Brownwang0426 commited on
Commit
45aa1cd
1 Parent(s): df0ca5b

Upload tokenizer

Browse files
Files changed (1) hide show
  1. tokenizer_config.json +1 -8
tokenizer_config.json CHANGED
@@ -2069,18 +2069,11 @@
2069
  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
2070
  "clean_up_tokenization_spaces": true,
2071
  "eos_token": "<|eot_id|>",
2072
- "max_length": 512,
2073
  "model_input_names": [
2074
  "input_ids",
2075
  "attention_mask"
2076
  ],
2077
  "model_max_length": 1000000000000000019884624838656,
2078
- "pad_to_multiple_of": null,
2079
  "pad_token": "<|end_of_text|>",
2080
- "pad_token_type_id": 0,
2081
- "padding_side": "right",
2082
- "stride": 0,
2083
- "tokenizer_class": "PreTrainedTokenizerFast",
2084
- "truncation_side": "right",
2085
- "truncation_strategy": "longest_first"
2086
  }
 
2069
  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
2070
  "clean_up_tokenization_spaces": true,
2071
  "eos_token": "<|eot_id|>",
 
2072
  "model_input_names": [
2073
  "input_ids",
2074
  "attention_mask"
2075
  ],
2076
  "model_max_length": 1000000000000000019884624838656,
 
2077
  "pad_token": "<|end_of_text|>",
2078
+ "tokenizer_class": "PreTrainedTokenizerFast"
 
 
 
 
 
2079
  }