Release: v4.39.1

[SuperPoint] Fix doc example

.circleci/create_circleci_config.py

@@ -57,9 +57,9 @@ class CircleCIJob:
     install_steps: List[str] = None
     marker: Optional[str] = None
     parallelism: Optional[int] = 1
-    pytest_num_workers: int = 8
+    pytest_num_workers: int = 12
     pytest_options: Dict[str, Any] = None
-    resource_class: Optional[str] = "xlarge"
+    resource_class: Optional[str] = "2xlarge"
     tests_to_run: Optional[List[str]] = None
     working_directory: str = "~/transformers"
     # This should be only used for doctest job!
@@ -317,7 +317,7 @@ torch_job = CircleCIJob(
         "pip install -U --upgrade-strategy eager -e git+https://github.com/huggingface/accelerate@main#egg=accelerate",
     ],
     parallelism=1,
-    pytest_num_workers=6,
+    pytest_num_workers=12,
 )
 
 
@@ -353,7 +353,7 @@ pipelines_torch_job = CircleCIJob(
         "pip install -U --upgrade-strategy eager .[sklearn,torch,testing,sentencepiece,torch-speech,vision,timm,video]",
     ],
     marker="is_pipeline_test",
-    pytest_num_workers=6,
+    pytest_num_workers=12,
 )
 
 

README.md

@@ -331,7 +331,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
@@ -376,7 +376,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (from Microsoft Research) released with the paper [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) by Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (from ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. Released with the paper [blog post](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
@@ -409,6 +409,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (from The FAIR team of Meta AI) released with the paper [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) by Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.
 1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (from The FAIR team of Meta AI) released with the paper [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
@@ -416,7 +417,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (from Facebook) released with the paper [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Machine translation models trained using [OPUS](http://opus.nlpl.eu/) data by Jörg Tiedemann. The [Marian Framework](https://marian-nmt.github.io/) is being developed by the Microsoft Translator Team.
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (from FAIR and UIUC) released with the paper [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) by Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar.
@@ -442,7 +443,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (from the University of Wisconsin - Madison) released with the paper [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) by Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
@@ -469,7 +470,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi and Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/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.
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (from Nanjing University, The University of Hong Kong etc.) released with the paper [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (from the Qwen team, Alibaba Group) released with the paper [Qwen Technical Report](https://arxiv.org/abs/2309.16609) by Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu.
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
@@ -486,7 +487,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI)) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI)) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (from Meta AI) released with the paper [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) by Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -497,7 +498,8 @@ 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. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report)](https://stability.wandb.io/stability-llm/stable-lm/reports/StableLM-3B-4E1T--VmlldzoyMjU4?accessToken=u3zujipenkx5g7rtcj9qojjgxpconyjktjkli2po09nffrffdhhchq045vp0wyfo) by Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu.
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (from MBZUAI) released with the paper [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) by Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.
 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/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.
@@ -515,7 +517,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft), released together with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (from UNC Chapel Hill) released with the paper [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) by Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (from Intel) released with the paper [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) by Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (from Google Research) released with the paper [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) by Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.
 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.

README_de.md

@@ -327,7 +327,7 @@ Aktuelle Anzahl der Checkpoints: ![](https://img.shields.io/endpoint?url=https:/
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
@@ -372,7 +372,7 @@ Aktuelle Anzahl der Checkpoints: ![](https://img.shields.io/endpoint?url=https:/
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (from Microsoft Research) released with the paper [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) by Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (from ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. Released with the paper [blog post](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
@@ -405,6 +405,7 @@ Aktuelle Anzahl der Checkpoints: ![](https://img.shields.io/endpoint?url=https:/
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (from The FAIR team of Meta AI) released with the paper [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) by Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.
 1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (from The FAIR team of Meta AI) released with the paper [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
@@ -412,7 +413,7 @@ Aktuelle Anzahl der Checkpoints: ![](https://img.shields.io/endpoint?url=https:/
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (from Facebook) released with the paper [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Machine translation models trained using [OPUS](http://opus.nlpl.eu/) data by Jörg Tiedemann. The [Marian Framework](https://marian-nmt.github.io/) is being developed by the Microsoft Translator Team.
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (from FAIR and UIUC) released with the paper [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) by Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar.
@@ -438,7 +439,7 @@ Aktuelle Anzahl der Checkpoints: ![](https://img.shields.io/endpoint?url=https:/
 1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (from the University of Wisconsin - Madison) released with the paper [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) by Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
@@ -465,7 +466,7 @@ Aktuelle Anzahl der Checkpoints: ![](https://img.shields.io/endpoint?url=https:/
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi and Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/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.
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (from Nanjing University, The University of Hong Kong etc.) released with the paper [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (from the Qwen team, Alibaba Group) released with the paper [Qwen Technical Report](https://arxiv.org/abs/2309.16609) by Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu.
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
@@ -482,7 +483,7 @@ Aktuelle Anzahl der Checkpoints: ![](https://img.shields.io/endpoint?url=https:/
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (from Meta AI) released with the paper [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) by Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -493,7 +494,8 @@ Aktuelle Anzahl der Checkpoints: ![](https://img.shields.io/endpoint?url=https:/
 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. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report) by Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu.
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (from MBZUAI) released with the paper [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) by Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.
 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/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.
@@ -511,7 +513,7 @@ Aktuelle Anzahl der Checkpoints: ![](https://img.shields.io/endpoint?url=https:/
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft), released together with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (from UNC Chapel Hill) released with the paper [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) by Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (from Intel) released with the paper [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) by Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (from Google Research) released with the paper [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) by Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.
 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.

README_es.md

@@ -304,7 +304,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
@@ -349,7 +349,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (from Microsoft Research) released with the paper [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) by Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (from ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. Released with the paper [blog post](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
@@ -380,8 +380,9 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (from Meta AI) released with the paper [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) by Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze.
 1. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (from South China University of Technology) released with the paper [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) by Jiapeng Wang, Lianwen Jin, Kai Ding.
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (from The FAIR team of Meta AI) released with the paper [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) by Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.
-1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (from The FAIR team of Meta AI) released with the paper [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/XXX) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom..
+1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (from The FAIR team of Meta AI) released with the paper [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom..
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
@@ -389,7 +390,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (from Facebook) released with the paper [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Machine translation models trained using [OPUS](http://opus.nlpl.eu/) data by Jörg Tiedemann. The [Marian Framework](https://marian-nmt.github.io/) is being developed by the Microsoft Translator Team.
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (from FAIR and UIUC) released with the paper [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) by Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar.
@@ -412,10 +413,10 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[MobileViTV2](https://huggingface.co/docs/transformers/model_doc/mobilevitv2)** (from Apple) released with the paper [Separable Self-attention for Mobile Vision Transformers](https://arxiv.org/abs/2206.02680) by Sachin Mehta and Mohammad Rastegari.
 1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MPT](https://huggingface.co/docs/transformers/model_doc/mpt)** (from MosaiML) released with the repository [llm-foundry](https://github.com/mosaicml/llm-foundry/) by the MosaicML NLP Team.
-1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (from the University of Wisconsin - Madison) released with the paper [Multi Resolution Analysis (MRA)](https://arxiv.org/abs/2207.10284) by Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
+1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (from the University of Wisconsin - Madison) released with the paper [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) by Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
@@ -429,7 +430,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 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. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (from Google AI) released with the paper [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683) by Matthias Minderer, Alexey Gritsenko, Neil Houlsby.
 1. **[PatchTSMixer](https://huggingface.co/docs/transformers/model_doc/patchtsmixer)** (from  IBM Research) released with the paper [TSMixer: Lightweight MLP-Mixer Model for Multivariate Time Series Forecasting](https://arxiv.org/pdf/2306.09364.pdf) by Vijay Ekambaram, Arindam Jati, Nam Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
-1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (from IBM) released with the paper [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf) by Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
+1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (from IBM) released with the paper [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/abs/2211.14730) by Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
 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/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.
@@ -442,7 +443,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi, Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/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.
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (from Nanjing University, The University of Hong Kong etc.) released with the paper [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao. 
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao. 
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (from the Qwen team, Alibaba Group) released with the paper [Qwen Technical Report](https://arxiv.org/abs/2309.16609) by Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu.
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
@@ -459,7 +460,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (from Meta AI) released with the paper [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) by Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -470,7 +471,8 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 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. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report)](https://stability.wandb.io/stability-llm/stable-lm/reports/StableLM-3B-4E1T--VmlldzoyMjU4?accessToken=u3zujipenkx5g7rtcj9qojjgxpconyjktjkli2po09nffrffdhhchq045vp0wyfo) by  Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu. 
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich. 
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (from MBZUAI) released with the paper [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) by Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.
 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/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.
@@ -488,7 +490,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft), released together with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (from UNC Chapel Hill) released with the paper [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) by Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (from Intel) released with the paper [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) by Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (from Google Research) released with the paper [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) by Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.
 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.

README_fr.md

@@ -143,7 +143,7 @@ Pour utiliser immédiatement un modèle sur une entrée donnée (texte, image, a
 
