Loop ForgeLaughing
Limit, from MOSS-Image add-ons, tested locally
- Model
- MOSS-Image SFT (moss_image_sft_int8_convrot.safetensors)
- Output
- 880×1184
- Sampling
- 30 steps, CFG 4, euler / normal
- Seed
- 42
- Edit LoRA
- 61.5 s, peak VRAM 10.6 GiB
- Ran on
- RTX 3060, 12GB (local)
What was wired in
It got the laugh, and the outfit and background are still there, but the face is no longer hers. It is someone else in her clothes.
Change only her expression: she is laughing with her eyes half closed and her head tilted back a little. Keep her face, identity, hair, outfit, pose and the background unchanged.
nsfw, worst quality, bad anatomy, watermark, jpeg artifact
Notes
ComfyUI's MOSS-Image edit template as it ships, with the edit LoRA, 30 steps and CFG 4. Only the photo, the instruction and the seed are mine.
Comfy-Org/MOSS-Image, with the edit LoRA (from chinoll/Moss-Image-Edit)
Comfy-Org workflow templates (the MOSS-Image templates ship with ComfyUI)
Run it yourself
Download everything for this test (.zip, 1.2 MB). It holds the workflow, every input it loads and a README. Copy the inputs into ComfyUI/input, put the models below in place and drag the workflow onto ComfyUI.
| File | Put it in | Size |
|---|---|---|
| moss_image_sft_int8_convrot.safetensors | models/diffusion_models | 5.4 GB |
| moss_image_edit_lora_rank_64_bf16.safetensors | models/loras | 297.4 MB |
| qwen3.5_9b_w4a8_gptq.safetensors | models/text_encoders | 7.3 GB |
| flux2-vae_small.safetensors | models/vae | 249.5 MB |
No custom nodes. Every node is part of ComfyUI 0.43.1 (frontend 1.57.6, templates 0.11.100).