james-kramer/tabular-autogluon-predictor

license: mit language:

• en pipeline_tag: tabular-classification tags:
• pokemon datasets:
• ecopus/pokemon_cards metrics:
• accuracy
• f1 library_name: autogluon

Model Details: -This model was trained using AutoGluon's best_quality preset, which automatically selects the best of various models to maximize accuracy

Training Data -This model was trained on the ecopus/pokemon_cards dataset.

Evaluation Data -This model was evaluated on the test split of the augmented data (df_synth_test) and original data (df_orig) from ecopus/pokemon_cards.

Performance -On the augmented test set, both the accuracy and weighted F1 were 1.0000. -On the original data, the accuracy was 0.8235 and the weighted F1 was 0.8135

Usage -This model can be loaded and used with AutoGluon and the huggingface_hub library.