riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-07-31 02:02:21 +06:00
Code Issues Actions 33 Packages Projects Releases Wiki Activity
5,152 Commits 887 Branches 232 Tags 2.1 GiB
0fab39695a
Commit Graph

5152 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sam Shleifer
0fab39695a
[s2s distill] allow pegasus-12-12 (#7104) 2020-09-14 00:03:59 -04:00
Sam Shleifer
de9e297964
[s2s] distributed eval cleanup (#7110) 2020-09-13 23:40:38 -04:00
Sam Shleifer
54395d87a6
Update xsum length penalty to better values (#7107) 2020-09-13 20:48:47 -04:00
Sam Shleifer
e7f8d2ab64
[s2s] two stage run_distributed_eval.py (#7105) 2020-09-13 17:28:18 -04:00
Sam Shleifer
0ec63afec2
fix bug in pegasus converter (#7094) 2020-09-13 15:11:47 -04:00
Sam Shleifer
b76cb1c3df
[s2s] run_eval supports --prefix clarg. (#6953) 2020-09-12 01:08:21 -04:00
李明浩
563ffb3dc3
Create README.md (#7066) 2020-09-11 15:21:05 -04:00
李明浩
1ad49cde3a
Create README.md (#7067) 2020-09-11 15:20:54 -04:00
Sagor Sarker
4753816e39
added bangla-bert-base model card and also modified other model cards (#7071) 
* added bangla-bert-base

* Apply suggestions from code review

Co-authored-by: Julien Chaumond <chaumond@gmail.com>
 2020-09-11 15:17:25 -04:00
Suraj Patil
0a8c17d53c
[T5Tokenizer] remove prefix_tokens (#7078) 2020-09-11 14:18:45 -04:00
Sylvain Gugger
4cbd50e611
Compute loss method (#7074) 2020-09-11 12:06:31 -04:00
Sylvain Gugger
ae736163d0
Add tests and fix various bugs in ModelOutput (#7073) 
* Add tests and fix various bugs in ModelOutput

* Update tests/test_model_output.py

Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
 2020-09-11 12:01:33 -04:00
Sylvain Gugger
e841b75dec
Automate the lists in auto-xxx docs (#7061) 
* More readable dict

* More nlp -> datasets

* Revert "More nlp -> datasets"

This reverts commit 3cd1883d22.

* Automate the lists in auto-xxx docs

* More readable dict

* Revert "More nlp -> datasets"

This reverts commit 3cd1883d22.

* Automate the lists in auto-xxx docs

* nlp -> datasets

* Fix new key
 2020-09-11 10:42:09 -04:00
Sylvain Gugger
0054a48cdd
Add dep on datasets (#7058) 2020-09-11 04:43:19 -04:00
Patrick von Platen
221d4c63a3
clean naming (#7068) 2020-09-11 09:57:53 +02:00
Stas Bekman
8fcbe486e1
these tests require non-multigpu env (#7059) 
* these tests require non-multigpu env

* cleanup

* clarify
 2020-09-10 18:52:55 -04:00
Sam Shleifer
77950c485a
[wip/s2s] DistributedSortishSampler (#7056) 2020-09-10 15:23:44 -04:00
Sylvain Gugger
514486739c
Fix CI with change of name of nlp (#7054) 
* nlp -> datasets

* More nlp -> datasets

* Woopsie

* More nlp -> datasets

* One last
 2020-09-10 14:51:08 -04:00
Sam Shleifer
e9a2f772bc
[s2s] --eval_max_generate_length (#7018) 2020-09-10 14:11:34 -04:00
Stas Bekman
df4594a9da
[xlm tok] config dict: fix str into int to match definition (#7034) 2020-09-10 19:31:01 +02:00
Julien Chaumond
d6c08b07a0 [AutoTokenizer] Correct error message 2020-09-10 17:19:01 +02:00
Patrick von Platen
db38f7ce29
[BertGeneration, Docs] Fix another old name in docs (#7050) 
* correct docs for bert generation

