Home Artists Posts Import Register

Content

First of all, I apologize for the lack of recent posts, but I have to say that I've been tinkering with the Rife-App model all this time. In the past few weeks, I've tested, tested, and tested some more various configurations and logics to improve the model, and I must say that all this work has paid off. I am currently finishing training the new models for it, and after releasing this version, I intend to work a bit more on them and then release them as open-source on GitHub. I can say without exaggeration that my model works differently from any public model that can currently be found on GitHub (to the best of my Google-fu skills), and I hope that this model can help other groups of people who work with image transformation. It works basically for any purpose, as long as the output is an image, such as upscaling or image generation. I'm not sure how useful it would be as a diffusion model; that's something I plan to test in the future. However, it is much lighter than self-attention models.

The model is still in training, but here's an example I quickly generated today. Most of the artifacts will likely be fixed upon completing the training:

I like using this video as an example because it's very chaotic:

Original input:
https://drive.google.com/file/d/1OzLGaXP4seHRfwp1tFGxQZ77spwNGdce/view?usp=share_link

Interpolated (download may play better?):
https://drive.google.com/file/d/1et2ppkIGioHFZN5iwGr-slvjE-X1sxny/view


Bark-GUI:
I worked hard to release the Bark-Gui today, but I'll have to finish it tomorrow because I ran out of time. With it, you can create voices, sounds, audios, and other things using text. It's far from perfect, but it's already possible to create some interesting results. Maybe someone will find it useful for something.

Other projects:
90% of my time is being focused on the Rife-App, and the remaining 10% was spent creating this GUI for Bark. Once this next version is ready, I'll focus on finishing the other projects, like releasing the Stable Diffusion GUI as open source. Although, this field is evolving so quickly that it becomes outdated every passing day, hehe.

Comments

Rai T00n

Nice :D

Alien Anthony

Can't wait for bark that sounds super useful