detr_finetuned_kitti_mots-bright

This model is a fine-tuned version of microsoft/conditional-detr-resnet-50 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6703
• Map: 0.4803
• Map 50: 0.7799
• Map 75: 0.494
• Map Small: 0.2239
• Map Medium: 0.4794
• Map Large: 0.7442
• Mar 1: 0.1586
• Mar 10: 0.5224
• Mar 100: 0.6196
• Mar Small: 0.4422
• Mar Medium: 0.6281
• Mar Large: 0.8123
• Map Car: 0.6182
• Mar 100 Car: 0.7081
• Map Pedestrian: 0.3423
• Mar 100 Pedestrian: 0.5311

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• num_epochs: 60

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 Car	Mar 100 Car	Map Pedestrian	Mar 100 Pedestrian
0.9436	1.0	743	0.8850	0.3518	0.6491	0.3381	0.1025	0.3437	0.6351	0.1336	0.4063	0.5187	0.3121	0.5288	0.7428	0.4742	0.5947	0.2293	0.4428
0.8864	2.0	1486	0.8514	0.3714	0.6685	0.3602	0.1196	0.3636	0.6536	0.1384	0.4226	0.5403	0.3347	0.5536	0.7549	0.5002	0.6203	0.2426	0.4603
0.8577	3.0	2229	0.8346	0.3815	0.6809	0.3741	0.1251	0.3727	0.6703	0.1428	0.4299	0.5424	0.3401	0.5521	0.7616	0.5075	0.6206	0.2555	0.4642
0.8454	4.0	2972	0.8225	0.3872	0.6848	0.3845	0.1313	0.3804	0.671	0.142	0.4376	0.5496	0.3441	0.5631	0.763	0.5173	0.6291	0.2571	0.47
0.8258	5.0	3715	0.8161	0.3929	0.6901	0.3893	0.1383	0.3859	0.6733	0.1434	0.4402	0.5486	0.3469	0.561	0.7605	0.5246	0.6332	0.2613	0.464
0.8186	6.0	4458	0.8088	0.3963	0.6967	0.3967	0.1404	0.3882	0.6784	0.1441	0.4437	0.5516	0.3549	0.5603	0.7673	0.5232	0.6351	0.2695	0.4681
0.8058	7.0	5201	0.7959	0.4054	0.7096	0.4078	0.1448	0.4036	0.6818	0.1444	0.4565	0.5626	0.3653	0.5749	0.7706	0.5338	0.6433	0.2771	0.4819
0.7957	8.0	5944	0.7860	0.4115	0.7072	0.4119	0.1455	0.4088	0.6917	0.1469	0.4561	0.5666	0.3603	0.5817	0.7772	0.543	0.6507	0.28	0.4825
0.7842	9.0	6687	0.7766	0.4155	0.7136	0.4165	0.1508	0.4105	0.6974	0.1479	0.459	0.5711	0.3734	0.5819	0.7826	0.547	0.6543	0.2841	0.4879
0.7814	10.0	7430	0.7803	0.4128	0.7122	0.4133	0.1481	0.41	0.6995	0.1467	0.4578	0.5646	0.3611	0.577	0.7794	0.5464	0.6516	0.2793	0.4776
0.771	11.0	8173	0.7707	0.4145	0.7134	0.4199	0.1524	0.4116	0.696	0.1459	0.4624	0.5691	0.3691	0.5823	0.7773	0.5474	0.6548	0.2816	0.4834
0.7674	12.0	8916	0.7643	0.4228	0.7207	0.4258	0.1581	0.4223	0.6966	0.1477	0.4668	0.5762	0.3753	0.5912	0.7799	0.5566	0.6607	0.289	0.4916
0.7584	13.0	9659	0.7559	0.4246	0.7265	0.4316	0.1629	0.4213	0.7056	0.1472	0.4698	0.5772	0.3817	0.5872	0.7875	0.5574	0.6611	0.2919	0.4933
0.7516	14.0	10402	0.7555	0.4269	0.7277	0.4334	0.1615	0.4246	0.7072	0.1481	0.4716	0.5787	0.3765	0.5922	0.7879	0.5619	0.6641	0.2919	0.4933
0.7456	15.0	11145	0.7448	0.4328	0.7333	0.4432	0.1679	0.4316	0.7122	0.1507	0.4785	0.5819	0.3883	0.5916	0.7915	0.5686	0.6714	0.297	0.4923
0.7404	16.0	11888	0.7475	0.4336	0.7359	0.4363	0.1668	0.4304	0.7181	0.1501	0.4798	0.579	0.3815	0.5885	0.7936	0.5668	0.6671	0.3005	0.4909
0.7404	17.0	12631	0.7393	0.4361	0.7386	0.4473	0.1752	0.4337	0.7128	0.1509	0.4821	0.583	0.3997	0.5905	0.787	0.5745	0.6745	0.2978	0.4915
