More changes

app.py

@@ -8,7 +8,7 @@ import spaces
 import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig, TextIteratorStreamer
 
-subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
+#subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
 
 MODEL_ID = os.environ.get("MODEL_ID")
 CHAT_TEMPLATE = os.environ.get("CHAT_TEMPLATE")

requirements.txt

@@ -1,4 +1,5 @@
 transformers==4.39.2
 accelerate
 bitsandbytes
+flash-attn
 optimum