rtdetr-v2-r50-finetune-13

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: 6.5019
Map: 0.523
Map 50: 0.8174
Map 75: 0.6077
Map Small: 0.4894
Map Medium: 0.6276
Map Large: -1.0
Mar 1: 0.3317
Mar 10: 0.6531
Mar 100: 0.6845
Mar Small: 0.6595
Mar Medium: 0.7483
Mar Large: -1.0
Map Artemia: 0.523
Mar 100 Artemia: 0.6845

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: 80

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	8.6911	0.1717	0.3341	0.1607	0.1039	0.4238	-1.0	0.2199	0.5935	0.6558	0.5951	0.7396	-1.0	0.1717	0.6558
16.7559	2.0	500	9.4289	0.2465	0.4755	0.2212	0.157	0.4397	-1.0	0.2268	0.5598	0.6246	0.5728	0.6964	-1.0	0.2465	0.6246
16.7559	3.0	750	8.6787	0.3559	0.6698	0.3192	0.2664	0.564	-1.0	0.3118	0.5632	0.61	0.5614	0.6748	-1.0	0.3559	0.61
14.238	4.0	1000	8.6466	0.3638	0.6839	0.3361	0.3584	0.4687	-1.0	0.353	0.5794	0.6243	0.5652	0.7043	-1.0	0.3638	0.6243
14.238	5.0	1250	8.5819	0.4219	0.7617	0.3877	0.3178	0.5785	-1.0	0.3492	0.6109	0.6355	0.581	0.7086	-1.0	0.4219	0.6355
13.5958	6.0	1500	8.2032	0.4467	0.817	0.4301	0.375	0.5581	-1.0	0.3623	0.5723	0.6093	0.5533	0.6842	-1.0	0.4467	0.6093
13.5958	7.0	1750	8.2276	0.2162	0.4155	0.184	0.1467	0.5516	-1.0	0.2308	0.5813	0.6312	0.5929	0.6835	-1.0	0.2162	0.6312
12.4655	8.0	2000	8.4735	0.2072	0.4046	0.1894	0.2735	0.4419	-1.0	0.3234	0.5673	0.5984	0.5489	0.6655	-1.0	0.2072	0.5984
12.4655	9.0	2250	8.4963	0.2862	0.5416	0.2482	0.2377	0.5643	-1.0	0.2651	0.5826	0.6019	0.5462	0.677	-1.0	0.2862	0.6019
11.4111	10.0	2500	8.9231	0.1209	0.2328	0.0985	0.2457	0.3157	-1.0	0.2981	0.5664	0.5863	0.5342	0.6576	-1.0	0.1209	0.5863
11.4111	11.0	2750	8.8147	0.3115	0.6029	0.2848	0.2301	0.5664	-1.0	0.2888	0.5701	0.5947	0.5391	0.6705	-1.0	0.3115	0.5947
10.7381	12.0	3000	8.9513	0.0813	0.1564	0.0738	0.0456	0.3201	-1.0	0.1333	0.5536	0.5832	0.5185	0.6712	-1.0	0.0813	0.5832
10.7381	13.0	3250	8.6451	0.384	0.7314	0.3388	0.309	0.5441	-1.0	0.3405	0.5548	0.5869	0.5337	0.6597	-1.0	0.384	0.5869
10.2413	14.0	3500	9.2982	0.2659	0.5204	0.2301	0.1942	0.5159	-1.0	0.2583	0.5692	0.6028	0.5533	0.6698	-1.0	0.2659	0.6028
10.2413	15.0	3750	9.0844	0.3617	0.6869	0.334	0.2632	0.5678	-1.0	0.3062	0.5751	0.5841	0.5228	0.6662	-1.0	0.3617	0.5841
9.6157	16.0	4000	9.2734	0.26	0.512	0.232	0.1909	0.5004	-1.0	0.2648	0.552	0.5667	0.5103	0.6424	-1.0	0.26	0.5667
9.6157	17.0	4250	9.5975	0.2483	0.4893	0.2251	0.1648	0.5153	-1.0	0.2212	0.5489	0.5611	0.4962	0.6489	-1.0	0.2483	0.5611
8.9942	18.0	4500	9.3368	0.1837	0.3703	0.1552	0.1666	0.3349	-1.0	0.2885	0.5402	0.5414	0.4902	0.6115	-1.0	0.1837	0.5414
