Skip to content

Latest commit

 

History

History
87 lines (59 loc) · 2.23 KB

stable-diffusion.md

File metadata and controls

87 lines (59 loc) · 2.23 KB

Stable Diffusion

QuickStart

Prerequisites

  • An ASK cluster is created. For more information, see Create an ASK cluster
  • The cluster runs as expected. You can log on to the Container Service for Kubernetes (ACK) console, navigate to the Clusters page, and then check whether the cluster is in the Running state.

Deploy Stable Diffusion

  1. create deployment & service
# for cpu
kubectl apply -f stable-diffusion-cpu.yaml
# for gpu
kubectl apply -f stable-diffusion-gpu.yaml
  1. wait deployment ready
$ kubectl get po|grep stable-diffusion
---
NAME                                READY   STATUS              RESTARTS   AGE
stable-diffusion-5469d87d76-fjg9f   1/1     Running             0          2m55s
  1. Using Stable Diffusion WebUI

Run the following command to port-forward:

kubectl port-forward -n <namespace> service/stable-diffusion-svc 7860:7860

And then open the console using the following URL:

http://localhost:7860

stable-diffusion

Deploy Stable Diffusion Dreambooth

  1. create deploy & svc
kubectl apply -f sd-dreambooth.yaml
  1. wait pod ready
$ kubectl get po|grep stable-diffusion-train
---
stable-diffusion-train-5469d87d76-fjg9f   1/1     Running             0          2m55s
  1. Using Stable Diffusion WebUI with dreambooth

Run the following command to port-forward:

kubectl port-forward -n <namespace> service/stable-diffusion-train 7860:7860

And then open the console using the following URL:

http://localhost:7860

See the doc for using Dreambooth to train your own models.

Release Tag

tag date release
v1.4.0 2023-12 model: v1-5-pruned-emaonly.safetensors; WebUI版本:v1.6.0
v2.0.0-gpu 2023-12 model: v1-5-pruned-emaonly.safetensors;support dreambooth