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: 0.8075
• Map: 0.5492
• Map 50: 0.8129
• Map 75: 0.6184
• Map Small: -1.0
• Map Medium: 0.5412
• Map Large: 0.5745
• Mar 1: 0.4367
• Mar 10: 0.7285
• Mar 100: 0.7829
• Mar Small: -1.0
• Mar Medium: 0.7643
• Mar Large: 0.7895
• Map Banana: 0.4035
• Mar 100 Banana: 0.73
• Map Orange: 0.5513
• Mar 100 Orange: 0.7929
• Map Apple: 0.6929
• Mar 100 Apple: 0.8257

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	2.1930	0.0025	0.0072	0.0016	-1.0	0.0007	0.0035	0.0161	0.0825	0.2242	-1.0	0.0429	0.2487	0.0015	0.2925	0.0	0.0	0.0061	0.38
No log	2.0	120	1.9326	0.011	0.0299	0.0064	-1.0	0.0044	0.0134	0.0758	0.2107	0.3813	-1.0	0.1214	0.416	0.0137	0.45	0.0062	0.1738	0.0131	0.52
No log	3.0	180	1.6307	0.0352	0.0947	0.0195	-1.0	0.0531	0.0355	0.1342	0.303	0.503	-1.0	0.3214	0.5277	0.0413	0.5375	0.0424	0.3714	0.0218	0.6
No log	4.0	240	1.6542	0.0558	0.1344	0.0522	-1.0	0.1515	0.0482	0.0944	0.2671	0.4604	-1.0	0.35	0.4725	0.0524	0.5075	0.0662	0.3881	0.0487	0.4857
No log	5.0	300	1.6691	0.0388	0.1063	0.0274	-1.0	0.0944	0.0364	0.1583	0.2932	0.4751	-1.0	0.35	0.4917	0.043	0.5	0.0359	0.3452	0.0374	0.58
No log	6.0	360	1.1086	0.0826	0.1345	0.0841	-1.0	0.2117	0.0861	0.2797	0.4782	0.7029	-1.0	0.5929	0.7177	0.071	0.7225	0.1168	0.6262	0.06	0.76
No log	7.0	420	1.1675	0.0814	0.165	0.072	-1.0	0.2427	0.086	0.2683	0.4722	0.6522	-1.0	0.4929	0.6746	0.0837	0.6575	0.0851	0.5619	0.0754	0.7371
No log	8.0	480	1.0365	0.1206	0.207	0.1248	-1.0	0.2508	0.1171	0.3042	0.5348	0.7123	-1.0	0.6143	0.7282	0.0767	0.6875	0.1441	0.681	0.1411	0.7686
1.512	9.0	540	1.0794	0.1506	0.2487	0.1703	-1.0	0.2685	0.1558	0.3506	0.5771	0.6842	-1.0	0.4857	0.7162	0.0875	0.645	0.1822	0.6619	0.1823	0.7457
1.512	10.0	600	0.9685	0.2052	0.3178	0.2417	-1.0	0.3075	0.2088	0.3638	0.5795	0.713	-1.0	0.5571	0.7386	0.1142	0.66	0.2011	0.6619	0.3002	0.8171
1.512	11.0	660	1.0193	0.2702	0.4348	0.3242	-1.0	0.3423	0.287	0.3652	0.6083	0.6889	-1.0	0.6429	0.699	0.1441	0.64	0.263	0.6952	0.4036	0.7314
1.512	12.0	720	0.9402	0.3339	0.5175	0.3808	-1.0	0.358	0.3523	0.3898	0.637	0.7244	-1.0	0.6286	0.7421	0.2116	0.67	0.3413	0.7262	0.4489	0.7771
1.512	13.0	780	0.9065	0.4067	0.6265	0.4574	-1.0	0.5061	0.4159	0.3831	0.6531	0.7409	-1.0	0.6286	0.76	0.2899	0.705	0.3526	0.7262	0.5776	0.7914
1.512	14.0	840	0.8992	0.4333	0.6571	0.4951	-1.0	0.5391	0.4405	0.3823	0.679	0.7469	-1.0	0.6929	0.7585	0.2879	0.6975	0.4142	0.7405	0.5978	0.8029
1.512	15.0	900	0.9158	0.4523	0.6711	0.5006	-1.0	0.567	0.457	0.3885	0.6792	0.7503	-1.0	0.7429	0.7554	0.3111	0.6875	0.4015	0.7548	0.6444	0.8086
1.512	16.0	960	0.8610	0.4903	0.7499	0.5371	-1.0	0.5782	0.4965	0.4083	0.6934	0.7603	-1.0	0.75	0.7656	0.3468	0.6975	0.4646	0.769	0.6594	0.8143
0.8176	17.0	1020	0.8541	0.5026	0.7497	0.5756	-1.0	0.6024	0.509	0.4079	0.7004	0.7741	-1.0	0.7429	0.783	0.363	0.7175	0.5092	0.7905	0.6356	0.8143
0.8176	18.0	1080	0.8627	0.4944	0.7614	0.5615	-1.0	0.58	0.5081	0.4067	0.6975	0.7571	-1.0	0.7	0.7686	0.3636	0.715	0.501	0.7476	0.6185	0.8086
0.8176	19.0	1140	0.8270	0.5227	0.7928	0.5967	-1.0	0.589	0.5339	0.4137	0.7212	0.7767	-1.0	0.7143	0.789	0.3864	0.735	0.5444	0.781	0.6372	0.8143
0.8176	20.0	1200	0.8100	0.5428	0.807	0.629	-1.0	0.5925	0.561	0.4291	0.7177	0.7721	-1.0	0.7571	0.7787	0.4188	0.7125	0.553	0.7952	0.6567	0.8086
0.8176	21.0	1260	0.8255	0.5424	0.8012	0.6145	-1.0	0.5723	0.5611	0.4269	0.7175	0.7674	-1.0	0.7286	0.7775	0.3995	0.7075	0.5572	0.7833	0.6703	0.8114
0.8176	22.0	1320	0.8203	0.5447	0.8214	0.6081	-1.0	0.5544	0.567	0.4308	0.7186	0.7785	-1.0	0.75	0.7863	0.3999	0.7275	0.5527	0.7881	0.6815	0.82
0.8176	23.0	1380	0.8116	0.555	0.8196	0.6291	-1.0	0.5953	0.569	0.4345	0.7297	0.7793	-1.0	0.75	0.7874	0.4045	0.725	0.5768	0.7929	0.6836	0.82
0.8176	24.0	1440	0.8178	0.5431	0.7922	0.6252	-1.0	0.5631	0.5629	0.4217	0.7217	0.7755	-1.0	0.7357	0.7849	0.3973	0.7275	0.5501	0.7905	0.6819	0.8086
0.6165	25.0	1500	0.8056	0.5533	0.8126	0.6213	-1.0	0.569	0.5718	0.43	0.7249	0.779	-1.0	0.75	0.787	0.4049	0.7275	0.5546	0.781	0.7002	0.8286
0.6165	26.0	1560	0.7900	0.556	0.8143	0.6332	-1.0	0.552	0.5828	0.4417	0.7364	0.7812	-1.0	0.7643	0.7873	0.418	0.7325	0.55	0.7881	0.7002	0.8229
0.6165	27.0	1620	0.8072	0.5466	0.8125	0.6105	-1.0	0.5431	0.5733	0.4367	0.7327	0.7787	-1.0	0.7571	0.786	0.4051	0.725	0.5465	0.7881	0.688	0.8229
0.6165	28.0	1680	0.8077	0.5481	0.8135	0.6199	-1.0	0.5418	0.5725	0.4351	0.7263	0.783	-1.0	0.7643	0.7899	0.4018	0.7275	0.5497	0.7929	0.6927	0.8286
0.6165	29.0	1740	0.8085	0.5488	0.812	0.618	-1.0	0.541	0.574	0.4367	0.7277	0.7837	-1.0	0.7643	0.7905	0.4021	0.73	0.5514	0.7952	0.6929	0.8257
0.6165	30.0	1800	0.8075	0.5492	0.8129	0.6184	-1.0	0.5412	0.5745	0.4367	0.7285	0.7829	-1.0	0.7643	0.7895	0.4035	0.73	0.5513	0.7929	0.6929	0.8257

Framework versions

• Transformers 4.51.1
• Pytorch 2.6.0+cu124
• Datasets 2.14.4
• Tokenizers 0.21.1