Add support for XLM-R XL and XXL models by modeling_xlm_roberta_xl.py (#13727)

* add xlm roberta xl

* add convert xlm xl fairseq checkpoint to pytorch

* fix init and documents for xlm-roberta-xl

* fix indention

* add test for XLM-R xl,xxl

* fix model hub name

* fix some stuff

* up

* correct init

* fix more

* fix as suggestions

* add torch_device

* fix default values of doc strings

* fix leftovers

* merge to master

* up

* correct hub names

* fix docs

* fix model

* up

* finalize

* last fix

* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* add copied from

* make style
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

README.md

@@ -323,6 +323,7 @@ AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) by Sanyuan Chen, Yu Wu, Ch
 1. **[XLM](https://huggingface.co/docs/transformers/model_doc/xlm)** (from Facebook) released together with the paper [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) by Guillaume Lample and Alexis Conneau.
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (from Microsoft Research) released with the paper [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) by Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
 1. **[XLM-RoBERTa](https://huggingface.co/docs/transformers/model_doc/xlm-roberta)** (from Facebook AI), released together with the paper [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) by Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer and Veselin Stoyanov.
+1. **[XLM-RoBERTa-XL](https://huggingface.co/docs/transformers/master/model_doc/xlm-roberta-xl)** (from Facebook AI), released together with the paper [Larger-Scale Transformers for Multilingual Masked Language Modeling](https://arxiv.org/abs/2105.00572) by Naman Goyal, Jingfei Du, Myle Ott, Giri Anantharaman, Alexis Conneau.
 1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (from Google/CMU) released with the paper [​XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) by Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
 1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
 1. **[XLS-R](https://huggingface.co/docs/master/transformers/model_doc/xls_r)** (from Facebook AI) released with the paper [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) by Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli.

README_ko.md

@@ -301,6 +301,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[XLM](https://huggingface.co/docs/transformers/model_doc/xlm)** (from Facebook) released together with the paper [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) by Guillaume Lample and Alexis Conneau.
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (from Microsoft Research) released with the paper [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) by Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
 1. **[XLM-RoBERTa](https://huggingface.co/docs/transformers/model_doc/xlm-roberta)** (from Facebook AI), released together with the paper [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) by Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer and Veselin Stoyanov.
+1. **[XLM-RoBERTa-XL](https://huggingface.co/docs/transformers/master/model_doc/xlm-roberta-xl)** (from Facebook AI) released with the paper [Larger-Scale Transformers for Multilingual Masked Language Modeling](https://arxiv.org/abs/2105.00572) by Naman Goyal, Jingfei Du, Myle Ott, Giri Anantharaman, Alexis Conneau.
 1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (from Google/CMU) released with the paper [​XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) by Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
 1. **[XLS-R](https://huggingface.co/docs/master/transformers/model_doc/xls_r)** (from Facebook AI) released with the paper [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) by Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli.
 1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.

README_zh-hans.md

@@ -325,6 +325,7 @@ conda install -c huggingface transformers
 1. **[XLM](https://huggingface.co/docs/transformers/model_doc/xlm)** (来自 Facebook) 伴随论文 [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) 由 Guillaume Lample and Alexis Conneau 发布。
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (来自 Microsoft Research) 伴随论文 [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) 由 Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou 发布。
 1. **[XLM-RoBERTa](https://huggingface.co/docs/transformers/model_doc/xlm-roberta)** (来自 Facebook AI), 伴随论文 [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) 由 Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer and Veselin Stoyanov 发布。
+1. **[XLM-RoBERTa-XL](https://huggingface.co/docs/transformers/master/model_doc/xlm-roberta-xl)** (来自 Facebook AI) 伴随论文 [Larger-Scale Transformers for Multilingual Masked Language Modeling](https://arxiv.org/abs/2105.00572) 由 Naman Goyal, Jingfei Du, Myle Ott, Giri Anantharaman, Alexis Conneau 发布。
 1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (来自 Google/CMU) 伴随论文 [XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) 由 Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le 发布。
 1. **[XLS-R](https://huggingface.co/docs/master/transformers/model_doc/xls_r)** (来自 Facebook AI) 伴随论文 [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) 由 Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli 发布。
 1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (来自 Facebook AI) 伴随论文 [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) 由 Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli 发布。

README_zh-hant.md

@@ -337,6 +337,7 @@ conda install -c huggingface transformers
 1. **[XLM](https://huggingface.co/docs/transformers/model_doc/xlm)** (from Facebook) released together with the paper [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) by Guillaume Lample and Alexis Conneau.
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (from Microsoft Research) released with the paper [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) by Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
 1. **[XLM-RoBERTa](https://huggingface.co/docs/transformers/model_doc/xlm-roberta)** (from Facebook AI), released together with the paper [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) by Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer and Veselin Stoyanov.
+1. **[XLM-RoBERTa-XL](https://huggingface.co/docs/transformers/master/model_doc/xlm-roberta-xl)** (from Facebook AI) released with the paper [Larger-Scale Transformers for Multilingual Masked Language Modeling](https://arxiv.org/abs/2105.00572) by Naman Goyal, Jingfei Du, Myle Ott, Giri Anantharaman, Alexis Conneau.
 1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (from Google/CMU) released with the paper [​XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) by Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
 1. **[XLS-R](https://huggingface.co/docs/master/transformers/model_doc/xls_r)** (from Facebook AI) released with the paper [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) by Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli.
 1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.

docs/source/_toctree.yml

@@ -312,6 +312,8 @@
       title: XLM-ProphetNet
     - local: model_doc/xlm-roberta
       title: XLM-RoBERTa
+    - local: model_doc/xlm-roberta-xl
+      title: XLM-RoBERTa-XL
     - local: model_doc/xlnet
       title: XLNet
     - local: model_doc/xlsr_wav2vec2

docs/source/index.mdx

@@ -146,6 +146,7 @@ conversion utilities for the following models.
 1. **[XLM](model_doc/xlm)** (from Facebook) released together with the paper [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) by Guillaume Lample and Alexis Conneau.
 1. **[XLM-ProphetNet](model_doc/xlm-prophetnet)** (from Microsoft Research) released with the paper [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) by Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
 1. **[XLM-RoBERTa](model_doc/xlm-roberta)** (from Facebook AI), released together with the paper [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) by Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer and Veselin Stoyanov.
+1. **[XLM-RoBERTa-XL](model_doc/xlm-roberta-xl)** (from Facebook AI), released together with the paper [Larger-Scale Transformers for Multilingual Masked Language Modeling](https://arxiv.org/abs/2105.00572) by Naman Goyal, Jingfei Du, Myle Ott, Giri Anantharaman, Alexis Conneau.
 1. **[XLNet](model_doc/xlnet)** (from Google/CMU) released with the paper [​XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) by Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
 1. **[XLSR-Wav2Vec2](model_doc/xlsr_wav2vec2)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
 1. **[XLS-R](https://huggingface.co/docs/master/transformers/model_doc/xls_r)** (from Facebook AI) released with the paper [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) by Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli.
@@ -246,6 +247,7 @@ Flax), PyTorch, and/or TensorFlow.
 |            XGLM             |       ✅       |       ✅       |       ✅        |         ❌         |      ✅      |
 |             XLM             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
 |         XLM-RoBERTa         |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|       XLM-RoBERTa-XL        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |        XLMProphetNet        |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            XLNet            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 |            YOSO             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |

docs/source/model_doc/xlm-roberta-xl.mdx

+<!--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.
+-->
+
+# XLM-RoBERTa-XL
+
+## Overview
+
+The XLM-RoBERTa-XL model was proposed in [Larger-Scale Transformers for Multilingual Masked Language Modeling](https://arxiv.org/abs/2105.00572) by Naman Goyal, Jingfei Du, Myle Ott, Giri Anantharaman, Alexis Conneau. 
+
+The abstract from the paper is the following:
+
+*Recent work has demonstrated the effectiveness of cross-lingual language model pretraining for cross-lingual understanding. In this study, we present the results of two larger multilingual masked language models, with 3.5B and 10.7B parameters. Our two new models dubbed XLM-R XL and XLM-R XXL outperform XLM-R by 1.8% and 2.4% average accuracy on XNLI. Our model also outperforms the RoBERTa-Large model on several English tasks of the GLUE benchmark by 0.3% on average while handling 99 more languages. This suggests pretrained models with larger capacity may obtain both strong performance on high-resource languages while greatly improving low-resource languages. We make our code and models publicly available.*
+
+Tips:
+
+- XLM-RoBERTa-XL is a multilingual model trained on 100 different languages. Unlike some XLM multilingual models, it does
+  not require `lang` tensors to understand which language is used, and should be able to determine the correct
+  language from the input ids.
+
+This model was contributed by [Soonhwan-Kwon](https://github.com/Soonhwan-Kwon) and [stefan-it](https://huggingface.co/stefan-it). The original code can be found [here](https://github.com/pytorch/fairseq/tree/master/examples/xlmr).
+
+
+## XLMRobertaXLConfig
+
+[[autodoc]] XLMRobertaXLConfig
+
+## XLMRobertaXLModel
+
+[[autodoc]] XLMRobertaXLModel
+    - forward
+
+## XLMRobertaXLForCausalLM
+
+[[autodoc]] XLMRobertaXLForCausalLM
+    - forward
+
+## XLMRobertaXLForMaskedLM
+
+[[autodoc]] XLMRobertaXLForMaskedLM
+    - forward
+
+## XLMRobertaXLForSequenceClassification
+
+[[autodoc]] XLMRobertaXLForSequenceClassification
+    - forward
+
+## XLMRobertaXLForMultipleChoice
+
+[[autodoc]] XLMRobertaXLForMultipleChoice
+    - forward
+
+## XLMRobertaXLForTokenClassification
+
+[[autodoc]] XLMRobertaXLForTokenClassification
+    - forward
+
+## XLMRobertaXLForQuestionAnswering
+
+[[autodoc]] XLMRobertaXLForQuestionAnswering
+    - forward

docs/source/serialization.mdx

@@ -60,6 +60,7 @@ Ready-made configurations include the following architectures:
 - RoBERTa
 - T5
 - XLM-RoBERTa
+- XLM-RoBERTa-XL
 
 The ONNX conversion is supported for the PyTorch versions of the models. If you
 would like to be able to convert a TensorFlow model, please let us know by

src/transformers/__init__.py

@@ -332,6 +332,7 @@ _import_structure = {
     "models.xlm": ["XLM_PRETRAINED_CONFIG_ARCHIVE_MAP", "XLMConfig", "XLMTokenizer"],
     "models.xlm_prophetnet": ["XLM_PROPHETNET_PRETRAINED_CONFIG_ARCHIVE_MAP", "XLMProphetNetConfig"],
     "models.xlm_roberta": ["XLM_ROBERTA_PRETRAINED_CONFIG_ARCHIVE_MAP", "XLMRobertaConfig"],
+    "models.xlm_roberta_xl": ["XLM_ROBERTA_XL_PRETRAINED_CONFIG_ARCHIVE_MAP", "XLMRobertaXLConfig"],
     "models.xlnet": ["XLNET_PRETRAINED_CONFIG_ARCHIVE_MAP", "XLNetConfig"],
     "models.yoso": ["YOSO_PRETRAINED_CONFIG_ARCHIVE_MAP", "YosoConfig"],
     "onnx": [],
@@ -1507,6 +1508,19 @@ if is_torch_available():
             "XLMRobertaModel",
         ]
     )
+    _import_structure["models.xlm_roberta_xl"].extend(
+        [
+            "XLM_ROBERTA_XL_PRETRAINED_MODEL_ARCHIVE_LIST",
+            "XLMRobertaXLForCausalLM",
+            "XLMRobertaXLForMaskedLM",
+            "XLMRobertaXLForMultipleChoice",
+            "XLMRobertaXLForQuestionAnswering",
+            "XLMRobertaXLForSequenceClassification",
+            "XLMRobertaXLForTokenClassification",
+            "XLMRobertaXLModel",
+            "XLMRobertaXLPreTrainedModel",
+        ]
+    )
     _import_structure["models.xlnet"].extend(
         [
             "XLNET_PRETRAINED_MODEL_ARCHIVE_LIST",
@@ -2485,6 +2499,7 @@ if TYPE_CHECKING:
     from .models.xlm import XLM_PRETRAINED_CONFIG_ARCHIVE_MAP, XLMConfig, XLMTokenizer
     from .models.xlm_prophetnet import XLM_PROPHETNET_PRETRAINED_CONFIG_ARCHIVE_MAP, XLMProphetNetConfig
     from .models.xlm_roberta import XLM_ROBERTA_PRETRAINED_CONFIG_ARCHIVE_MAP, XLMRobertaConfig
+    from .models.xlm_roberta_xl import XLM_ROBERTA_XL_PRETRAINED_CONFIG_ARCHIVE_MAP, XLMRobertaXLConfig
     from .models.xlnet import XLNET_PRETRAINED_CONFIG_ARCHIVE_MAP, XLNetConfig
     from .models.yoso import YOSO_PRETRAINED_CONFIG_ARCHIVE_MAP, YosoConfig
 
@@ -3455,6 +3470,17 @@ if TYPE_CHECKING:
             XLMRobertaForTokenClassification,
             XLMRobertaModel,
         )
+        from .models.xlm_roberta_xl import (
+            XLM_ROBERTA_XL_PRETRAINED_MODEL_ARCHIVE_LIST,
+            XLMRobertaXLForCausalLM,
+            XLMRobertaXLForMaskedLM,
+            XLMRobertaXLForMultipleChoice,
+            XLMRobertaXLForQuestionAnswering,
+            XLMRobertaXLForSequenceClassification,
+            XLMRobertaXLForTokenClassification,
+            XLMRobertaXLModel,
+            XLMRobertaXLPreTrainedModel,
+        )
         from .models.xlnet import (
             XLNET_PRETRAINED_MODEL_ARCHIVE_LIST,
             XLNetForMultipleChoice,

src/transformers/models/__init__.py

@@ -120,6 +120,7 @@ from . import (
     xlm,
     xlm_prophetnet,
     xlm_roberta,
+    xlm_roberta_xl,
     xlnet,
     yoso,
 )

src/transformers/models/auto/configuration_auto.py

@@ -76,6 +76,7 @@ CONFIG_MAPPING_NAMES = OrderedDict(
         ("albert", "AlbertConfig"),
         ("bert-generation", "BertGenerationConfig"),
         ("camembert", "CamembertConfig"),
+        ("xlm-roberta-xl", "XLMRobertaXLConfig"),
         ("xlm-roberta", "XLMRobertaConfig"),
         ("pegasus", "PegasusConfig"),
         ("marian", "MarianConfig"),
@@ -250,6 +251,7 @@ MODEL_NAMES_MAPPING = OrderedDict(
         ("bert-generation", "Bert Generation"),
         ("camembert", "CamemBERT"),
         ("xlm-roberta", "XLM-RoBERTa"),
+        ("xlm-roberta-xl", "XLM-RoBERTa-XL"),
         ("pegasus", "Pegasus"),
         ("blenderbot", "Blenderbot"),
         ("marian", "Marian"),

src/transformers/models/auto/modeling_auto.py

@@ -75,6 +75,7 @@ MODEL_MAPPING_NAMES = OrderedDict(
         ("distilbert", "DistilBertModel"),
         ("albert", "AlbertModel"),
         ("camembert", "CamembertModel"),
+        ("xlm-roberta-xl", "XLMRobertaXLModel"),
         ("xlm-roberta", "XLMRobertaModel"),
         ("bart", "BartModel"),
         ("longformer", "LongformerModel"),
@@ -123,6 +124,7 @@ MODEL_FOR_PRETRAINING_MAPPING_NAMES = OrderedDict(
         ("distilbert", "DistilBertForMaskedLM"),
         ("albert", "AlbertForPreTraining"),
         ("camembert", "CamembertForMaskedLM"),
+        ("xlm-roberta-xl", "XLMRobertaXLForMaskedLM"),
         ("xlm-roberta", "XLMRobertaForMaskedLM"),
         ("bart", "BartForConditionalGeneration"),
         ("fsmt", "FSMTForConditionalGeneration"),
@@ -178,6 +180,7 @@ MODEL_WITH_LM_HEAD_MAPPING_NAMES = OrderedDict(
         ("distilbert", "DistilBertForMaskedLM"),
         ("albert", "AlbertForMaskedLM"),
         ("camembert", "CamembertForMaskedLM"),
+        ("xlm-roberta-xl", "XLMRobertaXLForMaskedLM"),
         ("xlm-roberta", "XLMRobertaForMaskedLM"),
         ("marian", "MarianMTModel"),
         ("fsmt", "FSMTForConditionalGeneration"),
@@ -220,6 +223,7 @@ MODEL_FOR_CAUSAL_LM_MAPPING_NAMES = OrderedDict(
         ("gpt_neo", "GPTNeoForCausalLM"),
         ("big_bird", "BigBirdForCausalLM"),
         ("camembert", "CamembertForCausalLM"),
+        ("xlm-roberta-xl", "XLMRobertaXLForCausalLM"),
         ("xlm-roberta", "XLMRobertaForCausalLM"),
         ("roberta", "RobertaForCausalLM"),
         ("bert", "BertLMHeadModel"),
@@ -304,6 +308,7 @@ MODEL_FOR_MASKED_LM_MAPPING_NAMES = OrderedDict(
         ("bart", "BartForConditionalGeneration"),
         ("mbart", "MBartForConditionalGeneration"),
         ("camembert", "CamembertForMaskedLM"),
+        ("xlm-roberta-xl", "XLMRobertaXLForMaskedLM"),
         ("xlm-roberta", "XLMRobertaForMaskedLM"),
         ("longformer", "LongformerForMaskedLM"),
         ("roberta", "RobertaForMaskedLM"),
@@ -379,6 +384,7 @@ MODEL_FOR_SEQUENCE_CLASSIFICATION_MAPPING_NAMES = OrderedDict(
         ("distilbert", "DistilBertForSequenceClassification"),
         ("albert", "AlbertForSequenceClassification"),
         ("camembert", "CamembertForSequenceClassification"),
+        ("xlm-roberta-xl", "XLMRobertaXLForSequenceClassification"),
         ("xlm-roberta", "XLMRobertaForSequenceClassification"),
         ("mbart", "MBartForSequenceClassification"),
         ("bart", "BartForSequenceClassification"),
@@ -430,6 +436,7 @@ MODEL_FOR_QUESTION_ANSWERING_MAPPING_NAMES = OrderedDict(
         ("bart", "BartForQuestionAnswering"),
         ("mbart", "MBartForQuestionAnswering"),
         ("longformer", "LongformerForQuestionAnswering"),
+        ("xlm-roberta-xl", "XLMRobertaXLForQuestionAnswering"),
         ("xlm-roberta", "XLMRobertaForQuestionAnswering"),
         ("roberta", "RobertaForQuestionAnswering"),
         ("squeezebert", "SqueezeBertForQuestionAnswering"),
@@ -476,6 +483,7 @@ MODEL_FOR_TOKEN_CLASSIFICATION_MAPPING_NAMES = OrderedDict(
         ("camembert", "CamembertForTokenClassification"),
         ("flaubert", "FlaubertForTokenClassification"),
         ("xlm", "XLMForTokenClassification"),
+        ("xlm-roberta-xl", "XLMRobertaXLForTokenClassification"),
         ("xlm-roberta", "XLMRobertaForTokenClassification"),
         ("longformer", "LongformerForTokenClassification"),
         ("roberta", "RobertaForTokenClassification"),
@@ -509,6 +517,7 @@ MODEL_FOR_MULTIPLE_CHOICE_MAPPING_NAMES = OrderedDict(
         ("convbert", "ConvBertForMultipleChoice"),
         ("camembert", "CamembertForMultipleChoice"),
         ("electra", "ElectraForMultipleChoice"),
+        ("xlm-roberta-xl", "XLMRobertaXLForMultipleChoice"),
         ("xlm-roberta", "XLMRobertaForMultipleChoice"),
         ("longformer", "LongformerForMultipleChoice"),
         ("roberta", "RobertaForMultipleChoice"),

src/transformers/models/xlm_roberta_xl/__init__.py

+# flake8: noqa
+# There's no way to ignore "F401 '...' imported but unused" warnings in this
+# module, but to preserve other warnings. So, don't check this module at all.
+
+# 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.
+
+from typing import TYPE_CHECKING
+
+from ...file_utils import _LazyModule, is_torch_available
+
+
+_import_structure = {
+    "configuration_xlm_roberta_xl": [
+        "XLM_ROBERTA_XL_PRETRAINED_CONFIG_ARCHIVE_MAP",
+        "XLMRobertaXLConfig",
+        "XLMRobertaXLOnnxConfig",
+    ],
+}
+
+if is_torch_available():
+    _import_structure["modeling_xlm_roberta_xl"] = [
+        "XLM_ROBERTA_XL_PRETRAINED_MODEL_ARCHIVE_LIST",
+        "XLMRobertaXLForCausalLM",
+        "XLMRobertaXLForMaskedLM",
+        "XLMRobertaXLForMultipleChoice",
+        "XLMRobertaXLForQuestionAnswering",
+        "XLMRobertaXLForSequenceClassification",
+        "XLMRobertaXLForTokenClassification",
+        "XLMRobertaXLModel",
+        "XLMRobertaXLPreTrainedModel",
+    ]
+
+if TYPE_CHECKING:
+    from .configuration_xlm_roberta_xl import (
+        XLM_ROBERTA_XL_PRETRAINED_CONFIG_ARCHIVE_MAP,
+        XLMRobertaXLConfig,
+        XLMRobertaXLOnnxConfig,
+    )
+
+    if is_torch_available():
+        from .modeling_xlm_roberta_xl import (
+            XLM_ROBERTA_XL_PRETRAINED_MODEL_ARCHIVE_LIST,
+            XLMRobertaXLForCausalLM,
+            XLMRobertaXLForMaskedLM,
+            XLMRobertaXLForMultipleChoice,
+            XLMRobertaXLForQuestionAnswering,
+            XLMRobertaXLForSequenceClassification,
+            XLMRobertaXLForTokenClassification,
+            XLMRobertaXLModel,
+            XLMRobertaXLPreTrainedModel,
+        )
+
+else:
+    import sys
+
+    sys.modules[__name__] = _LazyModule(__name__, globals()["__file__"], _import_structure)

src/transformers/models/xlm_roberta_xl/configuration_xlm_roberta_xl.py

+# coding=utf-8
+# Copyright 2022 The HuggingFace Inc. team.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+""" XLM_ROBERTa_XL configuration"""
+
+from collections import OrderedDict
+from typing import Mapping
+
+from ...configuration_utils import PretrainedConfig
+from ...onnx import OnnxConfig
+from ...utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+XLM_ROBERTA_XL_PRETRAINED_CONFIG_ARCHIVE_MAP = {
+    "xlm-roberta-xl": "https://huggingface.co/facebook/xlm-roberta-xl/resolve/main/config.json",
+    "xlm-roberta-xxl": "https://huggingface.co/facebook/xlm-roberta-xxl/resolve/main/config.json",
+    # See all XLM-RoBERTa-XL models at https://huggingface.co/models?filter=xlm-roberta-xl
+}
+
+
+class XLMRobertaXLConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`XLMRobertaXLModel`] or a [`TFXLMRobertaXLModel`].
+    It is used to instantiate a XLM_ROBERTA_XL model according to the specified arguments, defining the model
+    architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the
+    XLM_ROBERTA_XL [bert-base-uncased](https://huggingface.co/bert-base-uncased) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+
+    Args:
+        vocab_size (`int`, *optional*, defaults to 250880):
+            Vocabulary size of the XLM_ROBERTA_XL model. Defines the number of different tokens that can be represented
+            by the `inputs_ids` passed when calling [`XLMRobertaXLModel`].
+        hidden_size (`int`, *optional*, defaults to 2560):
+            Dimensionality of the encoder layers and the pooler layer.
+        num_hidden_layers (`int`, *optional*, defaults to 36):
+            Number of hidden layers in the Transformer encoder.
+        num_attention_heads (`int`, *optional*, defaults to 32):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        intermediate_size (`int`, *optional*, defaults to 10240):
+            Dimensionality of the "intermediate" (often named feed-forward) layer in the Transformer encoder.
+        hidden_act (`str` or `Callable`, *optional*, defaults to `"gelu"`):
+            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
+            `"relu"`, `"silu"` and `"gelu_new"` are supported.
+        hidden_dropout_prob (`float`, *optional*, defaults to 0.1):
+            The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
+        attention_probs_dropout_prob (`float`, *optional*, defaults to 0.1):
+            The dropout ratio for the attention probabilities.
+        max_position_embeddings (`int`, *optional*, defaults to 514):
+            The maximum sequence length that this model might ever be used with. Typically set this to something large
+            just in case (e.g., 512 or 1024 or 2048).
+        type_vocab_size (`int`, *optional*, defaults to 1):
+            The vocabulary size of the `token_type_ids` passed when calling [`XLMRobertaXLModel`] or
+            [`TFXLMRobertaXLModel`].
+        initializer_range (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        layer_norm_eps (`float`, *optional*, defaults to 1e-5):
+            The epsilon used by the layer normalization layers.
+        position_embedding_type (`str`, *optional*, defaults to `"absolute"`):
+            Type of position embedding. Choose one of `"absolute"`, `"relative_key"`, `"relative_key_query"`. For
+            positional embeddings use `"absolute"`. For more information on `"relative_key"`, please refer to
+            [Self-Attention with Relative Position Representations (Shaw et al.)](https://arxiv.org/abs/1803.02155).
+            For more information on `"relative_key_query"`, please refer to *Method 4* in [Improve Transformer Models
+            with Better Relative Position Embeddings (Huang et al.)](https://arxiv.org/abs/2009.13658).
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should return the last key/values attentions (not used by all models). Only
+            relevant if `config.is_decoder=True`.
+        classifier_dropout (`float`, *optional*):
+            The dropout ratio for the classification head.
+
+    Examples:
+
+    ```python
+    >>> from transformers import XLMRobertaXLModel, XLMRobertaXLConfig
+
+    >>> # Initializing a XLM_ROBERTA_XL bert-base-uncased style configuration
+    >>> configuration = XLMRobertaXLConfig()
+
+    >>> # Initializing a model from the bert-base-uncased style configuration
+    >>> model = XLMRobertaXLModel(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+    model_type = "xlm-roberta-xl"
+
+    def __init__(
+        self,
+        vocab_size=250880,
+        hidden_size=2560,
+        num_hidden_layers=36,
+        num_attention_heads=32,
+        intermediate_size=10240,
+        hidden_act="gelu",
+        hidden_dropout_prob=0.1,
+        attention_probs_dropout_prob=0.1,
+        max_position_embeddings=514,
+        type_vocab_size=1,
+        initializer_range=0.02,
+        layer_norm_eps=1e-05,
+        pad_token_id=1,
+        bos_token_id=0,
+        eos_token_id=2,
+        position_embedding_type="absolute",
+        use_cache=True,
+        classifier_dropout=None,
+        **kwargs
+    ):
+        super().__init__(pad_token_id=pad_token_id, bos_token_id=bos_token_id, eos_token_id=eos_token_id, **kwargs)
+        self.vocab_size = vocab_size
+        self.hidden_size = hidden_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.hidden_act = hidden_act
+        self.intermediate_size = intermediate_size
+        self.hidden_dropout_prob = hidden_dropout_prob
+        self.attention_probs_dropout_prob = attention_probs_dropout_prob
+        self.max_position_embeddings = max_position_embeddings
+        self.type_vocab_size = type_vocab_size
+        self.initializer_range = initializer_range
+        self.layer_norm_eps = layer_norm_eps
+        self.position_embedding_type = position_embedding_type
+        self.use_cache = use_cache
+        self.classifier_dropout = classifier_dropout
+
+
+# Copied from transformers.models.roberta.configuration_roberta.RobertaOnnxConfig with Roberta->XLMRobertaXL
+class XLMRobertaXLOnnxConfig(OnnxConfig):
+    @property
+    def inputs(self) -> Mapping[str, Mapping[int, str]]:
+        return OrderedDict(
+            [
+                ("input_ids", {0: "batch", 1: "sequence"}),
+                ("attention_mask", {0: "batch", 1: "sequence"}),
+            ]
+        )

src/transformers/models/xlm_roberta_xl/convert_xlm_roberta_xl_original_pytorch_checkpoint_to_pytorch.py

+# coding=utf-8
+# Copyright 2018 The HuggingFace Inc. team.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""Convert RoBERTa checkpoint."""
+
+import argparse
+import pathlib
+
+import fairseq
+import torch
+from fairseq.models.roberta import RobertaModel as FairseqRobertaModel
+from fairseq.modules import TransformerSentenceEncoderLayer
+from packaging import version
+
+from transformers import XLMRobertaConfig, XLMRobertaXLForMaskedLM, XLMRobertaXLForSequenceClassification
+from transformers.models.bert.modeling_bert import (
+    BertIntermediate,
+    BertLayer,
+    BertOutput,
+    BertSelfAttention,
+    BertSelfOutput,
+)
+from transformers.models.roberta.modeling_roberta import RobertaAttention
+from transformers.utils import logging
+
+
+if version.parse(fairseq.__version__) < version.parse("1.0.0a"):
+    raise Exception("requires fairseq >= 1.0.0a")
+
+logging.set_verbosity_info()
+logger = logging.get_logger(__name__)
+
+SAMPLE_TEXT = "Hello world! cécé herlolip"
+
+
+def convert_xlm_roberta_xl_checkpoint_to_pytorch(
+    roberta_checkpoint_path: str, pytorch_dump_folder_path: str, classification_head: bool
+):
+    """
+    Copy/paste/tweak roberta's weights to our BERT structure.
+    """
+    roberta = FairseqRobertaModel.from_pretrained(roberta_checkpoint_path)
+    roberta.eval()  # disable dropout
+    roberta_sent_encoder = roberta.model.encoder.sentence_encoder
+    config = XLMRobertaConfig(
+        vocab_size=roberta_sent_encoder.embed_tokens.num_embeddings,
+        hidden_size=roberta.cfg.model.encoder_embed_dim,
+        num_hidden_layers=roberta.cfg.model.encoder_layers,
+        num_attention_heads=roberta.cfg.model.encoder_attention_heads,
+        intermediate_size=roberta.cfg.model.encoder_ffn_embed_dim,
+        max_position_embeddings=514,
+        type_vocab_size=1,
+        layer_norm_eps=1e-5,  # PyTorch default used in fairseq
+    )
+    if classification_head:
+        config.num_labels = roberta.model.classification_heads["mnli"].out_proj.weight.shape[0]
+
+    print("Our RoBERTa config:", config)
+
+    model = XLMRobertaXLForSequenceClassification(config) if classification_head else XLMRobertaXLForMaskedLM(config)
+    model.eval()
+
+    # Now let's copy all the weights.
+    # Embeddings
+    model.roberta.embeddings.word_embeddings.weight = roberta_sent_encoder.embed_tokens.weight
+    model.roberta.embeddings.position_embeddings.weight = roberta_sent_encoder.embed_positions.weight
+    model.roberta.embeddings.token_type_embeddings.weight.data = torch.zeros_like(
+        model.roberta.embeddings.token_type_embeddings.weight
+    )  # just zero them out b/c RoBERTa doesn't use them.
+
+    model.roberta.encoder.LayerNorm.weight = roberta_sent_encoder.layer_norm.weight
+    model.roberta.encoder.LayerNorm.bias = roberta_sent_encoder.layer_norm.bias
+
+    for i in range(config.num_hidden_layers):
+        # Encoder: start of layer
+        layer: BertLayer = model.roberta.encoder.layer[i]
+        roberta_layer: TransformerSentenceEncoderLayer = roberta_sent_encoder.layers[i]
+
+        attention: RobertaAttention = layer.attention
+        attention.self_attn_layer_norm.weight = roberta_layer.self_attn_layer_norm.weight
+        attention.self_attn_layer_norm.bias = roberta_layer.self_attn_layer_norm.bias
+
+        # self attention
+        self_attn: BertSelfAttention = layer.attention.self
+        assert (
+            roberta_layer.self_attn.k_proj.weight.data.shape
+            == roberta_layer.self_attn.q_proj.weight.data.shape
+            == roberta_layer.self_attn.v_proj.weight.data.shape
+            == torch.Size((config.hidden_size, config.hidden_size))
+        )
+
+        self_attn.query.weight.data = roberta_layer.self_attn.q_proj.weight
+        self_attn.query.bias.data = roberta_layer.self_attn.q_proj.bias
+        self_attn.key.weight.data = roberta_layer.self_attn.k_proj.weight
+        self_attn.key.bias.data = roberta_layer.self_attn.k_proj.bias
+        self_attn.value.weight.data = roberta_layer.self_attn.v_proj.weight
+        self_attn.value.bias.data = roberta_layer.self_attn.v_proj.bias
+
+        # self-attention output
+        self_output: BertSelfOutput = layer.attention.output
+        assert self_output.dense.weight.shape == roberta_layer.self_attn.out_proj.weight.shape
+        self_output.dense.weight = roberta_layer.self_attn.out_proj.weight
+        self_output.dense.bias = roberta_layer.self_attn.out_proj.bias
+
+        # this one is final layer norm
+        layer.LayerNorm.weight = roberta_layer.final_layer_norm.weight
+        layer.LayerNorm.bias = roberta_layer.final_layer_norm.bias
+
+        # intermediate
+        intermediate: BertIntermediate = layer.intermediate
+        assert intermediate.dense.weight.shape == roberta_layer.fc1.weight.shape
+        intermediate.dense.weight = roberta_layer.fc1.weight
+        intermediate.dense.bias = roberta_layer.fc1.bias
+
+        # output
+        bert_output: BertOutput = layer.output
+        assert bert_output.dense.weight.shape == roberta_layer.fc2.weight.shape
+        bert_output.dense.weight = roberta_layer.fc2.weight
+        bert_output.dense.bias = roberta_layer.fc2.bias
+        # end of layer
+
+    if classification_head:
+        model.classifier.dense.weight = roberta.model.classification_heads["mnli"].dense.weight
+        model.classifier.dense.bias = roberta.model.classification_heads["mnli"].dense.bias
+        model.classifier.out_proj.weight = roberta.model.classification_heads["mnli"].out_proj.weight
+        model.classifier.out_proj.bias = roberta.model.classification_heads["mnli"].out_proj.bias
+    else:
+        # LM Head
+        model.lm_head.dense.weight = roberta.model.encoder.lm_head.dense.weight
+        model.lm_head.dense.bias = roberta.model.encoder.lm_head.dense.bias
+        model.lm_head.layer_norm.weight = roberta.model.encoder.lm_head.layer_norm.weight
+        model.lm_head.layer_norm.bias = roberta.model.encoder.lm_head.layer_norm.bias
+        model.lm_head.decoder.weight = roberta.model.encoder.lm_head.weight
+        model.lm_head.decoder.bias = roberta.model.encoder.lm_head.bias
+
+    # Let's check that we get the same results.
+    input_ids: torch.Tensor = roberta.encode(SAMPLE_TEXT).unsqueeze(0)  # batch of size 1
+
+    our_output = model(input_ids)[0]
+    if classification_head:
+        their_output = roberta.model.classification_heads["mnli"](roberta.extract_features(input_ids))
+    else:
+        their_output = roberta.model(input_ids)[0]
+    print(our_output.shape, their_output.shape)
+    max_absolute_diff = torch.max(torch.abs(our_output - their_output)).item()
+    print(f"max_absolute_diff = {max_absolute_diff}")  # ~ 1e-7
+    success = torch.allclose(our_output, their_output, atol=1e-3)
+    print("Do both models output the same tensors?", "🔥" if success else "💩")
+    if not success:
+        raise Exception("Something went wRoNg")
+
+    pathlib.Path(pytorch_dump_folder_path).mkdir(parents=True, exist_ok=True)
+    print(f"Saving model to {pytorch_dump_folder_path}")
+    model.save_pretrained(pytorch_dump_folder_path)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    # Required parameters
+    parser.add_argument(
+        "--roberta_checkpoint_path", default=None, type=str, required=True, help="Path the official PyTorch dump."
+    )
+    parser.add_argument(
+        "--pytorch_dump_folder_path", default=None, type=str, required=True, help="Path to the output PyTorch model."
+    )
+    parser.add_argument(
+        "--classification_head", action="store_true", help="Whether to convert a final classification head."
+    )
+    args = parser.parse_args()
+    convert_xlm_roberta_xl_checkpoint_to_pytorch(
+        args.roberta_checkpoint_path, args.pytorch_dump_folder_path, args.classification_head
+    )

src/transformers/models/xlm_roberta_xl/modeling_xlm_roberta_xl.py

+# coding=utf-8
+# Copyright 2022 The HuggingFace Inc. team.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""PyTorch XLM RoBERTa xl,xxl model."""
+
+import math
+
+import torch
+import torch.utils.checkpoint
+from packaging import version
+from torch import nn
+from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
+
+from ...activations import ACT2FN, gelu
+from ...file_utils import (
+    add_code_sample_docstrings,
+    add_start_docstrings,
+    add_start_docstrings_to_model_forward,
+    replace_return_docstrings,
+)
+from ...modeling_outputs import (
+    BaseModelOutputWithPastAndCrossAttentions,
+    BaseModelOutputWithPoolingAndCrossAttentions,
+    CausalLMOutputWithCrossAttentions,
+    MaskedLMOutput,
+    MultipleChoiceModelOutput,
+    QuestionAnsweringModelOutput,
+    SequenceClassifierOutput,
+    TokenClassifierOutput,
+)
+from ...modeling_utils import (
+    PreTrainedModel,
+    apply_chunking_to_forward,
+    find_pruneable_heads_and_indices,
+    prune_linear_layer,
+)
+from ...utils import logging
+from .configuration_xlm_roberta_xl import XLMRobertaXLConfig
+
+
+logger = logging.get_logger(__name__)
+
+_CHECKPOINT_FOR_DOC = "xlm-roberta-xlarge"
+_CONFIG_FOR_DOC = "XLMRobertaXLConfig"
+_TOKENIZER_FOR_DOC = "RobertaTokenizer"
+
+XLM_ROBERTA_XL_PRETRAINED_MODEL_ARCHIVE_LIST = [
+    "facebook/xlm-roberta-xl",
+    "facebook/xlm-roberta-xxl",
+    # See all RoBERTa models at https://huggingface.co/models?filter=xlm-roberta-xl
+]
+
+
+class XLMRobertaXLEmbeddings(nn.Module):
+    """
+    Same as BertEmbeddings with a tiny tweak for positional embeddings indexing.
+    """
+
+    def __init__(self, config):
+        super().__init__()
+        self.word_embeddings = nn.Embedding(config.vocab_size, config.hidden_size, padding_idx=config.pad_token_id)
+        self.position_embeddings = nn.Embedding(config.max_position_embeddings, config.hidden_size)
+        self.token_type_embeddings = nn.Embedding(config.type_vocab_size, config.hidden_size)
+
+        # self.LayerNorm is not snake-cased to stick with TensorFlow model variable name and be able to load
+        # any TensorFlow checkpoint file
+        self.dropout = nn.Dropout(config.hidden_dropout_prob)
+        # position_ids (1, len position emb) is contiguous in memory and exported when serialized
+        self.position_embedding_type = getattr(config, "position_embedding_type", "absolute")
+        self.register_buffer("position_ids", torch.arange(config.max_position_embeddings).expand((1, -1)))
+        if version.parse(torch.__version__) > version.parse("1.6.0"):
+            self.register_buffer(
+                "token_type_ids",
+                torch.zeros(self.position_ids.size(), dtype=torch.long),
+                persistent=False,
+            )
+
+        # End copy
+        self.padding_idx = config.pad_token_id
+        self.position_embeddings = nn.Embedding(
+            config.max_position_embeddings, config.hidden_size, padding_idx=self.padding_idx
+        )
+
+    def forward(
+        self, input_ids=None, token_type_ids=None, position_ids=None, inputs_embeds=None, past_key_values_length=0
+    ):
+        if position_ids is None:
+            if input_ids is not None:
+                # Create the position ids from the input token ids. Any padded tokens remain padded.
+                position_ids = create_position_ids_from_input_ids(input_ids, self.padding_idx, past_key_values_length)
+            else:
+                position_ids = self.create_position_ids_from_inputs_embeds(inputs_embeds)
+
+        if input_ids is not None:
+            input_shape = input_ids.size()
+        else:
+            input_shape = inputs_embeds.size()[:-1]
+
+        seq_length = input_shape[1]
+
+        # Setting the token_type_ids to the registered buffer in constructor where it is all zeros, which usually occurs
+        # when its auto-generated, registered buffer helps users when tracing the model without passing token_type_ids, solves
+        # issue #5664
+        if token_type_ids is None:
+            if hasattr(self, "token_type_ids"):
+                buffered_token_type_ids = self.token_type_ids[:, :seq_length]
+                buffered_token_type_ids_expanded = buffered_token_type_ids.expand(input_shape[0], seq_length)
+                token_type_ids = buffered_token_type_ids_expanded
+            else:
+                token_type_ids = torch.zeros(input_shape, dtype=torch.long, device=self.position_ids.device)
+
+        if inputs_embeds is None:
+            inputs_embeds = self.word_embeddings(input_ids)
+        token_type_embeddings = self.token_type_embeddings(token_type_ids)
+
+        embeddings = inputs_embeds + token_type_embeddings
+        if self.position_embedding_type == "absolute":
+            position_embeddings = self.position_embeddings(position_ids)
+            embeddings += position_embeddings
+
+        embeddings = self.dropout(embeddings)
+        return embeddings
+
+    # Copied from transformers.models.roberta.modeling_roberta.RobertaEmbeddings.create_position_ids_from_inputs_embeds
+    def create_position_ids_from_inputs_embeds(self, inputs_embeds):
+        """
+        We are provided embeddings directly. We cannot infer which are padded so just generate sequential position ids.
+
+        Args:
+            inputs_embeds: torch.Tensor
+
+        Returns: torch.Tensor
+        """
+        input_shape = inputs_embeds.size()[:-1]
+        sequence_length = input_shape[1]
+
+        position_ids = torch.arange(
+            self.padding_idx + 1, sequence_length + self.padding_idx + 1, dtype=torch.long, device=inputs_embeds.device
+        )
+        return position_ids.unsqueeze(0).expand(input_shape)
+
+
+# Copied from transformers.models.bert.modeling_bert.BertSelfAttention with Bert->XLMRobertaXL
+class XLMRobertaXLSelfAttention(nn.Module):
+    def __init__(self, config, position_embedding_type=None):
+        super().__init__()
+        if config.hidden_size % config.num_attention_heads != 0 and not hasattr(config, "embedding_size"):
+            raise ValueError(
+                f"The hidden size ({config.hidden_size}) is not a multiple of the number of attention "
+                f"heads ({config.num_attention_heads})"
+            )
+
+        self.num_attention_heads = config.num_attention_heads
+        self.attention_head_size = int(config.hidden_size / config.num_attention_heads)
+        self.all_head_size = self.num_attention_heads * self.attention_head_size
+
+        self.query = nn.Linear(config.hidden_size, self.all_head_size)
+        self.key = nn.Linear(config.hidden_size, self.all_head_size)
+        self.value = nn.Linear(config.hidden_size, self.all_head_size)
+
+        self.dropout = nn.Dropout(config.attention_probs_dropout_prob)
+        self.position_embedding_type = position_embedding_type or getattr(
+            config, "position_embedding_type", "absolute"
+        )
+        if self.position_embedding_type == "relative_key" or self.position_embedding_type == "relative_key_query":
+            self.max_position_embeddings = config.max_position_embeddings
+            self.distance_embedding = nn.Embedding(2 * config.max_position_embeddings - 1, self.attention_head_size)
+
+        self.is_decoder = config.is_decoder
+
+    def transpose_for_scores(self, x):
+        new_x_shape = x.size()[:-1] + (self.num_attention_heads, self.attention_head_size)
+        x = x.view(*new_x_shape)
+        return x.permute(0, 2, 1, 3)
+
+    def forward(
+        self,
+        hidden_states,
+        attention_mask=None,
+        head_mask=None,
+        encoder_hidden_states=None,
+        encoder_attention_mask=None,
+        past_key_value=None,
+        output_attentions=False,
+    ):
+        mixed_query_layer = self.query(hidden_states)
+
+        # If this is instantiated as a cross-attention module, the keys
+        # and values come from an encoder; the attention mask needs to be
+        # such that the encoder's padding tokens are not attended to.
+        is_cross_attention = encoder_hidden_states is not None
+
+        if is_cross_attention and past_key_value is not None:
+            # reuse k,v, cross_attentions
+            key_layer = past_key_value[0]
+            value_layer = past_key_value[1]
+            attention_mask = encoder_attention_mask
+        elif is_cross_attention:
+            key_layer = self.transpose_for_scores(self.key(encoder_hidden_states))
+            value_layer = self.transpose_for_scores(self.value(encoder_hidden_states))
+            attention_mask = encoder_attention_mask
+        elif past_key_value is not None:
+            key_layer = self.transpose_for_scores(self.key(hidden_states))
+            value_layer = self.transpose_for_scores(self.value(hidden_states))
+            key_layer = torch.cat([past_key_value[0], key_layer], dim=2)
+            value_layer = torch.cat([past_key_value[1], value_layer], dim=2)
+        else:
+            key_layer = self.transpose_for_scores(self.key(hidden_states))
+            value_layer = self.transpose_for_scores(self.value(hidden_states))
+
+        query_layer = self.transpose_for_scores(mixed_query_layer)
+
+        if self.is_decoder:
+            # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states.
+            # Further calls to cross_attention layer can then reuse all cross-attention
+            # key/value_states (first "if" case)
+            # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of
+            # all previous decoder key/value_states. Further calls to uni-directional self-attention
+            # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case)
+            # if encoder bi-directional self-attention `past_key_value` is always `None`
+            past_key_value = (key_layer, value_layer)
+
+        # Take the dot product between "query" and "key" to get the raw attention scores.
+        attention_scores = torch.matmul(query_layer, key_layer.transpose(-1, -2))
+
+        if self.position_embedding_type == "relative_key" or self.position_embedding_type == "relative_key_query":
+            seq_length = hidden_states.size()[1]
+            position_ids_l = torch.arange(seq_length, dtype=torch.long, device=hidden_states.device).view(-1, 1)
+            position_ids_r = torch.arange(seq_length, dtype=torch.long, device=hidden_states.device).view(1, -1)
+            distance = position_ids_l - position_ids_r
+            positional_embedding = self.distance_embedding(distance + self.max_position_embeddings - 1)
+            positional_embedding = positional_embedding.to(dtype=query_layer.dtype)  # fp16 compatibility
+
+            if self.position_embedding_type == "relative_key":
+                relative_position_scores = torch.einsum("bhld,lrd->bhlr", query_layer, positional_embedding)
+                attention_scores = attention_scores + relative_position_scores
+            elif self.position_embedding_type == "relative_key_query":
+                relative_position_scores_query = torch.einsum("bhld,lrd->bhlr", query_layer, positional_embedding)
+                relative_position_scores_key = torch.einsum("bhrd,lrd->bhlr", key_layer, positional_embedding)
+                attention_scores = attention_scores + relative_position_scores_query + relative_position_scores_key
+
+        attention_scores = attention_scores / math.sqrt(self.attention_head_size)
+        if attention_mask is not None:
+            # Apply the attention mask is (precomputed for all layers in XLMRobertaXLModel forward() function)
+            attention_scores = attention_scores + attention_mask
+
+        # Normalize the attention scores to probabilities.
+        attention_probs = nn.functional.softmax(attention_scores, dim=-1)
+
+        # This is actually dropping out entire tokens to attend to, which might
+        # seem a bit unusual, but is taken from the original Transformer paper.
+        attention_probs = self.dropout(attention_probs)
+
+        # Mask heads if we want to
+        if head_mask is not None:
+            attention_probs = attention_probs * head_mask
+
+        context_layer = torch.matmul(attention_probs, value_layer)
+
+        context_layer = context_layer.permute(0, 2, 1, 3).contiguous()
+        new_context_layer_shape = context_layer.size()[:-2] + (self.all_head_size,)
+        context_layer = context_layer.view(*new_context_layer_shape)
+
+        outputs = (context_layer, attention_probs) if output_attentions else (context_layer,)
+
+        if self.is_decoder:
+            outputs = outputs + (past_key_value,)
+        return outputs
+
+
+class XLMRobertaXLSelfOutput(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.dense = nn.Linear(config.hidden_size, config.hidden_size)
+        self.dropout = nn.Dropout(config.hidden_dropout_prob)
+
+    def forward(self, hidden_states, input_tensor):
+        hidden_states = self.dense(hidden_states)
+        hidden_states = self.dropout(hidden_states)
+        hidden_states = hidden_states + input_tensor
+        return hidden_states
+
+
+class XLMRobertaXLAttention(nn.Module):
+    def __init__(self, config, position_embedding_type=None):
+        super().__init__()
+        self.self_attn_layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.self = XLMRobertaXLSelfAttention(config, position_embedding_type=position_embedding_type)
+        self.output = XLMRobertaXLSelfOutput(config)
+        self.pruned_heads = set()
+
+    def prune_heads(self, heads):
+        if len(heads) == 0:
+            return
+        heads, index = find_pruneable_heads_and_indices(
+            heads, self.self.num_attention_heads, self.self.attention_head_size, self.pruned_heads
+        )
+
+        # Prune linear layers
+        self.self.query = prune_linear_layer(self.self.query, index)
+        self.self.key = prune_linear_layer(self.self.key, index)
+        self.self.value = prune_linear_layer(self.self.value, index)
+        self.output.dense = prune_linear_layer(self.output.dense, index, dim=1)
+
+        # Update hyper params and store pruned heads
+        self.self.num_attention_heads = self.self.num_attention_heads - len(heads)
+        self.self.all_head_size = self.self.attention_head_size * self.self.num_attention_heads
+        self.pruned_heads = self.pruned_heads.union(heads)
+
+    def forward(
+        self,
+        hidden_states,
+        attention_mask=None,
+        head_mask=None,
+        encoder_hidden_states=None,
+        encoder_attention_mask=None,
+        past_key_value=None,
+        output_attentions=False,
+    ):
+        intermediate = self.self_attn_layer_norm(hidden_states)
+        self_outputs = self.self(
+            intermediate,
+            attention_mask,
+            head_mask,
+            encoder_hidden_states,
+            encoder_attention_mask,
+            past_key_value,
+            output_attentions,
+        )
+        attention_output = self.output(self_outputs[0], hidden_states)
+        outputs = (attention_output,) + self_outputs[1:]  # add attentions if we output them
+        return outputs
+
+
+# Copied from transformers.models.bert.modeling_bert.BertIntermediate
+class XLMRobertaXLIntermediate(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.dense = nn.Linear(config.hidden_size, config.intermediate_size)
+        if isinstance(config.hidden_act, str):
+            self.intermediate_act_fn = ACT2FN[config.hidden_act]
+        else:
+            self.intermediate_act_fn = config.hidden_act
+
+    def forward(self, hidden_states):
+        hidden_states = self.dense(hidden_states)
+        hidden_states = self.intermediate_act_fn(hidden_states)
+        return hidden_states
+
+
+class XLMRobertaXLOutput(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.dense = nn.Linear(config.intermediate_size, config.hidden_size)
+
+    def forward(self, hidden_states, input_tensor):
+        hidden_states = self.dense(hidden_states)
+        hidden_states = hidden_states + input_tensor
+        return hidden_states
+
+
+class XLMRobertaXLLayer(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.chunk_size_feed_forward = config.chunk_size_feed_forward
+        self.seq_len_dim = 1
+        self.attention = XLMRobertaXLAttention(config)
+        self.is_decoder = config.is_decoder
+        self.add_cross_attention = config.add_cross_attention
+        if self.add_cross_attention:
+            if not self.is_decoder:
+                raise ValueError(f"{self} should be used as a decoder model if cross attention is added")
+            self.crossattention = XLMRobertaXLAttention(config, position_embedding_type="absolute")
+        self.intermediate = XLMRobertaXLIntermediate(config)
+        self.output = XLMRobertaXLOutput(config)
+        self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+
+    def forward(
+        self,
+        hidden_states,
+        attention_mask=None,
+        head_mask=None,
+        encoder_hidden_states=None,
+        encoder_attention_mask=None,
+        past_key_value=None,
+        output_attentions=False,
+    ):
+        # decoder uni-directional self-attention cached key/values tuple is at positions 1,2
+        self_attn_past_key_value = past_key_value[:2] if past_key_value is not None else None
+        self_attention_outputs = self.attention(
+            hidden_states,
+            attention_mask,
+            head_mask,
+            output_attentions=output_attentions,
+            past_key_value=self_attn_past_key_value,
+        )
+        attention_output = self_attention_outputs[0]
+
+        # if decoder, the last output is tuple of self-attn cache
+        if self.is_decoder:
+            outputs = self_attention_outputs[1:-1]
+            present_key_value = self_attention_outputs[-1]
+        else:
+            outputs = self_attention_outputs[1:]  # add self attentions if we output attention weights
+
+        cross_attn_present_key_value = None
+        if self.is_decoder and encoder_hidden_states is not None:
+            if not hasattr(self, "crossattention"):
+                raise ValueError(
+                    f"If `encoder_hidden_states` are passed, {self} has to be instantiated with cross-attention layers by setting `config.add_cross_attention=True`"
