1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
|
// META: global=window,dedicatedworker
// META: script=/resources/WebIDLParser.js
// META: script=/resources/idlharness.js
// META: script=./resources/utils.js
// META: timeout=long
// https://webmachinelearning.github.io/webnn/
'use strict';
idl_test(
['webnn'],
['html', 'webidl', 'webgpu'],
async (idl_array) => {
if (self.GLOBAL.isWindow()) {
idl_array.add_objects({ Navigator: ['navigator'] });
} else if (self.GLOBAL.isWorker()) {
idl_array.add_objects({ WorkerNavigator: ['navigator'] });
}
idl_array.add_objects({
NavigatorML: ['navigator'],
ML: ['navigator.ml'],
MLContext: ['context'],
MLOperand: ['input', 'filter', 'output'],
MLOperator: ['relu'],
MLGraphBuilder: ['builder'],
MLGraph: ['graph']
});
for (const executionType of ExecutionArray) {
const isSync = executionType === 'sync';
if (self.GLOBAL.isWindow() && isSync) {
continue;
}
for (const deviceType of DeviceTypeArray) {
if (isSync) {
self.context = navigator.ml.createContextSync({deviceType});
} else {
self.context = await navigator.ml.createContext({deviceType});
}
self.builder = new MLGraphBuilder(self.context);
self.input = builder.input('input', {type: 'float32', dimensions: [1, 1, 5, 5]});
self.filter = builder.constant({type: 'float32', dimensions: [1, 1, 3, 3]}, new Float32Array(9).fill(1));
self.relu = builder.relu();
self.output = builder.conv2d(input, filter, {activation: relu, inputLayout: "nchw"});
if (isSync) {
self.graph = builder.buildSync({output});
} else {
self.graph = await builder.build({output});
}
}
}
}
);
|
