vit-base-patch16-224-in21k_brain_tumor_diagnosis

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: 0.2591
• Accuracy: 0.9216
• F1: 0.9375
• Recall: 1.0
• Precision: 0.8824

Model description

This is a binary classification model to distinguish between if the MRI images detect a brain tumor or not.

For more information on how it was created, check out the following link: https://github.com/DunnBC22/Vision_Audio_and_Multimodal_Projects/blob/main/Computer%20Vision/Image%20Classification/Binary%20Classification/Brain%20Tumor%20MRI%20Images/brain_tumor_MRI_Images_ViT.ipynb

Intended uses & limitations

This model is intended to demonstrate my ability to solve a complex problem using technology.

Training and evaluation data

Dataset Source: https://www.kaggle.com/datasets/navoneel/brain-mri-images-for-brain-tumor-detection

Sample Images From Dataset:

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Recall	Precision
0.7101	1.0	13	0.3351	0.9412	0.9474	0.9	1.0
0.7101	2.0	26	0.3078	0.9020	0.9231	1.0	0.8571
0.7101	3.0	39	0.2591	0.9216	0.9375	1.0	0.8824
0.7101	4.0	52	0.2702	0.9020	0.9123	0.8667	0.9630
0.7101	5.0	65	0.2855	0.9020	0.9123	0.8667	0.9630

Framework versions

• Transformers 4.25.1
• Pytorch 1.12.1
• Datasets 2.8.0
• Tokenizers 0.12.1