AI-Hypercomputer/maxtext
A high-performance, scalable open-source LLM library written in JAX for training models on TPUs and GPUs.
Not currently ranked — collecting fresh signals.
star history
MaxText provides a library of transformer-based large language models including Gemma, Llama, DeepSeek, Qwen, and Mistral. It supports pre-training at scale (tens of thousands of chips) and post-training techniques like Supervised Fine-Tuning (SFT), Group Relative Policy Optimization (GRPO), and Group Sequence Policy Optimization (GSPO). The library achieves high Model FLOPs Utilization through JAX and XLA compiler optimizations while remaining simple and largely optimization-free.
Frequently asked
- What is AI-Hypercomputer/maxtext?
- A high-performance, scalable open-source LLM library written in JAX for training models on TPUs and GPUs.
- Is maxtext open source?
- Yes — AI-Hypercomputer/maxtext is open source, released under the Apache-2.0 license.
- What language is maxtext written in?
- AI-Hypercomputer/maxtext is primarily written in Python.
- How popular is maxtext?
- AI-Hypercomputer/maxtext has 2.4k stars on GitHub.
- Where can I find maxtext?
- AI-Hypercomputer/maxtext is on GitHub at https://github.com/AI-Hypercomputer/maxtext.