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: 0.7836
• Map: 0.5785
• Map 50: 0.8356
• Map 75: 0.6723
• Map Small: -1.0
• Map Medium: 0.5125
• Map Large: 0.605
• Mar 1: 0.4248
• Mar 10: 0.7284
• Mar 100: 0.7686
• Mar Small: -1.0
• Mar Medium: 0.6125
• Mar Large: 0.7829
• Map Banana: 0.448
• Mar 100 Banana: 0.72
• Map Orange: 0.6045
• Mar 100 Orange: 0.7857
• Map Apple: 0.6831
• Mar 100 Apple: 0.8

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	2.2392	0.0133	0.0374	0.0065	-1.0	0.0006	0.0174	0.0367	0.1159	0.2228	-1.0	0.075	0.2375	0.0033	0.295	0.0055	0.019	0.031	0.3543
No log	2.0	120	1.8045	0.0433	0.094	0.035	-1.0	0.0841	0.0463	0.1148	0.2667	0.4661	-1.0	0.3708	0.4806	0.0131	0.425	0.0335	0.419	0.0834	0.5543
No log	3.0	180	1.7343	0.0758	0.1809	0.0542	-1.0	0.0666	0.0765	0.1559	0.3357	0.473	-1.0	0.3708	0.4901	0.0802	0.39	0.0401	0.4548	0.107	0.5743
No log	4.0	240	1.5930	0.0667	0.1545	0.0477	-1.0	0.0345	0.0729	0.1339	0.3051	0.4819	-1.0	0.2167	0.5061	0.0823	0.4875	0.0565	0.3524	0.0614	0.6057
No log	5.0	300	1.4399	0.08	0.1519	0.0659	-1.0	0.0812	0.0899	0.1599	0.327	0.5297	-1.0	0.35	0.5466	0.0811	0.4925	0.0724	0.4595	0.0867	0.6371
No log	6.0	360	1.2057	0.1493	0.2472	0.1804	-1.0	0.1378	0.1618	0.2595	0.4663	0.6235	-1.0	0.3542	0.6502	0.0964	0.5825	0.1548	0.6167	0.1967	0.6714
No log	7.0	420	1.1930	0.2454	0.4068	0.2628	-1.0	0.1931	0.2652	0.2975	0.4886	0.6008	-1.0	0.3625	0.6243	0.1301	0.53	0.2107	0.5952	0.3953	0.6771
No log	8.0	480	1.1520	0.3021	0.5017	0.3603	-1.0	0.2696	0.3272	0.3091	0.5556	0.6268	-1.0	0.4083	0.6477	0.136	0.57	0.2458	0.5905	0.5244	0.72
1.4531	9.0	540	1.0371	0.3781	0.5892	0.4062	-1.0	0.3088	0.3964	0.3496	0.6028	0.6662	-1.0	0.3958	0.6901	0.2285	0.63	0.3607	0.6429	0.5451	0.7257
1.4531	10.0	600	1.0391	0.3811	0.6249	0.4312	-1.0	0.2525	0.4061	0.3532	0.6144	0.6606	-1.0	0.4167	0.6837	0.2649	0.625	0.2871	0.631	0.5912	0.7257
1.4531	11.0	660	0.9947	0.4314	0.6884	0.4616	-1.0	0.2102	0.4734	0.3681	0.6204	0.678	-1.0	0.4	0.7046	0.2683	0.6025	0.449	0.7	0.5768	0.7314
1.4531	12.0	720	1.0551	0.4382	0.7558	0.4724	-1.0	0.2711	0.4696	0.339	0.6118	0.6658	-1.0	0.475	0.6833	0.2939	0.6325	0.4729	0.6762	0.5477	0.6886
1.4531	13.0	780	0.9251	0.4752	0.7361	0.5321	-1.0	0.3079	0.5056	0.3823	0.6394	0.7055	-1.0	0.4667	0.7265	0.333	0.6375	0.4894	0.6905	0.6033	0.7886
1.4531	14.0	840	0.8957	0.4906	0.7363	0.5688	-1.0	0.34	0.5195	0.3813	0.6715	0.7187	-1.0	0.5208	0.7345	0.3125	0.66	0.52	0.7333	0.6394	0.7629
