From d90e1118a6cee61d6de0cb2a2534d70462e034e4 Mon Sep 17 00:00:00 2001 From: vepricov Date: Tue, 26 Nov 2024 02:05:08 +0300 Subject: [PATCH] readme fix --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 539e369..cf0faad 100644 --- a/README.md +++ b/README.md @@ -57,11 +57,11 @@ In this project we implement different alternatives to it. ## 🗃 Algorithms to implement (from simplest to hardest) - [x] [Relaxed Bernoulli](http://proceedings.mlr.press/v119/yamada20a/yamada20a.pdf) - [x] [Correlated relaxed Bernoulli](https://openreview.net/pdf?id=oDFvtxzPOx) -- [ ] [Gumbel-softmax TOP-K](https://arxiv.org/pdf/1903.06059) +- [x] [Gumbel-softmax TOP-K](https://arxiv.org/pdf/1903.06059) - [x] [Straight-Through Bernoulli, distribution (don't mix with Relaxed distribution from pyro)](https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=62c76ca0b2790c34e85ba1cce09d47be317c7235) - [x] [Invertible Gaussian reparametrization](https://arxiv.org/abs/1912.09588) with KL implemented - [x] [Hard concrete](https://arxiv.org/pdf/1712.01312) -- [ ] [REINFORCE](http://www.cs.toronto.edu/~tingwuwang/REINFORCE.pdf) (not a distribution actually, think how to integrate it with other distributions) +- [x] [REINFORCE](http://www.cs.toronto.edu/~tingwuwang/REINFORCE.pdf) (not a distribution actually, think how to integrate it with other distributions) - [x] [Logit-normal distribution](https://en.wikipedia.org/wiki/Logit-normal_distribution) and [Laplace-form approximation of Dirichlet](https://stats.stackexchange.com/questions/535560/approximating-the-logit-normal-by-dirichlet) ## 📚 Recommended stack