Advanced features on TensorFlow Serving @TensorFlow

Uploaded August 2022 | Updated September 2026, 1 week ago

Wei Wei, Developer Advocate at Google, shares several advanced TensorFlow Serving features (experimental). Learn how TF Serving can tend JAX models, serve non-TensorFlow models with new servables, and remotely predict RPC with distributed serving.

Resources:
TF Serving model server parameters → https://goo.gle/3tQ99Qq
Intro to JAX: Accelerating Machine Learning research → https://goo.gle/3xOKBsq
Convert JAX models to SavedModel with jax2tf → https://goo.gle/3NiCCtk
Convert JAX models to TFLite → https://goo.gle/3Ngqo4n
Creating a new kind of servable → https://goo.gle/3A1xuqq
XGBoost Serving using TF Serving (community project) → https://goo.gle/3A08iAO
Hashmap servable reference on Github → https://goo.gle/3HUy4bp
TensorFlow servable on Github → https://goo.gle/3xOZScG
Remote predict TensorFlow operator → https://goo.gle/3Nim5pl

Deploying Production ML Models with TensorFlow Serving playlist → https://goo.gle/tf-serving
Subscribe to TensorFlow → https://goo.gle/TensorFlow

#TensorFlow #MachineLearning #ML

Shaping the future of AI to be more inclusive and approachable

Making OSS Machine Learning communities more diverse | Keynote

Building an inclusive community in the field of Machine Learning

Top 5 takeaways from Women in Machine Learning Symposium 2022

Kind and helpful Machine Learning through UX research

Grocery store recommendation systems with HarperDB and Web ML - Made with TensorFlow.js

Reinforcement learning with Snake-RL - Made with TensorFlow.js

Simplified Machine Learning on mobile, browsers, and edge devices with on-device ML

Super resolution with UpscalerJS - Made with TensorFlow.js