rtdetr-v2-r101-finetune

This model is a fine-tuned version of PekingU/rtdetr_v2_r101vd on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 6.1243
Map: 0.559
Map 50: 0.8915
Map 75: 0.6353
Map Small: 0.5298
Map Medium: 0.6441
Map Large: -1.0
Mar 1: 0.3369
Mar 10: 0.6612
Mar 100: 0.6981
Mar Small: 0.6784
Mar Medium: 0.7483
Mar Large: -1.0
Map Artemia: 0.559
Mar 100 Artemia: 0.6981

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	11.5036	0.4546	0.8448	0.4307	0.3738	0.5719	-1.0	0.3486	0.5632	0.6274	0.5543	0.7292	-1.0	0.4546	0.6274
128.8923	2.0	500	8.1984	0.4766	0.8663	0.4534	0.4074	0.5828	-1.0	0.3664	0.5938	0.6383	0.5839	0.7131	-1.0	0.4766	0.6383
128.8923	3.0	750	8.2372	0.4717	0.8718	0.4354	0.3835	0.5984	-1.0	0.3713	0.5782	0.6508	0.5914	0.7336	-1.0	0.4717	0.6508
14.8313	4.0	1000	7.9904	0.4705	0.8427	0.4589	0.3927	0.5987	-1.0	0.3679	0.5704	0.6486	0.5989	0.7182	-1.0	0.4705	0.6486
14.8313	5.0	1250	7.8513	0.4733	0.8529	0.4731	0.3936	0.5991	-1.0	0.3726	0.5891	0.6383	0.5817	0.7161	-1.0	0.4733	0.6383
13.6236	6.0	1500	7.7301	0.4847	0.8866	0.4591	0.4082	0.6069	-1.0	0.3791	0.5947	0.6458	0.5871	0.7277	-1.0	0.4847	0.6458
13.6236	7.0	1750	8.0222	0.4615	0.8531	0.4617	0.3995	0.55	-1.0	0.3654	0.5642	0.5882	0.5452	0.6482	-1.0	0.4615	0.5882
13.3225	8.0	2000	8.0110	0.4525	0.8394	0.4626	0.3733	0.583	-1.0	0.3558	0.5763	0.6262	0.5618	0.7146	-1.0	0.4525	0.6262
13.3225	9.0	2250	7.8077	0.4734	0.8719	0.4365	0.4085	0.5754	-1.0	0.3773	0.5741	0.5972	0.5446	0.6701	-1.0	0.4734	0.5972
12.8163	10.0	2500	8.1337	0.4723	0.8717	0.4703	0.3961	0.5966	-1.0	0.3819	0.5857	0.6209	0.5661	0.6964	-1.0	0.4723	0.6209
12.8163	11.0	2750	8.0850	0.4809	0.8822	0.4314	0.4071	0.5994	-1.0	0.3822	0.5766	0.6103	0.5446	0.7015	-1.0	0.4809	0.6103
12.6997	12.0	3000	8.0487	0.4524	0.8616	0.4011	0.3702	0.58	-1.0	0.3651	0.5639	0.5916	0.5199	0.6905	-1.0	0.4524	0.5916
12.6997	13.0	3250	8.0089	0.4687	0.8587	0.4545	0.3917	0.5948	-1.0	0.3723	0.5676	0.586	0.5108	0.6898	-1.0	0.4687	0.586
12.3637	14.0	3500	8.0282	0.4711	0.8638	0.4583	0.4074	0.5731	-1.0	0.385	0.5748	0.5769	0.5199	0.6562	-1.0	0.4711	0.5769
12.3637	15.0	3750	8.3123	0.4525	0.8575	0.3928	0.3785	0.5778	-1.0	0.3673	0.5726	0.5847	0.5274	0.665	-1.0	0.4525	0.5847
