Hugging Face's logo

Phips
/
4xmssim_realplksr_dysample_pretrain

Image-to-Image
PyTorch
super-resolution
pretrain
Model card Files Community
Phips commited on
Commit
13be7b3
·
verified ·
1 Parent(s): 0cf9735

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +48 -3
README.md CHANGED
@@ -1,3 +1,48 @@
1
- ---
2
- license: cc-by-4.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ pipeline_tag: image-to-image
4
+ tags:
5
+ - pytorch
6
+ - super-resolution
7
+ - pretrain
8
+ ---
9
+
10
+ [Link to Github Release](https://github.com/Phhofm/models/releases/tag/4xmssim_realplksr_dysample_pretrain)
11
+
12
+ # 4xmssim_realplksr_dysample_pretrain
13
+ Scale: 4
14
+ Architecture: [RealPLKSR with Dysample](https://github.com/muslll/neosr/?tab=readme-ov-file#supported-archs)
15
+ Architecture Option: [realplksr](https://github.com/muslll/neosr/blob/master/neosr/archs/realplksr_arch.py)
16
+
17
+ Author: Philip Hofmann
18
+ License: CC-BY-0.4
19
+ Purpose: Pretrained
20
+ Subject: Photography
21
+ Input Type: Images
22
+ Release Date: 27.06.2024
23
+
24
+ Dataset: [nomosv2](https://github.com/muslll/neosr/?tab=readme-ov-file#-datasets)
25
+ Dataset Size: 6000
26
+ OTF (on the fly augmentations): No
27
+ Pretrained Model: None (=From Scratch)
28
+ Iterations: 200'000
29
+ Batch Size: 8
30
+ GT Size: 192, 512
31
+
32
+ Description: [Dysample](https://arxiv.org/pdf/2308.15085) had been recently added to RealPLKSR, which from what I had seen can resolve or help avoid the checkerboard / grid pattern on inference outputs. So with the [commits from three days ago, the 24.06.24, on neosr](https://github.com/muslll/neosr/commits/master/?since=2024-06-24&until=2024-06-24), I wanted to create a 4x photo pretrain I can then use to train more realplksr models with dysample specifically to stabilize training at the beginning.
33
+
34
+ Showcase:
35
+ [Imgsli](https://imgsli.com/Mjc0OTA1)
36
+ [Slowpics](https://slow.pics/c/I9grkcqM)
37
+
38
+ ![Example1](https://github.com/Phhofm/models/assets/14755670/25406570-3388-4d22-ae68-68560c8bd917)
39
+ ![Example2](https://github.com/Phhofm/models/assets/14755670/bf3e946b-9646-441e-a15e-9bbc290a8885)
40
+ ![Example3](https://github.com/Phhofm/models/assets/14755670/7e5ccec4-f485-4e02-a76b-9ec8827ee663)
41
+ ![Example4](https://github.com/Phhofm/models/assets/14755670/9c665c12-c30f-4b7f-a0a6-46a3645633fe)
42
+ ![Example5](https://github.com/Phhofm/models/assets/14755670/4868ba82-fe8a-468c-bfd4-1f471d3ba361)
43
+ ![Example6](https://github.com/Phhofm/models/assets/14755670/3ee42167-721f-4866-8529-d0a19f121ff1)
44
+ ![Example7](https://github.com/Phhofm/models/assets/14755670/a3a23246-8f58-4810-823b-10b701bdbced)
45
+ ![Example8](https://github.com/Phhofm/models/assets/14755670/f6665676-0845-4019-a476-ed253306f838)
46
+ ![Example9](https://github.com/Phhofm/models/assets/14755670/3da3a0d9-ef5e-4e66-a1d1-03dae2bf437b)
47
+ ![Example10](https://github.com/Phhofm/models/assets/14755670/3468d468-09a0-42e4-945b-f63e48d9744b)
48
+ ![Example11](https://github.com/Phhofm/models/assets/14755670/83cfa6d2-3fc9-4099-834a-129a6e87e4de)