tensorplay.compiler.api.compile

tensorplay.compiler.api.compile(model: Callable[[...], Any] | None = None, *, fullgraph: bool = False, dynamic: bool | None = None, backend: str | Callable[[...], CompiledFn] | None = None, mode: str | None = None, options: dict[str, Any] | None = None, name: str | None = None, disable: bool = False, recompile_limit: int | None = None, isolate_recompiles: bool = False, strict_native: bool = False, dynamic_shapes: Any | None = None) → Callable[[...], Any][source]

Compile a callable through the TensorPlay compiler frontend.

backend may be a registered name or a callable with the contract backend(graph_module, example_inputs, **kwargs) -> callable. The frontend caches specializations by input metadata and falls back to the original callable when a non-fullgraph capture cannot represent the Python region.