Transformers
tobiges commited on
Commit
b2422ee
·
verified ·
1 Parent(s): 94c1678

Upload processor

Browse files
Files changed (2) hide show
  1. processing_action_tokenizer.py +3 -9
  2. tokenizer.json +0 -0
processing_action_tokenizer.py CHANGED
@@ -164,15 +164,9 @@ class UniversalActionProcessor(ProcessorMixin):
164
 
165
  # Train BPE tokenizer
166
  tokenizer = Tokenizer(BPE())
167
- tokenizer.pre_tokenizer = pre_tokenizers.ByteLevel(
168
- add_prefix_space=False, use_regex=False, trim_offsets=False
169
- )
170
- tokenizer.decoder = decoders.ByteLevel(
171
- add_prefix_space=False, trim_offsets=False, use_regex=False
172
- )
173
- tokenizer.post_processor = processors.ByteLevel(
174
- add_prefix_space=False, trim_offsets=False, use_regex=False
175
- )
176
 
177
  trainer = BpeTrainer(
178
  vocab_size=vocab_size,
 
164
 
165
  # Train BPE tokenizer
166
  tokenizer = Tokenizer(BPE())
167
+ tokenizer.pre_tokenizer = pre_tokenizers.ByteLevel(add_prefix_space=False, use_regex=False)
168
+ tokenizer.decoder = decoders.ByteLevel()
169
+ tokenizer.post_processor = processors.ByteLevel(trim_offsets=False)
 
 
 
 
 
 
170
 
171
  trainer = BpeTrainer(
172
  vocab_size=vocab_size,
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff