NotSHAFT版本v2.0 (3.5) (3.5)

=== The following is related to v2.0 (3.5) ===

This is a style trained on (checks) Uhh... 1665 of a certain SHAFT anime (Here is a hint. anime aired in 2022). Before anyone asks, this is not a LyCORIS (LoCon) on Monogatari.

Additionally, Salient cropping was used to improve other details such as eyes.

For NSFW Content, you will need to prompt it harder but it can do NSFW.

Try not to use it for commercial uses. No one is gonna stop you but I believe some people will be rather upset over it.

=== The following is related to v1.0 ===

This is a style trained on (checks) Uhh... 1665 of a certain SHAFT anime (Here is a hint. anime aired in 2022). Before anyone asks, this is not a LyCORIS (LoCon) on Monogatari.

NSFW? Uh no. The model strong refuses NSFW content. I tried, and only bringing down the weight does the model swallows the fact you want to gen horny.

Try not to use it for commercial uses. No one is gonna stop you but I believe some people will be rather upset over it.

IMO this model is not good, I'm just posting because I could.

描述:

Retrained the model. This time, there is some special sauce (Which I call crop augmentation).
Essentially, I ran the images through a model to get the saliency for each model and cropped them to a 1:1 ratio. In my initIal testing, square pictures are now much more defined and close up and the over quality appears to have improved. I described a bit of my process on twitter.
The epoch uploaded is epoch 4, which isn't the final epoch but is so far the best performing one.
N.B. the trigger word "notshaft style", doesn't strongly affect the image too much, I messed up and forgot to add keep tokens.
N.B 2: Consider letting the model stew a bit longer to get more the the fine details out (More steps. Euler a with 80 seems like a decent spot). Also, set the text encoder to 0.5. I'll probably find out how to properly train a text encoder eventually lol.
Hint: The examples are your friend. :3

训练词语: notshaft style