Dreambooth trained and extracted to LyCORIS.
Full body model trained with sd.15 pruned (~200 images)
Face model trained with Realistic_Vision_v2.0 (~20 images)
Must use both models for a more accurate portrayal (0.75 for DM-full; 0.6 for DM-Face)
Use with hiresfix and Adetailer
Best checkpoint models:
cyberrealistic_v32
analogMadness_v70
juggernaut_final
realisticVisionV60B1_v51VAE
Reliberate_v3
some images where inpainted to keep it Civitai-friendly.
trigger word activates both models.