http://programmer.ie/tags/training-loop/ Recent content in Training Loop on Programmer.ie Hugo en-us Sat, 08 Aug 2026 12:40:00 +0100 http://programmer.ie/books/pytorch-from-first-principles/01-chapter/ Sat, 08 Aug 2026 12:40:00 +0100 http://programmer.ie/books/pytorch-from-first-principles/01-chapter/ <p>It has never been easier to obtain working PyTorch code. Describe a model to an assistant, paste the result into a file, and a training loop will usually appear and usually run. For a large fraction of everyday work this is fine, and this book does not ask you to pretend otherwise.</p> <p>The difficulty arrives afterwards.</p> <p>The code runs, but the loss sits at exactly the same value for two hundred steps. Or one tensor is <code>[32, 128]</code> where the next operation wanted <code>[128, 32]</code>. Or the gradient of a weight is <code>None</code> and nothing says why. Or the model trained yesterday and today it does not, with no change you can point to. None of these are syntax errors. They do not produce a helpful traceback, and often they produce no traceback at all. They are questions about what the program is doing, and to answer them you need somewhere to look.</p>