finetuned-vietnamese-food

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the indian_vietnam_images dataset. It achieves the following results on the evaluation set:

• Loss: 0.3760
• Accuracy: 0.8958

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: 0.0002
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.1058	0.0910	100	1.9974	0.5694
1.4012	0.1820	200	1.4076	0.6855
1.3551	0.2730	300	1.1650	0.7264
1.1111	0.3640	400	1.0998	0.7062
1.0038	0.4550	500	0.9087	0.7483
0.9599	0.5460	600	0.8278	0.7682
1.0932	0.6369	700	0.9115	0.7360
0.7807	0.7279	800	0.8011	0.7730
0.8237	0.8189	900	0.8345	0.7726
0.7288	0.9099	1000	0.6427	0.8258
0.7982	1.0009	1100	0.6427	0.8215
0.7331	1.0919	1200	0.6423	0.8183
0.6849	1.1829	1300	0.6820	0.8151
0.671	1.2739	1400	0.6325	0.8191
0.7307	1.3649	1500	0.6079	0.8286
0.7499	1.4559	1600	0.5832	0.8346
0.7004	1.5469	1700	0.6048	0.8342
0.7543	1.6379	1800	0.5612	0.8394
0.5557	1.7288	1900	0.5740	0.8318
0.5019	1.8198	2000	0.5064	0.8561
0.7043	1.9108	2100	0.5513	0.8441
0.519	2.0018	2200	0.5862	0.8350
0.3366	2.0928	2300	0.5159	0.8517
0.4167	2.1838	2400	0.5386	0.8469
0.402	2.2748	2500	0.5614	0.8374
0.4133	2.3658	2600	0.4756	0.8652
0.4751	2.4568	2700	0.4882	0.8612
0.3108	2.5478	2800	0.4946	0.8648
0.3218	2.6388	2900	0.4707	0.8680
0.282	2.7298	3000	0.4407	0.8712
0.2823	2.8207	3100	0.4843	0.8712
0.3498	2.9117	3200	0.4609	0.8744
0.3196	3.0027	3300	0.4369	0.8763
0.2822	3.0937	3400	0.4662	0.8748
0.4166	3.1847	3500	0.4539	0.8779
0.1904	3.2757	3600	0.4205	0.8887
0.388	3.3667	3700	0.4163	0.8863
0.2851	3.4577	3800	0.4168	0.8891
0.2455	3.5487	3900	0.4004	0.8930
0.2804	3.6397	4000	0.4044	0.8938
0.2008	3.7307	4100	0.3833	0.8950
0.2487	3.8217	4200	0.3812	0.8958
0.2077	3.9126	4300	0.3760	0.8958

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3