AI-Hypercomputer
Reference implementations, benchmarks, recipes, and all things Google Cloud AI Hypercomputer
Built from AI-Hypercomputer's public GitHub profile. Not on Take My Repo yet, so there are no posts or updates here.
JetStream is a throughput and memory optimized engine for LLM inference on XLA devices, starting with TPUs (and GPUs in future -- PRs welcome).
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 GPUs on GKE.
Recipes for reproducing training and serving benchmarks for large machine learning models using GPUs on Google Cloud.
Hardware-grounded model architecture design principles, MoE co-design patterns, low-precision numerics, and parallelism guides for training and serving on Google Cloud TPUs.
PyTorch/XLA integration with JetStream (https://github.com/google/JetStream) for LLM inference"
TPUMS (TPU Microbenchmark Suite) is a modular, high-fidelity benchmarking and profiling framework designed to evaluate the compute, memory, and interconnect performance of Cloud TPUs using JAX.
This project provides resilient training logic to improve the training efficiency of large-scale distributed workloads on Google Cloud.