Release: v4.4.2

* Fix distributed evaluation

* Use logger

.circleci/config.yml

@@ -77,9 +77,10 @@ jobs:
                   keys:
                       - v0.4-torch_and_tf-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
+            - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev
             - run: pip install --upgrade pip
-            - run: pip install .[sklearn,tf-cpu,torch,testing,sentencepiece]
-            - run: pip install tapas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.7.0+cpu.html
+            - run: pip install .[sklearn,tf-cpu,torch,testing,sentencepiece,speech]
+            - run: pip install tapas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.8.0+cpu.html
             - save_cache:
                 key: v0.4-{{ checksum "setup.py" }}
                 paths:
@@ -90,6 +91,34 @@ jobs:
             - store_artifacts:
                   path: ~/transformers/reports
 
+    run_tests_torch_and_flax:
+        working_directory: ~/transformers
+        docker:
+            - image: circleci/python:3.6
+        environment:
+            OMP_NUM_THREADS: 1
+        resource_class: xlarge
+        parallelism: 1
+        steps:
+            - checkout
+            - restore_cache:
+                  keys:
+                      - v0.4-torch_and_flax-{{ checksum "setup.py" }}
+                      - v0.4-{{ checksum "setup.py" }}
+            - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev
+            - run: pip install --upgrade pip
+            - run: pip install .[sklearn,flax,torch,testing,sentencepiece,speech]
+            - run: pip install tapas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.8.0+cpu.html
+            - save_cache:
+                key: v0.4-{{ checksum "setup.py" }}
+                paths:
+                    - '~/.cache/pip'
+            - run: RUN_PT_FLAX_CROSS_TESTS=1 python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_torch_and_flax ./tests/ -m is_pt_flax_cross_test --durations=0 | tee tests_output.txt
+            - store_artifacts:
+                  path: ~/transformers/tests_output.txt
+            - store_artifacts:
+                  path: ~/transformers/reports
+
     run_tests_torch:
         working_directory: ~/transformers
         docker:
@@ -104,9 +133,10 @@ jobs:
                   keys:
                       - v0.4-torch-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
+            - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev
             - run: pip install --upgrade pip
-            - run: pip install .[sklearn,torch,testing,sentencepiece]
-            - run: pip install tapas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.7.0+cpu.html
+            - run: pip install .[sklearn,torch,testing,sentencepiece,speech]
+            - run: pip install tapas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.8.0+cpu.html
             - save_cache:
                   key: v0.4-torch-{{ checksum "setup.py" }}
                   paths:
@@ -158,7 +188,7 @@ 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,sentencepiece]
+            - run: sudo pip install .[flax,testing,sentencepiece]
             - save_cache:
                   key: v0.4-flax-{{ checksum "setup.py" }}
                   paths:
@@ -183,9 +213,10 @@ jobs:
                   keys:
                       - v0.4-torch-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
+            - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev
             - run: pip install --upgrade pip
-            - run: pip install .[sklearn,torch,testing,sentencepiece]
-            - run: pip install tapas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.7.0+cpu.html
+            - run: pip install .[sklearn,torch,testing,sentencepiece,speech]
+            - run: pip install tapas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.8.0+cpu.html
             - save_cache:
                   key: v0.4-torch-{{ checksum "setup.py" }}
                   paths:
@@ -300,13 +331,14 @@ jobs:
                   keys:
                       - v0.4-build_doc-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
+            - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev
             - run: pip install --upgrade pip
             - run: pip install ."[all, docs]"
             - save_cache:
                   key: v0.4-build_doc-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
-            - run: cd docs && make html SPHINXOPTS="-W"
+            - run: cd docs && make html SPHINXOPTS="-W -j 4"
             - store_artifacts:
                 path: ./docs/_build
 
@@ -413,6 +445,7 @@ workflows:
             - run_examples_torch
             - run_tests_custom_tokenizers
             - run_tests_torch_and_tf
+            - run_tests_torch_and_flax
             - run_tests_torch
             - run_tests_tf
             - run_tests_flax
@@ -421,15 +454,15 @@ workflows:
             - run_tests_git_lfs
             - build_doc
             - deploy_doc: *workflow_filters
-    tpu_testing_jobs:
-        triggers:
-            - schedule:
-                # Set to run at the first minute of every hour.
-                cron: "0 8 * * *"
-                filters:
-                    branches:
-                        only:
-                            - master
-        jobs:
-            - cleanup-gke-jobs
-            - run_examples_tpu
+#    tpu_testing_jobs:
+#        triggers:
+#            - schedule:
+#                # Set to run at the first minute of every hour.
+#                cron: "0 8 * * *"
+#                filters:
+#                    branches:
+#                        only:
+#                            - master
+#        jobs:
+#            - cleanup-gke-jobs
+#            - run_examples_tpu

.circleci/deploy.sh

@@ -3,6 +3,7 @@ cd docs
 function deploy_doc(){
 	echo "Creating doc at commit $1 and pushing to folder $2"
 	git checkout $1
+	pip install -U ..
 	if [ ! -z "$2" ]
 	then
 		if [ "$2" == "master" ]; then
@@ -45,7 +46,7 @@ deploy_doc "6f5a12a" v2.7.0
 deploy_doc "11c3257" v2.8.0
 deploy_doc "e7cfc1a" v2.9.0
 deploy_doc "7cb203f" v2.9.1
-deploy_doc "10d7239" v2.10.0 
+deploy_doc "10d7239" v2.10.0
 deploy_doc "b42586e" v2.11.0
 deploy_doc "7fb8bdf" v3.0.2
 deploy_doc "4b3ee9c" v3.1.0
@@ -53,5 +54,7 @@ deploy_doc "3ebb1b3" v3.2.0
 deploy_doc "0613f05" v3.3.1
 deploy_doc "eb0e0ce" v3.4.0
 deploy_doc "818878d" v3.5.1
-deploy_doc "c781171" v4.0.0
-deploy_doc "bfa4ccf" # v4.1.1 Latest stable release
+deploy_doc "c781171" v4.0.1
+deploy_doc "bfa4ccf" v4.1.1
+deploy_doc "7d9a9d0" v4.2.2
+deploy_doc "bae0c79"  # v4.3.3 Latest stable release

.gitattributes

@@ -0,0 +1,3 @@
+*.py	eol=lf
+*.rst	eol=lf
+*.md	eol=lf

.github/ISSUE_TEMPLATE/bug-report.md

@@ -25,32 +25,40 @@ assignees: ''
  If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
  Please tag fewer than 3 people.
 
- albert, bert, GPT2, XLM: @LysandreJik
- tokenizers: @mfuntowicz
- Trainer: @sgugger
- Speed and Memory Benchmarks: @patrickvonplaten
- Model Cards: @julien-c
- TextGeneration: @TevenLeScao
- examples/distillation: @VictorSanh
- nlp datasets: [different repo](https://github.com/huggingface/nlp)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
- Text Generation: @patrickvonplaten @TevenLeScao
- Blenderbot: @patrickvonplaten
- Bart: @patrickvonplaten
- Marian: @patrickvonplaten
- Pegasus: @patrickvonplaten
- mBART: @patrickvonplaten
- T5: @patrickvonplaten
- Longformer/Reformer: @patrickvonplaten
- TransfoXL/XLNet: @TevenLeScao
- RAG: @patrickvonplaten, @lhoestq
- FSMT: @stas00
- examples/seq2seq: @patil-suraj
- examples/bert-loses-patience: @JetRunner
- ray/raytune: @richardliaw @amogkam
- tensorflow: @jplu
- examples/token-classification: @stefan-it
- documentation: @sgugger
+Models:
+
+- albert, bert, xlm: @LysandreJik
+- blenderbot, bart, marian, pegasus, encoderdecoder,  t5: @patrickvonplaten, @patil-suraj
+- longformer, reformer, transfoxl, xlnet: @patrickvonplaten
+- fsmt: @stas00
+- funnel: @sgugger
+- gpt2: @patrickvonplaten, @LysandreJik
+- rag: @patrickvonplaten, @lhoestq
+- tensorflow: @jplu
+
+Library:
+
+- benchmarks: @patrickvonplaten
+- deepspeed: @stas00
+- ray/raytune: @richardliaw, @amogkam
+- text generation: @patrickvonplaten
+- tokenizers: @LysandreJik
+- trainer: @sgugger
+- pipelines: @LysandreJik
+
+Documentation: @sgugger
+
+HF projects:
+
+- nlp datasets: [different repo](https://github.com/huggingface/nlp)
+- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
+
+Examples:
+
+- maintained examples (not research project or legacy): @sgugger, @patil-suraj
+- research_projects/bert-loses-patience: @JetRunner
+- research_projects/distillation: @VictorSanh
+
  -->
 
 ## Information

.github/PULL_REQUEST_TEMPLATE.md

@@ -37,26 +37,38 @@ members/contributors which may be interested in your PR.
  If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
  Please tag fewer than 3 people.
 
- albert, bert, XLM: @LysandreJik
- GPT2: @LysandreJik, @patrickvonplaten
- tokenizers: @mfuntowicz
- Trainer: @sgugger
- Benchmarks: @patrickvonplaten
- Model Cards: @julien-c
- examples/distillation: @VictorSanh
- nlp datasets: [different repo](https://github.com/huggingface/nlp)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
- Text Generation: @patrickvonplaten, @TevenLeScao
- Blenderbot, Bart, Marian, Pegasus: @patrickvonplaten
- T5: @patrickvonplaten
- Rag: @patrickvonplaten, @lhoestq
- EncoderDecoder: @patrickvonplaten
- Longformer, Reformer: @patrickvonplaten
- TransfoXL, XLNet: @TevenLeScao, @patrickvonplaten
- examples/seq2seq: @patil-suraj
- examples/bert-loses-patience: @JetRunner
- tensorflow: @jplu
- examples/token-classification: @stefan-it
- documentation: @sgugger
- FSMT: @stas00
+Models:
+
+- albert, bert, xlm: @LysandreJik
+- blenderbot, bart, marian, pegasus, encoderdecoder,  t5: @patrickvonplaten, @patil-suraj
+- longformer, reformer, transfoxl, xlnet: @patrickvonplaten
+- fsmt: @stas00
+- funnel: @sgugger
+- gpt2: @patrickvonplaten, @LysandreJik
+- rag: @patrickvonplaten, @lhoestq
+- tensorflow: @jplu
+
+Library:
+
+- benchmarks: @patrickvonplaten
+- deepspeed: @stas00
+- ray/raytune: @richardliaw, @amogkam
+- text generation: @patrickvonplaten
+- tokenizers: @n1t0, @LysandreJik
+- trainer: @sgugger
+- pipelines: @LysandreJik
+
+Documentation: @sgugger
+
+HF projects:
+
+- nlp datasets: [different repo](https://github.com/huggingface/nlp)
+- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
+
+Examples:
+
+- maintained examples (not research project or legacy): @sgugger, @patil-suraj
+- research_projects/bert-loses-patience: @JetRunner
+- research_projects/distillation: @VictorSanh
+
  -->

.github/conda/meta.yaml

@@ -14,7 +14,7 @@ requirements:
   host:
     - python
     - pip
-    - numpy
+    - numpy >=1.17
     - dataclasses
     - packaging
     - filelock
@@ -23,10 +23,10 @@ requirements:
     - sacremoses
     - regex !=2019.12.17
     - protobuf
-    - tokenizers ==0.9.4
+    - tokenizers >=0.10.1,<0.11.0
   run:
     - python
-    - numpy
+    - numpy >=1.17
     - dataclasses
     - packaging
     - filelock
@@ -35,7 +35,7 @@ requirements:
     - sacremoses
     - regex !=2019.12.17
     - protobuf
-    - tokenizers ==0.9.4
+    - tokenizers >=0.10.1,<0.11.0
 
 test:
   imports:

.github/stale.yml

@@ -1,18 +0,0 @@
-# Number of days of inactivity before an issue becomes stale
-daysUntilStale: 60
-# Number of days of inactivity before a stale issue is closed
-daysUntilClose: 7
-# Issues with these labels will never be considered stale
-exemptLabels:
-  - pinned
-  - security
-  - Feature request
-# Label to use when marking an issue as stale
-staleLabel: wontfix
-# Comment to post when marking an issue as stale. Set to `false` to disable
-markComment: >
-  This issue has been automatically marked as stale because it has not had
-  recent activity. It will be closed if no further activity occurs. Thank you
-  for your contributions.
-# Comment to post when closing a stale issue. Set to `false` to disable
-closeComment: false

.github/workflows/model-templates.yml

@@ -1,12 +1,13 @@
 name: Model templates runner
 
 on:
-  push:
+  pull_request:
     paths:
       - "src/**"
       - "tests/**"
       - ".github/**"
       - "templates/**"
+    types: [assigned, opened, synchronize, reopened]
 
 jobs:
   run_tests_templates:

.github/workflows/release-conda.yml

@@ -37,7 +37,8 @@ jobs:
       - name: Build conda packages
         run: |
           conda info
-          conda build .github/conda
+          conda list
+          conda-build .github/conda
 
       - name: Upload to Anaconda
-        run: anaconda upload `conda build .github/conda --output` --force
+        run: anaconda upload `conda-build .github/conda --output` --force

.github/workflows/self-push.yml

@@ -10,143 +10,91 @@ on:
       - "tests/**"
       - ".github/**"
       - "templates/**"
-  # pull_request:
   repository_dispatch:
 
-
 jobs:
   run_tests_torch_gpu:
-    runs-on: [self-hosted, gpu, single-gpu]
+    runs-on: [self-hosted, docker-gpu, single-gpu]
+    container:
+      image: pytorch/pytorch:1.8.0-cuda11.1-cudnn8-runtime
+      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - uses: actions/checkout@v2
-      - name: Python version
+      - name: Launcher docker
+        uses: actions/checkout@v2
+
+      - name: NVIDIA-SMI
         run: |
-          which python
-          python --version
-          pip --version
-
-      - name: Current dir
-        run: pwd
-      - run: nvidia-smi
-
-      - name: Loading cache.
-        uses: actions/cache@v2
-        id: cache
-        with:
-          path: .env
-          key: v1.1-tests_torch_gpu-${{ hashFiles('setup.py') }}
-
-      - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
-        run: |
-          python -m venv .env
-          source .env/bin/activate
-          which python
-          python --version
-          pip --version
+          nvidia-smi
 
       - name: Install dependencies
         run: |
-          source .env/bin/activate
+          apt -y update && apt install -y libsndfile1-dev
           pip install --upgrade pip
-          pip install .[torch,sklearn,testing,onnxruntime,sentencepiece]
-          pip install git+https://github.com/huggingface/datasets
-          pip install pandas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.7.0+cu102.html
+          pip install .[sklearn,testing,onnxruntime,sentencepiece,speech]
 
       - name: Are GPUs recognized by our DL frameworks
         run: |
-          source .env/bin/activate
           python -c "import torch; print('Cuda available:', torch.cuda.is_available())"
+          python -c "import torch; print('Cuda version:', torch.version.cuda)"
+          python -c "import torch; print('CuDNN version:', torch.backends.cudnn.version())"
           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
-          CUDA_VISIBLE_DEVICES: 0
+          OMP_NUM_THREADS: 8
+          MKL_NUM_THREADS: 8
+          HF_HOME: /mnt/cache
         run: |
-          source .env/bin/activate
-          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_torch_gpu tests
+          python -m pytest -n 2 --dist=loadfile --make-reports=tests_torch_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
         run: cat reports/tests_torch_gpu_failures_short.txt
-        
+
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
         with:
           name: run_all_tests_torch_gpu_test_reports
           path: reports
-                  
 
   run_tests_tf_gpu:
-    runs-on: [self-hosted, gpu, single-gpu]
+    runs-on: [self-hosted, docker-gpu, single-gpu]
+    container:
+      image: tensorflow/tensorflow:2.4.1-gpu
+      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - uses: actions/checkout@v2
-      - name: Python version
-        run: |
-          which python
-          python --version
-          pip --version
-      - name: Current dir
-        run: pwd
-      - run: nvidia-smi
+      - name: Launcher docker
+        uses: actions/checkout@v2
 
-      - name: Loading cache.
-        uses: actions/cache@v2
-        id: cache
-        with:
-          path: .env
-          key: v1.1-tests_tf_gpu-${{ hashFiles('setup.py') }}
-
-      - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
+      - name: NVIDIA-SMI
         run: |
-          python -m venv .env
-          source .env/bin/activate
-          which python
-          python --version
-          pip --version
+          nvidia-smi
 
       - name: Install dependencies
         run: |
-          source .env/bin/activate
           pip install --upgrade pip
-          pip install .[tf,sklearn,testing,onnxruntime,sentencepiece]
-          pip install git+https://github.com/huggingface/datasets
+          pip install .[sklearn,testing,onnxruntime,sentencepiece]
 
       - name: Are GPUs recognized by our DL frameworks
         run: |
-          source .env/bin/activate
           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
-          CUDA_VISIBLE_DEVICES: 0
+          OMP_NUM_THREADS: 8
+          MKL_NUM_THREADS: 8
+          TF_NUM_INTRAOP_THREADS: 8
+          TF_NUM_INTEROP_THREADS: 1
+          HF_HOME: /mnt/cache
         run: |
-          source .env/bin/activate
-          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_tf_gpu tests
+          python -m pytest -n 2 --dist=loadfile --make-reports=tests_tf_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
         run: cat reports/tests_tf_gpu_failures_short.txt
-        
+
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
@@ -154,58 +102,45 @@ jobs:
           name: run_all_tests_tf_gpu_test_reports
           path: reports
 
+
   run_tests_torch_multi_gpu:
-    runs-on: [self-hosted, gpu, multi-gpu]
+    runs-on: [self-hosted, docker-gpu, multi-gpu]
+    container:
+      image: pytorch/pytorch:1.8.0-cuda11.1-cudnn8-runtime
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - uses: actions/checkout@v2
-      - name: Python version
+      - name: Launcher docker
+        uses: actions/checkout@v2
+
+      - name: NVIDIA-SMI
         run: |
-          which python
-          python --version
-          pip --version
+          nvidia-smi
 
-      - name: Current dir
-        run: pwd
-      - run: nvidia-smi
-
-      - name: Loading cache.
-        uses: actions/cache@v2
-        id: cache
-        with:
-          path: .env
-          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: |
-          python -m venv .env
-          source .env/bin/activate
-          which python
-          python --version
-          pip --version
       - name: Install dependencies
         run: |
-          source .env/bin/activate
+          apt -y update && apt install -y libsndfile1-dev
           pip install --upgrade pip
-          pip install .[torch,sklearn,testing,onnxruntime,sentencepiece]
-          pip install git+https://github.com/huggingface/datasets
-          pip install pandas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.7.0+cu102.html
+          pip install .[sklearn,testing,onnxruntime,sentencepiece,speech]
 
       - name: Are GPUs recognized by our DL frameworks
         run: |
-          source .env/bin/activate
           python -c "import torch; print('Cuda available:', torch.cuda.is_available())"
+          python -c "import torch; print('Cuda version:', torch.version.cuda)"
+          python -c "import torch; print('CuDNN version:', torch.backends.cudnn.version())"
           python -c "import torch; print('Number of GPUs available:', torch.cuda.device_count())"
 
       - name: Run all non-slow tests on GPU
         env:
-          OMP_NUM_THREADS: 1
+          OMP_NUM_THREADS: 8
+          MKL_NUM_THREADS: 8
+          MKL_SERVICE_FORCE_INTEL: 1
+          HF_HOME: /mnt/cache
         run: |
-          source .env/bin/activate
-          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_torch_multi_gpu tests
+          python -m pytest -n 2 --dist=loadfile --make-reports=tests_torch_multi_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
-        run: cat reports/tests_torch_multi_gpu_failures_short.txt          
+        run: cat reports/tests_torch_multi_gpu_failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
@@ -215,52 +150,37 @@ jobs:
           path: reports
 
   run_tests_tf_multi_gpu:
-    runs-on: [self-hosted, gpu, multi-gpu]
+    runs-on: [self-hosted, docker-gpu, multi-gpu]
+    container:
+      image: tensorflow/tensorflow:2.4.1-gpu
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - uses: actions/checkout@v2
-      - name: Python version
+      - name: Launcher docker
+        uses: actions/checkout@v2
+
+      - name: NVIDIA-SMI
         run: |
-          which python
-          python --version
-          pip --version
+          nvidia-smi
 
-      - name: Current dir
-        run: pwd
-      - run: nvidia-smi
-
-      - name: Loading cache.
-        uses: actions/cache@v2
-        id: cache
-        with:
-          path: .env
-          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: |
-          python -m venv .env
-          source .env/bin/activate
-          which python
-          python --version
-          pip --version
       - name: Install dependencies
         run: |
-          source .env/bin/activate
           pip install --upgrade pip
-          pip install .[tf,sklearn,testing,onnxruntime,sentencepiece]
-          pip install git+https://github.com/huggingface/datasets
+          pip install .[sklearn,testing,onnxruntime,sentencepiece]
 
       - name: Are GPUs recognized by our DL frameworks
         run: |
-          source .env/bin/activate
           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: Run all non-slow tests on GPU
         env:
-          OMP_NUM_THREADS: 1
+          OMP_NUM_THREADS: 8
+          MKL_NUM_THREADS: 8
+          TF_NUM_INTRAOP_THREADS: 8
+          TF_NUM_INTEROP_THREADS: 1
+          HF_HOME: /mnt/cache
         run: |
-          source .env/bin/activate
-          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_tf_multi_gpu tests
+          python -m pytest -n 2 --dist=loadfile --make-reports=tests_tf_multi_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
@@ -272,4 +192,22 @@ jobs:
         with:
           name: run_all_tests_tf_multi_gpu_test_reports
           path: reports
-          
+
+  send_results:
+    name: Send results to webhook
+    runs-on: ubuntu-latest
+    if: always()
+    needs: [run_tests_torch_gpu, run_tests_tf_gpu, run_tests_torch_multi_gpu, run_tests_tf_multi_gpu]
+    steps:
+      - uses: actions/checkout@v2
+
+      - uses: actions/download-artifact@v2
+
+      - name: Send message to Slack
+        env:
+          CI_SLACK_BOT_TOKEN: ${{ secrets.CI_SLACK_BOT_TOKEN }}
+          CI_SLACK_CHANNEL_ID: ${{ secrets.CI_SLACK_CHANNEL_ID }}
+
+        run: |
+          pip install slack_sdk
+          python utils/notification_service.py push

.github/workflows/self-scheduled.yml

@@ -1,82 +1,63 @@
-# configuration notes:
-#
-# - `source .env/bin/activate` is currently needed to be run first thing first in each step. Otherwise
-# the step uses the system-wide python interpreter.
-
 name: Self-hosted runner (scheduled)
 
 on:
+  push:
+    branches:
+      - multi_ci_*
   repository_dispatch:
   schedule:
     - cron: "0 0 * * *"
 
 jobs:
   run_all_tests_torch_gpu:
-    runs-on: [self-hosted, gpu, single-gpu]
+    runs-on: [self-hosted, docker-gpu, single-gpu]
+    container:
+      image: pytorch/pytorch:1.8.0-cuda11.1-cudnn8-runtime
+      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - uses: actions/checkout@v2
+      - name: Launcher docker
+        uses: actions/checkout@v2
 
-      - name: Loading cache.
-        uses: actions/cache@v2
-        id: cache
-        with:
-          path: .env
-          key: v  1.1-slow_tests_torch_gpu-${{ hashFiles('setup.py') }}
-
-      - name: Python version
+      - name: NVIDIA-SMI
         run: |
-          which python
-          python --version
-          pip --version
-
-      - name: Current dir
-        run: pwd
-      - run: nvidia-smi
-
-      - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
-        if: steps.cache.outputs.cache-hit != 'true'
-        run: |
-          python -m venv .env
-          source .env/bin/activate
-          which python
-          python --version
-          pip --version
+          nvidia-smi
 
       - name: Install dependencies
         run: |
-          source .env/bin/activate
+          apt -y update && apt install -y libsndfile1-dev
           pip install --upgrade pip
-          pip install .[torch,sklearn,testing,onnxruntime,sentencepiece]
-          pip install git+https://github.com/huggingface/datasets
-          pip list
+          pip install .[sklearn,testing,onnxruntime,sentencepiece,speech]
 
       - name: Are GPUs recognized by our DL frameworks
         run: |
-          source .env/bin/activate
           python -c "import torch; print('Cuda available:', torch.cuda.is_available())"
+          python -c "import torch; print('Cuda version:', torch.version.cuda)"
+          python -c "import torch; print('CuDNN version:', torch.backends.cudnn.version())"
           python -c "import torch; print('Number of GPUs available:', torch.cuda.device_count())"
 
       - name: Run all tests on GPU
         env:
-          OMP_NUM_THREADS: 1
+          OMP_NUM_THREADS: 16
+          MKL_NUM_THREADS: 16
           RUN_SLOW: yes
+          HF_HOME: /mnt/cache
         run: |
-          source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_torch_gpu tests
+          python -m pytest -n 1 --dist=loadfile --make-reports=tests_torch_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
         run: cat reports/tests_torch_gpu_failures_short.txt
-        
+
       - name: Run examples tests on GPU
         if: ${{ always() }}
         env:
-          OMP_NUM_THREADS: 1
+          OMP_NUM_THREADS: 16
+          MKL_NUM_THREADS: 16
           RUN_SLOW: yes
+          HF_HOME: /mnt/cache
         run: |
-          source .env/bin/activate
           pip install -r examples/_tests_requirements.txt
-          python -m pytest -n 1 --dist=loadfile -s --make-reports=examples_torch_gpu examples
+          python -m pytest -n 1 --dist=loadfile --make-reports=examples_torch_gpu examples
 
       - name: Failure short reports
         if: ${{ always() }}
@@ -85,13 +66,13 @@ jobs:
       - name: Run all pipeline tests on GPU
         if: ${{ always() }}
         env:
-          TF_FORCE_GPU_ALLOW_GROWTH: "true"
-          OMP_NUM_THREADS: 1
+          OMP_NUM_THREADS: 16
+          MKL_NUM_THREADS: 16
           RUN_SLOW: yes
           RUN_PIPELINE_TESTS: yes
+          HF_HOME: /mnt/cache
         run: |
-          source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s -m is_pipeline_test --make-reports=tests_torch_pipeline_gpu tests
+          python -m pytest -n 1 --dist=loadfile -m is_pipeline_test --make-reports=tests_torch_pipeline_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
@@ -104,60 +85,40 @@ jobs:
           name: run_all_tests_torch_gpu_test_reports
           path: reports
 
-
   run_all_tests_tf_gpu:
-    runs-on: [self-hosted, gpu, single-gpu]
+    runs-on: [self-hosted, docker-gpu, single-gpu]
+    container:
+      image: tensorflow/tensorflow:2.4.1-gpu
+      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - uses: actions/checkout@v2
+      - name: Launcher docker
+        uses: actions/checkout@v2
 
-      - name: Loading cache.
-        uses: actions/cache@v2
-        id: cache
-        with:
-          path: .env
-          key: v1.1-slow_tests_tf_gpu-${{ hashFiles('setup.py') }}
-
-      - name: Python version
+      - name: NVIDIA-SMI
         run: |
-          which python
-          python --version
-          pip --version
-
-      - name: Current dir
-        run: pwd
-      - run: nvidia-smi
-
-      - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
-        if: steps.cache.outputs.cache-hit != 'true'
-        run: |
-          python -m venv .env
-          source .env/bin/activate
-          which python
-          python --version
-          pip --version
+          nvidia-smi
 
       - name: Install dependencies
         run: |
-          source .env/bin/activate
           pip install --upgrade pip
-          pip install .[tf,sklearn,testing,onnxruntime,sentencepiece]
-          pip install git+https://github.com/huggingface/datasets
-          pip list
+          pip install .[sklearn,testing,onnx,sentencepiece]
 
       - name: Are GPUs recognized by our DL frameworks
         run: |
-          source .env/bin/activate
           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: Run all tests on GPU
         env:
-          OMP_NUM_THREADS: 1
           RUN_SLOW: yes
+          HF_HOME: /mnt/cache
+          OMP_NUM_THREADS: 16
+          TF_NUM_INTEROP_THREADS: 1
+          TF_NUM_INTRAOP_THREADS: 16
+          MKL_NUM_THREADS: 16
         run: |
-          source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_tf_gpu tests
-          
+          python -m pytest -n 1 --dist=loadfile --make-reports=tests_tf_gpu tests
+
       - name: Failure short reports
         if: ${{ always() }}
         run: cat reports/tests_tf_gpu_failures_short.txt
@@ -165,17 +126,19 @@ jobs:
       - name: Run all pipeline tests on GPU
         if: ${{ always() }}
         env:
-          TF_FORCE_GPU_ALLOW_GROWTH: "true"
-          OMP_NUM_THREADS: 1
           RUN_SLOW: yes
+          HF_HOME: /mnt/cache
+          OMP_NUM_THREADS: 16
           RUN_PIPELINE_TESTS: yes
+          TF_NUM_INTEROP_THREADS: 1
+          TF_NUM_INTRAOP_THREADS: 16
+          MKL_NUM_THREADS: 16
         run: |
-          source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s -m is_pipeline_test --make-reports=tests_tf_pipelines_gpu tests
+          python -m pytest -n 1 --dist=loadfile -m is_pipeline_test --make-reports=tests_tf_pipeline_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
-        run: cat reports/tests_tf_pipelines_gpu_failures_short.txt
+        run: cat reports/tests_tf_pipeline_gpu_failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
@@ -183,86 +146,57 @@ jobs:
         with:
           name: run_all_tests_tf_gpu_test_reports
           path: reports
-          
+
   run_all_tests_torch_multi_gpu:
-    runs-on: [self-hosted, gpu, multi-gpu]
+    runs-on: [self-hosted, docker-gpu, multi-gpu]
+    container:
+      image: pytorch/pytorch:1.8.0-cuda11.1-cudnn8-runtime
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - uses: actions/checkout@v2
+      - name: Launcher docker
+        uses: actions/checkout@v2
 
-      - name: Loading cache.
-        uses: actions/cache@v2
-        id: cache
-        with:
-          path: .env
-          key: v1.1-slow_tests_torch_multi_gpu-${{ hashFiles('setup.py') }}
-
-      - name: Python version
+      - name: NVIDIA-SMI
         run: |
-          which python
-          python --version
-          pip --version
-
-      - name: Current dir
-        run: pwd
-      - run: nvidia-smi
-
-      - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
-        if: steps.cache.outputs.cache-hit != 'true'
-        run: |
-          python -m venv .env
-          source .env/bin/activate
-          which python
-          python --version
-          pip --version
+          nvidia-smi
 
       - name: Install dependencies
         run: |
-          source .env/bin/activate
+          apt -y update && apt install -y libsndfile1-dev
           pip install --upgrade pip
-          pip install .[torch,sklearn,testing,onnxruntime,sentencepiece]
-          pip install git+https://github.com/huggingface/datasets
-          pip list
+          pip install .[sklearn,testing,onnxruntime,sentencepiece,speech]
 
       - name: Are GPUs recognized by our DL frameworks
         run: |
-          source .env/bin/activate
           python -c "import torch; print('Cuda available:', torch.cuda.is_available())"
+          python -c "import torch; print('Cuda version:', torch.version.cuda)"
+          python -c "import torch; print('CuDNN version:', torch.backends.cudnn.version())"
           python -c "import torch; print('Number of GPUs available:', torch.cuda.device_count())"
 
-      - name: Run all tests on multi-GPU
+      - name: Run all tests on GPU
         env:
-          OMP_NUM_THREADS: 1
           RUN_SLOW: yes
+          HF_HOME: /mnt/cache
+          OMP_NUM_THREADS: 16
+          MKL_NUM_THREADS: 16
+          MKL_SERVICE_FORCE_INTEL: 1
         run: |
-          source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_torch_multi_gpu tests
+          python -m pytest -n 1 --dist=loadfile --make-reports=tests_torch_multi_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
         run: cat reports/tests_torch_multi_gpu_failures_short.txt
 
-      - name: Run examples tests on multi-GPU
-        env:
-          OMP_NUM_THREADS: 1
-          RUN_SLOW: yes
-        run: |
-          source .env/bin/activate
-          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/tests_torch_examples_multi_gpu_failures_short.txt
-
-      - name: Run all pipeline tests on multi-GPU
+      - name: Run all pipeline tests on GPU
         if: ${{ always() }}
         env:
-          TF_FORCE_GPU_ALLOW_GROWTH: "true"
-          OMP_NUM_THREADS: 1
+          OMP_NUM_THREADS: 16
+          MKL_NUM_THREADS: 16
           RUN_SLOW: yes
           RUN_PIPELINE_TESTS: yes
+          HF_HOME: /mnt/cache
         run: |
-          source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s -m is_pipeline_test --make-reports=tests_torch_pipeline_multi_gpu tests
+          python -m pytest -n 1 --dist=loadfile -m is_pipeline_test --make-reports=tests_torch_pipeline_multi_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
@@ -276,73 +210,56 @@ jobs:
           path: reports
 
   run_all_tests_tf_multi_gpu:
-    runs-on: [self-hosted, gpu, multi-gpu]
+    runs-on: [self-hosted, docker-gpu, multi-gpu]
+    container:
+      image: tensorflow/tensorflow:2.4.1-gpu
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - uses: actions/checkout@v2
+      - name: Launcher docker
+        uses: actions/checkout@v2
 
-      - name: Loading cache.
-        uses: actions/cache@v2
-        id: cache
-        with:
-          path: .env
-          key: v1.1-slow_tests_tf_multi_gpu-${{ hashFiles('setup.py') }}
-
-      - name: Python version
+      - name: NVIDIA-SMI
         run: |
-          which python
-          python --version
-          pip --version
-
-      - name: Current dir
-        run: pwd
-      - run: nvidia-smi
-
-      - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
-        if: steps.cache.outputs.cache-hit != 'true'
-        run: |
-          python -m venv .env
-          source .env/bin/activate
-          which python
-          python --version
-          pip --version
+          nvidia-smi
 
       - name: Install dependencies
         run: |
-          source .env/bin/activate
           pip install --upgrade pip
-          pip install .[tf,sklearn,testing,onnxruntime,sentencepiece]
-          pip install git+https://github.com/huggingface/datasets
-          pip list
+          pip install .[sklearn,testing,onnx,sentencepiece]
 
       - name: Are GPUs recognized by our DL frameworks
         run: |
-          source .env/bin/activate
           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: Run all tests on multi-GPU
+      - name: Run all tests on GPU
         env:
-          OMP_NUM_THREADS: 1
+          OMP_NUM_THREADS: 16
           RUN_SLOW: yes
+          MKL_NUM_THREADS: 16
+          TF_NUM_INTEROP_THREADS: 1
+          TF_NUM_INTRAOP_THREADS: 16
+          HF_HOME: /mnt/cache
         run: |
-          source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_tf_multi_gpu tests
+          python -m pytest -n 1 --dist=loadfile --make-reports=tests_tf_multi_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
         run: cat reports/tests_tf_multi_gpu_failures_short.txt
 
-      - name: Run all pipeline tests on multi-GPU
+      - name: Run all pipeline tests on GPU
         if: ${{ always() }}
         env:
-          TF_FORCE_GPU_ALLOW_GROWTH: "true"
-          OMP_NUM_THREADS: 1
+          OMP_NUM_THREADS: 16
           RUN_SLOW: yes
           RUN_PIPELINE_TESTS: yes
+          MKL_NUM_THREADS: 16
+          TF_NUM_INTEROP_THREADS: 1
+          TF_NUM_INTRAOP_THREADS: 16
+          HF_HOME: /mnt/cache
         run: |
-          source .env/bin/activate
-          python -m pytest -n 1 --dist=loadfile -s -m is_pipeline_test --make-reports=tests_tf_pipeline_multi_gpu tests
-          
+          python -m pytest -n 1 --dist=loadfile -m is_pipeline_test --make-reports=tests_tf_pipeline_multi_gpu tests
+
       - name: Failure short reports
         if: ${{ always() }}
         run: cat reports/tests_tf_pipeline_multi_gpu_failures_short.txt
@@ -353,4 +270,23 @@ jobs:
         with:
           name: run_all_tests_tf_multi_gpu_test_reports
           path: reports
-          
+
+  send_results:
+    name: Send results to webhook
+    runs-on: ubuntu-latest
+    if: always()
+    needs: [run_all_tests_torch_gpu, run_all_tests_tf_gpu, run_all_tests_torch_multi_gpu, run_all_tests_tf_multi_gpu]
+    steps:
+      - uses: actions/checkout@v2
+
+      - uses: actions/download-artifact@v2
+
+      - name: Send message to Slack
+        env:
+          CI_SLACK_BOT_TOKEN: ${{ secrets.CI_SLACK_BOT_TOKEN }}
+          CI_SLACK_CHANNEL_ID: ${{ secrets.CI_SLACK_CHANNEL_ID }}
+
+
+        run: |
+          pip install slack_sdk
+          python utils/notification_service.py scheduled

.github/workflows/stale.yml

@@ -0,0 +1,27 @@
+name: Stale Bot
+
+on:
+  schedule:
+    - cron: "0 0 * * *"
+
+jobs:
+  close_stale_issues:
+    name: Close Stale Issues
+    if: github.repository == 'huggingface/transformers'
+    runs-on: ubuntu-latest
+    env:
+      GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+    steps:
+    - uses: actions/checkout@v2
+
+    - name: Setup Python
+      uses: actions/setup-python@v1
+      with:
+        python-version: 3.7
+
+    - name: Install requirements
+      run: |
+        pip install PyGithub
+    - name: Close stale issues
+      run: |
+        python scripts/stale.py

ISSUES.md

@@ -207,6 +207,8 @@ You are not required to read the following guidelines before opening an issue. H
 
    Do not dispair if you can't figure it out from the begining, just share what you can and perhaps someone else will be able to help you at the forums.
 
+   If your setup involves any custom datasets, the best way to help us reproduce the problem is to create a [Google Colab notebook](https://colab.research.google.com/) that demonstrates the issue and once you verify that the issue still exists, include a link to that notebook in the Issue. Just make sure that you don't copy and paste the location bar url of the open notebook - as this is private and we won't be able to open it. Instead, you need to click on `Share` in the right upper corner of the notebook, select `Get Link` and then copy and paste the public link it will give to you.
+
 7. If you forked off some of this project's code or example applications, please, do not ask us to go into your code repository and figure out what you may have done. The code is already very complex and unless there is an easy way to do a diff and it's a small diff, it won't be possible to find someone with time on their hands to make a lengthy investigation. Albeit, you might find someone at the forums who will be generous to do this for you.
 
