* Init FNet
* Update config
* Fix config
* Update model classes
* Update tokenizers to use sentencepiece
* Fix errors in model
* Fix defaults in config
* Remove position embedding type completely
* Fix typo and take only real numbers
* Fix type vocab size in configuration
* Add projection layer to embeddings
* Fix position ids bug in embeddings
* Add minor changes
* Add conversion script and remove CausalLM vestiges
* Fix conversion script
* Fix conversion script
* Remove CausalLM Test
* Update checkpoint names to dummy checkpoints
* Add tokenizer mapping
* Fix modeling file and corresponding tests
* Add tokenization test file
* Add PreTraining model test
* Make style and quality
* Make tokenization base tests work
* Update docs
* Add FastTokenizer tests
* Fix fast tokenizer special tokens
* Fix style and quality
* Remove load_tf_weights vestiges
* Add FNet to main README
* Fix configuration example indentation
* Comment tokenization slow test
* Fix style
* Add changes from review
* Fix style
* Remove bos and eos tokens from tokenizers
* Add tokenizer slow test, TPU transforms, NSP
* Add scipy check
* Add scipy availabilty check to test
* Fix tokenizer and use correct inputs
* Remove remaining TODOs
* Fix tests
* Fix tests
* Comment Fourier Test
* Uncomment Fourier Test
* Change to google checkpoint
* Add changes from review
* Fix activation function
* Fix model integration test
* Add more integration tests
* Add comparison steps to MLM integration test
* Fix style
* Add masked tokenization fix
* Improve mask tokenization fix
* Fix index docs
* Add changes from review
* Fix issue
* Fix failing import in test
* some more fixes
* correct fast tokenizer
* finalize
* make style
* Remove additional tokenization logic
* Set do_lower_case to False
* Allow keeping accents
* Fix tokenization test
* Fix FNet Tokenizer Fast
* fix tests
* make style
* Add tips to FNet docs
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>
* Enabling dataset iteration on pipelines.
Enabling dataset iteration on pipelines.
Unifying parameters under `set_parameters` function.
Small fix.
Last fixes after rebase
Remove print.
Fixing text2text `generate_kwargs`
No more `self.max_length`.
Fixing tf only conversational.
Consistency in start/stop index over TF/PT.
Speeding up drastically on TF (nasty bug where max_length would increase
a ton.)
Adding test for support for non fast tokenizers.
Fixign GPU usage on zero-shot.
Fix working on Tf.
Update src/transformers/pipelines/base.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Update src/transformers/pipelines/base.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Small cleanup.
Remove all asserts + simple format.
* Fixing audio-classification for large PR.
* Overly explicity null checking.
* Encapsulating GPU/CPU pytorch manipulation directly within `base.py`.
* Removed internal state for parameters of the pipeline.
Instead of overriding implicitly internal state, we moved
to real named arguments on every `preprocess`, `_forward`,
`postprocess` function.
Instead `_sanitize_parameters` will be used to split all kwargs
of both __init__ and __call__ into the 3 kinds of named parameters.
* Move import warnings.
* Small fixes.
* Quality.
* Another small fix, using the CI to debug faster.
* Last fixes.
* Last fix.
* Small cleanup of tensor moving.
* is not None.
* Adding a bunch of docs + a iteration test.
* Fixing doc style.
* KeyDataset = None guard.
* RRemoving the Cuda test for pipelines (was testing).
* Even more simple iteration test.
* Correct import .
* Long day.
* Fixes in docs.
* [WIP] migrating object detection.
* Fixed the target_size bug.
* Fixup.
* Bad variable name.
* Fixing `ensure_on_device` respects original ModelOutput.
* [docs] update dead quickstart link on resuing past for GPT2
Thed dead link have been replaced by two links of forward and call methods of the GPT2 class for torch and tensorflow respectively.
* [docs] fix formatting for gpt2 page update
* refactor GPT Config to allow dyn. properties
* make attribute_map a class attribute
* remove old code
* update unit test to test config: Add test for common properties setter
* update unit test to test config: Add test for common properties passed as parameters to __init__
* update to black code format
* Allow that setters are not defined for certain config classes
* update config classes to implement attribute_map
* bugfix lxmert config - id2labels was not defined when num_labels was set
* update broken configs - add attribute_maps
* update bart config
* update black codestyle
* update documentation on common config attributes
* update GPTJ config to new attribute map
* update docs on common attributes
* gptj config: add max_position_embeddings
* gptj config: format with black
* update speech to text 2 config
* format doc file to max_len 119
* update config template
* [docs] Update perplexity.rst to use negative log likelihood
Model `forward` returns the negative log likelihood. The document correctly defines and calculates perplexity, but the description and variable names are inconsistent, which might cause confusion.
* [docs] restyle perplexity.rst
* fix_torch_device_generate_test
* remove @
* up
* correct some bugs
* correct model
* finish speech2text extension
* up
* up
* up
* up
* Update utils/custom_init_isort.py
* up
* up
* update with tokenizer
* correct old tok
* correct old tok
* fix bug
* up
* up
* add more tests
* up
* fix docs
* up
* fix some more tests
* add better config
* correct some more things
"
* fix tests
* improve docs
* Apply suggestions from code review
* Apply suggestions from code review
* final fixes
* finalize
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* apply suggestions Lysandre and Sylvain
* apply nicos suggestions
* upload everything
* finish
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
Co-authored-by: your_github_username <your_github_email>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Add the audio classification pipeline
* Remove autoconfig exception
* Mark ffmpeg test as slow
* Rearrange pipeline tests
* Add small test
* Replace asserts with ValueError
* Adding a TF variant of the DataCollatorForTokenClassification to get feedback
* Added a Numpy variant and a post_init check to fail early if a missing import is found
* Fixed call to Numpy variant
* Added a couple more of the collators
* Update src/transformers/data/data_collator.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Fixes, style pass, finished DataCollatorForSeqToSeq
* Added all the LanguageModeling DataCollators, except SOP and PermutationLanguageModeling
* Adding DataCollatorForPermutationLanguageModeling
* Style pass
* Add missing `__call__` for PLM
* Remove `post_init` checks for frameworks because the imports inside them were making us fail code quality checks
* Remove unused imports
* First attempt at some TF tests
* A second attempt to make any of those tests actually work
* TF tests, round three
* TF tests, round four
* TF tests, round five
* TF tests, all enabled!
* Style pass
* Merging tests into `test_data_collator.py`
* Merging tests into `test_data_collator.py`
* Fixing up test imports
* Fixing up test imports
* Trying shuffling the conditionals around
* Commenting out non-functional old tests
* Completed all tests for all three frameworks
* Style pass
* Fixed test typo
* Style pass
* Move standard `__call__` method to mixin
* Rearranged imports for `test_data_collator`
* Fix data collator typo "torch" -> "pt"
* Fixed the most embarrassingly obvious bug
* Update src/transformers/data/data_collator.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Renaming mixin
* Updating docs
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Dalton Walker <dalton_walker@icloud.com>
Co-authored-by: Andrew Romans <andrew.romans@hotmail.com>
* Deberta_v2 tf
* added new line at the end of file, make style
* +V2, typo
* remove never executed branch of code
* rm cmnt and fixed typo in url filter
* cleanup according to review comments
* added #Copied from
* First commit
* Make style
* Fix dummy objects
* Add Detectron2 config
* Add LayoutLMv2 pooler
* More improvements, add documentation
* More improvements
* Add model tests
* Add clarification regarding image input
* Improve integration test
* Fix bug
* Fix another bug
* Fix another bug
* Fix another bug
* More improvements
* Make more tests pass
* Make more tests pass
* Improve integration test
* Remove gradient checkpointing and add head masking
* Add integration test
* Add LayoutLMv2ForSequenceClassification to the tests
* Add LayoutLMv2ForQuestionAnswering
* More improvements
* More improvements
* Small improvements
* Fix _LazyModule
* Fix fast tokenizer
* Move sync_batch_norm to a separate method
* Replace dummies by requires_backends
* Move calculation of visual bounding boxes to separate method + update README
* Add models to main init
* First draft
* More improvements
* More improvements
* More improvements
* More improvements
* More improvements
* Remove is_split_into_words
* More improvements
* Simply tesseract - no use of pandas anymore
* Add LayoutLMv2Processor
* Update is_pytesseract_available
* Fix bugs
* Improve feature extractor
* Fix bug
* Add print statement
* Add truncation of bounding boxes
* Add tests for LayoutLMv2FeatureExtractor and LayoutLMv2Tokenizer
* Improve tokenizer tests
* Make more tokenizer tests pass
* Make more tests pass, add integration tests
* Finish integration tests
* More improvements
* More improvements - update API of the tokenizer
* More improvements
* Remove support for VQA training
* Remove some files
* Improve feature extractor
* Improve documentation and one more tokenizer test
* Make quality and small docs improvements
* Add batched tests for LayoutLMv2Processor, remove fast tokenizer
* Add truncation of labels
* Apply suggestions from code review
* Improve processor tests
* Fix failing tests and add suggestion from code review
* Fix tokenizer test
* Add detectron2 CI job
* Simplify CI job
* Comment out non-detectron2 jobs and specify number of processes
* Add pip install torchvision
* Add durations to see which tests are slow
* Fix tokenizer test and make model tests smaller
* Frist draft
* Use setattr
* Possible fix
* Proposal with configuration
* First draft of fast tokenizer
* More improvements
* Enable fast tokenizer tests
* Make more tests pass
* Make more tests pass
* More improvements
* Addd padding to fast tokenizer
* Mkae more tests pass
* Make more tests pass
* Make all tests pass for fast tokenizer
* Make fast tokenizer support overflowing boxes and labels
* Add support for overflowing_labels to slow tokenizer
* Add support for fast tokenizer to the processor
* Update processor tests for both slow and fast tokenizers
* Add head models to model mappings
* Make style & quality
* Remove Detectron2 config file
* Add configurable option to label all subwords
* Fix test
* Skip visual segment embeddings in test
* Use ResNet-18 backbone in tests instead of ResNet-101
* Proposal
* Re-enable all jobs on CI
* Fix installation of tesseract
* Fix failing test
* Fix index table
* Add LayoutXLM doc page, first draft of code examples
* Improve documentation a lot
* Update expected boxes for Tesseract 4.0.0 beta
* Use offsets to create labels instead of checking if they start with ##
* Update expected boxes for Tesseract 4.1.1
* Fix conflict
* Make variable names cleaner, add docstring, add link to notebooks
* Revert "Fix conflict"
This reverts commit a9b46ce9afe47ebfcfe7b45e6a121d49e74ef2c5.
* Revert to make integration test pass
* Apply suggestions from @LysandreJik's review
* Address @patrickvonplaten's comments
* Remove fixtures DocVQA in favor of dataset on the hub
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
* Add hubert classifier + tests
* Add hubert classifier + tests
* Dummies for all classification tests
* Wav2Vec2 classifier + ER test
* Fix hubert integration tests
* Add hubert IC
* Pass tests for all classification tasks on Hubert
* Pass all tests + copies
* Move models to the SUPERB org
* First commit
* Add interpolation of patch embeddings
* Comment out code
* Fix bug
* Fix another bug
* Fix bug
* Fix another bug
* Remove print statements
* Update conversion script
* Use the official vit implementation
* Add support for converting dino_vits8
* Add DINO to docs of ViT
* Remove assertion
* Add interpolation of position encodings
* Fix bug
* Add align_corners
* Add interpolate_pos_encoding option to forward pass of ViTModel
* Improve interpolate_pos_encoding method
* Add docstring
* make flax gpt2 working with cross attention
* Remove encoder->decoder projection layer
* A draft (incomplete) for FlaxEncoderDecoderModel
* Add the method from_encoder_decoder_pretrained + the docstrings
* Fix the mistakes of using EncoderDecoderModel
* Fix style
* Add FlaxEncoderDecoderModel to the library
* Fix cyclic imports
* Add FlaxEncoderDecoderModel to modeling_flax_auto.py
* Remove question comments
* add tests for FlaxEncoderDecoderModel
* add flax_encoder_decoder to the lists of ignored entries in check_repo.py
* fix missing required positional arguments
* Remove **kwargs when creating FlaxEncoderDecoderModel in from_encoder_decoder_pretrained()
Also fix generation eos/pad tokens issue
* Fix: Use sequences from the generated_output
* Change a check from assert to raise ValueError
* Fix examples and token ids issues
* Fix missing all_cross_attentions when outputting tuple in modeling_gpt2
* Remove the changes in configuration docstrings.