+                )
+
+            # cross_attn cached key/values tuple is at positions 3,4 of past_key_value tuple
+            cross_attn_past_key_value = past_key_value[-2:] if past_key_value is not None else None
+            cross_attention_outputs = self.crossattention(
+                attention_output,
+                attention_mask,
+                head_mask,
+                encoder_hidden_states,
+                encoder_attention_mask,
+                cross_attn_past_key_value,
+                output_attentions,
+            )
+            attention_output = cross_attention_outputs[0]
+            outputs = outputs + cross_attention_outputs[1:-1]  # add cross attentions if we output attention weights
+
+            # add cross-attn cache to positions 3,4 of present_key_value tuple
+            cross_attn_present_key_value = cross_attention_outputs[-1]
+            present_key_value = present_key_value + cross_attn_present_key_value
+
+        layer_output = apply_chunking_to_forward(
+            self.feed_forward_chunk, self.chunk_size_feed_forward, self.seq_len_dim, attention_output
+        )
+        outputs = (layer_output,) + outputs
+
+        # if decoder, return the attn key/values as the last output
+        if self.is_decoder:
+            outputs = outputs + (present_key_value,)
+
+        return outputs
+
+    def feed_forward_chunk(self, attention_output):
+        intermediate_output = self.LayerNorm(attention_output)
+        intermediate_output = self.intermediate(intermediate_output)
+        layer_output = self.output(intermediate_output, attention_output)
+        return layer_output
+
+
+class XLMRobertaXLEncoder(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.config = config
+        self.layer = nn.ModuleList([XLMRobertaXLLayer(config) for _ in range(config.num_hidden_layers)])
+        self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.gradient_checkpointing = False
+
+    def forward(
+        self,
+        hidden_states,
+        attention_mask=None,
+        head_mask=None,
+        encoder_hidden_states=None,
+        encoder_attention_mask=None,
+        past_key_values=None,
+        use_cache=None,
+        output_attentions=False,
+        output_hidden_states=False,
+        return_dict=True,
+    ):
+        all_hidden_states = () if output_hidden_states else None
+        all_self_attentions = () if output_attentions else None
+        all_cross_attentions = () if output_attentions and self.config.add_cross_attention else None
+
+        next_decoder_cache = () if use_cache else None
+        for i, layer_module in enumerate(self.layer):
+            if output_hidden_states:
+                all_hidden_states = all_hidden_states + (hidden_states,)
+
+            layer_head_mask = head_mask[i] if head_mask is not None else None
+            past_key_value = past_key_values[i] if past_key_values is not None else None
+
+            if self.gradient_checkpointing and self.training:
+
+                if use_cache:
+                    logger.warning(
+                        "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
+                    )
+                    use_cache = False
+
+                def create_custom_forward(module):
+                    def custom_forward(*inputs):
+                        return module(*inputs, past_key_value, output_attentions)
+
+                    return custom_forward
+
+                layer_outputs = torch.utils.checkpoint.checkpoint(
+                    create_custom_forward(layer_module),
+                    hidden_states,
+                    attention_mask,
+                    layer_head_mask,
+                    encoder_hidden_states,
+                    encoder_attention_mask,
+                )
+            else:
+                layer_outputs = layer_module(
+                    hidden_states,
+                    attention_mask,
+                    layer_head_mask,
+                    encoder_hidden_states,
+                    encoder_attention_mask,
+                    past_key_value,
+                    output_attentions,
+                )
+
+            hidden_states = layer_outputs[0]
+            if use_cache:
+                next_decoder_cache += (layer_outputs[-1],)
+            if output_attentions:
+                all_self_attentions = all_self_attentions + (layer_outputs[1],)
+                if self.config.add_cross_attention:
+                    all_cross_attentions = all_cross_attentions + (layer_outputs[2],)
+
+        hidden_states = self.LayerNorm(hidden_states)
+
+        if output_hidden_states:
+            all_hidden_states = all_hidden_states + (hidden_states,)
+
+        if not return_dict:
+            return tuple(
+                v
+                for v in [
+                    hidden_states,
+                    next_decoder_cache,
+                    all_hidden_states,
+                    all_self_attentions,
+                    all_cross_attentions,
+                ]
+                if v is not None
+            )
+        return BaseModelOutputWithPastAndCrossAttentions(
+            last_hidden_state=hidden_states,
+            past_key_values=next_decoder_cache,
+            hidden_states=all_hidden_states,
+            attentions=all_self_attentions,
+            cross_attentions=all_cross_attentions,
+        )
+
+
+# Copied from transformers.models.bert.modeling_bert.BertPooler
+class XLMRobertaXLPooler(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.dense = nn.Linear(config.hidden_size, config.hidden_size)
+        self.activation = nn.Tanh()
+
+    def forward(self, hidden_states):
+        # We "pool" the model by simply taking the hidden state corresponding
+        # to the first token.
+        first_token_tensor = hidden_states[:, 0]
+        pooled_output = self.dense(first_token_tensor)
+        pooled_output = self.activation(pooled_output)
+        return pooled_output
+
+
+class XLMRobertaXLPreTrainedModel(PreTrainedModel):
+    """
+    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
+    models.
+    """
+
+    config_class = XLMRobertaXLConfig
+    base_model_prefix = "roberta"
+
+    # Copied from transformers.models.bert.modeling_bert.BertPreTrainedModel._init_weights
+    def _init_weights(self, module):
+        """Initialize the weights"""
+        if isinstance(module, nn.Linear):
+            # Slightly different from the TF version which uses truncated_normal for initialization
+            # cf https://github.com/pytorch/pytorch/pull/5617
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+        elif isinstance(module, nn.LayerNorm):
+            module.bias.data.zero_()
+            module.weight.data.fill_(1.0)
+
+    def update_keys_to_ignore(self, config, del_keys_to_ignore):
+        """Remove some keys from ignore list"""
+        if not config.tie_word_embeddings:
+            # must make a new list, or the class variable gets modified!
+            self._keys_to_ignore_on_save = [k for k in self._keys_to_ignore_on_save if k not in del_keys_to_ignore]
+            self._keys_to_ignore_on_load_missing = [
+                k for k in self._keys_to_ignore_on_load_missing if k not in del_keys_to_ignore
+            ]
+
+
+XLM_ROBERTA_XL_START_DOCSTRING = r"""
+    This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the
+    library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
+    etc.) This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module)
+    subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to
+    general usage and behavior.
+
+    Parameters:
+        config ([`XLMRobertaXLConfig`]): Model configuration class with all the parameters of the
+            model. Initializing with a config file does not load the weights associated with the model, only the
+            configuration. Check out the [`~PreTrainedModel.from_pretrained`] method to load the model weights.
+"""
+
+XLM_ROBERTA_XL_INPUTS_DOCSTRING = r"""
+    Args:
+        input_ids (`torch.LongTensor` of shape `({0})`):
+            Indices of input sequence tokens in the vocabulary. Indices can be obtained using [`RobertaTokenizer`]. See
+            [`PreTrainedTokenizer.encode`] and [`PreTrainedTokenizer.__call__`] for details. [What are input
+            IDs?](../glossary#input-ids)
+        attention_mask (`torch.FloatTensor` of shape `({0})`, *optional*):
+            Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:
+
+            - 1 for tokens that are **not masked**,
+            - 0 for tokens that are **masked**.
+            [What are attention masks?](../glossary#attention-mask)
+        token_type_ids (`torch.LongTensor` of shape `({0})`, *optional*):
+            Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0,
+            1]`:
+
+            - 0 corresponds to a *sentence A* token,
+            - 1 corresponds to a *sentence B* token.
+            [What are token type IDs?](../glossary#token-type-ids)
+        position_ids (`torch.LongTensor` of shape `({0})`, *optional*):
+            Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
+            config.max_position_embeddings - 1]`. [What are position IDs?](../glossary#position-ids)
+        head_mask (`torch.FloatTensor` of shape `(num_heads,)` or `(num_layers, num_heads)`, *optional*):
+            Mask to nullify selected heads of the self-attention modules. Mask values selected in `[0, 1]`:
+
+            - 1 indicates the head is **not masked**,
+            - 0 indicates the head is **masked**.
+        inputs_embeds (`torch.FloatTensor` of shape `({0}, hidden_size)`, *optional*):
+            Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This
+            is useful if you want more control over how to convert `input_ids` indices into associated vectors than the
+            model's internal embedding lookup matrix.
+        output_attentions (`bool`, *optional*):
+            Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned
+            tensors for more detail.
+        output_hidden_states (`bool`, *optional*):
+            Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
+            more detail.
+        return_dict (`bool`, *optional*):
+            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+"""
+
+
+@add_start_docstrings(
+    "The bare XLM-RoBERTa-xlarge Model transformer outputting raw hidden-states without any specific head on top.",
+    XLM_ROBERTA_XL_START_DOCSTRING,
+)
+class XLMRobertaXLModel(XLMRobertaXLPreTrainedModel):
+    """
+    The model can behave as an encoder (with only self-attention) as well as a decoder, in which case a layer of
+    cross-attention is added between the self-attention layers, following the architecture described in *Attention is
+    all you need*_ by Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz
+    Kaiser and Illia Polosukhin. To behave as an decoder the model needs to be initialized with the `is_decoder`
+    argument of the configuration set to `True`. To be used in a Seq2Seq model, the model needs to initialized with
+    both `is_decoder` argument and `add_cross_attention` set to `True`; an `encoder_hidden_states` is then expected as
+    an input to the forward pass. .. _*Attention is all you need*: https://arxiv.org/abs/1706.03762
+    """
+
+    _keys_to_ignore_on_load_missing = [r"position_ids"]
+
+    # Copied from transformers.models.bert.modeling_bert.BertModel.__init__ with Bert->XLMRobertaXL
+    def __init__(self, config, add_pooling_layer=True):
+        super().__init__(config)
+        self.config = config
+
+        self.embeddings = XLMRobertaXLEmbeddings(config)
+        self.encoder = XLMRobertaXLEncoder(config)
+
+        self.pooler = XLMRobertaXLPooler(config) if add_pooling_layer else None
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_input_embeddings(self):
+        return self.embeddings.word_embeddings
+
+    def set_input_embeddings(self, value):
+        self.embeddings.word_embeddings = value
+
+    def _prune_heads(self, heads_to_prune):
+        """
+        Prunes heads of the model. heads_to_prune: dict of {layer_num: list of heads to prune in this layer} See base
+        class PreTrainedModel
+        """
+        for layer, heads in heads_to_prune.items():
+            self.encoder.layer[layer].attention.prune_heads(heads)
+
+    @add_start_docstrings_to_model_forward(XLM_ROBERTA_XL_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
+    @add_code_sample_docstrings(
+        processor_class=_TOKENIZER_FOR_DOC,
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=BaseModelOutputWithPoolingAndCrossAttentions,
+        config_class=_CONFIG_FOR_DOC,
+    )
+    # Copied from transformers.models.bert.modeling_bert.BertModel.forward
+    def forward(
+        self,
+        input_ids=None,
+        attention_mask=None,
+        token_type_ids=None,
+        position_ids=None,
+        head_mask=None,
+        inputs_embeds=None,
+        encoder_hidden_states=None,
+        encoder_attention_mask=None,
+        past_key_values=None,
+        use_cache=None,
+        output_attentions=None,
+        output_hidden_states=None,
+        return_dict=None,
+    ):
+        r"""
+        encoder_hidden_states  (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
+            Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if
+            the model is configured as a decoder.
+        encoder_attention_mask (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in
+            the cross-attention if the model is configured as a decoder. Mask values selected in `[0, 1]`:
+
+            - 1 for tokens that are **not masked**,
+            - 0 for tokens that are **masked**.
+        past_key_values (`tuple(tuple(torch.FloatTensor))` of length `config.n_layers` with each tuple having 4 tensors of shape `(batch_size, num_heads, sequence_length - 1, embed_size_per_head)`):
+            Contains precomputed key and value hidden states of the attention blocks. Can be used to speed up decoding.
+
+            If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
+            don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
+            `decoder_input_ids` of shape `(batch_size, sequence_length)`.
+        use_cache (`bool`, *optional*):
+            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
+            `past_key_values`).
+        """
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if self.config.is_decoder:
+            use_cache = use_cache if use_cache is not None else self.config.use_cache
+        else:
+            use_cache = False
+
+        if input_ids is not None and inputs_embeds is not None:
+            raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
+        elif input_ids is not None:
+            input_shape = input_ids.size()
+        elif inputs_embeds is not None:
+            input_shape = inputs_embeds.size()[:-1]
+        else:
+            raise ValueError("You have to specify either input_ids or inputs_embeds")
+
+        batch_size, seq_length = input_shape
+        device = input_ids.device if input_ids is not None else inputs_embeds.device
+
+        # past_key_values_length
+        past_key_values_length = past_key_values[0][0].shape[2] if past_key_values is not None else 0
+
+        if attention_mask is None:
+            attention_mask = torch.ones(((batch_size, seq_length + past_key_values_length)), device=device)
+
+        if token_type_ids is None:
+            if hasattr(self.embeddings, "token_type_ids"):
+                buffered_token_type_ids = self.embeddings.token_type_ids[:, :seq_length]
+                buffered_token_type_ids_expanded = buffered_token_type_ids.expand(batch_size, seq_length)
+                token_type_ids = buffered_token_type_ids_expanded
+            else:
+                token_type_ids = torch.zeros(input_shape, dtype=torch.long, device=device)
+
+        # We can provide a self-attention mask of dimensions [batch_size, from_seq_length, to_seq_length]
+        # ourselves in which case we just need to make it broadcastable to all heads.
+        extended_attention_mask: torch.Tensor = self.get_extended_attention_mask(attention_mask, input_shape, device)
+
