426 lines
15 KiB
JavaScript
426 lines
15 KiB
JavaScript
// META: title=validation tests for WebNN API lstm operation
|
|
// META: global=window
|
|
// META: variant=?cpu
|
|
// META: variant=?gpu
|
|
// META: variant=?npu
|
|
// META: script=../resources/utils_validation.js
|
|
|
|
'use strict';
|
|
|
|
const steps = 10, batchSize = 5, inputSize = 3, hiddenSize = 8,
|
|
numDirections = 1;
|
|
|
|
// Dimensions required of required inputs.
|
|
const kValidInputShape = [steps, batchSize, inputSize];
|
|
const kValidWeightShape = [numDirections, 4 * hiddenSize, inputSize];
|
|
const kValidRecurrentWeightShape = [numDirections, 4 * hiddenSize, hiddenSize];
|
|
// Dimensions required of optional inputs.
|
|
const kValidBiasShape = [numDirections, 4 * hiddenSize];
|
|
const kValidPeepholeWeightShape = [numDirections, 3 * hiddenSize];
|
|
const kValidInitialHiddenStateShape = [numDirections, batchSize, hiddenSize];
|
|
|
|
// Example descriptors which are valid according to the above dimensions.
|
|
const kExampleInputDescriptor = {
|
|
dataType: 'float32',
|
|
shape: kValidInputShape
|
|
};
|
|
const kExampleWeightDescriptor = {
|
|
dataType: 'float32',
|
|
shape: kValidWeightShape
|
|
};
|
|
const kExampleRecurrentWeightDescriptor = {
|
|
dataType: 'float32',
|
|
shape: kValidRecurrentWeightShape
|
|
};
|
|
const kExampleBiasDescriptor = {
|
|
dataType: 'float32',
|
|
shape: kValidBiasShape
|
|
};
|
|
const kExamplePeepholeWeightDescriptor = {
|
|
dataType: 'float32',
|
|
shape: kValidPeepholeWeightShape
|
|
};
|
|
const kExampleInitialHiddenStateDescriptor = {
|
|
dataType: 'float32',
|
|
shape: kValidInitialHiddenStateShape
|
|
};
|
|
|
|
const tests = [
|
|
{
|
|
name: '[lstm] Test with default options',
|
|
input: {dataType: 'float16', shape: kValidInputShape},
|
|
weight: {dataType: 'float16', shape: kValidWeightShape},
|
|
recurrentWeight: {dataType: 'float16', shape: kValidRecurrentWeightShape},
|
|
steps: steps,
|
|
hiddenSize: hiddenSize,
|
|
outputs: [
|
|
{dataType: 'float16', shape: [numDirections, batchSize, hiddenSize]},
|
|
{dataType: 'float16', shape: [numDirections, batchSize, hiddenSize]}
|
|
]
|
|
},
|
|
{
|
|
name: '[lstm] Test with given options',
|
|
input: kExampleInputDescriptor,
|
|
weight: {
|
|
dataType: 'float32',
|
|
shape: [/*numDirections=*/ 2, 4 * hiddenSize, inputSize]
|
|
},
|
|
recurrentWeight: {
|
|
dataType: 'float32',
|
|
shape: [/*numDirections=*/ 2, 4 * hiddenSize, hiddenSize]
|
|
},
|
|
steps: steps,
|
|
hiddenSize: hiddenSize,
|
|
options: {
|
|
bias:
|
|
{dataType: 'float32', shape: [/*numDirections=*/ 2, 4 * hiddenSize]},
|
|
recurrentBias:
|
|
{dataType: 'float32', shape: [/*numDirections=*/ 2, 4 * hiddenSize]},
|
|
peepholeWeight:
|
|
{dataType: 'float32', shape: [/*numDirections=*/ 2, 3 * hiddenSize]},
|
|
initialHiddenState: {
|
|
dataType: 'float32',
|
|
shape: [/*numDirections=*/ 2, batchSize, hiddenSize]
|
|
},
|
|
initialCellState: {
|
|
dataType: 'float32',
|
|
shape: [/*numDirections=*/ 2, batchSize, hiddenSize]
|
|
},
|
|
returnSequence: true,
|
|
direction: 'both',
|
|
layout: 'ifgo',
|
|
activations: ['sigmoid', 'relu', 'tanh']
|
|
},
|
|
outputs: [
|
|
{
|
|
dataType: 'float32',
|
|
shape: [/*numDirections=*/ 2, batchSize, hiddenSize]
|
|
},
|
|
{
|
|
dataType: 'float32',
|
|
shape: [/*numDirections=*/ 2, batchSize, hiddenSize]
|
|
},
|
|
{
|
|
dataType: 'float32',
|
|
shape: [steps, /*numDirections=*/ 2, batchSize, hiddenSize]
|
|
}
|
|
]
|
|
},
|
|
{
|
|
name: '[lstm] TypeError is expected if hiddenSize equals to zero',
|
|
input: kExampleInputDescriptor,
|
|
weight: kExampleWeightDescriptor,
|
|
recurrentWeight: kExampleRecurrentWeightDescriptor,
|
|
steps: steps,
|
|
hiddenSize: 0
|
|
},
|
|
{
|
|
name: '[lstm] TypeError is expected if hiddenSize is too large',
|
|
input: kExampleInputDescriptor,
|
|
weight: kExampleWeightDescriptor,
|
|
recurrentWeight: kExampleRecurrentWeightDescriptor,
|
|
steps: steps,
|
|
hiddenSize: 4294967295,
|
|
},
|
|
{
|
|
name: '[lstm] TypeError is expected if steps equals to zero',
|
|
input: kExampleInputDescriptor,
|
|
weight: kExampleWeightDescriptor,
|
|
recurrentWeight: kExampleRecurrentWeightDescriptor,
|
|
steps: 0,
|
|
hiddenSize: hiddenSize,
|
|
},
|
|
{
|
|
name:
|
|
'[lstm] TypeError is expected if the data type is not one of the floating point types',
|
|
input: {dataType: 'uint32', shape: kValidInputShape},
|
|
weight: {dataType: 'uint32', shape: kValidWeightShape},
|
|
recurrentWeight: {dataType: 'uint32', shape: kValidRecurrentWeightShape},
|
|
steps: steps,
|
|
hiddenSize: hiddenSize
|
|
},
|
|
{
|
|
name: '[lstm] TypeError is expected if the rank of input is not 3',
|
|
input: {dataType: 'float32', shape: [steps, batchSize]},
|
|
weight: kExampleWeightDescriptor,
|
|
recurrentWeight: kExampleRecurrentWeightDescriptor,
|
|
steps: steps,
|
|
hiddenSize: hiddenSize
|
|
},
|
|
{
|
|
name:
|
|
'[lstm] TypeError is expected if input.shape[0] is not equal to steps',
|
|
input: {dataType: 'float32', shape: [1000, batchSize, inputSize]},
|
|
weight: kExampleWeightDescriptor,
|
|
recurrentWeight: kExampleRecurrentWeightDescriptor,
|
|
steps: steps,
|
|
hiddenSize: hiddenSize
|
|
},
|
|
{
|
|
name: '[lstm] TypeError is expected if the shape of weight is incorrect',
|
|
input: kExampleInputDescriptor,
|
|
weight: {dataType: 'float32', shape: [numDirections, 4 * hiddenSize, 1000]},
|
|
recurrentWeight: kExampleRecurrentWeightDescriptor,
|
|
steps: steps,
|
|
hiddenSize: hiddenSize
|
|
},
|
|
{
|
|
name:
|
|
'[lstm] TypeError is expected if the rank of recurrentWeight is not 3',
|
|
input: kExampleInputDescriptor,
|
|
weight: kExampleWeightDescriptor,
|
|
recurrentWeight:
|
|
{dataType: 'float32', shape: [numDirections, 4 * hiddenSize]},
|
|
steps: steps,
|
|
hiddenSize: hiddenSize
|
|
},
|
|
{
|
|
name:
|
|
'[lstm] TypeError is expected if the size of options.activations is not 3',
|
|
input: kExampleInputDescriptor,
|
|
weight: kExampleWeightDescriptor,
|
|
recurrentWeight: kExampleRecurrentWeightDescriptor,
|
|
steps: steps,
|
|
hiddenSize: hiddenSize,
|
|
options: {activations: ['sigmoid', 'tanh']}
|
|
},
|
|
{
|
|
name: '[lstm] TypeError is expected if the rank of options.bias is not 2',
|
|
