For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).
Mojo struct
ImageData
struct ImageData[LayoutType: TensorLayout, //, dtype: DType, static_image_layout: Image2DLayout, origin: MutOrigin]
Utility class that generalizes conv2d data and filter tensor with a given data layout.
Fields
- data (
TileTensor[dtype, LayoutType, origin]): - dynamic_image_layout (
Image2DLayout):
Implemented traits
AnyType,
Copyable,
ImplicitlyCopyable,
ImplicitlyDeletable,
Movable,
RegisterPassable,
TrivialRegisterPassable
Methods
__init__
def __init__(data: TileTensor[dtype, LayoutType, origin], _layout: Image2DLayout) -> Self
Construct of an image data instance with dynamic layout param.
Args:
- data (
TileTensor[dtype, LayoutType, origin]): A 4d buffer containing the actual data. - _layout (
Image2DLayout): Data layout tag.
def __init__(data: TileTensor[dtype, LayoutType, origin]) -> Self
__getitem__
def __getitem__(self, n: Int, c: Int, h: Int, w: Int) -> Scalar[dtype]
Reads the underlying data buffer based on the tensor index and under- lying data layout.
Args:
- n (
Int): Index on the batch dimension. - c (
Int): Index on the channel dimension. - h (
Int): Index on the height dimension. - w (
Int): Index on the width dimension.
Returns:
Scalar[dtype]: The value stored at the given index position.
__setitem__
def __setitem__(self, n: Int, c: Int, h: Int, w: Int, value: Scalar[dtype])
Writes the underlying data buffer based on the tensor index and under- lying data layout.
Args:
- n (
Int): Index on the batch dimension. - c (
Int): Index on the channel dimension. - h (
Int): Index on the height dimension. - w (
Int): Index on the width dimension. - value (
Scalar[dtype]): The value to store at the given index position.
to_static_layout
def to_static_layout[new_static_image_layout: Image2DLayout](self) -> ImageData[dtype, new_static_image_layout, origin]
Conversion utility from a fully dynamic data structure, e.g. from c shim to one with compile-time known data layout.
Returns:
ImageData[dtype, new_static_image_layout, origin]: The image data with static data layout.
get_image_layout
def get_image_layout(self) -> Image2DLayout
The getter function of the underlying data layout, resolving from either statically or dynamically provided information.
Returns:
Image2DLayout: The resolved data layout tag for this image instance.
get_flat_index
def get_flat_index(self, n: Int, c: Int, h: Int, w: Int) -> Int
Converts the dimension index to the flat index of the underlying data based on the tensor layout.
Args:
- n (
Int): Index on the batch dimension. - c (
Int): Index on the channel dimension. - h (
Int): Index on the height dimension. - w (
Int): Index on the width dimension.
Returns:
Int: An integer containing the index based on the underlying
data layout.
get_tuple_index
def get_tuple_index(self, idx: Int) -> IndexList[Int(4)]
Converts the flat index to the dimension index of the underlying data based on the tensor layout.
Args:
- idx (
Int): Flat index.
Returns:
IndexList[Int(4)]: A IndexList containing the index in NCHW order.
num_elements
Was this page helpful?
Thank you! We'll create more content like this.
Thank you for helping us improve!