license: cc-by-4.0
pipeline_tag: image-to-image
tags:
- pytorch
- super-resolution
4xNomos2_hq_dat2
Scale: 4
Architecture: DAT
Architecture Option: dat2
Author: Philip Hofmann
License: CC-BY-0.4
Purpose: Upscaler
Subject: Photography
Input Type: Images
Release Date: 29.08.2024
Dataset: nomosv2
Dataset Size: 6000
OTF (on the fly augmentations): No
Pretrained Model: DAT_2_x4
Iterations: 140'000
Batch Size: 4
Patch Size: 48
Description:
A dat2 4x upscaling model, similiar to the 4xNomos2_hq_mosr model, trained and for usage on non-degraded input to give good quality output.
I scored 7 validation outputs of each of the 21 checkpoints (10k-210k) of this model training with 68 metrics.
The metric scores can be found in this google sheet.
The corresponding image files for this scoring can be found here
Screenshot of the google sheet:
Release checkpoint has been selected by looking at the scores, manually inspecting, and then getting responses on discord which chose B to this quick visual test, A B or C, which denote different checkpoints: https://slow.pics/c/8Akzj6rR
Checkpoint B is 140k which is 4xNomos2_hq_dat2. But I added checkpoint A (4xNomos2_hq_dat2_150000) and checkpoint C (4xNomos2_hq_dat2_10000) model files additionally here if people want to try them out).
Model Showcase:
(Click on image for better view)