Property addon#

The openvino-node package exports addon which contains the following properties:

Interface NodeAddon

export interface NodeAddon {
    Core: CoreConstructor;
    Tensor: TensorConstructor;
    PartialShape: PartialShapeConstructor;

    preprocess: {
      resizeAlgorithm: typeof resizeAlgorithm;
      PrePostProcessor: PrePostProcessorConstructor;
    };
    saveModelSync(model: Model, path: string, compressToFp16?: boolean): void;
    element: typeof element;
  }

Defined in: addon.ts:669

Properties#

Core

```
Core: CoreConstructor
```
- Type declaration:
  - CoreConstructor: CoreConstructor
- Defined in: addon.ts:670

PartialShape

```
PartialShape: PartialShapeConstructor
```
- Type declaration:
  - PartialShapeConstructor: PartialShapeConstructor
- Defined in: addon.ts:672

Tensor

```
Tensor: TensorConstructor
```
- Type declaration:
  - TensorConstructor: TensorConstructor
- Defined in: addon.ts:671

element

```
element: typeof element
```
- Type declaration:
  - element: typeof element
- Defined in: addon.ts:678

preprocess

preprocess: {
    resizeAlgorithm: typeof resizeAlgorithm;
    PrePostProcessor: PrePostProcessorConstructor;
}

Type declaration:
- resizeAlgorithm: typeof resizeAlgorithm
- PrePostProcessor: PrePostProcessorConstructor

Defined in: addon.ts:674

saveModelSync

```
saveModelSync(model: Model, path: string, compressToFp16?: boolean): void;
```
This method saves a model to IR (xml and bin files), applying all necessary transformations that are usually added during model conversion. Particularly, weights are compressed to FP16 by default, and debug information in model nodes is cleaned up.
- Parameters:
  - model: Model
    
    A model which will be converted to IR and saved.
  - path: string
    
    A path for saving the model.
  - Optional
    - compressToFp16: boolean
      
      Compression of weights to FP16 floating point precision. The default value is true .
- Returns: void
- Defined in: addon.ts:692