GitLab

* 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>

* Script & Manual edition

* Update

* Enable instantiating model with pretrained backbone weights

* Clarify pretrained import

* Use load_backbone instead

* Add backbone_kwargs to config

* Pass kwargs to constructors

* Fix up

* Input verification

* Add tests

* Tidy up

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

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

* fix / skip some tests before we can switch to torch 2.2

* style

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

* Port core files + ESM (because ESM code is odd)

* Search-replace in modelling code

* Fix up transfo_xl as well

* Fix other core files + tests (still need to add correct import to tests)

* Fix cookiecutter

* make fixup, fix imports in some more core files

* Auto-add imports to tests

* Cleanup, add imports to sagemaker tests

* Use correct exception for importing tf_keras

* Fixes in modeling_tf_utils

* make fixup

* Correct version parsing code

* Ensure the pipeline tests correctly revert to float32 after each test

* Ensure the pipeline tests correctly revert to float32 after each test

* More tf.keras -> keras

* Add dtype cast

* Better imports of tf_keras

* Add a cast for tf.assign, just in case

* Fix callback imports

* Add serialized type name to pytrees

* Modify context

* add serde test

* Enable instantiating model with pretrained backbone weights

* Update tests so backbone checkpoint isn't passed in

* Remove doc updates until changes made in modeling code

* Clarify pretrained import

* Update configs - docs and validation check

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

* Clarify exception message

* Update config init in tests

* Add test for when use_timm_backbone=True

* Small test updates

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

* improve dev setup comments and hints

* fix tests for new dev setup hints

* add check_support_list.py

* fix

* fix

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

* Iteratre over out_features instead of stage_names

* Update for all backbones

* Add tests

* Fix

* Align timm backbone behaviour with other backbones

* Fix tests

* Stricter checks on set out_features and out_indices

* Revert back stage selection logic

* Remove out-of-order logic

* Document restriction in docstrings

* More build_in_name_scope()

* Make sure we set the save spec now we don't do it with dummies anymore

* make fixup

* fix hf-internal-testing/fixtures_image_utils

* fix test

* comments

* add sdpa

* wip

* cleaning

* add ref

* yet more cleaning

* and more :)

* wip llama

* working llama

* add output_attentions=True support

* bigcode sdpa support

* fixes

* gpt-bigcode support, require torch>=2.1.1

* add falcon support

* fix conflicts falcon

* style

* fix attention_mask definition

* remove output_attentions from attnmaskconverter

* support whisper without removing any Copied from statement

* fix mbart default to eager renaming

* fix typo in falcon

* fix is_causal in SDPA

* check is_flash_attn_2_available in the models init as well in case the model is not initialized through from_pretrained

* add warnings when falling back on the manual implementation

* precise doc

* wip replace _flash_attn_enabled by config.attn_implementation

* fix typo

* add tests

* style

* add a copy.deepcopy on the config in from_pretrained, as we do not want to modify it inplace

* obey to config.attn_implementation if a config is passed in from_pretrained

* fix is_torch_sdpa_available when torch is not installed

* remove dead code

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

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

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

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

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

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

* remove duplicate pretraining_tp code

* add dropout in llama

* precise comment on attn_mask

* add fmt: off for _unmask_unattended docstring

* precise num_masks comment

* nuke pretraining_tp in LlamaSDPAAttention following Arthur's suggestion

* cleanup modeling_utils

* backward compatibility

* fix style as requested

* style

* improve documentation

* test pass

* style

* add _unmask_unattended tests

* skip meaningless tests for idefics

* hard_check SDPA requirements when specifically requested

* standardize the use if XXX_ATTENTION_CLASSES

* fix SDPA bug with mem-efficient backend on CUDA when using fp32

* fix test

* rely on SDPA is_causal parameter to handle the causal mask in some cases

* fix FALCON_ATTENTION_CLASSES

* remove _flash_attn_2_enabled occurences

* fix test

* add OPT to the list of supported flash models

* improve test

* properly test on different SDPA backends, on different dtypes & properly handle separately the pad tokens in the test

* remove remaining _flash_attn_2_enabled occurence

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

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

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

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

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

* remove use_attn_implementation

* fix docstring & slight bug

* make attn_implementation internal (_attn_implementation)

* typos

* fix tests

* deprecate use_flash_attention_2=True

* fix test

* add back llama that was removed by mistake

* fix tests

* remove _flash_attn_2_enabled occurences bis

* add check & test that passed attn_implementation is valid

* fix falcon torchscript export

* fix device of mask in tests

* add tip about torch.jit.trace and move bt doc below sdpa

* fix parameterized.expand order

* move tests from test_modeling_attn_mask_utils to test_modeling_utils as a relevant test class is already there

* update sdpaattention class with the new cache

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

* Update src/transformers/models/bark/modeling_bark.py

* address review comments

* WIP torch.jit.trace fix. left: test both eager & sdpa

* add test for torch.jit.trace for both eager/sdpa

* fix falcon with torch==2.0 that needs to use sdpa

* fix doc

* hopefully last fix

* fix key_value_length that has no default now in mask converter

* is it flacky?

