Update token config
#2
by
dualism2
- opened
- special_tokens_map.json +6 -1
- tokenizer_config.json +5 -1
special_tokens_map.json
CHANGED
|
@@ -1 +1,6 @@
|
|
| 1 |
-
{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<|begin_of_text|>",
|
| 3 |
+
"eos_token": "<|end_of_text|>",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"unk_token": "[UNK]"
|
| 6 |
+
}
|
tokenizer_config.json
CHANGED
|
@@ -33,8 +33,12 @@
|
|
| 33 |
"special": true
|
| 34 |
}
|
| 35 |
},
|
|
|
|
| 36 |
"clean_up_tokenization_spaces": false,
|
|
|
|
| 37 |
"extra_special_tokens": {},
|
| 38 |
"model_max_length": 1000000000000000019884624838656,
|
| 39 |
-
"
|
|
|
|
|
|
|
| 40 |
}
|
|
|
|
| 33 |
"special": true
|
| 34 |
}
|
| 35 |
},
|
| 36 |
+
"bos_token": "<|begin_of_text|>",
|
| 37 |
"clean_up_tokenization_spaces": false,
|
| 38 |
+
"eos_token": "<|end_of_text|>",
|
| 39 |
"extra_special_tokens": {},
|
| 40 |
"model_max_length": 1000000000000000019884624838656,
|
| 41 |
+
"pad_token": "[PAD]",
|
| 42 |
+
"tokenizer_class": "PreTrainedTokenizer",
|
| 43 |
+
"unk_token": "[UNK]"
|
| 44 |
}
|