Jin Yong Yoo
3a375a3d8e
finish trainer with cmd-line command
2021-06-04 01:47:00 -04:00
Jin Yong Yoo
4fbb09c9ac
WIP: Trainer class feature complete
2021-06-04 01:47:00 -04:00
Mint Lin
429254d873
add output scale factor assignment
2021-06-04 01:45:11 -04:00
Will Peterson
1f709ff0fa
Adding Tutorial 4: Tensorflow, Tutorial 8: Attacking BERT models w/ CAptum
2021-06-04 01:45:11 -04:00
Will Peterson
25d3296979
Adapted Examples: 2, 6 (AllenNLP), 7 (Multilingual) to be compatible with attack-api. Still need to finish revising code for Examples 0, 8
2021-06-04 01:33:17 -04:00
Jin Yong Yoo
21054c99c2
fix wandb logging bug
2021-06-04 01:33:17 -04:00
Jin Yong Yoo
cc5ff46226
fix formatting
2021-06-04 01:33:16 -04:00
Jin Yong Yoo
3c40ce8605
add text window for masked LM transformations
2021-06-04 01:33:16 -04:00
Jin Yong Yoo
a32263a20e
fix bug with HFDataset
2021-06-04 01:33:15 -04:00
Jin Yong Yoo
fccf60dfd4
WIP: Add an example for parallel attack
2021-06-04 01:32:33 -04:00
Jin Yong Yoo
58d6ea6652
WIP: Get parallel attacks working
2021-06-04 01:32:33 -04:00
Jin Yong Yoo
de88a5ee23
WIP: finish attack args, attacker, cli commands
2021-06-04 01:32:33 -04:00
Jin Yong Yoo
d394d048ac
WIP: completed attack, dataset, model, logging args
2021-06-04 01:31:09 -04:00
Jin Yong Yoo
65010cf562
WIP: reactor cmd line args
2021-06-04 01:31:09 -04:00
Jin Yong Yoo
2744fd3ca8
WIP: add datasets and attacker
2021-06-04 01:25:49 -04:00
Jin Yong Yoo
10f54a16f4
add datasets
2021-06-04 01:25:48 -04:00
Jack Morris
4015ad23b4
Merge pull request #448 from Ashiq5/Ashiq5-patch-1
...
Proposed solution to Issue #447
2021-05-30 10:28:28 -04:00
Md. Ishtiaq Ashiq
3f224bce50
formatting fixed
2021-05-29 13:59:27 -04:00
Jack Morris
937ba0d6b8
release v0.2.16
2021-05-23 22:18:45 -04:00
Jack Morris
d5e9048a48
Merge pull request #466 from QData/qwerty-bug
...
return [] when not in adj list
2021-05-23 20:50:10 -04:00
Jack Morris
5030d52997
Merge pull request #467 from QData/max-logic-fix
...
UntargetedClassification: Don't check self.target_max_score when it is already known to be None
2021-05-23 20:49:54 -04:00
Jack Morris
c7e15279a0
simplify (self.x or y) to just y when self.x is None
2021-05-23 20:08:05 -04:00
Jack Morris
723bf6c191
return [] when not in adj list
2021-05-23 19:59:54 -04:00
Jack Morris
6950621a95
Merge pull request #465 from QData/run-training-bug
...
Run training bug
2021-05-23 19:49:59 -04:00
Jack Morris
e9fde9912e
Merge pull request #464 from QData/fast-alzantot-doc-fix
...
Rename all instances of faster-alzantot to fast-alzantot
2021-05-23 19:49:47 -04:00
Jack Morris
fe8bd660e4
Merge pull request #461 from FrancoMuniz/patch-1
...
