Advanced hands-on training in Deep Learning and GPU-accelerated computing with NVIDIA CUDA, covering the design, training and optimization of neural networks using Python, PyTorch, TensorFlow and NVIDIA GPU technologies.
Key topics include neural networks, forward and backpropagation, gradient-based optimization, loss functions, logistic regression, CNNs (Convolutional Neural Networks), supervised and unsupervised learning, sparse coding, RNNs, LSTM and GRU architectures, sequence modeling, NLP (Natural Language Processing), supervised and unsupervised embeddings, and multirelational data representations.
The repository follows a course that also covers GPU-accelerated model training, CUDA, cuDNN, computational efficiency, hyperparameter optimization and model scalability, with practical exercises using NVIDIA training materials and NVIDIA computing infrastructure.