Release v4.25.1

* Fix Data2VecTextForCasualLM example code documentation

* Change RobertaTokenizer to AutoTokenizer in data2vectext example code

.circleci/create_circleci_config.py

@@ -0,0 +1,406 @@
+# coding=utf-8
+# Copyright 2022 The HuggingFace Inc. team.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import argparse
+import copy
+import os
+from dataclasses import dataclass
+from typing import Any, Dict, List, Optional
+
+import yaml
+
+
+COMMON_ENV_VARIABLES = {"OMP_NUM_THREADS": 1, "TRANSFORMERS_IS_CI": True, "PYTEST_TIMEOUT": 120}
+COMMON_PYTEST_OPTIONS = {"max-worker-restart": 0, "dist": "loadfile", "s": None}
+DEFAULT_DOCKER_IMAGE = [{"image": "cimg/python:3.7.12"}]
+
+
+@dataclass
+class CircleCIJob:
+    name: str
+    additional_env: Dict[str, Any] = None
+    cache_name: str = None
+    cache_version: str = "0.5"
+    docker_image: List[Dict[str, str]] = None
+    install_steps: List[str] = None
+    marker: Optional[str] = None
+    parallelism: Optional[int] = 1
+    pytest_num_workers: int = 8
+    pytest_options: Dict[str, Any] = None
+    resource_class: Optional[str] = "xlarge"
+    tests_to_run: Optional[List[str]] = None
+    working_directory: str = "~/transformers"
+
+    def __post_init__(self):
+        # Deal with defaults for mutable attributes.
+        if self.additional_env is None:
+            self.additional_env = {}
+        if self.cache_name is None:
+            self.cache_name = self.name
+        if self.docker_image is None:
+            # Let's avoid changing the default list and make a copy.
+            self.docker_image = copy.deepcopy(DEFAULT_DOCKER_IMAGE)
+        if self.install_steps is None:
+            self.install_steps = []
+        if self.pytest_options is None:
+            self.pytest_options = {}
+        if isinstance(self.tests_to_run, str):
+            self.tests_to_run = [self.tests_to_run]
+
+    def to_dict(self):
+        job = {
+            "working_directory": self.working_directory,
+            "docker": self.docker_image,
+            "environment": {**COMMON_ENV_VARIABLES, **self.additional_env},
+        }
+        if self.resource_class is not None:
+            job["resource_class"] = self.resource_class
+        if self.parallelism is not None:
+            job["parallelism"] = self.parallelism
+        steps = [
+            "checkout",
+            {"attach_workspace": {"at": "~/transformers/test_preparation"}},
+            {
+                "restore_cache": {
+                    "keys": [
+                        f"v{self.cache_version}-{self.cache_name}-" + '{{ checksum "setup.py" }}',
+                        f"v{self.cache_version}-{self.cache_name}-",
+                    ]
+                }
+            },
+        ]
+        steps.extend([{"run": l} for l in self.install_steps])
+        steps.append(
+            {
+                "save_cache": {
+                    "key": f"v{self.cache_version}-{self.cache_name}-" + '{{ checksum "setup.py" }}',
+                    "paths": ["~/.cache/pip"],
+                }
+            }
+        )
+        steps.append({"run": {"name": "Show installed libraries and their versions", "command": "pip freeze | tee installed.txt"}})
+        steps.append({"store_artifacts": {"path": "~/transformers/installed.txt"}})
+
+        all_options = {**COMMON_PYTEST_OPTIONS, **self.pytest_options}
+        pytest_flags = [f"--{key}={value}" if value is not None else f"-{key}" for key, value in all_options.items()]
+        pytest_flags.append(
+            f"--make-reports={self.name}" if "examples" in self.name else f"--make-reports=tests_{self.name}"
+        )
+        test_command = f"python -m pytest -n {self.pytest_num_workers} " + " ".join(pytest_flags)
+        if self.tests_to_run is None:
+            test_command += " << pipeline.parameters.tests_to_run >>"
+        else:
+            test_command += " " + " ".join(self.tests_to_run)
+        if self.marker is not None:
+            test_command += f" -m {self.marker}"
+        test_command += " | tee tests_output.txt"
+        steps.append({"run": {"name": "Run tests", "command": test_command}})
+        steps.append({"store_artifacts": {"path": "~/transformers/tests_output.txt"}})
+        steps.append({"store_artifacts": {"path": "~/transformers/reports"}})
+        job["steps"] = steps
+        return job
+
+    @property
+    def job_name(self):
+        return self.name if "examples" in self.name else f"tests_{self.name}"
+
+
+# JOBS
+torch_and_tf_job = CircleCIJob(
+    "torch_and_tf",
+    additional_env={"RUN_PT_TF_CROSS_TESTS": True},
+    install_steps=[
+        "sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng git-lfs",
+        "git lfs install",
+        "pip install --upgrade pip",
+        "pip install .[sklearn,tf-cpu,torch,testing,sentencepiece,torch-speech,vision]",
+        "pip install tensorflow_probability",
+        "pip install git+https://github.com/huggingface/accelerate",
+    ],
+    marker="is_pt_tf_cross_test",
+    pytest_options={"rA": None, "durations": 0},
+)
+
+
+torch_and_flax_job = CircleCIJob(
+    "torch_and_flax",
+    additional_env={"RUN_PT_FLAX_CROSS_TESTS": True},
+    install_steps=[
+        "sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng",
+        "pip install --upgrade pip",
+        "pip install .[sklearn,flax,torch,testing,sentencepiece,torch-speech,vision]",
+        "pip install git+https://github.com/huggingface/accelerate",
+    ],
+    marker="is_pt_flax_cross_test",
+    pytest_options={"rA": None, "durations": 0},
+)
+
+
+torch_job = CircleCIJob(
+    "torch",
+    install_steps=[
+        "sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng time",
+        "pip install --upgrade pip",
+        "pip install .[sklearn,torch,testing,sentencepiece,torch-speech,vision,timm]",
+        "pip install git+https://github.com/huggingface/accelerate",
+    ],
+    pytest_num_workers=3,
+)
+
+
+tf_job = CircleCIJob(
+    "tf",
+    install_steps=[
+        "sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng",
+        "pip install --upgrade pip",
+        "pip install .[sklearn,tf-cpu,testing,sentencepiece,tf-speech,vision]",
+        "pip install tensorflow_probability",
+    ],
+    pytest_options={"rA": None},
+)
+
+
+flax_job = CircleCIJob(
+    "flax",
+    install_steps=[
+        "sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng",
+        "pip install --upgrade pip",
+        "pip install .[flax,testing,sentencepiece,flax-speech,vision]",
+    ],
+    pytest_options={"rA": None},
+)
+
+
+pipelines_torch_job = CircleCIJob(
+    "pipelines_torch",
+    install_steps=[
+        "sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng",
+        "pip install --upgrade pip",
+        "pip install .[sklearn,torch,testing,sentencepiece,torch-speech,vision,timm]",
+    ],
+    pytest_options={"rA": None},
+    tests_to_run="tests/pipelines/"
+)
+
+
+pipelines_tf_job = CircleCIJob(
+    "pipelines_tf",
+    install_steps=[
+        "pip install --upgrade pip",
+        "pip install .[sklearn,tf-cpu,testing,sentencepiece]",
+        "pip install tensorflow_probability",
+    ],
+    pytest_options={"rA": None},
+    tests_to_run="tests/pipelines/"
+)
+
+
+custom_tokenizers_job = CircleCIJob(
+    "custom_tokenizers",
+    additional_env={"RUN_CUSTOM_TOKENIZERS": True},
+    install_steps=[
+        "sudo apt-get -y update && sudo apt-get install -y cmake",
+        {
+            "name": "install jumanpp",
+            "command":
+                "wget https://github.com/ku-nlp/jumanpp/releases/download/v2.0.0-rc3/jumanpp-2.0.0-rc3.tar.xz\n"
+                "tar xvf jumanpp-2.0.0-rc3.tar.xz\n"
+                "mkdir jumanpp-2.0.0-rc3/bld\n"
+                "cd jumanpp-2.0.0-rc3/bld\n"
+                "sudo cmake .. -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr/local\n"
+                "sudo make install\n",
+        },
+        "pip install --upgrade pip",
+        "pip install .[ja,testing,sentencepiece,jieba,spacy,ftfy,rjieba]",
+        "python -m unidic download",
+    ],
+    parallelism=None,
+    resource_class=None,
+    tests_to_run=[
+        "./tests/models/bert_japanese/test_tokenization_bert_japanese.py",
+        "./tests/models/openai/test_tokenization_openai.py",
+        "./tests/models/clip/test_tokenization_clip.py",
+    ],
+)
+
+
+examples_torch_job = CircleCIJob(
+    "examples_torch",
+    cache_name="torch_examples",
+    install_steps=[
+        "sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev espeak-ng",
+        "pip install --upgrade pip",
+        "pip install .[sklearn,torch,sentencepiece,testing,torch-speech]",
+        "pip install -r examples/pytorch/_tests_requirements.txt",
+    ],
+    tests_to_run="./examples/pytorch/",
+)
+
+
+examples_tensorflow_job = CircleCIJob(
+    "examples_tensorflow",
+    cache_name="tensorflow_examples",
+    install_steps=[
+        "pip install --upgrade pip",
+        "pip install .[sklearn,tensorflow,sentencepiece,testing]",
+        "pip install -r examples/tensorflow/_tests_requirements.txt",
+    ],
+    tests_to_run="./examples/tensorflow/",
+)
+
+
+examples_flax_job = CircleCIJob(
+    "examples_flax",
+    cache_name="flax_examples",
+    install_steps=[
+        "pip install --upgrade pip",
+        "pip install .[flax,testing,sentencepiece]",
+        "pip install -r examples/flax/_tests_requirements.txt",
+    ],
+    tests_to_run="./examples/flax/",
+)
+
+
+hub_job = CircleCIJob(
+    "hub",
+    install_steps=[
+        "sudo apt-get -y update && sudo apt-get install git-lfs",
+        'git config --global user.email "ci@dummy.com"',
+        'git config --global user.name "ci"',
+        "pip install --upgrade pip",
+        "pip install .[torch,sentencepiece,testing]",
+    ],
+    marker="is_staging_test",
+    pytest_num_workers=1,
+)
+
+
+onnx_job = CircleCIJob(
+    "onnx",
+    install_steps=[
+        "pip install --upgrade pip",
+        "pip install .[torch,tf,testing,sentencepiece,onnxruntime,vision,rjieba]",
+    ],
+    pytest_options={"k onnx": None},
+    pytest_num_workers=1,
+)
+
+
+layoutlm_job = CircleCIJob(
+    "layoutlmv2_and_v3",
+    install_steps=[
+        "sudo apt-get -y update && sudo apt-get install -y libsndfile1-dev",
+        "pip install --upgrade pip",
+        "pip install .[torch,testing,vision]",
+        "pip install torchvision",
+        "pip install 'git+https://github.com/facebookresearch/detectron2.git'",
+        "sudo apt install tesseract-ocr",
+        "pip install pytesseract",
+        "pip install natten",
+    ],
+    tests_to_run=[
+        "tests/models/*layoutlmv*",
+        "tests/models/*nat",
+    ],
+    pytest_num_workers=1,
+    pytest_options={"durations": 100},
+)
+
+
+repo_utils_job = CircleCIJob(
+    "repo_utils",
+    install_steps=[
+        "pip install --upgrade pip",
+        "pip install .[quality,testing]",
+    ],
+    parallelism=None,
+    pytest_num_workers=1,
+    resource_class=None,
+    tests_to_run="tests/repo_utils",
+)
+
+REGULAR_TESTS = [
+    torch_and_tf_job,
+    torch_and_flax_job,
+    torch_job,
+    tf_job,
+    flax_job,
+    custom_tokenizers_job,
+    hub_job,
+    onnx_job,
+    layoutlm_job,
+]
+EXAMPLES_TESTS = [
+    examples_torch_job,
+    examples_tensorflow_job,
+    examples_flax_job,
+]
+PIPELINE_TESTS = [
+    pipelines_torch_job,
+    pipelines_tf_job,
+]
+REPO_UTIL_TESTS = [repo_utils_job]
+
+def create_circleci_config(folder=None):
+    if folder is None:
+        folder = os.getcwd()
+    jobs = []
+    all_test_file = os.path.join(folder, "test_list.txt")
+    if os.path.exists(all_test_file):
+        with open(all_test_file) as f:
+            all_test_list = f.read()
+    else:
+        all_test_list = []
+    if len(all_test_list) > 0:
+        jobs.extend(PIPELINE_TESTS)
+
+    test_file = os.path.join(folder, "filtered_test_list.txt")
+    if os.path.exists(test_file):
+        with open(test_file) as f:
+            test_list = f.read()
+    else:
+        test_list = []
+    if len(test_list) > 0:
+        jobs.extend(REGULAR_TESTS)
+
+    example_file = os.path.join(folder, "examples_test_list.txt")
+    if os.path.exists(example_file) and os.path.getsize(example_file) > 0:
+        jobs.extend(EXAMPLES_TESTS)
+    
+    repo_util_file = os.path.join(folder, "test_repo_utils.txt")
+    if os.path.exists(repo_util_file) and os.path.getsize(repo_util_file) > 0:
+        jobs.extend(REPO_UTIL_TESTS)
+
+    if len(jobs) > 0:
+        config = {"version": "2.1"}
+        config["parameters"] = {
+            # Only used to accept the parameters from the trigger
+            "nightly": {"type": "boolean", "default": False},
+            "tests_to_run": {"type": "string", "default": test_list},
+        }
+        config["jobs"] = {j.job_name: j.to_dict() for j in jobs}
+        config["workflows"] = {"version": 2, "run_tests": {"jobs": [j.job_name for j in jobs]}}
+        with open(os.path.join(folder, "generated_config.yml"), "w") as f:
+            f.write(yaml.dump(config, indent=2, width=1000000, sort_keys=False))
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--fetcher_folder", type=str, default=None, help="Only test that all tests and modules are accounted for."
+    )
+    args = parser.parse_args()
+
+    create_circleci_config(args.fetcher_folder)

.github/ISSUE_TEMPLATE/bug-report.yml

@@ -1,6 +1,5 @@
 name: "\U0001F41B Bug Report"
 description: Submit a bug report to help us improve transformers
-labels: [ "bug" ]
 body:
   - type: textarea
     id: system-info

.github/ISSUE_TEMPLATE/i18n.md

@@ -0,0 +1,46 @@
+---
+name: 🌐 Translating a new language?
+about: Start a new translation effort in your language
+title: '[i18n-<languageCode>] Translating docs to <languageName>'
+labels: WIP
+assignees: ''
+
+---
+
+<!--
+Note: Please search to see if an issue already exists for the language you are trying to translate.
+-->
+
+Hi!
+
+Let's bring the documentation to all the <languageName>-speaking community 🌐 (currently 0 out of 267 complete)
+
+Who would want to translate? Please follow the 🤗 [TRANSLATING guide](https://github.com/huggingface/transformers/blob/main/docs/TRANSLATING.md). Here is a list of the files ready for translation. Let us know in this issue if you'd like to translate any, and we'll add your name to the list.
+
+Some notes:
+
+* Please translate using an informal tone (imagine you are talking with a friend about transformers 🤗).
+* Please translate in a gender-neutral way.
+* Add your translations to the folder called `<languageCode>` inside the [source folder](https://github.com/huggingface/transformers/tree/main/docs/source).
+* Register your translation in `<languageCode>/_toctree.yml`; please follow the order of the [English version](https://github.com/huggingface/transformers/blob/main/docs/source/en/_toctree.yml).
+* Once you're finished, open a pull request and tag this issue by including #issue-number in the description, where issue-number is the number of this issue. Please ping @ArthurZucker, @sgugger for review.
+* 🙋 If you'd like others to help you with the translation, you can also post in the 🤗 [forums](https://discuss.huggingface.co/).
+
+## Get Started section
+
+- [ ] [index.mdx](https://github.com/huggingface/transformers/blob/main/docs/source/en/index.mdx) https://github.com/huggingface/transformers/pull/20180
+- [ ] [quicktour.mdx](https://github.com/huggingface/transformers/blob/main/docs/source/en/quicktour.mdx) (waiting for initial PR to go through)
+- [ ] [installation.mdx](https://github.com/huggingface/transformers/blob/main/docs/source/en/installation.mdx).
+
+## Tutorial section
+- [ ] [pipeline_tutorial.mdx](https://github.com/huggingface/transformers/blob/main/docs/source/en/pipeline_tutorial.mdx)
+- [ ]  [autoclass_tutorial.mdx](https://github.com/huggingface/transformers/blob/master/docs/source/autoclass_tutorial.mdx)
+- [ ]  [preprocessing.mdx](https://github.com/huggingface/transformers/blob/main/docs/source/en/preprocessing.mdx)
+- [ ]  [training.mdx](https://github.com/huggingface/transformers/blob/main/docs/source/en/training.mdx)
+- [ ]  [accelerate.mdx](https://github.com/huggingface/transformers/blob/main/docs/source/en/accelerate.mdx)
+- [ ]  [model_sharing.mdx](https://github.com/huggingface/transformers/blob/main/docs/source/en/model_sharing.mdx)
+- [ ]  [multilingual.mdx](https://github.com/huggingface/transformers/blob/main/docs/source/en/multilingual.mdx)
+
+<!--
+Keep on adding more as you go 🔥
+-->

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

@@ -6,6 +6,10 @@ on:
       - docker-image*
   repository_dispatch:
   workflow_call:
+    inputs:
+      image_postfix:
+        required: true
+        type: string
   schedule:
     - cron: "0 1 * * *"
 
@@ -23,7 +27,7 @@ jobs:
         uses: docker/setup-buildx-action@v1
       -
         name: Check out code
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
       -
         name: Login to DockerHub
         uses: docker/login-action@v1
@@ -38,10 +42,25 @@ jobs:
           build-args: |
             REF=main
           push: true
-          tags: huggingface/transformers-all-latest-gpu
+          tags: huggingface/transformers-all-latest-gpu${{ inputs.image_postfix }}
+      # Push CI images still need to be re-built daily
+      -
+        name: Build and push (for Push CI) in a daily basis
+        # This condition allows `schedule` events, or `push` events that trigger this workflow NOT via `workflow_call`.
+        # The later case is useful for manual image building for debugging purpose. Use another tag in this case!
+        if: inputs.image_postfix != '-push-ci'
+        uses: docker/build-push-action@v2
+        with:
+          context: ./docker/transformers-all-latest-gpu
+          build-args: |
+            REF=main
+          push: true
+          tags: huggingface/transformers-all-latest-gpu-push-ci
 
   latest-with-torch-nightly-docker:
     name: "Nightly PyTorch + Stable TensorFlow"
+    # Push CI doesn't need this image
+    if: inputs.image_postfix != '-push-ci'
     runs-on: ubuntu-latest
     steps:
       -
@@ -91,10 +110,46 @@ jobs:
           build-args: |
             REF=main
           push: true
-          tags: huggingface/transformers-pytorch-deepspeed-latest-gpu
+          tags: huggingface/transformers-pytorch-deepspeed-latest-gpu${{ inputs.image_postfix }}
+
+  # Can't build 2 images in a single job `latest-torch-deepspeed-docker` (for `nvcr.io/nvidia`)
+  latest-torch-deepspeed-docker-for-push-ci-daily-build:
+    name: "Latest PyTorch + DeepSpeed (Push CI - Daily Build)"
+    # Can't run in parallel, otherwise get an error:
+    #   `Error response from daemon: Get "https://registry-1.docker.io/v2/": received unexpected HTTP status: 503 Service Unavailable`
+    needs: latest-torch-deepspeed-docker
+    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 }}
+      # Push CI images still need to be re-built daily
+      -
+        name: Build and push (for Push CI) in a daily basis
+        # This condition allows `schedule` events, or `push` events that trigger this workflow NOT via `workflow_call`.
+        # The later case is useful for manual image building for debugging purpose. Use another tag in this case!
+        if: inputs.image_postfix != '-push-ci'
+        uses: docker/build-push-action@v2
+        with:
+          context: ./docker/transformers-pytorch-deepspeed-latest-gpu
+          build-args: |
+            REF=main
+          push: true
+          tags: huggingface/transformers-pytorch-deepspeed-latest-gpu-push-ci
 
   nightly-torch-deepspeed-docker:
     name: "Nightly PyTorch + DeepSpeed"
+    # Push CI doesn't need this image
+    if: inputs.image_postfix != '-push-ci'
     runs-on: ubuntu-latest
     steps:
       -
@@ -121,6 +176,8 @@ jobs:
 
   doc-builder:
     name: "Doc builder"
+    # Push CI doesn't need this image
+    if: inputs.image_postfix != '-push-ci'
     runs-on: ubuntu-latest
     steps:
       -
@@ -145,6 +202,8 @@ jobs:
 
   latest-pytorch:
     name: "Latest PyTorch [dev]"
+    # Push CI doesn't need this image
+    if: inputs.image_postfix != '-push-ci'
     runs-on: ubuntu-latest
     steps:
       -
@@ -171,6 +230,8 @@ jobs:
 
   latest-tensorflow:
     name: "Latest TensorFlow [dev]"
+    # Push CI doesn't need this image
+    if: inputs.image_postfix != '-push-ci'
     runs-on: ubuntu-latest
     steps:
       -

.github/workflows/build_documentation.yml

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

.github/workflows/check_runner_status.yml

@@ -0,0 +1,67 @@
+name: Self-hosted runner (check runner status)
+
+# 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:
+  repository_dispatch:
+  schedule:
+    # run per hour
+    - cron: "0 */1 * * *"
+
+env:
+  TRANSFORMERS_IS_CI: yes
+
+jobs:
+  check_runner_status:
+    name: Check Runner Status
+    runs-on: ubuntu-latest
+    outputs:
+      offline_runners: ${{ steps.set-offline_runners.outputs.offline_runners }}
+    steps:
+      - name: Checkout transformers
+        uses: actions/checkout@v2
+        with:
+          fetch-depth: 2
+
+      - name: Check Runner Status
+        run: python utils/check_self_hosted_runner.py --target_runners single-gpu-ci-runner-docker,multi-gpu-ci-runner-docker,single-gpu-scheduled-ci-runner-docker,multi-scheduled-scheduled-ci-runner-docker,single-gpu-doctest-ci-runner-docker --token ${{ secrets.ACCESS_REPO_INFO_TOKEN }}
+
+      - id: set-offline_runners
+        name: Set output for offline runners
+        if: ${{ always() }}
+        run: |
+          offline_runners=$(python3 -c 'fp = open("offline_runners.txt"); failed = fp.read(); fp.close(); print(failed)')
+          echo "::set-output name=offline_runners::$offline_runners"
+
+  send_results:
+    name: Send results to webhook
+    runs-on: ubuntu-latest
+    needs: check_runner_status
+    if: ${{ failure() }}
+    steps:
+      - name: Preliminary job status
+        shell: bash
+        run: |
+          echo "Runner availability: ${{ needs.check_runner_status.result }}"
+
+      - 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_DAILY }}
+          CI_EVENT: runner status check
+          RUNNER_STATUS: ${{ needs.check_runner_status.result }}
+          OFFLINE_RUNNERS: ${{ needs.check_runner_status.outputs.offline_runners }}
+        # 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

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

@@ -23,8 +23,21 @@ env:
   RUN_PT_TF_CROSS_TESTS: 1
 
 jobs:
-  run_check_runners:
+  check_runner_status:
+    name: Check Runner Status
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout transformers
+        uses: actions/checkout@v2
+        with:
+          fetch-depth: 2
+
+      - name: Check Runner Status
+        run: python utils/check_self_hosted_runner.py --target_runners single-gpu-scheduled-ci-runner-docker,multi-gpu-scheduled-ci-runner-docker --token ${{ secrets.ACCESS_REPO_INFO_TOKEN }}
+
+  check_runners:
     name: Check Runners
+    needs: check_runner_status
     strategy:
       matrix:
         machine_type: [single-gpu, multi-gpu]
@@ -39,7 +52,7 @@ jobs:
 
   setup:
     name: Setup
-    needs: run_check_runners
+    needs: check_runners
     strategy:
       matrix:
         machine_type: [single-gpu, multi-gpu]
@@ -62,6 +75,10 @@ jobs:
           rm -rf tests/models/__pycache__
           rm -rf reports
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - id: set-matrix
         name: Identify models to test
         working-directory: /transformers/tests
@@ -83,7 +100,7 @@ jobs:
     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: [run_check_runners, setup]
+    needs: setup
     steps:
       - name: Echo folder ${{ matrix.folders }}
         shell: bash
@@ -109,6 +126,10 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - 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 }}
@@ -136,7 +157,7 @@ jobs:
     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: [run_check_runners, setup]
+    needs: setup
     steps:
       - name: Echo folder ${{ matrix.folders }}
         shell: bash
@@ -162,6 +183,10 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - 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 }}
@@ -185,7 +210,7 @@ jobs:
       matrix:
         machine_type: [single-gpu, multi-gpu]
     runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
-    needs: [run_check_runners, setup]
+    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/
@@ -215,6 +240,10 @@ jobs:
         run: |
           python utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /workspace/transformers
+        run: pip freeze
+
       - name: Run all tests on GPU
         working-directory: /workspace/transformers
         run: |
@@ -236,13 +265,21 @@ jobs:
     name: Send results to webhook
     runs-on: ubuntu-latest
     if: always()
-    needs: [run_check_runners, setup, run_tests_single_gpu, run_tests_multi_gpu, run_all_tests_torch_cuda_extensions_gpu]
+    needs: [
+      check_runner_status,
+      check_runners,
+      setup,
+      run_tests_single_gpu,
+      run_tests_multi_gpu,
+      run_all_tests_torch_cuda_extensions_gpu
+    ]
     steps:
       - name: Preliminary job status
         shell: bash
         # For the meaning of these environment variables, see the job `Setup`
         run: |
-          echo "Runner status: ${{ needs.run_check_runners.result }}"
+          echo "Runner availability: ${{ needs.check_runner_status.result }}"
+          echo "Runner status: ${{ needs.check_runners.result }}"
           echo "Setup status: ${{ needs.setup.result }}"
 
       - uses: actions/checkout@v2
@@ -255,10 +292,12 @@ jobs:
           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
+          RUNNER_STATUS: ${{ needs.check_runner_status.result }}
+          RUNNER_ENV_STATUS: ${{ needs.check_runners.result }}
           SETUP_STATUS: ${{ needs.setup.result }}
-          RUNNER_STATUS: ${{ needs.run_check_runners.result }}
         # We pass `needs.setup.outputs.matrix` as the argument. A processing in `notification_service.py` to change
         # `models/bert` to `models_bert` is required, as the artifact names use `_` instead of `/`.
         run: |
           pip install slack_sdk
+          pip show slack_sdk
           python utils/notification_service.py "${{ needs.setup.outputs.matrix }}"

.github/workflows/self-past.yml

@@ -15,6 +15,11 @@ on:
       version:
         required: true
         type: string
+      # Use this to control the commit to test against
+      sha:
+        default: 'main'
+        required: false
+        type: string
 
 env:
   HF_HOME: /mnt/cache
@@ -27,32 +32,21 @@ env:
   RUN_PT_TF_CROSS_TESTS: 1
 
 jobs:
-  setup:
-    name: Setup
+  check_runner_status:
+    name: Check Runner Status
     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
+      - name: Check Runner Status
+        run: python utils/check_self_hosted_runner.py --target_runners single-gpu-past-ci-runner-docker,multi-gpu-past-ci-runner-docker --token ${{ secrets.ACCESS_REPO_INFO_TOKEN }}
 
-      - 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_check_runners:
+  check_runners:
     name: Check Runners
-    needs: setup
+    needs: check_runner_status
     strategy:
       matrix:
         machine_type: [single-gpu, multi-gpu]
@@ -65,6 +59,41 @@ jobs:
         run: |
           nvidia-smi
 
+  setup:
+    name: Setup
+    needs: check_runners
+    strategy:
+      matrix:
+        machine_type: [single-gpu, 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 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 ${{ inputs.sha }}
+
+      - name: Cleanup
+        working-directory: /transformers
+        run: |
+          rm -rf tests/__pycache__
+          rm -rf tests/models/__pycache__
+          rm -rf reports
+
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
+      - id: set-matrix
+        working-directory: /transformers
+        name: Identify models to test
+        run: |
+          cd tests
+          echo "::set-output name=matrix::$(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:
@@ -76,11 +105,11 @@ jobs:
     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, run_check_runners]
+    needs: setup
     steps:
       - name: Update clone
         working-directory: /transformers
-        run: git fetch && git checkout ${{ github.sha }}
+        run: git fetch && git checkout ${{ inputs.sha }}
 
       - name: Echo folder ${{ matrix.folders }}
         shell: bash
@@ -102,6 +131,10 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - name: Run all tests on GPU
         working-directory: /transformers
         run: python3 -m pytest -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} tests/${{ matrix.folders }}
@@ -111,6 +144,15 @@ jobs:
         continue-on-error: true
         run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
 
+      - name: Save job name
+        if: ${{ always() }}
+        shell: bash
+        run: |
+          matrix_folders=${matrix_folders/'models_'/'models/'}
+          job_name="Model tests ($matrix_folders, ${{ matrix.machine_type }})"
+          echo "$job_name"
+          echo "$job_name" > /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/job_name.txt
+
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
@@ -129,11 +171,11 @@ jobs:
     container:
       image: huggingface/transformers-${{ inputs.framework }}-past-${{ inputs.version }}-gpu
       options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-    needs: [setup, run_check_runners]
+    needs: setup
     steps:
       - name: Update clone
         working-directory: /transformers
-        run: git fetch && git checkout ${{ github.sha }}
+        run: git fetch && git checkout ${{ inputs.sha }}
 
       - name: Echo folder ${{ matrix.folders }}
         shell: bash
@@ -155,6 +197,10 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - name: Run all tests on GPU
         working-directory: /transformers
         run: python3 -m pytest -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} tests/${{ matrix.folders }}
@@ -164,6 +210,15 @@ jobs:
         continue-on-error: true
         run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
 
+      - name: Save job name
+        if: ${{ always() }}
+        shell: bash
+        run: |
+          matrix_folders=${matrix_folders/'models_'/'models/'}
+          job_name="Model tests ($matrix_folders, ${{ matrix.machine_type }})"
+          echo "$job_name"
+          echo "$job_name" > /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/job_name.txt
+
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
@@ -175,13 +230,14 @@ jobs:
     name: Send results to webhook
     runs-on: ubuntu-latest
     if: always()
-    needs: [setup, run_check_runners, run_tests_single_gpu, run_tests_multi_gpu]
+    needs: [check_runner_status, check_runners, setup, run_tests_single_gpu, run_tests_multi_gpu]
     steps:
       - name: Preliminary job status
         shell: bash
         # For the meaning of these environment variables, see the job `Setup`
         run: |
-          echo "Runner status: ${{ needs.run_check_runners.result }}"
+          echo "Runner availability: ${{ needs.check_runner_status.result }}"
+          echo "Runner status: ${{ needs.check_runners.result }}"
           echo "Setup status: ${{ needs.setup.result }}"
 
       - uses: actions/checkout@v2
@@ -199,12 +255,14 @@ jobs:
           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 }}
+          RUNNER_STATUS: ${{ needs.check_runner_status.result }}
+          RUNNER_ENV_STATUS: ${{ needs.check_runners.result }}
           SETUP_STATUS: ${{ needs.setup.result }}
-          RUNNER_STATUS: ${{ needs.run_check_runners.result }}
         # We pass `needs.setup.outputs.matrix` as the argument. A processing in `notification_service.py` to change
         # `models/bert` to `models_bert` is required, as the artifact names use `_` instead of `/`.
         run: |
           pip install slack_sdk
+          pip show slack_sdk
           python utils/notification_service.py "${{ needs.setup.outputs.matrix }}"
 
       # Upload complete failure tables, as they might be big and only truncated versions could be sent to Slack.

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

@@ -40,6 +40,8 @@ jobs:
     needs: check-for-setup
     if: (github.event_name == 'push') && (needs.check-for-setup.outputs.changed == '1')
     uses: ./.github/workflows/build-docker-images.yml
+    with:
+      image_postfix: "-push-ci"
     secrets: inherit
 
   run_push_ci:

.github/workflows/self-push.yml

@@ -27,9 +27,43 @@ env:
   RUN_PT_TF_CROSS_TESTS: 1
 
 jobs:
+  check_runner_status:
+    name: Check Runner Status
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout transformers
+        uses: actions/checkout@v2
+        with:
+          fetch-depth: 2
+
+      - name: Check Runner Status
+        run: python utils/check_self_hosted_runner.py --target_runners single-gpu-ci-runner-docker,multi-gpu-ci-runner-docker --token ${{ secrets.ACCESS_REPO_INFO_TOKEN }}
+
+  check_runners:
+    name: Check Runners
+    needs: check_runner_status
+    strategy:
+      matrix:
+        machine_type: [single-gpu, multi-gpu]
+    runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
+    container:
+      image: huggingface/transformers-all-latest-gpu-push-ci
+      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+    steps:
+      - name: NVIDIA-SMI
+        run: |
+          nvidia-smi
+
   setup:
     name: Setup
-    runs-on: ubuntu-latest
+    needs: check_runners
+    strategy:
+      matrix:
+        machine_type: [single-gpu, multi-gpu]
+    runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
+    container:
+      image: huggingface/transformers-all-latest-gpu-push-ci
+      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     outputs:
       matrix: ${{ steps.set-matrix.outputs.matrix }}
       test_map: ${{ steps.set-matrix.outputs.test_map }}
@@ -62,12 +96,8 @@ jobs:
           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: Update clone using environment variables
+        working-directory: /transformers
         run: |
           echo "original branch = $(git branch --show-current)"
           git fetch && git checkout ${{ env.CI_BRANCH }}
@@ -76,25 +106,32 @@ jobs:
           echo "log = $(git log -n 1)"
 
       - name: Cleanup
+        working-directory: /transformers
         run: |
           rm -rf tests/__pycache__
           rm -rf tests/models/__pycache__
           rm -rf reports
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - name: Fetch the tests to run
+        working-directory: /transformers
         # 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
+          python3 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
+          path: /transformers/test_preparation.txt
 
       - id: set-matrix
         name: Organize tests into models
+        working-directory: /transformers
         # 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)
@@ -111,24 +148,9 @@ jobs:
           echo "::set-output name=matrix::$keys"
           echo "::set-output name=test_map::$test_map"
 
-  run_check_runners:
-    name: Check Runners
-    needs: setup
-    strategy:
-      matrix:
-        machine_type: [single-gpu, multi-gpu]
-    runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
-    container:
-      image: huggingface/transformers-all-latest-gpu
-      options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-    steps:
-      - name: NVIDIA-SMI
-        run: |
-          nvidia-smi
-
   run_tests_single_gpu:
     name: Model tests
-    needs: [setup, run_check_runners]
+    needs: setup
     # `dummy` means there is no test to run
     if: contains(fromJson(needs.setup.outputs.matrix), 'dummy') != true
     strategy:
@@ -138,7 +160,7 @@ jobs:
         machine_type: [single-gpu]
     runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
     container:
-      image: huggingface/transformers-all-latest-gpu
+      image: huggingface/transformers-all-latest-gpu-push-ci
       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
@@ -194,6 +216,10 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - name: Run all non-slow selected tests on GPU
         working-directory: /transformers
         run: |
@@ -213,7 +239,7 @@ jobs:
 
   run_tests_multi_gpu:
     name: Model tests
-    needs: [setup, run_check_runners]
+    needs: setup
     # `dummy` means there is no test to run
     if: contains(fromJson(needs.setup.outputs.matrix), 'dummy') != true
     strategy:
@@ -223,7 +249,7 @@ jobs:
         machine_type: [multi-gpu]
     runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
     container:
-      image: huggingface/transformers-all-latest-gpu
+      image: huggingface/transformers-all-latest-gpu-push-ci
       options: --gpus all --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
@@ -279,6 +305,10 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - name: Run all non-slow selected tests on GPU
         env:
           MKL_SERVICE_FORCE_INTEL: 1
@@ -300,7 +330,7 @@ jobs:
 
   run_tests_torch_cuda_extensions_single_gpu:
     name: Torch CUDA extension tests
-    needs: [setup, run_check_runners]
+    needs: setup
     if: contains(fromJson(needs.setup.outputs.matrix), 'deepspeed') || contains(fromJson(needs.setup.outputs.matrix), 'extended')
     strategy:
       fail-fast: false
@@ -308,7 +338,7 @@ jobs:
         machine_type: [single-gpu]
     runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
     container:
-      image: huggingface/transformers-pytorch-deepspeed-latest-gpu
+      image: huggingface/transformers-pytorch-deepspeed-latest-gpu-push-ci
       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
@@ -362,6 +392,10 @@ jobs:
         run: |
           python utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /workspace/transformers
+        run: pip freeze
+
       - 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.
@@ -382,7 +416,7 @@ jobs:
 
   run_tests_torch_cuda_extensions_multi_gpu:
     name: Torch CUDA extension tests
-    needs: [setup, run_check_runners]
+    needs: setup
     if: contains(fromJson(needs.setup.outputs.matrix), 'deepspeed') || contains(fromJson(needs.setup.outputs.matrix), 'extended')
     strategy:
       fail-fast: false
@@ -390,7 +424,7 @@ jobs:
         machine_type: [multi-gpu]
     runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
     container:
-      image: huggingface/transformers-pytorch-deepspeed-latest-gpu
+      image: huggingface/transformers-pytorch-deepspeed-latest-gpu-push-ci
       options: --gpus all --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
@@ -444,6 +478,10 @@ jobs:
         run: |
           python utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /workspace/transformers
+        run: pip freeze
+
       - 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.
@@ -467,8 +505,9 @@ jobs:
     runs-on: ubuntu-latest
     if: always()
     needs: [
+        check_runner_status,
+        check_runners,
         setup,
-        run_check_runners,
         run_tests_single_gpu,
         run_tests_multi_gpu,
         run_tests_torch_cuda_extensions_single_gpu,
@@ -479,8 +518,9 @@ jobs:
         shell: bash
         # For the meaning of these environment variables, see the job `Setup`
         run: |
+          echo "Runner availability: ${{ needs.check_runner_status.result }}"
           echo "Setup status: ${{ needs.setup.result }}"
-          echo "Runner status: ${{ needs.run_check_runners.result }}"
+          echo "Runner status: ${{ needs.check_runners.result }}"
 
       # 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)
@@ -506,6 +546,11 @@ jobs:
           echo "env.CI_SHA = ${{ env.CI_SHA }}"
 
       - uses: actions/checkout@v2
+        # To avoid failure when multiple commits are merged into `main` in a short period of time.
+        # Checking out to an old commit beyond the fetch depth will get an error `fatal: reference is not a tree: ...
+        # (Only required for `workflow_run` event, where we get the latest HEAD on `main` instead of the event commit)
+        with:
+          fetch-depth: 20
 
       - name: Update clone using environment variables
         run: |
@@ -527,11 +572,13 @@ jobs:
           CI_TITLE_PUSH: ${{ github.event.head_commit.message }}
           CI_TITLE_WORKFLOW_RUN: ${{ github.event.workflow_run.head_commit.message }}
           CI_SHA: ${{ env.CI_SHA }}
+          RUNNER_STATUS: ${{ needs.check_runner_status.result }}
+          RUNNER_ENV_STATUS: ${{ needs.check_runners.result }}
           SETUP_STATUS: ${{ needs.setup.result }}
-          RUNNER_STATUS: ${{ needs.run_check_runners.result }}
 
         # 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
+          pip show slack_sdk
           python utils/notification_service.py "${{ needs.setup.outputs.matrix }}"

.github/workflows/self-scheduled.yml

@@ -22,8 +22,21 @@ env:
   RUN_PT_TF_CROSS_TESTS: 1
 
 jobs:
-  run_check_runners:
+  check_runner_status:
+    name: Check Runner Status
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout transformers
+        uses: actions/checkout@v2
+        with:
+          fetch-depth: 2
+
+      - name: Check Runner Status
+        run: python utils/check_self_hosted_runner.py --target_runners single-gpu-scheduled-ci-runner-docker,multi-gpu-scheduled-ci-runner-docker --token ${{ secrets.ACCESS_REPO_INFO_TOKEN }}
+
+  check_runners:
     name: Check Runners
+    needs: check_runner_status
     strategy:
       matrix:
         machine_type: [single-gpu, multi-gpu]
@@ -38,7 +51,7 @@ jobs:
 
   setup:
     name: Setup
-    needs: run_check_runners
+    needs: check_runners
     strategy:
       matrix:
         machine_type: [single-gpu, multi-gpu]
@@ -61,6 +74,10 @@ jobs:
           rm -rf tests/models/__pycache__
           rm -rf reports
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - id: set-matrix
         name: Identify models to test
         working-directory: /transformers/tests
@@ -82,7 +99,7 @@ jobs:
     container:
       image: huggingface/transformers-all-latest-gpu
       options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-    needs: [run_check_runners, setup]
+    needs: setup
     steps:
       - name: Echo folder ${{ matrix.folders }}
         shell: bash
@@ -108,6 +125,10 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - 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 }}
@@ -135,7 +156,7 @@ jobs:
     container:
       image: huggingface/transformers-all-latest-gpu
       options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-    needs: [run_check_runners, setup]
+    needs: setup
     steps:
       - name: Echo folder ${{ matrix.folders }}
         shell: bash
@@ -161,6 +182,10 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - 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 }}
@@ -179,11 +204,15 @@ jobs:
 
   run_examples_gpu:
     name: Examples directory
-    runs-on: [self-hosted, single-gpu-docker]
+    strategy:
+      fail-fast: false
+      matrix:
+        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: [run_check_runners, setup]
+    needs: setup
     steps:
       - name: Update clone
         working-directory: /transformers
@@ -198,23 +227,27 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - name: Run examples tests on GPU
         working-directory: /transformers
         run: |
           pip install -r examples/pytorch/_tests_requirements.txt
-          python3 -m pytest -v --make-reports=single-gpu_examples_gpu examples/pytorch
+          python3 -m pytest -v --make-reports=${{ matrix.machine_type }}_examples_gpu examples/pytorch
 
       - name: Failure short reports
         if: ${{ failure() }}
         continue-on-error: true
-        run: cat /transformers/reports/single-gpu_examples_gpu/failures_short.txt
+        run: cat /transformers/reports/${{ matrix.machine_type }}_examples_gpu/failures_short.txt
 
       - name: Test suite reports artifacts
         if: ${{ always() }}
         uses: actions/upload-artifact@v2
         with:
-          name: single-gpu_run_examples_gpu
-          path: /transformers/reports/single-gpu_examples_gpu
+          name: ${{ matrix.machine_type }}_run_examples_gpu
+          path: /transformers/reports/${{ matrix.machine_type }}_examples_gpu
 
   run_pipelines_torch_gpu:
     name: PyTorch pipelines
@@ -226,7 +259,7 @@ jobs:
     container:
       image: huggingface/transformers-pytorch-gpu
       options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-    needs: [run_check_runners, setup]
+    needs: setup
     steps:
       - name: Update clone
         working-directory: /transformers
@@ -241,12 +274,14 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - 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=${{ matrix.machine_type }}_tests_torch_pipeline_gpu tests
+          python3 -m pytest -n 1 -v --dist=loadfile --make-reports=${{ matrix.machine_type }}_tests_torch_pipeline_gpu tests/pipelines
 
       - name: Failure short reports
         if: ${{ failure() }}
@@ -270,7 +305,7 @@ jobs:
     container:
       image: huggingface/transformers-tensorflow-gpu
       options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
-    needs: [run_check_runners, setup]
+    needs: setup
     steps:
       - name: Update clone
         working-directory: /transformers
@@ -286,12 +321,14 @@ jobs:
         run: |
           python3 utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /transformers
+        run: pip freeze
+
       - 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=${{ matrix.machine_type }}_tests_tf_pipeline_gpu tests
+          python3 -m pytest -n 1 -v --dist=loadfile --make-reports=${{ matrix.machine_type }}_tests_tf_pipeline_gpu tests/pipelines
 
       - name: Failure short reports
         if: ${{ always() }}
@@ -312,7 +349,7 @@ jobs:
       matrix:
         machine_type: [single-gpu, multi-gpu]
     runs-on: ${{ format('{0}-{1}', matrix.machine_type, 'docker') }}
-    needs: [run_check_runners, setup]
+    needs: setup
     container:
       image: huggingface/transformers-pytorch-deepspeed-latest-gpu
       options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
@@ -340,6 +377,10 @@ jobs:
         run: |
           python utils/print_env.py
 
+      - name: Show installed libraries and their versions
+        working-directory: /workspace/transformers
+        run: pip freeze
+
       - name: Run all tests on GPU
         working-directory: /workspace/transformers
         run: |
@@ -357,12 +398,13 @@ jobs:
           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
+  run_extract_warnings:
+    name: Extract warnings in CI artifacts
     runs-on: ubuntu-latest
     if: always()
     needs: [
-      run_check_runners,
+      check_runner_status,
+      check_runners,
       setup,
       run_tests_single_gpu,
       run_tests_multi_gpu,
@@ -371,12 +413,64 @@ jobs:
       run_pipelines_torch_gpu,
       run_all_tests_torch_cuda_extensions_gpu
     ]
+    steps:
+      - name: Checkout transformers
+        uses: actions/checkout@v2
+        with:
+          fetch-depth: 2
+
+      - name: Install transformers
+        run: pip install transformers
+
+      - name: Show installed libraries and their versions
+        run: pip freeze
+
+      - name: Create output directory
+        run: mkdir warnings_in_ci
+
+      - uses: actions/download-artifact@v2
+        with:
+          path: warnings_in_ci
+
+      - name: Show artifacts
+        run: echo "$(python3 -c 'import os; d = os.listdir(); print(d)')"
+        working-directory: warnings_in_ci
+
+      - name: Extract warnings in CI artifacts
+        run: |
+          python3 utils/extract_warnings.py --workflow_run_id ${{ github.run_id }} --output_dir warnings_in_ci --token ${{ secrets.ACCESS_REPO_INFO_TOKEN }} --from_gh
+          echo "$(python3 -c 'import os; import json; fp = open("warnings_in_ci/selected_warnings.json"); d = json.load(fp); d = "\n".join(d) ;print(d)')"
+
+      - name: Upload artifact
+        if: ${{ always() }}
+        uses: actions/upload-artifact@v2
+        with:
+          name: warnings_in_ci
+          path: warnings_in_ci/selected_warnings.json
+
+  send_results:
+    name: Send results to webhook
+    runs-on: ubuntu-latest
+    if: always()
+    needs: [
+      check_runner_status,
+      check_runners,
+      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,
+      run_extract_warnings
+    ]
     steps:
       - name: Preliminary job status
         shell: bash
         # For the meaning of these environment variables, see the job `Setup`
         run: |
-          echo "Runner status: ${{ needs.run_check_runners.result }}"
+          echo "Runner availability: ${{ needs.check_runner_status.result }}"
+          echo "Runner status: ${{ needs.check_runners.result }}"
           echo "Setup status: ${{ needs.setup.result }}"
 
       - uses: actions/checkout@v2
@@ -389,10 +483,12 @@ jobs:
           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
+          RUNNER_STATUS: ${{ needs.check_runner_status.result }}
+          RUNNER_ENV_STATUS: ${{ needs.check_runners.result }}
           SETUP_STATUS: ${{ needs.setup.result }}
-          RUNNER_STATUS: ${{ needs.run_check_runners.result }}
         # We pass `needs.setup.outputs.matrix` as the argument. A processing in `notification_service.py` to change
         # `models/bert` to `models_bert` is required, as the artifact names use `_` instead of `/`.
         run: |
           pip install slack_sdk
+          pip show slack_sdk
           python utils/notification_service.py "${{ needs.setup.outputs.matrix }}"

.github/workflows/stale.yml

@@ -12,10 +12,10 @@ jobs:
     env:
       GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
     steps:
-    - uses: actions/checkout@v2
+    - uses: actions/checkout@v3
 
     - name: Setup Python
-      uses: actions/setup-python@v1
+      uses: actions/setup-python@v4
       with:
         python-version: 3.7
 
@@ -24,4 +24,4 @@ jobs:
         pip install PyGithub
     - name: Close stale issues
       run: |
-        python scripts/stale.py
+        python scripts/stale.py

CODE_OF_CONDUCT.md

@@ -7,8 +7,8 @@ We as members, contributors, and leaders pledge to make participation in our
 community a harassment-free experience for everyone, regardless of age, body
 size, visible or invisible disability, ethnicity, sex characteristics, gender
 identity and expression, level of experience, education, socio-economic status,
-nationality, personal appearance, race, religion, or sexual identity
-and orientation.
+nationality, personal appearance, race, caste, color, religion, or sexual
+identity and orientation.
 
 We pledge to act and interact in ways that contribute to an open, welcoming,
 diverse, inclusive, and healthy community.
@@ -23,17 +23,17 @@ community include:
 * Giving and gracefully accepting constructive feedback
 * Accepting responsibility and apologizing to those affected by our mistakes,
   and learning from the experience
-* Focusing on what is best not just for us as individuals, but for the
-  overall community
+* Focusing on what is best not just for us as individuals, but for the overall
+  community
 
 Examples of unacceptable behavior include:
 
-* The use of sexualized language or imagery, and sexual attention or
-  advances of any kind
+* The use of sexualized language or imagery, and sexual attention or advances of
+  any kind
 * Trolling, insulting or derogatory comments, and personal or political attacks
 * Public or private harassment
-* Publishing others' private information, such as a physical or email
-  address, without their explicit permission
+* Publishing others' private information, such as a physical or email address,
+  without their explicit permission
 * Other conduct which could reasonably be considered inappropriate in a
   professional setting
 
@@ -83,15 +83,15 @@ behavior was inappropriate. A public apology may be requested.
 
 ### 2. Warning
 
-**Community Impact**: A violation through a single incident or series
-of actions.
+**Community Impact**: A violation through a single incident or series of
+actions.
 
 **Consequence**: A warning with consequences for continued behavior. No
 interaction with the people involved, including unsolicited interaction with
 those enforcing the Code of Conduct, for a specified period of time. This
 includes avoiding interactions in community spaces as well as external channels
-like social media. Violating these terms may lead to a temporary or
-permanent ban.
+like social media. Violating these terms may lead to a temporary or permanent
+ban.
 
 ### 3. Temporary Ban
 
@@ -107,23 +107,27 @@ Violating these terms may lead to a permanent ban.
 ### 4. Permanent Ban
 
 **Community Impact**: Demonstrating a pattern of violation of community
-standards, including sustained inappropriate behavior,  harassment of an
+standards, including sustained inappropriate behavior, harassment of an
 individual, or aggression toward or disparagement of classes of individuals.
 
-**Consequence**: A permanent ban from any sort of public interaction within
-the community.
+**Consequence**: A permanent ban from any sort of public interaction within the
+community.
 
 ## Attribution
 
 This Code of Conduct is adapted from the [Contributor Covenant][homepage],
-version 2.0, available at
-https://www.contributor-covenant.org/version/2/0/code_of_conduct.html.
+version 2.1, available at
+[https://www.contributor-covenant.org/version/2/1/code_of_conduct.html][v2.1].
 
-Community Impact Guidelines were inspired by [Mozilla's code of conduct
-enforcement ladder](https://github.com/mozilla/diversity).
-
-[homepage]: https://www.contributor-covenant.org
+Community Impact Guidelines were inspired by
+[Mozilla's code of conduct enforcement ladder][Mozilla CoC].
 
 For answers to common questions about this code of conduct, see the FAQ at
-https://www.contributor-covenant.org/faq. Translations are available at
-https://www.contributor-covenant.org/translations.
+[https://www.contributor-covenant.org/faq][FAQ]. Translations are available at
+[https://www.contributor-covenant.org/translations][translations].
+
+[homepage]: https://www.contributor-covenant.org
+[v2.1]: https://www.contributor-covenant.org/version/2/1/code_of_conduct.html
+[Mozilla CoC]: https://github.com/mozilla/diversity
+[FAQ]: https://www.contributor-covenant.org/faq
+[translations]: https://www.contributor-covenant.org/translations

CONTRIBUTING.md

@@ -14,124 +14,126 @@ See the License for the specific language governing permissions and
 limitations under the License.
 -->
 
-# How to contribute to transformers?
+# Contribute to 🤗 Transformers
 
 Everyone is welcome to contribute, and we value everybody's contribution. Code
-is thus not the only way to help the community. Answering questions, helping
-others, reaching out and improving the documentations are immensely valuable to
-the community.
+contributions are not the only way to help the community. Answering questions, helping
+others, and improving the documentation are also immensely valuable.
 
-It also helps us if you spread the word: reference the library from blog posts
-on the awesome projects it made possible, shout out on Twitter every time it has
-helped you, or simply star the repo to say "thank you".
+It also helps us if you spread the word! Reference the library in blog posts
+about the awesome projects it made possible, shout out on Twitter every time it has
+helped you, or simply ⭐️ the repository to say thank you.
 
-Whichever way you choose to contribute, please be mindful to respect our
+However you choose to contribute, please be mindful and respect our
 [code of conduct](https://github.com/huggingface/transformers/blob/main/CODE_OF_CONDUCT.md).
 
-## You can contribute in so many ways!
+**This guide was heavily inspired by the awesome [scikit-learn guide to contributing](https://github.com/scikit-learn/scikit-learn/blob/main/CONTRIBUTING.md).**
 
-There are 4 ways you can contribute to transformers:
-* Fixing outstanding issues with the existing code;
-* Implementing new models;
-* Contributing to the examples or to the documentation;
-* Submitting issues related to bugs or desired new features.
+## Ways to contribute
 
-In particular there is a special [Good First
+There are several ways you can contribute to 🤗 Transformers:
+
+* Fix outstanding issues with the existing code.
+* Submit issues related to bugs or desired new features.
+* Implement new models.
+* Contribute to the examples or to the documentation.
+
+If you don't know where to start, there is a special [Good First
 Issue](https://github.com/huggingface/transformers/contribute) listing. It will give you a list of
-open Issues that are open to anybody to work on. Just comment in the issue that you'd like to work
-on it. In that same listing you will also find some Issues with `Good Second Issue` label. These are
-typically slightly more complicated than the Issues with just `Good First Issue` label. But if you
-feel you know what you're doing, go for it.
+open issues that are beginner-friendly and help you start contributing to open-source. Just comment in the issue that you'd like to work
+on it. 
 
-*All are equally valuable to the community.*
+For something slightly more challenging, you can also take a look at the [Good Second Issue](https://github.com/huggingface/transformers/labels/Good%20Second%20Issue) list. In general though, if you feel like you know what you're doing, go for it and we'll help you get there! 🚀
 
-## Submitting a new issue or feature request
+> All contributions are equally valuable to the community. 🥰
 
-Do your best to follow these guidelines when submitting an issue or a feature
+## Fixing outstanding issues
+
+If you notice an issue with the existing code and have a fix in mind, feel free to [start contributing](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md/#create-a-pull-request) and open a Pull Request!
+
+## Submitting a bug-related issue or feature request
+
+Do your best to follow these guidelines when submitting a bug-related issue or a feature
 request. It will make it easier for us to come back to you quickly and with good
 feedback.
 
 ### Did you find a bug?
 
-The 🤗 Transformers library is robust and reliable thanks to the users who notify us of
-the problems they encounter. So thank you for reporting an issue.
+The 🤗 Transformers library is robust and reliable thanks to users who report the problems they encounter.
 
-First, we would really appreciate it if you could **make sure the bug was not
-already reported** (use the search bar on Github under Issues).
+Before you report an issue, we would really appreciate it if you could **make sure the bug was not
+already reported** (use the search bar on GitHub under Issues). Your issue should also be related to bugs in the library itself, and not your code. If you're unsure whether the bug is in your code or the library, please ask on the [forum](https://discuss.huggingface.co/) first. This helps us respond quicker to fixing issues related to the library versus general questions.
 
-Did not find it? :( So we can act quickly on it, please follow these steps:
+Once you've confirmed the bug hasn't already been reported, please include the following information in your issue so we can quickly resolve it:
 
-* Include your **OS type and version**, the versions of **Python**, **PyTorch** and
-  **Tensorflow** when applicable;
+* Your **OS type and version** and **Python**, **PyTorch** and
+  **TensorFlow** versions when applicable.
 * A short, self-contained, code snippet that allows us to reproduce the bug in
-  less than 30s;
-* Provide the *full* traceback if an exception is raised.
+  less than 30s.
+* The *full* traceback if an exception is raised.
+* Attach any other additional information, like screenshots, you think may help.
 
