PyTorch for Beginners, by a Beginner

PyTorch for Beginners, by a Beginner

Ever felt like machine learning frameworks were someone else's territory? This talk is for you.

We'll cover where PyTorch came from (a Lua library, an intern's project, and a bet on letting Python just be Python), why it now powers ChatGPT, Tesla Autopilot, and most of modern AI research. And then we'll look at the actual code: tensors, automatic differentiation, and a complete training loop in about 50 lines. We'll even peek under the hood at how a + b routes through multiple layers before anything actually runs.

By the end you'll understand the four concepts that make PyTorch tick, why it beat TensorFlow (hint: it's not the benchmarks), and how to go from zero to running a real neural network in an evening.

No prior ML experience required. Python basics are helpful. Laptop optional but encouraged — the examples are all pip install torch away.

#VegasTech