Instructions to use thangkt/PCB-Prune-YOLO-P10-DepGraph with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- ultralytics
How to use thangkt/PCB-Prune-YOLO-P10-DepGraph with ultralytics:
from ultralytics import YOLOvv8 model = YOLOvv8.from_pretrained("thangkt/PCB-Prune-YOLO-P10-DepGraph") source = 'http://images.cocodataset.org/val2017/000000039769.jpg' model.predict(source=source, save=True) - Notebooks
- Google Colab
- Kaggle
| { | |
| "selection_split": "val", | |
| "test_used_for_selection": false, | |
| "sparse_training": { | |
| "checkpoint": "outputs/sparse/depgraph_sparse_p10_reg5e4/weights/best.pt", | |
| "regularization_strength": 0.0005, | |
| "epochs_completed": 30, | |
| "best_epoch": 10, | |
| "precision": 0.97646, | |
| "recall": 0.95673, | |
| "mAP50": 0.98539, | |
| "mAP50_95": 0.78938, | |
| "regularizer_gradient_nonzero_all_epochs": true, | |
| "regularizer_new_nonfinite_total": 0, | |
| "near_zero_fraction_max": 0.0, | |
| "norm_change_percent_epoch1_to_epoch30": { | |
| "min": -0.095445, | |
| "mean": -0.004166, | |
| "median": -0.018078, | |
| "max": 0.049815 | |
| } | |
| }, | |
| "p10_before_finetune": { | |
| "checkpoint": "outputs/pruning_sparse_reg5e4/p10/pruned.pt", | |
| "parameters": 2415613, | |
| "macs": 3232802400, | |
| "precision": 0.0045010777, | |
| "recall": 0.0512265512, | |
| "mAP50": 0.0024274871, | |
| "mAP50_95": 0.0003512981, | |
| "mean_latency_ms": 10.126654, | |
| "fps": 98.749299, | |
| "reload_inference": true | |
| }, | |
| "p10_after_finetune": { | |
| "checkpoint": "outputs/finetune_sparse_reg5e4/p10/weights/best.pt", | |
| "epochs_completed": 37, | |
| "best_epoch": 27, | |
| "parameters": 2415613, | |
| "macs": 3232802400, | |
| "precision": 0.9590941498, | |
| "recall": 0.948691531, | |
| "mAP50": 0.9812432157, | |
| "mAP50_95": 0.7631779975, | |
| "mean_latency_ms": 9.718659, | |
| "fps": 102.894852, | |
| "model_size_mb": 4.850336, | |
| "peak_gpu_memory_mb": 40.015625, | |
| "reload_inference": true | |
| }, | |
| "comparison_to_baseline": { | |
| "parameter_reduction_percent": 19.800844, | |
| "mac_reduction_percent": 20.633661, | |
| "mAP50_95_change_points": -2.2062, | |
| "latency_change_percent": 17.247668, | |
| "fps_change_percent": -14.709175 | |
| } | |
| } | |