rtdetr-v2-r50-finetune-18

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: 6.2140
Map: 0.5792
Map 50: 0.9062
Map 75: 0.6927
Map Small: 0.5499
Map Medium: 0.6586
Map Large: -1.0
Mar 1: 0.3485
Mar 10: 0.665
Mar 100: 0.7061
Mar Small: 0.6766
Mar Medium: 0.7816
Mar Large: -1.0
Map Artemia: 0.5792
Mar 100 Artemia: 0.7061

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: 70

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	14.9156	0.2677	0.5009	0.2541	0.1906	0.4855	-1.0	0.3037	0.605	0.6642	0.5866	0.7708	-1.0	0.2677	0.6642
177.9373	2.0	500	8.5364	0.4833	0.8728	0.4873	0.4202	0.5904	-1.0	0.3688	0.59	0.6455	0.5876	0.7255	-1.0	0.4833	0.6455
177.9373	3.0	750	7.9292	0.508	0.8772	0.5383	0.4318	0.6172	-1.0	0.3879	0.5947	0.6498	0.5925	0.7285	-1.0	0.508	0.6498
13.8839	4.0	1000	8.0375	0.4786	0.8568	0.4833	0.3907	0.5981	-1.0	0.3788	0.586	0.6315	0.5699	0.7161	-1.0	0.4786	0.6315
13.8839	5.0	1250	8.6785	0.472	0.8573	0.4344	0.3884	0.5955	-1.0	0.3657	0.5869	0.6252	0.578	0.6898	-1.0	0.472	0.6252
12.3143	6.0	1500	8.1100	0.4655	0.8508	0.4471	0.3825	0.5996	-1.0	0.3735	0.5984	0.6371	0.5817	0.7139	-1.0	0.4655	0.6371
12.3143	7.0	1750	8.0911	0.4592	0.8707	0.3957	0.387	0.5737	-1.0	0.3639	0.5766	0.6019	0.5398	0.6869	-1.0	0.4592	0.6019
11.5396	8.0	2000	8.1778	0.4418	0.8254	0.4287	0.3621	0.5876	-1.0	0.3636	0.5785	0.6181	0.5645	0.692	-1.0	0.4418	0.6181
11.5396	9.0	2250	8.4973	0.4392	0.8118	0.4342	0.3567	0.5837	-1.0	0.3533	0.5779	0.6003	0.5478	0.6723	-1.0	0.4392	0.6003
10.7203	10.0	2500	8.5260	0.4341	0.8037	0.4154	0.3393	0.5951	-1.0	0.3636	0.5807	0.5978	0.5387	0.6796	-1.0	0.4341	0.5978
10.7203	11.0	2750	8.5815	0.435	0.8195	0.38	0.3556	0.5834	-1.0	0.3511	0.5769	0.5988	0.5473	0.6701	-1.0	0.435	0.5988
10.1351	12.0	3000	8.7181	0.4211	0.7924	0.3826	0.3365	0.5697	-1.0	0.3424	0.5595	0.571	0.486	0.6876	-1.0	0.4211	0.571
10.1351	13.0	3250	8.8507	0.3734	0.7101	0.3307	0.2948	0.5799	-1.0	0.324	0.5794	0.5944	0.528	0.6854	-1.0	0.3734	0.5944
9.5146	14.0	3500	8.7414	0.3723	0.713	0.3318	0.2899	0.5869	-1.0	0.3097	0.5769	0.585	0.5167	0.6781	-1.0	0.3723	0.585
9.5146	15.0	3750	9.1167	0.3965	0.7505	0.3816	0.3087	0.5787	-1.0	0.3262	0.5483	0.5502	0.4758	0.6518	-1.0	0.3965	0.5502
8.9041	16.0	4000	9.0343	0.4	0.7501	0.3696	0.3079	0.573	-1.0	0.3336	0.5467	0.548	0.4581	0.6715	-1.0	0.4	0.548
8.9041	17.0	4250	9.0600	0.3947	0.7454	0.3669	0.3066	0.5811	-1.0	0.3243	0.5583	0.5607	0.4876	0.6613	-1.0	0.3947	0.5607
8.5156	18.0	4500	9.2045	0.4059	0.7495	0.4026	0.3113	0.5856	-1.0	0.3268	0.5564	0.5573	0.479	0.665	-1.0	0.4059	0.5573
8.5156	19.0	4750	9.1470	0.379	0.7153	0.3435	0.2844	0.5811	-1.0	0.3156	0.5514	0.5523	0.4742	0.6599	-1.0	0.379	0.5523
8.1768	20.0	5000	9.3248	0.3669	0.6835	0.3373	0.2644	0.5717	-1.0	0.3153	0.5467	0.5474	0.4672	0.6569	-1.0	0.3669	0.5474
8.1768	21.0	5250	9.4792	0.3765	0.7227	0.3466	0.286	0.5684	-1.0	0.3196	0.5433	0.5452	0.4704	0.6482	-1.0	0.3765	0.5452
7.869	22.0	5500	9.2443	0.3771	0.7192	0.373	0.2872	0.5745	-1.0	0.314	0.5514	0.5523	0.4828	0.6482	-1.0	0.3771	0.5523
7.869	23.0	5750	9.9580	0.373	0.6973	0.3408	0.2805	0.5707	-1.0	0.3118	0.5442	0.5467	0.4726	0.6496	-1.0	0.373	0.5467
7.4736	24.0	6000	9.5836	0.3747	0.7054	0.3515	0.2829	0.5793	-1.0	0.3181	0.5486	0.5498	0.472	0.6569	-1.0	0.3747	0.5498
7.4736	25.0	6250	9.4607	0.3773	0.7019	0.3469	0.2649	0.5879	-1.0	0.3037	0.5411	0.5411	0.4505	0.6664	-1.0	0.3773	0.5411
7.3698	26.0	6500	9.6946	0.3871	0.7305	0.3386	0.2855	0.565	-1.0	0.3199	0.5336	0.5336	0.4554	0.6416	-1.0	0.3871	0.5336
7.3698	27.0	6750	9.4733	0.383	0.7296	0.3394	0.2855	0.5721	-1.0	0.3159	0.5502	0.5539	0.4817	0.6533	-1.0	0.383	0.5539
7.0649	28.0	7000	10.1288	0.3579	0.6647	0.3629	0.2498	0.5743	-1.0	0.295	0.5492	0.5517	0.4763	0.6555	-1.0	0.3579	0.5517
7.0649	29.0	7250	9.9305	0.3669	0.6803	0.3514	0.2578	0.5856	-1.0	0.3065	0.5477	0.5511	0.4667	0.6664	-1.0	0.3669	0.5511
6.7768	30.0	7500	10.2239	0.3711	0.6831	0.3507	0.2691	0.5797	-1.0	0.3109	0.5542	0.5542	0.4753	0.6628	-1.0	0.3711	0.5542
6.7768	31.0	7750	10.1363	0.3513	0.6476	0.3266	0.2389	0.5771	-1.0	0.2838	0.5417	0.5417	0.4565	0.6599	-1.0	0.3513	0.5417
6.6165	32.0	8000	9.8693	0.3941	0.7303	0.3711	0.288	0.577	-1.0	0.3234	0.5377	0.5377	0.4565	0.6504	-1.0	0.3941	0.5377
6.6165	33.0	8250	9.9937	0.3664	0.6824	0.3472	0.257	0.5777	-1.0	0.3044	0.5433	0.5433	0.4586	0.6599	-1.0	0.3664	0.5433
6.4464	34.0	8500	9.9507	0.3671	0.6839	0.3368	0.2573	0.5759	-1.0	0.3031	0.5386	0.5411	0.4581	0.6555	-1.0	0.3671	0.5411
