Add Nystromformer (huggingface#14659)

* Initial commit * Config and modelling changes Added Nystromformer-specific attributes to config and removed all decoder functionality from modelling. * Modelling and test changes Added Nystrom approximation and removed decoder tests. * Code quality fixes * Modeling changes and conversion script Initial commits to conversion script, modeling changes. * Minor modeling changes and conversion script * Modeling changes * Correct modeling, add tests and documentation * Code refactor * Remove tokenizers * Code refactor * Update __init__.py * Fix bugs * Update src/transformers/__init__.py Co-authored-by: NielsRogge <[email protected]> * Update src/transformers/__init__.py Co-authored-by: NielsRogge <[email protected]> * Update src/transformers/models/nystromformer/__init__.py Co-authored-by: NielsRogge <[email protected]> * Update docs/source/model_doc/nystromformer.mdx Co-authored-by: NielsRogge <[email protected]> * Update src/transformers/models/nystromformer/configuration_nystromformer.py Co-authored-by: NielsRogge <[email protected]> * Update src/transformers/models/nystromformer/configuration_nystromformer.py Co-authored-by: NielsRogge <[email protected]> * Update src/transformers/models/nystromformer/configuration_nystromformer.py Co-authored-by: NielsRogge <[email protected]> * Update src/transformers/models/nystromformer/configuration_nystromformer.py Co-authored-by: NielsRogge <[email protected]> * Update src/transformers/models/nystromformer/convert_nystromformer_original_pytorch_checkpoint_to_pytorch.py Co-authored-by: NielsRogge <[email protected]> * Update src/transformers/models/nystromformer/configuration_nystromformer.py Co-authored-by: NielsRogge <[email protected]> * Update modeling and test_modeling * Code refactor * .rst to .mdx * doc changes * Doc changes * Update modeling_nystromformer.py * Doc changes * Fix copies * Apply suggestions from code review Co-authored-by: NielsRogge <[email protected]> * Apply suggestions from code review Co-authored-by: NielsRogge <[email protected]> * Update configuration_nystromformer.py * Fix copies * Update tests/test_modeling_nystromformer.py Co-authored-by: NielsRogge <[email protected]> * Update test_modeling_nystromformer.py * Apply suggestions from code review Co-authored-by: Lysandre Debut <[email protected]> * Fix code style * Update modeling_nystromformer.py * Update modeling_nystromformer.py * Fix code style * Reformat modeling file * Update modeling_nystromformer.py * Modify NystromformerForMultipleChoice * Fix code quality * Apply suggestions from code review Co-authored-by: Sylvain Gugger <[email protected]> * Code style changes and torch.no_grad() * make style * Apply suggestions from code review Co-authored-by: NielsRogge <[email protected]> Co-authored-by: Lysandre Debut <[email protected]> Co-authored-by: Sylvain Gugger <[email protected]>
avain · Jan 11, 2022 · 28e0914 · 28e0914
1 parent 444ea95
commit 28e0914
Show file tree

Hide file tree

Showing 17 changed files with 1,970 additions and 0 deletions.
diff --git a/README.md b/README.md
@@ -285,6 +285,7 @@ Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih.
 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. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
+1. **[Nyströmformer](https://huggingface.co/docs/transformers/master/model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
 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. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.

diff --git a/README_ko.md b/README_ko.md
@@ -264,6 +264,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
+1. **[Nyströmformer](https://huggingface.co/docs/transformers/master/model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
 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. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.

diff --git a/README_zh-hans.md b/README_zh-hans.md
@@ -288,6 +288,7 @@ conda install -c huggingface transformers
 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. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (来自 Microsoft Research) 伴随论文 [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) 由 Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu 发布。
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (来自 Google AI) 伴随论文 [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) 由 Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel 发布。
+1. **[Nyströmformer](https://huggingface.co/docs/transformers/master/model_doc/nystromformer)** (来自 the University of Wisconsin - Madison) 伴随论文 [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) 由 Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh 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. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (来自 Deepmind) 伴随论文 [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) 由 Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira 发布。
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (来自 VinAI Research) 伴随论文 [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) 由 Dat Quoc Nguyen and Anh Tuan Nguyen 发布。

diff --git a/README_zh-hant.md b/README_zh-hant.md
@@ -300,6 +300,7 @@ conda install -c huggingface transformers
 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. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
+1. **[Nyströmformer](https://huggingface.co/docs/transformers/master/model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
 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. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.

