Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,75 @@
|
|
| 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 |
+
[Github Link](https://github.com/Phhofm/models/releases/tag/4xmssim_hma_pretrains)
|
| 11 |
+
|
| 12 |
+
Since no official [HMA](https://github.com/korouuuuu/HMA) model releases exist yet, I am releasing my [hma](https://github.com/muslll/neosr/blob/a50b56de418c938425fd3606697f62d61dbddfc3/neosr/archs/hma_arch.py#L1148) and [hma_medium](https://github.com/muslll/neosr/blob/a50b56de418c938425fd3606697f62d61dbddfc3/neosr/archs/hma_arch.py#L1386) mssim pretrains.
|
| 13 |
+
These can be used to speed up and stabilize early training stages when training new hma models.
|
| 14 |
+
Trained with mssim on [nomosv2](https://github.com/muslll/neosr/tree/a50b56de418c938425fd3606697f62d61dbddfc3?tab=readme-ov-file#-datasets).
|
| 15 |
+
|
| 16 |
+
|
| 17 |
+
## 4xmssim_hma_pretrain
|
| 18 |
+
|
| 19 |
+
**Scale:** 4
|
| 20 |
+
**Architecture:** [HMA](https://github.com/korouuuuu/HMA)
|
| 21 |
+
**Architecture Option:** [hma](https://github.com/muslll/neosr/blob/a50b56de418c938425fd3606697f62d61dbddfc3/neosr/archs/hma_arch.py#L1148)
|
| 22 |
+
|
| 23 |
+
**Author:** Philip Hofmann
|
| 24 |
+
**License:** CC-BY-0.4
|
| 25 |
+
**Purpose:** Pretrained
|
| 26 |
+
**Subject:** Photography
|
| 27 |
+
**Input Type:** Images
|
| 28 |
+
**Release Date:** 19.07.2024
|
| 29 |
+
|
| 30 |
+
**Dataset:** [nomosv2](https://github.com/muslll/neosr/tree/a50b56de418c938425fd3606697f62d61dbddfc3?tab=readme-ov-file#-datasets)
|
| 31 |
+
**Dataset Size:** 6000
|
| 32 |
+
**OTF (on the fly augmentations):** No
|
| 33 |
+
**Pretrained Model:** None (=From Scratch)
|
| 34 |
+
**Iterations:** 205'000
|
| 35 |
+
**Batch Size:** 4
|
| 36 |
+
**Patch Size:** 96
|
| 37 |
+
|
| 38 |
+
**Description:** A pretrain to start hma model training.
|
| 39 |
+
|
| 40 |
+
---
|
| 41 |
+
|
| 42 |
+
## 4xmssim_hma_medium_pretrain
|
| 43 |
+
|
| 44 |
+
**Scale:** 4
|
| 45 |
+
**Architecture:** [HMA](https://github.com/korouuuuu/HMA)
|
| 46 |
+
**Architecture Option:** [hma_medium](https://github.com/muslll/neosr/blob/a50b56de418c938425fd3606697f62d61dbddfc3/neosr/archs/hma_arch.py#L1386)
|
| 47 |
+
|
| 48 |
+
**Author:** Philip Hofmann
|
| 49 |
+
**License:** CC-BY-0.4
|
| 50 |
+
**Purpose:** Pretrained
|
| 51 |
+
**Subject:** Photography
|
| 52 |
+
**Input Type:** Images
|
| 53 |
+
**Release Date:** 19.07.2024
|
| 54 |
+
|
| 55 |
+
**Dataset:** [nomosv2](https://github.com/muslll/neosr/tree/a50b56de418c938425fd3606697f62d61dbddfc3?tab=readme-ov-file#-datasets)
|
| 56 |
+
**Dataset Size:** 6000
|
| 57 |
+
**OTF (on the fly augmentations):** No
|
| 58 |
+
**Pretrained Model:** None (=From Scratch)
|
| 59 |
+
**Iterations:** 150'000
|
| 60 |
+
**Batch Size:** 4
|
| 61 |
+
**Patch Size:** 48
|
| 62 |
+
|
| 63 |
+
**Description:** A pretrain to start hma_medium model training.
|
| 64 |
+
|
| 65 |
+
---
|
| 66 |
+
|
| 67 |
+
**Showcase:**
|
| 68 |
+
|
| 69 |
+
[slow.pics](https://slow.pics/c/Eoq18VDx)
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
|
| 73 |
+
|
| 74 |
+
|
| 75 |
+
|