Release: v4.21.0

Add function to the submodule init

.circleci/config.yml

@@ -65,27 +65,30 @@ jobs:
     run_tests_torch_and_tf:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             RUN_PT_TF_CROSS_TESTS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   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 espeak-ng
+                      - v0.5-torch_and_tf-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
+            - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng git-lfs
+            - run: git lfs install
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,tf-cpu,torch,testing,sentencepiece,torch-speech,vision]
             - run: pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.11.0+cpu.html
             - run: pip install tensorflow_probability
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
+            - run: pip install git+https://github.com/huggingface/accelerate
             - save_cache:
-                key: v0.4-{{ checksum "setup.py" }}
+                key: v0.5-{{ checksum "setup.py" }}
                 paths:
                     - '~/.cache/pip'
             - run: python utils/tests_fetcher.py | tee test_preparation.txt
@@ -93,7 +96,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_torch_and_tf $(cat test_list.txt) -m is_pt_tf_cross_test --durations=0 | tee tests_output.txt
+                    python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_torch_and_tf $(cat test_list.txt) -m is_pt_tf_cross_test --durations=0 | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -103,31 +106,34 @@ jobs:
     run_tests_torch_and_tf_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             RUN_PT_TF_CROSS_TESTS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   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 espeak-ng
+                      - v0.5-torch_and_tf-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
+            - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng git-lfs
+            - run: git lfs install
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,tf-cpu,torch,testing,sentencepiece,torch-speech,vision]
             - run: pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.11.0+cpu.html
             - run: pip install tensorflow_probability
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
+            - run: pip install git+https://github.com/huggingface/accelerate
             - save_cache:
-                key: v0.4-{{ checksum "setup.py" }}
+                key: v0.5-{{ checksum "setup.py" }}
                 paths:
                     - '~/.cache/pip'
             - run: |
-                  python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_torch_and_tf tests -m is_pt_tf_cross_test --durations=0 | tee tests_output.txt
+                  python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_torch_and_tf tests -m is_pt_tf_cross_test --durations=0 | tee tests_output.txt
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
@@ -136,26 +142,28 @@ jobs:
     run_tests_torch_and_flax:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.6
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             RUN_PT_FLAX_CROSS_TESTS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-torch_and_flax-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-torch_and_flax-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,flax,torch,testing,sentencepiece,torch-speech,vision]
             - run: pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.11.0+cpu.html
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
+            - run: pip install git+https://github.com/huggingface/accelerate
             - save_cache:
-                key: v0.4-{{ checksum "setup.py" }}
+                key: v0.5-{{ checksum "setup.py" }}
                 paths:
                     - '~/.cache/pip'
             - run: python utils/tests_fetcher.py | tee test_preparation.txt
@@ -163,7 +171,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_torch_and_flax $(cat test_list.txt) -m is_pt_flax_cross_test --durations=0 | tee tests_output.txt
+                    python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_torch_and_flax $(cat test_list.txt) -m is_pt_flax_cross_test --durations=0 | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -173,30 +181,32 @@ jobs:
     run_tests_torch_and_flax_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.6
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             RUN_PT_FLAX_CROSS_TESTS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-torch_and_flax-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-torch_and_flax-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,flax,torch,testing,sentencepiece,torch-speech,vision]
             - run: pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.11.0+cpu.html
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
+            - run: pip install git+https://github.com/huggingface/accelerate
             - save_cache:
-                key: v0.4-{{ checksum "setup.py" }}
+                key: v0.5-{{ checksum "setup.py" }}
                 paths:
                     - '~/.cache/pip'
             - run: |
-                  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
+                  python -m pytest -n 8 --max-worker-restart=0 --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:
@@ -205,25 +215,27 @@ jobs:
     run_tests_torch:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   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 espeak-ng
+                      - v0.5-torch-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
+            - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng time
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,torch,testing,sentencepiece,torch-speech,vision,timm]
             - run: pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.11.0+cpu.html
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
+            - run: pip install git+https://github.com/huggingface/accelerate
             - save_cache:
-                  key: v0.4-torch-{{ checksum "setup.py" }}
+                  key: v0.5-torch-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/tests_fetcher.py | tee test_preparation.txt
@@ -231,7 +243,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 3 --dist=loadfile -s --make-reports=tests_torch $(cat test_list.txt) | tee tests_output.txt
+                    python -m pytest -n 3 --max-worker-restart=0 --dist=loadfile -s --make-reports=tests_torch $(cat test_list.txt) | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -241,29 +253,31 @@ jobs:
     run_tests_torch_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-torch-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-torch-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,torch,testing,sentencepiece,torch-speech,vision,timm]
             - run: pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.11.0+cpu.html
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
+            - run: pip install git+https://github.com/huggingface/accelerate
             - save_cache:
-                  key: v0.4-torch-{{ checksum "setup.py" }}
+                  key: v0.5-torch-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: |
-                  python -m pytest -n 3 --dist=loadfile -s --make-reports=tests_torch tests | tee tests_output.txt
+                  python -m pytest -n 3 --max-worker-restart=0 --dist=loadfile -s --make-reports=tests_torch tests | tee tests_output.txt
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
@@ -272,25 +286,26 @@ jobs:
     run_tests_tf:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-tf-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-tf-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,tf-cpu,testing,sentencepiece,tf-speech,vision]
             - run: pip install tensorflow_probability
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
             - save_cache:
-                  key: v0.4-tf-{{ checksum "setup.py" }}
+                  key: v0.5-tf-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/tests_fetcher.py | tee test_preparation.txt
@@ -298,7 +313,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_tf $(cat test_list.txt) | tee tests_output.txt
+                    python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_tf $(cat test_list.txt) | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -308,29 +323,30 @@ jobs:
     run_tests_tf_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-tf-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-tf-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,tf-cpu,testing,sentencepiece,tf-speech,vision]
             - run: pip install tensorflow_probability
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
             - save_cache:
-                  key: v0.4-tf-{{ checksum "setup.py" }}
+                  key: v0.5-tf-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: |
-                  python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_tf tests | tee tests_output.txt
+                  python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_tf tests | tee tests_output.txt
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
@@ -339,24 +355,25 @@ jobs:
     run_tests_flax:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                 keys:
-                    - v0.4-flax-{{ checksum "setup.py" }}
-                    - v0.4-{{ checksum "setup.py" }}
+                    - v0.5-flax-{{ checksum "setup.py" }}
+                    - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[flax,testing,sentencepiece,flax-speech,vision]
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
             - save_cache:
-                  key: v0.4-flax-{{ checksum "setup.py" }}
+                  key: v0.5-flax-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/tests_fetcher.py | tee test_preparation.txt
@@ -364,7 +381,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_flax $(cat test_list.txt) | tee tests_output.txt
+                    python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_flax $(cat test_list.txt) | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -374,28 +391,29 @@ jobs:
     run_tests_flax_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                 keys:
-                    - v0.4-flax-{{ checksum "setup.py" }}
-                    - v0.4-{{ checksum "setup.py" }}
+                    - v0.5-flax-{{ checksum "setup.py" }}
+                    - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[flax,testing,sentencepiece,vision,flax-speech]
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
             - save_cache:
-                  key: v0.4-flax-{{ checksum "setup.py" }}
+                  key: v0.5-flax-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: |
-                  python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_flax tests | tee tests_output.txt
+                  python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_flax tests | tee tests_output.txt
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
@@ -404,26 +422,27 @@ jobs:
     run_tests_pipelines_torch:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             RUN_PIPELINE_TESTS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-torch-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-torch-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,torch,testing,sentencepiece,torch-speech,vision,timm]
             - run: pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.11.0+cpu.html
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
             - save_cache:
-                  key: v0.4-torch-{{ checksum "setup.py" }}
+                  key: v0.5-torch-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/tests_fetcher.py | tee test_preparation.txt
@@ -431,7 +450,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_pipelines_torch -m is_pipeline_test $(cat test_list.txt) | tee tests_output.txt
+                    python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_pipelines_torch -m is_pipeline_test $(cat test_list.txt) | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -441,30 +460,31 @@ jobs:
     run_tests_pipelines_torch_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             RUN_PIPELINE_TESTS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-torch-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-torch-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,torch,testing,sentencepiece,torch-speech,vision,timm]
             - run: pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.11.0+cpu.html
             - run: pip install https://github.com/kpu/kenlm/archive/master.zip
             - save_cache:
-                  key: v0.4-torch-{{ checksum "setup.py" }}
+                  key: v0.5-torch-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: |
-                  python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_pipelines_torch -m is_pipeline_test tests | tee tests_output.txt
+                  python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_pipelines_torch -m is_pipeline_test tests | tee tests_output.txt
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
@@ -473,24 +493,25 @@ jobs:
     run_tests_pipelines_tf:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             RUN_PIPELINE_TESTS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-tf-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-tf-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,tf-cpu,testing,sentencepiece]
             - run: pip install tensorflow_probability
             - save_cache:
-                  key: v0.4-tf-{{ checksum "setup.py" }}
+                  key: v0.5-tf-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/tests_fetcher.py | tee test_preparation.txt
@@ -498,7 +519,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_pipelines_tf $(cat test_list.txt) -m is_pipeline_test | tee tests_output.txt
+                    python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_pipelines_tf $(cat test_list.txt) -m is_pipeline_test | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -508,28 +529,29 @@ jobs:
     run_tests_pipelines_tf_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             RUN_PIPELINE_TESTS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-tf-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-tf-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,tf-cpu,testing,sentencepiece]
             - run: pip install tensorflow_probability
             - save_cache:
-                  key: v0.4-tf-{{ checksum "setup.py" }}
+                  key: v0.5-tf-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: |
-                  python -m pytest -n 8 --dist=loadfile -rA -s --make-reports=tests_pipelines_tf tests -m is_pipeline_test | tee tests_output.txt
+                  python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -rA -s --make-reports=tests_pipelines_tf tests -m is_pipeline_test | tee tests_output.txt
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
@@ -538,30 +560,31 @@ jobs:
     run_tests_custom_tokenizers:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             RUN_CUSTOM_TOKENIZERS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-custom_tokenizers-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-custom_tokenizers-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[ja,testing,sentencepiece,jieba,spacy,ftfy]
+            - run: pip install .[ja,testing,sentencepiece,jieba,spacy,ftfy,rjieba]
             - run: python -m unidic download
             - save_cache:
-                  key: v0.4-custom_tokenizers-{{ checksum "setup.py" }}
+                  key: v0.5-custom_tokenizers-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -s --make-reports=tests_custom_tokenizers ./tests/test_tokenization_bert_japanese.py ./tests/test_tokenization_openai.py | tee tests_output.txt
+                    python -m pytest --max-worker-restart=0 -s --make-reports=tests_custom_tokenizers ./tests/test_tokenization_bert_japanese.py ./tests/test_tokenization_openai.py | tee tests_output.txt
                   fi
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 1 tests/test_tokenization_clip.py --dist=loadfile -s --make-reports=tests_tokenization_clip --durations=100 | tee tests_output.txt
+                    python -m pytest -n 1 --max-worker-restart=0 tests/test_tokenization_clip.py --dist=loadfile -s --make-reports=tests_tokenization_clip --durations=100 | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -571,25 +594,25 @@ jobs:
     run_examples_torch:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.6
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-torch_examples-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-torch_examples-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,torch,sentencepiece,testing,torch-speech]
             - run: pip install -r examples/pytorch/_tests_requirements.txt
-            - run: pip install git+https://github.com/huggingface/accelerate
             - save_cache:
-                  key: v0.4-torch_examples-{{ checksum "setup.py" }}
+                  key: v0.5-torch_examples-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/tests_fetcher.py --filters examples tests | tee test_preparation.txt
@@ -597,7 +620,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 8 --dist=loadfile -s --make-reports=examples_torch ./examples/pytorch/ | tee tests_output.txt
+                    python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -s --make-reports=examples_torch ./examples/pytorch/ | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/examples_output.txt
@@ -607,28 +630,29 @@ jobs:
     run_examples_torch_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.6
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-torch_examples-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-torch_examples-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng
             - run: pip install --upgrade pip
             - run: pip install .[sklearn,torch,sentencepiece,testing,torch-speech]
             - run: pip install -r examples/pytorch/_tests_requirements.txt
             - save_cache:
-                  key: v0.4-torch_examples-{{ checksum "setup.py" }}
+                  key: v0.5-torch_examples-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: |
-                  TRANSFORMERS_IS_CI=1 python -m pytest -n 8 --dist=loadfile -s --make-reports=examples_torch ./examples/pytorch/ | tee examples_output.txt
+                  TRANSFORMERS_IS_CI=1 python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -s --make-reports=examples_torch ./examples/pytorch/ | tee examples_output.txt
             - store_artifacts:
                   path: ~/transformers/examples_output.txt
             - store_artifacts:
@@ -637,23 +661,24 @@ jobs:
     run_examples_flax:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                 keys:
-                    - v0.4-flax_examples-{{ checksum "setup.py" }}
-                    - v0.4-{{ checksum "setup.py" }}
+                    - v0.5-flax_examples-{{ checksum "setup.py" }}
+                    - v0.5-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: sudo pip install .[flax,testing,sentencepiece]
+            - run: pip install .[flax,testing,sentencepiece]
             - run: pip install -r examples/flax/_tests_requirements.txt
             - save_cache:
-                  key: v0.4-flax_examples-{{ checksum "setup.py" }}
+                  key: v0.5-flax_examples-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/tests_fetcher.py --filters examples tests | tee test_preparation.txt
@@ -661,7 +686,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 8 --dist=loadfile -s --make-reports=examples_flax ./examples/flax/ | tee tests_output.txt
+                    python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -s --make-reports=examples_flax ./examples/flax/ | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/flax_examples_output.txt
@@ -671,27 +696,28 @@ jobs:
     run_examples_flax_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                 keys:
-                    - v0.4-flax_examples-{{ checksum "setup.py" }}
-                    - v0.4-{{ checksum "setup.py" }}
+                    - v0.5-flax_examples-{{ checksum "setup.py" }}
+                    - v0.5-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: sudo pip install .[flax,testing,sentencepiece]
+            - run: pip install .[flax,testing,sentencepiece]
             - run: pip install -r examples/flax/_tests_requirements.txt
             - save_cache:
-                  key: v0.4-flax_examples-{{ checksum "setup.py" }}
+                  key: v0.5-flax_examples-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: |
-                  TRANSFORMERS_IS_CI=1 python -m pytest -n 8 --dist=loadfile -s --make-reports=examples_flax ./examples/flax/ | tee examples_output.txt
+                  TRANSFORMERS_IS_CI=1 python -m pytest -n 8 --max-worker-restart=0 --dist=loadfile -s --make-reports=examples_flax ./examples/flax/ | tee examples_output.txt
             - store_artifacts:
                   path: ~/transformers/flax_examples_output.txt
             - store_artifacts:
@@ -700,27 +726,28 @@ jobs:
     run_tests_hub:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             HUGGINGFACE_CO_STAGING: yes
             RUN_GIT_LFS_TESTS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-hub-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
-            - run: sudo apt-get install git-lfs
+                      - v0.5-hub-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
+            - run: sudo apt-get -y update && sudo apt-get install git-lfs
             - run: |
                 git config --global user.email "ci@dummy.com"
                 git config --global user.name "ci"
             - run: pip install --upgrade pip
             - run: pip install .[torch,sentencepiece,testing]
             - save_cache:
-                  key: v0.4-hub-{{ checksum "setup.py" }}
+                  key: v0.5-hub-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/tests_fetcher.py | tee test_preparation.txt
@@ -728,7 +755,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -sv --make-reports=tests_hub $(cat test_list.txt) -m is_staging_test | tee tests_output.txt
+                    python -m pytest --max-worker-restart=0 -sv --make-reports=tests_hub $(cat test_list.txt) -m is_staging_test | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -738,31 +765,32 @@ jobs:
     run_tests_hub_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             HUGGINGFACE_CO_STAGING: yes
             RUN_GIT_LFS_TESTS: yes
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-hub-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
-            - run: sudo apt-get install git-lfs
+                      - v0.5-hub-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
+            - run: sudo apt-get -y update && sudo apt-get install git-lfs
             - run: |
                 git config --global user.email "ci@dummy.com"
                 git config --global user.name "ci"
             - run: pip install --upgrade pip
             - run: pip install .[torch,sentencepiece,testing]
             - save_cache:
-                  key: v0.4-hub-{{ checksum "setup.py" }}
+                  key: v0.5-hub-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: |
-                  python -m pytest -sv --make-reports=tests_hub tests -m is_staging_test | tee tests_output.txt
+                  python -m pytest --max-worker-restart=0 -sv --make-reports=tests_hub tests -m is_staging_test | tee tests_output.txt
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
@@ -771,22 +799,23 @@ jobs:
     run_tests_onnxruntime:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-torch-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-torch-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[torch,testing,sentencepiece,onnxruntime,vision]
+            - run: pip install .[torch,testing,sentencepiece,onnxruntime,vision,rjieba]
             - save_cache:
-                  key: v0.4-onnx-{{ checksum "setup.py" }}
+                  key: v0.5-onnx-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/tests_fetcher.py | tee test_preparation.txt
@@ -794,7 +823,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_onnx $(cat test_list.txt) -k onnx | tee tests_output.txt
+                    python -m pytest -n 1 --max-worker-restart=0 --dist=loadfile -s --make-reports=tests_onnx $(cat test_list.txt) -k onnx | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -804,26 +833,27 @@ jobs:
     run_tests_onnxruntime_all:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-torch-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-torch-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
             - run: pip install .[torch,testing,sentencepiece,onnxruntime,vision]
             - save_cache:
-                  key: v0.4-onnx-{{ checksum "setup.py" }}
+                  key: v0.5-onnx-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: |
-                  python -m pytest -n 1 --dist=loadfile -s --make-reports=tests_onnx tests -k onnx | tee tests_output.txt
+                  python -m pytest -n 1 --max-worker-restart=0 --dist=loadfile -s --make-reports=tests_onnx tests -k onnx | tee tests_output.txt
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
@@ -832,47 +862,51 @@ jobs:
     check_code_quality:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.6
+            - image: cimg/python:3.7.12
         resource_class: large
         environment:
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-code_quality-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-code_quality-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
             - run: pip install .[all,quality]
             - save_cache:
-                  key: v0.4-code_quality-{{ checksum "setup.py" }}
+                  key: v0.5-code_quality-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
-            - run: black --check examples tests src utils
+            - run: black --check --preview examples tests src utils
             - run: isort --check-only examples tests src utils
             - run: python utils/custom_init_isort.py --check_only
+            - run: python utils/sort_auto_mappings.py --check_only
             - run: flake8 examples tests src utils
             - run: doc-builder style src/transformers docs/source --max_len 119 --check_only --path_to_docs docs/source
+            - run: python utils/check_doc_toc.py
 
     check_repository_consistency:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.6
+            - image: cimg/python:3.7.12
         resource_class: large
         environment:
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-repository_consistency-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-repository_consistency-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
             - run: pip install .[all,quality]
             - save_cache:
-                  key: v0.4-repository_consistency-{{ checksum "setup.py" }}
+                  key: v0.5-repository_consistency-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/check_copies.py
@@ -880,24 +914,26 @@ jobs:
             - run: python utils/check_dummies.py
             - run: python utils/check_repo.py
             - run: python utils/check_inits.py
+            - run: python utils/check_config_docstrings.py
             - run: make deps_table_check_updated
             - run: python utils/tests_fetcher.py --sanity_check
 
-    run_tests_layoutlmv2:
+    run_tests_layoutlmv2_and_v3:
         working_directory: ~/transformers
         docker:
-            - image: circleci/python:3.7
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
+            PYTEST_TIMEOUT: 120
         resource_class: xlarge
         parallelism: 1
         steps:
             - checkout
             - restore_cache:
                   keys:
-                      - v0.4-torch-{{ checksum "setup.py" }}
-                      - v0.4-{{ checksum "setup.py" }}
+                      - v0.5-torch-{{ checksum "setup.py" }}
+                      - v0.5-{{ checksum "setup.py" }}
             - run: sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev
             - run: pip install --upgrade pip
             - run: pip install .[torch,testing,vision]
@@ -906,7 +942,7 @@ jobs:
             - run: sudo apt install tesseract-ocr
             - run: pip install pytesseract
             - save_cache:
-                  key: v0.4-torch-{{ checksum "setup.py" }}
+                  key: v0.5-torch-{{ checksum "setup.py" }}
                   paths:
                       - '~/.cache/pip'
             - run: python utils/tests_fetcher.py | tee test_preparation.txt
@@ -914,7 +950,7 @@ jobs:
                   path: ~/transformers/test_preparation.txt
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -n 1 tests/*layoutlmv2* --dist=loadfile -s --make-reports=tests_layoutlmv2 --durations=100
+                    python -m pytest -n 1 --max-worker-restart=0 tests/models/*layoutlmv* --dist=loadfile -s --make-reports=tests_layoutlmv2_and_v3 --durations=100
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -924,7 +960,7 @@ jobs:
 # TPU JOBS
     run_examples_tpu:
         docker:
-            - image: circleci/python:3.6
+            - image: cimg/python:3.7.12
         environment:
             OMP_NUM_THREADS: 1
             TRANSFORMERS_IS_CI: yes
@@ -944,7 +980,7 @@ jobs:
 
     cleanup-gke-jobs:
         docker:
-            - image: circleci/python:3.6
+            - image: cimg/python:3.7.12
         steps:
             - gcp-gke/install
             - gcp-gke/update-kubeconfig-with-credentials:
@@ -975,7 +1011,7 @@ workflows:
             - run_tests_pipelines_tf
             - run_tests_onnxruntime
             - run_tests_hub
-            - run_tests_layoutlmv2
+            - run_tests_layoutlmv2_and_v3
     nightly:
         triggers:
             - schedule:

.gitattributes

@@ -1,3 +1,4 @@
 *.py	eol=lf
 *.rst	eol=lf
-*.md	eol=lf
+*.md	eol=lf
+*.mdx   eol=lf

.github/ISSUE_TEMPLATE/---new-benchmark.md

@@ -1,22 +0,0 @@
----
-name: "\U0001F5A5 New benchmark"
-about: Benchmark a part of this library and share your results
-title: "[Benchmark]"
-labels: ''
-assignees: ''
-
----
-
-# 🖥 Benchmarking `transformers`
-
-## Benchmark
-
-Which part of `transformers` did you benchmark?
-
-## Set-up
-
-What did you run your benchmarks on? Please include details, such as: CPU, GPU? If using multiple GPUs, which parallelization did you use?
-
-## Results
-
-Put your results here!

.github/ISSUE_TEMPLATE/--new-model-addition.md

@@ -1,20 +0,0 @@
----
-name: "\U0001F31F New model addition"
-about: Submit a proposal/request to implement a new Transformer-based model
-title: ''
-labels: New model
-assignees: ''
-
----
-
-# 🌟 New model addition
-
-## Model description
-
-<!-- Important information -->
-
-## Open source status
-
-* [ ] the model implementation is available: (give details)
-* [ ] the model weights are available: (give details)
-* [ ] who are the authors: (mention them, if possible by @gh-username)

.github/ISSUE_TEMPLATE/bug-report.md

@@ -1,107 +0,0 @@
----
-name: "\U0001F41B Bug Report"
-about: Submit a bug report to help us improve transformers
-title: ''
-labels: ''
-assignees: ''
-
----
-
-
-## Environment info
-<!-- You can run the command `transformers-cli env` and copy-and-paste its output below.
-     Don't forget to fill out the missing fields in that output! -->
-
-- `transformers` version:
-- Platform:
-- Python version:
-- PyTorch version (GPU?):
-- Tensorflow version (GPU?):
-- Using GPU in script?:
-- Using distributed or parallel set-up in script?:
-
-### Who can help
-<!-- Your issue will be replied to more quickly if you can figure out the right person to tag with @
- 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.
-
-Models:
-
-- ALBERT, BERT, XLM, DeBERTa, DeBERTa-v2, ELECTRA, MobileBert, SqueezeBert: @LysandreJik
-- T5, Pegasus, EncoderDecoder: @patrickvonplaten
-- Blenderbot, MBART, BART, Marian, Pegasus: @patil-suraj
-- Reformer, TransfoXL, XLNet, FNet: @patrickvonplaten
-- Longformer, BigBird: @ydshieh
-- FSMT: @stas00
-- Funnel: @sgugger
-- GPT-2, GPT: @patil-suraj, @patrickvonplaten, @LysandreJik
-- RAG, DPR: @patrickvonplaten, @lhoestq
-- TensorFlow: @Rocketknight1
-- JAX/Flax: @patil-suraj
-- TAPAS, LayoutLM, LayoutLMv2, LUKE, ViT, BEiT, DEiT, DETR, CANINE: @NielsRogge
-- GPT-Neo, GPT-J, CLIP: @patil-suraj
-- Wav2Vec2, HuBERT, SpeechEncoderDecoder, UniSpeech, UniSpeechSAT, SEW, SEW-D, Speech2Text: @patrickvonplaten, @anton-l
-
-If the model isn't in the list, ping @LysandreJik who will redirect you to the correct contributor.
-
-Library:
-
-- Benchmarks: @patrickvonplaten
-- Deepspeed: @stas00
-- Ray/raytune: @richardliaw, @amogkam
-- Text generation: @patrickvonplaten @narsil
-- Tokenizers: @SaulLu
-- Trainer: @sgugger
-- Pipelines: @Narsil
-- Speech: @patrickvonplaten, @anton-l
-- Vision: @NielsRogge, @sgugger
-
-Documentation: @sgugger
-
-Model hub:
-
-- for issues with a model, report at https://discuss.huggingface.co/ and tag the model's creator.
-
-HF projects:
-
-- datasets: [different repo](https://github.com/huggingface/datasets)
-- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
-
-Examples:
-
-- maintained examples (not research project or legacy): @sgugger, @patil-suraj
-
-For research projetcs, please ping the contributor directly. For example, on the following projects:
-
-- research_projects/bert-loses-patience: @JetRunner
-- research_projects/distillation: @VictorSanh
-
- -->
-
-## Information
-
-Model I am using (Bert, XLNet ...):
-
-The problem arises when using:
-* [ ] the official example scripts: (give details below)
-* [ ] my own modified scripts: (give details below)
-
-The tasks I am working on is:
-* [ ] an official GLUE/SQUaD task: (give the name)
-* [ ] my own task or dataset: (give details below)
-
-## To reproduce
-
-Steps to reproduce the behavior:
-
-1.
-2.
-3.
-
-<!-- If you have code snippets, error messages, stack traces please provide them here as well.
-     Important! Use code tags to correctly format your code. See https://help.github.com/en/github/writing-on-github/creating-and-highlighting-code-blocks#syntax-highlighting
-     Do not use screenshots, as they are hard to read and (more importantly) don't allow others to copy-and-paste your code.-->
-
-## Expected behavior
-
-<!-- A clear and concise description of what you would expect to happen. -->

.github/ISSUE_TEMPLATE/bug-report.yml

@@ -0,0 +1,119 @@
+name: "\U0001F41B Bug Report"
+description: Submit a bug report to help us improve transformers
+labels: [ "bug" ]
+body:
+  - type: textarea
+    id: system-info
+    attributes:
+      label: System Info
+      description: Please share your system info with us. You can run the command `transformers-cli env` and copy-paste its output below.
+      placeholder: transformers version, platform, python version, ...
+    validations:
+      required: true
+
+  - type: textarea
+    id: who-can-help
+    attributes:
+      label: Who can help?
+      description: |
+        Your issue will be replied to more quickly if you can figure out the right person to tag with @
+        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.
+        
+        Models:
+
+          - ALBERT, BERT, XLM, DeBERTa, DeBERTa-v2, ELECTRA, MobileBert, SqueezeBert: `@LysandreJik`
+          - T5, Pegasus, EncoderDecoder: `@patrickvonplaten`
+          - Blenderbot, MBART, BART, Marian, Pegasus: `@patil-suraj`
+          - Reformer, TransfoXL, XLNet, FNet: `@patrickvonplaten`
+          - Longformer, BigBird: `@ydshieh`
+          - FSMT: `@stas00`
+          - Funnel: `@sgugger`
+          - GPT-2, GPT: `@patil-suraj`, `@patrickvonplaten`, `@LysandreJik`
+          - RAG, DPR: `@patrickvonplaten`, `@lhoestq`
+          - TensorFlow: `@Rocketknight1`
+          - JAX/Flax: `@patil-suraj`
+          - TAPAS, LayoutLM, LayoutLMv2, LUKE, ViT, BEiT, DEiT, DETR, CANINE: `@NielsRogge`
+          - GPT-Neo, GPT-J, CLIP: `@patil-suraj`
+          - Wav2Vec2, HuBERT, UniSpeech, UniSpeechSAT, SEW, SEW-D: `@patrickvonplaten`, `@anton-l`
+          - SpeechEncoderDecoder, Speech2Text, Speech2Text2: `@sanchit-gandhi`, `@patrickvonplaten`, `@anton-l`
+          
+          If the model isn't in the list, ping `@LysandreJik` who will redirect you to the correct contributor.
+
+        Library:
+          - Benchmarks: `@patrickvonplaten`
+          - Deepspeed: `@stas00`
+          - Ray/raytune: `@richardliaw`, `@amogkam`
+          - Text generation: `@patrickvonplaten`, `@Narsil`, `@gante`
+          - Tokenizers: `@SaulLu`
+          - Trainer: `@sgugger`
+          - Pipelines: `@Narsil`
+          - Speech: `@patrickvonplaten`, `@anton-l`, `@sanchit-gandhi`
+          - Vision: `@NielsRogge`, `@sgugger`
+
+        Documentation: `@sgugger`, `@stevhliu`
+
+        Model hub:
+
+          - for issues with a model, report at https://discuss.huggingface.co/ and tag the model's creator.
+
+        HF projects:
+
+          - datasets: [different repo](https://github.com/huggingface/datasets)
+          - rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
+
+        Examples:
+
+          - maintained examples (not research project or legacy): `@sgugger`, `@patil-suraj`
+
+        For research projetcs, please ping the contributor directly. For example, on the following projects:
+
+          - research_projects/bert-loses-patience: `@JetRunner`
+          - research_projects/distillation: `@VictorSanh`
+      placeholder: "@Username ..."
+
+  - type: checkboxes
+    id: information-scripts-examples
+    attributes:
+      label: Information
+      description: 'The problem arises when using:'
+      options:
+        - label: "The official example scripts"
+        - label: "My own modified scripts"
+
+  - type: checkboxes
+    id: information-tasks
+    attributes:
+      label: Tasks
+      description: "The tasks I am working on are:"
+      options:
+        - label: "An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)"
+        - label: "My own task or dataset (give details below)"
+
+  - type: textarea
+    id: reproduction
+    validations:
+      required: true
+    attributes:
+      label: Reproduction
+      description: |
+        Please provide a code sample that reproduces the problem you ran into. It can be a Colab link or just a code snippet.
+        If you have code snippets, error messages, stack traces please provide them here as well.
+        Important! Use code tags to correctly format your code. See https://help.github.com/en/github/writing-on-github/creating-and-highlighting-code-blocks#syntax-highlighting
+        Do not use screenshots, as they are hard to read and (more importantly) don't allow others to copy-and-paste your code.
+
+      placeholder: |
+        Steps to reproduce the behavior:
+          
+          1.
+          2.
+          3.
+          
+
+  - type: textarea
+    id: expected-behavior
+    validations:
+      required: true
+    attributes:
+      label: Expected behavior
+      description: "A clear and concise description of what you would expect to happen."

.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,12 @@
+blank_issues_enabled: true
+version: 2.1
+contact_links:
+  - name: Model checkpoints on the Hugging Face Hub
+    url: https://huggingface.co/models
+    about: Open a Pull request / Discussion related to a specific model checkpoint directly on the Hugging Face Hub
+  - name: Website Related
+    url: https://github.com/huggingface/hub-docs/issues
+    about: Feature requests and bug reports related to the website
+  - name: Forum
+    url: https://discuss.huggingface.co/
+    about: General usage questions and community discussions

.github/ISSUE_TEMPLATE/feature-request.md

@@ -1,25 +0,0 @@
----
-name: "\U0001F680 Feature request"
-about: Submit a proposal/request for a new transformers feature
-title: ''
-labels: ''
-assignees: ''
-
----
-
-# 🚀 Feature request
-
-<!-- A clear and concise description of the feature proposal.
-     Please provide a link to the paper and code in case they exist. -->
-
-## Motivation
-
-<!-- Please outline the motivation for the proposal. Is your feature request
-     related to a problem? e.g., I'm always frustrated when [...]. If this is related
-     to another GitHub issue, please link here too. -->
-
-## Your contribution
-
-<!-- Is there any way that you could help, e.g. by submitting a PR?
-     Make sure to read the CONTRIBUTING.MD readme:
-     https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md -->

.github/ISSUE_TEMPLATE/feature-request.yml

@@ -0,0 +1,31 @@
+name: "\U0001F680 Feature request"
+description: Submit a proposal/request for a new transformers feature
+labels: [ "feature" ]
+body:
+  - type: textarea
+    id: feature-request
+    validations:
+      required: true
+    attributes:
+      label: Feature request
+      description: |
+        A clear and concise description of the feature proposal. Please provide a link to the paper and code in case they exist.
+
+  - type: textarea
+    id: motivation
+    validations:
+      required: true
+    attributes:
+      label: Motivation
+      description: |
+        Please outline the motivation for the proposal. Is your feature request related to a problem? e.g., I'm always frustrated when [...]. If this is related to another GitHub issue, please link here too.
+        
+
+  - type: textarea
+    id: contribution
+    validations:
+      required: true
+    attributes:
+      label: Your contribution
+      description: |
+        Is there any way that you could help, e.g. by submitting a PR? Make sure to read the CONTRIBUTING.MD [readme](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md)

.github/ISSUE_TEMPLATE/migration.md

@@ -1,58 +0,0 @@
----
-name: "\U0001F4DA Migration from pytorch-pretrained-bert or pytorch-transformers"
-about: Report a problem when migrating from pytorch-pretrained-bert or pytorch-transformers
-  to transformers
-title: ''
-labels: Migration
-assignees: ''
-
----
-
-# 📚 Migration
-
-## Information
-
-<!-- Important information -->
-
-Model I am using (Bert, XLNet ...):
-
-Language I am using the model on (English, Chinese ...):
-
-The problem arises when using:
-* [ ] the official example scripts: (give details below)
-* [ ] my own modified scripts: (give details below)
-
-The tasks I am working on is:
-* [ ] an official GLUE/SQUaD task: (give the name)
-* [ ] my own task or dataset: (give details below)
-
-## Details
-
-<!-- A clear and concise description of the migration issue.
-    If you have code snippets, please provide it here as well.
-    Important! Use code tags to correctly format your code. See https://help.github.com/en/github/writing-on-github/creating-and-highlighting-code-blocks#syntax-highlighting
-    Do not use screenshots, as they are hard to read and (more importantly) don't allow others to copy-and-paste your code.
-    -->
-
-## Environment info
-<!-- You can run the command `python transformers-cli env` and copy-and-paste its output below.
-     Don't forget to fill out the missing fields in that output! -->
- 
-- `transformers` version:
-- Platform:
-- Python version:
-- PyTorch version (GPU?):
-- Tensorflow version (GPU?):
-- Using GPU in script?:
-- Using distributed or parallel set-up in script?:
-
-<!-- IMPORTANT: which version of the former library do you use? -->
-* `pytorch-transformers` or `pytorch-pretrained-bert` version (or branch):
-
-
-## Checklist
-
-- [ ] I have read the migration guide in the readme.
- ([pytorch-transformers](https://github.com/huggingface/transformers#migrating-from-pytorch-transformers-to-transformers);
-  [pytorch-pretrained-bert](https://github.com/huggingface/transformers#migrating-from-pytorch-pretrained-bert-to-transformers))
-- [ ] I checked if a related official extension example runs on my machine.

.github/ISSUE_TEMPLATE/migration.yml

@@ -0,0 +1,72 @@
+name: "\U0001F4DA Migration from pytorch-pretrained-bert or pytorch-transformers"
+description: Report a problem when migrating from pytorch-pretrained-bert or pytorch-transformers to transformers
+labels: [ "migration" ]
+body:
+  - type: textarea
+    id: system-info
+    attributes:
+      label: System Info
+      description: Please share your system info with us. You can run the command `transformers-cli env` and copy-paste its output below.
+      render: shell
+      placeholder: transformers version, platform, python version, ...
+    validations:
+      required: true
+
+  - type: checkboxes
+    id: information-scripts-examples
+    attributes:
+      label: Information
+      description: 'The problem arises when using:'
+      options:
+        - label: "The official example scripts"
+        - label: "My own modified scripts"
+
+  - type: checkboxes
+    id: information-tasks
+    attributes:
+      label: Tasks
+      description: "The tasks I am working on are:"
+      options:
+        - label: "An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)"
+        - label: "My own task or dataset (give details below)"
+
+  - type: textarea
+    id: reproduction
+    validations:
+      required: true
+    attributes:
+      label: Reproduction
+      description: |
+        Please provide a code sample that reproduces the problem you ran into. It can be a Colab link or just a code snippet.
+        If you have code snippets, error messages, stack traces please provide them here as well.
+        Important! Use code tags to correctly format your code. See https://help.github.com/en/github/writing-on-github/creating-and-highlighting-code-blocks#syntax-highlighting
+        Do not use screenshots, as they are hard to read and (more importantly) don't allow others to copy-and-paste your code.
+
+      placeholder: |
+        Steps to reproduce the behavior:
+          
+          1.
+          2.
+          3.
+          
+
+  - type: textarea
+    id: expected-behavior
+    validations:
+      required: true
+    attributes:
+      label: Expected behavior
+      description: "A clear and concise description of what you would expect to happen."
+      render: shell
+
+  - type: checkboxes
+    id: checklist
+    attributes:
+      label: Checklist
+      options:
+        - label: "I have read the migration guide in the readme.
+ ([pytorch-transformers](https://github.com/huggingface/transformers#migrating-from-pytorch-transformers-to-transformers);
+  [pytorch-pretrained-bert](https://github.com/huggingface/transformers#migrating-from-pytorch-pretrained-bert-to-transformers))"
+          required: true
+        - label: "I checked if a related official extension example runs on my machine."
+          required: true

.github/ISSUE_TEMPLATE/new-model-addition.yml

@@ -0,0 +1,31 @@
+name: "\U0001F31F New model addition"
+description: Submit a proposal/request to implement a new model
+labels: [ "New model" ]
+
+body:
+  - type: textarea
+    id: description-request
+    validations:
+      required: true
+    attributes:
+      label: Model description
+      description: |
+        Put any and all important information relative to the model
+
+  - type: checkboxes
+    id: information-tasks
+    attributes:
+      label: Open source status
+      description: |
+          Please note that if the model implementation isn't available or if the weights aren't open-source, we are less likely to implement it in `transformers`.
+      options:
+        - label: "The model implementation is available"
+        - label: "The model weights are available"
+
+  - type: textarea
+    id: additional-info
+    attributes:
+      label: Provide useful links for the implementation
+      description: |
+        Please provide information regarding the implementation, the weights, and the authors.
+        Please mention the authors by @gh-username if you're aware of their usernames.

.github/ISSUE_TEMPLATE/question-help.md

@@ -1,26 +0,0 @@
----
-name: "❓ Questions & Help"
-about: Post your general questions on the Hugging Face forum: https://discuss.huggingface.co/
-title: ''
-labels: ''
-assignees: ''
-
----
-
-# ❓ Questions & Help
-
-<!-- The GitHub issue tracker is primarly intended for bugs, feature requests,
-     new models, benchmarks, and migration questions. For all other questions,
-     we direct you to the Hugging Face forum: https://discuss.huggingface.co/ .
-     -->
-
-## Details
-
-<!-- Description of your issue -->
-
-<!-- You should first ask your question on the forum, and only if
-     you didn't get an answer after a few days ask it here on GitHub. -->
-
-**A link to original question on the forum**:
-
-<!-- Your issue will be closed if you don't fill this part. -->

.github/conda/meta.yaml

@@ -25,7 +25,7 @@ requirements:
     - sacremoses
     - regex !=2019.12.17
     - protobuf
-    - tokenizers >=0.10.1,<0.11.0
+    - tokenizers >=0.11.1,!=0.11.3,<0.13
     - pyyaml >=5.1
   run:
     - python
@@ -40,7 +40,7 @@ requirements:
     - sacremoses
     - regex !=2019.12.17
     - protobuf
-    - tokenizers >=0.10.1,<0.11.0
+    - tokenizers >=0.11.1,!=0.11.3,<0.13
     - pyyaml >=5.1
 
 test:

.github/workflows/add-model-like.yml

@@ -18,32 +18,52 @@ jobs:
     steps:
       - uses: actions/checkout@v2
 
-      - name: Loading cache.
+      - name: Install dependencies
+        run: |
+          sudo apt -y update && sudo apt install -y libsndfile1-dev
+
+      - name: Load cached virtual environment
         uses: actions/cache@v2
         id: cache
         with:
-          path: ~/.cache/pip
-          key: v1-tests_model_like-${{ hashFiles('setup.py') }}
+          path: ~/venv/
+          key: v4-tests_model_like-${{ hashFiles('setup.py') }}
 
-      - name: Install dependencies
+      - name: Create virtual environment on cache miss
+        if: steps.cache.outputs.cache-hit != 'true'
         run: |
+          python -m venv ~/venv && . ~/venv/bin/activate
           pip install --upgrade pip!=21.3
-          pip install -U click  # Click 7 is installed in the environment by default, but we need at least version 8 for Black
-          sudo apt -y update && sudo apt install -y libsndfile1-dev
-          pip install .[dev]
+          pip install -e .[dev]
+
+      - name: Check transformers location
+        # make `transformers` available as package (required since we use `-e` flag) and check it's indeed from the repo.
+        run: |
+          . ~/venv/bin/activate
+          python setup.py develop
+          transformer_loc=$(pip show transformers | grep "Location: " | cut -c11-)
+          transformer_repo_loc=$(pwd .)
+          if [ "$transformer_loc" != "$transformer_repo_loc/src" ]; then
+              echo "transformers is from $transformer_loc but it shoud be from $transformer_repo_loc/src."
+              echo "A fix is required. Stop testing."
+              exit 1
+          fi
 
       - name: Create model files
         run: |
+          . ~/venv/bin/activate
           transformers-cli add-new-model-like --config_file tests/fixtures/add_distilbert_like_config.json --path_to_repo .
           make style
           make fix-copies
 
       - name: Run all PyTorch modeling test
         run: |
+          . ~/venv/bin/activate
           python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_new_models tests/bert_new/test_modeling_bert_new.py
 
       - name: Run style changes
         run: |
+          . ~/venv/bin/activate
           make style && make quality && make repo-consistency
 
       - name: Failure short reports

.github/workflows/build-docker-images.yml

@@ -5,6 +5,7 @@ on:
     branches:
       - docker-image*
   repository_dispatch:
+  workflow_call:
   schedule:
     - cron: "0 1 * * *"
 
@@ -39,9 +40,35 @@ jobs:
           push: true
           tags: huggingface/transformers-all-latest-gpu
 
+  latest-with-torch-nightly-docker:
+    name: "Nightly PyTorch + Stable TensorFlow"
+    runs-on: ubuntu-latest
+    steps:
+      -
+        name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v1
+      -
+        name: Check out code
+        uses: actions/checkout@v2
+      -
+        name: Login to DockerHub
+        uses: docker/login-action@v1
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_PASSWORD }}
+      -
+        name: Build and push
+        uses: docker/build-push-action@v2
+        with:
+          context: ./docker/transformers-all-latest-gpu
+          build-args: |
+            REF=main
+            PYTORCH=pre
+          push: true
+          tags: huggingface/transformers-all-latest-torch-nightly-gpu
+
   latest-torch-deepspeed-docker:
     name: "Latest PyTorch + DeepSpeed"
-    needs: latest-docker
     runs-on: ubuntu-latest
     steps:
       -
@@ -66,6 +93,32 @@ jobs:
           push: true
           tags: huggingface/transformers-pytorch-deepspeed-latest-gpu
 
+  nightly-torch-deepspeed-docker:
+    name: "Nightly PyTorch + DeepSpeed"
+    runs-on: ubuntu-latest
+    steps:
+      -
+        name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v1
+      -
+        name: Check out code
+        uses: actions/checkout@v2
+      -
+        name: Login to DockerHub
+        uses: docker/login-action@v1
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_PASSWORD }}
+      -
+        name: Build and push
+        uses: docker/build-push-action@v2
+        with:
+          context: ./docker/transformers-pytorch-deepspeed-nightly-gpu
+          build-args: |
+            REF=main
+          push: true
+          tags: huggingface/transformers-pytorch-deepspeed-nightly-gpu
+
   doc-builder:
     name: "Doc builder"
     runs-on: ubuntu-latest
@@ -93,7 +146,6 @@ jobs:
   latest-pytorch:
     name: "Latest PyTorch [dev]"
     runs-on: ubuntu-latest
-    needs: latest-torch-deepspeed-docker
     steps:
       -
         name: Set up Docker Buildx
@@ -118,7 +170,6 @@ jobs:
           tags: huggingface/transformers-pytorch-gpu
 
   latest-tensorflow:
-    needs: latest-pytorch
     name: "Latest TensorFlow [dev]"
     runs-on: ubuntu-latest
     steps:

.github/workflows/build-past-ci-docker-images.yml

@@ -0,0 +1,108 @@
+name: Build docker images (Past CI)
+
+on:
+  push:
+    branches:
+      - past-ci-docker-image*
+
+concurrency:
+  group: docker-images-builds
+  cancel-in-progress: false
+
+jobs:
+  past-pytorch-docker:
+    name: "Past PyTorch Docker"
+    strategy:
+      fail-fast: false
+      matrix:
+        version: ["1.11", "1.10", "1.9", "1.8", "1.7", "1.6", "1.5", "1.4"]
+    runs-on: ubuntu-latest
+    steps:
+      -
+        name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v1
+      -
+        name: Check out code
+        uses: actions/checkout@v2
+      -
+        name: Login to DockerHub
+        uses: docker/login-action@v1
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_PASSWORD }}
+      -
+        name: Build and push
+        uses: docker/build-push-action@v2
+        with:
+          context: ./docker/transformers-past-gpu
+          build-args: |
+            REF=main
+            FRAMEWORK=pytorch
+            VERSION=${{ matrix.version }}
+          push: true
+          tags: huggingface/transformers-pytorch-past-${{ matrix.version }}-gpu
+
+  past-tensorflow-docker:
+    name: "Past TensorFlow Docker"
+    strategy:
+      fail-fast: false
+      matrix:
+        version: ["2.8", "2.7", "2.6", "2.5"]
+    runs-on: ubuntu-latest
+    steps:
+      -
+        name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v1
+      -
+        name: Check out code
+        uses: actions/checkout@v2
+      -
+        name: Login to DockerHub
+        uses: docker/login-action@v1
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_PASSWORD }}
+      -
+        name: Build and push
+        uses: docker/build-push-action@v2
+        with:
+          context: ./docker/transformers-past-gpu
+          build-args: |
+            REF=main
+            FRAMEWORK=tensorflow
+            VERSION=${{ matrix.version }}
+          push: true
+          tags: huggingface/transformers-tensorflow-past-${{ matrix.version }}-gpu
+
+  past-tensorflow-docker-2-4:
+    name: "Past TensorFlow Docker"
+    strategy:
+      fail-fast: false
+      matrix:
+        version: ["2.4"]
+    runs-on: ubuntu-latest
+    steps:
+      -
+        name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v1
+      -
+        name: Check out code
+        uses: actions/checkout@v2
+      -
+        name: Login to DockerHub
+        uses: docker/login-action@v1
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_PASSWORD }}
+      -
+        name: Build and push
+        uses: docker/build-push-action@v2
+        with:
+          context: ./docker/transformers-past-gpu
+          build-args: |
+            REF=main
+            BASE_DOCKER_IMAGE=nvidia/cuda:11.0.3-cudnn8-devel-ubuntu20.04
+            FRAMEWORK=tensorflow
+            VERSION=${{ matrix.version }}
+          push: true
+          tags: huggingface/transformers-tensorflow-past-${{ matrix.version }}-gpu

.github/workflows/build_documentation.yml

@@ -15,6 +15,6 @@ jobs:
       commit_sha: ${{ github.sha }}
       package: transformers
       notebook_folder: transformers_doc
-      languages: en es
+      languages: en es it pt
     secrets:
       token: ${{ secrets.HUGGINGFACE_PUSH }}

.github/workflows/build_pr_documentation.yml

@@ -14,4 +14,4 @@ jobs:
       commit_sha: ${{ github.event.pull_request.head.sha }}
       pr_number: ${{ github.event.number }}
       package: transformers
-      languages: en es
+      languages: en es it pt

.github/workflows/doctests.yml

@@ -32,9 +32,7 @@ jobs:
 
       - name: GPU visibility
         run: |
-          utils/print_env_pt.py
-          TF_CPP_MIN_LOG_LEVEL=3 python3 -c "import tensorflow as tf; print('TF GPUs available:', bool(tf.config.list_physical_devices('GPU')))"
-          TF_CPP_MIN_LOG_LEVEL=3 python3 -c "import tensorflow as tf; print('Number of TF GPUs available:', len(tf.config.list_physical_devices('GPU')))"
+          python3 utils/print_env.py
 
