lw-detr-medium-tray-detection

This model is a fine-tuned version of AnnaZhang/lwdetr_medium_60e_coco on the nielsr/tray-cart-detection dataset. It achieves the following results on the evaluation set:

• Loss: 8.6378
• Map: 0.4419
• Map 50: 0.7634
• Map 75: 0.4321
• Map Small: 0.6228
• Map Medium: 0.4287
• Map Large: 0.5024
• Mar 1: 0.0566
• Mar 10: 0.3224
• Mar 100: 0.5434
• Mar Small: 0.6238
• Mar Medium: 0.5245
• Mar Large: 0.6596
• Map Per Class: -1.0
• Mar 100 Per Class: -1.0

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
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• 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
• num_epochs: 300.0

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 Per Class	Mar 100 Per Class
8.0517	1.0	25	7.5501	0.1695	0.4286	0.1112	0.2996	0.1636	0.2694	0.0337	0.1580	0.3180	0.4286	0.2886	0.4978	-1.0	-1.0
5.7334	2.0	50	8.1382	0.2317	0.4736	0.2073	0.4546	0.2245	0.3582	0.0351	0.1654	0.3319	0.5452	0.2917	0.5543	-1.0	-1.0
5.0764	3.0	75	8.3725	0.3248	0.6463	0.2534	0.5197	0.3029	0.4731	0.0490	0.2409	0.4048	0.5714	0.3706	0.6008	-1.0	-1.0
4.7140	4.0	100	8.2542	0.3632	0.6653	0.3197	0.5918	0.3403	0.4378	0.0506	0.2716	0.4429	0.6024	0.4128	0.6121	-1.0	-1.0
4.4517	5.0	125	7.5693	0.3250	0.6562	0.2605	0.5206	0.2965	0.4704	0.0469	0.2654	0.4334	0.5952	0.4009	0.6208	-1.0	-1.0
4.1539	6.0	150	8.6084	0.3616	0.6758	0.2926	0.5325	0.3458	0.4462	0.0463	0.2752	0.4371	0.5643	0.4189	0.5285	-1.0	-1.0
4.0950	7.0	175	8.7664	0.3819	0.7026	0.3594	0.5704	0.3568	0.5346	0.0525	0.2822	0.4631	0.5762	0.4350	0.6359	-1.0	-1.0
3.9274	8.0	200	8.8551	0.4033	0.7251	0.3662	0.5597	0.3810	0.5447	0.0559	0.2977	0.4828	0.5881	0.4598	0.6231	-1.0	-1.0
3.7842	9.0	225	8.5001	0.4101	0.7291	0.4116	0.6180	0.3954	0.5028	0.0554	0.2933	0.4974	0.6524	0.4695	0.6552	-1.0	-1.0
3.6846	10.0	250	8.5214	0.3728	0.6703	0.3901	0.5360	0.3613	0.4350	0.0442	0.2828	0.4719	0.6381	0.4574	0.5248	-1.0	-1.0
3.8286	11.0	275	8.6306	0.4163	0.7228	0.3890	0.5807	0.4072	0.4799	0.0543	0.3032	0.4942	0.5929	0.4823	0.5604	-1.0	-1.0
3.6285	12.0	300	8.4890	0.4145	0.7332	0.3865	0.5893	0.3916	0.5381	0.0531	0.3092	0.4921	0.6095	0.4691	0.6255	-1.0	-1.0
3.3944	13.0	325	8.6758	0.4081	0.7184	0.3577	0.5308	0.4049	0.4576	0.0549	0.2953	0.4959	0.5524	0.4857	0.5604	-1.0	-1.0
