rtdetr-v2-r50-kitti2-finetune-2

This model is a fine-tuned version of PekingU/rtdetr_v2_r50vd on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 10.8052
• Map: 0.3423
• Map 50: 0.5623
• Map 75: 0.3606
• Map Small: 0.2161
• Map Medium: 0.3631
• Map Large: 0.4172
• Mar 1: 0.2926
• Mar 10: 0.5137
• Mar 100: 0.5886
• Mar Small: 0.4156
• Mar Medium: 0.5976
• Mar Large: 0.6841
• Map Car: 0.6052
• Mar 100 Car: 0.7464
• Map Pedestrian: 0.3483
• Mar 100 Pedestrian: 0.5172
• Map Cyclist: 0.2062
• Mar 100 Cyclist: 0.4523
• Map Van: 0.5231
• Mar 100 Van: 0.7377
• Map Truck: 0.6026
• Mar 100 Truck: 0.7417
• Map Misc: 0.1678
• Mar 100 Misc: 0.516
• Map Tram: 0.3851
• Mar 100 Tram: 0.6984
• Map Person Sitting: 0.1978
• Mar 100 Person Sitting: 0.5314
• Map Dontcare: 0.0442
• Mar 100 Dontcare: 0.3566

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 Car	Mar 100 Car	Map Pedestrian	Mar 100 Pedestrian	Map Cyclist	Mar 100 Cyclist	Map Van	Mar 100 Van	Map Truck	Mar 100 Truck	Map Misc	Mar 100 Misc	Map Tram	Mar 100 Tram	Map Person Sitting	Mar 100 Person Sitting	Map Dontcare	Mar 100 Dontcare
34.3989	1.0	655	10.4608	0.2538	0.4309	0.2527	0.1757	0.2643	0.3546	0.2566	0.4571	0.5322	0.3842	0.5368	0.6569	0.6141	0.743	0.2816	0.4418	0.1456	0.4579	0.4092	0.6612	0.4751	0.7195	0.0566	0.4101	0.133	0.6307	0.1334	0.4357	0.0356	0.2902
15.6516	2.0	1310	10.8110	0.3342	0.5511	0.3468	0.2589	0.3513	0.4378	0.2863	0.5125	0.5775	0.4382	0.5818	0.7005	0.6159	0.7491	0.3197	0.4894	0.249	0.4594	0.4919	0.6995	0.6118	0.7552	0.1736	0.502	0.3949	0.7102	0.1077	0.4786	0.0433	0.354
14.1316	3.0	1965	11.9174	0.3267	0.5128	0.3591	0.2156	0.3255	0.4735	0.2828	0.5065	0.5748	0.4402	0.5811	0.7018	0.5899	0.7322	0.3121	0.4754	0.2163	0.4766	0.4165	0.6865	0.5878	0.7526	0.1879	0.5507	0.429	0.6807	0.1537	0.45	0.0474	0.3687
12.9222	4.0	2620	12.5913	0.3124	0.4931	0.3376	0.2457	0.3275	0.46	0.279	0.4928	0.5516	0.4054	0.5606	0.688	0.5478	0.7176	0.2902	0.4381	0.13	0.413	0.4401	0.6785	0.5964	0.7468	0.2421	0.5601	0.4089	0.7057	0.108	0.3429	0.0482	0.3613
12.6036	5.0	3275	13.0327	0.3117	0.4742	0.3476	0.2483	0.3282	0.4539	0.2901	0.4865	0.5418	0.4431	0.5536	0.6582	0.5549	0.7381	0.292	0.4435	0.1428	0.3874	0.4093	0.6968	0.5729	0.7494	0.2757	0.6088	0.4991	0.6989	0.0056	0.1714	0.0528	0.3825
12.2062	6.0	3930	13.2957	0.3151	0.4824	0.3456	0.2292	0.3166	0.465	0.2908	0.4839	0.5362	0.4192	0.5437	0.663	0.5304	0.7357	0.2899	0.4352	0.1494	0.3556	0.4531	0.6872	0.5778	0.7442	0.3199	0.598	0.4628	0.7034	0.004	0.1714	0.0491	0.3949
11.847	7.0	4585	13.1970	0.309	0.4753	0.3435	0.2333	0.3199	0.4422	0.2983	0.4868	0.5398	0.4161	0.5544	0.6635	0.5359	0.7321	0.2909	0.4272	0.099	0.359	0.4248	0.6806	0.5835	0.7247	0.2946	0.6203	0.4941	0.7102	0.0087	0.2214	0.0495	0.3826

Framework versions

• Transformers 4.50.0.dev0
• Pytorch 2.6.0+cu124
• Datasets 3.4.0
• Tokenizers 0.21.1