GitLab

* don't reconvert when the type is already right

* better name

* adjust logic as suggested

* merge

* `disable_ngram_loss` fix for prophetnet

* add changes documentation

* fix _compute_loss to use mean reduction and -100 to masked tokens & remove unnecessary arguments

* mean label smoothing loss

* small refactor

* fix test
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>

* Better filtering of the model outputs in Trainer

* Fix examples tests

* Add test for Lysandre

* CI should install `sentencepiece`

* Requiring TF

* Fixing some TFDPR bugs

* remove return_dict=False/True hack
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>

* working on LongformerForSequenceClassification

* add TFLongformerForMultipleChoice

* add TFLongformerForTokenClassification

* use add_start_docstrings_to_model_forward

* test TFLongformerForSequenceClassification

* test TFLongformerForMultipleChoice

* test TFLongformerForTokenClassification

* remove test from repo

* add test and doc for TFLongformerForSequenceClassification, TFLongformerForTokenClassification, TFLongformerForMultipleChoice

* add requested classes to modeling_tf_auto.py
update dummy_tf_objects
fix tests
fix bugs in requested classes

* pass all tests except test_inputs_embeds

* sync with master

* pass all tests except test_inputs_embeds

* pass all tests

* pass all tests

* work on test_inputs_embeds

* fix style and quality

* make multi choice work

* fix TFLongformerForTokenClassification signature

* fix TFLongformerForMultipleChoice, TFLongformerForSequenceClassification signature

* fix mult choice

* fix mc hint

* fix input embeds

* fix input embeds

* refactor input embeds

* fix copy issue

* apply sylvains changes and clean more
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Adding PrefixConstrainedLogitsProcessor

* fixing RAG and style_doc

* fixing black (v20 instead of v19)

* Improving doc in generation_logits_process.py

* Improving docs and typing in generation_utils.py

* docs improvement

* adding test and fixing doc typo

* fixing doc_len

* isort on test

* fixed test

* improve docstring a bit
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Remove old deprecated arguments
Co-authored-by: LysandreJik <lysandre.debut@reseau.eseo.fr>

* Remove needless imports

* Fix tests
Co-authored-by: LysandreJik <lysandre.debut@reseau.eseo.fr>

* Tokenizers should be framework agnostic

* Run the slow tests

* Not testing

* Fix documentation

* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* add mt5 and t5v1_1 model

* fix tests

* correct some imports

* add tf model

* finish tf t5

* improve examples

* fix copies

* clean doc

* Put models in subfolders

* Styling

* Fix imports in tests

* More fixes in test imports

* Sneaky hidden imports

* Fix imports in doc files

* More sneaky imports

* Finish fixing tests

* Fix examples

* Fix path for copies

* More fixes for examples

* Fix dummy files

* More fixes for example

* More model import fixes

* Is this why you're unhappy GitHub?

* Fix imports in conver command

* Use the CI to identify failing tests

* Remove from all examples and tests

* More default switch

* Fixes

* More test fixes

* More fixes

* Last fixes hopefully

* Use the CI to identify failing tests

* Remove from all examples and tests

* More default switch

* Fixes

* More test fixes

* More fixes

* Last fixes hopefully

* Run on the real suite

* Fix slow tests

* Fix passing token_type_ids during GPT2DoubleHeadsModel.generate() if used

and for GPT2LMHeadModel too

* Update tests to check token_type_ids usage in GPT2 models

* Simply insert T5Tokenizer's prepare_seq2seq_batch

* Update/Add some 'import'

* fix RunTimeError caused by '.view'

* Moves .view related error avoidance from seq2seq_trainer to inside prophetnet

* Update test_tokenization_prophetnet.py

* Format the test code with black

* Re-format the test code

* Update test_tokenization_prophetnet.py

* Add importing require_torch in the test code

* Add importing BatchEncoding in the test code

* Re-format the test code on Colab

[breaking|pipelines|tokenizers] Adding slow-fast tokenizers equivalence tests pipelines - Removing sentencepiece as a required dependency (#8073)

