First test with flux concept.
Trained on 35mm photographs from multiple photographers mostly on landscapes, items and buildings. It took on the lighting and esthetics of the dataset pretty well but needs to be pushed into doing grainy pictures if you want those.
Use 35mm Photo, film grain in prompt for best results although it should work without.
Z-Turbo:
First real lora training for this model. This was trained using the biggest bucket available in AI-toolkit (from 768 to 1532px).
Used the same dataset as I did with Flux but added some more pictures and described the ones I had a lot better to train this one.
Whilst the model is already very capable to do photography, this has to ability to have a bit more contrast and darkness in images.
I keep the value at around 0.8
35mm is trigger word. I use high contrast or film grain / grainy shot depending on what i want to achieve as well.
Description
FAQ
Comments (6)
This is the closest we have ever been to creating real life photos that never happened. Amazing what the LORA is able to do. Great job!
nice one, but it seems it fries flux' prompt adherence a tiny bit. how were the training captions constructed?
Mostly manually but given how flux works a bit too short and I'll need to do a redo with more detailed descriptions. Given the dataset, and that I'm busy working on another one. It'll take some time untill I can retrain.
can u share your wf ,thx
Looks great, but can you share at least one prompt with parameters to start with? I would be very grateful!
advice：find some not 35mm pics，and tag them 50 mm,20 mm,100 mm，it will help ai konw what excatly 35mm looks like
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.