Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn the fundamentals of AI research through a comprehensive course that covers essential mathematics, PyTorch programming, neural networks, and transformer architectures. Begin with foundational mathematical concepts including functions, derivatives, vectors, gradients, matrices, and probability theory that underpin modern artificial intelligence. Master PyTorch fundamentals by working with tensors, learning reshaping and viewing operations, dimension manipulation, indexing and slicing techniques, and creating special tensor types. Build neural networks from scratch, starting with single neurons and progressing through activation functions like sigmoid, ReLU, and tanh, before tackling multi-layer networks and backpropagation algorithms. Dive deep into transformer architectures that power large language models, exploring attention mechanisms with query, key, and value components, self-attention and causal self-attention patterns, rotary positional embeddings (RoPE), multi-head attention systems, and complete transformer blocks with feed-forward networks and normalization layers. Conclude with tokenization techniques used in GPT architectures, providing a complete foundation for understanding and implementing modern AI systems.
