AITOPAI热报
How-toOfficialSelected

Amazon SageMaker 推出 training session API，用 Python 脚本自定义 LLM 训练循环

Write your own training loop — without managing GPUs — with Amazon SageMaker training session APIs

官方一手AWS Machine Learning Blog@Nicolas Jourdan
Blog/Media
Read the original
Why we picked it

AWS 出了个新 API，写个 Python 脚本就能跑自己的 SFT 或 RL 训练循环，GPU 不用自己管，文中还有 GRPO 跑 GSM8K 的完整例子。

AWS 为 Amazon SageMaker AI 推出 training session APIs，开发者可以用 Python 脚本自己写 LLM 训练循环，支持 SFT、RL 和自定义目标函数，GPU 由服务托管。文章介绍了编程模型和三个 SDK 客户端，并演示了在 GSM8K 上跑 GRPO 强化学习循环，以及在竞赛编程任务上的 RL 工作负载。

Image · AWS Machine Learning Blog
Original · AWS Machine Learning Blog

Write your own training loop — without managing GPUs — with Amazon SageMaker training session APIs

Amazon SageMaker AI training session APIs let you write your own LLM training loop (SFT, RL, or custom objectives) in a Python script while the service runs the GPUs. This post introduces the programming model and three SDK clients, then walks through a GRPO loop on GSM8K and a reinforcement learning workload on competitive programming.

#Amazon SageMaker#GRPO#GSM8K#微调#强化学习

Keep reading

Recent stories with the same tag or section
ProductsBuildersOfficialSelected00:27
Amazon SageMaker HyperPod Cluster Sharing

This solution addresses a critical challenge in resource management for AI teams, preventing resource hoarding and ensuring fair allocation. For organizations using AWS, this provides a concrete blueprint for multi-team GPU cluster management with built-in cost tracking and security controls.

#Amazon SageMaker#HyperPod#AWS#GPU clusters
官方一手AWS Machine Learning Blog@Giuseppe Angelo PorcelliSource