vit-retail-finetuned

This model is a fine-tuned version of google/vit-base-patch16-224 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3202
• Accuracy: 0.9142

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: 32
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1377	0.0806	100	1.0902	0.7582
0.6781	0.1613	200	0.7387	0.8147
0.7764	0.2419	300	0.6461	0.8265
0.6088	0.3226	400	0.5804	0.8385
0.5444	0.4032	500	0.5562	0.8496
0.4501	0.4839	600	0.4864	0.8575
0.4771	0.5645	700	0.4509	0.8718
0.4475	0.6452	800	0.4453	0.8668
0.3776	0.7258	900	0.4300	0.8745
0.4127	0.8065	1000	0.4002	0.8823
0.3597	0.8871	1100	0.4132	0.8725
0.3164	0.9677	1200	0.3987	0.8784
0.3018	1.0484	1300	0.3707	0.8907
0.2185	1.1290	1400	0.3781	0.8852
0.2867	1.2097	1500	0.3803	0.8829
0.2212	1.2903	1600	0.3814	0.8882
0.2206	1.3710	1700	0.3577	0.8959
0.2099	1.4516	1800	0.3572	0.8943
0.2366	1.5323	1900	0.3582	0.8934
0.2610	1.6129	2000	0.3346	0.9040
0.2403	1.6935	2100	0.3367	0.8970
0.1953	1.7742	2200	0.3325	0.9009
0.1959	1.8548	2300	0.3309	0.9047
0.2398	1.9355	2400	0.3193	0.9079
0.1428	2.0161	2500	0.3120	0.9063
0.1302	2.0968	2600	0.3178	0.9097
0.1019	2.1774	2700	0.3200	0.9083
0.1076	2.2581	2800	0.3258	0.9093
0.0654	2.3387	2900	0.3305	0.9056
0.0923	2.4194	3000	0.3355	0.9095
0.0778	2.5	3100	0.3247	0.9083
0.0681	2.5806	3200	0.3287	0.9097
0.1084	2.6613	3300	0.3253	0.9106
0.1011	2.7419	3400	0.3226	0.9113
0.0895	2.8226	3500	0.3251	0.9122
0.0768	2.9032	3600	0.3210	0.9142
0.0834	2.9839	3700	0.3202	0.9142

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2