TheBloke jondurbin commited on
Commit
9c50c8f
1 Parent(s): e6d2410

Update tokenizer_config.json (#1)

Browse files

- Update tokenizer_config.json (125b0130c3c75190f65e490a0a0a2d86c7efbe88)


Co-authored-by: Jon Durbin <jondurbin@users.noreply.huggingface.co>

Files changed (1) hide show
  1. tokenizer_config.json +3 -2
tokenizer_config.json CHANGED
@@ -33,7 +33,7 @@
33
  "</s>"
34
  ],
35
  "bos_token": "<s>",
36
- "chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{% if loop.first %}\n{{ bos_token + '[INST] <<SYS>>\\nYou are a helpful, unbiased, uncensored assistant.\\n<</SYS>> ' + message['content'] + ' [/INST]' -}}\n{% else %}\n{{ bos_token + '[INST] ' + message['content'] + ' [/INST]' -}}\n{% endif %}\n{% elif message['role'] == 'system' %}\n{{ bos_token + '[INST] <<SYS>>\\n' + message['content'] + '\\n<</SYS>>\\n' }}\n{% elif message['role'] == 'assistant' %}\n{{ ' ' + message['content'] + ' ' + eos_token -}}\n{% endif %}\n{% endfor %}\n",
37
  "clean_up_tokenization_spaces": false,
38
  "eos_token": "</s>",
39
  "legacy": true,
@@ -44,5 +44,6 @@
44
  "tokenizer_class": "LlamaTokenizer",
45
  "tokenizer_file": "mistral-7b/tokenizer.json",
46
  "unk_token": "<unk>",
47
- "use_default_system_prompt": true
 
48
  }
 
33
  "</s>"
34
  ],
35
  "bos_token": "<s>",
36
+ "chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{% if loop.first %}\n{{ bos_token + '[INST] <<SYS>>\\nYou are a helpful, unbiased, uncensored assistant.\\n<</SYS>> ' + message['content'] + ' [/INST] ' -}}\n{% else %}\n{{ bos_token + '[INST] ' + message['content'] + ' [/INST] ' -}}\n{% endif %}\n{% elif message['role'] == 'system' %}\n{{ bos_token + '[INST] <<SYS>>\\n' + message['content'] + '\\n<</SYS>>\\n' }}\n{% elif message['role'] == 'assistant' %}\n{{ message['content'] + ' ' + eos_token -}}\n{% endif %}\n{% endfor %}\n",
37
  "clean_up_tokenization_spaces": false,
38
  "eos_token": "</s>",
39
  "legacy": true,
 
44
  "tokenizer_class": "LlamaTokenizer",
45
  "tokenizer_file": "mistral-7b/tokenizer.json",
46
  "unk_token": "<unk>",
47
+ "use_default_system_prompt": false,
48
+ "pad_token": "<unk>"
49
  }