 8. Before reporting an issue, first, always try to update your environment to the latest official version of this library. We have no resources to go and debug older revisions, which could easily have bugs that have been fixed in the latest released version.

Makefile

@@ -27,6 +27,7 @@ extra_quality_checks: deps_table_update
 	python utils/check_dummies.py
 	python utils/check_repo.py
 	python utils/style_doc.py src/transformers docs/source --max_len 119
+	python utils/class_mapping_update.py
 
 # this target runs checks on all files
 quality:
@@ -68,3 +69,18 @@ test-examples:
 
 docs:
 	cd docs && make html SPHINXOPTS="-W -j 4"
+
+# Release stuff
+
+pre-release:
+	python utils/release.py
+
+pre-patch:
+	python utils/release.py --patch
+
+post-release:
+	python utils/release.py --post_release
+
+post-patch:
+	python utils/release.py --post_release --patch
+

README.md

@@ -55,7 +55,7 @@ Here are a few examples:
 - [Masked word completion with BERT](https://huggingface.co/bert-base-uncased?text=Paris+is+the+%5BMASK%5D+of+France)
 - [Name Entity Recognition with Electra](https://huggingface.co/dbmdz/electra-large-discriminator-finetuned-conll03-english?text=My+name+is+Sarah+and+I+live+in+London+city)
 - [Text generation with GPT-2](https://huggingface.co/gpt2?text=A+long+time+ago%2C+)
-- [Natural Langugage Inference with RoBERTa](https://huggingface.co/roberta-large-mnli?text=The+dog+was+lost.+Nobody+lost+any+animal)
+- [Natural Language Inference with RoBERTa](https://huggingface.co/roberta-large-mnli?text=The+dog+was+lost.+Nobody+lost+any+animal)
 - [Summarization with BART](https://huggingface.co/facebook/bart-large-cnn?text=The+tower+is+324+metres+%281%2C063+ft%29+tall%2C+about+the+same+height+as+an+81-storey+building%2C+and+the+tallest+structure+in+Paris.+Its+base+is+square%2C+measuring+125+metres+%28410+ft%29+on+each+side.+During+its+construction%2C+the+Eiffel+Tower+surpassed+the+Washington+Monument+to+become+the+tallest+man-made+structure+in+the+world%2C+a+title+it+held+for+41+years+until+the+Chrysler+Building+in+New+York+City+was+finished+in+1930.+It+was+the+first+structure+to+reach+a+height+of+300+metres.+Due+to+the+addition+of+a+broadcasting+aerial+at+the+top+of+the+tower+in+1957%2C+it+is+now+taller+than+the+Chrysler+Building+by+5.2+metres+%2817+ft%29.+Excluding+transmitters%2C+the+Eiffel+Tower+is+the+second+tallest+free-standing+structure+in+France+after+the+Millau+Viaduct)
 - [Question answering with DistilBERT](https://huggingface.co/distilbert-base-uncased-distilled-squad?text=Which+name+is+also+used+to+describe+the+Amazon+rainforest+in+English%3F&context=The+Amazon+rainforest+%28Portuguese%3A+Floresta+Amaz%C3%B4nica+or+Amaz%C3%B4nia%3B+Spanish%3A+Selva+Amaz%C3%B3nica%2C+Amazon%C3%ADa+or+usually+Amazonia%3B+French%3A+For%C3%AAt+amazonienne%3B+Dutch%3A+Amazoneregenwoud%29%2C+also+known+in+English+as+Amazonia+or+the+Amazon+Jungle%2C+is+a+moist+broadleaf+forest+that+covers+most+of+the+Amazon+basin+of+South+America.+This+basin+encompasses+7%2C000%2C000+square+kilometres+%282%2C700%2C000+sq+mi%29%2C+of+which+5%2C500%2C000+square+kilometres+%282%2C100%2C000+sq+mi%29+are+covered+by+the+rainforest.+This+region+includes+territory+belonging+to+nine+nations.+The+majority+of+the+forest+is+contained+within+Brazil%2C+with+60%25+of+the+rainforest%2C+followed+by+Peru+with+13%25%2C+Colombia+with+10%25%2C+and+with+minor+amounts+in+Venezuela%2C+Ecuador%2C+Bolivia%2C+Guyana%2C+Suriname+and+French+Guiana.+States+or+departments+in+four+nations+contain+%22Amazonas%22+in+their+names.+The+Amazon+represents+over+half+of+the+planet%27s+remaining+rainforests%2C+and+comprises+the+largest+and+most+biodiverse+tract+of+tropical+rainforest+in+the+world%2C+with+an+estimated+390+billion+individual+trees+divided+into+16%2C000+species)
 - [Translation with T5](https://huggingface.co/t5-base?text=My+name+is+Wolfgang+and+I+live+in+Berlin)
@@ -167,7 +167,7 @@ When TensorFlow 2.0 and/or PyTorch has been installed, 🤗 Transformers can be
 pip install transformers
 ```
 
-If you'd like to play with the examples, you must [install the library from source](https://huggingface.co/transformers/installation.html#installing-from-source).
+If you'd like to play with the examples or need the bleeding edge of the code and can't wait for a new release, you must [install the library from source](https://huggingface.co/transformers/installation.html#installing-from-source).
 
 ### With conda
 
@@ -179,7 +179,7 @@ Since Transformers version v4.0.0, we now have a conda channel: `huggingface`.
 conda install -c huggingface transformers
 ```
 
-Follow the installation pages of TensorFlow, PyTorch or Flax to see how to install them with conda. 
+Follow the installation pages of TensorFlow, PyTorch or Flax to see how to install them with conda.
 
 ## Models architectures
 
@@ -196,9 +196,12 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[BERT For Sequence Generation](https://huggingface.co/transformers/model_doc/bertgeneration.html)** (from Google) released with the paper [Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) by Sascha Rothe, Shashi Narayan, Aliaksei Severyn.
 1. **[Blenderbot](https://huggingface.co/transformers/model_doc/blenderbot.html)** (from Facebook) released with the paper [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
 1. **[BlenderbotSmall](https://huggingface.co/transformers/model_doc/blenderbot_small.html)** (from Facebook) released with the paper [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
+1. **[BORT](https://huggingface.co/transformers/model_doc/bort.html)** (from Alexa) released with the paper [Optimal Subarchitecture Extraction For BERT](https://arxiv.org/abs/2010.10499) by Adrian de Wynter and Daniel J. Perry.
 1. **[CamemBERT](https://huggingface.co/transformers/model_doc/camembert.html)** (from Inria/Facebook/Sorbonne) released with the paper [CamemBERT: a Tasty French Language Model](https://arxiv.org/abs/1911.03894) by Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah and Benoît Sagot.
+1. **[ConvBERT](https://huggingface.co/transformers/model_doc/convbert.html)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
 1. **[CTRL](https://huggingface.co/transformers/model_doc/ctrl.html)** (from Salesforce) released with the paper [CTRL: A Conditional Transformer Language Model for Controllable Generation](https://arxiv.org/abs/1909.05858) by Nitish Shirish Keskar*, Bryan McCann*, Lav R. Varshney, Caiming Xiong and Richard Socher.
-1. **[DeBERTa](https://huggingface.co/transformers/model_doc/deberta.html)** (from Microsoft Research) released with the paper [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://arxiv.org/abs/2006.03654) by Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen.
+1. **[DeBERTa](https://huggingface.co/transformers/model_doc/deberta.html)** (from Microsoft) released with the paper [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://arxiv.org/abs/2006.03654) by Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen.
+1. **[DeBERTa-v2](https://huggingface.co/transformers/model_doc/deberta_v2.html)** (from Microsoft) released with the paper [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://arxiv.org/abs/2006.03654) by Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen.
 1. **[DialoGPT](https://huggingface.co/transformers/model_doc/dialogpt.html)** (from Microsoft Research) released with the paper [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://arxiv.org/abs/1911.00536) by Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan.
 1. **[DistilBERT](https://huggingface.co/transformers/model_doc/distilbert.html)** (from HuggingFace), released together with the paper [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108) by Victor Sanh, Lysandre Debut and Thomas Wolf. The same method has been applied to compress GPT2 into [DistilGPT2](https://github.com/huggingface/transformers/tree/master/examples/distillation), RoBERTa into [DistilRoBERTa](https://github.com/huggingface/transformers/tree/master/examples/distillation), Multilingual BERT into [DistilmBERT](https://github.com/huggingface/transformers/tree/master/examples/distillation) and a German version of DistilBERT.
 1. **[DPR](https://huggingface.co/transformers/model_doc/dpr.html)** (from Facebook) released with the paper [Dense Passage Retrieval
@@ -209,27 +212,32 @@ Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih.
 1. **[Funnel Transformer](https://huggingface.co/transformers/model_doc/funnel.html)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[GPT](https://huggingface.co/transformers/model_doc/gpt.html)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
 1. **[GPT-2](https://huggingface.co/transformers/model_doc/gpt2.html)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/) by Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** and Ilya Sutskever**.
+1. **[I-BERT](https://huggingface.co/transformers/model_doc/ibert.html)** (from Berkeley) released with the paper [I-BERT: Integer-only BERT Quantization](https://arxiv.org/abs/2101.01321) by Sehoon Kim, Amir Gholami, Zhewei Yao, Michael W. Mahoney, Kurt Keutzer
 1. **[LayoutLM](https://huggingface.co/transformers/model_doc/layoutlm.html)** (from Microsoft Research Asia) released with 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, Ming Zhou.
 1. **[LED](https://huggingface.co/transformers/model_doc/led.html)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 1. **[Longformer](https://huggingface.co/transformers/model_doc/longformer.html)** (from AllenAI) released with the paper [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) by Iz Beltagy, Matthew E. Peters, Arman Cohan.
 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. **[M2M100](https://huggingface.co/transformers/model_doc/m2m_100.html)** (from Facebook) released with the paper [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) by by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
 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. **[MBart-50](https://huggingface.co/transformers/model_doc/mbart.html)** (from Facebook) released with the paper [Multilingual Translation with Extensible Multilingual Pretraining and Finetuning](https://arxiv.org/abs/2008.00401) by Yuqing Tang, Chau Tran, Xian Li, Peng-Jen Chen, Naman Goyal, Vishrav Chaudhary, Jiatao Gu, Angela Fan.
 1. **[MPNet](https://huggingface.co/transformers/model_doc/mpnet.html)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MT5](https://huggingface.co/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.
 1. **[RoBERTa](https://huggingface.co/transformers/model_doc/roberta.html)** (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.
+1. **[SpeechToTextTransformer](https://huggingface.co/transformers/model_doc/speech_to_text.html)** (from Facebook), released together with the paper [fairseq S2T: Fast Speech-to-Text Modeling with fairseq](https://arxiv.org/abs/2010.05171) by Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Dmytro Okhonko, Juan Pino.
 1. **[SqueezeBert](https://huggingface.co/transformers/model_doc/squeezebert.html)** 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.
 1. **[T5](https://huggingface.co/transformers/model_doc/t5.html)** (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.
 1. **[TAPAS](https://huggingface.co/transformers/model_doc/tapas.html)** (from Google AI) released with the paper [TAPAS: Weakly Supervised Table Parsing via Pre-training](https://arxiv.org/abs/2004.02349) by Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Müller, Francesco Piccinno and Julian Martin Eisenschlos.
 1. **[Transformer-XL](https://huggingface.co/transformers/model_doc/transformerxl.html)** (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.
+1. **[Wav2Vec2](https://huggingface.co/transformers/model_doc/wav2vec2.html)** (from Facebook AI) released with the paper [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://arxiv.org/abs/2006.11477) by Alexei Baevski, Henry Zhou, Abdelrahman Mohamed, Michael Auli.
 1. **[XLM](https://huggingface.co/transformers/model_doc/xlm.html)** (from Facebook) released together with the paper [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) by Guillaume Lample and Alexis Conneau.
 1. **[XLM-ProphetNet](https://huggingface.co/transformers/model_doc/xlmprophetnet.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. **[XLM-RoBERTa](https://huggingface.co/transformers/model_doc/xlmroberta.html)** (from Facebook AI), released together with the paper [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) by Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer and Veselin Stoyanov.
 1. **[XLNet](https://huggingface.co/transformers/model_doc/xlnet.html)** (from Google/CMU) released with the paper [​XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) by Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
+1. **[XLSR-Wav2Vec2](https://huggingface.co/transformers/model_doc/xlsr_wav2vec2.html)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
 1. 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 check 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)

docs/README.md

@@ -26,7 +26,7 @@ pip install -e ".[docs]"
 ---
 **NOTE**
 
-You only need to generate the documentation to inspect it locally (if you're planning changes and want to 
+You only need to generate the documentation to inspect it locally (if you're planning changes and want to
 check how they look like before committing for instance). You don't have to commit the built documentation.
 
 ---
@@ -65,7 +65,7 @@ make html
 ```
 
 A folder called ``_build/html`` should have been created. You can now open the file ``_build/html/index.html`` in your
-browser. 
+browser.
 
 ---
 **NOTE**
@@ -95,15 +95,15 @@ following these steps:
   expand them).
 - Click on "details" next to the `ci/circleci: build_doc` check.
 - In the new window, click on the "Artifacts" tab.
-- Locate the file "docs/_build/html/index.html" (or any specific page you want to check) and click on it to get a 
+- Locate the file "docs/_build/html/index.html" (or any specific page you want to check) and click on it to get a
   preview.
 
 ## Writing Documentation - Specification
 
 The `huggingface/transformers` documentation follows the
 [Google documentation](https://sphinxcontrib-napoleon.readthedocs.io/en/latest/example_google.html) style. It is
-mostly written in ReStructuredText 
-([Sphinx simple documentation](https://www.sphinx-doc.org/en/master/usage/restructuredtext/index.html), 
+mostly written in ReStructuredText
+([Sphinx simple documentation](https://www.sphinx-doc.org/en/master/usage/restructuredtext/index.html),
 [Sourceforge complete documentation](https://docutils.sourceforge.io/docs/ref/rst/restructuredtext.html)).
 
 
@@ -121,8 +121,8 @@ four.
 ### Adding a new model
 
 When adding a new model:
- 
-- Create a file `xxx.rst` under `./source/model_doc` (don't hesitate to copy an existing file as template). 
+
+- Create a file `xxx.rst` under `./source/model_doc` (don't hesitate to copy an existing file as template).
 - Link that file in `./source/index.rst` on the `model_doc` toc-tree.
 - Write a short overview of the model:
     - Overview with paper & authors
@@ -130,8 +130,8 @@ When adding a new model:
     - Tips and tricks and how to use it best
 - Add the classes that should be linked in the model. This generally includes the configuration, the tokenizer, and
   every model of that class (the base model, alongside models with additional heads), both in PyTorch and TensorFlow.
-  The order is generally: 
-    - Configuration, 
+  The order is generally:
+    - Configuration,
     - Tokenizer
     - PyTorch base model
     - PyTorch head models
@@ -179,7 +179,7 @@ Links should be done as so (note the double underscore at the end): \`text for t
 
 #### Defining arguments in a method
 
-Arguments should be defined with the `Args:` prefix, followed by a line return and an indentation. 
+Arguments should be defined with the `Args:` prefix, followed by a line return and an indentation.
 The argument should be followed by its type, with its shape if it is a tensor, and a line return.
 Another indentation is necessary before writing the description of the argument.
 
@@ -216,9 +216,9 @@ then its documentation should look like this:
 
 Note that we always omit the "defaults to :obj:\`None\`" when None is the default for any argument. Also note that even
 if the first line describing your argument type and its default gets long, you can't break it on several lines. You can
-however write as many lines as you want in the indented description (see the example above with `input_ids`). 
+however write as many lines as you want in the indented description (see the example above with `input_ids`).
 
-#### Writing a multi-line code block 
+#### Writing a multi-line code block
 
 Multi-line code blocks can be useful for displaying examples. They are done like so:
 
@@ -237,7 +237,7 @@ the results stay consistent with the library.
 
 #### Writing a return block
 
-Arguments should be defined with the `Args:` prefix, followed by a line return and an indentation. 
+Arguments should be defined with the `Args:` prefix, followed by a line return and an indentation.
 The first line should be the type of the return, followed by a line return. No need to indent further for the elements
 building the return.
 
@@ -258,3 +258,43 @@ Here's an example for a single value return:
     Returns:
         :obj:`List[int]`: A list of integers in the range [0, 1] --- 1 for a special token, 0 for a sequence token.
 ```
+
+#### Adding a new section
+
+In ReST section headers are designated as such with the help of a line of underlying characters, e.g.,:
+
+```
+Section 1
+^^^^^^^^^^^^^^^^^^
+
+Sub-section 1
+~~~~~~~~~~~~~~~~~~
+```
+
+ReST allows the use of any characters to designate different section levels, as long as they are used consistently within the same document. For details see [sections doc](https://www.sphinx-doc.org/en/master/usage/restructuredtext/basics.html#sections). Because there is no standard different documents often end up using different characters for the same levels which makes it very difficult to know which character to use when creating a new section.
+
+Specifically, if when running `make docs` you get an error like:
+```
+docs/source/main_classes/trainer.rst:127:Title level inconsistent:
+```
+you picked an inconsistent character for some of the levels.
+
+But how do you know which characters you must use for an already existing level or when adding a new level?
+
+You can use this helper script:
+```
+perl -ne '/^(.)\1{100,}/ && do { $h{$1}=++$c if !$h{$1} }; END { %h = reverse %h ; print "$_ $h{$_}\n" for sort keys %h}' docs/source/main_classes/trainer.rst
+1 -
+2 ~
+3 ^
+4 =
+5 "
+```
+
+This tells you which characters have already been assigned for each level.
+
+So using this particular example's output -- if your current section's header uses `=` as its underline character, you now know you're at level 4, and if you want to add a sub-section header you know you want `"` as it'd level 5.
+
+If you needed to add yet another sub-level, then pick a character that is not used already. That is you must pick a character that is not in the output of that script.
+
+Here is the full list of characters that can be used in this context: `= - ` : ' " ~ ^ _ * + # < >`

docs/source/_static/js/custom.js

@@ -1,10 +1,12 @@
 // These two things need to be updated at each release for the version selector.
 // Last stable version
-const stableVersion = "v4.1.1"
+const stableVersion = "v4.3.2"
 // Dictionary doc folder to label. The last stable version should have an empty key.
 const versionMapping = {
     "master": "master",
-    "": "v4.1.1 (stable)",
+    "": "v4.3.0/v4.3.1/v4.3.2/v4.3.3 (stable)",
+    "v4.2.2": "v4.2.0/v4.2.1/v4.2.2",
+    "v4.1.1": "v4.1.0/v4.1.1",
     "v4.0.1": "v4.0.0/v4.0.1",
     "v3.5.1": "v3.5.0/v3.5.1",
     "v3.4.0": "v3.4.0",
@@ -126,11 +128,11 @@ function addVersionControl() {
     const parts = location.toString().split('/');
     let versionIndex = parts.length - 2;
     // Index page may not have a last part with filename.html so we need to go up
-    if (parts[parts.length - 1] != "" && ! parts[parts.length - 1].match(/\.html$|^search.html?/)) {
+    if (parts[parts.length - 1] != "" && ! parts[parts.length - 1].match(/\.html/)) {
         versionIndex = parts.length - 1;
     }
     // Main classes and models are nested so we need to go deeper
-    else if (parts[versionIndex] == "main_classes" || parts[versionIndex] == "model_doc") {
+    else if (parts[versionIndex] == "main_classes" || parts[versionIndex] == "model_doc" || parts[versionIndex] == "internal") {
         versionIndex = versionIndex - 1;
     } 
     const version = parts[versionIndex];

