Release: v4.0.0

* Remove deprecated `evalutate_during_training`

* Update src/transformers/training_args_tf.py

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

.circleci/config.yml

@@ -77,7 +77,7 @@ jobs:
                       - v0.4-torch_and_tf-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[sklearn,tf-cpu,torch,testing]
+            - run: pip install .[sklearn,tf-cpu,torch,testing,sentencepiece]
             - save_cache:
                 key: v0.4-{{ checksum "setup.py" }}
                 paths:
@@ -103,7 +103,7 @@ jobs:
                       - v0.4-torch-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[sklearn,torch,testing]
+            - run: pip install .[sklearn,torch,testing,sentencepiece]
             - save_cache:
                   key: v0.4-torch-{{ checksum "setup.py" }}
                   paths:
@@ -129,7 +129,7 @@ jobs:
                       - v0.4-tf-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[sklearn,tf-cpu,testing]
+            - run: pip install .[sklearn,tf-cpu,testing,sentencepiece]
             - save_cache:
                   key: v0.4-tf-{{ checksum "setup.py" }}
                   paths:
@@ -155,7 +155,7 @@ jobs:
                     - v0.4-flax-{{ checksum "setup.py" }}
                     - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: sudo pip install .[flax,sklearn,torch,testing]
+            - run: sudo pip install .[flax,sklearn,torch,testing,sentencepiece]
             - save_cache:
                   key: v0.4-flax-{{ checksum "setup.py" }}
                   paths:
@@ -181,7 +181,7 @@ jobs:
                       - v0.4-torch-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[sklearn,torch,testing]
+            - run: pip install .[sklearn,torch,testing,sentencepiece]
             - save_cache:
                   key: v0.4-torch-{{ checksum "setup.py" }}
                   paths:
@@ -207,7 +207,7 @@ jobs:
                       - v0.4-tf-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[sklearn,tf-cpu,testing]
+            - run: pip install .[sklearn,tf-cpu,testing,sentencepiece]
             - save_cache:
                   key: v0.4-tf-{{ checksum "setup.py" }}
                   paths:
@@ -231,7 +231,7 @@ jobs:
                       - v0.4-custom_tokenizers-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[ja,testing]
+            - run: pip install .[ja,testing,sentencepiece]
             - run: python -m unidic download
             - save_cache:
                   key: v0.4-custom_tokenizers-{{ checksum "setup.py" }}
@@ -258,7 +258,7 @@ jobs:
                       - v0.4-torch_examples-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[sklearn,torch,testing]
+            - run: pip install .[sklearn,torch,sentencepiece,testing]
             - run: pip install -r examples/requirements.txt
             - save_cache:
                   key: v0.4-torch_examples-{{ checksum "setup.py" }}
@@ -281,7 +281,7 @@ jobs:
                       - v0.4-build_doc-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[tf,torch,sentencepiece,docs]
+            - run: pip install ."[all, docs]"
             - save_cache:
                   key: v0.4-build_doc-{{ checksum "setup.py" }}
                   paths:
@@ -303,7 +303,7 @@ jobs:
                   keys:
                       - v0.4-deploy_doc-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
-            - run: pip install .[tf,torch,sentencepiece,docs]
+            - run: pip install ."[all,docs]"
             - save_cache:
                   key: v0.4-deploy_doc-{{ checksum "setup.py" }}
                   paths:
@@ -324,7 +324,7 @@ jobs:
                       - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
             - run: pip install isort
-            - run: pip install .[tf,torch,flax,quality]
+            - run: pip install .[all,quality]
             - save_cache:
                   key: v0.4-code_quality-{{ checksum "setup.py" }}
                   paths:

.circleci/deploy.sh

@@ -51,4 +51,5 @@ deploy_doc "7fb8bdf" v3.0.2
 deploy_doc "4b3ee9c" v3.1.0
 deploy_doc "3ebb1b3" v3.2.0
 deploy_doc "0613f05" v3.3.1
-deploy_doc "eb0e0ce" # v3.4.0 Latest stable release
+deploy_doc "eb0e0ce" v3.4.0
+deploy_doc "818878d" # v3.5.1 Latest stable release

.github/ISSUE_TEMPLATE/question-help.md

@@ -1,6 +1,6 @@
 ---
 name: "❓ Questions & Help"
-about: Post your general questions on the Hugging Face forum or Stack Overflow tagged huggingface-transformers
+about: Post your general questions on the Hugging Face forum: https://discuss.huggingface.co/
 title: ''
 labels: ''
 assignees: ''
@@ -10,18 +10,17 @@ assignees: ''
 # ❓ Questions & Help
 
 <!-- The GitHub issue tracker is primarly intended for bugs, feature requests,
-     new models and benchmarks, and migration questions. For all other questions,
+     new models, benchmarks, and migration questions. For all other questions,
      we direct you to the Hugging Face forum: https://discuss.huggingface.co/ .
-     You can also try Stack Overflow (SO) where a whole community of PyTorch and
-     Tensorflow enthusiast can help you out. In this case, make sure to tag your
-     question with the right deep learning framework as well as the
-     huggingface-transformers tag: 
-     https://stackoverflow.com/questions/tagged/huggingface-transformers 
      -->
 
 ## Details
+
 <!-- Description of your issue -->
 
-<!-- You should first ask your question on the forum or SO, and only if
-     you didn't get an answer ask it here on GitHub. -->
-**A link to original question on the forum/Stack Overflow**:
+<!-- You should first ask your question on the forum, and only if
+     you didn't get an answer after a few days ask it here on GitHub. -->
+
+**A link to original question on the forum**:
+
+<!-- Your issue will be closed if you don't fill this part. -->

.github/PULL_REQUEST_TEMPLATE.md

@@ -20,7 +20,7 @@ Fixes # (issue)
 - [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md#start-contributing-pull-requests),
       Pull Request section?
 - [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
-      to the it if that's the case.
+      to it if that's the case.
 - [ ] Did you make sure to update the documentation with your changes? Here are the
       [documentation guidelines](https://github.com/huggingface/transformers/tree/master/docs), and
       [here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/master/docs#writing-source-documentation).

.github/workflows/github-torch-hub.yml

@@ -8,6 +8,9 @@ on:
 jobs:
   torch_hub_integration:
     runs-on: ubuntu-latest
+    env:
+      # TODO quickfix but may need more investigation
+      ACTIONS_ALLOW_UNSECURE_COMMANDS: True
     steps:
     # no checkout necessary here.
     - name: Extract branch name

.github/workflows/self-push.yml

@@ -4,17 +4,19 @@ on:
   push:
     branches:
       - master
+      - model-templates
     paths:
       - "src/**"
       - "tests/**"
       - ".github/**"
+      - "templates/**"
   # pull_request:
   repository_dispatch:
 
 
 jobs:
   run_tests_torch_gpu:
-    runs-on: [self-hosted, single-gpu]
+    runs-on: [self-hosted, gpu, single-gpu]
     steps:
       - uses: actions/checkout@v2
       - name: Python version
@@ -46,7 +48,7 @@ jobs:
         run: |
           source .env/bin/activate
           pip install --upgrade pip
-          pip install .[torch,sklearn,testing,onnxruntime]
+          pip install .[torch,sklearn,testing,onnxruntime,sentencepiece]
           pip install git+https://github.com/huggingface/datasets
 
       - name: Are GPUs recognized by our DL frameworks
@@ -55,6 +57,14 @@ jobs:
           python -c "import torch; print('Cuda available:', torch.cuda.is_available())"
           python -c "import torch; print('Number of GPUs available:', torch.cuda.device_count())"
 
+      - name: Create model files
+        run: |
+          source .env/bin/activate
+          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/encoder-bert-tokenizer.json --path=templates/adding_a_new_model
+          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/pt-encoder-bert-tokenizer.json --path=templates/adding_a_new_model
+          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/standalone.json --path=templates/adding_a_new_model
+          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/tf-encoder-bert-tokenizer.json --path=templates/adding_a_new_model
+
       - name: Run all non-slow tests on GPU
         env:
           OMP_NUM_THREADS: 1
@@ -76,7 +86,7 @@ jobs:
                   
 
   run_tests_tf_gpu:
-    runs-on: [self-hosted, single-gpu]
+    runs-on: [self-hosted, gpu, single-gpu]
     steps:
       - uses: actions/checkout@v2
       - name: Python version
@@ -107,7 +117,7 @@ jobs:
         run: |
           source .env/bin/activate
           pip install --upgrade pip
-          pip install .[tf,sklearn,testing,onnxruntime]
+          pip install .[tf,sklearn,testing,onnxruntime,sentencepiece]
           pip install git+https://github.com/huggingface/datasets
 
       - name: Are GPUs recognized by our DL frameworks
@@ -116,6 +126,14 @@ jobs:
           TF_CPP_MIN_LOG_LEVEL=3 python -c "import tensorflow as tf; print('TF GPUs available:', bool(tf.config.list_physical_devices('GPU')))"
           TF_CPP_MIN_LOG_LEVEL=3 python -c "import tensorflow as tf; print('Number of TF GPUs available:', len(tf.config.list_physical_devices('GPU')))"
 
+      - name: Create model files
+        run: |
+          source .env/bin/activate
+          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/encoder-bert-tokenizer.json --path=templates/adding_a_new_model
+          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/pt-encoder-bert-tokenizer.json --path=templates/adding_a_new_model
+          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/standalone.json --path=templates/adding_a_new_model
+          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/tf-encoder-bert-tokenizer.json --path=templates/adding_a_new_model
+
       - name: Run all non-slow tests on GPU
         env:
           OMP_NUM_THREADS: 1
@@ -135,8 +153,8 @@ jobs:
           name: run_all_tests_tf_gpu_test_reports
           path: reports
 
-  run_tests_torch_multiple_gpu:
-    runs-on: [self-hosted, multi-gpu]
+  run_tests_torch_multi_gpu:
+    runs-on: [self-hosted, gpu, multi-gpu]
     steps:
       - uses: actions/checkout@v2
       - name: Python version
@@ -154,7 +172,7 @@ jobs:
         id: cache
         with:
           path: .env
-          key: v1.1-tests_torch_multiple_gpu-${{ hashFiles('setup.py') }}
+          key: v1.1-tests_torch_multi_gpu-${{ hashFiles('setup.py') }}
 
       - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
         run: |
@@ -167,7 +185,7 @@ jobs:
         run: |
           source .env/bin/activate
           pip install --upgrade pip
-          pip install .[torch,sklearn,testing,onnxruntime]
+          pip install .[torch,sklearn,testing,onnxruntime,sentencepiece]
           pip install git+https://github.com/huggingface/datasets
 
       - name: Are GPUs recognized by our DL frameworks
@@ -181,11 +199,11 @@ jobs:
           OMP_NUM_THREADS: 1
         run: |
           source .env/bin/activate
-          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_torch_multiple_gpu tests
+          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_torch_multi_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
-        run: cat reports/tests_torch_multiple_gpu_failures_short.txt          
+        run: cat reports/tests_torch_multi_gpu_failures_short.txt          
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
@@ -194,8 +212,8 @@ jobs:
           name: run_all_tests_torch_multi_gpu_test_reports
           path: reports
 
-  run_tests_tf_multiple_gpu:
-    runs-on: [self-hosted, multi-gpu]
+  run_tests_tf_multi_gpu:
+    runs-on: [self-hosted, gpu, multi-gpu]
     steps:
       - uses: actions/checkout@v2
       - name: Python version
@@ -213,7 +231,7 @@ jobs:
         id: cache
         with:
           path: .env
-          key: v1.1-tests_tf_multiple_gpu-${{ hashFiles('setup.py') }}
+          key: v1.1-tests_tf_multi_gpu-${{ hashFiles('setup.py') }}
 
       - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
         run: |
@@ -226,7 +244,7 @@ jobs:
         run: |
           source .env/bin/activate
           pip install --upgrade pip
-          pip install .[tf,sklearn,testing,onnxruntime]
+          pip install .[tf,sklearn,testing,onnxruntime,sentencepiece]
           pip install git+https://github.com/huggingface/datasets
 
       - name: Are GPUs recognized by our DL frameworks
@@ -240,11 +258,11 @@ jobs:
           OMP_NUM_THREADS: 1
         run: |
           source .env/bin/activate
-          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_tf_multiple_gpu tests
+          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_tf_multi_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
-        run: cat reports/tests_tf_multiple_gpu_failures_short.txt
+        run: cat reports/tests_tf_multi_gpu_failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}

.github/workflows/self-scheduled.yml

@@ -15,7 +15,7 @@ on:
 
 jobs:
   run_all_tests_torch_gpu:
-    runs-on: [self-hosted, single-gpu]
+    runs-on: [self-hosted, gpu, single-gpu]
     steps:
       - uses: actions/checkout@v2
 
@@ -49,7 +49,7 @@ jobs:
         run: |
           source .env/bin/activate
           pip install --upgrade pip
-          pip install .[torch,sklearn,testing,onnxruntime]
+          pip install .[torch,sklearn,testing,onnxruntime,sentencepiece]
           pip install git+https://github.com/huggingface/datasets
           pip list
 
@@ -109,7 +109,7 @@ jobs:
 
 
   run_all_tests_tf_gpu:
-    runs-on: [self-hosted, single-gpu]
+    runs-on: [self-hosted, gpu, single-gpu]
     steps:
       - uses: actions/checkout@v2
 
@@ -143,7 +143,7 @@ jobs:
         run: |
           source .env/bin/activate
           pip install --upgrade pip
-          pip install .[tf,sklearn,testing,onnxruntime]
+          pip install .[tf,sklearn,testing,onnxruntime,sentencepiece]
           pip install git+https://github.com/huggingface/datasets
           pip list
 
@@ -187,8 +187,8 @@ jobs:
           name: run_all_tests_tf_gpu_test_reports
           path: reports
           
-  run_all_tests_torch_multiple_gpu:
-    runs-on: [self-hosted, multi-gpu]
+  run_all_tests_torch_multi_gpu:
+    runs-on: [self-hosted, gpu, multi-gpu]
     steps:
       - uses: actions/checkout@v2
 
@@ -222,7 +222,7 @@ jobs:
         run: |
           source .env/bin/activate
           pip install --upgrade pip
-          pip install .[torch,sklearn,testing,onnxruntime]
+          pip install .[torch,sklearn,testing,onnxruntime,sentencepiece]
           pip install git+https://github.com/huggingface/datasets
           pip list
 
@@ -238,11 +238,11 @@ jobs:
           RUN_SLOW: yes
         run: |
           source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_torch_multiple_gpu tests
+          python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_torch_multi_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
-        run: cat reports/tests_torch_multiple_gpu_failures_short.txt
+        run: cat reports/tests_torch_multi_gpu_failures_short.txt
 
       - name: Run examples tests on multi-GPU
         env:
@@ -250,11 +250,11 @@ jobs:
           RUN_SLOW: yes
         run: |
           source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s --make-reports=examples_torch_multiple_gpu examples
+          python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_torch_examples_multi_gpu examples
 
       - name: Failure short reports
         if: ${{ always() }}
-        run: cat reports/examples_torch_multiple_gpu_failures_short.txt
+        run: cat reports/tests_torch_examples_multi_gpu_failures_short.txt
 
       - name: Run all pipeline tests on multi-GPU
         if: ${{ always() }}
@@ -265,11 +265,11 @@ jobs:
           RUN_PIPELINE_TESTS: yes
         run: |
           source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s -m is_pipeline_test --make-reports=tests_torch_pipeline_multiple_gpu tests
+          python -m pytest -n 1 --dist=loadfile -s -m is_pipeline_test --make-reports=tests_torch_pipeline_multi_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
-        run: cat reports/tests_torch_pipeline_multiple_gpu_failures_short.txt
+        run: cat reports/tests_torch_pipeline_multi_gpu_failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
@@ -278,8 +278,8 @@ jobs:
           name: run_all_tests_torch_multi_gpu_test_reports
           path: reports
 
-  run_all_tests_tf_multiple_gpu:
-    runs-on: [self-hosted, multi-gpu]
+  run_all_tests_tf_multi_gpu:
+    runs-on: [self-hosted, gpu, multi-gpu]
     steps:
       - uses: actions/checkout@v2
 
@@ -313,7 +313,7 @@ jobs:
         run: |
           source .env/bin/activate
           pip install --upgrade pip
-          pip install .[tf,sklearn,testing,onnxruntime]
+          pip install .[tf,sklearn,testing,onnxruntime,sentencepiece]
           pip install git+https://github.com/huggingface/datasets
           pip list
 
@@ -329,11 +329,11 @@ jobs:
           RUN_SLOW: yes
         run: |
           source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_tf_multiple_gpu tests
+          python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_tf_multi_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
-        run: cat reports/tests_tf_multiple_gpu_failures_short.txt
+        run: cat reports/tests_tf_multi_gpu_failures_short.txt
 
       - name: Run all pipeline tests on multi-GPU
         if: ${{ always() }}
@@ -344,11 +344,11 @@ jobs:
           RUN_PIPELINE_TESTS: yes
         run: |
           source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s -m is_pipeline_test --make-reports=tests_tf_pipelines_multiple_gpu tests
+          python -m pytest -n 1 --dist=loadfile -s -m is_pipeline_test --make-reports=tests_tf_pipeline_multi_gpu tests
           
       - name: Failure short reports
         if: ${{ always() }}
-        run: cat reports/tests_tf_multiple_gpu_pipelines_failures_short.txt
+        run: cat reports/tests_tf_pipeline_multi_gpu_failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}

.gitignore

@@ -133,7 +133,6 @@ dmypy.json
 tensorflow_code
 
 # Models
-models
 proc_data
 
 # examples

CONTRIBUTING.md

@@ -308,3 +308,12 @@ Check our [documentation writing guide](https://github.com/huggingface/transform
 for more information.
 
 #### This guide was heavily inspired by the awesome [scikit-learn guide to contributing](https://github.com/scikit-learn/scikit-learn/blob/master/CONTRIBUTING.md)
+
+
+### Develop on Windows
+
+One way one can run the make command on Window is to pass by MSYS2:
+
+1. [Download MSYS2](https://www.msys2.org/), we assume to have it installed in C:\msys64
+2. Open the command line C:\msys64\msys2.exe (it should be available from the start menu)
+3. Run in the shell: `pacman -Syu` and install make with `pacman -S make`

README.md

@@ -181,6 +181,7 @@ Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih.
 1. **[LXMERT](https://huggingface.co/transformers/model_doc/lxmert.html)** (from UNC Chapel Hill) released with the paper [LXMERT: Learning Cross-Modality Encoder Representations from Transformers for Open-Domain Question Answering](https://arxiv.org/abs/1908.07490) by Hao Tan and Mohit Bansal.
 1. **[MarianMT](https://huggingface.co/transformers/model_doc/marian.html)** Machine translation models trained using [OPUS](http://opus.nlpl.eu/) data by Jörg Tiedemann. The [Marian Framework](https://marian-nmt.github.io/) is being developed by the Microsoft Translator Team.
 1. **[MBart](https://huggingface.co/transformers/model_doc/mbart.html)** (from Facebook) released with the paper [Multilingual Denoising Pre-training for Neural Machine Translation](https://arxiv.org/abs/2001.08210) by Yinhan Liu, Jiatao Gu, Naman Goyal, Xian Li, Sergey Edunov, Marjan Ghazvininejad, Mike Lewis, Luke Zettlemoyer.
+1. **[MT5](https://huggingface.co/transformers/model_doc/mt5.html)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
 1. **[Pegasus](https://huggingface.co/transformers/model_doc/pegasus.html)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777)> by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
 1. **[ProphetNet](https://huggingface.co/transformers/model_doc/prophetnet.html)** (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. **[Reformer](https://huggingface.co/transformers/model_doc/reformer.html)** (from Google Research) released with the paper [Reformer: The Efficient Transformer](https://arxiv.org/abs/2001.04451) by Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya.
@@ -196,6 +197,8 @@ ultilingual BERT into [DistilmBERT](https://github.com/huggingface/transformers/
 1. **[Other community models](https://huggingface.co/models)**, contributed by the [community](https://huggingface.co/users).
 1. Want to contribute a new model? We have added a **detailed guide and templates** to guide you in the process of adding a new model. You can find them in the [`templates`](./templates) folder of the repository. Be sure to check the [contributing guidelines](./CONTRIBUTING.md) and contact the maintainers or open an issue to collect feedbacks before starting your PR.
 
+To cehck if each model has an implementation in PyTorch/TensorFlow/Flax or has an associated tokenizer backed by the 🤗 Tokenizers library, refer to [this table](https://huggingface.co/transformers/index.html#bigtable)
+
 These implementations have been tested on several datasets (see the example scripts) and should match the performances of the original implementations. You can find more details on the performances in the Examples section of the [documentation](https://huggingface.co/transformers/examples.html).
 
 

docs/source/_static/css/huggingface.css

@@ -2,6 +2,15 @@
 
 /* Colab dropdown */
 
+table.center-aligned-table td {
+    text-align: center;
+}
+
+table.center-aligned-table th {
+    text-align: center;
+    vertical-align: middle;
+}
+
 .colab-dropdown {
     position: relative;
     display: inline-block;

docs/source/_static/js/custom.js

@@ -1,10 +1,11 @@
 // These two things need to be updated at each release for the version selector.
 // Last stable version
-const stableVersion = "v3.4.0"
+const stableVersion = "v3.5.0"
 // Dictionary doc folder to label
 const versionMapping = {
     "master": "master",
-    "": "v3.4.0",
+    "": "v3.5.0/v3.5.1",
+    "v3.4.0": "v3.4.0",
     "v3.3.1": "v3.3.0/v3.3.1",
     "v3.2.0": "v3.2.0",
     "v3.1.0": "v3.1.0 (stable)",

docs/source/conf.py

@@ -26,7 +26,7 @@ author = u'huggingface'
 # The short X.Y version
 version = u''
 # The full version, including alpha/beta/rc tags
-release = u'3.5.0'
+release = u'4.0.0'
 
 
 # -- General configuration ---------------------------------------------------

docs/source/index.rst

@@ -35,6 +35,8 @@ Choose the right framework for every part of a model's lifetime:
 - Move a single model between TF2.0/PyTorch frameworks at will
 - Seamlessly pick the right framework for training, evaluation, production
 
+Experimental support for Flax with a few models right now, expected to grow in the coming months.
+
 Contents
 -----------------------------------------------------------------------------------------------------------------------
 
@@ -44,7 +46,7 @@ The documentation is organized in five parts:
   and a glossary.
 - **USING 🤗 TRANSFORMERS** contains general tutorials on how to use the library.
 - **ADVANCED GUIDES** contains more advanced guides that are more specific to a given script or part of the library.
-- **RESEARCH** focuses on tutorials that have less to do with how to use the library but more about general resarch in
+- **RESEARCH** focuses on tutorials that have less to do with how to use the library but more about general research in
   transformers model
 - The three last section contain the documentation of each public class and function, grouped in:
 
@@ -52,8 +54,8 @@ The documentation is organized in five parts:
     - **MODELS** for the classes and functions related to each model implemented in the library.
     - **INTERNAL HELPERS** for the classes and functions we use internally.
 
