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

Fix circular imports in gptq.py and unblock test explorer (#4736)

Browse Source
This commit is contained in:
Stefan He
2025-03-24 18:07:08 -07:00
committed by GitHub
parent 77cf771ebe
commit 4c584fc632
3 changed files with 3 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
python/sglang/srt/layers/quantization/gptq.py
View File

@@ -6,7 +6,6 @@ import torch
from sglang.srt.layers.linear import LinearBase from sglang.srt.layers.linear import LinearBase
from sglang.srt.layers.quantization.base_config import QuantizationConfig from sglang.srt.layers.quantization.base_config import QuantizationConfig
from sglang.srt.layers.vocab_parallel_embedding import ParallelLMHead
from sglang.srt.utils import is_cuda from sglang.srt.utils import is_cuda
_is_cuda = is_cuda() _is_cuda = is_cuda()
@@ -434,6 +433,9 @@ class MarlinConfig(QuantizationConfig):
from vllm.model_executor.layers.quantization.marlin import MarlinLinearMethod from vllm.model_executor.layers.quantization.marlin import MarlinLinearMethod
# Delay import to avoid circular dependency
from sglang.srt.layers.vocab_parallel_embedding import ParallelLMHead
if isinstance(layer, LinearBase) or ( if isinstance(layer, LinearBase) or (
isinstance(layer, ParallelLMHead) and self.lm_head_quantized isinstance(layer, ParallelLMHead) and self.lm_head_quantized
): ):

0
python/sglang/test/__init__.py Normal file
View File

0
python/sglang/test/attention/__init__.py Normal file
View File