Edit model card

This was perhaps one of my greatest follies and one of my greatest challenges.

I would like to start by pointing out that this model is not at its ideal quality. Stable Diffusion's lack of skill with words makes it difficult to make a model in this subject of high quality.

This was the best quality I could get after testing over 8 different models.

It took more than 30 images and 9200 steps to try to get some relatively good result. The best model turned out to be the 5500 steps model.

This model will never replace designers and developers, because that was never my goal. I tried to create something that would serve as inspiration for them and as an incentive for other people to work on developing new models on the subject. I just wanted to give a kick-start to this area that is still not as explored.

I honestly don't see this model being widely used outside of the testing enthusiasts, designers and developers. You can get a lot of design inspiration for your application from this model. Many times it doesn't respond the way we want it to, so we have to readjust the words. Stable Diffusion is also not so good at words.

I am providing two templates that you can test according to your wishes and needs.

  • The 2000 model delivers more or less quality, but has more creativity. It can be used in some cases.
  • The 5500 model is the one that I found the best. It has the best quality.

To use it you have to use the word ''MobiUiUX'' in the prompt.

From my tests the images look better with this prompt:

highly detailed, trending on artstation, behance, ios app, MobiUiUx

For negative prompts I got better results when I used: out of frame, duplicate, watermark, signature, text, ugly, sketch, deformed, mutated, blurry, mutilated, ugly sketch

The model is available only on Huggingface.

You can make a collaborative donation at the following sites. All money raised goes to pay GPU Rent and Colab.

Patreon:https://www.patreon.com/user?u=81570187

Ko-Fi:https://ko-fi.com/jvkape

buy me a coffe:https://www.buymeacoffee.com/JVKAPE

Hopefully better models on the subject will come along!

Enjoy :)

Downloads last month
0
Inference API
Unable to determine this model's library. Check the docs .