Skip to content

[Kernel][Triton][AMD] Use block size heuristic for avg 2.8x speedup for int8 models #2112

[Kernel][Triton][AMD] Use block size heuristic for avg 2.8x speedup for int8 models

[Kernel][Triton][AMD] Use block size heuristic for avg 2.8x speedup for int8 models #2112

Annotations

1 warning

update-description

succeeded Jan 3, 2025 in 6s