Daniel Bogdoll
de8a0b7547
Option to set 'non_blocking' for to(device) in BatchEncoding and BatchFeature ( #34883 )
...
* Option to set 'non_blocking' for to(device) operation for performance improvements. Defaults to 'false', thus no behavioral changes.
* Enabling non_blocking in to() operation of BatchFeature.
* Improved docstring on utilization of non_blocking
* Force non_blocking as keyword argument
Co-authored-by: Pavel Iakubovskii <qubvel@gmail.com>
---------
Co-authored-by: Daniel Bogdoll <dbogdoll@umich.edu>
Co-authored-by: Pavel Iakubovskii <qubvel@gmail.com>
2024-12-09 11:29:04 +01:00
UV
1452dc2514
Corrected typo in agent system prompts ( #35143 )
2024-12-09 10:42:23 +01:00
NielsRogge
9e420e0269
[I-JEPA] Update docs ( #35148 )
...
Update docs
2024-12-09 10:01:31 +01:00
kang sheng
1ccca8f48c
Fix GA loss bugs and add unit test ( #35121 )
...
* fix GA bugs and add unit test
* narrow down model loss unit test diff gap
* format code to make ruff happy
* send num_items_in_batch argument to decoder
* fix GA loss bug in BertLMHeadModel
* use TinyStories-33M to narrow down diff gap
* fotmat code
* missing .config
* avoid add extra args
---------
Co-authored-by: kangsheng <kangsheng@meituan.com>
2024-12-09 09:57:41 +01:00
Pavel Iakubovskii
c8c8dffbe4
Update I-JEPA checkpoints path ( #35120 )
...
Update checkpoints path
2024-12-06 13:42:51 +00:00
Victor Agostinelli
7f95372c62
Add feature dim attributes to BitLinear for easier PEFT integration ( #34946 )
...
Update bitnet.py, extremely small change to allow for easier PEFT integration
Co-authored-by: Mohamed Mekkouri <93391238+MekkCyber@users.noreply.github.com>
2024-12-06 13:39:45 +01:00
Aymeric Roucher
9ad4c93536
Add Aria ( #34157 )
...
* Add Aria
---------
Co-authored-by: Cyril Vallez <cyril.vallez@gmail.com>
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
2024-12-06 12:17:34 +01:00
Yih-Dar
15ab310c3a
Fix private forked repo. CI ( #35114 )
...
fix
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2024-12-06 12:03:31 +01:00
Steven Liu
98e8062df3
[docs] top_p, top_k, temperature docstrings ( #35065 )
...
clarify
2024-12-05 11:24:51 -08:00
Jacky Lee
44f88d8ccb
[docs] Update Python version in translations ( #35096 )
...
update: doc version
2024-12-05 11:06:54 -08:00
Lysandre
66ab300aaf
Dev version
2024-12-05 19:12:22 +01:00
Pablo Montalvo
a5bb528471
Fix signatures for processing kwargs ( #35105 )
...
* add conversion script
* remove pg2 refs
* fixup style
* small update
* get correct scaling
* add back missing bos
* fix missing config keys
* might revert this pos_embeddings
* fixup 9b config
* fix 9b
* fixup 9b conversion for good + add back num_hidden_layers
* add correct query scaling for 2b, 9b, 27b
* fixup 27b conversion
* Additional variant: 27b-896
* Use CPU for conversion to reduce GPU RAM requirements
* fix causal mask generation + formatting
* fix in-training causal mask generation edge case
* trigger CI
* update config
* update config
* update config
* update config
* update config
* update config
* update config
* update config
* update config
* move conversion file to main model dir
* handle multi-images + bos token
* address comments for input ids
* revert ci fixes
* [run-slow] paligemma
* fix
* [run-slow] paligemma
* skip end 2 end
* [run-slow] paligemma
---------
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2024-12-05 18:15:48 +01:00
Jonathan Mamou
e27465c801
Adaptive dynamic number of speculative tokens ( #34156 )
...
* initial commit
* update strategy
* add tradeoff FPR TPR with cost
* all probs
* fix
* fix
* fix style
* Update src/transformers/generation/configuration_utils.py
shorter docstring
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
* import guard
* fix style
* add is_sklearn_available condition
* vectorizing to flatten the for-loop
* fix style
* disable adaptation for UAG
* update doc
* add TestAssistedCandidateGeneratorUpdateStrategy
* fix style
* protect import
* fix style
---------
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
2024-12-05 17:07:33 +01:00
Yih-Dar
b0a51e5cff
Fix flaky Hub CI (test_trainer.py) ( #35062 )
...