* upload
 2020-09-10 17:12:33 +02:00
Patrick von Platen
3bd95b0faf
correct docs for bert generation (#7048) 2020-09-10 17:08:40 +02:00
Patrick von Platen
eb2feb5d90
Create README.md 2020-09-10 17:05:50 +02:00
Ashwin Geet Dsa
66a5a6fda8
fix to ensure that returned tensors after the tokenization is Long (#7039) 
* fix to ensure that returned tensors after the tokenization is Long

* fix to ensure that returned tensors after the tokenization is Long

Co-authored-by: Ashwin Geet Dsa <adsa@grvingt-6.nancy.grid5000.fr>
 2020-09-10 11:04:03 -04:00
Patrick von Platen
9ccdb1d517
Update README.md 2020-09-10 17:01:19 +02:00
Patrick von Platen
60698936fc
Create README.md 2020-09-10 17:00:10 +02:00
Patrick von Platen
e0c3bc8ee0
Create README.md 2020-09-10 16:51:15 +02:00
Patrick von Platen
c356b9878d
Create README.md 2020-09-10 16:45:44 +02:00
Patrick von Platen
5afd3f6196
Create README.md 2020-09-10 16:44:47 +02:00
Sylvain Gugger
15a189049e
Add TF Funnel Transformer (#7029) 
* Add TF Funnel Transformer

* Proper dummy input

* Formatting

* Update src/transformers/modeling_tf_funnel.py

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

* Address review comments

* One review comment forgotten

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
 2020-09-10 10:41:56 -04:00
Patrick von Platen
7fd1febf38
Add "Leveraging Pretrained Checkpoints for Generation" Seq2Seq models. (#6594) 
* add conversion script

* improve conversion script

* make style

* add tryout files

* fix

* update

* add causal bert

* better names

* add tokenizer file as well

* finish causal_bert

* fix small bugs

* improve generate

* change naming

* renaming

* renaming

* renaming

* remove leftover files

* clean files

* add fix tokenizer

* finalize

* correct slow test

* update docs

* small fixes

* fix link

* adapt check repo

* apply sams and sylvains recommendations

* fix import

* implement Lysandres recommendations

* fix logger warn
 2020-09-10 16:40:51 +02:00
Sylvain Gugger
d1691d90e5
Samell fixed in tf template (#7044) 2020-09-10 10:36:02 -04:00
Patrick von Platen
63e539459d
Update README.md 2020-09-10 16:34:28 +02:00
Patrick von Platen
054db06b1b
Create README.md 2020-09-10 16:30:46 +02:00
Lysandre Debut
b482ad474a
Fix template (#7040) 2020-09-10 08:45:52 -04:00
Yu Liu
762cba3bda
Albert pretrain datasets/ datacollator (#6168) 
* add dataset for albert pretrain

* datacollator for albert pretrain

* naming, comprehension, file reading change

* data cleaning is no needed after this modification

* delete prints

* fix a bug

* file structure change

* add tests for albert datacollator

* remove random seed

* add back len and get item function

* sample file for testing and test code added

* format change for black

* more format change

* Style

* var assignment issue resolve

* add back wrongly deleted DataCollatorWithPadding in init file

* Style

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Lysandre <lysandre.debut@reseau.eseo.fr>
 2020-09-10 07:56:29 -04:00
Johann C. Rocholl
49e9be0639
Fix confusing warnings during TF2 import from PyTorch (#6623) 
1. Swapped missing_keys and unexpected_keys.

