Skip to content

Commit d73131c

Browse files
axingingaxinging
authored
[js/webgpu] Use DataType as uniform cpu type (#19281)
This saves turning data type to string by tensorDataTypeEnumToString.
1 parent 85cef0a commit d73131c

37 files changed

+148
-108
lines changed

‎js/web/lib/wasm/jsep/backend-webgpu.ts

+10-8
Original file line numberDiff line numberDiff line change
@@ -3,7 +3,7 @@
33

44
import {Env, Tensor, TRACE, TRACE_FUNC_BEGIN, TRACE_FUNC_END} from 'onnxruntime-common';
55

6-
import {tensorDataTypeEnumToString} from '../wasm-common';
6+
import {DataType, tensorDataTypeEnumToString} from '../wasm-common';
77

88
import {configureLogger, LOG_DEBUG} from './log';
99
import {createView, TensorView} from './tensor-view';
@@ -453,10 +453,10 @@ export class WebGpuBackend {
453453
return;
454454
}
455455
// https://www.w3.org/TR/WGSL/#alignof
456-
const sizeOfElement = v.type === 'float16' ? 2 : 4;
456+
const sizeOfElement = v.type === DataType.float16 ? 2 : 4;
457457
let sizeOfVecOrMat;
458458
let baseAlignment;
459-
if (v.type === 'float16') {
459+
if (v.type === DataType.float16) {
460460
baseAlignment = data.length > 4 ? 16 : (data.length > 2 ? 8 : data.length * sizeOfElement);
461461
sizeOfVecOrMat = data.length > 4 ? 16 : sizeOfElement * data.length;
462462
} else {
@@ -470,7 +470,7 @@ export class WebGpuBackend {
470470
// SizeOf(vec4<i32|u32|f32>). For float16 type, when data.length > 4, the uniform variable is of type
471471
// array<mat2x4<f16>,N>, where N = Math.ceil(data.length / 8) and SizeOf(mat2x4<f16>) = 16. The total byte
472472
// length is N * SizeOf(mat2x4<f16>).
473-
const elementPerVecOrMat = v.type === 'float16' ? 8 : 4;
473+
const elementPerVecOrMat = v.type === DataType.float16 ? 8 : 4;
474474
currentOffset += data.length > 4 ? Math.ceil(data.length / elementPerVecOrMat) * sizeOfVecOrMat :
475475
data.length * sizeOfElement;
476476
});
@@ -483,15 +483,17 @@ export class WebGpuBackend {
483483
programUniforms.forEach((v, i) => {
484484
const offset = offsets[i];
485485
const data = typeof v.data === 'number' ? [v.data] : v.data;
486-
if (v.type === 'int32') {
486+
if (v.type === DataType.int32) {
487487
new Int32Array(arrayBuffer, offset, data.length).set(data);
488-
} else if (v.type === 'uint32') {
488+
} else if (v.type === DataType.uint32) {
489489
new Uint32Array(arrayBuffer, offset, data.length).set(data);
490-
} else if (v.type === 'float16') {
490+
} else if (v.type === DataType.float16) {
491491
// TODO: use Float16Array.
492492
new Uint16Array(arrayBuffer, offset, data.length).set(data);
493-
} else {
493+
} else if (v.type === DataType.float) {
494494
new Float32Array(arrayBuffer, offset, data.length).set(data);
495+
} else {
496+
throw new Error(`Unsupported uniform type: ${tensorDataTypeEnumToString(v.type)}`);
495497
}
496498
});
497499

‎js/web/lib/wasm/jsep/webgpu/ops/3rd-party/conv2d_mm_webgpu.ts

+4-3
Original file line numberDiff line numberDiff line change
@@ -19,6 +19,7 @@
1919
//
2020
// modified to fit the needs of the project
2121

22+
import {DataType} from '../../../../wasm-common';
2223
import {LOG_DEBUG} from '../../../log';
2324
import {TensorView} from '../../../tensor-view';
2425
import {ProgramInfo, ProgramInputTensorInfoDependency, ProgramUniform} from '../../types';
@@ -189,9 +190,9 @@ export const createConv2DMatMulProgramInfo =
189190
const elementsSize = isVec4 ? [innerElementSize, 4, 4] : [1, 1, 1];
190191

