Skip to content

DataTunner 4.0.0

Latest

Choose a tag to compare

@LeandroCRX LeandroCRX released this 18 Sep 17:15
· 1 commit to main since this release

English

DataTunner 4.0.0 is the first stable release of the 4.x line and the version published on
PyPI: https://pypi.org/project/datatunner/4.0.0/

Highlights

  • Two selected alphas per experiment — alpha* (best cost-benefit) and alpha**
    (best effective), each validated on the held-out test set.
  • Explicit public APIs — DataTunnerTabular and DataTunnerImage, with lazy imports
    so tabular installs do not require PyTorch.
  • Model persistence — set artifact_dir to save every trained model per alpha and
    repetition plus the final selected models, with a models_index.json catalog and
    experiment_report.json; datatunner.load_model reloads any artifact.
  • Python 3.10+, 27 tests passing, ruff and mypy clean.
  • Bilingual documentation — README.md (English) and README.pt-BR.md (Portuguese).

Português

DataTunner 4.0.0 é a primeira versão estável da linha 4.x e a versão publicada no PyPI:
https://pypi.org/project/datatunner/4.0.0/

Destaques

  • Dois alphas selecionados por experimento — alpha* (melhor custo-benefício) e
    alpha** (melhor resultado efetivo), ambos validados no conjunto de teste.
  • APIs públicas explícitas — DataTunnerTabular e DataTunnerImage, com importação
    tardia para que instalações tabulares não exijam PyTorch.
  • Persistência de modelos — defina artifact_dir para salvar cada modelo treinado por
    alpha e repetição e os modelos finais selecionados, com o catálogo models_index.json e
    experiment_report.json; datatunner.load_model recarrega qualquer artefato.
  • Python 3.10+, 27 testes passando, ruff e mypy limpos.
  • Documentação bilíngue — README.md (inglês) e README.pt-BR.md (português).

Full changelog: V.1.0.0...v4.0.0