Release: v4.36.2

* add multi-node traning setting

* fix style

.github/workflows/delete_doc_comment.yml

@@ -1,14 +0,0 @@
-name: Delete doc comment
-
-on:
-  workflow_run:
-    workflows: ["Delete doc comment trigger"]
-    types:
-      - completed
-
-
-jobs:
-  delete:
-    uses: huggingface/doc-builder/.github/workflows/delete_doc_comment.yml@main
-    secrets:
-      comment_bot_token: ${{ secrets.COMMENT_BOT_TOKEN }}

.github/workflows/self-scheduled.yml

@@ -494,5 +494,5 @@ jobs:
         if: ${{ always() }}
         uses: actions/upload-artifact@v3
         with:
-          name: test_failure_tables
-          path: test_failure_tables
+          name: prev_ci_results
+          path: prev_ci_results

README.md

@@ -321,7 +321,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[CLAP](https://huggingface.co/docs/transformers/model_doc/clap)** (from LAION-AI) released with the paper [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://arxiv.org/abs/2211.06687) by Yusong Wu, Ke Chen, Tianyu Zhang, Yuchen Hui, Taylor Berg-Kirkpatrick, Shlomo Dubnov.
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (from OpenAI) released with the paper [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) by 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)** (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/main/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
+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. **[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.
@@ -397,6 +397,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 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/) 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. **[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.
@@ -416,6 +417,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
 1. **[MGP-STR](https://huggingface.co/docs/transformers/model_doc/mgp-str)** (from Alibaba Research) released with the paper [Multi-Granularity Prediction for Scene Text Recognition](https://arxiv.org/abs/2209.03592) by Peng Wang, Cheng Da, and Cong Yao.
 1. **[Mistral](https://huggingface.co/docs/transformers/model_doc/mistral)** (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. **[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)** (from Studio Ousia) released with the paper [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) by Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka.
 1. **[MMS](https://huggingface.co/docs/transformers/model_doc/mms)** (from Facebook) released with the paper [Scaling Speech Technology to 1,000+ Languages](https://arxiv.org/abs/2305.13516) by 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)** (from CMU/Google Brain) released with the paper [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) by Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou.
@@ -440,12 +442,13 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[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. **[PatchTST](https://huggingface.co/docs/transformers/main/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. **[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/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.
 1. **[Persimmon](https://huggingface.co/docs/transformers/model_doc/persimmon)** (from ADEPT) released in a [blog post](https://www.adept.ai/blog/persimmon-8b) by Erich Elsen, Augustus Odena, Maxwell Nye, Sağnak Taşırlar, Tri Dao, Curtis Hawthorne, Deepak Moparthi, Arushi Somani.
-1. **[Phi](https://huggingface.co/docs/transformers/main/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. **[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)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.
 1. **[Pix2Struct](https://huggingface.co/docs/transformers/model_doc/pix2struct)** (from Google) released with the paper [Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding](https://arxiv.org/abs/2210.03347) by 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)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
@@ -466,7 +469,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (from ZhuiyiTechnology), released together with the paper [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 on [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/main/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. **[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. **[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.
@@ -497,7 +500,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 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/main/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. **[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/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.

README_es.md

@@ -296,7 +296,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[CLAP](https://huggingface.co/docs/transformers/model_doc/clap)** (from LAION-AI) released with the paper [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://arxiv.org/abs/2211.06687) by Yusong Wu, Ke Chen, Tianyu Zhang, Yuchen Hui, Taylor Berg-Kirkpatrick, Shlomo Dubnov.
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (from OpenAI) released with the paper [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) by 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)** (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/main/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker. 
+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. **[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.
@@ -372,6 +372,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 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. **[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. **[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.
@@ -391,6 +392,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
 1. **[MGP-STR](https://huggingface.co/docs/transformers/model_doc/mgp-str)** (from Alibaba Research) released with the paper [Multi-Granularity Prediction for Scene Text Recognition](https://arxiv.org/abs/2209.03592) by Peng Wang, Cheng Da, and Cong Yao.
 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)** (from Studio Ousia) released with the paper [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) by Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka.
 1. **[MMS](https://huggingface.co/docs/transformers/model_doc/mms)** (from Facebook) released with the paper [Scaling Speech Technology to 1,000+ Languages](https://arxiv.org/abs/2305.13516) by 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)** (from CMU/Google Brain) released with the paper [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) by Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou.
@@ -415,12 +417,13 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[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. **[PatchTST](https://huggingface.co/docs/transformers/main/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. **[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. **[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.
 1. **[Persimmon](https://huggingface.co/docs/transformers/model_doc/persimmon)** (from ADEPT) released with the paper [blog post](https://www.adept.ai/blog/persimmon-8b) by Erich Elsen, Augustus Odena, Maxwell Nye, Sağnak Taşırlar, Tri Dao, Curtis Hawthorne, Deepak Moparthi, Arushi Somani.
-1. **[Phi](https://huggingface.co/docs/transformers/main/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. **[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)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.
 1. **[Pix2Struct](https://huggingface.co/docs/transformers/model_doc/pix2struct)** (from Google) released with the paper [Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding](https://arxiv.org/abs/2210.03347) by 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)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
@@ -441,7 +444,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (from ZhuiyiTechnology), released together with the paper [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/main/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. **[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. **[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.
@@ -472,7 +475,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 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/main/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. **[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/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.

README_hd.md

@@ -270,7 +270,7 @@ conda install -c huggingface transformers
 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. **[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/main/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker. 
+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. **[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. **[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।
@@ -346,6 +346,7 @@ conda install -c huggingface transformers
 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. **[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. **[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. **[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।
@@ -365,6 +366,7 @@ conda install -c huggingface transformers
 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. **[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. **[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।
@@ -389,12 +391,13 @@ conda install -c huggingface transformers
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (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. **[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. **[PatchTST](https://huggingface.co/docs/transformers/main/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. **[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. **[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. **[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/main/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. **[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. **[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।
@@ -415,7 +418,7 @@ conda install -c huggingface transformers
 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. **[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/main/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. **[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. **[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।
@@ -446,7 +449,7 @@ conda install -c huggingface transformers
 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. **[UnivNet](https://huggingface.co/docs/transformers/main/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. **[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।

README_ja.md

@@ -330,7 +330,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[CLAP](https://huggingface.co/docs/transformers/model_doc/clap)** (LAION-AI から) Yusong Wu, Ke Chen, Tianyu Zhang, Yuchen Hui, Taylor Berg-Kirkpatrick, Shlomo Dubnov. から公開された研究論文 [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://arxiv.org/abs/2211.06687)
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (OpenAI から) 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 から公開された研究論文: [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020)
 1. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (University of Göttingen から) Timo Lüddecke and Alexander Ecker から公開された研究論文: [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003)
-1. **[CLVP](https://huggingface.co/docs/transformers/main/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker. 
+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. **[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)
@@ -406,6 +406,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 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. **[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. **[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)
@@ -425,6 +426,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (NVIDIA から) Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro から公開された研究論文: [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053)
 1. **[MGP-STR](https://huggingface.co/docs/transformers/model_doc/mgp-str)** (Alibaba Research から) Peng Wang, Cheng Da, and Cong Yao. から公開された研究論文 [Multi-Granularity Prediction for Scene Text Recognition](https://arxiv.org/abs/2209.03592)
 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)** (Studio Ousia から) Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka から公開された研究論文: [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151)
 1. **[MMS](https://huggingface.co/docs/transformers/model_doc/mms)** (Facebook から) 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. から公開された研究論文 [Scaling Speech Technology to 1,000+ Languages](https://arxiv.org/abs/2305.13516)
 1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (CMU/Google Brain から) Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou から公開された研究論文: [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984)
@@ -449,12 +451,13 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (Meta AI から) Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al から公開された研究論文: [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068)
 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. **[PatchTST](https://huggingface.co/docs/transformers/main/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. **[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. **[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)
 1. **[Persimmon](https://huggingface.co/docs/transformers/model_doc/persimmon)** (ADEPT から) Erich Elsen, Augustus Odena, Maxwell Nye, Sağnak Taşırlar, Tri Dao, Curtis Hawthorne, Deepak Moparthi, Arushi Somani. から公開された研究論文 [blog post](https://www.adept.ai/blog/persimmon-8b)
-1. **[Phi](https://huggingface.co/docs/transformers/main/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. **[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 から) Dat Quoc Nguyen and Anh Tuan Nguyen から公開された研究論文: [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/)
 1. **[Pix2Struct](https://huggingface.co/docs/transformers/model_doc/pix2struct)** (Google から) Kenton Lee, Mandar Joshi, Iulia Turc, Hexiang Hu, Fangyu Liu, Julian Eisenschlos, Urvashi Khandelwal, Peter Shaw, Ming-Wei Chang, Kristina Toutanova. から公開された研究論文 [Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding](https://arxiv.org/abs/2210.03347)
 1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (UCLA NLP から) Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang から公開された研究論文: [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333)
@@ -475,7 +478,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (ZhuiyiTechnology から), Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu から公開された研究論文: [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/abs/2104.09864)
 1. **[RWKV](https://huggingface.co/docs/transformers/model_doc/rwkv)** (Bo Peng から) Bo Peng. から公開された研究論文 [this repo](https://github.com/BlinkDL/RWKV-LM)
 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/main/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. **[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. **[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)
@@ -506,7 +509,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 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)
 1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (Microsoft Research から) Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu から公開された研究論文: [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752)
-1. **[UnivNet](https://huggingface.co/docs/transformers/main/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. **[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 から) Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, Jian Sun. から公開された研究論文 [Unified Perceptual Parsing for Scene Understanding](https://arxiv.org/abs/1807.10221)
 1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (Tsinghua University and Nankai University から) Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu から公開された研究論文: [Visual Attention Network](https://arxiv.org/abs/2202.09741)
 1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (Multimedia Computing Group, Nanjing University から) Zhan Tong, Yibing Song, Jue Wang, Limin Wang から公開された研究論文: [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602)

README_ko.md

@@ -245,7 +245,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[CLAP](https://huggingface.co/docs/transformers/model_doc/clap)** (LAION-AI 에서 제공)은 Yusong Wu, Ke Chen, Tianyu Zhang, Yuchen Hui, Taylor Berg-Kirkpatrick, Shlomo Dubnov.의 [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://arxiv.org/abs/2211.06687)논문과 함께 발표했습니다.
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (OpenAI 에서) 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 의 [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) 논문과 함께 발표했습니다.
 1. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (University of Göttingen 에서) Timo Lüddecke and Alexander Ecker 의 [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) 논문과 함께 발표했습니다.
-1. **[CLVP](https://huggingface.co/docs/transformers/main/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker. 
+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. **[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.
@@ -321,6 +321,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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. **[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. **[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.
@@ -340,6 +341,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (NVIDIA 에서) Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro 의 [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) 논문과 함께 발표했습니다.
 1. **[MGP-STR](https://huggingface.co/docs/transformers/model_doc/mgp-str)** (Alibaba Research 에서 제공)은 Peng Wang, Cheng Da, and Cong Yao.의 [Multi-Granularity Prediction for Scene Text Recognition](https://arxiv.org/abs/2209.03592)논문과 함께 발표했습니다.
 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)** (Studio Ousia 에서) Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka 의 [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) 논문과 함께 발표했습니다.
 1. **[MMS](https://huggingface.co/docs/transformers/model_doc/mms)** (Facebook 에서 제공)은 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.의 [Scaling Speech Technology to 1,000+ Languages](https://arxiv.org/abs/2305.13516)논문과 함께 발표했습니다.
 1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (CMU/Google Brain 에서) Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou 의 [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) 논문과 함께 발표했습니다.
@@ -364,12 +366,13 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (Meta AI 에서) Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al 의 [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) 논문과 함께 발표했습니다.
 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. **[PatchTST](https://huggingface.co/docs/transformers/main/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. **[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. **[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.
 1. **[Persimmon](https://huggingface.co/docs/transformers/model_doc/persimmon)** (ADEPT 에서 제공)은 Erich Elsen, Augustus Odena, Maxwell Nye, Sağnak Taşırlar, Tri Dao, Curtis Hawthorne, Deepak Moparthi, Arushi Somani.의 [blog post](https://www.adept.ai/blog/persimmon-8b)논문과 함께 발표했습니다.
-1. **[Phi](https://huggingface.co/docs/transformers/main/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. **[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 에서) Dat Quoc Nguyen and Anh Tuan Nguyen 의 [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) 논문과 함께 발표했습니다.
 1. **[Pix2Struct](https://huggingface.co/docs/transformers/model_doc/pix2struct)** (Google 에서 제공)은 Kenton Lee, Mandar Joshi, Iulia Turc, Hexiang Hu, Fangyu Liu, Julian Eisenschlos, Urvashi Khandelwal, Peter Shaw, Ming-Wei Chang, Kristina Toutanova.의 [Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding](https://arxiv.org/abs/2210.03347)논문과 함께 발표했습니다.
 1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (UCLA NLP 에서) Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang 의 [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) 논문과 함께 발표했습니다.
@@ -390,7 +393,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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. **[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/main/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. **[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. **[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.
@@ -421,7 +424,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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/main/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. **[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. **[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.

README_zh-hans.md

@@ -269,7 +269,7 @@ conda install -c huggingface transformers
 1. **[CLAP](https://huggingface.co/docs/transformers/model_doc/clap)** (来自 LAION-AI) 伴随论文 [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://arxiv.org/abs/2211.06687) 由 Yusong Wu, Ke Chen, Tianyu Zhang, Yuchen Hui, Taylor Berg-Kirkpatrick, Shlomo Dubnov 发布。
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (来自 OpenAI) 伴随论文 [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) 由 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 发布。
 1. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (来自 University of Göttingen) 伴随论文 [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) 由 Timo Lüddecke and Alexander Ecker 发布。
-1. **[CLVP](https://huggingface.co/docs/transformers/main/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker. 
+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. **[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。
@@ -345,6 +345,7 @@ conda install -c huggingface transformers
 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. **[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. **[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。
@@ -364,6 +365,7 @@ conda install -c huggingface transformers
 1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (来自 NVIDIA) 伴随论文 [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) 由 Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro 发布。
 1. **[MGP-STR](https://huggingface.co/docs/transformers/model_doc/mgp-str)** (来自 Alibaba Research) 伴随论文 [Multi-Granularity Prediction for Scene Text Recognition](https://arxiv.org/abs/2209.03592) 由 Peng Wang, Cheng Da, and Cong Yao 发布。
 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)** (来自 Studio Ousia) 伴随论文 [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) 由 Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka 发布。
 1. **[MMS](https://huggingface.co/docs/transformers/model_doc/mms)** (来自 Facebook) 伴随论文 [Scaling Speech Technology to 1,000+ Languages](https://arxiv.org/abs/2305.13516) 由 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 发布。
 1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (来自 CMU/Google Brain) 伴随论文 [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) 由 Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou 发布。
@@ -388,12 +390,13 @@ conda install -c huggingface transformers
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (来自 Meta AI) 伴随论文 [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) 由 Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al 发布。
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (来自 Google AI) 伴随论文 [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) 由 Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby 发布。
 1. **[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. **[PatchTST](https://huggingface.co/docs/transformers/main/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. **[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. **[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。
 1. **[Persimmon](https://huggingface.co/docs/transformers/model_doc/persimmon)** (来自 ADEPT) 伴随论文 [blog post](https://www.adept.ai/blog/persimmon-8b) 由 Erich Elsen, Augustus Odena, Maxwell Nye, Sağnak Taşırlar, Tri Dao, Curtis Hawthorne, Deepak Moparthi, Arushi Somani 发布。
-1. **[Phi](https://huggingface.co/docs/transformers/main/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. **[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)** (来自 VinAI Research) 伴随论文 [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) 由 Dat Quoc Nguyen and Anh Tuan Nguyen 发布。
 1. **[Pix2Struct](https://huggingface.co/docs/transformers/model_doc/pix2struct)** (来自 Google) 伴随论文 [Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding](https://arxiv.org/abs/2210.03347) 由 Kenton Lee, Mandar Joshi, Iulia Turc, Hexiang Hu, Fangyu Liu, Julian Eisenschlos, Urvashi Khandelwal, Peter Shaw, Ming-Wei Chang, Kristina Toutanova 发布。
 1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (来自 UCLA NLP) 伴随论文 [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) 由 Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang 发布。
@@ -414,7 +417,7 @@ conda install -c huggingface transformers
 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. **[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/main/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. **[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. **[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。
@@ -445,7 +448,7 @@ conda install -c huggingface transformers
 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/main/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. **[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. **[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。

README_zh-hant.md

@@ -281,7 +281,7 @@ conda install -c huggingface transformers
 1. **[CLAP](https://huggingface.co/docs/transformers/model_doc/clap)** (from LAION-AI) released with the paper [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://arxiv.org/abs/2211.06687) by Yusong Wu, Ke Chen, Tianyu Zhang, Yuchen Hui, Taylor Berg-Kirkpatrick, Shlomo Dubnov.
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (from OpenAI) released with the paper [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) by 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)** (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/main/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker. 
+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. **[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.
@@ -357,6 +357,7 @@ conda install -c huggingface transformers
 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. **[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. **[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.
@@ -376,6 +377,7 @@ conda install -c huggingface transformers
 1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
 1. **[MGP-STR](https://huggingface.co/docs/transformers/model_doc/mgp-str)** (from Alibaba Research) released with the paper [Multi-Granularity Prediction for Scene Text Recognition](https://arxiv.org/abs/2209.03592) by Peng Wang, Cheng Da, and Cong Yao.
 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)** (from Studio Ousia) released with the paper [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) by Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka.
 1. **[MMS](https://huggingface.co/docs/transformers/model_doc/mms)** (from Facebook) released with the paper [Scaling Speech Technology to 1,000+ Languages](https://arxiv.org/abs/2305.13516) by 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)** (from CMU/Google Brain) released with the paper [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) by Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou.
@@ -400,12 +402,13 @@ conda install -c huggingface transformers
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[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. **[PatchTST](https://huggingface.co/docs/transformers/main/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. **[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. **[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.
 1. **[Persimmon](https://huggingface.co/docs/transformers/model_doc/persimmon)** (from ADEPT) released with the paper [blog post](https://www.adept.ai/blog/persimmon-8b) by Erich Elsen, Augustus Odena, Maxwell Nye, Sağnak Taşırlar, Tri Dao, Curtis Hawthorne, Deepak Moparthi, Arushi Somani.
-1. **[Phi](https://huggingface.co/docs/transformers/main/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. **[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)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.
 1. **[Pix2Struct](https://huggingface.co/docs/transformers/model_doc/pix2struct)** (from Google) released with the paper [Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding](https://arxiv.org/abs/2210.03347) by 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)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
@@ -426,7 +429,7 @@ conda install -c huggingface transformers
 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. **[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/main/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. **[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. **[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.
@@ -457,7 +460,7 @@ conda install -c huggingface transformers
 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/main/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. **[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. **[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.

docker/transformers-all-latest-gpu/Dockerfile

@@ -56,7 +56,7 @@ RUN python3 -m pip install --no-cache-dir auto-gptq --extra-index-url https://hu
 RUN python3 -m pip install --no-cache-dir einops
 
 # Add autoawq for quantization testing
-RUN python3 -m pip install --no-cache-dir https://github.com/casper-hansen/AutoAWQ/releases/download/v0.1.6/autoawq-0.1.6+cu118-cp38-cp38-linux_x86_64.whl
+RUN python3 -m pip install --no-cache-dir https://github.com/casper-hansen/AutoAWQ/releases/download/v0.1.7/autoawq-0.1.7+cu118-cp38-cp38-linux_x86_64.whl
 
 # For bettertransformer + gptq 
 RUN python3 -m pip install --no-cache-dir git+https://github.com/huggingface/optimum@main#egg=optimum

docker/transformers-pytorch-deepspeed-latest-gpu/Dockerfile

@@ -1,12 +1,12 @@
-# https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-22-12.html#rel-22-12
-FROM nvcr.io/nvidia/pytorch:22.12-py3
+# https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-23-11.html#rel-23-11
+FROM nvcr.io/nvidia/pytorch:23.11-py3
 LABEL maintainer="Hugging Face"
 
 ARG DEBIAN_FRONTEND=noninteractive
 
 ARG PYTORCH='2.1.0'
 # Example: `cu102`, `cu113`, etc.
-ARG CUDA='cu118'
+ARG CUDA='cu121'
 
 RUN apt -y update
 RUN apt install -y libaio-dev

docker/transformers-pytorch-deepspeed-nightly-gpu/Dockerfile

@@ -1,11 +1,11 @@
-# https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-22-12.html#rel-22-12
-FROM nvcr.io/nvidia/pytorch:22.12-py3
+# https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-23-11.html#rel-23-11
+FROM nvcr.io/nvidia/pytorch:23.11-py3
 LABEL maintainer="Hugging Face"
 
 ARG DEBIAN_FRONTEND=noninteractive
 
 # Example: `cu102`, `cu113`, etc.
-ARG CUDA='cu118'
+ARG CUDA='cu121'
 
 RUN apt -y update
 RUN apt install -y libaio-dev

docker/transformers-pytorch-gpu/Dockerfile

@@ -1,4 +1,4 @@
-FROM nvidia/cuda:11.8.0-cudnn8-devel-ubuntu20.04
+FROM nvidia/cuda:12.1.0-cudnn8-devel-ubuntu20.04
 LABEL maintainer="Hugging Face"
 
 ARG DEBIAN_FRONTEND=noninteractive
@@ -15,7 +15,7 @@ ARG PYTORCH='2.1.0'
 ARG TORCH_VISION=''
 ARG TORCH_AUDIO=''
 # Example: `cu102`, `cu113`, etc.
-ARG CUDA='cu118'
+ARG CUDA='cu121'
 
 RUN [ ${#PYTORCH} -gt 0 ] && VERSION='torch=='$PYTORCH'.*' ||  VERSION='torch'; python3 -m pip install --no-cache-dir -U $VERSION --extra-index-url https://download.pytorch.org/whl/$CUDA
 RUN [ ${#TORCH_VISION} -gt 0 ] && VERSION='torchvision=='TORCH_VISION'.*' ||  VERSION='torchvision'; python3 -m pip install --no-cache-dir -U $VERSION --extra-index-url https://download.pytorch.org/whl/$CUDA

docs/source/en/_toctree.yml

@@ -396,6 +396,8 @@
         title: MegatronGPT2
       - local: model_doc/mistral
         title: Mistral
+      - local: model_doc/mixtral
+        title: Mixtral
       - local: model_doc/mluke
         title: mLUKE
       - local: model_doc/mobilebert
@@ -703,6 +705,8 @@
         title: LayoutXLM
       - local: model_doc/lilt
         title: LiLT
+      - local: model_doc/llava
+        title: Llava
       - local: model_doc/lxmert
         title: LXMERT
       - local: model_doc/matcha
@@ -757,6 +761,8 @@
         title: Autoformer
       - local: model_doc/informer
         title: Informer
+      - local: model_doc/patchtsmixer
+        title: PatchTSMixer
       - local: model_doc/patchtst
         title: PatchTST
       - local: model_doc/time_series_transformer

docs/source/en/index.md

@@ -1,4 +1,4 @@
-<!--Copyright 2020 The HuggingFace Team. All rights reserved.
+        <!--Copyright 2020 The HuggingFace Team. All rights reserved.
 
 Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
 the License. You may obtain a copy of the License at
@@ -94,7 +94,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                       [CLIPSeg](model_doc/clipseg)                       |       ✅        |         ❌         |      ❌      |
 |                          [CLVP](model_doc/clvp)                          |       ✅        |         ❌         |      ❌      |
 |                       [CodeGen](model_doc/codegen)                       |       ✅        |         ❌         |      ❌      |
-|                    [CodeLlama](model_doc/code_llama)                     |       ✅        |         ❌         |      ❌      |
+|                    [CodeLlama](model_doc/code_llama)                     |       ✅        |         ❌         |      ✅      |
 |              [Conditional DETR](model_doc/conditional_detr)              |       ✅        |         ❌         |      ❌      |
 |                      [ConvBERT](model_doc/convbert)                      |       ✅        |         ✅         |      ❌      |
 |                      [ConvNeXT](model_doc/convnext)                      |       ✅        |         ✅         |      ❌      |
@@ -167,8 +167,9 @@ Flax), PyTorch, and/or TensorFlow.
 |                           [LED](model_doc/led)                           |       ✅        |         ✅         |      ❌      |
 |                         [LeViT](model_doc/levit)                         |       ✅        |         ❌         |      ❌      |
 |                          [LiLT](model_doc/lilt)                          |       ✅        |         ❌         |      ❌      |
-|                         [LLaMA](model_doc/llama)                         |       ✅        |         ❌         |      ❌      |
-|                        [Llama2](model_doc/llama2)                        |       ✅        |         ❌         |      ❌      |
+|                         [LLaMA](model_doc/llama)                         |       ✅        |         ❌         |      ✅      |
+|                        [Llama2](model_doc/llama2)                        |       ✅        |         ❌         |      ✅      |
+|                         [LLaVa](model_doc/llava)                         |       ✅        |         ❌         |      ❌      |
 |                    [Longformer](model_doc/longformer)                    |       ✅        |         ✅         |      ❌      |
 |                        [LongT5](model_doc/longt5)                        |       ✅        |         ❌         |      ✅      |
 |                          [LUKE](model_doc/luke)                          |       ✅        |         ❌         |      ❌      |
@@ -188,6 +189,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                 [Megatron-GPT2](model_doc/megatron_gpt2)                 |       ✅        |         ✅         |      ✅      |
 |                       [MGP-STR](model_doc/mgp-str)                       |       ✅        |         ❌         |      ❌      |
 |                       [Mistral](model_doc/mistral)                       |       ✅        |         ❌         |      ❌      |
+|                       [Mixtral](model_doc/mixtral)                       |       ✅        |         ❌         |      ❌      |
 |                         [mLUKE](model_doc/mluke)                         |       ✅        |         ❌         |      ❌      |
 |                           [MMS](model_doc/mms)                           |       ✅        |         ✅         |      ✅      |
 |                    [MobileBERT](model_doc/mobilebert)                    |       ✅        |         ✅         |      ❌      |
@@ -214,6 +216,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                           [OPT](model_doc/opt)                           |       ✅        |         ✅         |      ✅      |
 |                       [OWL-ViT](model_doc/owlvit)                        |       ✅        |         ❌         |      ❌      |
 |                         [OWLv2](model_doc/owlv2)                         |       ✅        |         ❌         |      ❌      |
+|                  [PatchTSMixer](model_doc/patchtsmixer)                  |       ✅        |         ❌         |      ❌      |
 |                      [PatchTST](model_doc/patchtst)                      |       ✅        |         ❌         |      ❌      |
 |                       [Pegasus](model_doc/pegasus)                       |       ✅        |         ✅         |      ✅      |
 |                     [PEGASUS-X](model_doc/pegasus_x)                     |       ✅        |         ❌         |      ❌      |

docs/source/en/internal/generation_utils.md

@@ -368,3 +368,20 @@ A [`Constraint`] can be used to force the generation to include specific tokens
 [[autodoc]] TextStreamer
 
 [[autodoc]] TextIteratorStreamer
+
+## Caches
+
+[[autodoc]] Cache
+    - update
+
+[[autodoc]] DynamicCache
+    - update
+    - get_seq_length
+    - reorder_cache
+    - to_legacy_cache
+    - from_legacy_cache
+
+[[autodoc]] SinkCache
+    - update
+    - get_seq_length
+    - reorder_cache

docs/source/en/llm_tutorial.md

@@ -250,7 +250,7 @@ While the autoregressive generation process is relatively straightforward, makin
 1. [Guide](generation_strategies) on how to control different generation methods, how to set up the generation configuration file, and how to stream the output;
 2. [Guide](chat_templating) on the prompt template for chat LLMs;
 3. [Guide](tasks/prompting) on to get the most of prompt design;
-4. API reference on [`~generation.GenerationConfig`], [`~generation.GenerationMixin.generate`], and [generate-related classes](internal/generation_utils).
+4. API reference on [`~generation.GenerationConfig`], [`~generation.GenerationMixin.generate`], and [generate-related classes](internal/generation_utils). Most of the classes, including the logits processors, have usage examples!
 
 ### LLM leaderboards
 

docs/source/en/llm_tutorial_optimization.md

@@ -441,7 +441,7 @@ flush()
 ```
 
 For comparison, let's run the same function, but enable Flash Attention instead.
-To do so, we convert the model to [BetterTransformers](https://huggingface.co/docs/optimum/bettertransformer/overview) and by doing so enabling PyTorch's [SDPA self-attention](https://pytorch.org/docs/master/generated/torch.nn.functional.scaled_dot_product_attention) which in turn is based on Flash Attention.
+To do so, we convert the model to [BetterTransformer](https://huggingface.co/docs/optimum/bettertransformer/overview) and by doing so enabling PyTorch's [SDPA self-attention](https://pytorch.org/docs/master/generated/torch.nn.functional.scaled_dot_product_attention) which in turn is able to use Flash Attention.
 
 ```python
 model.to_bettertransformer()

docs/source/en/model_doc/bark.md

@@ -83,10 +83,10 @@ pip install -U flash-attn --no-build-isolation
 
 ##### Usage
 
-To load a model using Flash Attention 2, we can pass the `use_flash_attention_2` flag to [`.from_pretrained`](https://huggingface.co/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel.from_pretrained). We'll also load the model in half-precision (e.g. `torch.float16`), since it results in almost no degradation to audio quality but significantly lower memory usage and faster inference:
+To load a model using Flash Attention 2, we can pass the `attn_implementation="flash_attention_2"` flag to [`.from_pretrained`](https://huggingface.co/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel.from_pretrained). We'll also load the model in half-precision (e.g. `torch.float16`), since it results in almost no degradation to audio quality but significantly lower memory usage and faster inference:
 
 ```python
-model = BarkModel.from_pretrained("suno/bark-small", torch_dtype=torch.float16, use_flash_attention_2=True).to(device)
+model = BarkModel.from_pretrained("suno/bark-small", torch_dtype=torch.float16, attn_implementation="flash_attention_2").to(device)
 ```
 
 ##### Performance comparison
@@ -114,7 +114,7 @@ import torch
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
 # load in fp16 and use Flash Attention 2
-model = BarkModel.from_pretrained("suno/bark-small", torch_dtype=torch.float16, use_flash_attention_2=True).to(device)
+model = BarkModel.from_pretrained("suno/bark-small", torch_dtype=torch.float16, attn_implementation="flash_attention_2").to(device)
 
 # enable CPU offload
 model.enable_cpu_offload()

docs/source/en/model_doc/biogpt.md

@@ -18,8 +18,7 @@ rendered properly in your Markdown viewer.
 
 ## Overview
 
-The BioGPT model was proposed in [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. BioGPT is a domain-specific generative pre-trained Transformer language model for biomedical text generation and mining. BioGPT follows the Transformer language model backbone, and is pre-trained on 15M PubMed abstracts from scratch.
+The BioGPT model was proposed in [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. BioGPT is a domain-specific generative pre-trained Transformer language model for biomedical text generation and mining. BioGPT follows the Transformer language model backbone, and is pre-trained on 15M PubMed abstracts from scratch.
 
