Add GPTQ support (#916)

vllm/model_executor/models/mpt.py

@@ -297,6 +297,9 @@ class MPTForCausalLM(nn.Module):
         params_dict = dict(self.named_parameters(remove_duplicate=False))
         for name, loaded_weight in hf_model_weights_iterator(
                 model_name_or_path, cache_dir, load_format, revision):
+            # Skip loading extra bias for GPTQ models.
+            if name.endswith(".bias") and name not in params_dict:
+                continue
             param = params_dict[name]
             weight_loader = getattr(param, "weight_loader",
                                     default_weight_loader)