swin-tiny-patch4-window7-224-finetuned-lungs-disease

This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.2817
• Accuracy: 0.8746

Model description

This model was created by importing the dataset of the chest x-rays images into Google Colab from kaggle here:

https://www.kaggle.com/datasets/omkarmanohardalvi/lungs-disease-dataset-4-types .

I then used the image classification tutorial here:

https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/image_classification.ipynb

obtaining the following notebook:

https://colab.research.google.com/drive/1rNKeA25BR05iMUvKFvRD8SkySBOlO4AC?usp=sharing

The possible classified data are:

• Viral Pneumonia
• Corona Virus Disease
• Normal
• Tuberculosis
• Bacterial Pneumonia

X-rays image example:

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: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7851	0.98	21	0.4674	0.8152
0.4335	2.0	43	0.3662	0.8515
0.3231	2.98	64	0.3361	0.8581
0.3014	4.0	86	0.2817	0.8746
0.252	4.88	105	0.3071	0.8713

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2