12.0121	16.0	4000	8.5927	0.4568	0.8508	0.4479	0.381	0.574	-1.0	0.376	0.5738	0.5769	0.5065	0.6745	-1.0	0.4568	0.5769
12.0121	17.0	4250	8.2875	0.4665	0.8724	0.4265	0.3949	0.5766	-1.0	0.3685	0.5782	0.586	0.5199	0.6766	-1.0	0.4665	0.586
11.6483	18.0	4500	8.2437	0.4558	0.831	0.4556	0.3857	0.5654	-1.0	0.3667	0.5567	0.5598	0.4968	0.6474	-1.0	0.4558	0.5598
11.6483	19.0	4750	8.4029	0.441	0.8304	0.4202	0.3675	0.5833	-1.0	0.3629	0.586	0.605	0.5495	0.6818	-1.0	0.441	0.605
11.4112	20.0	5000	8.2684	0.4565	0.8483	0.4196	0.3849	0.572	-1.0	0.3741	0.5729	0.5819	0.5156	0.6737	-1.0	0.4565	0.5819
11.4112	21.0	5250	8.4136	0.474	0.8662	0.4467	0.4083	0.5734	-1.0	0.3822	0.5713	0.5723	0.507	0.6628	-1.0	0.474	0.5723
11.1786	22.0	5500	8.4458	0.4636	0.8684	0.438	0.3863	0.5886	-1.0	0.3639	0.5919	0.5991	0.5258	0.7	-1.0	0.4636	0.5991
11.1786	23.0	5750	8.5703	0.4458	0.8422	0.4095	0.3715	0.5655	-1.0	0.3573	0.5558	0.5592	0.4839	0.6628	-1.0	0.4458	0.5592
10.6932	24.0	6000	8.4739	0.4382	0.834	0.4036	0.3767	0.5346	-1.0	0.3614	0.5455	0.547	0.4849	0.6321	-1.0	0.4382	0.547
10.6932	25.0	6250	8.4428	0.4434	0.8321	0.4481	0.3627	0.5742	-1.0	0.3551	0.5402	0.5408	0.4597	0.6533	-1.0	0.4434	0.5408
10.7234	26.0	6500	8.3418	0.4559	0.8699	0.4276	0.3793	0.5777	-1.0	0.3704	0.5701	0.572	0.4978	0.6752	-1.0	0.4559	0.572
10.7234	27.0	6750	8.6772	0.4347	0.847	0.4315	0.359	0.5678	-1.0	0.3639	0.5436	0.5439	0.4731	0.6409	-1.0	0.4347	0.5439
10.3517	28.0	7000	8.6828	0.4335	0.8203	0.4182	0.3576	0.5615	-1.0	0.3601	0.5489	0.5502	0.4747	0.654	-1.0	0.4335	0.5502
10.3517	29.0	7250	8.4278	0.427	0.8348	0.3879	0.3474	0.5681	-1.0	0.3586	0.5573	0.5579	0.4833	0.6613	-1.0	0.427	0.5579
10.1122	30.0	7500	8.3581	0.4472	0.8396	0.4156	0.3731	0.5702	-1.0	0.3745	0.5561	0.5564	0.4871	0.6526	-1.0	0.4472	0.5564
10.1122	31.0	7750	8.7164	0.4165	0.7827	0.3993	0.3403	0.5655	-1.0	0.3651	0.5583	0.5601	0.4973	0.6474	-1.0	0.4165	0.5601
9.8785	32.0	8000	8.7283	0.4338	0.828	0.3783	0.3601	0.5513	-1.0	0.3611	0.5268	0.5268	0.4468	0.6387	-1.0	0.4338	0.5268
9.8785	33.0	8250	8.8290	0.444	0.848	0.3973	0.3659	0.5652	-1.0	0.353	0.5411	0.5411	0.4677	0.6423	-1.0	0.444	0.5411
9.6861	34.0	8500	9.1182	0.4247	0.8164	0.3691	0.3505	0.557	-1.0	0.3586	0.5455	0.5455	0.4806	0.6358	-1.0	0.4247	0.5455
