Commit
•
42eb3b8
1
Parent(s):
a0655a7
Upload docs
Browse files- .gitattributes +2 -0
- README.md +23 -0
.gitattributes
CHANGED
@@ -32,3 +32,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
32 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
33 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
34 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
32 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
33 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
34 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
35 |
+
variables/variables.data-00000-of-00001 filter=lfs diff=lfs merge=lfs -text
|
36 |
+
variables/variables.index filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
@@ -1,3 +1,26 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
---
|
4 |
+
# WD 1.4 MOAT Tagger V2
|
5 |
+
|
6 |
+
Supports ratings, characters and general tags.
|
7 |
+
|
8 |
+
Trained using https://github.com/SmilingWolf/SW-CV-ModelZoo.
|
9 |
+
TPUs used for training kindly provided by the [TRC program](https://sites.research.google/trc/about/).
|
10 |
+
|
11 |
+
## Dataset
|
12 |
+
Last image id: 5944504
|
13 |
+
Trained on Danbooru images with IDs modulo 0000-0899.
|
14 |
+
Validated on images with IDs modulo 0950-0999.
|
15 |
+
Images with less than 10 general tags were filtered out.
|
16 |
+
Tags with less than 600 images were filtered out.
|
17 |
+
|
18 |
+
## Validation results
|
19 |
+
`P=R: threshold = 0.3771, F1 = 0.6911`
|
20 |
+
|
21 |
+
## Paper
|
22 |
+
[`MOAT: Alternating Mobile Convolution and Attention Brings Strong Vision Models`](https://arxiv.org/abs/2210.01820)
|
23 |
+
|
24 |
+
## Final words
|
25 |
+
Subject to change and updates.
|
26 |
+
Downstream users are encouraged to use tagged releases rather than relying on the head of the repo.
|