-To get the OS and software versions automatically, you can run the following command:
+To get the OS and software versions automatically, run the following command:
 
 ```bash
 transformers-cli env
 ```
 
-or from the root of the repository the following command:
+You can also run the same command from the root of the repository:
 
 ```bash
 python src/transformers/commands/transformers_cli.py env
 ```
 
+### Do you want a new feature?
 
-### Do you want to implement a new model?
+If there is a new feature you'd like to see in 🤗 Transformers, please open an issue and describe:
 
-Awesome! Please provide the following information:
+1. What is the *motivation* behind this feature? Is it related to a problem or frustration with the library? Is it a feature related to something you need for a project? Is it something you worked on and think it could benefit the community?
 
-* Short description of the model and link to the paper;
-* Link to the implementation if it is open-source;
+   Whatever it is, we'd love to hear about it!
+
+2. Describe your requested feature in as much detail as possible. The more you can tell us about it, the better we'll be able to help you.
+3. Provide a *code snippet* that demonstrates the features usage.
+4. If the feature is related to a paper, please include a link.
+
+If your issue is well written we're already 80% of the way there by the time you create it.
+
+We have added [templates](https://github.com/huggingface/transformers/tree/main/templates) to help you get started with your issue.
+
+## Do you want to implement a new model?
+
+New models are constantly released and if you want to implement a new model, please provide the following information
+
+* A short description of the model and link to the paper.
+* Link to the implementation if it is open-sourced.
 * Link to the model weights if they are available.
 
-If you are willing to contribute the model yourself, let us know so we can best
-guide you.
+If you are willing to contribute the model yourself, let us know so we can help you add it to 🤗 Transformers!
 
-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`](https://github.com/huggingface/transformers/tree/main/templates) folder.
+We have added a [detailed guide and templates](https://github.com/huggingface/transformers/tree/main/templates) to help you get started with adding a new model, and we also have a more technical guide for [how to add a model to 🤗 Transformers](https://huggingface.co/docs/transformers/add_new_model).
 
-### Do you want a new feature (that is not a model)?
+## Do you want to add documentation?
 
-A world-class feature request addresses the following points:
+We're always looking for improvements to the documentation that make it more clear and accurate. Please let us know how the documentation can be improved such as typos and any content that is missing, unclear or inaccurate. We'll be happy to make the changes or help you make a contribution if you're interested!
 
-1. Motivation first:
-  * Is it related to a problem/frustration with the library? If so, please explain
-    why. Providing a code snippet that demonstrates the problem is best.
-  * Is it related to something you would need for a project? We'd love to hear
-    about it!
-  * Is it something you worked on and think could benefit the community?
-    Awesome! Tell us what problem it solved for you.
-2. Write a *full paragraph* describing the feature;
-3. Provide a **code snippet** that demonstrates its future use;
-4. In case this is related to a paper, please attach a link;
-5. Attach any additional information (drawings, screenshots, etc.) you think may help.
+For more details about how to generate, build, and write the documentation, take a look at the documentation [README](https://github.com/huggingface/transformers/tree/main/docs).
 
-If your issue is well written we're already 80% of the way there by the time you
-post it.
+## Create a Pull Request
 
-We have added **templates** to guide you in the process of adding a new example script for training or testing the
-models in the library. You can find them in the [`templates`](https://github.com/huggingface/transformers/tree/main/templates)
-folder.
-
-## Start contributing! (Pull Requests)
-
-Before writing code, we strongly advise you to search through the existing PRs or
-issues to make sure that nobody is already working on the same thing. If you are
+Before writing any code, we strongly advise you to search through the existing PRs or
+issues to make sure nobody is already working on the same thing. If you are
 unsure, it is always a good idea to open an issue to get some feedback.
 
-You will need basic `git` proficiency to be able to contribute to
-🤗 Transformers. `git` is not the easiest tool to use but it has the greatest
-manual. Type `git --help` in a shell and enjoy. If you prefer books, [Pro
+You will need basic `git` proficiency to contribute to
+🤗 Transformers. While `git` is not the easiest tool to use, it has the greatest
+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 ([supported Python versions](https://github.com/huggingface/transformers/blob/main/setup.py#L426)):
+You'll need **[Python 3.7]((https://github.com/huggingface/transformers/blob/main/setup.py#L426))** or above to contribute to 🤗 Transformers. Follow the steps below to start contributing:
 
 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
+   clicking on the **[Fork](https://github.com/huggingface/transformers/fork)** button on the repository's page. This creates a copy of the code
    under your GitHub user account.
 
 2. Clone your fork to your local disk, and add the base repository as a remote:
@@ -148,7 +150,7 @@ Follow these steps to start contributing ([supported Python versions](https://gi
    $ git checkout -b a-descriptive-name-for-my-changes
    ```
 
-   **Do not** work on the `main` branch.
+   🚨 **Do not** work on the `main` branch!
 
 4. Set up a development environment by running the following command in a virtual environment:
 
@@ -156,25 +158,13 @@ Follow these steps to start contributing ([supported Python versions](https://gi
    $ pip install -e ".[dev]"
    ```
 
-   (If transformers was already installed in the virtual environment, remove
+   If 🤗 Transformers was already installed in the virtual environment, remove
    it with `pip uninstall transformers` before reinstalling it in editable
-   mode with the `-e` flag.)
-
-   To run the full test suite, you might need the additional dependency on `datasets` which requires a separate source
-   install:
-
-   ```bash
-   $ git clone https://github.com/huggingface/datasets
-   $ cd datasets
-   $ pip install -e .
-   ```
-
-   If you have already cloned that repo, you might need to `git pull` to get the most recent changes in the `datasets`
-   library.
+   mode with the `-e` flag.
    
-   Depending on your OS, you might need to install some external libraries, as well, if the `pip` installation fails.
+   Depending on your OS, you may need to install some external libraries as well if the `pip` installation fails.
    
-   For macOS, you will likely need [MeCab](https://taku910.github.io/mecab/), which can be installed from Homebrew:
+   For macOS, you will likely need [MeCab](https://taku910.github.io/mecab/) which can be installed from Homebrew:
    
    ```bash
    brew install mecab
@@ -182,23 +172,15 @@ Follow these steps to start contributing ([supported Python versions](https://gi
 
 5. Develop the features on your branch.
 
-   As you work on the features, you should make sure that the test suite
-   passes. You should run the tests impacted by your changes like this:
+   As you work on your code, you should make sure the test suite
+   passes. Run the tests impacted by your changes like this:
 
    ```bash
    $ pytest tests/<TEST_TO_RUN>.py
    ```
 
-   You can also run the full suite with the following command, but it takes
-   a beefy machine to produce a result in a decent amount of time now that
-   Transformers has grown a lot. Here is the command for it:
-
-   ```bash
-   $ make test
-   ```
-
    For more information about tests, check out the
-   [dedicated documentation](https://huggingface.co/docs/transformers/testing)
+   [Testing](https://huggingface.co/docs/transformers/testing) guide.
 
    🤗 Transformers relies on `black` and `isort` to format its source code
    consistently. After you make changes, apply automatic style corrections and code verifications
@@ -210,7 +192,7 @@ Follow these steps to start contributing ([supported Python versions](https://gi
 
    This target is also optimized to only work with files modified by the PR you're working on.
 
-   If you prefer to run the checks one after the other, the following command apply the
+   If you prefer to run the checks one after the other, the following command applies the
    style corrections:
 
    ```bash
@@ -218,145 +200,144 @@ Follow these steps to start contributing ([supported Python versions](https://gi
    ```
 
    🤗 Transformers also uses `flake8` and a few custom scripts to check for coding mistakes. Quality
-   control runs in CI, however you can also run the same checks with:
+   controls are run by the CI, but you can run the same checks with:
 
    ```bash
    $ make quality
    ```
 
-   Finally we have a lot of scripts that check we didn't forget to update
-   some files when adding a new model, that you can run with
+   Finally, we have a lot of scripts to make sure we didn't forget to update
+   some files when adding a new model. You can run these scripts with:
 
    ```bash
    $ make repo-consistency
    ```
 
-   To learn more about those checks and how to fix any issue with them, check out the
-   [documentation](https://huggingface.co/docs/transformers/pr_checks)
+   To learn more about those checks and how to fix any issues with them, check out the
+   [Checks on a Pull Request](https://huggingface.co/docs/transformers/pr_checks) guide.
 
-   If you're modifying documents under `docs/source`, make sure to validate that
-   they can still be built. This check also runs in CI. To run a local check
-   make sure you have installed the documentation builder requirements. First you will need to clone the
-   repository containing our tools to build the documentation:
-   
-   ```bash
-   $ pip install git+https://github.com/huggingface/doc-builder
-   ```
-
-   Then, make sure you have all the dependencies to be able to build the doc with:
+   If you're modifying documents under `docs/source` directory, make sure the documentation can still be built. This check will also run in the CI when you open a pull request. To run a local check
+   make sure you install the documentation builder:
    
    ```bash
    $ pip install ".[docs]"
    ```
 
-   Finally run the following command from the root of the repository:
+   Run the following command from the root of the repository:
 
    ```bash
-   $ doc-builder build transformers docs/source/ --build_dir ~/tmp/test-build
+   $ doc-builder build transformers docs/source/en --build_dir ~/tmp/test-build
    ```
 
    This will build the documentation in the `~/tmp/test-build` folder where you can inspect the generated
-   Markdown files with your favorite editor. You won't be able to see the final rendering on the website
-   before your PR is merged, we are actively working on adding a tool for this.
+   Markdown files with your favorite editor. You can also preview the docs on GitHub when you open a pull request.
 
-   Once you're happy with your changes, add changed files using `git add` and
-   make a commit with `git commit` to record your changes locally:
+   Once you're happy with your changes, add changed files with `git add` and
+   record your changes locally with `git commit`:
 
    ```bash
    $ git add modified_file.py
    $ git commit
    ```
 
-   Please write [good commit
-   messages](https://chris.beams.io/posts/git-commit/).
+   Please remember to write [good commit
+   messages](https://chris.beams.io/posts/git-commit/) to clearly communicate the changes you made!
 
-   It is a good idea to sync your copy of the code with the original
-   repository regularly. This way you can quickly account for changes:
+   To keep your copy of the code up to date with the original
+   repository, rebase your branch on `upstream/branch` *before* you open a pull request or if requested by a maintainer:
 
    ```bash
    $ git fetch upstream
    $ git rebase upstream/main
    ```
 
-   Push the changes to your account using:
+   Push your changes to your branch:
 
    ```bash
    $ git push -u origin a-descriptive-name-for-my-changes
    ```
 
-6. Once you are satisfied (**and the checklist below is happy too**), go to the
-   webpage of your fork on GitHub. Click on 'Pull request' to send your changes
-   to the project maintainers for review.
+   If you've already opened a pull request, you'll need to force push with the `--force` flag. Otherwise, if the pull request hasn't been opened yet, you can just push your changes normally.
 
-7. It's ok if maintainers ask you for changes. It happens to core contributors
-   too! So everyone can see the changes in the Pull request, work in your local
+6. Now you can go to your fork of the repository on GitHub and click on **Pull request** to open a pull request. Make sure you tick off all the boxes in our [checklist](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md/#pull-request-checklist) below. When you're ready, you can send your changes to the project maintainers for review.
+
+7. It's ok if maintainers request changes, it happens to our core contributors
+   too! So everyone can see the changes in the pull request, work in your local
    branch and push the changes to your fork. They will automatically appear in
    the pull request.
 
+### Pull request checklist
 
-### Checklist
-
-1. The title of your pull request should be a summary of its contribution;
-2. If your pull request addresses an issue, please mention the issue number in
-   the pull request description to make sure they are linked (and people
-   consulting the issue know you are working on it);
-3. To indicate a work in progress please prefix the title with `[WIP]`. These
-   are useful to avoid duplicated work, and to differentiate it from PRs ready
-   to be merged;
-4. Make sure existing tests pass;
-5. Add high-coverage tests. No quality testing = no merge.
-   - If you are adding a new model, make sure that you use
-     `ModelTester.all_model_classes = (MyModel, MyModelWithLMHead,...)`, which triggers the common tests.
+☐ The pull request title should summarize your contribution.<br>
+☐ If your pull request addresses an issue, please mention the issue number in the pull
+request description to make sure they are linked (and people viewing the issue know you
+are working on it).<br>
+☐ To indicate a work in progress please prefix the title with `[WIP]`. These are
+useful to avoid duplicated work, and to differentiate it from PRs ready to be merged.
+☐ Make sure existing tests pass.<br>
+☐ If adding a new feature, also add tests for it.<br>
+   - If you are adding a new model, make sure you use
+     `ModelTester.all_model_classes = (MyModel, MyModelWithLMHead,...)` to trigger the common tests.
    - If you are adding new `@slow` tests, make sure they pass using
-     `RUN_SLOW=1 python -m pytest tests/test_my_new_model.py`.
-   - If you are adding a new tokenizer, write tests, and make sure
-     `RUN_SLOW=1 python -m pytest tests/test_tokenization_{your_model_name}.py` passes.
-   CircleCI does not run the slow tests, but github actions does every night!
-6. All public methods must have informative docstrings that work nicely with sphinx. See `modeling_bert.py` for an
-   example.
-7. Due to the rapidly growing repository, it is important to make sure that no files that would significantly weigh down the repository are added. This includes images, videos and other non-text files. We prefer to leverage a hf.co hosted `dataset` like
-   the ones hosted on [`hf-internal-testing`](https://huggingface.co/hf-internal-testing) in which to place these files and reference 
-   them by URL. We recommend putting them in the following dataset: [huggingface/documentation-images](https://huggingface.co/datasets/huggingface/documentation-images).
-   If an external contribution, feel free to add the images to your PR and ask a Hugging Face member to migrate your images
-   to this dataset.
+     `RUN_SLOW=1 python -m pytest tests/models/my_new_model/test_my_new_model.py`.
+   - If you are adding a new tokenizer, write tests and make sure
+     `RUN_SLOW=1 python -m pytest tests/models/{your_model_name}/test_tokenization_{your_model_name}.py` passes.
+   CircleCI does not run the slow tests, but GitHub Actions does every night!<br>
 
-See more about the checks run on a pull request in our [PR guide](pr_checks)
+☐ All public methods must have informative docstrings (see
+[`modeling_bert.py`](https://github.com/huggingface/transformers/blob/main/src/transformers/models/bert/modeling_bert.py)
+for an example).<br>
+☐ Due to the rapidly growing repository, don't add any images, videos and other
+non-text files that'll significantly weigh down the repository. Instead, use a Hub
+repository such as [`hf-internal-testing`](https://huggingface.co/hf-internal-testing)
+to host these files and reference them by URL. We recommend placing documentation
+related images in the following repository:
+[huggingface/documentation-images](https://huggingface.co/datasets/huggingface/documentation-images).
+You can open a PR on this dataset repostitory and ask a Hugging Face member to merge it.
+
+For more information about the checks run on a pull request, take a look at our [Checks on a Pull Request](https://huggingface.co/docs/transformers/pr_checks) guide.
 
 ### Tests
 
 An extensive test suite is included to test the library behavior and several examples. Library tests can be found in
-the [tests folder](https://github.com/huggingface/transformers/tree/main/tests) and examples tests in the
-[examples folder](https://github.com/huggingface/transformers/tree/main/examples).
+the [tests](https://github.com/huggingface/transformers/tree/main/tests) folder and examples tests in the
+[examples](https://github.com/huggingface/transformers/tree/main/examples) folder.
 
 We like `pytest` and `pytest-xdist` because it's faster. From the root of the
-repository, here's how to run tests with `pytest` for the library:
+repository, specify a *path to a subfolder or a test file* to run the test.
 
 ```bash
-$ python -m pytest -n auto --dist=loadfile -s -v ./tests/
+$ python -m pytest -n auto --dist=loadfile -s -v ./tests/models/my_new_model
 ```
 
-and for the examples:
+Similarly, for the `examples` directory, specify a *path to a subfolder or test file* to run the test. For example, the following command tests the text classification subfolder in the PyTorch `examples` directory:
 
 ```bash
 $ pip install -r examples/xxx/requirements.txt  # only needed the first time
-$ python -m pytest -n auto --dist=loadfile -s -v ./examples/
+$ python -m pytest -n auto --dist=loadfile -s -v ./examples/pytorch/text-classification
 ```
-In fact, that's how `make test` and `make test-examples` are implemented (sans the `pip install` line)!
 
-You can specify a smaller set of tests in order to test only the feature
+In fact, this is actually how our `make test` and `make test-examples` commands are implemented (not including the `pip install`)!
+
+You can also specify a smaller set of tests in order to test only the feature
 you're working on.
 
-By default, slow tests are skipped. Set the `RUN_SLOW` environment variable to
-`yes` to run them. This will download many gigabytes of models — make sure you
-have enough disk space and a good Internet connection, or a lot of patience!
+By default, slow tests are skipped but you can set the `RUN_SLOW` environment variable to
+`yes` to run them. This will download many gigabytes of models so make sure you
+have enough disk space, a good internet connection or a lot of patience!
+
+<Tip warning={true}>
+
+Remember to specify a *path to a subfolder or a test file* to run the test. Otherwise, you'll run all the tests in the `tests` or `examples` folder, which will take a very long time!
+
+</Tip>
 
 ```bash
-$ RUN_SLOW=yes python -m pytest -n auto --dist=loadfile -s -v ./tests/
-$ RUN_SLOW=yes python -m pytest -n auto --dist=loadfile -s -v ./examples/
+$ RUN_SLOW=yes python -m pytest -n auto --dist=loadfile -s -v ./tests/models/my_new_model
+$ RUN_SLOW=yes python -m pytest -n auto --dist=loadfile -s -v ./examples/pytorch/text-classification
 ```
 
-Likewise, set the `RUN_CUSTOM_TOKENIZERS` environment variable to `yes` to run
-tests for custom tokenizers, which don't run by default either.
+Like the slow tests, custom tokenizer tests are skipped but you can set the `RUN_CUSTOM_TOKENIZERS` environment variable to `yes` to run them.
 
 🤗 Transformers uses `pytest` as a test runner only. It doesn't use any
 `pytest`-specific features in the test suite itself.
@@ -369,37 +350,37 @@ $ python -m unittest discover -s tests -t . -v
 $ python -m unittest discover -s examples -t examples -v
 ```
 
-
 ### Style guide
 
-For documentation strings, 🤗 Transformers follows the [google style](https://google.github.io/styleguide/pyguide.html).
+For documentation strings, 🤗 Transformers follows the [Google Python Style Guide](https://google.github.io/styleguide/pyguide.html).
 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).**
-
 ### Develop on Windows
 
-On windows, you need to configure git to transform Windows `CRLF` line endings to Linux `LF` line endings:
+On Windows (unless you're working in [Windows Subsytem for Linux](https://learn.microsoft.com/en-us/windows/wsl/) or WSL), you need to configure git to transform Windows `CRLF` line endings to Linux `LF` line endings:
 
-`git config core.autocrlf input`
+```bash
+git config core.autocrlf input
+```
 
-One way one can run the make command on Window is to pass by MSYS2:
+One way to run the `make` command on Windows is with MSYS2:
 
-1. [Download MSYS2](https://www.msys2.org/), we assume to have it installed in C:\msys64
-2. Open the command line C:\msys64\msys2.exe (it should be available from the start menu)
-3. Run in the shell: `pacman -Syu` and install make with `pacman -S make`
+1. [Download MSYS2](https://www.msys2.org/), and we assume it's installed in `C:\msys64`.
+2. Open the command line `C:\msys64\msys2.exe` (it should be available from the **Start** menu).
+3. Run in the shell: `pacman -Syu` and install `make` with `pacman -S make`.
 4. Add `C:\msys64\usr\bin` to your PATH environment variable.
 
-You can now use `make` from any terminal (Powershell, cmd.exe, etc) 🎉
+You can now use `make` from any terminal (Powershell, cmd.exe, etc.)! 🎉
 
-### Syncing forked main with upstream (HuggingFace) main
+### Sync a forked repository with upstream main (the Hugging Face repository)
 
-To avoid pinging the upstream repository which adds reference notes to each upstream PR and sends unnecessary notifications to the developers involved in these PRs,
-when syncing the main branch of a forked repository, please, follow these steps:
-1. When possible, avoid syncing with the upstream using a branch and PR on the forked repository. Instead merge directly into the forked main.
+When updating the main branch of a forked repository, please follow these steps to avoid pinging the upstream repository which adds reference notes to each upstream PR, and sends unnecessary notifications to the developers involved in these PRs.
+
+1. When possible, avoid syncing with the upstream using a branch and PR on the forked repository. Instead, merge directly into the forked main.
 2. If a PR is absolutely necessary, use the following steps after checking out your branch:
-```
+
+```bash
 $ git checkout -b your-branch-for-syncing
 $ git pull --squash --no-commit upstream main
 $ git commit -m '<your message without GitHub references>'

ISSUES.md

@@ -18,7 +18,7 @@ limitations under the License.
 
 This is an Open Source Project so please be mindful that like in any other project of this kind there is no obligation to answer all requests for help.
 
-However, we want to encourage you to ask for help whenever you think it's needed! We are happy about every  question we get because it allows us to better understand your needs, possible misunderstandings, and most importantly a way for you to help us make this library better. That being said, this document's main purpose is to provide guidelines at how you can formulate your requests to increase your chances to be understood and to get support.
+However, we want to encourage you to ask for help whenever you think it's needed! We are happy about every question we get because it allows us to better understand your needs, possible misunderstandings, and most importantly a way for you to help us make this library better. That being said, this document's main purpose is to provide guidelines at how you can formulate your requests to increase your chances to be understood and to get support.
 
 There are two main venues to receive support: [the forums](https://discuss.huggingface.co/) and [the GitHub issues](https://github.com/huggingface/transformers/issues).
 

Makefile

@@ -41,6 +41,7 @@ repo-consistency:
 	python utils/check_inits.py
 	python utils/check_config_docstrings.py
 	python utils/tests_fetcher.py --sanity_check
+	python utils/update_metadata.py --check-only
 
 # this target runs checks on all files
 

README.md

@@ -43,7 +43,9 @@ limitations under the License.
         <b>English</b> |
         <a href="https://github.com/huggingface/transformers/blob/main/README_zh-hans.md">简体中文</a> |
         <a href="https://github.com/huggingface/transformers/blob/main/README_zh-hant.md">繁體中文</a> |
-        <a href="https://github.com/huggingface/transformers/blob/main/README_ko.md">한국어</a>
+        <a href="https://github.com/huggingface/transformers/blob/main/README_ko.md">한국어</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_es.md">Español</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_ja.md">日本語</a>
     <p>
 </h4>
 
@@ -55,13 +57,13 @@ limitations under the License.
     <a href="https://hf.co/course"><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/course_banner.png"></a>
 </h3>
 
-🤗 Transformers provides thousands of pretrained models to perform tasks on different modalities such as text, vision, and audio. 
+🤗 Transformers provides thousands of pretrained models to perform tasks on different modalities such as text, vision, and audio.
 
 These models can be applied on:
 
-* 📝 Text, for tasks like text classification, information extraction, question answering, summarization, translation, text generation, in over 100 languages. 
-* 🖼️ Images, for tasks like image classification, object detection, and segmentation. 
-* 🗣️ Audio, for tasks like speech recognition and audio classification. 
+* 📝 Text, for tasks like text classification, information extraction, question answering, summarization, translation, text generation, in over 100 languages.
+* 🖼️ Images, for tasks like image classification, object detection, and segmentation.
+* 🗣️ Audio, for tasks like speech recognition and audio classification.
 
 Transformer models can also perform tasks on **several modalities combined**, such as table question answering, optical character recognition, information extraction from scanned documents, video classification, and visual question answering.
 
@@ -133,7 +135,7 @@ Many tasks have a pre-trained `pipeline` ready to go, in NLP but also in compute
 >>> image = Image.open(image_data)
 
 # Allocate a pipeline for object detection
->>> object_detector = pipeline('object_detection')
+>>> object_detector = pipeline('object-detection')
 >>> object_detector(image)
 [{'score': 0.9982201457023621,
   'label': 'remote',
@@ -152,7 +154,7 @@ Many tasks have a pre-trained `pipeline` ready to go, in NLP but also in compute
   'box': {'xmin': 345, 'ymin': 23, 'xmax': 640, 'ymax': 368}}]
 ```
 
-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:
+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 left, with the predictions displayed on the right:
 
 <h3 align="center">
     <a><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/coco_sample.png" width="400"></a>
@@ -227,7 +229,7 @@ You should install 🤗 Transformers in a [virtual environment](https://docs.pyt
 First, create a virtual environment with the version of Python you're going to use and activate it.
 
 Then, you will need to install at least one of Flax, PyTorch or TensorFlow.
-Please refer to [TensorFlow installation page](https://www.tensorflow.org/install/), [PyTorch installation page](https://pytorch.org/get-started/locally/#start-locally) and/or [Flax](https://github.com/google/flax#quick-install) and [Jax](https://github.com/google/jax#installation) installation pages regarding the specific install command for your platform.
+Please refer to [TensorFlow installation page](https://www.tensorflow.org/install/), [PyTorch installation page](https://pytorch.org/get-started/locally/#start-locally) and/or [Flax](https://github.com/google/flax#quick-install) and [Jax](https://github.com/google/jax#installation) installation pages regarding the specific installation command for your platform.
 
 When one of those backends has been installed, 🤗 Transformers can be installed using pip as follows:
 
@@ -249,6 +251,8 @@ conda install -c huggingface transformers
 
 Follow the installation pages of Flax, PyTorch or TensorFlow to see how to install them with conda.
 
+> **_NOTE:_**  On Windows, you may be prompted to activate Developer Mode in order to benefit from caching. If this is not an option for you, please let us know in [this issue](https://github.com/huggingface/huggingface_hub/issues/1062).
+
 ## Model architectures
 
 **[All the model checkpoints](https://huggingface.co/models)** provided by 🤗 Transformers are seamlessly integrated from the huggingface.co [model hub](https://huggingface.co) where they are uploaded directly by [users](https://huggingface.co/users) and [organizations](https://huggingface.co/organizations).
@@ -258,6 +262,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 🤗 Transformers currently provides the following architectures (see [here](https://huggingface.co/docs/transformers/model_summary) for a high-level summary of each them):
 
 1. **[ALBERT](https://huggingface.co/docs/transformers/model_doc/albert)** (from Google Research and the Toyota Technological Institute at Chicago) released with the paper [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://arxiv.org/abs/1909.11942), by Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin Gimpel, Piyush Sharma, Radu Soricut.
+1. **[Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer)** (from MIT) released with the paper [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) by Yuan Gong, Yu-An Chung, James Glass.
 1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (from Facebook) released with the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) by Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer.
 1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (from École polytechnique) released with the paper [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) by Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis.
 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.
@@ -269,13 +274,16 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 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. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (from BigScience workshop) released by the [BigScience 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. **[Chinese-CLIP](https://huggingface.co/docs/transformers/model_doc/chinese_clip)** (from OFA-Sys) released with the paper [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) by An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou.
 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. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (from University of Göttingen) released with the paper [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) by Timo Lüddecke and Alexander Ecker.
 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. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 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.
@@ -285,17 +293,21 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 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. **[Deformable DETR](https://huggingface.co/docs/transformers/model_doc/deformable_detr)** (from SenseTime Research) released with the paper [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://arxiv.org/abs/2010.04159) by Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, Jifeng Dai.
 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. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (from SHI Labs) released with the paper [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001) by Ali Hassani and Humphrey Shi.
 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. **[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. **[Donut](https://huggingface.co/docs/transformers/main/model_doc/donut)** (from NAVER), released together with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
+1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (from NAVER), released together with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
 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. **[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. **[ERNIE](https://huggingface.co/docs/transformers/main/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (from Meta AI) are transformer protein language models.  **ESM-1b** was released with the paper [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus. **ESM-1v** was released with the paper [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648) by Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives. **ESM-2 and ESMFold** were released with the paper [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
+1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 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.
@@ -304,18 +316,21 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 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 NeoX Japanese](https://huggingface.co/docs/transformers/model_doc/gpt_neox_japanese)** (from ABEJA) released by Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori.
 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/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. **[Jukebox](https://huggingface.co/docs/transformers/model_doc/jukebox)** (from OpenAI) released with the paper [Jukebox: A Generative Model for Music](https://arxiv.org/pdf/2005.00341.pdf) by Prafulla Dhariwal, Heewoo Jun, Christine Payne, Jong Wook Kim, Alec Radford, 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/layoutxlm)** (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. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (from South China University of Technology) released with the paper [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) by Jiapeng Wang, Lianwen Jin, Kai Ding.
 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.
@@ -323,6 +338,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 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. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 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.
@@ -330,17 +346,20 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 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. **[MobileNetV1](https://huggingface.co/docs/transformers/model_doc/mobilenet_v1)** (from Google Inc.) released with the paper [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://arxiv.org/abs/1704.04861) by Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam.
+1. **[MobileNetV2](https://huggingface.co/docs/transformers/model_doc/mobilenet_v2)** (from Google Inc.) released with the paper [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381) by Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen.
 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. **[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. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 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. **[PEGASUS-X](https://huggingface.co/docs/transformers/main/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, and Peter J. Liu.
+1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, 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/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.
@@ -354,6 +373,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 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. **[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. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (from WeChatAI) released with the paper [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) by HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou.
 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.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -363,11 +383,14 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 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/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. **[Swin Transformer V2](https://huggingface.co/docs/transformers/main/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[SwitchTransformers](https://huggingface.co/docs/transformers/model_doc/switch_transformers)** (from Google) released with the paper [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://arxiv.org/abs/2101.03961) by William Fedus, Barret Zoph, Noam Shazeer.
 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. **[Table Transformer](https://huggingface.co/docs/transformers/model_doc/table-transformer)** (from Microsoft Research) released with the paper [PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents](https://arxiv.org/abs/2110.00061) by Brandon Smock, Rohith Pesala, Robin Abraham.
 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/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. **[Time Series Transformer](https://huggingface.co/docs/transformers/model_doc/time_series_transformer)**  (from HuggingFace).
 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.
@@ -375,16 +398,18 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 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/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. **[VideoMAE](https://huggingface.co/docs/transformers/main/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
+1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
 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/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. **[ViTMSN](https://huggingface.co/docs/transformers/model_doc/vit_msn)** (from Meta AI) released with the paper [Masked Siamese Networks for Label-Efficient Learning](https://arxiv.org/abs/2204.07141) by Mahmoud Assran, Mathilde Caron, Ishan Misra, Piotr Bojanowski, Florian Bordes, Pascal Vincent, Armand Joulin, Michael Rabbat, Nicolas Ballas.
 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. **[X-CLIP](https://huggingface.co/docs/transformers/main/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
+1. **[Whisper](https://huggingface.co/docs/transformers/model_doc/whisper)** (from OpenAI) released with the paper [Robust Speech Recognition via Large-Scale Weak Supervision](https://cdn.openai.com/papers/whisper.pdf) by Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, Ilya Sutskever.
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
 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.
@@ -399,7 +424,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 
 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).
 
-These implementations have been tested on several datasets (see the example scripts) and should match the performance of the original implementations. You can find more details on performance in the Examples section of the [documentation](https://huggingface.co/docs/transformers/examples).
+These implementations have been tested on several datasets (see the example scripts) and should match the performance of the original implementations. You can find more details on performance in the Examples section of the [documentation](https://github.com/huggingface/transformers/tree/main/examples).
 
 
 ## Learn more

README_es.md

@@ -0,0 +1,457 @@
+<!---
+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.
+-->
+
+<p align="center">
+    <br>
+    <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers_logo_name.png" width="400"/>
+    <br>
+<p>
+<p align="center">
+    <a href="https://circleci.com/gh/huggingface/transformers">
+        <img alt="Build" src="https://img.shields.io/circleci/build/github/huggingface/transformers/main">
+    </a>
+    <a href="https://github.com/huggingface/transformers/blob/main/LICENSE">
+        <img alt="GitHub" src="https://img.shields.io/github/license/huggingface/transformers.svg?color=blue">
+    </a>
+    <a href="https://huggingface.co/docs/transformers/index">
+        <img alt="Documentation" src="https://img.shields.io/website/http/huggingface.co/docs/transformers/index.svg?down_color=red&down_message=offline&up_message=online">
+    </a>
+    <a href="https://github.com/huggingface/transformers/releases">
+        <img alt="GitHub release" src="https://img.shields.io/github/release/huggingface/transformers.svg">
+    </a>
+    <a href="https://github.com/huggingface/transformers/blob/main/CODE_OF_CONDUCT.md">
+        <img alt="Contributor Covenant" src="https://img.shields.io/badge/Contributor%20Covenant-v2.0%20adopted-ff69b4.svg">
+    </a>
+    <a href="https://zenodo.org/badge/latestdoi/155220641"><img src="https://zenodo.org/badge/155220641.svg" alt="DOI"></a>
+</p>
+
+<h4 align="center">
+    <p>
+        <a href="https://github.com/huggingface/transformers/">English</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_zh-hans.md">简体中文</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_zh-hant.md">繁體中文</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_ko.md">한국어</a> |
+        <b>Español</b> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_ja.md">日本語</a>
+    <p>
+</h4>
+
+<h3 align="center">
+    <p>Lo último de Machine Learning para JAX, PyTorch y TensorFlow</p>
+</h3>
+
+<h3 align="center">
+    <a href="https://hf.co/course"><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/course_banner.png"></a>
+</h3>
+
+🤗 Transformers aporta miles de modelos preentrenados Para realizar tareas en diferentes modalidades como texto, vision, y audio.
+
+Estos modelos pueden ser aplicados en:
+
+* 📝 Texto, Para tareas como clasificación de texto, extracción de información, responder preguntas, resumir, traducir, generación de texto, en más de 100 idiomas.
+* 🖼️ Imágenes, para tareas como clasificación de imágenes, detección the objetos, y segmentación.
+* 🗣️ Audio, para tareas como reconocimiento de voz y clasificación de audio.
+
+Los modelos de Transformer también pueden realizar tareas en **muchas modalidades combinadas**, como responder pregunstas, reconocimiento de carácteres ópticos,extracción de información de documentos escaneados, clasificación de video, y respuesta de preguntas visuales.
+
+🤗 Transformers aporta APIs para descargar rápidamente y usar estos modelos preentrenados en un texto dado, afinarlos en tus propios sets de datos y compartirlos con la comunidad en nuestro [centro de modelos](https://huggingface.co/models). Al mismo tiempo, cada módulo de Python que define una arquitectura es completamente independiente y se puede modificar para permitir experimentos de investigación rápidos.
+
+🤗 Transformers está respaldado por las tres bibliotecas de deep learning más populares — [Jax](https://jax.readthedocs.io/en/latest/), [PyTorch](https://pytorch.org/) y [TensorFlow](https://www.tensorflow.org/) — con una perfecta integración entre ellos. Es sencillo entrenar sus modelos con uno antes de cargarlos para la inferencia con el otro.
+
+## Demostraciones en línea
+
+Puedes probar la mayoría de nuestros modelos directamente en sus páginas desde el [centro de modelos](https://huggingface.co/models). También ofrecemos [alojamiento de modelos privados, control de versiones y una API de inferencia](https://huggingface.co/pricing) para modelos públicos y privados.
+
+Aquí hay algunos ejemplos:
+
+ En procesamiento del lenguaje natural:
+- [Terminación de palabras enmascaradas con BERT](https://huggingface.co/bert-base-uncased?text=Paris+is+the+%5BMASK%5D+of+France)
+- [Reconocimiento del nombre de la entidad con Electra](https://huggingface.co/dbmdz/electra-large-discriminator-finetuned-conll03-english?text=My+name+is+Sarah+and+I+live+in+London+city)
+- [Generación de texto con GPT-2](https://huggingface.co/gpt2?text=A+long+time+ago%2C+)
+- [Inferencia del lenguaje natural con RoBERTa](https://huggingface.co/roberta-large-mnli?text=The+dog+was+lost.+Nobody+lost+any+animal)
+- [Resumen con BART](https://huggingface.co/facebook/bart-large-cnn?text=The+tower+is+324+metres+%281%2C063+ft%29+tall%2C+about+the+same+height+as+an+81-storey+building%2C+and+the+tallest+structure+in+Paris.+Its+base+is+square%2C+measuring+125+metres+%28410+ft%29+on+each+side.+During+its+construction%2C+the+Eiffel+Tower+surpassed+the+Washington+Monument+to+become+the+tallest+man-made+structure+in+the+world%2C+a+title+it+held+for+41+years+until+the+Chrysler+Building+in+New+York+City+was+finished+in+1930.+It+was+the+first+structure+to+reach+a+height+of+300+metres.+Due+to+the+addition+of+a+broadcasting+aerial+at+the+top+of+the+tower+in+1957%2C+it+is+now+taller+than+the+Chrysler+Building+by+5.2+metres+%2817+ft%29.+Excluding+transmitters%2C+the+Eiffel+Tower+is+the+second+tallest+free-standing+structure+in+France+after+the+Millau+Viaduct)
+- [Responder a preguntas con DistilBERT](https://huggingface.co/distilbert-base-uncased-distilled-squad?text=Which+name+is+also+used+to+describe+the+Amazon+rainforest+in+English%3F&context=The+Amazon+rainforest+%28Portuguese%3A+Floresta+Amaz%C3%B4nica+or+Amaz%C3%B4nia%3B+Spanish%3A+Selva+Amaz%C3%B3nica%2C+Amazon%C3%ADa+or+usually+Amazonia%3B+French%3A+For%C3%AAt+amazonienne%3B+Dutch%3A+Amazoneregenwoud%29%2C+also+known+in+English+as+Amazonia+or+the+Amazon+Jungle%2C+is+a+moist+broadleaf+forest+that+covers+most+of+the+Amazon+basin+of+South+America.+This+basin+encompasses+7%2C000%2C000+square+kilometres+%282%2C700%2C000+sq+mi%29%2C+of+which+5%2C500%2C000+square+kilometres+%282%2C100%2C000+sq+mi%29+are+covered+by+the+rainforest.+This+region+includes+territory+belonging+to+nine+nations.+The+majority+of+the+forest+is+contained+within+Brazil%2C+with+60%25+of+the+rainforest%2C+followed+by+Peru+with+13%25%2C+Colombia+with+10%25%2C+and+with+minor+amounts+in+Venezuela%2C+Ecuador%2C+Bolivia%2C+Guyana%2C+Suriname+and+French+Guiana.+States+or+departments+in+four+nations+contain+%22Amazonas%22+in+their+names.+The+Amazon+represents+over+half+of+the+planet%27s+remaining+rainforests%2C+and+comprises+the+largest+and+most+biodiverse+tract+of+tropical+rainforest+in+the+world%2C+with+an+estimated+390+billion+individual+trees+divided+into+16%2C000+species)
+- [Traducción con T5](https://huggingface.co/t5-base?text=My+name+is+Wolfgang+and+I+live+in+Berlin)
+
+En visión de ordenador:
+- [Clasificación de imágenes con ViT](https://huggingface.co/google/vit-base-patch16-224)
+- [Detección de objetos con DETR](https://huggingface.co/facebook/detr-resnet-50)
+- [Segmentación semántica con SegFormer](https://huggingface.co/nvidia/segformer-b0-finetuned-ade-512-512)
+- [Segmentación panóptica con DETR](https://huggingface.co/facebook/detr-resnet-50-panoptic)
+
+En Audio:
+- [Reconocimiento de voz automático con Wav2Vec2](https://huggingface.co/facebook/wav2vec2-base-960h)
+- [Detección de palabras clave con Wav2Vec2](https://huggingface.co/superb/wav2vec2-base-superb-ks)
+
+En tareas multimodales:
+- [Respuesta visual a preguntas con ViLT](https://huggingface.co/dandelin/vilt-b32-finetuned-vqa)
+
+**[Escribe con Transformer](https://transformer.huggingface.co)**, construido por el equipo de Hugging Face, es la demostración oficial de las capacidades de generación de texto de este repositorio.
+
+## Si está buscando soporte personalizado del equipo de Hugging Face
+
+<a target="_blank" href="https://huggingface.co/support">
+    <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>
+
+## Tour rápido
+
+Para usar inmediatamente un modelo en una entrada determinada (texto, imagen, audio, ...), proporcionamos la API de `pipeline`. Los pipelines agrupan un modelo previamente entrenado con el preprocesamiento que se usó durante el entrenamiento de ese modelo. Aquí se explica cómo usar rápidamente un pipeline para clasificar textos positivos frente a negativos:
+
+```python
+>>> from transformers import pipeline
+
+# Allocate a pipeline for sentiment-analysis
+>>> classifier = pipeline('sentiment-analysis')
+>>> classifier('We are very happy to introduce pipeline to the transformers repository.')
+[{'label': 'POSITIVE', 'score': 0.9996980428695679}]
+```
+
+La segunda línea de código descarga y almacena en caché el modelo previamente entrenado que usa la canalización, mientras que la tercera lo evalúa en el texto dado. Aquí la respuesta es "positiva" con una confianza del 99,97%.
+
+Muchas tareas tienen un `pipeline` preentrenado listo para funcionar, en NLP pero también en visión por ordenador y habla. Por ejemplo, podemos extraer fácilmente los objetos detectados en una imagen:
+
+``` python
+>>> import requests
+>>> from PIL import Image
+>>> from transformers import pipeline
+
+# 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}}]
+```
+
+Aquí obtenemos una lista de objetos detectados en la imagen, con un cuadro que rodea el objeto y una puntuación de confianza. Aquí está la imagen original a la derecha, con las predicciones mostradas a la izquierda:
+
+<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>
+
+Puedes obtener más información sobre las tareas admitidas por la API de `pipeline` en [este tutorial](https://huggingface.co/docs/transformers/task_summary).
+
+Además de `pipeline`, para descargar y usar cualquiera de los modelos previamente entrenados en su tarea dada, todo lo que necesita son tres líneas de código. Aquí está la versión de PyTorch:
+```python
+>>> from transformers import AutoTokenizer, AutoModel
+
+>>> tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
+>>> model = AutoModel.from_pretrained("bert-base-uncased")
+
+>>> inputs = tokenizer("Hello world!", return_tensors="pt")
+>>> outputs = model(**inputs)
+```
+
+Y aquí está el código equivalente para TensorFlow:
+```python
+>>> from transformers import AutoTokenizer, TFAutoModel
+
+>>> tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
+>>> model = TFAutoModel.from_pretrained("bert-base-uncased")
+
+>>> inputs = tokenizer("Hello world!", return_tensors="tf")
+>>> outputs = model(**inputs)
+```
+
+El tokenizador es responsable de todo el preprocesamiento que espera el modelo preentrenado y se puede llamar directamente en una sola cadena (como en los ejemplos anteriores) o en una lista. Dará como resultado un diccionario que puedes usar en el código descendente o simplemente pasarlo directamente a su modelo usando el operador de desempaquetado de argumento **.
+
+El modelo en si es un [Pytorch `nn.Module`](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) normal o un [TensorFlow `tf.keras.Model`](https://www.tensorflow.org/api_docs/python/tf/keras/Model) (dependiendo De tu backend) que puedes usar de forma habitual. [Este tutorial](https://huggingface.co/docs/transformers/training) explica cómo integrar un modelo de este tipo en un ciclo de entrenamiento PyTorch o TensorFlow clásico, o como usar nuestra API `Trainer` para ajustar rápidamente un nuevo conjunto de datos.
+
+## ¿Por qué debo usar transformers?
+
+1. Modelos de última generación fáciles de usar:
+    - Alto rendimiento en comprensión y generación de lenguaje natural, visión artificial y tareas de audio.
+    - Baja barrera de entrada para educadores y profesionales.
+    - Pocas abstracciones de cara al usuario con solo tres clases para aprender.
+    - Una API unificada para usar todos nuestros modelos preentrenados.
+
+1. Menores costes de cómputo, menor huella de carbono:
+    - Los investigadores pueden compartir modelos entrenados en lugar de siempre volver a entrenar.
+    - Los profesionales pueden reducir el tiempo de cómputo y los costos de producción.
+    - Docenas de arquitecturas con más de 60 000 modelos preentrenados en todas las modalidades.
+
+1. Elija el marco adecuado para cada parte de la vida útil de un modelo:
+    - Entrene modelos de última generación en 3 líneas de código.
+    - Mueva un solo modelo entre los marcos TF2.0/PyTorch/JAX a voluntad.
+    - Elija sin problemas el marco adecuado para la formación, la evaluación y la producción.
+
+1. Personalice fácilmente un modelo o un ejemplo según sus necesidades:
+    - Proporcionamos ejemplos de cada arquitectura para reproducir los resultados publicados por sus autores originales..
+    - Los internos del modelo están expuestos lo más consistentemente posible..
+    - Los archivos modelo se pueden usar independientemente de la biblioteca para experimentos rápidos.
+
+## ¿Por qué no debería usar transformers?
+
+- Esta biblioteca no es una caja de herramientas modular de bloques de construcción para redes neuronales. El código en los archivos del modelo no se refactoriza con abstracciones adicionales a propósito, de modo que los investigadores puedan iterar rápidamente en cada uno de los modelos sin sumergirse en abstracciones/archivos adicionales.
+- La API de entrenamiento no está diseñada para funcionar en ningún modelo, pero está optimizada para funcionar con los modelos proporcionados por la biblioteca. Para bucles genéricos de aprendizaje automático, debe usar otra biblioteca (posiblemente, [Accelerate](https://huggingface.co/docs/accelerate)).
+- Si bien nos esforzamos por presentar tantos casos de uso como sea posible, los scripts en nuestra [carpeta de ejemplos](https://github.com/huggingface/transformers/tree/main/examples) son solo eso: ejemplos. Se espera que no funcionen de forma inmediata en su problema específico y que deba cambiar algunas líneas de código para adaptarlas a sus necesidades.
+
+## Instalación
+
+### Con pip
+
+Este repositorio está probado en Python 3.6+, Flax 0.3.2+, PyTorch 1.3.1+ y TensorFlow 2.3+.
+
+Deberías instalar 🤗 Transformers en un [ambiente virtual](https://docs.python.org/3/library/venv.html). Si no estas familiarizado con los entornos virtuales de Python, consulta la [guía de usuario](https://packaging.python.org/guides/installing-using-pip-and-virtual-environments/).
+
+Primero, crea un entorno virtual con la versión de Python que vas a usar y actívalo.
+
+Luego, deberás instalar al menos uno de Flax, PyTorch o TensorFlow.
+Por favor, ve a la [página de instalación de TensorFlow](https://www.tensorflow.org/install/), [página de instalación de PyTorch](https://pytorch.org/get-started/locally/#start-locally) y/o las páginas de instalación de [Flax](https://github.com/google/flax#quick-install) y [Jax](https://github.com/google/jax#installation) con respecto al comando de instalación específico para tu plataforma.
+
+Cuando se ha instalado uno de esos backends, los 🤗 Transformers se pueden instalar usando pip de la siguiente manera:
+
+```bash
+pip install transformers
+```
+
+Si deseas jugar con los ejemplos o necesitas la última versión del código y no puedes esperar a una nueva versión, tienes que [instalar la librería de la fuente](https://huggingface.co/docs/transformers/installation#installing-from-source).
+
+### Con conda
+
+Desde la versión v4.0.0 de Transformers, ahora tenemos un canal conda: `huggingface`.
+
+🤗 Transformers se puede instalar usando conda de la siguiente manera:
+
+```shell script
+conda install -c huggingface transformers
+```
+
+Sigue las páginas de instalación de Flax, PyTorch o TensorFlow para ver cómo instalarlos con conda.
+
+> **_NOTA:_**  En Windows, es posible que se le pida que active el modo de desarrollador para beneficiarse del almacenamiento en caché. Si esta no es una opción para usted, háganoslo saber en [esta issue](https://github.com/huggingface/huggingface_hub/issues/1062).
+
+## Arquitecturas modelo
+
+**[Todos los puntos de control del modelo](https://huggingface.co/models)** aportados por 🤗 Transformers están perfectamente integrados desde huggingface.co [Centro de modelos](https://huggingface.co) donde son subidos directamente por los [usuarios](https://huggingface.co/users) y [organizaciones](https://huggingface.co/organizations).
+
+Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=https://huggingface.co/api/shields/models&color=brightgreen)
+
+🤗 Transformers actualmente proporciona las siguientes arquitecturas (ver [aquí](https://huggingface.co/docs/transformers/model_summary) para un resumen de alto nivel de cada uno de ellas.):
+
+1. **[ALBERT](https://huggingface.co/docs/transformers/model_doc/albert)** (from Google Research and the Toyota Technological Institute at Chicago) released with the paper [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://arxiv.org/abs/1909.11942), by Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin Gimpel, Piyush Sharma, Radu Soricut.
+1. **[Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer)** (from MIT) released with the paper [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) by Yuan Gong, Yu-An Chung, James Glass.
+1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (from Facebook) released with the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) by Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer.
+1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (from École polytechnique) released with the paper [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) by Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis.
+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. **[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. **[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 [BigScience 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. **[Chinese-CLIP](https://huggingface.co/docs/transformers/model_doc/chinese_clip)** (from OFA-Sys) released with the paper [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) by An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou. 
+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. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (from University of Göttingen) released with the paper [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) by Timo Lüddecke and Alexander Ecker.
+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. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
+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. **[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. **[Deformable DETR](https://huggingface.co/docs/transformers/model_doc/deformable_detr)** (from SenseTime Research) released with the paper [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://arxiv.org/abs/2010.04159) by Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, Jifeng Dai.
+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. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (from SHI Labs) released with the paper [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001) by Ali Hassani and Humphrey Shi.
+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. **[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. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (from NAVER), released together with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
+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. **[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. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (from Meta AI) are transformer protein language models.  **ESM-1b** was released with the paper [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus. **ESM-1v** was released with the paper [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648) by Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives. **ESM-2** was released with the paper [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
+1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
+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/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 NeoX Japanese](https://huggingface.co/docs/transformers/model_doc/gpt_neox_japanese)** (from ABEJA) released by Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori.
+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/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. **[Jukebox](https://huggingface.co/docs/transformers/model_doc/jukebox)** (from OpenAI) released with the paper [Jukebox: A Generative Model for Music](https://arxiv.org/pdf/2005.00341.pdf) by Prafulla Dhariwal, Heewoo Jun, Christine Payne, Jong Wook Kim, Alec Radford, 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/layoutxlm)** (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. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (from South China University of Technology) released with the paper [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) by Jiapeng Wang, Lianwen Jin, Kai Ding.
+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. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
+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. **[MobileNetV1](https://huggingface.co/docs/transformers/model_doc/mobilenet_v1)** (from Google Inc.) released with the paper [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://arxiv.org/abs/1704.04861) by Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam.
+1. **[MobileNetV2](https://huggingface.co/docs/transformers/model_doc/mobilenet_v2)** (from Google Inc.) released with the paper [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381) by Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen.
+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. **[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. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
+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. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, 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/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. **[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. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (from WeChatAI) released with the paper [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) by HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou.
+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.
+1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
+1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
+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/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. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[SwitchTransformers](https://huggingface.co/docs/transformers/model_doc/switch_transformers)** (from Google) released with the paper [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://arxiv.org/abs/2101.03961) by William Fedus, Barret Zoph, Noam Shazeer.
+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. **[Table Transformer](https://huggingface.co/docs/transformers/model_doc/table-transformer)** (from Microsoft Research) released with the paper [PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents](https://arxiv.org/abs/2110.00061) by Brandon Smock, Rohith Pesala, Robin Abraham.
+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/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. **[Time Series Transformer](https://huggingface.co/docs/transformers/model_doc/time_series_transformer)**  (from HuggingFace).
+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/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. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
+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/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. **[ViTMSN](https://huggingface.co/docs/transformers/model_doc/vit_msn)** (from Meta AI) released with the paper [Masked Siamese Networks for Label-Efficient Learning](https://arxiv.org/abs/2204.07141) by Mahmoud Assran, Mathilde Caron, Ishan Misra, Piotr Bojanowski, Florian Bordes, Pascal Vincent, Armand Joulin, Michael Rabbat, Nicolas Ballas.
+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. **[Whisper](https://huggingface.co/docs/transformers/model_doc/whisper)** (from OpenAI) released with the paper [Robust Speech Recognition via Large-Scale Weak Supervision](https://cdn.openai.com/papers/whisper.pdf) by Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, Ilya Sutskever.
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
+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. **[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. **[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. ¿Quieres aportar un nuevo modelo? Hemos agregado una **guía detallada y plantillas** para guiarte en el proceso de agregar un nuevo modelo. Puedes encontrarlos en la carpeta de [`templates`](./templates) del repositorio. Asegúrate de revisar las [pautas de contribución](./CONTRIBUTING.md) y comunícate con los mantenedores o abra un problema para recopilar comentarios antes de comenzar su PR.
+
+Para comprobar si cada modelo tiene una implementación en Flax, PyTorch o TensorFlow, o tiene un tokenizador asociado respaldado por la librería 🤗 Tokenizers , ve a [esta tabla](https://huggingface.co/docs/transformers/index#supported-frameworks).
+
+Estas implementaciones se han probado en varios conjuntos de datos (consulte los scripts de ejemplo) y deberían coincidir con el rendimiento de las implementaciones originales. Puede encontrar más detalles sobre el rendimiento en la sección Examples de la [documentación](https://github.com/huggingface/transformers/tree/main/examples).
+
+
+## Aprender más
+
+| Sección | Descripción |
+|-|-|
+| [Documentación](https://huggingface.co/docs/transformers/) | Toda la documentación de la API y tutoriales |
+| [Resumen de tareas](https://huggingface.co/docs/transformers/task_summary) | Tareas soportadas 🤗 Transformers |
+| [Tutorial de preprocesAmiento](https://huggingface.co/docs/transformers/preprocessing) | Usando la clase `Tokenizer` para preparar datos para los modelos |
+| [Entrenamiento y puesta a punto](https://huggingface.co/docs/transformers/training) | Usando los modelos aportados por 🤗 Transformers en un bucle de entreno de PyTorch/TensorFlow y la API de `Trainer` |
+| [Recorrido rápido: secuencias de comandos de ajuste/uso](https://github.com/huggingface/transformers/tree/main/examples) | Scripts de ejemplo para ajustar modelos en una amplia gama de tareas |
+| [Compartir y subir modelos](https://huggingface.co/docs/transformers/model_sharing) | Carga y comparte tus modelos perfeccionados con la comunidad |
+| [Migración](https://huggingface.co/docs/transformers/migration) | Migra a 🤗 Transformers desde `pytorch-transformers` o `pytorch-pretrained-bert` |
+
+## Citación
+
+Ahora nosotros tenemos un [papel](https://www.aclweb.org/anthology/2020.emnlp-demos.6/) que puedes citar para la librería de  🤗 Transformers:
+```bibtex
+@inproceedings{wolf-etal-2020-transformers,
+    title = "Transformers: State-of-the-Art Natural Language Processing",
+    author = "Thomas Wolf and Lysandre Debut and Victor Sanh and Julien Chaumond and Clement Delangue and Anthony Moi and Pierric Cistac and Tim Rault and Rémi Louf and Morgan Funtowicz and Joe Davison and Sam Shleifer and Patrick von Platen and Clara Ma and Yacine Jernite and Julien Plu and Canwen Xu and Teven Le Scao and Sylvain Gugger and Mariama Drame and Quentin Lhoest and Alexander M. Rush",
+    booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
+    month = oct,
+    year = "2020",
+    address = "Online",
+    publisher = "Association for Computational Linguistics",
+    url = "https://www.aclweb.org/anthology/2020.emnlp-demos.6",
+    pages = "38--45"
+}
+```

README_ja.md

@@ -0,0 +1,492 @@
+<!---
+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.
+-->
+
+<!---
+A useful guide for English-Traditional Japanese translation of Hugging Face documentation
+- Use square quotes, e.g.,「引用」
+
+Dictionary
+
+API: API(翻訳しない)
+add: 追加
+checkpoint: チェックポイント
+code: コード
+community: コミュニティ
+confidence: 信頼度
+dataset: データセット
+documentation: ドキュメント
+example: 例
+finetune: 微調整
+Hugging Face: Hugging Face(翻訳しない)
+implementation: 実装
+inference: 推論
+library: ライブラリ
+module: モジュール
+NLP/Natural Language Processing: NLPと表示される場合は翻訳されず、Natural Language Processingと表示される場合は翻訳される
+online demos: オンラインデモ
+pipeline: pipeline(翻訳しない)
+pretrained/pretrain: 学習済み
+Python data structures (e.g., list, set, dict): リスト、セット、ディクショナリと訳され、括弧内は原文英語
+repository: repository(翻訳しない)
+summary: 概要
+token-: token-(翻訳しない)
+Trainer: Trainer(翻訳しない)
+transformer: transformer(翻訳しない)
+tutorial: チュートリアル
+user: ユーザ
+-->
+
+<p align="center">
+    <br>
+    <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers_logo_name.png" width="400"/>
+    <br>
+<p>
+<p align="center">
+    <a href="https://circleci.com/gh/huggingface/transformers">
+        <img alt="Build" src="https://img.shields.io/circleci/build/github/huggingface/transformers/main">
+    </a>
+    <a href="https://github.com/huggingface/transformers/blob/main/LICENSE">
+        <img alt="GitHub" src="https://img.shields.io/github/license/huggingface/transformers.svg?color=blue">
+    </a>
+    <a href="https://huggingface.co/docs/transformers/index">
+        <img alt="Documentation" src="https://img.shields.io/website/http/huggingface.co/docs/transformers/index.svg?down_color=red&down_message=offline&up_message=online">
+    </a>
+    <a href="https://github.com/huggingface/transformers/releases">
+        <img alt="GitHub release" src="https://img.shields.io/github/release/huggingface/transformers.svg">
+    </a>
+    <a href="https://github.com/huggingface/transformers/blob/main/CODE_OF_CONDUCT.md">
+        <img alt="Contributor Covenant" src="https://img.shields.io/badge/Contributor%20Covenant-v2.0%20adopted-ff69b4.svg">
+    </a>
+    <a href="https://zenodo.org/badge/latestdoi/155220641"><img src="https://zenodo.org/badge/155220641.svg" alt="DOI"></a>
+</p>
+
+<h4 align="center">
+    <p>
+        <a href="https://github.com/huggingface/transformers/">English</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_zh-hans.md">简体中文</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_zh-hant.md">繁體中文</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_ko.md">한국어</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_es.md">Español</a> |
+        <b>日本語</b>
+    <p>
+</h4>
+
+<h3 align="center">
+    <p>JAX、PyTorch、TensorFlowのための最先端機械学習</p>
+</h3>
+
+<h3 align="center">
+    <a href="https://hf.co/course"><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/course_banner.png"></a>
+</h3>
+
+🤗Transformersは、テキスト、視覚、音声などの異なるモダリティに対してタスクを実行するために、事前に学習させた数千のモデルを提供します。
+
+これらのモデルは次のような場合に適用できます:
+
+* 📝 テキストは、テキストの分類、情報抽出、質問応答、要約、翻訳、テキスト生成などのタスクのために、100以上の言語に対応しています。
+* 🖼️ 画像分類、物体検出、セグメンテーションなどのタスクのための画像。
+* 🗣️ 音声は、音声認識や音声分類などのタスクに使用します。
+
+トランスフォーマーモデルは、テーブル質問応答、光学文字認識、スキャン文書からの情報抽出、ビデオ分類、視覚的質問応答など、**複数のモダリティを組み合わせた**タスクも実行可能です。
+
+🤗Transformersは、与えられたテキストに対してそれらの事前学習されたモデルを素早くダウンロードして使用し、あなた自身のデータセットでそれらを微調整し、私たちの[model hub](https://huggingface.co/models)でコミュニティと共有するためのAPIを提供します。同時に、アーキテクチャを定義する各Pythonモジュールは完全にスタンドアロンであり、迅速な研究実験を可能にするために変更することができます。
+
+🤗Transformersは[Jax](https://jax.readthedocs.io/en/latest/)、[PyTorch](https://pytorch.org/)、[TensorFlow](https://www.tensorflow.org/)という3大ディープラーニングライブラリーに支えられ、それぞれのライブラリをシームレスに統合しています。片方でモデルを学習してから、もう片方で推論用にロードするのは簡単なことです。
+
+## オンラインデモ
+
+[model hub](https://huggingface.co/models)から、ほとんどのモデルのページで直接テストすることができます。また、パブリックモデル、プライベートモデルに対して、[プライベートモデルのホスティング、バージョニング、推論API](https://huggingface.co/pricing)を提供しています。
+
+以下はその一例です:
+
+ 自然言語処理にて:
+- [BERTによるマスクドワード補完](https://huggingface.co/bert-base-uncased?text=Paris+is+the+%5BMASK%5D+of+France)
+- [Electraによる名前実体認識](https://huggingface.co/dbmdz/electra-large-discriminator-finetuned-conll03-english?text=My+name+is+Sarah+and+I+live+in+London+city)
+- [GPT-2によるテキスト生成](https://huggingface.co/gpt2?text=A+long+time+ago%2C+)
+- [RoBERTaによる自然言語推論](https://huggingface.co/roberta-large-mnli?text=The+dog+was+lost.+Nobody+lost+any+animal)
+- [BARTによる要約](https://huggingface.co/facebook/bart-large-cnn?text=The+tower+is+324+metres+%281%2C063+ft%29+tall%2C+about+the+same+height+as+an+81-storey+building%2C+and+the+tallest+structure+in+Paris.+Its+base+is+square%2C+measuring+125+metres+%28410+ft%29+on+each+side.+During+its+construction%2C+the+Eiffel+Tower+surpassed+the+Washington+Monument+to+become+the+tallest+man-made+structure+in+the+world%2C+a+title+it+held+for+41+years+until+the+Chrysler+Building+in+New+York+City+was+finished+in+1930.+It+was+the+first+structure+to+reach+a+height+of+300+metres.+Due+to+the+addition+of+a+broadcasting+aerial+at+the+top+of+the+tower+in+1957%2C+it+is+now+taller+than+the+Chrysler+Building+by+5.2+metres+%2817+ft%29.+Excluding+transmitters%2C+the+Eiffel+Tower+is+the+second+tallest+free-standing+structure+in+France+after+the+Millau+Viaduct)
+- [DistilBERTによる質問応答](https://huggingface.co/distilbert-base-uncased-distilled-squad?text=Which+name+is+also+used+to+describe+the+Amazon+rainforest+in+English%3F&context=The+Amazon+rainforest+%28Portuguese%3A+Floresta+Amaz%C3%B4nica+or+Amaz%C3%B4nia%3B+Spanish%3A+Selva+Amaz%C3%B3nica%2C+Amazon%C3%ADa+or+usually+Amazonia%3B+French%3A+For%C3%AAt+amazonienne%3B+Dutch%3A+Amazoneregenwoud%29%2C+also+known+in+English+as+Amazonia+or+the+Amazon+Jungle%2C+is+a+moist+broadleaf+forest+that+covers+most+of+the+Amazon+basin+of+South+America.+This+basin+encompasses+7%2C000%2C000+square+kilometres+%282%2C700%2C000+sq+mi%29%2C+of+which+5%2C500%2C000+square+kilometres+%282%2C100%2C000+sq+mi%29+are+covered+by+the+rainforest.+This+region+includes+territory+belonging+to+nine+nations.+The+majority+of+the+forest+is+contained+within+Brazil%2C+with+60%25+of+the+rainforest%2C+followed+by+Peru+with+13%25%2C+Colombia+with+10%25%2C+and+with+minor+amounts+in+Venezuela%2C+Ecuador%2C+Bolivia%2C+Guyana%2C+Suriname+and+French+Guiana.+States+or+departments+in+four+nations+contain+%22Amazonas%22+in+their+names.+The+Amazon+represents+over+half+of+the+planet%27s+remaining+rainforests%2C+and+comprises+the+largest+and+most+biodiverse+tract+of+tropical+rainforest+in+the+world%2C+with+an+estimated+390+billion+individual+trees+divided+into+16%2C000+species)
+- [T5による翻訳](https://huggingface.co/t5-base?text=My+name+is+Wolfgang+and+I+live+in+Berlin)
+
+コンピュータビジョンにて:
+- [ViTによる画像分類](https://huggingface.co/google/vit-base-patch16-224)
+- [DETRによる物体検出](https://huggingface.co/facebook/detr-resnet-50)
+- [SegFormerによるセマンティックセグメンテーション](https://huggingface.co/nvidia/segformer-b0-finetuned-ade-512-512)
+- [DETRによるパノプティックセグメンテーション](https://huggingface.co/facebook/detr-resnet-50-panoptic)
+
+オーディオにて:
+- [Wav2Vec2による自動音声認識](https://huggingface.co/facebook/wav2vec2-base-960h)
+- [Wav2Vec2によるキーワード検索](https://huggingface.co/superb/wav2vec2-base-superb-ks)
+
+マルチモーダルなタスクにて:
+- [ViLTによる視覚的質問応答](https://huggingface.co/dandelin/vilt-b32-finetuned-vqa)
+
+Hugging Faceチームによって作られた **[トランスフォーマーを使った書き込み](https://transformer.huggingface.co)** は、このリポジトリのテキスト生成機能の公式デモである。
+
+## Hugging Faceチームによるカスタム・サポートをご希望の場合
+
+<a target="_blank" href="https://huggingface.co/support">
+    <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>
+
+## クイックツアー
+
+与えられた入力（テキスト、画像、音声、...）に対してすぐにモデルを使うために、我々は`pipeline`というAPIを提供しております。pipelineは、学習済みのモデルと、そのモデルの学習時に使用された前処理をグループ化したものです。以下は、肯定的なテキストと否定的なテキストを分類するためにpipelineを使用する方法です:
+
+```python
+>>> from transformers import pipeline
+
+# Allocate a pipeline for sentiment-analysis
+>>> classifier = pipeline('sentiment-analysis')
+>>> classifier('We are very happy to introduce pipeline to the transformers repository.')
+[{'label': 'POSITIVE', 'score': 0.9996980428695679}]
+```
+
+2行目のコードでは、pipelineで使用される事前学習済みモデルをダウンロードしてキャッシュし、3行目では与えられたテキストに対してそのモデルを評価します。ここでは、答えは99.97%の信頼度で「ポジティブ」です。
+
+自然言語処理だけでなく、コンピュータビジョンや音声処理においても、多くのタスクにはあらかじめ訓練された`pipeline`が用意されている。例えば、画像から検出された物体を簡単に抽出することができる:
+
+``` python
+>>> import requests
+>>> from PIL import Image
+>>> from transformers import pipeline
+
+# 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}}]
+```
+
+ここでは、画像から検出されたオブジェクトのリストが得られ、オブジェクトを囲むボックスと信頼度スコアが表示されます。左側が元画像、右側が予測結果を表示したものです:
+
+<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>
+
+[このチュートリアル](https://huggingface.co/docs/transformers/task_summary)では、`pipeline`APIでサポートされているタスクについて詳しく説明しています。
+
+`pipeline`に加えて、与えられたタスクに学習済みのモデルをダウンロードして使用するために必要なのは、3行のコードだけです。以下はPyTorchのバージョンです:
+```python
+>>> from transformers import AutoTokenizer, AutoModel
+
+>>> tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
+>>> model = AutoModel.from_pretrained("bert-base-uncased")
+
+>>> inputs = tokenizer("Hello world!", return_tensors="pt")
+>>> outputs = model(**inputs)
+```
+
+And here is the equivalent code for TensorFlow:
+```python
+>>> from transformers import AutoTokenizer, TFAutoModel
+
+>>> tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
+>>> model = TFAutoModel.from_pretrained("bert-base-uncased")
+
+>>> inputs = tokenizer("Hello world!", return_tensors="tf")
+>>> outputs = model(**inputs)
+```
+
+トークナイザは学習済みモデルが期待するすべての前処理を担当し、単一の文字列 (上記の例のように) またはリストに対して直接呼び出すことができます。これは下流のコードで使用できる辞書を出力します。また、単純に ** 引数展開演算子を使用してモデルに直接渡すこともできます。
+
+モデル自体は通常の[Pytorch `nn.Module`](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) または [TensorFlow `tf.keras.Model`](https://www.tensorflow.org/api_docs/python/tf/keras/Model) (バックエンドによって異なる)で、通常通り使用することが可能です。[このチュートリアル](https://huggingface.co/docs/transformers/training)では、このようなモデルを従来のPyTorchやTensorFlowの学習ループに統合する方法や、私たちの`Trainer`APIを使って新しいデータセットで素早く微調整を行う方法について説明します。
+
+## なぜtransformersを使う必要があるのでしょうか？
+
+1. 使いやすい最新モデル:
+    - 自然言語理解・生成、コンピュータビジョン、オーディオの各タスクで高いパフォーマンスを発揮します。
+    - 教育者、実務者にとっての低い参入障壁。
+    - 学習するクラスは3つだけで、ユーザが直面する抽象化はほとんどありません。
+    - 学習済みモデルを利用するための統一されたAPI。
+
+1. 低い計算コスト、少ないカーボンフットプリント:
+    - 研究者は、常に再トレーニングを行うのではなく、トレーニングされたモデルを共有することができます。
+    - 実務家は、計算時間や生産コストを削減することができます。
+    - すべてのモダリティにおいて、60,000以上の事前学習済みモデルを持つ数多くのアーキテクチャを提供します。
+
+1. モデルのライフタイムのあらゆる部分で適切なフレームワークを選択可能:
+    - 3行のコードで最先端のモデルをトレーニング。
+    - TF2.0/PyTorch/JAXフレームワーク間で1つのモデルを自在に移動させる。
+    - 学習、評価、生産に適したフレームワークをシームレスに選択できます。
+
+1. モデルやサンプルをニーズに合わせて簡単にカスタマイズ可能:
+    - 原著者が発表した結果を再現するために、各アーキテクチャの例を提供しています。
+    - モデル内部は可能な限り一貫して公開されています。
+    - モデルファイルはライブラリとは独立して利用することができ、迅速な実験が可能です。
+
+## なぜtransformersを使ってはいけないのでしょうか？
+
+- このライブラリは、ニューラルネットのためのビルディングブロックのモジュール式ツールボックスではありません。モデルファイルのコードは、研究者が追加の抽象化/ファイルに飛び込むことなく、各モデルを素早く反復できるように、意図的に追加の抽象化でリファクタリングされていません。
+- 学習APIはどのようなモデルでも動作するわけではなく、ライブラリが提供するモデルで動作するように最適化されています。一般的な機械学習のループには、別のライブラリ(おそらく[Accelerate](https://huggingface.co/docs/accelerate))を使用する必要があります。
+- 私たちはできるだけ多くの使用例を紹介するよう努力していますが、[examples フォルダ](https://github.com/huggingface/transformers/tree/main/examples) にあるスクリプトはあくまで例です。あなたの特定の問題に対してすぐに動作するわけではなく、あなたのニーズに合わせるために数行のコードを変更する必要があることが予想されます。
+
+## インストール
+
+### pipにて
+
+このリポジトリは、Python 3.6+, Flax 0.3.2+, PyTorch 1.3.1+, TensorFlow 2.3+ でテストされています。
+
+🤗Transformersは[仮想環境](https://docs.python.org/3/library/venv.html)にインストールする必要があります。Pythonの仮想環境に慣れていない場合は、[ユーザーガイド](https://packaging.python.org/guides/installing-using-pip-and-virtual-environments/)を確認してください。
+
+まず、使用するバージョンのPythonで仮想環境を作成し、アクティベートします。
+
+その後、Flax, PyTorch, TensorFlowのうち少なくとも1つをインストールする必要があります。
+[TensorFlowインストールページ](https://www.tensorflow.org/install/)、[PyTorchインストールページ](https://pytorch.org/get-started/locally/#start-locally)、[Flax](https://github.com/google/flax#quick-install)、[Jax](https://github.com/google/jax#installation)インストールページで、お使いのプラットフォーム別のインストールコマンドを参照してください。
+
+これらのバックエンドのいずれかがインストールされている場合、🤗Transformersは以下のようにpipを使用してインストールすることができます:
+
+```bash
+pip install transformers
+```
+
+もしサンプルを試したい、またはコードの最先端が必要で、新しいリリースを待てない場合は、[ライブラリをソースからインストール](https://huggingface.co/docs/transformers/installation#installing-from-source)する必要があります。
+
+### condaにて
+
+Transformersバージョン4.0.0から、condaチャンネルを搭載しました: `huggingface`。
+
+🤗Transformersは以下のようにcondaを使って設置することができます:
+
+```shell script
+conda install -c huggingface transformers
+```
+
+Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それぞれのインストールページに従ってください。
+
+> **_注意:_**  Windowsでは、キャッシュの恩恵を受けるために、デベロッパーモードを有効にするよう促されることがあります。このような場合は、[このissue](https://github.com/huggingface/huggingface_hub/issues/1062)でお知らせください。
+
+## モデルアーキテクチャ
+
+🤗Transformersが提供する **[全モデルチェックポイント](https://huggingface.co/models)** は、[ユーザー](https://huggingface.co/users)や[組織](https://huggingface.co/organizations)によって直接アップロードされるhuggingface.co [model hub](https://huggingface.co)からシームレスに統合されています。
+
+現在のチェックポイント数: ![](https://img.shields.io/endpoint?url=https://huggingface.co/api/shields/models&color=brightgreen)
+
+🤗Transformersは現在、以下のアーキテクチャを提供しています（それぞれのハイレベルな要約は[こちら](https://huggingface.co/docs/transformers/model_summary)を参照してください）:
+
+1. **[ALBERT](https://huggingface.co/docs/transformers/model_doc/albert)** (from Google Research and the Toyota Technological Institute at Chicago) released with the paper [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://arxiv.org/abs/1909.11942), by Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin Gimpel, Piyush Sharma, Radu Soricut.
+1. **[Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer)** (from MIT) released with the paper [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) by Yuan Gong, Yu-An Chung, James Glass.
+1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (from Facebook) released with the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) by Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer.
+1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (from École polytechnique) released with the paper [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) by Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis.
+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. **[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. **[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 [BigScience 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. **[Chinese-CLIP](https://huggingface.co/docs/transformers/model_doc/chinese_clip)** (from OFA-Sys) released with the paper [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) by An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou.
+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. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (from University of Göttingen) released with the paper [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) by Timo Lüddecke and Alexander Ecker.
+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. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
+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. **[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. **[Deformable DETR](https://huggingface.co/docs/transformers/model_doc/deformable_detr)** (from SenseTime Research) released with the paper [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://arxiv.org/abs/2010.04159) by Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, Jifeng Dai.
+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. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (from SHI Labs) released with the paper [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001) by Ali Hassani and Humphrey Shi.
+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. **[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. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (from NAVER), released together with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
+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. **[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. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (from Meta AI) are transformer protein language models.  **ESM-1b** was released with the paper [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus. **ESM-1v** was released with the paper [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648) by Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives. **ESM-2** was released with the paper [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
+1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
+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/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 NeoX Japanese](https://huggingface.co/docs/transformers/model_doc/gpt_neox_japanese)** (from ABEJA) released by Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori.
+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/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. **[Jukebox](https://huggingface.co/docs/transformers/model_doc/jukebox)** (from OpenAI) released with the paper [Jukebox: A Generative Model for Music](https://arxiv.org/pdf/2005.00341.pdf) by Prafulla Dhariwal, Heewoo Jun, Christine Payne, Jong Wook Kim, Alec Radford, 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/layoutxlm)** (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. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (from South China University of Technology) released with the paper [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) by Jiapeng Wang, Lianwen Jin, Kai Ding.
+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. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
+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. **[MobileNetV1](https://huggingface.co/docs/transformers/model_doc/mobilenet_v1)** (from Google Inc.) released with the paper [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://arxiv.org/abs/1704.04861) by Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam.
+1. **[MobileNetV2](https://huggingface.co/docs/transformers/model_doc/mobilenet_v2)** (from Google Inc.) released with the paper [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381) by Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen.
+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. **[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. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
+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. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, 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/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. **[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. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (from WeChatAI) released with the paper [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) by HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou.
+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.
+1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
+1. **[SEW-D](https://huggingface.co/docs/transformers/model_doc/sew_d)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
+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/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. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[SwitchTransformers](https://huggingface.co/docs/transformers/model_doc/switch_transformers)** (from Google) released with the paper [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://arxiv.org/abs/2101.03961) by William Fedus, Barret Zoph, Noam Shazeer.
+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. **[Table Transformer](https://huggingface.co/docs/transformers/model_doc/table-transformer)** (from Microsoft Research) released with the paper [PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents](https://arxiv.org/abs/2110.00061) by Brandon Smock, Rohith Pesala, Robin Abraham.
+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/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. **[Time Series Transformer](https://huggingface.co/docs/transformers/model_doc/time_series_transformer)**  (from HuggingFace).
+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/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. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
+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/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. **[ViTMSN](https://huggingface.co/docs/transformers/model_doc/vit_msn)** (from Meta AI) released with the paper [Masked Siamese Networks for Label-Efficient Learning](https://arxiv.org/abs/2204.07141) by Mahmoud Assran, Mathilde Caron, Ishan Misra, Piotr Bojanowski, Florian Bordes, Pascal Vincent, Armand Joulin, Michael Rabbat, Nicolas Ballas.
+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. **[Whisper](https://huggingface.co/docs/transformers/model_doc/whisper)** (from OpenAI) released with the paper [Robust Speech Recognition via Large-Scale Weak Supervision](https://cdn.openai.com/papers/whisper.pdf) by Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, Ilya Sutskever.
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
+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. **[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. **[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. 新しいモデルを投稿したいですか？新しいモデルを追加するためのガイドとして、**詳細なガイドとテンプレート**が追加されました。これらはリポジトリの[`templates`](./templates)フォルダにあります。PRを始める前に、必ず[コントリビューションガイド](./CONTRIBUTING.md)を確認し、メンテナに連絡するか、フィードバックを収集するためにissueを開いてください。
+
+各モデルがFlax、PyTorch、TensorFlowで実装されているか、🤗Tokenizersライブラリに支えられた関連トークナイザを持っているかは、[この表](https://huggingface.co/docs/transformers/index#supported-frameworks)を参照してください。
+
+これらの実装はいくつかのデータセットでテストされており(サンプルスクリプトを参照)、オリジナルの実装の性能と一致するはずである。性能の詳細は[documentation](https://github.com/huggingface/transformers/tree/main/examples)のExamplesセクションで見ることができます。
+
+
+## さらに詳しく
+
+| セクション | 概要 |
+|-|-|
+| [ドキュメント](https://huggingface.co/docs/transformers/) | 完全なAPIドキュメントとチュートリアル |
+| [タスク概要](https://huggingface.co/docs/transformers/task_summary) | 🤗Transformersがサポートするタスク |
+| [前処理チュートリアル](https://huggingface.co/docs/transformers/preprocessing) | モデル用のデータを準備するために`Tokenizer`クラスを使用 |
+| [トレーニングと微調整](https://huggingface.co/docs/transformers/training) | PyTorch/TensorFlowの学習ループと`Trainer`APIで🤗Transformersが提供するモデルを使用 |
+| [クイックツアー: 微調整/使用方法スクリプト](https://github.com/huggingface/transformers/tree/main/examples) | 様々なタスクでモデルの微調整を行うためのスクリプト例 |
+| [モデルの共有とアップロード](https://huggingface.co/docs/transformers/model_sharing) | 微調整したモデルをアップロードしてコミュニティで共有する |
+| [マイグレーション](https://huggingface.co/docs/transformers/migration) | `pytorch-transformers`または`pytorch-pretrained-bert`から🤗Transformers に移行する |
+
+## 引用
+
+🤗 トランスフォーマーライブラリに引用できる[論文](https://www.aclweb.org/anthology/2020.emnlp-demos.6/)が出来ました:
+```bibtex
+@inproceedings{wolf-etal-2020-transformers,
+    title = "Transformers: State-of-the-Art Natural Language Processing",
+    author = "Thomas Wolf and Lysandre Debut and Victor Sanh and Julien Chaumond and Clement Delangue and Anthony Moi and Pierric Cistac and Tim Rault and Rémi Louf and Morgan Funtowicz and Joe Davison and Sam Shleifer and Patrick von Platen and Clara Ma and Yacine Jernite and Julien Plu and Canwen Xu and Teven Le Scao and Sylvain Gugger and Mariama Drame and Quentin Lhoest and Alexander M. Rush",
+    booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
+    month = oct,
+    year = "2020",
+    address = "Online",
+    publisher = "Association for Computational Linguistics",
+    url = "https://www.aclweb.org/anthology/2020.emnlp-demos.6",
+    pages = "38--45"
+}
+```

README_ko.md

@@ -43,7 +43,9 @@ limitations under the License.
         <a href="https://github.com/huggingface/transformers/">English</a> |
         <a href="https://github.com/huggingface/transformers/blob/main/README_zh-hans.md">简体中文</a> |
         <a href="https://github.com/huggingface/transformers/blob/main/README_zh-hant.md">繁體中文</a> |
-        <b>한국어</b>
+        <b>한국어</b> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_es.md">Español</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_ja.md">日本語</a>
     <p>
 </h4>
 
@@ -59,7 +61,7 @@ limitations under the License.
 
 🤗 Transformers는 이러한 사전학습 모델을 빠르게 다운로드해 특정 텍스트에 사용하고, 원하는 데이터로 fine-tuning해 커뮤니티나 우리의 [모델 허브](https://huggingface.co/models)에 공유할 수 있도록 API를 제공합니다. 또한, 모델 구조를 정의하는 각 파이썬 모듈은 완전히 독립적이여서 연구 실험을 위해 손쉽게 수정할 수 있습니다.
 
-🤗 Transformers는 가장 유명한 3개의 딥러닝 라이브러리를 지원합니다. 이들은 서로 완벽히 연동됩니다 — [Jax](https://jax.readthedocs.io/en/latest/), [PyTorch](https://pytorch.org/), [TensorFlow](https://www.tensorflow.org/). 간단하게 이 라이브러리 중 하나로 모델을 학습하고, 또 다른 라이브러리로 추론을 위해 모델을 불러올 수 있습니다. 
+🤗 Transformers는 가장 유명한 3개의 딥러닝 라이브러리를 지원합니다. 이들은 서로 완벽히 연동됩니다 — [Jax](https://jax.readthedocs.io/en/latest/), [PyTorch](https://pytorch.org/), [TensorFlow](https://www.tensorflow.org/). 간단하게 이 라이브러리 중 하나로 모델을 학습하고, 또 다른 라이브러리로 추론을 위해 모델을 불러올 수 있습니다.
 
 ## 온라인 데모
 
@@ -74,7 +76,7 @@ limitations under the License.
 - [DistilBERT를 이용한 질문 답변](https://huggingface.co/distilbert-base-uncased-distilled-squad?text=Which+name+is+also+used+to+describe+the+Amazon+rainforest+in+English%3F&context=The+Amazon+rainforest+%28Portuguese%3A+Floresta+Amaz%C3%B4nica+or+Amaz%C3%B4nia%3B+Spanish%3A+Selva+Amaz%C3%B3nica%2C+Amazon%C3%ADa+or+usually+Amazonia%3B+French%3A+For%C3%AAt+amazonienne%3B+Dutch%3A+Amazoneregenwoud%29%2C+also+known+in+English+as+Amazonia+or+the+Amazon+Jungle%2C+is+a+moist+broadleaf+forest+that+covers+most+of+the+Amazon+basin+of+South+America.+This+basin+encompasses+7%2C000%2C000+square+kilometres+%282%2C700%2C000+sq+mi%29%2C+of+which+5%2C500%2C000+square+kilometres+%282%2C100%2C000+sq+mi%29+are+covered+by+the+rainforest.+This+region+includes+territory+belonging+to+nine+nations.+The+majority+of+the+forest+is+contained+within+Brazil%2C+with+60%25+of+the+rainforest%2C+followed+by+Peru+with+13%25%2C+Colombia+with+10%25%2C+and+with+minor+amounts+in+Venezuela%2C+Ecuador%2C+Bolivia%2C+Guyana%2C+Suriname+and+French+Guiana.+States+or+departments+in+four+nations+contain+%22Amazonas%22+in+their+names.+The+Amazon+represents+over+half+of+the+planet%27s+remaining+rainforests%2C+and+comprises+the+largest+and+most+biodiverse+tract+of+tropical+rainforest+in+the+world%2C+with+an+estimated+390+billion+individual+trees+divided+into+16%2C000+species)
 - [T5로 번역하기](https://huggingface.co/t5-base?text=My+name+is+Wolfgang+and+I+live+in+Berlin)
 
-**[Transformer와 글쓰기](https://transformer.huggingface.co)** 는 이 저장소의 텍스트 생성 능력에 관한 Hugging Face 팀의 공식 데모입니다. 
+**[Transformer와 글쓰기](https://transformer.huggingface.co)** 는 이 저장소의 텍스트 생성 능력에 관한 Hugging Face 팀의 공식 데모입니다.
 
 ## Hugging Face 팀의 커스텀 지원을 원한다면
 
@@ -210,6 +212,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 🤗 Transformers는 다음 모델들을 제공합니다 (각 모델의 요약은 [여기](https://huggingface.co/docs/transformers/model_summary)서 확인하세요):
 
 1. **[ALBERT](https://huggingface.co/docs/transformers/model_doc/albert)** (from Google Research and the Toyota Technological Institute at Chicago) released with the paper [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://arxiv.org/abs/1909.11942), by Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin Gimpel, Piyush Sharma, Radu Soricut.
+1. **[Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer)** (from MIT) released with the paper [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) by Yuan Gong, Yu-An Chung, James Glass.
 1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (from Facebook) released with the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/pdf/1910.13461.pdf) by Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer.
 1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (from École polytechnique) released with the paper [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) by Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis.
 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.
@@ -221,13 +224,16 @@ 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. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (from BigScience workshop) released by the [BigScience 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. **[Chinese-CLIP](https://huggingface.co/docs/transformers/model_doc/chinese_clip)** (from OFA-Sys) released with the paper [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) by An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou.
 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. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (from University of Göttingen) released with the paper [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) by Timo Lüddecke and Alexander Ecker.
 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. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 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.
@@ -237,17 +243,21 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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. **[Deformable DETR](https://huggingface.co/docs/transformers/model_doc/deformable_detr)** (from SenseTime Research) released with the paper [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://arxiv.org/abs/2010.04159) by Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, Jifeng Dai.
 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. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (from SHI Labs) released with the paper [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001) by Ali Hassani and Humphrey Shi.
 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/distillation), RoBERTa into [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/distillation), Multilingual BERT into [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/distillation) and a German version of DistilBERT.
 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. **[Donut](https://huggingface.co/docs/transformers/main/model_doc/donut)** (from NAVER) released with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
+1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (from NAVER) released with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
 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. **[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. **[ERNIE](https://huggingface.co/docs/transformers/main/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (from Meta AI) are transformer protein language models.  **ESM-1b** was released with the paper [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus. **ESM-1v** was released with the paper [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648) by Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives. **ESM-2** was released with the paper [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
+1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 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.
@@ -255,19 +265,22 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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 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 NeoX Japanese](https://huggingface.co/docs/transformers/model_doc/gpt_neox_japanese)** (from ABEJA) released by Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori.
 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/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. **[Jukebox](https://huggingface.co/docs/transformers/model_doc/jukebox)** (from OpenAI) released with the paper [Jukebox: A Generative Model for Music](https://arxiv.org/pdf/2005.00341.pdf) by Prafulla Dhariwal, Heewoo Jun, Christine Payne, Jong Wook Kim, Alec Radford, 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/layoutxlm)** (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. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (from South China University of Technology) released with the paper [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) by Jiapeng Wang, Lianwen Jin, Kai Ding.
 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.
@@ -275,6 +288,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 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.
@@ -282,17 +296,20 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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. **[MobileNetV1](https://huggingface.co/docs/transformers/model_doc/mobilenet_v1)** (from Google Inc.) released with the paper [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://arxiv.org/abs/1704.04861) by Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam.
+1. **[MobileNetV2](https://huggingface.co/docs/transformers/model_doc/mobilenet_v2)** (from Google Inc.) released with the paper [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381) by Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen.
 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. **[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. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 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. **[PEGASUS-X](https://huggingface.co/docs/transformers/main/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, Peter J. Liu. 
+1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, 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/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.
@@ -302,10 +319,11 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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 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/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. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (from WeChatAI) released with the paper [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) by HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou.
 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.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -315,29 +333,34 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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/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. **[Swin Transformer V2](https://huggingface.co/docs/transformers/main/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[SwitchTransformers](https://huggingface.co/docs/transformers/model_doc/switch_transformers)** (from Google) released with the paper [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://arxiv.org/abs/2101.03961) by William Fedus, Barret Zoph, Noam Shazeer.
 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. **[Table Transformer](https://huggingface.co/docs/transformers/model_doc/table-transformer)** (from Microsoft Research) released with the paper [PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents](https://arxiv.org/abs/2110.00061) by Brandon Smock, Rohith Pesala, Robin Abraham.
 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/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. **[Time Series Transformer](https://huggingface.co/docs/transformers/model_doc/time_series_transformer)** (from HuggingFace).
+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. **[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/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. **[VideoMAE](https://huggingface.co/docs/transformers/main/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
+1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
 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/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. **[ViTMSN](https://huggingface.co/docs/transformers/model_doc/vit_msn)** (from Meta AI) released with the paper [Masked Siamese Networks for Label-Efficient Learning](https://arxiv.org/abs/2204.07141) by Mahmoud Assran, Mathilde Caron, Ishan Misra, Piotr Bojanowski, Florian Bordes, Pascal Vincent, Armand Joulin, Michael Rabbat, Nicolas Ballas.
 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. **[X-CLIP](https://huggingface.co/docs/transformers/main/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
-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. **[Whisper](https://huggingface.co/docs/transformers/model_doc/whisper)** (from OpenAI) released with the paper [Robust Speech Recognition via Large-Scale Weak Supervision](https://cdn.openai.com/papers/whisper.pdf) by Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, Ilya Sutskever.
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
+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.
@@ -347,7 +370,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 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) 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. 
+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

@@ -68,7 +68,9 @@ checkpoint: 检查点
         <a href="https://github.com/huggingface/transformers/">English</a> |
         <b>简体中文</b> |
         <a href="https://github.com/huggingface/transformers/blob/main/README_zh-hant.md">繁體中文</a> |
-        <a href="https://github.com/huggingface/transformers/blob/main/README_ko.md">한국어</a>
+        <a href="https://github.com/huggingface/transformers/blob/main/README_ko.md">한국어</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_es.md">Español</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_ja.md">日本語</a>
     <p>
 </h4>
 
@@ -234,6 +236,7 @@ conda install -c huggingface transformers
 🤗 Transformers 目前支持如下的架构（模型概述请阅[这里](https://huggingface.co/docs/transformers/model_summary)）：
 
 1. **[ALBERT](https://huggingface.co/docs/transformers/model_doc/albert)** (来自 Google Research and the Toyota Technological Institute at Chicago) 伴随论文 [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://arxiv.org/abs/1909.11942), 由 Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin Gimpel, Piyush Sharma, Radu Soricut 发布。
+1. **[Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer)** (来自 MIT) 伴随论文 [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) 由 Yuan Gong, Yu-An Chung, James Glass 发布。
 1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (来自 Facebook) 伴随论文 [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/pdf/1910.13461.pdf) 由 Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer 发布。
 1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (来自 École polytechnique) 伴随论文 [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) 由 Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis 发布。
 1. **[BARTpho](https://huggingface.co/docs/transformers/model_doc/bartpho)** (来自 VinAI Research) 伴随论文 [BARTpho: Pre-trained Sequence-to-Sequence Models for Vietnamese](https://arxiv.org/abs/2109.09701) 由 Nguyen Luong Tran, Duong Minh Le and Dat Quoc Nguyen 发布。
@@ -245,13 +248,16 @@ 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. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (from BigScience workshop) released by the [BigScience 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. **[Chinese-CLIP](https://huggingface.co/docs/transformers/model_doc/chinese_clip)** (来自 OFA-Sys) 伴随论文 [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) 由 An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou 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. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (来自 University of Göttingen) 伴随论文 [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) 由 Timo Lüddecke and Alexander Ecker 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. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (来自 Microsoft Research Asia) 伴随论文 [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) 由 Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang 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/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。
@@ -261,17 +267,21 @@ conda install -c huggingface transformers
 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。
+1. **[Deformable DETR](https://huggingface.co/docs/transformers/model_doc/deformable_detr)** (来自 SenseTime Research) 伴随论文 [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://arxiv.org/abs/2010.04159) 由 Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, Jifeng Dai 发布。
 1. **[DeiT](https://huggingface.co/docs/transformers/model_doc/deit)** (来自 Facebook) 伴随论文 [Training data-efficient image transformers & distillation through attention](https://arxiv.org/abs/2012.12877) 由 Hugo Touvron, Matthieu Cord, Matthijs Douze, Francisco Massa, Alexandre Sablayrolles, Hervé Jégou 发布。
 1. **[DETR](https://huggingface.co/docs/transformers/model_doc/detr)** (来自 Facebook) 伴随论文 [End-to-End Object Detection with Transformers](https://arxiv.org/abs/2005.12872) 由 Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, Sergey Zagoruyko 发布。
 1. **[DialoGPT](https://huggingface.co/docs/transformers/model_doc/dialogpt)** (来自 Microsoft Research) 伴随论文 [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://arxiv.org/abs/1911.00536) 由 Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan 发布。
+1. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (来自 SHI Labs) 伴随论文 [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001) 由 Ali Hassani and Humphrey Shi 发布。
 1. **[DistilBERT](https://huggingface.co/docs/transformers/model_doc/distilbert)** (来自 HuggingFace), 伴随论文 [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://arxiv.org/abs/1910.01108) 由 Victor Sanh, Lysandre Debut and Thomas Wolf 发布。 同样的方法也应用于压缩 GPT-2 到 [DistilGPT2](https://github.com/huggingface/transformers/tree/main/examples/distillation), RoBERTa 到 [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/distillation), Multilingual BERT 到 [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/distillation) 和德语版 DistilBERT。
 1. **[DiT](https://huggingface.co/docs/transformers/model_doc/dit)** (来自 Microsoft Research) 伴随论文 [DiT: Self-supervised Pre-training for Document Image Transformer](https://arxiv.org/abs/2203.02378) 由 Junlong Li, Yiheng Xu, Tengchao Lv, Lei Cui, Cha Zhang, Furu Wei 发布。
-1. **[Donut](https://huggingface.co/docs/transformers/main/model_doc/donut)** (来自 NAVER) 伴随论文 [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) 由 Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park 发布。
+1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (来自 NAVER) 伴随论文 [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) 由 Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park 发布。
 1. **[DPR](https://huggingface.co/docs/transformers/model_doc/dpr)** (来自 Facebook) 伴随论文 [Dense Passage Retrieval for Open-Domain Question Answering](https://arxiv.org/abs/2004.04906) 由 Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih 发布。
 1. **[DPT](https://huggingface.co/docs/transformers/master/model_doc/dpt)** (来自 Intel Labs) 伴随论文 [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) 由 René Ranftl, Alexey Bochkovskiy, Vladlen Koltun 发布。
 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. **[ERNIE](https://huggingface.co/docs/transformers/main/model_doc/ernie)** (来自 Baidu) 伴随论文 [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu 发布。
+1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (来自 Baidu) 伴随论文 [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu 发布。
+1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (from Meta AI) are transformer protein language models.  **ESM-1b** was released with the paper [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus. **ESM-1v** was released with the paper [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648) by Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives. **ESM-2** was released with the paper [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
+1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 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。
@@ -279,19 +289,22 @@ conda install -c huggingface transformers
 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 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 NeoX Japanese](https://huggingface.co/docs/transformers/model_doc/gpt_neox_japanese)** (来自 ABEJA) 由 Shinya Otani, Takayoshi Makabe, Anuj Arora, Kyo Hattori。
 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/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. **[Jukebox](https://huggingface.co/docs/transformers/model_doc/jukebox)** (from OpenAI) released with the paper [Jukebox: A Generative Model for Music](https://arxiv.org/pdf/2005.00341.pdf) by Prafulla Dhariwal, Heewoo Jun, Christine Payne, Jong Wook Kim, Alec Radford, Ilya Sutskever.
 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/layoutxlm)** (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. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (来自 South China University of Technology) 伴随论文 [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) 由 Jiapeng Wang, Lianwen Jin, Kai Ding 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。
@@ -299,24 +312,28 @@ conda install -c huggingface transformers
 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/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. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (来自 Microsoft Research Asia) 伴随论文 [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) 由 Junlong Li, Yiheng Xu, Lei Cui, Furu Wei 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  >>>>>>> Fix rebase
 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. **[MobileNetV1](https://huggingface.co/docs/transformers/model_doc/mobilenet_v1)** (来自 Google Inc.) 伴随论文 [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://arxiv.org/abs/1704.04861) 由 Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam 发布。
+1. **[MobileNetV2](https://huggingface.co/docs/transformers/model_doc/mobilenet_v2)** (来自 Google Inc.) 伴随论文 [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381) 由 Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen 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. **[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. **[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. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (来自 SHI Labs) 伴随论文 [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) 由 Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi 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. **[PEGASUS-X](https://huggingface.co/docs/transformers/main/model_doc/pegasus_x)** (来自 Google) 伴随论文 [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) 由 Jason Phang, Yao Zhao, Peter J. Liu 发布。  
+1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (来自 Google) 伴随论文 [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) 由 Jason Phang, Yao Zhao, 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/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。
@@ -326,10 +343,11 @@ conda install -c huggingface transformers
 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/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/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. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (来自 WeChatAI), 伴随论文 [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) 由 HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou 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。
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (来自 ASAPP) 伴随论文 [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) 由 Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi 发布。
@@ -339,29 +357,34 @@ conda install -c huggingface transformers
 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/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. **[Swin Transformer V2](https://huggingface.co/docs/transformers/main/model_doc/swinv2)** (来自 Microsoft) 伴随论文 [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) 由 Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo 发布。
+1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (来自 Microsoft) 伴随论文 [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) 由 Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo 发布。
+1. **[SwitchTransformers](https://huggingface.co/docs/transformers/model_doc/switch_transformers)** (from Google) released with the paper [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://arxiv.org/abs/2101.03961) by William Fedus, Barret Zoph, Noam Shazeer.
 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. **[Table Transformer](https://huggingface.co/docs/transformers/model_doc/table-transformer)** (来自 Microsoft Research) 伴随论文 [PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents](https://arxiv.org/abs/2110.00061) 由 Brandon Smock, Rohith Pesala, Robin Abraham 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/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. **[Time Series Transformer](https://huggingface.co/docs/transformers/model_doc/time_series_transformer)** (from HuggingFace).
+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. **[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/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. **[VideoMAE](https://huggingface.co/docs/transformers/main/model_doc/videomae)** (来自 Multimedia Computing Group, Nanjing University) 伴随论文 [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) 由 Zhan Tong, Yibing Song, Jue Wang, Limin Wang 发布。
+1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (来自 Multimedia Computing Group, Nanjing University) 伴随论文 [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) 由 Zhan Tong, Yibing Song, Jue Wang, Limin Wang 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/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. **[ViTMSN](https://huggingface.co/docs/transformers/model_doc/vit_msn)** (来自 Meta AI) 伴随论文 [Masked Siamese Networks for Label-Efficient Learning](https://arxiv.org/abs/2204.07141) by Mahmoud Assran, Mathilde Caron, Ishan Misra, Piotr Bojanowski, Florian Bordes, Pascal Vincent, Armand Joulin, Michael Rabbat, Nicolas Ballas 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/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. **[X-CLIP](https://huggingface.co/docs/transformers/main/model_doc/xclip)** (来自 Microsoft Research) 伴随论文 [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) 由 Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling 发布。
-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. **[Whisper](https://huggingface.co/docs/transformers/model_doc/whisper)** (来自 OpenAI) 伴随论文 [Robust Speech Recognition via Large-Scale Weak Supervision](https://cdn.openai.com/papers/whisper.pdf) 由 Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, Ilya Sutskever 发布。
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (来自 Microsoft Research) 伴随论文 [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) 由 Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling 发布。
+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。
 1. **[XLM-RoBERTa](https://huggingface.co/docs/transformers/model_doc/xlm-roberta)** (来自 Facebook AI), 伴随论文 [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) 由 Alexis Conneau*, Kartikay Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer and Veselin Stoyanov 发布。

README_zh-hant.md

@@ -80,7 +80,9 @@ user: 使用者
         <a href="https://github.com/huggingface/transformers/">English</a> |
         <a href="https://github.com/huggingface/transformers/blob/main/README_zh-hans.md">简体中文</a> |
         <b>繁體中文</b> |
-        <a href="https://github.com/huggingface/transformers/blob/main/README_ko.md">한국어</a>
+        <a href="https://github.com/huggingface/transformers/blob/main/README_ko.md">한국어</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_es.md">Español</a> |
+        <a href="https://github.com/huggingface/transformers/blob/main/README_ja.md">日本語</a>
     <p>
 </h4>
 
@@ -246,6 +248,7 @@ conda install -c huggingface transformers
 🤗 Transformers 目前支援以下的架構（模型概覽請參閱[這裡](https://huggingface.co/docs/transformers/model_summary)）：
 
 1. **[ALBERT](https://huggingface.co/docs/transformers/model_doc/albert)** (from Google Research and the Toyota Technological Institute at Chicago) released with the paper [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://arxiv.org/abs/1909.11942), by Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin Gimpel, Piyush Sharma, Radu Soricut.
+1. **[Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer)** (from MIT) released with the paper [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) by Yuan Gong, Yu-An Chung, James Glass.
 1. **[BART](https://huggingface.co/docs/transformers/model_doc/bart)** (from Facebook) released with the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/pdf/1910.13461.pdf) by Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer.
 1. **[BARThez](https://huggingface.co/docs/transformers/model_doc/barthez)** (from École polytechnique) released with the paper [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) by Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis.
 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.
@@ -257,13 +260,16 @@ 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. **[BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)** (from BigScience workshop) released by the [BigScience 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. **[Chinese-CLIP](https://huggingface.co/docs/transformers/model_doc/chinese_clip)** (from OFA-Sys) released with the paper [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) by An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou.
 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. **[CLIPSeg](https://huggingface.co/docs/transformers/model_doc/clipseg)** (from University of Göttingen) released with the paper [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) by Timo Lüddecke and Alexander Ecker.
 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. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 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.
@@ -273,17 +279,21 @@ conda install -c huggingface transformers
 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. **[Deformable DETR](https://huggingface.co/docs/transformers/model_doc/deformable_detr)** (from SenseTime Research) released with the paper [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://arxiv.org/abs/2010.04159) by Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, Jifeng Dai.
 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. **[DiNAT](https://huggingface.co/docs/transformers/model_doc/dinat)** (from SHI Labs) released with the paper [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001) by Ali Hassani and Humphrey Shi.
 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/distillation), RoBERTa into [DistilRoBERTa](https://github.com/huggingface/transformers/tree/main/examples/distillation), Multilingual BERT into [DistilmBERT](https://github.com/huggingface/transformers/tree/main/examples/distillation) and a German version of DistilBERT.
 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. **[Donut](https://huggingface.co/docs/transformers/main/model_doc/donut)** (from NAVER) released with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
+1. **[Donut](https://huggingface.co/docs/transformers/model_doc/donut)** (from NAVER) released with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
 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. **[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. **[ERNIE](https://huggingface.co/docs/transformers/main/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ERNIE](https://huggingface.co/docs/transformers/model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ESM](https://huggingface.co/docs/transformers/model_doc/esm)** (from Meta AI) are transformer protein language models.  **ESM-1b** was released with the paper [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus. **ESM-1v** was released with the paper [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648) by Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives. **ESM-2** was released with the paper [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
+1. **[FLAN-T5](https://huggingface.co/docs/transformers/model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 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.
@@ -291,19 +301,22 @@ conda install -c huggingface transformers
 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 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 NeoX Japanese](https://huggingface.co/docs/transformers/model_doc/gpt_neox_japanese)** (from ABEJA) released by Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori.
 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/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. **[Jukebox](https://huggingface.co/docs/transformers/model_doc/jukebox)** (from OpenAI) released with the paper [Jukebox: A Generative Model for Music](https://arxiv.org/pdf/2005.00341.pdf) by Prafulla Dhariwal, Heewoo Jun, Christine Payne, Jong Wook Kim, Alec Radford, 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/layoutxlm)** (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. **[LiLT](https://huggingface.co/docs/transformers/model_doc/lilt)** (from South China University of Technology) released with the paper [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) by Jiapeng Wang, Lianwen Jin, Kai Ding.
 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.
@@ -311,24 +324,28 @@ conda install -c huggingface transformers
 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/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. **[MarkupLM](https://huggingface.co/docs/transformers/model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
+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. **[MobileNetV1](https://huggingface.co/docs/transformers/model_doc/mobilenet_v1)** (from Google Inc.) released with the paper [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://arxiv.org/abs/1704.04861) by Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam.
+1. **[MobileNetV2](https://huggingface.co/docs/transformers/model_doc/mobilenet_v2)** (from Google Inc.) released with the paper [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381) by Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen.
 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. **[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. **[NAT](https://huggingface.co/docs/transformers/model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 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. **[PEGASUS-X](https://huggingface.co/docs/transformers/main/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, Peter J. Liu.
+1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, 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/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.
@@ -338,10 +355,11 @@ conda install -c huggingface transformers
 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 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/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. **[RoCBert](https://huggingface.co/docs/transformers/model_doc/roc_bert)** (from WeChatAI) released with the paper [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) by HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou.
 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.
 1. **[SEW](https://huggingface.co/docs/transformers/model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -351,29 +369,34 @@ conda install -c huggingface transformers
 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/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. **[Swin Transformer V2](https://huggingface.co/docs/transformers/main/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[Swin Transformer V2](https://huggingface.co/docs/transformers/model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[SwitchTransformers](https://huggingface.co/docs/transformers/model_doc/switch_transformers)** (from Google) released with the paper [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://arxiv.org/abs/2101.03961) by William Fedus, Barret Zoph, Noam Shazeer.
 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. **[Table Transformer](https://huggingface.co/docs/transformers/model_doc/table-transformer)** (from Microsoft Research) released with the paper [PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents](https://arxiv.org/abs/2110.00061) by Brandon Smock, Rohith Pesala, Robin Abraham.
 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/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. **[Time Series Transformer](https://huggingface.co/docs/transformers/model_doc/time_series_transformer)** (from HuggingFace).
+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. **[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/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. **[VideoMAE](https://huggingface.co/docs/transformers/main/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
+1. **[VideoMAE](https://huggingface.co/docs/transformers/model_doc/videomae)** (from Multimedia Computing Group, Nanjing University) released with the paper [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://arxiv.org/abs/2203.12602) by Zhan Tong, Yibing Song, Jue Wang, Limin Wang.
 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. **[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/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. **[ViTMSN](https://huggingface.co/docs/transformers/model_doc/vit_msn)** (from Meta AI) released with the paper [Masked Siamese Networks for Label-Efficient Learning](https://arxiv.org/abs/2204.07141) by Mahmoud Assran, Mathilde Caron, Ishan Misra, Piotr Bojanowski, Florian Bordes, Pascal Vincent, Armand Joulin, Michael Rabbat, Nicolas Ballas.
 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. **[X-CLIP](https://huggingface.co/docs/transformers/main/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
-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. **[Whisper](https://huggingface.co/docs/transformers/model_doc/whisper)** (from OpenAI) released with the paper [Robust Speech Recognition via Large-Scale Weak Supervision](https://cdn.openai.com/papers/whisper.pdf) by Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, Ilya Sutskever.
+1. **[X-CLIP](https://huggingface.co/docs/transformers/model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
+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.

conftest.py

@@ -32,7 +32,6 @@ warnings.simplefilter(action="ignore", category=FutureWarning)
 
 
 def pytest_configure(config):
-    config.addinivalue_line("markers", "is_pipeline_test: mark test to run only when pipeline are tested")
     config.addinivalue_line(
         "markers", "is_pt_tf_cross_test: mark test to run only when PT and TF interactions are tested"
     )

docker/transformers-all-latest-gpu/Dockerfile

@@ -32,15 +32,18 @@ RUN echo torch=$VERSION
 # TODO: We might need to specify proper versions that work with a specific torch version (especially for past CI).
 RUN [ "$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 install --no-cache-dir -U tensorflow==2.10.1
+RUN python3 -m pip install --no-cache-dir -U tensorflow_probability
 RUN python3 -m pip uninstall -y flax jax
 
-# 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
+# To include the change in this commit https://github.com/onnx/tensorflow-onnx/commit/ddca3a5eb2d912f20fe7e0568dd1a3013aee9fa3
+# Otherwise, we get tf2onnx==1.8 (caused by `flatbuffers` version),  and some tests fail with `ValueError: from_keras requires input_signature`.
+# TODO: remove this line once the conflict is resolved in these libraries.
+RUN python3 -m pip install --no-cache-dir git+https://github.com/onnx/tensorflow-onnx.git@ddca3a5eb2d912f20fe7e0568dd1a3013aee9fa3
+
 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 --no-cache-dir git+https://github.com/facebookresearch/detectron2.git pytesseract
 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
@@ -50,6 +53,9 @@ RUN python3 -m pip install --no-cache-dir bitsandbytes
 
 RUN python3 -m pip install --no-cache-dir decord
 
+# For `dinat` model
+RUN python3 -m pip install --no-cache-dir natten
+
 # 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

@@ -10,8 +10,7 @@ RUN apt-get -y update && apt-get install -y libsndfile1-dev && apt install -y te
 # Torch needs to be installed before deepspeed
 RUN python3 -m pip install --no-cache-dir ./transformers[deepspeed]
 
-RUN python3 -m pip install --no-cache-dir torch-scatter -f https://data.pyg.org/whl/torch-$(python -c "from torch import version; print(version.__version__.split('+')[0])")+cpu.html
-RUN python3 -m pip install --no-cache-dir torchvision git+https://github.com/facebookresearch/detectron2.git pytesseract https://github.com/kpu/kenlm/archive/master.zip
+RUN python3 -m pip install --no-cache-dir torchvision git+https://github.com/facebookresearch/detectron2.git pytesseract
 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"
 

docker/transformers-past-gpu/Dockerfile

@@ -34,10 +34,4 @@ RUN python3 ./transformers/utils/past_ci_versions.py --framework $FRAMEWORK --ve
 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-nightly-gpu/Dockerfile

@@ -27,6 +27,24 @@ RUN python3 -m pip uninstall -y deepspeed
 # RUN git clone https://github.com/microsoft/DeepSpeed && cd DeepSpeed && rm -rf build && \
 #    DS_BUILD_CPU_ADAM=1 DS_BUILD_FUSED_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
 
+# For `torchdynamo` tests
+# (see https://github.com/huggingface/transformers/pull/17765)
+RUN git clone https://github.com/pytorch/functorch
+RUN python3 -m pip install --no-cache-dir ./functorch[aot]
+RUN cd functorch && python3 setup.py develop
+
+RUN git clone https://github.com/pytorch/torchdynamo
+RUN python3 -m pip install -r ./torchdynamo/requirements.txt
+RUN cd torchdynamo && python3 setup.py develop
+
+# install TensorRT
+RUN python3 -m pip install --no-cache-dir -U nvidia-pyindex
+RUN python3 -m pip install --no-cache-dir -U nvidia-tensorrt==8.2.4.2
+
+# install torch_tensorrt (fx path)
+RUN git clone https://github.com/pytorch/TensorRT.git
+RUN cd TensorRT/py && python3 setup.py install --fx-only
+
 # 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-pytorch-gpu/Dockerfile

@@ -23,7 +23,7 @@ RUN [ ${#TORCH_AUDIO} -gt 0 ] && VERSION='torchaudio=='TORCH_AUDIO'.*' ||  VERSI
 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])")+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 --no-cache-dir git+https://github.com/facebookresearch/detectron2.git pytesseract
 RUN python3 -m pip install -U "itsdangerous<2.1.0"
 
 # When installing in editable mode, `transformers` is not recognized as a package.

docker/transformers-tensorflow-gpu/Dockerfile

@@ -12,7 +12,7 @@ RUN git clone https://github.com/huggingface/transformers && cd transformers &&
 RUN python3 -m pip install --no-cache-dir -e ./transformers[dev-tensorflow,testing]
 
 # If set to nothing, will install the latest version
-ARG TENSORFLOW=''
+ARG TENSORFLOW='2.10'
 
 RUN [ ${#TENSORFLOW} -gt 0 ] && VERSION='tensorflow=='$TENSORFLOW'.*' ||  VERSION='tensorflow'; python3 -m pip install --no-cache-dir -U $VERSION
 RUN python3 -m pip uninstall -y torch flax

docs/README.md

@@ -16,7 +16,7 @@ limitations under the License.
 
