Upload tokenizer
Browse files- added_tokens.json +1 -1
- tokenizer_config.json +17 -1
- vocab.json +2 -2
added_tokens.json
CHANGED
|
@@ -1,4 +1,4 @@
|
|
| 1 |
{
|
| 2 |
"</s>": 67,
|
| 3 |
-
"
|
| 4 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"</s>": 67,
|
| 3 |
+
"<s>": 66
|
| 4 |
}
|
tokenizer_config.json
CHANGED
|
@@ -1,6 +1,14 @@
|
|
| 1 |
{
|
| 2 |
"added_tokens_decoder": {
|
| 3 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
"content": "[PAD]",
|
| 5 |
"lstrip": true,
|
| 6 |
"normalized": false,
|
|
@@ -8,6 +16,14 @@
|
|
| 8 |
"single_word": false,
|
| 9 |
"special": false
|
| 10 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
"67": {
|
| 12 |
"content": "</s>",
|
| 13 |
"lstrip": false,
|
|
|
|
| 1 |
{
|
| 2 |
"added_tokens_decoder": {
|
| 3 |
+
"64": {
|
| 4 |
+
"content": "[UNK]",
|
| 5 |
+
"lstrip": true,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": true,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": false
|
| 10 |
+
},
|
| 11 |
+
"65": {
|
| 12 |
"content": "[PAD]",
|
| 13 |
"lstrip": true,
|
| 14 |
"normalized": false,
|
|
|
|
| 16 |
"single_word": false,
|
| 17 |
"special": false
|
| 18 |
},
|
| 19 |
+
"66": {
|
| 20 |
+
"content": "<s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
"67": {
|
| 28 |
"content": "</s>",
|
| 29 |
"lstrip": false,
|
vocab.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"!": 1,
|
| 3 |
-
"[PAD]":
|
| 4 |
-
"[UNK]":
|
| 5 |
"|": 0,
|
| 6 |
"ँ": 2,
|
| 7 |
"ं": 3,
|
|
|
|
| 1 |
{
|
| 2 |
"!": 1,
|
| 3 |
+
"[PAD]": 65,
|
| 4 |
+
"[UNK]": 64,
|
| 5 |
"|": 0,
|
| 6 |
"ँ": 2,
|
| 7 |
"ं": 3,
|