* change sequence_bias type of SequenceBiasLogitsProcessor tp list, add config tests for all processors
* fix format
* small fix for all_token_bias_pairs_are_valid internal func
* small typo fix in description
* improve test impl, some SequenceBiasLogitsProcessor refactoring
* add tests
* fix whisper
* update
* nit
* add qwen2-vl
* more updates!
* better this way
* fix this one
* fix more tests
* fix final tests, hope so
* fix led
* Update tests/generation/test_utils.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
* pr comments
* not pass pixels and extra for low-mem tests, very flaky because of visio tower
---------
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
* clean mimi commit
* some nits suggestions from Arthur
* make fixup
* rename repo id + change readme
* Update docs/source/en/model_doc/mimi.md
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
* add flaky flag to batching equivalence due to audio_codes failing sometimes
---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
* _decode signature change and quick return
* added bunch of decoding tests
* signature match and return
* added tests for decoding
* merged decoding test
* more tests for special tokens
* cosmetics
* fixed param
* ruffed the file
* refinement for single special tokens
* added test for single special tokens
* slight change to test name
Co-authored-by: Ita Zaporozhets <31893021+itazap@users.noreply.github.com>
* minor change test name for skip tokens
Co-authored-by: Ita Zaporozhets <31893021+itazap@users.noreply.github.com>
* killed already defined var
Co-authored-by: Ita Zaporozhets <31893021+itazap@users.noreply.github.com>
* minor update with vars
Co-authored-by: Ita Zaporozhets <31893021+itazap@users.noreply.github.com>
* killed already defined var once more
Co-authored-by: Ita Zaporozhets <31893021+itazap@users.noreply.github.com>
---------
Co-authored-by: Ita Zaporozhets <31893021+itazap@users.noreply.github.com>
* fix patch_attention_mask incorrect setting which leads to the difference in the generated text if batch > 1
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
* fix format
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
* [run_slow] idefics2
---------
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
* added sequences_scores to the output
* added beam_indices to output
* added test to check for beam_indices, sequences_scores and their shape
* removed redundant whitespaces
* make fixup
* idefics2 enable_input_require_grads not aligned with disable_input_require_grads
make peft+idefics2 checkpoints disable fail
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
* split test case
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
* fix ci failure
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
* refine test
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
---------
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
* refactor weight_norm + propose uniformed solution to reconcile meta load_state_dict with classic loading
* make style
* fix sew
* fix sew and sew_d tests
* Fix failing tensor placement in Whisper
* fix long form generation tests
* more return_timestamps=True
* make fixup
* [run_slow] whisper
* [run_slow] whisper
* Uniformize kwargs for LlaVa and update docs
* Change order of processor inputs in docstring
* Improve BC support for reversed images and text inputs
* cleanup llava processor call docstring
* Add encoded inputs as valid text inputs in reverse input check, add deprecation version in warning
* Put function check reversed images text outside base processor class
* Refactor _validate_images_text_input_order
* Add ProcessingUtilTester
* fix processing and test_processing
* initial commit
* gloups
* updates
* work
* weights match
* nits
* nits
* updates to support the tokenizer :)
* updates
* Pixtral processor (#33454)
* rough outline
* Add in image break and end tokens
* Fix
* Udo some formatting changes
* Set patch_size default
* Fix
* Fix token expansion
* nit in conversion script
* Fix image token list creation
* done
* add expected results
* Process list of list of images (#33465)
* updates
* working image and processor
* this is the expected format
* some fixes
* push current updated
* working mult images!
* add a small integration test
* Uodate configuration docstring
* Formatting
* Config docstring fix
* simplify model test
* fixup modeling and etests
* Return BatchMixFeature in image processor
* fix some copies
* update
* nits
* Update model docstring
* Apply suggestions from code review
* Fix up
* updates
* revert modeling changes
* update
* update
* fix load safe
* addd liscence
* update
* use pixel_values as required by the model
* skip some tests and refactor
* Add pixtral image processing tests (#33476)
* Image processing tests
* Add processing tests
* woops
* defaults reflect pixtral image processor
* fixup post merge
* images -> pixel values
* oups sorry Mr docbuilder
* isort
* fix
* fix processor tests
* small fixes
* nit
* update
* last nits
* oups this was really breaking!
* nits
* is composition needs to be true
---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
* fix long seq bug
* fixed format
* fixed fn copy inconsistency
* fix long seq bug
* fixed format
* fixed fn copy inconsistency
* Addressed comments
* added a unit test
* fixed cache position
* Added a warning msg to the forward fn
* fixed test case
* test(tokenizers): add a test showing conflict with sentencepiece
This is due to the fact that protobuf C implementation uses a global
pool for all added descriptors, so if two different files add
descriptors, they will end up conflicting.
