rtdetr-v2-r50-finetune-17

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: 5.8163
Map: 0.5846
Map 50: 0.8995
Map 75: 0.6952
Map Small: 0.5485
Map Medium: 0.6752
Map Large: -1.0
Mar 1: 0.3408
Mar 10: 0.6505
Mar 100: 0.6631
Mar Small: 0.6405
Mar Medium: 0.7207
Mar Large: -1.0
Map Artemia: 0.5846
Mar 100 Artemia: 0.6631

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	13.4833	0.4609	0.8239	0.4532	0.3697	0.5745	-1.0	0.3523	0.5835	0.6417	0.5726	0.7365	-1.0	0.4609	0.6417
98.512	2.0	500	8.1338	0.4816	0.8681	0.4733	0.4061	0.5851	-1.0	0.3673	0.5695	0.6143	0.5608	0.6891	-1.0	0.4816	0.6143
98.512	3.0	750	7.8676	0.4886	0.8826	0.486	0.3983	0.6132	-1.0	0.3623	0.5732	0.6106	0.55	0.6934	-1.0	0.4886	0.6106
14.4472	4.0	1000	7.5051	0.4856	0.8874	0.4631	0.418	0.5792	-1.0	0.3819	0.566	0.5953	0.5495	0.6599	-1.0	0.4856	0.5953
14.4472	5.0	1250	7.3685	0.5014	0.8831	0.5397	0.4268	0.6023	-1.0	0.3807	0.5794	0.5969	0.5414	0.6737	-1.0	0.5014	0.5969
13.4927	6.0	1500	7.4613	0.4979	0.8739	0.5153	0.4207	0.6005	-1.0	0.3763	0.5741	0.5916	0.5398	0.6635	-1.0	0.4979	0.5916
13.4927	7.0	1750	7.4155	0.4883	0.8896	0.4602	0.4146	0.5846	-1.0	0.3741	0.5826	0.6	0.5527	0.6657	-1.0	0.4883	0.6
13.1819	8.0	2000	7.3981	0.4877	0.8863	0.454	0.4266	0.5713	-1.0	0.376	0.5701	0.5875	0.5462	0.6438	-1.0	0.4877	0.5875
13.1819	9.0	2250	7.7891	0.4929	0.8808	0.4832	0.4329	0.5724	-1.0	0.386	0.5816	0.59	0.5478	0.6474	-1.0	0.4929	0.59
12.7799	10.0	2500	7.6775	0.4866	0.8915	0.4488	0.4127	0.5844	-1.0	0.381	0.5723	0.5844	0.5328	0.6562	-1.0	0.4866	0.5844
12.7799	11.0	2750	7.5852	0.4981	0.8788	0.5056	0.4235	0.5984	-1.0	0.3879	0.581	0.5978	0.5441	0.6723	-1.0	0.4981	0.5978
12.6518	12.0	3000	7.8547	0.4766	0.8661	0.4646	0.3893	0.5912	-1.0	0.3664	0.5607	0.5841	0.5199	0.6723	-1.0	0.4766	0.5841
12.6518	13.0	3250	7.6425	0.4765	0.8525	0.4895	0.4009	0.5828	-1.0	0.3639	0.5589	0.5726	0.5172	0.6496	-1.0	0.4765	0.5726
12.3862	14.0	3500	7.4230	0.4862	0.8827	0.4771	0.421	0.5789	-1.0	0.3913	0.5785	0.5813	0.5344	0.6453	-1.0	0.4862	0.5813
12.3862	15.0	3750	7.9584	0.4686	0.8641	0.4604	0.3828	0.5915	-1.0	0.3773	0.5642	0.5713	0.5043	0.6635	-1.0	0.4686	0.5713
12.0196	16.0	4000	8.1336	0.4757	0.8562	0.4901	0.3903	0.592	-1.0	0.3801	0.5698	0.5717	0.5097	0.6577	-1.0	0.4757	0.5717
12.0196	17.0	4250	7.5777	0.4708	0.8591	0.4544	0.3965	0.5786	-1.0	0.3738	0.5632	0.5664	0.5048	0.6504	-1.0	0.4708	0.5664
11.6984	18.0	4500	7.5543	0.4656	0.8544	0.4538	0.3858	0.5895	-1.0	0.3738	0.5483	0.5517	0.471	0.6642	-1.0	0.4656	0.5517
11.6984	19.0	4750	7.8076	0.46	0.8527	0.4625	0.3827	0.5718	-1.0	0.3685	0.5461	0.5558	0.4919	0.6445	-1.0	0.46	0.5558
11.5058	20.0	5000	7.8777	0.4592	0.8579	0.4237	0.3782	0.5757	-1.0	0.3657	0.5445	0.5452	0.4677	0.6518	-1.0	0.4592	0.5452
11.5058	21.0	5250	7.8221	0.4799	0.8678	0.4849	0.4051	0.5783	-1.0	0.3879	0.5738	0.5751	0.5199	0.6518	-1.0	0.4799	0.5751
11.3416	22.0	5500	7.7471	0.4648	0.8536	0.4574	0.3743	0.5935	-1.0	0.3738	0.552	0.5561	0.4828	0.6569	-1.0	0.4648	0.5561
11.3416	23.0	5750	7.9992	0.4633	0.8716	0.4251	0.3861	0.5635	-1.0	0.3807	0.5614	0.5614	0.4984	0.6482	-1.0	0.4633	0.5614
10.8843	24.0	6000	8.0876	0.4497	0.8444	0.3947	0.3536	0.5798	-1.0	0.3698	0.5455	0.5467	0.464	0.6599	-1.0	0.4497	0.5467
10.8843	25.0	6250	7.6154	0.4554	0.8505	0.4111	0.3694	0.5769	-1.0	0.366	0.534	0.5358	0.4462	0.6599	-1.0	0.4554	0.5358
10.8928	26.0	6500	7.5963	0.4533	0.8511	0.4193	0.3763	0.5613	-1.0	0.3645	0.5402	0.5402	0.464	0.6445	-1.0	0.4533	0.5402
10.8928	27.0	6750	7.9062	0.4482	0.8479	0.3959	0.365	0.5691	-1.0	0.3639	0.5327	0.5327	0.4597	0.6343	-1.0	0.4482	0.5327
10.5503	28.0	7000	7.9219	0.4442	0.849	0.4234	0.3643	0.5597	-1.0	0.3642	0.5445	0.5445	0.4796	0.6343	-1.0	0.4442	0.5445
10.5503	29.0	7250	7.9543	0.4532	0.8582	0.4212	0.3689	0.5762	-1.0	0.3692	0.5414	0.5414	0.4602	0.6533	-1.0	0.4532	0.5414
10.305	30.0	7500	7.8435	0.4502	0.8482	0.4402	0.3597	0.5847	-1.0	0.3654	0.534	0.534	0.4468	0.654	-1.0	0.4502	0.534
10.305	31.0	7750	7.9075	0.4664	0.8562	0.4384	0.3876	0.5729	-1.0	0.3754	0.5442	0.5442	0.4694	0.6467	-1.0	0.4664	0.5442
10.0286	32.0	8000	7.9997	0.4553	0.8615	0.4367	0.3671	0.5816	-1.0	0.3763	0.5424	0.5424	0.4613	0.6526	-1.0	0.4553	0.5424
10.0286	33.0	8250	8.2244	0.4616	0.8717	0.444	0.3842	0.5697	-1.0	0.372	0.547	0.547	0.4715	0.6511	-1.0	0.4616	0.547
9.822	34.0	8500	8.3654	0.4493	0.8626	0.4076	0.3641	0.5698	-1.0	0.352	0.5234	0.5234	0.4392	0.6394	-1.0	0.4493	0.5234
