MaxText v0.2.4 adds long-context training for Cloud TPU
Google announced that MaxText v0.2.4 supports long-context training on Cloud TPU through four context-parallel strategies: All-gather, A2A (DeepSpeed Ulysses), P2P (Ring Attention), and A2A + P2P (USP). The framework can extend context to 14 million tokens across TPU 5p and TPU 7, and Google reports that Llama 3.1 8B reached 70% Model FLOPs Utilization on TPU 5p for sequences up to 4 million tokens. The release is available on PyPI, Docker images, and GitHub main.
The release gives AI developers an available open-source JAX framework for training models with million-token sequences on Cloud TPU, addressing the memory and communication bottlenecks that emerge during long-context training.
Announcing Long-Context Training on TPUs with MaxText Context Parallelism - Google Developers Blog
Announcing Long-Context Training on TPUs with MaxText Context Parallelism
NOV. 9, 2026
Michael Tran Software Engineering Intern Google Cloud
Nuojin Cheng Software Engineer Google Cloud
Deepak Patil Group Product Manager