selected_rows.cc 6.3 KB
Newer Older
1 2
/* Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.

Q
qijun 已提交
3 4 5
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
6

Q
qijun 已提交
7
    http://www.apache.org/licenses/LICENSE-2.0
8

Q
qijun 已提交
9 10 11 12 13 14
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. */

Y
Yi Wang 已提交
15
#include "paddle/fluid/framework/selected_rows.h"
Q
qijun 已提交
16 17

namespace paddle {
18
namespace framework {
19

Y
Yancey1989 已提交
20
struct ReAllocateVisitor {
21 22
  ReAllocateVisitor(const framework::DDim& dims, framework::Tensor* tensor)
      : dims_(dims), tensor_(tensor) {}
Y
Yancey1989 已提交
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

  template <typename T>
  void operator()() const {
    framework::Tensor cpu_tensor;
    platform::CPUPlace cpu;
    T* ptr = cpu_tensor.mutable_data<T>(dims_, cpu);
    const T* old_ptr =
        tensor_->memory_size() == 0 ? nullptr : tensor_->data<T>();
    if (old_ptr != nullptr) {
      std::copy(old_ptr, old_ptr + tensor_->numel(), ptr);
    }
    tensor_->ShareDataWith(cpu_tensor);
  }

  framework::DDim dims_;
Q
qiaolongfei 已提交
38
  framework::Tensor* tensor_;
Y
Yancey1989 已提交
39 40
};

Y
update  
Yancey1989 已提交
41
struct TensorCopyVisitor {
Y
Yancey1989 已提交
42 43 44 45
  TensorCopyVisitor(framework::Tensor* dst, int64_t dst_offset,
                    const framework::Tensor src, int64_t src_offset,
                    int64_t size)
      : dst_(dst),
Y
Yancey1989 已提交
46 47 48 49 50 51
        dst_offset_(dst_offset),
        src_(src),
        src_offset_(src_offset),
        size_(size) {}

  template <typename T>
D
dzhwinter 已提交
52
  void apply() const {
Y
Yancey1989 已提交
53 54 55 56
    // TODO(Yancey1989): support other place
    platform::CPUPlace cpu;
    memory::Copy(cpu, dst_->mutable_data<T>(cpu) + dst_offset_, cpu,
                 src_.data<T>() + src_offset_, size_ * sizeof(T));
Y
Yancey1989 已提交
57 58 59 60 61 62 63 64 65
  }

  framework::Tensor* dst_;
  int64_t dst_offset_;
  framework::Tensor src_;
  int64_t src_offset_;
  int64_t size_;
};

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
void SerializeToStream(std::ostream& os, const SelectedRows& selected_rows,
                       const platform::DeviceContext& dev_ctx) {
  {  // the 1st field, uint32_t version
    constexpr uint32_t version = 0;
    os.write(reinterpret_cast<const char*>(&version), sizeof(version));
  }
  {
    // the 2st field, rows information
    auto& rows = selected_rows.rows();
    uint64_t size = rows.size();
    os.write(reinterpret_cast<const char*>(&size), sizeof(size));
    for (uint64_t i = 0; i < size; ++i) {
      os.write(reinterpret_cast<const char*>(&rows[i]), sizeof(rows[i]));
    }
  }
  {
    // the 3st field, the height of SelectedRows
    int64_t height = selected_rows.height();
    os.write(reinterpret_cast<const char*>(&height), sizeof(height));
  }
  // the 4st field, Tensor data
Y
Yi Wang 已提交
87
  TensorToStream(os, selected_rows.value(), dev_ctx);
88 89
}

