Sign In

Kirit0Vision

276
2.2k
63
Updated: Feb 14, 2024
base model
Verified:
SafeTensor
Type
Checkpoint Merge
Stats
1,574
Reviews
Published
Feb 1, 2024
Base Model
SD 1.5
Hash
AutoV2
6327ED5FE1
default creator card background decoration
kirit0's Avatar
kirit0

Before i would have said this is just a merger of some of my most frequently used models.
RevAnimated, CartoonMadness, Epic photogasmDodgeandBurn, SexyToon3d, and a dozen+ more.

But it has come a LONG way since then, and id like to say its its own unique model now, its been heavily diluted from its original form, and tailored to the generalized aesthetic that i like.

It is Extremely Good with semi to realistic photos, better with simple negative prompts, or even no negative prompt.

Leave me a review with feedback on what you did or didnt like or experience with it.

Suggestion Time:
For Anime the following (first) works best:
(((((anime style, drawn, japanese_animation, lineart, manga)))))

My Preferred detail prompt is
highly realistic, realistic, masterpiece, absurdres, intircate, highly detailed, detailed, high quality, best quality, highres, photo_(medium), reality, 8k, cinematic lighting, dimly lit, low light, dim, dark, dim lighting, volumetric lighting, professional photography,

Detail Prompts that work well:
Concept: perfect eyes, https://civitai.com/models/65941 (:1) seems sufficient
Add More Details - Detail Enhancer, https://civitai.com/models/82098 (0.6-1) seems good

For Full body images use: (Full body shot)

Preferred Negatives (when i do use them, cause as stated above its not really required)

difConsistency Negative v2, https://civitai.com/models/87375

Beyond Negative Embedding v3, https://civitai.com/models/108821?modelVersionId=173718

bad-picture-chill 75v, https://civitai.com/models/17083

Sometimes ill used aDetailer, face_yolov8n.pt denoise 0.4 to smooth out some face artifacting caused by poor lora compatibility, i use with blank prompt or with single word emotion prompts

And if i really want the eyes to pop ill use aDetailer, mediapipe_face_mesh_eyes_only, i use with the above mentioned lora, <lora:locon_perfecteyes_v1_from_v1_64_32:1>, perfecteyes eyes, photo of perfecteyes eyes,

For low light stuff, you will need to find yourself a copy of LowRa here: https://huggingface.co/XpucT/Loras/tree/main