Simple but useful layers for Pytorch
Project description
TorchSUL
This package is created for better experience while using Pytorch.
Why making this
-
For fun.
-
Path-dependence. I am addicted to my own wrap-ups.
-
Multi-platform. I have made the same APIs for pytorch, TF, MXNet, and a conversion tool to Caffe.
Installation
You need to install the newest version of pytorch.
Good, then just
pip install torchsul
Projects
You can find some examples in the "example" folder.
-
ArcFace (Deng, Jiankang, et al. "Arcface: Additive angular margin loss for deep face recognition." arXiv preprint arXiv:1801.07698 (2018))
-
HR Net (Sun, Ke, et al. "Deep High-Resolution Representation Learning for Human Pose Estimation." arXiv preprint arXiv:1902.09212 (2019))
-
AutoDeepLab (Liu, Chenxi, et al. "Auto-deeplab: Hierarchical neural architecture search for semantic image segmentation." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2019)
-
Knowledge distillation (Hinton, Geoffrey, Oriol Vinyals, and Jeff Dean. "Distilling the knowledge in a neural network." arXiv preprint arXiv:1503.02531 (2015))
-
3DCNN (Ji, Shuiwang, et al. "3D convolutional neural networks for human action recognition." IEEE transactions on pattern analysis and machine intelligence 35.1 (2012): 221-231)
-
Temporal Convolutional Network (Not the same) (Pavllo, Dario, et al. "3D human pose estimation in video with temporal convolutions and semi-supervised training." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2019)
-
Model conversions
-
Batch_norm compression to speed-up models
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for TorchSUL-0.1.12-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 987cba0ac6e550b659e97bb05d8217cfaaef2b0e72436bd13eaee082fa5fcd9c |
|
MD5 | 7363321bc811314e0de3e0eb00cc1ef5 |
|
BLAKE2b-256 | 56eab50644488dba8d59869724fbb18339475ad3dbadc045297885acd553636c |