Repositories list

• maxtext

  Public

  AI-Hypercomputer/maxtext’s past year of commit activity
  A simple, performant and scalable Jax LLM!

  gptmistralfine-tuning+ 11

  gptmistralfine-tuningjaxsftlarge-language-modelsllmllama2mixtraldeepseek+ 4

  Python

  •

  Apache License 2.0

  •506506 forks•2.3k2.3k stars•7777 issues•165165 pull requests•Updated Apr 23, 2026Apr 23, 2026

• maxdiffusion

  Public

  AI-Hypercomputer/maxdiffusion’s past year of commit activity
  Python

  •

  Apache License 2.0

  •6969 forks•341341 stars•66 issues•66 pull requests•Updated Apr 23, 2026Apr 23, 2026

• accelerator-agents

  Public

  AI-Hypercomputer/accelerator-agents’s past year of commit activity
  Python

  •

  Apache License 2.0

  •22 forks•1313 stars•11 issue•1111 pull requests•Updated Apr 22, 2026Apr 22, 2026

• tpu-recipes

  Public

  AI-Hypercomputer/tpu-recipes’s past year of commit activity
  Shell

  •

  Apache License 2.0

  •5757 forks•9292 stars•66 issues•1717 pull requests•Updated Apr 22, 2026Apr 22, 2026

• accelerator-microbenchmarks

  Public

  AI-Hypercomputer/accelerator-microbenchmarks’s past year of commit activity
  Python

  •

  Apache License 2.0

  •3434 forks•1616 stars•33 issues•77 pull requests•Updated Apr 22, 2026Apr 22, 2026

• pathways-utils

  Public

  AI-Hypercomputer/pathways-utils’s past year of commit activity
  Package of Pathways-on-Cloud utilities

  Python

  •

  Apache License 2.0

  •1212 forks•2727 stars•00 issues•66 pull requests•Updated Apr 22, 2026Apr 22, 2026

• gpu-recipes

  Public

  AI-Hypercomputer/gpu-recipes’s past year of commit activity
  Recipes for reproducing training and serving benchmarks for large machine learning models using GPUs on Google Cloud.

  gpubenchmarksgoogle-cloud-platform+ 2

  gpubenchmarksgoogle-cloud-platformservingdistributed-training

  Python

  •

  Apache License 2.0

  •7474 forks•131131 stars•88 issues•1919 pull requests•Updated Apr 22, 2026Apr 22, 2026

• cloud-accelerator-diagnostics

  Public

  AI-Hypercomputer/cloud-accelerator-diagnostics’s past year of commit activity
  Python

  •

  Apache License 2.0

  •1414 forks•2929 stars•66 issues•00 pull requests•Updated Apr 21, 2026Apr 21, 2026

• xpk

  Public

  AI-Hypercomputer/xpk’s past year of commit activity
  xpk (Accelerated Processing Kit, pronounced x-p-k,) is a software tool to help Cloud developers to orchestrate training jobs on accelerators such as TPUs and GP…

  gkegcloudtpu

  gkegcloudtpu

  Python

  •

  Apache License 2.0

  •8787 forks•180180 stars•2525 issues•22 pull requests•Updated Apr 21, 2026Apr 21, 2026

• RecML

  Public

  AI-Hypercomputer/RecML’s past year of commit activity
  Python

  •

  Apache License 2.0

  •2222 forks•223223 stars•33 issues•8484 pull requests•Updated Apr 14, 2026Apr 14, 2026

• cloud-tpu-monitoring-debugging

  Public
  HCL

  •

  Apache License 2.0

  •55 forks•1515 stars•00 issues•11 pull request•Updated Apr 8, 2026Apr 8, 2026

• google-cloud-mldiagnostics

  Public
  Python

  •

  Apache License 2.0

  •22 forks•1010 stars•00 issues•00 pull requests•Updated Apr 7, 2026Apr 7, 2026

• inference-benchmark

  Public archive

  AI-Hypercomputer/inference-benchmark’s past year of commit activity
  Python

  •

  Apache License 2.0

  •1515 forks•2020 stars•77 issues•44 pull requests•Updated Mar 11, 2026Mar 11, 2026

• torchprime

  Public archive
  torchprime is a reference model implementation for PyTorch on TPU.

  Python

  •

  Other

  •1414 forks•4747 stars•8888 issues•1717 pull requests•Updated Mar 3, 2026Mar 3, 2026

• kithara

  Public archive
  Python

  •

  Apache License 2.0

  •99 forks•1717 stars•00 issues•22 pull requests•Updated Feb 18, 2026Feb 18, 2026

• ml-goodput-measurement

  Public
  Python

  •

  Apache License 2.0

  •66 forks•3535 stars•11 issue•00 pull requests•Updated Feb 4, 2026Feb 4, 2026

• cloud-diagnostics-xprof

  Public
  Python

  •

  Apache License 2.0

  •88 forks•1313 stars•00 issues•11 pull request•Updated Jan 23, 2026Jan 23, 2026

• JetStream

  Public
  JetStream is a throughput and memory optimized engine for LLM inference on XLA devices, starting with TPUs (and GPUs in future -- PRs welcome).

  gpuinferencepytorch+ 13

  gpuinferencepytorchtransformerllamagptgemmamodel-servingtpujax+ 6

  Python

  •

  Apache License 2.0

  •6363 forks•428428 stars•1414 issues•1313 pull requests•Updated Jan 5, 2026Jan 5, 2026

• jetstream-pytorch

  Public archive
  PyTorch/XLA integration with JetStream (https://github.com/google/JetStream) for LLM inference"

  inferencepytorchbatching+ 8

  inferencepytorchbatchingattentionllamagemmamodel-servingtpullmllm-inference+ 1

  Python

  •

  Apache License 2.0

  •2121 forks•8282 stars•1010 issues•33 pull requests•Updated Dec 18, 2025Dec 18, 2025

• resiliency

  Public
  This project provides resilient training logic to improve the training efficiency of large-scale distributed workloads on Google Cloud.

  Python

  •

  Apache License 2.0

  •33 forks•1414 stars•11 issue•11 pull request•Updated Sep 29, 2025Sep 29, 2025

• ray-tpu

  Public
  Python

  •

  Apache License 2.0

  •88 forks•1616 stars•33 issues•11 pull request•Updated May 11, 2025May 11, 2025

• aotc

  Public archive
  Python

  •

  Other

  •55 forks•11 star•11 issue•11 pull request•Updated May 8, 2025May 8, 2025