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: 0.7771
• Map: 0.5882
• Map 50: 0.8376
• Map 75: 0.6723
• Map Small: -1.0
• Map Medium: 0.6116
• Map Large: 0.5966
• Mar 1: 0.4201
• Mar 10: 0.7111
• Mar 100: 0.7683
• Mar Small: -1.0
• Mar Medium: 0.7071
• Mar Large: 0.7767
• Map Banana: 0.4758
• Mar 100 Banana: 0.7425
• Map Orange: 0.6281
• Mar 100 Orange: 0.8024
• Map Apple: 0.6608
• Mar 100 Apple: 0.76

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	1.9700	0.0096	0.0268	0.0038	-1.0	0.0155	0.0132	0.078	0.2026	0.3463	-1.0	0.2343	0.3714	0.0132	0.2975	0.0096	0.3786	0.0058	0.3629
No log	2.0	120	1.6517	0.0553	0.1516	0.0414	-1.0	0.111	0.0556	0.1359	0.2777	0.4308	-1.0	0.3186	0.4454	0.0647	0.5175	0.0406	0.1976	0.0608	0.5771
No log	3.0	180	1.2778	0.1262	0.2428	0.1168	-1.0	0.1877	0.1303	0.2519	0.5055	0.6286	-1.0	0.5814	0.634	0.1024	0.6225	0.0983	0.4976	0.1778	0.7657
No log	4.0	240	1.0948	0.2377	0.4041	0.2352	-1.0	0.4084	0.2402	0.3266	0.5759	0.7115	-1.0	0.6371	0.7237	0.182	0.695	0.1717	0.7024	0.3596	0.7371
No log	5.0	300	1.0477	0.2746	0.4623	0.2895	-1.0	0.2475	0.3142	0.3285	0.609	0.7315	-1.0	0.6257	0.7458	0.221	0.7075	0.1828	0.7214	0.42	0.7657
No log	6.0	360	1.0028	0.3661	0.6059	0.4064	-1.0	0.4221	0.3982	0.3651	0.6231	0.7251	-1.0	0.6229	0.7379	0.2698	0.7	0.3568	0.7238	0.4716	0.7514
No log	7.0	420	0.9809	0.3532	0.5656	0.4002	-1.0	0.4557	0.3731	0.3569	0.6472	0.7488	-1.0	0.6829	0.7591	0.3239	0.715	0.3333	0.7714	0.4025	0.76
No log	8.0	480	0.9679	0.4348	0.6762	0.4868	-1.0	0.5782	0.4375	0.3547	0.6527	0.7254	-1.0	0.7343	0.7269	0.2877	0.68	0.4769	0.7619	0.5397	0.7343
1.2471	9.0	540	0.9173	0.4434	0.7005	0.5049	-1.0	0.5147	0.4475	0.3646	0.6443	0.7348	-1.0	0.6771	0.7408	0.3288	0.7225	0.4683	0.7619	0.5332	0.72
1.2471	10.0	600	0.8875	0.4834	0.7654	0.5497	-1.0	0.5051	0.4991	0.369	0.6925	0.7589	-1.0	0.6957	0.7689	0.3668	0.73	0.497	0.7952	0.5864	0.7514
1.2471	11.0	660	0.9261	0.4803	0.7507	0.5799	-1.0	0.4907	0.4971	0.3818	0.6745	0.7525	-1.0	0.6957	0.7629	0.3567	0.7175	0.5014	0.7714	0.5828	0.7686
1.2471	12.0	720	0.8520	0.4974	0.7451	0.5567	-1.0	0.6198	0.4976	0.3946	0.691	0.7489	-1.0	0.7157	0.7532	0.3709	0.7025	0.5588	0.7929	0.5626	0.7514
1.2471	13.0	780	0.8630	0.4998	0.7799	0.5682	-1.0	0.546	0.5213	0.3848	0.6848	0.7519	-1.0	0.6443	0.768	0.4078	0.7575	0.5624	0.7952	0.5292	0.7029
1.2471	14.0	840	0.8469	0.5071	0.776	0.5801	-1.0	0.6247	0.5104	0.3913	0.7049	0.7579	-1.0	0.6971	0.7682	0.3635	0.71	0.5271	0.781	0.6306	0.7829
1.2471	15.0	900	0.7995	0.5311	0.8059	0.5856	-1.0	0.6156	0.5327	0.3958	0.7068	0.7576	-1.0	0.7429	0.7592	0.3951	0.7175	0.5739	0.8095	0.6244	0.7457
1.2471	16.0	960	0.8150	0.5342	0.8046	0.6189	-1.0	0.6285	0.5346	0.3974	0.7012	0.7505	-1.0	0.7043	0.7556	0.4157	0.73	0.584	0.7929	0.603	0.7286
0.7135	17.0	1020	0.7887	0.5532	0.8155	0.6643	-1.0	0.5982	0.5619	0.4184	0.7122	0.7656	-1.0	0.6929	0.7758	0.4475	0.7425	0.5754	0.8	0.6365	0.7543
0.7135	18.0	1080	0.7961	0.5545	0.8237	0.6426	-1.0	0.6024	0.5606	0.4042	0.7056	0.7583	-1.0	0.6971	0.7648	0.4583	0.7425	0.6036	0.8095	0.6014	0.7229
0.7135	19.0	1140	0.7936	0.5726	0.8321	0.6599	-1.0	0.6004	0.5838	0.4203	0.7209	0.7776	-1.0	0.7071	0.7878	0.4648	0.75	0.5835	0.8	0.6695	0.7829
0.7135	20.0	1200	0.7948	0.5543	0.8208	0.638	-1.0	0.5928	0.5617	0.4001	0.7032	0.7665	-1.0	0.7	0.7747	0.4439	0.7525	0.5944	0.8071	0.6246	0.74
0.7135	21.0	1260	0.7850	0.5808	0.8357	0.6736	-1.0	0.5831	0.5941	0.4118	0.7229	0.7766	-1.0	0.7	0.7863	0.4928	0.765	0.6112	0.8048	0.6386	0.76
0.7135	22.0	1320	0.8025	0.5813	0.8356	0.6729	-1.0	0.6177	0.5906	0.4188	0.7138	0.771	-1.0	0.6871	0.7812	0.4719	0.755	0.6277	0.7952	0.6442	0.7629
0.7135	23.0	1380	0.7886	0.5795	0.83	0.6743	-1.0	0.5957	0.589	0.4076	0.7065	0.7598	-1.0	0.69	0.7679	0.4784	0.75	0.624	0.7952	0.6362	0.7343
0.7135	24.0	1440	0.8081	0.5787	0.8341	0.6563	-1.0	0.5982	0.5875	0.4117	0.7084	0.7679	-1.0	0.7114	0.7748	0.463	0.745	0.6192	0.7929	0.6538	0.7657
0.5383	25.0	1500	0.7858	0.5865	0.8318	0.6691	-1.0	0.6285	0.5935	0.4216	0.7144	0.7729	-1.0	0.7186	0.7792	0.473	0.75	0.624	0.8	0.6626	0.7686
0.5383	26.0	1560	0.7777	0.5935	0.8462	0.6778	-1.0	0.6176	0.6011	0.4216	0.7151	0.7709	-1.0	0.7143	0.7784	0.4799	0.7475	0.6363	0.8024	0.6643	0.7629
0.5383	27.0	1620	0.7821	0.5914	0.8388	0.6746	-1.0	0.6231	0.5982	0.4209	0.7128	0.7685	-1.0	0.7043	0.7771	0.4773	0.7375	0.6304	0.8024	0.6665	0.7657
0.5383	28.0	1680	0.7803	0.5918	0.8401	0.6739	-1.0	0.6233	0.5987	0.4201	0.7129	0.7684	-1.0	0.7143	0.7759	0.4768	0.74	0.6328	0.8024	0.6658	0.7629
0.5383	29.0	1740	0.7800	0.5886	0.8382	0.6727	-1.0	0.6116	0.5971	0.4201	0.7111	0.7683	-1.0	0.7071	0.7767	0.476	0.7425	0.629	0.8024	0.6608	0.76
0.5383	30.0	1800	0.7771	0.5882	0.8376	0.6723	-1.0	0.6116	0.5966	0.4201	0.7111	0.7683	-1.0	0.7071	0.7767	0.4758	0.7425	0.6281	0.8024	0.6608	0.76

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1