+        # If a 2D or 3D attention mask is provided for the cross-attention
+        # we need to make broadcastable to [batch_size, num_heads, seq_length, seq_length]
+        if self.config.is_decoder and encoder_hidden_states is not None:
+            encoder_batch_size, encoder_sequence_length, _ = encoder_hidden_states.size()
+            encoder_hidden_shape = (encoder_batch_size, encoder_sequence_length)
+            if encoder_attention_mask is None:
+                encoder_attention_mask = torch.ones(encoder_hidden_shape, device=device)
+            encoder_extended_attention_mask = self.invert_attention_mask(encoder_attention_mask)
+        else:
+            encoder_extended_attention_mask = None
+
+        # Prepare head mask if needed
+        # 1.0 in head_mask indicate we keep the head
+        # attention_probs has shape bsz x n_heads x N x N
+        # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads]
+        # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length]
+        head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers)
+
+        embedding_output = self.embeddings(
+            input_ids=input_ids,
+            position_ids=position_ids,
+            token_type_ids=token_type_ids,
+            inputs_embeds=inputs_embeds,
+            past_key_values_length=past_key_values_length,
+        )
+        encoder_outputs = self.encoder(
+            embedding_output,
+            attention_mask=extended_attention_mask,
+            head_mask=head_mask,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_extended_attention_mask,
+            past_key_values=past_key_values,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+        sequence_output = encoder_outputs[0]
+        pooled_output = self.pooler(sequence_output) if self.pooler is not None else None
+
+        if not return_dict:
+            return (sequence_output, pooled_output) + encoder_outputs[1:]
+
+        return BaseModelOutputWithPoolingAndCrossAttentions(
+            last_hidden_state=sequence_output,
+            pooler_output=pooled_output,
+            past_key_values=encoder_outputs.past_key_values,
+            hidden_states=encoder_outputs.hidden_states,
+            attentions=encoder_outputs.attentions,
+            cross_attentions=encoder_outputs.cross_attentions,
+        )
+
+
+@add_start_docstrings(
+    """XLM-RoBERTa-xlarge Model with a `language modeling` head on top for CLM fine-tuning.""",
+    XLM_ROBERTA_XL_START_DOCSTRING,
+)
+class XLMRobertaXLForCausalLM(XLMRobertaXLPreTrainedModel):
+    _keys_to_ignore_on_save = [r"lm_head.decoder.weight", r"lm_head.decoder.bias"]
+    _keys_to_ignore_on_load_missing = [r"position_ids", r"lm_head.decoder.weight", r"lm_head.decoder.bias"]
+    _keys_to_ignore_on_load_unexpected = [r"pooler"]
+
+    def __init__(self, config):
+        super().__init__(config)
+
+        if not config.is_decoder:
+            logger.warning("If you want to use `RobertaLMHeadModel` as a standalone, add `is_decoder=True.`")
+
+        self.roberta = XLMRobertaXLModel(config, add_pooling_layer=False)
+        self.lm_head = XLMRobertaXLLMHead(config)
+
+        # The LM head weights require special treatment only when they are tied with the word embeddings
+        self.update_keys_to_ignore(config, ["lm_head.decoder.weight"])
+
+        self.init_weights()
+
+    def get_output_embeddings(self):
+        return self.lm_head.decoder
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head.decoder = new_embeddings
+
+    @add_start_docstrings_to_model_forward(XLM_ROBERTA_XL_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
+    @replace_return_docstrings(output_type=CausalLMOutputWithCrossAttentions, config_class=_CONFIG_FOR_DOC)
+    def forward(
+        self,
+        input_ids=None,
+        attention_mask=None,
+        token_type_ids=None,
+        position_ids=None,
+        head_mask=None,
+        inputs_embeds=None,
+        encoder_hidden_states=None,
+        encoder_attention_mask=None,
+        labels=None,
+        past_key_values=None,
+        use_cache=None,
+        output_attentions=None,
+        output_hidden_states=None,
+        return_dict=None,
+    ):
+        r"""
+        encoder_hidden_states  (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
+            Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if
+            the model is configured as a decoder.
+        encoder_attention_mask (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in
+            the cross-attention if the model is configured as a decoder. Mask values selected in `[0, 1]`:
+
+            - 1 for tokens that are **not masked**,
+            - 0 for tokens that are **masked**.
+        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Labels for computing the left-to-right language modeling loss (next word prediction). Indices should be in
+            `[-100, 0, ..., config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are
+            ignored (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`
+        past_key_values (`tuple(tuple(torch.FloatTensor))` of length `config.n_layers` with each tuple having 4 tensors of shape `(batch_size, num_heads, sequence_length - 1, embed_size_per_head)`):
+            Contains precomputed key and value hidden states of the attention blocks. Can be used to speed up decoding.
+            If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
+            don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
+            `decoder_input_ids` of shape `(batch_size, sequence_length)`.
+        use_cache (`bool`, *optional*):
+            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
+            `past_key_values`).
+
+        Returns:
+
+        Example:
+
+        ```python
+        >>> from transformers import RobertaTokenizer, RobertaForCausalLM, RobertaConfig
+        >>> import torch
+
+        >>> tokenizer = RobertaTokenizer.from_pretrained("roberta-base")
+        >>> config = RobertaConfig.from_pretrained("roberta-base")
+        >>> config.is_decoder = True
+        >>> model = RobertaForCausalLM.from_pretrained("roberta-base", config=config)
+        >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
+        >>> outputs = model(**inputs)
+        >>> prediction_logits = outputs.logits
+        ```
+        """
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+        if labels is not None:
+            use_cache = False
+
+        outputs = self.roberta(
+            input_ids,
+            attention_mask=attention_mask,
+            token_type_ids=token_type_ids,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            past_key_values=past_key_values,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+
+        sequence_output = outputs[0]
+        prediction_scores = self.lm_head(sequence_output)
+
+        lm_loss = None
+        if labels is not None:
+            # we are doing next-token prediction; shift prediction scores and input ids by one
+            shifted_prediction_scores = prediction_scores[:, :-1, :].contiguous()
+            labels = labels[:, 1:].contiguous()
+            loss_fct = CrossEntropyLoss()
+            lm_loss = loss_fct(shifted_prediction_scores.view(-1, self.config.vocab_size), labels.view(-1))
+
+        if not return_dict:
+            output = (prediction_scores,) + outputs[2:]
+            return ((lm_loss,) + output) if lm_loss is not None else output
+
+        return CausalLMOutputWithCrossAttentions(
+            loss=lm_loss,
+            logits=prediction_scores,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            cross_attentions=outputs.cross_attentions,
+        )
+
+    def prepare_inputs_for_generation(self, input_ids, past=None, attention_mask=None, **model_kwargs):
+        input_shape = input_ids.shape
+        # if model is used as a decoder in encoder-decoder model, the decoder attention mask is created on the fly
+        if attention_mask is None:
+            attention_mask = input_ids.new_ones(input_shape)
+
+        # cut decoder_input_ids if past is used
+        if past is not None:
+            input_ids = input_ids[:, -1:]
+
+        return {"input_ids": input_ids, "attention_mask": attention_mask, "past_key_values": past}
+
+    def _reorder_cache(self, past, beam_idx):
+        reordered_past = ()
+        for layer_past in past:
+            reordered_past += (tuple(past_state.index_select(0, beam_idx) for past_state in layer_past),)
+        return reordered_past
+
+
+@add_start_docstrings(
+    """XLM-RoBERTa-xlarge Model with a `language modeling` head on top.""", XLM_ROBERTA_XL_START_DOCSTRING
+)
+class XLMRobertaXLForMaskedLM(XLMRobertaXLPreTrainedModel):
+    _keys_to_ignore_on_save = [r"lm_head.decoder.weight", r"lm_head.decoder.bias"]
+    _keys_to_ignore_on_load_missing = [r"position_ids", r"lm_head.decoder.weight", r"lm_head.decoder.bias"]
+    _keys_to_ignore_on_load_unexpected = [r"pooler"]
+
+    def __init__(self, config):
+        super().__init__(config)
+
+        if config.is_decoder:
+            logger.warning(
+                "If you want to use `RobertaForMaskedLM` make sure `config.is_decoder=False` for "
+                "bi-directional self-attention."
+            )
+
+        self.roberta = XLMRobertaXLModel(config, add_pooling_layer=False)
+        self.lm_head = XLMRobertaXLLMHead(config)
+
+        # The LM head weights require special treatment only when they are tied with the word embeddings
+        self.update_keys_to_ignore(config, ["lm_head.decoder.weight"])
+
+        self.init_weights()
+
+    def get_output_embeddings(self):
+        return self.lm_head.decoder
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head.decoder = new_embeddings
+
+    @add_start_docstrings_to_model_forward(XLM_ROBERTA_XL_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
+    @add_code_sample_docstrings(
+        processor_class=_TOKENIZER_FOR_DOC,
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=MaskedLMOutput,
+        config_class=_CONFIG_FOR_DOC,
+        mask="<mask>",
+    )
+    def forward(
+        self,
+        input_ids=None,
+        attention_mask=None,
+        token_type_ids=None,
+        position_ids=None,
+        head_mask=None,
+        inputs_embeds=None,
+        encoder_hidden_states=None,
+        encoder_attention_mask=None,
+        labels=None,
+        output_attentions=None,
+        output_hidden_states=None,
+        return_dict=None,
+    ):
+        r"""
+        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ...,
+            config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the
+            loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`
+        kwargs (`Dict[str, any]`, optional, defaults to *{}*):
+            Used to hide legacy arguments that have been deprecated.
+        """
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        outputs = self.roberta(
+            input_ids,
+            attention_mask=attention_mask,
+            token_type_ids=token_type_ids,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+        sequence_output = outputs[0]
+        prediction_scores = self.lm_head(sequence_output)
+
+        masked_lm_loss = None
+        if labels is not None:
+            loss_fct = CrossEntropyLoss()
+            masked_lm_loss = loss_fct(prediction_scores.view(-1, self.config.vocab_size), labels.view(-1))
+
+        if not return_dict:
+            output = (prediction_scores,) + outputs[2:]
+            return ((masked_lm_loss,) + output) if masked_lm_loss is not None else output
+
+        return MaskedLMOutput(
+            loss=masked_lm_loss,
+            logits=prediction_scores,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )
+
+
+class XLMRobertaXLLMHead(nn.Module):
+    """XLM-Roberta-xlarge Head for masked language modeling."""
+
+    def __init__(self, config):
+        super().__init__()
+        self.dense = nn.Linear(config.hidden_size, config.hidden_size)
+        self.layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+
+        self.decoder = nn.Linear(config.hidden_size, config.vocab_size)
+        self.bias = nn.Parameter(torch.zeros(config.vocab_size))
+        self.decoder.bias = self.bias
+
+    def forward(self, features, **kwargs):
+        x = self.dense(features)
+        x = gelu(x)
+        x = self.layer_norm(x)
+
+        # project back to size of vocabulary with bias
+        x = self.decoder(x)
+
+        return x
+
+    def _tie_weights(self):
+        # To tie those two weights if they get disconnected (on TPU or when the bias is resized)
+        self.bias = self.decoder.bias
+
+
+@add_start_docstrings(
+    """
+    XLM-RoBERTa-xlarge Model transformer with a sequence classification/regression head on top (a linear layer on top
+    of the pooled output) e.g. for GLUE tasks.
+    """,
+    XLM_ROBERTA_XL_START_DOCSTRING,
+)
+class XLMRobertaXLForSequenceClassification(XLMRobertaXLPreTrainedModel):
+    _keys_to_ignore_on_load_missing = [r"position_ids"]
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.num_labels = config.num_labels
+        self.config = config
+
+        self.roberta = XLMRobertaXLModel(config, add_pooling_layer=False)
+        self.classifier = XLMRobertaXLClassificationHead(config)
+
+        self.init_weights()
+
+    @add_start_docstrings_to_model_forward(XLM_ROBERTA_XL_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
+    @add_code_sample_docstrings(
+        processor_class=_TOKENIZER_FOR_DOC,
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=SequenceClassifierOutput,
+        config_class=_CONFIG_FOR_DOC,
+    )
+    def forward(
+        self,
+        input_ids=None,
+        attention_mask=None,
+        token_type_ids=None,
+        position_ids=None,
+        head_mask=None,
+        inputs_embeds=None,
+        labels=None,
+        output_attentions=None,
+        output_hidden_states=None,
+        return_dict=None,
+    ):
+        r"""
+        labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
+            Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
+            config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
+            `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
+        """
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        outputs = self.roberta(
+            input_ids,
+            attention_mask=attention_mask,
+            token_type_ids=token_type_ids,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+        sequence_output = outputs[0]
+        logits = self.classifier(sequence_output)
+
+        loss = None
+        if labels is not None:
+            if self.config.problem_type is None:
+                if self.num_labels == 1:
+                    self.config.problem_type = "regression"
+                elif self.num_labels > 1 and (labels.dtype == torch.long or labels.dtype == torch.int):
+                    self.config.problem_type = "single_label_classification"
+                else:
+                    self.config.problem_type = "multi_label_classification"
+
+            if self.config.problem_type == "regression":
+                loss_fct = MSELoss()
+                if self.num_labels == 1:
+                    loss = loss_fct(logits.squeeze(), labels.squeeze())
+                else:
+                    loss = loss_fct(logits, labels)
+            elif self.config.problem_type == "single_label_classification":
+                loss_fct = CrossEntropyLoss()
+                loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))
+            elif self.config.problem_type == "multi_label_classification":
+                loss_fct = BCEWithLogitsLoss()
+                loss = loss_fct(logits, labels)
+
+        if not return_dict:
+            output = (logits,) + outputs[2:]
+            return ((loss,) + output) if loss is not None else output
+
+        return SequenceClassifierOutput(
+            loss=loss,
+            logits=logits,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )
+
+
+@add_start_docstrings(
+    """
+    XLM-Roberta-xlarge Model with a multiple choice classification head on top (a linear layer on top of the pooled
+    output and a softmax) e.g. for RocStories/SWAG tasks.
+    """,
+    XLM_ROBERTA_XL_START_DOCSTRING,
+)