8.9942	19.0	4750	9.3903	0.4204	0.8147	0.3888	0.3283	0.5745	-1.0	0.3449	0.5545	0.5801	0.5152	0.6683	-1.0	0.4204	0.5801
8.7433	20.0	5000	9.2044	0.4156	0.7742	0.3808	0.3241	0.5585	-1.0	0.3533	0.5567	0.5632	0.4984	0.6496	-1.0	0.4156	0.5632
8.7433	21.0	5250	9.2036	0.3809	0.7539	0.3507	0.2759	0.5566	-1.0	0.3196	0.5495	0.5564	0.4935	0.6417	-1.0	0.3809	0.5564
8.3487	22.0	5500	9.8918	0.1445	0.2827	0.1379	0.0722	0.562	-1.0	0.1903	0.5579	0.5822	0.5212	0.664	-1.0	0.1445	0.5822
8.3487	23.0	5750	9.4369	0.3495	0.6816	0.3205	0.2517	0.5532	-1.0	0.3009	0.5408	0.5483	0.4848	0.6345	-1.0	0.3495	0.5483
8.0076	24.0	6000	9.5877	0.355	0.6893	0.3269	0.2519	0.5436	-1.0	0.31	0.5417	0.5464	0.4788	0.6388	-1.0	0.355	0.5464
8.0076	25.0	6250	8.9577	0.4112	0.7815	0.3663	0.3067	0.5674	-1.0	0.3355	0.5477	0.5483	0.4761	0.6453	-1.0	0.4112	0.5483
7.8147	26.0	6500	9.6258	0.389	0.7366	0.358	0.2725	0.5801	-1.0	0.3361	0.581	0.5969	0.5353	0.6791	-1.0	0.389	0.5969
7.8147	27.0	6750	9.4621	0.4175	0.7835	0.4131	0.3157	0.5712	-1.0	0.3427	0.5748	0.5785	0.512	0.6676	-1.0	0.4175	0.5785
7.4382	28.0	7000	10.4193	0.3391	0.6597	0.3232	0.2309	0.5559	-1.0	0.2963	0.5511	0.5523	0.4842	0.6439	-1.0	0.3391	0.5523
7.4382	29.0	7250	10.5008	0.1499	0.2955	0.135	0.0809	0.5352	-1.0	0.2234	0.547	0.5498	0.4826	0.641	-1.0	0.1499	0.5498
7.2444	30.0	7500	9.9424	0.3539	0.6873	0.3332	0.241	0.5657	-1.0	0.2984	0.547	0.5492	0.4717	0.6532	-1.0	0.3539	0.5492
7.2444	31.0	7750	11.0772	0.3295	0.6527	0.2902	0.2237	0.5463	-1.0	0.2931	0.5349	0.5349	0.4592	0.6374	-1.0	0.3295	0.5349
7.0042	32.0	8000	10.0542	0.4034	0.7649	0.3688	0.3033	0.5673	-1.0	0.3368	0.538	0.538	0.4582	0.6453	-1.0	0.4034	0.538
7.0042	33.0	8250	10.2424	0.3748	0.7088	0.3515	0.2659	0.5692	-1.0	0.3255	0.5573	0.5595	0.4853	0.6597	-1.0	0.3748	0.5595
6.8233	34.0	8500	9.6359	0.3792	0.7143	0.3644	0.2755	0.5731	-1.0	0.3283	0.5421	0.5421	0.4668	0.6424	-1.0	0.3792	0.5421
6.8233	35.0	8750	10.4864	0.34	0.6444	0.3338	0.2294	0.5728	-1.0	0.3097	0.5498	0.5502	0.4745	0.6518	-1.0	0.34	0.5502
6.656	36.0	9000	10.7000	0.3453	0.6644	0.3272	0.2422	0.5579	-1.0	0.3044	0.5445	0.5445	0.4707	0.6439	-1.0	0.3453	0.5445
6.656	37.0	9250	10.5485	0.1923	0.3666	0.1796	0.1131	0.564	-1.0	0.2305	0.553	0.5536	0.4783	0.6547	-1.0	0.1923	0.5536
6.6012	38.0	9500	11.3336	0.2934	0.5704	0.2625	0.1924	0.5579	-1.0	0.2645	0.5402	0.5402	0.462	0.6453	-1.0	0.2934	0.5402
6.6012	39.0	9750	11.6340	0.342	0.6523	0.3216	0.2316	0.5722	-1.0	0.3031	0.5511	0.5514	0.4804	0.6468	-1.0	0.342	0.5514
