{ROCm]: gpt-oss fusion/padding fixes (#38043)

Signed-off-by: Andreas Karatzas <akaratza@amd.com>
Signed-off-by: Rohan138 <rohanpotdar138@gmail.com>
Signed-off-by: Rohan Potdar <66227218+Rohan138@users.noreply.github.com>
Co-authored-by: Andreas Karatzas <akaratza@amd.com>
This commit is contained in:
Rohan Potdar
2026-03-27 12:19:15 -04:00
committed by GitHub
co-authored by Andreas Karatzas
parent 21d2b53f88
commit 0e9358c11d
3 changed files with 4 additions and 19 deletions
+1 -3
View File
@@ -152,13 +152,11 @@ def enable_rope_kvcache_fusion(cfg: "VllmConfig") -> bool:
def enable_norm_pad_fusion(cfg: "VllmConfig") -> bool:
"""Enable if using AITER RMSNorm and AITER Triton GEMMs
and hidden size is 2880 i.e. gpt-oss; otherwise Inductor handles fusion."""
"""Enable if using AITER RMSNorm and hidden size is 2880 i.e. gpt-oss."""
from vllm._aiter_ops import rocm_aiter_ops
return (
rocm_aiter_ops.is_rmsnorm_enabled()
and not rocm_aiter_ops.is_triton_gemm_enabled()
and cfg.model_config is not None
and cfg.model_config.get_hidden_size() == 2880
)