* fix
* Update src/transformers/testing_utils.py
Co-authored-by: Lucain <lucainp@gmail.com>
* fix
* fix
* fix
* fix
* fix
* fix
* fix
* fix
* check
* check
* check
* check
* check
* check
* Update src/transformers/testing_utils.py
Co-authored-by: Lucain <lucainp@gmail.com>
* Update src/transformers/testing_utils.py
Co-authored-by: Lucain <lucainp@gmail.com>
* check
* check
* check
* Final space
* Final adjustment
---------
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
Co-authored-by: Lucain <lucainp@gmail.com>
2024-12-05 17:02:27 +01:00
Arthur
a928d9c128
[trainer] fix the GA model_accepts_loss_kwargs ( #34915 )
...
* fix
* style
* values
* fix
2024-12-05 16:37:46 +01:00
Raushan Turganbay
e682c17e4a
BLIP: this is correct now ( #35081 )
...
this is correct now
2024-12-05 16:30:09 +01:00
João Marcelo
50189e36a6
Add I-JEPA ( #33125 )
...
* first draft
* add IJepaEmbeddings class
* fix copy-from for IJepa model
* add weight conversion script
* update attention class names in IJepa model
* style changes
* Add push_to_hub option to convert_ijepa_checkpoint function
* add initial tests for I-JEPA
* minor style changes to conversion script
* make fixup related
* rename conversion script
* Add I-JEPA to sdpa docs
* minor fixes
* adjust conversion script
* update conversion script
* adjust sdpa docs
* [run_slow] ijepa
* [run-slow] ijepa
* [run-slow] ijepa
* [run-slow] ijepa
* [run-slow] ijepa
* [run-slow] ijepa
* formatting issues
* adjust modeling to modular code
* add IJepaModel to objects to ignore in docstring checks
* [run-slow] ijepa
* fix formatting issues
* add usage instruction snippet to docs
* change pos encoding, add checkpoint for doc
* add verify logits for all models
* [run-slow] ijepa
* update docs to include image feature extraction instructions
* remove pooling layer from IJepaModel in image classification class
* [run-slow] ijepa
* remove pooling layer from IJepaModel constructor
* update docs
* [run-slow] ijepa
* [run-slow] ijepa
* small changes
* [run-slow] ijepa
* style adjustments
* update copyright in init file
* adjust modular ijepa
* [run-slow] ijepa
2024-12-05 16:14:46 +01:00
Mohamed Mekkouri
95a855e212
Deprecate quanto and switch to optimum-quanto ( #35001 )
...
* deprecate quanto
* fix style
2024-12-05 16:11:09 +01:00
Isotr0py
482cb28a18
Fix tie_word_embeddings handling for GGUF models ( #35085 )
...
* fix tie_word_embeddings
Signed-off-by: Isotr0py <2037008807@qq.com>
* fix
Signed-off-by: Isotr0py <2037008807@qq.com>
---------
Signed-off-by: Isotr0py <2037008807@qq.com>
2024-12-05 16:00:41 +01:00
Cyril Vallez
35447054f5
Update Mistral conversion script ( #34829 )
...
* Update convert_mistral_weights_to_hf.py
* Update convert_mistral_weights_to_hf.py
* Update convert_mistral_weights_to_hf.py
2024-12-05 15:47:20 +01:00
Arthur
93f87d3cf5
[tokenizers] bump to 0.21 ( #34972 )
...
bump to 0.21
2024-12-05 15:46:02 +01:00
eustlb
54aae121eb
[Whisper] Fix whisper tokenizer ( #34537 )
...
* handle single timestamp ending
* include last timestamp token
* handle single timestamp ending
* avoid floating points arithm limitations
* ensure float64 operations
* new test
* make fixup
* make copies
* handle edge case double tokens ending with different tokens
* handle single timestamp ending
* make fixup
* handle conditioning on prev segments
* fix
* Update src/transformers/models/whisper/generation_whisper.py
Co-authored-by: Yoach Lacombe <52246514+ylacombe@users.noreply.github.com>
* [run-slow] whisper
* don't call item() to avoid unnecessary sync
* fix
---------
Co-authored-by: Yoach Lacombe <52246514+ylacombe@users.noreply.github.com>
Co-authored-by: Eustache Le Bihan <eustlb@users.noreply.huggingface.co>
2024-12-05 13:46:29 +01:00
Yih-Dar
beb2c66ec3
Informative ( #35059 )
...