-The library currently contains PyTorch and Tensorflow implementations, pre-trained model weights, usage scripts and
-conversion utilities for the following models:
+The library currently contains PyTorch, Tensorflow and Flax implementations, pretrained model weights, usage scripts
+and conversion utilities for the following models:
 
 ..
     This list is updated automatically from the README with `make fix-copies`. Do not update manually!
@@ -126,43 +128,137 @@ conversion utilities for the following models:
 21. :doc:`MBart <model_doc/mbart>` (from Facebook) released with the paper `Multilingual Denoising Pre-training for
     Neural Machine Translation <https://arxiv.org/abs/2001.08210>`__ by Yinhan Liu, Jiatao Gu, Naman Goyal, Xian Li,
     Sergey Edunov, Marjan Ghazvininejad, Mike Lewis, Luke Zettlemoyer.
-22. :doc:`Pegasus <model_doc/pegasus>` (from Google) released with the paper `PEGASUS: Pre-training with Extracted
+22. :doc:`MT5 <model_doc/mt5>` (from Google AI) released with the paper `mT5: A massively multilingual pre-trained
+    text-to-text transformer <https://arxiv.org/abs/2010.11934>`__ by Linting Xue, Noah Constant, Adam Roberts, Mihir
+    Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
+23. :doc:`Pegasus <model_doc/pegasus>` (from Google) released with the paper `PEGASUS: Pre-training with Extracted
     Gap-sentences for Abstractive Summarization <https://arxiv.org/abs/1912.08777>`__> by Jingqing Zhang, Yao Zhao,
     Mohammad Saleh and Peter J. Liu.
-23. :doc:`ProphetNet <model_doc/prophetnet>` (from Microsoft Research) released with the paper `ProphetNet: Predicting
+24. :doc:`ProphetNet <model_doc/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.
-24. :doc:`Reformer <model_doc/reformer>` (from Google Research) released with the paper `Reformer: The Efficient
+25. :doc:`Reformer <model_doc/reformer>` (from Google Research) released with the paper `Reformer: The Efficient
     Transformer <https://arxiv.org/abs/2001.04451>`__ by Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya.
-25. :doc:`RoBERTa <model_doc/roberta>` (from Facebook), released together with the paper a `Robustly Optimized BERT
+26. :doc:`RoBERTa <model_doc/roberta>` (from Facebook), released together with the paper a `Robustly Optimized BERT
     Pretraining Approach <https://arxiv.org/abs/1907.11692>`__ by Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar
     Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov. ultilingual BERT into `DistilmBERT
     <https://github.com/huggingface/transformers/tree/master/examples/distillation>`__ and a German version of
     DistilBERT.
-26. :doc:`SqueezeBert <model_doc/squeezebert>` released with the paper `SqueezeBERT: What can computer vision teach NLP
+27. :doc:`SqueezeBert <model_doc/squeezebert>` released with the paper `SqueezeBERT: What can computer vision teach NLP
     about efficient neural networks? <https://arxiv.org/abs/2006.11316>`__ by Forrest N. Iandola, Albert E. Shaw, Ravi
     Krishna, and Kurt W. Keutzer.
-27. :doc:`T5 <model_doc/t5>` (from Google AI) released with the paper `Exploring the Limits of Transfer Learning with a
+28. :doc:`T5 <model_doc/t5>` (from Google AI) released with the paper `Exploring the Limits of Transfer Learning with a
     Unified Text-to-Text Transformer <https://arxiv.org/abs/1910.10683>`__ by Colin Raffel and Noam Shazeer and Adam
     Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu.
-28. :doc:`Transformer-XL <model_doc/transformerxl>` (from Google/CMU) released with the paper `Transformer-XL:
+29. :doc:`Transformer-XL <model_doc/transformerxl>` (from Google/CMU) released with the paper `Transformer-XL:
     Attentive Language Models Beyond a Fixed-Length Context <https://arxiv.org/abs/1901.02860>`__ by Zihang Dai*,
     Zhilin Yang*, Yiming Yang, Jaime Carbonell, Quoc V. Le, Ruslan Salakhutdinov.
-29. :doc:`XLM <model_doc/xlm>` (from Facebook) released together with the paper `Cross-lingual Language Model
+30. :doc:`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.
-30. :doc:`XLM-ProphetNet <model_doc/xlmprophetnet>` (from Microsoft Research) released with the paper `ProphetNet:
+31. :doc:`XLM-ProphetNet <model_doc/xlmprophetnet>` (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.
-31. :doc:`XLM-RoBERTa <model_doc/xlmroberta>` (from Facebook AI), released together with the paper `Unsupervised
+32. :doc:`XLM-RoBERTa <model_doc/xlmroberta>` (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.
-32. :doc:`XLNet <model_doc/xlnet>` (from Google/CMU) released with the paper `​XLNet: Generalized Autoregressive
+33. :doc:`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.
-33. `Other community models <https://huggingface.co/models>`__, contributed by the `community
+34. `Other community models <https://huggingface.co/models>`__, contributed by the `community
     <https://huggingface.co/users>`__.
 
+
+.. _bigtable:
+
+The table below represents the current support in the library for each of those models, whether they have a Python
+tokenizer (called "slow"). A "fast" tokenizer backed by the 🤗 Tokenizers library, whether they have support in PyTorch,
+TensorFlow and/or Flax.
+
+..
+    This table is updated automatically from the auto modules with `make fix-copies`. Do not update manually!
+
+.. rst-class:: center-aligned-table
+
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|            Model            | Tokenizer slow | Tokenizer fast | PyTorch support | TensorFlow support | Flax Support |
++=============================+================+================+=================+====================+==============+
+|           ALBERT            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|            BART             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|            BERT             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|       Bert Generation       |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|         Blenderbot          |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|            CTRL             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|          CamemBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|             DPR             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|           DeBERTa           |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|         DistilBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|           ELECTRA           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|       Encoder decoder       |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+| FairSeq Machine-Translation |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|          FlauBERT           |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|     Funnel Transformer      |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|           LXMERT            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|          LayoutLM           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|         Longformer          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|           Marian            |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|         MobileBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|         OpenAI GPT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|        OpenAI GPT-2         |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|           Pegasus           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|         ProphetNet          |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|             RAG             |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|          Reformer           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|          RetriBERT          |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|           RoBERTa           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|         SqueezeBERT         |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|             T5              |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|       Transformer-XL        |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|             XLM             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|         XLM-RoBERTa         |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|        XLMProphetNet        |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|            XLNet            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|            mBART            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|             mT5             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+
+
 .. toctree::
     :maxdepth: 2
     :caption: Get started
@@ -248,6 +344,7 @@ conversion utilities for the following models:
     model_doc/marian
     model_doc/mbart
     model_doc/mobilebert
+    model_doc/mt5
     model_doc/gpt
     model_doc/gpt2
     model_doc/pegasus

docs/source/installation.md

@@ -70,15 +70,15 @@ to check 🤗 Transformers is properly installed.
 
 This library provides pretrained models that will be downloaded and cached locally. Unless you specify a location with
 `cache_dir=...` when you use methods like `from_pretrained`, these models will automatically be downloaded in the
-folder given by the shell environment variable ``TRANSFORMERS_CACHE``. The default value for it will be the PyTorch
-cache home followed by ``/transformers/`` (even if you don't have PyTorch installed). This is (by order of priority):
+folder given by the shell environment variable ``TRANSFORMERS_CACHE``. The default value for it will be the Hugging
+Face cache home followed by ``/transformers/``. This is (by order of priority):
 
-  * shell environment variable ``TORCH_HOME``
-  * shell environment variable ``XDG_CACHE_HOME`` + ``/torch/``
-  * default: ``~/.cache/torch/``
+  * shell environment variable ``HF_HOME`` 
+  * shell environment variable ``XDG_CACHE_HOME`` + ``/huggingface/``
+  * default: ``~/.cache/huggingface/``
 
 So if you don't have any specific environment variable set, the cache directory will be at
-``~/.cache/torch/transformers/``.
+``~/.cache/huggingface/transformers/``.
 
 **Note:** If you have set a shell environment variable for one of the predecessors of this library
 (``PYTORCH_TRANSFORMERS_CACHE`` or ``PYTORCH_PRETRAINED_BERT_CACHE``), those will be used if there is no shell
@@ -97,6 +97,6 @@ You should check out our [swift-coreml-transformers](https://github.com/huggingf
 It contains a set of tools to convert PyTorch or TensorFlow 2.0 trained Transformer models (currently contains `GPT-2`, 
 `DistilGPT-2`, `BERT`, and `DistilBERT`) to CoreML models that run on iOS devices.
 
-At some point in the future, you'll be able to seamlessly move from pre-training or fine-tuning models in PyTorch or
+At some point in the future, you'll be able to seamlessly move from pretraining or fine-tuning models in PyTorch or
 TensorFlow 2.0 to productizing them in CoreML, or prototype a model or an app in CoreML then research its
 hyperparameters or architecture from PyTorch or TensorFlow 2.0. Super exciting!

docs/source/main_classes/trainer.rst

@@ -35,7 +35,7 @@ Here is an example of how to customize :class:`~transformers.Trainer` using a cu
     class MyTrainer(Trainer):
         def compute_loss(self, model, inputs):
             labels = inputs.pop("labels")
-            outputs = models(**inputs)
+            outputs = model(**inputs)
             logits = outputs[0]
             return my_custom_loss(logits, labels)
 

docs/source/migration.md

@@ -1,5 +1,170 @@
 # Migrating from previous packages
 
+## Migrating from transformers `v3.x` to `v4.x`
+
+A couple of changes were introduced when the switch from version 3 to version 4 was done. Below is a summary of the
+expected changes:
+
+#### 1. AutoTokenizers and pipelines now use fast (rust) tokenizers by default.
+
+The python and rust tokenizers have roughly the same API, but the rust tokenizers have a more complete feature set. 
+
+This introduces two breaking changes:
+- The handling of overflowing tokens between the python and rust tokenizers is different.
+- The rust tokenizers do not accept integers in the encoding methods.
+
+##### How to obtain the same behavior as v3.x in v4.x
+
+- The pipelines now contain additional features out of the box. See the [token-classification pipeline with the `grouped_entities` flag](https://huggingface.co/transformers/main_classes/pipelines.html?highlight=textclassification#tokenclassificationpipeline).
+- The auto-tokenizers now return rust tokenizers. In order to obtain the python tokenizers instead, the user may use the `use_fast` flag by setting it to `False`:
+
+In version `v3.x`:
+```py
+from transformers import AutoTokenizer
+
+tokenizer = AutoTokenizer.from_pretrained("bert-base-cased")
+```
+to obtain the same in version `v4.x`:
+```py
+from transformers import AutoTokenizer
+
+tokenizer = AutoTokenizer.from_pretrained("bert-base-cased", use_fast=False)
+```
+
+#### 2. SentencePiece is removed from the required dependencies
+
+The requirement on the SentencePiece dependency has been lifted from the `setup.py`. This is done so that we may have a channel on anaconda cloud without relying on `conda-forge`. This means that the tokenizers that depend on the SentencePiece library will not be available with a standard `transformers` installation.
+
+This includes the **slow** versions of:
+- `XLNetTokenizer`
+- `AlbertTokenizer`
+- `CamembertTokenizer`
+- `MBartTokenizer`
+- `PegasusTokenizer`
+- `T5Tokenizer`
+- `ReformerTokenizer`
+- `XLMRobertaTokenizer`
+
+##### How to obtain the same behavior as v3.x in v4.x
+
+In order to obtain the same behavior as version `v3.x`, you should install `sentencepiece` additionally:
+
+In version `v3.x`:
+```bash
+pip install transformers
+```
+to obtain the same in version `v4.x`:
+```bash
+pip install transformers[sentencepiece]
+```
+or
+```bash
+pip install transformers sentencepiece
+```
+#### 3. The architecture of the repo has been updated so that each model resides in its folder
+
+The past and foreseeable addition of new models means that the number of files in the directory `src/transformers` keeps growing and becomes harder to navigate and understand. We made the choice to put each model and the files accompanying it in their own sub-directories.
+
+This is a breaking change as importing intermediary layers using a model's module directly needs to be done via a different path.
+
+##### How to obtain the same behavior as v3.x in v4.x
+
+In order to obtain the same behavior as version `v3.x`, you should update the path used to access the layers. 
+
+In version `v3.x`:
+```bash
+from transformers.modeling_bert import BertLayer
+```
+to obtain the same in version `v4.x`:
+```bash
+from transformers.models.bert.modeling_bert import BertLayer
+```
+
+#### 4. Switching the `return_dict` argument to `True` by default
+
+The [`return_dict` argument](https://huggingface.co/transformers/main_classes/output.html) enables the return of dict-like python objects containing the model outputs, instead of the standard tuples. This object is self-documented as keys can be used to retrieve values, while also behaving as a tuple as users may retrieve objects by index or by slice.
+
+This is a breaking change as the limitation of that tuple is that it cannot be unpacked: `value0, value1 = outputs` will not work.
+
+##### How to obtain the same behavior as v3.x in v4.x
+
+In order to obtain the same behavior as version `v3.x`, you should specify the `return_dict` argument to `False`, either in the model configuration or during the forward pass.
+
+In version `v3.x`:
+```bash
+model = BertModel.from_pretrained("bert-base-cased")
+outputs = model(**inputs)
+```
+to obtain the same in version `v4.x`:
+```bash
+model = BertModel.from_pretrained("bert-base-cased")
+outputs = model(**inputs, return_dict=False)
+```
+or
+```bash
+model = BertModel.from_pretrained("bert-base-cased", return_dict=False)
+outputs = model(**inputs)
+```
+
+#### 5. Removed some deprecated attributes
+
+Attributes that were deprecated have been removed if they had been deprecated for at least a month. The full list of deprecated attributes can be found in [#8604](https://github.com/huggingface/transformers/pull/8604).
+
+Here is a list of these attributes/methods/arguments and what their replacements should be:
+
+In several models, the labels become consistent with the other models:
+- `masked_lm_labels` becomes `labels` in `AlbertForMaskedLM` and `AlbertForPreTraining`.
+- `masked_lm_labels` becomes `labels` in `BertForMaskedLM` and `BertForPreTraining`.
+- `masked_lm_labels` becomes `labels` in `DistilBertForMaskedLM`.
+- `masked_lm_labels` becomes `labels` in `ElectraForMaskedLM`.
+- `masked_lm_labels` becomes `labels` in `LongformerForMaskedLM`.
+- `masked_lm_labels` becomes `labels` in `MobileBertForMaskedLM`.
+- `masked_lm_labels` becomes `labels` in `RobertaForMaskedLM`.
+- `lm_labels` becomes `labels` in `BartForConditionalGeneration`.
+- `lm_labels` becomes `labels` in `GPT2DoubleHeadsModel`.
+- `lm_labels` becomes `labels` in `OpenAIGPTDoubleHeadsModel`.
+- `lm_labels` becomes `labels` in `T5ForConditionalGeneration`.
+
+In several models, the caching mechanism becomes consistent with the other models:
+- `decoder_cached_states` becomes `past_key_values` in all BART-like, FSMT and T5 models.
+- `decoder_past_key_values` becomes `past_key_values` in all BART-like, FSMT and T5 models.
+- `past` becomes `past_key_values` in all CTRL models.
+- `past` becomes `past_key_values` in all GPT-2 models.
+
+Regarding the tokenizer classes:
+- The tokenizer attribute `max_len` becomes `model_max_length`.
+- The tokenizer attribute `return_lengths` becomes `return_length`.
+- The tokenizer encoding argument `is_pretokenized` becomes `is_split_into_words`.
+
+Regarding the `Trainer` class:
+- The `Trainer` argument `tb_writer` is removed in favor of the callback `TensorBoardCallback(tb_writer=...)`.
+- The `Trainer` argument `prediction_loss_only` is removed in favor of the class argument `args.prediction_loss_only`.
+- The `Trainer` attribute `data_collator` should be a callable.
+- The `Trainer` method `_log` is deprecated in favor of `log`.
+- The `Trainer` method `_training_step` is deprecated in favor of `training_step`.
+- The `Trainer` method `_prediction_loop` is deprecated in favor of `prediction_loop`.
+- The `Trainer` method `is_local_master` is deprecated in favor of `is_local_process_zero`.
+- The `Trainer` method `is_world_master` is deprecated in favor of `is_world_process_zero`.
+
+Regarding the `TFTrainer` class:
+- The `TFTrainer` argument `prediction_loss_only` is removed in favor of the class argument `args.prediction_loss_only`.
+- The `Trainer` method `_log` is deprecated in favor of `log`.
+- The `TFTrainer` method `_prediction_loop` is deprecated in favor of `prediction_loop`.
+- The `TFTrainer` method `_setup_wandb` is deprecated in favor of `setup_wandb`.
+- The `TFTrainer` method `_run_model` is deprecated in favor of `run_model`.
+
+Regarding the `TrainerArgument` class:
+- The `TrainerArgument` argument `evaluate_during_training` is deprecated in favor of `evaluation_strategy`.
+
+Regarding the Transfo-XL model:
+- The Transfo-XL configuration attribute `tie_weight` becomes `tie_words_embeddings`.
+- The Transfo-XL modeling method `reset_length` becomes `reset_memory_length`.
+
+Regarding pipelines:
+- The `FillMaskPipeline` argument `topk` becomes `top_k`.
+
+
+
 ## Migrating from pytorch-transformers to 🤗 Transformers
 
 Here is a quick summary of what you should take care of when migrating from `pytorch-transformers` to 🤗 Transformers.

docs/source/model_doc/albert.rst

@@ -51,10 +51,10 @@ AlbertTokenizer
 Albert specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_albert.AlbertForPreTrainingOutput
+.. autoclass:: transformers.models.albert.modeling_albert.AlbertForPreTrainingOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_albert.TFAlbertForPreTrainingOutput
+.. autoclass:: transformers.models.albert.modeling_tf_albert.TFAlbertForPreTrainingOutput
     :members:
 
 

docs/source/model_doc/bart.rst

@@ -34,6 +34,8 @@ ________________________________________________________________________________
 - An example of how to train :class:`~transformers.BartForConditionalGeneration` with a Hugging Face :obj:`datasets`
   object can be found in this `forum discussion
   <https://discuss.huggingface.co/t/train-bart-for-conditional-generation-e-g-summarization/1904>`__.
+- `Distilled checkpoints <https://huggingface.co/models?search=distilbart>`__ are described in this `paper
+  <https://arxiv.org/abs/2010.13002>`__.
 
 
 Implementation Notes
@@ -42,16 +44,33 @@ Implementation Notes
 - Bart doesn't use :obj:`token_type_ids` for sequence classification. Use :class:`~transformers.BartTokenizer` or
   :meth:`~transformers.BartTokenizer.encode` to get the proper splitting.
 - The forward pass of :class:`~transformers.BartModel` will create decoder inputs (using the helper function
-  :func:`transformers.modeling_bart._prepare_bart_decoder_inputs`) if they are not passed. This is different than some
-  other modeling APIs.
-- Model predictions are intended to be identical to the original implementation. This only works, however, if the
-  string you pass to :func:`fairseq.encode` starts with a space.
+  :func:`transformers.models.bart.modeling_bart._prepare_bart_decoder_inputs`) if they are not passed. This is
+  different than some other modeling APIs.
+- Model predictions are intended to be identical to the original implementation when
+  :obj:`force_bos_token_to_be_generated=True`. This only works, however, if the string you pass to
+  :func:`fairseq.encode` starts with a space.
 - :meth:`~transformers.BartForConditionalGeneration.generate` should be used for conditional generation tasks like
   summarization, see the example in that docstrings.
 - Models that load the `facebook/bart-large-cnn` weights will not have a :obj:`mask_token_id`, or be able to perform
   mask-filling tasks.
 - For training/forward passes that don't involve beam search, pass :obj:`use_cache=False`.
 
+Mask Filling
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The :obj:`facebook/bart-base` and :obj:`facebook/bart-large` checkpoints can be used to fill multi-token masks.
+
+.. code-block::
+
+    from transformers import BartForConditionalGeneration, BartTokenizer
+    model = BartForConditionalGeneration.from_pretrained("facebook/bart-large", force_bos_token_to_be_generated=True)
+    tok = BartTokenizer.from_pretrained("facebook/bart-large")
+    example_english_phrase = "UN Chief Says There Is No <mask> in Syria"
+    batch = tok(example_english_phrase, return_tensors='pt')
+    generated_ids = model.generate(batch['input_ids'])
+    assert tok.batch_decode(generated_ids, skip_special_tokens=True) == ['UN Chief Says There Is No Plan to Stop Chemical Weapons in Syria']
+
+
 
 BartConfig
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@@ -74,7 +93,7 @@ BartModel
 .. autoclass:: transformers.BartModel
     :members: forward
 
-.. autofunction:: transformers.modeling_bart._prepare_bart_decoder_inputs
+.. autofunction:: transformers.models.bart.modeling_bart._prepare_bart_decoder_inputs
 
 
 BartForConditionalGeneration

docs/source/model_doc/bert.rst

@@ -57,10 +57,10 @@ BertTokenizerFast
 Bert specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_bert.BertForPreTrainingOutput
+.. autoclass:: transformers.models.bert.modeling_bert.BertForPreTrainingOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_bert.TFBertForPreTrainingOutput
+.. autoclass:: transformers.models.bert.modeling_tf_bert.TFBertForPreTrainingOutput
     :members:
 
 
@@ -188,3 +188,10 @@ TFBertForQuestionAnswering
 
 .. autoclass:: transformers.TFBertForQuestionAnswering
     :members: call
+
+
+FlaxBertModel
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.FlaxBertModel
+    :members: __call__

docs/source/model_doc/bertgeneration.rst

@@ -10,7 +10,7 @@ Tasks <https://arxiv.org/abs/1907.12461>`__ by Sascha Rothe, Shashi Narayan, Ali
 
 The abstract from the paper is the following:
 
-*Unsupervised pre-training of large neural models has recently revolutionized Natural Language Processing. By
+*Unsupervised pretraining of large neural models has recently revolutionized Natural Language Processing. By
 warm-starting from the publicly released checkpoints, NLP practitioners have pushed the state-of-the-art on multiple
 benchmarks while saving significant amounts of compute time. So far the focus has been mainly on the Natural Language
 Understanding tasks. In this paper, we demonstrate the efficacy of pre-trained checkpoints for Sequence Generation. We
@@ -40,7 +40,7 @@ Usage:
   labels = tokenizer('This is a short summary', return_tensors="pt").input_ids
 
   # train...
-  loss = bert2bert(input_ids=input_ids, decoder_input_ids=labels, labels=labels, return_dict=True).loss
+  loss = bert2bert(input_ids=input_ids, decoder_input_ids=labels, labels=labels).loss
   loss.backward()
 
 

docs/source/model_doc/deberta.rst

@@ -20,8 +20,8 @@ disentangled attention mechanism, where each word is represented using two vecto
 position, respectively, and the attention weights among words are computed using disentangled matrices on their
 contents and relative positions. Second, an enhanced mask decoder is used to replace the output softmax layer to
 predict the masked tokens for model pretraining. We show that these two techniques significantly improve the efficiency
-of model pre-training and performance of downstream tasks. Compared to RoBERTa-Large, a DeBERTa model trained on half
-of the training data performs consistently better on a wide range of NLP tasks, achieving improvements on MNLI by +0.9%
+of model pretraining and performance of downstream tasks. Compared to RoBERTa-Large, a DeBERTa model trained on half of
+the training data performs consistently better on a wide range of NLP tasks, achieving improvements on MNLI by +0.9%
 (90.2% vs. 91.1%), on SQuAD v2.0 by +2.3% (88.4% vs. 90.7%) and RACE by +3.6% (83.2% vs. 86.8%). The DeBERTa code and
 pre-trained models will be made publicly available at https://github.com/microsoft/DeBERTa.*
 

docs/source/model_doc/distilbert.rst

@@ -18,9 +18,9 @@ operating these large models in on-the-edge and/or under constrained computation
 remains challenging. In this work, we propose a method to pre-train a smaller general-purpose language representation
 model, called DistilBERT, which can then be fine-tuned with good performances on a wide range of tasks like its larger
 counterparts. While most prior work investigated the use of distillation for building task-specific models, we leverage
-knowledge distillation during the pre-training phase and show that it is possible to reduce the size of a BERT model by
+knowledge distillation during the pretraining phase and show that it is possible to reduce the size of a BERT model by
 40%, while retaining 97% of its language understanding capabilities and being 60% faster. To leverage the inductive
-biases learned by larger models during pre-training, we introduce a triple loss combining language modeling,
+biases learned by larger models during pretraining, we introduce a triple loss combining language modeling,
 distillation and cosine-distance losses. Our smaller, faster and lighter model is cheaper to pre-train and we
 demonstrate its capabilities for on-device computations in a proof-of-concept experiment and a comparative on-device
 study.*

docs/source/model_doc/dpr.rst

@@ -5,7 +5,7 @@ Overview
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 Dense Passage Retrieval (DPR) is a set of tools and models for state-of-the-art open-domain Q&A research. It was
-intorduced in `Dense Passage Retrieval for Open-Domain Question Answering <https://arxiv.org/abs/2004.04906>`__ by
+introduced in `Dense Passage Retrieval for Open-Domain Question Answering <https://arxiv.org/abs/2004.04906>`__ by
 Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, Wen-tau Yih.
 
