Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[JS/WebGPU] Preserve zero size input tensor dims. #19737

Merged
merged 12 commits into from
Mar 8, 2024
Merged

[JS/WebGPU] Preserve zero size input tensor dims. #19737

Show file tree
Hide file tree
Changes from 4 commits
Commits
Show all changes
12 commits
Select commit Hold shift + click to select a range
da6e9f5
Preserve zero size input tensor dims.
satyajandhyala Mar 1, 2024
ae01d11
Use adjustedAxis instead of axis
satyajandhyala Mar 1, 2024
2c0d67c
Refactor; add more tests.
satyajandhyala Mar 1, 2024
ce6a004
Moved the new concat test cases from concat.jsonc to concat_zero-size…
satyajandhyala Mar 1, 2024
51c83d7
Removed calculateOutputShape function.
satyajandhyala Mar 4, 2024
2bfc23e
Removed special case handling of all inputs zero-sized
satyajandhyala Mar 5, 2024
e8fb4d2
Use adjusted axis to validate.
satyajandhyala Mar 5, 2024
f19ece3
Rollback code allowing zero-sized input non-concat axes dims mismatch…
satyajandhyala Mar 6, 2024
c0fab59
Keep the variable name adjustedAxis
satyajandhyala Mar 6, 2024
ed048b0
Removed referenceInput.
satyajandhyala Mar 7, 2024
971b612
Fixed comment
satyajandhyala Mar 7, 2024
adfca56
Check if the axis is with in the range.
satyajandhyala Mar 8, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
70 changes: 49 additions & 21 deletions js/web/lib/wasm/jsep/webgpu/ops/concat.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,22 +13,23 @@ export interface ConcatAttributes extends AttributeWithCacheKey {
readonly axis: number;
}

const validateInputs = (inputs: readonly TensorView[]): void => {
const validateInputs = (inputs: readonly TensorView[], referenceIndex: number): void => {
if (!inputs || inputs.length < 1) {
throw new Error('too few inputs');
}
fs-eire marked this conversation as resolved.
Show resolved Hide resolved

const inputType = inputs[0].dataType;
const inputDimensionality = inputs[0].dims.length;

const inputType = inputs[referenceIndex].dataType;
const inputDimensionality = inputs[referenceIndex].dims.length;
const referenceInput = inputs[referenceIndex];
for (const input of inputs) {
// make sure types of all inputs match
if (input.dataType !== inputType) {
throw new Error('input tensors should be one type');
}

// make sure the dimensionality of all inputs are the same
if (input.dims.length !== inputDimensionality) {
if (input.dims.length !== inputDimensionality && ShapeUtil.size(input.dims) > 0 &&
ShapeUtil.size(referenceInput.dims) > 0) {
throw new Error('input tensors should have the same shape');
}
}
Expand Down Expand Up @@ -64,29 +65,35 @@ const assignOutputData = (inputs: readonly IndicesHelper[], output: IndicesHelpe
return codeLines.join('\n');
};

const createConcatProgramInfo = (inputs: readonly TensorView[], axis: number): ProgramInfo => {
const inputShape = inputs[0].dims.slice();

const computeOutputShape = (inputs: readonly TensorView[], axis: number, referenceIndex: number): number[] => {
const inputShape = inputs[referenceIndex].dims.slice();
if (axis >= inputShape.length || axis < (-1 * inputShape.length)) {
throw new Error('axis specified for concat doesn\'t match input dimensionality');
}
const adjustedAxis = (axis < 0) ? inputShape.length + axis : axis;
// ensure all of the non-concatenated axes match each other
// calculate the shape of the output tensor while we do that
const outputShape = inputShape.slice(0);
for (let i = 1; i < inputs.length; i++) {
for (let i = 0; i < inputs.length; i++) {
if (i === referenceIndex) {
continue;
}
const dataNShape = inputs[i].dims.slice();
for (let axisIndex = 0; axisIndex < inputShape.length; axisIndex++) {
// add to the placeholder for computing output shape
if (axisIndex === adjustedAxis) {
outputShape[adjustedAxis] += dataNShape[axisIndex];
if (axisIndex === axis) {
outputShape[axis] += dataNShape[axisIndex];
}
// ensure all non-cancatenated axes match each other
else if (inputShape[axisIndex] !== dataNShape[axisIndex]) {
else if (inputShape[axisIndex] !== dataNShape[axisIndex] && ShapeUtil.size(dataNShape) > 0) {
throw new Error('non concat dimensions must match');
}
}
}
return outputShape;
};

