From ddce04ea9d6d41079b9a7bec048a054ade8667b8 Mon Sep 17 00:00:00 2001 From: Mike J Innes Date: Tue, 23 Aug 2016 14:59:35 +0100 Subject: [PATCH] repetition --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index c93d8ee1..28214ba0 100644 --- a/README.md +++ b/README.md @@ -12,7 +12,7 @@ Flux decouples the model (what you'd find in a paper) from the implementation (d It's also designed to be extremely flexible. Flux supports multiple backends (like MXNet and TensorFlow) and can transparently take advantage of features unique to the backend. Custom layer types can be implemented in pure Julia, and you can even mix and match different backends together. -Finally, Flux is hackable. Using Julia enables custom kernels, including GPU code, to be written in an interactive and high-level way. Flux itself is very hackable, and its whole implementation – including all built-in layers and utilities – is under 500 lines of pure Julia code. +Finally, Flux is hackable. Using Julia enables custom kernels, including GPU code, to be written in an interactive and high-level way. Flux's whole implementation – including all built-in layers and utilities – is under 500 lines of pure Julia code. ## How?