 The abstract from the paper is the following:
@@ -71,13 +71,13 @@ DPRReaderTokenizerFast
 DPR specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_dpr.DPRContextEncoderOutput
+.. autoclass:: transformers.models.dpr.modeling_dpr.DPRContextEncoderOutput
     :members:
 
-.. autoclass:: transformers.modeling_dpr.DPRQuestionEncoderOutput
+.. autoclass:: transformers.models.dpr.modeling_dpr.DPRQuestionEncoderOutput
     :members:
 
-.. autoclass:: transformers.modeling_dpr.DPRReaderOutput
+.. autoclass:: transformers.models.dpr.modeling_dpr.DPRReaderOutput
     :members:
 
 
@@ -99,3 +99,22 @@ DPRReader
 
 .. autoclass:: transformers.DPRReader
     :members: forward
+
+TFDPRContextEncoder
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFDPRContextEncoder
+    :members: call
+
+TFDPRQuestionEncoder
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFDPRQuestionEncoder
+    :members: call
+
+
+TFDPRReader
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFDPRReader
+    :members: call

docs/source/model_doc/electra.rst

@@ -12,14 +12,14 @@ identify which tokens were replaced by the generator in the sequence.
 
 The abstract from the paper is the following:
 
-*Masked language modeling (MLM) pre-training methods such as BERT corrupt the input by replacing some tokens with
-[MASK] and then train a model to reconstruct the original tokens. While they produce good results when transferred to
+*Masked language modeling (MLM) pretraining methods such as BERT corrupt the input by replacing some tokens with [MASK]
+and then train a model to reconstruct the original tokens. While they produce good results when transferred to
 downstream NLP tasks, they generally require large amounts of compute to be effective. As an alternative, we propose a
-more sample-efficient pre-training task called replaced token detection. Instead of masking the input, our approach
+more sample-efficient pretraining task called replaced token detection. Instead of masking the input, our approach
 corrupts it by replacing some tokens with plausible alternatives sampled from a small generator network. Then, instead
 of training a model that predicts the original identities of the corrupted tokens, we train a discriminative model that
 predicts whether each token in the corrupted input was replaced by a generator sample or not. Thorough experiments
-demonstrate this new pre-training task is more efficient than MLM because the task is defined over all input tokens
+demonstrate this new pretraining task is more efficient than MLM because the task is defined over all input tokens
 rather than just the small subset that was masked out. As a result, the contextual representations learned by our
 approach substantially outperform the ones learned by BERT given the same model size, data, and compute. The gains are
 particularly strong for small models; for example, we train a model on one GPU for 4 days that outperforms GPT (trained
@@ -69,10 +69,10 @@ ElectraTokenizerFast
 Electra specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_electra.ElectraForPreTrainingOutput
+.. autoclass:: transformers.models.electra.modeling_electra.ElectraForPreTrainingOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_electra.TFElectraForPreTrainingOutput
+.. autoclass:: transformers.models.electra.modeling_tf_electra.TFElectraForPreTrainingOutput
     :members:
 
 

docs/source/model_doc/flaubert.rst

@@ -19,7 +19,7 @@ representations (Dai and Le, 2015; Peters et al., 2018; Howard and Ruder, 2018;
 heterogeneous French corpus. Models of different sizes are trained using the new CNRS (French National Centre for
 Scientific Research) Jean Zay supercomputer. We apply our French language models to diverse NLP tasks (text
 classification, paraphrasing, natural language inference, parsing, word sense disambiguation) and show that most of the
-time they outperform other pre-training approaches. Different versions of FlauBERT as well as a unified evaluation
+time they outperform other pretraining approaches. Different versions of FlauBERT as well as a unified evaluation
 protocol for the downstream tasks, called FLUE (French Language Understanding Evaluation), are shared to the research
 community for further reproducible experiments in French NLP.*
 

docs/source/model_doc/funnel.rst

@@ -65,10 +65,10 @@ FunnelTokenizerFast
 Funnel specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_funnel.FunnelForPreTrainingOutput
+.. autoclass:: transformers.models.funnel.modeling_funnel.FunnelForPreTrainingOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_funnel.TFFunnelForPreTrainingOutput
+.. autoclass:: transformers.models.funnel.modeling_tf_funnel.TFFunnelForPreTrainingOutput
     :members:
 
 

docs/source/model_doc/gpt.rst

@@ -14,7 +14,7 @@ The abstract from the paper is the following:
 *Natural language understanding comprises a wide range of diverse tasks such as textual entailment, question answering,
 semantic similarity assessment, and document classification. Although large unlabeled text corpora are abundant,
 labeled data for learning these specific tasks is scarce, making it challenging for discriminatively trained models to
-perform adequately. We demonstrate that large gains on these tasks can be realized by generative pre-training of a
+perform adequately. We demonstrate that large gains on these tasks can be realized by generative pretraining of a
 language model on a diverse corpus of unlabeled text, followed by discriminative fine-tuning on each specific task. In
 contrast to previous approaches, we make use of task-aware input transformations during fine-tuning to achieve
 effective transfer while requiring minimal changes to the model architecture. We demonstrate the effectiveness of our
@@ -72,10 +72,10 @@ OpenAIGPTTokenizerFast
 OpenAI specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_openai.OpenAIGPTDoubleHeadsModelOutput
+.. autoclass:: transformers.models.openai.modeling_openai.OpenAIGPTDoubleHeadsModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_openai.TFOpenAIGPTDoubleHeadsModelOutput
+.. autoclass:: transformers.models.openai.modeling_tf_openai.TFOpenAIGPTDoubleHeadsModelOutput
     :members:
 
 

docs/source/model_doc/gpt2.rst

@@ -60,10 +60,10 @@ GPT2TokenizerFast
 GPT2 specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_gpt2.GPT2DoubleHeadsModelOutput
+.. autoclass:: transformers.models.gpt2.modeling_gpt2.GPT2DoubleHeadsModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_gpt2.TFGPT2DoubleHeadsModelOutput
+.. autoclass:: transformers.models.gpt2.modeling_tf_gpt2.TFGPT2DoubleHeadsModelOutput
     :members:
 
 

docs/source/model_doc/layoutlm.rst

@@ -6,19 +6,19 @@ Overview
 
 The LayoutLM model was proposed in the paper `LayoutLM: Pre-training of Text and Layout for Document Image
 Understanding <https://arxiv.org/abs/1912.13318>`__ by Yiheng Xu, Minghao Li, Lei Cui, Shaohan Huang, Furu Wei, and
-Ming Zhou. It's a simple but effective pre-training method of text and layout for document image understanding and
+Ming Zhou. It's a simple but effective pretraining method of text and layout for document image understanding and
 information extraction tasks, such as form understanding and receipt understanding.
 
 The abstract from the paper is the following:
 
 *Pre-training techniques have been verified successfully in a variety of NLP tasks in recent years. Despite the
-widespread use of pre-training models for NLP applications, they almost exclusively focus on text-level manipulation,
+widespread use of pretraining models for NLP applications, they almost exclusively focus on text-level manipulation,
 while neglecting layout and style information that is vital for document image understanding. In this paper, we propose
 the \textbf{LayoutLM} to jointly model interactions between text and layout information across scanned document images,
 which is beneficial for a great number of real-world document image understanding tasks such as information extraction
 from scanned documents. Furthermore, we also leverage image features to incorporate words' visual information into
 LayoutLM. To the best of our knowledge, this is the first time that text and layout are jointly learned in a single
-framework for document-level pre-training. It achieves new state-of-the-art results in several downstream tasks,
+framework for document-level pretraining. It achieves new state-of-the-art results in several downstream tasks,
 including form understanding (from 70.72 to 79.27), receipt understanding (from 94.02 to 95.24) and document image
 classification (from 93.07 to 94.42).*
 

docs/source/model_doc/longformer.rst

@@ -93,29 +93,47 @@ LongformerTokenizerFast
 Longformer specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_longformer.LongformerBaseModelOutput
+.. autoclass:: transformers.models.longformer.modeling_longformer.LongformerBaseModelOutput
     :members: 
 
-.. autoclass:: transformers.modeling_longformer.LongformerBaseModelOutputWithPooling
+.. autoclass:: transformers.models.longformer.modeling_longformer.LongformerBaseModelOutputWithPooling
     :members: 
 
-.. autoclass:: transformers.modeling_longformer.LongformerMultipleChoiceModelOutput
+.. autoclass:: transformers.models.longformer.modeling_longformer.LongformerMaskedLMOutput
     :members: 
 
-.. autoclass:: transformers.modeling_longformer.LongformerQuestionAnsweringModelOutput
+.. autoclass:: transformers.models.longformer.modeling_longformer.LongformerQuestionAnsweringModelOutput
     :members: 
 
-.. autoclass:: transformers.modeling_tf_longformer.TFLongformerBaseModelOutput
+.. autoclass:: transformers.models.longformer.modeling_longformer.LongformerSequenceClassifierOutput
     :members: 
 
-.. autoclass:: transformers.modeling_tf_longformer.TFLongformerBaseModelOutputWithPooling
+.. autoclass:: transformers.models.longformer.modeling_longformer.LongformerMultipleChoiceModelOutput
     :members: 
 
-.. autoclass:: transformers.modeling_tf_longformer.TFLongformerQuestionAnsweringModelOutput
+.. autoclass:: transformers.models.longformer.modeling_longformer.LongformerTokenClassifierOutput
     :members: 
 
-LongformerModel
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+.. autoclass:: transformers.models.longformer.modeling_tf_longformer.TFLongformerBaseModelOutput
+    :members: 
+
+.. autoclass:: transformers.models.longformer.modeling_tf_longformer.TFLongformerBaseModelOutputWithPooling
+    :members: 
+
+.. autoclass:: transformers.models.longformer.modeling_tf_longformer.TFLongformerMaskedLMOutput
+    :members: 
+
+.. autoclass:: transformers.models.longformer.modeling_tf_longformer.TFLongformerQuestionAnsweringModelOutput
+    :members: 
+
+.. autoclass:: transformers.models.longformer.modeling_tf_longformer.TFLongformerSequenceClassifierOutput
+    :members: 
+
+.. autoclass:: transformers.models.longformer.modeling_tf_longformer.TFLongformerMultipleChoiceModelOutput
+    :members: 
+
+.. autoclass:: transformers.models.longformer.modeling_tf_longformer.TFLongformerTokenClassifierOutput
+    :members: 
 
 LongformerModel
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@@ -179,3 +197,24 @@ TFLongformerForQuestionAnswering
 .. autoclass:: transformers.TFLongformerForQuestionAnswering
     :members: call
 
+
+TFLongformerForSequenceClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFLongformerForSequenceClassification
+    :members: call
+
+
+TFLongformerForTokenClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFLongformerForTokenClassification
+    :members: call
+
+
+TFLongformerForMultipleChoice
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFLongformerForMultipleChoice
+    :members: call
+

docs/source/model_doc/lxmert.rst

@@ -19,7 +19,7 @@ Encoder Representations from Transformers) framework to learn these vision-and-l
 build a large-scale Transformer model that consists of three encoders: an object relationship encoder, a language
 encoder, and a cross-modality encoder. Next, to endow our model with the capability of connecting vision and language
 semantics, we pre-train the model with large amounts of image-and-sentence pairs, via five diverse representative
-pre-training tasks: masked language modeling, masked object prediction (feature regression and label classification),
+pretraining tasks: masked language modeling, masked object prediction (feature regression and label classification),
 cross-modality matching, and image question answering. These tasks help in learning both intra-modality and
 cross-modality relationships. After fine-tuning from our pretrained parameters, our model achieves the state-of-the-art
 results on two visual question answering datasets (i.e., VQA and GQA). We also show the generalizability of our
@@ -67,19 +67,19 @@ LxmertTokenizerFast
 Lxmert specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_lxmert.LxmertModelOutput
+.. autoclass:: transformers.models.lxmert.modeling_lxmert.LxmertModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_lxmert.LxmertForPreTrainingOutput
+.. autoclass:: transformers.models.lxmert.modeling_lxmert.LxmertForPreTrainingOutput
     :members:
 
-.. autoclass:: transformers.modeling_lxmert.LxmertForQuestionAnsweringOutput
+.. autoclass:: transformers.models.lxmert.modeling_lxmert.LxmertForQuestionAnsweringOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_lxmert.TFLxmertModelOutput
+.. autoclass:: transformers.models.lxmert.modeling_tf_lxmert.TFLxmertModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_lxmert.TFLxmertForPreTrainingOutput
+.. autoclass:: transformers.models.lxmert.modeling_tf_lxmert.TFLxmertForPreTrainingOutput
     :members:
 
 

docs/source/model_doc/marian.rst

@@ -5,7 +5,7 @@ MarianMT
 <https://github.com/huggingface/transformers/issues/new?assignees=sshleifer&labels=&template=bug-report.md&title>`__
 and assign @patrickvonplaten.
 
-Translations should be similar, but not identical to, output in the test set linked to in each model card.
+Translations should be similar, but not identical to output in the test set linked to in each model card.
 
 Implementation Notes
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@@ -35,24 +35,109 @@ Naming
   <https://developers.google.com/admin-sdk/directory/v1/languages>`__, three digit codes require googling "language
   code {code}".
 - Codes formatted like :obj:`es_AR` are usually :obj:`code_{region}`. That one is Spanish from Argentina.
+- The models were converted in two stages. The first 1000 models use ISO-639-2 codes to identify languages, the second
+  group use a combination of ISO-639-5 codes and ISO-639-2 codes.
 
 
+Examples
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+- Since Marian models are smaller than many other translation models available in the library, they can be useful for
+  fine-tuning experiments and integration tests.
+- `Fine-tune on TPU
+  <https://github.com/huggingface/transformers/blob/master/examples/seq2seq/builtin_trainer/train_distil_marian_enro_tpu.sh>`__
+- `Fine-tune on GPU
+  <https://github.com/huggingface/transformers/blob/master/examples/seq2seq/builtin_trainer/train_distil_marian_enro.sh>`__
+- `Fine-tune on GPU with pytorch-lightning
+  <https://github.com/huggingface/transformers/blob/master/examples/seq2seq/distil_marian_no_teacher.sh>`__
+
 Multilingual Models
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-All model names use the following format: :obj:`Helsinki-NLP/opus-mt-{src}-{tgt}`:
+- All model names use the following format: :obj:`Helsinki-NLP/opus-mt-{src}-{tgt}`:
+- If a model can output multiple languages, and you should specify a language code by prepending the desired output
+  language to the :obj:`src_text`.
+- You can see a models's supported language codes in its model card, under target constituents, like in `opus-mt-en-roa
+  <https://huggingface.co/Helsinki-NLP/opus-mt-en-roa>`__.
+- Note that if a model is only multilingual on the source side, like :obj:`Helsinki-NLP/opus-mt-roa-en`, no language
+  codes are required.
 
-    - If :obj:`src` is in all caps, the model supports multiple input languages, you can figure out which ones by
-      looking at the model card, or the Group Members `mapping
-      <https://gist.github.com/sshleifer/6d20e7761931b08e73c3219027b97b8a>`_ .
-    - If :obj:`tgt` is in all caps, the model can output multiple languages, and you should specify a language code by
-      prepending the desired output language to the :obj:`src_text`.
-    - You can see a tokenizer's supported language codes in ``tokenizer.supported_language_codes``
-
-Example of translating english to many romance languages, using language codes:
+New multi-lingual models from the `Tatoeba-Challenge repo <https://github.com/Helsinki-NLP/Tatoeba-Challenge>`__
+require 3 character language codes:
 
 .. code-block:: python
 
+    from transformers import MarianMTModel, MarianTokenizer
+    src_text = [
+        '>>fra<< this is a sentence in english that we want to translate to french',
+        '>>por<< This should go to portuguese',
+        '>>esp<< And this to Spanish'
+    ]
+
+    model_name = 'Helsinki-NLP/opus-mt-en-roa'
+    tokenizer = MarianTokenizer.from_pretrained(model_name)
+    print(tokenizer.supported_language_codes)
+    model = MarianMTModel.from_pretrained(model_name)
+    translated = model.generate(**tokenizer.prepare_seq2seq_batch(src_text, return_tensors="pt"))
+    tgt_text = [tokenizer.decode(t, skip_special_tokens=True) for t in translated]
+    # ["c'est une phrase en anglais que nous voulons traduire en français",
+    # 'Isto deve ir para o português.',
+    # 'Y esto al español']
+
+
+
+
+Code to see available pretrained models:
+
+.. code-block:: python
+
+    from transformers.hf_api import HfApi
+    model_list = HfApi().model_list()
+    org = "Helsinki-NLP"
+    model_ids = [x.modelId for x in model_list if x.modelId.startswith(org)]
+    suffix = [x.split('/')[1] for x in model_ids]
+    old_style_multi_models = [f'{org}/{s}' for s in suffix if s != s.lower()]
+
+
+
+Old Style Multi-Lingual Models
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+These are the old style multi-lingual models ported from the OPUS-MT-Train repo: and the members of each language
+group:
+
+.. code-block:: python
+
+    ['Helsinki-NLP/opus-mt-NORTH_EU-NORTH_EU',
+     'Helsinki-NLP/opus-mt-ROMANCE-en',
+     'Helsinki-NLP/opus-mt-SCANDINAVIA-SCANDINAVIA',
+     'Helsinki-NLP/opus-mt-de-ZH',
+     'Helsinki-NLP/opus-mt-en-CELTIC',
+     'Helsinki-NLP/opus-mt-en-ROMANCE',
+     'Helsinki-NLP/opus-mt-es-NORWAY',
+     'Helsinki-NLP/opus-mt-fi-NORWAY',
+     'Helsinki-NLP/opus-mt-fi-ZH',
+     'Helsinki-NLP/opus-mt-fi_nb_no_nn_ru_sv_en-SAMI',
+     'Helsinki-NLP/opus-mt-sv-NORWAY',
+     'Helsinki-NLP/opus-mt-sv-ZH']
+    GROUP_MEMBERS = {
+     'ZH': ['cmn', 'cn', 'yue', 'ze_zh', 'zh_cn', 'zh_CN', 'zh_HK', 'zh_tw', 'zh_TW', 'zh_yue', 'zhs', 'zht', 'zh'],
+     'ROMANCE': ['fr', 'fr_BE', 'fr_CA', 'fr_FR', 'wa', 'frp', 'oc', 'ca', 'rm', 'lld', 'fur', 'lij', 'lmo', 'es', 'es_AR', 'es_CL', 'es_CO', 'es_CR', 'es_DO', 'es_EC', 'es_ES', 'es_GT', 'es_HN', 'es_MX', 'es_NI', 'es_PA', 'es_PE', 'es_PR', 'es_SV', 'es_UY', 'es_VE', 'pt', 'pt_br', 'pt_BR', 'pt_PT', 'gl', 'lad', 'an', 'mwl', 'it', 'it_IT', 'co', 'nap', 'scn', 'vec', 'sc', 'ro', 'la'],
+     'NORTH_EU': ['de', 'nl', 'fy', 'af', 'da', 'fo', 'is', 'no', 'nb', 'nn', 'sv'],
+     'SCANDINAVIA': ['da', 'fo', 'is', 'no', 'nb', 'nn', 'sv'],
+     'SAMI': ['se', 'sma', 'smj', 'smn', 'sms'],
+     'NORWAY': ['nb_NO', 'nb', 'nn_NO', 'nn', 'nog', 'no_nb', 'no'],
+     'CELTIC': ['ga', 'cy', 'br', 'gd', 'kw', 'gv']
+    }
+
+
+
+
+Example of translating english to many romance languages, using old-style 2 character language codes
+
+
+.. code-block::python
+
     from transformers import MarianMTModel, MarianTokenizer
     src_text = [
         '>>fr<< this is a sentence in english that we want to translate to french',
@@ -63,52 +148,12 @@ Example of translating english to many romance languages, using language codes:
     model_name = 'Helsinki-NLP/opus-mt-en-ROMANCE'
     tokenizer = MarianTokenizer.from_pretrained(model_name)
     print(tokenizer.supported_language_codes)
+
     model = MarianMTModel.from_pretrained(model_name)
-    translated = model.generate(**tokenizer.prepare_seq2seq_batch(src_text))
+    translated = model.generate(**tokenizer.prepare_seq2seq_batch(src_text, return_tensors="pt"))
     tgt_text = [tokenizer.decode(t, skip_special_tokens=True) for t in translated]
-    # ["c'est une phrase en anglais que nous voulons traduire en français",
-    # 'Isto deve ir para o português.',
-    # 'Y esto al español']
+    # ["c'est une phrase en anglais que nous voulons traduire en français", 'Isto deve ir para o português.',  'Y esto al español']
 
-Sometimes, models were trained on collections of languages that do not resolve to a group. In this case, _ is used as a
-separator for src or tgt, as in :obj:`Helsinki-NLP/opus-mt-en_el_es_fi-en_el_es_fi`. These still require language
-codes.
-
-There are many supported regional language codes, like :obj:`>>es_ES<<` (Spain) and :obj:`>>es_AR<<` (Argentina), that
-do not seem to change translations. I have not found these to provide different results than just using :obj:`>>es<<`.
-
-For example:
-
-    - `Helsinki-NLP/opus-mt-NORTH_EU-NORTH_EU`: translates from all NORTH_EU languages (see `mapping
-      <https://gist.github.com/sshleifer/6d20e7761931b08e73c3219027b97b8a>`_) to all NORTH_EU languages. Use a special
-      language code like :obj:`>>de<<` to specify output language.
-    - `Helsinki-NLP/opus-mt-ROMANCE-en`: translates from many romance languages to english, no codes needed since there
-      is only one target language.
-
-
-
-.. code-block:: python
-
-    GROUP_MEMBERS = {
-     'ZH': ['cmn', 'cn', 'yue', 'ze_zh', 'zh_cn', 'zh_CN', 'zh_HK', 'zh_tw', 'zh_TW', 'zh_yue', 'zhs', 'zht', 'zh'],
-     'ROMANCE': ['fr', 'fr_BE', 'fr_CA', 'fr_FR', 'wa', 'frp', 'oc', 'ca', 'rm', 'lld', 'fur', 'lij', 'lmo', 'es', 'es_AR', 'es_CL', 'es_CO', 'es_CR', 'es_DO', 'es_EC', 'es_ES', 'es_GT', 'es_HN', 'es_MX', 'es_NI', 'es_PA', 'es_PE', 'es_PR', 'es_SV', 'es_UY', 'es_VE', 'pt', 'pt_br', 'pt_BR', 'pt_PT', 'gl', 'lad', 'an', 'mwl', 'it', 'it_IT', 'co', 'nap', 'scn', 'vec', 'sc', 'ro', 'la'],
-     'NORTH_EU': ['de', 'nl', 'fy', 'af', 'da', 'fo', 'is', 'no', 'nb', 'nn', 'sv'],
-     'SCANDINAVIA': ['da', 'fo', 'is', 'no', 'nb', 'nn', 'sv'],
-     'SAMI': ['se', 'sma', 'smj', 'smn', 'sms'],
-     'NORWAY': ['nb_NO', 'nb', 'nn_NO', 'nn', 'nog', 'no_nb', 'no'],
-     'CELTIC': ['ga', 'cy', 'br', 'gd', 'kw', 'gv']
-    }
-
-Code to see available pretrained models:
-
-.. code-block:: python
-
-    from transformers.hf_api import HfApi
-    model_list = HfApi().model_list()
-    org = "Helsinki-NLP"
-    model_ids = [x.modelId for x in model_list if x.modelId.startswith(org)]
-    suffix = [x.split('/')[1] for x in model_ids]
-    multi_models = [f'{org}/{s}' for s in suffix if s != s.lower()]
 
 
 MarianConfig

docs/source/model_doc/mbart.rst

@@ -13,12 +13,19 @@ The MBart model was presented in `Multilingual Denoising Pre-training for Neural
 Ghazvininejad, Mike Lewis, Luke Zettlemoyer.
 
