Unverified Commit fbb5f229 authored by Yineng Zhang's avatar Yineng Zhang Committed by GitHub
Browse files

fix awq_dequantize import (#5669)

parent 15fabcc0
......@@ -3,7 +3,6 @@ import logging
from typing import Any, Dict, List, Optional
import torch
from sgl_kernel import awq_dequantize
from sglang.srt.layers.linear import (
LinearBase,
......@@ -12,6 +11,11 @@ from sglang.srt.layers.linear import (
)
from sglang.srt.layers.parameter import GroupQuantScaleParameter, PackedvLLMParameter
from sglang.srt.layers.quantization.base_config import QuantizationConfig
from sglang.srt.utils import is_cuda
_is_cuda = is_cuda()
if _is_cuda:
from sgl_kernel import awq_dequantize
logger = logging.getLogger(__name__)
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment