.webp)
關於「tamamo cross (umamusume)」標籤
tamamo cross (umamusume) 是 NovelAI Diffusion 可以使用的角色標籤。Danbooru 上有 1,999 篇投稿，靜態圖的最高分數為 672。上方範例把這個標籤放在固定基準提示詞的開頭，並以 seed 1234 生成，因此與其他標籤的差異都來自這個標籤。
各模型畫出的 tamamo cross (umamusume)
忠實遵循構圖與表情的指定，標籤效果呈現得較整齊。
對標籤的反應更強，效果更容易看出來。
本機模型 WAI illustrious SDXL v17。可以看到同一標籤在 NovelAI 以外的表現。
tamamo cross (umamusume) 在提示詞中的寫法
先寫主體，再寫這個標籤，最後接品質標籤。
tamamo cross (umamusume), solo, best quality, masterpiecewlop, tamamo cross (umamusume), solo負面提示詞與完整設定請見 使用條款。
範例的生成條件
所有圖片都以 832×1216、seed 1234、23 step、k_euler_ancestral 生成。基準提示詞與負面提示詞對所有標籤相同，只替換要測試的標籤。被 NSFW 檢查判定的圖片會換 seed 重新生成或隱藏。
Danbooru wiki 摘錄
展開 wiki 內文 ▾收合 ▴
Umamusume character, voiced by Oozora Naomi.
A short (140cm), feisty, Kansai-accented horse girl nicknamed 'Tama' by friends and known as 'The White Lightning', who grew up in rural poverty and runs to support her family. Close friends and rivals with Oguri Cap, whom brings out the most of Tamamo's natural-born tsukkomi tendencies with her straight-faced oddities.
She has blue eyes and long grey hair with hair between her eyes. She wears red ear covers, as well as a red headband and a red and blue hairband. A spherical ball-like ornament is attached to the hairband, in front of her right ear.
Appearance
- : Tracen School Uniform
- : Fast As Lighting
- : Fast As Lighting (Draft)
- : Starting Future
- : Casual
- : Fierce Thunder God
- : Beware! Halloween Night!
Related characters
- Gold City
- Inari One
- Oguri Cap - Roommate, friend, rival and natural manzai partner
- Super Creek - Friend and rival. She also fears Creek due to her tendency to treat Tamamo like a child.
- Komiyama Masami - Trainer in Umamusume: Cinderella Gray
See also
- Tamamo Cross (racehorse)
- Midori Makibaoo (Another fictional character based on the same racehorse)
- Nervous Tama-chan Pose
External links
常見問題
NovelAI 能辨識 tamamo cross (umamusume) 嗎？
請對照本頁的範例。本站所有標籤都使用相同的 seed 與基準提示詞，畫面上的差異都來自 tamamo cross (umamusume)。
tamamo cross (umamusume) 在提示詞中怎麼寫？
把「tamamo cross (umamusume)」寫在提示詞前段，並用逗號與其他標籤分開。
V5 和 V4.5 該用哪個？
V4.5 更能表現標籤的特色，V5 的姿勢更穩定、破綻更少。