Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Boltzmann Machines in TensorFlow with examples
| Date | Stars |
|---|---|
| 2026-07-24 | 851 |
| 2026-07-25 | 851 |
| 2026-07-28 | 851 |
| 2026-07-30 | 851 |
| 2026-08-06 | 851 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
<p float="left"> <img src="img/dbm_mnist/rbm1.png" width="129" /> <img src="img/dbm_mnist/samples.png" width="132" /> <img src="img/dbm_cifar2/rbm_small_0.png" width="129" /> <img src="img/dbm_cifar2/grbm.png" width="129" /> <img src="img/dbm_cifar2/mrbm.png" width="129" /> <img src="img/dbm_cifar/samples.png" width="129" /> </p> # Boltzmann Machines This repository implements generic and flexible RBM and DBM models with lots of features and reproduces some experiments from *"Deep boltzmann machines"* [**[1]**](#1), *"Learning with hierarchical-deep models"* [**[2]**](#2), *"Learning multiple layers of features from tiny images"* [**[3]**](#3), and some others. ## Table of contents * [What's Implemented](#whats-implemented) * [Restricted Boltzmann Machines (RBM)](#restricted-boltzmann-machines-rbm) * [Deep Boltzmann Machines (DBM)](#deep-boltzmann-machines-dbm) * [Common features](#common-features) * [Examples](#examples) * [#1 RBM MNIST: <a href="examples/rbm_mnist.py">script</a>, <a href="https://nbviewer.jupyter.org/github/monsta-hd/boltzmann-machines/blob/master/notebooks/rbm_mnist.ipynb">notebook</a>](#1-rbm-mnist-script-notebook) * [#2 DBM MNIST: <a href="examples/dbm_mnist.py">script</a>, <a href="https://nbviewer.jupyter.org/github/monsta-hd/boltzmann-machines/blob/master/notebooks/dbm_mnist.ipynb">notebook</a>](#2-dbm-mnist-script-notebook) * [#3 DBM CIFAR-10 "Naïve": <a href="examples/dbm_cifar_naive.py">script</a>, <a href="https://nbviewer.jupyter.org/github/monsta-hd/boltzmann-machines/blob/master/notebooks/dbm_cifar_naive.ipynb">notebook</a>](#3-dbm-cifar-10-naïve-script-notebook) * [#4 DBM CIFAR-10: <a href="examples/dbm_cifar.py">script</a>, <a href="https://nbviewer.jupyter.org/github/monsta-hd/boltzmann-machines/blob/master/notebooks/dbm_cifar.ipynb">notebook</a>](#4-dbm-cifar-10-script-notebook) * [How to use examples](#how-to-use-examples) * [Memory requirements](#memory-requirements) * [Download models and stuff](#download-models-and-stuff) * [TeX notes](#tex-notes) * [How to install](#how-to-install) * [Common installation issues](#common-installation-issues) * [Possible future work](#possible-future-work) * [Contributing](#contributing) * [References](#references) ## What's Implemented ### Restricted Boltzmann Machines (RBM) * [[computational graph]](img/tensorboard_rbm/tf_graph.png) * k-step Contrastive Divergence; * whether to sample or use probabilities for visible and hidden units; * *variable* learning rate, momentum and number of Gibbs steps per weight update; * *regularization*: L2 weight decay, dropout, sparsity targets; * *different types of stochastic layers and RBMs*: implement new type of stochastic units or create new RBM from existing types of units; * *predefined stochastic layers*: Bernoulli, Multinomial, Gaussian; * *predefined RBMs*: Bernoulli-Bernoulli, Bernoulli-Multinomial, Gaussian-Bernoulli; * initialize weights randomly, from `np.ndarray`-s or from another RBM; * can be modified for greedy layer-wise pretraining of DBM (see [notes](#tex-notes) or [**[1]**](#1) for details); * *visualizations in Tensorboard* (hover images for details) and more: <p align="center"> <img src="img/tensorboard_rbm/msre.png" height="156" title="Mean squared reconstruction error" /> <img src="img/tensorboard_rbm/pll.png" height="156" title="Pseudo log-likelihood" /> <img src="img/tensorboard_rbm/feg.png" height="156" title="Free energy gap [4]" /> </p> <p align="center"> <img src="img/tensorboard_rbm/l2_loss.png" height="156" title="L2 loss (weight decay cost times 0.5||W||^2)" /> <img src="img/tensorboard_rbm/dist_W.png" width="256" title="Distribution of weights and biases" /> <img src="img/tensorboard_rbm/dist_hb.png" width="256" title="Distribution of weights and biases" /> </p> <p align="center"> <img src="img/tensorboard_rbm/dist_dW.png" width="258" title="Distribution of weights and biases updates" /> <img src="img/tensorboard_rbm/di
Excerpt of 28,859 characters
Read on GitHub
769
2
1
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:57388b2878a90912, topic:deep-learning, topic:tensorflow, readme:pretraining