[js/common] fix typedoc warnings (#19933)

### Description
Fix a few warnings in typedoc (for generating JS API):
```
[warning] The signature TrainingSession.loadParametersBuffer has an @param with name "buffer", which was not used.
[warning] NonTensorType, defined in ./lib/onnx-value.ts, is referenced by OnnxValue but not included in the documentation.
[warning] TensorFactory, defined in ./lib/tensor-factory.ts, is referenced by Tensor but not included in the documentation.
[warning] ExternalDataFileType, defined in ./lib/onnx-model.ts, is referenced by InferenceSession.SessionOptions.externalData but not included in the documentation.
[warning] TensorToDataUrlOptions, defined in ./lib/tensor-conversion.ts, is referenced by Tensor.toDataURL.toDataURL.options but not included in the documentation.
[warning] TensorToImageDataOptions, defined in ./lib/tensor-conversion.ts, is referenced by Tensor.toImageData.toImageData.options but not included in the documentation.
[warning] Failed to resolve link to "GpuBufferType" in comment for Env.WebGpuFlags.adapter.
[warning] Failed to resolve link to "GpuBufferType" in comment for Env.WebGpuFlags.device.
```

Changes highlighted:
- Merge `CoreMlExecutionProviderOption` and
`CoreMLExecutionProviderOption`. They expose 2 set of different options
for React-native and ORT nodejs binding. This should be fixed in future.
- Fix a few inconsistency of names between JSDoc and parameters
- Fix broken type links
- Exclude trace functions

js/common/lib/backend.ts

@@ -58,7 +58,7 @@ export interface TrainingSessionHandler extends SessionHandler {
       options: InferenceSession.RunOptions): Promise<SessionHandler.ReturnType>;
 
   getParametersSize(trainableOnly: boolean): Promise<number>;
-  loadParametersBuffer(array: Uint8Array, trainableOnly: boolean): Promise<void>;
+  loadParametersBuffer(buffer: Uint8Array, trainableOnly: boolean): Promise<void>;
   getContiguousParameters(trainableOnly: boolean): Promise<OnnxValue>;
 }
 
@@ -77,8 +77,8 @@ export interface Backend {
       Promise<InferenceSessionHandler>;
 
   createTrainingSessionHandler?
-      (checkpointStateUriOrBuffer: TrainingSession.URIorBuffer, trainModelUriOrBuffer: TrainingSession.URIorBuffer,
-       evalModelUriOrBuffer: TrainingSession.URIorBuffer, optimizerModelUriOrBuffer: TrainingSession.URIorBuffer,
+      (checkpointStateUriOrBuffer: TrainingSession.UriOrBuffer, trainModelUriOrBuffer: TrainingSession.UriOrBuffer,
+       evalModelUriOrBuffer: TrainingSession.UriOrBuffer, optimizerModelUriOrBuffer: TrainingSession.UriOrBuffer,
        options: InferenceSession.SessionOptions): Promise<TrainingSessionHandler>;
 }
 

js/common/lib/env.ts

@@ -173,7 +173,7 @@ export declare namespace Env {
      * When use with TypeScript, the type of this property is `GPUAdapter` defined in "@webgpu/types".
      * Use `const adapter = env.webgpu.adapter as GPUAdapter;` in TypeScript to access this property with correct type.
      *
-     * see comments on {@link GpuBufferType}
+     * see comments on {@link Tensor.GpuBufferType}
      */
     readonly adapter: unknown;
     /**
@@ -184,7 +184,7 @@ export declare namespace Env {
      * When use with TypeScript, the type of this property is `GPUDevice` defined in "@webgpu/types".
      * Use `const device = env.webgpu.device as GPUDevice;` in TypeScript to access this property with correct type.
      *
-     * see comments on {@link GpuBufferType} for more details about why not use types defined in "@webgpu/types".
+     * see comments on {@link Tensor.GpuBufferType} for more details about why not use types defined in "@webgpu/types".
      */
     readonly device: unknown;
     /**

js/common/lib/index.ts

@@ -21,6 +21,9 @@ export * from './backend.js';
 export * from './env.js';
 export * from './inference-session.js';
 export * from './tensor.js';
+export * from './tensor-conversion.js';
+export * from './tensor-factory.js';
 export * from './trace.js';
+export * from './onnx-model.js';
 export * from './onnx-value.js';
 export * from './training-session.js';

js/common/lib/inference-session.ts

@@ -186,22 +186,22 @@ export declare namespace InferenceSession {
   // #region execution providers
 
