Unit test for TensorWeightedAvg RecMetric #3139
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
facebook-github-bot
added
the
CLA Signed
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D77169334 |
jeffkbkim
force-pushed
the
export-D77169334
branch
from
c9839b1 to
7557083
Compare
jeffkbkim
added a commit
to jeffkbkim/torchrec
that referenced
this pull request
Jun 26, 2025
Summary: Main test lies in test_tensor_weighted_avg.py This patch adds a unit test for TensorWeightedAvg metric. The existing framework for the rec metric testing module did not support processing tensors tied to a task's tensor_name, which the tensor weighted avg metric uses. Main modifications to other tests include changes to the TestMetric's _get_states() method signature to support processing arbitrary tensors in the testing framework. Rollback Plan: Differential Revision: D77169334
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D77169334 |
jeffkbkim
added a commit
to jeffkbkim/torchrec
that referenced
this pull request
Jun 26, 2025
Summary: Main test lies in `test_tensor_weighted_avg.py` This patch adds a unit test for TensorWeightedAvg metric. The existing framework for the rec metric testing module did not support processing tensors tied to a task's tensor_name, which the tensor weighted avg metric uses. Increases code coverage to 97.7%. [Previous external tests covered 35%](https://www.internalfb.com/quality/coverage?query=%7B%22key%22%3A%22AND%22%2C%22children%22%3A[%7B%22key%22%3A%22CONTAINS_ANY_OF_FBIDS%22%2C%22field%22%3A%22CODE_QUALITY_ONCALL%22%2C%22value%22%3A[%7B%22title%22%3A%22torchrec%20(torchrec%20%3A%20Zain%20Huda)%22%2C%22fbid%22%3A%22711323163133477%22%2C%22photo%22%3A%22%22%7D]%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Ffb%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22tests%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22__init__.py%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fgithub%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fcsrc%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fdatasets%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Finference%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fpt2%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22benchmark%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fmetrics%2F%22%7D]%7D&is_e2e=false&show_all_partitions=false&split_graph=true) Main modifications to other tests include changes to the TestMetric's _get_states() method signature to support processing arbitrary tensors in the testing framework. Differential Revision: D77169334
jeffkbkim
force-pushed
the
export-D77169334
branch
from
7557083 to
33b6180
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D77169334 |
jeffkbkim
added a commit
to jeffkbkim/torchrec
that referenced
this pull request
Jun 26, 2025
Summary: Main test lies in `test_tensor_weighted_avg.py` (format similar to `test_weighted_avg.py`) This patch adds a unit test for TensorWeightedAvgMetric. The existing framework for the rec metric testing module does not support processing tensors tied to a task's tensor_name, which the tensor weighted avg metric uses. Main modifications to other tests include changes to the TestMetric's _get_states() method signature to support processing arbitrary tensors in the testing framework. Increases code coverage to 97.7%. [Previous external tests covered 35%](https://www.internalfb.com/quality/coverage?query=%7B%22key%22%3A%22AND%22%2C%22children%22%3A[%7B%22key%22%3A%22CONTAINS_ANY_OF_FBIDS%22%2C%22field%22%3A%22CODE_QUALITY_ONCALL%22%2C%22value%22%3A[%7B%22title%22%3A%22torchrec%20(torchrec%20%3A%20Zain%20Huda)%22%2C%22fbid%22%3A%22711323163133477%22%2C%22photo%22%3A%22%22%7D]%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Ffb%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22tests%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22__init__.py%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fgithub%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fcsrc%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fdatasets%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Finference%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fpt2%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22benchmark%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fmetrics%2F%22%7D]%7D&is_e2e=false&show_all_partitions=false&split_graph=true) Differential Revision: D77169334
jeffkbkim
force-pushed
the
export-D77169334
branch
from
33b6180 to
24a2fd6
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D77169334 |
jeffkbkim
added a commit
to jeffkbkim/torchrec
that referenced
this pull request
Jun 26, 2025
Summary: Main test lies in `test_tensor_weighted_avg.py` (format similar to `test_weighted_avg.py`) This patch adds a unit test for TensorWeightedAvgMetric. The existing framework for the rec metric testing module does not support processing tensors tied to a task's tensor_name, which the tensor weighted avg metric uses. Main modifications to other tests include changes to the TestMetric's _get_states() method signature to support processing arbitrary tensors in the testing framework. Increases code coverage to 97.7%. [Previous external tests covered 35%](https://www.internalfb.com/quality/coverage?query=%7B%22key%22%3A%22AND%22%2C%22children%22%3A[%7B%22key%22%3A%22CONTAINS_ANY_OF_FBIDS%22%2C%22field%22%3A%22CODE_QUALITY_ONCALL%22%2C%22value%22%3A[%7B%22title%22%3A%22torchrec%20(torchrec%20%3A%20Zain%20Huda)%22%2C%22fbid%22%3A%22711323163133477%22%2C%22photo%22%3A%22%22%7D]%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Ffb%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22tests%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22__init__.py%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fgithub%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fcsrc%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fdatasets%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Finference%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fpt2%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22benchmark%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fmetrics%2F%22%7D]%7D&is_e2e=false&show_all_partitions=false&split_graph=true) Differential Revision: D77169334
jeffkbkim
force-pushed
the
export-D77169334
branch
from
24a2fd6 to
e3ca582
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D77169334 |
jeffkbkim
added a commit
to jeffkbkim/torchrec
that referenced
this pull request
Jun 27, 2025
Summary: Main test lies in `test_tensor_weighted_avg.py` (format similar to `test_weighted_avg.py`) This patch adds a unit test for TensorWeightedAvgMetric. TensorWeightedAvgMetric allows users to pass in a tensor that is not a prediction/weight/label with a corresponding tensor name, into a dict called "required_inputs". The purpose is to compute and emit the weighted average of a tensor of their choice. However, the existing framework for the rec metric testing module does not support processing tensors under "required_inputs". Main modifications to other tests include changes to the TestMetric's _get_states() method signature to support processing tensor passed into "required_inputs". Increases code coverage to 97.7%. [Previous external tests covered 35%](https://www.internalfb.com/quality/coverage?query=%7B%22key%22%3A%22AND%22%2C%22children%22%3A[%7B%22key%22%3A%22CONTAINS_ANY_OF_FBIDS%22%2C%22field%22%3A%22CODE_QUALITY_ONCALL%22%2C%22value%22%3A[%7B%22title%22%3A%22torchrec%20(torchrec%20%3A%20Zain%20Huda)%22%2C%22fbid%22%3A%22711323163133477%22%2C%22photo%22%3A%22%22%7D]%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Ffb%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22tests%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22__init__.py%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fgithub%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fcsrc%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fdatasets%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Finference%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fpt2%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22benchmark%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fmetrics%2F%22%7D]%7D&is_e2e=false&show_all_partitions=false&split_graph=true) Reviewed By: kausv Differential Revision: D77169334
jeffkbkim
force-pushed
the
export-D77169334
branch
from
e3ca582 to
0dd138e
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D77169334 |
jeffkbkim
added a commit
to jeffkbkim/torchrec
that referenced
this pull request
Jun 30, 2025
Summary: Main test lies in `test_tensor_weighted_avg.py` (format similar to `test_weighted_avg.py`) This patch adds a unit test for TensorWeightedAvgMetric. TensorWeightedAvgMetric allows users to pass in a tensor that is not a prediction/weight/label with a corresponding tensor name, into a dict called "required_inputs". The purpose is to compute and emit the weighted average of a tensor of their choice. However, the existing framework for the rec metric testing module does not support processing tensors under "required_inputs". Main modifications to other tests include changes to the TestMetric's _get_states() method signature to support processing tensor passed into "required_inputs". Increases code coverage to 97.7%. [Previous external tests covered 35%](https://www.internalfb.com/quality/coverage?query=%7B%22key%22%3A%22AND%22%2C%22children%22%3A[%7B%22key%22%3A%22CONTAINS_ANY_OF_FBIDS%22%2C%22field%22%3A%22CODE_QUALITY_ONCALL%22%2C%22value%22%3A[%7B%22title%22%3A%22torchrec%20(torchrec%20%3A%20Zain%20Huda)%22%2C%22fbid%22%3A%22711323163133477%22%2C%22photo%22%3A%22%22%7D]%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Ffb%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22tests%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22__init__.py%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fgithub%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fcsrc%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fdatasets%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Finference%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fpt2%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22benchmark%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fmetrics%2F%22%7D]%7D&is_e2e=false&show_all_partitions=false&split_graph=true) Reviewed By: kausv Differential Revision: D77169334
jeffkbkim
force-pushed
the
export-D77169334
branch
from
0dd138e to
5ce344e
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D77169334 |
jeffkbkim
added a commit
to jeffkbkim/torchrec
that referenced
this pull request
Jun 30, 2025
Summary: Pull Request resolved: meta-pytorch#3139 Main test lies in `test_tensor_weighted_avg.py` (format similar to `test_weighted_avg.py`) This patch adds a unit test for TensorWeightedAvgMetric. TensorWeightedAvgMetric allows users to pass in a tensor that is not a prediction/weight/label with a corresponding tensor name, into a dict called "required_inputs". The purpose is to compute and emit the weighted average of a tensor of their choice. However, the existing framework for the rec metric testing module does not support processing tensors under "required_inputs". Main modifications to other tests include changes to the TestMetric's _get_states() method signature to support processing tensor passed into "required_inputs". Increases code coverage to 97.7%. [Previous external tests covered 35%](https://www.internalfb.com/quality/coverage?query=%7B%22key%22%3A%22AND%22%2C%22children%22%3A[%7B%22key%22%3A%22CONTAINS_ANY_OF_FBIDS%22%2C%22field%22%3A%22CODE_QUALITY_ONCALL%22%2C%22value%22%3A[%7B%22title%22%3A%22torchrec%20(torchrec%20%3A%20Zain%20Huda)%22%2C%22fbid%22%3A%22711323163133477%22%2C%22photo%22%3A%22%22%7D]%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Ffb%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22tests%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22__init__.py%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fgithub%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fcsrc%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fdatasets%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Finference%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fpt2%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22benchmark%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fmetrics%2F%22%7D]%7D&is_e2e=false&show_all_partitions=false&split_graph=true) Reviewed By: kausv Differential Revision: D77169334
jeffkbkim
force-pushed
the
export-D77169334
branch
from
5ce344e to
b9737a2
Compare
Summary: Main test lies in `test_tensor_weighted_avg.py` (format similar to `test_weighted_avg.py`) This patch adds a unit test for TensorWeightedAvgMetric. TensorWeightedAvgMetric allows users to pass in a tensor that is not a prediction/weight/label with a corresponding tensor name, into a dict called "required_inputs". The purpose is to compute and emit the weighted average of a tensor of their choice. However, the existing framework for the rec metric testing module does not support processing tensors under "required_inputs". Main modifications to other tests include changes to the TestMetric's _get_states() method signature to support processing tensor passed into "required_inputs". Increases code coverage to 97.7%. [Previous external tests covered 35%](https://www.internalfb.com/quality/coverage?query=%7B%22key%22%3A%22AND%22%2C%22children%22%3A[%7B%22key%22%3A%22CONTAINS_ANY_OF_FBIDS%22%2C%22field%22%3A%22CODE_QUALITY_ONCALL%22%2C%22value%22%3A[%7B%22title%22%3A%22torchrec%20(torchrec%20%3A%20Zain%20Huda)%22%2C%22fbid%22%3A%22711323163133477%22%2C%22photo%22%3A%22%22%7D]%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Ffb%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22tests%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22__init__.py%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fgithub%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fcsrc%2F%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fdatasets%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Finference%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fpt2%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_NOT_CONTAINS_TEXT%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22benchmark%22%7D%2C%7B%22key%22%3A%22CODE_QUALITY_STARTS_WITH%22%2C%22field%22%3A%22CODE_QUALITY_FILE_PATH%22%2C%22value%22%3A%22fbcode%2Ftorchrec%2Fmetrics%2F%22%7D]%7D&is_e2e=false&show_all_partitions=false&split_graph=true) Reviewed By: kausv Differential Revision: D77169334
jeffkbkim
force-pushed
the
export-D77169334
branch
from
b9737a2 to
0b7a9d3
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D77169334 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Main test lies in
test_tensor_weighted_avg.py(format similar totest_weighted_avg.py)This patch adds a unit test for TensorWeightedAvgMetric.
TensorWeightedAvgMetric allows users to pass in a tensor that is not a prediction/weight/label with a corresponding tensor name, into a dict called "required_inputs". The purpose is to compute and emit the weighted average of a tensor of their choice.
However, the existing framework for the rec metric testing module does not support processing tensors under "required_inputs". Main modifications to other tests include changes to the TestMetric's _get_states() method signature to support processing tensor passed into "required_inputs".
Increases code coverage from 0% to 97.5%