Inference module (Inference Overlay): Difference between revisions
Jump to navigation
Jump to search
No edit summary |
No edit summary |
||
| Line 1: | Line 1: | ||
The Inference Module of the [[Inference Overlay]] is responsible for applying the convolution network on subsections of [[Prequels (Inference Overlay)|prequel grid]]s configured for the [[Inference Overlay]]. The size of the subsection is defined by the [[input tensor]]s of the neural network. The marching speed of this subsection is defined by the [[Stride fraction (Inference Overlay)|STRIDE_FRACTION]] attribute. | The Inference Module of the [[Inference Overlay]] is responsible for applying the convolution network on subsections of [[Prequels (Inference Overlay)|prequel grid]]s configured for the [[Inference Overlay]]. The size of the subsection is defined by the [[Input tensor (Inference Overlay)|input tensor]]s of the neural network. The marching speed of this subsection is defined by the [[Stride fraction (Inference Overlay)|STRIDE_FRACTION]] attribute. | ||
{{InferenceOverlay nav}} | {{InferenceOverlay nav}} | ||
Revision as of 12:34, 9 October 2024
The Inference Module of the Inference Overlay is responsible for applying the convolution network on subsections of prequel grids configured for the Inference Overlay. The size of the subsection is defined by the input tensors of the neural network. The marching speed of this subsection is defined by the STRIDE_FRACTION attribute.
