Create app.py

app.py

@@ -0,0 +1,51 @@
+import gradio as gr
+import torch
+from transformers import AutoProcessor, AutoModelForCausalLM
+
+# Load the processor and model from Hugging Face
+processor = AutoProcessor.from_pretrained("lmms-lab/LLaVA-Video-7B-Qwen2")
+model = AutoModelForCausalLM.from_pretrained("lmms-lab/LLaVA-Video-7B-Qwen2")
+
+# Set the device (use GPU if available)
+device = "cuda" if torch.cuda.is_available() else "cpu"
+model.to(device)
+
+def analyze_video(video_path):
+    """
+    This function accepts the path to a video file,
+    then uses the LLaVA-Video-7B-Qwen2 model to analyze the video.
+    
+    The prompt instructs the model to analyze the video and return 
+    the moment when the crowd is most engaged.
+    """
+    # Define the prompt for the model
+    prompt = "Analyze this video of a concert and determine the moment when the crowd is most engaged."
+    
+    # Process the video and prompt.
+    # Note: The processor is expected to handle the video input (e.g., by reading frames).
+    inputs = processor(text=prompt, video=video_path, return_tensors="pt")
+    
+    # Move all tensor inputs to the selected device
+    inputs = {key: value.to(device) for key, value in inputs.items()}
+    
+    # Generate the model's response
+    outputs = model.generate(**inputs, max_new_tokens=100)
+    
+    # Decode the generated tokens to a human-readable string
+    answer = processor.decode(outputs[0], skip_special_tokens=True)
+    return answer
+
+# Create the Gradio Interface
+iface = gr.Interface(
+    fn=analyze_video,
+    inputs=gr.Video(label="Upload Concert/Event Video", type="filepath"),
+    outputs=gr.Textbox(label="Engagement Analysis"),
+    title="Crowd Engagement Analyzer",
+    description=(
+        "Upload a video of a concert or event and the model will analyze "
+        "the video to identify the moment when the crowd is most engaged."
+    )
+)
+
+if __name__ == "__main__":
+    iface.launch()