suas-2025-rtdetr-finetuned-b16-lr1e-4

This model is a fine-tuned version of PekingU/rtdetr_r50vd on the mfly-auton/suas-2025-synthetic-data dataset. It achieves the following results on the evaluation set:

• Loss: 7.4789
• Map: 0.636
• Map 50: 0.7833
• Map 75: 0.6825
• Map Small: 0.5292
• Map Medium: 0.6761
• Map Large: 0.6243
• Mar 1: 0.6914
• Mar 10: 0.8319
• Mar 100: 0.8459
• Mar Small: 0.6526
• Mar Medium: 0.861
• Mar Large: 0.9511
• Map Baseball-bat: 0.7327
• Mar 100 Baseball-bat: 0.8479
• Map Basketball: 0.7378
• Mar 100 Basketball: 0.8453
• Map Car: -1.0
• Mar 100 Car: -1.0
• Map Football: 0.515
• Mar 100 Football: 0.5874
• Map Human: 0.2845
• Mar 100 Human: 0.6493
• Map Luggage: 0.3927
• Mar 100 Luggage: 0.9199
• Map Mattress: 0.4605
• Mar 100 Mattress: 0.9731
• Map Motorcycle: 0.8921
• Mar 100 Motorcycle: 0.9379
• Map Skis: 0.9356
• Mar 100 Skis: 0.9772
• Map Snowboard: 0.4858
• Mar 100 Snowboard: 0.9648
• Map Soccer-ball: 0.291
• Mar 100 Soccer-ball: 0.5365
• Map Stop-sign: 0.8721
• Mar 100 Stop-sign: 0.9777
• Map Tennis-racket: 0.8872
• Mar 100 Tennis-racket: 0.9275
• Map Umbrella: 0.7926
• Mar 100 Umbrella: 0.945
• Map Volleyball: 0.6243
• Mar 100 Volleyball: 0.7525

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: 16
• eval_batch_size: 16
• seed: 1337
• optimizer: Use adamw_torch 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 Baseball-bat	Mar 100 Baseball-bat	Map Basketball	Mar 100 Basketball	Map Car	Mar 100 Car	Map Football	Mar 100 Football	Map Human	Mar 100 Human	Map Luggage	Mar 100 Luggage	Map Mattress	Mar 100 Mattress	Map Motorcycle	Mar 100 Motorcycle	Map Skis	Mar 100 Skis	Map Snowboard	Mar 100 Snowboard	Map Soccer-ball	Mar 100 Soccer-ball	Map Stop-sign	Mar 100 Stop-sign	Map Tennis-racket	Mar 100 Tennis-racket	Map Umbrella	Mar 100 Umbrella	Map Volleyball	Mar 100 Volleyball
14.9473	1.0	438	6.0393	0.6929	0.8006	0.7826	0.6924	0.6693	0.7428	0.7205	0.8507	0.8584	0.7888	0.8574	0.8979	0.7927	0.8492	0.8387	0.9263	-1.0	-1.0	0.5419	0.8203	0.2032	0.5136	0.4715	0.931	0.6218	0.9887	0.866	0.9199	0.8182	0.8782	0.5796	0.9515	0.6169	0.6753	0.9283	0.9732	0.814	0.8668	0.8116	0.8802	0.7961	0.8431
6.8045	2.0	876	5.2885	0.7468	0.8307	0.8234	0.7611	0.7249	0.6791	0.7565	0.8986	0.9052	0.8475	0.877	0.9424	0.8403	0.8959	0.8533	0.9358	-1.0	-1.0	0.7497	0.8638	0.1319	0.532	0.4374	0.9456	0.5892	0.9958	0.9306	0.9599	0.9652	0.9802	0.6246	0.995	0.7541	0.8175	0.9788	0.9958	0.8972	0.9316	0.9091	0.9701	0.7936	0.8532
6.122	3.0	1314	6.0592	0.7128	0.8135	0.7987	0.6996	0.7187	0.665	0.739	0.8748	0.8891	0.8104	0.8731	0.9347	0.8141	0.8864	0.8214	0.9197	-1.0	-1.0	0.7119	0.8346	0.1712	0.5716	0.3367	0.9129	0.4596	0.9784	0.9166	0.9472	0.9445	0.9718	0.6822	0.9922	0.6055	0.7355	0.9341	0.9659	0.8819	0.9207	0.9083	0.9698	0.7906	0.8409
5.7542	4.0	1752	6.7229	0.7089	0.8239	0.7984	0.6721	0.7153	0.6452	0.752	0.8732	0.8817	0.7591	0.8769	0.9587	0.7857	0.8574	0.835	0.9052	-1.0	-1.0	0.6519	0.7566	0.3228	0.6242	0.5068	0.9147	0.4086	0.9842	0.9254	0.9588	0.9363	0.9639	0.6089	0.9904	0.6176	0.7185	0.8906	0.9765	0.8719	0.9197	0.8104	0.9656	0.7532	0.8083
5.4197	5.0	2190	6.9705	0.6705	0.7896	0.7393	0.6309	0.6775	0.6347	0.7021	0.8427	0.8567	0.7156	0.8715	0.9375	0.8044	0.874	0.8447	0.8896	-1.0	-1.0	0.6092	0.6991	0.1786	0.5444	0.3152	0.8589	0.3916	0.946	0.943	0.9708	0.957	0.9782	0.521	0.9759	0.4336	0.5923	0.8949	0.9757	0.8747	0.9171	0.8842	0.9704	0.7354	0.8017
5.1598	6.0	2628	6.7201	0.6509	0.7806	0.7067	0.5621	0.6931	0.6332	0.6927	0.826	0.839	0.6904	0.8633	0.8993	0.7365	0.8412	0.8438	0.9021	-1.0	-1.0	0.5378	0.639	0.1533	0.4137	0.412	0.91	0.3546	0.9008	0.9302	0.967	0.9724	0.9881	0.5606	0.9739	0.3666	0.5918	0.8779	0.9642	0.843	0.8969	0.8855	0.9633	0.6379	0.7946
5.1626	7.0	3066	7.2974	0.644	0.7767	0.6936	0.5262	0.6787	0.6513	0.7043	0.8445	0.8587	0.6746	0.8942	0.9601	0.7639	0.8663	0.7068	0.8149	-1.0	-1.0	0.5452	0.6602	0.3086	0.6825	0.41	0.9299	0.4634	0.9747	0.9327	0.9679	0.9505	0.9797	0.4214	0.9677	0.2936	0.5483	0.852	0.9818	0.8823	0.9181	0.8924	0.9678	0.5933	0.7615
5.15	8.0	3504	7.2201	0.6572	0.7908	0.7113	0.5559	0.6962	0.6276	0.7048	0.8396	0.8534	0.6795	0.8702	0.9327	0.7574	0.8526	0.7803	0.8718	-1.0	-1.0	0.5741	0.6547	0.2742	0.5961	0.418	0.9233	0.4893	0.9697	0.91	0.9479	0.9345	0.9807	0.5642	0.9692	0.3294	0.5602	0.8803	0.9821	0.8794	0.9264	0.7936	0.9476	0.616	0.7659
4.9521	9.0	3942	7.5060	0.6374	0.7847	0.6838	0.5132	0.6921	0.6331	0.6868	0.8276	0.8419	0.6487	0.8619	0.9341	0.7388	0.8431	0.715	0.8273	-1.0	-1.0	0.5194	0.592	0.2577	0.624	0.3945	0.9234	0.4599	0.966	0.9031	0.942	0.9235	0.9752	0.5532	0.9694	0.2822	0.5368	0.9131	0.9855	0.8769	0.9244	0.8091	0.9438	0.5766	0.7336
4.8252	10.0	4380	7.4789	0.636	0.7833	0.6825	0.5292	0.6761	0.6243	0.6914	0.8319	0.8459	0.6526	0.861	0.9511	0.7327	0.8479	0.7378	0.8453	-1.0	-1.0	0.515	0.5874	0.2845	0.6493	0.3927	0.9199	0.4605	0.9731	0.8921	0.9379	0.9356	0.9772	0.4858	0.9648	0.291	0.5365	0.8721	0.9777	0.8872	0.9275	0.7926	0.945	0.6243	0.7525

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.21.0