191192
const programUniforms: ProgramUniform[] = [
192-
{type: 'int32', data: dimAOuter}, {type: 'int32', data: dimBOuter}, {type: 'int32', data: dimInner},
193-
{type: 'int32', data: [attributes.pads[0], attributes.pads[1]]}, {type: 'int32', data: attributes.strides},
194-
{type: 'int32', data: attributes.dilations}
193+
{type: DataType.int32, data: dimAOuter}, {type: DataType.int32, data: dimBOuter},
194+
{type: DataType.int32, data: dimInner}, {type: DataType.int32, data: [attributes.pads[0], attributes.pads[1]]},
195+
{type: DataType.int32, data: attributes.strides}, {type: DataType.int32, data: attributes.dilations}
195196
];
196197
appendActivationUniformsData(attributes, programUniforms);
197198
programUniforms.push(

‎js/web/lib/wasm/jsep/webgpu/ops/3rd-party/conv_backprop_mm_webgpu.ts

+5-3
Original file line numberDiff line numberDiff line change
@@ -19,6 +19,7 @@
1919
//
2020
// modified to fit the needs of the project
2121

22+
import {DataType} from '../../../../wasm-common';
2223
import {LOG_DEBUG} from '../../../log';
2324
import {TensorView} from '../../../tensor-view';
2425
import {ProgramInfo, ProgramInputTensorInfoDependency, ProgramUniform} from '../../types';
@@ -197,9 +198,10 @@ export const createConv2DTransposeMatMulProgramInfo =
197198
];
198199

199200
const programUniforms: ProgramUniform[] = [
200-
{type: 'int32', data: dimAOuter}, {type: 'int32', data: dimBOuter}, {type: 'int32', data: dimInner},
201-
{type: 'int32', data: attributes.strides}, {type: 'int32', data: attributes.dilations},
202-
{type: 'int32', data: filterDims}, {type: 'int32', data: pads}
201+
{type: DataType.int32, data: dimAOuter}, {type: DataType.int32, data: dimBOuter},
202+
{type: DataType.int32, data: dimInner}, {type: DataType.int32, data: attributes.strides},
203+
{type: DataType.int32, data: attributes.dilations}, {type: DataType.int32, data: filterDims},
204+
{type: DataType.int32, data: pads}
203205
];
204206
appendActivationUniformsData(attributes, programUniforms);
205207
programUniforms.push(

‎js/web/lib/wasm/jsep/webgpu/ops/3rd-party/conv_backprop_webgpu.ts

+5-3
Original file line numberDiff line numberDiff line change
@@ -17,6 +17,7 @@
1717

1818
// sampled from [@tensorflow/tfjs] tfjs-backend-webgpu/src/conv_backprop_webgpu.ts
1919

20+
import {DataType} from '../../../../wasm-common';
2021
import {LOG_DEBUG} from '../../../log';
2122
import {TensorView} from '../../../tensor-view';
2223
import {ShapeUtil} from '../../../util';
@@ -264,9 +265,10 @@ export const createConvTranspose2DProgramInfo =
264265
const outputChannelsPerGroup = wShape[1];
265266

266267
const programUniforms: ProgramUniform[] = [
267-
{type: 'int32', data: outputSize}, {type: 'uint32', data: strides}, {type: 'uint32', data: filterDims},
268-
{type: 'uint32', data: dilations}, {type: 'uint32', data: effectiveFilterDims}, {type: 'int32', data: pads},
269-
{type: 'uint32', data: inputChannelsPerGroup}, {type: 'uint32', data: outputChannelsPerGroup},
268+
{type: DataType.int32, data: outputSize}, {type: DataType.uint32, data: strides},
269+
{type: DataType.uint32, data: filterDims}, {type: DataType.uint32, data: dilations},
270+
{type: DataType.uint32, data: effectiveFilterDims}, {type: DataType.int32, data: pads},
271+
{type: DataType.uint32, data: inputChannelsPerGroup}, {type: DataType.uint32, data: outputChannelsPerGroup},
270272
...createTensorShapeVariables(inputs[0].dims), ...createTensorShapeVariables(inputs[1].dims)
271273
];
272274
if (hasBias) {

‎js/web/lib/wasm/jsep/webgpu/ops/3rd-party/matmul_packed_webgpu.ts

+5-2
Original file line numberDiff line numberDiff line change
@@ -19,6 +19,7 @@
1919
//
2020
// modified to fit the needs of the project
2121

