Sign In

Hairstyle - Slicked Back

52

Updated: Oct 17, 2026

conceptweyfluxwomenhairears

Not recently loaded

Download

1 variant available

bf16 SafeTensor

Slicked Back Hair Style - Krea2.safetensors

BF16, good balance • 218 MB

Verified:

Type
LoRA
Stats

433

40

113

Reviews
Published

Aug 8, 2026

Base Model
Training
Steps: 2,000Epochs: 10
Usage Tips
Strength: 1
Hash
AutoV281D1A0306F
00002-1761328132.jpg

SLICKED BACK HAIR STYLE - FLUX.1

This LoRA helps to generate images of women with that iconic slicked back hair style. Trained with the on-site CivitAI Trainer on 44 images containing women with all kinds of hair colors/skin colors/hair lengths. Please share your awesome creations here. Your feedback is highly appreciated! Thanks a Lot!!

👩‍🦳 INFO ----- 20 Epochs / 20 Repeats / 4 Batch Size / 512px / AdamW8bit (Altogether ~4400 Steps)

👩‍🦳 USAGE ------ I used Flux.1 Dev (FP8) and Flux.1 Krea Dev (FP8) Checkpoints with Euler + Beta / Normal at 20 steps, 896x1152 Resolution and 3.5 Distilled CFG for my image generations.

👩‍🦳 WEIGHT ------ Seems to work well within 0.8-0.9. Please experiment yourselves!

👩‍🦳 PROMPTING ------ You can add "slicked back" to the prompt if it's not working as intended. Also. add the word "wet hair" to get that wet combed look. (It doesn't work all the time though).

👩‍🦳 LIMITATIONS ------ This LoRA is not perfect. It seems the model has a bit of bias towards certain clothing style and makeup style. You could try using this with other LoRAs such as character/clothing LoRAs but as per my testing, the results may vary.

❗❗❗ DISCLAIMER ❗❗❗
This LoRA Model was created for educational and personal entertainment purposes only. I do not wish to gain any commercial benefit. I used publicly available images from the internet as source data for the model and I do not posses any rights to those images. All the credit must go to the respective original owners. Any resemblance to an actual person in a generated image is purely coincidental. I kindly ask all of you to responsibly use the model. Thank you very much!