Add examples for JS (#43)

* Add examples for JS

* resolve comments

Author: fs-eire

Diff for: js/README.md

@@ -25,3 +25,7 @@ Click links for README of each examples.
 * [API usage - Tensor](api-usage_tensor) - a demonstration of basic usage of `Tensor`.
 
 * [API usage - InferenceSession](api-usage_inference-session) - a demonstration of basic usage of `InferenceSession`.
+
+* [API usage - SessionOptions](api-usage_session-options) - a demonstration of how to configure creation of an `InferenceSession` instance.
+
+* [API usage - `ort.env` flags](api-usage_ort-env-flags) - a demonstration of how to configure a set of global flags.

Diff for: js/api-usage_inference-session/README.md

@@ -10,6 +10,8 @@ This example is a demonstration of basic usage of `InferenceSession`.
 
 For more information about `SessionOptions` and `RunOptions`, please refer to other examples.
 
+See also [`InferenceSession.create`](https://onnxruntime.ai/docs/api/js/interfaces/InferenceSessionFactory.html#create) and [`InferenceSession` interface](https://onnxruntime.ai/docs/api/js/interfaces/InferenceSession.html) in API reference document.
+
 ## Usage
 
 ```sh

Diff for: js/api-usage_ort-env-flags/README.md

@@ -0,0 +1,73 @@
+# 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:
+
+```js
+// enable DEBUG flag
+ort.env.debug = true;
+
+// set global logging level
+ort.env.logLevel = 'info';
+```
+
+See also [`Env` interface](https://onnxruntime.ai/docs/api/js/interfaces/Env.html) 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:
+
+```js
+// 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](https://onnxruntime.ai/docs/api/js/interfaces/Env.WebAssemblyFlags.html) 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:
+
+```js
+// enable packed texture. This helps to improve inference performance for some models
+ort.env.webgl.pack = true;
+```
+
+See also [WebGL flags](https://onnxruntime.ai/docs/api/js/interfaces/Env.WebGLFlags.html) 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`](../api-usage_session-options) 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.

Diff for: js/api-usage_session-options/README.md

@@ -0,0 +1,101 @@
+# API usage - SessionOptions
+
+## Summary
+
+This example is a demonstration of how to configure an `InferenceSession` instance using `SessionOptions`.
+
+A `SessionOptions` is an object with properties to instruct the creation of an `InferenceSession` instance. See [type declaration](https://github.com/microsoft/onnxruntime/blob/master/js/common/lib/inference-session.ts) for schema definition. `SessionOptions` is passed to `InferenceSession.create()` as the last parameter:
+
+```js
+const mySession = await InferenceSession.create(..., mySessionOptions);
+```
+
+### Execution providers
+
+An [execution provider](https://onnxruntime.ai/docs/reference/execution-providers/) (EP) defines how operators get resolved to specific kernel implementation. Following is a table of supported EP in different environments:
+
+| EP name | Hardware          | available in                      |
+| ------- | ----------------- | --------------------------------- |
+| `cpu`   | CPU (default CPU) | onnxruntime-node                  |
+| `cuda`  | GPU (NVIDIA CUDA) | onnxruntime-node                  |
+| `dml`   | GPU (Direct ML)   | onnxruntime-node (Windows)        |
+| `wasm`  | CPU (WebAssembly) | onnxruntime-web, onnxruntime-node |
+| `webgl` | GPU (WebGL)       | onnxruntime-web                   |
+
+Execution provider is specified by `sessionOptions.executionProviders`. Multiple EPs can be specified and the first available one will be used.
+
+Following are some example code snippets:
+
+```js
+// [Node.js binding example] Use CPU EP.
+const sessionOption = { executionProviders: ['cpu'] };
+```
+
+```js
+// [Node.js binding example] Use CUDA EP.
+const sessionOption = { executionProviders: ['cuda'] };
+```
+
+```js
+// [Node.js binding example] Use CUDA EP, specifying device ID.
+const sessionOption = {
+  executionProviders: [
+    {
+      name: 'cuda',
+      deviceId: 0
+    }
+  ]
+};
+```
+
+```js
+// [Node.js binding example] Try multiple EPs using an execution provider list.
+// The first successfully initialized one will be used. Use CUDA EP if it is available, otherwise fallback to CPU EP.
+const sessionOption = { executionProviders: ['cuda', 'cpu'] };
+```
+
+```js
+// [ONNX Runtime Web example] Use WebAssembly (CPU) EP.
+const sessionOption = { executionProviders: ['wasm'] };
+```
+
+```js
+// [ONNX Runtime Web example] Use WebGL EP.
+const sessionOption = { executionProviders: ['webgl'] };
+```
+
+### other common options
+
+There are also some other options available for all EPs.
+
+Following are some example code snippets:
+
+```js
+// [Node.js binding example] Use CPU EP with single-thread and enable verbose logging.
+const sessionOption = {
+  executionProviders: ['cpu'],
+  interOpNumThreads: 1,
+  intraOpNumThreads: 1,
+  logSeverityLevel: 0
+};
+```
+
+```js
+// [ONNX Runtime Web example] Use WebAssembly EP and enable profiling.
+const sessionOptions = {
+  executionProviders: ['wasm'],
+  enableProfiling: true
+};
+```
+
+See also [`SessionOptions` interface](https://onnxruntime.ai/docs/api/js/interfaces/InferenceSession.SessionOptions.html) 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 - `ort.env` flags](../api-usage_ort-env-flags) for an example of using `ort.env`.
+
+## Usage
+
+The code snippets demonstrated above cannot run standalone. Put the code into one of the "Quick Start" examples and try it out.

Diff for: js/api-usage_tensor/README.md

@@ -7,6 +7,7 @@ This example is a demonstration of basic usage of `Tensor`.
 - `tensor-create.js`: In this example, we create tensors in different ways.
 - `tensor-properties.js`: In this example, we get tensor properties from a Tensor object.
 
+See also [`Tensor` constructor](https://onnxruntime.ai/docs/api/js/interfaces/TensorConstructor.html) and [`Tensor` interface](https://onnxruntime.ai/docs/api/js/interfaces/Tensor.html) in API reference document.
 ## Usage
 
 ```sh