* fix speculative decoding bug

* tests do pass

* fix following #27907

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

* update

* fix

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

* don't use `use_auth_token`internally

* let's use token everywhere

* fixup

* Safetensors serialization by default

* First pass on the tests

* Second pass on the tests

* Third pass on the tests

* Fix TF weight loading from TF-format safetensors

* Specific encoder-decoder fixes for weight crossloading

* Add VisionEncoderDecoder fixes for TF too

* Change filename test for pt-to-tf

* One missing fix for TFVisionEncoderDecoder

* Fix the other crossload test

* Support for flax + updated tests

* Apply suggestions from code review
Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

* Sanchit's comments

* Sanchit's comments 2

* Nico's comments

* Fix tests

* cleanup

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

---------
Co-authored-by: Matt <rocketknight1@gmail.com>
Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Register ModelOutput as supported torch pytree nodes

* Test ModelOutput as supported torch pytree nodes

* Update type hints for pytree unflatten functions

* fix

* fix

* Fix

* Fix

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

* Update tiny model summary file

* add to pipeline tests

* revert

* fix import

* fix import

* fix

* fix

* update

* update

* update

* fix

* remove BarkModelTest

* fix

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

* add kaldi fbank

* make style

* add herz_to_mel_kaldi tests

* add mel to hertz kaldi test

* integration tests

* correct test and remove comment

* make style

* Apply suggestions from code review
Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

* change parameter name

* Apply suggestions from Arthur review
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update remove_dc_offset description

* fix bug  + make style

* fix error in using np.exp instead of np.power

* make style

---------
Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Fix gated repo tests

* Apply suggestions from code review

Add missing Maskformer dataclass decorator, add dataclass check in ModelOutput for subclasses (#25638)

* Add @dataclass to MaskFormerPixelDecoderOutput

* Add dataclass check if subclass of ModelOutout

* Use unittest assertRaises rather than pytest per contribution doc

* Update src/transformers/utils/generic.py per suggested change
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

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

* fix

* fix

* fix

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

* support loading base64 images

* add test

* mention in docs

* remove the logging

* sort imports

* update error message

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

* restructure to catch base64 exception

* doesn't like the newline

* download files

* format

* optimize imports

* guess it needs a space?

* support loading base64 images

* add test

* remove the logging

* sort imports

* restructure to catch base64 exception

* doesn't like the newline

* download files

* optimize imports

* guess it needs a space?

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

Make CI less brittle

* Register ModelOutput subclasses as supported torch.utils._pytree nodes

Fixes #25357 where DDP with static_graph=True does not sync gradients when calling backward() over tensors contained in ModelOutput subclasses

* Add test for torch pytree ModelOutput serialization and deserialization

* Add timeout parameter to load_image function.

* Remove line.

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

* Add parameter to docs.

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

* update tiny_model_summary.json

* update

* update

* update

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

* fix

* fix

* fix

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

* fix

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

* Rename index.mdx to index.md

* With saved modifs

* Address review comment

* Treat all files

* .mdx -> .md

* Remove special char

* Update utils/tests_fetcher.py
Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>

---------
Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>

* fix

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

* Fix saved_model_creation_extended

* Skip the BLIP model creation test for now

* Fix TF SAM test

* Fix longformer tests

* Fix Wav2Vec2

* Add a skip for XLNet

* make fixup

* make fix-copies

* Add comments

* Add test for proper input signatures

* No more signature pruning

* Test the dummy inputs are valid too

* fine-tine -> fine-tune

* Fix indent in test_dataset_conversion

* Stop storing references to bound methods in tf.functions

* Remove the gc.collect calls now that we resolved the underlying problem

* Remove the default signature from model.serving entirely, big cleanup

* Remove _prune_signature as self.input_signature can prune itself

* Restore serving docstring

* Update int support test to check the input signature

* Make sure other tests also use model.input_signature and not serving.input_signature

* Restore _prune_signature

* Remove the doctest GC now it's no longer needed

* Correct core tests to use the pruned sig

* order lines correctly in core tests

* Add eager_serving back with a deprecation warning

* add trust_remote_code option

* require_torch