AI) 伴随论文 [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) 由 Qiantong Xu, Alexei Baevski, Michael Auli 发布。
 1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
-1. **[X-CLIP](https://huggingface.co/docs/transformers/main/model_doc/xclip)** (来自 Microsoft Research) 伴随论文 [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) 由 Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling 发布。
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (来自 Microsoft Research) 伴随论文 [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) 由 Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling 发布。
 1. **[XGLM](https://huggingface.co/docs/transformers/model_doc/xglm)** (From Facebook AI) released with the paper [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) by Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li. 
 1. **[XLM](https://huggingface.co/docs/transformers/model_doc/xlm)** (来自 Facebook) 伴随论文 [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) 由 Guillaume Lample and Alexis Conneau 发布。
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (来自 Microsoft Research) 伴随论文 [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) 由 Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou 发布。

README_zh-hant.md

@@ -278,12 +278,12 @@ conda install -c huggingface transformers
 1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (from Microsoft Research) released with the paper [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://arxiv.org/abs/1911.00536) by Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan.
 1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (from HuggingFace), released together with the paper [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108) by Victor Sanh, Lysandre Debut and Thomas Wolf. The same method has been applied to compress GPT2 into [DistilGPT2](https://github.com/huggingface/transformers/tree/main/examples/distillation), RoBERTa into [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/distillation), Multilingual BERT into [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/distillation) and a German version of DistilBERT.
 1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (from Microsoft Research) released with the paper [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) by Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei.
-1. **[Donut](https://huggingface.co/docs/transformers/main/model_doc/donut)** (from NAVER) released with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
+1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (from NAVER) released with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
 1. **[DPR](https://huggingface.co/docs/transformers/model_doc/dpr)** (from Facebook) released with the paper [Dense Passage Retrieval for Open-Domain Question Answering](https://arxiv.org/abs/2004.04906) by Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih.
 1. **[DPT](https://huggingface.co/docs/transformers/master/model_doc/dpt)** (from Intel Labs) released with the paper [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) by René Ranftl, Alexey Bochkovskiy, Vladlen Koltun.
 1. **[ELECTRA](https://huggingface.co/docs/transformers/model_doc/electra)** (from Google Research/Stanford University) released with the paper [ELECTRA: Pre-training text encoders as discriminators rather than generators](https://arxiv.org/abs/2003.10555) by Kevin Clark, Minh-Thang Luong, Quoc V. Le, Christopher D. Manning.
 1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (from Google Research) released with the paper [Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) by Sascha Rothe, Shashi Narayan, Aliaksei Severyn.
-1. **[ERNIE](https://huggingface.co/docs/transformers/main/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
 1. **[FlauBERT](https://huggingface.co/docs/transformers/model_doc/flaubert)** (from CNRS) released with the paper [FlauBERT: Unsupervised Language Model Pre-training for French](https://arxiv.org/abs/1912.05372) by Hang Le, Loïc Vial, Jibril Frej, Vincent Segonne, Maximin Coavoux, Benjamin Lecouteux, Alexandre Allauzen, Benoît Crabbé, Laurent Besacier, Didier Schwab.
 1. **[FLAVA](https://huggingface.co/docs/transformers/model_doc/flava)** (from Facebook AI) released with the paper [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) by Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach, and Douwe Kiela.
 1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (from Google Research) released with the paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
@@ -328,7 +328,7 @@ conda install -c huggingface transformers
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
-1. **[PEGASUS-X](https://huggingface.co/docs/transformers/main/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, Peter J. Liu.
+1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, Peter J. Liu.
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.
 1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
@@ -351,7 +351,7 @@ conda install -c huggingface transformers
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (from Tel Aviv University) released with the paper [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) by Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy.
 1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (from Berkeley) released with the paper [SqueezeBERT: What can computer vision teach NLP about efficient neural networks?](https://arxiv.org/abs/2006.11316) by Forrest N. Iandola, Albert E. Shaw, Ravi Krishna, and Kurt W. Keutzer.
 1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (from Microsoft) released with the paper [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
-1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/main/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
 1. **[T5](https://huggingface.co/docs/transformers/model_doc/t5)** (from Google AI) released with the paper [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/abs/1910.10683) by Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu.
 1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (from Google AI) released with the paper [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) by Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu.
 1. **[TAPAS](https://huggingface.co/docs/transformers/model_doc/tapas)** (from Google AI) released with the paper [TAPAS: Weakly Supervised Table Parsing via Pre-training](https://arxiv.org/abs/2004.02349) by Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Müller, Francesco Piccinno and Julian Martin Eisenschlos.
@@ -363,7 +363,7 @@ conda install -c huggingface transformers
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (from Microsoft Research) released with the paper [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) by Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang.
 1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (from Microsoft Research) released with the paper [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) by Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu.
 1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (from Tsinghua University and Nankai University) released with the paper [Visual Attention Network](https://arxiv.org/pdf/2202.09741.pdf) by Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
-1. **[VideoMAE](https://huggingface.co/docs/transformers/main/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
+1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
 1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (from NAVER AI Lab/Kakao Enterprise/Kakao Brain) released with the paper [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) by Wonjae Kim, Bokyung Son, Ildoo Kim.
 1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/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. 
 1. **[VisualBERT](https://huggingface.co/docs/transformers/model_doc/visual_bert)** (from UCLA NLP) released with the paper [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) by Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang.
@@ -372,7 +372,7 @@ conda install -c huggingface transformers
 1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (from Facebook AI) released with the paper [FAIRSEQ S2T: Fast Speech-to-Text Modeling with FAIRSEQ](https://arxiv.org/abs/2010.05171) by Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Sravya Popuri, Dmytro Okhonko, Juan Pino.
 1. **[Wav2Vec2Phoneme](https://huggingface.co/docs/transformers/model_doc/wav2vec2_phoneme)** (from Facebook AI) released with the paper [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) by Qiantong Xu, Alexei Baevski, Michael Auli.
 1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
-1. **[X-CLIP](https://huggingface.co/docs/transformers/main/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
 1. **[XGLM](https://huggingface.co/docs/transformers/model_doc/xglm)** (From Facebook AI) released with the paper [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) by Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li. 
 1. **[XLM](https://huggingface.co/docs/transformers/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.
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (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.

examples/flax/question-answering/run_qa.py

@@ -61,7 +61,7 @@ from utils_qa import postprocess_qa_predictions
 logger = logging.getLogger(__name__)
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 Array = Any
 Dataset = datasets.arrow_dataset.Dataset

examples/flax/text-classification/run_flax_glue.py

@@ -54,7 +54,7 @@ from transformers.utils import check_min_version, get_full_repo_name, send_examp
 
 logger = logging.getLogger(__name__)
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 Array = Any
 Dataset = datasets.arrow_dataset.Dataset

examples/flax/token-classification/run_flax_ner.py

@@ -55,7 +55,7 @@ from transformers.utils.versions import require_version
 
 logger = logging.getLogger(__name__)
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/token-classification/requirements.txt")
 

examples/pytorch/audio-classification/run_audio_classification.py

@@ -45,7 +45,7 @@ from transformers.utils.versions import require_version
 logger = logging.getLogger(__name__)
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.14.0", "To fix: pip install -r examples/pytorch/audio-classification/requirements.txt")
 

examples/pytorch/contrastive-image-text/run_clip.py

@@ -54,7 +54,7 @@ from transformers.utils.versions import require_version
 logger = logging.getLogger(__name__)
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/contrastive-image-text/requirements.txt")
 

examples/pytorch/image-classification/run_image_classification.py

@@ -55,7 +55,7 @@ from transformers.utils.versions import require_version
 logger = logging.getLogger(__name__)
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/image-classification/requirements.txt")
 

examples/pytorch/image-classification/run_image_classification_no_trainer.py

@@ -53,7 +53,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = get_logger(__name__)
 

examples/pytorch/image-pretraining/run_mae.py

@@ -43,7 +43,7 @@ from transformers.utils.versions import require_version
 logger = logging.getLogger(__name__)
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/image-pretraining/requirements.txt")
 

examples/pytorch/image-pretraining/run_mim.py

@@ -48,7 +48,7 @@ Any model supported by the AutoModelForMaskedImageModeling API can be used.
 logger = logging.getLogger(__name__)
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/image-pretraining/requirements.txt")
 

examples/pytorch/language-modeling/run_clm.py

@@ -54,7 +54,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/language-modeling/requirements.txt")
 

examples/pytorch/language-modeling/run_clm_no_trainer.py

@@ -57,7 +57,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = get_logger(__name__)
 

examples/pytorch/language-modeling/run_mlm.py

@@ -53,7 +53,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/language-modeling/requirements.txt")
 

examples/pytorch/language-modeling/run_mlm_no_trainer.py

@@ -57,7 +57,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = get_logger(__name__)
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/language-modeling/requirements.txt")

examples/pytorch/language-modeling/run_plm.py

@@ -47,7 +47,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/language-modeling/requirements.txt")
 

examples/pytorch/multiple-choice/run_swag.py

@@ -47,7 +47,7 @@ from transformers.utils import PaddingStrategy, check_min_version, send_example_
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = logging.getLogger(__name__)
 

examples/pytorch/multiple-choice/run_swag_no_trainer.py

@@ -56,7 +56,7 @@ from transformers.utils import PaddingStrategy, check_min_version, get_full_repo
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = get_logger(__name__)
 # You should update this to your particular problem to have better documentation of `model_type`

examples/pytorch/question-answering/run_qa.py

@@ -49,7 +49,7 @@ from utils_qa import postprocess_qa_predictions
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/question-answering/requirements.txt")
 

examples/pytorch/question-answering/run_qa_beam_search.py

@@ -48,7 +48,7 @@ from utils_qa import postprocess_qa_predictions_with_beam_search
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/question-answering/requirements.txt")
 

examples/pytorch/question-answering/run_qa_beam_search_no_trainer.py

@@ -56,7 +56,7 @@ from utils_qa import postprocess_qa_predictions_with_beam_search
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/question-answering/requirements.txt")
 

examples/pytorch/question-answering/run_qa_no_trainer.py

@@ -57,7 +57,7 @@ from utils_qa import postprocess_qa_predictions
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/question-answering/requirements.txt")
 

examples/pytorch/question-answering/run_seq2seq_qa.py

@@ -45,7 +45,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/question-answering/requirements.txt")
 

examples/pytorch/semantic-segmentation/run_semantic_segmentation.py

@@ -51,7 +51,7 @@ from transformers.utils.versions import require_version
 logger = logging.getLogger(__name__)
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=2.0.0", "To fix: pip install -r examples/pytorch/semantic-segmentation/requirements.txt")
 

examples/pytorch/semantic-segmentation/run_semantic_segmentation_no_trainer.py

@@ -50,7 +50,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = get_logger(__name__)
 

examples/pytorch/speech-recognition/run_speech_recognition_ctc.py

@@ -50,7 +50,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.18.0", "To fix: pip install -r examples/pytorch/speech-recognition/requirements.txt")
 

examples/pytorch/speech-recognition/run_speech_recognition_seq2seq.py

@@ -48,7 +48,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.18.0", "To fix: pip install -r examples/pytorch/speech-recognition/requirements.txt")
 

examples/pytorch/summarization/run_summarization.py

@@ -52,7 +52,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/summarization/requirements.txt")
 

examples/pytorch/summarization/run_summarization_no_trainer.py

@@ -56,7 +56,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = get_logger(__name__)
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/summarization/requirements.txt")

examples/pytorch/text-classification/run_glue.py

@@ -48,7 +48,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/text-classification/requirements.txt")
 

examples/pytorch/text-classification/run_glue_no_trainer.py

@@ -48,7 +48,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = get_logger(__name__)
 

examples/pytorch/text-classification/run_xnli.py

@@ -48,7 +48,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/text-classification/requirements.txt")
 

examples/pytorch/token-classification/run_ner.py

@@ -49,7 +49,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/token-classification/requirements.txt")
 

examples/pytorch/token-classification/run_ner_no_trainer.py

@@ -55,7 +55,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = get_logger(__name__)
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/token-classification/requirements.txt")

examples/pytorch/translation/run_translation.py

@@ -52,7 +52,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/translation/requirements.txt")
 

