[Bugfix] Fix transformers model impl ignored for mixtral quant (#18602)

Signed-off-by: Tristan Leclercq <tristanleclercq@gmail.com>

vllm/model_executor/model_loader/utils.py

@@ -225,17 +225,16 @@ def get_model_architecture(
         "fp8", "compressed-tensors", "gptq_marlin", "awq_marlin", "quark"
     ]
 
-    if (model_config.quantization is not None
-            and model_config.quantization not in mixtral_supported
-            and "MixtralForCausalLM" in architectures):
-        architectures = ["QuantMixtralForCausalLM"]
-
     vllm_supported_archs = ModelRegistry.get_supported_archs()
     vllm_not_supported = not any(arch in vllm_supported_archs
                                  for arch in architectures)
     if (model_config.model_impl == ModelImpl.TRANSFORMERS or
             model_config.model_impl != ModelImpl.VLLM and vllm_not_supported):
         architectures = resolve_transformers_arch(model_config, architectures)
+    elif (model_config.quantization is not None
+          and model_config.quantization not in mixtral_supported
+          and "MixtralForCausalLM" in architectures):
+        architectures = ["QuantMixtralForCausalLM"]
 
     model_cls, arch = ModelRegistry.resolve_model_cls(architectures)
     if model_config.task == "embed":