input: {dataType: 'float16', shape: kValidInputShape},
|
|
weight: {dataType: 'float16', shape: kValidWeightShape},
|
|
recurrentWeight: {dataType: 'float16', shape: kValidRecurrentWeightShape},
|
|
steps: steps,
|
|
hiddenSize: hiddenSize,
|
|
options: {bias: {dataType: 'float16', shape: [numDirections]}}
|
|
},
|
|
{
|
|
name:
|
|
'[lstm] TypeError is expected if the shape of options.recurrentBias.shape is incorrect',
|
|
input: {dataType: 'float16', shape: kValidInputShape},
|
|
weight: {dataType: 'float16', shape: kValidWeightShape},
|
|
recurrentWeight: {dataType: 'float16', shape: kValidRecurrentWeightShape},
|
|
steps: steps,
|
|
hiddenSize: hiddenSize,
|
|
options:
|
|
{recurrentBias: {dataType: 'float16', shape: [numDirections, 1000]}}
|
|
},
|
|
{
|
|
name:
|
|
'[lstm] TypeError is expected if the dataType of options.peepholeWeight is incorrect',
|
|
input: {dataType: 'float16', shape: kValidInputShape},
|
|
weight: {dataType: 'float16', shape: kValidWeightShape},
|
|
recurrentWeight: {dataType: 'float16', shape: kValidRecurrentWeightShape},
|
|
steps: steps,
|
|
hiddenSize: hiddenSize,
|
|
options: {
|
|
peepholeWeight:
|
|
{dataType: 'float32', shape: [numDirections, 3 * hiddenSize]}
|
|
}
|
|
},
|
|
{
|
|
name:
|
|
'[lstm] TypeError is expected if the dataType of options.initialHiddenState is incorrect',
|
|
input: {dataType: 'float16', shape: kValidInputShape},
|
|
weight: {dataType: 'float16', shape: kValidWeightShape},
|
|
recurrentWeight: {dataType: 'float16', shape: kValidRecurrentWeightShape},
|
|
steps: steps,
|
|
hiddenSize: hiddenSize,
|
|
options: {
|
|
initialHiddenState:
|
|
{dataType: 'uint64', shape: [numDirections, batchSize, hiddenSize]}
|
|
}
|
|
},
|
|
{
|
|
name:
|
|
'[lstm] TypeError is expected if the shape of options.initialCellState is incorrect',
|
|
input: kExampleInputDescriptor,
|
|
weight: kExampleWeightDescriptor,
|
|
recurrentWeight: kExampleRecurrentWeightDescriptor,
|
|
steps: steps,
|
|
hiddenSize: hiddenSize,
|
|
options: {
|
|
initialCellState:
|
|
{dataType: 'float32', shape: [numDirections, batchSize, 1000]}
|
|
}
|
|
}
|
|
];
|
|
|
|
tests.forEach(
|
|
test => promise_test(async t => {
|
|
const builder = new MLGraphBuilder(context);
|
|
const input = builder.input('input', test.input);
|
|
const weight = builder.input('weight', test.weight);
|
|
const recurrentWeight =
|
|
builder.input('recurrentWeight', test.recurrentWeight);
|
|
|
|
const options = {};
|
|
if (test.options) {
|
|
if (test.options.bias) {
|
|
options.bias = builder.input('bias', test.options.bias);
|
|
}
|
|
if (test.options.recurrentBias) {
|
|
options.recurrentBias =
|
|
builder.input('recurrentBias', test.options.recurrentBias);
|
|
}
|
|
if (test.options.peepholeWeight) {
|
|
options.peepholeWeight =
|
|
builder.input('peepholeWeight', test.options.peepholeWeight);
|
|
}
|
|
if (test.options.initialHiddenState) {
|
|
options.initialHiddenState = builder.input(
|
|
'initialHiddenState', test.options.initialHiddenState);
|
|
}
|
|
if (test.options.initialCellState) {
|
|
options.initialCellState =
|
|
