In Development - Borzoi, a simple way to communicate with Stable Diffusion WebUI API!

WARNING: This code was created by a non-professional software programmer, so i love feedbacks, etc. It's just a thing that i created to test/extend/consume the API in a external program, just discord bots or other thing, in a simple way. I'm not skilled or anything, but i will try.

Requirements

• Axios
• and nothing more;

Installation

npm i --save github:nozomihub/borzoi
# OR
yarn add https://github.com/nozomihub/borzoi

Things that i added by now:

• Text-to-image inference
  • One drawback is that you have to decode the base64 image later. i will change that. gimme time.
  • it has a convenient Class Builder.

Things i need to add later:

• Image-to-image inference
  • It's theoretically easy, just extends one class, add more things...
  • I have a internal Buffer-to-Base64. i'm just lazy right now.
• Lora support
  • Of course i want Mai Sakurajima, with a blue bikini in the beach. i need Lora, okay?
• Upscaling
  • Images are nice, but in 1024x1024 are way better.
• Options for blacklisting nsfw/some words
  • The age of consent is 18 years old.

Code Example:

const borzoi = require("borzoi")

    const bor = new borzoi.Borzoi("http://127.0.0.1:8080")
    const payload = new borzoi.BorzoiPayloadBuilder(0)
        .setPrompt("masterpiece, solo, 1girl")
        .setSteps(5)
    bor.inference(payload).then(res => console.log(res))