rtdetr-r50-drone-detection-0205

This model is a fine-tuned version of PekingU/rtdetr_r50vd_coco_o365 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 6.7461
• Precision: 1.0
• Recall: 0.0169
• F1: 0.0333
• Total Gt: 59
• Tp: 1
• Fp: 0
• Fn: 58
• Map: 1.0

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30

Training results

Training Loss	Epoch	Step	F1	Fn	Fp	Validation Loss	Map	Precision	Recall	Total Gt	Tp
58.72	1.0	229	0.0094	51	1629	50.8780	0.0049	0.0049	0.1356	59	8
14.2567	2.0	458	0.0935	54	43	7.7872	0.1042	0.1042	0.0847	59	5
12.8301	3.0	687	0.1519	53	14	6.7555	0.3	0.3	0.1017	59	6
10.5614	4.0	916	0.1429	54	6	6.1924	0.4545	0.4545	0.0847	59	5
11.0246	5.0	1145	0.1714	53	5	5.9438	0.5455	0.5455	0.1017	59	6
11.2422	6.0	1374	0.1429	54	6	5.8866	0.4545	0.4545	0.0847	59	5
8.5436	7.0	1603	0.1471	54	4	5.6235	0.5556	0.5556	0.0847	59	5
9.086	8.0	1832	0.1471	54	4	5.7372	0.5556	0.5556	0.0847	59	5
8.0738	9.0	2061	0.1471	54	4	5.7808	0.5556	0.5556	0.0847	59	5
8.2842	10.0	2290	0.1429	54	6	5.7614	0.4545	0.4545	0.0847	59	5
9.667	11.0	2519	0.0	59	0	6.0685	0.0	0.0	0.0	59	0
10.0863	12.0	2748	0.0	59	0	5.9248	0.0	0.0	0.0	59	0
9.488	13.0	2977	0.0	59	0	6.0921	0.0	0.0	0.0	59	0
8.6154	14.0	3206	5.7059	1.0	0.0847	0.1562	59	5	0	54	1.0
8.3708	15.0	3435	5.8087	1.0	0.0847	0.1562	59	5	0	54	1.0
7.8841	16.0	3664	5.8849	1.0	0.0678	0.127	59	4	0	55	1.0
7.9895	17.0	3893	6.0315	1.0	0.0508	0.0968	59	3	0	56	1.0
7.4122	18.0	4122	5.8645	1.0	0.0339	0.0656	59	2	0	57	1.0
7.0292	19.0	4351	6.1574	1.0	0.0339	0.0656	59	2	0	57	1.0
6.7216	20.0	4580	5.9338	1.0	0.0339	0.0656	59	2	0	57	1.0
6.7722	21.0	4809	6.2863	0.0	0.0	0.0	59	0	0	59	0.0
6.7148	22.0	5038	6.3802	1.0	0.0169	0.0333	59	1	0	58	1.0
6.2273	23.0	5267	6.3039	1.0	0.0169	0.0333	59	1	0	58	1.0
6.4498	24.0	5496	6.4376	1.0	0.0169	0.0333	59	1	0	58	1.0
5.7548	25.0	5725	6.3513	1.0	0.0169	0.0333	59	1	0	58	1.0
6.1244	26.0	5954	6.4635	1.0	0.0169	0.0333	59	1	0	58	1.0
5.3083	27.0	6183	6.5638	1.0	0.0169	0.0333	59	1	0	58	1.0
5.5927	28.0	6412	6.5546	1.0	0.0169	0.0333	59	1	0	58	1.0
5.1132	29.0	6641	6.7564	1.0	0.0169	0.0333	59	1	0	58	1.0
5.3641	30.0	6870	6.7461	1.0	0.0169	0.0333	59	1	0	58	1.0

Framework versions

• Transformers 4.57.1
• Pytorch 2.8.0+cu126
• Datasets 4.2.0
• Tokenizers 0.22.1