9.822	35.0	8750	8.1392	0.4548	0.8591	0.421	0.3673	0.5791	-1.0	0.3707	0.5467	0.5467	0.4694	0.6533	-1.0	0.4548	0.5467
9.6952	36.0	9000	8.0096	0.4641	0.8654	0.4447	0.3887	0.5761	-1.0	0.366	0.548	0.548	0.4769	0.6453	-1.0	0.4641	0.548
9.6952	37.0	9250	8.4491	0.439	0.8512	0.3937	0.3519	0.5595	-1.0	0.3629	0.5308	0.5308	0.4462	0.6467	-1.0	0.439	0.5308
9.4144	38.0	9500	8.6595	0.4499	0.8527	0.4391	0.3663	0.5807	-1.0	0.3741	0.5508	0.5508	0.4731	0.6569	-1.0	0.4499	0.5508
9.4144	39.0	9750	8.4940	0.4415	0.8508	0.405	0.3493	0.5719	-1.0	0.3632	0.5321	0.5321	0.4462	0.6504	-1.0	0.4415	0.5321
9.0742	40.0	10000	8.1819	0.4423	0.8628	0.3711	0.3565	0.5677	-1.0	0.3592	0.5321	0.5321	0.4489	0.6474	-1.0	0.4423	0.5321
9.0742	41.0	10250	8.4433	0.445	0.8682	0.3842	0.3661	0.568	-1.0	0.3632	0.543	0.543	0.4694	0.6445	-1.0	0.445	0.543
8.8479	42.0	10500	8.4365	0.4481	0.8535	0.4342	0.3667	0.5688	-1.0	0.3604	0.5246	0.5246	0.4398	0.6423	-1.0	0.4481	0.5246
8.8479	43.0	10750	8.7075	0.4385	0.8622	0.377	0.3644	0.5478	-1.0	0.3611	0.5299	0.5299	0.4554	0.6328	-1.0	0.4385	0.5299
8.6742	44.0	11000	8.4554	0.4352	0.8517	0.406	0.3522	0.5585	-1.0	0.3576	0.5352	0.5352	0.4608	0.6358	-1.0	0.4352	0.5352
8.6742	45.0	11250	8.3417	0.4506	0.8556	0.4097	0.3673	0.5706	-1.0	0.3704	0.5442	0.5442	0.4688	0.6474	-1.0	0.4506	0.5442
8.4248	46.0	11500	8.6795	0.4539	0.8476	0.4225	0.3678	0.5773	-1.0	0.3673	0.5405	0.5405	0.4634	0.6467	-1.0	0.4539	0.5405
8.4248	47.0	11750	8.5291	0.4489	0.8542	0.4158	0.369	0.5717	-1.0	0.3657	0.5393	0.5393	0.457	0.6526	-1.0	0.4489	0.5393
8.2016	48.0	12000	8.6930	0.4403	0.8538	0.3723	0.359	0.5611	-1.0	0.3617	0.534	0.534	0.4559	0.6416	-1.0	0.4403	0.534
8.2016	49.0	12250	8.9282	0.4353	0.8426	0.3987	0.3473	0.5617	-1.0	0.3598	0.5283	0.5283	0.4484	0.638	-1.0	0.4353	0.5283
7.9596	50.0	12500	8.7151	0.4468	0.843	0.4203	0.3687	0.5631	-1.0	0.3651	0.5336	0.5336	0.4565	0.6409	-1.0	0.4468	0.5336
7.9596	51.0	12750	8.7857	0.4286	0.856	0.3588	0.3436	0.5499	-1.0	0.3598	0.5283	0.5283	0.4505	0.6358	-1.0	0.4286	0.5283
7.6242	52.0	13000	8.8183	0.4395	0.8525	0.3915	0.365	0.5561	-1.0	0.3654	0.5224	0.5224	0.4462	0.6277	-1.0	0.4395	0.5224
7.6242	53.0	13250	9.2425	0.4395	0.8583	0.4199	0.3478	0.5723	-1.0	0.3579	0.5252	0.5252	0.4349	0.6511	-1.0	0.4395	0.5252
