Sign In

Igarashi Rika/五十嵐梨花 | hunyuan video

2

Download

1 variant available

SafeTensor

225.06 MB

Verified:

Type
LoRA
Stats

116

2

5

Reviews
Published

Jan 26, 2025

Base Model

Hunyuan Video

Training
Steps: 320
Epochs: 16
Usage Tips
Strength: 0.6
Hash
AutoV2
4F20EA4CFB
Trigger Words
Igarashi rika, An anime-style girl with light orange hair styled in a single bun, with a white scrunchie, and vibrant green eyes. She is wearing a stylish outfit consisting of a teal-colored overall dress layered over a black top. Over this, she wears an oversized orange and black jacket

Tencent Hunyuan is licensed under the Tencent Hunyuan Community License Agreement, Copyright © 2024 Tencent. All Rights Reserved. The trademark rights of “Tencent Hunyuan” are owned by Tencent or its affiliate.

Powered by Tencent Hunyuan

ComfyUI_03833_.webp

This lora depicts the character Igarashi Rika(五十嵐梨花) from Nijisanji

Trigger Word:Igarashi rika, An anime-style girl with light orange hair styled in a single bun, with a white scrunchie, and vibrant green eyes. She is wearing a stylish outfit consisting of a teal-colored overall dress layered over a black top. Over this, she wears an oversized orange and black jacket

Trained: musubi-tuner, 20 images

Caption: adjusted caption from CahtGPT (includes described background)

epoch:16

train_step:320

lora_strength:0.6

sample prompt:Igarashi rika, An anime-style girl with light orange hair styled in a single bun, with a white scrunchie, and vibrant green eyes. She is wearing a stylish outfit consisting of a teal-colored overall dress layered over a black top. Over this, she wears an oversized orange and black jacket, (she is dancing:3.0), (chibi Nendoroid:3.0) in bedroom.

11~20epochのモデルを確認したが16epochのモデルの造形の出来が一番良かった。

I checked models from 11 to 20 epochs and the 16 epoch model was the best work.

アクションは強調promptによって2.0~4.0に調整する必要があります。画像から学習したためか強調promptを使わないとあまり動きません。またはモーションloraを併用ください。

Action prompt should be adjusted to 2.0 to 4.0 depending on emphasis prompt. It does not move much without the emphasis prompt because it was learned from the image. Or motion lora use together.