HamAndCheese82 commited on
Commit
77707f7
·
verified ·
1 Parent(s): 5f34972

On validation process, epoch 4, global step 164236

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
@@ -130,11 +130,18 @@
130
  "cls_token": "<s>",
131
  "eos_token": "</s>",
132
  "mask_token": "<mask>",
 
133
  "model_max_length": 1000000000000000019884624838656,
 
134
  "pad_token": "<pad>",
 
 
135
  "processor_class": "DonutProcessor",
136
  "sep_token": "</s>",
137
  "sp_model_kwargs": {},
 
138
  "tokenizer_class": "XLMRobertaTokenizer",
 
 
139
  "unk_token": "<unk>"
140
  }
 
130
  "cls_token": "<s>",
131
  "eos_token": "</s>",
132
  "mask_token": "<mask>",
133
+ "max_length": 512,
134
  "model_max_length": 1000000000000000019884624838656,
135
+ "pad_to_multiple_of": null,
136
  "pad_token": "<pad>",
137
+ "pad_token_type_id": 0,
138
+ "padding_side": "right",
139
  "processor_class": "DonutProcessor",
140
  "sep_token": "</s>",
141
  "sp_model_kwargs": {},
142
+ "stride": 0,
143
  "tokenizer_class": "XLMRobertaTokenizer",
144
+ "truncation_side": "right",
145
+ "truncation_strategy": "longest_first",
146
  "unk_token": "<unk>"
147
  }