Or: How I learned to stop worrying and love debugging neural networks

The Grand Plan (That Almost Failed Spectacularly)

I recently completed what I’d call my first “full-stack” MLOps project – and by “full-stack” I mean the entire pipeline from raw data to deployed …