 # Generating the documentation
 
-To generate the documentation, you first have to build it. Several packages are necessary to build the doc,
+To generate the documentation, you first have to build it. Several packages are necessary to build the doc, 
 you can install them with the following command, at the root of the code repository:
 
 ```bash
@@ -33,7 +33,7 @@ pip install git+https://github.com/huggingface/doc-builder
 **NOTE**
 
 You only need to generate the documentation to inspect it locally (if you're planning changes and want to
-check how they look like before committing for instance). You don't have to commit the built documentation.
+check how they look before committing for instance). You don't have to commit the built documentation.
 
 ---
 
@@ -88,9 +88,9 @@ the filename without the extension in the [`_toctree.yml`](https://github.com/hu
 
 ## Renaming section headers and moving sections
 
-It helps to keep the old links working when renaming section header and/or moving sections from one document to another. This is because the old links are likely to be used in Issues, Forums and Social media and it'd be make for a much more superior user experience if users reading those months later could still easily navigate to the originally intended information.
+It helps to keep the old links working when renaming the section header and/or moving sections from one document to another. This is because the old links are likely to be used in Issues, Forums, and Social media and it'd make for a much more superior user experience if users reading those months later could still easily navigate to the originally intended information.
 
-Therefore we simply keep a little map of moved sections at the end of the document where the original section was. The key is to preserve the original anchor.
+Therefore, we simply keep a little map of moved sections at the end of the document where the original section was. The key is to preserve the original anchor.
 
 So if you renamed a section from: "Section A" to "Section B", then you can add at the end of the file:
 
@@ -99,7 +99,7 @@ Sections that were moved:
 
 [ <a href="#section-b">Section A</a><a id="section-a"></a> ]
 ```
-and of course if you moved it to another file, then:
+and of course, if you moved it to another file, then:
 
 ```
 Sections that were moved:
@@ -109,7 +109,7 @@ Sections that were moved:
 
 Use the relative style to link to the new file so that the versioned docs continue to work.
 
-For an example of a rich moved sections set please see the very end of [the Trainer doc](https://github.com/huggingface/transformers/blob/main/docs/source/main_classes/trainer.mdx).
+For an example of a rich moved section set please see the very end of [the Trainer doc](https://github.com/huggingface/transformers/blob/main/docs/source/en/main_classes/trainer.mdx).
 
 
 ## Writing Documentation - Specification
@@ -126,7 +126,7 @@ Adding a new tutorial or section is done in two steps:
 - Link that file in `./source/_toctree.yml` on the correct toc-tree.
 
 Make sure to put your new file under the proper section. It's unlikely to go in the first section (*Get Started*), so
-depending on the intended targets (beginners, more advanced users or researchers) it should go in section two, three or
+depending on the intended targets (beginners, more advanced users, or researchers) it should go in sections two, three, or
 four.
 
 ### Translating
@@ -177,8 +177,8 @@ not to be displayed in the documentation, you can do so by specifying which meth
     - save_vocabulary
 ```
 
-If you just want to add a method that is not documented (for instance magic method like `__call__` are not documented
-byt default) you can put the list of methods to add in a list that contains `all`:
+If you just want to add a method that is not documented (for instance magic methods like `__call__` are not documented
+by default) you can put the list of methods to add in a list that contains `all`:
 
 ```
 ## XXXTokenizer
@@ -191,9 +191,9 @@ byt default) you can put the list of methods to add in a list that contains `all
 ### Writing source documentation
 
 Values that should be put in `code` should either be surrounded by backticks: \`like so\`. Note that argument names
-and objects like True, None or any strings should usually be put in `code`.
+and objects like True, None, or any strings should usually be put in `code`.
 
-When mentioning a class, function or method, it is recommended to use our syntax for internal links so that our tool
+When mentioning a class, function, or method, it is recommended to use our syntax for internal links so that our tool
 adds a link to its documentation with this syntax: \[\`XXXClass\`\] or \[\`function\`\]. This requires the class or 
 function to be in the main package.
 
@@ -207,7 +207,7 @@ The same works for methods so you can either use \[\`XXXClass.method\`\] or \[~\
 #### Defining arguments in a method
 
 Arguments should be defined with the `Args:` (or `Arguments:` or `Parameters:`) prefix, followed by a line return and
-an indentation. The argument should be followed by its type, with its shape if it is a tensor, a colon and its
+an indentation. The argument should be followed by its type, with its shape if it is a tensor, a colon, and its
 description:
 
 ```
@@ -216,7 +216,7 @@ description:
 ```
 
 If the description is too long to fit in one line, another indentation is necessary before writing the description
-after th argument.
+after the argument.
 
 Here's an example showcasing everything so far:
 
@@ -266,7 +266,7 @@ Multi-line code blocks can be useful for displaying examples. They are done betw
 ````
 
 We follow the [doctest](https://docs.python.org/3/library/doctest.html) syntax for the examples to automatically test
-the results stay consistent with the library.
+the results to stay consistent with the library.
 
 #### Writing a return block
 
@@ -274,27 +274,27 @@ The return block should be introduced with the `Returns:` prefix, followed by a
 The first line should be the type of the return, followed by a line return. No need to indent further for the elements
 building the return.
 
-Here's an example for a single value return:
+Here's an example of a single value return:
 
 ```
     Returns:
         `List[int]`: A list of integers in the range [0, 1] --- 1 for a special token, 0 for a sequence token.
 ```
 
-Here's an example for tuple return, comprising several objects:
+Here's an example of a tuple return, comprising several objects:
 
 ```
     Returns:
         `tuple(torch.FloatTensor)` comprising various elements depending on the configuration ([`BertConfig`]) and inputs:
         - ** loss** (*optional*, returned when `masked_lm_labels` is provided) `torch.FloatTensor` of shape `(1,)` --
-          Total loss as the sum of the masked language modeling loss and the next sequence prediction (classification) loss.
+          Total loss is the sum of the masked language modeling loss and the next sequence prediction (classification) loss.
         - **prediction_scores** (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`) --
           Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
 ```
 
 #### Adding an image
 
-Due to the rapidly growing repository, it is important to make sure that no files that would significantly weigh down the repository are added. This includes images, videos and other non-text files. We prefer to leverage a hf.co hosted `dataset` like
+Due to the rapidly growing repository, it is important to make sure that no files that would significantly weigh down the repository are added. This includes images, videos, and other non-text files. We prefer to leverage a hf.co hosted `dataset` like
 the ones hosted on [`hf-internal-testing`](https://huggingface.co/hf-internal-testing) in which to place these files and reference
 them by URL. We recommend putting them in the following dataset: [huggingface/documentation-images](https://huggingface.co/datasets/huggingface/documentation-images).
 If an external contribution, feel free to add the images to your PR and ask a Hugging Face member to migrate your images
@@ -312,13 +312,13 @@ easily.
 
 # Testing documentation examples
 
-Good documentation oftens comes with an example of how a specific function or class should be used. 
+Good documentation often comes with an example of how a specific function or class should be used. 
 Each model class should contain at least one example showcasing
 how to use this model class in inference. *E.g.* the class [Wav2Vec2ForCTC](https://huggingface.co/docs/transformers/model_doc/wav2vec2#transformers.Wav2Vec2ForCTC) 
 includes an example of how to transcribe speech to text in the 
 [docstring of its forward function](https://huggingface.co/docs/transformers/model_doc/wav2vec2#transformers.Wav2Vec2ForCTC.forward).
 
-## Writing documenation examples
+## Writing documentation examples
 
 The syntax for Example docstrings can look as follows:
 
@@ -354,7 +354,7 @@ The docstring should give a minimal, clear example of how the respective model
 is to be used in inference and also include the expected (ideally sensible)
 output.
 Often, readers will try out the example before even going through the function 
-or class definitions. Therefore it is of utmost importance that the example 
+or class definitions. Therefore, it is of utmost importance that the example 
 works as expected.
 
 ## Docstring testing
@@ -364,7 +364,7 @@ We use pytests' [doctest integration](https://docs.pytest.org/doctest.html) to v
 For Transformers, the doctests are run on a daily basis via GitHub Actions as can be 
 seen [here](https://github.com/huggingface/transformers/actions/workflows/doctests.yml).
 
-To include your example in the daily doctests, you need add the filename that
+To include your example in the daily doctests, you need to add the filename that
 contains the example docstring to the [documentation_tests.txt](../utils/documentation_tests.txt).
 
 ### For Python files
@@ -426,6 +426,6 @@ Here are a few tips to help you debug the doctests and make them pass:
 
 - The outputs of the code need to match the expected output **exactly**, so make sure you have the same outputs. In particular doctest will see a difference between single quotes and double quotes, or a missing parenthesis. The only exceptions to that rule are:
   * 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).
+  * 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 configured 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 producing it.

docs/source/de/_toctree.yml

@@ -9,4 +9,14 @@
 - sections:
   - local: pipeline_tutorial
     title: Pipelines für Inferenzen
+  - local: autoclass_tutorial
+    title: Laden von vortrainierten Instanzen mit einer AutoClass
+  - local: preprocessing
+    title: Vorverarbeiten
+  - local: training
+    title: Optimierung eines vortrainierten Modells
+  - local: accelerate
+    title: Verteiltes Training mit 🤗 Accelerate
+  - local: model_sharing
+    title: Ein Modell teilen
   title: Tutorials

docs/source/de/accelerate.mdx

@@ -0,0 +1,132 @@
+<!--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.
+-->
+
+# Verteiltes Training mit 🤗 Accelerate
+
+Da die Modelle immer größer werden, hat sich die Parallelität als Strategie zum Trainieren größerer Modelle auf begrenzter Hardware und zur Beschleunigung der Trainingsgeschwindigkeit um mehrere Größenordnungen erwiesen. Bei Hugging Face haben wir die Bibliothek [🤗 Accelerate](https://huggingface.co/docs/accelerate) entwickelt, um Nutzern zu helfen, ein 🤗 Transformers-Modell auf jeder Art von verteiltem Setup zu trainieren, egal ob es sich um mehrere GPUs auf einer Maschine oder mehrere GPUs auf mehreren Maschinen handelt. In diesem Tutorial lernen Sie, wie Sie Ihre native PyTorch-Trainingsschleife anpassen, um das Training in einer verteilten Umgebung zu ermöglichen.
+
+## Einrichtung
+
+Beginnen Sie mit der Installation von 🤗 Accelerate:
+
+```bash
+pip install accelerate
+```
+
+Dann importieren und erstellen Sie ein [`~accelerate.Accelerator`]-Objekt. Der [`~accelerate.Accelerator`] wird automatisch Ihre Art der verteilten Einrichtung erkennen und alle notwendigen Komponenten für das Training initialisieren. Sie müssen Ihr Modell nicht explizit auf einem Gerät platzieren.
+
+```py
+>>> from accelerate import Accelerator
+
+>>> accelerator = Accelerator()
+```
+
+## Vorbereiten auf die Beschleunigung
+
+Der nächste Schritt ist die Übergabe aller relevanten Trainingsobjekte an die Methode [`~accelerate.Accelerator.prepare`]. Dazu gehören Ihre Trainings- und Evaluierungs-DataLoader, ein Modell und ein Optimierer:
+
+```py
+>>> train_dataloader, eval_dataloader, model, optimizer = accelerator.prepare(
+...     train_dataloader, eval_dataloader, model, optimizer
+... )
+```
+
+## Rückwärts
+
+Die letzte Ergänzung besteht darin, das typische `loss.backward()` in der Trainingsschleife durch die 🤗 Accelerate-Methode [`~accelerate.Accelerator.backward`] zu ersetzen:
+
+```py
+>>> for epoch in range(num_epochs):
+...     for batch in train_dataloader:
+...         outputs = model(**batch)
+...         loss = outputs.loss
+...         accelerator.backward(loss)
+
+...         optimizer.step()
+...         lr_scheduler.step()
+...         optimizer.zero_grad()
+...         progress_bar.update(1)
+```
+
+Wie Sie im folgenden Code sehen können, müssen Sie nur vier zusätzliche Codezeilen zu Ihrer Trainingsschleife hinzufügen, um verteiltes Training zu ermöglichen!
+
+```diff
++ from accelerate import Accelerator
+  from transformers import AdamW, AutoModelForSequenceClassification, get_scheduler
+
++ accelerator = Accelerator()
+
+  model = AutoModelForSequenceClassification.from_pretrained(checkpoint, num_labels=2)
+  optimizer = AdamW(model.parameters(), lr=3e-5)
+
+- device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
+- model.to(device)
+
++ train_dataloader, eval_dataloader, model, optimizer = accelerator.prepare(
++     train_dataloader, eval_dataloader, model, optimizer
++ )
+
+  num_epochs = 3
+  num_training_steps = num_epochs * len(train_dataloader)
+  lr_scheduler = get_scheduler(
+      "linear",
+      optimizer=optimizer,
+      num_warmup_steps=0,
+      num_training_steps=num_training_steps
+  )
+
+  progress_bar = tqdm(range(num_training_steps))
+
+  model.train()
+  for epoch in range(num_epochs):
+      for batch in train_dataloader:
+-         batch = {k: v.to(device) for k, v in batch.items()}
+          outputs = model(**batch)
+          loss = outputs.loss
+-         loss.backward()
++         accelerator.backward(loss)
+
+          optimizer.step()
+          lr_scheduler.step()
+          optimizer.zero_grad()
+          progress_bar.update(1)
+```
+
+## Trainieren
+
+Sobald Sie die entsprechenden Codezeilen hinzugefügt haben, starten Sie Ihr Training in einem Skript oder einem Notebook wie Colaboratory.
+
+### Trainieren mit einem Skript
+
+Wenn Sie Ihr Training mit einem Skript durchführen, führen Sie den folgenden Befehl aus, um eine Konfigurationsdatei zu erstellen und zu speichern:
+
+```bash
+accelerate config
+```
+
+Dann starten Sie Ihr Training mit:
+
+```bash
+accelerate launch train.py
+```
+
+### Trainieren mit einem Notebook
+
+🤗 Accelerate kann auch in einem Notebook laufen, wenn Sie planen, die TPUs von Colaboratory zu verwenden. Verpacken Sie den gesamten Code, der für das Training verantwortlich ist, in eine Funktion und übergeben Sie diese an [`~accelerate.notebook_launcher`]:
+
+```py
+>>> from accelerate import notebook_launcher
+
+>>> notebook_launcher(training_function)
+```
+
+Weitere Informationen über 🤗 Accelerate und seine umfangreichen Funktionen finden Sie in der [Dokumentation](https://huggingface.co/docs/accelerate).

docs/source/de/autoclass_tutorial.mdx

@@ -0,0 +1,127 @@
+<!--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.
+-->
+
+# Vortrainierte Instanzen mit einer AutoClass laden
+
+Bei so vielen verschiedenen Transformator-Architekturen kann es eine Herausforderung sein, eine für Ihren Checkpoint zu erstellen. Als Teil der 🤗 Transformers Kernphilosophie, die Bibliothek leicht, einfach und flexibel nutzbar zu machen, leitet eine `AutoClass` automatisch die richtige Architektur aus einem gegebenen Checkpoint ab und lädt sie. Mit der Methode `from_pretrained()` kann man schnell ein vortrainiertes Modell für eine beliebige Architektur laden, so dass man keine Zeit und Ressourcen aufwenden muss, um ein Modell von Grund auf zu trainieren. Die Erstellung dieser Art von Checkpoint-agnostischem Code bedeutet, dass Ihr Code, wenn er für einen Checkpoint funktioniert, auch mit einem anderen Checkpoint funktionieren wird - solange er für eine ähnliche Aufgabe trainiert wurde - selbst wenn die Architektur unterschiedlich ist.
+
+<Tip>
+
+Denken Sie daran, dass sich die Architektur auf das Skelett des Modells bezieht und die Checkpoints die Gewichte für eine bestimmte Architektur sind. Zum Beispiel ist [BERT](https://huggingface.co/bert-base-uncased) eine Architektur, während `bert-base-uncased` ein Checkpoint ist. Modell ist ein allgemeiner Begriff, der entweder Architektur oder Prüfpunkt bedeuten kann.
+
+</Tip>
+
+In dieser Anleitung lernen Sie, wie man:
+
+* Einen vortrainierten Tokenizer lädt.
+* Einen vortrainierten Merkmalsextraktor lädt.
+* Einen vortrainierten Prozessor lädt.
+* Ein vortrainiertes Modell lädt.
+
+## AutoTokenizer
+
+Nahezu jede NLP-Aufgabe beginnt mit einem Tokenizer. Ein Tokenizer wandelt Ihre Eingabe in ein Format um, das vom Modell verarbeitet werden kann.
+
+Laden Sie einen Tokenizer mit [`AutoTokenizer.from_pretrained`]:
+
+```py
+>>> from transformers import AutoTokenizer
+
+>>> tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
+```
+
+Dann tokenisieren Sie Ihre Eingabe wie unten gezeigt:
+
+```py
+>>> sequence = "In a hole in the ground there lived a hobbit."
+>>> print(tokenizer(sequence))
+{'input_ids': [101, 1999, 1037, 4920, 1999, 1996, 2598, 2045, 2973, 1037, 7570, 10322, 4183, 1012, 102], 
+ 'token_type_ids': [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
+ 'attention_mask': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]}
+```
+
+## AutoFeatureExtractor
+
+Für Audio- und Bildverarbeitungsaufgaben verarbeitet ein Merkmalsextraktor das Audiosignal oder Bild in das richtige Eingabeformat.
+
+Laden Sie einen Merkmalsextraktor mit [`AutoFeatureExtractor.from_pretrained`]:
+
+```py
+>>> from transformers import AutoFeatureExtractor
+
+>>> feature_extractor = AutoFeatureExtractor.from_pretrained(
+...     "ehcalabres/wav2vec2-lg-xlsr-en-speech-emotion-recognition"
+... )
+```
+
+## AutoProcessor
+
+Multimodale Aufgaben erfordern einen Prozessor, der zwei Arten von Vorverarbeitungswerkzeugen kombiniert. Das Modell [LayoutLMV2](model_doc/layoutlmv2) beispielsweise benötigt einen Feature-Extraktor für Bilder und einen Tokenizer für Text; ein Prozessor kombiniert beide.
+
+Laden Sie einen Prozessor mit [`AutoProcessor.from_pretrained`]:
+
+```py
+>>> from transformers import AutoProcessor
+
+>>> processor = AutoProcessor.from_pretrained("microsoft/layoutlmv2-base-uncased")
+```
+
+## AutoModel
+
+<frameworkcontent>
+<pt>
+Mit den `AutoModelFor`-Klassen können Sie schließlich ein vortrainiertes Modell für eine bestimmte Aufgabe laden (siehe [hier](model_doc/auto) für eine vollständige Liste der verfügbaren Aufgaben). Laden Sie zum Beispiel ein Modell für die Sequenzklassifikation mit [`AutoModelForSequenceClassification.from_pretrained`]:
+
+```py
+>>> from transformers import AutoModelForSequenceClassification
+
+>>> model = AutoModelForSequenceClassification.from_pretrained("distilbert-base-uncased")
+```
+
+Sie können denselben Prüfpunkt problemlos wiederverwenden, um eine Architektur für eine andere Aufgabe zu laden:
+
+```py
+>>> from transformers import AutoModelForTokenClassification
+
+>>> model = AutoModelForTokenClassification.from_pretrained("distilbert-base-uncased")
+```
+
+<Tip warning={true}>
+
+Für PyTorch-Modelle verwendet die Methode `from_pretrained()` `torch.load()`, die intern `pickle` verwendet und als unsicher bekannt ist. Generell sollte man niemals ein Modell laden, das aus einer nicht vertrauenswürdigen Quelle stammen könnte, oder das manipuliert worden sein könnte. Dieses Sicherheitsrisiko wird für öffentliche Modelle, die auf dem Hugging Face Hub gehostet werden, teilweise gemildert, da diese bei jeder Übertragung [auf Malware](https://huggingface.co/docs/hub/security-malware) gescannt werden. Siehe die [Hub-Dokumentation](https://huggingface.co/docs/hub/security) für Best Practices wie [signierte Commit-Verifizierung](https://huggingface.co/docs/hub/security-gpg#signing-commits-with-gpg) mit GPG.
+
+TensorFlow- und Flax-Checkpoints sind nicht betroffen und können in PyTorch-Architekturen mit den Kwargs `from_tf` und `from_flax` für die Methode `from_pretrained` geladen werden, um dieses Problem zu umgehen.
+
+</Tip>
+
+Im Allgemeinen empfehlen wir die Verwendung der Klasse "AutoTokenizer" und der Klasse "AutoModelFor", um trainierte Instanzen von Modellen zu laden. Dadurch wird sichergestellt, dass Sie jedes Mal die richtige Architektur laden. Im nächsten [Tutorial] (Vorverarbeitung) erfahren Sie, wie Sie Ihren neu geladenen Tokenizer, Feature Extractor und Prozessor verwenden, um einen Datensatz für die Feinabstimmung vorzuverarbeiten.
+</pt>
+<tf>
+Mit den Klassen `TFAutoModelFor` schließlich können Sie ein vortrainiertes Modell für eine bestimmte Aufgabe laden (siehe [hier](model_doc/auto) für eine vollständige Liste der verfügbaren Aufgaben). Laden Sie zum Beispiel ein Modell für die Sequenzklassifikation mit [`TFAutoModelForSequenceClassification.from_pretrained`]:
+
+```py
+>>> from transformers import TFAutoModelForSequenceClassification
+
+>>> model = TFAutoModelForSequenceClassification.from_pretrained("distilbert-base-uncased")
+```
+
+Sie können denselben Prüfpunkt problemlos wiederverwenden, um eine Architektur für eine andere Aufgabe zu laden:
+
+```py
+>>> from transformers import TFAutoModelForTokenClassification
+
+>>> model = TFAutoModelForTokenClassification.from_pretrained("distilbert-base-uncased")
+```
+
+Im Allgemeinen empfehlen wir, die Klasse "AutoTokenizer" und die Klasse "TFAutoModelFor" zu verwenden, um vortrainierte Instanzen von Modellen zu laden. Dadurch wird sichergestellt, dass Sie jedes Mal die richtige Architektur laden. Im nächsten [Tutorial] (Vorverarbeitung) erfahren Sie, wie Sie Ihren neu geladenen Tokenizer, Feature Extractor und Prozessor verwenden, um einen Datensatz für die Feinabstimmung vorzuverarbeiten.
+</tf>
+</frameworkcontent>

docs/source/de/index.mdx

@@ -28,8 +28,8 @@ Jede 🤗 Transformers-Architektur ist in einem eigenständigen Python-Modul def
 ## Wenn Sie auf der Suche nach individueller Unterstützung durch das Hugging Face-Team sind
 
 <a target="_blank" href="https://huggingface.co/support">
-    <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>
+    <img alt="HuggingFace Expert Acceleration Program" src="https://cdn-media.huggingface.co/marketing/transformers/new-support-improved.png" style="width: 100%; max-width: 600px; border: 1px solid #eee; border-radius: 4px; box-shadow: 0 1px 2px 0 rgba(0, 0, 0, 0.05);">
+</a>
 
 ## Inhalt
 
@@ -63,7 +63,7 @@ Die Bibliothek enthält derzeit JAX-, PyTorch- und TensorFlow-Implementierungen,
 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. **[BLOOM](model_doc/bloom)** (from BigScience workshop) released by the [BigScience 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.

docs/source/de/model_sharing.mdx

@@ -0,0 +1,228 @@
+<!--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.
+-->
+
+# Ein Modell teilen
+
+Die letzten beiden Tutorials haben gezeigt, wie man ein Modell mit PyTorch, Keras und 🤗 Accelerate für verteilte Setups feinabstimmen kann. Der nächste Schritt besteht darin, Ihr Modell mit der Community zu teilen! Bei Hugging Face glauben wir an den offenen Austausch von Wissen und Ressourcen, um künstliche Intelligenz für alle zu demokratisieren. Wir ermutigen Sie, Ihr Modell mit der Community zu teilen, um anderen zu helfen, Zeit und Ressourcen zu sparen.
+
+In diesem Tutorial lernen Sie zwei Methoden kennen, wie Sie ein trainiertes oder verfeinertes Modell auf dem [Model Hub](https://huggingface.co/models) teilen können:
+
+- Programmgesteuertes Übertragen Ihrer Dateien auf den Hub.
+- Ziehen Sie Ihre Dateien per Drag-and-Drop über die Weboberfläche in den Hub.
+
+<iframe width="560" height="315" src="https://www.youtube.com/embed/XvSGPZFEjDY" title="YouTube video player"
+frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope;
+picture-in-picture" allowfullscreen></iframe>
+
+<Tip>
+
+Um ein Modell mit der Öffentlichkeit zu teilen, benötigen Sie ein Konto auf [huggingface.co](https://huggingface.co/join). Sie können auch einer bestehenden Organisation beitreten oder eine neue Organisation gründen.
+
+</Tip>
+
+## Repository-Funktionen
+
+Jedes Repository im Model Hub verhält sich wie ein typisches GitHub-Repository. Unsere Repositorys bieten Versionierung, Commit-Historie und die Möglichkeit, Unterschiede zu visualisieren.
+
+Die integrierte Versionierung des Model Hub basiert auf Git und [git-lfs](https://git-lfs.github.com/). Mit anderen Worten: Sie können ein Modell als ein Repository behandeln, was eine bessere Zugriffskontrolle und Skalierbarkeit ermöglicht. Die Versionskontrolle ermöglicht *Revisionen*, eine Methode zum Anheften einer bestimmten Version eines Modells mit einem Commit-Hash, Tag oder Branch.
+
+Folglich können Sie eine bestimmte Modellversion mit dem Parameter "Revision" laden:
+
+```py
+>>> model = AutoModel.from_pretrained(
+...     "julien-c/EsperBERTo-small", revision="v2.0.1"  # tag name, or branch name, or commit hash
+... )
+```
+
+Dateien lassen sich auch in einem Repository leicht bearbeiten, und Sie können die Commit-Historie sowie die Unterschiede einsehen:
+
+![vis_diff](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/vis_diff.png)
+
+## Einrichtung
+
+Bevor Sie ein Modell für den Hub freigeben, benötigen Sie Ihre Hugging Face-Anmeldedaten. Wenn Sie Zugang zu einem Terminal haben, führen Sie den folgenden Befehl in der virtuellen Umgebung aus, in der 🤗 Transformers installiert ist. Dadurch werden Ihre Zugangsdaten in Ihrem Hugging Face-Cache-Ordner (standardmäßig `~/.cache/`) gespeichert:
+
+```bash
+huggingface-cli login
+```
+
+Wenn Sie ein Notebook wie Jupyter oder Colaboratory verwenden, stellen Sie sicher, dass Sie die [`huggingface_hub`](https://huggingface.co/docs/hub/adding-a-library) Bibliothek installiert haben. Diese Bibliothek ermöglicht Ihnen die programmatische Interaktion mit dem Hub.
+
+```bash
+pip install huggingface_hub
+```
+
+Verwenden Sie dann `notebook_login`, um sich beim Hub anzumelden, und folgen Sie dem Link [hier](https://huggingface.co/settings/token), um ein Token für die Anmeldung zu generieren:
+
+```py
+>>> from huggingface_hub import notebook_login
+
+>>> notebook_login()
+```
+
+## Ein Modell für alle Frameworks konvertieren
+
+Um sicherzustellen, dass Ihr Modell von jemandem verwendet werden kann, der mit einem anderen Framework arbeitet, empfehlen wir Ihnen, Ihr Modell sowohl mit PyTorch- als auch mit TensorFlow-Checkpoints zu konvertieren und hochzuladen. Während Benutzer immer noch in der Lage sind, Ihr Modell von einem anderen Framework zu laden, wenn Sie diesen Schritt überspringen, wird es langsamer sein, weil 🤗 Transformers den Checkpoint on-the-fly konvertieren müssen.
+
+Die Konvertierung eines Checkpoints für ein anderes Framework ist einfach. Stellen Sie sicher, dass Sie PyTorch und TensorFlow installiert haben (siehe [hier](installation) für Installationsanweisungen), und finden Sie dann das spezifische Modell für Ihre Aufgabe in dem anderen Framework. 
+
+<frameworkcontent>
+<pt>
+Geben Sie `from_tf=True` an, um einen Prüfpunkt von TensorFlow nach PyTorch zu konvertieren:
+
+```py
+>>> pt_model = DistilBertForSequenceClassification.from_pretrained("path/to/awesome-name-you-picked", from_tf=True)
+>>> pt_model.save_pretrained("path/to/awesome-name-you-picked")
+```
+</pt>
+<tf>
+Geben Sie `from_pt=True` an, um einen Prüfpunkt von PyTorch nach TensorFlow zu konvertieren:
+
+```py
+>>> tf_model = TFDistilBertForSequenceClassification.from_pretrained("path/to/awesome-name-you-picked", from_pt=True)
+```
+
+Dann können Sie Ihr neues TensorFlow-Modell mit seinem neuen Checkpoint speichern:
+
+```py
+>>> tf_model.save_pretrained("path/to/awesome-name-you-picked")
+```
+</tf>
+<jax>
+Wenn ein Modell in Flax verfügbar ist, können Sie auch einen Kontrollpunkt von PyTorch nach Flax konvertieren:
+
+```py
+>>> flax_model = FlaxDistilBertForSequenceClassification.from_pretrained(
+...     "path/to/awesome-name-you-picked", from_pt=True
+... )
+```
+</jax>
+</frameworkcontent>
+
+## Ein Modell während des Trainings hochladen
+
+<frameworkcontent>
+<pt>
+<Youtube id="Z1-XMy-GNLQ"/>
+
+Die Weitergabe eines Modells an den Hub ist so einfach wie das Hinzufügen eines zusätzlichen Parameters oder Rückrufs. Erinnern Sie sich an das [Feinabstimmungs-Tutorial](training), in der Klasse [`TrainingArguments`] geben Sie Hyperparameter und zusätzliche Trainingsoptionen an. Eine dieser Trainingsoptionen beinhaltet die Möglichkeit, ein Modell direkt an den Hub zu pushen. Setzen Sie `push_to_hub=True` in Ihrer [`TrainingArguments`]:
+
+```py
+>>> training_args = TrainingArguments(output_dir="my-awesome-model", push_to_hub=True)
+```
+
+Übergeben Sie Ihre Trainingsargumente wie gewohnt an [`Trainer`]:
+
+```py
+>>> trainer = Trainer(
+...     model=model,
+...     args=training_args,
+...     train_dataset=small_train_dataset,
+...     eval_dataset=small_eval_dataset,
+...     compute_metrics=compute_metrics,
+... )
+```
+
+Nach der Feinabstimmung Ihres Modells rufen Sie [`~transformers.Trainer.push_to_hub`] auf [`Trainer`] auf, um das trainierte Modell an den Hub zu übertragen. Transformers fügt sogar automatisch Trainings-Hyperparameter, Trainingsergebnisse und Framework-Versionen zu Ihrer Modellkarte hinzu!
+
+```py
+>>> trainer.push_to_hub()
+```
+</pt>
+<tf>
+Geben Sie ein Modell mit [`PushToHubCallback`] an den Hub weiter. In der [`PushToHubCallback`] Funktion, fügen Sie hinzu:
+
+- Ein Ausgabeverzeichnis für Ihr Modell.
+- Einen Tokenizer.
+- Die `hub_model_id`, die Ihr Hub-Benutzername und Modellname ist.
+
+```py
+>>> from transformers.keras.callbacks import PushToHubCallback
+
+>>> push_to_hub_callback = PushToHubCallback(
+...     output_dir="./your_model_save_path", tokenizer=tokenizer, hub_model_id="your-username/my-awesome-model"
+... )
+```
+
+Fügen Sie den Callback zu [`fit`](https://keras.io/api/models/model_training_apis/) hinzu, und 🤗 Transformers wird das trainierte Modell an den Hub weiterleiten:
+
+```py
+>>> model.fit(tf_train_dataset, validation_data=tf_validation_dataset, epochs=3, callbacks=push_to_hub_callback)
+```
+</tf>
+</frameworkcontent>
+
+## Verwenden Sie die Funktion `push_to_hub`.
+
+Sie können `push_to_hub` auch direkt für Ihr Modell aufrufen, um es in den Hub hochzuladen.
+
+Geben Sie den Namen Ihres Modells in "push_to_hub" an:
+
+```py
+>>> pt_model.push_to_hub("my-awesome-model")
+```
+
+Dadurch wird ein Repository unter Ihrem Benutzernamen mit dem Modellnamen `my-awesome-model` erstellt. Benutzer können nun Ihr Modell mit der Funktion `from_pretrained` laden:
+
+```py
+>>> from transformers import AutoModel
+
+>>> model = AutoModel.from_pretrained("your_username/my-awesome-model")
+```
+
+Wenn Sie zu einer Organisation gehören und Ihr Modell stattdessen unter dem Namen der Organisation pushen wollen, fügen Sie diesen einfach zur `repo_id` hinzu:
+
+```py
+>>> pt_model.push_to_hub("my-awesome-org/my-awesome-model")
+```
+
+Die Funktion "push_to_hub" kann auch verwendet werden, um andere Dateien zu einem Modell-Repository hinzuzufügen. Zum Beispiel kann man einen Tokenizer zu einem Modell-Repository hinzufügen:
+
+```py
+>>> tokenizer.push_to_hub("my-awesome-model")
+```
+
+Oder vielleicht möchten Sie die TensorFlow-Version Ihres fein abgestimmten PyTorch-Modells hinzufügen:
+
+```py
+>>> tf_model.push_to_hub("my-awesome-model")
+```
+
+Wenn Sie nun zu Ihrem Hugging Face-Profil navigieren, sollten Sie Ihr neu erstelltes Modell-Repository sehen. Wenn Sie auf die Registerkarte **Dateien** klicken, werden alle Dateien angezeigt, die Sie in das Repository hochgeladen haben.
+
+Weitere Einzelheiten zum Erstellen und Hochladen von Dateien in ein Repository finden Sie in der Hub-Dokumentation [hier](https://huggingface.co/docs/hub/how-to-upstream).
+
+## Hochladen mit der Weboberfläche
+
+Benutzer, die einen no-code Ansatz bevorzugen, können ein Modell über das Webinterface des Hubs hochladen. Besuchen Sie [huggingface.co/new](https://huggingface.co/new) um ein neues Repository zu erstellen:
+
+![new_model_repo](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/new_model_repo.png)
+
+Fügen Sie von hier aus einige Informationen über Ihr Modell hinzu:
+
+- Wählen Sie den **Besitzer** des Repositorys. Dies können Sie selbst oder eine der Organisationen sein, denen Sie angehören.
+- Wählen Sie einen Namen für Ihr Modell, der auch der Name des Repositorys sein wird.
+- Wählen Sie, ob Ihr Modell öffentlich oder privat ist.
+- Geben Sie die Lizenzverwendung für Ihr Modell an.
+
+Klicken Sie nun auf die Registerkarte **Dateien** und klicken Sie auf die Schaltfläche **Datei hinzufügen**, um eine neue Datei in Ihr Repository hochzuladen. Ziehen Sie dann eine Datei per Drag-and-Drop hoch und fügen Sie eine Übergabemeldung hinzu.
+
+![upload_file](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/upload_file.png)
+
+## Hinzufügen einer Modellkarte
+
+Um sicherzustellen, dass die Benutzer die Fähigkeiten, Grenzen, möglichen Verzerrungen und ethischen Aspekte Ihres Modells verstehen, fügen Sie bitte eine Modellkarte zu Ihrem Repository hinzu. Die Modellkarte wird in der Datei `README.md` definiert. Sie können eine Modellkarte hinzufügen, indem Sie:
+
+* Manuelles Erstellen und Hochladen einer "README.md"-Datei.
+* Klicken Sie auf die Schaltfläche **Modellkarte bearbeiten** in Ihrem Modell-Repository.
+
+Werfen Sie einen Blick auf die DistilBert [model card](https://huggingface.co/distilbert-base-uncased) als gutes Beispiel für die Art von Informationen, die eine Modellkarte enthalten sollte. Weitere Details über andere Optionen, die Sie in der Datei "README.md" einstellen können, wie z.B. den Kohlenstoff-Fußabdruck eines Modells oder Beispiele für Widgets, finden Sie in der Dokumentation [hier](https://huggingface.co/docs/hub/models-cards).

docs/source/de/pipeline_tutorial.mdx

@@ -56,7 +56,7 @@ Wenn Sie mehr als eine Eingabe haben, übergeben Sie die Eingabe als Liste:
 ... )  # doctest: +SKIP
 ```
 
-Alle zusätzlichen Parameter für Ihre Aufgabe können auch in die [`pipeline`] aufgenommen werden. Die Aufgabe `Text-Generierung` hat eine [`~generation_utils.GenerationMixin.generate`]-Methode mit mehreren Parametern zur Steuerung der Ausgabe. Wenn Sie zum Beispiel mehr als eine Ausgabe erzeugen wollen, setzen Sie den Parameter `num_return_sequences`:
+Alle zusätzlichen Parameter für Ihre Aufgabe können auch in die [`pipeline`] aufgenommen werden. Die Aufgabe `Text-Generierung` hat eine [`~generation.GenerationMixin.generate`]-Methode mit mehreren Parametern zur Steuerung der Ausgabe. Wenn Sie zum Beispiel mehr als eine Ausgabe erzeugen wollen, setzen Sie den Parameter `num_return_sequences`:
 
 ```py
 >>> generator(

docs/source/de/preprocessing.mdx

@@ -0,0 +1,502 @@
+<!--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.
+-->
+
+# Vorverarbeiten
+
+[[open-in-colab]]
+
+Bevor Sie Ihre Daten in einem Modell verwenden können, müssen die Daten in ein für das Modell akzeptables Format gebracht werden. Ein Modell versteht keine Rohtexte, Bilder oder Audiodaten. Diese Eingaben müssen in Zahlen umgewandelt und zu Tensoren zusammengesetzt werden. In dieser Anleitung werden Sie:
+
+* Textdaten mit einem Tokenizer vorverarbeiten.
+* Bild- oder Audiodaten mit einem Feature Extractor vorverarbeiten.
+* Daten für eine multimodale Aufgabe mit einem Prozessor vorverarbeiten.
+
+## NLP
+
+<Youtube id="Yffk5aydLzg"/>
+
+Das wichtigste Werkzeug zur Verarbeitung von Textdaten ist ein [Tokenizer](main_classes/tokenizer). Ein Tokenizer zerlegt Text zunächst nach einer Reihe von Regeln in *Token*. Die Token werden in Zahlen umgewandelt, die zum Aufbau von Tensoren als Eingabe für ein Modell verwendet werden. Alle zusätzlichen Eingaben, die ein Modell benötigt, werden ebenfalls vom Tokenizer hinzugefügt.
+
+<Tip>
+
+Wenn Sie ein vortrainiertes Modell verwenden möchten, ist es wichtig, den zugehörigen vortrainierten Tokenizer zu verwenden. Dadurch wird sichergestellt, dass der Text auf die gleiche Weise aufgeteilt wird wie das Pretraining-Korpus und die gleichen entsprechenden Token-zu-Index (in der Regel als *vocab* bezeichnet) während des Pretrainings verwendet werden.
+
+</Tip>
+
+Laden Sie einen vortrainierten Tokenizer mit der Klasse [AutoTokenizer], um schnell loszulegen. Damit wird das *vocab* heruntergeladen, das verwendet wird, wenn ein Modell vortrainiert wird.
+
+### Tokenize
+
+Laden Sie einen vortrainierten Tokenizer mit [`AutoTokenizer.from_pretrained`]:
+
+```py
+>>> from transformers import AutoTokenizer
+
+>>> tokenizer = AutoTokenizer.from_pretrained("bert-base-cased")
+```
+
+Dann übergeben Sie Ihren Satz an den Tokenizer:
+
+```py
+>>> encoded_input = tokenizer("Do not meddle in the affairs of wizards, for they are subtle and quick to anger.")
+>>> print(encoded_input)
+{'input_ids': [101, 2079, 2025, 19960, 10362, 1999, 1996, 3821, 1997, 16657, 1010, 2005, 2027, 2024, 11259, 1998, 4248, 2000, 4963, 1012, 102], 
+ 'token_type_ids': [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
+ 'attention_mask': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]}
+```
+
+Der Tokenizer gibt ein Wörterbuch mit drei wichtigen Elementen zurück:
+
+* [input_ids](glossary#input-ids) sind die Indizes, die den einzelnen Token im Satz entsprechen.
+* [attention_mask](glossary#attention-mask) gibt an, ob ein Token beachtet werden soll oder nicht.
+* [token_type_ids](glossary#token-type-ids) gibt an, zu welcher Sequenz ein Token gehört, wenn es mehr als eine Sequenz gibt.
+
+Sie können die `input_ids` dekodieren, um die ursprüngliche Eingabe zurückzugeben:
+
+```py
+>>> tokenizer.decode(encoded_input["input_ids"])
+'[CLS] Do not meddle in the affairs of wizards, for they are subtle and quick to anger. [SEP]'
+```
+
+Wie Sie sehen können, hat der Tokenisierer zwei spezielle Token - `CLS` und `SEP` (Klassifikator und Separator) - zum Satz hinzugefügt. Nicht alle Modelle benötigen
+spezielle Token, aber wenn dies der Fall ist, fügt der Tokenisierer sie automatisch für Sie hinzu.
+
+Wenn Sie mehrere Sätze verarbeiten wollen, übergeben Sie die Sätze als Liste an den Tokenizer:
+
+```py
+>>> batch_sentences = [
+...     "But what about second breakfast?",
+...     "Don't think he knows about second breakfast, Pip.",
+...     "What about elevensies?",
+... ]
+>>> encoded_inputs = tokenizer(batch_sentences)
+>>> print(encoded_inputs)
+{'input_ids': [[101, 1252, 1184, 1164, 1248, 6462, 136, 102], 
+               [101, 1790, 112, 189, 1341, 1119, 3520, 1164, 1248, 6462, 117, 21902, 1643, 119, 102], 
+               [101, 1327, 1164, 5450, 23434, 136, 102]], 
+ 'token_type_ids': [[0, 0, 0, 0, 0, 0, 0, 0], 
+                    [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
+                    [0, 0, 0, 0, 0, 0, 0]], 
+ 'attention_mask': [[1, 1, 1, 1, 1, 1, 1, 1], 
+                    [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], 
+                    [1, 1, 1, 1, 1, 1, 1]]}
+```
+
+### Pad
+
+Dies bringt uns zu einem wichtigen Thema. Wenn Sie einen Haufen von Sätzen verarbeiten, sind diese nicht immer gleich lang. Das ist ein Problem, weil Tensoren, die Eingabe für das Modell, eine einheitliche Form haben müssen. Padding ist eine Strategie, die sicherstellt, dass Tensoren rechteckig sind, indem ein spezielles *Padding-Token* zu Sätzen mit weniger Token hinzugefügt wird.
+
+Setzen Sie den Parameter "padding" auf "true", um die kürzeren Sequenzen im Stapel so aufzufüllen, dass sie der längsten Sequenz entsprechen:
+
+```py
+>>> batch_sentences = [
+...     "But what about second breakfast?",
+...     "Don't think he knows about second breakfast, Pip.",
+...     "What about elevensies?",
+... ]
+>>> encoded_input = tokenizer(batch_sentences, padding=True)
+>>> print(encoded_input)
+{'input_ids': [[101, 1252, 1184, 1164, 1248, 6462, 136, 102, 0, 0, 0, 0, 0, 0, 0], 
+               [101, 1790, 112, 189, 1341, 1119, 3520, 1164, 1248, 6462, 117, 21902, 1643, 119, 102], 
+               [101, 1327, 1164, 5450, 23434, 136, 102, 0, 0, 0, 0, 0, 0, 0, 0]], 
+ 'token_type_ids': [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
+                    [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
+                    [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], 
+ 'attention_mask': [[1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0], 
+                    [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], 
+                    [1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0]]}
+```
+
+Beachten Sie, dass der Tokenizer den ersten und den dritten Satz mit einer "0" aufgefüllt hat, weil sie kürzer sind!
+
+### Kürzung
+
+Auf der anderen Seite des Spektrums kann es vorkommen, dass eine Sequenz zu lang für ein Modell ist. In diesem Fall müssen Sie die Sequenz auf eine kürzere Länge kürzen.
+
+Setzen Sie den Parameter "truncation" auf "true", um eine Sequenz auf die vom Modell akzeptierte Höchstlänge zu kürzen:
+
+```py
+>>> batch_sentences = [
+...     "But what about second breakfast?",
+...     "Don't think he knows about second breakfast, Pip.",
+...     "What about elevensies?",
+... ]
+>>> encoded_input = tokenizer(batch_sentences, padding=True, truncation=True)
+>>> print(encoded_input)
+{'input_ids': [[101, 1252, 1184, 1164, 1248, 6462, 136, 102, 0, 0, 0, 0, 0, 0, 0], 
+               [101, 1790, 112, 189, 1341, 1119, 3520, 1164, 1248, 6462, 117, 21902, 1643, 119, 102], 
+               [101, 1327, 1164, 5450, 23434, 136, 102, 0, 0, 0, 0, 0, 0, 0, 0]], 
+ 'token_type_ids': [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
+                    [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
+                    [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], 
+ 'attention_mask': [[1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0], 
+                    [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], 
+                    [1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0]]}
+```
+
+### Tensoren erstellen
+
+Schließlich möchten Sie, dass der Tokenizer die tatsächlichen Tensoren zurückgibt, die dem Modell zugeführt werden.
+
+Setzen Sie den Parameter `return_tensors` entweder auf `pt` für PyTorch, oder `tf` für TensorFlow:
+
+<frameworkcontent>
+<pt>
+
+```py
+>>> batch_sentences = [
+...     "But what about second breakfast?",
+...     "Don't think he knows about second breakfast, Pip.",
+...     "What about elevensies?",
+... ]
+>>> encoded_input = tokenizer(batch_sentences, padding=True, truncation=True, return_tensors="pt")
+>>> print(encoded_input)
+{'input_ids': tensor([[101, 1252, 1184, 1164, 1248, 6462, 136, 102, 0, 0, 0, 0, 0, 0, 0],
+                      [101, 1790, 112, 189, 1341, 1119, 3520, 1164, 1248, 6462, 117, 21902, 1643, 119, 102],
+                      [101, 1327, 1164, 5450, 23434, 136, 102, 0, 0, 0, 0, 0, 0, 0, 0]]), 
+ 'token_type_ids': tensor([[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+                           [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+                           [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]), 
+ 'attention_mask': tensor([[1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0],
+                           [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
+                           [1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0]])}
+```
+</pt>
+<tf>
+```py
+>>> batch_sentences = [
+...     "But what about second breakfast?",
+...     "Don't think he knows about second breakfast, Pip.",
+...     "What about elevensies?",
+... ]
+>>> encoded_input = tokenizer(batch_sentences, padding=True, truncation=True, return_tensors="tf")
+>>> print(encoded_input)
+{'input_ids': <tf.Tensor: shape=(2, 9), dtype=int32, numpy=
+array([[101, 1252, 1184, 1164, 1248, 6462, 136, 102, 0, 0, 0, 0, 0, 0, 0],
+       [101, 1790, 112, 189, 1341, 1119, 3520, 1164, 1248, 6462, 117, 21902, 1643, 119, 102],
+       [101, 1327, 1164, 5450, 23434, 136, 102, 0, 0, 0, 0, 0, 0, 0, 0]],
+      dtype=int32)>, 
+ 'token_type_ids': <tf.Tensor: shape=(2, 9), dtype=int32, numpy=
+array([[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], dtype=int32)>, 
+ 'attention_mask': <tf.Tensor: shape=(2, 9), dtype=int32, numpy=
+array([[1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0],
+       [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
+       [1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0]], dtype=int32)>}
+```
+</tf>
+</frameworkcontent>
+
+## Audio
+
+Audioeingaben werden anders vorverarbeitet als Texteingaben, aber das Endziel bleibt dasselbe: numerische Sequenzen zu erstellen, die das Modell verstehen kann. Ein [feature extractor](main_classes/feature_extractor) dient dem ausdrücklichen Zweck, Merkmale aus Rohbild- oder Audiodaten zu extrahieren und in Tensoren zu konvertieren. Bevor Sie beginnen, installieren Sie 🤗 Datasets, um einen Audio-Datensatz zu laden, mit dem Sie experimentieren können:
+
+```bash
+pip install datasets
+```
+
+Laden Sie den [MInDS-14](https://huggingface.co/datasets/PolyAI/minds14) Datensatz (weitere Informationen zum Laden eines Datensatzes finden Sie im 🤗 [Datasets tutorial](https://huggingface.co/docs/datasets/load_hub.html)):
+
+```py
+>>> from datasets import load_dataset, Audio
+
+>>> dataset = load_dataset("PolyAI/minds14", name="en-US", split="train")
+```
+
+Greifen Sie auf das erste Element der `audio`-Spalte zu, um einen Blick auf die Eingabe zu werfen. Durch den Aufruf der Spalte "audio" wird die Audiodatei automatisch geladen und neu gesampelt:
+
+```py
+>>> dataset[0]["audio"]
+{'array': array([ 0.        ,  0.00024414, -0.00024414, ..., -0.00024414,
+         0.        ,  0.        ], dtype=float32),
+ 'path': '/root/.cache/huggingface/datasets/downloads/extracted/f14948e0e84be638dd7943ac36518a4cf3324e8b7aa331c5ab11541518e9368c/en-US~JOINT_ACCOUNT/602ba55abb1e6d0fbce92065.wav',
+ 'sampling_rate': 8000}
+```
+
+Dies gibt drei Elemente zurück:
+
+* "array" ist das Sprachsignal, das als 1D-Array geladen - und möglicherweise neu gesampelt - wurde.
+* Pfad" zeigt auf den Speicherort der Audiodatei.
+* `sampling_rate` bezieht sich darauf, wie viele Datenpunkte im Sprachsignal pro Sekunde gemessen werden.
+
+### Resample
+
+Für dieses Tutorial werden Sie das Modell [Wav2Vec2](https://huggingface.co/facebook/wav2vec2-base) verwenden. Wie Sie aus der Modellkarte ersehen können, ist das Wav2Vec2-Modell auf 16kHz abgetastetes Sprachaudio vortrainiert. Es ist wichtig, dass die Abtastrate Ihrer Audiodaten mit der Abtastrate des Datensatzes übereinstimmt, der für das Pre-Training des Modells verwendet wurde. Wenn die Abtastrate Ihrer Daten nicht dieselbe ist, müssen Sie Ihre Audiodaten neu abtasten. 
+
+Der Datensatz [MInDS-14](https://huggingface.co/datasets/PolyAI/minds14) hat zum Beispiel eine Abtastrate von 8000 kHz. Um das Wav2Vec2-Modell mit diesem Datensatz verwenden zu können, müssen Sie die Abtastrate auf 16 kHz erhöhen:
+
+```py
+>>> dataset = load_dataset("PolyAI/minds14", name="en-US", split="train")
+>>> dataset[0]["audio"]
+{'array': array([ 0.        ,  0.00024414, -0.00024414, ..., -0.00024414,
+         0.        ,  0.        ], dtype=float32),
+ 'path': '/root/.cache/huggingface/datasets/downloads/extracted/f14948e0e84be638dd7943ac36518a4cf3324e8b7aa331c5ab11541518e9368c/en-US~JOINT_ACCOUNT/602ba55abb1e6d0fbce92065.wav',
+ 'sampling_rate': 8000}
+```
+
+1. Verwenden Sie die Methode [~datasets.Dataset.cast_column] von 🤗 Datasets, um die Abtastrate auf 16kHz zu erhöhen:
+
+```py
+>>> dataset = dataset.cast_column("audio", Audio(sampling_rate=16_000))
+```
+
+2. Laden Sie die Audiodatei:
+
+```py
+>>> dataset[0]["audio"]
+{'array': array([ 2.3443763e-05,  2.1729663e-04,  2.2145823e-04, ...,
+         3.8356509e-05, -7.3497440e-06, -2.1754686e-05], dtype=float32),
+ 'path': '/root/.cache/huggingface/datasets/downloads/extracted/f14948e0e84be638dd7943ac36518a4cf3324e8b7aa331c5ab11541518e9368c/en-US~JOINT_ACCOUNT/602ba55abb1e6d0fbce92065.wav',
+ 'sampling_rate': 16000}
+```
+
+Wie Sie sehen können, ist die Abtastrate jetzt 16kHz!
+
+### Merkmalsextraktor
+
+Der nächste Schritt ist das Laden eines Merkmalsextraktors, um die Eingabe zu normalisieren und aufzufüllen. Beim Auffüllen von Textdaten wird für kürzere Sequenzen ein `0` hinzugefügt. Die gleiche Idee gilt für Audiodaten, und der Audio-Feature-Extraktor fügt eine `0` - interpretiert als Stille - zu `array` hinzu.
+
+Laden Sie den Merkmalsextraktor mit [`AutoFeatureExtractor.from_pretrained`]:
+
+```py
+>>> from transformers import AutoFeatureExtractor
+
+>>> feature_extractor = AutoFeatureExtractor.from_pretrained("facebook/wav2vec2-base")
+```
+
+Übergeben Sie das Audio-"Array" an den Feature-Extraktor. Wir empfehlen auch, das Argument `sampling_rate` im Feature Extractor hinzuzufügen, um eventuell auftretende stille Fehler besser zu beheben.
+
+```py
+>>> audio_input = [dataset[0]["audio"]["array"]]
+>>> feature_extractor(audio_input, sampling_rate=16000)
+{'input_values': [array([ 3.8106556e-04,  2.7506407e-03,  2.8015103e-03, ...,
+        5.6335266e-04,  4.6588284e-06, -1.7142107e-04], dtype=float32)]}
+```
+
+### Auffüllen und Kürzen
+
+Genau wie beim Tokenizer können Sie variable Sequenzen in einem Stapel durch Auffüllen oder Abschneiden behandeln. Werfen Sie einen Blick auf die Sequenzlänge dieser beiden Audiobeispiele:
+
+```py
+>>> dataset[0]["audio"]["array"].shape
+(173398,)
+
+>>> dataset[1]["audio"]["array"].shape
+(106496,)
+```
+
+Wie Sie sehen können, hat das erste Beispiel eine längere Sequenz als das zweite Beispiel. Lassen Sie uns eine Funktion erstellen, die den Datensatz vorverarbeitet. Geben Sie eine maximale Länge der Probe an, und der Feature-Extraktor wird die Sequenzen entweder auffüllen oder abschneiden, damit sie dieser Länge entsprechen:
+
+```py
+>>> def preprocess_function(examples):
+...     audio_arrays = [x["array"] for x in examples["audio"]]
+...     inputs = feature_extractor(
+...         audio_arrays,
+...         sampling_rate=16000,
+...         padding=True,
+...         max_length=100000,
+...         truncation=True,
+...     )
+...     return inputs
+```
+
+Wenden Sie die Funktion auf die ersten paar Beispiele im Datensatz an:
+
+```py
+>>> processed_dataset = preprocess_function(dataset[:5])
+```
+
+Schauen Sie sich nun noch einmal die verarbeiteten Beispiel-Längen an:
+
+```py
+>>> processed_dataset["input_values"][0].shape
+(100000,)
+
+>>> processed_dataset["input_values"][1].shape
+(100000,)
+```
+
+Die Länge der ersten beiden Beispiele entspricht nun der von Ihnen angegebenen Maximallänge.
+
+## Bildverarbeitung
+
+Ein Merkmalsextraktor wird auch verwendet, um Bilder für Bildverarbeitungsaufgaben zu verarbeiten. Auch hier besteht das Ziel darin, das Rohbild in eine Reihe von Tensoren als Eingabe zu konvertieren.
+
+Laden wir den [food101](https://huggingface.co/datasets/food101) Datensatz für dieses Tutorial. Verwenden Sie den Parameter 🤗 Datasets `split`, um nur eine kleine Stichprobe aus dem Trainingssplit zu laden, da der Datensatz recht groß ist:
+
+```py
+>>> from datasets import load_dataset
+
+>>> dataset = load_dataset("food101", split="train[:100]")
+```
+
+Als Nächstes sehen Sie sich das Bild mit dem Merkmal 🤗 Datensätze [Bild] (https://huggingface.co/docs/datasets/package_reference/main_classes.html?highlight=image#datasets.Image) an:
+
+```py
+>>> dataset[0]["image"]
+```
+
+![vision-preprocess-tutorial.png](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/vision-preprocess-tutorial.png)
+
+### Merkmalsextraktor
+
+Laden Sie den Merkmalsextraktor mit [`AutoFeatureExtractor.from_pretrained`]:
+
+```py
+>>> from transformers import AutoFeatureExtractor
+
+>>> feature_extractor = AutoFeatureExtractor.from_pretrained("google/vit-base-patch16-224")
+```
+
+### Datenerweiterung
+
+Bei Bildverarbeitungsaufgaben ist es üblich, den Bildern als Teil der Vorverarbeitung eine Art von Datenerweiterung hinzuzufügen. Sie können Erweiterungen mit jeder beliebigen Bibliothek hinzufügen, aber in diesem Tutorial werden Sie das Modul [`transforms`](https://pytorch.org/vision/stable/transforms.html) von torchvision verwenden.
+
+1. Normalisieren Sie das Bild und verwenden Sie [`Compose`](https://pytorch.org/vision/master/generated/torchvision.transforms.Compose.html), um einige Transformationen - [`RandomResizedCrop`](https://pytorch.org/vision/main/generated/torchvision.transforms.RandomResizedCrop.html) und [`ColorJitter`](https://pytorch.org/vision/main/generated/torchvision.transforms.ColorJitter.html) - miteinander zu verknüpfen:
+
+```py
+>>> from torchvision.transforms import Compose, Normalize, RandomResizedCrop, ColorJitter, ToTensor
+
+>>> normalize = Normalize(mean=feature_extractor.image_mean, std=feature_extractor.image_std)
+>>> _transforms = Compose(
+...     [RandomResizedCrop(feature_extractor.size), ColorJitter(brightness=0.5, hue=0.5), ToTensor(), normalize]
+... )
+```
+
+2. Das Modell akzeptiert [`pixel_values`](model_doc/visionencoderdecoder#transformers.VisionEncoderDecoderModel.forward.pixel_values) als Eingabe. Dieser Wert wird vom Merkmalsextraktor erzeugt. Erstellen Sie eine Funktion, die `pixel_values` aus den Transformationen erzeugt:
+
+```py
+>>> def transforms(examples):
+...     examples["pixel_values"] = [_transforms(image.convert("RGB")) for image in examples["image"]]
+...     return examples
+```
+
+3. Dann verwenden Sie 🤗 Datasets [`set_transform`](https://huggingface.co/docs/datasets/process.html#format-transform), um die Transformationen im laufenden Betrieb anzuwenden:
+
+```py
+>>> dataset.set_transform(transforms)
+```
+
+4. Wenn Sie nun auf das Bild zugreifen, werden Sie feststellen, dass der Feature Extractor die Modelleingabe "pixel_values" hinzugefügt hat:
+
+```py
+>>> dataset[0]["image"]
+{'image': <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=384x512 at 0x7F1A7B0630D0>,
+ 'label': 6,
+ 'pixel_values': tensor([[[ 0.0353,  0.0745,  0.1216,  ..., -0.9922, -0.9922, -0.9922],
+          [-0.0196,  0.0667,  0.1294,  ..., -0.9765, -0.9843, -0.9922],
+          [ 0.0196,  0.0824,  0.1137,  ..., -0.9765, -0.9686, -0.8667],
+          ...,
+          [ 0.0275,  0.0745,  0.0510,  ..., -0.1137, -0.1216, -0.0824],
+          [ 0.0667,  0.0824,  0.0667,  ..., -0.0588, -0.0745, -0.0980],
+          [ 0.0353,  0.0353,  0.0431,  ..., -0.0039, -0.0039, -0.0588]],
+ 
+         [[ 0.2078,  0.2471,  0.2863,  ..., -0.9451, -0.9373, -0.9451],
+          [ 0.1608,  0.2471,  0.3098,  ..., -0.9373, -0.9451, -0.9373],
+          [ 0.2078,  0.2706,  0.3020,  ..., -0.9608, -0.9373, -0.8275],
+          ...,
+          [-0.0353,  0.0118, -0.0039,  ..., -0.2392, -0.2471, -0.2078],
+          [ 0.0196,  0.0353,  0.0196,  ..., -0.1843, -0.2000, -0.2235],
+          [-0.0118, -0.0039, -0.0039,  ..., -0.0980, -0.0980, -0.1529]],
+ 
+         [[ 0.3961,  0.4431,  0.4980,  ..., -0.9216, -0.9137, -0.9216],
+          [ 0.3569,  0.4510,  0.5216,  ..., -0.9059, -0.9137, -0.9137],
+          [ 0.4118,  0.4745,  0.5216,  ..., -0.9137, -0.8902, -0.7804],
+          ...,
+          [-0.2314, -0.1922, -0.2078,  ..., -0.4196, -0.4275, -0.3882],
+          [-0.1843, -0.1686, -0.2000,  ..., -0.3647, -0.3804, -0.4039],
+          [-0.1922, -0.1922, -0.1922,  ..., -0.2941, -0.2863, -0.3412]]])}
+```
+
+Hier sehen Sie, wie das Bild nach der Vorverarbeitung aussieht. Wie von den angewandten Transformationen zu erwarten, wurde das Bild willkürlich beschnitten und seine Farbeigenschaften sind anders.
+
+```py
+>>> import numpy as np
+>>> import matplotlib.pyplot as plt
+
+>>> img = dataset[0]["pixel_values"]
+>>> plt.imshow(img.permute(1, 2, 0))
+```
+
+![preprocessed_image](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/preprocessed_image.png)
+
+## Multimodal
+
+Für multimodale Aufgaben werden Sie eine Kombination aus allem, was Sie bisher gelernt haben, verwenden und Ihre Fähigkeiten auf eine Aufgabe der automatischen Spracherkennung (ASR) anwenden. Dies bedeutet, dass Sie einen:
+
+* Feature Extractor zur Vorverarbeitung der Audiodaten.
+* Tokenizer, um den Text zu verarbeiten.
+
+Kehren wir zum [LJ Speech](https://huggingface.co/datasets/lj_speech) Datensatz zurück:
+
+```py
+>>> from datasets import load_dataset
+
+>>> lj_speech = load_dataset("lj_speech", split="train")
+```
+
+Da Sie hauptsächlich an den Spalten "Audio" und "Text" interessiert sind, entfernen Sie die anderen Spalten:
+
+```py
+>>> lj_speech = lj_speech.map(remove_columns=["file", "id", "normalized_text"])
+```
+
+Schauen Sie sich nun die Spalten "Audio" und "Text" an:
+
+```py
+>>> lj_speech[0]["audio"]
+{'array': array([-7.3242188e-04, -7.6293945e-04, -6.4086914e-04, ...,
+         7.3242188e-04,  2.1362305e-04,  6.1035156e-05], dtype=float32),
+ 'path': '/root/.cache/huggingface/datasets/downloads/extracted/917ece08c95cf0c4115e45294e3cd0dee724a1165b7fc11798369308a465bd26/LJSpeech-1.1/wavs/LJ001-0001.wav',
+ 'sampling_rate': 22050}
+
+>>> lj_speech[0]["text"]
+'Printing, in the only sense with which we are at present concerned, differs from most if not from all the arts and crafts represented in the Exhibition'
+```
+
+Erinnern Sie sich an den früheren Abschnitt über die Verarbeitung von Audiodaten: Sie sollten immer die Abtastrate Ihrer Audiodaten [resample](preprocessing#audio), damit sie mit der Abtastrate des Datensatzes übereinstimmt, der für das Vortraining eines Modells verwendet wird:
+
+```py
+>>> lj_speech = lj_speech.cast_column("audio", Audio(sampling_rate=16_000))
+```
+
+### Prozessor
+
+Ein Processor kombiniert einen Feature-Extraktor und einen Tokenizer. Laden Sie einen Processor mit [`AutoProcessor.from_pretrained]:
+
+```py
+>>> from transformers import AutoProcessor
+
+>>> processor = AutoProcessor.from_pretrained("facebook/wav2vec2-base-960h")
+```
+
+1. Erstellen Sie eine Funktion, die die Audiodaten zu `input_values` verarbeitet und den Text zu `labels` tokenisiert. Dies sind Ihre Eingaben für das Modell:
+
+```py
+>>> def prepare_dataset(example):
+...     audio = example["audio"]
+
+...     example.update(processor(audio=audio["array"], text=example["text"], sampling_rate=16000))
+
+...     return example
+```
+
+2. Wenden Sie die Funktion "prepare_dataset" auf ein Beispiel an:
+
+```py
+>>> prepare_dataset(lj_speech[0])
+```
+
+Beachten Sie, dass der Processor `input_values` und `labels` hinzugefügt hat. Auch die Abtastrate wurde korrekt auf 16kHz heruntergerechnet.
+
+Toll, Sie sollten jetzt in der Lage sein, Daten für jede Modalität vorzuverarbeiten und sogar verschiedene Modalitäten zu kombinieren! Im nächsten Kurs lernen Sie, wie Sie ein Modell mit Ihren neu aufbereiteten Daten feinabstimmen können.

