* Add {decoder_,}head_mask to LED
* Fix create_custom_forward signatue in encoder
* Add head_mask to longformer
* Add head_mask to longformer to fix dependencies
of LED on Longformer.
* Not working yet
* Add mising one input in longofrmer_modeling.py
* make fix-copies
* Add head_mask/decoder_head_mask for BART
This branch implement head_mask and decoder_head_mask
for BART-based models. Full list below:
- BART
- MBart
- Blenderbot
- BlenderbotSmall
- Marian
- Pegasus
Everything is accompanied with updated testing.
* Fix test_headmasking for BART models
* Fix text_headmasking for BART-like models
which has only 2 layers in each modules.
The condition
```
self.assertNotEqual(attentions[1][..., 0, :, :].flatten().sum().item(), 0.0)
```
is, therefore, invalid for encoder-decoder models considering
the `head_mask`
```
head_mask = torch.ones(
self.model_tester.num_hidden_layers,
self.model_tester.num_attention_heads,
device=torch_device,
)
head_mask[0, 0] = 0
head_mask[-1, :-1] = 0
```
specified in the `test_headmasking` test/function.
* Adjust test_modeling_common.py to reflect T5 input args
* Update tests/test_modeling_common.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* make style
* make fix-copies
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* remove make on the fly linear embedding
* start refactor
* big first refactor
* save intermediate
* save intermediat
* correct mask issue
* save tests
* refactor padding masks
* make all tests pass
* further refactor
* make pegasus test pass
* fix bool if
* fix leftover tests
* continue
* bart renaming
* delete torchscript test hack
* fix imports in tests
* correct shift
* fix docs and repo cons
* re-add fix for FSTM
* typo in test
* fix typo
* fix another typo
* continue
* hot fix 2 for tf
* small fixes
* refactor types linting
* continue
* finish refactor
* fix import in tests
* better bart names
* further refactor and add test
* delete hack
* apply sylvains and lysandres commens
* small perf improv
* further perf improv
* improv perf
* fix typo
* make style
* small perf improv
* Fix decoder not returning hidden states from the last layer
* Resolve conflict
* Change the way to gather hidden states
* Add decoder hidden states test
* Make pytest and black happy
* Remove redundant line
* remove new line
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>
* bart output hidden states upstream
* same w/ decoder
* add tests
* fix prophetnet
* fix gpt2 and ctrl
* fix fstm and skip test for reformer and longformer
* fix all models
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* 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
* 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
* 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>
* 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>
* first draft
* show design proposition for new generate method
* up
* make better readable
* make first version
* gpt2 tests pass
* make beam search for gpt2 work
* add first encoder-decoder code
* delete typo
* make t5 work
* save indermediate
* make bart work with beam search
* finish beam search bart / t5
* add default kwargs
* make more tests pass
* fix no bad words sampler
* some fixes and tests for all distribution processors
* fix test
* fix rag slow tests
* merge to master
* add nograd to generate
* make all slow tests pass
* speed up generate
* fix edge case bug
* small fix
* correct typo
* add type hints and docstrings
* fix typos in tests
* add beam search tests
* add tests for beam scorer
* fix test rag
* finish beam search tests
* move generation tests in seperate file
* fix generation tests
* more tests
* add aggressive generation tests
* fix tests
* add gpt2 sample test
* add more docstring
* add more docs
* finish doc strings
* apply some more of sylvains and sams comments
* fix some typos
* make fix copies
* apply lysandres and sylvains comments
* final corrections on examples
* small fix for reformer
* Test TF GPU CI
* Change cache
* Fix missing torch requirement
* Fix some model tests
Style
* LXMERT
* MobileBERT
* Longformer skip test
* XLNet
* The rest of the tests
* RAG goes OOM in multi gpu setup
* YAML test files
* Last fixes
* Skip doctests
* Fill mask tests
* Yaml files
* Last test fix
* Style
* Update cache
* Change ONNX tests to slow + use tiny model
* make the save_load special key tests common
* handle mbart
* cleaner solution
* fix
* move test_save_load_missing_keys back into fstm for now
* restore
* style
* add marian
* add pegasus
* blenderbot
* revert - no static embed
* Initial model
* Fix upsampling
* Add special cls token id and test
* Formatting
* Test and fist FunnelTokenizerFast
* Common tests
* Fix the check_repo script and document Funnel
* Doc fixes
* Add all models
* Write doc
* Fix test
* Initial model
* Fix upsampling
* Add special cls token id and test
* Formatting
* Test and fist FunnelTokenizerFast
* Common tests
* Fix the check_repo script and document Funnel
* Doc fixes
* Add all models
* Write doc
* Fix test
* Fix copyright
* Forgot some layers can be repeated
* Apply suggestions from code review
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Update src/transformers/modeling_funnel.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Address review comments
* Update src/transformers/modeling_funnel.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Address review comments
* Update src/transformers/modeling_funnel.py
Co-authored-by: Sam Shleifer <sshleifer@gmail.com>
* Slow integration test
* Make small integration test
* Formatting
* Add checkpoint and separate classification head
* Formatting
* Expand list, fix link and add in pretrained models
* Styling
* Add the model in all summaries
* Typo fixes
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Sam Shleifer <sshleifer@gmail.com>
Since `generate()` does:
```
num_beams = num_beams if num_beams is not None else self.config.num_beams
```
This test fails if `model.config.num_beams > 1` (which is the case in the model I'm porting).
This fix makes the test setup unambiguous by passing an explicit `num_beams=1` to `generate()`.
Thanks.
* Feed forward chunking for Distilbert & Albert
* Added ff chunking for many other models
* Change model signature
* Added chunking for XLM
* Cleaned up by removing some variables.
* remove test_chunking flag
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>
* Chunked feed forward for Bert
This is an initial implementation to test applying feed forward chunking for BERT.
Will need additional modifications based on output and benchmark results.
* Black and cleanup
* Feed forward chunking in BertLayer class.
* Isort
* add chunking for all models
* fix docs
* Fix typo
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>
* Switch from return_tuple to return_dict
* Fix test
* [WIP] Test TF Flaubert + Add {XLM, Flaubert}{TokenClassification, MultipleC… (#5614)
* Test TF Flaubert + Add {XLM, Flaubert}{TokenClassification, MultipleChoice} models and tests
* AutoModels
Tiny tweaks
* Style
* Final changes before merge
* Re-order for simpler review
* Final fixes
* Addressing @sgugger's comments
* Test MultipleChoice
* Rework TF trainer (#6038)
* Fully rework training/prediction loops
* fix method name
* Fix variable name
* Fix property name
* Fix scope
* Fix method name
* Fix tuple index
* Fix tuple index
* Fix indentation
* Fix variable name
* fix eval before log
* Add drop remainder for test dataset
* Fix step number + fix logging datetime
* fix eval loss value
* use global step instead of step + fix logging at step 0
* Fix logging datetime
* Fix global_step usage
* Fix breaking loop + logging datetime
* Fix step in prediction loop
* Fix step breaking
* Fix train/test loops
* Force TF at least 2.2 for the trainer
* Use assert_cardinality to facilitate the dataset size computation
* Log steps per epoch
* Make tfds compliant with TPU
* Make tfds compliant with TPU
* Use TF dataset enumerate instead of the Python one
* revert previous commit
* Fix data_dir
* Apply style
* rebase on master
* Address Sylvain's comments
* Address Sylvain's and Lysandre comments
* Trigger CI
* Remove unused import
* Switch from return_tuple to return_dict
* Fix test
* Add recent model
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Julien Plu <plu.julien@gmail.com>
* DataParallel fixes:
1. switched to a more precise check
- if self.args.n_gpu > 1:
+ if isinstance(model, nn.DataParallel):
2. fix tests - require the same fixup under DataParallel as the training module
* another fix
* fix-5181
Padding to max sequence length while truncation to another length was wrong on slow tokenizers
* clean up and fix#5155
* fix XLM test
* Fix tests for Transfo-XL
* logging only above WARNING in tests
* switch slow tokenizers tests in @slow
* fix Marian truncation tokenization test
* style and quality
* make the test a lot faster by limiting the sequence length used in tests
* Configure all models to use output_hidden_states as argument passed to foward()
* Pass all tests
* Remove cast_bool_to_primitive in TF Flaubert model
* correct tf xlnet
* add pytorch test
* add tf test
* Fix broken tests
* Configure all models to use output_hidden_states as argument passed to foward()
* Pass all tests
* Remove cast_bool_to_primitive in TF Flaubert model
* correct tf xlnet
* add pytorch test
* add tf test
* Fix broken tests
* Refactor output_hidden_states for mobilebert
* Reset and remerge to master
Co-authored-by: Joseph Liu <joseph.liu@coinflex.com>
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>
