We are aware of the instability and are actively investigating. Our team is working to restore full functionality as quickly as possible, and we appreciate your patience while we work through this!

Ongoing Site Issues

The Krea 2 LoRA training contest is over with winners announced!

⚡ Krea 2 Training Contest Winners⚡

We're a Day-0 launch partner for MiniMax H3, a next-gen open-weights multimodal video model - and it's already in the generator. No waitlist. Open weights land in a few days.

MiniMax H3 is live: open-weights video, generate now

Simple LoRA for analog/film photo style. Works best at resolutions above 1024x, but I don't have the memory to train above that, so it's not perfect.Example images were generated with 8step lightning lora on q4ks gguf.

Same images, same captions, same settings, same seed. Only change is using 640p + 1328p buckets instead of only 1024p buckets, as the latest musubi-tuner + activation offloading lowered my memory usage by a lot.The dataset is copied to do both resolutions, so the step count is doubled from 1000 to 2000.I didn't save the list of buckets for v1.0, but this version's buckets are:<pre><code>INFO:musubi_tuner.dataset.image_video_dataset:bucket: (512, 800), count: 6
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (528, 768), count: 18
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (544, 752), count: 6
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (560, 720), count: 7
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (576, 704), count: 9
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (624, 656), count: 1
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (640, 640), count: 14
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (656, 624), count: 1
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (672, 608), count: 1
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (704, 576), count: 5
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (720, 560), count: 3
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (752, 544), count: 3
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (768, 528), count: 26
INFO:musubi_tuner.dataset.image_video_dataset:total batches: 100
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1072, 1632), count: 7
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1088, 1616), count: 16
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1104, 1584), count: 1
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1120, 1568), count: 3
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1136, 1552), count: 3
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1152, 1520), count: 1
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1168, 1504), count: 4
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1184, 1488), count: 9
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1200, 1456), count: 2
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1296, 1360), count: 1
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1312, 1344), count: 5
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1328, 1328), count: 6
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1344, 1312), count: 4
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1392, 1264), count: 1
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1456, 1200), count: 2
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1488, 1184), count: 3
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1504, 1168), count: 1
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1520, 1152), count: 3
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1552, 1136), count: 1
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1568, 1120), count: 1
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1616, 1088), count: 17
INFO:musubi_tuner.dataset.image_video_dataset:bucket: (1632, 1072), count: 9
INFO:musubi_tuner.dataset.image_video_dataset:total batches: 100</code></pre>

v1.1

Very simple training config for this. Most are captioned starting with "An analog photo of...", and the rest were only captioned with "An analog photo." Uses 1024*1024 buckets. Next version I will likely remove the trigger words.

v1.0

Analog (Style LoRA) | Qwen Image

1aa80c88615d91ddaf6fbeec7301d554.png

physical violence

weapon violence

wide hips

revealing clothes

downblouse

convenient censoring

pg-13

corpses

suggestive

oral invitation

pg13

sexy

huge breasts

thick thighs

sexual situations

male nudity

disturbing

male swimwear or underwear

female swimwear or underwear

partial nudity

undressed

female nudity

breasts out

exposed female nipple

breast out

lingerie

male underwear

hair over breasts

female swimwear

gigantic breasts

no panties

graphic violence or gore

covered nipples

huge butt

strapless leotard

sitting on face

emaciated bodies

one breast out

female underwear

nude

nsfw

graphic male nudity

adult toys

illustrated explicit nudity

nudity

graphic female nudity

hentai

futanari

porn

sexual intent

genitals

peeing

vore

oral

sexual activity

anal

blowjob

dildo riding

incest

hanging

hate symbols

nazi party

white supremacy

diapers

scat

self injury

hate speech

urine

extremist

child on child

latex clothing

swimwear

bukkake

fellatio

cumshot

implied fellatio

eat_cum

cumdrip

cum in pussy

cum on face

after fellatio

cum on hair

cum on body

cum on tongue

cum on hands

cum in mouth

triple fellatio

autofellatio

fucked silly

cum on pussy