rtdetr-v2-r50-finetune-19

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: 5.7893
Map: 0.5958
Map 50: 0.913
Map 75: 0.6854
Map Small: 0.5668
Map Medium: 0.6783
Map Large: -1.0
Mar 1: 0.346
Mar 10: 0.6715
Mar 100: 0.7071
Mar Small: 0.6784
Mar Medium: 0.7805
Mar Large: -1.0
Map Artemia: 0.5958
Mar 100 Artemia: 0.7071

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	15.4471	0.4602	0.8135	0.4712	0.36	0.5794	-1.0	0.3551	0.6012	0.6505	0.5677	0.7642	-1.0	0.4602	0.6505
294.1038	2.0	500	8.2104	0.4944	0.8664	0.5264	0.425	0.589	-1.0	0.3723	0.585	0.6355	0.5801	0.7124	-1.0	0.4944	0.6355
294.1038	3.0	750	7.5714	0.5041	0.8852	0.5233	0.4306	0.6117	-1.0	0.3813	0.5969	0.6523	0.5855	0.7453	-1.0	0.5041	0.6523
13.5778	4.0	1000	7.5424	0.5045	0.8843	0.5397	0.4305	0.6108	-1.0	0.3866	0.6181	0.6682	0.6097	0.7474	-1.0	0.5045	0.6682
13.5778	5.0	1250	7.7068	0.4804	0.8815	0.4347	0.4079	0.5825	-1.0	0.3776	0.5773	0.6159	0.5613	0.6912	-1.0	0.4804	0.6159
11.7918	6.0	1500	7.9212	0.4811	0.8438	0.4835	0.4009	0.6004	-1.0	0.3804	0.5782	0.5947	0.5253	0.6898	-1.0	0.4811	0.5947
11.7918	7.0	1750	7.7355	0.4615	0.8629	0.4159	0.3815	0.5769	-1.0	0.3645	0.5667	0.5798	0.5043	0.6839	-1.0	0.4615	0.5798
10.9583	8.0	2000	8.4821	0.4548	0.8762	0.4127	0.3792	0.5757	-1.0	0.362	0.5673	0.5969	0.5355	0.681	-1.0	0.4548	0.5969
10.9583	9.0	2250	7.8946	0.4731	0.8724	0.4631	0.3976	0.5894	-1.0	0.3695	0.5794	0.591	0.5253	0.6818	-1.0	0.4731	0.591
10.0548	10.0	2500	7.9248	0.4751	0.8794	0.4336	0.399	0.5874	-1.0	0.366	0.585	0.6097	0.5473	0.6949	-1.0	0.4751	0.6097
10.0548	11.0	2750	8.2730	0.464	0.8638	0.4356	0.3839	0.5764	-1.0	0.3692	0.5548	0.5611	0.4903	0.6591	-1.0	0.464	0.5611
9.3904	12.0	3000	8.8837	0.4419	0.8341	0.4202	0.3549	0.5709	-1.0	0.3567	0.5393	0.5486	0.4634	0.6657	-1.0	0.4419	0.5486
9.3904	13.0	3250	8.2422	0.4708	0.8468	0.4589	0.3878	0.585	-1.0	0.3713	0.5583	0.567	0.4978	0.662	-1.0	0.4708	0.567
8.8342	14.0	3500	9.2072	0.4348	0.8142	0.4401	0.3439	0.5833	-1.0	0.3592	0.5551	0.5664	0.4909	0.6708	-1.0	0.4348	0.5664
8.8342	15.0	3750	8.9809	0.4658	0.8561	0.4366	0.3873	0.5748	-1.0	0.366	0.5555	0.5576	0.4806	0.6613	-1.0	0.4658	0.5576
8.2416	16.0	4000	8.8302	0.4618	0.84	0.4362	0.3804	0.5791	-1.0	0.3698	0.5445	0.5464	0.4645	0.6599	-1.0	0.4618	0.5464
8.2416	17.0	4250	10.3475	0.417	0.8004	0.3573	0.3239	0.5598	-1.0	0.3458	0.5421	0.5452	0.4677	0.6511	-1.0	0.417	0.5452
7.6898	18.0	4500	9.3375	0.4586	0.8616	0.411	0.3764	0.576	-1.0	0.3685	0.5486	0.5517	0.4828	0.6482	-1.0	0.4586	0.5517
7.6898	19.0	4750	9.6257	0.4633	0.8587	0.4292	0.3733	0.5844	-1.0	0.372	0.5642	0.5685	0.4946	0.6693	-1.0	0.4633	0.5685
7.2914	20.0	5000	9.2906	0.4613	0.8437	0.433	0.3754	0.582	-1.0	0.3685	0.5573	0.5589	0.4855	0.6599	-1.0	0.4613	0.5589
7.2914	21.0	5250	10.5863	0.4557	0.8389	0.4384	0.3741	0.5727	-1.0	0.3548	0.5408	0.5439	0.4645	0.6526	-1.0	0.4557	0.5439
6.8948	22.0	5500	9.9043	0.4597	0.8377	0.4363	0.3731	0.5844	-1.0	0.3595	0.5433	0.5442	0.4624	0.6562	-1.0	0.4597	0.5442
6.8948	23.0	5750	10.8579	0.4638	0.851	0.4306	0.3773	0.5817	-1.0	0.3611	0.5558	0.5564	0.479	0.6628	-1.0	0.4638	0.5564
6.5026	24.0	6000	9.6172	0.4717	0.8415	0.4534	0.3863	0.5865	-1.0	0.3707	0.5436	0.5439	0.4602	0.6591	-1.0	0.4717	0.5439
