[Roadmap] 2025 Q4 Milestones #542
Description
AReaL 2025 Q4 Milestone Tracker
Introduction
This document tracks major planned enhancements for AReaL through January 31, 2026. Our development roadmap is organized into two categories to help contributors identify where they can make the most impact:
On-going sections contain features currently under active development by the core AReaL team. These represent our immediate priorities.
Planned but not in progress sections list features with concrete implementation plans that we currently lack bandwidth to pursue. We actively welcome community contributions for these items! If you're interested in contributing to any planned feature, please reach out to discuss implementation details.
Backends
On-going
- Single-controller mode [Feature] Add single-controller mode #260
- Detailed profiling for optimal performance across different scales fix: enhance rollout statistics tracking with enqueued state #522 chore: refactor boba GRPO for tracing #527 feat: introduces session-centric tracing APIs #539 etc.
- RL training with cross-node vLLM pipeline/context parallelism
Planned but not in progress
- Multi-LLM training (different agents with different parameters)
- Data transfer optimization in single-controller mode
- Auto-scaling inference engines in single-controller mode
- Elastic weight update setup and acceleration
- Low-precision RL training
Usability
Done
- Add CI pipeline to build Docker images upon release ci: build docker images with GCP #564 ci: automatically tag the dev image upon new releases #574
On-going
N/A
Planned but not in progress
- Wrap training scripts into trainers
- Fully respect allocation mode in trainers/training scripts
- Support distributed training and debugging in Jupyter notebooks
- Refactor FSDP/Megatron engine/controller APIs to finer granularity
- Example of using a generative or critic-like reward model
- Rename
RemoteSGLang/vLLMEngineasSGLang/vLLMEngine - Support directly constructing inference/training engines without config objects
- Flatten the import structure of areal modules
Documentation
On-going
N/A
Planned but not in progress
- Tutorial on how to write efficient async rollout workflows
- Benchmarking and profiling guide
- Use case guides: offline inference, offline evaluation, multi-agent training
- AReaL performance tuning guide
- Device allocation strategies for training and inference
- Parallelism strategy configuration for training and inference