Update app.py

app.py

@@ -1,31 +1,25 @@
 import gradio as gr
 import torch
-from transformers import AutoProcessor, AutoModel
+from transformers import AutoProcessor, AutoModelForCausalLM
 
 # Load the processor and model with remote code enabled.
 processor = AutoProcessor.from_pretrained(
     "lmms-lab/LLaVA-Video-7B-Qwen2",
     trust_remote_code=True
 )
-model = AutoModel.from_pretrained(
+model = AutoModelForCausalLM.from_pretrained(
     "lmms-lab/LLaVA-Video-7B-Qwen2",
     trust_remote_code=True
 )
 
-# Use GPU if available.
 device = "cuda" if torch.cuda.is_available() else "cpu"
 model.to(device)
 
 def analyze_video(video_path):
     prompt = "Analyze this video of a concert and determine the moment when the crowd is most engaged."
-    # Process text and video.
     inputs = processor(text=prompt, video=video_path, return_tensors="pt")
     inputs = {k: v.to(device) for k, v in inputs.items()}
-    
-    # Generate a response (this assumes the remote code has added a generate method).
     outputs = model.generate(**inputs, max_new_tokens=100)
-    
-    # Decode the output tokens.
     answer = processor.decode(outputs[0], skip_special_tokens=True)
     return answer