6.4464	35.0	8750	10.4605	0.3613	0.6771	0.3363	0.2561	0.5822	-1.0	0.3087	0.5495	0.5495	0.4688	0.6606	-1.0	0.3613	0.5495
6.3113	36.0	9000	10.0878	0.3733	0.7043	0.3271	0.2743	0.5684	-1.0	0.3093	0.5464	0.5464	0.4731	0.6482	-1.0	0.3733	0.5464
6.3113	37.0	9250	10.8871	0.3571	0.666	0.3277	0.2507	0.5785	-1.0	0.3034	0.5455	0.5455	0.4683	0.6518	-1.0	0.3571	0.5455
6.143	38.0	9500	10.3063	0.3454	0.6421	0.325	0.2368	0.5728	-1.0	0.296	0.5439	0.5439	0.4618	0.6569	-1.0	0.3454	0.5439
6.143	39.0	9750	11.3595	0.3221	0.6059	0.2863	0.2155	0.5834	-1.0	0.2822	0.553	0.553	0.4747	0.6613	-1.0	0.3221	0.553
5.8745	40.0	10000	10.6999	0.3427	0.6347	0.3284	0.2335	0.5689	-1.0	0.2913	0.5389	0.5389	0.4548	0.6547	-1.0	0.3427	0.5389
5.8745	41.0	10250	10.4883	0.3597	0.6529	0.3545	0.2451	0.5938	-1.0	0.2972	0.5458	0.5458	0.4586	0.6657	-1.0	0.3597	0.5458
5.7384	42.0	10500	10.6574	0.3482	0.6599	0.3102	0.2397	0.5815	-1.0	0.2919	0.5526	0.5526	0.4737	0.6613	-1.0	0.3482	0.5526
5.7384	43.0	10750	11.3304	0.3463	0.6459	0.3145	0.2432	0.5757	-1.0	0.2944	0.5526	0.5526	0.4763	0.6577	-1.0	0.3463	0.5526
5.6412	44.0	11000	10.8019	0.3623	0.6627	0.337	0.2529	0.5941	-1.0	0.3137	0.5567	0.5567	0.4753	0.6686	-1.0	0.3623	0.5567
5.6412	45.0	11250	10.6010	0.3668	0.6727	0.3474	0.2598	0.5824	-1.0	0.3025	0.5489	0.5489	0.4688	0.6591	-1.0	0.3668	0.5489
5.515	46.0	11500	10.9244	0.3652	0.6679	0.3505	0.2642	0.5695	-1.0	0.3028	0.5483	0.5483	0.4737	0.6511	-1.0	0.3652	0.5483
5.515	47.0	11750	11.4745	0.3431	0.6425	0.3148	0.2383	0.5755	-1.0	0.3025	0.5514	0.5514	0.4769	0.654	-1.0	0.3431	0.5514
5.3244	48.0	12000	10.8319	0.3676	0.691	0.3341	0.263	0.5834	-1.0	0.3003	0.5474	0.5474	0.471	0.6526	-1.0	0.3676	0.5474
5.3244	49.0	12250	11.1756	0.3419	0.6409	0.3019	0.2373	0.57	-1.0	0.2919	0.5542	0.5542	0.4801	0.6569	-1.0	0.3419	0.5542
5.1944	50.0	12500	11.2864	0.3467	0.6446	0.3348	0.2395	0.5813	-1.0	0.3056	0.5502	0.5502	0.4742	0.6555	-1.0	0.3467	0.5502
5.1944	51.0	12750	11.4134	0.3627	0.6767	0.345	0.2519	0.5912	-1.0	0.3031	0.5576	0.5576	0.4785	0.6664	-1.0	0.3627	0.5576
5.0184	52.0	13000	11.6002	0.3394	0.6226	0.3178	0.2315	0.585	-1.0	0.2935	0.5567	0.5567	0.479	0.6635	-1.0	0.3394	0.5567
5.0184	53.0	13250	11.6213	0.3542	0.6505	0.3245	0.2454	0.5758	-1.0	0.2907	0.5486	0.5486	0.4694	0.6577	-1.0	0.3542	0.5486
