biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Bugfix] Get available quantization methods from quantization registry (#4098)

Browse Source
This commit is contained in:
Michael Goin
2024-04-18 03:21:55 -04:00
committed by GitHub
parent 66ded03067
commit 53b018edcb
6 changed files with 18 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
vllm/model_executor/layers/quantization/__init__.py
View File

@@ -7,7 +7,7 @@ from vllm.model_executor.layers.quantization.gptq import GPTQConfig
from vllm.model_executor.layers.quantization.marlin import MarlinConfig
from vllm.model_executor.layers.quantization.squeezellm import SqueezeLLMConfig
_QUANTIZATION_CONFIG_REGISTRY = {
QUANTIZATION_METHODS = {
"awq": AWQConfig,
"gptq": GPTQConfig,
"squeezellm": SqueezeLLMConfig,
@@ -16,12 +16,13 @@ _QUANTIZATION_CONFIG_REGISTRY = {
def get_quantization_config(quantization: str) -> Type[QuantizationConfig]:
if quantization not in _QUANTIZATION_CONFIG_REGISTRY:
if quantization not in QUANTIZATION_METHODS:
raise ValueError(f"Invalid quantization method: {quantization}")
return _QUANTIZATION_CONFIG_REGISTRY[quantization]
return QUANTIZATION_METHODS[quantization]
__all__ = [
"QuantizationConfig",
"get_quantization_config",
"QUANTIZATION_METHODS",
]