add flash atten properly

app.py

@@ -1,4 +1,11 @@
 import spaces
+import subprocess
+# Install flash attention, skipping CUDA build if necessary
+subprocess.run(
+    "pip install flash-attn --no-build-isolation",
+    env={"FLASH_ATTENTION_SKIP_CUDA_BUILD": "TRUE"},
+    shell=True,
+)
 import time
 import logging
 import gradio as gr

requirements.txt

@@ -6,4 +6,4 @@ termcolor
 transformers
 torch
 num2words
-flash-attn
+# flash-attn