Tsuchinaga Hiyori (Blue Archive)

槌永ヒヨリ (ブルーアーカイブ) / 츠치나가 히요리 (블루 아카이브) / 槌永日和 (碧蓝档案)

Usage

• Use any or all of the following tags to summon Hiyori: hiyori, 1girl, hair over one eye, side ponytail, aqua eyes, hairclip, halo

• For her normal outfit: cabbie hat, black sweater, white capelet, white skirt, belt, long sleeves, thigh pouch, kneehighs, loafers, black footwear

  • Without cape: remove white capelet, add grey shirt, collared shirt, wing collar

  • Without sweater: in the negative prompt, use white capelet, grey shirt, collared shirt, wing collar

  • For some reason, the model usually fails to render the black stripe on her skirt, despite being quite accurate otherwise. 😔

• For her daydreaming expression: smile, naughty face, nervous smile, open mouth

• For crying:

  • Mild: tearing up, closed mouth, wavy mouth

  • Moderate: crying with eyes open, flustered, tears, wavy mouth, open mouth

  • Maximum: closed eyes, crying, bawling, tears, flustered, (wide open mouth:1.3)

• Sweater lift: sweater lift, clothes lift, lifted by self, midriff, navel

  • Optional: weight conscious

• Accessories:

  • weapon case, strap, sling for her rifle case

  • ammunition belt, bullet for her ammo belt

  • fashion magazine, book for her magazines

    • The AI will struggle to render a recognizable magazine cover

  • backpack, holding straps for her equipment pack

The AI may be stubborn with trying to always show her with a scarf or lifted shirt, since she has many images with them. Try using the following negative prompt to counteract this behavior: (black sweater, collared shirt, white cape, grey shirt, white scarf, detached collar:1.2)

Not all tags may be necessary.

I try to avoid overfitting my LoRAs. As a result, your prompt might require more tags to perfectly reproduce the character, but in exchange the model should be much more flexible when changing clothing or hairstyles and mixes well with other LoRAs and art styles. I recommend setting weight to 1.0 and using some of the tags provided above.

私はLoRAモデルのオーバーフィッティングを避けるようにしています。そのため、キャラクターを完全に再現するには、より多くのタグが必要になる場合がありますが、その代わり、モデルはキャラクターの衣服や髪型の変更に柔軟に対応でき、他のLoRAモデルやアートスタイルとも相性が良くなります。必要に応じて、重みを1.0に設定し、上記で提供したタグのいくつかを使用することをお勧めします。

저는 LoRA 모델을 오버피팅하는 것을 피하기 위해 노력합니다. 그 결과로, 캐릭터를 완벽하게 재현하기 위해서는 더 많은 태그가 필요할 수 있지만, 대신 모델은 캐릭터의 의상이나 머리 스타일을 쉽게 변경할 수 있으며, 다른 LoRA 모델 및 아트 스타일과도 잘 어울립니다. 가능한 태그 중 일부를 사용하고, 가중치를 1.0으로 설정하는 것을 권장합니다.

我尽量避免LoRA模型的过拟合。因此，你的提示可能需要更多的标签来完美地复现角色，但是相应地，该模型在更改角色服装或发型方面更加灵活，与其他LoRA模型和艺术风格混合得更好。我建议将权重设置为1.0，并使用上面提供的标签中的一些。

Refer to https://huggingface.co/khanon/lora-training for other Blue Archive characters, training parameters, and datasets.