Orbit - Ultra-efficient RL Pipeline

Orbit — Ultra-efficient RL Pipeline

Orbit is an open-source RL framework for memory-efficient post-training of trillion-parameter LLMs. By minimizing the train-rollout gap in RL, Orbit is able to effortlessly train Kimi-K2.6 and DeepSeek V4-pro at a single-node setting . Five rollout architectures, one shared clock, clear win for Orbit's design Watch the bottleneck move. We start fully serial with a full-weight push (basesline 1 below), then test the tokens "is the heavy push the bottleneck?" . We gradually turn the baseline into Orbit's final design (+ overlap (async); + adapter-native push; + a non-blocking double-buffered swa

saved by

• Alex K

related reading

• You Only Need Minimal RLVR Training: Extrapolating LLMs via Rank-1 Trajectoriesarxiv.org
• LoRA Without Regret - Thinking Machines Labthinkingmachines.ai
• How We Build Trillion Parameter Reasoning RL with 10% GPUsmacaron.im
• Keep the Tokens Flowing: Lessons from 16 Open-Source RL Librarieshuggingface.co
• RL at 1T Scale: prime-rl Performance Deep Diveprimeintellect.ai
• Composer2.pdfcursor.com
• The 4-bitter Lesson | humans&humansand.ai
• RL Post-Training on Macs | Pluralis Researchpluralis.ai
• Frontier-scale RL with Kimi K3appliedcompute.com
• frontier model training methodologies | Alex Wa's Blogdjdumpling.github.io
• 2305.14314arxiv.org
• >10x More Efficient Pretraining — Magicmagic.dev