sandeep-krishnamurthy opened a new pull request #12786: Set correct update on kvstore flag in dist_device_sync mode URL: https://github.com/apache/incubator-mxnet/pull/12786 ## Description ## * Fix 1st part of the issue - https://github.com/apache/incubator-mxnet/issues/12713 * In dist mode, update_on_kvstore was getting overridden based on async/sparse weight, grad only. In this fix we handle - "dist_device_sync" and not ignoring kvstore_params i.e, kvstore_params['update_on_kvstore'] value. * Credits @eric-haibin-lin - https://github.com/eric-haibin-lin/mxnet/commit/856b5034e57633d888fcc1e3d82b8703945511ce ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [X] Changes are complete (i.e. I finished coding on this PR) - [X] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [X] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at http://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [X] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change ### Changes ### - Fix update_on_kvstore flag setting bug in Gluon trainer in dist mode. ## Comments ## - Test cases are passing. - [TODO] Run segmentation example in distributed mode to verify the fix in real-world case. @eric-haibin-lin
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
