docs: explain trained localization model versions
Browse files- README.md +21 -0
- model-versions.json +13 -0
README.md
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: opensportslib
|
| 3 |
+
tags:
|
| 4 |
+
- action-spotting
|
| 5 |
+
- sports
|
| 6 |
+
---
|
| 7 |
+
# OSL-loc-NetVLADpp
|
| 8 |
+
|
| 9 |
+
OpenSportsLib localization models. `main` is the version catalog; load an explicit experiment revision using `LocalizationModel.from_pretrained`.
|
| 10 |
+
|
| 11 |
+
<!-- osl-model-versions:start -->
|
| 12 |
+
## Trained model versions
|
| 13 |
+
|
| 14 |
+
Select a version below to read its recipe, data, measured metrics and usage instructions. Each weight/config pair belongs to the linked experiment and seed.
|
| 15 |
+
|
| 16 |
+
| Version / README | Dataset | Seed | Scope | Measured test scores | Artifacts |
|
| 17 |
+
|---|---|---:|---|---|---|
|
| 18 |
+
| [soccernet_netvladpp_resnet_pca512](https://huggingface.co/OpenSportsLab/OSL-loc-NetVLADpp/tree/soccernet_netvladpp_resnet_pca512) | SoccerNet-v2 | 1 | released-code PCA ablation | loose average-mAP: 50.4990%; tight average-mAP: 11.9517% | [weights](https://huggingface.co/OpenSportsLab/OSL-loc-NetVLADpp/resolve/05e1b644fb77668c2ac49daeaef01779cddeb360/model.pth) 路 [config](https://huggingface.co/OpenSportsLab/OSL-loc-NetVLADpp/resolve/05e1b644fb77668c2ac49daeaef01779cddeb360/config.yaml) |
|
| 19 |
+
|
| 20 |
+
Per-seed measurements do not establish exact paper reproduction. Read each version's protocol and aggregation notes.
|
| 21 |
+
<!-- osl-model-versions:end -->
|
model-versions.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"soccernet_netvladpp_resnet_pca512": {
|
| 3 |
+
"card_url": "https://huggingface.co/OpenSportsLab/OSL-loc-NetVLADpp/tree/soccernet_netvladpp_resnet_pca512",
|
| 4 |
+
"config_url": "https://huggingface.co/OpenSportsLab/OSL-loc-NetVLADpp/resolve/05e1b644fb77668c2ac49daeaef01779cddeb360/config.yaml",
|
| 5 |
+
"dataset": "SoccerNet-v2",
|
| 6 |
+
"experiment": "soccernet_netvladpp_resnet_pca512",
|
| 7 |
+
"model_url": "https://huggingface.co/OpenSportsLab/OSL-loc-NetVLADpp/resolve/05e1b644fb77668c2ac49daeaef01779cddeb360/model.pth",
|
| 8 |
+
"revision": "soccernet_netvladpp_resnet_pca512",
|
| 9 |
+
"scope": "released-code PCA ablation",
|
| 10 |
+
"scores": "loose average-mAP: 50.4990%; tight average-mAP: 11.9517%",
|
| 11 |
+
"seed": 1
|
| 12 |
+
}
|
| 13 |
+
}
|