rtdetr-r50-bench-finetune

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: 15.1486
• Map: 0.217
• Map 50: 0.3852
• Map 75: 0.1895
• Map Small: 0.206
• Map Medium: 0.2572
• Map Large: 0.0515
• Mar 1: 0.2172
• Mar 10: 0.3591
• Mar 100: 0.3796
• Mar Small: 0.3184
• Mar Medium: 0.4616
• Mar Large: 0.05
• Map Elbows-shoulders-bar-athlete: -1.0
• Mar 100 Elbows-shoulders-bar-athlete: -1.0
• Map Barbell: 0.1248
• Mar 100 Barbell: 0.305
• Map Benchpresschest: -1.0
• Mar 100 Benchpresschest: -1.0
• Map Benchpressshoulder: 0.0
• Mar 100 Benchpressshoulder: 0.0
• Map Elbow: 0.01
• Mar 100 Elbow: 0.3238
• Map Hand: 0.3141
• Mar 100 Hand: 0.5248
• Map Head: 0.636
• Mar 100 Head: 0.7447

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 300
• 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 Elbows-shoulders-bar-athlete	Mar 100 Elbows-shoulders-bar-athlete	Map Barbell	Mar 100 Barbell	Map Benchpresschest	Mar 100 Benchpresschest	Map Benchpressshoulder	Mar 100 Benchpressshoulder	Map Elbow	Mar 100 Elbow	Map Hand	Mar 100 Hand	Map Head	Mar 100 Head
No log	1.0	37	58.1399	0.0001	0.0001	0.0001	0.0	0.0002	0.0	0.0021	0.0028	0.0044	0.002	0.0059	0.0	-1.0	-1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.009	0.0	0.005	0.0003	0.0123
No log	2.0	74	42.4180	0.0038	0.0056	0.0054	0.0001	0.0103	0.0	0.0062	0.0156	0.0156	0.0021	0.0218	0.0	-1.0	-1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0007	0.0256	0.022	0.0679
No log	3.0	111	34.4996	0.015	0.0217	0.0178	0.0002	0.0327	0.0	0.0245	0.0383	0.0385	0.0032	0.0552	0.0	-1.0	-1.0	0.0003	0.0043	0.0	0.0	0.0	0.0	0.0	0.0	0.0031	0.0525	0.0867	0.1741
No log	4.0	148	29.1792	0.0254	0.0435	0.0239	0.0083	0.0486	0.0014	0.0349	0.0626	0.0626	0.0245	0.0885	0.02	-1.0	-1.0	0.0004	0.0021	0.0	0.0	0.0	0.0	0.0	0.0	0.0587	0.2044	0.0933	0.1691
No log	5.0	185	22.1037	0.0867	0.1572	0.0786	0.06	0.1234	0.0	0.0902	0.1323	0.1324	0.1011	0.1681	0.0	-1.0	-1.0	0.0022	0.0209	0.0	0.0	0.0	0.0	0.0	0.0108	0.2576	0.4181	0.2605	0.3444
No log	6.0	222	18.5807	0.1367	0.2393	0.13	0.1254	0.1771	0.0027	0.1285	0.2144	0.2357	0.2282	0.3049	0.04	-1.0	-1.0	0.0143	0.1235	0.0	0.0	0.0	0.0	0.0005	0.1175	0.3634	0.5731	0.4419	0.6
No log	7.0	259	17.0227	0.1642	0.2798	0.1596	0.1853	0.2005	0.0153	0.1531	0.2538	0.2663	0.2964	0.3338	0.04	-1.0	-1.0	0.0456	0.1674	0.0	0.0	0.0	0.0	0.0014	0.1687	0.3894	0.5825	0.5486	0.679
No log	8.0	296	15.6265	0.1849	0.3282	0.1716	0.2236	0.2215	0.0208	0.1801	0.2874	0.2967	0.3501	0.3619	0.02	-1.0	-1.0	0.1076	0.2984	0.0	0.0	0.0	0.0	0.0017	0.197	0.4029	0.5838	0.5973	0.7012
No log	9.0	333	15.1906	0.1871	0.3393	0.166	0.2442	0.2201	0.0401	0.1826	0.3077	0.3425	0.4268	0.3931	0.1	-1.0	-1.0	0.1194	0.3396	0.0	0.0	0.002	0.05	0.0071	0.3747	0.3975	0.5638	0.5969	0.7272
No log	10.0	370	14.8474	0.1954	0.3502	0.1848	0.2464	0.2337	0.0208	0.1818	0.2982	0.311	0.3738	0.3728	0.02	-1.0	-1.0	0.1522	0.331	0.0	0.0	0.0	0.0	0.006	0.2627	0.4073	0.565	0.6066	0.7074

Framework versions

• Transformers 4.54.0.dev0
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2