add init

app.py

@@ -1,6 +1,13 @@
+import torch
 import gradio as gr
+from vlm_captions import VLMCaptioning
 from llm_inference_video import VideoLLMInferenceNode
 
+# Initialize the VLMCaptioning model once at startup
+print("Initializing Video Prompt Generator...")
+vlm_captioner = VLMCaptioning()
+print("Video Prompt Generator initialized successfully!")
+
 title = """<h1 align="center">AI Video Prompt Generator</h1>
 <p align="center">Generate creative video prompts with technical specifications</p>
 <p align="center">You can use prompts with Kling, MiniMax, Hunyuan, Haiper, CogVideoX, Luma, LTX, Runway, PixVerse. </p>"""
@@ -156,6 +163,34 @@ def create_video_interface():
 
     return demo
 
+def describe_image_interface(image, question="Describe this image in detail.", temperature=0.7, top_p=0.9, top_k=40, max_new_tokens=512):
+    """Interface function for image description"""
+    if image is None:
+        return "Please upload an image."
+    
+    return vlm_captioner.describe_image(
+        image=image,
+        question=question,
+        temperature=temperature,
+        top_p=top_p,
+        top_k=top_k,
+        max_new_tokens=max_new_tokens
+    )
+
+def describe_video_interface(video, frame_interval=30, temperature=0.7, top_p=0.9, top_k=40, max_new_tokens=512):
+    """Interface function for video description"""
+    if video is None:
+        return "Please upload a video."
+    
+    return vlm_captioner.describe_video(
+        video_path=video,
+        frame_interval=frame_interval,
+        temperature=temperature,
+        top_p=top_p,
+        top_k=top_k,
+        max_new_tokens=max_new_tokens
+    )
+
 if __name__ == "__main__":
     demo = create_video_interface()
     demo.launch(share=True)