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.

Sign up for GitHub

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

Avoid limited memory adaptor issue in balanced KMeans #2570

Merged
merged 6 commits into from
Feb 13, 2025
Merged

Avoid limited memory adaptor issue in balanced KMeans #2570

Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
c4814ab
Use large workspace resource instead of workspace resource.
csadorf Feb 4, 2025
ca49029
Use current device resource instead of explicit managed memory resource.
csadorf Feb 4, 2025
da0ecb8
Rename corresponding variable names.
csadorf Feb 4, 2025
9f59617
Revert "Rename corresponding variable names."
csadorf Feb 5, 2025
b218c94
Revert "Use current device resource instead of explicit managed memor…
csadorf Feb 5, 2025
02be3c3
Update corresponding inline comment.
csadorf Feb 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions cpp/include/raft/cluster/detail/kmeans_balanced.cuh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -967,9 +967,10 @@ void build_hierarchical(const raft::resources& handle,
IdxT n_mesoclusters = std::min(n_clusters, static_cast<IdxT>(std::sqrt(n_clusters) + 0.5));
RAFT_LOG_DEBUG("build_hierarchical: n_mesoclusters: %u", n_mesoclusters);

// TODO: Remove the explicit managed memory- we shouldn't be creating this on the user's behalf.
// Need to use explicit managed_memory here since corresponding allocations
// must be both host and device accessible.
rmm::mr::managed_memory_resource managed_memory;
rmm::device_async_resource_ref device_memory = resource::get_workspace_resource(handle);
rmm::device_async_resource_ref device_memory = resource::get_large_workspace_resource(handle);
auto [max_minibatch_size, mem_per_row] =
calc_minibatch_size<MathT>(n_clusters, n_rows, dim, params.metric, std::is_same_v<T, MathT>);

Expand Down
Loading