Instructions to use abbuibuibui/MinamiNanami-Lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use abbuibuibui/MinamiNanami-Lora with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("waiIllustriousSDXL-v17", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("abbuibuibui/MinamiNanami-Lora") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
Minami Nanami (Bottom-tier Character Tomozaki) — Illustrious SDXL Character LoRA
中文说明 | English (current)
An unofficial character LoRA for Minami Nanami (七海南) from Bottom-tier Character Tomozaki, trained on waiIllustriousSDXL v17. Formal E1–E7. Comparative evaluation recommends E6 at strength 0.8. Dataset: abbuibuibui/MinamiNanami-Dataset.
Fan-made derivative, not official.
Files
| Epoch | File | Role |
|---|---|---|
| E1–E4 | minami_nanami-000001 … 000004 |
Early / alternate. Blonde / blue already appears. |
| E5 | minami_nanami-000005.safetensors |
Matrix candidate. Portrait brow is a bit stern. |
| E6 | minami_nanami-000006.safetensors |
Recommended default. |
| E7 | minami_nanami.safetensors |
Final epoch. Face close to E6. |
No minami_nanami-000007.safetensors — unnumbered file is E7.
E6 SHA-256: C8F4E012D4EE0E354EDC45045C1F4B8235E96CDB659C482DBD276A88B6DE6D40
Each file is 456497868 bytes.
Showcase
E6 @ 0.8, ComfyUI, Euler a / 28 / CFG 4.5 / CLIP skip 2 / 832×1216. No hires / ADetailer.
Trigger Words and Settings
Captions often omit hair / eye color — add them.
| Role | Token | Typical look |
|---|---|---|
| Identity | minami_nanami |
long blonde hair, blue eyes |
| Outfit | minami_nanami_blazer_uniform |
Navy blazer with crest, white shirt, solid red necktie, blue-grey plaid skirt, dark knee socks. Main eval. |
| Outfit | minami_nanami_summer_uniform |
Summer school look; eval often keeps blazer-off / lighter silhouette with red necktie (sometimes striped). Alt eval. |
| Outfit | minami_nanami_casual_cardigan / minami_nanami_home_loungewear / minami_nanami_sports_t_shirt / minami_nanami_yukata |
Other trained outfits |
Example:
<lora:minami_nanami-000006:0.8>, masterpiece, best quality, very aesthetic, absurdres, anime screencap, 1girl, solo, minami_nanami, minami_nanami_blazer_uniform, long hair, blonde hair, blue eyes, looking at viewer
Negative: add brown hair, black hair, pink hair, brown eyes, green eyes, red eyes, twintails, short hair as needed.
Why E6
Round1 screened E1–E7 @ 0.8 (face / blazer / summer). Round3 matrix on E5–E7.
| Candidate | Observed | Decision |
|---|---|---|
| E5 | Blonde / blue locked; portrait a bit stern. | Alternate. |
| E6 | Portrait clearer; blazer full body is crest blazer + red tie + plaid + knee socks, arms at sides. Night street single subject with fists lightly clenched. | Recommended @ 0.8. |
| E7 | Face nearly E6. | Final alternate. |
Without outfit token, 0.8 becomes a loose light-blue T-shirt and white sneakers — casual, same blonde / blue. Same series as Hinami / Kikuchi / Hanabi: keep the name trigger.
Chinese notes: evaluation/横向测评记录.md.
Public Horizontal Evaluation
Round 1: E1–E7 @ 0.8
Round 3: E5 / E6 / E7 × 0.4 / 0.6 / 0.8 / 1.0
Training
- rank 64 / alpha 32; 71 pairs; repeats 2; DGX Spark + ComfyUI
Related
- HF: https://huggingface.co/abbuibuibui/MinamiNanami-Lora · Dataset same name
- ModelScope: https://www.modelscope.cn/models/abbuibuibui/MinamiNanami-Lora
License
creativeml-openrail-m
- Downloads last month
- 94

