 # Allouer un pipeline pour l'analyse de sentiment
 >>> classifieur = pipeline('sentiment-analysis')
->>> classifieur('Nous sommes très heureux d'introduire le pipeline dans le référentiel transformers.')
+>>> classifieur("Nous sommes très heureux d'introduire le pipeline dans le référentiel transformers.")
 [{'label': 'POSITIF', 'score': 0.9996980428695679}]
 ```
 
@@ -299,255 +299,257 @@ Nombre actuel de points de contrôle : ![](https://img.shields.io/endpoint?url=h
 1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (de Facebook) publié dans l'article [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) de Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov et Luke Zettlemoyer.
 1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (de l'École polytechnique) publié dans l'article [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) de Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis.
 1. **[BARTpho](https://huggingface.co/docs/transformers/model_doc/bartpho)** (de VinAI Research) publié dans l'article [BARTpho: Pre-trained Sequence-to-Sequence Models for Vietnamese](https://arxiv.org/abs/2109.09701) de Nguyen Luong Tran, Duong Minh Le et Dat Quoc Nguyen.
-1. **[BEiT](https://huggingface.co/docs/transformers/model_doc/beit)** (de Microsoft) publié dans l'article [BEiT: Pré-entraînement BERT des transformateurs d'images](https://arxiv.org/abs/2106.08254) par Hangbo Bao, Li Dong, Furu Wei.
-1. **[BERT](https://huggingface.co/docs/transformers/model_doc/bert)** (de Google) publié dans l'article [BERT : Pré-entraînement de transformateurs bidirectionnels profonds pour la compréhension du langage](https://arxiv.org/abs/1810.04805) par Jacob Devlin, Ming-Wei Chang, Kenton Lee et Kristina Toutanova.
+1. **[BEiT](https://huggingface.co/docs/transformers/model_doc/beit)** (de Microsoft) publié dans l'article [BEiT: BERT Pre-Training of Image Transformers](https://arxiv.org/abs/2106.08254) par Hangbo Bao, Li Dong, Furu Wei.
+1. **[BERT](https://huggingface.co/docs/transformers/model_doc/bert)** (de Google) publié dans l'article [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://arxiv.org/abs/1810.04805) par Jacob Devlin, Ming-Wei Chang, Kenton Lee et Kristina Toutanova.
 1. **[BERT For Sequence Generation](https://huggingface.co/docs/transformers/model_doc/bert-generation)** (de Google) publié dans l'article [Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) parSascha Rothe, Shashi Narayan, Aliaksei Severyn.
-1. **[BERTweet](https://huggingface.co/docs/transformers/model_doc/bertweet)** (de VinAI Research) publié dans l'article [BERTweet : un modèle de langage pré-entraîné pour les Tweets en anglais](https://aclanthology.org/2020.emnlp-demos.2/) par Dat Quoc Nguyen, Thanh Vu et Anh Tuan Nguyen.
-1. **[BigBird-Pegasus](https://huggingface.co/docs/transformers/model_doc/bigbird_pegasus)** (de Google Research) publié dans l'article [Big Bird: Transformateurs pour des séquences plus longues](https://arxiv.org/abs/2007.14062) par Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
-1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (de Google Research) publié dans l'article [Big Bird: Transformateurs pour des séquences plus longues](https://arxiv.org/abs/2007.14062) par Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
-1. **[BioGpt](https://huggingface.co/docs/transformers/model_doc/biogpt)** (de Microsoft Research AI4Science) publié dans l'article [BioGPT : transformateur génératif pré-entraîné pour la génération et l'extraction de texte biomédical](https://academic.oup.com/bib/advance-article/doi/10.1093/bib/bbac409/6713511?guestAccessKey=a66d9b5d-4f83-4017-bb52-405815c907b9) par Renqian Luo, Liai Sun, Yingce Xia, Tao Qin, Sheng Zhang, Hoifung Poon et Tie-Yan Liu.
-1. **[BiT](https://huggingface.co/docs/transformers/model_doc/bit)** (de Google AI) publié dans l'article [Big Transfer (BiT) : Apprentissage général de la représentation visuelle](https://arxiv.org/abs/1912.11370) par Alexander Kolesnikov, Lucas Beyer, Xiaohua Zhai, Joan Puigcerver, Jessica Yung, Sylvain Gelly, Neil Houlsby.
+1. **[BERTweet](https://huggingface.co/docs/transformers/model_doc/bertweet)** (de VinAI Research) publié dans l'article [BERTweet: A pre-trained language model for English Tweets](https://aclanthology.org/2020.emnlp-demos.2/) par Dat Quoc Nguyen, Thanh Vu et Anh Tuan Nguyen.
+1. **[BigBird-Pegasus](https://huggingface.co/docs/transformers/model_doc/bigbird_pegasus)** (de Google Research) publié dans l'article [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) par Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
+1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (de Google Research) publié dans l'article [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) par Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
+1. **[BioGpt](https://huggingface.co/docs/transformers/model_doc/biogpt)** (de Microsoft Research AI4Science) publié dans l'article [BioGPT: generative pre-trained transformer for biomedical text generation and mining](https://academic.oup.com/bib/advance-article/doi/10.1093/bib/bbac409/6713511?guestAccessKey=a66d9b5d-4f83-4017-bb52-405815c907b9) par Renqian Luo, Liai Sun, Yingce Xia, Tao Qin, Sheng Zhang, Hoifung Poon et Tie-Yan Liu.
+1. **[BiT](https://huggingface.co/docs/transformers/model_doc/bit)** (de Google AI) publié dans l'article [Big Transfer (BiT): General Visual Representation Learning](https://arxiv.org/abs/1912.11370) par Alexander Kolesnikov, Lucas Beyer, Xiaohua Zhai, Joan Puigcerver, Jessica Yung, Sylvain Gelly, Neil Houlsby.
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (de Facebook) publié dans l'article [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) par Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
 1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (de Facebook) publié dans l'article [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) par Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
-1. **[BLIP](https://huggingface.co/docs/transformers/model_doc/blip)** (de Salesforce) publié dans l'article [BLIP : Pré-entraînement de la langue et de l'image par bootstrap pour une compréhension et une génération unifiées de la vision et du langage](https://arxiv.org/abs/2201.12086) par Junnan Li, Dongxu Li, Caiming Xiong, Steven Hoi.
-1. **[BLIP-2](https://huggingface.co/docs/transformers/model_doc/blip-2)** (de Salesforce) publié dans l'article [BLIP-2 : Pré-entraînement de la langue et de l'image avec des encodeurs d'images gelés et de grands modèles de langage](https://arxiv.org/abs/2301.12597) par Junnan Li, Dongxu Li, Silvio Savarese, Steven Hoi.
+1. **[BLIP](https://huggingface.co/docs/transformers/model_doc/blip)** (de Salesforce) publié dans l'article [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://arxiv.org/abs/2201.12086) par Junnan Li, Dongxu Li, Caiming Xiong, Steven Hoi.
+1. **[BLIP-2](https://huggingface.co/docs/transformers/model_doc/blip-2)** (de Salesforce) publié dans l'article [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://arxiv.org/abs/2301.12597) par Junnan Li, Dongxu Li, Silvio Savarese, Steven Hoi.
 1. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (de l'atelier BigScience) publié par l'[atelier BigScience](https://bigscience.huggingface.co/).
 1. **[BORT](https://huggingface.co/docs/transformers/model_doc/bort)** (d'Alexa) publié dans l'article [Optimal Subarchitecture Extraction For BERT](https://arxiv.org/abs/2010.10499) par Adrian de Wynter et Daniel J. Perry.
-1. **[BridgeTower](https://huggingface.co/docs/transformers/model_doc/bridgetower)** (de l'Institut de technologie de Harbin/Microsoft Research Asia/Intel Labs) publié dans l'article [BridgeTower : Construire des ponts entre les encodeurs dans l'apprentissage de la représentation vision-langage](https://arxiv.org/abs/2206.08657) par Xiao Xu, Chenfei Wu, Shachar Rosenman, Vasudev Lal, Wanxiang Che, Nan Duan.
-1. **[BROS](https://huggingface.co/docs/transformers/model_doc/bros)** (de NAVER CLOVA) publié dans l'article [BROS : un modèle de langage pré-entraîné axé sur le texte et la mise en page pour une meilleure extraction des informations clés des documents](https://arxiv.org/abs/2108.04539) par Teakgyu Hong, Donghyun Kim, Mingi Ji, Wonseok Hwang, Daehyun Nam, Sungrae Park.
-1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (de Google Research) publié dans l'article [ByT5 : Vers un futur sans jeton avec des modèles pré-entraînés byte-to-byte](https://arxiv.org/abs/2105.13626) par Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel.
-1. **[CamemBERT](https://huggingface.co/docs/transformers/model_doc/camembert)** (d'Inria/Facebook/Sorbonne) publié dans l'article [CamemBERT : un modèle de langue français savoureux](https://arxiv.org/abs/1911.03894) par Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah et Benoît Sagot.
-1. **[CANINE](https://huggingface.co/docs/transformers/model_doc/canine)** (de Google Research) publié dans l'article [CANINE : Pré-entraînement d'un encodeur sans tokenisation efficace pour la représentation du langage](https://arxiv.org/abs/2103.06874) par Jonathan H. Clark, Dan Garrette, Iulia Turc, John Wieting.
-1. **[Chinese-CLIP](https://huggingface.co/docs/transformers/model_doc/chinese_clip)** (d'OFA-Sys) publié dans l'article [Chinese CLIP : Pré-entraînement contrastif vision-langage en chinois](https://arxiv.org/abs/2211.01335) par An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou.
+1. **[BridgeTower](https://huggingface.co/docs/transformers/model_doc/bridgetower)** (de l'Institut de technologie de Harbin/Microsoft Research Asia/Intel Labs) publié dans l'article [BridgeTower: Building Bridges Between Encoders in Vision-Language Representation Learning](https://arxiv.org/abs/2206.08657) par Xiao Xu, Chenfei Wu, Shachar Rosenman, Vasudev Lal, Wanxiang Che, Nan Duan.
+1. **[BROS](https://huggingface.co/docs/transformers/model_doc/bros)** (de NAVER CLOVA) publié dans l'article [BROS: A Pre-trained Language Model Focusing on Text and Layout for Better Key Information Extraction from Documents](https://arxiv.org/abs/2108.04539) par Teakgyu Hong, Donghyun Kim, Mingi Ji, Wonseok Hwang, Daehyun Nam, Sungrae Park.
+1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (de Google Research) publié dans l'article [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) par Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel.
+1. **[CamemBERT](https://huggingface.co/docs/transformers/model_doc/camembert)** (d'Inria/Facebook/Sorbonne) publié dans l'article [CamemBERT: a Tasty French Language Model](https://arxiv.org/abs/1911.03894) par Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah et Benoît Sagot.
+1. **[CANINE](https://huggingface.co/docs/transformers/model_doc/canine)** (de Google Research) publié dans l'article [CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language Representation](https://arxiv.org/abs/2103.06874) par Jonathan H. Clark, Dan Garrette, Iulia Turc, John Wieting.
+1. **[Chinese-CLIP](https://huggingface.co/docs/transformers/model_doc/chinese_clip)** (d'OFA-Sys) publié dans l'article [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) par An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou.
 1. **[CLAP](https://huggingface.co/docs/transformers/model_doc/clap)** (de LAION-AI) publié dans l'article [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://arxiv.org/abs/2211.06687) par Yusong Wu, Ke Chen, Tianyu Zhang, Yuchen Hui, Taylor Berg-Kirkpatrick, Shlomo Dubnov.
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (d'OpenAI) publié dans l'article [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) par Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, Ilya Sutskever.
 1. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (de l'Université de Göttingen) publié dans l'article [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) par Timo Lüddecke et Alexander Ecker.
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** publié dans l'article [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) par James Betker.
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (de Salesforce) publié dans l'article [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) par Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
-1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (de MetaAI) publié dans l'article [Code Llama : Modèles ouverts fondamentaux pour le code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) par Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (de Cohere) publié dans l'article [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) parCohere. 
+1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (de MetaAI) publié dans l'article [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) par Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (de Cohere) publié dans l'article [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) parCohere.
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (de Microsoft Research Asia) publié dans l'article [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) par Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
-1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (de YituTech) publié dans l'article [ConvBERT : Amélioration de BERT avec une convolution dynamique basée sur des plages](https://arxiv.org/abs/2008.02496) par Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
+1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (de YituTech) publié dans l'article [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) par Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (de Facebook AI) publié dans l'article [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) par Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
-1. **[ConvNeXTV2](https://huggingface.co/docs/transformers/model_doc/convnextv2)** (de Facebook AI) publié dans l'article [ConvNeXt V2 : Conception conjointe et mise à l'échelle de ConvNets avec des autoencodeurs masqués](https://arxiv.org/abs/2301.00808) par Sanghyun Woo, Shoubhik Debnath, Ronghang Hu, Xinlei Chen, Zhuang Liu, In So Kweon, Saining Xie.
-1. **[CPM](https://huggingface.co/docs/transformers/model_doc/cpm)** (de l'Université de Tsinghua) publié dans l'article [CPM : Un modèle de langue chinois pré-entraîné génératif à grande échelle](https://arxiv.org/abs/2012.00413) par Zhengyan Zhang, Xu Han, Hao Zhou, Pei Ke, Yuxian Gu, Deming Ye, Yujia Qin, Yusheng Su, Haozhe Ji, Jian Guan, Fanchao Qi, Xiaozhi Wang, Yanan Zheng, Guoyang Zeng, Huanqi Cao, Shengqi Chen, Daixuan Li, Zhenbo Sun, Zhiyuan Liu, Minlie Huang, Wentao Han, Jie Tang, Juanzi Li, Xiaoyan Zhu, Maosong Sun.
+1. **[ConvNeXTV2](https://huggingface.co/docs/transformers/model_doc/convnextv2)** (de Facebook AI) publié dans l'article [ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoders](https://arxiv.org/abs/2301.00808) par Sanghyun Woo, Shoubhik Debnath, Ronghang Hu, Xinlei Chen, Zhuang Liu, In So Kweon, Saining Xie.
+1. **[CPM](https://huggingface.co/docs/transformers/model_doc/cpm)** (de l'Université de Tsinghua) publié dans l'article [CPM: A Large-scale Generative Chinese Pre-trained Language Model](https://arxiv.org/abs/2012.00413) par Zhengyan Zhang, Xu Han, Hao Zhou, Pei Ke, Yuxian Gu, Deming Ye, Yujia Qin, Yusheng Su, Haozhe Ji, Jian Guan, Fanchao Qi, Xiaozhi Wang, Yanan Zheng, Guoyang Zeng, Huanqi Cao, Shengqi Chen, Daixuan Li, Zhenbo Sun, Zhiyuan Liu, Minlie Huang, Wentao Han, Jie Tang, Juanzi Li, Xiaoyan Zhu, Maosong Sun.
 1. **[CPM-Ant](https://huggingface.co/docs/transformers/model_doc/cpmant)** (d'OpenBMB) publié par l'[OpenBMB](https://www.openbmb.org/).
-1. **[CTRL](https://huggingface.co/docs/transformers/model_doc/ctrl)** (de Salesforce) publié dans l'article [CTRL : Un modèle de langage conditionnel de type Transformer pour une génération contrôlable](https://arxiv.org/abs/1909.05858) par Nitish Shirish Keskar*, Bryan McCann*, Lav R. Varshney, Caiming Xiong et Richard Socher.
-1. **[CvT](https://huggingface.co/docs/transformers/model_doc/cvt)** (de Microsoft) publié dans l'article [CvT : Introduction de convolutions aux transformateurs visuels](https://arxiv.org/abs/2103.15808) par Haiping Wu, Bin Xiao, Noel Codella, Mengchen Liu, Xiyang Dai, Lu Yuan, Lei Zhang.
-1. **[Data2Vec](https://huggingface.co/docs/transformers/model_doc/data2vec)** (de Facebook) publié dans l'article [Data2Vec : Un cadre général pour l'apprentissage auto-supervisé en parole, vision et langage](https://arxiv.org/abs/2202.03555) par Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli.
-1. **[DeBERTa](https://huggingface.co/docs/transformers/model_doc/deberta)** (de Microsoft) publié dans l'article [DeBERTa : BERT amélioré avec attention désentrelacée](https://arxiv.org/abs/2006.03654) par Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen.
-1. **[DeBERTa-v2](https://huggingface.co/docs/transformers/model_doc/deberta-v2)** (de Microsoft) publié dans l'article [DeBERTa : BERT amélioré avec attention désentrelacée](https://arxiv.org/abs/2006.03654) par Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen.
-1. **[Decision Transformer](https://huggingface.co/docs/transformers/model_doc/decision_transformer)** (de Berkeley/Facebook/Google) publié dans l'article [Decision Transformer : Apprentissage par renforcement via la modélisation de séquences](https://arxiv.org/abs/2106.01345) par Lili Chen, Kevin Lu, Aravind Rajeswaran, Kimin Lee, Aditya Grover, Michael Laskin, Pieter Abbeel, Aravind Srinivas, Igor Mordatch.
-1. **[Deformable DETR](https://huggingface.co/docs/transformers/model_doc/deformable_detr)** (de SenseTime Research) publié dans l'article [Deformable DETR : Transformateurs déformables pour la détection d'objets de bout en bout](https://arxiv.org/abs/2010.04159) par Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, Jifeng Dai.
-1. **[DeiT](https://huggingface.co/docs/transformers/model_doc/deit)** (de Facebook) publié dans l'article [Entraînement d'images efficace et distillation par l'attention](https://arxiv.org/abs/2012.12877) par Hugo Touvron, Matthieu Cord, Matthijs Douze, Francisco Massa, Alexandre Sablayrolles, Hervé Jégou.
-1. **[DePlot](https://huggingface.co/docs/transformers/model_doc/deplot)** (de Google AI) publié dans l'article [DePlot : Raisonnement visuel en une étape par traduction de l'intrigue en tableau](https://arxiv.org/abs/2212.10505) par Fangyu Liu, Julian Martin Eisenschlos, Francesco Piccinno, Syrine Krichene, Chenxi Pang, Kenton Lee, Mandar Joshi, Wenhu Chen, Nigel Collier, Yasemin Altun.
+1. **[CTRL](https://huggingface.co/docs/transformers/model_doc/ctrl)** (de Salesforce) publié dans l'article [CTRL: A Conditional Transformer Language Model for Controllable Generation](https://arxiv.org/abs/1909.05858) par Nitish Shirish Keskar*, Bryan McCann*, Lav R. Varshney, Caiming Xiong et Richard Socher.
+1. **[CvT](https://huggingface.co/docs/transformers/model_doc/cvt)** (de Microsoft) publié dans l'article [CvT: Introducing Convolutions to Vision Transformers](https://arxiv.org/abs/2103.15808) par Haiping Wu, Bin Xiao, Noel Codella, Mengchen Liu, Xiyang Dai, Lu Yuan, Lei Zhang.
+1. **[Data2Vec](https://huggingface.co/docs/transformers/model_doc/data2vec)** (de Facebook) publié dans l'article [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) par Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli.
+1. **[DeBERTa](https://huggingface.co/docs/transformers/model_doc/deberta)** (de Microsoft) publié dans l'article [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://arxiv.org/abs/2006.03654) par Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen.
+1. **[DeBERTa-v2](https://huggingface.co/docs/transformers/model_doc/deberta-v2)** (de Microsoft) publié dans l'article [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://arxiv.org/abs/2006.03654) par Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen.
+1. **[Decision Transformer](https://huggingface.co/docs/transformers/model_doc/decision_transformer)** (de Berkeley/Facebook/Google) publié dans l'article [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://arxiv.org/abs/2106.01345) par Lili Chen, Kevin Lu, Aravind Rajeswaran, Kimin Lee, Aditya Grover, Michael Laskin, Pieter Abbeel, Aravind Srinivas, Igor Mordatch.
+1. **[Deformable DETR](https://huggingface.co/docs/transformers/model_doc/deformable_detr)** (de SenseTime Research) publié dans l'article [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://arxiv.org/abs/2010.04159) par Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, Jifeng Dai.
+1. **[DeiT](https://huggingface.co/docs/transformers/model_doc/deit)** (de Facebook) publié dans l'article [Training data-efficient image transformers & distillation through attention](https://arxiv.org/abs/2012.12877) par Hugo Touvron, Matthieu Cord, Matthijs Douze, Francisco Massa, Alexandre Sablayrolles, Hervé Jégou.
+1. **[DePlot](https://huggingface.co/docs/transformers/model_doc/deplot)** (de Google AI) publié dans l'article [DePlot: One-shot visual language reasoning by plot-to-table translation](https://arxiv.org/abs/2212.10505) par Fangyu Liu, Julian Martin Eisenschlos, Francesco Piccinno, Syrine Krichene, Chenxi Pang, Kenton Lee, Mandar Joshi, Wenhu Chen, Nigel Collier, Yasemin Altun.
 1. **[Depth Anything](https://huggingface.co/docs/transformers/model_doc/depth_anything)** (de l'université d'Hong Kong et TikTok) publié dans l'article    [Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data](https://arxiv.org/abs/2401.10891) by Lihe Yang, Bingyi Kang, Zilong Huang, Xiaogang Xu, Jiashi Feng, Hengshuang Zhao.
 1. **[DETA](https://huggingface.co/docs/transformers/model_doc/deta)** (de l'Université du Texas à Austin) publié dans l'article [NMS Strikes Back](https://arxiv.org/abs/2212.06137) par Jeffrey Ouyang-Zhang, Jang Hyun Cho, Xingyi Zhou, Philipp Krähenbühl.
-1. **[DETR](https://huggingface.co/docs/transformers/model_doc/detr)** (de Facebook) publié dans l'article [Détection d'objets de bout en bout avec des transformateurs](https://arxiv.org/abs/2005.12872) par Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, Sergey Zagoruyko.
-1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (de Microsoft Research) publié dans l'article [DialoGPT : Pré-entraînement génératif à grande échelle pour la génération de réponses conversationnelles](https://arxiv.org/abs/1911.00536) par Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan.
-1. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (de SHI Labs) publié dans l'article [Transformateur d'attention dilatée pour l'attention aux quartiers](https://arxiv.org/abs/2209.15001) par Ali Hassani et Humphrey Shi.
-1. **[DINOv2](https://huggingface.co/docs/transformers/model_doc/dinov2)** (de Meta AI) publié dans l'article [DINOv2 : Apprentissage de fonctionnalités visuelles robustes sans supervision](https://arxiv.org/abs/2304.07193) par Maxime Oquab, Timothée Darcet, Théo Moutakanni, Huy Vo, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, Mahmoud Assran, Nicolas Ballas, Wojciech Galuba, Russell Howes, Po-Yao Huang, Shang-Wen Li, Ishan Misra, Michael Rabbat, Vasu Sharma, Gabriel Synnaeve, Hu Xu, Hervé Jegou, Julien Mairal, Patrick Labatut, Armand Joulin, Piotr Bojanowski.
-1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (de HuggingFace), publié dans l'article [DistilBERT, une version condensée de BERT : plus petit, plus rapide, moins cher et plus léger](https://arxiv.org/abs/1910.01108) par Victor Sanh, Lysandre Debut et Thomas Wolf. La même méthode a été appliquée pour compresser GPT2 en [DistilGPT2](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation), RoBERTa en [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation), Multilingual BERT en [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation) et une version allemande de DistilBERT.
-1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (de Microsoft Research) publié dans l'article [DiT : Auto-pré-entraînement pour le transformateur d'images de documents](https://arxiv.org/abs/2203.02378) par Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei.
-1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (de NAVER), publié dans l'article [Transformation de compréhension de documents sans OCR](https://arxiv.org/abs/2111.15664) par 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)** (de Facebook) publié dans l'article [Passage dense pour la recherche de réponses à des questions en domaine ouvert](https://arxiv.org/abs/2004.04906) par Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen et Wen-tau Yih.
-1. **[DPT](https://huggingface.co/docs/transformers/master/model_doc/dpt)** (d'Intel Labs) publié dans l'article [Transformateurs de vision pour la prédiction dense](https://arxiv.org/abs/2103.13413) par René Ranftl, Alexey Bochkovskiy, Vladlen Koltun.
-1. **[EfficientFormer](https://huggingface.co/docs/transformers/model_doc/efficientformer)** (de Snap Research) publié dans l'article [EfficientFormer : Transformateurs de vision à la vitesse de MobileNet](https://arxiv.org/abs/2206.01191) par Yanyu Li, Geng Yuan, Yang Wen, Ju Hu, Georgios Evangelidis, Sergey Tulyakov, Yanzhi Wang, Jian Ren.
-1. **[EfficientNet](https://huggingface.co/docs/transformers/model_doc/efficientnet)** (de Google Brain) publié dans l'article [EfficientNet: Repenser l'échelle des modèles pour les réseaux de neurones convolutionnels](https://arxiv.org/abs/1905.11946) par Mingxing Tan, Quoc V. Le.
-1. **[ELECTRA](https://huggingface.co/docs/transformers/model_doc/electra)** (de Google Research/Université Stanford) publié dans l'article [ELECTRA: Pré-entraîner les encodeurs de texte en tant que discriminateurs plutôt que des générateurs](https://arxiv.org/abs/2003.10555) par Kevin Clark, Minh-Thang Luong, Quoc V. Le, Christopher D. Manning.
-1. **[EnCodec](https://huggingface.co/docs/transformers/model_doc/encodec)** (de Meta AI) publié dans l'article [Compression neuronale audio de haute fidélité](https://arxiv.org/abs/2210.13438) par Alexandre Défossez, Jade Copet, Gabriel Synnaeve, Yossi Adi.
-1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (de Google Research) publié dans l'article [Exploiter des points de contrôle pré-entraînés pour les tâches de génération de séquences](https://arxiv.org/abs/1907.12461) par Sascha Rothe, Shashi Narayan, Aliaksei Severyn.
-1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (de Baidu) publié dans l'article [ERNIE: Intégration améliorée des représentations par la connaissance](https://arxiv.org/abs/1904.09223) par Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
-1. **[ErnieM](https://huggingface.co/docs/transformers/model_doc/ernie_m)** (de Baidu) publié dans l'article [ERNIE-M: Représentation multilingue améliorée en alignant les sémantiques interlingues avec des corpus monolingues](https://arxiv.org/abs/2012.15674) par Xuan Ouyang, Shuohuan Wang, Chao Pang, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang.
-1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (de Meta AI) sont des modèles de langage de protéines de type transformateur. **ESM-1b** a été publié dans l'article [La structure et la fonction biologiques émergent de la mise à l'échelle de l'apprentissage non supervisé à 250 millions de séquences de protéines](https://www.pnas.org/content/118/15/e2016239118) par Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma et Rob Fergus. **ESM-1v** a été publié dans l'article [Les modèles de langage permettent une prédiction hors champ des effets des mutations sur la fonction des protéines](https://doi.org/10.1101/2021.07.09.450648) par Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu et Alexander Rives. **ESM-2 et ESMFold** ont été publiés avec l'article [Les modèles de langage des séquences de protéines à l'échelle de l'évolution permettent une prédiction précise de la structure](https://doi.org/10.1101/2022.07.20.500902) par Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
+1. **[DETR](https://huggingface.co/docs/transformers/model_doc/detr)** (de Facebook) publié dans l'article [End-to-End Object Detection with Transformers](https://arxiv.org/abs/2005.12872) par Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, Sergey Zagoruyko.
+1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (de Microsoft Research) publié dans l'article [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://arxiv.org/abs/1911.00536) par Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan.
+1. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (de SHI Labs) publié dans l'article [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001) par Ali Hassani et Humphrey Shi.
+1. **[DINOv2](https://huggingface.co/docs/transformers/model_doc/dinov2)** (de Meta AI) publié dans l'article [DINOv2: Learning Robust Visual Features without Supervision](https://arxiv.org/abs/2304.07193) par Maxime Oquab, Timothée Darcet, Théo Moutakanni, Huy Vo, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, Mahmoud Assran, Nicolas Ballas, Wojciech Galuba, Russell Howes, Po-Yao Huang, Shang-Wen Li, Ishan Misra, Michael Rabbat, Vasu Sharma, Gabriel Synnaeve, Hu Xu, Hervé Jegou, Julien Mairal, Patrick Labatut, Armand Joulin, Piotr Bojanowski.
+1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (de HuggingFace), publié dans l'article [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108) par Victor Sanh, Lysandre Debut et Thomas Wolf. La même méthode a été appliquée pour compresser GPT2 en [DistilGPT2](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation), RoBERTa en [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation), Multilingual BERT en [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation) et une version allemande de DistilBERT.
+1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (de Microsoft Research) publié dans l'article [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) par Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei.
+1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (de NAVER), publié dans l'article [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) par 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)** (de Facebook) publié dans l'article [Dense Passage Retrieval for Open-Domain Question Answering](https://arxiv.org/abs/2004.04906) par Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen et Wen-tau Yih.
+1. **[DPT](https://huggingface.co/docs/transformers/master/model_doc/dpt)** (d'Intel Labs) publié dans l'article [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) par René Ranftl, Alexey Bochkovskiy, Vladlen Koltun.
+1. **[EfficientFormer](https://huggingface.co/docs/transformers/model_doc/efficientformer)** (de Snap Research) publié dans l'article [EfficientFormer: Vision Transformers at MobileNetSpeed](https://arxiv.org/abs/2206.01191) par Yanyu Li, Geng Yuan, Yang Wen, Ju Hu, Georgios Evangelidis, Sergey Tulyakov, Yanzhi Wang, Jian Ren.
+1. **[EfficientNet](https://huggingface.co/docs/transformers/model_doc/efficientnet)** (de Google Brain) publié dans l'article [EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks](https://arxiv.org/abs/1905.11946) par Mingxing Tan, Quoc V. Le.
+1. **[ELECTRA](https://huggingface.co/docs/transformers/model_doc/electra)** (de Google Research/Université Stanford) publié dans l'article [ELECTRA: Pre-training text encoders as discriminators rather than generators](https://arxiv.org/abs/2003.10555) par Kevin Clark, Minh-Thang Luong, Quoc V. Le, Christopher D. Manning.
+1. **[EnCodec](https://huggingface.co/docs/transformers/model_doc/encodec)** (de Meta AI) publié dans l'article [High Fidelity Neural Audio Compression](https://arxiv.org/abs/2210.13438) par Alexandre Défossez, Jade Copet, Gabriel Synnaeve, Yossi Adi.
+1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (de Google Research) publié dans l'article [Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) par Sascha Rothe, Shashi Narayan, Aliaksei Severyn.
+1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (de Baidu) publié dans l'article [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) par Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ErnieM](https://huggingface.co/docs/transformers/model_doc/ernie_m)** (de Baidu) publié dans l'article [ERNIE-M: Enhanced Multilingual Representation by Aligning Cross-lingual Semantics with Monolingual Corpora](https://arxiv.org/abs/2012.15674) par Xuan Ouyang, Shuohuan Wang, Chao Pang, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang.
+1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (de Meta AI) sont des modèles de langage de protéines de type transformateur. **ESM-1b** a été publié dans l'article [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) par Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma et Rob Fergus. **ESM-1v** a été publié dans l'article [Les modèles de langage permettent une prédiction hors champ des effets des mutations sur la fonction des protéines](https://doi.org/10.1101/2021.07.09.450648) par Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu et Alexander Rives. **ESM-2 et ESMFold** ont été publiés avec l'article [Les modèles de langage des séquences de protéines à l'échelle de l'évolution permettent une prédiction précise de la structure](https://doi.org/10.1101/2022.07.20.500902) par Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
 1. **[Falcon](https://huggingface.co/docs/transformers/model_doc/falcon)** (de Technology Innovation Institute) par Almazrouei, Ebtesam et Alobeidli, Hamza et Alshamsi, Abdulaziz et Cappelli, Alessandro et Cojocaru, Ruxandra et Debbah, Merouane et Goffinet, Etienne et Heslow, Daniel et Launay, Julien et Malartic, Quentin et Noune, Badreddine et Pannier, Baptiste et Penedo, Guilherme.
-1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (d'ESPnet) publié dans l'article [Développements récents sur la boîte à outils Espnet boostés par Conformer](https://arxiv.org/abs/2010.13956) par Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang et Yuekai Zhang.
+1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (d'ESPnet) publié dans l'article [Recent Developments On Espnet Toolkit Boosted By Conformer](https://arxiv.org/abs/2010.13956) par Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang et Yuekai Zhang.
 1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (de Google AI) publié dans le référentiel [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) par Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le et Jason Wei
 1. **[FLAN-UL2](https://huggingface.co/docs/transformers/model_doc/flan-ul2)** (de Google AI) publié dans le référentiel [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-ul2-checkpoints) par Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le et Jason Wei
-1. **[FlauBERT](https://huggingface.co/docs/transformers/model_doc/flaubert)** (du CNRS) publié dans l'article [FlauBERT : Pré-entraînement de modèle de langue non supervisé pour le français](https://arxiv.org/abs/1912.05372) par 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)** (de Facebook AI) publié dans l'article [FLAVA : Un modèle fondamental d'alignement de la langue et de la vision](https://arxiv.org/abs/2112.04482) par Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach et Douwe Kiela.
-1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (de Google Research) publié dans l'article [FNet : Mélanger les jetons avec des transformations de Fourier](https://arxiv.org/abs/2105.03824) par James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
-1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (de Microsoft Research) publié dans l'article [Réseaux de modulation focale](https://arxiv.org/abs/2203.11926) par Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
-1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (de l'Université Carnegie Mellon/Google Brain) publié dans l'article [Funnel-Transformer : Filtrer la redondance séquentielle pour un traitement efficace du langage](https://arxiv.org/abs/2006.03236) par Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
-1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (de ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. Publié dans l'article [billet de blog](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (de Google) publié dans l'article [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) parthe Gemma Google team.
-1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (de Microsoft Research) publié dans l'article [GIT : Un transformateur génératif d'images en texte pour la vision et le langage](https://arxiv.org/abs/2205.14100) par Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
-1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (de la KAIST) publié dans l'article [Réseaux de chemins globaux-locaux pour l'estimation de profondeur monoculaire avec Vertical CutDepth](https://arxiv.org/abs/2201.07436) par Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
-1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (d'OpenAI) publié dans l'article [Améliorer la compréhension du langage par l'apprentissage préalable génératif](https://openai.com/research/language-unsupervised/) par Alec Radford, Karthik Narasimhan, Tim Salimans et Ilya Sutskever.
+1. **[FlauBERT](https://huggingface.co/docs/transformers/model_doc/flaubert)** (du CNRS) publié dans l'article [FlauBERT: Unsupervised Language Model Pre-training for French](https://arxiv.org/abs/1912.05372) par 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)** (de Facebook AI) publié dans l'article [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) par Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach et Douwe Kiela.
+1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (de Google Research) publié dans l'article [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) par James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
+1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (de Microsoft Research) publié dans l'article [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) par Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
+1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (de l'Université Carnegie Mellon/Google Brain) publié dans l'article [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) par Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
+1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (de ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. Publié dans l'article [blog post](https://www.adept.ai/blog/fuyu-8b)
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (de Google) publié dans l'article [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) parthe Gemma Google team.
+1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (de Microsoft Research) publié dans l'article [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) par Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
+1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (de la KAIST) publié dans l'article [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) par Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
+1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (d'OpenAI) publié dans l'article [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) par Alec Radford, Karthik Narasimhan, Tim Salimans et Ilya Sutskever.
 1. **[GPT Neo](https://huggingface.co/docs/transformers/model_doc/gpt_neo)** (d'EleutherAI) publié dans le référentiel [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) par Sid Black, Stella Biderman, Leo Gao, Phil Wang et Connor Leahy.
-1. **[GPT NeoX](https://huggingface.co/docs/transformers/model_doc/gpt_neox)** (d'EleutherAI) publié dans l'article [GPT-NeoX-20B: Un modèle de langue autonome open source](https://arxiv.org/abs/2204.06745) par Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach
+1. **[GPT NeoX](https://huggingface.co/docs/transformers/model_doc/gpt_neox)** (d'EleutherAI) publié dans l'article [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) par Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach
 1. **[GPT NeoX Japanese](https://huggingface.co/docs/transformers/model_doc/gpt_neox_japanese)** (de ABEJA) publié par Shinya Otani, Takayoshi Makabe, Anuj Arora et Kyo Hattori.
-1. **[GPT-2](https://huggingface.co/docs/transformers/model_doc/gpt2)** (d'OpenAI) a été publié dans l'article [Les modèles de langage sont des apprenants multitâches non supervisés](https://openai.com/research/better-language-models/) par Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei et Ilya Sutskever.
+1. **[GPT-2](https://huggingface.co/docs/transformers/model_doc/gpt2)** (d'OpenAI) a été publié dans l'article [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/) par Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei et Ilya Sutskever.
 1. **[GPT-J](https://huggingface.co/docs/transformers/model_doc/gptj)** (d'EleutherAI) a été publié dans le dépôt [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) par Ben Wang et Aran Komatsuzaki.
-1. **[GPT-Sw3](https://huggingface.co/docs/transformers/model_doc/gpt-sw3)** (d'AI-Sweden) a été publié dans l'article [Leçons apprises de GPT-SW3 : Construction du premier modèle de langage génératif à grande échelle pour le suédois](http://www.lrec-conf.org/proceedings/lrec2022/pdf/2022.lrec-1.376.pdf) par Ariel Ekgren, Amaru Cuba Gyllensten, Evangelia Gogoulou, Alice Heiman, Severine Verlinden, Joey Öhman, Fredrik Carlsson, Magnus Sahlgren.
-1. **[GPTBigCode](https://huggingface.co/docs/transformers/model_doc/gpt_bigcode)** (de BigCode) a été publié dans l'article [SantaCoder: ne visez pas les étoiles !](https://arxiv.org/abs/2301.03988) par Loubna Ben Allal, Raymond Li, Denis Kocetkov, Chenghao Mou, Christopher Akiki, Carlos Munoz Ferrandis, Niklas Muennighoff, Mayank Mishra, Alex Gu, Manan Dey, Logesh Kumar Umapathi, Carolyn Jane Anderson, Yangtian Zi, Joel Lamy Poirier, Hailey Schoelkopf, Sergey Troshin, Dmitry Abulkhanov, Manuel Romero, Michael Lappert, Francesco De Toni, Bernardo García del Río, Qian Liu, Shamik Bose, Urvashi Bhattacharyya, Terry Yue Zhuo, Ian Yu, Paulo Villegas, Marco Zocca, Sourab Mangrulkar, David Lansky, Huu Nguyen, Danish Contractor, Luis Villa, Jia Li, Dzmitry Bahdanau, Yacine Jernite, Sean Hughes, Daniel Fried, Arjun Guha, Harm de Vries, Leandro von Werra.
+1. **[GPT-Sw3](https://huggingface.co/docs/transformers/model_doc/gpt-sw3)** (d'AI-Sweden) a été publié dans l'article [Lessons Learned from GPT-SW3: Building the First Large-Scale Generative Language Model for Swedish](http://www.lrec-conf.org/proceedings/lrec2022/pdf/2022.lrec-1.376.pdf) par Ariel Ekgren, Amaru Cuba Gyllensten, Evangelia Gogoulou, Alice Heiman, Severine Verlinden, Joey Öhman, Fredrik Carlsson, Magnus Sahlgren.
+1. **[GPTBigCode](https://huggingface.co/docs/transformers/model_doc/gpt_bigcode)** (de BigCode) a été publié dans l'article [SantaCoder: don't reach for the stars!](https://arxiv.org/abs/2301.03988) par Loubna Ben Allal, Raymond Li, Denis Kocetkov, Chenghao Mou, Christopher Akiki, Carlos Munoz Ferrandis, Niklas Muennighoff, Mayank Mishra, Alex Gu, Manan Dey, Logesh Kumar Umapathi, Carolyn Jane Anderson, Yangtian Zi, Joel Lamy Poirier, Hailey Schoelkopf, Sergey Troshin, Dmitry Abulkhanov, Manuel Romero, Michael Lappert, Francesco De Toni, Bernardo García del Río, Qian Liu, Shamik Bose, Urvashi Bhattacharyya, Terry Yue Zhuo, Ian Yu, Paulo Villegas, Marco Zocca, Sourab Mangrulkar, David Lansky, Huu Nguyen, Danish Contractor, Luis Villa, Jia Li, Dzmitry Bahdanau, Yacine Jernite, Sean Hughes, Daniel Fried, Arjun Guha, Harm de Vries, Leandro von Werra.
 1. **[GPTSAN-japanese](https://huggingface.co/docs/transformers/model_doc/gptsan-japanese)** a été publié dans le dépôt [tanreinama/GPTSAN](https://github.com/tanreinama/GPTSAN/blob/main/report/model.md) par Toshiyuki Sakamoto (tanreinama).
-1. **[Graphormer](https://huggingface.co/docs/transformers/model_doc/graphormer)** (de Microsoft) a été publié dans l'article [Les Transformers sont-ils vraiment inefficaces pour la représentation graphique ?](https://arxiv.org/abs/2106.05234) par Chengxuan Ying, Tianle Cai, Shengjie Luo, Shuxin Zheng, Guolin Ke, Di He, Yanming Shen, Tie-Yan Liu.
-1. **[GroupViT](https://huggingface.co/docs/transformers/model_doc/groupvit)** (de l'UCSD, NVIDIA) a été publié dans l'article [GroupViT : la segmentation sémantique émerge de la supervision textuelle](https://arxiv.org/abs/2202.11094) par Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang.
-1. **[HerBERT](https://huggingface.co/docs/transformers/model_doc/herbert)** (d'Allegro.pl, AGH University of Science and Technology) a été publié dans l'article [KLEJ : référentiel complet pour la compréhension du langage polonais](https://www.aclweb.org/anthology/2020.acl-main.111.pdf) par Piotr Rybak, Robert Mroczkowski, Janusz Tracz, Ireneusz Gawlik.
-1. **[Hubert](https://huggingface.co/docs/transformers/model_doc/hubert)** (de Facebook) a été publié dans l'article [HuBERT : Apprentissage de la représentation autonome de la parole par prédiction masquée des unités cachées](https://arxiv.org/abs/2106.07447) par Wei-Ning Hsu, Benjamin Bolte, Yao-Hung Hubert Tsai, Kushal Lakhotia, Ruslan Salakhutdinov, Abdelrahman Mohamed.
-1. **[I-BERT](https://huggingface.co/docs/transformers/model_doc/ibert)** (de Berkeley) a été publié dans l'article [I-BERT : Quantification entière de BERT avec des entiers uniquement](https://arxiv.org/abs/2101.01321) par Sehoon Kim, Amir Gholami, Zhewei Yao, Michael W. Mahoney, Kurt Keutzer.
-1. **[IDEFICS](https://huggingface.co/docs/transformers/model_doc/idefics)** (de HuggingFace) a été publié dans l'article [OBELICS : Un ensemble de données filtré à l'échelle du Web d'intercalation de documents texte-image](https://huggingface.co/papers/2306.16527) par Hugo Laurençon, Lucile Saulnier, Léo Tronchon, Stas Bekman, Amanpreet Singh, Anton Lozhkov, Thomas Wang, Siddharth Karamcheti, Alexander M. Rush, Douwe Kiela, Matthieu Cord, Victor Sanh.
-1. **[ImageGPT](https://huggingface.co/docs/transformers/model_doc/imagegpt)** (d'OpenAI) a été publié dans l'article [Pré-entraînement génératif à partir de pixels](https://openai.com/blog/image-gpt/) par Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever.
+1. **[Graphormer](https://huggingface.co/docs/transformers/model_doc/graphormer)** (de Microsoft) a été publié dans l'article [Do Transformers Really Perform Bad for Graph Representation?](https://arxiv.org/abs/2106.05234) par Chengxuan Ying, Tianle Cai, Shengjie Luo, Shuxin Zheng, Guolin Ke, Di He, Yanming Shen, Tie-Yan Liu.
+1. **[GroupViT](https://huggingface.co/docs/transformers/model_doc/groupvit)** (de l'UCSD, NVIDIA) a été publié dans l'article [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) par Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang.
+1. **[HerBERT](https://huggingface.co/docs/transformers/model_doc/herbert)** (d'Allegro.pl, AGH University of Science and Technology) a été publié dans l'article [KLEJ: Comprehensive Benchmark for Polish Language Understanding](https://www.aclweb.org/anthology/2020.acl-main.111.pdf) par Piotr Rybak, Robert Mroczkowski, Janusz Tracz, Ireneusz Gawlik.
+1. **[Hubert](https://huggingface.co/docs/transformers/model_doc/hubert)** (de Facebook) a été publié dans l'article [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://arxiv.org/abs/2106.07447) par Wei-Ning Hsu, Benjamin Bolte, Yao-Hung Hubert Tsai, Kushal Lakhotia, Ruslan Salakhutdinov, Abdelrahman Mohamed.
+1. **[I-BERT](https://huggingface.co/docs/transformers/model_doc/ibert)** (de Berkeley) a été publié dans l'article [I-BERT: Integer-only BERT Quantization](https://arxiv.org/abs/2101.01321) par Sehoon Kim, Amir Gholami, Zhewei Yao, Michael W. Mahoney, Kurt Keutzer.
+1. **[IDEFICS](https://huggingface.co/docs/transformers/model_doc/idefics)** (de HuggingFace) a été publié dans l'article [OBELICS: An Open Web-Scale Filtered Dataset of Interleaved Image-Text Documents](https://huggingface.co/papers/2306.16527) par Hugo Laurençon, Lucile Saulnier, Léo Tronchon, Stas Bekman, Amanpreet Singh, Anton Lozhkov, Thomas Wang, Siddharth Karamcheti, Alexander M. Rush, Douwe Kiela, Matthieu Cord, Victor Sanh.
+1. **[ImageGPT](https://huggingface.co/docs/transformers/model_doc/imagegpt)** (d'OpenAI) a été publié dans l'article [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) par Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever.
 1. **[Informer](https://huggingface.co/docs/transformers/model_doc/informer)** (de l'Université de Beihang, UC Berkeley, Rutgers University, SEDD Company) a été publié dans l'article [Informer : Au-delà du Transformer efficace pour la prévision de séries temporel
-1. **[InstructBLIP](https://huggingface.co/docs/transformers/model_doc/instructblip)** (de Salesforce) a été publié dans l'article [InstructBLIP : Vers des modèles vision-langage polyvalents avec un réglage d'instructions](https://arxiv.org/abs/2305.06500) de Wenliang Dai, Junnan Li, Dongxu Li, Anthony Meng Huat Tiong, Junqi Zhao, Weisheng Wang, Boyang Li, Pascale Fung, Steven Hoi.
-1. **[Jukebox](https://huggingface.co/docs/transformers/model_doc/jukebox)** (d'OpenAI) a été publié dans l'article [Jukebox : Un modèle génératif pour la musique](https://arxiv.org/pdf/2005.00341.pdf) de Prafulla Dhariwal, Heewoo Jun, Christine Payne, Jong Wook Kim, Alec Radford, Ilya Sutskever.
-1. **[KOSMOS-2](https://huggingface.co/docs/transformers/model_doc/kosmos-2)** (de Microsoft Research Asia) a été publié dans l'article [Kosmos-2 : Ancrage de modèles linguistiques multimodaux à travers le monde](https://arxiv.org/abs/2306.14824) de Zhiliang Peng, Wenhui Wang, Li Dong, Yaru Hao, Shaohan Huang, Shuming Ma, Furu Wei.
-1. **[LayoutLM](https://huggingface.co/docs/transformers/model_doc/layoutlm)** (de Microsoft Research Asia) a été publié dans l'article [LayoutLM : Pré-entraînement de texte et de mise en page pour la compréhension d'images de documents](https://arxiv.org/abs/1912.13318) de Yiheng Xu, Minghao Li, Lei Cui, Shaohan Huang, Furu Wei, Ming Zhou.
-1. **[LayoutLMv2](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (de Microsoft Research Asia) a été publié dans l'article [LayoutLMv2 : Pré-entraînement multimodal pour la compréhension visuellement riche de documents](https://arxiv.org/abs/2012.14740) de Yang Xu, Yiheng Xu, Tengchao Lv, Lei Cui, Furu Wei, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Wanxiang Che, Min Zhang, Lidong Zhou.
-1. **[LayoutLMv3](https://huggingface.co/docs/transformers/model_doc/layoutlmv3)** (de Microsoft Research Asia) a été publié dans l'article [LayoutLMv3 : Pré-entraînement pour l'IA de documents avec un masquage de texte et d'image unifié](https://arxiv.org/abs/2204.08387) de Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, Furu Wei.
-1. **[LayoutXLM](https://huggingface.co/docs/transformers/model_doc/layoutxlm)** (de Microsoft Research Asia) a été publié dans l'article [LayoutXLM : Pré-entraînement multimodal pour la compréhension de documents visuellement riches et multilingues](https://arxiv.org/abs/2104.08836) de Yiheng Xu, Tengchao Lv, Lei Cui, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Furu Wei.
-1. **[LED](https://huggingface.co/docs/transformers/model_doc/led)** (d'AllenAI) a été publié dans l'article [Longformer : Le transformateur de documents longs](https://arxiv.org/abs/2004.05150) de Iz Beltagy, Matthew E. Peters, Arman Cohan.
-1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (de Meta AI) a été publié dans l'article [LeViT : Un transformateur de vision déguisé en réseau de neurones convolutionnel pour une inférence plus rapide](https://arxiv.org/abs/2104.01136) de Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze.
-1. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (de l'Université de technologie du Sud de la Chine) a été publié dans l'article [LiLT : Un transformateur de mise en page simple mais efficace et indépendant de la langue pour la compréhension de documents structurés](https://arxiv.org/abs/2202.13669) de Jiapeng Wang, Lianwen Jin, Kai Ding.
-1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (de l'équipe FAIR de Meta AI) a été publié dans l'article [LLaMA : Modèles linguistiques de base ouverts et efficaces](https://arxiv.org/abs/2302.13971) de Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.
-1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (de l'équipe FAIR de Meta AI) a été publié dans l'article [Llama2 : Modèles de base ouverts et affinés pour le chat](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) de Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.
+1. **[InstructBLIP](https://huggingface.co/docs/transformers/model_doc/instructblip)** (de Salesforce) a été publié dans l'article [InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning](https://arxiv.org/abs/2305.06500) de Wenliang Dai, Junnan Li, Dongxu Li, Anthony Meng Huat Tiong, Junqi Zhao, Weisheng Wang, Boyang Li, Pascale Fung, Steven Hoi.
+1. **[Jukebox](https://huggingface.co/docs/transformers/model_doc/jukebox)** (d'OpenAI) a été publié dans l'article [Jukebox: A Generative Model for Music](https://arxiv.org/pdf/2005.00341.pdf) de Prafulla Dhariwal, Heewoo Jun, Christine Payne, Jong Wook Kim, Alec Radford, Ilya Sutskever.
+1. **[KOSMOS-2](https://huggingface.co/docs/transformers/model_doc/kosmos-2)** (de Microsoft Research Asia) a été publié dans l'article [Kosmos-2: Grounding Multimodal Large Language Models to the World](https://arxiv.org/abs/2306.14824) de Zhiliang Peng, Wenhui Wang, Li Dong, Yaru Hao, Shaohan Huang, Shuming Ma, Furu Wei.
+1. **[LayoutLM](https://huggingface.co/docs/transformers/model_doc/layoutlm)** (de Microsoft Research Asia) a été publié dans l'article [LayoutLM: Pre-training of Text and Layout for Document Image Understanding](https://arxiv.org/abs/1912.13318) de Yiheng Xu, Minghao Li, Lei Cui, Shaohan Huang, Furu Wei, Ming Zhou.
+1. **[LayoutLMv2](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (de Microsoft Research Asia) a été publié dans l'article [LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding](https://arxiv.org/abs/2012.14740) de Yang Xu, Yiheng Xu, Tengchao Lv, Lei Cui, Furu Wei, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Wanxiang Che, Min Zhang, Lidong Zhou.
+1. **[LayoutLMv3](https://huggingface.co/docs/transformers/model_doc/layoutlmv3)** (de Microsoft Research Asia) a été publié dans l'article [LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking](https://arxiv.org/abs/2204.08387) de Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, Furu Wei.
+1. **[LayoutXLM](https://huggingface.co/docs/transformers/model_doc/layoutxlm)** (de Microsoft Research Asia) a été publié dans l'article [LayoutXLM: Multimodal Pre-training for Multilingual Visually-rich Document Understanding](https://arxiv.org/abs/2104.08836) de Yiheng Xu, Tengchao Lv, Lei Cui, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Furu Wei.
+1. **[LED](https://huggingface.co/docs/transformers/model_doc/led)** (d'AllenAI) a été publié dans l'article [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) de Iz Beltagy, Matthew E. Peters, Arman Cohan.
+1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (de Meta AI) a été publié dans l'article [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) de Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze.
+1. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (de l'Université de technologie du Sud de la Chine) a été publié dans l'article [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) de Jiapeng Wang, Lianwen Jin, Kai Ding.
+1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (de l'équipe FAIR de Meta AI) a été publié dans l'article [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) de Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.
+1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (de l'équipe FAIR de Meta AI) a été publié dans l'article [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) de Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (de Microsoft Research & University of Wisconsin-Madison) a été publié dans l'article [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) de Haotian Liu, Chunyuan Li, Yuheng Li et Yong Jae Lee.
-1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (d'AllenAI) a été publié dans l'article [Longformer : Le transformateur de documents longs](https://arxiv.org/abs/2004.05150) de Iz Beltagy, Matthew E. Peters, Arman Cohan.
-1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (de Google AI) a été publié dans l'article [LongT5 : Transformateur de texte-à-texte efficace pour de longues séquences](https://arxiv.org/abs/2112.07916) de Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
-1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (de Studio Ousia) a été publié dans l'article [LUKE : Représentations contextuelles profondes d'entités avec auto-attention consciente des entités](https://arxiv.org/abs/2010.01057) de Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
-1. **[LXMERT](https://huggingface.co/docs/transformers/model_doc/lxmert)** (de l'UNC Chapel Hill) a été publié dans l'article [LXMERT : Apprentissage de représentations d'encodeur cross-modal à partir de transformateurs pour le questionnement en domaine ouvert](https://arxiv.org/abs/1908.07490) de Hao Tan et Mohit Bansal.
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (de Microsoft Research & University of Wisconsin-Madison) publié dans l'article [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) parHaotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
+1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (d'AllenAI) a été publié dans l'article [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) de Iz Beltagy, Matthew E. Peters, Arman Cohan.
+1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (de Google AI) a été publié dans l'article [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) de Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
+1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (de Studio Ousia) a été publié dans l'article [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) de Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
+1. **[LXMERT](https://huggingface.co/docs/transformers/model_doc/lxmert)** (de l'UNC Chapel Hill) a été publié dans l'article [LXMERT: Learning Cross-Modality Encoder Representations from Transformers for Open-Domain Question Answering](https://arxiv.org/abs/1908.07490) de Hao Tan et Mohit Bansal.
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (de Facebook) a été publié dans l'article [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) de Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve et Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (de Facebook) a été publié dans l'article [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) de Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
-1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (de Google) a été publié dans l'article [MADLAD-400 : Un ensemble de données multilingue et de niveau document](https://arxiv.org/abs/2309.04662) de Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (de Albert Gu and Tri Dao) publié dans l'article [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) parAlbert Gu and Tri Dao.
+1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (de Google) a été publié dans l'article [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) de Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (de Albert Gu and Tri Dao) publié dans l'article [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) parAlbert Gu and Tri Dao.
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Des modèles de traduction automatique formés avec les données [OPUS](http://opus.nlpl.eu/) par Jörg Tiedemann. Le [cadre Marian](https://marian-nmt.github.io/) est en cours de développement par l'équipe Microsoft Translator.
-1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (de Microsoft Research Asia) a été publié dans l'article [MarkupLM : Pré-entraînement de texte et de langage de balisage pour la compréhension visuellement riche de documents](https://arxiv.org/abs/2110.08518) de Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
+1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (de Microsoft Research Asia) a été publié dans l'article [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) de Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (de FAIR et UIUC) a été publié dans l'article [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) de Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar.
 1. **[MaskFormer](https://huggingface.co/docs/transformers/model_doc/maskformer)** (de Meta et UIUC) a été publié dans l'article [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) de Bowen Cheng, Alexander G. Schwing, Alexander Kirillov.
-1. **[MatCha](https://huggingface.co/docs/transformers/model_doc/matcha)** (de Google AI) a été publié dans l'article [MatCha : Amélioration du pré-entraînement de langage visuel avec raisonnement mathématique et décomposition de graphiques](https://arxiv.org/abs/2212.09662) de Fangyu Liu, Francesco Piccinno, Syrine Krichene, Chenxi Pang, Kenton Lee, Mandar Joshi, Yasemin Altun, Nigel Collier, Julian Martin Eisenschlos.
+1. **[MatCha](https://huggingface.co/docs/transformers/model_doc/matcha)** (de Google AI) a été publié dans l'article [MatCha: Enhancing Visual Language Pretraining with Math Reasoning and Chart Derendering](https://arxiv.org/abs/2212.09662) de Fangyu Liu, Francesco Piccinno, Syrine Krichene, Chenxi Pang, Kenton Lee, Mandar Joshi, Yasemin Altun, Nigel Collier, Julian Martin Eisenschlos.
 1. **[mBART](https://huggingface.co/docs/transformers/model_doc/mbart)** (de Facebook) a été publié dans l'article [Pré-entraînement de débruitage multilingue pour la traduction automatique neuronale
-1. **[mBART-50](https://huggingface.co/docs/transformers/model_doc/mbart)** (de Facebook) a été publié dans l'article [Traduction multilingue avec un pré-entraînement et un fine-tuning multilingues extensibles](https://arxiv.org/abs/2008.00401) par Yuqing Tang, Chau Tran, Xian Li, Peng-Jen Chen, Naman Goyal, Vishrav Chaudhary, Jiatao Gu, Angela Fan.
-1. **[MEGA](https://huggingface.co/docs/transformers/model_doc/mega)** (de Meta/USC/CMU/SJTU) a été publié dans l'article [Mega : Attention équipée d'une moyenne mobile](https://arxiv.org/abs/2209.10655) par Xuezhe Ma, Chunting Zhou, Xiang Kong, Junxian He, Liangke Gui, Graham Neubig, Jonathan May et Luke Zettlemoyer.
-1. **[Megatron-BERT](https://huggingface.co/docs/transformers/model_doc/megatron-bert)** (de NVIDIA) a été publié dans l'article [Megatron-LM : Entraînement de modèles linguistiques de plusieurs milliards de paramètres en utilisant le parallélisme de modèle](https://arxiv.org/abs/1909.08053) par Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper et Bryan Catanzaro.
-1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (de NVIDIA) a été publié dans l'article [Megatron-LM : Entraînement de modèles linguistiques de plusieurs milliards de paramètres en utilisant le parallélisme de modèle](https://arxiv.org/abs/1909.08053) par Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper et Bryan Catanzaro.
-1. **[MGP-STR](https://huggingface.co/docs/transformers/model_doc/mgp-str)** (d'Alibaba Research) a été publié dans l'article [Prédiction multi-granularité pour la reconnaissance de texte de scène](https://arxiv.org/abs/2209.03592) par Peng Wang, Cheng Da et Cong Yao.
+1. **[mBART-50](https://huggingface.co/docs/transformers/model_doc/mbart)** (de Facebook) a été publié dans l'article [Multilingual Translation with Extensible Multilingual Pretraining and Finetuning](https://arxiv.org/abs/2008.00401) par Yuqing Tang, Chau Tran, Xian Li, Peng-Jen Chen, Naman Goyal, Vishrav Chaudhary, Jiatao Gu, Angela Fan.
+1. **[MEGA](https://huggingface.co/docs/transformers/model_doc/mega)** (de Meta/USC/CMU/SJTU) a été publié dans l'article [Mega: Moving Average Equipped Gated Attention](https://arxiv.org/abs/2209.10655) par Xuezhe Ma, Chunting Zhou, Xiang Kong, Junxian He, Liangke Gui, Graham Neubig, Jonathan May et Luke Zettlemoyer.
+1. **[Megatron-BERT](https://huggingface.co/docs/transformers/model_doc/megatron-bert)** (de NVIDIA) a été publié dans l'article [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) par Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper et Bryan Catanzaro.
+1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (de NVIDIA) a été publié dans l'article [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) par Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper et Bryan Catanzaro.
+1. **[MGP-STR](https://huggingface.co/docs/transformers/model_doc/mgp-str)** (d'Alibaba Research) a été publié dans l'article [Multi-Granularity Prediction for Scene Text Recognition](https://arxiv.org/abs/2209.03592) par Peng Wang, Cheng Da et Cong Yao.
 1. **[Mistral](https://huggingface.co/docs/transformers/model_doc/mistral)** (de Mistral AI) par l'équipe [Mistral AI](https://mistral.ai) : Albert Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lélio Renard Lavaud, Lucile Saulnier, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix, William El Sayed.
 1. **[Mixtral](https://huggingface.co/docs/transformers/model_doc/mixtral)** (de Mistral AI) par l'équipe [Mistral AI](https://mistral.ai) : Albert Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lélio Renard Lavaud, Lucile Saulnier, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix, William El Sayed.
-1. **[mLUKE](https://huggingface.co/docs/transformers/model_doc/mluke)** (de Studio Ousia) a été publié dans l'article [mLUKE : La puissance des représentations d'entités dans les modèles linguistiques pré-entraînés multilingues](https://arxiv.org/abs/2110.08151) par Ryokan Ri, Ikuya Yamada et Yoshimasa Tsuruoka.
-1. **[MMS](https://huggingface.co/docs/transformers/model_doc/mms)** (de Facebook) a été publié dans l'article [Mise à l'échelle de la technologie de la parole à plus de 1 000 langues](https://arxiv.org/abs/2305.13516) par Vineel Pratap, Andros Tjandra, Bowen Shi, Paden Tomasello, Arun Babu, Sayani Kundu, Ali Elkahky, Zhaoheng Ni, Apoorv Vyas, Maryam Fazel-Zarandi, Alexei Baevski, Yossi Adi, Xiaohui Zhang, Wei-Ning Hsu, Alexis Conneau, Michael Auli.
-1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (de CMU/Google Brain) a été publié dans l'article [MobileBERT : un BERT compact et agnostique pour les tâches sur les appareils à ressources limitées](https://arxiv.org/abs/2004.02984) par Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang et Denny Zhou.
-1. **[MobileNetV1](https://huggingface.co/docs/transformers/model_doc/mobilenet_v1)** (de Google Inc.) a été publié dans l'article [MobileNets : Réseaux neuronaux convolutifs efficaces pour les applications de vision mobile](https://arxiv.org/abs/1704.04861) par Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam.
-1. **[MobileNetV2](https://huggingface.co/docs/transformers/model_doc/mobilenet_v2)** (de Google Inc.) a été publié dans l'article [MobileNetV2 : Résidus inversés et coudes linéaires](https://arxiv.org/abs/1801.04381) par Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen.
-1. **[MobileViT](https://huggingface.co/docs/transformers/model_doc/mobilevit)** (d'Apple) a été publié dans l'article [MobileViT : Vision Transformer léger, polyvalent et adapté aux mobiles](https://arxiv.org/abs/2110.02178) par Sachin Mehta et Mohammad Rastegari.
-1. **[MobileViTV2](https://huggingface.co/docs/transformers/model_doc/mobilevitv2)** (d'Apple) a été publié dans l'article [Auto-attention séparable pour les Vision Transformers mobiles](https://arxiv.org/abs/2206.02680) par Sachin Mehta et Mohammad Rastegari.
-1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (de Microsoft Research) a été publié dans l'article [MPNet : Pré-entraînement masqué et permuté pour la compréhension du langage](https://arxiv.org/abs/2004.09297) par Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
+1. **[mLUKE](https://huggingface.co/docs/transformers/model_doc/mluke)** (de Studio Ousia) a été publié dans l'article [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) par Ryokan Ri, Ikuya Yamada et Yoshimasa Tsuruoka.
+1. **[MMS](https://huggingface.co/docs/transformers/model_doc/mms)** (de Facebook) a été publié dans l'article [Scaling Speech Technology to 1,000+ Languages](https://arxiv.org/abs/2305.13516) par Vineel Pratap, Andros Tjandra, Bowen Shi, Paden Tomasello, Arun Babu, Sayani Kundu, Ali Elkahky, Zhaoheng Ni, Apoorv Vyas, Maryam Fazel-Zarandi, Alexei Baevski, Yossi Adi, Xiaohui Zhang, Wei-Ning Hsu, Alexis Conneau, Michael Auli.
+1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (de CMU/Google Brain) a été publié dans l'article [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) par Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang et Denny Zhou.
+1. **[MobileNetV1](https://huggingface.co/docs/transformers/model_doc/mobilenet_v1)** (de Google Inc.) a été publié dans l'article [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://arxiv.org/abs/1704.04861) par Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam.
+1. **[MobileNetV2](https://huggingface.co/docs/transformers/model_doc/mobilenet_v2)** (de Google Inc.) a été publié dans l'article [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381) par Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen.
+1. **[MobileViT](https://huggingface.co/docs/transformers/model_doc/mobilevit)** (d'Apple) a été publié dans l'article [MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer](https://arxiv.org/abs/2110.02178) par Sachin Mehta et Mohammad Rastegari.
+1. **[MobileViTV2](https://huggingface.co/docs/transformers/model_doc/mobilevitv2)** (d'Apple) a été publié dans l'article [Separable Self-attention for Mobile Vision Transformers](https://arxiv.org/abs/2206.02680) par Sachin Mehta et Mohammad Rastegari.
+1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (de Microsoft Research) a été publié dans l'article [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) par Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MPT](https://huggingface.co/docs/transformers/model_doc/mpt)** (de MosaiML) a été publié avec le référentiel [llm-foundry](https://github.com/mosaicml/llm-foundry/) par l'équipe MosaiML NLP.
-1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (de l'Université du Wisconsin - Madison) a été publié dans l'article [Analyse multi-résolution (MRA) pour une auto-attention approximative](https://arxiv.org/abs/2207.10284) par Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
-1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (de Google AI) a été publié dans l'article [mT5 : un transformateur texte-à-texte pré-entraîné massivement multilingue](https://arxiv.org/abs/2010.11934) par Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
-1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (de Meta) a été publié dans l'article [Génération de musique simple et contrôlable](https://arxiv.org/abs/2306.05284) par Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi et Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (de Meta) publié dans l'article [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) parJade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (de RUC AI Box) a été publié dans l'article [MVP : Pré-entraînement supervisé multi-tâche pour la génération de langage naturel](https://arxiv.org/abs/2206.12131) par Tianyi Tang, Junyi Li, Wayne Xin Zhao et Ji-Rong Wen.
-1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (de SHI Labs) a été publié dans l'article [Transformateur d'attention de voisinage](https://arxiv.org/abs/2204.07143) par Ali Hassani, Steven Walton, Jiachen Li, Shen Li et Humphrey Shi.
-1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (du laboratoire Noah's Ark de Huawei) a été publié dans l'article [NEZHA : Représentation contextualisée neurale pour la compréhension du langage chinois](https://arxiv.org/abs/1909.00204) par Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen et Qun Liu.
-1. **[NLLB](https://huggingface.co/docs/transformers/model_doc/nllb)** (de Meta) a été publié dans l'article [No Language Left Behind : Mise à l'échelle de la traduction automatique centrée sur l'humain](https://arxiv.org/abs/2207.04672) par l'équipe NLLB.
-1. **[NLLB-MOE](https://huggingface.co/docs/transformers/model_doc/nllb-moe)** (de Meta) a été publié dans l'article [No Language Left Behind : Mise à l'échelle de la traduction automatique centrée sur l'humain](https://arxiv.org/abs/2207.04672) par l'équipe NLLB.
-1. **[Nougat](https://huggingface.co/docs/transformers/model_doc/nougat)** (de Meta AI) a été publié dans l'article [Nougat : Compréhension Optique Neuronale pour les Documents Académiques](https://arxiv.org/abs/2308.13418) par Lukas Blecher, Guillem Cucurull, Thomas Scialom, Robert Stojnic.
-1. **[Nyströmformer](https://huggingface.co/docs/transformers/model_doc/nystromformer)** (de l'Université du Wisconsin - Madison) a été publié dans l'article [Nyströmformer : Un algorithme basé sur Nyström pour approximer l'auto-attention](https://arxiv.org/abs/2102.03902) par Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
-1. **[OneFormer](https://huggingface.co/docs/transformers/model_doc/oneformer)** (de SHI Labs) a été publié dans l'article [OneFormer : Un Transformer pour dominer la segmentation universelle d'images](https://arxiv.org/abs/2211.06220) par Jitesh Jain, Jiachen Li, MangTik Chiu, Ali Hassani, Nikita Orlov, Humphrey Shi.
+1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (de l'Université du Wisconsin - Madison) a été publié dans l'article [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) par Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
+1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (de Google AI) a été publié dans l'article [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) par Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
+1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (de Meta) a été publié dans l'article [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) par Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi et Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (de Meta) publié dans l'article [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) parJade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (de RUC AI Box) a été publié dans l'article [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) par Tianyi Tang, Junyi Li, Wayne Xin Zhao et Ji-Rong Wen.
+1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (de SHI Labs) a été publié dans l'article [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) par Ali Hassani, Steven Walton, Jiachen Li, Shen Li et Humphrey Shi.
+1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (du laboratoire Noah's Ark de Huawei) a été publié dans l'article [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) par Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen et Qun Liu.
+1. **[NLLB](https://huggingface.co/docs/transformers/model_doc/nllb)** (de Meta) a été publié dans l'article [No Language Left Behind: Scaling Human-Centered Machine Translation](https://arxiv.org/abs/2207.04672) par l'équipe NLLB.
+1. **[NLLB-MOE](https://huggingface.co/docs/transformers/model_doc/nllb-moe)** (de Meta) a été publié dans l'article [No Language Left Behind: Scaling Human-Centered Machine Translation](https://arxiv.org/abs/2207.04672) par l'équipe NLLB.
+1. **[Nougat](https://huggingface.co/docs/transformers/model_doc/nougat)** (de Meta AI) a été publié dans l'article [Nougat: Neural Optical Understanding for Academic Documents](https://arxiv.org/abs/2308.13418) par Lukas Blecher, Guillem Cucurull, Thomas Scialom, Robert Stojnic.
+1. **[Nyströmformer](https://huggingface.co/docs/transformers/model_doc/nystromformer)** (de l'Université du Wisconsin - Madison) a été publié dans l'article [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) par Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
+1. **[OneFormer](https://huggingface.co/docs/transformers/model_doc/oneformer)** (de SHI Labs) a été publié dans l'article [OneFormer: One Transformer to Rule Universal Image Segmentation](https://arxiv.org/abs/2211.06220) par Jitesh Jain, Jiachen Li, MangTik Chiu, Ali Hassani, Nikita Orlov, Humphrey Shi.
 1. **[OpenLlama](https://huggingface.co/docs/transformers/model_doc/open-llama)** (de [s-JoL](https://huggingface.co/s-JoL)) publié sur GitHub (maintenant supprimé).
-1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (de Meta AI) a été publié dans l'article [OPT : Modèles linguistiques Transformer pré-entraînés ouverts](https://arxiv.org/abs/2205.01068) par 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)** (de Google AI) a été publié dans l'article [OWL-ViT : Détection d'objets simple à vocabulaire ouvert avec des transformateurs de vision](https://arxiv.org/abs/2205.06230) par 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 et Neil Houlsby.
+1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (de Meta AI) a été publié dans l'article [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) par 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)** (de Google AI) a été publié dans l'article [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) par 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 et Neil Houlsby.
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (de Google AI) a été publié dans l'article [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683) par Matthias Minderer, Alexey Gritsenko, Neil Houlsby.
-1. **[PatchTSMixer](https://huggingface.co/docs/transformers/model_doc/patchtsmixer)** (d'IBM Research) a été publié dans l'article [TSMixer : Modèle MLP-Mixer léger pour la prévision multivariée de séries temporelles](https://arxiv.org/pdf/2306.09364.pdf) par Vijay Ekambaram, Arindam Jati, Nam Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
-1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (d'IBM) a été publié dans l'article [Une série temporelle vaut 64 mots : Prévision à long terme avec des Transformers](https://arxiv.org/abs/2211.14730) par Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
-1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (de Google) a été publié dans l'article [PEGASUS : Pré-entraînement avec Phrases-Écart Extraites pour la Résumé Abstrait](https://arxiv.org/abs/1912.08777) par Jingqing Zhang, Yao Zhao, Mohammad Saleh et Peter J. Liu.
-1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (de Google) a été publié dans l'article [Investiguer l'Extension Efficace des Transformers pour la Longue Summarization d'Entrée](https://arxiv.org/abs/2208.04347) par Jason Phang, Yao Zhao et Peter J. Liu.
-1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (de Deepmind) a été publié dans l'article [Perceiver IO : Une architecture générale pour les entrées et sorties structurées](https://arxiv.org/abs/2107.14795) par 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 et João Carreira.
-1. **[Persimmon](https://huggingface.co/docs/transformers/model_doc/persimmon)** (d'ADEPT) a été publié dans un [billet de blog](https://www.adept.ai/blog/persimmon-8b) par Erich Elsen, Augustus Odena, Maxwell Nye, Sağnak Taşırlar, Tri Dao, Curtis Hawthorne, Deepak Moparthi, Arushi Somani.
+1. **[PatchTSMixer](https://huggingface.co/docs/transformers/model_doc/patchtsmixer)** (d'IBM Research) a été publié dans l'article [TSMixer: Lightweight MLP-Mixer Model for Multivariate Time Series Forecasting](https://arxiv.org/pdf/2306.09364.pdf) par Vijay Ekambaram, Arindam Jati, Nam Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
+1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (d'IBM) a été publié dans l'article [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/abs/2211.14730) par Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
+1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (de Google) a été publié dans l'article [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) par Jingqing Zhang, Yao Zhao, Mohammad Saleh et Peter J. Liu.
+1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (de Google) a été publié dans l'article [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) par Jason Phang, Yao Zhao et Peter J. Liu.
+1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (de Deepmind) a été publié dans l'article [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) par 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 et João Carreira.
+1. **[Persimmon](https://huggingface.co/docs/transformers/model_doc/persimmon)** (d'ADEPT) a été publié dans un [blog post](https://www.adept.ai/blog/persimmon-8b) par Erich Elsen, Augustus Odena, Maxwell Nye, Sağnak Taşırlar, Tri Dao, Curtis Hawthorne, Deepak Moparthi, Arushi Somani.
 1. **[Phi](https://huggingface.co/docs/transformers/model_doc/phi)** (de Microsoft) a été publié avec les articles - [Textbooks Are All You Need](https://arxiv.org/abs/2306.11644) par Suriya Gunasekar, Yi Zhang, Jyoti Aneja, Caio César Teodoro Mendes, Allie Del Giorno, Sivakanth Gopi, Mojan Javaheripi, Piero Kauffmann, Gustavo de Rosa, Olli Saarikivi, Adil Salim, Shital Shah, Harkirat Singh Behl, Xin Wang, Sébastien Bubeck, Ronen Eldan, Adam Tauman Kalai, Yin Tat Lee et Yuanzhi Li, [Textbooks Are All You Need II : Rapport technique phi-1.5](https://arxiv.org/abs/2309.05463) par Yuanzhi Li, Sébastien Bubeck, Ronen Eldan, Allie Del Giorno, Suriya Gunasekar et Yin Tat Lee.
-1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (de VinAI Research) a été publié dans l'article [PhoBERT : Modèles linguistiques pré-entraînés pour le vietnamien](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) par Dat Quoc Nguyen et Anh Tuan Nguyen.
-1. **[Pix2Struct](https://huggingface.co/docs/transformers/model_doc/pix2struct)** (de Google) a été publié dans l'article [Pix2Struct : Analyse d'images d'écran en tant que pré-entraînement pour la compréhension du langage visuel](https://arxiv.org/abs/2210.03347) par Kenton Lee, Mandar Joshi, Iulia Turc, Hexiang Hu, Fangyu Liu, Julian Eisenschlos, Urvashi Khandelwal, Peter Shaw, Ming-Wei Chang, Kristina Toutanova.
+1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (de VinAI Research) a été publié dans l'article [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) par Dat Quoc Nguyen et Anh Tuan Nguyen.
+1. **[Pix2Struct](https://huggingface.co/docs/transformers/model_doc/pix2struct)** (de Google) a été publié dans l'article [Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding](https://arxiv.org/abs/2210.03347) par Kenton Lee, Mandar Joshi, Iulia Turc, Hexiang Hu, Fangyu Liu, Julian Eisenschlos, Urvashi Khandelwal, Peter Shaw, Ming-Wei Chang, Kristina Toutanova.
 1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (de UCLA NLP) a été publié dans l'article [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) par Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
 1. **[PoolFormer](https://huggingface.co/docs/transformers/model_doc/poolformer)** (de Sea AI Labs) a été publié dans l'article [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) par Yu, Weihao et Luo, Mi et Zhou, Pan et Si, Chenyang et Zhou, Yichen et Wang, Xinchao et Feng, Jiashi et Yan, Shuicheng.
-1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** a été publié dans l'article [Pop2Piano : Génération de reprises de morceaux de piano basée sur l'audio pop](https://arxiv.org/abs/2211.00895) par Jongho Choi et Kyogu Lee.
-1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/prophetnet)** (de Microsoft Research) a été publié dans l'article [ProphetNet : Prédire les N-grammes futurs pour l'entraînement préalable de séquences à séquences](https://arxiv.org/abs/2001.04063) par Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang et Ming Zhou.
-1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (de l'Université de Nankin, l'Université de Hong Kong, etc.) a été publié dans l'article [Pyramid Vision Transformer : Une colonne vertébrale polyvalente pour la prédiction dense sans convolutions](https://arxiv.org/pdf/2102.12122.pdf) par Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo et Ling Shao.
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (de Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) publié dans l'article [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) parWenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
-1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (de NVIDIA) a été publié dans l'article [Quantification entière pour l'inférence d'apprentissage profond : Principes et évaluation empirique](https://arxiv.org/abs/2004.09602) par Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev et Paulius Micikevicius.
-1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (de l'équipe Qwen, Alibaba Group) a été publié avec le rapport technique [Qwen](https://arxiv.org/abs/2309.16609) par Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou et Tianhang Zhu.
+1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** a été publié dans l'article [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) par Jongho Choi et Kyogu Lee.
+1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/prophetnet)** (de Microsoft Research) a été publié dans l'article [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) par Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang et Ming Zhou.
+1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (de l'Université de Nankin, l'Université de Hong Kong, etc.) a été publié dans l'article [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) par Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo et Ling Shao.
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (de Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) publié dans l'article [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) parWenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
+1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (de NVIDIA) a été publié dans l'article [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) par Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev et Paulius Micikevicius.
+1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (de l'équipe Qwen, Alibaba Group) a été publié avec le rapport technique [Qwen Technical Report](https://arxiv.org/abs/2309.16609) par Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou et Tianhang Zhu.
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (de Facebook) a été publié dans l'article [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) par Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
-1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (de Google Research) a été publié dans l'article [REALM : Pré-entraînement de modèle linguistique augmenté par la récupération](https://arxiv.org/abs/2002.08909) par Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat et Ming-Wei Chang.
-1. **[Reformer](https://huggingface.co/docs/transformers/model_doc/reformer)** (de Google Research) a été publié dans l'article [Reformer : Le transformateur efficace](https://arxiv.org/abs/2001.04451) par Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya.
+1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (de Google Research) a été publié dans l'article [REALM: Retrieval-Augmented Language Model Pre-Training](https://arxiv.org/abs/2002.08909) par Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat et Ming-Wei Chang.
+1. **[Reformer](https://huggingface.co/docs/transformers/model_doc/reformer)** (de Google Research) a été publié dans l'article [Reformer: The Efficient Transformer](https://arxiv.org/abs/2001.04451) par Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya.
