rtdetr-v2-r50-finetune-12

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: 5.9887
Map: 0.5924
Map 50: 0.9088
Map 75: 0.7155
Map Small: 0.5708
Map Medium: 0.6663
Map Large: -1.0
Mar 1: 0.3453
Mar 10: 0.6738
Mar 100: 0.6939
Mar Small: 0.6743
Mar Medium: 0.7437
Mar Large: -1.0
Map Artemia: 0.5924
Mar 100 Artemia: 0.6939

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 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	14.3634	0.4559	0.8203	0.4344	0.3587	0.5855	-1.0	0.3542	0.5832	0.6374	0.5582	0.7432	-1.0	0.4559	0.6374
136.1634	2.0	500	8.1958	0.4766	0.8566	0.4654	0.408	0.5732	-1.0	0.3667	0.5748	0.6109	0.5582	0.6835	-1.0	0.4766	0.6109
136.1634	3.0	750	7.8232	0.5003	0.8815	0.5146	0.4292	0.5933	-1.0	0.381	0.5925	0.6215	0.5712	0.6899	-1.0	0.5003	0.6215
13.7109	4.0	1000	7.5473	0.4849	0.8893	0.4567	0.4133	0.5833	-1.0	0.3791	0.5875	0.6097	0.5554	0.6835	-1.0	0.4849	0.6097
13.7109	5.0	1250	9.1724	0.4708	0.8714	0.4684	0.3975	0.5706	-1.0	0.3695	0.5698	0.5956	0.5484	0.6597	-1.0	0.4708	0.5956
12.0475	6.0	1500	7.6710	0.4792	0.8517	0.4831	0.4009	0.5846	-1.0	0.3773	0.5844	0.6084	0.5571	0.677	-1.0	0.4792	0.6084
12.0475	7.0	1750	7.8450	0.4653	0.8729	0.4027	0.389	0.5678	-1.0	0.3657	0.5682	0.5819	0.5261	0.6576	-1.0	0.4653	0.5819
11.1842	8.0	2000	7.6341	0.4638	0.8721	0.4395	0.3862	0.5737	-1.0	0.3676	0.5664	0.5844	0.5326	0.6554	-1.0	0.4638	0.5844
11.1842	9.0	2250	10.1731	0.4434	0.8557	0.39	0.3506	0.578	-1.0	0.3682	0.5763	0.5897	0.5342	0.6647	-1.0	0.4434	0.5897
10.2581	10.0	2500	7.8777	0.4777	0.8598	0.476	0.4029	0.5815	-1.0	0.3804	0.5726	0.5807	0.5228	0.6597	-1.0	0.4777	0.5807
10.2581	11.0	2750	7.9899	0.4739	0.8646	0.4573	0.3989	0.5816	-1.0	0.3779	0.571	0.5826	0.5217	0.6662	-1.0	0.4739	0.5826
9.6163	12.0	3000	7.9703	0.4496	0.8496	0.4146	0.3656	0.5675	-1.0	0.3632	0.553	0.5614	0.4793	0.6719	-1.0	0.4496	0.5614
9.6163	13.0	3250	9.1562	0.4827	0.8699	0.4751	0.3967	0.5934	-1.0	0.3751	0.5713	0.5757	0.5038	0.6734	-1.0	0.4827	0.5757
9.1076	14.0	3500	8.6625	0.4644	0.8611	0.4586	0.3803	0.5914	-1.0	0.3729	0.5801	0.586	0.5304	0.6626	-1.0	0.4644	0.586
9.1076	15.0	3750	9.1641	0.4598	0.8586	0.4162	0.3849	0.5674	-1.0	0.3682	0.5551	0.5579	0.4859	0.6554	-1.0	0.4598	0.5579
8.5514	16.0	4000	8.7067	0.4435	0.8449	0.397	0.3645	0.5569	-1.0	0.362	0.5327	0.534	0.45	0.6475	-1.0	0.4435	0.534
8.5514	17.0	4250	8.1835	0.4684	0.8674	0.454	0.3929	0.5768	-1.0	0.376	0.5548	0.5548	0.4788	0.6576	-1.0	0.4684	0.5548
8.0201	18.0	4500	8.4612	0.4562	0.8495	0.4414	0.3858	0.5588	-1.0	0.3607	0.5393	0.5393	0.475	0.6266	-1.0	0.4562	0.5393
