rtdetr-v2-r18-final

This model is a fine-tuned version of PekingU/rtdetr_v2_r18vd on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 5.1520
Map: 0.3649
Map 50: 0.6288
Map 75: 0.4056
Map Small: 0.3424
Map Medium: 0.4806
Map Large: -1.0
Mar 1: 0.2595
Mar 10: 0.6168
Mar 100: 0.6197
Mar Small: 0.5986
Mar Medium: 0.6736
Mar Large: -1.0
Map Artemia: 0.3649
Mar 100 Artemia: 0.6197

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	8.9989	0.228	0.4228	0.2165	0.1636	0.3911	-1.0	0.2614	0.5234	0.6016	0.5059	0.7336	-1.0	0.228	0.6016
72.6973	2.0	500	5.6413	0.4119	0.743	0.4161	0.3218	0.5456	-1.0	0.3555	0.5785	0.638	0.5892	0.7058	-1.0	0.4119	0.638
72.6973	3.0	750	5.3121	0.4655	0.8136	0.4693	0.3578	0.5948	-1.0	0.3682	0.5944	0.6567	0.5952	0.7409	-1.0	0.4655	0.6567
9.0229	4.0	1000	5.3701	0.4621	0.8186	0.4747	0.3575	0.5968	-1.0	0.3763	0.6003	0.6667	0.6161	0.7365	-1.0	0.4621	0.6667
9.0229	5.0	1250	5.2021	0.4247	0.7752	0.4147	0.3491	0.5913	-1.0	0.3583	0.5794	0.6336	0.5672	0.7248	-1.0	0.4247	0.6336
8.0582	6.0	1500	5.3194	0.4533	0.8219	0.4402	0.3502	0.5857	-1.0	0.3717	0.5832	0.6296	0.5823	0.6949	-1.0	0.4533	0.6296
8.0582	7.0	1750	5.3094	0.4641	0.8506	0.4456	0.3662	0.5866	-1.0	0.3682	0.596	0.6355	0.5817	0.7095	-1.0	0.4641	0.6355
7.6102	8.0	2000	5.4058	0.4451	0.8295	0.4381	0.334	0.5937	-1.0	0.3595	0.5847	0.6181	0.5672	0.6883	-1.0	0.4451	0.6181
7.6102	9.0	2250	5.2928	0.4478	0.8336	0.4677	0.3488	0.578	-1.0	0.3698	0.5738	0.6009	0.5565	0.662	-1.0	0.4478	0.6009
7.1832	10.0	2500	5.4281	0.4439	0.8457	0.3989	0.3472	0.5703	-1.0	0.3642	0.5794	0.5872	0.5339	0.6606	-1.0	0.4439	0.5872
7.1832	11.0	2750	5.4312	0.4221	0.8145	0.3739	0.324	0.5697	-1.0	0.3436	0.5726	0.5879	0.5274	0.6715	-1.0	0.4221	0.5879
6.8644	12.0	3000	5.5109	0.394	0.7696	0.3671	0.2844	0.5654	-1.0	0.3483	0.581	0.5907	0.521	0.6869	-1.0	0.394	0.5907
6.8644	13.0	3250	5.6830	0.4131	0.7966	0.3826	0.3123	0.5689	-1.0	0.3511	0.5835	0.596	0.5317	0.6847	-1.0	0.4131	0.596
6.5527	14.0	3500	5.8084	0.3875	0.7528	0.3669	0.2836	0.5751	-1.0	0.3255	0.5885	0.5984	0.5344	0.6869	-1.0	0.3875	0.5984
6.5527	15.0	3750	5.6620	0.4144	0.8001	0.3633	0.3095	0.577	-1.0	0.3424	0.5763	0.5844	0.5199	0.6737	-1.0	0.4144	0.5844
6.2953	16.0	4000	5.4284	0.4235	0.8105	0.4036	0.315	0.5828	-1.0	0.3402	0.5885	0.5944	0.5301	0.6832	-1.0	0.4235	0.5944
6.2953	17.0	4250	5.6714	0.4139	0.7801	0.3975	0.3053	0.5843	-1.0	0.3421	0.5916	0.5963	0.5349	0.6803	-1.0	0.4139	0.5963
5.9900	18.0	4500	5.7959	0.385	0.7596	0.3489	0.2758	0.5676	-1.0	0.329	0.5607	0.5704	0.4984	0.6701	-1.0	0.385	0.5704
5.9900	19.0	4750	5.7535	0.3605	0.7034	0.3299	0.2486	0.5844	-1.0	0.3174	0.5826	0.5882	0.5167	0.6869	-1.0	0.3605	0.5882
5.7903	20.0	5000	5.8551	0.3678	0.721	0.3245	0.2667	0.5651	-1.0	0.315	0.5745	0.5819	0.5161	0.6723	-1.0	0.3678	0.5819
5.7903	21.0	5250	6.2028	0.332	0.6412	0.3023	0.2251	0.5798	-1.0	0.2879	0.5804	0.5869	0.5194	0.6803	-1.0	0.332	0.5869
5.6119	22.0	5500	5.9598	0.3706	0.7262	0.3434	0.2618	0.5735	-1.0	0.324	0.5785	0.5844	0.5199	0.673	-1.0	0.3706	0.5844
5.6119	23.0	5750	5.8399	0.3501	0.7011	0.307	0.2444	0.5676	-1.0	0.3009	0.5838	0.5963	0.5274	0.6905	-1.0	0.3501	0.5963
5.4305	24.0	6000	5.9999	0.3788	0.7311	0.3388	0.2779	0.569	-1.0	0.3159	0.5819	0.585	0.5172	0.6781	-1.0	0.3788	0.585
5.4305	25.0	6250	5.8356	0.3841	0.7491	0.3609	0.2784	0.581	-1.0	0.3224	0.5751	0.5785	0.4968	0.6905	-1.0	0.3841	0.5785
5.3127	26.0	6500	6.1741	0.3538	0.7	0.3133	0.2479	0.5654	-1.0	0.3037	0.5857	0.5907	0.5113	0.6993	-1.0	0.3538	0.5907
5.3127	27.0	6750	6.3570	0.3248	0.6443	0.2872	0.2163	0.5665	-1.0	0.2844	0.5791	0.5854	0.5108	0.6883	-1.0	0.3248	0.5854
5.2137	28.0	7000	6.2257	0.3318	0.6573	0.3058	0.2317	0.5425	-1.0	0.2903	0.5751	0.5801	0.5097	0.6759	-1.0	0.3318	0.5801
5.2137	29.0	7250	6.2910	0.3334	0.654	0.2933	0.2348	0.5384	-1.0	0.2844	0.5844	0.5875	0.5054	0.7	-1.0	0.3334	0.5875
5.0860	30.0	7500	6.1008	0.3375	0.6594	0.3187	0.2367	0.5652	-1.0	0.3	0.5841	0.5879	0.5167	0.6854	-1.0	0.3375	0.5879
5.0860	31.0	7750	6.3340	0.3316	0.653	0.3019	0.2346	0.554	-1.0	0.305	0.576	0.5785	0.4973	0.6898	-1.0	0.3316	0.5785
4.9273	32.0	8000	6.0959	0.3518	0.6769	0.3315	0.2565	0.5501	-1.0	0.3165	0.5757	0.5763	0.4946	0.6883	-1.0	0.3518	0.5763
4.9273	33.0	8250	6.5223	0.3222	0.6283	0.3047	0.2283	0.5602	-1.0	0.3006	0.5707	0.5717	0.4903	0.6832	-1.0	0.3222	0.5717
4.8827	34.0	8500	6.3633	0.3536	0.6949	0.317	0.2516	0.5544	-1.0	0.3072	0.5685	0.5695	0.4903	0.6788	-1.0	0.3536	0.5695
4.8827	35.0	8750	6.4564	0.3451	0.6825	0.3275	0.2408	0.5785	-1.0	0.3109	0.5704	0.5704	0.4925	0.6774	-1.0	0.3451	0.5704
4.7840	36.0	9000	6.2205	0.3759	0.7295	0.3699	0.2791	0.5761	-1.0	0.3146	0.5673	0.5676	0.4882	0.6766	-1.0	0.3759	0.5676
4.7840	37.0	9250	6.3559	0.3708	0.7232	0.3424	0.2675	0.5721	-1.0	0.3174	0.567	0.567	0.486	0.6781	-1.0	0.3708	0.567
4.7406	38.0	9500	6.4708	0.3514	0.6887	0.3342	0.2486	0.5716	-1.0	0.3065	0.5676	0.5676	0.4866	0.6788	-1.0	0.3514	0.5676
4.7406	39.0	9750	6.6405	0.3424	0.6761	0.3114	0.2454	0.5709	-1.0	0.3003	0.5704	0.5707	0.4898	0.681	-1.0	0.3424	0.5707
