awslabs / sagemaker-debugger

Amazon SageMaker Debugger provides functionality to save tensors during training of machine learning jobs and analyze those tensors
Apache License 2.0
158 stars 82 forks source link

Fix rules #565

Closed mariumof closed 2 years ago

mariumof commented 2 years ago

Description of changes:

Style and formatting:

I have run pre-commit install && pre-commit run --all-files to ensure that auto-formatting happens with every commit.

Issue number, if available

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

codecov-commenter commented 2 years ago

Codecov Report

Merging #565 (2ff37db) into master (22fbec2) will decrease coverage by 4.16%. The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #565      +/-   ##
==========================================
- Coverage   73.40%   69.24%   -4.17%     
==========================================
  Files         127      117      -10     
  Lines       11171    10693     -478     
==========================================
- Hits         8200     7404     -796     
- Misses       2971     3289     +318     
Impacted Files Coverage Δ
smdebug/mxnet/__init__.py 0.00% <0.00%> (-100.00%) :arrow_down:
smdebug/mxnet/singleton_utils.py 0.00% <0.00%> (-100.00%) :arrow_down:
smdebug/mxnet/hook.py 0.00% <0.00%> (-80.30%) :arrow_down:
smdebug/rules/action/message_action.py 13.25% <0.00%> (-72.29%) :arrow_down:
smdebug/mxnet/collection.py 3.12% <0.00%> (-68.75%) :arrow_down:
smdebug/rules/action/action.py 30.61% <0.00%> (-61.23%) :arrow_down:
smdebug/tensorflow/reductions.py 0.00% <0.00%> (-47.06%) :arrow_down:
smdebug/rules/action/stop_training_action.py 15.62% <0.00%> (-39.07%) :arrow_down:
smdebug/mxnet/graph.py 0.00% <0.00%> (-22.59%) :arrow_down:
smdebug/pytorch/utils.py 54.16% <0.00%> (-20.84%) :arrow_down:
... and 26 more

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 22fbec2...2ff37db. Read the comment docs.