Upload tokenizer
Browse files- added_tokens.json +2 -3
- special_tokens_map.json +0 -7
- vocab.json +2 -2
added_tokens.json
CHANGED
|
@@ -1,5 +1,4 @@
|
|
| 1 |
{
|
| 2 |
-
"</s>":
|
| 3 |
-
"<s>":
|
| 4 |
-
"[PAD]": 65
|
| 5 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"</s>": 65,
|
| 3 |
+
"<s>": 64
|
|
|
|
| 4 |
}
|
special_tokens_map.json
CHANGED
|
@@ -13,13 +13,6 @@
|
|
| 13 |
"normalized": true,
|
| 14 |
"rstrip": false,
|
| 15 |
"single_word": false
|
| 16 |
-
},
|
| 17 |
-
{
|
| 18 |
-
"content": "[PAD]",
|
| 19 |
-
"lstrip": false,
|
| 20 |
-
"normalized": true,
|
| 21 |
-
"rstrip": false,
|
| 22 |
-
"single_word": false
|
| 23 |
}
|
| 24 |
],
|
| 25 |
"bos_token": "<s>",
|
|
|
|
| 13 |
"normalized": true,
|
| 14 |
"rstrip": false,
|
| 15 |
"single_word": false
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
}
|
| 17 |
],
|
| 18 |
"bos_token": "<s>",
|
vocab.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"ben": {
|
| 3 |
-
"[PAD]":
|
| 4 |
-
"[UNK]":
|
| 5 |
"|": 0,
|
| 6 |
"ঁ": 1,
|
| 7 |
"ং": 2,
|
|
|
|
| 1 |
{
|
| 2 |
"ben": {
|
| 3 |
+
"[PAD]": 63,
|
| 4 |
+
"[UNK]": 62,
|
| 5 |
"|": 0,
|
| 6 |
"ঁ": 1,
|
| 7 |
"ং": 2,
|