Moss character sheet workflow: double-loading an acceleration LoRA on purpose
tomByrer · reddit · 2026-10-29
Reddit user tomByrer shared a ComfyUI workflow for generating character sheets in Moss (Pastebin, v1.0, 1Mb square only).
Key trick: the double-loading of the 4-step acceleration LoRA is intentional — the author discovered it by accident; turning one off made both the character and poses go wild. For character sheets, only the direct prompt and LoRAs should control the character's look, not random chance.
The author is also looking for a better Bounding Box LoRA: ComfyUI's default BB node is bad — hard-set to square, boxes don't follow resolution changes, set by pixels rather than percentages. The attached image shows the multi-view character sheet output.
More from Multimodal
- Redditor seeks open-source VLM to reverse-engineer cinematic prompts from stills, says ChatGPT falls short — Complete-Box-3030 · 2026-10-29
- Magnific launches Move: agentic motion graphics you can edit by talking to it — LudovicCreator · 2026-10-29
- Midjourney v8.2 style test: sref 2717961968 nails the Peanuts look — azed_ai · 2026-10-29
- Updated open-source workflow turns Minimax + ComfyUI into an MV maker — TheDerminator1337 · 2026-10-29
- AI-generated Harry Potter 'looksmaxxing' music video goes viral on Reddit — Upper-Mountain-3397 · 2026-10-29
- A 15-second AI-generated motorcycle escape sequence is being called the coolest yet — umesh_ai · 2026-10-29