riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-08-02 19:21:31 +06:00
Code Issues Actions 3 Packages Projects Releases Wiki Activity
5,601 Commits 892 Branches 232 Tags 2.1 GiB
06fc3954a1
Commit Graph

5601 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sylvain Gugger
06fc3954a1
Only log total_flos at the end of training (#7981) 
* Only log total_flos at the end of training

* Fix test
 2020-10-22 14:26:55 -04:00
Julien Chaumond
ff65beafa3
FillMaskPipeline: support passing top_k on __call__ (#7971) 
* FillMaskPipeline: support passing top_k on __call__

Also move from topk to top_k

* migrate to new param name in tests

* Review from @sgugger
 2020-10-22 12:54:25 -04:00
Sylvain Gugger
2e5052d4f1
New run glue script (#7917) 
* Start simplification

* More progress

* Finished script

* Address comments and update tests instructions

* Wrong test

* Accept files as inputs and fix test

* Update src/transformers/trainer_utils.py

Co-authored-by: Julien Chaumond <chaumond@gmail.com>

* Fix labels and add combined score

* Add special labels

* Update TPU command

* Revert to old label strategy

* Use model labels

* Fix for STT-B

* Styling

* Apply suggestions from code review

Co-authored-by: Thomas Wolf <thomwolf@users.noreply.github.com>

* Code styling

* Fix review comments

Co-authored-by: Julien Chaumond <chaumond@gmail.com>
Co-authored-by: Thomas Wolf <thomwolf@users.noreply.github.com>
 2020-10-22 11:42:22 -04:00
Nicolas Patry
18ce6b8ff3
Fixing the "translation", "translation_XX_to_YY" pipelines. (#7975) 
* Actually make the "translation", "translation_XX_to_YY" task behave correctly.

Background:
- Currently "translation_cn_to_ar" does not work. (only 3 pairs are
supported)
- Some models, contain in their config the correct values for the (src,
tgt) pair they can translate. It's usually just one pair, and we can
infer it automatically from the `model.config.task_specific_params`. If
it's not defined we can still probably load the TranslationPipeline
nevertheless.

Proposed fix:
- A simplified version of what could become more general which is
a `parametrized` task. "translation" + (src, tgt) in this instance
it what we need in the general case. The way we go about it for now
is simply parsing "translation_XX_to_YY". If cases of parametrized task arise
we should preferably go in something closer to what `datasets` propose
which is having a secondary argument `task_options`? that will be close
to what that task requires.
- Should be backward compatible in all cases for instance
`pipeline(task="translation_en_to_de") should work out of the box.
- Should provide a warning when a specific translation pair has been
selected on behalf of the user using
`model.config.task_specific_params`.

* Update src/transformers/pipelines.py

Co-authored-by: Julien Chaumond <chaumond@gmail.com>

Co-authored-by: Julien Chaumond <chaumond@gmail.com>
 2020-10-22 17:16:21 +02:00
Funtowicz Morgan
901e9b8eda
Remove the else branch adding 0 to the hidden state if token_type_embeds is None. (#7977) 
Signed-off-by: Morgan Funtowicz <funtowiczmo@gmail.com>
 2020-10-22 16:41:41 +02:00
Patrick von Platen
f34372a9ff
[PretrainedConfig] Fix save pretrained config for edge case (#7943) 
* fix config save

* add test

* add config class variable and another test

* line break

* fix fsmt and typo

* god am I making many errors today :-/

* Update src/transformers/configuration_utils.py

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

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
 2020-10-22 15:39:01 +02:00
Peter Bayerle
cc2e312ca3
adding text classification with DistilBERT/tf notebook (#7964) 
Looking at the current community notebooks, it seems that few are targeted for absolute beginners and even fewer are written with TensorFlow. This notebook describes absolutely everything a beginner would need to know, including how to save/load their model and use it for new predictions (this is often omitted in tutorials)

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
 2020-10-22 09:30:50 -04:00
wlhgtc
a16e568f22
# Add whole word mask support for lm fine-tune (#7925) 
* ADD: add whole word mask proxy for both eng and chinese

* MOD: adjust format

* MOD: reformat code

* MOD: update import

* MOD: fix bug

* MOD: add import

* MOD: fix bug

* MOD: decouple code and update readme

* MOD: reformat code

* Update examples/language-modeling/README.md

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

* Update examples/language-modeling/README.md

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

* Update examples/language-modeling/run_language_modeling.py

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

* Update examples/language-modeling/run_language_modeling.py

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

* Update examples/language-modeling/run_language_modeling.py

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

* Update examples/language-modeling/run_language_modeling.py

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

* change wwm to whole_word_mask

* reformat code

* reformat

* format

* Code quality

* ADD: update chinese ref readme

* MOD: small changes

* MOD: small changes2

* update readme

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <sylvain.gugger@gmail.com>
 2020-10-22 09:19:00 -04:00
Stas Bekman
64b4d25cf3
[fsmt test] basic config test with online model + super tiny model (#7860) 
* basic config test with online model

