Upload tokenizer
Browse files- special_tokens_map.json +1 -0
- tokenizer.json +9 -0
special_tokens_map.json
CHANGED
|
@@ -2,5 +2,6 @@
|
|
| 2 |
"bos_token": "<|endoftext|>",
|
| 3 |
"eos_token": "<|endoftext|>",
|
| 4 |
"pad_token": "[PAD]",
|
|
|
|
| 5 |
"unk_token": "<|endoftext|>"
|
| 6 |
}
|
|
|
|
| 2 |
"bos_token": "<|endoftext|>",
|
| 3 |
"eos_token": "<|endoftext|>",
|
| 4 |
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
"unk_token": "<|endoftext|>"
|
| 7 |
}
|
tokenizer.json
CHANGED
|
@@ -9236,6 +9236,15 @@
|
|
| 9236 |
"rstrip": false,
|
| 9237 |
"normalized": false,
|
| 9238 |
"special": true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9239 |
}
|
| 9240 |
],
|
| 9241 |
"normalizer": {
|
|
|
|
| 9236 |
"rstrip": false,
|
| 9237 |
"normalized": false,
|
| 9238 |
"special": true
|
| 9239 |
+
},
|
| 9240 |
+
{
|
| 9241 |
+
"id": 1026,
|
| 9242 |
+
"content": "[SEP]",
|
| 9243 |
+
"single_word": false,
|
| 9244 |
+
"lstrip": false,
|
| 9245 |
+
"rstrip": false,
|
| 9246 |
+
"normalized": false,
|
| 9247 |
+
"special": true
|
| 9248 |
}
|
| 9249 |
],
|
| 9250 |
"normalizer": {
|