Coin Sampling: Gradient-Based Bayesian Inference without Learning Rates

Open in new window