eandujar/classical_automl_notebook

Pokémon 151 AutoGluon Classifier

This repository contains the best-performing AutoGluon model selected using validation balanced accuracy.

Dataset

pakiino/2026-24679-pokemon-151-tabular-hw1

Model

• Framework: AutoGluon
• Problem type: Binary classification
• Target: label
• Positive class: 1
• Selection metric: Balanced accuracy
• Best model: LightGBM
• Testing performance: 1

The model was selected using the provided validation split.

Only the best-performing model and any model files required for its inference are retained in this repository.

Loading

from huggingface_hub import snapshot_download
from autogluon.tabular import TabularPredictor

model_path = snapshot_download(
    repo_id="eandujar/classical_automl_notebook"
)

predictor = TabularPredictor.load(model_path)