Build Your Own AI Image Editing Assistant with Free Prompt Editing on Alibaba Cloud
This tutorial guides you through using Alibaba Cloud's PAI‑DSW platform with the Free Prompt Editing algorithm (a CVPR 2024 paper) to set up a single‑GPU AIGC drawing assistant that enables text‑driven image editing, covering resource acquisition, instance creation, notebook execution, WebUI parameters, and cleanup.
Tutorial Overview
In this tutorial you will learn how to use Alibaba Cloud Interactive Modeling Platform PAI‑DSW with the Free Prompt Editing algorithm (selected paper from CVPR 2024) to develop a personal AIGC drawing assistant that enables text‑driven image editing on a single GPU.
What You Will Achieve
Free trial resources for the PAI platform
Create a PAI‑DSW instance
Compile code and launch WebUI in DSW
Develop a personal AIGC drawing assistant
Step 1: Obtain Free Trial Resources
New users receive 5000 CU·h free trial for PAI‑DSW. Claim the resource pack on the activity page or purchase a package (starting at 59 CNY).
Step 2: Create a PAI‑DSW Instance
Go to the PAI console (https://pai.console.aliyun.com/)
Activate PAI and create a default workspace
Select the Interactive Modeling PAI‑DSW service
Click “Create Instance”, give it a name, choose a GPU (A10 or V100 16 GB+), and select the image pytorch-develop:1.12-gpu-py39-cu113-ubuntu20.04 Confirm resource configuration and create the instance (takes 3‑5 minutes)
Step 3: Open the Tutorial Notebook in DSW
In the instance list, click “Open” to enter the DSW development environment.
In the Notebook tab, click “Launcher”, then “Go to Gallery”.
Search for “AI Artist: Unlimited Possibilities for Image Editing” and open the notebook.
Step 4: Run the Notebook
The notebook contains three sections: preparation, environment installation, and WebUI launch. Execute each code cell sequentially.
Prepare work
Download and install environment
Run WebUI
After the WebUI starts, click the provided URL to access the interface.
WebUI Adjustable Parameters
Attention map Replacing Scale (0.0‑1.0, recommended 0.4‑0.8)
DDIM Steps (denoising steps)
Layers to Edit (recommended 16‑32)
Image Resolution (recommended 512)
CFG Scale (text guidance strength, recommended 7.5)
Seed (random seed)
Example Edits
Changing background: Edit Prompt “a Chinese building on the forest”.
Changing location: Edit Prompt “a Chinese building in New York City”.
Changing style: Edit Prompt “a kids drawing of a cat”.
Van Gogh style: Edit Prompt “Van Gogh style architectural picture”.
Resource Cleanup
After experiments, stop or delete the instance to avoid extra charges.
If you plan to reuse, stop the instance; otherwise delete it.
Further Use
During the trial period you may continue using the DSW instance for model training and inference.
Alibaba Cloud Big Data AI Platform
The Alibaba Cloud Big Data AI Platform builds on Alibaba’s leading cloud infrastructure, big‑data and AI engineering capabilities, scenario algorithms, and extensive industry experience to offer enterprises and developers a one‑stop, cloud‑native big‑data and AI capability suite. It boosts AI development efficiency, enables large‑scale AI deployment across industries, and drives business value.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.