+class XLMRobertaXLForMultipleChoice(XLMRobertaXLPreTrainedModel):
+    _keys_to_ignore_on_load_missing = [r"position_ids"]
+
+    def __init__(self, config):
+        super().__init__(config)
+
+        self.roberta = XLMRobertaXLModel(config)
+        self.dropout = nn.Dropout(config.hidden_dropout_prob)
+        self.classifier = nn.Linear(config.hidden_size, 1)
+
+        self.init_weights()
+
+    @add_start_docstrings_to_model_forward(
+        XLM_ROBERTA_XL_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length")
+    )
+    @add_code_sample_docstrings(
+        processor_class=_TOKENIZER_FOR_DOC,
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=MultipleChoiceModelOutput,
+        config_class=_CONFIG_FOR_DOC,
+    )
+    def forward(
+        self,
+        input_ids=None,
+        token_type_ids=None,
+        attention_mask=None,
+        labels=None,
+        position_ids=None,
+        head_mask=None,
+        inputs_embeds=None,
+        output_attentions=None,
+        output_hidden_states=None,
+        return_dict=None,
+    ):
+        r"""
+        labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
+            Labels for computing the multiple choice classification loss. Indices should be in `[0, ...,
+            num_choices-1]` where `num_choices` is the size of the second dimension of the input tensors. (See
+            `input_ids` above)
+        """
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+        num_choices = input_ids.shape[1] if input_ids is not None else inputs_embeds.shape[1]
+
+        flat_input_ids = input_ids.view(-1, input_ids.size(-1)) if input_ids is not None else None
+        flat_position_ids = position_ids.view(-1, position_ids.size(-1)) if position_ids is not None else None
+        flat_token_type_ids = token_type_ids.view(-1, token_type_ids.size(-1)) if token_type_ids is not None else None
+        flat_attention_mask = attention_mask.view(-1, attention_mask.size(-1)) if attention_mask is not None else None
+        flat_inputs_embeds = (
+            inputs_embeds.view(-1, inputs_embeds.size(-2), inputs_embeds.size(-1))
+            if inputs_embeds is not None
+            else None
+        )
+
+        outputs = self.roberta(
+            flat_input_ids,
+            position_ids=flat_position_ids,
+            token_type_ids=flat_token_type_ids,
+            attention_mask=flat_attention_mask,
+            head_mask=head_mask,
+            inputs_embeds=flat_inputs_embeds,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+        pooled_output = outputs[1]
+
+        pooled_output = self.dropout(pooled_output)
+        logits = self.classifier(pooled_output)
+        reshaped_logits = logits.view(-1, num_choices)
+
+        loss = None
+        if labels is not None:
+            loss_fct = CrossEntropyLoss()
+            loss = loss_fct(reshaped_logits, labels)
+
+        if not return_dict:
+            output = (reshaped_logits,) + outputs[2:]
+            return ((loss,) + output) if loss is not None else output
+
+        return MultipleChoiceModelOutput(
+            loss=loss,
+            logits=reshaped_logits,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )
+
+
+@add_start_docstrings(
+    """
+    XLM-Roberta-xlarge Model with a token classification head on top (a linear layer on top of the hidden-states
+    output) e.g. for Named-Entity-Recognition (NER) tasks.
+    """,
+    XLM_ROBERTA_XL_START_DOCSTRING,
+)
+class XLMRobertaXLForTokenClassification(XLMRobertaXLPreTrainedModel):
+    _keys_to_ignore_on_load_unexpected = [r"pooler"]
+    _keys_to_ignore_on_load_missing = [r"position_ids"]
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.num_labels = config.num_labels
+
+        self.roberta = XLMRobertaXLModel(config, add_pooling_layer=False)
+        classifier_dropout = (
+            config.classifier_dropout if config.classifier_dropout is not None else config.hidden_dropout_prob
+        )
+        self.dropout = nn.Dropout(classifier_dropout)
+        self.classifier = nn.Linear(config.hidden_size, config.num_labels)
+
+        self.init_weights()
+
+    @add_start_docstrings_to_model_forward(XLM_ROBERTA_XL_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
+    @add_code_sample_docstrings(
+        processor_class=_TOKENIZER_FOR_DOC,
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=TokenClassifierOutput,
+        config_class=_CONFIG_FOR_DOC,
+    )
+    def forward(
+        self,
+        input_ids=None,
+        attention_mask=None,
+        token_type_ids=None,
+        position_ids=None,
+        head_mask=None,
+        inputs_embeds=None,
+        labels=None,
+        output_attentions=None,
+        output_hidden_states=None,
+        return_dict=None,
+    ):
+        r"""
+        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Labels for computing the token classification loss. Indices should be in `[0, ..., config.num_labels - 1]`.
+        """
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        outputs = self.roberta(
+            input_ids,
+            attention_mask=attention_mask,
+            token_type_ids=token_type_ids,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+
+        sequence_output = outputs[0]
+
+        sequence_output = self.dropout(sequence_output)
+        logits = self.classifier(sequence_output)
+
+        loss = None
+        if labels is not None:
+            loss_fct = CrossEntropyLoss()
+            # Only keep active parts of the loss
+            if attention_mask is not None:
+                active_loss = attention_mask.view(-1) == 1
+                active_logits = logits.view(-1, self.num_labels)
+                active_labels = torch.where(
+                    active_loss, labels.view(-1), torch.tensor(loss_fct.ignore_index).type_as(labels)
+                )
+                loss = loss_fct(active_logits, active_labels)
+            else:
+                loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))
+
+        if not return_dict:
+            output = (logits,) + outputs[2:]
+            return ((loss,) + output) if loss is not None else output
+
+        return TokenClassifierOutput(
+            loss=loss,
+            logits=logits,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )
+
+
+class XLMRobertaXLClassificationHead(nn.Module):
+    """Head for sentence-level classification tasks."""
+
+    def __init__(self, config):
+        super().__init__()
+        self.dense = nn.Linear(config.hidden_size, config.hidden_size)
+        classifier_dropout = (
+            config.classifier_dropout if config.classifier_dropout is not None else config.hidden_dropout_prob
+        )
+        self.dropout = nn.Dropout(classifier_dropout)
+        self.out_proj = nn.Linear(config.hidden_size, config.num_labels)
+
+    def forward(self, features, **kwargs):
+        x = features[:, 0, :]  # take <s> token (equiv. to [CLS])
+        x = self.dropout(x)
+        x = self.dense(x)
+        x = torch.tanh(x)
+        x = self.dropout(x)
+        x = self.out_proj(x)
+        return x
+
+
+@add_start_docstrings(
+    """
+    XLM-Roberta-xlarge Model with a span classification head on top for extractive question-answering tasks like SQuAD
+    (a linear layers on top of the hidden-states output to compute `span start logits` and `span end logits`).
+    """,
+    XLM_ROBERTA_XL_START_DOCSTRING,
+)
+class XLMRobertaXLForQuestionAnswering(XLMRobertaXLPreTrainedModel):
+    _keys_to_ignore_on_load_unexpected = [r"pooler"]
+    _keys_to_ignore_on_load_missing = [r"position_ids"]
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.num_labels = config.num_labels
+
+        self.roberta = XLMRobertaXLModel(config, add_pooling_layer=False)
+        self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels)
+
+        self.init_weights()
+
+    @add_start_docstrings_to_model_forward(XLM_ROBERTA_XL_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
+    @add_code_sample_docstrings(
+        processor_class=_TOKENIZER_FOR_DOC,
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=QuestionAnsweringModelOutput,
+        config_class=_CONFIG_FOR_DOC,
+    )
+    def forward(
+        self,
+        input_ids=None,
+        attention_mask=None,
+        token_type_ids=None,
+        position_ids=None,
+        head_mask=None,
+        inputs_embeds=None,
+        start_positions=None,
+        end_positions=None,
+        output_attentions=None,
+        output_hidden_states=None,
+        return_dict=None,
+    ):
+        r"""
+        start_positions (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
+            Labels for position (index) of the start of the labelled span for computing the token classification loss.
+            Positions are clamped to the length of the sequence (`sequence_length`). Position outside of the sequence
+            are not taken into account for computing the loss.
+        end_positions (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
+            Labels for position (index) of the end of the labelled span for computing the token classification loss.
+            Positions are clamped to the length of the sequence (`sequence_length`). Position outside of the sequence
+            are not taken into account for computing the loss.
+        """
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        outputs = self.roberta(
+            input_ids,
+            attention_mask=attention_mask,
+            token_type_ids=token_type_ids,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+
+        sequence_output = outputs[0]
+
+        logits = self.qa_outputs(sequence_output)
+        start_logits, end_logits = logits.split(1, dim=-1)
+        start_logits = start_logits.squeeze(-1).contiguous()
+        end_logits = end_logits.squeeze(-1).contiguous()
+
+        total_loss = None
+        if start_positions is not None and end_positions is not None:
+            # If we are on multi-GPU, split add a dimension
+            if len(start_positions.size()) > 1:
+                start_positions = start_positions.squeeze(-1)
+            if len(end_positions.size()) > 1:
+                end_positions = end_positions.squeeze(-1)
+            # sometimes the start/end positions are outside our model inputs, we ignore these terms
+            ignored_index = start_logits.size(1)
+            start_positions = start_positions.clamp(0, ignored_index)
+            end_positions = end_positions.clamp(0, ignored_index)
+
+            loss_fct = CrossEntropyLoss(ignore_index=ignored_index)
+            start_loss = loss_fct(start_logits, start_positions)
+            end_loss = loss_fct(end_logits, end_positions)
+            total_loss = (start_loss + end_loss) / 2
+
+        if not return_dict:
+            output = (start_logits, end_logits) + outputs[2:]
+            return ((total_loss,) + output) if total_loss is not None else output
+
+        return QuestionAnsweringModelOutput(
+            loss=total_loss,
+            start_logits=start_logits,
+            end_logits=end_logits,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )
+
+
+# Copied from transformers.models.roberta.modeling_roberta.create_position_ids_from_input_ids
+def create_position_ids_from_input_ids(input_ids, padding_idx, past_key_values_length=0):
+    """
+    Replace non-padding symbols with their position numbers. Position numbers begin at padding_idx+1. Padding symbols
+    are ignored. This is modified from fairseq's `utils.make_positions`.
+
+    Args:
+        x: torch.Tensor x:
+
+    Returns: torch.Tensor
+    """
+    # The series of casts and type-conversions here are carefully balanced to both work with ONNX export and XLA.
+    mask = input_ids.ne(padding_idx).int()
+    incremental_indices = (torch.cumsum(mask, dim=1).type_as(mask) + past_key_values_length) * mask
+    return incremental_indices.long() + padding_idx

src/transformers/utils/dummy_pt_objects.py

@@ -4021,6 +4021,65 @@ class XLMRobertaModel(metaclass=DummyObject):
         requires_backends(self, ["torch"])
 
 
+XLM_ROBERTA_XL_PRETRAINED_MODEL_ARCHIVE_LIST = None
+
+
+class XLMRobertaXLForCausalLM(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
+class XLMRobertaXLForMaskedLM(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
+class XLMRobertaXLForMultipleChoice(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
+class XLMRobertaXLForQuestionAnswering(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
+class XLMRobertaXLForSequenceClassification(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
+class XLMRobertaXLForTokenClassification(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
+class XLMRobertaXLModel(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
+class XLMRobertaXLPreTrainedModel(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
 XLNET_PRETRAINED_MODEL_ARCHIVE_LIST = None
 
 

tests/test_modeling_xlm_roberta_xl.py

+# coding=utf-8
+# 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.
+
+
+import unittest
+
+from transformers import XLMRobertaXLConfig, is_torch_available
+from transformers.testing_utils import require_torch, slow, torch_device
+
+from .test_configuration_common import ConfigTester
+from .test_generation_utils import GenerationTesterMixin
+from .test_modeling_common import ModelTesterMixin, floats_tensor, ids_tensor, random_attention_mask
+
+
+if is_torch_available():
+    import torch
+
+    from transformers import (
+        XLMRobertaXLForCausalLM,
+        XLMRobertaXLForMaskedLM,
+        XLMRobertaXLForMultipleChoice,
+        XLMRobertaXLForQuestionAnswering,
+        XLMRobertaXLForSequenceClassification,
+        XLMRobertaXLForTokenClassification,
+        XLMRobertaXLModel,
+    )
+    from transformers.models.xlm_roberta_xl.modeling_xlm_roberta_xl import (
+        XLMRobertaXLEmbeddings,
+        create_position_ids_from_input_ids,
+    )
+
+
+class XLMRobertaXLModelTester:
+    def __init__(
+        self,
+        parent,
+    ):
+        self.parent = parent
+        self.batch_size = 13
+        self.seq_length = 7
+        self.is_training = True
+        self.use_input_mask = True
+        self.use_token_type_ids = True
+        self.use_labels = True
+        self.vocab_size = 99
+        self.hidden_size = 32
+        self.num_hidden_layers = 5
+        self.num_attention_heads = 4
+        self.intermediate_size = 37
+        self.hidden_act = "gelu"
+        self.hidden_dropout_prob = 0.1
+        self.attention_probs_dropout_prob = 0.1
+        self.max_position_embeddings = 512
+        self.type_vocab_size = 16
+        self.type_sequence_label_size = 2
+        self.initializer_range = 0.02
+        self.num_labels = 3
+        self.num_choices = 4
+        self.scope = None
+
+    def prepare_config_and_inputs(self):
+        input_ids = ids_tensor([self.batch_size, self.seq_length], self.vocab_size)
+
+        input_mask = None
+        if self.use_input_mask:
+            input_mask = random_attention_mask([self.batch_size, self.seq_length])
+
+        token_type_ids = None
+        if self.use_token_type_ids:
+            token_type_ids = ids_tensor([self.batch_size, self.seq_length], self.type_vocab_size)
+
+        sequence_labels = None
+        token_labels = None
+        choice_labels = None
+        if self.use_labels:
+            sequence_labels = ids_tensor([self.batch_size], self.type_sequence_label_size)
+            token_labels = ids_tensor([self.batch_size, self.seq_length], self.num_labels)
+            choice_labels = ids_tensor([self.batch_size], self.num_choices)
+
+        config = self.get_config()
+
+        return config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels
+
+    def get_config(self):
+        return XLMRobertaXLConfig(
+            vocab_size=self.vocab_size,
+            hidden_size=self.hidden_size,
+            num_hidden_layers=self.num_hidden_layers,
+            num_attention_heads=self.num_attention_heads,
+            intermediate_size=self.intermediate_size,
+            hidden_act=self.hidden_act,
+            hidden_dropout_prob=self.hidden_dropout_prob,
+            attention_probs_dropout_prob=self.attention_probs_dropout_prob,
+            max_position_embeddings=self.max_position_embeddings,
+            type_vocab_size=self.type_vocab_size,
+            initializer_range=self.initializer_range,
+        )
+
+    def prepare_config_and_inputs_for_decoder(self):
+        (
+            config,
+            input_ids,
+            token_type_ids,
+            input_mask,
+            sequence_labels,
+            token_labels,
+            choice_labels,
+        ) = self.prepare_config_and_inputs()
+
+        config.is_decoder = True
+        encoder_hidden_states = floats_tensor([self.batch_size, self.seq_length, self.hidden_size])
+        encoder_attention_mask = ids_tensor([self.batch_size, self.seq_length], vocab_size=2)
+
+        return (
+            config,
+            input_ids,
+            token_type_ids,
+            input_mask,
+            sequence_labels,
+            token_labels,
+            choice_labels,
+            encoder_hidden_states,
+            encoder_attention_mask,
+        )
+
+    def create_and_check_model(
+        self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels
+    ):
+        model = XLMRobertaXLModel(config=config)
+        model.to(torch_device)
+        model.eval()
+        result = model(input_ids, attention_mask=input_mask, token_type_ids=token_type_ids)
+        result = model(input_ids, token_type_ids=token_type_ids)
+        result = model(input_ids)
+
+        self.parent.assertEqual(result.last_hidden_state.shape, (self.batch_size, self.seq_length, self.hidden_size))
+        self.parent.assertEqual(result.pooler_output.shape, (self.batch_size, self.hidden_size))
+
+    def create_and_check_model_as_decoder(
+        self,
+        config,
+        input_ids,
+        token_type_ids,
+        input_mask,
+        sequence_labels,
+        token_labels,
+        choice_labels,
+        encoder_hidden_states,
+        encoder_attention_mask,
+    ):
+        config.add_cross_attention = True
+        model = XLMRobertaXLModel(config)
+        model.to(torch_device)
+        model.eval()
+        result = model(
+            input_ids,
+            attention_mask=input_mask,
+            token_type_ids=token_type_ids,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+        )
+        result = model(
+            input_ids,
+            attention_mask=input_mask,
+            token_type_ids=token_type_ids,
+            encoder_hidden_states=encoder_hidden_states,
+        )
+        result = model(input_ids, attention_mask=input_mask, token_type_ids=token_type_ids)
+        self.parent.assertEqual(result.last_hidden_state.shape, (self.batch_size, self.seq_length, self.hidden_size))
+        self.parent.assertEqual(result.pooler_output.shape, (self.batch_size, self.hidden_size))
+
+    def create_and_check_for_causal_lm(
+        self,
+        config,
+        input_ids,
+        token_type_ids,
+        input_mask,
+        sequence_labels,
+        token_labels,
+        choice_labels,
+        encoder_hidden_states,
+        encoder_attention_mask,
+    ):
+        model = XLMRobertaXLForCausalLM(config=config)
+        model.to(torch_device)
+        model.eval()
+        result = model(input_ids, attention_mask=input_mask, token_type_ids=token_type_ids, labels=token_labels)
+        self.parent.assertEqual(result.logits.shape, (self.batch_size, self.seq_length, self.vocab_size))
+
+    def create_and_check_decoder_model_past_large_inputs(
+        self,
+        config,
+        input_ids,
+        token_type_ids,
+        input_mask,
+        sequence_labels,
+        token_labels,
+        choice_labels,
+        encoder_hidden_states,
+        encoder_attention_mask,
+    ):
+        config.is_decoder = True
+        config.add_cross_attention = True
+        model = XLMRobertaXLForCausalLM(config=config).to(torch_device).eval()
+
+        # make sure that ids don't start with pad token
+        mask = input_ids.ne(config.pad_token_id).long()
+        input_ids = input_ids * mask
+
+        # first forward pass
+        outputs = model(
+            input_ids,
+            attention_mask=input_mask,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            use_cache=True,
+        )
+        past_key_values = outputs.past_key_values
+
+        # create hypothetical multiple next token and extent to next_input_ids
+        next_tokens = ids_tensor((self.batch_size, 3), config.vocab_size)
+
+        # make sure that ids don't start with pad token
+        mask = next_tokens.ne(config.pad_token_id).long()
+        next_tokens = next_tokens * mask
+        next_mask = ids_tensor((self.batch_size, 3), vocab_size=2)
+
+        # append to next input_ids and
+        next_input_ids = torch.cat([input_ids, next_tokens], dim=-1)
+        next_attention_mask = torch.cat([input_mask, next_mask], dim=-1)
+
+        output_from_no_past = model(
+            next_input_ids,
+            attention_mask=next_attention_mask,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            output_hidden_states=True,
+        )["hidden_states"][0]
+        output_from_past = model(
+            next_tokens,
+            attention_mask=next_attention_mask,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            past_key_values=past_key_values,
+            output_hidden_states=True,
+        )["hidden_states"][0]
+
+        # select random slice
+        random_slice_idx = ids_tensor((1,), output_from_past.shape[-1]).item()
+        output_from_no_past_slice = output_from_no_past[:, -3:, random_slice_idx].detach()
+        output_from_past_slice = output_from_past[:, :, random_slice_idx].detach()
+
+        self.parent.assertTrue(output_from_past_slice.shape[1] == next_tokens.shape[1])
+
+        # test that outputs are equal for slice
+        self.parent.assertTrue(torch.allclose(output_from_past_slice, output_from_no_past_slice, atol=1e-3))
+
+    def create_and_check_for_masked_lm(
+        self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels
+    ):
+        model = XLMRobertaXLForMaskedLM(config=config)
+        model.to(torch_device)
+        model.eval()
+        result = model(input_ids, attention_mask=input_mask, token_type_ids=token_type_ids, labels=token_labels)
+        self.parent.assertEqual(result.logits.shape, (self.batch_size, self.seq_length, self.vocab_size))
+
+    def create_and_check_for_token_classification(
+        self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels
+    ):
+        config.num_labels = self.num_labels
+        model = XLMRobertaXLForTokenClassification(config=config)
+        model.to(torch_device)
+        model.eval()
+        result = model(input_ids, attention_mask=input_mask, token_type_ids=token_type_ids, labels=token_labels)
+        self.parent.assertEqual(result.logits.shape, (self.batch_size, self.seq_length, self.num_labels))
+
+    def create_and_check_for_multiple_choice(
+        self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels
+    ):
+        config.num_choices = self.num_choices
+        model = XLMRobertaXLForMultipleChoice(config=config)
+        model.to(torch_device)
+        model.eval()
+        multiple_choice_inputs_ids = input_ids.unsqueeze(1).expand(-1, self.num_choices, -1).contiguous()
+        multiple_choice_token_type_ids = token_type_ids.unsqueeze(1).expand(-1, self.num_choices, -1).contiguous()
+        multiple_choice_input_mask = input_mask.unsqueeze(1).expand(-1, self.num_choices, -1).contiguous()
+        result = model(
+            multiple_choice_inputs_ids,
+            attention_mask=multiple_choice_input_mask,
+            token_type_ids=multiple_choice_token_type_ids,
+            labels=choice_labels,
+        )
+        self.parent.assertEqual(result.logits.shape, (self.batch_size, self.num_choices))
+
+    def create_and_check_for_question_answering(
+        self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels
+    ):
+        model = XLMRobertaXLForQuestionAnswering(config=config)
+        model.to(torch_device)
+        model.eval()
+        result = model(
+            input_ids,
+            attention_mask=input_mask,
+            token_type_ids=token_type_ids,
+            start_positions=sequence_labels,
+            end_positions=sequence_labels,
+        )
+        self.parent.assertEqual(result.start_logits.shape, (self.batch_size, self.seq_length))
+        self.parent.assertEqual(result.end_logits.shape, (self.batch_size, self.seq_length))
+
+    def prepare_config_and_inputs_for_common(self):
+        config_and_inputs = self.prepare_config_and_inputs()
+        (
+            config,
+            input_ids,
+            token_type_ids,
+            input_mask,
+            sequence_labels,
+            token_labels,
+            choice_labels,
+        ) = config_and_inputs
+        inputs_dict = {"input_ids": input_ids, "token_type_ids": token_type_ids, "attention_mask": input_mask}
+        return config, inputs_dict
+
+
+@require_torch
+class XLMRobertaXLModelTest(ModelTesterMixin, GenerationTesterMixin, unittest.TestCase):
+
+    all_model_classes = (
+        (
+            XLMRobertaXLForCausalLM,
+            XLMRobertaXLForMaskedLM,
+            XLMRobertaXLModel,
+            XLMRobertaXLForSequenceClassification,
+            XLMRobertaXLForTokenClassification,
+            XLMRobertaXLForMultipleChoice,
+            XLMRobertaXLForQuestionAnswering,
+        )
+        if is_torch_available()
+        else ()
+    )
+    all_generative_model_classes = (XLMRobertaXLForCausalLM,) if is_torch_available() else ()
+
+    def setUp(self):
+        self.model_tester = XLMRobertaXLModelTester(self)
+        self.config_tester = ConfigTester(self, config_class=XLMRobertaXLConfig, hidden_size=37)
+
+    def test_config(self):
+        self.config_tester.run_common_tests()
+
+    def test_model(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_model(*config_and_inputs)
+
+    def test_model_various_embeddings(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        for type in ["absolute", "relative_key", "relative_key_query"]:
+            config_and_inputs[0].position_embedding_type = type
+            self.model_tester.create_and_check_model(*config_and_inputs)
+
+    def test_model_as_decoder(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs_for_decoder()
+        self.model_tester.create_and_check_model_as_decoder(*config_and_inputs)
+
+    def test_model_as_decoder_with_default_input_mask(self):
+        # This regression test was failing with PyTorch < 1.3
+        (
+            config,
+            input_ids,
+            token_type_ids,
+            input_mask,
+            sequence_labels,
+            token_labels,
+            choice_labels,
+            encoder_hidden_states,
+            encoder_attention_mask,
+        ) = self.model_tester.prepare_config_and_inputs_for_decoder()
+
+        input_mask = None
+
+        self.model_tester.create_and_check_model_as_decoder(
+            config,
+            input_ids,
+            token_type_ids,
+            input_mask,
+            sequence_labels,
+            token_labels,
+            choice_labels,
+            encoder_hidden_states,
+            encoder_attention_mask,
+        )
+
+    def test_for_causal_lm(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs_for_decoder()
+        self.model_tester.create_and_check_for_causal_lm(*config_and_inputs)
+
+    def test_decoder_model_past_with_large_inputs(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs_for_decoder()
+        self.model_tester.create_and_check_decoder_model_past_large_inputs(*config_and_inputs)
+
+    def test_for_masked_lm(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_for_masked_lm(*config_and_inputs)
+
+    def test_for_token_classification(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_for_token_classification(*config_and_inputs)
+
+    def test_for_multiple_choice(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_for_multiple_choice(*config_and_inputs)
+
+    def test_for_question_answering(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_for_question_answering(*config_and_inputs)
+
+    def test_create_position_ids_respects_padding_index(self):
+        """Ensure that the default position ids only assign a sequential . This is a regression
+        test for https://github.com/huggingface/transformers/issues/1761
+
+        The position ids should be masked with the embedding object's padding index. Therefore, the
+        first available non-padding position index is XLMRobertaXLEmbeddings.padding_idx + 1
+        """
+        config = self.model_tester.prepare_config_and_inputs()[0]
+        model = XLMRobertaXLEmbeddings(config=config)
+
+        input_ids = torch.as_tensor([[12, 31, 13, model.padding_idx]])
+        expected_positions = torch.as_tensor(
+            [[0 + model.padding_idx + 1, 1 + model.padding_idx + 1, 2 + model.padding_idx + 1, model.padding_idx]]
+        )
+
+        position_ids = create_position_ids_from_input_ids(input_ids, model.padding_idx)
+        self.assertEqual(position_ids.shape, expected_positions.shape)
+        self.assertTrue(torch.all(torch.eq(position_ids, expected_positions)))
+
+    def test_create_position_ids_from_inputs_embeds(self):
+        """Ensure that the default position ids only assign a sequential . This is a regression
+        test for https://github.com/huggingface/transformers/issues/1761
+
+        The position ids should be masked with the embedding object's padding index. Therefore, the
+        first available non-padding position index is XLMRobertaXLEmbeddings.padding_idx + 1
+        """
+        config = self.model_tester.prepare_config_and_inputs()[0]
+        embeddings = XLMRobertaXLEmbeddings(config=config)
+
+        inputs_embeds = torch.empty(2, 4, 30)
+        expected_single_positions = [
+            0 + embeddings.padding_idx + 1,
+            1 + embeddings.padding_idx + 1,
+            2 + embeddings.padding_idx + 1,
+            3 + embeddings.padding_idx + 1,
+        ]
+        expected_positions = torch.as_tensor([expected_single_positions, expected_single_positions])
+        position_ids = embeddings.create_position_ids_from_inputs_embeds(inputs_embeds)
+        self.assertEqual(position_ids.shape, expected_positions.shape)
+        self.assertTrue(torch.all(torch.eq(position_ids, expected_positions)))
+
+
+@require_torch
+class XLMRobertaModelXLIntegrationTest(unittest.TestCase):
+    @slow
+    def test_xlm_roberta_xl(self):
+        model = XLMRobertaXLModel.from_pretrained("facebook/xlm-roberta-xl").to(torch_device)
+        input_ids = torch.tensor(
+            [[0, 581, 10269, 83, 99942, 136, 60742, 23, 70, 80583, 18276, 2]], device=torch_device
+        )
+        # The dog is cute and lives in the garden house
+
+        expected_output_shape = torch.Size((1, 12, 2560))  # batch_size, sequence_length, embedding_vector_dim
+        expected_output_values_last_dim = torch.tensor(
+            [[0.0110, 0.0605, 0.0354, 0.0689, 0.0066, 0.0691, 0.0302, 0.0412, 0.0860, 0.0036, 0.0405, 0.0170]],
+            device=torch_device,
+        )
+
+        output = model(input_ids)["last_hidden_state"].detach()
+        self.assertEqual(output.shape, expected_output_shape)
+        # compare the actual values for a slice of last dim
+        self.assertTrue(torch.allclose(output[:, :, -1], expected_output_values_last_dim, atol=1e-3))
+
+    @unittest.skip(reason="Model is too large to be tested on the CI")
+    def test_xlm_roberta_xxl(self):
+        model = XLMRobertaXLModel.from_pretrained("facebook/xlm-roberta-xxl").to(torch_device)
+        input_ids = torch.tensor(
+            [[0, 581, 10269, 83, 99942, 136, 60742, 23, 70, 80583, 18276, 2]], device=torch_device
+        )
+        # The dog is cute and lives in the garden house
+
+        expected_output_shape = torch.Size((1, 12, 4096))  # batch_size, sequence_length, embedding_vector_dim
+        expected_output_values_last_dim = torch.tensor(
+            [[0.0046, 0.0146, 0.0227, 0.0126, 0.0219, 0.0175, -0.0101, 0.0006, 0.0124, 0.0209, -0.0063, 0.0096]],
+            device=torch_device,
+        )
+
+        output = model(input_ids)["last_hidden_state"].detach()
+        self.assertEqual(output.shape, expected_output_shape)
+        # compare the actual values for a slice of last dim
+        self.assertTrue(torch.allclose(output[:, :, -1], expected_output_values_last_dim, atol=1e-3))