Computes the Word Error Rate, defined as the edit distance between the two provided sentences after tokenizing to words. Arguments: s1 (string): space-separated sentence s2 (string): space-separated sentence
(self, s1, s2)
source not stored for this graph (policy: none)