 1. **[RegNet](https://huggingface.co/docs/transformers/model_doc/regnet)** (de META Platforms) a été publié dans l'article [Designing Network Design Space](https://arxiv.org/abs/2003.13678) par Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár.
 1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (de Google Research) a été publié dans l'article [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/abs/2010.12821) par Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder.
 1. **[ResNet](https://huggingface.co/docs/transformers/model_doc/resnet)** (de Microsoft Research) a été publié dans l'article [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) par Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun.
-1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (de Facebook), publié dans l'article [RoBERTa : Une approche d'entraînement préalable BERT robuste](https://arxiv.org/abs/1907.11692) par Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov.
-1. **[RoBERTa-PreLayerNorm](https://huggingface.co/docs/transformers/model_doc/roberta-prelayernorm)** (de Facebook) a été publié dans l'article [fairseq : Une boîte à outils rapide et extensible pour la modélisation de séquences](https://arxiv.org/abs/1904.01038) par Myle Ott, Sergey Edunov, Alexei Baevski, Angela Fan, Sam Gross, Nathan Ng, David Grangier, Michael Auli.
-1. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (de WeChatAI) a été publié dans l'article [RoCBert : BERT chinois robuste avec pré-entraînement contrastif multimodal](https://aclanthology.org/2022.acl-long.65.pdf) par HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou.
-1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (de ZhuiyiTechnology), publié dans l'article [RoFormer : Transformateur amélioré avec insertion rotative de position](https://arxiv.org/abs/2104.09864) par Jianlin Su et Yu Lu et Shengfeng Pan et Bo Wen et Yunfeng Liu.
-1. **[RWKV](https://huggingface.co/docs/transformers/model_doc/rwkv)** (de Bo Peng), publié sur [ce référentiel](https://github.com/BlinkDL/RWKV-LM) par Bo Peng.
-1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (de Meta AI) a été publié dans l'article [SeamlessM4T — Traduction multimodale et massivement multilingue](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) par l'équipe de communication transparente.
-1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (de Meta AI) a été publié dans l'article [Seamless: Traduction de la parole multilingue, expressive et en continu](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) par l'équipe de communication transparente.
-1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (de NVIDIA) a été publié dans l'article [SegFormer : Conception simple et efficace pour la segmentation sémantique avec des transformateurs](https://arxiv.org/abs/2105.15203) par Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (de Beijing Academy of Artificial Intelligence (BAAI) publié dans l'article [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) parXinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
+1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (de Facebook), publié dans l'article [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://arxiv.org/abs/1907.11692) par Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov.
+1. **[RoBERTa-PreLayerNorm](https://huggingface.co/docs/transformers/model_doc/roberta-prelayernorm)** (de Facebook) a été publié dans l'article [fairseq: A Fast, Extensible Toolkit for Sequence Modeling](https://arxiv.org/abs/1904.01038) par Myle Ott, Sergey Edunov, Alexei Baevski, Angela Fan, Sam Gross, Nathan Ng, David Grangier, Michael Auli.
+1. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (de WeChatAI) a été publié dans l'article [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) par HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou.
+1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (de ZhuiyiTechnology), publié dans l'article [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/abs/2104.09864) par Jianlin Su et Yu Lu et Shengfeng Pan et Bo Wen et Yunfeng Liu.
+1. **[RWKV](https://huggingface.co/docs/transformers/model_doc/rwkv)** (de Bo Peng), publié sur [this repo](https://github.com/BlinkDL/RWKV-LM) par Bo Peng.
+1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (de Meta AI) a été publié dans l'article [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) par l'équipe de communication transparente.
+1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (de Meta AI) a été publié dans l'article [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) par l'équipe de communication transparente.
+1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (de NVIDIA) a été publié dans l'article [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) par Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (de Beijing Academy of Artificial Intelligence (BAAI) publié dans l'article [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) parXinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (de Meta AI) a été publié dans l'article [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) par Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.
-1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (de ASAPP) a été publié dans l'article [Compromis entre performances et efficacité dans l'entraînement non supervisé pour la reconnaissance vocale](https://arxiv.org/abs/2109.06870) par Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
-1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (de ASAPP) a été publié dans l'article [Compromis entre performances et efficacité dans l'entraînement non supervisé pour la reconnaissance vocale](https://arxiv.org/abs/2109.06870) par Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
+1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (de ASAPP) a été publié dans l'article [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) par Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
+1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (de ASAPP) a été publié dans l'article [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) par Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
 1. **[SigLIP](https://huggingface.co/docs/transformers/model_doc/siglip)** (de Google AI) a été publié dans l'article [Sigmoid Loss for Language Image Pre-Training](https://arxiv.org/abs/2303.15343) par Xiaohua Zhai, Basil Mustafa, Alexander Kolesnikov, Lucas Beyer.
-1. **[SpeechT5](https://huggingface.co/docs/transformers/model_doc/speecht5)** (de Microsoft Research) a été publié dans l'article [SpeechT5 : Pré-entraînement unifié encodeur-décodeur pour le traitement du langage parlé](https://arxiv.org/abs/2110.07205) par Junyi Ao, Rui Wang, Long Zhou, Chengyi Wang, Shuo Ren, Yu Wu, Shujie Liu, Tom Ko, Qing Li, Yu Zhang, Zhihua Wei, Yao Qian, Jinyu Li, Furu Wei.
-1. **[SpeechToTextTransformer](https://huggingface.co/docs/transformers/model_doc/speech_to_text)** (de Facebook), publié dans l'article [fairseq S2T : Modélisation rapide de la parole à texte avec fairseq](https://arxiv.org/abs/2010.05171) par Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Dmytro Okhonko, Juan Pino.
-1. **[SpeechToTextTransformer2](https://huggingface.co/docs/transformers/model_doc/speech_to_text_2)** (de Facebook), publié dans l'article [Apprentissage auto-supervisé et semi-supervisé à grande échelle pour la traduction de la parole](https://arxiv.org/abs/2104.06678) par Changhan Wang, Anne Wu, Juan Pino, Alexei Baevski, Michael Auli, Alexis Conneau.
-1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (de l'Université de Tel Aviv), publié dans l'article [Réponse à quelques questions avec peu d'exemples par la pré-sélection des spans](https://arxiv.org/abs/2101.00438) par Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy.
-1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (de Berkeley) a été publié dans l'article [SqueezeBERT : Que l'apprentissage automatique peut-il apprendre au traitement du langage naturel sur les réseaux neuronaux efficaces ?](https://arxiv.org/abs/2006.11316) par Forrest N. Iandola, Albert E. Shaw, Ravi Krishna et Kurt W. Keutzer.
-1. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report)](https://stability.wandb.io/stability-llm/stable-lm/reports/StableLM-3B-4E1T--VmlldzoyMjU4?accessToken=u3zujipenkx5g7rtcj9qojjgxpconyjktjkli2po09nffrffdhhchq045vp0wyfo) by  Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu. 
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
-1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (de MBZUAI) a été publié dans l'article [SwiftFormer : Attention additive efficace pour les applications de vision mobile en temps réel basées sur des transformateurs](https://arxiv.org/abs/2303.15446) par Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.
-1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (de Microsoft) a été publié dans l'article [Swin Transformer : Transformateur hiérarchique de la vision utilisant des fenêtres décalées](https://arxiv.org/abs/2103.14030) par 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/model_doc/swinv2)** (de Microsoft) a été publié dans l'article [Swin Transformer V2 : Augmentation de la capacité et de la résolution](https://arxiv.org/abs/2111.09883) par 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. **[Swin2SR](https://huggingface.co/docs/transformers/model_doc/swin2sr)** (de l'Université de Würzburg) a été publié dans l'article [Swin2SR : Transformateur SwinV2 pour la super-résolution et la restauration d'images compressées](https://arxiv.org/abs/2209.11345) par Marcos V. Conde, Ui-Jin Choi, Maxime Burchi, Radu Timofte.
-1. **[SwitchTransformers](https://huggingface.co/docs/transformers/model_doc/switch_transformers)** (de Google) a été publié dans l'article [Switch Transformers : Passage à des modèles de trillions de paramètres avec une parcimonie simple et efficace](https://arxiv.org/abs/2101.03961) par William Fedus, Barret Zoph, Noam Shazeer.
-1. **[T5](https://huggingface.co/docs/transformers/model_doc/t5)** (de Google AI) a été publié dans l'article [Exploration des limites de l'apprentissage par transfert avec un transformateur de texte à texte unifié](https://arxiv.org/abs/1910.10683) par Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li et Peter J. Liu.
+1. **[SpeechT5](https://huggingface.co/docs/transformers/model_doc/speecht5)** (de Microsoft Research) a été publié dans l'article [SpeechT5: Unified-Modal Encoder-Decoder Pre-Training for Spoken Language Processing](https://arxiv.org/abs/2110.07205) par Junyi Ao, Rui Wang, Long Zhou, Chengyi Wang, Shuo Ren, Yu Wu, Shujie Liu, Tom Ko, Qing Li, Yu Zhang, Zhihua Wei, Yao Qian, Jinyu Li, Furu Wei.
+1. **[SpeechToTextTransformer](https://huggingface.co/docs/transformers/model_doc/speech_to_text)** (de Facebook), publié dans l'article [fairseq S2T: Fast Speech-to-Text Modeling with fairseq](https://arxiv.org/abs/2010.05171) par Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Dmytro Okhonko, Juan Pino.
+1. **[SpeechToTextTransformer2](https://huggingface.co/docs/transformers/model_doc/speech_to_text_2)** (de Facebook), publié dans l'article [Large-Scale Self- and Semi-Supervised Learning for Speech Translation](https://arxiv.org/abs/2104.06678) par Changhan Wang, Anne Wu, Juan Pino, Alexei Baevski, Michael Auli, Alexis Conneau.
+1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (de l'Université de Tel Aviv), publié dans l'article [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) par Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy.
+1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (de Berkeley) a été publié dans l'article [SqueezeBERT: What can computer vision teach NLP about efficient neural networks?](https://arxiv.org/abs/2006.11316) par Forrest N. Iandola, Albert E. Shaw, Ravi Krishna et Kurt W. Keutzer.
+1. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report)](https://stability.wandb.io/stability-llm/stable-lm/reports/StableLM-3B-4E1T--VmlldzoyMjU4?accessToken=u3zujipenkx5g7rtcj9qojjgxpconyjktjkli2po09nffrffdhhchq045vp0wyfo) by  Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (de MagicLeap) publié dans l'article [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) parDaniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
+1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (de MBZUAI) a été publié dans l'article [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) par Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.
+1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (de Microsoft) a été publié dans l'article [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) par 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/model_doc/swinv2)** (de Microsoft) a été publié dans l'article [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) par 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. **[Swin2SR](https://huggingface.co/docs/transformers/model_doc/swin2sr)** (de l'Université de Würzburg) a été publié dans l'article [Swin2SR: SwinV2 Transformer for Compressed Image Super-Resolution and Restoration](https://arxiv.org/abs/2209.11345) par Marcos V. Conde, Ui-Jin Choi, Maxime Burchi, Radu Timofte.
+1. **[SwitchTransformers](https://huggingface.co/docs/transformers/model_doc/switch_transformers)** (de Google) a été publié dans l'article [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://arxiv.org/abs/2101.03961) par William Fedus, Barret Zoph, Noam Shazeer.
+1. **[T5](https://huggingface.co/docs/transformers/model_doc/t5)** (de Google AI) a été publié dans l'article [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/abs/1910.10683) par Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li et Peter J. Liu.
 1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (de Google AI) a été publié dans le dépôt [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) par Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li et Peter J. Liu.
-1. **[Table Transformer](https://huggingface.co/docs/transformers/model_doc/table-transformer)** (de Microsoft Research) a été publié dans l'article [PubTables-1M : Vers une extraction complète des tables à partir de documents non structurés](https://arxiv.org/abs/2110.00061) par Brandon Smock, Rohith Pesala, Robin Abraham.
-1. **[TAPAS](https://huggingface.co/docs/transformers/model_doc/tapas)** (de Google AI) a été publié dans l'article [TAPAS : Analyse faiblement supervisée des tables via le pré-entraînement](https://arxiv.org/abs/2004.02349) par Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Müller, Francesco Piccinno et Julian Martin Eisenschlos.
-1. **[TAPEX](https://huggingface.co/docs/transformers/model_doc/tapex)** (de Microsoft Research) a été publié dans l'article [TAPEX : Pré-entraînement des tables via l'apprentissage d'un exécuteur SQL neuronal](https://arxiv.org/abs/2107.07653) par Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen et Jian-Guang Lou.
+1. **[Table Transformer](https://huggingface.co/docs/transformers/model_doc/table-transformer)** (de Microsoft Research) a été publié dans l'article [PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents](https://arxiv.org/abs/2110.00061) par Brandon Smock, Rohith Pesala, Robin Abraham.
+1. **[TAPAS](https://huggingface.co/docs/transformers/model_doc/tapas)** (de Google AI) a été publié dans l'article [TAPAS: Weakly Supervised Table Parsing via Pre-training](https://arxiv.org/abs/2004.02349) par Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Müller, Francesco Piccinno et Julian Martin Eisenschlos.
+1. **[TAPEX](https://huggingface.co/docs/transformers/model_doc/tapex)** (de Microsoft Research) a été publié dans l'article [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) par Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen et Jian-Guang Lou.
 1. **[Time Series Transformer](https://huggingface.co/docs/transformers/model_doc/time_series_transformer)** (de HuggingFace).
 1. **[TimeSformer](https://huggingface.co/docs/transformers/model_doc/timesformer)** (de Facebook) a été publié dans l'article [Is Space-Time Attention All You Need for Video Understanding?](https://arxiv.org/abs/2102.05095) par Gedas Bertasius, Heng Wang, Lorenzo Torresani.
-1. **[Trajectory Transformer](https://huggingface.co/docs/transformers/model_doc/trajectory_transformers)** (de l'Université de Californie à Berkeley) a été publié dans l'article [L'apprentissage par renforcement hors ligne comme un grand problème de modélisation de séquence](https://arxiv.org/abs/2106.02039) par Michael Janner, Qiyang Li, Sergey Levine.
-1. **[Transformer-XL](https://huggingface.co/docs/transformers/model_doc/transfo-xl)** (de Google/CMU) a été publié dans l'article [Transformer-XL : Modèles de langage attentifs au-delà d'un contexte de longueur fixe](https://arxiv.org/abs/1901.02860) par Zihang Dai*, Zhilin Yang*, Yiming Yang, Jaime Carbonell, Quoc V. Le, Ruslan Salakhutdinov.
-1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (de Microsoft), publié dans l'article [TrOCR : Reconnaissance optique de caractères basée sur un transformateur avec des modèles pré-entraînés](https://arxiv.org/abs/2109.10282) par Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
-1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (de l'UNC Chapel Hill) a été publié dans l'article [TVLT : Transformer Vision-Language sans texte](https://arxiv.org/abs/2209.14156) par Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
+1. **[Trajectory Transformer](https://huggingface.co/docs/transformers/model_doc/trajectory_transformers)** (de l'Université de Californie à Berkeley) a été publié dans l'article [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://arxiv.org/abs/2106.02039) par Michael Janner, Qiyang Li, Sergey Levine.
+1. **[Transformer-XL](https://huggingface.co/docs/transformers/model_doc/transfo-xl)** (de Google/CMU) a été publié dans l'article [Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context](https://arxiv.org/abs/1901.02860) par Zihang Dai*, Zhilin Yang*, Yiming Yang, Jaime Carbonell, Quoc V. Le, Ruslan Salakhutdinov.
+1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (de Microsoft), publié dans l'article [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) par Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
+1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (de l'UNC Chapel Hill) a été publié dans l'article [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) par Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (d'Intel) a été publié dans l'article [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) par Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (de Microsoft Research) publié dans l'article [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) parZineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (de Microsoft Research) publié dans l'article [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) parZineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (de Google Research) a été publié dans l'article [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) par Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler.
-1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (de Google Research) a été publié dans l'article [UniMax : Échantillonnage linguistique plus équitable et plus efficace pour l'entraînement préalable multilingue à grande échelle](https://openreview.net/forum?id=kXwdL1cWOAi) par Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.
-1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (de Microsoft Research) a été publié dans l'article [UniSpeech : Apprentissage unifié de la représentation de la parole avec des données étiquetées et non étiquetées](https://arxiv.org/abs/2101.07597) par 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)** (de Microsoft Research) a été publié dans l'article [UNISPEECH-SAT : Apprentissage universel de la représentation de la parole avec la préformation consciente du locuteur](https://arxiv.org/abs/2110.05752) par Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu.
-1. **[UnivNet](https://huggingface.co/docs/transformers/model_doc/univnet)** (de Kakao Corporation) a été publié dans l'article [UnivNet : un vocodeur neuronal avec des discriminateurs de spectrogramme multi-résolution pour la génération de formes d'onde haute fidélité](https://arxiv.org/abs/2106.07889) par Won Jang, Dan Lim, Jaesam Yoon, Bongwan Kim et Juntae Kim.
-1. **[UPerNet](https://huggingface.co/docs/transformers/model_doc/upernet)** (de l'Université de Pékin) a été publié dans l'article [Analyse perceptuelle unifiée pour la compréhension de scènes](https://arxiv.org/abs/1807.10221) par Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, Jian Sun.
+1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (de Google Research) a été publié dans l'article [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) par Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.
+1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (de Microsoft Research) a été publié dans l'article [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) par 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)** (de Microsoft Research) a été publié dans l'article [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) par Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu.
+1. **[UnivNet](https://huggingface.co/docs/transformers/model_doc/univnet)** (de Kakao Corporation) a été publié dans l'article [UnivNet: A Neural Vocoder with Multi-Resolution Spectrogram Discriminators for High-Fidelity Waveform Generation](https://arxiv.org/abs/2106.07889) par Won Jang, Dan Lim, Jaesam Yoon, Bongwan Kim et Juntae Kim.
+1. **[UPerNet](https://huggingface.co/docs/transformers/model_doc/upernet)** (de l'Université de Pékin) a été publié dans l'article [Unified Perceptual Parsing for Scene Understanding](https://arxiv.org/abs/1807.10221) par Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, Jian Sun.
 1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (de l'Université Tsinghua et de l'Université Nankai) publié dans l'article [Visual Attention Network](https://arxiv.org/abs/2202.09741) par Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
-1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (du groupe d'informatique multimédia, Université de Nankin) publié dans l'article [VideoMAE : Les autoencodeurs masqués sont des apprenants efficaces en données pour l'auto-pré-entraînement vidéo supervisé](https://arxiv.org/abs/2203.12602) par Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
-1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (du NAVER AI Lab/Kakao Enterprise/Kakao Brain) publié dans l'article [ViLT : Vision-and-Language Transformer sans convolution ni supervision de région](https://arxiv.org/abs/2102.03334) par Wonjae Kim, Bokyung Son, Ildoo Kim.
-1. **[VipLlava](https://huggingface.co/docs/transformers/model_doc/vipllava)** (de l'Université du Wisconsin–Madison) publié dans l'article [Rendre les grands modèles multimodaux comprenant des incitations visuelles arbitraires](https://arxiv.org/abs/2312.00784) par Mu Cai, Haotian Liu, Siva Karthik Mustikovela, Gregory P. Meyer, Yuning Chai, Dennis Park, Yong Jae Lee.
-1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/model_doc/vit)** (de Google AI) publié dans l'article [Une image vaut 16x16 mots : Transformers pour la reconnaissance d'images à grande échelle](https://arxiv.org/abs/2010.11929) par 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)** (de UCLA NLP) publié dans l'article [VisualBERT : Une référence simple et performante pour la vision et le langage](https://arxiv.org/pdf/1908.03557) par Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang.
-1. **[ViT Hybrid](https://huggingface.co/docs/transformers/model_doc/vit_hybrid)** (de Google AI) publié dans l'article [Une image vaut 16x16 mots : Transformers pour la reconnaissance d'images à grande échelle](https://arxiv.org/abs/2010.11929) par 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. **[VitDet](https://huggingface.co/docs/transformers/model_doc/vitdet)** (de Meta AI) publié dans l'article [Exploration des transformateurs de vision plain pour la détection d'objets](https://arxiv.org/abs/2203.16527) par Yanghao Li, Hanzi Mao, Ross Girshick, Kaiming He.
-1. **[ViTMAE](https://huggingface.co/docs/transformers/model_doc/vit_mae)** (de Meta AI) publié dans l'article [Les autoencodeurs masqués sont des apprenants évolutifs de la vision](https://arxiv.org/abs/2111.06377) par Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick.
-1. **[ViTMatte](https://huggingface.co/docs/transformers/model_doc/vitmatte)** (de HUST-VL) publié dans l'article [ViTMatte : Renforcer le détourage d'image avec des transformateurs de vision plain pré-entraînés](https://arxiv.org/abs/2305.15272) par Jingfeng Yao, Xinggang Wang, Shusheng Yang, Baoyuan Wang.
+1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (du groupe d'informatique multimédia, Université de Nankin) publié dans l'article [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) par Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
+1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (du NAVER AI Lab/Kakao Enterprise/Kakao Brain) publié dans l'article [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) par Wonjae Kim, Bokyung Son, Ildoo Kim.
+1. **[VipLlava](https://huggingface.co/docs/transformers/model_doc/vipllava)** (de l'Université du Wisconsin–Madison) publié dans l'article [Making Large Multimodal Models Understand Arbitrary Visual Prompts](https://arxiv.org/abs/2312.00784) par Mu Cai, Haotian Liu, Siva Karthik Mustikovela, Gregory P. Meyer, Yuning Chai, Dennis Park, Yong Jae Lee.
+1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/model_doc/vit)** (de Google AI) publié dans l'article [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://arxiv.org/abs/2010.11929) par 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)** (de UCLA NLP) publié dans l'article [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) par Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang.
+1. **[ViT Hybrid](https://huggingface.co/docs/transformers/model_doc/vit_hybrid)** (de Google AI) publié dans l'article [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://arxiv.org/abs/2010.11929) par 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. **[VitDet](https://huggingface.co/docs/transformers/model_doc/vitdet)** (de Meta AI) publié dans l'article [Exploring Plain Vision Transformer Backbones for Object Detection](https://arxiv.org/abs/2203.16527) par Yanghao Li, Hanzi Mao, Ross Girshick, Kaiming He.
+1. **[ViTMAE](https://huggingface.co/docs/transformers/model_doc/vit_mae)** (de Meta AI) publié dans l'article [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) par Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick.
+1. **[ViTMatte](https://huggingface.co/docs/transformers/model_doc/vitmatte)** (de HUST-VL) publié dans l'article [ViTMatte: Boosting Image Matting with Pretrained Plain Vision Transformers](https://arxiv.org/abs/2305.15272) par Jingfeng Yao, Xinggang Wang, Shusheng Yang, Baoyuan Wang.
 1. **[ViTMSN](https://huggingface.co/docs/transformers/model_doc/vit_msn)** (de Meta AI) publié dans l'article [Masked Siamese Networks for Label-Efficient Learning](https://arxiv.org/abs/2204.07141) par Mahmoud Assran, Mathilde Caron, Ishan Misra, Piotr Bojanowski, Florian Bordes, Pascal Vincent, Armand Joulin, Michael Rabbat, Nicolas Ballas.
-1. **[VITS](https://huggingface.co/docs/transformers/model_doc/vits)** (de Kakao Enterprise) publié dans l'article [Auto-encodeur variationnel conditionnel avec apprentissage adversarial pour la conversion texte-parole de bout en bout](https://arxiv.org/abs/2106.06103) par Jaehyeon Kim, Jungil Kong, Juhee Son.
-1. **[ViViT](https://huggingface.co/docs/transformers/model_doc/vivit)** (de Google Research) publié dans l'article [ViViT : Un transformateur de vision vidéo](https://arxiv.org/abs/2103.15691) par Anurag Arnab, Mostafa Dehghani, Georg Heigold, Chen Sun, Mario Lučić, Cordelia Schmid.
-1. **[Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/wav2vec2)** (de Facebook AI) publié dans l'article [wav2vec 2.0 : Un cadre pour l'apprentissage auto-supervisé des représentations de la parole](https://arxiv.org/abs/2006.11477) par Alexei Baevski, Henry Zhou, Abdelrahman Mohamed, Michael Auli.
-1. **[Wav2Vec2-BERT](https://huggingface.co/docs/transformers/model_doc/wav2vec2-bert)** (de Meta AI) publié dans l'article [Seamless : Traduction de la parole multilingue, expressive et en continu](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) par l'équipe Seamless Communication.
-1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (de Facebook AI) a été publié dans l'article [FAIRSEQ S2T : Modélisation rapide de la parole au texte avec FAIRSEQ](https://arxiv.org/abs/2010.05171) par 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)** (de Facebook AI) a été publié dans l'article [Reconnaissance de phonèmes interlingues simple et efficace sans apprentissage préalable](https://arxiv.org/abs/2109.11680) par Qiantong Xu, Alexei Baevski, Michael Auli.
-1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (de Microsoft Research) a été publié dans l'article [WavLM : Pré-entraînement auto-supervisé à grande échelle pour le traitement complet de la parole](https://arxiv.org/abs/2110.13900) par 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. **[Whisper](https://huggingface.co/docs/transformers/model_doc/whisper)** (d'OpenAI) a été publié dans l'article [Reconnaissance robuste de la parole via une supervision faible à grande échelle](https://cdn.openai.com/papers/whisper.pdf) par Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, Ilya Sutskever.
-1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (de Microsoft Research) a été publié dans l'article [Expansion des modèles pré-entraînés langage-image pour la reconnaissance vidéo générale](https://arxiv.org/abs/2208.02816) par Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
-1. **[X-MOD](https://huggingface.co/docs/transformers/model_doc/xmod)** (de Meta AI) a été publié dans l'article [Lever le sort de la multilinguité par le pré-entraînement des transformateurs modulaires](http://dx.doi.org/10.18653/v1/2022.naacl-main.255) par Jonas Pfeiffer, Naman Goyal, Xi Lin, Xian Li, James Cross, Sebastian Riedel, Mikel Artetxe.
-1. **[XGLM](https://huggingface.co/docs/transformers/model_doc/xglm)** (de Facebook AI) a été publié dans l'article [Apprentissage à quelques échantillons avec des modèles de langues multilingues](https://arxiv.org/abs/2112.10668) par 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)** (de Facebook) a été publié dans l'article [Pré-entraînement de modèles linguistiques multilingues](https://arxiv.org/abs/1901.07291) par Guillaume Lample et Alexis Conneau.
-1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (de Microsoft Research) a été publié dans l'article [ProphetNet : Prédire l'avenir N-gramme pour le pré-entraînement séquence-séquence](https://arxiv.org/abs/2001.04063) par Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang et Ming Zhou.
-1. **[XLM-RoBERTa](https://huggingface.co/docs/transformers/model_doc/xlm-roberta)** (de Facebook AI), publié dans l'article [Apprentissage non supervisé de la représentation croisée-lingue à grande échelle](https://arxiv.org/abs/1911.02116) par Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer et Veselin Stoyanov.
-1. **[XLM-RoBERTa-XL](https://huggingface.co/docs/transformers/model_doc/xlm-roberta-xl)** (de Facebook AI), publié dans l'article [Transformateurs à plus grande échelle pour la modélisation du langage masqué multilingue](https://arxiv.org/abs/2105.00572) par Naman Goyal, Jingfei Du, Myle Ott, Giri Anantharaman, Alexis Conneau.
-1. **[XLM-V](https://huggingface.co/docs/transformers/model_doc/xlm-v)** (de Meta AI) a été publié dans l'article [XLM-V : Surmonter le goulot d'étranglement du vocabulaire dans les modèles de langage masqués multilingues](https://arxiv.org/abs/2301.10472) par Davis Liang, Hila Gonen, Yuning Mao, Rui Hou, Naman Goyal, Marjan Ghazvininejad, Luke Zettlemoyer, Madian Khabsa.
-1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (de Google/CMU) a été publié dans l'article [XLNet : Préentraînement autorégressif généralisé pour la compréhension du langage](https://arxiv.org/abs/1906.08237) par Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
-1. **[XLS-R](https://huggingface.co/docs/transformers/model_doc/xls_r)** (de Facebook AI) publié dans l'article [XLS-R : Apprentissage d'une représentation de la parole autonome et multilingue à grande échelle](https://arxiv.org/abs/2111.09296) par Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli.
-1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (de Facebook AI) publié dans l'article [Apprentissage non supervisé de représentations multilingues pour la reconnaissance vocale](https://arxiv.org/abs/2006.13979) par Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
-1. **[YOLOS](https://huggingface.co/docs/transformers/model_doc/yolos)** (de l'Université Huazhong des sciences et technologies) publié dans l'article [You Only Look at One Sequence : Repenser le Transformer dans la vision à travers la détection d'objets](https://arxiv.org/abs/2106.00666) par Yuxin Fang, Bencheng Liao, Xinggang Wang, Jiemin Fang, Jiyang Qi, Rui Wu, Jianwei Niu, Wenyu Liu.
-1. **[YOSO](https://huggingface.co/docs/transformers/model_doc/yoso)** (de l'Université du Wisconsin - Madison) publié dans l'article [You Only Sample (Almost) Once : Coût linéaire Self-Attention via l'échantillonnage Bernoulli](https://arxiv.org/abs/2111.09714) par Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh.
+1. **[VITS](https://huggingface.co/docs/transformers/model_doc/vits)** (de Kakao Enterprise) publié dans l'article [Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech](https://arxiv.org/abs/2106.06103) par Jaehyeon Kim, Jungil Kong, Juhee Son.
+1. **[ViViT](https://huggingface.co/docs/transformers/model_doc/vivit)** (de Google Research) publié dans l'article [ViViT: A Video Vision Transformer](https://arxiv.org/abs/2103.15691) par Anurag Arnab, Mostafa Dehghani, Georg Heigold, Chen Sun, Mario Lučić, Cordelia Schmid.
+1. **[Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/wav2vec2)** (de Facebook AI) publié dans l'article [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://arxiv.org/abs/2006.11477) par Alexei Baevski, Henry Zhou, Abdelrahman Mohamed, Michael Auli.
+1. **[Wav2Vec2-BERT](https://huggingface.co/docs/transformers/model_doc/wav2vec2-bert)** (de Meta AI) publié dans l'article [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) par l'équipe Seamless Communication.
+1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (de Facebook AI) a été publié dans l'article [FAIRSEQ S2T: Fast Speech-to-Text Modeling with FAIRSEQ](https://arxiv.org/abs/2010.05171) par 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)** (de Facebook AI) a été publié dans l'article [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) par Qiantong Xu, Alexei Baevski, Michael Auli.
+1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (de Microsoft Research) a été publié dans l'article [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) par 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. **[Whisper](https://huggingface.co/docs/transformers/model_doc/whisper)** (d'OpenAI) a été publié dans l'article [Robust Speech Recognition via Large-Scale Weak Supervision](https://cdn.openai.com/papers/whisper.pdf) par Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, Ilya Sutskever.
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (de Microsoft Research) a été publié dans l'article [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) par Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
+1. **[X-MOD](https://huggingface.co/docs/transformers/model_doc/xmod)** (de Meta AI) a été publié dans l'article [Lifting the Curse of Multilinguality by Pre-training Modular Transformers](http://dx.doi.org/10.18653/v1/2022.naacl-main.255) par Jonas Pfeiffer, Naman Goyal, Xi Lin, Xian Li, James Cross, Sebastian Riedel, Mikel Artetxe.
+1. **[XGLM](https://huggingface.co/docs/transformers/model_doc/xglm)** (de Facebook AI) a été publié dans l'article [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) par 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)** (de Facebook) a été publié dans l'article [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) par Guillaume Lample et Alexis Conneau.
+1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (de Microsoft Research) a été publié dans l'article [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) par Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang et Ming Zhou.
+1. **[XLM-RoBERTa](https://huggingface.co/docs/transformers/model_doc/xlm-roberta)** (de Facebook AI), publié dans l'article [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) par Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer et Veselin Stoyanov.
+1. **[XLM-RoBERTa-XL](https://huggingface.co/docs/transformers/model_doc/xlm-roberta-xl)** (de Facebook AI), publié dans l'article [Larger-Scale Transformers for Multilingual Masked Language Modeling](https://arxiv.org/abs/2105.00572) par Naman Goyal, Jingfei Du, Myle Ott, Giri Anantharaman, Alexis Conneau.
+1. **[XLM-V](https://huggingface.co/docs/transformers/model_doc/xlm-v)** (de Meta AI) a été publié dans l'article [XLM-V: Overcoming the Vocabulary Bottleneck in Multilingual Masked Language Models](https://arxiv.org/abs/2301.10472) par Davis Liang, Hila Gonen, Yuning Mao, Rui Hou, Naman Goyal, Marjan Ghazvininejad, Luke Zettlemoyer, Madian Khabsa.
+1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (de Google/CMU) a été publié dans l'article [XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) par Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
+1. **[XLS-R](https://huggingface.co/docs/transformers/model_doc/xls_r)** (de Facebook AI) publié dans l'article [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) par Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli.
+1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (de Facebook AI) publié dans l'article [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) par Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
+1. **[YOLOS](https://huggingface.co/docs/transformers/model_doc/yolos)** (de l'Université Huazhong des sciences et technologies) publié dans l'article [You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection](https://arxiv.org/abs/2106.00666) par Yuxin Fang, Bencheng Liao, Xinggang Wang, Jiemin Fang, Jiyang Qi, Rui Wu, Jianwei Niu, Wenyu Liu.
+1. **[YOSO](https://huggingface.co/docs/transformers/model_doc/yoso)** (de l'Université du Wisconsin - Madison) publié dans l'article [You Only Sample (Almost) Once: Linear Cost Self-Attention Via Bernoulli Sampling](https://arxiv.org/abs/2111.09714) par Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh.
 1. Vous souhaitez contribuer avec un nouveau modèle ? Nous avons ajouté un **guide détaillé et des modèles types** pour vous guider dans le processus d'ajout d'un nouveau modèle. Vous pouvez les trouver dans le dossier [`templates`](./templates) du référentiel. Assurez-vous de consulter les [directives de contribution](./CONTRIBUTING.md) et de contacter les mainteneurs ou d'ouvrir un ticket pour recueillir des commentaires avant de commencer votre pull request.
 
 Pour vérifier si chaque modèle a une implémentation en Flax, PyTorch ou TensorFlow, ou s'il a un tokenizer associé pris en charge par la bibliothèque 🤗 Tokenizers, consultez [ce tableau](https://huggingface.co/docs/transformers/index#supported-frameworks).

README_hd.md

@@ -243,102 +243,102 @@ conda install conda-forge::transformers
 
 🤗 ट्रांसफॉर्मर वर्तमान में निम्नलिखित आर्किटेक्चर का समर्थन करते हैं (मॉडल के अवलोकन के लिए [यहां देखें](https://huggingface.co/docs/transformers/model_summary))：
 
-1. **[ALBERT](https://huggingface.co/docs/transformers/model_doc/albert)** (Google Research and the Toyota Technological Institute at Chicago) साथ थीसिस [ALBERT: A Lite BERT for Self-supervised भाषा प्रतिनिधित्व सीखना](https://arxiv.org/abs/1909.11942), झेंझोंग लैन, मिंगदा चेन, सेबेस्टियन गुडमैन, केविन गिम्पेल, पीयूष शर्मा, राडू सोरिकट
+1. **[ALBERT](https://huggingface.co/docs/transformers/model_doc/albert)** (Google Research and the Toyota Technological Institute at Chicago) साथ थीसिस [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://arxiv.org/abs/1909.11942), झेंझोंग लैन, मिंगदा चेन, सेबेस्टियन गुडमैन, केविन गिम्पेल, पीयूष शर्मा, राडू सोरिकट
 1. **[ALIGN](https://huggingface.co/docs/transformers/model_doc/align)** (Google Research से) Chao Jia, Yinfei Yang, Ye Xia, Yi-Ting Chen, Zarana Parekh, Hieu Pham, Quoc V. Le, Yunhsuan Sung, Zhen Li, Tom Duerig. द्वाराअनुसंधान पत्र [Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision](https://arxiv.org/abs/2102.05918) के साथ जारी किया गया
 1. **[AltCLIP](https://huggingface.co/docs/transformers/model_doc/altclip)** (from BAAI) released with the paper [AltCLIP: Altering the Language Encoder in CLIP for Extended Language Capabilities](https://arxiv.org/abs/2211.06679) by Chen, Zhongzhi and Liu, Guang and Zhang, Bo-Wen and Ye, Fulong and Yang, Qinghong and Wu, Ledell.
 1. **[Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer)** (from MIT) released with the paper [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) by Yuan Gong, Yu-An Chung, James Glass.
 1. **[Autoformer](https://huggingface.co/docs/transformers/model_doc/autoformer)** (from Tsinghua University) released with the paper [Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting](https://arxiv.org/abs/2106.13008) by Haixu Wu, Jiehui Xu, Jianmin Wang, Mingsheng Long.
 1. **[Bark](https://huggingface.co/docs/transformers/model_doc/bark)** (from Suno) released in the repository [suno-ai/bark](https://github.com/suno-ai/bark) by Suno AI team.
-1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (फेसबुक) साथ थीसिस [बार्ट: प्राकृतिक भाषा निर्माण, अनुवाद के लिए अनुक्रम-से-अनुक्रम पूर्व प्रशिक्षण , और समझ](https://arxiv.org/pdf/1910.13461.pdf) पर निर्भर माइक लुईस, यिनहान लियू, नमन गोयल, मार्जन ग़ज़विनिनेजाद, अब्देलरहमान मोहम्मद, ओमर लेवी, वेस स्टोयानोव और ल्यूक ज़ेटलमॉयर
+1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (फेसबुक) साथ थीसिस [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) पर निर्भर माइक लुईस, यिनहान लियू, नमन गोयल, मार्जन ग़ज़विनिनेजाद, अब्देलरहमान मोहम्मद, ओमर लेवी, वेस स्टोयानोव और ल्यूक ज़ेटलमॉयर
 1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (से École polytechnique) साथ थीसिस [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) पर निर्भर Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis रिहाई।
 1. **[BARTpho](https://huggingface.co/docs/transformers/model_doc/bartpho)** (VinAI Research से) साथ में पेपर [BARTpho: Pre-trained Sequence-to-Sequence Models for Vietnamese](https://arxiv.org/abs/2109.09701)गुयेन लुओंग ट्रान, डुओंग मिन्ह ले और डाट क्वोक गुयेन द्वारा पोस्ट किया गया।
-1. **[BEiT](https://huggingface.co/docs/transformers/model_doc/beit)** (Microsoft से) साथ में कागज [BEiT: BERT इमेज ट्रांसफॉर्मर्स का प्री-ट्रेनिंग](https://arxiv.org/abs/2106.08254) Hangbo Bao, Li Dong, Furu Wei द्वारा।
-1. **[BERT](https://huggingface.co/docs/transformers/model_doc/bert)** (गूगल से) साथ वाला पेपर [बीईआरटी: प्री-ट्रेनिंग ऑफ डीप बिडायरेक्शनल ट्रांसफॉर्मर्स फॉर लैंग्वेज अंडरस्टैंडिंग](https://arxiv.org/abs/1810.04805) जैकब डेवलिन, मिंग-वेई चांग, केंटन ली और क्रिस्टीना टौटानोवा द्वारा प्रकाशित किया गया था। .
-1. **[BERT For Sequence Generation](https://huggingface.co/docs/transformers/model_doc/bert-generation)** (गूगल से) साथ देने वाला पेपर [सीक्वेंस जेनरेशन टास्क के लिए प्री-ट्रेंड चेकपॉइंट का इस्तेमाल करना](https://arxiv.org/abs/1907.12461) साशा रोठे, शशि नारायण, अलियाक्सि सेवेरिन द्वारा।
-1. **[BERTweet](https://huggingface.co/docs/transformers/model_doc/bertweet)** (VinAI Research से) साथ में पेपर [BERTweet: अंग्रेजी ट्वीट्स के लिए एक पूर्व-प्रशिक्षित भाषा मॉडल](https://aclanthology.org/2020.emnlp-demos.2/) डाट क्वोक गुयेन, थान वु और अन्ह तुआन गुयेन द्वारा प्रकाशित।
-1. **[BigBird-Pegasus](https://huggingface.co/docs/transformers/model_doc/bigbird_pegasus)** (गूगल रिसर्च से) साथ वाला पेपर [बिग बर्ड: ट्रांसफॉर्मर्स फॉर लॉन्गर सीक्वेंस](https://arxiv.org/abs/2007.14062) मंज़िल ज़हीर, गुरु गुरुगणेश, अविनावा दुबे, जोशुआ आइंस्ली, क्रिस अल्बर्टी, सैंटियागो ओंटानोन, फिलिप फाम, अनिरुद्ध रावुला, किफ़ान वांग, ली यांग, अमर अहमद द्वारा।
-1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (गूगल रिसर्च से) साथ में पेपर [बिग बर्ड: ट्रांसफॉर्मर्स फॉर लॉन्गर सीक्वेंस](https://arxiv.org/abs/2007.14062) मंज़िल ज़हीर, गुरु गुरुगणेश, अविनावा दुबे, जोशुआ आइंस्ली, क्रिस अल्बर्टी, सैंटियागो ओंटानन, फिलिप फाम द्वारा , अनिरुद्ध रावुला, किफ़ान वांग, ली यांग, अमर अहमद द्वारा पोस्ट किया गया।
+1. **[BEiT](https://huggingface.co/docs/transformers/model_doc/beit)** (Microsoft से) साथ में कागज [BEiT: BERT Pre-Training of Image Transformers](https://arxiv.org/abs/2106.08254) Hangbo Bao, Li Dong, Furu Wei द्वारा।
+1. **[BERT](https://huggingface.co/docs/transformers/model_doc/bert)** (गूगल से) साथ वाला पेपर [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://arxiv.org/abs/1810.04805) जैकब डेवलिन, मिंग-वेई चांग, केंटन ली और क्रिस्टीना टौटानोवा द्वारा प्रकाशित किया गया था। .
+1. **[BERT For Sequence Generation](https://huggingface.co/docs/transformers/model_doc/bert-generation)** (गूगल से) साथ देने वाला पेपर [Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) साशा रोठे, शशि नारायण, अलियाक्सि सेवेरिन द्वारा।
+1. **[BERTweet](https://huggingface.co/docs/transformers/model_doc/bertweet)** (VinAI Research से) साथ में पेपर [BERTweet: A pre-trained language model for English Tweets](https://aclanthology.org/2020.emnlp-demos.2/) डाट क्वोक गुयेन, थान वु और अन्ह तुआन गुयेन द्वारा प्रकाशित।
+1. **[BigBird-Pegasus](https://huggingface.co/docs/transformers/model_doc/bigbird_pegasus)** (गूगल रिसर्च से) साथ वाला पेपर [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) मंज़िल ज़हीर, गुरु गुरुगणेश, अविनावा दुबे, जोशुआ आइंस्ली, क्रिस अल्बर्टी, सैंटियागो ओंटानोन, फिलिप फाम, अनिरुद्ध रावुला, किफ़ान वांग, ली यांग, अमर अहमद द्वारा।
+1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (गूगल रिसर्च से) साथ में पेपर [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) मंज़िल ज़हीर, गुरु गुरुगणेश, अविनावा दुबे, जोशुआ आइंस्ली, क्रिस अल्बर्टी, सैंटियागो ओंटानन, फिलिप फाम द्वारा , अनिरुद्ध रावुला, किफ़ान वांग, ली यांग, अमर अहमद द्वारा पोस्ट किया गया।
 1. **[BioGpt](https://huggingface.co/docs/transformers/model_doc/biogpt)** (from Microsoft Research AI4Science) released with the paper [BioGPT: generative pre-trained transformer for biomedical text generation and mining](https://academic.oup.com/bib/advance-article/doi/10.1093/bib/bbac409/6713511?guestAccessKey=a66d9b5d-4f83-4017-bb52-405815c907b9) by Renqian Luo, Liai Sun, Yingce Xia, Tao Qin, Sheng Zhang, Hoifung Poon and Tie-Yan Liu.
 1. **[BiT](https://huggingface.co/docs/transformers/model_doc/bit)** (from Google AI) released with the paper [Big Transfer (BiT) by Alexander Kolesnikov, Lucas Beyer, Xiaohua Zhai, Joan Puigcerver, Jessica Yung, Sylvain Gelly, Neil Houlsby.
-1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (फेसबुक से) साथ में कागज [एक ओपन-डोमेन चैटबॉट बनाने की विधि](https://arxiv.org/abs/2004.13637) स्टीफन रोलर, एमिली दीनन, नमन गोयल, दा जू, मैरी विलियमसन, यिनहान लियू, जिंग जू, मायल ओट, कर्ट शस्टर, एरिक एम। स्मिथ, वाई-लैन बॉरो, जेसन वेस्टन द्वारा।
-1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (फेसबुक से) साथ में पेपर [एक ओपन-डोमेन चैटबॉट बनाने की रेसिपी](https://arxiv.org/abs/2004.13637) स्टीफन रोलर, एमिली दीनन, नमन गोयल, दा जू, मैरी विलियमसन, यिनहान लियू, जिंग जू, मायल ओट, कर्ट शस्टर, एरिक एम स्मिथ, वाई-लैन बॉरो, जेसन वेस्टन द्वारा।
+1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (फेसबुक से) साथ में कागज [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) स्टीफन रोलर, एमिली दीनन, नमन गोयल, दा जू, मैरी विलियमसन, यिनहान लियू, जिंग जू, मायल ओट, कर्ट शस्टर, एरिक एम। स्मिथ, वाई-लैन बॉरो, जेसन वेस्टन द्वारा।
+1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (फेसबुक से) साथ में पेपर [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) स्टीफन रोलर, एमिली दीनन, नमन गोयल, दा जू, मैरी विलियमसन, यिनहान लियू, जिंग जू, मायल ओट, कर्ट शस्टर, एरिक एम स्मिथ, वाई-लैन बॉरो, जेसन वेस्टन द्वारा।
 1. **[BLIP](https://huggingface.co/docs/transformers/model_doc/blip)** (from Salesforce) released with the paper [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://arxiv.org/abs/2201.12086) by Junnan Li, Dongxu Li, Caiming Xiong, Steven Hoi.
 1. **[BLIP-2](https://huggingface.co/docs/transformers/model_doc/blip-2)** (Salesforce से) Junnan Li, Dongxu Li, Silvio Savarese, Steven Hoi. द्वाराअनुसंधान पत्र [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://arxiv.org/abs/2301.12597) के साथ जारी किया गया
-1. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/).
-1. **[BORT](https://huggingface.co/docs/transformers/model_doc/bort)** (एलेक्सा से) कागज के साथ [बीईआरटी के लिए ऑप्टिमल सबआर्किटेक्चर एक्सट्रैक्शन](https://arxiv.org/abs/2010.10499) एड्रियन डी विंटर और डैनियल जे पेरी द्वारा।
-1. **[BridgeTower](https://huggingface.co/docs/transformers/model_doc/bridgetower)** (हरबिन इंस्टिट्यूट ऑफ़ टेक्नोलॉजी/माइक्रोसॉफ्ट रिसर्च एशिया/इंटेल लैब्स से) कागज के साथ [ब्रिजटॉवर: विजन-लैंग्वेज रिप्रेजेंटेशन लर्निंग में एनकोडर्स के बीच ब्रिज बनाना](<https://arxiv.org/abs/2206.08657>) by Xiao Xu, Chenfei Wu, Shachar Rosenman, Vasudev Lal, Wanxiang Che, Nan Duan.
+1. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (from BigScience workshop) released by the [BigScience Workshop](https://bigscience.huggingface.co/).
+1. **[BORT](https://huggingface.co/docs/transformers/model_doc/bort)** (एलेक्सा से) कागज के साथ [Optimal Subarchitecture Extraction For BERT](https://arxiv.org/abs/2010.10499) एड्रियन डी विंटर और डैनियल जे पेरी द्वारा।
+1. **[BridgeTower](https://huggingface.co/docs/transformers/model_doc/bridgetower)** (हरबिन इंस्टिट्यूट ऑफ़ टेक्नोलॉजी/माइक्रोसॉफ्ट रिसर्च एशिया/इंटेल लैब्स से) कागज के साथ [BridgeTower: Building Bridges Between Encoders in Vision-Language Representation Learning](https://arxiv.org/abs/2206.08657) by Xiao Xu, Chenfei Wu, Shachar Rosenman, Vasudev Lal, Wanxiang Che, Nan Duan.
 1. **[BROS](https://huggingface.co/docs/transformers/model_doc/bros)** (NAVER CLOVA से) Teakgyu Hong, Donghyun Kim, Mingi Ji, Wonseok Hwang, Daehyun Nam, Sungrae Park. द्वाराअनुसंधान पत्र [BROS: A Pre-trained Language Model Focusing on Text and Layout for Better Key Information Extraction from Documents](https://arxiv.org/abs/2108.04539) के साथ जारी किया गया
-1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (Google अनुसंधान से) साथ में कागज [ByT5: पूर्व-प्रशिक्षित बाइट-टू-बाइट मॉडल के साथ एक टोकन-मुक्त भविष्य की ओर](https://arxiv.org/abs/2105.13626) Linting Xue, Aditya Barua, Noah Constant, रामी अल-रफू, शरण नारंग, मिहिर काले, एडम रॉबर्ट्स, कॉलिन रैफेल द्वारा पोस्ट किया गया।
-1. **[CamemBERT](https://huggingface.co/docs/transformers/model_doc/camembert)** (इनरिया/फेसबुक/सोरबोन से) साथ में कागज [CamemBERT: एक टेस्टी फ्रेंच लैंग्वेज मॉडल](https://arxiv.org/abs/1911.03894) लुई मार्टिन*, बेंजामिन मुलर*, पेड्रो जेवियर ऑर्टिज़ सुआरेज़*, योआन ड्यूपॉन्ट, लॉरेंट रोमरी, एरिक विलेमोन्टे डे ला क्लर्जरी, जैमे सेडाह और बेनोइट सगोट द्वारा।
-1. **[CANINE](https://huggingface.co/docs/transformers/model_doc/canine)** (Google रिसर्च से) साथ में दिया गया पेपर [कैनाइन: प्री-ट्रेनिंग ए एफिशिएंट टोकनाइजेशन-फ्री एनकोडर फॉर लैंग्वेज रिप्रेजेंटेशन](https://arxiv.org/abs/2103.06874) जोनाथन एच क्लार्क, डैन गैरेट, यूलिया टर्क, जॉन विएटिंग द्वारा।
+1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (Google अनुसंधान से) साथ में कागज [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) Linting Xue, Aditya Barua, Noah Constant, रामी अल-रफू, शरण नारंग, मिहिर काले, एडम रॉबर्ट्स, कॉलिन रैफेल द्वारा पोस्ट किया गया।
+1. **[CamemBERT](https://huggingface.co/docs/transformers/model_doc/camembert)** (इनरिया/फेसबुक/सोरबोन से) साथ में कागज [CamemBERT: a Tasty French Language Model](https://arxiv.org/abs/1911.03894) लुई मार्टिन*, बेंजामिन मुलर*, पेड्रो जेवियर ऑर्टिज़ सुआरेज़*, योआन ड्यूपॉन्ट, लॉरेंट रोमरी, एरिक विलेमोन्टे डे ला क्लर्जरी, जैमे सेडाह और बेनोइट सगोट द्वारा।
+1. **[CANINE](https://huggingface.co/docs/transformers/model_doc/canine)** (Google रिसर्च से) साथ में दिया गया पेपर [CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language Representation](https://arxiv.org/abs/2103.06874) जोनाथन एच क्लार्क, डैन गैरेट, यूलिया टर्क, जॉन विएटिंग द्वारा।
 1. **[Chinese-CLIP](https://huggingface.co/docs/transformers/model_doc/chinese_clip)** (from OFA-Sys) released with the paper [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) by An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou.
 1. **[CLAP](https://huggingface.co/docs/transformers/model_doc/clap)** (LAION-AI से) Yusong Wu, Ke Chen, Tianyu Zhang, Yuchen Hui, Taylor Berg-Kirkpatrick, Shlomo Dubnov. द्वाराअनुसंधान पत्र [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://arxiv.org/abs/2211.06687) के साथ जारी किया गया
-1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (OpenAI से) साथ वाला पेपर [लर्निंग ट्रांसफरेबल विजुअल मॉडल फ्रॉम नेचुरल लैंग्वेज सुपरविजन](https://arxiv.org/abs/2103.00020) एलेक रैडफोर्ड, जोंग वूक किम, क्रिस हैलासी, आदित्य रमेश, गेब्रियल गोह, संध्या अग्रवाल, गिरीश शास्त्री, अमांडा एस्केल, पामेला मिश्किन, जैक क्लार्क, ग्रेचेन क्रुएगर, इल्या सुत्स्केवर द्वारा।
+1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (OpenAI से) साथ वाला पेपर [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) एलेक रैडफोर्ड, जोंग वूक किम, क्रिस हैलासी, आदित्य रमेश, गेब्रियल गोह, संध्या अग्रवाल, गिरीश शास्त्री, अमांडा एस्केल, पामेला मिश्किन, जैक क्लार्क, ग्रेचेन क्रुएगर, इल्या सुत्स्केवर द्वारा।
 1. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (from University of Göttingen) released with the paper [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) by Timo Lüddecke and Alexander Ecker.
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
-1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (सेल्सफोर्स से) साथ में पेपर [प्रोग्राम सिंथेसिस के लिए एक संवादात्मक प्रतिमान](https://arxiv.org/abs/2203.13474) एरिक निजकैंप, बो पैंग, हिरोआकी हयाशी, लिफू तू, हुआन वांग, यिंगबो झोउ, सिल्वियो सावरेस, कैमिंग जिओंग रिलीज।
+1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (सेल्सफोर्स से) साथ में पेपर [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) एरिक निजकैंप, बो पैंग, हिरोआकी हयाशी, लिफू तू, हुआन वांग, यिंगबो झोउ, सिल्वियो सावरेस, कैमिंग जिओंग रिलीज।
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (MetaAI से) Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve. द्वाराअनुसंधान पत्र [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) के साथ जारी किया गया
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (Cohere से) Cohere.  द्वाराअनुसंधान पत्र [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) के साथ जारी किया गया
-1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (माइक्रोसॉफ्ट रिसर्च एशिया से) कागज के साथ [फास्ट ट्रेनिंग कन्वर्जेंस के लिए सशर्त डीईटीआर](https://arxiv.org/abs/2108.06152) डेपू मेंग, ज़ियाओकांग चेन, ज़ेजिया फैन, गैंग ज़ेंग, होउकियांग ली, युहुई युआन, लेई सन, जिंगडोंग वांग द्वारा।
-1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (YituTech से) साथ में कागज [ConvBERT: स्पैन-आधारित डायनेमिक कनवल्शन के साथ BERT में सुधार](https://arxiv.org/abs/2008.02496) जिहांग जियांग, वीहाओ यू, डाकान झोउ, युनपेंग चेन, जियाशी फेंग, शुइचेंग यान द्वारा।
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (Cohere से) Cohere.  द्वाराअनुसंधान पत्र [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) के साथ जारी किया गया
+1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (माइक्रोसॉफ्ट रिसर्च एशिया से) कागज के साथ [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) डेपू मेंग, ज़ियाओकांग चेन, ज़ेजिया फैन, गैंग ज़ेंग, होउकियांग ली, युहुई युआन, लेई सन, जिंगडोंग वांग द्वारा।
+1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (YituTech से) साथ में कागज [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) जिहांग जियांग, वीहाओ यू, डाकान झोउ, युनपेंग चेन, जियाशी फेंग, शुइचेंग यान द्वारा।
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (Facebook AI से) साथ वाला पेपर [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) ज़ुआंग लियू, हेंज़ी माओ, चाओ-युआन वू, क्रिस्टोफ़ फीचटेनहोफ़र, ट्रेवर डेरेल, सैनिंग ज़ी द्वारा।
 1. **[ConvNeXTV2](https://huggingface.co/docs/transformers/model_doc/convnextv2)** (from Facebook AI) released with the paper [ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoders](https://arxiv.org/abs/2301.00808) by Sanghyun Woo, Shoubhik Debnath, Ronghang Hu, Xinlei Chen, Zhuang Liu, In So Kweon, Saining Xie.
-1. **[CPM](https://huggingface.co/docs/transformers/model_doc/cpm)** (सिंघुआ यूनिवर्सिटी से) साथ में पेपर [सीपीएम: ए लार्ज-स्केल जेनेरेटिव चाइनीज प्री-ट्रेंड लैंग्वेज मॉडल](https://arxiv.org/abs/2012.00413) झेंग्यान झांग, जू हान, हाओ झोउ, पेई के, युक्सियन गु, डेमिंग ये, युजिया किन, युशेंग सु, हाओझे जी, जियान गुआन, फैंचाओ क्यूई, ज़ियाओझी वांग, यानान झेंग द्वारा , गुओयांग ज़ेंग, हुआनकी काओ, शेंगकी चेन, डाइक्सुआन ली, ज़ेनबो सन, ज़ियुआन लियू, मिनली हुआंग, वेंटाओ हान, जी तांग, जुआनज़ी ली, ज़ियाओयान झू, माओसोंग सन।
+1. **[CPM](https://huggingface.co/docs/transformers/model_doc/cpm)** (सिंघुआ यूनिवर्सिटी से) साथ में पेपर [CPM: A Large-scale Generative Chinese Pre-trained Language Model](https://arxiv.org/abs/2012.00413) झेंग्यान झांग, जू हान, हाओ झोउ, पेई के, युक्सियन गु, डेमिंग ये, युजिया किन, युशेंग सु, हाओझे जी, जियान गुआन, फैंचाओ क्यूई, ज़ियाओझी वांग, यानान झेंग द्वारा , गुओयांग ज़ेंग, हुआनकी काओ, शेंगकी चेन, डाइक्सुआन ली, ज़ेनबो सन, ज़ियुआन लियू, मिनली हुआंग, वेंटाओ हान, जी तांग, जुआनज़ी ली, ज़ियाओयान झू, माओसोंग सन।
 1. **[CPM-Ant](https://huggingface.co/docs/transformers/model_doc/cpmant)** (from OpenBMB) released by the [OpenBMB](https://www.openbmb.org/).
-1. **[CTRL](https://huggingface.co/docs/transformers/model_doc/ctrl)** (सेल्सफोर्स से) साथ में पेपर [CTRL: ए कंडिशनल ट्रांसफॉर्मर लैंग्वेज मॉडल फॉर कंट्रोलेबल जेनरेशन](https://arxiv.org/abs/1909.05858) नीतीश शिरीष केसकर*, ब्रायन मैककैन*, लव आर. वार्ष्णेय, कैमिंग जिओंग और रिचर्ड द्वारा सोचर द्वारा जारी किया गया।
-1. **[CvT](https://huggingface.co/docs/transformers/model_doc/cvt)** (Microsoft से) साथ में दिया गया पेपर [CvT: इंट्रोड्यूसिंग कनवॉल्यूशन टू विजन ट्रांसफॉर्मर्स](https://arxiv.org/एब्स/2103.15808) हैपिंग वू, बिन जिओ, नोएल कोडेला, मेंगचेन लियू, जियांग दाई, लू युआन, लेई झांग द्वारा।
-1. **[Data2Vec](https://huggingface.co/docs/transformers/model_doc/data2vec)** (फेसबुक से) साथ में कागज [Data2Vec: भाषण, दृष्टि और भाषा में स्व-पर्यवेक्षित सीखने के लिए एक सामान्य ढांचा](https://arxiv.org/abs/2202.03555) एलेक्सी बाएव्स्की, वेई-निंग सू, कियानटोंग जू, अरुण बाबू, जियाताओ गु, माइकल औली द्वारा पोस्ट किया गया।
-1. **[DeBERTa](https://huggingface.co/docs/transformers/model_doc/deberta)** (Microsoft से) साथ में दिया गया पेपर [DeBERta: डिकोडिंग-एन्हांस्ड BERT विद डिसेंटैंगल्ड अटेंशन](https://arxiv.org/abs/2006.03654) पेंगचेंग हे, ज़ियाओडोंग लियू, जियानफेंग गाओ, वीज़ू चेन द्वारा।
-1. **[DeBERTa-v2](https://huggingface.co/docs/transformers/model_doc/deberta-v2)** (Microsoft से) साथ में दिया गया पेपर [DeBERTa: डिकोडिंग-एन्हांस्ड BERT विथ डिसेंन्गल्ड अटेंशन](https://arxiv.org/abs/2006.03654) पेंगचेंग हे, ज़ियाओडोंग लियू, जियानफेंग गाओ, वीज़ू चेन द्वारा पोस्ट किया गया।
-1. **[Decision Transformer](https://huggingface.co/docs/transformers/model_doc/decision_transformer)** (बर्कले/फेसबुक/गूगल से) पेपर के साथ [डिसीजन ट्रांसफॉर्मर: रीनफोर्समेंट लर्निंग वाया सीक्वेंस मॉडलिंग](https://arxiv.org/abs/2106.01345) लिली चेन, केविन लू, अरविंद राजेश्वरन, किमिन ली, आदित्य ग्रोवर, माइकल लास्किन, पीटर एबील, अरविंद श्रीनिवास, इगोर मोर्डच द्वारा पोस्ट किया गया।
-1. **[Deformable DETR](https://huggingface.co/docs/transformers/model_doc/deformable_detr)** (सेंसटाइम रिसर्च से) साथ में पेपर [डिफॉर्मेबल डीईटीआर: डिफॉर्मेबल ट्रांसफॉर्मर्स फॉर एंड-टू-एंड ऑब्जेक्ट डिटेक्शन](https://arxiv.org/abs/2010.04159) Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, जिफेंग दाई द्वारा पोस्ट किया गया।
-1. **[DeiT](https://huggingface.co/docs/transformers/model_doc/deit)** (फेसबुक से) साथ में पेपर [ट्रेनिंग डेटा-एफिशिएंट इमेज ट्रांसफॉर्मर और डिस्टिलेशन थ्रू अटेंशन](https://arxiv.org/abs/2012.12877) ह्यूगो टौव्रोन, मैथ्यू कॉर्ड, मैथिज्स डूज़, फ़्रांसिस्को मस्सा, एलेक्ज़ेंडर सबलेरोल्स, हर्वे जेगौ द्वारा।
+1. **[CTRL](https://huggingface.co/docs/transformers/model_doc/ctrl)** (सेल्सफोर्स से) साथ में पेपर [CTRL: A Conditional Transformer Language Model for Controllable Generation](https://arxiv.org/abs/1909.05858) नीतीश शिरीष केसकर*, ब्रायन मैककैन*, लव आर. वार्ष्णेय, कैमिंग जिओंग और रिचर्ड द्वारा सोचर द्वारा जारी किया गया।
+1. **[CvT](https://huggingface.co/docs/transformers/model_doc/cvt)** (Microsoft से) साथ में दिया गया पेपर [CvT: Introducing Convolutions to Vision Transformers](https://arxiv.org/abs/2103.15808) हैपिंग वू, बिन जिओ, नोएल कोडेला, मेंगचेन लियू, जियांग दाई, लू युआन, लेई झांग द्वारा।
+1. **[Data2Vec](https://huggingface.co/docs/transformers/model_doc/data2vec)** (फेसबुक से) साथ में कागज [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) एलेक्सी बाएव्स्की, वेई-निंग सू, कियानटोंग जू, अरुण बाबू, जियाताओ गु, माइकल औली द्वारा पोस्ट किया गया।
+1. **[DeBERTa](https://huggingface.co/docs/transformers/model_doc/deberta)** (Microsoft से) साथ में दिया गया पेपर [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://arxiv.org/abs/2006.03654) पेंगचेंग हे, ज़ियाओडोंग लियू, जियानफेंग गाओ, वीज़ू चेन द्वारा।
+1. **[DeBERTa-v2](https://huggingface.co/docs/transformers/model_doc/deberta-v2)** (Microsoft से) साथ में दिया गया पेपर [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://arxiv.org/abs/2006.03654) पेंगचेंग हे, ज़ियाओडोंग लियू, जियानफेंग गाओ, वीज़ू चेन द्वारा पोस्ट किया गया।
+1. **[Decision Transformer](https://huggingface.co/docs/transformers/model_doc/decision_transformer)** (बर्कले/फेसबुक/गूगल से) पेपर के साथ [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://arxiv.org/abs/2106.01345) लिली चेन, केविन लू, अरविंद राजेश्वरन, किमिन ली, आदित्य ग्रोवर, माइकल लास्किन, पीटर एबील, अरविंद श्रीनिवास, इगोर मोर्डच द्वारा पोस्ट किया गया।
+1. **[Deformable DETR](https://huggingface.co/docs/transformers/model_doc/deformable_detr)** (सेंसटाइम रिसर्च से) साथ में पेपर [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://arxiv.org/abs/2010.04159) Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, जिफेंग दाई द्वारा पोस्ट किया गया।
+1. **[DeiT](https://huggingface.co/docs/transformers/model_doc/deit)** (फेसबुक से) साथ में पेपर [Training data-efficient image transformers & distillation through attention](https://arxiv.org/abs/2012.12877) ह्यूगो टौव्रोन, मैथ्यू कॉर्ड, मैथिज्स डूज़, फ़्रांसिस्को मस्सा, एलेक्ज़ेंडर सबलेरोल्स, हर्वे जेगौ द्वारा।
 1. **[DePlot](https://huggingface.co/docs/transformers/model_doc/deplot)** (Google AI से) Fangyu Liu, Julian Martin Eisenschlos, Francesco Piccinno, Syrine Krichene, Chenxi Pang, Kenton Lee, Mandar Joshi, Wenhu Chen, Nigel Collier, Yasemin Altun. द्वाराअनुसंधान पत्र [DePlot: One-shot visual language reasoning by plot-to-table translation](https://arxiv.org/abs/2212.10505) के साथ जारी किया गया
 1. **[Depth Anything](https://huggingface.co/docs/transformers/model_doc/depth_anything)** (University of Hong Kong and TikTok से) Lihe Yang, Bingyi Kang, Zilong Huang, Xiaogang Xu, Jiashi Feng, Hengshuang Zhao. द्वाराअनुसंधान पत्र [Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data](https://arxiv.org/abs/2401.10891) के साथ जारी किया गया
 1. **[DETA](https://huggingface.co/docs/transformers/model_doc/deta)** (from The University of Texas at Austin) released with the paper [NMS Strikes Back](https://arxiv.org/abs/2212.06137) by Jeffrey Ouyang-Zhang, Jang Hyun Cho, Xingyi Zhou, Philipp Krähenbühl.
-1. **[DETR](https://huggingface.co/docs/transformers/model_doc/detr)** (फेसबुक से) साथ में कागज [ट्रांसफॉर्मर्स के साथ एंड-टू-एंड ऑब्जेक्ट डिटेक्शन](https://arxiv.org/abs/2005.12872) निकोलस कैरियन, फ़्रांसिस्को मस्सा, गेब्रियल सिनेव, निकोलस उसुनियर, अलेक्जेंडर किरिलोव, सर्गेई ज़ागोरुयको द्वारा।
-1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (माइक्रोसॉफ्ट रिसर्च से) कागज के साथ [DialoGPT: बड़े पैमाने पर जनरेटिव प्री-ट्रेनिंग फॉर कन्वर्सेशनल रिस्पांस जेनरेशन](https://arxiv.org/abs/1911.00536) यिज़े झांग, सिकी सन, मिशेल गैली, येन-चुन चेन, क्रिस ब्रोकेट, जियांग गाओ, जियानफेंग गाओ, जिंगजिंग लियू, बिल डोलन द्वारा।
+1. **[DETR](https://huggingface.co/docs/transformers/model_doc/detr)** (फेसबुक से) साथ में कागज [End-to-End Object Detection with Transformers](https://arxiv.org/abs/2005.12872) निकोलस कैरियन, फ़्रांसिस्को मस्सा, गेब्रियल सिनेव, निकोलस उसुनियर, अलेक्जेंडर किरिलोव, सर्गेई ज़ागोरुयको द्वारा।
+1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (माइक्रोसॉफ्ट रिसर्च से) कागज के साथ [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://arxiv.org/abs/1911.00536) यिज़े झांग, सिकी सन, मिशेल गैली, येन-चुन चेन, क्रिस ब्रोकेट, जियांग गाओ, जियानफेंग गाओ, जिंगजिंग लियू, बिल डोलन द्वारा।
 1. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (from SHI Labs) released with the paper [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001) by Ali Hassani and Humphrey Shi.
 1. **[DINOv2](https://huggingface.co/docs/transformers/model_doc/dinov2)** (Meta AI से) Maxime Oquab, Timothée Darcet, Théo Moutakanni, Huy Vo, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, Mahmoud Assran, Nicolas Ballas, Wojciech Galuba, Russell Howes, Po-Yao Huang, Shang-Wen Li, Ishan Misra, Michael Rabbat, Vasu Sharma, Gabriel Synnaeve, Hu Xu, Hervé Jegou, Julien Mairal, Patrick Labatut, Armand Joulin, Piotr Bojanowski. द्वाराअनुसंधान पत्र [DINOv2: Learning Robust Visual Features without Supervision](https://arxiv.org/abs/2304.07193) के साथ जारी किया गया
-1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (हगिंगफेस से), साथ में कागज [डिस्टिलबर्ट, बीईआरटी का डिस्टिल्ड वर्जन: छोटा, तेज, सस्ता और हल्का](https://arxiv.org/abs/1910.01108) विक्टर सनह, लिसांड्रे डेब्यू और थॉमस वुल्फ द्वारा पोस्ट किया गया। यही तरीका GPT-2 को [DistilGPT2](https://github.com/huggingface/transformers/tree/main/examples/distillation), RoBERta से [DistilRoBERta](https://github.com) पर कंप्रेस करने के लिए भी लागू किया जाता है। / हगिंगफेस/ट्रांसफॉर्मर्स/ट्री/मेन/उदाहरण/डिस्टिलेशन), बहुभाषी BERT से [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/distillation) और डिस्टिलबर्ट का जर्मन संस्करण।
-1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (माइक्रोसॉफ्ट रिसर्च से) साथ में पेपर [DiT: सेल्फ सुपरवाइज्ड प्री-ट्रेनिंग फॉर डॉक्यूमेंट इमेज ट्रांसफॉर्मर](https://arxiv.org/abs/2203.02378) जुनलॉन्ग ली, यिहेंग जू, टेंगचाओ लव, लेई कुई, चा झांग द्वारा फुरु वेई द्वारा पोस्ट किया गया।
-1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (NAVER से) साथ में कागज [OCR-मुक्त डॉक्यूमेंट अंडरस्टैंडिंग ट्रांसफॉर्मर](https://arxiv.org/abs/2111.15664) गीवूक किम, टीकग्यू होंग, मूनबिन यिम, जियोंग्योन नाम, जिनयॉन्ग पार्क, जिनयॉन्ग यिम, वोनसेओक ह्वांग, सांगडू यूं, डोंगयून हान, सेउंग्युन पार्क द्वारा।
-1. **[DPR](https://huggingface.co/docs/transformers/model_doc/dpr)** (फेसबुक से) साथ में पेपर [ओपन-डोमेन क्वेश्चन आंसरिंग के लिए डेंस पैसेज रिट्रीवल](https://arxiv.org/abs/2004.04906) व्लादिमीर करपुखिन, बरलास ओज़ुज़, सेवन मिन, पैट्रिक लुईस, लेडेल वू, सर्गेई एडुनोव, डैनकी चेन, और वेन-ताऊ यिह द्वारा।
-1. **[DPT](https://huggingface.co/docs/transformers/master/model_doc/dpt)** (इंटेल लैब्स से) साथ में कागज [विज़न ट्रांसफॉर्मर्स फॉर डेंस प्रेडिक्शन](https://arxiv.org/abs/2103.13413) रेने रैनफ्टल, एलेक्सी बोचकोवस्की, व्लादलेन कोल्टन द्वारा।
+1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (हगिंगफेस से), साथ में कागज [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108) विक्टर सनह, लिसांड्रे डेब्यू और थॉमस वुल्फ द्वारा पोस्ट किया गया। यही तरीका GPT-2 को [DistilGPT2](https://github.com/huggingface/transformers/tree/main/examples/distillation), RoBERta से [DistilRoBERta](https://github.com) पर कंप्रेस करने के लिए भी लागू किया जाता है। / हगिंगफेस/ट्रांसफॉर्मर्स/ट्री/मेन/उदाहरण/डिस्टिलेशन), बहुभाषी BERT से [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/distillation) और डिस्टिलबर्ट का जर्मन संस्करण।
+1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (माइक्रोसॉफ्ट रिसर्च से) साथ में पेपर [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) जुनलॉन्ग ली, यिहेंग जू, टेंगचाओ लव, लेई कुई, चा झांग द्वारा फुरु वेई द्वारा पोस्ट किया गया।
+1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (NAVER से) साथ में कागज [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) गीवूक किम, टीकग्यू होंग, मूनबिन यिम, जियोंग्योन नाम, जिनयॉन्ग पार्क, जिनयॉन्ग यिम, वोनसेओक ह्वांग, सांगडू यूं, डोंगयून हान, सेउंग्युन पार्क द्वारा।
+1. **[DPR](https://huggingface.co/docs/transformers/model_doc/dpr)** (फेसबुक से) साथ में पेपर [Dense Passage Retrieval for Open-Domain Question Answering](https://arxiv.org/abs/2004.04906) व्लादिमीर करपुखिन, बरलास ओज़ुज़, सेवन मिन, पैट्रिक लुईस, लेडेल वू, सर्गेई एडुनोव, डैनकी चेन, और वेन-ताऊ यिह द्वारा।
+1. **[DPT](https://huggingface.co/docs/transformers/master/model_doc/dpt)** (इंटेल लैब्स से) साथ में कागज [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) रेने रैनफ्टल, एलेक्सी बोचकोवस्की, व्लादलेन कोल्टन द्वारा।
 1. **[EfficientFormer](https://huggingface.co/docs/transformers/model_doc/efficientformer)** (from Snap Research) released with the paper [EfficientFormer: Vision Transformers at MobileNetSpeed](https://arxiv.org/abs/2206.01191) by Yanyu Li, Geng Yuan, Yang Wen, Ju Hu, Georgios Evangelidis, Sergey Tulyakov, Yanzhi Wang, Jian Ren.
 1. **[EfficientNet](https://huggingface.co/docs/transformers/model_doc/efficientnet)** (from Google Brain) released with the paper [EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks](https://arxiv.org/abs/1905.11946) by Mingxing Tan, Quoc V. Le.
-1. **[ELECTRA](https://huggingface.co/docs/transformers/model_doc/electra)** (Google रिसर्च/स्टैनफोर्ड यूनिवर्सिटी से) साथ में दिया गया पेपर [इलेक्ट्रा: जेनरेटर के बजाय भेदभाव करने वाले के रूप में टेक्स्ट एन्कोडर्स का पूर्व-प्रशिक्षण](https://arxiv.org/abs/2003.10555) केविन क्लार्क, मिन्ह-थांग लुओंग, क्वोक वी. ले, क्रिस्टोफर डी. मैनिंग द्वारा पोस्ट किया गया।
+1. **[ELECTRA](https://huggingface.co/docs/transformers/model_doc/electra)** (Google रिसर्च/स्टैनफोर्ड यूनिवर्सिटी से) साथ में दिया गया पेपर [ELECTRA: Pre-training text encoders as discriminators rather than generators](https://arxiv.org/abs/2003.10555) केविन क्लार्क, मिन्ह-थांग लुओंग, क्वोक वी. ले, क्रिस्टोफर डी. मैनिंग द्वारा पोस्ट किया गया।
 1. **[EnCodec](https://huggingface.co/docs/transformers/model_doc/encodec)** (Meta AI से) Alexandre Défossez, Jade Copet, Gabriel Synnaeve, Yossi Adi. द्वाराअनुसंधान पत्र [High Fidelity Neural Audio Compression](https://arxiv.org/abs/2210.13438) के साथ जारी किया गया
-1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (Google रिसर्च से) साथ में दिया गया पेपर [सीक्वेंस जेनरेशन टास्क के लिए प्री-ट्रेंड चेकपॉइंट का इस्तेमाल करना](https://arxiv.org/abs/1907.12461) साशा रोठे, शशि नारायण, अलियाक्सि सेवेरिन द्वारा।
-1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)**(Baidu से) साथ देने वाला पेपर [ERNIE: एन्हांस्ड रिप्रेजेंटेशन थ्रू नॉलेज इंटीग्रेशन](https://arxiv.org/abs/1904.09223) यू सन, शुओहुआन वांग, युकुन ली, शिकुन फेंग, ज़ुई चेन, हान झांग, शिन तियान, डैनक्सियांग झू, हाओ तियान, हुआ वू द्वारा पोस्ट किया गया।
+1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (Google रिसर्च से) साथ में दिया गया पेपर [Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) साशा रोठे, शशि नारायण, अलियाक्सि सेवेरिन द्वारा।
+1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)**(Baidu से) साथ देने वाला पेपर [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) यू सन, शुओहुआन वांग, युकुन ली, शिकुन फेंग, ज़ुई चेन, हान झांग, शिन तियान, डैनक्सियांग झू, हाओ तियान, हुआ वू द्वारा पोस्ट किया गया।
 1. **[ErnieM](https://huggingface.co/docs/transformers/model_doc/ernie_m)** (Baidu से) Xuan Ouyang, Shuohuan Wang, Chao Pang, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang. द्वाराअनुसंधान पत्र [ERNIE-M: Enhanced Multilingual Representation by Aligning Cross-lingual Semantics with Monolingual Corpora](https://arxiv.org/abs/2012.15674) के साथ जारी किया गया
-1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (मेटा AI से) ट्रांसफॉर्मर प्रोटीन भाषा मॉडल हैं। **ESM-1b** पेपर के साथ जारी किया गया था [अलेक्जेंडर राइव्स, जोशुआ मेयर, टॉम सर्कु, सिद्धार्थ गोयल, ज़ेमिंग लिन द्वारा जैविक संरचना और कार्य असुरक्षित सीखने को 250 मिलियन प्रोटीन अनुक्रमों तक स्केल करने से उभरता है](https://www.pnas.org/content/118/15/e2016239118) जेसन लियू, डेमी गुओ, मायल ओट, सी. लॉरेंस ज़िटनिक, जेरी मा और रॉब फर्गस। **ESM-1v** को पेपर के साथ जारी किया गया था [भाषा मॉडल प्रोटीन फ़ंक्शन पर उत्परिवर्तन के प्रभावों की शून्य-शॉट भविष्यवाणी को सक्षम करते हैं](https://doi.org/10.1101/2021.07.09.450648) जोशुआ मेयर, रोशन राव, रॉबर्ट वेरकुइल, जेसन लियू, टॉम सर्कु और अलेक्जेंडर राइव्स द्वारा। **ESM-2** को पेपर के साथ जारी किया गया था [भाषा मॉडल विकास के पैमाने पर प्रोटीन अनुक्रम सटीक संरचना भविष्यवाणी को सक्षम करते हैं](https://doi.org/10.1101/2022.07.20.500902) ज़ेमिंग लिन, हलील अकिन, रोशन राव, ब्रायन ही, झोंगकाई झू, वेंटिंग लू, ए द्वारा लान डॉस सैंटोस कोस्टा, मरियम फ़ज़ल-ज़रंडी, टॉम सर्कू, साल कैंडिडो, अलेक्जेंडर राइव्स।
+1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (मेटा AI से) ट्रांसफॉर्मर प्रोटीन भाषा मॉडल हैं। **ESM-1b** पेपर के साथ जारी किया गया था [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) जेसन लियू, डेमी गुओ, मायल ओट, सी. लॉरेंस ज़िटनिक, जेरी मा और रॉब फर्गस। **ESM-1v** को पेपर के साथ जारी किया गया था [भाषा मॉडल प्रोटीन फ़ंक्शन पर उत्परिवर्तन के प्रभावों की शून्य-शॉट भविष्यवाणी को सक्षम करते हैं](https://doi.org/10.1101/2021.07.09.450648) जोशुआ मेयर, रोशन राव, रॉबर्ट वेरकुइल, जेसन लियू, टॉम सर्कु और अलेक्जेंडर राइव्स द्वारा। **ESM-2** को पेपर के साथ जारी किया गया था [भाषा मॉडल विकास के पैमाने पर प्रोटीन अनुक्रम सटीक संरचना भविष्यवाणी को सक्षम करते हैं](https://doi.org/10.1101/2022.07.20.500902) ज़ेमिंग लिन, हलील अकिन, रोशन राव, ब्रायन ही, झोंगकाई झू, वेंटिंग लू, ए द्वारा लान डॉस सैंटोस कोस्टा, मरियम फ़ज़ल-ज़रंडी, टॉम सर्कू, साल कैंडिडो, अलेक्जेंडर राइव्स।
 1. **[Falcon](https://huggingface.co/docs/transformers/model_doc/falcon)** (from Technology Innovation Institute) by Almazrouei, Ebtesam and Alobeidli, Hamza and Alshamsi, Abdulaziz and Cappelli, Alessandro and Cojocaru, Ruxandra and Debbah, Merouane and Goffinet, Etienne and Heslow, Daniel and Launay, Julien and Malartic, Quentin and Noune, Badreddine and Pannier, Baptiste and Penedo, Guilherme.
-1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (ESPnet and Microsoft Research से) Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang, and Yuekai Zhang. द्वाराअनुसंधान पत्र [Fastspeech 2: Fast And High-quality End-to-End Text To Speech](https://arxiv.org/pdf/2006.04558.pdf) के साथ जारी किया गया
+1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (ESPnet and Microsoft Research से) Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang, and Yuekai Zhang. द्वाराअनुसंधान पत्र [Recent Developments On Espnet Toolkit Boosted By Conformer](https://arxiv.org/abs/2010.13956) के साथ जारी किया गया
 1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 1. **[FLAN-UL2](https://huggingface.co/docs/transformers/model_doc/flan-ul2)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-ul2-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
-1. **[FlauBERT](https://huggingface.co/docs/transformers/model_doc/flaubert)** (CNRS से) साथ वाला पेपर [FlauBERT: Unsupervised Language Model Pre-training for फ़्रेंच](https://arxiv.org/abs/1912.05372) Hang Le, Loïc Vial, Jibril Frej, Vincent Segonne, Maximin Coavoux, बेंजामिन लेकोउटेक्स, अलेक्जेंड्रे अल्लाउज़ेन, बेनोइट क्रैबे, लॉरेंट बेसेसियर, डिडिएर श्वाब द्वारा।
-1. **[FLAVA](https://huggingface.co/docs/transformers/model_doc/flava)** [FLAVA: A फाउंडेशनल लैंग्वेज एंड विजन अलाइनमेंट मॉडल](https://arxiv.org/abs/2112.04482) साथ वाला पेपर अमनप्रीत सिंह, रोंगहांग हू, वेदानुज गोस्वामी, गुइल्यूम कुएरॉन, वोज्शिएक गालुबा, मार्कस रोहरबैक, और डौवे कीला द्वारा।
-1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (गूगल रिसर्च से) साथ वाला पेपर [FNet: मिक्सिंग टोकन विद फूरियर ट्रांसफॉर्म्स](https://arxiv.org/abs/2105.03824) जेम्स ली-थॉर्प, जोशुआ आइंस्ली, इल्या एकस्टीन, सैंटियागो ओंटानन द्वारा।
+1. **[FlauBERT](https://huggingface.co/docs/transformers/model_doc/flaubert)** (CNRS से) साथ वाला पेपर [FlauBERT: Unsupervised Language Model Pre-training for French](https://arxiv.org/abs/1912.05372) Hang Le, Loïc Vial, Jibril Frej, Vincent Segonne, Maximin Coavoux, बेंजामिन लेकोउटेक्स, अलेक्जेंड्रे अल्लाउज़ेन, बेनोइट क्रैबे, लॉरेंट बेसेसियर, डिडिएर श्वाब द्वारा।
+1. **[FLAVA](https://huggingface.co/docs/transformers/model_doc/flava)** [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) साथ वाला पेपर अमनप्रीत सिंह, रोंगहांग हू, वेदानुज गोस्वामी, गुइल्यूम कुएरॉन, वोज्शिएक गालुबा, मार्कस रोहरबैक, और डौवे कीला द्वारा।
+1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (गूगल रिसर्च से) साथ वाला पेपर [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) जेम्स ली-थॉर्प, जोशुआ आइंस्ली, इल्या एकस्टीन, सैंटियागो ओंटानन द्वारा।
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (Microsoft Research से) Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao. द्वाराअनुसंधान पत्र [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) के साथ जारी किया गया
-1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (सीएमयू/गूगल ब्रेन से) साथ में कागज [फ़नल-ट्रांसफॉर्मर: कुशल भाषा प्रसंस्करण के लिए अनुक्रमिक अतिरेक को छानना](https://arxiv.org/abs/2006.03236) जिहांग दाई, गुओकुन लाई, यिमिंग यांग, क्वोक वी. ले द्वारा रिहाई।
+1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (सीएमयू/गूगल ब्रेन से) साथ में कागज [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) जिहांग दाई, गुओकुन लाई, यिमिंग यांग, क्वोक वी. ले द्वारा रिहाई।
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (ADEPT से) रोहन बाविशी, एरिच एलसेन, कर्टिस हॉथोर्न, मैक्सवेल नी, ऑगस्टस ओडेना, अरुशी सोमानी, सागनाक तासिरलार  [blog post](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (Google से) the Gemma Google team. द्वाराअनुसंधान पत्र [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) के साथ जारी किया गया
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (Google से) the Gemma Google team. द्वाराअनुसंधान पत्र [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) के साथ जारी किया गया
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
-1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (KAIST से) साथ वाला पेपर [वर्टिकल कटडेप्थ के साथ मोनोकुलर डेप्थ एस्टीमेशन के लिए ग्लोबल-लोकल पाथ नेटवर्क्स](https://arxiv.org/abs/2201.07436) डोयोन किम, वूंगह्युन गा, प्युंगवान आह, डोंगग्यू जू, सेहवान चुन, जुनमो किम द्वारा।
-1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (OpenAI से) साथ में दिया गया पेपर [जेनरेटिव प्री-ट्रेनिंग द्वारा भाषा की समझ में सुधार](https://openai.com/research/language-unsupervised/) एलेक रैडफोर्ड, कार्तिक नरसिम्हन, टिम सालिमन्स और इल्या सुत्स्केवर द्वारा।
+1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (KAIST से) साथ वाला पेपर [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) डोयोन किम, वूंगह्युन गा, प्युंगवान आह, डोंगग्यू जू, सेहवान चुन, जुनमो किम द्वारा।
+1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (OpenAI से) साथ में दिया गया पेपर [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) एलेक रैडफोर्ड, कार्तिक नरसिम्हन, टिम सालिमन्स और इल्या सुत्स्केवर द्वारा।
 1. **[GPT Neo](https://huggingface.co/docs/transformers/model_doc/gpt_neo)** (EleutherAI से) रिपॉजिटरी के साथ [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) रिलीज। सिड ब्लैक, स्टेला बिडरमैन, लियो गाओ, फिल वांग और कॉनर लेही द्वारा पोस्ट किया गया।
-1. **[GPT NeoX](https://huggingface.co/docs/transformers/model_doc/gpt_neox)** (EleutherAI से) पेपर के साथ जारी किया गया [GPT-NeoX-20B: एक ओपन-सोर्स ऑटोरेग्रेसिव लैंग्वेज मॉडल](https://arxiv.org/abs/2204.06745) सिड ब्लैक, स्टेला बिडरमैन, एरिक हैलाहन, क्वेंटिन एंथोनी, लियो गाओ, लॉरेंस गोल्डिंग, होरेस हे, कॉनर लेही, काइल मैकडोनेल, जेसन फांग, माइकल पाइलर, यूएसवीएसएन साई प्रशांत द्वारा , शिवांशु पुरोहित, लारिया रेनॉल्ड्स, जोनाथन टो, बेन वांग, सैमुअल वेनबैक
+1. **[GPT NeoX](https://huggingface.co/docs/transformers/model_doc/gpt_neox)** (EleutherAI से) पेपर के साथ जारी किया गया [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) सिड ब्लैक, स्टेला बिडरमैन, एरिक हैलाहन, क्वेंटिन एंथोनी, लियो गाओ, लॉरेंस गोल्डिंग, होरेस हे, कॉनर लेही, काइल मैकडोनेल, जेसन फांग, माइकल पाइलर, यूएसवीएसएन साई प्रशांत द्वारा , शिवांशु पुरोहित, लारिया रेनॉल्ड्स, जोनाथन टो, बेन वांग, सैमुअल वेनबैक
 1. **[GPT NeoX Japanese](https://huggingface.co/docs/transformers/model_doc/gpt_neox_japanese)** (अबेजा के जरिए) शिन्या ओटानी, ताकायोशी मकाबे, अनुज अरोड़ा, क्यो हटोरी द्वारा।
-1. **[GPT-2](https://huggingface.co/docs/transformers/model_doc/gpt2)** (ओपनएआई से) साथ में पेपर [लैंग्वेज मॉडल्स अनसुपरवाइज्ड मल्टीटास्क लर्नर्स हैं](https://openai.com/research/better-language-models/) एलेक रैडफोर्ड, जेफरी वू, रेवन चाइल्ड, डेविड लुआन, डारियो एमोडी द्वारा और इल्या सुत्सकेवर ने पोस्ट किया।
+1. **[GPT-2](https://huggingface.co/docs/transformers/model_doc/gpt2)** (ओपनएआई से) साथ में पेपर [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/) एलेक रैडफोर्ड, जेफरी वू, रेवन चाइल्ड, डेविड लुआन, डारियो एमोडी द्वारा और इल्या सुत्सकेवर ने पोस्ट किया।
 1. **[GPT-J](https://huggingface.co/docs/transformers/model_doc/gptj)** (EleutherAI से) साथ वाला पेपर [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) बेन वांग और अरन कोमात्सुजाकी द्वारा।
 1. **[GPT-Sw3](https://huggingface.co/docs/transformers/model_doc/gpt-sw3)** (from AI-Sweden) released with the paper [Lessons Learned from GPT-SW3: Building the First Large-Scale Generative Language Model for Swedish](http://www.lrec-conf.org/proceedings/lrec2022/pdf/2022.lrec-1.376.pdf) by Ariel Ekgren, Amaru Cuba Gyllensten, Evangelia Gogoulou, Alice Heiman, Severine Verlinden, Joey Öhman, Fredrik Carlsson, Magnus Sahlgren.
 1. **[GPTBigCode](https://huggingface.co/docs/transformers/model_doc/gpt_bigcode)** (BigCode से) Loubna Ben Allal, Raymond Li, Denis Kocetkov, Chenghao Mou, Christopher Akiki, Carlos Munoz Ferrandis, Niklas Muennighoff, Mayank Mishra, Alex Gu, Manan Dey, Logesh Kumar Umapathi, Carolyn Jane Anderson, Yangtian Zi, Joel Lamy Poirier, Hailey Schoelkopf, Sergey Troshin, Dmitry Abulkhanov, Manuel Romero, Michael Lappert, Francesco De Toni, Bernardo García del Río, Qian Liu, Shamik Bose, Urvashi Bhattacharyya, Terry Yue Zhuo, Ian Yu, Paulo Villegas, Marco Zocca, Sourab Mangrulkar, David Lansky, Huu Nguyen, Danish Contractor, Luis Villa, Jia Li, Dzmitry Bahdanau, Yacine Jernite, Sean Hughes, Daniel Fried, Arjun Guha, Harm de Vries, Leandro von Werra. द्वाराअनुसंधान पत्र [SantaCoder: don't reach for the stars!](https://arxiv.org/abs/2301.03988) के साथ जारी किया गया
 1. **[GPTSAN-japanese](https://huggingface.co/docs/transformers/model_doc/gptsan-japanese)** released in the repository [tanreinama/GPTSAN](https://github.com/tanreinama/GPTSAN/blob/main/report/model.md) by Toshiyuki Sakamoto(tanreinama).
 1. **[Graphormer](https://huggingface.co/docs/transformers/model_doc/graphormer)** (from Microsoft) released with the paper [Do Transformers Really Perform Bad for Graph Representation?](https://arxiv.org/abs/2106.05234) by Chengxuan Ying, Tianle Cai, Shengjie Luo, Shuxin Zheng, Guolin Ke, Di He, Yanming Shen, Tie-Yan Liu.
-1. **[GroupViT](https://huggingface.co/docs/transformers/model_doc/groupvit)** (UCSD, NVIDIA से) साथ में कागज [GroupViT: टेक्स्ट सुपरविजन से सिमेंटिक सेगमेंटेशन इमर्जेस](https://arxiv.org/abs/2202.11094) जियारुई जू, शालिनी डी मेलो, सिफ़ी लियू, वोनमिन बायन, थॉमस ब्रेउएल, जान कौट्ज़, ज़ियाओलोंग वांग द्वारा।
+1. **[GroupViT](https://huggingface.co/docs/transformers/model_doc/groupvit)** (UCSD, NVIDIA से) साथ में कागज [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) जियारुई जू, शालिनी डी मेलो, सिफ़ी लियू, वोनमिन बायन, थॉमस ब्रेउएल, जान कौट्ज़, ज़ियाओलोंग वांग द्वारा।
 1. **[HerBERT](https://huggingface.co/docs/transformers/model_doc/herbert)** (Allegro.pl, AGH University of Science and Technology से) Piotr Rybak, Robert Mroczkowski, Janusz Tracz, Ireneusz Gawlik. द्वाराअनुसंधान पत्र [KLEJ: Comprehensive Benchmark for Polish Language Understanding](https://www.aclweb.org/anthology/2020.acl-main.111.pdf) के साथ जारी किया गया
-1. **[Hubert](https://huggingface.co/docs/transformers/model_doc/hubert)** (फेसबुक से) साथ में पेपर [ह्यूबर्ट: सेल्फ सुपरवाइज्ड स्पीच रिप्रेजेंटेशन लर्निंग बाय मास्क्ड प्रेडिक्शन ऑफ हिडन यूनिट्स](https://arxiv.org/abs/2106.07447) वेई-निंग सू, बेंजामिन बोल्टे, याओ-हंग ह्यूबर्ट त्साई, कुशाल लखोटिया, रुस्लान सालाखुतदीनोव, अब्देलरहमान मोहम्मद द्वारा।
+1. **[Hubert](https://huggingface.co/docs/transformers/model_doc/hubert)** (फेसबुक से) साथ में पेपर [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://arxiv.org/abs/2106.07447) वेई-निंग सू, बेंजामिन बोल्टे, याओ-हंग ह्यूबर्ट त्साई, कुशाल लखोटिया, रुस्लान सालाखुतदीनोव, अब्देलरहमान मोहम्मद द्वारा।
 1. **[I-BERT](https://huggingface.co/docs/transformers/model_doc/ibert)** (बर्कले से) साथ में कागज [I-BERT: Integer-only BERT Quantization](https://arxiv.org/abs/2101.01321) सेहून किम, अमीर घोलमी, ज़ेवेई याओ, माइकल डब्ल्यू महोनी, कर्ट केटज़र द्वारा।
 1. **[IDEFICS](https://huggingface.co/docs/transformers/model_doc/idefics)** (from HuggingFace) released with the paper [OBELICS: An Open Web-Scale Filtered Dataset of Interleaved Image-Text Documents](https://huggingface.co/papers/2306.16527) by Hugo Laurençon, Lucile Saulnier, Léo Tronchon, Stas Bekman, Amanpreet Singh, Anton Lozhkov, Thomas Wang, Siddharth Karamcheti, Alexander M. Rush, Douwe Kiela, Matthieu Cord, Victor Sanh.
 1. **[ImageGPT](https://huggingface.co/docs/transformers/model_doc/imagegpt)** (from OpenAI) released with the paper [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) by Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever.
@@ -348,158 +348,160 @@ conda install conda-forge::transformers
 1. **[KOSMOS-2](https://huggingface.co/docs/transformers/model_doc/kosmos-2)** (from Microsoft Research Asia) released with the paper [Kosmos-2: Grounding Multimodal Large Language Models to the World](https://arxiv.org/abs/2306.14824) by Zhiliang Peng, Wenhui Wang, Li Dong, Yaru Hao, Shaohan Huang, Shuming Ma, Furu Wei.
 1. **[LayoutLM](https://huggingface.co/docs/transformers/model_doc/layoutlm)** (from Microsoft Research Asia) released with the paper [LayoutLM: Pre-training of Text and Layout for Document Image Understanding](https://arxiv.org/abs/1912.13318) by Yiheng Xu, Minghao Li, Lei Cui, Shaohan Huang, Furu Wei, Ming Zhou.
 1. **[LayoutLMv2](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (from Microsoft Research Asia) released with the paper [LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding](https://arxiv.org/abs/2012.14740) by Yang Xu, Yiheng Xu, Tengchao Lv, Lei Cui, Furu Wei, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Wanxiang Che, Min Zhang, Lidong Zhou.
-1. **[LayoutLMv3](https://huggingface.co/docs/transformers/model_doc/layoutlmv3)** (माइक्रोसॉफ्ट रिसर्च एशिया से) साथ देने वाला पेपर [लेआउटएलएमवी3: यूनिफाइड टेक्स्ट और इमेज मास्किंग के साथ दस्तावेज़ एआई के लिए पूर्व-प्रशिक्षण](https://arxiv.org/abs/2204.08387) युपन हुआंग, टेंगचाओ लव, लेई कुई, युटोंग लू, फुरु वेई द्वारा पोस्ट किया गया।
+1. **[LayoutLMv3](https://huggingface.co/docs/transformers/model_doc/layoutlmv3)** (माइक्रोसॉफ्ट रिसर्च एशिया से) साथ देने वाला पेपर [LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking](https://arxiv.org/abs/2204.08387) युपन हुआंग, टेंगचाओ लव, लेई कुई, युटोंग लू, फुरु वेई द्वारा पोस्ट किया गया।
 1. **[LayoutXLM](https://huggingface.co/docs/transformers/model_doc/layoutxlm)** (from Microsoft Research Asia) released with the paper [LayoutXLM: Multimodal Pre-training for Multilingual Visually-rich Document Understanding](https://arxiv.org/abs/2104.08836) by Yiheng Xu, Tengchao Lv, Lei Cui, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Furu Wei.