docs/source/add_new_model.rst

@@ -0,0 +1,844 @@
+.. 
+    Copyright 2020 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+
+How to add a model to 🤗 Transformers?
+=======================================================================================================================
+
+Adding a new model is often difficult and requires an in-depth knowledge of the 🤗 Transformers library and ideally also
+of the model's original repository. At Hugging Face, we are trying to empower the community more and more to add models
+independently. Thus, for some new models that the community wants to be added to 🤗 Transformers, we create a customized
+*call-for-model-addition* that explains step-by-step how to add the requested model. With this
+*call-for-model-addition*, we want to teach a motivated and experienced contributor of the community how to port a
+model to 🤗 Transformers.
+
+If this sounds like something you would be interested in, feel free to check out the currently open
+“calls-for-model-addition” `here
+<https://github.com/huggingface/transformers/tree/master/templates/adding_a_new_model/open_model_proposals/README.md>`__
+and to contact us.
+
+If selected, you will then work closely with one member of the Hugging Face team to integrate the model into 🤗
+Transformers. By doing so, you will both gain a theoretical and deep practical understanding of the proposed model. But
+more importantly, you will have made a major open-source contribution to 🤗 Transformers. Along the way, you will:
+
+-  get insights into open-source best practices
+-  understand the design principles of one of the most popular NLP libraries
+-  learn how to do efficiently test large NLP models
+-  learn how to integrate Python utilities like ``black``, ``isort``, ``make fix-copies`` into a library to always
+   ensure clean and readable code
+
+We are also more than happy if you want to add a model that cannot be found in the “calls-for-model-addition” folder.
+The following sections explain in detail how to add a new model. It might also be very helpful to check out already
+added models to see if those resemble the model you would like to add `here
+<https://github.com/huggingface/transformers/pulls?q=is%3Apr+label%3A%22PR+for+Model+Addition%22+is%3Aclosed>`__.
+
+To start, let's try to get a general overview of the Transformers library.
+
+General overview of 🤗 Transformers
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+First, you should get a general overview of 🤗 Transformers. 🤗 Transformers is a very opinionated library, so there is a
+chance that you don't agree with some of the library's philosophies or design choices. From our experience, however, we
+found that the fundamental design choices and philosophies of the library are crucial to efficiently scale 🤗
+Transformers while keeping maintenance costs at a reasonable level.
+
+A good first starting point to better understand the library is to read the :doc:`documentation of our philosophy
+<philosophy>`. As a result of our way of working, there are some choices that we try to apply to all models:
+
+-  Composition is generally favored over-abstraction
+-  Duplicating code is not always bad if it strongly improves the readability or accessibility of a model
+-  Model files are as self-contained as possible so that when you read the code of a specific model, you ideally only
+   have to look into the respective ``modeling_....py`` file.
+
+In our opinion, the library's code is not just a means to provide a product, *e.g.* the ability to use BERT for
+inference, but also as the very product that we want to improve. Hence, when adding a model, the user is not only the
+person that will use your model, but also everybody that will read, try to understand, and possibly tweak your code.
+
+With this in mind, let's go a bit deeper into the general library design.
+
+Overview of models
+-----------------------------------------------------------------------------------------------------------------------
+
+To successfully add a model, it is important to understand the interaction between your model and its config,
+:class:`~transformers.PreTrainedModel`, and :class:`~transformers.PretrainedConfig`. For exemplary purposes, we will
+call the model to be added to 🤗 Transformers ``BrandNewBert``.
+
+Let's take a look:
+
+.. image:: ./imgs/transformers_overview.png
+
+As you can see, we do make use of inheritance in 🤗 Transformers, but we keep the level of abstraction to an absolute
+minimum. There are never more than two levels of abstraction for any model in the library. :obj:`BrandNewBertModel`
+inherits from :obj:`BrandNewBertPreTrainedModel` which in turn inherits from :class:`~transformres.PreTrainedModel` and
+that's it. As a general rule, we want to make sure that a new model only depends on
+:class:`~transformers.PreTrainedModel`. The important functionalities that are automatically provided to every new
+model are :meth:`~transformers.PreTrainedModel.from_pretrained` and
+:meth:`~transformers.PreTrainedModel.save_pretrained`, which are used for serialization and deserialization. All of the
+other important functionalities, such as :meth:`BrandNewBertModel.forward` should be completely defined in the new
+``modeling_brand_new_bert.py`` script. Next, we want to make sure that a model with a specific head layer, such as
+:obj:`BrandNewBertForMaskedLM` does not inherit from :obj:`BrandNewBertModel`, but rather uses :obj:`BrandNewBertModel`
+as a component that can be called in its forward pass to keep the level of abstraction low. Every new model requires a
+configuration class, called :obj:`BrandNewBertConfig`. This configuration is always stored as an attribute in
+:class:`~transformers.PreTrainedModel`, and thus can be accessed via the ``config`` attribute for all classes
+inheriting from :obj:`BrandNewBertPreTrainedModel`:
+
+   .. code:: python
+
+      model = BrandNewBertModel.from_pretrained("brandy/brand_new_bert")
+      model.config  # model has access to its config
+
+Similar to the model, the configuration inherits basic serialization and deserialization functionalities from
+:class:`~transformers.PretrainedConfig`. Note that the configuration and the model are always serialized into two
+different formats - the model to a `pytorch_model.bin` file and the configuration to a `config.json` file. Calling
+:meth:`~transformers.PreTrainedModel.save_pretrained` will automatically call
+:meth:`~transformers.PretrainedConfig.save_pretrained`, so that both model and configuration are saved.
+
+
+Overview of tokenizers
+-----------------------------------------------------------------------------------------------------------------------
+
+Not quite ready yet :-( This section will be added soon!
+
+Step-by-step recipe to add a model to 🤗 Transformers
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Everyone has different preferences of how to port a model so it can be very helpful for you to take a look at summaries
+of how other contributors ported models to Hugging Face. Here is a list of community blog posts on how to port a model:
+
+1. `Porting GPT2 Model <https://medium.com/huggingface/from-tensorflow-to-pytorch-265f40ef2a28>`__ by `Thomas
+   <https://huggingface.co/thomwolf>`__
+2. `Porting WMT19 MT Model <https://huggingface.co/blog/porting-fsmt>`__ by `Stas <https://huggingface.co/stas>`__
+
+From experience, we can tell you that the most important things to keep in mind when adding a model are:
+
+-  Don't reinvent the wheel! Most parts of the code you will add for the new 🤗 Transformers model already exist
+   somewhere in 🤗 Transformers. Take some time to find similar, already existing models and tokenizers you can copy
+   from. `grep <https://www.gnu.org/software/grep/>`__ and `rg <https://github.com/BurntSushi/ripgrep>`__ are your
+   friends. Note that it might very well happen that your model's tokenizer is based on one model implementation, and
+   your model's modeling code on another one. *E.g.* FSMT's modeling code is based on BART, while FSMT's tokenizer code
+   is based on XLM.
+-  It's more of an engineering challenge than a scientific challenge. You should spend more time on creating an
+   efficient debugging environment than trying to understand all theoretical aspects of the model in the paper.
+-  Ask for help, when you're stuck! Models are the core component of 🤗 Transformers so that we at Hugging Face are more
+   than happy to help you at every step to add your model. Don't hesitate to ask if you notice you are not making
+   progress.
+
+In the following, we try to give you a general recipe that we found most useful when porting a model to 🤗 Transformers.
+
+The following list is a summary of everything that has to be done to add a model and can be used by you as a To-Do
+List:
+
+-  1. ☐ (Optional) Understood theoretical aspects
+-  2. ☐ Prepared transformers dev environment
+-  3. ☐ Set up debugging environment of the original repository
+-  4. ☐ Created script that successfully runs forward pass using original repository and checkpoint
+-  5. ☐ Successfully added the model skeleton to Transformers
+-  6. ☐ Successfully converted original checkpoint to Transformers checkpoint
+-  7. ☐ Successfully ran forward pass in Transformers that gives identical output to original checkpoint
+-  8. ☐ Finished model tests in Transformers
+-  9. ☐ Successfully added Tokenizer in Transformers
+-  10. ☐ Run end-to-end integration tests
+-  11. ☐ Finished docs
+-  12. ☐ Uploaded model weights to the hub
+-  13. ☐ Submitted the pull request
+-  14. ☐ (Optional) Added a demo notebook
+
+To begin with, we usually recommend to start by getting a good theoretical understanding of ``BrandNewBert``. However,
+if you prefer to understand the theoretical aspects of the model *on-the-job*, then it is totally fine to directly dive
+into the ``BrandNewBert``'s code-base. This option might suit you better, if your engineering skills are better than
+your theoretical skill, if you have trouble understanding ``BrandNewBert``'s paper, or if you just enjoy programming
+much more than reading scientific papers.
+
+1. (Optional) Theoretical aspects of BrandNewBert
+-----------------------------------------------------------------------------------------------------------------------
+
+You should take some time to read *BrandNewBert's* paper, if such descriptive work exists. There might be large
+sections of the paper that are difficult to understand. If this is the case, this is fine - don't worry! The goal is
+not to get a deep theoretical understanding of the paper, but to extract the necessary information required to
+effectively re-implement the model in 🤗 Transformers. That being said, you don't have to spend too much time on the
+theoretical aspects, but rather focus on the practical ones, namely:
+
+-  What type of model is *brand_new_bert*? BERT-like encoder-only model? GPT2-like decoder-only model? BART-like
+   encoder-decoder model? Look at the :doc:`model_summary` if you're not familiar with the differences between those.
+-  What are the applications of *brand_new_bert*? Text classification? Text generation? Seq2Seq tasks, *e.g.,*
+   summarization?
+-  What is the novel feature of the model making it different from BERT/GPT-2/BART?
+-  Which of the already existing `🤗 Transformers models <https://huggingface.co/transformers/#contents>`__ is most
+   similar to *brand_new_bert*?
+-  What type of tokenizer is used? A sentencepiece tokenizer? Word piece tokenizer? Is it the same tokenizer as used
+   for BERT or BART?
+
+After you feel like you have gotten a good overview of the architecture of the model, you might want to write to the
+Hugging Face team with any questions you might have. This might include questions regarding the model's architecture,
+its attention layer, etc. We will be more than happy to help you.
+
+2. Next prepare your environment
+-----------------------------------------------------------------------------------------------------------------------
+
+1. Fork the `repository <https://github.com/huggingface/transformers>`__ by clicking on the ‘Fork' button on the
+   repository's page. This creates a copy of the code under your GitHub user account.
+
+2. Clone your ``transformers`` fork to your local disk, and add the base repository as a remote:
+
+   .. code:: bash
+
+      git clone https://github.com/[your Github handle]/transformers.git
+      cd transformers
+      git remote add upstream https://github.com/huggingface/transformers.git
+
+3. Set up a development environment, for instance by running the following command:
+
+   .. code:: bash
+
+      python -m venv .env
+      source .env/bin/activate
+      pip install -e ".[dev]"
+
+and return to the parent directory
+
+.. code:: bash
+
+   cd ..
+
+4. We recommend adding the PyTorch version of *brand_new_bert* to Transformers. To install PyTorch, please follow the
+   instructions on https://pytorch.org/get-started/locally/.
+
+**Note:** You don't need to have CUDA installed. Making the new model work on CPU is sufficient.
+
+5. To port *brand_new_bert*, you will also need access to its original repository:
+
+.. code:: bash
+
+   git clone https://github.com/org_that_created_brand_new_bert_org/brand_new_bert.git 
+   cd brand_new_bert
+   pip install -e .
+
+Now you have set up a development environment to port *brand_new_bert* to 🤗 Transformers.
+
+3.-4. Run a pretrained checkpoint using the original repository
+-----------------------------------------------------------------------------------------------------------------------
+
+At first, you will work on the original *brand_new_bert* repository. Often, the original implementation is very
+“researchy”. Meaning that documentation might be lacking and the code can be difficult to understand. But this should
+be exactly your motivation to reimplement *brand_new_bert*. At Hugging Face, one of our main goals is to *make people
+stand on the shoulders of giants* which translates here very well into taking a working model and rewriting it to make
+it as **accessible, user-friendly, and beautiful** as possible. This is the number-one motivation to re-implement
+models into 🤗 Transformers - trying to make complex new NLP technology accessible to **everybody**.
+
+You should start thereby by diving into the original repository.
+
+Successfully running the official pretrained model in the original repository is often **the most difficult** step.
+From our experience, it is very important to spend some time getting familiar with the original code-base. You need to
+figure out the following:
+
+-  Where to find the pretrained weights?
+-  How to load the pretrained weights into the corresponding model?
+-  How to run the tokenizer independently from the model?
+-  Trace one forward pass so that you know which classes and functions are required for a simple forward pass. Usually,
+   you only have to reimplement those functions.
+-  Be able to locate the important components of the model: Where is the model's class? Are there model sub-classes,
+   *e.g.* EncoderModel, DecoderModel? Where is the self-attention layer? Are there multiple different attention layers,
+   *e.g.* *self-attention*, *cross-attention*...?
+-  How can you debug the model in the original environment of the repo? Do you have to add `print` statements, can you
+   work with an interactive debugger like `ipdb`, or should you use an efficient IDE to debug the model, like PyCharm?
+
+It is very important that before you start the porting process, that you can **efficiently** debug code in the original
+repository! Also, remember that you are working with an open-source library, so do not hesitate to open an issue, or
+even a pull request in the original repository. The maintainers of this repository are most likely very happy about
+someone looking into their code!
+
+At this point, it is really up to you which debugging environment and strategy you prefer to use to debug the original
+model. We strongly advise against setting up a costly GPU environment, but simply work on a CPU both when starting to
+dive into the original repository and also when starting to write the 🤗 Transformers implementation of the model. Only
+at the very end, when the model has already been successfully ported to 🤗 Transformers, one should verify that the
+model also works as expected on GPU.
+
+In general, there are two possible debugging environments for running the original model
+
+-  `Jupyter notebooks <https://jupyter.org/>`__ / `google colab
+   <https://colab.research.google.com/notebooks/intro.ipynb>`__
+-  Local python scripts.
+
+Jupyter notebooks have the advantage that they allow for cell-by-cell execution which can be helpful to better split
+logical components from one another and to have faster debugging cycles as intermediate results can be stored. Also,
+notebooks are often easier to share with other contributors, which might be very helpful if you want to ask the Hugging
+Face team for help. If you are familiar with Jupiter notebooks, we strongly recommend you to work with them.
+
+The obvious disadvantage of Jupyther notebooks is that if you are not used to working with them you will have to spend
+some time adjusting to the new programming environment and that you might not be able to use your known debugging tools
+anymore, like ``ipdb``.
+
+For each code-base, a good first step is always to load a **small** pretrained checkpoint and to be able to reproduce a
+single forward pass using a dummy integer vector of input IDs as an input. Such a script could look like this (in
+pseudocode):
+
+.. code:: bash
+
+   model = BrandNewBertModel.load_pretrained_checkpoint(/path/to/checkpoint/)
+   input_ids = [0, 4, 5, 2, 3, 7, 9]  # vector of input ids
+   original_output = model.predict(input_ids)
+
+Next, regarding the debugging strategy, there are generally a few from which to choose from:
+
+-  Decompose the original model into many small testable components and run a forward pass on each of those for
+   verification
+-  Decompose the original model only into the original *tokenizer* and the original *model*, run a forward pass on
+   those, and use intermediate print statements or breakpoints for verification
+
+Again, it is up to you which strategy to choose. Often, one or the other is advantageous depending on the original code
+base.
+
+If the original code-base allows you to decompose the model into smaller sub-components, *e.g.* if the original
+code-base can easily be run in eager mode, it is usually worth the effort to do so. There are some important advantages
+to taking the more difficult road in the beginning:
+
+- at a later stage when comparing the original model to the Hugging Face implementation, you can verify automatically
+  for each component individually that the corresponding component of the 🤗 Transformers implementation matches instead
+  of relying on visual comparison via print statements
+- it can give you some rope to decompose the big problem of porting a model into smaller problems of just porting
+  individual components and thus structure your work better
+- separating the model into logical meaningful components will help you to get a better overview of the model's design
+  and thus to better understand the model
+- at a later stage those component-by-component tests help you to ensure that no regression occurs as you continue
+  changing your code
+
+`Lysandre's <https://gist.github.com/LysandreJik/db4c948f6b4483960de5cbac598ad4ed>`__ integration checks for ELECTRA
+gives a nice example of how this can be done.
+
+However, if the original code-base is very complex or only allows intermediate components to be run in a compiled mode,
+it might be too time-consuming or even impossible to separate the model into smaller testable sub-components. A good
+example is `T5's MeshTensorFlow <https://github.com/tensorflow/mesh/tree/master/mesh_tensorflow>`__ library which is
+very complex and does not offer a simple way to decompose the model into its sub-components. For such libraries, one
+often relies on verifying print statements.
+
+No matter which strategy you choose, the recommended procedure is often the same in that you should start to debug the
+starting layers first and the ending layers last.
+
+It is recommended that you retrieve the output, either by print statements or sub-component functions, of the following
+layers in the following order:
+
+1.  Retrieve the input IDs passed to the model
+2.  Retrieve the word embeddings
+3.  Retrieve the input of the first Transformer layer
+4.  Retrieve the output of the first Transformer layer
+5.  Retrieve the output of the following n - 1 Transformer layers
+6.  Retrieve the output of the whole BrandNewBert Model
+
+Input IDs should thereby consists of an array of integers, *e.g.* ``input_ids = [0, 4, 4, 3, 2, 4, 1, 7, 19]``
+
+The outputs of the following layers often consist of multi-dimensional float arrays and can look like this:
+
+.. code:: bash
+
+   [[
+    [-0.1465, -0.6501,  0.1993,  ...,  0.1451,  0.3430,  0.6024],
+    [-0.4417, -0.5920,  0.3450,  ..., -0.3062,  0.6182,  0.7132],
+    [-0.5009, -0.7122,  0.4548,  ..., -0.3662,  0.6091,  0.7648],
+    ...,
+    [-0.5613, -0.6332,  0.4324,  ..., -0.3792,  0.7372,  0.9288],
+    [-0.5416, -0.6345,  0.4180,  ..., -0.3564,  0.6992,  0.9191],
+    [-0.5334, -0.6403,  0.4271,  ..., -0.3339,  0.6533,  0.8694]]],
+
+We expect that every model added to 🤗 Transformers passes a couple of integration tests, meaning that the original
+model and the reimplemented version in 🤗 Transformers have to give the exact same output up to a precision of 0.001!
+Since it is normal that the exact same model written in different libraries can give a slightly different output
+depending on the library framework, we accept an error tolerance of 1e-3 (0.001). It is not enough if the model gives
+nearly the same output, they have to be the almost identical. Therefore, you will certainly compare the intermediate
+outputs of the 🤗 Transformers version multiple times against the intermediate outputs of the original implementation of
+*brand_new_bert* in which case an **efficient** debugging environment of the original repository is absolutely
+important. Here is some advice is to make your debugging environment as efficient as possible.
+
+-  Find the best way of debugging intermediate results. Is the original repository written in PyTorch? Then you should
+   probably take the time to write a longer script that decomposes the original model into smaller sub-components to
+   retrieve intermediate values. Is the original repository written in Tensorflow 1? Then you might have to rely on
+   TensorFlow print operations like `tf.print <https://www.tensorflow.org/api_docs/python/tf/print>`__ to output
+   intermediate values. Is the original repository written in Jax? Then make sure that the model is **not jitted** when
+   running the forward pass, *e.g.* check-out `this link <https://github.com/google/jax/issues/196>`__.
+-  Use the smallest pretrained checkpoint you can find. The smaller the checkpoint, the faster your debug cycle
+   becomes. It is not efficient if your pretrained model is so big that your forward pass takes more than 10 seconds.
+   In case only very large checkpoints are available, it might make more sense to create a dummy model in the new
+   environment with randomly initialized weights and save those weights for comparison with the 🤗 Transformers version
+   of your model
+-  Make sure you are using the easiest way of calling a forward pass in the original repository. Ideally, you want to
+   find the function in the original repository that **only** calls a single forward pass, *i.e.* that is often called
+   ``predict``, ``evaluate``, ``forward`` or ``__call__``. You don't want to debug a function that calls ``forward``
+   multiple times, *e.g.* to generate text, like ``autoregressive_sample``, ``generate``.
+-  Try to separate the tokenization from the model's `forward` pass. If the original repository shows examples where
+   you have to input a string, then try to find out where in the forward call the string input is changed to input ids
+   and start from this point. This might mean that you have to possibly write a small script yourself or change the
+   original code so that you can directly input the ids instead of an input string.
+-  Make sure that the model in your debugging setup is **not** in training mode, which often causes the model to yield
+   random outputs due to multiple dropout layers in the model. Make sure that the forward pass in your debugging
+   environment is **deterministic** so that the dropout layers are not used. Or use `transformers.file_utils.set_seed`
+   if the old and new implementations are in the same framework.
+
+The following section gives you more specific details/tips on how you can do this for *brand_new_bert*.
+
+5.-14. Port BrandNewBert to 🤗 Transformers
+-----------------------------------------------------------------------------------------------------------------------
+
+Next, you can finally start adding new code to 🤗 Transformers. Go into the clone of your 🤗 Transformers' fork:
+
+::
+
+   cd transformers
+
+In the special case that you are adding a model whose architecture exactly matches the model architecture of an
+existing model you only have to add a conversion script as described in `this section <#write-a-conversion-script>`__.
+In this case, you can just re-use the whole model architecture of the already existing model.
+
+Otherwise, let's start generating a new model with the amazing Cookiecutter!
+
+**Use the Cookiecutter to automatically generate the model's code**
+
+To begin with head over to the `🤗 Transformers templates
+<https://github.com/huggingface/transformers/tree/master/templates/adding_a_new_model>`__ to make use of our
+``cookiecutter`` implementation to automatically generate all the relevant files for your model. Again, we recommend
+only adding the PyTorch version of the model at first. Make sure you follow the instructions of the ``README.md`` on
+the `🤗 Transformers templates <https://github.com/huggingface/transformers/tree/master/templates/adding_a_new_model>`__
+carefully.
+
+**Open a Pull Request on the main huggingface/transformers repo**
+
+Before starting to adapt the automatically generated code, now is the time to open a “Work in progress (WIP)” pull
+request, *e.g.* “[WIP] Add *brand_new_bert*”, in 🤗 Transformers so that you and the Hugging Face team can work
+side-by-side on integrating the model into 🤗 Transformers.
+
+You should do the following:
+
+1. Create a branch with a descriptive name from your master branch
+
+::
+
+   git checkout -b add_brand_new_bert
+
+2. Commit the automatically generated code:
+
+::
+
+   git add .
+   git commit
+
+3. Fetch and rebase to current master
+
+::
+
+   git fetch upstream
+   git rebase upstream/master
+
+4. Push the changes to your account using:
+
+::
+
+   git push -u origin a-descriptive-name-for-my-changes
+
+5. Once you are satisfied, go to the webpage of your fork on GitHub. Click on “Pull request”. Make sure to add the
+   GitHub handle of some members of the Hugging Face team as reviewers, so that the Hugging Face team gets notified for
+   future changes.
+
+6. Change the PR into a draft by clicking on “Convert to draft” on the right of the GitHub pull request web page.
+
+In the following, whenever you have done some progress, don't forget to commit your work and push it to your account so
+that it shows in the pull request. Additionally, you should make sure to update your work with the current master from
+time to time by doing:
+
+::
+
+   git fetch upstream
+   git merge upstream/master
+
+In general, all questions you might have regarding the model or your implementation should be asked in your PR and
+discussed/solved in the PR. This way, the Hugging Face team will always be notified when you are committing new code or
+if you have a question. It is often very helpful to point the Hugging Face team to your added code so that the Hugging
+Face team can efficiently understand your problem or question.
+
+To do so, you can go to the “Files changed” tab where you see all of your changes, go to a line regarding which you
+want to ask a question, and click on the “+” symbol to add a comment. Whenever a question or problem has been solved,
+you can click on the “Resolve” button of the created comment.
+
+In the same way, the Hugging Face team will open comments when reviewing your code. We recommend asking most questions
+on GitHub on your PR. For some very general questions that are not very useful for the public, feel free to ping the
+Hugging Face team by Slack or email.
+
+**5. Adapt the generated models code for brand_new_bert**
+
+At first, we will focus only on the model itself and not care about the tokenizer. All the relevant code should be
+found in the generated files ``src/transformers/models/brand_new_bert/modeling_brand_new_bert.py`` and
+``src/transformers/models/brand_new_bert/configuration_brand_new_bert.py``.
+
+Now you can finally start coding :). The generated code in
+``src/transformers/models/brand_new_bert/modeling_brand_new_bert.py`` will either have the same architecture as BERT if
+it's an encoder-only model or BART if it's an encoder-decoder model. At this point, you should remind yourself what
+you've learned in the beginning about the theoretical aspects of the model: *How is the model different from BERT or
+BART?*". Implement those changes which often means to change the *self-attention* layer, the order of the normalization
+layer, etc… Again, it is often useful to look at the similar architecture of already existing models in Transformers to
+get a better feeling of how your model should be implemented.
+
+**Note** that at this point, you don't have to be very sure that your code is fully correct or clean. Rather, it is
+advised to add a first *unclean*, copy-pasted version of the original code to
+``src/transformers/models/brand_new_bert/modeling_brand_new_bert.py`` until you feel like all the necessary code is
+added. From our experience, it is much more efficient to quickly add a first version of the required code and
+improve/correct the code iteratively with the conversion script as described in the next section. The only thing that
+has to work at this point is that you can instantiate the 🤗 Transformers implementation of *brand_new_bert*, *i.e.* the
+following command should work:
+
+.. code:: python
+
+   from transformers import BrandNewBertModel, BrandNewBertConfig
+   model = BrandNewBertModel(BrandNewBertConfig())
+
+The above command will create a model according to the default parameters as defined in ``BrandNewBertConfig()`` with
+random weights, thus making sure that the ``init()`` methods of all components works.
+
+**6. Write a conversion script**
+
+Next, you should write a conversion script that lets you convert the checkpoint you used to debug *brand_new_bert* in
+the original repository to a checkpoint compatible with your just created 🤗 Transformers implementation of
+*brand_new_bert*. It is not advised to write the conversion script from scratch, but rather to look through already
+existing conversion scripts in 🤗 Transformers for one that has been used to convert a similar model that was written in
+the same framework as *brand_new_bert*. Usually, it is enough to copy an already existing conversion script and
+slightly adapt it for your use case. Don't hesitate to ask the Hugging Face team to point you to a similar already
+existing conversion script for your model.
+
+-  If you are porting a model from TensorFlow to PyTorch, a good starting point might be BERT's conversion script `here
+   <https://github.com/huggingface/transformers/blob/7acfa95afb8194f8f9c1f4d2c6028224dbed35a2/src/transformers/models/bert/modeling_bert.py#L91>`__
+-  If you are porting a model from PyTorch to PyTorch, a good starting point might be BART's conversion script `here
+   <https://github.com/huggingface/transformers/blob/master/src/transformers/models/bart/convert_bart_original_pytorch_checkpoint_to_pytorch.py>`__
+
+In the following, we'll quickly explain how PyTorch models store layer weights and define layer names. In PyTorch, the
+name of a layer is defined by the name of the class attribute you give the layer. Let's define a dummy model in
+PyTorch, called ``SimpleModel`` as follows:
+
+.. code:: python
+
+   import torch.nn as nn
+
+   class SimpleModel(nn.Module):
+       def __init__(self):
+               super().__init__()
+               self.dense = nn.Linear(10, 10)
+               self.intermediate = nn.Linear(10, 10)
+               self.layer_norm = nn.LayerNorm(10)
+
+Now we can create an instance of this model definition which will fill all weights: ``dense``, ``intermediate``,
+``layer_norm`` with random weights. We can print the model to see its architecture
+
+.. code:: python
+
+   model = SimpleModel()
+
+   print(model)
+
+This will print out the following:
+
+.. code:: bash
+
+   SimpleModel(
+     (dense): Linear(in_features=10, out_features=10, bias=True)
+     (intermediate): Linear(in_features=10, out_features=10, bias=True)
+     (layer_norm): LayerNorm((10,), eps=1e-05, elementwise_affine=True)
+   )
+
+We can see that the layer names are defined by the name of the class attribute in PyTorch. You can print out the weight
+values of a specific layer:
+
+.. code:: python
+
+   print(model.dense.weight.data)
+
+to see that the weights were randomly initialized
+
+.. code:: bash
+
+   tensor([[-0.0818,  0.2207, -0.0749, -0.0030,  0.0045, -0.1569, -0.1598,  0.0212,
+            -0.2077,  0.2157],
+           [ 0.1044,  0.0201,  0.0990,  0.2482,  0.3116,  0.2509,  0.2866, -0.2190,
+             0.2166, -0.0212],
+           [-0.2000,  0.1107, -0.1999, -0.3119,  0.1559,  0.0993,  0.1776, -0.1950,
+            -0.1023, -0.0447],
+           [-0.0888, -0.1092,  0.2281,  0.0336,  0.1817, -0.0115,  0.2096,  0.1415,
+            -0.1876, -0.2467],
+           [ 0.2208, -0.2352, -0.1426, -0.2636, -0.2889, -0.2061, -0.2849, -0.0465,
+             0.2577,  0.0402],
+           [ 0.1502,  0.2465,  0.2566,  0.0693,  0.2352, -0.0530,  0.1859, -0.0604,
+             0.2132,  0.1680],
+           [ 0.1733, -0.2407, -0.1721,  0.1484,  0.0358, -0.0633, -0.0721, -0.0090,
+             0.2707, -0.2509],
+           [-0.1173,  0.1561,  0.2945,  0.0595, -0.1996,  0.2988, -0.0802,  0.0407,
+             0.1829, -0.1568],
+           [-0.1164, -0.2228, -0.0403,  0.0428,  0.1339,  0.0047,  0.1967,  0.2923,
+             0.0333, -0.0536],
+           [-0.1492, -0.1616,  0.1057,  0.1950, -0.2807, -0.2710, -0.1586,  0.0739,
+             0.2220,  0.2358]]).
+
+In the conversion script, you should fill those randomly initialized weights with the exact weights of the
+corresponding layer in the checkpoint. *E.g.*
+
+.. code:: python
+
+   # retrieve matching layer weights, e.g. by 
+   # recursive algorithm
+   layer_name = "dense"
+   pretrained_weight = array_of_dense_layer
+
+   model_pointer = getattr(model, "dense")
+
+   model_pointer.weight.data = torch.from_numpy(pretrained_weight)
+
+While doing so, you must verify that each randomly initialized weight of your PyTorch model and its corresponding
+pretrained checkpoint weight exactly match in both **shape and name**. To do so, it is **necessary** to add assert
+statements for the shape and print out the names of the checkpoints weights. E.g. you should add statements like:
+
+.. code:: python
+
+   assert (
+        model_pointer.weight.shape == pretrained_weight.shape
+   ), f"Pointer shape of random weight {model_pointer.shape} and array shape of checkpoint weight {pretrained_weight.shape} mismatched"
+
+Besides, you should also print out the names of both weights to make sure they match, *e.g.*
+
+.. code:: python
+
+   logger.info(f"Initialize PyTorch weight {layer_name} from {pretrained_weight.name}")
+
+If either the shape or the name doesn't match, you probably assigned the wrong checkpoint weight to a randomly
+initialized layer of the 🤗 Transformers implementation.
+
+An incorrect shape is most likely due to an incorrect setting of the config parameters in ``BrandNewBertConfig()`` that
+do not exactly match those that were used for the checkpoint you want to convert. However, it could also be that
+PyTorch's implementation of a layer requires the weight to be transposed beforehand.
+
+Finally, you should also check that **all** required weights are initialized and print out all checkpoint weights that
+were not used for initialization to make sure the model is correctly converted. It is completely normal, that the
+conversion trials fail with either a wrong shape statement or wrong name assignment. This is most likely because either
+you used incorrect parameters in ``BrandNewBertConfig()``, have a wrong architecture in the 🤗 Transformers
+implementation, you have a bug in the ``init()`` functions of one of the components of the 🤗 Transformers
+implementation or you need to transpose one of the checkpoint weights.
+
+This step should be iterated with the previous step until all weights of the checkpoint are correctly loaded in the
+Transformers model. Having correctly loaded the checkpoint into the 🤗 Transformers implementation, you can then save
+the model under a folder of your choice ``/path/to/converted/checkpoint/folder`` that should then contain both a
+``pytorch_model.bin`` file and a ``config.json`` file:
+
+.. code:: python
+
+   model.save_pretrained("/path/to/converted/checkpoint/folder")
+
+**7. Implement the forward pass**
+
+Having managed to correctly load the pretrained weights into the 🤗 Transformers implementation, you should now make
+sure that the forward pass is correctly implemented. In `Get familiar with the original repository
+<#run-a-pretrained-checkpoint-using-the-original-repository>`__, you have already created a script that runs a forward
+pass of the model using the original repository. Now you should write an analogous script using the 🤗 Transformers
+implementation instead of the original one. It should look as follows:
+
+.. code:: python
+
+   model = BrandNewBertModel.from_pretrained(/path/to/converted/checkpoint/folder)
+   input_ids = [0, 4, 4, 3, 2, 4, 1, 7, 19]
+   output = model(input_ids).last_hidden_states
+
+It is very likely that the 🤗 Transformers implementation and the original model implementation don't give the exact
+same output the very first time or that the forward pass throws an error. Don't be disappointed - it's expected! First,
+you should make sure that the forward pass doesn't throw any errors. It often happens that the wrong dimensions are
+used leading to a `Dimensionality mismatch` error or that the wrong data type object is used, *e.g.* ``torch.long``
+instead of ``torch.float32``. Don't hesitate to ask the Hugging Face team for help, if you don't manage to solve
+certain errors.
+
+The final part to make sure the 🤗 Transformers implementation works correctly is to ensure that the outputs are
+equivalent to a precision of ``1e-3``. First, you should ensure that the output shapes are identical, *i.e.*
+``outputs.shape`` should yield the same value for the script of the 🤗 Transformers implementation and the original
+implementation. Next, you should make sure that the output values are identical as well. This one of the most difficult
+parts of adding a new model. Common mistakes why the outputs are not identical are:
+
+-  Some layers were not added, *i.e.* an `activation` layer was not added, or the residual connection was forgotten
+-  The word embedding matrix was not tied
+-  The wrong positional embeddings are used because the original implementation uses on offset
+-  Dropout is applied during the forward pass. To fix this make sure `model.training is False` and that no dropout
+   layer is falsely activated during the forward pass, *i.e.* pass `self.training` to `PyTorch's functional dropout
+   <https://pytorch.org/docs/stable/nn.functional.html?highlight=dropout#torch.nn.functional.dropout>`_
+
+The best way to fix the problem is usually to look at the forward pass of the original implementation and the 🤗
+Transformers implementation side-by-side and check if there are any differences. Ideally, you should debug/print out
+intermediate outputs of both implementations of the forward pass to find the exact position in the network where the 🤗
+Transformers implementation shows a different output than the original implementation. First, make sure that the
+hard-coded ``input_ids`` in both scripts are identical. Next, verify that the outputs of the first transformation of
+the ``input_ids`` (usually the word embeddings) are identical. And then work your way up to the very last layer of the
+network. At some point, you will notice a difference between the two implementations, which should point you to the bug
+in the 🤗 Transformers implementation. From our experience, a simple and efficient way is to add many print statements
+in both the original implementation and 🤗 Transformers implementation, at the same positions in the network
+respectively, and to successively remove print statements showing the same values for intermediate presentions.
+
+When you're confident that both implementations yield the same output, verifying the outputs with
+``torch.allclose(original_output, output, atol=1e-3)``, you're done with the most difficult part! Congratulations - the
+work left to be done should be a cakewalk 😊.
+
+**8. Adding all necessary model tests**
+
+At this point, you have successfully added a new model. However, it is very much possible that the model does not yet
+fully comply with the required design. To make sure, the implementation is fully compatible with 🤗 Transformers, all
+common tests should pass. The Cookiecutter should have automatically added a test file for your model, probably under
+the same ``tests/test_modeling_brand_new_bert.py``. Run this test file to verify that all common tests pass:
+
+.. code:: python
+
+   pytest tests/test_modeling_brand_new_bert.py
+
+Having fixed all common tests, it is now crucial to ensure that all the nice work you have done is well tested, so that
+
+-  
+
+   a) The community can easily understand your work by looking at specific tests of *brand_new_bert*
+
+-  
+
+   b) Future changes to your model will not break any important feature of the model.
+
+At first, integration tests should be added. Those integration tests essentially do the same as the debugging scripts
+you used earlier to implement the model to 🤗 Transformers. A template of those model tests is already added by the
+Cookiecutter, called ``BrandNewBertModelIntegrationTests`` and only has to be filled out by you. To ensure that those
+tests are passing, run
+
+.. code:: python
+
+   RUN_SLOW=1 pytest -sv tests/test_modeling_brand_new_bert.py::BrandNewBertModelIntegrationTests
+
+.. note::
+
+  In case you are using Windows, you should replace ``RUN_SLOW=1`` with ``SET RUN_SLOW=1``
+
+Second, all features that are special to *brand_new_bert* should be tested additionally in a separate test under
+``BrandNewBertModelTester``/``BrandNewBertModelTest``. This part is often forgotten but is extremely useful in two
+ways:
+
+-  It helps to transfer the knowledge you have acquired during the model addition to the community by showing how the
+   special features of *brand_new_bert* should work.
+-  Future contributors can quickly test changes to the model by running those special tests.
+
+
+**9. Implement the tokenizer**
+
+Next, we should add the tokenizer of *brand_new_bert*. Usually, the tokenizer is equivalent or very similar to an
+already existing tokenizer of 🤗 Transformers.
+
+It is very important to find/extract the original tokenizer file and to manage to load this file into the 🤗
+Transformers' implementation of the tokenizer.
+
+To ensure that the tokenizer works correctly, it is recommended to first create a script in the original repository
+that inputs a string and returns the ``input_ids``. It could look similar to this (in pseudo-code):
+
+.. code:: bash
+
+   input_str = "This is a long example input string containing special characters .$?-, numbers 2872 234 12 and words."
+   model = BrandNewBertModel.load_pretrained_checkpoint(/path/to/checkpoint/)
+   input_ids = model.tokenize(input_str)
+
+You might have to take a deeper look again into the original repository to find the correct tokenizer function or you
+might even have to do changes to your clone of the original repository to only output the ``input_ids``. Having written
+a functional tokenization script that uses the original repository, an analogous script for 🤗 Transformers should be
+created. It should look similar to this:
+
+.. code:: python
+
+   from transformers import BrandNewBertTokenizer
+   input_str = "This is a long example input string containing special characters .$?-, numbers 2872 234 12 and words."
+
+   tokenizer = BrandNewBertTokenizer.from_pretrained(/path/to/tokenizer/folder/)
+
+   input_ids = tokenizer(input_str).input_ids
+
+When both ``input_ids`` yield the same values, as a final step a tokenizer test file should also be added.
+
+Analogous to the modeling test files of *brand_new_bert*, the tokenization test files of *brand_new_bert* should
+contain a couple of hard-coded integration tests.
+
+**10. Run End-to-end integration tests**
+
+Having added the tokenizer, you should also add a couple of end-to-end integration tests using both the model and the
+tokenizer to ``tests/test_modeling_brand_new_bert.py`` in 🤗 Transformers. Such a test should show on a meaningful
+text-to-text sample that the 🤗 Transformers implementation works as expected. A meaningful text-to-text sample can
+include *e.g.* a source-to-target-translation pair, an article-to-summary pair, a question-to-answer pair, etc… If none
+of the ported checkpoints has been fine-tuned on a downstream task it is enough to simply rely on the model tests. In a
+final step to ensure that the model is fully functional, it is advised that you also run all tests on GPU. It can
+happen that you forgot to add some ``.to(self.device)`` statements to internal tensors of the model, which in such a
+test would show in an error. In case you have no access to a GPU, the Hugging Face team can take care of running those
+tests for you.
+
+**11. Add Docstring**
+
+Now, all the necessary functionality for *brand_new_bert* is added - you're almost done! The only thing left to add is
+a nice docstring and a doc page. The Cookiecutter should have added a template file called
+``docs/source/model_doc/brand_new_bert.rst`` that you should fill out. Users of your model will usually first look at
+this page before using your model. Hence, the documentation must be understandable and concise. It is very useful for
+the community to add some *Tips* to show how the model should be used. Don't hesitate to ping the Hugging Face team
+regarding the docstrings.
+
+Next, make sure that the docstring added to ``src/transformers/models/brand_new_bert/modeling_brand_new_bert.py`` is
+correct and included all necessary inputs and outputs. It is always to good to remind oneself that documentation should
+be treated at least as carefully as the code in 🤗 Transformers since the documentation is usually the first contact
+point of the community with the model.
+
+**Code refactor**
+
+Great, now you have added all the necessary code for *brand_new_bert*. At this point, you should correct some potential
+incorrect code style by running:
+
+.. code:: bash
+
+   make style
+
+and verify that your coding style passes the quality check:
+
+.. code:: bash
+
+   make quality
+
+There are a couple of other very strict design tests in 🤗 Transformers that might still be failing, which shows up in
+the tests of your pull request. This is often because of some missing information in the docstring or some incorrect
+naming. The Hugging Face team will surely help you if you're stuck here.
+
+Lastly, it is always a good idea to refactor one's code after having ensured that the code works correctly. With all
+tests passing, now it's a good time to go over the added code again and do some refactoring.
+
+You have now finished the coding part, congratulation! 🎉 You are Awesome! 😎
+
+**12. Upload the models to the model hub**
+
+In this final part, you should convert and upload all checkpoints to the model hub and add a model card for each
+uploaded model checkpoint. You should work alongside the Hugging Face team here to decide on a fitting name for each
+checkpoint and to get the required access rights to be able to upload the model under the author's organization of
+*brand_new_bert*.
+
+It is worth spending some time to create fitting model cards for each checkpoint. The model cards should highlight the
+specific characteristics of this particular checkpoint, *e.g.* On which dataset was the checkpoint
+pretrained/fine-tuned on? On what down-stream task should the model be used? And also include some code on how to
+correctly use the model.
+
+**13. (Optional) Add notebook**
+
+It is very helpful to add a notebook that showcases in-detail how *brand_new_bert* can be used for inference and/or
+fine-tuned on a downstream task. This is not mandatory to merge your PR, but very useful for the community.
+
+**14. Submit your finished PR**
+
+You're done programming now and can move to the last step, which is getting your PR merged into master. Usually, the
+Hugging Face team should have helped you already at this point, but it is worth taking some time to give your finished
+PR a nice description and eventually add comments to your code, if you want to point out certain design choices to your
+reviewer.
+
+Share your work!!
+-----------------------------------------------------------------------------------------------------------------------
+
+Now, it's time to get some credit from the community for your work! Having completed a model addition is a major
+contribution to Transformers and the whole NLP community. Your code and the ported pre-trained models will certainly be
+used by hundreds and possibly even thousands of developers and researchers. You should be proud of your work and share
+your achievement with the community.
+
+**You have made another model that is super easy to access for everyone in the community! 🤯**

docs/source/community.md

@@ -0,0 +1,50 @@
+# Community
+
+This page regroups resources around 🤗 Transformers developed by the community.
+
+## Community resources:
+
+| Resource     |      Description      |      Author      |
+|:----------|:-------------|------:|
+| [Hugging Face Transformers Glossary Flashcards](https://www.darigovresearch.com/huggingface-transformers-glossary-flashcards) | A set of flashcards based on the [Transformers Docs Glossary](https://huggingface.co/transformers/master/glossary.html) that has been put into a form which can be easily learnt/revised using [Anki ](https://apps.ankiweb.net/) an open source, cross platform app specifically designed for long term knowledge retention. See this [Introductory video on how to use the flashcards](https://www.youtube.com/watch?v=Dji_h7PILrw). | [Darigov Research](https://www.darigovresearch.com/) |
+
+## Community notebooks:
+
+| Notebook     |      Description      |      Author      |      |
+|:----------|:-------------|:-------------|------:|
+| [Train T5 in Tensorflow 2 ](https://github.com/snapthat/TF-T5-text-to-text) | How to train T5 for any task using Tensorflow 2. This notebook demonstrates a Question & Answer task implemented in Tensorflow 2 using SQUAD | [Muhammad Harris](https://github.com/HarrisDePerceptron) |[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/snapthat/TF-T5-text-to-text/blob/master/snapthatT5/notebooks/TF-T5-Datasets%20Training.ipynb) |
+| [Train T5 on TPU](https://github.com/patil-suraj/exploring-T5/blob/master/T5_on_TPU.ipynb)  | How to train T5 on SQUAD with Transformers and Nlp | [Suraj Patil](https://github.com/patil-suraj) |[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/patil-suraj/exploring-T5/blob/master/T5_on_TPU.ipynb#scrollTo=QLGiFCDqvuil) |
+| [Fine-tune T5 for Classification and Multiple Choice](https://github.com/patil-suraj/exploring-T5/blob/master/t5_fine_tuning.ipynb)  | How to fine-tune T5 for classification and multiple choice tasks using a text-to-text format with PyTorch Lightning |  [Suraj Patil](https://github.com/patil-suraj) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/patil-suraj/exploring-T5/blob/master/t5_fine_tuning.ipynb) |
+| [Fine-tune DialoGPT on New Datasets and Languages](https://github.com/ncoop57/i-am-a-nerd/blob/master/_notebooks/2020-05-12-chatbot-part-1.ipynb)  | How to fine-tune the DialoGPT model on a new dataset for open-dialog conversational chatbots |  [Nathan Cooper](https://github.com/ncoop57) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/ncoop57/i-am-a-nerd/blob/master/_notebooks/2020-05-12-chatbot-part-1.ipynb) |
+| [Long Sequence Modeling with Reformer](https://github.com/patrickvonplaten/notebooks/blob/master/PyTorch_Reformer.ipynb)  | How to train on sequences as long as 500,000 tokens with Reformer |  [Patrick von Platen](https://github.com/patrickvonplaten) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/patrickvonplaten/notebooks/blob/master/PyTorch_Reformer.ipynb)  |
+| [Fine-tune BART for Summarization](https://github.com/ohmeow/ohmeow_website/blob/master/_notebooks/2020-05-23-text-generation-with-blurr.ipynb) | How to fine-tune BART for summarization with fastai using blurr | [Wayde Gilliam](https://ohmeow.com/) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/ohmeow/ohmeow_website/blob/master/_notebooks/2020-05-23-text-generation-with-blurr.ipynb) |
+| [Fine-tune a pre-trained Transformer on anyone's tweets](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb) | How to generate tweets in the style of your favorite Twitter account by fine-tuning a GPT-2 model |  [Boris Dayma](https://github.com/borisdayma) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb) |
+| [Optimize 🤗 Hugging Face models with Weights & Biases](https://colab.research.google.com/github/wandb/examples/blob/master/colabs/huggingface/Optimize_Hugging_Face_models_with_Weights_%26_Biases.ipynb) | A complete tutorial showcasing W&B integration with Hugging Face | [Boris Dayma](https://github.com/borisdayma) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/wandb/examples/blob/master/colabs/huggingface/Optimize_Hugging_Face_models_with_Weights_%26_Biases.ipynb) |
+| [Pretrain Longformer](https://github.com/allenai/longformer/blob/master/scripts/convert_model_to_long.ipynb)  | How to build a "long" version of existing pretrained models |  [Iz Beltagy](https://beltagy.net) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/allenai/longformer/blob/master/scripts/convert_model_to_long.ipynb) |
+| [Fine-tune Longformer for QA](https://github.com/patil-suraj/Notebooks/blob/master/longformer_qa_training.ipynb) | How to fine-tune longformer model for QA task | [Suraj Patil](https://github.com/patil-suraj) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/patil-suraj/Notebooks/blob/master/longformer_qa_training.ipynb) |
+| [Evaluate Model with 🤗nlp](https://github.com/patrickvonplaten/notebooks/blob/master/How_to_evaluate_Longformer_on_TriviaQA_using_NLP.ipynb) | How to evaluate longformer on TriviaQA with `nlp` | [Patrick von Platen](https://github.com/patrickvonplaten) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1m7eTGlPmLRgoPkkA7rkhQdZ9ydpmsdLE?usp=sharing) |
+| [Fine-tune T5 for Sentiment Span Extraction](https://github.com/enzoampil/t5-intro/blob/master/t5_qa_training_pytorch_span_extraction.ipynb)  | How to fine-tune T5 for sentiment span extraction using a text-to-text format with PyTorch Lightning |  [Lorenzo Ampil](https://github.com/enzoampil) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/enzoampil/t5-intro/blob/master/t5_qa_training_pytorch_span_extraction.ipynb) |
+| [Fine-tune DistilBert for Multiclass Classification](https://github.com/abhimishra91/transformers-tutorials/blob/master/transformers_multiclass_classification.ipynb) | How to fine-tune DistilBert for multiclass classification with PyTorch | [Abhishek Kumar Mishra](https://github.com/abhimishra91) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/abhimishra91/transformers-tutorials/blob/master/transformers_multiclass_classification.ipynb)|
+|[Fine-tune BERT for Multi-label Classification](https://github.com/abhimishra91/transformers-tutorials/blob/master/transformers_multi_label_classification.ipynb)|How to fine-tune BERT for multi-label classification using PyTorch|[Abhishek Kumar Mishra](https://github.com/abhimishra91) |[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/abhimishra91/transformers-tutorials/blob/master/transformers_multi_label_classification.ipynb)|
+|[Fine-tune T5 for Summarization](https://github.com/abhimishra91/transformers-tutorials/blob/master/transformers_summarization_wandb.ipynb)|How to fine-tune T5 for summarization in PyTorch and track experiments with WandB|[Abhishek Kumar Mishra](https://github.com/abhimishra91) |[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/abhimishra91/transformers-tutorials/blob/master/transformers_summarization_wandb.ipynb)|
+|[Speed up Fine-Tuning in Transformers with Dynamic Padding / Bucketing](https://github.com/ELS-RD/transformers-notebook/blob/master/Divide_Hugging_Face_Transformers_training_time_by_2_or_more.ipynb)|How to speed up fine-tuning by a factor of 2 using dynamic padding / bucketing|[Michael Benesty](https://github.com/pommedeterresautee) |[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1CBfRU1zbfu7-ijiOqAAQUA-RJaxfcJoO?usp=sharing)|
+|[Pretrain Reformer for Masked Language Modeling](https://github.com/patrickvonplaten/notebooks/blob/master/Reformer_For_Masked_LM.ipynb)| How to train a Reformer model with bi-directional self-attention layers | [Patrick von Platen](https://github.com/patrickvonplaten) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1tzzh0i8PgDQGV3SMFUGxM7_gGae3K-uW?usp=sharing)|
+|[Expand and Fine Tune Sci-BERT](https://github.com/lordtt13/word-embeddings/blob/master/COVID-19%20Research%20Data/COVID-SciBERT.ipynb)| How to increase vocabulary of a pretrained SciBERT model from AllenAI on the CORD dataset and pipeline it. | [Tanmay Thakur](https://github.com/lordtt13) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1rqAR40goxbAfez1xvF3hBJphSCsvXmh8)|
+|[Fine Tune BlenderBotSmall for Summarization using the Trainer API](https://github.com/lordtt13/transformers-experiments/blob/master/Custom%20Tasks/fine-tune-blenderbot_small-for-summarization.ipynb)| How to fine tune BlenderBotSmall for summarization on a custom dataset, using the Trainer API. | [Tanmay Thakur](https://github.com/lordtt13) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/19Wmupuls7mykSGyRN_Qo6lPQhgp56ymq?usp=sharing)|
+|[Fine-tune Electra and interpret with Integrated Gradients](https://github.com/elsanns/xai-nlp-notebooks/blob/master/electra_fine_tune_interpret_captum_ig.ipynb) | How to fine-tune Electra for sentiment analysis and interpret predictions with Captum Integrated Gradients | [Eliza Szczechla](https://elsanns.github.io) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/elsanns/xai-nlp-notebooks/blob/master/electra_fine_tune_interpret_captum_ig.ipynb)|
+|[fine-tune a non-English GPT-2 Model with Trainer class](https://github.com/philschmid/fine-tune-GPT-2/blob/master/Fine_tune_a_non_English_GPT_2_Model_with_Huggingface.ipynb) | How to fine-tune a non-English GPT-2 Model with Trainer class | [Philipp Schmid](https://www.philschmid.de) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/philschmid/fine-tune-GPT-2/blob/master/Fine_tune_a_non_English_GPT_2_Model_with_Huggingface.ipynb)|
+|[Fine-tune a DistilBERT Model for Multi Label Classification task](https://github.com/DhavalTaunk08/Transformers_scripts/blob/master/Transformers_multilabel_distilbert.ipynb) | How to fine-tune a DistilBERT Model for Multi Label Classification task | [Dhaval Taunk](https://github.com/DhavalTaunk08) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/DhavalTaunk08/Transformers_scripts/blob/master/Transformers_multilabel_distilbert.ipynb)|
+|[Fine-tune ALBERT for sentence-pair classification](https://github.com/NadirEM/nlp-notebooks/blob/master/Fine_tune_ALBERT_sentence_pair_classification.ipynb) | How to fine-tune an ALBERT model or another BERT-based model for the sentence-pair classification task | [Nadir El Manouzi](https://github.com/NadirEM) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/NadirEM/nlp-notebooks/blob/master/Fine_tune_ALBERT_sentence_pair_classification.ipynb)|
+|[Fine-tune Roberta for sentiment analysis](https://github.com/DhavalTaunk08/NLP_scripts/blob/master/sentiment_analysis_using_roberta.ipynb) | How to fine-tune an Roberta model for sentiment analysis | [Dhaval Taunk](https://github.com/DhavalTaunk08) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/DhavalTaunk08/NLP_scripts/blob/master/sentiment_analysis_using_roberta.ipynb)|
+|[Evaluating Question Generation Models](https://github.com/flexudy-pipe/qugeev) | How accurate are the answers to questions generated by your seq2seq transformer model? | [Pascal Zoleko](https://github.com/zolekode) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1bpsSqCQU-iw_5nNoRm_crPq6FRuJthq_?usp=sharing)|
+|[Classify text with DistilBERT and Tensorflow](https://github.com/peterbayerle/huggingface_notebook/blob/main/distilbert_tf.ipynb) | How to fine-tune DistilBERT for text classification in TensorFlow | [Peter Bayerle](https://github.com/peterbayerle) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/peterbayerle/huggingface_notebook/blob/main/distilbert_tf.ipynb)|
+|[Leverage BERT for Encoder-Decoder Summarization on CNN/Dailymail](https://github.com/patrickvonplaten/notebooks/blob/master/BERT2BERT_for_CNN_Dailymail.ipynb) | How to warm-start a *EncoderDecoderModel* with a *bert-base-uncased* checkpoint for summarization on CNN/Dailymail | [Patrick von Platen](https://github.com/patrickvonplaten) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/patrickvonplaten/notebooks/blob/master/BERT2BERT_for_CNN_Dailymail.ipynb)|
+|[Leverage RoBERTa for Encoder-Decoder Summarization on BBC XSum](https://github.com/patrickvonplaten/notebooks/blob/master/RoBERTaShared_for_BBC_XSum.ipynb) | How to warm-start a shared *EncoderDecoderModel* with a *roberta-base* checkpoint for summarization on BBC/XSum | [Patrick von Platen](https://github.com/patrickvonplaten) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/patrickvonplaten/notebooks/blob/master/RoBERTaShared_for_BBC_XSum.ipynb)|
+|[Fine-tune TAPAS on Sequential Question Answering (SQA)](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/TAPAS/Fine_tuning_TapasForQuestionAnswering_on_SQA.ipynb) | How to fine-tune *TapasForQuestionAnswering* with a *tapas-base* checkpoint on the Sequential Question Answering (SQA) dataset | [Niels Rogge](https://github.com/nielsrogge) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/TAPAS/Fine_tuning_TapasForQuestionAnswering_on_SQA.ipynb)|
+|[Evaluate TAPAS on Table Fact Checking (TabFact)](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/TAPAS/Evaluating_TAPAS_on_the_Tabfact_test_set.ipynb) | How to evaluate a fine-tuned *TapasForSequenceClassification* with a *tapas-base-finetuned-tabfact* checkpoint using a combination of the 🤗 datasets and 🤗 transformers libraries | [Niels Rogge](https://github.com/nielsrogge) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/TAPAS/Evaluating_TAPAS_on_the_Tabfact_test_set.ipynb)|
+|[Fine-tuning mBART for translation](https://colab.research.google.com/github/vasudevgupta7/huggingface-tutorials/blob/main/translation_training.ipynb) | How to fine-tune mBART using Seq2SeqTrainer for Hindi to English translation | [Vasudev Gupta](https://github.com/vasudevgupta7) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vasudevgupta7/huggingface-tutorials/blob/main/translation_training.ipynb)|
+|[Fine-tune LayoutLM on FUNSD (a form understanding dataset)](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/LayoutLM/Fine_tuning_LayoutLMForTokenClassification_on_FUNSD.ipynb) | How to fine-tune *LayoutLMForTokenClassification* on the FUNSD dataset for information extraction from scanned documents | [Niels Rogge](https://github.com/nielsrogge) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/LayoutLM/Fine_tuning_LayoutLMForTokenClassification_on_FUNSD.ipynb)|
+|[Fine-Tune DistilGPT2 and Generate Text](https://colab.research.google.com/github/tripathiaakash/DistilGPT2-Tutorial/blob/main/distilgpt2_fine_tuning.ipynb) | How to fine-tune DistilGPT2 and generate text | [Aakash Tripathi](https://github.com/tripathiaakash) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/tripathiaakash/DistilGPT2-Tutorial/blob/main/distilgpt2_fine_tuning.ipynb)|
+|[Fine-Tune LED on up to 8K tokens](https://github.com/patrickvonplaten/notebooks/blob/master/Fine_tune_Longformer_Encoder_Decoder_(LED)_for_Summarization_on_pubmed.ipynb) | How to fine-tune LED on pubmed for long-range summarization | [Patrick von Platen](https://github.com/patrickvonplaten) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/patrickvonplaten/notebooks/blob/master/Fine_tune_Longformer_Encoder_Decoder_(LED)_for_Summarization_on_pubmed.ipynb)|
+|[Evaluate LED on Arxiv](https://github.com/patrickvonplaten/notebooks/blob/master/LED_on_Arxiv.ipynb) | How to effectively evaluate LED on long-range summarization | [Patrick von Platen](https://github.com/patrickvonplaten) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/patrickvonplaten/notebooks/blob/master/LED_on_Arxiv.ipynb)|
+|[Fine-tune LayoutLM on RVL-CDIP (a document image classification dataset)](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/LayoutLM/Fine_tuning_LayoutLMForSequenceClassification_on_RVL_CDIP.ipynb) | How to fine-tune *LayoutLMForSequenceClassification* on the RVL-CDIP dataset for scanned document classification | [Niels Rogge](https://github.com/nielsrogge) | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/LayoutLM/Fine_tuning_LayoutLMForSequenceClassification_on_RVL_CDIP.ipynb)|

docs/source/conf.py

@@ -26,7 +26,10 @@ author = u'huggingface'
 # The short X.Y version
 version = u''
 # The full version, including alpha/beta/rc tags
-release = u'4.2.0'
+release = u'4.4.2'
+
+
+
 # Prefix link to point to master, comment this during version release and uncomment below line
 extlinks = {'prefix_link': ('https://github.com/huggingface/transformers/blob/master/%s', '')}
 # Prefix link to always point to corresponding version, uncomment this during version release
@@ -95,7 +98,8 @@ html_theme = 'sphinx_rtd_theme'
 # documentation.
 #
 html_theme_options = {
-    'analytics_id': 'UA-83738774-2'
+    'analytics_id': 'UA-83738774-2',
+    'navigation_with_keys': True
 }
 
 # Add any paths that contain custom static files (such as style sheets) here,

docs/source/converting_tensorflow_models.rst

@@ -28,17 +28,13 @@ BERT
 You can convert any TensorFlow checkpoint for BERT (in particular `the pre-trained models released by Google
 <https://github.com/google-research/bert#pre-trained-models>`_\ ) in a PyTorch save file by using the
 :prefix_link:`convert_bert_original_tf_checkpoint_to_pytorch.py
-<src/transformers/convert_bert_original_tf_checkpoint_to_pytorch.py>` script.
+<src/transformers/models/bert/convert_bert_original_tf_checkpoint_to_pytorch.py>` script.
 
 This CLI takes as input a TensorFlow checkpoint (three files starting with ``bert_model.ckpt``\ ) and the associated
 configuration file (\ ``bert_config.json``\ ), and creates a PyTorch model for this configuration, loads the weights
 from the TensorFlow checkpoint in the PyTorch model and saves the resulting model in a standard PyTorch save file that
-can be imported using ``torch.load()`` (see examples in `run_bert_extract_features.py
-<https://github.com/huggingface/pytorch-pretrained-BERT/tree/master/examples/run_bert_extract_features.py>`_\ ,
-`run_bert_classifier.py
-<https://github.com/huggingface/pytorch-pretrained-BERT/tree/master/examples/run_bert_classifier.py>`_ and
-`run_bert_squad.py <https://github.com/huggingface/pytorch-pretrained-BERT/tree/master/examples/run_bert_squad.py>`_\
-).
+can be imported using ``from_pretrained()`` (see example in :doc:`quicktour` , `run_glue.py
+<https://github.com/huggingface/transformers/blob/master/examples/text-classification/run_glue.py>`_\ ).
 
 You only need to run this conversion script **once** to get a PyTorch model. You can then disregard the TensorFlow
 checkpoint (the three files starting with ``bert_model.ckpt``\ ) but be sure to keep the configuration file (\
@@ -66,7 +62,7 @@ ALBERT
 
 Convert TensorFlow model checkpoints of ALBERT to PyTorch using the
 :prefix_link:`convert_albert_original_tf_checkpoint_to_pytorch.py
-<src/transformers/convert_bert_original_tf_checkpoint_to_pytorch.py>` script.
+<src/transformers/models/albert/convert_albert_original_tf_checkpoint_to_pytorch.py>` script.
 
 The CLI takes as input a TensorFlow checkpoint (three files starting with ``model.ckpt-best``\ ) and the accompanying
 configuration file (\ ``albert_config.json``\ ), then creates and saves a PyTorch model. To run this conversion you
@@ -168,3 +164,18 @@ Here is an example of the conversion process for a pre-trained XLM model:
      --pytorch_dump_output $PYTORCH_DUMP_OUTPUT
     [--config XML_CONFIG] \
     [--finetuning_task_name XML_FINETUNED_TASK]
+
+
+T5
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+Here is an example of the conversion process for a pre-trained T5 model:
+
+.. code-block:: shell
+
+   export T5=/path/to/t5/uncased_L-12_H-768_A-12
+
+   transformers-cli convert --model_type t5 \
+     --tf_checkpoint $T5/t5_model.ckpt \
+     --config $T5/t5_config.json \
+     --pytorch_dump_output $T5/pytorch_model.bin

docs/source/custom_datasets.rst

@@ -75,7 +75,7 @@ read this in.
     test_texts, test_labels = read_imdb_split('aclImdb/test')
 
 We now have a train and test dataset, but let's also also create a validation set which we can use for for evaluation
-and tuning without training our test set results. Sklearn has a convenient utility for creating such splits:
+and tuning without tainting our test set results. Sklearn has a convenient utility for creating such splits:
 
 .. code-block:: python
 
@@ -558,15 +558,14 @@ we can use the built in :func:`~transformers.BatchEncoding.char_to_token` method
         end_positions = []
         for i in range(len(answers)):
             start_positions.append(encodings.char_to_token(i, answers[i]['answer_start']))
-            end_positions.append(encodings.char_to_token(i, answers[i]['answer_end']))
+            end_positions.append(encodings.char_to_token(i, answers[i]['answer_end'] - 1))
 
             # if start position is None, the answer passage has been truncated
             if start_positions[-1] is None:
                 start_positions[-1] = tokenizer.model_max_length
-
-            # if end position is None, the 'char_to_token' function points to the space before the correct token - > add + 1
             if end_positions[-1] is None:
-                end_positions[-1] = encodings.char_to_token(i, answers[i]['answer_end'] + 1)
+                end_positions[-1] = tokenizer.model_max_length
+
         encodings.update({'start_positions': start_positions, 'end_positions': end_positions})
 
     add_token_positions(train_encodings, train_answers)

docs/source/glossary.rst

@@ -21,22 +21,25 @@ General terms
 - CLM: causal language modeling, a pretraining task where the model reads the texts in order and has to predict the
   next word. It's usually done by reading the whole sentence but using a mask inside the model to hide the future
   tokens at a certain timestep.
+- deep learning: machine learning algorithms which uses neural networks with several layers.
 - MLM: masked language modeling, a pretraining task where the model sees a corrupted version of the texts, usually done
   by masking some tokens randomly, and has to predict the original text.
 - multimodal: a task that combines texts with another kind of inputs (for instance images).
-- NLG: natural language generation, all tasks related to generating text ( for instance talk with transformers,
-  translation)
+- NLG: natural language generation, all tasks related to generating text (for instance talk with transformers,
+  translation).
 - NLP: natural language processing, a generic way to say "deal with texts".
 - NLU: natural language understanding, all tasks related to understanding what is in a text (for instance classifying
-  the whole text, individual words)
+  the whole text, individual words).
 - pretrained model: a model that has been pretrained on some data (for instance all of Wikipedia). Pretraining methods
   involve a self-supervised objective, which can be reading the text and trying to predict the next word (see CLM) or
   masking some words and trying to predict them (see MLM).
 - RNN: recurrent neural network, a type of model that uses a loop over a layer to process texts.
+- self-attention: each element of the input finds out which other elements of the input they should attend to.
 - seq2seq or sequence-to-sequence: models that generate a new sequence from an input, like translation models, or
   summarization models (such as :doc:`Bart </model_doc/bart>` or :doc:`T5 </model_doc/t5>`).
 - token: a part of a sentence, usually a word, but can also be a subword (non-common words are often split in subwords)
   or a punctuation symbol.
+- transformer: self-attention based deep learning model architecture.
 
 Model inputs
 -----------------------------------------------------------------------------------------------------------------------

docs/source/index.rst

@@ -103,102 +103,126 @@ and conversion utilities for the following models:
 7. :doc:`BlenderbotSmall <model_doc/blenderbot_small>` (from Facebook) released with the paper `Recipes for building an
    open-domain chatbot <https://arxiv.org/abs/2004.13637>`__ by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary
    Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
-8. :doc:`CamemBERT <model_doc/camembert>` (from Inria/Facebook/Sorbonne) released with the paper `CamemBERT: a Tasty
+8. :doc:`BORT <model_doc/bort>` (from Alexa) released with the paper `Optimal Subarchitecture Extraction For BERT
+   <https://arxiv.org/abs/2010.10499>`__ by Adrian de Wynter and Daniel J. Perry.
+9. :doc:`CamemBERT <model_doc/camembert>` (from Inria/Facebook/Sorbonne) released with the paper `CamemBERT: a Tasty
    French Language Model <https://arxiv.org/abs/1911.03894>`__ by Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz
    Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah and Benoît Sagot.
-9. :doc:`CTRL <model_doc/ctrl>` (from Salesforce) released with the paper `CTRL: A Conditional Transformer Language
-   Model for Controllable Generation <https://arxiv.org/abs/1909.05858>`__ by Nitish Shirish Keskar*, Bryan McCann*,
-   Lav R. Varshney, Caiming Xiong and Richard Socher.
-10. :doc:`DeBERTa <model_doc/deberta>` (from Microsoft Research) released with the paper `DeBERTa: Decoding-enhanced
-    BERT with Disentangled Attention <https://arxiv.org/abs/2006.03654>`__ by Pengcheng He, Xiaodong Liu, Jianfeng Gao,
+10. :doc:`ConvBERT <model_doc/convbert>` (from YituTech) released with the paper `ConvBERT: Improving BERT with
+    Span-based Dynamic Convolution <https://arxiv.org/abs/2008.02496>`__ by Zihang Jiang, Weihao Yu, Daquan Zhou,
+    Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
+11. :doc:`CTRL <model_doc/ctrl>` (from Salesforce) released with the paper `CTRL: A Conditional Transformer Language
+    Model for Controllable Generation <https://arxiv.org/abs/1909.05858>`__ by Nitish Shirish Keskar*, Bryan McCann*,
+    Lav R. Varshney, Caiming Xiong and Richard Socher.
+12. :doc:`DeBERTa <model_doc/deberta>` (from Microsoft) released with the paper `DeBERTa: Decoding-enhanced BERT with
+    Disentangled Attention <https://arxiv.org/abs/2006.03654>`__ by Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu
+    Chen.
+13. :doc:`DeBERTa-v2 <model_doc/deberta_v2>` (from Microsoft) released with the paper `DeBERTa: Decoding-enhanced BERT
+    with Disentangled Attention <https://arxiv.org/abs/2006.03654>`__ by Pengcheng He, Xiaodong Liu, Jianfeng Gao,
     Weizhu Chen.
-11. :doc:`DialoGPT <model_doc/dialogpt>` (from Microsoft Research) released with the paper `DialoGPT: Large-Scale
+14. :doc:`DialoGPT <model_doc/dialogpt>` (from Microsoft Research) released with the paper `DialoGPT: Large-Scale
     Generative Pre-training for Conversational Response Generation <https://arxiv.org/abs/1911.00536>`__ by Yizhe
     Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan.
-12. :doc:`DistilBERT <model_doc/distilbert>` (from HuggingFace), released together with the paper `DistilBERT, a
+15. :doc:`DistilBERT <model_doc/distilbert>` (from HuggingFace), released together with the paper `DistilBERT, a
     distilled version of BERT: smaller, faster, cheaper and lighter <https://arxiv.org/abs/1910.01108>`__ by Victor
     Sanh, Lysandre Debut and Thomas Wolf. The same method has been applied to compress GPT2 into `DistilGPT2
     <https://github.com/huggingface/transformers/tree/master/examples/distillation>`__, RoBERTa into `DistilRoBERTa
     <https://github.com/huggingface/transformers/tree/master/examples/distillation>`__, Multilingual BERT into
     `DistilmBERT <https://github.com/huggingface/transformers/tree/master/examples/distillation>`__ and a German
     version of DistilBERT.
-13. :doc:`DPR <model_doc/dpr>` (from Facebook) released with the paper `Dense Passage Retrieval for Open-Domain
+16. :doc:`DPR <model_doc/dpr>` (from Facebook) released with the paper `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, and Wen-tau Yih.
-14. :doc:`ELECTRA <model_doc/electra>` (from Google Research/Stanford University) released with the paper `ELECTRA:
+17. :doc:`ELECTRA <model_doc/electra>` (from Google Research/Stanford University) released with the paper `ELECTRA:
     Pre-training text encoders as discriminators rather than generators <https://arxiv.org/abs/2003.10555>`__ by Kevin
     Clark, Minh-Thang Luong, Quoc V. Le, Christopher D. Manning.
-15. :doc:`FlauBERT <model_doc/flaubert>` (from CNRS) released with the paper `FlauBERT: Unsupervised Language Model
+18. :doc:`FlauBERT <model_doc/flaubert>` (from CNRS) released with the paper `FlauBERT: Unsupervised Language Model
     Pre-training for French <https://arxiv.org/abs/1912.05372>`__ by Hang Le, Loïc Vial, Jibril Frej, Vincent Segonne,
     Maximin Coavoux, Benjamin Lecouteux, Alexandre Allauzen, Benoît Crabbé, Laurent Besacier, Didier Schwab.
-16. :doc:`Funnel Transformer <model_doc/funnel>` (from CMU/Google Brain) released with the paper `Funnel-Transformer:
+19. :doc:`Funnel Transformer <model_doc/funnel>` (from CMU/Google Brain) released with the paper `Funnel-Transformer:
     Filtering out Sequential Redundancy for Efficient Language Processing <https://arxiv.org/abs/2006.03236>`__ by
     Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
-17. :doc:`GPT <model_doc/gpt>` (from OpenAI) released with the paper `Improving Language Understanding by Generative
+20. :doc:`GPT <model_doc/gpt>` (from OpenAI) released with the paper `Improving Language Understanding by Generative
     Pre-Training <https://blog.openai.com/language-unsupervised/>`__ by Alec Radford, Karthik Narasimhan, Tim Salimans
     and Ilya Sutskever.
-18. :doc:`GPT-2 <model_doc/gpt2>` (from OpenAI) released with the paper `Language Models are Unsupervised Multitask
+21. :doc:`GPT-2 <model_doc/gpt2>` (from OpenAI) released with the paper `Language Models are Unsupervised Multitask
     Learners <https://blog.openai.com/better-language-models/>`__ by Alec Radford*, Jeffrey Wu*, Rewon Child, David
     Luan, Dario Amodei** and Ilya Sutskever**.
-19. :doc:`LayoutLM <model_doc/layoutlm>` (from Microsoft Research Asia) released with the paper `LayoutLM: Pre-training
+22. :doc:`I-BERT <model_doc/ibert>` (from Berkeley) released with the paper `I-BERT: Integer-only BERT Quantization
+    <https://arxiv.org/abs/2101.01321>`__ by Sehoon Kim, Amir Gholami, Zhewei Yao, Michael W. Mahoney, Kurt Keutzer
+23. :doc:`LayoutLM <model_doc/layoutlm>` (from Microsoft Research Asia) released with 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, Ming Zhou.
-20. :doc:`LED <model_doc/led>` (from AllenAI) released with the paper `Longformer: The Long-Document Transformer
+24. :doc:`LED <model_doc/led>` (from AllenAI) released with the paper `Longformer: The Long-Document Transformer
     <https://arxiv.org/abs/2004.05150>`__ by Iz Beltagy, Matthew E. Peters, Arman Cohan.
-21. :doc:`Longformer <model_doc/longformer>` (from AllenAI) released with the paper `Longformer: The Long-Document
+25. :doc:`Longformer <model_doc/longformer>` (from AllenAI) released with the paper `Longformer: The Long-Document
     Transformer <https://arxiv.org/abs/2004.05150>`__ by Iz Beltagy, Matthew E. Peters, Arman Cohan.
-22. :doc:`LXMERT <model_doc/lxmert>` (from UNC Chapel Hill) released with the paper `LXMERT: Learning Cross-Modality
+26. :doc:`LXMERT <model_doc/lxmert>` (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.
-23. :doc:`MarianMT <model_doc/marian>` Machine translation models trained using `OPUS <http://opus.nlpl.eu/>`__ data by
+27. :doc:`M2M100 <model_doc/m2m_100>` (from Facebook) released with the paper `Beyond English-Centric Multilingual
+    Machine Translation <https://arxiv.org/abs/2010.11125>`__ by by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi
+    Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman
+    Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
+28. :doc:`MarianMT <model_doc/marian>` 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.
-24. :doc:`MBart <model_doc/mbart>` (from Facebook) released with the paper `Multilingual Denoising Pre-training for
+29. :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.
-25. :doc:`MPNet <model_doc/mpnet>` (from Microsoft Research) released with the paper `MPNet: Masked and Permuted
+30. :doc:`MBart-50 <model_doc/mbart>` (from Facebook) released with the paper `Multilingual Translation with Extensible
+    Multilingual Pretraining and Finetuning <https://arxiv.org/abs/2008.00401>`__ by Yuqing Tang, Chau Tran, Xian Li,
+    Peng-Jen Chen, Naman Goyal, Vishrav Chaudhary, Jiatao Gu, Angela Fan.
+31. :doc:`MPNet <model_doc/mpnet>` (from Microsoft Research) released with the paper `MPNet: Masked and Permuted
     Pre-training for Language Understanding <https://arxiv.org/abs/2004.09297>`__ by Kaitao Song, Xu Tan, Tao Qin,
     Jianfeng Lu, Tie-Yan Liu.
-26. :doc:`MT5 <model_doc/mt5>` (from Google AI) released with the paper `mT5: A massively multilingual pre-trained
+32. :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.
-27. :doc:`Pegasus <model_doc/pegasus>` (from Google) released with the paper `PEGASUS: Pre-training with Extracted
+33. :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.
-28. :doc:`ProphetNet <model_doc/prophetnet>` (from Microsoft Research) released with the paper `ProphetNet: Predicting
+34. :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.
-29. :doc:`Reformer <model_doc/reformer>` (from Google Research) released with the paper `Reformer: The Efficient
+35. :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.
-30. :doc:`RoBERTa <model_doc/roberta>` (from Facebook), released together with the paper a `Robustly Optimized BERT
+36. :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.
-31. :doc:`SqueezeBert <model_doc/squeezebert>` released with the paper `SqueezeBERT: What can computer vision teach NLP
+    Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov.
+37. :doc:`SpeechToTextTransformer <model_doc/speech_to_text>` (from Facebook), released together with the paper
+    `fairseq S2T: Fast Speech-to-Text Modeling with fairseq <https://arxiv.org/abs/2010.05171>`__ by Changhan Wang, Yun
+    Tang, Xutai Ma, Anne Wu, Dmytro Okhonko, Juan Pino.
+38. :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.
-32. :doc:`T5 <model_doc/t5>` (from Google AI) released with the paper `Exploring the Limits of Transfer Learning with a
+39. :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.
-33. :doc:`TAPAS <model_doc/tapas>` (from Google AI) released with the paper `TAPAS: Weakly Supervised Table Parsing via
+40. :doc:`TAPAS <model_doc/tapas>` (from Google AI) released with the paper `TAPAS: Weakly Supervised Table Parsing via
     Pre-training <https://arxiv.org/abs/2004.02349>`__ by Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Müller,
     Francesco Piccinno and Julian Martin Eisenschlos.
-34. :doc:`Transformer-XL <model_doc/transformerxl>` (from Google/CMU) released with the paper `Transformer-XL:
+41. :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.
-35. :doc:`XLM <model_doc/xlm>` (from Facebook) released together with the paper `Cross-lingual Language Model
+42. :doc:`Wav2Vec2 <model_doc/wav2vec2>` (from Facebook AI) released with the paper `wav2vec 2.0: A Framework for
+    Self-Supervised Learning of Speech Representations <https://arxiv.org/abs/2006.11477>`__ by Alexei Baevski, Henry
+    Zhou, Abdelrahman Mohamed, Michael Auli.
+43. :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.
-36. :doc:`XLM-ProphetNet <model_doc/xlmprophetnet>` (from Microsoft Research) released with the paper `ProphetNet:
+44. :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.
-37. :doc:`XLM-RoBERTa <model_doc/xlmroberta>` (from Facebook AI), released together with the paper `Unsupervised
+45. :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.
-38. :doc:`XLNet <model_doc/xlnet>` (from Google/CMU) released with the paper `​XLNet: Generalized Autoregressive
+46. :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.
+47. :doc:`XLSR-Wav2Vec2 <model_doc/xlsr_wav2vec2>` (from Facebook AI) released with the paper `Unsupervised
+    Cross-Lingual Representation Learning For Speech Recognition <https://arxiv.org/abs/2006.13979>`__ by Alexis
+    Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
 
 
 .. _bigtable:
@@ -231,10 +255,14 @@ TensorFlow and/or Flax.
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |          CamemBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|          ConvBERT           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |             DPR             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |           DeBERTa           |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|         DeBERTa-v2          |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |         DistilBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |           ELECTRA           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
@@ -247,6 +275,8 @@ TensorFlow and/or Flax.
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |     Funnel Transformer      |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|           I-BERT            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |             LED             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |           LXMERT            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
@@ -255,6 +285,8 @@ TensorFlow and/or Flax.
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |         Longformer          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|           M2M100            |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |            MPNet            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |           Marian            |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
@@ -269,7 +301,7 @@ TensorFlow and/or Flax.
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |         ProphetNet          |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
-|             RAG             |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             RAG             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |          Reformer           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
@@ -277,6 +309,8 @@ TensorFlow and/or Flax.
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |           RoBERTa           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|         Speech2Text         |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |         SqueezeBERT         |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |             T5              |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
@@ -285,6 +319,8 @@ TensorFlow and/or Flax.
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |       Transformer-XL        |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
+|          Wav2Vec2           |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
++-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |             XLM             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
 +-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
 |         XLM-RoBERTa         |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
@@ -327,9 +363,11 @@ TensorFlow and/or Flax.
     examples
     custom_datasets
     notebooks
+    community
     converting_tensorflow_models
     migration
     contributing
+    add_new_model
     testing
     serialization
 
@@ -355,6 +393,7 @@ TensorFlow and/or Flax.
     main_classes/processors
     main_classes/tokenizer
     main_classes/trainer
+    main_classes/feature_extractor
 
 .. toctree::
     :maxdepth: 2
@@ -369,9 +408,12 @@ TensorFlow and/or Flax.
     model_doc/bertgeneration
     model_doc/blenderbot
     model_doc/blenderbot_small
+    model_doc/bort
     model_doc/camembert
+    model_doc/convbert
     model_doc/ctrl
     model_doc/deberta
+    model_doc/deberta_v2
     model_doc/dialogpt
     model_doc/distilbert
     model_doc/dpr
@@ -381,11 +423,13 @@ TensorFlow and/or Flax.
     model_doc/fsmt
     model_doc/funnel
     model_doc/herbert
+    model_doc/ibert
     model_doc/layoutlm
     model_doc/led
     model_doc/longformer
     model_doc/lxmert
     model_doc/marian
+    model_doc/m2m_100
     model_doc/mbart
     model_doc/mobilebert
     model_doc/mpnet
@@ -399,14 +443,17 @@ TensorFlow and/or Flax.
     model_doc/reformer
     model_doc/retribert
     model_doc/roberta
+    model_doc/speech_to_text
     model_doc/squeezebert
     model_doc/t5
     model_doc/tapas
     model_doc/transformerxl
+    model_doc/wav2vec2
     model_doc/xlm
     model_doc/xlmprophetnet
     model_doc/xlmroberta
     model_doc/xlnet
+    model_doc/xlsr_wav2vec2
 
 .. toctree::
     :maxdepth: 2
@@ -417,3 +464,4 @@ TensorFlow and/or Flax.
     internal/tokenization_utils
     internal/trainer_utils
     internal/generation_utils
+    internal/file_utils

docs/source/installation.md

@@ -19,7 +19,7 @@ limitations under the License.
 🤗 Transformers is tested on Python 3.6+, and PyTorch 1.1.0+ or TensorFlow 2.0+.
 
 You should install 🤗 Transformers in a [virtual environment](https://docs.python.org/3/library/venv.html). If you're
-unfamiliar with Python virtual environments, check out the [user guide](https://packaging.python.org/guides/installing-using-pip-and-virtual-environments/). Create a virtual environment with the version of Python you're going 
+unfamiliar with Python virtual environments, check out the [user guide](https://packaging.python.org/guides/installing-using-pip-and-virtual-environments/). Create a virtual environment with the version of Python you're going
 to use and activate it.
 
 Now, if you want to use 🤗 Transformers, you can install it with pip. If you'd like to play with the examples, you
@@ -28,8 +28,8 @@ must install it from source.
 ## Installation with pip
 
 First you need to install one of, or both, TensorFlow 2.0 and PyTorch.
-Please refer to [TensorFlow installation page](https://www.tensorflow.org/install/pip#tensorflow-2.0-rc-is-available), 
-[PyTorch installation page](https://pytorch.org/get-started/locally/#start-locally) and/or 
+Please refer to [TensorFlow installation page](https://www.tensorflow.org/install/pip#tensorflow-2.0-rc-is-available),
+[PyTorch installation page](https://pytorch.org/get-started/locally/#start-locally) and/or
 [Flax installation page](https://github.com/google/flax#quick-install)
 regarding the specific install command for your platform.
 
@@ -73,7 +73,27 @@ It should download a pretrained model then print something like
 
 ## Installing from source
 
-To install from source, clone the repository and install with the following commands:
+Here is how to quickly install `transformers` from source:
+
+```bash
+pip install git+https://github.com/huggingface/transformers
+```
+
+Note that this will install not the latest released version, but the bleeding edge `master` version, which you may want to use in case a bug has been fixed since the last official release and a new release hasn't  been yet rolled out.
+
+While we strive to keep `master` operational at all times, if you notice some issues, they usually get fixed within a few hours or a day and and you're more than welcome to help us detect any problems by opening an [Issue](https://github.com/huggingface/transformers/issues) and this way, things will get fixed even sooner.
+
+Again, you can run:
+
+```bash
+python -c "from transformers import pipeline; print(pipeline('sentiment-analysis')('I hate you'))"
+```
+
+to check 🤗 Transformers is properly installed.
+
+## Editable install
+
+If you want to constantly use the bleeding edge `master` version of the source code, or if you want to contribute to the library and need to test the changes in the code you're making, you will need an editable install. This is done by cloning the repository and installing with the following commands:
 
 ``` bash
 git clone https://github.com/huggingface/transformers.git
@@ -81,13 +101,22 @@ cd transformers
 pip install -e .
 ```
 
-Again, you can run 
+This command performs a magical link between the folder you cloned the repository to and your python library paths, and it'll look inside this folder in addition to the normal library-wide paths. So if normally your python packages get installed into:
+```
+~/anaconda3/envs/main/lib/python3.7/site-packages/
+```
+now this editable install will reside where you clone the folder to, e.g. `~/transformers/` and python will search it too.
 
-```bash
-python -c "from transformers import pipeline; print(pipeline('sentiment-analysis')('I hate you'))"
+Do note that you have to keep that `transformers` folder around and not delete it to continue using the  `transfomers` library.
+
+Now, let's get to the real benefit of this installation approach. Say, you saw some new feature has been just committed into `master`. If you have already performed all the steps above, to update your transformers to include all the latest commits, all you need to do is to `cd` into that cloned repository folder and update the clone to the latest version:
+
+```
+cd ~/transformers/
+git pull
 ```
 
-to check 🤗 Transformers is properly installed.
+There is nothing else to do. Your python environment will find the bleeding edge version of `transformers` on the next run.
 
 
 ## With conda
@@ -100,7 +129,7 @@ Since Transformers version v4.0.0, we now have a conda channel: `huggingface`.
 conda install -c huggingface transformers
 ```
 
-Follow the installation pages of TensorFlow, PyTorch or Flax to see how to install them with conda. 
+Follow the installation pages of TensorFlow, PyTorch or Flax to see how to install them with conda.
 
 ## Caching models
 
@@ -109,7 +138,7 @@ This library provides pretrained models that will be downloaded and cached local
 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 ``HF_HOME`` 
+  * shell environment variable ``HF_HOME``
   * shell environment variable ``XDG_CACHE_HOME`` + ``/huggingface/``
   * default: ``~/.cache/huggingface/``
 
@@ -126,11 +155,36 @@ If you expect to be downloading large volumes of models (more than 1,000) from o
 your CI setup, or a large-scale production deployment), please cache the model files on your end. It will be way
 faster, and cheaper. Feel free to contact us privately if you need any help.
 
+### Offline mode
+
+It's possible to run 🤗 Transformers in a firewalled or a no-network environment.
+
+Setting environment variable `TRANSFORMERS_OFFLINE=1` will tell 🤗 Transformers to use local files only and will not try to look things up.
+
+Most likely you may want to couple this with `HF_DATASETS_OFFLINE=1` that performs the same for 🤗 Datasets if you're using the latter.
+
+Here is an example of how this can be used on a filesystem that is shared between a normally networked and a firewalled to the external world instances.
+
+On the instance with the normal network run your program which will download and cache models (and optionally datasets if you use 🤗 Datasets). For example:
+
+```
+python examples/seq2seq/run_translation.py --model_name_or_path t5-small --dataset_name wmt16 --dataset_config ro-en ...
+```
+
+and then with the same filesystem you can now run the same program on a firewalled instance:
+```
+HF_DATASETS_OFFLINE=1 TRANSFORMERS_OFFLINE=1 \
+python examples/seq2seq/run_translation.py --model_name_or_path t5-small --dataset_name wmt16 --dataset_config ro-en ...
+```
+and it should succeed without any hanging waiting to timeout.
+
+
+
 ## Do you want to run a Transformer model on a mobile device?
 
 You should check out our [swift-coreml-transformers](https://github.com/huggingface/swift-coreml-transformers) repo.
 
-It contains a set of tools to convert PyTorch or TensorFlow 2.0 trained Transformer models (currently contains `GPT-2`, 
+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 pretraining or fine-tuning models in PyTorch or

docs/source/internal/file_utils.rst

@@ -0,0 +1,54 @@
+.. 
+    Copyright 2021 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+    specific language governing permissions and limitations under the License.
+
+General Utilities
+-----------------------------------------------------------------------------------------------------------------------
+
+This page lists all of Transformers general utility functions that are found in the file ``file_utils.py``.
+
+Most of those are only useful if you are studying the general code in the library.
+
+
+Enums and namedtuples
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.file_utils.ExplicitEnum
+
+.. autoclass:: transformers.file_utils.PaddingStrategy
+
+.. autoclass:: transformers.file_utils.TensorType
+
+
+Special Decorators
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autofunction:: transformers.file_utils.add_start_docstrings
+
+.. autofunction:: transformers.file_utils.add_start_docstrings_to_model_forward
+
+.. autofunction:: transformers.file_utils.add_end_docstrings
+
+.. autofunction:: transformers.file_utils.add_code_sample_docstrings
+
+.. autofunction:: transformers.file_utils.replace_return_docstrings
+
+
+Special Properties
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.file_utils.cached_property
+
+
+Other Utilities
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.file_utils._BaseLazyModule

docs/source/internal/generation_utils.rst

@@ -151,6 +151,23 @@ generation.
 .. autoclass:: transformers.HammingDiversityLogitsProcessor
     :members: __call__
 
+StoppingCriteria
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+A :class:`~transformers.StoppingCriteria` can be used to change when to stop generation (other than EOS token).
+
+.. autoclass:: transformers.StoppingCriteria
+    :members: __call__
+
+.. autoclass:: transformers.StoppingCriteriaList
+    :members: __call__
+
+.. autoclass:: transformers.MaxLengthCriteria
+    :members: __call__
+
+.. autoclass:: transformers.MaxTimeCriteria
+    :members: __call__
+
 BeamSearch
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

docs/source/internal/tokenization_utils.rst

@@ -38,12 +38,6 @@ SpecialTokensMixin
 Enums and namedtuples
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-.. autoclass:: transformers.tokenization_utils_base.ExplicitEnum
-
-.. autoclass:: transformers.tokenization_utils_base.PaddingStrategy
-
-.. autoclass:: transformers.tokenization_utils_base.TensorType
-
 .. autoclass:: transformers.tokenization_utils_base.TruncationStrategy
 
 .. autoclass:: transformers.tokenization_utils_base.CharSpan

docs/source/internal/trainer_utils.rst

@@ -22,7 +22,7 @@ Utilities
 
 .. autoclass:: transformers.EvalPrediction
 
-.. autoclass:: transformers.EvaluationStrategy
+.. autoclass:: transformers.IntervalStrategy
 
 .. autofunction:: transformers.set_seed
 

docs/source/main_classes/feature_extractor.rst

@@ -0,0 +1,41 @@
+.. 
+    Copyright 2021 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+    specific language governing permissions and limitations under the License.
+
+
+Feature Extractor
+-----------------------------------------------------------------------------------------------------------------------
+
+A feature extractor is in charge of preparing input features for a multi-modal model. This includes feature extraction
+from sequences, *e.g.*, pre-processing audio files to Log-Mel Spectrogram features, feature extraction from images
+*e.g.* cropping image image files, but also padding, normalization, and conversion to Numpy, PyTorch, and TensorFlow
+tensors.
+
+
+FeatureExtractionMixin
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.feature_extraction_utils.FeatureExtractionMixin
+    :members: from_pretrained, save_pretrained
+
+
+SequenceFeatureExtractor
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.SequenceFeatureExtractor
+    :members: pad
+
+
+BatchFeature
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.BatchFeature
+    :members:

docs/source/main_classes/logging.rst

@@ -65,6 +65,10 @@ Other functions
 
 .. autofunction:: transformers.logging.get_logger
 
+.. autofunction:: transformers.logging.enable_default_handler
+
+.. autofunction:: transformers.logging.disable_default_handler
+
 .. autofunction:: transformers.logging.enable_explicit_format
 
 .. autofunction:: transformers.logging.reset_format

docs/source/main_classes/output.rst

@@ -60,7 +60,7 @@ ModelOutput
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 .. autoclass:: transformers.file_utils.ModelOutput
-    :members:
+    :members: to_tuple
 
 
 BaseModelOutput

docs/source/main_classes/tokenizer.rst

@@ -54,7 +54,9 @@ PreTrainedTokenizer
 
 .. autoclass:: transformers.PreTrainedTokenizer
     :special-members: __call__
-    :members:
+    :members: batch_decode, convert_ids_to_tokens, convert_tokens_to_ids, convert_tokens_to_string, decode, encode, 
+        get_added_vocab, get_special_tokens_mask, num_special_tokens_to_add, prepare_for_tokenization, tokenize,
+        vocab_size
 
 
 PreTrainedTokenizerFast
@@ -62,7 +64,9 @@ PreTrainedTokenizerFast
 
 .. autoclass:: transformers.PreTrainedTokenizerFast
     :special-members: __call__
-    :members:
+    :members: batch_decode, convert_ids_to_tokens, convert_tokens_to_ids, convert_tokens_to_string, decode, encode, 
+        get_added_vocab, get_special_tokens_mask, num_special_tokens_to_add,
+        set_truncation_and_padding,tokenize, vocab_size
 
 
 BatchEncoding

docs/source/main_classes/trainer.rst

@@ -21,16 +21,16 @@ Before instantiating your :class:`~transformers.Trainer`/:class:`~transformers.T
 customization during training.
 
 The API supports distributed training on multiple GPUs/TPUs, mixed precision through `NVIDIA Apex
-<https://github.com/NVIDIA/apex>`__ for PyTorch and :obj:`tf.keras.mixed_precision` for TensorFlow.
+<https://github.com/NVIDIA/apex>`__ and Native AMP for PyTorch and :obj:`tf.keras.mixed_precision` for TensorFlow.
 
-Both :class:`~transformers.Trainer` and :class:`~transformers.TFTrainer` contain the basic training loop supporting the
-previous features. To inject custom behavior you can subclass them and override the following methods:
+Both :class:`~transformers.Trainer` and :class:`~transformers.TFTrainer` contain the basic training loop which supports
+the above features. To inject custom behavior you can subclass them and override the following methods:
 
 - **get_train_dataloader**/**get_train_tfdataset** -- Creates the training DataLoader (PyTorch) or TF Dataset.
 - **get_eval_dataloader**/**get_eval_tfdataset** -- Creates the evaluation DataLoader (PyTorch) or TF Dataset.
 - **get_test_dataloader**/**get_test_tfdataset** -- Creates the test DataLoader (PyTorch) or TF Dataset.
 - **log** -- Logs information on the various objects watching training.
-- **create_optimizer_and_scheduler** -- Setups the optimizer and learning rate scheduler if they were not passed at
+- **create_optimizer_and_scheduler** -- Sets up the optimizer and learning rate scheduler if they were not passed at
   init.
 - **compute_loss** - Computes the loss on a batch of training inputs.
 - **training_step** -- Performs a training step.
@@ -39,17 +39,35 @@ previous features. To inject custom behavior you can subclass them and override
 - **evaluate** -- Runs an evaluation loop and returns metrics.
 - **predict** -- Returns predictions (with metrics if labels are available) on a test set.
 
-Here is an example of how to customize :class:`~transformers.Trainer` using a custom loss function:
+.. warning::
+
+    The :class:`~transformers.Trainer` class is optimized for 🤗 Transformers models and can have surprising behaviors
+    when you use it on other models. When using it on your own model, make sure:
+
+    - your model always return tuples or subclasses of :class:`~transformers.file_utils.ModelOutput`.
+    - your model can compute the loss if a :obj:`labels` argument is provided and that loss is returned as the first
+      element of the tuple (if your model returns tuples)
+    - your model can accept multiple label arguments (use the :obj:`label_names` in your
+      :class:`~transformers.TrainingArguments` to indicate their name to the :class:`~transformers.Trainer`) but none
+      of them should be named :obj:`"label"`.
+
+Here is an example of how to customize :class:`~transformers.Trainer` using a custom loss function for multi-label
+classification:
 
 .. code-block:: python
 
+    import torch
     from transformers import Trainer
-    class MyTrainer(Trainer):
-        def compute_loss(self, model, inputs):
+
+    class MultilabelTrainer(Trainer):
+        def compute_loss(self, model, inputs, return_outputs=False):
             labels = inputs.pop("labels")
             outputs = model(**inputs)
-            logits = outputs[0]
-            return my_custom_loss(logits, labels)
+            logits = outputs.logits
+            loss_fct = torch.nn.BCEWithLogitsLoss()
+            loss = loss_fct(logits.view(-1, self.model.config.num_labels),
+                            labels.float().view(-1, self.model.config.num_labels))
+            return (loss, outputs) if return_outputs else loss
 
 Another way to customize the training loop behavior for the PyTorch :class:`~transformers.Trainer` is to use
 :doc:`callbacks <callback>` that can inspect the training loop state (for progress reporting, logging on TensorBoard or
@@ -113,7 +131,125 @@ Toward Training Trillion Parameter Models, by Samyam Rajbhandari, Jeff Rasley, O
 
 This provided support is new and experimental as of this writing.
 
-You will need at least 2 GPUs to benefit from these features.
+Installation Notes
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+As of this writing, both FairScale and Deepspeed require compilation of CUDA C++ code, before they can be used.
+
+While all installation issues should be dealt with through the corresponding GitHub Issues of `FairScale
+<https://github.com/facebookresearch/fairscale/issues>`__ and `Deepspeed
+<https://github.com/microsoft/DeepSpeed/issues>`__, there are a few common issues that one may encounter while building
+any PyTorch extension that needs to build CUDA extensions.
+
+Therefore, if you encounter a CUDA-related build issue while doing one of the following or both:
+
+.. code-block:: bash
+
+    pip install fairscale
+    pip install deepspeed
+
+please, read the following notes first.
+
+In these notes we give examples for what to do when ``pytorch`` has been built with CUDA ``10.2``. If your situation is
+different remember to adjust the version number to the one you are after.
+
+**Possible problem #1:**
+
+While, Pytorch comes with its own CUDA toolkit, to build these two projects you must have an identical version of CUDA
+installed system-wide.
+
+For example, if you installed ``pytorch`` with ``cudatoolkit==10.2`` in the Python environment, you also need to have
+CUDA ``10.2`` installed system-wide.
+
+The exact location may vary from system to system, but ``/usr/local/cuda-10.2`` is the most common location on many
+Unix systems. When CUDA is correctly set up and added to the ``PATH`` environment variable, one can find the
+installation location by doing:
+
+.. code-block:: bash
+
+    which nvcc
+
+If you don't have CUDA installed system-wide, install it first. You will find the instructions by using your favorite
+search engine. For example, if you're on Ubuntu you may want to search for: `ubuntu cuda 10.2 install
+<https://www.google.com/search?q=ubuntu+cuda+10.2+install>`__.
+
+**Possible problem #2:**
+
+Another possible common problem is that you may have more than one CUDA toolkit installed system-wide. For example you
+may have:
+
+.. code-block:: bash
+
+    /usr/local/cuda-10.2
+    /usr/local/cuda-11.0
+
+Now, in this situation you need to make sure that your ``PATH`` and ``LD_LIBRARY_PATH`` environment variables contain
+the correct paths to the desired CUDA version. Typically, package installers will set these to contain whatever the
+last version was installed. If you encounter the problem, where the package build fails because it can't find the right
+CUDA version despite you having it installed system-wide, it means that you need to adjust the 2 aforementioned
+environment variables.
+
+First, you may look at their contents:
+
+.. code-block:: bash
+
+    echo $PATH
+    echo $LD_LIBRARY_PATH
+
+so you get an idea of what is inside.
+
+It's possible that ``LD_LIBRARY_PATH`` is empty.
+
+``PATH`` lists the locations of where executables can be found and ``LD_LIBRARY_PATH`` is for where shared libraries
+are to looked for. In both cases, earlier entries have priority over the later ones. ``:`` is used to separate multiple
+entries.
+
+Now, to tell the build program where to find the specific CUDA toolkit, insert the desired paths to be listed first by
+doing:
+
+.. code-block:: bash
+
+    export PATH=/usr/local/cuda-10.2/bin:$PATH
+    export LD_LIBRARY_PATH=/usr/local/cuda-10.2/lib64:$LD_LIBRARY_PATH
+
+Note that we aren't overwriting the existing values, but prepending instead.
+
+Of course, adjust the version number, the full path if need be. Check that the directories you assign actually do
+exist. ``lib64`` sub-directory is where the various CUDA ``.so`` objects, like ``libcudart.so`` reside, it's unlikely
+that your system will have it named differently, but if it is adjust it to reflect your reality.
+
+
+**Possible problem #3:**
+
+Some older CUDA versions may refuse to build with newer compilers. For example, you my have ``gcc-9`` but it wants
+``gcc-7``.
+
+There are various ways to go about it.
+
+If you can install the latest CUDA toolkit it typically should support the newer compiler.
+
+Alternatively, you could install the lower version of the compiler in addition to the one you already have, or you may
+already have it but it's not the default one, so the build system can't see it. If you have ``gcc-7`` installed but the
+build system complains it can't find it, the following might do the trick:
+
+.. code-block:: bash
+
+    sudo ln -s /usr/bin/gcc-7  /usr/local/cuda-10.2/bin/gcc
+    sudo ln -s /usr/bin/g++-7  /usr/local/cuda-10.2/bin/g++
+
+
+Here, we are making a symlink to ``gcc-7`` from ``/usr/local/cuda-10.2/bin/gcc`` and since
+``/usr/local/cuda-10.2/bin/`` should be in the ``PATH`` environment variable (see the previous problem's solution), it
+should find ``gcc-7`` (and ``g++7``) and then the build will succeed.
+
+As always make sure to edit the paths in the example to match your situation.
+
+**If still unsuccessful:**
+
+If after addressing these you still encounter build issues, please, proceed with the GitHub Issue of `FairScale
+<https://github.com/facebookresearch/fairscale/issues>`__ and `Deepspeed
+<https://github.com/microsoft/DeepSpeed/issues>`__, depending on the project you have the problem with.
+
 
 FairScale
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
@@ -123,6 +259,10 @@ provides support for the following features from `the ZeRO paper <https://arxiv.
 
 1. Optimizer State Sharding
 2. Gradient Sharding
+3. Model Parameters Sharding (new and very experimental)
+4. CPU offload (new and very experimental)
+
+You will need at least two GPUs to use this feature.
 
 To deploy this feature:
 
@@ -132,102 +272,346 @@ To deploy this feature:
 
        pip install fairscale
 
-   or find more details on `the FairScale's github page
+   or find more details on `the FairScale's GitHub page
    <https://github.com/facebookresearch/fairscale/#installation>`__.
 
-2. Add ``--sharded_ddp`` to the command line arguments, and make sure you have added the distributed launcher ``-m
-   torch.distributed.launch --nproc_per_node=NUMBER_OF_GPUS_YOU_HAVE`` if you haven't been using it already.
+2. To use the first version of Sharded data-parallelism, add ``--sharded_ddp simple`` to the command line arguments,
+   and make sure you have added the distributed launcher ``-m torch.distributed.launch
+   --nproc_per_node=NUMBER_OF_GPUS_YOU_HAVE`` if you haven't been using it already.
 
-For example here is how you could use it for ``finetune_trainer.py`` with 2 GPUs:
+For example here is how you could use it for ``run_translation.py`` with 2 GPUs:
 
 .. code-block:: bash
 
-    cd examples/seq2seq
-    python -m torch.distributed.launch --nproc_per_node=2 ./finetune_trainer.py \
-    --model_name_or_path sshleifer/distill-mbart-en-ro-12-4 --data_dir wmt_en_ro \
+    python -m torch.distributed.launch --nproc_per_node=2 examples/seq2seq/run_translation.py \
+    --model_name_or_path t5-small --per_device_train_batch_size 1   \
     --output_dir output_dir --overwrite_output_dir \
-    --do_train --n_train 500 --num_train_epochs 1 \
-    --per_device_train_batch_size 1  --freeze_embeds \
-    --src_lang en_XX --tgt_lang ro_RO --task translation \
-    --fp16 --sharded_ddp
+    --do_train --max_train_samples 500 --num_train_epochs 1 \
+    --dataset_name wmt16 --dataset_config "ro-en" \
+    --source_lang en --target_lang ro \
+    --fp16 --sharded_ddp simple
 
 Notes:
 
 - This feature requires distributed training (so multiple GPUs).
 - It is not implemented for TPUs.
 - It works with ``--fp16`` too, to make things even faster.
-- One of the main benefits of enabling ``--sharded_ddp`` is that it uses a lot less GPU memory, so you should be able
-  to use significantly larger batch sizes using the same hardware (e.g. 3x and even bigger) which should lead to
+- One of the main benefits of enabling ``--sharded_ddp simple`` is that it uses a lot less GPU memory, so you should be
+  able to use significantly larger batch sizes using the same hardware (e.g. 3x and even bigger) which should lead to
   significantly shorter training time.
 
+3. To use the second version of Sharded data-parallelism, add ``--sharded_ddp zero_dp_2`` or ``--sharded_ddp zero_dp_3`
+   to the command line arguments, and make sure you have added the distributed launcher ``-m torch.distributed.launch
+   --nproc_per_node=NUMBER_OF_GPUS_YOU_HAVE`` if you haven't been using it already.
+
+For example here is how you could use it for ``run_translation.py`` with 2 GPUs:
+
+.. code-block:: bash
+
+    python -m torch.distributed.launch --nproc_per_node=2 examples/seq2seq/run_translation.py \
+    --model_name_or_path t5-small --per_device_train_batch_size 1   \
+    --output_dir output_dir --overwrite_output_dir \
+    --do_train --max_train_samples 500 --num_train_epochs 1 \
+    --dataset_name wmt16 --dataset_config "ro-en" \
+    --source_lang en --target_lang ro \
+    --fp16 --sharded_ddp zero_dp_2
+
+:obj:`zero_dp_2` is an optimized version of the simple wrapper, while :obj:`zero_dp_3` fully shards model weights,
+gradients and optimizer states.
+
+Both are compatible with adding :obj:`cpu_offload` to enable ZeRO-offload (activate it like this: :obj:`--sharded_ddp
+"zero_dp_2 cpu_offload"`).
+
+Notes:
+
+- This feature requires distributed training (so multiple GPUs).
+- It is not implemented for TPUs.
+- It works with ``--fp16`` too, to make things even faster.
+- The ``cpu_offload`` additional option requires ``--fp16``.
+- This is an area of active development, so make sure you have a source install of fairscale to use this feature as
+  some bugs you encounter may have been fixed there already.
+
+Known caveats:
+
+- This feature is incompatible with :obj:`--predict_with_generate` in the `run_translation.py` script.
+- Using :obj:`--sharded_ddp zero_dp_3` requires wrapping each layer of the model in the special container
+  :obj:`FullyShardedDataParallelism` of fairscale. It should be used with the option :obj:`auto_wrap` if you are not
+  doing this yourself: :obj:`--sharded_ddp "zero_dp_3 auto_wrap"`.
+
 
 DeepSpeed
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
-
 `DeepSpeed <https://github.com/microsoft/DeepSpeed>`__ implements everything described in the `ZeRO paper
 <https://arxiv.org/abs/1910.02054>`__, except ZeRO's stage 3. "Parameter Partitioning (Pos+g+p)". Currently it provides
 full support for:
 
 1. Optimizer State Partitioning (ZeRO stage 1)
 2. Add Gradient Partitioning (ZeRO stage 2)
+3. Custom fp16 handling
+4. A range of fast Cuda-extension-based Optimizers
+5. ZeRO-Offload
 
-To deploy this feature:
+ZeRO-Offload has its own dedicated paper: `ZeRO-Offload: Democratizing Billion-Scale Model Training
+<https://arxiv.org/abs/2101.06840>`__.
 
-1. Install the library via pypi:
+DeepSpeed is currently used only for training, as all the currently available features are of no use to inference.
+
+
+
+Installation
+=======================================================================================================================
+
+Install the library via pypi:
+
+.. code-block:: bash
+
+    pip install deepspeed
+
+or find more details on `the DeepSpeed's GitHub page <https://github.com/microsoft/deepspeed#installation>`__.
+
+Deployment with multiple GPUs
+=======================================================================================================================
+
+To deploy this feature with multiple GPUs adjust the :class:`~transformers.Trainer` command line arguments as
+following:
+
+1. replace ``python -m torch.distributed.launch`` with ``deepspeed``.
+2. add a new argument ``--deepspeed ds_config.json``, where ``ds_config.json`` is the DeepSpeed configuration file as
+   documented `here <https://www.deepspeed.ai/docs/config-json/>`__. The file naming is up to you.
+
+Therefore, if your original command line looked as following:
+
+.. code-block:: bash
+
+    python -m torch.distributed.launch --nproc_per_node=2 your_program.py <normal cl args>
+
+Now it should be:
+
+.. code-block:: bash
+
+    deepspeed --num_gpus=2 your_program.py <normal cl args> --deepspeed ds_config.json
+
+Unlike, ``torch.distributed.launch`` where you have to specify how many GPUs to use with ``--nproc_per_node``, with the
+``deepspeed`` launcher you don't have to use the corresponding ``--num_gpus`` if you want all of your GPUs used. The
+full details on how to configure various nodes and GPUs can be found `here
+<https://www.deepspeed.ai/getting-started/#resource-configuration-multi-node>`__.
+
+In fact, you can continue using ``-m torch.distributed.launch`` with DeepSpeed as long as you don't need to use
+``deepspeed`` launcher-specific arguments. Typically if you don't need a multi-node setup you're not required to use
+the ``deepspeed`` launcher. But since in the DeepSpeed documentation it'll be used everywhere, for consistency we will
+use it here as well.
+
+Here is an example of running ``run_translation.py`` under DeepSpeed deploying all available GPUs:
+
+.. code-block:: bash
+
+    deepspeed examples/seq2seq/run_translation.py \
+    --deepspeed examples/tests/deepspeed/ds_config.json \
+    --model_name_or_path t5-small --per_device_train_batch_size 1   \
+    --output_dir output_dir --overwrite_output_dir --fp16 \
+    --do_train --max_train_samples 500 --num_train_epochs 1 \
+    --dataset_name wmt16 --dataset_config "ro-en" \
+    --source_lang en --target_lang ro
+
+
+Note that in the DeepSpeed documentation you are likely to see ``--deepspeed --deepspeed_config ds_config.json`` - i.e.
+two DeepSpeed-related arguments, but for the sake of simplicity, and since there are already so many arguments to deal
+with, we combined the two into a single argument.
+
+For some practical usage examples, please, see this `post
+<https://github.com/huggingface/transformers/issues/8771#issuecomment-759248400>`__.
+
+
+
+Deployment with one GPU
+=======================================================================================================================
+
+To deploy DeepSpeed with one GPU adjust the :class:`~transformers.Trainer` command line arguments as following:
+
+.. code-block:: bash
+
+    deepspeed --num_gpus=1 examples/seq2seq/run_translation.py \
+    --deepspeed examples/tests/deepspeed/ds_config.json \
+    --model_name_or_path t5-small --per_device_train_batch_size 1   \
+    --output_dir output_dir --overwrite_output_dir --fp16 \
+    --do_train --max_train_samples 500 --num_train_epochs 1 \
+    --dataset_name wmt16 --dataset_config "ro-en" \
+    --source_lang en --target_lang ro
+
+This is almost the same as with multiple-GPUs, but here we tell DeepSpeed explicitly to use just one GPU. By default,
+DeepSpeed deploys all GPUs it can see. If you have only 1 GPU to start with, then you don't need this argument. The
+following `documentation <https://www.deepspeed.ai/getting-started/#resource-configuration-multi-node>`__ discusses the
+launcher options.
+
+Why would you want to use DeepSpeed with just one GPU?
+
+1. It has a ZeRO-offload feature which can delegate some computations and memory to the host's CPU and RAM, and thus
+   leave more GPU resources for model's needs - e.g. larger batch size, or enabling a fitting of a very big model which
+   normally won't fit.
+2. It provides a smart GPU memory management system, that minimizes memory fragmentation, which again allows you to fit
+   bigger models and data batches.
+
+While we are going to discuss the configuration in details next, the key to getting a huge improvement on a single GPU
+with DeepSpeed is to have at least the following configuration in the configuration file:
+
+.. code-block:: json
+
+  {
+    "zero_optimization": {
+       "stage": 2,
+       "allgather_partitions": true,
+       "allgather_bucket_size": 2e8,
+       "reduce_scatter": true,
+       "reduce_bucket_size": 2e8,
+       "overlap_comm": true,
+       "contiguous_gradients": true,
+       "cpu_offload": true
+    },
+  }
+
+which enables ``cpu_offload`` and some other important features. You may experiment with the buffer sizes, you will
+find more details in the discussion below.
+
+For a practical usage example of this type of deployment, please, see this `post
+<https://github.com/huggingface/transformers/issues/8771#issuecomment-759176685>`__.
+
+Notes:
+
+- if you need to run on a specific GPU, which is different from GPU 0, you can't use ``CUDA_VISIBLE_DEVICES`` to limit
+  the visible scope of available GPUs. Instead, you have to use the following syntax:
 
    .. code-block:: bash
 
-       pip install deepspeed
+       deepspeed --include localhost:1 examples/seq2seq/run_translation.py ...
 
-   or find more details on `the DeepSpeed's github page <https://github.com/microsoft/deepspeed#installation>`__.
+   In this example, we tell DeepSpeed to use GPU 1 (second gpu).
 
-2. Adjust the :class:`~transformers.Trainer` command line arguments as following:
 
-   1. replace ``python -m torch.distributed.launch`` with ``deepspeed``.
-   2. add a new argument ``--deepspeed ds_config.json``, where ``ds_config.json`` is the DeepSpeed configuration file
-      as documented `here <https://www.deepspeed.ai/docs/config-json/>`__. The file naming is up to you.
 
-   Therefore, if your original command line looked as following:
+Deployment in Notebooks
+=======================================================================================================================
 
-   .. code-block:: bash
+The problem with running notebook cells as a script is that there is no normal ``deepspeed`` launcher to rely on, so
+under certain setups we have to emulate it.
 
-       python -m torch.distributed.launch --nproc_per_node=2 your_program.py <normal cl args>
+Here is how you'd have to adjust your training code in the notebook to use DeepSpeed.
 
-   Now it should be:
+.. code-block:: python
 
-   .. code-block:: bash
+    # DeepSpeed requires a distributed environment even when only one process is used.
+    # This emulates a launcher in the notebook
+    import os
+    os.environ['MASTER_ADDR'] = 'localhost'
+    os.environ['MASTER_PORT'] = '9994' # modify if RuntimeError: Address already in use
+    os.environ['RANK'] = "0"
+    os.environ['LOCAL_RANK'] = "0"
+    os.environ['WORLD_SIZE'] = "1"
 
-       deepspeed --num_gpus=2 your_program.py <normal cl args> --deepspeed ds_config.json
+    # Now proceed as normal, plus pass the deepspeed config file
+    training_args = TrainingArguments(..., deepspeed="ds_config.json")
+    trainer = Trainer(...)
+    trainer.train()
 
-   Unlike, ``torch.distributed.launch`` where you have to specify how many GPUs to use with ``--nproc_per_node``, with
-   the ``deepspeed`` launcher you don't have to use the corresponding ``--num_gpus`` if you want all of your GPUs used.
-   The full details on how to configure various nodes and GPUs can be found `here
-   <https://www.deepspeed.ai/getting-started/#resource-configuration-multi-node>`__.
+Note: `...` stands for the normal arguments that you'd pass to the functions.
 
-   Here is an example of running ``finetune_trainer.py`` under DeepSpeed deploying all available GPUs:
+If you want to create the config file on the fly in the notebook in the current directory, you could have a dedicated
+cell with:
 
-   .. code-block:: bash
+.. code-block:: python
 
-       cd examples/seq2seq
-       deepspeed ./finetune_trainer.py --deepspeed ds_config.json \
-       --model_name_or_path sshleifer/distill-mbart-en-ro-12-4 --data_dir wmt_en_ro \
-       --output_dir output_dir --overwrite_output_dir \
-       --do_train --n_train 500 --num_train_epochs 1 \
-       --per_device_train_batch_size 1  --freeze_embeds \
-       --src_lang en_XX --tgt_lang ro_RO --task translation
+    %%bash
+    cat <<'EOT' > ds_config.json
+    {
+        "fp16": {
+            "enabled": true,
+            "loss_scale": 0,
+            "loss_scale_window": 1000,
+            "hysteresis": 2,
+            "min_loss_scale": 1
+        },
 
-   Note that in the DeepSpeed documentation you are likely to see ``--deepspeed --deepspeed_config ds_config.json`` -
-   i.e. two DeepSpeed-related arguments, but for the sake of simplicity, and since there are already so many arguments
-   to deal with, we combined the two into a single argument.
+        "zero_optimization": {
+            "stage": 2,
+            "allgather_partitions": true,
+            "allgather_bucket_size": 2e8,
+            "overlap_comm": true,
+            "reduce_scatter": true,
+            "reduce_bucket_size": 2e8,
+            "contiguous_gradients": true,
+            "cpu_offload": true
+        },
 
-Before you can deploy DeepSpeed, let's discuss its configuration.
+        "zero_allow_untested_optimizer": true,
 
-**Configuration:**
+        "optimizer": {
+            "type": "AdamW",
+            "params": {
+                "lr": 3e-5,
+                "betas": [0.8, 0.999],
+                "eps": 1e-8,
+                "weight_decay": 3e-7
+            }
+        },
+
+        "scheduler": {
+            "type": "WarmupLR",
+            "params": {
+                "warmup_min_lr": 0,
+                "warmup_max_lr": 3e-5,
+                "warmup_num_steps": 500
+            }
+        },
+
+        "steps_per_print": 2000,
+        "wall_clock_breakdown": false
+    }
+    EOT
+
+
+That's said if the script is not in the notebook cells, you can launch ``deepspeed`` normally via shell from a cell
+with:
+
+.. code-block::
+
+   !deepspeed examples/seq2seq/run_translation.py ...
+
+or with bash magic, where you can write a multi-line code for the shell to run:
+
+.. code-block::
+
+   %%bash
+
+   cd /somewhere
+   deepspeed examples/seq2seq/run_translation.py ...
+
+
+
+
+Configuration
+=======================================================================================================================
 
 For the complete guide to the DeepSpeed configuration options that can be used in its configuration file please refer
 to the `following documentation <https://www.deepspeed.ai/docs/config-json/>`__.
 
+You can find dozens of DeepSpeed configuration examples that address various practical needs in `the DeepSpeedExamples
+repo <https://github.com/microsoft/DeepSpeedExamples>`__:
+
+.. code-block:: bash
+
+  git clone https://github.com/microsoft/DeepSpeedExamples
+  cd DeepSpeedExamples
+  find . -name '*json'
+
+Continuing the code from above, let's say you're looking to configure the Lamb optimizer. So you can search through the
+example ``.json`` files with:
+
+.. code-block:: bash
+
+  grep -i Lamb $(find . -name '*json')
+
+Some more examples are to be found in the `main repo <https://github.com/microsoft/DeepSpeed>`__ as well.
+
 While you always have to supply the DeepSpeed configuration file, you can configure the DeepSpeed integration in
 several ways:
 
@@ -271,7 +655,6 @@ enables FP16, uses AdamW optimizer and WarmupLR scheduler:
            "weight_decay": 3e-7
          }
        },
-       "zero_allow_untested_optimizer": true,
 
        "scheduler": {
          "type": "WarmupLR",
@@ -314,7 +697,8 @@ to achieve the same configuration as provided by the longer json file in the fir
 When you execute the program, DeepSpeed will log the configuration it received from the :class:`~transformers.Trainer`
 to the console, so you can see exactly what the final configuration was passed to it.
 
-**Shared Configuration:**
+Shared Configuration
+=======================================================================================================================
 
 Some configuration information is required by both the :class:`~transformers.Trainer` and DeepSpeed to function
 correctly, therefore, to prevent conflicting definitions, which could lead to hard to detect errors, we chose to
@@ -338,7 +722,8 @@ Of course, you will need to adjust the values in this example to your situation.
 
 
 
-**ZeRO:**
+ZeRO
+=======================================================================================================================
 
 The ``zero_optimization`` section of the configuration file is the most important part (`docs
 <https://www.deepspeed.ai/docs/config-json/#zero-optimizations-for-fp16-training>`__), since that is where you define
@@ -365,18 +750,23 @@ Notes:
 - ``"overlap_comm": true`` trades off increased GPU RAM usage to lower all-reduce latency. ``overlap_comm`` uses 4.5x
   the ``allgather_bucket_size`` and ``reduce_bucket_size`` values. So if they are set to 5e8, this requires a 9GB
   footprint (``5e8 x 2Bytes x 2 x 4.5``). Therefore, if you have a GPU with 8GB or less RAM, to avoid getting
-  OOM-errors you will need to reduce those parameters to about ``2e8``, which would require 3.6GB.
+  OOM-errors you will need to reduce those parameters to about ``2e8``, which would require 3.6GB. You will want to do
+  the same on larger capacity GPU as well, if you're starting to hit OOM.
+- when reducing these buffers you're trading communication speed to avail more GPU RAM. The smaller the buffer size,
+  the slower the communication, and the more GPU RAM will be available to other tasks. So if a bigger batch size is
+  important, getting a slightly slower training time could be a good trade.
 
 This section has to be configured exclusively via DeepSpeed configuration - the :class:`~transformers.Trainer` provides
 no equivalent command line arguments.
 
 
 
-**Optimizer:**
+Optimizer
+=======================================================================================================================
 
 
-DeepSpeed's main optimizers are Adam, OneBitAdam, and Lamb. These have been thoroughly tested with ZeRO and are thus
-recommended to be used. It, however, can import other optimizers from ``torch``. The full documentation is `here
+DeepSpeed's main optimizers are Adam, AdamW, OneBitAdam, and Lamb. These have been thoroughly tested with ZeRO and are
+thus recommended to be used. It, however, can import other optimizers from ``torch``. The full documentation is `here
 <https://www.deepspeed.ai/docs/config-json/#optimizer-parameters>`__.
 
 If you don't configure the ``optimizer`` entry in the configuration file, the :class:`~transformers.Trainer` will
@@ -388,7 +778,6 @@ Here is an example of the pre-configured ``optimizer`` entry for AdamW:
 .. code-block:: json
 
     {
-       "zero_allow_untested_optimizer": true,
        "optimizer": {
            "type": "AdamW",
            "params": {
@@ -400,14 +789,15 @@ Here is an example of the pre-configured ``optimizer`` entry for AdamW:
          }
     }
 
-Since AdamW isn't on the list of tested with DeepSpeed/ZeRO optimizers, we have to add
-``zero_allow_untested_optimizer`` flag.
+If you want to use another optimizer which is not listed above, you will have to add ``"zero_allow_untested_optimizer":
+true`` to the top level configuration.
 
 If you want to use one of the officially supported optimizers, configure them explicitly in the configuration file, and
 make sure to adjust the values. e.g. if use Adam you will want ``weight_decay`` around ``0.01``.
 
 
-**Scheduler:**
+Scheduler
+=======================================================================================================================
 
 DeepSpeed supports LRRangeTest, OneCycle, WarmupLR and WarmupDecayLR LR schedulers. The full documentation is `here
 <https://www.deepspeed.ai/docs/config-json/#scheduler-parameters>`__.
@@ -456,7 +846,8 @@ Here is an example of the pre-configured ``scheduler`` entry for WarmupLR (``con
          }
     }
 
-**Automatic Mixed Precision:**
+Automatic Mixed Precision
+=======================================================================================================================
 
 You can work with FP16 in one of the following ways:
 
@@ -464,7 +855,7 @@ You can work with FP16 in one of the following ways:
 2. NVIDIA's apex, as documented `here
    <https://www.deepspeed.ai/docs/config-json/#automatic-mixed-precision-amp-training-options>`__.
 
-If you want to use an equivalent of the pytorch native amp, you can either configure the ``fp16`` entry in the
+If you want to use an equivalent of the Pytorch native amp, you can either configure the ``fp16`` entry in the
 configuration file, or use the following command line arguments: ``--fp16 --fp16_backend amp``.
 
 Here is an example of the ``fp16`` configuration:
@@ -496,8 +887,31 @@ Here is an example of the ``amp`` configuration:
     }
 
 
+Gradient Accumulation
+=======================================================================================================================
 
-**Gradient Clipping:**
+While normally DeepSpeed gets gradient accumulation configured with:
+
+.. code-block:: json
+
+    {
+        "gradient_accumulation_steps": 3,
+    }
+
+in this case, to enable gradient accumulation, pass the command line `--gradient_accumulation_steps` argument as normal
+and it will get injected into the DeepSpeed configuration.
+
+If you try to add it directly to the configuration file, you will receive an error from the Trainer - this is because
+this setting is needed by the Trainer too, and so this approach ensures that there is a single way of setting this
+value and thus avoid potential subtle errors.
+
+
+
+
+
+
+Gradient Clipping
+=======================================================================================================================
 
 If you don't configure the ``gradient_clipping`` entry in the configuration file, the :class:`~transformers.Trainer`
 will use the value of the ``--max_grad_norm`` command line argument to set it.
@@ -512,7 +926,8 @@ Here is an example of the ``gradient_clipping`` configuration:
 
 
 
-**Notes:**
+Notes
+=======================================================================================================================
 
 * DeepSpeed works with the PyTorch :class:`~transformers.Trainer` but not TF :class:`~transformers.TFTrainer`.
 * While DeepSpeed has a pip installable PyPI package, it is highly recommended that it gets installed from `source
@@ -522,12 +937,19 @@ Here is an example of the ``gradient_clipping`` configuration:
   use any model with your own trainer, and you will have to adapt the latter according to `the DeepSpeed integration
   instructions <https://www.deepspeed.ai/getting-started/#writing-deepspeed-models>`__.
 
-**Main DeepSpeed Resources:**
+Main DeepSpeed Resources
+=======================================================================================================================
 
-- `github <https://github.com/microsoft/deepspeed>`__
+- `Project's github <https://github.com/microsoft/deepspeed>`__
 - `Usage docs <https://www.deepspeed.ai/getting-started/>`__
 - `API docs <https://deepspeed.readthedocs.io/en/latest/index.html>`__
+- `Blog posts <https://www.microsoft.com/en-us/research/search/?q=deepspeed>`__
+
+Papers:
+
+- `ZeRO: Memory Optimizations Toward Training Trillion Parameter Models <https://arxiv.org/abs/1910.02054>`__
+- `ZeRO-Offload: Democratizing Billion-Scale Model Training <https://arxiv.org/abs/2101.06840>`__
 
 Finally, please, remember that, HuggingFace :class:`~transformers.Trainer` only integrates DeepSpeed, therefore if you
-have any problems or questions with regards to DeepSpeed usage, please, file an issue with `DeepSpeed github
+have any problems or questions with regards to DeepSpeed usage, please, file an issue with `DeepSpeed GitHub
 <https://github.com/microsoft/DeepSpeed/issues>`__.

docs/source/model_doc/bart.rst

@@ -130,6 +130,12 @@ BartForQuestionAnswering
 .. autoclass:: transformers.BartForQuestionAnswering
     :members: forward
 
+BartForCausalLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.BartForCausalLM
+    :members: forward
+
 
 
 TFBartModel

docs/source/model_doc/blenderbot.rst

@@ -98,6 +98,13 @@ See :obj:`transformers.BartForConditionalGeneration` for arguments to `forward`
     :members: forward
 
 
+BlenderbotForCausalLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.BlenderbotForCausalLM
+    :members: forward
+
+
 TFBlenderbotModel
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

docs/source/model_doc/blenderbot_small.rst

@@ -70,6 +70,13 @@ BlenderbotSmallForConditionalGeneration
     :members: forward
 
 
+BlenderbotSmallForCausalLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.BlenderbotSmallForCausalLM
+    :members: forward
+
+
 TFBlenderbotSmallModel
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

docs/source/model_doc/bort.rst

@@ -0,0 +1,46 @@
+.. 
+    Copyright 2020 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+    specific language governing permissions and limitations under the License.
+
+BORT
+-----------------------------------------------------------------------------------------------------------------------
+
+Overview
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The BORT model was proposed in `Optimal Subarchitecture Extraction for BERT <https://arxiv.org/abs/2010.10499>`__ by
+Adrian de Wynter and Daniel J. Perry. It is an optimal subset of architectural parameters for the BERT, which the
+authors refer to as "Bort".
+
+The abstract from the paper is the following:
+
+*We extract an optimal subset of architectural parameters for the BERT architecture from Devlin et al. (2018) by
+applying recent breakthroughs in algorithms for neural architecture search. This optimal subset, which we refer to as
+"Bort", is demonstrably smaller, having an effective (that is, not counting the embedding layer) size of 5.5% the
+original BERT-large architecture, and 16% of the net size. Bort is also able to be pretrained in 288 GPU hours, which
+is 1.2% of the time required to pretrain the highest-performing BERT parametric architectural variant, RoBERTa-large
+(Liu et al., 2019), and about 33% of that of the world-record, in GPU hours, required to train BERT-large on the same
+hardware. It is also 7.9x faster on a CPU, as well as being better performing than other compressed variants of the
+architecture, and some of the non-compressed variants: it obtains performance improvements of between 0.3% and 31%,
+absolute, with respect to BERT-large, on multiple public natural language understanding (NLU) benchmarks.*
+
+Tips:
+
+- BORT's model architecture is based on BERT, so one can refer to :doc:`BERT's documentation page <bert>` for the
+  model's API as well as usage examples.
+- BORT uses the RoBERTa tokenizer instead of the BERT tokenizer, so one can refer to :doc:`RoBERTa's documentation page
+  <roberta>` for the tokenizer's API as well as usage examples.
+- BORT requires a specific fine-tuning algorithm, called `Agora
+  <https://adewynter.github.io/notes/bort_algorithms_and_applications.html#fine-tuning-with-algebraic-topology>`__ ,
+  that is sadly not open-sourced yet. It would be very useful for the community, if someone tries to implement the
+  algorithm to make BORT fine-tuning work.
+
+The original code can be found `here <https://github.com/alexa/bort/>`__.

docs/source/model_doc/convbert.rst

@@ -0,0 +1,144 @@
+.. 
+    Copyright 2020 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+    specific language governing permissions and limitations under the License.
+
+ConvBERT
+-----------------------------------------------------------------------------------------------------------------------
+
+Overview
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The ConvBERT model was proposed in `ConvBERT: Improving BERT with Span-based Dynamic Convolution
+<https://arxiv.org/abs/2008.02496>`__ by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng
+Yan.
+
+The abstract from the paper is the following:
+
+*Pre-trained language models like BERT and its variants have recently achieved impressive performance in various
+natural language understanding tasks. However, BERT heavily relies on the global self-attention block and thus suffers
+large memory footprint and computation cost. Although all its attention heads query on the whole input sequence for
+generating the attention map from a global perspective, we observe some heads only need to learn local dependencies,
+which means the existence of computation redundancy. We therefore propose a novel span-based dynamic convolution to
+replace these self-attention heads to directly model local dependencies. The novel convolution heads, together with the
+rest self-attention heads, form a new mixed attention block that is more efficient at both global and local context
+learning. We equip BERT with this mixed attention design and build a ConvBERT model. Experiments have shown that
+ConvBERT significantly outperforms BERT and its variants in various downstream tasks, with lower training cost and
+fewer model parameters. Remarkably, ConvBERTbase model achieves 86.4 GLUE score, 0.7 higher than ELECTRAbase, while
+using less than 1/4 training cost. Code and pre-trained models will be released.*
+
+ConvBERT training tips are similar to those of BERT. The original implementation can be found here:
+https://github.com/yitu-opensource/ConvBert
+
+ConvBertConfig
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.ConvBertConfig
+    :members:
+
+
+ConvBertTokenizer
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.ConvBertTokenizer
+    :members: build_inputs_with_special_tokens, get_special_tokens_mask,
+        create_token_type_ids_from_sequences, save_vocabulary
+
+
+ConvBertTokenizerFast
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.ConvBertTokenizerFast
+    :members: build_inputs_with_special_tokens, get_special_tokens_mask,
+        create_token_type_ids_from_sequences, save_vocabulary
+
+
+ConvBertModel
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.ConvBertModel
+    :members: forward
+
+
+ConvBertForMaskedLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.ConvBertForMaskedLM
+    :members: forward
+
+
+ConvBertForSequenceClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.ConvBertForSequenceClassification
+    :members: forward
+
+
+ConvBertForMultipleChoice
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.ConvBertForMultipleChoice
+    :members: forward
+
+
+ConvBertForTokenClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.ConvBertForTokenClassification
+    :members: forward
+
+
+ConvBertForQuestionAnswering
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.ConvBertForQuestionAnswering
+    :members: forward
+
+
+TFConvBertModel
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFConvBertModel
+    :members: call
+
+
+TFConvBertForMaskedLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFConvBertForMaskedLM
+    :members: call
+
+
+TFConvBertForSequenceClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFConvBertForSequenceClassification
+    :members: call
+
+
+TFConvBertForMultipleChoice
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFConvBertForMultipleChoice
+    :members: call
+
+
+TFConvBertForTokenClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFConvBertForTokenClassification
+    :members: call
+
+
+TFConvBertForQuestionAnswering
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFConvBertForQuestionAnswering
+    :members: call

docs/source/model_doc/deberta.rst

@@ -60,7 +60,7 @@ DebertaModel
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 .. autoclass:: transformers.DebertaModel
-    :members:
+    :members: forward
 
 
 DebertaPreTrainedModel
@@ -70,8 +70,29 @@ DebertaPreTrainedModel
     :members:
 
 
+DebertaForMaskedLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaForMaskedLM
+    :members: forward
+
+
 DebertaForSequenceClassification
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 .. autoclass:: transformers.DebertaForSequenceClassification
-    :members:
+    :members: forward
+
+
+DebertaForTokenClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaForTokenClassification
+    :members: forward
+
+
+DebertaForQuestionAnswering
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaForQuestionAnswering
+    :members: forward

docs/source/model_doc/deberta_v2.rst

@@ -0,0 +1,118 @@
+.. 
+    Copyright 2020 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+    specific language governing permissions and limitations under the License.
+
+DeBERTa-v2
+-----------------------------------------------------------------------------------------------------------------------
+
+Overview
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The DeBERTa model was proposed in `DeBERTa: Decoding-enhanced BERT with Disentangled Attention
+<https://arxiv.org/abs/2006.03654>`__ by Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen It is based on Google's
+BERT model released in 2018 and Facebook's RoBERTa model released in 2019.
+
+It builds on RoBERTa with disentangled attention and enhanced mask decoder training with half of the data used in
+RoBERTa.
+
+The abstract from the paper is the following:
+
+*Recent progress in pre-trained neural language models has significantly improved the performance of many natural
+language processing (NLP) tasks. In this paper we propose a new model architecture DeBERTa (Decoding-enhanced BERT with
+disentangled attention) that improves the BERT and RoBERTa models using two novel techniques. The first is the
+disentangled attention mechanism, where each word is represented using two vectors that encode its content and
+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 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.*
+
+
+The following information is visible directly on the [original implementation
+repository](https://github.com/microsoft/DeBERTa). DeBERTa v2 is the second version of the DeBERTa model. It includes
+the 1.5B model used for the SuperGLUE single-model submission and achieving 89.9, versus human baseline 89.8. You can
+find more details about this submission in the authors'
+[blog](https://www.microsoft.com/en-us/research/blog/microsoft-deberta-surpasses-human-performance-on-the-superglue-benchmark/)
+
+New in v2:
+
+- **Vocabulary** In v2 the tokenizer is changed to use a new vocabulary of size 128K built from the training data.
+  Instead of a GPT2-based tokenizer, the tokenizer is now
+  [sentencepiece-based](https://github.com/google/sentencepiece) tokenizer.
+- **nGiE(nGram Induced Input Encoding)** The DeBERTa-v2 model uses an additional convolution layer aside with the first
+  transformer layer to better learn the local dependency of input tokens.
+- **Sharing position projection matrix with content projection matrix in attention layer** Based on previous
+  experiments, this can save parameters without affecting the performance.
+- **Apply bucket to encode relative postions** The DeBERTa-v2 model uses log bucket to encode relative positions
+  similar to T5.
+- **900M model & 1.5B model** Two additional model sizes are available: 900M and 1.5B, which significantly improves the
+  performance of downstream tasks.
+
+The original code can be found `here <https://github.com/microsoft/DeBERTa>`__.
+
+
+DebertaV2Config
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaV2Config
+    :members:
+
+
+DebertaV2Tokenizer
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaV2Tokenizer
+    :members: build_inputs_with_special_tokens, get_special_tokens_mask,
+        create_token_type_ids_from_sequences, save_vocabulary
+
+
+DebertaV2Model
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaV2Model
+    :members: forward
+
+
+DebertaV2PreTrainedModel
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaV2PreTrainedModel
+    :members: forward
+
+
+DebertaV2ForMaskedLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaV2ForMaskedLM
+    :members: forward
+
+
+DebertaV2ForSequenceClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaV2ForSequenceClassification
+    :members: forward
+
+
+DebertaV2ForTokenClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaV2ForTokenClassification
+    :members: forward
+
+
+DebertaV2ForQuestionAnswering
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.DebertaV2ForQuestionAnswering
+    :members: forward

docs/source/model_doc/dialogpt.rst

@@ -48,7 +48,6 @@ modeling. We first concatenate all dialog turns within a dialogue session into a
 sequence length), ended by the end-of-text token.* For more information please confer to the original paper.
 
