mirror of
https://github.com/saymrwulf/onnxruntime.git
synced 2026-06-26 03:00:54 +00:00
abdc31de40
### Description
See
454996d496
for manual changes (excluded auto-generated formatting changes)
### Why
Because the toolsets for old clang-format is out-of-date. This reduces
the development efficiency.
- The NPM package `clang-format` is already in maintenance mode. not
updated since 2 years ago.
- The VSCode extension for clang-format is not maintained for a while,
and a recent Node.js security update made it not working at all in
Windows.
No one in community seems interested in fixing those.
Choose Prettier as it is the most popular TS/JS formatter.
### How to merge
It's easy to break the build:
- Be careful of any new commits on main not included in this PR.
- Be careful that after this PR is merged, other PRs that already passed
CI can merge.
So, make sure there is no new commits before merging this one, and
invalidate js PRs that already passed CI, force them to merge to latest.
941 lines
23 KiB
TypeScript
941 lines
23 KiB
TypeScript
// Copyright (c) Microsoft Corporation. All rights reserved.
|
|
// Licensed under the MIT License.
|
|
|
|
import { Attribute } from '../../../../lib/onnxjs/attribute';
|
|
import { Backend, InferenceHandler, resolveBackend, SessionHandler } from '../../../../lib/onnxjs/backend';
|
|
import { Profiler } from '../../../../lib/onnxjs/instrument';
|
|
import { Tensor } from '../../../../lib/onnxjs/tensor';
|
|
import { PoolConvUtil } from '../../../../lib/onnxjs/util';
|
|
import { TensorResultValidator } from '../../../test-runner';
|
|
import { createMockGraph } from '../../../test-shared';
|
|
|
|
import { conv2d } from './test-conv-utils';
|
|
|
|
function createRandomArray(size: number): Float32Array {
|
|
const randomTable = [0, 3, 6, 9, 2, 5, 8, 1, 4, 7];
|
|
return new Float32Array(
|
|
Array.from({ length: size }, (_v, k) => randomTable[k % 10] * 0.1 + randomTable[Math.trunc(k / 10) % 10] * 0.01),
|
|
);
|
|
}
|
|
interface TestData {
|
|
inputShape: number[];
|
|
kernelShape: number[];
|
|
biasShape: number[];
|
|
autoPad?: string;
|
|
pads?: number[];
|
|
dilations: number[];
|
|
strides: number[];
|
|
group: number;
|
|
}
|
|
function getTestData(): TestData[] {
|
|
return [
|
|
{
|
|
inputShape: [1, 3, 416, 416],
|
|
kernelShape: [64, 3, 3, 3],
|
|
biasShape: [],
|
|
autoPad: 'SAME_UPPER',
|
|
dilations: [1, 1],
|
|
strides: [1, 1],
|
|
group: 1,
|
|
},
|
|
{
|
|
inputShape: [1, 3, 224, 224],
|
|
kernelShape: [64, 3, 3, 3],
|
|
biasShape: [64],
|
|
pads: [0, 0, 0, 0],
|
|
dilations: [1, 1],
|
|
strides: [2, 2],
|
|
group: 1,
|
|
},
|
|
{
|
|
inputShape: [1, 64, 55, 55],
|
|
kernelShape: [16, 64, 1, 1],
|
|
biasShape: [16],
|
|
pads: [0, 0, 0, 0],
|
|
dilations: [1, 1],
|
|
strides: [1, 1],
|
|
group: 1,
|
|
},
|
|
// {
|
|
// inputShape: [1, 16, 55, 55],
|
|
// kernelShape: [64, 16, 1, 1],
|
|
// biasShape: [64],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 16, 55, 55],
|
|
// kernelShape: [64, 16, 3, 3],
|
|
// biasShape: [64],
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 128, 55, 55],
|
|
// kernelShape: [16, 128, 1, 1],
|
|
// biasShape: [16],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 16, 55, 55],
|
|
// kernelShape: [64, 16, 1, 1],
|
|
// biasShape: [64],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 16, 55, 55],
|
|
// kernelShape: [64, 16, 3, 3],
|
|
// biasShape: [64],
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 128, 27, 27],
|
|
// kernelShape: [32, 128, 1, 1],
|
|
// biasShape: [32],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 32, 27, 27],
|
|
// kernelShape: [128, 32, 1, 1],
|
|
// biasShape: [128],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 32, 27, 27],
|
|
// kernelShape: [128, 32, 3, 3],
|
|
// biasShape: [128],
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 27, 27],
|
|
// kernelShape: [32, 256, 1, 1],
|
|
// biasShape: [32],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 32, 27, 27],
|
|
// kernelShape: [128, 32, 1, 1],
|
|
