From 9d89cbca131125aa73110c6987b34a54cece3dc9 Mon Sep 17 00:00:00 2001 From: Alexander Chalk Date: Fri, 28 Jun 2019 11:36:18 -0400 Subject: [PATCH] Change fasttext references in readme --- .../examples/cnn-text-classification/README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/contrib/clojure-package/examples/cnn-text-classification/README.md b/contrib/clojure-package/examples/cnn-text-classification/README.md index aa7640ff3190..cdae5ff0d308 100644 --- a/contrib/clojure-package/examples/cnn-text-classification/README.md +++ b/contrib/clojure-package/examples/cnn-text-classification/README.md @@ -54,11 +54,11 @@ Using fastText instead of glove is fairly straightforward, as the pretrained emb Download the 'Simple English' pretrained wiki word vectors (text) from the fastText [site](https://fasttext.cc/docs/en/pretrained-vectors.html) and place them in the -`data/fastText` directory. +`data/fasttext` directory. Then you can run training on a subset of examples through the repl using: ``` -(train-convnet {:devs [(context/default-context)] :embedding-size 300 :batch-size 100 :test-size 100 :num-epoch 10 :max-examples 1000 :pretrained-embedding :fastText}) +(train-convnet {:devs [(context/default-context)] :embedding-size 300 :batch-size 100 :test-size 100 :num-epoch 10 :max-examples 1000 :pretrained-embedding :fasttext}) ``` Expect a validation accuracy of `~0.67` with the above parameters.