 1. **[LED](https://huggingface.co/docs/transformers/model_doc/led)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (मेटा AI से) साथ वाला पेपर [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) बेन ग्राहम, अलाएल्डिन एल-नौबी, ह्यूगो टौवरन, पियरे स्टॉक, आर्मंड जौलिन, हर्वे जेगौ, मैथिज डूज़ द्वारा।
-1. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (दक्षिण चीन प्रौद्योगिकी विश्वविद्यालय से) साथ में कागज [LiLT: एक सरल लेकिन प्रभावी भाषा-स्वतंत्र लेआउट ट्रांसफार्मर संरचित दस्तावेज़ समझ के लिए](https://arxiv.org/abs/2202.13669) जियापेंग वांग, लियानवेन जिन, काई डिंग द्वारा पोस्ट किया गया।
+1. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (दक्षिण चीन प्रौद्योगिकी विश्वविद्यालय से) साथ में कागज [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) जियापेंग वांग, लियानवेन जिन, काई डिंग द्वारा पोस्ट किया गया।
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (The FAIR team of Meta AI से) Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample. द्वाराअनुसंधान पत्र [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) के साथ जारी किया गया
-1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (The FAIR team of Meta AI से) Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.. द्वाराअनुसंधान पत्र [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/XXX) के साथ जारी किया गया
+1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (The FAIR team of Meta AI से) Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.. द्वाराअनुसंधान पत्र [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) के साथ जारी किया गया
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (Microsoft Research & University of Wisconsin-Madison से) Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee. द्वाराअनुसंधान पत्र [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) के साथ जारी किया गया
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (Microsoft Research & University of Wisconsin-Madison से) Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee. द्वाराअनुसंधान पत्र [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) के साथ जारी किया गया
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (मैंडी गुओ, जोशुआ आइंस्ली, डेविड यूथस, सैंटियागो ओंटानन, जियानमो नि, यूं-हुआन सुंग, यिनफेई यांग द्वारा पोस्ट किया गया।
-1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (स्टूडियो औसिया से) साथ में पेपर [LUKE: डीप कॉन्टेक्स्टुअलाइज्ड एंटिटी रिप्रेजेंटेशन विद एंटिटी-अवेयर सेल्फ-अटेंशन](https://arxiv.org/abs/2010.01057) Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto द्वारा।
-1. **[LXMERT](https://huggingface.co/docs/transformers/model_doc/lxmert)** (UNC चैपल हिल से) साथ में पेपर [LXMERT: ओपन-डोमेन क्वेश्चन के लिए ट्रांसफॉर्मर से क्रॉस-मोडलिटी एनकोडर रिप्रेजेंटेशन सीखना Answering](https://arxiv.org/abs/1908.07490) हाओ टैन और मोहित बंसल द्वारा।
+1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (स्टूडियो औसिया से) साथ में पेपर [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto द्वारा।
+1. **[LXMERT](https://huggingface.co/docs/transformers/model_doc/lxmert)** (UNC चैपल हिल से) साथ में पेपर [LXMERT: Learning Cross-Modality Encoder Representations from Transformers for Open-Domain Question Answering](https://arxiv.org/abs/1908.07490) हाओ टैन और मोहित बंसल द्वारा।
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
-1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (फेसबुक से) साथ देने वाला पेपर [बियॉन्ड इंग्लिश-सेंट्रिक मल्टीलिंगुअल मशीन ट्रांसलेशन](https://arxiv.org/एब्स/2010.11125) एंजेला फैन, श्रुति भोसले, होल्गर श्वेन्क, झी मा, अहमद अल-किश्की, सिद्धार्थ गोयल, मनदीप बैनेस, ओनूर सेलेबी, गुइल्लाम वेन्जेक, विश्रव चौधरी, नमन गोयल, टॉम बर्च, विटाली लिपचिंस्की, सर्गेई एडुनोव, एडौर्ड द्वारा ग्रेव, माइकल औली, आर्मंड जौलिन द्वारा पोस्ट किया गया।
+1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (फेसबुक से) साथ देने वाला पेपर [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) एंजेला फैन, श्रुति भोसले, होल्गर श्वेन्क, झी मा, अहमद अल-किश्की, सिद्धार्थ गोयल, मनदीप बैनेस, ओनूर सेलेबी, गुइल्लाम वेन्जेक, विश्रव चौधरी, नमन गोयल, टॉम बर्च, विटाली लिपचिंस्की, सर्गेई एडुनोव, एडौर्ड द्वारा ग्रेव, माइकल औली, आर्मंड जौलिन द्वारा पोस्ट किया गया।
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (Albert Gu and Tri Dao से) Albert Gu and Tri Dao. द्वाराअनुसंधान पत्र [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) के साथ जारी किया गया
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (Albert Gu and Tri Dao से) Albert Gu and Tri Dao. द्वाराअनुसंधान पत्र [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) के साथ जारी किया गया
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Jörg द्वारा [OPUS](http://opus.nlpl.eu/) डेटा से प्रशिक्षित मशीनी अनुवाद मॉडल पोस्ट किया गया टाइडेमैन द्वारा। [मैरियन फ्रेमवर्क](https://marian-nmt.github.io/) माइक्रोसॉफ्ट ट्रांसलेटर टीम द्वारा विकसित।
-1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (माइक्रोसॉफ्ट रिसर्च एशिया से) साथ में पेपर [मार्कअपएलएम: विजुअली-रिच डॉक्यूमेंट अंडरस्टैंडिंग के लिए टेक्स्ट और मार्कअप लैंग्वेज का प्री-ट्रेनिंग](https://arxiv.org/abs/2110.08518) जुनलॉन्ग ली, यिहेंग जू, लेई कुई, फुरु द्वारा वी द्वारा पोस्ट किया गया।
+1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (माइक्रोसॉफ्ट रिसर्च एशिया से) साथ में पेपर [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) जुनलॉन्ग ली, यिहेंग जू, लेई कुई, फुरु द्वारा वी द्वारा पोस्ट किया गया।
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (FAIR and UIUC से) Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar. द्वाराअनुसंधान पत्र [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) के साथ जारी किया गया
-1. **[MaskFormer](https://huggingface.co/docs/transformers/model_doc/maskformer)** (मेटा और UIUC से) पेपर के साथ जारी किया गया [प्रति-पिक्सेल वर्गीकरण वह सब नहीं है जिसकी आपको सिमेंटिक सेगमेंटेशन की आवश्यकता है](https://arxiv.org/abs/2107.06278) बोवेन चेंग, अलेक्जेंडर जी. श्विंग, अलेक्जेंडर किरिलोव द्वारा >>>>>> रिबेस ठीक करें
+1. **[MaskFormer](https://huggingface.co/docs/transformers/model_doc/maskformer)** (मेटा और UIUC से) पेपर के साथ जारी किया गया [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) बोवेन चेंग, अलेक्जेंडर जी. श्विंग, अलेक्जेंडर किरिलोव द्वारा >>>>>> रिबेस ठीक करें
 1. **[MatCha](https://huggingface.co/docs/transformers/model_doc/matcha)** (Google AI से) Fangyu Liu, Francesco Piccinno, Syrine Krichene, Chenxi Pang, Kenton Lee, Mandar Joshi, Yasemin Altun, Nigel Collier, Julian Martin Eisenschlos. द्वाराअनुसंधान पत्र [MatCha: Enhancing Visual Language Pretraining with Math Reasoning and Chart Derendering](https://arxiv.org/abs/2212.09662) के साथ जारी किया गया
-1. **[mBART](https://huggingface.co/docs/transformers/model_doc/mbart)** (फेसबुक से) साथ में पेपर [न्यूरल मशीन ट्रांसलेशन के लिए मल्टीलिंगुअल डीनोइजिंग प्री-ट्रेनिंग](https://arxiv.org/abs/2001.08210) यिनहान लियू, जियाताओ गु, नमन गोयल, जियान ली, सर्गेई एडुनोव, मार्जन ग़ज़विनिनेजाद, माइक लुईस, ल्यूक ज़ेटलमॉयर द्वारा।
-1. **[mBART-50](https://huggingface.co/docs/transformers/model_doc/mbart)** (फेसबुक से) साथ में पेपर [एक्स्टेंसिबल बहुभाषी प्रीट्रेनिंग और फाइनट्यूनिंग के साथ बहुभाषी अनुवाद](https://arxiv.org/abs/2008.00401) युकिंग टैंग, चाउ ट्रान, जियान ली, पेंग-जेन चेन, नमन गोयल, विश्रव चौधरी, जियाताओ गु, एंजेला फैन द्वारा।
+1. **[mBART](https://huggingface.co/docs/transformers/model_doc/mbart)** (फेसबुक से) साथ में पेपर [Multilingual Denoising Pre-training for Neural Machine Translation](https://arxiv.org/abs/2001.08210) यिनहान लियू, जियाताओ गु, नमन गोयल, जियान ली, सर्गेई एडुनोव, मार्जन ग़ज़विनिनेजाद, माइक लुईस, ल्यूक ज़ेटलमॉयर द्वारा।
+1. **[mBART-50](https://huggingface.co/docs/transformers/model_doc/mbart)** (फेसबुक से) साथ में पेपर [Multilingual Translation with Extensible Multilingual Pretraining and Finetuning](https://arxiv.org/abs/2008.00401) युकिंग टैंग, चाउ ट्रान, जियान ली, पेंग-जेन चेन, नमन गोयल, विश्रव चौधरी, जियाताओ गु, एंजेला फैन द्वारा।
 1. **[MEGA](https://huggingface.co/docs/transformers/model_doc/mega)** (Facebook से) Xuezhe Ma, Chunting Zhou, Xiang Kong, Junxian He, Liangke Gui, Graham Neubig, Jonathan May, and Luke Zettlemoyer. द्वाराअनुसंधान पत्र [Mega: Moving Average Equipped Gated Attention](https://arxiv.org/abs/2209.10655) के साथ जारी किया गया
-1. **[Megatron-BERT](https://huggingface.co/docs/transformers/model_doc/megatron-bert)** (NVIDIA से) कागज के साथ [Megatron-LM: मॉडल का उपयोग करके बहु-अरब पैरामीटर भाषा मॉडल का प्रशिक्षण Parallelism](https://arxiv.org/abs/1909.08053) मोहम्मद शोएबी, मोस्टोफा पटवारी, राउल पुरी, पैट्रिक लेग्रेस्ले, जेरेड कैस्पर और ब्रायन कैटानज़ारो द्वारा।
-1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (NVIDIA से) साथ वाला पेपर [Megatron-LM: ट्रेनिंग मल्टी-बिलियन पैरामीटर लैंग्वेज मॉडल्स यूजिंग मॉडल पैरेललिज़्म](https://arxiv.org/abs/1909.08053) मोहम्मद शोएबी, मोस्टोफा पटवारी, राउल पुरी, पैट्रिक लेग्रेस्ले, जेरेड कैस्पर और ब्रायन कैटानज़ारो द्वारा पोस्ट किया गया।
+1. **[Megatron-BERT](https://huggingface.co/docs/transformers/model_doc/megatron-bert)** (NVIDIA से) कागज के साथ [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) मोहम्मद शोएबी, मोस्टोफा पटवारी, राउल पुरी, पैट्रिक लेग्रेस्ले, जेरेड कैस्पर और ब्रायन कैटानज़ारो द्वारा।
+1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (NVIDIA से) साथ वाला पेपर [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) मोहम्मद शोएबी, मोस्टोफा पटवारी, राउल पुरी, पैट्रिक लेग्रेस्ले, जेरेड कैस्पर और ब्रायन कैटानज़ारो द्वारा पोस्ट किया गया।
 1. **[MGP-STR](https://huggingface.co/docs/transformers/model_doc/mgp-str)** (Alibaba Research से) Peng Wang, Cheng Da, and Cong Yao. द्वाराअनुसंधान पत्र [Multi-Granularity Prediction for Scene Text Recognition](https://arxiv.org/abs/2209.03592) के साथ जारी किया गया
 1. **[Mistral](https://huggingface.co/docs/transformers/model_doc/mistral)** (from Mistral AI) by The Mistral AI team: Albert Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lélio Renard Lavaud, Lucile Saulnier, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix, William El Sayed..
 1. **[Mixtral](https://huggingface.co/docs/transformers/model_doc/mixtral)** (from Mistral AI) by The [Mistral AI](https://mistral.ai) team: Albert Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lélio Renard Lavaud, Lucile Saulnier, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix, William El Sayed.
-1. **[mLUKE](https://huggingface.co/docs/transformers/model_doc/mluke)** (फ्रॉम Studio Ousia) साथ में पेपर [mLUKE: द पावर ऑफ एंटिटी रिप्रेजेंटेशन इन मल्टीलिंगुअल प्रीट्रेन्ड लैंग्वेज मॉडल्स](https://arxiv.org/abs/2110.08151) रयोकन री, इकुया यामाडा, और योशिमासा त्सुरोका द्वारा।
+1. **[mLUKE](https://huggingface.co/docs/transformers/model_doc/mluke)** (फ्रॉम Studio Ousia) साथ में पेपर [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) रयोकन री, इकुया यामाडा, और योशिमासा त्सुरोका द्वारा।
 1. **[MMS](https://huggingface.co/docs/transformers/model_doc/mms)** (Facebook से) Vineel Pratap, Andros Tjandra, Bowen Shi, Paden Tomasello, Arun Babu, Sayani Kundu, Ali Elkahky, Zhaoheng Ni, Apoorv Vyas, Maryam Fazel-Zarandi, Alexei Baevski, Yossi Adi, Xiaohui Zhang, Wei-Ning Hsu, Alexis Conneau, Michael Auli. द्वाराअनुसंधान पत्र [Scaling Speech Technology to 1,000+ Languages](https://arxiv.org/abs/2305.13516) के साथ जारी किया गया
-1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (सीएमयू/गूगल ब्रेन से) साथ में कागज [मोबाइलबर्ट: संसाधन-सीमित उपकरणों के लिए एक कॉम्पैक्ट टास्क-अज्ञेय बीईआरटी](https://arxiv.org/abs/2004.02984) Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, और Denny Zhou द्वारा पोस्ट किया गया।
+1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (सीएमयू/गूगल ब्रेन से) साथ में कागज [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, और Denny Zhou द्वारा पोस्ट किया गया।
 1. **[MobileNetV1](https://huggingface.co/docs/transformers/model_doc/mobilenet_v1)** (from Google Inc.) released with the paper [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://arxiv.org/abs/1704.04861) by Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam.
 1. **[MobileNetV2](https://huggingface.co/docs/transformers/model_doc/mobilenet_v2)** (from Google Inc.) released with the paper [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381) by Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen.
-1. **[MobileViT](https://huggingface.co/docs/transformers/model_doc/mobilevit)** (Apple से) साथ में कागज [MobileViT: लाइट-वेट, जनरल-पर्पस, और मोबाइल-फ्रेंडली विजन ट्रांसफॉर्मर](https://arxiv.org/abs/2110.02178) सचिन मेहता और मोहम्मद रस्तगरी द्वारा पोस्ट किया गया।
+1. **[MobileViT](https://huggingface.co/docs/transformers/model_doc/mobilevit)** (Apple से) साथ में कागज [MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer](https://arxiv.org/abs/2110.02178) सचिन मेहता और मोहम्मद रस्तगरी द्वारा पोस्ट किया गया।
 1. **[MobileViTV2](https://huggingface.co/docs/transformers/model_doc/mobilevitv2)** (Apple से) Sachin Mehta and Mohammad Rastegari. द्वाराअनुसंधान पत्र [Separable Self-attention for Mobile Vision Transformers](https://arxiv.org/abs/2206.02680) के साथ जारी किया गया
 1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MPT](https://huggingface.co/docs/transformers/model_doc/mpt)** (MosaiML से) the MosaicML NLP Team. द्वाराअनुसंधान पत्र [llm-foundry](https://github.com/mosaicml/llm-foundry/) के साथ जारी किया गया
-1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (the University of Wisconsin - Madison से) Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh. द्वाराअनुसंधान पत्र [Multi Resolution Analysis (MRA)](https://arxiv.org/abs/2207.10284) के साथ जारी किया गया
-1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (Google AI से) साथ वाला पेपर [mT5: एक व्यापक बहुभाषी पूर्व-प्रशिक्षित टेक्स्ट-टू-टेक्स्ट ट्रांसफॉर्मर](https://arxiv.org/abs/2010.11934) लिंटिंग ज़ू, नोआ कॉन्सटेंट, एडम रॉबर्ट्स, मिहिर काले, रामी अल-रफू, आदित्य सिद्धांत, आदित्य बरुआ, कॉलिन रैफेल द्वारा पोस्ट किया गया।
+1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (the University of Wisconsin - Madison से) Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh. द्वाराअनुसंधान पत्र [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) के साथ जारी किया गया
+1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (Google AI से) साथ वाला पेपर [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) लिंटिंग ज़ू, नोआ कॉन्सटेंट, एडम रॉबर्ट्स, मिहिर काले, रामी अल-रफू, आदित्य सिद्धांत, आदित्य बरुआ, कॉलिन रैफेल द्वारा पोस्ट किया गया।
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
-1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (हुआवेई नूह के आर्क लैब से) साथ में कागज़ [NEZHA: चीनी भाषा समझ के लिए तंत्रिका प्रासंगिक प्रतिनिधित्व](https://arxiv.org/abs/1909.00204) जुन्किउ वेई, ज़ियाओज़े रेन, ज़िआओगुआंग ली, वेनयोंग हुआंग, यी लियाओ, याशेंग वांग, जियाशू लिन, शिन जियांग, जिओ चेन और कुन लियू द्वारा।
-1. **[NLLB](https://huggingface.co/docs/transformers/model_doc/nllb)** (फ्रॉम मेटा) साथ में पेपर [नो लैंग्वेज लेफ्ट बिहाइंड: स्केलिंग ह्यूमन-सेंटेड मशीन ट्रांसलेशन](https://arxiv.org/abs/2207.04672) एनएलएलबी टीम द्वारा प्रकाशित।
+1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (हुआवेई नूह के आर्क लैब से) साथ में कागज़ [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) जुन्किउ वेई, ज़ियाओज़े रेन, ज़िआओगुआंग ली, वेनयोंग हुआंग, यी लियाओ, याशेंग वांग, जियाशू लिन, शिन जियांग, जिओ चेन और कुन लियू द्वारा।
+1. **[NLLB](https://huggingface.co/docs/transformers/model_doc/nllb)** (फ्रॉम मेटा) साथ में पेपर [No Language Left Behind: Scaling Human-Centered Machine Translation](https://arxiv.org/abs/2207.04672) एनएलएलबी टीम द्वारा प्रकाशित।
 1. **[NLLB-MOE](https://huggingface.co/docs/transformers/model_doc/nllb-moe)** (Meta से) the NLLB team. द्वाराअनुसंधान पत्र [No Language Left Behind: Scaling Human-Centered Machine Translation](https://arxiv.org/abs/2207.04672) के साथ जारी किया गया
 1. **[Nougat](https://huggingface.co/docs/transformers/model_doc/nougat)** (Meta AI से) Lukas Blecher, Guillem Cucurull, Thomas Scialom, Robert Stojnic. द्वाराअनुसंधान पत्र [Nougat: Neural Optical Understanding for Academic Documents](https://arxiv.org/abs/2308.13418) के साथ जारी किया गया
-1. **[Nyströmformer](https://huggingface.co/docs/transformers/model_doc/nystromformer)** (विस्कॉन्सिन विश्वविद्यालय - मैडिसन से) साथ में कागज [Nyströmformer: A Nyström- आधारित एल्गोरिथम आत्म-ध्यान का अनुमान लगाने के लिए](https://arxiv.org/abs/2102.03902) युनयांग ज़िओंग, झानपेंग ज़ेंग, रुद्रसिस चक्रवर्ती, मिंगक्सिंग टैन, ग्लेन फंग, यिन ली, विकास सिंह द्वारा पोस्ट किया गया।
+1. **[Nyströmformer](https://huggingface.co/docs/transformers/model_doc/nystromformer)** (विस्कॉन्सिन विश्वविद्यालय - मैडिसन से) साथ में कागज [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) युनयांग ज़िओंग, झानपेंग ज़ेंग, रुद्रसिस चक्रवर्ती, मिंगक्सिंग टैन, ग्लेन फंग, यिन ली, विकास सिंह द्वारा पोस्ट किया गया।
 1. **[OneFormer](https://huggingface.co/docs/transformers/model_doc/oneformer)** (SHI Labs से) पेपर [OneFormer: One Transformer to Rule Universal Image Segmentation](https://arxiv.org/abs/2211.06220) जितेश जैन, जिआचेन ली, मांगटिक चिउ, अली हसनी, निकिता ओरलोव, हम्फ्री शि के द्वारा जारी किया गया है।
 1. **[OpenLlama](https://huggingface.co/docs/transformers/model_doc/open-llama)** (from [s-JoL](https://huggingface.co/s-JoL)) released on GitHub (now removed).
 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)** (Google AI से) साथ में कागज [विज़न ट्रांसफॉर्मर्स के साथ सिंपल ओपन-वोकैबुलरी ऑब्जेक्ट डिटेक्शन](https://arxiv.org/abs/2205.06230) मैथियास मिंडरर, एलेक्सी ग्रिट्सेंको, ऑस्टिन स्टोन, मैक्सिम न्यूमैन, डिर्क वीसेनबोर्न, एलेक्सी डोसोवित्स्की, अरविंद महेंद्रन, अनुराग अर्नब, मुस्तफा देहघानी, ज़ुओरन शेन, जिओ वांग, ज़ियाओहुआ झाई, थॉमस किफ़, और नील हॉल्सबी द्वारा पोस्ट किया गया।
+1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (Google AI से) साथ में कागज [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) मैथियास मिंडरर, एलेक्सी ग्रिट्सेंको, ऑस्टिन स्टोन, मैक्सिम न्यूमैन, डिर्क वीसेनबोर्न, एलेक्सी डोसोवित्स्की, अरविंद महेंद्रन, अनुराग अर्नब, मुस्तफा देहघानी, ज़ुओरन शेन, जिओ वांग, ज़ियाओहुआ झाई, थॉमस किफ़, और नील हॉल्सबी द्वारा पोस्ट किया गया।
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (Google AI से) Matthias Minderer, Alexey Gritsenko, Neil Houlsby. द्वाराअनुसंधान पत्र [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683) के साथ जारी किया गया
 1. **[PatchTSMixer](https://huggingface.co/docs/transformers/model_doc/patchtsmixer)** ( IBM Research से) Vijay Ekambaram, Arindam Jati, Nam Nguyen, Phanwadee Sinthong, Jayant Kalagnanam. द्वाराअनुसंधान पत्र [TSMixer: Lightweight MLP-Mixer Model for Multivariate Time Series Forecasting](https://arxiv.org/pdf/2306.09364.pdf) के साथ जारी किया गया
-1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (IBM से) Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam. द्वाराअनुसंधान पत्र [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf) के साथ जारी किया गया
+1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (IBM से) Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam. द्वाराअनुसंधान पत्र [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/abs/2211.14730) के साथ जारी किया गया
 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/model_doc/pegasus_x)** (Google की ओर से) साथ में दिया गया पेपर [लंबे इनपुट सारांश के लिए ट्रांसफ़ॉर्मरों को बेहतर तरीके से एक्सटेंड करना](https://arxiv.org/abs/2208.04347) जेसन फांग, याओ झाओ, पीटर जे लियू द्वारा।
-1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (दीपमाइंड से) साथ में पेपर [पर्सीवर आईओ: संरचित इनपुट और आउटपुट के लिए एक सामान्य वास्तुकला](https://arxiv.org/abs/2107.14795) एंड्रयू जेगल, सेबेस्टियन बोरग्यूड, जीन-बैप्टिस्ट अलायराक, कार्ल डोर्श, कैटलिन इओनेस्कु, डेविड द्वारा डिंग, स्कंद कोप्पुला, डैनियल ज़ोरान, एंड्रयू ब्रॉक, इवान शेलहैमर, ओलिवियर हेनाफ, मैथ्यू एम। बोट्विनिक, एंड्रयू ज़िसरमैन, ओरिओल विनियल्स, जोआओ कैरेरा द्वारा पोस्ट किया गया।
+1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (Google की ओर से) साथ में दिया गया पेपर [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) जेसन फांग, याओ झाओ, पीटर जे लियू द्वारा।
+1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (दीपमाइंड से) साथ में पेपर [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) एंड्रयू जेगल, सेबेस्टियन बोरग्यूड, जीन-बैप्टिस्ट अलायराक, कार्ल डोर्श, कैटलिन इओनेस्कु, डेविड द्वारा डिंग, स्कंद कोप्पुला, डैनियल ज़ोरान, एंड्रयू ब्रॉक, इवान शेलहैमर, ओलिवियर हेनाफ, मैथ्यू एम। बोट्विनिक, एंड्रयू ज़िसरमैन, ओरिओल विनियल्स, जोआओ कैरेरा द्वारा पोस्ट किया गया।
 1. **[Persimmon](https://huggingface.co/docs/transformers/model_doc/persimmon)** (ADEPT से) Erich Elsen, Augustus Odena, Maxwell Nye, Sağnak Taşırlar, Tri Dao, Curtis Hawthorne, Deepak Moparthi, Arushi Somani. द्वाराअनुसंधान पत्र [blog post](https://www.adept.ai/blog/persimmon-8b) के साथ जारी किया गया
 1. **[Phi](https://huggingface.co/docs/transformers/model_doc/phi)** (from Microsoft) released with the papers - [Textbooks Are All You Need](https://arxiv.org/abs/2306.11644) by Suriya Gunasekar, Yi Zhang, Jyoti Aneja, Caio César Teodoro Mendes, Allie Del Giorno, Sivakanth Gopi, Mojan Javaheripi, Piero Kauffmann, Gustavo de Rosa, Olli Saarikivi, Adil Salim, Shital Shah, Harkirat Singh Behl, Xin Wang, Sébastien Bubeck, Ronen Eldan, Adam Tauman Kalai, Yin Tat Lee and Yuanzhi Li, [Textbooks Are All You Need II: phi-1.5 technical report](https://arxiv.org/abs/2309.05463) by Yuanzhi Li, Sébastien Bubeck, Ronen Eldan, Allie Del Giorno, Suriya Gunasekar and Yin Tat Lee.
-1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (VinAI Research से) कागज के साथ [PhoBERT: वियतनामी के लिए पूर्व-प्रशिक्षित भाषा मॉडल](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) डैट क्वोक गुयेन और अन्ह तुआन गुयेन द्वारा पोस्ट किया गया।
+1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (VinAI Research से) कागज के साथ [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) डैट क्वोक गुयेन और अन्ह तुआन गुयेन द्वारा पोस्ट किया गया।
 1. **[Pix2Struct](https://huggingface.co/docs/transformers/model_doc/pix2struct)** (Google से) Kenton Lee, Mandar Joshi, Iulia Turc, Hexiang Hu, Fangyu Liu, Julian Eisenschlos, Urvashi Khandelwal, Peter Shaw, Ming-Wei Chang, Kristina Toutanova. द्वाराअनुसंधान पत्र [Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding](https://arxiv.org/abs/2210.03347) के साथ जारी किया गया
-1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (UCLA NLP से) साथ वाला पेपर [प्रोग्राम अंडरस्टैंडिंग एंड जेनरेशन के लिए यूनिफाइड प्री-ट्रेनिंग](https://arxiv.org/abs/2103.06333) वसी उद्दीन अहमद, सैकत चक्रवर्ती, बैशाखी रे, काई-वेई चांग द्वारा।
+1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (UCLA NLP से) साथ वाला पेपर [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) वसी उद्दीन अहमद, सैकत चक्रवर्ती, बैशाखी रे, काई-वेई चांग द्वारा।
 1. **[PoolFormer](https://huggingface.co/docs/transformers/model_doc/poolformer)** (from Sea AI Labs) released with the paper [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) by Yu, Weihao and Luo, Mi and Zhou, Pan and Si, Chenyang and Zhou, Yichen and Wang, Xinchao and Feng, Jiashi and Yan, Shuicheng.
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi, Kyogu Lee.
-1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/prophetnet)** (माइक्रोसॉफ्ट रिसर्च से) साथ में पेपर [ProphetNet: प्रेडिक्टिंग फ्यूचर एन-ग्राम फॉर सीक्वेंस-टू-सीक्वेंस प्री-ट्रेनिंग](https://arxiv.org/abs/2001.04063) यू यान, वीज़ेन क्यूई, येयुन गोंग, दयाहेंग लियू, नान डुआन, जिउशेंग चेन, रुओफ़ेई झांग और मिंग झोउ द्वारा पोस्ट किया गया।
+1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/prophetnet)** (माइक्रोसॉफ्ट रिसर्च से) साथ में पेपर [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) यू यान, वीज़ेन क्यूई, येयुन गोंग, दयाहेंग लियू, नान डुआन, जिउशेंग चेन, रुओफ़ेई झांग और मिंग झोउ द्वारा पोस्ट किया गया।
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (Nanjing University, The University of Hong Kong etc. से) Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao. द्वाराअनुसंधान पत्र [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) के साथ जारी किया गया
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc. से) Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao. द्वाराअनुसंधान पत्र [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) के साथ जारी किया गया 
-1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (NVIDIA से) साथ वाला पेपर [डीप लर्निंग इंफ़ेक्शन के लिए इंटीजर क्वांटिज़ेशन: प्रिंसिपल्स एंड एम्पिरिकल इवैल्यूएशन](https://arxiv.org/abs/2004.09602) हाओ वू, पैट्रिक जुड, जिआओजी झांग, मिखाइल इसेव और पॉलियस माइकेविसियस द्वारा।
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc. से) Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao. द्वाराअनुसंधान पत्र [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) के साथ जारी किया गया 
+1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (NVIDIA से) साथ वाला पेपर [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) हाओ वू, पैट्रिक जुड, जिआओजी झांग, मिखाइल इसेव और पॉलियस माइकेविसियस द्वारा।
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (the Qwen team, Alibaba Group से) Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu. द्वाराअनुसंधान पत्र [Qwen Technical Report](https://arxiv.org/abs/2309.16609) के साथ जारी किया गया
-1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (फेसबुक से) साथ में कागज [रिट्रीवल-ऑगमेंटेड जेनरेशन फॉर नॉलेज-इंटेंसिव एनएलपी टास्क](https://arxiv.org/abs/2005.11401) पैट्रिक लुईस, एथन पेरेज़, अलेक्जेंड्रा पिक्टस, फैबियो पेट्रोनी, व्लादिमीर कारपुखिन, नमन गोयल, हेनरिक कुटलर, माइक लुईस, वेन-ताउ यिह, टिम रॉकटाशेल, सेबस्टियन रिडेल, डौवे कीला द्वारा।
-1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (Google अनुसंधान से) केल्विन गु, केंटन ली, ज़ोरा तुंग, पानुपोंग पसुपत और मिंग-वेई चांग द्वारा साथ में दिया गया पेपर [REALM: रिट्रीवल-ऑगमेंटेड लैंग्वेज मॉडल प्री-ट्रेनिंग](https://arxiv.org/abs/2002.08909)।
+1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (फेसबुक से) साथ में कागज [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) पैट्रिक लुईस, एथन पेरेज़, अलेक्जेंड्रा पिक्टस, फैबियो पेट्रोनी, व्लादिमीर कारपुखिन, नमन गोयल, हेनरिक कुटलर, माइक लुईस, वेन-ताउ यिह, टिम रॉकटाशेल, सेबस्टियन रिडेल, डौवे कीला द्वारा।
+1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (Google अनुसंधान से) केल्विन गु, केंटन ली, ज़ोरा तुंग, पानुपोंग पसुपत और मिंग-वेई चांग द्वारा साथ में दिया गया पेपर [REALM: Retrieval-Augmented Language Model Pre-Training](https://arxiv.org/abs/2002.08909)।
 1. **[Reformer](https://huggingface.co/docs/transformers/model_doc/reformer)** (from Google Research) released with the paper [Reformer: The Efficient Transformer](https://arxiv.org/abs/2001.04451) by Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya.
-1. **[RegNet](https://huggingface.co/docs/transformers/model_doc/regnet)** (META रिसर्च से) [डिज़ाइनिंग नेटवर्क डिज़ाइन स्पेस](https://arxiv.org/) पेपर के साथ जारी किया गया एब्स/2003.13678) इलिजा राडोसावोविक, राज प्रतीक कोसाराजू, रॉस गिर्शिक, कैमिंग ही, पिओटर डॉलर द्वारा।
-1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (गूगल रिसर्च से) साथ वाला पेपर [पूर्व-प्रशिक्षित भाषा मॉडल में एम्बेडिंग कपलिंग पर पुनर्विचार](https://arxiv.org/pdf/2010.12821.pdf) ह्युंग वोन चुंग, थिबॉल्ट फ़ेवरी, हेनरी त्साई, एम. जॉनसन, सेबेस्टियन रुडर द्वारा।
-1. **[ResNet](https://huggingface.co/docs/transformers/model_doc/resnet)** (माइक्रोसॉफ्ट रिसर्च से) [डीप रेसिडुअल लर्निंग फॉर इमेज रिकग्निशन](https://arxiv.org/abs/1512.03385) कैमिंग हे, जियांग्यु झांग, शाओकिंग रेन, जियान सन द्वारा।
-1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (फेसबुक से), साथ में कागज [मजबूत रूप से अनुकूलित BERT प्रीट्रेनिंग दृष्टिकोण](https://arxiv.org/abs/1907.11692) यिनहान लियू, मायल ओट, नमन गोयल, जिंगफेई डू, मंदार जोशी, डैनकी चेन, ओमर लेवी, माइक लुईस, ल्यूक ज़ेटलमॉयर, वेसेलिन स्टोयानोव द्वारा।
+1. **[RegNet](https://huggingface.co/docs/transformers/model_doc/regnet)** (META रिसर्च से) [Designing Network Design Space](https://arxiv.org/abs/2003.13678) पेपर के साथ जारी किया गया एब्स/2003.13678) इलिजा राडोसावोविक, राज प्रतीक कोसाराजू, रॉस गिर्शिक, कैमिंग ही, पिओटर डॉलर द्वारा।
+1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (गूगल रिसर्च से) साथ वाला पेपर [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/abs/2010.12821) ह्युंग वोन चुंग, थिबॉल्ट फ़ेवरी, हेनरी त्साई, एम. जॉनसन, सेबेस्टियन रुडर द्वारा।
+1. **[ResNet](https://huggingface.co/docs/transformers/model_doc/resnet)** (माइक्रोसॉफ्ट रिसर्च से) [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) कैमिंग हे, जियांग्यु झांग, शाओकिंग रेन, जियान सन द्वारा।
+1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (फेसबुक से), साथ में कागज [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://arxiv.org/abs/1907.11692) यिनहान लियू, मायल ओट, नमन गोयल, जिंगफेई डू, मंदार जोशी, डैनकी चेन, ओमर लेवी, माइक लुईस, ल्यूक ज़ेटलमॉयर, वेसेलिन स्टोयानोव द्वारा।
 1. **[RoBERTa-PreLayerNorm](https://huggingface.co/docs/transformers/model_doc/roberta-prelayernorm)** (from Facebook) released with the paper [fairseq: A Fast, Extensible Toolkit for Sequence Modeling](https://arxiv.org/abs/1904.01038) by Myle Ott, Sergey Edunov, Alexei Baevski, Angela Fan, Sam Gross, Nathan Ng, David Grangier, Michael Auli.
 1. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (from WeChatAI) released with the paper [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) by HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou.
-1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (झुईई टेक्नोलॉजी से), साथ में पेपर [रोफॉर्मर: रोटरी पोजिशन एंबेडिंग के साथ एन्हांस्ड ट्रांसफॉर्मर](https://arxiv.org/pdf/2104.09864v1.pdf) जियानलिन सु और यू लू और शेंगफेंग पैन और बो वेन और युनफेंग लियू द्वारा प्रकाशित।
+1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (झुईई टेक्नोलॉजी से), साथ में पेपर [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/abs/2104.09864) जियानलिन सु और यू लू और शेंगफेंग पैन और बो वेन और युनफेंग लियू द्वारा प्रकाशित।
 1. **[RWKV](https://huggingface.co/docs/transformers/model_doc/rwkv)** (Bo Peng से) Bo Peng. द्वाराअनुसंधान पत्र [this repo](https://github.com/BlinkDL/RWKV-LM) के साथ जारी किया गया
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (Beijing Academy of Artificial Intelligence (BAAI से) Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang. द्वाराअनुसंधान पत्र [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) के साथ जारी किया गया
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (Beijing Academy of Artificial Intelligence (BAAI से) Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang. द्वाराअनुसंधान पत्र [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) के साथ जारी किया गया
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (Meta AI से) Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick. द्वाराअनुसंधान पत्र [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) के साथ जारी किया गया
-1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (ASAPP से) साथ देने वाला पेपर [भाषण पहचान के लिए अनसुपरवाइज्ड प्री-ट्रेनिंग में परफॉर्मेंस-एफिशिएंसी ट्रेड-ऑफ्स](https://arxiv.org/abs/2109.06870) फेलिक्स वू, क्वांगयुन किम, जिंग पैन, क्यू हान, किलियन क्यू. वेनबर्गर, योव आर्टज़ी द्वारा।
-1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (ASAPP से) साथ में पेपर [भाषण पहचान के लिए अनसुपरवाइज्ड प्री-ट्रेनिंग में परफॉर्मेंस-एफिशिएंसी ट्रेड-ऑफ्स](https://arxiv.org/abs/2109.06870) फेलिक्स वू, क्वांगयुन किम, जिंग पैन, क्यू हान, किलियन क्यू. वेनबर्गर, योआव आर्टज़ी द्वारा पोस्ट किया गया।
+1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (ASAPP से) साथ देने वाला पेपर [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) फेलिक्स वू, क्वांगयुन किम, जिंग पैन, क्यू हान, किलियन क्यू. वेनबर्गर, योव आर्टज़ी द्वारा।
+1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (ASAPP से) साथ में पेपर [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) फेलिक्स वू, क्वांगयुन किम, जिंग पैन, क्यू हान, किलियन क्यू. वेनबर्गर, योआव आर्टज़ी द्वारा पोस्ट किया गया।
 1. **[SigLIP](https://huggingface.co/docs/transformers/model_doc/siglip)** (Google AI से) Xiaohua Zhai, Basil Mustafa, Alexander Kolesnikov, Lucas Beyer. द्वाराअनुसंधान पत्र [Sigmoid Loss for Language Image Pre-Training](https://arxiv.org/abs/2303.15343) के साथ जारी किया गया
 1. **[SpeechT5](https://huggingface.co/docs/transformers/model_doc/speecht5)** (from Microsoft Research) released with the paper [SpeechT5: Unified-Modal Encoder-Decoder Pre-Training for Spoken Language Processing](https://arxiv.org/abs/2110.07205) by Junyi Ao, Rui Wang, Long Zhou, Chengyi Wang, Shuo Ren, Yu Wu, Shujie Liu, Tom Ko, Qing Li, Yu Zhang, Zhihua Wei, Yao Qian, Jinyu Li, Furu Wei.
-1. **[SpeechToTextTransformer](https://huggingface.co/docs/transformers/model_doc/speech_to_text)** (फेसबुक से), साथ में पेपर [फेयरसेक S2T: फास्ट स्पीच-टू-टेक्स्ट मॉडलिंग विद फेयरसेक](https://arxiv.org/abs/2010.05171) चांगहान वांग, यूं तांग, जुताई मा, ऐनी वू, दिमित्रो ओखोनको, जुआन पिनो द्वारा पोस्ट किया गया。
-1. **[SpeechToTextTransformer2](https://huggingface.co/docs/transformers/model_doc/speech_to_text_2)** (फेसबुक से) साथ में पेपर [लार्ज-स्केल सेल्फ- एंड सेमी-सुपरवाइज्ड लर्निंग फॉर स्पीच ट्रांसलेशन](https://arxiv.org/abs/2104.06678) चांगहान वांग, ऐनी वू, जुआन पिनो, एलेक्सी बेवस्की, माइकल औली, एलेक्सिस द्वारा Conneau द्वारा पोस्ट किया गया।
-1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (तेल अवीव यूनिवर्सिटी से) साथ में पेपर [स्पैन सिलेक्शन को प्री-ट्रेनिंग करके कुछ-शॉट क्वेश्चन आंसरिंग](https://arxiv.org/abs/2101.00438) ओरि राम, युवल कर्स्टन, जोनाथन बेरेंट, अमीर ग्लोबर्सन, ओमर लेवी द्वारा।
-1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (बर्कले से) कागज के साथ [SqueezeBERT: कुशल तंत्रिका नेटवर्क के बारे में NLP को कंप्यूटर विज़न क्या सिखा सकता है?](https://arxiv.org/abs/2006.11316) फॉरेस्ट एन. इनडोला, अल्बर्ट ई. शॉ, रवि कृष्णा, और कर्ट डब्ल्यू. केटज़र द्वारा।
+1. **[SpeechToTextTransformer](https://huggingface.co/docs/transformers/model_doc/speech_to_text)** (फेसबुक से), साथ में पेपर [fairseq S2T: Fast Speech-to-Text Modeling with fairseq](https://arxiv.org/abs/2010.05171) चांगहान वांग, यूं तांग, जुताई मा, ऐनी वू, दिमित्रो ओखोनको, जुआन पिनो द्वारा पोस्ट किया गया。
+1. **[SpeechToTextTransformer2](https://huggingface.co/docs/transformers/model_doc/speech_to_text_2)** (फेसबुक से) साथ में पेपर [Large-Scale Self- and Semi-Supervised Learning for Speech Translation](https://arxiv.org/abs/2104.06678) चांगहान वांग, ऐनी वू, जुआन पिनो, एलेक्सी बेवस्की, माइकल औली, एलेक्सिस द्वारा Conneau द्वारा पोस्ट किया गया।
+1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (तेल अवीव यूनिवर्सिटी से) साथ में पेपर [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) ओरि राम, युवल कर्स्टन, जोनाथन बेरेंट, अमीर ग्लोबर्सन, ओमर लेवी द्वारा।
+1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (बर्कले से) कागज के साथ [SqueezeBERT: What can computer vision teach NLP about efficient neural networks?](https://arxiv.org/abs/2006.11316) फॉरेस्ट एन. इनडोला, अल्बर्ट ई. शॉ, रवि कृष्णा, और कर्ट डब्ल्यू. केटज़र द्वारा।
 1. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report)](https://stability.wandb.io/stability-llm/stable-lm/reports/StableLM-3B-4E1T--VmlldzoyMjU4?accessToken=u3zujipenkx5g7rtcj9qojjgxpconyjktjkli2po09nffrffdhhchq045vp0wyfo) by  Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu. 
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (MBZUAI से) Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan. द्वाराअनुसंधान पत्र [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) के साथ जारी किया गया
-1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (माइक्रोसॉफ्ट से) साथ में कागज [स्वाइन ट्रांसफॉर्मर: शिफ्टेड विंडोज का उपयोग कर पदानुक्रमित विजन ट्रांसफॉर्मर](https://arxiv.org/abs/2103.14030) ज़ी लियू, युटोंग लिन, यू काओ, हान हू, यिक्सुआन वेई, झेंग झांग, स्टीफन लिन, बैनिंग गुओ द्वारा।
-1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (Microsoft से) साथ वाला पेपर [Swin Transformer V2: स्केलिंग अप कैपेसिटी एंड रेजोल्यूशन](https://arxiv.org/abs/2111.09883) ज़ी लियू, हान हू, युटोंग लिन, ज़ुलिआंग याओ, ज़ेंडा ज़ी, यिक्सुआन वेई, जिया निंग, यू काओ, झेंग झांग, ली डोंग, फुरु वेई, बैनिंग गुओ द्वारा।
+1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (माइक्रोसॉफ्ट से) साथ में कागज [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) ज़ी लियू, युटोंग लिन, यू काओ, हान हू, यिक्सुआन वेई, झेंग झांग, स्टीफन लिन, बैनिंग गुओ द्वारा।
+1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (Microsoft से) साथ वाला पेपर [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) ज़ी लियू, हान हू, युटोंग लिन, ज़ुलिआंग याओ, ज़ेंडा ज़ी, यिक्सुआन वेई, जिया निंग, यू काओ, झेंग झांग, ली डोंग, फुरु वेई, बैनिंग गुओ द्वारा।
 1. **[Swin2SR](https://huggingface.co/docs/transformers/model_doc/swin2sr)** (from University of Würzburg) released with the paper [Swin2SR: SwinV2 Transformer for Compressed Image Super-Resolution and Restoration](https://arxiv.org/abs/2209.11345) by Marcos V. Conde, Ui-Jin Choi, Maxime Burchi, Radu Timofte.
 1. **[SwitchTransformers](https://huggingface.co/docs/transformers/model_doc/switch_transformers)** (from Google) released with the paper [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://arxiv.org/abs/2101.03961) by William Fedus, Barret Zoph, Noam Shazeer.
-1. **[T5](https://huggingface.co/docs/transformers/model_doc/t5)** (来自 Google AI)कॉलिन रैफेल और नोम शज़ीर और एडम रॉबर्ट्स और कैथरीन ली और शरण नारंग और माइकल मटेना द्वारा साथ में पेपर [एक एकीकृत टेक्स्ट-टू-टेक्स्ट ट्रांसफॉर्मर के साथ स्थानांतरण सीखने की सीमा की खोज](https://arxiv.org/abs/1910.10683) और यांकी झोउ और वेई ली और पीटर जे लियू।
-1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (Google AI से) साथ वाला पेपर [google-research/text-to-text-transfer- ट्रांसफॉर्मर](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) कॉलिन रैफेल और नोम शज़ीर और एडम रॉबर्ट्स और कैथरीन ली और शरण नारंग द्वारा और माइकल मटेना और यांकी झोउ और वेई ली और पीटर जे लियू।
-1. **[Table Transformer](https://huggingface.co/docs/transformers/model_doc/table-transformer)** (माइक्रोसॉफ्ट रिसर्च से) साथ में पेपर [पबटेबल्स-1एम: टूवर्ड्स कॉम्प्रिहेंसिव टेबल एक्सट्रैक्शन फ्रॉम अनस्ट्रक्चर्ड डॉक्यूमेंट्स](https://arxiv.org/abs/2110.00061) ब्रैंडन स्मॉक, रोहित पेसाला, रॉबिन अब्राहम द्वारा पोस्ट किया गया।
-1. **[TAPAS](https://huggingface.co/docs/transformers/model_doc/tapas)** (Google AI से) साथ में कागज [TAPAS: पूर्व-प्रशिक्षण के माध्यम से कमजोर पर्यवेक्षण तालिका पार्सिंग](https://arxiv.org/abs/2004.02349) जोनाथन हर्ज़िग, पावेल क्रिज़िस्तोफ़ नोवाक, थॉमस मुलर, फ्रांसेस्को पिकिन्नो और जूलियन मार्टिन ईसेन्च्लोस द्वारा।
-1. **[TAPEX](https://huggingface.co/docs/transformers/model_doc/tapex)** (माइक्रोसॉफ्ट रिसर्च से) साथ में पेपर [TAPEX: टेबल प्री-ट्रेनिंग थ्रू लर्निंग अ न्यूरल SQL एक्ज़ीक्यूटर](https://arxiv.org/abs/2107.07653) कियान लियू, बेई चेन, जियाकी गुओ, मोर्टेज़ा ज़ियादी, ज़ेकी लिन, वीज़ू चेन, जियान-गुआंग लू द्वारा पोस्ट किया गया।
+1. **[T5](https://huggingface.co/docs/transformers/model_doc/t5)** (来自 Google AI)कॉलिन रैफेल और नोम शज़ीर और एडम रॉबर्ट्स और कैथरीन ली और शरण नारंग और माइकल मटेना द्वारा साथ में पेपर [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/abs/1910.10683) और यांकी झोउ और वेई ली और पीटर जे लियू।
+1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (Google AI से) साथ वाला पेपर [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) कॉलिन रैफेल और नोम शज़ीर और एडम रॉबर्ट्स और कैथरीन ली और शरण नारंग द्वारा और माइकल मटेना और यांकी झोउ और वेई ली और पीटर जे लियू।
+1. **[Table Transformer](https://huggingface.co/docs/transformers/model_doc/table-transformer)** (माइक्रोसॉफ्ट रिसर्च से) साथ में पेपर [PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents](https://arxiv.org/abs/2110.00061) ब्रैंडन स्मॉक, रोहित पेसाला, रॉबिन अब्राहम द्वारा पोस्ट किया गया।
+1. **[TAPAS](https://huggingface.co/docs/transformers/model_doc/tapas)** (Google AI से) साथ में कागज [TAPAS: Weakly Supervised Table Parsing via Pre-training](https://arxiv.org/abs/2004.02349) जोनाथन हर्ज़िग, पावेल क्रिज़िस्तोफ़ नोवाक, थॉमस मुलर, फ्रांसेस्को पिकिन्नो और जूलियन मार्टिन ईसेन्च्लोस द्वारा।
+1. **[TAPEX](https://huggingface.co/docs/transformers/model_doc/tapex)** (माइक्रोसॉफ्ट रिसर्च से) साथ में पेपर [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) कियान लियू, बेई चेन, जियाकी गुओ, मोर्टेज़ा ज़ियादी, ज़ेकी लिन, वीज़ू चेन, जियान-गुआंग लू द्वारा पोस्ट किया गया।
 1. **[Time Series Transformer](https://huggingface.co/docs/transformers/model_doc/time_series_transformer)** (from HuggingFace).
 1. **[TimeSformer](https://huggingface.co/docs/transformers/model_doc/timesformer)** (from Facebook) released with the paper [Is Space-Time Attention All You Need for Video Understanding?](https://arxiv.org/abs/2102.05095) by Gedas Bertasius, Heng Wang, Lorenzo Torresani.
 1. **[Trajectory Transformer](https://huggingface.co/docs/transformers/model_doc/trajectory_transformers)** (from the University of California at Berkeley) released with the paper [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://arxiv.org/abs/2106.02039) by Michael Janner, Qiyang Li, Sergey Levine
-1. **[Transformer-XL](https://huggingface.co/docs/transformers/model_doc/transfo-xl)** (Google/CMU की ओर से) कागज के साथ [संस्करण-एक्स: एक ब्लॉग मॉडल चौकस चौक मॉडल मॉडल](https://arxivorg/abs/1901.02860) क्वोकोक वी. ले, रुस्लैन सलाखुतदी
+1. **[Transformer-XL](https://huggingface.co/docs/transformers/model_doc/transfo-xl)** (Google/CMU की ओर से) कागज के साथ [Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context](https://arxiv.org/abs/1901.02860) क्वोकोक वी. ले, रुस्लैन सलाखुतदी
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft) released with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (from UNC Chapel Hill) released with the paper [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) by Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (from Intel) released with the paper [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) by Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (Microsoft Research से) Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal. द्वाराअनुसंधान पत्र [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) के साथ जारी किया गया
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (Microsoft Research से) Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal. द्वाराअनुसंधान पत्र [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) के साथ जारी किया गया
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (Google Research से) Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant. द्वाराअनुसंधान पत्र [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) के साथ जारी किया गया
-1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (माइक्रोसॉफ्ट रिसर्च से) साथ में दिया गया पेपर [UniSpeech: यूनिफाइड स्पीच रिप्रेजेंटेशन लर्निंग विद लेबलेड एंड अनलेबल्ड डेटा](https://arxiv.org/abs/2101.07597) चेंगई वांग, यू वू, याओ कियान, केनिची कुमातानी, शुजी लियू, फुरु वेई, माइकल ज़ेंग, ज़ुएदोंग हुआंग द्वारा।
-1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (माइक्रोसॉफ्ट रिसर्च से) कागज के साथ [UNISPEECH-SAT: यूनिवर्सल स्पीच रिप्रेजेंटेशन लर्निंग विद स्पीकर अवेयर प्री-ट्रेनिंग](https://arxiv.org/abs/2110.05752) सानयुआन चेन, यू वू, चेंग्यी वांग, झेंगयांग चेन, झूओ चेन, शुजी लियू, जियान वू, याओ कियान, फुरु वेई, जिन्यु ली, जियांगज़ान यू द्वारा पोस्ट किया गया।
+1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (माइक्रोसॉफ्ट रिसर्च से) साथ में दिया गया पेपर [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) चेंगई वांग, यू वू, याओ कियान, केनिची कुमातानी, शुजी लियू, फुरु वेई, माइकल ज़ेंग, ज़ुएदोंग हुआंग द्वारा।
+1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (माइक्रोसॉफ्ट रिसर्च से) कागज के साथ [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) सानयुआन चेन, यू वू, चेंग्यी वांग, झेंगयांग चेन, झूओ चेन, शुजी लियू, जियान वू, याओ कियान, फुरु वेई, जिन्यु ली, जियांगज़ान यू द्वारा पोस्ट किया गया।
 1. **[UnivNet](https://huggingface.co/docs/transformers/model_doc/univnet)** (from Kakao Corporation) released with the paper [UnivNet: A Neural Vocoder with Multi-Resolution Spectrogram Discriminators for High-Fidelity Waveform Generation](https://arxiv.org/abs/2106.07889) by Won Jang, Dan Lim, Jaesam Yoon, Bongwan Kim, and Juntae Kim.
 1. **[UPerNet](https://huggingface.co/docs/transformers/model_doc/upernet)** (from Peking University) released with the paper [Unified Perceptual Parsing for Scene Understanding](https://arxiv.org/abs/1807.10221) by Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, Jian Sun.
-1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (सिंघुआ यूनिवर्सिटी और ननकाई यूनिवर्सिटी से) साथ में पेपर [विजुअल अटेंशन नेटवर्क](https://arxiv.org/pdf/2202.09741.pdf) मेंग-हाओ गुओ, चेंग-ज़े लू, झेंग-निंग लियू, मिंग-मिंग चेंग, शि-मिन हू द्वारा।
-1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (मल्टीमीडिया कम्प्यूटिंग ग्रुप, नानजिंग यूनिवर्सिटी से) साथ में पेपर [वीडियोएमएई: मास्क्ड ऑटोएन्कोडर स्व-पर्यवेक्षित वीडियो प्री-ट्रेनिंग के लिए डेटा-कुशल सीखने वाले हैं](https://arxiv.org/abs/2203.12602) ज़ान टोंग, यिबिंग सॉन्ग, जुए द्वारा वांग, लिमिन वांग द्वारा पोस्ट किया गया।
-1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (NAVER AI Lab/Kakao Enterprise/Kakao Brain से) साथ में कागज [ViLT: Vision-and-Language Transformer बिना कनवल्शन या रीजन सुपरविजन](https://arxiv.org/abs/2102.03334) वोनजे किम, बोक्यूंग सोन, इल्डू किम द्वारा पोस्ट किया गया।
+1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (सिंघुआ यूनिवर्सिटी और ननकाई यूनिवर्सिटी से) साथ में पेपर [Visual Attention Network](https://arxiv.org/abs/2202.09741) मेंग-हाओ गुओ, चेंग-ज़े लू, झेंग-निंग लियू, मिंग-मिंग चेंग, शि-मिन हू द्वारा।
+1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (मल्टीमीडिया कम्प्यूटिंग ग्रुप, नानजिंग यूनिवर्सिटी से) साथ में पेपर [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) ज़ान टोंग, यिबिंग सॉन्ग, जुए द्वारा वांग, लिमिन वांग द्वारा पोस्ट किया गया।
+1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (NAVER AI Lab/Kakao Enterprise/Kakao Brain से) साथ में कागज [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) वोनजे किम, बोक्यूंग सोन, इल्डू किम द्वारा पोस्ट किया गया।
 1. **[VipLlava](https://huggingface.co/docs/transformers/model_doc/vipllava)** (University of Wisconsin–Madison से) Mu Cai, Haotian Liu, Siva Karthik Mustikovela, Gregory P. Meyer, Yuning Chai, Dennis Park, Yong Jae Lee. द्वाराअनुसंधान पत्र [Making Large Multimodal Models Understand Arbitrary Visual Prompts](https://arxiv.org/abs/2312.00784) के साथ जारी किया गया
-1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/model_doc/vit)** (गूगल एआई से) कागज के साथ [एक इमेज इज़ वर्थ 16x16 वर्ड्स: ट्रांसफॉर्मर्स फॉर इमेज रिकॉग्निशन एट स्केल](https://arxiv.org/abs/2010.11929) एलेक्सी डोसोवित्स्की, लुकास बेयर, अलेक्जेंडर कोलेसनिकोव, डिर्क वीसेनबोर्न, शियाओहुआ झाई, थॉमस अनटरथिनर, मुस्तफा देहघानी, मैथियास मिंडरर, जॉर्ज हेगोल्ड, सिल्वेन गेली, जैकब उस्ज़कोरेइट द्वारा हॉल्सबी द्वारा पोस्ट किया गया।
+1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/model_doc/vit)** (गूगल एआई से) कागज के साथ [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://arxiv.org/abs/2010.11929) एलेक्सी डोसोवित्स्की, लुकास बेयर, अलेक्जेंडर कोलेसनिकोव, डिर्क वीसेनबोर्न, शियाओहुआ झाई, थॉमस अनटरथिनर, मुस्तफा देहघानी, मैथियास मिंडरर, जॉर्ज हेगोल्ड, सिल्वेन गेली, जैकब उस्ज़कोरेइट द्वारा हॉल्सबी द्वारा पोस्ट किया गया।
 1. **[VisualBERT](https://huggingface.co/docs/transformers/model_doc/visual_bert)** (UCLA NLP से) साथ वाला पेपर [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) लियुनियन हेरोल्ड ली, मार्क यात्स्कर, दा यिन, चो-जुई हसीह, काई-वेई चांग द्वारा।
 1. **[ViT Hybrid](https://huggingface.co/docs/transformers/model_doc/vit_hybrid)** (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. **[VitDet](https://huggingface.co/docs/transformers/model_doc/vitdet)** (Meta AI से) Yanghao Li, Hanzi Mao, Ross Girshick, Kaiming He. द्वाराअनुसंधान पत्र [Exploring Plain Vision Transformer Backbones for Object Detection](https://arxiv.org/abs/2203.16527) के साथ जारी किया गया
-1. **[ViTMAE](https://huggingface.co/docs/transformers/model_doc/vit_mae)** (मेटा एआई से) साथ में कागज [मास्कड ऑटोएन्कोडर स्केलेबल विजन लर्नर्स हैं](https://arxiv.org/एब्स/2111.06377) कैमिंग हे, ज़िनेली चेन, सेनिंग ज़ी, यांगहो ली, पिओट्र डॉलर, रॉस गिर्शिक द्वारा।
+1. **[ViTMAE](https://huggingface.co/docs/transformers/model_doc/vit_mae)** (मेटा एआई से) साथ में कागज [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) कैमिंग हे, ज़िनेली चेन, सेनिंग ज़ी, यांगहो ली, पिओट्र डॉलर, रॉस गिर्शिक द्वारा।
 1. **[ViTMatte](https://huggingface.co/docs/transformers/model_doc/vitmatte)** (HUST-VL से) Jingfeng Yao, Xinggang Wang, Shusheng Yang, Baoyuan Wang. द्वाराअनुसंधान पत्र [ViTMatte: Boosting Image Matting with Pretrained Plain Vision Transformers](https://arxiv.org/abs/2305.15272) के साथ जारी किया गया
-1. **[ViTMSN](https://huggingface.co/docs/transformers/model_doc/vit_msn)** (मेटा एआई से) साथ में कागज [लेबल-कुशल सीखने के लिए मास्क्ड स्याम देश के नेटवर्क](https://arxiv.org/abs/2204.07141) महमूद असरान, मथिल्डे कैरन, ईशान मिश्रा, पियोट्र बोजानोवस्की, फ्लोरियन बोर्डेस, पास्कल विंसेंट, आर्मंड जौलिन, माइकल रब्बत, निकोलस बल्लास द्वारा।
+1. **[ViTMSN](https://huggingface.co/docs/transformers/model_doc/vit_msn)** (मेटा एआई से) साथ में कागज [Masked Siamese Networks for Label-Efficient Learning](https://arxiv.org/abs/2204.07141) महमूद असरान, मथिल्डे कैरन, ईशान मिश्रा, पियोट्र बोजानोवस्की, फ्लोरियन बोर्डेस, पास्कल विंसेंट, आर्मंड जौलिन, माइकल रब्बत, निकोलस बल्लास द्वारा।
 1. **[VITS](https://huggingface.co/docs/transformers/model_doc/vits)** (Kakao Enterprise से) Jaehyeon Kim, Jungil Kong, Juhee Son. द्वाराअनुसंधान पत्र [Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech](https://arxiv.org/abs/2106.06103) के साथ जारी किया गया
 1. **[ViViT](https://huggingface.co/docs/transformers/model_doc/vivit)** (from Google Research) released with the paper [ViViT: A Video Vision Transformer](https://arxiv.org/abs/2103.15691) by Anurag Arnab, Mostafa Dehghani, Georg Heigold, Chen Sun, Mario Lučić, Cordelia Schmid.
-1. **[Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/wav2vec2)** (फेसबुक एआई से) साथ में पेपर [wav2vec 2.0: ए फ्रेमवर्क फॉर सेल्फ-सुपरवाइज्ड लर्निंग ऑफ स्पीच रिप्रेजेंटेशन](https://arxiv.org/abs/2006.11477) एलेक्सी बेवस्की, हेनरी झोउ, अब्देलरहमान मोहम्मद, माइकल औली द्वारा।
+1. **[Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/wav2vec2)** (फेसबुक एआई से) साथ में पेपर [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://arxiv.org/abs/2006.11477) एलेक्सी बेवस्की, हेनरी झोउ, अब्देलरहमान मोहम्मद, माइकल औली द्वारा।
 1. **[Wav2Vec2-BERT](https://huggingface.co/docs/transformers/model_doc/wav2vec2-bert)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
-1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (Facebook AI से) साथ वाला पेपर [FAIRSEQ S2T: FAIRSEQ के साथ फास्ट स्पीच-टू-टेक्स्ट मॉडलिंग](https://arxiv.org/abs/2010.05171) चांगहान वांग, यूं तांग, जुताई मा, ऐनी वू, सरव्या पोपुरी, दिमित्रो ओखोनको, जुआन पिनो द्वारा पोस्ट किया गया।
-1. **[Wav2Vec2Phoneme](https://huggingface.co/docs/transformers/model_doc/wav2vec2_phoneme)** (Facebook AI से) साथ वाला पेपर [सरल और प्रभावी जीरो-शॉट क्रॉस-लिंगुअल फोनेम रिकॉग्निशन](https://arxiv.org/abs/2109.11680) कियानटोंग जू, एलेक्सी बाएव्स्की, माइकल औली द्वारा।
-1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (माइक्रोसॉफ्ट रिसर्च से) पेपर के साथ जारी किया गया [WavLM: फुल स्टैक के लिए बड़े पैमाने पर स्व-पर्यवेक्षित पूर्व-प्रशिक्षण स्पीच प्रोसेसिंग](https://arxiv.org/abs/2110.13900) सानयुआन चेन, चेंगयी वांग, झेंगयांग चेन, यू वू, शुजी लियू, ज़ुओ चेन, जिन्यु ली, नाओयुकी कांडा, ताकुया योशियोका, ज़िओंग जिओ, जियान वू, लॉन्ग झोउ, शुओ रेन, यानमिन कियान, याओ कियान, जियान वू, माइकल ज़ेंग, फुरु वेई।
-1. **[Whisper](https://huggingface.co/docs/transformers/model_doc/whisper)** (OpenAI से) साथ में कागज [बड़े पैमाने पर कमजोर पर्यवेक्षण के माध्यम से मजबूत भाषण पहचान](https://cdn.openai.com/papers/whisper.pdf) एलेक रैडफोर्ड, जोंग वूक किम, ताओ जू, ग्रेग ब्रॉकमैन, क्रिस्टीन मैकलीवे, इल्या सुत्स्केवर द्वारा।
-1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (माइक्रोसॉफ्ट रिसर्च से) कागज के साथ [एक्सपैंडिंग लैंग्वेज-इमेज प्रीट्रेन्ड मॉडल फॉर जनरल वीडियो रिकग्निशन](https://arxiv.org/abs/2208.02816) बोलिन नी, होउवेन पेंग, मिंगाओ चेन, सोंगयांग झांग, गाओफेंग मेंग, जियानलोंग फू, शिमिंग जियांग, हैबिन लिंग द्वारा।
+1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (Facebook AI से) साथ वाला पेपर [FAIRSEQ S2T: Fast Speech-to-Text Modeling with FAIRSEQ](https://arxiv.org/abs/2010.05171) चांगहान वांग, यूं तांग, जुताई मा, ऐनी वू, सरव्या पोपुरी, दिमित्रो ओखोनको, जुआन पिनो द्वारा पोस्ट किया गया।
+1. **[Wav2Vec2Phoneme](https://huggingface.co/docs/transformers/model_doc/wav2vec2_phoneme)** (Facebook AI से) साथ वाला पेपर [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) कियानटोंग जू, एलेक्सी बाएव्स्की, माइकल औली द्वारा।
+1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (माइक्रोसॉफ्ट रिसर्च से) पेपर के साथ जारी किया गया [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) सानयुआन चेन, चेंगयी वांग, झेंगयांग चेन, यू वू, शुजी लियू, ज़ुओ चेन, जिन्यु ली, नाओयुकी कांडा, ताकुया योशियोका, ज़िओंग जिओ, जियान वू, लॉन्ग झोउ, शुओ रेन, यानमिन कियान, याओ कियान, जियान वू, माइकल ज़ेंग, फुरु वेई।
+1. **[Whisper](https://huggingface.co/docs/transformers/model_doc/whisper)** (OpenAI से) साथ में कागज [Robust Speech Recognition via Large-Scale Weak Supervision](https://cdn.openai.com/papers/whisper.pdf) एलेक रैडफोर्ड, जोंग वूक किम, ताओ जू, ग्रेग ब्रॉकमैन, क्रिस्टीन मैकलीवे, इल्या सुत्स्केवर द्वारा।
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (माइक्रोसॉफ्ट रिसर्च से) कागज के साथ [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) बोलिन नी, होउवेन पेंग, मिंगाओ चेन, सोंगयांग झांग, गाओफेंग मेंग, जियानलोंग फू, शिमिंग जियांग, हैबिन लिंग द्वारा।
 1. **[X-MOD](https://huggingface.co/docs/transformers/model_doc/xmod)** (Meta AI से) Jonas Pfeiffer, Naman Goyal, Xi Lin, Xian Li, James Cross, Sebastian Riedel, Mikel Artetxe. द्वाराअनुसंधान पत्र [Lifting the Curse of Multilinguality by Pre-training Modular Transformers](http://dx.doi.org/10.18653/v1/2022.naacl-main.255) के साथ जारी किया गया
 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)** (फेसबुक से) साथ में पेपर [क्रॉस-लिंगुअल लैंग्वेज मॉडल प्रीट्रेनिंग](https://arxiv.org/abs/1901.07291) गिलाउम लैम्पल और एलेक्सिस कोनो द्वारा।
-1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (माइक्रोसॉफ्ट रिसर्च से) साथ में कागज [ProphetNet: प्रेडिक्टिंग फ्यूचर एन-ग्राम फॉर सीक्वेंस-टू- सीक्वेंस प्री-ट्रेनिंग](https://arxiv.org/abs/2001.04063) यू यान, वीज़ेन क्यूई, येयुन गोंग, दयाहेंग लियू, नान डुआन, जिउशेंग चेन, रुओफ़ेई झांग और मिंग झोउ द्वारा।
-1. **[XLM-RoBERTa](https://huggingface.co/docs/transformers/model_doc/xlm-roberta)** (फेसबुक एआई से), साथ में पेपर [अनसुपरवाइज्ड क्रॉस-लिंगुअल रिप्रेजेंटेशन लर्निंग एट स्केल](https://arxiv.org/abs/1911.02116) एलेक्सिस कोन्यू*, कार्तिकेय खंडेलवाल*, नमन गोयल, विश्रव चौधरी, गिलाउम वेनज़ेक, फ्रांसिस्को गुज़मैन द्वारा , एडौर्ड ग्रेव, मायल ओट, ल्यूक ज़ेटलमॉयर और वेसेलिन स्टोयानोव द्वारा।
-1. **[XLM-RoBERTa-XL](https://huggingface.co/docs/transformers/model_doc/xlm-roberta-xl)** (Facebook AI से) साथ में कागज [बहुभाषी नकाबपोश भाषा के लिए बड़े पैमाने पर ट्रांसफॉर्मर मॉडलिंग](https://arxiv.org/abs/2105.00572) नमन गोयल, जिंगफेई डू, मायल ओट, गिरि अनंतरामन, एलेक्सिस कोनो द्वारा पोस्ट किया गया।
+1. **[XLM](https://huggingface.co/docs/transformers/model_doc/xlm)** (फेसबुक से) साथ में पेपर [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) गिलाउम लैम्पल और एलेक्सिस कोनो द्वारा।
+1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (माइक्रोसॉफ्ट रिसर्च से) साथ में कागज [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) यू यान, वीज़ेन क्यूई, येयुन गोंग, दयाहेंग लियू, नान डुआन, जिउशेंग चेन, रुओफ़ेई झांग और मिंग झोउ द्वारा।
+1. **[XLM-RoBERTa](https://huggingface.co/docs/transformers/model_doc/xlm-roberta)** (फेसबुक एआई से), साथ में पेपर [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) एलेक्सिस कोन्यू*, कार्तिकेय खंडेलवाल*, नमन गोयल, विश्रव चौधरी, गिलाउम वेनज़ेक, फ्रांसिस्को गुज़मैन द्वारा , एडौर्ड ग्रेव, मायल ओट, ल्यूक ज़ेटलमॉयर और वेसेलिन स्टोयानोव द्वारा।
+1. **[XLM-RoBERTa-XL](https://huggingface.co/docs/transformers/model_doc/xlm-roberta-xl)** (Facebook AI से) साथ में कागज [Larger-Scale Transformers for Multilingual Masked Language Modeling](https://arxiv.org/abs/2105.00572) नमन गोयल, जिंगफेई डू, मायल ओट, गिरि अनंतरामन, एलेक्सिस कोनो द्वारा पोस्ट किया गया।
 1. **[XLM-V](https://huggingface.co/docs/transformers/model_doc/xlm-v)** (from Meta AI) released with the paper [XLM-V: Overcoming the Vocabulary Bottleneck in Multilingual Masked Language Models](https://arxiv.org/abs/2301.10472) by Davis Liang, Hila Gonen, Yuning Mao, Rui Hou, Naman Goyal, Marjan Ghazvininejad, Luke Zettlemoyer, Madian Khabsa.
-1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (Google/CMU से) साथ वाला पेपर [XLNet: जनरलाइज्ड ऑटोरेग्रेसिव प्रीट्रेनिंग फॉर लैंग्वेज अंडरस्टैंडिंग](https://arxiv.org/abs/1906.08237) ज़ीलिन यांग*, ज़िहांग दाई*, यिमिंग यांग, जैम कार्बोनेल, रुस्लान सलाखुतदीनोव, क्वोक वी. ले द्वारा।
-1. **[XLS-R](https://huggingface.co/docs/transformers/model_doc/xls_r)** (Facebook AI से) साथ वाला पेपर [XLS-R: सेल्फ सुपरवाइज्ड क्रॉस-लिंगुअल स्पीच रिप्रेजेंटेशन लर्निंग एट स्केल](https://arxiv.org/abs/2111.09296) अरुण बाबू, चांगहान वांग, एंड्रोस तजंद्रा, कुशाल लखोटिया, कियानटोंग जू, नमन गोयल, कृतिका सिंह, पैट्रिक वॉन प्लैटन, याथार्थ सराफ, जुआन पिनो, एलेक्सी बेवस्की, एलेक्सिस कोन्यू, माइकल औली द्वारा पोस्ट किया गया।
-1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (फेसबुक एआई से) साथ में पेपर [अनसुपरवाइज्ड क्रॉस-लिंगुअल रिप्रेजेंटेशन लर्निंग फॉर स्पीच रिकग्निशन](https://arxiv.org/abs/2006.13979) एलेक्सिस कोन्यू, एलेक्सी बेवस्की, रोनन कोलोबर्ट, अब्देलरहमान मोहम्मद, माइकल औली द्वारा।
-1. **[YOLOS](https://huggingface.co/docs/transformers/model_doc/yolos)** (हुआझोंग यूनिवर्सिटी ऑफ साइंस एंड टेक्नोलॉजी से) साथ में पेपर [यू ओनली लुक एट वन सीक्वेंस: रीथिंकिंग ट्रांसफॉर्मर इन विज़न थ्रू ऑब्जेक्ट डिटेक्शन](https://arxiv.org/abs/2106.00666) युक्सिन फेंग, बेनचेंग लियाओ, जिंगगैंग वांग, जेमिन फेंग, जियांग क्यूई, रुई वू, जियानवेई नीयू, वेन्यू लियू द्वारा पोस्ट किया गया।
+1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (Google/CMU से) साथ वाला पेपर [XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) ज़ीलिन यांग*, ज़िहांग दाई*, यिमिंग यांग, जैम कार्बोनेल, रुस्लान सलाखुतदीनोव, क्वोक वी. ले द्वारा।
+1. **[XLS-R](https://huggingface.co/docs/transformers/model_doc/xls_r)** (Facebook AI से) साथ वाला पेपर [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) अरुण बाबू, चांगहान वांग, एंड्रोस तजंद्रा, कुशाल लखोटिया, कियानटोंग जू, नमन गोयल, कृतिका सिंह, पैट्रिक वॉन प्लैटन, याथार्थ सराफ, जुआन पिनो, एलेक्सी बेवस्की, एलेक्सिस कोन्यू, माइकल औली द्वारा पोस्ट किया गया।
+1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (फेसबुक एआई से) साथ में पेपर [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) एलेक्सिस कोन्यू, एलेक्सी बेवस्की, रोनन कोलोबर्ट, अब्देलरहमान मोहम्मद, माइकल औली द्वारा।
+1. **[YOLOS](https://huggingface.co/docs/transformers/model_doc/yolos)** (हुआझोंग यूनिवर्सिटी ऑफ साइंस एंड टेक्नोलॉजी से) साथ में पेपर [You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection](https://arxiv.org/abs/2106.00666) युक्सिन फेंग, बेनचेंग लियाओ, जिंगगैंग वांग, जेमिन फेंग, जियांग क्यूई, रुई वू, जियानवेई नीयू, वेन्यू लियू द्वारा पोस्ट किया गया।
 1. **[YOSO](https://huggingface.co/docs/transformers/model_doc/yoso)** (विस्कॉन्सिन विश्वविद्यालय - मैडिसन से) साथ में पेपर [यू ओनली सैंपल (लगभग) ज़ानपेंग ज़ेंग, युनयांग ज़िओंग द्वारा , सत्य एन. रवि, शैलेश आचार्य, ग्लेन फंग, विकास सिंह द्वारा पोस्ट किया गया।
 1. एक नए मॉडल में योगदान देना चाहते हैं? नए मॉडल जोड़ने में आपका मार्गदर्शन करने के लिए हमारे पास एक **विस्तृत मार्गदर्शिका और टेम्प्लेट** है। आप उन्हें [`टेम्पलेट्स`](./templates) निर्देशिका में पा सकते हैं। पीआर शुरू करने से पहले [योगदान दिशानिर्देश](./CONTRIBUTING.md) देखना और अनुरक्षकों से संपर्क करना या प्रतिक्रिया प्राप्त करने के लिए एक नया मुद्दा खोलना याद रखें।
 

README_ja.md

@@ -319,7 +319,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[BigBird-Pegasus](https://huggingface.co/docs/transformers/model_doc/bigbird_pegasus)** (Google Research から) Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed から公開された研究論文: [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062)
 1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (Google Research から) Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed から公開された研究論文: [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062)
 1. **[BioGpt](https://huggingface.co/docs/transformers/model_doc/biogpt)** (Microsoft Research AI4Science から) Renqian Luo, Liai Sun, Yingce Xia, Tao Qin, Sheng Zhang, Hoifung Poon and Tie-Yan Liu から公開された研究論文: [BioGPT: generative pre-trained transformer for biomedical text generation and mining](https://academic.oup.com/bib/advance-article/doi/10.1093/bib/bbac409/6713511?guestAccessKey=a66d9b5d-4f83-4017-bb52-405815c907b9)
-1. **[BiT](https://huggingface.co/docs/transformers/model_doc/bit)** (Google AI から) Alexander Kolesnikov, Lucas Beyer, Xiaohua Zhai, Joan Puigcerver, Jessica Yung, Sylvain Gelly, Neil から公開された研究論文: [Big Transfer (BiT)](https://arxiv.org/abs/1912.11370)Houlsby.
+1. **[BiT](https://huggingface.co/docs/transformers/model_doc/bit)** (Google AI から) Alexander Kolesnikov, Lucas Beyer, Xiaohua Zhai, Joan Puigcerver, Jessica Yung, Sylvain Gelly, Neil から公開された研究論文: [Big Transfer (BiT): General Visual Representation Learning](https://arxiv.org/abs/1912.11370)Houlsby.
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (Facebook から) Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston から公開された研究論文: [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637)
 1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (Facebook から) Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston から公開された研究論文: [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637)
 1. **[BLIP](https://huggingface.co/docs/transformers/model_doc/blip)** (Salesforce から) Junnan Li, Dongxu Li, Caiming Xiong, Steven Hoi から公開された研究論文: [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://arxiv.org/abs/2201.12086)
@@ -338,7 +338,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (Salesforce から) Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong から公開された研究論文: [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474)
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (MetaAI から) Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve. から公開された研究論文 [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/)
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (Cohere から) Cohere.  から公開された研究論文 [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>)
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (Cohere から) Cohere.  から公開された研究論文 [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>)
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (Microsoft Research Asia から) Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang から公開された研究論文: [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152)
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (YituTech から) Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan から公開された研究論文: [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496)
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (Facebook AI から) Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie から公開された研究論文: [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545)
@@ -360,7 +360,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (Microsoft Research から) Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan から公開された研究論文: [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://arxiv.org/abs/1911.00536)
 1. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (SHI Labs から) Ali Hassani and Humphrey Shi から公開された研究論文: [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001)
 1. **[DINOv2](https://huggingface.co/docs/transformers/model_doc/dinov2)** (Meta AI から) Maxime Oquab, Timothée Darcet, Théo Moutakanni, Huy Vo, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, Mahmoud Assran, Nicolas Ballas, Wojciech Galuba, Russell Howes, Po-Yao Huang, Shang-Wen Li, Ishan Misra, Michael Rabbat, Vasu Sharma, Gabriel Synnaeve, Hu Xu, Hervé Jegou, Julien Mairal, Patrick Labatut, Armand Joulin, Piotr Bojanowski. から公開された研究論文 [DINOv2: Learning Robust Visual Features without Supervision](https://arxiv.org/abs/2304.07193)
-1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (HuggingFace から), Victor Sanh, Lysandre Debut and Thomas Wolf. 同じ手法で GPT2, RoBERTa と Multilingual BERT の圧縮を行いました.圧縮されたモデルはそれぞれ [DistilGPT2](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation)、[DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation)、[DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation) と名付けられました. 公開された研究論文: [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108)
+1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (HuggingFace から), Victor Sanh, Lysandre Debut and Thomas Wolf. 同じ手法で GPT2, RoBERTa と Multilingual BERT の圧縮を行いました.圧縮されたモデルはそれぞれ [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108)、[DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation)、[DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation) と名付けられました. 公開された研究論文: [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108)
 1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (Microsoft Research から) Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei から公開された研究論文: [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378)
 1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (NAVER から), Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park から公開された研究論文: [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664)
 1. **[DPR](https://huggingface.co/docs/transformers/model_doc/dpr)** (Facebook から) Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih から公開された研究論文: [Dense Passage Retrieval for Open-Domain Question Answering](https://arxiv.org/abs/2004.04906)
@@ -374,7 +374,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[ErnieM](https://huggingface.co/docs/transformers/model_doc/ernie_m)** (Baidu から) Xuan Ouyang, Shuohuan Wang, Chao Pang, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang. から公開された研究論文 [ERNIE-M: Enhanced Multilingual Representation by Aligning Cross-lingual Semantics with Monolingual Corpora](https://arxiv.org/abs/2012.15674)
 1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (Meta AI から) はトランスフォーマープロテイン言語モデルです.  **ESM-1b** は Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus から公開された研究論文: [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118). **ESM-1v** は Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives　から公開された研究論文: [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648). **ESM-2** と　**ESMFold** は Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives から公開された研究論文: [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902)
 1. **[Falcon](https://huggingface.co/docs/transformers/model_doc/falcon)** (from Technology Innovation Institute) by Almazrouei, Ebtesam and Alobeidli, Hamza and Alshamsi, Abdulaziz and Cappelli, Alessandro and Cojocaru, Ruxandra and Debbah, Merouane and Goffinet, Etienne and Heslow, Daniel and Launay, Julien and Malartic, Quentin and Noune, Badreddine and Pannier, Baptiste and Penedo, Guilherme.
-1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (ESPnet and Microsoft Research から) Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang, and Yuekai Zhang. から公開された研究論文 [Fastspeech 2: Fast And High-quality End-to-End Text To Speech](https://arxiv.org/pdf/2006.04558.pdf)
+1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (ESPnet and Microsoft Research から) Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang, and Yuekai Zhang. から公開された研究論文 [Recent Developments On Espnet Toolkit Boosted By Conformer](https://arxiv.org/abs/2010.13956)
 1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (Google AI から) Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V から公開されたレポジトリー [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) Le, and Jason Wei
 1. **[FLAN-UL2](https://huggingface.co/docs/transformers/model_doc/flan-ul2)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-ul2-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 1. **[FlauBERT](https://huggingface.co/docs/transformers/model_doc/flaubert)** (CNRS から) Hang Le, Loïc Vial, Jibril Frej, Vincent Segonne, Maximin Coavoux, Benjamin Lecouteux, Alexandre Allauzen, Benoît Crabbé, Laurent Besacier, Didier Schwab から公開された研究論文: [FlauBERT: Unsupervised Language Model Pre-training for French](https://arxiv.org/abs/1912.05372)
@@ -383,7 +383,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (Microsoft Research から) Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao. から公開された研究論文 [Focal Modulation Networks](https://arxiv.org/abs/2203.11926)
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (CMU/Google Brain から) Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le から公開された研究論文: [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236)
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (ADEPT から) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. から公開された研究論文 [blog post](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (Google から) the Gemma Google team. から公開された研究論文 [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/)
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (Google から) the Gemma Google team. から公開された研究論文 [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/)
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (Microsoft Research から) Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang. から公開された研究論文 [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100)
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (KAIST から) Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim から公開された研究論文: [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436)
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (OpenAI から) Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever から公開された研究論文: [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/)
@@ -414,8 +414,9 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (Meta AI から) Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze から公開された研究論文: [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136)
 1. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (South China University of Technology から) Jiapeng Wang, Lianwen Jin, Kai Ding から公開された研究論文: [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669)
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (The FAIR team of Meta AI から) Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample. から公開された研究論文 [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971)
-1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (The FAIR team of Meta AI から) Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.. から公開された研究論文 [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/XXX)
+1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (The FAIR team of Meta AI から) Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.. から公開された研究論文 [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/)
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (Microsoft Research & University of Wisconsin-Madison から) Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee. から公開された研究論文 [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485)
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (Microsoft Research & University of Wisconsin-Madison から) Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee. から公開された研究論文 [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744)
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (AllenAI から) Iz Beltagy, Matthew E. Peters, Arman Cohan から公開された研究論文: [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150)
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (Google AI から) Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang から公開された研究論文: [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916)
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (Studio Ousia から) Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto から公開された研究論文: [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057)
@@ -423,7 +424,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (Facebook から) Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert から公開された研究論文: [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161)
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (Facebook から) Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin から公開された研究論文: [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125)
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (Albert Gu and Tri Dao から) Albert Gu and Tri Dao. から公開された研究論文 [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752)
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (Albert Gu and Tri Dao から) Albert Gu and Tri Dao. から公開された研究論文 [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752)
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Jörg Tiedemann から. [OPUS](http://opus.nlpl.eu/) を使いながら学習された "Machine translation" (マシントランスレーション) モデル. [Marian Framework](https://marian-nmt.github.io/) はMicrosoft Translator Team　が現在開発中です.
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (Microsoft Research Asia から) Junlong Li, Yiheng Xu, Lei Cui, Furu Wei から公開された研究論文: [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518)
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (FAIR and UIUC から) Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar. から公開された研究論文 [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527)
@@ -446,10 +447,10 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[MobileViTV2](https://huggingface.co/docs/transformers/model_doc/mobilevitv2)** (Apple から) Sachin Mehta and Mohammad Rastegari. から公開された研究論文 [Separable Self-attention for Mobile Vision Transformers](https://arxiv.org/abs/2206.02680)
 1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (Microsoft Research から) Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu から公開された研究論文: [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297)
 1. **[MPT](https://huggingface.co/docs/transformers/model_doc/mpt)** (MosaiML から) the MosaicML NLP Team. から公開された研究論文 [llm-foundry](https://github.com/mosaicml/llm-foundry/)
-1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (the University of Wisconsin - Madison から) Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh. から公開された研究論文 [Multi Resolution Analysis (MRA)](https://arxiv.org/abs/2207.10284)
+1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (the University of Wisconsin - Madison から) Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh. から公開された研究論文 [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284)
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (Google AI から) Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel から公開された研究論文: [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934)
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (RUC AI Box から) Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen から公開された研究論文: [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131)
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (SHI Labs から) Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi から公開された研究論文: [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143)
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (Huawei Noah’s Ark Lab から) Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu から公開された研究論文: [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204)
@@ -463,7 +464,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (Google AI から) 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 から公開された研究論文: [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230)
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (Google AI から) Matthias Minderer, Alexey Gritsenko, Neil Houlsby. から公開された研究論文 [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683)
 1. **[PatchTSMixer](https://huggingface.co/docs/transformers/model_doc/patchtsmixer)** ( IBM Research から) Vijay Ekambaram, Arindam Jati, Nam Nguyen, Phanwadee Sinthong, Jayant Kalagnanam. から公開された研究論文 [TSMixer: Lightweight MLP-Mixer Model for Multivariate Time Series Forecasting](https://arxiv.org/pdf/2306.09364.pdf)
-1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (IBM から) Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam. から公開された研究論文 [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf)
+1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (IBM から) Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam. から公開された研究論文 [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/abs/2211.14730)
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (Google から) Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu から公開された研究論文: [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777)
 1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (Google から) Jason Phang, Yao Zhao, and Peter J. Liu から公開された研究論文: [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347)
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (Deepmind から) 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 から公開された研究論文: [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795)
@@ -476,7 +477,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi, Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/prophetnet)** (Microsoft Research から) Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou から公開された研究論文: [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063)
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (Nanjing University, The University of Hong Kong etc. から) Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao. から公開された研究論文 [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf)
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc. から) Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao. から公開された研究論文 [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) 
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc. から) Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao. から公開された研究論文 [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) 
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (NVIDIA から) Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius から公開された研究論文: [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602)
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (the Qwen team, Alibaba Group から) Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu. から公開された研究論文 [Qwen Technical Report](https://arxiv.org/abs/2309.16609)
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (Facebook から) Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela から公開された研究論文: [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401)
@@ -493,7 +494,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (NVIDIA から) Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo から公開された研究論文: [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203)
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (Beijing Academy of Artificial Intelligence (BAAI から) Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang. から公開された研究論文 [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284)
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (Beijing Academy of Artificial Intelligence (BAAI から) Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang. から公開された研究論文 [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284)
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (Meta AI から) Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick. から公開された研究論文 [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf)
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (ASAPP から) Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi から公開された研究論文: [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870)
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (ASAPP から) Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi から公開された研究論文: [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870)
@@ -504,7 +505,8 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (Tel Aviv University から), Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy から公開された研究論文: [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438)
 1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (Berkeley から) Forrest N. Iandola, Albert E. Shaw, Ravi Krishna, and Kurt W. Keutzer から公開された研究論文: [SqueezeBERT: What can computer vision teach NLP about efficient neural networks?](https://arxiv.org/abs/2006.11316)
 1. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report)](https://stability.wandb.io/stability-llm/stable-lm/reports/StableLM-3B-4E1T--VmlldzoyMjU4?accessToken=u3zujipenkx5g7rtcj9qojjgxpconyjktjkli2po09nffrffdhhchq045vp0wyfo) by  Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu. 
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (MBZUAI から) Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan. から公開された研究論文 [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446)
 1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (Microsoft から) Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo から公開された研究論文: [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030)
 1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (Microsoft から) Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo から公開された研究論文: [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883)
@@ -522,7 +524,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (Microsoft から), Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei から公開された研究論文: [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282)
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (from UNC Chapel Hill から), Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal から公開された研究論文: [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156)
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (Intel から), Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding から公開された研究論文: [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995)
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (Microsoft Research から) Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal. から公開された研究論文 [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623)
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (Microsoft Research から) Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal. から公開された研究論文 [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623)
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (Google Research から) Yi Tay, Mostafa Dehghani, Vinh Q から公開された研究論文: [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (Google Research から) Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant. から公開された研究論文 [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi)
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (Microsoft Research から) Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang から公開された研究論文: [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597)

README_ko.md

@@ -224,7 +224,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer)** (from MIT) released with the paper [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) by Yuan Gong, Yu-An Chung, James Glass.
 1. **[Autoformer](https://huggingface.co/docs/transformers/model_doc/autoformer)** (from Tsinghua University) released with the paper [Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting](https://arxiv.org/abs/2106.13008) by Haixu Wu, Jiehui Xu, Jianmin Wang, Mingsheng Long.
 1. **[Bark](https://huggingface.co/docs/transformers/model_doc/bark)** (from Suno) released in the repository [suno-ai/bark](https://github.com/suno-ai/bark) by Suno AI team.
-1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (from Facebook) released with the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/pdf/1910.13461.pdf) by Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer.
+1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (from Facebook) released with the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) by Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer.
 1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (from École polytechnique) released with the paper [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) by Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis.
 1. **[BARTpho](https://huggingface.co/docs/transformers/model_doc/bartpho)** (from VinAI Research) released with the paper [BARTpho: Pre-trained Sequence-to-Sequence Models for Vietnamese](https://arxiv.org/abs/2109.09701) by Nguyen Luong Tran, Duong Minh Le and Dat Quoc Nguyen.
 1. **[BEiT](https://huggingface.co/docs/transformers/model_doc/beit)** (from Microsoft) released with the paper [BEiT: BERT Pre-Training of Image Transformers](https://arxiv.org/abs/2106.08254) by Hangbo Bao, Li Dong, Furu Wei.
@@ -253,7 +253,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (Salesforce 에서) Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong 의 [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) 논문과 함께 발표했습니다.
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (MetaAI 에서 제공)은 Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.의 [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/)논문과 함께 발표했습니다.
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (Cohere 에서 제공)은 Cohere. 의 [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>)논문과 함께 발표했습니다.
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (Cohere 에서 제공)은 Cohere. 의 [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>)논문과 함께 발표했습니다.
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (Microsoft Research Asia 에서) Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang 의 [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) 논문과 함께 발표했습니다.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (YituTech 에서) Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan 의 [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) 논문과 함께 발표했습니다.
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (Facebook AI 에서) Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie 의 [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) 논문과 함께 발표했습니다.
@@ -275,7 +275,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (Microsoft Research 에서) Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan 의 [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://arxiv.org/abs/1911.00536) 논문과 함께 발표했습니다.
 1. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (SHI Labs 에서) Ali Hassani and Humphrey Shi 의 [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001) 논문과 함께 발표했습니다.
 1. **[DINOv2](https://huggingface.co/docs/transformers/model_doc/dinov2)** (Meta AI 에서 제공)은 Maxime Oquab, Timothée Darcet, Théo Moutakanni, Huy Vo, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, Mahmoud Assran, Nicolas Ballas, Wojciech Galuba, Russell Howes, Po-Yao Huang, Shang-Wen Li, Ishan Misra, Michael Rabbat, Vasu Sharma, Gabriel Synnaeve, Hu Xu, Hervé Jegou, Julien Mairal, Patrick Labatut, Armand Joulin, Piotr Bojanowski.의 [DINOv2: Learning Robust Visual Features without Supervision](https://arxiv.org/abs/2304.07193)논문과 함께 발표했습니다.
-1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (HuggingFace 에서) 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 의 [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108) 논문과 함께 발표했습니다.
+1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (HuggingFace 에서) Victor Sanh, Lysandre Debut and Thomas Wolf. The same method has been applied to compress GPT2 into [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108), 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 의 [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108) 논문과 함께 발표했습니다.
 1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (Microsoft Research 에서) Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei 의 [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) 논문과 함께 발표했습니다.
 1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (NAVER 에서) Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park 의 [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) 논문과 함께 발표했습니다.
 1. **[DPR](https://huggingface.co/docs/transformers/model_doc/dpr)** (Facebook 에서) Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih 의 [Dense Passage Retrieval for Open-Domain Question Answering](https://arxiv.org/abs/2004.04906) 논문과 함께 발표했습니다.
@@ -289,7 +289,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[ErnieM](https://huggingface.co/docs/transformers/model_doc/ernie_m)** (Baidu 에서 제공)은 Xuan Ouyang, Shuohuan Wang, Chao Pang, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang.의 [ERNIE-M: Enhanced Multilingual Representation by Aligning Cross-lingual Semantics with Monolingual Corpora](https://arxiv.org/abs/2012.15674)논문과 함께 발표했습니다.
