Skip to main content

Фреймворк для работы с пайплайном ML моделей

Project description

ML Pipeline Engine

Графовый движок для конвейеров ML-моделей

Table of Contents

Usage

Что нужно, чтобы сделать свой пайплайн?

  1. Написать классы узлов
  2. Связать узлы посредством указания зависимости

Поддерживаемые типы узлов

Протоколы

  1. DataSource
  2. FeatureBase
  3. MLModelBase
  4. ProcessorBase
  5. FeatureVectorizerBase

Примеры использования описаны в файле docs/usage_examples.md

Development

Environment setup

Clone the project

git clone https://github.com/tochka-public/ml-pipeline-engine.git

Go to the project directory

cd ml-pipeline-engine

Use Python>=3.8 and the package manager poetry to install ml-pipeline-engine dependencies

poetry install --no-root

For further contribution, use pre-commit hooks to maintain consistent code format

pre-commit install -f --hook-type pre-commit --hook-type pre-push

Run tests

python -m pytest tests

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ml_pipeline_engine-2.0.0a3.tar.gz (1.3 MB view details)

Uploaded Source

Built Distribution

ml_pipeline_engine-2.0.0a3-py3-none-any.whl (1.4 MB view details)

Uploaded Python 3

File details

Details for the file ml_pipeline_engine-2.0.0a3.tar.gz.

File metadata

  • Download URL: ml_pipeline_engine-2.0.0a3.tar.gz
  • Upload date:
  • Size: 1.3 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/5.1.0 CPython/3.9.19

File hashes

Hashes for ml_pipeline_engine-2.0.0a3.tar.gz
Algorithm Hash digest
SHA256 a86c33ee1d829cce4cef3da27bf494c6f980f3a6b647d36fc15eeb14210c2458
MD5 c5b59fab44fc3ebef6627e2583103925
BLAKE2b-256 5abd4f4e5ae0f127d504b4984de77e94b2418fb59552680fddd918ee91308f68

See more details on using hashes here.

File details

Details for the file ml_pipeline_engine-2.0.0a3-py3-none-any.whl.

File metadata

File hashes

Hashes for ml_pipeline_engine-2.0.0a3-py3-none-any.whl
Algorithm Hash digest
SHA256 84732bc9c3e1d47d8c6269d7cc72e6390d6ae5ee1a91f43dcfda58782f0f13bf
MD5 50f61fa2a3828f929c8e792f8f739aa1
BLAKE2b-256 eeb4bf2525a04905c5507414baa7d49bc67113c3082f358b6c69b8c2fe549a0c

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page