c-my/transformers
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
4,316 Commits 171 Branches 49 Tags
7ac91107119f95a9034e5404bd5af34355d0ffa5
Commit Graph
4316 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
Thomas Wolf 7ac9110711 Add more tests on tokenizers serialization - fix bugs (#5056) 
* update tests for fast tokenizers + fix small bug in saving/loading

* better tests on serialization

* fixing serialization

* comment cleanup
 2020-06-24 21:53:08 +02:00
Sylvain Gugger 0148c262e7 Fix first test (#5255) 2020-06-24 15:16:04 -04:00
Sylvain Gugger 70c1e1d2d5 Use master _static (#5253) 
* Use _static from master everywhere

* Copy to existing too
 2020-06-24 15:06:14 -04:00
Victor SANH 4965aee064 [HANS] Fix label_list for RoBERTa/BART (class flipping) (#5196) 
* fix weirdness in roberta/bart for mnli trained checkpoints

* black compliance

* isort code check
 2020-06-24 14:38:15 -04:00
Julien Chaumond fc24a93e64 [HfApi] Add support for pipeline_tag 2020-06-24 16:54:00 +00:00
Setu Shah 0a3d0e02c5 Replace labels with -100 to skip loss calc (#4718) 2020-06-24 12:14:50 -04:00
Sylvain Gugger 6894b486d0 Fix version controller links (for realsies) (#5251) 2020-06-24 12:13:43 -04:00
Sai Saketh Aluru 1121ce9f98 Model cards for Hate-speech-CNERG models (#5236) 
* Add dehatebert-mono-arabic readme card

* Update dehatebert-mono-arabic model card

* model cards for Hate-speech-CNERG models
 2020-06-24 11:41:08 -04:00
Lysandre Debut cf10d4cfdd Cleaning TensorFlow models (#5229) 
* Cleaning TensorFlow models

Update all classes


stylr

* Don't average loss
 2020-06-24 11:37:20 -04:00
Sylvain Gugger 609e0c583f Fix links (#5248) 2020-06-24 11:35:55 -04:00
Ali Modarressi c9163a8d5a delay decay schedule until the end of warmup (#4940) 2020-06-24 11:18:29 -04:00
Sylvain Gugger f216b60671 Fix deploy doc (#5246) 
* Try with the same command

* Try like this
 2020-06-24 10:59:06 -04:00
Sylvain Gugger 49f6e7a3c6 Add some prints to debug (#5244) 2020-06-24 10:37:01 -04:00
Patrick von Platen c2a26ec8a6 [Use cache] Align logic of use_cache with output_attentions and output_hidden_states (#5194) 
* fix use cache

* add bart use cache

* fix bart

* finish bart
 2020-06-24 16:09:17 +02:00
Sylvain Gugger 64c393ee74 Don't recreate old docs (#5243) 2020-06-24 09:59:07 -04:00
Patrick von Platen b29683736a fix print in benchmark (#5242) 2020-06-24 15:58:49 +02:00
Patrick von Platen 9fe09cec76 [Benchmark] Extend Benchmark to all model type extensions (#5241) 
* add benchmark for all kinds of models

* improved import

* delete bogus files

* make style
 2020-06-24 15:11:42 +02:00
Sylvain Gugger 7c41057d50 Add hugs (#5225) 2020-06-24 07:56:14 -04:00
Sylvain Gugger 5e85b324ec Use the script in utils (#5224) 2020-06-24 07:55:58 -04:00
flozi00andJulien Chaumond 5e31a98ab7 Create README.md (#5108) 
* Create README.md

* Update model_cards/a-ware/roberta-large-squad-classification/README.md

Co-authored-by: Julien Chaumond <chaumond@gmail.com>
 2020-06-24 04:45:51 -04:00
Adriano Diniz 033124e5f8 Update README.md (#5199) 
Fix/add information in README.md
 2020-06-24 04:42:46 -04:00
ahotrod 7ca6627ec3 Create README.md (#5217) 
electra_large_discriminator_squad2_512 Question Answering LM
 2020-06-24 04:40:50 -04:00
Kevin Canwen Xu 54e9ce785d Fix PABEE division by zero error (#5233) 
* Fix PABEE division by zero error

* patience=0 by default
 2020-06-24 16:10:36 +08:00
Sylvain Gugger 9022ef021a Only put tensors on a device (#5223) 
* Only put tensors on a device

* Type hint and unpack list comprehension
 2020-06-23 17:30:17 -04:00
Sylvain GuggerLysandre DebutJulien Chaumond
 		173528e368 Add version control menu (#5222) 
* Add version control menu

* Constify things

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

* Apply suggestions from code review

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

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Julien Chaumond <chaumond@gmail.com>
 2020-06-23 17:05:12 -04:00
Sam Shleifer 76e5af4cfd [pl_examples] revert deletion of optimizer_step (#5227) 2020-06-23 16:40:45 -04:00
Julien Chaumond c01480bba3 [file_utils] Type user-agent 2020-06-23 18:31:13 +02:00
Sam Shleifer 58918c76f4 [bart] add config.extra_pos_embeddings to facilitate reuse (#5190) 2020-06-23 11:35:42 -04:00
Thomas Wolf b28b537131 More clear error message in the use-case of #5169 (#5184) 2020-06-23 13:37:29 +02:00
Thomas WolfandLysandre Debut 11fdde0271 Tokenizers API developments (#5103) 
* Add return lengths

* make pad a bit more flexible so it can be used as collate_fn

* check all kwargs sent to encoding method are known

* fixing kwargs in encodings

* New AddedToken class in python

This class let you specify specifique tokenization behaviors for some special tokens. Used in particular for GPT2 and Roberta, to control how white spaces are stripped around special tokens.

* style and quality

* switched to hugginface tokenizers library for AddedTokens

* up to tokenizer 0.8.0-rc3 - update API to use AddedToken state

* style and quality

* do not raise an error on additional or unused kwargs for tokenize() but only a warning

* transfo-xl pretrained model requires torch

* Update src/transformers/tokenization_utils.py

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

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
 2020-06-23 13:36:57 +02:00
Patrick von Platen 1ae132a07d [Reformer] Axial Pos Emb Improve mem usage reformer (#5209) 
* improve mem handling

* improve mem for pos ax encodings
 2020-06-23 10:49:18 +02:00
Sam Shleifer 5144104070 [fix] remove unused import (#5206) 2020-06-22 23:39:04 -04:00
Sam Shleifer 0d158e38c9 [fix] mobilebert had wrong path, causing slow test failure (#5205) 2020-06-22 23:31:36 -04:00
Sam Shleifer f5c2a122e3 Upgrade examples to pl=0.8.1(#5146) 2020-06-22 20:40:10 -04:00
flozi00 06b60c8b05 [Modelcard] bart-squadv2 (#5011) 
* [Modelcard] bart-squadv2

* Update README.md

* Update README.md
 2020-06-22 18:40:19 -04:00
flozi00 35e0687256 Create README.md (#5013) 2020-06-22 18:40:00 -04:00
Fran Martinez 22d2c8ea2f Create README.md for finetuned BERT model (#5009) 
* Create README.md

* changes in model usage section

* minor changes in output visualization

* minor errata in readme
 2020-06-22 18:39:29 -04:00
furunkel 2589505693 Add model card for StackOBERTflow-comments-small (#5008) 
* Create README.md

* Update README.md
 2020-06-22 18:39:22 -04:00
bogdankostic d8c26ed139 Specify dataset used for crossvalidation (#5175) 2020-06-22 18:26:12 -04:00
Adriano Diniz a34fb91d54 Create README.md (#5149) 2020-06-22 18:00:53 -04:00
Adriano Diniz ffabcf5249 Create README.md (#5160) 2020-06-22 17:59:54 -04:00
Adriano DinizandJulien Chaumond 3363a19b12 Create README.md (#5152) 
* Create README.md

* Apply suggestions from code review

Co-authored-by: Julien Chaumond <chaumond@gmail.com>
 2020-06-22 17:59:33 -04:00
Michaël BenestyandPatrick von Platen 0cca61925c Add link to new comunity notebook (optimization) (#5195) 
* Add link to new comunity notebook (optimization)

related to https://github.com/huggingface/transformers/issues/4842#event-3469184635

This notebook is about benchmarking model training with/without dynamic padding optimization. 
https://github.com/ELS-RD/transformers-notebook 

Using dynamic padding on MNLI provides a **4.7 times training time reduction**, with max pad length set to 512. The effect is strong because few examples are >> 400 tokens in this dataset. IRL, it will depend of the dataset, but it always bring improvement and, after more than 20 experiments listed in this [article](https://towardsdatascience.com/divide-hugging-face-transformers-training-time-by-2-or-more-21bf7129db9q-21bf7129db9e?source=friends_link&sk=10a45a0ace94b3255643d81b6475f409), it seems to not hurt performance.

Following advice from @patrickvonplaten I do the PR myself :-)

* Update notebooks/README.md

Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
 2020-06-22 23:47:33 +02:00
Lee Haau-SingandAlex Warstadt 1c5cd8e5f5 Add README.md (nyu-mll) (#5174) 
* nyu-mll: roberta on smaller datasets

* Update README.md

* Update README.md

Co-authored-by: Alex Warstadt <alexwarstadt@gmail.com>
 2020-06-22 17:24:27 -04:00
Sylvain Gugger c439752482 Switch master/stable doc and add older releases (#5193) 2020-06-22 16:38:53 -04:00
Sylvain GuggerThomas WolfLysandre Debut
 		417e492f1e Quick tour (#5145) 
* Quicktour part 1

* Update

* All done

* Typos

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

* Address comments in quick tour

* Update docs/source/quicktour.rst

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

* Update from feedback

Co-authored-by: Thomas Wolf <thomwolf@users.noreply.github.com>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
 2020-06-22 16:08:09 -04:00
Thomas WolfandJulien Chaumond 75e1eed8d1 Cleaner warning when loading pretrained models (#4557) 
* Cleaner warning when loading pretrained models

This make more explicit logging messages when using the various `from_pretrained` methods. It also make these messages as `logging.warning` because it's a common source of silent mistakes.

* Update src/transformers/modeling_utils.py

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

* Update src/transformers/modeling_utils.py

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

* style and quality

Co-authored-by: Julien Chaumond <chaumond@gmail.com>
 2020-06-22 21:58:47 +02:00
Lysandre Debut 4e741efa92 Have documentation fail on warning (#5189) 
* Have documentation fail on warning

* Force ci failure

* Revert "Force ci failure"

This reverts commit f0a4666ec2eb4cd00a4da48af3357defc63324a0.
 2020-06-22 15:49:50 -04:00
Sylvain Gugger 1262495a91 Add TF auto model to the docs + fix sphinx warnings (#5187) 2020-06-22 14:43:52 -04:00
Adriano Diniz 88429c57bc Create README.md (#5165) 2020-06-22 13:49:14 -04:00