Add OneFormer Model (#20577)

* Add Oneformer Model

* Add OneFormer Tests

* Add UNIVERSAL_SEGMENTATION_MAPPING

* Fix config

* 🐛 Fix error encountered while writing tests

* 🔨 Fix instance segmentation post processing

* Format Files and Add Documentation

* Add Documentation mdx file

* Run make fixup

* Run make fix-copies

* Remove unnecessary code

* Format modeling_oneformer.py

* Add OneFormer to ImageSegmentationPipeline

* Format files

* Add Demo link to Readme

* Fix fomatting errors

* Fix test failures

* Update Table in index.mdx

* Fix version

* Fix style

* Remove OneFormer from TF

* Fix Imports

* Fix dummy objects

* Fix tests

* Add newline

* Remove OneFormerFeatureExtractor

* Remove CUDA Kernels

* Use AutoBackbone for Swin

* Fix description

* Use Image Processor

* Fix copies

* Fix formatting

* Fix import order

* Fix flake8 errors

* Fix doc errors

* Add Hindi Readme entry

* Update supported backbones

* Update supported backbones

* Undo Changes

* Fix type of config

* Fix isort

* Fix auto.mdx

* Fix swin config

* Replace DinatBackbone with AutoBackbone

* Use SwinBackbone

* Use SwinBackbone

* Fix conversion script

* Fix arguments

* Add argument description

* Fix style

* Add OneFormerProcessor

* Fix OneFormerProcessor Tests

* Fix mapping

* Fix imports

* Fix inits

* Fix style

* Fix comment

* Fix docstring

* Move OneFormer to MultiModal

* Fix Copies

* Remove size divisor

* Fix check_repo.py

* Fix copies

* Add Processor for Testing Pipeline

* Fix padding for tokens

* Fix variables

* Fix formatting with correct black version

* Add Image Processor Test

* Apply suggestions

* Revert common modeling

* Add check for task

* Fix conversion script

* Fix initialization order

* Fix tests

* Undo Pipeline Changes

* Fix layers in MLP

* Fix copies

* Update image paths

* Fix copies

* Apply suggestions

docs/source/es/index.mdx

@@ -109,6 +109,7 @@ La biblioteca actualmente contiene implementaciones de JAX, PyTorch y TensorFlow
 1. **[MPNet](model_doc/mpnet)** (de Microsoft Research) publicado con el paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) por Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MT5](model_doc/mt5)** (de Google AI) publicado con el paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) por Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
 1. **[Nyströmformer](model_doc/nystromformer)** (de la Universidad de Wisconsin - Madison) publicado con el paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) por Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
+1. **[OneFormer](model_doc/oneformer)** (de la SHI Labs) publicado con el paper [OneFormer: One Transformer to Rule Universal Image Segmentation](https://arxiv.org/abs/2211.06220) por Jitesh Jain, Jiachen Li, MangTik Chiu, Ali Hassani, Nikita Orlov, Humphrey Shi.
 1. **[Pegasus](model_doc/pegasus)** (de Google) publicado con el paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) por Jingqing Zhang, Yao Zhao, Mohammad Saleh y Peter J. Liu.
 1. **[Perceiver IO](model_doc/perceiver)** (de Deepmind) publicado con el paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) por Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](model_doc/phobert)** (de VinAI Research) publicado con el paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) por Dat Quoc Nguyen y Anh Tuan Nguyen.