yolo_finetuned_fruits

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

• Loss: 0.8208
• Map: 0.5539
• Map 50: 0.8071
• Map 75: 0.6043
• Map Small: -1.0
• Map Medium: 0.4804
• Map Large: 0.5761
• Mar 1: 0.409
• Mar 10: 0.7106
• Mar 100: 0.7748
• Mar Small: -1.0
• Mar Medium: 0.6829
• Mar Large: 0.7861
• Map Banana: 0.4114
• Mar 100 Banana: 0.775
• Map Orange: 0.6102
• Mar 100 Orange: 0.781
• Map Apple: 0.6401
• Mar 100 Apple: 0.7686

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: 4
• 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
• num_epochs: 30

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 Banana	Mar 100 Banana	Map Orange	Mar 100 Orange	Map Apple	Mar 100 Apple
No log	1.0	60	2.1986	0.0068	0.0254	0.0016	-1.0	0.0068	0.0079	0.0246	0.0997	0.2776	-1.0	0.24	0.283	0.0109	0.2575	0.0002	0.0095	0.0092	0.5657
No log	2.0	120	1.9727	0.0088	0.03	0.0036	-1.0	0.0201	0.0089	0.0521	0.1605	0.3185	-1.0	0.26	0.3186	0.0163	0.4325	0.0	0.0	0.0103	0.5229
No log	3.0	180	1.9117	0.0353	0.114	0.0137	-1.0	0.0335	0.0411	0.1015	0.2692	0.4279	-1.0	0.28	0.4458	0.0185	0.415	0.0278	0.3714	0.0596	0.4971
No log	4.0	240	1.6734	0.0659	0.1642	0.0544	-1.0	0.1162	0.0783	0.1647	0.3225	0.4596	-1.0	0.28	0.4787	0.0818	0.485	0.0324	0.2452	0.0836	0.6486
No log	5.0	300	1.3011	0.1225	0.2534	0.1145	-1.0	0.1155	0.156	0.2833	0.4893	0.5985	-1.0	0.42	0.6231	0.0858	0.5575	0.0939	0.5238	0.1879	0.7143
No log	6.0	360	1.2643	0.2057	0.356	0.2293	-1.0	0.2614	0.2286	0.3177	0.5069	0.6091	-1.0	0.4843	0.6289	0.1166	0.5425	0.1363	0.5333	0.3641	0.7514
No log	7.0	420	1.1581	0.281	0.4787	0.2868	-1.0	0.3952	0.2874	0.3263	0.577	0.6758	-1.0	0.54	0.6973	0.139	0.6025	0.2267	0.619	0.4773	0.8057
No log	8.0	480	1.1026	0.3086	0.524	0.3347	-1.0	0.2653	0.3326	0.3576	0.58	0.6648	-1.0	0.5943	0.6766	0.2161	0.615	0.2935	0.631	0.4162	0.7486
1.4697	9.0	540	1.0055	0.3516	0.5724	0.3781	-1.0	0.3764	0.3613	0.3457	0.6023	0.7044	-1.0	0.6629	0.7125	0.2457	0.645	0.3506	0.7024	0.4585	0.7657
1.4697	10.0	600	0.9545	0.4136	0.6261	0.4555	-1.0	0.3712	0.4388	0.3688	0.6483	0.73	-1.0	0.6671	0.7413	0.2924	0.68	0.4384	0.75	0.51	0.76
1.4697	11.0	660	0.9475	0.423	0.6493	0.4547	-1.0	0.5066	0.4345	0.3763	0.662	0.7468	-1.0	0.6429	0.7622	0.2579	0.71	0.456	0.7476	0.5551	0.7829
1.4697	12.0	720	0.9563	0.4131	0.6719	0.4431	-1.0	0.4135	0.4285	0.3598	0.6447	0.7194	-1.0	0.5957	0.7354	0.3076	0.71	0.4745	0.731	0.4573	0.7171
1.4697	13.0	780	0.8893	0.4472	0.6689	0.4985	-1.0	0.4739	0.4567	0.3983	0.6573	0.7334	-1.0	0.6443	0.7447	0.3567	0.735	0.4538	0.7595	0.5309	0.7057