docs/source/de/training.mdx

@@ -0,0 +1,427 @@
+<!--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.
+-->
+
+# Optimierung eines vortrainierten Modells
+
+[[open-in-colab]]
+
+Die Verwendung eines vorab trainierten Modells hat erhebliche Vorteile. Es reduziert die Rechenkosten und den CO2-Fußabdruck und ermöglicht Ihnen die Verwendung von Modellen, die dem neuesten Stand der Technik entsprechen, ohne dass Sie ein Modell von Grund auf neu trainieren müssen. Transformers bietet Zugang zu Tausenden von vortrainierten Modellen für eine Vielzahl von Aufgaben. Wenn Sie ein vorab trainiertes Modell verwenden, trainieren Sie es auf einem für Ihre Aufgabe spezifischen Datensatz. Dies wird als Feinabstimmung bezeichnet und ist eine unglaublich leistungsfähige Trainingstechnik. In diesem Tutorial werden Sie ein vortrainiertes Modell mit einem Deep-Learning-Framework Ihrer Wahl feinabstimmen:
+
+* Feinabstimmung eines vorab trainierten Modells mit 🤗 Transformers [`Trainer`].
+* Feinabstimmung eines vorab trainierten Modells in TensorFlow mit Keras.
+* Feinabstimmung eines vorab trainierten Modells in nativem PyTorch.
+
+<a id='data-processing'></a>
+
+## Vorbereitung eines Datensatzes
+
+<Youtube id="_BZearw7f0w"/>
+
+Bevor Sie die Feinabstimmung eines vortrainierten Modells vornehmen können, müssen Sie einen Datensatz herunterladen und für das Training vorbereiten. Im vorangegangenen Leitfaden haben Sie gelernt, wie man Daten für das Training aufbereitet, und jetzt haben Sie die Gelegenheit, diese Fähigkeiten zu testen!
+
+Laden Sie zunächst den Datensatz [Yelp Reviews](https://huggingface.co/datasets/yelp_review_full):
+
+```py
+>>> from datasets import load_dataset
+
+>>> dataset = load_dataset("yelp_review_full")
+>>> dataset["train"][100]
+{'label': 0,
+ 'text': 'My expectations for McDonalds are t rarely high. But for one to still fail so spectacularly...that takes something special!\\nThe cashier took my friends\'s order, then promptly ignored me. I had to force myself in front of a cashier who opened his register to wait on the person BEHIND me. I waited over five minutes for a gigantic order that included precisely one kid\'s meal. After watching two people who ordered after me be handed their food, I asked where mine was. The manager started yelling at the cashiers for \\"serving off their orders\\" when they didn\'t have their food. But neither cashier was anywhere near those controls, and the manager was the one serving food to customers and clearing the boards.\\nThe manager was rude when giving me my order. She didn\'t make sure that I had everything ON MY RECEIPT, and never even had the decency to apologize that I felt I was getting poor service.\\nI\'ve eaten at various McDonalds restaurants for over 30 years. I\'ve worked at more than one location. I expect bad days, bad moods, and the occasional mistake. But I have yet to have a decent experience at this store. It will remain a place I avoid unless someone in my party needs to avoid illness from low blood sugar. Perhaps I should go back to the racially biased service of Steak n Shake instead!'}
+```
+
+Wie Sie nun wissen, benötigen Sie einen Tokenizer, um den Text zu verarbeiten und eine Auffüll- und Abschneidungsstrategie einzubauen, um mit variablen Sequenzlängen umzugehen. Um Ihren Datensatz in einem Schritt zu verarbeiten, verwenden Sie die 🤗 Methode Datasets [`map`](https://huggingface.co/docs/datasets/process.html#map), um eine Vorverarbeitungsfunktion auf den gesamten Datensatz anzuwenden:
+
+```py
+>>> from transformers import AutoTokenizer
+
+>>> tokenizer = AutoTokenizer.from_pretrained("bert-base-cased")
+
+
+>>> def tokenize_function(examples):
+...     return tokenizer(examples["text"], padding="max_length", truncation=True)
+
+
+>>> tokenized_datasets = dataset.map(tokenize_function, batched=True)
+```
+
+Wenn Sie möchten, können Sie eine kleinere Teilmenge des gesamten Datensatzes für die Feinabstimmung erstellen, um den Zeitaufwand zu verringern:
+
+```py
+>>> small_train_dataset = tokenized_datasets["train"].shuffle(seed=42).select(range(1000))
+>>> small_eval_dataset = tokenized_datasets["test"].shuffle(seed=42).select(range(1000))
+```
+
+<a id='trainer'></a>
+
+## Training
+
+An dieser Stelle sollten Sie dem Abschnitt folgen, der dem Rahmen entspricht, den Sie verwenden möchten. Sie können über die Links
+in der rechten Seitenleiste können Sie zu dem gewünschten Abschnitt springen - und wenn Sie den gesamten Inhalt eines bestimmten Frameworks ausblenden möchten,
+klicken Sie einfach auf die Schaltfläche oben rechts im Block des jeweiligen Frameworks!
+
+<frameworkcontent>
+<pt>
+<Youtube id="nvBXf7s7vTI"/>
+
+## Trainieren mit PyTorch Trainer
+
+🤗 Transformers bietet eine [`Trainer`]-Klasse, die für das Training von 🤗 Transformers-Modellen optimiert ist und es einfacher macht, mit dem Training zu beginnen, ohne manuell eine eigene Trainingsschleife zu schreiben. Die [`Trainer`]-API unterstützt eine breite Palette von Trainingsoptionen und Funktionen wie Logging, Gradientenakkumulation und gemischte Präzision.
+
+Beginnen Sie mit dem Laden Ihres Modells und geben Sie die Anzahl der erwarteten Labels an. Aus dem Yelp Review [dataset card](https://huggingface.co/datasets/yelp_review_full#data-fields) wissen Sie, dass es fünf Labels gibt:
+
+```py
+>>> from transformers import AutoModelForSequenceClassification
+
+>>> model = AutoModelForSequenceClassification.from_pretrained("bert-base-cased", num_labels=5)
+```
+
+<Tip>
+
+Es wird eine Warnung angezeigt, dass einige der trainierten Parameter nicht verwendet werden und einige Parameter zufällig
+initialisiert werden. Machen Sie sich keine Sorgen, das ist völlig normal! Der vorher trainierte Kopf des BERT-Modells wird verworfen und durch einen zufällig initialisierten Klassifikationskopf ersetzt. Sie werden diesen neuen Modellkopf in Ihrer Sequenzklassifizierungsaufgabe feinabstimmen, indem Sie das Wissen des vortrainierten Modells auf ihn übertragen.
+
+</Tip>
+
+### Hyperparameter für das Training
+
+Als Nächstes erstellen Sie eine Klasse [`TrainingArguments`], die alle Hyperparameter enthält, die Sie einstellen können, sowie Flags zur Aktivierung verschiedener Trainingsoptionen. Für dieses Lernprogramm können Sie mit den Standard- [Hyperparametern](https://huggingface.co/docs/transformers/main_classes/trainer#transformers.TrainingArguments) beginnen, aber Sie können mit diesen experimentieren, um Ihre optimalen Einstellungen zu finden.
+
+Geben Sie an, wo die Kontrollpunkte Ihres Trainings gespeichert werden sollen:
+
+```py
+>>> from transformers import TrainingArguments
+
+>>> training_args = TrainingArguments(output_dir="test_trainer")
+```
+
+### Auswerten
+
+Der [`Trainer`] wertet die Leistung des Modells während des Trainings nicht automatisch aus. Sie müssen [`Trainer`] eine Funktion übergeben, um Metriken zu berechnen und zu berichten. Die [🤗 Evaluate](https://huggingface.co/docs/evaluate/index) Bibliothek bietet eine einfache [`accuracy`](https://huggingface.co/spaces/evaluate-metric/accuracy) Funktion, die Sie mit der [`evaluate.load`] Funktion laden können (siehe diese [quicktour](https://huggingface.co/docs/evaluate/a_quick_tour) für weitere Informationen):
+
+```py
+>>> import numpy as np
+>>> import evaluate
+
+>>> metric = evaluate.load("accuracy")
+```
+
+Rufen Sie [`~evaluate.compute`] auf `metric` auf, um die Genauigkeit Ihrer Vorhersagen zu berechnen. Bevor Sie Ihre Vorhersagen an `compute` übergeben, müssen Sie die Vorhersagen in Logits umwandeln (denken Sie daran, dass alle 🤗 Transformers-Modelle Logits zurückgeben):
+
+```py
+>>> def compute_metrics(eval_pred):
+...     logits, labels = eval_pred
+...     predictions = np.argmax(logits, axis=-1)
+...     return metric.compute(predictions=predictions, references=labels)
+```
+
+Wenn Sie Ihre Bewertungsmetriken während der Feinabstimmung überwachen möchten, geben Sie den Parameter `evaluation_strategy` in Ihren Trainingsargumenten an, um die Bewertungsmetrik am Ende jeder Epoche zu ermitteln:
+
+```py
+>>> from transformers import TrainingArguments, Trainer
+
+>>> training_args = TrainingArguments(output_dir="test_trainer", evaluation_strategy="epoch")
+```
+
+### Trainer
+
+Erstellen Sie ein [`Trainer`]-Objekt mit Ihrem Modell, Trainingsargumenten, Trainings- und Testdatensätzen und einer Evaluierungsfunktion:
+
+```py
+>>> trainer = Trainer(
+...     model=model,
+...     args=training_args,
+...     train_dataset=small_train_dataset,
+...     eval_dataset=small_eval_dataset,
+...     compute_metrics=compute_metrics,
+... )
+```
+
+Anschließend können Sie Ihr Modell durch den Aufruf von [`~transformers.Trainer.train`] optimieren:
+
+```py
+>>> trainer.train()
+```
+</pt>
+<tf>
+<a id='keras'></a>
+
+<Youtube id="rnTGBy2ax1c"/>
+
+## Trainieren Sie ein TensorFlow-Modell mit Keras
+
+Sie können auch 🤗 Transformers Modelle in TensorFlow mit der Keras API trainieren!
+
+### Laden von Daten für Keras
+
+Wenn Sie ein 🤗 Transformers Modell mit der Keras API trainieren wollen, müssen Sie Ihren Datensatz in ein Format konvertieren, das
+Keras versteht. Wenn Ihr Datensatz klein ist, können Sie das Ganze einfach in NumPy-Arrays konvertieren und an Keras übergeben.
+Probieren wir das zuerst aus, bevor wir etwas Komplizierteres tun.
+
+Laden Sie zunächst ein Dataset. Wir werden den CoLA-Datensatz aus dem [GLUE-Benchmark](https://huggingface.co/datasets/glue) verwenden,
+da es sich um eine einfache Aufgabe zur Klassifizierung von binärem Text handelt, und nehmen vorerst nur den Trainingssplit.
+
+```py
+from datasets import load_dataset
+
+dataset = load_dataset("glue", "cola")
+dataset = dataset["train"]  # Just take the training split for now
+```
+
+Als nächstes laden Sie einen Tokenizer und tokenisieren die Daten als NumPy-Arrays. Beachten Sie, dass die Beschriftungen bereits eine Liste von 0 und 1en sind,
+Wir können sie also ohne Tokenisierung direkt in ein NumPy-Array konvertieren!
+
+```py
+from transformers import AutoTokenizer
+
+tokenizer = AutoTokenizer.from_pretrained("bert-base-cased")
+tokenized_data = tokenizer(dataset["text"], return_tensors="np", padding=True)
+
+labels = np.array(dataset["label"])  # Label is already an array of 0 and 1
+```
+
+Schließlich laden, [`compile`](https://keras.io/api/models/model_training_apis/#compile-method) und [`fit`](https://keras.io/api/models/model_training_apis/#fit-method) Sie das Modell:
+
+```py
+from transformers import TFAutoModelForSequenceClassification
+from tensorflow.keras.optimizers import Adam
+
+# Load and compile our model
+model = TFAutoModelForSequenceClassification.from_pretrained("bert-base-cased")
+# Lower learning rates are often better for fine-tuning transformers
+model.compile(optimizer=Adam(3e-5))
+
+model.fit(tokenized_data, labels)
+```
+
+<Tip>
+
+Sie müssen Ihren Modellen kein Verlustargument übergeben, wenn Sie sie `compile()`! Hugging-Face-Modelle wählen automatisch
+einen Loss, der für ihre Aufgabe und Modellarchitektur geeignet ist, wenn dieses Argument leer gelassen wird. Sie können jederzeit außer Kraft setzen, indem Sie selbst einen Loss angeben, wenn Sie das möchten!
+
+</Tip>
+
+Dieser Ansatz eignet sich hervorragend für kleinere Datensätze, aber bei größeren Datensätzen kann er zu einem Problem werden. Warum?
+Weil das tokenisierte Array und die Beschriftungen vollständig in den Speicher geladen werden müssten, und weil NumPy nicht mit
+"gezackte" Arrays nicht verarbeiten kann, so dass jedes tokenisierte Sample auf die Länge des längsten Samples im gesamten Datensatz aufgefüllt werden müsste.
+Datensatzes aufgefüllt werden. Dadurch wird das Array noch größer, und all die aufgefüllten Token verlangsamen auch das Training!
+
+### Laden von Daten als tf.data.Dataset
+
+Wenn Sie eine Verlangsamung des Trainings vermeiden wollen, können Sie Ihre Daten stattdessen als `tf.data.Dataset` laden. Sie können zwar Ihre eigene
+tf.data"-Pipeline schreiben können, wenn Sie wollen, haben wir zwei bequeme Methoden, um dies zu tun:
+
+- [`~TFPreTrainedModel.prepare_tf_dataset`]: Dies ist die Methode, die wir in den meisten Fällen empfehlen. Da es sich um eine Methode
+Ihres Modells ist, kann sie das Modell inspizieren, um automatisch herauszufinden, welche Spalten als Modelleingaben verwendet werden können, und
+verwirft die anderen, um einen einfacheren, leistungsfähigeren Datensatz zu erstellen.
+- [~datasets.Dataset.to_tf_dataset`]: Diese Methode ist eher auf niedriger Ebene angesiedelt und ist nützlich, wenn Sie genau kontrollieren wollen, wie
+Dataset erstellt wird, indem man genau angibt, welche `columns` und `label_cols` einbezogen werden sollen.
+
+Bevor Sie [~TFPreTrainedModel.prepare_tf_dataset`] verwenden können, müssen Sie die Tokenizer-Ausgaben als Spalten zu Ihrem Datensatz hinzufügen, wie in
+dem folgenden Codebeispiel:
+
+```py
+def tokenize_dataset(data):
+    # Keys of the returned dictionary will be added to the dataset as columns
+    return tokenizer(data["text"])
+
+
+dataset = dataset.map(tokenize_dataset)
+```
+
+Denken Sie daran, dass Hugging Face-Datensätze standardmäßig auf der Festplatte gespeichert werden, so dass dies nicht zu einem erhöhten Arbeitsspeicherbedarf führen wird! Sobald die
+Spalten hinzugefügt wurden, können Sie Batches aus dem Datensatz streamen und zu jedem Batch Auffüllungen hinzufügen, was die Anzahl der Auffüllungs-Token im Vergleich zum Auffüllen des gesamten Datensatzes reduziert.
+
+
+```py
+>>> tf_dataset = model.prepare_tf_dataset(dataset, batch_size=16, shuffle=True, tokenizer=tokenizer)
+```
+
+Beachten Sie, dass Sie im obigen Codebeispiel den Tokenizer an `prepare_tf_dataset` übergeben müssen, damit die Stapel beim Laden korrekt aufgefüllt werden können.
+Wenn alle Stichproben in Ihrem Datensatz die gleiche Länge haben und kein Auffüllen erforderlich ist, können Sie dieses Argument weglassen.
+Wenn Sie etwas Komplexeres als nur das Auffüllen von Stichproben benötigen (z. B. das Korrumpieren von Token für die maskierte Sprachmodellierung), können Sie das Argument
+Modellierung), können Sie stattdessen das Argument `collate_fn` verwenden, um eine Funktion zu übergeben, die aufgerufen wird, um die
+Liste von Stichproben in einen Stapel umwandelt und alle gewünschten Vorverarbeitungen vornimmt. Siehe unsere
+[examples](https://github.com/huggingface/transformers/tree/main/examples) oder
+[notebooks](https://huggingface.co/docs/transformers/notebooks), um diesen Ansatz in Aktion zu sehen.
+
+Sobald Sie einen `tf.data.Dataset` erstellt haben, können Sie das Modell wie zuvor kompilieren und anpassen:
+
+```py
+model.compile(optimizer=Adam(3e-5))
+
+model.fit(tf_dataset)
+```
+
+</tf>
+</frameworkcontent>
+
+<a id='pytorch_native'></a>
+
+## Trainieren in nativem PyTorch
+
+<frameworkcontent>
+<pt>
+<Youtube id="Dh9CL8fyG80"/>
+
+[`Trainer`] kümmert sich um die Trainingsschleife und ermöglicht die Feinabstimmung eines Modells in einer einzigen Codezeile. Für Benutzer, die es vorziehen, ihre eigene Trainingsschleife zu schreiben, können Sie auch eine Feinabstimmung eines 🤗 Transformers-Modells in nativem PyTorch vornehmen.
+
+An diesem Punkt müssen Sie möglicherweise Ihr Notebook neu starten oder den folgenden Code ausführen, um etwas Speicher freizugeben:
+
+```py
+del model
+del pytorch_model
+del trainer
+torch.cuda.empty_cache()
+```
+
+Als Nächstes müssen Sie den Datensatz `tokenized_dataset` manuell nachbearbeiten, um ihn für das Training vorzubereiten.
+
+1. Entfernen Sie die Spalte "Text", da das Modell keinen Rohtext als Eingabe akzeptiert:
+
+    ```py
+    >>> tokenized_datasets = tokenized_datasets.remove_columns(["text"])
+    ```
+
+2. Benennen Sie die Spalte "Label" in "Labels" um, da das Modell erwartet, dass das Argument "Labels" genannt wird:
+
+    ```py
+    >>> tokenized_datasets = tokenized_datasets.rename_column("label", "labels")
+    ```
+
+3. Stellen Sie das Format des Datensatzes so ein, dass PyTorch-Tensoren anstelle von Listen zurückgegeben werden:
+
+    ```py
+    >>> tokenized_datasets.set_format("torch")
+    ```
+
+Erstellen Sie dann eine kleinere Teilmenge des Datensatzes, wie zuvor gezeigt, um die Feinabstimmung zu beschleunigen:
+
+```py
+>>> small_train_dataset = tokenized_datasets["train"].shuffle(seed=42).select(range(1000))
+>>> small_eval_dataset = tokenized_datasets["test"].shuffle(seed=42).select(range(1000))
+```
+
+### DataLoader
+
+Erstellen Sie einen `DataLoader` für Ihre Trainings- und Testdatensätze, damit Sie über die Datenstapel iterieren können:
+
+```py
+>>> from torch.utils.data import DataLoader
+
+>>> train_dataloader = DataLoader(small_train_dataset, shuffle=True, batch_size=8)
+>>> eval_dataloader = DataLoader(small_eval_dataset, batch_size=8)
+```
+
+Laden Sie Ihr Modell mit der Anzahl der erwarteten Kennzeichnungen:
+
+```py
+>>> from transformers import AutoModelForSequenceClassification
+
+>>> model = AutoModelForSequenceClassification.from_pretrained("bert-base-cased", num_labels=5)
+```
+
+### Optimierer und Lernratensteuerung
+
+Erstellen Sie einen Optimierer und einen Scheduler für die Lernrate, um das Modell fein abzustimmen. Wir verwenden den Optimierer [`AdamW`](https://pytorch.org/docs/stable/generated/torch.optim.AdamW.html) aus PyTorch:
+
+```py
+>>> from torch.optim import AdamW
+
+>>> optimizer = AdamW(model.parameters(), lr=5e-5)
+```
+
+Erstellen Sie den Standard-Lernratenplaner aus [`Trainer`]:
+
+```py
+>>> from transformers import get_scheduler
+
+>>> num_epochs = 3
+>>> num_training_steps = num_epochs * len(train_dataloader)
+>>> lr_scheduler = get_scheduler(
+...     name="linear", optimizer=optimizer, num_warmup_steps=0, num_training_steps=num_training_steps
+... )
+```
+
+Geben Sie schließlich `device` an, um einen Grafikprozessor zu verwenden, wenn Sie Zugang zu einem solchen haben. Andernfalls kann das Training auf einer CPU mehrere Stunden statt ein paar Minuten dauern.
+
+```py
+>>> import torch
+
+>>> device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
+>>> model.to(device)
+```
+
+<Tip>
+
+Holen Sie sich mit einem gehosteten Notebook wie [Colaboratory](https://colab.research.google.com/) oder [SageMaker StudioLab](https://studiolab.sagemaker.aws/) kostenlosen Zugang zu einem Cloud-GPU, wenn Sie noch keinen haben.
+
+</Tip>
+
+Großartig, Sie sind bereit für das Training! 🥳 
+
+### Trainingsschleife
+
+Um Ihren Trainingsfortschritt zu verfolgen, verwenden Sie die [tqdm](https://tqdm.github.io/) Bibliothek, um einen Fortschrittsbalken über die Anzahl der Trainingsschritte hinzuzufügen:
+
+```py
+>>> from tqdm.auto import tqdm
+
+>>> progress_bar = tqdm(range(num_training_steps))
+
+>>> model.train()
+>>> for epoch in range(num_epochs):
+...     for batch in train_dataloader:
+...         batch = {k: v.to(device) for k, v in batch.items()}
+...         outputs = model(**batch)
+...         loss = outputs.loss
+...         loss.backward()
+
+...         optimizer.step()
+...         lr_scheduler.step()
+...         optimizer.zero_grad()
+...         progress_bar.update(1)
+```
+
+### Auswertung
+
+Genauso wie Sie eine Bewertungsfunktion zu [`Trainer`] hinzugefügt haben, müssen Sie dasselbe tun, wenn Sie Ihre eigene Trainingsschleife schreiben. Aber anstatt die Metrik am Ende jeder Epoche zu berechnen und zu melden, werden Sie dieses Mal alle Stapel mit [`~evaluate.add_batch`] akkumulieren und die Metrik ganz am Ende berechnen.
+
+```py
+>>> import evaluate
+
+>>> metric = evaluate.load("accuracy")
+>>> model.eval()
+>>> for batch in eval_dataloader:
+...     batch = {k: v.to(device) for k, v in batch.items()}
+...     with torch.no_grad():
+...         outputs = model(**batch)
+
+...     logits = outputs.logits
+...     predictions = torch.argmax(logits, dim=-1)
+...     metric.add_batch(predictions=predictions, references=batch["labels"])
+
+>>> metric.compute()
+```
+</pt>
+</frameworkcontent>
+
+<a id='additional-resources'></a>
+
+## Zusätzliche Ressourcen
+
+Weitere Beispiele für die Feinabstimmung finden Sie unter:
+
+- [🤗 Transformers Examples](https://github.com/huggingface/transformers/tree/main/examples) enthält Skripte
+  um gängige NLP-Aufgaben in PyTorch und TensorFlow zu trainieren.
+
+- [🤗 Transformers Notebooks](notebooks) enthält verschiedene Notebooks zur Feinabstimmung eines Modells für bestimmte Aufgaben in PyTorch und TensorFlow.

docs/source/en/_toctree.yml

@@ -31,9 +31,11 @@
     - local: sagemaker
       title: Run training on Amazon SageMaker
     - local: converting_tensorflow_models
-      title: Converting TensorFlow Checkpoints
+      title: Converting from TensorFlow checkpoints
     - local: serialization
-      title: Export 🤗 Transformers models
+      title: Export to ONNX
+    - local: torchscript
+      title: Export to TorchScript
     - local: troubleshooting
       title: Troubleshoot
     title: General usage
@@ -101,12 +103,16 @@
       title: Instantiating a big model
     - local: debugging
       title: Debugging
+    - local: hpo_train
+      title: Hyperparameter Search using Trainer API
     title: Performance and scalability
   - sections:
     - local: contributing
       title: How to contribute to transformers?
     - local: add_new_model
       title: How to add a model to 🤗 Transformers?
+    - local: add_tensorflow_model
+      title: How to convert a 🤗 Transformers model to TensorFlow?
     - local: add_new_pipeline
       title: How to add a pipeline to 🤗 Transformers?
     - local: testing
@@ -143,6 +149,8 @@
   title: Conceptual guides
 - sections:
   - sections:
+    - local: model_doc/auto
+      title: Auto Classes
     - local: main_classes/callback
       title: Callbacks
     - local: main_classes/configuration
@@ -175,10 +183,10 @@
       title: DeepSpeed Integration
     - local: main_classes/feature_extractor
       title: Feature Extractor
+    - local: main_classes/image_processor
+      title: Image Processor
     title: Main Classes
   - sections:
-    - local: model_doc/auto
-      title: Auto Classes
     - isExpanded: false
       sections:
       - local: model_doc/albert
@@ -239,6 +247,10 @@
         title: Encoder Decoder Models
       - local: model_doc/ernie
         title: ERNIE
+      - local: model_doc/esm
+        title: ESM
+      - local: model_doc/flan-t5
+        title: FLAN-T5
       - local: model_doc/flaubert
         title: FlauBERT
       - local: model_doc/fnet
@@ -253,6 +265,8 @@
         title: GPT Neo
       - local: model_doc/gpt_neox
         title: GPT NeoX
+      - local: model_doc/gpt_neox_japanese
+        title: GPT NeoX Japanese
       - local: model_doc/gptj
         title: GPT-J
       - local: model_doc/gpt2
@@ -261,10 +275,14 @@
         title: HerBERT
       - local: model_doc/ibert
         title: I-BERT
+      - local: model_doc/jukebox
+        title: Jukebox
       - local: model_doc/layoutlm
         title: LayoutLM
       - local: model_doc/led
         title: LED
+      - local: model_doc/lilt
+        title: LiLT
       - local: model_doc/longformer
         title: Longformer
       - local: model_doc/longt5
@@ -275,6 +293,8 @@
         title: M2M100
       - local: model_doc/marian
         title: MarianMT
+      - local: model_doc/markuplm
+        title: MarkupLM
       - local: model_doc/mbart
         title: MBart and MBart-50
       - local: model_doc/megatron-bert
@@ -323,12 +343,16 @@
         title: RetriBERT
       - local: model_doc/roberta
         title: RoBERTa
+      - local: model_doc/roc_bert
+        title: RoCBert
       - local: model_doc/roformer
         title: RoFormer
       - local: model_doc/splinter
         title: Splinter
       - local: model_doc/squeezebert
         title: SqueezeBERT
+      - local: model_doc/switch_transformers
+        title: SwitchTransformers
       - local: model_doc/t5
         title: T5
       - local: model_doc/t5v1.1
@@ -360,14 +384,20 @@
       sections:
       - local: model_doc/beit
         title: BEiT
+      - local: model_doc/conditional_detr
+        title: Conditional DETR
       - local: model_doc/convnext
         title: ConvNeXT
       - local: model_doc/cvt
         title: CvT
+      - local: model_doc/deformable_detr
+        title: Deformable DETR
       - local: model_doc/deit
         title: DeiT
       - local: model_doc/detr
         title: DETR
+      - local: model_doc/dinat
+        title: DiNAT
       - local: model_doc/dit
         title: DiT
       - local: model_doc/dpt
@@ -380,8 +410,14 @@
         title: LeViT
       - local: model_doc/maskformer
         title: MaskFormer
+      - local: model_doc/mobilenet_v1
+        title: MobileNetV1
+      - local: model_doc/mobilenet_v2
+        title: MobileNetV2
       - local: model_doc/mobilevit
         title: MobileViT
+      - local: model_doc/nat
+        title: NAT
       - local: model_doc/poolformer
         title: PoolFormer
       - local: model_doc/regnet
@@ -394,6 +430,8 @@
         title: Swin Transformer
       - local: model_doc/swinv2
         title: Swin Transformer V2
+      - local: model_doc/table-transformer
+        title: Table Transformer
       - local: model_doc/van
         title: VAN
       - local: model_doc/videomae
@@ -402,11 +440,15 @@
         title: Vision Transformer (ViT)
       - local: model_doc/vit_mae
         title: ViTMAE
+      - local: model_doc/vit_msn
+        title: ViTMSN
       - local: model_doc/yolos
         title: YOLOS
       title: Vision models
     - isExpanded: false
       sections:
+      - local: model_doc/audio-spectrogram-transformer
+        title: Audio Spectrogram Transformer
       - local: model_doc/hubert
         title: Hubert
       - local: model_doc/mctct
@@ -431,6 +473,8 @@
         title: Wav2Vec2Phoneme
       - local: model_doc/wavlm
         title: WavLM
+      - local: model_doc/whisper
+        title: Whisper
       - local: model_doc/xls_r
         title: XLS-R
       - local: model_doc/xlsr_wav2vec2
@@ -438,8 +482,12 @@
       title: Audio models
     - isExpanded: false
       sections:
+      - local: model_doc/chinese_clip
+        title: Chinese-CLIP
       - local: model_doc/clip
         title: CLIP
+      - local: model_doc/clipseg
+        title: CLIPSeg
       - local: model_doc/data2vec
         title: Data2Vec
       - local: model_doc/donut
@@ -482,6 +530,11 @@
       - local: model_doc/trajectory_transformer
         title: Trajectory Transformer
       title: Reinforcement learning models
+    - isExpanded: false
+      sections:
+      - local: model_doc/time_series_transformer
+        title: Time Series Transformer
+      title: Time series models
     title: Models
   - sections:
     - local: internal/modeling_utils
@@ -494,7 +547,9 @@
       title: Utilities for Trainer
     - local: internal/generation_utils
       title: Utilities for Generation
+    - local: internal/image_processing_utils
+      title: Utilities for Image Processors
     - local: internal/file_utils
       title: General Utilities
     title: Internal Helpers
-  title: API
+  title: API

docs/source/en/add_new_model.mdx

@@ -11,32 +11,26 @@ an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express o
 
 # How to add a model to 🤗 Transformers?
 
-Adding a new model is often difficult and requires an in-depth knowledge of the 🤗 Transformers library and ideally also
-of the model's original repository. At Hugging Face, we are trying to empower the community more and more to add models
-independently. Thus, for some new models that the community wants to be added to 🤗 Transformers, we create a customized
-*call-for-model-addition* that explains step-by-step how to add the requested model. With this
-*call-for-model-addition*, we want to teach a motivated and experienced contributor of the community how to port a
-model to 🤗 Transformers.
+The 🤗 Transformers library is often able to offer new models thanks to community contributors. But this can be a challenging project and requires an in-depth knowledge of the 🤗 Transformers library and the model to implement. At Hugging Face, we're trying to empower more of the community to actively add models and we've put together this guide to walk you through the process of adding a PyTorch model (make sure you have [PyTorch installed](https://pytorch.org/get-started/locally/)).
 
-If this sounds like something you would be interested in, feel free to check out the currently open
-“calls-for-model-addition” [here](https://github.com/huggingface/transformers/tree/main/templates/adding_a_new_model/open_model_proposals/README.md)
-and to contact us.
+<Tip>
 
-If selected, you will then work closely with one member of the Hugging Face team to integrate the model into 🤗
-Transformers. By doing so, you will both gain a theoretical and deep practical understanding of the proposed model. But
-more importantly, you will have made a major open-source contribution to 🤗 Transformers. Along the way, you will:
+If you're interested in implementing a TensorFlow model, take a look at the [How to convert a 🤗 Transformers model to TensorFlow](add_tensorflow_model) guide!
 
--  get insights into open-source best practices
--  understand the design principles of one of the most popular NLP libraries
--  learn how to do efficiently test large NLP models
--  learn how to integrate Python utilities like `black`, `isort`, `make fix-copies` into a library to always
-  ensure clean and readable code
+</Tip>
 
-We are also more than happy if you want to add a model that cannot be found in the “calls-for-model-addition” folder.
-The following sections explain in detail how to add a new model. It might also be very helpful to check out already
-added models to see if those resemble the model you would like to add [here](https://github.com/huggingface/transformers/pulls?q=is%3Apr+label%3A%22PR+for+Model+Addition%22+is%3Aclosed).
+Along the way, you'll:
 
-To start, let's try to get a general overview of the Transformers library.
+- get insights into open-source best practices
+- understand the design principles behind one of the most popular deep learning libraries
+- learn how to efficiently test large models
+- learn how to integrate Python utilities like `black`, `isort`, and `make fix-copies` to ensure clean and readable code
+
+A Hugging Face team member will be available to help you along the way so you'll never be alone. 🤗 ❤️
+
+To get started, open a [New model addition](https://github.com/huggingface/transformers/issues/new?assignees=&labels=New+model&template=new-model-addition.yml) issue for the model you want to see in 🤗 Transformers. If you're not especially picky about contributing a specific model, you can filter by the [New model label](https://github.com/huggingface/transformers/labels/New%20model) to see if there are any unclaimed model requests and work on it.
+
+Once you've opened a new model request, the first step is to get familiar with 🤗 Transformers if you aren't already!
 
 ## General overview of 🤗 Transformers
 
@@ -106,7 +100,7 @@ own regarding how code should be written :-)
    for a good example).
 2. The code should be fully understandable, even by a non-native English speaker. This means you should pick
    descriptive variable names and avoid abbreviations. As an example, `activation` is preferred to `act`.
-   One-letter variable names are strongly discouraged unless it's an index in a for loop. 
+   One-letter variable names are strongly discouraged unless it's an index in a for loop.
 3. More generally we prefer longer explicit code to short magical one.
 4. Avoid subclassing `nn.Sequential` in PyTorch but subclass `nn.Module` and write the forward pass, so that anyone
    using your code can quickly debug it by adding print statements or breaking points.
@@ -144,20 +138,20 @@ In the following, we try to give you a general recipe that we found most useful
 The following list is a summary of everything that has to be done to add a model and can be used by you as a To-Do
 List:
 
--  1. ☐ (Optional) Understood theoretical aspects
--  2. ☐ Prepared transformers dev environment
--  3. ☐ Set up debugging environment of the original repository
--  4. ☐ Created script that successfully runs forward pass using original repository and checkpoint
--  5. ☐ Successfully added the model skeleton to Transformers
--  6. ☐ Successfully converted original checkpoint to Transformers checkpoint
--  7. ☐ Successfully ran forward pass in Transformers that gives identical output to original checkpoint
--  8. ☐ Finished model tests in Transformers
--  9. ☐ Successfully added Tokenizer in Transformers
--  10. ☐ Run end-to-end integration tests
--  11. ☐ Finished docs
--  12. ☐ Uploaded model weights to the hub
--  13. ☐ Submitted the pull request
--  14. ☐ (Optional) Added a demo notebook
+☐ (Optional) Understood the model's theoretical aspects<br>
+☐ Prepared 🤗 Transformers dev environment<br>
+☐ Set up debugging environment of the original repository<br>
+☐ Created script that successfully runs the `forward()` pass using the original repository and checkpoint<br>
+☐ Successfully added the model skeleton to 🤗 Transformers<br>
+☐ Successfully converted original checkpoint to 🤗 Transformers checkpoint<br>
+☐ Successfully ran `forward()` pass in 🤗 Transformers that gives identical output to original checkpoint<br>
+☐ Finished model tests in 🤗 Transformers<br>
+☐ Successfully added tokenizer in 🤗 Transformers<br>
+☐ Run end-to-end integration tests<br>
+☐ Finished docs<br>
+☐ Uploaded model weights to the Hub<br>
+☐ Submitted the pull request<br>
+☐ (Optional) Added a demo notebook
 
 To begin with, we usually recommend to start by getting a good theoretical understanding of `BrandNewBert`. However,
 if you prefer to understand the theoretical aspects of the model *on-the-job*, then it is totally fine to directly dive
@@ -222,7 +216,7 @@ cd ..
 5. To port *brand_new_bert*, you will also need access to its original repository:
 
 ```bash
-git clone https://github.com/org_that_created_brand_new_bert_org/brand_new_bert.git 
+git clone https://github.com/org_that_created_brand_new_bert_org/brand_new_bert.git
 cd brand_new_bert
 pip install -e .
 ```
@@ -683,10 +677,11 @@ work left to be done should be a cakewalk 😊.
 At this point, you have successfully added a new model. However, it is very much possible that the model does not yet
 fully comply with the required design. To make sure, the implementation is fully compatible with 🤗 Transformers, all
 common tests should pass. The Cookiecutter should have automatically added a test file for your model, probably under
-the same `tests/test_modeling_brand_new_bert.py`. Run this test file to verify that all common tests pass:
+the same `tests/models/brand_new_bert/test_modeling_brand_new_bert.py`. Run this test file to verify that all common
+tests pass:
 
 ```bash
-pytest tests/test_modeling_brand_new_bert.py
+pytest tests/models/brand_new_bert/test_modeling_brand_new_bert.py
 ```
 
 Having fixed all common tests, it is now crucial to ensure that all the nice work you have done is well tested, so that
@@ -700,7 +695,7 @@ Cookiecutter, called `BrandNewBertModelIntegrationTests` and only has to be fill
 tests are passing, run
 
 ```bash
-RUN_SLOW=1 pytest -sv tests/test_modeling_brand_new_bert.py::BrandNewBertModelIntegrationTests
+RUN_SLOW=1 pytest -sv tests/models/brand_new_bert/test_modeling_brand_new_bert.py::BrandNewBertModelIntegrationTests
 ```
 
 <Tip>
@@ -758,7 +753,8 @@ contain a couple of hard-coded integration tests.
 **10. Run End-to-end integration tests**
 
 Having added the tokenizer, you should also add a couple of end-to-end integration tests using both the model and the
-tokenizer to `tests/test_modeling_brand_new_bert.py` in 🤗 Transformers. Such a test should show on a meaningful
+tokenizer to `tests/models/brand_new_bert/test_modeling_brand_new_bert.py` in 🤗 Transformers.
+Such a test should show on a meaningful
 text-to-text sample that the 🤗 Transformers implementation works as expected. A meaningful text-to-text sample can
 include *e.g.* a source-to-target-translation pair, an article-to-summary pair, a question-to-answer pair, etc… If none
 of the ported checkpoints has been fine-tuned on a downstream task it is enough to simply rely on the model tests. In a
@@ -771,7 +767,7 @@ tests for you.
 
 Now, all the necessary functionality for *brand_new_bert* is added - you're almost done! The only thing left to add is
 a nice docstring and a doc page. The Cookiecutter should have added a template file called
-`docs/source/model_doc/brand_new_bert.rst` that you should fill out. Users of your model will usually first look at
+`docs/source/model_doc/brand_new_bert.mdx` that you should fill out. Users of your model will usually first look at
 this page before using your model. Hence, the documentation must be understandable and concise. It is very useful for
 the community to add some *Tips* to show how the model should be used. Don't hesitate to ping the Hugging Face team
 regarding the docstrings.

docs/source/en/add_tensorflow_model.mdx

@@ -0,0 +1,346 @@
+<!--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
+-->
+
+# How to convert a 🤗 Transformers model to TensorFlow?
+
+Having multiple frameworks available to use with 🤗 Transformers gives you flexibility to play their strengths when
+designing your application, but it implies that compatibility must be added on a per-model basis. The good news is that
+adding TensorFlow compatibility to an existing model is simpler than [adding a new model from scratch](add_new_model)!
+Whether you wish to have a deeper understanding of large TensorFlow models, make a major open-source contribution, or
+enable TensorFlow for your model of choice, this guide is for you.
+
+This guide empowers you, a member of our community, to contribute TensorFlow model weights and/or
+architectures to be used in 🤗 Transformers, with minimal supervision from the Hugging Face team. Writing a new model
+is no small feat, but hopefully this guide will make it less of a rollercoaster 🎢 and more of a walk in the park 🚶.
+Harnessing our collective experiences is absolutely critical to make this process increasingly easier, and thus we
+highly encourage that you suggest improvements to this guide!
+
+Before you dive deeper, it is recommended that you check the following resources if you're new to 🤗 Transformers:
+- [General overview of 🤗 Transformers](add_new_model#general-overview-of-transformers)
+- [Hugging Face's TensorFlow Philosophy](https://huggingface.co/blog/tensorflow-philosophy)
+
+In the remainder of this guide, you will learn what's needed to add a new TensorFlow model architecture, the
+procedure to convert PyTorch into TensorFlow model weights, and how to efficiently debug mismatches across ML
+frameworks. Let's get started!
+
+<Tip>
+
+Are you unsure whether the model you wish to use already has a corresponding TensorFlow architecture?
+
+&nbsp;
+
+Check the `model_type` field of the `config.json` of your model of choice
+([example](https://huggingface.co/bert-base-uncased/blob/main/config.json#L14)). If the corresponding model folder in
+🤗 Transformers has a file whose name starts with "modeling_tf", it means that it has a corresponding TensorFlow
+architecture ([example](https://github.com/huggingface/transformers/tree/main/src/transformers/models/bert)).
+
+</Tip>
+
+
+## Step-by-step guide to add TensorFlow model architecture code
+
+There are many ways to design a large model architecture, and multiple ways of implementing said design. However,
+you might recall from our [general overview of 🤗 Transformers](add_new_model#general-overview-of-transformers)
+that we are an opinionated bunch - the ease of use of 🤗 Transformers relies on consistent design choices. From
+experience, we can tell you a few important things about adding TensorFlow models:
+
+- Don't reinvent the wheel! More often that not, there are at least two reference implementations you should check: the
+PyTorch equivalent of the model you are implementing and other TensorFlow models for the same class of problems.
+- Great model implementations survive the test of time. This doesn't happen because the code is pretty, but rather
+because the code is clear, easy to debug and build upon. If you make the life of the maintainers easy with your
+TensorFlow implementation, by replicating the same patterns as in other TensorFlow models and minimizing the mismatch
+to the PyTorch implementation, you ensure your contribution will be long lived.
+- Ask for help when you're stuck! The 🤗 Transformers team is here to help, and we've probably found solutions to the same
+problems you're facing.
+
+Here's an overview of the steps needed to add a TensorFlow model architecture:
+1. Select the model you wish to convert
+2. Prepare transformers dev environment
+3. (Optional) Understand theoretical aspects and the existing implementation
+4. Implement the model architecture
+5. Implement model tests
+6. Submit the pull request
+7. (Optional) Build demos and share with the world
+
+### 1.-3. Prepare your model contribution
+
+**1. Select the model you wish to convert**
+
+Let's start off with the basics: the first thing you need to know is the architecture you want to convert. If you
+don't have your eyes set on a specific architecture, asking the 🤗 Transformers team for suggestions is a great way to
+maximize your impact - we will guide you towards the most prominent architectures that are missing on the TensorFlow
+side. If the specific model you want to use with TensorFlow already has a TensorFlow architecture implementation in
+🤗 Transformers but is lacking weights, feel free to jump straight into the
+[weight conversion section](#adding-tensorflow-weights-to-hub)
+of this page.
+
+For simplicity, the remainder of this guide assumes you've decided to contribute with the TensorFlow version of
+*BrandNewBert* (the same example as in the [guide](add_new_model) to add a new model from scratch).
+
+<Tip>
+
+Before starting the work on a TensorFlow model architecture, double-check that there is no ongoing effort to do so.
+You can search for `BrandNewBert` on the
+[pull request GitHub page](https://github.com/huggingface/transformers/pulls?q=is%3Apr) to confirm that there is no
+TensorFlow-related pull request.
+
+</Tip>
+
+
+**2. Prepare transformers dev environment**
+
+Having selected the model architecture, open an draft PR to signal your intention to work on it. Follow the
+instructions below to set up your environment and open a draft PR.
+
+1. Fork the [repository](https://github.com/huggingface/transformers) by clicking on the 'Fork' button on the
+   repository's page. This creates a copy of the code under your GitHub user account.
+
+2. Clone your `transformers` fork to your local disk, and add the base repository as a remote:
+
+```bash
+git clone https://github.com/[your Github handle]/transformers.git
+cd transformers
+git remote add upstream https://github.com/huggingface/transformers.git
+```
+
+3. Set up a development environment, for instance by running the following command:
+
+```bash
+python -m venv .env
+source .env/bin/activate
+pip install -e ".[dev]"
+```
+
+**Note:** You don't need to have CUDA installed. Making the new model work on CPU is sufficient.
+
+4. Create a branch with a descriptive name from your main branch
+
+```bash
+git checkout -b add_tf_brand_new_bert
+```
+
+5. Fetch and rebase to current main
+
+```bash
+git fetch upstream
+git rebase upstream/main
+```
+
+6. Add an empty `.py` file in `transformers/src/models/brandnewbert/` named `modeling_tf_brandnewbert.py`. This will
+be your TensorFlow model file.
+
+7. Push the changes to your account using:
+
+```bash
+git add .
+git commit -m "initial commit"
+git push -u origin add_tf_brand_new_bert
+```
+
+8. Once you are satisfied, go to the webpage of your fork on GitHub. Click on “Pull request”. Make sure to add the
+   GitHub handle of some members of the Hugging Face team as reviewers, so that the Hugging Face team gets notified for
+   future changes.
+
+9. Change the PR into a draft by clicking on “Convert to draft” on the right of the GitHub pull request web page.
+
+
+Now you have set up a development environment to port *BrandNewBert* to TensorFlow in 🤗 Transformers.
+
+
+**3. (Optional) Understand theoretical aspects and the existing implementation**
+
+You should take some time to read *BrandNewBert's* paper, if such descriptive work exists. There might be large
+sections of the paper that are difficult to understand. If this is the case, this is fine - don't worry! The goal is
+not to get a deep theoretical understanding of the paper, but to extract the necessary information required to
+effectively re-implement the model in 🤗 Transformers using TensorFlow. That being said, you don't have to spend too
+much time on the theoretical aspects, but rather focus on the practical ones, namely the existing model documentation
+page (e.g. [model docs for BERT](model_doc/bert)).
+
+After you've grasped the basics of the models you are about to implement, it's important to understand the existing
+implementation. This is a great chance to confirm that a working implementation matches your expectations for the
+model, as well as to foresee technical challenges on the TensorFlow side.
+
+It's perfectly natural that you feel overwhelmed with the amount of information that you've just absorbed. It is
+definitely not a requirement that you understand all facets of the model at this stage. Nevertheless, we highly
+encourage you to clear any pressing questions in our [forum](https://discuss.huggingface.co/).
+
+
+### 4. Model implementation
+
+Now it's time to finally start coding. Our suggested starting point is the PyTorch file itself: copy the contents of
+`modeling_brand_new_bert.py` inside `src/transformers/models/brand_new_bert/` into
+`modeling_tf_brand_new_bert.py`. The goal of this section is to modify the file and update the import structure of
+🤗 Transformers such that you can import `TFBrandNewBert` and
+`TFBrandNewBert.from_pretrained(model_repo, from_pt=True)` successfully loads a working TensorFlow *BrandNewBert* model.
+
+Sadly, there is no prescription to convert a PyTorch model into TensorFlow. You can, however, follow our selection of
+tips to make the process as smooth as possible:
+- Prepend `TF` to the name of all classes (e.g. `BrandNewBert` becomes `TFBrandNewBert`).
+- Most PyTorch operations have a direct TensorFlow replacement. For example, `torch.nn.Linear` corresponds to
+  `tf.keras.layers.Dense`, `torch.nn.Dropout` corresponds to `tf.keras.layers.Dropout`, etc. If you're not sure
+  about a specific operation, you can use the [TensorFlow documentation](https://www.tensorflow.org/api_docs/python/tf)
+  or the [PyTorch documentation](https://pytorch.org/docs/stable/).
+- Look for patterns in the 🤗 Transformers codebase. If you come across a certain operation that doesn't have a direct
+   replacement, the odds are that someone else already had the same problem.
+- By default, keep the same variable names and structure as in PyTorch. This will make it easier to debug, track
+   issues, and add fixes down the line.
+- Some layers have different default values in each framework. A notable example is the batch normalization layer's
+   epsilon (`1e-5` in [PyTorch](https://pytorch.org/docs/stable/generated/torch.nn.BatchNorm2d.html#torch.nn.BatchNorm2d)
+   and `1e-3` in [TensorFlow](https://www.tensorflow.org/api_docs/python/tf/keras/layers/BatchNormalization)).
+   Double-check the documentation!
+- PyTorch's `nn.Parameter` variables typically need to be initialized within TF Layer's `build()`. See the following
+   example: [PyTorch](https://github.com/huggingface/transformers/blob/655f72a6896c0533b1bdee519ed65a059c2425ac/src/transformers/models/vit_mae/modeling_vit_mae.py#L212) /
+   [TensorFlow](https://github.com/huggingface/transformers/blob/655f72a6896c0533b1bdee519ed65a059c2425ac/src/transformers/models/vit_mae/modeling_tf_vit_mae.py#L220)
+- If the PyTorch model has a `#copied from ...` on top of a function, the odds are that your TensorFlow model can also
+   borrow that function from the architecture it was copied from, assuming it has a TensorFlow architecture.
+- Assigning the `name` attribute correctly in TensorFlow functions is critical to do the `from_pt=True` weight
+   cross-loading. `name` is almost always the name of the corresponding variable in the PyTorch code. If `name` is not
+   properly set, you will see it in the error message when loading the model weights.
+- The logic of the base model class, `BrandNewBertModel`, will actually reside in `TFBrandNewBertMainLayer`, a Keras
+   layer subclass ([example](https://github.com/huggingface/transformers/blob/4fd32a1f499e45f009c2c0dea4d81c321cba7e02/src/transformers/models/bert/modeling_tf_bert.py#L719)).
+   `TFBrandNewBertModel` will simply be a wrapper around this layer.
+- Keras models need to be built in order to load pretrained weights. For that reason, `TFBrandNewBertPreTrainedModel`
+   will need to hold an example of inputs to the model, the `dummy_inputs`
+   ([example](https://github.com/huggingface/transformers/blob/4fd32a1f499e45f009c2c0dea4d81c321cba7e02/src/transformers/models/bert/modeling_tf_bert.py#L916)).
+- If you get stuck, ask for help - we're here to help you! 🤗
+
+In addition to the model file itself, you will also need to add the pointers to the model classes and related
+documentation pages. You can complete this part entirely following the patterns in other PRs
+([example](https://github.com/huggingface/transformers/pull/18020/files)). Here's a list of the needed manual
+changes:
+- Include all public classes of *BrandNewBert* in `src/transformers/__init__.py`
+- Add *BrandNewBert* classes to the corresponding Auto classes in `src/transformers/models/auto/modeling_tf_auto.py`
+- Include the modeling file in the documentation test file list in `utils/documentation_tests.txt`
+- Add the lazy loading classes related to *BrandNewBert* in `src/transformers/utils/dummy_tf_objects.py`
+- Update the import structures for the public classes in `src/transformers/models/brand_new_bert/__init__.py`
+- Add the documentation pointers to the public methods of *BrandNewBert* in `docs/source/en/model_doc/brand_new_bert.mdx`
+- Add yourself to the list of contributors to *BrandNewBert* in `docs/source/en/model_doc/brand_new_bert.mdx`
+- Finally, add a green tick ✅ to the TensorFlow column of *BrandNewBert* in `docs/source/en/index.mdx`
+
+When you're happy with your implementation, run the following checklist to confirm that your model architecture is
+ready:
+1. All layers that behave differently at train time (e.g. Dropout) are called with a `training` argument, which is
+propagated all the way from the top-level classes
+2. You have used `#copied from ...` whenever possible
+3. `TFBrandNewBertMainLayer` and all classes that use it have their `call` function decorated with `@unpack_inputs`
+4. `TFBrandNewBertMainLayer` is decorated with `@keras_serializable`
+5. A TensorFlow model can be loaded from PyTorch weights using `TFBrandNewBert.from_pretrained(model_repo, from_pt=True)`
+6. You can call the TensorFlow model using the expected input format
+
+
+### 5. Add model tests
+
+Hurray, you've implemented a TensorFlow model! Now it's time to add tests to make sure that your model behaves as
+expected. As in the previous section, we suggest you start by copying the `test_modeling_brand_new_bert.py` file in
+`tests/models/brand_new_bert/` into `test_modeling_tf_brand_new_bert.py`, and continue by making the necessary
+TensorFlow replacements. For now, in all `.from_pretrained()` calls, you should use the `from_pt=True` flag to load
+the existing PyTorch weights.
+
+After you're done, it's time for the moment of truth: run the tests! 😬
+
+```bash
+NVIDIA_TF32_OVERRIDE=0 RUN_SLOW=1 RUN_PT_TF_CROSS_TESTS=1 \
+py.test -vv tests/models/brand_new_bert/test_modeling_tf_brand_new_bert.py
+```
+
+The most likely outcome is that you'll see a bunch of errors. Don't worry, this is expected! Debugging ML models is
+notoriously hard, and the key ingredient to success is patience (and `breakpoint()`). In our experience, the hardest
+problems arise from subtle mismatches between ML frameworks, for which we have a few pointers at the end of this guide.
+In other cases, a general test might not be directly applicable to your model, in which case we suggest an override
+at the model test class level. Regardless of the issue, don't hesitate to ask for help in your draft pull request if
+you're stuck.
+
+When all tests pass, congratulations, your model is nearly ready to be added to the 🤗 Transformers library! 🎉
+
+### 6.-7. Ensure everyone can use your model
+
+**6. Submit the pull request**
+
+Once you're done with the implementation and the tests, it's time to submit a pull request. Before pushing your code,
+run our code formatting utility, `make fixup` 🪄. This will automatically fix any formatting issues, which would cause
+our automatic checks to fail.
+
+It's now time to convert your draft pull request into a real pull request. To do so, click on the "Ready for
+review" button and add Joao (`@gante`) and Matt (`@Rocketknight1`) as reviewers. A model pull request will need
+at least 3 reviewers, but they will take care of finding appropriate additional reviewers for your model.
+
+After all reviewers are happy with the state of your PR, the final action point is to remove the `from_pt=True` flag in
+`.from_pretrained()` calls. Since there are no TensorFlow weights, you will have to add them! Check the section
+below for instructions on how to do it.
+
+Finally, when the TensorFlow weights get merged, you have at least 3 reviewer approvals, and all CI checks are
+green, double-check the tests locally one last time
+
+```bash
+NVIDIA_TF32_OVERRIDE=0 RUN_SLOW=1 RUN_PT_TF_CROSS_TESTS=1 \
+py.test -vv tests/models/brand_new_bert/test_modeling_tf_brand_new_bert.py
+```
+
+and we will merge your PR! Congratulations on the milestone 🎉
+
+**7. (Optional) Build demos and share with the world**
+
+One of the hardest parts about open-source is discovery. How can the other users learn about the existence of your
+fabulous TensorFlow contribution? With proper communication, of course! 📣
+
+There are two main ways to share your model with the community:
+- Build demos. These include Gradio demos, notebooks, and other fun ways to show off your model. We highly
+   encourage you to add a notebook to our [community-driven demos](https://huggingface.co/docs/transformers/community).
+- Share stories on social media like Twitter and LinkedIn. You should be proud of your work and share
+   your achievement with the community - your model can now be used by thousands of engineers and researchers around
+   the world 🌍! We will be happy to retweet your posts and help you share your work with the community.
+
+
+## Adding TensorFlow weights to 🤗 Hub
+
+Assuming that the TensorFlow model architecture is available in 🤗 Transformers, converting PyTorch weights into
+TensorFlow weights is a breeze!
+
+Here's how to do it:
+1. Make sure you are logged into your Hugging Face account in your terminal. You can log in using the command
+   `huggingface-cli login` (you can find your access tokens [here](https://huggingface.co/settings/tokens))
+2. Run `transformers-cli pt-to-tf --model-name foo/bar`, where `foo/bar` is the name of the model repository
+   containing the PyTorch weights you want to convert
+3. Tag `@joaogante` and `@Rocketknight1` in the 🤗 Hub PR the command above has just created
+
+That's it! 🎉
+
+
+## Debugging mismatches across ML frameworks 🐛
+
+At some point, when adding a new architecture or when creating TensorFlow weights for an existing architecture, you
+might come across errors compaining about mismatches between PyTorch and TensorFlow. You might even decide to open the
+model architecture code for the two frameworks, and find that they look identical. What's going on? 🤔
+
+First of all, let's talk about why understanding these mismatches matters. Many community members will use 🤗
+Transformers models out of the box, and trust that our models behave as expected. When there is a large mismatch
+between the two frameworks, it implies that the model is not following the reference implementation for at least one
+of the frameworks. This might lead to silent failures, in which the model runs but has poor performance. This is
+arguably worse than a model that fails to run at all! To that end, we aim at having a framework mismatch smaller than
+`1e-5` at all stages of the model.
+
+As in other numerical problems, the devil is in the details. And as in any detail-oriented craft, the secret
+ingredient here is patience. Here is our suggested workflow for when you come across this type of issues:
+1. Locate the source of mismatches. The model you're converting probably has near identical inner variables up to a
+   certain point. Place `breakpoint()` statements in the two frameworks' architectures, and compare the values of the
+   numerical variables in a top-down fashion until you find the source of the problems.
+2. Now that you've pinpointed the source of the issue, get in touch with the 🤗 Transformers team. It is possible
+   that we've seen a similar problem before and can promptly provide a solution. As a fallback, scan popular pages
+   like StackOverflow and GitHub issues.
+3. If there is no solution in sight, it means you'll have to go deeper. The good news is that you've located the
+   issue, so you can focus on the problematic instruction, abstracting away the rest of the model! The bad news is
+   that you'll have to venture into the source implementation of said instruction. In some cases, you might find an
+   issue with a reference implementation - don't abstain from opening an issue in the upstream repository.
+
+In some cases, in dicussion with the 🤗 Transformers team, we might find that the fixing the mismatch is infeasible.
+When the mismatch is very small in the output layers of the model (but potentially large in the hidden states), we
+might decide to ignore it in favor of distributing the model. The `pt-to-tf` CLI mentioned above has a `--max-error`
+flag to override the error message at weight conversion time.

docs/source/en/autoclass_tutorial.mdx

@@ -23,6 +23,7 @@ Remember, architecture refers to the skeleton of the model and checkpoints are t
 In this tutorial, learn to:
 
 * Load a pretrained tokenizer.
+* Load a pretrained image processor
 * Load a pretrained feature extractor.
 * Load a pretrained processor.
 * Load a pretrained model.
@@ -49,9 +50,20 @@ Then tokenize your input as shown below:
  'attention_mask': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]}
 ```
 
+## AutoImageProcessor
+
+For vision tasks, an image processor processes the image into the correct input format.
+
+```py
+>>> from transformers import AutoImageProcessor
+
+>>> image_processor = AutoImageProcessor.from_pretrained("google/vit-base-patch16-224")
+```
+
+
 ## AutoFeatureExtractor
 
-For audio and vision tasks, a feature extractor processes the audio signal or image into the correct input format.
+For audio tasks, a feature extractor processes the audio signal the correct input format.
 
 Load a feature extractor with [`AutoFeatureExtractor.from_pretrained`]:
 
@@ -65,7 +77,7 @@ Load a feature extractor with [`AutoFeatureExtractor.from_pretrained`]:
 
 ## AutoProcessor
 
-Multimodal tasks require a processor that combines two types of preprocessing tools. For example, the [LayoutLMV2](model_doc/layoutlmv2) model requires a feature extractor to handle images and a tokenizer to handle text; a processor combines both of them.
+Multimodal tasks require a processor that combines two types of preprocessing tools. For example, the [LayoutLMV2](model_doc/layoutlmv2) model requires an image processor to handle images and a tokenizer to handle text; a processor combines both of them.
 
 Load a processor with [`AutoProcessor.from_pretrained`]:
 
@@ -103,7 +115,7 @@ TensorFlow and Flax checkpoints are not affected, and can be loaded within PyTor
 
 </Tip>
 
-Generally, we recommend using the `AutoTokenizer` class and the `AutoModelFor` class to load pretrained instances of models. This will ensure you load the correct architecture every time. In the next [tutorial](preprocessing), learn how to use your newly loaded tokenizer, feature extractor and processor to preprocess a dataset for fine-tuning.
+Generally, we recommend using the `AutoTokenizer` class and the `AutoModelFor` class to load pretrained instances of models. This will ensure you load the correct architecture every time. In the next [tutorial](preprocessing), learn how to use your newly loaded tokenizer, image processor, feature extractor and processor to preprocess a dataset for fine-tuning.
 </pt>
 <tf>
 Finally, the `TFAutoModelFor` classes let you load a pretrained model for a given task (see [here](model_doc/auto) for a complete list of available tasks). For example, load a model for sequence classification with [`TFAutoModelForSequenceClassification.from_pretrained`]:
@@ -122,6 +134,6 @@ Easily reuse the same checkpoint to load an architecture for a different task:
 >>> model = TFAutoModelForTokenClassification.from_pretrained("distilbert-base-uncased")
 ```
 
-Generally, we recommend using the `AutoTokenizer` class and the `TFAutoModelFor` class to load pretrained instances of models. This will ensure you load the correct architecture every time. In the next [tutorial](preprocessing), learn how to use your newly loaded tokenizer, feature extractor and processor to preprocess a dataset for fine-tuning.
+Generally, we recommend using the `AutoTokenizer` class and the `TFAutoModelFor` class to load pretrained instances of models. This will ensure you load the correct architecture every time. In the next [tutorial](preprocessing), learn how to use your newly loaded tokenizer, image processor, feature extractor and processor to preprocess a dataset for fine-tuning.
 </tf>
 </frameworkcontent>

docs/source/en/big_models.mdx

@@ -72,7 +72,7 @@ On top of the configuration of the model, we see three different weights files,
 
 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:
+Behind 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
@@ -86,7 +86,7 @@ Beind the scenes, the index file is used to determine which keys are in the chec
 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:
+The metadata just consists of the total size of the model for now. We plan to add other information in the future:
 
 ```py
 >>> index["metadata"]

docs/source/en/converting_tensorflow_models.mdx

@@ -10,7 +10,7 @@ an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express o
 specific language governing permissions and limitations under the License.
 -->
 
-# Converting Tensorflow Checkpoints
+# Converting From Tensorflow Checkpoints
 
 A command-line interface is provided to convert original Bert/GPT/GPT-2/Transformer-XL/XLNet/XLM checkpoints to models
 that can be loaded using the `from_pretrained` methods of the library.

docs/source/en/create_a_model.mdx

@@ -17,7 +17,8 @@ An [`AutoClass`](model_doc/auto) automatically infers the model architecture and
 - Load and customize a model configuration.
 - Create a model architecture.
 - Create a slow and fast tokenizer for text.
-- Create a feature extractor for audio or image tasks.
+- Create an image processor for vision tasks.
+- Create a feature extractor for audio tasks.
 - Create a processor for multimodal tasks.
 
 ## Configuration
@@ -244,21 +245,21 @@ By default, [`AutoTokenizer`] will try to load a fast tokenizer. You can disable
 
 </Tip>
 
-## Feature Extractor
+## Image Processor
 
-A feature extractor processes audio or image inputs. It inherits from the base [`~feature_extraction_utils.FeatureExtractionMixin`] class, and may also inherit from the [`ImageFeatureExtractionMixin`] class for processing image features or the [`SequenceFeatureExtractor`] class for processing audio inputs.
+An image processor processes vision inputs. It inherits from the base [`~image_processing_utils.ImageProcessingMixin`] class.
 
-Depending on whether you are working on an audio or vision task, create a feature extractor associated with the model you're using. For example, create a default [`ViTFeatureExtractor`] if you are using [ViT](model_doc/vit) for image classification:
+To use, create an image processor associated with the model you're using. For example, create a default [`ViTImageProcessor`] if you are using [ViT](model_doc/vit) for image classification:
 
 ```py
->>> from transformers import ViTFeatureExtractor
+>>> from transformers import ViTImageProcessor
 
->>> vit_extractor = ViTFeatureExtractor()
+>>> vit_extractor = ViTImageProcessor()
 >>> print(vit_extractor)
-ViTFeatureExtractor {
+ViTImageProcessor {
   "do_normalize": true,
   "do_resize": true,
-  "feature_extractor_type": "ViTFeatureExtractor",
+  "feature_extractor_type": "ViTImageProcessor",
   "image_mean": [
     0.5,
     0.5,
@@ -276,21 +277,21 @@ ViTFeatureExtractor {
 
 <Tip>
 
-If you aren't looking for any customization, just use the `from_pretrained` method to load a model's default feature extractor parameters.
+If you aren't looking for any customization, just use the `from_pretrained` method to load a model's default image processor parameters.
 
 </Tip>
 
-Modify any of the [`ViTFeatureExtractor`] parameters to create your custom feature extractor:
+Modify any of the [`ViTImageProcessor`] parameters to create your custom image processor:
 
 ```py
->>> from transformers import ViTFeatureExtractor
+>>> from transformers import ViTImageProcessor
 
->>> my_vit_extractor = ViTFeatureExtractor(resample="PIL.Image.BOX", do_normalize=False, image_mean=[0.3, 0.3, 0.3])
+>>> my_vit_extractor = ViTImageProcessor(resample="PIL.Image.BOX", do_normalize=False, image_mean=[0.3, 0.3, 0.3])
 >>> print(my_vit_extractor)
-ViTFeatureExtractor {
+ViTImageProcessor {
   "do_normalize": false,
   "do_resize": true,
-  "feature_extractor_type": "ViTFeatureExtractor",
+  "feature_extractor_type": "ViTImageProcessor",
   "image_mean": [
     0.3,
     0.3,
@@ -306,7 +307,11 @@ ViTFeatureExtractor {
 }
 ```
 
-For audio inputs, you can create a [`Wav2Vec2FeatureExtractor`] and customize the parameters in a similar way:
+## Feature Extractor
+
+A feature extractor processes audio inputs. It inherits from the base [`~feature_extraction_utils.FeatureExtractionMixin`] class, and may also inherit from the [`SequenceFeatureExtractor`] class for processing audio inputs.
+
+To use, create a feature extractor associated with the model you're using. For example, create a default [`Wav2Vec2FeatureExtractor`] if you are using [Wav2Vec2](model_doc/wav2vec2) for audio classification:
 
 ```py
 >>> from transformers import Wav2Vec2FeatureExtractor
@@ -324,9 +329,34 @@ Wav2Vec2FeatureExtractor {
 }
 ```
 
+<Tip>
+
+If you aren't looking for any customization, just use the `from_pretrained` method to load a model's default feature extractor parameters.
+
+</Tip>
+
+Modify any of the [`Wav2Vec2FeatureExtractor`] parameters to create your custom feature extractor:
+
+```py
+>>> from transformers import Wav2Vec2FeatureExtractor
+
+>>> w2v2_extractor = Wav2Vec2FeatureExtractor(sampling_rate=8000, do_normalize=False)
+>>> print(w2v2_extractor)
+Wav2Vec2FeatureExtractor {
+  "do_normalize": false,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "return_attention_mask": false,
+  "sampling_rate": 8000
+}
+```
+
+
 ## Processor
 
-For models that support multimodal tasks, 🤗 Transformers offers a processor class that conveniently wraps a feature extractor and tokenizer into a single object. For example, let's use the [`Wav2Vec2Processor`] for an automatic speech recognition task (ASR). ASR transcribes audio to text, so you will need a feature extractor and a tokenizer.
+For models that support multimodal tasks, 🤗 Transformers offers a processor class that conveniently wraps processing classes such as a feature extractor and a tokenizer into a single object. For example, let's use the [`Wav2Vec2Processor`] for an automatic speech recognition task (ASR). ASR transcribes audio to text, so you will need a feature extractor and a tokenizer.
 
 Create a feature extractor to handle the audio inputs:
 
@@ -352,4 +382,4 @@ Combine the feature extractor and tokenizer in [`Wav2Vec2Processor`]:
 >>> processor = Wav2Vec2Processor(feature_extractor=feature_extractor, tokenizer=tokenizer)
 ```
 
-With two basic classes - configuration and model - and an additional preprocessing class (tokenizer, feature extractor, or processor), you can create any of the models supported by 🤗 Transformers. Each of these base classes are configurable, allowing you to use the specific attributes you want. You can easily setup a model for training or modify an existing pretrained model to fine-tune.
+With two basic classes - configuration and model - and an additional preprocessing class (tokenizer, image processor, feature extractor, or processor), you can create any of the models supported by 🤗 Transformers. Each of these base classes are configurable, allowing you to use the specific attributes you want. You can easily setup a model for training or modify an existing pretrained model to fine-tune.

docs/source/en/custom_models.mdx

@@ -21,7 +21,7 @@ with the community (with the code it relies on) so that anyone can use it, even
 Transformers library.
 
 We will illustrate all of this on a ResNet model, by wrapping the ResNet class of the
-[timm library](https://github.com/rwightman/pytorch-image-models/tree/master/timm) into a [`PreTrainedModel`].
+[timm library](https://github.com/rwightman/pytorch-image-models) into a [`PreTrainedModel`].
 
 ## Writing a custom configuration
 
@@ -55,9 +55,9 @@ class ResnetConfig(PretrainedConfig):
         **kwargs,
     ):
         if block_type not in ["basic", "bottleneck"]:
-            raise ValueError(f"`block` must be 'basic' or bottleneck', got {block}.")
+            raise ValueError(f"`block_type` must be 'basic' or bottleneck', got {block_type}.")
         if stem_type not in ["", "deep", "deep-tiered"]:
-            raise ValueError(f"`stem_type` must be '', 'deep' or 'deep-tiered', got {block}.")
+            raise ValueError(f"`stem_type` must be '', 'deep' or 'deep-tiered', got {stem_type}.")
 
         self.block_type = block_type
         self.layers = layers
@@ -146,6 +146,9 @@ class ResnetModel(PreTrainedModel):
 For the model that will classify images, we just change the forward method:
 
 ```py
+import torch
+
+
 class ResnetModelForImageClassification(PreTrainedModel):
     config_class = ResnetConfig
 

docs/source/en/debugging.mdx

@@ -77,7 +77,7 @@ to the normal command line arguments, or pass `debug="underflow_overflow"` when
 If you're using your own training loop or another Trainer you can accomplish the same with:
 
 ```python
-from .debug_utils import DebugUnderflowOverflow
+from transformers.debug_utils import DebugUnderflowOverflow
 
 debug_overflow = DebugUnderflowOverflow(model)
 ```
@@ -271,7 +271,7 @@ Additionally, if you're instantiating the debugger in your own code, you can adj
 its default, e.g.:
 
 ```python
-from .debug_utils import DebugUnderflowOverflow
+from transformers.debug_utils import DebugUnderflowOverflow
 
 debug_overflow = DebugUnderflowOverflow(model, max_frames_to_save=100)
 ```

docs/source/en/glossary.mdx

@@ -12,110 +12,12 @@ specific language governing permissions and limitations under the License.
 
 # Glossary
 
-## General terms
+This glossary defines general machine learning and 🤗 Transformers terms to help you better understand the
+documentation.
 
-- autoencoding models: see MLM
-- autoregressive models: see CLM
-- CLM: causal language modeling, a pretraining task where the model reads the texts in order and has to predict the
-  next word. It's usually done by reading the whole sentence but using a mask inside the model to hide the future
-  tokens at a certain timestep.
-- deep learning: machine learning algorithms which uses neural networks with several layers.
-- MLM: masked language modeling, a pretraining task where the model sees a corrupted version of the texts, usually done
-  by masking some tokens randomly, and has to predict the original text.
-- multimodal: a task that combines texts with another kind of inputs (for instance images).
-- NLG: natural language generation, all tasks related to generating text (for instance talk with transformers,
-  translation).
-- NLP: natural language processing, a generic way to say "deal with texts".
-- NLU: natural language understanding, all tasks related to understanding what is in a text (for instance classifying
-  the whole text, individual words).
-- pretrained model: a model that has been pretrained on some data (for instance all of Wikipedia). Pretraining methods
-  involve a self-supervised objective, which can be reading the text and trying to predict the next word (see CLM) or
-  masking some words and trying to predict them (see MLM).
-- RNN: recurrent neural network, a type of model that uses a loop over a layer to process texts.
-- self-attention: each element of the input finds out which other elements of the input they should attend to.
-- seq2seq or sequence-to-sequence: models that generate a new sequence from an input, like translation models, or
-  summarization models (such as [Bart](model_doc/bart) or [T5](model_doc/t5)).
-- token: a part of a sentence, usually a word, but can also be a subword (non-common words are often split in subwords)
-  or a punctuation symbol.
-- transformer: self-attention based deep learning model architecture.
+## A
 
-## Model inputs
-
-Every model is different yet bears similarities with the others. Therefore most models use the same inputs, which are
-detailed here alongside usage examples.
-
-
-
-### Input IDs
-
-The input ids are often the only required parameters to be passed to the model as input. *They are token indices,
-numerical representations of tokens building the sequences that will be used as input by the model*.
-
-<Youtube id="VFp38yj8h3A"/>
-
-Each tokenizer works differently but the underlying mechanism remains the same. Here's an example using the BERT
-tokenizer, which is a [WordPiece](https://arxiv.org/pdf/1609.08144.pdf) tokenizer:
-
-```python
->>> from transformers import BertTokenizer
-
->>> tokenizer = BertTokenizer.from_pretrained("bert-base-cased")
-
->>> sequence = "A Titan RTX has 24GB of VRAM"
-```
-
-The tokenizer takes care of splitting the sequence into tokens available in the tokenizer vocabulary.
-
-```python
->>> tokenized_sequence = tokenizer.tokenize(sequence)
-```
-
-The tokens are either words or subwords. Here for instance, "VRAM" wasn't in the model vocabulary, so it's been split
-in "V", "RA" and "M". To indicate those tokens are not separate words but parts of the same word, a double-hash prefix
-is added for "RA" and "M":
-
-```python
->>> print(tokenized_sequence)
-['A', 'Titan', 'R', '##T', '##X', 'has', '24', '##GB', 'of', 'V', '##RA', '##M']
-```
-
-These tokens can then be converted into IDs which are understandable by the model. This can be done by directly feeding
-the sentence to the tokenizer, which leverages the Rust implementation of [🤗 Tokenizers](https://github.com/huggingface/tokenizers) for peak performance.
-
-```python
->>> inputs = tokenizer(sequence)
-```
-
-The tokenizer returns a dictionary with all the arguments necessary for its corresponding model to work properly. The
-token indices are under the key "input_ids":
-
-```python
->>> encoded_sequence = inputs["input_ids"]
->>> print(encoded_sequence)
-[101, 138, 18696, 155, 1942, 3190, 1144, 1572, 13745, 1104, 159, 9664, 2107, 102]
-```
-
-Note that the tokenizer automatically adds "special tokens" (if the associated model relies on them) which are special
-IDs the model sometimes uses.
-
-If we decode the previous sequence of ids,
-
-```python
->>> decoded_sequence = tokenizer.decode(encoded_sequence)
-```
-
-we will see
-
-```python
->>> print(decoded_sequence)
-[CLS] A Titan RTX has 24GB of VRAM [SEP]
-```
-
-because this is the way a [`BertModel`] is going to expect its inputs.
-
-
-
-### Attention mask
+### attention mask
 
 The attention mask is an optional argument used when batching sequences together.
 
@@ -162,26 +64,310 @@ We can see that 0s have been added on the right of the first sentence to make it
 ```
 
 This can then be converted into a tensor in PyTorch or TensorFlow. The attention mask is a binary tensor indicating the
-position of the padded indices so that the model does not attend to them. For the [`BertTokenizer`],
-`1` indicates a value that should be attended to, while `0` indicates a padded value. This attention mask is
-in the dictionary returned by the tokenizer under the key "attention_mask":
+position of the padded indices so that the model does not attend to them. For the [`BertTokenizer`], `1` indicates a
+value that should be attended to, while `0` indicates a padded value. This attention mask is in the dictionary returned
+by the tokenizer under the key "attention_mask":
 
 ```python
 >>> padded_sequences["attention_mask"]
 [[1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]]
 ```
 
+### autoencoding models 
+
+see [masked language modeling](#masked-language-modeling)
+
+### autoregressive models
+
+see [causal language modeling](#causal-language-modeling)
+
+## B
+
+### backbone
+
+The backbone is the network (embeddings and layers) that outputs the raw hidden states or features. It is usually connected to a [head](#head) which accepts the features as its input to make a prediction. For example, [`ViTModel`] is a backbone without a specific head on top. Other models can also use [`VitModel`] as a backbone such as [DPT](model_doc/dpt).
+
+## C
+
+### channel
+
+Color images are made up of some combination of values in three channels - red, green, and blue (RGB) - and grayscale images only have one channel. In 🤗 Transformers, the channel can be the first or last dimension of an image's tensor: [`n_channels`, `height`, `width`] or [`height`, `width`, `n_channels`].
+
+### causal language modeling
+
+A pretraining task where the model reads the texts in order and has to predict the next word. It's usually done by
+reading the whole sentence but using a mask inside the model to hide the future tokens at a certain timestep.
+
+### connectionist temporal classification (CTC)
+
+An algorithm which allows a model to learn without knowing exactly how the input and output are aligned; CTC calculates the distribution of all possible outputs for a given input and chooses the most likely output from it. CTC is commonly used in speech recognition tasks because speech doesn't always cleanly align with the transcript for a variety of reasons such as a speaker's different speech rates.
+
+### convolution
+
+A type of layer in a neural network where the input matrix is multiplied element-wise by a smaller matrix (kernel or filter) and the values are summed up in a new matrix. This is known as a convolutional operation which is repeated over the entire input matrix. Each operation is applied to a different segment of the input matrix. Convolutional neural networks (CNNs) are commonly used in computer vision.
+
+## D
+
+### decoder input IDs
+
+This input is specific to encoder-decoder models, and contains the input IDs that will be fed to the decoder. These
+inputs should be used for sequence to sequence tasks, such as translation or summarization, and are usually built in a
+way specific to each model.
+
+Most encoder-decoder models (BART, T5) create their `decoder_input_ids` on their own from the `labels`. In such models,
+passing the `labels` is the preferred way to handle training.
+
+Please check each model's docs to see how they handle these input IDs for sequence to sequence training.
+
+### deep learning
+
+Machine learning algorithms which uses neural networks with several layers.
+
+## F
+
+### feed forward chunking
+
+In each residual attention block in transformers the self-attention layer is usually followed by 2 feed forward layers.
+The intermediate embedding size of the feed forward layers is often bigger than the hidden size of the model (e.g., for
+`bert-base-uncased`).
+
+For an input of size `[batch_size, sequence_length]`, the memory required to store the intermediate feed forward
+embeddings `[batch_size, sequence_length, config.intermediate_size]` can account for a large fraction of the memory
+use. The authors of [Reformer: The Efficient Transformer](https://arxiv.org/abs/2001.04451) noticed that since the
+computation is independent of the `sequence_length` dimension, it is mathematically equivalent to compute the output
+embeddings of both feed forward layers `[batch_size, config.hidden_size]_0, ..., [batch_size, config.hidden_size]_n`
+individually and concat them afterward to `[batch_size, sequence_length, config.hidden_size]` with `n =
+sequence_length`, which trades increased computation time against reduced memory use, but yields a mathematically
+**equivalent** result.
+
+For models employing the function [`apply_chunking_to_forward`], the `chunk_size` defines the number of output
+embeddings that are computed in parallel and thus defines the trade-off between memory and time complexity. If
+`chunk_size` is set to 0, no feed forward chunking is done.
+
+## H
+
+### head
+
+The model head refers to the last layer of a neural network that accepts the raw hidden states and projects them onto a different dimension. There is a different model head for each task. For example:
+
+  * [`GPT2ForSequenceClassification`] is a sequence classification head - a linear layer - on top of the base [`GPT2Model`].
+  * [`ViTForImageClassification`] is an image classification head - a linear layer on top of the final hidden state of the `CLS` token - on top of the base [`ViTModel`].
+  * [`Wav2Vec2ForCTC`] ia a language modeling head with [CTC](#connectionist-temporal-classification-(CTC)) on top of the base [`Wav2Vec2Model`].
+
+## I
+
+### image patch
+
+Vision-based Transformers models split an image into smaller patches which are linearly embedded, and then passed as a sequence to the model. You can find the `patch_size` - or resolution - of the model in it's configuration.
+
+### input IDs
+
+The input ids are often the only required parameters to be passed to the model as input. They are token indices,
+numerical representations of tokens building the sequences that will be used as input by the model.
+
+<Youtube id="VFp38yj8h3A"/>
+
+Each tokenizer works differently but the underlying mechanism remains the same. Here's an example using the BERT
+tokenizer, which is a [WordPiece](https://arxiv.org/pdf/1609.08144.pdf) tokenizer:
+
+```python
+>>> from transformers import BertTokenizer
+
+>>> tokenizer = BertTokenizer.from_pretrained("bert-base-cased")
+
+>>> sequence = "A Titan RTX has 24GB of VRAM"
+```
+
+The tokenizer takes care of splitting the sequence into tokens available in the tokenizer vocabulary.
+
+```python
+>>> tokenized_sequence = tokenizer.tokenize(sequence)
+```
+
+The tokens are either words or subwords. Here for instance, "VRAM" wasn't in the model vocabulary, so it's been split
+in "V", "RA" and "M". To indicate those tokens are not separate words but parts of the same word, a double-hash prefix
+is added for "RA" and "M":
+
+```python
+>>> print(tokenized_sequence)
+['A', 'Titan', 'R', '##T', '##X', 'has', '24', '##GB', 'of', 'V', '##RA', '##M']
+```
+
+These tokens can then be converted into IDs which are understandable by the model. This can be done by directly feeding
+the sentence to the tokenizer, which leverages the Rust implementation of [🤗
+Tokenizers](https://github.com/huggingface/tokenizers) for peak performance.
+
+```python
+>>> inputs = tokenizer(sequence)
+```
+
+The tokenizer returns a dictionary with all the arguments necessary for its corresponding model to work properly. The
+token indices are under the key `input_ids`:
+
+```python
+>>> encoded_sequence = inputs["input_ids"]
+>>> print(encoded_sequence)
+[101, 138, 18696, 155, 1942, 3190, 1144, 1572, 13745, 1104, 159, 9664, 2107, 102]
+```
+
+Note that the tokenizer automatically adds "special tokens" (if the associated model relies on them) which are special
+IDs the model sometimes uses.
+
+If we decode the previous sequence of ids,
+
+```python
+>>> decoded_sequence = tokenizer.decode(encoded_sequence)
+```
+
+we will see
+
+```python
+>>> print(decoded_sequence)
+[CLS] A Titan RTX has 24GB of VRAM [SEP]
+```
+
+because this is the way a [`BertModel`] is going to expect its inputs.
+
+## L
+
+### labels
+
+The labels are an optional argument which can be passed in order for the model to compute the loss itself. These labels
+should be the expected prediction of the model: it will use the standard loss in order to compute the loss between its
+predictions and the expected value (the label).
+
+These labels are different according to the model head, for example:
+
+- For sequence classification models, ([`BertForSequenceClassification`]), the model expects a tensor of dimension
+  `(batch_size)` with each value of the batch corresponding to the expected label of the entire sequence.
+- For token classification models, ([`BertForTokenClassification`]), the model expects a tensor of dimension
+  `(batch_size, seq_length)` with each value corresponding to the expected label of each individual token.
+- For masked language modeling, ([`BertForMaskedLM`]), the model expects a tensor of dimension `(batch_size,
+  seq_length)` with each value corresponding to the expected label of each individual token: the labels being the token
+  ID for the masked token, and values to be ignored for the rest (usually -100).
+- For sequence to sequence tasks, ([`BartForConditionalGeneration`], [`MBartForConditionalGeneration`]), the model
+  expects a tensor of dimension `(batch_size, tgt_seq_length)` with each value corresponding to the target sequences
+  associated with each input sequence. During training, both BART and T5 will make the appropriate
+  `decoder_input_ids` and decoder attention masks internally. They usually do not need to be supplied. This does not
+  apply to models leveraging the Encoder-Decoder framework.
+- For image classification models, ([`ViTForImageClassification`]), the model expects a tensor of dimension
+  `(batch_size)` with each value of the batch corresponding to the expected label of each individual image.
+- For semantic segmentation models, ([`SegformerForSemanticSegmentation`]), the model expects a tensor of dimension
+  `(batch_size, height, width)` with each value of the batch corresponding to the expected label of each individual pixel.
+- For object detection models, ([`DetrForObjectDetection`]), the model expects a list of dictionaries with a
+  `class_labels` and `boxes` key where each value of the batch corresponds to the expected label and number of bounding boxes of each individual image.
+- For automatic speech recognition models, ([`Wav2Vec2ForCTC`]), the model expects a tensor of dimension `(batch_size,
+  target_length)` with each value corresponding to the expected label of each individual token.
+  
+<Tip>
+
+Each model's labels may be different, so be sure to always check the documentation of each model for more information
+about their specific labels!
+
+</Tip>
+
+The base models ([`BertModel`]) do not accept labels, as these are the base transformer models, simply outputting
+features.
+
+## M
+
+### masked language modeling
+
+A pretraining task where the model sees a corrupted version of the texts, usually done by
+masking some tokens randomly, and has to predict the original text.
+
+### multimodal
+
+A task that combines texts with another kind of inputs (for instance images).
+
+## N
+
+### Natural language generation
+
+All tasks related to generating text (for instance talk with transformers, translation).
+
+### Natural language processing
+
+A generic way to say "deal with texts".
+
+### Natural language understanding
+
+All tasks related to understanding what is in a text (for instance classifying the
+whole text, individual words).
+
+## P
+
+### pixel values
+
+A tensor of the numerical representations of an image that is passed to a model. The pixel values have a shape of [`batch_size`, `num_channels`, `height`, `width`], and are generated from an image processor.
+
+### pooling
+
+An operation that reduces a matrix into a smaller matrix, either by taking the maximum or average of the pooled dimension(s). Pooling layers are commonly found between convolutional layers to downsample the feature representation.
+
+### position IDs
+
+Contrary to RNNs that have the position of each token embedded within them, transformers are unaware of the position of
+each token. Therefore, the position IDs (`position_ids`) are used by the model to identify each token's position in the
+list of tokens.
+
+They are an optional parameter. If no `position_ids` are passed to the model, the IDs are automatically created as
+absolute positional embeddings.
+
+Absolute positional embeddings are selected in the range `[0, config.max_position_embeddings - 1]`. Some models use
+other types of positional embeddings, such as sinusoidal position embeddings or relative position embeddings.
 
 
-### Token Type IDs
+### pretrained model
+
+A model that has been pretrained on some data (for instance all of Wikipedia). Pretraining methods involve a
+self-supervised objective, which can be reading the text and trying to predict the next word (see [causal language
+modeling](#causal-language-modeling)) or masking some words and trying to predict them (see [masked language
+modeling](#masked-language-modeling)). 
+
+  Speech and vision models have their own pretraining objectives. For example, Wav2Vec2 is a speech model pretrained on a contrastive task which requires the model to identify the "true" speech representation from a set of "false" speech representations. On the other hand, BEiT is a vision model pretrained on a masked image modeling task which masks some of the image patches and requires the model to predict the masked patches (similar to the masked language modeling objective).
+
+## R
+
+### recurrent neural network
+
+A type of model that uses a loop over a layer to process texts.
+
+## S
+
+### sampling rate
+
+A measurement in hertz of the number of samples (the audio signal) taken per second. The sampling rate is a result of discretizing a continuous signal such as speech.
+
+### self-attention
+
+Each element of the input finds out which other elements of the input they should attend to.
+
+### sequence-to-sequence (seq2seq)
+
+Models that generate a new sequence from an input, like translation models, or summarization models (such as
+[Bart](model_doc/bart) or [T5](model_doc/t5)).
+
+### stride
+
+In [convolution](#convolution) or [pooling](#pooling), the stride refers to the distance the kernel is moved over a matrix. A stride of 1 means the kernel is moved one pixel over at a time, and a stride of 2 means the kernel is moved two pixels over at a time.
+
+## T
+
+### token
+
+A part of a sentence, usually a word, but can also be a subword (non-common words are often split in subwords) or a
+punctuation symbol.
+
+### token Type IDs
 
 Some models' purpose is to do classification on pairs of sentences or question answering.
 
 <Youtube id="0u3ioSwev3s"/>
 
 These require two different sequences to be joined in a single "input_ids" entry, which usually is performed with the
-help of special tokens, such as the classifier (`[CLS]`) and separator (`[SEP]`) tokens. For example, the BERT
-model builds its two sequence input as such:
+help of special tokens, such as the classifier (`[CLS]`) and separator (`[SEP]`) tokens. For example, the BERT model
+builds its two sequence input as such:
 
 ```python
 >>> # [CLS] SEQUENCE_A [SEP] SEQUENCE_B [SEP]
@@ -219,81 +405,11 @@ The tokenizer returns this mask as the "token_type_ids" entry:
 [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1]
 ```
 
-The first sequence, the "context" used for the question, has all its tokens represented by a `0`, whereas the
-second sequence, corresponding to the "question", has all its tokens represented by a `1`.
+The first sequence, the "context" used for the question, has all its tokens represented by a `0`, whereas the second
+sequence, corresponding to the "question", has all its tokens represented by a `1`.
 
 Some models, like [`XLNetModel`] use an additional token represented by a `2`.
 
+### transformer
 
-
-### Position IDs
-
-Contrary to RNNs that have the position of each token embedded within them, transformers are unaware of the position of
-each token. Therefore, the position IDs (`position_ids`) are used by the model to identify each token's position in
-the list of tokens.
-
-They are an optional parameter. If no `position_ids` are passed to the model, the IDs are automatically created as
-absolute positional embeddings.
-
-Absolute positional embeddings are selected in the range `[0, config.max_position_embeddings - 1]`. Some models use
-other types of positional embeddings, such as sinusoidal position embeddings or relative position embeddings.
-
-
-
-### Labels
-
-The labels are an optional argument which can be passed in order for the model to compute the loss itself. These labels
-should be the expected prediction of the model: it will use the standard loss in order to compute the loss between its
-predictions and the expected value (the label).
-
-These labels are different according to the model head, for example:
-
-- For sequence classification models (e.g., [`BertForSequenceClassification`]), the model expects a
-  tensor of dimension `(batch_size)` with each value of the batch corresponding to the expected label of the
-  entire sequence.
-- For token classification models (e.g., [`BertForTokenClassification`]), the model expects a tensor
-  of dimension `(batch_size, seq_length)` with each value corresponding to the expected label of each individual
-  token.
-- For masked language modeling (e.g., [`BertForMaskedLM`]), the model expects a tensor of dimension
-  `(batch_size, seq_length)` with each value corresponding to the expected label of each individual token: the
-  labels being the token ID for the masked token, and values to be ignored for the rest (usually -100).
-- For sequence to sequence tasks,(e.g., [`BartForConditionalGeneration`],
-  [`MBartForConditionalGeneration`]), the model expects a tensor of dimension `(batch_size, tgt_seq_length)` with each value corresponding to the target sequences associated with each input sequence. During
-  training, both *BART* and *T5* will make the appropriate *decoder_input_ids* and decoder attention masks internally.
-  They usually do not need to be supplied. This does not apply to models leveraging the Encoder-Decoder framework. See
-  the documentation of each model for more information on each specific model's labels.
-
-The base models (e.g., [`BertModel`]) do not accept labels, as these are the base transformer
-models, simply outputting features.
-
-
-
-### Decoder input IDs
-
-This input is specific to encoder-decoder models, and contains the input IDs that will be fed to the decoder. These
-inputs should be used for sequence to sequence tasks, such as translation or summarization, and are usually built in a
-way specific to each model.
-
-Most encoder-decoder models (BART, T5) create their `decoder_input_ids` on their own from the `labels`. In
-such models, passing the `labels` is the preferred way to handle training.
-
-Please check each model's docs to see how they handle these input IDs for sequence to sequence training.
-
-
-### Feed Forward Chunking
-
-In each residual attention block in transformers the self-attention layer is usually followed by 2 feed forward layers.
-The intermediate embedding size of the feed forward layers is often bigger than the hidden size of the model (e.g., for
-`bert-base-uncased`).
-
-For an input of size `[batch_size, sequence_length]`, the memory required to store the intermediate feed forward
-embeddings `[batch_size, sequence_length, config.intermediate_size]` can account for a large fraction of the memory
-use. The authors of [Reformer: The Efficient Transformer](https://arxiv.org/abs/2001.04451) noticed that since the
-computation is independent of the `sequence_length` dimension, it is mathematically equivalent to compute the output
-embeddings of both feed forward layers `[batch_size, config.hidden_size]_0, ..., [batch_size, config.hidden_size]_n`
-individually and concat them afterward to `[batch_size, sequence_length, config.hidden_size]` with `n = sequence_length`, which trades increased computation time against reduced memory use, but yields a mathematically
-**equivalent** result.
-
-For models employing the function [`apply_chunking_to_forward`], the `chunk_size` defines the
-number of output embeddings that are computed in parallel and thus defines the trade-off between memory and time
-complexity. If `chunk_size` is set to 0, no feed forward chunking is done.
+Self-attention based deep learning model architecture.

docs/source/en/hpo_train.mdx

@@ -0,0 +1,120 @@
+<!--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
+-->
+
+# Hyperparameter Search using Trainer API
+
+🤗 Transformers provides a [`Trainer`] class optimized for training 🤗 Transformers models, making it easier to start training without manually writing your own training loop. The [`Trainer`] provides API for hyperparameter search. This doc shows how to enable it in example. 
+
+## Hyperparameter Search backend
+
+[`Trainer`] supports four hyperparameter search backends currently:
+[optuna](https://optuna.org/), [sigopt](https://sigopt.com/), [raytune](https://docs.ray.io/en/latest/tune/index.html) and [wandb](https://wandb.ai/site/sweeps).
+
+you should install them before using them as the hyperparameter search backend
+```bash
+pip install optuna/sigopt/wandb/ray[tune] 
+```
+
+## How to enable Hyperparameter search in example
+
+Define the hyperparameter search space, different backends need different format.
+
+For sigopt, see sigopt [object_parameter](https://docs.sigopt.com/ai-module-api-references/api_reference/objects/object_parameter), it's like following:
+```py
+>>> def sigopt_hp_space(trial):
+...     return [
+...         {"bounds": {"min": 1e-6, "max": 1e-4}, "name": "learning_rate", "type": "double"},
+...         {
+...             "categorical_values": ["16", "32", "64", "128"],
+...             "name": "per_device_train_batch_size",
+...             "type": "categorical",
+...         },
+...     ]
+```
+
+For optuna, see optuna [object_parameter](https://optuna.readthedocs.io/en/stable/tutorial/10_key_features/002_configurations.html#sphx-glr-tutorial-10-key-features-002-configurations-py), it's like following:
+
+```py
+>>> def optuna_hp_space(trial):
+...     return {
+...         "learning_rate": trial.suggest_float("learning_rate", 1e-6, 1e-4, log=True),
+...         "per_device_train_batch_size": trial.suggest_categorical("per_device_train_batch_size", [16, 32, 64, 128]),
+...     }
+```
+
+For raytune, see raytune [object_parameter](https://docs.ray.io/en/latest/tune/api_docs/search_space.html), it's like following:
+
+```py
+>>> def ray_hp_space(trial):
+...     return {
+...         "learning_rate": tune.loguniform(1e-6, 1e-4),
+...         "per_device_train_batch_size": tune.choice([16, 32, 64, 128]),
+...     }
+```
+
+For wandb, see wandb [object_parameter](https://docs.wandb.ai/guides/sweeps/configuration), it's like following:
+
+```py
+>>> def wandb_hp_space(trial):
+...     return {
+...         "method": "random",
+...         "metric": {"name": "objective", "goal": "minimize"},
+...         "parameters": {
+...             "learning_rate": {"distribution": "uniform", "min": 1e-6, "max": 1e-4},
+...             "per_device_train_batch_size": {"values": [16, 32, 64, 128]},
+...         },
+...     }
+```
+
+Define a `model_init` function and pass it to the [`Trainer`], as an example:
+```py
+>>> def model_init(trial):
+...     return AutoModelForSequenceClassification.from_pretrained(
+...         model_args.model_name_or_path,
+...         from_tf=bool(".ckpt" in model_args.model_name_or_path),
+...         config=config,
+...         cache_dir=model_args.cache_dir,
+...         revision=model_args.model_revision,
+...         use_auth_token=True if model_args.use_auth_token else None,
+...     )
+```
+
+Create a [`Trainer`] with your `model_init` function, training arguments, training and test datasets, and evaluation function:
+
+```py
+>>> trainer = Trainer(
+...     model=None,
+...     args=training_args,
+...     train_dataset=small_train_dataset,
+...     eval_dataset=small_eval_dataset,
+...     compute_metrics=compute_metrics,
+...     tokenizer=tokenizer,
+...     model_init=model_init,
+...     data_collator=data_collator,
+... )
+```
+
+Call hyperparameter search, get the best trial parameters, backend could be `"optuna"`/`"sigopt"`/`"wandb"`/`"ray"`. direction can be`"minimize"` or `"maximize"`, which indicates whether to optimize greater or lower objective.
+
+You could define your own compute_objective function, if not defined, the default compute_objective will be called, and the sum of eval metric like f1 is returned as objective value.
+
+```py
+>>> best_trial = trainer.hyperparameter_search(
+...     direction="maximize",
+...     backend="optuna",
+...     hp_space=optuna_hp_space,
+...     n_trials=20,
+...     compute_objective=compute_objective,
+... )
+```
+
+## Hyperparameter search For DDP finetune
+Currently, Hyperparameter search for DDP is enabled for optuna and sigopt. Only the rank-zero process will generate the search trial and pass the argument to other ranks.

docs/source/en/index.mdx

@@ -28,8 +28,8 @@ Join the growing community on the [Hub](https://huggingface.co/models), [forum](
 ## 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://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>
+    <img alt="HuggingFace Expert Acceleration Program" src="https://cdn-media.huggingface.co/marketing/transformers/new-support-improved.png" style="width: 100%; max-width: 600px; border: 1px solid #eee; border-radius: 4px; box-shadow: 0 1px 2px 0 rgba(0, 0, 0, 0.05);">
+</a>
 
 ## Contents
 
@@ -50,6 +50,7 @@ The documentation is organized into five sections:
 <!--This list is updated automatically from the README with _make fix-copies_. Do not update manually! -->
 
 1. **[ALBERT](model_doc/albert)** (from Google Research and the Toyota Technological Institute at Chicago) released with the paper [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://arxiv.org/abs/1909.11942), by Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin Gimpel, Piyush Sharma, Radu Soricut.
+1. **[Audio Spectrogram Transformer](model_doc/audio-spectrogram-transformer)** (from MIT) released with the paper [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) by Yuan Gong, Yu-An Chung, James Glass.
 1. **[BART](model_doc/bart)** (from Facebook) released with the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) by Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov and Luke Zettlemoyer.
 1. **[BARThez](model_doc/barthez)** (from École polytechnique) released with the paper [BARThez: a Skilled Pretrained French Sequence-to-Sequence Model](https://arxiv.org/abs/2010.12321) by Moussa Kamal Eddine, Antoine J.-P. Tixier, Michalis Vazirgiannis.
 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.
@@ -61,13 +62,16 @@ The documentation is organized into five sections:
 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. **[BLOOM](model_doc/bloom)** (from BigScience workshop) released by the [BigScience 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. **[Chinese-CLIP](model_doc/chinese_clip)** (from OFA-Sys) released with the paper [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) by An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou.
 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. **[CLIPSeg](model_doc/clipseg)** (from University of Göttingen) released with the paper [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) by Timo Lüddecke and Alexander Ecker.
 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. **[Conditional DETR](model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
 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.
@@ -77,9 +81,11 @@ The documentation is organized into five sections:
 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. **[Deformable DETR](model_doc/deformable_detr)** (from SenseTime Research) released with the paper [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://arxiv.org/abs/2010.04159) by Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, Jifeng Dai.
 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. **[DiNAT](model_doc/dinat)** (from SHI Labs) released with the paper [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001) by Ali Hassani and Humphrey Shi.
 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. **[Donut](model_doc/donut)** (from NAVER), released together with the paper [OCR-free Document Understanding Transformer](https://arxiv.org/abs/2111.15664) by Geewook Kim, Teakgyu Hong, Moonbin Yim, Jeongyeon Nam, Jinyoung Park, Jinyeong Yim, Wonseok Hwang, Sangdoo Yun, Dongyoon Han, Seunghyun Park.
@@ -88,6 +94,8 @@ The documentation is organized into five sections:
 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. **[ERNIE](model_doc/ernie)** (from Baidu) released with the paper [ERNIE: Enhanced Representation through Knowledge Integration](https://arxiv.org/abs/1904.09223) by Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Xuyi Chen, Han Zhang, Xin Tian, Danxiang Zhu, Hao Tian, Hua Wu.
+1. **[ESM](model_doc/esm)** (from Meta AI) are transformer protein language models.  **ESM-1b** was released with the paper [Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, C. Lawrence Zitnick, Jerry Ma, and Rob Fergus. **ESM-1v** was released with the paper [Language models enable zero-shot prediction of the effects of mutations on protein function](https://doi.org/10.1101/2021.07.09.450648) by Joshua Meier, Roshan Rao, Robert Verkuil, Jason Liu, Tom Sercu and Alexander Rives. **ESM-2 and ESMFold** were released with the paper [Language models of protein sequences at the scale of evolution enable accurate structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido, Alexander Rives.
+1. **[FLAN-T5](model_doc/flan-t5)** (from Google AI) released in the repository [google-research/t5x](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints) by Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Shane Gu, Zhuyun Dai, Mirac Suzgun, Xinyun Chen, Aakanksha Chowdhery, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew Dai, Hongkun Yu, Slav Petrov, Ed H. Chi, Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, and Jason Wei
 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.
@@ -96,18 +104,21 @@ The documentation is organized into five sections:
 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 NeoX Japanese](model_doc/gpt_neox_japanese)** (from ABEJA) released by Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori.
 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. **[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. **[Jukebox](model_doc/jukebox)** (from OpenAI) released with the paper [Jukebox: A Generative Model for Music](https://arxiv.org/pdf/2005.00341.pdf) by Prafulla Dhariwal, Heewoo Jun, Christine Payne, Jong Wook Kim, Alec Radford, 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/layoutxlm)** (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. **[LiLT](model_doc/lilt)** (from South China University of Technology) released with the paper [LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding](https://arxiv.org/abs/2202.13669) by Jiapeng Wang, Lianwen Jin, Kai Ding.
 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.
@@ -115,6 +126,7 @@ The documentation is organized into five sections:
 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. **[MarkupLM](model_doc/markuplm)** (from Microsoft Research Asia) released with the paper [MarkupLM: Pre-training of Text and Markup Language for Visually-rich Document Understanding](https://arxiv.org/abs/2110.08518) by Junlong Li, Yiheng Xu, Lei Cui, Furu Wei.
 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.
@@ -122,10 +134,13 @@ The documentation is organized into five sections:
 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. **[MobileNetV1](model_doc/mobilenet_v1)** (from Google Inc.) released with the paper [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://arxiv.org/abs/1704.04861) by Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam.
+1. **[MobileNetV2](model_doc/mobilenet_v2)** (from Google Inc.) released with the paper [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381) by Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen.
 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. **[NAT](model_doc/nat)** (from SHI Labs) released with the paper [Neighborhood Attention Transformer](https://arxiv.org/abs/2204.07143) by Ali Hassani, Steven Walton, Jiachen Li, Shen Li, and Humphrey Shi.
 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.
@@ -146,6 +161,7 @@ The documentation is organized into five sections:
 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. **[RoCBert](model_doc/roc_bert)** (from WeChatAI) released with the paper [RoCBert: Robust Chinese Bert with Multimodal Contrastive Pretraining](https://aclanthology.org/2022.acl-long.65.pdf) by HuiSu, WeiweiShi, XiaoyuShen, XiaoZhou, TuoJi, JiaruiFang, JieZhou.
 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.
 1. **[SegFormer](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.
 1. **[SEW](model_doc/sew)** (from ASAPP) released with the paper [Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition](https://arxiv.org/abs/2109.06870) by Felix Wu, Kwangyoun Kim, Jing Pan, Kyu Han, Kilian Q. Weinberger, Yoav Artzi.
@@ -156,10 +172,13 @@ The documentation is organized into five sections:
 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. **[Swin Transformer V2](model_doc/swinv2)** (from Microsoft) released with the paper [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883) by Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo.
+1. **[SwitchTransformers](model_doc/switch_transformers)** (from Google) released with the paper [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://arxiv.org/abs/2101.03961) by William Fedus, Barret Zoph, Noam Shazeer.
 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. **[Table Transformer](model_doc/table-transformer)** (from Microsoft Research) released with the paper [PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents](https://arxiv.org/abs/2110.00061) by Brandon Smock, Rohith Pesala, Robin Abraham.
 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. **[Time Series Transformer](model_doc/time_series_transformer)**  (from HuggingFace).
 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.
@@ -172,10 +191,12 @@ The documentation is organized into five sections:
 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. **[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. **[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. **[ViTMSN](model_doc/vit_msn)** (from Meta AI) released with the paper [Masked Siamese Networks for Label-Efficient Learning](https://arxiv.org/abs/2204.07141) by Mahmoud Assran, Mathilde Caron, Ishan Misra, Piotr Bojanowski, Florian Bordes, Pascal Vincent, Armand Joulin, Michael Rabbat, Nicolas Ballas.
 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. **[Whisper](model_doc/whisper)** (from OpenAI) released with the paper [Robust Speech Recognition via Large-Scale Weak Supervision](https://cdn.openai.com/papers/whisper.pdf) by Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, Ilya Sutskever.
 1. **[X-CLIP](model_doc/xclip)** (from Microsoft Research) released with the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Bolin Ni, Houwen Peng, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling.
 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.
@@ -197,132 +218,153 @@ Flax), PyTorch, and/or TensorFlow.
 
 <!--This table is updated automatically from the auto modules with _make fix-copies_. Do not update manually!-->
 
-|            Model            | Tokenizer slow | Tokenizer fast | PyTorch support | TensorFlow support | Flax Support |
-|:---------------------------:|:--------------:|:--------------:|:---------------:|:------------------:|:------------:|
-|           ALBERT            |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|            BART             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|            BEiT             |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
-|            BERT             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|       Bert Generation       |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
-|           BigBird           |       ✅       |       ✅       |       ✅        |         ❌         |      ✅      |
-|       BigBird-Pegasus       |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|         Blenderbot          |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|       BlenderbotSmall       |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|            BLOOM            |       ❌       |       ✅       |       ✅        |         ❌         |      ❌      |
-|          CamemBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|           CANINE            |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
-|            CLIP             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|           CodeGen           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
-|          ConvBERT           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|          ConvNeXT           |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
-|            CTRL             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
-|             CvT             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|        Data2VecAudio        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|        Data2VecText         |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|       Data2VecVision        |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
-|           DeBERTa           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|         DeBERTa-v2          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|    Decision Transformer     |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|            DeiT             |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
-|            DETR             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|         DistilBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|          DonutSwin          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|             DPR             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|             DPT             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|           ELECTRA           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|       Encoder decoder       |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
-|            ERNIE            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-| 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            |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|        Megatron-BERT        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|         MobileBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|          MobileViT          |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
-|            MPNet            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|             MT5             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|             MVP             |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
-|            Nezha            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|        Nyströmformer        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|         OpenAI GPT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|        OpenAI GPT-2         |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|             OPT             |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
-|           OWL-ViT           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|           Pegasus           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|          PEGASUS-X          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|          Perceiver          |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
-|           PLBart            |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
-|         PoolFormer          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|         ProphetNet          |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
-|           QDQBert           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|             RAG             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
-|            REALM            |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
-|          Reformer           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
-|           RegNet            |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
-|           RemBERT           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|           ResNet            |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
-|          RetriBERT          |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
-|           RoBERTa           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|          RoFormer           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|          SegFormer          |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
-|             SEW             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|            SEW-D            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|   Speech Encoder decoder    |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
-|         Speech2Text         |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
-|        Speech2Text2         |       ✅       |       ❌       |       ❌        |         ❌         |      ❌      |
-|          Splinter           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
-|         SqueezeBERT         |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
-|      Swin Transformer       |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
-|     Swin Transformer V2     |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|             T5              |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|            TAPAS            |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
-|   Trajectory Transformer    |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|       Transformer-XL        |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
-|            TrOCR            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|          UniSpeech          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|        UniSpeechSat         |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|             VAN             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|          VideoMAE           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|            ViLT             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|   Vision Encoder decoder    |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
-|    VisionTextDualEncoder    |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
-|         VisualBERT          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|             ViT             |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
-|           ViTMAE            |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
-|          Wav2Vec2           |       ✅       |       ❌       |       ✅        |         ✅         |      ✅      |
-|     Wav2Vec2-Conformer      |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|            WavLM            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|           X-CLIP            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|            XGLM             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|             XLM             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
-|       XLM-ProphetNet        |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
-|         XLM-RoBERTa         |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
-|       XLM-RoBERTa-XL        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|            XLNet            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
-|            YOLOS            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
-|            YOSO             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             Model             | Tokenizer slow | Tokenizer fast | PyTorch support | TensorFlow support | Flax Support |
+|:-----------------------------:|:--------------:|:--------------:|:---------------:|:------------------:|:------------:|
+|            ALBERT             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+| Audio Spectrogram Transformer |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             BART              |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|             BEiT              |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
+|             BERT              |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|        Bert Generation        |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            BigBird            |       ✅       |       ✅       |       ✅        |         ❌         |      ✅      |
+|        BigBird-Pegasus        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|          Blenderbot           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|        BlenderbotSmall        |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|             BLOOM             |       ❌       |       ✅       |       ✅        |         ❌         |      ❌      |
+|           CamemBERT           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|            CANINE             |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|         Chinese-CLIP          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             CLIP              |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|            CLIPSeg            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            CodeGen            |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|       Conditional DETR        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           ConvBERT            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|           ConvNeXT            |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|             CTRL              |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|              CvT              |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|         Data2VecAudio         |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|         Data2VecText          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|        Data2VecVision         |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|            DeBERTa            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|          DeBERTa-v2           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|     Decision Transformer      |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|        Deformable DETR        |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             DeiT              |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|             DETR              |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             DiNAT             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|          DistilBERT           |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|           DonutSwin           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|              DPR              |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|              DPT              |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            ELECTRA            |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|        Encoder decoder        |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
+|             ERNIE             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|              ESM              |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|  FairSeq Machine-Translation  |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           FlauBERT            |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|             FLAVA             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             FNet              |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|      Funnel Transformer       |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|             GLPN              |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            GPT Neo            |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
+|           GPT NeoX            |       ❌       |       ✅       |       ✅        |         ❌         |      ❌      |
+|       GPT NeoX Japanese       |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             GPT-J             |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
+|           GroupViT            |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|            Hubert             |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|            I-BERT             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           ImageGPT            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            Jukebox            |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           LayoutLM            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|          LayoutLMv2           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|          LayoutLMv3           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|              LED              |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|             LeViT             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             LiLT              |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|          Longformer           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|            LongT5             |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
+|             LUKE              |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            LXMERT             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|            M-CTC-T            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            M2M100             |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            Marian             |       ✅       |       ❌       |       ✅        |         ✅         |      ✅      |
+|           MarkupLM            |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|          MaskFormer           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|        MaskFormerSwin         |       ❌       |       ❌       |       ❌        |         ❌         |      ❌      |
+|             mBART             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|         Megatron-BERT         |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|          MobileBERT           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|          MobileNetV1          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|          MobileNetV2          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           MobileViT           |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|             MPNet             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|              MT5              |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|              MVP              |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|              NAT              |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             Nezha             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|         Nyströmformer         |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|          OpenAI GPT           |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|         OpenAI GPT-2          |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|              OPT              |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
+|            OWL-ViT            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            Pegasus            |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|           PEGASUS-X           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           Perceiver           |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            PLBart             |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|          PoolFormer           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|          ProphetNet           |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            QDQBert            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|              RAG              |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|             REALM             |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|           Reformer            |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|            RegNet             |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|            RemBERT            |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|            ResNet             |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|           RetriBERT           |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|            RoBERTa            |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|            RoCBert            |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           RoFormer            |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|           SegFormer           |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|              SEW              |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             SEW-D             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|    Speech Encoder decoder     |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
+|          Speech2Text          |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|         Speech2Text2          |       ✅       |       ❌       |       ❌        |         ❌         |      ❌      |
+|           Splinter            |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|          SqueezeBERT          |       ✅       |       ✅       |       ✅        |         ❌         |      ❌      |
+|       Swin Transformer        |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|      Swin Transformer V2      |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|      SwitchTransformers       |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|              T5               |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|       Table Transformer       |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             TAPAS             |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|    Time Series Transformer    |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|    Trajectory Transformer     |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|        Transformer-XL         |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|             TrOCR             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           UniSpeech           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|         UniSpeechSat          |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|              VAN              |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           VideoMAE            |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             ViLT              |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|    Vision Encoder decoder     |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
+|     VisionTextDualEncoder     |       ❌       |       ❌       |       ✅        |         ❌         |      ✅      |
+|          VisualBERT           |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|              ViT              |       ❌       |       ❌       |       ✅        |         ✅         |      ✅      |
+|            ViTMAE             |       ❌       |       ❌       |       ✅        |         ✅         |      ❌      |
+|            ViTMSN             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|           Wav2Vec2            |       ✅       |       ❌       |       ✅        |         ✅         |      ✅      |
+|      Wav2Vec2-Conformer       |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             WavLM             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|            Whisper            |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|            X-CLIP             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             XGLM              |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|              XLM              |       ✅       |       ❌       |       ✅        |         ✅         |      ❌      |
+|        XLM-ProphetNet         |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
+|          XLM-RoBERTa          |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|        XLM-RoBERTa-XL         |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             XLNet             |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
+|             YOLOS             |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
+|             YOSO              |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 
-<!-- End table-->
+<!-- End table-->

docs/source/en/internal/generation_utils.mdx

@@ -12,21 +12,22 @@ specific language governing permissions and limitations under the License.
 
 # Utilities for Generation
 
-This page lists all the utility functions used by [`~generation_utils.GenerationMixin.generate`],
-[`~generation_utils.GenerationMixin.greedy_search`],
-[`~generation_utils.GenerationMixin.sample`],
-[`~generation_utils.GenerationMixin.beam_search`],
-[`~generation_utils.GenerationMixin.beam_sample`],
-[`~generation_utils.GenerationMixin.group_beam_search`], and
-[`~generation_utils.GenerationMixin.constrained_beam_search`].
+This page lists all the utility functions used by [`~generation.GenerationMixin.generate`],
+[`~generation.GenerationMixin.greedy_search`],
+[`~generation.GenerationMixin.contrastive_search`],
+[`~generation.GenerationMixin.sample`],
+[`~generation.GenerationMixin.beam_search`],
+[`~generation.GenerationMixin.beam_sample`],
+[`~generation.GenerationMixin.group_beam_search`], and
+[`~generation.GenerationMixin.constrained_beam_search`].
 
 Most of those are only useful if you are studying the code of the generate methods in the library.
 
 ## Generate Outputs
 
-The output of [`~generation_utils.GenerationMixin.generate`] is an instance of a subclass of
+The output of [`~generation.GenerationMixin.generate`] is an instance of a subclass of
 [`~utils.ModelOutput`]. This output is a data structure containing all the information returned
-by [`~generation_utils.GenerationMixin.generate`], but that can also be used as tuple or dictionary.
+by [`~generation.GenerationMixin.generate`], but that can also be used as tuple or dictionary.
 
 Here's an example:
 
@@ -40,7 +41,7 @@ inputs = tokenizer("Hello, my dog is cute and ", return_tensors="pt")
 generation_output = model.generate(**inputs, return_dict_in_generate=True, output_scores=True)
 ```
 
-The `generation_output` object is a [`~generation_utils.GreedySearchDecoderOnlyOutput`], as we can
+The `generation_output` object is a [`~generation.GreedySearchDecoderOnlyOutput`], as we can
 see in the documentation of that class below, it means it has the following attributes:
 
 - `sequences`: the generated sequences of tokens
@@ -72,31 +73,31 @@ We document here all output types.
 
 ### GreedySearchOutput
 
-[[autodoc]] generation_utils.GreedySearchDecoderOnlyOutput
+[[autodoc]] generation.GreedySearchDecoderOnlyOutput
 
-[[autodoc]] generation_utils.GreedySearchEncoderDecoderOutput
+[[autodoc]] generation.GreedySearchEncoderDecoderOutput
 
-[[autodoc]] generation_flax_utils.FlaxGreedySearchOutput
+[[autodoc]] generation.FlaxGreedySearchOutput
 
 ### SampleOutput
 
-[[autodoc]] generation_utils.SampleDecoderOnlyOutput
+[[autodoc]] generation.SampleDecoderOnlyOutput
 
-[[autodoc]] generation_utils.SampleEncoderDecoderOutput
+[[autodoc]] generation.SampleEncoderDecoderOutput
 
-[[autodoc]] generation_flax_utils.FlaxSampleOutput
+[[autodoc]] generation.FlaxSampleOutput
 
 ### BeamSearchOutput
 
-[[autodoc]] generation_utils.BeamSearchDecoderOnlyOutput
+[[autodoc]] generation.BeamSearchDecoderOnlyOutput
 
-[[autodoc]] generation_utils.BeamSearchEncoderDecoderOutput
+[[autodoc]] generation.BeamSearchEncoderDecoderOutput
 
 ### BeamSampleOutput
 
-[[autodoc]] generation_utils.BeamSampleDecoderOnlyOutput
+[[autodoc]] generation.BeamSampleDecoderOnlyOutput
 
-[[autodoc]] generation_utils.BeamSampleEncoderDecoderOutput
+[[autodoc]] generation.BeamSampleEncoderDecoderOutput
 
 ## LogitsProcessor
 

docs/source/en/internal/image_processing_utils.mdx

@@ -0,0 +1,44 @@
+<!--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.
+-->
+
+# Utilities for Image Processors
+
+This page lists all the utility functions used by the image processors, mainly the functional
+transformations used to process the images.
+
+Most of those are only useful if you are studying the code of the image processors in the library.
+
+## Image Transformations
+
+[[autodoc]] image_transforms.center_crop
+
+[[autodoc]] image_transforms.center_to_corners_format
+
+[[autodoc]] image_transforms.corners_to_center_format
+
+[[autodoc]] image_transforms.id_to_rgb
+
+[[autodoc]] image_transforms.normalize
+
+[[autodoc]] image_transforms.pad
+
+[[autodoc]] image_transforms.rgb_to_id
+
+[[autodoc]] image_transforms.rescale
+
+[[autodoc]] image_transforms.resize
+
+[[autodoc]] image_transforms.to_pil_image
+
+## ImageProcessingMixin
+
+[[autodoc]] image_processing_utils.ImageProcessingMixin

docs/source/en/main_classes/callback.mdx

@@ -37,6 +37,7 @@ By default a [`Trainer`] will use the following callbacks:
   installed.
 - [`~integrations.CodeCarbonCallback`] if [codecarbon](https://pypi.org/project/codecarbon/) is
   installed.
+- [`~integrations.ClearMLCallback`] if [clearml](https://github.com/allegroai/clearml) is installed.
 
 The main class that implements callbacks is [`TrainerCallback`]. It gets the
 [`TrainingArguments`] used to instantiate the [`Trainer`], can access that
@@ -73,6 +74,8 @@ Here is the list of the available [`TrainerCallback`] in the library:
 
 [[autodoc]] integrations.NeptuneCallback
 
+[[autodoc]] integrations.ClearMLCallback
+
 ## TrainerCallback
 
 [[autodoc]] TrainerCallback

docs/source/en/main_classes/deepspeed.mdx

@@ -49,7 +49,7 @@ Inference:
 
 1. DeepSpeed ZeRO Inference supports ZeRO stage 3 with ZeRO-Infinity. It uses the same ZeRO protocol as training, but
    it doesn't use an optimizer and a lr scheduler and only stage 3 is relevant. For more details see:
-   [deepspeed-zero-inference](#deepspeed-zero-inference).
+   [zero-inference](#zero-inference).
 
 There is also DeepSpeed Inference - this is a totally different technology which uses Tensor Parallelism instead of
 ZeRO (coming soon).
@@ -81,7 +81,7 @@ pip install transformers[deepspeed]
 or find more details on [the DeepSpeed's GitHub page](https://github.com/microsoft/deepspeed#installation) and
 [advanced install](https://www.deepspeed.ai/tutorials/advanced-install/).
 
-If you're still struggling with the build, first make sure to read [zero-install-notes](#zero-install-notes).
+If you're still struggling with the build, first make sure to read [CUDA Extension Installation Notes](trainer#cuda-extension-installation-notes).
 
 If you don't prebuild the extensions and rely on them to be built at run time and you tried all of the above solutions
 to no avail, the next thing to try is to pre-build the modules before installing them.
@@ -1499,7 +1499,7 @@ fp32_model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
 
 <Tip>
 
-Note, that once `load_state_dict_from_zero_checkpoint` was run, the `model` will no longer be useable in the
+Note, that once `load_state_dict_from_zero_checkpoint` was run, the `model` will no longer be usable in the
 DeepSpeed context of the same application. i.e. you will need to re-initialize the deepspeed engine, since
 `model.load_state_dict(state_dict)` will remove all the DeepSpeed magic from it. So do this only at the very end
 of the training.

docs/source/en/main_classes/image_processor.mdx

@@ -0,0 +1,30 @@
+<!--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.
+-->
+
+# Image Processor
+
+An image processor is in charge of preparing input features for vision models and post processing their outputs. This includes transformations such as resizing, normalization, and conversion to PyTorch, TensorFlow, Flax and Numpy tensors. It may also include model specific post-processing such as converting logits to segmentation masks.
+
+
+## ImageProcessingMixin
+
+[[autodoc]] image_processing_utils.ImageProcessingMixin
+    - from_pretrained
+    - save_pretrained
+
+## BatchFeature
+
+[[autodoc]] BatchFeature
+
+## BaseImageProcessor
+
+[[autodoc]] image_processing_utils.BaseImageProcessor

docs/source/en/main_classes/model.mdx

@@ -25,9 +25,9 @@ are common among all the models to:
 
 The other methods that are common to each model are defined in [`~modeling_utils.ModuleUtilsMixin`]
 (for the PyTorch models) and [`~modeling_tf_utils.TFModuleUtilsMixin`] (for the TensorFlow models) or
-for text generation, [`~generation_utils.GenerationMixin`] (for the PyTorch models),
-[`~generation_tf_utils.TFGenerationMixin`] (for the TensorFlow models) and
-[`~generation_flax_utils.FlaxGenerationMixin`] (for the Flax/JAX models).
+for text generation, [`~generation.GenerationMixin`] (for the PyTorch models),
+[`~generation.TFGenerationMixin`] (for the TensorFlow models) and
+[`~generation.FlaxGenerationMixin`] (for the Flax/JAX models).
 
 
 ## PreTrainedModel

docs/source/en/main_classes/pipelines.mdx

@@ -20,29 +20,7 @@ Recognition, Masked Language Modeling, Sentiment Analysis, Feature Extraction an
 There are two categories of pipeline abstractions to be aware about:
 
 - The [`pipeline`] which is the most powerful object encapsulating all other pipelines.
-- The other task-specific pipelines:
-
-  - [`AudioClassificationPipeline`]
-  - [`AutomaticSpeechRecognitionPipeline`]
-  - [`ConversationalPipeline`]
-  - [`DocumentQuestionAnsweringPipeline`]
-  - [`FeatureExtractionPipeline`]
-  - [`FillMaskPipeline`]
-  - [`ImageClassificationPipeline`]
-  - [`ImageSegmentationPipeline`]
-  - [`ImageToTextPipeline`]
-  - [`ObjectDetectionPipeline`]
-  - [`QuestionAnsweringPipeline`]
-  - [`SummarizationPipeline`]
-  - [`TableQuestionAnsweringPipeline`]
-  - [`TextClassificationPipeline`]
-  - [`TextGenerationPipeline`]
-  - [`Text2TextGenerationPipeline`]
-  - [`TokenClassificationPipeline`]
-  - [`TranslationPipeline`]
-  - [`VisualQuestionAnsweringPipeline`]
-  - [`ZeroShotClassificationPipeline`]
-  - [`ZeroShotImageClassificationPipeline`]
+- Task-specific pipelines are available for [audio](#audio), [computer vision](#computer-vision), [natural language processing](#natural-language-processing), and [multimodal](#multimodal) tasks.
 
 ## The pipeline abstraction
 
@@ -89,7 +67,7 @@ pipe = pipeline("automatic-speech-recognition", model="facebook/wav2vec2-base-96
 dataset = datasets.load_dataset("superb", name="asr", split="test")
 
 # KeyDataset (only *pt*) will simply return the item in the dict returned by the dataset item
-# as we're not interested in the *target* part of the dataset.
+# as we're not interested in the *target* part of the dataset. For sentence pair use KeyPairDataset
 for out in tqdm(pipe(KeyDataset(dataset, "file"))):
     print(out)
     # {"text": "NUMBER TEN FRESH NELLY IS WAITING ON YOU GOOD NIGHT HUSBAND"}
@@ -320,8 +298,9 @@ That should enable you to do all the custom code you want.
 
 [Implementing a new pipeline](../add_new_pipeline)
 
-## The task specific pipelines
+## Audio
 
+Pipelines available for audio tasks include the following.
 
 ### AudioClassificationPipeline
 
@@ -335,29 +314,12 @@ That should enable you to do all the custom code you want.
     - __call__
     - all
 
-### ConversationalPipeline
+## Computer vision
 
-[[autodoc]] Conversation
+Pipelines available for computer vision tasks include the following.
 
-[[autodoc]] ConversationalPipeline
-    - __call__
-    - all
-
-### DocumentQuestionAnsweringPipeline
-
-[[autodoc]] DocumentQuestionAnsweringPipeline
-    - __call__
-    - all
-
-### FeatureExtractionPipeline
-
-[[autodoc]] FeatureExtractionPipeline
-    - __call__
-    - all
-
-### FillMaskPipeline
-
-[[autodoc]] FillMaskPipeline
+### DepthEstimationPipeline
+[[autodoc]] DepthEstimationPipeline
     - __call__
     - all
 
@@ -373,9 +335,39 @@ That should enable you to do all the custom code you want.
     - __call__
     - all
 
-### ImageToTextPipeline
+### ObjectDetectionPipeline
 
-[[autodoc]] ImageToTextPipeline
+[[autodoc]] ObjectDetectionPipeline
+    - __call__
+    - all
+
+### ZeroShotImageClassificationPipeline
+
+[[autodoc]] ZeroShotImageClassificationPipeline
+    - __call__
+    - all
+
+### ZeroShotObjectDetectionPipeline
+
+[[autodoc]] ZeroShotObjectDetectionPipeline
+    - __call__
+    - all
+
+## Natural Language Processing
+
+Pipelines available for natural language processing tasks include the following.
+
+### ConversationalPipeline
+
+[[autodoc]] Conversation
+
+[[autodoc]] ConversationalPipeline
+    - __call__
+    - all
+
+### FillMaskPipeline
+
+[[autodoc]] FillMaskPipeline
     - __call__
     - all
 
@@ -385,12 +377,6 @@ That should enable you to do all the custom code you want.
 
 See [`TokenClassificationPipeline`] for all details.
 
-### ObjectDetectionPipeline
-
-[[autodoc]] ObjectDetectionPipeline
-    - __call__
-    - all
-
 ### QuestionAnsweringPipeline
 
 [[autodoc]] QuestionAnsweringPipeline
@@ -438,21 +424,37 @@ See [`TokenClassificationPipeline`] for all details.
     - __call__
     - all
 
-### VisualQuestionAnsweringPipeline
-
-[[autodoc]] VisualQuestionAnsweringPipeline
-    - __call__
-    - all
-
 ### ZeroShotClassificationPipeline
 
 [[autodoc]] ZeroShotClassificationPipeline
     - __call__
     - all
 
-### ZeroShotImageClassificationPipeline
+## Multimodal
 
-[[autodoc]] ZeroShotImageClassificationPipeline
+Pipelines available for multimodal tasks include the following.
+
+### DocumentQuestionAnsweringPipeline
+
+[[autodoc]] DocumentQuestionAnsweringPipeline
+    - __call__
+    - all
+
+### FeatureExtractionPipeline
+
+[[autodoc]] FeatureExtractionPipeline
+    - __call__
+    - all
+
+### ImageToTextPipeline
+
+[[autodoc]] ImageToTextPipeline
+    - __call__
+    - all
+
+### VisualQuestionAnsweringPipeline
+
+[[autodoc]] VisualQuestionAnsweringPipeline
     - __call__
     - all
 

docs/source/en/main_classes/processors.mdx

@@ -20,8 +20,8 @@ Processors can mean two different things in the Transformers library:
 ## Multi-modal processors
 
 Any multi-modal model will require an object to encode or decode the data that groups several modalities (among text,
-vision and audio). This is handled by objects called processors, which group tokenizers (for the text modality) and
-feature extractors (for vision and audio).
+vision and audio). This is handled by objects called processors, which group together two or more processing objects
+such as tokenizers (for the text modality), image processors (for vision) and feature extractors (for audio).
 
 Those processors inherit from the following base class that implements the saving and loading functionality:
 
@@ -112,7 +112,7 @@ Additionally, the following method can be used to convert SQuAD examples into
 [[autodoc]] data.processors.squad.squad_convert_examples_to_features
 
 
-These processors as well as the aforementionned method can be used with files containing the data as well as with the
+These processors as well as the aforementioned method can be used with files containing the data as well as with the
 *tensorflow_datasets* package. Examples are given below.
 
 

docs/source/en/main_classes/text_generation.mdx

@@ -14,27 +14,36 @@ specific language governing permissions and limitations under the License.
 
 Each framework has a generate method for auto-regressive text generation implemented in their respective `GenerationMixin` class:
 
-- PyTorch [`~generation_utils.GenerationMixin.generate`] is implemented in [`~generation_utils.GenerationMixin`].
-- TensorFlow [`~generation_tf_utils.TFGenerationMixin.generate`] is implemented in [`~generation_tf_utils.TFGenerationMixin`].
-- Flax/JAX [`~generation_flax_utils.FlaxGenerationMixin.generate`] is implemented in [`~generation_flax_utils.FlaxGenerationMixin`].
+- PyTorch [`~generation.GenerationMixin.generate`] is implemented in [`~generation.GenerationMixin`].
+- TensorFlow [`~generation.TFGenerationMixin.generate`] is implemented in [`~generation.TFGenerationMixin`].
+- Flax/JAX [`~generation.FlaxGenerationMixin.generate`] is implemented in [`~generation.FlaxGenerationMixin`].
+
+<!--- TODO: add a brief description of GenerationConfig (with examples) when it becomes usable with generate --->
+
+## GenerationConfig
+
+[[autodoc]] generation.GenerationConfig
+	- from_pretrained
+	- save_pretrained
 
 ## GenerationMixin
 
-[[autodoc]] generation_utils.GenerationMixin
+[[autodoc]] generation.GenerationMixin
 	- generate
 	- greedy_search
 	- sample
 	- beam_search
 	- beam_sample
+	- contrastive_search
 	- group_beam_search
 	- constrained_beam_search
 
 ## TFGenerationMixin
 
-[[autodoc]] generation_tf_utils.TFGenerationMixin
+[[autodoc]] generation.TFGenerationMixin
 	- generate
 
 ## FlaxGenerationMixin
 
-[[autodoc]] generation_flax_utils.FlaxGenerationMixin
+[[autodoc]] generation.FlaxGenerationMixin
 	- generate

docs/source/en/main_classes/trainer.mdx

@@ -579,7 +579,7 @@ add `--fsdp "full_shard offload auto_wrap"` or `--fsdp "shard_grad_op offload au
   This specifies the transformer layer class name (case-sensitive) to wrap ,e.g, `BertLayer`, `GPTJBlock`, `T5Block` ....
   This is important because submodules that share weights (e.g., embedding layer) should not end up in different FSDP wrapped units. 
   Using this policy, wrapping happens for each block containing Multi-Head Attention followed by couple of MLP layers. 
-  Remaining layers including the shared embeddings are conviniently wrapped in same outermost FSDP unit.
+  Remaining layers including the shared embeddings are conveniently wrapped in same outermost FSDP unit.
   Therefore, use this for transformer based models.
   - For size based auto wrap policy, please add `--fsdp_min_num_params <number>` to command line arguments.
   It specifies FSDP's minimum number of parameters for auto wrapping.
@@ -620,7 +620,7 @@ please follow this nice medium article [GPU-Acceleration Comes to PyTorch on M1
 
 **Usage**:
 User has to just pass `--use_mps_device` argument. 
-For example, you can run the offical Glue text classififcation task (from the root folder) using Apple Silicon GPU with below command:
+For example, you can run the official Glue text classififcation task (from the root folder) using Apple Silicon GPU with below command:
 
 ```bash
 export TASK_NAME=mrpc

docs/source/en/model_doc/audio-spectrogram-transformer.mdx

@@ -0,0 +1,60 @@
+<!--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.
+-->
+
+# Audio Spectrogram Transformer
+
+## Overview
+
+The Audio Spectrogram Transformer model was proposed in [AST: Audio Spectrogram Transformer](https://arxiv.org/abs/2104.01778) by Yuan Gong, Yu-An Chung, James Glass.
+The Audio Spectrogram Transformer applies a [Vision Transformer](vit) to audio, by turning audio into an image (spectrogram). The model obtains state-of-the-art results
+for audio classification.
+
+The abstract from the paper is the following:
+
+*In the past decade, convolutional neural networks (CNNs) have been widely adopted as the main building block for end-to-end audio classification models, which aim to learn a direct mapping from audio spectrograms to corresponding labels. To better capture long-range global context, a recent trend is to add a self-attention mechanism on top of the CNN, forming a CNN-attention hybrid model. However, it is unclear whether the reliance on a CNN is necessary, and if neural networks purely based on attention are sufficient to obtain good performance in audio classification. In this paper, we answer the question by introducing the Audio Spectrogram Transformer (AST), the first convolution-free, purely attention-based model for audio classification. We evaluate AST on various audio classification benchmarks, where it achieves new state-of-the-art results of 0.485 mAP on AudioSet, 95.6% accuracy on ESC-50, and 98.1% accuracy on Speech Commands V2.*
+
+Tips:
+
+- When fine-tuning the Audio Spectrogram Transformer (AST) on your own dataset, it's recommended to take care of the input normalization (to make
+sure the input has mean of 0 and std of 0.5). [`ASTFeatureExtractor`] takes care of this. Note that it uses the AudioSet
+mean and std by default. You can check [`ast/src/get_norm_stats.py`](https://github.com/YuanGongND/ast/blob/master/src/get_norm_stats.py) to see how
+the authors compute the stats for a downstream dataset.
+- Note that the AST needs a low learning rate (the authors use a 10 times smaller learning rate compared to their CNN model proposed in the
+[PSLA paper](https://arxiv.org/abs/2102.01243)) and converges quickly, so please search for a suitable learning rate and learning rate scheduler for your task.
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/audio_spectogram_transformer_architecture.png"
+alt="drawing" width="600"/>
+
+<small> Audio pectrogram Transformer architecture. Taken from the <a href="https://arxiv.org/abs/2104.01778">original paper</a>.</small>
+
+This model was contributed by [nielsr](https://huggingface.co/nielsr).
+The original code can be found [here](https://github.com/YuanGongND/ast).
+
+
+## ASTConfig
+
+[[autodoc]] ASTConfig
+
+## ASTFeatureExtractor
+
+[[autodoc]] ASTFeatureExtractor
+    - __call__
+
+## ASTModel
+
+[[autodoc]] ASTModel
+    - forward
+
+## ASTForAudioClassification
+
+[[autodoc]] ASTForAudioClassification
+    - forward

docs/source/en/model_doc/auto.mdx

@@ -66,6 +66,10 @@ Likewise, if your `NewModel` is a subclass of [`PreTrainedModel`], make sure its
 
 [[autodoc]] AutoFeatureExtractor
 
+## AutoImageProcessor
+
+[[autodoc]] AutoImageProcessor
+
 ## AutoProcessor
 
 [[autodoc]] AutoProcessor
@@ -82,6 +86,10 @@ Likewise, if your `NewModel` is a subclass of [`PreTrainedModel`], make sure its
 
 [[autodoc]] AutoModelForCausalLM
 
+## AutoModelForDepthEstimation
+
+[[autodoc]] AutoModelForDepthEstimation
+
 ## AutoModelForMaskedLM
 
 [[autodoc]] AutoModelForMaskedLM
@@ -174,6 +182,10 @@ Likewise, if your `NewModel` is a subclass of [`PreTrainedModel`], make sure its
 
 [[autodoc]] AutoModelForInstanceSegmentation
 
+## AutoModelForZeroShotObjectDetection
+
+[[autodoc]] AutoModelForZeroShotObjectDetection
+
 ## TFAutoModel
 
 [[autodoc]] TFAutoModel

docs/source/en/model_doc/bart.mdx

@@ -32,6 +32,11 @@ According to the abstract,
   state-of-the-art results on a range of abstractive dialogue, question answering, and summarization tasks, with gains
   of up to 6 ROUGE.
 
+Tips:
+
+- BART is a model with absolute position embeddings so it's usually advised to pad the inputs on the right rather than
+  the left.
+
 This model was contributed by [sshleifer](https://huggingface.co/sshleifer). The Authors' code can be found [here](https://github.com/pytorch/fairseq/tree/master/examples/bart).
 
 
@@ -53,7 +58,7 @@ This model was contributed by [sshleifer](https://huggingface.co/sshleifer). The
 - Model predictions are intended to be identical to the original implementation when
   `forced_bos_token_id=0`. This only works, however, if the string you pass to
   [`fairseq.encode`] starts with a space.
-- [`~generation_utils.GenerationMixin.generate`] should be used for conditional generation tasks like
+- [`~generation.GenerationMixin.generate`] should be used for conditional generation tasks like
   summarization, see the example in that docstrings.
 - Models that load the *facebook/bart-large-cnn* weights will not have a `mask_token_id`, or be able to perform
   mask-filling tasks.
@@ -75,6 +80,33 @@ assert tok.batch_decode(generated_ids, skip_special_tokens=True) == [
 ]
 ```
 
+## Resources
+
+A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with BART. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
+
+<PipelineTag pipeline="summarization"/>
+
+- A blog post on [Distributed Training: Train BART/T5 for Summarization using 🤗 Transformers and Amazon SageMaker](https://huggingface.co/blog/sagemaker-distributed-training-seq2seq).
+- A notebook on how to [finetune BART for summarization with fastai using blurr](https://colab.research.google.com/github/ohmeow/ohmeow_website/blob/master/_notebooks/2020-05-23-text-generation-with-blurr.ipynb). 🌎
+- A notebook on how to [finetune BART for summarization in two languages with Trainer class](https://colab.research.google.com/github/elsanns/xai-nlp-notebooks/blob/master/fine_tune_bart_summarization_two_langs.ipynb). 🌎
+- [`BartForConditionalGeneration`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/summarization) and [noteboook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/summarization.ipynb).
+- [`TFBartForConditionalGeneration`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/summarization) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/summarization-tf.ipynb).
+- [`FlaxBartForConditionalGeneration`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/flax/summarization).
+- [Summarization](https://huggingface.co/course/chapter7/5?fw=pt#summarization) chapter of the 🤗 Hugging Face course.
+
+<PipelineTag pipeline="fill-mask"/>
+
+- [`BartForConditionalGeneration`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling#robertabertdistilbert-and-masked-language-modeling) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling.ipynb).
+- [`TFBartForConditionalGeneration`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/language-modeling#run_mlmpy) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling-tf.ipynb).
+- [`FlaxBartForConditionalGeneration`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/flax/language-modeling#masked-language-modeling) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/masked_language_modeling_flax.ipynb).
+- [Masked language modeling](https://huggingface.co/course/chapter7/3?fw=pt) chapter of the 🤗 Hugging Face Course.
+
+<PipelineTag pipeline="translation"/>
+
+- A notebook on how to [finetune mBART using Seq2SeqTrainer for Hindi to English translation](https://colab.research.google.com/github/vasudevgupta7/huggingface-tutorials/blob/main/translation_training.ipynb). 🌎
+- [`BartForConditionalGeneration`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/translation) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/translation.ipynb).
+- [`TFBartForConditionalGeneration`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/translation) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/translation-tf.ipynb).
+
 ## BartConfig
 
 [[autodoc]] BartConfig
@@ -156,4 +188,4 @@ assert tok.batch_decode(generated_ids, skip_special_tokens=True) == [
 ## FlaxBartForCausalLM
 
 [[autodoc]] FlaxBartForCausalLM
-    - __call__
+    - __call__

docs/source/en/model_doc/beit.mdx

@@ -40,12 +40,12 @@ Tips:
 - BEiT models are regular Vision Transformers, but pre-trained in a self-supervised way rather than supervised. They
   outperform both the [original model (ViT)](vit) as well as [Data-efficient Image Transformers (DeiT)](deit) 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 [`BeitFeatureExtractor`] and
+  [`ViTFeatureExtractor`] by [`BeitImageProcessor`] and
   [`ViTForImageClassification`] by [`BeitForImageClassification`]).
 - There's also a demo notebook available which showcases how to combine DALL-E's image tokenizer with BEiT for
   performing masked image modeling. You can find it [here](https://github.com/NielsRogge/Transformers-Tutorials/tree/master/BEiT).
 - As the BEiT models expect each image to be of the same size (resolution), one can use
-  [`BeitFeatureExtractor`] to resize (or rescale) and normalize images for the model.
+  [`BeitImageProcessor`] to resize (or rescale) and normalize images for the model.
 - Both the patch resolution and image resolution used during pre-training or fine-tuning are reflected in the name of
   each checkpoint. For example, `microsoft/beit-base-patch16-224` refers to a base-sized architecture with patch
   resolution of 16x16 and fine-tuning resolution of 224x224. All checkpoints can be found on the [hub](https://huggingface.co/models?search=microsoft/beit).
@@ -59,6 +59,11 @@ Tips:
   `use_relative_position_bias` attribute of [`BeitConfig`] to `True` in order to add
   position embeddings.
 
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/beit_architecture.jpg"
+alt="drawing" width="600"/>
+
+<small> BEiT pre-training. Taken from the <a href="https://arxiv.org/abs/2106.08254">original paper.</a> </small>
+
 This model was contributed by [nielsr](https://huggingface.co/nielsr). The JAX/FLAX version of this model was
 contributed by [kamalkraj](https://huggingface.co/kamalkraj). The original code can be found [here](https://github.com/microsoft/unilm/tree/master/beit).
 
@@ -77,6 +82,13 @@ contributed by [kamalkraj](https://huggingface.co/kamalkraj). The original code
 
 [[autodoc]] BeitFeatureExtractor
     - __call__
+    - post_process_semantic_segmentation
+
+## BeitImageProcessor
+
+[[autodoc]] BeitImageProcessor
+    - preprocess
+    - post_process_semantic_segmentation
 
 ## BeitModel
 

docs/source/en/model_doc/bert.mdx

@@ -41,6 +41,62 @@ Tips:
 
 This model was contributed by [thomwolf](https://huggingface.co/thomwolf). The original code can be found [here](https://github.com/google-research/bert).
 
+## Resources
+
+A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with BERT. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
+
+<PipelineTag pipeline="text-classification"/>
+
+- A blog post on [BERT Text Classification in a different language](https://www.philschmid.de/bert-text-classification-in-a-different-language).
+- A notebook for [Finetuning BERT (and friends) for multi-label text classification](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/BERT/Fine_tuning_BERT_(and_friends)_for_multi_label_text_classification.ipynb).
+- A notebook on how to [Finetune BERT for multi-label classification using PyTorch](https://colab.research.google.com/github/abhimishra91/transformers-tutorials/blob/master/transformers_multi_label_classification.ipynb). 🌎
+- A notebook on how to [warm-start an EncoderDecoder model with BERT for summarization](https://colab.research.google.com/github/patrickvonplaten/notebooks/blob/master/BERT2BERT_for_CNN_Dailymail.ipynb).
+- [`BertForSequenceClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/text-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification.ipynb).
+- [`TFBertForSequenceClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/text-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification-tf.ipynb).
+- [`FlaxBertForSequenceClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/flax/text-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification_flax.ipynb).
+
+<PipelineTag pipeline="token-classification"/>
+
+- A blog post on how to use [Hugging Face Transformers with Keras: Fine-tune a non-English BERT for Named Entity Recognition](https://www.philschmid.de/huggingface-transformers-keras-tf).
+- A notebook for [Finetuning BERT for named-entity recognition](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/Custom_Named_Entity_Recognition_with_BERT_only_first_wordpiece.ipynb) using only the first wordpiece of each word in the word label during tokenization. To propagate the label of the word to all wordpieces, see this [version](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/BERT/Custom_Named_Entity_Recognition_with_BERT.ipynb) of the notebook instead.
+- [`BertForTokenClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/token-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/token_classification.ipynb).
+- [`TFBertForTokenClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/token-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/token_classification-tf.ipynb).
+- [`FlaxBertForTokenClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/flax/token-classification).
+- [Token classification](https://huggingface.co/course/chapter7/2?fw=pt) chapter of the 🤗 Hugging Face Course.
+
+<PipelineTag pipeline="fill-mask"/>
+
+- [`BertForMaskedLM`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling#robertabertdistilbert-and-masked-language-modeling) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling.ipynb).
+- [`TFBertForMaskedLM`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/language-modeling#run_mlmpy) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling-tf.ipynb).
+- [`FlaxBertForMaskedLM`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/flax/language-modeling#masked-language-modeling) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/masked_language_modeling_flax.ipynb).
+- [Masked language modeling](https://huggingface.co/course/chapter7/3?fw=pt) chapter of the 🤗 Hugging Face Course.
+
+<PipelineTag pipeline="question-answering"/>
+
+- [`BertForQuestionAnswering`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/question-answering) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/question_answering.ipynb).
+- [`TFBertForQuestionAnswering`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/question-answering) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/question_answering-tf.ipynb).
+- [`FlaxBertForQuestionAnswering`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/flax/question-answering).
+- [Question answering](https://huggingface.co/course/chapter7/7?fw=pt) chapter of the 🤗 Hugging Face Course.
+
+**Multiple choice**
+- [`BertForMultipleChoice`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/multiple-choice) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/multiple_choice.ipynb).
+- [`TFBertForMultipleChoice`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/multiple-choice) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/multiple_choice-tf.ipynb).
+
+⚡️ **Inference**
+- A blog post on how to [Accelerate BERT inference with Hugging Face Transformers and AWS Inferentia](https://huggingface.co/blog/bert-inferentia-sagemaker).
+- A blog post on how to [Accelerate BERT inference with DeepSpeed-Inference on GPUs](https://www.philschmid.de/bert-deepspeed-inference).
+
+⚙️ **Pretraining**
+- A blog post on [Pre-Training BERT with Hugging Face Transformers and Habana Gaudi](https://www.philschmid.de/pre-training-bert-habana).
+
+🚀 **Deploy**
+- A blog post on how to [Convert Transformers to ONNX with Hugging Face Optimum](https://www.philschmid.de/convert-transformers-to-onnx).
+- A blog post on how to [Setup Deep Learning environment for Hugging Face Transformers with Habana Gaudi on AWS](https://www.philschmid.de/getting-started-habana-gaudi#conclusion).
+- A blog post on [Autoscaling BERT with Hugging Face Transformers, Amazon SageMaker and Terraform module](https://www.philschmid.de/terraform-huggingface-amazon-sagemaker-advanced).
+- A blog post on [Serverless BERT with HuggingFace, AWS Lambda, and Docker](https://www.philschmid.de/serverless-bert-with-huggingface-aws-lambda-docker).
+- A blog post on [Hugging Face Transformers BERT fine-tuning using Amazon SageMaker and Training Compiler](https://www.philschmid.de/huggingface-amazon-sagemaker-training-compiler).
+- A blog post on [Task-specific knowledge distillation for BERT using Transformers & Amazon SageMaker](https://www.philschmid.de/knowledge-distillation-bert-transformers).
+
 ## BertConfig
 
 [[autodoc]] BertConfig

docs/source/en/model_doc/big_bird.mdx

@@ -46,6 +46,8 @@ Tips:
 - Sequence length must be divisible by block size.
 - Current implementation supports only **ITC**.
 - Current implementation doesn't support **num_random_blocks = 0**
+- BigBird is a model with absolute position embeddings so it's usually advised to pad the inputs on the right rather than
+  the left.
 
 This model was contributed by [vasudevgupta](https://huggingface.co/vasudevgupta). The original code can be found
 [here](https://github.com/google-research/bigbird).

docs/source/en/model_doc/bigbird_pegasus.mdx

@@ -47,6 +47,8 @@ Tips:
 - Current implementation supports only **ITC**.
 - Current implementation doesn't support **num_random_blocks = 0**.
 - BigBirdPegasus uses the [PegasusTokenizer](https://github.com/huggingface/transformers/blob/main/src/transformers/models/pegasus/tokenization_pegasus.py).
+- BigBird is a model with absolute position embeddings so it's usually advised to pad the inputs on the right rather than
+  the left.
 
 The original code can be found [here](https://github.com/google-research/bigbird).
 

docs/source/en/model_doc/blenderbot-small.mdx

@@ -36,6 +36,11 @@ and code publicly available. Human evaluations show our best models are superior
 dialogue in terms of engagingness and humanness measurements. We then discuss the limitations of this work by analyzing
 failure cases of our models.*
 
+Tips:
+
+- Blenderbot Small is a model with absolute position embeddings so it's usually advised to pad the inputs on the right rather than
+  the left.
+
 This model was contributed by [patrickvonplaten](https://huggingface.co/patrickvonplaten). The authors' code can be
 found [here](https://github.com/facebookresearch/ParlAI) .
 

docs/source/en/model_doc/blenderbot.mdx

@@ -32,6 +32,11 @@ and code publicly available. Human evaluations show our best models are superior
 dialogue in terms of engagingness and humanness measurements. We then discuss the limitations of this work by analyzing
 failure cases of our models.*
 
+Tips:
+
+- Blenderbot is a model with absolute position embeddings so it's usually advised to pad the inputs on the right rather than
+  the left.
+
 This model was contributed by [sshleifer](https://huggingface.co/sshleifer). The authors' code can be found [here](https://github.com/facebookresearch/ParlAI) .
 
 

docs/source/en/model_doc/bloom.mdx

@@ -25,6 +25,21 @@ Several smaller versions of the models have been trained on the same dataset. BL
 - [bloom-7b1](https://huggingface.co/bigscience/bloom-7b1)
 - [bloom](https://huggingface.co/bigscience/bloom) (176B parameters)
 
+## Resources
+
+
+A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with BLOOM. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
+
+<PipelineTag pipeline="text-generation"/>
+
+- [`BloomForCausalLM`] is supported by this [causal language modeling example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling#gpt-2gpt-and-causal-language-modeling) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling.ipynb).
+
+⚡️ Inference
+- A blog on [Optimization story: Bloom inference](https://huggingface.co/blog/bloom-inference-optimization).
+- A blog on [Incredibly Fast BLOOM Inference with DeepSpeed and Accelerate](https://huggingface.co/blog/bloom-inference-pytorch-scripts).
+
+⚙️ Training
+- A blog on [The Technology Behind BLOOM Training](https://huggingface.co/blog/bloom-megatron-deepspeed).
 
 ## BloomConfig
 
@@ -55,3 +70,8 @@ Several smaller versions of the models have been trained on the same dataset. BL
 
 [[autodoc]] BloomForTokenClassification
     - forward
+
+## BloomForQuestionAnswering
+
+[[autodoc]] BloomForQuestionAnswering
+    - forward

docs/source/en/model_doc/chinese_clip.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.
+-->
+
+# Chinese-CLIP
+
+## Overview
+
+The Chinese-CLIP model was proposed in [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://arxiv.org/abs/2211.01335) by An Yang, Junshu Pan, Junyang Lin, Rui Men, Yichang Zhang, Jingren Zhou, Chang Zhou.
+Chinese-CLIP is an implementation of CLIP (Radford et al., 2021) on a large-scale dataset of Chinese image-text pairs. It is capable of performing cross-modal retrieval and also playing as a vision backbone for vision tasks like zero-shot image classification, open-domain object detection, etc. The original Chinese-CLIP code is released [at this link](https://github.com/OFA-Sys/Chinese-CLIP).
+
+The abstract from the paper is the following:
+
+*The tremendous success of CLIP (Radford et al., 2021) has promoted the research and application of contrastive learning for vision-language pretraining. In this work, we construct a large-scale dataset of image-text pairs in Chinese, where most data are retrieved from publicly available datasets, and we pretrain Chinese CLIP models on the new dataset. We develop 5 Chinese CLIP models of multiple sizes, spanning from 77 to 958 million parameters. Furthermore, we propose a two-stage pretraining method, where the model is first trained with the image encoder frozen and then trained with all parameters being optimized, to achieve enhanced model performance. Our comprehensive experiments demonstrate that Chinese CLIP can achieve the state-of-the-art performance on MUGE, Flickr30K-CN, and COCO-CN in the setups of zero-shot learning and finetuning, and it is able to achieve competitive performance in zero-shot image classification based on the evaluation on the ELEVATER benchmark (Li et al., 2022). Our codes, pretrained models, and demos have been released.*
+
+## Usage
+
+The code snippet below shows how to compute image & text features and similarities:
+
+```python
+>>> from PIL import Image
+>>> import requests
+>>> from transformers import ChineseCLIPProcessor, ChineseCLIPModel
+
+>>> model = ChineseCLIPModel.from_pretrained("OFA-Sys/chinese-clip-vit-base-patch16")
+>>> processor = ChineseCLIPProcessor.from_pretrained("OFA-Sys/chinese-clip-vit-base-patch16")
+
+>>> url = "https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/pokemon.jpeg"
+>>> image = Image.open(requests.get(url, stream=True).raw)
+>>> # Squirtle, Bulbasaur, Charmander, Pikachu in English
+>>> texts = ["杰尼龟", "妙蛙种子", "小火龙", "皮卡丘"]
+
+>>> # compute image feature
+>>> inputs = processor(images=image, return_tensors="pt")
+>>> image_features = model.get_image_features(**inputs)
+>>> image_features = image_features / image_features.norm(p=2, dim=-1, keepdim=True)  # normalize
+
+>>> # compute text features
+>>> inputs = processor(text=texts, padding=True, return_tensors="pt")
+>>> text_features = model.get_text_features(**inputs)
+>>> text_features = text_features / text_features.norm(p=2, dim=-1, keepdim=True)  # normalize
+
+>>> # compute image-text similarity scores
+>>> inputs = processor(text=texts, images=image, return_tensors="pt", padding=True)
+>>> outputs = model(**inputs)
+>>> logits_per_image = outputs.logits_per_image  # this is the image-text similarity score
+>>> probs = logits_per_image.softmax(dim=1)  # probs: [[1.2686e-03, 5.4499e-02, 6.7968e-04, 9.4355e-01]]
+```
+
+Currently, we release the following scales of pretrained Chinese-CLIP models at HF Model Hub:
+
+- [OFA-Sys/chinese-clip-vit-base-patch16](https://huggingface.co/OFA-Sys/chinese-clip-vit-base-patch16)
+- [OFA-Sys/chinese-clip-vit-large-patch14](https://huggingface.co/OFA-Sys/chinese-clip-vit-large-patch14)
+- [OFA-Sys/chinese-clip-vit-large-patch14-336px](https://huggingface.co/OFA-Sys/chinese-clip-vit-large-patch14-336px)
+- [OFA-Sys/chinese-clip-vit-huge-patch14](https://huggingface.co/OFA-Sys/chinese-clip-vit-huge-patch14)
+
+The Chinese-CLIP model was contributed by [OFA-Sys](https://huggingface.co/OFA-Sys). 
+
+## ChineseCLIPConfig
+
+[[autodoc]] ChineseCLIPConfig
+    - from_text_vision_configs
+
+## ChineseCLIPTextConfig
+
+[[autodoc]] ChineseCLIPTextConfig
+
+## ChineseCLIPVisionConfig
+
+[[autodoc]] ChineseCLIPVisionConfig
+
+## ChineseCLIPImageProcessor
+
+[[autodoc]] ChineseCLIPImageProcessor
+    - preprocess
+
+## ChineseCLIPFeatureExtractor
+
+[[autodoc]] ChineseCLIPFeatureExtractor
+
+## ChineseCLIPProcessor
+
+[[autodoc]] ChineseCLIPProcessor
+
+## ChineseCLIPModel
+
+[[autodoc]] ChineseCLIPModel
+    - forward
+    - get_text_features
+    - get_image_features
+
+## ChineseCLIPTextModel
+
+[[autodoc]] ChineseCLIPTextModel
+    - forward
+
+## ChineseCLIPVisionModel
+
+[[autodoc]] ChineseCLIPVisionModel
+    - forward

docs/source/en/model_doc/clip.mdx

@@ -75,6 +75,25 @@ encode the text and prepare the images. The following example shows how to get t
 
 This model was contributed by [valhalla](https://huggingface.co/valhalla). The original code can be found [here](https://github.com/openai/CLIP).
 
+## Resources
+
+A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with CLIP. If you're
+interested in submitting a resource to be included here, please feel free to open a Pull Request and we will review it.
+The resource should ideally demonstrate something new instead of duplicating an existing resource.
+
+<PipelineTag pipeline="text-to-image"/>
+- A blog post on [How to use CLIP to retrieve images from text](https://huggingface.co/blog/fine-tune-clip-rsicd).
+- A blog bost on [How to use CLIP for Japanese text to image generation](https://huggingface.co/blog/japanese-stable-diffusion).
+
+
+<PipelineTag pipeline="image-to-text"/>
+- A notebook showing [Video to text matching with CLIP for videos](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/X-CLIP/Video_text_matching_with_X_CLIP.ipynb).
+
+
+<PipelineTag pipeline="zero-shot-classification"/>
+- A notebook showing [Zero shot video classification using CLIP for video](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/X-CLIP/Zero_shot_classify_a_YouTube_video_with_X_CLIP.ipynb).
+
+
 ## CLIPConfig
 
 [[autodoc]] CLIPConfig
@@ -100,6 +119,11 @@ This model was contributed by [valhalla](https://huggingface.co/valhalla). The o
 
 [[autodoc]] CLIPTokenizerFast
 
+## CLIPImageProcessor
+
+[[autodoc]] CLIPImageProcessor
+    - preprocess
+
 ## CLIPFeatureExtractor
 
 [[autodoc]] CLIPFeatureExtractor
@@ -120,6 +144,17 @@ This model was contributed by [valhalla](https://huggingface.co/valhalla). The o
 [[autodoc]] CLIPTextModel
     - forward
 
+## CLIPTextModelWithProjection
+
+[[autodoc]] CLIPTextModelWithProjection
+    - forward
+
+## CLIPVisionModelWithProjection
+
+[[autodoc]] CLIPVisionModelWithProjection
+    - forward
+
+
 ## CLIPVisionModel
 
 [[autodoc]] CLIPVisionModel

docs/source/en/model_doc/clipseg.mdx

@@ -0,0 +1,100 @@
+<!--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.
+-->
+
+# CLIPSeg
+
+## Overview
+
+The CLIPSeg model was proposed in [Image Segmentation Using Text and Image Prompts](https://arxiv.org/abs/2112.10003) by Timo Lüddecke
+and Alexander Ecker. CLIPSeg adds a minimal decoder on top of a frozen [CLIP](clip) model for zero- and one-shot image segmentation.
+
+The abstract from the paper is the following:
+
+*Image segmentation is usually addressed by training a
+model for a fixed set of object classes. Incorporating additional classes or more complex queries later is expensive
+as it requires re-training the model on a dataset that encompasses these expressions. Here we propose a system
+that can generate image segmentations based on arbitrary
+prompts at test time. A prompt can be either a text or an
+image. This approach enables us to create a unified model
+(trained once) for three common segmentation tasks, which
+come with distinct challenges: referring expression segmentation, zero-shot segmentation and one-shot segmentation.
+We build upon the CLIP model as a backbone which we extend with a transformer-based decoder that enables dense
+prediction. After training on an extended version of the
+PhraseCut dataset, our system generates a binary segmentation map for an image based on a free-text prompt or on
+an additional image expressing the query. We analyze different variants of the latter image-based prompts in detail.
+This novel hybrid input allows for dynamic adaptation not
+only to the three segmentation tasks mentioned above, but
+to any binary segmentation task where a text or image query
+can be formulated. Finally, we find our system to adapt well
+to generalized queries involving affordances or properties*
+
+Tips:
+
+- [`CLIPSegForImageSegmentation`] adds a decoder on top of [`CLIPSegModel`]. The latter is identical to [`CLIPModel`].
+- [`CLIPSegForImageSegmentation`] can generate image segmentations based on arbitrary prompts at test time. A prompt can be either a text
+(provided to the model as `input_ids`) or an image (provided to the model as `conditional_pixel_values`). One can also provide custom
+conditional embeddings (provided to the model as `conditional_embeddings`).
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/clipseg_architecture.png"
+alt="drawing" width="600"/> 
+
+<small> CLIPSeg overview. Taken from the <a href="https://arxiv.org/abs/2112.10003">original paper.</a> </small>
+
+This model was contributed by [nielsr](https://huggingface.co/nielsr).
+The original code can be found [here](https://github.com/timojl/clipseg).
+
+## Resources
+
+A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with CLIPSeg. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
+
+<PipelineTag pipeline="image-segmentation"/>
+
+- A notebook that illustrates [zero-shot image segmentation with CLIPSeg](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/CLIPSeg/Zero_shot_image_segmentation_with_CLIPSeg.ipynb).
+
+## CLIPSegConfig
+
+[[autodoc]] CLIPSegConfig
+    - from_text_vision_configs
+
+## CLIPSegTextConfig
+
+[[autodoc]] CLIPSegTextConfig
+
+## CLIPSegVisionConfig
+
+[[autodoc]] CLIPSegVisionConfig
+
+## CLIPSegProcessor
+
+[[autodoc]] CLIPSegProcessor
+
+## CLIPSegModel
+
+[[autodoc]] CLIPSegModel
+    - forward
+    - get_text_features
+    - get_image_features
+
+## CLIPSegTextModel
+
+[[autodoc]] CLIPSegTextModel
+    - forward
+
+## CLIPSegVisionModel
+
+[[autodoc]] CLIPSegVisionModel
+    - forward
+
+## CLIPSegForImageSegmentation
+
+[[autodoc]] CLIPSegForImageSegmentation
+    - forward

docs/source/en/model_doc/conditional_detr.mdx

@@ -0,0 +1,68 @@
+<!--Copyright 2022 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# Conditional DETR
+
+## Overview
+
+The Conditional DETR model was proposed in [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang. Conditional DETR presents a conditional cross-attention mechanism for fast DETR training. Conditional DETR converges 6.7× to 10× faster than DETR.
+
+The abstract from the paper is the following:
+
+*The recently-developed DETR approach applies the transformer encoder and decoder architecture to object detection and achieves promising performance. In this paper, we handle the critical issue, slow training convergence, and present a conditional cross-attention mechanism for fast DETR training. Our approach is motivated by that the cross-attention in DETR relies highly on the content embeddings for localizing the four extremities and predicting the box, which increases the need for high-quality content embeddings and thus the training difficulty. Our approach, named conditional DETR, learns a conditional spatial query from the decoder embedding for decoder multi-head cross-attention. The benefit is that through the conditional spatial query, each cross-attention head is able to attend to a band containing a distinct region, e.g., one object extremity or a region inside the object box. This narrows down the spatial range for localizing the distinct regions for object classification and box regression, thus relaxing the dependence on the content embeddings and easing the training. Empirical results show that conditional DETR converges 6.7× faster for the backbones R50 and R101 and 10× faster for stronger backbones DC5-R50 and DC5-R101. Code is available at https://github.com/Atten4Vis/ConditionalDETR.*
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/conditional_detr_curve.jpg"
+alt="drawing" width="600"/>
+
+<small> Conditional DETR shows much faster convergence compared to the original DETR. Taken from the <a href="https://arxiv.org/abs/2108.06152">original paper</a>.</small>
+
+This model was contributed by [DepuMeng](https://huggingface.co/DepuMeng). The original code can be found [here](https://github.com/Atten4Vis/ConditionalDETR).
+
+
+## ConditionalDetrConfig
+
+[[autodoc]] ConditionalDetrConfig
+
+## ConditionalDetrImageProcessor
+
+[[autodoc]] ConditionalDetrImageProcessor
+    - preprocess
+    - pad_and_create_pixel_mask
+    - post_process_object_detection
+    - post_process_instance_segmentation
+    - post_process_semantic_segmentation
+    - post_process_panoptic_segmentation
+
+## ConditionalDetrFeatureExtractor
+
+[[autodoc]] ConditionalDetrFeatureExtractor
+    - __call__
+    - pad_and_create_pixel_mask
+    - post_process_object_detection
+    - post_process_instance_segmentation
+    - post_process_semantic_segmentation
+    - post_process_panoptic_segmentation
+
+## ConditionalDetrModel
+
+[[autodoc]] ConditionalDetrModel
+    - forward
+
+## ConditionalDetrForObjectDetection
+
+[[autodoc]] ConditionalDetrForObjectDetection
+    - forward
+
+## ConditionalDetrForSegmentation
+
+[[autodoc]] ConditionalDetrForSegmentation
+    - forward

docs/source/en/model_doc/convnext.mdx

@@ -33,7 +33,7 @@ Tips:
 - See the code examples below each model regarding usage.
 