* allow for bert 2 gpt2
* make fix-copies
* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Change remaining examples to bert2gpt2
* Change the test to Bert2GPT2
* Fix examples
* Fix import
* Fix unpack bug
* Rename to FlaxEncoderDecoderModelTest and change the test to bert2gpt2
* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Fix: NotImplentedError -> NotImplementedError
* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* up
* finalize
Co-authored-by: ydshieh <ydshieh@user.noreply>
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Doctests
* Limit to 4 decimals
* Try with separate PT/TF tests
* Remove test for TF
* Ellips the predictions
* Doctest continue on failure
Co-authored-by: Sylvain Gugger <sylvain.gugger@gmail.com>
* Fix doctests for quicktour
* Adapt causal LM exemple
* Remove space
* Fix until summarization
* End of task summary
* Style
* With last changes in quicktour
* Adding HuggingArtists to Community Notebooks
* Adding HuggingArtists to Community Notebooks
* Adding HuggingArtists to Community Notebooks
* docs: add HuggingArtists to community notebooks
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* First pass
* Make conversion script work
* Improve conversion script
* Fix bug, conversion script working
* Improve conversion script, implement BEiTFeatureExtractor
* Make conversion script work based on URL
* Improve conversion script
* Add tests, add documentation
* Fix bug in conversion script
* Fix another bug
* Add support for converting masked image modeling model
* Add support for converting masked image modeling
* Fix bug
* Add print statement for debugging
* Fix another bug
* Make conversion script finally work for masked image modeling models
* Move id2label for datasets to JSON files on the hub
* Make sure id's are read in as integers
* Add integration tests
* Make style & quality
* Fix test, add BEiT to README
* Apply suggestions from @sgugger's review
* Apply suggestions from code review
* Make quality
* Replace nielsr by microsoft in tests, add docs
* Rename BEiT to Beit
* Minor fix
* Fix docs of BeitForMaskedImageModeling
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Faster list concat for trainer_pt_utils.get_length_grouped_indices() (#11825)
get_length_grouped_indices() in LengthGroupedSampler and DistributedLengthGroupedSampler
is prohibitively slow for large number of megabatches (in test case takes hours for ~270k
megabatches with 100 items each) due to slow list concatenation with sum(megabatches, []).
Resolves: #11795
Co-authored-by: ctheodoris <cvtheodo@ds.dfci.harvard.edu>
* Replace double occurrences as the last step (#11367)
* [Flax] Fix PyTorch import error (#11839)
* fix_torch_device_generate_test
* remove @
* change pytorch import to flax import
* Fix reference to XLNet (#11846)
* Switch mem metrics flag (#11851)
* Switch mem metrics flag
* Update src/transformers/training_args.py
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>
* Fix flos single node (#11844)
* fixing flos bug/typo in non-distributed setting
* storing flos every logging_interval
* Fix two typos in docs (#11852)
* typo2
* fix typo
* [Trainer] Report both steps and num samples per second (#11818)
* [Trainer] Report both steps and num samples per second
* Fix batch number
* Update src/transformers/trainer_utils.py
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>
* Address review comments
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>
* Add some tests to the slow suite #11860
* Enable memory metrics in tests that need it (#11859)
* fixed a small typo in the doc (#11856)
* typo (#11858)
* Add option to log only once in multinode training (#11819)
* Add option to long only once in multinode training
* Use an alternate property
* [Wav2Vec2] SpecAugment Fast (#11764)
* first try
* finish
* [lm examples] fix overflow in perplexity calc (#11855)
* fix overflow in perplexity calc
* use inf
* fix
* [Examples] create model with custom config on the fly (#11798)
* create custom model on the flight
* better wording
* add update_from_string
* cleanup
* cleanup
* Update src/transformers/configuration_utils.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* more bool options
* style
* fix logger
* add test
* add the doc
* assert on conflict of options
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* [Wav2Vec2ForCTC] example typo fixed (#11878)
* Ensure input tensor are on device. (#11874)
The feature extractor does not create tensors on the appropriate device,
so we call `ensure_tensor_on_device` before feeding the processed inputs
to the model.
* Fix usage of head masks by TF encoder-decoder models' `generate()` function (#11775)
* Fix Bart
* Fix Blenderbot{,_small}
* Fix LED
* Fix Marian
* Fix MBart
* Fix Pegasus
* Fix T5
* Add test for generation with head_mask
* Add a common TF test
* Override a test for the LED model as head masking is not yet properly implemented
* Remove all head_masks from input preparation for LED
* Drop masking for T5 as it needs a bit of refactor
* Correcting comments in T5Stack to reflect correct tuple order (#11330)
* Correcting comments to reflect correct tuple order
In order to match the actual order (line 513 and 516, and as accessed in 968), I've changed the order mentioned in comments L962 and L966-967.
* Update modeling_t5.py
Updating another comment as well
* Removing extra space
* Fixing style and quality
* style & quality
* Update src/transformers/models/t5/modeling_t5.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* [Flax] Allow dataclasses to be jitted (#11886)
* fix_torch_device_generate_test
* remove @
* change dataclasses to flax ones
* fix typo
* fix jitted tests
* fix bert & electra
* changing find_batch_size to work with tokenizer outputs (#11890)
* changing find_batch_size to work with tokenizer outputs
trainer_pt_utils.find_batch_size does not recognize the batch size of BatchEncoding objects. This can cause an error when a trainer relies on find_batch_size to report the number of observed examples in the evaluation loop.
* Trigger CI
Co-authored-by: jrenner <joseph.renner@inria.fr>
* Link official Cloud TPU JAX docs (#11892)
* Flax Generate (#11777)
* fix_torch_device_generate_test
* remove @
* add
* indexing
* correct a couple of tests
* fix tests
* add logits processor
* finish top_k, top_p, temp
* add docs
* correct flax prng key default
* improve generate
* add generation docs
* add docs
* make style
* revert model outputs change
* make style
* correct typo
* fix tests
* fix slow test
* add raise
* finish generation
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
* Add Emotion Speech Noteboook (#11900)
* Update deepspeed config to reflect hyperparameter search parameters (#11896)
* rebuild deepspeed config for hyperparameter search
* reformat code to fix style issues
* Adding new argument `max_new_tokens` for generate. (#11476)
* Adding new argument `max_new_tokens` for generate.
This is a proposal to add a new argument `max_new_tokens` to `generate`.
This include a `MaxNewTokensCriteria` that enables callers that don't
know about the token length ahead (like pipelines callers) to manage
more easily the length of their generated output.
* Adding a test for the user warning when both`max_length` and
`max_new_tokens` are used together.
* Removed redundant `no_grad`.
* Added Sequence Classification class in GPTNeo (#11906)
* seq classification changes
* fix tests
* [Flax] Return Attention from BERT, ELECTRA, RoBERTa and GPT2 (#11918)
* Added logic to return attention from flax-bert model and added test cases to check that
* Added new line at the end of file to test_modeling_flax_common.py
* fixing code style
* Fixing Roberta and Elextra models too from cpoying bert
* Added temporary hack to not run test_attention_outputs for FlaxGPT2
* Returning attention weights from GPT2 and changed the tests accordingly.
* last fixes
* bump flax dependency
Co-authored-by: jayendra <jayendra@infocusp.in>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Test optuna and ray (#11924)
* Remove `datasets` submodule
* fix assert (#11935)
* Remove redundant `nn.log_softmax` in `run_flax_glue.py` (#11920)
* Remove redundant `nn.log_softmax` in `run_flax_glue.py`
`optax.softmax_cross_entropy` expects unnormalized logits, and so it already calls `nn.log_softmax`, so I believe it is not needed here. `nn.log_softmax` is idempotent so mathematically it shouldn't have made a difference.
* Remove unused 'flax.linen' import
* Add MT5ForConditionalGeneration as supported arch. to summarization README (#11961)
* Add MT5ForConditionalGeneration as supported arch.
* Update README.md
* Add FlaxCLIP (#11883)
* add flax CLIP
* default input_shape
* add tests
* fix test
* fix name
* fix docs
* fix shapes
* attend at least 1 token
* flax conv to torch conv
* return floats
* fix equivalence tests
* fix import
* return attention_weights and update tests
* fix dosctrings
* address patricks comments
* input_shape arg
* add tests for get_image_features and get_text_features methods
* fix tests
* RAG-2nd2end-revamp (#11893)
* initial
* code quality test
* code quality
* added test functions in test_modeling_rag.py and test_retrieval_rag.py to test end2end retreiver
* minor change in test_modeling_rag
* fixed tests
* Update examples/research_projects/rag-end2end-retriever/README.md
typo corrected as suggested by lhoestq
Co-authored-by: Quentin Lhoest <42851186+lhoestq@users.noreply.github.com>
* Update examples/research_projects/rag-end2end-retriever/finetune_rag.py
type change suggested by lhoestq
Co-authored-by: Quentin Lhoest <42851186+lhoestq@users.noreply.github.com>
* Update src/transformers/models/rag/retrieval_rag.py
Adding this change as mentioned by lhoestq.
Co-authored-by: Quentin Lhoest <42851186+lhoestq@users.noreply.github.com>
* completed the minor changes suggested by the reviewers
Co-authored-by: Quentin Lhoest <42851186+lhoestq@users.noreply.github.com>
* modify qa-trainer (#11872)
* modify qa-trainer
* fix flax model
* bugfixes training_args.py (#11922)
modified according to:
https://pytorch.org/xla/release/1.8.1/_modules/torch_xla/core/xla_model.html
* reinitialize wandb config for each hyperparameter search run (#11945)
* Add regression tests for slow sentencepiece tokenizers. (#11737)
* add test_vocab_size for sentencepiece tok.
* add test_get_vocab for sentencepiece tok.
* add test_convert_token_and_id for sentencepiece tok.
* add test_tokenize_and_convert_tokens_to_string for all tok.
* improve test_tokenize_and_convert_tokens_to_string for sp. tok.
* add common tokenizer integration tests
- for albert
- for barthez
* add tokenizer integration tests to bert gen.
* add most tokenizer integration tests
* fix camembert tokenizer integration test
* add tokenizer integration test to marian
* add tokenizer integration test to reformer
* add typing and doc to tokenizer_integration_test_util
* fix tokenizer integration test of reformer
* improve test_sentencepiece_tokenize_and_convert_tokens_to_string
* empty commit to trigger CI
* fix tokenizer integration test of reformer
* remove code not needed anymore
* empty commit to trigger CI
* empty commit to trigger CI
* Authorize args when instantiating an AutoModel (#11956)
* Neptune.ai integration (#11937)
An option that turns on neptune.ai logging
--report_to 'neptune'
Additional ENV variables:
NEPTUNE_PROJECT
NEPTUNE_API_TOKEN
NEPTUNE_RUN_NAME (optional)
NEPTUNE_STOP_TIMEOUT (optional)
* Run the integration tests on schedule tests instead of master tests
* [deepspeed] docs (#11940)
* deepspeed docs
* cleanup
* cleanup
* typo correction (#11973)
* typo correction
* type corrections
* ByT5 model (#11971)
* allow tf to use uneven num of layers
* add tokenizer
* finish docs
* finish docs
* Apply suggestions from code review
* include in index
* finish
* Update docs/source/model_doc/byt5.rst
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>
* apply sylvais suggestions
* make style
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>
* Typo in usage example, changed to device instead of torch_device (#11979)
* [DeepSpeed] decouple `DeepSpeedConfigHF` from `Trainer` (#11966)
* decouple DeepSpeedConfigHF from Trainer
* add LoggingLevel ctx manager; add new test
* cleanup
* add docs
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* implemented suggested renames
* formatter workaround
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* [Trainer] add train loss and flops metrics reports (#11980)
* add train loss and flops metrics reports
* consistency
* add train_loss to skip keys
* restore on_train_end call timing
* Bump urllib3 from 1.25.8 to 1.26.5 in /examples/research_projects/lxmert (#11983)
Bumps [urllib3](https://github.com/urllib3/urllib3) from 1.25.8 to 1.26.5.
- [Release notes](https://github.com/urllib3/urllib3/releases)
- [Changelog](https://github.com/urllib3/urllib3/blob/main/CHANGES.rst)
- [Commits](https://github.com/urllib3/urllib3/compare/1.25.8...1.26.5)
---
updated-dependencies:
- dependency-name: urllib3
dependency-type: direct:production
...
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
* [RAG] Fix rag from pretrained question encoder generator behavior (#11962)
* fix_torch_device_generate_test
* remove @
* fix rag from pretrained loading
* add test
* uplaod
* finish
* VisualBERT (#10534)
* Init VisualBERT
* Add cookie-cutter, Config, and Embeddings
* Add preliminary Model
* Add Bert analogous classes
* Add basic code for NLVR, VQA, Flickr
* Update Init
* Fix VisualBert Downstream Models
* Rename classifier to cls
* Comment position_ids buffer
* Remove sentence image predictor output
* Update output dicts
* Remove unnecessary files
* Fix Auto Modeling
* Fix transformers init
* Add conversion script
* Add conversion script
* Fix docs
* Update visualbert modelling
* Update configuration
* Style fixes
* Add model and integration tests
* Add all tests
* Update model mapping
* Add simple detector from original repository
* Update docs and configs
* Fix style
* Fix style
* Update docs
* Fix style
* Fix import issues in style
* Fix style
* Add changes from review
* Fix style
* Fix style
* Update docs
* Fix style
* Fix style
* Update docs/source/model_doc/visual_bert.rst
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/models/visual_bert/modeling_visual_bert.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update tests/test_modeling_visual_bert.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/models/visual_bert/modeling_visual_bert.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/models/visual_bert/modeling_visual_bert.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/models/visual_bert/modeling_visual_bert.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Add changes from review
* Remove convert run script
* Add changes from review
* Update src/transformers/models/visual_bert/modeling_visual_bert.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/models/visual_bert/modeling_visual_bert.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/models/visual_bert/modeling_visual_bert.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/models/visual_bert/modeling_visual_bert.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/models/visual_bert/modeling_visual_bert.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Add changes from review
* Add changes from review
* Add visual embedding example in docs
* Fix "copied from" comments
* Add changes from review
* Fix error, style, checkpoints
* Update docs
* Fix integration tests
* Fix style
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Fix examples (#11990)
* [docs] fix xref to `PreTrainedModel.generate` (#11049)
* fix xref to generate
* do the same for search methods
* style
* style
* Update return introduction (#11976)
Make it clear that the `forward` method now returns a dict instead of tuple.
Fix style
* [deepspeed] Move code and doc into standalone files (#11984)
* move code and docs
* style
* moved
* restore
* [deepspeed] add nvme test skip rule (#11997)
* add nvme skip rule
* fix
* Fix weight decay masking in `run_flax_glue.py` (#11964)
* Fix weight decay masking in `run_flax_glue.py`
Issues with the previous implementation:
- The `dict` from `traverse_util.flatten_dict` has keys which are tuples of strings, not one long string with the path separated by periods.
- `optax.masked` applies the transformation wherever the mask is True, so the masks are flipped.
- Flax's LayerNorm calls the scale parameter `scale` not `weight`
* Fix formatting with black
* adapt results
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
* [Flax] Refactor MLM (#12013)
* fix_torch_device_generate_test
* remove @
* finish refactor
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
* [Deepspeed] Assert on mismatches between ds and hf args (#12021)
* wip
* add mismatch validation + test
* renames
* Update docs/source/main_classes/deepspeed.rst
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* renames
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* [TrainerArguments] format and sort __repr__, add __str__ (#12018)
* format and sort __repr__, add __str__
* typo
* use __str__ directly
* alias __repr__ = __str__
* Fixed Typo in modeling_bart.py (#12035)
* Fixed Typo in modeling_bart.py - Issue #11895
* Fixed Typo in modeling_bart.py
* fix deberta 2 tokenizer integration test (#12017)
* fix docs of past_key_values (#12049)
* [JAX] Bump jax lib (#12053)
* fix_torch_device_generate_test
* remove @
* bump up jax lib
* Fixes bug that appears when using QA bert and distilation. (#12026)
* Fixing bug that appears when using distilation (and potentially other uses).
During backward pass Pytorch complains with:
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation
This happens because the QA model code modifies the start_positions and end_positions input tensors, using clamp_ function: as a consequence the teacher and the student both modifies the inputs, and backward pass fails.
* Fixing all models QA clamp_ bug.
* Extend pipelines for automodel tupels (#12025)
* fix_torch_device_generate_test
* remove @
* finish
* refactor
* add test
* fix test
* Attempt at simplification.
* Small fix.
* Fixing non existing AutoModel for TF.
* Naming.
* Remove extra condition.
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>
* Add optional grouped parsers description to HfArgumentParser (#12042)
* Adding optional argument group to HfArgumentParser
* Minor
* remove whitespace
* Minor styling
* adds metric prefix. (#12057)
* adds metric prefix.
* update tests to include prefix
* skip failing test (#12059)
* Fix integration tests (#12066)
* Fix tapas issue (#12063)
* Fix scatter function to be compatible with torch-scatter 2.7.0
* Allow test again
* updated the original RAG implementation to be compatible with latest Pytorch-Lightning (#11806)
* updated the original RAG implementation to be compatible with the latest PL version
* updated the requirements.txt file
* execute make style
* code quality test
* code quality
* conflix resolved in requirement.txt
* code quality
* changed the MyDDP class name to CustomDDP
* Replace legacy tensor.Tensor with torch.tensor/torch.empty (#12027)
* Replace legacy torch.Tensor constructor with torch.{tensor, empty}
* Remove torch.Tensor in examples
* Add torch to requirements.txt in language-modeling (#12040)
* Add torch to requirements.txt in language-modeling
* Update examples/pytorch/language-modeling/requirements.txt
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Properly indent block_size (#12070)
* [Deepspeed] various fixes (#12058)
* replace deprecated config
* sub_group_size was too big
* complete deprecation removal
* [Deepspeed Wav2vec2] integration (#11638)
* wip
* wip - but working with https://github.com/microsoft/DeepSpeed/pull/1044
* cleanup
* workaround
* working 5/8 modes
* solve fp32 distributed zero3
* style
* sync
* sync
* rework
* deprecation
* cleanup
* https://github.com/microsoft/DeepSpeed/pull/1044 pr was merged
* clean up
* add a guide
* more prose
* more prose
* fix
* more prose
* sub_group_size was too big
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* refactor
* bug fix
* make the true check explicit
* new deepspeed release
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* typo
* Update run_ner.py with id2label config (#12001)
* sync LayerDrop for Wav2Vec2Encoder + tests (#12076)
* Add DETR (#11653)
* Squash all commits of modeling_detr_v7 branch into one
* Improve docs
* Fix tests
* Style
* Improve docs some more and fix most tests
* Fix slow tests of ViT, DeiT and DETR
* Improve replacement of batch norm
* Restructure timm backbone forward
* Make DetrForSegmentation support any timm backbone
* Fix name of output
* Address most comments by @LysandreJik
* Give better names for variables
* Conditional imports + timm in setup.py
* Address additional comments by @sgugger
* Make style, add require_timm and require_vision to testsé
* Remove train_backbone attribute of DetrConfig, add methods to freeze/unfreeze backbone
* Add png files to fixtures
* Fix type hint
* Add timm to workflows
* Add `BatchNorm2d` to the weight initialization
* Fix retain_grad test
* Replace model checkpoints by Facebook namespace
* Fix name of checkpoint in test
* Add user-friendly message when scipy is not available
* Address most comments by @patrickvonplaten
* Remove return_intermediate_layers attribute of DetrConfig and simplify Joiner
* Better initialization
* Scipy is necessary to get sklearn metrics
* Rename TimmBackbone to DetrTimmConvEncoder and rename DetrJoiner to DetrConvModel
* Make style
* Improve docs and add 2 community notebooks
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
* [test] support more than 2 gpus (#12074)
* support more than 2 gpus
* style
* Wav2Vec2 Pretraining (#11306)
* Working quantizer forward
* Working quantizer forward
* Clean up unused model parts, test reproducibility
* Working quantizer forward
* Clean up unused model parts, test reproducibility
* Remove custom outputs from the shared ones
* correct conversion
* correct bug
* add first pretrain script
* save intermediate
* static shapes
* save intermediate
* finish first pretrain script version
* more refactor
* remove wanddb
* refactor more
* improve test
* correct perplexity compute bug
* finish model implementation
* add to docs
* finish docs
* finish pretraining script
* finish pretraining script
* remove wandb
* finish PR for merge
* finish config
* finish
* make deepspeed work
* Apply suggestions from code review
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* apply suggestions
* fix flaky test
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* pass decay_mask fn to optimizer (#12087)
* rm require_version_examples (#12088)
* [Wav2Vec2ForPretraining] Correct checkpoints wav2vec2 & fix tests (#12089)
* fix_torch_device_generate_test
* remove @
* fix tests
* Add text_column_name and label_column_name to run_ner and run_ner_no_trainer args (#12083)
* Add text_column_name and label_column_name to run_ner args
* Minor fix: grouping for text and label column name
* CLIPFeatureExtractor should resize images with kept aspect ratio (#11994)
* Resize with kept aspect ratio
* Fixed failed test
* Overload center_crop and resize methods instead
* resize should handle non-PIL images
* update slow test
* Tensor => tensor
Co-authored-by: patil-suraj <surajp815@gmail.com>
* New TF GLUE example (#12028)
* Pushing partially-complete new GLUE example
* First draft of the new TF GLUE example! Needs a little more testing to be sure but it's almost ready.
* Fix to the fit() call
* Bugfixes, making sure TPU and multi-GPU support is ready
* Remove logger line that depends on Pytorch
* Style pass
* Deleting old TF GLUE example
* Include label2id and id2label in the saved model config
* Don't clobber the existing model.config.label2id
* Style fixes
* Update examples/tensorflow/text-classification/run_glue.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Fix quality
* Update README.md to cover the TF GLUE example.
* Minor style edits
* Appending label2id and id2label to models to ensure inference works properly (#12102)
* Fix a condition in test_generate_with_head_masking (#11911)
* Fix a condition in test_generate_with_head_masking
* Fix usage of head_mask in bigbirg_pegasus
* Fix head masking for speech2text
* Resolve copy mismatch + drop unwanted print statement
* Fix the condition
* Flax VisionTransformer (#11951)
* adding vit for flax
* added test for Flax-vit and some bug-fixes
* overrided methods where variable changes were necessary for flax_vit test
* added FlaxViTForImageClassification for test
* Update src/transformers/models/vit/modeling_flax_vit.py
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* made changes suggested in PR
* Adding jax-vit models for autoimport
* swapping num_channels and height,width dimension
* fixing the docstring for torch-like inputs for VIT
* add model to main init
* add docs
* doc, fix-copies
* docstrings
* small test fixes
* fix docs
* fix docstr
* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* style
Co-authored-by: jayendra <jayendra@infocusp.in>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* add relevant description to tqdm in examples (#11927)
* add relevant `desc` in examples
* require_version datasets>=1.8.0
* Fix head masking generate tests (#12110)
* fix_torch_device_generate_test
* remove @
* fix tests
* Flax CLM script (#12023)
* first draft
* max_seq_length => block_size
* fix arg names
* fix typos
* fix loss calculation
* add max examples, fix train eval steps, metrics
* optimizer mask
* fix perpelexity, metric logging
* fix logging
* data_collator = > data_loader
* refactor loss_fn
* support single GPU
* pass distributed to write_metric
* fix jitting
* fix single device training
* fix single device metrics
* close inner progress bars once finished
* add overwrite_cache arg
* ifx dataset caching issue
* add more logs
* few small fixes,
* address nicholas suggestions
* fix docstr
* address patricks suggestions
* make flake happy
* pass new new_dropout_rng to apply_gradients
* reset train metrics after every epoc
* remove distributed logis, small fixes
* Add from_pretrained to dummy timm objects (#12097)
* Add from_pretrained to dummy timm
* Fix at the source
* Update utils/check_dummies.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Missing pretrained dummies
* Style
Co-authored-by: Sylvain Gugger <sylvain.gugger@gmail.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Fix t5 error message (#12136)
* Fix t5 error message
* Fix again
* Fix megatron_gpt2 attention block's causal mask (#12007)
* Fix megatron_gpt2 attention block's causal mask.
* compatibility with checkpoints created with recent versions of Megatron-LM
* added integration test for the released Megatron-GPT2 model
* code style changes
* added option to megatron conversion script to read from config file
Co-authored-by: Guido Novati <gnovati@nvidia.com>
* Add mlm pretraining xla torch readme (#12011)
* fix_torch_device_generate_test
* remove @
* upload
* Apply suggestions from code review
* Apply suggestions from code review
* Apply suggestions from code review
* Update examples/flax/language-modeling/README.md
* add more info
* finish
* fix
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
* add readme for flax clm (#12111)
* add readme for flax clm
* use section link for tokenizer
* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* update metrics
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* FlaxBart (#11537)
* Start working on FlaxBart
* Create modeling_flax_bart.py
* Write FlaxBartAttention
* Add FlaxBartEncoderLayer
* Add FlaxBartDecoderLayer and some typing
* Add helepr function for FlaxBart
* shift_tokens_right
* _make_causal_mask
* _expand_mask
* Add PositionalEmbedding and fix init_std naming
* Add FlaxBartPretrainedModel
* Add FlaxBartEncoder
* Add FlaxBartEncoder
* Add FlaxBartEncoder among modules to be imported
* YET WE CANNOT INITIALIZE THAT!! :(
* Make BartEncoder working
Change BartEncoder to instance of nn.Module so far
* Add FlaxBartDecoder
* Add FlaxBartModel
* TODO to make model run -> Prepapre model inputs
* Resolve padding
* Add FlaxBartModel
* Add FlaxBartModel into importable modules
* Remove FlaxBartEncoder and FlaxBartDecoder from importable modules
* make style; not properly working
* make style; make quality not pass due to some import I left
* Remove TODO for padding_idx in nn.Embed so far
* Add FlaxBartForConditionalGeneration
* Incorporate Flax model output classes, i.e. return_dict
* Add another models and incorporate use_cache arg
* Add FlaxBartForSequenceClassification and FlaxBartForQuestionAnswering
* Incorporate use_cache arg from PyTorch implementation
* Add all necessary Flax output utils
* Add FlaxBartForCausalLM; not working yet'
* Add minor improvements; still lacks some functionality
* Update docs, src and tests
* Add support of FlaxBart to docs/source
* Fix some bugs in FlaxBart souce code
* Add some neccessary tests for FlaxBart models - jit_compilation not passing
* Fix tests and add test_head_masking
* Fix tests for @jax.jit computation
* Add test_head_masking
* Migrate FlaxBart tests from jax.numpy to numpy
* Remove FlaxBartForCausalLM
* Clean repo
* fix bart model weight structure
* Fix FlaxBartForSequenceClassification
Slicing is not possible to use below jit, therefore, selecting sentence
representation from hidden_states must be changed.
* Allow FlaxBartForSequenceClassification for testing pt_flax equivalence
* Allow testing for FlaxBartForQA for pt_flax equivalence
* Add a comment to FlaxBartForSequenceClassification + change noise from 1e-3 to 1e-6
* remove past_key_values
* remove inputs_mebeds and make input_ids required
* add position ids
* re-write attention layer
* fix dataclass
* fix pos embeds and attention output
* fix pos embeds
* expose encode method
* expose decode method
* move docstring to top
* add cache for causal attn layer
* remove head masking for now
* s2s greedy search first pass
* boom boom
* fix typos
* fix greedy generate for bart
* use encoder, decoder layers instead of num_hidden_layers
* handle encoder_outputs
* cleanup
* simplify decoding
* more clean-up
* typos
* Change header + add {decoder_,}position_ids into 2 models
* add BartConfig
* fix existing tests
* add encode, decode methods
* Fix shift_tokens_right for JIT compilation + clarify one condition
* fix decode
* encoder => encode
* simplify generate
* add tests for encode and decode
* style
* add tests for cache
* fix equivalence tests
* sample generate now works with seq2seq
* generation tests
* initialize dense layers
* docstring and cleanup
* quality
* remove get/set input_embeddings
* address Patricks suggestions
* decode for every model, remove encoder_outputs from call
* update tests accordingly
* decode returns only decoder outputs and logits
* fix arguments
* doc encode, decode methods
* correct base_model_prefix
* fix test for seq classif model
* fix docs
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* Feature to use the PreTrainedTokenizerFast class as a stand-alone tokenizer (#11810)
* feature for tokenizer without slow/legacy version
* format
* modify common test
* add tests
* add PreTrainedTokenizerFast to AutoTokenizer
* format
* change tokenizer common test in order to be able to run test without a slow version
* update tokenizer fast test in order to use `rust_tokenizer_class` attribute instead of `tokenizer_class`
* add autokenizer test
* replace `if self.tokenizer_class is not None` with ` if self.tokenizer_class is None`
* remove obsolete change in comment
* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Update src/transformers/tokenization_utils_fast.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* change `get_main_tokenizer` into `get_tokenizers`
* clarify `get_tokenizers` method
* homogenize with `test_slow_tokenizer` and `test_rust_tokenizer`
* add `test_rust_tokenizer = False` to tokenizer which don't define a fast version
* `test_rust_tokenizer = False` for BertJapaneseTokenizer
* `test_rust_tokenizer = False` for BertJapaneseCharacterTokenizationTest
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* [Flax] Add links to google colabs (#12146)
* fix_torch_device_generate_test
* remove @
* add colab links
* Don't log anything before logging is setup in examples (#12121)
* Don't log anything before logging is setup in examples
* Last example
* Use text_column_name variable instead of "text" (#12132)
* Use text_column_name variable instead of "text"
`text_column_name` was already defined above where I made the changes and it was also used below where I made changes.
This is a very minor change. If a dataset does not use "text" as the column name, then the `tokenize_function` will now use whatever column is assigned to `text_column_name`. `text_column_name` is just the first column name if "text" is not a column name. It makes the function a little more robust, though I would assume that 90% + of datasets use "text" anyway.
* black formatting
* make style
Co-authored-by: Nicholas Broad <nicholas@nmbroad.com>
* [lm examples] Replicate --config_overrides addition to other LM examples (#12135)
* [lm examples] Replicate --config_overrides addition to other LM examples
* Removing no trainer files changes
* Update README
Co-authored-by: Kumar Abhishek <kabhishek@expedia.com>
* fix error message (#12148)
* [optim] implement AdafactorSchedule (#12123)
* implement AdafactorSchedule
* typo
* fix
* Update src/transformers/optimization.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* [style] consistent nn. and nn.functional (#12124)
* consistent nn. and nn.functional
* fix glitch
* fix glitch #2
* Adding TFWav2Vec2Model (#11617)
* [WIP] Add TFWav2Vec2Model
Work in progress for adding a tensorflow version of Wav2Vec2
* feedback changes
* small fix
* Test Feedback Round 1
* Add SpecAugment and CTC Loss
* correct spec augment mask creation
* docstring and correct copyright
* correct bugs
* remove bogus file
* finish tests correction
* del unnecessary layers
* Update src/transformers/models/wav2vec2/modeling_tf_wav2vec2.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* make style
* correct final bug
* Feedback Changes
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* [Flax] Fix flax pt equivalence tests (#12154)
* fix_torch_device_generate_test
* remove @
* upload
* consistent nn. and nn.functional: p2 templates (#12153)
* Flax Big Bird (#11967)
* add flax bert
* bert -> bigbird
* original_full ported
* add debugger
* init block sparse
* fix copies ; gelu_fast -> gelu_new
* block sparse port
* fix block sparse
* block sparse working
* all ckpts working
* fix-copies
* make quality
* init tests
* temporary fix for FlaxBigBirdForMultipleChoice
* skip test_attention_outputs
* fix
* gelu_fast -> gelu_new ; fix multiple choice model
* remove nsp
* fix sequence classifier
* fix
* make quality
* make fix-copies
* finish
* Delete debugger.ipynb
* Update src/transformers/models/big_bird/modeling_flax_big_bird.py
* make style
* finish
* bye bye jit flax tests
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* [style] consistent nn. and nn.functional: part 3 `tests` (#12155)
* consistent nn. and nn.functional: p3 templates
* restore
* [style] consistent nn. and nn.functional: part 4 `examples` (#12156)
* consistent nn. and nn.functional: p4 examples
* restore
* consistent nn. and nn.functional: part 5 docs (#12161)
* Add video links to the documentation (#12162)
* [Flax generate] Add params to generate (#12171)
* fix_torch_device_generate_test
* remove @
* add params as input
* finish
* Use a released version of optax rather than installing from Git. (#12173)
Use a released version of optax rather than installing from Git
* Have dummy processors have a `from_pretrained` method (#12145)
* Add course banner (#12157)
* Add course banner
* Update course banner
* Adjust banner width
* Enable add_prefix_space if model_type is roberta or gpt2 (#12116)
* Update AutoModel classes in summarization example (#12178)
- Convert use of deprecated AutoModelWithLMHead to AutoModelForSeq2SeqLM
- Add newly required `truncation=True` to `tokenizer.encode` with `max_length`
This silences all warnings.
* Ray Tune Integration Updates (#12134)
* fix
* fixes
* add back to scheduled tests
* formatting
* Update integrations.py
* [testing] ensure concurrent pytest workers use a unique port for torch.dist (#12166)
* ensure concurrent pytest workers use a unique port for torch.distributed.launch
* reword
* Model card defaults (#12122)
* [WIP] Model card defaults
* finetuned_from default value
* Add all mappings to the mapping file
* Be more defensive on finetuned_from arg
* Add default task tag
* Separate tags from tasks
* Edge case for dataset
* Apply suggestions from code review
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Temporarily deactivate torch-scatter while we wait for new release (#12181)
* Temporarily deactivate torch-scatter while we wait for new release
* torch-1.8.1 binary for scatter
* Revert to 1.8.0
* Pin torch dependency
* torchaudio and torchvision
* Temporarily deactivate torchhub test (#12184)
* [Flax] Add Beam Search (#12131)
* fix_torch_device_generate_test
* remove @
* push new logit processors
* add processors
* save first working version
* save intermediate
* finish
* make style
* make fix-copies
* finish
* Update tests/test_modeling_flax_bart.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Suraj Patil <surajp815@gmail.com>
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* Hubert (#11889)
* fix_torch_device_generate_test
* remove @
* add hubert
* add first test file
* more docs
* fix bugs
* fix bug
* finish
* finish
* finish docstring
* fix
* fix
* finalize
* add to ignored
* finish
* Apply suggestions from code review
* correct naming
* finish
* fix auto config
* finish
* correct convert script
* Apply suggestions from code review
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* apply suggestions lysandre & suraj
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* updated DLC images and sample notebooks (#12191)
* Enabling AutoTokenizer for HubertConfig. (#12198)
* Use yaml to create metadata (#12185)
* Use yaml to create metadata
* Fix typo
* Remove pin
* [Docs] fixed broken link (#12205)
* fixed broken link
* Update docs/source/benchmarks.rst
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update docs/source/benchmarks.rst
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Pipeline update & tests (#12207)
* Improve detr (#12147)
* Remove unused variables
* Improve docs
* Fix docs of segmentation masks
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Add link to the course (#12229)
* Support for torch 1.9.0 (#12224)
* Support for torch 1.9.0
* Torch scatter for 1.9.0
* Github Actions run on 1.9.0
* fix pt-1.9.0 `add_` deprecation (#12217)
* fix pt-1.9.0 add_ deprecation
* add () for clarity
* Trigger CI
* require_version(torch
* Release: v4.7.0
* Docs for v4.8.0
* AutoTokenizer: infer the class from the tokenizer config if possible (#12208)
* AutoTokenizer: infer the class from the tokenizer config if possible
* Add tests
* Update src/transformers/models/auto/tokenization_auto.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* update desc for map in all examples (#12226)
* update desc for map in all examples
* added plm
* suggestions
* [Flax] FlaxAutoModelForSeq2SeqLM (#12228)
* add FlaxAutoModelForSeq2SeqLM
* [FlaxBart] few small fixes (#12247)
* boom boom
* remove flax clip example
* few small fixes
* Depreciate pythonic Mish and support PyTorch 1.9 version of Mish (#12240)
* Moved Mish to Torch 1.9 version
* Run black formatting
* [t5 doc] make the example work out of the box (#12239)
* [run_clm.py] restore caching
* style
* [t5 doc] make the example work out of the box
This PR expands the training example to include the correct model type for the example to work, e.g. with `T5Model` this example will break.
* Update docs/source/model_doc/t5.rst
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* expand the other example
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* Fix the scheduled CI
* Better CI feedback (#12279)
* Better run ID
* Only part of CI
* Revert "Only part of CI"
This reverts commit 29f7f248d2.
* Fix for making student ProphetNet for Seq2Seq Distillation (#12130)
* make_student.py: fix to make student ProphetNet
* reformat
* [FlaxClip] fix test from/save pretrained test (#12284)
* boom boom
* remove flax clip example
* fix from_save_pretrained
* [Flax] [WIP] allow loading head model with base model weights (#12255)
* boom boom
* remove flax clip example
* allow loading head model with base model weights
* add test
* fix imports
* disable save, load test for clip
* add test_save_load_to_base
* [DeepSpeed] don't ignore --adafactor (#12257)
* [Flax] Fix flax test save pretrained (#12256)
* fix_torch_device_generate_test
* remove @
* fix flax save pretrained test
* Tensorflow QA example (#12252)
* New Tensorflow QA example!
* Style pass
* Updating README.md for the new example
* flake8 fixes
* Update examples/tensorflow/question-answering/README.md
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* [Flax] Add jax flax to env command (#12251)
* fix_torch_device_generate_test
* remove @
* add commands for flax/jax
* reset report_to to none, avoid deprecation warning (#12293)
* [trainer + examples] set log level from CLI (#12276)
* set log level from CLI
* add log_level_replica + test + extended docs
* cleanup
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* rename datasets objects to allow datasets module
* improve the doc
* style
* doc improve
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* [tests] multiple improvements (#12294)
* [tests] multiple improvements
* cleanup
* style
* todo to investigate
* fix
* Fix for the issue of device-id getting hardcoded for token_type_ids during Tracing [WIP] (#11252)
* registering a buffer for token_type_ids, to pass the error of device-id getting hardcoded when tracing
* sytle format
* adding persistent flag to the resgitered buffers that prevent from adding them to the state_dict and addresses the Backward compatibility issue
* adding the try catch to the fix as persistent flag is only available from PT >1.6
* adding version check
* added the condition to only use the token_type_ids buffer when its autogenerated not passed by user
* adding comments and making the conidtion where token_type_ids are None to use the registered buffer
* taking out position-embeddding from the if block
* adding comments
* handling the case if buffer for position_ids was not registered
* reverted the changes on position_ids, fix the issue with size of token_type_ids buffer, moved the modification for generated token_type_ids to Bertmodel, instead of Embeddings
* reverting the token_type_ids in case of None to the previous version
* reverting changes on position_ids adding back the if block
* changes added by running make fix-copies
* changes added by running make fix-copies and added the import version as it was getting used
* changes added by running make fix-copies
* changes added by running make fix-copies
* fixing the import format
* fixing the import format
* modified to use temp tensor for trimed and expanded token_type_ids buffer
* changes made by fix-copies after temp tensor modifications
* changes made by fix-copies after temp tensor modifications
* changes made by fix-copies after temp tensor modifications
* clean up
* clean up
* clean up
* clean up
* Nit
* Nit
* Nit
* modified according to support device conversion on traced models
* modified according to support device conversion on traced models
* modified according to support device conversion on traced models
* modified according to support device conversion on traced models
* changes based on latest in master
* Adapt templates
* Add version import
Co-authored-by: Ubuntu <ubuntu@ip-172-31-32-81.us-west-2.compute.internal>
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
* trainer_tf: adjust wandb installation command (#12291)
* add FlaxAutoModelForImageClassification in main init (#12298)
* Fix and improve documentation for LEDForConditionalGeneration (#12303)
* Replace conditional generation example (fixes#12268)
* Replace model in summarization example with finetuned checkpoint, adapt example text
* Fix typo in new summarization example
* Fix docstring formatting, add missing import statement to example
* [Flax] Main doc for event orga (#12305)
* fix_torch_device_generate_test
* remove @
* push
* finish
* some typos
* add more info on communication
* add suggestions
* [trainer] 2 bug fixes and a rename (#12309)
* bug fixes and a rename
* add extended DDP test
* FlaxBartPretrainedModel -> FlaxBartPreTrainedModel (#12313)
* [docs] performance (#12258)
* initial performance document
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* rewrites based on suggestions
* 8x multiple is for AMP only
* add contribute section
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Add CodeCarbon Integration (#12304)
* Add optional dependency
* Add CodeCarbon integration
* Add CodeCarbon integration
* Add CodeCarbon integration
* typo
* Optimizing away the `fill-mask` pipeline. (#12113)
* Optimizing away the `fill-mask` pipeline.
- Don't send anything to the tokenizer unless needed. Vocab check is
much faster
- Keep BC by sending data to the tokenizer when needed. User handling warning messages will see performance benefits again
- Make `targets` and `top_k` work together better `top_k` cannot be
higher than `len(targets)` but can be smaller still.
- Actually simplify the `target_ids` in case of duplicate (it can happen
because we're parsing raw strings)
- Removed useless code to fail on empty strings. It works only if empty
string is in first position, moved to ignoring them instead.
- Changed the related tests as only the tests would fail correctly
(having incorrect value in first position)
* Make tests compatible for 2 different vocabs... (at the price of a
warning).
Co-authored-by: @EtaoinWu
* ValueError working globally
* Update src/transformers/pipelines/fill_mask.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* `tokenizer.vocab` -> `tokenizer.get_vocab()` for more compatiblity +
fallback.
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Add output in a dictionary for TF `generate` method (#12139)
* Add output args to greedy search
* Fix critical typo + make style quality
* Handle generate_beam_search
* Add dict_specific tests and fix the placement of encoder outputs
* Add specific outputs
* Update doc
* Fix typo
* Adjust handling encoder_outputs + Fix generating for T5
* Fix generate for RAG
* Fix handling ouptut_attentions when target_mapping is not None
Take care of situations when target_mapping is provided
as there are 2-tuple of attentions
Change from:
if inputs["output_attentions"]:
attentions = tuple(tf.transpose(t, perm(2, 3, 0, 1)) for t in attentions)
to:
if inputs["output_attentions"]:
if inputs["target_mapping"] is not None:
# when target_mapping is provided, there are 2-tuple of attentions
attentions = tuple(
tuple(tf.transpose(attn_stream, perm=(2, 3, 0, 1)) for attn_stream in t) for t in attentions
)
else:
attentions = tuple(tf.transpose(t, perm=(2, 3, 0, 1)) for t in attentions)
* Rename kwargs to model_kwargs
* make style quality
* Move imports in test_modeling_tf_common.py
Move ModelOutput-related imports in test_modeling_tf_common.py
into the `is_tf_available():` statement.
* Rewrite nested if-statements
* Fix added tests
* Flax summarization script (#12230)
* add summrization script
* fix arguments, preprocessing, metrics
* add generation and metrics
* auto model, prediction loop
* prettify
* label smoothing
* adress Sylvain and Patricks suggestions
* dynamically import shift_tokens_right
* fix shift_tokens_right_fn call
* Rewrite ProphetNet to adapt converting ONNX friendly (#11981)
* Rewrite
* [ONNX] rewrite
* Flax T5 (#12150)
* copy pytorch-t5
* init
* boom boom
* forward pass same
* make generation work
* add more tests
* make test work
* finish normal tests
* make fix-copies
* finish quality
* correct slow example
* correct slow test
* version table
* upload models
* Update tests/test_modeling_flax_t5.py
* correct incorrectly deleted line
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
* Add mention of the huggingface_hub methods for offline mode (#12320)
* [Flax/JAX] Add how to propose projects markdown (#12311)
* fix_torch_device_generate_test
* remove @
* finish
* make style
* [TFWav2Vec2] Fix docs (#12283)
* fix error
* make style check happy
Co-authored-by: chenhaitao <chenhaitao@qiyi.com>
* Clean push to hub API (#12187)
* Clean push to hub API
* Create working dir if it does not exist
* Different tweak
* New API + all models + test Flax
* Adds the Trainer clean up
* Update src/transformers/file_utils.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Address review comments
* (nit) output types
* No need to set clone_from when folder exists
* Update src/transformers/trainer.py
Co-authored-by: Julien Chaumond <julien@huggingface.co>
* Add generated_from_trainer tag
* Update to new version
* Fixes
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Julien Chaumond <julien@huggingface.co>
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
* Add all XxxPreTrainedModel to the main init (#12314)
* Add all XxxPreTrainedModel to the main init
* Add to template
* Add to template bis
* Add FlaxT5
* Conda build (#12323)
* Temporarily revert the `fill-mask` improvements.
* changed modeling_fx_utils.py to utils/fx.py for clarity (#12326)
Co-authored-by: Michael Benayoun <michael@huggingface.co>
* Pin good version of huggingface_hub
* [Flax T5] Fix weight initialization and fix docs (#12327)
* finish t5 flax fixes
* improve naming
* Release: v4.8.0
* v4.9.0.dev0
* Update training_args.py (#12328)
mention in `save_strategy` param description that `load_best_model_at_end` can override
* [Deepspeed] new docs (#12077)
* document sub_group_size
* style
* install + issues reporting
* style
* style
* Update docs/source/main_classes/deepspeed.rst
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* indent 4
* restore
* style
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Fix default to logging_dir lost in merge conflict
* try-this (#12338)
Signed-off-by: Richard Liaw <rliaw@berkeley.edu>
* [examples/Flax] move the examples table up (#12341)
* Fix torchscript tests (#12336)
* Fix torchscript tests
* Better test
* Remove bogus print
* Document patch release v4.8.1
* Add flax/jax quickstart (#12342)
* Update README.md
* fixed typo (#12356)
* Fix exception in prediction loop occurring for certain batch sizes (#12350)
* fix distributed_concat for scalar outputs
* Update README.md
* fixed typo (#12356)
* simplify fix with terser syntax
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Trigger CI
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: michal pitr <21157924+MichalPitr@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Add FlaxBigBird QuestionAnswering script (#12233)
* port bigbird script
* adapt script a bit
* change location
* adapt more
* save progress
* init commit
* style
* dataset script tested
* readme add
* Replace NotebookProgressReporter by ProgressReporter in Ray Tune run (#12357)
* Replace NotebookProgressReporter by ProgressReporter in Ray Tune run
* Move to local import
* Style
* remove extra white space from log format (#12360)
* fixed multiplechoice tokenization (#12362)
* fixed multiplechoice tokenization
The model would have seen two sequences:
1. [CLS]prompt[SEP]prompt[SEP]
2. [CLS]choice0[SEP]choice1[SEP]
that is not correct as we want a contextualized embedding of prompt and choice
* removed outer brackets for proper sequence generation
* [trainer] add main_process_first context manager (#12351)
* main_process_first context manager
* handle multi-node, add context description
* sync desc
* [Examples] Replicates the new --log_level feature to all trainer-based pytorch (#12359)
* added log_level
* fix comment
* fixed log_level
* Trigger CI
* Unfied logging
* simplified args for log_level
* updated example template (#12365)
* replace print with logger (#12368)
* [Documentation] Warn that DataCollatorForWholeWordMask is limited to BertTokenizer-like tokenizers (#12371)
* Notify users that DataCollatorForWholeWordMask is limited to BertTokenier-like tokenizers
* Fix code formatting
* Update run_mlm.py (#12344)
Before the code could not be used for validation only because of this line:
extension = data_args.train_file.split(".")[-1]
was assuming that extension must be extracted from the training dataset. This line would run regardless of the training or validation options of the user. This would lead to an error if the user only wants to run an evaluation only and does not want to do train (because the training file does not exist). I modified it to extract extension from the training file if the user wants to do train and extract it from the validation file if the user wants to run eval. This way the code can be used for both training and validation separately.
* Add possibility to maintain full copies of files (#12312)
* [CI] add dependency table sync verification (#12364)
* add dependency table sync verification
* improve the message
* improve the message
* revert
* ready to merge
* [Examples] Added context manager to datasets map (#12367)
* added cotext manager to datasets map
* fixed style and spaces
* fixed warning of deprecation
* changed desc
* [Flax community event] Add more description to readme (#12398)
* fix_torch_device_generate_test
* remove @
* boom boom
* correct typos
* Apply suggestions from code review
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* Apply suggestions from code review
Co-authored-by: Suzana Ilić <io.suzanai@gmail.com>
* Apply suggestions from code review
Co-authored-by: Suraj Patil <surajp815@gmail.com>
Co-authored-by: Suzana Ilić <io.suzanai@gmail.com>
* Update README.md
* Fix copies
* Remove the need for `einsum` in Albert's attention computation (#12394)
* debug albert einsum
* Fix matmul computation
* Let's use torch linear layer.
* Style.
* [Flax] Adapt flax examples to include `push_to_hub` (#12391)
* fix_torch_device_generate_test
* remove @
* finish
* correct summary writer
* correct push to hub
* fix indent
* finish
* finish
* finish
* finish
* finish
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
* Tensorflow LM examples (#12358)
* Tensorflow MLM example
* Add CLM example
* Style fixes, adding missing checkpoint code from the CLM example
* Fix TPU training, avoid massive dataset warnings
* Fix incorrect training length calculation for multi-GPU training
* Fix incorrect training length calculation for multi-GPU training
* Refactors and nitpicks from the review
* Style pass
* Adding README
* pass the matching trainer log level to deepspeed (#12401)
* [Flax] Add T5 pretraining script (#12355)
* fix_torch_device_generate_test
* remove @
* add length computatan
* finish masking
* finish
* upload
* fix some bugs
* finish
* fix dependency table
* correct tensorboard
* Apply suggestions from code review
* correct processing
* slight change init
* correct some more mistakes
* apply suggestions
* improve readme
* fix indent
* Apply suggestions from code review
Co-authored-by: SaulLu <55560583+SaulLu@users.noreply.github.com>
* correct tokenizer
* finish
* finish
* finish
* finish
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
Co-authored-by: SaulLu <55560583+SaulLu@users.noreply.github.com>
* [models] respect dtype of the model when instantiating it (#12316)
* [models] respect dtype of the model when instantiating it
* cleanup
* cleanup
* rework to handle non-float dtype
* fix
* switch to fp32 tiny model
* improve
* use dtype.is_floating_point
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* fix the doc
* recode to use explicit torch_dtype_auto_detect, torch_dtype args
* docs and tweaks
* docs and tweaks
* docs and tweaks
* merge 2 args, add docs
* fix
* fix
* better doc
* better doc
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Rename detr targets to labels (#12280)
* Rename target to labels in DetrFeatureExtractor
* Update DetrFeatureExtractor tests accordingly
* Improve docs of DetrFeatureExtractor
* Improve docs
* Make style
* Add out of vocabulary error to ASR models (#12288)
* Add OOV error to ASR models
* Feedback changes
* Fix TFWav2Vec2 SpecAugment (#12289)
* Fix TFWav2Vec2 SpecAugment
* Invert masks
* Feedback changes
* [example/flax] add summarization readme (#12393)
* add readme
* update readme and add requirements
* Update examples/flax/summarization/README.md
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* [Flax] Example scripts - correct weight decay (#12409)
* fix_torch_device_generate_test
* remove @
* finish
* finish
* correct style
* fix ids_to_tokens naming error in tokenizer of deberta v2 (#12412)
Co-authored-by: Jipeng Huang <jihuan@microsoft.com>
* minor fixes in original RAG training (#12395)
* Added talks (#12415)
* Easily train a new fast tokenizer from a given one (#12361)
* [WIP] Easily train a new fast tokenizer from a given one
* Fix test
* Roll out to other tokenizers and add tests
* Fix bug with unk id and add emoji to test
* Really use something different in test
* Implement special tokens map
* Map special tokens in the Transformers tokenizers
* Fix test
* Make test more robust
* Fix test for BPE
* More robust map and test
Co-authored-by SaulLu
* Test file
* Stronger tests
Co-authored-by: SaulLu <lucilesaul.com@gmail.com>
* Map unk token for Wordpiece and address review comment
* Fix lowercase test and address review comment
* Fix all tests
* Simplify test
* Fix tests for realsies
* Easily train a new fast tokenizer from a given one - tackle the special tokens format (str or AddedToken) (#12420)
* Propose change in tests regarding lower case
* add new test for special tokens types
* put back the test part about decoding
* add feature: the AddedToken is re-build with the different mapped content
* Address review comment: simplify AddedToken building
Co-authored-by: sgugger <sylvain.gugger@gmail.com>
* Update src/transformers/tokenization_utils_fast.py
Co-authored-by: sgugger <sylvain.gugger@gmail.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: SaulLu <lucilesaul.com@gmail.com>
Co-authored-by: SaulLu <55560583+SaulLu@users.noreply.github.com>
* [modelcard] fix (#12422)
this PR is fixing an incorrect attribute - probably some tests are needed?
* Add option to save on each training node (#12421)
* Add option to save on each training node
* Apply suggestions from code review
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>
* Address review comments
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>
* Added to talks section (#12433)
Added one more confirmed speaker, zoom links and gcal event links
* Fix default bool in argparser (#12424)
* Fix default bool in argparser
* Add more to test
* Add default bos_token and eos_token for tokenizer of deberta_v2 (#12429)
* fix ids_to_tokens naming error in tokenizer of deberta v2
* Update tokenization_deberta_v2.py
Add bos_token and eos_token.
* format code
Co-authored-by: Jipeng Huang <jihuan@microsoft.com>
* Add CANINE (#12024)
* First pass
* More progress
* Add support for local attention
* More improvements
* More improvements
* Conversion script working
* Add CanineTokenizer
* Make style & quality
* First draft of integration test
* Remove decoder test
* Improve tests
* Add documentation
* Mostly docs improvements
* Add CanineTokenizer tests
* Fix most tests on GPU, improve upsampling projection
* Address most comments by @dhgarrette
* Remove decoder logic
* Improve Canine tests, improve docs of CanineConfig
* All tokenizer tests passing
* Make fix-copies and fix tokenizer tests
* Fix test_model_outputs_equivalence test
* Apply suggestions from @sgugger's review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Address some more comments
* Add support for hidden_states and attentions of shallow encoders
* Define custom CanineModelOutputWithPooling, tests pass
* First pass
* More progress
* Add support for local attention
* More improvements
* More improvements
* Conversion script working
* Add CanineTokenizer
* Make style & quality
* First draft of integration test
* Remove decoder test
* Improve tests
* Add documentation
* Mostly docs improvements
* Add CanineTokenizer tests
* Fix most tests on GPU, improve upsampling projection
* Address most comments by @dhgarrette
* Remove decoder logic
* Improve Canine tests, improve docs of CanineConfig
* All tokenizer tests passing
* Make fix-copies and fix tokenizer tests
* Fix test_model_outputs_equivalence test
* Apply suggestions from @sgugger's review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Address some more comments
* Make conversion script work for Canine-c too
* Fix tokenizer tests
* Remove file
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Document patch release v4.8.2
* fix typo in mt5 configuration docstring (#12432)
* Add to talks section (#12442)
* [JAX/Flax readme] add philosophy doc (#12419)
* add philosophy doc
* fix typos
* update doc
* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* address Patricks suggestions
* add a training example and fix typos
* jit the training step
* jit train step
* fix example code
* typo
* 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>
* [Flax] Add wav2vec2 (#12271)
* fix_torch_device_generate_test
* remove @
* start flax wav2vec2
* save intermediate
* forward pass has correct shape
* add weight norm
* add files
* finish ctc
* make style
* finish gumbel quantizer
* correct docstrings
* correct some more files
* fix vit
* finish quality
* correct tests
* correct docstring
* correct tests
* start wav2vec2 pretraining script
* save intermediate
* start pretraining script
* finalize pretraining script
* finish
* finish
* small typo
* finish
* correct
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* make style
* push
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* Add missing Copied from statements
* Reference model uploaded under Google org
* Fix various duplicates from merging
* Rembert-large -> rembert, fix overeager Copied from, return type
* Incorporate PR comments from Patrick and Sylvain
Co-authored-by: ctheodoris <seanymphoceana@yahoo.com>
Co-authored-by: ctheodoris <cvtheodo@ds.dfci.harvard.edu>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>
Co-authored-by: Teven <teven.lescao@gmail.com>
Co-authored-by: Nick Lane-Smith <nlanesmith@gmail.com>
Co-authored-by: Shiro T <stsuchi@users.noreply.github.com>
Co-authored-by: Wang Ran (汪然) <wrran@outlook.com>
Co-authored-by: Ahmet Akkoç <themadprogramer@gmail.com>
Co-authored-by: francescorubbo <francescorubbo@users.noreply.github.com>
Co-authored-by: Daniel Stancl <46073029+stancld@users.noreply.github.com>
Co-authored-by: talkhaldi <tareq.alkhaldi@gmail.com>
Co-authored-by: joerenner <joepeterrenner@gmail.com>
Co-authored-by: jrenner <joseph.renner@inria.fr>
Co-authored-by: Avital Oliver <avitalo@google.com>
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
Co-authored-by: Josh Tanner <mindful.jt@gmail.com>
Co-authored-by: Nicolas Patry <patry.nicolas@protonmail.com>
Co-authored-by: Bhadresh Savani <bhadreshpsavani@gmail.com>
Co-authored-by: Jayendra <jayendra0parmar@gmail.com>
Co-authored-by: jayendra <jayendra@infocusp.in>
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
Co-authored-by: Philip May <philip@may.la>
Co-authored-by: Nicholas Vadivelu <nicholas.vadivelu@gmail.com>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
Co-authored-by: Shamane Siri <shamane@ahlab.org>
Co-authored-by: Quentin Lhoest <42851186+lhoestq@users.noreply.github.com>
Co-authored-by: Fan Zhang <zhangfan.tju@gmail.com>
Co-authored-by: Riccardo Bassani <48254418+BassaniRiccardo@users.noreply.github.com>
Co-authored-by: Volodymyr Byno <volodymyr.byno@gmail.com>
Co-authored-by: Jeoung-Minju <51041861+JminJ@users.noreply.github.com>
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>
Co-authored-by: Alberto Villa <a.villa.diez@gmail.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Gunjan Chhablani <chhablani.gunjan@gmail.com>
Co-authored-by: Kou Yong Kang <kou.yongkang@dhs.sg>
Co-authored-by: Shiva Pundir <36535845+ceevaaa@users.noreply.github.com>
Co-authored-by: François Lagunas <francois.lagunas@gmail.com>
Co-authored-by: Peter Izsak <232524+peteriz@users.noreply.github.com>
Co-authored-by: Russell Klopfer <russell@klopfer.us>
Co-authored-by: Mario Šaško <mariosasko777@gmail.com>
Co-authored-by: cdleong <4109253+cdleong@users.noreply.github.com>
Co-authored-by: Koichi Yasuoka <yasuoka@kanji.zinbun.kyoto-u.ac.jp>
Co-authored-by: Anton Lozhkov <aglozhkov@gmail.com>
Co-authored-by: kumapo <kumapo@users.noreply.github.com>
Co-authored-by: Tobias Norlund <tobias@norlund.se>
Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <sylvain.gugger@gmail.com>
Co-authored-by: Bhavitvya Malik <bhavitvya.malik@gmail.com>
Co-authored-by: Jonathan Chang <31893406+cccntu@users.noreply.github.com>
Co-authored-by: Guido Novati <16716298+novatig@users.noreply.github.com>
Co-authored-by: Guido Novati <gnovati@nvidia.com>
Co-authored-by: SaulLu <55560583+SaulLu@users.noreply.github.com>
Co-authored-by: Nicholas Broad <nbroad94@gmail.com>
Co-authored-by: Nicholas Broad <nicholas@nmbroad.com>
Co-authored-by: Kumar Abhishek <kr.abhish@gmail.com>
Co-authored-by: Kumar Abhishek <kabhishek@expedia.com>
Co-authored-by: Will Rice <will@spokestack.io>
Co-authored-by: Vasudev Gupta <7vasudevgupta@gmail.com>
Co-authored-by: Kilian Kluge <32523967+ionicsolutions@users.noreply.github.com>
Co-authored-by: Amog Kamsetty <amogkam@users.noreply.github.com>
Co-authored-by: Philipp Schmid <32632186+philschmid@users.noreply.github.com>
Co-authored-by: Xa9aX ツ <mishradiganta91@gmail.com>
Co-authored-by: Vishal Burman <vishal.a.burman23@gmail.com>
Co-authored-by: Hamid Shojanazeri <hamid.nazeri2010@gmail.com>
Co-authored-by: Ubuntu <ubuntu@ip-172-31-32-81.us-west-2.compute.internal>
Co-authored-by: Stefan Schweter <stefan@schweter.it>
Co-authored-by: Kevin Canwen Xu <canwenxu@126.com>
Co-authored-by: David Fan <30608893+jiafatom@users.noreply.github.com>
Co-authored-by: chenht2010 <chenht2010@yahoo.com>
Co-authored-by: chenhaitao <chenhaitao@qiyi.com>
Co-authored-by: Julien Chaumond <julien@huggingface.co>
Co-authored-by: Michael Benayoun <mickbenayoun@gmail.com>
Co-authored-by: Michael Benayoun <michael@huggingface.co>
Co-authored-by: Sam Havens <47401552+sam-qordoba@users.noreply.github.com>
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
Co-authored-by: Marc van Zee <marcvanzee@gmail.com>
Co-authored-by: michal pitr <21157924+MichalPitr@users.noreply.github.com>
Co-authored-by: jglaser <glaserj@ornl.gov>
Co-authored-by: Kai Fricke <krfricke@users.noreply.github.com>
Co-authored-by: cronoik <johannes.schaffrath@mail.de>
Co-authored-by: Taha ValizadehAslani <47432410+TahaAslani@users.noreply.github.com>
Co-authored-by: Suzana Ilić <io.suzanai@gmail.com>
Co-authored-by: Funtowicz Morgan <mfuntowicz@users.noreply.github.com>
Co-authored-by: Will Rice <wrice20@gmail.com>
Co-authored-by: Jabin Huang <huangjipengnju@gmail.com>
Co-authored-by: Jipeng Huang <jihuan@microsoft.com>
Co-authored-by: SaulLu <lucilesaul.com@gmail.com>
Co-authored-by: fcakyon <34196005+fcakyon@users.noreply.github.com>
* fix_torch_device_generate_test
* remove @
* correct greedy search
* save intertmed
* add final logits bias
* correct
* up
* add more tests
* fix another bug
* finish tests
* finish marian tests
* up
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
* Wrong model is used, should be character instead of subword
In the original Google repo for CANINE there was mixup in the model names in the README.md, which was fixed 2 weeks ago. Since this transformer model was created before, it probably resulted in wrong use in this example.
s = subword, c = character
* canine.rst style fix
* Update docs/source/model_doc/canine.rst
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Styling canine.rst
* Added links to model cards.
* Fixed links to model cards.
Co-authored-by: Jeroen Steggink <978411+jsteggink@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Laying down building stone for more flexible ONNX export capabilities
* Ability to provide a map of config key to override before exporting.
* Makes it possible to export BART with/without past keys.
* Supports simple mathematical syntax for OnnxVariable.repeated
* Effectively apply value override from onnx config for model
* Supports export with additional features such as with-past for seq2seq
* Store the output path directly in the args for uniform usage across.
* Make BART_ONNX_CONFIG_* constants and fix imports.
* Support BERT model.
* Use tokenizer for more flexibility in defining the inputs of a model.
* Add TODO as remainder to provide the batch/sequence_length as CLI args
* Enable optimizations to be done on the model.
* Enable GPT2 + past
* Improve model validation with outputs containing nested structures
* Enable Roberta
* Enable Albert
* Albert requires opset >= 12
* BERT-like models requires opset >= 12
* Remove double printing.
* Enable XLM-Roberta
* Enable DistilBERT
* Disable optimization by default
* Fix missing setattr when applying optimizer_features
* Add value field to OnnxVariable to define constant input (not from tokenizers)
* Add T5 support.
* Simplify model type retrieval
* Example exporting token_classification pipeline for DistilBERT.
* Refactoring to package `transformers.onnx`
* Solve circular dependency & __main__
* Remove unnecessary imports in `__init__`
* Licences
* Use @Narsil's suggestion to forward the model's configuration to the ONNXConfig to avoid interpolation.
* Onnx export v2 fixes (#12388)
* Tiny fixes
Remove `convert_pytorch` from onnxruntime-less runtimes
Correct reference to model
* Style
* Fix Copied from
* LongFormer ONNX config.
* Removed optimizations
* Remvoe bad merge relicas.
* Remove unused constants.
* Remove some deleted constants from imports.
* Fix unittest to remove usage of PyTorch model for onnx.utils.
* Fix distilbert export
* Enable ONNX export test for supported model.
* Style.
* Fix lint.
* Enable all supported default models.
* GPT2 only has one output
* Fix bad property name when overriding config.
* Added unittests and docstrings.
* Disable with_past tests for now.
* Enable outputs validation for default export.
* Remove graph opt lvls.
* Last commit with on-going past commented.
* Style.
* Disabled `with_past` for now
* Remove unused imports.
* Remove framework argument
* Remove TFPreTrainedModel reference
* Add documentation
* Add onnxruntime tests to CircleCI
* Add test
* Rename `convert_pytorch` to `export`
* Use OrderedDict for dummy inputs
* WIP Wav2Vec2
* Revert "WIP Wav2Vec2"
This reverts commit f665efb04c92525c3530e589029f0ae7afdf603e.
* Style
* Use OrderedDict for I/O
* Style.
* Specify OrderedDict documentation.
* Style :)
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Copy BART to MBart and rename some stuff
* Add copy statements pointing to FlaxBart
* Update/add some common files
* Update shift_tokens_rigth + fix imports
* Fix shift_tokens_right method according to MBart implementation
* Update shift_tokens_right in tests accordingly
* Fix the import issue and update docs file
* make style quality
* Do some minor changes according to patil-suraj suggestions
* Change the order of normalization layer and attention
* Add some copu statementes
* Update generate method and add integration test for mBart
* Make a few updates after a review
Besides, add `lang_code_to_id` to MBartTokenizeFast
* fix-copies; make style quality
* Apply suggestions from code review
* Apply suggestions from code review
* Apply suggestions from code review
* fix output type, style
* add copied from
* resolve conflicts
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* First pass
* More progress
* Add support for local attention
* More improvements
* More improvements
* Conversion script working
* Add CanineTokenizer
* Make style & quality
* First draft of integration test
* Remove decoder test
* Improve tests
* Add documentation
* Mostly docs improvements
* Add CanineTokenizer tests
* Fix most tests on GPU, improve upsampling projection
* Address most comments by @dhgarrette
* Remove decoder logic
* Improve Canine tests, improve docs of CanineConfig
* All tokenizer tests passing
* Make fix-copies and fix tokenizer tests
* Fix test_model_outputs_equivalence test
* Apply suggestions from @sgugger's review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Address some more comments
* Add support for hidden_states and attentions of shallow encoders
* Define custom CanineModelOutputWithPooling, tests pass
* First pass
* More progress
* Add support for local attention
* More improvements
* More improvements
* Conversion script working
* Add CanineTokenizer
* Make style & quality
* First draft of integration test
* Remove decoder test
* Improve tests
* Add documentation
* Mostly docs improvements
* Add CanineTokenizer tests
* Fix most tests on GPU, improve upsampling projection
* Address most comments by @dhgarrette
* Remove decoder logic
* Improve Canine tests, improve docs of CanineConfig
* All tokenizer tests passing
* Make fix-copies and fix tokenizer tests
* Fix test_model_outputs_equivalence test
* Apply suggestions from @sgugger's review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Address some more comments
* Make conversion script work for Canine-c too
* Fix tokenizer tests
* Remove file
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* copy pytorch-t5
* init
* boom boom
* forward pass same
* make generation work
* add more tests
* make test work
* finish normal tests
* make fix-copies
* finish quality
* correct slow example
* correct slow test
* version table
* upload models
* Update tests/test_modeling_flax_t5.py
* correct incorrectly deleted line
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
* [run_clm.py] restore caching
* style
* [t5 doc] make the example work out of the box
This PR expands the training example to include the correct model type for the example to work, e.g. with `T5Model` this example will break.
* Update docs/source/model_doc/t5.rst
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* expand the other example
Co-authored-by: Suraj Patil <surajp815@gmail.com>
- Convert use of deprecated AutoModelWithLMHead to AutoModelForSeq2SeqLM
- Add newly required `truncation=True` to `tokenizer.encode` with `max_length`
This silences all warnings.
* [WIP] Add TFWav2Vec2Model
Work in progress for adding a tensorflow version of Wav2Vec2
* feedback changes
* small fix
* Test Feedback Round 1
* Add SpecAugment and CTC Loss
* correct spec augment mask creation
* docstring and correct copyright
* correct bugs
* remove bogus file
* finish tests correction
* del unnecessary layers
* Update src/transformers/models/wav2vec2/modeling_tf_wav2vec2.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* make style
* correct final bug
* Feedback Changes
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Start working on FlaxBart
* Create modeling_flax_bart.py
* Write FlaxBartAttention
* Add FlaxBartEncoderLayer
* Add FlaxBartDecoderLayer and some typing
* Add helepr function for FlaxBart
* shift_tokens_right
* _make_causal_mask
* _expand_mask
* Add PositionalEmbedding and fix init_std naming
* Add FlaxBartPretrainedModel
* Add FlaxBartEncoder
* Add FlaxBartEncoder
* Add FlaxBartEncoder among modules to be imported
* YET WE CANNOT INITIALIZE THAT!! :(
* Make BartEncoder working
Change BartEncoder to instance of nn.Module so far
* Add FlaxBartDecoder
* Add FlaxBartModel
* TODO to make model run -> Prepapre model inputs
* Resolve padding
* Add FlaxBartModel
* Add FlaxBartModel into importable modules
* Remove FlaxBartEncoder and FlaxBartDecoder from importable modules
* make style; not properly working
* make style; make quality not pass due to some import I left
* Remove TODO for padding_idx in nn.Embed so far
* Add FlaxBartForConditionalGeneration
* Incorporate Flax model output classes, i.e. return_dict
* Add another models and incorporate use_cache arg
* Add FlaxBartForSequenceClassification and FlaxBartForQuestionAnswering
* Incorporate use_cache arg from PyTorch implementation
* Add all necessary Flax output utils
* Add FlaxBartForCausalLM; not working yet'
* Add minor improvements; still lacks some functionality
* Update docs, src and tests
* Add support of FlaxBart to docs/source
* Fix some bugs in FlaxBart souce code
* Add some neccessary tests for FlaxBart models - jit_compilation not passing
* Fix tests and add test_head_masking
* Fix tests for @jax.jit computation
* Add test_head_masking
* Migrate FlaxBart tests from jax.numpy to numpy
* Remove FlaxBartForCausalLM
* Clean repo
* fix bart model weight structure
* Fix FlaxBartForSequenceClassification
Slicing is not possible to use below jit, therefore, selecting sentence
representation from hidden_states must be changed.
* Allow FlaxBartForSequenceClassification for testing pt_flax equivalence
* Allow testing for FlaxBartForQA for pt_flax equivalence
* Add a comment to FlaxBartForSequenceClassification + change noise from 1e-3 to 1e-6
* remove past_key_values
* remove inputs_mebeds and make input_ids required
* add position ids
* re-write attention layer
* fix dataclass
* fix pos embeds and attention output
* fix pos embeds
* expose encode method
* expose decode method
* move docstring to top
* add cache for causal attn layer
* remove head masking for now
* s2s greedy search first pass
* boom boom
* fix typos
* fix greedy generate for bart
* use encoder, decoder layers instead of num_hidden_layers
* handle encoder_outputs
* cleanup
* simplify decoding
* more clean-up
* typos
* Change header + add {decoder_,}position_ids into 2 models
* add BartConfig
* fix existing tests
* add encode, decode methods
* Fix shift_tokens_right for JIT compilation + clarify one condition
* fix decode
* encoder => encode
* simplify generate
* add tests for encode and decode
* style
* add tests for cache
* fix equivalence tests
* sample generate now works with seq2seq
* generation tests
* initialize dense layers
* docstring and cleanup
* quality
* remove get/set input_embeddings
* address Patricks suggestions
* decode for every model, remove encoder_outputs from call
* update tests accordingly
* decode returns only decoder outputs and logits
* fix arguments
* doc encode, decode methods
* correct base_model_prefix
* fix test for seq classif model
* fix docs
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* adding vit for flax
* added test for Flax-vit and some bug-fixes
* overrided methods where variable changes were necessary for flax_vit test
* added FlaxViTForImageClassification for test
* Update src/transformers/models/vit/modeling_flax_vit.py
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* made changes suggested in PR
* Adding jax-vit models for autoimport
* swapping num_channels and height,width dimension
* fixing the docstring for torch-like inputs for VIT
* add model to main init
* add docs
* doc, fix-copies
* docstrings
* small test fixes
* fix docs
* fix docstr
* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* style
Co-authored-by: jayendra <jayendra@infocusp.in>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Squash all commits of modeling_detr_v7 branch into one
* Improve docs
* Fix tests
* Style
* Improve docs some more and fix most tests
* Fix slow tests of ViT, DeiT and DETR
* Improve replacement of batch norm
* Restructure timm backbone forward
* Make DetrForSegmentation support any timm backbone
* Fix name of output
* Address most comments by @LysandreJik
* Give better names for variables
* Conditional imports + timm in setup.py
* Address additional comments by @sgugger
* Make style, add require_timm and require_vision to testsé
* Remove train_backbone attribute of DetrConfig, add methods to freeze/unfreeze backbone
* Add png files to fixtures
* Fix type hint
* Add timm to workflows
* Add `BatchNorm2d` to the weight initialization
* Fix retain_grad test
* Replace model checkpoints by Facebook namespace
* Fix name of checkpoint in test
* Add user-friendly message when scipy is not available
* Address most comments by @patrickvonplaten
* Remove return_intermediate_layers attribute of DetrConfig and simplify Joiner
* Better initialization
* Scipy is necessary to get sklearn metrics
* Rename TimmBackbone to DetrTimmConvEncoder and rename DetrJoiner to DetrConvModel
* Make style
* Improve docs and add 2 community notebooks
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
* Improve docs of DeiT and ViT, add community notebook
* Add gitignore for test_samples
* Add notebook with Trainer
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Added Big Bird Fast Tokenizer initial file
* style fixes
* flake fixes
* Added big bird fast tokenizer to init files
* Added big bird fast to Auto tokenization
* fix styles
* minor quality fixes
* Added initial test code
* Fix SpmConverter when precompiled_charsmap doesn't exist
* fixed post processor
* minor style fix
* minor fix input names
* Actually fix identity normalization
* style
* Added token type ids to fast tokenizer
* style
* flake fix
* fix copies
Co-authored-by: Anthony MOI <m.anthony.moi@gmail.com>
* Add the ImageClassificationPipeline
* Code review
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>
* Have `load_image` at the module level
Co-authored-by: patrickvonplaten <patrick.v.platen@gmail.com>
* add electra model to flax
* Remove Electra Next Sentence Prediction model added by mistake
* fix parameter sharing and loosen equality threshold
* fix styling issues
* add mistaken removen imports
* fix electra table
* Add FlaxElectra to automodels and fixe docs
* fix issues pointed out the PR
* fix flax electra to comply with latest changes
* remove stale class
* add copied from
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* add flax roberta
* make style
* correct initialiazation
* modify model to save weights
* fix copied from
* fix copied from
* correct some more code
* add more roberta models
* Apply suggestions from code review
* merge from master
* finish
* finish docs
Co-authored-by: Patrick von Platen <patrick@huggingface.co>
* Rebase with master
* Minor bug fix in docs
* Copy files from adding_luke_v2 and improve docs
* change the default value of use_entity_aware_attention to True
* remove word_hidden_states
* fix head models
* fix tests
* fix the conversion script
* add integration tests for the pretrained large model
* improve docstring
* Improve docs, make style
* fix _init_weights for pytorch 1.8
* improve docs
* fix tokenizer to construct entity sequence with [MASK] entity when entities=None
* Make fix-copies
* Make style & quality
* Bug fixes
* Add LukeTokenizer to init
* Address most comments by @patil-suraj and @LysandreJik
* rename _compute_extended_attention_mask to get_extended_attention_mask
* add comments to LukeSelfAttention
* fix the documentation of the tokenizer
* address comments by @patil-suraj, @LysandreJik, and @sgugger
* improve docs
* Make style, quality and fix-copies
* Improve docs
* fix docs
* add "entity_span_classification" task
* update example code for LukeForEntitySpanClassification
* improve docs
* improve docs
* improve the code example in luke.rst
* rename the classification layer in LukeForEntityClassification from typing to classifier
* add bias to the classifier in LukeForEntitySpanClassification
* update docs to use fine-tuned hub models in code examples of the head models
* update the example sentences
* Make style & quality
* Add require_torch to tokenizer tests
* Add require_torch to tokenizer tests
* Address comments by @sgugger and add community notebooks
* Make fix-copies
Co-authored-by: Ikuya Yamada <ikuya@ikuya.net>
* prep for deepspeed==0.3.16
* new version
* too soon
* support and test fp32 mode
* troubleshooting doc start
* workaround no longer needed
* add fp32 doc
* style
* cleanup, add tf32 note
* clarify
* release was made
* Adding `AutomaticSpeechRecognitionPipeline`.
