FLUX.1-dev-LoRA-playful-metropolis

This is a LoRA (Playful Metropolis) trained on FLUX.1-dev for illustration style by JHAlchemy, combining urban environments with imaginative cartoon elements.

Showcases

Prompt

A giant tiger dressed in retro clothing strolls between the skyscrapers of Tokyo. The tiger's eyes, peering through sunglasses, exude a cool vibe. On its back are a few cartoon-style birds playing small instruments, providing a soundtrack for the tiger's walk,

Prompt

In a London subway station, a blue octopus with multiple eyes is holding a pile of books that have fallen out of a subway car. Its tentacles are busy flipping through each book, while the passengers inside the station are both surprised and curious about the scene.

Prompt

On a bustling beach, a giant cartoon alligator with bright turquoise skin is carrying a surfboard under its arm, towering over the sunbathing beachgoers. The alligator is wearing black sunglasses and a relaxed expression, blending in with the beach vibes. Palm trees and people enjoying the sun surround the scene, creating a playful juxtaposition of scale and reality.

Trigger words

Trigger is not required. The recommended scale is 0.7 in diffusers.

Inference

import torch
from diffusers import FluxPipeline

pipe = FluxPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16)
pipe.load_lora_weights("Shakker-Labs/FLUX.1-dev-LoRA-playful-metropolis", weight_name="FLUX-dev-lora-playful_metropolis.safetensors")
pipe.fuse_lora(lora_scale=0.7)
pipe.to("cuda")

prompt = "A giant tiger dressed in retro clothing strolls between the skyscrapers of Tokyo. The tiger's eyes, peering through sunglasses, exude a cool vibe. On its back are a few cartoon-style birds playing small instruments, providing a soundtrack for the tiger's walk,"

image = pipe(prompt, 
             num_inference_steps=24, 
             guidance_scale=3.5,
             width=768, height=1024,
            ).images[0]
image.save(f"example.png")

Online Inference

You can also download this model at Shakker AI, where we provide an online interface to generate images.

Acknowledgements

This model is trained by our copyrighted users JHAlchemy. We release this model under permissions. The model follows flux-1-dev-non-commercial-license.