fhelium.backend.triton._ntt

Schedule NTT endpoints, pure expressions, and synchronized native interiors.

NttExecution

class View source

NttExecution(graph: BoundGraph, schedules: tuple[NttStages, ...], steps: tuple[_Step, ...], output_roots: tuple[ValueExpr, ...], stores: tuple[tuple[ValueExpr, ...], ...], dtype: torch.dtype, device: torch.device, table_strides: tuple[tuple[tuple[int, ...], ...], ...], complete: tuple[Callable | None, ...])

Execute dependency-selected endpoints with live tables and planned storage.

Attributes

Name	Type	Default/value
graph	BoundGraph
schedules	tuple[NttStages, ...]
steps	tuple[_Step, ...]
output_roots	tuple[ValueExpr, ...]
stores	tuple[tuple[ValueExpr, ...], ...]
dtype	torch.dtype
device	torch.device
table_strides	tuple[tuple[tuple[int, ...], ...], ...]
complete	tuple[Callable | None, ...]
execution_source	str	field(init=False, repr=False, compare=False)

build

method

def build(graph: BoundGraph, values: tuple[torch.Tensor, ...], tables: tuple[torch.Tensor, ...], *, max_endpoints: int | None=None) -> NttExecution: ...

execute

method

def execute(values: tuple[torch.Tensor, ...], tables: tuple[torch.Tensor, ...]) -> tuple[torch.Tensor, ...]: ...