9.6861	35.0	8750	9.1736	0.4079	0.7757	0.3848	0.3224	0.5554	-1.0	0.3511	0.5346	0.5346	0.4661	0.6299	-1.0	0.4079	0.5346
9.4407	36.0	9000	8.9869	0.4325	0.8164	0.3957	0.354	0.5748	-1.0	0.3604	0.543	0.543	0.4688	0.6467	-1.0	0.4325	0.543
9.4407	37.0	9250	9.2346	0.4185	0.7956	0.3695	0.3396	0.5604	-1.0	0.3583	0.5455	0.5455	0.4758	0.6431	-1.0	0.4185	0.5455
9.2431	38.0	9500	8.9343	0.4163	0.7939	0.3635	0.342	0.5567	-1.0	0.3467	0.5455	0.5455	0.4833	0.6314	-1.0	0.4163	0.5455
9.2431	39.0	9750	9.3849	0.4081	0.8006	0.3432	0.3279	0.5524	-1.0	0.3523	0.5414	0.5414	0.4785	0.6285	-1.0	0.4081	0.5414
8.963	40.0	10000	9.3751	0.3997	0.787	0.3486	0.309	0.5593	-1.0	0.3445	0.5259	0.5259	0.4446	0.6387	-1.0	0.3997	0.5259
8.963	41.0	10250	9.2140	0.4129	0.8007	0.3811	0.3226	0.5801	-1.0	0.3548	0.5408	0.5408	0.4645	0.6474	-1.0	0.4129	0.5408
8.7901	42.0	10500	9.3524	0.406	0.7675	0.3532	0.3109	0.5767	-1.0	0.347	0.538	0.538	0.4575	0.6489	-1.0	0.406	0.538
8.7901	43.0	10750	9.5249	0.407	0.7609	0.411	0.3211	0.5589	-1.0	0.3417	0.534	0.534	0.4634	0.6328	-1.0	0.407	0.534
8.6147	44.0	11000	9.2849	0.4177	0.7914	0.3959	0.3273	0.5672	-1.0	0.3536	0.5355	0.5355	0.4565	0.6467	-1.0	0.4177	0.5355
8.6147	45.0	11250	9.2095	0.4204	0.811	0.3977	0.3298	0.5841	-1.0	0.3492	0.5399	0.5399	0.4618	0.6482	-1.0	0.4204	0.5399
8.3995	46.0	11500	9.3734	0.4143	0.7975	0.4042	0.3362	0.5603	-1.0	0.3514	0.5402	0.5402	0.4753	0.6314	-1.0	0.4143	0.5402
8.3995	47.0	11750	9.2753	0.4057	0.773	0.3817	0.3224	0.5651	-1.0	0.3495	0.5371	0.5371	0.4672	0.6343	-1.0	0.4057	0.5371
8.1402	48.0	12000	9.4861	0.4091	0.7768	0.3839	0.3241	0.5644	-1.0	0.3436	0.5321	0.5321	0.4597	0.6328	-1.0	0.4091	0.5321
8.1402	49.0	12250	9.6633	0.3955	0.7634	0.3763	0.296	0.5787	-1.0	0.334	0.5324	0.5324	0.4527	0.6445	-1.0	0.3955	0.5324
7.8936	50.0	12500	9.7859	0.4068	0.778	0.3728	0.3163	0.5641	-1.0	0.3436	0.5333	0.5333	0.4624	0.6336	-1.0	0.4068	0.5333
7.8936	51.0	12750	9.7867	0.3813	0.7626	0.3264	0.2904	0.5552	-1.0	0.329	0.5199	0.5199	0.4409	0.6299	-1.0	0.3813	0.5199
7.7308	52.0	13000	9.5287	0.4075	0.7775	0.372	0.3177	0.5723	-1.0	0.3364	0.5396	0.5396	0.4672	0.6394	-1.0	0.4075	0.5396
7.7308	53.0	13250	9.5851	0.4113	0.7893	0.3899	0.3206	0.5743	-1.0	0.3461	0.533	0.533	0.4575	0.6387	-1.0	0.4113	0.533
