rtdetr-tray-cart

This model is a fine-tuned version of PekingU/rtdetr_r101vd_coco_o365 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 8.0013
• Map: 0.3254
• Map 50: 0.5528
• Map 75: 0.3324
• Map Small: 0.6086
• Map Medium: 0.3016
• Map Large: 0.5975
• Mar 1: 0.0588
• Mar 10: 0.3009
• Mar 100: 0.5992
• Mar Small: 0.675
• Mar Medium: 0.5598
• Mar Large: 0.8468
• Map Tray: 0.2404
• Mar 100 Tray: 0.505
• Map Cart: 0.4104
• Mar 100 Cart: 0.6935

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: 8
• eval_batch_size: 8
• seed: 42
• 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: cosine
• num_epochs: 50
• 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 Tray	Mar 100 Tray	Map Cart	Mar 100 Cart
No log	1.0	13	19.9337	0.002	0.0076	0.0004	0.0	0.0027	0.0154	0.0	0.0057	0.0623	0.0	0.0603	0.0895	0.0034	0.066	0.0006	0.0587
No log	2.0	26	22.1722	0.0027	0.0084	0.0012	0.0001	0.0052	0.0111	0.0	0.0057	0.0701	0.0139	0.0649	0.1323	0.0049	0.0685	0.0004	0.0717
No log	3.0	39	17.1149	0.0113	0.0298	0.0068	0.0591	0.0127	0.042	0.0008	0.0178	0.105	0.3056	0.0713	0.2169	0.0201	0.1273	0.0026	0.0826
No log	4.0	52	10.9760	0.1099	0.1888	0.1182	0.2553	0.1004	0.2333	0.0137	0.0908	0.2935	0.2722	0.2621	0.5581	0.2089	0.4088	0.0108	0.1783
No log	5.0	65	10.3979	0.1323	0.2611	0.1276	0.5042	0.1206	0.2902	0.0227	0.1687	0.376	0.5306	0.3309	0.6456	0.2317	0.4215	0.0328	0.3304
No log	6.0	78	10.0574	0.183	0.3502	0.1716	0.5487	0.1395	0.4027	0.0274	0.1741	0.4055	0.6528	0.3589	0.6032	0.2429	0.4392	0.123	0.3717
No log	7.0	91	9.9984	0.2236	0.4243	0.2019	0.5683	0.1733	0.4474	0.0493	0.2187	0.419	0.6167	0.3656	0.7052	0.2473	0.4336	0.2	0.4043
No log	8.0	104	9.6693	0.2453	0.4586	0.2238	0.6032	0.268	0.2727	0.0481	0.2637	0.4929	0.6444	0.4468	0.7573	0.1751	0.4401	0.3156	0.5457
No log	9.0	117	8.7372	0.3333	0.5783	0.3575	0.6196	0.3265	0.3584	0.0581	0.2821	0.5441	0.6639	0.5068	0.7573	0.2562	0.4731	0.4103	0.6152
No log	10.0	130	8.8186	0.2482	0.4278	0.2468	0.5817	0.2468	0.3644	0.0468	0.2827	0.558	0.6028	0.5278	0.7593	0.2212	0.4639	0.2753	0.6522
No log	11.0	143	8.5718	0.2821	0.501	0.2732	0.6277	0.2546	0.4392	0.057	0.2703	0.5479	0.6278	0.5104	0.7843	0.2289	0.4762	0.3353	0.6196
No log	12.0	156	8.7045	0.3346	0.5686	0.3482	0.6173	0.3073	0.394	0.0506	0.2959	0.5566	0.6472	0.5207	0.7589	0.2881	0.4806	0.3811	0.6326
No log	13.0	169	8.3602	0.3594	0.592	0.3995	0.6538	0.3268	0.4867	0.0641	0.2966	0.5952	0.675	0.5618	0.7895	0.3103	0.5013	0.4086	0.6891
