Skip to content

Conversation

@akhilg-nv
Copy link
Collaborator

@akhilg-nv akhilg-nv commented Jun 4, 2025

Naive performance test results indicate the new implementation is about 50% faster

LayerNorm2d_TP1 (Direct Implementation) took 0.0059s for 100 iterations
LayerNorm2d_TP2 (LayerNorm-based) took 0.0038s for 100 iterations

@akhilg-nv akhilg-nv force-pushed the dev-akhilg-layernorm2d branch from 345c099 to f0d3cb6 Compare June 4, 2025 23:08
@akhilg-nv akhilg-nv merged commit 3cab972 into main Jun 5, 2025
1 of 2 checks passed
@akhilg-nv akhilg-nv deleted the dev-akhilg-layernorm2d branch June 5, 2025 00:29
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants