Skip to content

Conversation

nipung90
Copy link
Contributor

@nipung90 nipung90 commented Oct 9, 2025

Summary:
In the KVZCH scenarios, the input_dist for row-wise sharding will have to deal with KVZCH and non-KVZCH features together. This means that the virtual_table_feature_num_buckets will have to represent non-KVZCH features too.
In this diff, we default the value of the num_buckets to world_size for non-KVZCH features.

Differential Revision: D84094039

@meta-cla meta-cla bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Oct 9, 2025
Copy link
Contributor

meta-codesync bot commented Oct 9, 2025

@nipung90 has exported this pull request. If you are a Meta employee, you can view the originating Diff in D84094039.

…ogether for row-wise sharding (meta-pytorch#3444)

Summary:

In the KVZCH scenarios, the input_dist for row-wise sharding will have to deal with KVZCH and non-KVZCH features together. This means that the virtual_table_feature_num_buckets will have to represent non-KVZCH features too.
In this diff, we default the value of the num_buckets to world_size for non-KVZCH features.

Differential Revision: D84094039
Copy link
Contributor

meta-codesync bot commented Oct 9, 2025

@nipung90 has exported this pull request. If you are a Meta employee, you can view the originating Diff in D84094039.

@nipung90 nipung90 added CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported meta-exported labels Oct 9, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported meta-exported
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant