alihan/textattack-nlp-transformer
1
0
Fork 0
mirror of https://github.com/QData/TextAttack.git synced 2021-10-13 00:05:06 +03:00
Code Issues Projects Releases Wiki Activity
1,763 Commits 20 Branches 13 Tags
278bb33b587a110d9cc64a37bf78d1d49b21c97c
Commit Graph

217 Commits

Author SHA1 Message Date
sanchit97
b0892ce7fe Remove model wrapping module 2020-11-12 21:17:55 -05:00
sanchit97
f0e27d129c Change model to Huggingface 2020-11-12 14:50:40 -05:00
sanchit97
e04c3777b3 Fix for tut0 2020-11-12 14:49:44 -05:00
Jin Yong Yoo
2df8833965 fix training code bug 2020-11-11 13:47:10 -05:00
Jin Yong Yoo
b148bd2818 fix bugs 2020-11-10 21:45:38 -05:00
Jin Yong Yoo
2328b7f76c Merge branch 'master' into lazy-loading 2020-11-10 19:21:17 -05:00
Yanjun Qi
de0b5b8b8b add in the docstring for textattack_cli main func 2020-11-10 13:32:21 -05:00
Jin Yong Yoo
6273b19c19 make WordSwapGradientBased work with get_grad 2020-11-01 00:58:16 -04:00
Jin Yong Yoo
7a89aba559 make some dependencies optional 2020-10-30 17:31:00 -04:00
Jin Yong Yoo
7dcf27d7bd add lazy loading 2020-10-30 12:17:52 -04:00
Yanjun Qi
c78aa66980 fix all flake8 found whitespacing trailing issues 2020-10-26 08:18:55 -04:00
Yanjun Qi
23d0de6b43 clean up the docstring of each module and module content section 2020-10-25 23:15:07 -04:00
Yanjun Qi
2ea690bbc6 a major shift to rst files generated by sphinx-apidoc 
major docstring clean up / plus reorganize the folder structure under docs
 2020-10-25 20:21:43 -04:00
Jack Morris
5a2f7b3bd5 Merge pull request #298 from QData/new-augmenter 
New augmenter
 2020-10-18 22:24:06 -04:00
Hanyu Liu
9b2fc8df24 more tests 2020-10-18 02:09:09 -04:00
Hanyu Liu
f3557c4180 updates 2020-10-18 01:45:10 -04:00
k-ivey
c9b1bcabdd Add CoLA based constraint 
Added a constraint that uses a model pre-trained on the CoLA dataset to check that the attacked text has a certain number of linguistically acceptable sentences as compared to the reference text
 2020-10-16 17:15:03 -04:00
Florian Over
1aa8bac7ba Added a new Constrain MultilingualUniversalSentenceEncoder 2020-10-11 21:13:14 +02:00
Hanyu Liu
90bfe90e83 better presentation 2020-10-07 02:41:59 -04:00
Hanyu Liu
22f36b594c added helpful interative function 2020-10-07 02:27:03 -04:00
Jack Morris
47d15acea9 Merge pull request #284 from QData/INV-Augmentation 
CheckList-Inv augmentation
 2020-09-30 11:30:25 -04:00
Jin Yong Yoo
e496280286 fix missed change 2020-09-30 00:49:46 -04:00
Jin Yong Yoo
88024e1f49 change --dataset-from-datasets to --dataset-from-huggingface 2020-09-30 00:33:08 -04:00
Jin Yong Yoo
428b19a511 change all mentions of nlp to datasets 2020-09-28 22:37:21 -04:00
Hanyu Liu
cd611887dc New Augmentation Recipe 2020-09-18 00:35:45 -04:00
Hanyu Liu
9fa5874144 Improvements after Review 2020-09-14 10:40:21 -04:00
Hanyu Liu
96c5feac64 Add CheckList Recipe 2020-09-01 10:35:25 -04:00
Jack Morris
af7f31d265 update black formatting 2020-08-30 15:25:56 -04:00
Jack Morris
b014a75f9d consistent random seed 2020-08-30 14:49:50 -04:00
Jack Morris
4bdaff94cb eval test stderr 2020-08-25 14:28:00 -04:00
Jack Morris
e2217914b6 fix eval and add tests 2020-08-25 13:46:29 -04:00
Jin Yong Yoo
f691ede779 fix gpu allocation 2020-08-08 14:32:25 -04:00
Jack Morris
436b79de66 update docs for PR; bump version for pypi; fix tqdm in colab 2020-08-07 12:33:38 -04:00
Jack Morris
3416cd3828 format imports 2020-08-06 11:46:05 -04:00
Jack Morris
0291a6fbdb | -> ^ 2020-08-06 10:30:30 -04:00
Jack Morris
ae575cfeb5 attack recipes -> classes 2020-08-06 07:42:47 -04:00
Jack Morris
6cd3a2d9a5 parallel fix; add --check-robustness training option 2020-08-04 13:49:20 -04:00
Jin Yong (Jeffrey) Yoo
aacbed4009 Add reseeding for parallel workers 2020-07-30 14:35:08 +09:00
Jack Morris
b589baf985 update test - : -> | 2020-07-29 15:40:03 -04:00
Jack Morris
ea3911fb4c Merge pull request #226 from QData/augmentation-interactive 
Augment Interactively
 2020-07-29 15:37:57 -04:00
Jack Morris
fe90ceaf9a fix imports 2020-07-29 15:10:44 -04:00
Hanyu Liu
c7cfa42bcf Better Visuals & Commands 
added spaces, lines, tabs for a clearer interface

more detailed user instructions

when in mode 'c' and asked to change recipe, users can type in 'r' to see all available recipes (do we need this?)
 2020-07-27 22:54:47 -04:00
Hanyu Liu
6271fe9edc Relocate augmenter statement 
Create augmenter before loops or when "c" is typed by the user; this saves time since creating augmenter in each loop could be expensive
 2020-07-27 21:57:05 -04:00
Hanyu Liu
ee05ca0f17 If-else Change 
Change from "if not -- else" to " if --else" for clearness

Change "if" to "elif"
 2020-07-27 21:21:21 -04:00
Hanyu-Liu-123
0560d4f80f Fix Error 
make --csv and --input-column optional when running in interactive mode
 2020-07-27 13:28:18 -04:00
Jack Morris
618e815eea update model training code 2020-07-27 10:40:19 -04:00
Jack Morris
29432d88c8 fix hotflip & model wrappers 2020-07-26 15:56:36 -04:00
Jack Morris
e6d671e918 remove dead imports; model wrappers work 2020-07-26 13:57:27 -04:00
Hanyu-Liu-123
7aa407496f Initial Functions 
Run augmentations interactively just like running attacks
 2020-07-24 18:42:18 -04:00
Hanyu-Liu-123
f5e1a731d2 Update Comments 2020-07-24 18:16:53 -04:00