GitLab

* Update trainer.py

* format

update docs around mixing hf scheduler with deepspeed optimizer

Update modeling_utils.py

* v1

* add docstring

* add tests

* add awq 0.1.8

* oops

* fix test

* fix llava index errors

* forward contrib credits from original implementation and fix

* better fix

* final fixes and fix all tests

* fix

* fix nit

* fix tests

* add regression tests

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

Co-authored-by: yudong.lin <yudong.lin@funplus.com>

* fix: minor enhancement and fix in bounding box visualization example

The example that was trying to visualize the bounding box was not considering an edge case,
where the bounding box can be un-normalized. So using the same set of code, we can not get
results with a different dataset with un-normalized bounding box. This commit fixes that.

* run make clean

* add an additional note on the scenarios where the box viz code works

---------
Co-authored-by: Anindyadeep <anindya@pop-os.localdomain>

* fix frames

* use smaller chunk length

* correct beam search + tentative stride

* fix whisper word timestamp in batch

* add test batch generation with return token timestamps

* Apply suggestions from code review
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

* clean a test

* make style + correct typo

* write clearer comments

* explain test in comment

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

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

* add check_support_list.py

* fix

* fix

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

Bug: `training_args.py` fix missing import with accelerate with version `accelerate==0.20.1` (#28171)

* fix-accelerate-version

* updated with exported ACCELERATE_MIN_VERSION,

* update string in ACCELERATE_MIN_VERSION

* First draft

* More improvements

* More improvements

* Make all tests pass

* Remove script

* Update image processor

* Address comments

* Use new gradient checkpointing method

* Convert checkpoints, add integration test

* Do not keep aspect ratio for now

* Set keep_aspect_ratio=False for beit, add integration test

* Remove print statement

Fix: [SeamlessM4T - S2TT] Bug in batch loading of audio in torch.Tensor format in the SeamlessM4TFeatureExtractor class (#27914)

* fixes: code fixes on is_batched condition to also check for batched audio data in torch.Tensor format instead of only just checking for batched audio data in np.ndarray format

* Update src/transformers/models/seamless_m4t/feature_extraction_seamless_m4t.py
Co-authored-by: Yoach Lacombe <52246514+ylacombe@users.noreply.github.com>

* refactor: code refactoring to remove torch framework dependency

* docs: updated docstring to add torch tensor compatibility

* test: add test cases to incorporate torch tensor inputs

* test: ran make fix-copies for code conformity

* test: refactor test to separate the test_call into test_call_numpy and test_call_torch

---------
Co-authored-by: Yoach Lacombe <52246514+ylacombe@users.noreply.github.com>

* normalize reverse indexing for causal lm sequence classifiers

* normalize reverse indexing for causal lm sequence classifiers

* normalize reverse indexing for causal lm sequence classifiers

* use modulo instead

* unify modulo-based sequence lengths

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

* add 4-bit serialization docs

* up

* up

Update test values

Indices must match stage name ordering

* Fix vision text dual encoder

* Small cleanup for wav2vec2 (not fixed yet)

* Small fix for vision_encoder_decoder

* Fix SAM builds

* Update TFBertTokenizer test with modern exporting + tokenizer

* Fix DeBERTa

* Fix DeBERTav2

* Try RAG fix but it's impossible to test locally

* Actually fix RAG now that I got FAISS working somehow

* Fix Wav2Vec2, add sermon

* Fix Hubert

* some nits

* update test

* add support d\sd[a

* remove some dummy inputs

* all good

* style

* nits

* fixes

* fix more copies

* nits

* styling

* fix

* Update src/transformers/models/mistral/modeling_mistral.py
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* add a slow test just to be sure

* fixup

---------
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* [Whisper] Use torch for stft if available

* update docstring

* mock patch decorator

* fit on one line

* updated bitsandbytes.py

* rm test_raise_* from test_4bit.py

* add test_4bit_serialization.py

* modeling_utils bulk edits

* bnb_ver 0.41.3 in integrations/bitsandbytes.py
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* @slow reinstated
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* bnb ver 0.41.3 in  src/transformers/modeling_utils.py
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* rm bnb version todo in  integrations/bitsandbytes.py
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* moved 4b serialization tests to test_4bit

* tests upd for opt

* to torch_device
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* ruff fixes to tests

* rm redundant bnb version check in mod_utils
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* restore _hf_peft_config_loaded  modeling_utils.py::2188
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* restore _hf_peft_config_loaded  test in modeling_utils.py::2199
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* fixed NOT getattr(self, "is_8bit_serializable")
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* setting model.is_4bit_serializable

* rm separate fp16_statistics arg from set_module...

* rm else branch in integrations::bnb::set_module

* bnb 4bit dtype check

* upd comment on 4bit weights

* upd tests for FP4 safe

---------
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

disable retain_grad_hidden_states_attentions on SeamlessM4TModelWithTextInputTest

* Fix yolos resizing

* Update tests

* Add a test

Co-authored-by: Merve Noyan <merveenoyan@gmail.com>

* fsdp, debugging, gpu selection

* fix hfoption

* fix

* 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

* Update FA2 exception msg to point to hub discussions

* Use path for hub url

* fix

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

* fix

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

* move code to Trainer.evaluate to enable use of that function with multiple datasets

* test

* update doc string

* and a tip

* forgot the type

---------
Co-authored-by: Prof. Peter Schneider-Kamp <jps@ordbogen.com>

* add attention_bias hparam for a model trained without attention biases

* fix argument documentation error

* fix fa2

* fix FA2 for popular models

* improve warning and add Younes as co-author
Co-Authored-By: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

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

* fix the warning

* Add Tip

* typo fix

* nit

---------
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

Signed-off-by: Abolfazl Shahbazi <abolfazl.shahbazi@intel.com>

Fix mistral link in mixtral.md

In docstring for PreTrainedModel.resize_token_embeddings, correct definition of new_num_tokens parameter to read "the new number of tokens" (meaning the new size of the vocab) rather than "the number of new tokens" (number of newly added tokens only).

* default config should not use sliding window

* update the doc

* nits

* add a proper test

* update

* update

* update expected value

* Update src/transformers/tokenization_utils_fast.py
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* convert to float

* average then N**2

* comment

* revert nit

* good to fo

* fixup

* Update tests/models/mixtral/test_modeling_mixtral.py
Co-authored-by: Lysandre Debut <hi@lysand.re>

* revert unrelated change

---------
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>
Co-authored-by: Lysandre Debut <hi@lysand.re>

* speculative decoding

* fix test

* space

* better comments

* remove redundant test

* test nit

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

* PR comments

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