Upload folder using huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: pytorch
|
| 4 |
+
tags:
|
| 5 |
+
- video-classification
|
| 6 |
+
- lpcv
|
| 7 |
+
- qevd
|
| 8 |
+
- slowfast
|
| 9 |
+
- r2plus1d
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# LPCV Track2 EfficientAI Models
|
| 13 |
+
|
| 14 |
+
This repository provides model weights for our LPCV Track 2 video classification solution.
|
| 15 |
+
|
| 16 |
+
## Files
|
| 17 |
+
|
| 18 |
+
- `slowfast_r101_16x4_QEVD_sup.pth`: MMAction2 SlowFast-R101 teacher model.
|
| 19 |
+
- `r2plus1d_r18_16x4_QEVD_sup.pth`: distilled TorchVision R(2+1)D-18 student model.
|
| 20 |
+
- `SHA256SUMS`: checksums for released weights.
|
| 21 |
+
|
| 22 |
+
## Dataset Notice
|
| 23 |
+
|
| 24 |
+
These weights were trained with supplemental QEVD data and the original Qualcomm QEVD dataset.
|
| 25 |
+
|
| 26 |
+
## Usage
|
| 27 |
+
|
| 28 |
+
See the project repository README for training, distillation, and export instructions.
|