       - name: Prepare files for doctests
         run: |

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

@@ -1,46 +0,0 @@
-name: Torch hub integration
-
-on:
-  push:
-    branches:
-      - "*"
-
-jobs:
-  torch_hub_integration:
-    runs-on: ubuntu-latest
-    env:
-      # TODO quickfix but may need more investigation
-      ACTIONS_ALLOW_UNSECURE_COMMANDS: True
-    steps:
-    # no checkout necessary here.
-    - name: Extract branch name
-      run: echo "::set-env name=BRANCH::${GITHUB_REF#refs/heads/}"
-    - name: Check branch name
-      run: echo $BRANCH
-    - name: Set up Python
-      uses: actions/setup-python@v1
-      with:
-        python-version: 3.7
-
-    - name: Loading cache
-      uses: actions/cache@v2
-      id: cache
-      with:
-        path: ~/.cache/pip
-        key: v0-torch_hub-${{ hashFiles('setup.py') }}
-
-    - name: Install dependencies
-      run: |
-        pip install --upgrade pip
-        # install torch-hub specific dependencies
-        pip install -e git+https://github.com/huggingface/transformers.git#egg=transformers[torchhub]
-        # no longer needed
-        pip uninstall -y transformers
-
-    #- name: Torch hub list
-    #  run: |
-    #    python -c "import torch; print(torch.hub.list('huggingface/transformers:$BRANCH'))"
-
-    #- name: Torch hub help
-    #  run: |
-    #    python -c "import torch; print(torch.hub.help('huggingface/transformers:$BRANCH', 'modelForSequenceClassification'))"

.github/workflows/model-templates.yml

@@ -1,43 +1,51 @@
 name: Model templates runner
 
 on:
-  push:
-    branches:
-      - main
-  pull_request:
-    paths:
-      - "src/**"
-      - "tests/**"
-      - ".github/**"
-      - "templates/**"
-    types: [assigned, opened, synchronize, reopened]
+  repository_dispatch:
+  schedule:
+    - cron: "0 2 * * *"
 
 jobs:
   run_tests_templates:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout repository
-        uses: actions/checkout@v1
-
-      - name: Install Python
-        uses: actions/setup-python@v1
-        with:
-          python-version: 3.6
-
-      - name: Loading cache.
-        uses: actions/cache@v2
-        id: cache
-        with:
-          path: ~/.cache/pip
-          key: v1.2-tests_templates-${{ hashFiles('setup.py') }}
+        uses: actions/checkout@v2
 
       - name: Install dependencies
         run: |
-          pip install --upgrade pip!=21.3
           sudo apt -y update && sudo apt install -y libsndfile1-dev
-          pip install .[dev]
+
+      - name: Load cached virtual environment
+        uses: actions/cache@v2
+        id: cache
+        with:
+          path: ~/venv/
+          key: v4-tests_templates-${{ hashFiles('setup.py') }}
+
+      - name: Create virtual environment on cache miss
+        if: steps.cache.outputs.cache-hit != 'true'
+        run: |
+          python -m venv ~/venv && . ~/venv/bin/activate
+          pip install --upgrade pip!=21.3
+          pip install -e .[dev]
+
+      - name: Check transformers location
+        # make `transformers` available as package (required since we use `-e` flag) and check it's indeed from the repo.
+        run: |
+          . ~/venv/bin/activate
+          python setup.py develop
+          transformer_loc=$(pip show transformers | grep "Location: " | cut -c11-)
+          transformer_repo_loc=$(pwd .)
+          if [ "$transformer_loc" != "$transformer_repo_loc/src" ]; then
+              echo "transformers is from $transformer_loc but it shoud be from $transformer_repo_loc/src."
+              echo "A fix is required. Stop testing."
+              exit 1
+          fi
+
       - name: Create model files
         run: |
+          . ~/venv/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
@@ -53,11 +61,12 @@ jobs:
 
       - name: Run all non-slow tests
         run: |
+          . ~/venv/bin/activate
           python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_templates tests/*template*
 
       - name: Run style changes
         run: |
-          git fetch origin main:main
+          . ~/venv/bin/activate
           make style && make quality && make repo-consistency
 
       - name: Failure short reports

.github/workflows/self-nightly-scheduled.yml

@@ -1,250 +1,236 @@
-name: Self-hosted runner; Nightly (scheduled)
+name: Self-hosted runner (nightly)
+
+# Note that each job's dependencies go into a corresponding docker file.
+#
+# For example for `run_all_tests_torch_cuda_extensions_gpu` the docker image is
+# `huggingface/transformers-pytorch-deepspeed-latest-gpu`, which can be found at
+# `docker/transformers-pytorch-deepspeed-latest-gpu/Dockerfile`
 
 on:
-    push:
-        branches:
-            - nightly_ci*
-    repository_dispatch:
-    schedule:
-        - cron: "0 0 */3 * *"
+  repository_dispatch:
+  schedule:
+    - cron: "0 16 * * *"
 
 env:
-    HF_HOME: /mnt/cache
-    TRANSFORMERS_IS_CI: yes
-    RUN_SLOW: yes
-    OMP_NUM_THREADS: 16
-    MKL_NUM_THREADS: 16
-    PYTEST_TIMEOUT: 600
-    SIGOPT_API_TOKEN: ${{ secrets.SIGOPT_API_TOKEN }}
+  HF_HOME: /mnt/cache
+  TRANSFORMERS_IS_CI: yes
+  OMP_NUM_THREADS: 8
+  MKL_NUM_THREADS: 8
+  RUN_SLOW: yes
+  SIGOPT_API_TOKEN: ${{ secrets.SIGOPT_API_TOKEN }}
+  TF_FORCE_GPU_ALLOW_GROWTH: true
+  RUN_PT_TF_CROSS_TESTS: 1
 
 jobs:
-    run_all_tests_torch_gpu:
-        runs-on: [self-hosted, docker-gpu, single-gpu]
-        container:
-            image: pytorch/pytorch:1.10.0-cuda11.3-cudnn8-runtime
-            options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-        steps:
-            - name: Launcher docker
-              uses: actions/checkout@v2
+  setup:
+    name: Setup
+    strategy:
+      matrix:
+        machine_type: [single-gpu, multi-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
+    container:
+      image: huggingface/transformers-all-latest-torch-nightly-gpu
+      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+    outputs:
+      matrix: ${{ steps.set-matrix.outputs.matrix }}
+    steps:
+      - name: Update clone
+        working-directory: /transformers
+        run: |
+          git fetch && git checkout ${{ github.sha }}
 
-            - name: NVIDIA-SMI
-              run: |
-                  nvidia-smi
+      - name: Cleanup
+        working-directory: /transformers
+        run: |
+          rm -rf tests/__pycache__
+          rm -rf tests/models/__pycache__
+          rm -rf reports
 
-            - name: Install dependencies
-              run: |
-                  apt -y update && apt install -y libsndfile1-dev git espeak-ng
-                  pip install --upgrade pip
-                  pip install .[integrations,sklearn,testing,onnxruntime,sentencepiece,torch-speech,vision,timm]
-                  pip install https://github.com/kpu/kenlm/archive/master.zip
-                  pip install --pre torch torchvision torchaudio -f https://download.pytorch.org/whl/nightly/cu113/torch_nightly.html -U
+      - id: set-matrix
+        name: Identify models to test
+        working-directory: /transformers/tests
+        run: |
+          echo "::set-output name=matrix::$(python3 -c 'import os; tests = os.getcwd(); model_tests = os.listdir(os.path.join(tests, "models")); d1 = sorted(list(filter(os.path.isdir, os.listdir(tests)))); d2 = sorted(list(filter(os.path.isdir, [f"models/{x}" for x in model_tests]))); d1.remove("models"); d = d2 + d1; print(d)')"
 
-            - name: Are GPUs recognized by our DL frameworks
-              run: |
-                utils/print_env_pt.py
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
 
-            - name: Run all tests on GPU
-              run: |
-                  python -m pytest -n 1 -v --dist=loadfile --make-reports=tests_torch_gpu tests
+  run_tests_single_gpu:
+    name: Model tests
+    strategy:
+      fail-fast: false
+      matrix:
+        folders: ${{ fromJson(needs.setup.outputs.matrix) }}
+        machine_type: [single-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
+    container:
+      image: huggingface/transformers-all-latest-torch-nightly-gpu
+      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+    needs: setup
+    steps:
+      - name: Echo folder ${{ matrix.folders }}
+        shell: bash
+        # For folders like `models/bert`, set an env. var. (`matrix_folders`) to `models_bert`, which will be used to
+        # set the artifact folder names (because the character `/` is not allowed).
+        run: |
+          echo "${{ matrix.folders }}"
+          matrix_folders=${{ matrix.folders }}
+          matrix_folders=${matrix_folders/'models/'/'models_'}
+          echo "$matrix_folders"
+          echo "matrix_folders=$matrix_folders" >> $GITHUB_ENV
 
-            - name: Failure short reports
-              if: ${{ always() }}
-              run: cat reports/tests_torch_gpu_failures_short.txt
+      - name: Update clone
+        working-directory: /transformers
+        run: git fetch && git checkout ${{ github.sha }}
 
-            - name: Run examples tests on GPU
-              if: ${{ always() }}
-              env:
-                  OMP_NUM_THREADS: 16
-                  MKL_NUM_THREADS: 16
-                  RUN_SLOW: yes
-                  HF_HOME: /mnt/cache
-                  TRANSFORMERS_IS_CI: yes
-              run: |
-                  pip install -r examples/pytorch/_tests_requirements.txt
-                  python -m pytest -n 1 -v --dist=loadfile --make-reports=examples_torch_gpu examples
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
 
-            - name: Failure short reports
-              if: ${{ always() }}
-              run: cat reports/examples_torch_gpu_failures_short.txt
+      - name: Environment
+        working-directory: /transformers
+        run: |
+          python3 utils/print_env.py
 
-            - name: Run all pipeline tests on GPU
-              if: ${{ always() }}
-              env:
-                  RUN_PIPELINE_TESTS: yes
-              run: |
-                  python -m pytest -n 1 -v --dist=loadfile -m is_pipeline_test --make-reports=tests_torch_pipeline_gpu tests
+      - name: Run all tests on GPU
+        working-directory: /transformers
+        run: python3 -m pytest -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} tests/${{ matrix.folders }}
 
-            - name: Failure short reports
-              if: ${{ always() }}
-              run: cat reports/tests_torch_pipeline_gpu_failures_short.txt
+      - name: Failure short reports
+        if: ${{ failure() }}
+        continue-on-error: true
+        run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/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
+      - name: Test suite reports artifacts
+        if: ${{ always() }}
+        uses: actions/upload-artifact@v2
+        with:
+          name: ${{ matrix.machine_type }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
+          path: /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}
 
-    run_all_tests_torch_multi_gpu:
-        runs-on: [self-hosted, docker-gpu, multi-gpu]
-        container:
-            image: pytorch/pytorch:1.10.0-cuda11.3-cudnn8-runtime
-            options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-        steps:
-            - name: Launcher docker
-              uses: actions/checkout@v2
+  run_tests_multi_gpu:
+    name: Model tests
+    strategy:
+      fail-fast: false
+      matrix:
+        folders: ${{ fromJson(needs.setup.outputs.matrix) }}
+        machine_type: [multi-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
+    container:
+      image: huggingface/transformers-all-latest-torch-nightly-gpu
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+    needs: setup
+    steps:
+      - name: Echo folder ${{ matrix.folders }}
+        shell: bash
+        # For folders like `models/bert`, set an env. var. (`matrix_folders`) to `models_bert`, which will be used to
+        # set the artifact folder names (because the character `/` is not allowed).
+        run: |
+          echo "${{ matrix.folders }}"
+          matrix_folders=${{ matrix.folders }}
+          matrix_folders=${matrix_folders/'models/'/'models_'}
+          echo "$matrix_folders"
+          echo "matrix_folders=$matrix_folders" >> $GITHUB_ENV
 
-            - name: NVIDIA-SMI
-              continue-on-error: true
-              run: |
-                  nvidia-smi
+      - name: Update clone
+        working-directory: /transformers
+        run: git fetch && git checkout ${{ github.sha }}
 
-            - name: Install dependencies
-              run: |
-                  apt -y update && apt install -y libsndfile1-dev git espeak-ng
-                  pip install --upgrade pip
-                  pip install .[integrations,sklearn,testing,onnxruntime,sentencepiece,torch-speech,vision,timm]
-                  pip install https://github.com/kpu/kenlm/archive/master.zip
-                  pip install --pre torch torchvision torchaudio -f https://download.pytorch.org/whl/nightly/cu113/torch_nightly.html -U
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
 
-            - name: Are GPUs recognized by our DL frameworks
-              run: |
-                utils/print_env_pt.py
+      - name: Environment
+        working-directory: /transformers
+        run: |
+          python3 utils/print_env.py
 
-            - name: Run all tests on GPU
-              env:
-                  MKL_SERVICE_FORCE_INTEL: 1
-              run: |
-                  python -m pytest -n 1 -v --dist=loadfile --make-reports=tests_torch_multi_gpu tests
+      - name: Run all tests on GPU
+        working-directory: /transformers
+        run: python3 -m pytest -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} tests/${{ matrix.folders }}
 
-            - name: Failure short reports
-              if: ${{ always() }}
-              run: cat reports/tests_torch_multi_gpu_failures_short.txt
+      - name: Failure short reports
+        if: ${{ failure() }}
+        continue-on-error: true
+        run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
 
-            - name: Run all pipeline tests on GPU
-              if: ${{ always() }}
-              env:
-                  RUN_PIPELINE_TESTS: yes
-              run: |
-                  python -m pytest -n 1 -v --dist=loadfile -m is_pipeline_test --make-reports=tests_torch_pipeline_multi_gpu tests
+      - name: Test suite reports artifacts
+        if: ${{ always() }}
+        uses: actions/upload-artifact@v2
+        with:
+          name: ${{ matrix.machine_type }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
+          path: /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}
 
-            - name: Failure short reports
-              if: ${{ always() }}
-              run: cat reports/tests_torch_pipeline_multi_gpu_failures_short.txt
+  run_all_tests_torch_cuda_extensions_gpu:
+    name: Torch CUDA extension tests
+    strategy:
+      fail-fast: false
+      matrix:
+        machine_type: [single-gpu, multi-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
+    needs: setup
+    container:
+      image: huggingface/transformers-pytorch-deepspeed-nightly-gpu
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+    steps:
+      - name: Update clone
+        working-directory: /workspace/transformers
+        run: git fetch && git checkout ${{ github.sha }}
 
-            - name: Test suite reports artifacts
-              if: ${{ always() }}
-              uses: actions/upload-artifact@v2
-              with:
-                  name: run_all_tests_torch_multi_gpu_test_reports
-                  path: reports
+      # To avoid unknown test failures
+      - name: Pre build DeepSpeed *again*
+        working-directory: /workspace
+        run: |
+          python3 -m pip uninstall -y deepspeed
+          rm -rf DeepSpeed
+          git clone https://github.com/microsoft/DeepSpeed && cd DeepSpeed && rm -rf build
+          DS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 python3 -m pip install . --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check
 
-    run_all_tests_torch_cuda_extensions_gpu:
-        runs-on: [self-hosted, docker-gpu, single-gpu]
-        container:
-            image: nvcr.io/nvidia/pytorch:21.03-py3
-            options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-        steps:
-            - name: Launcher docker
-              uses: actions/checkout@v2
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
 
-            - name: NVIDIA-SMI
-              run: |
-                  nvidia-smi
+      - name: Environment
+        working-directory: /workspace/transformers
+        run: |
+          python utils/print_env.py
 
-            - name: Install dependencies
-              run: |
-                  apt -y update && apt install -y libaio-dev libsndfile1-dev git espeak-ng
-                  pip install --upgrade pip
-                  pip install --pre torch torchvision torchaudio -f https://download.pytorch.org/whl/nightly/cu113/torch_nightly.html -U
-                  pip install .[testing,deepspeed]
-                  pip install https://github.com/kpu/kenlm/archive/master.zip
-                  pip install git+https://github.com/microsoft/DeepSpeed
+      - name: Run all tests on GPU
+        working-directory: /workspace/transformers
+        run: |
+          python -m pytest -v --make-reports=${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu tests/deepspeed tests/extended
 
-            - name: Are GPUs recognized by our DL frameworks
-              run: |
-                utils/print_env_pt.py
+      - name: Failure short reports
+        if: ${{ failure() }}
+        continue-on-error: true
+        run: cat /workspace/transformers/reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu/failures_short.txt
 
-            - name: Run all tests on GPU
-              run: |
-                  python -m pytest -n 1 -v --dist=loadfile --make-reports=tests_torch_cuda_extensions_gpu tests/deepspeed tests/extended
+      - name: Test suite reports artifacts
+        if: ${{ always() }}
+        uses: actions/upload-artifact@v2
+        with:
+          name: ${{ matrix.machine_type }}_run_tests_torch_cuda_extensions_gpu_test_reports
+          path: /workspace/transformers/reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu
 
-            - name: Failure short reports
-              if: ${{ always() }}
-              run: cat reports/tests_torch_cuda_extensions_gpu_failures_short.txt
-
-            - name: Test suite reports artifacts
-              if: ${{ always() }}
-              uses: actions/upload-artifact@v2
-              with:
-                  name: run_tests_torch_cuda_extensions_gpu_test_reports
-                  path: reports
-
-    run_all_tests_torch_cuda_extensions_multi_gpu:
-        runs-on: [self-hosted, docker-gpu, multi-gpu]
-        container:
-            image: nvcr.io/nvidia/pytorch:21.03-py3
-            options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-        steps:
-            - name: Launcher docker
-              uses: actions/checkout@v2
-
-            - name: NVIDIA-SMI
-              continue-on-error: true
-              run: |
-                  nvidia-smi
-
-            - name: Install dependencies
-              run: |
-                  apt -y update && apt install -y libaio-dev libsndfile1-dev git espeak-ng
-                  pip install --upgrade pip
-                  pip install --pre torch torchvision torchaudio -f https://download.pytorch.org/whl/nightly/cu113/torch_nightly.html -U
-                  rm -rf ~/.cache/torch_extensions/ # shared between conflicting builds
-                  pip install .[testing,fairscale]
-                  pip install https://github.com/kpu/kenlm/archive/master.zip
-                  pip install git+https://github.com/microsoft/DeepSpeed # testing bleeding edge
-
-            - name: Are GPUs recognized by our DL frameworks
-              run: |
-                utils/print_env_pt.py
-
-            - name: Run all tests on GPU
-              run: |
-                  python -m pytest -n 1 -v --dist=loadfile --make-reports=tests_torch_cuda_extensions_multi_gpu tests/deepspeed tests/extended
-
-            - name: Failure short reports
-              if: ${{ always() }}
-              run: cat reports/tests_torch_cuda_extensions_multi_gpu_failures_short.txt
-
-            - name: Test suite reports artifacts
-              if: ${{ always() }}
-              uses: actions/upload-artifact@v2
-              with:
-                  name: run_tests_torch_cuda_extensions_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_torch_multi_gpu,
-                run_all_tests_torch_cuda_extensions_gpu,
-                run_all_tests_torch_cuda_extensions_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 }}
-                  CI_SLACK_CHANNEL_ID_DAILY: ${{ secrets.CI_SLACK_CHANNEL_ID_DAILY }}
-                  CI_SLACK_CHANNEL_ID_PAST_FUTURE: ${{ secrets.CI_SLACK_CHANNEL_ID_PAST_FUTURE }}
-
-              run: |
-                  pip install slack_sdk
-                  python utils/notification_service.py scheduled nightly-torch
+  send_results:
+    name: Send results to webhook
+    runs-on: ubuntu-latest
+    if: always()
+    needs: [setup, run_tests_single_gpu, run_tests_multi_gpu, run_all_tests_torch_cuda_extensions_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 }}
+          CI_SLACK_CHANNEL_ID_DAILY: ${{ secrets.CI_SLACK_CHANNEL_ID_DAILY }}
+          CI_SLACK_CHANNEL_DUMMY_TESTS: ${{ secrets.CI_SLACK_CHANNEL_DUMMY_TESTS }}
+          CI_SLACK_REPORT_CHANNEL_ID: ${{ secrets.CI_SLACK_CHANNEL_ID_PAST_FUTURE }}
+          CI_EVENT: nightly-build
+        # We pass `needs.setup.outputs.matrix` as the argument. A processing in `notification_service.py` to change
+        # `models/bert` to `models_bert` is required, as the artifact names use `_` instead of `/`.
+        run: |
+          pip install slack_sdk
+          python utils/notification_service.py "${{ needs.setup.outputs.matrix }}"

.github/workflows/self-past-caller.yml

@@ -0,0 +1,136 @@
+name: Self-hosted runner (past-ci-caller)
+
+on:
+  push:
+    branches:
+      - run-past-ci*
+
+jobs:
+  run_past_ci_pytorch_1-11:
+    name: PyTorch 1.11
+    if: always()
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: pytorch
+      version: "1.11"
+    secrets: inherit
+
+  run_past_ci_pytorch_1-10:
+    name: PyTorch 1.10
+    if: always()
+    needs: [run_past_ci_pytorch_1-11]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: pytorch
+      version: "1.10"
+    secrets: inherit
+
+  run_past_ci_pytorch_1-9:
+    name: PyTorch 1.9
+    if: always()
+    needs: [run_past_ci_pytorch_1-10]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: pytorch
+      version: "1.9"
+    secrets: inherit
+
+  run_past_ci_pytorch_1-8:
+    name: PyTorch 1.8
+    if: always()
+    needs: [run_past_ci_pytorch_1-9]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: pytorch
+      version: "1.8"
+    secrets: inherit
+
+  run_past_ci_pytorch_1-7:
+    name: PyTorch 1.7
+    if: always()
+    needs: [run_past_ci_pytorch_1-8]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: pytorch
+      version: "1.7"
+    secrets: inherit
+
+  run_past_ci_pytorch_1-6:
+    name: PyTorch 1.6
+    if: always()
+    needs: [run_past_ci_pytorch_1-7]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: pytorch
+      version: "1.6"
+    secrets: inherit
+
+  run_past_ci_pytorch_1-5:
+    name: PyTorch 1.5
+    if: always()
+    needs: [run_past_ci_pytorch_1-6]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: pytorch
+      version: "1.5"
+    secrets: inherit
+
+  run_past_ci_pytorch_1-4:
+    name: PyTorch 1.4
+    if: always()
+    needs: [run_past_ci_pytorch_1-5]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: pytorch
+      version: "1.4"
+    secrets: inherit
+
+  run_past_ci_tensorflow_2-8:
+    name: TensorFlow 2.8
+    if: always()
+    needs: [run_past_ci_pytorch_1-4]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: tensorflow
+      version: "2.8"
+    secrets: inherit
+
+  run_past_ci_tensorflow_2-7:
+    name: TensorFlow 2.7
+    if: always()
+    needs: [run_past_ci_tensorflow_2-8]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: tensorflow
+      version: "2.7"
+    secrets: inherit
+
+  run_past_ci_tensorflow_2-6:
+    name: TensorFlow 2.6
+    if: always()
+    needs: [run_past_ci_tensorflow_2-7]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: tensorflow
+      version: "2.6"
+    secrets: inherit
+
+  run_past_ci_tensorflow_2-5:
+    name: TensorFlow 2.5
+    if: always()
+    needs: [run_past_ci_tensorflow_2-6]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: tensorflow
+      version: "2.5"
+    secrets: inherit
+
+  run_past_ci_tensorflow_2-4:
+    name: TensorFlow 2.4
+    if: always()
+    needs: [run_past_ci_tensorflow_2-5]
+    uses: ./.github/workflows/self-past.yml
+    with:
+      framework: tensorflow
+      version: "2.4"
+    secrets: inherit

.github/workflows/self-past.yml

@@ -0,0 +1,192 @@
+name: Self-hosted runner (past)
+
+# Note that each job's dependencies go into a corresponding docker file.
+#
+# For example for `run_all_tests_torch_cuda_extensions_gpu` the docker image is
+# `huggingface/transformers-pytorch-deepspeed-latest-gpu`, which can be found at
+# `docker/transformers-pytorch-deepspeed-latest-gpu/Dockerfile`
+
+on:
+  workflow_call:
+    inputs:
+      framework:
+        required: true
+        type: string
+      version:
+        required: true
+        type: string
+
+env:
+  HF_HOME: /mnt/cache
+  TRANSFORMERS_IS_CI: yes
+  OMP_NUM_THREADS: 8
+  MKL_NUM_THREADS: 8
+  RUN_SLOW: yes
+  SIGOPT_API_TOKEN: ${{ secrets.SIGOPT_API_TOKEN }}
+  TF_FORCE_GPU_ALLOW_GROWTH: true
+  RUN_PT_TF_CROSS_TESTS: 1
+
+jobs:
+  setup:
+    name: Setup
+    runs-on: ubuntu-latest
+    outputs:
+      matrix: ${{ steps.set-matrix.outputs.matrix }}
+    steps:
+      - name: Checkout transformers
+        uses: actions/checkout@v2
+        with:
+          fetch-depth: 2
+
+      - name: Cleanup
+        run: |
+          rm -rf tests/__pycache__
+          rm -rf tests/models/__pycache__
+          rm -rf reports
+
+      - id: set-matrix
+        name: Identify models to test
+        run: |
+          cd tests
+          echo "::set-output name=matrix::$(python3 -c 'import os; tests = os.getcwd(); model_tests = os.listdir(os.path.join(tests, "models")); d1 = sorted(list(filter(os.path.isdir, os.listdir(tests)))); d2 = sorted(list(filter(os.path.isdir, [f"models/{x}" for x in model_tests]))); d1.remove("models"); d = d2 + d1; print(d)')"
+
+  run_tests_single_gpu:
+    name: Model tests
+    strategy:
+      fail-fast: false
+      matrix:
+        folders: ${{ fromJson(needs.setup.outputs.matrix) }}
+        machine_type: [single-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker-past-ci') }}
+    container:
+      image: huggingface/transformers-${{ inputs.framework }}-past-${{ inputs.version }}-gpu
+      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+    needs: setup
+    steps:
+      - name: Update clone
+        working-directory: /transformers
+        run: git fetch && git checkout ${{ github.sha }}
+
+      - name: Echo folder ${{ matrix.folders }}
+        shell: bash
+        # For folders like `models/bert`, set an env. var. (`matrix_folders`) to `models_bert`, which will be used to
+        # set the artifact folder names (because the character `/` is not allowed).
+        run: |
+          echo "${{ matrix.folders }}"
+          matrix_folders=${{ matrix.folders }}
+          matrix_folders=${matrix_folders/'models/'/'models_'}
+          echo "$matrix_folders"
+          echo "matrix_folders=$matrix_folders" >> $GITHUB_ENV
+
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
+
+      - name: Environment
+        working-directory: /transformers
+        run: |
+          python3 utils/print_env.py
+
+      - name: Run all tests on GPU
+        working-directory: /transformers
+        run: python3 -m pytest -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} tests/${{ matrix.folders }}
+
+      - name: Failure short reports
+        if: ${{ failure() }}
+        continue-on-error: true
+        run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
+
+      - name: Test suite reports artifacts
+        if: ${{ always() }}
+        uses: actions/upload-artifact@v2
+        with:
+          name: ${{ matrix.machine_type }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
+          path: /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}
+
+  run_tests_multi_gpu:
+    name: Model tests
+    strategy:
+      fail-fast: false
+      matrix:
+        folders: ${{ fromJson(needs.setup.outputs.matrix) }}
+        machine_type: [multi-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker-past-ci') }}
+    container:
+      image: huggingface/transformers-${{ inputs.framework }}-past-${{ inputs.version }}-gpu
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+    needs: setup
+    steps:
+      - name: Update clone
+        working-directory: /transformers
+        run: git fetch && git checkout ${{ github.sha }}
+
+      - name: Echo folder ${{ matrix.folders }}
+        shell: bash
+        # For folders like `models/bert`, set an env. var. (`matrix_folders`) to `models_bert`, which will be used to
+        # set the artifact folder names (because the character `/` is not allowed).
+        run: |
+          echo "${{ matrix.folders }}"
+          matrix_folders=${{ matrix.folders }}
+          matrix_folders=${matrix_folders/'models/'/'models_'}
+          echo "$matrix_folders"
+          echo "matrix_folders=$matrix_folders" >> $GITHUB_ENV
+
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
+
+      - name: Environment
+        working-directory: /transformers
+        run: |
+          python3 utils/print_env.py
+
+      - name: Run all tests on GPU
+        working-directory: /transformers
+        run: python3 -m pytest -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} tests/${{ matrix.folders }}
+
+      - name: Failure short reports
+        if: ${{ failure() }}
+        continue-on-error: true
+        run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
+
+      - name: Test suite reports artifacts
+        if: ${{ always() }}
+        uses: actions/upload-artifact@v2
+        with:
+          name: ${{ matrix.machine_type }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
+          path: /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}
+
+  send_results:
+    name: Send results to webhook
+    runs-on: ubuntu-latest
+    if: always()
+    needs: [setup, run_tests_single_gpu, run_tests_multi_gpu]
+    steps:
+      - uses: actions/checkout@v2
+      - uses: actions/download-artifact@v2
+
+      # Create a directory to store test failure tables in the next step
+      - name: Create directory
+        run: mkdir test_failure_tables
+
+      - name: Send message to Slack
+        env:
+          CI_SLACK_BOT_TOKEN: ${{ secrets.CI_SLACK_BOT_TOKEN }}
+          CI_SLACK_CHANNEL_ID: ${{ secrets.CI_SLACK_CHANNEL_ID }}
+          CI_SLACK_CHANNEL_ID_DAILY: ${{ secrets.CI_SLACK_CHANNEL_ID_DAILY }}
+          CI_SLACK_CHANNEL_DUMMY_TESTS: ${{ secrets.CI_SLACK_CHANNEL_DUMMY_TESTS }}
+          CI_SLACK_REPORT_CHANNEL_ID: ${{ secrets.CI_SLACK_CHANNEL_ID_PAST_FUTURE }}
+          CI_EVENT: Past CI - ${{ inputs.framework }}-${{ inputs.version }}
+        # We pass `needs.setup.outputs.matrix` as the argument. A processing in `notification_service.py` to change
+        # `models/bert` to `models_bert` is required, as the artifact names use `_` instead of `/`.
+        run: |
+          pip install slack_sdk
+          python utils/notification_service.py "${{ needs.setup.outputs.matrix }}"
+
+      # Upload complete failure tables, as they might be big and only truncated versions could be sent to Slack.
+      - name: Failure table artifacts
+        if: ${{ always() }}
+        uses: actions/upload-artifact@v2
+        with:
+          name: test_failure_tables_${{ inputs.framework }}-${{ inputs.version }}
+          path: test_failure_tables

.github/workflows/self-push-caller.yml

@@ -0,0 +1,52 @@
+# Used to trigger self-push CI
+name: Self-hosted runner (push-caller)
+
+on:
+  push:
+    branches:
+      - main
+    paths:
+      - "src/**"
+      - "tests/**"
+      - ".github/**"
+      - "templates/**"
+      - "utils/**"
+
+jobs:
+  check-for-setup:
+      runs-on: ubuntu-latest
+      name: Check if setup was changed
+      outputs:
+        changed: ${{ steps.was_changed.outputs.changed }}
+      steps:
+        - uses: actions/checkout@v3
+          with: 
+            fetch-depth: "2"
+        
+        - name: Get changed files
+          id: changed-files
+          uses: tj-actions/changed-files@v22.2
+        
+        - name: Was setup changed 
+          id: was_changed
+          run: |
+            for file in ${{ steps.changed-files.outputs.all_changed_files }}; do
+              if [ `basename "${file}"` = "setup.py" ]; then
+                echo ::set-output name=changed::"1"
+              fi
+            done
+
+  build-docker-containers:
+    needs: check-for-setup
+    if: (github.event_name == 'push') && (needs.check-for-setup.outputs.changed == '1')
+    uses: ./.github/workflows/build-docker-images.yml
+    secrets: inherit
+
+  run_push_ci:
+    name: Trigger Push CI
+    runs-on: ubuntu-latest
+    if: ${{ always() }}
+    needs: build-docker-containers
+    steps:
+      - name: Trigger push CI via workflow_run
+        run: echo "Trigger push CI via workflow_run"

.github/workflows/self-push.yml

@@ -1,9 +1,12 @@
 name: Self-hosted runner (push)
 
 on:
+  workflow_run:
+    workflows: ["Self-hosted runner (push-caller)"]
+    branches: ["main"]
+    types: [completed]
   push:
     branches:
-      - main
       - ci_*
       - ci-*
     paths:
@@ -20,37 +23,68 @@ env:
   OMP_NUM_THREADS: 8
   MKL_NUM_THREADS: 8
   PYTEST_TIMEOUT: 60
+  TF_FORCE_GPU_ALLOW_GROWTH: true
+  RUN_PT_TF_CROSS_TESTS: 1
 
 jobs:
-  run_tests_torch_gpu:
-    runs-on: [self-hosted, docker-gpu, single-gpu]
-    container:
-      image: pytorch/pytorch:1.9.0-cuda11.1-cudnn8-runtime
-      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+  setup:
+    name: Setup
+    runs-on: ubuntu-latest
+    outputs:
+      matrix: ${{ steps.set-matrix.outputs.matrix }}
+      test_map: ${{ steps.set-matrix.outputs.test_map }}
     steps:
-      - name: Install dependencies
+      # Necessary to get the correct branch name and commit SHA for `workflow_run` event
+      # We also take into account the `push` event (we might want to test some changes in a branch)
+      - name: Prepare custom environment variables
+        shell: bash
+        # `CI_BRANCH_PUSH`: The branch name from the push event
+        # `CI_BRANCH_WORKFLOW_RUN`: The name of the branch on which this workflow is triggered by `workflow_run` event
+        # `CI_BRANCH`: The non-empty branch name from the above two (one and only one of them is empty)
+        # `CI_SHA_PUSH`: The commit SHA from the push event
+        # `CI_SHA_WORKFLOW_RUN`: The commit SHA that triggers this workflow by `workflow_run` event
+        # `CI_SHA`: The non-empty commit SHA from the above two (one and only one of them is empty)
         run: |
-          apt -y update && apt install -y software-properties-common && apt -y update && add-apt-repository -y ppa:git-core/ppa && apt -y update && apt install -y git
-          apt install -y libsndfile1-dev espeak-ng
-          pip install --upgrade pip
-          pip install .[sklearn,testing,onnxruntime,sentencepiece,torch-speech,vision,timm]
-          pip install https://github.com/kpu/kenlm/archive/master.zip
+          CI_BRANCH_PUSH=${{ github.event.ref }}
+          CI_BRANCH_PUSH=${CI_BRANCH_PUSH/'refs/heads/'/''}
+          CI_BRANCH_WORKFLOW_RUN=${{ github.event.workflow_run.head_branch }}
+          CI_SHA_PUSH=${{ github.event.head_commit.id }}
+          CI_SHA_WORKFLOW_RUN=${{ github.event.workflow_run.head_sha }}
+          echo $CI_BRANCH_PUSH
+          echo $CI_BRANCH_WORKFLOW_RUN
+          echo $CI_SHA_PUSH
+          echo $CI_SHA_WORKFLOW_RUN
+          [[ ! -z "$CI_BRANCH_PUSH" ]] && echo "CI_BRANCH=$CI_BRANCH_PUSH" >> $GITHUB_ENV || echo "CI_BRANCH=$CI_BRANCH_WORKFLOW_RUN" >> $GITHUB_ENV
+          [[ ! -z "$CI_SHA_PUSH" ]] && echo "CI_SHA=$CI_SHA_PUSH" >> $GITHUB_ENV || echo "CI_SHA=$CI_SHA_WORKFLOW_RUN" >> $GITHUB_ENV
 
-      - name: Launcher docker
+      - name: print environment variables
+        run: |
+          echo "env.CI_BRANCH = ${{ env.CI_BRANCH }}"
+          echo "env.CI_SHA = ${{ env.CI_SHA }}"
+
+      - name: Checkout transformers
         uses: actions/checkout@v2
         with:
           fetch-depth: 2
 
-      - name: NVIDIA-SMI
+      - name: Update clone using environment variables
         run: |
-          nvidia-smi
+          echo "original branch = $(git branch --show-current)"
+          git fetch && git checkout ${{ env.CI_BRANCH }}
+          echo "updated branch = $(git branch --show-current)"
+          git checkout ${{ env.CI_SHA }}
+          echo "log = $(git log -n 1)"
 
-      - name: Are GPUs recognized by our DL frameworks
+      - name: Cleanup
         run: |
-          utils/print_env_pt.py
+          rm -rf tests/__pycache__
+          rm -rf tests/models/__pycache__
+          rm -rf reports
 
       - name: Fetch the tests to run
+        # TODO: add `git-python` in the docker images
         run: |
+          pip install --upgrade git-python
           python utils/tests_fetcher.py --diff_with_last_commit | tee test_preparation.txt
 
       - name: Report fetched tests
@@ -59,441 +93,413 @@ jobs:
           name: test_fetched
           path: test_preparation.txt
 
-      - name: Run all non-slow tests on GPU
+      - id: set-matrix
+        name: Organize tests into models
+        # The `keys` is used as GitHub actions matrix for jobs, i.e. `models/bert`, `tokenization`, `pipeline`, etc.
+        # The `test_map` is used to get the actual identified test files under each key.
+        # If no test to run (so no `test_map.json` file), create a dummy map (empty matrix will fail)
         run: |
-          if [ -f test_list.txt ]; then
-            python -m pytest -n 2 --dist=loadfile -v --make-reports=tests_torch_gpu $(cat test_list.txt)
+          if [ -f test_map.json ]; then
+              keys=$(python3 -c 'import json; fp = open("test_map.json"); test_map = json.load(fp); fp.close(); d = list(test_map.keys()); print(d)')
+              test_map=$(python3 -c 'import json; fp = open("test_map.json"); test_map = json.load(fp); fp.close(); print(test_map)')
+          else
+              keys=$(python3 -c 'keys = ["dummy"]; print(keys)')
+              test_map=$(python3 -c 'test_map = {"dummy": []}; print(test_map)')
           fi
+          echo $keys
+          echo $test_map
+          echo "::set-output name=matrix::$keys"
+          echo "::set-output name=test_map::$test_map"
+
+  run_tests_single_gpu:
+    name: Model tests
+    needs: setup
+    # `dummy` means there is no test to run
+    if: contains(fromJson(needs.setup.outputs.matrix), 'dummy') != true
+    strategy:
+      fail-fast: false
+      matrix:
+        folders: ${{ fromJson(needs.setup.outputs.matrix) }}
+        machine_type: [single-gpu]
+    runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
+    container:
+      image: huggingface/transformers-all-latest-gpu
+      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+    steps:
+      # Necessary to get the correct branch name and commit SHA for `workflow_run` event
+      # We also take into account the `push` event (we might want to test some changes in a branch)
+      - name: Prepare custom environment variables
+        shell: bash
+        # For the meaning of these environment variables, see the job `Setup`
+        run: |
+          CI_BRANCH_PUSH=${{ github.event.ref }}
+          CI_BRANCH_PUSH=${CI_BRANCH_PUSH/'refs/heads/'/''}
+          CI_BRANCH_WORKFLOW_RUN=${{ github.event.workflow_run.head_branch }}
+          CI_SHA_PUSH=${{ github.event.head_commit.id }}
+          CI_SHA_WORKFLOW_RUN=${{ github.event.workflow_run.head_sha }}
+          echo $CI_BRANCH_PUSH
+          echo $CI_BRANCH_WORKFLOW_RUN
+          echo $CI_SHA_PUSH
+          echo $CI_SHA_WORKFLOW_RUN
+          [[ ! -z "$CI_BRANCH_PUSH" ]] && echo "CI_BRANCH=$CI_BRANCH_PUSH" >> $GITHUB_ENV || echo "CI_BRANCH=$CI_BRANCH_WORKFLOW_RUN" >> $GITHUB_ENV
+          [[ ! -z "$CI_SHA_PUSH" ]] && echo "CI_SHA=$CI_SHA_PUSH" >> $GITHUB_ENV || echo "CI_SHA=$CI_SHA_WORKFLOW_RUN" >> $GITHUB_ENV
+
+      - name: print environment variables
+        run: |
+          echo "env.CI_BRANCH = ${{ env.CI_BRANCH }}"
+          echo "env.CI_SHA = ${{ env.CI_SHA }}"
+
+      - name: Update clone using environment variables
+        working-directory: /transformers
+        run: |
+          echo "original branch = $(git branch --show-current)"
+          git fetch && git checkout ${{ env.CI_BRANCH }}
+          echo "updated branch = $(git branch --show-current)"
+          git checkout ${{ env.CI_SHA }}
+          echo "log = $(git log -n 1)"
+
+      - name: Echo folder ${{ matrix.folders }}
+        shell: bash
+        # For folders like `models/bert`, set an env. var. (`matrix_folders`) to `models_bert`, which will be used to
+        # set the artifact folder names (because the character `/` is not allowed).
+        run: |
+          echo "${{ matrix.folders }}"
+          echo "${{ fromJson(needs.setup.outputs.test_map)[matrix.folders] }}"
+          matrix_folders=${{ matrix.folders }}
+          matrix_folders=${matrix_folders/'models/'/'models_'}
+          echo "$matrix_folders"
+          echo "matrix_folders=$matrix_folders" >> $GITHUB_ENV
+
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
+
+      - name: Environment
+        working-directory: /transformers
+        run: |
+          python3 utils/print_env.py
+
+      - name: Run all non-slow selected tests on GPU
+        working-directory: /transformers
+        run: |
+          python3 -m pytest -n 2 --dist=loadfile -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} ${{ fromJson(needs.setup.outputs.test_map)[matrix.folders] }}
 
       - name: Failure short reports
         if: ${{ failure() }}
-        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_flax_gpu:
-    runs-on: [self-hosted, docker-gpu-test, 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:
-      - name: Set up Python 3.7
-        uses: actions/setup-python@v2
-        with:
-          python-version: 3.7
-
-      - name: Install dependencies
-        run: |
-          apt -y update && apt install -y software-properties-common && apt -y update && add-apt-repository -y ppa:git-core/ppa && apt -y update && apt install -y git espeak-ng
-          pip install --upgrade "jax[cuda111]" -f https://storage.googleapis.com/jax-releases/jax_releases.html
-          pip install --upgrade pip
-          pip install .[sklearn,testing,sentencepiece,flax,flax-speech,vision]
-
-      - name: Launcher docker
-        uses: actions/checkout@v2
-        with:
-          fetch-depth: 2
-
-      - name: NVIDIA-SMI
         continue-on-error: true
-        run: |
-          nvidia-smi
-
-      - name: Are GPUs recognized by our DL frameworks
-        run: |
-          python -c "from jax.lib import xla_bridge; print('GPU available:', xla_bridge.get_backend().platform)"
-          python -c "import jax; print('Number of GPUs available:', len(jax.local_devices()))"
-
-      - name: Fetch the tests to run
-        run: |
-          python utils/tests_fetcher.py --diff_with_last_commit | tee test_preparation.txt
-
-      - name: Report fetched tests
-        uses: actions/upload-artifact@v2
-        with:
-          name: test_fetched
-          path: test_preparation.txt
-
-      - name: Run all non-slow tests on GPU
-        run: |
-          if [ -f test_list.txt ]; then
-            python -m pytest -n 2 --dist=loadfile -v --make-reports=tests_flax_gpu $(cat test_list.txt)
-          fi
-
-      - name: Failure short reports
-        if: ${{ failure() }}
-        run: cat reports/tests_flax_gpu_failures_short.txt
+        run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
         with:
-          name: run_all_tests_flax_gpu_test_reports
-          path: reports
+          name: ${{ matrix.machine_type }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
+          path: /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}
 
-#  run_tests_tf_gpu:
-#    runs-on: [self-hosted, docker-gpu, single-gpu]
-#    timeout-minutes: 120
-#    container:
-#      image: tensorflow/tensorflow:2.4.1-gpu
-#      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-#    steps:
-#      - name: Install dependencies
-#        run: |
-#          apt -y update && apt install -y software-properties-common && apt -y update && add-apt-repository -y ppa:git-core/ppa && apt -y update && apt install -y git espeak-ng
-#          pip install --upgrade pip
-#          pip install .[sklearn,testing,onnxruntime,sentencepiece,tf-speech]
-#          pip install https://github.com/kpu/kenlm/archive/master.zip
-#
-#      - name: Launcher docker
-#        uses: actions/checkout@v2
-#        with:
-#          fetch-depth: 2
-#
-#      - name: NVIDIA-SMI
-#        run: |
-#          nvidia-smi
-#
-#      - name: Are GPUs recognized by our DL frameworks
-#        run: |
-#          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: Fetch the tests to run
-#        run: |
-#          python utils/tests_fetcher.py --diff_with_last_commit | tee test_preparation.txt
-#
-#      - name: Report fetched tests
-#        uses: actions/upload-artifact@v2
-#        with:
-#          name: test_fetched
-#          path: test_preparation.txt
-#
-#      - name: Run all non-slow tests on GPU
-#        env:
-#          TF_NUM_INTRAOP_THREADS: 8
-#          TF_NUM_INTEROP_THREADS: 1
-#        run: |
-#          if [ -f test_list.txt ]; then
-#            python -m pytest -n 1 --dist=loadfile --make-reports=tests_tf_gpu $(cat test_list.txt)
-#          fi
-#
-#      - name: Failure short reports
-#        if: ${{ failure() }}
-#        run: cat reports/tests_tf_gpu_failures_short.txt
-#
-#      - name: Test suite reports artifacts
-#        if: ${{ always() }}
-#        uses: actions/upload-artifact@v2
-#        with:
-#          name: run_all_tests_tf_gpu_test_reports
-#          path: reports
-
-
-  run_tests_torch_multi_gpu:
-    runs-on: [self-hosted, docker-gpu, multi-gpu]
+  run_tests_multi_gpu:
+    name: Model tests
+    needs: setup
+    # `dummy` means there is no test to run
+    if: contains(fromJson(needs.setup.outputs.matrix), 'dummy') != true
+    strategy:
+      fail-fast: false
+      matrix:
+        folders: ${{ fromJson(needs.setup.outputs.matrix) }}
+        machine_type: [multi-gpu]
+    runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
     container:
-      image: pytorch/pytorch:1.9.0-cuda11.1-cudnn8-runtime
+      image: huggingface/transformers-all-latest-gpu
       options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - name: Install dependencies
+      # Necessary to get the correct branch name and commit SHA for `workflow_run` event
+      # We also take into account the `push` event (we might want to test some changes in a branch)
+      - name: Prepare custom environment variables
+        shell: bash
+        # For the meaning of these environment variables, see the job `Setup`
         run: |
-          apt -y update && apt install -y software-properties-common && apt -y update && add-apt-repository -y ppa:git-core/ppa && apt -y update && apt install -y git espeak-ng
-          apt install -y libsndfile1-dev espeak-ng
-          pip install --upgrade pip
-          pip install .[sklearn,testing,onnxruntime,sentencepiece,torch-speech,vision,timm]
-          pip install https://github.com/kpu/kenlm/archive/master.zip
-      - name: Launcher docker
-        uses: actions/checkout@v2
-        with:
-          fetch-depth: 2
+          CI_BRANCH_PUSH=${{ github.event.ref }}
+          CI_BRANCH_PUSH=${CI_BRANCH_PUSH/'refs/heads/'/''}
+          CI_BRANCH_WORKFLOW_RUN=${{ github.event.workflow_run.head_branch }}
+          CI_SHA_PUSH=${{ github.event.head_commit.id }}
+          CI_SHA_WORKFLOW_RUN=${{ github.event.workflow_run.head_sha }}
+          echo $CI_BRANCH_PUSH
+          echo $CI_BRANCH_WORKFLOW_RUN
+          echo $CI_SHA_PUSH
+          echo $CI_SHA_WORKFLOW_RUN
+          [[ ! -z "$CI_BRANCH_PUSH" ]] && echo "CI_BRANCH=$CI_BRANCH_PUSH" >> $GITHUB_ENV || echo "CI_BRANCH=$CI_BRANCH_WORKFLOW_RUN" >> $GITHUB_ENV
+          [[ ! -z "$CI_SHA_PUSH" ]] && echo "CI_SHA=$CI_SHA_PUSH" >> $GITHUB_ENV || echo "CI_SHA=$CI_SHA_WORKFLOW_RUN" >> $GITHUB_ENV
+
+      - name: print environment variables
+        run: |
+          echo "env.CI_BRANCH = ${{ env.CI_BRANCH }}"
+          echo "env.CI_SHA = ${{ env.CI_SHA }}"
+
+      - name: Update clone using environment variables
+        working-directory: /transformers
+        run: |
+          echo "original branch = $(git branch --show-current)"
+          git fetch && git checkout ${{ env.CI_BRANCH }}
+          echo "updated branch = $(git branch --show-current)"
+          git checkout ${{ env.CI_SHA }}
+          echo "log = $(git log -n 1)"
+
+      - name: Echo folder ${{ matrix.folders }}
+        shell: bash
+        # For folders like `models/bert`, set an env. var. (`matrix_folders`) to `models_bert`, which will be used to
+        # set the artifact folder names (because the character `/` is not allowed).
+        run: |
+          echo "${{ matrix.folders }}"
+          echo "${{ fromJson(needs.setup.outputs.test_map)[matrix.folders] }}"
+          matrix_folders=${{ matrix.folders }}
+          matrix_folders=${matrix_folders/'models/'/'models_'}
+          echo "$matrix_folders"
+          echo "matrix_folders=$matrix_folders" >> $GITHUB_ENV
 
       - name: NVIDIA-SMI
-        continue-on-error: true
         run: |
           nvidia-smi
 
-      - name: Are GPUs recognized by our DL frameworks
+      - name: Environment
+        working-directory: /transformers
         run: |
-          utils/print_env_pt.py
+          python3 utils/print_env.py
 
