Upload processor
Browse files- tokenizer.json +0 -0
- tokenizer_config.json +7 -0
tokenizer.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
CHANGED
@@ -11273,8 +11273,10 @@
|
|
11273 |
"eos_token": "</s>",
|
11274 |
"extra_ids": 100,
|
11275 |
"legacy": true,
|
|
|
11276 |
"model_max_length": 512,
|
11277 |
"only_label_first_subword": true,
|
|
|
11278 |
"pad_token": "<pad>",
|
11279 |
"pad_token_box": [
|
11280 |
0,
|
@@ -11283,6 +11285,8 @@
|
|
11283 |
0
|
11284 |
],
|
11285 |
"pad_token_label": -100,
|
|
|
|
|
11286 |
"processor_class": "UdopProcessor",
|
11287 |
"sep_token": "</s>",
|
11288 |
"sep_token_box": [
|
@@ -11292,7 +11296,10 @@
|
|
11292 |
1000
|
11293 |
],
|
11294 |
"sp_model_kwargs": {},
|
|
|
11295 |
"tokenizer_class": "UdopTokenizer",
|
|
|
|
|
11296 |
"unk_token": "<unk>",
|
11297 |
"use_fast": true
|
11298 |
}
|
|
|
11273 |
"eos_token": "</s>",
|
11274 |
"extra_ids": 100,
|
11275 |
"legacy": true,
|
11276 |
+
"max_length": 128,
|
11277 |
"model_max_length": 512,
|
11278 |
"only_label_first_subword": true,
|
11279 |
+
"pad_to_multiple_of": null,
|
11280 |
"pad_token": "<pad>",
|
11281 |
"pad_token_box": [
|
11282 |
0,
|
|
|
11285 |
0
|
11286 |
],
|
11287 |
"pad_token_label": -100,
|
11288 |
+
"pad_token_type_id": 0,
|
11289 |
+
"padding_side": "right",
|
11290 |
"processor_class": "UdopProcessor",
|
11291 |
"sep_token": "</s>",
|
11292 |
"sep_token_box": [
|
|
|
11296 |
1000
|
11297 |
],
|
11298 |
"sp_model_kwargs": {},
|
11299 |
+
"stride": 0,
|
11300 |
"tokenizer_class": "UdopTokenizer",
|
11301 |
+
"truncation_side": "right",
|
11302 |
+
"truncation_strategy": "longest_first",
|
11303 |
"unk_token": "<unk>",
|
11304 |
"use_fast": true
|
11305 |
}
|