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

Revert "Add mrope op fusion (#3509)" (#3562)

Browse Source 
This reverts commit 646c1db5d7.

this new ops may lead accuracy problem

### What this PR does / why we need it?

### Does this PR introduce _any_ user-facing change?

### How was this patch tested?

- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
This commit is contained in:
wangxiyuan
2025-10-20 20:19:24 +08:00
committed by GitHub
parent 068ed706c8
commit 0bf3f21a98
3 changed files with 4 additions and 123 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
vllm_ascend/utils.py
View File

@@ -517,8 +517,8 @@ def register_ascend_customop(vllm_config: Optional[VllmConfig] = None):
AscendReplicatedLinear,
AscendRowParallelLinear)
from vllm_ascend.ops.rotary_embedding import (
AscendDeepseekScalingRotaryEmbedding, AscendMRotaryEmbedding,
AscendRotaryEmbedding, AscendYaRNRotaryEmbedding)
AscendDeepseekScalingRotaryEmbedding, AscendRotaryEmbedding,
AscendYaRNRotaryEmbedding)
from vllm_ascend.ops.vocab_parallel_embedding import (
AscendLogitsProcessor, AscendParallelLMHead,
AscendVocabParallelEmbedding)
@@ -528,7 +528,6 @@ def register_ascend_customop(vllm_config: Optional[VllmConfig] = None):
"QuickGELU": AscendQuickGELU,
"SiluAndMul": AscendSiluAndMul,
"RotaryEmbedding": AscendRotaryEmbedding,
"MRotaryEmbedding": AscendMRotaryEmbedding,
"ColumnParallelLinear": AscendColumnParallelLinear,
"RowParallelLinear": AscendRowParallelLinear,
"YaRNScalingRotaryEmbedding": AscendYaRNRotaryEmbedding,