full 10th epoch of a braces lora. testing to see how it goes

Braces

This is a version of the lora - the the same data set as the xl version, but trained in Flux. Flux seems to be so much easier to get loras trained. I will keep trying with others

users_LORA_849314969710281425_08e7783b-b26e-4a0e-85a4-64f997bc9b26.TA_trained.safetensors

v1.0

This was my very first LORA train, and I've recently come back to it to try it again, and I'm actually getting pretty good results.I'm finding by running Upscale on the image, you get some really good results. Please if you like this, share your work, I'd love to see others results.

I just updated this to a new LORA i trained on a new set of high resolution images that are much higher quality than the previous set I used. I still need to run some tests my self, but this is looking very positive!

Files with SHA256: 68a00059f7a4897999c8be6e811479eaec42e3aa733b72618e70978885ca9cdc