Y
Yancey 已提交
90 91
void DeserializeFromStream(std::istream& is, SelectedRows* selected_rows,
                           const platform::DeviceContext& dev_ctx) {
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114
  {
    // the 1st field, unit32_t version for SelectedRows
    uint32_t version;
    is.read(reinterpret_cast<char*>(&version), sizeof(version));
    PADDLE_ENFORCE_EQ(version, 0U, "Only version 0 is supported");
  }
  {
    // the 2st field, rows information
    uint64_t size;
    is.read(reinterpret_cast<char*>(&size), sizeof(size));
    auto& rows = *selected_rows->mutable_rows();
    rows.resize(size);
    for (uint64_t i = 0; i < size; ++i) {
      is.read(reinterpret_cast<char*>(&rows[i]), sizeof(int64_t));
    }
  }
  {
    // the 3st field, the height of the SelectedRows
    int64_t height;
    is.read(reinterpret_cast<char*>(&height), sizeof(int64_t));
    selected_rows->set_height(height);
  }
  // the 4st field, tensor which contains the data
Y
Yi Wang 已提交
115
  TensorFromStream(is, selected_rows->mutable_value(), dev_ctx);
116 117
}

Y
Yancey1989 已提交
118 119 120 121 122
bool SelectedRows::HasKey(int64_t key) const {
  return std::find(rows_.begin(), rows_.end(), key) == rows_.end() ? false
                                                                   : true;
}

123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141
int64_t SelectedRows::AutoGrownIndex(int64_t key, bool auto_grown) {
  rwlock_->RDLock();
  auto iter = id_to_index_.find(key);
  if (iter == id_to_index_.end()) {
    rwlock_->UNLock();
    if (!auto_grown) {
      PADDLE_THROW("key %d not found", key);
    }
    rwlock_->WRLock();
    auto map_size = id_to_index_.size();
    auto vector_size = rows_.size();
    if (map_size != vector_size) {
      rwlock_->UNLock();
      PADDLE_THROW(
          "id_to_index_ size %d should have the same size with rows_ %d",
          map_size, vector_size);
    }
    auto write_iter = id_to_index_.find(key);
    if (write_iter == id_to_index_.end()) {
142
      int row_num = rows_.size();
143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175
      if (row_num == value_->dims()[0]) {
        rwlock_->UNLock();
        PADDLE_THROW("selected rows is full, then length exceed %d", row_num);
      }
      // key logic to put a key into id_to_index_
      rows_.push_back(key);
      auto index = static_cast<int64_t>(rows_.size() - 1);
      id_to_index_[key] = index;
      rwlock_->UNLock();
      return index;
    } else {
      auto index = write_iter->second;
      rwlock_->UNLock();
      return index;
    }
  } else {
    auto index = iter->second;
    rwlock_->UNLock();
    return index;
  }
}

void SelectedRows::SyncIndex() {
  rwlock_->WRLock();
  id_to_index_.clear();
  for (size_t i = 0; i < rows_.size(); ++i) {
    id_to_index_[rows_[i]] = i;
  }
  rwlock_->UNLock();
}

void SelectedRows::Get(const framework::Tensor& ids, framework::Tensor* value,
                       bool auto_grown) {
Y
update  
Yancey1989 已提交
176 177
  PADDLE_ENFORCE(value->IsInitialized(),
                 "The value tensor should be initialized.");
178
  if (ids.numel() == 0) {
179
    VLOG(30) << "keys is empty, please check data!";
Q
qiaolongfei 已提交
180 181 182 183 184
  } else {
    int64_t value_width = value_->numel() / value_->dims()[0];
    PADDLE_ENFORCE_EQ(value_width, value->numel() / value->dims()[0],
                      "output tensor should have the same shape with table "
                      "except the dims[0].");
185
    for (int i = 0; i < ids.numel(); ++i) {
186 187 188 189 190
      int64_t index = AutoGrownIndex(ids.data<int64_t>()[i], auto_grown);
      framework::VisitDataType(
          framework::ToDataType(value_->type()),
          TensorCopyVisitor(value, i * value_width, *value_.get(),
                            index * value_width, value_width));
Y
Yancey1989 已提交
191 192
    }
  }
Y
Yancey1989 已提交
193 194
}

195
}  // namespace framework
Q
qijun 已提交
196
}  // namespace paddle