IMPORTANT: To view this page as Markdown, append `.md` to the URL (e.g. /get-started.md). For the complete documentation index, see llms.txt.
Skip to main content
For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).

Mojo struct

ConvTransposeGPU

struct ConvTransposeGPU

Registers the mo.conv_transpose graph op for GPU devices.

Two execute overloads under one registration: the graph compiler calls the fused one when any output has a fused epilogue and the plain one when none has. cuDNN writes a whole output buffer, so the fused form has it write scratch and stores from there into the fused output, while the plain form has cuDNN write the output directly. The pair is grandfathered on the graph compiler's allow list; new kernels declare one execute with a StoreAccess output.

Implemented traitsโ€‹

AnyType, Deinitable, Movable

Methodsโ€‹

executeโ€‹

static def execute[input_layout: StaticString, filter_layout: StaticString](output: T, input: T, filter: T, strides: T, dilation: T, paddings: T, output_paddings: T, ctx: DeviceContext)

static def execute[input_layout: StaticString, filter_layout: StaticString](output: T, input: T, filter: T, strides: T, dilation: T, paddings: T, output_paddings: T, ctx: DeviceContext) where not conforms_to(T, Fused).__bool__()

Was this page helpful?