Imperceptible asr bug detach

[davidslater]

Description

When running imperceptible ASR attack on pytorch_deep_speech on GPU, I sometimes run into this error:

File "/opt/conda/lib/python3.8/site-packages/art/attacks/evasion/imperceptible_asr/imperceptible_asr_pytorch.py", line 567, in _attack_2nd_stage[best_loss_2nd_stage[local_batch_size_idx] = loss_2nd_stage[local_batch_size_idx].numpy()
		RuntimeError^[[0m:^[[1m Can't call numpy() on Tensor that requires grad. Use tensor.detach().numpy() instead

I believe it only occurs sometimes due to specific branching conditions being hit.
See also twosixlabs/armory#1472

This update just ensures that when moving from Tensor to numpy array, .detach().cpu().numpy() is always called. This should always be safe (detach and cpu are no-ops if they are already true).

Fixes # (issue)

Type of change

Please check all relevant options.

• Improvement (non-breaking)
• Bug fix (non-breaking)
• New feature (non-breaking)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

Testing

Please describe the tests that you ran to verify your changes. Consider listing any relevant details of your test configuration.

• Test A
• Test B

Test Configuration:

• OS
• Python version
• ART version or commit number
• TensorFlow / Keras / PyTorch / MXNet version

Checklist

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

[codecov-commenter]

Codecov Report

Merging #1677 (b3b8cf9) into dev_1.10.2 (1fd1f92) will decrease coverage by 0.74%.
The diff coverage is 0.00%.

@@              Coverage Diff               @@
##           dev_1.10.2    #1677      +/-   ##
==============================================
- Coverage       88.10%   87.35%   -0.75%     
==============================================
  Files             259      259              
  Lines           21349    21349              
  Branches         3789     3789              
==============================================
- Hits            18809    18649     -160     
- Misses           1597     1755     +158     
- Partials          943      945       +2     

Impacted Files	Coverage Δ
...ion/imperceptible_asr/imperceptible_asr_pytorch.py	86.79% <0.00%> (ø)
...estimators/speech_recognition/tensorflow_lingvo.py	18.27% <0.00%> (-73.10%)	⬇️
art/estimators/certification/abstain.py	90.90% <0.00%> (-9.10%)	⬇️
art/estimators/poison_mitigation/strip/strip.py	94.44% <0.00%> (-5.56%)	⬇️
art/estimators/tensorflow.py	77.77% <0.00%> (-2.03%)	⬇️
art/attacks/evasion/brendel_bethge.py	85.99% <0.00%> (-1.63%)	⬇️
art/attacks/evasion/boundary.py	92.77% <0.00%> (-1.21%)	⬇️
art/utils.py	74.95% <0.00%> (-0.51%)	⬇️
art/defences/detector/poison/activation_defence.py	84.28% <0.00%> (-0.34%)	⬇️

[beat-buesser]

Hi @davidslater Thank you very much! Your solution looks good to me.