No log	14.0	182	8.4438	0.3271	0.5673	0.3338	0.5406	0.3084	0.4213	0.0661	0.28	0.5717	0.6528	0.535	0.7927	0.2647	0.4848	0.3895	0.6587
No log	15.0	195	8.2520	0.335	0.5597	0.3426	0.6311	0.3032	0.5211	0.0562	0.3073	0.58	0.65	0.5397	0.8435	0.2978	0.4948	0.3722	0.6652
No log	16.0	208	8.1700	0.3065	0.5151	0.308	0.6639	0.2903	0.4197	0.0568	0.3025	0.5802	0.7167	0.5423	0.7815	0.2632	0.4973	0.3498	0.663
No log	17.0	221	8.0011	0.3557	0.5819	0.3839	0.6318	0.3133	0.5577	0.061	0.3042	0.5911	0.6333	0.5585	0.796	0.2978	0.5084	0.4136	0.6739
No log	18.0	234	7.9459	0.3566	0.5752	0.3869	0.6684	0.3313	0.5498	0.0578	0.3123	0.614	0.7139	0.5762	0.8351	0.2848	0.5171	0.4284	0.7109
No log	19.0	247	7.9807	0.3556	0.5988	0.3768	0.6516	0.323	0.5591	0.0581	0.3112	0.5848	0.6694	0.5468	0.8177	0.2875	0.5088	0.4237	0.6609
No log	20.0	260	8.2510	0.3793	0.6172	0.3966	0.648	0.3517	0.527	0.0651	0.3108	0.5999	0.7139	0.5596	0.8294	0.2839	0.4846	0.4746	0.7152
No log	21.0	273	8.0173	0.3769	0.6308	0.3901	0.6581	0.348	0.5505	0.0537	0.3094	0.5865	0.7028	0.5526	0.7629	0.2872	0.4992	0.4666	0.6739
No log	22.0	286	8.1768	0.3486	0.5852	0.3254	0.6257	0.3176	0.5186	0.0645	0.3031	0.589	0.6611	0.5523	0.8109	0.2765	0.491	0.4207	0.687
No log	23.0	299	8.1836	0.3417	0.5674	0.3576	0.6202	0.3257	0.5427	0.0599	0.3157	0.5813	0.6639	0.5458	0.7895	0.2469	0.4931	0.4365	0.6696
No log	24.0	312	8.1504	0.3545	0.5978	0.3692	0.6435	0.3284	0.5889	0.058	0.3149	0.5825	0.6639	0.5447	0.8093	0.2753	0.491	0.4338	0.6739
No log	25.0	325	7.8980	0.3769	0.6108	0.4214	0.6247	0.3486	0.5719	0.0639	0.3131	0.5976	0.6583	0.5631	0.8085	0.2873	0.5017	0.4666	0.6935
No log	26.0	338	7.9985	0.3495	0.5762	0.3213	0.6342	0.3189	0.5759	0.0515	0.3032	0.6023	0.6583	0.5678	0.8294	0.2508	0.5046	0.4483	0.7
No log	27.0	351	7.8948	0.3738	0.6191	0.4456	0.6332	0.3502	0.5408	0.0643	0.3083	0.5954	0.6389	0.5632	0.8	0.2775	0.5104	0.4701	0.6804
No log	28.0	364	7.9396	0.3658	0.6121	0.3443	0.5331	0.3397	0.5885	0.0645	0.3105	0.5972	0.6444	0.5625	0.821	0.2744	0.5052	0.4571	0.6891
No log	29.0	377	7.9732	0.3522	0.5863	0.3728	0.5942	0.3264	0.5782	0.0627	0.3028	0.5956	0.65	0.5609	0.8177	0.2709	0.5086	0.4335	0.6826
No log	30.0	390	8.1223	0.3404	0.5936	0.3081	0.6316	0.3191	0.5283	0.05	0.2941	0.5913	0.6861	0.5534	0.8153	0.2474	0.4935	0.4334	0.6891
No log	31.0	403	7.9385	0.3279	0.5698	0.2987	0.5828	0.3084	0.5489	0.0579	0.2968	0.5883	0.65	0.5511	0.8294	0.2384	0.4962	0.4174	0.6804
No log	32.0	416	7.9805	0.3328	0.5745	0.3031	0.5977	0.3111	0.5616	0.0594	0.2946	0.5881	0.6667	0.5528	0.796	0.24	0.5023	0.4257	0.6739
No log	33.0	429	7.9656	0.3372	0.5745	0.3294	0.5334	0.3199	0.5599	0.0503	0.3032	0.5959	0.6639	0.5629	0.7911	0.2371	0.5092	0.4374	0.6826
No log	34.0	442	7.7492	0.3418	0.5842	0.3344	0.6367	0.3147	0.606	0.059	0.3082	0.5934	0.6667	0.5576	0.8101	0.2523	0.5106	0.4313	0.6761
No log	35.0	455	7.9115	0.3336	0.5782	0.3039	0.6315	0.3075	0.6255	0.0594	0.3012	0.5928	0.6639	0.5537	0.8435	0.2453	0.5073	0.4219	0.6783
No log	36.0	468	7.8646	0.3414	0.5791	0.3556	0.6197	0.3159	0.6261	0.0547	0.301	0.5994	0.6639	0.5627	0.8327	0.2447	0.5119	0.4381	0.687
No log	37.0	481	7.9918	0.3504	0.5905	0.3812	0.6226	0.3246	0.6256	0.0507	0.2982	0.5922	0.65	0.5576	0.8093	0.2403	0.5104	0.4604	0.6739
No log	38.0	494	7.9649	0.3396	0.5715	0.3579	0.615	0.317	0.584	0.0589	0.3046	0.6018	0.6611	0.5628	0.8585	0.2443	0.5079	0.4348	0.6957
13.3406	39.0	507	7.9482	0.328	0.5557	0.3271	0.5964	0.3056	0.5675	0.059	0.2983	0.5949	0.6444	0.5575	0.8444	0.2501	0.5115	0.4058	0.6783
13.3406	40.0	520	7.9461	0.3255	0.5618	0.3063	0.5913	0.305	0.5877	0.058	0.299	0.5888	0.6583	0.5516	0.8202	0.2427	0.5038	0.4084	0.6739
13.3406	41.0	533	7.9860	0.3074	0.5349	0.2944	0.6136	0.2866	0.5951	0.0507	0.2954	0.5946	0.6611	0.5582	0.821	0.2397	0.5067	0.3751	0.6826
13.3406	42.0	546	7.9835	0.3154	0.5395	0.3094	0.6094	0.2927	0.6031	0.0584	0.2971	0.5951	0.6583	0.559	0.8202	0.2425	0.5075	0.3883	0.6826
13.3406	43.0	559	8.0506	0.3225	0.5535	0.3255	0.5259	0.3027	0.6104	0.0584	0.296	0.5891	0.6444	0.5528	0.8218	0.2359	0.5021	0.409	0.6761
13.3406	44.0	572	8.0112	0.3321	0.5729	0.3363	0.6113	0.3089	0.5915	0.0516	0.2966	0.5892	0.6611	0.552	0.8194	0.2409	0.5023	0.4233	0.6761
13.3406	45.0	585	7.9716	0.3259	0.5551	0.3343	0.6304	0.3031	0.6173	0.0592	0.2969	0.5952	0.6583	0.559	0.821	0.2432	0.5056	0.4086	0.6848
13.3406	46.0	598	7.9923	0.3291	0.5642	0.3335	0.6106	0.3043	0.6146	0.0589	0.2943	0.5954	0.6806	0.5563	0.8343	0.2424	0.5061	0.4158	0.6848
13.3406	47.0	611	8.0093	0.3266	0.5565	0.3375	0.6017	0.3035	0.6104	0.0512	0.2999	0.5948	0.6583	0.5561	0.8452	0.2408	0.5048	0.4124	0.6848
13.3406	48.0	624	7.9778	0.3309	0.5626	0.3342	0.603	0.3062	0.6507	0.0588	0.2981	0.5993	0.675	0.5602	0.8444	0.2396	0.5029	0.4222	0.6957
13.3406	49.0	637	7.9963	0.3285	0.5612	0.3368	0.6042	0.3051	0.6123	0.0589	0.3028	0.5981	0.6583	0.5598	0.8468	0.2432	0.5071	0.4137	0.6891
13.3406	50.0	650	8.0013	0.3254	0.5528	0.3324	0.6086	0.3016	0.5975	0.0588	0.3009	0.5992	0.675	0.5598	0.8468	0.2404	0.505	0.4104	0.6935

Framework versions

• Transformers 5.2.0
• Pytorch 2.10.0+cu128
• Datasets 4.6.1
• Tokenizers 0.22.2