Quick fix for issue #424
2021-05-23 19:45:13 -04:00
Jack Morris
facad249d8
args.global_step -> global_step
2021-05-23 19:44:35 -04:00
Jack Morris
ed303a58a1
Rename all instances of faster-alzantot to fast-alzantot
2021-05-23 19:37:59 -04:00
Franco Muñiz
b8a5830c7f
Quick fix for issue #424
...
https://github.com/QData/TextAttack/issues/424
2021-05-10 17:53:28 -03:00
Jin Yong Yoo
10309a32de
Merge pull request #452 from QData/doc0414
...
fix minor documentation errors
2021-04-23 23:12:20 +09:00
U-DESKTOP-99E94MF\matrix
fa4ccb2f97
fix all_words_diff problem
2021-04-22 08:54:47 +08:00
Yanjun Qi
1021c5643e
Update benchmark-search.md
2021-04-21 12:42:18 -04:00
ashiq
dc45bb346d
minor
2021-04-15 22:54:13 -04:00
Md. Ishtiaq Ashiq
5453e1e05f
Issue #447
2021-04-14 19:16:34 -04:00
Md. Ishtiaq Ashiq
ef72936a83
Issue #447
...
Checking if the model_from_file argument exists or not
2021-04-14 19:14:28 -04:00
Yanjun Qi / Jane
4da3a257ab
Merge pull request #446 from QData/doc0414
...
add in talks-visualization.md for our tutorial talks on TextAttack
2021-04-14 15:17:45 -04:00
Yanjun Qi
12b76d3c3b
add in talks-visualization.md for our tutorial talks on TextAttack
2021-04-14 14:04:57 -04:00
Yanjun Qi / Jane
ae68c81411
Update README.md
...
update citation to EMNLP version
2021-04-09 09:05:54 -04:00
Yanjun Qi / Jane
8e18c3c850
Merge pull request #442 from P3n9W31/example_5_explain_BERT_fix
...
Fix errors in Example_5_Explain_BERT
2021-04-08 23:09:05 -04:00
ZhanPw
aa736d26b4
Fix errors in Example_5_Explain_BERT
...
calculate(input_ids, token_type_ids, attention_mask) miss an important parameter 'position_ids' of Huggingface BERT model, lead to wrong results (logit and Visualization). Fix it.
get_text() is just never called. Delete it.
deepcopy of the model is unnecessary. Delete it.
colored text can not display correctly in notebook. Fix it.
A lot of useless test code. Delete them.
Attribution score of perturbed text is calculated wrongly. Fix it.
Some attribute methods that are commented out cannot be used directly and may be misleading. Deleted and provided a link to captum document.
2021-04-09 00:43:56 +08:00
Yanjun Qi / Jane
6c3359a33d
Merge pull request #441 from QData/doc_configure_issues
...
Update requirements.txt
2021-04-05 14:11:22 -04:00
Yanjun Qi
ac306c7973
Update setup.py
2021-04-05 13:03:48 -04:00
Yanjun Qi
3b6de0b28a
locally test all passed...
2021-04-05 12:56:59 -04:00
Yanjun Qi
62a99ae7de
Update requirements.txt
2021-04-05 12:07:12 -04:00
Yanjun Qi
e6b3709b31
Revert "correct wrong library mentions of docs"
...
This reverts commit 678cf4ba93
2021-04-05 11:55:32 -04:00
Yanjun Qi
9f84732045
Revert "Update requirements.txt"
...
This reverts commit d04c740f11
2021-04-05 11:55:28 -04:00
Yanjun Qi
d04c740f11
Update requirements.txt
2021-04-05 11:40:00 -04:00
Yanjun Qi
678cf4ba93
correct wrong library mentions of docs
2021-04-05 11:22:55 -04:00
Yanjun Qi / Jane
749c023c61
Merge pull request #440 from nithvijay/cuda-installation-documentation
...
CUDA installation documentation
2021-04-02 18:23:38 -04:00
Yanjun Qi / Jane
9b7fb83ab4
Merge pull request #433 from cogeid/new-tutorial-custom
...
New tutorial on custom dataset & word embedding
2021-04-02 18:23:14 -04:00
