Upload tokenizer
Browse files- special_tokens_map.json +7 -2
- tokenizer_config.json +2 -2
special_tokens_map.json
CHANGED
|
@@ -6,8 +6,13 @@
|
|
| 6 |
"rstrip": false,
|
| 7 |
"single_word": false
|
| 8 |
},
|
| 9 |
-
"eos_token":
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
"unk_token": {
|
| 12 |
"content": "<unk>",
|
| 13 |
"lstrip": false,
|
|
|
|
| 6 |
"rstrip": false,
|
| 7 |
"single_word": false
|
| 8 |
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
"unk_token": {
|
| 17 |
"content": "<unk>",
|
| 18 |
"lstrip": false,
|
tokenizer_config.json
CHANGED
|
@@ -29,9 +29,9 @@
|
|
| 29 |
},
|
| 30 |
"bos_token": "<s>",
|
| 31 |
"clean_up_tokenization_spaces": false,
|
| 32 |
-
"eos_token": "<
|
| 33 |
"model_max_length": 1000000000000000019884624838656,
|
| 34 |
-
"pad_token":
|
| 35 |
"sp_model_kwargs": {},
|
| 36 |
"tokenizer_class": "LlamaTokenizer",
|
| 37 |
"unk_token": "<unk>",
|
|
|
|
| 29 |
},
|
| 30 |
"bos_token": "<s>",
|
| 31 |
"clean_up_tokenization_spaces": false,
|
| 32 |
+
"eos_token": "</s>",
|
| 33 |
"model_max_length": 1000000000000000019884624838656,
|
| 34 |
+
"pad_token": null,
|
| 35 |
"sp_model_kwargs": {},
|
| 36 |
"tokenizer_class": "LlamaTokenizer",
|
| 37 |
"unk_token": "<unk>",
|