An adaptation of pytorch OneCycleLR that doesn't crash when max steps is reached
Project description
OneLongCycleLR
This is a simple adaptation from the OneCycleLR of pytorch.optim package. The default behaviour is to raise a ValueError when the maximum number of steps is reached. This one instead just keeps on returning the last value.
To install it, just use:
pip install onelongcyclelr
And to use it you need to import it like
from onelongcyclelr import OneLongCycleLR
The arguments and keywords needed are exactly the same as the official implementation.
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
onelongcyclelr-0.1.1.tar.gz
(2.3 kB
view hashes)
Built Distribution
Close
Hashes for onelongcyclelr-0.1.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | edab7dff68de134b716045ffa40dd3a830cf47226a1c4a61ccc03c9444ba3e03 |
|
MD5 | b24134c7affe526a8308c54ffbdd504e |
|
BLAKE2b-256 | 14b06a624ddf63de945b5b847a30d5dd5a342fcc4e08f83510dbe2466ce07a3b |