* gradient checkpointing for GPT-NeoX

* initialize gradient checkpointing flag

* must set flag before init

* Add Example docstring to DebertaConfig

* Add configuration_deberta to documentation_tests

* Add microsoft/deberta-base to example docstring

* Fix example docstring mistake

* donut -> donut-swin

* remove ("donut-swin", "DonutProcessor")

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

* Fix ONNX tests for ONNX Runtime v1.13.1

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Add postprocessing methods

* Update docs

* Add fix

* Add test

* Add test for deformable detr postprocessing

* Add post processing methods for segmentation

* Update code examples

* Add post_process to make the pipeline work

* Apply updates

Co-authored-by: Niels Rogge <nielsrogge@Nielss-MacBook-Pro.local>

* add wav2vec2 resources

* apply review

Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

* Update Code of Conduct to Contributor Covenant v2.1

* Update CODE_OF_CONDUCT.md

* add Onnx Config for ImageGPT

* add generate_dummy_inputs for onnx config

* add TYPE_CHECKING clause

* Update doc for generate_dummy_inputs

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Do not generate trial_name when trail is None

* Use (trial or self._trial) to generate trial_name

* Follow comments

* Support segformer fx

* Add fx_compatible attribute to test_modeling_segformer.py

* Update glpn model (fx support)

glpn model was copied from segformer.

* Update utils/fx.py | add semantic-segmentation

for SegformerForSemanticSegmentation model

* Fix minor import order(isort)

* Add random input generation for segformer fx

Co-authored-by: noelbird <lduldu00228@gmail.com>

* create dummy models

* quality

* update

* update

* Make Wav2Vec2Conformer work

* style

* deal with models with text_config and vision_config

* apply suggestions

* Composite models

* style

* style

* fix shape issue

* fix shape issue

* For VisionTextDualEncoderModel

* show_progress=False when converting tokenizers

* Fix for OwlViT

* Fix for VisualBert

* Update

* final

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

* forward contrib credits from suggestion

* add `accelerate` support for BART-like models

Co-authored-by: sgugger <sgugger@users.noreply.github.com>

* Let inputs of fast tokenizers be tuples as well as lists

* Update src/transformers/tokenization_utils_fast.py

Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>

* Style

Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>

* Wip

* Add safetensors support for TensorFlow

* First tests

* Add final test for now

* Retrigger CI like this

* Update src/transformers/modeling_tf_utils.py

Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>

Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>

* add resources for gpt2

* add pipeline icons and community resources

* add bloom resources

* add pipeline icon

* add resources for t5

* add pipeline icons and community resources

* add roberta resources

* fix typo

* add `accelerate` support for M2M100

* fix device set nit

* added typing to call in TFPegasusModel and TFPegasusForConditionalGeneration

* fixed type for TFPegasusForConditionalGeneration call

* Fix tests when running on GPU

* Fix tests that require mp.set_start_method

* Add pegasus_x

* ViTMSN

* ESM

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

* Convert None logits processor/stopping criteria to empty list.

* Initialize stopping_criteria, logits_processor in generate.

* Default stopping_criteria, logits_processor to None.

Co-authored-by: Chandler May <chandler.j.may@gmail.com>

* For CLIP

* Others

* update

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

* add first generation tutorial

* [Flax] Add subfolder functionality

* [Flax] Add subfolder functionality

* up

* finish

* delete file and re-add test

* add `flan-t5` documentation page

* Update README.md

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* add more content

* revert `_toctree` modif

* revert `toctree` modif - 2

* Update README.md

* Revert "Update README.md"

This reverts commit 56607144299c5fdf7b18abdb776efd0d03287727.

* Update README_es.md

* Update README_zh-hans.md

* Update README_zh-hant.md

* Update README_ko.md

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Fix test_save_load_fast_init_to_base

* Fix test_save_load_fast_init_to_base

* update

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

* Change the import of kenlm from github to pypi

* Change the import of kenlm from github to pypi in circleci config

* Fix code quality issues

* Fix isort issue, add kenlm in extras for audio

* Add kenlm to deps

* Add kenlm to deps

* Commit 'make fixup' changes

* Remove version from kenlm deps

* commit make fixup changes

* Remove manual installation of kenlm

* Remove manual installation of kenlm

* Remove manual installation of kenlm

* Add missing information on token_type_ids for roberta model

* Fix code format issues

* Fix code format issues

* Add more explicit document for token_type_ids for roberta

* Fix flake8 issues

* Fix flake8 issues

* Fix flake8 issues

* Fix flake8 issues

* Fix flake8 issues

* no conv bn folding in ipex

* no flag in training

* comment

Co-authored-by: Sander Land <sander@chatdesk.com>

swap "right" and "left" so description is correct.

* Update documentation_tests.txt

* Update configuration_mbart.py

* Update configuration_mctct.py

* Update configuration_layoutlm.py

* Update configuration_layoutlmv2.py

* Update configuration_layoutlmv3.py

* Update documentation_tests.txt