   // Currently, we have the following backends to support execution providers:
-  // Backend Node.js binding: supports 'cpu' and 'cuda'.
+  // Backend Node.js binding: supports 'cpu', 'dml' (win32), 'coreml' (macOS) and 'cuda' (linux).
   // Backend WebAssembly: supports 'cpu', 'wasm', 'webgpu' and 'webnn'.
   // Backend ONNX.js: supports 'webgl'.
   // Backend React Native: supports 'cpu', 'xnnpack', 'coreml' (iOS), 'nnapi' (Android).
   interface ExecutionProviderOptionMap {
+    coreml: CoreMLExecutionProviderOption;
     cpu: CpuExecutionProviderOption;
-    coreml: CoreMlExecutionProviderOption;
     cuda: CudaExecutionProviderOption;
     dml: DmlExecutionProviderOption;
+    nnapi: NnapiExecutionProviderOption;
     tensorrt: TensorRtExecutionProviderOption;
     wasm: WebAssemblyExecutionProviderOption;
     webgl: WebGLExecutionProviderOption;
-    xnnpack: XnnpackExecutionProviderOption;
     webgpu: WebGpuExecutionProviderOption;
     webnn: WebNNExecutionProviderOption;
-    nnapi: NnapiExecutionProviderOption;
+    xnnpack: XnnpackExecutionProviderOption;
   }
 
   type ExecutionProviderName = keyof ExecutionProviderOptionMap;
@@ -219,10 +219,6 @@ export declare namespace InferenceSession {
     readonly name: 'cuda';
     deviceId?: number;
   }
-  export interface CoreMlExecutionProviderOption extends ExecutionProviderOption {
-    readonly name: 'coreml';
-    coreMlFlags?: number;
-  }
   export interface DmlExecutionProviderOption extends ExecutionProviderOption {
     readonly name: 'dml';
     deviceId?: number;
@@ -253,8 +249,39 @@ export declare namespace InferenceSession {
   }
   export interface CoreMLExecutionProviderOption extends ExecutionProviderOption {
     readonly name: 'coreml';
+    /**
+     * The bit flags for CoreML execution provider.
+     *
+     * ```
+     * COREML_FLAG_USE_CPU_ONLY = 0x001
+     * COREML_FLAG_ENABLE_ON_SUBGRAPH = 0x002
+     * COREML_FLAG_ONLY_ENABLE_DEVICE_WITH_ANE = 0x004
+     * COREML_FLAG_ONLY_ALLOW_STATIC_INPUT_SHAPES = 0x008
+     * COREML_FLAG_CREATE_MLPROGRAM = 0x010
+     * ```
+     *
+     * See include/onnxruntime/core/providers/coreml/coreml_provider_factory.h for more details.
+     *
+     * This flag is available only in ONNXRuntime (Node.js binding).
+     */
+    coreMlFlags?: number;
+    /**
+     * Specify whether to use CPU only in CoreML EP.
+     *
+     * This setting is available only in ONNXRuntime (react-native).
+     */
     useCPUOnly?: boolean;
+    /**
+     * Specify whether to enable CoreML EP on subgraph.
+     *
+     * This setting is available only in ONNXRuntime (react-native).
+     */
     enableOnSubgraph?: boolean;
+    /**
+     * Specify whether to only enable CoreML EP for Apple devices with ANE (Apple Neural Engine).
+     *
+     * This setting is available only in ONNXRuntime (react-native).
+     */
     onlyEnableDeviceWithANE?: boolean;
   }
   export interface NnapiExecutionProviderOption extends ExecutionProviderOption {

js/common/lib/onnx-value.ts

@@ -3,7 +3,7 @@
 
 import {Tensor} from './tensor.js';
 
-type NonTensorType = never;
+export type NonTensorType = never;
 
 /**
  * Type OnnxValue Represents both tensors and non-tensors value for model's inputs/outputs.

js/common/lib/tensor-factory.ts

@@ -253,7 +253,7 @@ export interface TensorFactory {
   /**
    * create a tensor from an ImageBitmap object
    *
-   * @param bitMap - the ImageBitmap object to create tensor from
+   * @param bitmap - the ImageBitmap object to create tensor from
    * @param options - An optional object representing options for creating tensor from URL.
    *
    * The following default settings will be applied:

js/common/lib/tensor.ts

@@ -160,7 +160,7 @@ export interface Tensor extends TypedTensorBase<Tensor.Type>, TypedTensorUtils<T
 /**
  * type TensorConstructor defines the constructors of 'Tensor' to create CPU tensor instances.
  */
-export interface TensorConstructor {
+export interface TensorConstructor extends TensorFactory {
   // #region CPU tensor - specify element type
   /**
    * Construct a new string tensor object from the given type, data and dims.
@@ -326,4 +326,4 @@ export interface TensorConstructor {
 }
 
