update toy example to BPE use same representation again as learn_bpe.py

learn_bpe.py was changed in commit a749a7 to make end-of-word representation more consistent.

bpe_toy.py

@@ -35,12 +35,15 @@ def merge_vocab(pair, v_in):
     v_out[w_out] = v_in[word]
   return v_out
 
-vocab = {'l o w </w>' : 5, 'l o w e r </w>' : 2,
-         'n e w e s t </w>' : 6, 'w i d e s t </w>' : 3}
+vocab = {'l o w</w>' : 5, 'l o w e r</w>' : 2,
+         'n e w e s t</w>' : 6, 'w i d e s t</w>' : 3}
 num_merges = 15
 for i in range(num_merges):
   pairs = get_stats(vocab)
-  best = max(pairs, key=pairs.get)
+  try:
+    best = max(pairs, key=pairs.get)
+  except ValueError:
+    break
   if pairs[best] < 2:
      sys.stderr.write('no pair has frequency > 1. Stopping\n')
      break