* fix(tokenizers): mitigate sentencepiece/protobuf conflict
When sentencepiece is available, use that protobuf instead of the
internal one.
* chore(style): fix with ruff
* Update tokenization_whisper.py
Fix issue with flax whisper model
* Update tokenization_whisper_fast.py
Fix issue with flax whisper model
* Update tokenization_whisper.py
just check len of token_ids
* Update tokenization_whisper_fast.py
just use len of token_ids
* Update tokenization_whisper_fast.py and revert changes in _strip_prompt and add support to jax arrays in _convert_to_list
* Update tokenization_whisper.py and revert changes in _strip_prompt and add support to jax arrays in _convert_to_list
* Update test_tokenization_whisper.py to add test for _convert_to_list method
* Update test_tokenization_whisper.py to fix code style issues
* Fix code style
* Fix code check again
* Update test_tokenization)whisper.py to Improve code style
* Update test_tokenization_whisper.py to run each of jax, tf and flax modules if available
* Update tests/models/whisper/test_tokenization_whisper.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
* Update test_tokenization_whisper.py and use require_xxx decorators instead of `is_xxx_available()` method
* Revert the changes automatically applied by formatter and was unrelated to PR
* Format for minimal changes
---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
* add tests for linear shape behavior
* fix linear shape behavior
ended up adding the reshape at the end, after f8f8bf16_rowwise, because adding
it directly after quantize_fp8_per_row caused f8f8bf16_rowwise to drop the
seq_len dimension. (i.e., (17, 23, 1014) -> (17, 1024))
* save shape up front + comment
* Make StaticCache configurable at model construct time
* integrations import structure
* add new doc file to toc
---------
Co-authored-by: Guang Yang <guangyang@fb.com>
Co-authored-by: Joao Gante <joao@huggingface.co>
* Update docs for GGUF supported models
* Add tensor mappings and define class GGUFPhi3Converter
* Fix tokenizer
* Working version
* Attempt to fix some CI failures
* Run ruff format
* Add vocab, merges, decoder methods like LlamaConverter
* Resolve conflicts since Qwen2Moe was added to gguf
- I missed one place when resolving conflict
- I also made a mistake with tests_ggml.py and now has been fixed to reflect
its master version.
* Import structure & first three model refactors
* Register -> Export. Export all in __all__. Sensible defaults according to filename.
* Apply most comments from Amy and some comments from Lucain
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
Co-authored-by: Lucain Pouget <lucainp@gmail.com>
* Style
* Add comment
* Clearer .py management
* Raise if not in backend mapping
* More specific type
* More efficient listdir
* Misc fixes
---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
Co-authored-by: Lucain Pouget <lucainp@gmail.com>
* add self.head_dim for VisionAttention in Qwen2-VL
* add self.head_dim for VisionAttention in Qwen2-VL
* fix ci
* black the test_modeling_qwen2_vl.py
* use ruff to format test_modeling_qwen2_vl.py
* [run-slow] qwen2_vl
* use tying for python3.8
* fix the import format
* use ruff to fix the ci error I001
* [run-slow] qwen2_vl
* remove unused import
* commit for rebase
* use ruff fix ci
* [run-slow] qwen2_vl
---------
Co-authored-by: root <liji>
* Add validation for maximum sequence length in modeling_whisper.py
Added a validation check to ensure that the sequence length of labels does not exceed the maximum allowed length of 448 tokens. If the sequence length exceeds this limit, a ValueError is raised with a descriptive error message.
This change prevents the model from encountering errors or unexpected behavior due to excessively long sequences during training or fine-tuning, ensuring consistent input dimensions and improving overall robustness.
* Change exception message in src/transformers/models/whisper/modeling_whisper.py
The exception message is for whisper's label's sequence max length.
Co-authored-by: Yoach Lacombe <52246514+ylacombe@users.noreply.github.com>
* Change 448 to config.max_target_positions in src/transformers/models/whisper/modeling_whisper.py
It's for whisper's config.max_target_positions.
Co-authored-by: Yoach Lacombe <52246514+ylacombe@users.noreply.github.com>
* Change method's documentation in src/transformers/models/whisper/modeling_whisper.py
* Add test for maximum label's sequence length in test_modeling_whisper.py
* Add self to modeling_whisper.py
* Update test_modeling_whisper.py with respect to automatic validations
* Update modeling_whisper.py with respect to ci/circleci: check_code_quality
* Update test_modeling_whisper.py with respect to ci/circleci: check_code_quality
* Update test_modeling_whisper.py with respect to ci/circleci: tests_generate
* Update test_modeling_whisper.py with respect to ci/circleci: tests_generate
* Update test_modeling_whisper.py with respect to ci/circleci: check_code_quality
* Separate test_labels_sequence_max_length tests in test_modeling_whisper.py
* Update test_modeling_whisper.py with respect to ci/circleci: check_code_quality
* Remove assert from test_modeling_whisper.py
* Add max_target_positions to WhisperModelTester in test_modeling_whisper.py
* Update test_modeling_whisper.py with respect to ci/circleci: check_code_quality
* Update test_modeling_whisper.py with respect to ci/circleci: tests_generate
* Update test_modeling_whisper.py
* Change test_labels_sequence_max_length_error_after_changing_config in test_modeling_whisper.py
* Change self.config.max_target_positions to self.max_target_positions modeling_whisper.py
* Add new tests in test_modeling_whisper.py
* Update test_modeling_whisper.py
---------
Co-authored-by: Yoach Lacombe <52246514+ylacombe@users.noreply.github.com>
* Load remote code only once
* Use hash as load indicator
* Add a new option `force_reload` for old behavior (i.e. always reload)
* Add test for dynamic module is cached
* Add more type annotations to improve code readability
* Address comments from code review
* Add validate images and test processing utils
* Remove encoded text from possible inputs in tests
* Removed encoded inputs as valid in processing_utils
* change text input check to be recursive
* change text check to all element of lists and not just the first one in recursive checks
* [InstructBLIP] qformer_tokenizer is required input
* Bit safer
* Add to instructblipvideo processor
* Fix up
* Use video inputs
* Update tests/models/instructblipvideo/test_processor_instructblipvideo.py
* Fixing a bug in the way "attention_factor" is validated in ROPE utilities.
* Fixing a bug in the way "attention_factor" is validated in ROPE utilities.
* Fixing a bug in the way "attention_factor" is validated in ROPE utilities.
* use gguf internal dequantize
* add Q5_0 test
* add iq1 test
* add remained test
* remove duplicated test
* update docs
* add gguf version limit
* make style
* update gguf import catch
* revert vocab_size patch
* make style
* use GGUF_MIN_VERSION everywhere
* remove to restiction for 4-bit model
* Update src/transformers/modeling_utils.py
Co-authored-by: Matthew Douglas <38992547+matthewdouglas@users.noreply.github.com>
* bitsandbytes: prevent dtype casting while allowing device movement with .to or .cuda
* quality fix
* Improve warning message for .to() and .cuda() on bnb quantized models
---------
Co-authored-by: Matthew Douglas <38992547+matthewdouglas@users.noreply.github.com>
* don't run custom when not needed?
* update test fetcher filtering
* fixup and updates
* update
* update
* reduce burden
* nit
* nit
* mising comma
* this?
* this?
* more parallelism
* more
* nit for real parallelism on tf and torch examples
* update
* update
* update
* update
* update
* update
* update
* update
* update
* update
* update
* update
* update to make it more custom
* update to make it more custom
* update to make it more custom
* update to make it more custom
* update
* update
* update
* update
* update
* update
* use correct path
* fix path to test files and examples
* filter-tests
* filter?
* filter?
* filter?
* nits
* fix naming of the artifacts to be pushed
* list vs files
* list vs files
* fixup
* fix list of all tests
* fix the install steps
* fix the install steps
* fix the config
* fix the config
* only split if needed
* only split if needed
* extend should fix it
* extend should fix it
* arg
* arg
* update
* update
* run tests
* run tests
* run tests
* more nits
* update
* update
* update
* update
* update
* update
* update
* simpler way to show the test, reduces the complexity of the generated config
* simpler way to show the test, reduces the complexity of the generated config
* style
* oups
* oups
* fix import errors
* skip some tests for now
* update doctestjob
* more parallelism
* fixup
* test only the test in examples
* test only the test in examples
* nits
* from Arthur
* fix generated congi
* update
* update
* show tests
* oups
* oups
* fix torch job for now
* use single upload setp
* oups
* fu**k
* fix
* nit
* update
* nit
* fix
* fixes
* [test-all]
* add generate marker and generate job
* oups
* torch job runs not generate tests
* let repo utils test all utils
* UPdate
* styling
* fix repo utils test
* more parallel please
* don't test
* update
* bit more verbose sir
* more
* hub were skipped
* split by classname
* revert
* maybe?