7.5921	54.0	13500	9.6527	0.3892	0.7664	0.3222	0.3011	0.5486	-1.0	0.3343	0.5252	0.5252	0.4575	0.6197	-1.0	0.3892	0.5252
7.5921	55.0	13750	9.6497	0.4004	0.759	0.4043	0.305	0.5722	-1.0	0.3336	0.528	0.528	0.4538	0.6314	-1.0	0.4004	0.528
7.3406	56.0	14000	9.7962	0.3896	0.7405	0.3538	0.289	0.5799	-1.0	0.3399	0.5308	0.5308	0.4505	0.6423	-1.0	0.3896	0.5308
7.3406	57.0	14250	9.4300	0.4116	0.7865	0.3775	0.3244	0.5654	-1.0	0.3508	0.5336	0.5336	0.4608	0.6358	-1.0	0.4116	0.5336
7.3323	58.0	14500	9.9767	0.3856	0.7387	0.3443	0.2882	0.5598	-1.0	0.3315	0.5265	0.5265	0.4495	0.6336	-1.0	0.3856	0.5265
7.3323	59.0	14750	9.9109	0.3921	0.7598	0.3285	0.2957	0.5703	-1.0	0.3452	0.5296	0.5296	0.4522	0.6372	-1.0	0.3921	0.5296
7.0145	60.0	15000	9.7573	0.3929	0.7588	0.3479	0.3019	0.5586	-1.0	0.3442	0.533	0.533	0.464	0.6299	-1.0	0.3929	0.533
7.0145	61.0	15250	9.7050	0.4006	0.7712	0.3734	0.3102	0.5593	-1.0	0.3424	0.534	0.534	0.4645	0.6307	-1.0	0.4006	0.534
6.8638	62.0	15500	9.9508	0.3926	0.7398	0.3596	0.2911	0.5716	-1.0	0.338	0.5299	0.5299	0.4478	0.6445	-1.0	0.3926	0.5299
6.8638	63.0	15750	10.0933	0.3884	0.765	0.3178	0.2968	0.5445	-1.0	0.3383	0.5212	0.5212	0.4505	0.6212	-1.0	0.3884	0.5212
6.6495	64.0	16000	10.1369	0.3878	0.7489	0.3541	0.2934	0.5654	-1.0	0.3383	0.5249	0.5249	0.4522	0.6263	-1.0	0.3878	0.5249
6.6495	65.0	16250	10.2357	0.3915	0.7451	0.3369	0.2962	0.5626	-1.0	0.3324	0.5333	0.5333	0.4624	0.6314	-1.0	0.3915	0.5333
6.3762	66.0	16500	10.4613	0.3801	0.7348	0.3546	0.282	0.5619	-1.0	0.3355	0.5259	0.5259	0.4495	0.6321	-1.0	0.3801	0.5259
6.3762	67.0	16750	10.2216	0.3958	0.7422	0.3551	0.2955	0.5843	-1.0	0.3411	0.5417	0.5417	0.4683	0.6445	-1.0	0.3958	0.5417
6.3108	68.0	17000	10.1687	0.3878	0.7606	0.3432	0.2965	0.5603	-1.0	0.3361	0.5299	0.5299	0.4597	0.6285	-1.0	0.3878	0.5299
6.3108	69.0	17250	10.1483	0.379	0.7313	0.3382	0.2826	0.5523	-1.0	0.3336	0.5193	0.5193	0.4425	0.627	-1.0	0.379	0.5193
6.1575	70.0	17500	10.4216	0.3842	0.7416	0.3403	0.2892	0.5807	-1.0	0.3402	0.5315	0.5315	0.4516	0.6438	-1.0	0.3842	0.5315

Framework versions

Transformers 4.57.1
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.2

dariacuna/rtdetr-v2-r101-finetune

rtdetr-v2-r101-finetune

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

为此模型做种