zhihanyang2022 / alpha-zero

Minimal AlphaZero in PyTorch, trained on Connect4 on a 6x6 board.

Hi, my name is Zhihan. I work on machine learning research.

Paper implementations:

• [2024/2] maf-pytorch: Demo PyTorch code for "Masked Autoregressive Flow for Density Estimation" (NIPS 2018)
• [2024/1] sylvester-flows-pytorch: Demo PyTorch code for "Sylvester Normalizing Flows for Variational Inference" (UAI 2018)
• [2023/8] normalizing-flows-pytorch: Demo PyTorch code for "Variational Inference with Normalizing Flows" (ICML 2015)
• [2021/10] off-policy-continuous-control: Official PyTorch code for "Recurrent Off-policy Baselines for Memory-based Continuous Control" (DeepRL Workshop, NeurIPS 2021)

Mini projects:

• [2023/11] dior: Differentiable reparameterization of matrices with orthogonal columns
• [2023/9] gradient-gmm: Finite mixture of full-rank Gaussians for density estimation & variational inference
• [2022/8] aevb-tutorial: Training VAE, conditional VAE, Gaussian Mixture VAE and Variational RNN on MNIST
• [2022/1] pytorch-ppo: A minimal version of the proximal policy gradient algorithm using the clipped objective
• [2021/8] alpha-zero: A minimal version of DeepMind's AlphaZero that beats humans on Connect 4