 According to the abstract, MBART is a sequence-to-sequence denoising auto-encoder pretrained on large-scale monolingual
-corpora in many languages using the BART objective. mBART is one of the first methods for pre-training a complete
+corpora in many languages using the BART objective. mBART is one of the first methods for pretraining a complete
 sequence-to-sequence model by denoising full texts in multiple languages, while previous approaches have focused only
 on the encoder, decoder, or reconstructing parts of the text.
 
 The Authors' code can be found `here <https://github.com/pytorch/fairseq/tree/master/examples/mbart>`__
 
+Examples
+_______________________________________________________________________________________________________________________
+
+- Examples and scripts for fine-tuning mBART and other models for sequence to sequence tasks can be found in
+  `examples/seq2seq/ <https://github.com/huggingface/transformers/blob/master/examples/seq2seq/README.md>`__.
+- Given the large embeddings table, mBART consumes a large amount of GPU RAM, especially for fine-tuning.
+  :class:`MarianMTModel` is usually a better choice for bilingual machine translation.
 
 Training
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@@ -37,12 +44,8 @@ the sequences for sequence-to-sequence fine-tuning.
 
     example_english_phrase = "UN Chief Says There Is No Military Solution in Syria"
     expected_translation_romanian = "Şeful ONU declară că nu există o soluţie militară în Siria"
-    batch = tokenizer.prepare_seq2seq_batch(example_english_phrase, src_lang="en_XX", tgt_lang="ro_RO", tgt_texts=expected_translation_romanian)
-    input_ids = batch["input_ids"]
-    target_ids = batch["decoder_input_ids"]
-    decoder_input_ids = target_ids[:, :-1].contiguous()
-    labels = target_ids[:, 1:].clone()
-    model(input_ids=input_ids, decoder_input_ids=decoder_input_ids, labels=labels) #forward
+    batch = tokenizer.prepare_seq2seq_batch(example_english_phrase, src_lang="en_XX", tgt_lang="ro_RO", tgt_texts=expected_translation_romanian, return_tensors="pt")
+    model(input_ids=batch['input_ids'], labels=batch['labels']) # forward pass
 
 - Generation
 
@@ -55,7 +58,7 @@ the sequences for sequence-to-sequence fine-tuning.
     model = MBartForConditionalGeneration.from_pretrained("facebook/mbart-large-en-ro")
     tokenizer = MBartTokenizer.from_pretrained("facebook/mbart-large-en-ro")
     article = "UN Chief Says There Is No Military Solution in Syria"
-    batch = tokenizer.prepare_seq2seq_batch(src_texts=[article], src_lang="en_XX")
+    batch = tokenizer.prepare_seq2seq_batch(src_texts=[article], src_lang="en_XX", return_tensors="pt")
     translated_tokens = model.generate(**batch, decoder_start_token_id=tokenizer.lang_code_to_id["ro_RO"])
     translation = tokenizer.batch_decode(translated_tokens, skip_special_tokens=True)[0]
     assert translation == "Şeful ONU declară că nu există o soluţie militară în Siria"

docs/source/model_doc/mobilebert.rst

@@ -58,10 +58,10 @@ MobileBertTokenizerFast
 MobileBert specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_mobilebert.MobileBertForPreTrainingOutput
+.. autoclass:: transformers.models.mobilebert.modeling_mobilebert.MobileBertForPreTrainingOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_mobilebert.TFMobileBertForPreTrainingOutput
+.. autoclass:: transformers.models.mobilebert.modeling_tf_mobilebert.TFMobileBertForPreTrainingOutput
     :members:
 
 

docs/source/model_doc/mt5.rst

@@ -0,0 +1,53 @@
+MT5
+-----------------------------------------------------------------------------------------------------------------------
+
+Overview
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The mT5 model was presented in `mT5: A massively multilingual pre-trained text-to-text transformer
+<https://arxiv.org/abs/2010.11934>`_ by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya
+Siddhant, Aditya Barua, Colin Raffel.
+
+The abstract from the paper is the following:
+
+*The recent "Text-to-Text Transfer Transformer" (T5) leveraged a unified text-to-text format and scale to attain
+state-of-the-art results on a wide variety of English-language NLP tasks. In this paper, we introduce mT5, a
+multilingual variant of T5 that was pre-trained on a new Common Crawl-based dataset covering 101 languages. We describe
+the design and modified training of mT5 and demonstrate its state-of-the-art performance on many multilingual
+benchmarks. All of the code and model checkpoints*
+
+The original code can be found `here <https://github.com/google-research/multilingual-t5>`__.
+
+MT5Config
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.MT5Config
+    :members:
+
+
+MT5Model
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.MT5Model
+    :members:
+
+
+MT5ForConditionalGeneration
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.MT5ForConditionalGeneration
+    :members:
+
+
+TFMT5Model
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFMT5Model
+    :members:
+
+
+TFMT5ForConditionalGeneration
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFMT5ForConditionalGeneration
+    :members:

docs/source/model_doc/pegasus.rst

@@ -31,10 +31,19 @@ All the `checkpoints <https://huggingface.co/models?search=pegasus>`__ are fine-
 - Each checkpoint is 2.2 GB on disk and 568M parameters.
 - FP16 is not supported (help/ideas on this appreciated!).
 - Summarizing xsum in fp32 takes about 400ms/sample, with default parameters on a v100 GPU.
-- For XSUM, The paper reports rouge1,rouge2, rougeL of paper: 47.21/24.56/39.25. As of Aug 9, this port scores
-  46.91/24.34/39.1.
+- Full replication results and correctly pre-processed data can be found in this `Issue
+  <https://github.com/huggingface/transformers/issues/6844#issue-689259666>`__.
+- `Distilled checkpoints <https://huggingface.co/models?search=distill-pegasus>`__ are described in this `paper
+  <https://arxiv.org/abs/2010.13002>`__.
 
-The gap is likely because of different alpha/length_penalty implementations in beam search.
+Examples
+_______________________________________________________________________________________________________________________
+
+- `Script <https://github.com/huggingface/transformers/blob/master/examples/seq2seq/finetune_pegasus_xsum.sh>`__ to
+  fine-tune pegasus on the XSUM dataset. Data download instructions at `examples/seq2seq/
+  <https://github.com/huggingface/transformers/blob/master/examples/seq2seq/README.md>`__.
+- FP16 is not supported (help/ideas on this appreciated!).
+- The adafactor optimizer is recommended for pegasus fine-tuning.
 
 
 Implementation Notes
@@ -45,7 +54,7 @@ Implementation Notes
 - Some key configuration differences:
 
     - static, sinusoidal position embeddings
-    - no :obj:`layernorm_embedding` (:obj`PegasusConfig.normalize_embedding=False`)
+    - no :obj:`layernorm_embedding` (:obj:`PegasusConfig.normalize_embedding=False`)
     - the model starts generating with pad_token_id (which has 0 token_embedding) as the prefix.
     - more beams are used (:obj:`num_beams=8`)
 - All pretrained pegasus checkpoints are the same besides three attributes: :obj:`tokenizer.model_max_length` (maximum
@@ -69,7 +78,7 @@ Usage Example
     torch_device = 'cuda' if torch.cuda.is_available() else 'cpu'
     tokenizer = PegasusTokenizer.from_pretrained(model_name)
     model = PegasusForConditionalGeneration.from_pretrained(model_name).to(torch_device)
-    batch = tokenizer.prepare_seq2seq_batch(src_text, truncation=True, padding='longest').to(torch_device)
+    batch = tokenizer.prepare_seq2seq_batch(src_text, truncation=True, padding='longest', return_tensors="pt").to(torch_device)
     translated = model.generate(**batch)
     tgt_text = tokenizer.batch_decode(translated, skip_special_tokens=True)
     assert tgt_text[0] == "California's largest electricity provider has turned off power to hundreds of thousands of customers."

docs/source/model_doc/prophetnet.rst

@@ -17,7 +17,7 @@ the next token.
 
 The abstract from the paper is the following:
 
-*In this paper, we present a new sequence-to-sequence pre-training model called ProphetNet, which introduces a novel
+*In this paper, we present a new sequence-to-sequence pretraining model called ProphetNet, which introduces a novel
 self-supervised objective named future n-gram prediction and the proposed n-stream self-attention mechanism. Instead of
 the optimization of one-step ahead prediction in traditional sequence-to-sequence model, the ProphetNet is optimized by
 n-step ahead prediction which predicts the next n tokens simultaneously based on previous context tokens at each time
@@ -25,7 +25,7 @@ step. The future n-gram prediction explicitly encourages the model to plan for t
 overfitting on strong local correlations. We pre-train ProphetNet using a base scale dataset (16GB) and a large scale
 dataset (160GB) respectively. Then we conduct experiments on CNN/DailyMail, Gigaword, and SQuAD 1.1 benchmarks for
 abstractive summarization and question generation tasks. Experimental results show that ProphetNet achieves new
-state-of-the-art results on all these datasets compared to the models using the same scale pre-training corpus.*
+state-of-the-art results on all these datasets compared to the models using the same scale pretraining corpus.*
 
 The Authors' code can be found `here <https://github.com/microsoft/ProphetNet>`__.
 
@@ -47,16 +47,16 @@ ProphetNetTokenizer
 ProphetNet specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_prophetnet.ProphetNetSeq2SeqLMOutput
+.. autoclass:: transformers.models.prophetnet.modeling_prophetnet.ProphetNetSeq2SeqLMOutput
     :members:
 
-.. autoclass:: transformers.modeling_prophetnet.ProphetNetSeq2SeqModelOutput
+.. autoclass:: transformers.models.prophetnet.modeling_prophetnet.ProphetNetSeq2SeqModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_prophetnet.ProphetNetDecoderModelOutput
+.. autoclass:: transformers.models.prophetnet.modeling_prophetnet.ProphetNetDecoderModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_prophetnet.ProphetNetDecoderLMOutput
+.. autoclass:: transformers.models.prophetnet.modeling_prophetnet.ProphetNetDecoderLMOutput
     :members:
 
 ProphetNetModel

docs/source/model_doc/rag.rst

@@ -50,10 +50,10 @@ RagTokenizer
 Rag specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_rag.RetrievAugLMMarginOutput
+.. autoclass:: transformers.models.rag.modeling_rag.RetrievAugLMMarginOutput
     :members:
 
-.. autoclass:: transformers.modeling_rag.RetrievAugLMOutput
+.. autoclass:: transformers.models.rag.modeling_rag.RetrievAugLMOutput
     :members:
 
 RagRetriever

docs/source/model_doc/roberta.rst

@@ -146,3 +146,10 @@ TFRobertaForQuestionAnswering
 
 .. autoclass:: transformers.TFRobertaForQuestionAnswering
     :members: call
+
+
+FlaxRobertaModel
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.FlaxRobertaModel
+    :members: __call__

docs/source/model_doc/t5.rst

@@ -17,7 +17,7 @@ The abstract from the paper is the following:
 task, has emerged as a powerful technique in natural language processing (NLP). The effectiveness of transfer learning
 has given rise to a diversity of approaches, methodology, and practice. In this paper, we explore the landscape of
 transfer learning techniques for NLP by introducing a unified framework that converts every language problem into a
-text-to-text format. Our systematic study compares pre-training objectives, architectures, unlabeled datasets, transfer
+text-to-text format. Our systematic study compares pretraining objectives, architectures, unlabeled datasets, transfer
 approaches, and other factors on dozens of language understanding tasks. By combining the insights from our exploration
 with scale and our new "Colossal Clean Crawled Corpus", we achieve state-of-the-art results on many benchmarks covering
 summarization, question answering, text classification, and more. To facilitate future work on transfer learning for
@@ -64,7 +64,7 @@ token. T5 can be trained / fine-tuned both in a supervised and unsupervised fash
   input_ids = tokenizer('The <extra_id_0> walks in <extra_id_1> park', return_tensors='pt').input_ids
   labels = tokenizer('<extra_id_0> cute dog <extra_id_1> the <extra_id_2>', return_tensors='pt').input_ids
   # the forward function automatically creates the correct decoder_input_ids
-  loss = model(input_ids=input_ids, labels=labels, return_dict=True).loss
+  loss = model(input_ids=input_ids, labels=labels).loss
 
 - Supervised training
 
@@ -77,7 +77,7 @@ token. T5 can be trained / fine-tuned both in a supervised and unsupervised fash
   input_ids = tokenizer('translate English to German: The house is wonderful.', return_tensors='pt').input_ids
   labels = tokenizer('Das Haus ist wunderbar.', return_tensors='pt').input_ids
   # the forward function automatically creates the correct decoder_input_ids
-  loss = model(input_ids=input_ids, labels=labels, return_dict=True).loss
+  loss = model(input_ids=input_ids, labels=labels).loss
 
 
 T5Config

docs/source/model_doc/transformerxl.rst

@@ -49,16 +49,16 @@ TransfoXLTokenizer
 TransfoXL specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_transfo_xl.TransfoXLModelOutput
+.. autoclass:: transformers.models.transfo_xl.modeling_transfo_xl.TransfoXLModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_transfo_xl.TransfoXLLMHeadModelOutput
+.. autoclass:: transformers.models.transfo_xl.modeling_transfo_xl.TransfoXLLMHeadModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_transfo_xl.TFTransfoXLModelOutput
+.. autoclass:: transformers.models.transfo_xl.modeling_tf_transfo_xl.TFTransfoXLModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_transfo_xl.TFTransfoXLLMHeadModelOutput
+.. autoclass:: transformers.models.transfo_xl.modeling_tf_transfo_xl.TFTransfoXLLMHeadModelOutput
     :members:
 
 

docs/source/model_doc/xlm.rst

@@ -50,7 +50,7 @@ XLMTokenizer
 XLM specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_xlm.XLMForQuestionAnsweringOutput
+.. autoclass:: transformers.models.xlm.modeling_xlm.XLMForQuestionAnsweringOutput
     :members:
 
 

docs/source/model_doc/xlmprophetnet.rst

@@ -19,7 +19,7 @@ just the next token. Its architecture is identical to ProhpetNet, but the model
 
 The abstract from the paper is the following:
 
-*In this paper, we present a new sequence-to-sequence pre-training model called ProphetNet, which introduces a novel
+*In this paper, we present a new sequence-to-sequence pretraining model called ProphetNet, which introduces a novel
 self-supervised objective named future n-gram prediction and the proposed n-stream self-attention mechanism. Instead of
 the optimization of one-step ahead prediction in traditional sequence-to-sequence model, the ProphetNet is optimized by
 n-step ahead prediction which predicts the next n tokens simultaneously based on previous context tokens at each time
@@ -27,7 +27,7 @@ step. The future n-gram prediction explicitly encourages the model to plan for t
 overfitting on strong local correlations. We pre-train ProphetNet using a base scale dataset (16GB) and a large scale
 dataset (160GB) respectively. Then we conduct experiments on CNN/DailyMail, Gigaword, and SQuAD 1.1 benchmarks for
 abstractive summarization and question generation tasks. Experimental results show that ProphetNet achieves new
-state-of-the-art results on all these datasets compared to the models using the same scale pre-training corpus.*
+state-of-the-art results on all these datasets compared to the models using the same scale pretraining corpus.*
 
 The Authors' code can be found `here <https://github.com/microsoft/ProphetNet>`__.
 

docs/source/model_doc/xlnet.rst

@@ -53,43 +53,43 @@ XLNetTokenizer
 XLNet specific outputs
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.modeling_xlnet.XLNetModelOutput
+.. autoclass:: transformers.models.xlnet.modeling_xlnet.XLNetModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_xlnet.XLNetLMHeadModelOutput
+.. autoclass:: transformers.models.xlnet.modeling_xlnet.XLNetLMHeadModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_xlnet.XLNetForSequenceClassificationOutput
+.. autoclass:: transformers.models.xlnet.modeling_xlnet.XLNetForSequenceClassificationOutput
     :members:
 
-.. autoclass:: transformers.modeling_xlnet.XLNetForMultipleChoiceOutput
+.. autoclass:: transformers.models.xlnet.modeling_xlnet.XLNetForMultipleChoiceOutput
     :members:
 
-.. autoclass:: transformers.modeling_xlnet.XLNetForTokenClassificationOutput
+.. autoclass:: transformers.models.xlnet.modeling_xlnet.XLNetForTokenClassificationOutput
     :members:
 
-.. autoclass:: transformers.modeling_xlnet.XLNetForQuestionAnsweringSimpleOutput
+.. autoclass:: transformers.models.xlnet.modeling_xlnet.XLNetForQuestionAnsweringSimpleOutput
     :members:
 
-.. autoclass:: transformers.modeling_xlnet.XLNetForQuestionAnsweringOutput
+.. autoclass:: transformers.models.xlnet.modeling_xlnet.XLNetForQuestionAnsweringOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_xlnet.TFXLNetModelOutput
+.. autoclass:: transformers.models.xlnet.modeling_tf_xlnet.TFXLNetModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_xlnet.TFXLNetLMHeadModelOutput
+.. autoclass:: transformers.models.xlnet.modeling_tf_xlnet.TFXLNetLMHeadModelOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_xlnet.TFXLNetForSequenceClassificationOutput
+.. autoclass:: transformers.models.xlnet.modeling_tf_xlnet.TFXLNetForSequenceClassificationOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_xlnet.TFXLNetForMultipleChoiceOutput
+.. autoclass:: transformers.models.xlnet.modeling_tf_xlnet.TFXLNetForMultipleChoiceOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_xlnet.TFXLNetForTokenClassificationOutput
+.. autoclass:: transformers.models.xlnet.modeling_tf_xlnet.TFXLNetForTokenClassificationOutput
     :members:
 
-.. autoclass:: transformers.modeling_tf_xlnet.TFXLNetForQuestionAnsweringSimpleOutput
+.. autoclass:: transformers.models.xlnet.modeling_tf_xlnet.TFXLNetForQuestionAnsweringSimpleOutput
     :members:
 
 

docs/source/model_sharing.rst

@@ -37,7 +37,7 @@ For instance:
 
 .. code-block::
 
-    >>> tokenizer = AutoTokenizer.from_pretrained(
+    >>> model = AutoModel.from_pretrained(
     >>>   "julien-c/EsperBERTo-small",
     >>>   revision="v2.0.1" # tag name, or branch name, or commit hash
     >>> )
@@ -46,31 +46,42 @@ Basic steps
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 In order to upload a model, you'll need to first create a git repo. This repo will live on the model hub, allowing
-users to clone it and you (and your organization members) to push to it. First, you should ensure you are logged in the
-``transformers-cli``:
+users to clone it and you (and your organization members) to push to it.
 
-Go in a terminal and run the following command. It should be in the virtual environment where you installed 🤗
+You can create a model repo directly from the website, `here <https://huggingface.co/new>`.
+
+Alternatively, you can use the ``transformers-cli``. The next steps describe that process:
+
+Go to a terminal and run the following command. It should be in the virtual environment where you installed 🤗
 Transformers, since that command :obj:`transformers-cli` comes from the library.
 
-.. code-block::
+.. code-block:: bash
 
     transformers-cli login
 
 
 Once you are logged in with your model hub credentials, you can start building your repositories. To create a repo:
 
-.. code-block::
+.. code-block:: bash
 
     transformers-cli repo create your-model-name
 
-This creates a repo on the model hub, which can be cloned. You can then add/remove from that repo as you would with any
-other git repo.
+This creates a repo on the model hub, which can be cloned.
 
-.. code-block::
+.. code-block:: bash
 
     git clone https://huggingface.co/username/your-model-name
 
-    # Then commit as usual
+    # Make sure you have git-lfs installed
+    # (https://git-lfs.github.com/)
+    git lfs install
+
+When you have your local clone of your repo and lfs installed, you can then add/remove from that clone as you would
+with any other git repo.
+
+.. code-block:: bash
+
+    # Commit as usual
     cd your-model-name
     echo "hello" >> README.md
     git add . && git commit -m "Update from $USER"
@@ -159,24 +170,25 @@ Or, if you're using the Trainer API
 .. code-block::
 
     >>> trainer.save_model("path/to/awesome-name-you-picked")
+    >>> tokenizer.save_pretrained("path/to/repo/clone/your-model-name")
 
 You can then add these files to the staging environment and verify that they have been correctly staged with the ``git
 status`` command:
 
-.. code-block::
+.. code-block:: bash
 
     git add --all
     git status
 
 Finally, the files should be comitted:
 
-.. code-block::
+.. code-block:: bash
 
     git commit -m "First version of the your-model-name model and tokenizer."
 
 And pushed to the remote:
 
-.. code-block::
+.. code-block:: bash
 
     git push
 
@@ -199,7 +211,7 @@ don't forget to link to its model card so that people can fully trace how your m
 If you have never made a pull request to the 🤗 Transformers repo, look at the :doc:`contributing guide <contributing>`
 to see the steps to follow.
 
-.. Note::
+.. note::
 
     You can also send your model card in the folder you uploaded with the CLI by placing it in a `README.md` file
     inside `path/to/awesome-name-you-picked/`.
@@ -225,3 +237,47 @@ You may specify a revision by using the ``revision`` flag in the ``from_pretrain
     >>>   "julien-c/EsperBERTo-small",
     >>>   revision="v2.0.1" # tag name, or branch name, or commit hash
     >>> )
+
+Workflow in a Colab notebook
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+If you're in a Colab notebook (or similar) with no direct access to a terminal, here is the workflow you can use to
+upload your model. You can execute each one of them in a cell by adding a ! at the beginning.
+
+First you need to install `git-lfs` in the environment used by the notebook:
+
+.. code-block:: bash
+
+    sudo apt-get install git-lfs
+
+Then you can use the :obj:`transformers-cli` to create your new repo:
+
+
+.. code-block:: bash
+
+    transformers-cli login
+    transformers-cli repo create your-model-name
+
+Once it's created, you can clone it and configure it (replace username by your username on huggingface.co):
+
+.. code-block:: bash
+
+    git clone https://username:password@huggingface.co/username/your-model-name
+    # Alternatively if you have a token,
+    # you can use it instead of your password
+    git clone https://username:token@huggingface.co/username/your-model-name
+
+    cd your-model-name
+    git lfs install
+    git config --global user.email "email@example.com"
+    # Tip: using the same email than for your huggingface.co account will link your commits to your profile
+    git config --global user.name "Your name"
+
+Once you've saved your model inside, and your clone is setup with the right remote URL, you can add it and push it with
+usual git commands.
+
+.. code-block:: bash
+
+    git add .
+    git commit -m "Initial commit"
+    git push

docs/source/model_summary.rst

@@ -527,10 +527,10 @@ Pegasus
 <https://arxiv.org/pdf/1912.08777.pdf>`_, Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu on Dec 18, 2019.
 
 Sequence-to-sequence model with the same encoder-decoder model architecture as BART. Pegasus is pre-trained jointly on
-two self-supervised objective functions: Masked Language Modeling (MLM) and a novel summarization specific pre-training
+two self-supervised objective functions: Masked Language Modeling (MLM) and a novel summarization specific pretraining
 objective, called Gap Sentence Generation (GSG).
 
-  * MLM: encoder input tokens are randomely replaced by a mask tokens and have to be predicted by the encoder (like in
+  * MLM: encoder input tokens are randomly replaced by a mask tokens and have to be predicted by the encoder (like in
     BERT)
   * GSG: whole encoder input sentences are replaced by a second mask token and fed to the decoder, but which has a
     causal mask to hide the future words like a regular auto-regressive transformer decoder.
@@ -560,6 +560,7 @@ A framework for translation models, using the same models as BART
 
 The library provides a version of this model for conditional generation.
 
+
 T5
 -----------------------------------------------------------------------------------------------------------------------
 
@@ -592,6 +593,28 @@ For instance, if we have the sentence “My dog is very cute .”, and we decide
 
 The library provides a version of this model for conditional generation.
 
