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

[Model] Support common fused moe ops for moe model, such as Qwen3Moe (#709)

Browse Source 
vllm-ascend now only support moe for deepseek. We should add common moe
support back

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
This commit is contained in:
wangxiyuan
2025-04-28 21:57:01 +08:00
committed by GitHub
parent 40bd602485
commit 1fce70a2fb
2 changed files with 68 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
vllm_ascend/ops/__init__.py
View File

@@ -19,6 +19,7 @@ import torch
import torch_npu # noqa: F401
import vllm_ascend.ops.activation # noqa
import vllm_ascend.ops.common_fused_moe # noqa
import vllm_ascend.ops.fused_moe # noqa
import vllm_ascend.ops.layernorm # noqa
import vllm_ascend.ops.rotary_embedding # noqa