panels_detection_rtdetr_augmented_consolidated_labels

This model is a fine-tuned version of PekingU/rtdetr_r50vd_coco_o365 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 11.9166
• Map: 0.3323
• Map 50: 0.469
• Map 75: 0.3645
• Map Small: -1.0
• Map Medium: 0.3095
• Map Large: 0.4094
• Mar 1: 0.459
• Mar 10: 0.6541
• Mar 100: 0.7167
• Mar Small: -1.0
• Mar Medium: 0.5411
• Mar Large: 0.8481
• Map Radar: 0.3744
• Mar 100 Radar: 0.9049
• Map Ship management system: 0.4699
• Mar 100 Ship management system: 0.9591
• Map Ship management system (top): 0.4953
• Mar 100 Ship management system (top): 0.8538
• Map Ecdis: 0.3507
• Mar 100 Ecdis: 0.8893
• Map Visual observation: 0.2942
• Mar 100 Visual observation: 0.8507
• Map Ship management system (table top): 0.5411
• Mar 100 Ship management system (table top): 0.72
• Map Thruster control: 0.1788
• Mar 100 Thruster control: 0.4077
• Map Bow thruster: 0.156
• Mar 100 Bow thruster: 0.4034
• Map Me telegraph: 0.1302
• Mar 100 Me telegraph: 0.4615
• Classification Accuracy: 0.2282
• Classification Accuracy Ship management system: 0.2957
• Classification Accuracy Radar: 0.3297
• Classification Accuracy Visual observation: 0.2203
• Classification Accuracy Ship management system (table top): 0.0
• Classification Accuracy Thruster control: 0.0256
• Classification Accuracy Ship management system (top): 0.3173
• Classification Accuracy Ecdis: 0.1071
• Classification Accuracy Me telegraph: 0.1923
• Classification Accuracy Bow thruster: 0.069

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• num_epochs: 10

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 Radar	Mar 100 Radar	Map Ship management system	Mar 100 Ship management system	Map Ship management system (top)	Mar 100 Ship management system (top)	Map Ecdis	Mar 100 Ecdis	Map Visual observation	Mar 100 Visual observation	Map Ship management system (table top)	Mar 100 Ship management system (table top)	Map Thruster control	Mar 100 Thruster control	Map Bow thruster	Mar 100 Bow thruster	Map Me telegraph	Mar 100 Me telegraph	Classification Accuracy	Classification Accuracy Ship management system	Classification Accuracy Radar	Classification Accuracy Visual observation	Classification Accuracy Ship management system (table top)	Classification Accuracy Thruster control	Classification Accuracy Ship management system (top)	Classification Accuracy Ecdis	Classification Accuracy Me telegraph	Classification Accuracy Bow thruster
12.1253	1.0	596	9.0098	0.4593	0.5254	0.4993	-1.0	0.3057	0.5007	0.482	0.6769	0.6956	-1.0	0.5412	0.799	0.7974	0.9308	0.8298	0.9366	0.5269	0.8221	0.8698	0.9736	0.581	0.9381	0.1451	0.5571	0.3128	0.6923	0.0391	0.2172	0.0319	0.1923	0.2592	0.3226	0.3027	0.3252	0.0	0.1282	0.125	0.2857	0.0	0.0
8.3909	2.0	1192	9.0147	0.4982	0.62	0.5363	-1.0	0.3274	0.5545	0.5187	0.677	0.6981	-1.0	0.5581	0.8106	0.7253	0.9259	0.8928	0.957	0.7477	0.8904	0.759	0.9264	0.5413	0.8951	0.3766	0.6171	0.2168	0.5308	0.0941	0.2483	0.1301	0.2923	0.2485	0.2796	0.2378	0.2937	0.0	0.0256	0.4423	0.2	0.0	0.0345
7.7341	3.0	1788	9.4848	0.5194	0.646	0.5676	-1.0	0.4074	0.6135	0.5582	0.7028	0.7338	-1.0	0.581	0.8242	0.7274	0.933	0.8448	0.9559	0.6896	0.8981	0.7131	0.9379	0.3997	0.8909	0.4943	0.6257	0.4133	0.6385	0.2247	0.4207	0.168	0.3038	0.2408	0.2419	0.2054	0.2727	0.0286	0.1795	0.2981	0.2571	0.3077	0.1379
7.6104	4.0	2384	9.4468	0.5195	0.6623	0.5763	-1.0	0.4071	0.5821	0.5825	0.7279	0.7439	-1.0	0.5779	0.8443	0.6493	0.9232	0.7005	0.9532	0.7667	0.875	0.6255	0.9329	0.5905	0.9101	0.4918	0.7057	0.4777	0.6949	0.2204	0.3655	0.1532	0.3346	0.2447	0.1882	0.2378	0.3811	0.0	0.1026	0.2981	0.1643	0.1923	0.0345
7.0844	5.0	2980	9.9452	0.5275	0.6615	0.5679	-1.0	0.4377	0.632	0.5795	0.7384	0.7581	-1.0	0.6037	0.8954	0.7249	0.9405	0.7164	0.9457	0.5673	0.8663	0.7485	0.9436	0.6044	0.9392	0.604	0.7086	0.3434	0.6256	0.2514	0.4379	0.1874	0.4154	0.2097	0.0968	0.2703	0.3252	0.0	0.0513	0.2019	0.2071	0.0769	0.0345
6.6803	6.0	3576	10.6003	0.4635	0.5961	0.5036	-1.0	0.4156	0.5591	0.5808	0.7419	0.7748	-1.0	0.6358	0.8924	0.5136	0.947	0.6338	0.9629	0.69	0.8904	0.6112	0.9407	0.3836	0.9066	0.6059	0.8029	0.3949	0.6513	0.1625	0.4483	0.1762	0.4231	0.2388	0.1774	0.427	0.2133	0.0286	0.0256	0.4038	0.1786	0.1538	0.0
6.4981	7.0	4172	11.3455	0.3103	0.4335	0.3555	-1.0	0.333	0.3612	0.4548	0.6611	0.7139	-1.0	0.567	0.8517	0.3993	0.8973	0.4322	0.9489	0.4434	0.8394	0.3526	0.9057	0.2785	0.8479	0.4733	0.7143	0.1746	0.5385	0.0939	0.3138	0.1448	0.4192	0.2311	0.2581	0.3622	0.2028	0.0	0.0256	0.2788	0.1857	0.1923	0.1379
6.1177	8.0	4768	11.4110	0.3684	0.5032	0.4021	-1.0	0.3369	0.4837	0.4834	0.6829	0.7346	-1.0	0.5818	0.8732	0.4138	0.9162	0.5081	0.9651	0.5771	0.8721	0.4373	0.9307	0.3317	0.8734	0.5112	0.6886	0.2239	0.5128	0.1511	0.3793	0.1609	0.4731	0.2437	0.3172	0.3568	0.2378	0.0	0.0256	0.2212	0.2071	0.1538	0.0345
5.9352	9.0	5364	11.8779	0.3299	0.4592	0.3628	-1.0	0.3115	0.4131	0.4611	0.652	0.7139	-1.0	0.5542	0.8584	0.351	0.9086	0.4956	0.9613	0.4915	0.8462	0.3509	0.8936	0.326	0.8598	0.4875	0.6714	0.1978	0.4564	0.1452	0.3621	0.1234	0.4654	0.2262	0.328	0.3243	0.2168	0.0286	0.0256	0.2308	0.15	0.0769	0.0345
5.8941	10.0	5960	11.9166	0.3323	0.469	0.3645	-1.0	0.3095	0.4094	0.459	0.6541	0.7167	-1.0	0.5411	0.8481	0.3744	0.9049	0.4699	0.9591	0.4953	0.8538	0.3507	0.8893	0.2942	0.8507	0.5411	0.72	0.1788	0.4077	0.156	0.4034	0.1302	0.4615	0.2282	0.2957	0.3297	0.2203	0.0	0.0256	0.3173	0.1071	0.1923	0.069

Framework versions

• Transformers 4.46.0
• Pytorch 2.5.0+cu121
• Datasets 3.0.2
• Tokenizers 0.20.1