 1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (from Meta AI) are transformer protein language models.  **ESM-1b** was released with the paper [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus. **ESM-1v** was released with the paper [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648) by Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives. **ESM-2** was released with the paper [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
 1. **[Falcon](https://huggingface.co/docs/transformers/model_doc/falcon)** (from Technology Innovation Institute) by Almazrouei, Ebtesam and Alobeidli, Hamza and Alshamsi, Abdulaziz and Cappelli, Alessandro and Cojocaru, Ruxandra and Debbah, Merouane and Goffinet, Etienne and Heslow, Daniel and Launay, Julien and Malartic, Quentin and Noune, Badreddine and Pannier, Baptiste and Penedo, Guilherme.
-1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (ESPnet and Microsoft Research 에서 제공)은 Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang, and Yuekai Zhang.의 [Fastspeech 2: Fast And High-quality End-to-End Text To Speech](https://arxiv.org/pdf/2006.04558.pdf)논문과 함께 발표했습니다.
+1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (ESPnet and Microsoft Research 에서 제공)은 Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang, and Yuekai Zhang.의 [Recent Developments On Espnet Toolkit Boosted By Conformer](https://arxiv.org/abs/2010.13956)논문과 함께 발표했습니다.
 1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 1. **[FLAN-UL2](https://huggingface.co/docs/transformers/model_doc/flan-ul2)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-ul2-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 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.
@@ -298,7 +298,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (from Microsoft Research) released with the paper [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) by Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (from ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. 논문과 함께 공개 [blog post](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (Google 에서 제공)은 the Gemma Google team.의 [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/)논문과 함께 발표했습니다.
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (Google 에서 제공)은 the Gemma Google team.의 [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/)논문과 함께 발표했습니다.
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
@@ -329,8 +329,9 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (Meta AI 에서) Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze 의 [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) 논문과 함께 발표했습니다.
 1. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (South China University of Technology 에서) Jiapeng Wang, Lianwen Jin, Kai Ding 의 [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) 논문과 함께 발표했습니다.
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (The FAIR team of Meta AI 에서 제공)은 Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.의 [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971)논문과 함께 발표했습니다.
-1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (The FAIR team of Meta AI 에서 제공)은 Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom..의 [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/XXX)논문과 함께 발표했습니다.
+1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (The FAIR team of Meta AI 에서 제공)은 Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom..의 [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/)논문과 함께 발표했습니다.
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (Microsoft Research & University of Wisconsin-Madison 에서 제공)은 Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.의 [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485)논문과 함께 발표했습니다.
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (Microsoft Research & University of Wisconsin-Madison 에서 제공)은 Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.의 [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744)논문과 함께 발표했습니다.
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (AllenAI 에서) Iz Beltagy, Matthew E. Peters, Arman Cohan 의 [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) 논문과 함께 발표했습니다.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (Google AI 에서) Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang 의 [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) 논문과 함께 발표했습니다.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (Studio Ousia 에서) Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto 의 [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) 논문과 함께 발표했습니다.
@@ -338,7 +339,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (Facebook 에서) Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert 의 [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) 논문과 함께 발표했습니다.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (Facebook 에서) Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin 의 [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) 논문과 함께 발표했습니다.
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (Albert Gu and Tri Dao 에서 제공)은 Albert Gu and Tri Dao.의 [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752)논문과 함께 발표했습니다.
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (Albert Gu and Tri Dao 에서 제공)은 Albert Gu and Tri Dao.의 [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752)논문과 함께 발표했습니다.
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Machine translation models trained using [OPUS](http://opus.nlpl.eu/) data by Jörg Tiedemann. The [Marian Framework](https://marian-nmt.github.io/) is being developed by the Microsoft Translator Team.
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (Microsoft Research Asia 에서) Junlong Li, Yiheng Xu, Lei Cui, Furu Wei 의 [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) 논문과 함께 발표했습니다.
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (FAIR and UIUC 에서 제공)은 Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar.의 [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527)논문과 함께 발표했습니다.
@@ -361,10 +362,10 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[MobileViTV2](https://huggingface.co/docs/transformers/model_doc/mobilevitv2)** (Apple 에서 제공)은 Sachin Mehta and Mohammad Rastegari.의 [Separable Self-attention for Mobile Vision Transformers](https://arxiv.org/abs/2206.02680)논문과 함께 발표했습니다.
 1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (Microsoft Research 에서) Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu 의 [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) 논문과 함께 발표했습니다.
 1. **[MPT](https://huggingface.co/docs/transformers/model_doc/mpt)** (MosaiML 에서 제공)은 the MosaicML NLP Team.의 [llm-foundry](https://github.com/mosaicml/llm-foundry/)논문과 함께 발표했습니다.
-1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (the University of Wisconsin - Madison 에서 제공)은 Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.의 [Multi Resolution Analysis (MRA)](https://arxiv.org/abs/2207.10284) 논문과 함께 발표했습니다.
+1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (the University of Wisconsin - Madison 에서 제공)은 Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.의 [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) 논문과 함께 발표했습니다.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (Google AI 에서) Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel 의 [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) 논문과 함께 발표했습니다.
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (RUC AI Box 에서) Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen 의 [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) 논문과 함께 발표했습니다.
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (SHI Labs 에서) Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi 의 [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) 논문과 함께 발표했습니다.
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (Huawei Noah’s Ark Lab 에서) Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu 의 [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) 논문과 함께 발표했습니다.
@@ -378,7 +379,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (Google AI 에서) 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 의 [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) 논문과 함께 발표했습니다.
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (Google AI 에서 제공)은 Matthias Minderer, Alexey Gritsenko, Neil Houlsby.의 [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683)논문과 함께 발표했습니다.
 1. **[PatchTSMixer](https://huggingface.co/docs/transformers/model_doc/patchtsmixer)** ( IBM Research 에서 제공)은 Vijay Ekambaram, Arindam Jati, Nam Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.의 [TSMixer: Lightweight MLP-Mixer Model for Multivariate Time Series Forecasting](https://arxiv.org/pdf/2306.09364.pdf)논문과 함께 발표했습니다.
-1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (IBM 에서 제공)은 Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.의 [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf)논문과 함께 발표했습니다.
+1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (IBM 에서 제공)은 Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.의 [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/abs/2211.14730)논문과 함께 발표했습니다.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (Google 에서) Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu 의 [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) 논문과 함께 발표했습니다.
 1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (Google 에서) Jason Phang, Yao Zhao, Peter J. Liu 의 [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) 논문과 함께 발표했습니다.
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (Deepmind 에서) 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 의 [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) 논문과 함께 발표했습니다.
@@ -391,24 +392,24 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi, Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/prophetnet)** (Microsoft Research 에서) Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou 의 [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) 논문과 함께 발표했습니다.
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (Nanjing University, The University of Hong Kong etc. 에서 제공)은 Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.의 [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf)논문과 함께 발표했습니다.
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc. 에서 제공)은 Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.의 [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797)논문과 함께 발표했습니다. 
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc. 에서 제공)은 Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.의 [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797)논문과 함께 발표했습니다. 
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (NVIDIA 에서) Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius 의 [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) 논문과 함께 발표했습니다.
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (the Qwen team, Alibaba Group 에서 제공)은 Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu.의 [Qwen Technical Report](https://arxiv.org/abs/2309.16609)논문과 함께 발표했습니다.
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (Facebook 에서) Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela 의 [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) 논문과 함께 발표했습니다.
 1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (Google Research 에서) Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat and Ming-Wei Chang 의 [REALM: Retrieval-Augmented Language Model Pre-Training](https://arxiv.org/abs/2002.08909) 논문과 함께 발표했습니다.
 1. **[Reformer](https://huggingface.co/docs/transformers/model_doc/reformer)** (Google Research 에서) Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya 의 [Reformer: The Efficient Transformer](https://arxiv.org/abs/2001.04451) 논문과 함께 발표했습니다.
 1. **[RegNet](https://huggingface.co/docs/transformers/model_doc/regnet)** (META Research 에서) Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár 의 [Designing Network Design Space](https://arxiv.org/abs/2003.13678) 논문과 함께 발표했습니다.
-1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (Google Research 에서) Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder 의 [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/pdf/2010.12821.pdf) 논문과 함께 발표했습니다.
+1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (Google Research 에서) Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder 의 [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/abs/2010.12821) 논문과 함께 발표했습니다.
 1. **[ResNet](https://huggingface.co/docs/transformers/model_doc/resnet)** (Microsoft Research 에서) Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun 의 [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) 논문과 함께 발표했습니다.
-1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (Facebook 에서) Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov 의 a [Robustly Optimized BERT Pretraining Approach](https://arxiv.org/abs/1907.11692) 논문과 함께 발표했습니다.
+1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (Facebook 에서) Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov 의 a [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://arxiv.org/abs/1907.11692) 논문과 함께 발표했습니다.
 1. **[RoBERTa-PreLayerNorm](https://huggingface.co/docs/transformers/model_doc/roberta-prelayernorm)** (Facebook 에서) Myle Ott, Sergey Edunov, Alexei Baevski, Angela Fan, Sam Gross, Nathan Ng, David Grangier, Michael Auli 의 [fairseq: A Fast, Extensible Toolkit for Sequence Modeling](https://arxiv.org/abs/1904.01038) 논문과 함께 발표했습니다.
 1. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (WeChatAI 에서) HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou 의 [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) 논문과 함께 발표했습니다.
-1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (ZhuiyiTechnology 에서) Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu 의 a [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/pdf/2104.09864v1.pdf) 논문과 함께 발표했습니다.
+1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (ZhuiyiTechnology 에서) Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu 의 a [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/abs/2104.09864) 논문과 함께 발표했습니다.
 1. **[RWKV](https://huggingface.co/docs/transformers/model_doc/rwkv)** (Bo Peng 에서 제공)은 Bo Peng.의 [this repo](https://github.com/BlinkDL/RWKV-LM)논문과 함께 발표했습니다.
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (NVIDIA 에서) Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo 의 [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) 논문과 함께 발표했습니다.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (Beijing Academy of Artificial Intelligence (BAAI 에서 제공)은 Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.의 [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284)논문과 함께 발표했습니다.
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (Beijing Academy of Artificial Intelligence (BAAI 에서 제공)은 Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.의 [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284)논문과 함께 발표했습니다.
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (Meta AI 에서 제공)은 Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.의 [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf)논문과 함께 발표했습니다.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (ASAPP 에서) Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi 의 [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) 논문과 함께 발표했습니다.
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (ASAPP 에서) Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi 의 [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) 논문과 함께 발표했습니다.
@@ -419,7 +420,8 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (Tel Aviv University 에서) Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy 의 [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) 논문과 함께 발표했습니다.
 1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (Berkeley 에서) Forrest N. Iandola, Albert E. Shaw, Ravi Krishna, and Kurt W. Keutzer 의 [SqueezeBERT: What can computer vision teach NLP about efficient neural networks?](https://arxiv.org/abs/2006.11316) 논문과 함께 발표했습니다.
 1. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report)](https://stability.wandb.io/stability-llm/stable-lm/reports/StableLM-3B-4E1T--VmlldzoyMjU4?accessToken=u3zujipenkx5g7rtcj9qojjgxpconyjktjkli2po09nffrffdhhchq045vp0wyfo) by  Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu. 
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (MBZUAI 에서 제공)은 Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.의 [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446)논문과 함께 발표했습니다.
 1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (Microsoft 에서) Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo 의 [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) 논문과 함께 발표했습니다.
 1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (Microsoft 에서) Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo 의 [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) 논문과 함께 발표했습니다.
@@ -437,14 +439,14 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (Microsoft 에서) Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei 의 [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) 논문과 함께 발표했습니다.
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (from UNC Chapel Hill 에서) Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal 의 [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) 논문과 함께 발표했습니다.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (Intel 에서) Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding 의 [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) 논문과 함께 발표했습니다.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (Microsoft Research 에서 제공)은 Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.의 [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623)논문과 함께 발표했습니다.
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (Microsoft Research 에서 제공)은 Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.의 [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623)논문과 함께 발표했습니다.
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (Google Research 에서) Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzle 의 [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) 논문과 함께 발표했습니다.
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (Google Research 에서 제공)은 Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.의 [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi)논문과 함께 발표했습니다.
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (Microsoft Research 에서) Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang 의 [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) 논문과 함께 발표했습니다.
 1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (Microsoft Research 에서) Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu 의 [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) 논문과 함께 발표했습니다.
 1. **[UnivNet](https://huggingface.co/docs/transformers/model_doc/univnet)** (from Kakao Corporation) released with the paper [UnivNet: A Neural Vocoder with Multi-Resolution Spectrogram Discriminators for High-Fidelity Waveform Generation](https://arxiv.org/abs/2106.07889) by Won Jang, Dan Lim, Jaesam Yoon, Bongwan Kim, and Juntae Kim.
 1. **[UPerNet](https://huggingface.co/docs/transformers/model_doc/upernet)** (Peking University 에서 제공)은 Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, Jian Sun.의 [Unified Perceptual Parsing for Scene Understanding](https://arxiv.org/abs/1807.10221)논문과 함께 발표했습니다.
-1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (Tsinghua University and Nankai University 에서) Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu 의 [Visual Attention Network](https://arxiv.org/pdf/2202.09741.pdf) 논문과 함께 발표했습니다.
+1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (Tsinghua University and Nankai University 에서) Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu 의 [Visual Attention Network](https://arxiv.org/abs/2202.09741) 논문과 함께 발표했습니다.
 1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (Multimedia Computing Group, Nanjing University 에서) Zhan Tong, Yibing Song, Jue Wang, Limin Wang 의 [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) 논문과 함께 발표했습니다.
 1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (NAVER AI Lab/Kakao Enterprise/Kakao Brain 에서) Wonjae Kim, Bokyung Son, Ildoo Kim 의 [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) 논문과 함께 발표했습니다.
 1. **[VipLlava](https://huggingface.co/docs/transformers/model_doc/vipllava)** (University of Wisconsin–Madison 에서 제공)은 Mu Cai, Haotian Liu, Siva Karthik Mustikovela, Gregory P. Meyer, Yuning Chai, Dennis Park, Yong Jae Lee.의 [Making Large Multimodal Models Understand Arbitrary Visual Prompts](https://arxiv.org/abs/2312.00784)논문과 함께 발표했습니다.

README_pt-br.md

@@ -336,7 +336,7 @@ Número atual de pontos de verificação: ![](https://img.shields.io/endpoint?ur
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker. 
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
@@ -381,7 +381,7 @@ Número atual de pontos de verificação: ![](https://img.shields.io/endpoint?ur
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (from Microsoft Research) released with the paper [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) by Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (from ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. Released with the paper [blog post](https://www.adept.ai/blog/fuyu-8b) 
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
@@ -414,6 +414,7 @@ Número atual de pontos de verificação: ![](https://img.shields.io/endpoint?ur
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (from The FAIR team of Meta AI) released with the paper [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) by Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.
 1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (from The FAIR team of Meta AI) released with the paper [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
@@ -421,7 +422,7 @@ Número atual de pontos de verificação: ![](https://img.shields.io/endpoint?ur
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (from Facebook) released with the paper [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Machine translation models trained using [OPUS](http://opus.nlpl.eu/) data by Jörg Tiedemann. The [Marian Framework](https://marian-nmt.github.io/) is being developed by the Microsoft Translator Team.
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (from FAIR and UIUC) released with the paper [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) by Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar.
@@ -447,7 +448,7 @@ Número atual de pontos de verificação: ![](https://img.shields.io/endpoint?ur
 1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (from the University of Wisconsin - Madison) released with the paper [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) by Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
@@ -474,7 +475,7 @@ Número atual de pontos de verificação: ![](https://img.shields.io/endpoint?ur
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi and Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/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.
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (from Nanjing University, The University of Hong Kong etc.) released with the paper [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (from the Qwen team, Alibaba Group) released with the paper [Qwen Technical Report](https://arxiv.org/abs/2309.16609) by Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu.
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
@@ -491,7 +492,7 @@ Número atual de pontos de verificação: ![](https://img.shields.io/endpoint?ur
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (from Meta AI) released with the paper [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) by Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -502,7 +503,8 @@ Número atual de pontos de verificação: ![](https://img.shields.io/endpoint?ur
 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. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report) by Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu.
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (from MBZUAI) released with the paper [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) by Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.
 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/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.
@@ -520,7 +522,7 @@ Número atual de pontos de verificação: ![](https://img.shields.io/endpoint?ur
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft), released together with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (from UNC Chapel Hill) released with the paper [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) by Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (from Intel) released with the paper [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) by Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (from Google Research) released with the paper [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) by Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.
 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.

README_ru.md

@@ -326,7 +326,7 @@ conda install conda-forge::transformers
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker. 
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
@@ -371,7 +371,7 @@ conda install conda-forge::transformers
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (from Microsoft Research) released with the paper [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) by Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (from ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. Released with the paper [blog post](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
@@ -402,8 +402,9 @@ conda install conda-forge::transformers
 1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (from Meta AI) released with the paper [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) by Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze.
 1. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (from South China University of Technology) released with the paper [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) by Jiapeng Wang, Lianwen Jin, Kai Ding.
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (from The FAIR team of Meta AI) released with the paper [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) by Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.
-1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (from The FAIR team of Meta AI) released with the paper [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/XXX) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.
+1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (from The FAIR team of Meta AI) released with the paper [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
@@ -411,7 +412,7 @@ conda install conda-forge::transformers
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (from Facebook) released with the paper [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Machine translation models trained using [OPUS](http://opus.nlpl.eu/) data by Jörg Tiedemann. The [Marian Framework](https://marian-nmt.github.io/) is being developed by the Microsoft Translator Team.
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (from FAIR and UIUC) released with the paper [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) by Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar.
@@ -437,7 +438,7 @@ conda install conda-forge::transformers
 1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (from the University of Wisconsin - Madison) released with the paper [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) by Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
@@ -464,7 +465,7 @@ conda install conda-forge::transformers
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi and Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/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.
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (from Nanjing University, The University of Hong Kong etc.) released with the paper [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (from the Qwen team, Alibaba Group) released with the paper [Qwen Technical Report](https://arxiv.org/abs/2309.16609) by Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu.
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
@@ -481,7 +482,7 @@ conda install conda-forge::transformers
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (from Meta AI) released with the paper [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) by Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -492,7 +493,8 @@ conda install conda-forge::transformers
 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. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report) by Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu.
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (from MBZUAI) released with the paper [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) by Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.
 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/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.
@@ -510,7 +512,7 @@ conda install conda-forge::transformers
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft), released together with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (from UNC Chapel Hill) released with the paper [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) by Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (from Intel) released with the paper [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) by Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (from Google Research) released with the paper [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) by Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.
 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.

README_te.md

@@ -328,7 +328,7 @@ Flax, PyTorch లేదా TensorFlow యొక్క ఇన్‌స్టా
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker. 
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
@@ -373,7 +373,7 @@ Flax, PyTorch లేదా TensorFlow యొక్క ఇన్‌స్టా
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (from Microsoft Research) released with the paper [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) by Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (from ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. Released with the paper [blog post](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
@@ -406,6 +406,7 @@ Flax, PyTorch లేదా TensorFlow యొక్క ఇన్‌స్టా
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (from The FAIR team of Meta AI) released with the paper [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) by Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.
 1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (from The FAIR team of Meta AI) released with the paper [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
@@ -413,7 +414,7 @@ Flax, PyTorch లేదా TensorFlow యొక్క ఇన్‌స్టా
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (from Facebook) released with the paper [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Machine translation models trained using [OPUS](http://opus.nlpl.eu/) data by Jörg Tiedemann. The [Marian Framework](https://marian-nmt.github.io/) is being developed by the Microsoft Translator Team.
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (from FAIR and UIUC) released with the paper [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) by Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar.
@@ -439,7 +440,7 @@ Flax, PyTorch లేదా TensorFlow యొక్క ఇన్‌స్టా
 1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (from the University of Wisconsin - Madison) released with the paper [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) by Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
@@ -466,7 +467,7 @@ Flax, PyTorch లేదా TensorFlow యొక్క ఇన్‌స్టా
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi and Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/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.
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (from Nanjing University, The University of Hong Kong etc.) released with the paper [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (from the Qwen team, Alibaba Group) released with the paper [Qwen Technical Report](https://arxiv.org/abs/2309.16609) by Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu.
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
@@ -483,7 +484,7 @@ Flax, PyTorch లేదా TensorFlow యొక్క ఇన్‌స్టా
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (from Meta AI) released with the paper [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) by Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -494,7 +495,8 @@ Flax, PyTorch లేదా TensorFlow యొక్క ఇన్‌స్టా
 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. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report) by Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu.
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (from MBZUAI) released with the paper [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) by Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.
 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/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.
@@ -512,7 +514,7 @@ Flax, PyTorch లేదా TensorFlow యొక్క ఇన్‌స్టా
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft), released together with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (from UNC Chapel Hill) released with the paper [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) by Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (from Intel) released with the paper [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) by Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (from Google Research) released with the paper [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) by Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.
 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.

README_vi.md

@@ -308,9 +308,9 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
 1. **[BigBird-Pegasus](https://huggingface.co/docs/transformers/model_doc/bigbird_pegasus)** (từ Google Research) được phát hành với bài báo [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) của Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang và Amr Ahmed.
 1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (từ Google Research) được phát hành với bài báo [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) của Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang và Amr Ahmed.
 1. **[BioGpt](https://huggingface.co/docs/transformers/model_doc/biogpt)** (từ Microsoft Research AI4Science) được phát hành với bài báo [BioGPT: generative pre-trained transformer for biomedical text generation and mining](https://academic.oup.com/bib/advance-article/doi/10.1093/bib/bbac409/6713511?guestAccessKey=a66d9b5d-4f83-4017-bb52-405815c907b9) by Renqian Luo, Liai Sun, Yingce Xia, Tao Qin, Sheng Zhang, Hoifung Poon and Tie-Yan Liu.
-1. **[BiT](https://huggingface.co/docs/transformers/model_doc/bit)** (từ Google AI) được phát hành với bài báo [Big Transfer (BiT): Học biểu diễn hình ảnh tổng quát](https://arxiv.org/abs/1912.11370) của Alexander Kolesnikov, Lucas Beyer, Xiaohua Zhai, Joan Puigcerver, Jessica Yung, Sylvain Gelly, Neil Houlsby.
-1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (từ Facebook) được phát hành với bài báo [Công thức xây dựng một chatbot miền mở](https://arxiv.org/abs/2004.13637) của Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
-1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (từ Facebook) được phát hành với bài báo [Công thức xây dựng một chatbot miền mở](https://arxiv.org/abs/2004.13637) của Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
+1. **[BiT](https://huggingface.co/docs/transformers/model_doc/bit)** (từ Google AI) được phát hành với bài báo [Big Transfer (BiT): General Visual Representation Learning](https://arxiv.org/abs/1912.11370) của Alexander Kolesnikov, Lucas Beyer, Xiaohua Zhai, Joan Puigcerver, Jessica Yung, Sylvain Gelly, Neil Houlsby.
+1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (từ Facebook) được phát hành với bài báo [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) của Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
+1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (từ Facebook) được phát hành với bài báo [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) của Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
 1. **[BLIP](https://huggingface.co/docs/transformers/model_doc/blip)** (từ Salesforce) được phát hành với bài báo [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://arxiv.org/abs/2201.12086) của Junnan Li, Dongxu Li, Caiming Xiong, Steven Hoi.
 1. **[BLIP-2](https://huggingface.co/docs/transformers/model_doc/blip-2)** (từ Salesforce) được phát hành với bài báo [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://arxiv.org/abs/2301.12597) by Junnan Li, Dongxu Li, Silvio Savarese, Steven Hoi.
 1. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (từ BigScience workshop) released by the [BigScience Workshop](https://bigscience.huggingface.co/).
@@ -327,7 +327,7 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** được phát hành với bài báo [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (từ Salesforce) được phát hành với bài báo [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (từ MetaAI) được phát hành với bài báo [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (từ Cohere) được phát hành với bài báo [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (từ Cohere) được phát hành với bài báo [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (từ Microsoft Research Asia) được phát hành với bài báo [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (từ YituTech) được phát hành với bài báo [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (từ Facebook AI) được phát hành với bài báo [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
@@ -372,7 +372,7 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (từ Microsoft Research) được phát hành với bài báo [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) by Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (từ CMU/Google Brain) được phát hành với bài báo [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (từ ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. được phát hành với bài báo [blog post](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (từ Google) được phát hành với bài báo [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (từ Google) được phát hành với bài báo [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (từ Microsoft Research) được phát hành với bài báo [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (từ KAIST) được phát hành với bài báo [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (từ OpenAI) được phát hành với bài báo [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
@@ -405,6 +405,7 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (từ The FAIR team of Meta AI) được phát hành với bài báo [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) by Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.
 1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (từ The FAIR team of Meta AI) được phát hành với bài báo [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom.
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (từ Microsoft Research & University of Wisconsin-Madison) được phát hành với bài báo [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (từ Microsoft Research & University of Wisconsin-Madison) được phát hành với bài báo [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (từ AllenAI) được phát hành với bài báo [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (từ Google AI) được phát hành với bài báo [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (từ Studio Ousia) được phát hành với bài báo [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
@@ -412,7 +413,7 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (từ Facebook) được phát hành với bài báo [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (từ Facebook) được phát hành với bài báo [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (từ Google) được phát hành với bài báo [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (từ Albert Gu and Tri Dao) được phát hành với bài báo [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (từ Albert Gu and Tri Dao) được phát hành với bài báo [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Machine translation models trained using [OPUS](http://opus.nlpl.eu/) data by Jörg Tiedemann. The [Marian Framework](https://marian-nmt.github.io/) is being developed by the Microsoft Translator Team.
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (từ Microsoft Research Asia) được phát hành với bài báo [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (từ FAIR and UIUC) được phát hành với bài báo [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) by Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar.
@@ -438,7 +439,7 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
 1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (từ the University of Wisconsin - Madison) được phát hành với bài báo [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) by Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (từ Google AI) được phát hành với bài báo [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (từ Meta) được phát hành với bài báo [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (từ Meta) được phát hành với bài báo [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (từ Meta) được phát hành với bài báo [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (từ RUC AI Box) được phát hành với bài báo [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (từ SHI Labs) được phát hành với bài báo [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (từ Huawei Noah’s Ark Lab) được phát hành với bài báo [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
@@ -465,7 +466,7 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** được phát hành với bài báo [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi and Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/prophetnet)** (từ Microsoft Research) được phát hành với bài báo [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.
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (từ Nanjing University, The University of Hong Kong etc.) được phát hành với bài báo [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (từ Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) được phát hành với bài báo [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (từ Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) được phát hành với bài báo [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (từ NVIDIA) được phát hành với bài báo [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (từ the Qwen team, Alibaba Group) được phát hành với bài báo [Qwen Technical Report](https://arxiv.org/abs/2309.16609) by Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu.
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (từ Facebook) được phát hành với bài báo [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
@@ -482,7 +483,7 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (từ Meta AI) được phát hành với bài báo [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (từ Meta AI) được phát hành với bài báo [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (từ NVIDIA) được phát hành với bài báo [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (từ Beijing Academy of Artificial Intelligence (BAAI) được phát hành với bài báo [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (từ Beijing Academy of Artificial Intelligence (BAAI) được phát hành với bài báo [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (từ Meta AI) được phát hành với bài báo [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) by Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (từ ASAPP) được phát hành với bài báo [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (từ ASAPP) được phát hành với bài báo [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -493,7 +494,8 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (từ 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)** (từ Berkeley) được phát hành với bài báo [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. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (từ Stability AI) được phát hành với bài báo [StableLM 3B 4E1T (Technical Report)](https://stability.wandb.io/stability-llm/stable-lm/reports/StableLM-3B-4E1T--VmlldzoyMjU4?accessToken=u3zujipenkx5g7rtcj9qojjgxpconyjktjkli2po09nffrffdhhchq045vp0wyfo) by Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu.
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (từ BigCode team) được phát hành với bài báo [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (từ BigCode team) được phát hành với bài báo [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (từ MagicLeap) được phát hành với bài báo [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (từ MBZUAI) được phát hành với bài báo [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) by Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.
 1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (từ Microsoft) được phát hành với bài báo [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/model_doc/swinv2)** (từ Microsoft) được phát hành với bài báo [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.
@@ -511,7 +513,7 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (từ Microsoft), released together with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (từ UNC Chapel Hill) được phát hành với bài báo [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) by Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (từ Intel) được phát hành với bài báo [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) by Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (từ Microsoft Research) được phát hành với bài báo [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (từ Microsoft Research) được phát hành với bài báo [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (từ Google Research) được phát hành với bài báo [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (từ Google Research) được phát hành với bài báo [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) by Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (từ Microsoft Research) được phát hành với bài báo [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.

README_zh-hans.md

@@ -248,7 +248,7 @@ conda install conda-forge::transformers
 1. **[Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer)** (来自 MIT) 伴随论文 [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) 由 Yuan Gong, Yu-An Chung, James Glass 发布。
 1. **[Autoformer](https://huggingface.co/docs/transformers/model_doc/autoformer)** (from Tsinghua University) released with the paper [Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting](https://arxiv.org/abs/2106.13008) by Haixu Wu, Jiehui Xu, Jianmin Wang, Mingsheng Long.
 1. **[Bark](https://huggingface.co/docs/transformers/model_doc/bark)** (from Suno) released in the repository [suno-ai/bark](https://github.com/suno-ai/bark) by Suno AI team.
-1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (来自 Facebook) 伴随论文 [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/pdf/1910.13461.pdf) 由 Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer 发布。
+1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (来自 Facebook) 伴随论文 [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) 由 Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer 发布。
 1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (来自 École polytechnique) 伴随论文 [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) 由 Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis 发布。
 1. **[BARTpho](https://huggingface.co/docs/transformers/model_doc/bartpho)** (来自 VinAI Research) 伴随论文 [BARTpho: Pre-trained Sequence-to-Sequence Models for Vietnamese](https://arxiv.org/abs/2109.09701) 由 Nguyen Luong Tran, Duong Minh Le and Dat Quoc Nguyen 发布。
 1. **[BEiT](https://huggingface.co/docs/transformers/model_doc/beit)** (来自 Microsoft) 伴随论文 [BEiT: BERT Pre-Training of Image Transformers](https://arxiv.org/abs/2106.08254) 由 Hangbo Bao, Li Dong, Furu Wei 发布。
@@ -277,7 +277,7 @@ conda install conda-forge::transformers
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (来自 Salesforce) 伴随论文 [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) 由 Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong 发布。
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (来自 MetaAI) 伴随论文 [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) 由 Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve 发布。
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (来自 Cohere) 伴随论文 [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) 由 Cohere 发布。 
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (来自 Cohere) 伴随论文 [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) 由 Cohere 发布。 
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (来自 Microsoft Research Asia) 伴随论文 [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) 由 Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang 发布。
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (来自 YituTech) 伴随论文 [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) 由 Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan 发布。
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (来自 Facebook AI) 伴随论文 [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) 由 Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie 发布。
@@ -313,7 +313,7 @@ conda install conda-forge::transformers
 1. **[ErnieM](https://huggingface.co/docs/transformers/model_doc/ernie_m)** (来自 Baidu) 伴随论文 [ERNIE-M: Enhanced Multilingual Representation by Aligning Cross-lingual Semantics with Monolingual Corpora](https://arxiv.org/abs/2012.15674) 由 Xuan Ouyang, Shuohuan Wang, Chao Pang, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang 发布。
 1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (from Meta AI) are transformer protein language models.  **ESM-1b** was released with the paper [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus. **ESM-1v** was released with the paper [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648) by Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives. **ESM-2** was released with the paper [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
 1. **[Falcon](https://huggingface.co/docs/transformers/model_doc/falcon)** (from Technology Innovation Institute) by Almazrouei, Ebtesam and Alobeidli, Hamza and Alshamsi, Abdulaziz and Cappelli, Alessandro and Cojocaru, Ruxandra and Debbah, Merouane and Goffinet, Etienne and Heslow, Daniel and Launay, Julien and Malartic, Quentin and Noune, Badreddine and Pannier, Baptiste and Penedo, Guilherme.
-1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (来自 ESPnet and Microsoft Research) 伴随论文 [Fastspeech 2: Fast And High-quality End-to-End Text To Speech](https://arxiv.org/pdf/2006.04558.pdf) 由 Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang, and Yuekai Zhang 发布。
+1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (来自 ESPnet and Microsoft Research) 伴随论文 [Recent Developments On Espnet Toolkit Boosted By Conformer](https://arxiv.org/abs/2010.13956) 由 Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang, and Yuekai Zhang 发布。
 1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 1. **[FLAN-UL2](https://huggingface.co/docs/transformers/model_doc/flan-ul2)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-ul2-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 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。
@@ -322,7 +322,7 @@ conda install conda-forge::transformers
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (来自 Microsoft Research) 伴随论文 [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) 由 Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao 发布。
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (来自 CMU/Google Brain) 伴随论文 [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) 由 Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le 发布。
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (来自 ADEPT) 伴随论文 [blog post](https://www.adept.ai/blog/fuyu-8b) 由 Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar 发布。
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (来自 Google) 伴随论文 [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) 由 the Gemma Google team 发布。
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (来自 Google) 伴随论文 [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) 由 the Gemma Google team 发布。
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (来自 Microsoft Research) 伴随论文 [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) 由 Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang 发布。
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (来自 KAIST) 伴随论文 [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) 由 Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim 发布。
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (来自 OpenAI) 伴随论文 [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) 由 Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever 发布。
@@ -353,8 +353,9 @@ conda install conda-forge::transformers
 1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (来自 Meta AI) 伴随论文 [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) 由 Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze 发布。
 1. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (来自 South China University of Technology) 伴随论文 [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) 由 Jiapeng Wang, Lianwen Jin, Kai Ding 发布。
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (来自 The FAIR team of Meta AI) 伴随论文 [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) 由 Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample 发布。
-1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (来自 The FAIR team of Meta AI) 伴随论文 [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/XXX) 由 Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom. 发布。
+1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (来自 The FAIR team of Meta AI) 伴随论文 [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) 由 Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom. 发布。
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (来自 Microsoft Research & University of Wisconsin-Madison) 伴随论文 [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) 由 Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee 发布。
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (来自 Microsoft Research & University of Wisconsin-Madison) 伴随论文 [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) 由 Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee 发布。
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (来自 AllenAI) 伴随论文 [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) 由 Iz Beltagy, Matthew E. Peters, Arman Cohan 发布。
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (来自 Google AI) released 伴随论文 [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) 由 Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang 发布。
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (来自 Studio Ousia) 伴随论文 [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) 由 Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto 发布。
@@ -362,7 +363,7 @@ conda install conda-forge::transformers
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (来自 Facebook) 伴随论文 [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) 由 Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert 发布。
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (来自 Facebook) 伴随论文 [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) 由 Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin 发布。
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (来自 Albert Gu and Tri Dao) 伴随论文 [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) 由 Albert Gu and Tri Dao 发布。
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (来自 Albert Gu and Tri Dao) 伴随论文 [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) 由 Albert Gu and Tri Dao 发布。
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** 用 [OPUS](http://opus.nlpl.eu/) 数据训练的机器翻译模型由 Jörg Tiedemann 发布。[Marian Framework](https://marian-nmt.github.io/) 由微软翻译团队开发。
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (来自 Microsoft Research Asia) 伴随论文 [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) 由 Junlong Li, Yiheng Xu, Lei Cui, Furu Wei 发布。
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (来自 FAIR and UIUC) 伴随论文 [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) 由 Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar 发布。
@@ -385,10 +386,10 @@ conda install conda-forge::transformers
 1. **[MobileViTV2](https://huggingface.co/docs/transformers/model_doc/mobilevitv2)** (来自 Apple) 伴随论文 [Separable Self-attention for Mobile Vision Transformers](https://arxiv.org/abs/2206.02680) 由 Sachin Mehta and Mohammad Rastegari 发布。
 1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (来自 Microsoft Research) 伴随论文 [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) 由 Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu 发布。
 1. **[MPT](https://huggingface.co/docs/transformers/model_doc/mpt)** (来自 MosaiML) 伴随论文 [llm-foundry](https://github.com/mosaicml/llm-foundry/) 由 the MosaicML NLP Team 发布。
-1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (来自 the University of Wisconsin - Madison) 伴随论文 [Multi Resolution Analysis (MRA)](https://arxiv.org/abs/2207.10284) 由 Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh 发布。
+1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (来自 the University of Wisconsin - Madison) 伴随论文 [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) 由 Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh 发布。
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (来自 Google AI) 伴随论文 [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) 由 Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel 发布。
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (来自 中国人民大学 AI Box) 伴随论文 [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) 由 Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen 发布。
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (来自 SHI Labs) 伴随论文 [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) 由 Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi 发布。
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (来自华为诺亚方舟实验室) 伴随论文 [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) 由 Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu 发布。
@@ -402,7 +403,7 @@ conda install conda-forge::transformers
 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. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (来自 Google AI) 伴随论文 [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683) 由 Matthias Minderer, Alexey Gritsenko, Neil Houlsby 发布。
 1. **[PatchTSMixer](https://huggingface.co/docs/transformers/model_doc/patchtsmixer)** (来自  IBM Research) 伴随论文 [TSMixer: Lightweight MLP-Mixer Model for Multivariate Time Series Forecasting](https://arxiv.org/pdf/2306.09364.pdf) 由 Vijay Ekambaram, Arindam Jati, Nam Nguyen, Phanwadee Sinthong, Jayant Kalagnanam 发布。
-1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (来自 IBM) 伴随论文 [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf) 由 Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam 发布。
+1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (来自 IBM) 伴随论文 [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/abs/2211.14730) 由 Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam 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/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。
@@ -415,24 +416,24 @@ conda install conda-forge::transformers
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi, Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/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。
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (来自 Nanjing University, The University of Hong Kong etc.) 伴随论文 [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) 由 Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao 发布。
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (来自 Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) 伴随论文 [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) 由 Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao 发布。 
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (来自 Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) 伴随论文 [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) 由 Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao 发布。 
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (来自 NVIDIA) 伴随论文 [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) 由 Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius 发布。
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (来自 the Qwen team, Alibaba Group) 伴随论文 [Qwen Technical Report](https://arxiv.org/abs/2309.16609) 由 Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu 发布。
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (来自 Facebook) 伴随论文 [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) 由 Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela 发布。
 1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (来自 Google Research) 伴随论文 [REALM: Retrieval-Augmented Language Model Pre-Training](https://arxiv.org/abs/2002.08909) 由 Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat and Ming-Wei Chang 发布。
 1. **[Reformer](https://huggingface.co/docs/transformers/model_doc/reformer)** (来自 Google Research) 伴随论文 [Reformer: The Efficient Transformer](https://arxiv.org/abs/2001.04451) 由 Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya 发布。
 1. **[RegNet](https://huggingface.co/docs/transformers/model_doc/regnet)** (from META Research) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár.
-1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (来自 Google Research) 伴随论文 [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/pdf/2010.12821.pdf) 由 Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder 发布。
+1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (来自 Google Research) 伴随论文 [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/abs/2010.12821) 由 Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder 发布。
 1. **[ResNet](https://huggingface.co/docs/transformers/model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun.
-1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (来自 Facebook), 伴随论文 [Robustly Optimized BERT Pretraining Approach](https://arxiv.org/abs/1907.11692) 由 Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov 发布。
+1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (来自 Facebook), 伴随论文 [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://arxiv.org/abs/1907.11692) 由 Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov 发布。
 1. **[RoBERTa-PreLayerNorm](https://huggingface.co/docs/transformers/model_doc/roberta-prelayernorm)** (来自 Facebook) 伴随论文 [fairseq: A Fast, Extensible Toolkit for Sequence Modeling](https://arxiv.org/abs/1904.01038) 由 Myle Ott, Sergey Edunov, Alexei Baevski, Angela Fan, Sam Gross, Nathan Ng, David Grangier, Michael Auli 发布。
 1. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (来自 WeChatAI), 伴随论文 [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) 由 HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou 发布。
-1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (来自 ZhuiyiTechnology), 伴随论文 [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/pdf/2104.09864v1.pdf) 由 Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu 发布。
+1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (来自 ZhuiyiTechnology), 伴随论文 [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/abs/2104.09864) 由 Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu 发布。
 1. **[RWKV](https://huggingface.co/docs/transformers/model_doc/rwkv)** (来自 Bo Peng) 伴随论文 [this repo](https://github.com/BlinkDL/RWKV-LM) 由 Bo Peng 发布。
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (来自 NVIDIA) 伴随论文 [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) 由 Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo 发布。
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (来自 Beijing Academy of Artificial Intelligence (BAAI) 伴随论文 [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) 由 Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang 发布。
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (来自 Beijing Academy of Artificial Intelligence (BAAI) 伴随论文 [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) 由 Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang 发布。
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (来自 Meta AI) 伴随论文 [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) 由 Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick 发布。
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (来自 ASAPP) 伴随论文 [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) 由 Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi 发布。
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (来自 ASAPP) 伴随论文 [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) 由 Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi 发布。
@@ -443,7 +444,8 @@ conda install conda-forge::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. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)** (from Stability AI) released with the paper [StableLM 3B 4E1T (Technical Report)](https://stability.wandb.io/stability-llm/stable-lm/reports/StableLM-3B-4E1T--VmlldzoyMjU4?accessToken=u3zujipenkx5g7rtcj9qojjgxpconyjktjkli2po09nffrffdhhchq045vp0wyfo) by  Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu. 
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich. 
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (来自 MBZUAI) 伴随论文 [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) 由 Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan 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/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。
@@ -461,14 +463,14 @@ conda install conda-forge::transformers
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (来自 Microsoft) 伴随论文 [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) 由 Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei 发布。
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (来自 UNC Chapel Hill) 伴随论文 [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) 由 Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal 发布。
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (来自 Intel) 伴随论文 [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) 由 Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding 发布.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (来自 Microsoft Research) 伴随论文 [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) 由 Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal 发布。
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (来自 Microsoft Research) 伴随论文 [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) 由 Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal 发布。
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (来自 Google Research) 伴随论文 [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) 由 Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant 发布。
 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. **[UnivNet](https://huggingface.co/docs/transformers/model_doc/univnet)** (from Kakao Corporation) released with the paper [UnivNet: A Neural Vocoder with Multi-Resolution Spectrogram Discriminators for High-Fidelity Waveform Generation](https://arxiv.org/abs/2106.07889) by Won Jang, Dan Lim, Jaesam Yoon, Bongwan Kim, and Juntae Kim.
 1. **[UPerNet](https://huggingface.co/docs/transformers/model_doc/upernet)** (来自 Peking University) 伴随论文 [Unified Perceptual Parsing for Scene Understanding](https://arxiv.org/abs/1807.10221) 由 Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, Jian Sun 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. **[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/abs/2202.09741) 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/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. **[VipLlava](https://huggingface.co/docs/transformers/model_doc/vipllava)** (来自 University of Wisconsin–Madison) 伴随论文 [Making Large Multimodal Models Understand Arbitrary Visual Prompts](https://arxiv.org/abs/2312.00784) 由 Mu Cai, Haotian Liu, Siva Karthik Mustikovela, Gregory P. Meyer, Yuning Chai, Dennis Park, Yong Jae Lee 发布。

README_zh-hant.md

@@ -260,7 +260,7 @@ conda install conda-forge::transformers
 1. **[Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer)** (from MIT) released with the paper [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) by Yuan Gong, Yu-An Chung, James Glass.
 1. **[Autoformer](https://huggingface.co/docs/transformers/model_doc/autoformer)** (from Tsinghua University) released with the paper [Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting](https://arxiv.org/abs/2106.13008) by Haixu Wu, Jiehui Xu, Jianmin Wang, Mingsheng Long.
 1. **[Bark](https://huggingface.co/docs/transformers/model_doc/bark)** (from Suno) released in the repository [suno-ai/bark](https://github.com/suno-ai/bark) by Suno AI team.
-1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (from Facebook) released with the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/pdf/1910.13461.pdf) by Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer.
+1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (from Facebook) released with the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) by Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer.
 1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (from École polytechnique) released with the paper [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) by Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis.
 1. **[BARTpho](https://huggingface.co/docs/transformers/model_doc/bartpho)** (from VinAI Research) released with the paper [BARTpho: Pre-trained Sequence-to-Sequence Models for Vietnamese](https://arxiv.org/abs/2109.09701) by Nguyen Luong Tran, Duong Minh Le and Dat Quoc Nguyen.
 1. **[BEiT](https://huggingface.co/docs/transformers/model_doc/beit)** (from Microsoft) released with the paper [BEiT: BERT Pre-Training of Image Transformers](https://arxiv.org/abs/2106.08254) by Hangbo Bao, Li Dong, Furu Wei.
@@ -289,7 +289,7 @@ conda install conda-forge::transformers
 1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
 1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
-1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
+1. **[Cohere](https://huggingface.co/docs/transformers/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere. 
 1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
 1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
@@ -325,7 +325,7 @@ conda install conda-forge::transformers
 1. **[ErnieM](https://huggingface.co/docs/transformers/model_doc/ernie_m)** (from Baidu) released with the paper [ERNIE-M: Enhanced Multilingual Representation by Aligning Cross-lingual Semantics with Monolingual Corpora](https://arxiv.org/abs/2012.15674) by Xuan Ouyang, Shuohuan Wang, Chao Pang, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang.
 1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (from Meta AI) are transformer protein language models.  **ESM-1b** was released with the paper [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus. **ESM-1v** was released with the paper [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648) by Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives. **ESM-2** was released with the paper [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
 1. **[Falcon](https://huggingface.co/docs/transformers/model_doc/falcon)** (from Technology Innovation Institute) by Almazrouei, Ebtesam and Alobeidli, Hamza and Alshamsi, Abdulaziz and Cappelli, Alessandro and Cojocaru, Ruxandra and Debbah, Merouane and Goffinet, Etienne and Heslow, Daniel and Launay, Julien and Malartic, Quentin and Noune, Badreddine and Pannier, Baptiste and Penedo, Guilherme.
-1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (from ESPnet and Microsoft Research) released with the paper [Fastspeech 2: Fast And High-quality End-to-End Text To Speech](https://arxiv.org/pdf/2006.04558.pdf) by Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang, and Yuekai Zhang.
+1. **[FastSpeech2Conformer](https://huggingface.co/docs/transformers/model_doc/fastspeech2_conformer)** (from ESPnet and Microsoft Research) released with the paper [Recent Developments On Espnet Toolkit Boosted By Conformer](https://arxiv.org/abs/2010.13956) by Pengcheng Guo, Florian Boyer, Xuankai Chang, Tomoki Hayashi, Yosuke Higuchi, Hirofumi Inaguma, Naoyuki Kamo, Chenda Li, Daniel Garcia-Romero, Jiatong Shi, Jing Shi, Shinji Watanabe, Kun Wei, Wangyou Zhang, and Yuekai Zhang.
 1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 1. **[FLAN-UL2](https://huggingface.co/docs/transformers/model_doc/flan-ul2)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-ul2-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 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.
@@ -334,7 +334,7 @@ conda install conda-forge::transformers
 1. **[FocalNet](https://huggingface.co/docs/transformers/model_doc/focalnet)** (from Microsoft Research) released with the paper [Focal Modulation Networks](https://arxiv.org/abs/2203.11926) by Jianwei Yang, Chunyuan Li, Xiyang Dai, Lu Yuan, Jianfeng Gao.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[Fuyu](https://huggingface.co/docs/transformers/model_doc/fuyu)** (from ADEPT) Rohan Bavishi, Erich Elsen, Curtis Hawthorne, Maxwell Nye, Augustus Odena, Arushi Somani, Sağnak Taşırlar. Released with the paper [blog post](https://www.adept.ai/blog/fuyu-8b)
-1. **[Gemma](https://huggingface.co/docs/transformers/main/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
+1. **[Gemma](https://huggingface.co/docs/transformers/model_doc/gemma)** (from Google) released with the paper [Gemma: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/gemma-open-models/) by the Gemma Google team.
 1. **[GIT](https://huggingface.co/docs/transformers/model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
@@ -365,8 +365,9 @@ conda install conda-forge::transformers
 1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (from Meta AI) released with the paper [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) by Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze.
 1. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (from South China University of Technology) released with the paper [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) by Jiapeng Wang, Lianwen Jin, Kai Ding.
 1. **[LLaMA](https://huggingface.co/docs/transformers/model_doc/llama)** (from The FAIR team of Meta AI) released with the paper [LLaMA: Open and Efficient Foundation Language Models](https://arxiv.org/abs/2302.13971) by Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, Guillaume Lample.
-1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (from The FAIR team of Meta AI) released with the paper [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/XXX) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom..
+1. **[Llama2](https://huggingface.co/docs/transformers/model_doc/llama2)** (from The FAIR team of Meta AI) released with the paper [Llama2: Open Foundation and Fine-Tuned Chat Models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) by Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, Brian Fuller, Cynthia Gao, Vedanuj Goswami, Naman Goyal, Anthony Hartshorn, Saghar Hosseini, Rui Hou, Hakan Inan, Marcin Kardas, Viktor Kerkez Madian Khabsa, Isabel Kloumann, Artem Korenev, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Jenya Lee, Diana Liskovich, Yinghai Lu, Yuning Mao, Xavier Martinet, Todor Mihaylov, Pushka rMishra, Igor Molybog, Yixin Nie, Andrew Poulton, Jeremy Reizenstein, Rashi Rungta, Kalyan Saladi, Alan Schelten, Ruan Silva, Eric Michael Smith, Ranjan Subramanian, Xiaoqing EllenTan, Binh Tang, Ross Taylor, Adina Williams, Jian Xiang Kuan, Puxin Xu, Zheng Yan, Iliyan Zarov, Yuchen Zhang, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic, Sergey Edunov, Thomas Scialom..
 1. **[LLaVa](https://huggingface.co/docs/transformers/model_doc/llava)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
+1. **[LLaVA-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)** (from Microsoft Research & University of Wisconsin-Madison) released with the paper [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/abs/2310.03744) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
@@ -374,7 +375,7 @@ conda install conda-forge::transformers
 1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (from Facebook) released with the paper [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
 1. **[MADLAD-400](https://huggingface.co/docs/transformers/model_doc/madlad-400)** (from Google) released with the paper [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://arxiv.org/abs/2309.04662) by Sneha Kudugunta, Isaac Caswell, Biao Zhang, Xavier Garcia, Christopher A. Choquette-Choo, Katherine Lee, Derrick Xin, Aditya Kusupati, Romi Stella, Ankur Bapna, Orhan Firat.
-1. **[Mamba](https://huggingface.co/docs/transformers/main/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
+1. **[Mamba](https://huggingface.co/docs/transformers/model_doc/mamba)** (from Albert Gu and Tri Dao) released with the paper [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://arxiv.org/abs/2312.00752) by Albert Gu and Tri Dao.
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** Machine translation models trained using [OPUS](http://opus.nlpl.eu/) data by Jörg Tiedemann. The [Marian Framework](https://marian-nmt.github.io/) is being developed by the Microsoft Translator Team.
 1. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 1. **[Mask2Former](https://huggingface.co/docs/transformers/model_doc/mask2former)** (from FAIR and UIUC) released with the paper [Masked-attention Mask Transformer for Universal Image Segmentation](https://arxiv.org/abs/2112.01527) by Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar.
@@ -397,10 +398,10 @@ conda install conda-forge::transformers
 1. **[MobileViTV2](https://huggingface.co/docs/transformers/model_doc/mobilevitv2)** (from Apple) released with the paper [Separable Self-attention for Mobile Vision Transformers](https://arxiv.org/abs/2206.02680) by Sachin Mehta and Mohammad Rastegari.
 1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MPT](https://huggingface.co/docs/transformers/model_doc/mpt)** (from MosaiML) released with the paper [llm-foundry](https://github.com/mosaicml/llm-foundry/) by the MosaicML NLP Team.
-1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (from the University of Wisconsin - Madison) released with the paper [Multi Resolution Analysis (MRA)](https://arxiv.org/abs/2207.10284) by Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
+1. **[MRA](https://huggingface.co/docs/transformers/model_doc/mra)** (from the University of Wisconsin - Madison) released with the paper [Multi Resolution Analysis (MRA) for Approximate Self-Attention](https://arxiv.org/abs/2207.10284) by Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
 1. **[MusicGen](https://huggingface.co/docs/transformers/model_doc/musicgen)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
-1. **[MusicGen Melody](https://huggingface.co/docs/transformers/main/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
+1. **[MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody)** (from Meta) released with the paper [Simple and Controllable Music Generation](https://arxiv.org/abs/2306.05284) by Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi and Alexandre Défossez.
 1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
 1. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
@@ -414,7 +415,7 @@ conda install conda-forge::transformers
 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. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (from Google AI) released with the paper [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683) by Matthias Minderer, Alexey Gritsenko, Neil Houlsby.
 1. **[PatchTSMixer](https://huggingface.co/docs/transformers/model_doc/patchtsmixer)** (from  IBM Research) released with the paper [TSMixer: Lightweight MLP-Mixer Model for Multivariate Time Series Forecasting](https://arxiv.org/pdf/2306.09364.pdf) by Vijay Ekambaram, Arindam Jati, Nam Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
-1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (from IBM) released with the paper [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf) by Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
+1. **[PatchTST](https://huggingface.co/docs/transformers/model_doc/patchtst)** (from IBM) released with the paper [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/abs/2211.14730) by Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
 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/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.
@@ -427,24 +428,24 @@ conda install conda-forge::transformers
 1. **[Pop2Piano](https://huggingface.co/docs/transformers/model_doc/pop2piano)** released with the paper [Pop2Piano : Pop Audio-based Piano Cover Generation](https://arxiv.org/abs/2211.00895) by Jongho Choi, Kyogu Lee.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/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.
 1. **[PVT](https://huggingface.co/docs/transformers/model_doc/pvt)** (from Nanjing University, The University of Hong Kong etc.) released with the paper [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://arxiv.org/pdf/2102.12122.pdf) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao.
-1. **[PVTv2](https://huggingface.co/docs/transformers/main/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao. 
+1. **[PVTv2](https://huggingface.co/docs/transformers/model_doc/pvt_v2)** (from Shanghai AI Laboratory, Nanjing University, The University of Hong Kong etc.) released with the paper [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://arxiv.org/abs/2106.13797) by Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao. 
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
 1. **[Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2)** (from the Qwen team, Alibaba Group) released with the paper [Qwen Technical Report](https://arxiv.org/abs/2309.16609) by Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng Xu, Jin Xu, An Yang, Hao Yang, Jian Yang, Shusheng Yang, Yang Yao, Bowen Yu, Hongyi Yuan, Zheng Yuan, Jianwei Zhang, Xingxuan Zhang, Yichang Zhang, Zhenru Zhang, Chang Zhou, Jingren Zhou, Xiaohuan Zhou and Tianhang Zhu.
 1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
 1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (from Google Research) released with the paper [REALM: Retrieval-Augmented Language Model Pre-Training](https://arxiv.org/abs/2002.08909) by Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat and Ming-Wei Chang.
 1. **[Reformer](https://huggingface.co/docs/transformers/model_doc/reformer)** (from Google Research) released with the paper [Reformer: The Efficient Transformer](https://arxiv.org/abs/2001.04451) by Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya.
 1. **[RegNet](https://huggingface.co/docs/transformers/model_doc/regnet)** (from META Research) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár.
-1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (from Google Research) released with the paper [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/pdf/2010.12821.pdf) by Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder.
+1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (from Google Research) released with the paper [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/abs/2010.12821) by Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder.
 1. **[ResNet](https://huggingface.co/docs/transformers/model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun.
-1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (from Facebook), released together with the paper a [Robustly Optimized BERT Pretraining Approach](https://arxiv.org/abs/1907.11692) by Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov.
+1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (from Facebook), released together with the paper a [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://arxiv.org/abs/1907.11692) by Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov.
 1. **[RoBERTa-PreLayerNorm](https://huggingface.co/docs/transformers/model_doc/roberta-prelayernorm)** (from Facebook) released with the paper [fairseq: A Fast, Extensible Toolkit for Sequence Modeling](https://arxiv.org/abs/1904.01038) by Myle Ott, Sergey Edunov, Alexei Baevski, Angela Fan, Sam Gross, Nathan Ng, David Grangier, Michael Auli.
 1. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (from WeChatAI) released with the paper [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) by HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou.
-1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (from ZhuiyiTechnology), released together with the paper a [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/pdf/2104.09864v1.pdf) by Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu.
+1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (from ZhuiyiTechnology), released together with the paper a [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/abs/2104.09864) by Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu.
 1. **[RWKV](https://huggingface.co/docs/transformers/model_doc/rwkv)** (from Bo Peng) released with the paper [this repo](https://github.com/BlinkDL/RWKV-LM) by Bo Peng.
 1. **[SeamlessM4T](https://huggingface.co/docs/transformers/model_doc/seamless_m4t)** (from Meta AI) released with the paper [SeamlessM4T — Massively Multilingual & Multimodal Machine Translation](https://dl.fbaipublicfiles.com/seamless/seamless_m4t_paper.pdf) by the Seamless Communication team.
 1. **[SeamlessM4Tv2](https://huggingface.co/docs/transformers/model_doc/seamless_m4t_v2)** (from Meta AI) released with the paper [Seamless: Multilingual Expressive and Streaming Speech Translation](https://ai.meta.com/research/publications/seamless-multilingual-expressive-and-streaming-speech-translation/) by the Seamless Communication team.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
-1. **[SegGPT](https://huggingface.co/docs/transformers/main/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
+1. **[SegGPT](https://huggingface.co/docs/transformers/model_doc/seggpt)** (from Beijing Academy of Artificial Intelligence (BAAI) released with the paper [SegGPT: Segmenting Everything In Context](https://arxiv.org/abs/2304.03284) by Xinlong Wang, Xiaosong Zhang, Yue Cao, Wen Wang, Chunhua Shen, Tiejun Huang.
 1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (from Meta AI) released with the paper [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) by Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
 1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -455,7 +456,8 @@ conda install conda-forge::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. **[StableLm](https://huggingface.co/docs/transformers/model_doc/stablelm)**  released with the paper [StableLM 3B 4E1T (Technical Report)](https://stability.wandb.io/stability-llm/stable-lm/reports/StableLM-3B-4E1T--VmlldzoyMjU4?accessToken=u3zujipenkx5g7rtcj9qojjgxpconyjktjkli2po09nffrffdhhchq045vp0wyfo) by  Jonathan Tow, Marco Bellagente, Dakota Mahan, Carlos Riquelme Ruiz, Duy Phung, Maksym Zhuravinskyi, Nathan Cooper, Nikhil Pinnaparaju, Reshinth Adithyan, and James Baicoianu. 
