Update app.py

app.py

@@ -12,6 +12,7 @@ from langchain_community.llms.huggingface_pipeline import HuggingFacePipeline
 from langchain_core.prompts import PromptTemplate
 from langchain.chains import RetrievalQA
 
+os.system('pip install auto_gptq-0.4.1+cu118-cp310-cp310-linux_x86_64.whl')
 os.system('sudo apt-get install -y poppler-utils')
 
 DEVICE = "cuda:0" if torch.cuda.is_available() else "cpu"