Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[backport] Fix compiling with the latest CTX. #10263

Merged
merged 1 commit into from
May 10, 2024

Conversation

trivialfis
Copy link
Member

No description provided.

@trivialfis trivialfis merged commit 1fdcd2b into dmlc:release_2.0.0 May 10, 2024
12 of 26 checks passed
@trivialfis trivialfis deleted the backport-cuda-fix branch May 10, 2024 17:27
razdoburdin added a commit to IntelPython/xgboost_oneapi that referenced this pull request May 15, 2024
* [backport][sklearn] Fix loading model attributes. (dmlc#9808) (dmlc#9880)

* [backport][py] Use the first found native library. (dmlc#9860) (dmlc#9879)

* [backport] [CI] Upload libxgboost4j.dylib (M1) to S3 bucket (dmlc#9887)

* [CI] Set up CI for Mac M1 (dmlc#9699)

* [CI] Improve CI for Mac M1 (dmlc#9748)

* [CI] Build libxgboost4j.dylib with CMAKE_OSX_DEPLOYMENT_TARGET (dmlc#9749)

* [CI] Upload libxgboost4j.dylib (M1) to S3 bucket (dmlc#9886)

* [jvm-packages] Fix POM for xgboost-jvm metapackage (dmlc#9893)

* [jvm-packages] Fix POM for xgboost-jvm metapackage

* Add script for updating the Scala version

* Bump version to 2.0.3 (dmlc#9895)

* Update change_scala_version.py to also change scala.version property (dmlc#9897)

* [backport] Fix compiling with the latest CTX. (dmlc#10263)

---------

Co-authored-by: Jiaming Yuan <[email protected]>
Co-authored-by: Philip Hyunsu Cho <[email protected]>
trivialfis added a commit to ZiyueXu77/xgboost that referenced this pull request May 29, 2024
trivialfis added a commit to trivialfis/xgboost that referenced this pull request Jun 18, 2024
trivialfis added a commit that referenced this pull request Jun 18, 2024
* Revert "Update the federated branch. (#10409)"

This reverts commit 922cca3.

* Revert "Update federated branch. (#10317)"

This reverts commit b491e09.

* [backport] Fix compiling with the latest CTX. (#10263)
trivialfis added a commit that referenced this pull request Jun 18, 2024
* Add additional data split mode to cover the secure vertical pipeline

* Add IsSecure info and update corresponding functions

* Modify evaluate_splits to block non-label owners to perform hist compute under secure scenario

* Continue using Allgather for best split sync for secure vertical, equvalent to broadcast

* Modify histogram sync scheme for secure vertical case, can identify global best split, but need to further apply split correctly

* Sync cut informaiton across clients, full pipeline works for testing case

* Code cleanup, phase 1 of alternative vertical pipeline finished

* Code clean

* change kColS to kColSecure to avoid confusion with kCols

* Replace allreduce with allgather, functional but inefficient version

* Update AllGather behavior from individual pair to bulk by adopting new histogram transmission data structure of a flat vector

* comment out the record printing

* fix pointer bug for histsync with allgather

* identify the HE adding locations

* revise and simplify template code

* revise and simplify template code

* prepare aggregator for gh broadcast

* prepare histogram for histindex and row index for allgather

* fix conflicts

* fix conflicts

* fix format

* fix allgather logic and update unit test

* fix linting

* fix linting and other unit test issues

* fix linting and other unit test issues

* integration with interface initial attempt

* integration with interface initial attempt

* integration with interface initial attempt

* functional integration with interface

* remove debugging prints

* remove processor from another PR

* Update the processor functions according to new processor implementation

* Move processor interface init from learner to communicator

* Move processor interface init from learner to communicator functional

* switch to allgatherV for encrypted message with varying lenghts

* consolidate with processor interface PR

* remove prints and fix format

* fix linting over reference pass

* fix undefined symbol issue

* fix processor test

* secure vertical relies on processor, move the unit test

* type correction

* type correction

* extra linting from last change

* Added Windows support

* fix for cstdint types

* fix for cstdint types

* Added support for horizontal secure XGBoost

* update with mock plugin

* secure horizontal fully functional with mock plugin

* linting fix

* linting fix

* linting fix

* fix type

* change loader and proc params input pattern to align with std map

* update with secure vertical incorporation

* Update mock_processor to enable nvflare usage

* [backport] Fix compiling with the latest CTX. (#10263)

* fix secure horizontal inference

* initialized aggr context only once

* Added support for multiple plugins in a single lib

* remove redundant condition

* Added support for boolean in proc_params

* free buffer

* CUDA.

* Fix clean build.

* Fix include.

* tidy.

* lint.

* nolint.

* disable.

* disable sanitizer.

---------

Co-authored-by: Zhihong Zhang <[email protected]>
Co-authored-by: Jiaming Yuan <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant