rtdetr-v2-r50-finetune-7

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: 6.5296
Map: 0.5458
Map 50: 0.8587
Map 75: 0.6348
Map Small: 0.5149
Map Medium: 0.6432
Map Large: -1.0
Mar 1: 0.3327
Mar 10: 0.6605
Mar 100: 0.6945
Mar Small: 0.6694
Mar Medium: 0.7586
Mar Large: -1.0

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: 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
No log	1.0	250	11.5570	0.184	0.3705	0.1554	0.1368	0.3388	-1.0	0.2751	0.5801	0.6215	0.5771	0.6747	-1.0
81.2435	2.0	500	8.9895	0.3941	0.7086	0.3573	0.2889	0.5864	-1.0	0.3526	0.5941	0.6368	0.5749	0.711	-1.0
81.2435	3.0	750	9.6576	0.4588	0.7902	0.4848	0.3422	0.5957	-1.0	0.3745	0.5972	0.6405	0.6069	0.6808	-1.0
13.6125	4.0	1000	8.1103	0.4774	0.8568	0.4705	0.3786	0.5943	-1.0	0.3766	0.6081	0.6442	0.588	0.7116	-1.0
13.6125	5.0	1250	8.6176	0.4756	0.8512	0.4605	0.3828	0.5948	-1.0	0.3676	0.5919	0.6227	0.5829	0.6705	-1.0
11.9762	6.0	1500	8.4376	0.415	0.7818	0.3827	0.3141	0.5869	-1.0	0.3439	0.5894	0.6059	0.5503	0.6726	-1.0
11.9762	7.0	1750	8.1101	0.4434	0.829	0.3922	0.3358	0.5865	-1.0	0.3614	0.5882	0.6034	0.5217	0.7014	-1.0
11.2525	8.0	2000	8.5625	0.4585	0.8576	0.4609	0.3641	0.5907	-1.0	0.3757	0.5798	0.5835	0.5166	0.6637	-1.0
11.2525	9.0	2250	8.6193	0.4727	0.8545	0.4486	0.3772	0.5921	-1.0	0.3826	0.5869	0.5928	0.524	0.6753	-1.0
10.3846	10.0	2500	8.6932	0.442	0.8173	0.4168	0.339	0.5896	-1.0	0.366	0.566	0.5673	0.48	0.6719	-1.0
10.3846	11.0	2750	8.6027	0.4634	0.8358	0.4466	0.3666	0.5955	-1.0	0.3732	0.5788	0.581	0.4971	0.6815	-1.0
9.7743	12.0	3000	8.6512	0.4259	0.7972	0.3862	0.3198	0.5837	-1.0	0.3576	0.5604	0.5623	0.4634	0.6808	-1.0
9.7743	13.0	3250	8.4895	0.4321	0.8105	0.3998	0.3352	0.5846	-1.0	0.3592	0.5698	0.5704	0.4823	0.676	-1.0
9.2872	14.0	3500	8.9649	0.4047	0.7475	0.3677	0.3016	0.5894	-1.0	0.3302	0.5692	0.571	0.4874	0.6712	-1.0
9.2872	15.0	3750	9.4081	0.4205	0.7769	0.3962	0.3118	0.5923	-1.0	0.3564	0.5598	0.5598	0.4611	0.6781	-1.0
8.7304	16.0	4000	9.3530	0.4086	0.7617	0.3801	0.3025	0.5816	-1.0	0.3483	0.5567	0.5567	0.4571	0.676	-1.0
8.7304	17.0	4250	9.4145	0.4182	0.759	0.4053	0.3126	0.5984	-1.0	0.3523	0.5732	0.5732	0.4817	0.6829	-1.0
8.2523	18.0	4500	8.9666	0.4476	0.8112	0.4353	0.3465	0.5948	-1.0	0.3561	0.5604	0.5604	0.4617	0.6788	-1.0
8.2523	19.0	4750	9.4701	0.4186	0.7807	0.3821	0.3169	0.5859	-1.0	0.3536	0.5548	0.5548	0.4606	0.6678	-1.0
7.8352	20.0	5000	9.3753	0.4165	0.7674	0.3872	0.3099	0.5884	-1.0	0.3461	0.5545	0.5545	0.4606	0.6671	-1.0
7.8352	21.0	5250	9.4054	0.4361	0.7974	0.4301	0.3236	0.595	-1.0	0.3533	0.5551	0.5567	0.4549	0.6788	-1.0
7.5325	22.0	5500	9.7138	0.3917	0.7332	0.376	0.2826	0.5904	-1.0	0.3368	0.5601	0.5601	0.4731	0.6644	-1.0
7.5325	23.0	5750	10.1058	0.4169	0.7659	0.3716	0.3156	0.5948	-1.0	0.3555	0.5598	0.5598	0.4669	0.6712	-1.0
7.1113	24.0	6000	10.1551	0.3975	0.7235	0.3668	0.2925	0.5919	-1.0	0.352	0.5505	0.5505	0.452	0.6685	-1.0
7.1113	25.0	6250	9.6439	0.3913	0.7279	0.3585	0.2907	0.5762	-1.0	0.3467	0.5508	0.5508	0.4549	0.6658	-1.0
6.8666	26.0	6500	9.2025	0.426	0.7911	0.3941	0.3236	0.5765	-1.0	0.353	0.5508	0.5508	0.4503	0.6712	-1.0
6.8666	27.0	6750	9.6123	0.3811	0.703	0.3552	0.3003	0.5798	-1.0	0.3511	0.5542	0.5542	0.4583	0.6692	-1.0
6.6357	28.0	7000	10.1534	0.4004	0.7451	0.3626	0.2976	0.5855	-1.0	0.3439	0.5551	0.5551	0.4709	0.6562	-1.0
6.6357	29.0	7250	9.7884	0.4025	0.7353	0.3638	0.2945	0.5969	-1.0	0.348	0.5576	0.5576	0.46	0.6747	-1.0
6.3624	30.0	7500	10.2677	0.3828	0.7148	0.3456	0.276	0.5887	-1.0	0.3405	0.5539	0.5539	0.4606	0.6658	-1.0
6.3624	31.0	7750	10.3460	0.3853	0.7105	0.3597	0.2804	0.5987	-1.0	0.3439	0.5551	0.5551	0.4537	0.6767	-1.0
6.0826	32.0	8000	10.6511	0.3631	0.6621	0.3393	0.264	0.5991	-1.0	0.3349	0.5542	0.5542	0.452	0.6767	-1.0
6.0826	33.0	8250	10.9861	0.3509	0.641	0.3258	0.2433	0.5956	-1.0	0.319	0.5579	0.5579	0.4646	0.6699	-1.0
5.8633	34.0	8500	11.6602	0.3496	0.6435	0.3312	0.2468	0.5884	-1.0	0.3324	0.5595	0.5595	0.468	0.6692	-1.0
5.8633	35.0	8750	10.9800	0.3599	0.6543	0.3306	0.2538	0.595	-1.0	0.3355	0.5636	0.5636	0.4714	0.674	-1.0
5.6808	36.0	9000	11.2324	0.3532	0.6504	0.3134	0.2606	0.5924	-1.0	0.3355	0.5586	0.5586	0.4703	0.6644	-1.0
5.6808	37.0	9250	10.9914	0.3744	0.6932	0.3488	0.254	0.588	-1.0	0.3312	0.5595	0.5595	0.4743	0.6616	-1.0
5.4889	38.0	9500	12.3386	0.3418	0.6337	0.3148	0.2353	0.5933	-1.0	0.3299	0.5589	0.5589	0.4657	0.6705	-1.0
5.4889	39.0	9750	11.2158	0.3689	0.6776	0.3421	0.2568	0.5907	-1.0	0.3321	0.5586	0.5586	0.4691	0.6658	-1.0
5.1509	40.0	10000	11.2599	0.3543	0.6533	0.3345	0.2475	0.5934	-1.0	0.3389	0.5555	0.5555	0.4617	0.6678	-1.0
5.1509	41.0	10250	11.4931	0.3771	0.6889	0.3541	0.2553	0.5998	-1.0	0.3474	0.5573	0.5573	0.4606	0.6733	-1.0
4.96	42.0	10500	11.2832	0.3536	0.6514	0.3289	0.2458	0.5951	-1.0	0.3327	0.5526	0.5526	0.456	0.6685	-1.0
4.96	43.0	10750	11.8294	0.3575	0.6548	0.3334	0.2457	0.592	-1.0	0.3393	0.5533	0.5533	0.4571	0.6685	-1.0
4.8151	44.0	11000	12.3141	0.369	0.6787	0.3409	0.2463	0.5934	-1.0	0.3408	0.5545	0.5545	0.4629	0.6644	-1.0
4.8151	45.0	11250	12.4884	0.3258	0.6021	0.3026	0.2452	0.588	-1.0	0.3417	0.5564	0.5564	0.4594	0.6726	-1.0
4.6594	46.0	11500	11.9297	0.3728	0.6831	0.3498	0.2534	0.5955	-1.0	0.3405	0.5555	0.5555	0.4594	0.6705	-1.0
4.6594	47.0	11750	12.6762	0.3457	0.6309	0.3203	0.2365	0.5891	-1.0	0.3439	0.5555	0.5555	0.4617	0.6678	-1.0
4.4425	48.0	12000	12.5025	0.3651	0.6688	0.3352	0.2425	0.5911	-1.0	0.3417	0.5551	0.5551	0.4594	0.6699	-1.0
4.4425	49.0	12250	12.3912	0.349	0.6383	0.3293	0.2412	0.5923	-1.0	0.3417	0.5555	0.5555	0.4606	0.6692	-1.0
4.3247	50.0	12500	12.4655	0.3542	0.6433	0.3326	0.2416	0.5901	-1.0	0.3377	0.5573	0.5573	0.4646	0.6685	-1.0

Framework versions

Transformers 4.57.1
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.1

dariacuna/rtdetr-v2-r50-finetune-7

rtdetr-v2-r50-finetune-7

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

为此模型做种