Update app.py

app.py

@@ -47,18 +47,19 @@ MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 
 device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
 
-# Function to return an HTML snippet of a thin animated progress bar
+# Updated function with optimized progress UI
 def progress_bar_html(message: str) -> str:
     return f"""
-<div style="display: flex; align-items: center;">
-  <span style="margin-right: 8px;">{message}</span>
-  <div style="position: relative; width: 110px; height: 5px; background-color: #f8d7da; border-radius: 2px; overflow: hidden;">
-    <div style="position: absolute; width: 100%; height: 100%; background-color: #f5c6cb; animation: loading 1.5s linear infinite;"></div>
+<div style="display: flex; align-items: center; justify-content: center; margin: 10px 0;">
+  <span style="margin-right: 10px; font-weight: bold; color: #333;">{message}</span>
+  <div style="position: relative; width: 200px; height: 10px; background-color: #e0e0e0; border-radius: 5px; overflow: hidden;">
+    <div style="position: absolute; width: 100%; height: 100%; background: linear-gradient(90deg, #76c7c0, #4caf50); animation: loading 2s ease-in-out infinite;"></div>
   </div>
 </div>
 <style>
 @keyframes loading {{
     0% {{ transform: translateX(-100%); }}
+    50% {{ transform: translateX(0%); }}
     100% {{ transform: translateX(100%); }}
 }}
 </style>
@@ -79,7 +80,7 @@ TTS_VOICES = [
     "en-US-GuyNeural",    # @tts2
 ]
 
-MODEL_ID = "prithivMLmods/Qwen2-VL-OCR-2B-Instruct" 
+MODEL_ID = "prithivMLmods/Qwen2-VL-OCR-2B-Instruct"
 processor = AutoProcessor.from_pretrained(MODEL_ID, trust_remote_code=True)
 model_m = Qwen2VLForConditionalGeneration.from_pretrained(
     MODEL_ID,