1.4531	15.0	900	0.9153	0.4978	0.7646	0.5708	-1.0	0.41	0.5297	0.401	0.6679	0.7131	-1.0	0.5708	0.7275	0.3437	0.6275	0.5364	0.7548	0.6133	0.7571
1.4531	16.0	960	0.8663	0.5276	0.7993	0.576	-1.0	0.3697	0.5634	0.4088	0.6738	0.7315	-1.0	0.525	0.7493	0.3965	0.675	0.5225	0.731	0.6638	0.7886
0.7981	17.0	1020	0.8745	0.5359	0.8136	0.5912	-1.0	0.3684	0.5684	0.4217	0.6903	0.7463	-1.0	0.5458	0.765	0.3881	0.68	0.5621	0.7762	0.6575	0.7829
0.7981	18.0	1080	0.8692	0.5375	0.814	0.6356	-1.0	0.4627	0.5653	0.4139	0.6979	0.7461	-1.0	0.6083	0.76	0.3799	0.6825	0.5793	0.7786	0.6532	0.7771
0.7981	19.0	1140	0.8285	0.5488	0.8236	0.6288	-1.0	0.4448	0.5802	0.4215	0.7103	0.7608	-1.0	0.6542	0.7699	0.4209	0.7175	0.574	0.7762	0.6513	0.7886
0.7981	20.0	1200	0.8036	0.5544	0.8123	0.6339	-1.0	0.4699	0.5869	0.4227	0.7209	0.7735	-1.0	0.625	0.7859	0.4012	0.7175	0.5806	0.8	0.6815	0.8029
0.7981	21.0	1260	0.8163	0.5546	0.8194	0.6187	-1.0	0.4976	0.5843	0.426	0.7134	0.7648	-1.0	0.6083	0.781	0.3824	0.6925	0.6011	0.8048	0.6803	0.7971
0.7981	22.0	1320	0.8323	0.5608	0.8266	0.6316	-1.0	0.5279	0.5848	0.4161	0.711	0.7573	-1.0	0.6083	0.7706	0.4091	0.6975	0.5902	0.7857	0.6831	0.7886
0.7981	23.0	1380	0.8178	0.5621	0.83	0.6621	-1.0	0.4861	0.5881	0.4194	0.7124	0.7578	-1.0	0.6125	0.7707	0.4356	0.71	0.5775	0.7833	0.6733	0.78
0.7981	24.0	1440	0.8000	0.5615	0.8331	0.66	-1.0	0.5107	0.5872	0.4135	0.7153	0.7615	-1.0	0.5917	0.7765	0.4259	0.725	0.5974	0.7738	0.6611	0.7857
0.5872	25.0	1500	0.7918	0.5691	0.8323	0.6611	-1.0	0.5043	0.5945	0.4271	0.7258	0.7671	-1.0	0.6	0.7824	0.4274	0.7175	0.5935	0.781	0.6863	0.8029
0.5872	26.0	1560	0.7879	0.5846	0.839	0.674	-1.0	0.4845	0.611	0.4234	0.7313	0.7656	-1.0	0.6208	0.7789	0.457	0.7125	0.6081	0.7786	0.6888	0.8057
0.5872	27.0	1620	0.7810	0.5793	0.8423	0.664	-1.0	0.485	0.6038	0.4285	0.7251	0.7736	-1.0	0.6167	0.7865	0.4498	0.735	0.6025	0.7857	0.6857	0.8
0.5872	28.0	1680	0.7838	0.5779	0.8359	0.6719	-1.0	0.5125	0.6044	0.424	0.7256	0.7666	-1.0	0.6125	0.7803	0.4494	0.725	0.6017	0.7833	0.6827	0.7914
0.5872	29.0	1740	0.7841	0.5776	0.8363	0.6718	-1.0	0.5125	0.604	0.4248	0.7276	0.7678	-1.0	0.6125	0.782	0.4479	0.72	0.6019	0.7833	0.6829	0.8
0.5872	30.0	1800	0.7836	0.5785	0.8356	0.6723	-1.0	0.5125	0.605	0.4248	0.7284	0.7686	-1.0	0.6125	0.7829	0.448	0.72	0.6045	0.7857	0.6831	0.8

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1