* typo

* style

* better test
 2020-10-22 09:14:54 -04:00
Julien Chaumond
3479787edc
Disable inference API for t5-11b (#7978) 2020-10-22 09:08:37 -04:00
Julien Chaumond
a7db81c33f [model_card] t5-11b move disclaimer to top of page 
cc @Narsil @patrickvonplaten
 2020-10-22 14:35:31 +02:00
Haebin Shin
f774b2e8c4
support relative path for best_model_checkpoint (#7973) 2020-10-22 07:55:31 -04:00
Stas Bekman
8348105692
[testing] slow tests should be marked as slow (#7895) 
* slow tests should be slow

* exception note

* style

* integrate LysandreJik's notes with some expansions

* Apply suggestions from code review

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

* another slow test

* fix link, and prose

* clarify.

* note from Sam

* typo

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
 2020-10-22 06:34:05 -04:00
rmroczkowski
95792a948e
Herbert tokenizer auto load (#7968) 2020-10-22 05:48:29 -04:00
zolekode
4abb7ffc18
added qg evaluation notebook (#7958) 
* added qg evaluation notebook

* Update notebooks/README.md

Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
 2020-10-22 11:02:12 +02:00
Stas Bekman
8b38173398
[seq2seq testing] multigpu test run via subprocess (#7281) 
Co-authored-by: Sam Shleifer <sshleifer@gmail.com>
 2020-10-21 17:20:53 -04:00
Julien Chaumond
f8d3695e8c
[model_cards] camembert: dataset = oscar 
Hat/tip @pjox
 2020-10-21 14:17:56 -04:00
Evan Pete Walsh
16da877139
fix 'encode_plus' docstring for 'special_tokens_mask' (0s and 1s were reversed) (#7949) 
* fix docstring for 'special_tokens_mask'

* revert auto formatter changes

* revert another auto format

* revert another auto format
 2020-10-21 13:57:44 -04:00
Patrick von Platen
52decab371
fix test (#7947) 2020-10-21 19:06:23 +02:00
Patrick von Platen
9b6610f7f6
[ProphetNet] Correct Doc string example (#7944) 
* correct xlm prophetnet auto model and examples

* fix line-break docs
 2020-10-21 17:27:20 +02:00
François Lagunas
e174bfeb34
TensorBoard/Wandb/optuna/raytune integration improvements. (#7935) 
Improved TensorBoard and Wandb integration, as well as optuna and ray/tune support, with minor modifications to trainer core code.
 2020-10-21 17:18:52 +02:00
Ali Hamdi Ali Fadel
bf162ce8ca
Add AI-SOCO models (#7867) 2020-10-21 09:24:43 -04:00
Fangyu Liu
58fb25f25b
Create README.md (#7857) 
* Create README.md

model card for cambridgeltl/BioRedditBERT-uncased.

* Update model_cards/cambridgeltl/BioRedditBERT-uncased/README.md

Co-authored-by: Julien Chaumond <chaumond@gmail.com>
 2020-10-21 08:41:41 -04:00
Manuel Romero
2b07ec7823
Model card for German BERT fine-tuned for LER/NER (#7855) 2020-10-21 08:31:41 -04:00
MichalPleban
35d2ad5b83
Create README.md (#7819) 2020-10-21 08:30:01 -04:00
Wuwei Lan
bdda4f2249
Create README.md (#7625) 
* Create README.md

* Update model_cards/lanwuwei/GigaBERT-v3-Arabic-and-English/README.md

* Update model_cards/lanwuwei/GigaBERT-v3-Arabic-and-English/README.md

Co-authored-by: Julien Chaumond <chaumond@gmail.com>
 2020-10-21 08:29:39 -04:00
Manuel Romero
8e23749649
Add missing comma (#7870) 2020-10-21 08:24:12 -04:00
Manuel Romero
3eaa007d78
Create README.md (#7899) 2020-10-21 08:23:55 -04:00
Julien Chaumond
758572cad8 [model_cards] move hatmimoha/arabic-ner to correct location 
see 16d3cc187d and https://github.com/huggingface/transformers/pull/7836
 2020-10-21 14:13:17 +02:00
Stas Bekman
57516c0cc8
[multiple models] skip saving/loading deterministic state_dict keys (#7878) 
* make the save_load special key tests common

* handle mbart

* cleaner solution

* fix

* move test_save_load_missing_keys back into fstm for now

* restore

* style

* add marian

* add pegasus

* blenderbot

* revert - no static embed
 2020-10-21 08:06:07 -04:00
quentinheinrich
006a16483f
update model cards of Illuin models (#7930) 2020-10-21 08:05:53 -04:00
hatmimoha
16d3cc187d
model card for arabic-ner model (#7836) 
* Create README.md

