@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / 1ef13cf92f
Fork 0

A high-throughput and memory-efficient inference and serving engine for LLMs

Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git

Synced Nov 16, 18:15 UTC

[Misc]Fix BitAndBytes exception messages (#7626)

Jee Jee Li committed Aug 18, 2024 at 03:02 UTC 1ef13cf92f42a086d14a202a594d7dcae524e640
1 file changed +2 -2
vllm/model_executor/model_loader/loader.py
+2 -2
@@ -883,11 +883,11 @@ class BitsAndBytesModelLoader(BaseModelLoader):
883 if not hasattr(model, 'load_weights'):
884 raise AttributeError(
885 "The required method 'load_weights' is not defined in class"
886 - f" {type(self).__name__}.")
886 + f" {type(model).__name__}.")
887
888 if not hasattr(model, 'bitsandbytes_stacked_params_mapping'):
889 raise AttributeError(
890 - f"Model {type(self).__name__} does not support BitsAndBytes "
890 + f"Model {type(model).__name__} does not support BitsAndBytes "
891 "quantization yet.")
892
893 logger.info("Loading weights with BitsAndBytes quantization. "