istolemyownwheat Open-source Text-to-Image Model - Focus on Character Performance, Freely Generate Personalized Images

Istolemyownwheat

Developed by nadanainone

A text-to-image generation model based on the Anything model, trained on a personal artwork dataset, focusing on character representation rather than background details.

Image Generation EnglishOpen Source License:Openrail #Anime Character Generation #Low Background Detail #Multi-step Style Variation

Downloads 46

Release Time : 11/15/2022

Model Overview

This model is a text-to-image generation model based on Stable Diffusion technology, trained on a personal artwork dataset, specifically optimized for character representation.

Model Features

Character Representation Optimization

The model's output focuses more on character representation, with relatively simplified background details.

Multi-step Options

Provides model files with 5k, 10k, and 15k training steps, producing different style effects.

Personal Art Style

Trained on the creator's personal artwork, featuring a unique artistic style.

Model Capabilities

Text-to-Image Generation

Anime-style Image Generation

Character Portrait Generation

Use Cases

Art Creation

Anime Character Design

Generate anime character images with a personal style

Can generate character portraits in different styles, such as the pink-haired girl in the example

Concept Art Creation

Quickly generate concept art sketches

Can generate portraits like the underwater biopunk style shown in the example

🚀 Stable Diffusion Model: Customized Training

This project is a customized Stable Diffusion model, which uses the same dataset as istolemyownart (https://huggingface.co/nadanainone/istolemyownart) and adds 9 extra drawings that were not uploaded on Tumblr. It is based on the Anything model instead of a personal merge, and uses some tweaked settings with the latest updated Dreambooth extension for Voldy's (AUTOMATIC1111)'s UI.

✨ Features

Focus on Characters: Produces less random over - detailed backgrounds and focuses more on characters.
Multiple Training Steps: Trained at 15k steps instead of 10k. Files included are 5k, 10k, and 15k steps, each with different aesthetic results.
Same Prompt: The prompt is still "ludditesbtfo".

📦 Installation

No installation steps are provided in the original document, so this section is skipped.

💻 Usage Examples

Basic Usage

The basic usage of this model is to use the prompt "ludditesbtfo" to generate images. Here are some examples of images generated at different training steps:

5k steps

![00311 - 1131084824 - 1girl, ludditesbtfo, solo focus, pink hair, long hair, medium breasts, simple background.png](https://s3.amazonaws.com/moonup/production/uploads/1668544476783 - 63716cac15aafbe231371caa.png)

10k steps

![00310 - 1131084824 - 1girl, ludditesbtfo, solo focus, pink hair, long hair, medium breasts, simple background.png](https://s3.amazonaws.com/moonup/production/uploads/1668544481696 - 63716cac15aafbe231371caa.png)

15k steps

![00309 - 1131084824 - 1girl, ludditesbtfo, solo focus, pink hair, long hair, medium breasts, simple background.png](https://s3.amazonaws.com/moonup/production/uploads/1668544488452 - 63716cac15aafbe231371caa.png)

Additional Examples

Here are some additional randomly selected examples at random steps:

![00301 - 2701705781 - portrait of sansa dark crimson, intricate, elegant, highly detailed, digital painting, artstation, concept art, smooth, sharp fo.png](https://s3.amazonaws.com/moonup/production/uploads/1668544882956 - 63716cac15aafbe231371caa.png)

![00261 - 434223274 - highly detailed portrait of a sewer punk lady student, blue eyes, tartan hoodie and scarf, white hair by atey ghailan, by greg r.png](https://s3.amazonaws.com/moonup/production/uploads/1668544911556 - 63716cac15aafbe231371caa.png)

![00291 - 10062718 - beautiful full body portrait of a human like anime girl android with pale skin and fox ears, high quality, highly detailed, 4 k,.png](https://s3.amazonaws.com/moonup/production/uploads/1668544898006 - 63716cac15aafbe231371caa.png)

![00189 - 3229683580 - SFW, Masterpiece, best quality, high detail, (detailed face and eyes_1.1), (style of silver and black theme_1.1), pov, looking a.png](https://s3.amazonaws.com/moonup/production/uploads/1668544924132 - 63716cac15aafbe231371caa.png)

![00184 - 114906268 - SFW, 1girl, solo focus, Masterpiece, best quality, (style of a poster_1.3), from side, looking at viewer, teenager, mature, smal.png](https://s3.amazonaws.com/moonup/production/uploads/1668544936694 - 63716cac15aafbe231371caa.png)

![00221 - 1096241287 - lofi underwater bioshock biopunk steampunk portrait, wearing corset, Pixar style, by Tristan Eaton Stanley Artgerm and Tom Bagsh.png](https://s3.amazonaws.com/moonup/production/uploads/1668544953235 - 63716cac15aafbe231371caa.png)

📚 Documentation

Here are some frequently asked questions and answers:

⚠️ Important Note

Bad Pickles: Tested for bad pickles, and it is bad - pickle free.

Rights: You do not have the rights over the entire art, dataset, and all future artworks (both manual and AI - made).

Model Usage: You can't sell the model. All models should always be free and open - source. You can use it for fair use and derivative works. You can train another model using the art, but you don't own the new model.

💡 Usage Tip

You can wrangle it with better artists to improve results.

📄 License

The license for this project is creativeml - openrail - m.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご