dragon.ai.inference.llm_proxy.InferenceRequest

class InferenceRequest[source] 

Bases: NamedTuple

Typed request sent through the inference input queue.

Uses NamedTuple for zero overhead over a plain tuple — same memory layout, same pickle behavior — while providing named field access and catching field-ordering bugs at definition time.

__init__()

Methods

`__init__`()
`count`(value, /)	Return number of occurrences of value.
`index`(value[, start, stop])	Return first index of value.

Attributes

`continue_final_message`	Alias for field number 6
`formatted_messages`	Alias for field number 1
`messages`	Alias for field number 0
`response_queue`	Alias for field number 2
`sampling_override`	Alias for field number 5
`timestamp`	Alias for field number 3
`tools`	Alias for field number 4

messages: list : Alias for field number 0

formatted_messages: list : Alias for field number 1

response_queue: Any : Alias for field number 2

timestamp: float : Alias for field number 3

tools: list | None : Alias for field number 4

sampling_override: dict | None : Alias for field number 5

continue_final_message: bool : Alias for field number 6

__contains__(key, /): Return key in self.

__getitem__(key, /): Return self[key].

__len__(): Return len(self).

count(value, /): Return number of occurrences of value.

index(value, start=0, stop=9223372036854775807, /)

Return first index of value.

Raises ValueError if the value is not present.