-1. **[Starcoder2](https://huggingface.co/docs/transformers/main/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[Starcoder2](https://huggingface.co/docs/transformers/model_doc/starcoder2)** (from BigCode team) released with the paper [StarCoder 2 and The Stack v2: The Next Generation](https://arxiv.org/abs/2402.19173) by Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, Tianyang Liu, Max Tian, Denis Kocetkov, Arthur Zucker, Younes Belkada, Zijian Wang, Qian Liu, Dmitry Abulkhanov, Indraneil Paul, Zhuang Li, Wen-Ding Li, Megan Risdal, Jia Li, Jian Zhu, Terry Yue Zhuo, Evgenii Zheltonozhskii, Nii Osae Osae Dade, Wenhao Yu, Lucas Krauß, Naman Jain, Yixuan Su, Xuanli He, Manan Dey, Edoardo Abati, Yekun Chai, Niklas Muennighoff, Xiangru Tang, Muhtasham Oblokulov, Christopher Akiki, Marc Marone, Chenghao Mou, Mayank Mishra, Alex Gu, Binyuan Hui, Tri Dao, Armel Zebaze, Olivier Dehaene, Nicolas Patry, Canwen Xu, Julian McAuley, Han Hu, Torsten Scholak, Sebastien Paquet, Jennifer Robinson, Carolyn Jane Anderson, Nicolas Chapados, Mostofa Patwary, Nima Tajbakhsh, Yacine Jernite, Carlos Muñoz Ferrandis, Lingming Zhang, Sean Hughes, Thomas Wolf, Arjun Guha, Leandro von Werra, and Harm de Vries.
+1. **[SuperPoint](https://huggingface.co/docs/transformers/model_doc/superpoint)** (from MagicLeap) released with the paper [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
 1. **[SwiftFormer](https://huggingface.co/docs/transformers/model_doc/swiftformer)** (from MBZUAI) released with the paper [SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications](https://arxiv.org/abs/2303.15446) by Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan.
 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/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.
@@ -473,14 +475,14 @@ conda install conda-forge::transformers
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft) released with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
 1. **[TVLT](https://huggingface.co/docs/transformers/model_doc/tvlt)** (from UNC Chapel Hill) released with the paper [TVLT: Textless Vision-Language Transformer](https://arxiv.org/abs/2209.14156) by Zineng Tang, Jaemin Cho, Yixin Nie, Mohit Bansal.
 1. **[TVP](https://huggingface.co/docs/transformers/model_doc/tvp)** (from Intel) released with the paper [Text-Visual Prompting for Efficient 2D Temporal Video Grounding](https://arxiv.org/abs/2303.04995) by Yimeng Zhang, Xin Chen, Jinghan Jia, Sijia Liu, Ke Ding.
-1. **[UDOP](https://huggingface.co/docs/transformers/main/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
+1. **[UDOP](https://huggingface.co/docs/transformers/model_doc/udop)** (from Microsoft Research) released with the paper [Unifying Vision, Text, and Layout for Universal Document Processing](https://arxiv.org/abs/2212.02623) by Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal.
 1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UMT5](https://huggingface.co/docs/transformers/model_doc/umt5)** (from Google Research) released with the paper [UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining](https://openreview.net/forum?id=kXwdL1cWOAi) by Hyung Won Chung, Xavier Garcia, Adam Roberts, Yi Tay, Orhan Firat, Sharan Narang, Noah Constant.
 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. **[UnivNet](https://huggingface.co/docs/transformers/model_doc/univnet)** (from Kakao Corporation) released with the paper [UnivNet: A Neural Vocoder with Multi-Resolution Spectrogram Discriminators for High-Fidelity Waveform Generation](https://arxiv.org/abs/2106.07889) by Won Jang, Dan Lim, Jaesam Yoon, Bongwan Kim, and Juntae Kim.
 1. **[UPerNet](https://huggingface.co/docs/transformers/model_doc/upernet)** (from Peking University) released with the paper [Unified Perceptual Parsing for Scene Understanding](https://arxiv.org/abs/1807.10221) by Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, Jian Sun.
-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. **[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/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. **[VipLlava](https://huggingface.co/docs/transformers/model_doc/vipllava)** (from University of Wisconsin–Madison) released with the paper [Making Large Multimodal Models Understand Arbitrary Visual Prompts](https://arxiv.org/abs/2312.00784) by Mu Cai, Haotian Liu, Siva Karthik Mustikovela, Gregory P. Meyer, Yuning Chai, Dennis Park, Yong Jae Lee.

docs/source/en/_toctree.yml

@@ -134,7 +134,7 @@
   - local: custom_tools
     title: Custom Tools and Prompts
   - local: troubleshooting
-    title: Troubleshoot  
+    title: Troubleshoot
   - local: hf_quantizer
     title: Contribute new quantization method
   title: Developer guides
@@ -595,6 +595,8 @@
         title: SegFormer
       - local: model_doc/seggpt
         title: SegGpt
+      - local: model_doc/superpoint
+        title: SuperPoint
       - local: model_doc/swiftformer
         title: SwiftFormer
       - local: model_doc/swin
@@ -693,7 +695,7 @@
         title: VideoMAE
       - local: model_doc/vivit
         title: ViViT
-      title: Video models 
+      title: Video models
     - isExpanded: false
       sections:
       - local: model_doc/align
@@ -746,6 +748,8 @@
         title: LiLT
       - local: model_doc/llava
         title: Llava
+      - local: model_doc/llava_next
+        title: LLaVA-NeXT
       - local: model_doc/lxmert
         title: LXMERT
       - local: model_doc/matcha

docs/source/en/index.md

@@ -174,6 +174,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                         [LLaMA](model_doc/llama)                         |       ✅        |         ❌         |      ✅      |
 |                        [Llama2](model_doc/llama2)                        |       ✅        |         ❌         |      ✅      |
 |                         [LLaVa](model_doc/llava)                         |       ✅        |         ❌         |      ❌      |
+|                    [LLaVA-NeXT](model_doc/llava_next)                    |       ✅        |         ❌         |      ❌      |
 |                    [Longformer](model_doc/longformer)                    |       ✅        |         ✅         |      ❌      |
 |                        [LongT5](model_doc/longt5)                        |       ✅        |         ❌         |      ✅      |
 |                          [LUKE](model_doc/luke)                          |       ✅        |         ❌         |      ❌      |
@@ -266,6 +267,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                   [SqueezeBERT](model_doc/squeezebert)                   |       ✅        |         ❌         |      ❌      |
 |                      [StableLm](model_doc/stablelm)                      |       ✅        |         ❌         |      ❌      |
 |                    [Starcoder2](model_doc/starcoder2)                    |       ✅        |         ❌         |      ❌      |
+|                    [SuperPoint](model_doc/superpoint)                    |       ✅        |         ❌         |      ❌      |
 |                   [SwiftFormer](model_doc/swiftformer)                   |       ✅        |         ❌         |      ❌      |
 |                    [Swin Transformer](model_doc/swin)                    |       ✅        |         ✅         |      ❌      |
 |                 [Swin Transformer V2](model_doc/swinv2)                  |       ✅        |         ❌         |      ❌      |

docs/source/en/model_doc/auto.md

@@ -250,6 +250,10 @@ The following auto classes are available for the following computer vision tasks
 
 [[autodoc]] AutoModelForVideoClassification
 
+### AutoModelForKeypointDetection
+
+[[autodoc]] AutoModelForKeypointDetection
+
 ### AutoModelForMaskedImageModeling
 
 [[autodoc]] AutoModelForMaskedImageModeling

docs/source/en/model_doc/llava_next.md

@@ -0,0 +1,147 @@
+<!--Copyright 2024 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+
+⚠️ Note that this file is in Markdown but contain specific syntax for our doc-builder (similar to MDX) that may not be
+rendered properly in your Markdown viewer.
+
+-->
+
+# LLaVA-NeXT
+
+## Overview
+
+The LLaVA-NeXT model was proposed in [LLaVA-NeXT: Improved reasoning, OCR, and world knowledge](https://llava-vl.github.io/blog/2024-01-30-llava-next/) by Haotian Liu, Chunyuan Li, Yuheng Li, Bo Li, Yuanhan Zhang, Sheng Shen, Yong Jae Lee. LLaVa-NeXT (also called LLaVa-1.6) improves upon [LLaVa](llava) by increasing the input image resolution and training on an improved visual instruction tuning dataset to improve OCR and common sense reasoning.
+
+The introduction from the blog is the following:
+
+*In October 2023, we released LLaVA-1.5 with a simple and efficient design along with great performance on a benchmark suite of 12 datasets. It has since served as the foundation of many comprehensive studies of data, model, and capabilities of large multimodal models (LMM), and has enabled various new applications.
+
+Today, we are thrilled to present LLaVA-NeXT, with improved reasoning, OCR, and world knowledge. LLaVA-NeXT even exceeds Gemini Pro on several benchmarks.
+
+Compared with LLaVA-1.5, LLaVA-NeXT has several improvements:
+
+Increasing the input image resolution to 4x more pixels. This allows it to grasp more visual details. It supports three aspect ratios, up to 672x672, 336x1344, 1344x336 resolution.
+Better visual reasoning and OCR capability with an improved visual instruction tuning data mixture.
+Better visual conversation for more scenarios, covering different applications. Better world knowledge and logical reasoning.
+Efficient deployment and inference with SGLang.
+Along with performance improvements, LLaVA-NeXT maintains the minimalist design and data efficiency of LLaVA-1.5. It re-uses the pretrained connector of LLaVA-1.5, and still uses less than 1M visual instruction tuning samples. The largest 34B variant finishes training in ~1 day with 32 A100s.*
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/llava_next_overview.png"
+alt="drawing" width="600"/>
+
+<small> LLaVa-NeXT incorporates a higher input resolution by encoding various patches of the input image. Taken from the <a href="https://arxiv.org/abs/2310.03744">original paper.</a> </small>
+
+This model was contributed by [nielsr](https://huggingface.co/nielsr).
+The original code can be found [here](https://github.com/haotian-liu/LLaVA/tree/main).
+
+## Usage tips
+
+- We advise users to use `padding_side="left"` when computing batched generation as it leads to more accurate results. Simply make sure to call `processor.tokenizer.padding_side = "left"` before generating.
+
+- Note that each checkpoint has been trained with a specific prompt format, depending on which large language model (LLM) was used. Below, we list the correct prompt formats to use for the text prompt "What is shown in this image?":
+
+[llava-v1.6-mistral-7b-hf](https://huggingface.co/llava-hf/llava-v1.6-mistral-7b-hf) requires the following format:
+
+```bash
+"[INST] <image>\nWhat is shown in this image? [/INST]"
+```
+
+[llava-v1.6-vicuna-7b-hf](https://huggingface.co/llava-hf/llava-v1.6-vicuna-7b-hf) and [llava-v1.6-vicuna-13b-hf](https://huggingface.co/llava-hf/llava-v1.6-vicuna-13b-hf) require the following format:
+
+```bash
+"A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions. USER: <image>\nWhat is shown in this image? ASSISTANT:"
+```
+
+[llava-v1.6-34b-hf](https://huggingface.co/llava-hf/llava-v1.6-34b-hf) requires the following format:
+
+```bash
+"<|im_start|>system\nAnswer the questions.<|im_end|><|im_start|>user\n<image>\nWhat is shown in this image?<|im_end|><|im_start|>assistant\n"
+```
+
+## Usage example
+
+Here's how to load the model and perform inference in half-precision (`torch.float16`):
+
+```python
+from transformers import LlavaNextProcessor, LlavaNextForConditionalGeneration
+import torch
+from PIL import Image
+import requests
+
+processor = LlavaNextProcessor.from_pretrained("llava-hf/llava-v1.6-mistral-7b-hf")
+
+model = LlavaNextForConditionalGeneration.from_pretrained("llava-hf/llava-v1.6-mistral-7b-hf", torch_dtype=torch.float16, low_cpu_mem_usage=True) 
+model.to("cuda:0")
+
+# prepare image and text prompt, using the appropriate prompt template
+url = "https://github.com/haotian-liu/LLaVA/blob/1a91fc274d7c35a9b50b3cb29c4247ae5837ce39/images/llava_v1_5_radar.jpg?raw=true"
+image = Image.open(requests.get(url, stream=True).raw)
+prompt = "[INST] <image>\nWhat is shown in this image? [/INST]"
+
+inputs = processor(prompt, image, return_tensors="pt").to("cuda:0")
+
+# autoregressively complete prompt
+output = model.generate(**inputs, max_new_tokens=100)
+
+print(processor.decode(output[0], skip_special_tokens=True))
+```
+
+## Model optimization
+
+### Quantization using Bitsandbytes
+
+The model can be loaded in 8 or 4 bits, greatly reducing the memory requirements while maintaining the performance of the original model. First make sure to install bitsandbytes, `pip install bitsandbytes`` and make sure to have access to a CUDA compatible GPU device. Simply change the snippet above with:
+
+```python
+from transformers import LlavaNextForConditionalGeneration, BitsandBytesConfig
+
+# specify how to quantize the model
+quantization_config = BitsAndBytesConfig(
+    load_in_4bit=True,
+    bnb_4bit_quant_type="nf4",
+    bnb_4bit_compute_dtype="torch.float16",
+)
+
+model = LlavaNextForConditionalGeneration.from_pretrained("llava-hf/llava-v1.6-mistral-7b-hf", quantization_config=quantization_config, device_map="auto")
+```
+
+### Use Flash-Attention 2 to further speed-up generation
+
+First make sure to install flash-attn. Refer to the [original repository of Flash Attention](https://github.com/Dao-AILab/flash-attention) regarding that package installation. Simply change the snippet above with:
+
+```python
+from transformers import LlavaNextForConditionalGeneration
+
+model = LlavaNextForConditionalGeneration.from_pretrained(
+    model_id, 
+    torch_dtype=torch.float16, 
+    low_cpu_mem_usage=True,
+    use_flash_attention_2=True
+).to(0)
+```
+
+## LlavaNextConfig
+
+[[autodoc]] LlavaNextConfig
+
+## LlavaNextImageProcessor
+
+[[autodoc]] LlavaNextImageProcessor
+    - preprocess
+
+## LlavaNextProcessor
+
+[[autodoc]] LlavaNextProcessor
+
+## LlavaNextForConditionalGeneration
+
+[[autodoc]] LlavaNextForConditionalGeneration
+    - forward

docs/source/en/model_doc/superpoint.md

@@ -0,0 +1,120 @@
+<!--Copyright 2024 The HuggingFace Team. All rights reserved.
+
+Licensed under the MIT License; you may not use this file except in compliance with
+the License.
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+
+⚠️ Note that this file is in Markdown but contain specific syntax for our doc-builder (similar to MDX) that may not be
+rendered properly in your Markdown viewer.
+
+
+-->
+
+# SuperPoint
+
+## Overview
+
+The SuperPoint model was proposed
+in [SuperPoint: Self-Supervised Interest Point Detection and Description](https://arxiv.org/abs/1712.07629) by Daniel
+DeTone, Tomasz Malisiewicz and Andrew Rabinovich.
+
+This model is the result of a self-supervised training of a fully-convolutional network for interest point detection and
+description. The model is able to detect interest points that are repeatable under homographic transformations and
+provide a descriptor for each point. The use of the model in its own is limited, but it can be used as a feature
+extractor for other tasks such as homography estimation, image matching, etc.
+
+The abstract from the paper is the following:
+
+*This paper presents a self-supervised framework for training interest point detectors and descriptors suitable for a
+large number of multiple-view geometry problems in computer vision. As opposed to patch-based neural networks, our
+fully-convolutional model operates on full-sized images and jointly computes pixel-level interest point locations and
+associated descriptors in one forward pass. We introduce Homographic Adaptation, a multi-scale, multi-homography
+approach for boosting interest point detection repeatability and performing cross-domain adaptation (e.g.,
+synthetic-to-real). Our model, when trained on the MS-COCO generic image dataset using Homographic Adaptation, is able
+to repeatedly detect a much richer set of interest points than the initial pre-adapted deep model and any other
+traditional corner detector. The final system gives rise to state-of-the-art homography estimation results on HPatches
+when compared to LIFT, SIFT and ORB.*
+
+## How to use
+
+Here is a quick example of using the model to detect interest points in an image:
+
+```python
+from transformers import AutoImageProcessor, AutoModel
+import torch
+from PIL import Image
+import requests
+
+url = "http://images.cocodataset.org/val2017/000000039769.jpg"
+image = Image.open(requests.get(url, stream=True).raw)
+
+processor = AutoImageProcessor.from_pretrained("magic-leap-community/superpoint")
+model = AutoModel.from_pretrained("magic-leap-community/superpoint")
+
+inputs = processor(image, return_tensors="pt")
+outputs = model(**inputs)
+```
+
+The outputs contain the list of keypoint coordinates with their respective score and description (a 256-long vector).
+
+You can also feed multiple images to the model. Due to the nature of SuperPoint, to output a dynamic number of keypoints,
+you will need to use the mask attribute to retrieve the respective information :
+
+```python
+from transformers import AutoImageProcessor, AutoModel
+import torch
+from PIL import Image
+import requests
+
+url_image_1 = "http://images.cocodataset.org/val2017/000000039769.jpg"
+image_1 = Image.open(requests.get(url_image_1, stream=True).raw)
+url_image_2 = "http://images.cocodataset.org/test-stuff2017/000000000568.jpg"
+image_2 = Image.open(requests.get(url_image_2, stream=True).raw)
+
+images = [image_1, image_2]
+
+processor = AutoImageProcessor.from_pretrained("magic-leap-community/superpoint")
+model = AutoModel.from_pretrained("magic-leap-community/superpoint")
+
+inputs = processor(images, return_tensors="pt")
+outputs = model(**inputs)
+
+for i in range(len(images)):
+    image_mask = outputs.mask[i]
+    image_indices = torch.nonzero(image_mask).squeeze()
+    image_keypoints = outputs.keypoints[i][image_indices]
+    image_scores = outputs.scores[i][image_indices]
+    image_descriptors = outputs.descriptors[i][image_indices]
+```
+
+You can then print the keypoints on the image to visualize the result :
+```python
+import cv2
+for keypoint, score in zip(image_keypoints, image_scores):
+    keypoint_x, keypoint_y = int(keypoint[0].item()), int(keypoint[1].item())
+    color = tuple([score.item() * 255] * 3)
+    image = cv2.circle(image, (keypoint_x, keypoint_y), 2, color)
+cv2.imwrite("output_image.png", image)
+```
+
+This model was contributed by [stevenbucaille](https://huggingface.co/stevenbucaille).
+The original code can be found [here](https://github.com/magicleap/SuperPointPretrainedNetwork).
+
+## SuperPointConfig
+
+[[autodoc]] SuperPointConfig
+
+## SuperPointImageProcessor
+
+[[autodoc]] SuperPointImageProcessor
+
+- preprocess
+
+## SuperPointForKeypointDetection
+
+[[autodoc]] SuperPointForKeypointDetection
+
+- forward

docs/source/en/perf_infer_gpu_one.md

@@ -49,6 +49,7 @@ FlashAttention-2 is currently supported for the following architectures:
 * [Falcon](https://huggingface.co/docs/transformers/model_doc/falcon#transformers.FalconModel)
 * [Llama](https://huggingface.co/docs/transformers/model_doc/llama#transformers.LlamaModel)
 * [Llava](https://huggingface.co/docs/transformers/model_doc/llava)
+* [Llava-NeXT](https://huggingface.co/docs/transformers/model_doc/llava_next)
 * [VipLlava](https://huggingface.co/docs/transformers/model_doc/vipllava)
 * [MBart](https://huggingface.co/docs/transformers/model_doc/mbart#transformers.MBartModel)
 * [Mistral](https://huggingface.co/docs/transformers/model_doc/mistral#transformers.MistralModel)

examples/flax/question-answering/run_qa.py

@@ -62,7 +62,7 @@ from transformers.utils import check_min_version, send_example_telemetry
 logger = logging.getLogger(__name__)
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.39.0.dev0")
+check_min_version("4.39.0")
 
 Array = Any
 Dataset = datasets.arrow_dataset.Dataset

examples/flax/speech-recognition/run_flax_speech_recognition_seq2seq.py

@@ -60,7 +60,7 @@ from transformers.utils.versions import require_version
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risk.
-check_min_version("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=2.14.0", "To fix: pip install -r examples/flax/speech-recognition/requirements.txt")
 

examples/flax/text-classification/run_flax_glue.py

@@ -55,7 +55,7 @@ from transformers.utils import check_min_version, send_example_telemetry
 
 logger = logging.getLogger(__name__)
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.39.0.dev0")
+check_min_version("4.39.0")
 
 Array = Any
 Dataset = datasets.arrow_dataset.Dataset

examples/flax/token-classification/run_flax_ner.py

@@ -56,7 +56,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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

@@ -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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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

@@ -57,7 +57,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=2.14.0", "To fix: pip install -r examples/pytorch/image-classification/requirements.txt")
 

examples/pytorch/image-classification/run_image_classification_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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 logger = get_logger(__name__)
 

examples/pytorch/image-pretraining/run_mae.py

@@ -44,7 +44,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/image-pretraining/requirements.txt")
 

examples/pytorch/image-pretraining/run_mim.py

@@ -49,7 +49,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/image-pretraining/requirements.txt")
 

examples/pytorch/image-pretraining/run_mim_no_trainer.py

@@ -54,7 +54,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/image-pretraining/requirements.txt")
 

examples/pytorch/language-modeling/run_clm.py

@@ -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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 logger = get_logger(__name__)
 

examples/pytorch/language-modeling/run_fim.py

@@ -58,7 +58,7 @@ 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("4.36.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/language-modeling/requirements.txt")
 

examples/pytorch/language-modeling/run_fim_no_trainer.py

@@ -60,7 +60,7 @@ 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("4.36.0.dev0")
+check_min_version("4.39.0")
 
 logger = get_logger(__name__)
 

examples/pytorch/language-modeling/run_mlm.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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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

@@ -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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/language-modeling/requirements.txt")
 

examples/pytorch/multiple-choice/run_swag.py

@@ -48,7 +48,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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, send_example_
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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

@@ -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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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

@@ -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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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

@@ -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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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

@@ -52,7 +52,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 logger = get_logger(__name__)
 

examples/pytorch/speech-recognition/run_speech_recognition_ctc.py

@@ -51,7 +51,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.18.0", "To fix: pip install -r examples/pytorch/speech-recognition/requirements.txt")
 

examples/pytorch/speech-recognition/run_speech_recognition_ctc_adapter.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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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

@@ -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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.18.0", "To fix: pip install -r examples/pytorch/speech-recognition/requirements.txt")
 

examples/pytorch/summarization/run_summarization.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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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_classification.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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/text-classification/requirements.txt")
 

examples/pytorch/text-classification/run_glue.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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 logger = get_logger(__name__)
 

examples/pytorch/text-classification/run_xnli.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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/text-classification/requirements.txt")
 

examples/pytorch/token-classification/run_ner.py

@@ -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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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

@@ -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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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

@@ -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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 logger = get_logger(__name__)
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/translation/requirements.txt")

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

@@ -52,7 +52,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version(
     "datasets>=1.8.0", "To fix: pip install -r examples/tensorflow/contrastive-image-text/requirements.txt"

examples/tensorflow/image-classification/run_image_classification.py

@@ -56,7 +56,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/image-classification/requirements.txt")
 

examples/tensorflow/multiple-choice/run_swag.py

@@ -51,7 +51,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 logger = logging.getLogger(__name__)
 

examples/tensorflow/question-answering/run_qa.py

@@ -63,7 +63,7 @@ except (ModuleNotFoundError, ImportError):
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
-check_min_version("4.39.0.dev0")
+check_min_version("4.39.0")
 
 logger = logging.getLogger(__name__)
 

examples/tensorflow/summarization/run_summarization.py

@@ -54,7 +54,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/summarization/requirements.txt")
 

examples/tensorflow/text-classification/run_glue.py

@@ -48,7 +48,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 task_to_keys = {
     "cola": ("sentence", None),

examples/tensorflow/translation/run_translation.py

@@ -57,7 +57,7 @@ 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("4.39.0.dev0")
+check_min_version("4.39.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/summarization/requirements.txt")
 

setup.py

@@ -428,7 +428,7 @@ install_requires = [
 
 setup(
     name="transformers",
-    version="4.39.0.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="4.39.1",  # 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

@@ -18,7 +18,7 @@
 # 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__ = "4.39.0.dev0"
+__version__ = "4.39.1"
 
 from typing import TYPE_CHECKING
 
@@ -555,6 +555,12 @@ _import_structure = {
     "models.llava": [
         "LLAVA_PRETRAINED_CONFIG_ARCHIVE_MAP",
         "LlavaConfig",
+        "LlavaProcessor",
+    ],
+    "models.llava_next": [
+        "LLAVA_NEXT_PRETRAINED_CONFIG_ARCHIVE_MAP",
+        "LlavaNextConfig",
+        "LlavaNextProcessor",
     ],
     "models.longformer": [
         "LONGFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP",
@@ -819,6 +825,7 @@ _import_structure = {
     ],
     "models.stablelm": ["STABLELM_PRETRAINED_CONFIG_ARCHIVE_MAP", "StableLmConfig"],
     "models.starcoder2": ["STARCODER2_PRETRAINED_CONFIG_ARCHIVE_MAP", "Starcoder2Config"],
+    "models.superpoint": ["SUPERPOINT_PRETRAINED_CONFIG_ARCHIVE_MAP", "SuperPointConfig"],
     "models.swiftformer": [
         "SWIFTFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP",
         "SwiftFormerConfig",
@@ -1322,6 +1329,7 @@ else:
     _import_structure["models.layoutlmv2"].extend(["LayoutLMv2FeatureExtractor", "LayoutLMv2ImageProcessor"])
     _import_structure["models.layoutlmv3"].extend(["LayoutLMv3FeatureExtractor", "LayoutLMv3ImageProcessor"])
     _import_structure["models.levit"].extend(["LevitFeatureExtractor", "LevitImageProcessor"])
+    _import_structure["models.llava_next"].append("LlavaNextImageProcessor")
     _import_structure["models.mask2former"].append("Mask2FormerImageProcessor")
     _import_structure["models.maskformer"].extend(["MaskFormerFeatureExtractor", "MaskFormerImageProcessor"])
     _import_structure["models.mobilenet_v1"].extend(["MobileNetV1FeatureExtractor", "MobileNetV1ImageProcessor"])
@@ -1339,6 +1347,7 @@ else:
     _import_structure["models.segformer"].extend(["SegformerFeatureExtractor", "SegformerImageProcessor"])
     _import_structure["models.seggpt"].extend(["SegGptImageProcessor"])
     _import_structure["models.siglip"].append("SiglipImageProcessor")
+    _import_structure["models.superpoint"].extend(["SuperPointImageProcessor"])
     _import_structure["models.swin2sr"].append("Swin2SRImageProcessor")
     _import_structure["models.tvlt"].append("TvltImageProcessor")
     _import_structure["models.tvp"].append("TvpImageProcessor")
@@ -1485,6 +1494,7 @@ else:
             "MODEL_FOR_IMAGE_SEGMENTATION_MAPPING",
             "MODEL_FOR_IMAGE_TO_IMAGE_MAPPING",
             "MODEL_FOR_INSTANCE_SEGMENTATION_MAPPING",
+            "MODEL_FOR_KEYPOINT_DETECTION_MAPPING",
             "MODEL_FOR_MASKED_IMAGE_MODELING_MAPPING",
             "MODEL_FOR_MASKED_LM_MAPPING",
             "MODEL_FOR_MASK_GENERATION_MAPPING",
@@ -1525,6 +1535,7 @@ else:
             "AutoModelForImageSegmentation",
             "AutoModelForImageToImage",
             "AutoModelForInstanceSegmentation",
+            "AutoModelForKeypointDetection",
             "AutoModelForMaskedImageModeling",
             "AutoModelForMaskedLM",
             "AutoModelForMaskGeneration",
@@ -2531,7 +2542,13 @@ else:
             "LLAVA_PRETRAINED_MODEL_ARCHIVE_LIST",
             "LlavaForConditionalGeneration",
             "LlavaPreTrainedModel",
-            "LlavaProcessor",
+        ]
+    )
+    _import_structure["models.llava_next"].extend(
+        [
+            "LLAVA_NEXT_PRETRAINED_MODEL_ARCHIVE_LIST",
+            "LlavaNextForConditionalGeneration",
+            "LlavaNextPreTrainedModel",
         ]
     )
     _import_structure["models.longformer"].extend(
@@ -3336,6 +3353,13 @@ else:
             "Starcoder2PreTrainedModel",
         ]
     )
+    _import_structure["models.superpoint"].extend(
+        [
+            "SUPERPOINT_PRETRAINED_MODEL_ARCHIVE_LIST",
+            "SuperPointForKeypointDetection",
+            "SuperPointPreTrainedModel",
+        ]
+    )
     _import_structure["models.swiftformer"].extend(
         [
             "SWIFTFORMER_PRETRAINED_MODEL_ARCHIVE_LIST",
@@ -5405,6 +5429,12 @@ if TYPE_CHECKING:
     from .models.llava import (
         LLAVA_PRETRAINED_CONFIG_ARCHIVE_MAP,
         LlavaConfig,
+        LlavaProcessor,
+    )
+    from .models.llava_next import (
+        LLAVA_NEXT_PRETRAINED_CONFIG_ARCHIVE_MAP,
+        LlavaNextConfig,
+        LlavaNextProcessor,
     )
     from .models.longformer import (
         LONGFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP,
@@ -5670,6 +5700,7 @@ if TYPE_CHECKING:
     )
     from .models.stablelm import STABLELM_PRETRAINED_CONFIG_ARCHIVE_MAP, StableLmConfig
     from .models.starcoder2 import STARCODER2_PRETRAINED_CONFIG_ARCHIVE_MAP, Starcoder2Config
+    from .models.superpoint import SUPERPOINT_PRETRAINED_CONFIG_ARCHIVE_MAP, SuperPointConfig
     from .models.swiftformer import (
         SWIFTFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP,
         SwiftFormerConfig,
@@ -6153,6 +6184,7 @@ if TYPE_CHECKING:
             LayoutLMv3ImageProcessor,
         )
         from .models.levit import LevitFeatureExtractor, LevitImageProcessor
+        from .models.llava_next import LlavaNextImageProcessor
         from .models.mask2former import Mask2FormerImageProcessor
         from .models.maskformer import (
             MaskFormerFeatureExtractor,
@@ -6182,6 +6214,7 @@ if TYPE_CHECKING:
         from .models.segformer import SegformerFeatureExtractor, SegformerImageProcessor
         from .models.seggpt import SegGptImageProcessor
         from .models.siglip import SiglipImageProcessor
+        from .models.superpoint import SuperPointImageProcessor
         from .models.swin2sr import Swin2SRImageProcessor
         from .models.tvlt import TvltImageProcessor
         from .models.tvp import TvpImageProcessor
@@ -6308,6 +6341,7 @@ if TYPE_CHECKING:
             MODEL_FOR_IMAGE_SEGMENTATION_MAPPING,
             MODEL_FOR_IMAGE_TO_IMAGE_MAPPING,
             MODEL_FOR_INSTANCE_SEGMENTATION_MAPPING,
+            MODEL_FOR_KEYPOINT_DETECTION_MAPPING,
             MODEL_FOR_MASK_GENERATION_MAPPING,
             MODEL_FOR_MASKED_IMAGE_MODELING_MAPPING,
             MODEL_FOR_MASKED_LM_MAPPING,
@@ -6348,6 +6382,7 @@ if TYPE_CHECKING:
             AutoModelForImageSegmentation,
             AutoModelForImageToImage,
             AutoModelForInstanceSegmentation,
+            AutoModelForKeypointDetection,
             AutoModelForMaskedImageModeling,
             AutoModelForMaskedLM,
             AutoModelForMaskGeneration,
@@ -7179,7 +7214,11 @@ if TYPE_CHECKING:
             LLAVA_PRETRAINED_MODEL_ARCHIVE_LIST,
             LlavaForConditionalGeneration,
             LlavaPreTrainedModel,
-            LlavaProcessor,
+        )
+        from .models.llava_next import (
+            LLAVA_NEXT_PRETRAINED_MODEL_ARCHIVE_LIST,
+            LlavaNextForConditionalGeneration,
+            LlavaNextPreTrainedModel,
         )
         from .models.longformer import (
             LONGFORMER_PRETRAINED_MODEL_ARCHIVE_LIST,
@@ -7839,6 +7878,11 @@ if TYPE_CHECKING:
             Starcoder2Model,
             Starcoder2PreTrainedModel,
         )
+        from .models.superpoint import (
+            SUPERPOINT_PRETRAINED_MODEL_ARCHIVE_LIST,
+            SuperPointForKeypointDetection,
+            SuperPointPreTrainedModel,
+        )
         from .models.swiftformer import (
             SWIFTFORMER_PRETRAINED_MODEL_ARCHIVE_LIST,
             SwiftFormerForImageClassification,

src/transformers/image_processing_utils.py

@@ -748,6 +748,44 @@ def get_size_dict(
     return size_dict
 
 
+def select_best_resolution(original_size: tuple, possible_resolutions: list) -> tuple:
+    """
+    Selects the best resolution from a list of possible resolutions based on the original size.
+
+    This is done by calculating the effective and wasted resolution for each possible resolution.
+
+    The best fit resolution is the one that maximizes the effective resolution and minimizes the wasted resolution.
+
+    Args:
+        original_size (tuple):
+            The original size of the image in the format (height, width).
+        possible_resolutions (list):
+            A list of possible resolutions in the format [(height1, width1), (height2, width2), ...].
+
+    Returns:
+        tuple: The best fit resolution in the format (height, width).
+    """
+    original_height, original_width = original_size
+    best_fit = None
+    max_effective_resolution = 0
+    min_wasted_resolution = float("inf")
+
+    for height, width in possible_resolutions:
+        scale = min(width / original_width, height / original_height)
+        downscaled_width, downscaled_height = int(original_width * scale), int(original_height * scale)
+        effective_resolution = min(downscaled_width * downscaled_height, original_width * original_height)
+        wasted_resolution = (width * height) - effective_resolution
+
+        if effective_resolution > max_effective_resolution or (
+            effective_resolution == max_effective_resolution and wasted_resolution < min_wasted_resolution
+        ):
+            max_effective_resolution = effective_resolution
+            min_wasted_resolution = wasted_resolution
+            best_fit = (height, width)
+
+    return best_fit
+
+
 ImageProcessingMixin.push_to_hub = copy_func(ImageProcessingMixin.push_to_hub)
 if ImageProcessingMixin.push_to_hub.__doc__ is not None:
     ImageProcessingMixin.push_to_hub.__doc__ = ImageProcessingMixin.push_to_hub.__doc__.format(

src/transformers/models/__init__.py

@@ -124,6 +124,7 @@ from . import (
     lilt,
     llama,
     llava,
+    llava_next,
     longformer,
     longt5,
     luke,
@@ -210,6 +211,7 @@ from . import (
     squeezebert,
     stablelm,
     starcoder2,
+    superpoint,
     swiftformer,
     swin,
     swin2sr,

src/transformers/models/auto/__init__.py

@@ -52,6 +52,7 @@ else:
         "MODEL_FOR_IMAGE_MAPPING",
         "MODEL_FOR_IMAGE_SEGMENTATION_MAPPING",
         "MODEL_FOR_IMAGE_TO_IMAGE_MAPPING",
+        "MODEL_FOR_KEYPOINT_DETECTION_MAPPING",
         "MODEL_FOR_INSTANCE_SEGMENTATION_MAPPING",
         "MODEL_FOR_MASKED_IMAGE_MODELING_MAPPING",
         "MODEL_FOR_MASKED_LM_MAPPING",
@@ -92,6 +93,7 @@ else:
         "AutoModelForImageSegmentation",
         "AutoModelForImageToImage",
         "AutoModelForInstanceSegmentation",
+        "AutoModelForKeypointDetection",
         "AutoModelForMaskGeneration",
         "AutoModelForTextEncoding",
         "AutoModelForMaskedImageModeling",
@@ -238,6 +240,7 @@ if TYPE_CHECKING:
             MODEL_FOR_IMAGE_SEGMENTATION_MAPPING,
             MODEL_FOR_IMAGE_TO_IMAGE_MAPPING,
             MODEL_FOR_INSTANCE_SEGMENTATION_MAPPING,
+            MODEL_FOR_KEYPOINT_DETECTION_MAPPING,
             MODEL_FOR_MASK_GENERATION_MAPPING,
             MODEL_FOR_MASKED_IMAGE_MODELING_MAPPING,
             MODEL_FOR_MASKED_LM_MAPPING,
@@ -278,6 +281,7 @@ if TYPE_CHECKING:
             AutoModelForImageSegmentation,
             AutoModelForImageToImage,
             AutoModelForInstanceSegmentation,
+            AutoModelForKeypointDetection,
             AutoModelForMaskedImageModeling,
             AutoModelForMaskedLM,
             AutoModelForMaskGeneration,

src/transformers/models/auto/configuration_auto.py

@@ -133,6 +133,7 @@ CONFIG_MAPPING_NAMES = OrderedDict(
         ("lilt", "LiltConfig"),
         ("llama", "LlamaConfig"),
         ("llava", "LlavaConfig"),
+        ("llava_next", "LlavaNextConfig"),
         ("longformer", "LongformerConfig"),
         ("longt5", "LongT5Config"),
         ("luke", "LukeConfig"),
@@ -219,6 +220,7 @@ CONFIG_MAPPING_NAMES = OrderedDict(
         ("squeezebert", "SqueezeBertConfig"),
         ("stablelm", "StableLmConfig"),
         ("starcoder2", "Starcoder2Config"),
+        ("superpoint", "SuperPointConfig"),
         ("swiftformer", "SwiftFormerConfig"),
         ("swin", "SwinConfig"),
         ("swin2sr", "Swin2SRConfig"),
@@ -373,6 +375,7 @@ CONFIG_ARCHIVE_MAP_MAPPING_NAMES = OrderedDict(
         ("lilt", "LILT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("llama", "LLAMA_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("llava", "LLAVA_PRETRAINED_CONFIG_ARCHIVE_MAP"),
+        ("llava_next", "LLAVA_NEXT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("longformer", "LONGFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("longt5", "LONGT5_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("luke", "LUKE_PRETRAINED_CONFIG_ARCHIVE_MAP"),
@@ -450,6 +453,7 @@ CONFIG_ARCHIVE_MAP_MAPPING_NAMES = OrderedDict(
         ("squeezebert", "SQUEEZEBERT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("stablelm", "STABLELM_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("starcoder2", "STARCODER2_PRETRAINED_CONFIG_ARCHIVE_MAP"),
+        ("superpoint", "SUPERPOINT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("swiftformer", "SWIFTFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("swin", "SWIN_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("swin2sr", "SWIN2SR_PRETRAINED_CONFIG_ARCHIVE_MAP"),
@@ -617,6 +621,7 @@ MODEL_NAMES_MAPPING = OrderedDict(
         ("llama", "LLaMA"),
         ("llama2", "Llama2"),
         ("llava", "LLaVa"),
+        ("llava_next", "LLaVA-NeXT"),
         ("longformer", "Longformer"),
         ("longt5", "LongT5"),
         ("luke", "LUKE"),
@@ -711,6 +716,7 @@ MODEL_NAMES_MAPPING = OrderedDict(
         ("squeezebert", "SqueezeBERT"),
         ("stablelm", "StableLm"),
         ("starcoder2", "Starcoder2"),
+        ("superpoint", "SuperPoint"),
         ("swiftformer", "SwiftFormer"),
         ("swin", "Swin Transformer"),
         ("swin2sr", "Swin2SR"),

src/transformers/models/auto/image_processing_auto.py

@@ -77,6 +77,7 @@ IMAGE_PROCESSOR_MAPPING_NAMES = OrderedDict(
         ("layoutlmv3", "LayoutLMv3ImageProcessor"),
         ("levit", "LevitImageProcessor"),
         ("llava", "CLIPImageProcessor"),
+        ("llava_next", "LlavaNextImageProcessor"),
         ("mask2former", "Mask2FormerImageProcessor"),
         ("maskformer", "MaskFormerImageProcessor"),
         ("mgp-str", "ViTImageProcessor"),

src/transformers/models/auto/modeling_auto.py

@@ -286,6 +286,7 @@ MODEL_FOR_PRETRAINING_MAPPING_NAMES = OrderedDict(
         ("idefics", "IdeficsForVisionText2Text"),
         ("layoutlm", "LayoutLMForMaskedLM"),
         ("llava", "LlavaForConditionalGeneration"),
+        ("llava_next", "LlavaNextForConditionalGeneration"),
         ("longformer", "LongformerForMaskedLM"),
         ("luke", "LukeForMaskedLM"),
         ("lxmert", "LxmertForPreTraining"),
@@ -675,6 +676,7 @@ MODEL_FOR_VISION_2_SEQ_MAPPING_NAMES = OrderedDict(
         ("instructblip", "InstructBlipForConditionalGeneration"),
         ("kosmos-2", "Kosmos2ForConditionalGeneration"),
         ("llava", "LlavaForConditionalGeneration"),
+        ("llava_next", "LlavaNextForConditionalGeneration"),
         ("pix2struct", "Pix2StructForConditionalGeneration"),
         ("vipllava", "VipLlavaForConditionalGeneration"),
         ("vision-encoder-decoder", "VisionEncoderDecoderModel"),
@@ -1224,6 +1226,14 @@ MODEL_FOR_MASK_GENERATION_MAPPING_NAMES = OrderedDict(
     ]
 )
 
+
+MODEL_FOR_KEYPOINT_DETECTION_MAPPING_NAMES = OrderedDict(
+    [
+        ("superpoint", "SuperPointForKeypointDetection"),
+    ]
+)
+
+
 MODEL_FOR_TEXT_ENCODING_MAPPING_NAMES = OrderedDict(
     [
         ("albert", "AlbertModel"),
@@ -1359,6 +1369,10 @@ MODEL_FOR_BACKBONE_MAPPING = _LazyAutoMapping(CONFIG_MAPPING_NAMES, MODEL_FOR_BA
 
 MODEL_FOR_MASK_GENERATION_MAPPING = _LazyAutoMapping(CONFIG_MAPPING_NAMES, MODEL_FOR_MASK_GENERATION_MAPPING_NAMES)
 
+MODEL_FOR_KEYPOINT_DETECTION_MAPPING = _LazyAutoMapping(
+    CONFIG_MAPPING_NAMES, MODEL_FOR_KEYPOINT_DETECTION_MAPPING_NAMES
+)
+
 MODEL_FOR_TEXT_ENCODING_MAPPING = _LazyAutoMapping(CONFIG_MAPPING_NAMES, MODEL_FOR_TEXT_ENCODING_MAPPING_NAMES)
 
 MODEL_FOR_TIME_SERIES_CLASSIFICATION_MAPPING = _LazyAutoMapping(
@@ -1376,6 +1390,10 @@ class AutoModelForMaskGeneration(_BaseAutoModelClass):
     _model_mapping = MODEL_FOR_MASK_GENERATION_MAPPING
 
 
+class AutoModelForKeypointDetection(_BaseAutoModelClass):
+    _model_mapping = MODEL_FOR_KEYPOINT_DETECTION_MAPPING
+
+
 class AutoModelForTextEncoding(_BaseAutoModelClass):
     _model_mapping = MODEL_FOR_TEXT_ENCODING_MAPPING
 

src/transformers/models/auto/processing_auto.py

@@ -66,6 +66,7 @@ PROCESSOR_MAPPING_NAMES = OrderedDict(
         ("layoutlmv2", "LayoutLMv2Processor"),
         ("layoutlmv3", "LayoutLMv3Processor"),
         ("llava", "LlavaProcessor"),
+        ("llava_next", "LlavaNextProcessor"),
         ("markuplm", "MarkupLMProcessor"),
         ("mctct", "MCTCTProcessor"),
         ("mgp-str", "MgpstrProcessor"),

src/transformers/models/auto/tokenization_auto.py

@@ -223,6 +223,7 @@ else:
                 ),
             ),
             ("llava", ("LlamaTokenizer", "LlamaTokenizerFast" if is_tokenizers_available() else None)),
+            ("llava_next", ("LlamaTokenizer", "LlamaTokenizerFast" if is_tokenizers_available() else None)),
             ("longformer", ("LongformerTokenizer", "LongformerTokenizerFast" if is_tokenizers_available() else None)),
             (
                 "longt5",

src/transformers/models/cohere/modeling_cohere.py

@@ -274,9 +274,7 @@ class CohereAttention(nn.Module):
         attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) / math.sqrt(self.head_dim)
 
         if attention_mask is not None:  # no matter the length, we just slice it
-            causal_mask = attention_mask
-            if cache_position is not None:
-                causal_mask = attention_mask[:, :, cache_position, : key_states.shape[-2]]
+            causal_mask = attention_mask[:, :, :, : key_states.shape[-2]]
             attn_weights = attn_weights + causal_mask
 
         # upcast attention to fp32
@@ -559,8 +557,9 @@ class CohereSdpaAttention(CohereAttention):
         value_states = repeat_kv(value_states, self.num_key_value_groups)
 
         causal_mask = attention_mask
-        if attention_mask is not None and cache_position is not None:
-            causal_mask = causal_mask[:, :, cache_position, : key_states.shape[-2]]
+        # if attention_mask is not None and cache_position is not None:
+        if attention_mask is not None:
+            causal_mask = causal_mask[:, :, :, : key_states.shape[-2]]
 
         # SDPA with memory-efficient backend is currently (torch==2.1.2) bugged with non-contiguous inputs with custom attn_mask,
         # Reference: https://github.com/pytorch/pytorch/issues/112577.
@@ -692,7 +691,7 @@ class CoherePreTrainedModel(PreTrainedModel):
     base_model_prefix = "model"
     supports_gradient_checkpointing = True
     _no_split_modules = ["CohereDecoderLayer"]
-    _skip_keys_device_placement = ["past_key_values", "causal_mask"]
+    _skip_keys_device_placement = ["past_key_values"]
     _supports_flash_attn_2 = True
     _supports_sdpa = True
     _supports_cache_class = True
@@ -715,12 +714,6 @@ class CoherePreTrainedModel(PreTrainedModel):
                 "make sure to use `sdpa` in the mean time, and open an issue at https://github.com/huggingface/transformers"
             )
 
-        if max_cache_len > self.model.causal_mask.shape[-1] or self.device != self.model.causal_mask.device:
-            causal_mask = torch.full(
-                (max_cache_len, max_cache_len), fill_value=True, device=self.device, dtype=torch.bool
-            )
-            self.register_buffer("causal_mask", torch.triu(causal_mask, diagonal=1), persistent=False)
-
         for layer in self.model.layers:
             device = layer.input_layernorm.weight.device
             if hasattr(self.config, "_pre_quantization_dtype"):
@@ -832,12 +825,6 @@ class CohereModel(CoherePreTrainedModel):
         self.norm = CohereLayerNorm(config.hidden_size, eps=config.layer_norm_eps)
         self.gradient_checkpointing = False
 
-        # Register a causal mask to separate causal and padding mask creation. Merging happens in the attention class.
-        # NOTE: This is not friendly with TorchScript, ONNX, ExportedProgram serialization for very large `max_position_embeddings`.
-        causal_mask = torch.full(
-            (config.max_position_embeddings, config.max_position_embeddings), fill_value=True, dtype=torch.bool
-        )
-        self.register_buffer("causal_mask", torch.triu(causal_mask, diagonal=1), persistent=False)
         # Initialize weights and apply final processing
         self.post_init()
 
@@ -899,7 +886,7 @@ class CohereModel(CoherePreTrainedModel):
         if position_ids is None:
             position_ids = cache_position.unsqueeze(0)
 
-        causal_mask = self._update_causal_mask(attention_mask, inputs_embeds)
+        causal_mask = self._update_causal_mask(attention_mask, inputs_embeds, cache_position)
 
         # embed positions
         hidden_states = inputs_embeds
@@ -967,25 +954,27 @@ class CohereModel(CoherePreTrainedModel):
     # KV cache is used. This is an issue for torch.compile which then recaptures cudagraphs at each decode steps due to the dynamic shapes.
     # (`recording cudagraph tree for symint key 13`, etc.), which is VERY slow. A workaround is `@torch.compiler.disable`, but this prevents using
     # `fullgraph=True`. See more context in https://github.com/huggingface/transformers/pull/29114
-    def _update_causal_mask(self, attention_mask, input_tensor):
+    def _update_causal_mask(self, attention_mask, input_tensor, cache_position):
         if self.config._attn_implementation == "flash_attention_2":
             if attention_mask is not None and 0.0 in attention_mask:
                 return attention_mask
             return None
 
-        batch_size, seq_length = input_tensor.shape[:2]
-        dtype = input_tensor.dtype
-        device = input_tensor.device
-
-        # support going beyond cached `max_position_embedding`
-        if seq_length > self.causal_mask.shape[-1]:
-            causal_mask = torch.full((2 * self.causal_mask.shape[-1], 2 * self.causal_mask.shape[-1]), fill_value=1)
-            self.register_buffer("causal_mask", torch.triu(causal_mask, diagonal=1), persistent=False)
-
-        # We use the current dtype to avoid any overflows
+        dtype, device = input_tensor.dtype, input_tensor.device
         min_dtype = torch.finfo(dtype).min
-        causal_mask = self.causal_mask[None, None, :, :].to(dtype=dtype, device=device) * min_dtype
-        causal_mask = causal_mask.expand(batch_size, 1, -1, -1)
+        sequence_length = input_tensor.shape[1]
+        if hasattr(self.layers[0].self_attn, "past_key_value"):  # static cache
+            target_length = self.config.max_position_embeddings
+        else:  # dynamic cache
+            target_length = (
+                attention_mask.shape[-1] if isinstance(attention_mask, torch.Tensor) else cache_position[-1] + 1
+            )
+
+        causal_mask = torch.full((sequence_length, target_length), fill_value=min_dtype, dtype=dtype, device=device)
+        if sequence_length != 1:
+            causal_mask = torch.triu(causal_mask, diagonal=1)
+        causal_mask *= torch.arange(target_length, device=device) > cache_position.reshape(-1, 1)
+        causal_mask = causal_mask[None, None, :, :].expand(input_tensor.shape[0], 1, -1, -1)
         if attention_mask is not None:
             causal_mask = causal_mask.clone()  # copy to contiguous memory for in-place edit
             if attention_mask.dim() == 2:
@@ -993,9 +982,17 @@ class CohereModel(CoherePreTrainedModel):
                 padding_mask = causal_mask[..., :mask_length].eq(0.0) * attention_mask[:, None, None, :].eq(0.0)
                 causal_mask[..., :mask_length] = causal_mask[..., :mask_length].masked_fill(padding_mask, min_dtype)
             elif attention_mask.dim() == 4:
+                # backwards compatibility: we allow passing a 4D attention mask shorter than the input length with
+                # cache. In that case, the 4D attention mask attends to the newest tokens only.
+                if attention_mask.shape[-2] < cache_position[0] + sequence_length:
+                    offset = cache_position[0]
+                else:
+                    offset = 0
                 mask_shape = attention_mask.shape
                 mask_slice = (attention_mask.eq(0.0)).to(dtype=dtype) * min_dtype
-                causal_mask[: mask_shape[0], : mask_shape[1], : mask_shape[2], : mask_shape[3]] = mask_slice
+                causal_mask[
+                    : mask_shape[0], : mask_shape[1], offset : mask_shape[2] + offset, : mask_shape[3]
+                ] = mask_slice
 
         if (
             self.config._attn_implementation == "sdpa"

src/transformers/models/gemma/modeling_gemma.py

@@ -279,10 +279,7 @@ class GemmaAttention(nn.Module):
         attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) / math.sqrt(self.head_dim)
 
         if attention_mask is not None:  # no matter the length, we just slice it
-            if cache_position is not None:
-                causal_mask = attention_mask[:, :, cache_position, : key_states.shape[-2]]
-            else:
-                causal_mask = attention_mask
+            causal_mask = attention_mask[:, :, :, : key_states.shape[-2]]
             attn_weights = attn_weights + causal_mask
 
         # upcast attention to fp32
@@ -563,8 +560,8 @@ class GemmaSdpaAttention(GemmaAttention):
         value_states = repeat_kv(value_states, self.num_key_value_groups)
 
         causal_mask = attention_mask
-        if attention_mask is not None and cache_position is not None:
-            causal_mask = causal_mask[:, :, cache_position, : key_states.shape[-2]]
+        if attention_mask is not None:
+            causal_mask = causal_mask[:, :, :, : key_states.shape[-2]]
 
         # SDPA with memory-efficient backend is currently (torch==2.1.2) bugged with non-contiguous inputs with custom attn_mask,
         # Reference: https://github.com/pytorch/pytorch/issues/112577.
@@ -722,10 +719,6 @@ class GemmaPreTrainedModel(PreTrainedModel):
                 "make sure to use `sdpa` in the mean time, and open an issue at https://github.com/huggingface/transformers"
             )
 
-        if max_cache_len > self.model.causal_mask.shape[-1] or self.device != self.model.causal_mask.device:
-            causal_mask = torch.full((max_cache_len, max_cache_len), fill_value=1, device=self.device)
-            self.register_buffer("causal_mask", torch.triu(causal_mask, diagonal=1), persistent=False)
-
         for layer in self.model.layers:
             weights = layer.self_attn.o_proj.weight
             layer.self_attn.past_key_value = cache_cls(
@@ -836,12 +829,6 @@ class GemmaModel(GemmaPreTrainedModel):
         self.norm = GemmaRMSNorm(config.hidden_size, eps=config.rms_norm_eps)
         self.gradient_checkpointing = False
 
-        # Register a causal mask to separate causal and padding mask creation. Merging happens in the attention class.
-        # NOTE: This is not friendly with TorchScript, ONNX, ExportedProgram serialization for very large `max_position_embeddings`.
-        causal_mask = torch.full(
-            (config.max_position_embeddings, config.max_position_embeddings), fill_value=True, dtype=torch.bool
-        )
-        self.register_buffer("causal_mask", torch.triu(causal_mask, diagonal=1), persistent=False)
         # Initialize weights and apply final processing
         self.post_init()
 
@@ -901,7 +888,7 @@ class GemmaModel(GemmaPreTrainedModel):
         if position_ids is None:
             position_ids = cache_position.unsqueeze(0)
 
-        causal_mask = self._update_causal_mask(attention_mask, inputs_embeds)
+        causal_mask = self._update_causal_mask(attention_mask, inputs_embeds, cache_position)
 
         # embed positions
         hidden_states = inputs_embeds
@@ -975,26 +962,27 @@ class GemmaModel(GemmaPreTrainedModel):
     # KV cache is used. This is an issue for torch.compile which then recaptures cudagraphs at each decode steps due to the dynamic shapes.
     # (`recording cudagraph tree for symint key 13`, etc.), which is VERY slow. A workaround is `@torch.compiler.disable`, but this prevents using
     # `fullgraph=True`. See more context in https://github.com/huggingface/transformers/pull/29114
-    def _update_causal_mask(self, attention_mask, input_tensor):
+    def _update_causal_mask(self, attention_mask, input_tensor, cache_position):
         if self.config._attn_implementation == "flash_attention_2":
             if attention_mask is not None and 0.0 in attention_mask:
                 return attention_mask
             return None
 
-        batch_size, seq_length = input_tensor.shape[:2]
-        dtype = input_tensor.dtype
-        device = input_tensor.device
-
-        # support going beyond cached `max_position_embedding`
-        if seq_length > self.causal_mask.shape[-1]:
-            causal_mask = torch.full((2 * self.causal_mask.shape[-1], 2 * self.causal_mask.shape[-1]), fill_value=1)
-            self.register_buffer("causal_mask", torch.triu(causal_mask, diagonal=1), persistent=False)
-
-        # We use the current dtype to avoid any overflows
+        dtype, device = input_tensor.dtype, input_tensor.device
         min_dtype = torch.finfo(dtype).min
+        sequence_length = input_tensor.shape[1]
+        if hasattr(self.layers[0].self_attn, "past_key_value"):  # static cache
+            target_length = self.config.max_position_embeddings
+        else:  # dynamic cache
+            target_length = (
+                attention_mask.shape[-1] if isinstance(attention_mask, torch.Tensor) else cache_position[-1] + 1
+            )
 
-        causal_mask = self.causal_mask[None, None, :, :].to(dtype=dtype, device=device) * min_dtype
-        causal_mask = causal_mask.expand(batch_size, 1, -1, -1)
+        causal_mask = torch.full((sequence_length, target_length), fill_value=min_dtype, dtype=dtype, device=device)
+        if sequence_length != 1:
+            causal_mask = torch.triu(causal_mask, diagonal=1)
+        causal_mask *= torch.arange(target_length, device=device) > cache_position.reshape(-1, 1)
+        causal_mask = causal_mask[None, None, :, :].expand(input_tensor.shape[0], 1, -1, -1)
         if attention_mask is not None:
             causal_mask = causal_mask.clone()  # copy to contiguous memory for in-place edit
             if attention_mask.dim() == 2:
@@ -1002,9 +990,17 @@ class GemmaModel(GemmaPreTrainedModel):
                 padding_mask = causal_mask[..., :mask_length].eq(0.0) * attention_mask[:, None, None, :].eq(0.0)
                 causal_mask[..., :mask_length] = causal_mask[..., :mask_length].masked_fill(padding_mask, min_dtype)
             elif attention_mask.dim() == 4:
+                # backwards compatibility: we allow passing a 4D attention mask shorter than the input length with
+                # cache. In that case, the 4D attention mask attends to the newest tokens only.
+                if attention_mask.shape[-2] < cache_position[0] + sequence_length:
+                    offset = cache_position[0]
+                else:
+                    offset = 0
                 mask_shape = attention_mask.shape
                 mask_slice = (attention_mask.eq(0.0)).to(dtype=dtype) * min_dtype
-                causal_mask[: mask_shape[0], : mask_shape[1], : mask_shape[2], : mask_shape[3]] = mask_slice
+                causal_mask[
+                    : mask_shape[0], : mask_shape[1], offset : mask_shape[2] + offset, : mask_shape[3]
+                ] = mask_slice
 
         if (
             self.config._attn_implementation == "sdpa"

src/transformers/models/llama/modeling_llama.py

@@ -371,9 +371,7 @@ class LlamaAttention(nn.Module):
         attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) / math.sqrt(self.head_dim)
 
         if attention_mask is not None:  # no matter the length, we just slice it
-            causal_mask = attention_mask
-            if cache_position is not None:
-                causal_mask = attention_mask[:, :, cache_position, : key_states.shape[-2]]
+            causal_mask = attention_mask[:, :, :, : key_states.shape[-2]]
             attn_weights = attn_weights + causal_mask
 
         # upcast attention to fp32
@@ -658,8 +656,9 @@ class LlamaSdpaAttention(LlamaAttention):
         value_states = repeat_kv(value_states, self.num_key_value_groups)
 
         causal_mask = attention_mask
-        if attention_mask is not None and cache_position is not None:
-            causal_mask = causal_mask[:, :, cache_position, : key_states.shape[-2]]
+        # if attention_mask is not None and cache_position is not None:
+        if attention_mask is not None:
+            causal_mask = causal_mask[:, :, :, : key_states.shape[-2]]
 
         # SDPA with memory-efficient backend is currently (torch==2.1.2) bugged with non-contiguous inputs with custom attn_mask,
         # Reference: https://github.com/pytorch/pytorch/issues/112577.
@@ -792,7 +791,7 @@ class LlamaPreTrainedModel(PreTrainedModel):
     base_model_prefix = "model"
     supports_gradient_checkpointing = True
     _no_split_modules = ["LlamaDecoderLayer"]
-    _skip_keys_device_placement = ["past_key_values", "causal_mask"]
+    _skip_keys_device_placement = ["past_key_values"]
     _supports_flash_attn_2 = True
     _supports_sdpa = True
     _supports_cache_class = True
@@ -815,12 +814,6 @@ class LlamaPreTrainedModel(PreTrainedModel):
                 "make sure to use `sdpa` in the mean time, and open an issue at https://github.com/huggingface/transformers"
             )
 
-        if max_cache_len > self.model.causal_mask.shape[-1] or self.device != self.model.causal_mask.device:
-            causal_mask = torch.full(
-                (max_cache_len, max_cache_len), fill_value=True, device=self.device, dtype=torch.bool
-            )
-            self.register_buffer("causal_mask", torch.triu(causal_mask, diagonal=1), persistent=False)
-
         for layer in self.model.layers:
             device = layer.input_layernorm.weight.device
             if hasattr(self.config, "_pre_quantization_dtype"):
@@ -934,12 +927,6 @@ class LlamaModel(LlamaPreTrainedModel):
         self.norm = LlamaRMSNorm(config.hidden_size, eps=config.rms_norm_eps)
         self.gradient_checkpointing = False
 
-        # Register a causal mask to separate causal and padding mask creation. Merging happens in the attention class.
-        # NOTE: This is not friendly with TorchScript, ONNX, ExportedProgram serialization for very large `max_position_embeddings`.
-        causal_mask = torch.full(
-            (config.max_position_embeddings, config.max_position_embeddings), fill_value=True, dtype=torch.bool
-        )
-        self.register_buffer("causal_mask", torch.triu(causal_mask, diagonal=1), persistent=False)
         # Initialize weights and apply final processing
         self.post_init()
 
@@ -1000,7 +987,7 @@ class LlamaModel(LlamaPreTrainedModel):
         if position_ids is None:
             position_ids = cache_position.unsqueeze(0)
 
-        causal_mask = self._update_causal_mask(attention_mask, inputs_embeds)
+        causal_mask = self._update_causal_mask(attention_mask, inputs_embeds, cache_position)
 
         # embed positions
         hidden_states = inputs_embeds
@@ -1068,25 +1055,27 @@ class LlamaModel(LlamaPreTrainedModel):
     # KV cache is used. This is an issue for torch.compile which then recaptures cudagraphs at each decode steps due to the dynamic shapes.
     # (`recording cudagraph tree for symint key 13`, etc.), which is VERY slow. A workaround is `@torch.compiler.disable`, but this prevents using
     # `fullgraph=True`. See more context in https://github.com/huggingface/transformers/pull/29114
-    def _update_causal_mask(self, attention_mask, input_tensor):
+    def _update_causal_mask(self, attention_mask, input_tensor, cache_position):
         if self.config._attn_implementation == "flash_attention_2":
             if attention_mask is not None and 0.0 in attention_mask:
                 return attention_mask
             return None
 
-        batch_size, seq_length = input_tensor.shape[:2]
-        dtype = input_tensor.dtype
-        device = input_tensor.device
-
-        # support going beyond cached `max_position_embedding`
-        if seq_length > self.causal_mask.shape[-1]:
-            causal_mask = torch.full((2 * self.causal_mask.shape[-1], 2 * self.causal_mask.shape[-1]), fill_value=1)
-            self.register_buffer("causal_mask", torch.triu(causal_mask, diagonal=1), persistent=False)
-
-        # We use the current dtype to avoid any overflows
+        dtype, device = input_tensor.dtype, input_tensor.device
         min_dtype = torch.finfo(dtype).min
-        causal_mask = self.causal_mask[None, None, :, :].to(dtype=dtype, device=device) * min_dtype
-        causal_mask = causal_mask.expand(batch_size, 1, -1, -1)
+        sequence_length = input_tensor.shape[1]
+        if hasattr(self.layers[0].self_attn, "past_key_value"):  # static cache
+            target_length = self.config.max_position_embeddings
+        else:  # dynamic cache
+            target_length = (
+                attention_mask.shape[-1] if isinstance(attention_mask, torch.Tensor) else cache_position[-1] + 1
+            )
+
+        causal_mask = torch.full((sequence_length, target_length), fill_value=min_dtype, dtype=dtype, device=device)
+        if sequence_length != 1:
+            causal_mask = torch.triu(causal_mask, diagonal=1)
+        causal_mask *= torch.arange(target_length, device=device) > cache_position.reshape(-1, 1)
+        causal_mask = causal_mask[None, None, :, :].expand(input_tensor.shape[0], 1, -1, -1)
         if attention_mask is not None:
             causal_mask = causal_mask.clone()  # copy to contiguous memory for in-place edit
             if attention_mask.dim() == 2:
@@ -1094,9 +1083,17 @@ class LlamaModel(LlamaPreTrainedModel):
                 padding_mask = causal_mask[..., :mask_length].eq(0.0) * attention_mask[:, None, None, :].eq(0.0)
                 causal_mask[..., :mask_length] = causal_mask[..., :mask_length].masked_fill(padding_mask, min_dtype)
             elif attention_mask.dim() == 4:
+                # backwards compatibility: we allow passing a 4D attention mask shorter than the input length with
+                # cache. In that case, the 4D attention mask attends to the newest tokens only.
+                if attention_mask.shape[-2] < cache_position[0] + sequence_length:
+                    offset = cache_position[0]
+                else:
+                    offset = 0
                 mask_shape = attention_mask.shape
                 mask_slice = (attention_mask.eq(0.0)).to(dtype=dtype) * min_dtype
-                causal_mask[: mask_shape[0], : mask_shape[1], : mask_shape[2], : mask_shape[3]] = mask_slice
+                causal_mask[
+                    : mask_shape[0], : mask_shape[1], offset : mask_shape[2] + offset, : mask_shape[3]
+                ] = mask_slice
 
         if (
             self.config._attn_implementation == "sdpa"

src/transformers/models/llava/__init__.py

@@ -16,7 +16,10 @@ from typing import TYPE_CHECKING
 from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available
 
 
-_import_structure = {"configuration_llava": ["LLAVA_PRETRAINED_CONFIG_ARCHIVE_MAP", "LlavaConfig"]}
+_import_structure = {
+    "configuration_llava": ["LLAVA_PRETRAINED_CONFIG_ARCHIVE_MAP", "LlavaConfig"],
+    "processing_llava": ["LlavaProcessor"],
+}
 
 
 try:
@@ -30,11 +33,11 @@ else:
         "LlavaForConditionalGeneration",
         "LlavaPreTrainedModel",
     ]
-    _import_structure["processing_llava"] = ["LlavaProcessor"]
 
 
 if TYPE_CHECKING:
     from .configuration_llava import LLAVA_PRETRAINED_CONFIG_ARCHIVE_MAP, LlavaConfig
+    from .processing_llava import LlavaProcessor
 
     try:
         if not is_torch_available():
@@ -47,8 +50,6 @@ if TYPE_CHECKING:
             LlavaForConditionalGeneration,
             LlavaPreTrainedModel,
         )
-        from .processing_llava import LlavaProcessor
-
 
 else:
     import sys

src/transformers/models/llava/configuration_llava.py

@@ -13,6 +13,8 @@
 # limitations under the License.
 """ Llava model configuration"""
 
+import warnings
+
 from ...configuration_utils import PretrainedConfig
 from ...utils import logging
 from ..auto import CONFIG_MAPPING
@@ -37,10 +39,10 @@ class LlavaConfig(PretrainedConfig):
     documentation from [`PretrainedConfig`] for more information.
 
     Args:
-        vision_config (`LlavaVisionConfig`,  *optional*):
-            Custom vision config or dict
-        text_config (`Union[AutoConfig, dict]`, *optional*):
-            The config object of the text backbone. Can be any of `LlamaConfig` or `MistralConfig`.
+        vision_config (`Union[AutoConfig, dict]`,  *optional*, defaults to `CLIPVisionConfig`):
+            The config object or dictionary of the vision backbone.
+        text_config (`Union[AutoConfig, dict]`, *optional*, defaults to `LlamaConfig`):
+            The config object or dictionary of the text backbone.
         ignore_index (`int`, *optional*, defaults to -100):
             The ignore index for the loss function.
         image_token_index (`int`, *optional*, defaults to 32000):
@@ -48,12 +50,10 @@ class LlavaConfig(PretrainedConfig):
         projector_hidden_act (`str`, *optional*, defaults to `"gelu"`):
             The activation function used by the multimodal projector.
         vision_feature_select_strategy (`str`, *optional*, defaults to `"default"`):
-            The feature selection strategy used to select the vision feature from the CLIP backbone.
+            The feature selection strategy used to select the vision feature from the vision backbone.
+            Can be one of `"default"` or `"full"`.
         vision_feature_layer (`int`, *optional*, defaults to -2):
             The index of the layer to select the vision feature.
-        vocab_size (`int`, *optional*, defaults to 32000):
-            Vocabulary size of the Llava model. Defines the number of different tokens that can be represented by the
-            `inputs_ids` passed when calling [`~LlavaForConditionalGeneration`]
 
     Example:
 
@@ -88,25 +88,34 @@ class LlavaConfig(PretrainedConfig):
         projector_hidden_act="gelu",
         vision_feature_select_strategy="default",
         vision_feature_layer=-2,
-        vocab_size=32000,
         **kwargs,
     ):
         self.ignore_index = ignore_index
         self.image_token_index = image_token_index
         self.projector_hidden_act = projector_hidden_act
+
+        if vision_feature_select_strategy not in ["default", "full"]:
+            raise ValueError(
+                "vision_feature_select_strategy should be one of 'default', 'full'."
+                f"Got: {vision_feature_select_strategy}"
+            )
+
+        if "vocab_size" in kwargs:
+            warnings.warn(
+                "The `vocab_size` argument is deprecated and will be removed in v4.42, since it can be inferred from the `text_config`. Passing this argument has no effect",
+                FutureWarning,
+            )
+
         self.vision_feature_select_strategy = vision_feature_select_strategy
         self.vision_feature_layer = vision_feature_layer
-        self.vocab_size = vocab_size
 
-        self.vision_config = vision_config
-
-        if isinstance(self.vision_config, dict):
+        if isinstance(vision_config, dict):
             vision_config["model_type"] = (
                 vision_config["model_type"] if "model_type" in vision_config else "clip_vision_model"
             )
-            self.vision_config = CONFIG_MAPPING[vision_config["model_type"]](**vision_config)
+            vision_config = CONFIG_MAPPING[vision_config["model_type"]](**vision_config)
         elif vision_config is None:
-            self.vision_config = CONFIG_MAPPING["clip_vision_model"](
+            vision_config = CONFIG_MAPPING["clip_vision_model"](
                 intermediate_size=4096,
                 hidden_size=1024,
                 patch_size=14,
@@ -116,15 +125,33 @@ class LlavaConfig(PretrainedConfig):
                 vocab_size=32000,
                 projection_dim=768,
             )
-        self.vocab_size = self.vocab_size
+
+        self.vision_config = vision_config
+
+        if isinstance(text_config, dict):
+            text_config["model_type"] = text_config["model_type"] if "model_type" in text_config else "llama"
+            text_config = CONFIG_MAPPING[text_config["model_type"]](**text_config)
+        elif text_config is None:
+            text_config = CONFIG_MAPPING["llama"]()
 
         self.text_config = text_config
-
-        if isinstance(self.text_config, dict):
-            text_config["model_type"] = text_config["model_type"] if "model_type" in text_config else "llama"
-            self.text_config = CONFIG_MAPPING[text_config["model_type"]](**text_config)
-            self.vocab_size = self.text_config.vocab_size
-        elif text_config is None:
-            self.text_config = CONFIG_MAPPING["llama"]()
+        self._vocab_size = self.text_config.vocab_size
 
         super().__init__(**kwargs)
+
+    @property
+    def vocab_size(self):
+        warnings.warn(
+            "The `vocab_size` attribute is deprecated and will be removed in v4.42, Please use `text_config.vocab_size` instead.",
+            FutureWarning,
+        )
+        return self._vocab_size
+
+    @vocab_size.setter
+    def vocab_size(self, value):
+        self._vocab_size = value
+
+    def to_dict(self):
+        output = super().to_dict()
+        output.pop("_vocab_size", None)
+        return output

src/transformers/models/llava/modeling_llava.py

@@ -216,6 +216,11 @@ LLAVA_INPUTS_DOCSTRING = r"""
             Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This
             is useful if you want more control over how to convert `input_ids` indices into associated vectors than the
             model's internal embedding lookup matrix.
+        vision_feature_layer (`int`, *optional*, defaults to -2):
+            The index of the layer to select the vision feature.
+        vision_feature_select_strategy (`str`, *optional*, defaults to `"default"`):
+            The feature selection strategy used to select the vision feature from the vision backbone.
+            Can be one of `"default"` or `"full"`.
         use_cache (`bool`, *optional*):
             If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
             `past_key_values`).
@@ -240,7 +245,7 @@ class LlavaForConditionalGeneration(LlavaPreTrainedModel):
         self.vision_tower = AutoModel.from_config(config.vision_config)
 
         self.multi_modal_projector = LlavaMultiModalProjector(config)
-        self.vocab_size = config.vocab_size
+        self.vocab_size = config.text_config.vocab_size
         self.language_model = AutoModelForCausalLM.from_config(
             config.text_config, attn_implementation=config._attn_implementation
         )
@@ -272,7 +277,6 @@ class LlavaForConditionalGeneration(LlavaPreTrainedModel):
         model_embeds = self.language_model.resize_token_embeddings(new_num_tokens, pad_to_multiple_of)
         # update vocab size
         self.config.text_config.vocab_size = model_embeds.num_embeddings
-        self.config.vocab_size = model_embeds.num_embeddings
         self.vocab_size = model_embeds.num_embeddings
         return model_embeds
 
@@ -340,6 +344,12 @@ class LlavaForConditionalGeneration(LlavaPreTrainedModel):
         final_attention_mask |= image_to_overwrite
         position_ids = (final_attention_mask.cumsum(-1) - 1).masked_fill_((final_attention_mask == 0), 1)
 
+        # 6. Mask out the embedding at padding positions, as we later use the past_key_value value to determine the non-attended tokens.
+        batch_indices, pad_indices = torch.where(input_ids == self.pad_token_id)
+        indices_to_mask = new_token_positions[batch_indices, pad_indices]
+
+        final_embedding[batch_indices, indices_to_mask] = 0
+
         if labels is None:
             final_labels = None
 
@@ -433,38 +443,39 @@ class LlavaForConditionalGeneration(LlavaPreTrainedModel):
                 )
                 if labels is None:
                     labels = torch.full_like(attention_mask, self.config.ignore_index).to(torch.long)
-            else:
-                # In case input_ids.shape[1] == 1 & pixel_values==None & past_key_values != None, we are in the case of
-                # generation with cache
-                if past_key_values is not None and pixel_values is not None and input_ids.shape[1] == 1:
-                    # Retrieve the first layer to inspect the logits and mask out the hidden states
-                    # that are set to 0
-                    first_layer_past_key_value = past_key_values[0][0][:, :, :, 0]
 
-                    # Sum all dimensions of head_dim (-2) to avoid random errors such as: https://github.com/huggingface/transformers/pull/28032#issuecomment-1863691941
-                    batch_index, non_attended_tokens = torch.where(first_layer_past_key_value.float().sum(-2) == 0)
+            # In case input_ids.shape[1] == 1 & pixel_values==None & past_key_values != None, we are in the case of
+            # generation with cache
+            elif past_key_values is not None and pixel_values is not None and input_ids.shape[1] == 1:
+                # Retrieve the first layer to inspect the logits and mask out the hidden states
+                # that are set to 0
+                first_layer_past_key_value = past_key_values[0][0][:, :, :, 0]
 
-                    # Get the target length
-                    target_seqlen = first_layer_past_key_value.shape[-1] + 1
+                # Sum all dimensions of head_dim (-2) to avoid random errors such as: https://github.com/huggingface/transformers/pull/28032#issuecomment-1863691941
+                batch_index, non_attended_tokens = torch.where(first_layer_past_key_value.float().sum(-2) == 0)
 
-                    extended_attention_mask = torch.ones(
-                        (attention_mask.shape[0], target_seqlen - attention_mask.shape[1]),
-                        dtype=attention_mask.dtype,
-                        device=attention_mask.device,
-                    )
+                # Get the target length
+                target_length = input_ids.shape[1]
+                past_length = first_layer_past_key_value.shape[-1]
 
-                    # Filter out only the tokens that can be un-attended, this can happen
-                    # if one uses Llava + Fused modules where the cache on the
-                    # first iteration is already big enough, or if one passes custom cache
-                    valid_indices = non_attended_tokens < extended_attention_mask.size(-1)
-                    new_batch_index = batch_index[valid_indices]
-                    new_non_attended_tokens = non_attended_tokens[valid_indices]
+                extended_attention_mask = torch.ones(
+                    (attention_mask.shape[0], past_length),
+                    dtype=attention_mask.dtype,
+                    device=attention_mask.device,
+                )
 
-                    # Zero-out the places where we don't need to attend
-                    extended_attention_mask[new_batch_index, new_non_attended_tokens] = 0
+                # Filter out only the tokens that can be un-attended, this can happen
+                # if one uses Llava + Fused modules where the cache on the
+                # first iteration is already big enough, or if one passes custom cache
+                valid_indices = non_attended_tokens < extended_attention_mask.size(-1)
+                new_batch_index = batch_index[valid_indices]
+                new_non_attended_tokens = non_attended_tokens[valid_indices]
 
-                    attention_mask = torch.cat((attention_mask, extended_attention_mask), dim=1)
-                    position_ids = torch.sum(attention_mask, dim=1).unsqueeze(-1) - 1
+                # Zero-out the places where we don't need to attend
+                extended_attention_mask[new_batch_index, new_non_attended_tokens] = 0
+
+                attention_mask = torch.cat((extended_attention_mask, attention_mask[:, -target_length:]), dim=1)
+                position_ids = torch.sum(attention_mask, dim=1).unsqueeze(-1) - 1
 
         outputs = self.language_model(
             attention_mask=attention_mask,

src/transformers/models/llava_next/__init__.py

@@ -0,0 +1,74 @@
+# Copyright 2024 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from typing import TYPE_CHECKING
+
+from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available, is_vision_available
+
+
+_import_structure = {
+    "configuration_llava_next": ["LLAVA_NEXT_PRETRAINED_CONFIG_ARCHIVE_MAP", "LlavaNextConfig"],
+    "processing_llava_next": ["LlavaNextProcessor"],
+}
+
+
+try:
+    if not is_torch_available():
+        raise OptionalDependencyNotAvailable()
+except OptionalDependencyNotAvailable:
+    pass
+else:
+    _import_structure["modeling_llava_next"] = [
+        "LLAVA_NEXT_PRETRAINED_MODEL_ARCHIVE_LIST",
+        "LlavaNextForConditionalGeneration",
+        "LlavaNextPreTrainedModel",
+    ]
+
+try:
+    if not is_vision_available():
+        raise OptionalDependencyNotAvailable()
+except OptionalDependencyNotAvailable:
+    pass
+else:
+    _import_structure["image_processing_llava_next"] = ["LlavaNextImageProcessor"]
+
+
+if TYPE_CHECKING:
+    from .configuration_llava_next import LLAVA_NEXT_PRETRAINED_CONFIG_ARCHIVE_MAP, LlavaNextConfig
+    from .processing_llava_next import LlavaNextProcessor
+
+    try:
+        if not is_torch_available():
+            raise OptionalDependencyNotAvailable()
+    except OptionalDependencyNotAvailable:
+        pass
+    else:
+        from .modeling_llava_next import (
+            LLAVA_NEXT_PRETRAINED_MODEL_ARCHIVE_LIST,
+            LlavaNextForConditionalGeneration,
+            LlavaNextPreTrainedModel,
+        )
+
+    try:
+        if not is_vision_available():
+            raise OptionalDependencyNotAvailable()
+    except OptionalDependencyNotAvailable:
+        pass
+    else:
+        from .image_processing_llava_next import LlavaNextImageProcessor
+
+
+else:
+    import sys
+
+    sys.modules[__name__] = _LazyModule(__name__, globals()["__file__"], _import_structure)

src/transformers/models/llava_next/configuration_llava_next.py

@@ -0,0 +1,141 @@
+# coding=utf-8
+# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+""" Llava-NeXT model configuration"""
+
+from ...configuration_utils import PretrainedConfig
+from ...utils import logging
+from ..auto import CONFIG_MAPPING
+
+
+logger = logging.get_logger(__name__)
+
+LLAVA_NEXT_PRETRAINED_CONFIG_ARCHIVE_MAP = {
+    "llava-hf/llava-v1.6-mistral-7b-hf": "https://huggingface.co/llava-hf/llava-v1.6-mistral-7b-hf/resolve/main/config.json",
+}
+
+
+class LlavaNextConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`LlavaNextForConditionalGeneration`]. It is used to instantiate an
+    Llava-NeXT model according to the specified arguments, defining the model architecture. Instantiating a configuration
+    with the defaults will yield a similar configuration to that of the [llava-hf/llava-v1.6-mistral-7b-hf](https://huggingface.co/llava-hf/llava-v1.6-mistral-7b-hf)
+    model.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        vision_config (`Union[AutoConfig, dict]`,  *optional*, defaults to `CLIPVisionConfig`):
+            The config object or dictionary of the vision backbone.
+        text_config (`Union[AutoConfig, dict]`, *optional*, defaults to `LlamaConfig`):
+            The config object or dictionary of the text backbone.
+        ignore_index (`int`, *optional*, defaults to -100):
+            The ignore index for the loss function.
+        image_token_index (`int`, *optional*, defaults to 32000):
+            The image token index to encode the image prompt.
+        projector_hidden_act (`str`, *optional*, defaults to `"gelu"`):
+            The activation function used by the multimodal projector.
+        vision_feature_select_strategy (`str`, *optional*, defaults to `"default"`):
+            The feature selection strategy used to select the vision feature from the vision backbone.
+            Can be one of `"default"` or `"full"`. If `"default"`, the CLS token is removed from the vision features.
+            If `"full"`, the full vision features are used.
+        vision_feature_layer (`int`, *optional*, defaults to -2):
+            The index of the layer to select the vision feature.
+        image_grid_pinpoints (`List`, *optional*, defaults to `[[336, 672], [672, 336], [672, 672], [1008, 336], [336, 1008]]`):
+            A list of possible resolutions to use for processing high resolution images. Each item in the list should be a tuple or list
+            of the form `(height, width)`.
+
+    Example:
+
+    ```python
+    >>> from transformers import LlavaNextForConditionalGeneration, LlavaNextConfig, CLIPVisionConfig, LlamaConfig
+
+    >>> # Initializing a CLIP-vision config
+    >>> vision_config = CLIPVisionConfig()
+
+    >>> # Initializing a Llama config
+    >>> text_config = LlamaConfig()
+
+    >>> # Initializing a Llava-Next llava-hf/llava-v1.6-mistral-7b-hf style configuration
+    >>> configuration = LlavaNextConfig(vision_config, text_config)
+
+    >>> # Initializing a model from the llava-hf/llava-v1.6-mistral-7b-hf style configuration
+    >>> model = LlavaNextForConditionalGeneration(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "llava_next"
+    is_composition = False
+
+    def __init__(
+        self,
+        vision_config=None,
+        text_config=None,
+        ignore_index=-100,
+        image_token_index=32000,
+        projector_hidden_act="gelu",
+        vision_feature_select_strategy="default",
+        vision_feature_layer=-2,
+        image_grid_pinpoints=None,
+        **kwargs,
+    ):
+        self.ignore_index = ignore_index
+        self.image_token_index = image_token_index
+        self.projector_hidden_act = projector_hidden_act
+
+        if vision_feature_select_strategy not in ["default", "full"]:
+            raise ValueError(
+                "vision_feature_select_strategy should be one of 'default', 'full'."
+                f"Got: {vision_feature_select_strategy}"
+            )
+
+        self.vision_feature_select_strategy = vision_feature_select_strategy
+        self.vision_feature_layer = vision_feature_layer
+        image_grid_pinpoints = (
+            image_grid_pinpoints
+            if image_grid_pinpoints is not None
+            else [[336, 672], [672, 336], [672, 672], [1008, 336], [336, 1008]]
+        )
+        self.image_grid_pinpoints = image_grid_pinpoints
+
+        if isinstance(vision_config, dict):
+            vision_config["model_type"] = (
+                vision_config["model_type"] if "model_type" in vision_config else "clip_vision_model"
+            )
+            vision_config = CONFIG_MAPPING[vision_config["model_type"]](**vision_config)
+        elif vision_config is None:
+            vision_config = CONFIG_MAPPING["clip_vision_model"](
+                intermediate_size=4096,
+                hidden_size=1024,
+                patch_size=14,
+                image_size=336,
+                num_hidden_layers=24,
+                num_attention_heads=16,
+                vocab_size=32000,
+                projection_dim=768,
+            )
+
+        self.vision_config = vision_config
+
+        if isinstance(text_config, dict):
+            text_config["model_type"] = text_config["model_type"] if "model_type" in text_config else "llama"
+            text_config = CONFIG_MAPPING[text_config["model_type"]](**text_config)
+        elif text_config is None:
+            text_config = CONFIG_MAPPING["llama"]()
+
+        self.text_config = text_config
+
+        super().__init__(**kwargs)

