cvt-13

This model is a fine-tuned version of microsoft/cvt-13 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.0361
• Accuracy: 0.9886

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.4048	1.0	327	0.2161	0.9156
0.33	2.0	654	0.1320	0.9501
0.3147	3.0	981	0.1060	0.9612
0.2213	4.0	1309	0.0820	0.9742
0.3256	5.0	1636	0.0717	0.9750
0.3207	6.0	1963	0.1062	0.9626
0.2273	7.0	2290	0.0535	0.9797
0.2066	8.0	2618	0.0566	0.9817
0.2162	9.0	2945	0.0459	0.9828
0.2296	10.0	3272	0.0444	0.9851
0.187	11.0	3599	0.0348	0.9882
0.2208	12.0	3927	0.0505	0.9848
0.1855	13.0	4254	0.0371	0.9869
0.1875	14.0	4581	0.0384	0.9880
0.202	14.99	4905	0.0361	0.9886

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0