22+
import {DataType} from '../../../../wasm-common';
2223
import {TensorView} from '../../../tensor-view';
2324
import {ShapeUtil} from '../../../util';
2425
import {ProgramInfo, ProgramInputTensorInfoDependency, ProgramUniform} from '../../types';
@@ -447,8 +448,10 @@ export const createMatmulProgramInfo =
447448
const bShapeTemp = [...outerDimsB, dimInner, dimBOuter / components];
448449
const bRank = bShapeTemp.length;
449450
const outputShapeTemp = [batchSize, dimAOuter, dimBOuter / components];
450-
const programUniforms: ProgramUniform[] =
451-
[{type: 'int32', data: dimAOuter}, {type: 'int32', data: dimBOuter}, {type: 'int32', data: dimInner}];
451+
const programUniforms: ProgramUniform[] = [
452+
{type: DataType.int32, data: dimAOuter}, {type: DataType.int32, data: dimBOuter},
453+
{type: DataType.int32, data: dimInner}
454+
];
452455
appendActivationUniformsData(activationAttributes, programUniforms);
453456
programUniforms.push(
454457
...createTensorShapeVariables(outerDims), ...createTensorShapeVariables(aShapeTemp),

‎js/web/lib/wasm/jsep/webgpu/ops/attention.ts

+15-15
Original file line numberDiff line numberDiff line change
@@ -1,7 +1,7 @@
11
// Copyright (c) Microsoft Corporation. All rights reserved.
22
// Licensed under the MIT License.
33

4-
import {tensorDataTypeEnumToString} from '../../../wasm-common';
4+
import {DataType} from '../../../wasm-common';
55
import {TensorView} from '../../tensor-view';
66
import {ComputeContext, GpuDataType, ProgramUniform} from '../types';
77

@@ -241,9 +241,10 @@ export const computeInPlaceSoftmax = (context: ComputeContext, input: TensorView
241241
WG = Math.ceil(dComp / 8);
242242
}
243243
const elementsPerWG = Math.ceil(d / components / WG);
244-
const tensorDataType = tensorDataTypeEnumToString(input.dataType) as ProgramUniform['type'];
245-
const programUniforms: ProgramUniform[] =
246-
[{type: tensorDataType, data: 1 / d}, {type: 'uint32', data: dComp}, {type: 'uint32', data: elementsPerWG}];
244+
const programUniforms: ProgramUniform[] = [
245+
{type: input.dataType, data: 1 / d}, {type: DataType.uint32, data: dComp},
246+
{type: DataType.uint32, data: elementsPerWG}
247+
];
247248
const dataType = tensorTypeToWsglStorageType(input.dataType, components);
248249

249250
const getShaderSource = (shaderHelper: ShaderHelper) => {
@@ -336,11 +337,10 @@ const computeAttentionProbs =
336337
y: Math.ceil(parameters.sequenceLength / TILE_SIZE),
337338
z: parameters.batchSize * parameters.numHeads
338339
};
339-
const tensorDataType = tensorDataTypeEnumToString(q.dataType) as ProgramUniform['type'];
340340
const programUniforms: ProgramUniform[] = [
341-
{type: 'uint32', data: parameters.sequenceLength}, {type: 'uint32', data: vectorizedHeadSize},
342-
{type: 'uint32', data: parameters.totalSequenceLength}, {type: 'uint32', data: parameters.kvSequenceLength},
343-
{type: tensorDataType, data: alpha}
341+
{type: DataType.uint32, data: parameters.sequenceLength}, {type: DataType.uint32, data: vectorizedHeadSize},
342+
{type: DataType.uint32, data: parameters.totalSequenceLength},
343+
{type: DataType.uint32, data: parameters.kvSequenceLength}, {type: q.dataType, data: alpha}
344344
];
345345

