Oink
Inference Inference

Inference

Applies a .ocat model from a Train node to every frame (tiled, cached).

raw .md

Applies a .ocat model from a Train node to every frame (tiled, cached). tile_size and overlap set the tiles (they blend over the overlap); mix blends the result with the input. Export writes TorchScript or ONNX.

NodeInference
MenuAI › Inference
Names in each format
.oink comp (JSON)"type": "OC_N_aiinference"
.nk scriptnone: written as an Oink node
Pythonnuke.createNode("OC_N_aiinference")
Assistant (MCP)create_node(type="Inference")
Tab searchinference

Inference properties

Inputs and outputs

PortNameType
input 0Sourceimage
output 0Imageimage

Knobs

Panel tab Aiinference, in panel order.

KnobLabelTypeDefaultRange / values.nk knobDescription
checkpointmodel filefile.ocat checkpoint written by a Train node
mixmixfloat10 to 1
tile_sizetilefloat102464 to 8192 (integer)Tile size (px); tiles blend over the overlap
overlapoverlapfloat640 to 512 (integer)

Other knobs (Python, .nk and the assistant; not on the panel)

KnobLabelTypeDefaultRange / values.nk knobDescription
gl_colorViewer Controls Colorfloat ×30, 0, 00 to 1
postage_stamp_frameStatic Frameint1-2147483648 to 2147483647
lifetimeStartLifetime Startint0-2147483648 to 2147483647
lifetimeEndLifetime Endint100-2147483648 to 2147483647
useLifetimeUse Lifetimebooleanfalse
cachedCachedbooleanfalse
bookmarkBookmarkbooleanfalse
dope_sheetDope Sheetbooleanfalse
clone_groupClone Groupstring
link_sourceLink Sourcestring
link_overridesLink Overridesstring
split_viewsSplit Viewsstring
generationGenerationint0-2147483648 to 2147483647

Train

Connect before (Source) and after (Target) frames, list the frames, press Train.

ComfyUI

Run a ComfyUI workflow on this node's input (float EXR round trip).

On this page

Inputs and outputsKnobsOther knobs (Python, .nk and the assistant; not on the panel)