user@devops:~$ cat README.md
PyTorch Fundamentals
# Description
Complete introduction to PyTorch, the most popular deep learning framework in research. Covers 7 sections: from tensors (creation, operations, broadcasting, NumPy interop) to convolutional neural networks. Includes autograd for automatic differentiation, linear regression with manual SGD and nn.Linear, a fully-connected neural network with nn.Module and DataLoader for Fashion MNIST (86.84% accuracy), and a CNN with nn.Conv2d and MaxPool2d (90.57% accuracy in just 3 epochs). Includes detailed PyTorch vs Keras/TensorFlow comparison, confusion matrix, per-class metrics, model persistence, and 7 visualizations.
# Key features
$ PyTorch tensors: creation, operations, broadcasting, NumPy interoperability
$ Autograd: automatic differentiation with chain rule and gradient visualization
$ Linear regression: manual SGD vs nn.Linear with Adam optimizer
$ Neural network with nn.Module and DataLoader for Fashion MNIST (86.84% accuracy)
$ CNN with nn.Conv2d, MaxPool2d and Dropout (90.57% accuracy in 3 epochs)
$ PyTorch vs Keras/TensorFlow side-by-side comparison with detailed table
$ 7 visualizations: tensors, autograd, regression, training curves, inference, confusion matrix, CNN vs MLP
# Gallery
# Technologies used
