修复

.gitignore

@@ -0,0 +1 @@
+test.py

app.py

@@ -1,6 +1,5 @@
 from spaces import GPU
 
-from os import getenv
 from threading import Thread
 from transformers import Qwen2VLForConditionalGeneration, Qwen2VLProcessor, TextIteratorStreamer, AutoProcessor, BatchFeature
 from qwen_vl_utils import process_vision_info
@@ -60,10 +59,9 @@ def respond(
 You are an intelligent assistant developed by the School of Software at Hefei University of Technology.
 You like to chat with people and help them solve problems."""})
         history.append({"role": "user", "content": message})
-        text_inputs = processor.apply_chat_template(history, tokenize=False, add_generation_prompt=True)
-    else:
-        text_inputs = processor.apply_chat_template(message, tokenize=False, add_generation_prompt=True)
-    image_inputs, video_inputs = process_vision_info(history)
+        message = history
+    text_inputs = processor.apply_chat_template(message, tokenize=False, add_generation_prompt=True)
+    image_inputs, video_inputs = process_vision_info(message)
     for response in infer((text_inputs, image_inputs, video_inputs), max_tokens, temperature, top_p):
         yield response