vit-emotion_classification

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

• Loss: 1.2112
• Accuracy: 0.5938

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	40	1.8928	0.375
No log	2.0	80	1.5709	0.375
No log	3.0	120	1.4385	0.4938
No log	4.0	160	1.3183	0.5437
No log	5.0	200	1.2514	0.5813
No log	6.0	240	1.2412	0.5563
No log	7.0	280	1.2048	0.5875
No log	8.0	320	1.1530	0.6188
No log	9.0	360	1.1870	0.55
No log	10.0	400	1.2160	0.5563
No log	11.0	440	1.1182	0.5563
No log	12.0	480	1.1162	0.5938
1.0857	13.0	520	1.0960	0.6312
1.0857	14.0	560	1.1724	0.55
1.0857	15.0	600	1.1100	0.625

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.0
• Tokenizers 0.15.2