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

[CI] Fix broken CI (#6599)

Browse Source 
Revert
4fb3d5e1b2
it breaks E2E Test

- vLLM version: v0.15.0
- vLLM main:
d7e17aaacd
This commit is contained in:
wangxiyuan
2026-02-06 17:23:58 +08:00
committed by GitHub
parent 19b5d44ea8
commit 06c0aed124
17 changed files with 1147 additions and 947 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
vllm_ascend/ops/__init__.py
View File

@@ -27,7 +27,8 @@ if HAS_TRITON:
import vllm_ascend.ops.vocab_parallel_embedding # noqa
from vllm_ascend.ops.activation import AscendQuickGELU, AscendSiluAndMul
from vllm_ascend.ops.rotary_embedding import AscendDeepseekScalingRotaryEmbedding, AscendRotaryEmbedding
from vllm_ascend.ops.rotary_embedding import (
AscendDeepseekScalingRotaryEmbedding, AscendRotaryEmbedding)
class dummyFusionOp:
@@ -39,13 +40,23 @@ class dummyFusionOp:
def register_dummy_fusion_op() -> None:
torch.ops._C_ascend.rms_norm = dummyFusionOp(name="rms_norm")
torch.ops._C_ascend.fused_add_rms_norm = dummyFusionOp(name="fused_add_rms_norm")
torch.ops._C_ascend.static_scaled_fp8_quant = dummyFusionOp(name="static_scaled_fp8_quant")
torch.ops._C_ascend.dynamic_scaled_fp8_quant = dummyFusionOp(name="dynamic_scaled_fp8_quant")
torch.ops._C_ascend.dynamic_per_token_scaled_fp8_quant = dummyFusionOp(name="dynamic_per_token_scaled_fp8_quant")
torch.ops._C_ascend.rms_norm_static_fp8_quant = dummyFusionOp(name="rms_norm_static_fp8_quant")
torch.ops._C_ascend.fused_add_rms_norm_static_fp8_quant = dummyFusionOp(name="fused_add_rms_norm_static_fp8_quant")
torch.ops._C_ascend.rms_norm_dynamic_per_token_quant = dummyFusionOp(name="rms_norm_dynamic_per_token_quant")
torch.ops._C_ascend.fused_add_rms_norm = dummyFusionOp(
name="fused_add_rms_norm")
torch.ops._C_ascend.static_scaled_fp8_quant = dummyFusionOp(
name="static_scaled_fp8_quant")
torch.ops._C_ascend.dynamic_scaled_fp8_quant = dummyFusionOp(
name="dynamic_scaled_fp8_quant")
torch.ops._C_ascend.dynamic_per_token_scaled_fp8_quant = dummyFusionOp(
name="dynamic_per_token_scaled_fp8_quant")
torch.ops._C_ascend.rms_norm_static_fp8_quant = dummyFusionOp(
name="rms_norm_static_fp8_quant")
torch.ops._C_ascend.fused_add_rms_norm_static_fp8_quant = dummyFusionOp(
name="fused_add_rms_norm_static_fp8_quant")
torch.ops._C_ascend.rms_norm_dynamic_per_token_quant = dummyFusionOp(
name="rms_norm_dynamic_per_token_quant")
__all__ = ["AscendQuickGELU", "AscendSiluAndMul", "AscendRotaryEmbedding", "AscendDeepseekScalingRotaryEmbedding"]
__all__ = [
"AscendQuickGELU", "AscendSiluAndMul", "AscendRotaryEmbedding",
"AscendDeepseekScalingRotaryEmbedding"
]