Vasily-Lapinsky/Indoor-RT-Detrv2

rtdetr_v2_r18vd-finetuned-indoor-batch8-loss-finall

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

• Loss: 0.7451
• Map: 0.7714
• Map 50: 0.9402
• Map 75: 0.8807
• Map Small: 0.4279
• Map Medium: 0.6977
• Map Large: 0.7784
• Mar 1: 0.704
• Mar 10: 0.8268
• Mar 100: 0.8378
• Mar Small: 0.4723
• Mar Medium: 0.7949
• Mar Large: 0.8414
• Map Exit: 0.7325
• Mar 100 Exit: 0.7923
• Map Fireextinguisher: 0.7845
• Mar 100 Fireextinguisher: 0.8203
• Map Chair: 0.7354
• Mar 100 Chair: 0.8301
• Map Clock: 0.7962
• Mar 100 Clock: 0.837
• Map Trashbin: 0.5619
• Mar 100 Trashbin: 0.725
• Map Printer: 0.8914
• Mar 100 Printer: 0.9375
• Map Screen: 0.8976
• Mar 100 Screen: 0.9222

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: 16
• eval_batch_size: 8
• seed: 1337
• 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: linear
• num_epochs: 25.0
• mixed_precision_training: Native AMP

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 Exit	Mar 100 Exit	Map Fireextinguisher	Mar 100 Fireextinguisher	Map Chair	Mar 100 Chair	Map Clock	Mar 100 Clock	Map Trashbin	Mar 100 Trashbin	Map Printer	Mar 100 Printer	Map Screen	Mar 100 Screen
0.8218	1.0	112	0.9142	0.6911	0.8734	0.7937	0.2594	0.6244	0.7302	0.6551	0.7783	0.7887	0.3233	0.7267	0.8169	0.6682	0.7519	0.6719	0.7531	0.6624	0.7687	0.761	0.7963	0.3997	0.64	0.8667	0.9	0.8078	0.9111
0.801	2.0	224	0.8670	0.7147	0.8926	0.8306	0.2687	0.62	0.7324	0.6732	0.7871	0.7982	0.3294	0.7339	0.8116	0.6883	0.7577	0.7121	0.7797	0.6997	0.7982	0.7695	0.7889	0.4928	0.685	0.871	0.9	0.7696	0.8778
0.7682	3.0	336	0.9007	0.7075	0.9011	0.8123	0.2875	0.6152	0.7243	0.6701	0.7838	0.794	0.401	0.7275	0.8217	0.687	0.7442	0.7061	0.7741	0.6667	0.7804	0.7607	0.7963	0.4857	0.72	0.8775	0.8875	0.769	0.8556
0.7507	4.0	448	0.8265	0.7392	0.9242	0.8584	0.3972	0.6715	0.7561	0.6817	0.808	0.8202	0.4516	0.7653	0.8352	0.705	0.7654	0.7247	0.7811	0.709	0.8117	0.783	0.8333	0.5202	0.715	0.8842	0.9125	0.848	0.9222
0.7301	5.0	560	0.8271	0.7227	0.9141	0.8282	0.3326	0.6377	0.7385	0.6691	0.7986	0.8064	0.4591	0.7376	0.8487	0.7092	0.7942	0.7371	0.7902	0.7047	0.7969	0.7924	0.8481	0.5106	0.675	0.8083	0.8625	0.7967	0.8778
0.7197	6.0	672	0.8127	0.7561	0.9331	0.8718	0.3553	0.6725	0.7661	0.6901	0.8124	0.8204	0.4527	0.7664	0.8489	0.7203	0.7788	0.7444	0.8007	0.6974	0.8	0.7891	0.8148	0.5839	0.725	0.8725	0.9125	0.8853	0.9111
0.7041	7.0	784	0.7735	0.7637	0.9447	0.8864	0.4044	0.6757	0.7658	0.6935	0.8205	0.8326	0.5206	0.7725	0.8235	0.7446	0.8	0.7543	0.8154	0.7246	0.8209	0.7881	0.8407	0.5854	0.715	0.8675	0.925	0.8814	0.9111
0.7111	8.0	896	0.7977	0.7312	0.9297	0.8389	0.3618	0.6202	0.7366	0.6735	0.7931	0.8146	0.419	0.7552	0.8517	0.7106	0.7981	0.7485	0.8028	0.7261	0.8178	0.7699	0.8074	0.486	0.665	0.8737	0.9	0.8038	0.9111
0.6902	9.0	1008	0.7777	0.7398	0.9262	0.8423	0.4019	0.6672	0.7482	0.6855	0.8141	0.8267	0.4626	0.7678	0.849	0.7328	0.7981	0.7337	0.8035	0.7393	0.8442	0.777	0.8074	0.499	0.685	0.8638	0.9375	0.8332	0.9111
0.6712	10.0	1120	0.7881	0.7498	0.9418	0.8711	0.4102	0.6709	0.7691	0.6778	0.8058	0.815	0.4473	0.7647	0.843	0.6979	0.7712	0.7497	0.8014	0.7325	0.8325	0.7693	0.8111	0.5833	0.725	0.8505	0.875	0.8655	0.8889
