h2oai/h2o4gpu

View on GitHub

H2Oai GPU Edition

Stars: 468
Forks: 96
Open beginner issues: 0
Indexed issues: 146

Dominant language: C++
License: Apache-2.0
Last GitHub push: Jul 27, 2026
Latest indexed: Sep 17, 2026
Contributing guide: Contributing guide
Code of conduct: Code of conduct
Beginner labels: No beginner labels indexed
PR merge metrics: No merged PRs in 30d

Open All

146 open issues indexed

KMeans minibatch Open

C++ CUDA enhancement

h2oai/h2o4gpu#17 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
subkmeans for improved handling of high-dimensional data Open

C++ CUDA enhancement

h2oai/h2o4gpu#19 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Jenkins do all make tests Open

build

h2oai/h2o4gpu#21 · 3 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Jenkins TV board for accuracy and speed of h2o4gpu Open

build

h2oai/h2o4gpu#22 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
KMeans DAAL implementation. Open

C++ Python

h2oai/h2o4gpu#25 · 2 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Implement linear and general SVM Open

C++ CUDA Python R

h2oai/h2o4gpu#29 · 8 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Implement quantiles regression Open

C++ CUDA

h2oai/h2o4gpu#32 · 5 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Use Intel DAAL for CPU side of many algorithms Open

C++ enhancement Python

h2oai/h2o4gpu#33 · 4 comments · 0 reactions · 3 assignees · Indexed Sep 17, 2026
NCCL Support for multi GPU and multi machine Open

CUDA

h2oai/h2o4gpu#34 · 2 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Implement Kalman Filters Open

C++ CUDA

h2oai/h2o4gpu#35 · 5 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
GPU support for MOJO scoring Open

CUDA enhancement

h2oai/h2o4gpu#36 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Ying-Yang and other K-Means Open

CUDA enhancement kmeans

h2oai/h2o4gpu#43 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Connector to cuDF Open

CUDA

h2oai/h2o4gpu#44 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
GPU support for data.table Open

CUDA enhancement

h2oai/h2o4gpu#45 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
GLM C++/CUDA code Open

C++ CUDA glm refactoring

h2oai/h2o4gpu#47 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Logging framework Open

C++ CUDA Python

h2oai/h2o4gpu#48 · 13 comments · 0 reactions · 2 assignees · Indexed Sep 17, 2026
Progress bar for each solver Open

C++ enhancement

h2oai/h2o4gpu#51 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Use improved adaptive rho for GLM Open

enhancement glm

h2oai/h2o4gpu#52 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
automatic lambda_min_ratio search Open

enhancement glm

h2oai/h2o4gpu#55 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Improved automatic CPU/GPU selection Open

enhancement Python

h2oai/h2o4gpu#67 · 1 comment · 0 reactions · 3 assignees · Indexed Sep 17, 2026
Improved backend (CPU/GPU/TPU etc.) selection Open

enhancement Python

h2oai/h2o4gpu#68 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Docker image for developers Open

build community

h2oai/h2o4gpu#71 · 0 comments · 0 reactions · 2 assignees · Indexed Sep 17, 2026
Implement [ci skip] Open

build

h2oai/h2o4gpu#75 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Add backend/runtime configuration class Open

enhancement Python

h2oai/h2o4gpu#82 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Look into static code analysis tools Open

build enhancement

h2oai/h2o4gpu#92 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Need way to notify us when cub, py3nvml, xgboost, scikit-learn repo's are updated Open

build

h2oai/h2o4gpu#97 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
xgboost wheel should be built and tested and used separately Open

build

h2oai/h2o4gpu#112 · 1 comment · 0 reactions · 3 assignees · Indexed Sep 17, 2026
Also, can we have jenkins abort a build if another commit comes along -- especially useful for PRs when one makes minor fixes and updates the PR. Open

h2oai/h2o4gpu#114 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Include rest of API (including internals like examples show or is stated in code) for glm so more sklearn like Open

Python

h2oai/h2o4gpu#120 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Include rest of API (including internals like examples show or is stated in code) for kmeans, so more sklearn like Open

Python

h2oai/h2o4gpu#121 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Ensure no copies of data made in python if correct expected types Open

performance

h2oai/h2o4gpu#122 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
glm probabilities to classes Open

glm

h2oai/h2o4gpu#130 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Need glm to have score() Open

glm

h2oai/h2o4gpu#138 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
GPU kmeans needs to be compared with h2o-3 or sklearn or intel daal. Open

kmeans

h2oai/h2o4gpu#141 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Add testperf to Jenkins pipeline Open

h2oai/h2o4gpu#142 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
glm: return absolute best of alpha and lambda based upon appropriat error Open

glm

h2oai/h2o4gpu#146 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Need tests as part of code so can (like sklearn) just run model.tests.testname and runs the test fully self-contained Open

h2oai/h2o4gpu#150 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Notes from H2O4GPU meeting: 9/11/17 Open

h2oai/h2o4gpu#159 · 2 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Fallback to CPU/sklearn when GPU fails (e.g. out of memory) Open

h2oai/h2o4gpu#161 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
side-by-side scikit-learn and h2o4gpu Open

documentation

h2oai/h2o4gpu#162 · 2 comments · 0 reactions · 3 assignees · Indexed Sep 17, 2026
glm _ptr needs to save X or validPreds in GPU memory so can return GPU pointers Open

glm

h2oai/h2o4gpu#163 · 0 comments · 0 reactions · 2 assignees · Indexed Sep 17, 2026
kmeans needs _ptr and return results only on GPU as pointers as well Open

kmeans

h2oai/h2o4gpu#164 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
User should be able to specify early stopping metric in GLM Open

glm

h2oai/h2o4gpu#171 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Need to improve accuracy on make testperf (9/12/17) Open

performance

h2oai/h2o4gpu#173 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
need new jenkins to allow git submodule update --depth 1 <module name>? Open

h2oai/h2o4gpu#176 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
KMeans centroid averaging on GPUs Open

CUDA enhancement kmeans performance

h2oai/h2o4gpu#178 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Volta performance twice as slow as pascal for glm and kmeans Open

performance

h2oai/h2o4gpu#191 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Put xgboost and py3nvml into h2o4gpu module Open

h2oai/h2o4gpu#198 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
you can add a auditwheel check in jenkins to ensure that the wheel is portable Open

h2oai/h2o4gpu#202 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Benchmark Scikit tree models vs XGBoost Open

benchmark

h2oai/h2o4gpu#207 · 0 comments · 0 reactions · 2 assignees · Indexed Sep 17, 2026

h2oai/h2o4gpu

Get new issues in your inbox