[Misc] Improve error messages for unsupported types and parameters (#30593)

Signed-off-by: BlankR <hjyblanche@gmail.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>

vllm/model_executor/layers/quantization/auto_round.py

@@ -56,22 +56,22 @@ class AutoRoundConfig(QuantizationConfig):
         if weight_bits not in self.SUPPORTED_BITS:
             raise ValueError(
                 f"Unsupported weight_bits: {weight_bits}, "
-                f"currently only support  {self.SUPPORTED_BITS}"
+                f"currently only support {self.SUPPORTED_BITS}."
             )
         if data_type not in self.SUPPORTED_DTYPES:
             raise ValueError(
-                f"Unsupported data_type: {data_type},"
-                f" currently only support  {self.SUPPORTED_DTYPES}"
+                f"Unsupported data_type: {data_type}, "
+                f"currently only support {self.SUPPORTED_DTYPES}."
             )
         if packing_format not in self.SUPPORTED_FORMATS:
             raise ValueError(
                 f"Unsupported packing_format: {packing_format}, "
-                f"currently only support  {self.SUPPORTED_FORMATS}"
+                f"currently only support {self.SUPPORTED_FORMATS}."
             )
         if backend not in self.SUPPORTED_BACKENDS:
             raise ValueError(
-                f"Unsupported backend: {backend},  "
-                f"currently only support  {self.SUPPORTED_BACKENDS}"
+                f"Unsupported backend: {backend}, "
+                f"currently only support {self.SUPPORTED_BACKENDS}."
             )
 
         self.weight_bits = weight_bits

vllm/model_executor/layers/quantization/compressed_tensors/schemes/compressed_tensors_w8a8_fp8.py

@@ -158,7 +158,10 @@ class CompressedTensorsW8A8Fp8(CompressedTensorsScheme):
             input_scale = None
 
         else:
-            raise ValueError(f"Unknown quantization strategy {self.strategy}")
+            raise ValueError(
+                f"Unknown quantization strategy {self.strategy}: "
+                f"should be one of {list(QuantizationStrategy)}"
+            )
 
         # required by torch.compile to be torch.nn.Parameter
         layer.weight = Parameter(weight.data, requires_grad=False)

vllm/model_executor/layers/quantization/mxfp4.py

@@ -783,7 +783,10 @@ class Mxfp4MoEMethod(FusedMoEMethodBase):
             layer.w13_weight = w13_weight
             layer.w2_weight = w2_weight
         else:
-            raise ValueError(f"Unsupported backend: {self.mxfp4_backend}")
+            raise ValueError(
+                f"Unsupported mxfp4_backend: {self.mxfp4_backend}: "
+                f"should be one of: {list(Mxfp4Backend)}."
+            )
 
     def get_fused_moe_quant_config(
         self, layer: torch.nn.Module