* Fixing roberta for slow-fast tests

* WIP getting equivalence on pipelines

* slow-to-fast equivalence - working on question-answering pipeline

* optional FAISS tests

* Pipeline Q&A

* Move pipeline tests to their own test job again

* update tokenizer to add sequence id methods

* update to tokenizers 0.9.4

* set sentencepiecce as optional

* clean up squad

* clean up pipelines to use sequence_ids

* style/quality

* wording

* Switch to use_fast = True by default

* update tests for use_fast at True by default

* fix rag tokenizer test

* removing protobuf from required dependencies

* fix NER test for use_fast = True by default

* fixing example tests (Q&A examples use slow tokenizers for now)

* protobuf in main deps extras["sentencepiece"] and example deps

* fix protobug install test

* try to fix seq2seq by switching to slow tokenizers for now

* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Update some tests

* Small update

* Apply style

* Use max_position_embeddings

* Create a fake attribute

* Create a fake attribute

* Update wrong name

* Wrong TransfoXL model file

* Keep the common tests agnostic

* fix bug

* T5 refactor

* refactor tf

* apply sylvains suggestions

* Add pretraining loss computation for TF Bert pretraining

* Fix labels creation

* Fix T5 model

* restore T5 kwargs

* try a generic fix for pretraining models

* Apply style

* Overide the prepare method for the BERT tests

* Create modeling_tf_dpr.py

* Add TFDPR

* Add back TFPegasus, TFMarian, TFMBart, TFBlenderBot

last commit accidentally deleted these 4 lines, so I recover them back

* Add TFDPR

* Add TFDPR

* clean up some comments, add TF input-style doc string

* Add TFDPR

* Make return_dict=False as default

* Fix return_dict bug (in .from_pretrained)

* Add get_input_embeddings()

* Create test_modeling_tf_dpr.py

The current version is already passed all 27 tests!
Please see the test run at : 
https://colab.research.google.com/drive/1czS_m9zy5k-iSJbzA_DP1k1xAAC_sdkf?usp=sharing



* fix quality

* delete init weights

* run fix copies

* fix repo consis

* del config_class, load_tf_weights

They shoud be 'pytorch only'

* add config_class back

after removing it, test failed ... so totally only removing "use_tf_weights = None" on Lysandre suggestion

* newline after .. note::

* import tf, np (Necessary for ModelIntegrationTest)

* slow_test from_pretrained with from_pt=True

At the moment we don't have TF weights (since we don't have official official TF model)
Previously, I did not run slow test, so I missed this bug

* Add simple TFDPRModelIntegrationTest

Note that this is just a test that TF and Pytorch gives approx. the same output.
However, I could not test with the official DPR repo's output yet

* upload correct tf model

* remove position_ids as missing keys
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: patrickvonplaten <patrick@huggingface.co>

* Add next sentence prediction loss computation

* Apply style

* Fix tests

* Add forgotten import

* Add forgotten import

* Use a new parameter

* Remove kwargs and use positional arguments

* stash

* TF Integration testing for ELECTRA, BERT, Longformer

* Trigger slow tests

* Apply suggestions from code review

* s|multiple_gpu|multi_gpu|g; s|multigpu|multi_gpu|g'

* doc

* Patch token classification pipeline

* Some added tests for TokenClassificationArgumentHandler (#8366)
Co-authored-by: Nicolas Patry <patry.nicolas@protonmail.com>

* fix typo

* rm use_cdn & references, and implement new hf_bucket_url

* I'm pretty sure we don't need to `read` this file

* same here

* [BIG] file_utils.networking: do not gobble up errors anymore

* Fix CI 😇



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

* Tiny doc tweak

* Add doc + pass kwarg everywhere

* Add more tests and explain

cc @sshleifer let me know if better
Co-Authored-By: Sam Shleifer <sshleifer@gmail.com>

* Also implement revision in pipelines

In the case where we're passing a task name or a string model identifier

* Fix CI 😇



* Fix CI

* [hf_api] new methods + command line implem

* make style

* Final endpoints post-migration

* Fix post-migration

* Py3.6 compat

cc @stefan-it

Thank you @stas00
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sam Shleifer <sshleifer@gmail.com>

