sebulba

Python ★ 61 updated 2y ago

🪐 The Sebulba architecture to scale reinforcement learning on Cloud TPUs in JAX

No plain-English explanation yet — one is being written right now. Check back in a minute.