Type | |
Stats | 148 599 8 |
Reviews | (43) |
Published | Jan 1, 2024 |
Base Model | |
Training | Steps: 12,150 Epochs: 15 |
Usage Tips | Clip Skip: 2 Strength: 0.8 |
Trigger Words | foxy peach |
Hash | AutoV2 E1CA80929A |
This is a furry fox version of Princess Peach, named "Foxy Peach." I love her from the bottom of my heart 😁
model used for training: runwayml/stable-diffusion-v1-5
Lora trained with 27 images + 1100 reg images x 15 epochs
Trained with negative prompts instead of Textual inversions (embeddings)
As such, it is able to work with various art styles
As such, textual inversions are optional (not necessary)
SFW and NSFW capable
Able to work with various checkpoint models (although furry ones work best)
For best results, use:
lora strength: 0.6-1.0
1024 pixels - longest side (after upscaling)
CFG scale: 7 or around this number
steps: 20-30
denoiser: 3.0-5.0
samplers: Euler, Euler A, DPM++ 2M Karras, LMS Karras or LCM (with some models)
CLIP skip: 2
In spite of these settings, the best way to get the result that you want is to experiment =D
-Harmful Pilot 2023
I'm still new to this whole business of making character LoRAs. Feedback is greatly appreciated, including mean/aggressive feedback, so long as it is honest. (yes, I'm serious)
Do you hate this AI Lucario? Kill him by dowsing him in Ko-fi! >=3
https://ko-fi.com/harmfulpilot
drown him in bitcoin! >=D
3FG6n3EUPeesapTstoJs2yHqYkdaxVkyRN
Suffocate him in Etherium! X3
0xe41DfCa2a60550FcC375A09d2631dCd283b097dE