diff --git a/vllm/model_executor/kernels/linear/mixed_precision/exllama.py b/vllm/model_executor/kernels/linear/mixed_precision/exllama.py index 537a8e278a..3ad43a225f 100644 --- a/vllm/model_executor/kernels/linear/mixed_precision/exllama.py +++ b/vllm/model_executor/kernels/linear/mixed_precision/exllama.py @@ -59,6 +59,13 @@ class ExllamaLinearKernel(MPLinearKernel): f"{cls.SUPPORTED_QUANT_TYPES}", ) + if c.group_size <= 0: + return ( + False, + f"Group size ({c.group_size}) must be positive, " + "Exllama does not support channelwise quantization", + ) + if c.full_weight_shape[0] % c.group_size != 0: return ( False,