Kekas
Kek it easy.
Kekas is a simple tool for training neural networks on Pytorch.
I tried to keep it as simple as possible.
Installation
pip install kekas
Mixed Precision requirements
Kekas use https://github.com/NVIDIA/apex library for mixed precision training, so follow the installation instructions from its repo.
Features
- Awesome name
- Mixed Precision (FP16)
- Learning Rate Finder
- One Cycle policy
- Tensoboard logging
- Best checkpoints saving
- Early stopping
- TTA
- Freeze / unfreeze
- Easy customization
Quick start guide
I don't beieve in quick start guides, I think that they create more questions than answers.
Instead, I've created a detailed Tutorial notebook. Read it.
Contribution guide
Just contribute something good and don't contribute anything bad.
Citing
If you find this library useful for your research, please consider citing:
@misc{aleksandr belskikh_2019,
author = {Aleksandr Belskikh},
title = {{kekas: Just another DL library}},
month = dec,
year = 2019,
doi = {10.5281/zenodo.2577861},
version = {0.1.23},
publisher = {Zenodo},
url = {https://doi.org/10.5281/zenodo.2577861},
}
1 maintainer
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
kekas-0.1.23.tar.gz
(19.0 kB
view details)
File details
Details for the file kekas-0.1.23.tar.gz.
File metadata
- Download URL: kekas-0.1.23.tar.gz
- Upload date:
- Size: 19.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via:
twine/3.1.1 pkginfo/1.5.0.1 requests/2.22.0 setuptools/41.4.0 requests-toolbelt/0.9.1 tqdm/4.36.1 CPython/3.7.4
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
b0df1988542658531dedeef8976b986f857e91cfccafe9b6f385240be9d8d79f
|
|
| MD5 |
6b330bf5ed72c5ae7c2567e0c4f570cd
|
|
| BLAKE2b-256 |
039314a1e626773dbea43ced803fd2ebfb5534ec9d05f08b74ca23bc4a7341f4
|
