modify style & little change

Former-commit-id: c988477d14dc656450d5fec31895781b7f9f7dce

src/llamafactory/chat/hf_engine.py

@@ -165,8 +165,17 @@ class HuggingfaceEngine(BaseEngine):
         )
 
         mm_inputs = template.mm_plugin.get_mm_inputs(**mm_input_dict, seqlens=[prompt_length], processor=processor)
+
         for key, value in mm_inputs.items():
-            value = value if isinstance(value, torch.Tensor) else torch.tensor(value)
+            value = (
+                value
+                if isinstance(value, torch.Tensor)
+                else (
+                    torch.stack(value)
+                    if isinstance(value, list) and all(isinstance(v, torch.Tensor) for v in value)
+                    else torch.tensor(value)
+                )
+            )
             gen_kwargs[key] = value.to(model.device)
 
         return gen_kwargs, prompt_length