// biasShape: [128],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 32, 27, 27],
|
|
// kernelShape: [128, 32, 3, 3],
|
|
// biasShape: [128],
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 13, 13],
|
|
// kernelShape: [48, 256, 1, 1],
|
|
// biasShape: [48],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 48, 13, 13],
|
|
// kernelShape: [192, 48, 1, 1],
|
|
// biasShape: [192],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 48, 13, 13],
|
|
// kernelShape: [192, 48, 3, 3],
|
|
// biasShape: [192],
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 384, 13, 13],
|
|
// kernelShape: [48, 384, 1, 1],
|
|
// biasShape: [48],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 48, 13, 13],
|
|
// kernelShape: [192, 48, 1, 1],
|
|
// biasShape: [192],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 48, 13, 13],
|
|
// kernelShape: [192, 48, 3, 3],
|
|
// biasShape: [192],
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 384, 13, 13],
|
|
// kernelShape: [64, 384, 1, 1],
|
|
// biasShape: [64],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 64, 13, 13],
|
|
// kernelShape: [256, 64, 1, 1],
|
|
// biasShape: [256],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 64, 13, 13],
|
|
// kernelShape: [256, 64, 3, 3],
|
|
// biasShape: [256],
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 13, 13],
|
|
// kernelShape: [64, 512, 1, 1],
|
|
// biasShape: [64],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 64, 13, 13],
|
|
// kernelShape: [256, 64, 1, 1],
|
|
// biasShape: [256],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 64, 13, 13],
|
|
// kernelShape: [256, 64, 3, 3],
|
|
// biasShape: [256],
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 13, 13],
|
|
// kernelShape: [1000, 512, 1, 1],
|
|
// biasShape: [1000],
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
{
|
|
inputShape: [1, 3, 7, 7],
|
|
kernelShape: [8, 3, 3, 3],
|
|
biasShape: [],
|
|
pads: [1, 1, 1, 1],
|
|
dilations: [1, 1],
|
|
strides: [1, 1],
|
|
group: 1,
|
|
},
|
|
{
|
|
inputShape: [1, 2, 3, 3],
|
|
kernelShape: [4, 2, 1, 1],
|
|
biasShape: [],
|
|
pads: [0, 0, 0, 0],
|
|
dilations: [1, 1],
|
|
strides: [1, 1],
|
|
group: 1,
|
|
},
|
|
{
|
|
inputShape: [1, 3, 224, 224],
|
|
kernelShape: [64, 3, 7, 7],
|
|
biasShape: [],
|
|
pads: [3, 3, 3, 3],
|
|
dilations: [1, 1],
|
|
strides: [2, 2],
|
|
group: 1,
|
|
},
|
|
// {
|
|
// inputShape: [1, 64, 56, 56],
|
|
// kernelShape: [256, 64, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 64, 56, 56],
|
|
// kernelShape: [64, 64, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 64, 56, 56],
|
|
// kernelShape: [256, 64, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 56, 56],
|
|
// kernelShape: [64, 256, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 64, 56, 56],
|
|
// kernelShape: [64, 64, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 64, 56, 56],
|
|
// kernelShape: [256, 64, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 56, 56],
|
|
// kernelShape: [64, 256, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 64, 56, 56],
|
|
// kernelShape: [64, 64, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 64, 56, 56],
|
|
// kernelShape: [256, 64, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 56, 56],
|
|
// kernelShape: [128, 256, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 56, 56],
|
|
// kernelShape: [512, 256, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [2, 2],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 128, 56, 56],
|
|
// kernelShape: [128, 128, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [2, 2],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 128, 28, 28],
|
|
// kernelShape: [512, 128, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 28, 28],
|
|
// kernelShape: [128, 512, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 128, 28, 28],
|
|
// kernelShape: [128, 128, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 128, 28, 28],
|
|
// kernelShape: [512, 128, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 28, 28],
|
|
// kernelShape: [128, 512, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 128, 28, 28],
|
|
// kernelShape: [128, 128, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 128, 28, 28],
|
|
// kernelShape: [512, 128, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 28, 28],
|
|
// kernelShape: [128, 512, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 128, 28, 28],
|
|
// kernelShape: [128, 128, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 128, 28, 28],
|
|
// kernelShape: [512, 128, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 28, 28],
|
|
// kernelShape: [256, 512, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 28, 28],
|
|
// kernelShape: [1024, 512, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [2, 2],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 28, 28],
|
|
// kernelShape: [256, 256, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [2, 2],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [1024, 256, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 1024, 14, 14],
|
|
// kernelShape: [256, 1024, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [256, 256, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [1024, 256, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 1024, 14, 14],
|
|
// kernelShape: [256, 1024, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [256, 256, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [1024, 256, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 1024, 14, 14],
|
|
// kernelShape: [256, 1024, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [256, 256, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [1024, 256, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 1024, 14, 14],
|
|
// kernelShape: [256, 1024, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [256, 256, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [1024, 256, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 1024, 14, 14],
|
|
// kernelShape: [256, 1024, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [256, 256, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 256, 14, 14],
|
|
// kernelShape: [1024, 256, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 1024, 14, 14],
|
|
// kernelShape: [512, 1024, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 1024, 14, 14],
|
|
// kernelShape: [2048, 1024, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [2, 2],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 14, 14],
|
|
// kernelShape: [512, 512, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [2, 2],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 7, 7],
|
|
// kernelShape: [2048, 512, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 2048, 7, 7],
|
|
// kernelShape: [512, 2048, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
{
|
|
inputShape: [1, 512, 7, 7],
|
|
kernelShape: [512, 512, 3, 3],
|
|
biasShape: [],
|
|
pads: [1, 1, 1, 1],
|
|
dilations: [1, 1],
|
|
strides: [1, 1],
|
|
group: 1,
|
|
},
|
|
{
|
|
inputShape: [1, 512, 7, 7],
|
|
kernelShape: [2048, 512, 1, 1],
|
|
biasShape: [],
|
|
|
|
pads: [0, 0, 0, 0],
|
|
dilations: [1, 1],
|
|
strides: [1, 1],
|
|
group: 1,
|
|
},
|
|
// {
|
|
// inputShape: [1, 2048, 7, 7],
|
|
// kernelShape: [512, 2048, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 7, 7],
|
|
// kernelShape: [512, 512, 3, 3],
|
|
// biasShape:[],
|
|
|
|
// pads: [1, 1, 1, 1],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
// {
|
|
// inputShape: [1, 512, 7, 7],
|
|
// kernelShape: [2048, 512, 1, 1],
|
|
// biasShape:[],
|
|
|
|
// pads: [0, 0, 0, 0],
|
|
// dilations: [1, 1],
|
|
// strides: [1, 1],
|
|
// group: 1
|
|
// },
|
|
];
|
|
}
|
|
|
|
const validator = new TensorResultValidator('webgl');
|
|
let webglBackend: Backend | undefined;
|
|
let webglSessionhandler: SessionHandler | undefined;
|
|
let webglInferenceHandler: InferenceHandler | undefined;
|
|
|
|
function webglConv(
|
|
inputTensor: Tensor,
|
|
kernelTensor: Tensor,
|
|
biasTensor: Tensor | null,
|
|
autoPad: string | undefined,
|
|
dilations: number[],
|
|
pads: number[] | undefined,
|
|
strides: number[],
|
|
): Tensor {
|
|
const attributes = new Attribute(undefined);
|
|
attributes.set('dilations', 'ints', dilations);
|
|
attributes.set('auto_pad', 'string', autoPad ? autoPad : '');
|
|
attributes.set('kernel_shape', 'ints', kernelTensor.dims.slice(2));
|
|
if (pads) {
|
|
attributes.set('pads', 'ints', pads);
|
|
}
|
|
attributes.set('strides', 'ints', strides);
|
|
const graph = createMockGraph('Conv', attributes);
|
|
const op = webglSessionhandler!.resolve(graph.getNodes()[0], [{ domain: '', version: 7 }], graph);
|
|
const inputs = [inputTensor, kernelTensor];
|
|
if (biasTensor) {
|
|
inputs.push(biasTensor);
|
|
}
|
|
return op.impl(webglInferenceHandler!, inputs, op.context)[0];
|
|
}
|
|
function cpuConv(
|
|
inputTensor: Tensor,
|
|
kernelTensor: Tensor,
|
|
biasTensor: Tensor | null,
|
|
autoPad: string | undefined,
|
|
dilations: number[],
|
|
pads: number[] | undefined,
|
|
strides: number[],
|
|
): Tensor {
|
|
const attributes = new Attribute(undefined);
|
|
attributes.set('dilations', 'ints', dilations);
|
|
attributes.set('auto_pad', 'string', autoPad ? autoPad : '');
|
|
attributes.set('kernel_shape', 'ints', kernelTensor.dims.slice(2));
|
|
if (pads) {
|
|
attributes.set('pads', 'ints', pads);
|
|
}
|
|
attributes.set('strides', 'ints', strides);
|
|
|
|
const x = inputTensor;
|
|
const w = kernelTensor;
|
|
const b = biasTensor || undefined;
|
|
|
|
const adjustedPads = pads ? pads.slice(0) : [0, 0, 0, 0];
|
|
const outputDims = PoolConvUtil.computeConvOutputShape(
|
|
x.dims,
|
|
w.dims,
|
|
strides,
|
|
dilations,
|
|
kernelTensor.dims.slice(2),
|
|
adjustedPads,
|
|
autoPad,
|
|
);
|
|
const y = new Tensor(outputDims, x.type);
|
|
conv2d(y, x, w, b, dilations, 1, adjustedPads, strides);
|
|
return y;
|
|
}
|
|
describe('New Conv tests', () => {
|
|
before(async () => {
|
|
const profiler = Profiler.create();
|
|
webglBackend = await resolveBackend('webgl');
|
|
webglSessionhandler = webglBackend.createSessionHandler({ profiler });
|
|
webglInferenceHandler = webglSessionhandler.createInferenceHandler();
|
|
});
|
|
const testDataSet = getTestData();
|
|
for (let k = 0; k < testDataSet.length; ++k) {
|
|
const testData = testDataSet[k];
|
|
describe(`Testing ${JSON.stringify(testData)}`, () => {
|
|
const inputData = createRandomArray(testData.inputShape.reduce((a, b) => a * b));
|
|
const kernelData = createRandomArray(testData.kernelShape.reduce((a, b) => a * b));
|
|
const biasData = testData.biasShape.length === 1 ? createRandomArray(testData.biasShape[0]) : null;
|
|
const rgbas = [false];
|
|
rgbas.forEach((rgba) => {
|
|
describe(`RGBA: ${rgba}`, () => {
|
|
before(function () {
|
|
const patchSize = testData.kernelShape.slice(1).reduce((a, b) => a * b);
|
|
if (rgba && patchSize % 4 !== 0) {
|
|
// eslint-disable-next-line no-invalid-this
|
|
this.skip();
|
|
}
|
|
});
|
|
it('', () => {
|
|
// create new Tensors otherwise the session/inference level caching would cause issues
|
|
const inputTensor = new Tensor(testData.inputShape, 'float32', undefined, undefined, inputData);
|
|
const kernelTensor = new Tensor(testData.kernelShape, 'float32', undefined, undefined, kernelData);
|
|
const biasTensor = biasData
|
|
? new Tensor(testData.biasShape, 'float32', undefined, undefined, biasData)
|
|
: null;
|
|
const actual = webglConv(
|
|
inputTensor,
|
|
kernelTensor,
|
|
biasTensor,
|
|
testData.autoPad,
|
|
testData.dilations,
|
|
testData.pads,
|
|
testData.strides,
|
|
);
|
|
const expected = cpuConv(
|
|
inputTensor,
|
|
kernelTensor,
|
|
biasTensor,
|
|
testData.autoPad,
|
|
testData.dilations,
|
|
testData.pads,
|
|
testData.strides,
|
|
);
|
|
try {
|
|
validator.checkTensorResult([actual], [expected]);
|
|
} catch {
|
|
console.log(actual.dims, `[${actual.numberData.slice(0, 20).join(',')},...]`);
|
|
console.log(expected.dims, `[${expected.numberData.slice(0, 20).join(',')},...]`);
|
|
throw new Error('Expected and Actual did not match');
|
|
}
|
|
});
|
|
});
|
|
});
|
|
});
|
|
}
|
|
});
|