 The abstract from the paper is the following:
 

docs/source/en/model_doc/distilbert.md

@@ -153,7 +153,7 @@ To load and run a model using Flash Attention 2, refer to the snippet below:
 >>> device = "cuda" # the device to load the model onto
 
 >>> tokenizer = AutoTokenizer.from_pretrained('distilbert-base-uncased')
->>> model = AutoModel.from_pretrained("distilbert-base-uncased", torch_dtype=torch.float16, use_flash_attention_2=True)
+>>> model = AutoModel.from_pretrained("distilbert-base-uncased", torch_dtype=torch.float16, attn_implementation="flash_attention_2")
 
 >>> text = "Replace me by any text you'd like."
 

docs/source/en/model_doc/gpt_bigcode.md

@@ -59,7 +59,7 @@ To load and run a model using Flash Attention 2, refer to the snippet below:
 >>> from transformers import AutoModelForCausalLM, AutoTokenizer
 >>> device = "cuda" # the device to load the model onto
 
->>> model = AutoModelForCausalLM.from_pretrained("bigcode/gpt_bigcode-santacoder", torch_dtype=torch.float16, use_flash_attention_2=True)
+>>> model = AutoModelForCausalLM.from_pretrained("bigcode/gpt_bigcode-santacoder", torch_dtype=torch.float16, attn_implementation="flash_attention_2")
 >>> tokenizer = AutoTokenizer.from_pretrained("bigcode/gpt_bigcode-santacoder")
 
 >>> prompt = "def hello_world():"

docs/source/en/model_doc/gpt_neo.md

@@ -67,7 +67,7 @@ To load and run a model using Flash Attention 2, refer to the snippet below:
 >>> from transformers import AutoModelForCausalLM, AutoTokenizer
 >>> device = "cuda" # the device to load the model onto
 
->>> model = AutoModelForCausalLM.from_pretrained("EleutherAI/gpt-neo-2.7B", torch_dtype=torch.float16, use_flash_attention_2=True)
+>>> model = AutoModelForCausalLM.from_pretrained("EleutherAI/gpt-neo-2.7B", torch_dtype=torch.float16, attn_implementation="flash_attention_2")
 >>> tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neo-2.7B")
 
 >>> prompt = "def hello_world():"

docs/source/en/model_doc/gpt_neox.md

@@ -61,6 +61,40 @@ The `generate()` method can be used to generate text using GPT Neo model.
 >>> gen_text = tokenizer.batch_decode(gen_tokens)[0]
 ```
 
+## Using Flash Attention 2
+
+Flash Attention 2 is an faster, optimized version of the model.
+
+### Installation 
+
+First, check whether your hardware is compatible with Flash Attention 2. The latest list of compatible hardware can be found in the [official documentation](https://github.com/Dao-AILab/flash-attention#installation-and-features). If your hardware is not compatible with Flash Attention 2, you can still benefit from attention kernel optimisations through Better Transformer support covered [above](https://huggingface.co/docs/transformers/main/en/model_doc/bark#using-better-transformer).
+
+Next, [install](https://github.com/Dao-AILab/flash-attention#installation-and-features) the latest version of Flash Attention 2:
+
+```bash
+pip install -U flash-attn --no-build-isolation
+```
+
+### Usage
+
+To load a model using Flash Attention 2, we can pass the argument `attn_implementation="flash_attention_2"` to [`.from_pretrained`](https://huggingface.co/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel.from_pretrained). We'll also load the model in half-precision (e.g. `torch.float16`), since it results in almost no degradation to audio quality but significantly lower memory usage and faster inference:
+
+```python
+>>> from transformers import GPTNeoXForCausalLM, GPTNeoXTokenizerFast
+
+model = GPTNeoXForCausalLM.from_pretrained("EleutherAI/gpt-neox-20b", torch_dtype=torch.float16, attn_implementation="flash_attention_2").to(device)
+...
+```
+
+
+### Expected speedups
+
+Below is an expected speedup diagram that compares pure inference time between the native implementation in transformers using `stockmark/gpt-neox-japanese-1.4b` checkpoint and the Flash Attention 2 version of the model using a sequence length of 2048.
+
+<div style="text-align: center">
+<img src="https://huggingface.co/datasets/ybelkada/documentation-images/resolve/main/gpt-neox-1.8b-speedup.jpg">
+</div>
+
 ## Resources
 
 - [Causal language modeling task guide](../tasks/language_modeling)

docs/source/en/model_doc/llama.md

@@ -50,6 +50,9 @@ come in several checkpoints they each contain a part of each weight of the model
 
 - The LLaMA tokenizer is a BPE model based on [sentencepiece](https://github.com/google/sentencepiece). One quirk of sentencepiece is that when decoding a sequence, if the first token is the start of the word (e.g. "Banana"), the tokenizer does not prepend the prefix space to the string.
 
+This model was contributed by [zphang](https://huggingface.co/zphang) with contributions from [BlackSamorez](https://huggingface.co/BlackSamorez). The code of the implementation in Hugging Face is based on GPT-NeoX [here](https://github.com/EleutherAI/gpt-neox). The original code of the authors can be found [here](https://github.com/facebookresearch/llama). The Flax version of the implementation was contributed by [afmck](https://huggingface.co/afmck) with the code in the implementation based on Hugging Face's Flax GPT-Neo.
+
+
 Based on the original LLaMA model, Meta AI has released some follow-up works:
 
 - **Llama2**: Llama2 is an improved version of Llama with some architectural tweaks (Grouped Query Attention), and is pre-trained on 2Trillion tokens. Refer to the documentation of Llama2 which can be found [here](llama2).
@@ -112,3 +115,13 @@ A list of official Hugging Face and community (indicated by 🌎) resources to h
 
 [[autodoc]] LlamaForSequenceClassification
     - forward
+
+## FlaxLlamaModel
+
+[[autodoc]] FlaxLlamaModel
+    - __call__
+
+## FlaxLlamaForCausalLM
+
+[[autodoc]] FlaxLlamaForCausalLM
+    - __call__

docs/source/en/model_doc/llava.md

@@ -0,0 +1,80 @@
+<!--Copyright 2023 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
+
+## Overview
+
+LLaVa is an open-source chatbot trained by fine-tuning LlamA/Vicuna on GPT-generated multimodal instruction-following data. It is an auto-regressive language model, based on the transformer architecture. In other words, it is an multi-modal version of LLMs fine-tuned for chat / instructions.
+
+The LLaVa model was proposed in [Visual Instruction Tuning](https://arxiv.org/abs/2304.08485) and improved in [Improved Baselines with Visual Instruction Tuning](https://arxiv.org/pdf/2310.03744) by Haotian Liu, Chunyuan Li, Yuheng Li and Yong Jae Lee.
+
+The abstract from the paper is the following:
+
+*Large multimodal models (LMM) have recently shown encouraging progress with visual instruction tuning. In this note, we show that the fully-connected vision-language cross-modal connector in LLaVA is surprisingly powerful and data-efficient. With simple modifications to LLaVA, namely, using CLIP-ViT-L-336px with an MLP projection and adding academic-task-oriented VQA data with simple response formatting prompts, we establish stronger baselines that achieve state-of-the-art across 11 benchmarks. Our final 13B checkpoint uses merely 1.2M publicly available data, and finishes full training in ∼1 day on a single 8-A100 node. We hope this can make state-of-the-art LMM research more accessible. Code and model will be publicly available*
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/llava_architecture.jpg"
+alt="drawing" width="600"/>
+
+<small> LLaVa architecture. Taken from the <a href="https://arxiv.org/abs/2304.08485">original paper.</a> </small>
+
+This model was contributed by [ArthurZ](https://huggingface.co/ArthurZ) and [ybelkada](https://huggingface.co/ybelkada).
+The original code can be found [here](https://github.com/haotian-liu/LLaVA/tree/main/llava).
+
+## 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 the model has not been explicitly trained to process multiple images in the same prompt, although this is technically possible, you may experience inaccurate results.
+
+- For better results, we recommend users to prompt the model with the correct prompt format: 
+
+```bash
+"USER: <image>\n<prompt>ASSISTANT:"
+```
+
+For multiple turns conversation:
+
+```bash
+"USER: <image>\n<prompt1>ASSISTANT: <answer1>USER: <prompt2>ASSISTANT: <answer2>USER: <prompt3>ASSISTANT:"
+```
+
+### Using Flash Attention 2
+
+Flash Attention 2 is an even faster, optimized version of the previous optimization, please refer to the [Flash Attention 2 section of performance docs](https://huggingface.co/docs/transformers/perf_infer_gpu_one).
+
+## Resources
+
+A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with BEiT.
+
+<PipelineTag pipeline="image-to-text"/>
+
+- A [Google Colab demo](https://colab.research.google.com/drive/1qsl6cd2c8gGtEW1xV5io7S8NHh-Cp1TV?usp=sharing) on how to run Llava on a free-tier Google colab instance leveraging 4-bit inference.
+- A [similar notebook](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/LLaVa/Inference_with_LLaVa_for_multimodal_generation.ipynb) showcasing batched inference. 🌎
+
+
+## LlavaConfig
+
+[[autodoc]] LlavaConfig
+
+## LlavaProcessor
+
+[[autodoc]] LlavaProcessor
+
+## LlavaForConditionalGeneration
+
+[[autodoc]] LlavaForConditionalGeneration
+    - forward

docs/source/en/model_doc/mistral.md

@@ -99,7 +99,7 @@ To load and run a model using Flash Attention 2, refer to the snippet below:
 >>> from transformers import AutoModelForCausalLM, AutoTokenizer
 >>> device = "cuda" # the device to load the model onto
 
->>> model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-v0.1", torch_dtype=torch.float16, use_flash_attention_2=True)
+>>> model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-v0.1", torch_dtype=torch.float16, attn_implementation="flash_attention_2")
 >>> tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-v0.1")
 
 >>> prompt = "My favourite condiment is"

docs/source/en/model_doc/mixtral.md

@@ -0,0 +1,163 @@
+<!--Copyright 2023 Mistral AI and 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.
+
+-->
+
+# Mixtral
+
+## Overview
+
+Mixtral-8x7B is Mistral AI's second Large Language Model (LLM). 
+
+The Mixtral model was proposed in the by the [Mistral AI](https://mistral.ai/) team.
+
+It was introduced in the [Mixtral of Experts blogpost](https://mistral.ai/news/mixtral-of-experts/) with the following introduction:
+
+*Today, the team is proud to release Mixtral 8x7B, a high-quality sparse mixture of experts models (SMoE) with open weights. Licensed under Apache 2.0. Mixtral outperforms Llama 2 70B on most benchmarks with 6x faster inference. It is the strongest open-weight model with a permissive license and the best model overall regarding cost/performance trade-offs. In particular, it matches or outperforms GPT3.5 on most standard benchmarks.*
+
+Tips:
+
+
+- The model needs to be converted using the [conversion script](https://github.com/huggingface/transformers/blob/main/src/transformers/models/mixtral/convert_mixtral_weights_to_hf.py).
+- If the model is quantized to 4bits, a single A100 is enough to fit the entire 84B model.
+
+This model was contributed by [Younes Belkada](https://huggingface.co/ybelkada) and [Arthur Zucker](https://huggingface.co/ArthurZ) .
+The original code can be found [here](https://github.com/mistralai/mistral-src).
+
+
+### Model Details
+
+Mixtral-84B is a decoder-based LM with the following architectural choices:
+
+* Mixtral is a Mixture of Expert (MOE) model with 8 experts per MLP, with a total of 85B paramateres but the compute required is the same as a 14B model. This is because even though each experts have to be loaded in RAM (70B like ram requirement) each token from the hidden states are dipatched twice (top 2 routing) and thus the compute (the operation required at each foward computation) is just 2 X sequence_length. 
+
+The following implementation details are shared with Mistral AI's first model [mistral](~models/doc/mistral):
+* Sliding Window Attention - Trained with 8k context length and fixed cache size, with a theoretical attention span of 128K tokens
+* GQA (Grouped Query Attention) - allowing faster inference and lower cache size.
+* Byte-fallback BPE tokenizer - ensures that characters are never mapped to out of vocabulary tokens.
+
+They also provide an instruction fine-tuned model: `mistralai/Mixtral-8x7B-v0.1` which can be used for chat-based inference.
+
+For more details please read our [release blog post](https://mistral.ai/news/mixtral-of-experts/)
+
+### License
+
+`Mixtral-8x7B` is released under the Apache 2.0 license.
+
+## Usage tips
+
+`Mixtral-8x7B` can be found on the [Huggingface Hub](https://huggingface.co/mistralai)
+
+These ready-to-use checkpoints can be downloaded and used via the HuggingFace Hub:
+
+```python
+>>> from transformers import AutoModelForCausalLM, AutoTokenizer
+>>> device = "cuda" # the device to load the model onto
+
+>>> model = AutoModelForCausalLM.from_pretrained("mistralai/Mixtral-8x7B-v0.1")
+>>> tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-8x7B")
+
+>>> prompt = "My favourite condiment is"
+
+>>> model_inputs = tokenizer([prompt], return_tensors="pt").to(device)
+>>> model.to(device)
+
+>>> generated_ids = model.generate(**model_inputs, max_new_tokens=100, do_sample=True)
+>>> tokenizer.batch_decode(generated_ids)[0]
+"The expected output"
+```
+
+To use the raw checkpoints with HuggingFace you can use the `convert_mixtral_weights_to_hf.py` script to convert them to the HuggingFace format:
+
+```bash
+python src/transformers/models/mixtral/convert_mixtral_weights_to_hf.py \
+    --input_dir /path/to/downloaded/mistral/weights --output_dir /output/path
+```
+
+You can then load the converted model from the `output/path`:
+
+```python
+from transformers import MixtralForCausalLM, LlamaTokenizer
+
+tokenizer = LlamaTokenizer.from_pretrained("/output/path")
+model = MixtralForCausalLM.from_pretrained("/output/path")
+```
+
+## Combining Mixtral and Flash Attention 2
+
+First, make sure to install the latest version of Flash Attention 2 to include the sliding window attention feature.
+
+```bash
+pip install -U flash-attn --no-build-isolation
+```
+
+Make also sure that you have a hardware that is compatible with Flash-Attention 2. Read more about it in the official documentation of [`flash-attn`](https://github.com/Dao-AILab/flash-attention) repository. Make also sure to load your model in half-precision (e.g. `torch.float16`)
+
+To load and run a model using Flash Attention 2, refer to the snippet below:
+
+```python
+>>> import torch
+>>> from transformers import AutoModelForCausalLM, AutoTokenizer
+>>> device = "cuda" # the device to load the model onto
+
+>>> model = AutoModelForCausalLM.from_pretrained("mistralai/Mixtral-8x7B-v0.1", torch_dtype=torch.float16, attn_implementation="flash_attention_2")
+>>> tokenizer = AutoTokenizer.from_pretrained("mistralai/Mixtral-8x7B-v0.1")
+
+>>> prompt = "My favourite condiment is"
+
+>>> model_inputs = tokenizer([prompt], return_tensors="pt").to(device)
+>>> model.to(device)
+
+>>> generated_ids = model.generate(**model_inputs, max_new_tokens=100, do_sample=True)
+>>> tokenizer.batch_decode(generated_ids)[0]
+"The expected output"
+```
+
+### Expected speedups
+
+Below is a expected speedup diagram that compares pure inference time between the native implementation in transformers using `mistralai/Mixtral-8x7B-v0.1` checkpoint and the Flash Attention 2 version of the model.
+
+<div style="text-align: center">
+<img src="https://huggingface.co/datasets/ybelkada/documentation-images/resolve/main/mixtral-7b-inference-large-seqlen.png">
+</div>
+
+### Sliding window Attention
+
+The current implementation supports the sliding window attention mechanism and memory efficient cache management. 
+To enable sliding window attention, just make sure to have a `flash-attn` version that is compatible with sliding window attention (`>=2.3.0`). 
+
+The Flash Attention-2 model uses also a more memory efficient cache slicing mechanism - as recommended per the official implementation of Mistral model that use rolling cache mechanism we keep the cache size fixed (`self.config.sliding_window`), support batched generation only for `padding_side="left"` and use the absolute position of the current token to compute the positional embedding.
+
+## The Mistral 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.
+
+## MixtralConfig
+
+[[autodoc]] MixtralConfig
+
+## MixtralModel
+
+[[autodoc]] MixtralModel
+    - forward
+
+## MixtralForCausalLM
+
+[[autodoc]] MixtralForCausalLM
+    - forward
+
+## MixtralForSequenceClassification
+
+[[autodoc]] MixtralForSequenceClassification
+    - forward

docs/source/en/model_doc/opt.md

@@ -80,7 +80,7 @@ To load and run a model using Flash Attention 2, refer to the snippet below:
 >>> from transformers import OPTForCausalLM, GPT2Tokenizer
 >>> device = "cuda" # the device to load the model onto
 
->>> model = OPTForCausalLM.from_pretrained("facebook/opt-350m", torch_dtype=torch.float16, use_flash_attention_2=True)
+>>> model = OPTForCausalLM.from_pretrained("facebook/opt-350m", torch_dtype=torch.float16, attn_implementation="flash_attention_2")
 >>> tokenizer = GPT2Tokenizer.from_pretrained("facebook/opt-350m")
 
 >>> prompt = ("A chat between a curious human and the Statue of Liberty.\n\nHuman: What is your name?\nStatue: I am the "

docs/source/en/model_doc/patchtsmixer.md

@@ -0,0 +1,90 @@
+<!--Copyright 2023 IBM and 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.
+
+⚠️ 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.
+
+-->
+
+# PatchTSMixer
+
+## Overview
+
+The PatchTSMixer model was proposed in [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 and Jayant Kalagnanam.
+
+
+PatchTSMixer is a lightweight time-series modeling approach based on the MLP-Mixer architecture. In this HuggingFace implementation, we provide PatchTSMixer's capabilities to effortlessly facilitate lightweight mixing across patches, channels, and hidden features for effective multivariate time-series modeling. It also supports various attention mechanisms starting from simple gated attention to more complex self-attention blocks that can be customized accordingly. The model can be pretrained and subsequently used for various downstream tasks such as forecasting, classification and regression.
+
+
+The abstract from the paper is the following:
+
+*TSMixer is a lightweight neural architecture exclusively composed of multi-layer perceptron (MLP) modules designed for multivariate forecasting and representation learning on patched time series. Our model draws inspiration from the success of MLP-Mixer models in computer vision. We demonstrate the challenges involved in adapting Vision MLP-Mixer for time series and introduce empirically validated components to enhance accuracy. This includes a novel design paradigm of attaching online reconciliation heads to the MLP-Mixer backbone, for explicitly modeling the time-series properties such as hierarchy and channel-correlations. We also propose a Hybrid channel modeling approach to effectively handle noisy channel interactions and generalization across diverse datasets, a common challenge in existing patch channel-mixing methods. Additionally, a simple gated attention mechanism is introduced in the backbone to prioritize important features. By incorporating these lightweight components, we significantly enhance the learning capability of simple MLP structures, outperforming complex Transformer models with minimal computing usage. Moreover, TSMixer's modular design enables compatibility with both supervised and masked self-supervised learning methods, making it a promising building block for time-series Foundation Models. TSMixer outperforms state-of-the-art MLP and Transformer models in forecasting by a considerable margin of 8-60%. It also outperforms the latest strong benchmarks of Patch-Transformer models (by 1-2%) with a significant reduction in memory and runtime (2-3X).*
+
+
+
+This model was contributed by [ajati](https://huggingface.co/ajati), [vijaye12](https://huggingface.co/vijaye12), 
+[gsinthong](https://huggingface.co/gsinthong), [namctin](https://huggingface.co/namctin),
+[wmgifford](https://huggingface.co/wmgifford), [kashif](https://huggingface.co/kashif).
+
+
+## Sample usage 
+```python
+
+from transformers import PatchTSMixerConfig, PatchTSMixerForPrediction
+from transformers import Trainer, TrainingArguments,
+
+
+config = PatchTSMixerConfig(context_length = 512, prediction_length = 96)
+model = PatchTSMixerForPrediction(config)
+trainer = Trainer(model=model, args=training_args, 
+            train_dataset=train_dataset,
+            eval_dataset=valid_dataset)
+trainer.train()
+results = trainer.evaluate(test_dataset)
+```
+
+## Usage tips
+
+The model can also be used for time series classification and time series regression. See the respective [`PatchTSMixerForTimeSeriesClassification`] and [`PatchTSMixerForRegression`] classes.
+
+## PatchTSMixerConfig
+
+[[autodoc]] PatchTSMixerConfig
+
+
+## PatchTSMixerModel
+
+[[autodoc]] PatchTSMixerModel
+    - forward
+
+
+## PatchTSMixerForPrediction
+
+[[autodoc]] PatchTSMixerForPrediction
+    - forward
+
+
+## PatchTSMixerForTimeSeriesClassification
+
+[[autodoc]] PatchTSMixerForTimeSeriesClassification
+    - forward
+
+
+## PatchTSMixerForPretraining
+
+[[autodoc]] PatchTSMixerForPretraining
+    - forward
+
+
+## PatchTSMixerForRegression
+
+[[autodoc]] PatchTSMixerForRegression
+    - forward

docs/source/en/model_doc/phi.md

@@ -76,7 +76,7 @@ The original code for Phi-1 and Phi-1.5 can be found [here](https://huggingface.
 ```python
 >>> from transformers import PhiForCausalLM, AutoTokenizer
 
->>> # define the model and tokenzier.
+>>> # define the model and tokenizer.
 >>> model = PhiForCausalLM.from_pretrained("susnato/phi-1_5_dev")
 >>> tokenizer = AutoTokenizer.from_pretrained("susnato/phi-1_5_dev")
 
@@ -94,6 +94,46 @@ The original code for Phi-1 and Phi-1.5 can be found [here](https://huggingface.
 ```
 
 
+## Combining Phi and Flash Attention 2
+
+First, make sure to install the latest version of Flash Attention 2 to include the sliding window attention feature.
+
+```bash
+pip install -U flash-attn --no-build-isolation
+```
+
+Make also sure that you have a hardware that is compatible with Flash-Attention 2. Read more about it in the official documentation of flash-attn repository. Make also sure to load your model in half-precision (e.g. `torch.float16``)
+
+To load and run a model using Flash Attention 2, refer to the snippet below:
+
+```python
+>>> import torch
+>>> from transformers import PhiForCausalLM, AutoTokenizer
+
+>>> # define the model and tokenizer and push the model and tokens to the GPU.
+>>> model = PhiForCausalLM.from_pretrained("susnato/phi-1_5_dev", torch_dtype=torch.float16, attn_implementation="flash_attention_2").to("cuda")
+>>> tokenizer = AutoTokenizer.from_pretrained("susnato/phi-1_5_dev")
+
+>>> # feel free to change the prompt to your liking.
+>>> prompt = "If I were an AI that had just achieved"
+
+>>> # apply the tokenizer.
+>>> tokens = tokenizer(prompt, return_tensors="pt").to("cuda")
+
+>>> # use the model to generate new tokens.
+>>> generated_output = model.generate(**tokens, use_cache=True, max_new_tokens=10)
+
+>>> tokenizer.batch_decode(generated_output)[0]
+'If I were an AI that had just achieved a breakthrough in machine learning, I would be thrilled'
+```
+
+### Expected speedups
+Below is an expected speedup diagram that compares pure inference time between the native implementation in transformers using `susnato/phi-1_dev` checkpoint and the Flash Attention 2 version of the model using a sequence length of 2048.
+<div style="text-align: center">
+<img src="https://huggingface.co/datasets/ybelkada/documentation-images/resolve/main/phi_1_speedup_plot.jpg">
+</div>
+
+
 ## PhiConfig
 
 [[autodoc]] PhiConfig
@@ -123,4 +163,4 @@ The original code for Phi-1 and Phi-1.5 can be found [here](https://huggingface.
     - forward
 
 </pt>
-</frameworkcontent>
+</frameworkcontent>

docs/source/en/pad_truncation.md

@@ -54,7 +54,7 @@ The following table summarizes the recommended way to setup padding and truncati
 |                                      |                                   | `tokenizer(batch_sentences, padding='longest')`                                        |
 |                                      | padding to max model input length | `tokenizer(batch_sentences, padding='max_length')`                                     |
 |                                      | padding to specific length        | `tokenizer(batch_sentences, padding='max_length', max_length=42)`                      |
-|                                      | padding to a multiple of a value  | `tokenizer(batch_sentences, padding=True, pad_to_multiple_of=8)                        |
+|                                      | padding to a multiple of a value  | `tokenizer(batch_sentences, padding=True, pad_to_multiple_of=8)`                        |
 | truncation to max model input length | no padding                        | `tokenizer(batch_sentences, truncation=True)` or                                       |
 |                                      |                                   | `tokenizer(batch_sentences, truncation=STRATEGY)`                                      |
 |                                      | padding to max sequence in batch  | `tokenizer(batch_sentences, padding=True, truncation=True)` or                         |

docs/source/en/perf_infer_gpu_one.md

@@ -36,13 +36,30 @@ FlashAttention-2 is experimental and may change considerably in future versions.
 1. additionally parallelizing the attention computation over sequence length
 2. partitioning the work between GPU threads to reduce communication and shared memory reads/writes between them
 
-FlashAttention-2 supports inference with Llama, Mistral, Falcon and Bark models. You can request to add FlashAttention-2 support for another model by opening a GitHub Issue or Pull Request.
+FlashAttention-2 is currently supported for the following architectures:
+* [Bark](https://huggingface.co/docs/transformers/model_doc/bark#transformers.BarkModel)
+* [Bart](https://huggingface.co/docs/transformers/model_doc/bart#transformers.BartModel)
+* [DistilBert](https://huggingface.co/docs/transformers/model_doc/distilbert#transformers.DistilBertModel)
+* [GPTBigCode](https://huggingface.co/docs/transformers/model_doc/gpt_bigcode#transformers.GPTBigCodeModel)
+* [GPTNeo](https://huggingface.co/docs/transformers/model_doc/gpt_neo#transformers.GPTNeoModel)
+* [GPTNeoX](https://huggingface.co/docs/transformers/model_doc/gpt_neox#transformers.GPTNeoXModel)
+* [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)
+* [MBart](https://huggingface.co/docs/transformers/model_doc/mbart#transformers.MBartModel)
+* [Mistral](https://huggingface.co/docs/transformers/model_doc/mistral#transformers.MistralModel)
+* [Mixtral](https://huggingface.co/docs/transformers/model_doc/mixtral#transformers.MixtralModel)
+* [OPT](https://huggingface.co/docs/transformers/model_doc/opt#transformers.OPTModel)
+* [Phi](https://huggingface.co/docs/transformers/model_doc/phi#transformers.PhiModel)
+* [Whisper](https://huggingface.co/docs/transformers/model_doc/whisper#transformers.WhisperModel)
+
+You can request to add FlashAttention-2 support for another model by opening a GitHub Issue or Pull Request.
 
 Before you begin, make sure you have FlashAttention-2 installed. For NVIDIA GPUs, the library is installable through pip: `pip install flash-attn --no-build-isolation`. We strongly suggest to refer to the [detailed installation instructions](https://github.com/Dao-AILab/flash-attention?tab=readme-ov-file#installation-and-features).
 
 FlashAttention-2 is also supported on AMD GPUs, with the current support limited to **Instinct MI210 and Instinct MI250**. We strongly suggest to use the following [Dockerfile](https://github.com/huggingface/optimum-amd/tree/main/docker/transformers-pytorch-amd-gpu-flash/Dockerfile) to use FlashAttention-2 on AMD GPUs.
 
-To enable FlashAttention-2, add the `use_flash_attention_2` parameter to [`~AutoModelForCausalLM.from_pretrained`]:
+To enable FlashAttention-2, pass the argument `attn_implementation="flash_attention_2"` to [`~AutoModelForCausalLM.from_pretrained`]:
 
 ```python
 import torch
@@ -54,13 +71,15 @@ tokenizer = AutoTokenizer.from_pretrained(model_id)
 model = AutoModelForCausalLM.from_pretrained(
     model_id, 
     torch_dtype=torch.bfloat16, 
-    use_flash_attention_2=True,
+    attn_implementation="flash_attention_2",
 )
 ```
 
 <Tip>
 
 FlashAttention-2 can only be used when the model's dtype is `fp16` or `bf16`. Make sure to cast your model to the appropriate dtype and load them on a supported device before using FlashAttention-2.
+
+Note that `use_flash_attention_2=True` can also be used to enable Flash Attention 2, but is deprecated in favor of `attn_implementation="flash_attention_2"`.
   
 </Tip>
 
@@ -77,14 +96,14 @@ tokenizer = AutoTokenizer.from_pretrained(model_id)
 model = AutoModelForCausalLM.from_pretrained(
     model_id, 
     load_in_8bit=True,
-    use_flash_attention_2=True,
+    attn_implementation="flash_attention_2",
 )
 
 # load in 4bit
 model = AutoModelForCausalLM.from_pretrained(
     model_id, 
     load_in_4bit=True,
-    use_flash_attention_2=True,
+    attn_implementation="flash_attention_2",
 )
 ```
 
@@ -124,8 +143,58 @@ FlashAttention is more memory efficient, meaning you can train on much larger se
 <img src="https://huggingface.co/datasets/ybelkada/documentation-images/resolve/main/llama-2-large-seqlen-padding.png">
 </div>
 
+## FlashAttention and memory-efficient attention through PyTorch's scaled_dot_product_attention 
+
+PyTorch's [`torch.nn.functional.scaled_dot_product_attention`](https://pytorch.org/docs/master/generated/torch.nn.functional.scaled_dot_product_attention.html) (SDPA) can also call FlashAttention and memory-efficient attention kernels under the hood. SDPA support is currently being added natively in Transformers, and is used by default for `torch>=2.1.1` when an implementation is available.
+
+For now, Transformers supports inference and training through SDPA for the following architectures:
+* [Bart](https://huggingface.co/docs/transformers/model_doc/bart#transformers.BartModel)
+* [GPTBigCode](https://huggingface.co/docs/transformers/model_doc/gpt_bigcode#transformers.GPTBigCodeModel)
+* [Falcon](https://huggingface.co/docs/transformers/model_doc/falcon#transformers.FalconModel)
+* [Llama](https://huggingface.co/docs/transformers/model_doc/llama#transformers.LlamaModel)
+* [Idefics](https://huggingface.co/docs/transformers/model_doc/idefics#transformers.IdeficsModel)
+* [Whisper](https://huggingface.co/docs/transformers/model_doc/whisper#transformers.WhisperModel)
+
+Note that FlashAttention can only be used for models with the `fp16` or `bf16` torch type, so make sure to cast your model to the appropriate type before using it.
+
+By default, `torch.nn.functional.scaled_dot_product_attention` selects the most performant kernel available, but to check whether a backend is available in a given setting (hardware, problem size), you can use [`torch.backends.cuda.sdp_kernel`](https://pytorch.org/docs/master/backends.html#torch.backends.cuda.sdp_kernel) as a context manager:
+
+```diff
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+tokenizer = AutoTokenizer.from_pretrained("facebook/opt-350m")
+model = AutoModelForCausalLM.from_pretrained("facebook/opt-350m", torch_dtype=torch.float16).to("cuda")
+# convert the model to BetterTransformer
+model.to_bettertransformer()
+
+input_text = "Hello my dog is cute and"
+inputs = tokenizer(input_text, return_tensors="pt").to("cuda")
+
++ with torch.backends.cuda.sdp_kernel(enable_flash=True, enable_math=False, enable_mem_efficient=False):
+    outputs = model.generate(**inputs)
+
+print(tokenizer.decode(outputs[0], skip_special_tokens=True))
+```
+
+If you see a bug with the traceback below, try using nightly version of PyTorch which may have broader coverage for FlashAttention:
+
+```bash
+RuntimeError: No available kernel. Aborting execution.
+
+# install PyTorch nightly
+pip3 install -U --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu118
+```
+
 ## BetterTransformer
 
+<Tip warning={true}>
+
+Part of BetterTransformer features are being upstreamed in Transformers, with native `torch.nn.scaled_dot_product_attention` default support. BetterTransformer still has a wider coverage than the Transformers SDPA integration, but you can expect more and more architectures to support natively SDPA in Transformers.
+
+</Tip>
+
+
 <Tip>
 
 Check out our benchmarks with BetterTransformer and scaled dot product attention in the [Out of the box acceleration and memory savings of 🤗 decoder models with PyTorch 2.0](https://pytorch.org/blog/out-of-the-box-acceleration/) and learn more about the fastpath execution in the [BetterTransformer](https://medium.com/pytorch/bettertransformer-out-of-the-box-performance-for-huggingface-transformers-3fbe27d50ab2) blog post.
@@ -154,39 +223,6 @@ model = model.reverse_bettertransformer()
 model.save_pretrained("saved_model")
 ```
 
-### FlashAttention
-
-SDPA can also call FlashAttention kernels under the hood. FlashAttention can only be used for models using the `fp16` or `bf16` dtype, so make sure to cast your model to the appropriate dtype before using it.
-
-To enable FlashAttention or to check whether it is available in a given setting (hardware, problem size), use [`torch.backends.cuda.sdp_kernel`](https://pytorch.org/docs/master/backends.html#torch.backends.cuda.sdp_kernel) as a context manager:
-
-```diff
-import torch
-from transformers import AutoModelForCausalLM, AutoTokenizer
-
-tokenizer = AutoTokenizer.from_pretrained("facebook/opt-350m")
-model = AutoModelForCausalLM.from_pretrained("facebook/opt-350m", torch_dtype=torch.float16).to("cuda")
-# convert the model to BetterTransformer
-model.to_bettertransformer()
-
-input_text = "Hello my dog is cute and"
-inputs = tokenizer(input_text, return_tensors="pt").to("cuda")
-
-+ with torch.backends.cuda.sdp_kernel(enable_flash=True, enable_math=False, enable_mem_efficient=False):
-    outputs = model.generate(**inputs)
-
-print(tokenizer.decode(outputs[0], skip_special_tokens=True))
-```
-
-If you see a bug with the traceback below, try using nightly version of PyTorch which may have broader coverage for FlashAttention:
-
-```bash
-RuntimeError: No available kernel. Aborting execution.
-
-# install PyTorch nightly
-pip3 install -U --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu118
-```
-
 ## bitsandbytes
 
 bitsandbytes is a quantization library that includes support for 4-bit and 8-bit quantization. Quantization reduces your model size compared to its native full precision version, making it easier to fit large models onto GPUs with limited memory.

docs/source/en/perf_train_cpu_many.md

@@ -15,7 +15,8 @@ rendered properly in your Markdown viewer.
 
 # Efficient Training on Multiple CPUs
 
-When training on a single CPU is too slow, we can use multiple CPUs. This guide focuses on PyTorch-based DDP enabling distributed CPU training efficiently.
+When training on a single CPU is too slow, we can use multiple CPUs. This guide focuses on PyTorch-based DDP enabling
+distributed CPU training efficiently on [bare metal](#usage-in-trainer) and [Kubernetes](#usage-with-kubernetes).
 
 ## Intel® oneCCL Bindings for PyTorch
 
@@ -25,7 +26,7 @@ Module `oneccl_bindings_for_pytorch` (`torch_ccl` before version 1.12)  implemen
 
 Check more detailed information for [oneccl_bind_pt](https://github.com/intel/torch-ccl).
 
-### Intel® oneCCL Bindings for PyTorch installation:
+### Intel® oneCCL Bindings for PyTorch installation
 
 Wheel files are available for the following Python versions:
 
@@ -68,9 +69,9 @@ torch_ccl_path=$(python -c "import torch; import torch_ccl; import os;  print(os
 source $torch_ccl_path/env/setvars.sh
 ```
 
-#### IPEX installation:
+#### Intel® Extension for PyTorch installation
 
-IPEX provides performance optimizations for CPU training with both Float32 and BFloat16, you could refer [single CPU section](./perf_train_cpu).
+Intel Extension for PyTorch (IPEX) provides performance optimizations for CPU training with both Float32 and BFloat16 (refer to the [single CPU section](./perf_train_cpu) to learn more).
 
 
 The following "Usage in Trainer" takes mpirun in Intel® MPI library as an example.
@@ -132,3 +133,185 @@ Now, run the following command in node0 and **4DDP** will be enabled in node0 an
  --use_ipex \
  --bf16
 ```
+
+## Usage with Kubernetes
+
+The same distributed training job from the previous section can be deployed to a Kubernetes cluster using the
+[Kubeflow PyTorchJob training operator](https://www.kubeflow.org/docs/components/training/pytorch/).
+
+### Setup
+
+This example assumes that you have:
+* Access to a Kubernetes cluster with [Kubeflow installed](https://www.kubeflow.org/docs/started/installing-kubeflow/)
+* [`kubectl`](https://kubernetes.io/docs/tasks/tools/) installed and configured to access the Kubernetes cluster
+* A [Persistent Volume Claim (PVC)](https://kubernetes.io/docs/concepts/storage/persistent-volumes/) that can be used
+  to store datasets and model files. There are multiple options for setting up the PVC including using an NFS
+  [storage class](https://kubernetes.io/docs/concepts/storage/storage-classes/) or a cloud storage bucket.
+* A Docker container that includes your model training script and all the dependencies needed to run the script. For
+  distributed CPU training jobs, this typically includes PyTorch, Transformers, Intel Extension for PyTorch, Intel
+  oneCCL Bindings for PyTorch, and OpenSSH to communicate between the containers.
+
+The snippet below is an example of a Dockerfile that uses a base image that supports distributed CPU training and then
+extracts a Transformers release to the `/workspace` directory, so that the example scripts are included in the image:
+```
+FROM intel/ai-workflows:torch-2.0.1-huggingface-multinode-py3.9
+
+WORKDIR /workspace
+
+# Download and extract the transformers code
+ARG HF_TRANSFORMERS_VER="4.35.2"
+RUN mkdir transformers && \
+    curl -sSL --retry 5 https://github.com/huggingface/transformers/archive/refs/tags/v${HF_TRANSFORMERS_VER}.tar.gz | tar -C transformers --strip-components=1 -xzf -
+```
+The image needs to be built and copied to the cluster's nodes or pushed to a container registry prior to deploying the
+PyTorchJob to the cluster.
+
+### PyTorchJob Specification File
+
+The [Kubeflow PyTorchJob](https://www.kubeflow.org/docs/components/training/pytorch/) is used to run the distributed
+training job on the cluster. The yaml file for the PyTorchJob defines parameters such as:
+ * The name of the PyTorchJob
+ * The number of replicas (workers)
+ * The python script and it's parameters that will be used to run the training job
+ * The types of resources (node selector, memory, and CPU) needed for each worker
+ * The image/tag for the Docker container to use
+ * Environment variables
+ * A volume mount for the PVC
+
+The volume mount defines a path where the PVC will be mounted in the container for each worker pod. This location can be
+used for the dataset, checkpoint files, and the saved model after training completes.
+
+The snippet below is an example of a yaml file for a PyTorchJob with 4 workers running the
+[question-answering example](https://github.com/huggingface/transformers/tree/main/examples/pytorch/question-answering).
+```yaml
+apiVersion: "kubeflow.org/v1"
+kind: PyTorchJob
+metadata:
+  name: transformers-pytorchjob
+  namespace: kubeflow
+spec:
+  elasticPolicy:
+    rdzvBackend: c10d
+    minReplicas: 1
+    maxReplicas: 4
+    maxRestarts: 10
+  pytorchReplicaSpecs:
+    Worker:
+      replicas: 4  # The number of worker pods
+      restartPolicy: OnFailure
+      template:
+        spec:
+          containers:
+            - name: pytorch
+              image: <image name>:<tag>  # Specify the docker image to use for the worker pods
+              imagePullPolicy: IfNotPresent
+              command:
+                - torchrun
+                - /workspace/transformers/examples/pytorch/question-answering/run_qa.py
+                - --model_name_or_path
+                - "bert-large-uncased"
+                - --dataset_name
+                - "squad"
+                - --do_train
+                - --do_eval
+                - --per_device_train_batch_size
+                - "12"
+                - --learning_rate
+                - "3e-5"
+                - --num_train_epochs
+                - "2"
+                - --max_seq_length
+                - "384"
+                - --doc_stride
+                - "128"
+                - --output_dir
+                - "/tmp/pvc-mount/output"
+                - --no_cuda
+                - --ddp_backend
+                - "ccl"
+                - --use_ipex
+                - --bf16  # Specify --bf16 if your hardware supports bfloat16
+              env:
+              - name: LD_PRELOAD
+                value: "/usr/lib/x86_64-linux-gnu/libtcmalloc.so.4.5.9:/usr/local/lib/libiomp5.so"
+              - name: TRANSFORMERS_CACHE
+                value: "/tmp/pvc-mount/transformers_cache"
+              - name: HF_DATASETS_CACHE
+                value: "/tmp/pvc-mount/hf_datasets_cache"
+              - name: LOGLEVEL
+                value: "INFO"
+              - name: CCL_WORKER_COUNT
+                value: "1"
+              - name: OMP_NUM_THREADS  # Can be tuned for optimal performance
+-                value: "56"
+              resources:
+                limits:
+                  cpu: 200  # Update the CPU and memory limit values based on your nodes
+                  memory: 128Gi
+                requests:
+                  cpu: 200  # Update the CPU and memory request values based on your nodes
+                  memory: 128Gi
+              volumeMounts:
+              - name: pvc-volume
+                mountPath: /tmp/pvc-mount
+              - mountPath: /dev/shm
+                name: dshm
+          restartPolicy: Never
+          nodeSelector:  #  Optionally use the node selector to specify what types of nodes to use for the workers
+            node-type: spr
+          volumes:
+          - name: pvc-volume
+            persistentVolumeClaim:
+              claimName: transformers-pvc
+          - name: dshm
+            emptyDir:
+              medium: Memory
+```
+To run this example, update the yaml based on your training script and the nodes in your cluster.
+
+<Tip>
+
+The CPU resource limits/requests in the yaml are defined in [cpu units](https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#meaning-of-cpu)
+where 1 CPU unit is equivalent to 1 physical CPU core or 1 virtual core (depending on whether the node is a physical
+host or a VM). The amount of CPU and memory limits/requests defined in the yaml should be less than the amount of
+available CPU/memory capacity on a single machine. It is usually a good idea to not use the entire machine's capacity in
+order to leave some resources for the kubelet and OS. In order to get ["guaranteed"](https://kubernetes.io/docs/concepts/workloads/pods/pod-qos/#guaranteed)
+[quality of service](https://kubernetes.io/docs/tasks/configure-pod-container/quality-service-pod/) for the worker pods,
+set the same CPU and memory amounts for both the resource limits and requests.
+
+</Tip>
+
+### Deploy
+
+After the PyTorchJob spec has been updated with values appropriate for your cluster and training job, it can be deployed
+to the cluster using:
+```
+kubectl create -f pytorchjob.yaml
+```
+
+The `kubectl get pods -n kubeflow` command can then be used to list the pods in the `kubeflow` namespace. You should see
+the worker pods for the PyTorchJob that was just deployed. At first, they will probably have a status of "Pending" as
+the containers get pulled and created, then the status should change to "Running".
+```
+NAME                                                     READY   STATUS                  RESTARTS          AGE
+...
+transformers-pytorchjob-worker-0                         1/1     Running                 0                 7m37s
+transformers-pytorchjob-worker-1                         1/1     Running                 0                 7m37s
+transformers-pytorchjob-worker-2                         1/1     Running                 0                 7m37s
+transformers-pytorchjob-worker-3                         1/1     Running                 0                 7m37s
+...
+```
+
+The logs for worker can be viewed using `kubectl logs -n kubeflow <pod name>`. Add `-f` to stream the logs, for example:
+```
+kubectl logs -n kubeflow transformers-pytorchjob-worker-0 -f
+```
+
+After the training job completes, the trained model can be copied from the PVC or storage location. When you are done
+with the job, the PyTorchJob resource can be deleted from the cluster using `kubectl delete -f pytorchjob.yaml`.
+
+## Summary
+
+This guide covered running distributed PyTorch training jobs using multiple CPUs on bare metal and on a Kubernetes
+cluster. Both cases utilize Intel Extension for PyTorch and Intel oneCCL Bindings for PyTorch for optimal training
+performance, and can be used as a template to run your own workload on multiple nodes.

docs/source/en/quantization.md

@@ -82,9 +82,150 @@ AWQ quantization can also be combined with [FlashAttention-2](perf_infer_gpu_one
 ```py
 from transformers import AutoModelForCausalLM, AutoTokenizer
 
-model = AutoModelForCausalLM.from_pretrained("TheBloke/zephyr-7B-alpha-AWQ", use_flash_attention_2=True, device_map="cuda:0")
+model = AutoModelForCausalLM.from_pretrained("TheBloke/zephyr-7B-alpha-AWQ", attn_implementation="flash_attention_2", device_map="cuda:0")
 ```
 
+
+### Benchmarks
+
+We performed some speed, throughput and latency benchmarks using [`optimum-benchmark`](https://github.com/huggingface/optimum-benchmark) library. 
+
+Note at that time of writing this documentation section, the available quantization methods were: `awq`, `gptq` and `bitsandbytes`.
+
+The benchmark was run on a NVIDIA-A100 instance and the model used was [`TheBloke/Mistral-7B-v0.1-AWQ`](https://huggingface.co/TheBloke/Mistral-7B-v0.1-AWQ) for the AWQ model, [`TheBloke/Mistral-7B-v0.1-GPTQ`](https://huggingface.co/TheBloke/Mistral-7B-v0.1-GPTQ) for the GPTQ model. We also benchmarked it against `bitsandbytes` quantization methods and native `float16` model. Some results are shown below:
+
+<div style="text-align: center">
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/quantization/forward_memory_plot.png">
+</div>
+
+<div style="text-align: center">
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/quantization/generate_memory_plot.png">
+</div>
+
+<div style="text-align: center">
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/quantization/generate_throughput_plot.png">
+</div>
+
+<div style="text-align: center">
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/quantization/forward_latency_plot.png">
+</div>
+
+You can find the full results together with packages versions in [this link](https://github.com/huggingface/optimum-benchmark/tree/main/examples/running-mistrals).
+
+From the results it appears that AWQ quantization method is the fastest quantization method for inference, text generation and among the lowest peak memory for text generation. However, AWQ seems to have the largest forward latency per batch size. 
+
+
+### Make use of fused modules
+
+You can benefit from fused modules by passing an `AwqConfig` with `fuse_modules=True` and your expected maximum sequence length for generation to `fuse_max_seq_len`. For architectures that do not support `do_fuse=True`, you can still fuse the modules, however you need to pass a custom `fusing_mapping` to `AwqConfig()`. Let's dive into these specific usecases.
+
+Note that you cannot combine fusing modules and other optimization techniques such as Flash Attention 2.
+
+#### Fusing modules for supported architectures
+
+Currently we support out of the box AWQ module fusing for `llama` and `mistral`. 
+
+To enable this feature for supported architectures simply create an `AwqConfig` and pass the arguments `fuse_max_seq_len` and `do_fuse=True`.
+
+For example to enable module fusing for the model `TheBloke/Mistral-7B-OpenOrca-AWQ`, run:
+
+```python
+import torch
+from transformers import AwqConfig, AutoModelForCausalLM
+
+model_id = "TheBloke/Mistral-7B-OpenOrca-AWQ"
+
+quantization_config = AwqConfig(
+    bits=4,
+    fuse_max_seq_len=512,
+    do_fuse=True,
+)
+
+model = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=quantization_config).to(0)
+```
+
+Note that you need to define `fuse_max_seq_len` to `AwqConfig`. That total sequence length should include the context length and the expected generation length. You can set it to a large value to be on the safe zone.
+
+You can also apply module fusing for other architectures that are not supported.
+
+#### Fusing modules for unsupported architectures
+
+For architectures that do not support out of the box module fusing, you can pass a custom fusing mapping; simply pass a dictionnary `modules_to_fuse` to `AwqConfig`, let's take an example with the Yi model:
+
+
+```python
+import torch
+from transformers import AwqConfig, AutoModelForCausalLM
+
+model_id = "TheBloke/Yi-34B-AWQ"
+
+quantization_config = AwqConfig(
+    bits=4,
+    fuse_max_seq_len=512,
+    modules_to_fuse={
+        "attention": ["q_proj", "k_proj", "v_proj", "o_proj"],
+        "layernorm": ["ln1", "ln2", "norm"],
+        "mlp": ["gate_proj", "up_proj", "down_proj"],
+        "use_alibi": False,
+        "num_attention_heads": 56,
+        "num_key_value_heads": 8,
+        "hidden_size": 7168
+    }
+)
+
+model = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=quantization_config).to(0)
+```
+
+The parameter `modules_to_fuse` needs to have the following respective fields: 
+
+- `"attention"`: The names of the attention layers to fuse - in the order: query, key, value and output projection layer. In case you don't want to fuse the attention layers you can pass an empty list.
+- `"layernorm"`: The names of all the layernorm layers you want to replace with a custom fused layer norm. In case you don't want to fuse these layers you can also pass an empty list.
+- `"mlp"`: The names of the MLP layers you want to fuse into a single MLP layer in the order: (gate (dense layer post-attention) / up / down layers).
+- `"use_alibi"`: If you model uses alibi positional embedding
+- `"num_attention_heads"`: The number of attention heads
+- `"num_key_value_heads"`: This is the number of key value heads that should be used to implement Grouped Query Attention. If num_key_value_heads=num_attention_heads, the model will use Multi Head Attention (MHA), if num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. 
+- `"hidden_size"`: Dimension of the hidden representations.
+
+
+#### Benchmarks
+
+We benchmarked the model with and without fused modules first using only `batch_size=1` on the `TheBloke/Mistral-7B-OpenOrca-AWQ` model and below are the results:
+
+*unfused case*
+
+|   Batch Size |   Prefill Length |   Decode Length |   Prefill tokens/s |   Decode tokens/s | Memory (VRAM)   |
+|-------------:|-----------------:|----------------:|-------------------:|------------------:|:----------------|
+|            1 |               32 |              32 |            60.0984 |           38.4537 | 4.50 GB (5.68%) |
+|            1 |               64 |              64 |          1333.67   |           31.6604 | 4.50 GB (5.68%) |
+|            1 |              128 |             128 |          2434.06   |           31.6272 | 4.50 GB (5.68%) |
+|            1 |              256 |             256 |          3072.26   |           38.1731 | 4.50 GB (5.68%) |
+|            1 |              512 |             512 |          3184.74   |           31.6819 | 4.59 GB (5.80%) |
+|            1 |             1024 |            1024 |          3148.18   |           36.8031 | 4.81 GB (6.07%) |
+|            1 |             2048 |            2048 |          2927.33   |           35.2676 | 5.73 GB (7.23%) |
+
+*fused case*
+
+|   Batch Size |   Prefill Length |   Decode Length |   Prefill tokens/s |   Decode tokens/s | Memory (VRAM)   |
+|-------------:|-----------------:|----------------:|-------------------:|------------------:|:----------------|
+|            1 |               32 |              32 |            81.4899 |           80.2569 | 4.00 GB (5.05%) |
+|            1 |               64 |              64 |          1756.1    |          106.26   | 4.00 GB (5.05%) |
+|            1 |              128 |             128 |          2479.32   |          105.631  | 4.00 GB (5.06%) |
+|            1 |              256 |             256 |          1813.6    |           85.7485 | 4.01 GB (5.06%) |
+|            1 |              512 |             512 |          2848.9    |           97.701  | 4.11 GB (5.19%) |
+|            1 |             1024 |            1024 |          3044.35   |           87.7323 | 4.41 GB (5.57%) |
+|            1 |             2048 |            2048 |          2715.11   |           89.4709 | 5.57 GB (7.04%) |
+
+We also performed benchmarks with [`optimum-benchmark`](https://github.com/huggingface/optimum-benchmark) library. And below are the results:
+
+<div style="text-align: center">
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/quantization/fused_forward_memory_plot.png">
+</div>
+
+<div style="text-align: center">
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/quantization/fused_generate_throughput_plot.png">
+</div>
+
+
 ## AutoGPTQ
 
 <Tip>

docs/source/en/tasks/language_modeling.md

@@ -37,7 +37,7 @@ You can finetune other architectures for causal language modeling following the
 Choose one of the following architectures:
 
 <!--This tip is automatically generated by `make fix-copies`, do not fill manually!-->
-[BART](../model_doc/bart), [BERT](../model_doc/bert), [Bert Generation](../model_doc/bert-generation), [BigBird](../model_doc/big_bird), [BigBird-Pegasus](../model_doc/bigbird_pegasus), [BioGpt](../model_doc/biogpt), [Blenderbot](../model_doc/blenderbot), [BlenderbotSmall](../model_doc/blenderbot-small), [BLOOM](../model_doc/bloom), [CamemBERT](../model_doc/camembert), [CodeLlama](../model_doc/code_llama), [CodeGen](../model_doc/codegen), [CPM-Ant](../model_doc/cpmant), [CTRL](../model_doc/ctrl), [Data2VecText](../model_doc/data2vec-text), [ELECTRA](../model_doc/electra), [ERNIE](../model_doc/ernie), [Falcon](../model_doc/falcon), [Fuyu](../model_doc/fuyu), [GIT](../model_doc/git), [GPT-Sw3](../model_doc/gpt-sw3), [OpenAI GPT-2](../model_doc/gpt2), [GPTBigCode](../model_doc/gpt_bigcode), [GPT Neo](../model_doc/gpt_neo), [GPT NeoX](../model_doc/gpt_neox), [GPT NeoX Japanese](../model_doc/gpt_neox_japanese), [GPT-J](../model_doc/gptj), [LLaMA](../model_doc/llama), [Marian](../model_doc/marian), [mBART](../model_doc/mbart), [MEGA](../model_doc/mega), [Megatron-BERT](../model_doc/megatron-bert), [Mistral](../model_doc/mistral), [MPT](../model_doc/mpt), [MusicGen](../model_doc/musicgen), [MVP](../model_doc/mvp), [OpenLlama](../model_doc/open-llama), [OpenAI GPT](../model_doc/openai-gpt), [OPT](../model_doc/opt), [Pegasus](../model_doc/pegasus), [Persimmon](../model_doc/persimmon), [Phi](../model_doc/phi), [PLBart](../model_doc/plbart), [ProphetNet](../model_doc/prophetnet), [QDQBert](../model_doc/qdqbert), [Reformer](../model_doc/reformer), [RemBERT](../model_doc/rembert), [RoBERTa](../model_doc/roberta), [RoBERTa-PreLayerNorm](../model_doc/roberta-prelayernorm), [RoCBert](../model_doc/roc_bert), [RoFormer](../model_doc/roformer), [RWKV](../model_doc/rwkv), [Speech2Text2](../model_doc/speech_to_text_2), [Transformer-XL](../model_doc/transfo-xl), [TrOCR](../model_doc/trocr), [Whisper](../model_doc/whisper), [XGLM](../model_doc/xglm), [XLM](../model_doc/xlm), [XLM-ProphetNet](../model_doc/xlm-prophetnet), [XLM-RoBERTa](../model_doc/xlm-roberta), [XLM-RoBERTa-XL](../model_doc/xlm-roberta-xl), [XLNet](../model_doc/xlnet), [X-MOD](../model_doc/xmod)
+[BART](../model_doc/bart), [BERT](../model_doc/bert), [Bert Generation](../model_doc/bert-generation), [BigBird](../model_doc/big_bird), [BigBird-Pegasus](../model_doc/bigbird_pegasus), [BioGpt](../model_doc/biogpt), [Blenderbot](../model_doc/blenderbot), [BlenderbotSmall](../model_doc/blenderbot-small), [BLOOM](../model_doc/bloom), [CamemBERT](../model_doc/camembert), [CodeLlama](../model_doc/code_llama), [CodeGen](../model_doc/codegen), [CPM-Ant](../model_doc/cpmant), [CTRL](../model_doc/ctrl), [Data2VecText](../model_doc/data2vec-text), [ELECTRA](../model_doc/electra), [ERNIE](../model_doc/ernie), [Falcon](../model_doc/falcon), [Fuyu](../model_doc/fuyu), [GIT](../model_doc/git), [GPT-Sw3](../model_doc/gpt-sw3), [OpenAI GPT-2](../model_doc/gpt2), [GPTBigCode](../model_doc/gpt_bigcode), [GPT Neo](../model_doc/gpt_neo), [GPT NeoX](../model_doc/gpt_neox), [GPT NeoX Japanese](../model_doc/gpt_neox_japanese), [GPT-J](../model_doc/gptj), [LLaMA](../model_doc/llama), [Marian](../model_doc/marian), [mBART](../model_doc/mbart), [MEGA](../model_doc/mega), [Megatron-BERT](../model_doc/megatron-bert), [Mistral](../model_doc/mistral), [Mixtral](../model_doc/mixtral), [MPT](../model_doc/mpt), [MusicGen](../model_doc/musicgen), [MVP](../model_doc/mvp), [OpenLlama](../model_doc/open-llama), [OpenAI GPT](../model_doc/openai-gpt), [OPT](../model_doc/opt), [Pegasus](../model_doc/pegasus), [Persimmon](../model_doc/persimmon), [Phi](../model_doc/phi), [PLBart](../model_doc/plbart), [ProphetNet](../model_doc/prophetnet), [QDQBert](../model_doc/qdqbert), [Reformer](../model_doc/reformer), [RemBERT](../model_doc/rembert), [RoBERTa](../model_doc/roberta), [RoBERTa-PreLayerNorm](../model_doc/roberta-prelayernorm), [RoCBert](../model_doc/roc_bert), [RoFormer](../model_doc/roformer), [RWKV](../model_doc/rwkv), [Speech2Text2](../model_doc/speech_to_text_2), [Transformer-XL](../model_doc/transfo-xl), [TrOCR](../model_doc/trocr), [Whisper](../model_doc/whisper), [XGLM](../model_doc/xglm), [XLM](../model_doc/xlm), [XLM-ProphetNet](../model_doc/xlm-prophetnet), [XLM-RoBERTa](../model_doc/xlm-roberta), [XLM-RoBERTa-XL](../model_doc/xlm-roberta-xl), [XLNet](../model_doc/xlnet), [X-MOD](../model_doc/xmod)
 
 
 

docs/source/en/tasks/semantic_segmentation.md

@@ -23,8 +23,8 @@ rendered properly in your Markdown viewer.
 Image segmentation models separate areas corresponding to different areas of interest in an image. These models work by assigning a label to each pixel. There are several types of segmentation: semantic segmentation, instance segmentation, and panoptic segmentation.
 
 In this guide, we will:
-1. [Take a look at different types of segmentation](#Types-of-Segmentation),
-2. [Have an end-to-end fine-tuning example for semantic segmentation](#Fine-tuning-a-Model-for-Segmentation). 
+1. [Take a look at different types of segmentation](#types-of-segmentation).
+2. [Have an end-to-end fine-tuning example for semantic segmentation](#fine-tuning-a-model-for-segmentation).
 
 Before you begin, make sure you have all the necessary libraries installed:
 
@@ -256,6 +256,59 @@ You'll also want to create a dictionary that maps a label id to a label class wh
 >>> num_labels = len(id2label)
 ```
 
+#### Custom dataset
+
+You could also create and use your own dataset if you prefer to train with the [run_semantic_segmentation.py](https://github.com/huggingface/transformers/blob/main/examples/pytorch/semantic-segmentation/run_semantic_segmentation.py) script instead of a notebook instance. The script requires:
+
+1. a [`~datasets.DatasetDict`] with two [`~datasets.Image`] columns, "image" and "label"
+
+     ```py
+     from datasets import Dataset, DatasetDict, Image
+
+     image_paths_train = ["path/to/image_1.jpg/jpg", "path/to/image_2.jpg/jpg", ..., "path/to/image_n.jpg/jpg"]
+     label_paths_train = ["path/to/annotation_1.png", "path/to/annotation_2.png", ..., "path/to/annotation_n.png"]
+
+     image_paths_validation = [...]
+     label_paths_validation = [...]
+
+     def create_dataset(image_paths, label_paths):
+         dataset = Dataset.from_dict({"image": sorted(image_paths),
+                                     "label": sorted(label_paths)})
+         dataset = dataset.cast_column("image", Image())
+         dataset = dataset.cast_column("label", Image())
+
+     return dataset
+
+     # step 1: create Dataset objects
+     train_dataset = create_dataset(image_paths_train, label_paths_train)
+     validation_dataset = create_dataset(image_paths_validation, label_paths_validation)
+
+     # step 2: create DatasetDict
+     dataset = DatasetDict({
+          "train": train_dataset,
+          "validation": validation_dataset,
+          }
+     )
+
+     # step 3: push to Hub (assumes you have ran the huggingface-cli login command in a terminal/notebook)
+     dataset.push_to_hub("your-name/dataset-repo")
+
+     # optionally, you can push to a private repo on the Hub
+     # dataset.push_to_hub("name of repo on the hub", private=True)
+     ```
+
+2. an id2label dictionary mapping the class integers to their class names
+
+     ```py
+     import json
+     # simple example
+     id2label = {0: 'cat', 1: 'dog'}
+     with open('id2label.json', 'w') as fp:
+     json.dump(id2label, fp)
+     ```
+
+As an example, take a look at this [example dataset](https://huggingface.co/datasets/nielsr/ade20k-demo) which was created with the steps shown above.
+
 ### Preprocess
 