0.6763	11.0	1232	0.7937	0.7666	0.9359	0.8797	0.3943	0.6884	0.7803	0.6931	0.8147	0.8237	0.4368	0.7714	0.8322	0.7247	0.7788	0.7458	0.7944	0.727	0.8215	0.7673	0.8037	0.5958	0.695	0.9138	0.95	0.8916	0.9222
0.6792	12.0	1344	0.8028	0.7465	0.9163	0.8525	0.3173	0.6541	0.7807	0.6781	0.8056	0.8187	0.3531	0.7869	0.882	0.7167	0.775	0.7393	0.7874	0.7217	0.8166	0.7675	0.7889	0.5595	0.695	0.886	0.9125	0.8346	0.9556
0.6844	13.0	1456	0.7748	0.7528	0.9263	0.8639	0.3421	0.6686	0.776	0.6968	0.8168	0.8265	0.4386	0.7729	0.8377	0.7173	0.7923	0.7556	0.8021	0.7344	0.8252	0.7799	0.8222	0.5131	0.695	0.9105	0.9375	0.8591	0.9111
0.667	14.0	1568	0.7876	0.7551	0.9325	0.8741	0.3361	0.6796	0.7733	0.6921	0.8056	0.8199	0.3841	0.7683	0.8661	0.7108	0.7865	0.7369	0.793	0.7277	0.8141	0.7921	0.8259	0.5695	0.685	0.8835	0.9125	0.8654	0.9222
0.6596	15.0	1680	0.7691	0.7683	0.9294	0.8897	0.3874	0.6794	0.787	0.6952	0.8234	0.8342	0.4475	0.7897	0.8442	0.7322	0.7923	0.7666	0.8231	0.7281	0.8307	0.8055	0.837	0.5706	0.72	0.8986	0.925	0.8762	0.9111
0.6349	16.0	1792	0.7581	0.7683	0.9414	0.8772	0.3607	0.6942	0.7746	0.7047	0.8307	0.8391	0.4571	0.8117	0.8584	0.7252	0.8019	0.7592	0.8161	0.7307	0.8331	0.8069	0.8444	0.5626	0.72	0.9082	0.925	0.8853	0.9333
0.6433	17.0	1904	0.7571	0.7604	0.9339	0.881	0.3855	0.6851	0.7799	0.6981	0.8191	0.839	0.4629	0.8256	0.8329	0.7378	0.8077	0.7639	0.8168	0.7335	0.835	0.7824	0.8222	0.5558	0.725	0.8643	0.9	0.8849	0.9667
0.6348	18.0	2016	0.7545	0.7708	0.9345	0.8934	0.3228	0.6765	0.7954	0.7003	0.8238	0.8348	0.4469	0.7905	0.8576	0.7379	0.8019	0.7763	0.8182	0.7278	0.8245	0.7998	0.8407	0.5623	0.7	0.9131	0.925	0.8783	0.9333
0.6322	19.0	2128	0.7642	0.7651	0.931	0.8875	0.3233	0.6763	0.7673	0.6998	0.818	0.8329	0.4618	0.7852	0.835	0.7146	0.7827	0.7603	0.8273	0.7405	0.8294	0.795	0.8333	0.54	0.71	0.9184	0.925	0.887	0.9222
0.6219	20.0	2240	0.7587	0.7702	0.9434	0.8849	0.4091	0.6879	0.7744	0.6981	0.8254	0.8447	0.4631	0.8049	0.8641	0.734	0.8019	0.7715	0.8252	0.7387	0.8356	0.8136	0.8481	0.5692	0.72	0.8794	0.9375	0.8849	0.9444
0.6105	21.0	2352	0.7471	0.7701	0.9393	0.8868	0.4202	0.678	0.7805	0.6996	0.8218	0.8356	0.4765	0.7799	0.8389	0.7193	0.7923	0.7781	0.8259	0.7343	0.8337	0.8069	0.8444	0.5684	0.715	0.9144	0.9375	0.8691	0.9
0.6123	22.0	2464	0.7461	0.7691	0.9426	0.8919	0.4237	0.6834	0.7659	0.6897	0.8198	0.8336	0.4648	0.7924	0.8395	0.7271	0.7942	0.7794	0.828	0.7359	0.8301	0.7978	0.8333	0.5777	0.715	0.8965	0.9125	0.8695	0.9222
0.6125	23.0	2576	0.7466	0.7649	0.9344	0.8811	0.4351	0.6806	0.7737	0.6987	0.821	0.8323	0.4805	0.7883	0.8369	0.7274	0.7962	0.7802	0.8259	0.7339	0.8221	0.7908	0.8296	0.5378	0.705	0.911	0.925	0.8735	0.9222
0.616	24.0	2688	0.7563	0.7612	0.9365	0.872	0.4332	0.6812	0.7718	0.698	0.8246	0.8337	0.4809	0.7915	0.8422	0.7376	0.8019	0.7817	0.8273	0.7296	0.8184	0.793	0.8333	0.5472	0.72	0.8612	0.9125	0.8778	0.9222
0.6079	25.0	2800	0.7451	0.7714	0.9402	0.8807	0.4279	0.6977	0.7784	0.704	0.8268	0.8378	0.4723	0.7949	0.8414	0.7325	0.7923	0.7845	0.8203	0.7354	0.8301	0.7962	0.837	0.5619	0.725	0.8914	0.9375	0.8976	0.9222

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1