---
license: cc-by-4.0
pipeline_tag: image-to-image
tags:
- pytorch
- super-resolution
---

[Link to Github Release](https://github.com/Phhofm/models/releases/tag/4xFaceUpDAT_Series)

# 4xFaceUpDAT

Name: 4xFaceUpDAT  
Author: Philip Hofmann  
Release Date: 02.09.2023  
License: CC BY 4.0  
Network: DAT  
Scale: 4  
Purpose: 4x upscaling model for faces  
Iterations: 140000  
epoch: 54  
batch_size: 4  
HR_size: 128  
Dataset: FaceUp (see dataset-releases channel)  
Number of train images: 10'000  
OTF Training: Yes  
Pretrained_Model_G: DAT_x4.pth  

Description: 4x photo upscaler for faces, trained on the FaceUp dataset. These models are an improvement over the previously released 4xFFHQDAT and are its successors. These models are released together with the FaceUp dataset, plus the accompanying [youtube video](https://www.youtube.com/watch?v=TBiVIzQkptI)

This model comes in 4 different versions:  
4xFaceUpDAT (for good quality input)  
4xFaceUpLDAT (for lower quality input, can additionally denoise)  
4xFaceUpSharpDAT (for good quality input, produces sharper output, trained without USM but sharpened input images, good quality input)  
4xFaceUpSharpLDAT (for lower quality input, produces sharper output, trained without USM but sharpened input images, can additionally denoise)  

Web Examples (Slowpoke pics):  

High quality input:  
https://slow.pics/c/rsHKvfv3  
https://slow.pics/c/XMAcyBVV  
https://slow.pics/c/yWQKYSea  

Low quality input:  

https://slow.pics/c/QboAlS0t

-- additional info about the other model versions since I make a single entry here --
4xFaceUpLDAT - iters: 80k, pretrain: 4xFaceUpDAT  
4xFaceUpSharpDAT - iters: 100k, pretrain: 4xFaceUpDAT  
4xFaceUpSharpLDAT - iters: 80k, pretrain: 4xFaceUpSharpDAT