Tom9358 commited on
Commit
1dfb0a1
·
verified ·
1 Parent(s): 9e8dca0

Upload tokenizer

Browse files
Files changed (2) hide show
  1. tokenizer.json +2 -2
  2. tokenizer_config.json +7 -0
tokenizer.json CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7775ee533359d69ca21c708fd822bceb98d282ac189bf43a5b1a3de1f1ecc518
3
- size 31923943
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d8596f068fc4cb0aaee6a18903cd59ef84828f48c785ae8f1e9e59a84ddc036
3
+ size 32240574
tokenizer_config.json CHANGED
@@ -9,12 +9,19 @@
9
  "is_local": true,
10
  "legacy_behaviour": false,
11
  "mask_token": "<mask>",
 
12
  "model_max_length": 1024,
 
13
  "pad_token": "<pad>",
 
 
14
  "sep_token": "</s>",
15
  "sp_model_kwargs": {},
16
  "src_lang": "nld_Latn",
 
17
  "tgt_lang": "gos_Latn",
18
  "tokenizer_class": "NllbTokenizer",
 
 
19
  "unk_token": "<unk>"
20
  }
 
9
  "is_local": true,
10
  "legacy_behaviour": false,
11
  "mask_token": "<mask>",
12
+ "max_length": 43,
13
  "model_max_length": 1024,
14
+ "pad_to_multiple_of": null,
15
  "pad_token": "<pad>",
16
+ "pad_token_type_id": 0,
17
+ "padding_side": "right",
18
  "sep_token": "</s>",
19
  "sp_model_kwargs": {},
20
  "src_lang": "nld_Latn",
21
+ "stride": 0,
22
  "tgt_lang": "gos_Latn",
23
  "tokenizer_class": "NllbTokenizer",
24
+ "truncation_side": "right",
25
+ "truncation_strategy": "longest_first",
26
  "unk_token": "<unk>"
27
  }