 The next step is to load a SegFormer image processor to prepare the images and annotations for the model. Some datasets, like this one, use the zero-index as the background class. However, the background class isn't actually included in the 150 classes, so you'll need to set `reduce_labels=True` to subtract one from all the labels. The zero-index is replaced by `255` so it's ignored by SegFormer's loss function:

docs/source/en/tasks/sequence_classification.md

@@ -33,7 +33,7 @@ The task illustrated in this tutorial is supported by the following model archit
 <!--This tip is automatically generated by `make fix-copies`, do not fill manually!-->
 
 
-[ALBERT](../model_doc/albert), [BART](../model_doc/bart), [BERT](../model_doc/bert), [BigBird](../model_doc/big_bird), [BigBird-Pegasus](../model_doc/bigbird_pegasus), [BioGpt](../model_doc/biogpt), [BLOOM](../model_doc/bloom), [CamemBERT](../model_doc/camembert), [CANINE](../model_doc/canine), [CodeLlama](../model_doc/code_llama), [ConvBERT](../model_doc/convbert), [CTRL](../model_doc/ctrl), [Data2VecText](../model_doc/data2vec-text), [DeBERTa](../model_doc/deberta), [DeBERTa-v2](../model_doc/deberta-v2), [DistilBERT](../model_doc/distilbert), [ELECTRA](../model_doc/electra), [ERNIE](../model_doc/ernie), [ErnieM](../model_doc/ernie_m), [ESM](../model_doc/esm), [Falcon](../model_doc/falcon), [FlauBERT](../model_doc/flaubert), [FNet](../model_doc/fnet), [Funnel Transformer](../model_doc/funnel), [GPT-Sw3](../model_doc/gpt-sw3), [OpenAI GPT-2](../model_doc/gpt2), [GPTBigCode](../model_doc/gpt_bigcode), [GPT Neo](../model_doc/gpt_neo), [GPT NeoX](../model_doc/gpt_neox), [GPT-J](../model_doc/gptj), [I-BERT](../model_doc/ibert), [LayoutLM](../model_doc/layoutlm), [LayoutLMv2](../model_doc/layoutlmv2), [LayoutLMv3](../model_doc/layoutlmv3), [LED](../model_doc/led), [LiLT](../model_doc/lilt), [LLaMA](../model_doc/llama), [Longformer](../model_doc/longformer), [LUKE](../model_doc/luke), [MarkupLM](../model_doc/markuplm), [mBART](../model_doc/mbart), [MEGA](../model_doc/mega), [Megatron-BERT](../model_doc/megatron-bert), [Mistral](../model_doc/mistral), [MobileBERT](../model_doc/mobilebert), [MPNet](../model_doc/mpnet), [MPT](../model_doc/mpt), [MRA](../model_doc/mra), [MT5](../model_doc/mt5), [MVP](../model_doc/mvp), [Nezha](../model_doc/nezha), [Nyströmformer](../model_doc/nystromformer), [OpenLlama](../model_doc/open-llama), [OpenAI GPT](../model_doc/openai-gpt), [OPT](../model_doc/opt), [Perceiver](../model_doc/perceiver), [Persimmon](../model_doc/persimmon), [Phi](../model_doc/phi), [PLBart](../model_doc/plbart), [QDQBert](../model_doc/qdqbert), [Reformer](../model_doc/reformer), [RemBERT](../model_doc/rembert), [RoBERTa](../model_doc/roberta), [RoBERTa-PreLayerNorm](../model_doc/roberta-prelayernorm), [RoCBert](../model_doc/roc_bert), [RoFormer](../model_doc/roformer), [SqueezeBERT](../model_doc/squeezebert), [T5](../model_doc/t5), [TAPAS](../model_doc/tapas), [Transformer-XL](../model_doc/transfo-xl), [UMT5](../model_doc/umt5), [XLM](../model_doc/xlm), [XLM-RoBERTa](../model_doc/xlm-roberta), [XLM-RoBERTa-XL](../model_doc/xlm-roberta-xl), [XLNet](../model_doc/xlnet), [X-MOD](../model_doc/xmod), [YOSO](../model_doc/yoso)
+[ALBERT](../model_doc/albert), [BART](../model_doc/bart), [BERT](../model_doc/bert), [BigBird](../model_doc/big_bird), [BigBird-Pegasus](../model_doc/bigbird_pegasus), [BioGpt](../model_doc/biogpt), [BLOOM](../model_doc/bloom), [CamemBERT](../model_doc/camembert), [CANINE](../model_doc/canine), [CodeLlama](../model_doc/code_llama), [ConvBERT](../model_doc/convbert), [CTRL](../model_doc/ctrl), [Data2VecText](../model_doc/data2vec-text), [DeBERTa](../model_doc/deberta), [DeBERTa-v2](../model_doc/deberta-v2), [DistilBERT](../model_doc/distilbert), [ELECTRA](../model_doc/electra), [ERNIE](../model_doc/ernie), [ErnieM](../model_doc/ernie_m), [ESM](../model_doc/esm), [Falcon](../model_doc/falcon), [FlauBERT](../model_doc/flaubert), [FNet](../model_doc/fnet), [Funnel Transformer](../model_doc/funnel), [GPT-Sw3](../model_doc/gpt-sw3), [OpenAI GPT-2](../model_doc/gpt2), [GPTBigCode](../model_doc/gpt_bigcode), [GPT Neo](../model_doc/gpt_neo), [GPT NeoX](../model_doc/gpt_neox), [GPT-J](../model_doc/gptj), [I-BERT](../model_doc/ibert), [LayoutLM](../model_doc/layoutlm), [LayoutLMv2](../model_doc/layoutlmv2), [LayoutLMv3](../model_doc/layoutlmv3), [LED](../model_doc/led), [LiLT](../model_doc/lilt), [LLaMA](../model_doc/llama), [Longformer](../model_doc/longformer), [LUKE](../model_doc/luke), [MarkupLM](../model_doc/markuplm), [mBART](../model_doc/mbart), [MEGA](../model_doc/mega), [Megatron-BERT](../model_doc/megatron-bert), [Mistral](../model_doc/mistral), [Mixtral](../model_doc/mixtral), [MobileBERT](../model_doc/mobilebert), [MPNet](../model_doc/mpnet), [MPT](../model_doc/mpt), [MRA](../model_doc/mra), [MT5](../model_doc/mt5), [MVP](../model_doc/mvp), [Nezha](../model_doc/nezha), [Nyströmformer](../model_doc/nystromformer), [OpenLlama](../model_doc/open-llama), [OpenAI GPT](../model_doc/openai-gpt), [OPT](../model_doc/opt), [Perceiver](../model_doc/perceiver), [Persimmon](../model_doc/persimmon), [Phi](../model_doc/phi), [PLBart](../model_doc/plbart), [QDQBert](../model_doc/qdqbert), [Reformer](../model_doc/reformer), [RemBERT](../model_doc/rembert), [RoBERTa](../model_doc/roberta), [RoBERTa-PreLayerNorm](../model_doc/roberta-prelayernorm), [RoCBert](../model_doc/roc_bert), [RoFormer](../model_doc/roformer), [SqueezeBERT](../model_doc/squeezebert), [T5](../model_doc/t5), [TAPAS](../model_doc/tapas), [Transformer-XL](../model_doc/transfo-xl), [UMT5](../model_doc/umt5), [XLM](../model_doc/xlm), [XLM-RoBERTa](../model_doc/xlm-roberta), [XLM-RoBERTa-XL](../model_doc/xlm-roberta-xl), [XLNet](../model_doc/xlnet), [X-MOD](../model_doc/xmod), [YOSO](../model_doc/yoso)
 
 
 

docs/source/es/_toctree.yml

@@ -75,6 +75,10 @@
 - sections:
   - local: philosophy
     title: Filosofía
+  - local: pad_truncation
+    title: Relleno y truncamiento
   - local: bertology
     title: BERTología
+  - local: perplexity
+    title: Perplejidad de los modelos de longitud fija
   title: Guías conceptuales

docs/source/es/pad_truncation.md

@@ -0,0 +1,69 @@
+<!--Copyright 2022 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.
+
+-->
+
+# Relleno y truncamiento
+
+Las entradas agrupadas por lotes (batched) suelen tener longitudes diferentes, por lo que no se pueden convertir en tensores de tamaño fijo. El relleno (también conocido como "Padding") y el truncamiento (conocido como "Truncation") son estrategias para abordar este problema y crear tensores rectangulares a partir de lotes de longitudes variables. El relleno agrega un **padding token** especial para garantizar que las secuencias más cortas tengan la misma longitud que la secuencia más larga en un lote o la longitud máxima aceptada por el modelo. El truncamiento funciona en la otra dirección al truncar secuencias largas.
+
+En la mayoría de los casos, es bastante eficaz rellenar el lote hasta la longitud de la secuencia más larga y truncar hasta la longitud máxima que un modelo puede aceptar. Sin embargo, la API admite más estrategias si las necesitas. Los tres argumentos que necesitas son: `padding`, `truncation` y `max_length`.
+
+El argumento `padding` controla el relleno. Puede ser un booleano o una cadena:
+
+  - `True` o `'longest'`: rellena hasta la longitud de la secuencia más larga en el lote (no se aplica relleno si solo proporcionas una única secuencia).
+  - `'max_length'`: rellena hasta una longitud especificada por el argumento `max_length` o la longitud máxima aceptada
+    por el modelo si no se proporciona `max_length` (`max_length=None`). El relleno se aplicará incluso si solo proporcionas una única secuencia.
+  - `False` o `'do_not_pad'`: no se aplica relleno. Este es el comportamiento predeterminado.
+
+El argumento `truncation` controla el truncamiento. Puede ser un booleano o una cadena:
+
+  - `True` o `'longest_first'`: trunca hasta una longitud máxima especificada por el argumento `max_length` o
+    la longitud máxima aceptada por el modelo si no se proporciona `max_length` (`max_length=None`). Esto
+    truncará token por token, eliminando un token de la secuencia más larga en el par hasta alcanzar la longitud adecuada.
+  - `'only_second'`: trunca hasta una longitud máxima especificada por el argumento `max_length` o la longitud máxima
+    aceptada por el modelo si no se proporciona `max_length` (`max_length=None`). Esto solo truncará
+    la segunda oración de un par si se proporciona un par de secuencias (o un lote de pares de secuencias).
+  - `'only_first'`: trunca hasta una longitud máxima especificada por el argumento `max_length` o la longitud máxima
+    aceptada por el modelo si no se proporciona `max_length` (`max_length=None`). Esto solo truncará
+    la primera oración de un par si se proporciona un par de secuencias (o un lote de pares de secuencias).
+  - `False` o `'do_not_truncate'`: no se aplica truncamiento. Este es el comportamiento predeterminado.
+
+El argumento `max_length` controla la longitud del relleno y del truncamiento. Puede ser un número entero o `None`, en cuyo caso se establecerá automáticamente en la longitud máxima que el modelo puede aceptar. Si el modelo no tiene una longitud máxima de entrada específica, se desactiva el truncamiento o el relleno hasta `max_length`.
+
+La siguiente tabla resume la forma recomendada de configurar el relleno y el truncamiento. Si usas pares de secuencias de entrada en alguno de los siguientes ejemplos, puedes reemplazar `truncation=True` por una `ESTRATEGIA` seleccionada en
+`['only_first', 'only_second', 'longest_first']`, es decir, `truncation='only_second'` o `truncation='longest_first'` para controlar cómo se truncan ambas secuencias en el par, como se detalló anteriormente.
+
+| Truncation                           | Padding                             | Instrucción                                                                                 |
+|-----------------------------------------|--------------------------------------|---------------------------------------------------------------------------------------------|
+| sin truncamiento                       | sin relleno                          | `tokenizer(batch_sentences)`                                                           |
+|                                         | relleno hasta la longitud máxima del lote | `tokenizer(batch_sentences, padding=True)` o                                          |
+|                                         |                                      | `tokenizer(batch_sentences, padding='longest')`                                        |
+|                                         | relleno hasta la longitud máxima del modelo | `tokenizer(batch_sentences, padding='max_length')`                                     |
+|                                         | relleno hasta una longitud específica | `tokenizer(batch_sentences, padding='max_length', max_length=42)`                      |
+|                                         | relleno hasta un múltiplo de un valor | `tokenizer(batch_sentences, padding=True, pad_to_multiple_of=8)`                        |
+| truncamiento hasta la longitud máxima del modelo | sin relleno                          | `tokenizer(batch_sentences, truncation=True)` o                                       |
+|                                         |                                      | `tokenizer(batch_sentences, truncation=ESTRATEGIA)`                                      |
+|                                         | relleno hasta la longitud máxima del lote | `tokenizer(batch_sentences, padding=True, truncation=True)` o                         |
+|                                         |                                      | `tokenizer(batch_sentences, padding=True, truncation=ESTRATEGIA)`                        |
+|                                         | relleno hasta la longitud máxima del modelo | `tokenizer(batch_sentences, padding='max_length', truncation=True)` o                 |
+|                                         |                                      | `tokenizer(batch_sentences, padding='max_length', truncation=ESTRATEGIA)`                |
+|                                         | relleno hasta una longitud específica | No es posible                                                                                |
+| truncamiento hasta una longitud específica | sin relleno                          | `tokenizer(batch_sentences, truncation=True, max_length=42)` o                        |
+|                                         |                                      | `tokenizer(batch_sentences, truncation=ESTRATEGIA, max_length=42)`                       |
+|                                         | relleno hasta la longitud máxima del lote | `tokenizer(batch_sentences, padding=True, truncation=True, max_length=42)` o          |
+|                                         |                                      | `tokenizer(batch_sentences, padding=True, truncation=ESTRATEGIA, max_length=42)`         |
+|                                         | relleno hasta la longitud máxima del modelo | No es posible                                                                                |
+|                                         | relleno hasta una longitud específica | `tokenizer(batch_sentences, padding='max_length', truncation=True, max_length=42)` o  |
+|                                         |                                      | `tokenizer(batch_sentences, padding='max_length', truncation=ESTRATEGIA, max_length=42)` |

docs/source/es/perplexity.md

@@ -0,0 +1,116 @@
+<!--Copyright 2020 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+
+⚠️ 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.
+
+-->
+
+# Perplejidad de los modelos de longitud fija
+
+[[open-in-colab]]
+
+La perplejidad, perplexity en inglés (PPL), es una de las métricas más comunes para evaluar modelos de lenguaje. Antes de sumergirnos, debemos tener en cuenta que esta métrica se aplica específicamente a modelos de lenguaje clásicos (a veces llamados modelos autorregresivos o causales) y no está bien definida para modelos de lenguaje enmascarados como BERT (ver [resumen del modelo](model_summary)).
+
+La perplejidad se define como la media negativa exponenciada del log-likelihood de una secuencia. Si tenemos una secuencia tokenizada \\(X = (x_0, x_1, \dots, x_t)\\), entonces la perplejidad de \\(X\\) es,
+
+$$\text{PPL}(X) = \exp \left\{ {-\frac{1}{t}\sum_i^t \log p_\theta (x_i|x_{<i}) } \right\}$$
+
+donde \\(\log p_\theta (x_i|x_{<i})\\) es el log-likelihood del token i-ésimo condicionado a los tokens precedentes \\(x_{<i}\\) según nuestro modelo. De manera intuitiva, se puede pensar en esto como una evaluación de la capacidad del modelo para predecir de manera uniforme entre el conjunto de tokens especificados en un corpus. Es importante destacar que el procedimiento de tokenización tiene un impacto directo en la perplejidad de un modelo, lo cual siempre debe tenerse en cuenta al comparar diferentes modelos.
+
+Esto también es equivalente a la exponenciación de la entropía cruzada entre los datos y las predicciones del modelo. Para obtener más intuición sobre la perplejidad y su relación con los Bits Por Carácter (BPC) y la compresión de datos, echa un vistazo a esta [fantástica publicación en el blog de "The Gradient"](https://thegradient.pub/understanding-evaluation-metrics-for-language-models/).
+
+## Cálculo de PPL con modelos de longitud fija
+
+Si no estuviéramos limitados por el tamaño del contexto de un modelo, evaluaríamos la perplejidad (PPL) del modelo auto regresivamente factorizando una secuencia y condicionándonos en toda la subsecuencia precedente en cada paso, como se muestra a continuación.
+
+<img width="600" alt="Full decomposition of a sequence with unlimited context length" src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/ppl_full.gif"/>
+
+Sin embargo, al trabajar con modelos aproximados, generalmente tenemos una restricción en la cantidad de tokens que el modelo puede procesar. La versión más grande de [GPT-2](model_doc/gpt2), por ejemplo, tiene una longitud fija de 1024 tokens, por lo que no podemos calcular \\(p_\theta(x_t|x_{<t})\\) directamente cuando \\(t\\) es mayor que 1024.
+
+En cambio, la secuencia se divide típicamente en subsecuencias iguales al tamaño máximo de entrada del modelo. Si el tamaño máximo de entrada, de un modelo es \\(k\\), entonces aproximamos la probabilidad de un token \\(x_t\\) condicionándonos solo en los \\(k-1\\) tokens que lo preceden en lugar de todo el contexto. Al evaluar la perplejidad del modelo en una secuencia, un enfoque tentador pero sub óptimo es dividir la secuencia en fragmentos independientes y sumar los log-likelihood descompuestos de cada segmento de manera independiente.
+
+<img width="600" alt="Suboptimal PPL not taking advantage of full available context" src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/ppl_chunked.gif"/>
+
+Esto es rápido de calcular, ya que la perplejidad de cada segmento se puede calcular en un solo pase hacia adelante, pero sirve como una aproximación pobre de la perplejidad completamente factorizada y generalmente dará como resultado una PPL más alta (peor) porque el modelo tendrá menos contexto en la mayoría de los pasos de predicción.
+
+En cambio, la PPL de modelos de longitud fija debería evaluarse con una estrategia de ventana deslizante. Esto implica deslizar repetidamente la ventana de contexto para que el modelo tenga más contexto al hacer cada predicción.
+
+<img width="600" alt="Sliding window PPL taking advantage of all available context" src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/ppl_sliding.gif"/>
+
+Esta es una aproximación más cercana a la verdadera descomposición de la probabilidad de la secuencia y generalmente dará como resultado una puntuación más favorable. La desventaja es que requiere un pase hacia adelante separado para cada token en el corpus. Un buen compromiso práctico es emplear una ventana deslizante estratificada, moviendo el contexto con pasos más grandes en lugar de deslizarse de 1 token a la vez. Esto permite que la computación avance mucho más rápido, mientras le da al modelo un contexto amplio para hacer
+predicciones en cada paso.
+
+## Ejemplo: Cálculo de la perplejidad con GPT-2 en 🤗 Transformers
+
+Demostremos este proceso con GPT-2.
+
+```python
+from transformers import GPT2LMHeadModel, GPT2TokenizerFast
+
+device = "cuda"
+model_id = "gpt2-large"
+model = GPT2LMHeadModel.from_pretrained(model_id).to(device)
+tokenizer = GPT2TokenizerFast.from_pretrained(model_id)
+```
+
+Carguemos el conjunto de datos WikiText-2 y evaluemos la perplejidad utilizando algunas estrategias de ventana deslizante diferentes. Dado que este conjunto de datos es pequeño y solo estamos realizando un pase hacia adelante sobre el conjunto, podemos cargar y codificar todo el conjunto de datos en la memoria.
+
+```python
+from datasets import load_dataset
+
+test = load_dataset("wikitext", "wikitext-2-raw-v1", split="test")
+encodings = tokenizer("\n\n".join(test["text"]), return_tensors="pt")
+```
+
+Con 🤗 Transformers, simplemente podemos pasar los `input_ids` como las `labels` a nuestro modelo, y la media negativa del log-likelihood para cada token se devuelve como la pérdida. Sin embargo, con nuestro enfoque de ventana deslizante, hay superposición en los tokens que pasamos al modelo en cada iteración. No queremos que el log-likelihood de los tokens que estamos tratando solo como contexto se incluya en nuestra pérdida, por lo que podemos establecer estos objetivos en `-100` para que se ignoren. El siguiente es un ejemplo de cómo podríamos hacer esto con un paso de `512`. Esto significa que el modelo tendrá al menos `512` tokens como contexto al calcular el log-likelihood condicional de cualquier token (siempre que haya `512` tokens precedentes disponibles para condicionar).
+
+```python
+import torch
+from tqdm import tqdm
+
+max_length = model.config.n_positions
+stride = 512
+seq_len = encodings.input_ids.size(1)
+
+nlls = []
+prev_end_loc = 0
+for begin_loc in tqdm(range(0, seq_len, stride)):
+    end_loc = min(begin_loc + max_length, seq_len)
+    trg_len = end_loc - prev_end_loc  # puede ser diferente del paso en el último bucle
+    input_ids = encodings.input_ids[:, begin_loc:end_loc].to(device)
+    target_ids = input_ids.clone()
+    target_ids[:, :-trg_len] = -100
+
+    with torch.no_grad():
+        outputs = model(input_ids, labels=target_ids)
+
+        # la pérdida se calcula utilizando CrossEntropyLoss, que promedia las etiquetas válidas
+        # N.B. el modelo solo calcula la pérdida sobre trg_len - 1 etiquetas, porque desplaza las etiqueta internamente
+        # a la izquierda por 1.
+        neg_log_likelihood = outputs.loss
+
+    nlls.append(neg_log_likelihood)
+
+    prev_end_loc = end_loc
+    if end_loc == seq_len:
+        break
+
+ppl = torch.exp(torch.stack(nlls).mean())
+```
+
+Ejecuta esto con la longitud de paso igual a la longitud máxima de entrada es equivalente a la estrategia sub óptima,
+sin ventana deslizante, que discutimos anteriormente. Cuanto menor sea el paso, más contexto tendrá el modelo para
+realizar cada predicción y, por lo general, mejor será la perplejidad informada.
+
+Cuando ejecutamos lo anterior con `stride = 1024`, es decir, sin superposición, la PPL resultante es `19.44`, que es
+aproximadamente la misma que la `19.93` informada en el artículo de GPT-2. Al utilizar `stride = 512` y, por lo tanto,
+emplear nuestra estrategia de ventana deslizante, esto disminuye a `16.45`. Esto no solo es una puntuación más favorable, sino que se calcula de una manera más cercana a la verdadera descomposición autorregresiva de la probabilidad de una secuencia.

docs/source/fr/_toctree.yml

@@ -1,156 +1,30 @@
 - sections:
-  - local: index
-    title: 🤗 Transformers
-  - local: quicktour
-    title: Visite rapide
-  - local: installation
-    title: Installation
+    - local: index
+      title: 🤗 Transformers
+    - local: quicktour
+      title: Visite rapide
+    - local: installation
+      title: Installation
   title: Démarrer
 - sections:
-  - local: in_translation
-    title: Pipelines pour l'inférence
-  - local: in_translation
-    title: Chargement d'instances pré-entraînées avec une AutoClass
-  - local: in_translation
-    title: Préparation des données
-  - local: in_translation
-    title: Fine-tune un modèle pré-entraîné
-  - local: in_translation
-    title: Entraînement distribué avec 🤗 Accelerate
-  - local: in_translation
-    title: Partager un modèle
-  title: Tutoriels
-- sections:
-  - sections:
     - local: in_translation
-      title: Créer votre architecture
+      title: Pipelines pour l'inférence
+    - local: autoclass_tutorial
+      title: Chargement d'instances pré-entraînées avec une AutoClass
     - local: in_translation
-      title: Partager vos modèles
+      title: Préparation des données
+    - local: in_translation
+      title: Fine-tune un modèle pré-entraîné
     - local: in_translation
       title: Entraînement avec un script
     - local: in_translation
-      title: Entraînement avec Amazon SageMaker
+      title: Entraînement distribué avec 🤗 Accelerate
     - local: in_translation
-      title: Convertir depuis des checkpoints Tensorflow
+      title: Chargement et entraînement des adaptateurs avec 🤗 PEFT
     - local: in_translation
-      title: Exporter vers ONNX
+      title: Partager un modèle
     - local: in_translation
-      title: Exporter vers TorchScript
+      title: Agents
     - local: in_translation
-      title: Aide au dépannage
-    title: Usage général
-  - sections:
-    - local: in_translation
-      title: Utiliser les tokenizers de 🤗 Tokenizers
-    - local: in_translation
-      title: Inférence avec les modèles multilingues
-    - local: in_translation
-      title: Stratégies de génération de texte
-    - sections:
-      - isExpanded: false
-        local: in_translation
-        title: Classification de texte
-      - local: in_translation
-        title: Classification de token
-      - local: in_translation
-        title: Système de question-réponse
-      - local: in_translation
-        title: Modélisation causale du langage
-      - local: in_translation
-        title: Modélisation du langage avec masque
-      - local: in_translation
-        title: Traduction
-      - local: in_translation
-        title: Génération de résumé
-      - local: in_translation
-        title: Question à choix multiple
-      title: Guides des tâches
-    title: Traitement automatique des langues
-  - sections:
-    - local: in_translation
-      title: Classification audio
-    - local: in_translation
-      title: Reconnaissance automatique de la parole
-    title: Audio
-  - sections:
-    - local: in_translation
-      title: Classification d'images
-    - local: in_translation
-      title: Segmentation sémantique
-    - local: in_translation
-      title: Classification de vidéos
-    - local: in_translation
-      title: Détection d'objets
-    title: Vision par ordinateur
-  - sections:
-    - local: in_translation
-      title: Performance et extensibilité
-  - sections:
-    - local: in_translation
-      title: Comment contribuer à transformers?
-    - local: in_translation
-      title: Comment ajouter un modèle à 🤗 Transformers?
-    - local: in_translation
-      title: Comment convertir un modèle 🤗 Transformers vers TensorFlow?
-    - local: in_translation
-      title: Comment ajouter un pipeline à 🤗 Transformers?
-    - local: in_translation
-      title: Tester
-    - local: in_translation
-      title: Vérification pour une Pull Request
-    title: Contribuer
-  - local: in_translation
-    title: 🤗 Transformers Notebooks
-  - local: in_translation
-    title: Ressources communautaires
-  - local: in_translation
-    title: Benchmarks
-  - local: in_translation
-    title: Migration à partir de versions précédentes
-  title: Guides d'utilisation
-- sections:
-  - local: in_translation
-    title: Philosophie
-  - local: in_translation
-    title: Glossaire
-  - local: in_translation
-    title: Qu'est ce 🤗 Transformers peut faire ?
-  - local: in_translation
-    title: Quelles tâches 🤗 Transformers peut résoudre ?
-  - local: in_translation
-    title: Résumé des modèles
-  - local: in_translation
-    title: Résumé des tokenizers
-  - local: in_translation
-    title: Remplissage et troncature
-  - local: in_translation
-    title: BERTology
-  - local: in_translation
-    title: Perplexité des modèles à longueur fixe
-  - local: in_translation
-    title: Pipelines pour inférence avec des serveurs web
-  title: Guides conceptuels
-- sections:
-  - isExpanded: false
-    sections:
-    - local: in_translation
-      title: Classes principales
-    - local: in_translation
-      title: Modèles textuels
-    - local: in_translation
-      title: Modèles visuels
-    - local: in_translation
-      title: Modèles audio
-    - local: in_translation
-      title: Modèles multimodal
-    - local: in_translation
-      title: Modèles d'apprentissage par renforcement
-    - local: in_translation
-      title: Modèles de séries temporelles
-    - local: in_translation
-      title: Graph models
-    title: Modèles
-  - sections:
-    - local: in_translation
-      title: Utilitaires internes
-  title: API
+      title: Génération avec LLMs
+  title: Tutoriels

