Deep Learning with PyTorch Tutorial
Learn deep learning with PyTorch through tensors, training, CNNs, sequence models, transformers, evaluation, deployment, and responsible operation.
Start learning Deep Learning with PyTorch →
Course Contents
PyTorch Foundations
- Deep Learning and the PyTorch Workflow
- Tensors, Shapes, Dtypes, and Broadcasting
- Automatic Differentiation and Computation Graphs
- Modules, Parameters, Devices, and Reproducibility
Data Pipelines
- Datasets, DataLoaders, Batching, and Shuffling
- Transforms, Normalization, and Data Augmentation
- Train, Validation, and Test Splits
- Imbalanced Data, Sampling, and Label Quality
Training Neural Networks
- Build a Multilayer Perceptron
- Loss Functions and Task Alignment
- Optimizers, Learning Rates, and Schedulers
- Training Loops, Checkpoints, and Early Stopping
Computer Vision
- Convolutional Neural Networks
- Image Classification Pipeline
- Transfer Learning and Fine-Tuning
- Detection, Segmentation, and Vision Evaluation
Sequences and Transformers
- Embeddings and Sequence Representation
- Recurrent Networks, LSTMs, and GRUs
- Attention and Transformer Architecture
- Fine-Tuning a Pretrained Transformer
Evaluation and Improvement
- Metrics, Thresholds, and Calibration
- Error Analysis and Dataset Slices
- Regularization, Dropout, Normalization, and Generalization
- Hyperparameter Search and Experiment Tracking
Efficient and Responsible DL
- Mixed Precision, Gradient Accumulation, and Distributed Training
- Interpretability and Feature Attribution
- Bias, Privacy, Robustness, and Adversarial Risk
- Model Cards, Dataset Documentation, and Governance
