Things I've written down.
Ruby, Rails, performance, security and applied machine learning. Updated when something feels finished.
3 posts in MLOps
MLOPS
06 / 13 / 26
Your model didn't degrade, the world did
Model weights are frozen the day you deploy. Everything they were fitted to keeps moving, and nothing tells you unless you look.
2 min
MLOPS
05 / 08 / 26
The model was fine. The features were different.
A model that scores well offline and poorly in production usually isn't seeing the same inputs it was trained on.
2 min
MLOPS
01 / 21 / 26
I served ResNet-50 on a CPU and stopped paying for GPUs
Latency budgets, ONNX Runtime, and the arithmetic showing most inference workloads were never GPU problems in the first place.
4 min