builder.input('initialCellState', test.options.initialCellState);
|
|
}
|
|
if (test.options.returnSequence) {
|
|
options.returnSequence = test.options.returnSequence;
|
|
}
|
|
if (test.options.direction) {
|
|
options.direction = test.options.direction;
|
|
}
|
|
if (test.options.layout) {
|
|
options.layout = test.options.layout;
|
|
}
|
|
if (test.options.activations) {
|
|
options.activations = test.options.activations;
|
|
}
|
|
}
|
|
|
|
if (test.outputs &&
|
|
context.opSupportLimits().lstm.input.dataTypes.includes(
|
|
test.input.dataType)) {
|
|
const outputs = builder.lstm(
|
|
input, weight, recurrentWeight, test.steps, test.hiddenSize,
|
|
options);
|
|
assert_equals(outputs.length, test.outputs.length);
|
|
for (let i = 0; i < outputs.length; ++i) {
|
|
assert_equals(outputs[i].dataType, test.outputs[i].dataType);
|
|
assert_array_equals(outputs[i].shape, test.outputs[i].shape);
|
|
}
|
|
} else {
|
|
const label = 'lstm_xxx';
|
|
options.label = label;
|
|
const regrexp = new RegExp('\\[' + label + '\\]');
|
|
assert_throws_with_label(
|
|
() => builder.lstm(
|
|
input, weight, recurrentWeight, test.steps, test.hiddenSize,
|
|
options),
|
|
regrexp);
|
|
}
|
|
}, test.name));
|
|
|
|
multi_builder_test(async (t, builder, otherBuilder) => {
|
|
const inputFromOtherBuilder =
|
|
otherBuilder.input('input', kExampleInputDescriptor);
|
|
const weight = builder.input('weight', kExampleWeightDescriptor);
|
|
const recurrentWeight =
|
|
builder.input('recurrentWeight', kExampleRecurrentWeightDescriptor);
|
|
|
|
assert_throws_js(
|
|
TypeError,
|
|
() => builder.lstm(
|
|
inputFromOtherBuilder, weight, recurrentWeight, steps, hiddenSize));
|
|
}, '[lstm] throw if input is from another builder');
|
|
|
|
multi_builder_test(async (t, builder, otherBuilder) => {
|
|
const input = builder.input('input', kExampleInputDescriptor);
|
|
const weightFromOtherBuilder =
|
|
otherBuilder.input('weight', kExampleWeightDescriptor);
|
|
const recurrentWeight =
|
|
builder.input('recurrentWeight', kExampleRecurrentWeightDescriptor);
|
|
|
|
assert_throws_js(
|
|
TypeError,
|
|
() => builder.lstm(
|
|
input, weightFromOtherBuilder, recurrentWeight, steps, hiddenSize));
|
|
}, '[lstm] throw if weight is from another builder');
|
|
|
|
|
|
multi_builder_test(async (t, builder, otherBuilder) => {
|
|
const input = builder.input('input', kExampleInputDescriptor);
|
|
const weight = builder.input('weight', kExampleWeightDescriptor);
|
|
const recurrentWeightFromOtherBuilder =
|
|
otherBuilder.input('recurrentWeight', kExampleRecurrentWeightDescriptor);
|
|
|
|
assert_throws_js(
|
|
TypeError,
|
|
() => builder.lstm(
|
|
input, weight, recurrentWeightFromOtherBuilder, steps, hiddenSize));
|
|
}, '[lstm] throw if recurrentWeight is from another builder');
|
|
|
|
multi_builder_test(async (t, builder, otherBuilder) => {
|
|
const biasFromOtherBuilder =
|
|
otherBuilder.input('bias', kExampleBiasDescriptor);
|
|
const options = {bias: biasFromOtherBuilder};
|
|
|
|
const input = builder.input('input', kExampleInputDescriptor);
|
|
const weight = builder.input('weight', kExampleWeightDescriptor);
|
|
const recurrentWeight =
|
|
builder.input('recurrentWeight', kExampleRecurrentWeightDescriptor);
|
|
assert_throws_js(
|
|
TypeError,
|
|
() => builder.lstm(
|
|
input, weight, recurrentWeight, steps, hiddenSize, options));
|
|
}, '[lstm] throw if bias option is from another builder');
|
|
|
|
multi_builder_test(async (t, builder, otherBuilder) => {
|
|
const recurrentBiasFromOtherBuilder =
|
|
otherBuilder.input('bias', kExampleBiasDescriptor);
|
|
const options = {recurrentBias: recurrentBiasFromOtherBuilder};
|
|
|
|
const input = builder.input('input', kExampleInputDescriptor);
|
|
const weight = builder.input('weight', kExampleWeightDescriptor);
|
|
const recurrentWeight =
|
|
builder.input('recurrentWeight', kExampleRecurrentWeightDescriptor);
|
|
assert_throws_js(
|
|
TypeError,
|
|
() => builder.lstm(
|
|
input, weight, recurrentWeight, steps, hiddenSize, options));
|
|
}, '[lstm] throw if recurrentBias option is from another builder');
|
|
|
|
multi_builder_test(async (t, builder, otherBuilder) => {
|
|
const peepholeWeightFromOtherBuilder =
|
|
otherBuilder.input('peepholeWeight', kExamplePeepholeWeightDescriptor);
|
|
const options = {peepholeWeight: peepholeWeightFromOtherBuilder};
|
|
|
|
const input = builder.input('input', kExampleInputDescriptor);
|
|
const weight = builder.input('weight', kExampleWeightDescriptor);
|
|
const recurrentWeight =
|
|
builder.input('recurrentWeight', kExampleRecurrentWeightDescriptor);
|
|
assert_throws_js(
|
|
TypeError,
|
|
() => builder.lstm(
|
|
input, weight, recurrentWeight, steps, hiddenSize, options));
|
|
}, '[lstm] throw if peepholeWeight option is from another builder');
|
|
|
|
multi_builder_test(async (t, builder, otherBuilder) => {
|
|
const initialHiddenStateFromOtherBuilder = otherBuilder.input(
|
|
'initialHiddenState', kExampleInitialHiddenStateDescriptor);
|
|
const options = {initialHiddenState: initialHiddenStateFromOtherBuilder};
|
|
|
|
const input = builder.input('input', kExampleInputDescriptor);
|
|
const weight = builder.input('weight', kExampleWeightDescriptor);
|
|
const recurrentWeight =
|
|
builder.input('recurrentWeight', kExampleRecurrentWeightDescriptor);
|
|
assert_throws_js(
|
|
TypeError,
|
|
() => builder.lstm(
|
|
input, weight, recurrentWeight, steps, hiddenSize, options));
|
|
}, '[lstm] throw if initialHiddenState option is from another builder');
|
|
|
|
multi_builder_test(async (t, builder, otherBuilder) => {
|
|
const initialCellStateFromOtherBuilder = otherBuilder.input(
|
|
'initialCellState', kExampleInitialHiddenStateDescriptor);
|
|
const options = {initialCellState: initialCellStateFromOtherBuilder};
|
|
|
|
const input = builder.input('input', kExampleInputDescriptor);
|
|
const weight = builder.input('weight', kExampleWeightDescriptor);
|
|
const recurrentWeight =
|
|
builder.input('recurrentWeight', kExampleRecurrentWeightDescriptor);
|
|
assert_throws_js(
|
|
TypeError,
|
|
() => builder.lstm(
|
|
input, weight, recurrentWeight, steps, hiddenSize, options));
|
|
}, '[lstm] throw if initialCellState option is from another builder');
|