Update app.py

app.py

@@ -1,26 +1,39 @@
 import gradio as gr
 from transformers import AutoProcessor, AutoModelForCausalLM
 import torch
-from PIL import Image
+import os
+from huggingface_hub import login
 
-MODEL_ID = "Qwen/Qwen1.5-VL-Chat"
+# ✅ 设置你的 Hugging Face Token
+HF_TOKEN = os.environ.get("HF_TOKEN")# ← 这里替换为你的 token
+login(token=HF_TOKEN)
 
-processor = AutoProcessor.from_pretrained(MODEL_ID, trust_remote_code=True)
-model = AutoModelForCausalLM.from_pretrained(MODEL_ID, device_map="auto", trust_remote_code=True, torch_dtype=torch.bfloat16)
+# ✅ 指定模型
+MODEL_ID = "Qwen/Qwen-VL-Chat"
 
-def qwen_vl_chat(image, question):
-    inputs = processor(text=question, images=image, return_tensors="pt").to(model.device)
-    outputs = model.generate(**inputs, max_new_tokens=256)
-    answer = processor.batch_decode(outputs, skip_special_tokens=True)[0]
-    return answer
+# 加载模型
+processor = AutoProcessor.from_pretrained(MODEL_ID, trust_remote_code=True, token=HF_TOKEN)
+model = AutoModelForCausalLM.from_pretrained(MODEL_ID, trust_remote_code=True, device_map="auto", token=HF_TOKEN)
+model.eval()
 
+# 推理函数
+def ask(image, prompt):
+    inputs = processor.from_list_format([
+        {"image": image},
+        {"text": prompt}
+    ])
+    inputs = processor(inputs, return_tensors="pt").to(model.device)
+    outputs = model.generate(**inputs, max_new_tokens=512)
+    response = processor.batch_decode(outputs, skip_special_tokens=True)[0]
+    return response
+
+# Gradio 页面
 demo = gr.Interface(
-    fn=qwen_vl_chat,
+    fn=ask,
     inputs=[gr.Image(type="pil"), gr.Textbox(label="请输入问题")],
     outputs="text",
-    title="🧠 Qwen1.5-VL 图文问答 Demo",
-    description="上传一张图片，问一个问题，模型会给出答案。"
+    title="Qwen1.5-VL-Chat 在线体验"
 )
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()