Python API

TVM FFI Python package.

Object

Object	Base class of all TVM FFI objects.
register_object([type_key])	Register object type.

Function and Module

Function	Python class that wraps a function with tvm-ffi ABI.
Module	Module container for dynamically loaded Module.
register_global_func(func_name[, f, override])	Register global function.
get_global_func()	Get a global function by name.
system_lib([symbol_prefix])	Get system-wide library module singleton.
load_module(path)	Load module from file.
init_ffi_api(namespace[, target_module_name])	Initialize register ffi api functions into a given module.
register_error([name_or_cls, cls])	Register an error class so it can be recognized by the ffi error handler.
convert(value)	Convert a python object to ffi values.

Tensor

Shape(content)	Shape tuple that represents ffi::Shape returned by an FFI call.
Tensor	Tensor object that represents a managed n-dimensional array.
Device	Device represents a device in the ffi system.
from_dlpack(ext_tensor, *[, ...])	Convert an external tensor to an Tensor.

Containers

Array(input_list)	Array container that represents a sequence of values in ffi.
Map(input_dict)	Map container.

Stream Context

StreamContext(device, stream)	Represent a stream context in the FFI system.
use_torch_stream([context])	Create an FFI stream context with a Torch stream or graph.
use_raw_stream(device, stream)	Create a ffi stream context with given device and stream handle.

Utility

C++ integration helpers for building and loading inline modules.

cpp.load_inline(name, *[, cpp_sources, ...])	Compile, build and load a C++/CUDA module from inline source code.
cpp.build_inline(name, *[, cpp_sources, ...])	Compile and build a C++/CUDA module from inline source code.