The Production Machine Learning Blog

Google’s JAX: Flexible, High-Performance Machine Learning

Server-side batching: Scaling inference throughput in machine learning

How we served 1,000 models on GPUs for $0.47

Designing a machine learning platform for both data scientists and engineers

Netflix's Metaflow: Reproducible machine learning pipelines

How to serve batch predictions with TensorFlow Serving

How to build a pipeline to retrain and deploy models

How to deploy Transformer models for language tasks

Cortex 0.24: Announcing multi-cloud support

How we scale machine learning model deployment on Kubernetes

Why we built a serverless machine learning platform—instead of using AWS Lambda

Why we don’t deploy machine learning models with Flask

How to deploy machine learning models from a notebook to production

Why we use YAML—not notebooks—for machine learning

5 Lessons Learned Building an Open Source MLOps Platform

Machine learning doesn't have to be expensive

A/B testing machine learning models in production

How to deploy 1,000 models on one CPU with TensorFlow Serving

How to reduce the cost of machine learning inference

How to deploy PyTorch Lightning models to production

Improve NLP inference throughput 40x with ONNX and Hugging Face

Catching poachers with machine learning