* Fixing image segmentation for inference mode.

* Update src/transformers/pipelines/base.py

Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Torch 1.10

* torch scatter for 1.10

* style

* Skip tests
ok

* Fix docs

* Apply suggestions from review + fix bug

* First draft

* Make tuple output more readable

* Replace assertions by value errors

* Make it possible to predict_with_generate for vision and speech models

* Adapt Seq2SeqTrainer to work with VisionEncoderDecoder/SpeechEncoderDecoder

* Add deprecation warning

* Add copied from statements to vision and speech encoder decoders

* Fix failing test

* Apply @patrickvonplaten's suggestion

* Use reshape instead of view for consistency

* Fix SEW-D

* Update tests

* isort

* First draft

* Make style & quality

* Improve conversion script

* Add print statement to see actual slice

* Make absolute tolerance smaller

* Fix image classification models

* Add post_process_semantic method

* Disable padding

* Improve conversion script

* Rename to ForSemanticSegmentation, add integration test, remove post_process methods

* Improve docs

* Fix code quality

* Fix feature extractor tests

* Fix tests for image classification model

* Delete file

* Add is_torch_available to feature extractor

* Improve documentation of feature extractor methods

* Apply suggestions from @sgugger's code review

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

* Apply some more suggestions of code review

* Rebase with master

* Fix rebase issues

* Make sure model only outputs hidden states when the user wants to

* Apply suggestions from code review

* Add pad method

* Support padding of 2d images

* Add print statement

* Add print statement

* Move padding method to SegformerFeatureExtractor

* Fix issue

* Add casting of segmentation maps

* Add test for padding

* Add small note about padding

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

* respect dtype in _get_resized_lm_head

* Update src/transformers/modeling_utils.py

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

* consistency

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

* up

* up

* fix

* up

* Update examples/pytorch/test_xla_examples.py

* correct labels

* up

* up

* up

* up

* up

* up

* Add conversion

* Rename

* Add an integration test and remove layer_norm

* Remove layer_norm from the converter

* wording

* Fix imports

* Replace assert of data_collator.py by ValueError

* Replace assert of data_collator.py by ValueError

* up

* up

* finish

* up

* final changes

* Add SEW CTC models

* Update paths

* Update paths

* switch to inference_mode from no_gard
faster inference

* added switch to support older version of pytorch

Updated masked-language modeling examples in pytorch
with convention defined by #12789

* specify the text column name in the error message

* pluralize the word fields

* Add symbolic function for XSoftmax op for exporting to ONNX.

* Fix format issues.

* Fix a CI issue relative to copies.

* unispeech

* add copy from

* remove hubert copy from

* finish for today

* add unispeech-sat

* adapt more

* up

* up

* up

* up

* add modeling

* add tests

* up

* up

* finish

* up

* Apply suggestions from code review

* up

* up

* Apply suggestions from code review

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

* up

* up

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

* [megatron_gpt2] dynamic gelu, add tokenizer, save config

* cleanup

* Update src/transformers/models/megatron_gpt2/convert_megatron_gpt2_checkpoint.py

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

* apply suggestions

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

* Include KerasTensor in allowed types

- This allows propagating symbolic tensors through TFBert models and layers' call(),
  which allows converting the subclass models to functional models.

* Style pass

Co-authored-by: Sergio Valcarcel Macua <sergiov@graphcore.ai>
Co-authored-by: matt <rocketknight1@gmail.com>

[Speech Recognition] - Distributed training: Make sure vocab file removal and creation don't interfer  (#14161)

* up

* better

* Add vision_encoder_decoder

* Update _ignore_modules in get_model_modules()

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

Add Camembert to models exportable with ONNX

Co-authored-by: Thomas.Chaigneau <thomas.chaigneau@arkea.com>
Co-authored-by: Michael Benayoun <mickbenayoun@gmail.com>

* up

* up

* up

* up

* up

* up

* up

* add clip

* fix clip PyTorch

* fix clip PyTorch

* up

* up

* up

* up

* up

* up

* up