Add gradient monitor callback

[dxoigmn]

What does this PR do?

This adds a gradient monitor callback so that one can make monitor the norm of different gradients.

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.

• No tests

Before submitting

• The title is self-explanatory and the description concisely explains the PR
• My PR does only one thing, instead of bundling different changes together
• I list all the breaking changes introduced by this pull request
• I have commented my code
• 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
• I have run pre-commit hooks with pre-commit run -a command without errors

Did you have fun?

Make sure you had fun coding 🙃

[mzweilin]

Do you have an example? I tried to run this command, but I didn't see gradient logs in Tensorboard.

CUDA_VISIBLE_DEVICES=0 \
python -m mart \
experiment=CIFAR10_CNN_Adv \
fit=false \
trainer=gpu \
+trainer.limit_test_batches=1 \
+callbacks@model.modules.input_adv_test.callbacks=gradient_monitor \
+model.modules.input_adv_test.callbacks.gradient_monitor.frequency=1

Update: patch and example in #164

[dxoigmn]

There are no parameters in an Adversary ;) This is why I think it's better to modify the state_dict rather than playing tricks by hiding parameters in lists.

[mzweilin]

There are no parameters in an Adversary ;) This is why I think it's better to modify the state_dict rather than playing tricks by hiding parameters in lists.

Resolved in #160

[mzweilin]

LGTM