GRPO

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

Qwen-Cup-Multimodal-RL: Qwen Cup multimodal reasoning competition (20/337 → finals 9/20 →...

Fine-tuning GPT-OSS-20B for multilingual reasoning on Amazon SageMaker HyperPod

Scaling MoE reinforcement learning on Amazon EKS with EFA and DeepEP with 40% more throughput

Accelerate multimodal RL training with SkyRL on Amazon SageMaker HyperPod

DACA-GRPO: Denoising-Aware Credit Assignment for Reinforcement Learning in Diffusion Language Models

Async GRPO with LoRA across HF Jobs: a bucket, a proxy, and no NCCL

GRPO Beyond English: A Large-Scale Study of GRPO in Non-English and Multilingual Settings