docs/source/fr/autoclass_tutorial.md

@@ -0,0 +1,142 @@
+<!--Copyright 2022 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.
+
+-->
+
+# Chargement d'instances pré-entraînées avec une AutoClass
+
+Avec autant d'architectures Transformer différentes, il peut être difficile d'en créer une pour votre ensemble de poids (aussi appelés "weights" ou "checkpoint" en anglais). Dans l'idée de créer une librairie facile, simple et flexible à utiliser, 🤗 Transformers fournit une `AutoClass` qui infère et charge automatiquement l'architecture correcte à partir d'un ensemble de poids donné. La fonction `from_pretrained()` vous permet de charger rapidement un modèle pré-entraîné pour n'importe quelle architecture afin que vous n'ayez pas à consacrer du temps et des ressources à l'entraînement d'un modèle à partir de zéro. Produire un tel code indépendant d'un ensemble de poids signifie que si votre code fonctionne pour un ensemble de poids, il fonctionnera avec un autre ensemble - tant qu'il a été entraîné pour une tâche similaire - même si l'architecture est différente.
+
+<Tip>
+
+Rappel, l'architecture fait référence au squelette du modèle et l'ensemble de poids contient les poids pour une architecture donnée. Par exemple, [BERT](https://huggingface.co/bert-base-uncased) est une architecture, tandis que `bert-base-uncased` est un ensemble de poids. Le terme modèle est général et peut signifier soit architecture soit ensemble de poids.
+
+</Tip>
+
+Dans ce tutoriel, vous apprendrez à:
+
+  * Charger un tokenizer pré-entraîné.
+  * Charger un processeur d'image pré-entraîné.
+  * Charger un extracteur de caractéristiques pré-entraîné.
+  * Charger un processeur pré-entraîné.
+  * Charger un modèle pré-entraîné.
+
+## AutoTokenizer
+
+Quasiment toutes les tâches de traitement du langage (NLP) commencent avec un tokenizer. Un tokenizer convertit votre texte initial dans un format qui peut être traité par le modèle.
+
+Chargez un tokenizer avec [`AutoTokenizer.from_pretrained`]:
+
+```py
+>>> from transformers import AutoTokenizer
+
+>>> tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
+```
+
+Puis, transformez votre texte initial comme montré ci-dessous:
+
+```py
+>>> sequence = "In a hole in the ground there lived a hobbit."
+>>> print(tokenizer(sequence))
+{'input_ids': [101, 1999, 1037, 4920, 1999, 1996, 2598, 2045, 2973, 1037, 7570, 10322, 4183, 1012, 102], 
+ 'token_type_ids': [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
+ 'attention_mask': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]}
+```
+
+## AutoImageProcessor
+
+Pour les tâches de vision, un processeur d'image traite l'image pour la formater correctment.
+
+```py
+>>> from transformers import AutoImageProcessor
+
+>>> image_processor = AutoImageProcessor.from_pretrained("google/vit-base-patch16-224")
+```
+
+## AutoFeatureExtractor
+
+Pour les tâches audio, un extracteur de caractéristiques (aussi appelés "features" en anglais) traite le signal audio pour le formater correctement.
+
+Chargez un extracteur de caractéristiques avec [`AutoFeatureExtractor.from_pretrained`]:
+
+```py
+>>> from transformers import AutoFeatureExtractor
+
+>>> feature_extractor = AutoFeatureExtractor.from_pretrained(
+...     "ehcalabres/wav2vec2-lg-xlsr-en-speech-emotion-recognition"
+... )
+```
+
+## AutoProcessor
+
+Les tâches multimodales nécessitent un processeur qui combine deux types d'outils de prétraitement. Par exemple, le modèle [LayoutLMV2](model_doc/layoutlmv2) nécessite un processeur d'image pour traiter les images et un tokenizer pour traiter le texte ; un processeur combine les deux.
+
+Chargez un processeur avec [`AutoProcessor.from_pretrained`]:
+
+```py
+>>> from transformers import AutoProcessor
+
+>>> processor = AutoProcessor.from_pretrained("microsoft/layoutlmv2-base-uncased")
+```
+
+## AutoModel
+
+<frameworkcontent>
+<pt>
+Enfin, les classes `AutoModelFor` vous permettent de charger un modèle pré-entraîné pour une tâche donnée (voir [ici](model_doc/auto) pour une liste complète des tâches disponibles). Par exemple, chargez un modèle pour la classification de séquence avec [`AutoModelForSequenceClassification.from_pretrained`]:
+
+```py
+>>> from transformers import AutoModelForSequenceClassification
+
+>>> model = AutoModelForSequenceClassification.from_pretrained("distilbert-base-uncased")
+```
+
+Réutilisez facilement le même ensemble de poids pour charger une architecture pour une tâche différente :
+
+```py
+>>> from transformers import AutoModelForTokenClassification
+
+>>> model = AutoModelForTokenClassification.from_pretrained("distilbert-base-uncased")
+```
+
+<Tip warning={true}>
+
+Pour les modèles PyTorch, la fonction `from_pretrained()` utilise `torch.load()` qui utilise `pickle` en interne et est connu pour être non sécurisé. En général, ne chargez jamais un modèle qui pourrait provenir d'une source non fiable, ou qui pourrait avoir été altéré. Ce risque de sécurité est partiellement atténué pour les modèles hébergés publiquement sur le Hugging Face Hub, qui sont [scannés pour les logiciels malveillants](https://huggingface.co/docs/hub/security-malware) à chaque modification. Consultez la [documentation du Hub](https://huggingface.co/docs/hub/security) pour connaître les meilleures pratiques comme la [vérification des modifications signées](https://huggingface.co/docs/hub/security-gpg#signing-commits-with-gpg) avec GPG.
+
+Les points de contrôle TensorFlow et Flax ne sont pas concernés, et peuvent être chargés dans des architectures PyTorch en utilisant les arguments `from_tf` et `from_flax` de la fonction `from_pretrained` pour contourner ce problème.
+
+</Tip>
+
+En général, nous recommandons d'utiliser les classes `AutoTokenizer` et `AutoModelFor` pour charger des instances pré-entraînées de tokenizers et modèles respectivement. Cela vous permettra de charger la bonne architecture à chaque fois. Dans le prochain [tutoriel](preprocessing), vous apprenez à utiliser un tokenizer, processeur d'image, extracteur de caractéristiques et processeur pour pré-traiter un jeu de données pour le fine-tuning.
+</pt>
+<tf>
+Enfin, les classes `TFAutoModelFor` vous permettent de charger un modèle pré-entraîné pour une tâche donnée (voir [ici](model_doc/auto) pour une liste complète des tâches disponibles). Par exemple, chargez un modèle pour la classification de séquence avec [`TFAutoModelForSequenceClassification.from_pretrained`]:
+
+```py
+>>> from transformers import TFAutoModelForSequenceClassification
+
+>>> model = TFAutoModelForSequenceClassification.from_pretrained("distilbert-base-uncased")
+```
+
+Réutilisez facilement le même ensemble de poids pour charger une architecture pour une tâche différente :
+
+```py
+>>> from transformers import TFAutoModelForTokenClassification
+
+>>> model = TFAutoModelForTokenClassification.from_pretrained("distilbert-base-uncased")
+```
+
+En général, nous recommandons d'utiliser les classes `AutoTokenizer` et `TFAutoModelFor` pour charger des instances pré-entraînées de tokenizers et modèles respectivement. Cela vous permettra de charger la bonne architecture à chaque fois. Dans le prochain [tutoriel](preprocessing), vous apprenez à utiliser un tokenizer, processeur d'image, extracteur de caractéristiques et processeur pour pré-traiter un jeu de données pour le fine-tuning.
+</tf>
+</frameworkcontent>

docs/source/ja/_toctree.yml

@@ -274,6 +274,24 @@
         title: Blenderbot
       - local: model_doc/blenderbot-small
         title: Blenderbot Small
+      - local: model_doc/bloom
+        title: BLOOM
+      - local: model_doc/bort
+        title: BORT
+      - local: model_doc/byt5
+        title: ByT5
+      - local: model_doc/camembert
+        title: CamemBERT
+      - local: model_doc/canine
+        title: CANINE
+      - local: model_doc/codegen
+        title: CodeGen
+      - local: model_doc/code_llama
+        title: CodeLlama
+      - local: model_doc/convbert
+        title: ConvBERT
+      - local: model_doc/cpm
+        title: CPM
       title: 文章モデル
     - isExpanded: false
       sections:
@@ -281,6 +299,12 @@
         title: BEiT
       - local: model_doc/bit
         title: BiT
+      - local: model_doc/conditional_detr
+        title: Conditional DETR
+      - local: model_doc/convnext
+        title: ConvNeXT
+      - local: model_doc/convnextv2
+        title: ConvNeXTV2
       title: ビジョンモデル
     - isExpanded: false
       sections:
@@ -288,6 +312,8 @@
         title: Audio Spectrogram Transformer
       - local: model_doc/bark
         title: Bark
+      - local: model_doc/clap
+        title: CLAP
       title: 音声モデル
     - isExpanded: false
       sections:
@@ -295,6 +321,22 @@
         title: ALIGN
       - local: model_doc/altclip
         title: AltCLIP
+      - local: model_doc/blip
+        title: BLIP
+      - local: model_doc/blip-2
+        title: BLIP-2
+      - local: model_doc/bridgetower
+        title: BridgeTower
+      - local: model_doc/bros
+        title: BROS
+      - local: model_doc/chinese_clip
+        title: Chinese-CLIP
+      - local: model_doc/clip
+        title: CLIP
+      - local: model_doc/clipseg
+        title: CLIPSeg
+      - local: model_doc/clvp
+        title: CLVP
       title: マルチモーダルモデル
     - isExpanded: false
       sections:

docs/source/ja/model_doc/bert-generation.md

@@ -19,7 +19,7 @@ rendered properly in your Markdown viewer.
 ## Overview
 
 BertGeneration モデルは、次を使用してシーケンス間のタスクに利用できる BERT モデルです。
-[シーケンス生成のための事前トレーニング済みチェックポイントの活用](https://arxiv.org/abs/1907.12461) で提案されている [`EncoderDecoderModel`]
+[Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) で提案されている [`EncoderDecoderModel`]
 タスク、Sascha Rothe、Sishi Nagayan、Aliaksei Severyn 著。
 
 論文の要約は次のとおりです。

docs/source/ja/model_doc/biogpt.md

@@ -18,7 +18,7 @@ rendered properly in your Markdown viewer.
 
 ## Overview
 
-BioGPT モデルは、[BioGPT: 生物医学テキストの生成とマイニングのための生成事前トレーニング済みトランスフォーマー](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、Tie-Yan Liu。 BioGPT は、生物医学テキストの生成とマイニングのための、ドメイン固有の生成事前トレーニング済み Transformer 言語モデルです。 BioGPT は、Transformer 言語モデルのバックボーンに従い、1,500 万の PubMed 抄録で最初から事前トレーニングされています。
+BioGPT モデルは、[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、Tie-Yan Liu。 BioGPT は、生物医学テキストの生成とマイニングのための、ドメイン固有の生成事前トレーニング済み Transformer 言語モデルです。 BioGPT は、Transformer 言語モデルのバックボーンに従い、1,500 万の PubMed 抄録で最初から事前トレーニングされています。
 
 論文の要約は次のとおりです。
 

docs/source/ja/model_doc/blenderbot-small.md

@@ -24,7 +24,7 @@ rendered properly in your Markdown viewer.
 
 ## Overview
 
-Blender チャットボット モデルは、[オープンドメイン チャットボットを構築するためのレシピ](https://arxiv.org/pdf/2004.13637.pdf) Stephen Roller、Emily Dinan、Naman Goyal、Da Ju、Mary Williamson、yinghan Liu、で提案されました。
+Blender チャットボット モデルは、[Recipes for building an open-domain chatbot](https://arxiv.org/pdf/2004.13637.pdf) Stephen Roller、Emily Dinan、Naman Goyal、Da Ju、Mary Williamson、yinghan Liu、で提案されました。
 ジン・シュー、マイル・オット、カート・シャスター、エリック・M・スミス、Y-ラン・ブーロー、ジェイソン・ウェストン、2020年4月30日。
 
 論文の要旨は次のとおりです。

docs/source/ja/model_doc/blenderbot.md

@@ -20,7 +20,7 @@ rendered properly in your Markdown viewer.
 
 ## Overview
 
-Blender チャットボット モデルは、[オープンドメイン チャットボットを構築するためのレシピ](https://arxiv.org/pdf/2004.13637.pdf) Stephen Roller、Emily Dinan、Naman Goyal、Da Ju、Mary Williamson、yinghan Liu、で提案されました。
+Blender チャットボット モデルは、[Recipes for building an open-domain chatbot](https://arxiv.org/pdf/2004.13637.pdf) Stephen Roller、Emily Dinan、Naman Goyal、Da Ju、Mary Williamson、yinghan Liu、で提案されました。
 ジン・シュー、マイル・オット、カート・シャスター、エリック・M・スミス、Y-ラン・ブーロー、ジェイソン・ウェストン、2020年4月30日。
 
 論文の要旨は次のとおりです。

docs/source/ja/model_doc/blip-2.md

@@ -0,0 +1,90 @@
+<!--Copyright 2023 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.
+
+-->
+
+# BLIP-2
+
+## Overview
+
+BLIP-2 モデルは、[BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://arxiv.org/abs/2301.12597) で提案されました。
+Junnan Li, Dongxu Li, Silvio Savarese, Steven Hoi.・サバレーゼ、スティーブン・ホイ。 BLIP-2 は、軽量の 12 層 Transformer をトレーニングすることで、フリーズされた事前トレーニング済み画像エンコーダーと大規模言語モデル (LLM) を活用します。
+それらの間にエンコーダーを配置し、さまざまな視覚言語タスクで最先端のパフォーマンスを実現します。最も注目すべき点は、BLIP-2 が 800 億パラメータ モデルである [Flamingo](https://arxiv.org/abs/2204.14198) を 8.7% 改善していることです。
+ゼロショット VQAv2 ではトレーニング可能なパラメーターが 54 分の 1 に減少します。
+
+論文の要約は次のとおりです。
+
+*大規模モデルのエンドツーエンドのトレーニングにより、視覚と言語の事前トレーニングのコストはますます法外なものになってきています。この論文では、市販の凍結済み事前トレーニング画像エンコーダと凍結された大規模言語モデルから視覚言語の事前トレーニングをブートストラップする、汎用的で効率的な事前トレーニング戦略である BLIP-2 を提案します。 BLIP-2 は、2 段階で事前トレーニングされた軽量の Querying Transformer でモダリティのギャップを橋渡しします。最初のステージでは、フリーズされた画像エンコーダーから学習する視覚言語表現をブートストラップします。第 2 段階では、凍結された言語モデルから視覚から言語への生成学習をブートストラップします。 BLIP-2 は、既存の方法よりもトレーニング可能なパラメーターが大幅に少ないにもかかわらず、さまざまな視覚言語タスクで最先端のパフォーマンスを実現します。たとえば、私たちのモデルは、トレーニング可能なパラメーターが 54 分の 1 少ないゼロショット VQAv2 で、Flamingo80B を 8.7% 上回っています。また、自然言語の命令に従うことができる、ゼロショット画像からテキストへの生成というモデルの新しい機能も実証します*
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/blip2_architecture.jpg"
+alt="drawing" width="600"/> 
+
+<small> BLIP-2 アーキテクチャ。 <a href="https://arxiv.org/abs/2301.12597">元の論文から抜粋。</a> </small>
+
+このモデルは、[nielsr](https://huggingface.co/nielsr) によって提供されました。
+元のコードは [ここ](https://github.com/salesforce/LAVIS/tree/5ee63d688ba4cebff63acee04adaef2dee9af207) にあります。
+
+## Usage tips
+
+- BLIP-2 は、画像とオプションのテキスト プロンプトを指定して条件付きテキストを生成するために使用できます。推論時には、 [`generate`] メソッドを使用することをお勧めします。
+- [`Blip2Processor`] を使用してモデル用の画像を準備し、予測されたトークン ID をデコードしてテキストに戻すことができます。
+
+## Resources
+
+BLIP-2 の使用を開始するのに役立つ公式 Hugging Face およびコミュニティ (🌎 で示されている) リソースのリスト。
+
+- 画像キャプション、ビジュアル質問応答 (VQA)、およびチャットのような会話のための BLIP-2 のデモ ノートブックは、[こちら](https://github.com/NielsRogge/Transformers-Tutorials/tree/master/BLIP-2) にあります。
+
+ここに含めるリソースの送信に興味がある場合は、お気軽にプル リクエストを開いてください。審査させていただきます。リソースは、既存のリソースを複製するのではなく、何か新しいものを示すことが理想的です。
+
+## Blip2Config
+
+[[autodoc]] Blip2Config
+    - from_vision_qformer_text_configs
+
+## Blip2VisionConfig
+
+[[autodoc]] Blip2VisionConfig
+
+## Blip2QFormerConfig
+
+[[autodoc]] Blip2QFormerConfig
+
+## Blip2Processor
+
+[[autodoc]] Blip2Processor
+
+## Blip2VisionModel
+
+[[autodoc]] Blip2VisionModel
+    - forward
+
+## Blip2QFormerModel
+
+[[autodoc]] Blip2QFormerModel
+    - forward
+
+## Blip2Model
+
+[[autodoc]] Blip2Model
+    - forward
+    - get_text_features
+    - get_image_features
+    - get_qformer_features
+
+## Blip2ForConditionalGeneration
+
+[[autodoc]] Blip2ForConditionalGeneration
+    - forward
+    - generate

docs/source/ja/model_doc/blip.md

@@ -0,0 +1,134 @@
+<!--Copyright 2023 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.
+
+-->
+
+# BLIP
+
+## Overview
+
+BLIP モデルは、[BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://arxiv.org/abs/2201.12086) で Junnan Li、Dongxu Li、Caiming Xiong、Steven Hoi によって提案されました。 。
+
+BLIP は、次のようなさまざまなマルチモーダル タスクを実行できるモデルです。
+- 視覚的な質問応答
+- 画像とテキストの検索（画像とテキストのマッチング）
+- 画像キャプション
+
+論文の要約は次のとおりです。
+
+*視覚言語事前トレーニング (VLP) により、多くの視覚言語タスクのパフォーマンスが向上しました。
+ただし、既存の事前トレーニング済みモデルのほとんどは、理解ベースのタスクまたは世代ベースのタスクのいずれかでのみ優れています。さらに、最適ではない監視ソースである Web から収集されたノイズの多い画像とテキストのペアを使用してデータセットをスケールアップすることで、パフォーマンスの向上が大幅に達成されました。この論文では、視覚言語の理解と生成タスクの両方に柔軟に移行する新しい VLP フレームワークである BLIP を提案します。 BLIP は、キャプションをブートストラップすることでノイズの多い Web データを効果的に利用します。キャプショナーが合成キャプションを生成し、フィルターがノイズの多いキャプションを除去します。画像テキスト検索 (平均再現率 +2.7%@1)、画像キャプション作成 (CIDEr で +2.8%)、VQA ( VQA スコアは +1.6%)。 BLIP は、ゼロショット方式でビデオ言語タスクに直接転送した場合にも、強力な一般化能力を発揮します。コード、モデル、データセットがリリースされています。*
+
+![BLIP.gif](https://cdn-uploads.huggingface.co/production/uploads/1670928184033-62441d1d9fdefb55a0b7d12c.gif)
+
+このモデルは [ybelkada](https://huggingface.co/ybelkada) によって提供されました。
+元のコードは [ここ](https://github.com/salesforce/BLIP) にあります。
+
+## Resources
+
+- [Jupyter ノートブック](https://github.com/huggingface/notebooks/blob/main/examples/image_captioning_blip.ipynb) カスタム データセットの画像キャプション用に BLIP を微調整する方法
+
+## BlipConfig
+
+[[autodoc]] BlipConfig
+    - from_text_vision_configs
+
+## BlipTextConfig
+
+[[autodoc]] BlipTextConfig
+
+## BlipVisionConfig
+
+[[autodoc]] BlipVisionConfig
+
+## BlipProcessor
+
+[[autodoc]] BlipProcessor
+
+## BlipImageProcessor
+
+[[autodoc]] BlipImageProcessor
+    - preprocess
+
+<frameworkcontent>
+<pt>
+
+## BlipModel
+
+[[autodoc]] BlipModel
+    - forward
+    - get_text_features
+    - get_image_features
+
+## BlipTextModel
+
+[[autodoc]] BlipTextModel
+    - forward
+
+## BlipVisionModel
+
+[[autodoc]] BlipVisionModel
+    - forward
+
+## BlipForConditionalGeneration
+
+[[autodoc]] BlipForConditionalGeneration
+    - forward
+
+## BlipForImageTextRetrieval
+
+[[autodoc]] BlipForImageTextRetrieval
+    - forward
+
+## BlipForQuestionAnswering
+
+[[autodoc]] BlipForQuestionAnswering
+    - forward
+
+</pt>
+<tf>
+
+## TFBlipModel
+
+[[autodoc]] TFBlipModel
+    - call
+    - get_text_features
+    - get_image_features
+
+## TFBlipTextModel
+
+[[autodoc]] TFBlipTextModel
+    - call
+
+## TFBlipVisionModel
+
+[[autodoc]] TFBlipVisionModel
+    - call
+
+## TFBlipForConditionalGeneration
+
+[[autodoc]] TFBlipForConditionalGeneration
+    - call
+
+## TFBlipForImageTextRetrieval
+
+[[autodoc]] TFBlipForImageTextRetrieval
+    - call
+
+## TFBlipForQuestionAnswering
+
+[[autodoc]] TFBlipForQuestionAnswering
+    - call
+</tf>
+</frameworkcontent>

docs/source/ja/model_doc/bloom.md

@@ -0,0 +1,107 @@
+<!--Copyright 2022 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.
+
+-->
+
+# BLOOM
+
+## Overview
+
+BLOOM モデルは、[BigScience Workshop](https://bigscience.huggingface.co/) を通じてさまざまなバージョンで提案されています。 BigScience は、研究者が時間とリソースをプールして共同でより高い効果を達成する他のオープン サイエンス イニシアチブからインスピレーションを得ています。
+BLOOM のアーキテクチャは基本的に GPT3 (次のトークン予測のための自己回帰モデル) に似ていますが、46 の異なる言語と 13 のプログラミング言語でトレーニングされています。
+モデルのいくつかの小さいバージョンが同じデータセットでトレーニングされています。 BLOOM は次のバージョンで利用できます。
+
+- [bloom-560m](https://huggingface.co/bigscience/bloom-560m)
+- [bloom-1b1](https://huggingface.co/bigscience/bloom-1b1)
+- [bloom-1b7](https://huggingface.co/bigscience/bloom-1b7)
+- [bloom-3b](https://huggingface.co/bigscience/bloom-3b)
+- [bloom-7b1](https://huggingface.co/bigscience/bloom-7b1)
+- [bloom](https://huggingface.co/bigscience/bloom) (176B parameters)
+
+## Resources
+
+BLOOM を使い始めるのに役立つ公式 Hugging Face およびコミュニティ (🌎 で示されている) リソースのリスト。ここに含めるリソースの送信に興味がある場合は、お気軽にプル リクエストを開いてください。審査させていただきます。リソースは、既存のリソースを複製するのではなく、何か新しいものを示すことが理想的です。
+
+<PipelineTag pipeline="text-generation"/>
+
+- [`BloomForCausalLM`] これによってサポートされています [causal language modeling example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling#gpt-2gpt-and-causal-language-modeling) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling.ipynb).
+
+以下も参照してください。
+- [因果言語モデリング タスク ガイド](../tasks/language_modeling)
+- [テキスト分類タスクガイド](../tasks/sequence_classification)
+- [トークン分類タスクガイド](../tasks/token_classification)
+- [質問回答タスク ガイド](../tasks/question_answering)
+
+
+⚡️ 推論
+-  に関するブログ  [最適化の話: ブルーム推論](https://huggingface.co/blog/bloom-inference-optimization)。
+- に関するブログ [DeepSpeed と Accelerate を使用した信じられないほど高速な BLOOM 推論](https://huggingface.co/blog/bloom-inference-pytorch-scripts)。
+
+⚙️トレーニング
+- に関するブログ [BLOOM トレーニングの背後にあるテクノロジー](https://huggingface.co/blog/bloom-megatron-deepspeed)。
+
+## BloomConfig
+
+[[autodoc]] BloomConfig
+    - all
+
+## BloomTokenizerFast
+
+[[autodoc]] BloomTokenizerFast
+    - all
+
+
+<frameworkcontent>
+<pt>
+
+## BloomModel
+
+[[autodoc]] BloomModel
+    - forward
+
+## BloomForCausalLM
+
+[[autodoc]] BloomForCausalLM
+    - forward
+
+## BloomForSequenceClassification
+
+[[autodoc]] BloomForSequenceClassification
+    - forward
+
+## BloomForTokenClassification
+
+[[autodoc]] BloomForTokenClassification
+    - forward
+
+## BloomForQuestionAnswering
+
+[[autodoc]] BloomForQuestionAnswering
+    - forward
+
+</pt>
+<jax>
+
+## FlaxBloomModel
+
+[[autodoc]] FlaxBloomModel
+    - __call__
+
+## FlaxBloomForCausalLM
+
+[[autodoc]] FlaxBloomForCausalLM
+    - __call__
+
+</jax>
+</frameworkcontent>

docs/source/ja/model_doc/bort.md

@@ -0,0 +1,55 @@
+<!--Copyright 2020 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+
+⚠️ 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.
+
+-->
+
+# BORT
+
+<Tip warning={true}>
+
+このモデルはメンテナンス モードのみであり、コードを変更する新しい PR は受け付けられません。
+
+このモデルの実行中に問題が発生した場合は、このモデルをサポートしていた最後のバージョン (v4.30.0) を再インストールしてください。
+これを行うには、コマンド `pip install -U Transformers==4.30.0` を実行します。
+
+</Tip>
+
+## Overview
+
+BORT モデルは、[Optimal Subarchitecture Extraction for BERT](https://arxiv.org/abs/2010.10499) で提案されました。
+Adrian de Wynter and Daniel J. Perry.これは、BERT のアーキテクチャ パラメータの最適なサブセットです。
+著者は「ボルト」と呼んでいます。
+
+論文の要約は次のとおりです。
+
+*Devlin らから BERT アーキテクチャのアーキテクチャ パラメータの最適なサブセットを抽出します。 (2018)
+ニューラル アーキテクチャ検索のアルゴリズムにおける最近の画期的な技術を適用します。この最適なサブセットを次のように呼びます。
+"Bort" は明らかに小さく、有効 (つまり、埋め込み層を考慮しない) サイズは 5.5% です。
+オリジナルの BERT 大規模アーキテクチャ、およびネット サイズの 16%。 Bort は 288 GPU 時間で事前トレーニングすることもできます。
+最高パフォーマンスの BERT パラメトリック アーキテクチャ バリアントである RoBERTa-large の事前トレーニングに必要な時間の 1.2%
+(Liu et al., 2019)、同じマシンで BERT-large をトレーニングするのに必要な GPU 時間の世界記録の約 33%
+ハードウェア。また、CPU 上で 7.9 倍高速であるだけでなく、他の圧縮バージョンよりもパフォーマンスが優れています。
+アーキテクチャ、および一部の非圧縮バリアント: 0.3% ～ 31% のパフォーマンス向上が得られます。
+BERT-large に関して、複数の公開自然言語理解 (NLU) ベンチマークにおける絶対的な評価。*
+
+このモデルは [stefan-it](https://huggingface.co/stefan-it) によって提供されました。元のコードは[ここ](https://github.com/alexa/bort/)にあります。
+
+## Usage tips
+
+- BORT のモデル アーキテクチャは BERT に基づいています。詳細については、[BERT のドキュメント ページ](bert) を参照してください。
+  モデルの API リファレンスと使用例。
+- BORT は BERT トークナイザーの代わりに RoBERTa トークナイザーを使用します。トークナイザーの API リファレンスと使用例については、[RoBERTa のドキュメント ページ](roberta) を参照してください。
+- BORT には、 [Agora](https://adewynter.github.io/notes/bort_algorithms_and_applications.html#fine-tuning-with-algebraic-topology) と呼ばれる特定の微調整アルゴリズムが必要です。
+  残念ながらまだオープンソース化されていません。誰かが実装しようとすると、コミュニティにとって非常に役立ちます。
+  BORT の微調整を機能させるためのアルゴリズム。

docs/source/ja/model_doc/bridgetower.md

@@ -0,0 +1,171 @@
+<!--Copyright 2023 The Intel Labs Team Authors, The Microsoft Research Team Authors and 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.
+
+⚠️ 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.
+
+-->
+
+# BridgeTower
+
+## Overview
+
+BridgeTower モデルは、Xiao Xu、Chenfei Wu、Shachar Rosenman、Vasudev Lal、Wanxiang Che、Nan Duan [BridgeTower: Building Bridges Between Encoders in Vision-Language Representative Learning](https://arxiv.org/abs/2206.08657) で提案されました。ドゥアン。このモデルの目標は、
+各ユニモーダル エンコーダとクロスモーダル エンコーダの間のブリッジにより、クロスモーダル エンコーダの各層での包括的かつ詳細な対話が可能になり、追加のパフォーマンスと計算コストがほとんど無視できる程度で、さまざまな下流タスクで優れたパフォーマンスを実現します。
+
+この論文は [AAAI'23](https://aaai.org/Conferences/AAAI-23/) 会議に採択されました。
+
+論文の要約は次のとおりです。
+
+*TWO-TOWER アーキテクチャを備えたビジョン言語 (VL) モデルは、近年の視覚言語表現学習の主流となっています。
+現在の VL モデルは、軽量のユニモーダル エンコーダーを使用して、ディープ クロスモーダル エンコーダーで両方のモダリティを同時に抽出、位置合わせ、融合することを学習するか、事前にトレーニングされたディープ ユニモーダル エンコーダーから最終層のユニモーダル表現を上部のクロスモーダルエンコーダー。
+どちらのアプローチも、視覚言語表現の学習を制限し、モデルのパフォーマンスを制限する可能性があります。この論文では、ユニモーダル エンコーダの最上位層とクロスモーダル エンコーダの各層の間の接続を構築する複数のブリッジ層を導入する BRIDGETOWER を提案します。
+これにより、効果的なボトムアップのクロスモーダル調整と、クロスモーダル エンコーダー内の事前トレーニング済みユニモーダル エンコーダーのさまざまなセマンティック レベルの視覚表現とテキスト表現の間の融合が可能になります。 BRIDGETOWER は 4M 画像のみで事前トレーニングされており、さまざまな下流の視覚言語タスクで最先端のパフォーマンスを実現します。
+特に、VQAv2 テスト標準セットでは、BRIDGETOWER は 78.73% の精度を達成し、同じ事前トレーニング データとほぼ無視できる追加パラメータと計算コストで以前の最先端モデル METER を 1.09% 上回りました。
+特に、モデルをさらにスケーリングすると、BRIDGETOWER は 81.15% の精度を達成し、桁違いに大きなデータセットで事前トレーニングされたモデルを上回りました。*
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/bridgetower_architecture%20.jpg"
+alt="drawing" width="600"/>
+
+<small> ブリッジタワー アーキテクチャ。 <a href="https://arxiv.org/abs/2206.08657">元の論文から抜粋。</a> </small>
+
+このモデルは、[Anahita Bhiwandiwalla](https://huggingface.co/anahita-b)、[Tiep Le](https://huggingface.co/Tile)、[Shaoyen Tseng](https://huggingface.co/shaoyent）。元のコードは [ここ](https://github.com/microsoft/BridgeTower) にあります。
+
+## Usage tips and examples
+
+BridgeTower は、ビジュアル エンコーダー、テキスト エンコーダー、および複数の軽量ブリッジ レイヤーを備えたクロスモーダル エンコーダーで構成されます。
+このアプローチの目標は、各ユニモーダル エンコーダーとクロスモーダル エンコーダーの間にブリッジを構築し、クロスモーダル エンコーダーの各層で包括的かつ詳細な対話を可能にすることでした。
+原則として、提案されたアーキテクチャでは、任意のビジュアル、テキスト、またはクロスモーダル エンコーダを適用できます。
+
+[`BridgeTowerProcessor`] は、[`RobertaTokenizer`] と [`BridgeTowerImageProcessor`] を単一のインスタンスにラップし、両方の機能を実現します。
+テキストをエンコードし、画像をそれぞれ用意します。
+
+次の例は、[`BridgeTowerProcessor`] と [`BridgeTowerForContrastiveLearning`] を使用して対照学習を実行する方法を示しています。
+
+```python
+>>> from transformers import BridgeTowerProcessor, BridgeTowerForContrastiveLearning
+>>> import requests
+>>> from PIL import Image
+
+>>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
+>>> image = Image.open(requests.get(url, stream=True).raw)
+>>> texts = ["An image of two cats chilling on a couch", "A football player scoring a goal"]
+
+>>> processor = BridgeTowerProcessor.from_pretrained("BridgeTower/bridgetower-large-itm-mlm-itc")
+>>> model = BridgeTowerForContrastiveLearning.from_pretrained("BridgeTower/bridgetower-large-itm-mlm-itc")
+
+>>> # forward pass
+>>> scores = dict()
+>>> for text in texts:
+...     # prepare inputs
+...     encoding = processor(image, text, return_tensors="pt")
+...     outputs = model(**encoding)
+...     scores[text] = outputs
+```
+
+次の例は、[`BridgeTowerProcessor`] と [`BridgeTowerForImageAndTextRetrieval`] を使用して画像テキストの取得を実行する方法を示しています。
+
+```python
+>>> from transformers import BridgeTowerProcessor, BridgeTowerForImageAndTextRetrieval
+>>> import requests
+>>> from PIL import Image
+
+>>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
+>>> image = Image.open(requests.get(url, stream=True).raw)
+>>> texts = ["An image of two cats chilling on a couch", "A football player scoring a goal"]
+
+>>> processor = BridgeTowerProcessor.from_pretrained("BridgeTower/bridgetower-base-itm-mlm")
+>>> model = BridgeTowerForImageAndTextRetrieval.from_pretrained("BridgeTower/bridgetower-base-itm-mlm")
+
+>>> # forward pass
+>>> scores = dict()
+>>> for text in texts:
+...     # prepare inputs
+...     encoding = processor(image, text, return_tensors="pt")
+...     outputs = model(**encoding)
+...     scores[text] = outputs.logits[0, 1].item()
+```
+
+次の例は、[`BridgeTowerProcessor`] と [`BridgeTowerForMaskedLM`] を使用してマスクされた言語モデリングを実行する方法を示しています。
+
+```python
+>>> from transformers import BridgeTowerProcessor, BridgeTowerForMaskedLM
+>>> from PIL import Image
+>>> import requests
+
+>>> url = "http://images.cocodataset.org/val2017/000000360943.jpg"
+>>> image = Image.open(requests.get(url, stream=True).raw).convert("RGB")
+>>> text = "a <mask> looking out of the window"
+
+>>> processor = BridgeTowerProcessor.from_pretrained("BridgeTower/bridgetower-base-itm-mlm")
+>>> model = BridgeTowerForMaskedLM.from_pretrained("BridgeTower/bridgetower-base-itm-mlm")
+
+>>> # prepare inputs
+>>> encoding = processor(image, text, return_tensors="pt")
+
+>>> # forward pass
+>>> outputs = model(**encoding)
+
+>>> results = processor.decode(outputs.logits.argmax(dim=-1).squeeze(0).tolist())
+
+>>> print(results)
+.a cat looking out of the window.
+```
+
+チップ：
+
+- BridgeTower のこの実装では、[`RobertaTokenizer`] を使用してテキスト埋め込みを生成し、OpenAI の CLIP/ViT モデルを使用して視覚的埋め込みを計算します。
+- 事前トレーニングされた [bridgeTower-base](https://huggingface.co/BridgeTower/bridgetower-base) および [bridgetower マスクされた言語モデリングと画像テキスト マッチング](https://huggingface.co/BridgeTower/bridgetower--base-itm-mlm) のチェックポイント がリリースされました。
+- 画像検索およびその他の下流タスクにおける BridgeTower のパフォーマンスについては、[表 5](https://arxiv.org/pdf/2206.08657.pdf) を参照してください。
+- このモデルの PyTorch バージョンは、torch 1.10 以降でのみ使用できます。
+
+## BridgeTowerConfig
+
+[[autodoc]] BridgeTowerConfig
+
+## BridgeTowerTextConfig
+
+[[autodoc]] BridgeTowerTextConfig
+
+## BridgeTowerVisionConfig
+
+[[autodoc]] BridgeTowerVisionConfig
+
+## BridgeTowerImageProcessor
+
+[[autodoc]] BridgeTowerImageProcessor
+    - preprocess
+
+## BridgeTowerProcessor
+
+[[autodoc]] BridgeTowerProcessor
+    - __call__
+
+## BridgeTowerModel
+
+[[autodoc]] BridgeTowerModel
+    - forward
+
+## BridgeTowerForContrastiveLearning
+
+[[autodoc]] BridgeTowerForContrastiveLearning
+    - forward
+
+## BridgeTowerForMaskedLM
+
+[[autodoc]] BridgeTowerForMaskedLM
+    - forward
+
+## BridgeTowerForImageAndTextRetrieval
+
+[[autodoc]] BridgeTowerForImageAndTextRetrieval
+    - forward
+

docs/source/ja/model_doc/bros.md

@@ -0,0 +1,113 @@
+<!--Copyright 2023 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.
+-->
+
+# BROS
+
+## Overview
+
+BROS モデルは、Teakgyu Hon、Donghyun Kim、Mingi Ji, Wonseok Hwang, Daehyun Nam, Sungrae Park によって [BROS: A Pre-trained Language Model Focusing on Text and Layout for Better Key Information Extraction from Documents](https://arxiv.org/abs/2108.04539) で提案されました。 
+
+BROS は *BERT Relying On Spatality* の略です。これは、一連のトークンとその境界ボックスを入力として受け取り、一連の隠れ状態を出力するエンコーダー専用の Transformer モデルです。 BROS は、絶対的な空間情報を使用する代わりに、相対的な空間情報をエンコードします。
+
+BERT で使用されるトークンマスク言語モデリング目標 (TMLM) と新しいエリアマスク言語モデリング目標 (AMLM) の 2 つの目標で事前トレーニングされています。
+TMLM では、トークンはランダムにマスクされ、モデルは空間情報と他のマスクされていないトークンを使用してマスクされたトークンを予測します。
+AMLM は TMLM の 2D バージョンです。テキスト トークンをランダムにマスクし、TMLM と同じ情報で予測しますが、テキスト ブロック (領域) をマスクします。
+
+`BrosForTokenClassification`には、BrosModel の上に単純な線形層があります。各トークンのラベルを予測します。
+`BrosSpadeEEForTokenClassification`には、BrosModel の上に`initial_token_classifier`と`subsequent_token_classifier`があります。 `initial_token_classifier` は各エンティティの最初のトークンを予測するために使用され、`subsequent_token_classifier` はエンティティ内の次のトークンを予測するために使用されます。 `BrosSpadeELForTokenClassification`には BrosModel の上に`entity_linker`があります。 `entity_linker` は 2 つのエンティティ間の関係を予測するために使用されます。
+
+`BrosForTokenClassification`と`BrosSpadeEEForTokenClassification`は基本的に同じジョブを実行します。ただし、`BrosForTokenClassification`は入力トークンが完全にシリアル化されていることを前提としています (トークンは 2D 空間に存在するため、これは非常に困難な作業です)。一方、`BrosSpadeEEForTokenClassification`は 1 つのトークンから次の接続トークンを予測するため、シリアル化エラーの処理をより柔軟に行うことができます。
+
+`BrosSpadeELForTokenClassification` はエンティティ内のリンク タスクを実行します。これら 2 つのエンティティが何らかの関係を共有する場合、(あるエンティティの) 1 つのトークンから (別のエンティティの) 別のトークンへの関係を予測します。
+
+BROS は、明示的な視覚機能に依存せずに、FUNSD、SROIE、CORD、SciTSR などの Key Information Extraction (KIE) ベンチマークで同等以上の結果を達成します。
+
+論文の要約は次のとおりです。
+
+*文書画像からの重要情報抽出 (KIE) には、2 次元 (2D) 空間におけるテキストの文脈的および空間的意味論を理解する必要があります。最近の研究の多くは、文書画像の視覚的特徴とテキストおよびそのレイアウトを組み合わせることに重点を置いた事前トレーニング済み言語モデルを開発することで、この課題を解決しようとしています。一方、このペーパーでは、テキストとレイアウトの効果的な組み合わせという基本に立ち返ってこの問題に取り組みます。具体的には、BROS (BERT Relying On Spatality) という名前の事前トレーニング済み言語モデルを提案します。この言語モデルは、2D 空間内のテキストの相対位置をエンコードし、エリア マスキング戦略を使用してラベルのないドキュメントから学習します。 2D 空間内のテキストを理解するためのこの最適化されたトレーニング スキームにより、BROS は、視覚的な特徴に依存することなく、4 つの KIE ベンチマーク (FUNSD、SROIE*、CORD、および SciTSR) で以前の方法と比較して同等以上のパフォーマンスを示しました。また、この論文では、KIE タスクにおける 2 つの現実世界の課題 ((1) 間違ったテキスト順序によるエラーの最小化、および (2) 少数の下流例からの効率的な学習) を明らかにし、以前の方法に対する BROS の優位性を実証します。*
+
+このモデルは [jinho8345](https://huggingface.co/jinho8345) によって寄稿されました。元のコードは [ここ](https://github.com/clovaai/bros) にあります。
+
+## Usage tips and examples
+
+- [`~transformers.BrosModel.forward`] には、`input_ids` と `bbox` (バウンディング ボックス) が必要です。各境界ボックスは、(x0、y0、x1、y1) 形式 (左上隅、右下隅) である必要があります。境界ボックスの取得は外部 OCR システムに依存します。 「x」座標はドキュメント画像の幅で正規化する必要があり、「y」座標はドキュメント画像の高さで正規化する必要があります。
+
+```python
+def expand_and_normalize_bbox(bboxes, doc_width, doc_height):
+    # here, bboxes are numpy array
+
+    # Normalize bbox -> 0 ~ 1
+    bboxes[:, [0, 2]] = bboxes[:, [0, 2]] / width
+    bboxes[:, [1, 3]] = bboxes[:, [1, 3]] / height
+```
+
+- [`~transformers.BrosForTokenClassification.forward`、`~transformers.BrosSpadeEEForTokenClassification.forward`、`~transformers.BrosSpadeEEForTokenClassification.forward`] では、損失計算に `input_ids` と `bbox` だけでなく `box_first_token_mask` も必要です。これは、各ボックスの先頭以外のトークンを除外するためのマスクです。このマスクは、単語から `input_ids` を作成するときに境界ボックスの開始トークン インデックスを保存することで取得できます。次のコードで`box_first_token_mask`を作成できます。
+
+```python
+def make_box_first_token_mask(bboxes, words, tokenizer, max_seq_length=512):
+
+    box_first_token_mask = np.zeros(max_seq_length, dtype=np.bool_)
+
+    # encode(tokenize) each word from words (List[str])
+    input_ids_list: List[List[int]] = [tokenizer.encode(e, add_special_tokens=False) for e in words]
+
+    # get the length of each box
+    tokens_length_list: List[int] = [len(l) for l in input_ids_list]
+
+    box_end_token_indices = np.array(list(itertools.accumulate(tokens_length_list)))
+    box_start_token_indices = box_end_token_indices - np.array(tokens_length_list)
+
+    # filter out the indices that are out of max_seq_length
+    box_end_token_indices = box_end_token_indices[box_end_token_indices < max_seq_length - 1]
+    if len(box_start_token_indices) > len(box_end_token_indices):
+        box_start_token_indices = box_start_token_indices[: len(box_end_token_indices)]
+
+    # set box_start_token_indices to True
+    box_first_token_mask[box_start_token_indices] = True
+
+    return box_first_token_mask
+
+```
+
+## Resources
+
+- デモ スクリプトは [こちら](https://github.com/clovaai/bros) にあります。
+
+## BrosConfig
+
+[[autodoc]] BrosConfig
+
+## BrosProcessor
+
+[[autodoc]] BrosProcessor
+    - __call__
+
+## BrosModel
+
+[[autodoc]] BrosModel
+    - forward
+
+
+## BrosForTokenClassification
+
+[[autodoc]] BrosForTokenClassification
+    - forward
+
+## BrosSpadeEEForTokenClassification
+
+[[autodoc]] BrosSpadeEEForTokenClassification
+    - forward
+
+## BrosSpadeELForTokenClassification
+
+[[autodoc]] BrosSpadeELForTokenClassification
+    - forward

docs/source/ja/model_doc/byt5.md

@@ -0,0 +1,154 @@
+<!--Copyright 2021 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.
+
+-->
+
+# ByT5
+
+## Overview
+
+ByT5 モデルは、[ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) by Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir
+Kale, Adam Roberts, Colin Raffel.
+
+論文の要約は次のとおりです。
+
+*最も広く使用されている事前トレーニング済み言語モデルは、単語またはサブワード単位に対応するトークンのシーケンスで動作します。
+テキストをトークンのシーケンスとしてエンコードするには、トークナイザーが必要です。トークナイザーは通常、
+モデル。代わりに生のテキスト (バイトまたは文字) を直接操作するトークンフリー モデルには多くの利点があります。
+すぐに使用できるあらゆる言語のテキストを処理でき、ノイズに対してより堅牢であり、技術的負債を最小限に抑えます。
+複雑でエラーが発生しやすいテキスト前処理パイプラインを削除します。バイトまたは文字列がトークンより長いため
+トークンフリー モデルに関する過去の研究では、シーケンスのコストを償却するように設計された新しいモデル アーキテクチャが導入されることがよくありました。
+生のテキストを直接操作します。この論文では、標準的な Transformer アーキテクチャが次のようなもので使用できることを示します。
+バイトシーケンスを処理するための最小限の変更。パラメータ数の観点からトレードオフを注意深く特徴付けます。
+FLOP のトレーニングと推論速度を調べ、バイトレベルのモデルがトークンレベルと競合できることを示します。
+対応者。また、バイトレベルのモデルはノイズに対して大幅に堅牢であり、より優れたパフォーマンスを発揮することも示しています。
+スペルと発音に敏感なタスク。私たちの貢献の一環として、新しいセットをリリースします。
+T5 アーキテクチャに基づいた事前トレーニング済みのバイトレベルの Transformer モデルと、そこで使用されるすべてのコードとデータ
+実験。*
+
+このモデルは、[patrickvonplaten](https://huggingface.co/patrickvonplaten) によって提供されました。元のコードは次のとおりです
+[ここ](https://github.com/google-research/byt5) にあります。
+
+<Tip>
+
+ByT5 のアーキテクチャは T5v1.1 モデルに基づいています。API リファレンスについては、[T5v1.1 のドキュメント ページ](t5v1.1) を参照してください。彼らは
+モデルの入力を準備する方法が異なるだけです。以下のコード例を参照してください。
+
+</Tip>
+
+ByT5 は教師なしで事前トレーニングされているため、単一タスク中にタスク プレフィックスを使用する利点はありません。
+微調整。マルチタスクの微調整を行う場合は、プレフィックスを使用する必要があります。
+
+## Usage Examples
+
+ByT5 は生の UTF-8 バイトで動作するため、トークナイザーなしで使用できます。
+
+```python
+>>> from transformers import T5ForConditionalGeneration
+>>> import torch
+
+>>> model = T5ForConditionalGeneration.from_pretrained("google/byt5-small")
+
+>>> num_special_tokens = 3
+>>> # Model has 3 special tokens which take up the input ids 0,1,2 of ByT5.
+>>> # => Need to shift utf-8 character encodings by 3 before passing ids to model.
+
+>>> input_ids = torch.tensor([list("Life is like a box of chocolates.".encode("utf-8"))]) + num_special_tokens
+
+>>> labels = torch.tensor([list("La vie est comme une boîte de chocolat.".encode("utf-8"))]) + num_special_tokens
+
+>>> loss = model(input_ids, labels=labels).loss
+>>> loss.item()
+2.66
+```
+
+ただし、バッチ推論とトレーニングの場合は、トークナイザーを使用することをお勧めします。
+
+
+```python
+>>> from transformers import T5ForConditionalGeneration, AutoTokenizer
+
+>>> model = T5ForConditionalGeneration.from_pretrained("google/byt5-small")
+>>> tokenizer = AutoTokenizer.from_pretrained("google/byt5-small")
+
+>>> model_inputs = tokenizer(
+...     ["Life is like a box of chocolates.", "Today is Monday."], padding="longest", return_tensors="pt"
+... )
+>>> labels_dict = tokenizer(
+...     ["La vie est comme une boîte de chocolat.", "Aujourd'hui c'est lundi."], padding="longest", return_tensors="pt"
+... )
+>>> labels = labels_dict.input_ids
+
+>>> loss = model(**model_inputs, labels=labels).loss
+>>> loss.item()
+17.9
+```
+
+[T5](t5) と同様に、ByT5 はスパンマスクノイズ除去タスクでトレーニングされました。しかし、
+モデルはキャラクターに直接作用するため、事前トレーニングタスクは少し複雑です
+違う。のいくつかの文字を破損してみましょう
+`"The dog chases a ball in the park."`という文を入力し、ByT5 に予測してもらいます。
+わたしたちのため。
+
+```python
+>>> from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
+>>> import torch
+
+>>> tokenizer = AutoTokenizer.from_pretrained("google/byt5-base")
+>>> model = AutoModelForSeq2SeqLM.from_pretrained("google/byt5-base")
+
+>>> input_ids_prompt = "The dog chases a ball in the park."
+>>> input_ids = tokenizer(input_ids_prompt).input_ids
+
+>>> # Note that we cannot add "{extra_id_...}" to the string directly
+>>> # as the Byte tokenizer would incorrectly merge the tokens
+>>> # For ByT5, we need to work directly on the character level
+>>> # Contrary to T5, ByT5 does not use sentinel tokens for masking, but instead
+>>> # uses final utf character ids.
+>>> # UTF-8 is represented by 8 bits and ByT5 has 3 special tokens.
+>>> # => There are 2**8+2 = 259 input ids and mask tokens count down from index 258.
+>>> # => mask to "The dog [258]a ball [257]park."
+
+>>> input_ids = torch.tensor([input_ids[:8] + [258] + input_ids[14:21] + [257] + input_ids[28:]])
+>>> input_ids
+tensor([[ 87, 107, 104,  35, 103, 114, 106,  35, 258,  35, 100,  35, 101, 100, 111, 111, 257,  35, 115, 100, 117, 110,  49,   1]])
+
+>>> # ByT5 produces only one char at a time so we need to produce many more output characters here -> set `max_length=100`.
+>>> output_ids = model.generate(input_ids, max_length=100)[0].tolist()
+>>> output_ids
+[0, 258, 108, 118,  35, 119, 107, 104,  35, 114, 113, 104,  35, 122, 107, 114,  35, 103, 114, 104, 118, 257,  35, 108, 113,  35, 119, 107, 104,  35, 103, 108, 118, 102, 114, 256, 108, 113,  35, 119, 107, 104, 35, 115, 100, 117, 110,  49,  35,  87, 107, 104,  35, 103, 114, 106, 35, 108, 118,  35, 119, 107, 104,  35, 114, 113, 104,  35, 122, 107, 114,  35, 103, 114, 104, 118,  35, 100,  35, 101, 100, 111, 111,  35, 108, 113, 255,  35, 108, 113,  35, 119, 107, 104,  35, 115, 100, 117, 110,  49]
+
+>>> # ^- Note how 258 descends to 257, 256, 255
+
+>>> # Now we need to split on the sentinel tokens, let's write a short loop for this
+>>> output_ids_list = []
+>>> start_token = 0
+>>> sentinel_token = 258
+>>> while sentinel_token in output_ids:
+...     split_idx = output_ids.index(sentinel_token)
+...     output_ids_list.append(output_ids[start_token:split_idx])
+...     start_token = split_idx
+...     sentinel_token -= 1
+
+>>> output_ids_list.append(output_ids[start_token:])
+>>> output_string = tokenizer.batch_decode(output_ids_list)
+>>> output_string
+['<pad>', 'is the one who does', ' in the disco', 'in the park. The dog is the one who does a ball in', ' in the park.']
+```
+
+## ByT5Tokenizer
+
+[[autodoc]] ByT5Tokenizer
+
+詳細については、[`ByT5Tokenizer`] を参照してください。

docs/source/ja/model_doc/camembert.md

@@ -0,0 +1,135 @@
+<!--Copyright 2020 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+
+⚠️ 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.
+
+-->
+
+# CamemBERT
+
+## Overview
+
+CamemBERT モデルは、[CamemBERT: a Tasty French Language Model](https://arxiv.org/abs/1911.03894) で提案されました。
+Louis Martin, Benjamin Muller, Pedro Javier Ortiz Suárez, Yoann Dupont, Laurent Romary, Éric Villemonte de la
+Clergerie, Djamé Seddah, and Benoît Sagot. 2019年にリリースされたFacebookのRoBERTaモデルをベースにしたモデルです。
+138GBのフランス語テキストでトレーニングされました。
+
+論文の要約は次のとおりです。
+
+*事前トレーニングされた言語モデルは現在、自然言語処理で広く普及しています。成功にもかかわらず、利用可能なほとんどの
+モデルは英語のデータ、または複数言語のデータの連結でトレーニングされています。これにより、
+このようなモデルの実際の使用は、英語を除くすべての言語で非常に限られています。フランス人にとってこの問題に対処することを目指して、
+Bi-direction Encoders for Transformers (BERT) のフランス語版である CamemBERT をリリースします。測定します
+複数の下流タスク、つまり品詞タグ付けにおける多言語モデルと比較した CamemBERT のパフォーマンス
+依存関係解析、固有表現認識、自然言語推論。 CamemBERT は最先端技術を向上させます
+検討されているほとんどのタスクに対応します。私たちは、研究と
+フランス語 NLP の下流アプリケーション。*
+
+このモデルは [camembert](https://huggingface.co/camembert) によって提供されました。元のコードは [ここ](https://camembert-model.fr/) にあります。
+
+
+<Tip>
+
+この実装はRoBERTaと同じです。使用例については[RoBERTaのドキュメント](roberta)も参照してください。
+入力と出力に関する情報として。
+
+</Tip>
+
+## Resources
+
+- [テキスト分類タスクガイド](../tasks/sequence_classification)
+- [トークン分類タスクガイド](../tasks/token_classification)
+- [質問回答タスク ガイド](../tasks/question_answering)
+- [因果言語モデリング タスク ガイド](../tasks/language_modeling)
+- [マスク言語モデリング タスク ガイド](../tasks/masked_language_modeling)
+- [多肢選択タスク ガイド](../tasks/multiple_choice)
+
+## CamembertConfig
+
+[[autodoc]] CamembertConfig
+
+## CamembertTokenizer
+
+[[autodoc]] CamembertTokenizer
+    - build_inputs_with_special_tokens
+    - get_special_tokens_mask
+    - create_token_type_ids_from_sequences
+    - save_vocabulary
+
+## CamembertTokenizerFast
+
+[[autodoc]] CamembertTokenizerFast
+
+<frameworkcontent>
+<pt>
+
+## CamembertModel
+
+[[autodoc]] CamembertModel
+
+## CamembertForCausalLM
+
+[[autodoc]] CamembertForCausalLM
+
+## CamembertForMaskedLM
+
+[[autodoc]] CamembertForMaskedLM
+
+## CamembertForSequenceClassification
+
+[[autodoc]] CamembertForSequenceClassification
+
+## CamembertForMultipleChoice
+
+[[autodoc]] CamembertForMultipleChoice
+
+## CamembertForTokenClassification
+
+[[autodoc]] CamembertForTokenClassification
+
+## CamembertForQuestionAnswering
+
+[[autodoc]] CamembertForQuestionAnswering
+
+</pt>
+<tf>
+
+## TFCamembertModel
+
+[[autodoc]] TFCamembertModel
+
+## TFCamembertForCasualLM
+
+[[autodoc]] TFCamembertForCausalLM
+
+## TFCamembertForMaskedLM
+
+[[autodoc]] TFCamembertForMaskedLM
+
+## TFCamembertForSequenceClassification
+
+[[autodoc]] TFCamembertForSequenceClassification
+
+## TFCamembertForMultipleChoice
+
+[[autodoc]] TFCamembertForMultipleChoice
+
+## TFCamembertForTokenClassification
+
+[[autodoc]] TFCamembertForTokenClassification
+
+## TFCamembertForQuestionAnswering
+
+[[autodoc]] TFCamembertForQuestionAnswering
+
+</tf>
+</frameworkcontent>

docs/source/ja/model_doc/canine.md

@@ -0,0 +1,144 @@
+<!--Copyright 2021 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.
+
+-->
+
+# CANINE
+
+## Overview
+
+CANINE モデルは、[CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language
+Representation](https://arxiv.org/abs/2103.06874)、Jonathan H. Clark、Dan Garrette、Iulia Turc、John Wieting 著。その
+明示的なトークン化ステップ (バイト ペアなど) を使用せずに Transformer をトレーニングする最初の論文の 1 つ
+エンコーディング (BPE、WordPiece または SentencePiece)。代わりに、モデルは Unicode 文字レベルで直接トレーニングされます。
+キャラクターレベルでのトレーニングでは必然的にシーケンスの長さが長くなりますが、CANINE はこれを効率的な方法で解決します。
+ディープ Transformer エンコーダを適用する前に、ダウンサンプリング戦略を実行します。
+
+論文の要約は次のとおりです。
+
+*パイプライン NLP システムは、エンドツーエンドのニューラル モデリングに大部分が取って代わられていますが、一般的に使用されているほぼすべてのモデルは
+依然として明示的なトークン化手順が必要です。最近のトークン化アプローチはデータ由来のサブワードに基づいていますが、
+レキシコンは手動で作成されたトークナイザーよりも脆弱ではありませんが、これらの技術はすべての言語に等しく適しているわけではありません。
+言語や固定語彙の使用により、モデルの適応能力が制限される可能性があります。この論文では、CANINE を紹介します。
+明示的なトークン化や語彙を使用せずに、文字シーケンスを直接操作するニューラル エンコーダーと、
+文字に直接作用するか、オプションでサブワードをソフト誘導バイアスとして使用する事前トレーニング戦略。
+よりきめの細かい入力を効果的かつ効率的に使用するために、CANINE はダウンサンプリングを組み合わせて、入力を削減します。
+コンテキストをエンコードするディープトランスフォーマースタックを備えたシーケンスの長さ。 CANINE は、同等の mBERT モデルよりも次の点で優れています。
+TyDi QA の 2.8 F1 は、モデル パラメータが 28% 少ないにもかかわらず、困難な多言語ベンチマークです。*
+
+このモデルは、[nielsr](https://huggingface.co/nielsr) によって提供されました。元のコードは [ここ](https://github.com/google-research/language/tree/master/language/canine) にあります。
+
+## Usage tips
+
+- CANINE は内部で少なくとも 3 つの Transformer エンコーダーを使用します: 2 つの「浅い」エンコーダー (単一のエンコーダーのみで構成)
+  レイヤー) と 1 つの「ディープ」エンコーダー (通常の BERT エンコーダー)。まず、「浅い」エンコーダを使用してコンテキストを設定します。
+  ローカル アテンションを使用した文字の埋め込み。次に、ダウンサンプリングの後、「ディープ」エンコーダーが適用されます。ついに、
+  アップサンプリング後、「浅い」エンコーダを使用して最終的な文字埋め込みが作成されます。アップと
+  ダウンサンプリングについては論文に記載されています。
+- CANINE は、デフォルトで 2048 文字の最大シーケンス長を使用します。 [`CanineTokenizer`] を使用できます
+  モデル用のテキストを準備します。
+- 特別な [CLS] トークンの最終的な非表示状態の上に線形レイヤーを配置することで分類を行うことができます。
+  (事前定義された Unicode コード ポイントがあります)。ただし、トークン分類タスクの場合は、ダウンサンプリングされたシーケンス
+  トークンは、元の文字シーケンスの長さ (2048) と一致するように再度アップサンプリングする必要があります。の
+  詳細については、論文を参照してください。
+
+モデルのチェックポイント:
+
+  - [google/canine-c](https://huggingface.co/google/canine-c): 自己回帰文字損失で事前トレーニング済み、
+    12 レイヤー、768 隠し、12 ヘッド、121M パラメーター (サイズ ~500 MB)。
+  - [google/canine-s](https://huggingface.co/google/canine-s): サブワード損失で事前トレーニング済み、12 層、
+    768 個の非表示、12 ヘッド、121M パラメーター (サイズ ~500 MB)。
+
+## Usage example
+
+CANINE は生の文字で動作するため、**トークナイザーなし**で使用できます。
+
+```python
+>>> from transformers import CanineModel
+>>> import torch
+
+>>> model = CanineModel.from_pretrained("google/canine-c")  # model pre-trained with autoregressive character loss
+
+>>> text = "hello world"
+>>> # use Python's built-in ord() function to turn each character into its unicode code point id
+>>> input_ids = torch.tensor([[ord(char) for char in text]])
+
+>>> outputs = model(input_ids)  # forward pass
+>>> pooled_output = outputs.pooler_output
+>>> sequence_output = outputs.last_hidden_state
+```
+
+ただし、バッチ推論とトレーニングの場合は、トークナイザーを使用することをお勧めします（すべてをパディング/切り詰めるため）
+シーケンスを同じ長さにします):
+
+```python
+>>> from transformers import CanineTokenizer, CanineModel
+
+>>> model = CanineModel.from_pretrained("google/canine-c")
+>>> tokenizer = CanineTokenizer.from_pretrained("google/canine-c")
+
+>>> inputs = ["Life is like a box of chocolates.", "You never know what you gonna get."]
+>>> encoding = tokenizer(inputs, padding="longest", truncation=True, return_tensors="pt")
+
+>>> outputs = model(**encoding)  # forward pass
+>>> pooled_output = outputs.pooler_output
+>>> sequence_output = outputs.last_hidden_state
+```
+
+## Resources
+
+- [テキスト分類タスクガイド](../tasks/sequence_classification)
+- [トークン分類タスクガイド](../tasks/token_classification)
+- [質問回答タスク ガイド](../tasks/question_answering)
+- [多肢選択タスク ガイド](../tasks/multiple_choice)
+
+## CanineConfig
+
+[[autodoc]] CanineConfig
+
+## CanineTokenizer
+
+[[autodoc]] CanineTokenizer
+    - build_inputs_with_special_tokens
+    - get_special_tokens_mask
+    - create_token_type_ids_from_sequences
+
+## CANINE specific outputs
+
+[[autodoc]] models.canine.modeling_canine.CanineModelOutputWithPooling
+
+## CanineModel
+
+[[autodoc]] CanineModel
+    - forward
+
+## CanineForSequenceClassification
+
+[[autodoc]] CanineForSequenceClassification
+    - forward
+
+## CanineForMultipleChoice
+
+[[autodoc]] CanineForMultipleChoice
+    - forward
+
+## CanineForTokenClassification
+
+[[autodoc]] CanineForTokenClassification
+    - forward
+
+## CanineForQuestionAnswering
+
+[[autodoc]] CanineForQuestionAnswering
+    - forward

docs/source/ja/model_doc/chinese_clip.md

@@ -0,0 +1,112 @@
+<!--Copyright 2022 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.
+
+-->
+
+# Chinese-CLIP
+
+## Overview
+
+Chinese-CLIP An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) で提案されました。周、張周。
+Chinese-CLIP は、中国語の画像とテキストのペアの大規模なデータセットに対する CLIP (Radford et al., 2021) の実装です。クロスモーダル検索を実行できるほか、ゼロショット画像分類、オープンドメインオブジェクト検出などのビジョンタスクのビジョンバックボーンとしても機能します。オリジナルの中国語-CLIPコードは[このリンクで](https://github.com/OFA-Sys/Chinese-CLIP)。
+
+論文の要約は次のとおりです。
+
+*CLIP の大成功 (Radford et al., 2021) により、視覚言語の事前訓練のための対照学習の研究と応用が促進されました。この研究では、ほとんどのデータが公開されているデータセットから取得された中国語の画像とテキストのペアの大規模なデータセットを構築し、新しいデータセットで中国語の CLIP モデルを事前トレーニングします。当社では、7,700 万から 9 億 5,800 万のパラメータにわたる、複数のサイズの 5 つの中国 CLIP モデルを開発しています。さらに、モデルのパフォーマンスを向上させるために、最初に画像エンコーダーをフリーズさせてモデルをトレーニングし、次にすべてのパラメーターを最適化してトレーニングする 2 段階の事前トレーニング方法を提案します。私たちの包括的な実験では、中国の CLIP がゼロショット学習と微調整のセットアップで MUGE、Flickr30K-CN、および COCO-CN 上で最先端のパフォーマンスを達成でき、ゼロで競争力のあるパフォーマンスを達成できることを実証しています。 - ELEVATER ベンチマークでの評価に基づくショット画像の分類 (Li et al., 2022)。コード、事前トレーニング済みモデル、デモがリリースされました。*
+
+Chinese-CLIP モデルは、[OFA-Sys](https://huggingface.co/OFA-Sys) によって提供されました。
+
+## Usage example
+
+以下のコード スニペットは、画像とテキストの特徴と類似性を計算する方法を示しています。
+
+```python
+>>> from PIL import Image
+>>> import requests
+>>> from transformers import ChineseCLIPProcessor, ChineseCLIPModel
+
+>>> model = ChineseCLIPModel.from_pretrained("OFA-Sys/chinese-clip-vit-base-patch16")
+>>> processor = ChineseCLIPProcessor.from_pretrained("OFA-Sys/chinese-clip-vit-base-patch16")
+
+>>> url = "https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/pokemon.jpeg"
+>>> image = Image.open(requests.get(url, stream=True).raw)
+>>> # Squirtle, Bulbasaur, Charmander, Pikachu in English
+>>> texts = ["杰尼龟", "妙蛙种子", "小火龙", "皮卡丘"]
+
+>>> # compute image feature
+>>> inputs = processor(images=image, return_tensors="pt")
+>>> image_features = model.get_image_features(**inputs)
+>>> image_features = image_features / image_features.norm(p=2, dim=-1, keepdim=True)  # normalize
+
+>>> # compute text features
+>>> inputs = processor(text=texts, padding=True, return_tensors="pt")
+>>> text_features = model.get_text_features(**inputs)
+>>> text_features = text_features / text_features.norm(p=2, dim=-1, keepdim=True)  # normalize
+
+>>> # compute image-text similarity scores
+>>> inputs = processor(text=texts, images=image, return_tensors="pt", padding=True)
+>>> outputs = model(**inputs)
+>>> logits_per_image = outputs.logits_per_image  # this is the image-text similarity score
+>>> probs = logits_per_image.softmax(dim=1)  # probs: [[1.2686e-03, 5.4499e-02, 6.7968e-04, 9.4355e-01]]
+```
+
+現在、次のスケールの事前トレーニング済み Chinese-CLIP モデルが 🤗 Hub で利用可能です。
+
+- [OFA-Sys/chinese-clip-vit-base-patch16](https://huggingface.co/OFA-Sys/chinese-clip-vit-base-patch16)
+- [OFA-Sys/chinese-clip-vit-large-patch14](https://huggingface.co/OFA-Sys/chinese-clip-vit-large-patch14)
+- [OFA-Sys/chinese-clip-vit-large-patch14-336px](https://huggingface.co/OFA-Sys/chinese-clip-vit-large-patch14-336px)
+- [OFA-Sys/chinese-clip-vit-huge-patch14](https://huggingface.co/OFA-Sys/chinese-clip-vit-huge-patch14)
+
+## ChineseCLIPConfig
+
+[[autodoc]] ChineseCLIPConfig
+    - from_text_vision_configs
+
+## ChineseCLIPTextConfig
+
+[[autodoc]] ChineseCLIPTextConfig
+
+## ChineseCLIPVisionConfig
+
+[[autodoc]] ChineseCLIPVisionConfig
+
+## ChineseCLIPImageProcessor
+
+[[autodoc]] ChineseCLIPImageProcessor
+    - preprocess
+
+## ChineseCLIPFeatureExtractor
+
+[[autodoc]] ChineseCLIPFeatureExtractor
+
+## ChineseCLIPProcessor
+
+[[autodoc]] ChineseCLIPProcessor
+
+## ChineseCLIPModel
+
+[[autodoc]] ChineseCLIPModel
+    - forward
+    - get_text_features
+    - get_image_features
+
+## ChineseCLIPTextModel
+
+[[autodoc]] ChineseCLIPTextModel
+    - forward
+
+## ChineseCLIPVisionModel
+
+[[autodoc]] ChineseCLIPVisionModel
+    - forward

docs/source/ja/model_doc/clap.md

@@ -0,0 +1,80 @@
+<!--Copyright 2023 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.
+
+-->
+
+# CLAP
+
+## Overview
+
+CLAP モデルは、[Large Scale Contrastive Language-Audio pretraining with
+feature fusion and keyword-to-caption augmentation](https://arxiv.org/pdf/2211.06687.pdf)、Yusong Wu、Ke Chen、Tianyu Zhang、Yuchen Hui、Taylor Berg-Kirkpatrick、Shlomo Dubnov 著。
+
+CLAP (Contrastive Language-Audio Pretraining) は、さまざまな (音声、テキスト) ペアでトレーニングされたニューラル ネットワークです。タスクに合わせて直接最適化することなく、音声が与えられた場合に最も関連性の高いテキスト スニペットを予測するように指示できます。 CLAP モデルは、SWINTransformer を使用して log-Mel スペクトログラム入力からオーディオ特徴を取得し、RoBERTa モデルを使用してテキスト特徴を取得します。次に、テキストとオーディオの両方の特徴が、同じ次元の潜在空間に投影されます。投影されたオーディオとテキストの特徴の間のドット積が、同様のスコアとして使用されます。
+
+論文の要約は次のとおりです。
+
+*対照学習は、マルチモーダル表現学習の分野で目覚ましい成功を収めています。この論文では、音声データと自然言語記述を組み合わせて音声表現を開発する、対照的な言語音声事前トレーニングのパイプラインを提案します。この目標を達成するために、私たちはまず、さまざまなデータ ソースからの 633,526 個の音声とテキストのペアの大規模なコレクションである LAION-Audio-630K をリリースします。次に、さまざまなオーディオ エンコーダとテキスト エンコーダを考慮して、対照的な言語とオーディオの事前トレーニング モデルを構築します。機能融合メカニズムとキーワードからキャプションへの拡張をモデル設計に組み込んで、モデルが可変長の音声入力を処理できるようにし、パフォーマンスを向上させます。 3 番目に、包括的な実験を実行して、テキストから音声への取得、ゼロショット音声分類、教師付き音声分類の 3 つのタスクにわたってモデルを評価します。結果は、私たちのモデルがテキストから音声への検索タスクにおいて優れたパフォーマンスを達成していることを示しています。オーディオ分類タスクでは、モデルはゼロショット設定で最先端のパフォーマンスを達成し、非ゼロショット設定でもモデルの結果に匹敵するパフォーマンスを得ることができます。 LAION-オーディオ-6*
+
+このモデルは、[Younes Belkada](https://huggingface.co/ybelkada) および [Arthur Zucker](https://huggingface.co/ArthurZ) によって提供されました。
+元のコードは [こちら](https://github.com/LAION-AI/Clap) にあります。
+
+## ClapConfig
+
+[[autodoc]] ClapConfig
+    - from_text_audio_configs
+
+## ClapTextConfig
+
+[[autodoc]] ClapTextConfig
+
+## ClapAudioConfig
+
+[[autodoc]] ClapAudioConfig
+
+## ClapFeatureExtractor
+
+[[autodoc]] ClapFeatureExtractor
+
+## ClapProcessor
+
+[[autodoc]] ClapProcessor
+
+## ClapModel
+
+[[autodoc]] ClapModel
+    - forward
+    - get_text_features
+    - get_audio_features
+
+## ClapTextModel
+
+[[autodoc]] ClapTextModel
+    - forward
+
+## ClapTextModelWithProjection
+
+[[autodoc]] ClapTextModelWithProjection
+    - forward
+
+## ClapAudioModel
+
+[[autodoc]] ClapAudioModel
+    - forward
+
+## ClapAudioModelWithProjection
+
+[[autodoc]] ClapAudioModelWithProjection
+    - forward
+    

docs/source/ja/model_doc/clip.md

@@ -0,0 +1,220 @@
+<!--Copyright 2021 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.
+
+-->
+
+# CLIP
+
+## Overview
+
+CLIP モデルは、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 [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) で提案されました。
+サンディニ・アガルワル、ギリッシュ・サストリー、アマンダ・アスケル、パメラ・ミシュキン、ジャック・クラーク、グレッチェン・クルーガー、イリヤ・サツケヴァー。クリップ
+(Contrastive Language-Image Pre-Training) は、さまざまな (画像、テキスト) ペアでトレーニングされたニューラル ネットワークです。かもね
+直接最適化することなく、与えられた画像から最も関連性の高いテキスト スニペットを予測するように自然言語で指示されます。
+GPT-2 および 3 のゼロショット機能と同様に、タスクに対して。
+
+論文の要約は次のとおりです。
+
+*最先端のコンピューター ビジョン システムは、あらかじめ定められたオブジェクト カテゴリの固定セットを予測するようにトレーニングされています。これ
+制限された形式の監視では、指定するために追加のラベル付きデータが必要となるため、一般性と使いやすさが制限されます。
+その他の視覚的なコンセプト。画像に関する生のテキストから直接学習することは、
+より広範な監督源。どのキャプションが表示されるかを予測するという単純な事前トレーニング タスクが有効であることを示します。
+400 のデータセットで SOTA 画像表現を最初から学習するための効率的かつスケーラブルな方法はどの画像ですか
+インターネットから収集された数百万の（画像、テキスト）ペア。事前トレーニング後、自然言語を使用して参照します。
+視覚的な概念を学習し（または新しい概念を説明し）、下流のタスクへのモデルのゼロショット転送を可能にします。私たちは勉強します
+30 を超えるさまざまな既存のコンピューター ビジョン データセットでタスクをまたがってベンチマークを行うことにより、このアプローチのパフォーマンスを評価します。
+OCR、ビデオ内のアクション認識、地理的位置特定、およびさまざまな種類のきめ細かいオブジェクト分類など。の
+モデルはほとんどのタスクに簡単に移行でき、多くの場合、必要がなくても完全に監視されたベースラインと競合します。
+データセット固有のトレーニングに適しています。たとえば、ImageNet ゼロショットではオリジナルの ResNet-50 の精度と一致します。
+トレーニングに使用された 128 万のトレーニング サンプルを使用する必要はありません。コードをリリースし、事前トレーニング済み
+モデルの重みはこの https URL で確認できます。*
+
+このモデルは [valhalla](https://huggingface.co/valhalla) によって提供されました。元のコードは [ここ](https://github.com/openai/CLIP) にあります。
+
+## Usage tips and example
+
+CLIP は、マルチモーダルなビジョンおよび言語モデルです。画像とテキストの類似性やゼロショット画像に使用できます。
+分類。 CLIP は、ViT のようなトランスフォーマーを使用して視覚的特徴を取得し、因果言語モデルを使用してテキストを取得します
+特徴。次に、テキストと視覚の両方の特徴が、同じ次元の潜在空間に投影されます。ドット
+投影された画像とテキストの特徴間の積が同様のスコアとして使用されます。
+
+画像を Transformer エンコーダに供給するために、各画像は固定サイズの重複しないパッチのシーケンスに分割されます。
+これらは線形に埋め込まれます。 [CLS] トークンは、イメージ全体の表現として機能するために追加されます。作家たち
+また、絶対位置埋め込みを追加し、結果として得られるベクトルのシーケンスを標準の Transformer エンコーダに供給します。
+[`CLIPImageProcessor`] を使用して、モデルの画像のサイズ変更 (または再スケール) および正規化を行うことができます。
+
+[`CLIPTokenizer`] はテキストのエンコードに使用されます。 [`CLIPProcessor`] はラップします
+[`CLIPImageProcessor`] と [`CLIPTokenizer`] を両方の単一インスタンスに統合
+テキストをエンコードして画像を準備します。次の例は、次のメソッドを使用して画像とテキストの類似性スコアを取得する方法を示しています。
+[`CLIPProcessor`] と [`CLIPModel`]。
+
+```python
+>>> from PIL import Image
+>>> import requests
+
+>>> from transformers import CLIPProcessor, CLIPModel
+
+>>> model = CLIPModel.from_pretrained("openai/clip-vit-base-patch32")
+>>> processor = CLIPProcessor.from_pretrained("openai/clip-vit-base-patch32")
+
+>>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
+>>> image = Image.open(requests.get(url, stream=True).raw)
+
+>>> inputs = processor(text=["a photo of a cat", "a photo of a dog"], images=image, return_tensors="pt", padding=True)
+
+>>> outputs = model(**inputs)
+>>> logits_per_image = outputs.logits_per_image  # this is the image-text similarity score
+>>> probs = logits_per_image.softmax(dim=1)  # we can take the softmax to get the label probabilities
+```
+
+## Resources
+
+CLIP を使い始めるのに役立つ公式 Hugging Face およびコミュニティ (🌎 で示されている) リソースのリスト。
+
+- [リモート センシング (衛星) 画像とキャプションを使用した CLIP の微調整](https://huggingface.co/blog/fine-tune-clip-rsicd)、[RSICD データセット] を使用して CLIP を微調整する方法に関するブログ投稿(https://github.com/201528014227051/RSICD_optimal) と、データ拡張によるパフォーマンスの変化の比較。
+- この [サンプル スクリプト](https://github.com/huggingface/transformers/tree/main/examples/pytorch/contrastive-image-text) は、プレ- [COCO データセット](https://cocodataset.org/#home) を使用してトレーニングされたビジョンおよびテキスト エンコーダー。
+
+<PipelineTag pipeline="image-to-text"/>
+
+- 画像キャプションのビーム検索による推論に事前トレーニング済み CLIP を使用する方法に関する [ノートブック](https://colab.research.google.com/drive/1tuoAC5F4sC7qid56Z0ap-stR3rwdk0ZV?usp=sharing)。 🌎
+
+**画像検索**
+
+- 事前トレーニングされた CLIP を使用した画像検索と MRR (平均相互ランク) スコアの計算に関する [ノートブック](https://colab.research.google.com/drive/1bLVwVKpAndpEDHqjzxVPr_9nGrSbuOQd?usp=sharing)。 🌎
+- 画像の取得と類似性スコアの表示に関する [ノートブック](https://colab.research.google.com/github/deep-diver/image_search_with_natural_language/blob/main/notebooks/Image_Search_CLIP.ipynb)。 🌎
+- 多言語 CLIP を使用して画像とテキストを同じベクトル空間にマッピングする方法に関する [ノートブック](https://colab.research.google.com/drive/1xO-wC_m_GNzgjIBQ4a4znvQkvDoZJvH4?usp=sharing)。 🌎
+- を使用してセマンティック イメージ検索で CLIP を実行する方法に関する [ノートブック](https://colab.research.google.com/github/vivien000/clip-demo/blob/master/clip.ipynb#scrollTo=uzdFhRGqiWkR) [Unsplash](https://unsplash.com) および [TMBD](https://www.themoviedb.org/) データセット。 🌎
+
+**説明可能性**
+
+- 入力トークンと画像セグメントの類似性を視覚化する方法に関する [ノートブック](https://colab.research.google.com/github/hila-chefer/Transformer-MM-Explainability/blob/main/CLIP_explainability.ipynb)。 🌎
+
+ここに含めるリソースの送信に興味がある場合は、お気軽にプル リクエストを開いてください。審査させていただきます。
+リソースは、既存のリソースを複製するのではなく、何か新しいものを示すことが理想的です。
+
+## CLIPConfig
+
+[[autodoc]] CLIPConfig
+    - from_text_vision_configs
+
+## CLIPTextConfig
+
+[[autodoc]] CLIPTextConfig
+
+## CLIPVisionConfig
+
+[[autodoc]] CLIPVisionConfig
+
+## CLIPTokenizer
+
+[[autodoc]] CLIPTokenizer
+    - build_inputs_with_special_tokens
+    - get_special_tokens_mask
+    - create_token_type_ids_from_sequences
+    - save_vocabulary
+
+## CLIPTokenizerFast
+
+[[autodoc]] CLIPTokenizerFast
+
+## CLIPImageProcessor
+
+[[autodoc]] CLIPImageProcessor
+    - preprocess
+
+## CLIPFeatureExtractor
+
+[[autodoc]] CLIPFeatureExtractor
+
+## CLIPProcessor
+
+[[autodoc]] CLIPProcessor
+
+<frameworkcontent>
+<pt>
+
+## CLIPModel
+
+[[autodoc]] CLIPModel
+    - forward
+    - get_text_features
+    - get_image_features
+
+## CLIPTextModel
+
+[[autodoc]] CLIPTextModel
+    - forward
+
+## CLIPTextModelWithProjection
+
+[[autodoc]] CLIPTextModelWithProjection
+    - forward
+
+## CLIPVisionModelWithProjection
+
+[[autodoc]] CLIPVisionModelWithProjection
+    - forward
+
+## CLIPVisionModel
+
+[[autodoc]] CLIPVisionModel
+    - forward
+
+</pt>
+<tf>
+
+## TFCLIPModel
+
+[[autodoc]] TFCLIPModel
+    - call
+    - get_text_features
+    - get_image_features
+
+## TFCLIPTextModel
+
+[[autodoc]] TFCLIPTextModel
+    - call
+
+## TFCLIPVisionModel
+
+[[autodoc]] TFCLIPVisionModel
+    - call
+
+</tf>
+<jax>
+
+## FlaxCLIPModel
+
+[[autodoc]] FlaxCLIPModel
+    - __call__
+    - get_text_features
+    - get_image_features
+
+## FlaxCLIPTextModel
+
+[[autodoc]] FlaxCLIPTextModel
+    - __call__
+
+## FlaxCLIPTextModelWithProjection
+
+[[autodoc]] FlaxCLIPTextModelWithProjection
+    - __call__
+
+## FlaxCLIPVisionModel
+
+[[autodoc]] FlaxCLIPVisionModel
+    - __call__
+
+</jax>
+</frameworkcontent>

docs/source/ja/model_doc/clipseg.md

@@ -0,0 +1,104 @@
+<!--Copyright 2022 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.
+
+-->
+
+# CLIPSeg
+
+## Overview
+
+CLIPSeg モデルは、Timo Lüddecke, Alexander Ecker によって [Image Segmentation using Text and Image Prompts](https://arxiv.org/abs/2112.10003) で提案されました。
+そしてアレクサンダー・エッカー。 CLIPSeg は、ゼロショットおよびワンショット画像セグメンテーションのために、凍結された [CLIP](clip) モデルの上に最小限のデコーダを追加します。
+
+論文の要約は次のとおりです。
+
+*画像のセグメンテーションは通常、トレーニングによって解決されます。
+オブジェクト クラスの固定セットのモデル。後で追加のクラスやより複雑なクエリを組み込むとコストがかかります
+これらの式を含むデータセットでモデルを再トレーニングする必要があるためです。ここでシステムを提案します
+任意の情報に基づいて画像セグメンテーションを生成できます。
+テスト時にプロンプ​​トが表示されます。プロンプトはテキストまたは
+画像。このアプローチにより、統一されたモデルを作成できます。
+3 つの一般的なセグメンテーション タスクについて (1 回トレーニング済み)
+参照式のセグメンテーション、ゼロショット セグメンテーション、ワンショット セグメンテーションという明確な課題が伴います。
+CLIP モデルをバックボーンとして構築し、これをトランスベースのデコーダで拡張して、高密度なデータ通信を可能にします。
+予測。の拡張バージョンでトレーニングした後、
+PhraseCut データセット、私たちのシステムは、フリーテキスト プロンプトまたは
+クエリを表す追加の画像。後者の画像ベースのプロンプトのさまざまなバリエーションを詳細に分析します。
+この新しいハイブリッド入力により、動的適応が可能になります。
+前述の 3 つのセグメンテーション タスクのみですが、
+テキストまたは画像をクエリするバイナリ セグメンテーション タスクに
+定式化することができる。最後に、システムがうまく適応していることがわかりました
+アフォーダンスまたはプロパティを含む一般化されたクエリ*
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/clipseg_architecture.png"
+alt="描画" width="600"/>
+
+<small> CLIPSeg の概要。 <a href="https://arxiv.org/abs/2112.10003">元の論文から抜粋。</a> </small>
+
+このモデルは、[nielsr](https://huggingface.co/nielsr) によって提供されました。
+元のコードは [ここ](https://github.com/timojl/clipseg) にあります。
+
+## Usage tips
+
+- [`CLIPSegForImageSegmentation`] は、[`CLIPSegModel`] の上にデコーダを追加します。後者は [`CLIPModel`] と同じです。
+- [`CLIPSegForImageSegmentation`] は、テスト時に任意のプロンプトに基づいて画像セグメンテーションを生成できます。プロンプトはテキストのいずれかです
+(`input_ids` としてモデルに提供される) または画像 (`conditional_pixel_values` としてモデルに提供される)。カスタムを提供することもできます
+条件付き埋め込み (`conditional_embeddings`としてモデルに提供されます)。
+
+## Resources
+
+CLIPSeg の使用を開始するのに役立つ、公式 Hugging Face およびコミュニティ (🌎 で示されている) リソースのリスト。ここに含めるリソースの送信に興味がある場合は、お気軽にプル リクエストを開いてください。審査させていただきます。リソースは、既存のリソースを複製するのではなく、何か新しいものを示すことが理想的です。
+
+<PipelineTag pipeline="image-segmentation"/>
+
+- [CLIPSeg を使用したゼロショット画像セグメンテーション](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/CLIPSeg/Zero_shot_image_segmentation_with_CLIPSeg.ipynb) を説明するノートブック。
+
+## CLIPSegConfig
+
+[[autodoc]] CLIPSegConfig
+    - from_text_vision_configs
+
+## CLIPSegTextConfig
+
+[[autodoc]] CLIPSegTextConfig
+
+## CLIPSegVisionConfig
+
+[[autodoc]] CLIPSegVisionConfig
+
+## CLIPSegProcessor
+
+[[autodoc]] CLIPSegProcessor
+
+## CLIPSegModel
+
+[[autodoc]] CLIPSegModel
+    - forward
+    - get_text_features
+    - get_image_features
+
+## CLIPSegTextModel
+
+[[autodoc]] CLIPSegTextModel
+    - forward
+
+## CLIPSegVisionModel
+
+[[autodoc]] CLIPSegVisionModel
+    - forward
+
+## CLIPSegForImageSegmentation
+
+[[autodoc]] CLIPSegForImageSegmentation
+    - forward

docs/source/ja/model_doc/clvp.md

@@ -0,0 +1,123 @@
+<!--Copyright 2023 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.
+
+-->
+
+# CLVP
+
+## Overview
+
+CLVP (Contrastive Language-Voice Pretrained Transformer) モデルは、James Betker によって [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) で提案されました。
+
+論文の要約は次のとおりです。
+
+*近年、画像生成の分野は自己回帰変換器と DDPM の応用によって革命を起こしています。これらのアプローチは、画像生成のプロセスを段階的な確率的プロセスとしてモデル化し、大量のコンピューティングとデータを活用して画像の分布を学習します。パフォーマンスを向上させるこの方法論は、画像に限定される必要はありません。この論文では、画像生成ドメインの進歩を音声合成に適用する方法について説明します。その結果、表現力豊かなマルチ音声テキスト読み上げシステムである TorToise が誕生しました。
+
+
+このモデルは [Susnato Dhar](https://huggingface.co/susnato) によって提供されました。
+元のコードは [ここ](https://github.com/neonbjb/tortoise-tts) にあります。
+
+## Usage tips
+
+1. CLVP は Tortoise TTS モデルの不可欠な部分です。
+2. CLVP を使用して、生成されたさまざまな音声候補を提供されたテキストと比較することができ、最良の音声トークンが拡散モデルに転送されます。
+3. Tortoise の使用には、[`ClvpModelForConditionalGeneration.generate()`] メソッドの使用を強くお勧めします。
+4. 16 kHz を期待する他のオーディオ モデルとは対照的に、CLVP モデルはオーディオが 22.05 kHz でサンプリングされることを期待していることに注意してください。
+
+## Brief Explanation:
+
+- [`ClvpTokenizer`] はテキスト入力をトークン化し、[`ClvpFeatureExtractor`] は目的のオーディオからログ メル スペクトログラムを抽出します。
+- [`ClvpConditioningEncoder`] は、これらのテキスト トークンとオーディオ表現を取得し、テキストとオーディオに基づいて条件付けされた埋め込みに変換します。
+- [`ClvpForCausalLM`] は、これらの埋め込みを使用して複数の音声候補を生成します。
+- 各音声候補は音声エンコーダ ([`ClvpEncoder`]) を通過してベクトル表現に変換され、テキスト エンコーダ ([`ClvpEncoder`]) はテキスト トークンを同じ潜在空間に変換します。
+- 最後に、各音声ベクトルをテキスト ベクトルと比較して、どの音声ベクトルがテキスト ベクトルに最も類似しているかを確認します。
+- [`ClvpModelForConditionalGeneration.generate()`] は、上記のすべてのロジックを 1 つのメソッドに圧縮します。
+
+例 ：
+
+```python
+>>> import datasets
+>>> from transformers import ClvpProcessor, ClvpModelForConditionalGeneration
+
+>>> # Define the Text and Load the Audio (We are taking an audio example from HuggingFace Hub using `datasets` library).
+>>> text = "This is an example text."
+
+>>> ds = datasets.load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
+>>> ds = ds.cast_column("audio", datasets.Audio(sampling_rate=22050))
+>>> sample = ds[0]["audio"]
+
+>>> # Define processor and model.
+>>> processor = ClvpProcessor.from_pretrained("susnato/clvp_dev")
+>>> model = ClvpModelForConditionalGeneration.from_pretrained("susnato/clvp_dev")
+
+>>> # Generate processor output and model output.
+>>> processor_output = processor(raw_speech=sample["array"], sampling_rate=sample["sampling_rate"], text=text, return_tensors="pt")
+>>> generated_output = model.generate(**processor_output)
+```
+
+
+## ClvpConfig
+
+[[autodoc]] ClvpConfig
+    - from_sub_model_configs
+
+## ClvpEncoderConfig
+
+[[autodoc]] ClvpEncoderConfig
+
+## ClvpDecoderConfig
+
+[[autodoc]] ClvpDecoderConfig
+
+## ClvpTokenizer
+
+[[autodoc]] ClvpTokenizer
+    - save_vocabulary
+
+## ClvpFeatureExtractor
+
+[[autodoc]] ClvpFeatureExtractor
+    - __call__
+
+## ClvpProcessor
+
+[[autodoc]] ClvpProcessor
+    - __call__
+    - decode
+    - batch_decode
+
+## ClvpModelForConditionalGeneration
+
+[[autodoc]] ClvpModelForConditionalGeneration
+    - forward
+    - generate
+    - get_text_features
+    - get_speech_features
+
+## ClvpForCausalLM
+
+[[autodoc]] ClvpForCausalLM
+
+## ClvpModel
+
+[[autodoc]] ClvpModel
+
+## ClvpEncoder
+
+[[autodoc]] ClvpEncoder
+
+## ClvpDecoder
+
+[[autodoc]] ClvpDecoder
+

docs/source/ja/model_doc/code_llama.md

@@ -0,0 +1,125 @@
+<!--Copyright 2023 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 contains specific syntax for our doc-builder (similar to MDX) that may not be
+rendered properly in your Markdown viewer.
+
+-->
+
+# CodeLlama
+
+## Overview
+
+Code Llama モデルはによって [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) で提案されました。 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.
+論文の要約は次のとおりです。
+
+*私たちは Code Llama をリリースします。これは Llama 2 に基づくコードの大規模言語モデル ファミリであり、オープン モデルの中で最先端のパフォーマンス、埋め込み機能、大規模な入力コンテキストのサポート、プログラミング タスクのゼロショット命令追従機能を提供します。 。幅広いアプリケーションをカバーするための複数のフレーバーを提供しています。基盤モデル (Code Llama)、Python 特化 (Code Llama - Python)、およびそれぞれ 7B、13B、および 34B パラメーターを備えた命令追従モデル (Code Llama - Instruct) です。すべてのモデルは 16,000 トークンのシーケンスでトレーニングされ、最大 100,000 トークンの入力で改善が見られます。 7B および 13B コード ラマとコード ラマ - 命令バリアントは、周囲のコンテンツに基づいた埋め込みをサポートします。 Code Llama は、いくつかのコード ベンチマークでオープン モデルの中で最先端のパフォーマンスに達し、HumanEval と MBPP でそれぞれ最大 53% と 55% のスコアを獲得しました。特に、Code Llama - Python 7B は HumanEval および MBPP 上で Llama 2 70B よりも優れたパフォーマンスを示し、すべてのモデルは MultiPL-E 上で公開されている他のすべてのモデルよりも優れています。私たちは、研究と商業利用の両方を許可する寛容なライセンスに基づいて Code Llama をリリースしています。*
+
+すべての Code Llama モデル チェックポイントを [こちら](https://huggingface.co/models?search=code_llama) で確認し、[codellama org](https://huggingface.co/codellama) で正式にリリースされたチェックポイントを確認してください。
+
+このモデルは [ArthurZucker](https://huggingface.co/ArthurZ) によって提供されました。著者のオリジナルのコードは [こちら](https://github.com/facebookresearch/llama) にあります。
+
+## Usage tips and examples
+
+<Tip warning={true}>
+
+Code Llama のベースとなる`Llama2`ファミリー モデルは、`bfloat16`を使用してトレーニングされましたが、元の推論では`float16`を使用します。さまざまな精度を見てみましょう。
+
+* `float32`: モデルの初期化に関する PyTorch の規約では、モデルの重みがどの `dtype` で格納されたかに関係なく、モデルを `float32` にロードします。 「transformers」も、PyTorch との一貫性を保つためにこの規則に従っています。これはデフォルトで選択されます。 `AutoModel` API でストレージの重み付けタイプを使用してチェックポイントのロードをキャストする場合は、`torch_dtype="auto"` を指定する必要があります。 `model = AutoModelForCausalLM.from_pretrained("path", torch_dtype = "auto")`。
+* `bfloat16`: コード Llama はこの精度でトレーニングされているため、さらなるトレーニングや微調整に使用することをお勧めします。
+* `float16`: この精度を使用して推論を実行することをお勧めします。通常は `bfloat16` より高速であり、評価メトリクスには `bfloat16` と比べて明らかな低下が見られないためです。 bfloat16 を使用して推論を実行することもできます。微調整後、float16 と bfloat16 の両方で推論結果を確認することをお勧めします。
+
+上で述べたように、モデルを初期化するときに `torch_dtype="auto"` を使用しない限り、ストレージの重みの `dtype` はほとんど無関係です。その理由は、モデルが最初にダウンロードされ (オンラインのチェックポイントの `dtype` を使用)、次に `torch` のデフォルトの `dtype` にキャストされるためです (`torch.float32` になります)。指定された `torch_dtype` がある場合は、代わりにそれが使用されます。
+
+</Tip>
+
+チップ：
+- 充填タスクはすぐにサポートされます。入力を埋めたい場所には `tokenizer.fill_token` を使用する必要があります。
+- モデル変換スクリプトは、`Llama2` ファミリの場合と同じです。
+
+使用例は次のとおりです。
+
+```bash
+python src/transformers/models/llama/convert_llama_weights_to_hf.py \
+    --input_dir /path/to/downloaded/llama/weights --model_size 7B --output_dir /output/path
+```
+
+スクリプトを実行するには、(最大のバージョンであっても) float16 精度でモデル全体をホストするのに十分な CPU RAM が必要であることに注意してください。
+いくつかのチェックポイントがあり、それぞれにモデルの各重みの一部が含まれているため、すべてを RAM にロードする必要があります)。
+
+変換後、モデルとトークナイザーは次の方法でロードできます。
+
+```python
+>>> from transformers import LlamaForCausalLM, CodeLlamaTokenizer
+
+>>> tokenizer = CodeLlamaTokenizer.from_pretrained("codellama/CodeLlama-7b-hf")
+>>> model = LlamaForCausalLM.from_pretrained("codellama/CodeLlama-7b-hf")
+>>> PROMPT = '''def remove_non_ascii(s: str) -> str:
+    """ <FILL_ME>
+    return result
+'''
+>>> input_ids = tokenizer(PROMPT, return_tensors="pt")["input_ids"]
+>>> generated_ids = model.generate(input_ids, max_new_tokens=128)
+
+>>> filling = tokenizer.batch_decode(generated_ids[:, input_ids.shape[1]:], skip_special_tokens = True)[0]
+>>> print(PROMPT.replace("<FILL_ME>", filling))
+def remove_non_ascii(s: str) -> str:
+    """ Remove non-ASCII characters from a string.
+
+    Args:
+        s: The string to remove non-ASCII characters from.
+
+    Returns:
+        The string with non-ASCII characters removed.
+    """
+    result = ""
+    for c in s:
+        if ord(c) < 128:
+            result += c
+    return result
+```
+
+塗りつぶされた部分だけが必要な場合:
+
+```python
+>>> from transformers import pipeline
+>>> import torch
+
+>>> generator = pipeline("text-generation",model="codellama/CodeLlama-7b-hf",torch_dtype=torch.float16, device_map="auto")
+>>> generator('def remove_non_ascii(s: str) -> str:\n    """ <FILL_ME>\n    return result', max_new_tokens = 128, return_type = 1)
+```
+
+内部では、トークナイザーが [`<FILL_ME>` によって自動的に分割](https://huggingface.co/docs/transformers/main/model_doc/code_llama#transformers.CodeLlamaTokenizer.fill_token) して、[ に続く書式設定された入力文字列を作成します。オリジナルのトレーニング パターン](https://github.com/facebookresearch/codellama/blob/cb51c14ec761370ba2e2bc351374a79265d0465e/llama/generation.py#L402)。これは、パターンを自分で準備するよりも堅牢です。トークンの接着など、デバッグが非常に難しい落とし穴を回避できます。このモデルまたは他のモデルに必要な CPU および GPU メモリの量を確認するには、その値を決定するのに役立つ [この計算ツール](https://huggingface.co/spaces/hf-accelerate/model-memory-usage) を試してください。
+
+LLaMA トークナイザーは、[sentencepiece](https://github.com/google/sentencepiece) に基づく BPE モデルです。センテンスピースの癖の 1 つは、シーケンスをデコードするときに、最初のトークンが単語の先頭 (例: 「Banana」) である場合、トークナイザーは文字列の先頭にプレフィックス スペースを追加しないことです。
+
+<Tip>
+
+コード Llama は、`Llama2` モデルと同じアーキテクチャを持っています。API リファレンスについては、[Llama2 のドキュメント ページ](llama2) を参照してください。
+以下の Code Llama トークナイザーのリファレンスを見つけてください。
+</Tip>
+
+## CodeLlamaTokenizer
+
+[[autodoc]] CodeLlamaTokenizer
+    - build_inputs_with_special_tokens
+    - get_special_tokens_mask
+    - create_token_type_ids_from_sequences
+    - save_vocabulary
+
+## CodeLlamaTokenizerFast
+
+[[autodoc]] CodeLlamaTokenizerFast
+    - build_inputs_with_special_tokens
+    - get_special_tokens_mask
+    - create_token_type_ids_from_sequences
+    - update_post_processor
+    - save_vocabulary