diff --git a/docs/source/_toctree.yml b/docs/source/_toctree.yml
@@ -214,6 +214,8 @@
       title: MPNet
     - local: model_doc/mt5
       title: MT5
+    - local: model_doc/nystromformer
+      title: Nyströmformer
     - local: model_doc/openai-gpt
       title: OpenAI GPT
     - local: model_doc/gpt2

diff --git a/docs/source/index.mdx b/docs/source/index.mdx
@@ -145,6 +145,7 @@ conversion utilities for the following models.
 1. **[Megatron-GPT2](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. **[MPNet](model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MT5](model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
+1. **[Nyströmformer](model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
 1. **[Pegasus](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. **[Perceiver IO](model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](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.
@@ -235,6 +236,7 @@ Flax), PyTorch, and/or TensorFlow.
 |         MobileBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 |            MPNet            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 |             mT5             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|        Nystromformer        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |         OpenAI GPT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 |        OpenAI GPT-2         |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
 |           Pegasus           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |

diff --git a/docs/source/model_doc/nystromformer.mdx b/docs/source/model_doc/nystromformer.mdx
@@ -0,0 +1,68 @@
+<!--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.
+-->
+
+# Nyströmformer
+
+## Overview
+
+The Nyströmformer model was proposed in [*Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention*](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn
+Fung, Yin Li, and Vikas Singh.
+
+The abstract from the paper is the following:
+
+*Transformers have emerged as a powerful tool for a broad range of natural language processing tasks. A key component
+that drives the impressive performance of Transformers is the self-attention mechanism that encodes the influence or
+dependence of other tokens on each specific token. While beneficial, the quadratic complexity of self-attention on the
+input sequence length has limited its application to longer sequences -- a topic being actively studied in the
+community. To address this limitation, we propose Nyströmformer -- a model that exhibits favorable scalability as a
+function of sequence length. Our idea is based on adapting the Nyström method to approximate standard self-attention
+with O(n) complexity. The scalability of Nyströmformer enables application to longer sequences with thousands of
+tokens. We perform evaluations on multiple downstream tasks on the GLUE benchmark and IMDB reviews with standard
+sequence length, and find that our Nyströmformer performs comparably, or in a few cases, even slightly better, than
+standard self-attention. On longer sequence tasks in the Long Range Arena (LRA) benchmark, Nyströmformer performs
+favorably relative to other efficient self-attention methods. Our code is available at this https URL.*
+
+This model was contributed by [novice03](https://huggingface.co/novice03). The original code can be found [here](https://github.com/mlpen/Nystromformer).
+
+## NystromformerConfig
+
+[[autodoc]] NystromformerConfig
+
+## NystromformerModel
+
+[[autodoc]] NystromformerModel
+    - forward
+
+## NystromformerForMaskedLM
+
+[[autodoc]] NystromformerForMaskedLM
+    - forward
+
+## NystromformerForSequenceClassification
+
+[[autodoc]] NystromformerForSequenceClassification
+    - forward
+
+## NystromformerForMultipleChoice
+
+[[autodoc]] NystromformerForMultipleChoice
+    - forward
+
+## NystromformerForTokenClassification
+
+[[autodoc]] NystromformerForTokenClassification
+    - forward
+
+## NystromformerForQuestionAnswering
+
+[[autodoc]] NystromformerForQuestionAnswering
+    - forward
diff --git a/src/transformers/__init__.py b/src/transformers/__init__.py
@@ -241,6 +241,10 @@
     "models.mobilebert": ["MOBILEBERT_PRETRAINED_CONFIG_ARCHIVE_MAP", "MobileBertConfig", "MobileBertTokenizer"],
     "models.mpnet": ["MPNET_PRETRAINED_CONFIG_ARCHIVE_MAP", "MPNetConfig", "MPNetTokenizer"],
     "models.mt5": ["MT5Config"],
+    "models.nystromformer": [
+        "NYSTROMFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP",
+        "NystromformerConfig",
+    ],
     "models.openai": ["OPENAI_GPT_PRETRAINED_CONFIG_ARCHIVE_MAP", "OpenAIGPTConfig", "OpenAIGPTTokenizer"],
     "models.pegasus": ["PEGASUS_PRETRAINED_CONFIG_ARCHIVE_MAP", "PegasusConfig", "PegasusTokenizer"],
     "models.perceiver": ["PERCEIVER_PRETRAINED_CONFIG_ARCHIVE_MAP", "PerceiverConfig", "PerceiverTokenizer"],
@@ -1122,6 +1126,19 @@
         ]
     )
     _import_structure["models.mt5"].extend(["MT5EncoderModel", "MT5ForConditionalGeneration", "MT5Model"])
+    _import_structure["models.nystromformer"].extend(
+        [
+            "NYSTROMFORMER_PRETRAINED_MODEL_ARCHIVE_LIST",
+            "NystromformerForMaskedLM",
+            "NystromformerForMultipleChoice",
+            "NystromformerForQuestionAnswering",
+            "NystromformerForSequenceClassification",
+            "NystromformerForTokenClassification",
+            "NystromformerLayer",
+            "NystromformerModel",
+            "NystromformerPreTrainedModel",
+        ]
+    )
     _import_structure["models.openai"].extend(
         [
             "OPENAI_GPT_PRETRAINED_MODEL_ARCHIVE_LIST",
@@ -2308,6 +2325,7 @@
     from .models.mobilebert import MOBILEBERT_PRETRAINED_CONFIG_ARCHIVE_MAP, MobileBertConfig, MobileBertTokenizer
     from .models.mpnet import MPNET_PRETRAINED_CONFIG_ARCHIVE_MAP, MPNetConfig, MPNetTokenizer
     from .models.mt5 import MT5Config
+    from .models.nystromformer import NYSTROMFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP, NystromformerConfig
     from .models.openai import OPENAI_GPT_PRETRAINED_CONFIG_ARCHIVE_MAP, OpenAIGPTConfig, OpenAIGPTTokenizer
     from .models.pegasus import PEGASUS_PRETRAINED_CONFIG_ARCHIVE_MAP, PegasusConfig, PegasusTokenizer
     from .models.perceiver import PERCEIVER_PRETRAINED_CONFIG_ARCHIVE_MAP, PerceiverConfig, PerceiverTokenizer
@@ -3041,6 +3059,17 @@
             MPNetPreTrainedModel,
         )
         from .models.mt5 import MT5EncoderModel, MT5ForConditionalGeneration, MT5Model
+        from .models.nystromformer import (
+            NYSTROMFORMER_PRETRAINED_MODEL_ARCHIVE_LIST,
+            NystromformerForMaskedLM,
+            NystromformerForMultipleChoice,
+            NystromformerForQuestionAnswering,
+            NystromformerForSequenceClassification,
+            NystromformerForTokenClassification,
+            NystromformerLayer,
+            NystromformerModel,
+            NystromformerPreTrainedModel,
+        )
         from .models.openai import (
             OPENAI_GPT_PRETRAINED_MODEL_ARCHIVE_LIST,
             OpenAIGPTDoubleHeadsModel,

diff --git a/src/transformers/models/__init__.py b/src/transformers/models/__init__.py
@@ -76,6 +76,7 @@
     mobilebert,
     mpnet,
     mt5,
+    nystromformer,
     openai,
     pegasus,
     perceiver,

diff --git a/src/transformers/models/auto/configuration_auto.py b/src/transformers/models/auto/configuration_auto.py
@@ -30,6 +30,7 @@
 CONFIG_MAPPING_NAMES = OrderedDict(
     [
         # Add configs here
+        ("nystromformer", "NystromformerConfig"),
         ("imagegpt", "ImageGPTConfig"),
         ("qdqbert", "QDQBertConfig"),
         ("vision-encoder-decoder", "VisionEncoderDecoderConfig"),
@@ -116,6 +117,7 @@
 CONFIG_ARCHIVE_MAP_MAPPING_NAMES = OrderedDict(
     [
         # Add archive maps here
+        ("nystromformer", "NYSTROMFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("imagegpt", "IMAGEGPT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("qdqbert", "QDQBERT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("fnet", "FNET_PRETRAINED_CONFIG_ARCHIVE_MAP"),
@@ -190,6 +192,7 @@
 MODEL_NAMES_MAPPING = OrderedDict(
     [
         # Add full (and cased) model names here
+        ("nystromformer", "Nystromformer"),
         ("imagegpt", "ImageGPT"),
         ("qdqbert", "QDQBert"),
         ("vision-encoder-decoder", "Vision Encoder decoder"),