@ibab: The River API was tested in this blog post and outperformed Tinker on reinforcement learning runs with identical traini…
Summary
The River API was tested and outperformed Tinker on reinforcement learning runs with identical training code, featuring improvements in details like routing replay and being open-sourced with ablations.
View Cached Full Text
Cached at: 08/18/26, 12:27 AM
The River API was tested in this blog post and outperformed Tinker on reinforcement learning runs with identical training code. We spent a lot of effort to get details like routing replay right so you get the best possible results with the API.
Ashwinee Panda (@PandaAshwinee): We can now RL large MoEs with 0 train-infer mismatch! And doing so can improve performance (pictured task: teach Qwen3.6-35B-A3B to play Wordle). Everything is open-source and we did a bunch of ablations. 🧵
Similar Articles
@TheTuringPost: 10 open-source tools for the Agent RL stack ↓ OpenPipe ART verl-agent Agent Lightning Unsloth OpenRLHF SkyRL NVIDIA’s P…
A curated roundup of 10 open-source tools for training AI agents using reinforcement learning, covering frameworks like OpenPipe ART, verl-agent, Agent Lightning, and Unsloth, with details on their use cases and strengths.
@googledevs: A major update to Tunix for scaling Agentic RL is here The new asynchronous, decoupled rollout engine solves multi-turn…
Google announces a major update to Tunix, its post-training library, with an asynchronous decoupled rollout engine to scale agentic reinforcement learning on JAX/TPU, eliminating idle time and improving throughput.
@vivek_2332: Been exploring RL infra a lot lately, and it's something I want to go deep into. So I've been reading prime-rl codebase…
A technical blog post dissecting the main loop of the prime-rl reinforcement learning orchestrator, covering components like RolloutDispatcher, TrainSink, EvalSink, WeightWatcher, and PeriodicLogger.
Keep the Tokens Flowing: Lessons from 16 Open-Source RL Libraries
Hugging Face publishes a comprehensive analysis of 16 open-source reinforcement learning libraries, examining architectural patterns for asynchronous RL training and presenting design lessons for TRL's async trainer to address generation bottlenecks and weight synchronization challenges.
@hamishivi: Trained some terminal agents with friends! Introducing Tmax, open RL terminal agent models. Under default settings and …
Introducing Tmax, open reinforcement learning terminal agent models that outperform prior open work on terminal use. All data, weights, and rollouts are being released publicly.