examples/pytorch/translation/run_translation_no_trainer.py

@@ -57,7 +57,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = get_logger(__name__)
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/translation/requirements.txt")

examples/tensorflow/multiple-choice/run_swag.py

@@ -50,7 +50,7 @@ from transformers.utils import PaddingStrategy, check_min_version, send_example_
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = logging.getLogger(__name__)
 

examples/tensorflow/question-answering/run_qa.py

@@ -48,7 +48,7 @@ from utils_qa import postprocess_qa_predictions
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 logger = logging.getLogger(__name__)
 

examples/tensorflow/summarization/run_summarization.py

@@ -53,7 +53,7 @@ from transformers.utils.versions import require_version
 
 # region Checking dependencies
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/summarization/requirements.txt")
 

examples/tensorflow/text-classification/run_glue.py

@@ -47,7 +47,7 @@ from transformers.utils import check_min_version, send_example_telemetry
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 task_to_keys = {
     "cola": ("sentence", None),

examples/tensorflow/translation/run_translation.py

@@ -56,7 +56,7 @@ from transformers.utils.versions import require_version
 
 # region Dependencies and constants
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.22.0.dev0")
+check_min_version("4.22.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/summarization/requirements.txt")
 

setup.py

@@ -400,7 +400,7 @@ install_requires = [
 
 setup(
     name="transformers",
-    version="4.22.0.dev0",  # expected format is one of x.y.z.dev0, or x.y.z.rc1 or x.y.z (no to dashes, yes to dots)
+    version="4.22.1",  # expected format is one of x.y.z.dev0, or x.y.z.rc1 or x.y.z (no to dashes, yes to dots)
     author="The Hugging Face team (past and future) with the help of all our contributors (https://github.com/huggingface/transformers/graphs/contributors)",
     author_email="transformers@huggingface.co",
     description="State-of-the-art Machine Learning for JAX, PyTorch and TensorFlow",

src/transformers/__init__.py

@@ -22,7 +22,7 @@
 # to defer the actual importing for when the objects are requested. This way `import transformers` provides the names
 # in the namespace without actually importing anything (and especially none of the backends).
 
-__version__ = "4.22.0.dev0"
+__version__ = "4.22.1"
 
 from typing import TYPE_CHECKING
 

src/transformers/configuration_utils.py

@@ -32,7 +32,9 @@ from .utils import (
     PushToHubMixin,
     cached_file,
     copy_func,
+    download_url,
     extract_commit_hash,
+    is_remote_url,
     is_torch_available,
     logging,
 )
@@ -592,9 +594,12 @@ class PretrainedConfig(PushToHubMixin):
 
         is_local = os.path.isdir(pretrained_model_name_or_path)
         if os.path.isfile(os.path.join(subfolder, pretrained_model_name_or_path)):
-            # Soecial case when pretrained_model_name_or_path is a local file
+            # Special case when pretrained_model_name_or_path is a local file
             resolved_config_file = pretrained_model_name_or_path
             is_local = True
+        elif is_remote_url(pretrained_model_name_or_path):
+            configuration_file = pretrained_model_name_or_path
+            resolved_config_file = download_url(pretrained_model_name_or_path)
         else:
             configuration_file = kwargs.pop("_configuration_file", CONFIG_NAME)
 

src/transformers/convert_slow_tokenizer.py

@@ -282,8 +282,20 @@ class GPT2Converter(Converter):
 
         tokenizer.pre_tokenizer = pre_tokenizers.ByteLevel(add_prefix_space=self.original_tokenizer.add_prefix_space)
         tokenizer.decoder = decoders.ByteLevel()
+        if self.original_tokenizer.add_bos_token:
+            bos = self.original_tokenizer.bos_token
+            bos_token_id = self.original_tokenizer.bos_token_id
+            tokenizer.post_processor = processors.TemplateProcessing(
+                single=f"{bos}:0 $A:0",  # token_type_id is 2 for Funnel transformer
+                pair=f"{bos}:0 $A:0 $B:1",
+                special_tokens=[
+                    (bos, bos_token_id),
+                ],
+            )
+        else:
+            # XXX trim_offsets=False actually means this post_processor doesn't
+            # really do anything.
             tokenizer.post_processor = processors.ByteLevel(trim_offsets=False)
-
         return tokenizer
 
 

src/transformers/feature_extraction_utils.py

@@ -31,8 +31,10 @@ from .utils import (
     TensorType,
     cached_file,
     copy_func,
+    download_url,
     is_flax_available,
     is_offline_mode,
+    is_remote_url,
     is_tf_available,
     is_torch_available,
     logging,
@@ -386,6 +388,9 @@ class FeatureExtractionMixin(PushToHubMixin):
         if os.path.isfile(pretrained_model_name_or_path):
             resolved_feature_extractor_file = pretrained_model_name_or_path
             is_local = True
+        elif is_remote_url(pretrained_model_name_or_path):
+            feature_extractor_file = pretrained_model_name_or_path
+            resolved_feature_extractor_file = download_url(pretrained_model_name_or_path)
         else:
             feature_extractor_file = FEATURE_EXTRACTOR_NAME
             try:

src/transformers/modeling_flax_utils.py

@@ -47,8 +47,10 @@ from .utils import (
     add_start_docstrings_to_model_forward,
     cached_file,
     copy_func,
+    download_url,
     has_file,
     is_offline_mode,
+    is_remote_url,
     logging,
     replace_return_docstrings,
 )
@@ -677,6 +679,9 @@ class FlaxPreTrainedModel(PushToHubMixin, FlaxGenerationMixin):
             elif os.path.isfile(pretrained_model_name_or_path):
                 archive_file = pretrained_model_name_or_path
                 is_local = True
+            elif is_remote_url(pretrained_model_name_or_path):
+                filename = pretrained_model_name_or_path
+                resolved_archive_file = download_url(pretrained_model_name_or_path)
             else:
                 filename = WEIGHTS_NAME if from_pt else FLAX_WEIGHTS_NAME
                 try:

