rtdetr-r50-fruits3.1-finetune

This model is a fine-tuned version of PekingU/rtdetr_v2_r50vd on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 10.2825
• Map: 0.5141
• Map 50: 0.6438
• Map 75: 0.5328
• Map Small: 0.1834
• Map Medium: 0.4552
• Map Large: 0.6833
• Mar 1: 0.2766
• Mar 10: 0.6043
• Mar 100: 0.6896
• Mar Small: 0.3069
• Mar Medium: 0.653
• Mar Large: 0.8601
• Map Apple: 0.484
• Mar 100 Apple: 0.6951
• Map Banana: 0.5239
• Mar 100 Banana: 0.7047
• Map Grapes: 0.4436
• Mar 100 Grapes: 0.5755
• Map Orange: 0.4851
• Mar 100 Orange: 0.6356
• Map Pineapple: 0.6032
• Mar 100 Pineapple: 0.7605
• Map Watermelon: 0.5445
• Mar 100 Watermelon: 0.7659

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 with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 300
• num_epochs: 100
• 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 Apple	Mar 100 Apple	Map Banana	Mar 100 Banana	Map Grapes	Mar 100 Grapes	Map Orange	Mar 100 Orange	Map Pineapple	Mar 100 Pineapple	Map Watermelon	Mar 100 Watermelon
61.0927	1.0	750	12.3536	0.4105	0.5309	0.4348	0.1007	0.3523	0.5936	0.2516	0.5606	0.6943	0.3265	0.6804	0.8648	0.3696	0.6673	0.4721	0.7284	0.3272	0.5616	0.3893	0.626	0.4278	0.7525	0.4771	0.8298
16.2131	2.0	1500	11.0204	0.4591	0.5898	0.4801	0.1266	0.3931	0.6455	0.2608	0.5789	0.7038	0.343	0.6833	0.8654	0.4085	0.6891	0.4473	0.7299	0.3675	0.5923	0.4612	0.6384	0.5151	0.7489	0.5548	0.8241
15.0214	3.0	2250	10.6170	0.4864	0.6169	0.5035	0.1523	0.4532	0.651	0.2665	0.5951	0.7111	0.3693	0.6834	0.8752	0.4727	0.701	0.4647	0.7427	0.3902	0.5812	0.4707	0.6376	0.5272	0.7851	0.5933	0.8188
13.8951	4.0	3000	10.3428	0.5106	0.6355	0.5364	0.1468	0.4302	0.7136	0.2779	0.5933	0.6975	0.3156	0.6572	0.8758	0.4983	0.6914	0.4684	0.7223	0.4134	0.5737	0.4944	0.6373	0.5702	0.7678	0.6189	0.7926
13.1909	5.0	3750	10.0188	0.5284	0.6543	0.5485	0.1648	0.4766	0.7106	0.2813	0.6024	0.723	0.3977	0.6936	0.8861	0.5002	0.6963	0.5068	0.7582	0.4229	0.6168	0.5036	0.6445	0.6046	0.7812	0.6321	0.8406
12.6709	6.0	4500	9.9985	0.5344	0.6675	0.5558	0.1833	0.4766	0.7188	0.2757	0.6071	0.7241	0.3929	0.6945	0.8888	0.5139	0.7059	0.4884	0.7343	0.4506	0.6095	0.521	0.6597	0.5997	0.7891	0.6326	0.8457
12.2203	7.0	5250	10.2048	0.5242	0.6494	0.5433	0.178	0.4519	0.7161	0.2805	0.6061	0.7126	0.351	0.675	0.8877	0.4993	0.6946	0.5214	0.7222	0.4247	0.5999	0.4953	0.6465	0.5884	0.7826	0.6158	0.8295
12.0783	8.0	6000	10.1179	0.5268	0.6511	0.5493	0.1827	0.4863	0.7003	0.2771	0.6061	0.7276	0.3949	0.7093	0.8875	0.5119	0.7065	0.4949	0.7407	0.4276	0.6063	0.4987	0.6544	0.5789	0.7971	0.649	0.8608
11.7674	9.0	6750	10.7223	0.4952	0.6141	0.5192	0.1485	0.4494	0.6737	0.2707	0.6003	0.7203	0.397	0.69	0.8796	0.4633	0.6996	0.4973	0.7448	0.4313	0.5888	0.4631	0.6606	0.4962	0.7714	0.6198	0.8565
11.5219	10.0	7500	9.8389	0.5221	0.6564	0.5438	0.2016	0.461	0.7026	0.2753	0.6044	0.7138	0.3662	0.6795	0.8808	0.4921	0.6998	0.4981	0.731	0.4604	0.6223	0.5007	0.6329	0.5846	0.7768	0.5966	0.8202
11.1748	11.0	8250	10.1057	0.5227	0.6501	0.5488	0.2067	0.4508	0.7094	0.276	0.6117	0.7201	0.3771	0.6965	0.881	0.5098	0.7075	0.4953	0.7257	0.4535	0.6093	0.4955	0.6459	0.5596	0.7967	0.6225	0.8355
11.0897	12.0	9000	10.4848	0.5181	0.6473	0.5365	0.1853	0.4529	0.6962	0.2812	0.5989	0.6958	0.3296	0.6539	0.8736	0.4742	0.676	0.5002	0.7085	0.455	0.5944	0.4842	0.6337	0.5777	0.7373	0.6171	0.8247
10.8801	13.0	9750	10.0386	0.5219	0.6487	0.541	0.1772	0.4602	0.7022	0.2822	0.6146	0.7172	0.3145	0.6881	0.8917	0.4978	0.6926	0.5425	0.7467	0.4405	0.6068	0.4968	0.6402	0.5561	0.7967	0.5974	0.8205
10.723	14.0	10500	9.5538	0.5517	0.6795	0.5827	0.205	0.4932	0.7338	0.2815	0.6206	0.7203	0.3765	0.6777	0.8922	0.526	0.6924	0.5612	0.7605	0.4625	0.6355	0.5177	0.6492	0.6413	0.7783	0.6012	0.806
10.5762	15.0	11250	9.9150	0.5278	0.6531	0.5492	0.1654	0.4583	0.7103	0.2817	0.608	0.7124	0.328	0.6863	0.8814	0.5079	0.6937	0.5213	0.74	0.4296	0.5788	0.5046	0.6458	0.5681	0.788	0.6354	0.8278
10.5457	16.0	12000	9.7727	0.5418	0.6709	0.5595	0.2069	0.4866	0.7144	0.2766	0.6027	0.698	0.3201	0.6541	0.8743	0.4888	0.671	0.5528	0.7216	0.4597	0.5941	0.4876	0.6375	0.5958	0.7475	0.666	0.8165
10.3053	17.0	12750	9.9212	0.5392	0.6723	0.5627	0.1865	0.5009	0.7033	0.2806	0.6087	0.7101	0.3284	0.68	0.8813	0.4932	0.6922	0.5369	0.7298	0.458	0.598	0.5006	0.6409	0.5976	0.7681	0.6489	0.8315
10.2234	18.0	13500	9.7854	0.5276	0.6601	0.5511	0.1782	0.4737	0.6981	0.2792	0.61	0.7155	0.3208	0.6896	0.8851	0.5198	0.7052	0.5038	0.724	0.4565	0.6081	0.5157	0.6548	0.5671	0.7746	0.6024	0.8264
10.0754	19.0	14250	10.5111	0.5086	0.6329	0.531	0.1631	0.4495	0.6713	0.2769	0.6035	0.7092	0.3326	0.6738	0.8797	0.4743	0.6977	0.5051	0.7214	0.443	0.601	0.4704	0.6357	0.5976	0.7663	0.5609	0.8332
10.1197	20.0	15000	10.2968	0.5204	0.6463	0.5452	0.1705	0.4586	0.7002	0.2755	0.6036	0.7082	0.3294	0.6713	0.885	0.4837	0.6886	0.4481	0.7191	0.4491	0.5994	0.5103	0.6457	0.5882	0.7678	0.6432	0.8284
9.8649	21.0	15750	10.0439	0.5264	0.6576	0.5482	0.1852	0.4709	0.7035	0.2796	0.5943	0.697	0.3317	0.6524	0.8778	0.4649	0.6727	0.4915	0.7099	0.4386	0.5786	0.498	0.6446	0.6342	0.7696	0.631	0.8065
9.8367	22.0	16500	9.5550	0.5584	0.6869	0.5826	0.2292	0.5105	0.7348	0.2825	0.618	0.7228	0.3802	0.6957	0.8876	0.5113	0.6976	0.5429	0.7419	0.4542	0.6117	0.5285	0.6559	0.634	0.7841	0.6797	0.8455
9.5876	23.0	17250	9.8796	0.5391	0.6607	0.563	0.1863	0.4824	0.7163	0.2796	0.6184	0.7192	0.3546	0.6988	0.884	0.4933	0.6959	0.5088	0.7413	0.4553	0.6001	0.513	0.6559	0.6116	0.7779	0.6525	0.8443
9.5507	24.0	18000	10.0149	0.5351	0.6661	0.5565	0.2108	0.4703	0.7099	0.2782	0.6077	0.706	0.3393	0.6614	0.8794	0.5201	0.699	0.5335	0.7254	0.4731	0.6213	0.5071	0.6384	0.6214	0.7779	0.5554	0.7741
9.4775	25.0	18750	9.9049	0.538	0.6679	0.5583	0.1898	0.4768	0.7084	0.2808	0.6184	0.7181	0.3543	0.6864	0.8851	0.5175	0.6982	0.5342	0.7324	0.4569	0.6153	0.5212	0.6541	0.5982	0.7884	0.6002	0.8205
9.4565	26.0	19500	10.0631	0.5303	0.6553	0.5532	0.1805	0.4694	0.6984	0.281	0.6136	0.7016	0.325	0.6713	0.8758	0.5116	0.6836	0.5145	0.7229	0.4634	0.5911	0.4719	0.6312	0.5988	0.7678	0.6215	0.8128
9.3008	27.0	20250	9.9243	0.5324	0.6673	0.5509	0.1461	0.4834	0.6986	0.2792	0.6119	0.7063	0.3311	0.6805	0.8749	0.4854	0.7004	0.5318	0.7143	0.449	0.5907	0.4865	0.6382	0.5817	0.7641	0.6602	0.8304
9.1329	28.0	21000	10.0756	0.5404	0.675	0.5647	0.1808	0.4816	0.7115	0.2807	0.6161	0.7134	0.3472	0.6905	0.8782	0.5153	0.6949	0.5222	0.7309	0.4616	0.6116	0.5094	0.6427	0.6026	0.7808	0.631	0.8196
9.1302	29.0	21750	9.9070	0.548	0.6764	0.5735	0.1968	0.5044	0.7151	0.2826	0.6176	0.7138	0.3358	0.6884	0.8799	0.5129	0.7091	0.5441	0.7292	0.4818	0.6084	0.4875	0.6393	0.6194	0.7667	0.6423	0.8301
9.0189	30.0	22500	10.0898	0.5376	0.6634	0.5571	0.1898	0.489	0.705	0.2798	0.6117	0.7035	0.3247	0.6689	0.8763	0.5088	0.6922	0.5051	0.702	0.4494	0.588	0.4821	0.6381	0.6578	0.7928	0.6225	0.8082
8.9913	31.0	23250	10.9136	0.4988	0.6284	0.5167	0.128	0.452	0.667	0.2759	0.5892	0.6775	0.2474	0.643	0.8679	0.4517	0.6456	0.4671	0.6929	0.4257	0.5472	0.4507	0.611	0.5799	0.7757	0.6177	0.7923
8.7717	32.0	24000	10.2825	0.5141	0.6438	0.5328	0.1834	0.4552	0.6833	0.2766	0.6043	0.6896	0.3069	0.653	0.8601	0.484	0.6951	0.5239	0.7047	0.4436	0.5755	0.4851	0.6356	0.6032	0.7605	0.5445	0.7659

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1