uf0d0w_flux2klein9b_v1

LoRA fine-tune of black-forest-labs/FLUX.2-klein-base-9B trained with ai-toolkit.

Usage

Trigger word: uf0d0w

from diffusers import DiffusionPipeline
import torch

pipe = DiffusionPipeline.from_pretrained("black-forest-labs/FLUX.2-klein-base-9B", torch_dtype=torch.bfloat16)
pipe.load_lora_weights("hellorob/uf0d0w_flux2klein9b_v1")
image = pipe("uf0d0w ...your prompt here").images[0]

Training details

• Base model: black-forest-labs/FLUX.2-klein-base-9B
• Architecture: flux2_klein_9b
• LoRA rank: 32 (alpha: 32)
• Steps: 2000
• Learning rate: 0.0001
• Optimizer: adamw8bit
• Trained with: ostris/ai-toolkit on Modal

Sample prompts used during training

• [trigger] declassified UAP report image, a horse floating above a desert landscape
• [trigger] black and white surveillance photo of an unknown flying object above mountains at dusk
• [trigger] a wide shot photo of an island with a the tower of Modor, cloudy sky, washed colors
• [trigger] thermal imaging capture of a towering monolight exiting a portal over the open Pacific ocean
• [trigger] amateur cellphone photo of a a Clown in a suburban neighborhood street at night

License notice

This LoRA was trained on a FLUX.2 base model, which is released under a non-commercial license. Use of this LoRA inherits those terms — non-commercial use only. See the base model page for the full license text.