GitLab

* fix FA2 tests

* refactor inference test name

* Start rework

* Fix failing test

* Include max

* Update src/transformers/trainer.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

---------
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* fix

* revert for qwen2

* revert for qwen2

* update

* update

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

* fix

* fix test

* style

* nit

* rather rely on concert token to id

* fix quality

* Update src/transformers/convert_slow_tokenizer.py

* add gradient_accumulation_kwargs to AcceleratorConfig

* add suggestions from @muellerzr to docstrings, new behavior and tests

* Documentation suggestions from @muellerz
Co-authored-by: Zach Mueller <muellerzr@gmail.com>

* addressed @muellerzr comments regarding tests and test utils

* moved accelerate version to top of file.

* @muellerzr's variable fix
Co-authored-by: Zach Mueller <muellerzr@gmail.com>

* address @amyeroberts. fix tests and docstrings

* address @amyeroberts additional suggestions

---------
Co-authored-by: Yu Chin Fabian Lim <flim@sg.ibm.com>
Co-authored-by: Zach Mueller <muellerzr@gmail.com>

[ `TokenizationLlama`] fix the way we convert tokens to strings to keep leading spaces 🚨 breaking fix (#29453)

* nit

* update test and fix test

* fixup

* add hard rope scaling test

* make fixup

* quick rope scaling tests

* add copy statements

* add functions to get number of params which require grad, get optimizer group for parameters and get learning rates of param groups to trainer.py

* add tests and raise ValueError when optimizer is None

* add second layer to test and freeze its weigths

* check if torch is available before running tests

* use decorator to check if torch is available
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* fix test indentation
Co-authored-by: Zach Mueller <muellerzr@gmail.com>

---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
Co-authored-by: Zach Mueller <muellerzr@gmail.com>

* replace torch.testing.assert_allclose by torch.testing.assert_close

* missing atol rtol

* First commit to add flash attention 2 for GPT-2

* more improvements

* Make GPT2 pass tests and fixed Decison Transformers copies

* Fixed missing arg

* fix copies

* Added expected speedup

* Update src/transformers/models/gpt2/modeling_gpt2.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/models/gpt2/modeling_gpt2.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/models/gpt2/modeling_gpt2.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Added test

* Fixed attn attribute

* Update docs/source/en/model_doc/gpt2.md
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update docs/source/en/model_doc/gpt2.md
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update Decision transformer attentions

* More updates

* Passing tests

* Fix copies

* Fix copies part 2

* Decision transformer updates

* Update src/transformers/models/gpt2/modeling_gpt2.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Fix copies

* Decision transformer not supporting flash attn

* Addressed comments

* Addressed comments

* Addressed comments

---------
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

This commit adds gate jitter to MixtralSparseMoeBlock's input data
before passing it through the MoE layer, if turned on.

* add eos stopping criteria

* minor fix

* Update tests/generation/test_stopping_criteria.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* check eos is not None and fix tests

* make style and fixup

* Update src/transformers/generation/stopping_criteria.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* Update tests/generation/test_utils.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* Update tests/generation/test_utils.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* Update src/transformers/generation/__init__.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/generation/stopping_criteria.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/generation/stopping_criteria.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/generation/stopping_criteria.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* camel case everywhere

* call stopping criteria list for candidate ids

* make style  and fixup

* Empty commit

* Empty commit to pass flaky test

* set max length in PromptLookupCandidateGenerator

* Update src/transformers/generation/utils.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* lets fix this typo in docs

* Update src/transformers/generation/utils.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/generation/utils.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* update PR

* empty commit

---------
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Automatic safetensors conversion when lacking these files (#29390)

* Automatic safetensors conversion when lacking these files

* Remove debug

* Thread name

* Typo

* Ensure that raises do not affect the main thread

* Catch all errors

* Check for requires_grad when initing weights

* Add unit test

* Move sinusoidal positional encoding generation after post_init()

* Add modules to skip init list

* Move create_sinusoidal_embeddings to _init_weights

* FIX: Cached slow forward in mamba
- additionally added mamba cached test
- added unused test (mamba causal lm forward and backward)
- fixed typo: "causl" --> "causal"

* formatting

* fix: use real `slow_forward` call instead of torch module's

* add shape assertion for mixer block test

* adjust shape assertion

* add support for qwen2 MoE models

* update docs

* add support for qwen2 MoE models

* update docs

* update model name & test

* update readme

* update class names & readme & model_doc of Qwen2MoE.

* update architecture name

* fix qwen2_moe tests

* use Qwen2Tokenizer instead of Qwen2MoeTokenizer

* update modeling_qwen2_moe.py

* fix model architecture

* fix qwen2_moe tests

* use Qwen2Tokenizer instead of Qwen2MoeTokenizer

* update modeling_qwen2_moe.py

* fix model architecture

* fix style

* fix test when there are sparse and non sparse layers

* fixup

* Update README.md
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* fixup

* fixup

* add archive back

* add support for qwen2 MoE models

* update docs

* update model name & test

* update readme

* update class names & readme & model_doc of Qwen2MoE.

* update architecture name

* fix qwen2_moe tests

* use Qwen2Tokenizer instead of Qwen2MoeTokenizer

* update modeling_qwen2_moe.py

* fix model architecture

* fixup

* fix qwen2_moe tests

* use Qwen2Tokenizer instead of Qwen2MoeTokenizer

* fix style

* fix test when there are sparse and non sparse layers

* fixup

* add archive back

* fix integration test

* fixup

---------
Co-authored-by: bozheng-hit <dsoul0621@gmail.com>
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Add cosine_with_min_lr scheduler

* Update error message for missing min_lr or min_lr_rate

* update

* add ut

* update

* replace the 'decord' with 'av' in VideoClassificationPipeline

* fix the check of backend in VideoClassificationPipeline

* adjust the order of imports

* format 'video_classification.py'

* format 'video_classification.py' with ruff

---------
Co-authored-by: wanqiancheng <13541261013@163.com>

* add warnings if training args differ from checkpoint args stored in trainer_state.json

* run formatting and styling

* add a test

* format and styling

---------
Co-authored-by: Jonathan Flynn <jonl.flynn@guardian.co.uk>

* Populate torch_dtype from model to pipeline
Signed-off-by: B-Step62 <yuki.watanabe@databricks.com>

* use property
Signed-off-by: B-Step62 <yuki.watanabe@databricks.com>

* lint
Signed-off-by: B-Step62 <yuki.watanabe@databricks.com>

* Remove default handling
Signed-off-by: B-Step62 <yuki.watanabe@databricks.com>

---------
Signed-off-by: B-Step62 <yuki.watanabe@databricks.com>

* [test_all] Remove static pretrained maps from the library's internals

* Deprecate archive maps instead of removing them

* Revert init changes

* [test_all] Deprecate instead of removing

* [test_all] PVT v2 support

* [test_all] Tests should all pass

* [test_all] Style

* Address review comments

* Update src/transformers/models/deprecated/_archive_maps.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/models/deprecated/_archive_maps.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* [test_all] trigger tests

* [test_all] LLAVA

* [test_all] Bad rebase

---------
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* correct llava mask

* fix vipllava as wlel

* mask out embedding for padding tokens

* add test

* fix style

* add setter

* fix test on suggestion

* change in-place -> out-of-place

* add tests

* add more tests

* naming consistency

* fix doctest

* forgot min-length processors

* empty

* Revert "fix doctest"

This reverts commit 4772768457f9bc057f1d4d9d67ea94eb7224eb8d.

* revert change in docstring

* Update tests/generation/test_logits_process.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update tests/generation/test_logits_process.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* prepend "bos" to blip generation

* minor changes

* Update src/transformers/models/blip_2/modeling_blip_2.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* Update src/transformers/models/instructblip/modeling_instructblip.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* add generation tester mixin

---------
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Cast bfloat16 to float32 for Numpy conversions

* Add test

* attempt to fix

* the actual fix that works with compilation!

* this?

* temporary update

* nit?

* dispatcg to memory efficient?

* update both models that have static cache support

* fix copies fix compile

* make sure fix

* fix cohere and gemma

* fix beams?