src/transformers/modeling_tf_utils.py

@@ -54,9 +54,11 @@ from .utils import (
     ModelOutput,
     PushToHubMixin,
     cached_file,
+    download_url,
     find_labels,
     has_file,
     is_offline_mode,
+    is_remote_url,
     logging,
     requires_backends,
     working_or_temp_dir,
@@ -2345,6 +2347,9 @@ class TFPreTrainedModel(tf.keras.Model, TFModelUtilsMixin, TFGenerationMixin, Pu
             elif os.path.isfile(pretrained_model_name_or_path + ".index"):
                 archive_file = pretrained_model_name_or_path + ".index"
                 is_local = True
+            elif is_remote_url(pretrained_model_name_or_path):
+                filename = pretrained_model_name_or_path
+                resolved_archive_file = download_url(pretrained_model_name_or_path)
             else:
                 # set correct filename
                 filename = WEIGHTS_NAME if from_pt else TF2_WEIGHTS_NAME

src/transformers/modeling_utils.py

@@ -59,10 +59,12 @@ from .utils import (
     PushToHubMixin,
     cached_file,
     copy_func,
+    download_url,
     has_file,
     is_accelerate_available,
     is_bitsandbytes_available,
     is_offline_mode,
+    is_remote_url,
     logging,
     replace_return_docstrings,
 )
@@ -1998,6 +2000,9 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
                     )
                 archive_file = os.path.join(subfolder, pretrained_model_name_or_path + ".index")
                 is_local = True
+            elif is_remote_url(pretrained_model_name_or_path):
+                filename = pretrained_model_name_or_path
+                resolved_archive_file = download_url(pretrained_model_name_or_path)
             else:
                 # set correct filename
                 if from_tf:

src/transformers/models/auto/tokenization_auto.py

@@ -271,7 +271,13 @@ else:
                     "XLMRobertaTokenizerFast" if is_tokenizers_available() else None,
                 ),
             ),
-            ("xlm-roberta-xl", ("RobertaTokenizer", "RobertaTokenizerFast" if is_tokenizers_available() else None)),
+            (
+                "xlm-roberta-xl",
+                (
+                    "XLMRobertaTokenizer" if is_sentencepiece_available() else None,
+                    "XLMRobertaTokenizerFast" if is_tokenizers_available() else None,
+                ),
+            ),
             (
                 "xlnet",
                 (

src/transformers/models/gpt2/tokenization_gpt2.py

@@ -261,6 +261,38 @@ class GPT2Tokenizer(PreTrainedTokenizer):
 
         return output + bos_token_ids + token_ids_1
 
+    def get_special_tokens_mask(
+        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
+    ) -> List[int]:
+        """
+        Retrieves sequence ids from a token list that has no special tokens added. This method is called when adding
+        special tokens using the tokenizer `prepare_for_model` or `encode_plus` methods.
+
+        Args:
+            token_ids_0 (`List[int]`):
+                List of IDs.
+            token_ids_1 (`List[int]`, *optional*):
+                Optional second list of IDs for sequence pairs.
+            already_has_special_tokens (`bool`, *optional*, defaults to `False`):
+                Whether or not the token list is already formatted with special tokens for the model.
+
+        Returns:
+            `List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
+        """
+        if already_has_special_tokens:
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
+
+        if not self.add_bos_token:
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=False
+            )
+
+        if token_ids_1 is None:
+            return [1] + ([0] * len(token_ids_0))
+        return [1] + ([0] * len(token_ids_0)) + [1] + ([0] * len(token_ids_1))
+
     def _tokenize(self, text):
         """Tokenize a string."""
         bpe_tokens = []

src/transformers/models/gpt2/tokenization_gpt2_fast.py

@@ -146,16 +146,7 @@ class GPT2TokenizerFast(PreTrainedTokenizerFast):
             **kwargs,
         )
 
-        if kwargs.pop("add_bos_token", False):
-            model_id = kwargs.pop("name_or_path", "")
-            raise ValueError(
-                "Currenty GPT2's fast tokenizer does NOT support adding a BOS token."
-                "Instead you should use GPT2's slow tokenizer class `GPT2Tokenizer` as follows: \n"
-                f"`GPT2Tokenizer.from_pretrained('{model_id}')`\nor\n"
-                f"`AutoTokenizer.from_pretrained('{model_id}', use_fast=False)`\n"
-                "This issue will be fixed soon, see: https://github.com/huggingface/tokenizers/pull/1005."
-                " so that the fast tokenizer works correctly."
-            )
+        self.add_bos_token = kwargs.pop("add_bos_token", False)
 
         pre_tok_state = json.loads(self.backend_tokenizer.pre_tokenizer.__getstate__())
         if pre_tok_state.get("add_prefix_space", add_prefix_space) != add_prefix_space:

src/transformers/models/maskformer/feature_extraction_maskformer.py

@@ -69,11 +69,12 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
             The sequence of standard deviations for each channel, to be used when normalizing images. Defaults to the
             ImageNet std.
         ignore_index (`int`, *optional*):
-            Value of the index (label) to be removed from the segmentation maps.
+            Label to be assigned to background pixels in segmentation maps. If provided, segmentation map pixels
+            denoted with 0 (background) will be replaced with `ignore_index`.
         reduce_labels (`bool`, *optional*, defaults to `False`):
-            Whether or not to reduce all label values of segmentation maps by 1. Usually used for datasets where 0 is
-            used for background, and background itself is not included in all classes of a dataset (e.g. ADE20k). The
-            background label will be replaced by `ignore_index`.
+            Whether or not to decrement all label values of segmentation maps by 1. Usually used for datasets where 0
+            is used for background, and background itself is not included in all classes of a dataset (e.g. ADE20k).
+            The background label will be replaced by `ignore_index`.
 
