State-of-the-art Vision Transformer 🤗 for image classification, High Accuracy 🚀

Fashion product images dataset

44k products with multiple category labels, descriptions and high-res images. Dataset Link | | Kaggle Notebook

Training result

Epoch	Training Loss	Validation Loss	Accuracy
1	0.082200	0.059544	0.992506
2	0.042800	0.035637	0.995316
3	0.033300	0.030343	0.995472

>>> trainer.train()

# Evaluate the model on the test set
>>> outputs = trainer.predict(test_data)
>>> outputs.metrics
[{'test_loss': 0.03034268133342266, 'test_accuracy': 0.9954722872755659, 'test_runtime': 102.7551}]

5 Flower types classification dataset

Flower Images for Classification (1k images for each flower types) Dataset Link | Kaggle Notebook

Training result

Epoch	Training Loss	Validation Loss	Accuracy
1	No log	0.349625	0.974667
2	No log	0.184932	0.984000
3	No log	0.137928	0.989333
4	0.369800	0.122166	0.988000
5	0.369800	0.117171	0.988000

>>> trainer.train()

# Evaluate the model on the test set
>>> outputs = trainer.predict(test_data)
>>> outputs.metrics
[{'test_loss': 0.13792766630, 'test_accuracy': 0.98933333, 'test_runtime': 12.5379}]

Brain MRI Images dataset

It conists of MRI scans of two classes: NO - no tumor, encoded as 0 YES - tumor, encoded as 1 Dataset Link | Kaggle Notebook

Training result

Epoch	Training Loss	Validation Loss	Accuracy
1	No log	0.533956	0.769231
2	No log	0.454096	0.876923
3	No log	0.389601	0.969231
4	No log	0.347400	0.984615
5	No log	0.336078	0.984615

>>> trainer.train()

# Evaluate the model on the test set
>>> outputs = trainer.predict(test_data)
>>> outputs.metrics
[{'test_loss': 0.347400099, 'test_accuracy': 0.984615384, 'test_runtime': 0.8473}]