Update app.py

app.py

@@ -19,14 +19,6 @@ extract_model = YOLO("best.pt").to(device)
 detect_model  = YOLO("yolov8n.pt").to(device)
 
 
-def caption_image_with_llava(image_path):
-    image = Image.open(image_path).convert("RGB")
-    prompt = "[INST] <image>\nDescribe what is visible in the image in a concise, factual sentence. [/INST]"
-    inputs = processor(prompt, images=image, return_tensors="pt").to(device)
-    outputs = llava.generate(**inputs, max_new_tokens=100, do_sample=False)
-    caption = processor.decode(outputs[0], skip_special_tokens=True)
-    return caption
-
 @spaces.GPU
 def process_video(video_path):
     os.makedirs("frames", exist_ok=True)