Stylization by Valentin Gubarev.

The training was conducted on the CIVITAI website.

80 pictures were used.

Sometimes, when using LoRa models, there are still problems with limbs, whereas on a pure DEV model, there are much fewer of them, keep this in mind when generating.

Quite often, the generation turns out to be quite surreal both in terms of small details and how these details look. very often there are problems with limbs and their number. But, apparently, this is a problem with all models where there is a similar drawing in the original drawings, and similar details in the picture itself. FLUX is very often messing up here. But, again, any picture can be completed, the main thing is that there is a basis, and here it is.

we have to temporarily make early access to the model in order, if possible, to slightly compensate for the BUZZ spent on training.