Skip to content

[JSEP] adjust edge case logic for scatternd #24172

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

Merged
prathikr merged 3 commits into from
Mar 26, 2025
Merged

[JSEP] adjust edge case logic for scatternd #24172

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
139766c
working
prathikr Mar 25, 2025
7e36aab
clean up impl
prathikr Mar 25, 2025
a5204e6
formatting
prathikr Mar 25, 2025
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
88 changes: 49 additions & 39 deletions js/web/lib/wasm/jsep/webgpu/ops/scatter-nd.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -78,6 +78,36 @@ const atomicReductionSnippet = (reduction: string, ptr: string, v: string, type:
}
};

const calcDataOffsetSnippet = (dataRank: number, parallel: boolean) =>
`${
dataRank === 1
? `
let element_count_dim = uniforms.output_strides;
let dim_value = uniforms.output_shape;`
: `
let element_count_dim = uniforms.output_strides[${parallel ? 'i - indices_start' : 'i'}];
let dim_value = uniforms.output_shape[${parallel ? 'i - indices_start' : 'i'} + uniforms.last_index_dimension];`
}

if (index >= 0) {
if (index >= i32(dim_value)) {
index = i32(dim_value - 1);
}
} else {
if (index < -i32(dim_value)) {
index = 0;
} else {
index += i32(dim_value);
}
}
data_offset += u32((u32(index) * element_count_dim));`;

const updateElementsSnippet = (attributes: ScatterNDAttributes, outputTypeValue: ReductionType, parallel: boolean) =>
`for (var i = 0u; i < uniforms.num_updates_elements; i++) {
let value = updates[uniforms.num_updates_elements * ${parallel ? 'global_idx' : 'idx'} + i];
${atomicReductionSnippet(attributes.reduction, 'output[data_offset + i]', 'value', outputTypeValue)}
}`;

const createScatterNDProgramInfo = (inputs: readonly TensorView[], attributes: ScatterNDAttributes): ProgramInfo => {
const inputShape = inputs[0].dims;
const indicesShape = inputs[1].dims;
Expand All @@ -87,6 +117,7 @@ const createScatterNDProgramInfo = (inputs: readonly TensorView[], attributes: S
const outputSize = Math.ceil(ShapeUtil.size(indicesShape) / components);
const lastIndexDimension = indicesShape[indicesShape.length - 1];
const numUpdatesElements = ShapeUtil.sizeFromDimension(inputShape, lastIndexDimension);
const numIndicesElements = ShapeUtil.sizeFromDimension(indicesShape, 0) / lastIndexDimension;

const programUniforms: ProgramUniform[] = [
{ type: DataType.uint32, data: outputSize },
Expand All @@ -113,9 +144,8 @@ const createScatterNDProgramInfo = (inputs: readonly TensorView[], attributes: S
${shaderHelper.guardAgainstOutOfBoundsWorkgroupSizes('uniforms.output_size')}
var hasDuplicates = false;
if (${attributes.reduction === 'none'}) {
let n = ${ShapeUtil.size(indicesShape)};
for (var i = 0; i < n; i = i + 1) {
for (var j = i + 1; j < n; j = j + 1) {
for (var i = 0; i < ${numIndicesElements}; i = i + 1) {
for (var j = i + 1; j < ${numIndicesElements}; j = j + 1) {
var index_i = i32(indices[i].x);
var index_j = i32(indices[j].x);
if (index_i == index_j) {
Expand All @@ -129,51 +159,31 @@ const createScatterNDProgramInfo = (inputs: readonly TensorView[], attributes: S
}
}

var data_offset = 0u;
var indices_start = uniforms.last_index_dimension * global_idx;
if (${attributes.reduction === 'none'} && hasDuplicates) {
if (global_idx != 0u) {
return;
}
indices_start = 0u;
}
let indices_end = indices_start + uniforms.last_index_dimension;
for (var i = indices_start; i < indices_end; i++) {
var index = i32(indices[i].x);
${
inputs[0].dims.length === 1
? `
let element_count_dim = uniforms.output_strides;
let dim_value = uniforms.output_shape;`
: `
let element_count_dim = uniforms.output_strides[i - indices_start];
let dim_value = uniforms.output_shape[i - indices_start + uniforms.last_index_dimension];`
}
if (index >= 0) {
if (index >= i32(dim_value)) {
index = i32(dim_value - 1);
}
} else {
if (index < -i32(dim_value)) {
index = 0;
} else {
index += i32(dim_value);
// Process each index-update pair individually when duplicates exist
for (var idx = 0u; idx < ${numIndicesElements}u; idx++) {
var data_offset = 0u;
for (var i = 0u; i < uniforms.last_index_dimension; i++) {
var index = i32(indices[idx * uniforms.last_index_dimension + i].x);
${calcDataOffsetSnippet(inputShape.length, false)}
}
${updateElementsSnippet(attributes, output.type.value as ReductionType, false)}
}
data_offset += u32((u32(index) * element_count_dim));
return;
}

for (var i = 0u; i < uniforms.num_updates_elements; i++) {
let value = updates[uniforms.num_updates_elements * global_idx + i];
${atomicReductionSnippet(
attributes.reduction,
'output[data_offset + i]',
'value',
output.type.value as ReductionType,
)}
var data_offset = 0u;
var indices_start = uniforms.last_index_dimension * global_idx;
var indices_end = indices_start + uniforms.last_index_dimension;
for (var i = indices_start; i < indices_end; i++) {
var index = i32(indices[i].x);
${calcDataOffsetSnippet(inputShape.length, true)}
}

}`;
${updateElementsSnippet(attributes, output.type.value as ReductionType, true)}
}`;
};
return {
name: 'ScatterND',
Expand Down
Loading