update rai-core-flask by closing WSGIServer on separate thread to fix segfault errors and get the rai-core-flask release builds working again #1852
Conversation
This file contains 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
Codecov Report
@@ Coverage Diff @@
## main #1852 +/- ##
===========================================
+ Coverage 79.00% 89.35% +10.35%
===========================================
Files 19 118 +99
Lines 1324 5927 +4603
===========================================
+ Hits 1046 5296 +4250
- Misses 278 631 +353
Flags with carried forward coverage won't be shown. Click here to find out more.
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. |
1 similar comment
|
this pin might not be needed, just the other fix: |
imatiach-msft
force-pushed
the
ilmat/add-upper-bounds-greenlet-gevent-flask
branch
from
c320537
to
38ace98
Compare
…ase fails until they are more stable
imatiach-msft
force-pushed
the
ilmat/add-upper-bounds-greenlet-gevent-flask
branch
from
38ace98
to
5a0d209
Compare
imatiach-msft
changed the title
gaugup
approved these changes
Dec 16, 2022
imatiach-msft
deleted the
ilmat/add-upper-bounds-greenlet-gevent-flask
branch
gaugup
pushed a commit
that referenced
this pull request
Dec 19, 2022
…ase fails until they are more stable (#1852)
gaugup
pushed a commit
that referenced
this pull request
Jan 6, 2023
…ase fails until they are more stable (#1852)
3 tasks
gaugup
added a commit
that referenced
this pull request
Jan 6, 2023
* Add missing property 'id' in ICounterfactualData Signed-off-by: Gaurav Gupta <[email protected]> * Update some sample notebooks to include dropped_features (#1836) * update some notebooks to include dropped_features * update some notebooks to include dropped_features * update some notebooks to include dropped_features * remove unused changes * remove unused changes * address comments * upgrade setup-python github action from v2 to v4 and pin to older ubuntu image for python 3.6 to fix build error (#1838) * Gate call to probability distribution end point (#1839) Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Revert "Update some sample notebooks to include dropped_features (#1836)" (#1837) This reverts commit d89b03a. * fix cf error on unable to update text (#1842) * Disable partial dependence plots for large data scenarios (#1843) Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Remove redundant call to serialize_json_safe() in raiwidgets causal tests (#1840) Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Fetching global causal effects and global policy from SDK backend (#1753) * Add callback for global causal effects Signed-off-by: Gaurav Gupta <[email protected]> * Query SDK backend to get global causal effects Signed-off-by: Gaurav Gupta <[email protected]> * Fix bug in causal serialization Signed-off-by: Gaurav Gupta <[email protected]> * Adding callback for causal ploicy Signed-off-by: Gaurav Gupta <[email protected]> * propagate global and local effects only Signed-off-by: Gaurav Gupta <[email protected]> * Add support for getting global causal effects from SDK backend Signed-off-by: Gaurav Gupta <[email protected]> * lint fix Signed-off-by: Gaurav Gupta <[email protected]> * Fix lint Signed-off-by: Gaurav Gupta <[email protected]> * Fix Signed-off-by: Gaurav Gupta <[email protected]> * Fix bug Signed-off-by: Gaurav Gupta <[email protected]> * Modify the cohort message Signed-off-by: Gaurav Gupta <[email protected]> * Move to state Signed-off-by: Gaurav Gupta <[email protected]> * Fix imports Signed-off-by: Gaurav Gupta <[email protected]> * Remove console.log Signed-off-by: Gaurav Gupta <[email protected]> * Change the if condition Signed-off-by: Gaurav Gupta <[email protected]> * Add postga build trigger (#1755) (#1756) Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Add model wrapper for wrapping predictions and test data (#1762) * Add model wrapper for wrapping predictions and test data Signed-off-by: Gaurav Gupta <[email protected]> * Fix failing tests Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Fix lint Signed-off-by: Gaurav Gupta <[email protected]> * Remove force update Signed-off-by: Gaurav Gupta <[email protected]> * Fix lint Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Update requirements-linting.txt (#1847) * Remove typo casual and correcting it to causal (#1846) Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Disable local causal chart for large data support (#1845) * Disable local causal chart for large data support Signed-off-by: Gaurav Gupta <[email protected]> * Fix lint Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * add multilabel text example and fix correct and incorrect groups in data analysis (#1848) * Add number of cohort samples inside generateMetrics() (#1849) * Add number of cohort samples inside generateMetrics() Signed-off-by: Gaurav Gupta <[email protected]> * Yarn lintfix Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * add multilabel RAI vision example for fridge multilabel dataset (#1844) * fix install command for speech to text example (#1856) * Send the correct column name to SDK backend for model overview distribution charts (#1835) Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Derive fairness metrics from metric stats instead of cohort stats (#1854) Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * add upper bounds to greenlet and gevent in rai-core-flask as 5.0 release fails until they are more stable (#1852) * fix loading older versions of RAIInsights due to _FEATURE_METADATA key