A. Uploading Model

1. Come to Runpod website and go to secure cloud section. Choose a GPU. In this tutorial I am using 1x RTX A4500

1. Under Select a Template, choose RunPod Stable Diffusion v2.1

   

   Now we are ready to deploy

   

   You may click on MyPods and see that the SD docker container has started building

   

1. Once its done, and the connect button gets enables, click on the CONNECT button

   

   Click on the Start Web Terminal, and then click on Connect to Web Terminal

   

   1. In the terminal, run the following commands, 1. cd models/Stable-diffusion

      b. try to copy & paste this into the terminal

      curl -c cookies.txt -s -L "https:drive.google.com/uc?export=download&id=1Gl_AQ05GTBLUyNDy1WJOVxgPMcpL5A3f" > /dev/null

      c. What are these doing? —Tippi

      curl -Lb cookies.txt "[<https://drive.google.com/uc?export=download&confirm=$>](<https://drive.google.com/uc?export=download&confirm=$>)(awk '/download/ {print $NF}' cookies.txt)&id=1Gl_AQ05GTBLUyNDy1WJOVxgPMcpL5A3f" -o DnDDragonbornGeneral.ckpt

      After this our model gets download

      

      Yay! We have uploaded our model successfully! You can close the terminal window but don’t need to click Stop Web Terminal.

      B. Enabling SD-WebUI-Endpoints

      1. Go to Jupyter Lab

      

      Make these changes in the webui-user.sh file:

      

      after --enable-insecure-extension-access add --api --cors-allow-origins=* before the last “

      

      make sure to uncomment # export REQS_FILE="requirements_versions.txt”

      Save the file.

      1. Now go back to the pod and stop it. After stopping, restart it and click on Connect via HTTP [Port 3000] to start the stable diffusion.

         

      Our stable diffusion runpod has the custom DnDDragonbornGeneral model but we will have to load it manually by going to the connect to HTTP [port 3000] stable diffusion UI and change the model from the models dropdown menu.

      

      stable diffusion takes a while to load the new weights,

      when it appears like this, then its completed loading

      

      Ensure that you set height and width to 512 before you start using it.

      

To connect to the api click here to learn more.

To learn more about Somya enabled this sdapi/v1/txt2img see this other documentation.

You can read the documentation of sd UI from “API” in the footer.

You can also try out the API from this “API” documentation section.

Using on the Front End

1. Ensure the endpoint is up and running
2. In your .env.local file, you’ll add the url that ends with sdapi/v1/txt2img as your SD_API_ENDPOINT. You’ll end up with something like this:
   1. SD_API_ENDPOINT=[<https://2lqotoget9y4le-3000.proxy.runpod.net/sdapi/v1/txt2img>](<https://2lqotoget9y4le-3000.proxy.runpod.net/sdapi/v1/txt2img>)