Modelpublic

dri11heaD/rtdetr-vehicle-accident-detection

sourceHugging Faceapache-2.0updated 1mo agoView on Hugging Face

0likes17downloads

Model Card

rtdetr-vehicle-accident-detection

This model is a fine-tuned version of PekingU/rtdetr_r50vd_coco_o365 on the None dataset. It achieves the following results on the evaluation set:

—Loss: 12.2411
—Map: 0.4396
—Map 50: 0.5998
—Map 75: 0.4789
—Map Small: 0.3192
—Map Medium: 0.3353
—Map Large: 0.5358
—Mar 1: 0.3873
—Mar 10: 0.6599
—Mar 100: 0.7769
—Mar Small: 0.5733
—Mar Medium: 0.7379
—Mar Large: 0.846
—Map Accident: 0.5871
—Mar 100 Accident: 0.8212
—Map Non-accident: 0.2922
—Mar 100 Non-accident: 0.7326

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: 0.0001
—trainbatchsize: 4
—evalbatchsize: 4
—seed: 3407
—optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
—lrschedulertype: cosine
—lrschedulerwarmup_steps: 300
—num_epochs: 20
—mixedprecisiontraining: 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 Accident	Mar 100 Accident	Map Non-accident	Mar 100 Non-accident
34.3857	1.0	147	17.0306	0.1287	0.2164	0.1281	0.0152	0.0507	0.2043	0.1579	0.4775	0.6838	0.3933	0.6255	0.7562	0.2205	0.7451	0.0369	0.6225
23.2918	2.0	294	13.6842	0.2082	0.3272	0.2125	0.1027	0.1241	0.235	0.2401	0.5515	0.7359	0.4633	0.6905	0.7985	0.349	0.787	0.0674	0.6848
20.2538	3.0	441	15.5058	0.1086	0.1863	0.1037	0.162	0.1085	0.118	0.1705	0.4925	0.7317	0.48	0.7353	0.7556	0.1776	0.7611	0.0396	0.7022
18.6402	4.0	588	12.7415	0.2555	0.3774	0.2743	0.1527	0.1933	0.2796	0.2697	0.5975	0.7796	0.5133	0.7421	0.8368	0.4315	0.8207	0.0796	0.7384
18.0487	5.0	735	12.7144	0.3171	0.4477	0.3396	0.3157	0.2134	0.3409	0.3292	0.6712	0.7894	0.5533	0.7539	0.8465	0.491	0.8332	0.1431	0.7457
16.2050	6.0	882	12.1642	0.3583	0.5194	0.3917	0.392	0.294	0.4254	0.3362	0.658	0.7727	0.5	0.725	0.8361	0.5061	0.8171	0.2106	0.7283
15.4431	7.0	1029	11.9625	0.3941	0.5309	0.4366	0.313	0.31	0.4255	0.3691	0.6622	0.803	0.5033	0.7737	0.8682	0.5588	0.8358	0.2294	0.7703
13.9417	8.0	1176	11.7418	0.413	0.5827	0.4482	0.3636	0.2982	0.469	0.3611	0.6722	0.7738	0.58	0.732	0.8287	0.5484	0.8259	0.2776	0.7217
13.9773	9.0	1323	11.7699	0.4285	0.5919	0.4735	0.3905	0.3132	0.4907	0.3515	0.6647	0.7739	0.5567	0.7324	0.851	0.5641	0.8254	0.293	0.7225
13.5177	10.0	1470	11.5481	0.424	0.5776	0.4673	0.3623	0.2948	0.5155	0.3689	0.6762	0.7814	0.59	0.7399	0.8432	0.594	0.8295	0.254	0.7333
12.8414	11.0	1617	11.9601	0.4462	0.6045	0.4926	0.4114	0.3458	0.5228	0.3879	0.6826	0.7958	0.6067	0.7577	0.8593	0.6001	0.843	0.2924	0.7486
11.1679	12.0	1764	11.5837	0.4502	0.6095	0.4951	0.3434	0.3507	0.5305	0.3946	0.6792	0.7929	0.6433	0.7556	0.8412	0.6011	0.8337	0.2992	0.7522
10.4115	13.0	1911	12.0687	0.4393	0.6029	0.4926	0.2706	0.3286	0.5283	0.3735	0.6784	0.7765	0.5433	0.7361	0.8429	0.5885	0.8269	0.29	0.7261
10.1284	14.0	2058	11.9256	0.4471	0.6075	0.497	0.3799	0.3347	0.5352	0.3925	0.6833	0.7841	0.6	0.7351	0.8435	0.5957	0.829	0.2984	0.7391
9.7026	15.0	2205	11.9128	0.4461	0.6111	0.4978	0.3536	0.3341	0.5414	0.3823	0.6731	0.7796	0.6233	0.7321	0.8432	0.5997	0.8352	0.2926	0.7239
9.9951	16.0	2352	12.1509	0.4522	0.6193	0.4912	0.3841	0.3506	0.5359	0.3836	0.6733	0.785	0.6067	0.7305	0.8533	0.5956	0.8264	0.3088	0.7435
9.4406	17.0	2499	12.2262	0.4372	0.598	0.4804	0.3589	0.3363	0.525	0.385	0.6574	0.7792	0.5867	0.7358	0.8399	0.5817	0.8192	0.2927	0.7391
9.2950	18.0	2646	12.2601	0.4433	0.6035	0.481	0.333	0.3401	0.5386	0.3883	0.6713	0.776	0.5933	0.7282	0.8412	0.5918	0.8238	0.2948	0.7283
9.1047	19.0	2793	12.2527	0.4402	0.6011	0.4765	0.3246	0.3365	0.5359	0.3873	0.6678	0.7821	0.5767	0.7443	0.847	0.5878	0.8228	0.2926	0.7413
8.4988	20.0	2940	12.2411	0.4396	0.5998	0.4789	0.3192	0.3353	0.5358	0.3873	0.6599	0.7769	0.5733	0.7379	0.846	0.5871	0.8212	0.2922	0.7326

Framework versions

—Transformers 5.15.1
—Pytorch 2.11.0+cu128
—Datasets 5.0.1
—Tokenizers 0.22.2