Face detailer pass fixes character drift in SDXL LoRA workflows
petranova_ · reddit · 2026-09-05
A Reddit user shares hands-on ComfyUI experience with character LoRAs on SDXL: faces still drifted across batches until adding a face detailer pass at the end of the pipeline. Baking the LoRA into the UNET instead of loading it per run also cut VRAM churn significantly.
More from Multimodal
- Reddit user compares two MiniMax H3 Turbo LoRAs for video generation — No-Bee-231 · 2026-09-05
- MiniMax H3 workflow: REF2VA visuals + FL2VA audio + LightX2V speed, 198s on RTX 5090 — gabxav · 2026-09-05
- AI short film "Ishiro - The Last Blade" shared on Reddit — Nervous-Barnacle-888 · 2026-09-05
- Dev calls for a MiniMax H3 finetune that fills speech gaps with human filler words — cocktailpeanut · 2026-09-05
- Nvidia DLSS 5 frame interpolation discussed in Stable Diffusion community — KonoTheSavage1 · 2026-09-05
- Perceptron's Multilook API Prefills Video Context Once, Cuts Input Cost to 32% at 16 Prompts — AkshatS07 · 2026-09-05