GitLab

* fix special tokens in generatioon

* fix test

* add warning

* fix the check

* warn once

* fix

* make mamba use cache

* uss cache naming as in mamba

* fix musicgen

* SwitchTransformer MoE layer performance improvement

* make fixup

* comments about shapes

* make fixup

no need for explicit EXTRA_TOKENS

Skip failing tests for now

* Fix contrastive_search for new cache structure, and improve performance by removing inneficient torch.stack(torch.split(x, top_k, dim=0))

* Fix _contrastive_search for non-standard cache using ellipsis slicing

* Fix all outputs.logits memory leaks for all decoding strategies!

* Fix small error in _contrastive_search()

* Make all necessary change and revert for the new class

* Apply coding style

* Remove pipes in type hints for compatibility

* correct type hint

* apply style

* Use DynamicCache by default and solve conflicts

* Fix rebase issues

* Add `_supports_dynamic_cache_class` in models for models that support DynamicCache but not other caches to make DynamicCache the default for more models

* Create generation config to return legacy format by default, or to choose not to

* style

* Fix case when use_cache is False

* Remove default DynamicCache in assiste_decoding if assistant_model does not support it + fix _seen_tokens when cropping cache

* Update prepare_inputs_for_generation() for case with empty DynamicCache

* Correct return of args in _assisted_decoding

* Remove EfficientDynamicCache as it is no longer needed

* Correct mistake in generation config

* Move cache logic of assisted decoding to AssistedCandidateGenerator.__init__

* change DynamicCache function names from "split" to "batch_split" for readability + apply coding style

* Remove `_supports_dynamic_cache_class` attribute after rebase

* Correct missing line lost in conflict resolution during rebasing

* Add special case for Jamba

* Fix jamba test

* Coding style

* coding style

* Correct missing import in rebasing

* Simplify _validate_model_kwargs based on removal of _supports_dynamic_cache attribute

* Simplify code paths in _contrastive_search

* coding style

* Update docstrings of cache methods

* Update prepare_inputs_for_generation() -> past_key_values are always Cache objects

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

* add flaubert tokenization test, enrich inheritance in FlaubertTokenizer.

* fix quality code ci

* ensure parameter consistency

* fix ci

* fix copyright year and flatten vocab list.

* fix style

* doc: add info about wav2vec2 bert in older wav2vec2 models.

* apply suggestions from review.

* forward contrib credits from review

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

Bump transformers in /examples/research_projects/deebert

Bumps [transformers](https://github.com/huggingface/transformers) from 3.5.1 to 4.38.0.
- [Release notes](https://github.com/huggingface/transformers/releases)
- [Commits](https://github.com/huggingface/transformers/compare/v3.5.1...v4.38.0

)

---
updated-dependencies:
- dependency-name: transformers
  dependency-type: direct:production
...
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Move label validation checks - fail early

* Remove some formatting changes - add back labels change wav2vec2

* benchmark workflow

* benchmark workflow

* benchmark workflow

* benchmark workflow

* build

* build

* build

* build

* build

* build

* build

* build

* build

* build

* build

* build

* build

* build

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

Fixed torch definition error

The StoppingCriteriaList allocates is_done without specifying dtype=torch.bool. On XLA this allocates a float tensor and causes a failure on the following line:

is_done = is_done | criteria(input_ids, scores, **kwargs)

by attempting to OR float with bool.

Bump transformers in /examples/research_projects/vqgan-clip

Bumps [transformers](https://github.com/huggingface/transformers) from 4.26.0 to 4.38.0.
- [Release notes](https://github.com/huggingface/transformers/releases)
- [Commits](https://github.com/huggingface/transformers/compare/v4.26.0...v4.38.0

)

---
updated-dependencies:
- dependency-name: transformers
  dependency-type: direct:production
...
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* build

* build

* build

* build

* fix

* fix

* fix

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

* Move out common validation

* Add missing backbone config arguments

* deprecate blip

* mention deprecation on docs

* fix

* fix

* fix

* fix

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

✨

 Add new line switch before logging "***** Running {description} *****".
Signed-off-by: jacklanda <yonyonlau@gmail.com>

* Fix pipeline tests - torch imports

* Frameowrk dependant float conversion

* fix logits dtype

* Add bf16/fp16 tests for text_classification pipeline

* Update test_pipelines_text_classification.py

* fix

* fix

* Added interpolate pos encoding feature and test to deit

* Added interpolate pos encoding feature and test for deit TF model

* readded accidentally delted test for multi_gpu

* storing only patch_size instead of entire config and removed commented code

* Update modeling_tf_deit.py to remove extra line
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

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

video-llava can handle more frames

* fix(PatchTST): Wrong dropout used for PretainHead

* feat(PatchTST): remove unused config.dropout

---------
Co-authored-by: Strobel Maximilian (IFAG PSS SIS SCE ACM) <Maximilian.Strobel@infineon.com>

* add device in logits processor

* remove device when not needed

* codestyle

* tests

* forgot `melody` version

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

* codestyle

* updates

---------
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* add tokenizer_summary to es/_toctree.yml

* add tokenizer_summary to es/

* fix link to Transformes XL in en/

* translate until Subword tokenization section

* fix GPT link in en/

* fix other GPT link in en/

* fix typo in en/

* translate the doc

* run make fixup

* Remove .md in Transformer XL link

* fix some link issues in es/

* fix typo

* build

* build

* build

* build

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

* update to not(endswith(loss))

* ruff formatting

Update modeling_cohere.py

Update index.md - Contents = Contenu

French typo -
Contents = Contenu

* Rename sanity_evaluation to eval_on_start

* move arg back to last

fix typo

* fix the get_size_with_aspect_ratio in max_size situation

* make fix-up

* add more general solution

* consider when max_size is not defined

* fix typo

* fix typo

* simple fix

* fix error

* fix if else error

* fix error of size overwrite

* fix yolos image processing

* fix detr image processing

* make

* add longest related test script

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

* add more test

* add test script about longest size

* remove deprecated

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

* add qwen2 gguf support

* Update docs

* fix qwen2 tokenizer

* add qwen2 gguf test

* fix typo in qwen2 gguf test

* format code

* Remove mistral, clarify the error message

* format code

* add typing and update docstring