+
+MT5
+-----------------------------------------------------------------------------------------------------------------------
+
+.. raw:: html
+
+   <a href="https://huggingface.co/models?filter=mt5">
+       <img alt="Models" src="https://img.shields.io/badge/All_model_pages-mt5-blueviolet">
+   </a>
+   <a href="model_doc/mt5.html">
+       <img alt="Doc" src="https://img.shields.io/badge/Model_documentation-mt5-blueviolet">
+   </a>
+
+`mT5: A massively multilingual pre-trained text-to-text transformer <https://arxiv.org/abs/2010.11934>`_, Linting Xue
+et al.
+
+The model architecture is same as T5. mT5's pretraining objective includes T5's self-supervised training, but not T5's
+supervised training. mT5 is trained on 101 languages.
+
+The library provides a version of this model for conditional generation.
+
+
 MBart
 -----------------------------------------------------------------------------------------------------------------------
 
@@ -607,8 +630,8 @@ MBart
 `Multilingual Denoising Pre-training for Neural Machine Translation <https://arxiv.org/abs/2001.08210>`_ by Yinhan Liu,
 Jiatao Gu, Naman Goyal, Xian Li, Sergey Edunov Marjan Ghazvininejad, Mike Lewis, Luke Zettlemoyer.
 
-The model architecture and pre-training objective is same as BART, but MBart is trained on 25 languages and is intended
-for supervised and unsupervised machine translation. MBart is one of the first methods for pre-training a complete
+The model architecture and pretraining objective is same as BART, but MBart is trained on 25 languages and is intended
+for supervised and unsupervised machine translation. MBart is one of the first methods for pretraining a complete
 sequence-to-sequence model by denoising full texts in multiple languages,
 
 The library provides a version of this model for conditional generation.
@@ -635,7 +658,7 @@ ProphetNet
 `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, Ming Zhou.
 
-ProphetNet introduces a novel *sequence-to-sequence* pre-training objective, called *future n-gram prediction*. In
+ProphetNet introduces a novel *sequence-to-sequence* pretraining objective, called *future n-gram prediction*. In
 future n-gram prediction, the model predicts the next n tokens simultaneously based on previous context tokens at each
 time step instead instead of just the single next token. The future n-gram prediction explicitly encourages the model
 to plan for the future tokens and prevent overfitting on strong local correlations. The model architecture is based on
@@ -660,8 +683,8 @@ XLM-ProphetNet
 `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, Ming Zhou.
 
-XLM-ProphetNet's model architecture and pre-training objective is same as ProphetNet, but XLM-ProphetNet was
-pre-trained on the cross-lingual dataset `XGLUE <https://arxiv.org/abs/2004.01401>`__.
+XLM-ProphetNet's model architecture and pretraining objective is same as ProphetNet, but XLM-ProphetNet was pre-trained
+on the cross-lingual dataset `XGLUE <https://arxiv.org/abs/2004.01401>`__.
 
 The library provides a pre-trained version of this model for multi-lingual conditional generation and fine-tuned
 versions for headline generation and question generation, respectively.

docs/source/multilingual.rst

@@ -109,7 +109,7 @@ XLM-RoBERTa
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 XLM-RoBERTa was trained on 2.5TB of newly created clean CommonCrawl data in 100 languages. It provides strong gains
-over previously released multi-lingual models like mBERT or XLM on downstream taks like classification, sequence
+over previously released multi-lingual models like mBERT or XLM on downstream tasks like classification, sequence
 labeling and question answering.
 
 Two XLM-RoBERTa checkpoints can be used for multi-lingual tasks:

docs/source/perplexity.rst

@@ -62,7 +62,7 @@ sliding the context window so that the model has more context when making each p
 This is a closer approximation to the true decomposition of the sequence probability and will typically yield a more
 favorable score. The downside is that it requires a separate forward pass for each token in the corpus. A good
 practical compromise is to employ a strided sliding window, moving the context by larger strides rather than sliding by
-1 token a time. This allows computation to procede much faster while still giving the model a large context to make
+1 token a time. This allows computation to proceed much faster while still giving the model a large context to make
 predictions at each step.
 
 Example: Calculating perplexity with GPT-2 in 🤗 Transformers

docs/source/pretrained_models.rst

@@ -3,11 +3,11 @@ Pretrained models
 
 Here is the full list of the currently provided pretrained models together with a short presentation of each model.
 
-For a list that includes community-uploaded models, refer to `https://huggingface.co/models
+For a list that includes all community-uploaded models, refer to `https://huggingface.co/models
 <https://huggingface.co/models>`__.
 
 +--------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
-| Architecture       | Shortcut name                                              | Details of the model                                                                                                                  |
+| Architecture       | Model id                                                   | Details of the model                                                                                                                  |
 +====================+============================================================+=======================================================================================================================================+
 | BERT               | ``bert-base-uncased``                                      | | 12-layer, 768-hidden, 12-heads, 110M parameters.                                                                                    |
 |                    |                                                            | | Trained on lower-cased English text.                                                                                                |

docs/source/task_summary.rst

@@ -89,7 +89,7 @@ each other. The process is the following:
     >>> import torch
 
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-base-cased-finetuned-mrpc")
-    >>> model = AutoModelForSequenceClassification.from_pretrained("bert-base-cased-finetuned-mrpc", return_dict=True)
+    >>> model = AutoModelForSequenceClassification.from_pretrained("bert-base-cased-finetuned-mrpc")
 
     >>> classes = ["not paraphrase", "is paraphrase"]
 
@@ -122,7 +122,7 @@ each other. The process is the following:
     >>> import tensorflow as tf
 
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-base-cased-finetuned-mrpc")
-    >>> model = TFAutoModelForSequenceClassification.from_pretrained("bert-base-cased-finetuned-mrpc", return_dict=True)
+    >>> model = TFAutoModelForSequenceClassification.from_pretrained("bert-base-cased-finetuned-mrpc")
 
     >>> classes = ["not paraphrase", "is paraphrase"]
 
@@ -211,7 +211,7 @@ Here is an example of question answering using a model and a tokenizer. The proc
     >>> import torch
 
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad")
-    >>> model = AutoModelForQuestionAnswering.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad", return_dict=True)
+    >>> model = AutoModelForQuestionAnswering.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad")
 
     >>> text = r"""
     ... 🤗 Transformers (formerly known as pytorch-transformers and pytorch-pretrained-bert) provides general-purpose
@@ -231,7 +231,9 @@ Here is an example of question answering using a model and a tokenizer. The proc
     ...     input_ids = inputs["input_ids"].tolist()[0]
     ...
     ...     text_tokens = tokenizer.convert_ids_to_tokens(input_ids)
-    ...     answer_start_scores, answer_end_scores = model(**inputs)
+    ...     outputs = model(**inputs)
+    ...     answer_start_scores = outputs.start_logits
+    ...     answer_end_scores = outputs.end_logits
     ...
     ...     answer_start = torch.argmax(
     ...         answer_start_scores
@@ -253,7 +255,7 @@ Here is an example of question answering using a model and a tokenizer. The proc
     >>> import tensorflow as tf
 
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad")
-    >>> model = TFAutoModelForQuestionAnswering.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad", return_dict=True)
+    >>> model = TFAutoModelForQuestionAnswering.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad")
 
     >>> text = r"""
     ... 🤗 Transformers (formerly known as pytorch-transformers and pytorch-pretrained-bert) provides general-purpose
@@ -273,7 +275,9 @@ Here is an example of question answering using a model and a tokenizer. The proc
     ...     input_ids = inputs["input_ids"].numpy()[0]
     ...
     ...     text_tokens = tokenizer.convert_ids_to_tokens(input_ids)
-    ...     answer_start_scores, answer_end_scores = model(inputs)
+    ...     outputs = model(inputs)
+    ...     answer_start_scores = outputs.start_logits
+    ...     answer_end_scores = outputs.end_logits
     ...
     ...     answer_start = tf.argmax(
     ...         answer_start_scores, axis=1
@@ -301,7 +305,7 @@ Language modeling is the task of fitting a model to a corpus, which can be domai
 transformer-based models are trained using a variant of language modeling, e.g. BERT with masked language modeling,
 GPT-2 with causal language modeling.
 
-Language modeling can be useful outside of pre-training as well, for example to shift the model distribution to be
+Language modeling can be useful outside of pretraining as well, for example to shift the model distribution to be
 domain-specific: using a language model trained over a very large corpus, and then fine-tuning it to a news dataset or
 on scientific papers e.g. `LysandreJik/arxiv-nlp <https://huggingface.co/lysandre/arxiv-nlp>`__.
 
@@ -373,7 +377,7 @@ Here is an example of doing masked language modeling using a model and a tokeniz
     >>> import torch
 
     >>> tokenizer = AutoTokenizer.from_pretrained("distilbert-base-cased")
-    >>> model = AutoModelWithLMHead.from_pretrained("distilbert-base-cased", return_dict=True)
+    >>> model = AutoModelWithLMHead.from_pretrained("distilbert-base-cased")
 
     >>> sequence = f"Distilled models are smaller than the models they mimic. Using them instead of the large versions would help {tokenizer.mask_token} our carbon footprint."
 
@@ -389,7 +393,7 @@ Here is an example of doing masked language modeling using a model and a tokeniz
     >>> import tensorflow as tf
 
     >>> tokenizer = AutoTokenizer.from_pretrained("distilbert-base-cased")
-    >>> model = TFAutoModelWithLMHead.from_pretrained("distilbert-base-cased", return_dict=True)
+    >>> model = TFAutoModelWithLMHead.from_pretrained("distilbert-base-cased")
 
     >>> sequence = f"Distilled models are smaller than the models they mimic. Using them instead of the large versions would help {tokenizer.mask_token} our carbon footprint."
 
@@ -437,7 +441,7 @@ of tokens.
     >>> from torch.nn import functional as F
 
     >>> tokenizer = AutoTokenizer.from_pretrained("gpt2")
-    >>> model = AutoModelWithLMHead.from_pretrained("gpt2", return_dict=True)
+    >>> model = AutoModelWithLMHead.from_pretrained("gpt2")
 
     >>> sequence = f"Hugging Face is based in DUMBO, New York City, and "
 
@@ -461,7 +465,7 @@ of tokens.
     >>> import tensorflow as tf
 
     >>> tokenizer = AutoTokenizer.from_pretrained("gpt2")
-    >>> model = TFAutoModelWithLMHead.from_pretrained("gpt2", return_dict=True)
+    >>> model = TFAutoModelWithLMHead.from_pretrained("gpt2")
 
     >>> sequence = f"Hugging Face is based in DUMBO, New York City, and "
 
@@ -513,14 +517,14 @@ Here, the model generates a random text with a total maximal length of *50* toke
 concerned, I will"*. The default arguments of ``PreTrainedModel.generate()`` can be directly overridden in the
 pipeline, as is shown above for the argument ``max_length``.
 
-Here is an example of text generation using ``XLNet`` and its tokenzier.
+Here is an example of text generation using ``XLNet`` and its tokenizer.
 
 .. code-block::
 
     >>> ## PYTORCH CODE
     >>> from transformers import AutoModelWithLMHead, AutoTokenizer
 
-    >>> model = AutoModelWithLMHead.from_pretrained("xlnet-base-cased", return_dict=True)
+    >>> model = AutoModelWithLMHead.from_pretrained("xlnet-base-cased")
     >>> tokenizer = AutoTokenizer.from_pretrained("xlnet-base-cased")
 
     >>> # Padding text helps XLNet with short prompts - proposed by Aman Rusia in https://github.com/rusiaaman/XLNet-gen#methodology
@@ -545,7 +549,7 @@ Here is an example of text generation using ``XLNet`` and its tokenzier.
     >>> ## TENSORFLOW CODE
     >>> from transformers import TFAutoModelWithLMHead, AutoTokenizer
 
-    >>> model = TFAutoModelWithLMHead.from_pretrained("xlnet-base-cased", return_dict=True)
+    >>> model = TFAutoModelWithLMHead.from_pretrained("xlnet-base-cased")
     >>> tokenizer = AutoTokenizer.from_pretrained("xlnet-base-cased")
 
     >>> # Padding text helps XLNet with short prompts - proposed by Aman Rusia in https://github.com/rusiaaman/XLNet-gen#methodology
@@ -664,7 +668,7 @@ Here is an example of doing named entity recognition, using a model and a tokeni
     >>> from transformers import AutoModelForTokenClassification, AutoTokenizer
     >>> import torch
 
-    >>> model = AutoModelForTokenClassification.from_pretrained("dbmdz/bert-large-cased-finetuned-conll03-english", return_dict=True)
+    >>> model = AutoModelForTokenClassification.from_pretrained("dbmdz/bert-large-cased-finetuned-conll03-english")
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-base-cased")
 
     >>> label_list = [
@@ -692,7 +696,7 @@ Here is an example of doing named entity recognition, using a model and a tokeni
     >>> from transformers import TFAutoModelForTokenClassification, AutoTokenizer
     >>> import tensorflow as tf
 
-    >>> model = TFAutoModelForTokenClassification.from_pretrained("dbmdz/bert-large-cased-finetuned-conll03-english", return_dict=True)
+    >>> model = TFAutoModelForTokenClassification.from_pretrained("dbmdz/bert-large-cased-finetuned-conll03-english")
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-base-cased")
 
     >>> label_list = [
@@ -790,7 +794,7 @@ CNN / Daily Mail), it yields very good results.
     >>> ## PYTORCH CODE
     >>> from transformers import AutoModelWithLMHead, AutoTokenizer
 
-    >>> model = AutoModelWithLMHead.from_pretrained("t5-base", return_dict=True)
+    >>> model = AutoModelWithLMHead.from_pretrained("t5-base")
     >>> tokenizer = AutoTokenizer.from_pretrained("t5-base")
 
     >>> # T5 uses a max_length of 512 so we cut the article to 512 tokens.
@@ -799,7 +803,7 @@ CNN / Daily Mail), it yields very good results.
     >>> ## TENSORFLOW CODE
     >>> from transformers import TFAutoModelWithLMHead, AutoTokenizer
 
-    >>> model = TFAutoModelWithLMHead.from_pretrained("t5-base", return_dict=True)
+    >>> model = TFAutoModelWithLMHead.from_pretrained("t5-base")
     >>> tokenizer = AutoTokenizer.from_pretrained("t5-base")
 
     >>> # T5 uses a max_length of 512 so we cut the article to 512 tokens.
@@ -834,7 +838,7 @@ Here is an example of doing translation using a model and a tokenizer. The proce
 
 1. Instantiate a tokenizer and a model from the checkpoint name. Summarization is usually done using an encoder-decoder
    model, such as ``Bart`` or ``T5``.
-2. Define the article that should be summarizaed.
+2. Define the article that should be summarized.
 3. Add the T5 specific prefix "translate English to German: "
 4. Use the ``PreTrainedModel.generate()`` method to perform the translation.
 
@@ -843,7 +847,7 @@ Here is an example of doing translation using a model and a tokenizer. The proce
     >>> ## PYTORCH CODE
     >>> from transformers import AutoModelWithLMHead, AutoTokenizer
 
-    >>> model = AutoModelWithLMHead.from_pretrained("t5-base", return_dict=True)
+    >>> model = AutoModelWithLMHead.from_pretrained("t5-base")
     >>> tokenizer = AutoTokenizer.from_pretrained("t5-base")
 
     >>> inputs = tokenizer.encode("translate English to German: Hugging Face is a technology company based in New York and Paris", return_tensors="pt")
@@ -851,7 +855,7 @@ Here is an example of doing translation using a model and a tokenizer. The proce
     >>> ## TENSORFLOW CODE
     >>> from transformers import TFAutoModelWithLMHead, AutoTokenizer
 
-    >>> model = TFAutoModelWithLMHead.from_pretrained("t5-base", return_dict=True)
+    >>> model = TFAutoModelWithLMHead.from_pretrained("t5-base")
     >>> tokenizer = AutoTokenizer.from_pretrained("t5-base")
 
     >>> inputs = tokenizer.encode("translate English to German: Hugging Face is a technology company based in New York and Paris", return_tensors="tf")

docs/source/testing.rst

@@ -405,32 +405,32 @@ decorators are used to set the requirements of tests CPU/GPU/TPU-wise:
 
 * ``require_torch`` - this test will run only under torch
 * ``require_torch_gpu`` - as ``require_torch`` plus requires at least 1 GPU
-* ``require_torch_multigpu`` - as ``require_torch`` plus requires at least 2 GPUs
-* ``require_torch_non_multigpu`` - as ``require_torch`` plus requires 0 or 1 GPUs
+* ``require_torch_multi_gpu`` - as ``require_torch`` plus requires at least 2 GPUs
+* ``require_torch_non_multi_gpu`` - as ``require_torch`` plus requires 0 or 1 GPUs
 * ``require_torch_tpu`` - as ``require_torch`` plus requires at least 1 TPU
 
 Let's depict the GPU requirements in the following table:
 
 
-+----------+---------------------------------+
-| n gpus   |  decorator                      |
-+==========+=================================+
-| ``>= 0`` | ``@require_torch``              |
-+----------+---------------------------------+
-| ``>= 1`` | ``@require_torch_gpu``          |
-+----------+---------------------------------+
-| ``>= 2`` | ``@require_torch_multigpu``     |
-+----------+---------------------------------+
-| ``< 2``  | ``@require_torch_non_multigpu`` |
-+----------+---------------------------------+
++----------+----------------------------------+
+| n gpus   |  decorator                       |
++==========+==================================+
+| ``>= 0`` | ``@require_torch``               |
++----------+----------------------------------+
+| ``>= 1`` | ``@require_torch_gpu``           |
++----------+----------------------------------+
+| ``>= 2`` | ``@require_torch_multi_gpu``     |
++----------+----------------------------------+
+| ``< 2``  | ``@require_torch_non_multi_gpu`` |
++----------+----------------------------------+
 
 
 For example, here is a test that must be run only when there are 2 or more GPUs available and pytorch is installed:
 
 .. code-block:: python
 
-    @require_torch_multigpu
-    def test_example_with_multigpu():
+    @require_torch_multi_gpu
+    def test_example_with_multi_gpu():
 
 If a test requires ``tensorflow`` use the ``require_tf`` decorator. For example:
 
@@ -454,7 +454,7 @@ last for them to work correctly. Here is an example of the correct usage:
 .. code-block:: python
 
     @parameterized.expand(...)
-    @require_torch_multigpu
+    @require_torch_multi_gpu
     def test_integration_foo():
 
 This order problem doesn't exist with ``@pytest.mark.parametrize``, you can put it first or last and it will still
@@ -716,11 +716,11 @@ Temporary files and directories
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 Using unique temporary files and directories are essential for parallel test running, so that the tests won't overwrite
-each other's data. Also we want to get the temp files and directories removed at the end of each test that created
+each other's data. Also we want to get the temporary files and directories removed at the end of each test that created
 them. Therefore, using packages like ``tempfile``, which address these needs is essential.
 
-However, when debugging tests, you need to be able to see what goes into the temp file or directory and you want to
-know it's exact path and not having it randomized on every test re-run.
+However, when debugging tests, you need to be able to see what goes into the temporary file or directory and you want
+to know it's exact path and not having it randomized on every test re-run.
 
 A helper class :obj:`transformers.test_utils.TestCasePlus` is best used for such purposes. It's a sub-class of
 :obj:`unittest.TestCase`, so we can easily inherit from it in the test modules.
@@ -736,32 +736,33 @@ Here is an example of its usage:
 
 This code creates a unique temporary directory, and sets :obj:`tmp_dir` to its location.
 
-In this and all the following scenarios the temporary directory will be auto-removed at the end of test, unless
-``after=False`` is passed to the helper function.
-
-* Create a temporary directory of my choice and delete it at the end - useful for debugging when you want to monitor a
-  specific directory:
+* Create a unique temporary dir:
 
 .. code-block:: python
 
     def test_whatever(self):
-        tmp_dir = self.get_auto_remove_tmp_dir(tmp_dir="./tmp/run/test")
+        tmp_dir = self.get_auto_remove_tmp_dir()
 
-* Create a temporary directory of my choice and do not delete it at the end---useful for when you want to look at the
-  temp results:
+``tmp_dir`` will contain the path to the created temporary dir. It will be automatically removed at the end of the
+test.
+
+* Create a temporary dir of my choice, ensure it's empty before the test starts and don't empty it after the test.
 
 .. code-block:: python
 
     def test_whatever(self):
-        tmp_dir = self.get_auto_remove_tmp_dir(tmp_dir="./tmp/run/test", after=False)
+        tmp_dir = self.get_auto_remove_tmp_dir("./xxx")
 
-* Create a temporary directory of my choice and ensure to delete it right away---useful for when you disabled deletion
-  in the previous test run and want to make sure the that temporary directory is empty before the new test is run:
+This is useful for debug when you want to monitor a specific directory and want to make sure the previous tests didn't
+leave any data in there.
 
-.. code-block:: python
+* You can override the default behavior by directly overriding the ``before`` and ``after`` args, leading to one of the
+  following behaviors:
 
-   def test_whatever(self):
-        tmp_dir = self.get_auto_remove_tmp_dir(tmp_dir="./tmp/run/test", before=True)
+    - ``before=True``: the temporary dir will always be cleared at the beginning of the test.
+    - ``before=False``: if the temporary dir already existed, any existing files will remain there.
+    - ``after=True``: the temporary dir will always be deleted at the end of the test.
+    - ``after=False``: the temporary dir will always be left intact at the end of the test.
 
 .. note::
    In order to run the equivalent of ``rm -r`` safely, only subdirs of the project repository checkout are allowed if
@@ -815,7 +816,7 @@ or the ``xfail`` way:
     @pytest.mark.xfail
     def test_feature_x():
 
-Here is how to skip a test based on some internal check inside the test:
+- Here is how to skip a test based on some internal check inside the test:
 
 .. code-block:: python
 
@@ -838,7 +839,7 @@ or the ``xfail`` way:
     def test_feature_x():
         pytest.xfail("expected to fail until bug XYZ is fixed")
 
-Here is how to skip all tests in a module if some import is missing:
+- Here is how to skip all tests in a module if some import is missing:
 
 .. code-block:: python
 
@@ -1054,7 +1055,7 @@ If you need to validate the output of a logger, you can use :obj:`CaptureLogger`
 
     msg = "Testing 1, 2, 3"
     logging.set_verbosity_info()
-    logger = logging.get_logger("transformers.tokenization_bart")
+    logger = logging.get_logger("transformers.models.bart.tokenization_bart")
     with CaptureLogger(logger) as cl:
         logger.info(msg)
     assert cl.out, msg+"\n"

docs/source/tokenizer_summary.rst

@@ -1,223 +1,243 @@
-Tokenizer summary
+Summary of the tokenizers
 -----------------------------------------------------------------------------------------------------------------------
 
-In this page, we will have a closer look at tokenization. As we saw in :doc:`the preprocessing tutorial
-<preprocessing>`, tokenizing a text is splitting it into words or subwords, which then are converted to ids. The second
-part is pretty straightforward, here we will focus on the first part. More specifically, we will look at the three main
-different kinds of tokenizers used in 🤗 Transformers: :ref:`Byte-Pair Encoding (BPE) <byte-pair-encoding>`,
-:ref:`WordPiece <wordpiece>` and :ref:`SentencePiece <sentencepiece>`, and provide examples of models using each of
-those.
+On this page, we will have a closer look at tokenization. As we saw in :doc:`the preprocessing tutorial
+<preprocessing>`, tokenizing a text is splitting it into words or subwords, which then are converted to ids through a
+look-up table. Converting words or subwords to ids is straightforward, so in this summary, we will focus on splitting a
+text into words or subwords (i.e. tokenizing a text). More specifically, we will look at the three main types of
+tokenizers used in 🤗 Transformers: :ref:`Byte-Pair Encoding (BPE) <byte-pair-encoding>`, :ref:`WordPiece <wordpiece>`,
+and :ref:`SentencePiece <sentencepiece>`, and show exemplary which tokenizer type is used by which model.
 
-Note that on each model page, you can look at the documentation of the associated tokenizer to know which of those
-algorithms the pretrained model used. For instance, if we look at :class:`~transformers.BertTokenizer`, we can see it's
-using :ref:`WordPiece <wordpiece>`.
+Note that on each model page, you can look at the documentation of the associated tokenizer to know which tokenizer
+type was used by the pretrained model. For instance, if we look at :class:`~transformers.BertTokenizer`, we can see
+that the model uses :ref:`WordPiece <wordpiece>`.
 
-Introduction to tokenization
+Introduction
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-Splitting a text in smaller chunks is a task that's harder than it looks, and there are multiple ways of doing it. For
-instance, let's look at the sentence "Don't you love 🤗 Transformers? We sure do." A first simple way of tokenizing this
-text is just to split it by spaces, which would give:
+Splitting a text into smaller chunks is a task that is harder than it looks, and there are multiple ways of doing so.
+For instance, let's look at the sentence ``"Don't you love 🤗 Transformers? We sure do."`` A simple way of tokenizing
+this text is to split it by spaces, which would give:
 
 .. code-block::
 
     ["Don't", "you", "love", "🤗", "Transformers?", "We", "sure", "do."]
 
-This is a nice first step, but if we look at the tokens "Transformers?" or "do.", we can see we can do better. Those
-will be different than the tokens "Transformers" and "do" for our model, so we should probably take the punctuation
-into account. This would give:
+This is a sensible first step, but if we look at the tokens ``"Transformers?"`` and ``"do."``, we notice that the
+punctuation is attached to the words ``"Transformer"`` and ``"do"``, which is suboptimal. We should take the
+punctuation into account so that a model does not have to learn a different representation of a word and every possible
+punctuation symbol that could follow it, which would explode the number of representations the model has to learn.
+Taking punctuation into account, tokenizing our exemplary text would give:
 
 .. code-block::
 
     ["Don", "'", "t", "you", "love", "🤗", "Transformers", "?", "We", "sure", "do", "."]
 
-which is better already. One thing that is annoying though is how it dealt with "Don't". "Don't" stands for do not, so
-it should probably be better tokenized as ``["Do", "n't"]``. This is where things start getting more complicated, and
-part of the reason each kind of model has its own tokenizer class. Depending on the rules we apply to split our texts
-into tokens, we'll get different tokenized versions of the same text. And of course, a given pretrained model won't
-perform properly if you don't use the exact same rules as the persons who pretrained it.
+Better. However, it is disadvantageous, how the tokenization dealt with the word ``"Don't"``. ``"Don't"`` stands for
+``"do not"``, so it would be better tokenized as ``["Do", "n't"]``. This is where things start getting complicated, and
+part of the reason each model has its own tokenizer type. Depending on the rules we apply for tokenizing a text, a
+different tokenized output is generated for the same text. A pretrained model only performs properly if you feed it an
+input that was tokenized with the same rules that were used to tokenize its training data.
 
 `spaCy <https://spacy.io/>`__ and `Moses <http://www.statmt.org/moses/?n=Development.GetStarted>`__ are two popular
-rule-based tokenizers. On the text above, they'd output something like:
+rule-based tokenizers. Applying them on our example, *spaCy* and *Moses* would output something like:
 
 .. code-block::
 
     ["Do", "n't", "you", "love", "🤗", "Transformers", "?", "We", "sure", "do", "."]
 
-Space/punctuation-tokenization and rule-based tokenization are both examples of word tokenization, which is splitting a
-sentence into words. While it's the most intuitive way to separate texts in smaller chunks, it can have a problem when
-you have a huge corpus: it usually yields a very big vocabulary (the set of all unique tokens used). :doc:`Transformer
-XL <model_doc/transformerxl>` for instance uses space/punctuation-tokenization, and has a vocabulary size of 267,735!
+As can be seen space and punctuation tokenization, as well as rule-based tokenization, is used here. Space and
+punctuation tokenization and rule-based tokenization are both examples of word tokenization, which is loosely defined
+as splitting sentences into words. While it's the most intuitive way to split texts into smaller chunks, this
+tokenization method can lead to problems for massive text corpora. In this case, space and punctuation tokenization
+usually generates a very big vocabulary (the set of all unique words and tokens used). *E.g.*, :doc:`Transformer XL
+<model_doc/transformerxl>` uses space and punctuation tokenization, resulting in a vocabulary size of 267,735!
 
-A huge vocabulary size means a huge embedding matrix at the start of the model, which will cause memory problems.
-TransformerXL deals with it by using a special kind of embeddings called adaptive embeddings, but in general,
-transformers models rarely have a vocabulary size greater than 50,000, especially if they are trained on a single
-language.
+Such a big vocabulary size forces the model to have an enormous embedding matrix as the input and output layer, which
+causes both an increased memory and time complexity. In general, transformers models rarely have a vocabulary size
+greater than 50,000, especially if they are pretrained only on a single language.
 
-So if tokenizing on words is unsatisfactory, we could go on the opposite direction and simply tokenize on characters.
-While it's very simple and would save a lot of memory, this doesn't allow the model to learn representations of texts
-as meaningful as when using a word tokenization, leading to a loss of performance. So to get the best of both worlds,
-all transformers models use a hybrid between word-level and character-level tokenization called subword tokenization.
+So if simple space and punctuation tokenization is unsatisfactory, why not simply tokenize on characters? While
+character tokenization is very simple and would greatly reduce memory and time complexity it makes it much harder for
+the model to learn meaningful input representations. *E.g.* learning a meaningful context-independent representation
+for the letter ``"t"`` is much harder as learning a context-independent representation for the word ``"today"``.
+Therefore, character tokenization is often accompanied by a loss of performance. So to get the best of both worlds,
+transformers models use a hybrid between word-level and character-level tokenization called **subword** tokenization.
 
 Subword tokenization
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
-Subword tokenization algorithms rely on the principle that most common words should be left as is, but rare words
-should be decomposed in meaningful subword units. For instance "annoyingly" might be considered a rare word and
-decomposed as "annoying" and "ly". This is especially useful in agglutinative languages such as Turkish, where you can
-form (almost) arbitrarily long complex words by stringing together some subwords.
+Subword tokenization algorithms rely on the principle that frequently used words should not be split into smaller
+subwords, but rare words should be decomposed into meaningful subwords. For instance ``"annoyingly"`` might be
+considered a rare word and could be decomposed into ``"annoying"`` and ``"ly"``. Both ``"annoying"`` and ``"ly"`` as
+stand-alone subwords would appear more frequently while at the same time the meaning of ``"annoyingly"`` is kept by the
+composite meaning of ``"annoying"`` and ``"ly"``. This is especially useful in agglutinative languages such as Turkish,
+where you can form (almost) arbitrarily long complex words by stringing together subwords.
 
-This allows the model to keep a reasonable vocabulary while still learning useful representations for common words or
-subwords. This also enables the model to process words it has never seen before, by decomposing them into subwords it
-knows. For instance, the base :class:`~transformers.BertTokenizer` will tokenize "I have a new GPU!" like this:
+Subword tokenization allows the model to have a reasonable vocabulary size while being able to learn meaningful
+context-independent representations. In addition, subword tokenization enables the model to process words it has never
+seen before, by decomposing them into known subwords. For instance, the :class:`~transformers.BertTokenizer` tokenizes
+``"I have a new GPU!"`` as follows:
 
 .. code-block::
 
     >>> from transformers import BertTokenizer
-    >>> tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
+    >>> tokenizer = BertTokenizer.from_pretrained("bert-base-uncased")
     >>> tokenizer.tokenize("I have a new GPU!")
-    ['i', 'have', 'a', 'new', 'gp', '##u', '!']
+    ["i", "have", "a", "new", "gp", "##u", "!"]
 
-Since we are considering the uncased model, the sentence was lowercased first. Then all the words were present in the
-vocabulary of the tokenizer, except for "gpu", so the tokenizer splits it in subwords it knows: "gp" and "##u". The
-"##" means that the rest of the token should be attached to the previous one, without space (for when we need to decode
-predictions and reverse the tokenization).
+Because we are considering the uncased model, the sentence was lowercased first. We can see that the words ``["i",
+"have", "a", "new"]`` are present in the tokenizer's vocabulary, but the word ``"gpu"`` is not. Consequently, the
+tokenizer splits ``"gpu"`` into known subwords: ``["gp" and "##u"]``. ``"##"`` means that the rest of the token should
+be attached to the previous one, without space (for decoding or reversal of the tokenization).
 
-Another example is when we use the base :class:`~transformers.XLNetTokenizer` to tokenize our previous text:
+As another example, :class:`~transformers.XLNetTokenizer` tokenizes our previously exemplary text as follows:
 
 .. code-block::
 
     >>> from transformers import XLNetTokenizer
-    >>> tokenizer = XLNetTokenizer.from_pretrained('xlnet-base-cased')
+    >>> tokenizer = XLNetTokenizer.from_pretrained("xlnet-base-cased")
     >>> tokenizer.tokenize("Don't you love 🤗 Transformers? We sure do.")
-    ['▁Don', "'", 't', '▁you', '▁love', '▁', '🤗', '▁', 'Transform', 'ers', '?', '▁We', '▁sure', '▁do', '.']
+    ["▁Don", "'", "t", "▁you", "▁love", "▁", "🤗", "▁", "Transform", "ers", "?", "▁We", "▁sure", "▁do", "."]
 
-We'll get back to the meaning of those '▁' when we look at :ref:`SentencePiece <sentencepiece>` but you can see
-Transformers has been split into "Transform" and "ers".
+We'll get back to the meaning of those ``"▁"`` when we look at :ref:`SentencePiece <sentencepiece>`. As one can see,
+the rare word ``"Transformers"`` has been split into the more frequent subwords ``"Transform"`` and ``"ers"``.
 
-Let's now look at how the different subword tokenization algorithms work. Note that they all rely on some form of
-training which is usually done on the corpus the corresponding model will be trained on.
+Let's now look at how the different subword tokenization algorithms work. Note that all of those tokenization
+algorithms rely on some form of training which is usually done on the corpus the corresponding model will be trained
+on.
 
 .. _byte-pair-encoding:
 
-Byte-Pair Encoding
+Byte-Pair Encoding (BPE)
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-Byte-Pair Encoding was introduced in `this paper <https://arxiv.org/abs/1508.07909>`__. It relies on a pretokenizer
-splitting the training data into words, which can be a simple space tokenization (:doc:`GPT-2 <model_doc/gpt2>` and
-:doc:`Roberta <model_doc/roberta>` uses this for instance) or a rule-based tokenizer (:doc:`XLM <model_doc/xlm>` use
-Moses for most languages, as does :doc:`FlauBERT <model_doc/flaubert>`),
+Byte-Pair Encoding (BPE) was introduced in `Neural Machine Translation of Rare Words with Subword Units (Sennrich et
+al., 2015) <https://arxiv.org/abs/1508.07909>`__. BPE relies on a pre-tokenizer that splits the training data into
+words. Pretokenization can be as simple as space tokenization, e.g. :doc:`GPT-2 <model_doc/gpt2>`, :doc:`Roberta
+<model_doc/roberta>`. More advanced pre-tokenization include rule-based tokenization, e.g. :doc:`XLM <model_doc/xlm>`,
+:doc:`FlauBERT <model_doc/flaubert>` which uses Moses for most languages, or :doc:`GPT <model_doc/gpt>` which uses
+Spacy and ftfy, to count the frequency of each word in the training corpus.
 
-:doc:`GPT <model_doc/gpt>` uses Spacy and ftfy, and counts the frequency of each word in the training corpus.
+After pre-tokenization, a set of unique words has been created and the frequency of each word it occurred in the
+training data has been determined. Next, BPE creates a base vocabulary consisting of all symbols that occur in the set
+of unique words and learns merge rules to form a new symbol from two symbols of the base vocabulary. It does so until
+the vocabulary has attained the desired vocabulary size. Note that the desired vocabulary size is a hyperparameter to
+define before training the tokenizer.
 
-It then begins from the list of all characters and will learn merge rules to form a new token from two symbols in the
-vocabulary until it has learned a vocabulary of the desired size (this is a hyperparameter to pick).
-
-Let's say that after the pre-tokenization we have the following words (the number indicating the frequency of each
-word):
+As an example, let's assume that after pre-tokenization, the following set of words including their frequency has been
+determined:
 
 .. code-block::
 
-    ('hug', 10), ('pug', 5), ('pun', 12), ('bun', 4), ('hugs', 5)
+    ("hug", 10), ("pug", 5), ("pun", 12), ("bun", 4), ("hugs", 5)
 
-Then the base vocabulary is ['b', 'g', 'h', 'n', 'p', 's', 'u'] and all our words are first split by character:
+Consequently, the base vocabulary is ``["b", "g", "h", "n", "p", "s", "u"]``. Splitting all words into symbols of the
+base vocabulary, we obtain:
 
 .. code-block::
 
-    ('h' 'u' 'g', 10), ('p' 'u' 'g', 5), ('p' 'u' 'n', 12), ('b' 'u' 'n', 4), ('h' 'u' 'g' 's', 5)
+    ("h" "u" "g", 10), ("p" "u" "g", 5), ("p" "u" "n", 12), ("b" "u" "n", 4), ("h" "u" "g" "s", 5)
 
-We then take each pair of symbols and look at the most frequent. For instance 'hu' is present `10 + 5 = 15` times (10
-times in the 10 occurrences of 'hug', 5 times in the 5 occurrences of 'hugs'). The most frequent here is 'ug', present
-`10 + 5 + 5 = 20` times in total. So the first merge rule the tokenizer learns is to group all 'u' and 'g' together
-then it adds 'ug' to the vocabulary. Our corpus then becomes
+BPE then counts the frequency of each possible symbol pair and picks the symbol pair that occurs most frequently. In
+the example above ``"h"`` followed by ``"u"`` is present `10 + 5 = 15` times (10 times in the 10 occurrences of
+``"hug"``, 5 times in the 5 occurrences of "hugs"). However, the most frequent symbol pair is ``"u"`` followed by "g",
+occurring `10 + 5 + 5 = 20` times in total. Thus, the first merge rule the tokenizer learns is to group all ``"u"``
+symbols followed by a ``"g"`` symbol together. Next, "ug" is added to the vocabulary. The set of words then becomes
 
 .. code-block::
 
-    ('h' 'ug', 10), ('p' 'ug', 5), ('p' 'u' 'n', 12), ('b' 'u' 'n', 4), ('h' 'ug' 's', 5)
+    ("h" "ug", 10), ("p" "ug", 5), ("p" "u" "n", 12), ("b" "u" "n", 4), ("h" "ug" "s", 5)
 
-and we continue by looking at the next most common pair of symbols. It's 'un', present 16 times, so we merge those two
-and add 'un' to the vocabulary. Then it's 'hug' (as 'h' + 'ug'), present 15 times, so we merge those two and add 'hug'
-to the vocabulary.
+BPE then identifies the next most common symbol pair. It's ``"u"`` followed by ``"n"``, which occurs 16 times. ``"u"``,
+``"n"`` is merged to ``"un"`` and added to the vocabulary. The next most frequent symbol pair is ``"h"`` followed by
+``"ug"``, occurring 15 times. Again the pair is merged and ``"hug"`` can be added to the vocabulary.
 
-At this stage, the vocabulary is ``['b', 'g', 'h', 'n', 'p', 's', 'u', 'ug', 'un', 'hug']`` and our corpus is
-represented as
+At this stage, the vocabulary is ``["b", "g", "h", "n", "p", "s", "u", "ug", "un", "hug"]`` and our set of unique words
+is represented as
 
 .. code-block::
 
-    ('hug', 10), ('p' 'ug', 5), ('p' 'un', 12), ('b' 'un', 4), ('hug' 's', 5)
+    ("hug", 10), ("p" "ug", 5), ("p" "un", 12), ("b" "un", 4), ("hug" "s", 5)
 
-If we stop there, the tokenizer can apply the rules it learned to new words (as long as they don't contain characters
-that were not in the base vocabulary). For instance 'bug' would be tokenized as ``['b', 'ug']`` but mug would be
-tokenized as ``['<unk>', 'ug']`` since the 'm' is not in the base vocabulary. This doesn't happen to letters in general
-(since the base corpus uses all of them), but to special characters like emojis.
+Assuming, that the Byte-Pair Encoding training would stop at this point, the learned merge rules would then be applied
+to new words (as long as those new words do not include symbols that were not in the base vocabulary). For instance,
+the word ``"bug"`` would be tokenized to ``["b", "ug"]`` but ``"mug"`` would be tokenized as ``["<unk>", "ug"]`` since
+the symbol ``"m"`` is not in the base vocabulary. In general, single letters such as ``"m"`` are not replaced by the
+``"<unk>"`` symbol because the training data usually includes at least one occurrence of each letter, but it is likely
+to happen for very special characters like emojis.
 
-As we said before, the vocabulary size (which is the base vocabulary size + the number of merges) is a hyperparameter
+As mentioned earlier, the vocabulary size, *i.e.* the base vocabulary size + the number of merges, is a hyperparameter
 to choose. For instance :doc:`GPT <model_doc/gpt>` has a vocabulary size of 40,478 since they have 478 base characters
-and chose to stop the training of the tokenizer at 40,000 merges.
+and chose to stop training after 40,000 merges.
 
 Byte-level BPE
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
-To deal with the fact the base vocabulary needs to get all base characters, which can be quite big if one allows for
-all unicode characters, the `GPT-2 paper
-<https://cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf>`__ introduces a
-clever trick, which is to use bytes as the base vocabulary (which gives a size of 256). With some additional rules to
-deal with punctuation, this manages to be able to tokenize every text without needing an unknown token. For instance,
-the :doc:`GPT-2 model <model_doc/gpt>` has a vocabulary size of 50,257, which corresponds to the 256 bytes base tokens,
-a special end-of-text token and the symbols learned with 50,000 merges.
+A base vocabulary that includes all possible base characters can be quite large if *e.g.* all unicode characters are
+considered as base characters. To have a better base vocabulary, `GPT-2
+<https://cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf>`__ uses bytes
+as the base vocabulary, which is a clever trick to force the base vocabulary to be of size 256 while ensuring that
+every base character is included in the vocabulary. With some additional rules to deal with punctuation, the GPT2's
+tokenizer can tokenize every text without the need for the <unk> symbol. :doc:`GPT-2 <model_doc/gpt>` has a vocabulary
+size of 50,257, which corresponds to the 256 bytes base tokens, a special end-of-text token and the symbols learned
+with 50,000 merges.
 
 .. _wordpiece:
 
 WordPiece
 =======================================================================================================================
 
-WordPiece is the subword tokenization algorithm used for :doc:`BERT <model_doc/bert>` (as well as :doc:`DistilBERT
-<model_doc/distilbert>` and :doc:`Electra <model_doc/electra>`) and was outlined in `this paper
-<https://static.googleusercontent.com/media/research.google.com/ja//pubs/archive/37842.pdf>`__. It relies on the same
-base as BPE, which is to initialize the vocabulary to every character present in the corpus and progressively learn a
-given number of merge rules, the difference is that it doesn't choose the pair that is the most frequent but the one
-that will maximize the likelihood on the corpus once merged.
+WordPiece is the subword tokenization algorithm used for :doc:`BERT <model_doc/bert>`, :doc:`DistilBERT
+<model_doc/distilbert>`, and :doc:`Electra <model_doc/electra>`. The algorithm was outlined in `Japanese and Korean
+Voice Seach (Schuster et al., 2012)
+<https://static.googleusercontent.com/media/research.google.com/ja//pubs/archive/37842.pdf>`__ and is very similar to
+BPE. WordPiece first initializes the vocabulary to include every character present in the training data and
+progressively learn a given number of merge rules. In contrast to BPE, WordPiece does not choose the most frequent
+symbol pair, but the one that maximizes the likelihood of the training data once added to the vocabulary.
 
-What does this mean? Well, in the previous example, it means we would only merge 'u' and 'g' if the probability of
-having 'ug' divided by the probability of having 'u' then 'g' is greater than for any other pair of symbols. It's
-subtly different from what BPE does in the sense that it evaluates what it "loses" by merging two symbols and makes
-sure it's `worth it`.
+So what does this mean exactly? Referring to the previous example, maximizing the likelihood of the training data is
+equivalent to finding the symbol pair, whose probability divided by the probabilities of its first symbol followed by
+its second symbol is the greatest among all symbol pairs. *E.g.* ``"u"``, followed by ``"g"`` would have only been
+merged if the probability of ``"ug"`` divided by ``"u"``, ``"g"`` would have been greater than for any other symbol
+pair. Intuitively, WordPiece is slightly different to BPE in that it evaluates what it `loses` by merging two symbols
+to make ensure it's `worth it`.
 
 .. _unigram:
 
 Unigram
 =======================================================================================================================
 
-Unigram is a subword tokenization algorithm introduced in `this paper <https://arxiv.org/pdf/1804.10959.pdf>`__.
-Instead of starting with a group of base symbols and learning merges with some rule, like BPE or WordPiece, it starts
-from a large vocabulary (for instance, all pretokenized words and the most common substrings) that it will trim down
-progressively. It's not used directly for any of the pretrained models in the library, but it's used in conjunction
-with :ref:`SentencePiece <sentencepiece>`.
+Unigram is a subword tokenization algorithm introduced in `Subword Regularization: Improving Neural Network Translation
+Models with Multiple Subword Candidates (Kudo, 2018) <https://arxiv.org/pdf/1804.10959.pdf>`__. In contrast to BPE or
+WordPiece, Unigram initializes its base vocabulary to a large number of symbols and progressively trims down each
+symbol to obtain a smaller vocabulary. The base vocabulary could for instance correspond to all pre-tokenized words and
+the most common substrings. Unigram is not used directly for any of the models in the transformers, but it's used in
+conjunction with :ref:`SentencePiece <sentencepiece>`.
 
-More specifically, at a given step, unigram computes a loss from the corpus we have and the current vocabulary, then,
-for each subword, evaluate how much the loss would increase if the subword was removed from the vocabulary. It then
-sorts the subwords by this quantity (that represents how much worse the loss becomes if the token is removed) and
-removes all the worst p tokens (for instance p could be 10% or 20%). It then repeats the process until the vocabulary
-has reached the desired size, always keeping the base characters (to be able to tokenize any word written with them,
-like BPE or WordPiece).
+At each training step, the Unigram algorithm defines a loss (often defined as the log-likelihood) over the training
+data given the current vocabulary and a unigram language model. Then, for each symbol in the vocabulary, the algorithm
+computes how much the overall loss would increase if the symbol was to be removed from the vocabulary. Unigram then
+removes p (with p usually being 10% or 20%) percent of the symbols whose loss increase is the lowest, *i.e.* those
+symbols that least affect the overall loss over the training data. This process is repeated until the vocabulary has
+reached the desired size. The Unigram algorithm always keeps the base characters so that any word can be tokenized.
 
-Contrary to BPE and WordPiece that work out rules in a certain order that you can then apply in the same order when
-tokenizing new text, Unigram will have several ways of tokenizing a new text. For instance, if it ends up with the
-vocabulary
+Because Unigram is not based on merge rules (in contrast to BPE and WordPiece), the algorithm has several ways of
+tokenizing new text after training. As an example, if a trained Unigram tokenizer exhibits the vocabulary:
 
 .. code-block::
 
-    ['b', 'g', 'h', 'n', 'p', 's', 'u', 'ug', 'un', 'hug']
+    ["b", "g", "h", "n", "p", "s", "u", "ug", "un", "hug"],
 
-we had before, it could tokenize "hugs" as ``['hug', 's']``, ``['h', 'ug', 's']`` or ``['h', 'u', 'g', 's']``. So which
-one choose? On top of saving the vocabulary, the trained tokenizer will save the probability of each token in the
-training corpus. You can then give a probability to each tokenization (which is the product of the probabilities of the
-tokens forming it) and pick the most likely one (or if you want to apply some data augmentation, you could sample one
-of the tokenization according to their probabilities).
+``"hugs"`` could be tokenized both as ``["hug", "s"]``, ``["h", "ug", "s"]`` or ``["h", "u", "g", "s"]``. So which one
+to choose? Unigram saves the probability of each token in the training corpus on top of saving the vocabulary so that
+the probability of each possible tokenization can be computed after training. The algorithm simply picks the most
+likely tokenization in practice, but also offers the possibility to sample a possible tokenization according to their
+probabilities.
 
