Skip to content
build

utilize the cuda graph to optimize the kernel overhead #3493

Sign in to view logs

utilize the cuda graph to optimize the kernel overhead

utilize the cuda graph to optimize the kernel overhead #3493

Triggered via pull request February 27, 2024 13:05
@lambda7xxlambda7xx
synchronize #1313
cuda_graph_0215_0223
Status Success
Total duration 1h 21m 39s
Artifacts 1

build.yml

on: pull_request
Build FlexFlow with the Makefile
45m 31s
Build FlexFlow with the Makefile
Matrix: Build FlexFlow with CMake
Fit to window
Zoom out
Zoom in

Annotations

3 warnings
Build FlexFlow with the Makefile
Node.js 16 actions are deprecated. Please update the following actions to use Node.js 20: actions/checkout@v3, Jimver/[email protected], conda-incubator/setup-miniconda@v2. For more information see: https://github.blog/changelog/2023-09-22-github-actions-transitioning-from-node-16-to-node-20/.
Build FlexFlow with CMake (cuda)
Node.js 16 actions are deprecated. Please update the following actions to use Node.js 20: actions/checkout@v3, Jimver/[email protected], conda-incubator/setup-miniconda@v2. For more information see: https://github.blog/changelog/2023-09-22-github-actions-transitioning-from-node-16-to-node-20/.
Build FlexFlow with CMake (hip_rocm)
Node.js 16 actions are deprecated. Please update the following actions to use Node.js 20: actions/checkout@v3, conda-incubator/setup-miniconda@v2. For more information see: https://github.blog/changelog/2023-09-22-github-actions-transitioning-from-node-16-to-node-20/.

Artifacts

Produced during runtime
Name Size
install-log Expired
545 KB