* Amazing catch
Co-authored-by: Yih-Dar <2521628+ydshieh@users.noreply.github.com>
* fix
* update
* update
* maybe non capturing
* manual convert?
* pass artifacts as parameters as otherwise the config is too long
* artifact.json
* store output
* might not be safe?
* my token
* mmm?
* use CI job IS
* can't get a proper id?
* ups
* build num
* update
* echo url
* this?
* this!
* fix
* wget
* ish
* dang
* udpdate
* there we go
* update
* update
* pass all
* not .txt
* update
* fetcg
* fix naming
* fix
* up
* update
* update
* ??
* update
* more updates
* update
* more
* skip
* oups
* pr documentation tests are currently created differently
* update
* hmmmm
* oups
* curl -L
* update
* ????
* nit
* mmmm
* ish
* ouf
* update
* ish
* update
* update
* updatea
* nit
* nit
* up
* oups
* documentation_test fix
* test hub tests everything, just marker
* update
* fix
* test_hub is the only annoying one now
* tf threads?
* oups
* not sure what is happening?
* fix?
* just use folder for stating hub
* I am getting fucking annoyed
* fix the test?
* update
* uupdate
* ?
* fixes
* add comment!
* nit
---------
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
Co-authored-by: Yih-Dar <2521628+ydshieh@users.noreply.github.com>
* Adding SDPA support for RoBERTa-based models
* add not is_cross_attention
* fix copies
* fix test
* add minimal test for camembert and xlm_roberta as their test class does not inherit from ModelTesterMixin
* address some review comments
* use copied from
* style
* consistency
* fix lists
---------
Co-authored-by: fxmarty <9808326+fxmarty@users.noreply.github.com>
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
* init fix
* fix mask during cached forward, move mask related stuff to own function
* adjust tests as left padding does not change logits as much anymore + batch gen (with todo on logits comp)
* revert overwriting new integration tests
* move some comments to docstring
* add Blip2ForImageTextRetrieval
* use one line and remove unnecessary space in tests
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
* use value from the config, rather than hardcoded
* change order of params in Blip2QFormerModel.forward
* update docstring
* fix style
* update test_inference_opt
* move embeddings out of Blip2QFormerModel
* remove from_vision_qformer_configs
* remove autocast float16 in Blip2QFormerModel
* rename fiels into vision_projection,text_projection,use_image_text_matching_head
* use CLIPOutput for Blip2ImageTextMatchingModelOutput
* remove past_key_values_length from Blip2TextEmbeddings
* fix small typo in the CLIPOutput docstring
* add Blip2ForImageTextRetrieval to Zero Shot Image Classification mapping
* update docstring and add require_torch_fp16
* rollback test_inference_opt
* use use_image_text_matching_head=True in convert
* skip test_model_get_set_embeddings
* fix create_rename_keys error on new itm fields
* revert to do scale after dot product between "query" and "key"
* fix ValueError on convert script for blip2-opt-2.7b
* update org of paths to Salesforce
* add is_pipeline_test_to_skip for VisualQuestionAnsweringPipelineTests
* [run_slow] blip_2
* removed Blip2ForImageTextRetrieval from IGNORE_NON_AUTO_CONFIGURED
* fix docstring of Blip2ImageTextMatchingModelOutput
* [run_slow] blip_2
* fix multi-gpu tests
* [run_slow] blip_2
* [run_slow] blip_2
---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
* Add a fix for the case when tokenizers are passed as a string
* Support image processors and feature extractors as well
* Reverting load_feature_extractor and load_image_processor
* Add test
* Test is torch-only
* Add tests for preprocessors and feature extractors and move test
* Extremely experimental fix
* Revert that change, wrong branch!
* Typo!
