From c5761c7e6e223ad785d214a874526ad9b2d55027 Mon Sep 17 00:00:00 2001 From: Vlad Trubachov Date: Fri, 12 Feb 2021 19:44:39 -0800 Subject: [PATCH] Fixed typo in hw_wordentail.ipynb --- hw_wordentail.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw_wordentail.ipynb b/hw_wordentail.ipynb index d905b45..383ce47 100644 --- a/hw_wordentail.ipynb +++ b/hw_wordentail.ipynb @@ -109,7 +109,7 @@ "\n", "I've processed the data into a train/dev split that is designed to put some pressure on our models to actually learn these semantic relations, as opposed to exploiting regularities in the sample. \n", "\n", - "The defining feature of the dataset is that the `train` and `dev` __vocabularies__ are disjoint. That is, if a word `w` appears in a training pair, it does not occur in any text pair. It follows from this that there are also no word-pairs shared between train and dev, as you would expect. This should require your models to learn abstract relationships, as opposed to memorizing incidental properties of individual words in the dataset." + "The defining feature of the dataset is that the `train` and `dev` __vocabularies__ are disjoint. That is, if a word `w` appears in a training pair, it does not occur in any test pair. It follows from this that there are also no word-pairs shared between train and dev, as you would expect. This should require your models to learn abstract relationships, as opposed to memorizing incidental properties of individual words in the dataset." ] }, {