     """
 
@@ -162,7 +163,7 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
         images: ImageInput,
         segmentation_maps: ImageInput = None,
         pad_and_return_pixel_mask: Optional[bool] = True,
-        instance_id_to_semantic_id: Optional[Dict[int, int]] = None,
+        instance_id_to_semantic_id: Optional[Union[List[Dict[int, int]], Dict[int, int]]] = None,
         return_tensors: Optional[Union[str, TensorType]] = None,
         **kwargs,
     ) -> BatchFeature:
@@ -191,7 +192,9 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
                 number of channels, H and W are image height and width.
 
             segmentation_maps (`PIL.Image.Image`, `np.ndarray`, `torch.Tensor`, `List[PIL.Image.Image]`, `List[np.ndarray]`, `List[torch.Tensor]`, *optional*):
-                Optionally, the corresponding semantic segmentation maps with the pixel-wise annotations.
+                The corresponding semantic segmentation maps with the pixel-wise class id annotations or instance
+                segmentation maps with pixel-wise instance id annotations. Assumed to be semantic segmentation maps if
+                no `instance_id_to_semantic_id map` is provided.
 
             pad_and_return_pixel_mask (`bool`, *optional*, defaults to `True`):
                 Whether or not to pad images up to the largest image in a batch and create a pixel mask.
@@ -201,10 +204,11 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
                 - 1 for pixels that are real (i.e. **not masked**),
                 - 0 for pixels that are padding (i.e. **masked**).
 
-            instance_id_to_semantic_id (`Dict[int, int]`, *optional*):
-                If passed, we treat `segmentation_maps` as an instance segmentation map where each pixel represents an
-                instance id. To convert it to a binary mask of shape (`batch, num_labels, height, width`) we need a
-                dictionary mapping instance ids to label ids to create a semantic segmentation map.
+            instance_id_to_semantic_id (`List[Dict[int, int]]` or `Dict[int, int]`, *optional*):
+                A mapping between object instance ids and class ids. If passed, `segmentation_maps` is treated as an
+                instance segmentation map where each pixel represents an instance id. Can be provided as a single
+                dictionary with a global / dataset-level mapping or as a list of dictionaries (one per image), to map
+                instance ids in each image separately.
 
             return_tensors (`str` or [`~file_utils.TensorType`], *optional*):
                 If set, will return tensors instead of NumPy arrays. If set to `'pt'`, return PyTorch `torch.Tensor`
@@ -215,11 +219,11 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
 
             - **pixel_values** -- Pixel values to be fed to a model.
             - **pixel_mask** -- Pixel mask to be fed to a model (when `pad_and_return_pixel_mask=True` or if
-              *"pixel_mask"* is in `self.model_input_names`).
-            - **mask_labels** -- Optional list of mask labels of shape `(labels, height, width)` to be fed to a model
-              (when `annotations` are provided).
-            - **class_labels** -- Optional list of class labels of shape `(labels)` to be fed to a model (when
-              `annotations` are provided). They identify the labels of `mask_labels`, e.g. the label of
+              `pixel_mask` is in `self.model_input_names`).
+            - **mask_labels** -- Optional list of mask labels of shape `(num_class_labels, height, width)` to be fed to
+              a model (when `annotations` are provided).
+            - **class_labels** -- Optional list of class labels of shape `(num_class_labels)` to be fed to a model
+              (when `annotations` are provided). They identify the labels of `mask_labels`, e.g. the label of
               `mask_labels[i][j]` if `class_labels[i][j]`.
         """
         # Input type checking for clearer error
@@ -319,26 +323,31 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
         segmentation_map: "np.ndarray",
         instance_id_to_semantic_id: Optional[Dict[int, int]] = None,
     ):
-        if self.reduce_labels:
-            if self.ignore_index is None:
-                raise ValueError("`ignore_index` must be set when `reduce_labels` is `True`.")
-            segmentation_map[segmentation_map == 0] = self.ignore_index
-            # instances ids start from 1!
-            segmentation_map -= 1
-            segmentation_map[segmentation_map == self.ignore_index - 1] = self.ignore_index
+        # Get unique ids (class or instance ids based on input)
+        all_labels = np.unique(segmentation_map)
 
+        # Drop background label if applicable
+        if self.reduce_labels:
+            all_labels = all_labels[all_labels != 0]
+
+        # Generate a binary mask for each object instance
+        binary_masks = [np.ma.masked_where(segmentation_map == i, segmentation_map) for i in all_labels]
+        binary_masks = np.stack(binary_masks, axis=0)  # (num_labels, height, width)
+
+        # Convert instance ids to class ids
         if instance_id_to_semantic_id is not None:
-            # segmentation_map will be treated as an instance segmentation map where each pixel is a instance id
-            # thus it has to be converted to a semantic segmentation map
-            for instance_id, label_id in instance_id_to_semantic_id.items():
-                segmentation_map[segmentation_map == instance_id] = label_id
-        # get all the labels in the image
-        labels = np.unique(segmentation_map)
-        # remove ignore index (if we have one)
-        if self.ignore_index is not None:
-            labels = labels[labels != self.ignore_index]
-        # helping broadcast by making mask [1,W,H] and labels [C, 1, 1]
-        binary_masks = segmentation_map[None] == labels[:, None, None]
+            labels = np.zeros(all_labels.shape[0])
+
+            for label in all_labels:
+                class_id = instance_id_to_semantic_id[label]
+                labels[all_labels == label] = class_id
+        else:
+            labels = all_labels
+
+        # Decrement labels by 1
+        if self.reduce_labels:
+            labels -= 1
+
         return binary_masks.astype(np.float32), labels.astype(np.int64)
 
     def encode_inputs(
@@ -346,7 +355,7 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
         pixel_values_list: List["np.ndarray"],
         segmentation_maps: ImageInput = None,
         pad_and_return_pixel_mask: bool = True,
-        instance_id_to_semantic_id: Optional[Dict[int, int]] = None,
+        instance_id_to_semantic_id: Optional[Union[List[Dict[int, int]], Dict[int, int]]] = None,
         return_tensors: Optional[Union[str, TensorType]] = None,
     ):
         """
@@ -374,10 +383,11 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
                 - 1 for pixels that are real (i.e. **not masked**),
                 - 0 for pixels that are padding (i.e. **masked**).
 
