lmflow.utils.conversation_template
==================================

.. py:module:: lmflow.utils.conversation_template


Submodules
----------

.. toctree::
   :maxdepth: 1

   /autoapi/lmflow/utils/conversation_template/base/index
   /autoapi/lmflow/utils/conversation_template/chatglm/index
   /autoapi/lmflow/utils/conversation_template/chatml/index
   /autoapi/lmflow/utils/conversation_template/deepseek/index
   /autoapi/lmflow/utils/conversation_template/gemma/index
   /autoapi/lmflow/utils/conversation_template/hymba/index
   /autoapi/lmflow/utils/conversation_template/internlm/index
   /autoapi/lmflow/utils/conversation_template/llama/index
   /autoapi/lmflow/utils/conversation_template/phi/index
   /autoapi/lmflow/utils/conversation_template/qwen/index
   /autoapi/lmflow/utils/conversation_template/yi/index
   /autoapi/lmflow/utils/conversation_template/zephyr/index


Classes
-------

.. autoapisummary::

   lmflow.utils.conversation_template.ConversationTemplate
   lmflow.utils.conversation_template.ConversationTemplateForTool


Package Contents
----------------

.. py:class:: ConversationTemplate

   .. py:attribute:: user_formatter
      :type:  Formatter


   .. py:attribute:: assistant_formatter
      :type:  Formatter


   .. py:attribute:: function_formatter
      :type:  Optional[Formatter]
      :value: None



   .. py:attribute:: observation_formatter
      :type:  Optional[Formatter]
      :value: None



   .. py:attribute:: system_formatter
      :type:  Optional[Formatter]
      :value: None



   .. py:attribute:: force_system
      :type:  bool
      :value: False



   .. py:attribute:: tools_formatter
      :type:  Optional[Formatter]
      :value: None



   .. py:attribute:: separator
      :type:  Optional[TemplateComponent]
      :value: None



   .. py:attribute:: remove_last_sep
      :type:  bool
      :value: False



   .. py:attribute:: special_starter
      :type:  Optional[TemplateComponent]
      :value: None



   .. py:attribute:: special_stopper
      :type:  Optional[TemplateComponent]
      :value: None



   .. py:attribute:: template_name
      :type:  Optional[str]
      :value: None



   .. py:attribute:: system_default
      :type:  Optional[str]
      :value: None



   .. py:method:: __post_init__()


   .. py:method:: encode_conversation(tokenizer: transformers.PreTrainedTokenizer, messages: list[dict[str, str]], system: Optional[str] = None, tools: Optional[list[str]] = None, **kwargs) -> collections.abc.Sequence[tuple[list[int], list[int]]]

      
      Messages here should be guaranteed to be in pairs, with the first message being the user message and the second message being the system message.
      Data example:
      ```json
      {
          "conversation_id": 2,
          "system": "sysinfo1",
          "tools": ["tool_1_desc"],
          "messages": [
              {
                  "role": "user",
                  "content": "hi"
              },
              {
                  "role": "assistant",
                  "content": "Hello!"
              }
          ]
      }
      ```
















      ..
          !! processed by numpydoc !!


   .. py:method:: _encode(tokenizer: transformers.PreTrainedTokenizer, messages: list[dict[str, str]], system: Optional[str] = None, tools: Optional[str] = None, **kwargs) -> collections.abc.Sequence[tuple[list[int], list[int]]]


   .. py:method:: _encode_template(template: list[TemplateComponent], tokenizer: transformers.PreTrainedTokenizer, **kwargs) -> list[int]

      
      Encode template components into token ids.


      :Parameters:

          **template** : list[TemplateComponent]
              Formatted template components.

          **tokenizer** : PreTrainedTokenizer
              Tokenizer to convert tokens into token ids.



      :Returns:

          list[int]
              Encoded token ids.











      ..
          !! processed by numpydoc !!


   .. py:method:: post_process_pairs(encoded_pairs, tokenizer)


   .. py:method:: remove_last_separator(encoded_pairs: collections.abc.Sequence[tuple[list[int], list[int]]], tokenizer: transformers.PreTrainedTokenizer) -> collections.abc.Sequence[tuple[list[int], list[int]]]


   .. py:method:: add_special_starter(encoded_pairs: collections.abc.Sequence[tuple[list[int], list[int]]], tokenizer: transformers.PreTrainedTokenizer) -> collections.abc.Sequence[tuple[list[int], list[int]]]


   .. py:method:: add_special_stopper(encoded_pairs: collections.abc.Sequence[tuple[list[int], list[int]]], tokenizer: transformers.PreTrainedTokenizer) -> collections.abc.Sequence[tuple[list[int], list[int]]]


   .. py:method:: _ensure_id_list(obj: Union[int, list[int]]) -> list[int]

      
      Make sure the object is a list of integers. Useful for handling token ids.
















      ..
          !! processed by numpydoc !!


.. py:class:: ConversationTemplateForTool

   Bases: :py:obj:`ConversationTemplate`


   .. py:method:: encode_conversation(tokenizer: transformers.PreTrainedTokenizer, messages: list[dict[str, str]], system: Optional[str] = None, tools: Optional[list[str]] = None, **kwargs) -> collections.abc.Sequence[tuple[list[int], list[int]]]

      
      Messages here should be guaranteed to be in pairs, with the first message being the user message and the second message being the system message.
      Data example:
      ```json
      {
          "conversation_id": 2,
          "system": "sysinfo1",
          "tools": ["tool_1_desc"],
          "messages": [
              {
                  "role": "user",
                  "content": "hi"
              },
              {
                  "role": "assistant",
                  "content": "Hello!"
              }
          ]
      }
      ```
















      ..
          !! processed by numpydoc !!


   .. py:method:: _encode(tokenizer: transformers.PreTrainedTokenizer, messages: list[dict[str, str]], system: Optional[str] = None, tools: Optional[str] = None, **kwargs) -> collections.abc.Sequence[tuple[list[int], list[int]]]


   .. py:method:: _encode_template(template: list[TemplateComponent], tokenizer: transformers.PreTrainedTokenizer, **kwargs) -> list[int]

      
      Encode template components into token ids.


      :Parameters:

          **template** : list[TemplateComponent]
              Formatted template components.

          **tokenizer** : PreTrainedTokenizer
              Tokenizer to convert tokens into token ids.



      :Returns:

          list[int]
              Encoded token ids.











      ..
          !! processed by numpydoc !!


   .. py:method:: _handle_tools(tools: Optional[list[str]]) -> str