rtdetr-v2-r50-finetune-24

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: 6.3704
Map: 0.5366
Map 50: 0.8421
Map 75: 0.5952
Map Small: 0.4987
Map Medium: 0.6605
Map Large: -1.0
Mar 1: 0.332
Mar 10: 0.6793
Mar 100: 0.7019
Mar Small: 0.6743
Mar Medium: 0.7724
Mar Large: -1.0
Map Artemia: 0.5366
Mar 100 Artemia: 0.7019

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_FUSED 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: 50

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 Artemia	Mar 100 Artemia
No log	1.0	250	14.3520	0.3421	0.6431	0.3277	0.2616	0.491	-1.0	0.3174	0.5935	0.6623	0.5866	0.7657	-1.0	0.3421	0.6623
97.4404	2.0	500	9.6456	0.2961	0.5647	0.2679	0.2228	0.4804	-1.0	0.2953	0.5813	0.6174	0.5505	0.7102	-1.0	0.2961	0.6174
97.4404	3.0	750	8.2470	0.4845	0.8607	0.4961	0.4107	0.601	-1.0	0.3832	0.5981	0.6396	0.5919	0.7051	-1.0	0.4845	0.6396
13.9935	4.0	1000	7.9928	0.4773	0.8811	0.4594	0.3969	0.5948	-1.0	0.3692	0.5928	0.6187	0.5602	0.6993	-1.0	0.4773	0.6187
13.9935	5.0	1250	8.6240	0.4242	0.8189	0.3675	0.3434	0.5756	-1.0	0.3579	0.5779	0.6072	0.5591	0.673	-1.0	0.4242	0.6072
12.3453	6.0	1500	8.0713	0.4463	0.8232	0.4059	0.3616	0.5964	-1.0	0.3657	0.5713	0.5966	0.5371	0.6788	-1.0	0.4463	0.5966
12.3453	7.0	1750	8.3215	0.4318	0.8284	0.395	0.3571	0.5539	-1.0	0.3564	0.5617	0.5676	0.5081	0.6504	-1.0	0.4318	0.5676
11.4638	8.0	2000	8.6466	0.3923	0.787	0.3812	0.321	0.5649	-1.0	0.3324	0.5717	0.5879	0.5462	0.6453	-1.0	0.3923	0.5879
11.4638	9.0	2250	8.3070	0.4064	0.7746	0.3652	0.334	0.5886	-1.0	0.324	0.5907	0.5944	0.5387	0.6715	-1.0	0.4064	0.5944
10.5616	10.0	2500	8.5226	0.3911	0.7223	0.3539	0.3037	0.5957	-1.0	0.3262	0.5844	0.586	0.5226	0.6737	-1.0	0.3911	0.586
10.5616	11.0	2750	8.9261	0.3341	0.659	0.2884	0.2418	0.5799	-1.0	0.2941	0.5623	0.5745	0.5027	0.6737	-1.0	0.3341	0.5745
10.0255	12.0	3000	9.0511	0.3029	0.6061	0.2466	0.2193	0.5635	-1.0	0.2682	0.585	0.5988	0.5306	0.692	-1.0	0.3029	0.5988
10.0255	13.0	3250	8.6556	0.3675	0.6978	0.339	0.2725	0.5908	-1.0	0.3174	0.5782	0.5857	0.5172	0.6788	-1.0	0.3675	0.5857
9.3948	14.0	3500	8.9080	0.305	0.5841	0.2718	0.2052	0.576	-1.0	0.2639	0.5598	0.5648	0.4973	0.6569	-1.0	0.305	0.5648
9.3948	15.0	3750	9.1003	0.3551	0.6696	0.3237	0.2502	0.5841	-1.0	0.3199	0.5601	0.5648	0.4903	0.6679	-1.0	0.3551	0.5648
8.8491	16.0	4000	9.1328	0.3445	0.6649	0.2754	0.2476	0.554	-1.0	0.3028	0.5502	0.5548	0.4758	0.6642	-1.0	0.3445	0.5548
8.8491	17.0	4250	9.0727	0.3145	0.5933	0.2871	0.2229	0.5796	-1.0	0.2782	0.5748	0.5766	0.5199	0.654	-1.0	0.3145	0.5766
8.3038	18.0	4500	9.2502	0.3018	0.5752	0.2823	0.2021	0.5943	-1.0	0.2685	0.5773	0.5807	0.5161	0.6686	-1.0	0.3018	0.5807
8.3038	19.0	4750	9.0829	0.3448	0.6367	0.3228	0.2388	0.5793	-1.0	0.2773	0.5704	0.5745	0.5016	0.6737	-1.0	0.3448	0.5745
7.9314	20.0	5000	9.9463	0.3154	0.6031	0.2762	0.2177	0.5875	-1.0	0.2785	0.5773	0.5788	0.5108	0.6723	-1.0	0.3154	0.5788
7.9314	21.0	5250	9.5059	0.3692	0.7005	0.3349	0.2694	0.5698	-1.0	0.3103	0.5495	0.5508	0.4812	0.646	-1.0	0.3692	0.5508
7.5983	22.0	5500	10.3650	0.3107	0.5847	0.2924	0.2089	0.5933	-1.0	0.2713	0.5626	0.5629	0.4941	0.6569	-1.0	0.3107	0.5629
7.5983	23.0	5750	10.1495	0.292	0.5475	0.2763	0.1934	0.5902	-1.0	0.2343	0.5838	0.5872	0.5274	0.6693	-1.0	0.292	0.5872
7.1417	24.0	6000	10.8072	0.3141	0.5968	0.2926	0.2122	0.582	-1.0	0.2626	0.5533	0.5536	0.4796	0.6555	-1.0	0.3141	0.5536