-            instance_id_to_semantic_id (`Dict[int, int]`, *optional*):
-                If passed, we treat `segmentation_maps` as an instance segmentation map where each pixel represents an
-                instance id. To convert it to a binary mask of shape (`batch, num_labels, height, width`) we need a
-                dictionary mapping instance ids to label ids to create a semantic segmentation map.
+            instance_id_to_semantic_id (`List[Dict[int, int]]` or `Dict[int, int]`, *optional*):
+                A mapping between object instance ids and class ids. If passed, `segmentation_maps` is treated as an
+                instance segmentation map where each pixel represents an instance id. Can be provided as a single
+                dictionary with a global/dataset-level mapping or as a list of dictionaries (one per image), to map
+                instance ids in each image separately.
 
             return_tensors (`str` or [`~file_utils.TensorType`], *optional*):
                 If set, will return tensors instead of NumPy arrays. If set to `'pt'`, return PyTorch `torch.Tensor`
@@ -388,7 +398,7 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
 
             - **pixel_values** -- Pixel values to be fed to a model.
             - **pixel_mask** -- Pixel mask to be fed to a model (when `pad_and_return_pixel_mask=True` or if
-              *"pixel_mask"* is in `self.model_input_names`).
+              `pixel_mask` is in `self.model_input_names`).
             - **mask_labels** -- Optional list of mask labels of shape `(labels, height, width)` to be fed to a model
               (when `annotations` are provided).
             - **class_labels** -- Optional list of class labels of shape `(labels)` to be fed to a model (when
@@ -402,7 +412,14 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
         if segmentation_maps is not None:
             segmentation_maps = map(np.array, segmentation_maps)
             converted_segmentation_maps = []
-            for segmentation_map in segmentation_maps:
+
+            for i, segmentation_map in enumerate(segmentation_maps):
+                # Use instance2class_id mapping per image
+                if isinstance(instance_id_to_semantic_id, List):
+                    converted_segmentation_map = self.convert_segmentation_map_to_binary_masks(
+                        segmentation_map, instance_id_to_semantic_id[i]
+                    )
+                else:
                     converted_segmentation_map = self.convert_segmentation_map_to_binary_masks(
                         segmentation_map, instance_id_to_semantic_id
                     )
@@ -469,7 +486,7 @@ class MaskFormerFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
 
         Returns:
             `torch.Tensor`:
-                A tensor of shape (`batch_size, num_labels, height, width`).
+                A tensor of shape (`batch_size, num_class_labels, height, width`).
         """
         # class_queries_logits has shape [BATCH, QUERIES, CLASSES + 1]
         class_queries_logits = outputs.class_queries_logits

src/transformers/models/xlm_roberta_xl/modeling_xlm_roberta_xl.py

@@ -54,7 +54,7 @@ logger = logging.get_logger(__name__)
 
 _CHECKPOINT_FOR_DOC = "xlm-roberta-xlarge"
 _CONFIG_FOR_DOC = "XLMRobertaXLConfig"
-_TOKENIZER_FOR_DOC = "RobertaTokenizer"
+_TOKENIZER_FOR_DOC = "XLMRobertaTokenizer"
 
 XLM_ROBERTA_XL_PRETRAINED_MODEL_ARCHIVE_LIST = [
     "facebook/xlm-roberta-xl",

src/transformers/tokenization_utils_base.py

@@ -42,9 +42,11 @@ from .utils import (
     add_end_docstrings,
     cached_file,
     copy_func,
+    download_url,
     extract_commit_hash,
     is_flax_available,
     is_offline_mode,
+    is_remote_url,
     is_tf_available,
     is_tokenizers_available,
     is_torch_available,
@@ -1668,7 +1670,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin, PushToHubMixin):
         init_configuration = {}
 
         is_local = os.path.isdir(pretrained_model_name_or_path)
-        if os.path.isfile(pretrained_model_name_or_path):
+        if os.path.isfile(pretrained_model_name_or_path) or is_remote_url(pretrained_model_name_or_path):
             if len(cls.vocab_files_names) > 1:
                 raise ValueError(
                     f"Calling {cls.__name__}.from_pretrained() with the path to a single file or url is not "
@@ -1680,6 +1682,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin, PushToHubMixin):
                 FutureWarning,
             )
             file_id = list(cls.vocab_files_names.keys())[0]
+
             vocab_files[file_id] = pretrained_model_name_or_path
         else:
             # At this point pretrained_model_name_or_path is either a directory or a model identifier name
@@ -1723,6 +1726,10 @@ class PreTrainedTokenizerBase(SpecialTokensMixin, PushToHubMixin):
         for file_id, file_path in vocab_files.items():
             if file_path is None:
                 resolved_vocab_files[file_id] = None
+            elif os.path.isfile(file_path):
+                resolved_vocab_files[file_id] = file_path
+            elif is_remote_url(file_path):
+                resolved_vocab_files[file_id] = download_url(file_path, proxies=proxies)
             else:
                 resolved_vocab_files[file_id] = cached_file(
                     pretrained_model_name_or_path,

src/transformers/utils/__init__.py

@@ -63,6 +63,7 @@ from .hub import (
     cached_file,
     default_cache_path,
     define_sagemaker_information,
+    download_url,
     extract_commit_hash,
     get_cached_models,
     get_file_from_repo,
@@ -70,6 +71,7 @@ from .hub import (
     has_file,
     http_user_agent,
     is_offline_mode,
+    is_remote_url,
     move_cache,
     send_example_telemetry,
 )

src/transformers/utils/hub.py

@@ -19,10 +19,12 @@ import os
 import re
 import shutil
 import sys
+import tempfile
 import traceback
 import warnings
 from pathlib import Path
 from typing import Dict, List, Optional, Tuple, Union
+from urllib.parse import urlparse
 from uuid import uuid4
 
 import huggingface_hub
@@ -37,7 +39,7 @@ from huggingface_hub import (
     whoami,
 )
 from huggingface_hub.constants import HUGGINGFACE_HEADER_X_LINKED_ETAG, HUGGINGFACE_HEADER_X_REPO_COMMIT
-from huggingface_hub.file_download import REGEX_COMMIT_HASH
+from huggingface_hub.file_download import REGEX_COMMIT_HASH, http_get
 from huggingface_hub.utils import (
     EntryNotFoundError,
     LocalEntryNotFoundError,
@@ -124,6 +126,11 @@ HUGGINGFACE_CO_EXAMPLES_TELEMETRY = HUGGINGFACE_CO_RESOLVE_ENDPOINT + "/api/tele
 _CACHED_NO_EXIST = object()
 
