can.turan
commited on
Commit
·
7833aec
1
Parent(s):
04250db
Tokenizer fix
Browse files- tokenizer_config.json +8 -4
tokenizer_config.json
CHANGED
|
@@ -47,9 +47,9 @@
|
|
| 47 |
"do_lower_case": true,
|
| 48 |
"extra_special_tokens": {},
|
| 49 |
"mask_token": "[MASK]",
|
| 50 |
-
"max_len":
|
| 51 |
-
"max_length":
|
| 52 |
-
"model_max_length":
|
| 53 |
"never_split": null,
|
| 54 |
"pad_to_multiple_of": null,
|
| 55 |
"pad_token": "[PAD]",
|
|
@@ -59,8 +59,12 @@
|
|
| 59 |
"stride": 0,
|
| 60 |
"strip_accents": false,
|
| 61 |
"tokenize_chinese_chars": true,
|
| 62 |
-
"tokenizer_class": "
|
| 63 |
"truncation_side": "right",
|
| 64 |
"truncation_strategy": "longest_first",
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
"unk_token": "[UNK]"
|
| 66 |
}
|
|
|
|
| 47 |
"do_lower_case": true,
|
| 48 |
"extra_special_tokens": {},
|
| 49 |
"mask_token": "[MASK]",
|
| 50 |
+
"max_len": 8192,
|
| 51 |
+
"max_length": 8192,
|
| 52 |
+
"model_max_length": 8192,
|
| 53 |
"never_split": null,
|
| 54 |
"pad_to_multiple_of": null,
|
| 55 |
"pad_token": "[PAD]",
|
|
|
|
| 59 |
"stride": 0,
|
| 60 |
"strip_accents": false,
|
| 61 |
"tokenize_chinese_chars": true,
|
| 62 |
+
"tokenizer_class": "PreTrainedTokenizerFast",
|
| 63 |
"truncation_side": "right",
|
| 64 |
"truncation_strategy": "longest_first",
|
| 65 |
+
"model_input_names": [
|
| 66 |
+
"input_ids",
|
| 67 |
+
"attention_mask"
|
| 68 |
+
],
|
| 69 |
"unk_token": "[UNK]"
|
| 70 |
}
|