chenchenghao/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 11 Packages Projects Releases Wiki Activity

FP4 weight loading and inference (2/2) (#3972)

Browse Source
This commit is contained in:
Trevor Morris
2025-04-08 17:26:21 -07:00
committed by GitHub
parent 5039d54772
commit 11d760d56a
6 changed files with 262 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
python/sglang/srt/layers/quantization/__init__.py
View File

@@ -59,7 +59,10 @@ from sglang.srt.layers.quantization.compressed_tensors.compressed_tensors import
)
from sglang.srt.layers.quantization.fp8 import Fp8Config
from sglang.srt.layers.quantization.gptq import GPTQConfig, GPTQMarlinConfig
from sglang.srt.layers.quantization.modelopt_quant import ModelOptFp8Config
from sglang.srt.layers.quantization.modelopt_quant import (
ModelOptFp4Config,
ModelOptFp8Config,
)
from sglang.srt.layers.quantization.moe_wna16 import MoeWNA16Config
from sglang.srt.layers.quantization.w8a8_fp8 import W8A8Fp8Config
from sglang.srt.layers.quantization.w8a8_int8 import W8A8Int8Config
@@ -69,6 +72,7 @@ BASE_QUANTIZATION_METHODS: Dict[str, Type[QuantizationConfig]] = {
"fp8": Fp8Config,
"blockwise_int8": BlockInt8Config,
"modelopt": ModelOptFp8Config,
"modelopt_fp4": ModelOptFp4Config,
"w8a8_int8": W8A8Int8Config,
"w8a8_fp8": W8A8Fp8Config,
"moe_wna16": MoeWNA16Config,