Modelpublic

dri11heaD/rtdetr-accident-cctv

sourceHugging Faceapache-2.0updated 1mo agoView on Hugging Face

0likes8downloads

Model Card

rtdetr-accident-cctv

This model is a fine-tuned version of PekingU/rtdetr_r18vd on the None dataset. It achieves the following results on the evaluation set:

—Loss: 7.3092
—Map: 0.3966
—Map 50: 0.5632
—Map 75: 0.4384
—Map Small: 0.0078
—Map Medium: 0.3845
—Map Large: 0.438
—Mar 1: 0.4148
—Mar 10: 0.6538
—Mar 100: 0.7869
—Mar Small: 0.3633
—Mar Medium: 0.7854
—Mar Large: 0.7864
—Map Accident: 0.5535
—Mar 100 Accident: 0.8302
—Map Non-accident: 0.2396
—Mar 100 Non-accident: 0.7437

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
—trainbatchsize: 8
—evalbatchsize: 8
—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: 12
—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
13.9538	1.0	309	9.3017	0.1194	0.1975	0.1187	0.0005	0.0839	0.1319	0.2059	0.454	0.7321	0.1167	0.7256	0.755	0.212	0.7932	0.0267	0.671
10.9516	2.0	618	7.7241	0.1604	0.2364	0.1692	0.0046	0.1294	0.1721	0.2663	0.5475	0.7682	0.2033	0.7531	0.7882	0.2623	0.8098	0.0584	0.7265
9.7495	3.0	927	7.4023	0.2277	0.3323	0.2446	0.0027	0.2111	0.2466	0.2999	0.5443	0.7611	0.2167	0.7467	0.7647	0.402	0.8071	0.0534	0.7151
9.0010	4.0	1236	7.1968	0.2604	0.3866	0.2688	0.0033	0.2712	0.2747	0.3823	0.616	0.7804	0.2633	0.762	0.7846	0.4156	0.8163	0.1052	0.7445
8.1878	5.0	1545	7.3127	0.3201	0.4578	0.3438	0.0044	0.3368	0.3705	0.406	0.6326	0.8057	0.31	0.7894	0.8222	0.4969	0.838	0.1432	0.7735
7.6957	6.0	1854	7.1702	0.3381	0.4731	0.3851	0.0083	0.3387	0.4001	0.3991	0.6428	0.8036	0.3733	0.7875	0.8129	0.5361	0.841	0.1402	0.7661
7.2060	7.0	2163	7.1868	0.3748	0.5356	0.4141	0.0082	0.3765	0.4138	0.4068	0.6454	0.7854	0.3367	0.7818	0.7839	0.5359	0.8173	0.2136	0.7535
6.9991	8.0	2472	7.2713	0.3682	0.5264	0.4016	0.0091	0.3571	0.4085	0.4129	0.6504	0.7904	0.39	0.7895	0.7824	0.5473	0.8356	0.1892	0.7453
6.5298	9.0	2781	7.2003	0.399	0.5598	0.444	0.0092	0.3816	0.4531	0.4156	0.6427	0.7918	0.3467	0.7919	0.7889	0.5602	0.8339	0.2377	0.7498
6.3957	10.0	3090	7.3283	0.3835	0.5481	0.4221	0.0089	0.3827	0.4168	0.427	0.6367	0.7828	0.3567	0.7829	0.78	0.5509	0.8268	0.216	0.7388
6.2769	11.0	3399	7.2994	0.394	0.5562	0.434	0.0085	0.3767	0.4416	0.414	0.644	0.7889	0.3633	0.7847	0.7935	0.5563	0.8292	0.2316	0.7486
6.2947	12.0	3708	7.3092	0.3966	0.5632	0.4384	0.0078	0.3845	0.438	0.4148	0.6538	0.7869	0.3633	0.7854	0.7864	0.5535	0.8302	0.2396	0.7437

Framework versions

—Transformers 5.15.0
—Pytorch 2.11.0+cu128
—Datasets 5.0.1
—Tokenizers 0.22.2