Upload tokenizer
Browse files- added_tokens.json +1 -2
- special_tokens_map.json +1 -1
- tokenizer_config.json +9 -9
added_tokens.json
CHANGED
|
@@ -1,7 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"</s>": 305,
|
| 3 |
-
"<
|
| 4 |
-
"<mask>": 308,
|
| 5 |
"<s>": 304,
|
| 6 |
"<unk>": 306
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"</s>": 305,
|
| 3 |
+
"<mask>": 307,
|
|
|
|
| 4 |
"<s>": 304,
|
| 5 |
"<unk>": 306
|
| 6 |
}
|
special_tokens_map.json
CHANGED
|
@@ -28,7 +28,7 @@
|
|
| 28 |
"single_word": false
|
| 29 |
},
|
| 30 |
"pad_token": {
|
| 31 |
-
"content": "<
|
| 32 |
"lstrip": false,
|
| 33 |
"normalized": true,
|
| 34 |
"rstrip": false,
|
|
|
|
| 28 |
"single_word": false
|
| 29 |
},
|
| 30 |
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
"lstrip": false,
|
| 33 |
"normalized": true,
|
| 34 |
"rstrip": false,
|
tokenizer_config.json
CHANGED
|
@@ -1,6 +1,14 @@
|
|
| 1 |
{
|
| 2 |
"add_prefix_space": false,
|
| 3 |
"added_tokens_decoder": {
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
"304": {
|
| 5 |
"content": "<s>",
|
| 6 |
"lstrip": false,
|
|
@@ -26,14 +34,6 @@
|
|
| 26 |
"special": true
|
| 27 |
},
|
| 28 |
"307": {
|
| 29 |
-
"content": "<PAD>",
|
| 30 |
-
"lstrip": false,
|
| 31 |
-
"normalized": true,
|
| 32 |
-
"rstrip": false,
|
| 33 |
-
"single_word": false,
|
| 34 |
-
"special": true
|
| 35 |
-
},
|
| 36 |
-
"308": {
|
| 37 |
"content": "<mask>",
|
| 38 |
"lstrip": true,
|
| 39 |
"normalized": true,
|
|
@@ -49,7 +49,7 @@
|
|
| 49 |
"errors": "replace",
|
| 50 |
"mask_token": "<mask>",
|
| 51 |
"model_max_length": 1000000000000000019884624838656,
|
| 52 |
-
"pad_token": "<
|
| 53 |
"sep_token": "</s>",
|
| 54 |
"tokenizer_class": "BartTokenizer",
|
| 55 |
"unk_token": "<unk>"
|
|
|
|
| 1 |
{
|
| 2 |
"add_prefix_space": false,
|
| 3 |
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<pad>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
"304": {
|
| 13 |
"content": "<s>",
|
| 14 |
"lstrip": false,
|
|
|
|
| 34 |
"special": true
|
| 35 |
},
|
| 36 |
"307": {
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
"content": "<mask>",
|
| 38 |
"lstrip": true,
|
| 39 |
"normalized": true,
|
|
|
|
| 49 |
"errors": "replace",
|
| 50 |
"mask_token": "<mask>",
|
| 51 |
"model_max_length": 1000000000000000019884624838656,
|
| 52 |
+
"pad_token": "<pad>",
|
| 53 |
"sep_token": "</s>",
|
| 54 |
"tokenizer_class": "BartTokenizer",
|
| 55 |
"unk_token": "<unk>"
|