Verified: 10 months ago
SafeTensor
Type | |
Stats | 102 124 |
Reviews | (10) |
Published | Nov 7, 2023 |
Base Model | |
Training | Steps: 3,550 Epochs: 10 |
Usage Tips | Strength: 1 |
Trigger Words | heroes5_style |
Hash | AutoV2 B831226D0D |
This is a model that was trained on most concept arts and hero portraits found on the heroesofmightandmagic.com website. The goal was to make a lora that represents the Heroes of Might and Magic V style.
Weight: I usually go for 0.7 (I find it a good spot on every image I generate), sometimes lower if I decide to mix with other loras, but in this case you can set weight as 1, to capture the style more.
Trained on: AnyLora using this tutorial and collabs: https://civitai.com/models/22530/guide-make-your-own-loras-easy-and-free
Thank you!