* fix
* fix
* fix
* fix
---------
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2024-12-05 09:50:27 +01:00
Steven Liu
1ed1de2fec
[docs] Increase visibility of torch_dtype="auto" ( #35067 )
...
* auto-dtype
* feedback
2024-12-04 09:18:44 -08:00
Fanli Lin
baa3b22137
[docs] add a comment that offloading requires CUDA GPU ( #35055 )
...
* add commen to offloading
* Update docs/source/en/kv_cache.md
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
---------
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
2024-12-04 07:48:34 -08:00
Cyril Vallez
1da1e0d7f2
Support for easier multimodal use of modular ( #35056 )
...
* update modular and add examples
* style
* improve example comments
* style
* fix small logic issue for imports
* fix relative order issue when files do not make sense
* Improve comments
* trigger CIs
2024-12-04 15:13:11 +01:00
Anton Vlasjuk
46df859975
[GPTNeoX] Flex Attention + Refactor ( #34896 )
...
* gpt neox flex attention + refactor
* some formatting
* small fix on dropout
* add assertion on flex attn test
* flaky ci :(
* add head mask support
* style
* handle dtype, replace torch where
* fixup flex with output attns
* code review and several other fixes
* Update src/transformers/modeling_utils.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
* style
* remove unnecessary comment
* remove incorrect comment
* make flex attn check more agnostic tor versions and centralized
* change peft input dtype check to value since q and k could be affected by other stuff like RoPE
* i forgor
* flaky
* code review and small fixes
* Update src/transformers/models/gpt_neox/modeling_gpt_neox.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
---------
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
2024-12-04 14:48:28 +01:00
Vladislav Bronzov
accb7204f9
Add Pytorch Tensor Parallel support for Qwen2, Qwen2Moe, Starcoder2 ( #35007 )
...
* add base tp plan for qwen2 and qwen2moe
* add parallel tp for starcoder2
* fix modular conversion
* add infer dim for qkv states
* Update src/transformers/models/qwen2_moe/configuration_qwen2_moe.py
---------
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
2024-12-04 14:43:36 +01:00
Tianshu Wang
c7a109ec81
Fix pad_token_tensor is None in warning ( #34005 )
...
Fix pad_token_tensor is None in warning
2024-12-04 11:15:25 +01:00
Fanli Lin
329f5dbf97
[docs] use device-agnostic API instead of hard-coded cuda ( #35048 )
...
replace cuda
2024-12-03 10:54:15 -08:00
Fanli Lin
b8cdc262d5
[docs] use device-agnostic instead of cuda ( #35047 )
...
* fix on xpu
* [run_all]
* add the missing import for Image lib
* add more devices in comment
* bug fix
* replace cuda
2024-12-03 10:53:45 -08:00
wwwbai
346597b644
Translate community.md into Chinese ( #35013 )
...
* community translation
* Update docs/source/zh/community.md
Co-authored-by: Isotr0py <2037008807@qq.com>
---------
Co-authored-by: Isotr0py <2037008807@qq.com>
2024-12-03 10:22:02 -08:00
Fanli Lin
3deaa8179d
[docs] fix example code bug ( #35054 )
...
fix code bug
2024-12-03 09:18:39 -08:00
Wang, Yi
125de41643
fix speecht5 failure issue in test_peft_gradient_checkpointing_enable… ( #34454 )
...
* fix speecht5 failure issue in test_peft_gradient_checkpointing_enable_disable
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
* [run-slow] speecht5
---------
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
Co-authored-by: Matt <rocketknight1@gmail.com>
2024-12-03 13:58:54 +00:00
Yih-Dar
7a7f27697a
Fix BertGeneration ( #35043 )
...
fix
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2024-12-03 13:56:59 +01:00
Aymeric Roucher
901f504580
Add token cost + runtime monitoring to Agent and HfEngine children ( #34548 )
...
* Add monitoring to Agent and HfEngine children
2024-12-03 13:14:52 +01:00
Cyril Vallez
ee37bf0d95
Automatic compilation in generate: do not rely on inner function ( #34923 )
...
