"vscode:/vscode.git/clone" did not exist on "bdd85358cc52c4f1b64d801402827a2cbe8658fd"
Commit 09b0780d authored by Astha Rai's avatar Astha Rai
Browse files

removed extra files

parent 77a60235
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include "ck/ck.hpp"
#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
#include "ck/tensor_operation/gpu/device/impl/device_elementwise_3d_impl.hpp"
#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
namespace ck {
namespace tensor_operation {
namespace device {
namespace instance {
using namespace ck::tensor_layout::convolution;
using F16 = ck::half_t;
using F32 = float;
template <ck::index_t... Is>
using S = ck::Sequence<Is...>;
using device_transpose_f16_instances = std::tuple<
// clang-format off FOR 16, 32, 16, 32, 16
DeviceElementwise3dImpl<ck::Tuple<F16>,
ck::Tuple<F16>,
2,
2,
1,
8,
8,
8,
ck::Sequence<8>,
ck::Sequence<8>>,
DeviceElementwise3dImpl<ck::Tuple<F16>,
ck::Tuple<F16>,
2,
2,
1,
8,
8,
8,
ck::Sequence<8>,
ck::Sequence<1>>,
DeviceElementwise3dImpl<ck::Tuple<F16>,
ck::Tuple<F16>,
2,
2,
1,
8,
8,
8,
ck::Sequence<1>,
ck::Sequence<8>>,
DeviceElementwise3dImpl<ck::Tuple<F16>,
ck::Tuple<F16>,
2,
2,
1,
8,
8,
8,
ck::Sequence<1>,
ck::Sequence<1>>,
DeviceElementwise3dImpl<ck::Tuple<F16>,
ck::Tuple<F16>,
2,
2,
1,
8,
1,
1,
ck::Sequence<1>,
ck::Sequence<1>>,
DeviceElementwise3dImpl<ck::Tuple<F16>,
ck::Tuple<F16>,
2,
2,
1,
8,
1,
1,
ck::Sequence<8>,
ck::Sequence<1>>,
DeviceElementwise3dImpl<ck::Tuple<F16>,
ck::Tuple<F16>,
2,
2,
1,
8,
4,
4,
ck::Sequence<1>,
ck::Sequence<1>>,
DeviceElementwise3dImpl<ck::Tuple<F16>,
ck::Tuple<F16>,
2,
2,
1,
8,
4,
4,
ck::Sequence<8>,
ck::Sequence<8>>
// clang-format on
>;
using device_transpose_f32_instances = std::tuple<
// clang-format off // for 16, 8, 16, 32, 8 -> test with instances for fp16
DeviceElementwise3dImpl<ck::Tuple<F32>,
ck::Tuple<F32>,
2,
2,
1,
4,
4,
4,
ck::Sequence<1>,
ck::Sequence<1>>,
DeviceElementwise3dImpl<ck::Tuple<F32>,
ck::Tuple<F32>,
2,
2,
1,
4,
4,
4,
ck::Sequence<8>,
ck::Sequence<1>>,
DeviceElementwise3dImpl<ck::Tuple<F32>,
ck::Tuple<F32>,
2,
2,
1,
4,
4,
4,
ck::Sequence<8>,
ck::Sequence<8>>,
DeviceElementwise3dImpl<ck::Tuple<F32>,
ck::Tuple<F32>,
2,
2,
1,
4,
8,
4,
ck::Sequence<8>,
ck::Sequence<8>>,
DeviceElementwise3dImpl<ck::Tuple<F32>,
ck::Tuple<F32>,
2,
2,
1,
4,
8,
8,
ck::Sequence<8>,
ck::Sequence<8>>,
DeviceElementwise3dImpl<ck::Tuple<F32>,
ck::Tuple<F32>,
2,
2,
1,
4,
8,
8,
ck::Sequence<4>,
ck::Sequence<8>>,
DeviceElementwise3dImpl<ck::Tuple<F32>,
ck::Tuple<F32>,
2,
2,
1,
4,
8,
8,
ck::Sequence<4>,
ck::Sequence<4>>,
DeviceElementwise3dImpl<ck::Tuple<F32>,
ck::Tuple<F32>,
2,
2,
1,
4,
8,
8,
ck::Sequence<8>,
ck::Sequence<4>>,
DeviceElementwise3dImpl<ck::Tuple<F32>,
ck::Tuple<F32>,
2,
2,
1,
4,
4,
8,
ck::Sequence<8>,
ck::Sequence<8>>,
// clang-format on
>;
} // namespace instance
} // namespace device
} // namespace tensor_operation
