fix: Simplify model selection logic in message regeneration

Modify the model selection process to directly use the appropriate model based on the conversation context, removing redundant conditional logic

src/renderer/src/pages/home/Messages/MessageMenubar.tsx

@@ -282,7 +282,7 @@ const MessageMenubar: FC<Props> = (props) => {
   const onRegenerate = async (e: React.MouseEvent | undefined) => {
     e?.stopPropagation?.()
     await modelGenerating()
-    const selectedModel = message.model || (isGrouped ? model : assistantModel)
+    const selectedModel = isGrouped ? model : assistantModel
     const _message = resetAssistantMessage(message, selectedModel)
     onEditMessage?.(_message)
   }