Dec imports

app.py

@@ -1,6 +1,5 @@
-import gradio as gr
-import spaces
-
+from gradio import ChatInterface, Textbox, Slider
+from spaces import GPU
 from threading import Thread
 from torch import bfloat16
 from transformers import Qwen2VLForConditionalGeneration, Qwen2VLProcessor, TextIteratorStreamer, AutoProcessor
@@ -18,7 +17,7 @@ min_pixels = 256 * 28 * 28
 max_pixels = 1280 * 28 * 28
 processor: Qwen2VLProcessor = AutoProcessor.from_pretrained(model_path, min_pixels=min_pixels, max_pixels=max_pixels)
 
-@spaces.GPU
+@GPU
 def infer(
     messages,
     max_tokens,
@@ -66,14 +65,14 @@ def respond(
     for response in infer(messages, max_tokens, temperature, top_p):
         yield response
 
-app = gr.ChatInterface(
+app = ChatInterface(
     respond,
     type="messages",
     additional_inputs=[
-        gr.Textbox(value="You are Softie, a helpful assistant.", label="系统设定"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="最大生成长度"),
-        gr.Slider(minimum=0.01, maximum=4.0, value=0.75, step=0.01, label="温度系数（Temperature）"),
-        gr.Slider(minimum=0.01, maximum=1.0, value=0.5, step=0.01, label="核取样系数（Top-p）"),
+        Textbox(value="You are Softie, a helpful assistant.", label="系统设定"),
+        Slider(minimum=1, maximum=2048, value=512, step=1, label="最大生成长度"),
+        Slider(minimum=0.01, maximum=4.0, value=0.75, step=0.01, label="温度系数（Temperature）"),
+        Slider(minimum=0.01, maximum=1.0, value=0.5, step=0.01, label="核取样系数（Top-p）"),
     ],
 )