vit-base-patch16-224-dmae-va-da

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: 0.4510
• Accuracy: 0.8372

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: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 16

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.92	3	1.3352	0.3488
No log	1.85	6	1.1552	0.4186
1.2778	2.77	9	1.0285	0.4651
1.2778	4.0	13	0.9331	0.5116
0.8889	4.92	16	0.7961	0.6279
0.8889	5.85	19	0.7570	0.6977
0.8889	6.77	22	0.6943	0.6977
0.6605	8.0	26	0.6077	0.7442
0.6605	8.92	29	0.5718	0.7209
0.484	9.85	32	0.5346	0.7674
0.484	10.77	35	0.5174	0.7907
0.484	12.0	39	0.4780	0.8140
0.3193	12.92	42	0.4510	0.8372
0.3193	13.85	45	0.4390	0.8372
0.3161	14.77	48	0.4346	0.8372

Framework versions

• Transformers 4.34.1
• Pytorch 2.1.0+cu118
• Datasets 2.14.5
• Tokenizers 0.14.1