8.0201	19.0	4750	8.3623	0.464	0.8499	0.4546	0.3827	0.5811	-1.0	0.3667	0.5417	0.5417	0.469	0.6403	-1.0	0.464	0.5417
7.7227	20.0	5000	9.3416	0.4795	0.8753	0.4647	0.4013	0.5798	-1.0	0.3804	0.562	0.562	0.4973	0.6489	-1.0	0.4795	0.562
7.7227	21.0	5250	9.0153	0.4719	0.8678	0.451	0.3972	0.571	-1.0	0.3704	0.5551	0.5551	0.4853	0.6496	-1.0	0.4719	0.5551
7.4099	22.0	5500	8.8487	0.4717	0.8597	0.4281	0.4002	0.5691	-1.0	0.366	0.5449	0.5449	0.4761	0.6381	-1.0	0.4717	0.5449
7.4099	23.0	5750	8.8188	0.469	0.8595	0.4373	0.3981	0.5709	-1.0	0.3673	0.5477	0.5477	0.4793	0.641	-1.0	0.469	0.5477
6.978	24.0	6000	8.8877	0.4735	0.8523	0.4361	0.3937	0.5811	-1.0	0.3679	0.5495	0.5495	0.4647	0.6647	-1.0	0.4735	0.5495
6.978	25.0	6250	8.3720	0.4686	0.8313	0.4582	0.3884	0.5867	-1.0	0.3692	0.543	0.543	0.456	0.6604	-1.0	0.4686	0.543
6.8811	26.0	6500	9.0574	0.4634	0.8618	0.4264	0.382	0.579	-1.0	0.3688	0.5477	0.5477	0.4663	0.6583	-1.0	0.4634	0.5477
6.8811	27.0	6750	8.4806	0.4685	0.8566	0.4409	0.3975	0.5685	-1.0	0.3645	0.5498	0.5498	0.4728	0.654	-1.0	0.4685	0.5498
7.2591	28.0	7000	8.9210	0.4716	0.8603	0.4294	0.4052	0.5639	-1.0	0.362	0.5508	0.5511	0.4902	0.6331	-1.0	0.4716	0.5511
7.2591	29.0	7250	9.5020	0.4584	0.8562	0.4311	0.3773	0.5666	-1.0	0.3642	0.5402	0.5402	0.4549	0.6554	-1.0	0.4584	0.5402
6.7138	30.0	7500	9.6337	0.4752	0.8564	0.4477	0.3977	0.5788	-1.0	0.3754	0.5573	0.5573	0.4832	0.6576	-1.0	0.4752	0.5573
6.7138	31.0	7750	8.6602	0.4785	0.8582	0.4683	0.4026	0.5796	-1.0	0.3735	0.5542	0.5542	0.4717	0.6662	-1.0	0.4785	0.5542
6.1423	32.0	8000	9.1309	0.4681	0.845	0.4574	0.3863	0.5834	-1.0	0.3695	0.5445	0.5445	0.4576	0.6619	-1.0	0.4681	0.5445
6.1423	33.0	8250	8.9111	0.4692	0.8342	0.4653	0.388	0.5841	-1.0	0.3704	0.5458	0.5458	0.4603	0.6612	-1.0	0.4692	0.5458
6.0029	34.0	8500	9.9069	0.4712	0.8415	0.471	0.3894	0.5847	-1.0	0.3713	0.5508	0.5508	0.4701	0.6597	-1.0	0.4712	0.5508
6.0029	35.0	8750	8.9791	0.4717	0.8541	0.4443	0.3959	0.5782	-1.0	0.3682	0.5442	0.5442	0.4625	0.6547	-1.0	0.4717	0.5442
5.7711	36.0	9000	8.8139	0.4739	0.8448	0.4676	0.4019	0.5747	-1.0	0.3751	0.5514	0.5514	0.4783	0.6504	-1.0	0.4739	0.5514
5.7711	37.0	9250	9.8714	0.4775	0.863	0.4494	0.3971	0.5871	-1.0	0.3748	0.5564	0.5564	0.4783	0.6619	-1.0	0.4775	0.5564
5.6549	38.0	9500	10.0788	0.4697	0.8313	0.46	0.3918	0.5789	-1.0	0.3717	0.5486	0.5486	0.4652	0.6612	-1.0	0.4697	0.5486
5.6549	39.0	9750	9.4759	0.4769	0.8538	0.4646	0.4054	0.5756	-1.0	0.3738	0.5523	0.5523	0.4755	0.6561	-1.0	0.4769	0.5523
5.4266	40.0	10000	9.5951	0.4789	0.8525	0.4702	0.4056	0.5814	-1.0	0.376	0.553	0.553	0.4728	0.6619	-1.0	0.4789	0.553
5.4266	41.0	10250	9.8417	0.473	0.8416	0.453	0.3946	0.5817	-1.0	0.3695	0.5498	0.5498	0.469	0.659	-1.0	0.473	0.5498
5.3376	42.0	10500	10.3306	0.4717	0.8397	0.4688	0.3917	0.5809	-1.0	0.3692	0.5495	0.5495	0.4658	0.664	-1.0	0.4717	0.5495
5.3376	43.0	10750	9.7401	0.4752	0.8432	0.454	0.3974	0.5819	-1.0	0.3754	0.5495	0.5495	0.469	0.6583	-1.0	0.4752	0.5495
5.1927	44.0	11000	10.3028	0.48	0.8425	0.4645	0.4012	0.5861	-1.0	0.3735	0.552	0.552	0.4717	0.6612	-1.0	0.48	0.552
5.1927	45.0	11250	10.2989	0.4723	0.8255	0.4796	0.3911	0.5831	-1.0	0.3701	0.5442	0.5442	0.462	0.6554	-1.0	0.4723	0.5442
5.0593	46.0	11500	10.3159	0.4768	0.8434	0.4658	0.4012	0.5816	-1.0	0.3738	0.5533	0.5533	0.4755	0.6583	-1.0	0.4768	0.5533
5.0593	47.0	11750	9.8545	0.4807	0.8435	0.483	0.4035	0.5863	-1.0	0.3754	0.5508	0.5508	0.4696	0.6604	-1.0	0.4807	0.5508
4.8868	48.0	12000	10.4341	0.472	0.835	0.4711	0.3963	0.5779	-1.0	0.3717	0.5445	0.5445	0.463	0.6554	-1.0	0.472	0.5445
4.8868	49.0	12250	10.4958	0.4784	0.8441	0.4673	0.4037	0.584	-1.0	0.3766	0.5489	0.5489	0.4658	0.6619	-1.0	0.4784	0.5489
4.7461	50.0	12500	10.1639	0.4771	0.8445	0.47	0.3997	0.5793	-1.0	0.3754	0.5474	0.5474	0.4658	0.6576	-1.0	0.4771	0.5474
4.7461	51.0	12750	10.4623	0.4735	0.835	0.4511	0.3923	0.5832	-1.0	0.3695	0.5464	0.5464	0.4592	0.664	-1.0	0.4735	0.5464
4.6252	52.0	13000	10.3305	0.4808	0.8357	0.4803	0.4055	0.5851	-1.0	0.3766	0.5514	0.5514	0.4674	0.6647	-1.0	0.4808	0.5514
4.6252	53.0	13250	10.9429	0.475	0.8407	0.4605	0.398	0.5801	-1.0	0.3738	0.5486	0.5486	0.4658	0.6604	-1.0	0.475	0.5486
4.5892	54.0	13500	10.3024	0.4849	0.8434	0.4983	0.4077	0.5911	-1.0	0.3829	0.5561	0.5561	0.4734	0.6676	-1.0	0.4849	0.5561
4.5892	55.0	13750	10.2170	0.4811	0.8449	0.4682	0.4055	0.5875	-1.0	0.3779	0.5523	0.5523	0.469	0.6647	-1.0	0.4811	0.5523
4.4261	56.0	14000	10.4085	0.4815	0.8506	0.4752	0.4069	0.5842	-1.0	0.3757	0.5542	0.5542	0.4788	0.6561	-1.0	0.4815	0.5542
4.4261	57.0	14250	10.0891	0.4798	0.8369	0.4834	0.4059	0.5848	-1.0	0.3782	0.5526	0.5526	0.4707	0.6633	-1.0	0.4798	0.5526
4.3243	58.0	14500	10.3358	0.4775	0.8367	0.4668	0.4012	0.5817	-1.0	0.3745	0.5486	0.5486	0.4674	0.659	-1.0	0.4775	0.5486
4.3243	59.0	14750	10.5238	0.4805	0.8369	0.4834	0.4045	0.585	-1.0	0.376	0.5502	0.5502	0.4663	0.6633	-1.0	0.4805	0.5502
4.1296	60.0	15000	10.3719	0.4823	0.8449	0.4837	0.4062	0.5839	-1.0	0.376	0.5514	0.5514	0.4701	0.6612	-1.0	0.4823	0.5514
4.1296	61.0	15250	10.8361	0.4799	0.8441	0.47	0.4023	0.5876	-1.0	0.3788	0.553	0.553	0.4685	0.6669	-1.0	0.4799	0.553
