marigold334 commited on
Commit
5b2b713
1 Parent(s): f1e54a9

Update datautils.py (#22)

Browse files

- Update datautils.py (cc49fe56bc3ad6edbd157f5a6144e7d86fc04fd6)

Files changed (1) hide show
  1. datautils.py +2 -1
datautils.py CHANGED
@@ -18,8 +18,9 @@ SPACE = ' '
18
  JAMO_LEADS = "".join([chr(_) for _ in range(0x1100, 0x1113)])
19
  JAMO_VOWELS = "".join([chr(_) for _ in range(0x1161, 0x1176)])
20
  JAMO_TAILS = "".join([chr(_) for _ in range(0x11A8, 0x11C3)])
 
21
 
22
- VALID_CHARS = JAMO_LEADS + JAMO_VOWELS + JAMO_TAILS + SPACE
23
  symbols = PAD + EOS + VALID_CHARS
24
 
25
  _symbol_to_id = {s: i for i, s in enumerate(symbols)}
 
18
  JAMO_LEADS = "".join([chr(_) for _ in range(0x1100, 0x1113)])
19
  JAMO_VOWELS = "".join([chr(_) for _ in range(0x1161, 0x1176)])
20
  JAMO_TAILS = "".join([chr(_) for _ in range(0x11A8, 0x11C3)])
21
+ ETC = ".!?"
22
 
23
+ VALID_CHARS = JAMO_LEADS + JAMO_VOWELS + JAMO_TAILS + SPACE + ETC
24
  symbols = PAD + EOS + VALID_CHARS
25
 
26
  _symbol_to_id = {s: i for i, s in enumerate(symbols)}