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

feat: patch linear base (#2915)

Browse Source
This commit is contained in:
Yineng Zhang
2025-01-16 18:00:03 +08:00
committed by GitHub
parent ab31793661
commit bf8d07a6f9
6 changed files with 68 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
python/sglang/srt/layers/quantization/fp8.py
View File

@@ -9,7 +9,6 @@ from torch.nn import Module
from torch.nn.parameter import Parameter
from vllm import _custom_ops as ops
from vllm.distributed import get_tensor_model_parallel_world_size
from vllm.model_executor.layers.linear import LinearBase
from vllm.model_executor.layers.quantization.kv_cache import BaseKVCacheMethod
from vllm.model_executor.layers.quantization.utils.marlin_utils_fp8 import (
apply_fp8_marlin_linear,
@@ -25,7 +24,11 @@ from vllm.model_executor.layers.quantization.utils.w8a8_utils import (
requantize_with_max_scale,
)
from sglang.srt.layers.linear import LinearMethodBase, UnquantizedLinearMethod
from sglang.srt.layers.linear import (
LinearBase,
LinearMethodBase,
UnquantizedLinearMethod,
)
from sglang.srt.layers.parameter import ModelWeightParameter, PerTensorScaleParameter
from sglang.srt.layers.quantization.base_config import (
QuantizationConfig,