-      - name: Fetch the tests to run
-        run: |
-          python utils/tests_fetcher.py --diff_with_last_commit | tee test_preparation.txt
-
-      - name: Report fetched tests
-        uses: actions/upload-artifact@v2
-        with:
-          name: test_fetched
-          path: test_preparation.txt
-
-      - name: Run all non-slow tests on GPU
+      - name: Run all non-slow selected tests on GPU
         env:
           MKL_SERVICE_FORCE_INTEL: 1
+        working-directory: /transformers
         run: |
-          if [ -f test_list.txt ]; then
-            python -m pytest -n 2 --dist=loadfile -v --make-reports=tests_torch_multi_gpu $(cat test_list.txt)
-          fi
+          python3 -m pytest -n 2 --dist=loadfile -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} ${{ fromJson(needs.setup.outputs.test_map)[matrix.folders] }}
 
       - name: Failure short reports
         if: ${{ failure() }}
-        run: cat reports/tests_torch_multi_gpu_failures_short.txt
+        continue-on-error: true
+        run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
         with:
-          name: run_all_tests_torch_multi_gpu_test_reports
-          path: reports
+          name: ${{ matrix.machine_type }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
+          path: /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}
 
-#  run_tests_flax_multi_gpu:
-#    runs-on: [self-hosted, docker-gpu, multi-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:
-#      - name: Install dependencies
-#        run: |
-#          apt -y update && apt install -y software-properties-common && apt -y update && add-apt-repository -y ppa:git-core/ppa && apt -y update && apt install -y git espeak-ng
-#          pip install --upgrade "jax[cuda111]" -f https://storage.googleapis.com/jax-releases/jax_releases.html
-#          pip install --upgrade pip
-#          pip install .[sklearn,testing,sentencepiece,flax,flax-speech,vision]
-#          pip install https://github.com/kpu/kenlm/archive/master.zip
-#
-#      - name: Launcher docker
-#        uses: actions/checkout@v2
-#        with:
-#          fetch-depth: 2
-#
-#      - name: NVIDIA-SMI
-#        continue-on-error: true
-#        run: |
-#          nvidia-smi
-#
-#      - name: Are GPUs recognized by our DL frameworks
-#        run: |
-#          python -c "from jax.lib import xla_bridge; print('GPU available:', xla_bridge.get_backend().platform)"
-#          python -c "import jax; print('Number of GPUs available:', len(jax.local_devices()))"
-#
-#      - name: Fetch the tests to run
-#        run: |
-#          python utils/tests_fetcher.py --diff_with_last_commit | tee test_preparation.txt
-#
-#      - name: Report fetched tests
-#        uses: actions/upload-artifact@v2
-#        with:
-#          name: test_fetched
-#          path: test_preparation.txt
-#
-#      - name: Run all non-slow tests on GPU
-#        run: |
-#          if [ -f test_list.txt ]; then
-#            python -m pytest -n 2 --dist=loadfile -v --make-reports=tests_flax_multi_gpu $(cat test_list.txt)
-#          fi
-#
-#      - name: Failure short reports
-#        if: ${{ failure() }}
-#        run: cat reports/tests_flax_multi_gpu_failures_short.txt
-#
-#      - name: Test suite reports artifacts
-#        if: ${{ always() }}
-#        uses: actions/upload-artifact@v2
-#        with:
-#          name: run_all_tests_flax_multi_gpu_test_reports
-#          path: reports
-
-#  run_tests_tf_multi_gpu:
-#    runs-on: [self-hosted, docker-gpu, multi-gpu]
-#    timeout-minutes: 120
-#    container:
-#      image: tensorflow/tensorflow:2.4.1-gpu
-#      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-#    steps:
-#      - name: Install dependencies
-#        run: |
-#          apt -y update && apt install -y software-properties-common && apt -y update && add-apt-repository -y ppa:git-core/ppa && apt -y update && apt install -y git espeak-ng
-#          pip install --upgrade pip
-#          pip install .[sklearn,testing,onnxruntime,sentencepiece,tf-speech]
-#          pip install https://github.com/kpu/kenlm/archive/master.zip
-#
-#      - name: Launcher docker
-#        uses: actions/checkout@v2
-#        with:
-#          fetch-depth: 2
-#
-#      - name: NVIDIA-SMI
-#        run: |
-#          nvidia-smi
-#
-#      - name: Are GPUs recognized by our DL frameworks
-#        run: |
-#          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: Fetch the tests to run
-#        run: |
-#          python utils/tests_fetcher.py --diff_with_last_commit | tee test_preparation.txt
-#
-#      - name: Report fetched tests
-#        uses: actions/upload-artifact@v2
-#        with:
-#          name: test_fetched
-#          path: test_preparation.txt
-#
-#      - name: Run all non-slow tests on GPU
-#        env:
-#          TF_NUM_INTRAOP_THREADS: 8
-#          TF_NUM_INTEROP_THREADS: 1
-#        run: |
-#          if [ -f test_list.txt ]; then
-#            python -m pytest -n 1 --dist=loadfile --make-reports=tests_tf_multi_gpu $(cat test_list.txt)
-#          fi
-#
-#      - name: Failure short reports
-#        if: ${{ failure() }}
-#        run: cat reports/tests_tf_multi_gpu_failures_short.txt
-#
-#      - name: Test suite reports artifacts
-#        if: ${{ always() }}
-#        uses: actions/upload-artifact@v2
-#        with:
-#          name: run_all_tests_tf_multi_gpu_test_reports
-#          path: reports
-
-  run_tests_torch_cuda_extensions_gpu:
-    runs-on: [self-hosted, docker-gpu, single-gpu]
+  run_tests_torch_cuda_extensions_single_gpu:
+    name: Torch CUDA extension tests
+    needs: setup
+    if: contains(fromJson(needs.setup.outputs.matrix), 'deepspeed') || contains(fromJson(needs.setup.outputs.matrix), 'extended')
+    strategy:
+      fail-fast: false
+      matrix:
+        machine_type: [single-gpu]
+    runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
     container:
-      image: nvcr.io/nvidia/pytorch:21.03-py3
+      image: huggingface/transformers-pytorch-deepspeed-latest-gpu
       options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - name: Launcher docker
-        uses: actions/checkout@v2
-        with:
-          fetch-depth: 2
+      # Necessary to get the correct branch name and commit SHA for `workflow_run` event
+      # We also take into account the `push` event (we might want to test some changes in a branch)
+      - name: Prepare custom environment variables
+        shell: bash
+        # For the meaning of these environment variables, see the job `Setup`
+        run: |
+          CI_BRANCH_PUSH=${{ github.event.ref }}
+          CI_BRANCH_PUSH=${CI_BRANCH_PUSH/'refs/heads/'/''}
+          CI_BRANCH_WORKFLOW_RUN=${{ github.event.workflow_run.head_branch }}
+          CI_SHA_PUSH=${{ github.event.head_commit.id }}
+          CI_SHA_WORKFLOW_RUN=${{ github.event.workflow_run.head_sha }}
+          echo $CI_BRANCH_PUSH
+          echo $CI_BRANCH_WORKFLOW_RUN
+          echo $CI_SHA_PUSH
+          echo $CI_SHA_WORKFLOW_RUN
+          [[ ! -z "$CI_BRANCH_PUSH" ]] && echo "CI_BRANCH=$CI_BRANCH_PUSH" >> $GITHUB_ENV || echo "CI_BRANCH=$CI_BRANCH_WORKFLOW_RUN" >> $GITHUB_ENV
+          [[ ! -z "$CI_SHA_PUSH" ]] && echo "CI_SHA=$CI_SHA_PUSH" >> $GITHUB_ENV || echo "CI_SHA=$CI_SHA_WORKFLOW_RUN" >> $GITHUB_ENV
+
+      - name: print environment variables
+        run: |
+          echo "env.CI_BRANCH = ${{ env.CI_BRANCH }}"
+          echo "env.CI_SHA = ${{ env.CI_SHA }}"
+
+      - name: Update clone using environment variables
+        working-directory: /workspace/transformers
+        run: |
+          echo "original branch = $(git branch --show-current)"
+          git fetch && git checkout ${{ env.CI_BRANCH }}
+          echo "updated branch = $(git branch --show-current)"
+          git checkout ${{ env.CI_SHA }}
+          echo "log = $(git log -n 1)"
+
+      # To avoid unknown test failures
+      - name: Pre build DeepSpeed *again*
+        working-directory: /workspace
+        run: |
+          python3 -m pip uninstall -y deepspeed
+          DS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 python3 -m pip install deepspeed --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check
 
       - name: NVIDIA-SMI
         run: |
           nvidia-smi
 
-      - name: Install dependencies
+      - name: Environment
+        working-directory: /workspace/transformers
         run: |
-          apt -y update && apt install -y libaio-dev
-          pip install --upgrade pip
-          pip install .[testing,deepspeed]
+          python utils/print_env.py
 
-      - name: Are GPUs recognized by our DL frameworks
+      - name: Run all non-slow selected tests on GPU
+        working-directory: /workspace/transformers
+        # TODO: Here we pass all tests in the 2 folders for simplicity. It's better to pass only the identified tests.
         run: |
-          utils/print_env_pt.py
-
-      - name: Fetch the tests to run
-        run: |
-          python utils/tests_fetcher.py --diff_with_last_commit --filters tests/deepspeed tests/extended | tee test_preparation.txt
-
-      - name: Report fetched tests
-        uses: actions/upload-artifact@v2
-        with:
-          name: test_fetched
-          path: test_preparation.txt
-
-      - name: Run all tests on GPU
-        run: |
-          if [ -f test_list.txt ]; then
-            python -m pytest -n 1 --dist=loadfile -v --make-reports=tests_torch_cuda_extensions_gpu $(cat test_list.txt)
-          fi
+          python -m pytest -n 1 --dist=loadfile -v --make-reports=${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu tests/deepspeed tests/extended
 
       - name: Failure short reports
         if: ${{ failure() }}
-        run: cat reports/tests_torch_cuda_extensions_gpu_failures_short.txt
+        continue-on-error: true
+        run: cat /workspace/transformers/reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu/failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
         with:
-          name: run_tests_torch_cuda_extensions_gpu_test_reports
-          path: reports
+          name: ${{ matrix.machine_type }}_run_tests_torch_cuda_extensions_gpu_test_reports
+          path: /workspace/transformers/reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu
 
   run_tests_torch_cuda_extensions_multi_gpu:
-    runs-on: [self-hosted, docker-gpu, multi-gpu]
+    name: Torch CUDA extension tests
+    needs: setup
+    if: contains(fromJson(needs.setup.outputs.matrix), 'deepspeed') || contains(fromJson(needs.setup.outputs.matrix), 'extended')
+    strategy:
+      fail-fast: false
+      matrix:
+        machine_type: [multi-gpu]
+    runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
     container:
-      image: nvcr.io/nvidia/pytorch:21.03-py3
-      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+      image: huggingface/transformers-pytorch-deepspeed-latest-gpu
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     steps:
-      - name: Launcher docker
-        uses: actions/checkout@v2
-        with:
-          fetch-depth: 2
+      # Necessary to get the correct branch name and commit SHA for `workflow_run` event
+      # We also take into account the `push` event (we might want to test some changes in a branch)
+      - name: Prepare custom environment variables
+        shell: bash
+        # For the meaning of these environment variables, see the job `Setup`
+        run: |
+          CI_BRANCH_PUSH=${{ github.event.ref }}
+          CI_BRANCH_PUSH=${CI_BRANCH_PUSH/'refs/heads/'/''}
+          CI_BRANCH_WORKFLOW_RUN=${{ github.event.workflow_run.head_branch }}
+          CI_SHA_PUSH=${{ github.event.head_commit.id }}
+          CI_SHA_WORKFLOW_RUN=${{ github.event.workflow_run.head_sha }}
+          echo $CI_BRANCH_PUSH
+          echo $CI_BRANCH_WORKFLOW_RUN
+          echo $CI_SHA_PUSH
+          echo $CI_SHA_WORKFLOW_RUN
+          [[ ! -z "$CI_BRANCH_PUSH" ]] && echo "CI_BRANCH=$CI_BRANCH_PUSH" >> $GITHUB_ENV || echo "CI_BRANCH=$CI_BRANCH_WORKFLOW_RUN" >> $GITHUB_ENV
+          [[ ! -z "$CI_SHA_PUSH" ]] && echo "CI_SHA=$CI_SHA_PUSH" >> $GITHUB_ENV || echo "CI_SHA=$CI_SHA_WORKFLOW_RUN" >> $GITHUB_ENV
+
+      - name: print environment variables
+        run: |
+          echo "env.CI_BRANCH = ${{ env.CI_BRANCH }}"
+          echo "env.CI_SHA = ${{ env.CI_SHA }}"
+
+      - name: Update clone using environment variables
+        working-directory: /workspace/transformers
+        run: |
+          echo "original branch = $(git branch --show-current)"
+          git fetch && git checkout ${{ env.CI_BRANCH }}
+          echo "updated branch = $(git branch --show-current)"
+          git checkout ${{ env.CI_SHA }}
+          echo "log = $(git log -n 1)"
+
+      # To avoid unknown test failures
+      - name: Pre build DeepSpeed *again*
+        working-directory: /workspace
+        run: |
+          python3 -m pip uninstall -y deepspeed
+          DS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 python3 -m pip install deepspeed --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check
 
       - name: NVIDIA-SMI
-        continue-on-error: true
         run: |
           nvidia-smi
 
-      - name: Install dependencies
+      - name: Environment
+        working-directory: /workspace/transformers
         run: |
-          apt -y update && apt install -y libaio-dev
-          pip install --upgrade pip
-          rm -rf ~/.cache/torch_extensions/ # shared between conflicting builds
-          pip install .[testing,deepspeed,fairscale]
+          python utils/print_env.py
 
-      - name: Are GPUs recognized by our DL frameworks
+      - name: Run all non-slow selected tests on GPU
+        working-directory: /workspace/transformers
+        # TODO: Here we pass all tests in the 2 folders for simplicity. It's better to pass only the identified tests.
         run: |
-          utils/print_env_pt.py
-
-      - name: Fetch the tests to run
-        run: |
-          python utils/tests_fetcher.py --diff_with_last_commit --filters tests/deepspeed tests/extended | tee test_preparation.txt
-
-      - name: Report fetched tests
-        uses: actions/upload-artifact@v2
-        with:
-          name: test_fetched
-          path: test_preparation.txt
-
-      - name: Run all tests on GPU
-        run: |
-          if [ -f test_list.txt ]; then
-            python -m pytest -n 1 --dist=loadfile -v --make-reports=tests_torch_cuda_extensions_multi_gpu $(cat test_list.txt)
-          fi
+          python -m pytest -n 1 --dist=loadfile -v --make-reports=${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu tests/deepspeed tests/extended
 
       - name: Failure short reports
         if: ${{ failure() }}
-        run: cat reports/tests_torch_cuda_extensions_multi_gpu_failures_short.txt
+        continue-on-error: true
+        run: cat /workspace/transformers/reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu/failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
         with:
-          name: run_tests_torch_cuda_extensions_multi_gpu_test_reports
-          path: reports
-
+          name: ${{ matrix.machine_type }}_run_tests_torch_cuda_extensions_gpu_test_reports
+          path: /workspace/transformers/reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu
 
   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,
-        run_tests_torch_cuda_extensions_gpu,
+        setup,
+        run_tests_single_gpu,
+        run_tests_multi_gpu,
+        run_tests_torch_cuda_extensions_single_gpu,
         run_tests_torch_cuda_extensions_multi_gpu
     ]
     steps:
+      # Necessary to get the correct branch name and commit SHA for `workflow_run` event
+      # We also take into account the `push` event (we might want to test some changes in a branch)
+      - name: Prepare custom environment variables
+        shell: bash
+        # For the meaning of these environment variables, see the job `Setup`
+        run: |
+          CI_BRANCH_PUSH=${{ github.event.ref }}
+          CI_BRANCH_PUSH=${CI_BRANCH_PUSH/'refs/heads/'/''}
+          CI_BRANCH_WORKFLOW_RUN=${{ github.event.workflow_run.head_branch }}
+          CI_SHA_PUSH=${{ github.event.head_commit.id }}
+          CI_SHA_WORKFLOW_RUN=${{ github.event.workflow_run.head_sha }}
+          echo $CI_BRANCH_PUSH
+          echo $CI_BRANCH_WORKFLOW_RUN
+          echo $CI_SHA_PUSH
+          echo $CI_SHA_WORKFLOW_RUN
+          [[ ! -z "$CI_BRANCH_PUSH" ]] && echo "CI_BRANCH=$CI_BRANCH_PUSH" >> $GITHUB_ENV || echo "CI_BRANCH=$CI_BRANCH_WORKFLOW_RUN" >> $GITHUB_ENV
+          [[ ! -z "$CI_SHA_PUSH" ]] && echo "CI_SHA=$CI_SHA_PUSH" >> $GITHUB_ENV || echo "CI_SHA=$CI_SHA_WORKFLOW_RUN" >> $GITHUB_ENV
+
+      - name: print environment variables
+        run: |
+          echo "env.CI_BRANCH = ${{ env.CI_BRANCH }}"
+          echo "env.CI_SHA = ${{ env.CI_SHA }}"
+
       - uses: actions/checkout@v2
 
-      - uses: actions/download-artifact@v2
+      - name: Update clone using environment variables
+        run: |
+          echo "original branch = $(git branch --show-current)"
+          git fetch && git checkout ${{ env.CI_BRANCH }}
+          echo "updated branch = $(git branch --show-current)"
+          git checkout ${{ env.CI_SHA }}
+          echo "log = $(git log -n 1)"
 
+      - 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 }}
-
+          CI_SLACK_CHANNEL_ID_DAILY: ${{ secrets.CI_SLACK_CHANNEL_ID_DAILY }}
+          CI_SLACK_CHANNEL_DUMMY_TESTS: ${{ secrets.CI_SLACK_CHANNEL_DUMMY_TESTS }}
+          CI_SLACK_REPORT_CHANNEL_ID: ${{ secrets.CI_SLACK_CHANNEL_ID }}
+          CI_EVENT: push
+          CI_TITLE_PUSH: ${{ github.event.head_commit.message }}
+          CI_TITLE_WORKFLOW_RUN: ${{ github.event.workflow_run.head_commit.message }}
+          CI_SHA: ${{ env.CI_SHA }}
+        # We pass `needs.setup.outputs.matrix` as the argument. A processing in `notification_service.py` to change
+        # `models/bert` to `models_bert` is required, as the artifact names use `_` instead of `/`.
         run: |
           pip install slack_sdk
-          python utils/notification_service_deprecated.py push
+          python utils/notification_service.py "${{ needs.setup.outputs.matrix }}"

.github/workflows/self-scheduled.yml

@@ -26,8 +26,8 @@ jobs:
     name: Setup
     strategy:
       matrix:
-        machines: [multi-gpu-docker, single-gpu-docker]
-    runs-on: ${{ matrix.machines }}
+        machine_type: [single-gpu, multi-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
     container:
       image: huggingface/transformers-all-latest-gpu
       options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
@@ -43,60 +43,124 @@ jobs:
         working-directory: /transformers
         run: |
           rm -rf tests/__pycache__
+          rm -rf tests/models/__pycache__
           rm -rf reports
 
       - id: set-matrix
         name: Identify models to test
         working-directory: /transformers/tests
         run: |
-          echo "::set-output name=matrix::$(python3 -c 'import os; x = list(filter(os.path.isdir, os.listdir(os.getcwd()))); x.sort(); print(x)')"
+          echo "::set-output name=matrix::$(python3 -c 'import os; tests = os.getcwd(); model_tests = os.listdir(os.path.join(tests, "models")); d1 = sorted(list(filter(os.path.isdir, os.listdir(tests)))); d2 = sorted(list(filter(os.path.isdir, [f"models/{x}" for x in model_tests]))); d1.remove("models"); d = d2 + d1; print(d)')"
 
       - name: NVIDIA-SMI
         run: |
           nvidia-smi
 
-      - name: GPU visibility
-        working-directory: /transformers
-        run: |
-          utils/print_env_pt.py
-          TF_CPP_MIN_LOG_LEVEL=3 python3 -c "import tensorflow as tf; print('TF GPUs available:', bool(tf.config.list_physical_devices('GPU')))"
-          TF_CPP_MIN_LOG_LEVEL=3 python3 -c "import tensorflow as tf; print('Number of TF GPUs available:', len(tf.config.list_physical_devices('GPU')))"
-
-  run_tests_gpu:
+  run_tests_single_gpu:
     name: Model tests
     strategy:
       fail-fast: false
       matrix:
         folders: ${{ fromJson(needs.setup.outputs.matrix) }}
-        machines: [multi-gpu-docker, single-gpu-docker]
-    runs-on: ${{ matrix.machines }}
+        machine_type: [single-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
     container:
       image: huggingface/transformers-all-latest-gpu
       options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     needs: setup
     steps:
       - name: Echo folder ${{ matrix.folders }}
-        run: echo "${{ matrix.folders }}"
+        shell: bash
+        # For folders like `models/bert`, set an env. var. (`matrix_folders`) to `models_bert`, which will be used to
+        # set the artifact folder names (because the character `/` is not allowed).
+        run: |
+          echo "${{ matrix.folders }}"
+          matrix_folders=${{ matrix.folders }}
+          matrix_folders=${matrix_folders/'models/'/'models_'}
+          echo "$matrix_folders"
+          echo "matrix_folders=$matrix_folders" >> $GITHUB_ENV
 
       - name: Update clone
         working-directory: /transformers
         run: git fetch && git checkout ${{ github.sha }}
 
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
+
+      - name: Environment
+        working-directory: /transformers
+        run: |
+          python3 utils/print_env.py
+
       - name: Run all tests on GPU
         working-directory: /transformers
-        run: python3 -m pytest -v --make-reports=${{ matrix.machines }}_tests_gpu_${{ matrix.folders }} tests/${{ matrix.folders }}
+        run: python3 -m pytest -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} tests/${{ matrix.folders }}
 
       - name: Failure short reports
         if: ${{ failure() }}
         continue-on-error: true
-        run: cat /transformers/reports/${{ matrix.machines }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
+        run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
         with:
-          name: ${{ matrix.machines }}_run_all_tests_gpu_${{ matrix.folders }}_test_reports
-          path: /transformers/reports/${{ matrix.machines }}_tests_gpu_${{ matrix.folders }}
+          name: ${{ matrix.machine_type }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
+          path: /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}
+
+  run_tests_multi_gpu:
+    name: Model tests
+    strategy:
+      fail-fast: false
+      matrix:
+        folders: ${{ fromJson(needs.setup.outputs.matrix) }}
+        machine_type: [multi-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
+    container:
+      image: huggingface/transformers-all-latest-gpu
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+    needs: setup
+    steps:
+      - name: Echo folder ${{ matrix.folders }}
+        shell: bash
+        # For folders like `models/bert`, set an env. var. (`matrix_folders`) to `models_bert`, which will be used to
+        # set the artifact folder names (because the character `/` is not allowed).
+        run: |
+          echo "${{ matrix.folders }}"
+          matrix_folders=${{ matrix.folders }}
+          matrix_folders=${matrix_folders/'models/'/'models_'}
+          echo "$matrix_folders"
+          echo "matrix_folders=$matrix_folders" >> $GITHUB_ENV
+
+      - name: Update clone
+        working-directory: /transformers
+        run: git fetch && git checkout ${{ github.sha }}
+
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
+
+      - name: Environment
+        working-directory: /transformers
+        run: |
+          python3 utils/print_env.py
+
+      - name: Run all tests on GPU
+        working-directory: /transformers
+        run: python3 -m pytest -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} tests/${{ matrix.folders }}
+
+      - name: Failure short reports
+        if: ${{ failure() }}
+        continue-on-error: true
+        run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
+
+      - name: Test suite reports artifacts
+        if: ${{ always() }}
+        uses: actions/upload-artifact@v2
+        with:
+          name: ${{ matrix.machine_type }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
+          path: /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}
 
   run_examples_gpu:
     name: Examples directory
@@ -110,6 +174,15 @@ jobs:
         working-directory: /transformers
         run: git fetch && git checkout ${{ github.sha }}
 
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
+
+      - name: Environment
+        working-directory: /transformers
+        run: |
+          python3 utils/print_env.py
+
       - name: Run examples tests on GPU
         working-directory: /transformers
         run: |
@@ -133,46 +206,55 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        machines: [multi-gpu-docker, single-gpu-docker]
-    runs-on: ${{ matrix.machines }}
+        machine_type: [single-gpu, multi-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
     container:
       image: huggingface/transformers-pytorch-gpu
-      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     needs: setup
     steps:
       - name: Update clone
         working-directory: /transformers
         run: git fetch && git checkout ${{ github.sha }}
 
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
+
+      - name: Environment
+        working-directory: /transformers
+        run: |
+          python3 utils/print_env.py
+
       - name: Run all pipeline tests on GPU
         working-directory: /transformers
         env:
           RUN_PIPELINE_TESTS: yes
         run: |
-          python3 -m pytest -n 1 -v --dist=loadfile -m is_pipeline_test --make-reports=${{ matrix.machines }}_tests_torch_pipeline_gpu tests
+          python3 -m pytest -n 1 -v --dist=loadfile -m is_pipeline_test --make-reports=${{ matrix.machine_type }}_tests_torch_pipeline_gpu tests
 
       - name: Failure short reports
         if: ${{ failure() }}
         continue-on-error: true
-        run: cat /transformers/reports/${{ matrix.machines }}_tests_torch_pipeline_gpu/failures_short.txt
+        run: cat /transformers/reports/${{ matrix.machine_type }}_tests_torch_pipeline_gpu/failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
         with:
-          name: ${{ matrix.machines }}_run_tests_torch_pipeline_gpu
-          path: /transformers/reports/${{ matrix.machines }}_tests_torch_pipeline_gpu
+          name: ${{ matrix.machine_type }}_run_tests_torch_pipeline_gpu
+          path: /transformers/reports/${{ matrix.machine_type }}_tests_torch_pipeline_gpu
 
   run_pipelines_tf_gpu:
     name: TensorFlow pipelines
     strategy:
       fail-fast: false
       matrix:
-        machines: [multi-gpu-docker, single-gpu-docker]
-    runs-on: ${{ matrix.machines }}
+        machine_type: [single-gpu, multi-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
     container:
       image: huggingface/transformers-tensorflow-gpu
-      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+      options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     needs: setup
     steps:
       - name: Update clone
@@ -180,32 +262,41 @@ jobs:
         run: |
           git fetch && git checkout ${{ github.sha }}
 
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
+
+      - name: Environment
+        working-directory: /transformers
+        run: |
+          python3 utils/print_env.py
+
       - name: Run all pipeline tests on GPU
         working-directory: /transformers
         env:
           RUN_PIPELINE_TESTS: yes
         run: |
-          python3 -m pytest -n 1 -v --dist=loadfile -m is_pipeline_test --make-reports=${{ matrix.machines }}_tests_tf_pipeline_gpu tests
+          python3 -m pytest -n 1 -v --dist=loadfile -m is_pipeline_test --make-reports=${{ matrix.machine_type }}_tests_tf_pipeline_gpu tests
 
       - name: Failure short reports
         if: ${{ always() }}
         run: |
-          cat /transformers/reports/${{ matrix.machines }}_tests_tf_pipeline_gpu/failures_short.txt
+          cat /transformers/reports/${{ matrix.machine_type }}_tests_tf_pipeline_gpu/failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
         with:
-          name: ${{ matrix.machines }}_run_tests_tf_pipeline_gpu
-          path: /transformers/reports/${{ matrix.machines }}_tests_tf_pipeline_gpu
+          name: ${{ matrix.machine_type }}_run_tests_tf_pipeline_gpu
+          path: /transformers/reports/${{ matrix.machine_type }}_tests_tf_pipeline_gpu
 
   run_all_tests_torch_cuda_extensions_gpu:
     name: Torch CUDA extension tests
     strategy:
       fail-fast: false
       matrix:
-        machines: [multi-gpu-docker, single-gpu-docker]
-    runs-on: ${{ matrix.machines }}
+        machine_type: [single-gpu, multi-gpu]
+    runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
     needs: setup
     container:
       image: huggingface/transformers-pytorch-deepspeed-latest-gpu
@@ -215,37 +306,44 @@ jobs:
         working-directory: /workspace/transformers
         run: git fetch && git checkout ${{ github.sha }}
 
-      - name: Re-compile DeepSpeed
+      # To avoid unknown test failures
+      - name: Pre build DeepSpeed *again*
         working-directory: /workspace
         run: |
-          pip install deepspeed # installs the deps correctly
-          rm -rf DeepSpeed
-          git clone https://github.com/microsoft/DeepSpeed && cd DeepSpeed && rm -rf build
-          DS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 python3 -m pip install -e . --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check
+          python3 -m pip uninstall -y deepspeed
+          DS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 python3 -m pip install deepspeed --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check
+
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
+
+      - name: Environment
+        working-directory: /workspace/transformers
+        run: |
+          python utils/print_env.py
 
       - name: Run all tests on GPU
         working-directory: /workspace/transformers
         run: |
-          python -m pytest -v --make-reports=${{ matrix.machines }}_tests_torch_cuda_extensions_gpu tests/deepspeed tests/extended
+          python -m pytest -v --make-reports=${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu tests/deepspeed tests/extended
 
       - name: Failure short reports
         if: ${{ failure() }}
         continue-on-error: true
-        run: cat /workspace/transformers/reports/${{ matrix.machines }}_tests_torch_cuda_extensions_gpu/failures_short.txt
+        run: cat /workspace/transformers/reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu/failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
         with:
-          name: ${{ matrix.machines }}_run_tests_torch_cuda_extensions_gpu_test_reports
-          path: /workspace/transformers/reports/${{ matrix.machines }}_tests_torch_cuda_extensions_gpu
-
+          name: ${{ matrix.machine_type }}_run_tests_torch_cuda_extensions_gpu_test_reports
+          path: /workspace/transformers/reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu
 
   send_results:
     name: Send results to webhook
     runs-on: ubuntu-latest
     if: always()
-    needs: [setup, run_tests_gpu, run_examples_gpu, run_pipelines_tf_gpu, run_pipelines_torch_gpu, run_all_tests_torch_cuda_extensions_gpu]
+    needs: [setup, run_tests_single_gpu, run_tests_multi_gpu, run_examples_gpu, run_pipelines_tf_gpu, run_pipelines_torch_gpu, run_all_tests_torch_cuda_extensions_gpu]
     steps:
       - uses: actions/checkout@v2
       - uses: actions/download-artifact@v2
@@ -255,6 +353,10 @@ jobs:
           CI_SLACK_CHANNEL_ID: ${{ secrets.CI_SLACK_CHANNEL_ID }}
           CI_SLACK_CHANNEL_ID_DAILY: ${{ secrets.CI_SLACK_CHANNEL_ID_DAILY }}
           CI_SLACK_CHANNEL_DUMMY_TESTS: ${{ secrets.CI_SLACK_CHANNEL_DUMMY_TESTS }}
+          CI_SLACK_REPORT_CHANNEL_ID: ${{ secrets.CI_SLACK_CHANNEL_ID_DAILY }}
+          CI_EVENT: scheduled
+        # We pass `needs.setup.outputs.matrix` as the argument. A processing in `notification_service.py` to change
+        # `models/bert` to `models_bert` is required, as the artifact names use `_` instead of `/`.
         run: |
           pip install slack_sdk
           python utils/notification_service.py "${{ needs.setup.outputs.matrix }}"

.github/workflows/update_metdata.yml

@@ -16,17 +16,25 @@ jobs:
     steps:
       - uses: actions/checkout@v2
 
-      - name: Loading cache.
+      - name: Load cached virtual environment
         uses: actions/cache@v2
         id: cache
         with:
-          path: ~/.cache/pip
-          key: v1-metadata-${{ hashFiles('setup.py') }}
+          path: ~/venv/
+          key: v3-metadata-${{ hashFiles('setup.py') }}
+
+      - name: Create virtual environment on cache miss
+        if: steps.cache.outputs.cache-hit != 'true'
+        run: |
+          python -m venv ~/venv && . ~/venv/bin/activate
+          pip install --upgrade pip
 
       - name: Setup environment
         run: |
+          . ~/venv/bin/activate
           pip install git+https://github.com/huggingface/transformers#egg=transformers[dev]
 
       - name: Update metadata
         run: |
+          . ~/venv/bin/activate
           python utils/update_metadata.py --token ${{ secrets.SYLVAIN_HF_TOKEN }} --commit_sha ${{ github.sha }}

CONTRIBUTING.md

@@ -128,7 +128,7 @@ You will need basic `git` proficiency to be able to contribute to
 manual. Type `git --help` in a shell and enjoy. If you prefer books, [Pro
 Git](https://git-scm.com/book/en/v2) is a very good reference.
 
-Follow these steps to start contributing:
+Follow these steps to start contributing ([supported Python versions](https://github.com/huggingface/transformers/blob/main/setup.py#L426)):
 
 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
@@ -368,8 +368,7 @@ For documentation strings, 🤗 Transformers follows the [google style](https://
 Check our [documentation writing guide](https://github.com/huggingface/transformers/tree/main/docs#writing-documentation---specification)
 for more information.
 
-#### This guide was heavily inspired by the awesome [scikit-learn guide to contributing](https://github.com/scikit-learn/scikit-learn/blob/main/CONTRIBUTING.md)
-
+**This guide was heavily inspired by the awesome [scikit-learn guide to contributing](https://github.com/scikit-learn/scikit-learn/blob/main/CONTRIBUTING.md).**
 
 ### Develop on Windows
 

Makefile

@@ -9,7 +9,7 @@ modified_only_fixup:
 	$(eval modified_py_files := $(shell python utils/get_modified_files.py $(check_dirs)))
 	@if test -n "$(modified_py_files)"; then \
 		echo "Checking/fixing $(modified_py_files)"; \
-		black $(modified_py_files); \
+		black --preview $(modified_py_files); \
 		isort $(modified_py_files); \
 		flake8 $(modified_py_files); \
 	else \
@@ -39,27 +39,32 @@ repo-consistency:
 	python utils/check_dummies.py
 	python utils/check_repo.py
 	python utils/check_inits.py
+	python utils/check_config_docstrings.py
 	python utils/tests_fetcher.py --sanity_check
 
 # this target runs checks on all files
 
 quality:
-	black --check $(check_dirs)
+	black --check --preview $(check_dirs)
 	isort --check-only $(check_dirs)
 	python utils/custom_init_isort.py --check_only
+	python utils/sort_auto_mappings.py --check_only
 	flake8 $(check_dirs)
 	doc-builder style src/transformers docs/source --max_len 119 --check_only --path_to_docs docs/source
+	python utils/check_doc_toc.py
 
 # Format source code automatically and check is there are any problems left that need manual fixing
 
 extra_style_checks:
 	python utils/custom_init_isort.py
+	python utils/sort_auto_mappings.py
 	doc-builder style src/transformers docs/source --max_len 119 --path_to_docs docs/source
+	python utils/check_doc_toc.py --fix_and_overwrite
 
 # this target runs checks on all files and potentially modifies some of them
 
 style:
-	black $(check_dirs)
+	black --preview $(check_dirs)
 	isort $(check_dirs)
 	${MAKE} autogenerate_code
 	${MAKE} extra_style_checks

README.md

@@ -116,22 +116,46 @@ To immediately use a model on a given input (text, image, audio, ...), we provid
 
 The second line of code downloads and caches the pretrained model used by the pipeline, while the third evaluates it on the given text. Here the answer is "positive" with a confidence of 99.97%.
 
-Many NLP tasks have a pre-trained `pipeline` ready to go. For example, we can easily extract question answers given context:
+Many tasks have a pre-trained `pipeline` ready to go, in NLP but also in computer vision and speech. For example, we can easily extract detected objects in an image:
 
 ``` python
+>>> import requests
+>>> from PIL import Image
 >>> from transformers import pipeline
 
-# Allocate a pipeline for question-answering
->>> question_answerer = pipeline('question-answering')
->>> question_answerer({
-...     'question': 'What is the name of the repository ?',
-...     'context': 'Pipeline has been included in the huggingface/transformers repository'
-... })
-{'score': 0.30970096588134766, 'start': 34, 'end': 58, 'answer': 'huggingface/transformers'}
+# Download an image with cute cats
+>>> url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/coco_sample.png"
+>>> image_data = requests.get(url, stream=True).raw
+>>> image = Image.open(image_data)
 
+# Allocate a pipeline for object detection
+>>> object_detector = pipeline('object_detection')
+>>> object_detector(image)
+[{'score': 0.9982201457023621,
+  'label': 'remote',
+  'box': {'xmin': 40, 'ymin': 70, 'xmax': 175, 'ymax': 117}},
+ {'score': 0.9960021376609802,
+  'label': 'remote',
+  'box': {'xmin': 333, 'ymin': 72, 'xmax': 368, 'ymax': 187}},
+ {'score': 0.9954745173454285,
+  'label': 'couch',
+  'box': {'xmin': 0, 'ymin': 1, 'xmax': 639, 'ymax': 473}},
+ {'score': 0.9988006353378296,
+  'label': 'cat',
+  'box': {'xmin': 13, 'ymin': 52, 'xmax': 314, 'ymax': 470}},
+ {'score': 0.9986783862113953,
+  'label': 'cat',
+  'box': {'xmin': 345, 'ymin': 23, 'xmax': 640, 'ymax': 368}}]
 ```
 
-In addition to the answer, the pretrained model used here returned its confidence score, along with the start position and end position of the answer in the tokenized sentence. You can learn more about the tasks supported by the `pipeline` API in [this tutorial](https://huggingface.co/docs/transformers/task_summary).
+Here we get a list of objects detected in the image, with a box surrounding the object and a confidence score. Here is the original image on the right, with the predictions displayed on the left:
+
+<h3 align="center">
+    <a><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/coco_sample.png" width="400"></a>
+    <a><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/coco_sample_post_processed.png" width="400"></a>
+</h3>
+
+You can learn more about the tasks supported by the `pipeline` API in [this tutorial](https://huggingface.co/docs/transformers/task_summary).
 
 To download and use any of the pretrained models on your given task, all it takes is three lines of code. Here is the PyTorch version:
 ```python
@@ -143,6 +167,7 @@ To download and use any of the pretrained models on your given task, all it take
 >>> inputs = tokenizer("Hello world!", return_tensors="pt")
 >>> outputs = model(**inputs)
 ```
+
 And here is the equivalent code for TensorFlow:
 ```python
 >>> from transformers import AutoTokenizer, TFAutoModel
@@ -234,77 +259,93 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[BARTpho](https://huggingface.co/docs/transformers/model_doc/bartpho)** (from VinAI Research) released with the paper [BARTpho: Pre-trained Sequence-to-Sequence Models for Vietnamese](https://arxiv.org/abs/2109.09701) by Nguyen Luong Tran, Duong Minh Le and Dat Quoc Nguyen.
 1. **[BEiT](https://huggingface.co/docs/transformers/model_doc/beit)** (from Microsoft) released with the paper [BEiT: BERT Pre-Training of Image Transformers](https://arxiv.org/abs/2106.08254) by Hangbo Bao, Li Dong, Furu Wei.
 1. **[BERT](https://huggingface.co/docs/transformers/model_doc/bert)** (from Google) released with the paper [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://arxiv.org/abs/1810.04805) by Jacob Devlin, Ming-Wei Chang, Kenton Lee and Kristina Toutanova.
-1. **[BERTweet](https://huggingface.co/docs/transformers/model_doc/bertweet)** (from VinAI Research) released with the paper [BERTweet: A pre-trained language model for English Tweets](https://aclanthology.org/2020.emnlp-demos.2/) by Dat Quoc Nguyen, Thanh Vu and Anh Tuan Nguyen.
 1. **[BERT For Sequence Generation](https://huggingface.co/docs/transformers/model_doc/bert-generation)** (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. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
+1. **[BERTweet](https://huggingface.co/docs/transformers/model_doc/bertweet)** (from VinAI Research) released with the paper [BERTweet: A pre-trained language model for English Tweets](https://aclanthology.org/2020.emnlp-demos.2/) by Dat Quoc Nguyen, Thanh Vu and Anh Tuan Nguyen.
 1. **[BigBird-Pegasus](https://huggingface.co/docs/transformers/model_doc/bigbird_pegasus)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
+1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (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/docs/transformers/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.
+1. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/).
 1. **[BORT](https://huggingface.co/docs/transformers/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.
 1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (from Google Research) released with the paper [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) by Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel.
 1. **[CamemBERT](https://huggingface.co/docs/transformers/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.
 1. **[CANINE](https://huggingface.co/docs/transformers/model_doc/canine)** (from Google Research) released with the paper [CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language Representation](https://arxiv.org/abs/2103.06874) by Jonathan H. Clark, Dan Garrette, Iulia Turc, John Wieting.
-1. **[ConvNeXT](https://huggingface.co/docs/transformers/main/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (from OpenAI) released with the paper [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) by Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, Ilya Sutskever.
+1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/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.
+1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
 1. **[CPM](https://huggingface.co/docs/transformers/model_doc/cpm)** (from Tsinghua University) released with the paper [CPM: A Large-scale Generative Chinese Pre-trained Language Model](https://arxiv.org/abs/2012.00413) by Zhengyan Zhang, Xu Han, Hao Zhou, Pei Ke, Yuxian Gu, Deming Ye, Yujia Qin, Yusheng Su, Haozhe Ji, Jian Guan, Fanchao Qi, Xiaozhi Wang, Yanan Zheng, Guoyang Zeng, Huanqi Cao, Shengqi Chen, Daixuan Li, Zhenbo Sun, Zhiyuan Liu, Minlie Huang, Wentao Han, Jie Tang, Juanzi Li, Xiaoyan Zhu, Maosong Sun.
 1. **[CTRL](https://huggingface.co/docs/transformers/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.
-1. **[Data2Vec](https://huggingface.co/docs/transformers/main/model_doc/data2vec)** (from Facebook) released with the paper [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) by Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli.
+1. **[CvT](https://huggingface.co/docs/transformers/model_doc/cvt)** (from Microsoft) released with the paper [CvT: Introducing Convolutions to Vision Transformers](https://arxiv.org/abs/2103.15808) by Haiping Wu, Bin Xiao, Noel Codella, Mengchen Liu, Xiyang Dai, Lu Yuan, Lei Zhang.
+1. **[Data2Vec](https://huggingface.co/docs/transformers/model_doc/data2vec)** (from Facebook) released with the paper [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) by Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli.
 1. **[DeBERTa](https://huggingface.co/docs/transformers/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.
 1. **[DeBERTa-v2](https://huggingface.co/docs/transformers/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.
 1. **[Decision Transformer](https://huggingface.co/docs/transformers/model_doc/decision_transformer)** (from Berkeley/Facebook/Google) released with the paper [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://arxiv.org/abs/2106.01345) by Lili Chen, Kevin Lu, Aravind Rajeswaran, Kimin Lee, Aditya Grover, Michael Laskin, Pieter Abbeel, Aravind Srinivas, Igor Mordatch.
-1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (from Microsoft Research) released with the paper [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) by Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei.
 1. **[DeiT](https://huggingface.co/docs/transformers/model_doc/deit)** (from Facebook) released with the paper [Training data-efficient image transformers & distillation through attention](https://arxiv.org/abs/2012.12877) by Hugo Touvron, Matthieu Cord, Matthijs Douze, Francisco Massa, Alexandre Sablayrolles, Hervé Jégou.
 1. **[DETR](https://huggingface.co/docs/transformers/model_doc/detr)** (from Facebook) released with the paper [End-to-End Object Detection with Transformers](https://arxiv.org/abs/2005.12872) by Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, Sergey Zagoruyko.
 1. **[DialoGPT](https://huggingface.co/docs/transformers/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.
 1. **[DistilBERT](https://huggingface.co/docs/transformers/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/main/examples/research_projects/distillation), RoBERTa into [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation), Multilingual BERT into [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation) and a German version of DistilBERT.
-1. **[DPR](https://huggingface.co/docs/transformers/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.
+1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (from Microsoft Research) released with the paper [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) by Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei.
+1. **[DPR](https://huggingface.co/docs/transformers/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.
 1. **[DPT](https://huggingface.co/docs/transformers/master/model_doc/dpt)** (from Intel Labs) released with the paper [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) by René Ranftl, Alexey Bochkovskiy, Vladlen Koltun.
-1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (from Google Research) 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. **[ELECTRA](https://huggingface.co/docs/transformers/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.
+1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (from Google Research) 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. **[FlauBERT](https://huggingface.co/docs/transformers/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.
+1. **[FLAVA](https://huggingface.co/docs/transformers/model_doc/flava)** (from Facebook AI) released with the paper [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) by Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach, and Douwe Kiela.
 1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (from Google Research) released with the paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/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.
-1. **[GLPN](https://huggingface.co/docs/transformers/main/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
+1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-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.
+1. **[GPT Neo](https://huggingface.co/docs/transformers/model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
+1. **[GPT NeoX](https://huggingface.co/docs/transformers/model_doc/gpt_neox)** (from EleutherAI) released with the paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) by Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach
 1. **[GPT-2](https://huggingface.co/docs/transformers/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**.
 1. **[GPT-J](https://huggingface.co/docs/transformers/model_doc/gptj)** (from EleutherAI) released in the repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
-1. **[GPT Neo](https://huggingface.co/docs/transformers/model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
+1. **[GroupViT](https://huggingface.co/docs/transformers/model_doc/groupvit)** (from UCSD, NVIDIA) released with the paper [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) by Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang.
 1. **[Hubert](https://huggingface.co/docs/transformers/model_doc/hubert)** (from Facebook) released with the paper [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://arxiv.org/abs/2106.07447) by Wei-Ning Hsu, Benjamin Bolte, Yao-Hung Hubert Tsai, Kushal Lakhotia, Ruslan Salakhutdinov, Abdelrahman Mohamed.
 1. **[I-BERT](https://huggingface.co/docs/transformers/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.
-1. **[ImageGPT](https://huggingface.co/docs/transformers/main/model_doc/imagegpt)** (from OpenAI) released with the paper [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) by Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever.
+1. **[ImageGPT](https://huggingface.co/docs/transformers/model_doc/imagegpt)** (from OpenAI) released with the paper [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) by Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever.
 1. **[LayoutLM](https://huggingface.co/docs/transformers/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.
 1. **[LayoutLMv2](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (from Microsoft Research Asia) released with the paper [LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding](https://arxiv.org/abs/2012.14740) by Yang Xu, Yiheng Xu, Tengchao Lv, Lei Cui, Furu Wei, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Wanxiang Che, Min Zhang, Lidong Zhou.
+1. **[LayoutLMv3](https://huggingface.co/docs/transformers/model_doc/layoutlmv3)** (from Microsoft Research Asia) released with the paper [LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking](https://arxiv.org/abs/2204.08387) by Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, Furu Wei.
 1. **[LayoutXLM](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (from Microsoft Research Asia) released with the paper [LayoutXLM: Multimodal Pre-training for Multilingual Visually-rich Document Understanding](https://arxiv.org/abs/2104.08836) by Yiheng Xu, Tengchao Lv, Lei Cui, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Furu Wei.
 1. **[LED](https://huggingface.co/docs/transformers/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.
+1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (from Meta AI) released with the paper [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) by Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze.
 1. **[Longformer](https://huggingface.co/docs/transformers/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.
+1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
-1. **[mLUKE](https://huggingface.co/docs/transformers/model_doc/mluke)** (from Studio Ousia) released with the paper [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) by Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka.
 1. **[LXMERT](https://huggingface.co/docs/transformers/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.
+1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (from Facebook) released with the paper [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.
 1. **[MarianMT](https://huggingface.co/docs/transformers/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.
-1. **[MaskFormer](https://huggingface.co/docs/transformers/main/model_doc/maskformer)** (from Meta and UIUC) released with the paper [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) by Bowen Cheng, Alexander G. Schwing, Alexander Kirillov.
-1. **[MBart](https://huggingface.co/docs/transformers/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.
-1. **[MBart-50](https://huggingface.co/docs/transformers/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.
+1. **[MaskFormer](https://huggingface.co/docs/transformers/model_doc/maskformer)** (from Meta and UIUC) released with the paper [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) by Bowen Cheng, Alexander G. Schwing, Alexander Kirillov.
+1. **[mBART](https://huggingface.co/docs/transformers/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.
+1. **[mBART-50](https://huggingface.co/docs/transformers/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.
 1. **[Megatron-BERT](https://huggingface.co/docs/transformers/model_doc/megatron-bert)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
 1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
+1. **[mLUKE](https://huggingface.co/docs/transformers/model_doc/mluke)** (from Studio Ousia) released with the paper [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) by Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka.
+1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (from CMU/Google Brain) released with the paper [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) by Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou.
+1. **[MobileViT](https://huggingface.co/docs/transformers/model_doc/mobilevit)** (from Apple) released with the paper [MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer](https://arxiv.org/abs/2110.02178) by Sachin Mehta and Mohammad Rastegari.
 1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
-1. **[Nyströmformer](https://huggingface.co/docs/transformers/main/model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
+1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
+1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
+1. **[NLLB](https://huggingface.co/docs/transformers/model_doc/nllb)** (from Meta) released with the paper [No Language Left Behind: Scaling Human-Centered Machine Translation](https://arxiv.org/abs/2207.04672) by the NLLB team.
+1. **[Nyströmformer](https://huggingface.co/docs/transformers/model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
+1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
+1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.
-1. **[PLBart](https://huggingface.co/docs/transformers/main/model_doc/plbart)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
-1. **[PoolFormer](https://huggingface.co/docs/transformers/main/model_doc/poolformer)** (from Sea AI Labs) released with the paper [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) by Yu, Weihao and Luo, Mi and Zhou, Pan and Si, Chenyang and Zhou, Yichen and Wang, Xinchao and Feng, Jiashi and Yan, Shuicheng.
+1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
+1. **[PoolFormer](https://huggingface.co/docs/transformers/model_doc/poolformer)** (from Sea AI Labs) released with the paper [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) by Yu, Weihao and Luo, Mi and Zhou, Pan and Si, Chenyang and Zhou, Yichen and Wang, Xinchao and Feng, Jiashi and Yan, Shuicheng.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/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.
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
+1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
 1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (from Google Research) released with the paper [REALM: Retrieval-Augmented Language Model Pre-Training](https://arxiv.org/abs/2002.08909) by Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat and Ming-Wei Chang.
 1. **[Reformer](https://huggingface.co/docs/transformers/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.
+1. **[RegNet](https://huggingface.co/docs/transformers/model_doc/regnet)** (from META Platforms) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár.
 1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (from Google Research) released with the paper [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/abs/2010.12821) by Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder.
-1. **[RegNet](https://huggingface.co/docs/transformers/main/model_doc/regnet)** (from META Platforms) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár.
-1. **[ResNet](https://huggingface.co/docs/transformers/main/model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun.
+1. **[ResNet](https://huggingface.co/docs/transformers/model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun.
 1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (from Facebook), released together with the paper [RoBERTa: 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.
 1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (from ZhuiyiTechnology), released together with the paper [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/abs/2104.09864) by Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
@@ -313,34 +354,37 @@ Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih.
 1. **[SpeechToTextTransformer](https://huggingface.co/docs/transformers/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.
 1. **[SpeechToTextTransformer2](https://huggingface.co/docs/transformers/model_doc/speech_to_text_2)** (from Facebook), released together with the paper [Large-Scale Self- and Semi-Supervised Learning for Speech Translation](https://arxiv.org/abs/2104.06678) by Changhan Wang, Anne Wu, Juan Pino, Alexei Baevski, Michael Auli, Alexis Conneau.
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (from Tel Aviv University), released together with the paper [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) by Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy.
-1. **[SqueezeBert](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (from Berkeley) 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. **[Swin Transformer](https://huggingface.co/docs/transformers/main/model_doc/swin)** (from Microsoft) released with the paper [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
+1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (from Berkeley) 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. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (from Microsoft) released with the paper [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
 1. **[T5](https://huggingface.co/docs/transformers/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.
 1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (from Google AI) released in the repository [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) 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/docs/transformers/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.
-1. **[TAPEX](https://huggingface.co/docs/transformers/main/model_doc/tapex)** (from Microsoft Research) released with the paper [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) by Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, Jian-Guang Lou.
+1. **[TAPEX](https://huggingface.co/docs/transformers/model_doc/tapex)** (from Microsoft Research) released with the paper [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) by Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, Jian-Guang Lou.
+1. **[Trajectory Transformer](https://huggingface.co/docs/transformers/model_doc/trajectory_transformers)** (from the University of California at Berkeley) released with the paper [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://arxiv.org/abs/2106.02039) by Michael Janner, Qiyang Li, Sergey Levine
 1. **[Transformer-XL](https://huggingface.co/docs/transformers/model_doc/transfo-xl)** (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. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft), released together with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
+1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (from Microsoft Research) released with the paper [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) by Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang.
-1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (from Microsoft Research) released with the paper [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER
-AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) by Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu.
-1. **[VAN](https://huggingface.co/docs/transformers/main/model_doc/van)** (from Tsinghua University and Nankai University) released with the paper [Visual Attention Network](https://arxiv.org/abs/2202.09741) by Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
-1. **[ViLT](https://huggingface.co/docs/transformers/main/model_doc/vilt)** (from NAVER AI Lab/Kakao Enterprise/Kakao Brain) released with the paper [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) by Wonjae Kim, Bokyung Son, Ildoo Kim.
+1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (from Microsoft Research) released with the paper [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) by Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu.
+1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (from Tsinghua University and Nankai University) released with the paper [Visual Attention Network](https://arxiv.org/abs/2202.09741) by Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
+1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (from NAVER AI Lab/Kakao Enterprise/Kakao Brain) released with the paper [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) by Wonjae Kim, Bokyung Son, Ildoo Kim.
 1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/model_doc/vit)** (from Google AI) released with the paper [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://arxiv.org/abs/2010.11929) by Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby.
-1. **[ViTMAE](https://huggingface.co/docs/transformers/main/model_doc/vit_mae)** (from Meta AI) released with the paper [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) by Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick.
 1. **[VisualBERT](https://huggingface.co/docs/transformers/model_doc/visual_bert)** (from UCLA NLP) released with the paper [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) by Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang.
-1. **[WavLM](https://huggingface.co/docs/transformers/main/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
+1. **[ViTMAE](https://huggingface.co/docs/transformers/model_doc/vit_mae)** (from Meta AI) released with the paper [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) by Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick.
 1. **[Wav2Vec2](https://huggingface.co/docs/transformers/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.
+1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (from Facebook AI) released 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, Sravya Popuri, Dmytro Okhonko, Juan Pino.
 1. **[Wav2Vec2Phoneme](https://huggingface.co/docs/transformers/model_doc/wav2vec2_phoneme)** (from Facebook AI) released with the paper [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) by Qiantong Xu, Alexei Baevski, Michael Auli.
+1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
 1. **[XGLM](https://huggingface.co/docs/transformers/model_doc/xglm)** (From Facebook AI) released with the paper [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) by Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li.
 1. **[XLM](https://huggingface.co/docs/transformers/model_doc/xlm)** (from Facebook) released together with the paper [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) by Guillaume Lample and Alexis Conneau.
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (from Microsoft Research) released with the paper [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) by Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
 1. **[XLM-RoBERTa](https://huggingface.co/docs/transformers/model_doc/xlm-roberta)** (from Facebook AI), released together with the paper [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) by Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer and Veselin Stoyanov.
 1. **[XLM-RoBERTa-XL](https://huggingface.co/docs/transformers/model_doc/xlm-roberta-xl)** (from Facebook AI), released together with the paper [Larger-Scale Transformers for Multilingual Masked Language Modeling](https://arxiv.org/abs/2105.00572) by Naman Goyal, Jingfei Du, Myle Ott, Giri Anantharaman, Alexis Conneau.
 1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (from Google/CMU) released with the paper [​XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) by Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
-1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
 1. **[XLS-R](https://huggingface.co/docs/transformers/model_doc/xls_r)** (from Facebook AI) released with the paper [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) by Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli.
-1. **[YOSO](https://huggingface.co/docs/transformers/main/model_doc/yoso)** (from the University of Wisconsin - Madison) released with the paper [You Only Sample (Almost) Once: Linear Cost Self-Attention Via Bernoulli Sampling](https://arxiv.org/abs/2111.09714) by Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh.
+1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
+1. **[YOLOS](https://huggingface.co/docs/transformers/model_doc/yolos)** (from Huazhong University of Science & Technology) released with the paper [You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection](https://arxiv.org/abs/2106.00666) by Yuxin Fang, Bencheng Liao, Xinggang Wang, Jiemin Fang, Jiyang Qi, Rui Wu, Jianwei Niu, Wenyu Liu.
+1. **[YOSO](https://huggingface.co/docs/transformers/model_doc/yoso)** (from the University of Wisconsin - Madison) released with the paper [You Only Sample (Almost) Once: Linear Cost Self-Attention Via Bernoulli Sampling](https://arxiv.org/abs/2111.09714) by Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh.
 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 Flax, PyTorch or TensorFlow, or has an associated tokenizer backed by the 🤗 Tokenizers library, refer to [this table](https://huggingface.co/docs/transformers/index#supported-frameworks).

README_ko.md

@@ -221,16 +221,19 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (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/docs/transformers/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.
+1. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/).
 1. **[BORT](https://huggingface.co/docs/transformers/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.
 1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (from Google Research) released with the paper [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) by Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel.
 1. **[CamemBERT](https://huggingface.co/docs/transformers/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.
 1. **[CANINE](https://huggingface.co/docs/transformers/model_doc/canine)** (from Google Research) released with the paper [CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language Representation](https://arxiv.org/abs/2103.06874) by Jonathan H. Clark, Dan Garrette, Iulia Turc, John Wieting.
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (from OpenAI) released with the paper [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) by Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, Ilya Sutskever.
+1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/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.
-1. **[ConvNeXT](https://huggingface.co/docs/transformers/main/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
+1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
 1. **[CPM](https://huggingface.co/docs/transformers/model_doc/cpm)** (from Tsinghua University) released with the paper [CPM: A Large-scale Generative Chinese Pre-trained Language Model](https://arxiv.org/abs/2012.00413) by Zhengyan Zhang, Xu Han, Hao Zhou, Pei Ke, Yuxian Gu, Deming Ye, Yujia Qin, Yusheng Su, Haozhe Ji, Jian Guan, Fanchao Qi, Xiaozhi Wang, Yanan Zheng, Guoyang Zeng, Huanqi Cao, Shengqi Chen, Daixuan Li, Zhenbo Sun, Zhiyuan Liu, Minlie Huang, Wentao Han, Jie Tang, Juanzi Li, Xiaoyan Zhu, Maosong Sun.
 1. **[CTRL](https://huggingface.co/docs/transformers/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.
-1. **[Data2Vec](https://huggingface.co/docs/transformers/main/model_doc/data2vec)** (from Facebook) released with the paper [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) by Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli.
+1. **[CvT](https://huggingface.co/docs/transformers/model_doc/cvt)** (from Microsoft) released with the paper [CvT: Introducing Convolutions to Vision Transformers](https://arxiv.org/abs/2103.15808) by Haiping Wu, Bin Xiao, Noel Codella, Mengchen Liu, Xiyang Dai, Lu Yuan, Lei Zhang.
+1. **[Data2Vec](https://huggingface.co/docs/transformers/model_doc/data2vec)** (from Facebook) released with the paper [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) by Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli.
 1. **[DeBERTa](https://huggingface.co/docs/transformers/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.
 1. **[DeBERTa-v2](https://huggingface.co/docs/transformers/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.
 1. **[Decision Transformer](https://huggingface.co/docs/transformers/model_doc/decision_transformer)** (from Berkeley/Facebook/Google) released with the paper [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://arxiv.org/abs/2106.01345) by Lili Chen, Kevin Lu, Aravind Rajeswaran, Kimin Lee, Aditya Grover, Michael Laskin, Pieter Abbeel, Aravind Srinivas, Igor Mordatch.
@@ -244,46 +247,61 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[ELECTRA](https://huggingface.co/docs/transformers/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.
 1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (from Google Research) 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. **[FlauBERT](https://huggingface.co/docs/transformers/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.
+1. **[FLAVA](https://huggingface.co/docs/transformers/model_doc/flava)** (from Facebook AI) released with the paper [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) by Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach, and Douwe Kiela.
 1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (from Google Research) released with the paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/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.
-1. **[GLPN](https://huggingface.co/docs/transformers/main/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
+1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-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.
 1. **[GPT Neo](https://huggingface.co/docs/transformers/model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
+1. **[GPT NeoX](https://huggingface.co/docs/transformers/model_doc/gpt_neox)** (from EleutherAI) released with the paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) by Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach 
 1. **[GPT-2](https://huggingface.co/docs/transformers/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**.
 1. **[GPT-J](https://huggingface.co/docs/transformers/model_doc/gptj)** (from EleutherAI) released in the repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
+1. **[GroupViT](https://huggingface.co/docs/transformers/model_doc/groupvit)** (from UCSD, NVIDIA) released with the paper [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) by Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang.
 1. **[Hubert](https://huggingface.co/docs/transformers/model_doc/hubert)** (from Facebook) released with the paper [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://arxiv.org/abs/2106.07447) by Wei-Ning Hsu, Benjamin Bolte, Yao-Hung Hubert Tsai, Kushal Lakhotia, Ruslan Salakhutdinov, Abdelrahman Mohamed.
 1. **[I-BERT](https://huggingface.co/docs/transformers/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.
-1. **[ImageGPT](https://huggingface.co/docs/transformers/main/model_doc/imagegpt)** (from OpenAI) released with the paper [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) by Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever.
+1. **[ImageGPT](https://huggingface.co/docs/transformers/model_doc/imagegpt)** (from OpenAI) released with the paper [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) by Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever.
 1. **[LayoutLM](https://huggingface.co/docs/transformers/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.
 1. **[LayoutLMv2](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (from Microsoft Research Asia) released with the paper [LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding](https://arxiv.org/abs/2012.14740) by Yang Xu, Yiheng Xu, Tengchao Lv, Lei Cui, Furu Wei, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Wanxiang Che, Min Zhang, Lidong Zhou.
+1. **[LayoutLMv3](https://huggingface.co/docs/transformers/model_doc/layoutlmv3)** (from Microsoft Research Asia) released with the paper [LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking](https://arxiv.org/abs/2204.08387) by Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, Furu Wei.
 1. **[LayoutXLM](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (from Microsoft Research Asia) released with the paper [LayoutXLM: Multimodal Pre-training for Multilingual Visually-rich Document Understanding](https://arxiv.org/abs/2104.08836) by Yiheng Xu, Tengchao Lv, Lei Cui, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Furu Wei.
 1. **[LED](https://huggingface.co/docs/transformers/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.
+1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (from Meta AI) released with the paper [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) by Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze.
 1. **[Longformer](https://huggingface.co/docs/transformers/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.
+1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
 1. **[LXMERT](https://huggingface.co/docs/transformers/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.
+1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (from Facebook) released with the paper [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.
 1. **[MarianMT](https://huggingface.co/docs/transformers/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.
-1. **[MaskFormer](https://huggingface.co/docs/transformers/main/model_doc/maskformer)** (from Meta and UIUC) released with the paper [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) by Bowen Cheng, Alexander G. Schwing, Alexander Kirillov.
-1. **[MBart](https://huggingface.co/docs/transformers/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.
-1. **[MBart-50](https://huggingface.co/docs/transformers/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.
+1. **[MaskFormer](https://huggingface.co/docs/transformers/model_doc/maskformer)** (from Meta and UIUC) released with the paper [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) by Bowen Cheng, Alexander G. Schwing, Alexander Kirillov.
+1. **[mBART](https://huggingface.co/docs/transformers/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.
+1. **[mBART-50](https://huggingface.co/docs/transformers/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.
 1. **[Megatron-BERT](https://huggingface.co/docs/transformers/model_doc/megatron-bert)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
 1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
 1. **[mLUKE](https://huggingface.co/docs/transformers/model_doc/mluke)** (from Studio Ousia) released with the paper [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) by Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka.
+1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (from CMU/Google Brain) released with the paper [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) by Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou.
+1. **[MobileViT](https://huggingface.co/docs/transformers/model_doc/mobilevit)** (from Apple) released with the paper [MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer](https://arxiv.org/abs/2110.02178) by Sachin Mehta and Mohammad Rastegari.
 1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
-1. **[Nyströmformer](https://huggingface.co/docs/transformers/main/model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
+1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
+1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
+1. **[NLLB](https://huggingface.co/docs/transformers/model_doc/nllb)** (from Meta) released with the paper [No Language Left Behind: Scaling Human-Centered Machine Translation](https://arxiv.org/abs/2207.04672) by the NLLB team.
+1. **[Nyströmformer](https://huggingface.co/docs/transformers/model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
+1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
+1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.
-1. **[PLBart](https://huggingface.co/docs/transformers/main/model_doc/plbart)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
-1. **[PoolFormer](https://huggingface.co/docs/transformers/main/model_doc/poolformer)** (from Sea AI Labs) released with the paper [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) by Yu, Weihao and Luo, Mi and Zhou, Pan and Si, Chenyang and Zhou, Yichen and Wang, Xinchao and Feng, Jiashi and Yan, Shuicheng.
+1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
+1. **[PoolFormer](https://huggingface.co/docs/transformers/model_doc/poolformer)** (from Sea AI Labs) released with the paper [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) by Yu, Weihao and Luo, Mi and Zhou, Pan and Si, Chenyang and Zhou, Yichen and Wang, Xinchao and Feng, Jiashi and Yan, Shuicheng.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/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.
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
+1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
 1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (from Google Research) released with the paper [REALM: Retrieval-Augmented Language Model Pre-Training](https://arxiv.org/abs/2002.08909) by Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat and Ming-Wei Chang.
 1. **[Reformer](https://huggingface.co/docs/transformers/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.
-1. **[RegNet](https://huggingface.co/docs/transformers/main/model_doc/regnet)** (from META Research) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár. 
+1. **[RegNet](https://huggingface.co/docs/transformers/model_doc/regnet)** (from META Research) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár. 
 1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (from Google Research) released with the paper [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/pdf/2010.12821.pdf) by Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder.
-1. **[ResNet](https://huggingface.co/docs/transformers/main/model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun. 
+1. **[ResNet](https://huggingface.co/docs/transformers/model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun. 
 1. **[RoBERTa](https://huggingface.co/docs/transformers/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.
 1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (from ZhuiyiTechnology), released together with the paper a [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/pdf/2104.09864v1.pdf) by Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
@@ -292,24 +310,27 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[SpeechToTextTransformer](https://huggingface.co/docs/transformers/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.
 1. **[SpeechToTextTransformer2](https://huggingface.co/docs/transformers/model_doc/speech_to_text_2)** (from Facebook), released together with the paper [Large-Scale Self- and Semi-Supervised Learning for Speech Translation](https://arxiv.org/abs/2104.06678) by Changhan Wang, Anne Wu, Juan Pino, Alexei Baevski, Michael Auli, Alexis Conneau.
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (from Tel Aviv University), released together with the paper [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) by Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy.
-1. **[SqueezeBert](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (from Berkeley) 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. **[Swin Transformer](https://huggingface.co/docs/transformers/main/model_doc/swin)** (from Microsoft) released with the paper [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
+1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (from Berkeley) 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. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (from Microsoft) released with the paper [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
 1. **[T5](https://huggingface.co/docs/transformers/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.
 1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (from Google AI) released in the repository [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) 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/docs/transformers/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.
-1. **[TAPEX](https://huggingface.co/docs/transformers/main/model_doc/tapex)** (from Microsoft Research) released with the paper [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) by Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, Jian-Guang Lou.
+1. **[TAPEX](https://huggingface.co/docs/transformers/model_doc/tapex)** (from Microsoft Research) released with the paper [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) by Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, Jian-Guang Lou.
+1. **[Trajectory Transformer](https://huggingface.co/docs/transformers/model_doc/trajectory_transformers)** (from the University of California at Berkeley) released with the paper [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://arxiv.org/abs/2106.02039) by Michael Janner, Qiyang Li, Sergey Levine 
 1. **[Transformer-XL](https://huggingface.co/docs/transformers/model_doc/transfo-xl)** (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. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft), released together with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
+1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler 
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (from Microsoft Research) released with the paper [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) by Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang.
 1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (from Microsoft Research) released with the paper [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) by Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu.
-1. **[VAN](https://huggingface.co/docs/transformers/main/model_doc/van)** (from Tsinghua University and Nankai University) released with the paper [Visual Attention Network](https://arxiv.org/pdf/2202.09741.pdf) by Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
-1. **[ViLT](https://huggingface.co/docs/transformers/main/model_doc/vilt)** (from NAVER AI Lab/Kakao Enterprise/Kakao Brain) released with the paper [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) by Wonjae Kim, Bokyung Son, Ildoo Kim.
+1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (from Tsinghua University and Nankai University) released with the paper [Visual Attention Network](https://arxiv.org/pdf/2202.09741.pdf) by Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
+1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (from NAVER AI Lab/Kakao Enterprise/Kakao Brain) released with the paper [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) by Wonjae Kim, Bokyung Son, Ildoo Kim.
 1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/model_doc/vit)** (from Google AI) released with the paper [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://arxiv.org/abs/2010.11929) by Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby.
 1. **[VisualBERT](https://huggingface.co/docs/transformers/model_doc/visual_bert)** (from UCLA NLP) released with the paper [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) by Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang.
-1. **[ViTMAE](https://huggingface.co/docs/transformers/main/model_doc/vit_mae)** (from Meta AI) released with the paper [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) by Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick.
+1. **[ViTMAE](https://huggingface.co/docs/transformers/model_doc/vit_mae)** (from Meta AI) released with the paper [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) by Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick.
 1. **[Wav2Vec2](https://huggingface.co/docs/transformers/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.
+1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (from Facebook AI) released 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, Sravya Popuri, Dmytro Okhonko, Juan Pino.
 1. **[Wav2Vec2Phoneme](https://huggingface.co/docs/transformers/model_doc/wav2vec2_phoneme)** (from Facebook AI) released with the paper [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) by Qiantong Xu, Alexei Baevski, Michael Auli.
-1. **[WavLM](https://huggingface.co/docs/transformers/main/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
+1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
 1. **[XGLM](https://huggingface.co/docs/transformers/model_doc/xglm)** (From Facebook AI) released with the paper [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) by Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li. 
 1. **[XLM](https://huggingface.co/docs/transformers/model_doc/xlm)** (from Facebook) released together with the paper [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) by Guillaume Lample and Alexis Conneau.
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (from Microsoft Research) released with the paper [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) by Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
@@ -318,7 +339,8 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (from Google/CMU) released with the paper [​XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) by Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
 1. **[XLS-R](https://huggingface.co/docs/transformers/model_doc/xls_r)** (from Facebook AI) released with the paper [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) by Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli.
 1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
-1. **[YOSO](https://huggingface.co/docs/transformers/main/model_doc/yoso)** (from the University of Wisconsin - Madison) released with the paper [You Only Sample (Almost) by Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh.
+1. **[YOLOS](https://huggingface.co/docs/transformers/model_doc/yolos)** (from Huazhong University of Science & Technology) released with the paper [You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection](https://arxiv.org/abs/2106.00666) by Yuxin Fang, Bencheng Liao, Xinggang Wang, Jiemin Fang, Jiyang Qi, Rui Wu, Jianwei Niu, Wenyu Liu.
+1. **[YOSO](https://huggingface.co/docs/transformers/model_doc/yoso)** (from the University of Wisconsin - Madison) released with the paper [You Only Sample (Almost) by Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh.
 1. 새로운 모델을 올리고 싶나요? 우리가 **상세한 가이드와 템플릿** 으로 새로운 모델을 올리도록 도와드릴게요. 가이드와 템플릿은 이 저장소의 [`templates`](./templates) 폴더에서 확인하실 수 있습니다. [컨트리뷰션 가이드라인](./CONTRIBUTING.md)을 꼭 확인해주시고, PR을 올리기 전에 메인테이너에게 연락하거나 이슈를 오픈해 피드백을 받으시길 바랍니다. 
 
 각 모델이 Flax, PyTorch, TensorFlow으로 구현되었는지 또는 🤗 Tokenizers 라이브러리가 지원하는 토크나이저를 사용하는지 확인하려면, [이 표](https://huggingface.co/docs/transformers/index#supported-frameworks)를 확인하세요.

README_zh-hans.md

@@ -227,7 +227,7 @@ conda install -c huggingface transformers
 
 ## 模型架构
 
-**🤗 Transformers 支持的[所有的模型检查点](https://huggingface.co/models)** 由[用户](https://huggingface.co/users)和[组织](https://huggingface.co/organizations)上传，均与 huggingface.co [model hub](https://huggingface.co) 无缝整合。
+🤗 Transformers 支持的[**所有的模型检查点**](https://huggingface.co/models)由[用户](https://huggingface.co/users)和[组织](https://huggingface.co/organizations)上传，均与 huggingface.co [model hub](https://huggingface.co) 无缝整合。
 
 目前的检查点数量： ![](https://img.shields.io/endpoint?url=https://huggingface.co/api/shields/models&color=brightgreen)
 
@@ -245,16 +245,19 @@ conda install -c huggingface transformers
 1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (来自 Google Research) 伴随论文 [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) 由 Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed 发布。
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (来自 Facebook) 伴随论文 [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) 由 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 发布。
 1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (来自 Facebook) 伴随论文 [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) 由 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 发布。
+1. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/). 
 1. **[BORT](https://huggingface.co/docs/transformers/model_doc/bort)** (来自 Alexa) 伴随论文 [Optimal Subarchitecture Extraction For BERT](https://arxiv.org/abs/2010.10499) 由 Adrian de Wynter and Daniel J. Perry 发布。
 1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (来自 Google Research) 伴随论文 [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) 由 Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel 发布。
 1. **[CamemBERT](https://huggingface.co/docs/transformers/model_doc/camembert)** (来自 Inria/Facebook/Sorbonne) 伴随论文 [CamemBERT: a Tasty French Language Model](https://arxiv.org/abs/1911.03894) 由 Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah and Benoît Sagot 发布。
 1. **[CANINE](https://huggingface.co/docs/transformers/model_doc/canine)** (来自 Google Research) 伴随论文 [CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language Representation](https://arxiv.org/abs/2103.06874) 由 Jonathan H. Clark, Dan Garrette, Iulia Turc, John Wieting 发布。
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (来自 OpenAI) 伴随论文 [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) 由 Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, Ilya Sutskever 发布。
+1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (来自 Salesforce) 伴随论文 [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) 由 Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong 发布。
 1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (来自 YituTech) 伴随论文 [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) 由 Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan 发布。
-1. **[ConvNeXT](https://huggingface.co/docs/transformers/main/model_doc/convnext)** (来自 Facebook AI) 伴随论文 [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) 由 Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie 发布。
+1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (来自 Facebook AI) 伴随论文 [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) 由 Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie 发布。
 1. **[CPM](https://huggingface.co/docs/transformers/model_doc/cpm)** (来自 Tsinghua University) 伴随论文 [CPM: A Large-scale Generative Chinese Pre-trained Language Model](https://arxiv.org/abs/2012.00413) 由 Zhengyan Zhang, Xu Han, Hao Zhou, Pei Ke, Yuxian Gu, Deming Ye, Yujia Qin, Yusheng Su, Haozhe Ji, Jian Guan, Fanchao Qi, Xiaozhi Wang, Yanan Zheng, Guoyang Zeng, Huanqi Cao, Shengqi Chen, Daixuan Li, Zhenbo Sun, Zhiyuan Liu, Minlie Huang, Wentao Han, Jie Tang, Juanzi Li, Xiaoyan Zhu, Maosong Sun 发布。
 1. **[CTRL](https://huggingface.co/docs/transformers/model_doc/ctrl)** (来自 Salesforce) 伴随论文 [CTRL: A Conditional Transformer Language Model for Controllable Generation](https://arxiv.org/abs/1909.05858) 由 Nitish Shirish Keskar*, Bryan McCann*, Lav R. Varshney, Caiming Xiong and Richard Socher 发布。
-1. **[Data2Vec](https://huggingface.co/docs/transformers/main/model_doc/data2vec)** (来自 Facebook) 伴随论文 [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) 由 Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli 发布。
+1. **[CvT](https://huggingface.co/docs/transformers/model_doc/cvt)** (来自 Microsoft) 伴随论文 [CvT: Introducing Convolutions to Vision Transformers](https://arxiv.org/abs/2103.15808) 由 Haiping Wu, Bin Xiao, Noel Codella, Mengchen Liu, Xiyang Dai, Lu Yuan, Lei Zhang 发布。
+1. **[Data2Vec](https://huggingface.co/docs/transformers/model_doc/data2vec)** (来自 Facebook) 伴随论文 [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) 由 Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli 发布。
 1. **[DeBERTa](https://huggingface.co/docs/transformers/model_doc/deberta)** (来自 Microsoft) 伴随论文 [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://arxiv.org/abs/2006.03654) 由 Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen 发布。
 1. **[DeBERTa-v2](https://huggingface.co/docs/transformers/model_doc/deberta-v2)** (来自 Microsoft) 伴随论文 [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://arxiv.org/abs/2006.03654) 由 Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen 发布。
 1. **[Decision Transformer](https://huggingface.co/docs/transformers/model_doc/decision_transformer)** (来自 Berkeley/Facebook/Google) 伴随论文 [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://arxiv.org/abs/2106.01345) 由 Lili Chen, Kevin Lu, Aravind Rajeswaran, Kimin Lee, Aditya Grover, Michael Laskin, Pieter Abbeel, Aravind Srinivas, Igor Mordatch 发布。
@@ -268,46 +271,61 @@ conda install -c huggingface transformers
 1. **[ELECTRA](https://huggingface.co/docs/transformers/model_doc/electra)** (来自 Google Research/Stanford University) 伴随论文 [ELECTRA: Pre-training text encoders as discriminators rather than generators](https://arxiv.org/abs/2003.10555) 由 Kevin Clark, Minh-Thang Luong, Quoc V. Le, Christopher D. Manning 发布。
 1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (来自 Google Research) 伴随论文 [Leveraging Pre-trained Checkpoints for Sequence Generation Tasks](https://arxiv.org/abs/1907.12461) 由 Sascha Rothe, Shashi Narayan, Aliaksei Severyn 发布。
 1. **[FlauBERT](https://huggingface.co/docs/transformers/model_doc/flaubert)** (来自 CNRS) 伴随论文 [FlauBERT: Unsupervised Language Model Pre-training for French](https://arxiv.org/abs/1912.05372) 由 Hang Le, Loïc Vial, Jibril Frej, Vincent Segonne, Maximin Coavoux, Benjamin Lecouteux, Alexandre Allauzen, Benoît Crabbé, Laurent Besacier, Didier Schwab 发布。
+1. **[FLAVA](https://huggingface.co/docs/transformers/model_doc/flava)** (来自 Facebook AI) 伴随论文 [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) 由 Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach, and Douwe Kiela 发布。
 1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (来自 Google Research) 伴随论文 [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) 由 James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon 发布。
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (来自 CMU/Google Brain) 伴随论文 [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) 由 Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le 发布。
-1. **[GLPN](https://huggingface.co/docs/transformers/main/model_doc/glpn)** (来自 KAIST) 伴随论文 [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) 由 Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim 发布。
+1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (来自 KAIST) 伴随论文 [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) 由 Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim 发布。
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (来自 OpenAI) 伴随论文 [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/) 由 Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever 发布。
 1. **[GPT Neo](https://huggingface.co/docs/transformers/model_doc/gpt_neo)** (来自 EleutherAI) 随仓库 [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) 发布。作者为 Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy 发布。
+1. **[GPT NeoX](https://huggingface.co/docs/transformers/model_doc/gpt_neox)** (from EleutherAI) released with the paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) by Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach 
 1. **[GPT-2](https://huggingface.co/docs/transformers/model_doc/gpt2)** (来自 OpenAI) 伴随论文 [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/) 由 Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** and Ilya Sutskever** 发布。
 1. **[GPT-J](https://huggingface.co/docs/transformers/model_doc/gptj)** (来自 EleutherAI) 伴随论文 [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) 由 Ben Wang and Aran Komatsuzaki 发布。
+1. **[GroupViT](https://huggingface.co/docs/transformers/model_doc/groupvit)** (来自 UCSD, NVIDIA) 伴随论文 [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) 由 Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang 发布。
 1. **[Hubert](https://huggingface.co/docs/transformers/model_doc/hubert)** (来自 Facebook) 伴随论文 [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://arxiv.org/abs/2106.07447) 由 Wei-Ning Hsu, Benjamin Bolte, Yao-Hung Hubert Tsai, Kushal Lakhotia, Ruslan Salakhutdinov, Abdelrahman Mohamed 发布。
 1. **[I-BERT](https://huggingface.co/docs/transformers/model_doc/ibert)** (来自 Berkeley) 伴随论文 [I-BERT: Integer-only BERT Quantization](https://arxiv.org/abs/2101.01321) 由 Sehoon Kim, Amir Gholami, Zhewei Yao, Michael W. Mahoney, Kurt Keutzer 发布。
-1. **[ImageGPT](https://huggingface.co/docs/transformers/main/model_doc/imagegpt)** (来自 OpenAI) 伴随论文 [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) 由 Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever 发布。
+1. **[ImageGPT](https://huggingface.co/docs/transformers/model_doc/imagegpt)** (来自 OpenAI) 伴随论文 [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) 由 Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever 发布。
 1. **[LayoutLM](https://huggingface.co/docs/transformers/model_doc/layoutlm)** (来自 Microsoft Research Asia) 伴随论文 [LayoutLM: Pre-training of Text and Layout for Document Image Understanding](https://arxiv.org/abs/1912.13318) 由 Yiheng Xu, Minghao Li, Lei Cui, Shaohan Huang, Furu Wei, Ming Zhou 发布。
 1. **[LayoutLMv2](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (来自 Microsoft Research Asia) 伴随论文 [LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding](https://arxiv.org/abs/2012.14740) 由 Yang Xu, Yiheng Xu, Tengchao Lv, Lei Cui, Furu Wei, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Wanxiang Che, Min Zhang, Lidong Zhou 发布。
+1. **[LayoutLMv3](https://huggingface.co/docs/transformers/model_doc/layoutlmv3)** (来自 Microsoft Research Asia) 伴随论文 [LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking](https://arxiv.org/abs/2204.08387) 由 Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, Furu Wei 发布。
 1. **[LayoutXLM](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (来自 Microsoft Research Asia) 伴随论文 [LayoutXLM: Multimodal Pre-training for Multilingual Visually-rich Document Understanding](https://arxiv.org/abs/2104.08836) 由 Yiheng Xu, Tengchao Lv, Lei Cui, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Furu Wei 发布。
 1. **[LED](https://huggingface.co/docs/transformers/model_doc/led)** (来自 AllenAI) 伴随论文 [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) 由 Iz Beltagy, Matthew E. Peters, Arman Cohan 发布。
+1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (来自 Meta AI) 伴随论文 [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) 由 Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze 发布。
 1. **[Longformer](https://huggingface.co/docs/transformers/model_doc/longformer)** (来自 AllenAI) 伴随论文 [Longformer: The Long-Document Transformer](https://arxiv.org/abs/2004.05150) 由 Iz Beltagy, Matthew E. Peters, Arman Cohan 发布。
+1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (来自 Google AI) released 伴随论文 [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) 由 Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang 发布。
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (来自 Studio Ousia) 伴随论文 [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) 由 Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto 发布。
 1. **[LXMERT](https://huggingface.co/docs/transformers/model_doc/lxmert)** (来自 UNC Chapel Hill) 伴随论文 [LXMERT: Learning Cross-Modality Encoder Representations from Transformers for Open-Domain Question Answering](https://arxiv.org/abs/1908.07490) 由 Hao Tan and Mohit Bansal 发布。
+1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (来自 Facebook) 伴随论文 [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) 由 Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert 发布。
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (来自 Facebook) 伴随论文 [Beyond English-Centric Multilingual Machine Translation](https://arxiv.org/abs/2010.11125) 由 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 发布。
 1. **[MarianMT](https://huggingface.co/docs/transformers/model_doc/marian)** 用 [OPUS](http://opus.nlpl.eu/) 数据训练的机器翻译模型由 Jörg Tiedemann 发布。[Marian Framework](https://marian-nmt.github.io/) 由微软翻译团队开发。
-1. **[MaskFormer](https://huggingface.co/docs/transformers/main/model_doc/maskformer)** (from Meta and UIUC) released with the paper [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) by Bowen Cheng, Alexander G. Schwing, Alexander Kirillov 
-1. **[MBart](https://huggingface.co/docs/transformers/model_doc/mbart)** (来自 Facebook) 伴随论文 [Multilingual Denoising Pre-training for Neural Machine Translation](https://arxiv.org/abs/2001.08210) 由 Yinhan Liu, Jiatao Gu, Naman Goyal, Xian Li, Sergey Edunov, Marjan Ghazvininejad, Mike Lewis, Luke Zettlemoyer 发布。
-1. **[MBart-50](https://huggingface.co/docs/transformers/model_doc/mbart)** (来自 Facebook) 伴随论文 [Multilingual Translation with Extensible Multilingual Pretraining and Finetuning](https://arxiv.org/abs/2008.00401) 由 Yuqing Tang, Chau Tran, Xian Li, Peng-Jen Chen, Naman Goyal, Vishrav Chaudhary, Jiatao Gu, Angela Fan 发布。
+1. **[MaskFormer](https://huggingface.co/docs/transformers/model_doc/maskformer)** (from Meta and UIUC) released with the paper [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) by Bowen Cheng, Alexander G. Schwing, Alexander Kirillov 
+1. **[mBART](https://huggingface.co/docs/transformers/model_doc/mbart)** (来自 Facebook) 伴随论文 [Multilingual Denoising Pre-training for Neural Machine Translation](https://arxiv.org/abs/2001.08210) 由 Yinhan Liu, Jiatao Gu, Naman Goyal, Xian Li, Sergey Edunov, Marjan Ghazvininejad, Mike Lewis, Luke Zettlemoyer 发布。
+1. **[mBART-50](https://huggingface.co/docs/transformers/model_doc/mbart)** (来自 Facebook) 伴随论文 [Multilingual Translation with Extensible Multilingual Pretraining and Finetuning](https://arxiv.org/abs/2008.00401) 由 Yuqing Tang, Chau Tran, Xian Li, Peng-Jen Chen, Naman Goyal, Vishrav Chaudhary, Jiatao Gu, Angela Fan 发布。
 1. **[Megatron-BERT](https://huggingface.co/docs/transformers/model_doc/megatron-bert)** (来自 NVIDIA) 伴随论文 [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) 由 Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro 发布。
 1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (来自 NVIDIA) 伴随论文 [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) 由 Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro 发布。
 1. **[mLUKE](https://huggingface.co/docs/transformers/model_doc/mluke)** (来自 Studio Ousia) 伴随论文 [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) 由 Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka 发布。
+1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (来自 CMU/Google Brain) 伴随论文 [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) 由 Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou 发布。
+1. **[MobileViT](https://huggingface.co/docs/transformers/model_doc/mobilevit)** (来自 Apple) 伴随论文 [MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer](https://arxiv.org/abs/2110.02178) 由 Sachin Mehta and Mohammad Rastegari 发布。
 1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (来自 Microsoft Research) 伴随论文 [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) 由 Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu 发布。
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (来自 Google AI) 伴随论文 [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) 由 Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel 发布。
-1. **[Nyströmformer](https://huggingface.co/docs/transformers/main/model_doc/nystromformer)** (来自 the University of Wisconsin - Madison) 伴随论文 [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) 由 Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh 发布。
+1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (来自 中国人民大学 AI Box) 伴随论文 [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) 由 Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen 发布。 
+1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (来自华为诺亚方舟实验室) 伴随论文 [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) 由 Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu 发布。
+1. **[NLLB](https://huggingface.co/docs/transformers/model_doc/nllb)** (来自 Meta) 伴随论文 [No Language Left Behind: Scaling Human-Centered Machine Translation](https://arxiv.org/abs/2207.04672) 由 the NLLB team 发布。
+1. **[Nyströmformer](https://huggingface.co/docs/transformers/model_doc/nystromformer)** (来自 the University of Wisconsin - Madison) 伴随论文 [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) 由 Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh 发布。
+1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (来自 Meta AI) 伴随论文 [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) 由 Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al 发布。
+1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (来自 Google AI) 伴随论文 [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) 由 Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby 发布。
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (来自 Google) 伴随论文 [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) 由 Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu 发布。
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (来自 Deepmind) 伴随论文 [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) 由 Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira 发布。
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (来自 VinAI Research) 伴随论文 [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) 由 Dat Quoc Nguyen and Anh Tuan Nguyen 发布。
-1. **[PLBart](https://huggingface.co/docs/transformers/main/model_doc/plbart)** (来自 UCLA NLP) 伴随论文 [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) 由 Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang 发布。
-1. **[PoolFormer](https://huggingface.co/docs/transformers/main/model_doc/poolformer)** (来自 Sea AI Labs) 伴随论文 [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) 由 Yu, Weihao and Luo, Mi and Zhou, Pan and Si, Chenyang and Zhou, Yichen and Wang, Xinchao and Feng, Jiashi and Yan, Shuicheng 发布。
+1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (来自 UCLA NLP) 伴随论文 [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) 由 Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang 发布。
+1. **[PoolFormer](https://huggingface.co/docs/transformers/model_doc/poolformer)** (来自 Sea AI Labs) 伴随论文 [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) 由 Yu, Weihao and Luo, Mi and Zhou, Pan and Si, Chenyang and Zhou, Yichen and Wang, Xinchao and Feng, Jiashi and Yan, Shuicheng 发布。
 1. **[ProphetNet](https://huggingface.co/docs/transformers/model_doc/prophetnet)** (来自 Microsoft Research) 伴随论文 [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) 由 Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou 发布。
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (来自 NVIDIA) 伴随论文 [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) 由 Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius 发布。
+1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (来自 Facebook) 伴随论文 [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) 由 Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela 发布。
 1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (来自 Google Research) 伴随论文 [REALM: Retrieval-Augmented Language Model Pre-Training](https://arxiv.org/abs/2002.08909) 由 Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat and Ming-Wei Chang 发布。
 1. **[Reformer](https://huggingface.co/docs/transformers/model_doc/reformer)** (来自 Google Research) 伴随论文 [Reformer: The Efficient Transformer](https://arxiv.org/abs/2001.04451) 由 Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya 发布。
-1. **[RegNet](https://huggingface.co/docs/transformers/main/model_doc/regnet)** (from META Research) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár. 
+1. **[RegNet](https://huggingface.co/docs/transformers/model_doc/regnet)** (from META Research) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár. 
 1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (来自 Google Research) 伴随论文 [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/pdf/2010.12821.pdf) 由 Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder 发布。
-1. **[ResNet](https://huggingface.co/docs/transformers/main/model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun. 
+1. **[ResNet](https://huggingface.co/docs/transformers/model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun. 
 1. **[RoBERTa](https://huggingface.co/docs/transformers/model_doc/roberta)** (来自 Facebook), 伴随论文 [Robustly Optimized BERT Pretraining Approach](https://arxiv.org/abs/1907.11692) 由 Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov 发布。
 1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (来自 ZhuiyiTechnology), 伴随论文 [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/pdf/2104.09864v1.pdf) 由 Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu 发布。
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (来自 NVIDIA) 伴随论文 [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) 由 Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo 发布。
@@ -316,24 +334,27 @@ conda install -c huggingface transformers
 1. **[SpeechToTextTransformer](https://huggingface.co/docs/transformers/model_doc/speech_to_text)** (来自 Facebook), 伴随论文 [fairseq S2T: Fast Speech-to-Text Modeling with fairseq](https://arxiv.org/abs/2010.05171) 由 Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Dmytro Okhonko, Juan Pino 发布。
 1. **[SpeechToTextTransformer2](https://huggingface.co/docs/transformers/model_doc/speech_to_text_2)** (来自 Facebook) 伴随论文 [Large-Scale Self- and Semi-Supervised Learning for Speech Translation](https://arxiv.org/abs/2104.06678) 由 Changhan Wang, Anne Wu, Juan Pino, Alexei Baevski, Michael Auli, Alexis Conneau 发布。
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (来自 Tel Aviv University) 伴随论文 [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) 由 Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy 发布。
-1. **[SqueezeBert](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (来自 Berkeley) 伴随论文 [SqueezeBERT: What can computer vision teach NLP about efficient neural networks?](https://arxiv.org/abs/2006.11316) 由 Forrest N. Iandola, Albert E. Shaw, Ravi Krishna, and Kurt W. Keutzer 发布。
-1. **[Swin Transformer](https://huggingface.co/docs/transformers/main/model_doc/swin)** (来自 Microsoft) 伴随论文 [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) 由 Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo 发布。
+1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (来自 Berkeley) 伴随论文 [SqueezeBERT: What can computer vision teach NLP about efficient neural networks?](https://arxiv.org/abs/2006.11316) 由 Forrest N. Iandola, Albert E. Shaw, Ravi Krishna, and Kurt W. Keutzer 发布。
+1. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (来自 Microsoft) 伴随论文 [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) 由 Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo 发布。
 1. **[T5](https://huggingface.co/docs/transformers/model_doc/t5)** (来自 Google AI) 伴随论文 [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/abs/1910.10683) 由 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 发布。
 1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (来自 Google AI) 伴随论文 [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) 由 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 发布。
 1. **[TAPAS](https://huggingface.co/docs/transformers/model_doc/tapas)** (来自 Google AI) 伴随论文 [TAPAS: Weakly Supervised Table Parsing via Pre-training](https://arxiv.org/abs/2004.02349) 由 Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Müller, Francesco Piccinno and Julian Martin Eisenschlos 发布。
-1. **[TAPEX](https://huggingface.co/docs/transformers/main/model_doc/tapex)** (来自 Microsoft Research) 伴随论文 [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) 由 Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, Jian-Guang Lou 发布。
+1. **[TAPEX](https://huggingface.co/docs/transformers/model_doc/tapex)** (来自 Microsoft Research) 伴随论文 [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) 由 Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, Jian-Guang Lou 发布。
+1. **[Trajectory Transformer](https://huggingface.co/docs/transformers/model_doc/trajectory_transformers)** (from the University of California at Berkeley) released with the paper [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://arxiv.org/abs/2106.02039) by Michael Janner, Qiyang Li, Sergey Levine 
 1. **[Transformer-XL](https://huggingface.co/docs/transformers/model_doc/transfo-xl)** (来自 Google/CMU) 伴随论文 [Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context](https://arxiv.org/abs/1901.02860) 由 Zihang Dai*, Zhilin Yang*, Yiming Yang, Jaime Carbonell, Quoc V. Le, Ruslan Salakhutdinov 发布。
 1. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (来自 Microsoft) 伴随论文 [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) 由 Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei 发布。
+1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler 
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (来自 Microsoft Research) 伴随论文 [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) 由 Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang 发布。
 1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (来自 Microsoft Research) 伴随论文 [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) 由 Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu 发布。
-1. **[VAN](https://huggingface.co/docs/transformers/main/model_doc/van)** (来自 Tsinghua University and Nankai University) 伴随论文 [Visual Attention Network](https://arxiv.org/pdf/2202.09741.pdf) 由 Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu 发布。
-1. **[ViLT](https://huggingface.co/docs/transformers/main/model_doc/vilt)** (来自 NAVER AI Lab/Kakao Enterprise/Kakao Brain) 伴随论文 [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) 由 Wonjae Kim, Bokyung Son, Ildoo Kim 发布。
+1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (来自 Tsinghua University and Nankai University) 伴随论文 [Visual Attention Network](https://arxiv.org/pdf/2202.09741.pdf) 由 Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu 发布。
+1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (来自 NAVER AI Lab/Kakao Enterprise/Kakao Brain) 伴随论文 [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) 由 Wonjae Kim, Bokyung Son, Ildoo Kim 发布。
 1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/model_doc/vit)** (来自 Google AI) 伴随论文 [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://arxiv.org/abs/2010.11929) 由 Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby 发布。
 1. **[VisualBERT](https://huggingface.co/docs/transformers/model_doc/visual_bert)** (来自 UCLA NLP) 伴随论文 [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) 由 Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang 发布。
-1. **[ViTMAE](https://huggingface.co/docs/transformers/main/model_doc/vit_mae)** (来自 Meta AI) 伴随论文 [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) 由 Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick 发布。
+1. **[ViTMAE](https://huggingface.co/docs/transformers/model_doc/vit_mae)** (来自 Meta AI) 伴随论文 [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) 由 Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick 发布。
 1. **[Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/wav2vec2)** (来自 Facebook AI) 伴随论文 [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://arxiv.org/abs/2006.11477) 由 Alexei Baevski, Henry Zhou, Abdelrahman Mohamed, Michael Auli 发布。
+1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (来自 Facebook AI) 伴随论文 [FAIRSEQ S2T: Fast Speech-to-Text Modeling with FAIRSEQ](https://arxiv.org/abs/2010.05171) 由 Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Sravya Popuri, Dmytro Okhonko, Juan Pino 发布。
 1. **[Wav2Vec2Phoneme](https://huggingface.co/docs/transformers/model_doc/wav2vec2_phoneme)** (来自 Facebook AI) 伴随论文 [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) 由 Qiantong Xu, Alexei Baevski, Michael Auli 发布。
-1. **[WavLM](https://huggingface.co/docs/transformers/main/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
+1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
 1. **[XGLM](https://huggingface.co/docs/transformers/model_doc/xglm)** (From Facebook AI) released with the paper [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) by Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li. 
 1. **[XLM](https://huggingface.co/docs/transformers/model_doc/xlm)** (来自 Facebook) 伴随论文 [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) 由 Guillaume Lample and Alexis Conneau 发布。
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (来自 Microsoft Research) 伴随论文 [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) 由 Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou 发布。
@@ -342,7 +363,8 @@ conda install -c huggingface transformers
 1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (来自 Google/CMU) 伴随论文 [XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) 由 Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le 发布。
 1. **[XLS-R](https://huggingface.co/docs/transformers/model_doc/xls_r)** (来自 Facebook AI) 伴随论文 [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) 由 Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli 发布。
 1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (来自 Facebook AI) 伴随论文 [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) 由 Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli 发布。
-1. **[YOSO](https://huggingface.co/docs/transformers/main/model_doc/yoso)** (来自 the University of Wisconsin - Madison) 伴随论文 [You Only Sample (Almost) 由 Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh 发布。
+1. **[YOLOS](https://huggingface.co/docs/transformers/model_doc/yolos)** (来自 Huazhong University of Science & Technology) 伴随论文 [You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection](https://arxiv.org/abs/2106.00666) 由 Yuxin Fang, Bencheng Liao, Xinggang Wang, Jiemin Fang, Jiyang Qi, Rui Wu, Jianwei Niu, Wenyu Liu 发布。
+1. **[YOSO](https://huggingface.co/docs/transformers/model_doc/yoso)** (来自 the University of Wisconsin - Madison) 伴随论文 [You Only Sample (Almost) 由 Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh 发布。
 1. 想要贡献新的模型？我们这里有一份**详细指引和模板**来引导你添加新的模型。你可以在 [`templates`](./templates) 目录中找到他们。记得查看 [贡献指南](./CONTRIBUTING.md) 并在开始写 PR 前联系维护人员或开一个新的 issue 来获得反馈。
 
 要检查某个模型是否已有 Flax、PyTorch 或 TensorFlow 的实现，或其是否在 🤗 Tokenizers 库中有对应词符化器（tokenizer），敬请参阅[此表](https://huggingface.co/docs/transformers/index#supported-frameworks)。
@@ -357,7 +379,7 @@ conda install -c huggingface transformers
 | [文档](https://huggingface.co/transformers/) | 完整的 API 文档和教程 |
 | [任务总结](https://huggingface.co/docs/transformers/task_summary) | 🤗 Transformers 支持的任务 |
 | [预处理教程](https://huggingface.co/docs/transformers/preprocessing) | 使用 `Tokenizer` 来为模型准备数据 |
-| [训练和微调](https://huggingface.co/docstransformers/training) | 在 PyTorch/TensorFlow 的训练循环或 `Trainer` API 中使用 🤗 Transformers 提供的模型 |
+| [训练和微调](https://huggingface.co/docs/transformers/training) | 在 PyTorch/TensorFlow 的训练循环或 `Trainer` API 中使用 🤗 Transformers 提供的模型 |
 | [快速上手：微调和用例脚本](https://github.com/huggingface/transformers/tree/main/examples) | 为各种任务提供的用例脚本 |
 | [模型分享和上传](https://huggingface.co/docs/transformers/model_sharing) | 和社区上传和分享你微调的模型 |
 | [迁移](https://huggingface.co/docs/transformers/migration) | 从 `pytorch-transformers` 或 `pytorch-pretrained-bert` 迁移到 🤗 Transformers |

README_zh-hant.md

@@ -257,16 +257,19 @@ conda install -c huggingface transformers
 1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (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/docs/transformers/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.
+1. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/). 
 1. **[BORT](https://huggingface.co/docs/transformers/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.
 1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (from Google Research) released with the paper [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) by Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel.
 1. **[CamemBERT](https://huggingface.co/docs/transformers/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.
 1. **[CANINE](https://huggingface.co/docs/transformers/model_doc/canine)** (from Google Research) released with the paper [CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language Representation](https://arxiv.org/abs/2103.06874) by Jonathan H. Clark, Dan Garrette, Iulia Turc, John Wieting.
 1. **[CLIP](https://huggingface.co/docs/transformers/model_doc/clip)** (from OpenAI) released with the paper [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) by Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, Ilya Sutskever.
+1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[ConvBERT](https://huggingface.co/docs/transformers/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.
-1. **[ConvNeXT](https://huggingface.co/docs/transformers/main/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
+1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
 1. **[CPM](https://huggingface.co/docs/transformers/model_doc/cpm)** (from Tsinghua University) released with the paper [CPM: A Large-scale Generative Chinese Pre-trained Language Model](https://arxiv.org/abs/2012.00413) by Zhengyan Zhang, Xu Han, Hao Zhou, Pei Ke, Yuxian Gu, Deming Ye, Yujia Qin, Yusheng Su, Haozhe Ji, Jian Guan, Fanchao Qi, Xiaozhi Wang, Yanan Zheng, Guoyang Zeng, Huanqi Cao, Shengqi Chen, Daixuan Li, Zhenbo Sun, Zhiyuan Liu, Minlie Huang, Wentao Han, Jie Tang, Juanzi Li, Xiaoyan Zhu, Maosong Sun.
 1. **[CTRL](https://huggingface.co/docs/transformers/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.
-1. **[Data2Vec](https://huggingface.co/docs/transformers/main/model_doc/data2vec)** (from Facebook) released with the paper [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) by Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli.
+1. **[CvT](https://huggingface.co/docs/transformers/model_doc/cvt)** (from Microsoft) released with the paper [CvT: Introducing Convolutions to Vision Transformers](https://arxiv.org/abs/2103.15808) by Haiping Wu, Bin Xiao, Noel Codella, Mengchen Liu, Xiyang Dai, Lu Yuan, Lei Zhang.
+1. **[Data2Vec](https://huggingface.co/docs/transformers/model_doc/data2vec)** (from Facebook) released with the paper [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) by Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli.
 1. **[DeBERTa](https://huggingface.co/docs/transformers/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.
 1. **[DeBERTa-v2](https://huggingface.co/docs/transformers/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.
 1. **[Decision Transformer](https://huggingface.co/docs/transformers/model_doc/decision_transformer)** (from Berkeley/Facebook/Google) released with the paper [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://arxiv.org/abs/2106.01345) by Lili Chen, Kevin Lu, Aravind Rajeswaran, Kimin Lee, Aditya Grover, Michael Laskin, Pieter Abbeel, Aravind Srinivas, Igor Mordatch.
@@ -280,46 +283,61 @@ conda install -c huggingface transformers
 1. **[ELECTRA](https://huggingface.co/docs/transformers/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.
 1. **[EncoderDecoder](https://huggingface.co/docs/transformers/model_doc/encoder-decoder)** (from Google Research) 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. **[FlauBERT](https://huggingface.co/docs/transformers/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.
+1. **[FLAVA](https://huggingface.co/docs/transformers/model_doc/flava)** (from Facebook AI) released with the paper [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) by Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach, and Douwe Kiela.
 1. **[FNet](https://huggingface.co/docs/transformers/model_doc/fnet)** (from Google Research) released with the paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/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.
-1. **[GLPN](https://huggingface.co/docs/transformers/main/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
+1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-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.
 1. **[GPT Neo](https://huggingface.co/docs/transformers/model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
+1. **[GPT NeoX](https://huggingface.co/docs/transformers/model_doc/gpt_neox)** (from EleutherAI) released with the paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) by Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach 
 1. **[GPT-2](https://huggingface.co/docs/transformers/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**.
 1. **[GPT-J](https://huggingface.co/docs/transformers/model_doc/gptj)** (from EleutherAI) released with the paper [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
+1. **[GroupViT](https://huggingface.co/docs/transformers/model_doc/groupvit)** (from UCSD, NVIDIA) released with the paper [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) by Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang.
 1. **[Hubert](https://huggingface.co/docs/transformers/model_doc/hubert)** (from Facebook) released with the paper [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://arxiv.org/abs/2106.07447) by Wei-Ning Hsu, Benjamin Bolte, Yao-Hung Hubert Tsai, Kushal Lakhotia, Ruslan Salakhutdinov, Abdelrahman Mohamed.
 1. **[I-BERT](https://huggingface.co/docs/transformers/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.
-1. **[ImageGPT](https://huggingface.co/docs/transformers/main/model_doc/imagegpt)** (from OpenAI) released with the paper [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) by Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever.
+1. **[ImageGPT](https://huggingface.co/docs/transformers/model_doc/imagegpt)** (from OpenAI) released with the paper [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) by Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever.
 1. **[LayoutLM](https://huggingface.co/docs/transformers/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.
 1. **[LayoutLMv2](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (from Microsoft Research Asia) released with the paper [LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding](https://arxiv.org/abs/2012.14740) by Yang Xu, Yiheng Xu, Tengchao Lv, Lei Cui, Furu Wei, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Wanxiang Che, Min Zhang, Lidong Zhou.
+1. **[LayoutLMv3](https://huggingface.co/docs/transformers/model_doc/layoutlmv3)** (from Microsoft Research Asia) released with the paper [LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking](https://arxiv.org/abs/2204.08387) by Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, Furu Wei.
 1. **[LayoutXLM](https://huggingface.co/docs/transformers/model_doc/layoutlmv2)** (from Microsoft Research Asia) released with the paper [LayoutXLM: Multimodal Pre-training for Multilingual Visually-rich Document Understanding](https://arxiv.org/abs/2104.08836) by Yiheng Xu, Tengchao Lv, Lei Cui, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Furu Wei.
 1. **[LED](https://huggingface.co/docs/transformers/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.
+1. **[LeViT](https://huggingface.co/docs/transformers/model_doc/levit)** (from Meta AI) released with the paper [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) by Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze.
 1. **[Longformer](https://huggingface.co/docs/transformers/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.
+1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](https://huggingface.co/docs/transformers/model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
 1. **[LXMERT](https://huggingface.co/docs/transformers/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.
+1. **[M-CTC-T](https://huggingface.co/docs/transformers/model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](https://huggingface.co/docs/transformers/model_doc/m2m_100)** (from Facebook) released with the paper [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.
 1. **[MarianMT](https://huggingface.co/docs/transformers/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.
-1. **[MaskFormer](https://huggingface.co/docs/transformers/main/model_doc/maskformer)** (from Meta and UIUC) released with the paper [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) by Bowen Cheng, Alexander G. Schwing, Alexander Kirillov 
-1. **[MBart](https://huggingface.co/docs/transformers/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.
-1. **[MBart-50](https://huggingface.co/docs/transformers/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.
+1. **[MaskFormer](https://huggingface.co/docs/transformers/model_doc/maskformer)** (from Meta and UIUC) released with the paper [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) by Bowen Cheng, Alexander G. Schwing, Alexander Kirillov 
+1. **[mBART](https://huggingface.co/docs/transformers/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.
+1. **[mBART-50](https://huggingface.co/docs/transformers/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.
 1. **[Megatron-BERT](https://huggingface.co/docs/transformers/model_doc/megatron-bert)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
 1. **[Megatron-GPT2](https://huggingface.co/docs/transformers/model_doc/megatron_gpt2)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
 1. **[mLUKE](https://huggingface.co/docs/transformers/model_doc/mluke)** (from Studio Ousia) released with the paper [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) by Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka.
+1. **[MobileBERT](https://huggingface.co/docs/transformers/model_doc/mobilebert)** (from CMU/Google Brain) released with the paper [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) by Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou.
+1. **[MobileViT](https://huggingface.co/docs/transformers/model_doc/mobilevit)** (from Apple) released with the paper [MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer](https://arxiv.org/abs/2110.02178) by Sachin Mehta and Mohammad Rastegari.
 1. **[MPNet](https://huggingface.co/docs/transformers/model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MT5](https://huggingface.co/docs/transformers/model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
-1. **[Nyströmformer](https://huggingface.co/docs/transformers/main/model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
+1. **[MVP](https://huggingface.co/docs/transformers/model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
+1. **[Nezha](https://huggingface.co/docs/transformers/model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
+1. **[NLLB](https://huggingface.co/docs/transformers/model_doc/nllb)** (from Meta) released with the paper [No Language Left Behind: Scaling Human-Centered Machine Translation](https://arxiv.org/abs/2207.04672) by the NLLB team.
+1. **[Nyströmformer](https://huggingface.co/docs/transformers/model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
+1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
+1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](https://huggingface.co/docs/transformers/model_doc/phobert)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.
-1. **[PLBart](https://huggingface.co/docs/transformers/main/model_doc/plbart)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
-1. **[PoolFormer](https://huggingface.co/docs/transformers/main/model_doc/poolformer)** (from Sea AI Labs) released with the paper [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) by Yu, Weihao and Luo, Mi and Zhou, Pan and Si, Chenyang and Zhou, Yichen and Wang, Xinchao and Feng, Jiashi and Yan, Shuicheng.
+1. **[PLBart](https://huggingface.co/docs/transformers/model_doc/plbart)** (from UCLA NLP) released with the paper [Unified Pre-training for Program Understanding and Generation](https://arxiv.org/abs/2103.06333) by Wasi Uddin Ahmad, Saikat Chakraborty, Baishakhi Ray, Kai-Wei Chang.
+1. **[PoolFormer](https://huggingface.co/docs/transformers/model_doc/poolformer)** (from Sea AI Labs) released with the paper [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) by Yu, Weihao and Luo, Mi and Zhou, Pan and Si, Chenyang and Zhou, Yichen and Wang, Xinchao and Feng, Jiashi and Yan, Shuicheng.
 1. **[ProphetNet](https://huggingface.co/docs/transformers/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.
 1. **[QDQBert](https://huggingface.co/docs/transformers/model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
+1. **[RAG](https://huggingface.co/docs/transformers/model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
 1. **[REALM](https://huggingface.co/docs/transformers/model_doc/realm.html)** (from Google Research) released with the paper [REALM: Retrieval-Augmented Language Model Pre-Training](https://arxiv.org/abs/2002.08909) by Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat and Ming-Wei Chang.
 1. **[Reformer](https://huggingface.co/docs/transformers/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.
-1. **[RegNet](https://huggingface.co/docs/transformers/main/model_doc/regnet)** (from META Research) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár. 
+1. **[RegNet](https://huggingface.co/docs/transformers/model_doc/regnet)** (from META Research) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár. 
 1. **[RemBERT](https://huggingface.co/docs/transformers/model_doc/rembert)** (from Google Research) released with the paper [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/pdf/2010.12821.pdf) by Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder.
-1. **[ResNet](https://huggingface.co/docs/transformers/main/model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun. 
+1. **[ResNet](https://huggingface.co/docs/transformers/model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun. 
 1. **[RoBERTa](https://huggingface.co/docs/transformers/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.
 1. **[RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer)** (from ZhuiyiTechnology), released together with the paper a [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/pdf/2104.09864v1.pdf) by Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu.
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
@@ -328,24 +346,27 @@ conda install -c huggingface transformers
 1. **[SpeechToTextTransformer](https://huggingface.co/docs/transformers/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.
 1. **[SpeechToTextTransformer2](https://huggingface.co/docs/transformers/model_doc/speech_to_text_2)** (from Facebook) released with the paper [Large-Scale Self- and Semi-Supervised Learning for Speech Translation](https://arxiv.org/abs/2104.06678) by Changhan Wang, Anne Wu, Juan Pino, Alexei Baevski, Michael Auli, Alexis Conneau.
 1. **[Splinter](https://huggingface.co/docs/transformers/model_doc/splinter)** (from Tel Aviv University) released with the paper [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) by Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy.
-1. **[SqueezeBert](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (from Berkeley) 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. **[Swin Transformer](https://huggingface.co/docs/transformers/main/model_doc/swin)** (from Microsoft) released with the paper [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
+1. **[SqueezeBERT](https://huggingface.co/docs/transformers/model_doc/squeezebert)** (from Berkeley) 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. **[Swin Transformer](https://huggingface.co/docs/transformers/model_doc/swin)** (from Microsoft) released with the paper [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
 1. **[T5](https://huggingface.co/docs/transformers/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.
 1. **[T5v1.1](https://huggingface.co/docs/transformers/model_doc/t5v1.1)** (from Google AI) released with the paper [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) 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/docs/transformers/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.
-1. **[TAPEX](https://huggingface.co/docs/transformers/main/model_doc/tapex)** (from Microsoft Research) released with the paper [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) by Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, Jian-Guang Lou.
+1. **[TAPEX](https://huggingface.co/docs/transformers/model_doc/tapex)** (from Microsoft Research) released with the paper [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) by Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, Jian-Guang Lou.
+1. **[Trajectory Transformer](https://huggingface.co/docs/transformers/model_doc/trajectory_transformers)** (from the University of California at Berkeley) released with the paper [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://arxiv.org/abs/2106.02039) by Michael Janner, Qiyang Li, Sergey Levine 
 1. **[Transformer-XL](https://huggingface.co/docs/transformers/model_doc/transfo-xl)** (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. **[TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr)** (from Microsoft) released with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
+1. **[UL2](https://huggingface.co/docs/transformers/model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler 
 1. **[UniSpeech](https://huggingface.co/docs/transformers/model_doc/unispeech)** (from Microsoft Research) released with the paper [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) by Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang.
 1. **[UniSpeechSat](https://huggingface.co/docs/transformers/model_doc/unispeech-sat)** (from Microsoft Research) released with the paper [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) by Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu.
-1. **[VAN](https://huggingface.co/docs/transformers/main/model_doc/van)** (from Tsinghua University and Nankai University) released with the paper [Visual Attention Network](https://arxiv.org/pdf/2202.09741.pdf) by Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
-1. **[ViLT](https://huggingface.co/docs/transformers/main/model_doc/vilt)** (from NAVER AI Lab/Kakao Enterprise/Kakao Brain) released with the paper [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) by Wonjae Kim, Bokyung Son, Ildoo Kim.
+1. **[VAN](https://huggingface.co/docs/transformers/model_doc/van)** (from Tsinghua University and Nankai University) released with the paper [Visual Attention Network](https://arxiv.org/pdf/2202.09741.pdf) by Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
+1. **[ViLT](https://huggingface.co/docs/transformers/model_doc/vilt)** (from NAVER AI Lab/Kakao Enterprise/Kakao Brain) released with the paper [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) by Wonjae Kim, Bokyung Son, Ildoo Kim.
 1. **[Vision Transformer (ViT)](https://huggingface.co/docs/transformers/model_doc/vit)** (from Google AI) released with the paper [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://arxiv.org/abs/2010.11929) by Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby. 
 1. **[VisualBERT](https://huggingface.co/docs/transformers/model_doc/visual_bert)** (from UCLA NLP) released with the paper [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) by Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang.
-1. **[ViTMAE](https://huggingface.co/docs/transformers/main/model_doc/vit_mae)** (from Meta AI) released with the paper [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) by Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick.
+1. **[ViTMAE](https://huggingface.co/docs/transformers/model_doc/vit_mae)** (from Meta AI) released with the paper [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) by Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick.
 1. **[Wav2Vec2](https://huggingface.co/docs/transformers/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.
+1. **[Wav2Vec2-Conformer](https://huggingface.co/docs/transformers/model_doc/wav2vec2-conformer)** (from Facebook AI) released 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, Sravya Popuri, Dmytro Okhonko, Juan Pino.
 1. **[Wav2Vec2Phoneme](https://huggingface.co/docs/transformers/model_doc/wav2vec2_phoneme)** (from Facebook AI) released with the paper [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) by Qiantong Xu, Alexei Baevski, Michael Auli.
-1. **[WavLM](https://huggingface.co/docs/transformers/main/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
+1. **[WavLM](https://huggingface.co/docs/transformers/model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
 1. **[XGLM](https://huggingface.co/docs/transformers/model_doc/xglm)** (From Facebook AI) released with the paper [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) by Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li. 
 1. **[XLM](https://huggingface.co/docs/transformers/model_doc/xlm)** (from Facebook) released together with the paper [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) by Guillaume Lample and Alexis Conneau.
 1. **[XLM-ProphetNet](https://huggingface.co/docs/transformers/model_doc/xlm-prophetnet)** (from Microsoft Research) released with the paper [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) by Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
@@ -354,7 +375,8 @@ conda install -c huggingface transformers
 1. **[XLNet](https://huggingface.co/docs/transformers/model_doc/xlnet)** (from Google/CMU) released with the paper [​XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) by Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
 1. **[XLS-R](https://huggingface.co/docs/transformers/model_doc/xls_r)** (from Facebook AI) released with the paper [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) by Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli.
 1. **[XLSR-Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/xlsr_wav2vec2)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
-1. **[YOSO](https://huggingface.co/docs/transformers/main/model_doc/yoso)** (from the University of Wisconsin - Madison) released with the paper [You Only Sample (Almost) by Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh.
+1. **[YOLOS](https://huggingface.co/docs/transformers/model_doc/yolos)** (from Huazhong University of Science & Technology) released with the paper [You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection](https://arxiv.org/abs/2106.00666) by Yuxin Fang, Bencheng Liao, Xinggang Wang, Jiemin Fang, Jiyang Qi, Rui Wu, Jianwei Niu, Wenyu Liu.
+1. **[YOSO](https://huggingface.co/docs/transformers/model_doc/yoso)** (from the University of Wisconsin - Madison) released with the paper [You Only Sample (Almost) by Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh.
 1. 想要貢獻新的模型？我們這裡有一份**詳細指引和模板**來引導你加入新的模型。你可以在 [`templates`](./templates) 目錄中找到它們。記得查看[貢獻指引](./CONTRIBUTING.md)並在開始寫 PR 前聯繫維護人員或開一個新的 issue 來獲得 feedbacks。
 
 要檢查某個模型是否已有 Flax、PyTorch 或 TensorFlow 的實作，或其是否在🤗 Tokenizers 函式庫中有對應的 tokenizer，敬請參閱[此表](https://huggingface.co/docs/transformers/index#supported-frameworks)。

docker/transformers-all-latest-gpu/Dockerfile

@@ -3,20 +3,48 @@ LABEL maintainer="Hugging Face"
 
 ARG DEBIAN_FRONTEND=noninteractive
 
+# Use login shell to read variables from `~/.profile` (to pass dynamic created variables between RUN commands)
+SHELL ["sh", "-lc"]
+
+# The following `ARG` are mainly used to specify the versions explicitly & directly in this docker file, and not meant
+# to be used as arguments for docker build (so far).
+
+ARG PYTORCH='1.12.0'
+# (not always a valid torch version)
+ARG INTEL_TORCH_EXT='1.11.0'
+# Example: `cu102`, `cu113`, etc.
+ARG CUDA='cu113'
+
 RUN apt update
-RUN apt install -y git libsndfile1-dev tesseract-ocr espeak-ng python3 python3-pip ffmpeg
+RUN apt install -y git libsndfile1-dev tesseract-ocr espeak-ng python3 python3-pip ffmpeg git-lfs
+RUN git lfs install
 RUN python3 -m pip install --no-cache-dir --upgrade pip
 
 ARG REF=main
 RUN git clone https://github.com/huggingface/transformers && cd transformers && git checkout $REF
 RUN python3 -m pip install --no-cache-dir -e ./transformers[dev,onnxruntime]
 
-RUN python3 -m pip install --no-cache-dir -U torch tensorflow
+# TODO: Handle these in a python utility script
+RUN [ ${#PYTORCH} -gt 0 -a "$PYTORCH" != "pre" ] && VERSION='torch=='$PYTORCH'.*' ||  VERSION='torch'; echo "export VERSION='$VERSION'" >> ~/.profile
+RUN echo torch=$VERSION
+# `torchvision` and `torchaudio` should be installed along with `torch`, especially for nightly build.
+# Currently, let's just use their latest releases (when `torch` is installed with a release version)
+# TODO: We might need to specify proper versions that work with a specific torch version (especially for past CI).
+RUN [ "$PYTORCH" != "pre" ] && python3 -m pip install --no-cache-dir -U $VERSION torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/$CUDA || python3 -m pip install --no-cache-dir -U --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/$CUDA
+
+RUN python3 -m pip install --no-cache-dir -U tensorflow
 RUN python3 -m pip uninstall -y flax jax
-RUN python3 -m pip install --no-cache-dir torch-scatter -f https://data.pyg.org/whl/torch-$(python3 -c "from torch import version; print(version.__version__.split('+')[0])")+cu102.html
+
+# Use installed torch version for `torch-scatter` to avid to deal with PYTORCH='pre'.
+# If torch is nightly version, the link is likely to be invalid, but the installation falls back to the latest torch-scatter
+RUN python3 -m pip install --no-cache-dir torch-scatter -f https://data.pyg.org/whl/torch-$(python3 -c "from torch import version; print(version.__version__.split('+')[0])")+$CUDA.html
+RUN python3 -m pip install --no-cache-dir intel_extension_for_pytorch==$INTEL_TORCH_EXT+cpu -f https://software.intel.com/ipex-whl-stable
+
 RUN python3 -m pip install --no-cache-dir git+https://github.com/facebookresearch/detectron2.git pytesseract https://github.com/kpu/kenlm/archive/master.zip
 RUN python3 -m pip install -U "itsdangerous<2.1.0"
 
+RUN python3 -m pip install --no-cache-dir git+https://github.com/huggingface/accelerate@main#egg=accelerate
+
 # When installing in editable mode, `transformers` is not recognized as a package.
 # this line must be added in order for python to be aware of transformers.
 RUN cd transformers && python3 setup.py develop

docker/transformers-doc-builder/Dockerfile

@@ -15,5 +15,6 @@ RUN python3 -m pip install --no-cache-dir torchvision git+https://github.com/fac
 RUN python3 -m pip install --no-cache-dir pytorch-quantization --extra-index-url https://pypi.ngc.nvidia.com
 RUN python3 -m pip install -U "itsdangerous<2.1.0"
 
-RUN doc-builder build transformers transformers/docs/source --build_dir doc-build-dev --notebook_dir notebooks/transformers_doc --clean --version pr_$PR_NUMBER
+# Test if the image could successfully build the doc. before publishing the image
+RUN doc-builder build transformers transformers/docs/source/en --build_dir doc-build-dev --notebook_dir notebooks/transformers_doc --clean
 RUN rm -rf doc-build-dev

docker/transformers-past-gpu/Dockerfile

@@ -0,0 +1,43 @@
+ARG BASE_DOCKER_IMAGE="nvidia/cuda:11.2.2-cudnn8-devel-ubuntu20.04"
+FROM $BASE_DOCKER_IMAGE
+LABEL maintainer="Hugging Face"
+
+ARG DEBIAN_FRONTEND=noninteractive
+
+# Use login shell to read variables from `~/.profile` (to pass dynamic created variables between RUN commands)
+SHELL ["sh", "-lc"]
+
+RUN apt update
+RUN apt install -y git libsndfile1-dev tesseract-ocr espeak-ng python3 python3-pip ffmpeg git-lfs
+RUN git lfs install
+RUN python3 -m pip install --no-cache-dir --upgrade pip
+
+ARG REF=main
+RUN git clone https://github.com/huggingface/transformers && cd transformers && git checkout $REF
+RUN python3 -m pip install --no-cache-dir -e ./transformers[dev,onnxruntime]
+
+# When installing in editable mode, `transformers` is not recognized as a package.
+# this line must be added in order for python to be aware of transformers.
+RUN cd transformers && python3 setup.py develop
+
+ARG FRAMEWORK
+ARG VERSION
+
+# Remove all frameworks
+# (`accelerate` requires `torch`, and this causes import issues for TF-only testing)
+RUN python3 -m pip uninstall -y torch torchvision torchaudio accelerate tensorflow jax flax
+
+# Get the libraries and their versions to install, and write installation command to `~/.profile`.
+RUN python3 ./transformers/utils/past_ci_versions.py --framework $FRAMEWORK --version $VERSION
+
+# Install the target framework
+RUN echo "INSTALL_CMD = $INSTALL_CMD"
+RUN $INSTALL_CMD
+
+# Having installation problems for torch-scatter with torch <= 1.6. Disable so we have the same set of tests.
+# (This part will be removed once the logic of using `past_ci_versions.py` is used in other Dockerfile files.)
+# # Use installed torch version for `torch-scatter`.
+# # (The env. variable $CUDA is defined in `past_ci_versions.py`)
+# RUN [ "$FRAMEWORK" = "pytorch" ] && python3 -m pip install --no-cache-dir torch-scatter -f https://data.pyg.org/whl/torch-$(python3 -c "from torch import version; print(version.__version__.split('+')[0])")+$CUDA.html || echo "torch-scatter not to be installed"
+
+RUN python3 -m pip install -U "itsdangerous<2.1.0"

docker/transformers-pytorch-deepspeed-latest-gpu/Dockerfile

@@ -3,16 +3,30 @@ LABEL maintainer="Hugging Face"
 
 ARG DEBIAN_FRONTEND=noninteractive
 
+ARG PYTORCH='1.12.0'
+# Example: `cu102`, `cu113`, etc.
+ARG CUDA='cu113'
+
 RUN apt -y update
 RUN apt install -y libaio-dev
 RUN python3 -m pip install --no-cache-dir --upgrade pip
 
 ARG REF=main
 RUN git clone https://github.com/huggingface/transformers && cd transformers && git checkout $REF
-RUN python3 -m pip install --no-cache-dir -e ./transformers[testing,deepspeed]
 
-RUN git clone https://github.com/microsoft/DeepSpeed && cd DeepSpeed && rm -rf build && \
-    DS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 python3 -m pip install -e . --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check 2>&1
+# Install latest release PyTorch
+# (PyTorch must be installed before pre-compiling any DeepSpeed c++/cuda ops.)
+# (https://www.deepspeed.ai/tutorials/advanced-install/#pre-install-deepspeed-ops)
+RUN python3 -m pip install --no-cache-dir -U torch==$PYTORCH torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/$CUDA
+
+RUN python3 -m pip install --no-cache-dir ./transformers[deepspeed-testing]
+
+# Pre-build **latest** DeepSpeed, so it would be ready for testing (otherwise, the 1st deepspeed test will timeout)
+RUN python3 -m pip uninstall -y deepspeed
+# This has to be run (again) inside the GPU VMs running the tests.
+# The installation works here, but some tests fail, if we don't pre-build deepspeed again in the VMs running the tests.
+# TODO: Find out why test fail.
+RUN DS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 python3 -m pip install deepspeed --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check 2>&1
 
 # When installing in editable mode, `transformers` is not recognized as a package.
 # this line must be added in order for python to be aware of transformers.

docker/transformers-pytorch-deepspeed-nightly-gpu/Dockerfile

@@ -0,0 +1,35 @@
+FROM nvcr.io/nvidia/pytorch:21.03-py3
+LABEL maintainer="Hugging Face"
+
+ARG DEBIAN_FRONTEND=noninteractive
+
+# Example: `cu102`, `cu113`, etc.
+ARG CUDA='cu113'
+
+RUN apt -y update
+RUN apt install -y libaio-dev
+RUN python3 -m pip install --no-cache-dir --upgrade pip
+
+ARG REF=main
+RUN git clone https://github.com/huggingface/transformers && cd transformers && git checkout $REF
+
+# Install **nightly** release PyTorch (flag `--pre`)
+# (PyTorch must be installed before pre-compiling any DeepSpeed c++/cuda ops.)
+# (https://www.deepspeed.ai/tutorials/advanced-install/#pre-install-deepspeed-ops)
+RUN python3 -m pip install --no-cache-dir -U --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/$CUDA
+
+RUN python3 -m pip install --no-cache-dir ./transformers[deepspeed-testing]
+
+# Pre-build **nightly** release of DeepSpeed, so it would be ready for testing (otherwise, the 1st deepspeed test will timeout)
+RUN python3 -m pip uninstall -y deepspeed
+# This has to be run inside the GPU VMs running the tests. (So far, it fails here due to GPU checks during compilation.)
+# Issue: https://github.com/microsoft/DeepSpeed/issues/2010
+# RUN git clone https://github.com/microsoft/DeepSpeed && cd DeepSpeed && rm -rf build && \
+#    DS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 python3 -m pip install . --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check 2>&1
+
+# When installing in editable mode, `transformers` is not recognized as a package.
+# this line must be added in order for python to be aware of transformers.
+RUN cd transformers && python3 setup.py develop
+
+# Disable for now as deepspeed is not installed above. To be enabled once the issue is fixed.
+# RUN python3 -c "from deepspeed.launcher.runner import main"

docker/transformers-pytorch-gpu/Dockerfile

@@ -12,12 +12,17 @@ RUN git clone https://github.com/huggingface/transformers && cd transformers &&
 RUN python3 -m pip install --no-cache-dir -e ./transformers[dev-torch,testing]
 
 # If set to nothing, will install the latest version
-ARG PYTORCH=''
+ARG PYTORCH='1.12.0'
+ARG TORCH_VISION=''
+ARG TORCH_AUDIO=''
+
+RUN [ ${#PYTORCH} -gt 0 ] && VERSION='torch=='$PYTORCH'.*' ||  VERSION='torch'; python3 -m pip install --no-cache-dir -U $VERSION --extra-index-url https://download.pytorch.org/whl/cu113
+RUN [ ${#TORCH_VISION} -gt 0 ] && VERSION='torchvision=='TORCH_VISION'.*' ||  VERSION='torchvision'; python3 -m pip install --no-cache-dir -U $VERSION --extra-index-url https://download.pytorch.org/whl/cu113
+RUN [ ${#TORCH_AUDIO} -gt 0 ] && VERSION='torchaudio=='TORCH_AUDIO'.*' ||  VERSION='torchaudio'; python3 -m pip install --no-cache-dir -U $VERSION --extra-index-url https://download.pytorch.org/whl/cu113
 
-RUN [ ${#PYTORCH} -gt 0 ] && VERSION='torch=='$PYTORCH'.*' ||  VERSION='torch'; python3 -m pip install --no-cache-dir -U $VERSION
 RUN python3 -m pip uninstall -y tensorflow flax
 
-RUN python3 -m pip install --no-cache-dir torch-scatter -f https://data.pyg.org/whl/torch-$(python3 -c "from torch import version; print(version.__version__.split('+')[0])")+cu102.html
+RUN python3 -m pip install --no-cache-dir torch-scatter -f https://data.pyg.org/whl/torch-$(python3 -c "from torch import version; print(version.__version__.split('+')[0])")+cu113.html
 RUN python3 -m pip install --no-cache-dir git+https://github.com/facebookresearch/detectron2.git pytesseract https://github.com/kpu/kenlm/archive/master.zip
 RUN python3 -m pip install -U "itsdangerous<2.1.0"
 

docs/README.md

@@ -50,11 +50,32 @@ You can adapt the `--build_dir` to set any temporary folder that you prefer. Thi
 the MDX files that will be rendered as the documentation on the main website. You can inspect them in your favorite
 Markdown editor.
 
+## Previewing the documentation
+
+To preview the docs, first install the `watchdog` module with:
+
+```bash
+pip install watchdog
+```
+
+Then run the following command:
+
+```bash
+doc-builder preview {package_name} {path_to_docs}
+```
+
+For example:
+
+```bash
+doc-builder preview transformers docs/source/en/
+```
+
+The docs will be viewable at [http://localhost:3000](http://localhost:3000). You can also preview the docs once you have opened a PR. You will see a bot add a comment to a link where the documentation with your changes lives.
+
 ---
 **NOTE**
 
-It's not possible to see locally how the final documentation will look like for now. Once you have opened a PR, you
-will see a bot add a comment to a link where the documentation with your changes lives.
+The `preview` command only works with existing doc files. When you add a completely new file, you need to update `_toctree.yml` & restart `preview` command (`ctrl-c` to stop it & call `doc-builder preview ...` again).
 
 ---
 
@@ -108,6 +129,11 @@ Make sure to put your new file under the proper section. It's unlikely to go in
 depending on the intended targets (beginners, more advanced users or researchers) it should go in section two, three or
 four.
 
+### Translating
+
+When translating, refer to the guide at [./TRANSLATING.md](https://github.com/huggingface/transformers/blob/main/docs/TRANSLATING.md).
+
+
 ### Adding a new model
 
 When adding a new model:
@@ -402,4 +428,4 @@ Here are a few tips to help you debug the doctests and make them pass:
   * whitespace: one give whitespace (space, tabulation, new line) is equivalent to any number of whitespace, so you can add new lines where there are spaces to make your output more readable.
   * numerical values: you should never put more than 4 or 5 digits to expected results as different setups or library versions might get you slightly different results. `doctest` is configure to ignore any difference lower than the precision to which you wrote (so 1e-4 if you write 4 digits).
 - Don't leave a block of code that is very long to execute. If you can't make it fast, you can either not use the doctest syntax on it (so that it's ignored), or if you want to use the doctest syntax to show the results, you can add a comment `# doctest: +SKIP` at the end of the lines of code too long to execute
-- Each line of code that produces a result needs to have that result written below. You can ignore an output if you don't want to show it in your code example by adding a comment ` # doctest: +IGNORE_RESULT` at the end of the line of code produing it.
+- Each line of code that produces a result needs to have that result written below. You can ignore an output if you don't want to show it in your code example by adding a comment ` # doctest: +IGNORE_RESULT` at the end of the line of code producing it.

docs/TRANSLATING.md

@@ -0,0 +1,58 @@
+### Translating the Transformers documentation into your language
+
+As part of our mission to democratize machine learning, we'd love to make the Transformers library available in many more languages! Follow the steps below if you want to help translate the documentation into your language 🙏.
+
+**🗞️ Open an issue**
+
+To get started, navigate to the [Issues](https://github.com/huggingface/transformers/issues) page of this repo and check if anyone else has opened an issue for your language. If not, open a new issue by selecting the "Translation template" from the "New issue" button.
+
+Once an issue exists, post a comment to indicate which chapters you'd like to work on, and we'll add your name to the list.
+
+
+**🍴 Fork the repository**
+
+First, you'll need to [fork the Transformers repo](https://docs.github.com/en/get-started/quickstart/fork-a-repo). You can do this by clicking on the **Fork** button on the top-right corner of this repo's page.
+
+Once you've forked the repo, you'll want to get the files on your local machine for editing. You can do that by cloning the fork with Git as follows:
+
+```bash
+git clone https://github.com/YOUR-USERNAME/transformers.git
+```
+
+**📋 Copy-paste the English version with a new language code**
+
+The documentation files are in one leading directory:
+
+- [`docs/source`](https://github.com/huggingface/transformers/tree/main/docs/source): All the documentation materials are organized here by language.
+
+You'll only need to copy the files in the [`docs/source/en`](https://github.com/huggingface/transformers/tree/main/docs/source/en) directory, so first navigate to your fork of the repo and run the following:
+
+```bash
+cd ~/path/to/transformers/docs
+cp -r source/en source/LANG-ID
+```
+
+Here, `LANG-ID` should be one of the ISO 639-1 or ISO 639-2 language codes -- see [here](https://www.loc.gov/standards/iso639-2/php/code_list.php) for a handy table.
+
+**✍️ Start translating**
+
+The fun part comes - translating the text!
+
+The first thing we recommend is translating the part of the `_toctree.yml` file that corresponds to your doc chapter. This file is used to render the table of contents on the website. 
+
+> 🙋 If the `_toctree.yml` file doesn't yet exist for your language, you can create one by copy-pasting from the English version and deleting the sections unrelated to your chapter. Just make sure it exists in the `docs/source/LANG-ID/` directory!
+
+The fields you should add are `local` (with the name of the file containing the translation; e.g. `autoclass_tutorial`), and `title` (with the title of the doc in your language; e.g. `Load pretrained instances with an AutoClass`) -- as a reference, here is the `_toctree.yml` for [English](https://github.com/huggingface/transformers/blob/main/docs/source/en/_toctree.yml):
+
+```yaml
+- sections:
+  - local: pipeline_tutorial # Do not change this! Use the same name for your .md file
+    title: Pipelines for inference # Translate this!
+    ...
+  title: Tutorials # Translate this!
+```
+
+Once you have translated the `_toctree.yml` file, you can start translating the [MDX](https://mdxjs.com/) files associated with your docs chapter.
+
+> 🙋 If you'd like others to help you with the translation, you can either [open an issue](https://github.com/huggingface/transformers/issues) or tag @[espejelomar](https://twitter.com/espejelomar)
+ on Twitter to gain some visibility.

docs/source/en/_toctree.yml

@@ -1,4 +1,4 @@
-- sections: 
+- sections:
   - local: index
     title: 🤗 Transformers
   - local: quicktour
@@ -22,7 +22,7 @@
   title: Tutorials
 - sections:
   - local: fast_tokenizers
-    title: "Use tokenizers from 🤗 Tokenizers"
+    title: Use tokenizers from 🤗 Tokenizers
   - local: create_a_model
     title: Create a custom architecture
   - local: custom_models
@@ -59,10 +59,32 @@
     title: Converting TensorFlow Checkpoints
   - local: serialization
     title: Export 🤗 Transformers models
-  - local: performance
-    title: 'Performance and Scalability: How To Fit a Bigger Model and Train It Faster'
-  - local: parallelism
-    title: Model Parallelism
+  - sections:
+    - local: performance
+      title: Overview
+    - local: perf_train_gpu_one
+      title: Training on one GPU
+    - local: perf_train_gpu_many
+      title: Training on many GPUs
+    - local: perf_train_cpu
+      title: Training on CPU
+    - local: perf_train_tpu
+      title: Training on TPUs
+    - local: perf_train_special
+      title: Training on Specialized Hardware
+    - local: perf_infer_cpu
+      title: Inference on CPU
+    - local: perf_infer_gpu_one
+      title: Inference on one GPU
+    - local: perf_infer_gpu_many
+      title: Inference on many GPUs
+    - local: perf_infer_special
+      title: Inference on Specialized Hardware
+    - local: perf_hardware
+      title: Custom hardware for training
+    title: Performance and scalability
+  - local: big_models
+    title: Instantiating a big model
   - local: benchmarks
     title: Benchmarks
   - local: migration
@@ -72,15 +94,15 @@
   - local: debugging
     title: Debugging
   - local: notebooks
-    title: "🤗 Transformers Notebooks"
+    title: 🤗 Transformers Notebooks
   - local: community
     title: Community
   - local: contributing
     title: How to contribute to transformers?
   - local: add_new_model
-    title: "How to add a model to 🤗 Transformers?"
+    title: How to add a model to 🤗 Transformers?
   - local: add_new_pipeline
-    title: "How to add a pipeline to 🤗 Transformers?"
+    title: How to create a custom pipeline?
   - local: testing
     title: Testing
   - local: pr_checks
@@ -154,12 +176,12 @@
       title: BEiT
     - local: model_doc/bert
       title: BERT
-    - local: model_doc/bertweet
-      title: Bertweet
     - local: model_doc/bert-generation
       title: BertGeneration
     - local: model_doc/bert-japanese
       title: BertJapanese
+    - local: model_doc/bertweet
+      title: Bertweet
     - local: model_doc/big_bird
       title: BigBird
     - local: model_doc/bigbird_pegasus
@@ -168,6 +190,8 @@
       title: Blenderbot
     - local: model_doc/blenderbot-small
       title: Blenderbot Small
+    - local: model_doc/bloom
+      title: BLOOM
     - local: model_doc/bort
       title: BORT
     - local: model_doc/byt5
@@ -176,16 +200,20 @@
       title: CamemBERT
     - local: model_doc/canine
       title: CANINE
-    - local: model_doc/convnext
-      title: ConvNeXT
     - local: model_doc/clip
       title: CLIP
+    - local: model_doc/codegen
+      title: CodeGen
     - local: model_doc/convbert
       title: ConvBERT
+    - local: model_doc/convnext
+      title: ConvNeXT
     - local: model_doc/cpm
       title: CPM
     - local: model_doc/ctrl
       title: CTRL
+    - local: model_doc/cvt
+      title: CvT
     - local: model_doc/data2vec
       title: Data2Vec
     - local: model_doc/deberta
@@ -214,6 +242,8 @@
       title: Encoder Decoder Models
     - local: model_doc/flaubert
       title: FlauBERT
+    - local: model_doc/flava
+      title: FLAVA
     - local: model_doc/fnet
       title: FNet
     - local: model_doc/fsmt
@@ -222,8 +252,22 @@
       title: Funnel Transformer
     - local: model_doc/glpn
       title: GLPN
+    - local: model_doc/openai-gpt
+      title: GPT
+    - local: model_doc/gpt_neo
+      title: GPT Neo
+    - local: model_doc/gpt_neox
+      title: GPT NeoX
+    - local: model_doc/gptj
+      title: GPT-J
+    - local: model_doc/gpt2
+      title: GPT2
+    - local: model_doc/groupvit
+      title: GroupViT
     - local: model_doc/herbert
       title: HerBERT
+    - local: model_doc/hubert
+      title: Hubert
     - local: model_doc/ibert
       title: I-BERT
     - local: model_doc/imagegpt
@@ -232,24 +276,32 @@
       title: LayoutLM
     - local: model_doc/layoutlmv2
       title: LayoutLMV2
+    - local: model_doc/layoutlmv3
+      title: LayoutLMV3
     - local: model_doc/layoutxlm
       title: LayoutXLM
     - local: model_doc/led
       title: LED
+    - local: model_doc/levit
+      title: LeViT
     - local: model_doc/longformer
       title: Longformer
+    - local: model_doc/longt5
+      title: LongT5
     - local: model_doc/luke
       title: LUKE
     - local: model_doc/lxmert
       title: LXMERT
+    - local: model_doc/m2m_100
+      title: M2M100
     - local: model_doc/marian
       title: MarianMT
     - local: model_doc/maskformer
       title: MaskFormer
-    - local: model_doc/m2m_100
-      title: M2M100
     - local: model_doc/mbart
       title: MBart and MBart-50
+    - local: model_doc/mctct
+      title: MCTCT
     - local: model_doc/megatron-bert
       title: MegatronBERT
     - local: model_doc/megatron_gpt2
@@ -258,26 +310,28 @@
       title: mLUKE
     - local: model_doc/mobilebert
       title: MobileBERT
+    - local: model_doc/mobilevit
+      title: MobileViT
     - local: model_doc/mpnet
       title: MPNet
     - local: model_doc/mt5
       title: MT5
+    - local: model_doc/mvp
+      title: MVP
+    - local: model_doc/nezha
+      title: NEZHA
+    - local: model_doc/nllb
+      title: NLLB
     - local: model_doc/nystromformer
       title: Nyströmformer
-    - local: model_doc/openai-gpt
-      title: OpenAI GPT
-    - local: model_doc/gpt2
-      title: OpenAI GPT2
-    - local: model_doc/gptj
-      title: GPT-J
-    - local: model_doc/gpt_neo
-      title: GPT Neo
-    - local: model_doc/hubert
-      title: Hubert
-    - local: model_doc/perceiver
-      title: Perceiver
+    - local: model_doc/opt
+      title: OPT
+    - local: model_doc/owlvit
+      title: OWL-ViT
     - local: model_doc/pegasus
       title: Pegasus
+    - local: model_doc/perceiver
+      title: Perceiver
     - local: model_doc/phobert
       title: PhoBERT
     - local: model_doc/plbart
@@ -294,10 +348,10 @@
       title: REALM
     - local: model_doc/reformer
       title: Reformer
-    - local: model_doc/rembert
-      title: RemBERT
     - local: model_doc/regnet
       title: RegNet
+    - local: model_doc/rembert
+      title: RemBERT
     - local: model_doc/resnet
       title: ResNet
     - local: model_doc/retribert
@@ -332,10 +386,14 @@
       title: TAPAS
     - local: model_doc/tapex
       title: TAPEX
+    - local: model_doc/trajectory_transformer
+      title: Trajectory Transformer
     - local: model_doc/transfo-xl
       title: Transformer XL
     - local: model_doc/trocr
       title: TrOCR
+    - local: model_doc/ul2
+      title: UL2
     - local: model_doc/unispeech
       title: UniSpeech
     - local: model_doc/unispeech-sat
@@ -350,12 +408,14 @@
       title: Vision Text Dual Encoder
     - local: model_doc/vit
       title: Vision Transformer (ViT)
-    - local: model_doc/vit_mae
-      title: ViTMAE
     - local: model_doc/visual_bert
       title: VisualBERT
+    - local: model_doc/vit_mae
+      title: ViTMAE
     - local: model_doc/wav2vec2
       title: Wav2Vec2
+    - local: model_doc/wav2vec2-conformer
+      title: Wav2Vec2-Conformer
     - local: model_doc/wav2vec2_phoneme
       title: Wav2Vec2Phoneme
     - local: model_doc/wavlm
@@ -372,10 +432,12 @@
       title: XLM-RoBERTa-XL
     - local: model_doc/xlnet
       title: XLNet
-    - local: model_doc/xlsr_wav2vec2
-      title: XLSR-Wav2Vec2
     - local: model_doc/xls_r
       title: XLS-R
+    - local: model_doc/xlsr_wav2vec2
+      title: XLSR-Wav2Vec2
+    - local: model_doc/yolos
+      title: YOLOS
     - local: model_doc/yoso
       title: YOSO
     title: Models

docs/source/en/add_new_pipeline.mdx

@@ -9,7 +9,10 @@ Unless required by applicable law or agreed to in writing, software distributed
 an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
 -->
 
-# How to add a pipeline to 🤗 Transformers?
+# How to create a custom pipeline?
+
+In this guide, we will see how to create a custom pipeline and share it on the [Hub](hf.co/models) or add it to the
+Transformers library.
 
 First and foremost, you need to decide the raw entries the pipeline will be able to take. It can be strings, raw bytes,
 dictionaries or whatever seems to be the most likely desired input. Try to keep these inputs as pure Python as possible
@@ -99,7 +102,7 @@ def _sanitize_parameters(self, **kwargs):
 
     postprocess_kwargs = {}
     if "top_k" in kwargs:
-        preprocess_kwargs["top_k"] = kwargs["top_k"]
+        postprocess_kwargs["top_k"] = kwargs["top_k"]
     return preprocess_kwargs, {}, postprocess_kwargs
 ```
 
