Upload tokenizer
Browse files- added_tokens.json +2 -2
- tokenizer_config.json +4 -4
- vocab.json +29 -20
added_tokens.json
CHANGED
|
@@ -1,4 +1,4 @@
|
|
| 1 |
{
|
| 2 |
-
"</s>":
|
| 3 |
-
"<s>":
|
| 4 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"</s>": 49,
|
| 3 |
+
"<s>": 48
|
| 4 |
}
|
tokenizer_config.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"added_tokens_decoder": {
|
| 3 |
-
"
|
| 4 |
"content": "[UNK]",
|
| 5 |
"lstrip": true,
|
| 6 |
"normalized": false,
|
|
@@ -8,7 +8,7 @@
|
|
| 8 |
"single_word": false,
|
| 9 |
"special": false
|
| 10 |
},
|
| 11 |
-
"
|
| 12 |
"content": "[PAD]",
|
| 13 |
"lstrip": true,
|
| 14 |
"normalized": false,
|
|
@@ -16,7 +16,7 @@
|
|
| 16 |
"single_word": false,
|
| 17 |
"special": false
|
| 18 |
},
|
| 19 |
-
"
|
| 20 |
"content": "<s>",
|
| 21 |
"lstrip": false,
|
| 22 |
"normalized": false,
|
|
@@ -24,7 +24,7 @@
|
|
| 24 |
"single_word": false,
|
| 25 |
"special": true
|
| 26 |
},
|
| 27 |
-
"
|
| 28 |
"content": "</s>",
|
| 29 |
"lstrip": false,
|
| 30 |
"normalized": false,
|
|
|
|
| 1 |
{
|
| 2 |
"added_tokens_decoder": {
|
| 3 |
+
"46": {
|
| 4 |
"content": "[UNK]",
|
| 5 |
"lstrip": true,
|
| 6 |
"normalized": false,
|
|
|
|
| 8 |
"single_word": false,
|
| 9 |
"special": false
|
| 10 |
},
|
| 11 |
+
"47": {
|
| 12 |
"content": "[PAD]",
|
| 13 |
"lstrip": true,
|
| 14 |
"normalized": false,
|
|
|
|
| 16 |
"single_word": false,
|
| 17 |
"special": false
|
| 18 |
},
|
| 19 |
+
"48": {
|
| 20 |
"content": "<s>",
|
| 21 |
"lstrip": false,
|
| 22 |
"normalized": false,
|
|
|
|
| 24 |
"single_word": false,
|
| 25 |
"special": true
|
| 26 |
},
|
| 27 |
+
"49": {
|
| 28 |
"content": "</s>",
|
| 29 |
"lstrip": false,
|
| 30 |
"normalized": false,
|
vocab.json
CHANGED
|
@@ -17,25 +17,34 @@
|
|
| 17 |
"N": 16,
|
| 18 |
"O": 17,
|
| 19 |
"S": 18,
|
| 20 |
-
"[PAD]":
|
| 21 |
-
"[UNK]":
|
| 22 |
-
"
|
| 23 |
-
"
|
| 24 |
-
"
|
| 25 |
-
"
|
| 26 |
-
"
|
| 27 |
-
"
|
| 28 |
-
"
|
| 29 |
-
"
|
| 30 |
-
"
|
| 31 |
-
"
|
| 32 |
-
"
|
| 33 |
-
"
|
| 34 |
-
"
|
| 35 |
-
"
|
| 36 |
-
"
|
| 37 |
-
"
|
| 38 |
-
"
|
| 39 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
"|": 0
|
| 41 |
}
|
|
|
|
| 17 |
"N": 16,
|
| 18 |
"O": 17,
|
| 19 |
"S": 18,
|
| 20 |
+
"[PAD]": 47,
|
| 21 |
+
"[UNK]": 46,
|
| 22 |
+
"_": 19,
|
| 23 |
+
"a": 20,
|
| 24 |
+
"b": 21,
|
| 25 |
+
"c": 22,
|
| 26 |
+
"d": 23,
|
| 27 |
+
"e": 24,
|
| 28 |
+
"f": 25,
|
| 29 |
+
"g": 26,
|
| 30 |
+
"h": 27,
|
| 31 |
+
"i": 28,
|
| 32 |
+
"j": 29,
|
| 33 |
+
"k": 30,
|
| 34 |
+
"l": 31,
|
| 35 |
+
"m": 32,
|
| 36 |
+
"n": 33,
|
| 37 |
+
"o": 34,
|
| 38 |
+
"p": 35,
|
| 39 |
+
"q": 36,
|
| 40 |
+
"r": 37,
|
| 41 |
+
"s": 38,
|
| 42 |
+
"t": 39,
|
| 43 |
+
"u": 40,
|
| 44 |
+
"v": 41,
|
| 45 |
+
"w": 42,
|
| 46 |
+
"x": 43,
|
| 47 |
+
"y": 44,
|
| 48 |
+
"z": 45,
|
| 49 |
"|": 0
|
| 50 |
}
|