Image Classification
Distributed Processing on a Pod
Large Language Models
-
Train on a single-device TPU using Pax
A guide to train a SPMD model with Pax on a single-device Cloud TPU.
-
JetStream MaxText inference on v5e
A guide to set up and use JetStream with MaxText for inference.
-
JetStream PyTorch inference on v5e
A guide to set up and use JetStream with PyTorch for inference.
-
Serve an LLM using TPUs on GKE with vLLM
A guide to using vLLM to serve large language models (LLMs) using Tensor Processing Units (TPUs) on Google Kubernetes Engine (GKE).