Dejiao Z
commited on
Commit
•
2228afc
1
Parent(s):
a53aec6
fixed tokenizer
Browse files- tokenizer_config.json +6 -4
tokenizer_config.json
CHANGED
|
@@ -23,10 +23,12 @@
|
|
| 23 |
],
|
| 24 |
"bos_token": "<|endoftext|>",
|
| 25 |
"eos_token": "<|endoftext|>",
|
|
|
|
| 26 |
"model_max_length": 1000000000000000019884624838656,
|
| 27 |
-
"name_or_path": "/mnt/efs/people/dejiaoz/universal_embedding/codesage_v3/tokenizer/starcoder/",
|
| 28 |
-
"special_tokens_map_file": "/mnt/efs/people/dejiaoz/universal_embedding/codesage_v3/tokenizer/starcoder/special_tokens_map.json",
|
| 29 |
-
"tokenizer_class": "GPT2Tokenizer",
|
| 30 |
"unk_token": "<|endoftext|>",
|
| 31 |
-
"vocab_size": 49152
|
|
|
|
|
|
|
|
|
|
|
|
|
| 32 |
}
|
|
|
|
| 23 |
],
|
| 24 |
"bos_token": "<|endoftext|>",
|
| 25 |
"eos_token": "<|endoftext|>",
|
| 26 |
+
"add_eos_token": true,
|
| 27 |
"model_max_length": 1000000000000000019884624838656,
|
|
|
|
|
|
|
|
|
|
| 28 |
"unk_token": "<|endoftext|>",
|
| 29 |
+
"vocab_size": 49152,
|
| 30 |
+
"tokenizer_class": "CodeSageTokenizer",
|
| 31 |
+
"auto_map": {
|
| 32 |
+
"AutoTokenizer": ["tokenization_codesage.CodeSageTokenizer", null]
|
| 33 |
+
}
|
| 34 |
}
|