EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[Performance] Use forward_native for Conv3dLayer and add UT (#8375)

Browse Source 
What this PR does / why we need it?
switch Ascend conv3d forward_oot to use forward_native and  add ut

Does this PR introduce any user-facing change?
No

How was this patch tested?
by CI

---------

Signed-off-by: zouyizhou <zouyizhou@huawei.com>
This commit is contained in:
zyz111222
2026-04-20 17:20:40 +08:00
committed by GitHub
parent c124e8df07
commit dd7e08c6db
3 changed files with 64 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm_ascend/utils.py
View File

@@ -662,6 +662,7 @@ def register_ascend_customop(vllm_config: VllmConfig | None = None):
if is_310p():
from vllm_ascend._310p.fused_moe.fused_moe import AscendFusedMoE310, AscendSharedFusedMoE310
from vllm_ascend._310p.ops.activation import AscendSiluAndMul310
from vllm_ascend._310p.ops.conv import AscendConv3dLayer310
from vllm_ascend._310p.ops.layernorm import (
AscendGemmaRMSNorm310,
AscendRMSNorm310,
@@ -686,6 +687,7 @@ def register_ascend_customop(vllm_config: VllmConfig | None = None):
"ParallelLMHead": AscendParallelLMHead310,
"VocabParallelEmbedding": AscendVocabParallelEmbedding310,
"MMEncoderAttention": AscendMMEncoderAttention310,
"Conv3dLayer": AscendConv3dLayer310,
}
)