Verified: 2 years ago
SafeTensor
Type | |
Stats | 620 143 6 |
Reviews | (74) |
Published | Oct 2, 2023 |
Base Model | |
Hash | AutoV2 1234CAF147 |
Imagine one of those not-quite-realistic-enough uncanny valley western video game characters getting a driver's license photo. This LoRA does the opposite of that. It makes your character portraits (male and female) look nicer in a way that's noticeable but difficult to quantify. It also makes costumes more intricate and interesting.
It works with all types of models, including anime, cinematic, realistic, at so on (although if you're using a realistic model, this may push it away from what you want).