README.md

metadata

base_model:
  - apple/mobilevit-small
model_name: mobilevit_small-chest_xray
tags:
  - sft
  - vision
  - medical-imaging
  - chest-xray
  - pneumonia
  - covid19
  - tuberculosis
license: apache-2.0
language:
  - en
pipeline_tag: image-classification
library_name: transformers

🫁 MobileViT-Small — Chest X-Ray Classifier (LoRA Fine-tune)

This model is a LoRA fine-tuned version of apple/mobilevit-small
for chest X-ray classification on Chest X-ray Pneumonia, COVID-19 & Tuberculosis dataset.

---

🧩 Configuration

Attribute	Value
Base Model	apple/mobilevit-small
Fine-tuning Method	LoRA (Low-Rank Adaptation)
Dataset	Chest X-ray Pneumonia/COVID-19/Tuberculosis
Classes	COVID19, NORMAL, PNEUMONIA, TUBERCULOSIS
Epochs	10
Batch Size	32
Learning Rate	0.0005
Optimizer	AdamW
LoRA Config	r=4, α=32, dropout=0.1, target_modules=[query, value]
Trainable Params	30.9K / 4.97M (0.62%)

---

🫁 Example Image

---

📊 Evaluation Results

Metric	Avg	COVID19	NORMAL	PNEUMONIA	TUBERCULOSIS
Accuracy	0.9798	0.9909	0.9661	0.9648	0.9974
Precision	0.9616	0.9640	0.9130	0.9696	1.0000
Recall	0.9591	0.9727	0.9245	0.9673	0.9718
F1 Score	0.9603	0.9683	0.9188	0.9684	0.9857
AUC	0.9968	0.9991	0.9927	0.9959	0.9997

Test loss: 0.0938  Inference time: 0.476 s