danielhanchen commited on
Commit
d51fa04
·
verified ·
1 Parent(s): d75be36

Update tokenizer_config.json

Browse files
Files changed (1) hide show
  1. tokenizer_config.json +1 -2
tokenizer_config.json CHANGED
@@ -1010,6 +1010,5 @@
1010
  "pad_token": "<pad>",
1011
  "padding_side": "left",
1012
  "tokenizer_class": "TokenizersBackend",
1013
- "unk_token": "<unk>",
1014
- "chat_template": "{#- Unsloth template fixes #}\n{#- Default system message if no system prompt is passed. #}\n{%- set default_system_message = '' %}\n\n{#- Begin of sequence token. #}\n{{- bos_token }}\n\n{#- Handle system prompt if it exists. #}\n{#- System prompt supports text content or text chunks. #}\n{%- if messages[0]['role'] == 'system' %}\n {{- '[SYSTEM_PROMPT]' -}}\n {%- if messages[0]['content'] is string %}\n {{- messages[0]['content'] -}}\n {%- else %} \n {%- for block in messages[0]['content'] %}\n {%- if block['type'] == 'text' %}\n {{- block['text'] }}\n {%- else %}\n {{- raise_exception('Only text chunks are supported in system message contents.') }}\n {%- endif %}\n {%- endfor %}\n {%- endif %}\n {{- '[/SYSTEM_PROMPT]' -}}\n {%- set loop_messages = messages[1:] %}\n{%- else %}\n {%- set loop_messages = messages %}\n {%- if default_system_message != '' %}\n {{- '[SYSTEM_PROMPT]' + default_system_message + '[/SYSTEM_PROMPT]' }}\n {%- endif %}\n{%- endif %}\n\n\n{#- Tools definition #}\n{%- set tools_definition = '' %}\n{%- set has_tools = false %}\n{%- if tools is defined and tools is not none and tools|length > 0 %}\n {%- set has_tools = true %}\n {%- set tools_definition = '[AVAILABLE_TOOLS]' + (tools| tojson) + '[/AVAILABLE_TOOLS]' %}\n {{- tools_definition }}\n{%- endif %}\n\n{#- Checks for alternating user/assistant messages. #}\n{%- set ns = namespace() %}\n{%- set ns.index = 0 %}\n{%- for message in loop_messages %}\n {%- if message.role == 'user' or (message.role == 'assistant' and (message.tool_calls is not defined or message.tool_calls is none or message.tool_calls | length == 0)) %}\n {%- if (message['role'] == 'user') != (ns.index % 2 == 0) %}\n {{- raise_exception('After the optional system message, conversation roles must alternate user and assistant roles except for tool calls and results.') }}\n {%- endif %}\n {%- set ns.index = ns.index + 1 %}\n {%- endif %}\n{%- endfor %}\n\n{#- Handle conversation messages. #}\n{%- for message in loop_messages %}\n\n {#- User messages supports text content. #}\n {%- if message['role'] == 'user' %}\n {%- if message['content'] is string %}\n {{- '[INST]' + message['content'] + '[/INST]' }}\n {%- elif message['content'] | length > 0 %}\n {{- '[INST]' }}\n {%- set sorted_blocks = message['content'] | sort(attribute='type') %}\n {%- for block in sorted_blocks %}\n {%- if block['type'] == 'text' %}\n {{- block['text'] }}\n {%- else %}\n {{- raise_exception('Only text chunks are supported in user message content.') }}\n {%- endif %}\n {%- endfor %}\n {{- '[/INST]' }}\n {%- else %}\n {{- raise_exception('User message must have a string or a list of chunks in content') }}\n {%- endif %}\n\n {#- Assistant messages supports text content or text chunks. #}\n {%- elif message['role'] == 'assistant' %}\n {%- if (message['content'] is none or message['content'] == '' or message['content']|length == 0) and (message['tool_calls'] is not defined or message['tool_calls'] is none or message['tool_calls']|length == 0) %}\n {{- raise_exception('Assistant message must have a string or a list of chunks in content or a list of tool calls.') }}\n {%- endif %}\n\n {%- if message['content'] is string and message['content'] != '' %}\n {{- message['content'] }}\n {%- elif message['content'] is iterable and message['content'] | length > 0 %}\n {%- for block in message['content'] %}\n {%- if block['type'] == 'text' %}\n {{- block['text'] }}\n {%- else %}\n {{- raise_exception('Only text chunks are supported in assistant message contents.') }}\n {%- endif %}\n {%- endfor %}\n {%- endif %}\n\n {%- if message['tool_calls'] is defined and message['tool_calls'] is not none and message['tool_calls']|length > 0 %}\n {%- for tool in message['tool_calls'] %}\n {{- '[TOOL_CALLS]' }}\n {%- set name = tool['function']['name'] %}\n {%- set arguments = tool['function']['arguments'] %}\n {%- if arguments is not string %}\n {%- set arguments = arguments|tojson|safe %}\n {%- elif arguments == '' %}\n {%- set arguments = '{}' %}\n {%- endif %}\n {{- name + '[ARGS]' + arguments }}\n {%- endfor %}\n {%- endif %}\n\n {{- eos_token }}\n\n {#- Tool messages only supports text content. #}\n {%- elif message['role'] == 'tool' %}\n {{- '[TOOL_RESULTS]' + message['content']|string + '[/TOOL_RESULTS]' }}\n\n {#- Raise exception for unsupported roles. #}\n {%- else %}\n {{- raise_exception('Only user, assistant and tool roles are supported, got ' + message['role'] + '.') }}\n {%- endif %}\n{%- endfor %}\n{#- Copyright 2025-present Unsloth. Apache 2.0 License. #}"
1015
  }
 
1010
  "pad_token": "<pad>",
1011
  "padding_side": "left",
1012
  "tokenizer_class": "TokenizersBackend",
1013
+ "unk_token": "<unk>"
 
1014
  }