
触发词
文件清单 线路:镜像(hf-mirror.com) · 切换
下载后放进对应的 ComfyUI 目录。models/ 指 ComfyUI 的模型根目录。
说明
Senko-san_NAI-V5-Curated_ORIGINAL-ANIME_ULTRA (仙狐さん)
作者:JustInEchoes
Anime-focused character LoRA for Senko-san from The Helpful Fox Senko-san, trained on OpenMOSS MOSS-Image-SFT.
Built from an entirely new dataset of 1,000 native 1024×1024 NovelAI 5.0 Curated generations, this ORIGINAL-ANIME version brings the softer linework, familiar colors, and character design targeted by the original TV-anime dataset to MOSS Image. The training images were generated to resemble frames from an episode.
ORIGINAL-ANIME?
Where the HD version emphasizes ultra-crisp detail and polished rendering, ORIGINAL-ANIME prioritizes Senko's on-screen likeness and a softer, screencap-like finish. The new MOSS previews show different framing and softer rendering than HD SFT V1, with flexible natural-language scene and clothing prompts.
This ULTRA variant uses the same prompt set and dataset structure as the other ULTRA releases, with all images newly generated in NovelAI 5.0 Curated for the ORIGINAL-ANIME look. Its 598-image core character set establishes identity across angles, expressions, poses, and framing, while 402 activity scenes add varied activities, camera angles, shot types, and poses.
The Trigger Word
senko_san
Quick Generation Settings
- Base checkpoint — OpenMOSS-Team/MOSS-Image-SFT
- LoRA version — SFT V1, final 2,000-step checkpoint
- LoRA weight — 1.0
- Prompt style — Descriptive natural language with senko_san
- Sampler — Flow-matching Euler
- Denoising steps — 50
- CFG scale — 3
- Flow shift — 5
- Unconditional branch skip layer — 22, using zero-based indexing
- Negative prompt — Empty
- Resolution — 1024 × 1024
- Example seed — 38141
- Text encoder — Qwen3.5-9B, BF16
- VAE — OpenMOSS bundled small VAE, FP32
These are the settings used for the saved previews. Start with LoRA weight 1.0, then adjust it to suit your prompt. Use a MOSS workflow that exposes the flow shift and unconditional branch skip settings.
Prompting
Describe clothing, pose, composition, lighting, and environment in natural language. The trigger senko_san anchors the character. The examples below are the exact prompts from the training previews, each generated at seed 38141 with 50 denoising steps, CFG 3, and LoRA weight 1.0. All four used an empty negative prompt.
- Teal Raincoat in a Courtyard
senko_san, full-body anime illustration, taking a step across a stone courtyard in a teal raincoat and plain boots, fluffy fox tail visible, overcast light.
Negative prompt left empty.
- Window-Light Portrait
senko_san in a head-and-shoulders anime illustration, smiling gently in window light, both fox ears visible, clean line art and flat cel shading.
Negative prompt left empty.
- Setting Out Rice and Chopsticks
senko_san setting a bowl of rice and chopsticks on a low wooden table with both hands, medium-wide view, warm morning light, clean anime line art.
Negative prompt left empty.
- Sharing Tea
senko_san on the left, sharing tea with an adult human woman with short black hair on the right, two distinct faces, cozy room at dusk, cel-shaded anime illustration.
Negative prompt left empty.
Dataset and Captions
Training used all 1,000 image-caption pairs from Senko ORIGINAL-ANIME ULTRA V5 Captions, with images at their native 1024×1024 resolution. Each image has a matching .txt caption containing a descriptive natural-language caption. Captions were kept intact, with caption shuffling disabled and caption dropout set to zero.
Dataset available on Hugging Face.
Training Specifications
- Training engine — ostris/ai-toolkit, pinned commit d6a18ad72da9abca3ffb67ea8e42c25764eee61a
- Starting checkpoint — OpenMOSS-Team/MOSS-Image-SFT
- Model architecture — MOSS Image
- Adapter — Standard LoRA
- Trigger word — senko_san
- Training steps — 2,000 optimizer updates
- Batch size — 2
- Gradient accumulation — 1
- Approximate dataset passes — 4
- Network rank / alpha — 32 / 32
- Optimizer — Automagic
- Configured learning rate — 0.0001
- LR scheduler — Cosine
- Training precision — BF16
- Diffusion model weights — Unquantized BF16
- Text encoder — Frozen, unquantized BF16 Qwen3.5-9B
- VAE computation — FP32, OpenMOSS small VAE
- Saved LoRA precision — BF16
- Gradient checkpointing — Enabled
- Maximum gradient norm — 1.0
- Training scheduler — Flow matching with shifted timestep sampling
- Training resolution — 1024 × 1024
- Training image-caption pairs — 1,000
- Caption shuffling — Disabled
- Caption dropout — 0
- Horizontal flipping — Disabled
- Latent and text embedding caching — Enabled
- Excluded LoRA target — layers.22.
- Checkpoint and preview interval — Every 250 training updates
- Training hardware — NVIDIA RTX 6000 Ada, 48 GB, on RunPod
The uploaded adapter is the final ORIGINAL-ANIME SFT V1 checkpoint at step 2,000, trained under the internal name senko-original-anime-ultra-moss-sft-v1. The attached teal raincoat image is the exact final-step preview from this run. Training used 2,000 optimizer updates; the 50-step setting applies to rendering previews.