DeepeshAlwani commited on
Commit
a87acec
1 Parent(s): fe454b3

Training done

Browse files
Files changed (2) hide show
  1. tokenizer.json +0 -0
  2. 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
@@ -769,11 +769,18 @@
769
  "cls_token": "<s>",
770
  "eos_token": "</s>",
771
  "mask_token": "<mask>",
 
772
  "model_max_length": 1000000000000000019884624838656,
 
773
  "pad_token": "<pad>",
 
 
774
  "processor_class": "DonutProcessor",
775
  "sep_token": "</s>",
776
  "sp_model_kwargs": {},
 
777
  "tokenizer_class": "XLMRobertaTokenizer",
 
 
778
  "unk_token": "<unk>"
779
  }
 
769
  "cls_token": "<s>",
770
  "eos_token": "</s>",
771
  "mask_token": "<mask>",
772
+ "max_length": 768,
773
  "model_max_length": 1000000000000000019884624838656,
774
+ "pad_to_multiple_of": null,
775
  "pad_token": "<pad>",
776
+ "pad_token_type_id": 0,
777
+ "padding_side": "right",
778
  "processor_class": "DonutProcessor",
779
  "sep_token": "</s>",
780
  "sp_model_kwargs": {},
781
+ "stride": 0,
782
  "tokenizer_class": "XLMRobertaTokenizer",
783
+ "truncation_side": "right",
784
+ "truncation_strategy": "longest_first",
785
  "unk_token": "<unk>"
786
  }