mobileViTV2-64

This model is a fine-tuned version of apple/mobilevitv2-1.0-imagenet1k-256 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3307
• Accuracy: 0.9106
• F1: 0.9093
• Precision: 0.9118
• Recall: 0.9106

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: 3e-05
• train_batch_size: 16
• eval_batch_size: 16
• 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: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
1.6065	1.0	364	1.6103	0.1860	0.1866	0.1893	0.1860
1.5721	2.0	728	1.5804	0.2851	0.2857	0.2889	0.2851
1.4793	3.0	1092	1.5111	0.4229	0.4057	0.4006	0.4229
1.316	4.0	1456	1.2841	0.5028	0.4652	0.5157	0.5028
1.1397	5.0	1820	1.0520	0.5909	0.5474	0.6502	0.5909
0.8639	6.0	2184	0.8194	0.7163	0.6970	0.7325	0.7163
0.7371	7.0	2548	0.6773	0.7796	0.7711	0.7927	0.7796
0.6451	8.0	2912	0.5546	0.8292	0.8258	0.8334	0.8292
0.5299	9.0	3276	0.4800	0.8485	0.8461	0.8477	0.8485
0.387	10.0	3640	0.4091	0.8760	0.8737	0.8759	0.8760
0.3903	11.0	4004	0.3547	0.8884	0.8867	0.8887	0.8884
0.3513	12.0	4368	0.3207	0.9008	0.8983	0.9038	0.9008
0.3145	13.0	4732	0.3213	0.8967	0.8939	0.9009	0.8967
0.1838	14.0	5096	0.3013	0.8939	0.8924	0.8941	0.8939
0.3438	15.0	5460	0.3229	0.8857	0.8843	0.8850	0.8857
0.1913	16.0	5824	0.2568	0.9174	0.9159	0.9191	0.9174
0.2078	17.0	6188	0.2609	0.9187	0.9169	0.9206	0.9187
0.2061	18.0	6552	0.2811	0.9077	0.9061	0.9076	0.9077
0.2806	19.0	6916	0.2536	0.9242	0.9230	0.9260	0.9242
0.2495	20.0	7280	0.2881	0.9091	0.9076	0.9094	0.9091
0.0361	21.0	7644	0.2875	0.9311	0.9301	0.9331	0.9311
0.1811	22.0	8008	0.3067	0.9063	0.9050	0.9056	0.9063
0.1129	23.0	8372	0.2996	0.9050	0.9047	0.9053	0.9050
0.1138	24.0	8736	0.2970	0.9063	0.9060	0.9066	0.9063
0.3135	25.0	9100	0.3723	0.8967	0.8968	0.8972	0.8967
0.0828	26.0	9464	0.3574	0.9063	0.9060	0.9059	0.9063
0.0783	27.0	9828	0.4087	0.8939	0.8926	0.8926	0.8939
0.051	28.0	10192	0.3713	0.9063	0.9060	0.9068	0.9063
0.0744	29.0	10556	0.4470	0.8953	0.8951	0.8958	0.8953
0.0814	30.0	10920	0.4289	0.9077	0.9085	0.9099	0.9077
0.131	31.0	11284	0.4600	0.9008	0.8996	0.8997	0.9008
0.0245	32.0	11648	0.4818	0.8981	0.8978	0.8977	0.8981
0.0541	33.0	12012	0.4678	0.9050	0.9043	0.9040	0.9050
0.1011	34.0	12376	0.5298	0.8994	0.8985	0.8991	0.8994
0.17	35.0	12740	0.5093	0.9036	0.9026	0.9024	0.9036
0.0892	36.0	13104	0.5018	0.9063	0.9050	0.9050	0.9063
0.0246	37.0	13468	0.5520	0.9077	0.9058	0.9061	0.9077
0.0564	38.0	13832	0.5493	0.9077	0.9075	0.9077	0.9077
0.0817	39.0	14196	0.5607	0.9091	0.9084	0.9081	0.9091
0.0056	40.0	14560	0.5990	0.8939	0.8947	0.8961	0.8939
0.0653	41.0	14924	0.5870	0.9146	0.9136	0.9136	0.9146
0.1649	42.0	15288	0.5882	0.9050	0.9040	0.9039	0.9050
0.1057	43.0	15652	0.5924	0.9008	0.8999	0.9002	0.9008
0.0859	44.0	16016	0.5830	0.8994	0.8994	0.8999	0.8994
0.1809	45.0	16380	0.6357	0.8953	0.8939	0.8936	0.8953
0.1285	46.0	16744	0.6617	0.8967	0.8965	0.8975	0.8967
0.1018	47.0	17108	0.6006	0.9050	0.9044	0.9042	0.9050