-Those probabilities define the loss that trains the tokenizer: if our corpus consists of the words :math:`x_{1}, \dots,
-x_{N}` and if for the word :math:`x_{i}` we note :math:`S(x_{i})` the set of all possible tokenizations of
-:math:`x_{i}` (with the current vocabulary), then the loss is defined as
+Those probabilities are defined by the loss the tokenizer is trained on. Assuming that the training data consists of
+the words :math:`x_{1}, \dots, x_{N}` and that the set of all possible tokenizations for a word :math:`x_{i}` is
+defined as :math:`S(x_{i})`, then the overall loss is defined as
 
 .. math::
     \mathcal{L} = -\sum_{i=1}^{N} \log \left ( \sum_{x \in S(x_{i})} p(x) \right )
@@ -227,15 +247,18 @@ x_{N}` and if for the word :math:`x_{i}` we note :math:`S(x_{i})` the set of all
 SentencePiece
 =======================================================================================================================
 
-All the methods we have been looking at so far required some form of pretokenization, which has a central problem: not
-all languages use spaces to separate words. This is a problem :doc:`XLM <model_doc/xlm>` solves by using specific
-pretokenizers for each of those languages (in this case, Chinese, Japanese and Thai). To solve this problem,
-SentencePiece (introduced in `this paper <https://arxiv.org/pdf/1808.06226.pdf>`__) treats the input as a raw stream,
-includes the space in the set of characters to use, then uses BPE or unigram to construct the appropriate vocabulary.
+All tokenization algorithms described so far have the same problem: It is assumed that the input text uses spaces to
+separate words. However, not all languages use spaces to separate words. One possible solution is to use language
+specific pre-tokenizers, *e.g.* :doc:`XLM <model_doc/xlm>` uses a specific Chinese, Japanese, and Thai pre-tokenizer).
+To solve this problem more generally, `SentencePiece: A simple and language independent subword tokenizer and
+detokenizer for Neural Text Processing (Kudo et al., 2018) <https://arxiv.org/pdf/1808.06226.pdf>`__ treats the input
+as a raw input stream, thus including the space in the set of characters to use. It then uses the BPE or unigram
+algorithm to construct the appropriate vocabulary.
 
-That's why in the example we saw before using :class:`~transformers.XLNetTokenizer` (which uses SentencePiece), we had
-the '▁' character, that represents space. Decoding a tokenized text is then super easy: we just have to concatenate all
-of them together and replace '▁' with space.
+The :class:`~transformers.XLNetTokenizer` uses SentencePiece for example, which is also why in the example earlier the
+``"▁"`` character was included in the vocabulary. Decoding with SentencePiece is very easy since all tokens can just be
+concatenated and ``"▁"`` is replaced by a space.
 
-All transformers models in the library that use SentencePiece use it with unigram. Examples of models using it are
-:doc:`ALBERT <model_doc/albert>`, :doc:`XLNet <model_doc/xlnet>` or the :doc:`Marian framework <model_doc/marian>`.
+All transformers models in the library that use SentencePiece use it in combination with unigram. Examples of models
+using SentencePiece are :doc:`ALBERT <model_doc/albert>`, :doc:`XLNet <model_doc/xlnet>`, :doc:`Marian
+<model_doc/marian>`, and :doc:`T5 <model_doc/t5>`.

docs/source/training.rst

@@ -39,7 +39,7 @@ head on top of the encoder with an output size of 2. Models are initialized in `
 .. code-block:: python
 
     from transformers import BertForSequenceClassification
-    model = BertForSequenceClassification.from_pretrained('bert-base-uncased', return_dict=True)
+    model = BertForSequenceClassification.from_pretrained('bert-base-uncased')
     model.train()
 
 This is useful because it allows us to make use of the pre-trained BERT encoder and easily train it on whatever

examples/adversarial/run_hans.py

@@ -23,6 +23,7 @@ from typing import Dict, List, Optional
 import numpy as np
 import torch
 
+import transformers
 from transformers import (
     AutoConfig,
     AutoModelForSequenceClassification,
@@ -33,6 +34,7 @@ from transformers import (
     default_data_collator,
     set_seed,
 )
+from transformers.trainer_utils import is_main_process
 from utils_hans import HansDataset, InputFeatures, hans_processors, hans_tasks_num_labels
 
 
@@ -55,7 +57,8 @@ class ModelArguments:
         default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
     )
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
     )
 
 
@@ -124,6 +127,11 @@ def main():
         bool(training_args.local_rank != -1),
         training_args.fp16,
     )
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(training_args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     logger.info("Training/evaluation parameters %s", training_args)
 
     # Set seed

examples/benchmarking/plot_csv_file.py

@@ -25,7 +25,7 @@ class PlotArguments:
     )
     plot_along_batch: bool = field(
         default=False,
-        metadata={"help": "Whether to plot along batch size or sequence lengh. Defaults to sequence length."},
+        metadata={"help": "Whether to plot along batch size or sequence length. Defaults to sequence length."},
     )
     is_time: bool = field(
         default=False,

examples/bert-loses-patience/pabee/modeling_pabee_albert.py

@@ -21,7 +21,7 @@ import torch.nn as nn
 from torch.nn import CrossEntropyLoss, MSELoss
 
 from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
-from transformers.modeling_albert import (
+from transformers.models.albert.modeling_albert import (
     ALBERT_INPUTS_DOCSTRING,
     ALBERT_START_DOCSTRING,
     AlbertModel,

examples/bert-loses-patience/pabee/modeling_pabee_bert.py

@@ -23,7 +23,7 @@ from torch import nn
 from torch.nn import CrossEntropyLoss, MSELoss
 
 from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
-from transformers.modeling_bert import (
+from transformers.models.bert.modeling_bert import (
     BERT_INPUTS_DOCSTRING,
     BERT_START_DOCSTRING,
     BertEncoder,

examples/bert-loses-patience/run_glue_with_pabee.py

@@ -29,6 +29,7 @@ from torch.utils.data import DataLoader, RandomSampler, SequentialSampler, Tenso
 from torch.utils.data.distributed import DistributedSampler
 from tqdm import tqdm, trange
 
+import transformers
 from pabee.modeling_pabee_albert import AlbertForSequenceClassificationWithPabee
 from pabee.modeling_pabee_bert import BertForSequenceClassificationWithPabee
 from transformers import (
@@ -44,6 +45,7 @@ from transformers import glue_compute_metrics as compute_metrics
 from transformers import glue_convert_examples_to_features as convert_examples_to_features
 from transformers import glue_output_modes as output_modes
 from transformers import glue_processors as processors
+from transformers.trainer_utils import is_main_process
 
 
 try:
@@ -474,7 +476,7 @@ def main():
         "--cache_dir",
         default="",
         type=str,
-        help="Where do you want to store the pre-trained models downloaded from s3",
+        help="Where do you want to store the pre-trained models downloaded from huggingface.co",
     )
     parser.add_argument(
         "--max_seq_length",
@@ -630,7 +632,11 @@ def main():
         bool(args.local_rank != -1),
         args.fp16,
     )
-
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     # Set seed
     set_seed(args)
 

examples/bert-loses-patience/test_run_glue_with_pabee.py

@@ -4,7 +4,7 @@ import sys
 from unittest.mock import patch
 
 import run_glue_with_pabee
-from transformers.testing_utils import TestCasePlus, require_torch_non_multigpu_but_fix_me
+from transformers.testing_utils import TestCasePlus, require_torch_non_multi_gpu_but_fix_me
 
 
 logging.basicConfig(level=logging.DEBUG)
@@ -20,7 +20,7 @@ def get_setup_file():
 
 
 class PabeeTests(TestCasePlus):
-    @require_torch_non_multigpu_but_fix_me
+    @require_torch_non_multi_gpu_but_fix_me
     def test_run_glue(self):
         stream_handler = logging.StreamHandler(sys.stdout)
         logger.addHandler(stream_handler)

examples/bertology/run_bertology.py

@@ -30,6 +30,7 @@ from torch.utils.data import DataLoader, SequentialSampler, Subset
 from torch.utils.data.distributed import DistributedSampler
 from tqdm import tqdm
 
+import transformers
 from transformers import (
     AutoConfig,
     AutoModelForSequenceClassification,
@@ -41,6 +42,7 @@ from transformers import (
     glue_processors,
     set_seed,
 )
+from transformers.trainer_utils import is_main_process
 
 
 logger = logging.getLogger(__name__)
@@ -296,7 +298,7 @@ def main():
         "--cache_dir",
         default=None,
         type=str,
-        help="Where do you want to store the pre-trained models downloaded from s3",
+        help="Where do you want to store the pre-trained models downloaded from huggingface.co",
     )
     parser.add_argument(
         "--data_subset", type=int, default=-1, help="If > 0: limit the data to a subset of data_subset instances."
@@ -368,6 +370,11 @@ def main():
     # Setup logging
     logging.basicConfig(level=logging.INFO if args.local_rank in [-1, 0] else logging.WARN)
     logger.info("device: {} n_gpu: {}, distributed: {}".format(args.device, args.n_gpu, bool(args.local_rank != -1)))
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
 
     # Set seeds
     set_seed(args.seed)

examples/contrib/legacy/run_language_modeling.py

@@ -29,6 +29,7 @@ from typing import Optional
 
 from torch.utils.data import ConcatDataset
 
+import transformers
 from transformers import (
     CONFIG_MAPPING,
     MODEL_WITH_LM_HEAD_MAPPING,
@@ -47,6 +48,7 @@ from transformers import (
     TrainingArguments,
     set_seed,
 )
+from transformers.trainer_utils import is_main_process
 
 
 logger = logging.getLogger(__name__)
@@ -79,7 +81,8 @@ class ModelArguments:
         default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
     )
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
     )
 
 
@@ -219,6 +222,11 @@ def main():
         bool(training_args.local_rank != -1),
         training_args.fp16,
     )
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(training_args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     logger.info("Training/evaluation parameters %s", training_args)
 
     # Set seed

examples/contrib/mm-imdb/run_mmimdb.py

@@ -31,6 +31,7 @@ from torch.utils.data import DataLoader, RandomSampler, SequentialSampler
 from torch.utils.data.distributed import DistributedSampler
 from tqdm import tqdm, trange
 
+import transformers
 from transformers import (
     WEIGHTS_NAME,
     AdamW,
@@ -41,6 +42,7 @@ from transformers import (
     MMBTForClassification,
     get_linear_schedule_with_warmup,
 )
+from transformers.trainer_utils import is_main_process
 from utils_mmimdb import ImageEncoder, JsonlDataset, collate_fn, get_image_transforms, get_mmimdb_labels
 
 
@@ -348,7 +350,7 @@ def main():
         "--cache_dir",
         default=None,
         type=str,
-        help="Where do you want to store the pre-trained models downloaded from s3",
+        help="Where do you want to store the pre-trained models downloaded from huggingface.co",
     )
     parser.add_argument(
         "--max_seq_length",
@@ -476,7 +478,11 @@ def main():
         bool(args.local_rank != -1),
         args.fp16,
     )
-
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     # Set seed
     set_seed(args)
 

examples/contrib/run_camembert.py

@@ -1,7 +1,6 @@
 import torch
 
-from transformers.modeling_camembert import CamembertForMaskedLM
-from transformers.tokenization_camembert import CamembertTokenizer
+from transformers import CamembertForMaskedLM, CamembertTokenizer
 
 
 def fill_mask(masked_input, model, tokenizer, topk=5):

examples/contrib/run_chinese_ref.py

@@ -3,7 +3,7 @@ import json
 from typing import List
 
 from ltp import LTP
-from transformers.tokenization_bert import BertTokenizer
+from transformers import BertTokenizer
 
 
 def _is_chinese_char(cp):

examples/contrib/run_swag.py

@@ -31,8 +31,16 @@ from torch.utils.data import DataLoader, RandomSampler, SequentialSampler, Tenso
 from torch.utils.data.distributed import DistributedSampler
 from tqdm import tqdm, trange
 
-from transformers import WEIGHTS_NAME, AdamW, AutoConfig, AutoTokenizer, get_linear_schedule_with_warmup
-from transformers.modeling_auto import AutoModelForMultipleChoice
+import transformers
+from transformers import (
+    WEIGHTS_NAME,
+    AdamW,
+    AutoConfig,
+    AutoModelForMultipleChoice,
+    AutoTokenizer,
+    get_linear_schedule_with_warmup,
+)
+from transformers.trainer_utils import is_main_process
 
 
 try:
@@ -620,6 +628,11 @@ def main():
         bool(args.local_rank != -1),
         args.fp16,
     )
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
 
     # Set seed
     set_seed(args)

examples/deebert/run_glue_deebert.py

@@ -13,6 +13,7 @@ from torch.utils.data import DataLoader, RandomSampler, SequentialSampler, Tenso
 from torch.utils.data.distributed import DistributedSampler
 from tqdm import tqdm, trange
 
+import transformers
 from src.modeling_highway_bert import DeeBertForSequenceClassification
 from src.modeling_highway_roberta import DeeRobertaForSequenceClassification
 from transformers import (
@@ -28,6 +29,7 @@ from transformers import glue_compute_metrics as compute_metrics
 from transformers import glue_convert_examples_to_features as convert_examples_to_features
 from transformers import glue_output_modes as output_modes
 from transformers import glue_processors as processors
+from transformers.trainer_utils import is_main_process
 
 
 try:
@@ -450,7 +452,7 @@ def main():
         "--cache_dir",
         default="",
         type=str,
-        help="Where do you want to store the pre-trained models downloaded from s3",
+        help="Where do you want to store the pre-trained models downloaded from huggingface.co",
     )
     parser.add_argument(
         "--max_seq_length",
@@ -580,7 +582,11 @@ def main():
         bool(args.local_rank != -1),
         args.fp16,
     )
-
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     # Set seed
     set_seed(args)
 

examples/deebert/src/modeling_highway_bert.py

@@ -3,7 +3,7 @@ from torch import nn
 from torch.nn import CrossEntropyLoss, MSELoss
 
 from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
-from transformers.modeling_bert import (
+from transformers.models.bert.modeling_bert import (
     BERT_INPUTS_DOCSTRING,
     BERT_START_DOCSTRING,
     BertEmbeddings,

examples/deebert/src/modeling_highway_roberta.py

@@ -3,9 +3,13 @@ from __future__ import absolute_import, division, print_function, unicode_litera
 import torch.nn as nn
 from torch.nn import CrossEntropyLoss, MSELoss
 
-from transformers.configuration_roberta import RobertaConfig
+from transformers import RobertaConfig
 from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
-from transformers.modeling_roberta import ROBERTA_INPUTS_DOCSTRING, ROBERTA_START_DOCSTRING, RobertaEmbeddings
+from transformers.models.roberta.modeling_roberta import (
+    ROBERTA_INPUTS_DOCSTRING,
+    ROBERTA_START_DOCSTRING,
+    RobertaEmbeddings,
+)
 
 from .modeling_highway_bert import BertPreTrainedModel, DeeBertModel, HighwayException, entropy
 

examples/deebert/test_glue_deebert.py

@@ -5,7 +5,7 @@ import unittest
 from unittest.mock import patch
 
 import run_glue_deebert
-from transformers.testing_utils import require_torch_non_multigpu_but_fix_me, slow
+from transformers.testing_utils import require_torch_non_multi_gpu_but_fix_me, slow
 
 
 logging.basicConfig(level=logging.DEBUG)
@@ -26,7 +26,7 @@ class DeeBertTests(unittest.TestCase):
         logger.addHandler(stream_handler)
 
     @slow
-    @require_torch_non_multigpu_but_fix_me
+    @require_torch_non_multi_gpu_but_fix_me
     def test_glue_deebert_train(self):
 
         train_args = """

examples/distillation/README.md

@@ -17,7 +17,7 @@ This folder contains the original code used to train Distil* as well as examples
 
 ## What is Distil*
 
-Distil* is a class of compressed models that started with DistilBERT. DistilBERT stands for Distillated-BERT. DistilBERT is a small, fast, cheap and light Transformer model based on Bert architecture. It has 40% less parameters than `bert-base-uncased`, runs 60% faster while preserving 97% of BERT's performances as measured on the GLUE language understanding benchmark. DistilBERT is trained using knowledge distillation, a technique to compress a large model called the teacher into a smaller model called the student. By distillating Bert, we obtain a smaller Transformer model that bears a lot of similarities with the original BERT model while being lighter, smaller and faster to run. DistilBERT is thus an interesting option to put large-scaled trained Transformer model into production.
+Distil* is a class of compressed models that started with DistilBERT. DistilBERT stands for Distilled-BERT. DistilBERT is a small, fast, cheap and light Transformer model based on Bert architecture. It has 40% less parameters than `bert-base-uncased`, runs 60% faster while preserving 97% of BERT's performances as measured on the GLUE language understanding benchmark. DistilBERT is trained using knowledge distillation, a technique to compress a large model called the teacher into a smaller model called the student. By distillating Bert, we obtain a smaller Transformer model that bears a lot of similarities with the original BERT model while being lighter, smaller and faster to run. DistilBERT is thus an interesting option to put large-scaled trained Transformer model into production.
 
 We have applied the same method to other Transformer architectures and released the weights:
 - GPT2: on the [WikiText-103](https://blog.einstein.ai/the-wikitext-long-term-dependency-language-modeling-dataset/) benchmark, GPT2 reaches a perplexity on the test set of 16.3 compared to 21.1 for **DistilGPT2** (after fine-tuning on the train set).
@@ -57,7 +57,7 @@ Here are the results on the *test* sets for 6 of the languages available in XNLI
 
 This part of the library has only be tested with Python3.6+. There are few specific dependencies to install before launching a distillation, you can install them with the command `pip install -r requirements.txt`.
 
-**Important note:** The training scripts have been updated to support PyTorch v1.2.0 (there are breakings changes compared to v1.1.0).
+**Important note:** The training scripts have been updated to support PyTorch v1.2.0 (there are breaking changes compared to v1.1.0).
 
 
 ## How to use DistilBERT
@@ -111,7 +111,7 @@ python scripts/binarized_data.py \
     --dump_file data/binarized_text
 ```
 
-Our implementation of masked language modeling loss follows [XLM](https://github.com/facebookresearch/XLM)'s one and smoothes the probability of masking with a factor that put more emphasis on rare words. Thus we count the occurrences of each tokens in the data:
+Our implementation of masked language modeling loss follows [XLM](https://github.com/facebookresearch/XLM)'s one and smooths the probability of masking with a factor that put more emphasis on rare words. Thus we count the occurrences of each tokens in the data:
 
 ```bash
 python scripts/token_counts.py \
@@ -173,7 +173,7 @@ python -m torch.distributed.launch \
         --token_counts data/token_counts.bert-base-uncased.pickle
 ```
 
-**Tips:** Starting distillated training with good initialization of the model weights is crucial to reach decent performance. In our experiments, we initialized our model from a few layers of the teacher (Bert) itself! Please refer to `scripts/extract.py` and `scripts/extract_distilbert.py` to create a valid initialization checkpoint and use `--student_pretrained_weights` argument to use this initialization for the distilled training!
+**Tips:** Starting distilled training with good initialization of the model weights is crucial to reach decent performance. In our experiments, we initialized our model from a few layers of the teacher (Bert) itself! Please refer to `scripts/extract.py` and `scripts/extract_distilbert.py` to create a valid initialization checkpoint and use `--student_pretrained_weights` argument to use this initialization for the distilled training!
 
 Happy distillation!
 

examples/distillation/distiller.py

@@ -188,7 +188,7 @@ class Distiller:
 
     def prepare_batch_mlm(self, batch):
         """
-        Prepare the batch: from the token_ids and the lenghts, compute the attention mask and the masked label for MLM.
+        Prepare the batch: from the token_ids and the lengths, compute the attention mask and the masked label for MLM.
 
         Input:
         ------
@@ -200,7 +200,7 @@ class Distiller:
         -------
             token_ids: `torch.tensor(bs, seq_length)` - The token ids after the modifications for MLM.
             attn_mask: `torch.tensor(bs, seq_length)` - The attention mask for the self-attention.
-            mlm_labels: `torch.tensor(bs, seq_length)` - The masked languge modeling labels. There is a -100 where there is nothing to predict.
+            mlm_labels: `torch.tensor(bs, seq_length)` - The masked language modeling labels. There is a -100 where there is nothing to predict.
         """
         token_ids, lengths = batch
         token_ids, lengths = self.round_batch(x=token_ids, lengths=lengths)
@@ -253,7 +253,7 @@ class Distiller:
 
     def prepare_batch_clm(self, batch):
         """
-        Prepare the batch: from the token_ids and the lenghts, compute the attention mask and the labels for CLM.
+        Prepare the batch: from the token_ids and the lengths, compute the attention mask and the labels for CLM.
 
         Input:
         ------

examples/distillation/run_squad_w_distillation.py

@@ -30,6 +30,7 @@ from torch.utils.data import DataLoader, RandomSampler, SequentialSampler
 from torch.utils.data.distributed import DistributedSampler
 from tqdm import tqdm, trange
 
+import transformers
 from transformers import (
     WEIGHTS_NAME,
     AdamW,
@@ -57,6 +58,7 @@ from transformers.data.metrics.squad_metrics import (
     squad_evaluate,
 )
 from transformers.data.processors.squad import SquadResult, SquadV1Processor, SquadV2Processor
+from transformers.trainer_utils import is_main_process
 
 
 try:
@@ -576,7 +578,7 @@ def main():
         "--cache_dir",
         default="",
         type=str,
-        help="Where do you want to store the pre-trained models downloaded from s3",
+        help="Where do you want to store the pre-trained models downloaded from huggingface.co",
     )
 
     parser.add_argument(
@@ -745,7 +747,11 @@ def main():
         bool(args.local_rank != -1),
         args.fp16,
     )
-
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     # Set seed
     set_seed(args)
 

examples/distillation/scripts/extract_distilbert.py

@@ -86,7 +86,7 @@ if __name__ == "__main__":
             compressed_sd[f"vocab_layer_norm.{w}"] = state_dict[f"cls.predictions.transform.LayerNorm.{w}"]
 
     print(f"N layers selected for distillation: {std_idx}")
-    print(f"Number of params transfered for distillation: {len(compressed_sd.keys())}")
+    print(f"Number of params transferred for distillation: {len(compressed_sd.keys())}")
 
-    print(f"Save transfered checkpoint to {args.dump_checkpoint}.")
+    print(f"Save transferred checkpoint to {args.dump_checkpoint}.")
     torch.save(compressed_sd, args.dump_checkpoint)

examples/language-modeling/README.md

@@ -90,7 +90,7 @@ selected tokens (which may be part of words), they mask randomly selected words
 to that word). This technique has been refined for Chinese in [this paper](https://arxiv.org/abs/1906.08101).
 
 To fine-tune a model using whole word masking, use the following script:
-
+```bash
 python run_mlm_wwm.py \
     --model_name_or_path roberta-base \
     --dataset_name wikitext \
@@ -164,7 +164,7 @@ context length for permutation language modeling.
 The `--max_span_length` flag may also be used to limit the length of a span of masked tokens used 
 for permutation language modeling.
 
-Here is how to fine-tun XLNet on wikitext-2:
+Here is how to fine-tune XLNet on wikitext-2:
 
 ```bash
 python run_plm.py \

examples/language-modeling/run_clm.py

@@ -76,7 +76,8 @@ class ModelArguments:
         default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
     )
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
     )
     use_fast_tokenizer: bool = field(
         default=True,
@@ -168,6 +169,8 @@ def main():
     # Set the verbosity to info of the Transformers logger (on main process only):
     if is_main_process(training_args.local_rank):
         transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     logger.info("Training/evaluation parameters %s", training_args)
 
