api-usage_ort-env-flags

API usage - ort.env flags

Summary

This example is a demonstration of how to configure global flags by ort.env.

Following are some example code snippets:

// enable DEBUG flag
ort.env.debug = true;

// set global logging level
ort.env.logLevel = 'info';

See also Env interface in API reference document.

WebAssembly flags (ONNX Runtime Web)

WebAssembly flags are used to customize behaviors of WebAssembly execution provider.

Following are some example code snippets:

// set up-to-2-threads for multi-thread execution for WebAssembly
// may fallback to single-thread if multi-thread is not available in the current browser
ort.env.wasm.numThreads = 2;

// force single-thread for WebAssembly
ort.env.wasm.numThreads = 1;

// enable worker-proxy feature for WebAssembly
// this feature allows model inferencing to run in a web worker asynchronously.
ort.env.wasm.proxy = true;

// override path of wasm files - using a prefix
// in this example, ONNX Runtime Web will try to load file from https://example.com/my-example/ort-wasm*.wasm 
ort.env.wasm.wasmPaths = 'https://example.com/my-example/';

// override path of wasm files - for each file
ort.env.wasm.wasmPaths = {
    'ort-wasm.wasm': 'https://example.com/my-example/ort-wasm.wasm',
    'ort-wasm-simd.wasm': 'https://example.com/my-example/ort-wasm-simd.wasm',
    'ort-wasm-threaded.wasm': 'https://example.com/my-example/ort-wasm-threaded.wasm',
    'ort-wasm-simd-threaded.wasm': 'https://example.com/my-example/ort-wasm-simd-threaded.wasm'
};

See also WebAssembly flags in API reference document.

WebGL flags (ONNX Runtime Web)

WebGL flags are used to customize behaviors of WebGL execution provider.

Following are some example code snippets:

// enable packed texture. This helps to improve inference performance for some models
ort.env.webgl.pack = true;

See also WebGL flags in API reference document.

WebGPU flags (ONNX Runtime Web)

WebGPU flags are used to customize behaviors of WebGPU execution provider.

Following are some example code snippets:

// enable WebGPU profiling.
ort.env.webgpu.profilingMode = 'default';

// get the gpu device object.
const device = ort.env.webgpu.device;

See also WebGPU flags in API reference document.

SessionOptions vs. ort.env

Both SessionOptions and ort.env allow to specify configurations for inferencing behaviors. The biggest difference of them is: SessionOptions is set for one inference session instance, while ort.env is set global.

See also API usage - SessionOptions for an example of using SessionOptions.

Usage

The code snippets demonstrated above cannot run standalone. Put the code into one of the "Quick Start" examples and try it out.