Deep Learning

Hands-on Jupyter notebooks for deep learning with TensorFlow — from MNIST classification through gradient descent, TensorBoard, GPU benchmarks, and applied tabular projects.

Live site (GitHub Pages): https://lokeshpuma.github.io/Deep_Learning/

Repository layout

Path	Description
notebooks/	Jupyter notebooks (run these locally or on Colab)
data/	CSV datasets referenced by notebooks
src/	Static site source for the GitHub Pages frontend
frontend/	Deployable copy of src/ (synced before publish)
.github/workflows/deploy-pages.yml	Builds and deploys frontend/ to GitHub Pages

Run notebooks locally

python -m venv .venv
source .venv/bin/activate   # Windows: .venv\Scripts\activate
pip install tensorflow pandas matplotlib jupyter
jupyter lab

Open notebooks from the notebooks/ folder. CSV paths assume the working directory is the repo root (e.g. data/customer_churn.csv).

Edit and deploy the site

1. Change files under src/.

2. Sync the deployable folder:

   bash scripts/sync-frontend.sh

3. Push to main. The Deploy GitHub Pages workflow publishes frontend/.

Enable GitHub Pages (one-time)

In the repo on GitHub: Settings → Pages → Build and deployment → Source: GitHub Actions.

Notebooks

1. Digits Classification (MNIST)
2. Activation Functions
3. Matrix Operations
4. Loss Functions
5. Gradient Descent
6. GD & SGD
7. TensorBoard
8. GPU Benchmarking
9. Customer Churn Prediction
10. PPT Generator
11. Regression