@SergioPaniego: continuous batching just landed in TRL for GRPO at 64 generations it runs faster and uses less VRAM than plain generate…
Summary
Continuous batching has been added to TRL for GRPO, improving speed and VRAM usage without needing vLLM. The tweet explains how it works and when to use it.
View Cached Full Text
Cached at: 06/20/26, 02:36 PM
continuous batching just landed in TRL for GRPO
at 64 generations it runs faster and uses less VRAM than plain generate, no vLLM needed
how it works and when to reach for it, below
Similar Articles
@pallavishekhar_: Continuous Batching in LLMs Read here: https://outcomeschool.com/blog/continuous-batching-in-llms…
A blog post explaining continuous batching, a technique for improving LLM serving throughput by dynamically adding new requests to a batch as old ones finish, keeping the GPU busy and reducing idle time.
@QGallouedec: TRL v1.4 is out! two things I'm excited about: → chunked NLL loss for SFT. Way less VRAM, same loss, often faster. Qwen…
TRL v1.4 is released, featuring chunked NLL loss for SFT to reduce VRAM usage and first-class integration with OpenReward for GRPO.
LithoGRPO: Fast Inverse Lithography via GRPO Reinforced Flow Matching
LithoGRPO introduces a novel framework that combines flow matching with GRPO-based reinforcement learning for fast and high-quality inverse lithography mask optimization, achieving state-of-the-art performance while maintaining efficient generation.
Accelerating Disaggregated RL for Visual Generative LLMs with Diffusion-Based Parallelism and Trainer-Assisted Generation
This paper introduces DigenRL, a disaggregated RL framework for diffusion-based generative LLMs that uses generation-axis pipeline parallelism and trainer-assisted generation to improve throughput by 1.56-2.10x over existing systems.
@SergioPaniego: you can now train @liquidai's LFM2-VL in TRL GRPO and RLOO included, with an example script
You can now train Liquid AI's LFM2-VL model using TRL's GRPO and RLOO methods, with an example script provided.