2. Copy&paste error caused these warnings to say "from TF 2.0" when it's actually "from PyTorch".
 2020-09-10 05:31:59 -04:00
Stas Bekman
4ee1053dcf
add -y to bypass prompt for transformers-cli upload (#7035) 2020-09-10 04:58:29 -04:00
Patrick von Platen
76818cc4c6
Create README.md 2020-09-09 16:26:35 +02:00
Lysandre Debut
15478c1287
Batch encore plus and overflowing tokens fails when non existing overflowing tokens for a sequence (#6677) 
* Patch and test

* Fix tests
 2020-09-09 06:55:17 -04:00
Henry Dashwood
9fd11bf1a8
replace torch.triu with onnx compatible code (#6929) 2020-09-09 04:56:40 -04:00
Julien Chaumond
ed71c21d6a
[from_pretrained] Allow tokenizer_type ≠ model_type (#6995) 2020-09-09 04:22:59 -04:00
Stas Bekman
03e363f9ae
[generation] consistently add eos tokens (#6982) 
Currently beam search returns inconsistent outputs - if hypos have different lengths we get eos, if they are the same - we don't.

This PR makes the output consistent.

Also why not also replace:

```
            if sent_lengths[i] < max_length:
                decoded[i, sent_lengths[i]] = eos_token_id
```
with:
```
            decoded[i, sent_lengths[i]] = eos_token_id
```
Shouldn't eos always be there? If the data gets truncated, the caller needs to user a larger `max_length`.

Please correct me if my logic is flawed.
 2020-09-09 04:08:36 -04:00
Stas Bekman
d0963486c1
adding TRANSFORMERS_VERBOSITY env var (#6961) 
* introduce TRANSFORMERS_VERBOSITY env var + test + test helpers

* cleanup

* remove helper function
 2020-09-09 04:08:01 -04:00
Sam Shleifer
f0fc0aea6b
pegasus.rst: fix expected output (#7017) 2020-09-08 13:29:16 -04:00
Patrick von Platen
120176ea29
[Longformer] Fix longformer documentation (#7016) 
* fix longformer

* allow position ids to not be initialized
 2020-09-08 18:51:28 +02:00
Lysandre Debut
5c4eb4b1ac
Fixing FLOPS merge by checking if torch is available (#7013) 
* Should check if `torch` is available

* fixed samples_count error, distributed_concat arguments

* style

* Import torch at beginning of file

Co-authored-by: TevenLeScao <teven.lescao@gmail.com>
 2020-09-08 10:51:58 -04:00
Teven
01d340adfa
Floating-point operations logging in trainer (#6768) 
* neFLOs calculation, logging, and reloading (#1)

* testing distributed consecutive batches

* fixed AttributeError from DataParallel

* removed verbosity

* rotate with use_mtime=True

* removed print

* fixed interaction with gradient accumulation

* indent formatting

* distributed neflo counting

* fixed typo

* fixed typo

* mean distributed losses

* exporting log history

* moved a few functions

* floating_point_ops clarification for transformers with parameter-reuse

* code quality

* double import

* made flo estimation more task-agnostic

* only logging flos if computed

* code quality

* unused import

* Update src/transformers/trainer.py

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

* Update src/transformers/modeling_utils.py

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

* Sylvain review

* Update src/transformers/modeling_utils.py

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

* black

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
 2020-09-08 10:00:56 -04:00
Sylvain Gugger
d155b38d6e
Funnel transformer (#6908) 
* Initial model

* Fix upsampling

* Add special cls token id and test

* Formatting

* Test and fist FunnelTokenizerFast

* Common tests

* Fix the check_repo script and document Funnel

* Doc fixes

* Add all models

* Write doc

* Fix test

* Initial model

* Fix upsampling

* Add special cls token id and test

* Formatting

* Test and fist FunnelTokenizerFast

* Common tests

* Fix the check_repo script and document Funnel

* Doc fixes

* Add all models

* Write doc

* Fix test

* Fix copyright

* Forgot some layers can be repeated

* Apply suggestions from code review

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Update src/transformers/modeling_funnel.py

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

* Address review comments

* Update src/transformers/modeling_funnel.py

Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Address review comments

* Update src/transformers/modeling_funnel.py

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

* Slow integration test

* Make small integration test

* Formatting

* Add checkpoint and separate classification head

* Formatting

* Expand list, fix link and add in pretrained models

* Styling

* Add the model in all summaries

* Typo fixes

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Sam Shleifer <sshleifer@gmail.com>
 2020-09-08 08:08:08 -04:00