SGLang promotes an RL training session on TorchTitan in Miles

1. 5:43 PM
   LMSYS Org@lmsysorgOfficial

   Tianyu Liu explains TorchTitan as a training backend in Miles

   AITianyu Liu shows how TorchTitan works as a training backend in Miles, connecting large-scale training and RL with SGLang rollouts.

   

1. 3:57 PM
   SGLang@sgl_projectOfficial

   SGLang and Miles host a reinforcement learning training session with NVIDIA

   AISGLang invites developers to a live session on running a reinforcement learning training loop with NVIDIA, using SGLang for rollout generation and Miles for policy training. The session covers live weight syncing between the two systems and engineering choices for faster, more stable RL training, with a path to larger models and more complex tasks. Registration is open through a Luma link.

The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.