src/transformers/models/llava_next/convert_llava_next_weights_to_hf.py

@@ -0,0 +1,342 @@
+# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""Convert LLaVa-NeXT (LLaVa-1.6) checkpoints from the original repository.
+
+URL: https://github.com/haotian-liu/LLaVA/tree/main.
+
+
+The command used to obtain original logits is the following:
+python llava/eval/run_llava.py --model-path "liuhaotian/llava-v1.6-mistral-7b" --image-file "images/llava_v1_5_radar.jpg" --query "What is shown in this image?" --max_new_tokens 100 --temperature 0
+
+Note: logits are tested with torch==2.1.2.
+"""
+
+import argparse
+import glob
+import json
+from pathlib import Path
+
+import requests
+import torch
+from accelerate import init_empty_weights
+from huggingface_hub import hf_hub_download, snapshot_download
+from PIL import Image
+from safetensors import safe_open
+
+from transformers import (
+    AddedToken,
+    AutoConfig,
+    AutoTokenizer,
+    LlavaNextConfig,
+    LlavaNextForConditionalGeneration,
+    LlavaNextImageProcessor,
+    LlavaNextProcessor,
+)
+
+
+KEYS_TO_MODIFY_MAPPING = {
+    "model.vision_tower.": "",
+    "model.mm_projector": "multi_modal_projector",
+    "model": "model.model",
+    "vision_model.model": "vision_model",
+    "lm_head": "language_model.lm_head",
+    "model.model": "language_model.model",
+    "multi_modal_projector.0": "multi_modal_projector.linear_1",
+    "multi_modal_projector.2": "multi_modal_projector.linear_2",
+    "language_model.model.image_newline": "image_newline",
+}
+
+
+def load_original_state_dict(model_id):
+    directory_path = snapshot_download(repo_id=model_id, allow_patterns=["*.safetensors"])
+
+    original_state_dict = {}
+    for path in glob.glob(f"{directory_path}/*"):
+        if path.endswith(".safetensors"):
+            with safe_open(path, framework="pt", device="cpu") as f:
+                for key in f.keys():
+                    original_state_dict[key] = f.get_tensor(key)
+
+    return original_state_dict
+
+
+def convert_state_dict_to_hf(state_dict):
+    new_state_dict = {}
+    for key, value in state_dict.items():
+        if key.endswith(".inv_freq"):
+            continue
+        for key_to_modify, new_key in KEYS_TO_MODIFY_MAPPING.items():
+            if key_to_modify in key:
+                key = key.replace(key_to_modify, new_key)
+
+        new_state_dict[key] = value.to(torch.float16)
+    return new_state_dict
+
+
+def load_image():
+    url = "https://github.com/haotian-liu/LLaVA/blob/1a91fc274d7c35a9b50b3cb29c4247ae5837ce39/images/llava_v1_5_radar.jpg?raw=true"
+    image = Image.open(requests.get(url, stream=True).raw)
+    return image
+
+
+def convert_llava_to_hf(model_id, pytorch_dump_folder_path, push_to_hub=False):
+    # load original config
+    filepath = hf_hub_download(repo_id=model_id, filename="config.json", repo_type="model")
+    # read json
+    with open(filepath) as f:
+        data = json.load(f)
+        print(data)
+
+    if model_id == "liuhaotian/llava-v1.6-mistral-7b":
+        text_model_id = "mistralai/Mistral-7B-Instruct-v0.2"
+        image_token_index = 32000
+    elif model_id == "liuhaotian/llava-v1.6-vicuna-7b":
+        text_model_id = "lmsys/vicuna-7b-v1.5"
+        image_token_index = 32000
+    elif model_id == "liuhaotian/llava-v1.6-vicuna-13b":
+        text_model_id = "lmsys/vicuna-13b-v1.5"
+        image_token_index = 32000
+    elif model_id == "liuhaotian/llava-v1.6-34b":
+        text_model_id = "NousResearch/Nous-Hermes-2-Yi-34B"
+        image_token_index = 64000
+    vision_model_id = data["mm_vision_tower"]
+
+    torch.set_default_dtype(torch.float16)
+    text_config = AutoConfig.from_pretrained(text_model_id)
+
+    use_fast = False if model_id == "liuhaotian/llava-v1.6-34b" else True
+    tokenizer = AutoTokenizer.from_pretrained(text_model_id, use_fast=use_fast)
+    tokenizer.add_tokens(AddedToken("<image>", special=True, normalized=False), special_tokens=True)
+
+    if model_id == "liuhaotian/llava-v1.6-mistral-7b":
+        # Mistral-7B doesn't have a padding token set yet
+        tokenizer.add_special_tokens({"pad_token": "<pad>"})
+
+    image_processor = LlavaNextImageProcessor.from_pretrained(vision_model_id)
+    processor = LlavaNextProcessor(tokenizer=tokenizer, image_processor=image_processor)
+
+    config = LlavaNextConfig(
+        text_config=text_config.to_dict(),
+        image_grid_pinpoints=image_processor.image_grid_pinpoints,
+        use_image_newline_parameter=True,
+        image_token_index=image_token_index,
+    )
+
+    with init_empty_weights():
+        model = LlavaNextForConditionalGeneration(config)
+
+    # load original state dict
+    state_dict = load_original_state_dict(model_id)
+    state_dict = convert_state_dict_to_hf(state_dict)
+    model.load_state_dict(state_dict, assign=True)
+    model.eval()
+
+    pre_expansion_embeddings = model.language_model.model.embed_tokens.weight.data
+    mu = torch.mean(pre_expansion_embeddings, dim=0).float()
+    n = pre_expansion_embeddings.size()[0]
+    sigma = ((pre_expansion_embeddings - mu).T @ (pre_expansion_embeddings - mu)) / n
+    dist = torch.distributions.multivariate_normal.MultivariateNormal(mu, covariance_matrix=1e-5 * sigma)
+
+    # We add an image token so we resize the model
+    # Pad to 64 for performance reasons
+    pad_shape = 64
+    vocab_size = config.text_config.vocab_size
+    if model_id == "liuhaotian/llava-v1.6-34b":
+        # this one has 3 additional tokens, namely <|startoftext|>, <|endoftext|> and <image>
+        num_tokens = vocab_size + 3
+    else:
+        # this one has 2 additional tokens, namely <image> and <pad>
+        num_tokens = vocab_size + 2
+    model.resize_token_embeddings(num_tokens, pad_to_multiple_of=pad_shape)
+    model.language_model.model.embed_tokens.weight.data[vocab_size:] = torch.stack(
+        tuple(
+            (dist.sample() for _ in range(model.language_model.model.embed_tokens.weight.data[vocab_size:].shape[0]))
+        ),
+        dim=0,
+    )
+    model.language_model.lm_head.weight.data[vocab_size:] = torch.stack(
+        tuple((dist.sample() for _ in range(model.language_model.lm_head.weight.data[vocab_size:].shape[0]))),
+        dim=0,
+    )
+
+    device = "cuda:2"
+    model.to(device)
+
+    # prepare inputs
+    image = load_image()
+    if model_id == "liuhaotian/llava-v1.6-mistral-7b":
+        prompt = "[INST] <image>\nWhat is shown in this image? [/INST]"
+    elif model_id in ["liuhaotian/llava-v1.6-vicuna-7b", "liuhaotian/llava-v1.6-vicuna-13b"]:
+        prompt = "A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions. USER: <image>\nWhat is shown in this image? ASSISTANT:"
+    elif model_id == "liuhaotian/llava-v1.6-34b":
+        prompt = "<|im_start|>system\nAnswer the questions.<|im_end|><|im_start|>user\n<image>\nWhat is shown in this image?<|im_end|><|im_start|>assistant\n"
+    inputs = processor(images=image, text=prompt, return_tensors="pt")
+
+    # verify inputs
+    filepath = hf_hub_download(repo_id="nielsr/test-image", filename="llava_1_6_pixel_values.pt", repo_type="dataset")
+    original_pixel_values = torch.load(filepath, map_location="cpu")
+    assert torch.allclose(original_pixel_values, inputs.pixel_values.half())
+
+    if model_id == "liuhaotian/llava-v1.6-mistral-7b":
+        filepath = hf_hub_download(repo_id="nielsr/test-image", filename="llava_1_6_input_ids.pt", repo_type="dataset")
+        original_input_ids = torch.load(filepath, map_location="cpu")
+        # replace -200 by image_token_index (since we use token ID = 32000 for the image token)
+        original_input_ids[original_input_ids == -200] = image_token_index
+        print(tokenizer.decode([id for id in original_input_ids.tolist()[0] if id != -200]))
+
+        assert original_input_ids[0].tolist() == inputs.input_ids[0].tolist()
+
+    elif model_id == "liuhaotian/llava-v1.6-34b":
+        filepath = hf_hub_download(
+            repo_id="nielsr/test-image", filename="llava_1_6_34b_input_ids.pt", repo_type="dataset"
+        )
+        original_input_ids = torch.load(filepath, map_location="cpu")
+        # replace -200 by image_token_index
+        original_input_ids[original_input_ids == -200] = image_token_index
+
+        assert original_input_ids[0].tolist() == inputs.input_ids[0].tolist()
+
+    image_sizes = torch.tensor([[899, 1024]])
+    assert image_sizes[0].tolist() == inputs.image_sizes[0].tolist()
+
+    # verify single forward pass
+    print("Single forward pass")
+    with torch.inference_mode():
+        inputs = inputs.to(device)
+        outputs = model(**inputs)
+        print("Shape of logits:", outputs.logits.shape)
+        print("First values of logits:", outputs.logits[0, :3, :3])
+
+        if model_id == "liuhaotian/llava-v1.6-mistral-7b":
+            expected_slice = torch.tensor(
+                [[-4.8555, -4.6992, -0.1996], [-10.5703, -10.7344, -2.7246], [-7.0391, -7.3672, -0.2634]],
+                dtype=torch.float32,
+                device=device,
+            )
+        elif model_id == "liuhaotian/llava-v1.6-vicuna-7b":
+            expected_slice = torch.tensor(
+                [[1.4883, 0.9976, -0.6992], [-9.7031, -5.7031, -1.5557], [-5.1328, -5.5586, 8.8281]],
+                dtype=torch.float32,
+                device=device,
+            )
+        elif model_id == "liuhaotian/llava-v1.6-vicuna-13b":
+            expected_slice = torch.tensor(
+                [[-0.9614, 7.3125, 0.2106], [-7.2695, -8.5469, 3.6211], [-6.3750, -8.1875, 5.4688]],
+                dtype=torch.float32,
+                device=device,
+            )
+        elif model_id == "liuhaotian/llava-v1.6-34b":
+            expected_slice = torch.tensor(
+                [[-9.0859, -9.1406, 5.9453], [-5.9570, -5.9766, 2.2754], [-5.7305, -5.7539, 4.0000]],
+                dtype=torch.float32,
+                device=device,
+            )
+        else:
+            raise ValueError(f"Model {model_id} not supported")
+
+        assert torch.allclose(outputs.logits[0, :3, :3], expected_slice, atol=1e-4)
+        print("Logits are ok!")
+
+    # verify generation
+    output_ids = model.generate(
+        **inputs,
+        max_new_tokens=100,
+        use_cache=True,
+    )
+
+    generated_text = processor.batch_decode(output_ids, skip_special_tokens=True)[0].strip()
+
+    print("Generated text:", repr(generated_text))
+
+    if model_id == "liuhaotian/llava-v1.6-mistral-7b":
+        expected_text = '[INST]  \nWhat is shown in this image? [/INST] The image appears to be a radar chart, which is a type of multi-dimensional plot that displays data in the form of a two-dimensional chart of three or more quantitative variables represented on axes starting from the same point.\n\nIn this particular radar chart, there are several axes labeled with different metrics or benchmarks, such as "MMM-Vet," "MMM-Bench," "LLaVA-Bench," "SLED-Bench," "'
+    elif model_id == "liuhaotian/llava-v1.6-vicuna-7b":
+        expected_text = """A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human\'s questions. USER:  \nWhat is shown in this image? ASSISTANT: The image appears to be a graphical representation of a benchmarking study comparing the performance of various models or systems. It\'s a scatter plot with a circular layout, where each point represents a different model or system, and the axes represent different metrics or dimensions of comparison.\n\nThe metrics are likely related to machine learning or artificial intelligence performance, as indicated by the terms like "BLIP-2," "Instruct BLIP," "POE," "QWA," "V"""
+    elif model_id == "liuhaotian/llava-v1.6-vicuna-13b":
+        expected_text = "A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions. USER:  \nWhat is shown in this image? ASSISTANT: The image appears to be a radar chart, also known as a spider chart or star chart, which is a graphical method of displaying multivariate data in the form of a two-dimensional chart of three or more quantitative variables represented on axes starting from the same point.\n\nIn this particular radar chart, there are several variables represented:\n\n- MM-Vet\n- LLa-Va-Bench\n- SEED-Bench\n- MM"
+    elif model_id == "liuhaotian/llava-v1.6-34b":
+        expected_text = "<|im_start|> system\nAnswer the questions. <|im_start|> user\n\nWhat is shown in this image? <|im_start|> assistant\nThe image appears to be a radar chart, also known as a spider chart, which is a graphical method of displaying multivariate data in the form of a two-dimensional chart of three or more quantitative variables represented on axes starting from the same point.\n\nIn this particular chart, there are several datasets represented by different colors and labeled with various acronyms such as MM-Vet, LLaVA-Bench, SEED-Bench, MM-Bench-CN, MM-"
+    else:
+        raise ValueError(f"Model {model_id} not supported")
+
+    assert generated_text == expected_text
+    print("Generated text is ok!")
+
+    # verify batched generation
+    print("Batched generation...")
+    url = "http://images.cocodataset.org/val2017/000000039769.jpg"
+    cats_image = Image.open(requests.get(url, stream=True).raw)
+
+    inputs = processor(
+        images=[image, cats_image],
+        text=[prompt, "[INST] <image>\nHow many cats are there? [/INST]"],
+        padding=True,
+        return_tensors="pt",
+    ).to(device)
+
+    for k, v in inputs.items():
+        print(k, v.shape)
+
+    print("Image sizes:", inputs.image_sizes)
+
+    # make sure image_sizes are the same
+    # as otherwise batched generation doesn't work
+    inputs.image_sizes[1] = inputs.image_sizes[0]
+
+    print("Batched generation...")
+    output_ids = model.generate(
+        **inputs,
+        max_new_tokens=20,
+        use_cache=True,
+    )
+
+    outputs = tokenizer.batch_decode(output_ids, skip_special_tokens=True)
+    print(outputs)
+
+    if pytorch_dump_folder_path is not None:
+        print(f"Saving model and processor for {model_id} to {pytorch_dump_folder_path}")
+        Path(pytorch_dump_folder_path).mkdir(exist_ok=True)
+        model.save_pretrained(pytorch_dump_folder_path)
+        processor.save_pretrained(pytorch_dump_folder_path)
+
+    if push_to_hub:
+        repo_id = model_id.split("/")[-1]
+        model.push_to_hub(f"llava-hf/{repo_id}-hf")
+        processor.push_to_hub(f"llava-hf/{repo_id}-hf")
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model_id",
+        help="Hub location of the model to convert",
+        default="liuhaotian/llava-v1.6-mistral-7b",
+        choices=[
+            "liuhaotian/llava-v1.6-mistral-7b",
+            "liuhaotian/llava-v1.6-vicuna-7b",
+            "liuhaotian/llava-v1.6-vicuna-13b",
+            "liuhaotian/llava-v1.6-34b",
+        ],
+        required=False,
+    )
+    parser.add_argument(
+        "--pytorch_dump_folder_path", default=None, type=str, help="Path to the output PyTorch model directory."
+    )
+    parser.add_argument(
+        "--push_to_hub", action="store_true", help="Whether or not to push the converted model to the 🤗 hub."
+    )
+    args = parser.parse_args()
+
+    convert_llava_to_hf(args.model_id, args.pytorch_dump_folder_path, args.push_to_hub)

src/transformers/models/llava_next/image_processing_llava_next.py

@@ -0,0 +1,608 @@
+# coding=utf-8
+# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""Image processor class for LLaVa-NeXT."""
+
+import math
+from typing import Dict, List, Optional, Union
+
+import numpy as np
+
+from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict, select_best_resolution
+from ...image_transforms import (
+    convert_to_rgb,
+    get_resize_output_image_size,
+    pad,
+    resize,
+    to_channel_dimension_format,
+)
+from ...image_utils import (
+    OPENAI_CLIP_MEAN,
+    OPENAI_CLIP_STD,
+    ChannelDimension,
+    ImageInput,
+    PILImageResampling,
+    get_image_size,
+    infer_channel_dimension_format,
+    is_scaled_image,
+    make_list_of_images,
+    to_numpy_array,
+    valid_images,
+    validate_preprocess_arguments,
+)
+from ...utils import TensorType, is_vision_available, logging
+
+
+logger = logging.get_logger(__name__)
+
+
+if is_vision_available():
+    from PIL import Image
+
+
+def divide_to_patches(image: np.array, patch_size: int, input_data_format) -> List[np.array]:
+    """
+    Divides an image into patches of a specified size.
+
+    Args:
+        image (`np.array`):
+            The input image.
+        patch_size (`int`):
+            The size of each patch.
+        input_data_format (`ChannelDimension` or `str`):
+            The channel dimension format of the input image.
+
+    Returns:
+        list: A list of np.array representing the patches.
+    """
+    patches = []
+    height, width = get_image_size(image, channel_dim=input_data_format)
+    for i in range(0, height, patch_size):
+        for j in range(0, width, patch_size):
+            if input_data_format == ChannelDimension.LAST:
+                patch = image[i : i + patch_size, j : j + patch_size]
+            else:
+                patch = image[:, i : i + patch_size, j : j + patch_size]
+            patches.append(patch)
+
+    return patches
+
+
+def expand_to_square(image: np.array, background_color, input_data_format) -> np.array:
+    """
+    Expands an image to a square by adding a background color.
+    """
+
+    height, width = get_image_size(image, channel_dim=input_data_format)
+    if width == height:
+        return image
+    elif width > height:
+        result = np.ones((width, width, image.shape[2]), dtype=image.dtype) * background_color
+        result[(width - height) // 2 : (width - height) // 2 + height, :] = image
+        return result
+    else:
+        result = np.ones((height, height, image.shape[2]), dtype=image.dtype) * background_color
+        result[:, (height - width) // 2 : (height - width) // 2 + width] = image
+        return result
+
+
+def _get_patch_output_size(image, target_resolution, input_data_format):
+    original_height, original_width = get_image_size(image, channel_dim=input_data_format)
+    target_height, target_width = target_resolution
+
+    scale_w = target_width / original_width
+    scale_h = target_height / original_height
+
+    if scale_w < scale_h:
+        new_width = target_width
+        new_height = min(math.ceil(original_height * scale_w), target_height)
+    else:
+        new_height = target_height
+        new_width = min(math.ceil(original_width * scale_h), target_width)
+
+    return new_height, new_width
+
+
+class LlavaNextImageProcessor(BaseImageProcessor):
+    r"""
+    Constructs a LLaVa-NeXT image processor. Based on [`CLIPImageProcessor`] with incorporation of additional techniques
+    for processing high resolution images as explained in the [LLaVa paper](https://arxiv.org/abs/2310.03744).
+
+    Args:
+        do_resize (`bool`, *optional*, defaults to `True`):
+            Whether to resize the image's (height, width) dimensions to the specified `size`. Can be overridden by
+            `do_resize` in the `preprocess` method.
+        size (`Dict[str, int]` *optional*, defaults to `{"shortest_edge": 224}`):
+            Size of the image after resizing. The shortest edge of the image is resized to size["shortest_edge"], with
+            the longest edge resized to keep the input aspect ratio. Can be overridden by `size` in the `preprocess`
+            method.
+        image_grid_pinpoints (`List` *optional*, defaults to `[[672, 336], [336, 672], [672, 672], [336, 1008], [1008, 336]]`):
+            A list of possible resolutions to use for processing high resolution images. The best resolution is selected
+            based on the original size of the image. Can be overridden by `image_grid_pinpoints` in the `preprocess`
+            method.
+        resample (`PILImageResampling`, *optional*, defaults to `Resampling.BICUBIC`):
+            Resampling filter to use if resizing the image. Can be overridden by `resample` in the `preprocess` method.
+        do_center_crop (`bool`, *optional*, defaults to `True`):
+            Whether to center crop the image to the specified `crop_size`. Can be overridden by `do_center_crop` in the
+            `preprocess` method.
+        crop_size (`Dict[str, int]` *optional*, defaults to 224):
+            Size of the output image after applying `center_crop`. Can be overridden by `crop_size` in the `preprocess`
+            method.
+        do_rescale (`bool`, *optional*, defaults to `True`):
+            Whether to rescale the image by the specified scale `rescale_factor`. Can be overridden by `do_rescale` in
+            the `preprocess` method.
+        rescale_factor (`int` or `float`, *optional*, defaults to `1/255`):
+            Scale factor to use if rescaling the image. Can be overridden by `rescale_factor` in the `preprocess`
+            method.
+        do_normalize (`bool`, *optional*, defaults to `True`):
+            Whether to normalize the image. Can be overridden by `do_normalize` in the `preprocess` method.
+        image_mean (`float` or `List[float]`, *optional*, defaults to `[0.48145466, 0.4578275, 0.40821073]`):
+            Mean to use if normalizing the image. This is a float or list of floats the length of the number of
+            channels in the image. Can be overridden by the `image_mean` parameter in the `preprocess` method.
+        image_std (`float` or `List[float]`, *optional*, defaults to `[0.26862954, 0.26130258, 0.27577711]`):
+            Standard deviation to use if normalizing the image. This is a float or list of floats the length of the
+            number of channels in the image. Can be overridden by the `image_std` parameter in the `preprocess` method.
+            Can be overridden by the `image_std` parameter in the `preprocess` method.
+        do_convert_rgb (`bool`, *optional*, defaults to `True`):
+            Whether to convert the image to RGB.
+    """
+
+    model_input_names = ["pixel_values"]
+
+    def __init__(
+        self,
+        do_resize: bool = True,
+        size: Dict[str, int] = None,
+        image_grid_pinpoints: List = None,
+        resample: PILImageResampling = PILImageResampling.BICUBIC,
+        do_center_crop: bool = True,
+        crop_size: Dict[str, int] = None,
+        do_rescale: bool = True,
+        rescale_factor: Union[int, float] = 1 / 255,
+        do_normalize: bool = True,
+        image_mean: Optional[Union[float, List[float]]] = None,
+        image_std: Optional[Union[float, List[float]]] = None,
+        do_convert_rgb: bool = True,
+        **kwargs,
+    ) -> None:
+        super().__init__(**kwargs)
+        size = size if size is not None else {"shortest_edge": 224}
+        size = get_size_dict(size, default_to_square=False)
+        image_grid_pinpoints = (
+            image_grid_pinpoints
+            if image_grid_pinpoints is not None
+            else [[336, 672], [672, 336], [672, 672], [1008, 336], [336, 1008]]
+        )
+        crop_size = crop_size if crop_size is not None else {"height": 224, "width": 224}
+        crop_size = get_size_dict(crop_size, default_to_square=True, param_name="crop_size")
+
+        self.do_resize = do_resize
+        self.size = size
+        self.image_grid_pinpoints = image_grid_pinpoints
+        self.resample = resample
+        self.do_center_crop = do_center_crop
+        self.crop_size = crop_size
+        self.do_rescale = do_rescale
+        self.rescale_factor = rescale_factor
+        self.do_normalize = do_normalize
+        self.image_mean = image_mean if image_mean is not None else OPENAI_CLIP_MEAN
+        self.image_std = image_std if image_std is not None else OPENAI_CLIP_STD
+        self.do_convert_rgb = do_convert_rgb
+
+    # Copied from transformers.models.clip.image_processing_clip.CLIPImageProcessor.resize with CLIP->LLaVa
+    def resize(
+        self,
+        image: np.ndarray,
+        size: Dict[str, int],
+        resample: PILImageResampling = PILImageResampling.BICUBIC,
+        data_format: Optional[Union[str, ChannelDimension]] = None,
+        input_data_format: Optional[Union[str, ChannelDimension]] = None,
+        **kwargs,
+    ) -> np.ndarray:
+        """
+        Resize an image. The shortest edge of the image is resized to size["shortest_edge"], with the longest edge
+        resized to keep the input aspect ratio.
+
+        Args:
+            image (`np.ndarray`):
+                Image to resize.
+            size (`Dict[str, int]`):
+                Size of the output image.
+            resample (`PILImageResampling`, *optional*, defaults to `PILImageResampling.BICUBIC`):
+                Resampling filter to use when resiizing the image.
+            data_format (`str` or `ChannelDimension`, *optional*):
+                The channel dimension format of the image. If not provided, it will be the same as the input image.
+            input_data_format (`ChannelDimension` or `str`, *optional*):
+                The channel dimension format of the input image. If not provided, it will be inferred.
+        """
+        default_to_square = True
+        if "shortest_edge" in size:
+            size = size["shortest_edge"]
+            default_to_square = False
+        elif "height" in size and "width" in size:
+            size = (size["height"], size["width"])
+        else:
+            raise ValueError("Size must contain either 'shortest_edge' or 'height' and 'width'.")
+
+        output_size = get_resize_output_image_size(
+            image,
+            size=size,
+            default_to_square=default_to_square,
+            input_data_format=input_data_format,
+        )
+
+        return resize(
+            image,
+            size=output_size,
+            resample=resample,
+            data_format=data_format,
+            input_data_format=input_data_format,
+            **kwargs,
+        )
+
+    def _preprocess(
+        self,
+        images: ImageInput,
+        do_resize: bool = None,
+        size: Dict[str, int] = None,
+        resample: PILImageResampling = None,
+        do_center_crop: bool = None,
+        crop_size: int = None,
+        do_rescale: bool = None,
+        rescale_factor: float = None,
+        do_normalize: bool = None,
+        image_mean: Optional[Union[float, List[float]]] = None,
+        image_std: Optional[Union[float, List[float]]] = None,
+        data_format: Optional[ChannelDimension] = ChannelDimension.FIRST,
+        input_data_format: Optional[Union[str, ChannelDimension]] = None,
+    ) -> Image.Image:
+        """
+        Preprocess an image or batch of images. Copy of the `preprocess` method from `CLIPImageProcessor`.
+
+        Args:
+            images (`ImageInput`):
+                Image to preprocess. Expects a single or batch of images with pixel values ranging from 0 to 255. If
+                passing in images with pixel values between 0 and 1, set `do_rescale=False`.
+            do_resize (`bool`, *optional*, defaults to `self.do_resize`):
+                Whether to resize the image.
+            size (`Dict[str, int]`, *optional*, defaults to `self.size`):
+                Size of the image after resizing. Shortest edge of the image is resized to size["shortest_edge"], with
+                the longest edge resized to keep the input aspect ratio.
+            resample (`int`, *optional*, defaults to `self.resample`):
+                Resampling filter to use if resizing the image. This can be one of the enum `PILImageResampling`. Only
+                has an effect if `do_resize` is set to `True`.
+            do_center_crop (`bool`, *optional*, defaults to `self.do_center_crop`):
+                Whether to center crop the image.
+            crop_size (`Dict[str, int]`, *optional*, defaults to `self.crop_size`):
+                Size of the center crop. Only has an effect if `do_center_crop` is set to `True`.
+            do_rescale (`bool`, *optional*, defaults to `self.do_rescale`):
+                Whether to rescale the image.
+            rescale_factor (`float`, *optional*, defaults to `self.rescale_factor`):
+                Rescale factor to rescale the image by if `do_rescale` is set to `True`.
+            do_normalize (`bool`, *optional*, defaults to `self.do_normalize`):
+                Whether to normalize the image.
+            image_mean (`float` or `List[float]`, *optional*, defaults to `self.image_mean`):
+                Image mean to use for normalization. Only has an effect if `do_normalize` is set to `True`.
+            image_std (`float` or `List[float]`, *optional*, defaults to `self.image_std`):
+                Image standard deviation to use for normalization. Only has an effect if `do_normalize` is set to
+                `True`.
+            data_format (`ChannelDimension` or `str`, *optional*, defaults to `ChannelDimension.FIRST`):
+                The channel dimension format for the output image. Can be one of:
+                - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format.
+                - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format.
+                - Unset: Use the channel dimension format of the input image.
+            input_data_format (`ChannelDimension` or `str`, *optional*):
+                The channel dimension format for the input image. If unset, the channel dimension format is inferred
+                from the input image. Can be one of:
+                - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format.
+                - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format.
+                - `"none"` or `ChannelDimension.NONE`: image in (height, width) format.
+        """
+        images = make_list_of_images(images)
+
+        if do_resize:
+            images = [
+                self.resize(image=image, size=size, resample=resample, input_data_format=input_data_format)
+                for image in images
+            ]
+
+        if do_center_crop:
+            images = [
+                self.center_crop(image=image, size=crop_size, input_data_format=input_data_format) for image in images
+            ]
+
+        if do_rescale:
+            images = [
+                self.rescale(image=image, scale=rescale_factor, input_data_format=input_data_format)
+                for image in images
+            ]
+
+        if do_normalize:
+            images = [
+                self.normalize(image=image, mean=image_mean, std=image_std, input_data_format=input_data_format)
+                for image in images
+            ]
+
+        images = [
+            to_channel_dimension_format(image, data_format, input_channel_dim=input_data_format) for image in images
+        ]
+
+        return images
+
+    def _resize_for_patching(
+        self, image: np.array, target_resolution: tuple, resample, input_data_format: ChannelDimension
+    ) -> np.array:
+        """
+        Resizes an image to a target resolution while maintaining aspect ratio.
+
+        Args:
+            image (np.array):
+                The input image.
+            target_resolution (tuple):
+                The target resolution (height, width) of the image.
+            resample (`PILImageResampling`):
+                Resampling filter to use if resizing the image.
+            input_data_format (`ChannelDimension` or `str`):
+                The channel dimension format of the input image.
+
+        Returns:
+            np.array: The resized and padded image.
+        """
+        new_height, new_width = _get_patch_output_size(image, target_resolution, input_data_format)
+
+        # Resize the image
+        resized_image = resize(image, (new_height, new_width), resample=resample, input_data_format=input_data_format)
+
+        return resized_image
+
+    def _pad_for_patching(
+        self, image: np.array, target_resolution: tuple, input_data_format: ChannelDimension
+    ) -> np.array:
+        """
+        Pad an image to a target resolution while maintaining aspect ratio.
+        """
+        target_height, target_width = target_resolution
+        new_height, new_width = _get_patch_output_size(image, target_resolution, input_data_format)
+
+        paste_x = (target_width - new_width) // 2
+        paste_y = (target_height - new_height) // 2
+
+        padded_image = pad(image, padding=((paste_y, paste_y), (paste_x, paste_x)))
+
+        return padded_image
+
+    def get_image_patches(
+        self,
+        image: np.array,
+        grid_pinpoints,
+        size: tuple,
+        patch_size: int,
+        resample: PILImageResampling,
+        data_format: ChannelDimension,
+        input_data_format: ChannelDimension,
+    ) -> List[np.array]:
+        """
+        Process an image with variable resolutions by dividing it into patches.
+
+        Args:
+            image (np.array):
+                The input image to be processed.
+            grid_pinpoints (List):
+                A string representation of a list of possible resolutions.
+            size (`tuple`):
+                Size to resize the original image to.
+            patch_size (`int`):
+                Size of the patches to divide the image into.
+            resample (`PILImageResampling`):
+                Resampling filter to use if resizing the image.
+            data_format (`ChannelDimension` or `str`):
+                The channel dimension format for the output image.
+            input_data_format (`ChannelDimension` or `str`):
+                The channel dimension format of the input image.
+
+        Returns:
+            List[np.array]: A list of NumPy arrays containing the processed image patches.
+        """
+        if not isinstance(grid_pinpoints, list):
+            raise ValueError("grid_pinpoints must be a list of possible resolutions.")
+
+        possible_resolutions = grid_pinpoints
+
+        image_size = get_image_size(image, channel_dim=input_data_format)
+        best_resolution = select_best_resolution(image_size, possible_resolutions)
+        resized_image = self._resize_for_patching(
+            image, best_resolution, resample=resample, input_data_format=input_data_format
+        )
+        padded_image = self._pad_for_patching(resized_image, best_resolution, input_data_format=input_data_format)
+
+        patches = divide_to_patches(padded_image, patch_size=patch_size, input_data_format=input_data_format)
+
+        # make sure that all patches are in the input data format
+        patches = [
+            to_channel_dimension_format(patch, channel_dim=data_format, input_channel_dim=input_data_format)
+            for patch in patches
+        ]
+
+        resized_original_image = resize(
+            image,
+            size=size,
+            resample=resample,
+            data_format=data_format,
+            input_data_format=input_data_format,
+        )
+
+        image_patches = [resized_original_image] + patches
+
+        return image_patches
+
+    def preprocess(
+        self,
+        images: ImageInput,
+        do_resize: bool = None,
+        size: Dict[str, int] = None,
+        image_grid_pinpoints: List = None,
+        resample: PILImageResampling = None,
+        do_center_crop: bool = None,
+        crop_size: int = None,
+        do_rescale: bool = None,
+        rescale_factor: float = None,
+        do_normalize: bool = None,
+        image_mean: Optional[Union[float, List[float]]] = None,
+        image_std: Optional[Union[float, List[float]]] = None,
+        do_convert_rgb: bool = None,
+        return_tensors: Optional[Union[str, TensorType]] = None,
+        data_format: Optional[ChannelDimension] = ChannelDimension.FIRST,
+        input_data_format: Optional[Union[str, ChannelDimension]] = None,
+    ):
+        """
+        Args:
+            images (`ImageInput`):
+                Image to preprocess. Expects a single or batch of images with pixel values ranging from 0 to 255. If
+                passing in images with pixel values between 0 and 1, set `do_rescale=False`.
+            do_resize (`bool`, *optional*, defaults to `self.do_resize`):
+                Whether to resize the image.
+            size (`Dict[str, int]`, *optional*, defaults to `self.size`):
+                Size of the image after resizing. Shortest edge of the image is resized to size["shortest_edge"], with
+                the longest edge resized to keep the input aspect ratio.
+            image_grid_pinpoints (`List` *optional*, defaults to `self.image_grid_pinpoints`):
+                A list of possible resolutions to use for processing high resolution images. The best resolution is
+                selected based on the original size of the image.
+            resample (`int`, *optional*, defaults to `self.resample`):
+                Resampling filter to use if resizing the image. This can be one of the enum `PILImageResampling`. Only
+                has an effect if `do_resize` is set to `True`.
+            do_center_crop (`bool`, *optional*, defaults to `self.do_center_crop`):
+                Whether to center crop the image.
+            crop_size (`Dict[str, int]`, *optional*, defaults to `self.crop_size`):
+                Size of the center crop. Only has an effect if `do_center_crop` is set to `True`.
+            do_rescale (`bool`, *optional*, defaults to `self.do_rescale`):
+                Whether to rescale the image.
+            rescale_factor (`float`, *optional*, defaults to `self.rescale_factor`):
+                Rescale factor to rescale the image by if `do_rescale` is set to `True`.
+            do_normalize (`bool`, *optional*, defaults to `self.do_normalize`):
+                Whether to normalize the image.
+            image_mean (`float` or `List[float]`, *optional*, defaults to `self.image_mean`):
+                Image mean to use for normalization. Only has an effect if `do_normalize` is set to `True`.
+            image_std (`float` or `List[float]`, *optional*, defaults to `self.image_std`):
+                Image standard deviation to use for normalization. Only has an effect if `do_normalize` is set to
+                `True`.
+            do_convert_rgb (`bool`, *optional*, defaults to `self.do_convert_rgb`):
+                Whether to convert the image to RGB.
+            return_tensors (`str` or `TensorType`, *optional*):
+                The type of tensors to return. Can be one of:
+                - Unset: Return a list of `np.ndarray`.
+                - `TensorType.TENSORFLOW` or `'tf'`: Return a batch of type `tf.Tensor`.
+                - `TensorType.PYTORCH` or `'pt'`: Return a batch of type `torch.Tensor`.
+                - `TensorType.NUMPY` or `'np'`: Return a batch of type `np.ndarray`.
+                - `TensorType.JAX` or `'jax'`: Return a batch of type `jax.numpy.ndarray`.
+            data_format (`ChannelDimension` or `str`, *optional*, defaults to `ChannelDimension.FIRST`):
+                The channel dimension format for the output image. Can be one of:
+                - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format.
+                - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format.
+                - Unset: Use the channel dimension format of the input image.
+            input_data_format (`ChannelDimension` or `str`, *optional*):
+                The channel dimension format for the input image. If unset, the channel dimension format is inferred
+                from the input image. Can be one of:
+                - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format.
+                - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format.
+                - `"none"` or `ChannelDimension.NONE`: image in (height, width) format.
+        """
+        do_resize = do_resize if do_resize is not None else self.do_resize
+        size = size if size is not None else self.size
+        size = get_size_dict(size, param_name="size", default_to_square=False)
+        image_grid_pinpoints = image_grid_pinpoints if image_grid_pinpoints is not None else self.image_grid_pinpoints
+        resample = resample if resample is not None else self.resample
+        do_center_crop = do_center_crop if do_center_crop is not None else self.do_center_crop
+        crop_size = crop_size if crop_size is not None else self.crop_size
+        crop_size = get_size_dict(crop_size, param_name="crop_size", default_to_square=True)
+        do_rescale = do_rescale if do_rescale is not None else self.do_rescale
+        rescale_factor = rescale_factor if rescale_factor is not None else self.rescale_factor
+        do_normalize = do_normalize if do_normalize is not None else self.do_normalize
+        image_mean = image_mean if image_mean is not None else self.image_mean
+        image_std = image_std if image_std is not None else self.image_std
+        do_convert_rgb = do_convert_rgb if do_convert_rgb is not None else self.do_convert_rgb
+
+        images = make_list_of_images(images)
+
+        if not valid_images(images):
+            raise ValueError(
+                "Invalid image type. Must be of type PIL.Image.Image, numpy.ndarray, "
+                "torch.Tensor, tf.Tensor or jax.ndarray."
+            )
+
+        validate_preprocess_arguments(
+            do_rescale=do_rescale,
+            rescale_factor=rescale_factor,
+            do_normalize=do_normalize,
+            image_mean=image_mean,
+            image_std=image_std,
+            do_center_crop=do_center_crop,
+            crop_size=crop_size,
+            do_resize=do_resize,
+            size=size,
+            resample=resample,
+        )
+
+        if do_convert_rgb:
+            images = [convert_to_rgb(image) for image in images]
+
+        # All transformations expect numpy arrays.
+        images = [to_numpy_array(image) for image in images]
+
+        if is_scaled_image(images[0]) and do_rescale:
+            logger.warning_once(
+                "It looks like you are trying to rescale already rescaled images. If the input"
+                " images have pixel values between 0 and 1, set `do_rescale=False` to avoid rescaling them again."
+            )
+
+        if input_data_format is None:
+            # We assume that all images have the same channel dimension format.
+            input_data_format = infer_channel_dimension_format(images[0])
+
+        new_images = []
+        image_sizes = [get_image_size(image, channel_dim=input_data_format) for image in images]
+        for image in images:
+            # convert image into a list of patches
+            # we intentially use the same data format as the input data format
+            image_patches = self.get_image_patches(
+                image,
+                image_grid_pinpoints,
+                size=(size["shortest_edge"], size["shortest_edge"]),
+                patch_size=crop_size["height"],
+                resample=resample,
+                data_format=input_data_format,
+                input_data_format=input_data_format,
+            )
+
+            # preprocess patches
+            pixel_values = self._preprocess(
+                image_patches,
+                do_resize=do_resize,
+                size=size,
+                resample=resample,
+                do_center_crop=do_center_crop,
+                crop_size=crop_size,
+                do_rescale=do_rescale,
+                rescale_factor=rescale_factor,
+                do_normalize=do_normalize,
+                image_mean=image_mean,
+                image_std=image_std,
+                data_format=data_format,
+                input_data_format=input_data_format,
+            )
+            pixel_values = np.array(pixel_values)
+            new_images.append(pixel_values)
+
+        data = {"pixel_values": new_images, "image_sizes": image_sizes}
+
+        return BatchFeature(data=data, tensor_type=return_tensors)

src/transformers/models/llava_next/modeling_llava_next.py

@@ -0,0 +1,697 @@
+# coding=utf-8
+# Copyright 2024 the HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+""" PyTorch Llava-NeXT model."""
+
+from dataclasses import dataclass
+from typing import List, Optional, Tuple, Union
+
+import torch
+import torch.utils.checkpoint
+from torch import nn
+
+from ... import PreTrainedModel
+from ...activations import ACT2FN
+from ...cache_utils import Cache
+from ...image_processing_utils import select_best_resolution
+from ...modeling_outputs import ModelOutput
+from ...utils import (
+    add_start_docstrings,
+    add_start_docstrings_to_model_forward,
+    logging,
+    replace_return_docstrings,
+)
+from ..auto import AutoModel, AutoModelForCausalLM
+from .configuration_llava_next import LlavaNextConfig
+
+
+logger = logging.get_logger(__name__)
+
+_CONFIG_FOR_DOC = "LlavaNextConfig"
+
+LLAVA_NEXT_PRETRAINED_MODEL_ARCHIVE_LIST = [
+    "llava-hf/llava-v1.6-mistral-7b-hf",
+    # See all LLaVA-NeXT models at https://huggingface.co/models?filter=llava_next
+]
+
+
+def get_anyres_image_grid_shape(image_size, grid_pinpoints, patch_size):
+    """
+    Calculate the shape of the image patch grid after the preprocessing for images of any resolution.
+
+    Args:
+        image_size (`tuple`):
+            The size of the input image in the format (width, height).
+        grid_pinpoints (`List`):
+            A list containing possible resolutions. Each item in the list should be a tuple or list
+            of the form `(height, width)`.
+        patch_size (`int`):
+            The size of each image patch.
+
+    Returns:
+        tuple: The shape of the image patch grid in the format (width, height).
+    """
+    if not isinstance(grid_pinpoints, list):
+        raise ValueError("grid_pinpoints should be a list of tuples or lists")
+
+    height, width = select_best_resolution(image_size, grid_pinpoints)
+    return height // patch_size, width // patch_size
+
+
+def unpad_image(tensor, original_size):
+    """
+    Unpads a PyTorch tensor of a padded and resized image.
+
+    Args:
+        tensor (`torch.Tensor`):
+            The image tensor, assumed to be of shape (num_channels, height, width).
+        original_size (`tuple`):
+            The original size of the image (height, width).
+
+    Returns:
+        `torch.Tensor`: The unpadded image tensor.
+    """
+    original_height, original_width = original_size
+    current_height, current_width = tensor.shape[1:]
+
+    original_aspect_ratio = original_width / original_height
+    current_aspect_ratio = current_width / current_height
+
+    if original_aspect_ratio > current_aspect_ratio:
+        scale_factor = current_width / original_width
+        new_height = int(original_height * scale_factor)
+        padding = (current_height - new_height) // 2
+        unpadded_tensor = tensor[:, padding : current_height - padding, :]
+    else:
+        scale_factor = current_height / original_height
+        new_width = int(original_width * scale_factor)
+        padding = (current_width - new_width) // 2
+        unpadded_tensor = tensor[:, :, padding : current_width - padding]
+
+    return unpadded_tensor
+
+
+@dataclass
+# Copied from transformers.models.idefics.modeling_idefics.IdeficsCausalLMOutputWithPast with Idefics->LlavaNext
+class LlavaNextCausalLMOutputWithPast(ModelOutput):
+    """
+    Base class for LlavaNext causal language model (or autoregressive) outputs.
+
+    Args:
+        loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
+            Language modeling loss (for next-token prediction).
+        logits (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`):
+            Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
+            `(batch_size, num_heads, sequence_length, embed_size_per_head)`)
+
+            Contains pre-computed hidden-states (key and values in the self-attention blocks) that can be used (see
+            `past_key_values` input) to speed up sequential decoding.
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
+            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.
+
+            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
+        attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
+            Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
+            sequence_length)`.
+
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
+            heads.
+        image_hidden_states (`tuple(torch.FloatTensor)`, *optional*):
+            Tuple of `torch.FloatTensor` (one for the output of the image embeddings, `(batch_size, num_images,
+            sequence_length, hidden_size)`.
+
+            image_hidden_states of the model produced by the vision encoder, and optionally by the perceiver
+    """
+
+    loss: Optional[torch.FloatTensor] = None
+    logits: torch.FloatTensor = None
+    past_key_values: Optional[List[torch.FloatTensor]] = None
+    hidden_states: Optional[Tuple[torch.FloatTensor]] = None
+    attentions: Optional[Tuple[torch.FloatTensor]] = None
+    image_hidden_states: Optional[Tuple[torch.FloatTensor]] = None
+
+
+# Copied from transformers.models.llava.modeling_llava.LlavaMultiModalProjector with Llava->LlavaNext
+class LlavaNextMultiModalProjector(nn.Module):
+    def __init__(self, config: LlavaNextConfig):
+        super().__init__()
+
+        self.linear_1 = nn.Linear(config.vision_config.hidden_size, config.text_config.hidden_size, bias=True)
+        self.act = ACT2FN[config.projector_hidden_act]
+        self.linear_2 = nn.Linear(config.text_config.hidden_size, config.text_config.hidden_size, bias=True)
+
+    def forward(self, image_features):
+        hidden_states = self.linear_1(image_features)
+        hidden_states = self.act(hidden_states)
+        hidden_states = self.linear_2(hidden_states)
+        return hidden_states
+
+
+LLAVA_NEXT_START_DOCSTRING = r"""
+    This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the
+    library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
+    etc.)
+
+    This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
+    Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
+    and behavior.
+
+    Parameters:
+        config ([`LlavaNextConfig`] or [`LlavaNextVisionConfig`]):
+            Model configuration class with all the parameters of the model. Initializing with a config file does not
+            load the weights associated with the model, only the configuration. Check out the
+            [`~PreTrainedModel.from_pretrained`] method to load the model weights.
+"""
+
+
+@add_start_docstrings(
+    "The bare LLaMA Model outputting raw hidden-states without any specific head on top.",
+    LLAVA_NEXT_START_DOCSTRING,
+)
+# Copied from transformers.models.llava.modeling_llava.LlavaPreTrainedModel with Llava->LlavaNext,llava->llava_next
+class LlavaNextPreTrainedModel(PreTrainedModel):
+    config_class = LlavaNextConfig
+    base_model_prefix = "model"
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["LlavaNextVisionAttention"]
+    _skip_keys_device_placement = "past_key_values"
+    _supports_flash_attn_2 = True
+
+    def _init_weights(self, module):
+        # important: this ported version of LlavaNext isn't meant for training from scratch - only
+        # inference and fine-tuning - so the proper init weights code has been removed - the original codebase
+        # https://github.com/haotian-liu/LLaVA/tree/main/llava_next should serve for that purpose
+        std = (
+            self.config.initializer_range
+            if hasattr(self.config, "initializer_range")
+            else self.config.text_config.initializer_range
+        )
+
+        if hasattr(module, "class_embedding"):
+            module.class_embedding.data.normal_(mean=0.0, std=std)
+
+        if isinstance(module, (nn.Linear, nn.Conv2d)):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+
+    @property
+    def _supports_sdpa(self):
+        """
+        Retrieve language_model's attribute to check whether the model supports
+        SDPA or not.
+        """
+        return self.language_model._supports_sdpa
+
+
+LLAVA_NEXT_INPUTS_DOCSTRING = r"""
+    Args:
+        input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):
+            Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide
+            it.
+
+            Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
+            [`PreTrainedTokenizer.__call__`] for details.
+
+            [What are input IDs?](../glossary#input-ids)
+        pixel_values (`torch.FloatTensor` of shape `(batch_size, num_channels, image_size, image_size)):
+            The tensors corresponding to the input images. Pixel values can be obtained using
+            [`AutoImageProcessor`]. See [`LlavaNextImageProcessor.__call__`] for details. [`LlavaProcessor`] uses
+            [`LlavaNextImageProcessor`] for processing images.
+        image_sizes (`torch.LongTensor` of shape `(batch_size, 2)`, *optional*):
+            The sizes of the images in the batch, being (height, width) for each image.
+        attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:
+
+            - 1 for tokens that are **not masked**,
+            - 0 for tokens that are **masked**.
+
+            [What are attention masks?](../glossary#attention-mask)
+
+            Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
+            [`PreTrainedTokenizer.__call__`] for details.
+
+            If `past_key_values` is used, optionally only the last `decoder_input_ids` have to be input (see
+            `past_key_values`).
+
+            If you want to change padding behavior, you should read [`modeling_opt._prepare_decoder_attention_mask`]
+            and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more
+            information on the default strategy.
+
+            - 1 indicates the head is **not masked**,
+            - 0 indicates the head is **masked**.
+        position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
+            config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids)
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
+            `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of shape
+            `(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`.
+
+            Contains pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention
+            blocks) that can be used (see `past_key_values` input) to speed up sequential decoding.
+
+            If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
+            don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
+            `decoder_input_ids` of shape `(batch_size, sequence_length)`.
+        inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
+            Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This
+            is useful if you want more control over how to convert `input_ids` indices into associated vectors than the
+            model's internal embedding lookup matrix.
+        vision_feature_layer (`int`, *optional*, defaults to -2):
+            The index of the layer to select the vision feature.
+        vision_feature_select_strategy (`str`, *optional*, defaults to `"default"`):
+            The feature selection strategy used to select the vision feature from the vision backbone.
+            Can be one of `"default"` or `"full"`. If `"default"`, the CLS token is removed from the vision features.
+            If `"full"`, the full vision features are used.
+        use_cache (`bool`, *optional*):
+            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
+            `past_key_values`).
+        output_attentions (`bool`, *optional*):
+            Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned
+            tensors for more detail.
+        output_hidden_states (`bool`, *optional*):
+            Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
+            more detail.
+        return_dict (`bool`, *optional*):
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
+"""
+
+
+@add_start_docstrings(
+    """The LLAVA-NeXT model which consists of a vision backbone and a language model.""",
+    LLAVA_NEXT_START_DOCSTRING,
+)
+class LlavaNextForConditionalGeneration(LlavaNextPreTrainedModel):
+    def __init__(self, config: LlavaNextConfig):
+        super().__init__(config)
+        self.vision_tower = AutoModel.from_config(config.vision_config)
+
+        self.multi_modal_projector = LlavaNextMultiModalProjector(config)
+
+        self.image_newline = nn.Parameter(torch.empty(config.text_config.hidden_size, dtype=self.dtype))
+
+        self.vocab_size = config.text_config.vocab_size
+        self.language_model = AutoModelForCausalLM.from_config(
+            config.text_config, attn_implementation=config._attn_implementation
+        )
+        self.pad_token_id = self.config.pad_token_id if self.config.pad_token_id is not None else -1
+        self.post_init()
+
+    # Copied from transformers.models.llava.modeling_llava.LlavaForConditionalGeneration.get_input_embeddings
+    def get_input_embeddings(self):
+        return self.language_model.get_input_embeddings()
+
+    # Copied from transformers.models.llava.modeling_llava.LlavaForConditionalGeneration.set_input_embeddings
+    def set_input_embeddings(self, value):
+        self.language_model.set_input_embeddings(value)
+
+    # Copied from transformers.models.llava.modeling_llava.LlavaForConditionalGeneration.get_output_embeddings
+    def get_output_embeddings(self):
+        return self.language_model.get_output_embeddings()
+
+    # Copied from transformers.models.llava.modeling_llava.LlavaForConditionalGeneration.set_output_embeddings
+    def set_output_embeddings(self, new_embeddings):
+        self.language_model.set_output_embeddings(new_embeddings)
+
+    # Copied from transformers.models.llava.modeling_llava.LlavaForConditionalGeneration.set_decoder
+    def set_decoder(self, decoder):
+        self.language_model.set_decoder(decoder)
+
+    # Copied from transformers.models.llava.modeling_llava.LlavaForConditionalGeneration.get_decoder
+    def get_decoder(self):
+        return self.language_model.get_decoder()
+
+    # Copied from transformers.models.llava.modeling_llava.LlavaForConditionalGeneration.tie_weights
+    def tie_weights(self):
+        return self.language_model.tie_weights()
+
+    # Copied from transformers.models.llava.modeling_llava.LlavaForConditionalGeneration.resize_token_embeddings
+    def resize_token_embeddings(self, new_num_tokens: Optional[int] = None, pad_to_multiple_of=None) -> nn.Embedding:
+        model_embeds = self.language_model.resize_token_embeddings(new_num_tokens, pad_to_multiple_of)
+        # update vocab size
+        self.config.text_config.vocab_size = model_embeds.num_embeddings
+        self.vocab_size = model_embeds.num_embeddings
+        return model_embeds
+
+    # Copied from transformers.models.llava.modeling_llava.LlavaForConditionalGeneration._merge_input_ids_with_image_features
+    def _merge_input_ids_with_image_features(self, image_features, inputs_embeds, input_ids, attention_mask, labels):
+        num_images, num_image_patches, embed_dim = image_features.shape
+        batch_size, sequence_length = input_ids.shape
+        left_padding = not torch.sum(input_ids[:, -1] == torch.tensor(self.pad_token_id))
+        # 1. Create a mask to know where special image tokens are
+        special_image_token_mask = input_ids == self.config.image_token_index
+        num_special_image_tokens = torch.sum(special_image_token_mask, dim=-1)
+        # Compute the maximum embed dimension
+        max_embed_dim = (num_special_image_tokens.max() * (num_image_patches - 1)) + sequence_length
+        batch_indices, non_image_indices = torch.where(input_ids != self.config.image_token_index)
+
+        # 2. Compute the positions where text should be written
+        # Calculate new positions for text tokens in merged image-text sequence.
+        # `special_image_token_mask` identifies image tokens. Each image token will be replaced by `nb_text_tokens_per_images - 1` text tokens.
+        # `torch.cumsum` computes how each image token shifts subsequent text token positions.
+        # - 1 to adjust for zero-based indexing, as `cumsum` inherently increases indices by one.
+        new_token_positions = torch.cumsum((special_image_token_mask * (num_image_patches - 1) + 1), -1) - 1
+        nb_image_pad = max_embed_dim - 1 - new_token_positions[:, -1]
+        if left_padding:
+            new_token_positions += nb_image_pad[:, None]  # offset for left padding
+        text_to_overwrite = new_token_positions[batch_indices, non_image_indices]
+
+        # 3. Create the full embedding, already padded to the maximum position
+        final_embedding = torch.zeros(
+            batch_size, max_embed_dim, embed_dim, dtype=inputs_embeds.dtype, device=inputs_embeds.device
+        )
+        final_attention_mask = torch.zeros(
+            batch_size, max_embed_dim, dtype=attention_mask.dtype, device=inputs_embeds.device
+        )
+        if labels is not None:
+            final_labels = torch.full(
+                (batch_size, max_embed_dim), self.config.ignore_index, dtype=input_ids.dtype, device=input_ids.device
+            )
+        # In case the Vision model or the Language model has been offloaded to CPU, we need to manually
+        # set the corresponding tensors into their correct target device.
+        target_device = inputs_embeds.device
+        batch_indices, non_image_indices, text_to_overwrite = (
+            batch_indices.to(target_device),
+            non_image_indices.to(target_device),
+            text_to_overwrite.to(target_device),
+        )
+        attention_mask = attention_mask.to(target_device)
+
+        # 4. Fill the embeddings based on the mask. If we have ["hey" "<image>", "how", "are"]
+        # we need to index copy on [0, 577, 578, 579] for the text and [1:576] for the image features
+        final_embedding[batch_indices, text_to_overwrite] = inputs_embeds[batch_indices, non_image_indices]
+        final_attention_mask[batch_indices, text_to_overwrite] = attention_mask[batch_indices, non_image_indices]
+        if labels is not None:
+            final_labels[batch_indices, text_to_overwrite] = labels[batch_indices, non_image_indices]
+
+        # 5. Fill the embeddings corresponding to the images. Anything that is still zeros needs filling
+        image_to_overwrite = torch.all(final_embedding == 0, dim=-1)
+        image_to_overwrite &= image_to_overwrite.cumsum(-1) - 1 >= nb_image_pad[:, None].to(target_device)
+
+        if image_to_overwrite.sum() != image_features.shape[:-1].numel():
+            raise ValueError(
+                f"The input provided to the model are wrong. The number of image tokens is {torch.sum(special_image_token_mask)} while"
+                f" the number of image given to the model is {num_images}. This prevents correct indexing and breaks batch generation."
+            )
+
+        final_embedding[image_to_overwrite] = image_features.contiguous().reshape(-1, embed_dim).to(target_device)
+        final_attention_mask |= image_to_overwrite
+        position_ids = (final_attention_mask.cumsum(-1) - 1).masked_fill_((final_attention_mask == 0), 1)
+
+        # 6. Mask out the embedding at padding positions, as we later use the past_key_value value to determine the non-attended tokens.
+        batch_indices, pad_indices = torch.where(input_ids == self.pad_token_id)
+        indices_to_mask = new_token_positions[batch_indices, pad_indices]
+
+        final_embedding[batch_indices, indices_to_mask] = 0
+
+        if labels is None:
+            final_labels = None
+
+        return final_embedding, final_attention_mask, final_labels, position_ids
+
+    @add_start_docstrings_to_model_forward(LLAVA_NEXT_INPUTS_DOCSTRING)
+    @replace_return_docstrings(output_type=LlavaNextCausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC)
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        pixel_values: torch.FloatTensor = None,
+        image_sizes: Optional[torch.LongTensor] = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        vision_feature_layer: Optional[int] = None,
+        vision_feature_select_strategy: Optional[str] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Tuple, LlavaNextCausalLMOutputWithPast]:
+        r"""
+        Args:
+            labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+                Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
+                config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
+                (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+
+        Returns:
+
+        Example:
+
+        ```python
+        >>> from PIL import Image
+        >>> import requests
+        >>> from transformers import AutoProcessor, LlavaNextForConditionalGeneration
+
+        >>> model = LlavaNextForConditionalGeneration.from_pretrained("llava-hf/llava-v1.6-mistral-7b-hf")
+        >>> processor = AutoProcessor.from_pretrained("llava-hf/llava-v1.6-mistral-7b-hf")
+
+        >>> prompt = "[INST] <image>\nWhat is shown in this image? [/INST]"
+        >>> url = "https://www.ilankelman.org/stopsigns/australia.jpg"
+        >>> image = Image.open(requests.get(url, stream=True).raw)
+
+        >>> inputs = processor(text=prompt, images=image, return_tensors="pt")
+
+        >>> # Generate
+        >>> generate_ids = model.generate(**inputs, max_length=30)
+        >>> processor.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
+        "[INST]  \nWhat is shown in this image? [/INST] The image appears to be a radar chart, which is a type of multi-dimensional plot (...)"
+        ```"""
+
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+        vision_feature_layer = (
+            vision_feature_layer if vision_feature_layer is not None else self.config.vision_feature_layer
+        )
+        vision_feature_select_strategy = (
+            vision_feature_select_strategy
+            if vision_feature_select_strategy is not None
+            else self.config.vision_feature_select_strategy
+        )
+
+        if inputs_embeds is None:
+            # 1. Extract the input embeddings
+            inputs_embeds = self.get_input_embeddings()(input_ids)
+
+            # 2. Merge text and images
+            if pixel_values is not None and input_ids.shape[1] != 1:
+                batch_size, num_patches, num_channels, height, width = pixel_values.shape
+                reshaped_pixel_values = pixel_values.view(batch_size * num_patches, num_channels, height, width)
+                image_features = self.vision_tower(reshaped_pixel_values, output_hidden_states=True)
+
+                selected_image_feature = image_features.hidden_states[vision_feature_layer]
+
+                if vision_feature_select_strategy == "default":
+                    selected_image_feature = selected_image_feature[:, 1:]
+                elif vision_feature_select_strategy == "full":
+                    selected_image_feature = selected_image_feature
+
+                image_features = self.multi_modal_projector(selected_image_feature)
+
+                # split up image_features for each of the individual images
+                # hence we get a list of image_features, each of shape (5, num_patches, hidden_size)
+                # if we assume each image has 5 image features (base image + 4 patches)
+                split_sizes = [image.shape[0] for image in pixel_values]
+                image_features = torch.split(image_features, split_sizes, dim=0)
+
+                # NOTE we only support multimodal_patch_merge_type == "spatial_unpad"
+                height = width = self.config.vision_config.image_size // self.config.vision_config.patch_size
+
+                new_image_features = []
+                for image_idx, image_feature in enumerate(image_features):
+                    if image_feature.shape[0] > 1:
+                        base_image_feature = image_feature[0]
+                        image_feature = image_feature[1:]
+
+                        if height * width != base_image_feature.shape[0]:
+                            raise ValueError("The number of patches is not consistent with the image size.")
+                        num_patch_height, num_patch_width = get_anyres_image_grid_shape(
+                            image_sizes[image_idx],
+                            self.config.image_grid_pinpoints,
+                            self.config.vision_config.image_size,
+                        )
+                        image_feature = image_feature.view(num_patch_height, num_patch_width, height, width, -1)
+                        image_feature = image_feature.permute(4, 0, 2, 1, 3).contiguous()
+                        image_feature = image_feature.flatten(1, 2).flatten(2, 3)
+                        image_feature = unpad_image(image_feature, image_sizes[image_idx])
+                        image_feature = torch.cat(
+                            (
+                                image_feature,
+                                self.image_newline[:, None, None].expand(*image_feature.shape[:-1], 1),
+                            ),
+                            dim=-1,
+                        )
+                        image_feature = image_feature.flatten(1, 2).transpose(0, 1)
+                        image_feature = torch.cat((base_image_feature, image_feature), dim=0)
+                    else:
+                        image_feature = image_feature[0]
+                        image_feature = torch.cat((image_feature, self.image_newline[None]), dim=0)
+                    new_image_features.append(image_feature)
+                image_features = torch.stack(new_image_features, dim=0)
+
+                inputs_embeds, attention_mask, labels, position_ids = self._merge_input_ids_with_image_features(
+                    image_features, inputs_embeds, input_ids, attention_mask, labels
+                )
+                if labels is None:
+                    labels = torch.full_like(attention_mask, self.config.ignore_index).to(torch.long)
+
+            # In case input_ids.shape[1] == 1 & pixel_values==None & past_key_values != None, we are in the case of
+            # generation with cache
+            elif past_key_values is not None and pixel_values is not None and input_ids.shape[1] == 1:
+                # Retrieve the first layer to inspect the logits and mask out the hidden states
+                # that are set to 0
+                first_layer_past_key_value = past_key_values[0][0][:, :, :, 0]
+
+                # Sum all dimensions of head_dim (-2) to avoid random errors such as: https://github.com/huggingface/transformers/pull/28032#issuecomment-1863691941
+                batch_index, non_attended_tokens = torch.where(first_layer_past_key_value.float().sum(-2) == 0)
+
+                # Get the target length
+                target_seqlen = first_layer_past_key_value.shape[-1] + 1
+
+                extended_attention_mask = torch.ones(
+                    (attention_mask.shape[0], target_seqlen - attention_mask.shape[1]),
+                    dtype=attention_mask.dtype,
+                    device=attention_mask.device,
+                )
+
+                # Filter out only the tokens that can be un-attended, this can happen
+                # if one uses Llava + Fused modules where the cache on the
+                # first iteration is already big enough, or if one passes custom cache
+                valid_indices = non_attended_tokens < extended_attention_mask.size(-1)
+                new_batch_index = batch_index[valid_indices]
+                new_non_attended_tokens = non_attended_tokens[valid_indices]
+
+                # Zero-out the places where we don't need to attend
+                extended_attention_mask[new_batch_index, new_non_attended_tokens] = 0
+
+                attention_mask = torch.cat((attention_mask, extended_attention_mask), dim=1)
+                position_ids = torch.sum(attention_mask, dim=1).unsqueeze(-1) - 1
+
+        outputs = self.language_model(
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+
+        logits = outputs[0]
+
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            if attention_mask is not None:
+                shift_attention_mask = attention_mask[..., 1:]
+                shift_logits = logits[..., :-1, :][shift_attention_mask.to(logits.device) != 0].contiguous()
+                shift_labels = labels[..., 1:][shift_attention_mask.to(labels.device) != 0].contiguous()
+            else:
+                shift_logits = logits[..., :-1, :].contiguous()
+                shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = nn.CrossEntropyLoss()
+            loss = loss_fct(
+                shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1).to(shift_logits.device)
+            )
+
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            return (loss,) + output if loss is not None else output
+
+        return LlavaNextCausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )
+
+    def prepare_inputs_for_generation(
+        self,
+        input_ids,
+        past_key_values=None,
+        inputs_embeds=None,
+        pixel_values=None,
+        image_sizes=None,
+        attention_mask=None,
+        **kwargs,
+    ):
+        if past_key_values is not None:
+            if isinstance(past_key_values, Cache):
+                cache_length = past_key_values.get_seq_length()
+                past_length = past_key_values.seen_tokens
+            else:
+                cache_length = past_length = past_key_values[0][0].shape[2]
+
+            # Keep only the unprocessed tokens:
+            # 1 - If the length of the attention_mask exceeds the length of input_ids, then we are in a setting where
+            # some of the inputs are exclusively passed as part of the cache (e.g. when passing input_embeds as
+            # input)
+            if attention_mask is not None and attention_mask.shape[1] > input_ids.shape[1]:
+                input_ids = input_ids[:, -(attention_mask.shape[1] - past_length) :]
+            # 2 - If the past_length is smaller than input_ids', then input_ids holds all input tokens. We can discard
+            # input_ids based on the past_length.
+            elif past_length < input_ids.shape[1]:
+                input_ids = input_ids[:, past_length:]
+            # 3 - Otherwise (past_length >= input_ids.shape[1]), let's assume input_ids only has unprocessed tokens.
+            elif self.config.image_token_index in input_ids:
+                input_ids = input_ids[:, input_ids.shape[1] - 1 :]
+            # If the cache has seen more tokens than it can hold, then the cache has a size limit. Let's discard the
+            # older attention values, as their corresponding values are not part of the input.
+            if cache_length < past_length and attention_mask is not None:
+                attention_mask = attention_mask[:, -(cache_length + input_ids.shape[1]) :]
+
+        position_ids = kwargs.get("position_ids", None)
+        if attention_mask is not None and position_ids is None:
+            # create position_ids on the fly for batch generation
+            position_ids = attention_mask.long().cumsum(-1) - 1
+            position_ids.masked_fill_(attention_mask == 0, 1)
+            if past_key_values:
+                position_ids = position_ids[:, -input_ids.shape[1] :]
+
+        # if `inputs_embeds` are passed, we only want to use them in the 1st generation step
+        if inputs_embeds is not None and past_key_values is None:
+            model_inputs = {"inputs_embeds": inputs_embeds}
+        else:
+            model_inputs = {"input_ids": input_ids}
+
+        model_inputs.update(
+            {
+                "position_ids": position_ids,
+                "past_key_values": past_key_values,
+                "use_cache": kwargs.get("use_cache"),
+                "attention_mask": attention_mask,
+                "pixel_values": pixel_values,
+                "image_sizes": image_sizes,
+            }
+        )
+        return model_inputs
+
+    # Copied from transformers.models.llava.modeling_llava.LlavaForConditionalGeneration._reorder_cache
+    def _reorder_cache(self, *args, **kwargs):
+        return self.language_model._reorder_cache(*args, **kwargs)

