From 5d2d3758ade8fa5294b21b583ef34775f866f607 Mon Sep 17 00:00:00 2001
From: Rico Sennrich <rico.sennrich@gmx.ch>
Date: Thu, 3 Mar 2016 16:39:34 +0000
Subject: [PATCH] break condition for toy example

---
 bpe_toy.py | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/bpe_toy.py b/bpe_toy.py
index b930132..6e9bed3 100644
--- a/bpe_toy.py
+++ b/bpe_toy.py
@@ -13,7 +13,9 @@ Rico Sennrich, Barry Haddow and Alexandra Birch (2015). Neural Machine Translati
 """
 
 
-import re, collections
+import re
+import sys
+import collections
 
 def get_stats(vocab):
   pairs = collections.defaultdict(int)
@@ -38,5 +40,8 @@ num_merges = 15
 for i in range(num_merges):
   pairs = get_stats(vocab)
   best = max(pairs, key=pairs.get)
+  if pairs[best] < 2:
+     sys.stderr.write('no pair has frequency > 1. Stopping\n')
+     break
   vocab = merge_vocab(best, vocab)
   print(best)