Yet another amateur photo LoRA to improve realism.
Z-Image V1:
A new version for Z-Image trained on a small dataset.
V4 Beta 2:
Trained on more images.
V4 Beta 1:
I have started to create a completely new dataset with only high quality images. For Beta 1 I used 111 images, though I plan to significantly increase it for Beta 2. All images only have the caption "43stet1c", but it isn't needed to mention it in your prompt.
This combination seems to produce the best and most consistent results so far.
Increase the strength if the effect isn't strong enough.
V3:
Same dataset as V2 but completely recaptioned with Claude 3.5 Sonnet and upscaled to 1024 with Gigapixel.
V2:
Trained on 124 images with no captions at all. Let me know if this works better than V1. Results can be inconsistent, use HiRes fix and Beta scheduler.
V1:
This was trained on a dataset of 61 "aesthetic" images captioned with Claude 3.5 Sonnet.
Description
More than doubled the steps compared to V1
Stronger effect
Can be used with higher strength
FAQ
Comments (8)
usable with character lora?
I think so, haven’t tried it out.
Not really working for me, not even at 20%, it botches my model into an asian woman
Interesting to see json style prompts. They work with Flux 2 because they trained it that way but isn't it extra noise tokens for ZIT?
EDIT: Damn just tested your prompt format. Works really well. Thanks :)
EDIT2: Okay it kinda kills the prompt adherence and makes the results more stylized for some reason. Back to the drawing board haha
Yeah, just randomly tried it out and the results were fairly good. Still not sure what’s the best way to prompt for this model.
Sorry, but reality is v1 is overwhelmingly better than v2.
What do you like better in V1?
@powerpuff IMO, both are fine. I guess it depends on your definition of "Amateur". V1 feels more early 2000's with a grainy "first gen" digital camera feel that comes off as Amateurish. However, it's 2026, and most people's cell phone quality is similar to V2 if not better. Grainy was never meant to be a feature or filter, it was a byproduct of poor technology of the time period. Humanity has resolved that in the last 26 years for amateur photographers.
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.