 <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/convnext_architecture.jpg"
-alt="drawing" width="600"/> 
+alt="drawing" width="600"/>
 
 <small> ConvNeXT architecture. Taken from the <a href="https://arxiv.org/abs/2201.03545">original paper</a>.</small>
 
@@ -50,6 +50,11 @@ This model was contributed by [nielsr](https://huggingface.co/nielsr). TensorFlo
 [[autodoc]] ConvNextFeatureExtractor
 
 
+## ConvNextImageProcessor
+
+[[autodoc]] ConvNextImageProcessor
+    - preprocess
+
 ## ConvNextModel
 
 [[autodoc]] ConvNextModel
@@ -71,4 +76,4 @@ This model was contributed by [nielsr](https://huggingface.co/nielsr). TensorFlo
 ## TFConvNextForImageClassification
 
 [[autodoc]] TFConvNextForImageClassification
-    - call
+    - call

docs/source/en/model_doc/cvt.mdx

@@ -32,7 +32,7 @@ a crucial component in existing Vision Transformers, can be safely removed in ou
 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`]).
+- 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 [`AutoImageProcessor`] 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).
 
@@ -51,3 +51,14 @@ This model was contributed by [anugunj](https://huggingface.co/anugunj). The ori
 
 [[autodoc]] CvtForImageClassification
     - forward
+
+## TFCvtModel
+
+[[autodoc]] TFCvtModel
+    - call
+
+## TFCvtForImageClassification
+
+[[autodoc]] TFCvtForImageClassification
+    - call
+

docs/source/en/model_doc/deberta.mdx

@@ -38,6 +38,35 @@ pre-trained models will be made publicly available at https://github.com/microso
 This model was contributed by [DeBERTa](https://huggingface.co/DeBERTa). This model TF 2.0 implementation was
 contributed by [kamalkraj](https://huggingface.co/kamalkraj) . The original code can be found [here](https://github.com/microsoft/DeBERTa).
 
+## Resources
+
+A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with DeBERTa. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
+
+<PipelineTag pipeline="text-classification"/>
+
+- A blog post on how to [Accelerate Large Model Training using DeepSpeed](https://huggingface.co/blog/accelerate-deepspeed) with DeBERTa.
+- A blog post on [Supercharged Customer Service with Machine Learning](https://huggingface.co/blog/supercharge-customer-service-with-machine-learning) with DeBERTa.
+- [`DebertaForSequenceClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/text-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification.ipynb).
+- [`TFDebertaForSequenceClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/text-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification-tf.ipynb).
+
+<PipelineTag pipeline="token-classification" />
+
+- [`DebertaForTokenClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/token-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/token_classification.ipynb).
+- [`TFDebertaForTokenClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/token-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/token_classification-tf.ipynb).
+- [Token classification](https://huggingface.co/course/chapter7/2?fw=pt) chapter of the 🤗 Hugging Face Course.
+- [Byte-Pair Encoding tokenization](https://huggingface.co/course/chapter6/5?fw=pt) chapter of the 🤗 Hugging Face Course.
+
+<PipelineTag pipeline="fill-mask"/>
+
+- [`DebertaForMaskedLM`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling#robertabertdistilbert-and-masked-language-modeling) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling.ipynb).
+- [`TFDebertaForMaskedLM`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/language-modeling#run_mlmpy) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling-tf.ipynb).
+- [Masked language modeling](https://huggingface.co/course/chapter7/3?fw=pt) chapter of the 🤗 Hugging Face Course.
+
+<PipelineTag pipeline="question-answering"/>
+
+- [`DebertaForQuestionAnswering`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/question-answering) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/question_answering.ipynb).
+- [`TFDebertaForQuestionAnswering`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/question-answering) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/question_answering-tf.ipynb).
+- [Question answering](https://huggingface.co/course/chapter7/7?fw=pt) chapter of the 🤗 Hugging Face Course.
 
 ## DebertaConfig
 

docs/source/en/model_doc/deformable_detr.mdx

@@ -0,0 +1,65 @@
+<!--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.
+-->
+
+# Deformable DETR
+
+## Overview
+
+The Deformable DETR model was proposed in [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://arxiv.org/abs/2010.04159) by Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, Jifeng Dai.
+Deformable DETR mitigates the slow convergence issues and limited feature spatial resolution of the original [DETR](detr) by leveraging a new deformable attention module which only attends to a small set of key sampling points around a reference.
+
+The abstract from the paper is the following:
+
+*DETR has been recently proposed to eliminate the need for many hand-designed components in object detection while demonstrating good performance. However, it suffers from slow convergence and limited feature spatial resolution, due to the limitation of Transformer attention modules in processing image feature maps. To mitigate these issues, we proposed Deformable DETR, whose attention modules only attend to a small set of key sampling points around a reference. Deformable DETR can achieve better performance than DETR (especially on small objects) with 10 times less training epochs. Extensive experiments on the COCO benchmark demonstrate the effectiveness of our approach.*
+
+Tips:
+
+- One can use [`DeformableDetrImageProcessor`] to prepare images (and optional targets) for the model.
+- Training Deformable DETR is equivalent to training the original [DETR](detr) model. Demo notebooks can be found [here](https://github.com/NielsRogge/Transformers-Tutorials/tree/master/DETR).
+
+<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/deformable_detr_architecture.png"
+alt="drawing" width="600"/>
+
+<small> Deformable DETR architecture. Taken from the <a href="https://arxiv.org/abs/2010.04159">original paper</a>.</small>
+
+This model was contributed by [nielsr](https://huggingface.co/nielsr). The original code can be found [here](https://github.com/fundamentalvision/Deformable-DETR).
+
+## DeformableDetrImageProcessor
+
+[[autodoc]] DeformableDetrImageProcessor
+    - preprocess
+    - pad_and_create_pixel_mask
+    - post_process_object_detection
+
+## DeformableDetrFeatureExtractor
+
+[[autodoc]] DeformableDetrFeatureExtractor
+    - __call__
+    - pad_and_create_pixel_mask
+    - post_process_object_detection
+
+
+## DeformableDetrConfig
+
+[[autodoc]] DeformableDetrConfig
+
+
+## DeformableDetrModel
+
+[[autodoc]] DeformableDetrModel
+    - forward
+
+
+## DeformableDetrForObjectDetection
+
+[[autodoc]] DeformableDetrForObjectDetection
+    - forward

docs/source/en/model_doc/deit.mdx

@@ -66,7 +66,7 @@ Tips:
   augmentation, optimization, and regularization were used in order to simulate training on a much larger dataset
   (while only using ImageNet-1k for pre-training). There are 4 variants available (in 3 different sizes):
   *facebook/deit-tiny-patch16-224*, *facebook/deit-small-patch16-224*, *facebook/deit-base-patch16-224* and
-  *facebook/deit-base-patch16-384*. Note that one should use [`DeiTFeatureExtractor`] in order to
+  *facebook/deit-base-patch16-384*. Note that one should use [`DeiTImageProcessor`] in order to
   prepare images for the model.
 
 This model was contributed by [nielsr](https://huggingface.co/nielsr). The TensorFlow version of this model was added by [amyeroberts](https://huggingface.co/amyeroberts).
@@ -81,6 +81,11 @@ This model was contributed by [nielsr](https://huggingface.co/nielsr). The Tenso
 [[autodoc]] DeiTFeatureExtractor
     - __call__
 
+## DeiTImageProcessor
+
+[[autodoc]] DeiTImageProcessor
+    - preprocess
+
 ## DeiTModel
 
 [[autodoc]] DeiTModel

docs/source/en/model_doc/detr.mdx

@@ -105,21 +105,21 @@ Tips:
 - DETR resizes the input images such that the shortest side is at least a certain amount of pixels while the longest is
   at most 1333 pixels. At training time, scale augmentation is used such that the shortest side is randomly set to at
   least 480 and at most 800 pixels. At inference time, the shortest side is set to 800. One can use
-  [`~transformers.DetrFeatureExtractor`] to prepare images (and optional annotations in COCO format) for the
+  [`~transformers.DetrImageProcessor`] to prepare images (and optional annotations in COCO format) for the
   model. Due to this resizing, images in a batch can have different sizes. DETR solves this by padding images up to the
   largest size in a batch, and by creating a pixel mask that indicates which pixels are real/which are padding.
   Alternatively, one can also define a custom `collate_fn` in order to batch images together, using
-  [`~transformers.DetrFeatureExtractor.pad_and_create_pixel_mask`].
+  [`~transformers.DetrImageProcessor.pad_and_create_pixel_mask`].
 - 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")
+>>> model = DetrForObjectDetection.from_pretrained("facebook/detr-resnet-50")
 ```
 
 Option 2: Instantiate DETR with randomly initialized weights for Transformer, but pre-trained weights for backbone
@@ -142,14 +142,14 @@ As a summary, consider the following table:
 | **Description** | Predicting bounding boxes and class labels around objects in an image | Predicting masks around objects (i.e. instances) in an image | Predicting masks around both objects (i.e. instances) as well as "stuff" (i.e. background things like trees and roads) in an image |
 | **Model** | [`~transformers.DetrForObjectDetection`] | [`~transformers.DetrForSegmentation`] | [`~transformers.DetrForSegmentation`] |
 | **Example dataset** | COCO detection | COCO detection, COCO panoptic | COCO panoptic  |                                                                        |
-| **Format of annotations to provide to**  [`~transformers.DetrFeatureExtractor`] | {'image_id': `int`, 'annotations': `List[Dict]`} each Dict being a COCO object annotation  | {'image_id': `int`, 'annotations': `List[Dict]`}  (in case of COCO detection) or {'file_name': `str`, 'image_id': `int`, 'segments_info': `List[Dict]`} (in case of COCO panoptic) | {'file_name': `str`, 'image_id': `int`, 'segments_info': `List[Dict]`} and masks_path (path to directory containing PNG files of the masks) |
-| **Postprocessing** (i.e. converting the output of the model to COCO API) | [`~transformers.DetrFeatureExtractor.post_process`] | [`~transformers.DetrFeatureExtractor.post_process_segmentation`] | [`~transformers.DetrFeatureExtractor.post_process_segmentation`], [`~transformers.DetrFeatureExtractor.post_process_panoptic`] |
+| **Format of annotations to provide to**  [`~transformers.DetrImageProcessor`] | {'image_id': `int`, 'annotations': `List[Dict]`} each Dict being a COCO object annotation  | {'image_id': `int`, 'annotations': `List[Dict]`}  (in case of COCO detection) or {'file_name': `str`, 'image_id': `int`, 'segments_info': `List[Dict]`} (in case of COCO panoptic) | {'file_name': `str`, 'image_id': `int`, 'segments_info': `List[Dict]`} and masks_path (path to directory containing PNG files of the masks) |
+| **Postprocessing** (i.e. converting the output of the model to COCO API) | [`~transformers.DetrImageProcessor.post_process`] | [`~transformers.DetrImageProcessor.post_process_segmentation`] | [`~transformers.DetrImageProcessor.post_process_segmentation`], [`~transformers.DetrImageProcessor.post_process_panoptic`] |
 | **evaluators** | `CocoEvaluator` with `iou_types="bbox"` | `CocoEvaluator` with `iou_types="bbox"` or `"segm"` | `CocoEvaluator` with `iou_tupes="bbox"` or `"segm"`, `PanopticEvaluator` |
 
 In short, one should prepare the data either in COCO detection or COCO panoptic format, then use
-[`~transformers.DetrFeatureExtractor`] to create `pixel_values`, `pixel_mask` and optional
+[`~transformers.DetrImageProcessor`] to create `pixel_values`, `pixel_mask` and optional
 `labels`, which can then be used to train (or fine-tune) a model. For evaluation, one should first convert the
-outputs of the model using one of the postprocessing methods of [`~transformers.DetrFeatureExtractor`]. These can
+outputs of the model using one of the postprocessing methods of [`~transformers.DetrImageProcessor`]. These can
 be be provided to either `CocoEvaluator` or `PanopticEvaluator`, which allow you to calculate metrics like
 mean Average Precision (mAP) and Panoptic Quality (PQ). The latter objects are implemented in the [original repository](https://github.com/facebookresearch/detr). See the [example notebooks](https://github.com/NielsRogge/Transformers-Tutorials/tree/master/DETR) for more info regarding evaluation.
 
@@ -166,14 +166,24 @@ mean Average Precision (mAP) and Panoptic Quality (PQ). The latter objects are i
 
 [[autodoc]] DetrConfig
 
+## DetrImageProcessor
+
+[[autodoc]] DetrImageProcessor
+    - preprocess
+    - post_process_object_detection
+    - post_process_semantic_segmentation
+    - post_process_instance_segmentation
+    - post_process_panoptic_segmentation
+
 ## DetrFeatureExtractor
 
 [[autodoc]] DetrFeatureExtractor
     - __call__
     - pad_and_create_pixel_mask
-    - post_process
-    - post_process_segmentation
-    - post_process_panoptic
+    - post_process_object_detection
+    - post_process_semantic_segmentation
+    - post_process_instance_segmentation
+    - post_process_panoptic_segmentation
 
 ## DetrModel
 

docs/source/en/model_doc/dinat.mdx

@@ -0,0 +1,78 @@
+<!--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.
+-->
+
+# Dilated Neighborhood Attention Transformer
+
+## Overview
+
+DiNAT was proposed in [Dilated Neighborhood Attention Transformer](https://arxiv.org/abs/2209.15001)
+by Ali Hassani and Humphrey Shi.
+
+It extends [NAT](nat) by adding a Dilated Neighborhood Attention pattern to capture global context,
+and shows significant performance improvements over it.
+
+The abstract from the paper is the following:
+
+*Transformers are quickly becoming one of the most heavily applied deep learning architectures across modalities,
+domains, and tasks. In vision, on top of ongoing efforts into plain transformers, hierarchical transformers have
+also gained significant attention, thanks to their performance and easy integration into existing frameworks.
+These models typically employ localized attention mechanisms, such as the sliding-window Neighborhood Attention (NA)
+or Swin Transformer's Shifted Window Self Attention. While effective at reducing self attention's quadratic complexity,
+local attention weakens two of the most desirable properties of self attention: long range inter-dependency modeling,
+and global receptive field. In this paper, we introduce Dilated Neighborhood Attention (DiNA), a natural, flexible and
+efficient extension to NA that can capture more global context and expand receptive fields exponentially at no
+additional cost. NA's local attention and DiNA's sparse global attention complement each other, and therefore we
+introduce Dilated Neighborhood Attention Transformer (DiNAT), a new hierarchical vision transformer built upon both.
+DiNAT variants enjoy significant improvements over strong baselines such as NAT, Swin, and ConvNeXt.
+Our large model is faster and ahead of its Swin counterpart by 1.5% box AP in COCO object detection,
+1.3% mask AP in COCO instance segmentation, and 1.1% mIoU in ADE20K semantic segmentation.
+Paired with new frameworks, our large variant is the new state of the art panoptic segmentation model on COCO (58.2 PQ)
+and ADE20K (48.5 PQ), and instance segmentation model on Cityscapes (44.5 AP) and ADE20K (35.4 AP) (no extra data).
+It also matches the state of the art specialized semantic segmentation models on ADE20K (58.2 mIoU),
+and ranks second on Cityscapes (84.5 mIoU) (no extra data). *
+
+Tips:
+- One can use the [`AutoImageProcessor`] API to prepare images for the model.
+- DiNAT 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, height, width, num_channels)`.
+
+Notes:
+- DiNAT depends on [NATTEN](https://github.com/SHI-Labs/NATTEN/)'s implementation of Neighborhood Attention and Dilated Neighborhood Attention.
+You can install it with pre-built wheels for Linux by referring to [shi-labs.com/natten](https://shi-labs.com/natten), or build on your system by running `pip install natten`.
+Note that the latter will likely take time to compile. NATTEN does not support Windows devices yet.
+- Patch size of 4 is only supported at the moment.
+
+<img
+src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/dilated-neighborhood-attention-pattern.jpg"
+alt="drawing" width="600"/>
+
+<small> Neighborhood Attention with different dilation values.
+Taken from the <a href="https://arxiv.org/abs/2209.15001">original paper</a>.</small>
+
+This model was contributed by [Ali Hassani](https://huggingface.co/alihassanijr).
+The original code can be found [here](https://github.com/SHI-Labs/Neighborhood-Attention-Transformer).
+
+
+## DinatConfig
+
+[[autodoc]] DinatConfig
+
+
+## DinatModel
+
+[[autodoc]] DinatModel
+    - forward
+
+## DinatForImageClassification
+
+[[autodoc]] DinatForImageClassification
+    - forward

docs/source/en/model_doc/distilbert.mdx

@@ -45,6 +45,66 @@ Tips:
 This model was contributed by [victorsanh](https://huggingface.co/victorsanh). This model jax version was
 contributed by [kamalkraj](https://huggingface.co/kamalkraj). The original code can be found [here](https://github.com/huggingface/transformers/tree/main/examples/research_projects/distillation).
 
+## Resources
+
+A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with DistilBERT. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
+
+<PipelineTag pipeline="text-classification"/>
+
+- A blog post on [Getting Started with Sentiment Analysis using Python](https://huggingface.co/blog/sentiment-analysis-python) with DistilBERT.
+- A blog post on how to [train DistilBERT with Blurr for sequence classification](https://huggingface.co/blog/fastai).
+- A blog post on how to use [Ray to tune DistilBERT hyperparameters](https://huggingface.co/blog/ray-tune).
+- A blog post on how to [train DistilBERT with Hugging Face and Amazon SageMaker](https://huggingface.co/blog/the-partnership-amazon-sagemaker-and-hugging-face).
+- A notebook on how to [finetune DistilBERT for multi-label classification](https://colab.research.google.com/github/DhavalTaunk08/Transformers_scripts/blob/master/Transformers_multilabel_distilbert.ipynb). 🌎
+- A notebook on how to [finetune DistilBERT for multiclass classification with PyTorch](https://colab.research.google.com/github/abhimishra91/transformers-tutorials/blob/master/transformers_multiclass_classification.ipynb). 🌎
+- A notebook on how to [finetune DistilBERT for text classification in TensorFlow](https://colab.research.google.com/github/peterbayerle/huggingface_notebook/blob/main/distilbert_tf.ipynb). 🌎
+- [`DistilBertForSequenceClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/text-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification.ipynb).
+- [`TFDistilBertForSequenceClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/text-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification-tf.ipynb).
+- [`FlaxDistilBertForSequenceClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/flax/text-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification_flax.ipynb).
+
+
+<PipelineTag pipeline="token-classification"/>
+
+- [`DistilBertForTokenClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/token-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/token_classification.ipynb).
+- [`TFDistilBertForTokenClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/token-classification) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/token_classification-tf.ipynb).
+- [`FlaxDistilBertForTokenClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/flax/token-classification).
+- [Token classification](https://huggingface.co/course/chapter7/2?fw=pt) chapter of the 🤗 Hugging Face Course.
+
+
+<PipelineTag pipeline="fill-mask"/>
+
+- [`DistilBertForMaskedLM`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling#robertabertdistilbert-and-masked-language-modeling) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling.ipynb).
+- [`TFDistilBertForMaskedLM`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/language-modeling#run_mlmpy) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling-tf.ipynb).
+- [`FlaxDistilBertForMaskedLM`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/flax/language-modeling#masked-language-modeling) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/masked_language_modeling_flax.ipynb).
+- [Masked language modeling](https://huggingface.co/course/chapter7/3?fw=pt) chapter of the 🤗 Hugging Face Course.
+
+<PipelineTag pipeline="question-answering"/>
+
+- [`DistilBertForQuestionAnswering`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/question-answering) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/question_answering.ipynb).
+- [`TFDistilBertForQuestionAnswering`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/question-answering) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/question_answering-tf.ipynb).
+- [`FlaxDistilBertForQuestionAnswering`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/flax/question-answering).
+- [Question answering](https://huggingface.co/course/chapter7/7?fw=pt) chapter of the 🤗 Hugging Face Course.
+
+**Multiple choice**
+- [`DistilBertForMultipleChoice`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/multiple-choice) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/multiple_choice.ipynb).
+- [`TFDistilBertForMultipleChoice`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/multiple-choice) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/multiple_choice-tf.ipynb).
+
+⚗️ Optimization
+
+- A blog post on how to [quantize DistilBERT with 🤗 Optimum and Intel](https://huggingface.co/blog/intel).
+- A blog post on how [Optimizing Transformers for GPUs with 🤗 Optimum](https://www.philschmid.de/optimizing-transformers-with-optimum-gpu).
+- A blog post on [Optimizing Transformers with Hugging Face Optimum](https://www.philschmid.de/optimizing-transformers-with-optimum).
+
+⚡️ Inference
+
+- A blog post on how to [Accelerate BERT inference with Hugging Face Transformers and AWS Inferentia](https://huggingface.co/blog/bert-inferentia-sagemaker) with DistilBERT.
+- A blog post on [Serverless Inference with Hugging Face's Transformers, DistilBERT and Amazon SageMaker](https://www.philschmid.de/sagemaker-serverless-huggingface-distilbert).
+
+🚀 Deploy
+
+- A blog post on how to [deploy DistilBERT on Google Cloud](https://huggingface.co/blog/how-to-deploy-a-pipeline-to-google-clouds).
+- A blog post on how to [deploy DistilBERT with Amazon SageMaker](https://huggingface.co/blog/deploy-hugging-face-models-easily-with-amazon-sagemaker).
+- A blog post on how to [Deploy BERT with Hugging Face Transformers, Amazon SageMaker and Terraform module](https://www.philschmid.de/terraform-huggingface-amazon-sagemaker).
 
 ## DistilBertConfig
 

docs/source/en/model_doc/donut.mdx

@@ -23,7 +23,7 @@ The abstract from the paper is the following:
 *Understanding document images (e.g., invoices) is a core but challenging task since it requires complex functions such as reading text and a holistic understanding of the document. Current Visual Document Understanding (VDU) methods outsource the task of reading text to off-the-shelf Optical Character Recognition (OCR) engines and focus on the understanding task with the OCR outputs. Although such OCR-based approaches have shown promising performance, they suffer from 1) high computational costs for using OCR; 2) inflexibility of OCR models on languages or types of document; 3) OCR error propagation to the subsequent process. To address these issues, in this paper, we introduce a novel OCR-free VDU model named Donut, which stands for Document understanding transformer. As the first step in OCR-free VDU research, we propose a simple architecture (i.e., Transformer) with a pre-training objective (i.e., cross-entropy loss). Donut is conceptually simple yet effective. Through extensive experiments and analyses, we show a simple OCR-free VDU model, Donut, achieves state-of-the-art performances on various VDU tasks in terms of both speed and accuracy. In addition, we offer a synthetic data generator that helps the model pre-training to be flexible in various languages and domains.*
 
 <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/donut_architecture.jpg"
-alt="drawing" width="600"/> 
+alt="drawing" width="600"/>
 
 <small> Donut high-level overview. Taken from the <a href="https://arxiv.org/abs/2111.15664">original paper</a>. </small>
 
@@ -40,7 +40,7 @@ Tips:
 ## Inference
 
 Donut's [`VisionEncoderDecoder`] model accepts images as input and makes use of
-[`~generation_utils.GenerationMixin.generate`] to autoregressively generate text given the input image.
+[`~generation.GenerationMixin.generate`] to autoregressively generate text given the input image.
 
 The [`DonutFeatureExtractor`] class is responsible for preprocessing the input image and
 [`XLMRobertaTokenizer`/`XLMRobertaTokenizerFast`] decodes the generated target tokens to the target string. The
@@ -194,6 +194,11 @@ We refer to the [tutorial notebooks](https://github.com/NielsRogge/Transformers-
 
 [[autodoc]] DonutSwinConfig
 
+## DonutImageProcessor
+
+[[autodoc]] DonutImageProcessor
+    - preprocess
+
 ## DonutFeatureExtractor
 
 [[autodoc]] DonutFeatureExtractor
@@ -211,4 +216,4 @@ We refer to the [tutorial notebooks](https://github.com/NielsRogge/Transformers-
 ## DonutSwinModel
 
 [[autodoc]] DonutSwinModel
-    - forward
+    - forward

docs/source/en/model_doc/dpt.mdx

@@ -22,7 +22,7 @@ The abstract from the paper is the following:
 *We introduce dense vision transformers, an architecture that leverages vision transformers in place of convolutional networks as a backbone for dense prediction tasks. We assemble tokens from various stages of the vision transformer into image-like representations at various resolutions and progressively combine them into full-resolution predictions using a convolutional decoder. The transformer backbone processes representations at a constant and relatively high resolution and has a global receptive field at every stage. These properties allow the dense vision transformer to provide finer-grained and more globally coherent predictions when compared to fully-convolutional networks. Our experiments show that this architecture yields substantial improvements on dense prediction tasks, especially when a large amount of training data is available. For monocular depth estimation, we observe an improvement of up to 28% in relative performance when compared to a state-of-the-art fully-convolutional network. When applied to semantic segmentation, dense vision transformers set a new state of the art on ADE20K with 49.02% mIoU. We further show that the architecture can be fine-tuned on smaller datasets such as NYUv2, KITTI, and Pascal Context where it also sets the new state of the art.*
 
 <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/dpt_architecture.jpg"
-alt="drawing" width="600"/> 
+alt="drawing" width="600"/>
 
 <small> DPT architecture. Taken from the <a href="https://arxiv.org/abs/2103.13413" target="_blank">original paper</a>. </small>
 
@@ -37,6 +37,14 @@ This model was contributed by [nielsr](https://huggingface.co/nielsr). The origi
 
 [[autodoc]] DPTFeatureExtractor
     - __call__
+    - post_process_semantic_segmentation
+
+
+## DPTImageProcessor
+
+[[autodoc]] DPTImageProcessor
+    - preprocess
+    - post_process_semantic_segmentation
 
 
 ## DPTModel
@@ -54,4 +62,4 @@ This model was contributed by [nielsr](https://huggingface.co/nielsr). The origi
 ## DPTForSemanticSegmentation
 
 [[autodoc]] DPTForSemanticSegmentation
-    - forward
+    - forward

docs/source/en/model_doc/esm.mdx

@@ -0,0 +1,146 @@
+<!--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.
+-->
+
+# ESM
+
+## Overview
+This page provides code and pre-trained weights for Transformer protein language models from Meta AI's Fundamental 
+AI Research Team, providing the state-of-the-art ESMFold and ESM-2, and the previously released ESM-1b and ESM-1v.
+Transformer protein language models were introduced in the paper [Biological structure and function emerge from scaling
+unsupervised learning to 250 million protein sequences](https://www.pnas.org/content/118/15/e2016239118) by 
+Alexander Rives, Joshua Meier, Tom Sercu, Siddharth Goyal, Zeming Lin, Jason Liu, Demi Guo, Myle Ott, 
+C. Lawrence Zitnick, Jerry Ma, and Rob Fergus.
+The first version of this paper was [preprinted in 2019](https://www.biorxiv.org/content/10.1101/622803v1?versioned=true).
+
+ESM-2 outperforms all tested single-sequence protein language models across a range of structure prediction tasks,
+and enables atomic resolution structure prediction.
+It was released with the paper [Language models of protein sequences at the scale of evolution enable accurate
+structure prediction](https://doi.org/10.1101/2022.07.20.500902) by Zeming Lin, Halil Akin, Roshan Rao, Brian Hie,
+Zhongkai Zhu, Wenting Lu, Allan dos Santos Costa, Maryam Fazel-Zarandi, Tom Sercu, Sal Candido and Alexander Rives.
+
+Also introduced in this paper was ESMFold. It uses an ESM-2 stem with a head that can predict folded protein
+structures with state-of-the-art accuracy. Unlike [AlphaFold2](https://www.nature.com/articles/s41586-021-03819-2),
+it relies on the token embeddings from the large pre-trained protein language model stem and does not perform a multiple
+sequence alignment (MSA) step at inference time, which means that ESMFold checkpoints are fully "standalone" -
+they do not require a database of known protein sequences and structures with associated external query tools
+to make predictions, and are much faster as a result.
+
+
+The abstract from 
+"Biological structure and function emerge from scaling unsupervised learning to 250 
+million protein sequences" is
+
+
+*In the field of artificial intelligence, a combination of scale in data and model capacity enabled by unsupervised
+learning has led to major advances in representation learning and statistical generation. In the life sciences, the
+anticipated growth of sequencing promises unprecedented data on natural sequence diversity. Protein language modeling
+at the scale of evolution is a logical step toward predictive and generative artificial intelligence for biology. To
+this end, we use unsupervised learning to train a deep contextual language model on 86 billion amino acids across 250
+million protein sequences spanning evolutionary diversity. The resulting model contains information about biological
+properties in its representations. The representations are learned from sequence data alone. The learned representation
+space has a multiscale organization reflecting structure from the level of biochemical properties of amino acids to
+remote homology of proteins. Information about secondary and tertiary structure is encoded in the representations and
+can be identified by linear projections. Representation learning produces features that generalize across a range of
+applications, enabling state-of-the-art supervised prediction of mutational effect and secondary structure and
+improving state-of-the-art features for long-range contact prediction.*
+
+
+The abstract from
+"Language models of protein sequences at the scale of evolution enable accurate structure prediction" is
+
+*Large language models have recently been shown to develop emergent capabilities with scale, going beyond
+simple pattern matching to perform higher level reasoning and generate lifelike images and text. While
+language models trained on protein sequences have been studied at a smaller scale, little is known about
+what they learn about biology as they are scaled up. In this work we train models up to 15 billion parameters,
+the largest language models of proteins to be evaluated to date. We find that as models are scaled they learn
+information enabling the prediction of the three-dimensional structure of a protein at the resolution of
+individual atoms. We present ESMFold for high accuracy end-to-end atomic level structure prediction directly
+from the individual sequence of a protein. ESMFold has similar accuracy to AlphaFold2 and RoseTTAFold for
+sequences with low perplexity that are well understood by the language model. ESMFold inference is an
+order of magnitude faster than AlphaFold2, enabling exploration of the structural space of metagenomic
+proteins in practical timescales.*
+
+
+Tips:
+
+- ESM models are trained with a masked language modeling (MLM) objective.
+
+The original code can be found [here](https://github.com/facebookresearch/esm) and was
+was developed by the Fundamental AI Research team at Meta AI.
+ESM-1b, ESM-1v and ESM-2 were contributed to huggingface by [jasonliu](https://huggingface.co/jasonliu)
+and [Matt](https://huggingface.co/Rocketknight1).
+
+ESMFold was contributed to huggingface by [Matt](https://huggingface.co/Rocketknight1) and
+[Sylvain](https://huggingface.co/sgugger), with a big thank you to Nikita Smetanin, Roshan Rao and Tom Sercu for their
+help throughout the process!
+
+The HuggingFace port of ESMFold uses portions of the [openfold](https://github.com/aqlaboratory/openfold) library.
+The `openfold` library is licensed under the Apache License 2.0.
+
+## EsmConfig
+
+[[autodoc]] EsmConfig
+    - all
+
+## EsmTokenizer
+
+[[autodoc]] EsmTokenizer
+    - build_inputs_with_special_tokens
+    - get_special_tokens_mask
+    - create_token_type_ids_from_sequences
+    - save_vocabulary
+
+
+## EsmModel
+
+[[autodoc]] EsmModel
+    - forward
+
+## EsmForMaskedLM
+
+[[autodoc]] EsmForMaskedLM
+    - forward
+
+## EsmForSequenceClassification
+
+[[autodoc]] EsmForSequenceClassification
+    - forward
+
+## EsmForTokenClassification
+
+[[autodoc]] EsmForTokenClassification
+    - forward
+
+## EsmForProteinFolding
+
+[[autodoc]] EsmForProteinFolding
+    - forward
+
+## TFEsmModel
+
+[[autodoc]] TFEsmModel
+    - call
+
+## TFEsmForMaskedLM
+
+[[autodoc]] TFEsmForMaskedLM
+    - call
+
+## TFEsmForSequenceClassification
+
+[[autodoc]] TFEsmForSequenceClassification
+    - call
+
+## TFEsmForTokenClassification
+
+[[autodoc]] TFEsmForTokenClassification
+    - call

docs/source/en/model_doc/flan-t5.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.
+-->
+
+# FLAN-T5
+
+## Overview
+
+FLAN-T5 was released in the paper [Scaling Instruction-Finetuned Language Models](https://arxiv.org/pdf/2210.11416.pdf) - it is an enhanced version of T5 that has been finetuned in a mixture of tasks.
+
+One can directly use FLAN-T5 weights without finetuning the model:
+
+```python
+>>> from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
+
+>>> model = AutoModelForSeq2SeqLM.from_pretrained("google/flan-t5-small")
+>>> tokenizer = AutoTokenizer.from_pretrained("google/flan-t5-small")
+
+>>> inputs = tokenizer("A step by step recipe to make bolognese pasta:", return_tensors="pt")
+>>> outputs = model.generate(**inputs)
+>>> print(tokenizer.batch_decode(outputs, skip_special_tokens=True))
+['Pour a cup of bolognese into a large bowl and add the pasta']
+```
+
+FLAN-T5 includes the same improvements as T5 version 1.1 (see [here](https://huggingface.co/docs/transformers/model_doc/t5v1.1) for the full details of the model's improvements.)
+
+Google has released the following variants:
+
+- [google/flan-t5-small](https://huggingface.co/google/flan-t5-small)
+
+- [google/flan-t5-base](https://huggingface.co/google/flan-t5-base)
+
+- [google/flan-t5-large](https://huggingface.co/google/flan-t5-base)
+
+- [google/flan-t5-xl](https://huggingface.co/google/flan-t5-xl)
+
+- [google/flan-t5-xxl](https://huggingface.co/google/flan-t5-xxl).
+
+One can refer to [T5's documentation page](t5) for all tips, code examples and notebooks. As well as the FLAN-T5 model card for more details regarding training and evaluation of the model.
+
+The original checkpoints can be found [here](https://github.com/google-research/t5x/blob/main/docs/models.md#mixture-of-experts-moe-checkpoints).

docs/source/en/model_doc/flava.mdx

@@ -16,17 +16,17 @@ specific language governing permissions and limitations under the License.
 
 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 
+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 
+*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.*
 
 
@@ -61,6 +61,11 @@ This model was contributed by [aps](https://huggingface.co/aps). The original co
 
 [[autodoc]] FlavaFeatureExtractor
 
+## FlavaImageProcessor
+
+[[autodoc]] FlavaImageProcessor
+    - preprocess
+
 ## FlavaForPreTraining
 
 [[autodoc]] FlavaForPreTraining

docs/source/en/model_doc/glpn.mdx

@@ -32,10 +32,10 @@ The abstract from the paper is the following:
 Tips:
 
 - A notebook illustrating inference with [`GLPNForDepthEstimation`] can be found [here](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/GLPN/GLPN_inference_(depth_estimation).ipynb).
-- One can use [`GLPNFeatureExtractor`] to prepare images for the model.
+- One can use [`GLPNImageProcessor`] to prepare images for the model.
 
 <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/glpn_architecture.jpg"
-alt="drawing" width="600"/> 
+alt="drawing" width="600"/>
 
 <small> Summary of the approach. Taken from the <a href="https://arxiv.org/abs/2201.07436" target="_blank">original paper</a>. </small>
 
@@ -50,6 +50,11 @@ This model was contributed by [nielsr](https://huggingface.co/nielsr). The origi
 [[autodoc]] GLPNFeatureExtractor
     - __call__
 
+## GLPNImageProcessor
+
+[[autodoc]] GLPNImageProcessor
+    - preprocess
+
 ## GLPNModel
 
 [[autodoc]] GLPNModel
@@ -58,4 +63,4 @@ This model was contributed by [nielsr](https://huggingface.co/nielsr). The origi
 ## GLPNForDepthEstimation
 
 [[autodoc]] GLPNForDepthEstimation
-    - forward
+    - forward

docs/source/en/model_doc/gpt2.mdx

@@ -47,6 +47,24 @@ different sizes: small, medium, large, xl and a distilled version of the small c
 
 This model was contributed by [thomwolf](https://huggingface.co/thomwolf). The original code can be found [here](https://openai.com/blog/better-language-models/).
 
+## Resources
+
+A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with GPT2. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
+
+<PipelineTag pipeline="text-generation"/>
+
+- A blog on how to [Finetune a non-English GPT-2 Model with Hugging Face](https://www.philschmid.de/fine-tune-a-non-english-gpt-2-model-with-huggingface).
+- A blog on [How to generate text: using different decoding methods for language generation with Transformers](https://huggingface.co/blog/how-to-generate) with GPT-2.
+- A blog on [Training CodeParrot 🦜 from Scratch](https://huggingface.co/blog/codeparrot), a large GPT-2 model.
+- A blog on [Faster Text Generation with TensorFlow and XLA](https://huggingface.co/blog/tf-xla-generate) with GPT-2.
+- A blog on [How to train a Language Model with Megatron-LM](https://huggingface.co/blog/megatron-training) with a GPT-2 model.
+- A notebook on how to [finetune GPT2 to generate lyrics in the style of your favorite artist](https://colab.research.google.com/github/AlekseyKorshuk/huggingartists/blob/master/huggingartists-demo.ipynb). 🌎
+- A notebook on how to [finetune GPT2 to generate tweets in the style of your favorite Twitter user](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb). 🌎
+- [Causal language modeling](https://huggingface.co/course/en/chapter7/6?fw=pt#training-a-causal-language-model-from-scratch) chapter of the 🤗 Hugging Face Course.
+- [`GPT2LMHeadModel`] is supported by this [causal language modeling example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling#gpt-2gpt-and-causal-language-modeling), [text generation example script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/text-generation), and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling.ipynb).
+- [`TFGPT2LMHeadModel`] is supported by this [causal language modeling example script](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/language-modeling#run_clmpy) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling-tf.ipynb).
+- [`FlaxGPT2LMHeadModel`] is supported by this [causal language modeling example script](https://github.com/huggingface/transformers/tree/main/examples/flax/language-modeling#causal-language-modeling) and [notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/causal_language_modeling_flax.ipynb).
+
 
 ## GPT2Config
 
@@ -120,6 +138,10 @@ This model was contributed by [thomwolf](https://huggingface.co/thomwolf). The o
 
 [[autodoc]] modeling_tf_outputs.TFSequenceClassifierOutputWithPast
 
+## TFGPT2Tokenizer
+
+[[autodoc]] TFGPT2Tokenizer
+
 ## FlaxGPT2Model
 
 [[autodoc]] FlaxGPT2Model

docs/source/en/model_doc/gpt_neox_japanese.mdx

@@ -0,0 +1,66 @@
+<!--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-Japanese
+
+## Overview
+
+We introduce GPT-NeoX-Japanese, which is an autoregressive language model for Japanese, trained on top of [https://github.com/EleutherAI/gpt-neox](https://github.com/EleutherAI/gpt-neox).
+Japanese is a unique language with its large vocabulary and a combination of hiragana, katakana, and kanji writing scripts.
+To address this distinct structure of the Japanese language, we use a [special sub-word tokenizer](https://github.com/tanreinama/Japanese-BPEEncoder_V2). We are very grateful to *tanreinama* for open-sourcing this incredibly helpful tokenizer.
+Following the recommendations from Google's research on [PaLM](https://ai.googleblog.com/2022/04/pathways-language-model-palm-scaling-to.html), we have removed bias parameters from transformer blocks, achieving better model performance. Please refer [this article](https://medium.com/ml-abeja/training-a-better-gpt-2-93b157662ae4) in detail.
+
+Development of the model was led by [Shinya Otani](https://github.com/SO0529), [Takayoshi Makabe](https://github.com/spider-man-tm), [Anuj Arora](https://github.com/Anuj040), and [Kyo Hattori](https://github.com/go5paopao) from [ABEJA, Inc.](https://www.abejainc.com/). For more information on this model-building activity, please refer [here (ja)](https://tech-blog.abeja.asia/entry/abeja-gpt-project-202207).
+
+### Generation
+
+The `generate()` method can be used to generate text using GPT NeoX Japanese model.
+
+```python
+>>> from transformers import GPTNeoXJapaneseForCausalLM, GPTNeoXJapaneseTokenizer
+
+>>> model = GPTNeoXJapaneseForCausalLM.from_pretrained("abeja/gpt-neox-japanese-2.7b")
+>>> tokenizer = GPTNeoXJapaneseTokenizer.from_pretrained("abeja/gpt-neox-japanese-2.7b")
+
+>>> prompt = "人とAIが協調するためには、"
+
+>>> 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, skip_special_tokens=True)[0]
+
+>>> print(gen_text)
+人とAIが協調するためには、AIと人が共存し、AIを正しく理解する必要があります。
+```
+
+## GPTNeoXJapaneseConfig
+
+[[autodoc]] GPTNeoXJapaneseConfig
+
+## GPTNeoXJapaneseTokenizer
+
+[[autodoc]] GPTNeoXJapaneseTokenizer
+
+## GPTNeoXJapaneseModel
+
+[[autodoc]] GPTNeoXJapaneseModel
+    - forward
+
+## GPTNeoXJapaneseForCausalLM
+
+[[autodoc]] GPTNeoXJapaneseForCausalLM
+    - forward

docs/source/en/model_doc/gptj.mdx

@@ -47,13 +47,13 @@ Tips:
   that could be found [here](https://github.com/kingoflolz/mesh-transformer-jax/blob/master/howto_finetune.md)
 
 - Although the embedding matrix has a size of 50400, only 50257 entries are used by the GPT-2 tokenizer. These extra
-  tokens are added for the sake of efficiency on TPUs. To avoid the mis-match between embedding matrix size and vocab
+  tokens are added for the sake of efficiency on TPUs. To avoid the mismatch between embedding matrix size and vocab
   size, the tokenizer for [GPT-J](https://huggingface.co/EleutherAI/gpt-j-6B) contains 143 extra tokens
   `<|extratoken_1|>... <|extratoken_143|>`, so the `vocab_size` of tokenizer also becomes 50400.
 
 ### Generation
 
-The [`~generation_utils.GenerationMixin.generate`] method can be used to generate text using GPT-J
+The [`~generation.GenerationMixin.generate`] method can be used to generate text using GPT-J
 model.
 
 ```python

docs/source/en/model_doc/groupvit.mdx

@@ -26,7 +26,7 @@ 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).
+This model was contributed by [xvjiarui](https://huggingface.co/xvjiarui). The TensorFlow version was contributed by [ariG23498](https://huggingface.co/ariG23498) with the help of [Yih-Dar SHIEH](https://huggingface.co/ydshieh), [Amy Roberts](https://huggingface.co/amyeroberts), and [Joao Gante](https://huggingface.co/joaogante).
 The original code can be found [here](https://github.com/NVlabs/GroupViT).
 
 
@@ -59,3 +59,20 @@ The original code can be found [here](https://github.com/NVlabs/GroupViT).
 
 [[autodoc]] GroupViTVisionModel
     - forward
+
+## TFGroupViTModel
+
+[[autodoc]] TFGroupViTModel
+    - call
+    - get_text_features
+    - get_image_features
+
+## TFGroupViTTextModel
+
+[[autodoc]] TFGroupViTTextModel
+    - call
+
+## TFGroupViTVisionModel
+
+[[autodoc]] TFGroupViTVisionModel
+    - call

docs/source/en/model_doc/imagegpt.mdx

@@ -29,7 +29,7 @@ competitive with self-supervised benchmarks on ImageNet when substituting pixels
 top-1 accuracy on a linear probe of our features.*
 
 <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/imagegpt_architecture.png"
-alt="drawing" width="600"/> 
+alt="drawing" width="600"/>
 
 <small> Summary of the approach. Taken from the [original paper](https://cdn.openai.com/papers/Generative_Pretraining_from_Pixels_V2.pdf). </small>
 
@@ -49,7 +49,7 @@ Tips:
   applied k-means clustering to the (R,G,B) pixel values with k=512. This way, we only have a 32*32 = 1024-long
   sequence, but now of integers in the range 0..511. So we are shrinking the sequence length at the cost of a bigger
   embedding matrix. In other words, the vocabulary size of ImageGPT is 512, + 1 for a special "start of sentence" (SOS)
-  token, used at the beginning of every sequence. One can use [`ImageGPTFeatureExtractor`] to prepare
+  token, used at the beginning of every sequence. One can use [`ImageGPTImageProcessor`] to prepare
   images for the model.
 - Despite being pre-trained entirely unsupervised (i.e. without the use of any labels), ImageGPT produces fairly
   performant image features useful for downstream tasks, such as image classification. The authors showed that the
@@ -81,6 +81,11 @@ Tips:
 
     - __call__
 
+## ImageGPTImageProcessor
+
+[[autodoc]] ImageGPTImageProcessor
+    - preprocess
+
 ## ImageGPTModel
 
 [[autodoc]] ImageGPTModel
@@ -97,4 +102,4 @@ Tips:
 
 [[autodoc]] ImageGPTForImageClassification
 
-    - forward
+    - forward

docs/source/en/model_doc/jukebox.mdx

@@ -0,0 +1,79 @@
+<!--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.
+-->
+# Jukebox
+
+## Overview
+
+The Jukebox model was proposed in [Jukebox: A generative model for music](https://arxiv.org/pdf/2005.00341.pdf)
+by Prafulla Dhariwal, Heewoo Jun, Christine Payne, Jong Wook Kim, Alec Radford,
+Ilya Sutskever. It introduces a generative music model which can produce minute long samples that can be conditionned on
+an artist, genres and lyrics.
+
+The abstract from the paper is the following:
+
+*We introduce Jukebox, a model that generates music with singing in the raw audio domain. We tackle the long context of raw audio using a multiscale VQ-VAE to compress it to discrete codes, and modeling those using autoregressive Transformers. We show that the combined model at scale can generate high-fidelity and diverse songs with coherence up to multiple minutes. We can condition on artist and genre to steer the musical and vocal style, and on unaligned lyrics to make the singing more controllable. We are releasing thousands of non cherry-picked samples, along with model weights and code.*
+
+As shown on the following figure, Jukebox is made of 3 `priors` which are decoder only models. They follow the architecture described in [Generating Long Sequences with Sparse Transformers](https://arxiv.org/abs/1904.10509), modified to support longer context length.
+First, a autoencoder is used to encode the text lyrics. Next, the first (also called `top_prior`) prior attends to the last hidden states extracted from the lyrics encoder. The priors are linked to the previous priors respectively via an `AudioConditionner` module. The`AudioConditioner` upsamples the outputs of the previous prior to raw tokens at a certain audio frame per second resolution. 
+The metadata such as *artist, genre and timing* are passed to each prior, in the form of a start token and positionnal embedding for the timing data.  The hidden states are mapped to the closest codebook vector from the VQVAE in order to convert them to raw audio.
+
+![JukeboxModel](https://gist.githubusercontent.com/ArthurZucker/92c1acaae62ebf1b6a951710bdd8b6af/raw/c9c517bf4eff61393f6c7dec9366ef02bdd059a3/jukebox.svg)
+
+Tips:
+- This model only supports inference. This is for a few reasons, mostly because it requires a crazy amount of memory to train. Feel free to open a PR and add what's missing to have a full integration with the hugging face traineer!
+- This model is very slow, and takes 8h to generate a minute long audio using the 5b top prior on a V100 GPU. In order automaticallay handle the device on which the model should execute, use `accelerate`.
+- Contrary to the paper, the order of the priors goes from `0` to `1` as it felt more intuitive : we sample starting from `0`.
+- Primed sampling (conditionning the sampling on raw audio) requires more memory than ancestral sampling and should be used with `fp16` set to `True`.
+
+This model was contributed by [Arthur Zucker](https://huggingface.co/ArthurZ).
+The original code can be found [here](https://github.com/openai/jukebox).
+
+## JukeboxConfig
+
+[[autodoc]] JukeboxConfig
+
+## JukeboxPriorConfig
+
+[[autodoc]] JukeboxPriorConfig
+
+## JukeboxVQVAEConfig
+
+[[autodoc]] JukeboxVQVAEConfig
+
+## JukeboxTokenizer
+
+[[autodoc]] JukeboxTokenizer
+    - save_vocabulary
+
+## JukeboxModel
+
+[[autodoc]] JukeboxModel
+    - ancestral_sample
+    - primed_sample
+    - continue_sample
+    - upsample
+    - _sample
+
+
+## JukeboxPrior
+
+[[autodoc]] JukeboxPrior
+    - sample
+    - forward
+
+
+## JukeboxVQVAE
+
+[[autodoc]] JukeboxVQVAE
+    - forward
+    - encode
+    - decode

docs/source/en/model_doc/layoutlmv2.mdx

@@ -45,7 +45,7 @@ RVL-CDIP (0.9443 -> 0.9564), and DocVQA (0.7295 -> 0.8672). The pre-trained Layo
 this https URL.*
 
 LayoutLMv2 depends on `detectron2`, `torchvision` and `tesseract`. Run the
-following to install them: 
+following to install them:
 ```
 python -m pip install 'git+https://github.com/facebookresearch/detectron2.git'
 python -m pip install torchvision tesseract
@@ -275,6 +275,11 @@ print(encoding.keys())
 [[autodoc]] LayoutLMv2FeatureExtractor
     - __call__
 
+## LayoutLMv2ImageProcessor
+
+[[autodoc]] LayoutLMv2ImageProcessor
+    - preprocess
+
 ## LayoutLMv2Tokenizer
 
 [[autodoc]] LayoutLMv2Tokenizer

docs/source/en/model_doc/layoutlmv3.mdx

@@ -39,6 +39,30 @@ alt="drawing" width="600"/>
 
 This model was contributed by [nielsr](https://huggingface.co/nielsr). The TensorFlow version of this model was added by [chriskoo](https://huggingface.co/chriskoo), [tokec](https://huggingface.co/tokec), and [lre](https://huggingface.co/lre). The original code can be found [here](https://github.com/microsoft/unilm/tree/master/layoutlmv3).
 
+## Resources
+
+A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with LayoutLMv3. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
+
+<Tip>
+
+LayoutLMv3 is nearly identical to LayoutLMv2, so we've also included LayoutLMv2 resources you can adapt for LayoutLMv3 tasks. For these notebooks, take care to use [`LayoutLMv2Processor`] instead when preparing data for the model!
+
+</Tip>
+
+<PipelineTag pipeline="text-classification"/>
+
+- [`LayoutLMv2ForSequenceClassification`] is supported by this [notebook](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/LayoutLMv2/RVL-CDIP/Fine_tuning_LayoutLMv2ForSequenceClassification_on_RVL_CDIP.ipynb).
+
+<PipelineTag pipeline="token-classification"/>
+
+- [`LayoutLMv3ForTokenClassification`] is supported by this [example script](https://github.com/huggingface/transformers/tree/main/examples/research_projects/layoutlmv3) and [notebook](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/LayoutLMv3/Fine_tune_LayoutLMv3_on_FUNSD_(HuggingFace_Trainer).ipynb).
+- A [notebook](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/LayoutLMv2/FUNSD/Inference_with_LayoutLMv2ForTokenClassification.ipynb) for how to perform inference with [`LayoutLMv2ForTokenClassification`] and a [notebook](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/LayoutLMv2/FUNSD/True_inference_with_LayoutLMv2ForTokenClassification_%2B_Gradio_demo.ipynb) for how to perform inference when no labels are available with [`LayoutLMv2ForTokenClassification`].
+- A [notebook](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/LayoutLMv2/FUNSD/Fine_tuning_LayoutLMv2ForTokenClassification_on_FUNSD_using_HuggingFace_Trainer.ipynb) for how to finetune [`LayoutLMv2ForTokenClassification`] with the 🤗 Trainer.
+
+<PipelineTag pipeline="question-answering"/>
+
+- [`LayoutLMv2ForQuestionAnswering`] is supported by this [notebook](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/LayoutLMv2/DocVQA/Fine_tuning_LayoutLMv2ForQuestionAnswering_on_DocVQA.ipynb).
+
 
 ## LayoutLMv3Config
 
@@ -49,6 +73,11 @@ This model was contributed by [nielsr](https://huggingface.co/nielsr). The Tenso
 [[autodoc]] LayoutLMv3FeatureExtractor
     - __call__
 
+## LayoutLMv3ImageProcessor
+
+[[autodoc]] LayoutLMv3ImageProcessor
+    - preprocess
+
 ## LayoutLMv3Tokenizer
 
 [[autodoc]] LayoutLMv3Tokenizer

docs/source/en/model_doc/led.mdx

@@ -50,6 +50,8 @@ Tips:
   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).
+- LED is a model with absolute position embeddings so it's usually advised to pad the inputs on the right rather than
+  the left.
 
 This model was contributed by [patrickvonplaten](https://huggingface.co/patrickvonplaten).
 

docs/source/en/model_doc/levit.mdx

@@ -19,18 +19,18 @@ The LeViT model was proposed in [LeViT: Introducing Convolutions to Vision Trans
 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 
+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 
+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"/> 
+alt="drawing" width="600"/>
 
 <small> LeViT Architecture. Taken from the <a href="https://arxiv.org/abs/2104.01136">original paper</a>.</small>
 
@@ -38,26 +38,26 @@ 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 
+  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) 
+- 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`]. 
+- 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
+  *facebook/levit-384*. Note that one should use [`LevitImageProcessor`] 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`]).
+- 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 [`LevitImageProcessor`] 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).
 
@@ -71,6 +71,12 @@ This model was contributed by [anugunj](https://huggingface.co/anugunj). The ori
 [[autodoc]] LevitFeatureExtractor
     - __call__
 
+## LevitImageProcessor
+
+  [[autodoc]] LevitImageProcessor
+    - preprocess
+
+
 ## LevitModel
 
 [[autodoc]] LevitModel