Update app.py

app.py

@@ -13,6 +13,9 @@ import matplotlib.patches as patches
 import random
 import numpy as np
 
+import subprocess
+subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
+
 models = {
     'microsoft/Florence-2-large-ft': AutoModelForCausalLM.from_pretrained('microsoft/Florence-2-large-ft', trust_remote_code=True).to("cuda").eval(),
     'microsoft/Florence-2-large': AutoModelForCausalLM.from_pretrained('microsoft/Florence-2-large', trust_remote_code=True).to("cuda").eval(),