Class CompletionOutput

fastdeploy/engine/request.py:702–798

The output data of one completion output of a request. Args: index: The index of the output in the request. text: The generated output text. token_ids: The token IDs of the generated output text.

@dataclass(slots=True)
class CompletionOutput:
    """The output data of one completion output of a request.

    Args:
        index: The index of the output in the request.
        text: The generated output text.
        token_ids: The token IDs of the generated output text.
    """

    index: int
    send_idx: int
    token_ids: list[Any]
    decode_type: int = 0
    logprob: Optional[float] = None
    top_logprobs: Optional[LogprobsLists] = None
    draft_top_logprobs: Optional[LogprobsLists] = None
    logprobs: Optional[SampleLogprobs] = None
    draft_token_ids: list[int] = None
    text: Optional[str] = None
    reasoning_content: Optional[str] = None
    reasoning_token_num: Optional[int] = 0
    tool_calls: Optional[ToolCall] = None
    speculate_metrics: Optional[SpeculateMetrics] = None
    completion_tokens: Optional[str] = None
    delta_message: Optional[DeltaMessage] = None
    multipart: Optional[list[Any]] = None
    num_image_tokens: Optional[int] = None
    enable_parser: bool = False
    # Sparse indices of retained vocab ids:
    #   - Non-MTP: list[int]
    #   - MTP: list[list[int]]
    sampling_mask: Optional[Any] = None

    def to_dict(self):
        """
        convert CompletionOutput to a serialized dict
        """
        return {
            "index": self.index,
            "send_idx": self.send_idx,
            "token_ids": self.token_ids,
            "decode_type": self.decode_type,
            "logprob": self.logprob,
            "top_logprobs": self.top_logprobs,
            "draft_top_logprobs": self.draft_top_logprobs,
            "logprobs": self.logprobs,
            "draft_token_ids": self.draft_token_ids,
            "text": self.text,
            "reasoning_content": self.reasoning_content,
            "reasoning_token_num": self.reasoning_token_num,
            "sampling_mask": self.sampling_mask,
        }

    @classmethod
    def from_dict(cls, req_dict: dict[str, Any]) -> CompletionOutput:
        """Create instance from dict arguments"""
        return cls(
            **{