This lora model is trained on Pony Realism.

If you use animation-style checkpoints, the resulting image can also be animated

Well... The training seemed harder than I thought.

I trained it three times with different parameters. This one seems to work slightly better.Actually, it didn't have the effect I wanted.

This is the image in the training set.↓

This is the image generated by lora. ↓

This face doesn't look like it.

Maybe need to increase num_repeats.(This lora is 15).

Or you may need to train the portraits separately.

There may be a new version in the future.

But that's it for now.

There is a cost to training on colab, after all.