Fix tokenizer_config: move lang codes to additional_special_tokens for transformers<5 compat
Browse files- tokenizer_config.json +16 -15
tokenizer_config.json
CHANGED
|
@@ -3,7 +3,20 @@
|
|
| 3 |
"bos_token": "<s>",
|
| 4 |
"cls_token": "<s>",
|
| 5 |
"eos_token": "</s>",
|
| 6 |
-
"extra_special_tokens":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
"ace_Arab",
|
| 8 |
"ace_Latn",
|
| 9 |
"acm_Arab",
|
|
@@ -206,17 +219,5 @@
|
|
| 206 |
"zho_Hans",
|
| 207 |
"zho_Hant",
|
| 208 |
"zul_Latn"
|
| 209 |
-
]
|
| 210 |
-
|
| 211 |
-
"legacy_behaviour": false,
|
| 212 |
-
"local_files_only": false,
|
| 213 |
-
"mask_token": "<mask>",
|
| 214 |
-
"model_max_length": 1024,
|
| 215 |
-
"pad_token": "<pad>",
|
| 216 |
-
"sep_token": "</s>",
|
| 217 |
-
"sp_model_kwargs": {},
|
| 218 |
-
"src_lang": "eng_Latn",
|
| 219 |
-
"tgt_lang": null,
|
| 220 |
-
"tokenizer_class": "NllbTokenizer",
|
| 221 |
-
"unk_token": "<unk>"
|
| 222 |
-
}
|
|
|
|
| 3 |
"bos_token": "<s>",
|
| 4 |
"cls_token": "<s>",
|
| 5 |
"eos_token": "</s>",
|
| 6 |
+
"extra_special_tokens": {},
|
| 7 |
+
"is_local": false,
|
| 8 |
+
"legacy_behaviour": false,
|
| 9 |
+
"local_files_only": false,
|
| 10 |
+
"mask_token": "<mask>",
|
| 11 |
+
"model_max_length": 1024,
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"sp_model_kwargs": {},
|
| 15 |
+
"src_lang": "eng_Latn",
|
| 16 |
+
"tgt_lang": null,
|
| 17 |
+
"tokenizer_class": "NllbTokenizer",
|
| 18 |
+
"unk_token": "<unk>",
|
| 19 |
+
"additional_special_tokens": [
|
| 20 |
"ace_Arab",
|
| 21 |
"ace_Latn",
|
| 22 |
"acm_Arab",
|
|
|
|
| 219 |
"zho_Hans",
|
| 220 |
"zho_Hant",
|
| 221 |
"zul_Latn"
|
| 222 |
+
]
|
| 223 |
+
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|