This Checkpoint remains equally clear, clean, and creative, with stable body structures and poses, diverse compositions, and no need for AD.
這個Checkpoint還是一樣的清晰，乾淨，有創意，身體結構和姿勢穩定，構圖多樣，無需 AD。
Since there are quite a few novice users on Tensor, I made this Checkpoint. It's extremely easy to use. Just provide the character name, expression, and action, and you'll get the cosplay picture you desire. ((There's not even a need to add quality prompt words because I've already incorporated them for you)). For negative prompt words, you only need to add a simple Embedding.
因為tensor上的新手用戶比較多，我製作了這個Checkpoint，使用非常簡單，只需給出角色名，表情，動作，就能得到你想要的cosplay圖片。 ((甚至不需要加質量提示詞，因為我已經幫你寫進去了))。負面提示詞可以只需要加入簡單的Embedding。
Due to the powerful NSFW capability of Illustrious, I don't want too many Lolitas mixed in among the cosplayers. I integrated the Lora for age enhancement, so underage characters will look rather odd. Since this Lora isn't made by myself, it will affect the hands and limbs. The image - generation rate has decreased somewhat. Male characters are still a weak point because there aren't many males in my realistic training dataset........Maybe the next version will address this.
Wish you all a great time playing around.
由於Illustrious強大的NSFW能力，我不希望cosplayer裡混入太多Lolita，我融合了年齡提升的Lora，所以未成年的角色會看起來很奇怪。因為這個Lora不是我自己做的，所以會影響到手和肢體。出圖率有所下降。男性角色還是個弱項，因為我的realistic訓練集裡就沒有太多男性........也許下一個版本會有吧。
祝大家玩的愉快。
Steps: 15-40, CFG scale: 4-7, Sampler: Euler a/dpmpp_sde/dpmpp2m_sde, scheduler: karras/sgm_uniform/exponential/beat
The VAE is already integrated
My usual settings are: 25, 6, dpmpp_sde, beat, 832x1536,
Prompt words and Danbooru tags are the key to making the most of all Illustrious and NOOB features. Everyone, give it a try. Have a great time!