Add BLIP-2 (#21441)

* First draft

* More improvements

* More improvements

* Improve conversion script

* Convert all weights

* Make forward pass work

* Make logits match

* More improvements

* More improvements

* More improvements

* Use get_input_embeddings

* Improve some more

* Improve model tests

* Improve model tests

* More improvements

* Fix processor

* Update files

* Update prepare_inputs_for_generation

* More improvements

* Fix copies

* More fixes

* Make fixup

* More improvements

* Add support for seq2seq language model

* More improvements

* Fix test

* More improvements

* Improve conversion script

* Remove some todo's

* Fix README's

* Improve conversion script

* Fix generation

* Fix style and remove Blip2Model

* Fix model outputs

* More improvements

* Set eos_token_id in config

* Fix quality

* Small improvements

* Add processor tests

* More improvements

* Apply suggestions

* Apply suggestions

* Add integration test

* Update image URL

* Add integration test

* Fix model_type

* Update style

* Improve docs

* Add doc tests

* Fix copies

* Remove tests which are passing

* Improve some more

* Add tests for seq2seq language models

* Minor fix

* Convert more checkpoints

* finalize CI

* Fix blip and blip2 processors

* add `accelerate` support for `blip2`

* clean up

* make style

* Update conversion script

* Update conversion script some more

* Update organization

* revert toc file

* add blip-2 to toc file

* Some more improvements

* Fix docstring

* Improve docs

---------

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
Co-authored-by: younesbelkada <younesbelkada@gmail.com>

README_ja.md

@@ -314,6 +314,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (Facebook から) Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston から公開された研究論文: [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637)
 1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (Facebook から) Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston から公開された研究論文: [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637)
 1. **[BLIP](https://huggingface.co/docs/transformers/model_doc/blip)** (Salesforce から) Junnan Li, Dongxu Li, Caiming Xiong, Steven Hoi から公開された研究論文: [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://arxiv.org/abs/2201.12086)
+1. **[BLIP-2](https://huggingface.co/docs/transformers/main/model_doc/blip-2)** (Salesforce から) Junnan Li, Dongxu Li, Silvio Savarese, Steven Hoi. から公開された研究論文 [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://arxiv.org/abs/2301.12597)
 1. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (BigScience workshop から) [BigScience Workshop](https://bigscience.huggingface.co/) から公開されました.
 1. **[BORT](https://huggingface.co/docs/transformers/model_doc/bort)** (Alexa から) Adrian de Wynter and Daniel J. Perry から公開された研究論文: [Optimal Subarchitecture Extraction For BERT](https://arxiv.org/abs/2010.10499)
 1. **[BridgeTower](https://huggingface.co/docs/transformers/main/model_doc/bridgetower)** (Harbin Institute of Technology/Microsoft Research Asia/Intel Labs から) released with the paper [BridgeTower: Building Bridges Between Encoders in Vision-Language Representation Learning](https://arxiv.org/abs/2206.08657) by Xiao Xu, Chenfei Wu, Shachar Rosenman, Vasudev Lal, Wanxiang Che, Nan Duan.