Skip to main content

EGE Classification

Этот проект предназначен для классификации задач из профильного ЕГЭ по математике. Модель предсказывает номер задачи с точностью более 99%. Главная проблема - разделение 4 и 5 задач (теория вероятностей). Актуальная версия обучена на базе cointegrated/rubert-tiny2 (около 20M параметров). Старая версия (0.0.4 модуля; v1 модели на hf) обучена на базе distilbert-base-uncased (около 70M параметров). Настоятельно рекомендую актуальную версию. Задачи актуальны на Июль 2026 года.

Установка

Вы можете самостоятельно собрать библиотеку

poetry build
pip install ПУТЬ_К_АРХИВУ

Или воспользоваться pypi.org

pip install ege-classification

Использование

Обучение модели

Для обучения модели используйте следующий код:

from ege_classification.model import TaskClassifier

classifier = TaskClassifier()
classifier.train('path/to/dataset.csv')
classifier.save()

Тестирование модели

Для тестирования модели используйте следующий код:

classifier.test('path/to/dataset.csv')

Предсказание

Для предсказания номера задачи используйте следующий код:

predicted_class = classifier.predict("Текст задачи")
print(f"Предсказанный номер задачи: {predicted_class}")

Загрузка и сохранение модели

Для сохранения модели:

classifier.save('path/to/save/model')

Для загрузки модели:

classifier.load('path/to/saved/model')

Лицензия

Этот проект лицензирован под лицензией MIT.

Metadata

Release files for ege_classification 0.0.5

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for ege_classification 0.0.5
File Size Uploaded Provenance Actions
ege_classification-0.0.5.tar.gz 3.8 kB Without provenance Details for ege_classification-0.0.5.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for ege_classification 0.0.5
File Interpreter ABI Platform Actions
ege_classification-0.0.5-py3-none-any.whl Python 3 none any Details for ege_classification-0.0.5-py3-none-any.whl

Total release size: 8.9 kB

Release files / ege_classification-0.0.5.tar.gz

Download URL ege_classification-0.0.5.tar.gz
Size 3.8 kB
Tags Source
SHA-256 checksum
How to use checksums
2f7f315e84572e24edca7a17bbd1566bad7c005c5601aa49708d3da8ad727170
BLAKE2b-256 checksum
How to use checksums
360b9cbc4b4865e565f0b43b24086f3001ee80c98a3bd693aea94517449b3dd8
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via poetry/2.4.1 CPython/3.14.6 Linux/6.17.0-1020-azure

Release files / ege_classification-0.0.5-py3-none-any.whl

Download URL ege_classification-0.0.5-py3-none-any.whl
Size 5.1 kB
Tags Python 3
SHA-256 checksum
How to use checksums
fac6fcf8ef5fedd6c719d8b4196831c56524de8d8594f37f4e06809e01deba28
BLAKE2b-256 checksum
How to use checksums
dc55c5f0bcec79de342ec02be4cdba6b368b3357ced5cb94bd9a5a84a637ca31
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via poetry/2.4.1 CPython/3.14.6 Linux/6.17.0-1020-azure

Security for ege_classification 0.0.5

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page