From 22ac0d4096575aecbd35f85188a2ca7d0af4fad3 Mon Sep 17 00:00:00 2001
From: fmikaelian <39884124+fmikaelian@users.noreply.github.com>
Date: Thu, 14 Feb 2019 11:13:35 +0100
Subject: [PATCH] Implement download.py script (SQuAD fetch) #9

---
 reading-comprehension/pipeline/download.py | 11 +++++++++++
 requirements.txt                           |  1 +
 2 files changed, 12 insertions(+)

diff --git a/reading-comprehension/pipeline/download.py b/reading-comprehension/pipeline/download.py
index e69de29b..208f8274 100644
--- a/reading-comprehension/pipeline/download.py
+++ b/reading-comprehension/pipeline/download.py
@@ -0,0 +1,11 @@
+import wget
+
+squad_urls = [
+    'https://rajpurkar.github.io/SQuAD-explorer/dataset/train-v1.1.json',
+    'https://rajpurkar.github.io/SQuAD-explorer/dataset/dev-v1.1.json',
+]
+
+for squad_url in squad_urls:
+    wget.download(url=squad_url, out='data')
+
+wget.download(url='https://github.com/allenai/bi-att-flow/blob/master/squad/evaluate-v1.1.py')
diff --git a/requirements.txt b/requirements.txt
index abb5f833..8823b187 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -4,4 +4,5 @@ matplotlib
 pytorch-pretrained-bert
 beautifulsoup4
 tqdm
+wget
 prettytable