If you have some experience with AI, you may already know that both the Hugging Face website and Google Colab are treasure troves for generating uncensored images without the need for a powerful machine or installing many Python packages on your local system.
But finding the right resources can still be a challenge. Fortunately, there is a Github repository called stable-diffusion-webui-colab that collects ipynb files to import into Colab, enabling you to generate images in an array of wild styles using specialized models. If you are not comfortable with importing ipynb files, you can refer to the links in the repository’s readme to access third-party colabs or Huggin Face pages to achieve the same result.
Once you launch a Colab, you’ll be presented with a Stable Diffusion web interface where you can import your desired model and input crazy prompts to generate images. Don’t forget to switch to the latest branch (v2) on Github for the newest updates.
In a nutshell, the process is simple: just click the play button to start the Colab, and you’re ready to go. In this example, the author is using a specialized model to generate comics.
It will load a lot of stuff, then you will see a URL appear… Click on it…
Once you reach the web interface, you can input a prompt and then click on the Generate button. As a result, you will obtain an image that has a Comic Book appearance.
Once you’re at that point, you can experiment with a lot of artistic or ultra-realistic possibilities based on the model you select. Moreover, there are no censorship restrictions on the generated images as Stable Diffusion is utilized behind the scenes.
To sum up, you have plenty of options to keep yourself entertained for a while. However, be aware that it can be quite time-consuming. Additionally, I recommend checking out the other repositories, which include Colab projects enabling you to manipulate a person’s lips in videos or photos to make them say anything you want.