src/transformers/models/llava_next/processing_llava_next.py

@@ -0,0 +1,135 @@
+# coding=utf-8
+# Copyright 2024 The HuggingFace Inc. team.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+Processor class for LLaVa-NeXT.
+"""
+
+
+from typing import List, Optional, Union
+
+from ...feature_extraction_utils import BatchFeature
+from ...image_utils import ImageInput
+from ...processing_utils import ProcessorMixin
+from ...tokenization_utils_base import PaddingStrategy, PreTokenizedInput, TextInput, TruncationStrategy
+from ...utils import TensorType
+
+
+class LlavaNextProcessor(ProcessorMixin):
+    r"""
+    Constructs a LLaVa-NeXT processor which wraps a LLaVa-NeXT image processor and a LLaMa tokenizer into a single processor.
+
+    [`LlavaNextProcessor`] offers all the functionalities of [`LlavaNextImageProcessor`] and [`LlamaTokenizerFast`]. See the
+    [`~LlavaNextProcessor.__call__`] and [`~LlavaNextProcessor.decode`] for more information.
+
+    Args:
+        image_processor ([`LlavaNextImageProcessor`], *optional*):
+            The image processor is a required input.
+        tokenizer ([`LlamaTokenizerFast`], *optional*):
+            The tokenizer is a required input.
+    """
+
+    attributes = ["image_processor", "tokenizer"]
+    image_processor_class = "LlavaNextImageProcessor"
+    tokenizer_class = ("LlamaTokenizer", "LlamaTokenizerFast")
+
+    def __init__(self, image_processor=None, tokenizer=None):
+        super().__init__(image_processor, tokenizer)
+
+    def __call__(
+        self,
+        text: Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]],
+        images: ImageInput = None,
+        padding: Union[bool, str, PaddingStrategy] = False,
+        truncation: Union[bool, str, TruncationStrategy] = None,
+        max_length=None,
+        return_tensors: Optional[Union[str, TensorType]] = TensorType.PYTORCH,
+    ) -> BatchFeature:
+        """
+        Main method to prepare for the model one or several sequences(s) and image(s). This method forwards the `text`
+        and `kwargs` arguments to LlamaTokenizerFast's [`~LlamaTokenizerFast.__call__`] if `text` is not `None` to encode
+        the text. To prepare the image(s), this method forwards the `images` and `kwrags` arguments to
+        LlavaNextImageProcessor's [`~LlavaNextImageProcessor.__call__`] if `images` is not `None`. Please refer to the doctsring
+        of the above two methods for more information.
+
+        Args:
+            text (`str`, `List[str]`, `List[List[str]]`):
+                The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings
+                (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set
+                `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
+            images (`PIL.Image.Image`, `np.ndarray`, `torch.Tensor`, `List[PIL.Image.Image]`, `List[np.ndarray]`, `List[torch.Tensor]`):
+                The image or batch of images to be prepared. Each image can be a PIL image, NumPy array or PyTorch
+                tensor. Both channels-first and channels-last formats are supported.
+            padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `False`):
+                Select a strategy to pad the returned sequences (according to the model's padding side and padding
+                index) among:
+                - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
+                  sequence if provided).
+                - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum
+                  acceptable input length for the model if that argument is not provided.
+                - `False` or `'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of different
+                  lengths).
+            max_length (`int`, *optional*):
+                Maximum length of the returned list and optionally padding length (see above).
+            truncation (`bool`, *optional*):
+                Activates truncation to cut input sequences longer than `max_length` to `max_length`.
+            return_tensors (`str` or [`~utils.TensorType`], *optional*):
+                If set, will return tensors of a particular framework. Acceptable values are:
+
+                - `'tf'`: Return TensorFlow `tf.constant` objects.
+                - `'pt'`: Return PyTorch `torch.Tensor` objects.
+                - `'np'`: Return NumPy `np.ndarray` objects.
+                - `'jax'`: Return JAX `jnp.ndarray` objects.
+
+        Returns:
+            [`BatchFeature`]: A [`BatchFeature`] with the following fields:
+
+            - **input_ids** -- List of token ids to be fed to a model. Returned when `text` is not `None`.
+            - **attention_mask** -- List of indices specifying which tokens should be attended to by the model (when
+              `return_attention_mask=True` or if *"attention_mask"* is in `self.model_input_names` and if `text` is not
+              `None`).
+            - **pixel_values** -- Pixel values to be fed to a model. Returned when `images` is not `None`.
+        """
+        if images is not None:
+            image_inputs = self.image_processor(images, return_tensors=return_tensors)
+        else:
+            image_inputs = {}
+        text_inputs = self.tokenizer(
+            text, return_tensors=return_tensors, padding=padding, truncation=truncation, max_length=max_length
+        )
+
+        return BatchFeature(data={**text_inputs, **image_inputs})
+
+    # Copied from transformers.models.clip.processing_clip.CLIPProcessor.batch_decode with CLIP->Llama
+    def batch_decode(self, *args, **kwargs):
+        """
+        This method forwards all its arguments to LlamaTokenizerFast's [`~PreTrainedTokenizer.batch_decode`]. Please
+        refer to the docstring of this method for more information.
+        """
+        return self.tokenizer.batch_decode(*args, **kwargs)
+
+    # Copied from transformers.models.clip.processing_clip.CLIPProcessor.decode with CLIP->Llama
+    def decode(self, *args, **kwargs):
+        """
+        This method forwards all its arguments to LlamaTokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to
+        the docstring of this method for more information.
+        """
+        return self.tokenizer.decode(*args, **kwargs)
+
+    @property
+    # Copied from transformers.models.clip.processing_clip.CLIPProcessor.model_input_names
+    def model_input_names(self):
+        tokenizer_input_names = self.tokenizer.model_input_names
+        image_processor_input_names = self.image_processor.model_input_names
+        return list(dict.fromkeys(tokenizer_input_names + image_processor_input_names))

src/transformers/models/mamba/modeling_mamba.py

@@ -16,7 +16,7 @@
 
 import math
 from dataclasses import dataclass
-from typing import Any, Dict, List, Optional, Tuple, Union
+from typing import Any, Dict, Optional, Tuple, Union
 
 import torch
 import torch.utils.checkpoint
@@ -59,6 +59,24 @@ _CONFIG_FOR_DOC = "MambaConfig"
 MAMBA_PRETRAINED_MODEL_ARCHIVE_LIST = []  # See all Mamba models at https://huggingface.co/models?filter=mamba
 
 
+class MambaCache:
+    def __init__(self, config, batch_size, dtype=torch.float16, device=None):
+        self.seqlen_offset = 0
+        self.dtype = dtype
+        intermediate_size = config.intermediate_size
+        ssm_state_size = config.state_size
+        conv_kernel_size = config.conv_kernel
+
+        self.conv_states = {
+            i: torch.zeros(batch_size, intermediate_size, conv_kernel_size, device=device, dtype=dtype)
+            for i in range(config.num_hidden_layers)
+        }
+        self.ssm_states = {
+            i: torch.zeros(batch_size, intermediate_size, ssm_state_size, device=device, dtype=dtype)
+            for i in range(config.num_hidden_layers)
+        }
+
+
 class MambaMixer(nn.Module):
     """
     Compute ∆, A, B, C, and D the state space parameters and compute the `contextualized_states`.
@@ -112,7 +130,7 @@ class MambaMixer(nn.Module):
                 " https://github.com/Dao-AILab/causal-conv1d"
             )
 
-    def cuda_kernels_forward(self, hidden_states: torch.Tensor, cache_params=None):
+    def cuda_kernels_forward(self, hidden_states: torch.Tensor, cache_params: Optional[MambaCache] = None):
         # 1. Gated MLP's linear projection
         projected_states = self.in_proj(hidden_states).transpose(1, 2)
 
@@ -202,7 +220,7 @@ class MambaMixer(nn.Module):
         return contextualized_states
 
     # fmt: off
-    def slow_forward(self, input_states, cache_params=None):
+    def slow_forward(self, input_states, cache_params: Optional[MambaCache]=None):
         batch_size, seq_len, _ = input_states.shape
         dtype = input_states.dtype
         # 1. Gated MLP's linear projection
@@ -268,34 +286,16 @@ class MambaMixer(nn.Module):
         return contextualized_states
     # fmt: on
 
-    def forward(self, hidden_states, cache_params=None):
+    def forward(self, hidden_states, cache_params: Optional[MambaCache] = None):
         if is_fast_path_available and "cuda" in self.x_proj.weight.device.type:
             return self.cuda_kernels_forward(hidden_states, cache_params)
         return self.slow_forward(hidden_states, cache_params)
 
 
-class MambaCache:
-    def __init__(self, config, batch_size, dtype=torch.float16, device=None):
-        self.seqlen_offset = 0
-        self.dtype = dtype
-        intermediate_size = config.intermediate_size
-        ssm_state_size = config.state_size
-        conv_kernel_size = config.conv_kernel
-
-        self.conv_states = {
-            i: torch.zeros(batch_size, intermediate_size, conv_kernel_size, device=device, dtype=dtype)
-            for i in range(config.num_hidden_layers)
-        }
-        self.ssm_states = {
-            i: torch.zeros(batch_size, intermediate_size, ssm_state_size, device=device, dtype=dtype)
-            for i in range(config.num_hidden_layers)
-        }
-
-
 class MambaRMSNorm(nn.Module):
     def __init__(self, hidden_size, eps=1e-6):
         """
-        LlamaRMSNorm is equivalent to T5LayerNorm
+        MambaRMSNorm is equivalent to T5LayerNorm and LlamaRMSNorm
         """
         super().__init__()
         self.weight = nn.Parameter(torch.ones(hidden_size))
@@ -318,7 +318,7 @@ class MambaBlock(nn.Module):
         self.norm = MambaRMSNorm(config.hidden_size, eps=config.layer_norm_epsilon)
         self.mixer = MambaMixer(config, layer_idx=layer_idx)
 
-    def forward(self, hidden_states, cache_params=None):
+    def forward(self, hidden_states, cache_params: Optional[MambaCache] = None):
         residual = hidden_states
         hidden_states = self.norm(hidden_states.to(dtype=self.norm.weight.dtype))
         if self.residual_in_fp32:
@@ -396,11 +396,11 @@ class MambaOutput(ModelOutput):
     Args:
         last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
             Sequence of hidden-states at the output of the last layer of the model.
-        cache_params (list of five `torch.FloatTensor` of shape `(batch_size, hidden_size, num_hidden_layers)`):
+        cache_params (`MambaCache`):
             The state of the model at the last time step. Can be used in a forward method with the next `input_ids` to
             avoid providing the old `input_ids`.
 
-            Includes both the State space model states weights after the selective scan, and the Convolutional states
+            Includes both the State space model state matrices after the selective scan, and the Convolutional states
         hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
             Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
             one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.
@@ -408,8 +408,8 @@ class MambaOutput(ModelOutput):
             Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
     """
 
-    last_hidden_state: torch.FloatTensor = None
-    cache_params: Optional[List[torch.FloatTensor]] = None
+    last_hidden_state: Optional[torch.FloatTensor] = None
+    cache_params: Optional[MambaCache] = None
     hidden_states: Optional[Tuple[torch.FloatTensor]] = None
 
 
@@ -423,9 +423,11 @@ class MambaCausalLMOutput(ModelOutput):
             Language modeling loss (for next-token prediction).
         logits (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`):
             Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
-        cache_params (list of five `torch.FloatTensor` of shape `(batch_size, hidden_size, num_hidden_layers)`):
+        cache_params (`MambaCache`):
             The state of the model at the last time step. Can be used in a forward method with the next `input_ids` to
             avoid providing the old `input_ids`.
+
+            Includes both the State space model state matrices after the selective scan, and the Convolutional states
         hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
             Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
             one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.
@@ -434,8 +436,8 @@ class MambaCausalLMOutput(ModelOutput):
     """
 
     loss: Optional[torch.FloatTensor] = None
-    logits: torch.FloatTensor = None
-    cache_params: Optional[List[torch.FloatTensor]] = None
+    logits: Optional[torch.FloatTensor] = None
+    cache_params: Optional[MambaCache] = None
     hidden_states: Optional[Tuple[torch.FloatTensor]] = None
 
 
@@ -516,7 +518,7 @@ class MambaModel(MambaPreTrainedModel):
         self,
         input_ids: Optional[torch.LongTensor] = None,
         inputs_embeds: Optional[torch.LongTensor] = None,
-        cache_params: Optional[List[torch.FloatTensor]] = None,
+        cache_params: Optional[MambaCache] = None,
         use_cache: Optional[bool] = None,
         output_hidden_states: Optional[bool] = None,
         return_dict: Optional[bool] = None,
@@ -609,7 +611,7 @@ class MambaForCausalLM(MambaPreTrainedModel):
         return model_kwargs
 
     def prepare_inputs_for_generation(
-        self, input_ids, cache_params=None, inputs_embeds=None, attention_mask=None, **kwargs
+        self, input_ids, cache_params: Optional[MambaCache] = None, inputs_embeds=None, attention_mask=None, **kwargs
     ):
         # only last token for inputs_ids if the state is passed along.
         if cache_params is not None:
@@ -633,10 +635,11 @@ class MambaForCausalLM(MambaPreTrainedModel):
         self,
         input_ids: Optional[torch.LongTensor] = None,
         inputs_embeds: Optional[torch.FloatTensor] = None,
-        cache_params: Optional[torch.FloatTensor] = None,
+        cache_params: Optional[MambaCache] = None,
         labels: Optional[torch.LongTensor] = None,
         output_hidden_states: Optional[bool] = None,
         return_dict: Optional[bool] = None,
+        use_cache: Optional[bool] = None,
         **kwargs,  # for now we need this for generation
     ) -> Union[Tuple, MambaCausalLMOutput]:
         r"""
@@ -653,6 +656,7 @@ class MambaForCausalLM(MambaPreTrainedModel):
             inputs_embeds=inputs_embeds,
             output_hidden_states=output_hidden_states,
             return_dict=return_dict,
+            use_cache=use_cache,
         )
         hidden_states = mamba_outputs[0]
 

src/transformers/models/musicgen_melody/modeling_musicgen_melody.py

@@ -1294,7 +1294,7 @@ class MusicgenMelodyForCausalLM(MusicgenMelodyPreTrainedModel):
                 )
 
             # 11. run greedy search
-            outputs = self.greedy_search(
+            outputs = self._greedy_search(
                 input_ids,
                 logits_processor=logits_processor,
                 stopping_criteria=stopping_criteria,
@@ -1319,7 +1319,7 @@ class MusicgenMelodyForCausalLM(MusicgenMelodyPreTrainedModel):
             )
 
             # 12. run sample
-            outputs = self.sample(
+            outputs = self._sample(
                 input_ids,
                 logits_processor=logits_processor,
                 logits_warper=logits_warper,

src/transformers/models/superpoint/__init__.py

@@ -0,0 +1,77 @@
+# Copyright 2024 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from typing import TYPE_CHECKING
+
+# rely on isort to merge the imports
+from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available, is_vision_available
+
+
+_import_structure = {
+    "configuration_superpoint": [
+        "SUPERPOINT_PRETRAINED_CONFIG_ARCHIVE_MAP",
+        "SuperPointConfig",
+    ]
+}
+
+try:
+    if not is_vision_available():
+        raise OptionalDependencyNotAvailable()
+except OptionalDependencyNotAvailable:
+    pass
+else:
+    _import_structure["image_processing_superpoint"] = ["SuperPointImageProcessor"]
+
+try:
+    if not is_torch_available():
+        raise OptionalDependencyNotAvailable()
+except OptionalDependencyNotAvailable:
+    pass
+else:
+    _import_structure["modeling_superpoint"] = [
+        "SUPERPOINT_PRETRAINED_MODEL_ARCHIVE_LIST",
+        "SuperPointForKeypointDetection",
+        "SuperPointPreTrainedModel",
+    ]
+
+
+if TYPE_CHECKING:
+    from .configuration_superpoint import (
+        SUPERPOINT_PRETRAINED_CONFIG_ARCHIVE_MAP,
+        SuperPointConfig,
+    )
+
+    try:
+        if not is_vision_available():
+            raise OptionalDependencyNotAvailable()
+    except OptionalDependencyNotAvailable:
+        pass
+    else:
+        from .image_processing_superpoint import SuperPointImageProcessor
+
+    try:
+        if not is_torch_available():
+            raise OptionalDependencyNotAvailable()
+    except OptionalDependencyNotAvailable:
+        pass
+    else:
+        from .modeling_superpoint import (
+            SUPERPOINT_PRETRAINED_MODEL_ARCHIVE_LIST,
+            SuperPointForKeypointDetection,
+            SuperPointPreTrainedModel,
+        )
+
+else:
+    import sys
+
+    sys.modules[__name__] = _LazyModule(__name__, globals()["__file__"], _import_structure)

src/transformers/models/superpoint/configuration_superpoint.py

@@ -0,0 +1,91 @@
+# Copyright 2024 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from typing import List
+
+from ...configuration_utils import PretrainedConfig
+from ...utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+SUPERPOINT_PRETRAINED_CONFIG_ARCHIVE_MAP = {
+    "magic-leap-community/superpoint": "https://huggingface.co/magic-leap-community/superpoint/blob/main/config.json"
+}
+
+
+class SuperPointConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`SuperPointForKeypointDetection`]. It is used to instantiate a
+    SuperPoint model according to the specified arguments, defining the model architecture. Instantiating a
+    configuration with the defaults will yield a similar configuration to that of the SuperPoint
+    [magic-leap-community/superpoint](https://huggingface.co/magic-leap-community/superpoint) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        encoder_hidden_sizes (`List`, *optional*, defaults to `[64, 64, 128, 128]`):
+            The number of channels in each convolutional layer in the encoder.
+        decoder_hidden_size (`int`, *optional*, defaults to 256): The hidden size of the decoder.
+        keypoint_decoder_dim (`int`, *optional*, defaults to 65): The output dimension of the keypoint decoder.
+        descriptor_decoder_dim (`int`, *optional*, defaults to 256): The output dimension of the descriptor decoder.
+        keypoint_threshold (`float`, *optional*, defaults to 0.005):
+            The threshold to use for extracting keypoints.
+        max_keypoints (`int`, *optional*, defaults to -1):
+            The maximum number of keypoints to extract. If `-1`, will extract all keypoints.
+        nms_radius (`int`, *optional*, defaults to 4):
+            The radius for non-maximum suppression.
+        border_removal_distance (`int`, *optional*, defaults to 4):
+            The distance from the border to remove keypoints.
+        initializer_range (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+
+    Example:
+    ```python
+    >>> from transformers import SuperPointConfig, SuperPointForKeypointDetection
+
+    >>> # Initializing a SuperPoint superpoint style configuration
+    >>> configuration = SuperPointConfig()
+    >>> # Initializing a model from the superpoint style configuration
+    >>> model = SuperPointForKeypointDetection(configuration)
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "superpoint"
+
+    def __init__(
+        self,
+        encoder_hidden_sizes: List[int] = [64, 64, 128, 128],
+        decoder_hidden_size: int = 256,
+        keypoint_decoder_dim: int = 65,
+        descriptor_decoder_dim: int = 256,
+        keypoint_threshold: float = 0.005,
+        max_keypoints: int = -1,
+        nms_radius: int = 4,
+        border_removal_distance: int = 4,
+        initializer_range=0.02,
+        **kwargs,
+    ):
+        self.encoder_hidden_sizes = encoder_hidden_sizes
+        self.decoder_hidden_size = decoder_hidden_size
+        self.keypoint_decoder_dim = keypoint_decoder_dim
+        self.descriptor_decoder_dim = descriptor_decoder_dim
+        self.keypoint_threshold = keypoint_threshold
+        self.max_keypoints = max_keypoints
+        self.nms_radius = nms_radius
+        self.border_removal_distance = border_removal_distance
+        self.initializer_range = initializer_range
+
+        super().__init__(**kwargs)

src/transformers/models/superpoint/convert_superpoint_to_pytorch.py

@@ -0,0 +1,175 @@
+# Copyright 2024 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+import argparse
+import os
+
+import requests
+import torch
+from PIL import Image
+
+from transformers import SuperPointConfig, SuperPointForKeypointDetection, SuperPointImageProcessor
+
+
+def get_superpoint_config():
+    config = SuperPointConfig(
+        encoder_hidden_sizes=[64, 64, 128, 128],
+        decoder_hidden_size=256,
+        keypoint_decoder_dim=65,
+        descriptor_decoder_dim=256,
+        keypoint_threshold=0.005,
+        max_keypoints=-1,
+        nms_radius=4,
+        border_removal_distance=4,
+        initializer_range=0.02,
+    )
+
+    return config
+
+
+def create_rename_keys(config, state_dict):
+    rename_keys = []
+
+    # Encoder weights
+    rename_keys.append(("conv1a.weight", "encoder.conv_blocks.0.conv_a.weight"))
+    rename_keys.append(("conv1b.weight", "encoder.conv_blocks.0.conv_b.weight"))
+    rename_keys.append(("conv2a.weight", "encoder.conv_blocks.1.conv_a.weight"))
+    rename_keys.append(("conv2b.weight", "encoder.conv_blocks.1.conv_b.weight"))
+    rename_keys.append(("conv3a.weight", "encoder.conv_blocks.2.conv_a.weight"))
+    rename_keys.append(("conv3b.weight", "encoder.conv_blocks.2.conv_b.weight"))
+    rename_keys.append(("conv4a.weight", "encoder.conv_blocks.3.conv_a.weight"))
+    rename_keys.append(("conv4b.weight", "encoder.conv_blocks.3.conv_b.weight"))
+    rename_keys.append(("conv1a.bias", "encoder.conv_blocks.0.conv_a.bias"))
+    rename_keys.append(("conv1b.bias", "encoder.conv_blocks.0.conv_b.bias"))
+    rename_keys.append(("conv2a.bias", "encoder.conv_blocks.1.conv_a.bias"))
+    rename_keys.append(("conv2b.bias", "encoder.conv_blocks.1.conv_b.bias"))
+    rename_keys.append(("conv3a.bias", "encoder.conv_blocks.2.conv_a.bias"))
+    rename_keys.append(("conv3b.bias", "encoder.conv_blocks.2.conv_b.bias"))
+    rename_keys.append(("conv4a.bias", "encoder.conv_blocks.3.conv_a.bias"))
+    rename_keys.append(("conv4b.bias", "encoder.conv_blocks.3.conv_b.bias"))
+
+    # Keypoint Decoder weights
+    rename_keys.append(("convPa.weight", "keypoint_decoder.conv_score_a.weight"))
+    rename_keys.append(("convPb.weight", "keypoint_decoder.conv_score_b.weight"))
+    rename_keys.append(("convPa.bias", "keypoint_decoder.conv_score_a.bias"))
+    rename_keys.append(("convPb.bias", "keypoint_decoder.conv_score_b.bias"))
+
+    # Descriptor Decoder weights
+    rename_keys.append(("convDa.weight", "descriptor_decoder.conv_descriptor_a.weight"))
+    rename_keys.append(("convDb.weight", "descriptor_decoder.conv_descriptor_b.weight"))
+    rename_keys.append(("convDa.bias", "descriptor_decoder.conv_descriptor_a.bias"))
+    rename_keys.append(("convDb.bias", "descriptor_decoder.conv_descriptor_b.bias"))
+
+    return rename_keys
+
+
+def rename_key(dct, old, new):
+    val = dct.pop(old)
+    dct[new] = val
+
+
+def prepare_imgs():
+    url = "http://images.cocodataset.org/val2017/000000039769.jpg"
+    im1 = Image.open(requests.get(url, stream=True).raw)
+    url = "http://images.cocodataset.org/test-stuff2017/000000004016.jpg"
+    im2 = Image.open(requests.get(url, stream=True).raw)
+    return [im1, im2]
+
+
+@torch.no_grad()
+def convert_superpoint_checkpoint(checkpoint_url, pytorch_dump_folder_path, save_model, push_to_hub, test_mode=False):
+    """
+    Copy/paste/tweak model's weights to our SuperPoint structure.
+    """
+
+    print("Downloading original model from checkpoint...")
+    config = get_superpoint_config()
+
+    # load original state_dict from URL
+    original_state_dict = torch.hub.load_state_dict_from_url(checkpoint_url)
+
+    print("Converting model parameters...")
+    # rename keys
+    rename_keys = create_rename_keys(config, original_state_dict)
+    new_state_dict = original_state_dict.copy()
+    for src, dest in rename_keys:
+        rename_key(new_state_dict, src, dest)
+
+    # Load HuggingFace model
+    model = SuperPointForKeypointDetection(config)
+    model.load_state_dict(new_state_dict)
+    model.eval()
+    print("Successfully loaded weights in the model")
+
+    # Check model outputs
+    preprocessor = SuperPointImageProcessor()
+    inputs = preprocessor(images=prepare_imgs(), return_tensors="pt")
+    outputs = model(**inputs)
+
+    # If test_mode is True, we check that the model outputs match the original results
+    if test_mode:
+        torch.count_nonzero(outputs.mask[0])
+        expected_keypoints_shape = (2, 830, 2)
+        expected_scores_shape = (2, 830)
+        expected_descriptors_shape = (2, 830, 256)
+
+        expected_keypoints_values = torch.tensor([[480.0, 9.0], [494.0, 9.0], [489.0, 16.0]])
+        expected_scores_values = torch.tensor([0.0064, 0.0140, 0.0595, 0.0728, 0.5170, 0.0175, 0.1523, 0.2055, 0.0336])
+        expected_descriptors_value = torch.tensor(-0.1096)
+        assert outputs.keypoints.shape == expected_keypoints_shape
+        assert outputs.scores.shape == expected_scores_shape
+        assert outputs.descriptors.shape == expected_descriptors_shape
+
+        assert torch.allclose(outputs.keypoints[0, :3], expected_keypoints_values, atol=1e-3)
+        assert torch.allclose(outputs.scores[0, :9], expected_scores_values, atol=1e-3)
+        assert torch.allclose(outputs.descriptors[0, 0, 0], expected_descriptors_value, atol=1e-3)
+        print("Model outputs match the original results!")
+
+    if save_model:
+        print("Saving model to local...")
+        # Create folder to save model
+        if not os.path.isdir(pytorch_dump_folder_path):
+            os.mkdir(pytorch_dump_folder_path)
+
+        model.save_pretrained(pytorch_dump_folder_path)
+        preprocessor.save_pretrained(pytorch_dump_folder_path)
+
+        model_name = "superpoint"
+        if push_to_hub:
+            print(f"Pushing {model_name} to the hub...")
+        model.push_to_hub(model_name)
+        preprocessor.push_to_hub(model_name)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    # Required parameters
+    parser.add_argument(
+        "--checkpoint_url",
+        default="https://github.com/magicleap/SuperPointPretrainedNetwork/raw/master/superpoint_v1.pth",
+        type=str,
+        help="URL of the original SuperPoint checkpoint you'd like to convert.",
+    )
+    parser.add_argument(
+        "--pytorch_dump_folder_path",
+        default="model",
+        type=str,
+        help="Path to the output PyTorch model directory.",
+    )
+    parser.add_argument("--save_model", action="store_true", help="Save model to local")
+    parser.add_argument("--push_to_hub", action="store_true", help="Push model and image preprocessor to the hub")
+
+    args = parser.parse_args()
+    convert_superpoint_checkpoint(
+        args.checkpoint_url, args.pytorch_dump_folder_path, args.save_model, args.push_to_hub
+    )

src/transformers/models/superpoint/image_processing_superpoint.py

@@ -0,0 +1,272 @@
+# Copyright 2024 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""Image processor class for SuperPoint."""
+
+from typing import Dict, Optional, Union
+
+import numpy as np
+
+from ... import is_vision_available, requires_backends
+from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
+from ...image_transforms import resize, to_channel_dimension_format
+from ...image_utils import (
+    ChannelDimension,
+    ImageInput,
+    infer_channel_dimension_format,
+    is_scaled_image,
+    make_list_of_images,
+    to_numpy_array,
+    valid_images,
+)
+from ...utils import TensorType, logging
+
+
+if is_vision_available():
+    import PIL
+
+logger = logging.get_logger(__name__)
+
+
+def is_grayscale(
+    image: ImageInput,
+    input_data_format: Optional[Union[str, ChannelDimension]] = None,
+):
+    if input_data_format == ChannelDimension.FIRST:
+        return np.all(image[0, ...] == image[1, ...]) and np.all(image[1, ...] == image[2, ...])
+    elif input_data_format == ChannelDimension.LAST:
+        return np.all(image[..., 0] == image[..., 1]) and np.all(image[..., 1] == image[..., 2])
+
+
+def convert_to_grayscale(
+    image: ImageInput,
+    input_data_format: Optional[Union[str, ChannelDimension]] = None,
+) -> ImageInput:
+    """
+    Converts an image to grayscale format using the NTSC formula. Only support numpy and PIL Image. TODO support torch
+    and tensorflow grayscale conversion
+
+    This function is supposed to return a 1-channel image, but it returns a 3-channel image with the same value in each
+    channel, because of an issue that is discussed in :
+    https://github.com/huggingface/transformers/pull/25786#issuecomment-1730176446
+
+    Args:
+        image (Image):
+            The image to convert.
+        input_data_format (`ChannelDimension` or `str`, *optional*):
+            The channel dimension format for the input image.
+    """
+    requires_backends(convert_to_grayscale, ["vision"])
+
+    if isinstance(image, np.ndarray):
+        if input_data_format == ChannelDimension.FIRST:
+            gray_image = image[0, ...] * 0.2989 + image[1, ...] * 0.5870 + image[2, ...] * 0.1140
+            gray_image = np.stack([gray_image] * 3, axis=0)
+        elif input_data_format == ChannelDimension.LAST:
+            gray_image = image[..., 0] * 0.2989 + image[..., 1] * 0.5870 + image[..., 2] * 0.1140
+            gray_image = np.stack([gray_image] * 3, axis=-1)
+        return gray_image
+
+    if not isinstance(image, PIL.Image.Image):
+        return image
+
+    image = image.convert("L")
+    return image
+
+
+class SuperPointImageProcessor(BaseImageProcessor):
+    r"""
+    Constructs a SuperPoint image processor.
+
+    Args:
+        do_resize (`bool`, *optional*, defaults to `True`):
+            Controls whether to resize the image's (height, width) dimensions to the specified `size`. Can be overriden
+            by `do_resize` in the `preprocess` method.
+        size (`Dict[str, int]` *optional*, defaults to `{"height": 480, "width": 640}`):
+            Resolution of the output image after `resize` is applied. Only has an effect if `do_resize` is set to
+            `True`. Can be overriden by `size` in the `preprocess` method.
+        do_rescale (`bool`, *optional*, defaults to `True`):
+            Whether to rescale the image by the specified scale `rescale_factor`. Can be overriden by `do_rescale` in
+            the `preprocess` method.
+        rescale_factor (`int` or `float`, *optional*, defaults to `1/255`):
+            Scale factor to use if rescaling the image. Can be overriden by `rescale_factor` in the `preprocess`
+            method.
+    """
+
+    model_input_names = ["pixel_values"]
+
+    def __init__(
+        self,
+        do_resize: bool = True,
+        size: Dict[str, int] = None,
+        do_rescale: bool = True,
+        rescale_factor: float = 1 / 255,
+        **kwargs,
+    ) -> None:
+        super().__init__(**kwargs)
+        size = size if size is not None else {"height": 480, "width": 640}
+        size = get_size_dict(size, default_to_square=False)
+
+        self.do_resize = do_resize
+        self.size = size
+        self.do_rescale = do_rescale
+        self.rescale_factor = rescale_factor
+
+    def resize(
+        self,
+        image: np.ndarray,
+        size: Dict[str, int],
+        data_format: Optional[Union[str, ChannelDimension]] = None,
+        input_data_format: Optional[Union[str, ChannelDimension]] = None,
+        **kwargs,
+    ):
+        """
+        Resize an image.
+
+        Args:
+            image (`np.ndarray`):
+                Image to resize.
+            size (`Dict[str, int]`):
+                Dictionary of the form `{"height": int, "width": int}`, specifying the size of the output image.
+            data_format (`ChannelDimension` or `str`, *optional*):
+                The channel dimension format of the output image. If not provided, it will be inferred from the input
+                image. Can be one of:
+                - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format.
+                - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format.
+                - `"none"` or `ChannelDimension.NONE`: image in (height, width) format.
+            input_data_format (`ChannelDimension` or `str`, *optional*):
+                The channel dimension format for the input image. If unset, the channel dimension format is inferred
+                from the input image. Can be one of:
+                - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format.
+                - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format.
+                - `"none"` or `ChannelDimension.NONE`: image in (height, width) format.
+        """
+        size = get_size_dict(size, default_to_square=False)
+
+        return resize(
+            image,
+            size=(size["height"], size["width"]),
+            data_format=data_format,
+            input_data_format=input_data_format,
+            **kwargs,
+        )
+
+    def preprocess(
+        self,
+        images,
+        do_resize: bool = None,
+        size: Dict[str, int] = None,
+        do_rescale: bool = None,
+        rescale_factor: float = None,
+        return_tensors: Optional[Union[str, TensorType]] = None,
+        data_format: ChannelDimension = ChannelDimension.FIRST,
+        input_data_format: Optional[Union[str, ChannelDimension]] = None,
+        **kwargs,
+    ) -> BatchFeature:
+        """
+        Preprocess an image or batch of images.
+
+        Args:
+            images (`ImageInput`):
+                Image to preprocess. Expects a single or batch of images with pixel values ranging from 0 to 255. If
+                passing in images with pixel values between 0 and 1, set `do_rescale=False`.
+            do_resize (`bool`, *optional*, defaults to `self.do_resize`):
+                Whether to resize the image.
+            size (`Dict[str, int]`, *optional*, defaults to `self.size`):
+                Size of the output image after `resize` has been applied. If `size["shortest_edge"]` >= 384, the image
+                is resized to `(size["shortest_edge"], size["shortest_edge"])`. Otherwise, the smaller edge of the
+                image will be matched to `int(size["shortest_edge"]/ crop_pct)`, after which the image is cropped to
+                `(size["shortest_edge"], size["shortest_edge"])`. Only has an effect if `do_resize` is set to `True`.
+            do_rescale (`bool`, *optional*, defaults to `self.do_rescale`):
+                Whether to rescale the image values between [0 - 1].
+            rescale_factor (`float`, *optional*, defaults to `self.rescale_factor`):
+                Rescale factor to rescale the image by if `do_rescale` is set to `True`.
+            return_tensors (`str` or `TensorType`, *optional*):
+                The type of tensors to return. Can be one of:
+                    - Unset: Return a list of `np.ndarray`.
+                    - `TensorType.TENSORFLOW` or `'tf'`: Return a batch of type `tf.Tensor`.
+                    - `TensorType.PYTORCH` or `'pt'`: Return a batch of type `torch.Tensor`.
+                    - `TensorType.NUMPY` or `'np'`: Return a batch of type `np.ndarray`.
+                    - `TensorType.JAX` or `'jax'`: Return a batch of type `jax.numpy.ndarray`.
+            data_format (`ChannelDimension` or `str`, *optional*, defaults to `ChannelDimension.FIRST`):
+                The channel dimension format for the output image. Can be one of:
+                - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format.
+                - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format.
+                - Unset: Use the channel dimension format of the input image.
+            input_data_format (`ChannelDimension` or `str`, *optional*):
+                The channel dimension format for the input image. If unset, the channel dimension format is inferred
+                from the input image. Can be one of:
+                - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format.
+                - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format.
+                - `"none"` or `ChannelDimension.NONE`: image in (height, width) format.
+        """
+
+        do_resize = do_resize if do_resize is not None else self.do_resize
+        do_rescale = do_rescale if do_rescale is not None else self.do_rescale
+        rescale_factor = rescale_factor if rescale_factor is not None else self.rescale_factor
+
+        size = size if size is not None else self.size
+        size = get_size_dict(size, default_to_square=False)
+
+        images = make_list_of_images(images)
+
+        if not valid_images(images):
+            raise ValueError(
+                "Invalid image type. Must be of type PIL.Image.Image, numpy.ndarray, "
+                "torch.Tensor, tf.Tensor or jax.ndarray."
+            )
+
+        if do_resize and size is None:
+            raise ValueError("Size must be specified if do_resize is True.")
+
+        if do_rescale and rescale_factor is None:
+            raise ValueError("Rescale factor must be specified if do_rescale is True.")
+
+        # All transformations expect numpy arrays.
+        images = [to_numpy_array(image) for image in images]
+
+        if is_scaled_image(images[0]) and do_rescale:
+            logger.warning_once(
+                "It looks like you are trying to rescale already rescaled images. If the input"
+                " images have pixel values between 0 and 1, set `do_rescale=False` to avoid rescaling them again."
+            )
+
+        if input_data_format is None:
+            # We assume that all images have the same channel dimension format.
+            input_data_format = infer_channel_dimension_format(images[0])
+
+        if do_resize:
+            images = [self.resize(image=image, size=size, input_data_format=input_data_format) for image in images]
+
+        if do_rescale:
+            images = [
+                self.rescale(image=image, scale=rescale_factor, input_data_format=input_data_format)
+                for image in images
+            ]
+
+        if input_data_format is None:
+            # We assume that all images have the same channel dimension format.
+            input_data_format = infer_channel_dimension_format(images[0])
+
+        # Checking if image is RGB or grayscale
+        for i in range(len(images)):
+            if not is_grayscale(images[i], input_data_format):
+                images[i] = convert_to_grayscale(images[i], input_data_format=input_data_format)
+
+        images = [
+            to_channel_dimension_format(image, data_format, input_channel_dim=input_data_format) for image in images
+        ]
+
+        data = {"pixel_values": images}
+
+        return BatchFeature(data=data, tensor_type=return_tensors)

src/transformers/models/superpoint/modeling_superpoint.py

@@ -0,0 +1,507 @@
+# Copyright 2024 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""PyTorch SuperPoint model."""
+from dataclasses import dataclass
+from typing import Optional, Tuple, Union
+
+import torch
+from torch import nn
+
+from transformers import PreTrainedModel
+from transformers.modeling_outputs import (
+    BaseModelOutputWithNoAttention,
+)
+from transformers.models.superpoint.configuration_superpoint import SuperPointConfig
+
+from ...pytorch_utils import is_torch_greater_or_equal_than_1_13
+from ...utils import (
+    ModelOutput,
+    add_start_docstrings,
+    add_start_docstrings_to_model_forward,
+    logging,
+)
+
+
+logger = logging.get_logger(__name__)
+
+_CONFIG_FOR_DOC = "SuperPointConfig"
+
+_CHECKPOINT_FOR_DOC = "magic-leap-community/superpoint"
+
+SUPERPOINT_PRETRAINED_MODEL_ARCHIVE_LIST = ["magic-leap-community/superpoint"]
+
+
+def remove_keypoints_from_borders(
+    keypoints: torch.Tensor, scores: torch.Tensor, border: int, height: int, width: int
+) -> Tuple[torch.Tensor, torch.Tensor]:
+    """Removes keypoints (and their associated scores) that are too close to the border"""
+    mask_h = (keypoints[:, 0] >= border) & (keypoints[:, 0] < (height - border))
+    mask_w = (keypoints[:, 1] >= border) & (keypoints[:, 1] < (width - border))
+    mask = mask_h & mask_w
+    return keypoints[mask], scores[mask]
+
+
+def top_k_keypoints(keypoints: torch.Tensor, scores: torch.Tensor, k: int) -> Tuple[torch.Tensor, torch.Tensor]:
+    """Keeps the k keypoints with highest score"""
+    if k >= len(keypoints):
+        return keypoints, scores
+    scores, indices = torch.topk(scores, k, dim=0)
+    return keypoints[indices], scores
+
+
+def simple_nms(scores: torch.Tensor, nms_radius: int) -> torch.Tensor:
+    """Applies non-maximum suppression on scores"""
+    if nms_radius < 0:
+        raise ValueError("Expected positive values for nms_radius")
+
+    def max_pool(x):
+        return nn.functional.max_pool2d(x, kernel_size=nms_radius * 2 + 1, stride=1, padding=nms_radius)
+
+    zeros = torch.zeros_like(scores)
+    max_mask = scores == max_pool(scores)
+    for _ in range(2):
+        supp_mask = max_pool(max_mask.float()) > 0
+        supp_scores = torch.where(supp_mask, zeros, scores)
+        new_max_mask = supp_scores == max_pool(supp_scores)
+        max_mask = max_mask | (new_max_mask & (~supp_mask))
+    return torch.where(max_mask, scores, zeros)
+
+
+@dataclass
+class ImagePointDescriptionOutput(ModelOutput):
+    """
+    Base class for outputs of image point description models. Due to the nature of keypoint detection, the number of
+    keypoints is not fixed and can vary from image to image, which makes batching non-trivial. In the batch of images,
+    the maximum number of keypoints is set as the dimension of the keypoints, scores and descriptors tensors. The mask
+    tensor is used to indicate which values in the keypoints, scores and descriptors tensors are keypoint information
+    and which are padding.
+
+    Args:
+        last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
+            Sequence of hidden-states at the output of the last layer of the decoder of the model.
+        keypoints (`torch.FloatTensor` of shape `(batch_size, num_keypoints, 2)`):
+            Relative (x, y) coordinates of predicted keypoints in a given image.
+        scores (`torch.FloatTensor` of shape `(batch_size, num_keypoints)`):
+            Scores of predicted keypoints.
+        descriptors (`torch.FloatTensor` of shape `(batch_size, num_keypoints, descriptor_size)`):
+            Descriptors of predicted keypoints.
+        mask (`torch.BoolTensor` of shape `(batch_size, num_keypoints)`):
+            Mask indicating which values in keypoints, scores and descriptors are keypoint information.
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or
+        when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
+            one for the output of each stage) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states
+            (also called feature maps) of the model at the output of each stage.
+    """
+
+    last_hidden_state: torch.FloatTensor = None
+    keypoints: Optional[torch.IntTensor] = None
+    scores: Optional[torch.FloatTensor] = None
+    descriptors: Optional[torch.FloatTensor] = None
+    mask: Optional[torch.BoolTensor] = None
+    hidden_states: Optional[Tuple[torch.FloatTensor]] = None
+
+
+class SuperPointConvBlock(nn.Module):
+    def __init__(
+        self, config: SuperPointConfig, in_channels: int, out_channels: int, add_pooling: bool = False
+    ) -> None:
+        super().__init__()
+        self.conv_a = nn.Conv2d(
+            in_channels,
+            out_channels,
+            kernel_size=3,
+            stride=1,
+            padding=1,
+        )
+        self.conv_b = nn.Conv2d(
+            out_channels,
+            out_channels,
+            kernel_size=3,
+            stride=1,
+            padding=1,
+        )
+        self.relu = nn.ReLU(inplace=True)
+        self.pool = nn.MaxPool2d(kernel_size=2, stride=2) if add_pooling else None
+
+    def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
+        hidden_states = self.relu(self.conv_a(hidden_states))
+        hidden_states = self.relu(self.conv_b(hidden_states))
+        if self.pool is not None:
+            hidden_states = self.pool(hidden_states)
+        return hidden_states
+
+
+class SuperPointEncoder(nn.Module):
+    """
+    SuperPoint encoder module. It is made of 4 convolutional layers with ReLU activation and max pooling, reducing the
+     dimensionality of the image.
+    """
+
+    def __init__(self, config: SuperPointConfig) -> None:
+        super().__init__()
+        # SuperPoint uses 1 channel images
+        self.input_dim = 1
+
+        conv_blocks = []
+        conv_blocks.append(
+            SuperPointConvBlock(config, self.input_dim, config.encoder_hidden_sizes[0], add_pooling=True)
+        )
+        for i in range(1, len(config.encoder_hidden_sizes) - 1):
+            conv_blocks.append(
+                SuperPointConvBlock(
+                    config, config.encoder_hidden_sizes[i - 1], config.encoder_hidden_sizes[i], add_pooling=True
+                )
+            )
+        conv_blocks.append(
+            SuperPointConvBlock(
+                config, config.encoder_hidden_sizes[-2], config.encoder_hidden_sizes[-1], add_pooling=False
+            )
+        )
+        self.conv_blocks = nn.ModuleList(conv_blocks)
+
+    def forward(
+        self,
+        input,
+        output_hidden_states: Optional[bool] = False,
+        return_dict: Optional[bool] = True,
+    ) -> Union[Tuple, BaseModelOutputWithNoAttention]:
+        all_hidden_states = () if output_hidden_states else None
+
+        for conv_block in self.conv_blocks:
+            input = conv_block(input)
+            if output_hidden_states:
+                all_hidden_states = all_hidden_states + (input,)
+        output = input
+        if not return_dict:
+            return tuple(v for v in [output, all_hidden_states] if v is not None)
+
+        return BaseModelOutputWithNoAttention(
+            last_hidden_state=output,
+            hidden_states=all_hidden_states,
+        )
+
+
+class SuperPointInterestPointDecoder(nn.Module):
+    """
+    The SuperPointInterestPointDecoder uses the output of the SuperPointEncoder to compute the keypoint with scores.
+    The scores are first computed by a convolutional layer, then a softmax is applied to get a probability distribution
+    over the 65 possible keypoint classes. The keypoints are then extracted from the scores by thresholding and
+    non-maximum suppression. Post-processing is then applied to remove keypoints too close to the image borders as well
+    as to keep only the k keypoints with highest score.
+    """
+
+    def __init__(self, config: SuperPointConfig) -> None:
+        super().__init__()
+        self.keypoint_threshold = config.keypoint_threshold
+        self.max_keypoints = config.max_keypoints
+        self.nms_radius = config.nms_radius
+        self.border_removal_distance = config.border_removal_distance
+
+        self.relu = nn.ReLU(inplace=True)
+        self.pool = nn.MaxPool2d(kernel_size=2, stride=2)
+        self.conv_score_a = nn.Conv2d(
+            config.encoder_hidden_sizes[-1],
+            config.decoder_hidden_size,
+            kernel_size=3,
+            stride=1,
+            padding=1,
+        )
+        self.conv_score_b = nn.Conv2d(
+            config.decoder_hidden_size, config.keypoint_decoder_dim, kernel_size=1, stride=1, padding=0
+        )
+
+    def forward(self, encoded: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor]:
+        scores = self._get_pixel_scores(encoded)
+        keypoints, scores = self._extract_keypoints(scores)
+
+        return keypoints, scores
+
+    def _get_pixel_scores(self, encoded: torch.Tensor) -> torch.Tensor:
+        """Based on the encoder output, compute the scores for each pixel of the image"""
+        scores = self.relu(self.conv_score_a(encoded))
+        scores = self.conv_score_b(scores)
+        scores = nn.functional.softmax(scores, 1)[:, :-1]
+        batch_size, _, height, width = scores.shape
+        scores = scores.permute(0, 2, 3, 1).reshape(batch_size, height, width, 8, 8)
+        scores = scores.permute(0, 1, 3, 2, 4).reshape(batch_size, height * 8, width * 8)
+        scores = simple_nms(scores, self.nms_radius)
+        return scores
+
+    def _extract_keypoints(self, scores: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor]:
+        """Based on their scores, extract the pixels that represent the keypoints that will be used for descriptors computation"""
+        _, height, width = scores.shape
+
+        # Threshold keypoints by score value
+        keypoints = torch.nonzero(scores[0] > self.keypoint_threshold)
+        scores = scores[0][tuple(keypoints.t())]
+
+        # Discard keypoints near the image borders
+        keypoints, scores = remove_keypoints_from_borders(
+            keypoints, scores, self.border_removal_distance, height * 8, width * 8
+        )
+
+        # Keep the k keypoints with highest score
+        if self.max_keypoints >= 0:
+            keypoints, scores = top_k_keypoints(keypoints, scores, self.max_keypoints)
+
+        # Convert (y, x) to (x, y)
+        keypoints = torch.flip(keypoints, [1]).float()
+
+        return keypoints, scores
+
+
+class SuperPointDescriptorDecoder(nn.Module):
+    """
+    The SuperPointDescriptorDecoder uses the outputs of both the SuperPointEncoder and the
+    SuperPointInterestPointDecoder to compute the descriptors at the keypoints locations.
+
+    The descriptors are first computed by a convolutional layer, then normalized to have a norm of 1. The descriptors
+    are then interpolated at the keypoints locations.
+    """
+
+    def __init__(self, config: SuperPointConfig) -> None:
+        super().__init__()
+
+        self.relu = nn.ReLU(inplace=True)
+        self.pool = nn.MaxPool2d(kernel_size=2, stride=2)
+        self.conv_descriptor_a = nn.Conv2d(
+            config.encoder_hidden_sizes[-1],
+            config.decoder_hidden_size,
+            kernel_size=3,
+            stride=1,
+            padding=1,
+        )
+        self.conv_descriptor_b = nn.Conv2d(
+            config.decoder_hidden_size,
+            config.descriptor_decoder_dim,
+            kernel_size=1,
+            stride=1,
+            padding=0,
+        )
+
+    def forward(self, encoded: torch.Tensor, keypoints: torch.Tensor) -> torch.Tensor:
+        """Based on the encoder output and the keypoints, compute the descriptors for each keypoint"""
+        descriptors = self.conv_descriptor_b(self.relu(self.conv_descriptor_a(encoded)))
+        descriptors = nn.functional.normalize(descriptors, p=2, dim=1)
+
+        descriptors = self._sample_descriptors(keypoints[None], descriptors[0][None], 8)[0]
+
+        # [descriptor_dim, num_keypoints] -> [num_keypoints, descriptor_dim]
+        descriptors = torch.transpose(descriptors, 0, 1)
+
+        return descriptors
+
+    @staticmethod
+    def _sample_descriptors(keypoints, descriptors, scale: int = 8) -> torch.Tensor:
+        """Interpolate descriptors at keypoint locations"""
+        batch_size, num_channels, height, width = descriptors.shape
+        keypoints = keypoints - scale / 2 + 0.5
+        divisor = torch.tensor([[(width * scale - scale / 2 - 0.5), (height * scale - scale / 2 - 0.5)]])
+        divisor = divisor.to(keypoints)
+        keypoints /= divisor
+        keypoints = keypoints * 2 - 1  # normalize to (-1, 1)
+        kwargs = {"align_corners": True} if is_torch_greater_or_equal_than_1_13 else {}
+        # [batch_size, num_channels, num_keypoints, 2] -> [batch_size, num_channels, num_keypoints, 2]
+        keypoints = keypoints.view(batch_size, 1, -1, 2)
+        descriptors = nn.functional.grid_sample(descriptors, keypoints, mode="bilinear", **kwargs)
+        # [batch_size, descriptor_decoder_dim, num_channels, num_keypoints] -> [batch_size, descriptor_decoder_dim, num_keypoints]
+        descriptors = descriptors.reshape(batch_size, num_channels, -1)
+        descriptors = nn.functional.normalize(descriptors, p=2, dim=1)
+        return descriptors
+
+
+class SuperPointPreTrainedModel(PreTrainedModel):
+    """
+    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
+    models.
+    """
+
+    config_class = SuperPointConfig
+    base_model_prefix = "superpoint"
+    main_input_name = "pixel_values"
+    supports_gradient_checkpointing = False
+
+    def _init_weights(self, module: Union[nn.Linear, nn.Conv2d, nn.LayerNorm]) -> None:
+        """Initialize the weights"""
+        if isinstance(module, (nn.Linear, nn.Conv2d)):
+            # Slightly different from the TF version which uses truncated_normal for initialization
+            # cf https://github.com/pytorch/pytorch/pull/5617
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.LayerNorm):
+            module.bias.data.zero_()
+            module.weight.data.fill_(1.0)
+
+    def extract_one_channel_pixel_values(self, pixel_values: torch.FloatTensor) -> torch.FloatTensor:
+        """
+        Assuming pixel_values has shape (batch_size, 3, height, width), and that all channels values are the same,
+        extract the first channel value to get a tensor of shape (batch_size, 1, height, width) for SuperPoint. This is
+        a workaround for the issue discussed in :
+        https://github.com/huggingface/transformers/pull/25786#issuecomment-1730176446
+
+        Args:
+            pixel_values: torch.FloatTensor of shape (batch_size, 3, height, width)
+
+        Returns:
+            pixel_values: torch.FloatTensor of shape (batch_size, 1, height, width)
+
+        """
+        return pixel_values[:, 0, :, :][:, None, :, :]
+
+
+SUPERPOINT_START_DOCSTRING = r"""
+    This model is a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. Use it
+    as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and
+    behavior.
+
+    Parameters:
+        config ([`SuperPointConfig`]): Model configuration class with all the parameters of the model.
+            Initializing with a config file does not load the weights associated with the model, only the
+            configuration. Check out the [`~PreTrainedModel.from_pretrained`] method to load the model weights.
+    """
+
+SUPERPOINT_INPUTS_DOCSTRING = r"""
+Args:
+    pixel_values (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)`):
+        Pixel values. Pixel values can be obtained using [`SuperPointImageProcessor`]. See
+        [`SuperPointImageProcessor.__call__`] for details.
+    output_hidden_states (`bool`, *optional*):
+        Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for more
+        detail.
+    return_dict (`bool`, *optional*):
+        Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
+    """
+
+
+@add_start_docstrings(
+    "SuperPoint model outputting keypoints and descriptors.",
+    SUPERPOINT_START_DOCSTRING,
+)
+class SuperPointForKeypointDetection(SuperPointPreTrainedModel):
+    """
+    SuperPoint model. It consists of a SuperPointEncoder, a SuperPointInterestPointDecoder and a
+    SuperPointDescriptorDecoder. SuperPoint was proposed in `SuperPoint: Self-Supervised Interest Point Detection and
+    Description <https://arxiv.org/abs/1712.07629>`__ by Daniel DeTone, Tomasz Malisiewicz, and Andrew Rabinovich. It
+    is a fully convolutional neural network that extracts keypoints and descriptors from an image. It is trained in a
+    self-supervised manner, using a combination of a photometric loss and a loss based on the homographic adaptation of
+    keypoints. It is made of a convolutional encoder and two decoders: one for keypoints and one for descriptors.
+    """
+
+    def __init__(self, config: SuperPointConfig) -> None:
+        super().__init__(config)
+
+        self.config = config
+
+        self.encoder = SuperPointEncoder(config)
+        self.keypoint_decoder = SuperPointInterestPointDecoder(config)
+        self.descriptor_decoder = SuperPointDescriptorDecoder(config)
+
+        self.post_init()
+
+    @add_start_docstrings_to_model_forward(SUPERPOINT_INPUTS_DOCSTRING)
+    def forward(
+        self,
+        pixel_values: torch.FloatTensor = None,
+        labels: Optional[torch.LongTensor] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Tuple, ImagePointDescriptionOutput]:
+        """
+        Examples:
+
+        ```python
+        >>> from transformers import AutoImageProcessor, SuperPointForKeypointDetection
+        >>> import torch
+        >>> from PIL import Image
+        >>> import requests
+
+        >>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
+        >>> image = Image.open(requests.get(url, stream=True).raw)
+
+        >>> processor = AutoImageProcessor.from_pretrained("magic-leap-community/superpoint")
+        >>> model = SuperPointForKeypointDetection.from_pretrained("magic-leap-community/superpoint")
+
+        >>> inputs = processor(image, return_tensors="pt")
+        >>> outputs = model(**inputs)
+        ```"""
+
+        if labels is not None:
+            raise ValueError(
+                f"SuperPoint is not trainable, no labels should be provided.Therefore, labels should be None but were {type(labels)}"
+            )
+
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if pixel_values is None:
+            raise ValueError("You have to specify pixel_values")
+
+        pixel_values = self.extract_one_channel_pixel_values(pixel_values)
+
+        batch_size = pixel_values.shape[0]
+
+        encoder_outputs = self.encoder(
+            pixel_values,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+
+        last_hidden_state = encoder_outputs[0]
+
+        list_keypoints_scores = [
+            self.keypoint_decoder(last_hidden_state[None, ...]) for last_hidden_state in last_hidden_state
+        ]
+
+        list_keypoints = [keypoints_scores[0] for keypoints_scores in list_keypoints_scores]
+        list_scores = [keypoints_scores[1] for keypoints_scores in list_keypoints_scores]
+
+        list_descriptors = [
+            self.descriptor_decoder(last_hidden_state[None, ...], keypoints[None, ...])
+            for last_hidden_state, keypoints in zip(last_hidden_state, list_keypoints)
+        ]
+
+        maximum_num_keypoints = max(keypoints.shape[0] for keypoints in list_keypoints)
+
+        keypoints = torch.zeros((batch_size, maximum_num_keypoints, 2), device=pixel_values.device)
+        scores = torch.zeros((batch_size, maximum_num_keypoints), device=pixel_values.device)
+        descriptors = torch.zeros(
+            (batch_size, maximum_num_keypoints, self.config.descriptor_decoder_dim),
+            device=pixel_values.device,
+        )
+        mask = torch.zeros((batch_size, maximum_num_keypoints), device=pixel_values.device, dtype=torch.int)
+
+        for i, (_keypoints, _scores, _descriptors) in enumerate(zip(list_keypoints, list_scores, list_descriptors)):
+            keypoints[i, : _keypoints.shape[0]] = _keypoints
+            scores[i, : _scores.shape[0]] = _scores
+            descriptors[i, : _descriptors.shape[0]] = _descriptors
+            mask[i, : _scores.shape[0]] = 1
+
+        hidden_states = encoder_outputs[1] if output_hidden_states else None
+        if not return_dict:
+            return tuple(
+                v for v in [last_hidden_state, keypoints, scores, descriptors, mask, hidden_states] if v is not None
+            )
+
+        return ImagePointDescriptionOutput(
+            last_hidden_state=last_hidden_state,
+            keypoints=keypoints,
+            scores=scores,
+            descriptors=descriptors,
+            mask=mask,
+            hidden_states=hidden_states,
+        )

src/transformers/models/vipllava/configuration_vipllava.py

@@ -13,6 +13,8 @@
 # limitations under the License.
 """ VipLlava model configuration"""
 
+import warnings
+
 from ...configuration_utils import PretrainedConfig
 from ...utils import logging
 from ..auto import CONFIG_MAPPING
@@ -51,9 +53,6 @@ class VipLlavaConfig(PretrainedConfig):
             The layer norm epsilon of the projector layernorm
         vision_feature_layers (`List[int]`, *optional*, defaults to `[-2, -5, -8, -11, 6]`):
             The list of layers to select the vision features from.
-        vocab_size (`int`, *optional*, defaults to 32000):
-            Vocabulary size of the VipLlava model. Defines the number of different tokens that can be represented by the
-            `inputs_ids` passed when calling [`~VipLlavaForConditionalGeneration`]
 
     Example:
 
@@ -88,7 +87,6 @@ class VipLlavaConfig(PretrainedConfig):
         projector_hidden_act="gelu",
         projector_layernorm_eps=1e-5,
         vision_feature_layers=[-2, -5, -8, -11, 6],
-        vocab_size=32000,
         **kwargs,
     ):
         self.ignore_index = ignore_index
@@ -96,7 +94,12 @@ class VipLlavaConfig(PretrainedConfig):
         self.projector_hidden_act = projector_hidden_act
         self.projector_layernorm_eps = projector_layernorm_eps
         self.vision_feature_layers = vision_feature_layers
-        self.vocab_size = vocab_size
+
+        if "vocab_size" in kwargs:
+            warnings.warn(
+                "The `vocab_size` argument is deprecated and will be removed in v4.42, since it can be inferred from the `text_config`. Passing this argument has no effect",
+                FutureWarning,
+            )
 
         self.vision_config = vision_config
 
@@ -116,15 +119,27 @@ class VipLlavaConfig(PretrainedConfig):
                 vocab_size=32000,
                 projection_dim=768,
             )
-        self.vocab_size = self.vocab_size
+
+        if isinstance(text_config, dict):
+            text_config["model_type"] = text_config["model_type"] if "model_type" in text_config else "llama"
+            text_config = CONFIG_MAPPING[text_config["model_type"]](**text_config)
+        elif text_config is None:
+            text_config = CONFIG_MAPPING["llama"]()
 
         self.text_config = text_config
-
-        if isinstance(self.text_config, dict):
-            text_config["model_type"] = text_config["model_type"] if "model_type" in text_config else "llama"
-            self.text_config = CONFIG_MAPPING[text_config["model_type"]](**text_config)
-            self.vocab_size = self.text_config.vocab_size
-        elif text_config is None:
-            self.text_config = CONFIG_MAPPING["llama"]()
+        self._vocab_size = self.text_config.vocab_size
 
         super().__init__(**kwargs)
+
+        @property
+        def vocab_size(self):
+            warnings.warn(
+                "The `vocab_size` attribute is deprecated and will be removed in v4.42, Please use `text_config.vocab_size` instead.",
+                FutureWarning,
+            )
+            return self._vocab_size
+
+        def to_dict(self):
+            output = super().to_dict()
+            output.pop("_vocab_size", None)
+            return output

src/transformers/models/vipllava/modeling_vipllava.py

@@ -248,7 +248,7 @@ class VipLlavaForConditionalGeneration(VipLlavaPreTrainedModel):
         self.vision_tower = AutoModel.from_config(config.vision_config)
 
         self.multi_modal_projector = VipLlavaMultiModalProjector(config)
-        self.vocab_size = config.vocab_size
+        self.vocab_size = config.text_config.vocab_size
         self.language_model = AutoModelForCausalLM.from_config(
             config.text_config, attn_implementation=config._attn_implementation
         )
@@ -280,7 +280,6 @@ class VipLlavaForConditionalGeneration(VipLlavaPreTrainedModel):
         model_embeds = self.language_model.resize_token_embeddings(new_num_tokens, pad_to_multiple_of)
         # update vocab size
         self.config.text_config.vocab_size = model_embeds.num_embeddings
-        self.config.vocab_size = model_embeds.num_embeddings
         self.vocab_size = model_embeds.num_embeddings
         return model_embeds
 
@@ -348,6 +347,12 @@ class VipLlavaForConditionalGeneration(VipLlavaPreTrainedModel):
         final_attention_mask |= image_to_overwrite
         position_ids = (final_attention_mask.cumsum(-1) - 1).masked_fill_((final_attention_mask == 0), 1)
 
+        # 6. Mask out the embedding at padding positions, as we later use the past_key_value value to determine the non-attended tokens.
+        batch_indices, pad_indices = torch.where(input_ids == self.pad_token_id)
+        indices_to_mask = new_token_positions[batch_indices, pad_indices]
+
+        final_embedding[batch_indices, indices_to_mask] = 0
+
         if labels is None:
             final_labels = None
 
@@ -443,11 +448,11 @@ class VipLlavaForConditionalGeneration(VipLlavaPreTrainedModel):
                     # Sum all dimensions of head_dim (-1) to avoid random errors such as: https://github.com/huggingface/transformers/pull/28032#issuecomment-1863691941
                     batch_index, non_attended_tokens = torch.where(first_layer_past_key_value.float().sum(-1) == 0)
 
-                    # Get the target length
-                    target_seqlen = first_layer_past_key_value.shape[-2] + 1
+                    target_length = input_ids.shape[1]
+                    past_length = first_layer_past_key_value.shape[-1]
 
                     extended_attention_mask = torch.ones(
-                        (attention_mask.shape[0], target_seqlen - attention_mask.shape[1]),
+                        (attention_mask.shape[0], past_length),
                         dtype=attention_mask.dtype,
                         device=attention_mask.device,
                     )
@@ -462,7 +467,7 @@ class VipLlavaForConditionalGeneration(VipLlavaPreTrainedModel):
                     # Zero-out the places where we don't need to attend
                     extended_attention_mask[new_batch_index, new_non_attended_tokens] = 0
 
-                    attention_mask = torch.cat((attention_mask, extended_attention_mask), dim=1)
+                    attention_mask = torch.cat((extended_attention_mask, attention_mask[:, -target_length:]), dim=1)
                     position_ids = torch.sum(attention_mask, dim=1).unsqueeze(-1) - 1
 
         outputs = self.language_model(

src/transformers/optimization.py

@@ -385,7 +385,8 @@ def get_scheduler(
                 scheduler_dict[param].step()
 
         for param in optimizer_dict.keys():
-            param.register_post_accumulate_grad_hook(scheduler_hook)
+            if param.requires_grad:
+                param.register_post_accumulate_grad_hook(scheduler_hook)
 
         return LayerWiseDummyScheduler()
 

src/transformers/tokenization_utils_base.py

@@ -1819,7 +1819,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin, PushToHubMixin):
         except ImportError:
             raise ImportError("apply_chat_template requires jinja2 to be installed.")
 
-        if version.parse(jinja2.__version__) <= version.parse("3.0.0"):
+        if version.parse(jinja2.__version__) < version.parse("3.0.0"):
             raise ImportError(
                 "apply_chat_template requires jinja2>=3.0.0 to be installed. Your version is " f"{jinja2.__version__}."
             )