VeRL-Omni v0.3.0: Async Training and Broader Multimodal Post-Training
·12 min read
VeRL-Omni v0.3.0 expands async RL, makes the V1 trainer the default, and adds reproducible recipes for audio-video generation and multimodal understanding.
3 posts
VeRL-Omni v0.3.0 expands async RL, makes the V1 trainer the default, and adds reproducible recipes for audio-video generation and multimodal understanding.
A release focused on higher-throughput diffusion rollout, reusable omni adapters, and broader recipe coverage.
How VeRL-Omni extends verl with vLLM-Omni for reinforcement learning post-training of diffusion and multimodal generative models, including efficient rollouts, reward inference, trainers, hardware support, and recipes.