Propriétaire

AI-Hypercomputer

14 dépôts indexés · Voir sur GitHub

• maxtext

  A simple, performant, and scalable Jax LLM!

  Python · 2420 étoiles

• JetStream

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

  Python · 457 étoiles

• maxdiffusion

  Python · 386 étoiles

• RecML

  Python · 226 étoiles

• xpk

  xpk (Accelerated Processing Kit, pronounced x-p-k,) is a software tool to help Cloud developers to orchestrate training jobs on accelerat...

  Python · 193 étoiles

• gpu-recipes

  Recipes for reproducing training and serving benchmarks for large machine learning models using GPUs on Google Cloud.

  Python · 140 étoiles

• tpu-recipes

  Shell · 126 étoiles

• accelerator-agents

  Python · 78 étoiles

• ml-goodput-measurement

  Python · 41 étoiles

• cloud-accelerator-diagnostics

  Python · 32 étoiles

• accelerator-microbenchmarks

  Python · 24 étoiles

• ray-tpu

  Python · 15 étoiles

• resiliency

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

  Python · 14 étoiles

• google-cloud-mldiagnostics

  Python · 14 étoiles