Downloads · 30 days
2
2% of all-time downloads
kklors/DeepFocusLora
DeepFocusLora is a machine learning model from kklors. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as other.
This lora is trained on 80 hand picked high resolution f64 deep focus photographs. It does best with landscapes and architecture.
Downloads · 30 days
2
2% of all-time downloads
All-time downloads
110
Public
Repo size
345 MB
Likes
2
Public
Click a slice to open those files.
.pt173 MB · 50%
From the Hugging Face model README
This lora is trained on 80 hand picked high resolution f64 deep focus photographs. It does best with landscapes and architecture.
It is to overcome the shallow depth of field Flux often introduces. While a perfectly sharp image is not something you want to have all the time, it gives you more flexibility to add depth blur in post.
Example renders: <img src="https://i.ibb.co/JvqJDwV/Comfy-UI-temp-zbslp-00001-2.png"> <img src="https://i.ibb.co/7jcNjSZ/Comfy-UI-temp-zbslp-00002-1.png">
Kinda works on portraits too. It does not remove DOF entirely, but makes it less aggressive. F64 portrait photography doesn't exist anyway. <img src="https://i.ibb.co/LPLfC9n/Comfy-UI-temp-zbslp-00003.png">
Workflows should be embedded in the .pngs
Because the training data was shot on large scale film cameras using this lora will slightly stylize your image.