Fix side-effect to internvl (#2778)

* qwen2-vl

* internvl

* qwen2

* get image_tokens_per_patch for internvl2

* deepseek-vl

* cogvlm

* glm4v

* update internvl

* internvl_llava

* llava

* glm4v

* upate internvl

* cogvlm

* deepseek

* llava_hf

* rollback llava, internvl-llava

* refactor qwen

* update internvl

* update llava_hf

* update qwen2-vl

* llava_next

* update llava_next

* update llava

* update llava

* update llava

* qwen2

* fix internvl

lmdeploy/vl/model/internvl.py

@@ -245,7 +245,7 @@ def proc_messages(cls, messages, chat_template, sequence_start):
                 prompt = f'<img>{IMAGE_TOKEN * n_images}</img>\n' + prompt
             else:
                 pass
-            prompt_messages.append(dict(role='user', content=content))
+            prompt_messages.append(dict(role='user', content=prompt))
         prompt = chat_template.messages2prompt(prompt_messages, sequence_start)
         return prompt, IMAGE_TOKEN