/**
|
* @license
|
* Copyright 2020 Google LLC. All Rights Reserved.
|
* Licensed under the Apache License, Version 2.0 (the "License");
|
* you may not use this file except in compliance with the License.
|
* You may obtain a copy of the License at
|
*
|
* http://www.apache.org/licenses/LICENSE-2.0
|
*
|
* Unless required by applicable law or agreed to in writing, software
|
* distributed under the License is distributed on an "AS IS" BASIS,
|
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
* See the License for the specific language governing permissions and
|
* limitations under the License.
|
* =============================================================================
|
*/
|
|
import {BatchMatMul, BatchMatMulAttrs, BatchMatMulInputs, KernelConfig} from '@tensorflow/tfjs';
|
|
import {createTensorsTypeOpAttr, NodeJSKernelBackend} from '../nodejs_kernel_backend';
|
|
export const batchMatMulConfig: KernelConfig = {
|
kernelName: BatchMatMul,
|
backendName: 'tensorflow',
|
kernelFunc: (args) => {
|
const {a, b} = args.inputs as BatchMatMulInputs;
|
const backend = args.backend as NodeJSKernelBackend;
|
const {transposeA, transposeB} = args.attrs as unknown as BatchMatMulAttrs;
|
|
const opAttrs = [
|
createTensorsTypeOpAttr('T', a.dtype),
|
{name: 'adj_x', type: backend.binding.TF_ATTR_BOOL, value: transposeA},
|
{name: 'adj_y', type: backend.binding.TF_ATTR_BOOL, value: transposeB}
|
];
|
|
// libtensorflow's BatchMatMulV2 op performs the same behavior as other tfjs
|
// backends' BatchMatMul (supports broadcasting), so a string literal is
|
// used here to point to libtensorflow's BatchMatMulV2 op, instead of using
|
// const `BatchMatMul` ('BatchMatMul') to resolve node-backend's special
|
// mapping.
|
return backend.executeSingleOutput('BatchMatMulV2', opAttrs, [a, b]);
|
}
|
};
|