calib_compute.cc 4.1 KB
Newer Older
Y
Yan Chunwei 已提交
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 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114
// Copyright (c) 2019 PaddlePaddle Authors. 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.

#include "lite/kernels/fpga/calib_compute.h"
#include <vector>
#include "lite/core/op_registry.h"
#include "lite/core/type_system.h"
#include "lite/fpga/KD/float16.hpp"

namespace paddle {
namespace lite {
namespace kernels {
namespace fpga {
using float16 = zynqmp::float16;
void CalibComputeFp32ToFP16::Run() {
  auto& param = this->Param<operators::CalibParam>();
  const auto* din = param.input->data<float>();
  auto* dout = param.output->mutable_data<float16>(TARGET(kFPGA));

  for (int i = 0; i < param.input->numel(); ++i) {
    dout[i] = zynqmp::float_to_half(din[i]);
  }
  return;
}

void CalibComputeFP16ToFp32::Run() {
  auto& param = this->Param<operators::CalibParam>();
  const auto* din = param.input->data<float16>();
  auto* dout = param.output->mutable_data<float>(TARGET(kFPGA));
  for (int i = 0; i < param.input->numel(); ++i) {
    dout[i] = zynqmp::half_to_float(din[i]);
  }
  return;
}

}  // namespace fpga
}  // namespace kernels
}  // namespace lite
}  // namespace paddle

REGISTER_LITE_KERNEL(calib,
                     kFPGA,
                     kFP16,
                     kNHWC,
                     paddle::lite::kernels::fpga::CalibComputeFp32ToFP16,
                     fp32_to_fp16_fpga)
    .BindInput("Input",
               {LiteType::GetTensorTy(TARGET(kFPGA),
                                      PRECISION(kFloat),
                                      DATALAYOUT(kNCHW))})
    .BindOutput("Out",
                {LiteType::GetTensorTy(TARGET(kFPGA),
                                       PRECISION(kFP16),
                                       DATALAYOUT(kNCHW))})
    .Finalize();

REGISTER_LITE_KERNEL(calib,
                     kFPGA,
                     kFP16,
                     kNHWC,
                     paddle::lite::kernels::fpga::CalibComputeFP16ToFp32,
                     fp16_to_fp32_fpga)
    .BindInput("Input",
               {LiteType::GetTensorTy(TARGET(kFPGA),
                                      PRECISION(kFP16),
                                      DATALAYOUT(kNHWC))})
    .BindOutput("Out",
                {LiteType::GetTensorTy(TARGET(kFPGA),
                                       PRECISION(kFloat),
                                       DATALAYOUT(kNHWC))})
    .Finalize();

REGISTER_LITE_KERNEL(calib_once,
                     kFPGA,
                     kFP16,
                     kNHWC,
                     paddle::lite::kernels::fpga::CalibComputeFp32ToFP16,
                     fp32_to_fp16_fpga)
    .BindInput("Input",
               {LiteType::GetTensorTy(TARGET(kFPGA),
                                      PRECISION(kFloat),
                                      DATALAYOUT(kNCHW))})
    .BindOutput("Out",
                {LiteType::GetTensorTy(TARGET(kFPGA),
                                       PRECISION(kFP16),
                                       DATALAYOUT(kNCHW))})
    .Finalize();

REGISTER_LITE_KERNEL(calib_once,
                     kFPGA,
                     kFP16,
                     kNHWC,
                     paddle::lite::kernels::fpga::CalibComputeFP16ToFp32,
                     fp16_to_fp32_fpga)
    .BindInput("Input",
               {LiteType::GetTensorTy(TARGET(kFPGA),
                                      PRECISION(kFP16),
                                      DATALAYOUT(kNHWC))})
    .BindOutput("Out",
                {LiteType::GetTensorTy(TARGET(kFPGA),
                                       PRECISION(kFloat),
                                       DATALAYOUT(kNHWC))})
    .Finalize();