 // eslint-disable-next-line @typescript-eslint/naming-convention
-export const Tensor = TensorImpl as (TensorConstructor & TensorFactory);
+export const Tensor = TensorImpl as TensorConstructor;

js/common/lib/trace.ts

@@ -3,6 +3,9 @@
 
 import {env} from './env-impl.js';
 
+/**
+ * @ignore
+ */
 export const TRACE = (deviceType: string, label: string) => {
   if (typeof env.trace === 'undefined' ? !env.wasm.trace : !env.trace) {
     return;
@@ -29,13 +32,19 @@ const TRACE_FUNC = (msg: string, extraMsg?: string) => {
   }
 };
 
+/**
+ * @ignore
+ */
 export const TRACE_FUNC_BEGIN = (extraMsg?: string) => {
   if (typeof env.trace === 'undefined' ? !env.wasm.trace : !env.trace) {
     return;
   }
   TRACE_FUNC('BEGIN', extraMsg);
 };
 
+/**
+ * @ignore
+ */
 export const TRACE_FUNC_END = (extraMsg?: string) => {
   if (typeof env.trace === 'undefined' ? !env.wasm.trace : !env.trace) {
     return;

js/common/lib/training-session.ts

@@ -11,7 +11,7 @@ export declare namespace TrainingSession {
   /**
    * Either URI file path (string) or Uint8Array containing model or checkpoint information.
    */
-  type URIorBuffer = string|Uint8Array;
+  type UriOrBuffer = string|Uint8Array;
 }
 
 /**
@@ -98,13 +98,13 @@ export interface TrainingSession {
   getParametersSize(trainableOnly: boolean): Promise<number>;
 
   /**
-   * Copies parameter values from the given array to the training state. Currently, only supporting models with
+   * Copies parameter values from the given buffer to the training state. Currently, only supporting models with
    * parameters of type Float32.
    *
-   * @param buffer - Float32 buffer containing parameters converted to a Uint8Array.
+   * @param buffer - A Uint8Array representation of Float32 parameters.
    * @param trainableOnly - True if trainable parameters only to be modified, false otherwise. Default value is true.
    */
-  loadParametersBuffer(array: Uint8Array, trainableOnly: boolean): Promise<void>;
+  loadParametersBuffer(buffer: Uint8Array, trainableOnly: boolean): Promise<void>;
 
   /**
    * Copies the model parameters to a contiguous buffer. Usually used in the context of Federated Learning.
@@ -157,19 +157,19 @@ export interface TrainingSessionCreateOptions {
   /**
    * URI or buffer for a .ckpt file that contains the checkpoint for the training model.
    */
-  checkpointState: TrainingSession.URIorBuffer;
+  checkpointState: TrainingSession.UriOrBuffer;
   /**
    * URI or buffer for the .onnx training file.
    */
-  trainModel: TrainingSession.URIorBuffer;
+  trainModel: TrainingSession.UriOrBuffer;
   /**
    * Optional. URI or buffer for the .onnx optimizer model file.
    */
-  optimizerModel?: TrainingSession.URIorBuffer;
+  optimizerModel?: TrainingSession.UriOrBuffer;
   /**
    * Optional. URI or buffer for the .onnx eval model file.
    */
-  evalModel?: TrainingSession.URIorBuffer;
+  evalModel?: TrainingSession.UriOrBuffer;
 }
 
 /**

tools/ci_build/github/azure-pipelines/templates/linux-web-init-and-check.yml

@@ -31,6 +31,10 @@ steps:
     node -e "a=require('child_process').execSync('git diff --name-only').toString();if(a)throw new Error('Following source files are not formatted: (did you run \"npm run format\"?)\n'+a)"
   workingDirectory: '$(Build.SourcesDirectory)/js'
   displayName: 'Check unformatted files'
+- script: |
+    npx typedoc --emit none --treatWarningsAsErrors
+  workingDirectory: '$(Build.SourcesDirectory)/js/common'
+  displayName: 'TypeDoc Validation'
 - script: |
     npm run build:doc
   workingDirectory: '$(Build.SourcesDirectory)/js/web'