Deep Learning with PyTorch
Build neural networks with modern PyTorch: tensors and autograd, training loops, CNNs and transfer learning, RNNs/LSTMs for sequences, word embeddings and sentiment analysis, then autoencoders and GANs. Train interactively in the browser playground, run the real thing in Colab.
01Neural Networks
What deep learning is, tensors and autograd, the training loop, and your first networks in PyTorch.
- What Is Deep Learning?Where deep learning sits inside AI and ML, what a neuron actually computes, and why nonlinear activations give networks their power. 9 min read
- PyTorch & TensorsCreate and reshape tensors, bridge to NumPy, move to the GPU, and let autograd compute derivatives for you. 8 min read
- The Training LoopModel, criterion, optimizer — the three ingredients of training — and the forward-loss-backward-step dance that turns gradients into learning. 9 min read
- Building Networks with nn.ModuleA crash course in Python classes, the anatomy of nn.Module, and an end-to-end MLP classifier — from raw data to accuracy. 9 min read
- Minibatches, Datasets & DataLoadersWhy we train on small batches, how Dataset and DataLoader feed them to the model, and the full modern training recipe with validation, checkpointing, and early stopping. 8 min read
02Convolutional Neural Networks
Convolutions, pooling, and building CNN image classifiers with data augmentation.
- How Convolutions WorkWhy dense layers fail on images, what a convolution actually computes, and the arithmetic of channels, stride, padding, and pooling. 9 min read
- Building a CNN in PyTorchTrain a complete convolutional network on FashionMNIST — transforms, conv-relu-pool blocks with shape bookkeeping, data augmentation, and reading the mistakes. 9 min read
03Transfer Learning
Famous CNN architectures and how to fine-tune pretrained models on your own data.
- Famous CNN ArchitecturesThe ImageNet competition and the ideas it produced — AlexNet's GPUs, VGG's stacked 3×3 kernels, Inception's parallel branches, and ResNet's skip connections. 9 min read
- Transfer Learning in PracticeFreeze a pretrained ResNet, retrain its head on your own images, then fine-tune with a tiny learning rate — and see why this crushes training from scratch. 10 min read
04Recurrent Neural Networks
Sequence modeling with RNNs, LSTMs, and GRUs — plus bidirectionality and tuning for forecasting.
- Recurrent Neural NetworksGive a network memory — the recurrence behind RNNs, unrolling through time, vanishing gradients, and nn.RNN on a next-step prediction task. 9 min read
- LSTM & GRUGated memory cells that beat the vanishing gradient — how LSTM's forget/input/output gates work, GRU's streamlined variant, and both in PyTorch. 9 min read
- Sequence Models in PracticeBidirectional and stacked RNNs, windowing and scaling time series properly, multivariate inputs, and a full LSTM forecasting pipeline with early stopping. 10 min read
05Natural Language Processing
Word embeddings with Word2Vec and FastText, and an end-to-end sentiment analysis model.
- Word Embeddings: Word2Vec & FastTextWhy one-hot vectors fail, how Word2Vec learns meaning from context, and how FastText handles typos and words it has never seen. 9 min read
- Sentiment Analysis End to EndBuild a sentiment classifier three ways — a TF-IDF baseline you can run in the browser, an Embedding + LSTM model in PyTorch, and a modern transformer pipeline. 12 min read
06Generative Models
Autoencoders for compression and denoising, and GANs that learn to generate images.
- AutoencodersTrain a network to compress and reconstruct its own input — and use the bottleneck for dimensionality reduction, denoising, and anomaly detection. 10 min read
- Generative Adversarial NetworksTwo networks locked in a forgery contest — train a GAN from a 1-D toy demo in your browser up to a DCGAN that draws handwritten digits. 12 min read