 
+def is_remote_url(url_or_filename):
+    parsed = urlparse(url_or_filename)
+    return parsed.scheme in ("http", "https")
+
+
 def get_cached_models(cache_dir: Union[str, Path] = None) -> List[Tuple]:
     """
     Returns a list of tuples representing model binaries that are cached locally. Each tuple has shape `(model_url,
@@ -541,6 +548,32 @@ def get_file_from_repo(
     )
 
 
+def download_url(url, proxies=None):
+    """
+    Downloads a given url in a temporary file. This function is not safe to use in multiple processes. Its only use is
+    for deprecated behavior allowing to download config/models with a single url instead of using the Hub.
+
+    Args:
+        url (`str`): The url of the file to download.
+        proxies (`Dict[str, str]`, *optional*):
+            A dictionary of proxy servers to use by protocol or endpoint, e.g., `{'http': 'foo.bar:3128',
+            'http://hostname': 'foo.bar:4012'}.` The proxies are used on each request.
+
+    Returns:
+        `str`: The location of the temporary file where the url was downloaded.
+    """
+    warnings.warn(
+        f"Using `from_pretrained` with the url of a file (here {url}) is deprecated and won't be possible anymore in"
+        " v5 of Transformers. You should host your file on the Hub (hf.co) instead and use the repository ID. Note"
+        " that this is not compatible with the caching system (your file will be downloaded at each execution) or"
+        " multiple processes (each process will download the file in a different temporary file)."
+    )
+    tmp_file = tempfile.mktemp()
+    with open(tmp_file, "wb") as f:
+        http_get(url, f, proxies=proxies)
+    return tmp_file
+
+
 def has_file(
     path_or_repo: Union[str, os.PathLike],
     filename: str,
@@ -1095,9 +1128,9 @@ if cache_version < 1:
     except Exception as e:
         trace = "\n".join(traceback.format_tb(e.__traceback__))
         logger.error(
-            f"There was a problem when trying to move your cache:\n\n{trace}\n\nPlease file an issue at "
-            "https://github.com/huggingface/transformers/issues/new/choose and copy paste this whole message and we "
-            "will do our best to help."
+            f"There was a problem when trying to move your cache:\n\n{trace}\n{e.__class__.__name__}: {e}\n\nPlease "
+            "file an issue at https://github.com/huggingface/transformers/issues/new/choose and copy paste this whole "
+            "message and we will do our best to help."
         )
 
     try:

tests/models/gpt2/test_tokenization_gpt2.py

@@ -18,7 +18,7 @@ import json
 import os
 import unittest
 
-from transformers import GPT2Tokenizer, GPT2TokenizerFast
+from transformers import AutoTokenizer, GPT2Tokenizer, GPT2TokenizerFast
 from transformers.models.gpt2.tokenization_gpt2 import VOCAB_FILES_NAMES
 from transformers.testing_utils import require_tokenizers
 
@@ -250,3 +250,82 @@ class GPT2TokenizationTest(TokenizerTesterMixin, unittest.TestCase):
     # tokenizer has no padding token
     def test_padding_different_model_input_name(self):
         pass
+
+    def test_special_tokens_mask_input_pairs_and_bos_token(self):
+        # TODO: change to self.get_tokenizers() when the fast version is implemented
+        tokenizers = [self.get_tokenizer(do_lower_case=False, add_bos_token=True)]
+        for tokenizer in tokenizers:
+            with self.subTest(f"{tokenizer.__class__.__name__}"):
+                sequence_0 = "Encode this."
+                sequence_1 = "This one too please."
+                encoded_sequence = tokenizer.encode(sequence_0, add_special_tokens=False)
+                encoded_sequence += tokenizer.encode(sequence_1, add_special_tokens=False)
+                encoded_sequence_dict = tokenizer.encode_plus(
+                    sequence_0,
+                    sequence_1,
+                    add_special_tokens=True,
+                    return_special_tokens_mask=True,
+                )
+                encoded_sequence_w_special = encoded_sequence_dict["input_ids"]
+                special_tokens_mask = encoded_sequence_dict["special_tokens_mask"]
+                self.assertEqual(len(special_tokens_mask), len(encoded_sequence_w_special))
+
+                filtered_sequence = [
+                    (x if not special_tokens_mask[i] else None) for i, x in enumerate(encoded_sequence_w_special)
+                ]
+                filtered_sequence = [x for x in filtered_sequence if x is not None]
+                self.assertEqual(encoded_sequence, filtered_sequence)
+
+
+@require_tokenizers
+class OPTTokenizationTest(unittest.TestCase):
+    def test_serialize_deserialize_fast_opt(self):
+        # More context:
+        # https://huggingface.co/wjmcat/opt-350m-paddle/discussions/1
+        # https://huggingface.slack.com/archives/C01N44FJDHT/p1653511495183519
+        # https://github.com/huggingface/transformers/pull/17088#discussion_r871246439
+
+        tokenizer = AutoTokenizer.from_pretrained("facebook/opt-350m", from_slow=True)
+        text = "A photo of a cat"
+
+        tokens_ids = tokenizer.encode(
+            text,
+        )
+        self.assertEqual(tokens_ids, [2, 250, 1345, 9, 10, 4758])
+        tokenizer.save_pretrained("test_opt")
+
+        tokenizer = AutoTokenizer.from_pretrained("./test_opt")
+        tokens_ids = tokenizer.encode(
+            text,
+        )
+        self.assertEqual(tokens_ids, [2, 250, 1345, 9, 10, 4758])
+
+    def test_fast_slow_equivalence(self):
+        tokenizer = AutoTokenizer.from_pretrained("facebook/opt-350m", use_slow=True)
+        text = "A photo of a cat"
+
+        tokens_ids = tokenizer.encode(
+            text,
+        )
+        # Same as above
+        self.assertEqual(tokens_ids, [2, 250, 1345, 9, 10, 4758])
+
+    def test_users_can_modify_bos(self):
+        tokenizer = AutoTokenizer.from_pretrained("facebook/opt-350m", from_slow=True)
+
+        tokenizer.bos_token = "bos"
+        tokenizer.bos_token_id = tokenizer.get_vocab()["bos"]
+
+        text = "A photo of a cat"
+        tokens_ids = tokenizer.encode(
+            text,
+        )
+        # We changed the bos token
+        self.assertEqual(tokens_ids, [31957, 250, 1345, 9, 10, 4758])
+        tokenizer.save_pretrained("./tok")
+        tokenizer = AutoTokenizer.from_pretrained("./tok")
+        self.assertTrue(tokenizer.is_fast)
+        tokens_ids = tokenizer.encode(
+            text,
+        )
+        self.assertEqual(tokens_ids, [31957, 250, 1345, 9, 10, 4758])

tests/test_configuration_common.py

@@ -360,6 +360,12 @@ class ConfigTestUtils(unittest.TestCase):
             # This check we did call the fake head request
             mock_head.assert_called()
 
+    def test_legacy_load_from_url(self):
+        # This test is for deprecated behavior and can be removed in v5
+        _ = BertConfig.from_pretrained(
+            "https://huggingface.co/hf-internal-testing/tiny-random-bert/resolve/main/config.json"
+        )
+
 
 class ConfigurationVersioningTest(unittest.TestCase):
     def test_local_versioning(self):

tests/test_feature_extraction_common.py

@@ -182,6 +182,12 @@ class FeatureExtractorUtilTester(unittest.TestCase):
             # This check we did call the fake head request
             mock_head.assert_called()
 
+    def test_legacy_load_from_url(self):
+        # This test is for deprecated behavior and can be removed in v5
+        _ = Wav2Vec2FeatureExtractor.from_pretrained(
+            "https://huggingface.co/hf-internal-testing/tiny-random-wav2vec2/resolve/main/preprocessor_config.json"
+        )
+
 
 @is_staging_test
 class FeatureExtractorPushToHubTester(unittest.TestCase):

tests/test_modeling_common.py

@@ -33,6 +33,7 @@ import numpy as np
 
 import transformers
 from huggingface_hub import HfFolder, delete_repo, set_access_token
+from huggingface_hub.file_download import http_get
 from requests.exceptions import HTTPError
 from transformers import (
     AutoConfig,
@@ -2949,6 +2950,26 @@ class ModelUtilsTest(TestCasePlus):
             # This check we did call the fake head request
             mock_head.assert_called()
 
+    def test_load_from_one_file(self):
+        try:
+            tmp_file = tempfile.mktemp()
+            with open(tmp_file, "wb") as f:
+                http_get(
+                    "https://huggingface.co/hf-internal-testing/tiny-random-bert/resolve/main/pytorch_model.bin", f
+                )
+
+            config = BertConfig.from_pretrained("hf-internal-testing/tiny-random-bert")
+            _ = BertModel.from_pretrained(tmp_file, config=config)
+        finally:
+            os.remove(tmp_file)
+
+    def test_legacy_load_from_url(self):
+        # This test is for deprecated behavior and can be removed in v5
+        config = BertConfig.from_pretrained("hf-internal-testing/tiny-random-bert")
+        _ = BertModel.from_pretrained(
+            "https://huggingface.co/hf-internal-testing/tiny-random-bert/resolve/main/pytorch_model.bin", config=config
+        )
+
 
 @require_torch
 @is_staging_test

tests/test_modeling_tf_common.py

@@ -30,6 +30,7 @@ from typing import List, Tuple, get_type_hints
 from datasets import Dataset
 
 from huggingface_hub import HfFolder, Repository, delete_repo, set_access_token
+from huggingface_hub.file_download import http_get
 from requests.exceptions import HTTPError
 from transformers import is_tf_available, is_torch_available
 from transformers.configuration_utils import PretrainedConfig
@@ -1868,6 +1869,24 @@ class UtilsFunctionsTest(unittest.TestCase):
             # This check we did call the fake head request
             mock_head.assert_called()
 
+    def test_load_from_one_file(self):
+        try:
+            tmp_file = tempfile.mktemp()
+            with open(tmp_file, "wb") as f:
+                http_get("https://huggingface.co/hf-internal-testing/tiny-random-bert/resolve/main/tf_model.h5", f)
+
+            config = BertConfig.from_pretrained("hf-internal-testing/tiny-random-bert")
+            _ = TFBertModel.from_pretrained(tmp_file, config=config)
+        finally:
+            os.remove(tmp_file)
+
+    def test_legacy_load_from_url(self):
+        # This test is for deprecated behavior and can be removed in v5
+        config = BertConfig.from_pretrained("hf-internal-testing/tiny-random-bert")
+        _ = TFBertModel.from_pretrained(
+            "https://huggingface.co/hf-internal-testing/tiny-random-bert/resolve/main/tf_model.h5", config=config
+        )
+
     # tests whether the unpack_inputs function behaves as expected
     def test_unpack_inputs(self):
         class DummyModel:

tests/test_tokenization_common.py

@@ -31,6 +31,7 @@ from pathlib import Path
 from typing import TYPE_CHECKING, Any, Dict, List, Tuple, Union
 
 from huggingface_hub import HfFolder, delete_repo, set_access_token
+from huggingface_hub.file_download import http_get
 from parameterized import parameterized
 from requests.exceptions import HTTPError
 from transformers import (
@@ -3886,6 +3887,21 @@ class TokenizerUtilTester(unittest.TestCase):
             # This check we did call the fake head request
             mock_head.assert_called()
 
+    def test_legacy_load_from_one_file(self):
+        # This test is for deprecated behavior and can be removed in v5
+        try:
+            tmp_file = tempfile.mktemp()
+            with open(tmp_file, "wb") as f:
+                http_get("https://huggingface.co/albert-base-v1/resolve/main/spiece.model", f)
+
+            _ = AlbertTokenizer.from_pretrained(tmp_file)
+        finally:
+            os.remove(tmp_file)
+
+    def test_legacy_load_from_url(self):
+        # This test is for deprecated behavior and can be removed in v5
+        _ = AlbertTokenizer.from_pretrained("https://huggingface.co/albert-base-v1/resolve/main/spiece.model")
+
 
 @is_staging_test
 class TokenizerPushToHubTester(unittest.TestCase):