Keras implementation of NADE models
-
Updated
Jan 4, 2018 - Python
Keras implementation of NADE models
PyTorch implementations of FVSBN, NADE, and MADE — foundational autoregressive density estimators for binarised MNIST — with a vectorised NADE training loop, a 15-run hyper-parameter sweep, and publication-quality architecture figures.
基于单层神经网络的自回归生成模型,包含多种数据类型的概率建模、模型训练、样本生成与结果可视化。
To associate your repository with the nade topic, visit your repo's landing page and select "manage topics."