Update app.py

app.py

@@ -2,13 +2,17 @@ import gradio as gr
 from transformers import AutoProcessor, Qwen2_5_VLForConditionalGeneration, TextIteratorStreamer
 from qwen_vl_utils import process_vision_info
 from threading import Thread
-import torch
 import spaces
 
-MODEL_ID = "Qwen/Qwen2.5-VL-7B-Instruct"  
-processor = AutoProcessor.from_pretrained(MODEL_ID, trust_remote_code=True)
+file_path = "csfufu/Revisual-R1-final"
+
+processor = AutoProcessor.from_pretrained(
+    file_path,
+    min_pixels=256*28*28,
+    max_pixels=1280*28*28
+)
 model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
-    MODEL_ID, torch_dtype="auto", device_map="auto"
+    file_path, torch_dtype="auto", device_map="auto"
 )
 
 @spaces.GPU
@@ -28,6 +32,8 @@ def respond(
         "content": system_message
     }]
 
+    print(chat_history)
+
     for message in chat_history:
         if isinstance(message["content"], str):
             messages.append({
@@ -45,15 +51,22 @@ def respond(
                 ]
             })
 
-    messages.append(
-        {
+    if text:
+        messages.append({
             "role": "user",
             "content": [
                 { "type": "text", "text": text },
-                *[{"type": "image", "image": image} for image in files]
             ]
-        }
-    )
+        })
+    
+    for file in files:
+        messages.append({
+            "role": "user",
+            "content": [
+                { "type": "image", "image": file }
+            ]
+        })
+    
     image_inputs, video_inputs = process_vision_info(messages)
     prompt = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
     inputs = processor(
@@ -65,7 +78,7 @@ def respond(
     ).to(model.device)
     
     streamer = TextIteratorStreamer(processor, skip_prompt=True, skip_special_tokens=True)
-    generation_kwargs = dict(inputs, streamer=streamer, max_new_tokens=max_tokens, temperature=temperature, top_p=top_p)
+    generation_kwargs = dict(**inputs, streamer=streamer, max_new_tokens=max_tokens, temperature=temperature, top_p=top_p)
 
     thread = Thread(target=model.generate, kwargs=generation_kwargs)
     thread.start()
@@ -93,4 +106,4 @@ demo = gr.ChatInterface(
         ],
     )
 
-demo.launch(debug=True)
+demo.launch(debug=True)