4.9757	54.0	13500	11.9386	0.3432	0.6245	0.3212	0.2352	0.576	-1.0	0.2988	0.557	0.557	0.4763	0.6679	-1.0	0.3432	0.557
4.9757	55.0	13750	11.3950	0.3514	0.6459	0.3359	0.2478	0.5646	-1.0	0.3065	0.5477	0.5477	0.4699	0.6547	-1.0	0.3514	0.5477
4.748	56.0	14000	11.5925	0.3527	0.6564	0.3138	0.2468	0.5735	-1.0	0.3047	0.5492	0.5492	0.4715	0.6562	-1.0	0.3527	0.5492
4.748	57.0	14250	11.5311	0.3521	0.6495	0.3347	0.2446	0.582	-1.0	0.3053	0.5539	0.5539	0.4731	0.665	-1.0	0.3521	0.5539
4.7467	58.0	14500	11.1331	0.3681	0.6761	0.3491	0.2632	0.5806	-1.0	0.3125	0.5555	0.5555	0.4812	0.6577	-1.0	0.3681	0.5555
4.7467	59.0	14750	11.4894	0.3579	0.6607	0.3484	0.251	0.5795	-1.0	0.305	0.5526	0.5526	0.4747	0.6606	-1.0	0.3579	0.5526
4.5151	60.0	15000	11.6773	0.3545	0.6523	0.3351	0.2474	0.5829	-1.0	0.3112	0.553	0.553	0.4737	0.662	-1.0	0.3545	0.553
4.5151	61.0	15250	11.5171	0.3597	0.657	0.3377	0.2515	0.5837	-1.0	0.2994	0.5498	0.5498	0.4704	0.6591	-1.0	0.3597	0.5498
4.4284	62.0	15500	11.7295	0.3647	0.6688	0.3547	0.2572	0.583	-1.0	0.3078	0.5542	0.5542	0.4742	0.6642	-1.0	0.3647	0.5542
4.4284	63.0	15750	11.9836	0.3454	0.6308	0.3339	0.2376	0.5806	-1.0	0.2966	0.5545	0.5545	0.4763	0.6628	-1.0	0.3454	0.5545
4.3055	64.0	16000	11.8379	0.3469	0.6364	0.3406	0.2407	0.5677	-1.0	0.3009	0.5533	0.5533	0.4769	0.6584	-1.0	0.3469	0.5533
4.3055	65.0	16250	11.8967	0.3558	0.6511	0.3511	0.2498	0.5813	-1.0	0.3081	0.5558	0.5558	0.4785	0.662	-1.0	0.3558	0.5558
4.1396	66.0	16500	12.1633	0.3544	0.6516	0.3404	0.2456	0.5828	-1.0	0.309	0.5539	0.5539	0.4758	0.6613	-1.0	0.3544	0.5539
4.1396	67.0	16750	12.1185	0.3457	0.632	0.3415	0.241	0.5651	-1.0	0.3025	0.5545	0.5551	0.478	0.6613	-1.0	0.3457	0.5551
4.1061	68.0	17000	12.1791	0.3594	0.658	0.3505	0.2533	0.5795	-1.0	0.3103	0.5542	0.5542	0.4769	0.6606	-1.0	0.3594	0.5542
4.1061	69.0	17250	12.1376	0.3555	0.6499	0.3455	0.2488	0.5767	-1.0	0.3069	0.5542	0.5542	0.4753	0.6628	-1.0	0.3555	0.5542
3.9929	70.0	17500	12.2153	0.3564	0.6536	0.339	0.2486	0.5883	-1.0	0.3062	0.5536	0.5542	0.4769	0.6606	-1.0	0.3564	0.5542

Framework versions

Transformers 4.57.1
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.2

dariacuna/rtdetr-v2-r50-finetune-18

rtdetr-v2-r50-finetune-18

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Seed this model