multi_batch_merge_pass.cc 13.4 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
//   Copyright (c) 2018 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 "paddle/fluid/framework/ir/multi_batch_merge_pass.h"

#include <map>
#include <string>
19 20
#include <unordered_map>
#include <unordered_set>
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
#include <vector>
#include "paddle/fluid/framework/ir/graph_helper.h"
#include "paddle/fluid/framework/op_proto_maker.h"

namespace paddle {
namespace framework {
namespace ir {

static const char kNumRepeats[] = "num_repeats";
typedef std::unordered_map<std::string, std::vector<ir::Node*>> SSAVarList;

ir::Node* SameNameVar(std::unordered_set<ir::Node*> all, ir::Node* target) {
  for (auto n : all) {
    if (target->IsVar() && target->Name() == n->Name()) {
      return n;
    }
  }
  return nullptr;
}

VarDesc CopyVarDesc(VarDesc* var_desc) {
  VarDesc repeated_var(var_desc->Name());
  // copy other variable attributes
  if (var_desc->GetType() != proto::VarType::READER) {
    repeated_var.SetType(var_desc->GetType());
    repeated_var.SetShape(var_desc->GetShape());
    repeated_var.SetDataType(var_desc->GetDataType());
    repeated_var.SetLoDLevel(var_desc->GetLoDLevel());
    repeated_var.SetPersistable(var_desc->Persistable());
  } else {
    // TODO(typhoonzero): copy reader var
  }
  return repeated_var;
}

VarDesc UpdateGradVarDesc(
    VarDesc* var_desc, int repeat,
    const std::unordered_set<std::string>& grad_names,
    const std::unordered_set<std::string>& bn_vars_need_rename) {
  if (grad_names.find(var_desc->Name()) != grad_names.end() ||
      bn_vars_need_rename.find(var_desc->Name()) != bn_vars_need_rename.end()) {
    std::string new_gname =
        string::Sprintf("%s.repeat.%d", var_desc->Name(), repeat);
    VarDesc repeated_var = CopyVarDesc(var_desc);
    repeated_var.SetName(new_gname);
M
minqiyang 已提交
66
    VLOG(3) << "update " << var_desc->Name() << " to repeat " << repeat;
67 68 69 70 71
    return repeated_var;
  }
  return *var_desc;
}

72
void BatchMergePass::ApplyImpl(ir::Graph* graph) const {
73 74 75 76 77
  int num_repeats = Get<const int>(kNumRepeats);
  std::vector<Node*> forward_backward_ops;
  std::vector<Node*> optimize_ops;
  std::vector<Node*> lr_ops;  // ops other than forward/backward/optimize
  std::unordered_set<std::string> grad_names;
W
Wu Yi 已提交
78
  std::unordered_map<std::string, std::string> gradname2paramname;
79 80 81

  std::vector<ir::Node*> nodes = TopologySortOperations(*graph);
  auto origin_nodes = graph->ReleaseNodes();
M
minqiyang 已提交
82
  VLOG(3) << "origin nodes count: " << origin_nodes.size();
83 84 85 86
  ir::Graph& result = *graph;

  // 1. record op nodes of different roles
  for (auto node : nodes) {
87 88
    if (!node->IsOp()) continue;
    PADDLE_ENFORCE(node->Op(), "must find opdesc");
89 90 91 92 93 94 95 96 97 98 99 100 101 102 103
    int op_role = boost::get<int>(node->Op()->GetAttr(
        framework::OpProtoAndCheckerMaker::OpRoleAttrName()));
    if ((op_role == static_cast<int>(framework::OpRole::kForward)) ||
        (op_role & static_cast<int>(framework::OpRole::kBackward)) ||
        (op_role & static_cast<int>(framework::OpRole::kLoss))) {
      forward_backward_ops.push_back(node);
    } else if ((op_role & static_cast<int>(framework::OpRole::kOptimize)) ||
               (op_role & static_cast<int>(framework::OpRole::kDist)) ||
               (op_role & static_cast<int>(framework::OpRole::kRPC))) {
      optimize_ops.push_back(node);
      auto op_role_var = node->Op()->GetNullableAttr(
          OpProtoAndCheckerMaker::OpRoleVarAttrName());
      auto op_role_vars = boost::get<std::vector<std::string>>(op_role_var);
      for (size_t i = 0; i < op_role_vars.size(); i += 2) {
        grad_names.insert(op_role_vars[i + 1]);
W
Wu Yi 已提交
104
        gradname2paramname[op_role_vars[i + 1]] = op_role_vars[i];
105 106 107 108 109 110 111 112 113 114
      }
    } else if (op_role & static_cast<int>(framework::OpRole::kLRSched)) {
      lr_ops.push_back(node);
    } else {  // NOLINT
      PADDLE_THROW("Invalid op_role: %d", static_cast<int>(op_role));
    }
  }

