Pluralis Research seeks a senior RL post-training engineer to build the end-to-end post-training stack for large models that run on consumer devices over the internet. You will design rollout ingestion, reward computation, and weight synchronization across a geo-distributed setup.
You will invent algorithms for asynchronous, high-latency settings, deliver first decentralized post-trained models, and work with Python and PyTorch in a remote-first, globally distributed team.
J-*-Ljbffr
📌 Research Engineer Decentralized Rl Post Training Melbourne (Australia)
🏢 Pluralis Research
📍 Australia
Reply to this offer
Impress this employer describing Your skills and abilities, fill out the form below and leave Your personal touch in the presentation letter.