* add training tests

* correct longformer

* fix docs

* fix some tests

* fix some more train tests

* remove ipdb

* fix multiple edge case model training

* fix funnel and prophetnet

* clean gpt models

* undo renaming of albert

* Add new token classification example

* Remove txt file

* Add test

* With actual testing done

* Less warmup is better

* Update examples/token-classification/run_ner_new.py
Co-authored-by: Thomas Wolf <thomwolf@users.noreply.github.com>

* Address review comments

* Fix test

* Make Lysandre happy

* Last touches and rename

* Rename in tests

* Address review comments

* More run_ner -> run_ner_old
Co-authored-by: Thomas Wolf <thomwolf@users.noreply.github.com>

* support lowercase tokenizer

* fix arg pos

[All Seq2Seq model + CLM models that can be used with EncoderDecoder] Add cross-attention weights to outputs (#8071)

* Output cross-attention with decoder attention output

* Update src/transformers/modeling_bert.py

* add cross-attention for t5 and bart as well

* fix tests

* correct typo in docs

* add sylvains and sams comments

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

* Make Trainer evaluation handle dynamic seq_length

* Document behavior.

* Fix test

* Better fix

* Fixes for realsies this time

* Address review comments

* Without forgetting to save...

* Output global_attentions in Longformer models

* make style

* small refactoring

* fix tests

* make fix-copies

* add for tf as well

* remove comments in test

* make fix-copies

* make style

* add docs

* make docstring pretty
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>

* Clean up data collators and datasets

* Apply suggestions from code review
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Remove needless clone
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

- The issue is that with previous code we would have the following:

```python
qa_pipeline = (...)
qa_pipeline(question="Where was he born ?", context="")
-> IndexError: Dimension out of range (expected to be in range of [-1, 0], but got 1)
```

The goal here is to improve this to actually return a ValueError
wherever possible.

While at it, I tried to simplify QuestionArgumentHandler's code to
make it smaller and more compat while keeping backward compat.

* fix greedy generate test

* delet ipdb

* Bug fix: NER pipeline shouldn't group separate entities of same type

* style fix

* [Bug Fix] Shouldn't group entities that are both 'B' even if they are same type
	(B-type1 B-type1) != (B-type1 I-type1)
[Bug Fix] add an option `ignore_subwords` to ignore subsequent ##wordpieces in predictions. Because some models train on only the first token of a word and not on the subsequent wordpieces (BERT NER default). So it makes sense doing the same thing at inference time.
	The simplest fix is to just group the subwords with the first wordpiece.
	[TODO] how to handle ignored scores? just set them to 0 and calculate zero invariant mean ?
	[TODO] handle different wordpiece_prefix ## ? possible approaches:
		get it from tokenizer? but currently most tokenizers dont have a wordpiece_prefix property?
		have an _is_subword(token)
[Feature add] added option to `skip_special_tokens`. Cause It was harder to remove them after grouping.
[Additional Changes] remove B/I prefix on returned grouped_entities
[Feature Request/TODO] Return indexes?
[Bug TODO]  can't use fast tokenizer with grouped_entities ('BertTokenizerFast' object has no attribute 'convert_tokens_to_string')

* use offset_mapping to fix [UNK] token problem

* ignore score for subwords

* modify ner_pipeline test

* modify ner_pipeline test

* modify ner_pipeline test

* ner_pipeline change ignore_subwords default to true

* add ner_pipeline ignore_subword=False test case

* fix offset_mapping index

* fix style again duh

* change is_subword and convert_tokens_to_string logic

* merge tests with new test structure

* change test names

* remove old tests

* ner tests for fast tokenizer

* fast tokenizers have convert_tokens_to_string

* Fix the incorrect merge
Co-authored-by: Ceyda Cinarel <snu-ceyda@users.noreply.github.com>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>

* make it possible to invoke testconf.py in both test suites without crashing on having the same option added

* perl -pi -e 's|--make_reports|--make-reports|' to be consistent with other opts

* add `pytest --make-reports` to all CIs (and artifacts)

* fix

* Add DataCollatorForTokenClassification and clean tests

* Make quality