From 9afaa3a3236366d57cb1c3d771b25779ee76269b Mon Sep 17 00:00:00 2001
From: Eduardo Julian
Date: Fri, 31 Dec 2021 00:58:08 -0400
Subject: Fixes for the pure-Lux JVM compiler machinery.

---
 documentation/bookmark/machine_learning.md | 1 +
 1 file changed, 1 insertion(+)

(limited to 'documentation/bookmark/machine_learning.md')

diff --git a/documentation/bookmark/machine_learning.md b/documentation/bookmark/machine_learning.md
index 2c05864f7..6d11d7ca0 100644
--- a/documentation/bookmark/machine_learning.md
+++ b/documentation/bookmark/machine_learning.md
@@ -10,6 +10,7 @@
 
 # Reference
 
+1. [Why are ML Compilers so Hard?](https://petewarden.com/2021/12/24/why-are-ml-compilers-so-hard/)
 1. ["Multi-Level Intermediate Representation" Compiler Infrastructure](https://github.com/tensorflow/mlir)
 1. [Sampling can be faster than optimization](https://www.pnas.org/content/116/42/20881)
 1. [Layer rotation: a surprisingly powerful indicator of generalization in deep networks](https://arxiv.org/abs/1806.01603v2)
-- 
cgit v1.2.3