* Split tests
* fix param not being passed in tested; add exceptions
* better source of model name
* Update utils/create_dummy_models.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
* fix: multilingual midel convert to tflite get wrong token
* fix: modify test_force_tokens_logits_processor the checking value as scores.dtype.min
---------
Co-authored-by: kent.sc.hung <kent.sc.hung@benq.com>
Co-authored-by: Aya <[kent831217@gmail.com]>
* Add new Jinja features:
- Do extension
- Break/continue in loops
- Call strftime to get current datetime in any format
* Add new Jinja features:
- Do extension
- Break/continue in loops
- Call strftime to get current datetime in any format
* Fix strftime template
* Add template strip() just to be safe
* Remove the do extension to make porting easier, and also because it's the least useful
* Rename test
* strftime -> strftime_now
* Split test
* Update test to use strftime_now
* Refactor everything out into chat_template_utils
* Refactor everything out into chat_template_utils
* Refactor everything out into chat_template_utils
* Refactor everything out into chat_template_utils
* Refactor everything out into chat_template_utils
* Add .float() in all generation methods logit outputs
* Switch float-casting of logits to training only for main models
* Add `num_logits_to_keep` in Llama and add it by default in generate
* Apply style
* Add num_logits_to_keep as arg in prepare_input_for_generation
* Add support for Mistral
* Revert models except llama and mistral
* Fix default None value in _supports_num_logits_to_keep()
* Fix dimension of dummy input
* Add exception for prophetnet in _supports_num_logits_to_keep()
* Update _supports_num_logits_to_keep() to use inspect.signature()
* Add deprecation cycle + remove modification with pretraining_tp
* Apply style
* Add most used models
* Apply style
* Make `num_logits_to_keep` an int in all cases to remove if-else clause
* Add compile check for the warning
* Fix torch versions
* style
* Add gemma2
* Update warning version
* Add comment about .float operations in generation utils
* Add tests in GenerationTesterMixin and ModelTesterMixin
* Fix batch size for assisted decoding in tests
* fix small issues in test
* refacor test
* fix slicing removing dim issue
* Add nemotron support (should fix check-copy issue in CIs)
* Trigger new CIs
* Trigger new CIs
* Bump version
* Bump version in TODO
* Trigger CIs
* remove blank space
* Trigger CIs
* more precise name
* better docstrings
* Update src/transformers/cache_utils.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
---------
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
* Add TorchAOHfQuantizer
Summary:
Enable loading torchao quantized model in huggingface.
Test Plan:
local test
Reviewers:
Subscribers:
Tasks:
Tags:
* Fix a few issues
* style
* Added tests and addressed some comments about dtype conversion
* fix torch_dtype warning message
* fix tests
* style
* TorchAOConfig -> TorchAoConfig
* enable offload + fix memory with multi-gpu
* update torchao version requirement to 0.4.0
* better comments
* add torch.compile to torchao README, add perf number link
---------
Co-authored-by: Marc Sun <marc@huggingface.co>
* Add padding="max_length" to tokenizer kwargs and change crop_size to size for image_processor kwargs
* remove crop_size argument in align processor tests to be coherent with base tests
* Add pad_token when loading tokenizer if needed, change test override tokenizer kwargs, remove unnecessary test overwrites in grounding dino
* fix typo
* uniform kwargs
* make style
* add comments
* remove return_tensors
* remove common_kwargs from processor since it propagates
* make style
* return_token_type_ids to True
* revert the default imagekwargs since does not accept any value in the image processro
* revert processing_utils.py
* make style
* add molbap's commit
* fix typo
* fix common processor
* remain
* Revert "add molbap's commit"
This reverts commit a476c6ee88.
* add unsync PR
* revert
* make CI happy
* nit
* import annotationformat
* add new model like
* draft cuda forward - mismatched keys (sharding on conv1)
* match keys successfully
* fix split
* get generation/forward running (wrong gens, norm?)
* :update
* some refactoring
* fixes
* works up until copy to cache
* fix
* update
* NON WORKING VERSION
* version that work?
* nit
* fix config
* fix conversion script
* working cuda forward
* nit
* update
* simplifcation
* make mamba slow simple work
* no einops
* todo
* fix style
* no einops
* update fix no einsum
* nit
* remove einops
* bug: scan_output differs strongly
* add rms norm option
* fix fast + slow generation with and w/o cache ✔️
* draft integration tests
* remove a big chunk of the einsum
* fix slow, fast generations, without any einsum
* fix copies
* fix structure
* fix up modeling and tests
* fix tests
* clamping is indeed worse
* recover mamba2 cache test
* fix copies
* no cache position (yet)
* fix tf tests
* fix matmul for generate
* fixup
* skip cache tests for now
* [run-slow]mamba2
* tune out hidden states for padding
* test batched generation
* propagate attention mask changes
* fix past length
* fix integration test
* style
* address comments
* update readme
* add mamba2 version check
* fix tests
* [run-slow]mamba2
* skip edge tests
* [run-slow]mamba2
* last fixup
* [run-slow]mamba2
* update README
---------
Co-authored-by: Arthur Zucker <arthur.zucker@gmail.com>
* save total_vocab_size = vocab_size + user added tokens to speed up operation
* updating length when added_tokens_decoder is set
* add test len(tokenizer)
* Test this zach
* Test for improper init w/o zero3
* Move back
* Apply suggestions from code review
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
* Get rid of stars in warning
* Make private
* Make clear
---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>