ComfyPilotComfyPilot
模型库 / LoRA / Clairveil · 澄彩 | MOSS-Image Style LoRA
Clairveil · 澄彩 | MOSS-Image Style LoRA

Clairveil · 澄彩 | MOSS-Image Style LoRA

3049745

LoRA 6 次下载 1 喜欢 更新于 2026-11-13

底模:MOSS Image

mossstyle

在源站打开

触发词

clairveil

文件清单

下载后放进对应的 ComfyUI 目录。models/ 指 ComfyUI 的模型根目录。

权重 claireveil.safetensors 138 MB

说明

Clairveil · 澄彩 | MOSS-Image Style LoRA

作者:MengY_Madrigal

Fine lines, airy colors and soft character rendering — Clairveil brings a gentle illustration style to character art and scenic compositions.

Trigger: clairveil Suggested strength: 0.75 — try 1.0 for a stronger effect.

Trained on MOSS-Image-heal with AI Toolkit, and tested with the original MOSS-Image-SFT and MOSS-Image-Base INT8 ConvRot checkpoints. The gallery uses SFT at strength 0.75.

Add the trigger to your positive prompt and describe your character, action and setting. Natural-language prompts and MOSS layout JSON both work. The weight filename is claireveil.safetensors; the trigger is spelled clairveil.

This release is the 9750-step, rank-32 checkpoint. In my training run, a recognizable style began to emerge around 1000 steps.

Full comparisons, prompts and ComfyUI workflows on Hugging Face · Try the RunningHub comparison workflow

示例图