6.3092	40.0	10000	10.7695	0.3574	0.6784	0.3379	0.2481	0.5693	-1.0	0.3206	0.5477	0.5483	0.4728	0.6496	-1.0	0.3574	0.5483
6.3092	41.0	10250	10.8914	0.3596	0.6777	0.3491	0.2505	0.5749	-1.0	0.3156	0.5502	0.5502	0.4755	0.6511	-1.0	0.3596	0.5502
6.1174	42.0	10500	10.4539	0.1585	0.3052	0.1423	0.1621	0.2891	-1.0	0.2869	0.548	0.5523	0.4728	0.659	-1.0	0.1585	0.5523
6.1174	43.0	10750	10.4479	0.3002	0.5721	0.2834	0.2166	0.5055	-1.0	0.3075	0.5617	0.5682	0.4989	0.6612	-1.0	0.3002	0.5682
6.0535	44.0	11000	10.9388	0.3311	0.6372	0.3062	0.2214	0.5713	-1.0	0.29	0.5551	0.5558	0.4745	0.6647	-1.0	0.3311	0.5558
6.0535	45.0	11250	11.3116	0.3051	0.5736	0.2852	0.2282	0.4599	-1.0	0.3134	0.5614	0.5617	0.4897	0.6583	-1.0	0.3051	0.5617
5.8815	46.0	11500	11.2927	0.3288	0.622	0.3023	0.2226	0.5774	-1.0	0.2931	0.5632	0.5657	0.4924	0.664	-1.0	0.3288	0.5657
5.8815	47.0	11750	10.1384	0.3874	0.7226	0.3605	0.2861	0.5714	-1.0	0.3246	0.548	0.5486	0.4701	0.6547	-1.0	0.3874	0.5486
5.698	48.0	12000	10.8399	0.3833	0.7195	0.368	0.2735	0.575	-1.0	0.3287	0.5526	0.5526	0.4712	0.6626	-1.0	0.3833	0.5526
5.698	49.0	12250	10.2175	0.4013	0.7555	0.3777	0.2991	0.5742	-1.0	0.3374	0.5551	0.5551	0.4821	0.6532	-1.0	0.4013	0.5551
5.5552	50.0	12500	10.2580	0.3957	0.7264	0.3782	0.2902	0.5792	-1.0	0.334	0.5636	0.5636	0.4913	0.6612	-1.0	0.3957	0.5636
5.5552	51.0	12750	10.8993	0.3881	0.7204	0.3709	0.2844	0.5743	-1.0	0.3364	0.5536	0.5536	0.481	0.6518	-1.0	0.3881	0.5536
5.4306	52.0	13000	11.2916	0.3004	0.5728	0.2687	0.1941	0.5724	-1.0	0.2788	0.5539	0.5539	0.4815	0.6511	-1.0	0.3004	0.5539
5.4306	53.0	13250	11.6511	0.3488	0.6661	0.3259	0.241	0.5681	-1.0	0.2997	0.5545	0.5545	0.4842	0.6489	-1.0	0.3488	0.5545
5.3966	54.0	13500	12.0120	0.2932	0.5626	0.2685	0.1899	0.5509	-1.0	0.253	0.5542	0.5542	0.4821	0.6518	-1.0	0.2932	0.5542
5.3966	55.0	13750	11.1822	0.2903	0.5526	0.2668	0.1887	0.5536	-1.0	0.2732	0.562	0.562	0.488	0.6612	-1.0	0.2903	0.562
5.235	56.0	14000	11.4522	0.1958	0.3724	0.1849	0.1175	0.5253	-1.0	0.1938	0.5558	0.5558	0.4837	0.6525	-1.0	0.1958	0.5558
5.235	57.0	14250	12.4220	0.2688	0.5086	0.2511	0.1726	0.5663	-1.0	0.2455	0.5533	0.5536	0.4842	0.6468	-1.0	0.2688	0.5536
5.1989	58.0	14500	11.7566	0.2061	0.3938	0.1874	0.1179	0.5686	-1.0	0.2	0.5533	0.5536	0.4826	0.6489	-1.0	0.2061	0.5536
5.1989	59.0	14750	12.2170	0.2955	0.5607	0.2717	0.1864	0.5692	-1.0	0.253	0.5495	0.5495	0.4745	0.6504	-1.0	0.2955	0.5495
4.9526	60.0	15000	11.6398	0.3238	0.6104	0.3002	0.214	0.5691	-1.0	0.2891	0.5502	0.5502	0.4772	0.6489	-1.0	0.3238	0.5502
4.9526	61.0	15250	12.0900	0.3352	0.6329	0.3131	0.2253	0.5699	-1.0	0.2885	0.5564	0.557	0.4859	0.6525	-1.0	0.3352	0.557
