EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

Replace sglang.srt.layers.quantization.scalar_types with sgl_kernel.scalar_type (#8951)

Browse Source
This commit is contained in:
Hongbo Xu
2025-08-14 10:41:41 +08:00
committed by GitHub
parent 6b7c24712c
commit a669bc2f74
8 changed files with 44 additions and 362 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
python/sglang/test/test_marlin_moe.py
View File

@@ -4,9 +4,9 @@ from typing import Optional
import pytest
import torch
from sgl_kernel import fused_marlin_moe
from sgl_kernel.scalar_type import ScalarType, scalar_types
from sglang.srt.layers.activation import SiluAndMul
from sglang.srt.layers.quantization.scalar_type import ScalarType, scalar_types
from sglang.test.test_marlin_utils import awq_marlin_quantize, marlin_quantize

2
python/sglang/test/test_marlin_utils.py
View File

@@ -10,13 +10,13 @@ from typing import Optional
import numpy as np
import torch
from sgl_kernel.scalar_type import ScalarType
from sglang.srt.layers.quantization.marlin_utils import (
GPTQ_MARLIN_TILE,
marlin_permute_scales,
marlin_zero_points,
)
from sglang.srt.layers.quantization.scalar_type import ScalarType
from sglang.srt.layers.quantization.utils import (
get_pack_factor,
gptq_quantize_weights,