google/paxml

paxml

✅ Recently maintained

Pax is a Jax-based machine learning framework for training large scale models. Pax allows for advanced and fully configurable experimentation and parallelization, and has demonstrated industry leading model flop utilization rates.

560Stars
73Forks
07-day stars

Python · Apache-2.0 · Last push 2026-10-26

Official repository & installation ↗
c4jaxlarge-language-modelsllmmodel-flopsparallelismgpt

Star history

2025-09-292026-11-02
View historical data
DateStars
2026-10-04560
2026-10-05560
2026-10-06560
2026-10-07560
2026-10-08560
2026-10-09560
2026-10-10560
2026-10-11560
2026-10-12560
2026-10-13560
2026-10-14560
2026-10-15560
2026-10-16560
2026-10-17560
2026-10-18560
2026-10-19560
2026-10-20560
2026-10-21560
2026-10-22560
2026-10-23560
2026-10-24560
2026-10-25560
2026-10-26560
2026-10-27560
2026-10-28560
2026-10-29560
2026-10-30560
2026-10-31560
2026-11-01560
2026-11-02560

Get the project

git clone https://github.com/google/paxml.git

Related projects & alternative candidates

Related domains do not imply feature parity. Review official documentation and licenses.

LLMs-from-scratch

Step-by-step guide to building a GPT-like LLM from scratch in PyTorch.

★ 107,671LightRAG

[EMNLP2025] LightRAG: Simple and Fast Retrieval-Augmented Generation

★ 40,474dify

Build agentic workflows and RAG pipelines in one workspace, from prototype to self-hosted production.

★ 160,192