4.8838	62.0	15500	11.5574	0.3574	0.6697	0.3268	0.2503	0.572	-1.0	0.3125	0.5539	0.5539	0.4788	0.6554	-1.0	0.3574	0.5539
4.8838	63.0	15750	12.1932	0.334	0.6289	0.3085	0.2293	0.5699	-1.0	0.3034	0.5505	0.5505	0.4739	0.6532	-1.0	0.334	0.5505
4.7424	64.0	16000	11.7992	0.3396	0.6382	0.3157	0.2333	0.5712	-1.0	0.2975	0.5498	0.5498	0.4745	0.6511	-1.0	0.3396	0.5498
4.7424	65.0	16250	12.6417	0.3169	0.5998	0.2861	0.2152	0.5647	-1.0	0.2894	0.5508	0.5508	0.4761	0.6511	-1.0	0.3169	0.5508
4.5996	66.0	16500	12.6278	0.3233	0.6227	0.2925	0.2235	0.5667	-1.0	0.295	0.5486	0.5486	0.4745	0.6489	-1.0	0.3233	0.5486
4.5996	67.0	16750	12.2864	0.3506	0.6716	0.3182	0.243	0.5706	-1.0	0.3009	0.547	0.547	0.4696	0.6511	-1.0	0.3506	0.547
4.5648	68.0	17000	12.5421	0.3635	0.6826	0.3371	0.2569	0.5665	-1.0	0.3115	0.5492	0.5492	0.4739	0.6504	-1.0	0.3635	0.5492
4.5648	69.0	17250	12.7722	0.3511	0.6628	0.3218	0.2457	0.5685	-1.0	0.3065	0.5502	0.5502	0.4755	0.6511	-1.0	0.3511	0.5502
4.3825	70.0	17500	12.9456	0.344	0.6482	0.318	0.2383	0.5633	-1.0	0.3022	0.5495	0.5495	0.4723	0.6532	-1.0	0.344	0.5495
4.3825	71.0	17750	12.8847	0.3437	0.6457	0.316	0.2355	0.5653	-1.0	0.2919	0.5517	0.5517	0.4761	0.6532	-1.0	0.3437	0.5517
4.3724	72.0	18000	13.0869	0.3471	0.6556	0.3179	0.2406	0.5675	-1.0	0.3012	0.5502	0.5502	0.4739	0.6525	-1.0	0.3471	0.5502
4.3724	73.0	18250	13.2646	0.3528	0.6648	0.3223	0.2457	0.5633	-1.0	0.3006	0.5495	0.5495	0.4745	0.6511	-1.0	0.3528	0.5495
4.2321	74.0	18500	12.9260	0.3562	0.6729	0.3288	0.2501	0.5644	-1.0	0.3019	0.5486	0.5486	0.4734	0.6496	-1.0	0.3562	0.5486
4.2321	75.0	18750	13.1402	0.3535	0.6639	0.3226	0.2464	0.562	-1.0	0.3016	0.5505	0.5505	0.4739	0.6532	-1.0	0.3535	0.5505
4.1502	76.0	19000	13.1223	0.357	0.6727	0.3257	0.2526	0.5606	-1.0	0.3025	0.5483	0.5483	0.4723	0.6504	-1.0	0.357	0.5483
4.1502	77.0	19250	13.0908	0.3611	0.677	0.333	0.2522	0.5685	-1.0	0.3044	0.552	0.552	0.4755	0.6554	-1.0	0.3611	0.552
4.0569	78.0	19500	13.2942	0.3541	0.6695	0.3224	0.2465	0.5653	-1.0	0.3003	0.5486	0.5486	0.4717	0.6518	-1.0	0.3541	0.5486
4.0569	79.0	19750	13.3401	0.3572	0.6702	0.3252	0.2486	0.5658	-1.0	0.2981	0.5495	0.5495	0.475	0.6496	-1.0	0.3572	0.5495
4.034	80.0	20000	13.3229	0.3534	0.6634	0.3231	0.247	0.564	-1.0	0.2997	0.5495	0.5495	0.475	0.6496	-1.0	0.3534	0.5495

Framework versions

Transformers 4.57.1
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.2

dariacuna/rtdetr-v2-r50-finetune-13

rtdetr-v2-r50-finetune-13

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Seed this model