GitLab

* Add local and TensorFlow ONNX export examples to docs

* Use PyTorch - TensorFlow split

Co-authored-by: Niels Rogge <nielsrogge@Nielss-MBP.localdomain>

* Compute loss independent from decoder (as 14139)

* fix expected seq_len + style

* Apply the same change to TFVisionEncoderDecoderModel

* fix style

* Add case with labels in equivalence test

* uncomment

* Add case with labels in equivalence test

* add decoder_token_labels

* use hf_compute_loss

* Apply suggestions from code review
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>

* Add copied from
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>

* Add TensorFlow support for ONNX export

* Change documentation to mention conversion with Tensorflow

* Refactor export into export_pytorch and export_tensorflow

* Check model's type instead of framework installation to choose between TF and Pytorch
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Alberto Bégué <alberto.begue@della.ai>
Co-authored-by: lewtun <lewis.c.tunstall@gmail.com>

* Expand tutorial for custom models

* Style

* Apply suggestions from code review
Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>
Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>

Co-authored-by: Niels Rogge <nielsrogge@Nielss-MBP.localdomain>

* [trainer docs] document how to select specific gpus

* expand

* add urls

* add accelerate launcher

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

* Expose hub test problem

* Fix tests

* added classes to get started with constrained beam search

* in progress, think i can directly force tokens now but not yet with the round robin

* think now i have total control, now need to code the bank selection

* technically works as desired, need to optimize and fix design choices leading to undersirable outputs

* complete PR #1 without disjunctive decoding

* removed incorrect tests

* Delete k.txt

* Delete test.py

* Delete test.sh

* revert changes to test scripts

* genutils

* full implementation with testing, no disjunctive yet

* shifted docs

* passing all tests realistically ran locally

* removing accidentally included print statements

* fixed source of error in initial PR test

* fixing the get_device() vs device trap

* fixed documentation docstrings about constrained_beam_search

* fixed tests having failing for Speech2TextModel's floating point inputs

* fix cuda long tensor

* added examples and testing for them and founx & fixed a bug in beam_search and constrained_beam_search

* deleted accidentally added test halting code with assert False

* code reformat

* Update tests/test_generation_utils.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Update tests/test_generation_utils.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Update tests/test_generation_utils.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Update tests/test_generation_utils.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Update tests/test_generation_utils.py

* fixing based on comments on PR

* took out the testing code that should but work fails without the beam search moditification ; style changes

* fixing comments issues

* docstrings for ConstraintListState

* typo in PhrsalConstraint docstring

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

* typical decoding

* changing arg name

* add test config params

* forgotten arg rename

* fix edge case where scores are same

* test for typical logits warper

* code quality fixes

* Add Wav2Vec2 Adapter Weights to Flax

* Suggested changes

* Make sure custom configs work with Transformers

* Apply code review suggestions

* Upgrade black to version ~=22.0

* Check copies

* Fix code

* add model scaling section

* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* integrate reviewer feedback

* initialize GPU properly

* add note about BnB optimizer

* move doc from `scaling.mdx` to `performance.mdx`

* integrate reviewer feedback

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

* PoC for a ProcessorMixin class

* Documentation

* Apply suggestions from code review
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Roll out to other processors

* Add base feature extractor class in init

* Use args and kwargs
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* change logger.warn to logger.warning

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

* fix test_model_outputs_equivalence

* fix tuple outputs for blenderbot

* feat(flax): allow encoder_outputs in generate

* doc(flax): encoder_outputs in generate

* fix: style

* fix: style

* Add wrapper classes

* convert inner layers to tf

* Add TF Encoder and Decoder layers

* TFSpeech2Text models

* Loadable model

* TF model with same outputs as PT model

* test skeleton

* correct tests and run the fixup

* correct attention expansion

* TFSpeech2Text pask_key_values with TF format

* use_cache = False for PT models if labels is passed

* Fix for BigBirdPegasusForConditionalGeneration

* add warning if users specify use_cache=True

* Use logger.warning instead of warnings.warn
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

* electra is added to onnx supported model

* add google/electra-base-generator for test onnx module
Co-authored-by: Lewis Tunstall <lewis.c.tunstall@gmail.com>

* Change the way tracing happens, enabling dynamic axes out of the box

* Update the tests and modeling xlnet

* Add the non recoding of leaf modules to avoid recording more values for the methods to record than what will be seen at tracing time (which would otherwise desynchronize the recorded values and the values that need to be given to the proxies during tracing, causing errors).

* Comments and making tracing work for gpt-j and xlnet

* Refactore things related to num_choices (and batch_size, sequence_length)

* Update fx to work on PyTorch 1.10

* Postpone autowrap_function feature usage for later

* Add copyrights

* Remove unnecessary file

* Fix issue with add_new_model_like

* Apply suggestions

* 📝 add config section

* 📝 finish first draft

* 📝 add feature extractor and processor

* 🖍 apply feedback from review

* 📝 minor edits

* last review

* fix outputs

* fix for CTC

* fix doc

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

* add cross attn to outputs

* add cross attn to outputs for TFLED

* add undo padding

* remove unused import

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

* Wav2Vec2 models must either throw or deal with add_apater
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Add pre-add_adapter backwards compatibility

* Add pre-add_adapter backwards compatibility

* Fix issue in tests/test_modeling_wav2vec2.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>