0.7279	18.0	13374	0.7331	0.4384	0.7406	0.4417	0.1771	0.4376	0.7105	0.1508	0.4833	0.5879	0.3965	0.5984	0.7926	0.578	0.6775	0.2988	0.4982
0.722	19.0	14117	0.7329	0.4407	0.742	0.4443	0.1834	0.4393	0.7138	0.1505	0.4866	0.5898	0.4042	0.5995	0.7887	0.5748	0.6751	0.3067	0.5044
0.7184	20.0	14860	0.7240	0.4484	0.748	0.4557	0.1872	0.4476	0.7186	0.153	0.4915	0.5943	0.4045	0.6062	0.7926	0.5846	0.6819	0.3121	0.5066
0.7177	21.0	15603	0.7266	0.4447	0.75	0.4517	0.1856	0.4419	0.7154	0.1515	0.4883	0.5893	0.4061	0.5988	0.7866	0.58	0.6777	0.3095	0.5009
0.7077	22.0	16346	0.7172	0.4496	0.752	0.4618	0.1861	0.4486	0.7199	0.1524	0.4921	0.5935	0.4065	0.6031	0.7946	0.5856	0.6812	0.3137	0.5057
0.7073	23.0	17089	0.7199	0.4471	0.7489	0.4598	0.1882	0.4443	0.7203	0.1518	0.4898	0.5944	0.4094	0.6039	0.7936	0.5819	0.6807	0.3123	0.5081
0.7043	24.0	17832	0.7139	0.4525	0.7506	0.4618	0.1893	0.4508	0.7258	0.1542	0.4964	0.5994	0.4122	0.6084	0.8026	0.589	0.6827	0.316	0.516
0.6988	25.0	18575	0.7132	0.4527	0.7543	0.4627	0.19	0.4498	0.7296	0.1538	0.4957	0.5967	0.4039	0.6064	0.805	0.591	0.6854	0.3144	0.5081
0.6957	26.0	19318	0.7092	0.4545	0.7561	0.4626	0.1934	0.4516	0.7304	0.1539	0.4973	0.5984	0.4111	0.6069	0.8027	0.5887	0.6838	0.3203	0.513
0.6864	27.0	20061	0.7065	0.4559	0.7552	0.4667	0.1973	0.4536	0.7279	0.1542	0.4987	0.5998	0.4103	0.6117	0.7982	0.5941	0.6895	0.3178	0.5101
0.684	28.0	20804	0.7045	0.458	0.7582	0.4746	0.1966	0.4572	0.7311	0.1545	0.4997	0.6022	0.415	0.6116	0.8053	0.594	0.6893	0.322	0.5152
0.681	29.0	21547	0.7040	0.4574	0.7603	0.4715	0.1971	0.4563	0.7296	0.1536	0.4988	0.5987	0.4136	0.6073	0.8004	0.591	0.6872	0.3239	0.5102
0.6769	30.0	22290	0.7023	0.4585	0.7613	0.4703	0.2004	0.4565	0.7335	0.1539	0.5012	0.6019	0.4214	0.6084	0.8038	0.5922	0.6902	0.3247	0.5136
0.6774	31.0	23033	0.6974	0.4607	0.7646	0.4775	0.2032	0.4594	0.7304	0.1543	0.502	0.6048	0.4317	0.6094	0.8032	0.5963	0.6924	0.3251	0.5173
0.6678	32.0	23776	0.6914	0.4654	0.7623	0.4756	0.2076	0.4642	0.7337	0.1559	0.5067	0.6088	0.4287	0.6175	0.8047	0.6021	0.6976	0.3287	0.5201
0.6733	33.0	24519	0.6896	0.4664	0.767	0.4805	0.212	0.4653	0.7326	0.1552	0.5086	0.6078	0.4246	0.6166	0.8067	0.6038	0.6979	0.329	0.5177
0.6656	34.0	25262	0.6878	0.4687	0.769	0.4857	0.2133	0.4682	0.7353	0.1558	0.5112	0.6078	0.4241	0.6173	0.8055	0.6048	0.6975	0.3326	0.5181
0.6599	35.0	26005	0.6848	0.4716	0.7718	0.492	0.2121	0.4717	0.7364	0.156	0.5135	0.6121	0.4292	0.6218	0.8082	0.6081	0.7002	0.3351	0.524
0.6646	36.0	26748	0.6857	0.4709	0.7721	0.487	0.2129	0.4711	0.7369	0.1565	0.5137	0.6109	0.4316	0.6184	0.8092	0.6073	0.7001	0.3344	0.5217
0.6568	37.0	27491	0.6867	0.4707	0.7729	0.4843	0.2147	0.4694	0.7393	0.1564	0.5117	0.6102	0.4252	0.6195	0.8094	0.6065	0.6985	0.3349	0.5219
0.6493	38.0	28234	0.6830	0.4713	0.771	0.4835	0.2121	0.4734	0.7357	0.1573	0.5131	0.6118	0.4277	0.622	0.8083	0.6081	0.7002	0.3345	0.5234
0.6567	39.0	28977	0.6813	0.4724	0.771	0.4841	0.2117	0.4729	0.7396	0.1573	0.515	0.6135	0.4351	0.6213	0.8097	0.6098	0.701	0.3351	0.526
0.6532	40.0	29720	0.6797	0.4743	0.7751	0.4848	0.2137	0.4761	0.7369	0.1573	0.516	0.6149	0.4354	0.6243	0.8077	0.6101	0.7019	0.3384	0.5279
0.6475	41.0	30463	0.6769	0.4755	0.773	0.4903	0.219	0.4742	0.7397	0.1572	0.5193	0.6169	0.4418	0.6248	0.8088	0.6125	0.7044	0.3384	0.5295
0.6432	42.0	31206	0.6779	0.4762	0.7757	0.4926	0.2171	0.4777	0.739	0.158	0.5184	0.6168	0.4384	0.6262	0.8079	0.6122	0.703	0.3403	0.5305
0.6482	43.0	31949	0.6762	0.4759	0.7756	0.4897	0.218	0.4755	0.74	0.1579	0.5169	0.6141	0.4329	0.624	0.8071	0.6132	0.7042	0.3385	0.524
0.6427	44.0	32692	0.6744	0.4771	0.776	0.49	0.2167	0.4766	0.7445	0.1591	0.5195	0.6159	0.4333	0.6258	0.8112	0.616	0.7064	0.3382	0.5254
0.6409	45.0	33435	0.6758	0.4767	0.777	0.4882	0.2189	0.4762	0.7426	0.1581	0.5181	0.6155	0.437	0.6239	0.8099	0.6141	0.7046	0.3393	0.5264
0.6361	46.0	34178	0.6748	0.4758	0.7762	0.4888	0.2178	0.4744	0.7448	0.1577	0.5177	0.6139	0.4299	0.6234	0.8116	0.6135	0.704	0.338	0.5238
0.6383	47.0	34921	0.6757	0.475	0.7788	0.4883	0.217	0.4751	0.7424	0.158	0.5184	0.6139	0.4278	0.6244	0.8116	0.6115	0.7031	0.3384	0.5247
0.6421	48.0	35664	0.6717	0.4793	0.7796	0.4909	0.2217	0.4788	0.7447	0.1589	0.5208	0.6186	0.4413	0.627	0.8114	0.6161	0.7071	0.3426	0.5301
0.6357	49.0	36407	0.6712	0.4789	0.7787	0.4916	0.2215	0.4789	0.7425	0.1592	0.5219	0.6188	0.4403	0.6279	0.8114	0.6161	0.7069	0.3418	0.5308
0.6322	50.0	37150	0.6715	0.4792	0.7795	0.4922	0.2219	0.4792	0.7436	0.1587	0.5223	0.6188	0.4368	0.629	0.8124	0.6174	0.7074	0.3409	0.5302
0.6324	51.0	37893	0.6729	0.478	0.7787	0.4906	0.2206	0.4772	0.7447	0.1585	0.5202	0.6171	0.4379	0.6254	0.8126	0.6153	0.7048	0.3407	0.5293
0.6402	52.0	38636	0.6707	0.4806	0.7792	0.4978	0.2222	0.4795	0.747	0.1592	0.5221	0.6196	0.4419	0.6278	0.8135	0.6174	0.7076	0.3438	0.5317
0.6328	53.0	39379	0.6716	0.4796	0.7794	0.4964	0.2231	0.4789	0.7445	0.1587	0.5212	0.6184	0.4405	0.6269	0.812	0.6173	0.707	0.342	0.5299
0.6349	54.0	40122	0.6715	0.4795	0.7796	0.4941	0.223	0.4782	0.7453	0.1587	0.5216	0.6186	0.4399	0.6268	0.8135	0.6165	0.7066	0.3425	0.5305
0.6293	55.0	40865	0.6705	0.4798	0.779	0.4921	0.2232	0.479	0.7445	0.159	0.5222	0.6192	0.4408	0.628	0.8123	0.6177	0.7073	0.3419	0.5311
0.6324	56.0	41608	0.6705	0.4804	0.78	0.4939	0.2238	0.48	0.7446	0.1588	0.5222	0.6198	0.4418	0.6285	0.8127	0.618	0.7079	0.3428	0.5318
0.6293	57.0	42351	0.6702	0.4803	0.7796	0.4947	0.2235	0.4792	0.7452	0.159	0.5228	0.6197	0.4415	0.6283	0.813	0.6178	0.708	0.3428	0.5314
0.6353	58.0	43094	0.6701	0.4804	0.7798	0.4943	0.224	0.4795	0.7444	0.1588	0.5223	0.6198	0.4422	0.6284	0.8128	0.6183	0.7082	0.3424	0.5315
0.6323	59.0	43837	0.6703	0.4803	0.78	0.4935	0.2238	0.4794	0.7443	0.1586	0.5223	0.6196	0.4419	0.6282	0.8124	0.6183	0.7082	0.3423	0.5309
0.6384	60.0	44580	0.6703	0.4803	0.7799	0.494	0.2239	0.4794	0.7442	0.1586	0.5224	0.6196	0.4422	0.6281	0.8123	0.6182	0.7081	0.3423	0.5311

Framework versions

• Transformers 4.49.0
• Pytorch 2.5.1+cu121
• Datasets 3.3.2
• Tokenizers 0.21.0