* nit

* slipped through the cracks

* nit

* nits

* update

* fix-copies

* skip failing tests

* nits

* Update test reqs

* Clean

* First draft

* Fix tests, add docs

* Improve docstrings

* Fix test

* Address comments

* Address comments

* Remove vocab_size attribute

* Remove batch_size

* Address comment

* Add image processor tests

* Support fx

* Update docstring

* Add support for 34b

* Convert 34b model

* Add integration tests

* Update checkpoints

* Convert vicuna-13b, remove doc tests

* Remove script

* Remove file

* Address comments

* Improve docstrings

* Deprecate vocab_size

* Remove aspect_ratio_setting

* Address comments

* Update READMEs

* Add tips about chat templates

* Fix tests

* Deprecate vocab_size safely

* Update tests

---------
Co-authored-by: Amy Roberts <22614925+amyeroberts@users.noreply.github.com>

* Add correct batched handling for apply_chat_template

* Fix warning method

* Add error for incompatible options

* expand tests

* Add a skip for markuplm

* Add skips for other layout models

* Skip for LayoutLMv2

* Slightly update the warning message

* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* typo fix

* Update docstring for conversation kwarg

* Update return docstring

* Remove the warning, improve error message

* Update src/transformers/tokenization_utils_base.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update src/transformers/tokenization_utils_base.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update tests/test_tokenization_common.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update tests/test_tokenization_common.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Remove return_dict=None

* Fix up some merge cruft

* More merge cruft

* Add another skip

* Add another skip

---------
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Initial commit (still lots of unfinished bits)

* (Still untested) add safetensors sharding to save_pretrained

* Fix savetensors saving, update default shard size to match PT

* Add proper loading of TF-format safetensors

* Revert default size in case that changes things

* Fix incorrect index name

* Update loading priority

* Update tests

* Make the tests a little more stringent

* Expand tests

* Add sharded cross-test

* Fix argument name

* One more test fix

* Adding mlx to the list of allowed formats

* Remove irrelevant block for safetensors

* Refactor warning logging into a separate function

* Remove unused skip_logger_warnings arg

* Update src/transformers/modeling_tf_utils.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Move function def

---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

Remove unused code

* make fix-copies

* some tests fixed

* tests fixed

* partial 4d masks

* Apply suggestions from code review
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* clean-up tests

* refine comments

* fix musicgen tests

* make style

* remove slow decorator from a test

* more clean-up

* fix other failing tests

* Added SuperPoint docs

* Added tests

* Removed commented part

* Commit to create and fix add_superpoint branch with a new branch

* Fixed dummy_pt_objects

* Committed missing files

* Fixed README.md

* Apply suggestions from code review

Fixed small changes
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Moved ImagePointDescriptionOutput from modeling_outputs.py to modeling_superpoint.py

* Removed AutoModelForKeypointDetection and related stuff

* Fixed inconsistencies in image_processing_superpoint.py

* Moved infer_on_model logic simply in test_inference

* Fixed bugs, added labels to forward method with checks whether it is properly a None value, also added tests about this logic in test_modeling_superpoint.py

* Added tests to SuperPointImageProcessor to ensure that images are properly converted to grayscale

* Removed remaining mentions of MODEL_FOR_KEYPOINT_DETECTION_MAPPING

* Apply suggestions from code review
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Fixed from (w, h) to (h, w) as input for tests

* Removed unnecessary condition

* Moved last_hidden_state to be the first returned

* Moved last_hidden_state to be the first returned (bis)

* Moved last_hidden_state to be the first returned (ter)

* Switched image_width and image_height in tests to match recent changes

* Added config as first SuperPointConvBlock init argument

* Reordered README's after merge

* Added missing first config argument to SuperPointConvBlock instantiations

* Removed formatting error

* Added SuperPoint to README's de, pt-br, ru, te and vi

* Checked out README_fr.md

* Fixed README_fr.md

* Test fix README_fr.md

* Test fix README_fr.md

* Last make fix-copies !

* Updated checkpoint path

* Removed unused SuperPoint doc

* Added missing image

* Update src/transformers/models/superpoint/modeling_superpoint.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Removed unnecessary import

* Update src/transformers/models/superpoint/modeling_superpoint.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Added SuperPoint to _toctree.yml

---------
Co-authored-by: steven <steven.bucaillle@gmail.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
Co-authored-by: Steven Bucaille <steven.bucaille@buawei.com>

* use user_defined_symbols

* fixup

* nit

* add a very robust test

* make sure all models are tested with the `pretrained_tokenizer_to_test`

* should we make sure we test all of them?

* merge

* remove the id

* fix test

* update

* ousies

* oups

* fixup

* fix copies check

* remove `pretrained_tokenizer_to_test`

* add galore v1

* add import

* add tests and doc

* fix doctest

* forward contrib credits from discussions

* forward contrib credits from discussions

* Apply suggestions from code review
Co-authored-by: Zach Mueller <muellerzr@gmail.com>

* fix failing tests'

* switch to `optim_target_modules` and clarify docs

* more clarification

* enhance lookup logic

* update a test to add peak memory

* add regex, all-linear and single string support

* add layer-wise optimization through DummyOptimizers and LRSchedulers

* forward contrib credits from discussions and original idea

* add a section about DDP not supported in layerwise

* Update src/transformers/trainer.py
Co-authored-by: Zach Mueller <muellerzr@gmail.com>

* fix self

* check only if layer_wise

* Update src/transformers/training_args.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* oops

* make use of intervals

* clarify comment

* add matching tests

* GaLoRe -> GaLore

* move to `get_scheduler`

* add note on docs

* add a warning

* adapt a bit the docs

* update docstring

* support original API

* Update docs/source/en/trainer.md

* slightly refactor

* Update docs/source/en/trainer.md
Co-authored-by: Matthew Douglas <38992547+matthewdouglas@users.noreply.github.com>

* Update src/transformers/training_args.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* fix args parsing and add tests

* remove warning for regex

* fix type hint

* add note about extra args

* make `is_regex` return optional

---------

Co-authored-by: Maxime <maximegmd @users.noreply.github.com>
Co-authored-by: Wing Lian <winglian @users.noreply.github.com>
Co-authored-by: Zach Mueller <muellerzr@gmail.com>
Co-authored-by: hiyouga <hiyouga@users.noreply.github.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
Co-authored-by: Matthew Douglas <38992547+matthewdouglas@users.noreply.github.com>

* first modeling code

* make repository

* still WIP

* update model

* add tests

* add latest change

* clean docstrings and copied from

* update docstrings md and readme

* correct chroma function

* correct copied from and remove unreleated test

* add doc to toctree

* correct imports

* add convert script to notdoctested

* Add suggestion from Sanchit
Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

* correct get_uncoditional_inputs docstrings

* modify README according to SANCHIT feedback

* add chroma to audio utils

* clean librosa and torchaudio hard dependencies

* fix FE

* refactor audio decoder -> audio encoder for consistency with previous musicgen

* refactor conditional -> encoder

* modify sampling rate logics

* modify license at the beginning

* refactor all_self_attns->all_attentions

* remove ignore copy from causallm generate

* add copied from for from_sub_models

* fix make copies

* add warning if audio is truncated

* add copied from where relevant

* remove artefact

* fix convert script

* fix torchaudio and FE

* modify chroma method according to feedback-> better naming

* refactor input_values->input_features

* refactor input_values->input_features and fix import fe

* add input_features to docstrigs

* correct inputs_embeds logics

* remove dtype conversion

* refactor _prepare_conditional_hidden_states_kwargs_for_generation ->_prepare_encoder_hidden_states_kwargs_for_generation

* change warning for chroma length

* Update src/transformers/models/musicgen_melody/convert_musicgen_melody_transformers.py
Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

* change way to save wav, using soundfile

* correct docs and change to soundfile

* fix import

* fix init proj layers

* remove line breaks from md

* fix issue with docstrings

* add FE suggestions

* improve is in logics and remove useless imports

* remove custom from_pretrained

* simplify docstring code

* add suggestions for modeling tests

* make style

* update converting script with sanity check

* remove encoder attention mask from conditional generation

* replace musicgen melody checkpoints with official orga

* rename ylacombe->facebook in checkpoints

* fix copies

* remove unecessary warning

* add shape in code docstrings

* add files to slow doc tests

* fix md bug and add md to not_tested

* make fix-copies

* fix hidden states test and batching

---------
Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>