add back comment

This commit is contained in:
Kian Kyars 2025-11-23 08:09:28 -07:00
parent 1d719a7c94
commit 2f4f20862d

View File

@ -77,7 +77,7 @@ vocab_size = tokenizer.get_vocab_size()
special_set = set(tokenizer.get_special_tokens())
token_strings = [tokenizer.decode([token_id]) for token_id in range(vocab_size)]
token_bytes = []
for token_str in token_strings:
for token_str in token_strings: # the Python string representation of this token
if token_str in special_set:
token_bytes.append(0) # special characters are not counted
else: