Upload tokenizer
Browse files- vocab.json +30 -28
vocab.json
CHANGED
|
@@ -1,32 +1,34 @@
|
|
| 1 |
{
|
| 2 |
-
"[PAD]":
|
| 3 |
-
"[UNK]":
|
| 4 |
-
"a":
|
| 5 |
-
"b":
|
| 6 |
-
"c":
|
| 7 |
-
"d":
|
| 8 |
-
"e":
|
| 9 |
-
"f":
|
| 10 |
-
"g":
|
| 11 |
-
"h":
|
| 12 |
-
"i":
|
| 13 |
-
"j":
|
| 14 |
-
"k":
|
| 15 |
"l": 4,
|
| 16 |
-
"m":
|
| 17 |
"n": 13,
|
| 18 |
-
"o":
|
| 19 |
-
"p":
|
| 20 |
-
"q":
|
| 21 |
-
"r":
|
| 22 |
-
"s":
|
| 23 |
-
"t":
|
| 24 |
-
"u":
|
| 25 |
-
"v":
|
| 26 |
-
"w":
|
| 27 |
-
"x":
|
| 28 |
-
"y":
|
| 29 |
-
"z":
|
| 30 |
-
"|":
|
| 31 |
-
"
|
|
|
|
|
|
|
| 32 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"[PAD]": 31,
|
| 3 |
+
"[UNK]": 30,
|
| 4 |
+
"a": 18,
|
| 5 |
+
"b": 2,
|
| 6 |
+
"c": 0,
|
| 7 |
+
"d": 5,
|
| 8 |
+
"e": 26,
|
| 9 |
+
"f": 29,
|
| 10 |
+
"g": 27,
|
| 11 |
+
"h": 20,
|
| 12 |
+
"i": 15,
|
| 13 |
+
"j": 24,
|
| 14 |
+
"k": 16,
|
| 15 |
"l": 4,
|
| 16 |
+
"m": 23,
|
| 17 |
"n": 13,
|
| 18 |
+
"o": 17,
|
| 19 |
+
"p": 12,
|
| 20 |
+
"q": 28,
|
| 21 |
+
"r": 21,
|
| 22 |
+
"s": 3,
|
| 23 |
+
"t": 8,
|
| 24 |
+
"u": 11,
|
| 25 |
+
"v": 14,
|
| 26 |
+
"w": 7,
|
| 27 |
+
"x": 9,
|
| 28 |
+
"y": 10,
|
| 29 |
+
"z": 25,
|
| 30 |
+
"|": 22,
|
| 31 |
+
"ã": 1,
|
| 32 |
+
"ú": 6,
|
| 33 |
+
"ạ": 19
|
| 34 |
}
|