@@ -111,12 +114,123 @@ of arguments for ease of use (audio files, can be filenames, URLs or pure bytes)
 
 ## Adding it to the list of supported tasks
 
-Go to `src/transformers/pipelines/__init__.py` and fill in `SUPPORTED_TASKS` with your newly created pipeline.
-If possible it should provide a default model.
+To register your `new-task` to the list of supported tasks, you have to add it to the `PIPELINE_REGISTRY`:
 
-## Adding tests
+```python
+from transformers.pipelines import PIPELINE_REGISTRY
 
-Create a new file `tests/test_pipelines_MY_PIPELINE.py` with example with the other tests.
+PIPELINE_REGISTRY.register_pipeline(
+    "new-task",
+    pipeline_class=MyPipeline,
+    pt_model=AutoModelForSequenceClassification,
+)
+```
+
+You can specify a default model if you want, in which case it should come with a specific revision (which can be the name of a branch or a commit hash, here we took `"abcdef"`) as well was the type:
+
+```python
+PIPELINE_REGISTRY.register_pipeline(
+    "new-task",
+    pipeline_class=MyPipeline,
+    pt_model=AutoModelForSequenceClassification,
+    default={"pt": ("user/awesome_model", "abcdef")},
+    type="text",  # current support type: text, audio, image, multimodal
+)
+```
+
+## Share your pipeline on the Hub
+
+To share your custom pipeline on the Hub, you just have to save the custom code of your `Pipeline` subclass in a
+python file. For instance, let's say we want to use a custom pipeline for sentence pair classification like this:
+
+```py
+import numpy as np
+
+from transformers import Pipeline
+
+
+def softmax(outputs):
+    maxes = np.max(outputs, axis=-1, keepdims=True)
+    shifted_exp = np.exp(outputs - maxes)
+    return shifted_exp / shifted_exp.sum(axis=-1, keepdims=True)
+
+
+class PairClassificationPipeline(Pipeline):
+    def _sanitize_parameters(self, **kwargs):
+        preprocess_kwargs = {}
+        if "second_text" in kwargs:
+            preprocess_kwargs["second_text"] = kwargs["second_text"]
+        return preprocess_kwargs, {}, {}
+
+    def preprocess(self, text, second_text=None):
+        return self.tokenizer(text, text_pair=second_text, return_tensors=self.framework)
+
+    def _forward(self, model_inputs):
+        return self.model(**model_inputs)
+
+    def postprocess(self, model_outputs):
+        logits = model_outputs.logits[0].numpy()
+        probabilities = softmax(logits)
+
+        best_class = np.argmax(probabilities)
+        label = self.model.config.id2label[best_class]
+        score = probabilities[best_class].item()
+        logits = logits.tolist()
+        return {"label": label, "score": score, "logits": logits}
+```
+
+The implementation is framework agnostic, and will work for PyTorch and TensorFlow models. If we have saved this in
+a file named `pair_classification.py`, we can then import it and register it like this:
+
+```py
+from pair_classification import PairClassificationPipeline
+from transformers.pipelines import PIPELINE_REGISTRY
+from transformers import AutoModelForSequenceClassification, TFAutoModelForSequenceClassification
+
+PIPELINE_REGISTRY.register_pipeline(
+    "pair-classification",
+    pipeline_class=PairClassificationPipeline,
+    pt_model=AutoModelForSequenceClassification,
+    tf_model=TFAutoModelForSequenceClassification,
+)
+```
+
+Once this is done, we can use it with a pretrained model. For instance `sgugger/finetuned-bert-mrpc` has been
+fine-tuned on the MRPC dataset, which classifies pairs of sentences as paraphrases or not.
+
+```py
+from transformers import pipeline
+
+classifier = pipeline("pair-classification", model="sgugger/finetuned-bert-mrpc")
+```
+
+Then we can share it on the Hub by using the `save_pretrained` method in a `Repository`:
+
+```py
+from huggingface_hub import Repository
+
+repo = Repository("test-dynamic-pipeline", clone_from="{your_username}/test-dynamic-pipeline")
+classifier.save_pretrained("test-dynamic-pipeline")
+repo.push_to_hub()
+```
+
+This will copy the file where you defined `PairClassificationPipeline` inside the folder `"test-dynamic-pipeline"`,
+along with saving the model and tokenizer of the pipeline, before pushing everything in the repository
+`{your_username}/test-dynamic-pipeline`. After that anyone can use it as long as they provide the option
+`trust_remote_code=True`:
+
+```py
+from transformers import pipeline
+
+classifier = pipeline(model="{your_username}/test-dynamic-pipeline", trust_remote_code=True)
+```
+
+## Add the pipeline to Transformers
+
+If you want to contribute your pipeline to Transformers, you will need to add a new module in the `pipelines` submodule
+with the code of your pipeline, then add it in the list of tasks defined in `pipelines/__init__.py`.
+
+Then you will need to add tests. Create a new file `tests/test_pipelines_MY_PIPELINE.py` with example with the other tests.
 
 The `run_pipeline_test` function will be very generic and run on small random models on every possible
 architecture as defined by `model_mapping` and `tf_model_mapping`.

docs/source/en/big_models.mdx

@@ -0,0 +1,119 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# Instantiating a big model
+
+When you want to use a very big pretrained model, one challenge is to minimize the use of the RAM. The usual workflow
+from PyTorch is:
+
+1. Create your model with random weights.
+2. Load your pretrained weights.
+3. Put those pretrained weights in your random model.
+
+Step 1 and 2 both require a full version of the model in memory, which is not a problem in most cases, but if your model starts weighing several GigaBytes, those two copies can make you got our of RAM. Even worse, if you are using `torch.distributed` to launch a distributed training, each process will load the pretrained model and store these two copies in RAM.
+
+<Tip>
+
+Note that the randomly created model is initialized with "empty" tensors, which take the space in memory without filling it (thus the random values are whatever was in this chunk of memory at a given time). The random initialization following the appropriate distribution for the kind of model/parameters instatiated (like a normal distribution for instance) is only performed after step 3 on the non-initialized weights, to be as fast as possible! 
+
+</Tip>
+
+In this guide, we explore the solutions Transformers offer to deal with this issue. Note that this is an area of active development, so the APIs explained here may change slightly in the future.
+
+## Sharded checkpoints
+
+Since version 4.18.0, model checkpoints that end up taking more than 10GB of space are automatically sharded in smaller pieces. In terms of having one single checkpoint when you do `model.save_pretrained(save_dir)`, you will end up with several partial checkpoints (each of which being of size < 10GB) and an index that maps parameter names to the files they are stored in.
+
+You can control the maximum size before sharding with the `max_shard_size` parameter, so for the sake of an example, we'll use a normal-size models with a small shard size: let's take a traditional BERT model.
+
+```py
+from transformers import AutoModel
+
+model = AutoModel.from_pretrained("bert-base-cased")
+```
+
+If you save it using [`~PreTrainedModel.save_pretrained`], you will get a new folder with two files: the config of the model and its weights:
+
+```py
+>>> import os
+>>> import tempfile
+
+>>> with tempfile.TemporaryDirectory() as tmp_dir:
+...     model.save_pretrained(tmp_dir)
+...     print(sorted(os.listdir(tmp_dir)))
+['config.json', 'pytorch_model.bin']
+```
+
+Now let's use a maximum shard size of 200MB:
+
+```py
+>>> with tempfile.TemporaryDirectory() as tmp_dir:
+...     model.save_pretrained(tmp_dir, max_shard_size="200MB")
+...     print(sorted(os.listdir(tmp_dir)))
+['config.json', 'pytorch_model-00001-of-00003.bin', 'pytorch_model-00002-of-00003.bin', 'pytorch_model-00003-of-00003.bin', 'pytorch_model.bin.index.json']
+```
+
+On top of the configuration of the model, we see three different weights files, and an `index.json` file which is our index. A checkpoint like this can be fully reloaded using the [`~PreTrainedModel.from_pretrained`] method:
+
+```py
+>>> with tempfile.TemporaryDirectory() as tmp_dir:
+...     model.save_pretrained(tmp_dir, max_shard_size="200MB")
+...     new_model = AutoModel.from_pretrained(tmp_dir)
+```
+
+The main advantage of doing this for big models is that during step 2 of the workflow shown above, each shard of the checkpoint is loaded after the previous one, capping the memory usage in RAM to the model size plus the size of the biggest shard.
+
+Beind the scenes, the index file is used to determine which keys are in the checkpoint, and where the corresponding weights are stored. We can load that index like any json and get a dictionary:
+
+```py
+>>> import json
+
+>>> with tempfile.TemporaryDirectory() as tmp_dir:
+...     model.save_pretrained(tmp_dir, max_shard_size="200MB")
+...     with open(os.path.join(tmp_dir, "pytorch_model.bin.index.json"), "r") as f:
+...         index = json.load(f)
+
+>>> print(index.keys())
+dict_keys(['metadata', 'weight_map'])
+```
+
+The metadata just consists of the total size of the model for now. We plan to add several other informations in the future:
+
+```py
+>>> index["metadata"]
+{'total_size': 433245184}
+```
+
+The weights map is the main part of this index, which maps each parameter name (as usually found in a PyTorch model `state_dict`) to the file it's stored in:
+
+```py
+>>> index["weight_map"]
+{'embeddings.LayerNorm.bias': 'pytorch_model-00001-of-00003.bin',
+ 'embeddings.LayerNorm.weight': 'pytorch_model-00001-of-00003.bin',
+ ...
+```
+
+If you want to directly load such a sharded checkpoint inside a model without using [`~PreTrainedModel.from_pretrained`] (like you would do `model.load_state_dict()` for a full checkpoint) you should use [`~modeling_utils.load_sharded_checkpoint`]:
+
+```py
+>>> from transformers.modeling_utils import load_sharded_checkpoint
+
+>>> with tempfile.TemporaryDirectory() as tmp_dir:
+...     model.save_pretrained(tmp_dir, max_shard_size="200MB")
+...     load_sharded_checkpoint(model, tmp_dir)
+```
+
+## Low memory loading
+
+Sharded checkpoints reduce the memory usage during step 2 of the workflow mentioned above, but in order to use that model in a low memory setting, we recommend leveraging our tools based on the Accelerate library.
+
+Please read the following guide for more information: [Large model loading using Accelerate](./main_classes/model#large-model-loading)

docs/source/en/custom_models.mdx

@@ -106,7 +106,7 @@ directly upload your config to the Hub.
 
 Now that we have our ResNet configuration, we can go on writing the model. We will actually write two: one that
 extracts the hidden features from a batch of images (like [`BertModel`]) and one that is suitable for image
-classification (like [`BertModelForSequenceClassification`]).
+classification (like [`BertForSequenceClassification`]).
 
 As we mentioned before, we'll only write a loose wrapper of the model to keep it simple for this example. The only
 thing we need to do before writing this class is a map between the block types and actual block classes. Then the
@@ -289,7 +289,7 @@ from huggingface_hub import notebook_login
 notebook_login()
 ```
 
-You can then push to to your own namespace (or an organization you are a member of) like this:
+You can then push to your own namespace (or an organization you are a member of) like this:
 
 ```py
 resnet50d.push_to_hub("custom-resnet50d")

docs/source/en/index.mdx

@@ -28,7 +28,7 @@ Each 🤗 Transformers architecture is defined in a standalone Python module so
 ## If you are looking for custom support from the Hugging Face team
 
 <a target="_blank" href="https://huggingface.co/support">
-<img alt="HuggingFace Expert Acceleration Program" src="https://huggingface.co/front/thumbnails/support.png" style="max-width: 600px; border: 1px solid #eee; border-radius: 4px; box-shadow: 0 1px 2px 0 rgba(0, 0, 0, 0.05);">
+    <img alt="HuggingFace Expert Acceleration Program" src="https://cdn-media.huggingface.co/marketing/transformers/new-support-improved.png" style="max-width: 600px; border: 1px solid #eee; border-radius: 4px; box-shadow: 0 1px 2px 0 rgba(0, 0, 0, 0.05);">
 </a><br>
 
 ## Contents
@@ -57,63 +57,80 @@ The library currently contains JAX, PyTorch and TensorFlow implementations, pret
 1. **[BARTpho](model_doc/bartpho)** (from VinAI Research) released with the paper [BARTpho: Pre-trained Sequence-to-Sequence Models for Vietnamese](https://arxiv.org/abs/2109.09701) by Nguyen Luong Tran, Duong Minh Le and Dat Quoc Nguyen.
 1. **[BEiT](model_doc/beit)** (from Microsoft) released with the paper [BEiT: BERT Pre-Training of Image Transformers](https://arxiv.org/abs/2106.08254) by Hangbo Bao, Li Dong, Furu Wei.
 1. **[BERT](model_doc/bert)** (from Google) released with the paper [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://arxiv.org/abs/1810.04805) by Jacob Devlin, Ming-Wei Chang, Kenton Lee and Kristina Toutanova.
-1. **[BERTweet](model_doc/bertweet)** (from VinAI Research) released with the paper [BERTweet: A pre-trained language model for English Tweets](https://aclanthology.org/2020.emnlp-demos.2/) by Dat Quoc Nguyen, Thanh Vu and Anh Tuan Nguyen.
 1. **[BERT For Sequence Generation](model_doc/bert-generation)** (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. **[BigBird-RoBERTa](model_doc/big_bird)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
+1. **[BERTweet](model_doc/bertweet)** (from VinAI Research) released with the paper [BERTweet: A pre-trained language model for English Tweets](https://aclanthology.org/2020.emnlp-demos.2/) by Dat Quoc Nguyen, Thanh Vu and Anh Tuan Nguyen.
 1. **[BigBird-Pegasus](model_doc/bigbird_pegasus)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
+1. **[BigBird-RoBERTa](model_doc/big_bird)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
 1. **[Blenderbot](model_doc/blenderbot)** (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](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.
+1. **[BLOOM](model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/).
 1. **[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.
 1. **[ByT5](model_doc/byt5)** (from Google Research) released with the paper [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) by Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel.
 1. **[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.
 1. **[CANINE](model_doc/canine)** (from Google Research) released with the paper [CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language Representation](https://arxiv.org/abs/2103.06874) by Jonathan H. Clark, Dan Garrette, Iulia Turc, John Wieting.
-1. **[ConvNeXT](model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
 1. **[CLIP](model_doc/clip)** (from OpenAI) released with the paper [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) by Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, Ilya Sutskever.
+1. **[CodeGen](model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
 1. **[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.
+1. **[ConvNeXT](model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
 1. **[CPM](model_doc/cpm)** (from Tsinghua University) released with the paper [CPM: A Large-scale Generative Chinese Pre-trained Language Model](https://arxiv.org/abs/2012.00413) by Zhengyan Zhang, Xu Han, Hao Zhou, Pei Ke, Yuxian Gu, Deming Ye, Yujia Qin, Yusheng Su, Haozhe Ji, Jian Guan, Fanchao Qi, Xiaozhi Wang, Yanan Zheng, Guoyang Zeng, Huanqi Cao, Shengqi Chen, Daixuan Li, Zhenbo Sun, Zhiyuan Liu, Minlie Huang, Wentao Han, Jie Tang, Juanzi Li, Xiaoyan Zhu, Maosong Sun.
 1. **[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.
+1. **[CvT](model_doc/cvt)** (from Microsoft) released with the paper [CvT: Introducing Convolutions to Vision Transformers](https://arxiv.org/abs/2103.15808) by Haiping Wu, Bin Xiao, Noel Codella, Mengchen Liu, Xiyang Dai, Lu Yuan, Lei Zhang.
 1. **[Data2Vec](model_doc/data2vec)** (from Facebook) released with the paper [Data2Vec:  A General Framework for Self-supervised Learning in Speech, Vision and Language](https://arxiv.org/abs/2202.03555) by Alexei Baevski, Wei-Ning Hsu, Qiantong Xu, Arun Babu, Jiatao Gu, Michael Auli.
 1. **[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.
 1. **[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.
 1. **[Decision Transformer](model_doc/decision_transformer)** (from Berkeley/Facebook/Google) released with the paper [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://arxiv.org/abs/2106.01345) by Lili Chen, Kevin Lu, Aravind Rajeswaran, Kimin Lee, Aditya Grover, Michael Laskin, Pieter Abbeel, Aravind Srinivas, Igor Mordatch.
-1. **[DiT](model_doc/dit)** (from Microsoft Research) released with the paper [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) by Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei.
 1. **[DeiT](model_doc/deit)** (from Facebook) released with the paper [Training data-efficient image transformers & distillation through attention](https://arxiv.org/abs/2012.12877) by Hugo Touvron, Matthieu Cord, Matthijs Douze, Francisco Massa, Alexandre Sablayrolles, Hervé Jégou.
 1. **[DETR](model_doc/detr)** (from Facebook) released with the paper [End-to-End Object Detection with Transformers](https://arxiv.org/abs/2005.12872) by Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, Sergey Zagoruyko.
 1. **[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.
 1. **[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/main/examples/research_projects/distillation), RoBERTa into [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation), Multilingual BERT into [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation) and a German version of DistilBERT.
+1. **[DiT](model_doc/dit)** (from Microsoft Research) released with the paper [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) by Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei.
 1. **[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.
 1. **[DPT](master/model_doc/dpt)** (from Intel Labs) released with the paper [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) by René Ranftl, Alexey Bochkovskiy, Vladlen Koltun.
-1. **[EncoderDecoder](model_doc/encoder-decoder)** (from Google Research) 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. **[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.
+1. **[EncoderDecoder](model_doc/encoder-decoder)** (from Google Research) 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. **[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.
+1. **[FLAVA](model_doc/flava)** (from Facebook AI) released with the paper [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) by Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach, and Douwe Kiela.
 1. **[FNet](model_doc/fnet)** (from Google Research) released with the paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
 1. **[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.
 1. **[GLPN](model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
 1. **[GPT](model_doc/openai-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.
+1. **[GPT Neo](model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
+1. **[GPT NeoX](model_doc/gpt_neox)** (from EleutherAI) released with the paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) by Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach
 1. **[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**.
 1. **[GPT-J](model_doc/gptj)** (from EleutherAI) released in the repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
-1. **[GPT Neo](model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
+1. **[GroupViT](model_doc/groupvit)** (from UCSD, NVIDIA) released with the paper [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) by Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang.
 1. **[Hubert](model_doc/hubert)** (from Facebook) released with the paper [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://arxiv.org/abs/2106.07447) by Wei-Ning Hsu, Benjamin Bolte, Yao-Hung Hubert Tsai, Kushal Lakhotia, Ruslan Salakhutdinov, Abdelrahman Mohamed.
 1. **[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.
 1. **[ImageGPT](model_doc/imagegpt)** (from OpenAI) released with the paper [Generative Pretraining from Pixels](https://openai.com/blog/image-gpt/) by Mark Chen, Alec Radford, Rewon Child, Jeffrey Wu, Heewoo Jun, David Luan, Ilya Sutskever.
 1. **[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.
 1. **[LayoutLMv2](model_doc/layoutlmv2)** (from Microsoft Research Asia) released with the paper [LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding](https://arxiv.org/abs/2012.14740) by Yang Xu, Yiheng Xu, Tengchao Lv, Lei Cui, Furu Wei, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Wanxiang Che, Min Zhang, Lidong Zhou.
+1. **[LayoutLMv3](model_doc/layoutlmv3)** (from Microsoft Research Asia) released with the paper [LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking](https://arxiv.org/abs/2204.08387) by Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, Furu Wei.
 1. **[LayoutXLM](model_doc/layoutlmv2)** (from Microsoft Research Asia) released with the paper [LayoutXLM: Multimodal Pre-training for Multilingual Visually-rich Document Understanding](https://arxiv.org/abs/2104.08836) by Yiheng Xu, Tengchao Lv, Lei Cui, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Furu Wei.
 1. **[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.
+1. **[LeViT](model_doc/levit)** (from Meta AI) released with the paper [LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference](https://arxiv.org/abs/2104.01136) by Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze.
 1. **[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.
+1. **[LongT5](model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.
 1. **[LUKE](model_doc/luke)** (from Studio Ousia) released with the paper [LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention](https://arxiv.org/abs/2010.01057) by Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto.
-1. **[mLUKE](model_doc/mluke)** (from Studio Ousia) released with the paper [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) by Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka.
 1. **[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.
+1. **[M-CTC-T](model_doc/mctct)** (from Facebook) released with the paper [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert.
 1. **[M2M100](model_doc/m2m_100)** (from Facebook) released with the paper [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.
 1. **[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.
 1. **[MaskFormer](model_doc/maskformer)** (from Meta and UIUC) released with the paper [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://arxiv.org/abs/2107.06278) by Bowen Cheng, Alexander G. Schwing, Alexander Kirillov.
-1. **[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.
-1. **[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.
+1. **[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.
+1. **[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.
 1. **[Megatron-BERT](model_doc/megatron-bert)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
 1. **[Megatron-GPT2](model_doc/megatron_gpt2)** (from NVIDIA) released with the paper [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://arxiv.org/abs/1909.08053) by Mohammad Shoeybi, Mostofa Patwary, Raul Puri, Patrick LeGresley, Jared Casper and Bryan Catanzaro.
+1. **[mLUKE](model_doc/mluke)** (from Studio Ousia) released with the paper [mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models](https://arxiv.org/abs/2110.08151) by Ryokan Ri, Ikuya Yamada, and Yoshimasa Tsuruoka.
+1. **[MobileBERT](model_doc/mobilebert)** (from CMU/Google Brain) released with the paper [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://arxiv.org/abs/2004.02984) by Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou.
+1. **[MobileViT](model_doc/mobilevit)** (from Apple) released with the paper [MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer](https://arxiv.org/abs/2110.02178) by Sachin Mehta and Mohammad Rastegari.
 1. **[MPNet](model_doc/mpnet)** (from Microsoft Research) released with the paper [MPNet: Masked and Permuted Pre-training for Language Understanding](https://arxiv.org/abs/2004.09297) by Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, Tie-Yan Liu.
 1. **[MT5](model_doc/mt5)** (from Google AI) released with the paper [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934) by Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
+1. **[MVP](model_doc/mvp)** (from RUC AI Box) released with the paper [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
+1. **[Nezha](model_doc/nezha)** (from Huawei Noah’s Ark Lab) released with the paper [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, Wenyong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen and Qun Liu.
+1. **[NLLB](model_doc/nllb)** (from Meta) released with the paper [No Language Left Behind: Scaling Human-Centered Machine Translation](https://arxiv.org/abs/2207.04672) by the NLLB team.
 1. **[Nyströmformer](model_doc/nystromformer)** (from the University of Wisconsin - Madison) released with the paper [Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention](https://arxiv.org/abs/2102.03902) by Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh.
+1. **[OPT](master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
+1. **[OWL-ViT](model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[Pegasus](model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
 1. **[Perceiver IO](model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.
 1. **[PhoBERT](model_doc/phobert)** (from VinAI Research) released with the paper [PhoBERT: Pre-trained language models for Vietnamese](https://www.aclweb.org/anthology/2020.findings-emnlp.92/) by Dat Quoc Nguyen and Anh Tuan Nguyen.
@@ -121,10 +138,11 @@ The library currently contains JAX, PyTorch and TensorFlow implementations, pret
 1. **[PoolFormer](model_doc/poolformer)** (from Sea AI Labs) released with the paper [MetaFormer is Actually What You Need for Vision](https://arxiv.org/abs/2111.11418) by Yu, Weihao and Luo, Mi and Zhou, Pan and Si, Chenyang and Zhou, Yichen and Wang, Xinchao and Feng, Jiashi and Yan, Shuicheng.
 1. **[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.
 1. **[QDQBert](model_doc/qdqbert)** (from NVIDIA) released with the paper [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://arxiv.org/abs/2004.09602) by Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev and Paulius Micikevicius.
+1. **[RAG](model_doc/rag)** (from Facebook) released with the paper [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://arxiv.org/abs/2005.11401) by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela.
 1. **[REALM](model_doc/realm.html)** (from Google Research) released with the paper [REALM: Retrieval-Augmented Language Model Pre-Training](https://arxiv.org/abs/2002.08909) by Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat and Ming-Wei Chang.
 1. **[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.
-1. **[RemBERT](model_doc/rembert)** (from Google Research) released with the paper [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/abs/2010.12821) by Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder.
 1. **[RegNet](model_doc/regnet)** (from META Platforms) released with the paper [Designing Network Design Space](https://arxiv.org/abs/2003.13678) by Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, Piotr Dollár.
+1. **[RemBERT](model_doc/rembert)** (from Google Research) released with the paper [Rethinking embedding coupling in pre-trained language models](https://arxiv.org/abs/2010.12821) by Hyung Won Chung, Thibault Févry, Henry Tsai, M. Johnson, Sebastian Ruder.
 1. **[ResNet](model_doc/resnet)** (from Microsoft Research) released with the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) by Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun.
 1. **[RoBERTa](model_doc/roberta)** (from Facebook), released together with the paper [RoBERTa: 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.
 1. **[RoFormer](model_doc/roformer)** (from ZhuiyiTechnology), released together with the paper [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://arxiv.org/abs/2104.09864) by Jianlin Su and Yu Lu and Shengfeng Pan and Bo Wen and Yunfeng Liu.
@@ -134,32 +152,36 @@ The library currently contains JAX, PyTorch and TensorFlow implementations, pret
 1. **[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.
 1. **[SpeechToTextTransformer2](model_doc/speech_to_text_2)** (from Facebook), released together with the paper [Large-Scale Self- and Semi-Supervised Learning for Speech Translation](https://arxiv.org/abs/2104.06678) by Changhan Wang, Anne Wu, Juan Pino, Alexei Baevski, Michael Auli, Alexis Conneau.
 1. **[Splinter](model_doc/splinter)** (from Tel Aviv University), released together with the paper [Few-Shot Question Answering by Pretraining Span Selection](https://arxiv.org/abs/2101.00438) by Ori Ram, Yuval Kirstain, Jonathan Berant, Amir Globerson, Omer Levy.
-1. **[SqueezeBert](model_doc/squeezebert)** (from Berkeley) 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. **[SqueezeBERT](model_doc/squeezebert)** (from Berkeley) 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. **[Swin Transformer](model_doc/swin)** (from Microsoft) released with the paper [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030) by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
 1. **[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.
 1. **[T5v1.1](model_doc/t5v1.1)** (from Google AI) released in the repository [google-research/text-to-text-transfer-transformer](https://github.com/google-research/text-to-text-transfer-transformer/blob/main/released_checkpoints.md#t511) 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](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.
 1. **[TAPEX](model_doc/tapex)** (from Microsoft Research) released with the paper [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://arxiv.org/abs/2107.07653) by Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, Jian-Guang Lou.
+1. **[Trajectory Transformer](model_doc/trajectory_transformers)** (from the University of California at Berkeley) released with the paper [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://arxiv.org/abs/2106.02039) by Michael Janner, Qiyang Li, Sergey Levine
 1. **[Transformer-XL](model_doc/transfo-xl)** (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. **[TrOCR](model_doc/trocr)** (from Microsoft), released together with the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Minghao Li, Tengchao Lv, Lei Cui, Yijuan Lu, Dinei Florencio, Cha Zhang, Zhoujun Li, Furu Wei.
+1. **[UL2](model_doc/ul2)** (from Google Research) released with the paper [Unifying Language Learning Paradigms](https://arxiv.org/abs/2205.05131v1) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
 1. **[UniSpeech](model_doc/unispeech)** (from Microsoft Research) released with the paper [UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data](https://arxiv.org/abs/2101.07597) by Chengyi Wang, Yu Wu, Yao Qian, Kenichi Kumatani, Shujie Liu, Furu Wei, Michael Zeng, Xuedong Huang.
 1. **[UniSpeechSat](model_doc/unispeech-sat)** (from Microsoft Research) released with the paper [UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING](https://arxiv.org/abs/2110.05752) by Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu.
 1. **[VAN](model_doc/van)** (from Tsinghua University and Nankai University) released with the paper [Visual Attention Network](https://arxiv.org/abs/2202.09741) by Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, Shi-Min Hu.
 1. **[ViLT](model_doc/vilt)** (from NAVER AI Lab/Kakao Enterprise/Kakao Brain) released with the paper [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://arxiv.org/abs/2102.03334) by Wonjae Kim, Bokyung Son, Ildoo Kim.
 1. **[Vision Transformer (ViT)](model_doc/vit)** (from Google AI) released with the paper [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://arxiv.org/abs/2010.11929) by Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby.
-1. **[ViTMAE](model_doc/vit_mae)** (from Meta AI) released with the paper [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) by Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick.
 1. **[VisualBERT](model_doc/visual_bert)** (from UCLA NLP) released with the paper [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://arxiv.org/pdf/1908.03557) by Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, Kai-Wei Chang.
-1. **[WavLM](model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
+1. **[ViTMAE](model_doc/vit_mae)** (from Meta AI) released with the paper [Masked Autoencoders Are Scalable Vision Learners](https://arxiv.org/abs/2111.06377) by Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick.
 1. **[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.
+1. **[Wav2Vec2-Conformer](model_doc/wav2vec2-conformer)** (from Facebook AI) released 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, Sravya Popuri, Dmytro Okhonko, Juan Pino.
 1. **[Wav2Vec2Phoneme](model_doc/wav2vec2_phoneme)** (from Facebook AI) released with the paper [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://arxiv.org/abs/2109.11680) by Qiantong Xu, Alexei Baevski, Michael Auli.
+1. **[WavLM](model_doc/wavlm)** (from Microsoft Research) released with the paper [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://arxiv.org/abs/2110.13900) by Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Furu Wei.
 1. **[XGLM](model_doc/xglm)** (From Facebook AI) released with the paper [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) by Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li.
 1. **[XLM](model_doc/xlm)** (from Facebook) released together with the paper [Cross-lingual Language Model Pretraining](https://arxiv.org/abs/1901.07291) by Guillaume Lample and Alexis Conneau.
 1. **[XLM-ProphetNet](model_doc/xlm-prophetnet)** (from Microsoft Research) released with the paper [ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training](https://arxiv.org/abs/2001.04063) by Yu Yan, Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
 1. **[XLM-RoBERTa](model_doc/xlm-roberta)** (from Facebook AI), released together with the paper [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) by Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer and Veselin Stoyanov.
 1. **[XLM-RoBERTa-XL](model_doc/xlm-roberta-xl)** (from Facebook AI), released together with the paper [Larger-Scale Transformers for Multilingual Masked Language Modeling](https://arxiv.org/abs/2105.00572) by Naman Goyal, Jingfei Du, Myle Ott, Giri Anantharaman, Alexis Conneau.
 1. **[XLNet](model_doc/xlnet)** (from Google/CMU) released with the paper [​XLNet: Generalized Autoregressive Pretraining for Language Understanding](https://arxiv.org/abs/1906.08237) by Zhilin Yang*, Zihang Dai*, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
-1. **[XLSR-Wav2Vec2](model_doc/xlsr_wav2vec2)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
 1. **[XLS-R](model_doc/xls_r)** (from Facebook AI) released with the paper [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://arxiv.org/abs/2111.09296) by Arun Babu, Changhan Wang, Andros Tjandra, Kushal Lakhotia, Qiantong Xu, Naman Goyal, Kritika Singh, Patrick von Platen, Yatharth Saraf, Juan Pino, Alexei Baevski, Alexis Conneau, Michael Auli.
+1. **[XLSR-Wav2Vec2](model_doc/xlsr_wav2vec2)** (from Facebook AI) released with the paper [Unsupervised Cross-Lingual Representation Learning For Speech Recognition](https://arxiv.org/abs/2006.13979) by Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
+1. **[YOLOS](model_doc/yolos)** (from Huazhong University of Science & Technology) released with the paper [You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection](https://arxiv.org/abs/2106.00666) by Yuxin Fang, Bencheng Liao, Xinggang Wang, Jiemin Fang, Jiyang Qi, Rui Wu, Jianwei Niu, Wenyu Liu.
 1. **[YOSO](model_doc/yoso)** (from the University of Wisconsin - Madison) released with the paper [You Only Sample (Almost) Once: Linear Cost Self-Attention Via Bernoulli Sampling](https://arxiv.org/abs/2111.09714) by Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh.
 
 
@@ -179,21 +201,25 @@ Flax), PyTorch, and/or TensorFlow.
 |            BERT             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
 |       Bert Generation       |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |           BigBird           |       ✅       |       ✅       |       ✅        |         ❌         |      ✅      |
-|       BigBirdPegasus        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|       BigBird-Pegasus       |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |         Blenderbot          |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
 |       BlenderbotSmall       |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|            BLOOM            |       ❌       |       ✅       |       ✅        |         ❌         |      ❌      |
 |          CamemBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|           Canine            |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           CANINE            |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            CLIP             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|           CodeGen           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
 |          ConvBERT           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|          ConvNext           |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|          ConvNeXT           |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
 |            CTRL             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|             CvT             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |        Data2VecAudio        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |        Data2VecText         |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|       Data2VecVision        |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
 |           DeBERTa           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|         DeBERTa-v2          |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|         DeBERTa-v2          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 |    Decision Transformer     |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|            DeiT             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            DeiT             |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
 |            DETR             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |         DistilBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
 |             DPR             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
@@ -202,31 +228,43 @@ Flax), PyTorch, and/or TensorFlow.
 |       Encoder decoder       |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
 | FairSeq Machine-Translation |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |          FlauBERT           |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|            FLAVA            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            FNet             |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
 |     Funnel Transformer      |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 |            GLPN             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |           GPT Neo           |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
+|          GPT NeoX           |       ❌       |       ✅       |       ✅        |         ❌         |      ❌      |
 |            GPT-J            |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
+|          GroupViT           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |           Hubert            |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
 |           I-BERT            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |          ImageGPT           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |          LayoutLM           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 |         LayoutLMv2          |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|         LayoutLMv3          |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
 |             LED             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|            LeViT            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |         Longformer          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|           LongT5            |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
 |            LUKE             |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |           LXMERT            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|           M-CTC-T           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |           M2M100            |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |           Marian            |       ✅       |       ❌       |       ✅        |         ✅         |      ✅      |
 |         MaskFormer          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            mBART            |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|        MegatronBert         |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|        Megatron-BERT        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |         MobileBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|          MobileViT          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            MPNet            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|             mT5             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|        Nystromformer        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             MT5             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|             MVP             |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|            Nezha            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|        Nyströmformer        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |         OpenAI GPT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 |        OpenAI GPT-2         |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|             OPT             |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
+|           OWL-ViT           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |           Pegasus           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
 |          Perceiver          |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |           PLBart            |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
@@ -234,15 +272,15 @@ Flax), PyTorch, and/or TensorFlow.
 |         ProphetNet          |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |           QDQBert           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |             RAG             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
-|            Realm            |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|            REALM            |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
 |          Reformer           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
-|           RegNet            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           RegNet            |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
 |           RemBERT           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|           ResNet            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           ResNet            |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
 |          RetriBERT          |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
 |           RoBERTa           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
 |          RoFormer           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|          SegFormer          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|          SegFormer          |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
 |             SEW             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            SEW-D            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |   Speech Encoder decoder    |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
@@ -250,10 +288,10 @@ Flax), PyTorch, and/or TensorFlow.
 |        Speech2Text2         |       ✅       |       ❌       |       ❌        |         ❌         |      ❌      |
 |          Splinter           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
 |         SqueezeBERT         |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
-|            Swin             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|      Swin Transformer       |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
 |             T5              |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
 |            TAPAS            |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
-|            TAPEX            |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|   Trajectory Transformer    |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |       Transformer-XL        |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
 |            TrOCR            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |          UniSpeech          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
@@ -262,17 +300,19 @@ Flax), PyTorch, and/or TensorFlow.
 |            ViLT             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |   Vision Encoder decoder    |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
 |    VisionTextDualEncoder    |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
-|         VisualBert          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|         VisualBERT          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |             ViT             |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
 |           ViTMAE            |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
 |          Wav2Vec2           |       ✅       |       ❌       |       ✅        |         ✅         |      ✅      |
+|     Wav2Vec2-Conformer      |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            WavLM            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            XGLM             |       ✅       |       ✅       |       ✅        |         ❌         |      ✅      |
 |             XLM             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|       XLM-ProphetNet        |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |         XLM-RoBERTa         |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
 |       XLM-RoBERTa-XL        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|        XLMProphetNet        |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            XLNet            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|            YOLOS            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            YOSO             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 
 <!-- End table-->

docs/source/en/installation.mdx

@@ -34,11 +34,16 @@ Start by creating a virtual environment in your project directory:
 python -m venv .env
 ```
 
-Activate the virtual environment:
+Activate the virtual environment. On Linux and MacOs:
 
 ```bash
 source .env/bin/activate
 ```
+Activate Virtual environment on Windows
+
+```bash
+.env/Scripts/activate
+```
 
 Now you're ready to install 🤗 Transformers with the following command:
 

docs/source/en/internal/generation_utils.mdx

@@ -127,6 +127,9 @@ generation.
 [[autodoc]] TopKLogitsWarper
     - __call__
 
+[[autodoc]] TypicalLogitsWarper
+    - __call__
+
 [[autodoc]] NoRepeatNGramLogitsProcessor
     - __call__
 

docs/source/en/internal/trainer_utils.mdx

@@ -22,6 +22,8 @@ Most of those are only useful if you are studying the code of the Trainer in the
 
 [[autodoc]] IntervalStrategy
 
+[[autodoc]] enable_full_determinism
+
 [[autodoc]] set_seed
 
 [[autodoc]] torch_distributed_zero_first

docs/source/en/main_classes/logging.mdx

@@ -40,29 +40,17 @@ Additionally, some `warnings` can be disabled by setting the environment variabl
 TRANSFORMERS_NO_ADVISORY_WARNINGS=1 ./myprogram.py
 ```
 
-Here is an example of how to use `logging` in a module:
+Here is an example of how to use the same logger as the library in your own module or script:
 
 ```python
 from transformers.utils import logging
 
 logging.set_verbosity_info()
-logger = logging.get_logger(__name__)
+logger = logging.get_logger("transformers")
 logger.info("INFO")
 logger.warning("WARN")
 ```
 
-Above, a `logger` instance is created from `logging.get_logger(__name__)`. If you want to use `logging` in a script, you shouldn't pass `__name__` to `logging.get_logger`. For example:
-
-```python
-from transformers.utils import logging
-
-if __name__ == "__main__":
-    logging.set_verbosity_info()
-    # leave it empy or use a string
-    logger = logging.get_logger()
-    logger.info("INFO")
-    logger.warning("WARN")
-```
 
 All the methods of this logging module are documented below, the main ones are
 [`logging.get_verbosity`] to get the current level of verbosity in the logger and

docs/source/en/main_classes/model.mdx

@@ -38,6 +38,75 @@ for text generation, [`~generation_utils.GenerationMixin`] (for the PyTorch mode
 
 <a id='from_pretrained-torch-dtype'></a>
 
+### Large model loading
+
+In Transformers 4.20.0, the [`~PreTrainedModel.from_pretrained`] method has been reworked to accommodate large models using [Accelerate](https://huggingface.co/docs/accelerate/big_modeling). This requires Accelerate >= 0.9.0 and PyTorch >= 1.9.0. Instead of creating the full model, then loading the pretrained weights inside it (which takes twice the size of the model in RAM, one for the randomly initialized model, one for the weights), there is an option to create the model as an empty shell, then only materialize its parameters when the pretrained weights are loaded.
+
+This option can be activated with `low_cpu_mem_usage=True`. The model is first created on the Meta device (with empty weights) and the state dict is then loaded inside it (shard by shard in the case of a sharded checkpoint). This way the maximum RAM used is the full size of the model only.
+
+```py
+from transformers import AutoModelForSeq2SeqLM
+
+t0pp = AutoModelForSeq2SeqLM.from_pretrained("bigscience/T0pp", low_cpu_mem_usage=True)
+```
+
+Moreover, you can directly place the model on different devices if it doesn't fully fit in RAM (only works for inference for now). With `device_map="auto"`, Accelerate will determine where to put each layer to maximize the use of your fastest devices (GPUs) and offload the rest on the CPU, or even the hard drive if you don't have enough GPU RAM (or CPU RAM). Even if the model is split across several devices, it will run as you would normally expect.
+
+When passing a `device_map`, `low_cpu_mem_usage` is automatically set to `True`, so you don't need to specify it:
+
+```py
+from transformers import AutoModelForSeq2SeqLM
+
+t0pp = AutoModelForSeq2SeqLM.from_pretrained("bigscience/T0pp", device_map="auto")
+```
+
+You can inspect how the model was split across devices by looking at its `hf_device_map` attribute:
+
+```py
+t0pp.hf_device_map
+```
+
+```python out
+{'shared': 0,
+ 'decoder.embed_tokens': 0,
+ 'encoder': 0,
+ 'decoder.block.0': 0,
+ 'decoder.block.1': 1,
+ 'decoder.block.2': 1,
+ 'decoder.block.3': 1,
+ 'decoder.block.4': 1,
+ 'decoder.block.5': 1,
+ 'decoder.block.6': 1,
+ 'decoder.block.7': 1,
+ 'decoder.block.8': 1,
+ 'decoder.block.9': 1,
+ 'decoder.block.10': 1,
+ 'decoder.block.11': 1,
+ 'decoder.block.12': 1,
+ 'decoder.block.13': 1,
+ 'decoder.block.14': 1,
+ 'decoder.block.15': 1,
+ 'decoder.block.16': 1,
+ 'decoder.block.17': 1,
+ 'decoder.block.18': 1,
+ 'decoder.block.19': 1,
+ 'decoder.block.20': 1,
+ 'decoder.block.21': 1,
+ 'decoder.block.22': 'cpu',
+ 'decoder.block.23': 'cpu',
+ 'decoder.final_layer_norm': 'cpu',
+ 'decoder.dropout': 'cpu',
+ 'lm_head': 'cpu'}
+```
+
+You can also write your own device map following the same format (a dictionary layer name to device). It should map all parameters of the model to a given device, but you don't have to detail where all the submosules of one layer go if that layer is entirely on the same device. For instance, the following device map would work properly for T0pp (as long as you have the GPU memory):
+
+```python
+device_map = {"shared": 0, "encoder": 0, "decoder": 1, "lm_head": 1}
+```
+
+Another way to minimize the memory impact of your model is to instantiate it at a lower precision dtype (like `torch.float16`).
+
 ### Model Instantiation dtype
 
 Under Pytorch a model normally gets instantiated with `torch.float32` format. This can be an issue if one tries to
@@ -89,3 +158,7 @@ Due to Pytorch design, this functionality is only available for floating dtypes.
 ## Pushing to the Hub
 
 [[autodoc]] utils.PushToHubMixin
+
+## Sharded checkpoints
+
+[[autodoc]] modeling_utils.load_sharded_checkpoint

docs/source/en/main_classes/output.mdx

@@ -136,6 +136,30 @@ documented on their corresponding model page.
 
 [[autodoc]] modeling_outputs.Seq2SeqQuestionAnsweringModelOutput
 
+## SemanticSegmenterOutput
+
+[[autodoc]] modeling_outputs.SemanticSegmenterOutput
+
+## ImageClassifierOutput
+
+[[autodoc]] modeling_outputs.ImageClassifierOutput
+
+## ImageClassifierOutputWithNoAttention
+
+[[autodoc]] modeling_outputs.ImageClassifierOutputWithNoAttention
+
+## DepthEstimatorOutput
+
+[[autodoc]] modeling_outputs.DepthEstimatorOutput
+
+## Wav2Vec2BaseModelOutput
+
+[[autodoc]] modeling_outputs.Wav2Vec2BaseModelOutput
+
+## XVectorOutput
+
+[[autodoc]] modeling_outputs.XVectorOutput
+
 ## TFBaseModelOutput
 
 [[autodoc]] modeling_tf_outputs.TFBaseModelOutput

docs/source/en/main_classes/pipelines.mdx

@@ -38,6 +38,7 @@ There are two categories of pipeline abstractions to be aware about:
   - [`Text2TextGenerationPipeline`]
   - [`TokenClassificationPipeline`]
   - [`TranslationPipeline`]
+  - [`VisualQuestionAnsweringPipeline`]
   - [`ZeroShotClassificationPipeline`]
   - [`ZeroShotImageClassificationPipeline`]
 
@@ -423,6 +424,12 @@ See [`TokenClassificationPipeline`] for all details.
     - __call__
     - all
 
+### VisualQuestionAnsweringPipeline
+
+[[autodoc]] VisualQuestionAnsweringPipeline
+    - __call__
+    - all
+
 ### ZeroShotClassificationPipeline
 
 [[autodoc]] ZeroShotClassificationPipeline

docs/source/en/main_classes/tokenizer.mdx

@@ -18,9 +18,7 @@ Rust library [🤗 Tokenizers](https://github.com/huggingface/tokenizers). The "
 
 1. a significant speed-up in particular when doing batched tokenization and
 2. additional methods to map between the original string (character and words) and the token space (e.g. getting the
-   index of the token comprising a given character or the span of characters corresponding to a given token). Currently
-   no "Fast" implementation is available for the SentencePiece-based tokenizers (for T5, ALBERT, CamemBERT, XLM-RoBERTa
-   and XLNet models).
+   index of the token comprising a given character or the span of characters corresponding to a given token). 
 
 The base classes [`PreTrainedTokenizer`] and [`PreTrainedTokenizerFast`]
 implement the common methods for encoding string inputs in model inputs (see below) and instantiating/saving python and

docs/source/en/main_classes/trainer.mdx

@@ -12,7 +12,7 @@ specific language governing permissions and limitations under the License.
 
 # Trainer
 
-The [`Trainer`] class provides an API for feature-complete training in PyTorch for most standard use cases. It's used in most of the [example scripts](../examples).
+The [`Trainer`] class provides an API for feature-complete training in PyTorch for most standard use cases. It's used in most of the [example scripts](https://github.com/huggingface/transformers/tree/main/examples).
 
 Before instantiating your [`Trainer`], create a [`TrainingArguments`] to access all the points of customization during training.
 
@@ -291,10 +291,10 @@ Also if you do set this environment variable it's the best to set it in your `~/
 The [`Trainer`] has been extended to support libraries that may dramatically improve your training
 time and fit much bigger models.
 
-Currently it supports third party solutions, [DeepSpeed](https://github.com/microsoft/DeepSpeed) and [FairScale](https://github.com/facebookresearch/fairscale/), which implement parts of the paper [ZeRO: Memory Optimizations
+Currently it supports third party solutions, [DeepSpeed](https://github.com/microsoft/DeepSpeed), [PyTorch FSDP](https://pytorch.org/docs/stable/fsdp.html) and [FairScale](https://github.com/facebookresearch/fairscale/), which implement parts of the paper [ZeRO: Memory Optimizations
 Toward Training Trillion Parameter Models, by Samyam Rajbhandari, Jeff Rasley, Olatunji Ruwase, Yuxiong He](https://arxiv.org/abs/1910.02054).
 
-This provided support is new and experimental as of this writing.
+This provided support is new and experimental as of this writing. While the support for DeepSpeed and PyTorch FSDP is active and we welcome issues around it, we don't support the FairScale integration anymore since it has been integrated in PyTorch main (see the [PyTorch FSDP integration](#pytorch-fully-sharded-data-parallel))
 
 <a id='zero-install-notes'></a>
 
@@ -408,6 +408,12 @@ As always make sure to edit the paths in the example to match your situation.
 
 ### FairScale
 
+<Tip warning={true}>
+
+This integration is not supported anymore, we recommend you either use DeepSpeed or PyTorch FSDP.
+
+</Tip>
+
 By integrating [FairScale](https://github.com/facebookresearch/fairscale/) the [`Trainer`]
 provides support for the following features from [the ZeRO paper](https://arxiv.org/abs/1910.02054):
 
@@ -540,6 +546,42 @@ Known caveats:
   `FullyShardedDataParallelism` of fairscale. It should be used with the option `auto_wrap` if you are not
   doing this yourself: `--sharded_ddp "zero_dp_3 auto_wrap"`.
 
+### PyTorch Fully Sharded Data parallel
+
+To accelerate training huge models on larger batch sizes, we can use a fully sharded data parallel model.
+This type of data parallel paradigm enables fitting more data and larger models by sharding the optimizer states, gradients and parameters.
+To read more about it and the benefits, check out the [Fully Sharded Data Parallel blog](https://pytorch.org/blog/introducing-pytorch-fully-sharded-data-parallel-api/).
+We have integrated the latest PyTorch's Fully Sharded Data Parallel (FSDP) training feature.
+All you need to do is enable it through the config.
+
+**Required PyTorch version for FSDP support**: PyTorch Nightly (or 1.12.0 if you read this after it has been released)
+as the model saving with FSDP activated is only available with recent fixes.
+
+**Usage**:
+
+- 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.
+
+- **Sharding Strategy**: 
+  - FULL_SHARD : Shards optimizer states + gradients + model parameters across data parallel workers/GPUs.
+  For this, add `--fsdp full_shard` to the command line arguments. 
+  - SHARD_GRAD_OP : Shards optimizer states + gradients across data parallel workers/GPUs.
+    For this, add `--fsdp shard_grad_op` to the command line arguments.
+- To offload the parameters and gradients to the CPU, 
+add `--fsdp "full_shard offload"` or `--fsdp "shard_grad_op offload"` to the command line arguments.
+-  To automatically recursively wrap layers with FSDP using `default_auto_wrap_policy`, 
+add `--fsdp "full_shard auto_wrap"` or `--fsdp "shard_grad_op auto_wrap"` to the command line arguments.
+- To enable both CPU offloading and auto wrapping, 
+add `--fsdp "full_shard offload auto_wrap"` or `--fsdp "shard_grad_op offload auto_wrap"` to the command line arguments.
+- If auto wrapping is enabled, please add `--fsdp_min_num_params <number>` to command line arguments.
+It specifies FSDP's minimum number of parameters for Default Auto Wrapping.
+
+**Few caveats to be aware of**
+- Mixed precision is currently not supported with FSDP as we wait for PyTorch to fix support for it.
+More details in this [issues](https://github.com/pytorch/pytorch/issues/75676).
+- FSDP currently doesn't support multiple parameter groups. 
+More details mentioned in this [issue](https://github.com/pytorch/pytorch/issues/76501)
+(`The original model parameters' .grads are not set, meaning that they cannot be optimized separately (which is why we cannot support multiple parameter groups)`).
 
