Anupam comments

Results 33 comments of


                                            Anupam

Moving dice distance from raft to cuvs and support for half types

Adding @rhdong as a reviewer since he worked on half types for the other distances https://github.com/rapidsai/cuvs/pull/225

Moving dice distance from raft to cuvs and support for half types

Hi @cjnolet, sure where can I check the binary size/metrics for this pr?

Moving dice distance from raft to cuvs and support for half types

Thank you @bdice! It would be nice if there was a way to highlight in the html report which files are edited from the PR and the increase or decrease...

Moving dice distance from raft to cuvs and support for half types

To follow up here are the binary size reports. The `libcuvs` size increases by 6.06 MB * (1209.31 MB) Nightly https://downloads.rapids.ai/nightly/cuvs/2025-07-08/23c6dc0/cuda12_x86_64.compile_lib.html * (1215.37 MB) This PR https://downloads.rapids.ai/ci/cuvs/pull-request/971/d91ffdb/cuda12_x86_64.compile_lib.html ``` --- Added...

Moving dice distance from raft to cuvs and support for half types

This one is not targeting 25.12, and yep will update to add the new header format.

[dask-cuda] Memory corruption and worker crashes in cuML KMeans with large datasets

Looks like this one has a MRE now, so we can move to "Todo".

[FEA] Support Out of Core Kmeans Clustering

Any updates on this issue, and whether we still need more information?

Proposal: Accelerate Non-Negative Matrix Factorization (NMF) needed for Dimensionality Reduction

Thank you @aaalce for the detailed description and we are excited that you are interested in contributing! Please let us know of any further questions you may have as you...

[FEA] Support for Dice coefficient as a metric in UMAP

Hi @beckernick, do you know whether having the dice metric is still useful in UMAP or if we have any specific users that are asking for it? cc @csadorf

Prepare cuml for removal of deprecated raft apis

I've updated the PR so that pre and post processing code should functionally be the same. The goal is to remove the dependency on `processing.cuh` from raft which was implicitly...