1.4697	14.0	840	0.9049	0.4915	0.7427	0.5237	-1.0	0.415	0.5107	0.3922	0.6898	0.7536	-1.0	0.6529	0.7674	0.3643	0.7375	0.5229	0.7405	0.5872	0.7829
1.4697	15.0	900	0.8799	0.4884	0.7419	0.5376	-1.0	0.4822	0.5042	0.3963	0.6875	0.7565	-1.0	0.6614	0.7686	0.3481	0.7525	0.5076	0.7571	0.6095	0.76
1.4697	16.0	960	0.8778	0.5014	0.7714	0.5549	-1.0	0.5352	0.5127	0.4015	0.6808	0.744	-1.0	0.6329	0.7593	0.3398	0.725	0.5527	0.75	0.6116	0.7571
0.7568	17.0	1020	0.8810	0.5025	0.7664	0.5708	-1.0	0.506	0.5126	0.3919	0.6854	0.7424	-1.0	0.6743	0.7518	0.3768	0.7325	0.5336	0.7405	0.5973	0.7543
0.7568	18.0	1080	0.8716	0.4942	0.7505	0.5653	-1.0	0.4833	0.509	0.3965	0.6756	0.7391	-1.0	0.6357	0.7515	0.374	0.7525	0.5074	0.719	0.6011	0.7457
0.7568	19.0	1140	0.8007	0.5072	0.7516	0.5666	-1.0	0.4698	0.524	0.411	0.7079	0.757	-1.0	0.6486	0.7697	0.3868	0.7625	0.5498	0.7429	0.5849	0.7657
0.7568	20.0	1200	0.8122	0.5502	0.8115	0.594	-1.0	0.4834	0.575	0.4175	0.7223	0.7704	-1.0	0.6486	0.7855	0.436	0.765	0.6078	0.769	0.6067	0.7771
0.7568	21.0	1260	0.8067	0.5387	0.7907	0.5869	-1.0	0.505	0.5602	0.3976	0.72	0.7725	-1.0	0.6486	0.7874	0.3823	0.7775	0.6032	0.7857	0.6306	0.7543
0.7568	22.0	1320	0.8331	0.5408	0.7992	0.5769	-1.0	0.4986	0.5614	0.4017	0.71	0.7596	-1.0	0.6614	0.7726	0.4037	0.745	0.5779	0.7595	0.6408	0.7743
0.7568	23.0	1380	0.8336	0.5386	0.7938	0.5854	-1.0	0.4914	0.56	0.4017	0.713	0.7625	-1.0	0.6657	0.7751	0.3928	0.75	0.5954	0.769	0.6277	0.7686
0.7568	24.0	1440	0.8137	0.5391	0.7978	0.593	-1.0	0.4835	0.5612	0.4081	0.7134	0.7681	-1.0	0.6714	0.7807	0.3796	0.7625	0.6057	0.7762	0.6321	0.7657
0.5523	25.0	1500	0.8126	0.5518	0.8009	0.5998	-1.0	0.4901	0.5745	0.4082	0.7152	0.7745	-1.0	0.6757	0.7869	0.3933	0.7725	0.6199	0.7881	0.6423	0.7629
0.5523	26.0	1560	0.8205	0.5528	0.8115	0.6105	-1.0	0.4859	0.5733	0.4063	0.711	0.7727	-1.0	0.7	0.7819	0.4121	0.77	0.6125	0.7881	0.6338	0.76
0.5523	27.0	1620	0.8211	0.5503	0.8075	0.6082	-1.0	0.4756	0.5729	0.4081	0.7088	0.7748	-1.0	0.7	0.7843	0.4064	0.77	0.6134	0.7857	0.6312	0.7686
0.5523	28.0	1680	0.8223	0.5543	0.8091	0.6061	-1.0	0.4809	0.5771	0.4082	0.7081	0.7758	-1.0	0.6929	0.7862	0.4103	0.7725	0.6136	0.7833	0.639	0.7714
0.5523	29.0	1740	0.8171	0.5531	0.806	0.6037	-1.0	0.4803	0.5755	0.409	0.7106	0.774	-1.0	0.6829	0.7852	0.4113	0.775	0.6079	0.7786	0.6401	0.7686
0.5523	30.0	1800	0.8208	0.5539	0.8071	0.6043	-1.0	0.4804	0.5761	0.409	0.7106	0.7748	-1.0	0.6829	0.7861	0.4114	0.775	0.6102	0.781	0.6401	0.7686

Framework versions

• Transformers 4.50.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1