Hello welcome!

Here I leave a small introduction of the LoRA:

Providence is an amalgamation of styles, it is trained with the same idea as the detailers but based on style tags, things like sketches, pastel or comic are already included, you just mix what you like and generate good things.

For example, I like the mix of (sketch style + thick lines + hatched style).

Unfortunately I had to delete some data due to training problems, styles like oil and others are ruled out although they are certainly already included in the checkpoints.

But the good thing about providence is the main tags: (pvdc_il, PVDC_STL), by just using these, the style will be varied or mixed... in short, it's quite fun to use~

2025 | IL V1 (VAR-3)

(VAR-3) is the third version for IL, as I have already mentioned, this Lora has given me quite a few problems in training (many repetitions). I will also publish the latest version soon, both versions are good and I don't want to discard them~ 7w7)/

PVDC_IL, PVDC_STL (1.00)

SD1.5 | V1.0:

The idea of LORA arose while I was reviewing some USBs that I found from my school days. At that time I was quite fond of drawing and I downloaded many images to use as a reference and well... There are many images to search author by author with Google Images so I will only mix them in a style.

PS: I trained this model in Hassaku and tested it with Meinahentai. I'll put the instructions next to the examples, but the results are very good even without them: