The model includes data for three sets of clothing. The trigger words are as follows:
type 1: capelet, white thighighs
type 2: underwear, see-through
type 3: swimsuit, bikin, see-through
Description
-
FAQ
Comments (5)
Apologies if this was answered elsewhere in your other LORAs, but what checkpoint did you use to train these DOA LORAs?
I apologize for not specifying earlier. All the Lora models currently are trained using chilloutmix_NiPrunedFp32Fix. Moving forward, I will specify which base model is used for training in each Lora version. This will ensure that users are aware of the origin and related details of each model
@iiiiiphoenix That's surprising, I know it's a realistic model and these characters are off a 3D based game, but I didn't expect it to turn out like that with Chillout as the model for training.
Por que no da con otros modelos?..
Para lograr una fidelidad visual, te recomendaría usar modelos de 2.5d y semi-realistas, como los tres que recomendé anteriormente