 Sections that were moved:
 

docs/source/en/model_doc/auto.mdx

@@ -122,6 +122,10 @@ Likewise, if your `NewModel` is a subclass of [`PreTrainedModel`], make sure its
 
 [[autodoc]] AutoModelForVision2Seq
 
+## AutoModelForVisualQuestionAnswering
+
+[[autodoc]] AutoModelForVisualQuestionAnswering
+
 ## AutoModelForAudioClassification
 
 [[autodoc]] AutoModelForAudioClassification
@@ -194,6 +198,10 @@ Likewise, if your `NewModel` is a subclass of [`PreTrainedModel`], make sure its
 
 [[autodoc]] TFAutoModelForMultipleChoice
 
+## TFAutoModelForNextSentencePrediction
+
+[[autodoc]] TFAutoModelForNextSentencePrediction
+
 ## TFAutoModelForTableQuestionAnswering
 
 [[autodoc]] TFAutoModelForTableQuestionAnswering

docs/source/en/model_doc/bert-generation.mdx

@@ -49,7 +49,7 @@ Usage:
 
 >>> input_ids = tokenizer(
 ...     "This is a long article to summarize", add_special_tokens=False, return_tensors="pt"
->>> ).input_ids
+... ).input_ids
 >>> labels = tokenizer("This is a short summary", return_tensors="pt").input_ids
 
 >>> # train...
@@ -67,7 +67,7 @@ Usage:
 
 >>> input_ids = tokenizer(
 ...     "This is the first sentence. This is the second sentence.", add_special_tokens=False, return_tensors="pt"
->>> ).input_ids
+... ).input_ids
 
 >>> outputs = sentence_fuser.generate(input_ids)
 

docs/source/en/model_doc/bert.mdx

@@ -58,6 +58,10 @@ This model was contributed by [thomwolf](https://huggingface.co/thomwolf). The o
 
 [[autodoc]] BertTokenizerFast
 
+## TFBertTokenizer
+
+[[autodoc]] TFBertTokenizer
+
 ## Bert specific outputs
 
 [[autodoc]] models.bert.modeling_bert.BertForPreTrainingOutput
@@ -166,6 +170,11 @@ This model was contributed by [thomwolf](https://huggingface.co/thomwolf). The o
 [[autodoc]] FlaxBertForPreTraining
     - __call__
 
+## FlaxBertForCausalLM
+
+[[autodoc]] FlaxBertForCausalLM
+    - __call__
+
 ## FlaxBertForMaskedLM
 
 [[autodoc]] FlaxBertForMaskedLM

docs/source/en/model_doc/big_bird.mdx

@@ -120,6 +120,11 @@ This model was contributed by [vasudevgupta](https://huggingface.co/vasudevgupta
 [[autodoc]] FlaxBigBirdForPreTraining
     - __call__
 
+## FlaxBigBirdForCausalLM
+
+[[autodoc]] FlaxBigBirdForCausalLM
+    - __call__
+
 ## FlaxBigBirdForMaskedLM
 
 [[autodoc]] FlaxBigBirdForMaskedLM

docs/source/en/model_doc/bloom.mdx

@@ -0,0 +1,57 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# BLOOM
+
+## Overview
+
+The BLOOM model has been proposed with its various versions through the [BigScience Workshop](https://bigscience.huggingface.co/). BigScience is inspired by other open science initiatives where researchers have pooled their time and resources to collectively achieve a higher impact.
+The architecture of BLOOM is essentially similar to GPT3 (auto-regressive model for next token prediction), but has been trained on different 46 languages including code.
+Several smaller versions of the models have been trained on the same dataset. BLOOM is available in the following versions:
+
+- [bloom-350m](https://huggingface.co/bigscience/bloom-350m)
+- [bloom-760m](https://huggingface.co/bigscience/bloom-760m)
+- [bloom-1b3](https://huggingface.co/bigscience/bloom-1b3)
+- [bloom-2b5](https://huggingface.co/bigscience/bloom-2b5)
+- [bloom-6b3](https://huggingface.co/bigscience/bloom-6b3)
+- [bloom](https://huggingface.co/bigscience/bloom) (176B parameters)
+
+
+## BloomConfig
+
+[[autodoc]] BloomConfig
+    - all
+
+## BloomModel
+
+[[autodoc]] BloomModel
+    - forward
+
+## BloomTokenizerFast
+
+[[autodoc]] BloomTokenizerFast
+    - all
+
+## BloomForCausalLM
+
+[[autodoc]] BloomForCausalLM
+    - forward
+
+## BloomForSequenceClassification
+
+[[autodoc]] BloomForSequenceClassification
+    - forward
+
+## BloomForTokenClassification
+
+[[autodoc]] BloomForTokenClassification
+    - forward

docs/source/en/model_doc/codegen.mdx

@@ -0,0 +1,81 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# CodeGen
+
+## Overview
+
+The CodeGen model was proposed in [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, and Caiming Xiong.
+
+CodeGen is an autoregressive language model for program synthesis trained sequentially on [The Pile](https://pile.eleuther.ai/), BigQuery, and BigPython.
+
+The abstract from the paper is the following:
+
+*Program synthesis strives to generate a computer program as a solution to a given problem specification. We propose a conversational program synthesis approach via large language models, which addresses the challenges of searching over a vast program space and user intent specification faced in prior approaches. Our new approach casts the process of writing a specification and program as a multi-turn conversation between a user and a system. It treats program synthesis as a sequence prediction problem, in which the specification is expressed in natural language and the desired program is conditionally sampled. We train a family of large language models, called CodeGen, on natural language and programming language data. With weak supervision in the data and the scaling up of data size and model size, conversational capacities emerge from the simple autoregressive language modeling. To study the model behavior on conversational program synthesis, we develop a multi-turn programming benchmark (MTPB), where solving each problem requires multi-step synthesis via multi-turn conversation between the user and the model. Our findings show the emergence of conversational capabilities and the effectiveness of the proposed conversational program synthesis paradigm. In addition, our model CodeGen (with up to 16B parameters trained on TPU-v4) outperforms OpenAI's Codex on the HumanEval benchmark. We make the training library JaxFormer including checkpoints available as open source contribution: [this https URL](https://github.com/salesforce/codegen).* 
+
+This model was contributed by [Hiroaki Hayashi](https://huggingface.co/rooa).
+The original code can be found [here](https://github.com/salesforce/codegen).
+
+## Checkpoint Naming
+
+* CodeGen model [checkpoints](https://huggingface.co/models?other=codegen) are available on different pre-training data with variable sizes.
+* The format is: `Salesforce/codegen-{size}-{data}`, where
+  * `size`: `350M`, `2B`, `6B`, `16B`
+  * `data`: 
+    * `nl`: Pre-trained on the Pile
+    * `multi`: Initialized with `nl`, then further pre-trained on multiple programming languages data
+    * `mono`: Initialized with `multi`, then further pre-trained on Python data
+* For example, `Salesforce/codegen-350M-mono` offers a 350 million-parameter checkpoint pre-trained sequentially on the Pile, multiple programming languages, and Python.
+
+## How to use
+
+```python
+>>> from transformers import AutoModelForCausalLM, AutoTokenizer
+
+>>> checkpoint = "Salesforce/codegen-350M-mono"
+>>> model = AutoModelForCausalLM.from_pretrained(checkpoint)
+>>> tokenizer = AutoTokenizer.from_pretrained(checkpoint)
+
+>>> text = "def hello_world():"
+
+>>> completion = model.generate(**tokenizer(text, return_tensors="pt"))
+
+>>> print(tokenizer.decode(completion[0]))
+def hello_world():
+    print("Hello World")
+
+hello_world()
+```
+
+## CodeGenConfig
+
+[[autodoc]] CodeGenConfig
+    - all
+
+## CodeGenTokenizer
+
+[[autodoc]] CodeGenTokenizer
+    - save_vocabulary
+
+## CodeGenTokenizerFast
+
+[[autodoc]] CodeGenTokenizerFast
+
+## CodeGenModel
+
+[[autodoc]] CodeGenModel
+    - forward
+
+## CodeGenForCausalLM
+
+[[autodoc]] CodeGenForCausalLM
+    - forward

docs/source/en/model_doc/cpm.mdx

@@ -38,3 +38,7 @@ Note: We only have a tokenizer here, since the model architecture is the same as
 ## CpmTokenizer
 
 [[autodoc]] CpmTokenizer
+
+## CpmTokenizerFast
+
+[[autodoc]] CpmTokenizerFast

docs/source/en/model_doc/cvt.mdx

@@ -0,0 +1,53 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# Convolutional Vision Transformer (CvT)
+
+## Overview
+
+The CvT model was proposed in [CvT: Introducing Convolutions to Vision Transformers](https://arxiv.org/abs/2103.15808) by Haiping Wu, Bin Xiao, Noel Codella, Mengchen Liu, Xiyang Dai, Lu Yuan and Lei Zhang. The Convolutional vision Transformer (CvT) improves the [Vision Transformer (ViT)](vit) in performance and efficiency by introducing convolutions into ViT to yield the best of both designs.
+
+The abstract from the paper is the following:
+
+*We present in this paper a new architecture, named Convolutional vision Transformer (CvT), that improves Vision Transformer (ViT) 
+in performance and efficiency by introducing convolutions into ViT to yield the best of both designs. This is accomplished through 
+two primary modifications: a hierarchy of Transformers containing a new convolutional token embedding, and a convolutional Transformer 
+block leveraging a convolutional projection. These changes introduce desirable properties of convolutional neural networks (CNNs) 
+to the ViT architecture (\ie shift, scale, and distortion invariance) while maintaining the merits of Transformers (\ie dynamic attention, 
+global context, and better generalization). We validate CvT by conducting extensive experiments, showing that this approach achieves 
+state-of-the-art performance over other Vision Transformers and ResNets on ImageNet-1k, with fewer parameters and lower FLOPs. In addition, 
+performance gains are maintained when pretrained on larger datasets (\eg ImageNet-22k) and fine-tuned to downstream tasks. Pre-trained on 
+ImageNet-22k, our CvT-W24 obtains a top-1 accuracy of 87.7\% on the ImageNet-1k val set. Finally, our results show that the positional encoding, 
+a crucial component in existing Vision Transformers, can be safely removed in our model, simplifying the design for higher resolution vision tasks.*
+
+Tips:
+
+- CvT models are regular Vision Transformers, but trained with convolutions. They outperform the [original model (ViT)](vit) when fine-tuned on ImageNet-1K and CIFAR-100.
+- You can check out demo notebooks regarding inference as well as fine-tuning on custom data [here](https://github.com/NielsRogge/Transformers-Tutorials/tree/master/VisionTransformer) (you can just replace [`ViTFeatureExtractor`] by [`AutoFeatureExtractor`] and [`ViTForImageClassification`] by [`CvtForImageClassification`]).
+- The available checkpoints are either (1) pre-trained on [ImageNet-22k](http://www.image-net.org/) (a collection of 14 million images and 22k classes) only, (2) also fine-tuned on ImageNet-22k or (3) also fine-tuned on [ImageNet-1k](http://www.image-net.org/challenges/LSVRC/2012/) (also referred to as ILSVRC 2012, a collection of 1.3 million
+  images and 1,000 classes).
+
+This model was contributed by [anugunj](https://huggingface.co/anugunj). The original code can be found [here](https://github.com/microsoft/CvT).
+
+## CvtConfig
+
+[[autodoc]] CvtConfig
+
+## CvtModel
+
+[[autodoc]] CvtModel
+    - forward
+
+## CvtForImageClassification
+
+[[autodoc]] CvtForImageClassification
+    - forward

docs/source/en/model_doc/data2vec.mdx

@@ -33,11 +33,19 @@ Models and code are available at www.github.com/pytorch/fairseq/tree/master/exam
 
 Tips:
 
-- Both Data2VecAudio and Data2VecText have been trained using the same self-supervised learning method.
-  In the case of Data2VecAudio, preprocessing is identical to [`RobertaModel`], including tokenization.
+- Data2VecAudio, Data2VecText, and Data2VecVision have all been trained using the same self-supervised learning method.
+- For Data2VecAudio, preprocessing is identical to [`Wav2Vec2Model`], including feature extraction
+- For Data2VecText, preprocessing is identical to [`RobertaModel`], including tokenization.
+- For Data2VecVision, preprocessing is identical to [`BeitModel`], including feature extraction.
+- To know how a pre-trained Data2Vec vision model can be fine-tuned on the task of image classification, you can check out
+[this notebook](https://colab.research.google.com/github/sayakpaul/TF-2.0-Hacks/blob/master/data2vec_vision_image_classification.ipynb).
 
-This model was contributed by [edugp](https://huggingface.co/edugp).
-The original code can be found [here](https://github.com/pytorch/fairseq/tree/main/examples/data2vec).
+
+This model was contributed by [edugp](https://huggingface.co/edugp) and [patrickvonplaten](https://huggingface.co/patrickvonplaten).
+[sayakpaul](https://github.com/sayakpaul) and [Rocketknight1](https://github.com/Rocketknight1) contributed Data2Vec for vision in TensorFlow.
+
+The original code (for NLP and Speech) can be found [here](https://github.com/pytorch/fairseq/tree/main/examples/data2vec).
+The original code for vision can be found [here](https://github.com/facebookresearch/data2vec_vision/tree/main/beit).
 
 
 ## Data2VecTextConfig
@@ -48,12 +56,16 @@ The original code can be found [here](https://github.com/pytorch/fairseq/tree/ma
 
 [[autodoc]] Data2VecAudioConfig
 
+## Data2VecVisionConfig
+
+[[autodoc]] Data2VecVisionConfig
+
+
 ## Data2VecAudioModel
 
 [[autodoc]] Data2VecAudioModel
     - forward
 
-
 ## Data2VecAudioForAudioFrameClassification
 
 [[autodoc]] Data2VecAudioForAudioFrameClassification
@@ -108,3 +120,33 @@ The original code can be found [here](https://github.com/pytorch/fairseq/tree/ma
 
 [[autodoc]] Data2VecTextForQuestionAnswering
     - forward
+
+## Data2VecVisionModel
+
+[[autodoc]] Data2VecVisionModel
+    - forward
+
+## Data2VecVisionForImageClassification
+
+[[autodoc]] Data2VecVisionForImageClassification
+    - forward
+
+## Data2VecVisionForSemanticSegmentation
+
+[[autodoc]] Data2VecVisionForSemanticSegmentation
+    - forward
+
+## TFData2VecVisionModel
+
+[[autodoc]] TFData2VecVisionModel
+    - call
+
+## TFData2VecVisionForImageClassification
+
+[[autodoc]] TFData2VecVisionForImageClassification
+    - call
+
+## TFData2VecVisionForSemanticSegmentation
+
+[[autodoc]] TFData2VecVisionForSemanticSegmentation
+    - call

docs/source/en/model_doc/deberta-v2.mdx

@@ -71,6 +71,12 @@ contributed by [kamalkraj](https://huggingface.co/kamalkraj). The original code
     - create_token_type_ids_from_sequences
     - save_vocabulary
 
+## DebertaV2TokenizerFast
+
+[[autodoc]] DebertaV2TokenizerFast
+    - build_inputs_with_special_tokens
+    - create_token_type_ids_from_sequences
+
 ## DebertaV2Model
 
 [[autodoc]] DebertaV2Model
@@ -101,6 +107,11 @@ contributed by [kamalkraj](https://huggingface.co/kamalkraj). The original code
 [[autodoc]] DebertaV2ForQuestionAnswering
     - forward
 
+## DebertaV2ForMultipleChoice
+
+[[autodoc]] DebertaV2ForMultipleChoice
+    - forward
+
 ## TFDebertaV2Model
 
 [[autodoc]] TFDebertaV2Model

docs/source/en/model_doc/deit.mdx

@@ -69,7 +69,7 @@ Tips:
   *facebook/deit-base-patch16-384*. Note that one should use [`DeiTFeatureExtractor`] in order to
   prepare images for the model.
 
-This model was contributed by [nielsr](https://huggingface.co/nielsr).
+This model was contributed by [nielsr](https://huggingface.co/nielsr). The TensorFlow version of this model was added by [amyeroberts](https://huggingface.co/amyeroberts).
 
 
 ## DeiTConfig
@@ -100,3 +100,23 @@ This model was contributed by [nielsr](https://huggingface.co/nielsr).
 
 [[autodoc]] DeiTForImageClassificationWithTeacher
     - forward
+
+## TFDeiTModel
+
+[[autodoc]] TFDeiTModel
+    - call
+
+## TFDeiTForMaskedImageModeling
+
+[[autodoc]] TFDeiTForMaskedImageModeling
+    - call
+
+## TFDeiTForImageClassification
+
+[[autodoc]] TFDeiTForImageClassification
+    - call
+
+## TFDeiTForImageClassificationWithTeacher
+
+[[autodoc]] TFDeiTForImageClassificationWithTeacher
+    - call

docs/source/en/model_doc/detr.mdx

@@ -113,6 +113,28 @@ Tips:
 - The size of the images will determine the amount of memory being used, and will thus determine the `batch_size`.
   It is advised to use a batch size of 2 per GPU. See [this Github thread](https://github.com/facebookresearch/detr/issues/150) for more info.
 
+There are three ways to instantiate a DETR model (depending on what you prefer):
+  
+Option 1: Instantiate DETR with pre-trained weights for entire model
+```py
+>>> from transformers import DetrForObjectDetection
+
+>>> model = DetrForObjectDetection.from_pretrained("facebook/resnet-50")
+```
+
+Option 2: Instantiate DETR with randomly initialized weights for Transformer, but pre-trained weights for backbone
+```py
+>>> from transformers import DetrConfig, DetrForObjectDetection
+
+>>> config = DetrConfig()
+>>> model = DetrForObjectDetection(config)
+```
+Option 3: Instantiate DETR with randomly initialized weights for backbone + Transformer
+```py
+>>> config = DetrConfig(use_pretrained_backbone=False)
+>>> model = DetrForObjectDetection(config)
+```
+
 As a summary, consider the following table:
 
 | Task | Object detection | Instance segmentation | Panoptic segmentation |
@@ -166,4 +188,4 @@ mean Average Precision (mAP) and Panoptic Quality (PQ). The latter objects are i
 ## DetrForSegmentation
 
 [[autodoc]] DetrForSegmentation
-    - forward
+    - forward

docs/source/en/model_doc/electra.mdx

@@ -158,6 +158,11 @@ This model was contributed by [lysandre](https://huggingface.co/lysandre). The o
 [[autodoc]] FlaxElectraForPreTraining
     - __call__
 
+## FlaxElectraForCausalLM
+
+[[autodoc]] FlaxElectraForCausalLM
+    - __call__
+
 ## FlaxElectraForMaskedLM
 
 [[autodoc]] FlaxElectraForMaskedLM

docs/source/en/model_doc/encoder-decoder.mdx

@@ -12,6 +12,8 @@ specific language governing permissions and limitations under the License.
 
 # Encoder Decoder Models
 
+## Overview
+
 The [`EncoderDecoderModel`] can be used to initialize a sequence-to-sequence model with any
 pretrained autoencoding model as the encoder and any pretrained autoregressive model as the decoder.
 
@@ -25,15 +27,77 @@ any other models (see the examples for more information).
 An application of this architecture could be to leverage two pretrained [`BertModel`] as the encoder
 and decoder for a summarization model as was shown in: [Text Summarization with Pretrained Encoders](https://arxiv.org/abs/1908.08345) by Yang Liu and Mirella Lapata.
 
-The [`~TFEncoderDecoderModel.from_pretrained`] currently doesn't support initializing the model from a
+## Randomly initializing `EncoderDecoderModel` from model configurations.
+
+[`EncoderDecoderModel`] can be randomly initialized from an encoder and a decoder config. In the following example, we show how to do this using the default [`BertModel`] configuration for the encoder and the default [`BertForCausalLM`] configuration for the decoder.
+
+```python
+>>> from transformers import BertConfig, EncoderDecoderConfig, EncoderDecoderModel
+
+>>> config_encoder = BertConfig()
+>>> config_decoder = BertConfig()
+
+>>> config = EncoderDecoderConfig.from_encoder_decoder_configs(config_encoder, config_decoder)
+>>> model = EncoderDecoderModel(config=config)
+```
+
+## Initialising `EncoderDecoderModel` from a pretrained encoder and a pretrained decoder.
+
+[`EncoderDecoderModel`] can be initialized from a pretrained encoder checkpoint and a pretrained decoder checkpoint. Note that any pretrained auto-encoding model, *e.g.* BERT, can serve as the encoder and both pretrained auto-encoding models, *e.g.* BERT, pretrained causal language models, *e.g.* GPT2, as well as the pretrained decoder part of sequence-to-sequence models, *e.g.* decoder of BART, can be used as the decoder.
+Depending on which architecture you choose as the decoder, the cross-attention layers might be randomly initialized.
+Initializing [`EncoderDecoderModel`] from a pretrained encoder and decoder checkpoint requires the model to be fine-tuned on a downstream task, as has been shown in [the *Warm-starting-encoder-decoder blog post*](https://huggingface.co/blog/warm-starting-encoder-decoder).
+To do so, the `EncoderDecoderModel` class provides a [`EncoderDecoderModel.from_encoder_decoder_pretrained`] method.
+
+```python
+>>> from transformers import EncoderDecoderModel, BertTokenizer
+
+>>> tokenizer = BertTokenizer.from_pretrained("bert-base-uncased")
+>>> model = EncoderDecoderModel.from_encoder_decoder_pretrained("bert-base-uncased", "bert-base-uncased")
+```
+
+## Loading an existing `EncoderDecoderModel` checkpoint and perform inference.
+
+To load fine-tuned checkpoints of the `EncoderDecoderModel` class, [`EncoderDecoderModel`] provides the `from_pretrained(...)` method just like any other model architecture in Transformers.
+
+To perform inference, one uses the [`generate`] method, which allows to autoregressively generate text. This method supports various forms of decoding, such as greedy, beam search and multinomial sampling.
+
+```python
+>>> from transformers import AutoTokenizer, EncoderDecoderModel
+
+>>> # load a fine-tuned seq2seq model and corresponding tokenizer
+>>> model = EncoderDecoderModel.from_pretrained("patrickvonplaten/bert2bert_cnn_daily_mail")
+>>> tokenizer = AutoTokenizer.from_pretrained("patrickvonplaten/bert2bert_cnn_daily_mail")
+
+>>> # let's perform inference on a long piece of text
+>>> ARTICLE_TO_SUMMARIZE = (
+...     "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."
+... )
+>>> input_ids = tokenizer(ARTICLE_TO_SUMMARIZE, return_tensors="pt").input_ids
+
+>>> # autoregressively generate summary (uses greedy decoding by default)
+>>> generated_ids = model.generate(input_ids)
+>>> generated_text = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
+>>> print(generated_text)
+nearly 800 thousand customers were affected by the shutoffs. the aim is to reduce the risk of wildfires. nearly 800, 000 customers were expected to be affected by high winds amid dry conditions. pg & e said it scheduled the blackouts to last through at least midday tomorrow.
+```
+
+## Loading a PyTorch checkpoint into `TFEncoderDecoderModel`.
+
+[`TFEncoderDecoderModel.from_pretrained`] currently doesn't support initializing the model from a
 pytorch checkpoint. Passing `from_pt=True` to this method will throw an exception. If there are only pytorch
 checkpoints for a particular encoder-decoder model, a workaround is:
 
 ```python
 >>> # a workaround to load from pytorch checkpoint
+>>> from transformers import EncoderDecoderModel, TFEncoderDecoderModel
+
 >>> _model = EncoderDecoderModel.from_pretrained("patrickvonplaten/bert2bert-cnn_dailymail-fp16")
+
 >>> _model.encoder.save_pretrained("./encoder")
 >>> _model.decoder.save_pretrained("./decoder")
+
 >>> model = TFEncoderDecoderModel.from_encoder_decoder_pretrained(
 ...     "./encoder", "./decoder", encoder_from_pt=True, decoder_from_pt=True
 ... )
@@ -41,6 +105,38 @@ checkpoints for a particular encoder-decoder model, a workaround is:
 >>> model.config = _model.config
 ```
 
+## Training
+
+Once the model is created, it can be fine-tuned similar to BART, T5 or any other encoder-decoder model.
+As you can see, only 2 inputs are required for the model in order to compute a loss: `input_ids` (which are the
+`input_ids` of the encoded input sequence) and `labels` (which are the `input_ids` of the encoded
+target sequence).
+
+```python
+>>> from transformers import BertTokenizer, EncoderDecoderModel
+
+>>> tokenizer = BertTokenizer.from_pretrained("bert-base-uncased")
+>>> model = EncoderDecoderModel.from_encoder_decoder_pretrained("bert-base-uncased", "bert-base-uncased")
+
+>>> model.config.decoder_start_token_id = tokenizer.cls_token_id
+>>> model.config.pad_token_id = tokenizer.pad_token_id
+
+>>> input_ids = tokenizer(
+...     "The tower is 324 metres (1,063 ft) tall, about the same height as an 81-storey building, and the tallest structure in Paris. Its base is square, measuring 125 metres (410 ft) on each side.During its construction, the Eiffel Tower surpassed the Washington Monument to become the tallest man-made structure in the world, 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, it is now taller than the Chrysler Building by 5.2 metres (17 ft).Excluding transmitters, the Eiffel Tower is the second tallest free-standing structure in France after the Millau Viaduct.",
+...     return_tensors="pt",
+... ).input_ids
+
+>>> labels = tokenizer(
+...     "the eiffel tower surpassed the washington monument to become the tallest structure in the world. it was the first structure to reach a height of 300 metres in paris in 1930. it is now taller than the chrysler building by 5. 2 metres ( 17 ft ) and is the second tallest free - standing structure in paris.",
+...     return_tensors="pt",
+... ).input_ids
+
+>>> # the forward function automatically creates the correct decoder_input_ids
+>>> loss = model(input_ids=input_ids, labels=labels).loss
+```
+
+Detailed [colab](https://colab.research.google.com/drive/1WIk2bxglElfZewOHboPFNj8H44_VAyKE?usp=sharing#scrollTo=ZwQIEhKOrJpl) for training.
+
 This model was contributed by [thomwolf](https://github.com/thomwolf). This model's TensorFlow and Flax versions
 were contributed by [ydshieh](https://github.com/ydshieh).
 

docs/source/en/model_doc/flava.mdx

@@ -0,0 +1,96 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# FLAVA
+
+## Overview
+
+The FLAVA model was proposed in [FLAVA: A Foundational Language And Vision Alignment Model](https://arxiv.org/abs/2112.04482) by Amanpreet Singh, Ronghang Hu, Vedanuj Goswami, Guillaume Couairon, Wojciech Galuba, Marcus Rohrbach, and Douwe Kiela and is accepted at CVPR 2022.
+
+The paper aims at creating a single unified foundation model which can work across vision, language 
+as well as vision-and-language multimodal tasks.
+
+The abstract from the paper is the following:
+
+*State-of-the-art vision and vision-and-language models rely on large-scale visio-linguistic pretraining for obtaining good performance on a variety 
+of downstream tasks. Generally, such models are often either cross-modal (contrastive) or multi-modal 
+(with earlier fusion) but not both; and they often only target specific modalities or tasks. A promising 
+direction would be to use a single holistic universal model, as a "foundation", that targets all modalities 
+at once -- a true vision and language foundation model should be good at vision tasks, language tasks, and 
+cross- and multi-modal vision and language tasks. We introduce FLAVA as such a model and demonstrate 
+impressive performance on a wide range of 35 tasks spanning these target modalities.*
+
+
+This model was contributed by [aps](https://huggingface.co/aps). The original code can be found [here](https://github.com/facebookresearch/multimodal/tree/main/examples/flava).
+
+
+## FlavaConfig
+
+[[autodoc]] FlavaConfig
+
+## FlavaTextConfig
+
+[[autodoc]] FlavaTextConfig
+
+## FlavaImageConfig
+
+[[autodoc]] FlavaImageConfig
+
+## FlavaMultimodalConfig
+
+[[autodoc]] FlavaMultimodalConfig
+
+## FlavaImageCodebookConfig
+
+[[autodoc]] FlavaImageCodebookConfig
+
+## FlavaProcessor
+
+[[autodoc]] FlavaProcessor
+
+## FlavaFeatureExtractor
+
+[[autodoc]] FlavaFeatureExtractor
+
+## FlavaForPreTraining
+
+[[autodoc]] FlavaForPreTraining
+    - forward
+
+## FlavaModel
+
+[[autodoc]] FlavaModel
+    - forward
+    - get_text_features
+    - get_image_features
+
+## FlavaImageCodebook
+
+[[autodoc]] FlavaImageCodebook
+    - forward
+    - get_codebook_indices
+    - get_codebook_probs
+
+## FlavaTextModel
+
+[[autodoc]] FlavaTextModel
+    - forward
+
+## FlavaImageModel
+
+[[autodoc]] FlavaImageModel
+    - forward
+
+## FlavaMultimodalModel
+
+[[autodoc]] FlavaMultimodalModel
+    - forward

docs/source/en/model_doc/gpt_neox.mdx

@@ -0,0 +1,76 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# GPT-NeoX
+
+## Overview
+
+We introduce GPT-NeoX-20B, a 20 billion parameter autoregressive language model trained on the Pile, whose weights will
+be made freely and openly available to the public through a permissive license. It is, to the best of our knowledge,
+the largest dense autoregressive model that has publicly available weights at the time of submission. In this work,
+we describe GPT-NeoX-20B's architecture and training and evaluate its performance on a range of language-understanding,
+mathematics, and knowledge-based tasks. We find that GPT-NeoX-20B is a particularly powerful few-shot reasoner and
+gains far more in performance when evaluated five-shot than similarly sized GPT-3 and FairSeq models. We open-source
+the training and evaluation code, as well as the model weights, at [https://github.com/EleutherAI/gpt-neox](https://github.com/EleutherAI/gpt-neox).
+
+Development of the model was led by Sid Black, Stella Biderman and Eric Hallahan, and the model was trained with
+generous the support of [CoreWeave](https://www.coreweave.com/).
+
+GPT-NeoX-20B was trained with fp16, thus it is recommended to initialize the model as follows:
+
+```python
+model = GPTNeoXForCausalLM.from_pretrained("EleutherAI/gpt-neox-20b").half().cuda()
+```
+
+GPT-NeoX-20B also has a different tokenizer from the one used in GPT-J-6B and GPT-Neo. The new tokenizer allocates
+additional tokens to whitespace characters, making the model more suitable for certain tasks like code generation.
+
+### Generation
+
+The `generate()` method can be used to generate text using GPT Neo model.
+
+```python
+>>> from transformers import GPTNeoXForCausalLM, GPTNeoXTokenizerFast
+
+>>> model = GPTNeoXForCausalLM.from_pretrained("EleutherAI/gpt-neox-20b")
+>>> tokenizer = GPTNeoXTokenizerFast.from_pretrained("EleutherAI/gpt-neox-20b")
+
+>>> prompt = "GPTNeoX20B is a 20B-parameter autoregressive Transformer model developed by EleutherAI."
+
+>>> input_ids = tokenizer(prompt, return_tensors="pt").input_ids
+
+>>> gen_tokens = model.generate(
+...     input_ids,
+...     do_sample=True,
+...     temperature=0.9,
+...     max_length=100,
+... )
+>>> gen_text = tokenizer.batch_decode(gen_tokens)[0]
+```
+
+## GPTNeoXConfig
+
+[[autodoc]] GPTNeoXConfig
+
+## GPTNeoXTokenizerFast
+
+[[autodoc]] GPTNeoXTokenizerFast
+
+## GPTNeoXModel
+
+[[autodoc]] GPTNeoXModel
+    - forward
+
+## GPTNeoXForCausalLM
+
+[[autodoc]] GPTNeoXForCausalLM
+    - forward

docs/source/en/model_doc/groupvit.mdx

@@ -0,0 +1,61 @@
+<!--Copyright 2022 NVIDIA and 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.
+-->
+
+# GroupViT
+
+## Overview
+
+The GroupViT model was proposed in [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) by Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang.
+Inspired by [CLIP](clip), GroupViT is a vision-language model that can perform zero-shot semantic segmentation on any given vocabulary categories.
+
+The abstract from the paper is the following:
+
+*Grouping and recognition are important components of visual scene understanding, e.g., for object detection and semantic segmentation. With end-to-end deep learning systems, grouping of image regions usually happens implicitly via top-down supervision from pixel-level recognition labels. Instead, in this paper, we propose to bring back the grouping mechanism into deep networks, which allows semantic segments to emerge automatically with only text supervision. We propose a hierarchical Grouping Vision Transformer (GroupViT), which goes beyond the regular grid structure representation and learns to group image regions into progressively larger arbitrary-shaped segments. We train GroupViT jointly with a text encoder on a large-scale image-text dataset via contrastive losses. With only text supervision and without any pixel-level annotations, GroupViT learns to group together semantic regions and successfully transfers to the task of semantic segmentation in a zero-shot manner, i.e., without any further fine-tuning. It achieves a zero-shot accuracy of 52.3% mIoU on the PASCAL VOC 2012 and 22.4% mIoU on PASCAL Context datasets, and performs competitively to state-of-the-art transfer-learning methods requiring greater levels of supervision.*
+
+Tips:
+
+- You may specify `output_segmentation=True` in the forward of `GroupViTModel` to get the segmentation logits of input texts. 
+- The quickest way to get started with GroupViT is by checking the [example notebooks](https://github.com/xvjiarui/GroupViT/blob/main/demo/GroupViT_hf_inference_notebook.ipynb) (which showcase zero-shot segmentation inference). One can also check out the [HuggingFace Spaces demo](https://huggingface.co/spaces/xvjiarui/GroupViT) to play with GroupViT. 
+
+This model was contributed by [xvjiarui](https://huggingface.co/xvjiarui).
+The original code can be found [here](https://github.com/NVlabs/GroupViT).
+
+
+## GroupViTConfig
+
+[[autodoc]] GroupViTConfig
+    - from_text_vision_configs
+
+## GroupViTTextConfig
+
+[[autodoc]] GroupViTTextConfig
+
+## GroupViTVisionConfig
+
+[[autodoc]] GroupViTVisionConfig
+
+## GroupViTModel
+
+[[autodoc]] GroupViTModel
+    - forward
+    - get_text_features
+    - get_image_features
+
+## GroupViTTextModel
+
+[[autodoc]] GroupViTTextModel
+    - forward
+
+## GroupViTVisionModel
+
+[[autodoc]] GroupViTVisionModel
+    - forward

docs/source/en/model_doc/layoutlmv2.mdx

@@ -44,6 +44,14 @@ including FUNSD (0.7895 -> 0.8420), CORD (0.9493 -> 0.9601), SROIE (0.9524 -> 0.
 RVL-CDIP (0.9443 -> 0.9564), and DocVQA (0.7295 -> 0.8672). The pre-trained LayoutLMv2 model is publicly available at
 this https URL.*
 
+LayoutLMv2 depends on `detectron2`, `torchvision` and `tesseract`. Run the
+following to install them: 
+```
+python -m pip install 'git+https://github.com/facebookresearch/detectron2.git'
+python -m pip install torchvision tesseract
+```
+(If you are developing for LayoutLMv2, note that passing the doctests also requires the installation of these packages.)
+
 Tips:
 
 - The main difference between LayoutLMv1 and LayoutLMv2 is that the latter incorporates visual embeddings during

docs/source/en/model_doc/layoutlmv3.mdx

@@ -0,0 +1,85 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# LayoutLMv3
+
+## Overview
+
+The LayoutLMv3 model was proposed in [LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking](https://arxiv.org/abs/2204.08387) by Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, Furu Wei.
+LayoutLMv3 simplifies [LayoutLMv2](layoutlmv2) by using patch embeddings (as in [ViT](vit)) instead of leveraging a CNN backbone, and pre-trains the model on 3 objectives: masked language modeling (MLM), masked image modeling (MIM)
+and word-patch alignment (WPA).
+
+The abstract from the paper is the following:
+
+*Self-supervised pre-training techniques have achieved remarkable progress in Document AI. Most multimodal pre-trained models use a masked language modeling objective to learn bidirectional representations on the text modality, but they differ in pre-training objectives for the image modality. This discrepancy adds difficulty to multimodal representation learning. In this paper, we propose LayoutLMv3 to pre-train multimodal Transformers for Document AI with unified text and image masking. Additionally, LayoutLMv3 is pre-trained with a word-patch alignment objective to learn cross-modal alignment by predicting whether the corresponding image patch of a text word is masked. The simple unified architecture and training objectives make LayoutLMv3 a general-purpose pre-trained model for both text-centric and image-centric Document AI tasks. Experimental results show that LayoutLMv3 achieves state-of-the-art performance not only in text-centric tasks, including form understanding, receipt understanding, and document visual question answering, but also in image-centric tasks such as document image classification and document layout analysis.*
+
+Tips:
+
+- In terms of data processing, LayoutLMv3 is identical to its predecessor [LayoutLMv2](layoutlmv2), except that:
+    - images need to be resized and normalized with channels in regular RGB format. LayoutLMv2 on the other hand normalizes the images internally and expects the channels in BGR format.
+    - text is tokenized using byte-pair encoding (BPE), as opposed to WordPiece. 
+  Due to these differences in data preprocessing, one can use [`LayoutLMv3Processor`] which internally combines a [`LayoutLMv3FeatureExtractor`] (for the image modality) and a [`LayoutLMv3Tokenizer`]/[`LayoutLMv3TokenizerFast`] (for the text modality) to prepare all data for the model.
+- Regarding usage of [`LayoutLMv3Processor`], we refer to the [usage guide](layoutlmv2#usage-LayoutLMv2Processor) of its predecessor. 
+- Demo notebooks for LayoutLMv3 can be found [here](https://github.com/NielsRogge/Transformers-Tutorials/tree/master/LayoutLMv3).
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/layoutlmv3_architecture.png"
+alt="drawing" width="600"/> 
+
+<small> LayoutLMv3 architecture. Taken from the <a href="https://arxiv.org/abs/2204.08387">original paper</a>. </small>
+
+This model was contributed by [nielsr](https://huggingface.co/nielsr). The original code can be found [here](https://github.com/microsoft/unilm/tree/master/layoutlmv3).
+
+
+## LayoutLMv3Config
+
+[[autodoc]] LayoutLMv3Config
+
+## LayoutLMv3FeatureExtractor
+
+[[autodoc]] LayoutLMv3FeatureExtractor
+    - __call__
+
+## LayoutLMv3Tokenizer
+
+[[autodoc]] LayoutLMv3Tokenizer
+    - __call__
+    - save_vocabulary
+
+## LayoutLMv3TokenizerFast
+
+[[autodoc]] LayoutLMv3TokenizerFast
+    - __call__
+
+## LayoutLMv3Processor
+
+[[autodoc]] LayoutLMv3Processor
+    - __call__
+
+## LayoutLMv3Model
+
+[[autodoc]] LayoutLMv3Model
+    - forward
+
+## LayoutLMv3ForSequenceClassification
+
+[[autodoc]] LayoutLMv3ForSequenceClassification
+    - forward
+
+## LayoutLMv3ForTokenClassification
+
+[[autodoc]] LayoutLMv3ForTokenClassification
+    - forward
+
+## LayoutLMv3ForQuestionAnswering
+
+[[autodoc]] LayoutLMv3ForQuestionAnswering
+    - forward

docs/source/en/model_doc/led.mdx

@@ -44,8 +44,10 @@ Tips:
 - LED makes use of *global attention* by means of the `global_attention_mask` (see
   [`LongformerModel`]). For summarization, it is advised to put *global attention* only on the first
   `<s>` token. For question answering, it is advised to put *global attention* on all tokens of the question.
-- To fine-tune LED on all 16384, it is necessary to enable *gradient checkpointing* by executing
-  `model.gradient_checkpointing_enable()`.
+- To fine-tune LED on all 16384, *gradient checkpointing* can be enabled in case training leads to out-of-memory (OOM)
+  errors. This can be done by executing `model.gradient_checkpointing_enable()`. 
+ Moreover, the `use_cache=False`
+  flag can be used to disable the caching mechanism to save memory.
 - A notebook showing how to evaluate LED, can be accessed [here](https://colab.research.google.com/drive/12INTTR6n64TzS4RrXZxMSXfrOd9Xzamo?usp=sharing).
 - A notebook showing how to fine-tune LED, can be accessed [here](https://colab.research.google.com/drive/12LjJazBl7Gam0XBPy_y0CTOJZeZ34c2v?usp=sharing).
 

docs/source/en/model_doc/levit.mdx

@@ -0,0 +1,87 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# LeViT
+
+## Overview
+
+The LeViT model was proposed in [LeViT: Introducing Convolutions to Vision Transformers](https://arxiv.org/abs/2104.01136) by Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze. LeViT improves the [Vision Transformer (ViT)](vit) in performance and efficiency by a few architectural differences such as activation maps with decreasing resolutions in Transformers and the introduction of an attention bias to integrate positional information.
+
+The abstract from the paper is the following:
+
+*We design a family of image classification architectures that optimize the trade-off between accuracy
+and efficiency in a high-speed regime. Our work exploits recent findings in attention-based architectures, 
+which are competitive on highly parallel processing hardware. We revisit principles from the extensive 
+literature on convolutional neural networks to apply them to transformers, in particular activation maps 
+with decreasing resolutions. We also introduce the attention bias, a new way to integrate positional information
+in vision transformers. As a result, we propose LeVIT: a hybrid neural network for fast inference image classification. 
+We consider different measures of efficiency on different hardware platforms, so as to best reflect a wide range of 
+application scenarios. Our extensive experiments empirically validate our technical choices and show they are suitable 
+to most architectures. Overall, LeViT significantly outperforms existing convnets and vision transformers with respect 
+to the speed/accuracy tradeoff. For example, at 80% ImageNet top-1 accuracy, LeViT is 5 times faster than EfficientNet on CPU. *
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/levit_architecture.png"
+alt="drawing" width="600"/> 
+
+<small> LeViT Architecture. Taken from the <a href="https://arxiv.org/abs/2104.01136">original paper</a>.</small>
+
+Tips:
+
+- Compared to ViT, LeViT models use an additional distillation head to effectively learn from a teacher (which, in the LeViT paper, is a ResNet like-model). The distillation head is learned through backpropagation under supervision of a ResNet like-model. They also draw inspiration from convolution neural networks to use activation maps with decreasing resolutions to increase the efficiency.
+- There are 2 ways to fine-tune distilled models, either (1) in a classic way, by only placing a prediction head on top
+  of the final hidden state and not using the distillation head, or (2) by placing both a prediction head and distillation 
+  head on top of the final hidden state. In that case, the prediction head is trained using regular cross-entropy between 
+  the prediction of the head and the ground-truth label, while the distillation prediction head is trained using hard distillation 
+  (cross-entropy between the prediction of the distillation head and the label predicted by the teacher). At inference time, 
+  one takes the average prediction between both heads as final prediction. (2) is also called "fine-tuning with distillation", 
+  because one relies on a teacher that has already been fine-tuned on the downstream dataset. In terms of models, (1) corresponds 
+  to [`LevitForImageClassification`] and (2) corresponds to [`LevitForImageClassificationWithTeacher`].
+- All released checkpoints were pre-trained and fine-tuned on  [ImageNet-1k](https://huggingface.co/datasets/imagenet-1k) 
+  (also referred to as ILSVRC 2012, a collection of 1.3 million images and 1,000 classes). only. No external data was used. This is in
+  contrast with the original ViT model, which used external data like the JFT-300M dataset/Imagenet-21k for
+  pre-training.
+- The authors of LeViT released 5 trained LeViT models, which you can directly plug into [`LevitModel`] or [`LevitForImageClassification`]. 
+  Techniques like data augmentation, optimization, and regularization were used in order to simulate training on a much larger dataset
+  (while only using ImageNet-1k for pre-training). The 5 variants available are (all trained on images of size 224x224):
+  *facebook/levit-128S*, *facebook/levit-128*, *facebook/levit-192*, *facebook/levit-256* and
+  *facebook/levit-384*. Note that one should use [`LevitFeatureExtractor`] in order to
+  prepare images for the model.
+- [`LevitForImageClassificationWithTeacher`] currently supports only inference and not training or fine-tuning.
+- You can check out demo notebooks regarding inference as well as fine-tuning on custom data [here](https://github.com/NielsRogge/Transformers-Tutorials/tree/master/VisionTransformer) 
+  (you can just replace [`ViTFeatureExtractor`] by [`LevitFeatureExtractor`] and [`ViTForImageClassification`] by [`LevitForImageClassification`] or [`LevitForImageClassificationWithTeacher`]).
+
+This model was contributed by [anugunj](https://huggingface.co/anugunj). The original code can be found [here](https://github.com/facebookresearch/LeViT).
+
+
+## LevitConfig
+
+[[autodoc]] LevitConfig
+
+## LevitFeatureExtractor
+
+[[autodoc]] LevitFeatureExtractor
+    - __call__
+
+## LevitModel
+
+[[autodoc]] LevitModel
+    - forward
+
+## LevitForImageClassification
+
+[[autodoc]] LevitForImageClassification
+    - forward
+
+## LevitForImageClassificationWithTeacher
+
+[[autodoc]] LevitForImageClassificationWithTeacher
+    - forward

docs/source/en/model_doc/longt5.mdx

@@ -0,0 +1,121 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# LongT5
+
+## Overview
+
+The LongT5 model was proposed in [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916)
+by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung and Yinfei Yang. It's an
+encoder-decoder transformer pre-trained in a text-to-text denoising generative setting. LongT5 model is an extension of
+T5 model, and it enables using one of the two different efficient attention mechanisms - (1) Local attention, or (2)
+Transient-Global attention.
+
+
+The abstract from the paper is the following:
+
+*Recent work has shown that either (1) increasing the input length or (2) increasing model size can improve the
+performance of Transformer-based neural models. In this paper, we present a new model, called LongT5, with which we
+explore the effects of scaling both the input length and model size at the same time. Specifically, we integrated
+attention ideas from long-input transformers (ETC), and adopted pre-training strategies from summarization pre-training
+(PEGASUS) into the scalable T5 architecture. The result is a new attention mechanism we call {\em Transient Global}
+(TGlobal), which mimics ETC's local/global attention mechanism, but without requiring additional side-inputs. We are
+able to achieve state-of-the-art results on several summarization tasks and outperform the original T5 models on
+question answering tasks.*
+
+Tips:
+
+- [`LongT5ForConditionalGeneration`] is an extension of [`T5ForConditionalGeneration`] exchanging the traditional
+encoder *self-attention* layer with efficient either *local* attention or *transient-global* (*tglobal*) attention.
+- Unlike the T5 model, LongT5 does not use a task prefix. Furthermore, it uses a different pre-training objective
+inspired by the pre-training of `[PegasusForConditionalGeneration]`.
+- LongT5 model is designed to work efficiently and very well on long-range *sequence-to-sequence* tasks where the
+input sequence exceeds commonly used 512 tokens. It is capable of handling input sequences of a length up to 16,384 tokens.
+- For *Local Attention*, the sparse sliding-window local attention operation allows a given token to attend only `r`
+tokens to the left and right of it (with `r=127` by default). *Local Attention* does not introduce any new parameters
+to the model. The complexity of the mechanism is linear in input sequence length `l`: `O(l*r)`.
+- *Transient Global Attention* is an extension of the *Local Attention*. It, furthermore, allows each input token to
+interact with all other tokens in the layer. This is achieved via splitting an input sequence into blocks of a fixed
+length `k` (with a default `k=16`). Then, a global token for such a block is obtained via summing and normalizing the embeddings of every token
+in the block. Thanks to this, the attention allows each token to attend to both nearby tokens like in Local attention, and
+also every global token like in the case of standard global attention (*transient* represents the fact the global tokens
+are constructed dynamically within each attention operation).  As a consequence, *TGlobal* attention introduces
+a few new parameters -- global relative position biases and a layer normalization for global token's embedding.
+The complexity of this mechanism is `O(l(r + l/k))`.
+- An example showing how to evaluate a fine-tuned LongT5 model on the [pubmed dataset](https://huggingface.co/datasets/scientific_papers) is below.
+
+```python
+>>> import evaluate
+>>> from datasets import load_dataset
+>>> from transformers import AutoTokenizer, LongT5ForConditionalGeneration
+
+>>> dataset = load_dataset("scientific_papers", "pubmed", split="validation")
+>>> model = (
+...     LongT5ForConditionalGeneration.from_pretrained("Stancld/longt5-tglobal-large-16384-pubmed-3k_steps")
+...     .to("cuda")
+...     .half()
+... )
+>>> tokenizer = AutoTokenizer.from_pretrained("Stancld/longt5-tglobal-large-16384-pubmed-3k_steps")
+
+
+>>> def generate_answers(batch):
+...     inputs_dict = tokenizer(
+...         batch["article"], max_length=16384, padding="max_length", truncation=True, return_tensors="pt"
+...     )
+...     input_ids = inputs_dict.input_ids.to("cuda")
+...     attention_mask = inputs_dict.attention_mask.to("cuda")
+...     output_ids = model.generate(input_ids, attention_mask=attention_mask, max_length=512, num_beams=2)
+...     batch["predicted_abstract"] = tokenizer.batch_decode(output_ids, skip_special_tokens=True)
+...     return batch
+
+
+>>> result = dataset.map(generate_answer, batched=True, batch_size=2)
+>>> rouge = evaluate.load("rouge")
+>>> rouge.compute(predictions=result["predicted_abstract"], references=result["abstract"])
+```
+
+This model was contributed by [stancld](https://huggingface.co/stancld).
+The original code can be found [here](https://github.com/google-research/longt5).
+
+
+## LongT5Config
+
+[[autodoc]] LongT5Config
+
+## LongT5Model
+
+[[autodoc]] LongT5Model
+    - forward
+
+## LongT5ForConditionalGeneration
+
+[[autodoc]] LongT5ForConditionalGeneration
+    - forward
+
+## LongT5EncoderModel
+
+[[autodoc]] LongT5EncoderModel
+    - forward
+
+## FlaxLongT5Model
+
+[[autodoc]] FlaxLongT5Model
+    - __call__
+    - encode
+    - decode
+
+## FlaxLongT5ForConditionalGeneration
+
+[[autodoc]] FlaxLongT5ForConditionalGeneration
+    - __call__
+    - encode
+    - decode

docs/source/en/model_doc/luke.mdx

@@ -97,7 +97,7 @@ Example:
 >>> entities = [
 ...     "Beyoncé",
 ...     "Los Angeles",
->>> ]  # Wikipedia entity titles corresponding to the entity mentions "Beyoncé" and "Los Angeles"
+... ]  # Wikipedia entity titles corresponding to the entity mentions "Beyoncé" and "Los Angeles"
 >>> entity_spans = [(0, 7), (17, 28)]  # character-based entity spans corresponding to "Beyoncé" and "Los Angeles"
 >>> inputs = tokenizer(text, entities=entities, entity_spans=entity_spans, add_prefix_space=True, return_tensors="pt")
 >>> outputs = model(**inputs)

docs/source/en/model_doc/mctct.mdx

@@ -0,0 +1,62 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# M-CTC-T
+
+## Overview
+
+The M-CTC-T model was proposed in [Pseudo-Labeling For Massively Multilingual Speech Recognition](https://arxiv.org/abs/2111.00161) by Loren Lugosch, Tatiana Likhomanenko, Gabriel Synnaeve, and Ronan Collobert. The model is a 1B-param transformer encoder, with a CTC head over 8065 character labels and a language identification head over 60 language ID labels. It is trained on Common Voice (version 6.1, December 2020 release) and VoxPopuli. After training on Common Voice and VoxPopuli, the model is trained on Common Voice only. The labels are unnormalized character-level transcripts (punctuation and capitalization are not removed). The model takes as input Mel filterbank features from a 16Khz audio signal.
+
+The abstract from the paper is the following:
+
+*Semi-supervised learning through pseudo-labeling has become a staple of state-of-the-art monolingual
+speech recognition systems. In this work, we extend pseudo-labeling to massively multilingual speech
+recognition with 60 languages. We propose a simple pseudo-labeling recipe that works well even
+with low-resource languages: train a supervised multilingual model, fine-tune it with semi-supervised
+learning on a target language, generate pseudo-labels for that language, and train a final model using
+pseudo-labels for all languages, either from scratch or by fine-tuning. Experiments on the labeled
+Common Voice and unlabeled VoxPopuli datasets show that our recipe can yield a model with better
+performance for many languages that also transfers well to LibriSpeech.*
+
+
+
+This model was contributed by [cwkeam](https://huggingface.co/cwkeam). The original code can be found [here](https://github.com/flashlight/wav2letter/tree/main/recipes/mling_pl).
+
+## MCTCTConfig
+
+[[autodoc]] MCTCTConfig
+
+## MCTCTFeatureExtractor
+
+[[autodoc]] MCTCTFeatureExtractor
+    - __call__
+
+## MCTCTProcessor
+
+[[autodoc]] MCTCTProcessor
+    - __call__
+    - from_pretrained
+    - save_pretrained
+    - batch_decode
+    - decode
+    - as_target_processor
+
+
+## MCTCTModel
+
+[[autodoc]] MCTCTModel
+    - forward
+
+## MCTCTForCTC
+
+[[autodoc]] MCTCTForCTC
+    - forward

docs/source/en/model_doc/mobilevit.mdx

@@ -0,0 +1,55 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# MobileViT
+
+## Overview
+
+The MobileViT model was proposed in [MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer](https://arxiv.org/abs/2110.02178) by Sachin Mehta and Mohammad Rastegari. MobileViT introduces a new layer that replaces local processing in convolutions with global processing using transformers. 
+
+The abstract from the paper is the following:
+
+*Light-weight convolutional neural networks (CNNs) are the de-facto for mobile vision tasks. Their spatial inductive biases allow them to learn representations with fewer parameters across different vision tasks. However, these networks are spatially local. To learn global representations, self-attention-based vision trans-formers (ViTs) have been adopted. Unlike CNNs, ViTs are heavy-weight. In this paper, we ask the following question: is it possible to combine the strengths of CNNs and ViTs to build a light-weight and low latency network for mobile vision tasks? Towards this end, we introduce MobileViT, a light-weight and general-purpose vision transformer for mobile devices. MobileViT presents a different perspective for the global processing of information with transformers, i.e., transformers as convolutions. Our results show that MobileViT significantly outperforms CNN- and ViT-based networks across different tasks and datasets. On the ImageNet-1k dataset, MobileViT achieves top-1 accuracy of 78.4% with about 6 million parameters, which is 3.2% and 6.2% more accurate than MobileNetv3 (CNN-based) and DeIT (ViT-based) for a similar number of parameters. On the MS-COCO object detection task, MobileViT is 5.7% more accurate than MobileNetv3 for a similar number of parameters.*
+
+Tips:
+
+- MobileViT is more like a CNN than a Transformer model. It does not work on sequence data but on batches of images. Unlike ViT, there are no embeddings. The backbone model outputs a feature map.
+- One can use [`MobileViTFeatureExtractor`] to prepare images for the model. Note that if you do your own preprocessing, the pretrained checkpoints expect images to be in BGR pixel order (not RGB).
+- The available image classification checkpoints are pre-trained on [ImageNet-1k](https://huggingface.co/datasets/imagenet-1k) (also referred to as ILSVRC 2012, a collection of 1.3 million images and 1,000 classes).
+- The segmentation model uses a [DeepLabV3](https://arxiv.org/abs/1706.05587) head. The available semantic segmentation checkpoints are pre-trained on [PASCAL VOC](http://host.robots.ox.ac.uk/pascal/VOC/). 
+
+This model was contributed by [matthijs](https://huggingface.co/Matthijs). The original code and weights can be found [here](https://github.com/apple/ml-cvnets).
+
+
+## MobileViTConfig
+
+[[autodoc]] MobileViTConfig
+
+## MobileViTFeatureExtractor
+
+[[autodoc]] MobileViTFeatureExtractor
+    - __call__
+
+## MobileViTModel
+
+[[autodoc]] MobileViTModel
+    - forward
+
+## MobileViTForImageClassification
+
+[[autodoc]] MobileViTForImageClassification
+    - forward
+
+## MobileViTForSemanticSegmentation
+
+[[autodoc]] MobileViTForSemanticSegmentation
+    - forward

docs/source/en/model_doc/mt5.mdx

@@ -96,3 +96,7 @@ See [`T5TokenizerFast`] for all details.
 ## FlaxMT5ForConditionalGeneration
 
 [[autodoc]] FlaxMT5ForConditionalGeneration
+
+## FlaxMT5EncoderModel
+
+[[autodoc]] FlaxMT5EncoderModel

docs/source/en/model_doc/mvp.mdx

@@ -0,0 +1,138 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# MVP
+
+## Overview
+
+The MVP model was proposed in [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://arxiv.org/abs/2206.12131) by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen.
+
+
+According to the abstract,
+
+- MVP follows a standard Transformer encoder-decoder architecture.
+- MVP is supervised pre-trained using labeled datasets.
+- MVP also has task-specific soft prompts to stimulate the model's capacity in performing a certain task.
+- MVP is specially designed for natural language generation and can be adapted to a wide range of generation tasks, including but not limited to summarization, data-to-text generation, open-ended dialogue system, story generation, question answering, question generation, task-oriented dialogue system, commonsense generation, paraphrase generation, text style transfer, and text simplification. Our model can also be adapted to natural language understanding tasks such as sequence classification and (extractive) question answering.
+
+Tips:
+- We have released a series of models [here](https://huggingface.co/models?filter=mvp), including MVP, MVP with task-specific prompts, and multi-task pre-trained variants.
+- If you want to use a model without prompts (standard Transformer), you can load it through `MvpForConditionalGeneration.from_pretrained('RUCAIBox/mvp')`.
+- If you want to use a model with task-specific prompts, such as summarization, you can load it through `MvpForConditionalGeneration.from_pretrained('RUCAIBox/mvp-summarization')`.
+- Our model supports lightweight prompt tuning following [Prefix-tuning](https://arxiv.org/abs/2101.00190) with method `set_lightweight_tuning()`.
+
+This model was contributed by [Tianyi Tang](https://huggingface.co/StevenTang). The detailed information and instructions can be found [here](https://github.com/RUCAIBox/MVP).
+
+## Examples
+For summarization, it is an example to use MVP and MVP with summarization-specific prompts.
+
+```python
+>>> from transformers import MvpTokenizer, MvpForConditionalGeneration
+
+>>> tokenizer = MvpTokenizer.from_pretrained("RUCAIBox/mvp")
+>>> model = MvpForConditionalGeneration.from_pretrained("RUCAIBox/mvp")
+>>> model_with_prompt = MvpForConditionalGeneration.from_pretrained("RUCAIBox/mvp-summarization")
+
+>>> inputs = tokenizer(
+...     "Summarize: You may want to stick it to your boss and leave your job, but don't do it if these are your reasons.",
+...     return_tensors="pt",
+... )
+>>> generated_ids = model.generate(**inputs)
+>>> tokenizer.batch_decode(generated_ids, skip_special_tokens=True)
+["Why You Shouldn't Quit Your Job"]
+
+>>> generated_ids = model_with_prompt.generate(**inputs)
+>>> tokenizer.batch_decode(generated_ids, skip_special_tokens=True)
+["Don't do it if these are your reasons"]
+```
+
+For data-to-text generation, it is an example to use MVP and multi-task pre-trained variants.
+```python
+>>> from transformers import MvpTokenizerFast, MvpForConditionalGeneration
+
+>>> tokenizer = MvpTokenizerFast.from_pretrained("RUCAIBox/mvp")
+>>> model = MvpForConditionalGeneration.from_pretrained("RUCAIBox/mvp")
+>>> model_with_mtl = MvpForConditionalGeneration.from_pretrained("RUCAIBox/mtl-data-to-text")
+
+>>> inputs = tokenizer(
+...     "Describe the following data: Iron Man | instance of | Superhero [SEP] Stan Lee | creator | Iron Man",
+...     return_tensors="pt",
+... )
+>>> generated_ids = model.generate(**inputs)
+>>> tokenizer.batch_decode(generated_ids, skip_special_tokens=True)
+['Stan Lee created the character of Iron Man, a fictional superhero appearing in American comic']
+
+>>> generated_ids = model_with_mtl.generate(**inputs)
+>>> tokenizer.batch_decode(generated_ids, skip_special_tokens=True)
+['Iron Man is a fictional superhero appearing in American comic books published by Marvel Comics.']
+```
+
+For lightweight tuning, *i.e.*, fixing the model and only tuning prompts, you can load MVP with randomly initialized prompts or with task-specific prompts. Our code also supports Prefix-tuning with BART following the [original paper](https://arxiv.org/abs/2101.00190).
+
+```python
+>>> from transformers import MvpForConditionalGeneration
+
+>>> model = MvpForConditionalGeneration.from_pretrained("RUCAIBox/mvp", use_prompt=True)
+>>> # the number of trainable parameters (full tuning)
+>>> sum(p.numel() for p in model.parameters() if p.requires_grad)
+468116832
+
+>>> # lightweight tuning with randomly initialized prompts
+>>> model.set_lightweight_tuning()
+>>> # the number of trainable parameters (lightweight tuning)
+>>> sum(p.numel() for p in model.parameters() if p.requires_grad)
+61823328
+
+>>> # lightweight tuning with task-specific prompts
+>>> model = MvpForConditionalGeneration.from_pretrained("RUCAIBox/mtl-data-to-text")
+>>> model.set_lightweight_tuning()
+>>> # original lightweight Prefix-tuning
+>>> model = MvpForConditionalGeneration.from_pretrained("facebook/bart-large", use_prompt=True)
+>>> model.set_lightweight_tuning()
+```
+
+## MvpConfig
+
+[[autodoc]] MvpConfig
+
+## MvpTokenizer
+
+[[autodoc]] MvpTokenizer
+
+## MvpTokenizerFast
+
+[[autodoc]] MvpTokenizerFast
+
+## MvpModel
+
+[[autodoc]] MvpModel
+    - forward
+
+## MvpForConditionalGeneration
+
+[[autodoc]] MvpForConditionalGeneration
+    - forward
+
+## MvpForSequenceClassification
+
+[[autodoc]] MvpForSequenceClassification
+    - forward
+
+## MvpForQuestionAnswering
+
+[[autodoc]] MvpForQuestionAnswering
+    - forward
+
+## MvpForCausalLM
+
+[[autodoc]] MvpForCausalLM
+    - forward

docs/source/en/model_doc/nezha.mdx

@@ -0,0 +1,76 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# Nezha
+
+## Overview
+
+The Nezha model was proposed in [NEZHA: Neural Contextualized Representation for Chinese Language Understanding](https://arxiv.org/abs/1909.00204) by Junqiu Wei et al.
+
+The abstract from the paper is the following:
+
+*The pre-trained language models have achieved great successes in various natural language understanding (NLU) tasks
+due to its capacity to capture the deep contextualized information in text by pre-training on large-scale corpora.
+In this technical report, we present our practice of pre-training language models named NEZHA (NEural contextualiZed
+representation for CHinese lAnguage understanding) on Chinese corpora and finetuning for the Chinese NLU tasks. 
+The current version of NEZHA is based on BERT with a collection of proven improvements, which include Functional 
+Relative Positional Encoding as an effective positional encoding scheme, Whole Word Masking strategy,
+Mixed Precision Training and the LAMB Optimizer in training the models. The experimental results show that NEZHA
+achieves the state-of-the-art performances when finetuned on several representative Chinese tasks, including
+named entity recognition (People's Daily NER), sentence matching (LCQMC), Chinese sentiment classification (ChnSenti)
+and natural language inference (XNLI).*
+
+This model was contributed by [sijunhe](https://huggingface.co/sijunhe). The original code can be found [here](https://github.com/huawei-noah/Pretrained-Language-Model/tree/master/NEZHA-PyTorch).
+
+## NezhaConfig
+
+[[autodoc]] NezhaConfig
+
+## NezhaModel
+
+[[autodoc]] NezhaModel
+    - forward
+
+## NezhaForPreTraining
+
+[[autodoc]] NezhaForPreTraining
+    - forward
+
+## NezhaForMaskedLM
+
+[[autodoc]] NezhaForMaskedLM
+    - forward
+
+## NezhaForNextSentencePrediction
+
+[[autodoc]] NezhaForNextSentencePrediction
+    - forward
+
+## NezhaForSequenceClassification
+
+[[autodoc]] NezhaForSequenceClassification
+    - forward
+
+## NezhaForMultipleChoice
+
+[[autodoc]] NezhaForMultipleChoice
+    - forward
+
+## NezhaForTokenClassification
+
+[[autodoc]] NezhaForTokenClassification
+    - forward
+
+## NezhaForQuestionAnswering
+
+[[autodoc]] NezhaForQuestionAnswering
+    - forward

docs/source/en/model_doc/nllb.mdx

@@ -0,0 +1,99 @@
+<!--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.
+-->
+
+# NLLB
+
+**DISCLAIMER:** If you see something strange, file a [Github Issue](https://github.com/huggingface/transformers/issues/new?assignees=&labels=bug&template=bug-report.yml) and assign
+@LysandreJik
+
+## Overview of NLLB
+
+The NLLB model was presented in [No Language Left Behind: Scaling Human-Centered Machine Translation](https://arxiv.org/abs/2207.04672) by Marta R. Costa-jussà, James Cross, Onur Çelebi,
+Maha Elbayad, Kenneth Heafield, Kevin Heffernan, Elahe Kalbassi, Janice Lam, Daniel Licht, Jean Maillard, Anna Sun, Skyler Wang, Guillaume Wenzek, Al Youngblood, Bapi Akula,
+Loic Barrault, Gabriel Mejia Gonzalez, Prangthip Hansanti, John Hoffman, Semarley Jarrett, Kaushik Ram Sadagopan, Dirk Rowe, Shannon Spruit, Chau Tran, Pierre Andrews,
+Necip Fazil Ayan, Shruti Bhosale, Sergey Edunov, Angela Fan, Cynthia Gao, Vedanuj Goswami, Francisco Guzmán, Philipp Koehn, Alexandre Mourachko, Christophe Ropers,
+Safiyyah Saleem, Holger Schwenk, and Jeff Wang.
+
+The abstract of the paper is the following:
+
+*Driven by the goal of eradicating language barriers on a global scale, machine translation has solidified itself as a key focus of artificial intelligence research today.
+However, such efforts have coalesced around a small subset of languages, leaving behind the vast majority of mostly low-resource languages. What does it take to break the
+200 language barrier while ensuring safe, high quality results, all while keeping ethical considerations in mind? In No Language Left Behind, we took on this challenge by
+first contextualizing the need for low-resource language translation support through exploratory interviews with native speakers. Then, we created datasets and models aimed
+at narrowing the performance gap between low and high-resource languages. More specifically, we developed a conditional compute model based on Sparsely Gated Mixture of
+Experts that is trained on data obtained with novel and effective data mining techniques tailored for low-resource languages. We propose multiple architectural and training
+improvements to counteract overfitting while training on thousands of tasks. Critically, we evaluated the performance of over 40,000 different translation directions using
+a human-translated benchmark, Flores-200, and combined human evaluation with a novel toxicity benchmark covering all languages in Flores-200 to assess translation safety.
+Our model achieves an improvement of 44% BLEU relative to the previous state-of-the-art, laying important groundwork towards realizing a universal translation system.*
+
+This implementation contains the dense models available on release. Let us know via a GitHub issue if you would like to see the MoE models as well.
+
+This model was contributed by [Lysandre](https://huggingface.co/lysandre). The authors' code can be found [here](https://github.com/facebookresearch/fairseq/tree/nllb).
+
+## Generating with NLLB
+
+While generating the target text set the `forced_bos_token_id` to the target language id. The following
+example shows how to translate English to French using the *facebook/nllb-200-distilled-600M* model.
+
+Note that we're using the BCP-47 code for French `fra_Latn`. See [here](https://github.com/facebookresearch/flores/blob/main/flores200/README.md#languages-in-flores-200)
+for the list of all BCP-47 in the Flores 200 dataset.
+
+```python
+>>> from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
+
+>>> tokenizer = AutoTokenizer.from_pretrained("facebook/nllb-200-distilled-600M")
+>>> model = AutoModelForSeq2SeqLM.from_pretrained("facebook/nllb-200-distilled-600M")
+
+>>> article = "UN Chief says there is no military solution in Syria"
+>>> inputs = tokenizer(article, return_tensors="pt")
+
+>>> translated_tokens = model.generate(
+...     **inputs, forced_bos_token_id=tokenizer.lang_code_to_id["fra_Latn"], max_length=30
+... )
+>>> tokenizer.batch_decode(translated_tokens, skip_special_tokens=True)[0]
+Le chef de l'ONU dit qu'il n'y a pas de solution militaire en Syrie
+```
+
+### Generating from any other language than English
+
+English (`eng_Latn`) is set as the default language from which to translate. In order to specify that you'd like to translate from a different language,
+you should specify the BCP-47 code in the `src_lang` keyword argument of the tokenizer initialization.
+
+See example below for a translation from romanian to german:
+
+```py
+>>> from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
+
+>>> tokenizer = AutoTokenizer.from_pretrained(
+...     "facebook/nllb-200-distilled-600M", use_auth_token=True, src_lang="ron_Latn"
+... )
+>>> model = AutoModelForSeq2SeqLM.from_pretrained("facebook/nllb-200-distilled-600M", use_auth_token=True)
+
+>>> article = "Şeful ONU spune că nu există o soluţie militară în Siria"
+>>> inputs = tokenizer(article, return_tensors="pt")
+
+>>> translated_tokens = model.generate(
+...     **inputs, forced_bos_token_id=tokenizer.lang_code_to_id["deu_Latn"], max_length=30
+... )
+>>> tokenizer.batch_decode(translated_tokens, skip_special_tokens=True)[0]
+UN-Chef sagt, es gibt keine militärische Lösung in Syrien
+```
+
+## NllbTokenizer
+
+[[autodoc]] NllbTokenizer
+    - as_target_tokenizer
+    - build_inputs_with_special_tokens
+
+## NllbTokenizerFast
+
+[[autodoc]] NllbTokenizerFast

docs/source/en/model_doc/opt.mdx

@@ -0,0 +1,71 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# OPT
+
+## Overview
+
+The OPT model was proposed in [Open Pre-trained Transformer Language Models](https://arxiv.org/pdf/2205.01068) by Meta AI.
+OPT is a series of open-sourced large causal language models which perform similar in performance to GPT3. 
+
+
+The abstract from the paper is the following:
+
+*Large language models, which are often trained for hundreds of thousands of compute days, have shown remarkable capabilities for zero- and few-shot learning. Given their computational cost, these models are difficult to replicate without significant capital. For the few that are available through APIs, no access is granted to the full model weights, making them difficult to study. We present Open Pre-trained Transformers (OPT), a suite of decoder-only pre-trained transformers ranging from 125M to 175B parameters, which we aim to fully and responsibly share with interested researchers. We show that OPT-175B is comparable to GPT-3, while requiring only 1/7th the carbon footprint to develop. We are also releasing our logbook detailing the infrastructure challenges we faced, along with code for experimenting with all of the released models.*
+
+Tips:
+- OPT has the same architecture as [`BartDecoder`].
+- Contrary to GPT2, OPT adds the EOS token `</s>` to the beginning of every prompt. **Note**: Make sure to pass `use_fast=False` when loading OPT's tokenizer with [`AutoTokenizer`] to get the correct tokenizer.
+
+This model was contributed by [Arthur Zucker](https://huggingface.co/ArthurZ), [Younes Belkada](https://huggingface.co/ybelkada), and [Patrick Von Platen](https://huggingface.co/patrickvonplaten).
+The original code can be found [here](https://github.com/facebookresearch/metaseq).
+
+
+## OPTConfig
+
+[[autodoc]] OPTConfig
+
+## OPTModel
+
+[[autodoc]] OPTModel
+    - forward
+
+## OPTForCausalLM
+
+[[autodoc]] OPTForCausalLM
+    - forward
+
+## TFOPTModel
+
+[[autodoc]] TFOPTModel
+    - call
+
+## TFOPTForCausalLM
+
+[[autodoc]] TFOPTForCausalLM
+    - call
+
+## OPTForSequenceClassification
+
+[[autodoc]] OPTForSequenceClassification
+    - forward
+
+## FlaxOPTModel
+
+[[autodoc]] FlaxOPTModel
+    - __call__
+
+
+## FlaxOPTForCausalLM
+
+[[autodoc]] FlaxOPTForCausalLM
+    - __call__

docs/source/en/model_doc/owlvit.mdx

@@ -0,0 +1,108 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# OWL-ViT
+
+## Overview
+
+The OWL-ViT (short for Vision Transformer for Open-World Localization) was proposed in [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby. OWL-ViT is an open-vocabulary object detection network trained on a variety of (image, text) pairs. It can be used to query an image with one or multiple text queries to search for and detect target objects described in text.
+
+The abstract from the paper is the following:
+
+*Combining simple architectures with large-scale pre-training has led to massive improvements in image classification. For object detection, pre-training and scaling approaches are less well established, especially in the long-tailed and open-vocabulary setting, where training data is relatively scarce. In this paper, we propose a strong recipe for transferring image-text models to open-vocabulary object detection. We use a standard Vision Transformer architecture with minimal modifications, contrastive image-text pre-training, and end-to-end detection fine-tuning. Our analysis of the scaling properties of this setup shows that increasing image-level pre-training and model size yield consistent improvements on the downstream detection task. We provide the adaptation strategies and regularizations needed to attain very strong performance on zero-shot text-conditioned and one-shot image-conditioned object detection. Code and models are available on GitHub.*
+
+## Usage
+
+OWL-ViT is a zero-shot text-conditioned object detection model. OWL-ViT uses [CLIP](clip) as its multi-modal backbone, with a ViT-like Transformer to get visual features and a causal language model to get the text features. To use CLIP for detection, OWL-ViT removes the final token pooling layer of the vision model and attaches a lightweight classification and box head to each transformer output token. Open-vocabulary classification is enabled by replacing the fixed classification layer weights with the class-name embeddings obtained from the text model. The authors first train CLIP from scratch and fine-tune it end-to-end with the classification and box heads on standard detection datasets using a bipartite matching loss. One or multiple text queries per image can be used to perform zero-shot text-conditioned object detection. 
+
+[`OwlViTFeatureExtractor`] can be used to resize (or rescale) and normalize images for the model and [`CLIPTokenizer`] is used to encode the text. [`OwlViTProcessor`] wraps [`OwlViTFeatureExtractor`] and [`CLIPTokenizer`] into a single instance to both encode the text and prepare the images. The following example shows how to perform object detection using [`OwlViTProcessor`] and [`OwlViTForObjectDetection`].
+
+
+```python
+>>> import requests
+>>> from PIL import Image
+>>> import torch
+
+>>> from transformers import OwlViTProcessor, OwlViTForObjectDetection
+
+>>> processor = OwlViTProcessor.from_pretrained("google/owlvit-base-patch32")
+>>> model = OwlViTForObjectDetection.from_pretrained("google/owlvit-base-patch32")
+
+>>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
+>>> image = Image.open(requests.get(url, stream=True).raw)
+>>> texts = [["a photo of a cat", "a photo of a dog"]]
+>>> inputs = processor(text=texts, images=image, return_tensors="pt")
+>>> outputs = model(**inputs)
+
+>>> # Target image sizes (height, width) to rescale box predictions [batch_size, 2]
+>>> target_sizes = torch.Tensor([image.size[::-1]])
+>>> # Convert outputs (bounding boxes and class logits) to COCO API
+>>> results = processor.post_process(outputs=outputs, target_sizes=target_sizes)
+
+>>> i = 0  # Retrieve predictions for the first image for the corresponding text queries
+>>> text = texts[i]
+>>> boxes, scores, labels = results[i]["boxes"], results[i]["scores"], results[i]["labels"]
+
+>>> score_threshold = 0.1
+>>> for box, score, label in zip(boxes, scores, labels):
+...     box = [round(i, 2) for i in box.tolist()]
+...     if score >= score_threshold:
+...         print(f"Detected {text[label]} with confidence {round(score.item(), 3)} at location {box}")
+Detected a photo of a cat with confidence 0.243 at location [1.42, 50.69, 308.58, 370.48]
+Detected a photo of a cat with confidence 0.298 at location [348.06, 20.56, 642.33, 372.61]
+```
+
+This model was contributed by [adirik](https://huggingface.co/adirik). The original code can be found [here](https://github.com/google-research/scenic/tree/main/scenic/projects/owl_vit).
+
+## OwlViTConfig
+
+[[autodoc]] OwlViTConfig
+    - from_text_vision_configs
+
+## OwlViTTextConfig
+
+[[autodoc]] OwlViTTextConfig
+
+## OwlViTVisionConfig
+
+[[autodoc]] OwlViTVisionConfig
+
+## OwlViTFeatureExtractor
+
+[[autodoc]] OwlViTFeatureExtractor
+    - __call__
+
+## OwlViTProcessor
+
+[[autodoc]] OwlViTProcessor
+
+## OwlViTModel
+
+[[autodoc]] OwlViTModel
+    - forward
+    - get_text_features
+    - get_image_features
+
+## OwlViTTextModel
+
+[[autodoc]] OwlViTTextModel
+    - forward
+
+## OwlViTVisionModel
+
+[[autodoc]] OwlViTVisionModel
+    - forward
+
+## OwlViTForObjectDetection
+
+[[autodoc]] OwlViTForObjectDetection
+    - forward

docs/source/en/model_doc/regnet.mdx

@@ -27,7 +27,8 @@ Tips:
 - One can use [`AutoFeatureExtractor`] to prepare images for the model.
 - The huge 10B model from [Self-supervised Pretraining of Visual Features in the Wild](https://arxiv.org/abs/2103.01988), trained on one billion Instagram images, is available on the [hub](https://huggingface.co/facebook/regnet-y-10b-seer)
 
-This model was contributed by [Francesco](https://huggingface.co/Francesco).
+This model was contributed by [Francesco](https://huggingface.co/Francesco). The TensorFlow version of the model
+was contributed by [sayakpaul](https://huggingface.com/sayakpaul) and [ariG23498](https://huggingface.com/ariG23498).
 The original code can be found [here](https://github.com/facebookresearch/pycls).
 
 
@@ -45,4 +46,15 @@ The original code can be found [here](https://github.com/facebookresearch/pycls)
 ## RegNetForImageClassification
 
 [[autodoc]] RegNetForImageClassification
-    - forward
+    - forward
+
+## TFRegNetModel
+
+[[autodoc]] TFRegNetModel
+    - call
+
+
+## TFRegNetForImageClassification
+
+[[autodoc]] TFRegNetForImageClassification
+    - call

docs/source/en/model_doc/resnet.mdx

@@ -31,7 +31,7 @@ The figure below illustrates the architecture of ResNet. Taken from the [origina
 
 <img width="600" src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/resnet_architecture.png"/>
 
-This model was contributed by [Francesco](https://huggingface.co/Francesco). The original code can be found [here](https://github.com/KaimingHe/deep-residual-networks).
+This model was contributed by [Francesco](https://huggingface.co/Francesco). The TensorFlow version of this model was added by [amyeroberts](https://huggingface.co/amyeroberts). The original code can be found [here](https://github.com/KaimingHe/deep-residual-networks).
 
 ## ResNetConfig
 
@@ -47,4 +47,16 @@ This model was contributed by [Francesco](https://huggingface.co/Francesco). The
 ## ResNetForImageClassification
 
 [[autodoc]] ResNetForImageClassification
-    - forward
+    - forward
+
+
+## TFResNetModel
+
+[[autodoc]] TFResNetModel
+    - call
+
+
+## TFResNetForImageClassification
+
+[[autodoc]] TFResNetForImageClassification
+    - call

docs/source/en/model_doc/roberta.mdx

@@ -136,6 +136,11 @@ This model was contributed by [julien-c](https://huggingface.co/julien-c). The o
 [[autodoc]] FlaxRobertaModel
     - __call__
 
+## FlaxRobertaForCausalLM
+
+[[autodoc]] FlaxRobertaForCausalLM
+    - __call__
+
 ## FlaxRobertaForMaskedLM
 
 [[autodoc]] FlaxRobertaForMaskedLM

docs/source/en/model_doc/segformer.mdx

@@ -36,13 +36,14 @@ The figure below illustrates the architecture of SegFormer. Taken from the [orig
 
 <img width="600" src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/segformer_architecture.png"/>
 
-This model was contributed by [nielsr](https://huggingface.co/nielsr). The original code can be found [here](https://github.com/NVlabs/SegFormer).
+This model was contributed by [nielsr](https://huggingface.co/nielsr). The TensorFlow version 
+of the model was contributed by [sayakpaul](https://huggingface.co/sayakpaul). The original code can be found [here](https://github.com/NVlabs/SegFormer).
 
 Tips:
 
-- SegFormer consists of a hierarchical Transformer encoder, and a lightweight all-MLP decode head.
+- SegFormer consists of a hierarchical Transformer encoder, and a lightweight all-MLP decoder head.
   [`SegformerModel`] is the hierarchical Transformer encoder (which in the paper is also referred to
-  as Mix Transformer or MiT). [`SegformerForSemanticSegmentation`] adds the all-MLP decode head on
+  as Mix Transformer or MiT). [`SegformerForSemanticSegmentation`] adds the all-MLP decoder head on
   top to perform semantic segmentation of images. In addition, there's
   [`SegformerForImageClassification`] which can be used to - you guessed it - classify images. The
   authors of SegFormer first pre-trained the Transformer encoder on ImageNet-1k to classify images. Next, they throw
@@ -51,6 +52,9 @@ Tips:
   found on the [hub](https://huggingface.co/models?other=segformer).
 - The quickest way to get started with SegFormer is by checking the [example notebooks](https://github.com/NielsRogge/Transformers-Tutorials/tree/master/SegFormer) (which showcase both inference and
   fine-tuning on custom data). One can also check out the [blog post](https://huggingface.co/blog/fine-tune-segformer) introducing SegFormer and illustrating how it can be fine-tuned on custom data.
+- TensorFlow users should refer to [this repository](https://github.com/deep-diver/segformer-tf-transformers) that shows off-the-shelf inference and fine-tuning.
+- One can also check out [this interactive demo on Hugging Face Spaces](https://huggingface.co/spaces/chansung/segformer-tf-transformers)
+  to try out a SegFormer model on custom images.
 - SegFormer works on any input size, as it pads the input to be divisible by `config.patch_sizes`. 
 - One can use [`SegformerFeatureExtractor`] to prepare images and corresponding segmentation maps
   for the model. Note that this feature extractor is fairly basic and does not include all data augmentations used in
@@ -65,7 +69,8 @@ Tips:
   used by [`SegformerForSemanticSegmentation`]). However, other datasets use the 0 index as
   background class and include this class as part of all labels. In that case, `reduce_labels` should be set to
   `False`, as loss should also be computed for the background class.
-- As most models, SegFormer comes in different sizes, the details of which can be found in the table below.
+- As most models, SegFormer comes in different sizes, the details of which can be found in the table below
+  (taken from Table 7 of the [original paper](https://arxiv.org/abs/2105.15203)).
 
 | **Model variant** | **Depths**    | **Hidden sizes**    | **Decoder hidden size** | **Params (M)** | **ImageNet-1k Top 1** |
 | :---------------: | ------------- | ------------------- | :---------------------: | :------------: | :-------------------: |
@@ -76,6 +81,10 @@ Tips:
 | MiT-b4            | [3, 8, 27, 3] | [64, 128, 320, 512] | 768                     | 62.6           | 83.6                  |
 | MiT-b5            | [3, 6, 40, 3] | [64, 128, 320, 512] | 768                     | 82.0           | 83.8                  |
 
+Note that MiT in the above table refers to the Mix Transformer encoder backbone introduced in SegFormer. For
+SegFormer's results on the segmentation datasets like ADE20k, refer to the [paper](https://arxiv.org/abs/2105.15203).
+
+
 ## SegformerConfig
 
 [[autodoc]] SegformerConfig
@@ -104,3 +113,23 @@ Tips:
 
 [[autodoc]] SegformerForSemanticSegmentation
     - forward
+
+## TFSegformerDecodeHead
+
+[[autodoc]] TFSegformerDecodeHead
+    - call
+
+## TFSegformerModel
+
+[[autodoc]] TFSegformerModel
+    - call 
+
+## TFSegformerForImageClassification
+
+[[autodoc]] TFSegformerForImageClassification
+    - call 
+
+## TFSegformerForSemanticSegmentation
+
+[[autodoc]] TFSegformerForSemanticSegmentation
+    - call 

docs/source/en/model_doc/speech-encoder-decoder.mdx

@@ -12,7 +12,7 @@ specific language governing permissions and limitations under the License.
 
 # Speech Encoder Decoder Models
 
-The [`SpeechEncoderDecoderModel`] can be used to initialize a speech-sequence-to-text-sequence model
+The [`SpeechEncoderDecoderModel`] can be used to initialize a speech-to-text model
 with any pretrained speech autoencoding model as the encoder (*e.g.* [Wav2Vec2](wav2vec2), [Hubert](hubert)) and any pretrained autoregressive model as the decoder.
 
 The effectiveness of initializing speech-sequence-to-text-sequence models with pretrained checkpoints for speech
@@ -20,9 +20,95 @@ recognition and speech translation has *e.g.* been shown in [Large-Scale Self- a
 Translation](https://arxiv.org/abs/2104.06678) by Changhan Wang, Anne Wu, Juan Pino, Alexei Baevski, Michael Auli,
 Alexis Conneau.
 
-An example of how to use a [`SpeechEncoderDecoderModel`] for inference can be seen in
-[Speech2Text2](speech_to_text_2).
+An example of how to use a [`SpeechEncoderDecoderModel`] for inference can be seen in [Speech2Text2](speech_to_text_2).
 
+## Randomly initializing `SpeechEncoderDecoderModel` from model configurations.
+
+[`SpeechEncoderDecoderModel`] can be randomly initialized from an encoder and a decoder config. In the following example, we show how to do this using the default [`Wav2Vec2Model`] configuration for the encoder
+and the default [`BertForCausalLM`] configuration for the decoder.
+
+```python
+>>> from transformers import BertConfig, Wav2Vec2Config, SpeechEncoderDecoderConfig, SpeechEncoderDecoderModel
+
+>>> config_encoder = Wav2Vec2Config()
+>>> config_decoder = BertConfig()
+
+>>> config = SpeechEncoderDecoderConfig.from_encoder_decoder_configs(config_encoder, config_decoder)
+>>> model = SpeechEncoderDecoderModel(config=config)
+```
+
+## Initialising `SpeechEncoderDecoderModel` from a pretrained encoder and a pretrained decoder.
+
+[`SpeechEncoderDecoderModel`] can be initialized from a pretrained encoder checkpoint and a pretrained decoder checkpoint. Note that any pretrained Transformer-based speech model, *e.g.* [Wav2Vec2](wav2vec2), [Hubert](hubert) can serve as the encoder and both pretrained auto-encoding models, *e.g.* BERT, pretrained causal language models, *e.g.* GPT2, as well as the pretrained decoder part of sequence-to-sequence models, *e.g.* decoder of BART, can be used as the decoder.
+Depending on which architecture you choose as the decoder, the cross-attention layers might be randomly initialized.
+Initializing [`SpeechEncoderDecoderModel`] from a pretrained encoder and decoder checkpoint requires the model to be fine-tuned on a downstream task, as has been shown in [the *Warm-starting-encoder-decoder blog post*](https://huggingface.co/blog/warm-starting-encoder-decoder).
+To do so, the `SpeechEncoderDecoderModel` class provides a [`SpeechEncoderDecoderModel.from_encoder_decoder_pretrained`] method.
+
+```python
+>>> from transformers import SpeechEncoderDecoderModel
+
+>>> model = SpeechEncoderDecoderModel.from_encoder_decoder_pretrained(
+...     "facebook/hubert-large-ll60k", "bert-base-uncased"
+... )
+```
+
+## Loading an existing `SpeechEncoderDecoderModel` checkpoint and perform inference.
+
+To load fine-tuned checkpoints of the `SpeechEncoderDecoderModel` class, [`SpeechEncoderDecoderModel`] provides the `from_pretrained(...)` method just like any other model architecture in Transformers.
+
+To perform inference, one uses the [`generate`] method, which allows to autoregressively generate text. This method supports various forms of decoding, such as greedy, beam search and multinomial sampling.
+
+```python
+>>> from transformers import Wav2Vec2Processor, SpeechEncoderDecoderModel
+>>> from datasets import load_dataset
+>>> import torch
+
+>>> # load a fine-tuned speech translation model and corresponding processor
+>>> model = SpeechEncoderDecoderModel.from_pretrained("facebook/wav2vec2-xls-r-300m-en-to-15")
+>>> processor = Wav2Vec2Processor.from_pretrained("facebook/wav2vec2-xls-r-300m-en-to-15")
+
+>>> # let's perform inference on a piece of English speech (which we'll translate to German)
+>>> ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
+>>> input_values = processor(ds[0]["audio"]["array"], return_tensors="pt").input_values
+
+>>> # autoregressively generate transcription (uses greedy decoding by default)
+>>> generated_ids = model.generate(input_values)
+>>> generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
+>>> print(generated_text)
+Mr. Quilter ist der Apostel der Mittelschicht und wir freuen uns, sein Evangelium willkommen heißen zu können.
+```
+
+## Training
+
+Once the model is created, it can be fine-tuned similar to BART, T5 or any other encoder-decoder model on a dataset of (speech, text) pairs.
+As you can see, only 2 inputs are required for the model in order to compute a loss: `input_values` (which are the
+speech inputs) and `labels` (which are the `input_ids` of the encoded target sequence).
+
+```python
+>>> from transformers import Wav2Vec2Processor, SpeechEncoderDecoderModel
+>>> from datasets import load_dataset
+
+>>> processor = Wav2Vec2Processor.from_pretrained("facebook/wav2vec2-base-960h")
+>>> tokenizer = BertTokenizer.from_pretrained("bert-base-uncased")
+>>> model = SpeechEncoderDecoderModel.from_encoder_decoder_pretrained(
+...     "facebook/wav2vec2-base-960h", "bert-base-uncased"
+... )
+
+>>> model.config.decoder_start_token_id = processor.tokenizer.cls_token_id
+>>> model.config.pad_token_id = processor.tokenizer.pad_token_id
+
+>>> # load a speech input
+>>> ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
+>>> input_values = processor(ds[0]["audio"]["array"], return_tensors="pt").input_values
+
+>>> # load its corresponding transcription
+>>> with processor.as_target_processor():
+...     labels = processor(ds[0]["text"], return_tensors="pt").input_ids
+
+>>> # the forward function automatically creates the correct decoder_input_ids
+>>> loss = model(input_values, labels=labels).loss
+>>> loss.backward()
+```
 
 ## SpeechEncoderDecoderConfig
 

docs/source/en/model_doc/speech_to_text.mdx

@@ -37,7 +37,7 @@ predicted token ids.
 
 The feature extractor depends on `torchaudio` and the tokenizer depends on `sentencepiece` so be sure to
 install those packages before running the examples. You could either install those as extra speech dependencies with
-`pip install transformers"[speech, sentencepiece]"` or install the packages seperately 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
+`pip install transformers"[speech, sentencepiece]"` or install the packages separately 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`
 
 

docs/source/en/model_doc/splinter.mdx

@@ -72,3 +72,8 @@ This model was contributed by [yuvalkirstain](https://huggingface.co/yuvalkirsta
 
 [[autodoc]] SplinterForQuestionAnswering
     - forward
+
+## SplinterForPreTraining
+
+[[autodoc]] SplinterForPreTraining
+    - forward

docs/source/en/model_doc/swin.mdx

@@ -14,22 +14,22 @@ specific language governing permissions and limitations under the License.
 
 ## Overview
 
-The Swin Transformer was proposed in [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030)  
-by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo. 
+The Swin Transformer was proposed in [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://arxiv.org/abs/2103.14030)
+by Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo.
 
 The abstract from the paper is the following:
 
-*This paper presents a new vision Transformer, called Swin Transformer, that capably serves as a general-purpose backbone 
-for computer vision. Challenges in adapting Transformer from language to vision arise from differences between the two domains, 
-such as large variations in the scale of visual entities and the high resolution of pixels in images compared to words in text. 
-To address these differences, we propose a hierarchical Transformer whose representation is computed with \bold{S}hifted 
-\bold{win}dows. The shifted windowing scheme brings greater efficiency by limiting self-attention computation to non-overlapping 
-local windows while also allowing for cross-window connection. This hierarchical architecture has the flexibility to model at 
-various scales and has linear computational complexity with respect to image size. These qualities of Swin Transformer make it 
-compatible with a broad range of vision tasks, including image classification (87.3 top-1 accuracy on ImageNet-1K) and dense 
-prediction tasks such as object detection (58.7 box AP and 51.1 mask AP on COCO test-dev) and semantic segmentation 
-(53.5 mIoU on ADE20K val). Its performance surpasses the previous state-of-the-art by a large margin of +2.7 box AP and 
-+2.6 mask AP on COCO, and +3.2 mIoU on ADE20K, demonstrating the potential of Transformer-based models as vision backbones. 
+*This paper presents a new vision Transformer, called Swin Transformer, that capably serves as a general-purpose backbone
+for computer vision. Challenges in adapting Transformer from language to vision arise from differences between the two domains,
+such as large variations in the scale of visual entities and the high resolution of pixels in images compared to words in text.
+To address these differences, we propose a hierarchical Transformer whose representation is computed with \bold{S}hifted
+\bold{win}dows. The shifted windowing scheme brings greater efficiency by limiting self-attention computation to non-overlapping
+local windows while also allowing for cross-window connection. This hierarchical architecture has the flexibility to model at
+various scales and has linear computational complexity with respect to image size. These qualities of Swin Transformer make it
+compatible with a broad range of vision tasks, including image classification (87.3 top-1 accuracy on ImageNet-1K) and dense
+prediction tasks such as object detection (58.7 box AP and 51.1 mask AP on COCO test-dev) and semantic segmentation
+(53.5 mIoU on ADE20K val). Its performance surpasses the previous state-of-the-art by a large margin of +2.7 box AP and
++2.6 mask AP on COCO, and +3.2 mIoU on ADE20K, demonstrating the potential of Transformer-based models as vision backbones.
 The hierarchical design and the shifted window approach also prove beneficial for all-MLP architectures.*
 
 Tips:
@@ -38,11 +38,11 @@ Tips:
 - Swin can be used as a *backbone*. When `output_hidden_states = True`, it will output both `hidden_states` and `reshaped_hidden_states`. The `reshaped_hidden_states` have a shape of `(batch, num_channels, height, width)` rather than `(batch_size, sequence_length, num_channels)`.
 
 <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/swin_transformer_architecture.png"
-alt="drawing" width="600"/> 
+alt="drawing" width="600"/>
 
 <small> Swin Transformer architecture. Taken from the <a href="https://arxiv.org/abs/2102.03334">original paper</a>.</small>
 
-This model was contributed by [novice03](https://huggingface.co/novice03>). The original code can be found [here](https://github.com/microsoft/Swin-Transformer).
+This model was contributed by [novice03](https://huggingface.co/novice03>). The Tensorflow version of this model was contributed by [amyeroberts](https://huggingface.co/amyeroberts). The original code can be found [here](https://github.com/microsoft/Swin-Transformer).
 
 
 ## SwinConfig
@@ -63,4 +63,19 @@ This model was contributed by [novice03](https://huggingface.co/novice03>). The
 ## SwinForImageClassification
 
 [[autodoc]] transformers.SwinForImageClassification
-    - forward
+    - forward
+
+## TFSwinModel
+
+[[autodoc]] TFSwinModel
+    - call
+
+## TFSwinForMaskedImageModeling
+
+[[autodoc]] TFSwinForMaskedImageModeling
+    - call
+
+## TFSwinForImageClassification
+
+[[autodoc]] transformers.TFSwinForImageClassification
+    - call

docs/source/en/model_doc/t5.mdx

@@ -252,10 +252,9 @@ The example above only shows a single example. You can also do batched inference
 >>> model = T5ForConditionalGeneration.from_pretrained("t5-small")
 
 >>> task_prefix = "translate English to German: "
->>> sentences = [
-...     "The house is wonderful.",
-...     "I like to work in NYC.",
->>> ]  # use different length sentences to test batching
+>>> # use different length sentences to test batching
+>>> sentences = ["The house is wonderful.", "I like to work in NYC."]
+
 >>> inputs = tokenizer([task_prefix + sentence for sentence in sentences], return_tensors="pt", padding=True)
 
 >>> output_sequences = model.generate(
@@ -372,3 +371,8 @@ T5 is supported by several example scripts, both for pre-training and fine-tunin
     - __call__
     - encode
     - decode
+
+## FlaxT5EncoderModel
+
+[[autodoc]] FlaxT5EncoderModel
+    - __call__

docs/source/en/model_doc/trajectory_transformer.mdx

@@ -0,0 +1,49 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# Trajectory Transformer
+
+## Overview
+
+The Trajectory Transformer model was proposed in [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://arxiv.org/abs/2106.02039)  by Michael Janner, Qiyang Li, Sergey Levine.
+
+The abstract from the paper is the following:
+
+*Reinforcement learning (RL) is typically concerned with estimating stationary policies or single-step models,
+leveraging the Markov property to factorize problems in time. However, we can also view RL as a generic sequence
+modeling problem, with the goal being to produce a sequence of actions that leads to a sequence of high rewards.
+Viewed in this way, it is tempting to consider whether high-capacity sequence prediction models that work well
+in other domains, such as natural-language processing, can also provide effective solutions to the RL problem.
+To this end, we explore how RL can be tackled with the tools of sequence modeling, using a Transformer architecture
+to model distributions over trajectories and repurposing beam search as a planning algorithm. Framing RL as sequence
+modeling problem simplifies a range of design decisions, allowing us to dispense with many of the components common
+in offline RL algorithms. We demonstrate the flexibility of this approach across long-horizon dynamics prediction,
+imitation learning, goal-conditioned RL, and offline RL. Further, we show that this approach can be combined with
+existing model-free algorithms to yield a state-of-the-art planner in sparse-reward, long-horizon tasks.*
+
+Tips:
+
+This Transformer is used for deep reinforcement learning. To use it, you need to create sequences from
+actions, states and rewards from all previous timesteps. This model will treat all these elements together
+as one big sequence (a trajectory).
+
+This model was contributed by [CarlCochet](https://huggingface.co/CarlCochet). The original code can be found [here](https://github.com/jannerm/trajectory-transformer).
+
+## TrajectoryTransformerConfig
+
+[[autodoc]] TrajectoryTransformerConfig
+
+
+## TrajectoryTransformerModel
+
+[[autodoc]] TrajectoryTransformerModel
+    - forward

docs/source/en/model_doc/ul2.mdx

@@ -0,0 +1,31 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# UL2
+
+## Overview
+
+The T5 model was presented in [Unifying Language Learning Paradigms](https://arxiv.org/pdf/2205.05131v1.pdf) by Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler.
+
+The abstract from the paper is the following:
+
+*Existing pre-trained models are generally geared towards a particular class of problems. To date, there seems to be still no consensus on what the right architecture and pre-training setup should be. This paper presents a unified framework for pre-training models that are universally effective across datasets and setups. We begin by disentangling architectural archetypes with pre-training objectives -- two concepts that are commonly conflated. Next, we present a generalized and unified perspective for self-supervision in NLP and show how different pre-training objectives can be cast as one another and how interpolating between different objectives can be effective. We then propose Mixture-of-Denoisers (MoD), a pre-training objective that combines diverse pre-training paradigms together. We furthermore introduce a notion of mode switching, wherein downstream fine-tuning is associated with specific pre-training schemes. We conduct extensive ablative experiments to compare multiple pre-training objectives and find that our method pushes the Pareto-frontier by outperforming T5 and/or GPT-like models across multiple diverse setups. Finally, by scaling our model up to 20B parameters, we achieve SOTA performance on 50 well-established supervised NLP tasks ranging from language generation (with automated and human evaluation), language understanding, text classification, question answering, commonsense reasoning, long text reasoning, structured knowledge grounding and information retrieval. Our model also achieve strong results at in-context learning, outperforming 175B GPT-3 on zero-shot SuperGLUE and tripling the performance of T5-XXL on one-shot summarization.*
+
+Tips:
+
+- UL2 is an encoder-decoder model pre-trained on a mixture of denoising functions as well as fine-tuned on an array of downstream tasks.
+- UL2 has the same architecture as [T5v1.1](t5v1.1) but uses the Gated-SiLU activation function instead of Gated-GELU.
+- The authors release checkpoints of one architecture which can be seen [here](https://huggingface.co/google/ul2)
+
+The original code can be found [here](https://github.com/google-research/google-research/tree/master/ul2).
+
+This model was contributed by [DanielHesslow](https://huggingface.co/Seledorn).

docs/source/en/model_doc/unispeech-sat.mdx

@@ -51,8 +51,6 @@ found [here](https://github.com/microsoft/UniSpeech/tree/main/UniSpeech-SAT).
 
 ## UniSpeechSat specific outputs
 
-[[autodoc]] models.unispeech_sat.modeling_unispeech_sat.UniSpeechSatBaseModelOutput
-
 [[autodoc]] models.unispeech_sat.modeling_unispeech_sat.UniSpeechSatForPreTrainingOutput
 
 ## UniSpeechSatModel