I have a Openfaas deployed on my k8s cluster, though, low compute yet, do you need a lot of compute to run the backend? GPU?
https://www.openfaas.com/blog/openfaas-flask/

#@title # Setup, run this once
from IPython.display import clear_output
!sudo apt-get -y install llvm-9-dev cmake
!pip install Deepspeed
!npm install -g localtunnel
clear_output()
%cd /content/
!pip install keras==2.4.0 Flask==1.1.2 Flask-Cors==3.0.9 Flask-RESTful==0.3.8 dalle-pytorch==0.13.4 tqdm triton==0.4.2
!git clone https://github.com/rom1504/dalle-service.git
clear_output()
print("Finished setup.")

https://github.com/rom1504/dalle-service/tree/master/back
https://medium.com/@kshitijvijay271199/flask-on-google-colab-f6525986797b

dalle-pytorch version has been updated, and the weights from the checkpoint is different from the model.
Version dalle-pytorch==0.14.3 works. Consider updating requirements.txt

I've run the backend server, and entered the url into your front url.
The url loads fine, but when I hit enter, nothing comes up even though I'd waited quite a while