Adaptive Scheduling Pipeline for Multi-Instance Asynchronous Reinforcement Learning
cs.DC, cs.AI, cs.LG
Submitted: 2026-11-06
Updated: 2026-11-09
Code: https://github.com/verl-project/verl
Terminology
Sources
- Teaching Large Language Models to Reason with Reinforcement Learning
- Reinforcement Learning with Verifiable Rewards Implicitly Incentivizes Correct Reasoning in Base LLMs
- OpenRLHF: An Easy-to-use, Scalable and High-performance RLHF Framework
- StreamRL: Scalable, Heterogeneous, and Elastic RL for LLMs with Disaggregated Stream Generation
- AsyncFlow: An Asynchronous Streaming RL Framework for Efficient LLM Post-Training
- Asynchronous RLHF: Faster and More Efficient Off-Policy RL for Language Models
- AReaL: A Large-Scale Asynchronous Reinforcement Learning System for Language Reasoning
- Laminar: A Scalable Asynchronous RL Post-Training Framework
- MindSpeed RL: Distributed Dataflow for Scalable and Efficient RL Training on Ascend NPU Cluster
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- Qwen3 Technical Report
- Qwen2.5 Technical Report
- Revisiting Group Relative Policy Optimization: Insights into On-Policy and Off-Policy Training
- AlphaMaze: Enhancing Large Language Models' Spatial Intelligence via GRPO
- Proximal Policy Optimization Algorithms
- PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel
- Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism
- Ray: A Distributed Framework for Emerging AI Applications
- HuggingFace's Transformers: State-of-the-art Natural Language Processing
- IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
Related papers
- Programmable Probabilistic Computer with 1,000,000 p-bits
- BatchGen: An Architecture for Scalable and Efficient Batch Inference
- Seamless Reconfiguration for DAG-BFT (Technical Report)
- ELDR: Expert-Locality-Aware Decode Routing for PD-Disaggregated MoE Serving
- iScheduler: Reinforcement Learning-Driven Continual Optimization for Large-Scale Resource Investment Problems
- TIRx: A Unified Tile Primitive ML Compiler Across Heterogeneous Backends