Refactor model ID handling in app.py and update requirements.txt

app.py

@@ -3,6 +3,8 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStream
 import torch
 import gradio as gr
 from threading import Thread
+import subprocess
+subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
 
 models_available = [
     "MohamedRashad/Arabic-Orpo-Llama-3-8B-Instruct",

requirements.txt

@@ -1,5 +1,4 @@
 transformers
 torch
 accelerate
-spaces
-flash-attn
+spaces