error (#1857) * Update maintainers (#1860) * Update requirements.txt in `raiwidgets` and `responsibleai` to pin numpy upper bound (#1864) * Update requirements.txt * Update requirements.txt * Port `cohort.py` from `raiwidgets` to `raiutils` (#1859) * Port cohort.py from raiwidgets to raiutils Signed-off-by: Gaurav Gupta <[email protected]> * Add more constants Signed-off-by: Gaurav Gupta <[email protected]> * Add constants.py Signed-off-by: Gaurav Gupta <[email protected]> * Make constants public Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Add common APIs for model overview metrics (#1862) * Add metric SDK APIs for Model Overview metrics Signed-off-by: Gaurav Gupta <[email protected]> * remove console.log() Signed-off-by: Gaurav Gupta <[email protected]> * Port `cohort.py` from `raiwidgets` to `raiutils` (#1859) * Port cohort.py from raiwidgets to raiutils Signed-off-by: Gaurav Gupta <[email protected]> * Add more constants Signed-off-by: Gaurav Gupta <[email protected]> * Add constants.py Signed-off-by: Gaurav Gupta <[email protected]> * Make constants public Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Fix lint Signed-off-by: Gaurav Gupta <[email protected]> * lint fix Signed-off-by: Gaurav Gupta <[email protected]> * Fix lint Signed-off-by: Gaurav Gupta <[email protected]> * Fix lint Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Update requirements-linting.txt (#1861) * update raiwidgets to rai-core-flask v0.5.0 (#1851) * fix error in erroranalysis heatmap caused by max value over right side of max bin (#1866) * remove dead space in aggregate feature importance (#1869) * display tree branching decisions (#1870) * Remove warning for 'dropped_features are not in use currently.' (#1871) * display tree branching decisions * Remove warning for 'dropped_features are not in use currently.' * turn off fail-fast for python tests (#1878) * add multilabel model overview metrics (#1882) * Show branching decisions on hover over on unselected tree nodes (#1881) * show tree label on hover over * address comments * Update some sample notebooks to include dropped_features (#1836) * update some notebooks to include dropped_features * update some notebooks to include dropped_features * update some notebooks to include dropped_features * remove unused changes * remove unused changes * address comments * Revert "Update some sample notebooks to include dropped_features (#1836)" (#1837) This reverts commit d89b03a. * Update requirements-linting.txt (#1847) * Add number of cohort samples inside generateMetrics() (#1849) * Add number of cohort samples inside generateMetrics() Signed-off-by: Gaurav Gupta <[email protected]> * Yarn lintfix Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Update requirements-linting.txt (#1861) * add multilabel model overview metrics (#1882) * Update `raiutils` in `raiwidgets`, `responsibleai`, `erroranalysis` and `nlp_feature_extractors` (#1873) Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Fix test Signed-off-by: Gaurav Gupta <[email protected]> * Undo changes Signed-off-by: Gaurav Gupta <[email protected]> * Fix lint Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> Co-authored-by: tongy-msft <[email protected]> Co-authored-by: Ilya Matiach <[email protected]> Co-authored-by: Roman Lutz <[email protected]>
gaugup
added a commit
that referenced
this pull request
Jan 9, 2023
* Refactor metric computation in ModelOverview Signed-off-by: Gaurav Gupta <[email protected]> * Move metrics to state management Signed-off-by: Gaurav Gupta <[email protected]> * add multilabel RAI vision example for fridge multilabel dataset (#1844) * fix install command for speech to text example (#1856) * Send the correct column name to SDK backend for model overview distribution charts (#1835) Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * Derive fairness metrics from metric stats instead of cohort stats (#1854) Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> * add upper bounds to greenlet and gevent in rai-core-flask as 5.0 release fails until they are more stable (#1852) * fix loading older versions of RAIInsights due to _FEATURE_METADATA key error (#1857) * Fix build Signed-off-by: Gaurav Gupta <[email protected]> * Change cohort updation and deletion logic Signed-off-by: Gaurav Gupta <[email protected]> * Fix lint Signed-off-by: Gaurav Gupta <[email protected]> * Fix code review comments Signed-off-by: Gaurav Gupta <[email protected]> * Add delete logic for selected cohort index Signed-off-by: Gaurav Gupta <[email protected]> * Fix code review comments part-2 Signed-off-by: Gaurav Gupta <[email protected]> * Remove unnecessary condition Signed-off-by: Gaurav Gupta <[email protected]> Signed-off-by: Gaurav Gupta <[email protected]> Co-authored-by: Ilya Matiach <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Description
Update: instead of pinning, I found that closing WSGIServer on a separate thread seems to fix the segfaults, and doing the pinning is not necessary. See related stackoverflow about graceful shutdown of WSGIServer:
https://stackoverflow.com/questions/18277048/gevent-pywsgi-graceful-shutdown
=====
rai-core-flask package release seems to be failing on errors in tests from some segfaults: https://github.com/microsoft/responsible-ai-toolbox/actions/runs/3699568450
It seems the latest versions of greenlet v2.0 and gevent are less stable and fail with segfaults. Pinning here to <2.0 and corresponding gevent version until newer packages are released that are more stable.
Checklist