Make your app a product with the TFX team
summary
Announces TFX 1.0 with a stable public API and three years of critical fixes, demonstrates pipelines running on Kubernetes via Kubeflow Pipelines, on Airflow, locally, and on Vertex AI, and shows Docker images plus Apache Beam (Spark) execution. Gives concrete examples: pushing non‑TensorFlow models with the Pusher, scikit‑learn examples, added NLP support with BLEU and ROUGE metrics, and published step‑by‑step tutorials at tensorflow.org/tfx and a TFX add‑ons repo.
AI-written summary · about
chapters (14)
- 0:00Intro
- 1:00Why use TFX
- 2:15Who uses TFX
- 3:10Can I run TFX on Google Cloud
- 4:20Does TFX run on Kubernetes
- 5:20How easy is it to use TFX
- 6:30How can I use TFX for NLP
- 8:30Can TFX deploy other models
- 9:45Can I move TFX pipeline to a different compute environment
- 11:05Is TFX different from Kubernetes
- 12:00What is TFX 10
- 13:20TFX for research
- 20:35TFX with Airflow
- 23:25Getting started
related talks
08mManage MLOps and deploy Machine Learning to production with the new and improved TFX
38mTensorFlow Extended (TFX): Machine Learning Pipelines and Model Understanding
37mEnd-to-End ML Pipelines with KubeFlow and TensorFlow Extended (TFX)
13mAn introduction to MLOps with TensorFlow Extended (TFX)
22mThe Containerization of Machine Learning: TensorFlow: Kubernetes, and Kubeflow
38mFrom Experimentation to Products: The Production Machine Learning Journey
21mIntroducing Kubeflow (w. Special Guests Tensorflow and Apache Spark)
36mIntroducing Kubeflow (w. Special Guests Tensorflow and Apache Spark)