rtdetr-v2-mobile-ui-design

This model is a fine-tuned version of PekingU/rtdetr_v2_r50vd on the mrtoy/mobile-ui-design dataset. It achieves the following results on the evaluation set:

• Loss: 13.2616
• Map: 0.1534
• Map 50: 0.2287
• Map 75: 0.1542
• Map Small: 0.0865
• Map Medium: 0.2047
• Map Large: 0.3091
• Mar 1: 0.0442
• Mar 10: 0.2487
• Mar 100: 0.4962
• Mar Small: 0.2773
• Mar Medium: 0.6015
• Mar Large: 0.7944
• Map Group: 0.1249
• Mar 100 Group: 0.5172
• Map Image: 0.1624
• Mar 100 Image: 0.5815
• Map Rectangle: 0.1702
• Mar 100 Rectangle: 0.5057
• Map Text: 0.1562
• Mar 100 Text: 0.3804

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: 4
• 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: linear
• num_epochs: 10.0
• mixed_precision_training: 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 Group	Mar 100 Group	Map Image	Mar 100 Image	Map Rectangle	Mar 100 Rectangle	Map Text	Mar 100 Text
23.6602	1.0	197	14.0858	0.1528	0.2352	0.1481	0.0714	0.2019	0.2547	0.0459	0.2684	0.4782	0.291	0.5631	0.7147	0.1112	0.4851	0.2013	0.5721	0.1527	0.4655	0.1462	0.3903
22.2633	2.0	394	13.1787	0.1775	0.2653	0.1765	0.0891	0.2367	0.3279	0.0463	0.278	0.4988	0.2983	0.5899	0.7731	0.1364	0.5176	0.2159	0.6039	0.1877	0.4924	0.1699	0.3814
19.8778	3.0	591	13.0808	0.1671	0.2524	0.167	0.0908	0.2164	0.333	0.0448	0.2647	0.5016	0.3044	0.5882	0.7739	0.1129	0.5227	0.1943	0.6018	0.1666	0.4806	0.1946	0.4012
21.2068	4.0	788	12.9039	0.1501	0.2269	0.1496	0.0871	0.187	0.3033	0.043	0.2516	0.511	0.3137	0.6031	0.7796	0.1201	0.5328	0.1566	0.6087	0.1586	0.5057	0.165	0.3967
18.9370	5.0	985	13.0320	0.1692	0.2553	0.1685	0.0971	0.2165	0.3226	0.0481	0.2676	0.5146	0.3192	0.6079	0.7821	0.1356	0.5473	0.1766	0.6029	0.1785	0.4924	0.186	0.4159
19.2146	6.0	1182	12.6991	0.1719	0.2573	0.1733	0.1023	0.2208	0.3276	0.0461	0.2644	0.5164	0.306	0.619	0.7906	0.1341	0.5402	0.1708	0.5917	0.1881	0.5082	0.1948	0.4254
21.4358	7.0	1379	12.6713	0.1763	0.2598	0.1799	0.1045	0.2083	0.3432	0.0488	0.2691	0.5189	0.3103	0.6238	0.7839	0.1361	0.5482	0.1853	0.5935	0.1894	0.5146	0.1943	0.4192
20.0092	8.0	1576	12.6275	0.1634	0.2389	0.1669	0.0984	0.2029	0.303	0.0441	0.2547	0.5137	0.2994	0.6257	0.7886	0.13	0.5436	0.1648	0.5903	0.1962	0.5182	0.1628	0.4027
19.2347	9.0	1773	12.6504	0.1548	0.2268	0.1593	0.0984	0.191	0.2961	0.0428	0.2417	0.5066	0.2907	0.6185	0.7896	0.1268	0.5411	0.1574	0.5981	0.1788	0.5111	0.1561	0.3762
19.8882	10.0	1970	12.6196	0.1715	0.2498	0.1767	0.1056	0.2118	0.3165	0.0469	0.2576	0.5165	0.3077	0.6219	0.79	0.136	0.5488	0.1838	0.595	0.1928	0.5142	0.1735	0.4079

Framework versions

• Transformers 5.3.0.dev0
• Pytorch 2.10.0+cu128
• Datasets 4.5.0
• Tokenizers 0.22.2