raba - FLUX.1 [dev] LoRA

This is a LoRA model trained on top of FLUX.1 [dev] using Replicate: https://replicate.com/ostris/flux-dev-lora-trainer/train

Key Features

• Built on FLUX.1 [dev], a 12 billion parameter rectified flow transformer
• Designed for high-quality image generation from text descriptions
• Utilizes guidance distillation for efficient performance

Usage

To use this LoRA with FLUX.1 [dev]:

import torch
from diffusers import FluxPipeline

pipe = FluxPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16)
pipe.load_lora_weights("BartoUno/raba")
pipe.enable_model_cpu_offload()

prompt = "Your prompt here REALISTICHUMAN"
image = pipe(
    prompt,
    height=1024,
    width=1024,
    guidance_scale=3.5,
    num_inference_steps=50,
    max_sequence_length=512,
    generator=torch.Generator("cpu").manual_seed(0)
).images[0]
image.save("output.png")

Trigger Words

Use REALISTICHUMAN in your prompt to activate this LoRA's effect.

Model Information

• Base Model: FLUX.1 [dev] (black-forest-labs/FLUX.1-dev)
• Type: LoRA
• Trigger: REALISTICHUMAN

Limitations

• This model is not intended or able to provide factual information.
• As a statistical model, it might amplify existing societal biases.
• The model may fail to generate output that matches the prompts.
• Prompt following is heavily influenced by the prompting style.

License

This LoRA model falls under the FLUX.1 [dev] Non-Commercial License. Please refer to the license document for full details on usage and restrictions.

Out-of-Scope Use

Please refer to the Acceptable Use Policy for information on prohibited uses of this model and its derivatives.