File size: 1,761 Bytes
f12ab4c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 |
/*
* SPDX-FileCopyrightText: Copyright (c) 2021-2022 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: LicenseRef-NvidiaProprietary
*
* NVIDIA CORPORATION, its affiliates and licensors retain all intellectual
* property and proprietary rights in and to this material, related
* documentation and any modifications thereto. Any use, reproduction,
* disclosure or distribution of this material and related documentation
* without an express license agreement from NVIDIA CORPORATION or
* its affiliates is strictly prohibited.
*/
#include "filtered_lrelu.cu"
// Template/kernel specializations for no signs mode (no gradients required).
// Full op, 32-bit indexing.
template filtered_lrelu_kernel_spec choose_filtered_lrelu_kernel<c10::Half, int32_t, false, false>(const filtered_lrelu_kernel_params& p, int sharedKB);
template filtered_lrelu_kernel_spec choose_filtered_lrelu_kernel<float, int32_t, false, false>(const filtered_lrelu_kernel_params& p, int sharedKB);
// Full op, 64-bit indexing.
template filtered_lrelu_kernel_spec choose_filtered_lrelu_kernel<c10::Half, int64_t, false, false>(const filtered_lrelu_kernel_params& p, int sharedKB);
template filtered_lrelu_kernel_spec choose_filtered_lrelu_kernel<float, int64_t, false, false>(const filtered_lrelu_kernel_params& p, int sharedKB);
// Activation/signs only for generic variant. 64-bit indexing.
template void* choose_filtered_lrelu_act_kernel<c10::Half, false, false>(void);
template void* choose_filtered_lrelu_act_kernel<float, false, false>(void);
template void* choose_filtered_lrelu_act_kernel<double, false, false>(void);
// Copy filters to constant memory.
template cudaError_t copy_filters<false, false>(cudaStream_t stream);
|