} // namespace ck
\ No newline at end of file
...@@ -21,7 +21,7 @@ template <ck::index_t... Is> ...@@ -21,7 +21,7 @@ template <ck::index_t... Is>
using S = ck::Sequence<Is...>; using S = ck::Sequence<Is...>;
using device_transpose_f16_instances = std::tuple< using device_transpose_f16_instances = std::tuple<
// FOR 16, 32, 16, 32, 16 // FOR 16, 32, 16, 32, 16
// clang-format off // clang-format off
DeviceElementwise3dImpl<ck::Tuple<F16>, ck::Tuple<F16>, PassThrough, 2, 2, 1, 8, 8, 8, ck::Sequence<8>, ck::Sequence<8>>, DeviceElementwise3dImpl<ck::Tuple<F16>, ck::Tuple<F16>, PassThrough, 2, 2, 1, 8, 8, 8, ck::Sequence<8>, ck::Sequence<8>>,
DeviceElementwise3dImpl<ck::Tuple<F16>, ck::Tuple<F16>, PassThrough, 2, 2, 1, 8, 8, 8, ck::Sequence<8>, ck::Sequence<1>>, DeviceElementwise3dImpl<ck::Tuple<F16>, ck::Tuple<F16>, PassThrough, 2, 2, 1, 8, 8, 8, ck::Sequence<8>, ck::Sequence<1>>,
...@@ -35,7 +35,7 @@ using device_transpose_f16_instances = std::tuple< ...@@ -35,7 +35,7 @@ using device_transpose_f16_instances = std::tuple<
>; >;
using device_transpose_f32_instances = std::tuple< using device_transpose_f32_instances = std::tuple<
// for 16, 8, 16, 32, 8 -> test with instances for fp16 // for 16, 8, 16, 32, 8 -> test with instances for fp16
// clang-format off // clang-format off
DeviceElementwise3dImpl<ck::Tuple<F32>, ck::Tuple<F32>, PassThrough, 2, 2, 1, 4, 4, 4, ck::Sequence<1>, ck::Sequence<1>>, DeviceElementwise3dImpl<ck::Tuple<F32>, ck::Tuple<F32>, PassThrough, 2, 2, 1, 4, 4, 4, ck::Sequence<1>, ck::Sequence<1>>,
DeviceElementwise3dImpl<ck::Tuple<F32>, ck::Tuple<F32>, PassThrough, 2, 2, 1, 4, 4, 4, ck::Sequence<8>, ck::Sequence<1>>, DeviceElementwise3dImpl<ck::Tuple<F32>, ck::Tuple<F32>, PassThrough, 2, 2, 1, 4, 4, 4, ck::Sequence<8>, ck::Sequence<1>>,
......
// SPDX-License-Identifier: MIT
// Copyright (c) 2023, Advanced Micro Devices, Inc. All rights reserved.
#include "ck/library/tensor_operation_instance/gpu/conv_tensor_rearrange/device_column_to_image_instance.hpp"
#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
namespace ck {
namespace tensor_operation {
namespace device {
namespace instance {
using namespace ck::transpose_op;
void add_device_transpose_f16_instances(
std::vector<std::unique_ptr<DeviceElementwise3dImpl<F16, F16, NCDHW, 3>>>& instances)
{
#ifdef CK_ENABLE_FP16
add_device_operation_instances(instances, device_transpose_f16_instances<1, GNWC>{});
#else
ignore = instances;
#endif
}
void add_device_transpose_f32_instances(
std::vector<std::unique_ptr<DeviceElementwise3dImpl<F32, F32, NCDHW, 3>>>& instances)
{
#ifdef CK_ENABLE_FP32
add_device_operation_instances(instances, device_transpose_f32_instances<1, GNWC>{});
#else
ignore = instances;
#endif
}
} // namespace instance
} // namespace device
} // namespace tensor_operation
} // namespace ck
\ No newline at end of file
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment