biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[PluggableLayer][1/N] Define PluggableLayer (Fix ci) (#32744)

Browse Source 
Signed-off-by: whx-sjtu <2952154980@qq.com>
This commit is contained in:
whx
2026-01-22 00:38:04 +08:00
committed by GitHub
parent 4e31b7f228
commit 1861ae8aae
7 changed files with 108 additions and 48 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
tests/kernels/utils.py
View File

@@ -13,7 +13,7 @@ import torch
from torch._prims_common import TensorLikeType
from tests.kernels.quant_utils import native_w8a8_block_matmul
from vllm.model_executor.custom_op import CustomOp
from vllm.model_executor.custom_op import op_registry
from vllm.model_executor.layers.activation import SiluAndMul
from vllm.model_executor.layers.fused_moe.utils import moe_kernel_quantize_input
from vllm.utils.torch_utils import make_tensor_with_pad
@@ -883,7 +883,7 @@ def torch_experts(
f32 = torch.float32
act = CustomOp.op_registry[activation]
act = op_registry[activation]
for i in range(num_experts):
mask = topk_ids == i