Community Struggles to Train Character LoRAs on MOSS Model
A Reddit user reports that character LoRAs trained on the MOSS model achieve only about 70-80% likeness, with fine details failing to converge even when reusing datasets that worked well on other models, while style and concept LoRAs train fine.
2026-10-19 ~ 2026-10-20 · 2 related posts
- MOSS Character LoRas Only 70-80% Identity-Accurate, Community Seeks Fixes — Bostonparis · 2026-10-19
- Moss LoRA Training Struggles: Character Likeness Won't Converge — atakariax · 2026-10-20