const createConcatProgramInfo = (inputs: readonly TensorView[], axis: number, outputShape: number[]): ProgramInfo => {
const outputSize = ShapeUtil.size(outputShape);

const sizeInConcatAxis = new Array<number>(inputs.length);
Expand All @@ -98,7 +105,7 @@ const createConcatProgramInfo = (inputs: readonly TensorView[], axis: number): P
const inputRanks = [];
const programUniforms: ProgramUniform[] = [{type: DataType.uint32, data: outputSize}];
for (let i = 0; i < inputs.length; ++i) {
previousSum += inputs[i].dims[adjustedAxis];
previousSum += inputs[i].dims[axis];
sizeInConcatAxis[i] = previousSum;
inputRanks.push(inputs[i].dims.length);
inputVars[i] = inputVariable(`input${i}`, dataType, inputRanks[i]);
Expand All @@ -111,7 +118,7 @@ const createConcatProgramInfo = (inputs: readonly TensorView[], axis: number): P
programUniforms.push(...createTensorShapeVariables(outputShape));

const output = outputVariable('output', dataType, outputShape.length);
const indicesAxis = output.indicesGet('indices', adjustedAxis);
const indicesAxis = output.indicesGet('indices', axis);
const sizeInConcatAxisStr =
Array.from(Array(sizeInConcatAxis.length).keys()).map(i => `uniforms.sizeInConcatAxis${i}`).join(',');
const getShaderSource = (shaderHelper: ShaderHelper) => `
Expand All @@ -132,12 +139,16 @@ const createConcatProgramInfo = (inputs: readonly TensorView[], axis: number): P
var indices = ${output.offsetToIndices('global_idx')};

let inputIndex = calculateInputIndex(${indicesAxis});
if (inputIndex != 0u) {
let sizeInConcatAxis = array<u32, ${sizeInConcatAxis.length}u>(${sizeInConcatAxisStr});
${indicesAxis} -= sizeInConcatAxis[inputIndex - 1u];
}
if (inputIndex < ${inputs.length}u) {
fs-eire marked this conversation as resolved.
Show resolved Hide resolved
if (inputIndex != 0u) {
let sizeInConcatAxis = array<u32, ${sizeInConcatAxis.length}u>(${sizeInConcatAxisStr});
${indicesAxis} -= sizeInConcatAxis[inputIndex - 1u];
}

${assignOutputData(inputVars, output)}
${assignOutputData(inputVars, output)}
} else {
${output.setByOffset('global_idx', '0')}
}
}`;

return {
Expand All @@ -153,10 +164,27 @@ const createConcatProgramInfo = (inputs: readonly TensorView[], axis: number): P
};

export const concat = (context: ComputeContext, attributes: ConcatAttributes): void => {
validateInputs(context.inputs);
// find a none zero tensor to determine the output shape
// Choose input with max rank if all input tensors are zero size to make the output shape independent of the order of
// the inputs.
let referenceIndex = context.inputs.findIndex(input => ShapeUtil.size(input.dims) > 0);
if (referenceIndex === -1) {
referenceIndex =
context.inputs.map(input => input.dims.length)
.reduce((maxRankIndex, rank, index, array) => rank > array[maxRankIndex] ? index : maxRankIndex, 0);
}

validateInputs(context.inputs, referenceIndex);
const inputShape = context.inputs[referenceIndex].dims;
const adjustedAxis = attributes.axis + (attributes.axis < 0 ? inputShape.length : 0);
const outputShape = computeOutputShape(context.inputs, adjustedAxis, referenceIndex);
// 0 length tensors are valid for concat, remove them
const nonEmptyInputs = context.inputs.filter(input => ShapeUtil.size(input.dims) > 0);
fs-eire marked this conversation as resolved.
Show resolved Hide resolved
context.compute(createConcatProgramInfo(nonEmptyInputs, attributes.axis), {inputs: nonEmptyInputs});
if (nonEmptyInputs.length > 0) {
fs-eire marked this conversation as resolved.
Show resolved Hide resolved
context.compute(createConcatProgramInfo(nonEmptyInputs, adjustedAxis, outputShape), {inputs: nonEmptyInputs});
} else {
context.output(0, outputShape);
}
};

export const parseConcatAttributes = (attributes: Record<string, unknown>): ConcatAttributes =>
Expand Down
Loading
Loading