fd
Browse files- tokenizer_config.json +14 -16
tokenizer_config.json
CHANGED
|
@@ -1,19 +1,17 @@
|
|
|
|
|
| 1 |
{
|
| 2 |
"auto_map": {
|
| 3 |
-
"AutoTokenizer": [
|
|
|
|
|
|
|
|
|
|
| 4 |
},
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
|
| 13 |
-
|
| 14 |
-
"unk_token": "</s>",
|
| 15 |
-
"vocab_size": 10187
|
| 16 |
-
}
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
|
|
|
|
| 1 |
+
|
| 2 |
{
|
| 3 |
"auto_map": {
|
| 4 |
+
"AutoTokenizer": [
|
| 5 |
+
"tokenizeConfig.OBITokenizer",
|
| 6 |
+
null
|
| 7 |
+
]
|
| 8 |
},
|
| 9 |
+
"bos_token": "<s>",
|
| 10 |
+
"clean_up_tokenization_spaces": false,
|
| 11 |
+
"eos_token": "</s>",
|
| 12 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 13 |
+
"pad_token": "</s>",
|
| 14 |
+
"tokenizer_class": "OBITokenizer",
|
| 15 |
+
"unk_token": "<unk>"
|
| 16 |
+
}
|
| 17 |
+
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|