Update app.py

app.py

@@ -4,11 +4,15 @@ import gradio as gr
 from groq import Groq
 import base64
 import os
+import subprocess
 
 # Define models used in the process
 llava_model = 'llava-v1.5-7b-4096-preview'
 llama31_model = 'llama-3.1-70b-versatile'
 
+# Install flash-attn if not already installed
+subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
+
 # Image encoding function
 def encode_image(image_path):
     with open(image_path, "rb") as image_file: