[RA-L 26] Learning on the Fly: Rapid Policy Adaptation via Differentiable Simulation - View it on GitHub