4.6415	40.0	10000	6.5777	0.3649	0.7113	0.3574	0.2701	0.5403	-1.0	0.3193	0.5617	0.5617	0.4806	0.6737	-1.0	0.3649	0.5617
4.6415	41.0	10250	6.5270	0.3746	0.7187	0.3611	0.2825	0.5608	-1.0	0.329	0.5592	0.5595	0.4801	0.6686	-1.0	0.3746	0.5595
4.5823	42.0	10500	6.6756	0.3592	0.7048	0.3313	0.2556	0.5752	-1.0	0.3084	0.5573	0.5573	0.472	0.6745	-1.0	0.3592	0.5573
4.5823	43.0	10750	6.5801	0.3677	0.7235	0.3558	0.2691	0.5721	-1.0	0.3093	0.5576	0.5576	0.4753	0.6708	-1.0	0.3677	0.5576
4.5193	44.0	11000	6.6640	0.359	0.7079	0.3316	0.2624	0.5587	-1.0	0.3121	0.5548	0.5548	0.4742	0.6657	-1.0	0.359	0.5548
4.5193	45.0	11250	6.6996	0.3581	0.7021	0.3505	0.2662	0.557	-1.0	0.3174	0.5545	0.5545	0.4737	0.6657	-1.0	0.3581	0.5545
4.4735	46.0	11500	6.9097	0.3577	0.695	0.3465	0.2586	0.5676	-1.0	0.3075	0.5607	0.5607	0.4833	0.6664	-1.0	0.3577	0.5607
4.4735	47.0	11750	7.1354	0.3481	0.6788	0.3259	0.2522	0.5459	-1.0	0.3031	0.5526	0.5526	0.4699	0.6672	-1.0	0.3481	0.5526
4.3304	48.0	12000	6.8864	0.3532	0.6787	0.3316	0.2554	0.5721	-1.0	0.305	0.5607	0.5607	0.4753	0.6788	-1.0	0.3532	0.5607
4.3304	49.0	12250	6.7685	0.3585	0.6898	0.3428	0.2648	0.5429	-1.0	0.3199	0.5639	0.5639	0.4796	0.6803	-1.0	0.3585	0.5639
4.3225	50.0	12500	6.8697	0.3479	0.6821	0.3215	0.2595	0.5256	-1.0	0.3009	0.5545	0.5545	0.472	0.6679	-1.0	0.3479	0.5545
4.3225	51.0	12750	7.0907	0.3495	0.6803	0.3423	0.2513	0.5584	-1.0	0.3065	0.5657	0.5657	0.4833	0.6796	-1.0	0.3495	0.5657
4.2635	52.0	13000	6.9579	0.3497	0.679	0.3346	0.2576	0.5464	-1.0	0.3121	0.5604	0.5604	0.4774	0.6737	-1.0	0.3497	0.5604
4.2635	53.0	13250	7.1185	0.3483	0.6727	0.3409	0.2575	0.5498	-1.0	0.3159	0.5667	0.5667	0.4844	0.6796	-1.0	0.3483	0.5667
4.2576	54.0	13500	7.3033	0.341	0.6652	0.3282	0.2444	0.5461	-1.0	0.3093	0.5573	0.5573	0.4731	0.673	-1.0	0.341	0.5573
4.2576	55.0	13750	7.0839	0.3505	0.6817	0.3322	0.2599	0.5345	-1.0	0.31	0.5573	0.5573	0.4753	0.6701	-1.0	0.3505	0.5573
4.1523	56.0	14000	7.0539	0.3471	0.675	0.3418	0.2546	0.553	-1.0	0.3106	0.5589	0.5589	0.4747	0.6745	-1.0	0.3471	0.5589
4.1523	57.0	14250	7.0991	0.3433	0.6653	0.3266	0.2494	0.5443	-1.0	0.3087	0.5558	0.5558	0.472	0.6708	-1.0	0.3433	0.5558
4.1903	58.0	14500	7.0660	0.3508	0.6808	0.3392	0.2578	0.5448	-1.0	0.3137	0.5601	0.5601	0.479	0.6715	-1.0	0.3508	0.5601
4.1903	59.0	14750	7.0526	0.3506	0.6822	0.3337	0.2568	0.542	-1.0	0.3153	0.5586	0.5586	0.4753	0.673	-1.0	0.3506	0.5586
4.0932	60.0	15000	6.8802	0.3532	0.6904	0.3185	0.2595	0.557	-1.0	0.3143	0.5567	0.5567	0.4731	0.6715	-1.0	0.3532	0.5567
4.0932	61.0	15250	7.2525	0.3526	0.6887	0.3278	0.2562	0.573	-1.0	0.3249	0.5576	0.5576	0.4758	0.6701	-1.0	0.3526	0.5576
4.0472	62.0	15500	7.2023	0.3383	0.6532	0.3216	0.247	0.5467	-1.0	0.3128	0.5639	0.5639	0.4785	0.681	-1.0	0.3383	0.5639
4.0472	63.0	15750	7.1173	0.3507	0.6841	0.3272	0.2597	0.5536	-1.0	0.3212	0.5548	0.5548	0.472	0.6686	-1.0	0.3507	0.5548
3.9790	64.0	16000	7.3621	0.3485	0.6693	0.3252	0.2531	0.5572	-1.0	0.3134	0.5586	0.5586	0.4715	0.6781	-1.0	0.3485	0.5586
3.9790	65.0	16250	7.3887	0.3515	0.6802	0.3308	0.2542	0.5548	-1.0	0.3168	0.557	0.557	0.4715	0.6745	-1.0	0.3515	0.557
3.9523	66.0	16500	7.3179	0.346	0.6762	0.3259	0.25	0.5589	-1.0	0.3153	0.557	0.557	0.4699	0.6766	-1.0	0.346	0.557
3.9523	67.0	16750	7.2648	0.3488	0.6755	0.3294	0.2598	0.547	-1.0	0.3093	0.5589	0.5589	0.4731	0.6766	-1.0	0.3488	0.5589
3.9455	68.0	17000	7.2954	0.3446	0.6701	0.3356	0.2597	0.5395	-1.0	0.3174	0.557	0.557	0.4774	0.6664	-1.0	0.3446	0.557
3.9455	69.0	17250	7.3401	0.34	0.65	0.3316	0.2502	0.5492	-1.0	0.3159	0.557	0.557	0.4688	0.6781	-1.0	0.34	0.557
3.8855	70.0	17500	7.3309	0.3465	0.6696	0.3306	0.2598	0.5405	-1.0	0.3159	0.5583	0.5583	0.4715	0.6774	-1.0	0.3465	0.5583
3.8855	71.0	17750	7.4614	0.3406	0.6613	0.3185	0.2506	0.54	-1.0	0.3196	0.5589	0.5589	0.4737	0.6759	-1.0	0.3406	0.5589
3.8991	72.0	18000	7.4174	0.3497	0.6752	0.3338	0.2591	0.5487	-1.0	0.3131	0.5604	0.5604	0.4742	0.6788	-1.0	0.3497	0.5604
3.8991	73.0	18250	7.4782	0.3488	0.6752	0.3245	0.2555	0.5506	-1.0	0.3125	0.5589	0.5589	0.4742	0.6752	-1.0	0.3488	0.5589
3.8272	74.0	18500	7.4742	0.3458	0.673	0.3239	0.2564	0.5381	-1.0	0.3109	0.557	0.557	0.472	0.6737	-1.0	0.3458	0.557
3.8272	75.0	18750	7.4949	0.3437	0.6663	0.3332	0.253	0.5484	-1.0	0.3128	0.5586	0.5586	0.4742	0.6745	-1.0	0.3437	0.5586
3.8122	76.0	19000	7.4930	0.3452	0.6709	0.3267	0.2546	0.5421	-1.0	0.3178	0.5576	0.5576	0.472	0.6752	-1.0	0.3452	0.5576
3.8122	77.0	19250	7.6538	0.345	0.6682	0.3331	0.2564	0.5412	-1.0	0.3131	0.5589	0.5589	0.4726	0.6774	-1.0	0.345	0.5589
3.7544	78.0	19500	7.4972	0.3429	0.6639	0.3249	0.2529	0.5446	-1.0	0.3115	0.5579	0.5579	0.472	0.6759	-1.0	0.3429	0.5579
3.7544	79.0	19750	7.5362	0.3457	0.6749	0.3284	0.2562	0.5484	-1.0	0.3125	0.5573	0.5573	0.472	0.6745	-1.0	0.3457	0.5573
3.7789	80.0	20000	7.5626	0.3437	0.6659	0.3316	0.2542	0.5455	-1.0	0.3171	0.5573	0.5573	0.4726	0.6737	-1.0	0.3437	0.5573

Framework versions

Transformers 5.9.0
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.2

dariacuna/rtdetr-v2-r18-final

rtdetr-v2-r18-final

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions