A LoRA of Astolfo (from the Fate series). I recommend checking out both v4 and v3, I wouldn't necessarily say one is better than the other.While it does a good job of generating Astolfo without any physical descriptor tags, adding some can help. For instance, I've found that it struggles with doing his eye color naturally, so adding 'purple eyes' improves it.You may notice the absence of v1 and v2, this is because those were me learning how to even make these, so they aren't great.

AA's Astolfo LoRA

Trained on ~20 images. The tags in the data set were pruned. Though a lower version number, this is not necessarily worse than v4 (in fact, it may be better. Let me know your thoughts).

Files with SHA256: 93619a0bf81e815bfa64d3af699d07f49698f99a40c6fd8cc46388832615faa5