Overwatch - Mei版本fenn_mei (ID: 31572)

Mei, from Overwatch
(it's not perfect, I know, but I'm pretty happy with the results)
generation info still in the metadata for all images if you save them locally.

Lower CFG gives best results, I hang out around 4.5.

Lower CFG also gives more strength to the embedding vs your overall prompt. I’ve found a lower CFG works better for any embedding on this site.

Did some testing with prompts on this one, and for some reason it sometimes helps if you add "Mei_overwatch" to the prompt.
But without the embedding, "Mei_overwatch" looks nothing like her. Embedding by itself still works well though!

Tested this on a bunch of different models, some realistic, some anime.

TEST PROMPT
(I use a system the uses <> for TI triggers, remove those if your UI doesn't utilize that.)

<fenn_mei>, portrait, (masterpiece),(best quality:1.0), (ultra highres:1.0), detailed, 1girl, curvy, wearing glasses, messy bun, wearing a coat, kneeling, thong, large snow boots, snow outside, vibrant, vivid, ski lodge, smiling, booty, detailed beautiful skin, detailed eyes, mei_overwatch

Negative -
3d, sepia, thin, painting, cartoon, sketch, text, watermark, signature, logo, (worst quality:2), (low quality:2), (normal quality:2), lowres, bad anatomy, bad hands, normal quality, ((monochrome)), ((grayscale)), collapsed eyeshadow, multiple eyebrows,

Before anyone asks, it's a .bin file because it was trained using the Google Colab, and that's what it gives you.

This is my fourth upload, please, for the love of god, leave a review or something.

描述:

Lower CFG gives best results, I hang out around 4.5.

训练词语: fenn_mei