* compiled forward in PreTrainedModel
* update
* style
* update name
* trigger CIs
* Add way to use custom compile args
* style
* switch parameterization to generation_config
* Add to inits
* Update configuration_utils.py
* inits
* style
* docs
* style
* Update configuration_utils.py
* back without dataclass for repo consistency
* Update configuration_utils.py
* style
* style
* style once again
* add config serialization
* update
* true dataclass
* trigger CIs
* merge compile methods + remove serialization of compile config
2024-12-03 11:20:31 +01:00
wwwbai
f9c7e6021e
Translate bertlogy.md into Chinese ( #34908 )
...
* bertology translation
* Update docs/source/zh/_toctree.yml
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
* Update docs/source/zh/bertology.md
Co-authored-by: blueingman <15329507600@163.com>
* Update docs/source/zh/bertology.md
Co-authored-by: blueingman <15329507600@163.com>
* Update docs/source/zh/bertology.md
Co-authored-by: Isotr0py <2037008807@qq.com>
* Update docs/source/zh/bertology.md
Co-authored-by: Isotr0py <2037008807@qq.com>
---------
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
Co-authored-by: blueingman <15329507600@163.com>
Co-authored-by: Isotr0py <2037008807@qq.com>
2024-12-02 11:42:40 -08:00
Fanli Lin
527dc04e46
[docs] add the missing import for Image and bug fix ( #34776 )
...
* add the missing import for Image lib
* add more devices in comment
* bug fix
2024-12-02 11:40:20 -08:00
Ahmed Almaghz
4955e4e638
[i18n-ar] Translated file : docs/source/ar/notebooks.md into Arabic ( #33049 )
...
* Add docs/source/ar/notebooks.md to Add_docs_source_ar_notebooks.md
* Update notebooks.md
* Update _toctree.yml
2024-12-02 11:40:04 -08:00
secrettoad
f0dec874f0
add docstring example for compute_loss_func ( #35020 )
2024-12-02 11:39:09 -08:00
Henry Hyeonmok Ko
31299670cd
Multiple typo fixes in Tutorials docs ( #35035 )
...
* Fixed typo in multi gpu docs and OLMoE version
* Fixed typos in docs for agents, agents advanced, knowledge distillation, and image feature extraction
* Fixed incorrect usage of model.image_guided_detection in zero shot object detection docs
2024-12-02 15:26:34 +00:00
Dmitry Rogozhkin
31830474bf
Fix test_eager_matches_sdpa_inference for XPU backend ( #34889 )
...
* Use torch.nn.attention.sdpa_kernel instead of deprecated torch.backends.cuda.sdp_kernel
Signed-off-by: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
* Fix test_eager_matches_sdpa_inference for XPU backend
As of PyTorch 2.5 XPU backend supports only torch.nn.attention.SDPBackend.MATH
which is implemented on PyTorch level using aten operators and is device
agnostic with respect to implementation of each aten operator. Thus, we can
reuse CUDA (or CPU) MATH weights for XPU.
Fixes : #34888
Signed-off-by: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
* Use torch.amp.autocast instead of deprecated torch.cuda.amp.autocast in nemotron
Signed-off-by: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
---------
Signed-off-by: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
2024-12-02 16:21:04 +01:00
Jacky Lee
f41d5d8f74
Add type hints for forward functions in Gemma2 ( #35034 )
...
* feat: add gemma2 type hints
* fix: mask is optional
2024-12-02 14:03:36 +00:00
Bojun Feng
7b5f76e32e
Typo in warning switching to optimum-quanto ( #35028 )
...
fix typos
2024-12-02 13:47:05 +00:00
milesial
c24c79ebf9
Optimize memory usage of mllama encoder ( #34930 )
...
mllama encoder memory optimization
2024-12-02 11:46:45 +01:00
Weize Chen
9ab8c5b503
fix variable undefined bug when return_tensors is not specified in llava processing ( #34953 )
...
* fix variable undefined bug when return_tensors is not specified in llava processor
* improve readability
2024-12-02 11:44:42 +01:00
Joshua Lochner
3480cbb97e
Only cast cu_seqlens when tracing ( #35016 )
...
* Only cast `cu_seqlens` when tracing
* Formatting
2024-12-02 11:39:39 +01:00
Alvaro Bartolome
19dabe9636
Update FillMaskPipeline.__call__ signature and docstring ( #35006 )
...
Update `FillMaskPipeline.__call__`
- Remove unused `*args`
- Update docstring with `inputs` over `args`
2024-11-29 13:44:56 +00:00
Samuel Larkin
f7427f58ed
fix: double verbs ( #35008 )
2024-11-29 13:19:57 +00:00