- Because we added everything to enable this pipeline, we probably
should add it to `transformers`.
- This PR tries to limit the scope and focuses only on the pipeline part
(what should go in, and out).
- The tests are very specific for S2T and Wav2vec2 to make sure both
architectures are supported by the pipeline. We don't use the mixin for
tests right now, because that requires more work in the `pipeline`
function (will be done in a follow up PR).
- Unsure about the "helper" function `ffmpeg_read`. It makes a lot of
sense from a user perspective, it does not add any additional
dependencies (as in hard dependency, because users can always use their
own load mechanism). Meanwhile, it feels slightly clunky to have so much
optional preprocessing.
- The pipeline is not done to support streaming audio right now.
Future work:
- Add `automatic-speech-recognition` as a `task`. And add the
FeatureExtractor.from_pretrained within `pipeline` function.
- Add small models within tests
- Add the Mixin to tests.
- Make the logic between ForCTC vs ForConditionalGeneration better.
* Update tests/test_pipelines_automatic_speech_recognition.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Adding docs + main import + type checking + LICENSE.
* Doc style !.
* Fixing TYPE_HINT.
* Specifying waveform shape in the docs.
* Adding asserts + specify in the documentation the shape of the input
np.ndarray.
* Update src/transformers/pipelines/automatic_speech_recognition.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Adding require to tests + move the `feature_extractor` doc.
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Initial support for upload to hub
* push -> upload
* Fixes + examples
* Fix torchhub test
* Torchhub test I hate you
* push_model_to_hub -> push_to_hub
* Apply mixin to other pretrained models
* Remove ABC inheritance
* Add tests
* Typo
* Run tests
* Install git-lfs
* Change approach
* Add push_to_hub to all
* Staging test suite
* Typo
* Maybe like this?
* More deps
* Cache
* Adapt name
* Quality
* MOAR tests
* Put it in testing_utils
* Docs + torchhub last hope
* Styling
* Wrong method
* Typos
* Update src/transformers/file_utils.py
Co-authored-by: Julien Chaumond <julien@huggingface.co>
* Address review comments
* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Julien Chaumond <julien@huggingface.co>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Base move
* Examples reorganization
* Update references
* Put back test data
* Move conftest
* More fixes
* Move test data to test fixtures
* Update path
* Apply suggestions from code review
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Address review comments and clean
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Start writing BERT-Japanese doc
* Fix typo, Update toctree
* Modify model file to use comment for document, Add examples
* Clean bert_japanese by make style
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Split a big code block into two
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Add prefix >>> to all lines in code blocks
* Clean bert_japanese by make fixup
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* First draft of deit
* More improvements
* Remove DeiTTokenizerFast from init
* Conversion script works
* Add DeiT to ViT conversion script
* Add tests, add head model, add support for deit in vit conversion script
* Update model checkpoint names
* Update image_mean and image_std, set resample to bicubic
* Improve docs
* Docs improvements
* Add DeiTForImageClassificationWithTeacher to init
* Address comments by @sgugger
* Improve feature extractors
* Make fix-copies
* Minor fixes
* Address comments by @patil-suraj
* All models uploaded
* Fix tests
* Remove labels argument from DeiTForImageClassificationWithTeacher
* Fix-copies, style and quality
* Fix tests
* Fix typo
* Multiple docs improvements
* More docs fixes
* Added documentation for data collator.
* Update docs/source/data_collator.rst
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Added documentation for data collator.
* Added documentation for the data collator.
* Merge branch 'doc_DataCollator' of C:\Users\mahii\PycharmProjects\transformers with conflicts.
* Update documentation for the data collator.
* Update documentation for the data collator.
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Amna <A.A.Ahmad@student.tudelft.nl>
* Add a special tokenizer for CPM model
* make style
* fix
* Add docs
* styles
* cpm doc
* fix ci
* fix the overview
* add test
* make style
* typo
* Custom tokenizer flag
* Add REAMDE.md
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
* make fairscale and deepspeed setup extras
* fix default
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* no reason not to ask for the good version
* update the CIs
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* AutoFeatureExtractor
* Init and first tests
* Tests
* Damn you gitignore
* Quality
* Defensive test for when not all backends are here
* Use pattern for Speech2Text models
* Documentation about loading a fast tokenizer within Transformers
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* style
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Squash all commits into one
* Update ViTFeatureExtractor to use image_utils instead of torchvision
* Remove torchvision and add Pillow
* Small docs improvement
* Address most comments by @sgugger
* Fix tests
* Clean up conversion script
* Pooler first draft
* Fix quality
* Improve conversion script
* Make style and quality
* Make fix-copies
* Minor docs improvements
* Should use fix-copies instead of manual handling
* Revert "Should use fix-copies instead of manual handling"
This reverts commit fd4e591bce.
* Place ViT in alphabetical order
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Added embeddings layer
* Added layoutlm layers, main model, maskedlm and token classification classes
* Added model classes to tf auto models
* Added model to PT to TF conversion script
* Added model to doc README
* Added tests
* Removed unused imports
* Added layoutlm model, test, and doc for sequence classification, and fix imports in __init__.py
* Made tests pass!
* Fixed typos in imports and docs
* Fixed a typo in embeddings layer
* Removed imports
* Fixed formatting issues, imports, tests
* Added layoutlm layers, main model, maskedlm and token classification classes
* Added model classes to tf auto models
* Added model to PT to TF conversion script
* Removed unused imports
* Added layoutlm model, test, and doc for sequence classification, and fix imports in __init__.py
* Made tests pass!
* Fixed typos in imports and docs
* Removed imports
* Fixed small formatting issues
* Removed duplicates import from main __init__.py
* Chnaged deafult arg to true for adding pooling layer to tf layoutlm
* Fixed formatting issues
* Style
* Added copied from to classes copied from bert
* Fixed doc strings examples to work with layoutlm inputs
* Removed PyTorch reference in doc strings example
* Added integration tests
* Cleaned up initialization file
* Updated model checkpoint identifiers
* Fixed imports
Co-authored-by: Amir Tahmasbi <amir@ehsai.ca>
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
* [doc] [testing] extend -k section
This PR adds more examples on using `pytest -k` - I always forget that I want to use `-k A OR B` when I want several tests - I keep trying AND and it doesn't match any.
* style
* pass hf optimizer and scheduler to deepspeed if not specified in ds config
* pass hf optimizer and scheduler to deepspeed if not specified in ds config
* update
* make init_deepspeed support config dict
* fix docstring formatting
* clean up trainer's comments
* add new tests
* fix type
* composit argparse doesn't work
* style
* add a new test, rename others
* document new functionality
* complete tests, add docs
* style
* correct level
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* add new methods to the doc
* must tell DS we are using a non-native optimizer
* add protection against cpu_offload + HF optimizer combo
* fix the cli overrides
* sync docs + tests
* restore AdamW
* better docs
* need new version
* no longer needed
* remove outdate information
* refactor duplicated code
Co-authored-by: Stas Bekman <stas@stason.org>
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* [WIP] Adding new parameter to `generate`: `max_time`.
Generation by tokens number is sometimes a bit clunky because we don't
know how many tokens are good enough or even how many tokens are in
the payload (for pipelines users for instance). This leads to hard
to understand behavior.
This PR proposes a new argument `max_time` which is a float of seconds
for the allowed time for `generate` to run on.
Ideally combinations of `max_tokens=None`, `max_time=2` could be used to
generate as many tokens as possible within time budget.
NB: Another possible approach consists of passing a callback to `generate`
putting the caller in charge of the actual decision of when to stop
generating tokens. It opens the door to 'which args should we pass'
to this callback. It's hard to imagine other use-cases for this
early stopping behavior than time (that are not already covered by
parameters of generate)
* Revamp with StoppingCriteria
* Removing deprecated mentions.
* Forgot arguments to stopping criteria.
* Readding max_length it's not just used as a stopping criteria.
* Default value for `stopping_criteria`.
* Address @patrickvonplaten comments.
- More docstrings
- Actual doc
- Include in global namespace
- Remove TF work.
* Put back `max_length` (deprecation different PR).
* Doc quality.
* Fixing old behavior without `stopping_criteria` but with `max_length`.
Making sure we don't break that in the future.
* Adding more tests for possible inconsistencies between
`max_length` and `stopping_criteria`.
* Fixing the torch imports.
* 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
* create modeling_tf_rag
* add tests for tf
* add tf tests
* revert wrong pt commit
* further refactor
* further refactor
* refactor
* Update modeling_tf_rag.py
- input_processing
- fix prepare_input_for_generation (mostly fix generate bug)
- bring back from_pretrained hack in order to test generate
* delete colab pieces of code
* Show case of greedy "generate"
Temporarily change from beam_search test to greedy_search test to show case that TF and PT do get equivalent output.
* cosmetic update
* correct typos
* update
* push some progress
* make easy check
* fix rag save from pretrained
* Update src/transformers/modeling_tf_utils.py
* remove commented out lines
* delete unnecessary lines
* add simple test case for nq_checkpoint
Add nq_checkpoint test to show that current version without hack still fails
* temporarily put ugly hack back again
* Add TFRagSequenceForGeneration!!
* __init__.py , import TFRagSequenceForGeneration
* Add TFRagSequence tests!
* rag init.py - add TFRagSequenceForGeneration
* fix from_pretrained
* fix prepare_inputs_for_generation
* Beam search for RagToken!
* minor clean up
* add tf.cast in TFRagModel
* More tf.cast
* Add all remaining tests (still have issues)
* delete all T5 related
* make style
* fix load weight prefix
* fix bart
* fix return_dict for tf_rag
make all tests pass .. Hooray
* fix some tests
* fix code quality
* fix qualtiy check
* finish tests tf rag
* add tf rag to docs
* remove TFT5 from docstring
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* remove TFT5 from docstring
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Delete outdated comments
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* improve doc strings
* add generative model classes
* fix adjust token logic
* refactor generate for TFRag
* using shape_list, not _get_shape
Co-authored-by: Julien Plu <plu.julien@gmail.com>
* axis=[1]->axis=1
* delete NEED_HELP comment
* improve readability
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* improve readability
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* improve readability
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Indicating model is in a developing state in docstrings
As suggested by Julien
* small last changes
* apply sylvains suggestions
* finish tf rag
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: patrickvonplaten <patrick@huggingface.co>
Co-authored-by: Julien Plu <plu.julien@gmail.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* offline mode start
* add specific values
* fix fallback
* add test
* better values check and range
* test that actually works
* document the offline mode
* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* more strict check
* cleaner test
* pt-only test
* style
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* feat: Adds three definitions to glossary from @cronoik
Needed a definition for transformer which in turn needed 2 more definitions
To do with issue https://github.com/huggingface/transformers/issues/9078
* fix: Adjusts definition of neural network to make it easier to read
* Introduce save_strategy training argument
* deprecate EvaluationStrategy
* collapse EvaluationStrategy and LoggingStrategy into a single
IntervalStrategy enum
* modify tests to use modified enum
* Fix None in add_token_positions - issue #10210
Fix None in add_token_positions related to the issue #10210
* add_token_positions fix None values in end_positions vector
add_token_positions fix None in end_positions vector as proposed by @joeddav
