MLflow Projects: Make Training Runnable Anywhere

Day 37 of 100 Days of MLOps. 'Just run my script' hides a dozen assumptions — which entry point, which parameters, which environment. An MLflow Project packages your code into a self-describing, runnable unit: an MLproject file declares the entry points, parameters and environment, so anyone runs your training with one command, mlflow run . — reproducibly, even straight from a Git URL. Runs 100% locally on any OS.