I present another addition to my collection of Dragon Age men, Viago de Riva.
V2 is my revised version of this LoRA since I felt his face looked a bit off and his hair messier than I wanted so I added more training images and tried to refine the scenery a bit more too in the training data. I also increased the dim/alpha on this version from 16/8 to 32/16 so it could learn more detail than V1 which means it is around 197 MB vs 99 MB in size.
I do recommend an adetailer (or face fix) if the shots are more zoomed out.
V2 main tags:
dav_viago, 1boy, black hair, short hair, dark skin, beard, mustache, blue eyesV2 outfit tags:
leather armor, hooded cloak, single pauldron, leather gloves, leather belt, leather pants, leather bootsNote: the single pauldron on his right upper arm/shoulder doesn't like generating consistently. You'll see it on occasion but I got tired of fighting with it so it is what it is at this point The gloves also like to randomize on how they look.
Recommended LoRA strength: 0.7 - 0.8
All of my testing was done with Forge on 4-6 CFG with DPM++ 2M SDE - Karras at 20 Steps.
V1 is my initial version of this LoRA. The generations were fairly consistent (aside from the armor being a bit random here and there) and especially looked great on close-ups and NSFW.
I do recommend an adetailer (or face fix) if the shots are more zoomed out.
V1 main tags:
dav_viago, 1boy, black hair, short hair, dark skin, beard, mustache, blue eyesV1 outfit tags:
leather armor, hooded cloak, single pauldron, leather gloves, leather belt, leather pants, leather bootsNote: the single pauldron on his right upper arm/shoulder doesn't like generating consistently. You'll see it on occasion but I got tired of fighting with it so it is what it is at this point. The gloves also like to randomize on how they look.
Recommended LoRA strength: 0.5 - 0.7
All of my testing was done with Forge on 3-5 CFG with DPM++ 2M SDE - Karras at 20 Steps.
optional tag (for location-specific scenery):
dav_trevisoNote: I tried training scenery shots of Treviso into the LoRA so this tag makes the scenery look closer to Treviso.
Description
V1 is my initial version of this LoRA. The generations were fairly consistent (aside from the armor being a bit random here and there) and especially looked great on close-ups and NSFW.
I do recommend an adetailer (or face fix) if the shots are more zoomed out.