7.1417	25.0	6250	10.3555	0.3492	0.6701	0.3148	0.2424	0.5852	-1.0	0.2966	0.5502	0.5505	0.4667	0.6657	-1.0	0.3492	0.5505
6.9403	26.0	6500	10.1237	0.3132	0.5876	0.2859	0.2145	0.5776	-1.0	0.2629	0.5583	0.5586	0.4823	0.6642	-1.0	0.3132	0.5586
6.9403	27.0	6750	10.4347	0.336	0.6287	0.2956	0.2298	0.5859	-1.0	0.2885	0.5595	0.5595	0.4882	0.6584	-1.0	0.336	0.5595
6.6409	28.0	7000	10.4675	0.3264	0.6213	0.3087	0.2198	0.5855	-1.0	0.276	0.5626	0.5645	0.4946	0.6599	-1.0	0.3264	0.5645
6.6409	29.0	7250	10.4921	0.3174	0.5963	0.2764	0.2165	0.5764	-1.0	0.2698	0.5592	0.5592	0.4839	0.6628	-1.0	0.3174	0.5592
6.4083	30.0	7500	10.3468	0.3343	0.6349	0.311	0.2314	0.5757	-1.0	0.2791	0.5548	0.5555	0.4866	0.6504	-1.0	0.3343	0.5555
6.4083	31.0	7750	10.5682	0.317	0.6004	0.2861	0.2102	0.5828	-1.0	0.2754	0.5539	0.5539	0.4731	0.6657	-1.0	0.317	0.5539
6.0784	32.0	8000	10.5964	0.3468	0.6447	0.3035	0.2384	0.587	-1.0	0.3081	0.553	0.553	0.471	0.6657	-1.0	0.3468	0.553
6.0784	33.0	8250	11.0811	0.3106	0.5845	0.2963	0.2063	0.5909	-1.0	0.2701	0.5611	0.5614	0.4887	0.6613	-1.0	0.3106	0.5614
5.8548	34.0	8500	10.7206	0.3269	0.6212	0.2993	0.2205	0.5949	-1.0	0.2754	0.5567	0.5567	0.4801	0.6628	-1.0	0.3269	0.5567
5.8548	35.0	8750	11.1723	0.3078	0.5814	0.2873	0.2031	0.5907	-1.0	0.2651	0.5626	0.5626	0.4887	0.665	-1.0	0.3078	0.5626
5.7109	36.0	9000	11.6025	0.3059	0.5755	0.2711	0.2047	0.5853	-1.0	0.2735	0.5573	0.5573	0.4823	0.6606	-1.0	0.3059	0.5573
5.7109	37.0	9250	11.4261	0.3122	0.5878	0.2866	0.2049	0.5847	-1.0	0.2657	0.5598	0.5598	0.4828	0.6657	-1.0	0.3122	0.5598
5.5113	38.0	9500	10.8001	0.3484	0.6514	0.3222	0.2429	0.5928	-1.0	0.2907	0.5548	0.5548	0.4753	0.665	-1.0	0.3484	0.5548
5.5113	39.0	9750	11.4681	0.3174	0.5966	0.2837	0.2109	0.5915	-1.0	0.2698	0.5561	0.5564	0.478	0.665	-1.0	0.3174	0.5564
5.2335	40.0	10000	11.2990	0.3243	0.6081	0.3028	0.2173	0.5859	-1.0	0.2735	0.5558	0.5558	0.4801	0.6606	-1.0	0.3243	0.5558
5.2335	41.0	10250	12.1316	0.3148	0.5914	0.2874	0.2053	0.591	-1.0	0.2648	0.5539	0.5548	0.4715	0.6701	-1.0	0.3148	0.5548
5.0265	42.0	10500	11.8177	0.3237	0.6051	0.2939	0.2174	0.5892	-1.0	0.2798	0.5526	0.5526	0.4731	0.6628	-1.0	0.3237	0.5526
5.0265	43.0	10750	11.4751	0.3358	0.6295	0.3085	0.2284	0.586	-1.0	0.2841	0.5511	0.5511	0.4737	0.6584	-1.0	0.3358	0.5511
4.8217	44.0	11000	11.7414	0.333	0.6277	0.3011	0.2256	0.592	-1.0	0.2788	0.5545	0.5545	0.479	0.6591	-1.0	0.333	0.5545
4.8217	45.0	11250	12.0233	0.3209	0.598	0.2867	0.2136	0.5899	-1.0	0.2735	0.552	0.552	0.4747	0.6591	-1.0	0.3209	0.552
4.6744	46.0	11500	11.9372	0.3279	0.6121	0.2906	0.2187	0.5897	-1.0	0.2854	0.5511	0.5511	0.4704	0.6628	-1.0	0.3279	0.5511
4.6744	47.0	11750	12.5357	0.3254	0.6046	0.2946	0.2145	0.592	-1.0	0.2757	0.5523	0.5523	0.4715	0.6642	-1.0	0.3254	0.5523
4.441	48.0	12000	12.4076	0.329	0.6182	0.3051	0.2201	0.592	-1.0	0.2779	0.5555	0.5555	0.478	0.6628	-1.0	0.329	0.5555
4.441	49.0	12250	12.2815	0.3317	0.6189	0.3017	0.2255	0.589	-1.0	0.29	0.5542	0.5542	0.4763	0.662	-1.0	0.3317	0.5542
4.3324	50.0	12500	12.2671	0.3332	0.6206	0.3026	0.2267	0.5856	-1.0	0.2891	0.553	0.553	0.4769	0.6584	-1.0	0.3332	0.553

Framework versions

Transformers 4.57.1
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.2

dariacuna/rtdetr-v2-r50-finetune-24

rtdetr-v2-r50-finetune-24

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Seed this model