     # Set seed before initializing model.
@@ -254,7 +257,7 @@ def main():
         tokenize_function,
         batched=True,
         num_proc=data_args.preprocessing_num_workers,
-        remove_columns=[text_column_name],
+        remove_columns=column_names,
         load_from_cache_file=not data_args.overwrite_cache,
     )
 
@@ -310,9 +313,12 @@ def main():
 
     # Training
     if training_args.do_train:
-        trainer.train(
-            model_path=model_args.model_name_or_path if os.path.isdir(model_args.model_name_or_path) else None
+        model_path = (
+            model_args.model_name_or_path
+            if (model_args.model_name_or_path is not None and os.path.isdir(model_args.model_name_or_path))
+            else None
         )
+        trainer.train(model_path=model_path)
         trainer.save_model()  # Saves the tokenizer too for easy upload
 
     # Evaluation

examples/language-modeling/run_mlm.py

@@ -74,7 +74,8 @@ class ModelArguments:
         default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
     )
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
     )
     use_fast_tokenizer: bool = field(
         default=True,
@@ -179,6 +180,8 @@ def main():
     # Set the verbosity to info of the Transformers logger (on main process only):
     if is_main_process(training_args.local_rank):
         transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     logger.info("Training/evaluation parameters %s", training_args)
 
     # Set seed before initializing model.
@@ -292,7 +295,7 @@ def main():
             tokenize_function,
             batched=True,
             num_proc=data_args.preprocessing_num_workers,
-            remove_columns=[text_column_name],
+            remove_columns=column_names,
             load_from_cache_file=not data_args.overwrite_cache,
         )
 
@@ -351,9 +354,12 @@ def main():
 
     # Training
     if training_args.do_train:
-        trainer.train(
-            model_path=model_args.model_name_or_path if os.path.isdir(model_args.model_name_or_path) else None
+        model_path = (
+            model_args.model_name_or_path
+            if (model_args.model_name_or_path is not None and os.path.isdir(model_args.model_name_or_path))
+            else None
         )
+        trainer.train(model_path=model_path)
         trainer.save_model()  # Saves the tokenizer too for easy upload
 
     # Evaluation

examples/language-modeling/run_mlm_wwm.py

@@ -76,7 +76,8 @@ class ModelArguments:
         default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
     )
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
     )
     use_fast_tokenizer: bool = field(
         default=True,
@@ -186,6 +187,8 @@ def main():
     # Set the verbosity to info of the Transformers logger (on main process only):
     if is_main_process(training_args.local_rank):
         transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     logger.info("Training/evaluation parameters %s", training_args)
 
     # Set seed before initializing model.
@@ -299,9 +302,12 @@ def main():
 
     # Training
     if training_args.do_train:
-        trainer.train(
-            model_path=model_args.model_name_or_path if os.path.isdir(model_args.model_name_or_path) else None
+        model_path = (
+            model_args.model_name_or_path
+            if (model_args.model_name_or_path is not None and os.path.isdir(model_args.model_name_or_path))
+            else None
         )
+        trainer.train(model_path=model_path)
         trainer.save_model()  # Saves the tokenizer too for easy upload
 
     # Evaluation

examples/language-modeling/run_plm.py

@@ -64,7 +64,8 @@ class ModelArguments:
         default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
     )
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
     )
     use_fast_tokenizer: bool = field(
         default=True,
@@ -176,6 +177,8 @@ def main():
     # Set the verbosity to info of the Transformers logger (on main process only):
     if is_main_process(training_args.local_rank):
         transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     logger.info("Training/evaluation parameters %s", training_args)
 
     # Set seed before initializing model.
@@ -279,7 +282,7 @@ def main():
             tokenize_function,
             batched=True,
             num_proc=data_args.preprocessing_num_workers,
-            remove_columns=[text_column_name],
+            remove_columns=column_names,
             load_from_cache_file=not data_args.overwrite_cache,
         )
 
@@ -341,9 +344,12 @@ def main():
 
     # Training
     if training_args.do_train:
-        trainer.train(
-            model_path=model_args.model_name_or_path if os.path.isdir(model_args.model_name_or_path) else None
+        model_path = (
+            model_args.model_name_or_path
+            if (model_args.model_name_or_path is not None and os.path.isdir(model_args.model_name_or_path))
+            else None
         )
+        trainer.train(model_path=model_path)
         trainer.save_model()  # Saves the tokenizer too for easy upload
 
     # Evaluation

examples/lightning_base.py

@@ -4,6 +4,7 @@ import os
 from pathlib import Path
 from typing import Any, Dict
 
+import packaging
 import pytorch_lightning as pl
 from pytorch_lightning.utilities import rank_zero_info
 
@@ -33,15 +34,17 @@ from transformers.optimization import (
 
 logger = logging.getLogger(__name__)
 
-try:
-    pkg = "pytorch_lightning"
-    min_ver = "1.0.4"
-    pkg_resources.require(f"{pkg}>={min_ver}")
-except pkg_resources.VersionConflict:
-    logger.warning(
-        f"{pkg}>={min_ver} is required for a normal functioning of this module, but found {pkg}=={pkg_resources.get_distribution(pkg).version}. Try pip install -r examples/requirements.txt"
-    )
 
+def require_min_ver(pkg, min_ver):
+    got_ver = pkg_resources.get_distribution(pkg).version
+    if packaging.version.parse(got_ver) < packaging.version.parse(min_ver):
+        logger.warning(
+            f"{pkg}>={min_ver} is required for a normal functioning of this module, but found {pkg}=={got_ver}. "
+            "Try: pip install -r examples/requirements.txt"
+        )
+
+
+require_min_ver("pytorch_lightning", "1.0.4")
 
 MODEL_MODES = {
     "base": AutoModel,
@@ -233,7 +236,7 @@ class BaseTransformer(pl.LightningModule):
             "--cache_dir",
             default="",
             type=str,
-            help="Where do you want to store the pre-trained models downloaded from s3",
+            help="Where do you want to store the pre-trained models downloaded from huggingface.co",
         )
         parser.add_argument(
             "--encoder_layerdrop",

examples/longform-qa/eli5_app.py

@@ -107,7 +107,12 @@ def make_support(question, source="wiki40b", method="dense", n_results=10):
     return question_doc, support_list
 
 
-@st.cache(hash_funcs={torch.Tensor: (lambda _: None), transformers.tokenization_bart.BartTokenizer: (lambda _: None)})
+@st.cache(
+    hash_funcs={
+        torch.Tensor: (lambda _: None),
+        transformers.models.bart.tokenization_bart.BartTokenizer: (lambda _: None),
+    }
+)
 def answer_question(
     question_doc, s2s_model, s2s_tokenizer, min_len=64, max_len=256, sampling=False, n_beams=2, top_p=0.95, temp=0.8
 ):

examples/lxmert/demo.ipynb

@@ -210,7 +210,6 @@
     "        visual_feats=features,\n",
     "        visual_pos=normalized_boxes,\n",
     "        token_type_ids=inputs.token_type_ids,\n",
-    "        return_dict=True,\n",
     "        output_attentions=False,\n",
     "    )\n",
     "    output_vqa = lxmert_vqa(\n",
@@ -219,7 +218,6 @@
     "        visual_feats=features,\n",
     "        visual_pos=normalized_boxes,\n",
     "        token_type_ids=inputs.token_type_ids,\n",
-    "        return_dict=True,\n",
     "        output_attentions=False,\n",
     "    )\n",
     "    # get prediction\n",
@@ -266,4 +264,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

examples/movement-pruning/README.md

@@ -21,7 +21,7 @@ You can also have a look at this fun *Explain Like I'm Five* introductory [slide
 
 One promise of extreme pruning is to obtain extremely small models that can be easily sent (and stored) on edge devices. By setting weights to 0., we reduce the amount of information we need to store, and thus decreasing the memory size. We are able to obtain extremely sparse fine-pruned models with movement pruning: ~95% of the dense performance with ~5% of total remaining weights in the BERT encoder.
 
-In [this notebook](https://github.com/huggingface/transformers/blob/master/examples/movement-pruning/Saving_PruneBERT.ipynb), we showcase how we can leverage standard tools that exist out-of-the-box to efficiently store an extremely sparse question answering model (only 6% of total remaining weights in the encoder). We are able to reduce the memory size of the encoder **from the 340MB (the orignal dense BERT) to 11MB**, without any additional training of the model (every operation is performed *post fine-pruning*). It is sufficiently small to store it on a [91' floppy disk](https://en.wikipedia.org/wiki/Floptical) 📎!
+In [this notebook](https://github.com/huggingface/transformers/blob/master/examples/movement-pruning/Saving_PruneBERT.ipynb), we showcase how we can leverage standard tools that exist out-of-the-box to efficiently store an extremely sparse question answering model (only 6% of total remaining weights in the encoder). We are able to reduce the memory size of the encoder **from the 340MB (the original dense BERT) to 11MB**, without any additional training of the model (every operation is performed *post fine-pruning*). It is sufficiently small to store it on a [91' floppy disk](https://en.wikipedia.org/wiki/Floptical) 📎!
 
 While movement pruning does not directly optimize for memory footprint (but rather the number of non-null weights), we hypothetize that further memory compression ratios can be achieved with specific quantization aware trainings (see for instance [Q8BERT](https://arxiv.org/abs/1910.06188), [And the Bit Goes Down](https://arxiv.org/abs/1907.05686) or [Quant-Noise](https://arxiv.org/abs/2004.07320)).
 

examples/movement-pruning/emmental/modeling_bert_masked.py

@@ -16,7 +16,7 @@
 """Masked Version of BERT. It replaces the `torch.nn.Linear` layers with
 :class:`~emmental.MaskedLinear` and add an additional parameters in the forward pass to
 compute the adaptive mask.
-Built on top of `transformers.modeling_bert`"""
+Built on top of `transformers.models.bert.modeling_bert`"""
 
 
 import logging
@@ -29,8 +29,8 @@ from torch.nn import CrossEntropyLoss, MSELoss
 from emmental import MaskedBertConfig
 from emmental.modules import MaskedLinear
 from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
-from transformers.modeling_bert import ACT2FN, BertLayerNorm, load_tf_weights_in_bert
 from transformers.modeling_utils import PreTrainedModel, prune_linear_layer
+from transformers.models.bert.modeling_bert import ACT2FN, BertLayerNorm, load_tf_weights_in_bert
 
 
 logger = logging.getLogger(__name__)

examples/movement-pruning/emmental/modules/binarizer.py

@@ -14,7 +14,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """
-Binarizers take a (real value) matrice as input and produce a binary (values in {0,1}) mask of the same shape.
+Binarizers take a (real value) matrix as input and produce a binary (values in {0,1}) mask of the same shape.
 """
 
 import torch

examples/movement-pruning/masked_run_glue.py

@@ -620,7 +620,7 @@ def main():
         "--cache_dir",
         default="",
         type=str,
-        help="Where do you want to store the pre-trained models downloaded from s3",
+        help="Where do you want to store the pre-trained models downloaded from huggingface.co",
     )
     parser.add_argument(
         "--max_seq_length",

examples/movement-pruning/masked_run_squad.py

@@ -725,7 +725,7 @@ def main():
         "--cache_dir",
         default="",
         type=str,
-        help="Where do you want to store the pre-trained models downloaded from s3",
+        help="Where do you want to store the pre-trained models downloaded from huggingface.co",
     )
 
     parser.add_argument(

examples/multiple-choice/run_multiple_choice.py

@@ -23,6 +23,7 @@ from typing import Dict, Optional
 
 import numpy as np
 
+import transformers
 from transformers import (
     AutoConfig,
     AutoModelForMultipleChoice,
@@ -33,6 +34,7 @@ from transformers import (
     TrainingArguments,
     set_seed,
 )
+from transformers.trainer_utils import is_main_process
 from utils_multiple_choice import MultipleChoiceDataset, Split, processors
 
 
@@ -59,7 +61,8 @@ class ModelArguments:
         default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
     )
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
     )
 
 
@@ -115,6 +118,11 @@ def main():
         bool(training_args.local_rank != -1),
         training_args.fp16,
     )
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(training_args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     logger.info("Training/evaluation parameters %s", training_args)
 
     # Set seed

examples/multiple-choice/run_tf_multiple_choice.py

@@ -33,9 +33,15 @@ from transformers import (
     TFTrainingArguments,
     set_seed,
 )
+from transformers.utils import logging as hf_logging
 from utils_multiple_choice import Split, TFMultipleChoiceDataset, processors
 
 
+hf_logging.set_verbosity_info()
+hf_logging.enable_default_handler()
+hf_logging.enable_explicit_format()
+
+
 logger = logging.getLogger(__name__)
 
 
@@ -59,7 +65,8 @@ class ModelArguments:
         default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
     )
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
     )
 
 

examples/question-answering/run_squad.py

@@ -29,6 +29,7 @@ from torch.utils.data import DataLoader, RandomSampler, SequentialSampler
 from torch.utils.data.distributed import DistributedSampler
 from tqdm import tqdm, trange
 
+import transformers
 from transformers import (
     MODEL_FOR_QUESTION_ANSWERING_MAPPING,
     WEIGHTS_NAME,
@@ -45,6 +46,7 @@ from transformers.data.metrics.squad_metrics import (
     squad_evaluate,
 )
 from transformers.data.processors.squad import SquadResult, SquadV1Processor, SquadV2Processor
+from transformers.trainer_utils import is_main_process
 
 
 try:
@@ -319,7 +321,7 @@ def evaluate(args, model, tokenizer, prefix=""):
             eval_feature = features[feature_index.item()]
             unique_id = int(eval_feature.unique_id)
 
-            output = [to_list(output[i]) for output in outputs]
+            output = [to_list(output[i]) for output in outputs.to_tuple()]
 
             # Some models (XLNet, XLM) use 5 arguments for their predictions, while the other "simpler"
             # models only use two.
@@ -530,7 +532,7 @@ def main():
         "--cache_dir",
         default="",
         type=str,
-        help="Where do you want to store the pre-trained models downloaded from s3",
+        help="Where do you want to store the pre-trained models downloaded from huggingface.co",
     )
 
     parser.add_argument(
@@ -712,7 +714,11 @@ def main():
         bool(args.local_rank != -1),
         args.fp16,
     )
-
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     # Set seed
     set_seed(args)
 
@@ -730,6 +736,7 @@ def main():
         args.tokenizer_name if args.tokenizer_name else args.model_name_or_path,
         do_lower_case=args.do_lower_case,
         cache_dir=args.cache_dir if args.cache_dir else None,
+        use_fast=False,  # SquadDataset is not compatible with Fast tokenizers which have a smarter overflow handeling
     )
     model = AutoModelForQuestionAnswering.from_pretrained(
         args.model_name_or_path,
@@ -778,7 +785,10 @@ def main():
 
         # Load a trained model and vocabulary that you have fine-tuned
         model = AutoModelForQuestionAnswering.from_pretrained(args.output_dir)  # , force_download=True)
-        tokenizer = AutoTokenizer.from_pretrained(args.output_dir, do_lower_case=args.do_lower_case)
+
+        # SquadDataset is not compatible with Fast tokenizers which have a smarter overflow handeling
+        # So we use use_fast=False here for now until Fast-tokenizer-compatible-examples are out
+        tokenizer = AutoTokenizer.from_pretrained(args.output_dir, do_lower_case=args.do_lower_case, use_fast=False)
         model.to(args.device)
 
     # Evaluation - we can ask to evaluate all the checkpoints (sub-directories) in a directory

examples/question-answering/run_squad_trainer.py

@@ -22,9 +22,11 @@ import sys
 from dataclasses import dataclass, field
 from typing import Optional
 
+import transformers
 from transformers import AutoConfig, AutoModelForQuestionAnswering, AutoTokenizer, HfArgumentParser, SquadDataset
 from transformers import SquadDataTrainingArguments as DataTrainingArguments
 from transformers import Trainer, TrainingArguments
+from transformers.trainer_utils import is_main_process
 
 
 logger = logging.getLogger(__name__)
@@ -49,7 +51,8 @@ class ModelArguments:
     # If you want to tweak more attributes on your tokenizer, you should do it in a distinct script,
     # or just modify its tokenizer_config.json.
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
     )
 
 
@@ -91,6 +94,11 @@ def main():
         bool(training_args.local_rank != -1),
         training_args.fp16,
     )
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(training_args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+        transformers.utils.logging.enable_default_handler()
+        transformers.utils.logging.enable_explicit_format()
     logger.info("Training/evaluation parameters %s", training_args)
 
     # Prepare Question-Answering task
@@ -107,6 +115,7 @@ def main():
     tokenizer = AutoTokenizer.from_pretrained(
         model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path,
         cache_dir=model_args.cache_dir,
+        use_fast=False,  # SquadDataset is not compatible with Fast tokenizers which have a smarter overflow handeling
     )
     model = AutoModelForQuestionAnswering.from_pretrained(
         model_args.model_name_or_path,

examples/question-answering/run_tf_squad.py

@@ -33,6 +33,12 @@ from transformers import (
     squad_convert_examples_to_features,
 )
 from transformers.data.processors.squad import SquadV1Processor, SquadV2Processor
+from transformers.utils import logging as hf_logging
+
+
+hf_logging.set_verbosity_info()
+hf_logging.enable_default_handler()
+hf_logging.enable_explicit_format()
 
 
 logger = logging.getLogger(__name__)
@@ -57,7 +63,8 @@ class ModelArguments:
     # If you want to tweak more attributes on your tokenizer, you should do it in a distinct script,
     # or just modify its tokenizer_config.json.
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
     )
 
 

examples/rag/distributed_retriever.py

@@ -27,7 +27,7 @@ class RagPyTorchDistributedRetriever(RagRetriever):
             It is used to decode the question and then use the generator_tokenizer.
         generator_tokenizer (:class:`~transformers.PretrainedTokenizer`):
             The tokenizer used for the generator part of the RagModel.
-        index (:class:`~transformers.retrieval_rag.Index`, optional, defaults to the one defined by the configuration):
+        index (:class:`~transformers.models.rag.retrieval_rag.Index`, optional, defaults to the one defined by the configuration):
             If specified, use this index instead of the one built using the configuration
     """
 

examples/rag/eval_rag.py

@@ -95,7 +95,7 @@ def evaluate_batch_retrieval(args, rag_model, questions):
         truncation=True,
     )["input_ids"].to(args.device)
 
-    question_enc_outputs = rag_model.rag.question_encoder(retriever_input_ids, return_dict=True)
+    question_enc_outputs = rag_model.rag.question_encoder(retriever_input_ids)
     question_enc_pool_output = question_enc_outputs.pooler_output
 
     result = rag_model.retriever(

examples/rag/finetune.py

@@ -204,7 +204,6 @@ class GenerativeQAModule(BaseTransformer):
             decoder_input_ids=decoder_input_ids,
             use_cache=False,
             labels=lm_labels,
-            return_dict=True,
             **rag_kwargs,
         )
 

examples/rag/finetune.sh

@@ -7,7 +7,7 @@ export PYTHONPATH="../":"${PYTHONPATH}"
 python examples/rag/finetune.py \
     --data_dir $DATA_DIR \
     --output_dir $OUTPUT_DIR \
-    --model_name_or_path $MODLE_NAME_OR_PATH \
+    --model_name_or_path $MODEL_NAME_OR_PATH \
     --model_type rag_sequence \
     --fp16 \
     --gpus 8 \

examples/rag/test_distributed_retriever.py

@@ -11,16 +11,12 @@ import numpy as np
 from datasets import Dataset
 
 import faiss
-from transformers.configuration_bart import BartConfig
-from transformers.configuration_dpr import DPRConfig
-from transformers.configuration_rag import RagConfig
+from transformers import BartConfig, BartTokenizer, DPRConfig, DPRQuestionEncoderTokenizer, RagConfig
 from transformers.file_utils import is_datasets_available, is_faiss_available, is_psutil_available, is_torch_available
-from transformers.retrieval_rag import CustomHFIndex
-from transformers.testing_utils import require_torch_non_multigpu_but_fix_me
-from transformers.tokenization_bart import BartTokenizer
-from transformers.tokenization_bert import VOCAB_FILES_NAMES as DPR_VOCAB_FILES_NAMES
-from transformers.tokenization_dpr import DPRQuestionEncoderTokenizer
-from transformers.tokenization_roberta import VOCAB_FILES_NAMES as BART_VOCAB_FILES_NAMES
+from transformers.models.bert.tokenization_bert import VOCAB_FILES_NAMES as DPR_VOCAB_FILES_NAMES
+from transformers.models.rag.retrieval_rag import CustomHFIndex
+from transformers.models.roberta.tokenization_roberta import VOCAB_FILES_NAMES as BART_VOCAB_FILES_NAMES
+from transformers.testing_utils import require_torch_non_multi_gpu_but_fix_me
 
 
 sys.path.append(os.path.join(os.getcwd()))  # noqa: E402 # noqa: E402 # isort:skip
@@ -137,7 +133,7 @@ class RagRetrieverTest(TestCase):
             question_encoder=DPRConfig().to_dict(),
             generator=BartConfig().to_dict(),
         )
-        with patch("transformers.retrieval_rag.load_dataset") as mock_load_dataset:
+        with patch("transformers.models.rag.retrieval_rag.load_dataset") as mock_load_dataset:
             mock_load_dataset.return_value = dataset
             retriever = RagPyTorchDistributedRetriever(
                 config,
@@ -179,7 +175,7 @@ class RagRetrieverTest(TestCase):
             retriever.init_retrieval(port)
         return retriever
 
-    @require_torch_non_multigpu_but_fix_me
+    @require_torch_non_multi_gpu_but_fix_me
     def test_pytorch_distributed_retriever_retrieve(self):
         n_docs = 1
         retriever = self.get_dummy_pytorch_distributed_retriever(init_retrieval=True)
@@ -195,7 +191,7 @@ class RagRetrieverTest(TestCase):
         self.assertEqual(doc_dicts[1]["id"][0], "0")  # max inner product is reached with first doc
         self.assertListEqual(doc_ids.tolist(), [[1], [0]])
 
-    @require_torch_non_multigpu_but_fix_me
+    @require_torch_non_multi_gpu_but_fix_me
     def test_custom_hf_index_retriever_retrieve(self):
         n_docs = 1
         retriever = self.get_dummy_custom_hf_index_retriever(init_retrieval=True, from_disk=False)
@@ -211,7 +207,7 @@ class RagRetrieverTest(TestCase):
         self.assertEqual(doc_dicts[1]["id"][0], "0")  # max inner product is reached with first doc
         self.assertListEqual(doc_ids.tolist(), [[1], [0]])
 
-    @require_torch_non_multigpu_but_fix_me
+    @require_torch_non_multi_gpu_but_fix_me
     def test_custom_pytorch_distributed_retriever_retrieve_from_disk(self):
         n_docs = 1
         retriever = self.get_dummy_custom_hf_index_retriever(init_retrieval=True, from_disk=True)

examples/requirements.txt

@@ -18,3 +18,4 @@ fire
 pytest
 conllu
 sentencepiece != 0.1.92
+protobuf

examples/seq2seq/README.md

@@ -380,7 +380,7 @@ cp xsum/test* all_pl
 then use `all_pl` as DATA in the command above.
 
 #### Direct Knowledge Distillation (KD)
-+ In this method, we use try to enforce that the student and teacher produce similar encoder_outputs, logits, and hidden_states using `BartSummarizationDistiller`.
++ In this method, we use try to enforce that the student and teacher produce similar encoder_outputs, logits, and hidden_states using `SummarizationDistiller`.
 + This method was used for `sshleifer/distilbart-xsum-12-6`, `6-6`, and `9-6` checkpoints were produced.
 + You must use [`distillation.py`](./distillation.py). Note that this command initializes the student for you.
 

examples/seq2seq/builtin_trainer/finetune.sh

@@ -3,7 +3,8 @@
 python finetune_trainer.py \
     --learning_rate=3e-5 \
     --fp16 \
-    --do_train --do_eval --do_predict --evaluate_during_training \
+    --do_train --do_eval --do_predict \
+    --evaluation_strategy steps \
     --predict_with_generate \
     --n_val 1000 \
     "$@"