 
-DialoGPT's architecture is based on the GPT2 model, so one can refer to GPT2's `docstring
-<https://huggingface.co/transformers/model_doc/gpt2.html>`_.
+DialoGPT's architecture is based on the GPT2 model, so one can refer to :doc:`GPT2's documentation page <gpt2>`.
 
 The original code can be found `here <https://github.com/microsoft/DialoGPT>`_.

docs/source/model_doc/fsmt.rst

@@ -56,7 +56,7 @@ FSMTTokenizer
 
 .. autoclass:: transformers.FSMTTokenizer
     :members: build_inputs_with_special_tokens, get_special_tokens_mask,
-        create_token_type_ids_from_sequences, prepare_seq2seq_batch, save_vocabulary
+        create_token_type_ids_from_sequences, save_vocabulary
 
 
 FSMTModel

docs/source/model_doc/ibert.rst

@@ -0,0 +1,88 @@
+.. 
+    Copyright 2020 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+    specific language governing permissions and limitations under the License.
+
+I-BERT
+-----------------------------------------------------------------------------------------------------------------------
+
+Overview
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The I-BERT model was proposed in `I-BERT: Integer-only BERT Quantization <https://arxiv.org/abs/2101.01321>`__ by
+Sehoon Kim, Amir Gholami, Zhewei Yao, Michael W. Mahoney and Kurt Keutzer. It's a quantized version of RoBERTa running
+inference up to four times faster.
+
+The abstract from the paper is the following:
+
+*Transformer based models, like BERT and RoBERTa, have achieved state-of-the-art results in many Natural Language
+Processing tasks. However, their memory footprint, inference latency, and power consumption are prohibitive for
+efficient inference at the edge, and even at the data center. While quantization can be a viable solution for this,
+previous work on quantizing Transformer based models use floating-point arithmetic during inference, which cannot
+efficiently utilize integer-only logical units such as the recent Turing Tensor Cores, or traditional integer-only ARM
+processors. In this work, we propose I-BERT, a novel quantization scheme for Transformer based models that quantizes
+the entire inference with integer-only arithmetic. Based on lightweight integer-only approximation methods for
+nonlinear operations, e.g., GELU, Softmax, and Layer Normalization, I-BERT performs an end-to-end integer-only BERT
+inference without any floating point calculation. We evaluate our approach on GLUE downstream tasks using
+RoBERTa-Base/Large. We show that for both cases, I-BERT achieves similar (and slightly higher) accuracy as compared to
+the full-precision baseline. Furthermore, our preliminary implementation of I-BERT shows a speedup of 2.4 - 4.0x for
+INT8 inference on a T4 GPU system as compared to FP32 inference. The framework has been developed in PyTorch and has
+been open-sourced.*
+
+
+The original code can be found `here <https://github.com/kssteven418/I-BERT>`__.
+
+IBertConfig
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.IBertConfig
+    :members:
+
+
+IBertModel
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.IBertModel
+    :members: forward
+
+
+IBertForMaskedLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.IBertForMaskedLM
+    :members: forward
+
+
+IBertForSequenceClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.IBertForSequenceClassification
+    :members: forward
+
+
+IBertForMultipleChoice
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.IBertForMultipleChoice
+    :members: forward
+
+
+IBertForTokenClassification
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.IBertForTokenClassification
+    :members: forward
+
+
+IBertForQuestionAnswering
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.IBertForQuestionAnswering
+    :members: forward

docs/source/model_doc/m2m_100.rst

@@ -0,0 +1,128 @@
+.. 
+    Copyright 2020 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+    specific language governing permissions and limitations under the License.
+
+M2M100
+-----------------------------------------------------------------------------------------------------------------------
+
+Overview
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The M2M100 model was proposed in `Beyond English-Centric Multilingual Machine Translation
+<https://arxiv.org/abs/2010.11125>`__ by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky,
+Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy
+Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
+
+The abstract from the paper is the following:
+
+*Existing work in translation demonstrated the potential of massively multilingual machine translation by training a
+single model able to translate between any pair of languages. However, much of this work is English-Centric by training
+only on data which was translated from or to English. While this is supported by large sources of training data, it
+does not reflect translation needs worldwide. In this work, we create a true Many-to-Many multilingual translation
+model that can translate directly between any pair of 100 languages. We build and open source a training dataset that
+covers thousands of language directions with supervised data, created through large-scale mining. Then, we explore how
+to effectively increase model capacity through a combination of dense scaling and language-specific sparse parameters
+to create high quality models. Our focus on non-English-Centric models brings gains of more than 10 BLEU when directly
+translating between non-English directions while performing competitively to the best single systems of WMT. We
+open-source our scripts so that others may reproduce the data, evaluation, and final M2M-100 model.*
+
+
+Training and Generation
+_______________________________________________________________________________________________________________________
+
+M2M100 is a multilingual encoder-decoder (seq-to-seq) model primarily intended for translation tasks. As the model is
+multilingual it expects the sequences in a certain format: A special language id token is used as prefix in both the
+source and target text. The source text format is :obj:`[lang_code] X [eos]`, where :obj:`lang_code` is source language
+id for source text and target language id for target text, with :obj:`X` being the source or target text.
+
+The :class:`~transformers.M2M100Tokenizer` depends on :obj:`sentencepiece` so be sure to install it before running the
+examples. To install :obj:`sentencepiece` run ``pip install sentencepiece``.
+
+- Supervised Training
+
+.. code-block::
+
+    from transformers import M2M100Config, M2M100ForConditionalGeneration, M2M100Tokenizer
+
+    model = M2M100ForConditionalGeneration.from_pretrained('facebook/m2m100_418M')
+    tokenizer = M2M100Tokenizer.from_pretrained('facebook/m2m100_418M', src_lang="en", tgt_lang="fr")
+
+    src_text = "Life is like a box of chocolates."
+    tgt_lang = "La vie est comme une boîte de chocolat."
+
+    model_inputs = tokenizer(src_text, return_tensors="pt")
+    with tokenizer.as_target_tokenizer():
+        labels = tokenizer(tgt_text, return_tensors="pt").input_ids
+
+    loss = model(**model_inputs, labels=labels) # forward pass
+
+
+- Generation
+
+    M2M100 uses the :obj:`eos_token_id` as the :obj:`decoder_start_token_id` for generation with the target language id
+    being forced as the first generated token. To force the target language id as the first generated token, pass the
+    `forced_bos_token_id` parameter to the `generate` method. The following example shows how to translate between
+    Hindi to French and Chinese to English using the `facebook/m2m100_418M` checkpoint.
+
+.. code-block::
+
+    >>> from transformers import M2M100ForConditionalGeneration, M2M100Tokenizer
+
+    >>> hi_text = "जीवन एक चॉकलेट बॉक्स की तरह है।"
+    >>> chinese_text = "生活就像一盒巧克力。"
+
+    >>> model = M2M100ForConditionalGeneration.from_pretrained("facebook/m2m100_418M")
+    >>> tokenizer = M2M100Tokenizer.from_pretrained("facebook/m2m100_418M")
+
+    >>> # translate Hindi to French
+    >>> tokenizer.src_lang = "hi"
+    >>> encoded_hi = tokenizer(hi_text, return_tensors="pt")
+    >>> generated_tokens = model.generate(**encoded_hi, forced_bos_token_id=tokenizer.get_lang_id("fr"))
+    >>> tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
+    "La vie est comme une boîte de chocolat."
+
+    >>> # translate Chinese to English
+    >>> tokenizer.src_lang = "zh"
+    >>> encoded_zh = tokenizer(chinese_text, return_tensors="pt")
+    >>> generated_tokens = model.generate(**encoded_zh, forced_bos_token_id=tokenizer.get_lang_id("en"))
+    >>> tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
+    "Life is like a box of chocolate."
+
+
+M2M100Config
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.M2M100Config
+    :members:
+
+
+M2M100Tokenizer
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.M2M100Tokenizer
+    :members: build_inputs_with_special_tokens, get_special_tokens_mask,
+        create_token_type_ids_from_sequences, save_vocabulary
+
+
+M2M100Model
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.M2M100Model
+    :members: forward
+
+
+M2M100ForConditionalGeneration
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.M2M100ForConditionalGeneration
+    :members: forward
+
+

docs/source/model_doc/marian.rst

@@ -76,27 +76,29 @@ 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'
-    ]
+    >>> 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']
+    >>> model_name = 'Helsinki-NLP/opus-mt-en-roa'
+    >>> tokenizer = MarianTokenizer.from_pretrained(model_name)
+    >>> print(tokenizer.supported_language_codes)
+    ['>>zlm_Latn<<', '>>mfe<<', '>>hat<<', '>>pap<<', '>>ast<<', '>>cat<<', '>>ind<<', '>>glg<<', '>>wln<<', '>>spa<<', '>>fra<<', '>>ron<<', '>>por<<', '>>ita<<', '>>oci<<', '>>arg<<', '>>min<<']
+
+    >>> model = MarianMTModel.from_pretrained(model_name)
+    >>> translated = model.generate(**tokenizer(src_text, return_tensors="pt", padding=True))
+    >>> [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:
+Here is the code to see all available pretrained models on the hub:
 
 .. code-block:: python
 
@@ -147,21 +149,22 @@ Example of translating english to many romance languages, using old-style 2 char
 
 .. code-block::python
 
-    from transformers import MarianMTModel, MarianTokenizer
-    src_text = [
-        '>>fr<< this is a sentence in english that we want to translate to french',
-        '>>pt<< This should go to portuguese',
-        '>>es<< And this to Spanish'
-    ]
+    >>> from transformers import MarianMTModel, MarianTokenizer
+    >>> src_text = [
+    ...     '>>fr<< this is a sentence in english that we want to translate to french',
+    ...     '>>pt<< This should go to portuguese',
+    ...     '>>es<< And this to Spanish'
+    >>> ]
 
-    model_name = 'Helsinki-NLP/opus-mt-en-ROMANCE'
-    tokenizer = MarianTokenizer.from_pretrained(model_name)
-    print(tokenizer.supported_language_codes)
+    >>> model_name = 'Helsinki-NLP/opus-mt-en-ROMANCE'
+    >>> tokenizer = MarianTokenizer.from_pretrained(model_name)
 
-    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']
+    >>> model = MarianMTModel.from_pretrained(model_name)
+    >>> translated = model.generate(**tokenizer(src_text, return_tensors="pt", padding=True))
+    >>> 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']
 
 
 
@@ -176,7 +179,7 @@ MarianTokenizer
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 .. autoclass:: transformers.MarianTokenizer
-    :members: prepare_seq2seq_batch
+    :members: as_target_tokenizer
 
 
 MarianModel
@@ -193,6 +196,13 @@ MarianMTModel
     :members: forward
 
 
+MarianForCausalLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.MarianForCausalLM
+    :members: forward
+
+
 TFMarianModel
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

docs/source/model_doc/mbart.rst

@@ -10,14 +10,14 @@
     an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
     specific language governing permissions and limitations under the License.
 
-MBart
+MBart and MBart-50
 -----------------------------------------------------------------------------------------------------------------------
 
 **DISCLAIMER:** If you see something strange, file a `Github Issue
 <https://github.com/huggingface/transformers/issues/new?assignees=&labels=&template=bug-report.md&title>`__ and assign
 @patrickvonplaten
 
-Overview
+Overview of MBart
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 The MBart model was presented in `Multilingual Denoising Pre-training for Neural Machine Translation
@@ -31,33 +31,34 @@ 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
+Training of MBart
 _______________________________________________________________________________________________________________________
 
-- Examples and scripts for fine-tuning mBART and other models for sequence to sequence tasks can be found in
-  :prefix_link:`examples/seq2seq/ <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.
+MBart is a multilingual encoder-decoder (sequence-to-sequence) model primarily intended for translation task. As the
+model is multilingual it expects the sequences in a different format. A special language id token is added in both the
+source and target text. The source text format is :obj:`X [eos, src_lang_code]` where :obj:`X` is the source text. The
+target text format is :obj:`[tgt_lang_code] X [eos]`. :obj:`bos` is never used.
 
-Training
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-
-MBart is a multilingual encoder-decoder (seq-to-seq) model primarily intended for translation task. As the model is
-multilingual it expects the sequences in a different format. A special language id token is added in both the source
-and target text. The source text format is :obj:`X [eos, src_lang_code]` where :obj:`X` is the source text. The target
-text format is :obj:`[tgt_lang_code] X [eos]`. :obj:`bos` is never used.
-
-The :meth:`~transformers.MBartTokenizer.prepare_seq2seq_batch` handles this automatically and should be used to encode
-the sequences for sequence-to-sequence fine-tuning.
+The regular :meth:`~transformers.MBartTokenizer.__call__` will encode source text format, and it should be wrapped
+inside the context manager :meth:`~transformers.MBartTokenizer.as_target_tokenizer` to encode target text format.
 
 - Supervised training
 
 .. code-block::
 
-    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, return_tensors="pt")
-    model(input_ids=batch['input_ids'], labels=batch['labels']) # forward pass
+    >>> from transformers import MBartForConditionalGeneration, MBartTokenizer
+
+    >>> tokenizer = MBartTokenizer.from_pretrained("facebook/mbart-large-en-ro")
+    >>> 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"
+
+    >>> inputs = tokenizer(example_english_phrase, return_tensors="pt", src_lang="en_XX", tgt_lang="ro_RO")
+    >>> with tokenizer.as_target_tokenizer():
+    ...     labels = tokenizer(expected_translation_romanian, return_tensors="pt")
+
+    >>> model = MBartForConditionalGeneration.from_pretrained("facebook/mbart-large-en-ro")
+    >>> # forward pass
+    >>> model(**inputs, labels=batch['labels'])
 
 - Generation
 
@@ -66,14 +67,95 @@ the sequences for sequence-to-sequence fine-tuning.
 
 .. code-block::
 
-    from transformers import MBartForConditionalGeneration, MBartTokenizer
-    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", 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"
+    >>> from transformers import MBartForConditionalGeneration, MBartTokenizer
+
+    >>> tokenizer = MBartTokenizer.from_pretrained("facebook/mbart-large-en-ro", src_lang="en_XX")
+    >>> article = "UN Chief Says There Is No Military Solution in Syria"
+    >>> inputs = tokenizer(article, return_tensors="pt")
+    >>> translated_tokens = model.generate(**inputs, decoder_start_token_id=tokenizer.lang_code_to_id["ro_RO"])
+    >>> tokenizer.batch_decode(translated_tokens, skip_special_tokens=True)[0]
+    "Şeful ONU declară că nu există o soluţie militară în Siria"
+
+
+Overview of MBart-50
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+MBart-50 was introduced in the `Multilingual Translation with Extensible Multilingual Pretraining and Finetuning
+<https://arxiv.org/abs/2008.00401>` paper by Yuqing Tang, Chau Tran, Xian Li, Peng-Jen Chen, Naman Goyal, Vishrav
+Chaudhary, Jiatao Gu, Angela Fan. MBart-50 is created using the original `mbart-large-cc25` checkpoint by extendeding
+its embedding layers with randomly initialized vectors for an extra set of 25 language tokens and then pretrained on 50
+languages.
+
+According to the abstract
+
+*Multilingual translation models can be created through multilingual finetuning. Instead of finetuning on one
+direction, a pretrained model is finetuned on many directions at the same time. It demonstrates that pretrained models
+can be extended to incorporate additional languages without loss of performance. Multilingual finetuning improves on
+average 1 BLEU over the strongest baselines (being either multilingual from scratch or bilingual finetuning) while
+improving 9.3 BLEU on average over bilingual baselines from scratch.*
+
+
+Training of MBart-50
+_______________________________________________________________________________________________________________________
+
+The text format for MBart-50 is slightly different from mBART. For MBart-50 the language id token is used as a prefix
+for both source and target text i.e the text format is :obj:`[lang_code] X [eos]`, where :obj:`lang_code` is source
+language id for source text and target language id for target text, with :obj:`X` being the source or target text
+respectively.
+
+
+MBart-50 has its own tokenizer :class:`~transformers.MBart50Tokenizer`.
+
+-  Supervised training
+
+.. code-block::
+
+    from transformers import MBartForConditionalGeneration, MBart50TokenizerFast
+
+    model = MBartForConditionalGeneration.from_pretrained("facebook/mbart-large-50")
+    tokenizer = MBart50TokenizerFast.from_pretrained("facebook/mbart-large-50", src_lang="en_XX", tgt_lang="ro_RO")
+
+    src_text = " UN Chief Says There Is No Military Solution in Syria"
+    tgt_text =  "Şeful ONU declară că nu există o soluţie militară în Siria"
+
+    model_inputs = tokenizer(src_text, return_tensors="pt")
+    with tokenizer.as_target_tokenizer():
+        labels = tokenizer(tgt_text, return_tensors="pt").input_ids
+
+    model(**model_inputs, labels=labels) # forward pass
+
+
+- Generation
+
+    To generate using the mBART-50 multilingual translation models, :obj:`eos_token_id` is used as the
+    :obj:`decoder_start_token_id` and the target language id is forced as the first generated token. To force the
+    target language id as the first generated token, pass the `forced_bos_token_id` parameter to the `generate` method.
+    The following example shows how to translate between Hindi to French and Arabic to English using the
+    `facebook/mbart-50-large-many-to-many` checkpoint.
+
+.. code-block::
+
+    from transformers import MBartForConditionalGeneration, MBart50TokenizerFast
+
+    article_hi = "संयुक्त राष्ट्र के प्रमुख का कहना है कि सीरिया में कोई सैन्य समाधान नहीं है"
+    article_ar = "الأمين العام للأمم المتحدة يقول إنه لا يوجد حل عسكري في سوريا."
+
+    model = MBartForConditionalGeneration.from_pretrained("facebook/mbart-large-50-many-to-many-mmt")
+    tokenizer = MBart50TokenizerFast.from_pretrained("facebook/mbart-large-50-many-to-many-mmt")
+
+    # translate Hindi to French
+    tokenizer.src_lang = "hi_IN"
+    encoded_hi = tokenizer(article_hi, return_tensors="pt")
+    generated_tokens = model.generate(**encoded_hi, forced_bos_token_id=tokenizer.lang_code_to_id["fr_XX"])
+    tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
+    # => "Le chef de l 'ONU affirme qu 'il n 'y a pas de solution militaire en Syria."
+
+    # translate Arabic to English
+    tokenizer.src_lang = "ar_AR"
+    encoded_ar = tokenizer(article_ar, return_tensors="pt")
+    generated_tokens = model.generate(**encoded_ar, forced_bos_token_id=tokenizer.lang_code_to_id["en_XX"])
+    tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
+    # => "The Secretary-General of the United Nations says there is no military solution in Syria."
 
 
 MBartConfig
@@ -87,7 +169,7 @@ MBartTokenizer
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 .. autoclass:: transformers.MBartTokenizer
-    :members: build_inputs_with_special_tokens, prepare_seq2seq_batch
+    :members: as_target_tokenizer, build_inputs_with_special_tokens
 
 
 MBartTokenizerFast
@@ -97,6 +179,20 @@ MBartTokenizerFast
     :members:
 
 
+MBart50Tokenizer
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.MBart50Tokenizer
+    :members:
+
+
+MBart50TokenizerFast
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.MBart50TokenizerFast
+    :members:
+
+
 MBartModel
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
@@ -124,6 +220,13 @@ MBartForSequenceClassification
 .. autoclass:: transformers.MBartForSequenceClassification
 
 
+MBartForCausalLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.MBartForCausalLM
+    :members: forward
+
+
 TFMBartModel
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

docs/source/model_doc/pegasus.rst

@@ -51,8 +51,8 @@ All the `checkpoints <https://huggingface.co/models?search=pegasus>`__ are fine-
 Examples
 _______________________________________________________________________________________________________________________
 
-- :prefix_link:`Script <examples/seq2seq/finetune_pegasus_xsum.sh>` to fine-tune pegasus on the XSUM dataset. Data
-  download instructions at :prefix_link:`examples/seq2seq/ <examples/seq2seq/README.md>`.
+- :prefix_link:`Script <examples/research_projects/seq2seq-distillation/finetune_pegasus_xsum.sh>` to fine-tune pegasus
+  on the XSUM dataset. Data download instructions at :prefix_link:`examples/seq2seq/ <examples/seq2seq/README.md>`.
 - FP16 is not supported (help/ideas on this appreciated!).
 - The adafactor optimizer is recommended for pegasus fine-tuning.
 
@@ -78,20 +78,20 @@ Usage Example
 
 .. code-block:: python
 
-    from transformers import PegasusForConditionalGeneration, PegasusTokenizer
-    import torch
-    src_text = [
-        """ PG&E stated it scheduled the blackouts in response to forecasts for high winds amid dry conditions. The aim is to reduce the risk of wildfires. Nearly 800 thousand customers were scheduled to be affected by the shutoffs which were expected to last through at least midday tomorrow."""
-    ]
+    >>> from transformers import PegasusForConditionalGeneration, PegasusTokenizer
+    >>> import torch
+    >>> src_text = [
+    ...     """ PG&E stated it scheduled the blackouts in response to forecasts for high winds amid dry conditions. The aim is to reduce the risk of wildfires. Nearly 800 thousand customers were scheduled to be affected by the shutoffs which were expected to last through at least midday tomorrow."""
+    >>> ]
 
-    model_name = 'google/pegasus-xsum'
-    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', 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."
+    >>> model_name = 'google/pegasus-xsum'
+    >>> device = 'cuda' if torch.cuda.is_available() else 'cpu'
+    >>> tokenizer = PegasusTokenizer.from_pretrained(model_name)
+    >>> model = PegasusForConditionalGeneration.from_pretrained(model_name).to(device)
+    >>> batch = tokenizer(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."
 
 
 
@@ -107,7 +107,7 @@ PegasusTokenizer
 warning: ``add_tokens`` does not work at the moment.
 
 .. autoclass:: transformers.PegasusTokenizer
-    :members: __call__, prepare_seq2seq_batch
+    :members:
 
 
 PegasusTokenizerFast
@@ -131,6 +131,13 @@ PegasusForConditionalGeneration
     :members: forward
 
 
+PegasusForCausalLM
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.PegasusForCausalLM
+    :members: forward
+
+
 TFPegasusModel
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

docs/source/model_doc/rag.rst

@@ -56,7 +56,7 @@ RagTokenizer
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 .. autoclass:: transformers.RagTokenizer
-    :members: prepare_seq2seq_batch
+    :members:
 
 
 Rag specific outputs
@@ -94,3 +94,24 @@ RagTokenForGeneration
 
 .. autoclass:: transformers.RagTokenForGeneration
     :members: forward, generate
+
+
+TFRagModel
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFRagModel
+    :members: call
+
+
+TFRagSequenceForGeneration
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFRagSequenceForGeneration
+    :members: call, generate
+
+
+TFRagTokenForGeneration
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.TFRagTokenForGeneration
+    :members: call, generate

docs/source/model_doc/speech_to_text.rst

@@ -0,0 +1,152 @@
+.. 
+    Copyright 2021 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+    specific language governing permissions and limitations under the License.
+
+Speech2Text
+-----------------------------------------------------------------------------------------------------------------------
+
+Overview
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The Speech2Text model was proposed in `fairseq S2T: Fast Speech-to-Text Modeling with fairseq
+<https://arxiv.org/abs/2010.05171>`__ by Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Dmytro Okhonko, Juan Pino. It's a
+transformer-based seq2seq (encoder-decoder) model designed for end-to-end Automatic Speech Recognition (ASR) and Speech
+Translation (ST). It uses a convolutional downsampler to reduce the length of speech inputs by 3/4th before they are
+fed into the encoder. The model is trained with standard autoregressive cross-entropy loss and generates the
+transcripts/translations autoregressively. Speech2Text has been fine-tuned on several datasets for ASR and ST:
+`LibriSpeech <http://www.openslr.org/12>`__, `CoVoST 2 <https://github.com/facebookresearch/covost>`__, `MuST-C
+<https://ict.fbk.eu/must-c/>`__.
+
+The original code can be found `here <https://github.com/pytorch/fairseq/tree/master/examples/speech_to_text>`__.
+
+
+Inference
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Speech2Text is a speech model that accepts a float tensor of log-mel filter-bank features extracted from the speech
+signal. It's a transformer-based seq2seq model, so the transcripts/translations are generated autoregressively. The
+:obj:`generate()` method can be used for inference.
+
+The :class:`~transformers.Speech2TextFeatureExtractor` class is responsible for extracting the log-mel filter-bank
+features. The :class:`~transformers.Speech2TextProcessor` wraps :class:`~transformers.Speech2TextFeatureExtractor` and
+:class:`~transformers.Speech2TextTokenizer` into a single instance to both extract the input features and decode the
+predicted token ids.
+
+The feature extractor depends on :obj:`torchaudio` and the tokenizer depends on :obj:`sentencepiece` so be sure to
+install those packages before running the examples. You could either install those as extra speech dependancies with
+``pip install transformers"[speech, sentencepiece]"`` or install the packages seperatly with ``pip install torchaudio
+sentencepiece``. Also ``torchaudio`` requires the development version of the `libsndfile
+<http://www.mega-nerd.com/libsndfile/>`__ package which can be installed via a system package manager. On Ubuntu it can
+be installed as follows: ``apt install libsndfile1-dev``
+
+
+- ASR and Speech Translation
+
+.. code-block::
+
+        >>> import torch
+        >>> from transformers import Speech2TextProcessor, Speech2TextForConditionalGeneration
+        >>> from datasets import load_dataset
+        >>> import soundfile as sf
+
+        >>> model = Speech2TextForConditionalGeneration.from_pretrained("facebook/s2t-small-librispeech-asr")
+        >>> processor = Speech2TextProcessor.from_pretrained("facebook/s2t-small-librispeech-asr")
+
+        >>> def map_to_array(batch):
+        ...     speech, _ = sf.read(batch["file"])
+        ...     batch["speech"] = speech
+        ...     return batch
+
+        >>> ds = load_dataset("patrickvonplaten/librispeech_asr_dummy", "clean", split="validation")
+        >>> ds = ds.map(map_to_array)
+
+        >>> inputs = processor(ds["speech"][0], sampling_rate=16_000, return_tensors="pt")
+        >>> generated_ids = model.generate(input_ids=inputs["input_features"], attention_mask=inputs["attention_mask"])
+
+        >>> transcription = processor.batch_decode(generated_ids)
+
+
+- Multilingual speech translation
+
+    For multilingual speech translation models, :obj:`eos_token_id` is used as the :obj:`decoder_start_token_id` and
+    the target language id is forced as the first generated token. To force the target language id as the first
+    generated token, pass the :obj:`forced_bos_token_id` parameter to the :obj:`generate()` method. The following
+    example shows how to transate English speech to French text using the `facebook/s2t-medium-mustc-multilingual-st`
+    checkpoint.
+
+.. code-block::
+
+        >>> import torch
+        >>> from transformers import Speech2TextProcessor, Speech2TextForConditionalGeneration
+        >>> from datasets import load_dataset
+        >>> import soundfile as sf
+
+        >>> model = Speech2TextForConditionalGeneration.from_pretrained("facebook/s2t-medium-mustc-multilingual-st")
+        >>> processor = Speech2TextProcessor.from_pretrained("facebook/s2t-medium-mustc-multilingual-st")
+
+        >>> def map_to_array(batch):
+        ...     speech, _ = sf.read(batch["file"])
+        ...     batch["speech"] = speech
+        ...     return batch
+
+        >>> ds = load_dataset("patrickvonplaten/librispeech_asr_dummy", "clean", split="validation")
+        >>> ds = ds.map(map_to_array)
+
+        >>> inputs = processor(ds["speech"][0], sampling_rate=16_000, return_tensors="pt")
+        >>> generated_ids = model.generate(input_ids=inputs["input_features"], attention_mask=inputs["attention_mask], forced_bos_token_id=processor.tokenizer.lang_code_to_id["fr"])
+
+        >>> translation = processor.batch_decode(generated_ids)
+
+
+See the `model hub <https://huggingface.co/models?filter=speech_to_text>`__ to look for Speech2Text checkpoints.
+
+
+Speech2TextConfig
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Speech2TextConfig
+    :members:
+
+
+Speech2TextTokenizer
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Speech2TextTokenizer
+    :members: build_inputs_with_special_tokens, get_special_tokens_mask,
+        create_token_type_ids_from_sequences, save_vocabulary
+
+
+Speech2TextFeatureExtractor
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Speech2TextFeatureExtractor
+    :members: __call__
+
+
+Speech2TextProcessor
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Speech2TextProcessor
+    :members: __call__, from_pretrained, save_pretrained, batch_decode, decode, as_target_processor
+
+
+Speech2TextModel
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Speech2TextModel
+    :members: forward
+
+
+Speech2TextForConditionalGeneration
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Speech2TextForConditionalGeneration
+    :members: forward

docs/source/model_doc/t5.rst

@@ -104,7 +104,7 @@ T5Tokenizer
 
 .. autoclass:: transformers.T5Tokenizer
     :members: build_inputs_with_special_tokens, get_special_tokens_mask,
-        create_token_type_ids_from_sequences, prepare_seq2seq_batch, save_vocabulary
+        create_token_type_ids_from_sequences, save_vocabulary
 
 
 T5TokenizerFast

docs/source/model_doc/wav2vec2.rst

@@ -0,0 +1,79 @@
+.. 
+    Copyright 2021 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+    specific language governing permissions and limitations under the License.
+
+Wav2Vec2
+-----------------------------------------------------------------------------------------------------------------------
+
+Overview
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The Wav2Vec2 model was proposed in `wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations
+<https://arxiv.org/abs/2006.11477>`__ by Alexei Baevski, Henry Zhou, Abdelrahman Mohamed, Michael Auli.
+
+The abstract from the paper is the following:
+
+*We show for the first time that learning powerful representations from speech audio alone followed by fine-tuning on
+transcribed speech can outperform the best semi-supervised methods while being conceptually simpler. wav2vec 2.0 masks
+the speech input in the latent space and solves a contrastive task defined over a quantization of the latent
+representations which are jointly learned. Experiments using all labeled data of Librispeech achieve 1.8/3.3 WER on the
+clean/other test sets. When lowering the amount of labeled data to one hour, wav2vec 2.0 outperforms the previous state
+of the art on the 100 hour subset while using 100 times less labeled data. Using just ten minutes of labeled data and
+pre-training on 53k hours of unlabeled data still achieves 4.8/8.2 WER. This demonstrates the feasibility of speech
+recognition with limited amounts of labeled data.*
+
+Tips:
+
+- Wav2Vec2 is a speech model that accepts a float array corresponding to the raw waveform of the speech signal.
+- Wav2Vec2 model was trained using connectionist temporal classification (CTC) so the model output has to be decoded
+  using :class:`~transformers.Wav2Vec2CTCTokenizer`.
+
+
+Wav2Vec2Config
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Wav2Vec2Config
+    :members:
+
+
+Wav2Vec2CTCTokenizer
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Wav2Vec2CTCTokenizer
+    :members: __call__, save_vocabulary
+
+
+Wav2Vec2FeatureExtractor
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Wav2Vec2FeatureExtractor
+    :members: __call__
+
+
+Wav2Vec2Processor
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Wav2Vec2Processor
+    :members: __call__, pad, from_pretrained, save_pretrained, batch_decode, decode, as_target_processor
+
+
+Wav2Vec2Model
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Wav2Vec2Model
+    :members: forward
+
+
+Wav2Vec2ForCTC
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.Wav2Vec2ForCTC
+    :members: forward

docs/source/model_doc/xlsr_wav2vec2.rst

@@ -0,0 +1,45 @@
+.. 
+    Copyright 2021 The HuggingFace Team. All rights reserved.
+
+    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+    the License. You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+    an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+    specific language governing permissions and limitations under the License.
+
+XLSR-Wav2Vec2
+-----------------------------------------------------------------------------------------------------------------------
+
+Overview
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The XLSR-Wav2Vec2 model was proposed in `Unsupervised Cross-Lingual Representation Learning For Speech Recognition
+<https://arxiv.org/abs/2006.13979>`__ by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael
+Auli.
+
+The abstract from the paper is the following:
+
+*This paper presents XLSR which learns cross-lingual speech representations by pretraining a single model from the raw
+waveform of speech in multiple languages. We build on wav2vec 2.0 which is trained by solving a contrastive task over
+masked latent speech representations and jointly learns a quantization of the latents shared across languages. The
+resulting model is fine-tuned on labeled data and experiments show that cross-lingual pretraining significantly
+outperforms monolingual pretraining. On the CommonVoice benchmark, XLSR shows a relative phoneme error rate reduction
+of 72% compared to the best known results. On BABEL, our approach improves word error rate by 16% relative compared to
+a comparable system. Our approach enables a single multilingual speech recognition model which is competitive to strong
+individual models. Analysis shows that the latent discrete speech representations are shared across languages with
+increased sharing for related languages. We hope to catalyze research in low-resource speech understanding by releasing
+XLSR-53, a large model pretrained in 53 languages.*
+
+Tips:
+
+- XLSR-Wav2Vec2 is a speech model that accepts a float array corresponding to the raw waveform of the speech signal.
+- XLSR-Wav2Vec2 model was trained using connectionist temporal classification (CTC) so the model output has to be
+  decoded using :class:`~transformers.Wav2Vec2CTCTokenizer`.
+
+XLSR-Wav2Vec2's architecture is based on the Wav2Vec2 model, so one can refer to :doc:`Wav2Vec2's documentation page
+<wav2vec2>`.
+
+The original code can be found `here <https://github.com/pytorch/fairseq/tree/master/fairseq/models/wav2vec>`__.

docs/source/model_sharing.rst

@@ -60,7 +60,7 @@ 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.
 
-You can create a model repo **directly from `the /new page on the website <https://huggingface.co/new>`__.**
+You can create a model repo directly from `the /new page on the website <https://huggingface.co/new>`__.
 
 Alternatively, you can use the ``transformers-cli``. The next steps describe that process:
 

docs/source/model_summary.rst

@@ -330,6 +330,36 @@ the same probabilities as the larger model. The actual objective is a combinatio
 The library provides a version of the model for masked language modeling, token classification, sentence classification
 and question answering.
 
+ConvBERT
+-----------------------------------------------------------------------------------------------------------------------
+
+.. raw:: html
+
+   <a href="https://huggingface.co/models?filter=convbert">
+       <img alt="Models" src="https://img.shields.io/badge/All_model_pages-convbert-blueviolet">
+   </a>
+   <a href="model_doc/convbert.html">
+       <img alt="Doc" src="https://img.shields.io/badge/Model_documentation-convbert-blueviolet">
+   </a>
+
+`ConvBERT: Improving BERT with Span-based Dynamic Convolution <https://arxiv.org/abs/1910.01108>`_, Zihang Jiang,
+Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
+
+Pre-trained language models like BERT and its variants have recently achieved impressive performance in various natural
+language understanding tasks. However, BERT heavily relies on the global self-attention block and thus suffers large
+memory footprint and computation cost. Although all its attention heads query on the whole input sequence for
+generating the attention map from a global perspective, we observe some heads only need to learn local dependencies,
+which means the existence of computation redundancy. We therefore propose a novel span-based dynamic convolution to
+replace these self-attention heads to directly model local dependencies. The novel convolution heads, together with the
+rest self-attention heads, form a new mixed attention block that is more efficient at both global and local context
+learning. We equip BERT with this mixed attention design and build a ConvBERT model. Experiments have shown that
+ConvBERT significantly outperforms BERT and its variants in various downstream tasks, with lower training cost and
+fewer model parameters. Remarkably, ConvBERTbase model achieves 86.4 GLUE score, 0.7 higher than ELECTRAbase, while
+using less than 1/4 training cost.
+
+The library provides a version of the model for masked language modeling, token classification, sentence classification
+and question answering.
+
 XLM
 -----------------------------------------------------------------------------------------------------------------------
 

docs/source/pretrained_models.rst

@@ -365,6 +365,12 @@ For the full list, refer to `https://huggingface.co/models <https://huggingface.
 |                    | ``reformer-crime-and-punishment``                          | | 6-layer, 256-hidden, 2-heads, 3M parameters                                                                                         |
 |                    |                                                            | | Trained on English text: Crime and Punishment novel by Fyodor Dostoyevsky.                                                          |
 +--------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
+| M2M100             | ``facebook/m2m100_418M``                                   | | 24-layer, 1024-hidden, 16-heads, 418M parameters                                                                                    |
+|                    |                                                            | | multilingual machine translation model for 100 languages                                                                            |
+|                    +------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
+|                    | ``facebook/m2m100_1.2B``                                   | | 48-layer, 1024-hidden, 16-heads, 1.2B parameters                                                                                    |
+|                    |                                                            | | multilingual machine translation model for 100 languages                                                                            |
++--------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
 | MarianMT           | ``Helsinki-NLP/opus-mt-{src}-{tgt}``                       | | 12-layer, 512-hidden, 8-heads, ~74M parameter Machine translation models. Parameter counts vary depending on vocab size.            |
 |                    |                                                            | | (see `model list <https://huggingface.co/Helsinki-NLP>`_)                                                                           |
 +--------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
@@ -381,6 +387,15 @@ For the full list, refer to `https://huggingface.co/models <https://huggingface.
 |                    +------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
 |                    | ``facebook/mbart-large-en-ro``                             | | 24-layer, 1024-hidden, 16-heads, 610M parameters                                                                                    |
 |                    |                                                            | | mbart-large-cc25 model finetuned on WMT english romanian translation.                                                               |
+|                    +------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
+|                    | ``facebook/mbart-large-50``                                | | 24-layer, 1024-hidden, 16-heads,                                                                                                    |
+|                    |                                                            | | mBART model trained on 50 languages' monolingual corpus.                                                                            |
+|                    +------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
+|                    | ``facebook/mbart-large-50-one-to-many-mmt``                | | 24-layer, 1024-hidden, 16-heads,                                                                                                    |
+|                    |                                                            | | mbart-50-large model finetuned for one (English) to many multilingual machine translation covering 50 languages.                    |
+|                    +------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
+|                    | ``facebook/mbart-large-50-many-to-many-mmt``               | | 24-layer, 1024-hidden, 16-heads,                                                                                                    |
+|                    |                                                            | | mbart-50-large model finetuned for many to many multilingual machine translation covering 50 languages.                             |
 +--------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
 | Lxmert             | ``lxmert-base-uncased``                                    | | 9-language layers, 9-relationship layers, and 12-cross-modality layers                                                              |
 |                    |                                                            | | 768-hidden, 12-heads (for each layer) ~ 228M parameters                                                                             |
@@ -434,15 +449,30 @@ For the full list, refer to `https://huggingface.co/models <https://huggingface.
 |                    |                                                            |                                                                                                                                       |
 |                    |                                                            | (see `details <https://github.com/microsoft/unilm/tree/master/layoutlm>`__)                                                           |
 +--------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
-| DeBERTa            | ``microsoft/deberta-base``                                 | | 12-layer, 768-hidden, 12-heads, ~125M parameters                                                                                    |
+| DeBERTa            | ``microsoft/deberta-base``                                 | | 12-layer, 768-hidden, 12-heads, ~140M parameters                                                                                    |
 |                    |                                                            | | DeBERTa using the BERT-base architecture                                                                                            |
 |                    |                                                            |                                                                                                                                       |
 |                    |                                                            | (see `details <https://github.com/microsoft/DeBERTa>`__)                                                                              |
 |                    +------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
-|                    | ``microsoft/deberta-large``                                | | 24-layer, 1024-hidden, 16-heads, ~390M parameters                                                                                   |
+|                    | ``microsoft/deberta-large``                                | | 24-layer, 1024-hidden, 16-heads, ~400M parameters                                                                                   |
 |                    |                                                            | | DeBERTa using the BERT-large architecture                                                                                           |
 |                    |                                                            |                                                                                                                                       |
 |                    |                                                            | (see `details <https://github.com/microsoft/DeBERTa>`__)                                                                              |
+|                    +------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
+|                    | ``microsoft/deberta-xlarge``                               | | 48-layer, 1024-hidden, 16-heads, ~750M parameters                                                                                   |
+|                    |                                                            | | DeBERTa XLarge with similar BERT architecture                                                                                       |
+|                    |                                                            |                                                                                                                                       |
+|                    |                                                            | (see `details <https://github.com/microsoft/DeBERTa>`__)                                                                              |
+|                    +------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
+|                    | ``microsoft/deberta-xlarge-v2``                            | | 24-layer, 1536-hidden, 24-heads, ~900M parameters                                                                                   |
+|                    |                                                            | | DeBERTa XLarge V2 with similar BERT architecture                                                                                    |
+|                    |                                                            |                                                                                                                                       |
+|                    |                                                            | (see `details <https://github.com/microsoft/DeBERTa>`__)                                                                              |
+|                    +------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
+|                    | ``microsoft/deberta-xxlarge-v2``                           | | 48-layer, 1536-hidden, 24-heads, ~1.5B parameters                                                                                   |
+|                    |                                                            | | DeBERTa XXLarge V2 with similar BERT architecture                                                                                   |
+|                    |                                                            |                                                                                                                                       |
+|                    |                                                            | (see `details <https://github.com/microsoft/DeBERTa>`__)                                                                              |
 +--------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
 | SqueezeBERT        | ``squeezebert/squeezebert-uncased``                        | | 12-layer, 768-hidden, 12-heads, 51M parameters, 4.3x faster than bert-base-uncased on a smartphone.                                 |
 |                    |                                                            | | SqueezeBERT architecture pretrained from scratch on masked language model (MLM) and sentence order prediction (SOP) tasks.          |

docs/source/task_summary.rst

@@ -54,12 +54,11 @@ Sequence Classification
 
 Sequence classification is the task of classifying sequences according to a given number of classes. An example of
 sequence classification is the GLUE dataset, which is entirely based on that task. If you would like to fine-tune a
-model on a GLUE sequence classification task, you may leverage the `run_glue.py
-<https://github.com/huggingface/transformers/tree/master/examples/text-classification/run_glue.py>`__ and
-`run_pl_glue.py
-<https://github.com/huggingface/transformers/tree/master/examples/text-classification/run_pl_glue.py>`__ or
-`run_tf_glue.py
-<https://github.com/huggingface/transformers/tree/master/examples/text-classification/run_tf_glue.py>`__ scripts.
+model on a GLUE sequence classification task, you may leverage the :prefix_link:`run_glue.py
+<examples/text-classification/run_glue.py>`, :prefix_link:`run_tf_glue.py
+<examples/text-classification/run_tf_glue.py>`, :prefix_link:`run_tf_text_classification.py
+<examples/text-classification/run_tf_text_classification.py>` or :prefix_link:`run_xnli.py
+<examples/text-classification/run_xnli.py>` scripts.
 
 Here is an example of using pipelines to do sentiment analysis: identifying if a sequence is positive or negative. It
 leverages a fine-tuned model on sst2, which is a GLUE task.
@@ -168,9 +167,8 @@ Extractive Question Answering
 
 Extractive Question Answering is the task of extracting an answer from a text given a question. An example of a
 question answering dataset is the SQuAD dataset, which is entirely based on that task. If you would like to fine-tune a
-model on a SQuAD task, you may leverage the `run_squad.py
-<https://github.com/huggingface/transformers/tree/master/examples/question-answering/run_squad.py>`__ and
-`run_tf_squad.py
+model on a SQuAD task, you may leverage the `run_qa.py
+<https://github.com/huggingface/transformers/tree/master/examples/question-answering/run_qa.py>`__ and `run_tf_squad.py
 <https://github.com/huggingface/transformers/tree/master/examples/question-answering/run_tf_squad.py>`__ scripts.
 
 
@@ -242,7 +240,6 @@ Here is an example of question answering using a model and a tokenizer. The proc
     ...     inputs = tokenizer(question, text, add_special_tokens=True, return_tensors="pt")
     ...     input_ids = inputs["input_ids"].tolist()[0]
     ...
-    ...     text_tokens = tokenizer.convert_ids_to_tokens(input_ids)
     ...     outputs = model(**inputs)
     ...     answer_start_scores = outputs.start_logits
     ...     answer_end_scores = outputs.end_logits
@@ -286,7 +283,6 @@ Here is an example of question answering using a model and a tokenizer. The proc
     ...     inputs = tokenizer(question, text, add_special_tokens=True, return_tensors="tf")
     ...     input_ids = inputs["input_ids"].numpy()[0]
     ...
-    ...     text_tokens = tokenizer.convert_ids_to_tokens(input_ids)
     ...     outputs = model(inputs)
     ...     answer_start_scores = outputs.start_logits
     ...     answer_end_scores = outputs.end_logits
@@ -328,7 +324,9 @@ Masked language modeling is the task of masking tokens in a sequence with a mask
 fill that mask with an appropriate token. This allows the model to attend to both the right context (tokens on the
 right of the mask) and the left context (tokens on the left of the mask). Such a training creates a strong basis for
 downstream tasks requiring bi-directional context, such as SQuAD (question answering, see `Lewis, Lui, Goyal et al.
-<https://arxiv.org/abs/1910.13461>`__, part 4.2).
+<https://arxiv.org/abs/1910.13461>`__, part 4.2). If you would like to fine-tune a model on a masked language modeling
+task, you may leverage the `run_mlm.py
+<https://github.com/huggingface/transformers/tree/master/examples/language-modeling/run_mlm.py>`__ script.
 
 Here is an example of using pipelines to replace a mask from a sequence:
 
@@ -436,7 +434,8 @@ Causal Language Modeling
 
 Causal language modeling is the task of predicting the token following a sequence of tokens. In this situation, the
 model only attends to the left context (tokens on the left of the mask). Such a training is particularly interesting
-for generation tasks.
+for generation tasks. If you would like to fine-tune a model on a causal language modeling task, you may leverage the
+`run_clm.py <https://github.com/huggingface/transformers/tree/master/examples/language-modeling/run_clm.py>`__ script.
 
 Usually, the next token is predicted by sampling from the logits of the last hidden state the model produces from the
 input sequence.
@@ -604,11 +603,7 @@ Named Entity Recognition (NER) is the task of classifying tokens according to a
 as a person, an organisation or a location. An example of a named entity recognition dataset is the CoNLL-2003 dataset,
 which is entirely based on that task. If you would like to fine-tune a model on an NER task, you may leverage the
 `run_ner.py <https://github.com/huggingface/transformers/tree/master/examples/token-classification/run_ner.py>`__
-(PyTorch), `run_pl_ner.py
-<https://github.com/huggingface/transformers/tree/master/examples/token-classification/run_pl_ner.py>`__ (leveraging
-pytorch-lightning) or the `run_tf_ner.py
-<https://github.com/huggingface/transformers/tree/master/examples/token-classification/run_tf_ner.py>`__ (TensorFlow)
-scripts.
+script.
 
 Here is an example of using pipelines to do named entity recognition, specifically, trying to identify tokens as
 belonging to one of 9 classes:
@@ -746,7 +741,9 @@ token. The following array should be the output:
 Summarization
 -----------------------------------------------------------------------------------------------------------------------
 
-Summarization is the task of summarizing a document or an article into a shorter text.
+Summarization is the task of summarizing a document or an article into a shorter text. If you would like to fine-tune a
+model on a summarization task, you may leverage the `run_summarization.py
+<https://github.com/huggingface/transformers/tree/master/examples/seq2seq/run_summarization.py>`__ script.
 
 An example of a summarization dataset is the CNN / Daily Mail dataset, which consists of long news articles and was
 created for the task of summarization. If you would like to fine-tune a model on a summarization task, various
@@ -824,7 +821,9 @@ CNN / Daily Mail), it yields very good results.
 Translation
 -----------------------------------------------------------------------------------------------------------------------
 
-Translation is the task of translating a text from one language to another.
+Translation is the task of translating a text from one language to another. If you would like to fine-tune a model on a
+translation task, you may leverage the `run_translation.py
+<https://github.com/huggingface/transformers/tree/master/examples/seq2seq/run_translation.py>`__ script.
 
 An example of a translation dataset is the WMT English to German dataset, which has sentences in English as the input
 data and the corresponding sentences in German as the target data. If you would like to fine-tune a model on a

examples/README.md

@@ -1,6 +1,5 @@
 <!---
 Copyright 2020 The HuggingFace Team. All rights reserved.
-
 Licensed under the Apache License, Version 2.0 (the "License");
 you may not use this file except in compliance with the License.
 You may obtain a copy of the License at
@@ -34,10 +33,43 @@ Then cd in the example folder of your choice and run
 pip install -r requirements.txt
 ```
 
-Alternatively, you can run the version of the examples as they were for your current version of Transformers via (for instance with v3.5.1):
+To browse the examples corresponding to released versions of 🤗 Transformers, click on the line below and then on your desired version of the library:
+
+<details>
+  <summary>Examples for older versions of 🤗 Transformers</summary>
+
+  - [v4.3.3](https://github.com/huggingface/transformers/tree/v4.3.3/examples)
+  - [v4.2.2](https://github.com/huggingface/transformers/tree/v4.2.2/examples)
+  - [v4.1.1](https://github.com/huggingface/transformers/tree/v4.1.1/examples)
+  - [v4.0.1](https://github.com/huggingface/transformers/tree/v4.0.1/examples)
+  - [v3.5.1](https://github.com/huggingface/transformers/tree/v3.5.1/examples)
+  - [v3.4.0](https://github.com/huggingface/transformers/tree/v3.4.0/examples)
+  - [v3.3.1](https://github.com/huggingface/transformers/tree/v3.3.1/examples)
+  - [v3.2.0](https://github.com/huggingface/transformers/tree/v3.2.0/examples)
+  - [v3.1.0](https://github.com/huggingface/transformers/tree/v3.1.0/examples)
+  - [v3.0.2](https://github.com/huggingface/transformers/tree/v3.0.2/examples)
+  - [v2.11.0](https://github.com/huggingface/transformers/tree/v2.11.0/examples)
+  - [v2.10.0](https://github.com/huggingface/transformers/tree/v2.10.0/examples)
+  - [v2.9.1](https://github.com/huggingface/transformers/tree/v2.9.1/examples)
+  - [v2.8.0](https://github.com/huggingface/transformers/tree/v2.8.0/examples)
+  - [v2.7.0](https://github.com/huggingface/transformers/tree/v2.7.0/examples)
+  - [v2.6.0](https://github.com/huggingface/transformers/tree/v2.6.0/examples)
+  - [v2.5.1](https://github.com/huggingface/transformers/tree/v2.5.1/examples)
+  - [v2.4.0](https://github.com/huggingface/transformers/tree/v2.4.0/examples)
+  - [v2.3.0](https://github.com/huggingface/transformers/tree/v2.3.0/examples)
+  - [v2.2.0](https://github.com/huggingface/transformers/tree/v2.2.0/examples)
+  - [v2.1.1](https://github.com/huggingface/transformers/tree/v2.1.0/examples)
+  - [v2.0.0](https://github.com/huggingface/transformers/tree/v2.0.0/examples)
+  - [v1.2.0](https://github.com/huggingface/transformers/tree/v1.2.0/examples)
+  - [v1.1.0](https://github.com/huggingface/transformers/tree/v1.1.0/examples)
+  - [v1.0.0](https://github.com/huggingface/transformers/tree/v1.0.0/examples)
+</details>
+
+Alternatively, you can find switch your cloned 🤗 Transformers to a specific version (for instance with v3.5.1) with
 ```bash
 git checkout tags/v3.5.1
 ```
+and run the example command as usual afterward.
 
 ## The Big Table of Tasks
 
@@ -55,20 +87,14 @@ Coming soon!
 |---|---|:---:|:---:|:---:|:---:|
 | [**`language-modeling`**](https://github.com/huggingface/transformers/tree/master/examples/language-modeling)       | Raw text        | ✅ | -  | ✅ | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/blog/blob/master/notebooks/01_how_to_train.ipynb)
 | [**`multiple-choice`**](https://github.com/huggingface/transformers/tree/master/examples/multiple-choice)           | SWAG, RACE, ARC | ✅ | ✅ | ✅ | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/ViktorAlm/notebooks/blob/master/MPC_GPU_Demo_for_TF_and_PT.ipynb)
-| [**`question-answering`**](https://github.com/huggingface/transformers/tree/master/examples/question-answering)     | SQuAD           | ✅ | ✅ | ✅ | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://github.com/huggingface/notebooks/blob/master/examples/question_answering.ipynb)
+| [**`question-answering`**](https://github.com/huggingface/transformers/tree/master/examples/question-answering)     | SQuAD           | ✅ | ✅ | ✅ | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/notebooks/blob/master/examples/question_answering.ipynb)
 | [**`summarization`**](https://github.com/huggingface/transformers/tree/master/examples/seq2seq)                     | CNN/Daily Mail  | ✅  | - | - | -
-| [**`text-classification`**](https://github.com/huggingface/transformers/tree/master/examples/text-classification)   | GLUE, XNLI      | ✅ | ✅ | ✅ | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://github.com/huggingface/notebooks/blob/master/examples/text_classification.ipynb)
+| [**`text-classification`**](https://github.com/huggingface/transformers/tree/master/examples/text-classification)   | GLUE, XNLI      | ✅ | ✅ | ✅ | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/notebooks/blob/master/examples/text_classification.ipynb)
 | [**`text-generation`**](https://github.com/huggingface/transformers/tree/master/examples/text-generation)           | -               | n/a | n/a | - | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/blog/blob/master/notebooks/02_how_to_generate.ipynb)
-| [**`token-classification`**](https://github.com/huggingface/transformers/tree/master/examples/token-classification) | CoNLL NER       | ✅ | ✅ | ✅ | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://github.com/huggingface/notebooks/blob/master/examples/token_classification.ipynb)
+| [**`token-classification`**](https://github.com/huggingface/transformers/tree/master/examples/token-classification) | CoNLL NER       | ✅ | ✅ | ✅ | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/notebooks/blob/master/examples/token_classification.ipynb)
 | [**`translation`**](https://github.com/huggingface/transformers/tree/master/examples/seq2seq)                       | WMT             | ✅  | - | - | -
 
 
-<!--
-## One-click Deploy to Cloud (wip)
-
-**Coming soon!**
--->
-
 ## Distributed training and mixed precision
 
 All the PyTorch scripts mentioned above work out of the box with distributed training and mixed precision, thanks to
@@ -147,7 +173,7 @@ python xla_spawn.py --num_cores 8 \
 You can easily log and monitor your runs code. The following are currently supported:
 
 * [TensorBoard](https://www.tensorflow.org/tensorboard)
-* [Weights & Biases](https://docs.wandb.com/library/integrations/huggingface)
+* [Weights & Biases](https://docs.wandb.ai/integrations/huggingface)
 * [Comet ML](https://www.comet.ml/docs/python-sdk/huggingface/)
 
 ### Weights & Biases
@@ -171,9 +197,46 @@ import wandb
 wandb.login()
 ```
 
+To enable logging to W&B, include `"wandb"` in the `report_to` of your `TrainingArguments` or script. Or just pass along `--report_to all` if you have `wandb` installed.
+
 Whenever you use `Trainer` or `TFTrainer` classes, your losses, evaluation metrics, model topology and gradients (for `Trainer` only) will automatically be logged.
 
-When using 🤗 Transformers with PyTorch Lightning, runs can be tracked through `WandbLogger`. Refer to related [documentation & examples](https://docs.wandb.com/library/integrations/lightning).
+Advanced configuration is possible by setting environment variables:
+
+<table>
+  <thead>
+    <tr>
+      <th style="text-align:left">Environment Variables</th>
+      <th style="text-align:left">Options</th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td style="text-align:left">WANDB_LOG_MODEL</td>
+      <td style="text-align:left">Log the model as artifact at the end of training (<b>false</b> by default)</td>
+    </tr>
+    <tr>
+      <td style="text-align:left">WANDB_WATCH</td>
+      <td style="text-align:left">
+        <ul>
+          <li><b>gradients</b> (default): Log histograms of the gradients</li>
+          <li><b>all</b>: Log histograms of gradients and parameters</li>
+          <li><b>false</b>: No gradient or parameter logging</li>
+        </ul>
+      </td>
+    </tr>
+    <tr>
+      <td style="text-align:left">WANDB_PROJECT</td>
+      <td style="text-align:left">Organize runs by project</td>
+    </tr>
+  </tbody>
+</table>
+
+Set run names with `run_name` argument present in scripts or as part of `TrainingArguments`.
+
+Additional configuration options are available through generic [wandb environment variables](https://docs.wandb.com/library/environment-variables).
+
+Refer to related [documentation & examples](https://docs.wandb.ai/integrations/huggingface).
 
 ### Comet.ml
 

examples/_tests_requirements.txt

@@ -2,7 +2,7 @@ tensorboard
 scikit-learn
 seqeval
 psutil
-sacrebleu
+sacrebleu >= 1.4.12
 rouge-score
 tensorflow_datasets
 matplotlib

examples/benchmarking/run_benchmark.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 # coding=utf-8
 # Copyright 2020 The HuggingFace Inc. team.
 # Copyright (c) 2018, NVIDIA CORPORATION.  All rights reserved.

examples/benchmarking/run_benchmark_tf.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 # coding=utf-8
 # Copyright 2018 The HuggingFace Inc. team.
 # Copyright (c) 2020, NVIDIA CORPORATION.  All rights reserved.

examples/language-modeling/README.md

@@ -100,72 +100,7 @@ sure all your batches have the same length.
 
 ### Whole word masking
 
-The BERT authors released a new version of BERT using Whole Word Masking in May 2019. Instead of masking randomly
-selected tokens (which may be part of words), they mask randomly selected words (masking all the tokens corresponding
-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 \
-    --dataset_config_name wikitext-2-raw-v1 \
-    --do_train \
-    --do_eval \
-    --output_dir /tmp/test-mlm-wwm
-```
-
-For Chinese models, we need to generate a reference files (which requires the ltp library), because it's tokenized at
-the character level.
-
-**Q :** Why a reference file?
-
-**A :** Suppose we have a Chinese sentence like: `我喜欢你` The original Chinese-BERT will tokenize it as
-`['我','喜','欢','你']` (character level). But `喜欢` is a whole word. For whole word masking proxy, we need a result
-like `['我','喜','##欢','你']`, so we need a reference file to tell the model which position of the BERT original token
-should be added `##`.
-
-**Q :** Why LTP ?
-
-**A :** Cause the best known Chinese WWM BERT is [Chinese-BERT-wwm](https://github.com/ymcui/Chinese-BERT-wwm) by HIT.
-It works well on so many Chines Task like CLUE (Chinese GLUE). They use LTP, so if we want to fine-tune their model,
-we need LTP.
-
-Now LTP only only works well on `transformers==3.2.0`. So we don't add it to requirements.txt.
-You need to create a separate environment with this version of Transformers to run the `run_chinese_ref.py` script that
-will create the reference files. The script is in `examples/contrib`. Once in the proper environment, run the
-following:
-
-
-```bash
-export TRAIN_FILE=/path/to/dataset/wiki.train.raw
-export LTP_RESOURCE=/path/to/ltp/tokenizer
-export BERT_RESOURCE=/path/to/bert/tokenizer
-export SAVE_PATH=/path/to/data/ref.txt
-
-python examples/contrib/run_chinese_ref.py \
-    --file_name=path_to_train_or_eval_file \
-    --ltp=path_to_ltp_tokenizer \
-    --bert=path_to_bert_tokenizer \
-    --save_path=path_to_reference_file
-```
-
-Then you can run the script like this: 
-
-
-```bash
-python run_mlm_wwm.py \
-    --model_name_or_path roberta-base \
-    --train_file path_to_train_file \
-    --validation_file path_to_validation_file \
-    --train_ref_file path_to_train_chinese_ref_file \
-    --validation_ref_file path_to_validation_chinese_ref_file \
-    --do_train \
-    --do_eval \
-    --output_dir /tmp/test-mlm-wwm
-```
-
-**Note:** On TPU, you should the flag `--pad_to_max_length` to make sure all your batches have the same length.
+This part was moved to `examples/research_projects/mlm_wwm`. 
 
 ### XLNet and permutation language modeling
 

examples/language-modeling/run_clm.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 # coding=utf-8
 # Copyright 2020 The HuggingFace Inc. team. All rights reserved.
 #
@@ -42,9 +43,13 @@ from transformers import (
     default_data_collator,
     set_seed,
 )
-from transformers.trainer_utils import is_main_process
+from transformers.trainer_utils import get_last_checkpoint, is_main_process
+from transformers.utils import check_min_version
 
 
+# Will error if the minimal version of Transformers is not installed. Remove at your own risks.
+check_min_version("4.4.0")
+
 logger = logging.getLogger(__name__)
 
 
@@ -113,6 +118,21 @@ class DataTrainingArguments:
         default=None,
         metadata={"help": "An optional input evaluation data file to evaluate the perplexity on (a text file)."},
     )
+    max_train_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of training examples to this "
+            "value if set."
+        },
+    )
+    max_val_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of validation examples to this "
+            "value if set."
+        },
+    )
+
     block_size: Optional[int] = field(
         default=None,
         metadata={
@@ -160,23 +180,28 @@ def main():
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
 
-    if (
-        os.path.exists(training_args.output_dir)
-        and os.listdir(training_args.output_dir)
-        and training_args.do_train
-        and not training_args.overwrite_output_dir
-    ):
-        raise ValueError(
-            f"Output directory ({training_args.output_dir}) already exists and is not empty."
-            "Use --overwrite_output_dir to overcome."
-        )
+    # Detecting last checkpoint.
+    last_checkpoint = None
+    if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir:
+        last_checkpoint = get_last_checkpoint(training_args.output_dir)
+        if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0:
+            raise ValueError(
+                f"Output directory ({training_args.output_dir}) already exists and is not empty. "
+                "Use --overwrite_output_dir to overcome."
+            )
+        elif last_checkpoint is not None:
+            logger.info(
+                f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change "
+                "the `--output_dir` or add `--overwrite_output_dir` to train from scratch."
+            )
 
     # Setup logging
     logging.basicConfig(
         format="%(asctime)s - %(levelname)s - %(name)s -   %(message)s",
         datefmt="%m/%d/%Y %H:%M:%S",
-        level=logging.INFO if is_main_process(training_args.local_rank) else logging.WARN,
+        handlers=[logging.StreamHandler(sys.stdout)],
     )
+    logger.setLevel(logging.INFO if is_main_process(training_args.local_rank) else logging.WARN)
 
     # Log on each process the small summary:
     logger.warning(
@@ -222,7 +247,11 @@ def main():
             data_files["train"] = data_args.train_file
         if data_args.validation_file is not None:
             data_files["validation"] = data_args.validation_file
-        extension = data_args.train_file.split(".")[-1]
+        extension = (
+            data_args.train_file.split(".")[-1]
+            if data_args.train_file is not None
+            else data_args.validation_file.split(".")[-1]
+        )
         if extension == "txt":
             extension = "text"
         datasets = load_dataset(extension, data_files=data_files)
@@ -336,6 +365,7 @@ def main():
     #
     # To speed up this part, we use multiprocessing. See the documentation of the map method for more information:
     # https://huggingface.co/docs/datasets/package_reference/main_classes.html#datasets.Dataset.map
+
     lm_datasets = tokenized_datasets.map(
         group_texts,
         batched=True,
@@ -343,12 +373,26 @@ def main():
         load_from_cache_file=not data_args.overwrite_cache,
     )
 
+    if training_args.do_train:
+        if "train" not in tokenized_datasets:
+            raise ValueError("--do_train requires a train dataset")
+        train_dataset = lm_datasets["train"]
+        if data_args.max_train_samples is not None:
+            train_dataset = train_dataset.select(range(data_args.max_train_samples))
+
+    if training_args.do_eval:
+        if "validation" not in tokenized_datasets:
+            raise ValueError("--do_eval requires a validation dataset")
+        eval_dataset = lm_datasets["validation"]
+        if data_args.max_val_samples is not None:
+            eval_dataset = eval_dataset.select(range(data_args.max_val_samples))
+
     # Initialize our Trainer
     trainer = Trainer(
         model=model,
         args=training_args,
-        train_dataset=lm_datasets["train"] if training_args.do_train else None,
-        eval_dataset=lm_datasets["validation"] if training_args.do_eval else None,
+        train_dataset=train_dataset if training_args.do_train else None,
+        eval_dataset=eval_dataset if training_args.do_eval else None,
         tokenizer=tokenizer,
         # Data collator will default to DataCollatorWithPadding, so we change it.
         data_collator=default_data_collator,
@@ -356,44 +400,39 @@ def main():
 
     # Training
     if training_args.do_train:
-        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
-        )
-        train_result = trainer.train(model_path=model_path)
+        if last_checkpoint is not None:
+            checkpoint = last_checkpoint
+        elif model_args.model_name_or_path is not None and os.path.isdir(model_args.model_name_or_path):
+            checkpoint = model_args.model_name_or_path
+        else:
+            checkpoint = None
+        train_result = trainer.train(resume_from_checkpoint=checkpoint)
         trainer.save_model()  # Saves the tokenizer too for easy upload
 
-        output_train_file = os.path.join(training_args.output_dir, "train_results.txt")
-        if trainer.is_world_process_zero():
-            with open(output_train_file, "w") as writer:
-                logger.info("***** Train results *****")
-                for key, value in sorted(train_result.metrics.items()):
-                    logger.info(f"  {key} = {value}")
-                    writer.write(f"{key} = {value}\n")
+        metrics = train_result.metrics
 
-            # Need to save the state, since Trainer.save_model saves only the tokenizer with the model
-            trainer.state.save_to_json(os.path.join(training_args.output_dir, "trainer_state.json"))
+        max_train_samples = (
+            data_args.max_train_samples if data_args.max_train_samples is not None else len(train_dataset)
+        )
+        metrics["train_samples"] = min(max_train_samples, len(train_dataset))
+
+        trainer.log_metrics("train", metrics)
+        trainer.save_metrics("train", metrics)
+        trainer.save_state()
 
     # Evaluation
-    results = {}
     if training_args.do_eval:
         logger.info("*** Evaluate ***")
 
-        eval_output = trainer.evaluate()
+        metrics = trainer.evaluate()
 
-        perplexity = math.exp(eval_output["eval_loss"])
-        results["perplexity"] = perplexity
+        max_val_samples = data_args.max_val_samples if data_args.max_val_samples is not None else len(eval_dataset)
+        metrics["eval_samples"] = min(max_val_samples, len(eval_dataset))
+        perplexity = math.exp(metrics["eval_loss"])
+        metrics["perplexity"] = perplexity
 
-        output_eval_file = os.path.join(training_args.output_dir, "eval_results_clm.txt")
-        if trainer.is_world_process_zero():
-            with open(output_eval_file, "w") as writer:
-                logger.info("***** Eval results *****")
-                for key, value in sorted(results.items()):
-                    logger.info(f"  {key} = {value}")
-                    writer.write(f"{key} = {value}\n")
-
-    return results
+        trainer.log_metrics("eval", metrics)
+        trainer.save_metrics("eval", metrics)
 
 
 def _mp_fn(index):

examples/language-modeling/run_mlm.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 # coding=utf-8
 # Copyright 2020 The HuggingFace Team All rights reserved.
 #
@@ -42,9 +43,13 @@ from transformers import (
     TrainingArguments,
     set_seed,
 )
-from transformers.trainer_utils import is_main_process
+from transformers.trainer_utils import get_last_checkpoint, is_main_process
+from transformers.utils import check_min_version
 
 
+# Will error if the minimal version of Transformers is not installed. Remove at your own risks.
+check_min_version("4.4.0")
+
 logger = logging.getLogger(__name__)
 MODEL_CONFIG_CLASSES = list(MODEL_FOR_MASKED_LM_MAPPING.keys())
 MODEL_TYPES = tuple(conf.model_type for conf in MODEL_CONFIG_CLASSES)
@@ -145,6 +150,20 @@ class DataTrainingArguments:
             "If False, will pad the samples dynamically when batching to the maximum length in the batch."
         },
     )
+    max_train_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of training examples to this "
+            "value if set."
+        },
+    )
+    max_val_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of validation examples to this "
+            "value if set."
+        },
+    )
 
     def __post_init__(self):
         if self.dataset_name is None and self.train_file is None and self.validation_file is None:
@@ -171,23 +190,28 @@ def main():
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
 
-    if (
-        os.path.exists(training_args.output_dir)
-        and os.listdir(training_args.output_dir)
-        and training_args.do_train
-        and not training_args.overwrite_output_dir
-    ):
-        raise ValueError(
-            f"Output directory ({training_args.output_dir}) already exists and is not empty."
-            "Use --overwrite_output_dir to overcome."
-        )
+    # Detecting last checkpoint.
+    last_checkpoint = None
+    if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir:
+        last_checkpoint = get_last_checkpoint(training_args.output_dir)
+        if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0:
+            raise ValueError(
+                f"Output directory ({training_args.output_dir}) already exists and is not empty. "
+                "Use --overwrite_output_dir to overcome."
+            )
+        elif last_checkpoint is not None:
+            logger.info(
+                f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change "
+                "the `--output_dir` or add `--overwrite_output_dir` to train from scratch."
+            )
 
     # Setup logging
     logging.basicConfig(
         format="%(asctime)s - %(levelname)s - %(name)s -   %(message)s",
         datefmt="%m/%d/%Y %H:%M:%S",
-        level=logging.INFO if is_main_process(training_args.local_rank) else logging.WARN,
+        handlers=[logging.StreamHandler(sys.stdout)],
     )
+    logger.setLevel(logging.INFO if is_main_process(training_args.local_rank) else logging.WARN)
 
     # Log on each process the small summary:
     logger.warning(
@@ -297,6 +321,22 @@ def main():
         column_names = datasets["validation"].column_names
     text_column_name = "text" if "text" in column_names else column_names[0]
 
+    if data_args.max_seq_length is None:
+        max_seq_length = tokenizer.model_max_length
+        if max_seq_length > 1024:
+            logger.warn(
+                f"The tokenizer picked seems to have a very large `model_max_length` ({tokenizer.model_max_length}). "
+                "Picking 1024 instead. You can change that default value by passing --max_seq_length xxx."
+            )
+            max_seq_length = 1024
+    else:
+        if data_args.max_seq_length > tokenizer.model_max_length:
+            logger.warn(
+                f"The max_seq_length passed ({data_args.max_seq_length}) is larger than the maximum length for the"
+                f"model ({tokenizer.model_max_length}). Using max_seq_length={tokenizer.model_max_length}."
+            )
+        max_seq_length = min(data_args.max_seq_length, tokenizer.model_max_length)
+
     if data_args.line_by_line:
         # When using line_by_line, we just tokenize each nonempty line.
         padding = "max_length" if data_args.pad_to_max_length else False
@@ -308,7 +348,7 @@ def main():
                 examples["text"],
                 padding=padding,
                 truncation=True,
-                max_length=data_args.max_seq_length,
+                max_length=max_seq_length,
                 # We use this option because DataCollatorForLanguageModeling (see below) is more efficient when it
                 # receives the `special_tokens_mask`.
                 return_special_tokens_mask=True,
@@ -336,16 +376,6 @@ def main():
             load_from_cache_file=not data_args.overwrite_cache,
         )
 
-        if data_args.max_seq_length is None:
-            max_seq_length = tokenizer.model_max_length
-        else:
-            if data_args.max_seq_length > tokenizer.model_max_length:
-                logger.warn(
-                    f"The max_seq_length passed ({data_args.max_seq_length}) is larger than the maximum length for the"
-                    f"model ({tokenizer.model_max_length}). Using max_seq_length={tokenizer.model_max_length}."
-                )
-            max_seq_length = min(data_args.max_seq_length, tokenizer.model_max_length)
-
         # Main data processing function that will concatenate all texts from our dataset and generate chunks of
         # max_seq_length.
         def group_texts(examples):
@@ -368,6 +398,7 @@ def main():
         #
         # To speed up this part, we use multiprocessing. See the documentation of the map method for more information:
         # https://huggingface.co/docs/datasets/package_reference/main_classes.html#datasets.Dataset.map
+
         tokenized_datasets = tokenized_datasets.map(
             group_texts,
             batched=True,
@@ -375,6 +406,20 @@ def main():
             load_from_cache_file=not data_args.overwrite_cache,
         )
 
+    if training_args.do_train:
+        if "train" not in tokenized_datasets:
+            raise ValueError("--do_train requires a train dataset")
+        train_dataset = tokenized_datasets["train"]
+        if data_args.max_train_samples is not None:
+            train_dataset = train_dataset.select(range(data_args.max_train_samples))
+
+    if training_args.do_eval:
+        if "validation" not in tokenized_datasets:
+            raise ValueError("--do_eval requires a validation dataset")
+        eval_dataset = tokenized_datasets["validation"]
+        if data_args.max_val_samples is not None:
+            eval_dataset = eval_dataset.select(range(data_args.max_val_samples))
+
     # Data collator
     # This one will take care of randomly masking the tokens.
     data_collator = DataCollatorForLanguageModeling(tokenizer=tokenizer, mlm_probability=data_args.mlm_probability)
@@ -383,52 +428,46 @@ def main():
     trainer = Trainer(
         model=model,
         args=training_args,
-        train_dataset=tokenized_datasets["train"] if training_args.do_train else None,
-        eval_dataset=tokenized_datasets["validation"] if training_args.do_eval else None,
+        train_dataset=train_dataset if training_args.do_train else None,
+        eval_dataset=eval_dataset if training_args.do_eval else None,
         tokenizer=tokenizer,
         data_collator=data_collator,
     )
 
     # Training
     if training_args.do_train:
-        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
-        )
-        train_result = trainer.train(model_path=model_path)
+        if last_checkpoint is not None:
+            checkpoint = last_checkpoint
+        elif model_args.model_name_or_path is not None and os.path.isdir(model_args.model_name_or_path):
+            checkpoint = model_args.model_name_or_path
+        else:
+            checkpoint = None
+        train_result = trainer.train(resume_from_checkpoint=checkpoint)
         trainer.save_model()  # Saves the tokenizer too for easy upload
+        metrics = train_result.metrics
 
-        output_train_file = os.path.join(training_args.output_dir, "train_results.txt")
-        if trainer.is_world_process_zero():
-            with open(output_train_file, "w") as writer:
-                logger.info("***** Train results *****")
-                for key, value in sorted(train_result.metrics.items()):
-                    logger.info(f"  {key} = {value}")
-                    writer.write(f"{key} = {value}\n")
+        max_train_samples = (
+            data_args.max_train_samples if data_args.max_train_samples is not None else len(train_dataset)
+        )
+        metrics["train_samples"] = min(max_train_samples, len(train_dataset))
 
-            # Need to save the state, since Trainer.save_model saves only the tokenizer with the model
-            trainer.state.save_to_json(os.path.join(training_args.output_dir, "trainer_state.json"))
+        trainer.log_metrics("train", metrics)
+        trainer.save_metrics("train", metrics)
+        trainer.save_state()
 
     # Evaluation
-    results = {}
     if training_args.do_eval:
         logger.info("*** Evaluate ***")
 
-        eval_output = trainer.evaluate()
+        metrics = trainer.evaluate()
 
-        perplexity = math.exp(eval_output["eval_loss"])
-        results["perplexity"] = perplexity
+        max_val_samples = data_args.max_val_samples if data_args.max_val_samples is not None else len(eval_dataset)
+        metrics["eval_samples"] = min(max_val_samples, len(eval_dataset))
+        perplexity = math.exp(metrics["eval_loss"])
+        metrics["perplexity"] = perplexity
 
-        output_eval_file = os.path.join(training_args.output_dir, "eval_results_mlm.txt")
-        if trainer.is_world_process_zero():
-            with open(output_eval_file, "w") as writer:
-                logger.info("***** Eval results *****")
-                for key, value in sorted(results.items()):
-                    logger.info(f"  {key} = {value}")
-                    writer.write(f"{key} = {value}\n")
-
-    return results
+        trainer.log_metrics("eval", metrics)
+        trainer.save_metrics("eval", metrics)
 
 
 def _mp_fn(index):

examples/language-modeling/run_mlm_flax.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 # coding=utf-8
 # Copyright 2020 The HuggingFace Team All rights reserved.
 #

examples/language-modeling/run_plm.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 # coding=utf-8
 # Copyright 2020 The HuggingFace Team All rights reserved.
 #
@@ -38,9 +39,13 @@ from transformers import (
     XLNetLMHeadModel,
     set_seed,
 )
-from transformers.trainer_utils import is_main_process
+from transformers.trainer_utils import get_last_checkpoint, is_main_process
+from transformers.utils import check_min_version
 
 
+# Will error if the minimal version of Transformers is not installed. Remove at your own risks.
+check_min_version("4.4.0")
+
 logger = logging.getLogger(__name__)
 
 
@@ -142,6 +147,20 @@ class DataTrainingArguments:
             "If False, will pad the samples dynamically when batching to the maximum length in the batch."
         },
     )
+    max_train_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of training examples to this "
+            "value if set."
+        },
+    )
+    max_val_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of validation examples to this "
+            "value if set."
+        },
+    )
 
     def __post_init__(self):
         if self.dataset_name is None and self.train_file is None and self.validation_file is None:
@@ -168,23 +187,28 @@ def main():
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
 
-    if (
-        os.path.exists(training_args.output_dir)
-        and os.listdir(training_args.output_dir)
-        and training_args.do_train
-        and not training_args.overwrite_output_dir
-    ):
-        raise ValueError(
-            f"Output directory ({training_args.output_dir}) already exists and is not empty."
-            "Use --overwrite_output_dir to overcome."
-        )
+    # Detecting last checkpoint.
+    last_checkpoint = None
+    if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir:
+        last_checkpoint = get_last_checkpoint(training_args.output_dir)
+        if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0:
+            raise ValueError(
+                f"Output directory ({training_args.output_dir}) already exists and is not empty. "
+                "Use --overwrite_output_dir to overcome."
+            )
+        elif last_checkpoint is not None:
+            logger.info(
+                f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change "
+                "the `--output_dir` or add `--overwrite_output_dir` to train from scratch."
+            )
 
     # Setup logging
     logging.basicConfig(
         format="%(asctime)s - %(levelname)s - %(name)s -   %(message)s",
         datefmt="%m/%d/%Y %H:%M:%S",
-        level=logging.INFO if is_main_process(training_args.local_rank) else logging.WARN,
+        handlers=[logging.StreamHandler(sys.stdout)],
     )
+    logger.setLevel(logging.INFO if is_main_process(training_args.local_rank) else logging.WARN)
 
     # Log on each process the small summary:
     logger.warning(
@@ -294,6 +318,13 @@ def main():
         column_names = datasets["validation"].column_names
     text_column_name = "text" if "text" in column_names else column_names[0]
 
+    if data_args.max_seq_length > tokenizer.model_max_length:
+        logger.warn(
+            f"The max_seq_length passed ({data_args.max_seq_length}) is larger than the maximum length for the"
+            f"model ({tokenizer.model_max_length}). Using max_seq_length={tokenizer.model_max_length}."
+        )
+    max_seq_length = min(data_args.max_seq_length, tokenizer.model_max_length)
+
     if data_args.line_by_line:
         # When using line_by_line, we just tokenize each nonempty line.
         padding = "max_length" if data_args.pad_to_max_length else False
@@ -301,7 +332,7 @@ def main():
         def tokenize_function(examples):
             # Remove empty lines
             examples["text"] = [line for line in examples["text"] if len(line) > 0 and not line.isspace()]
-            return tokenizer(examples["text"], padding=padding, truncation=True, max_length=data_args.max_seq_length)
+            return tokenizer(examples["text"], padding=padding, truncation=True, max_length=max_seq_length)
 
         tokenized_datasets = datasets.map(
             tokenize_function,
@@ -323,13 +354,6 @@ def main():
             load_from_cache_file=not data_args.overwrite_cache,
         )
 
-        if data_args.max_seq_length > tokenizer.model_max_length:
-            logger.warn(
-                f"The max_seq_length passed ({data_args.max_seq_length}) is larger than the maximum length for the"
-                f"model ({tokenizer.model_max_length}). Using max_seq_length={tokenizer.model_max_length}."
-            )
-        max_seq_length = min(data_args.max_seq_length, tokenizer.model_max_length)
-
         # Main data processing function that will concatenate all texts from our dataset and generate chunks of
         # max_seq_length.
         def group_texts(examples):
@@ -352,6 +376,7 @@ def main():
         #
         # To speed up this part, we use multiprocessing. See the documentation of the map method for more information:
         # https://huggingface.co/docs/datasets/package_reference/main_classes.html#datasets.Dataset.map
+
         tokenized_datasets = tokenized_datasets.map(
             group_texts,
             batched=True,
@@ -359,6 +384,20 @@ def main():
             load_from_cache_file=not data_args.overwrite_cache,
         )
 
+    if training_args.do_train:
+        if "train" not in tokenized_datasets:
+            raise ValueError("--do_train requires a train dataset")
+        train_dataset = tokenized_datasets["train"]
+        if data_args.max_train_samples is not None:
+            train_dataset = train_dataset.select(range(data_args.max_train_samples))
+
+    if training_args.do_eval:
+        if "validation" not in tokenized_datasets:
+            raise ValueError("--do_eval requires a validation dataset")
+        eval_dataset = tokenized_datasets["validation"]
+        if data_args.max_val_samples is not None:
+            eval_dataset = eval_dataset.select(range(data_args.max_val_samples))
+
     # Data collator
     data_collator = DataCollatorForPermutationLanguageModeling(
         tokenizer=tokenizer,
@@ -370,52 +409,46 @@ def main():
     trainer = Trainer(
         model=model,
         args=training_args,
-        train_dataset=tokenized_datasets["train"] if training_args.do_train else None,
-        eval_dataset=tokenized_datasets["validation"] if training_args.do_eval else None,
+        train_dataset=train_dataset if training_args.do_train else None,
+        eval_dataset=eval_dataset if training_args.do_eval else None,
         tokenizer=tokenizer,
         data_collator=data_collator,
     )
 
     # Training
     if training_args.do_train:
-        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
-        )
-        train_result = trainer.train(model_path=model_path)
+        if last_checkpoint is not None:
+            checkpoint = last_checkpoint
+        elif model_args.model_name_or_path is not None and os.path.isdir(model_args.model_name_or_path):
+            checkpoint = model_args.model_name_or_path
+        else:
+            checkpoint = None
+        train_result = trainer.train(resume_from_checkpoint=checkpoint)
         trainer.save_model()  # Saves the tokenizer too for easy upload
+        metrics = train_result.metrics
 
-        output_train_file = os.path.join(training_args.output_dir, "train_results.txt")
-        if trainer.is_world_process_zero():
-            with open(output_train_file, "w") as writer:
-                logger.info("***** Train results *****")
-                for key, value in sorted(train_result.metrics.items()):
-                    logger.info(f"  {key} = {value}")
-                    writer.write(f"{key} = {value}\n")
+        max_train_samples = (
+            data_args.max_train_samples if data_args.max_train_samples is not None else len(train_dataset)
+        )
+        metrics["train_samples"] = min(max_train_samples, len(train_dataset))
 
-            # Need to save the state, since Trainer.save_model saves only the tokenizer with the model
-            trainer.state.save_to_json(os.path.join(training_args.output_dir, "trainer_state.json"))
+        trainer.log_metrics("train", metrics)
+        trainer.save_metrics("train", metrics)
+        trainer.save_state()
 
     # Evaluation
-    results = {}
     if training_args.do_eval:
         logger.info("*** Evaluate ***")
 
-        eval_output = trainer.evaluate()
+        metrics = trainer.evaluate()
 
-        perplexity = math.exp(eval_output["eval_loss"])
-        results["perplexity"] = perplexity
+        max_val_samples = data_args.max_val_samples if data_args.max_val_samples is not None else len(eval_dataset)
+        metrics["eval_samples"] = min(max_val_samples, len(eval_dataset))
+        perplexity = math.exp(metrics["eval_loss"])
+        metrics["perplexity"] = perplexity
 
-        output_eval_file = os.path.join(training_args.output_dir, "eval_results_plm.txt")
-        if trainer.is_world_process_zero():
-            with open(output_eval_file, "w") as writer:
-                logger.info("***** Eval results *****")
-                for key, value in sorted(results.items()):
-                    logger.info(f"  {key} = {value}")
-                    writer.write(f"{key} = {value}\n")
-
-    return results
+        trainer.log_metrics("eval", metrics)
+        trainer.save_metrics("eval", metrics)
 
 
 def _mp_fn(index):

examples/legacy/multiple_choice/run_multiple_choice.py

@@ -28,6 +28,7 @@ from transformers import (
     AutoConfig,
     AutoModelForMultipleChoice,
     AutoTokenizer,
+    DataCollatorWithPadding,
     EvalPrediction,
     HfArgumentParser,
     Trainer,
@@ -188,6 +189,9 @@ def main():
         preds = np.argmax(p.predictions, axis=1)
         return {"acc": simple_accuracy(preds, p.label_ids)}
 
+    # Data collator
+    data_collator = DataCollatorWithPadding(tokenizer, pad_to_multiple_of=8) if training_args.fp16 else None
+
     # Initialize our Trainer
     trainer = Trainer(
         model=model,
@@ -195,6 +199,7 @@ def main():
         train_dataset=train_dataset,
         eval_dataset=eval_dataset,
         compute_metrics=compute_metrics,
+        data_collator=data_collator,
     )
 
     # Training

examples/legacy/question-answering/run_squad_trainer.py

@@ -23,7 +23,14 @@ from dataclasses import dataclass, field
 from typing import Optional
 
 import transformers
-from transformers import AutoConfig, AutoModelForQuestionAnswering, AutoTokenizer, HfArgumentParser, SquadDataset
+from transformers import (
+    AutoConfig,
+    AutoModelForQuestionAnswering,
+    AutoTokenizer,
+    DataCollatorWithPadding,
+    HfArgumentParser,
+    SquadDataset,
+)
 from transformers import SquadDataTrainingArguments as DataTrainingArguments
 from transformers import Trainer, TrainingArguments
 from transformers.trainer_utils import is_main_process
@@ -145,12 +152,16 @@ def main():
         else None
     )
 
+    # Data collator
+    data_collator = DataCollatorWithPadding(tokenizer, pad_to_multiple_of=8) if training_args.fp16 else None
+
     # Initialize our Trainer
     trainer = Trainer(
         model=model,
         args=training_args,
         train_dataset=train_dataset,
         eval_dataset=eval_dataset,
+        data_collator=data_collator,
     )
 
     # Training

examples/legacy/run_camembert.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 import torch
 
 from transformers import CamembertForMaskedLM, CamembertTokenizer

examples/legacy/run_chinese_ref.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 import argparse
 import json
 from typing import List

examples/legacy/run_language_modeling.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 # coding=utf-8
 # Copyright 2018 The Google AI Language Team Authors and The HuggingFace Inc. team.
 # Copyright (c) 2018, NVIDIA CORPORATION.  All rights reserved.

examples/legacy/run_openai_gpt.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 # coding=utf-8
 # Copyright 2018 Google AI, Google Brain and Carnegie Mellon University Authors and the HuggingFace Inc. team.
 # Copyright (c) 2018, NVIDIA CORPORATION.  All rights reserved.

examples/legacy/run_swag.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 # coding=utf-8
 # Copyright 2018 The Google AI Language Team Authors and The HuggingFace Inc. team.
 # Copyright (c) 2018, NVIDIA CORPORATION.  All rights reserved.

examples/legacy/run_transfo_xl.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 # coding=utf-8
 # Copyright 2018 Google AI, Google Brain and Carnegie Mellon University Authors and the HuggingFace Inc. team.
 # Copyright (c) 2018, NVIDIA CORPORATION.  All rights reserved.

examples/legacy/seq2seq/README.md

@@ -0,0 +1,334 @@
+<!---
+Copyright 2020 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+-->
+
+# Sequence-to-Sequence Training and Evaluation
+
+This directory contains examples for finetuning and evaluating transformers on summarization and translation tasks.
+For deprecated `bertabs` instructions, see [`bertabs/README.md`](https://github.com/huggingface/transformers/blob/master/examples/research_projects/bertabs/README.md).
+
+### Supported Architectures
+
+- `BartForConditionalGeneration`
+- `MarianMTModel`
+- `PegasusForConditionalGeneration`
+- `MBartForConditionalGeneration`
+- `FSMTForConditionalGeneration`
+- `T5ForConditionalGeneration`
+
+### Downlowd the Datasets
+
+#### XSUM
+
+```bash
+cd examples/legacy/seq2seq
+wget https://cdn-datasets.huggingface.co/summarization/xsum.tar.gz
+tar -xzvf xsum.tar.gz
+export XSUM_DIR=${PWD}/xsum
+```
+this should make a directory called `xsum/` with files like `test.source`.
+To use your own data, copy that files format. Each article to be summarized is on its own line.
+
+#### CNN/DailyMail
+
+```bash
+cd examples/legacy/seq2seq
+wget https://cdn-datasets.huggingface.co/summarization/cnn_dm_v2.tgz
+tar -xzvf cnn_dm_v2.tgz  # empty lines removed
+mv cnn_cln cnn_dm
+export CNN_DIR=${PWD}/cnn_dm
+```
+this should make a directory called `cnn_dm/` with 6 files.
+
+#### WMT16 English-Romanian Translation Data
+
+download with this command:
+```bash
+wget https://cdn-datasets.huggingface.co/translation/wmt_en_ro.tar.gz
+tar -xzvf wmt_en_ro.tar.gz
+export ENRO_DIR=${PWD}/wmt_en_ro
+```
+this should make a directory called `wmt_en_ro/` with 6 files.
+
+#### WMT English-German
+
+```bash
+wget https://cdn-datasets.huggingface.co/translation/wmt_en_de.tgz
+tar -xzvf wmt_en_de.tgz
+export DATA_DIR=${PWD}/wmt_en_de
+```
+
+#### FSMT datasets (wmt)
+
+Refer to the scripts starting with `eval_` under:
+https://github.com/huggingface/transformers/tree/master/scripts/fsmt
+
+#### Pegasus (multiple datasets)
+
+Multiple eval datasets are available for download from:
+https://github.com/stas00/porting/tree/master/datasets/pegasus
+
+
+#### Your Data
+
+If you are using your own data, it must be formatted as one directory with 6 files:
+```
+train.source
+train.target
+val.source
+val.target
+test.source
+test.target
+```
+The `.source` files are the input, the `.target` files are the desired output.
+
+### Potential issues
+
+- native AMP (`--fp16` and no apex) may lead to a huge memory leak and require 10x gpu memory. This has been fixed in pytorch-nightly and the minimal official version to have this fix will be pytorch-1.7.1. Until then if you have to use mixed precision please use AMP only with pytorch-nightly or NVIDIA's apex. Reference: https://github.com/huggingface/transformers/issues/8403
+
+
+### Tips and Tricks
+
+General Tips:
+- since you need to run from `examples/legacy/seq2seq`, and likely need to modify code, the easiest workflow is fork transformers, clone your fork, and run `pip install -e .` before you get started.
+- try `--freeze_encoder` or `--freeze_embeds` for faster training/larger batch size.  (3hr per epoch with bs=8, see the "xsum_shared_task" command below)
+- `fp16_opt_level=O1` (the default works best).
+- In addition to the pytorch-lightning .ckpt checkpoint, a transformers checkpoint will be saved.
+Load it with `BartForConditionalGeneration.from_pretrained(f'{output_dir}/best_tfmr)`.
+- At the moment, `--do_predict` does not work in a multi-gpu setting. You need to use `evaluate_checkpoint` or the `run_eval.py` code.
+- This warning can be safely ignored:
+    > "Some weights of BartForConditionalGeneration were not initialized from the model checkpoint at facebook/bart-large-xsum and are newly initialized: ['final_logits_bias']"
+- Both finetuning and eval are 30% faster with `--fp16`. For that you need to [install apex](https://github.com/NVIDIA/apex#quick-start).
+- Read scripts before you run them!
+
+Summarization Tips:
+- (summ) 1 epoch at batch size 1 for bart-large takes 24 hours and requires 13GB GPU RAM with fp16 on an NVIDIA-V100.
+- If you want to run experiments on improving the summarization finetuning process, try the XSUM Shared Task (below). It's faster to train than CNNDM because the summaries are shorter.
+- For CNN/DailyMail, the default `val_max_target_length` and `test_max_target_length` will truncate the ground truth labels, resulting in slightly higher rouge scores. To get accurate rouge scores, you should rerun calculate_rouge on the `{output_dir}/test_generations.txt` file saved by `trainer.test()`
+- `--max_target_length=60 --val_max_target_length=60 --test_max_target_length=100 ` is a reasonable setting for XSUM.
+- `wandb` can be used by specifying `--logger_name wandb`. It is useful for reproducibility. Specify the environment variable `WANDB_PROJECT='hf_xsum'` to do the XSUM shared task.
+- If you are finetuning on your own dataset, start from `distilbart-cnn-12-6` if you want long summaries and `distilbart-xsum-12-6` if you want short summaries.
+(It rarely makes sense to start from `bart-large` unless you are a researching finetuning methods).
+
+**Update 2018-07-18**
+Datasets: `LegacySeq2SeqDataset` will be used for all tokenizers without a `prepare_seq2seq_batch` method. Otherwise, `Seq2SeqDataset` will be used.
+Future work/help wanted: A new dataset to support multilingual tasks.
+
+
+### Fine-tuning using Seq2SeqTrainer
+To use `Seq2SeqTrainer` for fine-tuning you should use the `finetune_trainer.py` script. It subclasses `Trainer` to extend it for seq2seq training. Except the `Trainer`-related `TrainingArguments`, it shares the same argument names as that of `finetune.py` file. One notable difference is that calculating generative metrics (BLEU, ROUGE) is optional and is controlled using the `--predict_with_generate` argument.
+
+With PyTorch 1.6+ it'll automatically use `native AMP` when `--fp16` is set.
+
+To see all the possible command line options, run:
+
+```bash
+python finetune_trainer.py --help
+```
+
+For multi-gpu training use `torch.distributed.launch`, e.g. with 2 gpus:
+```bash
+python -m torch.distributed.launch --nproc_per_node=2  finetune_trainer.py ...
+```
+
+**At the moment, `Seq2SeqTrainer` does not support *with teacher* distillation.**
+
+All `Seq2SeqTrainer`-based fine-tuning scripts are included in the `builtin_trainer` directory.
+
+#### TPU Training
+`Seq2SeqTrainer` supports TPU training with few caveats
+1. As `generate` method does not work on TPU at the moment, `predict_with_generate` cannot be used. You should use `--prediction_loss_only` to only calculate loss, and do not set `--do_predict` and `--predict_with_generate`.
+2. All sequences should be padded to be of equal length to avoid extremely slow training. (`finetune_trainer.py` does this automatically when running on TPU.)
+
+We provide a very simple launcher script named `xla_spawn.py` that lets you run our example scripts on multiple TPU cores without any boilerplate. Just pass a `--num_cores` flag to this script, then your regular training script with its arguments (this is similar to the `torch.distributed.launch` helper for `torch.distributed`).
+
+`builtin_trainer/finetune_tpu.sh` script provides minimal arguments needed for TPU training.
+
+The following command fine-tunes `sshleifer/student_marian_en_ro_6_3` on TPU V3-8 and should complete one epoch in ~5-6 mins.
+
+```bash
+./builtin_trainer/train_distil_marian_enro_tpu.sh
+```
+
+## Evaluation Commands
+
+To create summaries for each article in dataset, we use `run_eval.py`, here are a few commands that run eval for different tasks and models.
+If 'translation' is in your task name, the computed metric will be BLEU. Otherwise, ROUGE will be used.
+
+For t5, you need to specify --task translation_{src}_to_{tgt} as follows:
+```bash
+export DATA_DIR=wmt_en_ro
+./run_eval.py t5-base \
+    $DATA_DIR/val.source t5_val_generations.txt \
+    --reference_path $DATA_DIR/val.target \
+    --score_path enro_bleu.json \
+    --task translation_en_to_ro \
+    --n_obs 100 \
+    --device cuda \
+    --fp16 \
+    --bs 32
+```
+
+This command works for MBART, although the BLEU score is suspiciously low.
+```bash
+export DATA_DIR=wmt_en_ro
+./run_eval.py facebook/mbart-large-en-ro $DATA_DIR/val.source mbart_val_generations.txt \
+    --reference_path $DATA_DIR/val.target \
+    --score_path enro_bleu.json \
+    --task translation \
+    --n_obs 100 \
+    --device cuda \
+    --fp16 \
+    --bs 32
+```
+
+Summarization (xsum will be very similar):
+```bash
+export DATA_DIR=cnn_dm
+./run_eval.py sshleifer/distilbart-cnn-12-6 $DATA_DIR/val.source dbart_val_generations.txt \
+    --reference_path $DATA_DIR/val.target \
+    --score_path cnn_rouge.json \
+    --task summarization \
+    --n_obs 100 \
+
+th 56 \
+    --fp16 \
+    --bs 32
+```
+
+### Multi-GPU Evaluation
+here is a command to run xsum evaluation on 8 GPUS. It is more than linearly faster than run_eval.py in some cases
+because it uses SortishSampler to minimize padding. You can also use it on 1 GPU. `data_dir` must have
+`{type_path}.source` and `{type_path}.target`. Run `./run_distributed_eval.py --help` for all clargs.
+
+```bash
+python -m torch.distributed.launch --nproc_per_node=8  run_distributed_eval.py \
+    --model_name sshleifer/distilbart-large-xsum-12-3  \
+    --save_dir xsum_generations \
+    --data_dir xsum \
+    --fp16  # you can pass generate kwargs like num_beams here, just like run_eval.py
+```
+
+Contributions that implement this command for other distributed hardware setups are welcome!
+
+#### Single-GPU Eval: Tips and Tricks
+
+When using `run_eval.py`, the following features can be useful:
+
+* if you running the script multiple times and want to make it easier to track what arguments produced that output, use `--dump-args`. Along with the results it will also dump any custom params that were passed to the script. For example if you used: `--num_beams 8 --early_stopping true`, the output will be:
+   ```
+   {'bleu': 26.887, 'n_obs': 10, 'runtime': 1, 'seconds_per_sample': 0.1, 'num_beams': 8, 'early_stopping': True}
+   ```
+
+   `--info` is an additional argument available for the same purpose of tracking the conditions of the experiment. It's useful to pass things that weren't in the argument list, e.g. a language pair `--info "lang:en-ru"`. But also if you pass `--info` without a value it will fallback to the current date/time string, e.g. `2020-09-13 18:44:43`.
+
+   If using `--dump-args --info`, the output will be:
+
+   ```
+   {'bleu': 26.887, 'n_obs': 10, 'runtime': 1, 'seconds_per_sample': 0.1, 'num_beams': 8, 'early_stopping': True, 'info': '2020-09-13 18:44:43'}
+   ```
+
+   If using `--dump-args --info "pair:en-ru chkpt=best`, the output will be:
+
+   ```
+   {'bleu': 26.887, 'n_obs': 10, 'runtime': 1, 'seconds_per_sample': 0.1, 'num_beams': 8, 'early_stopping': True, 'info': 'pair=en-ru chkpt=best'}
+   ```
+
+
+* if you need to perform a parametric search in order to find the best ones that lead to the highest BLEU score, let `run_eval_search.py` to do the searching for you.
+
+   The script accepts the exact same arguments as `run_eval.py`, plus an additional argument `--search`. The value of `--search` is parsed, reformatted and fed to ``run_eval.py`` as additional args.
+
+   The format for the `--search` value is a simple string with hparams and colon separated values to try, e.g.:
+   ```
+    --search "num_beams=5:10 length_penalty=0.8:1.0:1.2 early_stopping=true:false"
+   ```
+   which will generate `12` `(2*3*2)` searches for a product of each hparam. For example the example that was just used will invoke `run_eval.py` repeatedly with:
+
+   ```
+    --num_beams 5 --length_penalty 0.8 --early_stopping true
+    --num_beams 5 --length_penalty 0.8 --early_stopping false
+    [...]
+    --num_beams 10 --length_penalty 1.2 --early_stopping false
+   ```
+
+   On completion, this function prints a markdown table of the results sorted by the best BLEU score and the winning arguments.
+
+```
+bleu  | num_beams | length_penalty | early_stopping
+----- | --------- | -------------- | --------------
+26.71 |         5 |            1.1 |              1
+26.66 |         5 |            0.9 |              1
+26.66 |         5 |            0.9 |              0
+26.41 |         5 |            1.1 |              0
+21.94 |         1 |            0.9 |              1
+21.94 |         1 |            0.9 |              0
+21.94 |         1 |            1.1 |              1
+21.94 |         1 |            1.1 |              0
+
+Best score args:
+stas/wmt19-en-ru data/en-ru/val.source data/en-ru/test_translations.txt --reference_path data/en-ru/val.target --score_path data/en-ru/test_bleu.json --bs 8 --task translation --num_beams 5 --length_penalty 1.1 --early_stopping True
+```
+
+If you pass `--info "some experiment-specific info"` it will get printed before the results table - this is useful for scripting and multiple runs, so one can tell the different sets of results from each other.
+
+
+### Contributing
+- follow the standard contributing guidelines and code of conduct.
+- add tests to `test_seq2seq_examples.py`
+- To run only the seq2seq tests, you must be in the root of the repository and run:
+```bash
+pytest examples/seq2seq/
+```
+
+### Converting pytorch-lightning checkpoints
+pytorch lightning ``-do_predict`` often fails, after you are done training, the best way to evaluate your model is to convert it.
+
+This should be done for you, with a file called `{save_dir}/best_tfmr`.
+
+If that file doesn't exist but you have a lightning `.ckpt` file, you can run
+```bash
+python convert_pl_checkpoint_to_hf.py PATH_TO_CKPT  randomly_initialized_hf_model_path save_dir/best_tfmr
+```
+Then either `run_eval` or `run_distributed_eval` with `save_dir/best_tfmr` (see previous sections)
+
+
+# Experimental Features
+These features are harder to use and not always useful.
+
+###  Dynamic Batch Size for MT
+`finetune.py` has a command line arg `--max_tokens_per_batch` that allows batches to be dynamically sized.
+This feature can only be used:
+- with fairseq installed
+- on 1 GPU
+- without sortish sampler
+- after calling `./save_len_file.py $tok $data_dir`
+
+For example,
+```bash
+./save_len_file.py Helsinki-NLP/opus-mt-en-ro  wmt_en_ro
+./dynamic_bs_example.sh --max_tokens_per_batch=2000 --output_dir benchmark_dynamic_bs
+```
+splits `wmt_en_ro/train` into 11,197 uneven lengthed batches and can finish 1 epoch in 8 minutes on a v100.
+
+For comparison,
+```bash
+./dynamic_bs_example.sh --sortish_sampler --train_batch_size 48
+```
+uses 12,723 batches of length 48 and takes slightly more time 9.5 minutes.
+
+The feature is still experimental, because:
++ we can make it much more robust if we have memory mapped/preprocessed datasets.
++ The speedup over sortish sampler is not that large at the moment.

examples/legacy/seq2seq/finetune_trainer.py

@@ -20,14 +20,15 @@ from dataclasses import dataclass, field
 from typing import Optional
 
 import transformers
+from seq2seq_trainer import Seq2SeqTrainer
+from seq2seq_training_args import Seq2SeqTrainingArguments
 from transformers import (
     AutoConfig,
     AutoModelForSeq2SeqLM,
     AutoTokenizer,
     HfArgumentParser,
     MBartTokenizer,
-    Seq2SeqTrainer,
-    Seq2SeqTrainingArguments,
+    MBartTokenizerFast,
     set_seed,
 )
 from transformers.trainer_utils import EvaluationStrategy, is_main_process
@@ -174,11 +175,11 @@ def main():
         bool(training_args.parallel_mode == ParallelMode.DISTRIBUTED),
         training_args.fp16,
     )
+    transformers.utils.logging.enable_default_handler()
+    transformers.utils.logging.enable_explicit_format()
     # 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
@@ -220,11 +221,14 @@ def main():
         data_args.eval_beams = model.config.num_beams
 
     # set decoder_start_token_id for MBart
-    if model.config.decoder_start_token_id is None and isinstance(tokenizer, MBartTokenizer):
+    if model.config.decoder_start_token_id is None and isinstance(tokenizer, (MBartTokenizer, MBartTokenizerFast)):
         assert (
             data_args.tgt_lang is not None and data_args.src_lang is not None
         ), "mBart requires --tgt_lang and --src_lang"
-        model.config.decoder_start_token_id = tokenizer.lang_code_to_id[data_args.tgt_lang]
+        if isinstance(tokenizer, MBartTokenizer):
+            model.config.decoder_start_token_id = tokenizer.lang_code_to_id[data_args.tgt_lang]
+        else:
+            model.config.decoder_start_token_id = tokenizer.convert_tokens_to_ids(data_args.tgt_lang)
 
     if model_args.freeze_embeds:
         freeze_embeds(model)
@@ -282,9 +286,12 @@ def main():
     trainer = Seq2SeqTrainer(
         model=model,
         args=training_args,
+        data_args=data_args,
         train_dataset=train_dataset,
         eval_dataset=eval_dataset,
-        data_collator=Seq2SeqDataCollator(tokenizer, data_args, training_args.tpu_num_cores),
+        data_collator=Seq2SeqDataCollator(
+            tokenizer, data_args, model.config.decoder_start_token_id, training_args.tpu_num_cores
+        ),
         compute_metrics=compute_metrics_fn,
         tokenizer=tokenizer,
     )
@@ -317,9 +324,7 @@ def main():
     if training_args.do_eval:
         logger.info("*** Evaluate ***")
 
-        metrics = trainer.evaluate(
-            metric_key_prefix="val", max_length=data_args.val_max_target_length, num_beams=data_args.eval_beams
-        )
+        metrics = trainer.evaluate(metric_key_prefix="val")
         metrics["val_n_objs"] = data_args.n_val
         metrics["val_loss"] = round(metrics["val_loss"], 4)
 
@@ -331,12 +336,7 @@ def main():
     if training_args.do_predict:
         logger.info("*** Predict ***")
 
-        test_output = trainer.predict(
-            test_dataset=test_dataset,
-            metric_key_prefix="test",
-            max_length=data_args.val_max_target_length,
-            num_beams=data_args.eval_beams,
-        )
+        test_output = trainer.predict(test_dataset=test_dataset, metric_key_prefix="test")
         metrics = test_output.metrics
         metrics["test_n_objs"] = data_args.n_test
 

examples/legacy/seq2seq/old_test_datasets.py

@@ -24,7 +24,7 @@ from parameterized import parameterized
 from save_len_file import save_len_file
 from transformers import AutoTokenizer
 from transformers.models.mbart.modeling_mbart import shift_tokens_right
-from transformers.testing_utils import TestCasePlus, require_torch_non_multi_gpu_but_fix_me, slow
+from transformers.testing_utils import TestCasePlus, slow
 from utils import FAIRSEQ_AVAILABLE, DistributedSortishSampler, LegacySeq2SeqDataset, Seq2SeqDataset
 
 
@@ -61,7 +61,6 @@ class TestAll(TestCasePlus):
         ],
     )
     @slow
-    @require_torch_non_multi_gpu_but_fix_me
     def test_seq2seq_dataset_truncation(self, tok_name):
         tokenizer = AutoTokenizer.from_pretrained(tok_name)
         tmp_dir = make_test_data_dir(tmp_dir=self.get_auto_remove_tmp_dir())
@@ -101,7 +100,6 @@ class TestAll(TestCasePlus):
             break  # No need to test every batch
 
     @parameterized.expand([BART_TINY, BERT_BASE_CASED])
-    @require_torch_non_multi_gpu_but_fix_me
     def test_legacy_dataset_truncation(self, tok):
         tokenizer = AutoTokenizer.from_pretrained(tok)
         tmp_dir = make_test_data_dir(tmp_dir=self.get_auto_remove_tmp_dir())
@@ -126,7 +124,6 @@ class TestAll(TestCasePlus):
             assert max_len_target > trunc_target  # Truncated
             break  # No need to test every batch
 
-    @require_torch_non_multi_gpu_but_fix_me
     def test_pack_dataset(self):
         tokenizer = AutoTokenizer.from_pretrained("facebook/mbart-large-cc25")
 
@@ -145,7 +142,6 @@ class TestAll(TestCasePlus):
         assert orig_paths == new_paths
 
     @pytest.mark.skipif(not FAIRSEQ_AVAILABLE, reason="This test requires fairseq")
-    @require_torch_non_multi_gpu_but_fix_me
     def test_dynamic_batch_size(self):
         if not FAIRSEQ_AVAILABLE:
             return
@@ -170,7 +166,6 @@ class TestAll(TestCasePlus):
         if failures:
             raise AssertionError(f"too many tokens in {len(failures)} batches")
 
-    @require_torch_non_multi_gpu_but_fix_me
     def test_sortish_sampler_reduces_padding(self):
         ds, _, tokenizer = self._get_dataset(max_len=512)
         bs = 2
@@ -210,7 +205,6 @@ class TestAll(TestCasePlus):
         )
         return ds, max_tokens, tokenizer
 
-    @require_torch_non_multi_gpu_but_fix_me
     def test_distributed_sortish_sampler_splits_indices_between_procs(self):
         ds, max_tokens, tokenizer = self._get_dataset()
         ids1 = set(DistributedSortishSampler(ds, 256, num_replicas=2, rank=0, add_extra_examples=False))
@@ -226,7 +220,6 @@ class TestAll(TestCasePlus):
             PEGASUS_XSUM,
         ],
     )
-    @require_torch_non_multi_gpu_but_fix_me
     def test_dataset_kwargs(self, tok_name):
         tokenizer = AutoTokenizer.from_pretrained(tok_name, use_fast=False)
         if tok_name == MBART_TINY:

examples/legacy/seq2seq/old_test_tatoeba_conversion.py

@@ -18,7 +18,7 @@ import unittest
 
 from transformers.file_utils import cached_property
 from transformers.models.marian.convert_marian_tatoeba_to_pytorch import DEFAULT_REPO, TatoebaConverter
-from transformers.testing_utils import require_torch_non_multi_gpu_but_fix_me, slow
+from transformers.testing_utils import slow
 
 
 @unittest.skipUnless(os.path.exists(DEFAULT_REPO), "Tatoeba directory does not exist.")
@@ -29,12 +29,10 @@ class TatoebaConversionTester(unittest.TestCase):
         return TatoebaConverter(save_dir=tmp_dir)
 
     @slow
-    @require_torch_non_multi_gpu_but_fix_me
     def test_resolver(self):
         self.resolver.convert_models(["heb-eng"])
 
     @slow
-    @require_torch_non_multi_gpu_but_fix_me
     def test_model_card(self):
         content, mmeta = self.resolver.write_model_card("opus-mt-he-en", dry_run=True)
         assert mmeta["long_pair"] == "heb-eng"

examples/legacy/seq2seq/requirements.txt

@@ -0,0 +1,20 @@
+tensorboard
+scikit-learn
+seqeval
+psutil
+sacrebleu
+rouge-score
+tensorflow_datasets
+matplotlib
+git-python==1.0.3
+faiss-cpu
+streamlit
+elasticsearch
+nltk
+pandas
+datasets >= 1.1.3
+fire
+pytest
+conllu
+sentencepiece != 0.1.92
+protobuf

examples/legacy/seq2seq/run_eval.py

@@ -132,8 +132,14 @@ def run_generate(verbose=True):
     if args.n_obs > 0:
         examples = examples[: args.n_obs]
     Path(args.save_path).parent.mkdir(exist_ok=True)
+
     if args.reference_path is None and Path(args.score_path).exists():
         warnings.warn(f"score_path {args.score_path} will be overwritten unless you type ctrl-c.")
+
+    if args.device == "cpu" and args.fp16:
+        # this mix leads to RuntimeError: "threshold_cpu" not implemented for 'Half'
+        raise ValueError("Can't mix --fp16 and --device cpu")
+
     runtime_metrics = generate_summaries_or_translations(
         examples,
         args.save_path,

examples/legacy/seq2seq/seq2seq_trainer.py

@@ -60,7 +60,7 @@ class Seq2SeqTrainer(Trainer):
             assert isinstance(
                 self.model, PreTrainedModel
             ), f"If no `config` is passed the model to be trained has to be of type `PreTrainedModel`, but is {self.model.__class__}"
-            self.config = self._actual_model(self.model).config
+            self.config = self.model.config
         else:
             self.config = config