docs/source/ja/model_doc/codegen.md

@@ -0,0 +1,90 @@
+<!--Copyright 2022 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.
+
+-->
+
+# CodeGen
+
+## Overview
+
+
+CodeGen モデルは、[A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) で Erik Nijkamp、Bo Pang、林宏明、Lifu Tu、Huan Wang、Yingbo Zhou、Silvio Savarese、Caiming Xiong およびカイミン・ションさん。
+
+CodeGen は、[The Pile](https://pile.eleuther.ai/)、BigQuery、BigPython で順次トレーニングされたプログラム合成用の自己回帰言語モデルです。
+
+論文の要約は次のとおりです。
+
+*プログラム合成は、与えられた問題仕様の解決策としてコンピューター プログラムを生成することを目的としています。我々は、大規模な言語モデルを介した会話型プログラム合成アプローチを提案します。これは、従来のアプローチで直面した広大なプログラム空間とユーザーの意図の仕様を検索するという課題に対処します。私たちの新しいアプローチでは、仕様とプログラムを作成するプロセスを、ユーザーとシステムの間の複数回の対話として捉えます。これはプログラム合成をシーケンス予測問題として扱い、仕様が自然言語で表現され、目的のプログラムが条件付きでサンプリングされます。私たちは、自然言語とプログラミング言語のデータに基づいて、CodeGen と呼ばれる大規模な言語モデルのファミリーをトレーニングします。データの監視が弱く、データ サイズとモデル サイズが拡大すると、単純な自己回帰言語モデリングから会話能力が生まれます。会話型プログラム合成におけるモデルの動作を研究するために、マルチターン プログラミング ベンチマーク (MTPB) を開発します。このベンチマークでは、各問題を解決するには、ユーザーとモデル間のマルチターン会話を介したマルチステップ合成が必要です。私たちの調査結果は、会話機能の出現と、提案されている会話プログラム合成パラダイムの有効性を示しています。さらに、私たちのモデル CodeGen (TPU-v4 でトレーニングされた最大 16B パラメーターを含む) は、HumanEval ベンチマークで OpenAI の Codex を上回ります。私たちはチェックポイントを含むトレーニング ライブラリ JaxFormer をオープン ソースのコントリビューションとして利用できるようにしています: [この https URL](https://github.com/salesforce/codegen)*。
+
+このモデルは [林 宏明](https://huggingface.co/rooa) によって寄稿されました。
+元のコードは [ここ](https://github.com/salesforce/codegen) にあります。
+
+## Checkpoint Naming
+
+* CodeGen モデル [チェックポイント](https://huggingface.co/models?other=codegen) は、可変サイズのさまざまな事前トレーニング データで利用できます。
+* 形式は「Salesforce/codegen-{size}-{data}」です。ここで、
+  * `size`: `350M`、`2B`、`6B`、`16B`
+  * `data`:
+    * `nl`: パイルで事前トレーニング済み
+    * `multi`: `nl` で初期化され、複数のプログラミング言語データでさらに事前トレーニングされます。
+    * `mono`: `multi` で初期化され、Python データでさらに事前トレーニングされます。
+* たとえば、`Salesforce/codegen-350M-mono` は、Pile、複数のプログラミング言語、および Python で順次事前トレーニングされた 3 億 5,000 万のパラメーターのチェックポイントを提供します。
+
+## Usage example
+
+```python
+>>> from transformers import AutoModelForCausalLM, AutoTokenizer
+
+>>> checkpoint = "Salesforce/codegen-350M-mono"
+>>> model = AutoModelForCausalLM.from_pretrained(checkpoint)
+>>> tokenizer = AutoTokenizer.from_pretrained(checkpoint)
+
+>>> text = "def hello_world():"
+
+>>> completion = model.generate(**tokenizer(text, return_tensors="pt"))
+
+>>> print(tokenizer.decode(completion[0]))
+def hello_world():
+    print("Hello World")
+
+hello_world()
+```
+
+## Resources
+
+- [因果言語モデリング タスク ガイド](../tasks/language_modeling)
+
+## CodeGenConfig
+
+[[autodoc]] CodeGenConfig
+    - all
+
+## CodeGenTokenizer
+
+[[autodoc]] CodeGenTokenizer
+    - save_vocabulary
+
+## CodeGenTokenizerFast
+
+[[autodoc]] CodeGenTokenizerFast
+
+## CodeGenModel
+
+[[autodoc]] CodeGenModel
+    - forward
+
+## CodeGenForCausalLM
+
+[[autodoc]] CodeGenForCausalLM
+    - forward

docs/source/ja/model_doc/conditional_detr.md

@@ -0,0 +1,75 @@
+<!--Copyright 2022 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.
+
+-->
+
+# Conditional DETR
+
+## Overview
+
+条件付き DETR モデルは、[Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) で Depu Meng、Xiaokang Chen、Zejia Fan、Gang Zeng、Houqiang Li、Yuhui Yuan、Lei Sun,  Jingdong Wang によって提案されました。王京東。条件付き DETR は、高速 DETR トレーニングのための条件付きクロスアテンション メカニズムを提供します。条件付き DETR は DETR よりも 6.7 倍から 10 倍速く収束します。
+
+論文の要約は次のとおりです。
+
+*最近開発された DETR アプローチは、トランスフォーマー エンコーダーおよびデコーダー アーキテクチャを物体検出に適用し、有望なパフォーマンスを実現します。この論文では、トレーニングの収束が遅いという重要な問題を扱い、高速 DETR トレーニングのための条件付きクロスアテンション メカニズムを紹介します。私たちのアプローチは、DETR におけるクロスアテンションが 4 つの四肢の位置特定とボックスの予測にコンテンツの埋め込みに大きく依存しているため、高品質のコンテンツの埋め込みの必要性が高まり、トレーニングの難易度が高くなるという点に動機づけられています。条件付き DETR と呼ばれる私たちのアプローチは、デコーダーのマルチヘッド クロスアテンションのためにデコーダーの埋め込みから条件付きの空間クエリを学習します。利点は、条件付き空間クエリを通じて、各クロスアテンション ヘッドが、個別の領域 (たとえば、1 つのオブジェクトの端またはオブジェクト ボックス内の領域) を含むバンドに注目できることです。これにより、オブジェクト分類とボックス回帰のための個別の領域をローカライズするための空間範囲が狭まり、コンテンツの埋め込みへの依存が緩和され、トレーニングが容易になります。実験結果は、条件付き DETR がバックボーン R50 および R101 で 6.7 倍速く収束し、より強力なバックボーン DC5-R50 および DC5-R101 で 10 倍速く収束することを示しています。コードは https://github.com/Atten4Vis/ConditionalDETR で入手できます。*
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/conditional_detr_curve.jpg"
+alt="描画" width="600"/>
+
+<small> 条件付き DETR は、元の DETR に比べてはるかに速い収束を示します。 <a href="https://arxiv.org/abs/2108.06152">元の論文</a>から引用。</small>
+
+このモデルは [DepuMeng](https://huggingface.co/DepuMeng) によって寄稿されました。元のコードは [ここ](https://github.com/Atten4Vis/ConditionalDETR) にあります。
+
+## Resources
+
+- [オブジェクト検出タスクガイド](../tasks/object_detection)
+
+## ConditionalDetrConfig
+
+[[autodoc]] ConditionalDetrConfig
+
+## ConditionalDetrImageProcessor
+
+[[autodoc]] ConditionalDetrImageProcessor
+    - preprocess
+    - post_process_object_detection
+    - post_process_instance_segmentation
+    - post_process_semantic_segmentation
+    - post_process_panoptic_segmentation
+
+## ConditionalDetrFeatureExtractor
+
+[[autodoc]] ConditionalDetrFeatureExtractor
+    - __call__
+    - post_process_object_detection
+    - post_process_instance_segmentation
+    - post_process_semantic_segmentation
+    - post_process_panoptic_segmentation
+
+## ConditionalDetrModel
+
+[[autodoc]] ConditionalDetrModel
+    - forward
+
+## ConditionalDetrForObjectDetection
+
+[[autodoc]] ConditionalDetrForObjectDetection
+    - forward
+
+## ConditionalDetrForSegmentation
+
+[[autodoc]] ConditionalDetrForSegmentation
+    - forward
+
+    

docs/source/ja/model_doc/convbert.md

@@ -0,0 +1,145 @@
+<!--Copyright 2020 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+
+⚠️ 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.
+
+-->
+
+# ConvBERT
+
+<div class="flex flex-wrap space-x-1">
+<a href="https://huggingface.co/models?filter=convbert">
+<img alt="Models" src="https://img.shields.io/badge/All_model_pages-convbert-blueviolet">
+</a>
+<a href="https://huggingface.co/spaces/docs-demos/conv-bert-base">
+<img alt="Spaces" src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue">
+</a>
+</div>
+
+## Overview
+
+ConvBERT モデルは、[ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) で Zihang Jiang、Weihao Yu、Daquan Zhou、Yunpeng Chen、Jiashi Feng、Shuicheng Yan によって提案されました。
+やん。
+
+論文の要約は次のとおりです。
+
+*BERT やそのバリアントなどの事前トレーニング済み言語モデルは、最近、さまざまな環境で目覚ましいパフォーマンスを達成しています。
+自然言語理解タスク。ただし、BERT はグローバルな自己注意ブロックに大きく依存しているため、問題が発生します。
+メモリ使用量と計算コストが大きくなります。すべての注意が入力シーケンス全体に対してクエリを実行しますが、
+グローバルな観点からアテンション マップを生成すると、一部のヘッドはローカルな依存関係のみを学習する必要があることがわかります。
+これは、計算の冗長性が存在することを意味します。したがって、我々は、新しいスパンベースの動的畳み込みを提案します。
+これらのセルフアテンション ヘッドを置き換えて、ローカルの依存関係を直接モデル化します。新しいコンボリューションヘッドと、
+自己注意の頭を休め、グローバルとローカルの両方の状況でより効率的な新しい混合注意ブロックを形成します
+学ぶ。この混合注意設計を BERT に装備し、ConvBERT モデルを構築します。実験でわかったことは、
+ConvBERT は、トレーニング コストが低く、さまざまな下流タスクにおいて BERT およびその亜種よりも大幅に優れたパフォーマンスを発揮します。
+モデルパラメータが少なくなります。注目すべきことに、ConvBERTbase モデルは 86.4 GLUE スコアを達成し、ELECTRAbase よりも 0.7 高いのに対し、
+トレーニングコストは 1/4 未満です。コードと事前トレーニングされたモデルがリリースされます。*
+
+このモデルは、[abhishek](https://huggingface.co/abhishek) によって提供されました。オリジナルの実装が見つかります
+ここ: https://github.com/yitu-opensource/ConvBert
+
+## Usage tips
+
+ConvBERT トレーニングのヒントは BERT のヒントと似ています。使用上のヒントについては、[BERT ドキュメント](bert) を参照してください。
+
+## Resources
+
+- [テキスト分類タスクガイド](../tasks/sequence_classification)
+- [トークン分類タスクガイド](../tasks/token_classification)
+- [質問回答タスク ガイド](../tasks/question_answering)
+- [マスクされた言語モデリング タスク ガイド](../tasks/masked_lang_modeling)
+- [多肢選択タスク ガイド](../tasks/multiple_choice)
+
+## ConvBertConfig
+
+[[autodoc]] ConvBertConfig
+
+## ConvBertTokenizer
+
+[[autodoc]] ConvBertTokenizer
+    - build_inputs_with_special_tokens
+    - get_special_tokens_mask
+    - create_token_type_ids_from_sequences
+    - save_vocabulary
+
+## ConvBertTokenizerFast
+
+[[autodoc]] ConvBertTokenizerFast
+
+<frameworkcontent>
+<pt>
+
+## ConvBertModel
+
+[[autodoc]] ConvBertModel
+    - forward
+
+## ConvBertForMaskedLM
+
+[[autodoc]] ConvBertForMaskedLM
+    - forward
+
+## ConvBertForSequenceClassification
+
+[[autodoc]] ConvBertForSequenceClassification
+    - forward
+
+## ConvBertForMultipleChoice
+
+[[autodoc]] ConvBertForMultipleChoice
+    - forward
+
+## ConvBertForTokenClassification
+
+[[autodoc]] ConvBertForTokenClassification
+    - forward
+
+## ConvBertForQuestionAnswering
+
+[[autodoc]] ConvBertForQuestionAnswering
+    - forward
+
+</pt>
+<tf>
+
+## TFConvBertModel
+
+[[autodoc]] TFConvBertModel
+    - call
+
+## TFConvBertForMaskedLM
+
+[[autodoc]] TFConvBertForMaskedLM
+    - call
+
+## TFConvBertForSequenceClassification
+
+[[autodoc]] TFConvBertForSequenceClassification
+    - call
+
+## TFConvBertForMultipleChoice
+
+[[autodoc]] TFConvBertForMultipleChoice
+    - call
+
+## TFConvBertForTokenClassification
+
+[[autodoc]] TFConvBertForTokenClassification
+    - call
+
+## TFConvBertForQuestionAnswering
+
+[[autodoc]] TFConvBertForQuestionAnswering
+    - call
+
+</tf>
+</frameworkcontent>

docs/source/ja/model_doc/convnext.md

@@ -0,0 +1,94 @@
+<!--Copyright 2022 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.
+
+-->
+
+# ConvNeXT
+
+## Overview
+
+ConvNeXT モデルは、[A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) で Zhuang Liu、Hanzi Mao、Chao-Yuan Wu、Christoph Feichtenhofer、Trevor Darrell、Saining Xie によって提案されました。
+ConvNeXT は、ビジョン トランスフォーマーの設計からインスピレーションを得た純粋な畳み込みモデル (ConvNet) であり、ビジョン トランスフォーマーよりも優れたパフォーマンスを発揮すると主張しています。
+
+論文の要約は次のとおりです。
+
+*視覚認識の「狂騒の 20 年代」は、最先端の画像分類モデルとして ConvNet にすぐに取って代わられた Vision Transformers (ViT) の導入から始まりました。
+一方、バニラ ViT は、オブジェクト検出やセマンティック セグメンテーションなどの一般的なコンピューター ビジョン タスクに適用すると困難に直面します。階層型トランスフォーマーです
+(Swin Transformers など) は、いくつかの ConvNet の以前の機能を再導入し、Transformers を汎用ビジョン バックボーンとして実用的に可能にし、幅広い環境で顕著なパフォーマンスを実証しました。
+さまざまな視覚タスク。ただし、このようなハイブリッド アプローチの有効性は、依然として、固有の誘導性ではなく、トランスフォーマーの本質的な優位性によるところが大きいと考えられています。
+畳み込みのバイアス。この作業では、設計空間を再検討し、純粋な ConvNet が達成できる限界をテストします。標準 ResNet を設計に向けて徐々に「最新化」します。
+ビジョン Transformer の概要を確認し、途中でパフォーマンスの違いに寄与するいくつかの重要なコンポーネントを発見します。この調査の結果は、純粋な ConvNet モデルのファミリーです。
+ConvNextと呼ばれます。 ConvNeXts は完全に標準の ConvNet モジュールから構築されており、精度と拡張性の点で Transformers と有利に競合し、87.8% の ImageNet トップ 1 精度を達成しています。
+標準 ConvNet のシンプルさと効率を維持しながら、COCO 検出と ADE20K セグメンテーションでは Swin Transformers よりも優れたパフォーマンスを発揮します。*
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/convnext_architecture.jpg"
+alt="描画" width="600"/>
+
+<small> ConvNeXT アーキテクチャ。 <a href="https://arxiv.org/abs/2201.03545">元の論文</a>から抜粋。</small>
+
+このモデルは、[nielsr](https://huggingface.co/nielsr) によって提供されました。 TensorFlow バージョンのモデルは [ariG23498](https://github.com/ariG23498) によって提供されました。
+[gante](https://github.com/gante)、および [sayakpaul](https://github.com/sayakpaul) (同等の貢献)。元のコードは [こちら](https://github.com/facebookresearch/ConvNeXt) にあります。
+
+## Resources
+
+ConvNeXT の使用を開始するのに役立つ公式 Hugging Face およびコミュニティ (🌎 で示される) リソースのリスト。
+
+<PipelineTag pipeline="image-classification"/>
+
+- [`ConvNextForImageClassification`] は、この [サンプル スクリプト](https://github.com/huggingface/transformers/tree/main/examples/pytorch/image-classification) および [ノートブック](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/image_classification.ipynb)。
+- 参照: [画像分類タスク ガイド](../tasks/image_classification)
+
+ここに含めるリソースの送信に興味がある場合は、お気軽にプル リクエストを開いてください。審査させていただきます。リソースは、既存のリソースを複製するのではなく、何か新しいものを示すことが理想的です。
+
+## ConvNextConfig
+
+[[autodoc]] ConvNextConfig
+
+## ConvNextFeatureExtractor
+
+[[autodoc]] ConvNextFeatureExtractor
+
+## ConvNextImageProcessor
+
+[[autodoc]] ConvNextImageProcessor
+    - preprocess
+
+<frameworkcontent>
+<pt>
+
+## ConvNextModel
+
+[[autodoc]] ConvNextModel
+    - forward
+
+## ConvNextForImageClassification
+
+[[autodoc]] ConvNextForImageClassification
+    - forward
+
+</pt>
+<tf>
+
+## TFConvNextModel
+
+[[autodoc]] TFConvNextModel
+    - call
+
+## TFConvNextForImageClassification
+
+[[autodoc]] TFConvNextForImageClassification
+    - call
+
+</tf>
+</frameworkcontent>

docs/source/ja/model_doc/convnextv2.md

@@ -0,0 +1,68 @@
+<!--Copyright 2023 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.
+
+-->
+
+# ConvNeXt V2
+
+## Overview
+
+ConvNeXt V2 モデルは、Sanghyun Woo、Shobhik Debnath、Ronghang Hu、Xinlei Chen、Zhuang Liu, In So Kweon, Saining Xie. によって [ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoders](https://arxiv.org/abs/2301.00808) で提案されました。
+ConvNeXt V2 は、Vision Transformers の設計からインスピレーションを得た純粋な畳み込みモデル (ConvNet) であり、[ConvNeXT](convnext) の後継です。
+
+論文の要約は次のとおりです。
+
+*アーキテクチャの改善と表現学習フレームワークの改善により、視覚認識の分野は 2020 年代初頭に急速な近代化とパフォーマンスの向上を実現しました。たとえば、ConvNeXt に代表される最新の ConvNet は、さまざまなシナリオで強力なパフォーマンスを実証しています。これらのモデルはもともと ImageNet ラベルを使用した教師あり学習用に設計されましたが、マスク オートエンコーダー (MAE) などの自己教師あり学習手法からも潜在的に恩恵を受けることができます。ただし、これら 2 つのアプローチを単純に組み合わせると、パフォーマンスが標準以下になることがわかりました。この論文では、完全畳み込みマスク オートエンコーダ フレームワークと、チャネル間の機能競合を強化するために ConvNeXt アーキテクチャに追加できる新しい Global Response Normalization (GRN) 層を提案します。この自己教師あり学習手法とアーキテクチャの改善の共同設計により、ConvNeXt V2 と呼ばれる新しいモデル ファミリが誕生しました。これにより、ImageNet 分類、COCO 検出、ADE20K セグメンテーションなどのさまざまな認識ベンチマークにおける純粋な ConvNet のパフォーマンスが大幅に向上します。また、ImageNet でトップ 1 の精度 76.7% を誇る効率的な 370 万パラメータの Atto モデルから、最先端の 88.9% を達成する 650M Huge モデルまで、さまざまなサイズの事前トレーニング済み ConvNeXt V2 モデルも提供しています。公開トレーニング データのみを使用した精度*。
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/convnextv2_architecture.png"
+alt="描画" width="600"/>
+
+<small> ConvNeXt V2 アーキテクチャ。 <a href="https://arxiv.org/abs/2301.00808">元の論文</a>から抜粋。</small>
+
+このモデルは [adirik](https://huggingface.co/adirik) によって提供されました。元のコードは [こちら](https://github.com/facebookresearch/ConvNeXt-V2) にあります。
+
+## Resources
+
+ConvNeXt V2 の使用を開始するのに役立つ公式 Hugging Face およびコミュニティ (🌎 で示される) リソースのリスト。
+
+<PipelineTag pipeline="image-classification"/>
+
+- [`ConvNextV2ForImageClassification`] は、この [サンプル スクリプト](https://github.com/huggingface/transformers/tree/main/examples/pytorch/image-classification) および [ノートブック](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/image_classification.ipynb)。
+
+ここに含めるリソースの送信に興味がある場合は、お気軽にプル リクエストを開いてください。審査させていただきます。リソースは、既存のリソースを複製するのではなく、何か新しいものを示すことが理想的です。
+
+## ConvNextV2Config
+
+[[autodoc]] ConvNextV2Config
+
+## ConvNextV2Model
+
+[[autodoc]] ConvNextV2Model
+    - forward
+
+## ConvNextV2ForImageClassification
+
+[[autodoc]] ConvNextV2ForImageClassification
+    - forward
+
+## TFConvNextV2Model
+
+[[autodoc]] TFConvNextV2Model
+    - call
+
+
+## TFConvNextV2ForImageClassification
+
+[[autodoc]] TFConvNextV2ForImageClassification
+    - call

docs/source/ja/model_doc/cpm.md

@@ -0,0 +1,54 @@
+<!--Copyright 2020 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+
+⚠️ 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.
+
+-->
+
+# CPM
+
+## Overview
+
+CPM モデルは、Zhengyan Zhang、Xu Han、Hao Zhou、Pei Ke、Yuxian Gu によって [CPM: A Large-scale Generative Chinese Pre-trained Language Model](https://arxiv.org/abs/2012.00413) で提案されました。葉徳明、秦裕佳、
+Yusheng Su、Haozhe Ji、Jian Guan、Fanchao Qi、Xiaozi 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。
+
+論文の要約は次のとおりです。
+
+*事前トレーニングされた言語モデル (PLM) は、さまざまな下流の NLP タスクに有益であることが証明されています。最近ではGPT-3、
+1,750億個のパラメータと570GBの学習データを備え、数回の撮影（1枚でも）の容量で大きな注目を集めました
+ゼロショット）学習。ただし、GPT-3 を適用して中国語の NLP タスクに対処することは依然として困難です。
+GPT-3 の言語は主に英語であり、パラメーターは公開されていません。この技術レポートでは、
+大規模な中国語トレーニング データに対する生成的事前トレーニングを備えた中国語事前トレーニング済み言語モデル (CPM)。最高に
+私たちの知識の限りでは、26 億のパラメータと 100GB の中国語トレーニング データを備えた CPM は、事前トレーニングされた中国語としては最大のものです。
+言語モデルは、会話、エッセイの作成、
+クローゼテストと言語理解。広範な実験により、CPM が多くの環境で優れたパフォーマンスを達成できることが実証されています。
+少数ショット (ゼロショットでも) 学習の設定での NLP タスク。*
+
+このモデルは [canwenxu](https://huggingface.co/canwenxu) によって提供されました。オリジナルの実装が見つかります
+ここ: https://github.com/TsinghuaAI/CPM-Generate
+
+
+<Tip>
+
+CPM のアーキテクチャは、トークン化方法を除いて GPT-2 と同じです。詳細については、[GPT-2 ドキュメント](gpt2) を参照してください。
+API リファレンス情報。
+
+</Tip>
+
+## CpmTokenizer
+
+[[autodoc]] CpmTokenizer
+
+## CpmTokenizerFast
+
+[[autodoc]] CpmTokenizerFast

docs/source/ja/perf_infer_gpu_one.md

@@ -44,7 +44,7 @@ Flash Attention 2は、モデルのdtypeが`fp16`または`bf16`の場合にの
 
 ### Quick usage
 
-モデルでFlash Attention 2を有効にするには、`from_pretrained`の引数に`use_flash_attention_2`を追加します。
+モデルでFlash Attention 2を有効にするには、`from_pretrained`の引数に`attn_implementation="flash_attention_2"`を追加します。
 
 
 ```python
@@ -57,7 +57,7 @@ tokenizer = AutoTokenizer.from_pretrained(model_id)
 model = AutoModelForCausalLM.from_pretrained(
     model_id, 
     torch_dtype=torch.bfloat16, 
-    use_flash_attention_2=True,
+    attn_implementation="flash_attention_2",
 )
 ```
 
@@ -114,7 +114,7 @@ tokenizer = AutoTokenizer.from_pretrained(model_id)
 model = AutoModelForCausalLM.from_pretrained(
     model_id, 
     load_in_8bit=True,
-    use_flash_attention_2=True,
+    attn_implementation="flash_attention_2",
 )
 ```
 
@@ -132,7 +132,7 @@ tokenizer = AutoTokenizer.from_pretrained(model_id)
 model = AutoModelForCausalLM.from_pretrained(
     model_id, 
     load_in_4bit=True,
-    use_flash_attention_2=True,
+    attn_implementation="flash_attention_2",
 )
 ```
 
@@ -151,7 +151,7 @@ tokenizer = AutoTokenizer.from_pretrained(model_id)
 model = AutoModelForCausalLM.from_pretrained(
     model_id, 
     load_in_4bit=True,
-    use_flash_attention_2=True,
+    attn_implementation="flash_attention_2",
 )
 
 lora_config = LoraConfig(

docs/source/zh/main_classes/quantization.md

@@ -66,12 +66,12 @@ model = AutoModelForCausalLM.from_pretrained(model_id).to("cuda:0")
 
 ### 结合 AWQ 和 Flash Attention
 
-您可以将AWQ量化与Flash Attention结合起来，得到一个既被量化又更快速的模型。只需使用`from_pretrained`加载模型，并传递`use_flash_attention_2=True`参数。
+您可以将AWQ量化与Flash Attention结合起来，得到一个既被量化又更快速的模型。只需使用`from_pretrained`加载模型，并传递`attn_implementation="flash_attention_2"`参数。
 
 ```python
 from transformers import AutoModelForCausalLM, AutoTokenizer
 
-model = AutoModelForCausalLM.from_pretrained("TheBloke/zephyr-7B-alpha-AWQ", use_flash_attention_2=True, device_map="cuda:0")
+model = AutoModelForCausalLM.from_pretrained("TheBloke/zephyr-7B-alpha-AWQ", attn_implementation="flash_attention_2", device_map="cuda:0")
 ```
 
 ### 基准测试

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("4.36.0.dev0")
+check_min_version("4.36.0")
 
 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("4.36.0.dev0")
+check_min_version("4.36.0")
 
 require_version("datasets>=2.14.0", "To fix: pip install -r examples/flax/speech-recogintion/requirements.txt")
 
@@ -640,7 +640,7 @@ def main():
 
     # Create learning rate schedule
     linear_decay_lr_schedule_fn = create_learning_rate_fn(
-        len(vectorized_datasets["train"]),
+        total_train_steps,
         training_args.warmup_steps,
         training_args.learning_rate,
     )

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("4.36.0.dev0")
+check_min_version("4.36.0")
 
 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("4.36.0.dev0")
+check_min_version("4.36.0")
 
 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("4.36.0.dev0")
+check_min_version("4.36.0")
 
 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("4.36.0.dev0")
+check_min_version("4.36.0")
 
 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("4.36.0.dev0")
+check_min_version("4.36.0")
 
 require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/image-classification/requirements.txt")
 

examples/pytorch/image-classification/run_image_classification_no_trainer.py

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