Merge pull request #4 from nuance/random-fixes

Two small fixes

Author: lukaszkaiser

tensor2tensor/data_generators/text_encoder.py

@@ -109,12 +109,12 @@ class TokenTextEncoder(TextEncoder):
 
   def __init__(self, vocab_filename, reverse=False, num_reserved_ids=2):
     """Initialize from a file, one token per line."""
+    super(TokenTextEncoder, self).__init__(num_reserved_ids=num_reserved_ids)
+
     self._reverse = reverse
-    if vocab_filename is None:
+    if vocab_filename is not None:
       self._load_vocab_from_file(vocab_filename)
 
-    super(TokenTextEncoder, self).__init__(num_reserved_ids=num_reserved_ids)
-
   def encode(self, sentence):
     """Converts a space-separated string of tokens to a list of ids."""
     ret = [self._token_to_id[tok] for tok in sentence.strip().split()]
@@ -285,7 +285,7 @@ def build_to_target_size(cls,
     subtokenizer.build_from_token_counts(token_counts, store_filename,
                                          present_count, num_iterations)
 
-    if min_val == max_val or subtokenizer.vocab_size == target_size:
+    if min_val >= max_val or subtokenizer.vocab_size == target_size:
       return subtokenizer
     elif subtokenizer.vocab_size > target_size:
       other_subtokenizer = cls.build_to_target_size(