  // 2. copy forward backward
  ir::Node* prev_repeat_last_op_node = nullptr;
W
Wu Yi 已提交
115
  // record origin_grad -> repeated_grad_list map.
116 117 118 119 120 121 122 123 124 125 126 127 128 129
  std::map<ir::Node*, std::vector<ir::Node*>> grad_repeated_map;
  std::map<std::string, std::vector<ir::Node*>> created;
  std::unordered_set<std::string> bn_vars_need_rename;
  for (int i = 0; i < num_repeats; ++i) {
    std::unordered_set<ir::Node*> copied;
    for (size_t node_idx = 0; node_idx < forward_backward_ops.size();
         ++node_idx) {
      auto node = forward_backward_ops[node_idx];
      OpDesc repeated_op(*(node->Op()), node->Op()->Block());
      // 3. rename grad outputs to current repeat.
      for (auto outname : repeated_op.OutputArgumentNames()) {
        if (grad_names.find(outname) != grad_names.end()) {
          std::string new_gname = string::Sprintf("%s.repeat.%d", outname, i);
          repeated_op.RenameOutput(outname, new_gname);
W
Wu Yi 已提交
130 131 132 133
          // remove op_role_var for backward ops that outputs grad for a
          // parameter.
          repeated_op.SetAttr(OpProtoAndCheckerMaker::OpRoleVarAttrName(),
                              std::vector<std::string>());
134 135 136
        }
      }
      // 3.5 let batch_norm ops use independent vars, note batch_norm_grad do
W
Wu Yi 已提交
137 138 139
      // not need this update, because only moving mean and variance should be
      // differ, trainable parameter scale and bias is the same as other
      // parameters.
140 141 142 143 144 145 146 147 148
      if (node->Name() == "batch_norm") {
        // NOTE: assume bn op created by layers use save var as output mean and
        // variance
        std::string new_mean_name =
            string::Sprintf("%s.repeat.%d", repeated_op.Input("Mean")[0], i);
        std::string new_var_name = string::Sprintf(
            "%s.repeat.%d", repeated_op.Input("Variance")[0], i);
        bn_vars_need_rename.insert(repeated_op.Input("Mean")[0]);
        bn_vars_need_rename.insert(repeated_op.Input("Variance")[0]);
M
minqiyang 已提交
149 150
        VLOG(3) << "renaming " << repeated_op.Input("Mean")[0] << " to "
                << new_mean_name;
151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235
        repeated_op.RenameInput(repeated_op.Input("Mean")[0], new_mean_name);
        repeated_op.RenameInput(repeated_op.Input("Variance")[0], new_var_name);
        repeated_op.RenameOutput(repeated_op.Output("MeanOut")[0],
                                 new_mean_name);
        repeated_op.RenameOutput(repeated_op.Output("VarianceOut")[0],
                                 new_var_name);
      }

      // 3.9 do copy
      auto repeated_node = result.CreateOpNode(&repeated_op);
      copied.insert(node);

      // 4. add deps between repeats
      if (node_idx == forward_backward_ops.size() - 1) {
        prev_repeat_last_op_node = repeated_node;
      }
      if (node_idx == 0 && prev_repeat_last_op_node) {
        auto* depvar = result.CreateControlDepVar();
        prev_repeat_last_op_node->outputs.push_back(depvar);
        depvar->inputs.push_back(prev_repeat_last_op_node);
        repeated_node->inputs.push_back(depvar);
        depvar->outputs.push_back(repeated_node);
      }

      for (auto in_node : node->inputs) {
        if (in_node->IsCtrlVar()) {
          continue;
        }
        ir::Node* var = nullptr;
        auto updated_var = UpdateGradVarDesc(in_node->Var(), i, grad_names,
                                             bn_vars_need_rename);
        // should be initialized by startup, how to initilize tensor in the
        // scope?
        if (node->Name() == "batch_norm" &&
            bn_vars_need_rename.find(in_node->Name()) !=
                bn_vars_need_rename.end()) {
          // Create bn mean/variance for each repeat
          var = result.CreateVarNode(&updated_var);
          created[updated_var.Name()].push_back(var);
          copied.insert(in_node);
          repeated_node->inputs.push_back(var);
          var->outputs.push_back(repeated_node);
          continue;
        }

        // for other ops
        if (in_node->inputs.empty() && i > 0) {
          // do not copy head vars (inputs, params) in repeats > 0
          var = created.at(in_node->Name()).back();
        } else {
          if (copied.find(in_node) == copied.end()) {
            var = result.CreateVarNode(&updated_var);
            if (grad_names.find(in_node->Var()->Name()) != grad_names.end()) {
              grad_repeated_map[in_node].push_back(var);
            }
            copied.insert(in_node);
            created[updated_var.Name()].push_back(var);
          } else {
            var = created.at(updated_var.Name()).back();
          }
        }
        repeated_node->inputs.push_back(var);
        var->outputs.push_back(repeated_node);
      }
      for (auto out_node : node->outputs) {
        if (out_node->IsCtrlVar()) {
          continue;
        }
        ir::Node* var = nullptr;
        auto updated_var = UpdateGradVarDesc(out_node->Var(), i, grad_names,
                                             bn_vars_need_rename);
        if (copied.find(out_node) == copied.end()) {
          var = result.CreateVarNode(&updated_var);
          if (grad_names.find(out_node->Var()->Name()) != grad_names.end()) {
            grad_repeated_map[out_node].push_back(var);
          }
          copied.insert(out_node);
          created[updated_var.Name()].push_back(var);
        } else {
          var = created.at(updated_var.Name()).back();
        }
        repeated_node->outputs.push_back(var);
        var->inputs.push_back(repeated_node);
      }
    }
W
Wu Yi 已提交
236
  }  // end copy forward backward
237

W
Wu Yi 已提交
238 239
  // 5. create GRAD merge op node: sum(repeat.0...repeat.n) ->
  // scale(1/num_repeats)
240 241 242 243
  for (auto kv : grad_repeated_map) {
    OpDesc sum_op;
    sum_op.SetType("sum");
    std::vector<std::string> repeated_grad_names;
W
Wu Yi 已提交
244
    std::vector<std::string> param_grad_op_role_var;
245 246 247
    for (auto r : kv.second) {
      repeated_grad_names.push_back(r->Var()->Name());
    }
W
Wu Yi 已提交
248 249 250 251 252 253 254
    // NOTE: use op_role_var to control allreduce op appending in
    //       multi_devices_graph_pass, we want to append op_role_var
    //       only once for the merged gradient, so break after first call.
    param_grad_op_role_var.push_back(
        gradname2paramname.at(kv.first->Var()->Name()));        // param
    param_grad_op_role_var.push_back(kv.first->Var()->Name());  // grad

255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276
    sum_op.SetInput("X", repeated_grad_names);
    sum_op.SetOutput("Out", {kv.first->Var()->Name()});
    sum_op.SetAttr(OpProtoAndCheckerMaker::OpRoleAttrName(),
                   static_cast<int>(OpRole::kBackward));
    auto sum_op_node = result.CreateOpNode(&sum_op);
    for (auto r : kv.second) {
      sum_op_node->inputs.push_back(r);
      r->outputs.push_back(sum_op_node);
    }
    auto sum_out_var_node = result.CreateVarNode(kv.first->Var());
    sum_op_node->outputs.push_back(sum_out_var_node);
    sum_out_var_node->inputs.push_back(sum_op_node);
    created[sum_out_var_node->Name()].push_back(sum_out_var_node);

    OpDesc scale_op;
    scale_op.SetType("scale");
    scale_op.SetInput("X", {sum_out_var_node->Var()->Name()});
    // NOTE: inplace scale.
    scale_op.SetOutput("Out", {sum_out_var_node->Var()->Name()});
    scale_op.SetAttr("scale", static_cast<float>(1.0f / num_repeats));
    scale_op.SetAttr(OpProtoAndCheckerMaker::OpRoleAttrName(),
                     static_cast<int>(OpRole::kBackward));
W
Wu Yi 已提交
277 278 279 280

    scale_op.SetAttr(OpProtoAndCheckerMaker::OpRoleVarAttrName(),
                     param_grad_op_role_var);

281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336
    auto scale_op_node = result.CreateOpNode(&scale_op);
    scale_op_node->inputs.push_back(sum_out_var_node);
    sum_out_var_node->outputs.push_back(scale_op_node);
    auto scale_out_var_node = result.CreateVarNode(sum_out_var_node->Var());
    scale_op_node->outputs.push_back(scale_out_var_node);
    scale_out_var_node->inputs.push_back(scale_op_node);
    created[scale_out_var_node->Name()].push_back(scale_out_var_node);
  }
  // 6. add optimize ops
  {
    auto copy_node = [&result, &created](ir::Node* node) {
      auto op_node = result.CreateOpNode(node->Op());
      // copy op ins/outs
      // NOTE: for send/recv ops, the OpDesc uses ctrldepvar to describe
      // dependencies, so create those depvars if OpDesc have in/outs.
      for (auto in_node : node->inputs) {
        if (in_node->IsCtrlVar() && !in_node->Var()) {
          continue;
        }
        ir::Node* var = nullptr;
        if (created.find(in_node->Name()) == created.end()) {
          var = result.CreateVarNode(in_node->Var());
          created[in_node->Name()].push_back(var);
        } else {
          var = created.at(in_node->Name()).back();
        }
        op_node->inputs.push_back(var);
        var->outputs.push_back(op_node);
      }
      for (auto out_node : node->outputs) {
        if (out_node->IsCtrlVar() && !out_node->Var()) {
          continue;
        }
        auto var = result.CreateVarNode(out_node->Var());
        created[out_node->Name()].push_back(var);
        op_node->outputs.push_back(var);
        var->inputs.push_back(op_node);
      }
    };
    for (auto node : lr_ops) {
      copy_node(node);
    }
    for (auto node : optimize_ops) {
      copy_node(node);
    }
  }

  result.ResolveHazard(created);
}

}  // namespace ir
}  // namespace framework
}  // namespace paddle

REGISTER_PASS(multi_batch_merge_pass, paddle::framework::ir::BatchMergePass)
    .RequirePassAttr(paddle::framework::ir::kNumRepeats);