0.0091	48.0	17472	0.5762	0.9091	0.9090	0.9094	0.9091
0.0368	49.0	17836	0.6097	0.9077	0.9067	0.9071	0.9077
0.0585	50.0	18200	0.6059	0.9063	0.9059	0.9061	0.9063
0.0373	51.0	18564	0.6621	0.8953	0.8953	0.8963	0.8953
0.1672	52.0	18928	0.6081	0.9022	0.9020	0.9019	0.9022
0.0344	53.0	19292	0.6145	0.8994	0.9002	0.9011	0.8994
0.0727	54.0	19656	0.6106	0.9036	0.9034	0.9034	0.9036
0.1997	55.0	20020	0.6037	0.9091	0.9082	0.9090	0.9091
0.0437	56.0	20384	0.5835	0.9105	0.9100	0.9105	0.9105
0.0263	57.0	20748	0.6032	0.9063	0.9062	0.9064	0.9063
0.056	58.0	21112	0.5828	0.9105	0.9101	0.9102	0.9105
0.0422	59.0	21476	0.6179	0.9105	0.9111	0.9129	0.9105
0.0377	60.0	21840	0.6400	0.8981	0.8997	0.9027	0.8981
0.1162	61.0	22204	0.5841	0.9105	0.9106	0.9108	0.9105
0.0407	62.0	22568	0.6017	0.9063	0.9064	0.9067	0.9063
0.0443	63.0	22932	0.6064	0.9036	0.9031	0.9029	0.9036
0.089	64.0	23296	0.6250	0.9008	0.9011	0.9018	0.9008
0.0971	65.0	23660	0.6729	0.9022	0.9011	0.9018	0.9022
0.046	66.0	24024	0.6445	0.9063	0.9060	0.9062	0.9063
0.0387	67.0	24388	0.6070	0.9036	0.9039	0.9046	0.9036
0.0709	68.0	24752	0.5890	0.9132	0.9131	0.9132	0.9132
0.0273	69.0	25116	0.6484	0.9008	0.9001	0.9007	0.9008
0.1951	70.0	25480	0.6336	0.9077	0.9075	0.9075	0.9077
0.0569	71.0	25844	0.6546	0.9105	0.9104	0.9105	0.9105
0.1145	72.0	26208	0.6964	0.9036	0.9026	0.9027	0.9036
0.0352	73.0	26572	0.6657	0.9118	0.9114	0.9115	0.9118
0.0375	74.0	26936	0.6417	0.9050	0.9054	0.9059	0.9050
0.0351	75.0	27300	0.6812	0.9091	0.9077	0.9081	0.9091
0.0675	76.0	27664	0.6445	0.9105	0.9100	0.9103	0.9105
0.0418	77.0	28028	0.7359	0.9091	0.9073	0.9084	0.9091
0.0353	78.0	28392	0.6701	0.9022	0.9021	0.9022	0.9022
0.0221	79.0	28756	0.6607	0.9036	0.9039	0.9051	0.9036
0.2435	80.0	29120	0.6487	0.9118	0.9115	0.9114	0.9118
0.0362	81.0	29484	0.7711	0.9077	0.9060	0.9064	0.9077
0.0116	82.0	29848	0.6276	0.9063	0.9067	0.9073	0.9063
0.001	83.0	30212	0.6564	0.9022	0.9020	0.9022	0.9022
0.013	84.0	30576	0.6576	0.9077	0.9071	0.9072	0.9077
0.0183	85.0	30940	0.7075	0.9036	0.9033	0.9038	0.9036
0.0367	86.0	31304	0.7168	0.9118	0.9100	0.9108	0.9118
0.027	87.0	31668	0.6892	0.9132	0.9133	0.9143	0.9132
0.015	88.0	32032	0.6886	0.9077	0.9069	0.9069	0.9077
0.0435	89.0	32396	0.6863	0.9008	0.9009	0.9012	0.9008
0.0049	90.0	32760	0.6883	0.9077	0.9072	0.9069	0.9077
0.1041	91.0	33124	0.7216	0.9008	0.9000	0.9002	0.9008
0.0465	92.0	33488	0.7032	0.9022	0.9021	0.9026	0.9022
0.0221	93.0	33852	0.7131	0.9036	0.9025	0.9023	0.9036
0.0091	94.0	34216	0.6886	0.8953	0.8963	0.8976	0.8953
0.0322	95.0	34580	0.7213	0.9022	0.9020	0.9024	0.9022
0.0348	96.0	34944	0.7005	0.9022	0.9016	0.9014	0.9022
0.0357	97.0	35308	0.7131	0.8967	0.8971	0.8980	0.8967
0.0363	98.0	35672	0.6947	0.9118	0.9112	0.9114	0.9118
0.0249	99.0	36036	0.6783	0.9132	0.9126	0.9126	0.9132
0.0179	100.0	36400	0.6614	0.9036	0.9037	0.9039	0.9036

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.4.0
• Tokenizers 0.21.0