Type | |
Stats | 491 |
Reviews | (74) |
Published | Jan 3, 2024 |
Base Model | |
Usage Tips | Strength: 0.8 |
Hash | AutoV2 872E128882 |
This is the LyCORIS version of the previously released LoRA. It's trained on the same dataset but should generate higher quality output while being more flexible.
How to use
The style will work even if you just prompt "1girl" or "1boy" but you can refine it further by using fantasy/Etrian Odyssey related tags. Or not if you want to use contemporary themes or just apply a general Himukai Yuuji style influence.
For more prompting tips see the sample images or the uploaded zip with training tags.
AUTOMATIC1111 users can utilize the wildcard prompt below to randomly prompt classes and races from the game. I recommend using "3d" and "realistic" in negatives on AOM2 and other models with mixed in realism. Not really needed in this version for AOM2, other models may still need it.
"Chibi" also seems to work fine for more chubby little adventurers.
<lora:EtrianV2:1>
({full|upper} body)
(solo)
(1girl)(facing viewer)
({human|elf|bunny_ear|goat_horns|horse_ears|cat_ears|demon|angel} girl)
in ({barbarian|bard|cleric|druid|fighter|monk|paladin|ranger|rogue|sorcerer|warlock|wizard|artificer} outfit)
Don't forget to review and upload your own images, it's always fun to see what you're creating!
MISC
Trained on 44 images on NAI at 512 resolution.
All sample images are raw txt2img upscaled, no inpainting or helper embeddings used.
The training was done on a small subset of the artists images, as such the output of this LoRA is not an accurate representation neither in quality or style of the source material. Go support the artist if you want the good stuff.