Add doc2vec to the list of things to try

Former-commit-id: a1a99b8ea4

run.py

@@ -102,7 +102,7 @@ def go(lemmatization=False):
     else:
         text_vectorizer = TfidfVectorizer
 
-    # TODO: Try bag-of-words with word/char 1-gram, 2-gram, 3-grams, word2vec, 1d-cnn (both using pretrained word embeddings and not)
+    # TODO: Try bag-of-words with word/char 1-gram, 2-gram, 3-grams, word2vec, doc2vec, 1d-cnn (both using pretrained word embeddings and not)
 
     # TODO: Text cleanup (replace stack traces with "STACK_TRACE", replace file references with "FILE_REFERENCE", etc.)