README file for the Arabic NER model

* Update README.md

* Update README.md

* Update hatmimoha/arabic-ner/README.md

Co-authored-by: Julien Chaumond <chaumond@gmail.com>
 2020-10-21 08:02:40 -04:00
Sam Shleifer
829842159e
Add TFBartForConditionalGeneration (#5411) 
* half done

* doc improvement

* Cp test file

* brokedn

* broken test

* undo some mess

* ckpt

* borked

* Halfway

* 6 passing

* boom boom

* Much progress but still 6

* boom boom

* merged master

* 10 passing

* boom boom

* Style

* no t5 changes

* 13 passing

* Integration test failing, but not gibberish

* Frustrated

* Merged master

* 4 fail

* 4 fail

* fix return_dict

* boom boom

* Still only 4

* prepare method

* prepare method

* before delete classif

* Skip tests to avoid adding boilerplate

* boom boom

* fast tests passing

* style

* boom boom

* Switch to supporting many input types

* remove FIXMENORM

* working

* Fixed past_key_values/decoder_cached_states confusion

* new broken test

* Fix attention mask kwarg name

* undo accidental

* Style and reviewers

* style

* Docs and common tests

* Cleaner assert messages

* copy docs

* style issues

* Sphinx fix

* Simplify caching logic

* test does not require torch

* copy _NoLayerEmbedTokens

* Update src/transformers/modeling_tf_bart.py

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Update tests/test_modeling_tf_bart.py

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Update src/transformers/modeling_tf_bart.py

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Update src/transformers/modeling_tf_bart.py

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Update src/transformers/modeling_tf_bart.py

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Line length and dont document None

* Add pipeline test coverage

* assert msg

* At parity

* Assert messages

* mark slow

* Update compile test

* back in init

* Merge master

* Fix tests

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
 2020-10-21 13:10:16 +02:00
Patrick von Platen
5cd9e2cba1
Update README.md 2020-10-21 12:43:42 +02:00
Patrick von Platen
220b5f97ca
Create README.md 2020-10-21 12:34:46 +02:00
Patrick von Platen
8ffd7fb12d
Update README.md 2020-10-21 12:27:09 +02:00
Patrick von Platen
613ab364eb
Update README.md 2020-10-21 12:23:17 +02:00
Patrick von Platen
f7eb17dc47
Update README.md 2020-10-21 12:19:44 +02:00
Patrick von Platen
29792864cb
[ProphetNet] Add Question Generation Model + Test (#7942) 
* new prophetnet model

* correct name

* make style
 2020-10-21 11:49:58 +02:00
Joe Davison
13842e413c
PPL guide minor code snippet fix (#7938) 2020-10-20 16:17:39 -06:00
Stas Bekman
0e24e4c136
[s2s] create doc for pegasus/fsmt replication (#7934) 2020-10-20 15:07:52 -04:00
Lysandre Debut
96f4828ace
Respect the 119 line chars (#7928) 2020-10-20 11:02:47 -04:00
Lysandre
ef0ac063c9 Docs for v3.4.0 2020-10-20 16:29:00 +02:00
Lysandre
eb0e0ce2ad Release: v3.4.0 2020-10-20 16:22:26 +02:00
Patrick von Platen
0264048660
Update README.md 2020-10-20 16:13:49 +02:00
Patrick von Platen
ffd675b42c
add summary (#7927) 2020-10-20 10:11:02 -04:00
Lysandre Debut
5547b40b13
labels and decoder_input_ids to Glossary (#7906) 
* labels and decoder_input_ids to Glossary

* Formatting fixes

* Update docs/source/glossary.rst

Co-authored-by: Sam Shleifer <sshleifer@gmail.com>

* sam's comments

Co-authored-by: Sam Shleifer <sshleifer@gmail.com>
 2020-10-20 09:50:47 -04:00
Patrick von Platen
f3312515b7
Add note for WikiSplit 2020-10-20 15:42:29 +02:00
Patrick von Platen
0724c0f3a2
Fix EncoderDecoder WikiSplit Example 2020-10-20 15:13:22 +02:00
Stas Bekman
ca37db0559
[flax] fix repo_check (#7914) 
* [flax] fix repo_check

Unless, this is actually a problem, this adds `modeling_flax_utils` to ignore list. otherwise currently it expects to have a 'tests/test_modeling_flax_utils.py' for it.
for context please see: https://github.com/huggingface/transformers/pull/3722#issuecomment-712360415

* fix 2 more issues

* merge https://github.com/huggingface/transformers/pull/7919/
 2020-10-20 07:55:40 -04:00