Modelpublic

KaushiGihan/detr-kids-and-adults-detection-finetuned

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face

1likes31downloads

Model Card

detr-kids-and-adults-detection-finetuned

This model is a fine-tuned version of facebook/detr-resnet-50-dc5 on the None dataset. It achieves the following results on the evaluation set:

—Loss: 1.7496
—Map: 0.0043
—Map 50: 0.0096
—Map 75: 0.0039
—Map Small: -1.0
—Map Medium: 0.0051
—Map Large: 0.0073
—Mar 1: 0.0175
—Mar 10: 0.0983
—Mar 100: 0.3165
—Mar Small: -1.0
—Mar Medium: 0.5364
—Mar Large: 0.3035
—Map Kid: 0.0029
—Mar 100 Kid: 0.2694
—Map Adult: 0.0056
—Mar 100 Adult: 0.3636

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

—learning_rate: 1e-05
—trainbatchsize: 4
—evalbatchsize: 4
—seed: 42
—optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—training_steps: 100
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Map	Map 50	Map 75	Map Small	Map Medium	Map Large	Mar 1	Mar 10	Mar 100	Mar Small	Mar Medium	Mar Large	Map Kid	Mar 100 Kid	Map Adult	Mar 100 Adult
1.9155	0.1852	20	1.8038	0.0047	0.0101	0.005	-1.0	0.0035	0.0071	0.0332	0.1231	0.3206	-1.0	0.4545	0.3067	0.0013	0.1685	0.0081	0.4727
1.8395	0.3704	40	1.7979	0.0045	0.0107	0.0035	-1.0	0.0042	0.0072	0.0223	0.1147	0.3081	-1.0	0.4909	0.2926	0.0014	0.1726	0.0076	0.4436
1.8477	0.5556	60	1.7656	0.0046	0.0107	0.0041	-1.0	0.0044	0.0077	0.0329	0.1148	0.3276	-1.0	0.5091	0.3137	0.002	0.2242	0.0072	0.4309
1.66	0.7407	80	1.7440	0.0044	0.0098	0.0039	-1.0	0.0049	0.0076	0.0256	0.1029	0.3266	-1.0	0.5182	0.3137	0.0026	0.2532	0.0062	0.4
2.0088	0.9259	100	1.7496	0.0043	0.0096	0.0039	-1.0	0.0051	0.0073	0.0175	0.0983	0.3165	-1.0	0.5364	0.3035	0.0029	0.2694	0.0056	0.3636

Framework versions

—Transformers 4.48.0
—Pytorch 2.5.1+cu121
—Datasets 3.2.0
—Tokenizers 0.21.0