The train/validation/test split is grouped by generated game round. Metrics describe the Mod-generated distribution and are not a claim of identical accuracy on arbitrary live captures.

Prediction examples

Batch 0	Batch 1	Batch 2

Training diagnostics

License

MIT.

proj-airi/games-balatro-2024-yolo-card-edition-detection-mod-ground-truth

Balatro card edition detection

Files

Independent test split

Prediction examples

Training diagnostics

License

proj-airi/
games-balatro-2024-yolo-card-edition-detection-mod-ground-truth