Toolbelt for PiePline training pipeline
Project description
PiePline toolbelt
Installation:
pip install pietoolbelt
Install latest version before it's published on PyPi
pip install -U git+https://github.com/PiePline/pietoolbelt
Functional
- Datasets
datasets.stratification
- stratification by histogramdatasets.utils
- set of datasets constructors that
- Losses
losses.common
- losses utilslosses.regression
- regression losseslosses.segmentation
- losses for single and multi-class segmentationlosses.detection
- losses for detection task
- Metrics
metrics.common
- common utils for metrics- CPU - metrics, that calculates by
numpy
metrics.cpu.classification
- classification metricsmetrics.cpu.detection
- detection metricsmetrics.cpu.regression
- regression metricsmetrics.cpu.segmentation
- segmentation metrics
- Torch - metrics, that calculates by
torch
metrics.torch.classification
- classification metricsmetrics.torch.detection
- detection metricsmetrics.torch.regression
- regression metricsmetrics.torch.segmentation
- segmentation metrics
- Models
decoders.unet
- UNet decoder, that automatically constructs by encoderencoders.common
- basic interfaces for encodersencoders.inception
- Inceptionv3 encoderencoders.mobile_net
- MobileNetv2 encoderencoders.resnet
- ResNet encodersalbunet
- albunet modelutils
- models utilsweights_storage
- pretrained weights storage
- Pipeline steps
regression.train
- train step for regression taskregression.bagging
- bagging step for regression task
img_matcher
- image comparision and matching tool based on descriptorsmask_composer
- mask composer tools that can effectively combine masks for regular, instance or multiclass segmentationutils
- some utilsviz
- image visualisation tools
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 Distributions
No source distribution files available for this release.See tutorial on generating distribution archives.
Built Distribution
Close
Hashes for pietoolbelt-0.2.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 6398131eea824c2e465b944370f8f241dfb21e65ad70c066a4033c8b4c3b4350 |
|
MD5 | e2cfae7b5dcae56ace65dc8b7bc26218 |
|
BLAKE2b-256 | bd66d4e1c166faf4037e5a5489124363532248572f31552f0a0dac5ef156270a |