Manage MLOps and deploy Machine Learning to production with the new and improved TFX
summary
TFX is a production MLOps stack that claims model code is only 5% of total code and focuses on pipelines, portability (Colab, notebook, Kubernetes, Vertex), and multiple executors and orchestrators (Spark, Flink, Dataflow; Airflow, Kubeflow). New SDK features include Conditional Control Flow and an Exit Handler for branching and success/failure actions. Community Firebase Publisher, Load Testing, XGBoost Evaluator, and Feast ExampleGen components add mobile publishing, pre-deploy load checks, XGBoost evaluation, and feature-store ingestion.
AI-written summary · about
chapters (7)
- 0:00Intro
- 2:08TFX portability
- 3:15Whats new
- 4:52Firebase Publisher
- 5:13Load Testing
- 5:38XGBoost Evaluation
- 6:07Feast Example Gen
related talks
13mAn introduction to MLOps with TensorFlow Extended (TFX)
14mDoes your app use ML? Make it a product with TFX
31mMake your app a product with the TFX team
38mTensorFlow Extended (TFX): Machine Learning Pipelines and Model Understanding
30mThe art of machine learning operations in TensorFlow ecosystem
38mFrom Experimentation to Products: The Production Machine Learning Journey
13mMLOps Pipelines to Deliver Tf.Lite Models to Distributed Compute Pools at the Edge
42mMLOps 2025: A Journey into the Past and the Future