6.5026	25.0	6250	9.2460	0.4678	0.8483	0.4336	0.3851	0.5823	-1.0	0.367	0.5433	0.5439	0.4597	0.6599	-1.0	0.4678	0.5439
6.1938	26.0	6500	10.2273	0.4624	0.8458	0.4168	0.3808	0.578	-1.0	0.3654	0.5449	0.5455	0.4613	0.6613	-1.0	0.4624	0.5455
6.1938	27.0	6750	11.2959	0.457	0.851	0.4147	0.3712	0.5724	-1.0	0.362	0.5439	0.5442	0.471	0.6453	-1.0	0.457	0.5442
5.912	28.0	7000	11.1027	0.4665	0.8473	0.447	0.384	0.5795	-1.0	0.3657	0.5474	0.5483	0.471	0.654	-1.0	0.4665	0.5483
5.912	29.0	7250	10.0175	0.4692	0.8574	0.4379	0.3922	0.5793	-1.0	0.3726	0.5539	0.5545	0.4823	0.6533	-1.0	0.4692	0.5545
5.6797	30.0	7500	9.6599	0.4758	0.8578	0.4468	0.3956	0.584	-1.0	0.3735	0.5467	0.548	0.4699	0.6555	-1.0	0.4758	0.548
5.6797	31.0	7750	9.6127	0.4724	0.8472	0.444	0.3891	0.5854	-1.0	0.3713	0.5514	0.552	0.4661	0.6701	-1.0	0.4724	0.552
5.4538	32.0	8000	9.8352	0.4739	0.8502	0.4465	0.3931	0.5831	-1.0	0.3713	0.5483	0.5483	0.4688	0.6569	-1.0	0.4739	0.5483
5.4538	33.0	8250	10.1440	0.4755	0.8541	0.4531	0.3957	0.5824	-1.0	0.3763	0.552	0.5533	0.4769	0.6584	-1.0	0.4755	0.5533
5.2083	34.0	8500	10.8088	0.4713	0.8385	0.4716	0.3881	0.585	-1.0	0.3704	0.5464	0.5467	0.4672	0.6569	-1.0	0.4713	0.5467
5.2083	35.0	8750	10.9905	0.4764	0.8481	0.4577	0.396	0.5858	-1.0	0.3692	0.5455	0.547	0.4672	0.6569	-1.0	0.4764	0.547
5.0279	36.0	9000	10.2119	0.4716	0.851	0.4321	0.3914	0.582	-1.0	0.3713	0.5458	0.547	0.4667	0.6577	-1.0	0.4716	0.547
5.0279	37.0	9250	10.2078	0.4768	0.8509	0.4622	0.3973	0.5875	-1.0	0.3773	0.553	0.5533	0.472	0.665	-1.0	0.4768	0.5533
4.7962	38.0	9500	10.8508	0.4716	0.8488	0.4513	0.3917	0.5804	-1.0	0.3698	0.547	0.547	0.4634	0.662	-1.0	0.4716	0.547
4.7962	39.0	9750	12.3011	0.4777	0.8577	0.4564	0.3959	0.5845	-1.0	0.372	0.552	0.552	0.472	0.662	-1.0	0.4777	0.552
4.4931	40.0	10000	11.4546	0.4756	0.8493	0.447	0.3942	0.5874	-1.0	0.371	0.552	0.552	0.4694	0.6657	-1.0	0.4756	0.552
4.4931	41.0	10250	11.3260	0.4763	0.8509	0.4567	0.3994	0.5809	-1.0	0.3698	0.5505	0.5511	0.471	0.6613	-1.0	0.4763	0.5511
4.2831	42.0	10500	10.9961	0.4702	0.8507	0.431	0.3915	0.5784	-1.0	0.3676	0.547	0.547	0.4667	0.6577	-1.0	0.4702	0.547
4.2831	43.0	10750	11.5155	0.473	0.8408	0.4375	0.3911	0.5864	-1.0	0.3723	0.5486	0.5486	0.464	0.6657	-1.0	0.473	0.5486
4.082	44.0	11000	11.4698	0.473	0.851	0.4366	0.3913	0.584	-1.0	0.3704	0.5492	0.5492	0.4694	0.6591	-1.0	0.473	0.5492
4.082	45.0	11250	11.8375	0.4802	0.8504	0.4552	0.3996	0.5904	-1.0	0.3723	0.5526	0.5526	0.4726	0.6635	-1.0	0.4802	0.5526
3.9037	46.0	11500	12.0670	0.4759	0.8427	0.4541	0.394	0.5873	-1.0	0.3701	0.5483	0.5483	0.464	0.665	-1.0	0.4759	0.5483
3.9037	47.0	11750	12.5318	0.4743	0.8498	0.4423	0.3906	0.5898	-1.0	0.3717	0.5492	0.5492	0.4645	0.6657	-1.0	0.4743	0.5492
3.653	48.0	12000	12.4333	0.4738	0.8505	0.4515	0.3902	0.5869	-1.0	0.3695	0.548	0.548	0.4651	0.662	-1.0	0.4738	0.548
3.653	49.0	12250	12.5932	0.4763	0.8421	0.4544	0.3921	0.5888	-1.0	0.372	0.5492	0.5492	0.4651	0.665	-1.0	0.4763	0.5492
3.5337	50.0	12500	12.4681	0.4745	0.8422	0.4532	0.3916	0.5898	-1.0	0.371	0.5489	0.5489	0.4645	0.665	-1.0	0.4745	0.5489

Framework versions

Transformers 4.57.1
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.2

dariacuna/rtdetr-v2-r50-finetune-19

rtdetr-v2-r50-finetune-19

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Seed this model