GitLab

* up

* up

* up

* finish

* Working encoder

* SEW-D and tests

* Further conv fixes

* Automodels and conv inits

* Update integration tests, add docs

* Docs cleanup, resolve todos

* Conf fix

* Fix docs

* Fix tests, apply suggestions

* Update src/transformers/models/sew/modeling_sew.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Model conversion and updated no-mask tests

* Remove copy of feature_proj

* Style

* Update src/transformers/models/auto/feature_extraction_auto.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Update src/transformers/models/auto/feature_extraction_auto.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

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

* Scatter dummies + skip pipeline tests

* Add torch scatter to build docs

* Add strong test for configuration attributes

* Add fake modif to trigger all tests

* Add a better fake modif

* Ignore is_encoder_decoder

* Fix faulty configs

* Remove fake modif

* First draft

* Update self-attention of RoBERTa as proposition

* Improve conversion script

* Add TrOCR decoder-only model

* More improvements

* Make forward pass with pretrained weights work

* More improvements

* Some more improvements

* More improvements

* Make conversion work

* Clean up print statements

* Add documentation, processor

* Add test files

* Small improvements

* Some more improvements

* Make fix-copies, improve docs

* Make all vision encoder decoder model tests pass

* Make conversion script support other models

* Update URL for OCR image

* Update conversion script

* Fix style & quality

* Add support for the large-printed model

* Fix some issues

* Add print statement for debugging

* Add print statements for debugging

* Make possible fix for sinusoidal embedding

* Further debugging

* Potential fix v2

* Add more print statements for debugging

* Add more print statements for debugging

* Deubg more

* Comment out print statements

* Make conversion of large printed model possible, address review comments

* Make it possible to convert the stage1 checkpoints

* Clean up code, apply suggestions from code review

* Apply suggestions from code review, use Microsoft models in tests

* Rename encoder_hidden_size to cross_attention_hidden_size

* Improve docs

* Add cross attentions to TFGPT2Model

* Add TFEncoderDecoderModel

* Add TFBaseModelOutputWithPoolingAndCrossAttentions

* Add cross attentions to TFBertModel

* Fix past or past_key_values argument issue

* Fix generation

* Fix save and load

* Add some checks and comments

* Clean the code that deals with past keys/values

* Add kwargs to processing_inputs

* Add serving_output to TFEncoderDecoderModel

* Some cleaning + fix use_cache value issue

* Fix tests + add bert2bert/bert2gpt2 tests

* Fix more tests

* Ignore crossattention.bias when loading GPT2 weights into TFGPT2

* Fix return_dict_in_generate in tf generation

* Fix is_token_logit_eos_token bug in tf generation

* Finalize the tests after fixing some bugs

* Fix another is_token_logit_eos_token bug in tf generation

* Add/Update docs

* Add TFBertEncoderDecoderModelTest

* Clean test script

* Add TFEncoderDecoderModel to the library

* Add cross attentions to TFRobertaModel

* Add TFRobertaEncoderDecoderModelTest

* make style

* Change the way of position_ids computation

* bug fix

* Fix copies in tf_albert

* Remove some copied from and apply some fix-copies

* Remove some copied

* Add cross attentions to some other TF models

* Remove encoder_hidden_states from TFLayoutLMModel.call for now

* Make style

* Fix TFRemBertForCausalLM

* Revert the change to longformer + Remove copies

* Revert the change to albert and convbert + Remove copies

* make quality

* make style

* Add TFRembertEncoderDecoderModelTest

* make quality and fix-copies

* test TFRobertaForCausalLM

* Fixes for failed tests

* Fixes for failed tests

* fix more tests

* Fixes for failed tests

* Fix Auto mapping order

* Fix TFRemBertEncoder return value

* fix tf_rembert

* Check copies are OK

* Fix missing TFBaseModelOutputWithPastAndCrossAttentions is not defined

* Add TFEncoderDecoderModelSaveLoadTests

* fix tf weight loading

* check the change of use_cache

* Revert the change

* Add missing test_for_causal_lm for TFRobertaModelTest

* Try cleaning past

* fix _reorder_cache

* Revert some files to original versions

* Keep as many copies as possible

* Apply suggested changes - Use raise ValueError instead of assert

* Move import to top

* Fix wrong require_torch

* Replace more assert by raise ValueError

* Add test_pt_tf_model_equivalence (the test won't pass for now)

* add test for loading/saving

* finish

* finish

* Remove test_pt_tf_model_equivalence

* Update tf modeling template

* Remove pooling, added in the prev. commit, from MainLayer

* Update tf modeling test template

* Move inputs["use_cache"] = False to modeling_tf_utils.py

* Fix torch.Tensor in the comment

* fix use_cache

* Fix missing use_cache in ElectraConfig

* Add a note to from_pretrained

* Fix style

* Change test_encoder_decoder_save_load_from_encoder_decoder_from_pt

* Fix TFMLP (in TFGPT2) activation issue

* Fix None past_key_values value in serving_output

* Don't call get_encoderdecoder_model in TFEncoderDecoderModelTest.test_configuration_tie until we have a TF checkpoint on Hub

* Apply review suggestions - style for cross_attns in serving_output

* Apply review suggestions - change assert + docstrings

* break the error message to respect the char limit

* deprecate the argument past

* fix docstring style

* Update the encoder-decoder rst file

* fix Unknown interpreted text role "method"

* fix typo
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* correct long to bool

* up

* correct code

* adapt wav2vec2

* add example

* add files

* adapt

* remove bogus file

* Apply suggestions from code review

* adapt files more

* upload changes

* del old files

* up

* up

* up

* up

* up

* correct gradient checkpoitning

* add readme

* finish

* finish

* up

* more fixes

* up

* up

* add demo run to readme

* up

[Gradient checkpoining] Correct disabling `find_unused_parameters` in Trainer when gradient checkpointing is enabled (#13961)

* up

* correct test

* Honor existing attention mask in tokenzier.pad

* Fix initialization of attention mask

* Roll the implem on all subclasses

* Fix tests

* up

* Update src/transformers/generation_stopping_criteria.py

* finish

* Adding support for tokens being suffixes or part of each other.

* Better test name.

* Implement img seg pipeline

* Update src/transformers/pipelines/image_segmentation.py
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>

* Update src/transformers/pipelines/image_segmentation.py
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>

* Update output shape with individual masks

* Rm dev change

* Remove loops in test
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>

* Fix issues with LED model

* Style pass

* Bugfixes

* correct attentions as well
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* up

* overwrite hubert

Fixes #13846

Co-authored-by: Pierre Snell <pierre.snell@botpress.com>
Co-authored-by: Pierre Snell <pierre.snell@botpress.com>

* Tmp.

* Fixing BC for question answering with long context.

* Capping model_max_length to avoid tf overflow.

* Bad workaround bugged roberta.

* Fixing name.

* Allow dataset to be an optional argument for (Distributed)LengthGroupedSampler

* Fix

* Symbolic trace dynamic axes support for BERT like models (albert, bert, distilbert, mobilebert, electra, megatron-bert)
* Sanity checks before tracing that make sure the model to trace is supported
* Adapted to PyTorch 1.9
Co-authored-by: Michael Benayoun <michael@huggingface.co>

* Fixing empty prompts for text-generation when BOS exists.

* Fixing odd case with Pegasus.

* Fixing Bert is Assertion Error.

* update no_* argument

Changes the order so that the no_* argument is created after the original argument AND sets the default for this no_* argument to False

* import copy

* update test

* make style

* Use kwargs to set default=False

* make style

* Add layer-wise scaling

* Add reorder & upcasting argument

* Add OpenAI GPT-2 weight initialization scheme

* start `layer_idx` count at zero for consistency

* disentangle attn and reordered and upscaled attn function

* rename `scale_attn_by_layer` to `scale_attn_by_layer_id`

* make autocast from amp compatible with pytorch<1.6

* fix docstring

* style fixes

* Add fixes from PR feedback, style tweaks

* Fix doc whitespace

* Reformat

* First pass scale_attn_by_layer_idx and reorder_and_upcast_attn tests

* Rename scale_attn_by_layer_idx, add tip

* Remove extra newline

* add test for weight initialization

* update code format

* add assert check weights are fp32

* remove assert

* Fix incorrect merge

* Fix shape mismatch in baddbmm

* Add generation test for Mistral flags
Co-authored-by: leandro <leandro.vonwerra@spoud.io>
Co-authored-by: Keshav Santhanam <keshav2@stanford.edu>
Co-authored-by: J38 <jebolton@stanford.edu>

* update

* add to docs and init

* make fix-copies

* Fix length of IterableDatasetShard and add test

* Add comments

* Enable readme link synchronization

* Style

* Reuse regex pattern

* Apply suggestions

* Update

Fixes #13697