Type | |
Stats | 463 4,107 1k |
Reviews | (70) |
Published | Sep 15, 2024 |
Base Model | |
Training | Steps: 10,000 Epochs: 10 |
Usage Tips | Clip Skip: 2 |
Trigger Words | detailed drawing ARKN |
Hash | AutoV2 AE30B6F3BB |
Hello welcome!
Here I leave a small introduction of the LoRA:
This is a lora trained to test word processors, I think I still prefer my DatasetHelpers, I hadn't noticed the updates and I see that they already have quite a few new versions, it's time to try them ~ 7w7)/
I have been testing some subtitles (tags) for the training, I used the images of the titansteng datasets for KDA and KDA x Arcane, for just Arcane I included a different data set but I included some of its captures. Their subtitling method is very different from mine so I couldn't be as vague as I wanted, anyway today they play three styles at once ~ XD
Only Arcane
Only arcane is something more personal, I included the characters but I focused on the style and well that's it ~ 7w7)/
KDA x Arcane
KDA x Arcane is a mix of the three styles but Pony is not usually very good at generating mixes and I had to separate the datasets, there I tried subtitling in sentences, I was just curious about the results 7o7)/
Only K/DA
Only KDA is trained with screenshots of a couple of MVs (I thank titansteng). This version does not have any alterations to its datasets and I only did it to test the results :3