7.6677	54.0	13500	9.3525	0.4385	0.853	0.4002	0.3536	0.5602	-1.0	0.3639	0.5364	0.5364	0.464	0.6372	-1.0	0.4385	0.5364
7.6677	55.0	13750	9.0027	0.4478	0.8567	0.425	0.371	0.5567	-1.0	0.3642	0.5396	0.5396	0.471	0.6358	-1.0	0.4478	0.5396
7.356	56.0	14000	8.9735	0.4522	0.8612	0.4025	0.377	0.5607	-1.0	0.3651	0.5414	0.5414	0.4731	0.6358	-1.0	0.4522	0.5414
7.356	57.0	14250	8.8415	0.4491	0.8518	0.4109	0.3668	0.5678	-1.0	0.3698	0.5408	0.5408	0.4656	0.6445	-1.0	0.4491	0.5408
7.2584	58.0	14500	9.4286	0.4313	0.8498	0.377	0.3511	0.5551	-1.0	0.3604	0.529	0.529	0.4538	0.6336	-1.0	0.4313	0.529
7.2584	59.0	14750	9.2132	0.4422	0.8393	0.4174	0.3602	0.5571	-1.0	0.3667	0.5299	0.5299	0.4575	0.6307	-1.0	0.4422	0.5299
6.9421	60.0	15000	9.2212	0.4417	0.8514	0.4049	0.3619	0.5526	-1.0	0.3626	0.5293	0.5293	0.4511	0.638	-1.0	0.4417	0.5293
6.9421	61.0	15250	9.4394	0.434	0.8406	0.3888	0.3466	0.5618	-1.0	0.3523	0.5268	0.5268	0.4441	0.6416	-1.0	0.434	0.5268
6.7683	62.0	15500	9.6471	0.4374	0.85	0.4249	0.3526	0.5622	-1.0	0.3589	0.5305	0.5305	0.4511	0.6401	-1.0	0.4374	0.5305
6.7683	63.0	15750	9.5563	0.4406	0.8554	0.3818	0.3653	0.5498	-1.0	0.3567	0.5361	0.5361	0.4602	0.6416	-1.0	0.4406	0.5361
6.5078	64.0	16000	9.5198	0.4489	0.8591	0.4068	0.3701	0.5606	-1.0	0.3682	0.5405	0.5405	0.4656	0.6445	-1.0	0.4489	0.5405
6.5078	65.0	16250	9.4543	0.4338	0.8453	0.3891	0.3466	0.5625	-1.0	0.362	0.529	0.529	0.4473	0.6416	-1.0	0.4338	0.529
6.2986	66.0	16500	9.1141	0.4458	0.8587	0.4144	0.3704	0.5595	-1.0	0.362	0.5393	0.5393	0.4688	0.6372	-1.0	0.4458	0.5393
6.2986	67.0	16750	9.7170	0.4333	0.834	0.3986	0.3511	0.5553	-1.0	0.3648	0.5336	0.5336	0.4608	0.635	-1.0	0.4333	0.5336
6.2156	68.0	17000	9.7496	0.4376	0.8386	0.4004	0.3546	0.5625	-1.0	0.3561	0.529	0.529	0.4516	0.6365	-1.0	0.4376	0.529
6.2156	69.0	17250	9.7007	0.4347	0.845	0.3513	0.3544	0.5488	-1.0	0.3502	0.5262	0.5262	0.4516	0.6299	-1.0	0.4347	0.5262
6.0353	70.0	17500	9.5200	0.4523	0.8565	0.4048	0.3742	0.5696	-1.0	0.3735	0.5421	0.5421	0.4677	0.6445	-1.0	0.4523	0.5421

Framework versions

Transformers 4.57.1
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.2

dariacuna/rtdetr-v2-r50-finetune-17

rtdetr-v2-r50-finetune-17

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Seed this model