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

[refactor] remove redundant code in linear

Browse Source
This commit is contained in:
Li Wei
2025-12-24 12:02:09 +08:00
parent 75d0bdae2f
commit 383eb5459a
8 changed files with 37 additions and 1570 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
vllm_kunlun/models/qwen3_next.py
View File

@@ -33,11 +33,11 @@ from vllm_kunlun.ops.fused_moe.layer import FusedMoE
from vllm.model_executor.layers.layernorm import (
GemmaRMSNorm as Qwen3NextRMSNorm)
# yapf: enable
from vllm_kunlun.ops.linear import (ColumnParallelLinear,
from vllm.model_executor.layers.linear import (ColumnParallelLinear,
MergedColumnParallelLinear,
QKVParallelLinear,
RowParallelLinear)
from vllm_kunlun.ops.linear import ReplicatedLinear
from vllm.model_executor.layers.linear import ReplicatedLinear
from vllm.model_executor.layers.logits_processor import LogitsProcessor
from vllm.model_executor.layers.mamba.abstract import MambaBase
from vllm.model_executor.layers.mamba.mamba_mixer2 import (