Bump model_max_length to 10000 (matches dnabert2/dnaberts; MosaicBERT/ALiBi extrapolates)
Browse files- tokenizer_config.json +7 -4
tokenizer_config.json
CHANGED
|
@@ -4,9 +4,9 @@
|
|
| 4 |
"do_basic_tokenize": true,
|
| 5 |
"do_lower_case": false,
|
| 6 |
"mask_token": "[MASK]",
|
| 7 |
-
"max_len":
|
| 8 |
-
"max_length":
|
| 9 |
-
"model_max_length":
|
| 10 |
"never_split": null,
|
| 11 |
"pad_to_multiple_of": null,
|
| 12 |
"pad_token": "[PAD]",
|
|
@@ -17,7 +17,10 @@
|
|
| 17 |
"strip_accents": null,
|
| 18 |
"tokenize_chinese_chars": true,
|
| 19 |
"auto_map": {
|
| 20 |
-
"AutoTokenizer": [
|
|
|
|
|
|
|
|
|
|
| 21 |
},
|
| 22 |
"tokenizer_class": "MRNABERTTokenizer",
|
| 23 |
"truncation_side": "right",
|
|
|
|
| 4 |
"do_basic_tokenize": true,
|
| 5 |
"do_lower_case": false,
|
| 6 |
"mask_token": "[MASK]",
|
| 7 |
+
"max_len": 10000,
|
| 8 |
+
"max_length": 10000,
|
| 9 |
+
"model_max_length": 10000,
|
| 10 |
"never_split": null,
|
| 11 |
"pad_to_multiple_of": null,
|
| 12 |
"pad_token": "[PAD]",
|
|
|
|
| 17 |
"strip_accents": null,
|
| 18 |
"tokenize_chinese_chars": true,
|
| 19 |
"auto_map": {
|
| 20 |
+
"AutoTokenizer": [
|
| 21 |
+
"tokenization_mrnabert.MRNABERTTokenizer",
|
| 22 |
+
null
|
| 23 |
+
]
|
| 24 |
},
|
| 25 |
"tokenizer_class": "MRNABERTTokenizer",
|
| 26 |
"truncation_side": "right",
|