A typed dataset abstraction toolkit for machine learning projects
Project description
InstanceLib
An abstract interface for datasets. This library is hosted on https://git.science.uu.nl/mpbron-phd/instancelib and available for usage according to the GNU LGPL v3 license. This library is derived from the data management part of Allib.
Installation
Clone this repository and then execute the following command within the repo directory:
pip install -e "./"
.
Usage
See usage.py
for examples on how to use this library.
TODO:
- Tokenization mapping (which characters were removed)
- Function f: Instance -> LimePertubation -> Sequence[Instance']
- Machine Learning Wrappers that can work with InstanceProviders (cf. allib)
- Feature Extraction (cf. allib?)
- Method that converts One-hot Encoding of labels to Human Readable format
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
instancelib-0.2.1.tar.gz
(30.5 kB
view hashes)
Built Distribution
Close
Hashes for instancelib-0.2.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | b3777c0a53bdfff5f91a1ea506b890662a5afde8363847a35756b3261bdd41b1 |
|
MD5 | 32d6e8710648155aac190f96906fa57b |
|
BLAKE2b-256 | 648dd0b1e9be6134e7174970901878cffa4c1887f69ea58ce2cef6a873be9742 |