g33-lokr β€” Krea 2 character LoKR

Character LoKR for Krea 2, trained on Krea-2-Raw (1500 steps, LoKR factor 4, 33 images).
Use with Krea-2-Turbo for inference.

Trigger

g3f

Strength

Start around 0.6–0.9.

Files

  • pytorch_lora_weights.safetensors β€” final adapter (~5.4 MB)
  • sample_*.jpg β€” step-1500 previews

Use with Diffusers / Krea 2 Turbo

import torch
from diffusers import Krea2Pipeline

pipe = Krea2Pipeline.from_pretrained(
    "krea/Krea-2-Turbo",
    torch_dtype=torch.bfloat16,
).to("cuda")
pipe.load_lora_weights("tmtmtm45/g33-lokr")
image = pipe(
    "g3f, portrait, soft window light",
    num_inference_steps=8,
    guidance_scale=0.0,
).images[0]
image.save("out.png")

ComfyUI

Drop into your Krea2 LoRAs folder and prompt with g3f.

Downloads last month
4
Inference Examples
Examples
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for tmtmtm45/g33-lokr

Base model

krea/Krea-2-Raw
Adapter
(590)
this model