NIXL (prime-rl) media is blocked
Allow external media to connect to the provider and play this content.
Description
A high-performance framework for large-scale, asynchronous reinforcement learning training designed to scale to 1000+ GPUs for 1T+ parameter MoE models. It features optimized kernels for expert and context parallelism, native integration with agentic environments like SWE, and support for distributed deployment via Slurm and Kubernetes.
Descriptions, tags, and model credits may be AI-generated or inferred from public sources and can be incomplete or wrong. Learn more.