Poke Grader YOLO Finetune

--- language: en license: apache-2.0 tags: - pokemon - object-detection - yolo - yolov11 - card-detection metrics: - map model-index: - name: pacoalberola/Poke-Grader-YOLO-Finetune results: - task: type: object-detection metrics: - name: mAP@50 type: map value: 99.50% - name: mAP@50-95 type: map_50_95 value: 98.92% --- # 🃏 PokeGrader — YOLO Card Detector Fine-tune de **YOLOv11n** para localizar cartas Pokémon TCG en imágenes. ## 📊 Métricas | Métrica | Valor | |---------|-------| | mAP@50 | 99.50% | | mAP@50-95 | 98.92% | | Precision | 99.48% | | Recall | 100.00% | ## ⚙️ Configuración de entrenamiento | Parámetro | Valor | |-----------|-------| | Epochs | 30 | | Batch size | 16 | | Image size | 640 | | LR inicial (lr0) | 0.001 | ## 🛠️ Uso en backend Este archivo `.onnx` es el que consume el servidor Railway en `ai_vision.py`. ```python import onnxruntime as ort sess = ort.InferenceSession('yolo_pokemon_card.onnx') ```

Source context: 0 downloads · 0 likes · Pipeline object-detection · Repo pacoalberola/Poke-Grader-YOLO-Finetune