346346
const inputs = [q, key];
@@ -430,9 +430,9 @@ const computeVxAttentionScore =
430430
z: params.batchSize * params.numHeads
431431
};
432432
const programUniforms: ProgramUniform[] = [
433-
{type: 'uint32', data: params.sequenceLength}, {type: 'uint32', data: params.totalSequenceLength},
434-
{type: 'uint32', data: params.vHeadSize}, {type: 'uint32', data: params.numHeads},
435-
{type: 'uint32', data: params.vHiddenSize}
433+
{type: DataType.uint32, data: params.sequenceLength}, {type: DataType.uint32, data: params.totalSequenceLength},
434+
{type: DataType.uint32, data: params.vHeadSize}, {type: DataType.uint32, data: params.numHeads},
435+
{type: DataType.uint32, data: params.vHiddenSize}
436436
];
437437

438438
const getShaderSource = (shaderHelper: ShaderHelper) => {
@@ -526,10 +526,10 @@ const prepare = (context: ComputeContext, parameters: AttentionParameters) => {
526526
};
527527
const inputs = [context.inputs[0], context.inputs[1], context.inputs[2]];
528528
const programUniforms: ProgramUniform[] = [
529-
{type: 'uint32', data: M}, {type: 'uint32', data: K}, {type: 'uint32', data: N},
530-
{type: 'uint32', data: parameters.numHeads}, {type: 'uint32', data: parameters.headSize},
531-
{type: 'uint32', data: parameters.hiddenSize},
532-
{type: 'uint32', data: parameters.hiddenSize + parameters.hiddenSize + parameters.vHiddenSize}
529+
{type: DataType.uint32, data: M}, {type: DataType.uint32, data: K}, {type: DataType.uint32, data: N},
530+
{type: DataType.uint32, data: parameters.numHeads}, {type: DataType.uint32, data: parameters.headSize},
531+
{type: DataType.uint32, data: parameters.hiddenSize},
532+
{type: DataType.uint32, data: parameters.hiddenSize + parameters.hiddenSize + parameters.vHiddenSize}
533533
];
534534

535535
const getShaderSource = (shaderHelper: ShaderHelper) => {

‎js/web/lib/wasm/jsep/webgpu/ops/batch-norm.ts

+3-2
Original file line numberDiff line numberDiff line change
@@ -3,6 +3,7 @@
33

44
import {env} from 'onnxruntime-common';
55

6+
import {DataType} from '../../../wasm-common';
67
import {TensorView} from '../../tensor-view';
78
import {ShapeUtil} from '../../util';
89
import {AttributeWithCacheKey, createAttributeWithCacheKey} from '../attribute-with-cache-key';
@@ -123,11 +124,11 @@ const createBatchNormInferenceProgramInfo =
123124
dispatchGroup: {x: Math.ceil(outputSize / 64 /* workgroup size */)},
124125
programUniforms: useShapesUniforms ?
125126
[
126-
{type: 'uint32', data: outputSize},
127+
{type: DataType.uint32, data: outputSize},
127128
...createTensorShapeVariables(yShape),
128129
] :
129130
[
130-
{type: 'uint32', data: outputSize},
131+
{type: DataType.uint32, data: outputSize},
131132
],
132133
}),
133134
};

‎js/web/lib/wasm/jsep/webgpu/ops/binary-op.ts

+1-1
Original file line numberDiff line numberDiff line change
@@ -179,7 +179,7 @@ const createBinaryOpProgramInfo =
179179
outputs: [{dims: outputShape, dataType: outputDataType}],
180180
dispatchGroup: {x: Math.ceil(outputSize / 64 /* workgroup size */ / 4 /* component size */)},
181181
programUniforms: [
182-
{type: 'uint32', data: Math.ceil(ShapeUtil.size(outputShape) / 4)},
182+
{type: DataType.uint32, data: Math.ceil(ShapeUtil.size(outputShape) / 4)},
183183
...createTensorShapeVariables(a.dims),
184184
...createTensorShapeVariables(b.dims),
185185
...createTensorShapeVariables(outputShape),

‎js/web/lib/wasm/jsep/webgpu/ops/common.ts

+3-2
Original file line numberDiff line numberDiff line change
@@ -259,8 +259,9 @@ export const tensorTypeToWsglValueType = (type: DataType, components: 1|2|3|4 =
259259
return typeof mappedType === 'string' ? mappedType : mappedType[1];
260260
};
261261

262-
export const createTensorShapeVariables = (dims: readonly number[]): ProgramUniform[] =>
263-
dims.length === 0 ? [] : [{type: 'uint32', data: dims}, {type: 'uint32', data: ShapeUtil.computeStrides(dims)}];
262+
export const createTensorShapeVariables = (dims: readonly number[]): ProgramUniform[] => dims.length === 0 ?
263+
[] :
264+
[{type: DataType.uint32, data: dims}, {type: DataType.uint32, data: ShapeUtil.computeStrides(dims)}];
264265

265266
/**
266267
* A helper function to get maximum vector size for specified data length

‎js/web/lib/wasm/jsep/webgpu/ops/concat.ts

+3-2
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,7 @@
11
// Copyright (c) Microsoft Corporation. All rights reserved.
22
// Licensed under the MIT License.
33

4+
import {DataType} from '../../../wasm-common';
45
import {TensorView} from '../../tensor-view';
56
import {ShapeUtil} from '../../util';
67
import {AttributeWithCacheKey, createAttributeWithCacheKey} from '../attribute-with-cache-key';
@@ -95,14 +96,14 @@ const createConcatProgramInfo = (inputs: readonly TensorView[], axis: number): P
9596
let previousSum = 0;
9697
const inputDependencies: ProgramInputTensorInfoDependency[] = [];
9798
const inputRanks = [];
98-
const programUniforms: ProgramUniform[] = [{type: 'uint32', data: outputSize}];
99+
const programUniforms: ProgramUniform[] = [{type: DataType.uint32, data: outputSize}];
99100
for (let i = 0; i < inputs.length; ++i) {
100101
previousSum += inputs[i].dims[adjustedAxis];
101102
sizeInConcatAxis[i] = previousSum;
102103
inputRanks.push(inputs[i].dims.length);
103104
inputVars[i] = inputVariable(`input${i}`, dataType, inputRanks[i]);
104105
inputDependencies.push('rank');
105-
programUniforms.push({type: 'uint32', data: sizeInConcatAxis[i]});
106+
programUniforms.push({type: DataType.uint32, data: sizeInConcatAxis[i]});
106107
}
107108
for (let i = 0; i < inputs.length; ++i) {
108109
programUniforms.push(...createTensorShapeVariables(inputs[i].dims));

‎js/web/lib/wasm/jsep/webgpu/ops/conv-grouped.ts

+8-5
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,7 @@
11
// Copyright (c) Microsoft Corporation. All rights reserved.
22
// Licensed under the MIT License.
33

4+
import {DataType} from '../../../wasm-common';
45
import {TensorView} from '../../tensor-view';
56
import {ShapeUtil} from '../../util';
67
import {ProgramInfo, ProgramInputTensorInfoDependency, ProgramUniform} from '../types';
@@ -28,9 +29,10 @@ export const createGroupedConvProgramInfo =
2829
const outputSize = ShapeUtil.size(outputShape);
2930

3031
const programUniforms: ProgramUniform[] = [
31-
{type: 'uint32', data: outputSize}, {type: 'uint32', data: attributes.dilations},
32-
{type: 'uint32', data: [attributes.strides[0], attributes.strides[1]]},
33-
{type: 'uint32', data: [attributes.pads[0], attributes.pads[1]]}, {type: 'uint32', data: outputChannelsPerGroup}
32+
{type: DataType.uint32, data: outputSize}, {type: DataType.uint32, data: attributes.dilations},
33+
{type: DataType.uint32, data: [attributes.strides[0], attributes.strides[1]]},
34+
{type: DataType.uint32, data: [attributes.pads[0], attributes.pads[1]]},
35+
{type: DataType.uint32, data: outputChannelsPerGroup}
3436
];
3537
appendActivationUniformsData(attributes, programUniforms);
3638
programUniforms.push(
@@ -127,8 +129,9 @@ export const createGroupedConvVectorizeProgramInfo =
127129
const outputShapeInShader = [outputShape[0], outputShape[1], outputShape[2], outputShape[3] / components];
128130

129131
const programUniforms: ProgramUniform[] = [
130-
{type: 'uint32', data: outputSize}, {type: 'int32', data: [attributes.strides[0], attributes.strides[1]]},
131-
{type: 'int32', data: [attributes.pads[0], attributes.pads[1]]}
132+
{type: DataType.uint32, data: outputSize},
133+
{type: DataType.int32, data: [attributes.strides[0], attributes.strides[1]]},
134+
{type: DataType.int32, data: [attributes.pads[0], attributes.pads[1]]}
132135
];
133136
appendActivationUniformsData(attributes, programUniforms);
134137
programUniforms.push(

‎js/web/lib/wasm/jsep/webgpu/ops/cumsum.ts

+1-1
Original file line numberDiff line numberDiff line change
@@ -54,7 +54,7 @@ const createCumsumProgramInfo =
5454
outputs: [{dims: inputShape, dataType: inputType}],
5555
dispatchGroup: {x: Math.ceil(outputSize / 64 /* workgroup size */)},
5656
programUniforms: [
57-
{type: 'uint32', data: outputSize}, {type: 'int32', data: axis},
57+
{type: DataType.uint32, data: outputSize}, {type: DataType.int32, data: axis},
5858
...createTensorShapeVariables(inputShape), ...createTensorShapeVariables(inputShape)
5959
]
6060

‎js/web/lib/wasm/jsep/webgpu/ops/einsum.ts

+5-2
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,7 @@
11
// Copyright (c) Microsoft Corporation. All rights reserved.
22
// Licensed under the MIT License.
33

4+
import {DataType} from '../../../wasm-common';
45
import {TensorView} from '../../tensor-view';
56
import {ShapeUtil} from '../../util';
67
import {AttributeWithCacheKey, createAttributeWithCacheKey} from '../attribute-with-cache-key';
@@ -272,8 +273,10 @@ const createEinsumProgramInfo =
272273
// filter is added to make sure that dimValue is never 0.
273274
const programUniformsInit: ProgramUniform[] =
274275
uniformsSymbols.filter((symbol) => einsumEquation.symbolToInfo.has(symbol))
275-
.map((symbol) => ({type: 'uint32', data: einsumEquation.symbolToInfo.get(symbol)?.dimValue || 0}));
276-
programUniformsInit.push({type: 'uint32', data: outputSize});
276+
.map(
277+
(symbol) =>
278+
({type: DataType.uint32, data: einsumEquation.symbolToInfo.get(symbol)?.dimValue || 0}));
279+
programUniformsInit.push({type: DataType.uint32, data: outputSize});
277280
const programUniforms: ProgramUniform[] =
278281
inputShapes.map((dims, _) => [...createTensorShapeVariables(dims)])
279282
.reduce((acc, inputProgramUniforms) => acc.concat(inputProgramUniforms), programUniformsInit);

‎js/web/lib/wasm/jsep/webgpu/ops/expand.ts

+1-1
Original file line numberDiff line numberDiff line change
@@ -85,7 +85,7 @@ const createExpandProgramInfo = (inputs: readonly TensorView[]): ProgramInfo =>
8585
};
8686

8787
const programUniforms: ProgramUniform[] = [
88-
{type: 'uint32', data: outputSize}, ...createTensorShapeVariables(inputShape),
88+
{type: DataType.uint32, data: outputSize}, ...createTensorShapeVariables(inputShape),
8989
...createTensorShapeVariables(outputShape)
9090
];
9191
return {

‎js/web/lib/wasm/jsep/webgpu/ops/fuse-utils.ts

+5-2
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,7 @@
11
// Copyright (c) Microsoft Corporation. All rights reserved.
22
// Licensed under the MIT License.
33

4+
import {DataType} from '../../../wasm-common';
45
import {MAX_CLIP, MIN_CLIP} from '../../util';
56
import {ProgramUniform} from '../types';
67

@@ -36,9 +37,11 @@ export const getActivationSnippet = (attributes: InternalActivationAttributes, v
3637
export const appendActivationUniformsData =
3738
(attributes: InternalActivationAttributes, programUniform: ProgramUniform[]) => {
3839
if (attributes.activation === 'Clip') {
39-
programUniform.push({type: 'float32', data: attributes.clipMax!}, {type: 'float32', data: attributes.clipMin!});
40+
programUniform.push(
41+
{type: DataType.float, data: attributes.clipMax!}, {type: DataType.float, data: attributes.clipMin!});
4042
} else if (attributes.activation === 'HardSigmoid') {
41-
programUniform.push({type: 'float32', data: attributes.alpha!}, {type: 'float32', data: attributes.beta!});
43+
programUniform.push(
44+
{type: DataType.float, data: attributes.alpha!}, {type: DataType.float, data: attributes.beta!});
4245
}
4346
};
4447

0 commit comments

Comments
 (0)