A DreamBooth-LoRA for Krea 2, trained on Krea 2 RAW and shown on Krea 2 Turbo. The samples below were generated with this LoRA on Turbo (8 steps).
Use the token TUMTUM@26 to invoke the concept.
"A futuristic cyberpunk city street drenched in neon rain, where a high-tech holographic TUMTUM@26 floats above a crowded marketplace."
"An ancient, overgrown stone temple deep in a tropical jungle, featuring a weathered statue of TUMTUM@26 reclaimed by vines and moss."
"A whimsical miniature world inside a glass terrarium, showcasing a tiny, porcelain TUMTUM@26 resting on a bed of glowing mushrooms."
import torch
from diffusers import Krea2Pipeline
pipe = Krea2Pipeline.from_pretrained("krea/Krea-2-Turbo", torch_dtype=torch.bfloat16).to("cuda")
pipe.load_lora_weights("Nanami231/tumtum-26")
image = pipe("A futuristic cyberpunk city street drenched in neon rain, where a high-tech holographic TUMTUM@26 floats above a crowded marketplace.", num_inference_steps=8, guidance_scale=0.0).images[0]
image.save("output.png")
checkpoint-1000/optimizer.bin 195 MBcheckpoint-1500/optimizer.bin 195 MBcheckpoint-2000/optimizer.bin 195 MBcheckpoint-500/optimizer.bin 195 MBcheckpoint-1000/pytorch_lora_weights.safetensors 192 MBcheckpoint-1500/pytorch_lora_weights.safetensors 192 MBcheckpoint-2000/pytorch_lora_weights.safetensors 192 MBcheckpoint-500/pytorch_lora_weights.safetensors 192 MBpytorch_lora_weights.safetensors 192 MBcheckpoint-1000/scheduler.bin 1.4 kBcheckpoint-1500/scheduler.bin 1.4 kBcheckpoint-2000/scheduler.bin 1.4 kBcheckpoint-500/scheduler.bin 1.4 kBkrea/Krea-2-Raw