4.0724	62.0	15500	11.2028	0.4777	0.8445	0.4634	0.402	0.582	-1.0	0.3713	0.5486	0.5486	0.4679	0.6576	-1.0	0.4777	0.5486
4.0724	63.0	15750	11.1891	0.4836	0.8438	0.4907	0.4067	0.5862	-1.0	0.3763	0.5502	0.5502	0.4658	0.664	-1.0	0.4836	0.5502
3.9138	64.0	16000	10.8475	0.4781	0.8351	0.474	0.4004	0.5859	-1.0	0.376	0.548	0.548	0.4614	0.6655	-1.0	0.4781	0.548
3.9138	65.0	16250	11.3144	0.4786	0.8351	0.4996	0.4025	0.5825	-1.0	0.3751	0.5489	0.5489	0.4679	0.6583	-1.0	0.4786	0.5489
3.7917	66.0	16500	11.4832	0.4833	0.8452	0.4946	0.4079	0.5863	-1.0	0.3776	0.552	0.552	0.4701	0.6626	-1.0	0.4833	0.552
3.7917	67.0	16750	11.5115	0.4786	0.845	0.4855	0.4036	0.5838	-1.0	0.3773	0.5498	0.5498	0.4685	0.6597	-1.0	0.4786	0.5498
3.7316	68.0	17000	11.4299	0.4837	0.8443	0.4898	0.4084	0.5851	-1.0	0.3773	0.5539	0.5539	0.4728	0.6633	-1.0	0.4837	0.5539
3.7316	69.0	17250	11.4267	0.4811	0.8361	0.4851	0.4039	0.5885	-1.0	0.3779	0.5533	0.5533	0.4685	0.6676	-1.0	0.4811	0.5533
3.589	70.0	17500	11.3175	0.4805	0.8359	0.5	0.404	0.5855	-1.0	0.3763	0.5505	0.5505	0.4679	0.6619	-1.0	0.4805	0.5505
3.589	71.0	17750	11.6336	0.4805	0.8367	0.4833	0.4039	0.5845	-1.0	0.3748	0.5492	0.5492	0.4652	0.6626	-1.0	0.4805	0.5492
3.5253	72.0	18000	11.5288	0.4814	0.8375	0.5054	0.4058	0.5862	-1.0	0.376	0.5498	0.5498	0.4663	0.6626	-1.0	0.4814	0.5498
3.5253	73.0	18250	11.6309	0.4844	0.8359	0.4943	0.4062	0.5889	-1.0	0.3769	0.5508	0.5508	0.4663	0.6647	-1.0	0.4844	0.5508
3.386	74.0	18500	11.5532	0.4811	0.8373	0.4918	0.4058	0.5833	-1.0	0.3776	0.5498	0.5498	0.4663	0.6626	-1.0	0.4811	0.5498
3.386	75.0	18750	11.9113	0.4829	0.8446	0.4902	0.4074	0.5842	-1.0	0.3776	0.552	0.552	0.469	0.664	-1.0	0.4829	0.552
3.2993	76.0	19000	12.0526	0.4839	0.837	0.4929	0.4093	0.5855	-1.0	0.3776	0.5514	0.5514	0.4685	0.6633	-1.0	0.4839	0.5514
3.2993	77.0	19250	12.0489	0.4838	0.837	0.5074	0.4067	0.5874	-1.0	0.3785	0.5505	0.5505	0.4652	0.6655	-1.0	0.4838	0.5505
3.1814	78.0	19500	12.1260	0.4846	0.8369	0.5039	0.4084	0.5868	-1.0	0.3791	0.5517	0.5517	0.4668	0.6662	-1.0	0.4846	0.5517
3.1814	79.0	19750	12.2362	0.4847	0.837	0.4995	0.4078	0.5868	-1.0	0.3785	0.5523	0.5523	0.4696	0.664	-1.0	0.4847	0.5523
3.1618	80.0	20000	12.2055	0.4861	0.8372	0.5049	0.4092	0.5877	-1.0	0.3791	0.553	0.553	0.4701	0.6647	-1.0	0.4861	0.553

Framework versions

Transformers 4.57.1
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.1

dariacuna/rtdetr-v2-r50-finetune-12

rtdetr-v2-r50-finetune-12

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Seed this model