3.4186	14.0	350	8.1625	0.4200	0.7018	0.4636	0.6063	0.4090	0.4809	0.0548	0.3053	0.5117	0.6595	0.4984	0.5648	-1.0	-1.0
3.3058	15.0	375	8.6068	0.4010	0.7081	0.3977	0.6034	0.3958	0.4440	0.0569	0.3131	0.5060	0.6262	0.4838	0.6347	-1.0	-1.0
3.2623	16.0	400	8.7036	0.3967	0.6959	0.3976	0.5572	0.3982	0.4124	0.0509	0.2900	0.4903	0.5738	0.4846	0.5147	-1.0	-1.0
3.2304	17.0	425	8.6188	0.4032	0.7127	0.3661	0.5780	0.3917	0.4922	0.0487	0.2946	0.5015	0.5833	0.4825	0.6202	-1.0	-1.0
3.1675	18.0	450	8.5400	0.4203	0.7311	0.4205	0.5791	0.3996	0.5425	0.0520	0.3126	0.5058	0.5905	0.4845	0.6413	-1.0	-1.0
3.0644	19.0	475	8.9040	0.4293	0.7462	0.4458	0.6098	0.4091	0.5483	0.0611	0.3135	0.5219	0.6286	0.5031	0.6297	-1.0	-1.0
3.1360	20.0	500	8.5330	0.4412	0.7499	0.4289	0.6234	0.4214	0.5239	0.0550	0.3362	0.5446	0.6548	0.5195	0.6996	-1.0	-1.0
3.0481	21.0	525	8.5831	0.4414	0.7512	0.4634	0.6079	0.4146	0.5592	0.0533	0.3086	0.5313	0.6167	0.5059	0.6946	-1.0	-1.0
3.0654	22.0	550	8.6378	0.4419	0.7634	0.4321	0.6228	0.4287	0.5024	0.0566	0.3224	0.5434	0.6238	0.5245	0.6596	-1.0	-1.0
2.9182	23.0	575	8.7346	0.3946	0.6975	0.3866	0.4916	0.3938	0.4398	0.0477	0.2899	0.4900	0.5024	0.4787	0.5803	-1.0	-1.0
2.8294	24.0	600	8.7305	0.3981	0.7152	0.3830	0.5847	0.3894	0.4705	0.0550	0.3006	0.4947	0.5857	0.4793	0.5807	-1.0	-1.0
2.8585	25.0	625	8.9077	0.3927	0.6933	0.3922	0.5287	0.3899	0.4685	0.0464	0.2964	0.4862	0.5381	0.4636	0.6402	-1.0	-1.0
2.6991	26.0	650	8.4823	0.4155	0.7124	0.4351	0.6226	0.4020	0.4909	0.0572	0.3089	0.5125	0.6286	0.4933	0.6180	-1.0	-1.0
2.8258	27.0	675	8.7625	0.4052	0.7436	0.3913	0.5667	0.3859	0.4814	0.0545	0.3006	0.4930	0.5786	0.4715	0.6240	-1.0	-1.0
2.7473	28.0	700	8.6783	0.4247	0.7304	0.4309	0.6396	0.4069	0.4929	0.0611	0.3044	0.5159	0.6405	0.5027	0.5725	-1.0	-1.0
2.7875	29.0	725	8.6406	0.4045	0.7111	0.4057	0.6242	0.4000	0.4394	0.0540	0.3005	0.5018	0.6500	0.4852	0.5746	-1.0	-1.0
2.6226	30.0	750	8.9299	0.4019	0.7086	0.4041	0.6233	0.3924	0.4977	0.0532	0.2996	0.4843	0.6238	0.4586	0.6271	-1.0	-1.0
2.5618	31.0	775	8.8131	0.4132	0.7018	0.4289	0.6137	0.3936	0.5068	0.0552	0.3073	0.5017	0.6214	0.4854	0.5825	-1.0	-1.0
2.5460	32.0	800	9.0914	0.3757	0.6862	0.3412	0.6066	0.3763	0.4124	0.0576	0.2936	0.4861	0.6143	0.4638	0.6069	-1.0	-1.0

Framework versions

• Transformers 5.3.0.dev0
• Pytorch 2.10.0+cu128
• Datasets 4.8.2
• Tokenizers 0.22.2