Welcome to my PyTorch deep learning practice repo!
This is a daily learning log based on the Zero to Mastery PyTorch course by Daniel Bourke + hands-on projects.
🎯 Mission: Build a strong, self-reliant foundation in deep learning — from fundamentals to real-world projects.
| Day | Notebook | Highlights |
|---|---|---|
| D01 | D1_tensors_operations.ipynb |
Tensor basics, creation, operations, GPU usage, shape manipulation |
| D02 | day02_reproducibility_device_agnostic.ipynb |
Reproducibility, random seeds, device-agnostic code, best practices, PyTorch docs, quickstart overview |
| D03 | day03_revision_exercises_docs_quickstart.ipynb |
Revision, solved exercises, explored PyTorch documentation, practiced with the Quickstart Guide |
| D04 | day04_model_building_essentials.ipynb |
Built first model, reviewed PyTorch model structure, revised Python OOP, explored extra learning resources |
| D05 | day05_pytorch_workflow_fundamentals.ipynb |
Learned PyTorch’s complete model training workflow: data → model → training → testing → saving & loading |
| D06 | day06_neural_network_data_preparation.ipynb |
Created custom data, converted it into tensors, set up groundwork for building a neural network |
| D07–08 | day07_activation_function_non_linear.ipynb |
Built linear & non-linear models, learned about activation functions, manually replicated non-linearity, completed full neural network training pipeline |
| D09 | FER2013_CNN_Inference.ipynb(day-09) |
Started a Facial Expression Recognition (FER2013) project using CNNs: loaded & explored FER2013 dataset, defined transforms, created data loaders, designed & trained CNN, evaluated on test set, saved model, and tested predictions on custom images the saved model is there in the MODELS/ folder in the github |
- 🗂️ Day-wise notebooks on PyTorch core concepts and projects
- 🔢 Tensor ops, model building, training, evaluation, CNNs
- 🧪 Mini-experiments, practical code, and problem-solving
- 📋 Clear commit history for learning traceability
- ✅ Master the core of PyTorch
- ✅ Build and train deep neural networks from scratch
- ✅ Apply for AI research assistant/intern roles
- ✅ Contribute to open-source AI/ML projects
- ✅ Become independent and confident in AI development
- Language: Python
- Framework: PyTorch
- Platform: Google Colab
- Version Control: Git + GitHub
- 🎓 Zero to Mastery – Deep Learning with PyTorch
- 🧾 PyTorch Cheatsheet
- 🧠 Mindset Boosters:
📌 I'm sharing updates, insights, and experiments on X (Twitter).
If you're on a similar path, feel free to fork this repo and join me!
Let’s learn, build, and grow — one day at a time. 🔥