SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity
13,113 Commits 5 Branches 239 Tags
66fd3a8d626a32989f4569260db32785c6cbf42a
Commit Graph

13113 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sylvain Gugger
66fd3a8d62 Patch: v4.30.2
Some checks failed
Release - Conda / build_and_package (push) Has been cancelled
Details
v4.30.2 		2023-06-13 14:24:02 -04:00
NielsRogge
8f9f1efaf8 Fix push to hub (#24187) 
Add fix
 2023-06-13 14:23:39 -04:00
Matt
497d66740b Fix how we detect the TF package (#24255) 
* Fix how we detect the TF package

* Add a comment as a talisman warding against future harm

* Actually put the comment in the right place
 2023-06-13 14:22:33 -04:00
Sylvain Gugger
65a1ec05ca Patch: v4.30.1
Some checks failed
Release - Conda / build_and_package (push) Has been cancelled
Details
v4.30.1 		2023-06-09 10:48:44 -04:00
Younes Belkada
fd59fc1a7f [bnb] Fix bnb config json serialization (#24137) 
* fix bnb config json serialization

* forward contrib credits from discussions

---------

Co-authored-by: Andrechang <Andrechang@users.noreply.github.com>
 2023-06-09 08:50:23 -04:00
Sourab Mangrulkar
a272e4135c fix bugs with trainer (#24134) 
* fix the deepspeed test failures

* apex fix

* FSDP save ckpt fix

* Update src/transformers/trainer.py

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

---------

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
 2023-06-09 08:31:09 -04:00
Matt
50ed79312d Correctly build models and import call_context for older TF versions (#24138) 2023-06-09 08:30:54 -04:00
Younes Belkada
fe861e578f [GPT2] Add correct keys on _keys_to_ignore_on_load_unexpected on all child classes of GPT2PreTrainedModel (#24113)
Some checks failed
Release - Conda / build_and_package (push) Has been cancelled
Details 
* add correct keys on `_keys_to_ignore_on_load_unexpected`

* oops
v4.30.0 		2023-06-08 10:22:12 -04:00
Sylvain Gugger
b3e27a8057 Update the pin on Accelerate (#24110) 2023-06-08 10:22:05 -04:00
Younes Belkada
53e1f5cf66 [Trainer] Correct behavior of _load_best_model for PEFT models (#24103) 
* v1

* some refactor

- add ST format as well

* fix

* add `ADAPTER_WEIGHTS_NAME` & `ADAPTER_SAFE_WEIGHTS_NAME`
 2023-06-08 09:40:01 -04:00
Sourab Mangrulkar
17db177714 reset accelerate env variables after each test (#24107) 2023-06-08 09:36:42 -04:00
Sylvain Gugger
905892f090 Release: v4.30.0 2023-06-07 16:48:28 -04:00
Sylvain Gugger
c3572e6bfb Add AzureOpenAiAgent (#24058) 
* Add AzureOpenAiAgent

* quality

* Update src/transformers/tools/agents.py

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

---------

Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>
 2023-06-07 16:34:53 -04:00
Zachary Mueller
5eb3d3c702 Up pinned accelerate version (#24089) 
* Min accelerate

* Also min version

* Min accelerate

* Also min version

* To different minor version

* Empty
 2023-06-07 16:21:51 -04:00
Sourab Mangrulkar
d1c039e398 fix accelerator prepare during eval only mode (#24014) 
* fix mixed precision prep during eval only mode

* update to address comments

* update to reflect the changes in accelerate
 2023-06-08 01:03:13 +05:30
Sylvain Gugger
2c887cf8e0 Do not prepare lr scheduler as it as the right number of steps (#24088) 
* Do not prepare lr scheduler as it as the right number of steps

* Trigger CI

* Trigger CI

* Trigger CI

* Add fake comment

* Remove fake comment

* Trigger CI please!
 2023-06-07 15:31:32 -04:00
Sourab Mangrulkar
12298cb65c fix executable batch size issue (#24067) 
* fix executable batch size issue

* fix

* undo
 2023-06-07 22:08:04 +05:30
Mishig
ef010071ee Update delete_doc_comment_trigger.yml (#24084) 
fix base workflow name
 2023-06-07 17:55:48 +02:00
Sylvain Gugger
89b00eef94 Fix expected value in tests of the test fetcher (#24077) 
* Fix expected value in tests of the test fetcher

* Fix trigger for repo util tests
 2023-06-07 11:38:56 -04:00
Mishig
5c9394b54c [doc build] Use secrets (#24079) 2023-06-07 17:33:39 +02:00
Matt
1fc832b454 Make the TF dummies even smaller (#24071) 
* Let's see if we can use the smallest possible dummies

* Make GPT-2's dummies a little longer

* Just use (1,2) as the default shape

* Update other dummies in sync

* Correct imports for Keras 2.13

* Shrink the Wav2Vec2 dummies
 2023-06-07 16:23:05 +01:00
Yih-Dar
092c14c37d Be nice to TF (#24076) 
* fix

* fix

---------

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
 2023-06-07 16:18:13 +02:00
Younes Belkada
4795219228 [bnb] Fix bnb skip modules (#24043) 
* fix skip modules test

* oops

* address comments
 2023-06-07 15:27:46 +02:00
Michael Benayoun
a1160185ff Fix is_optimum_neuron_available (#23961) 
Fix is_optimum_neuron_available
 2023-06-07 09:13:01 -04:00
Younes Belkada
6b548129b1 [Hub] Add safe_serialization in push_to_hub (#24074) 
add `safe_serialization` in push_to_hub
 2023-06-07 09:07:33 -04:00
Younes Belkada
6daf7c311b Support PEFT models when saving the model using trainer (#24073) 
* support PEFT models when saving the model using trainer

* fixup
 2023-06-07 14:30:55 +02:00
YangLiu
1e4a7737ed Add support for non-rust implemented tokenization for __getitem__ method. (#24039) 
* Add support for non-rust implemented tokenization for `__getitem__` method.

* Update for error message on adding new sub-branch for `__item__` method.

---------

Co-authored-by: liuyang17 <liuyang17@zhihu.com>
 2023-06-07 12:29:19 +01:00
Patrick von Platen
52972e70c7 [Wav2Vec2] Fix torch srcipt (#24062) 
* [Wav2Vec2] Fix torch srcipt

* fix more
 2023-06-07 07:27:07 -04:00
Joao Gante
612b2a1a6d Generate: increase left-padding test atol (#23448) 
increase atol
 2023-06-07 11:56:57 +01:00
Sylvain Gugger
f1660d7e23 Remote code improvements (#23959) 
* Fix model load when it has both code on the Hub and locally

* Add input check with timeout

* Add tests

* Apply suggestions from code review

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

* Some non-saved stuff

* Add feature extractors

* Add image processor

* Add model

* Add processor and tokenizer

* Reduce timeout

---------

Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>
 2023-06-06 14:31:14 -04:00
Sylvain Gugger
60825f2c6e Fix device placement for model-parallelism in generate for encoder/de… (#24025) 
* Fix device placement for model-parallelism in generate for encoder/decoders

* Remove debug statements
 2023-06-06 14:30:59 -04:00
Yih-Dar
02d255db26 bring back filtered_test_list_cross_tests.txt (#24055) 
fix

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
 2023-06-06 19:35:24 +02:00
Edward Z. Yang
bc9ecef942 Use new parametrization based weight norm if available (#24030) 
* Use new parametrization based weight norm if available

See https://github.com/pytorch/pytorch/pull/103001

Signed-off-by: Edward Z. Yang <ezyang@meta.com>

* handle copies

Signed-off-by: Edward Z. Yang <ezyang@meta.com>

* black

Signed-off-by: Edward Z. Yang <ezyang@meta.com>

---------

Signed-off-by: Edward Z. Yang <ezyang@meta.com>
 2023-06-06 13:34:57 -04:00
Matt
4a55e47877 Move TF building to an actual build() method (#23760) 
* A fun new PR where I break the entire codebase again

* A fun new PR where I break the entire codebase again

* Handle cross-attention

* Move calls to model(model.dummy_inputs) to the new build() method

* Seeing what fails with the build context thing

* make fix-copies

* Let's see what fails with new build methods

* Fix the pytorch crossload build calls

* Fix the overridden build methods in vision_text_dual_encoder

* Make sure all our build methods set self.built or call super().build(), which also sets it

* make fix-copies

* Remove finished TODO

* Tentatively remove unneeded (?) line

* Transpose b in deberta correctly and remove unused threading local

* Get rid of build_with_dummies and all it stands for

* Rollback some changes to TF-PT crossloading

* Correctly call super().build()
 2023-06-06 18:30:51 +01:00
Zachary Mueller
cbf6bc2350 Oops, missed one (#24054) 
Oops
 2023-06-06 13:30:19 -04:00
Matt
7203ea6797 Reduce memory usage in TF building (#24046) 
* Make the default dummies (2, 2) instead of (3, 3)

* Fix for Funnel

* Actually fix Funnel
 2023-06-06 18:29:54 +01:00
Zachary Mueller
072188d638 Act on deprecations in Accelerate no_trainer examples (#24053) 
Act on deprecation
 2023-06-06 13:04:38 -04:00
Yih-Dar
ff4c0fc7d2 Tiny fix for check_self_hosted_runner.py (#24052) 
fix

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
 2023-06-06 18:17:41 +02:00
amyeroberts
a717e0318c Add TimmBackbone model (#22619) 
* Add test_backbone for convnext

* Add TimmBackbone model

* Add check for backbone type

* Tidying up - config checks

* Update convnextv2

* Tidy up

* Fix indices & clearer comment

* Exceptions for config checks

* Correclty update config for tests

* Safer imports

* Safer safer imports

* Fix where decorators go

* Update import logic and backbone tests

* More import fixes

* Fixup

* Only import all_models if torch available

* Fix kwarg updates in from_pretrained & main rebase

* Tidy up

* Add tests for AutoBackbone

* Tidy up

* Fix import error

* Fix up

* Install nattan in doc_test_job

* Revert back to setting self._out_xxx directly

* Bug fix - out_indices mapping from out_features

* Fix tests

* Dont accept output_loading_info for Timm models

* Set out_xxx and don't remap

* Use smaller checkpoint for test

* Don't remap timm indices - check out_indices based on stage names

* Skip test as it's n/a

* Apply suggestions from code review

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Cleaner imports / spelling is hard

---------

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
 2023-06-06 17:11:30 +01:00
Sylvain Gugger
b8935980a2 Modification of one text example file should trigger said test (#24051) 2023-06-06 12:02:56 -04:00
Tom Aarsen
02fe3af275 Prevent ZeroDivisionError on trainer.evaluate if model and dataset are tiny (#24049) 
Prevent ZeroDivisionError if evaluation is too quick
 2023-06-06 11:31:05 -04:00
Roy Hvaara
d924390d5b Use TruncatedNormal from Keras initializers (#24036) 
Co-authored-by: Andrey Voynov <avoin@google.com>
 2023-06-06 14:51:44 +01:00
Nicolas Patry
c2e3fa0b2a Fixing single candidate_label return. (#24023) 2023-06-06 15:26:10 +02:00
Marc Sun
6307312dfc Add check for tied parameters (#24029) 
* Add check for tied parameters

* Fix style

* fix style

* Fix versioning

* Change if to elif
 2023-06-06 09:12:46 -04:00
Wonhyeong Seo
7da3ce04a6 🌐 [i18n-KO] Translated bertology.mdx to Korean (#23968) 
* docs: ko: `bertology.mdx`

* feat: nmt draft

* fix: manual edits

* fix: resolve suggestions

Co-authored-by: Hyeonseo Yun <0525yhs@gmail.com>

---------

Co-authored-by: Hyeonseo Yun <0525yhs@gmail.com>
 2023-06-06 09:08:45 -04:00
Wonhyeong Seo
c938597657 🌐 [i18n-KO] Translated language-modeling.mdx (#23969) 
* docs: ko: `language_modeling.mdx`

* feat: nmt draft

* fix: manual edits

* fix: add inline toc

* fix: typo in toc_tree.yml

* fix: resolve suggestions

Co-authored-by: Sohyun Sim <96299403+sim-so@users.noreply.github.com>

---------

Co-authored-by: Sohyun Sim <96299403+sim-so@users.noreply.github.com>
 2023-06-06 09:08:26 -04:00
Yih-Dar
7631db0fdc Pin deepspeed to 0.9.2 for now (#24024) 
fix

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
 2023-06-05 20:00:28 +02:00
Yih-Dar
17846646f2 Fix MobileViTV2 checkpoint name (#24018) 
* fix

* fix

* Apply suggestions from code review

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

---------

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
 2023-06-05 18:12:45 +02:00
Hyeonseo Yun
649ffbf575 🌐 [i18n-KO] Translated tasks_explained.mdx to Korean (#23844) 
* docs: ko: tasks_explained.mdx

* feat: nmt and manual edit `tasks_explained.mdx`

* revised: resolve suggestions task_explained.mdx

* fixed: added draft of reference docs

Co-Authored-By: Kihoon Son <75935546+KIHOON71@users.noreply.github.com>
Co-Authored-By: Nayeon Han <nayeon2.han@gmail.com>

* revised: resolve suggestions(voca, spell check) task_explained.mdx

Co-Authored-By: Sohyun Sim <96299403+sim-so@users.noreply.github.com>

* revised: remove duplicate sentence in task_explained.mdx

* fixed: remove draft of reference docs

- I think it will be confusing in the translation process.
- This issue is included in #23971.

---------

Co-authored-by: Kihoon Son <75935546+KIHOON71@users.noreply.github.com>
Co-authored-by: Nayeon Han <nayeon2.han@gmail.com>
Co-authored-by: Sohyun Sim <96299403+sim-so@users.noreply.github.com>
 2023-06-05 12:02:03 -04:00
Brian Yu
2872f9671b TensorBoard callback no longer adds hparams (#23999) 
tensorboard callback no longer adds hparams
 2023-06-05 11:53:45 -04:00