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: 0.8102
• Map: 0.5786
• Map 50: 0.8446
• Map 75: 0.6382
• Map Small: -1.0
• Map Medium: 0.4455
• Map Large: 0.612
• Mar 1: 0.4135
• Mar 10: 0.7102
• Mar 100: 0.7772
• Mar Small: -1.0
• Mar Medium: 0.6929
• Mar Large: 0.7901
• Map Banana: 0.4687
• Mar 100 Banana: 0.7725
• Map Orange: 0.5847
• Mar 100 Orange: 0.7762
• Map Apple: 0.6823
• Mar 100 Apple: 0.7829

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	1.8136	0.0276	0.0676	0.0239	-1.0	0.0078	0.0342	0.1148	0.2754	0.4185	-1.0	0.1714	0.4507	0.0408	0.485	0.0068	0.219	0.0351	0.5514
No log	2.0	120	1.7054	0.0355	0.1015	0.0111	-1.0	0.0269	0.0406	0.1291	0.2748	0.4096	-1.0	0.1786	0.4367	0.0311	0.515	0.0219	0.1595	0.0534	0.5543
No log	3.0	180	1.5536	0.0842	0.2199	0.0402	-1.0	0.0885	0.0896	0.1298	0.3777	0.523	-1.0	0.3643	0.546	0.0866	0.53	0.034	0.419	0.1319	0.62
No log	4.0	240	1.4179	0.1237	0.2493	0.1158	-1.0	0.1427	0.1403	0.202	0.435	0.584	-1.0	0.3786	0.6187	0.0677	0.5325	0.0733	0.5595	0.23	0.66
No log	5.0	300	1.3259	0.1914	0.3715	0.2017	-1.0	0.2276	0.1995	0.2675	0.4743	0.6013	-1.0	0.4214	0.6252	0.177	0.625	0.0701	0.4905	0.327	0.6886
No log	6.0	360	1.0490	0.3739	0.6071	0.4122	-1.0	0.275	0.4005	0.336	0.5812	0.6887	-1.0	0.5429	0.7081	0.2977	0.71	0.2726	0.6476	0.5513	0.7086
No log	7.0	420	1.0736	0.3657	0.6161	0.3907	-1.0	0.2635	0.3944	0.3273	0.5884	0.6913	-1.0	0.5071	0.7181	0.3167	0.695	0.2788	0.6619	0.5016	0.7171
No log	8.0	480	0.9415	0.4225	0.6716	0.4667	-1.0	0.3595	0.4549	0.3638	0.6513	0.7315	-1.0	0.6	0.7511	0.3356	0.7275	0.3724	0.7071	0.5596	0.76
1.3315	9.0	540	0.9699	0.4167	0.6212	0.4558	-1.0	0.3041	0.4531	0.3476	0.6281	0.7165	-1.0	0.5643	0.7385	0.3003	0.7175	0.3749	0.6976	0.5748	0.7343
1.3315	10.0	600	0.8980	0.4691	0.7169	0.5306	-1.0	0.3259	0.5058	0.3978	0.6631	0.7525	-1.0	0.6143	0.7731	0.3827	0.75	0.4542	0.7476	0.5705	0.76
1.3315	11.0	660	0.9776	0.4783	0.754	0.5379	-1.0	0.4189	0.5088	0.374	0.6489	0.7203	-1.0	0.6214	0.7373	0.3366	0.6975	0.4635	0.7119	0.6349	0.7514
1.3315	12.0	720	0.8537	0.4906	0.7546	0.5085	-1.0	0.3892	0.5281	0.3992	0.6637	0.7245	-1.0	0.5857	0.7456	0.3839	0.7225	0.4721	0.7167	0.6156	0.7343
1.3315	13.0	780	0.9508	0.489	0.771	0.5165	-1.0	0.3708	0.5242	0.3734	0.6578	0.7048	-1.0	0.5929	0.7231	0.3675	0.685	0.4847	0.7095	0.6147	0.72
1.3315	14.0	840	0.8707	0.5262	0.7754	0.5673	-1.0	0.4439	0.5601	0.3959	0.6783	0.7392	-1.0	0.6	0.7611	0.368	0.7275	0.5316	0.7214	0.6791	0.7686
1.3315	15.0	900	0.8676	0.524	0.7861	0.5509	-1.0	0.4334	0.5609	0.4157	0.6822	0.749	-1.0	0.6	0.7727	0.3959	0.7275	0.5311	0.7452	0.645	0.7743
1.3315	16.0	960	0.9307	0.5249	0.7949	0.5998	-1.0	0.5273	0.5483	0.4047	0.6706	0.7418	-1.0	0.7071	0.7476	0.3963	0.735	0.5469	0.7333	0.6314	0.7571
0.7366	17.0	1020	0.8927	0.5187	0.793	0.5634	-1.0	0.4054	0.5501	0.3921	0.6667	0.7375	-1.0	0.6357	0.7536	0.4097	0.725	0.5392	0.7476	0.6072	0.74
0.7366	18.0	1080	0.8342	0.5357	0.7734	0.5966	-1.0	0.4446	0.5642	0.4124	0.6984	0.7625	-1.0	0.6714	0.7765	0.4547	0.7575	0.5612	0.75	0.5913	0.78
0.7366	19.0	1140	0.8859	0.5298	0.7831	0.6049	-1.0	0.4868	0.5533	0.3974	0.6796	0.7596	-1.0	0.6929	0.7696	0.4249	0.7525	0.5349	0.7548	0.6295	0.7714
0.7366	20.0	1200	0.8419	0.541	0.7804	0.5784	-1.0	0.4396	0.5744	0.4146	0.7049	0.7669	-1.0	0.6786	0.7806	0.4128	0.7575	0.5499	0.7548	0.6602	0.7886
0.7366	21.0	1260	0.8121	0.5488	0.7972	0.5883	-1.0	0.4726	0.5808	0.4075	0.699	0.7654	-1.0	0.6429	0.7848	0.4376	0.75	0.5657	0.769	0.6431	0.7771
0.7366	22.0	1320	0.8244	0.5589	0.8228	0.5913	-1.0	0.4307	0.5956	0.4169	0.7031	0.7676	-1.0	0.65	0.7863	0.4263	0.7575	0.5662	0.7452	0.6844	0.8
0.7366	23.0	1380	0.8141	0.5664	0.8274	0.6123	-1.0	0.4366	0.6011	0.4235	0.7096	0.769	-1.0	0.6214	0.7913	0.4468	0.7675	0.5822	0.7595	0.6703	0.78
0.7366	24.0	1440	0.8117	0.5718	0.8355	0.6327	-1.0	0.4978	0.6039	0.42	0.7097	0.7812	-1.0	0.6571	0.7995	0.4668	0.785	0.5808	0.7643	0.6679	0.7943
0.5789	25.0	1500	0.8244	0.5733	0.8285	0.6347	-1.0	0.4932	0.6049	0.4133	0.706	0.7808	-1.0	0.6929	0.7943	0.4656	0.78	0.5824	0.7881	0.6719	0.7743
0.5789	26.0	1560	0.8051	0.5774	0.8352	0.6323	-1.0	0.4713	0.6104	0.4144	0.7068	0.7754	-1.0	0.6857	0.7892	0.4718	0.7725	0.5806	0.7738	0.6799	0.78
0.5789	27.0	1620	0.8131	0.5772	0.8438	0.6347	-1.0	0.4599	0.6098	0.4125	0.7069	0.7747	-1.0	0.6857	0.7887	0.468	0.7675	0.5868	0.7738	0.6769	0.7829
0.5789	28.0	1680	0.8100	0.5787	0.845	0.6392	-1.0	0.4471	0.6121	0.4134	0.7085	0.7741	-1.0	0.6857	0.7877	0.4728	0.7675	0.5839	0.769	0.6793	0.7857
0.5789	29.0	1740	0.8104	0.5786	0.8441	0.6391	-1.0	0.4455	0.6123	0.4135	0.7094	0.7747	-1.0	0.6929	0.7875	0.4699	0.7675	0.5845	0.7738	0.6814	0.7829
0.5789	30.0	1800	0.8102	0.5786	0.8446	0.6382	-1.0	0.4455	0.612	0.4135	0.7102	0.7772	-1.0	0.6929	0.7901	0.4687	0.7725	0.5847	0.7762	0.6823	0.7829

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.1
• Tokenizers 0.21.1