Add Fuse AllReduce Pass And Adam Pass #15497
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,213 @@ | ||
| // 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 <algorithm> | ||
| #include <string> | ||
| #include <utility> | ||
| #include <vector> | ||
|
|
||
| #include "paddle/fluid/framework/details/build_strategy.h" | ||
| #include "paddle/fluid/framework/details/multi_devices_helper.h" | ||
| #include "paddle/fluid/framework/ir/graph_helper.h" | ||
| #include "paddle/fluid/framework/op_registry.h" | ||
|
|
||
| namespace paddle { | ||
| namespace framework { | ||
| namespace details { | ||
|
|
||
| class AllocContinuousSpaceForGradPass : public ir::Pass { | ||
| protected: | ||
| std::unique_ptr<ir::Graph> ApplyImpl( | ||
| std::unique_ptr<ir::Graph> graph) const override { | ||
| ir::Graph& result = *graph; | ||
|
|
||
| ResetAttribute<ParamsAndGrads>(kParamsAndGrads, &result); | ||
| ResetAttribute<GroupGradsAndParams>(kGroupGradsAndParams, &result); | ||
|
|
||
| // NOTE: The operator nodes should be in topology order. | ||
| std::vector<ir::Node*> topo_nodes = ir::TopologySortOperations(result); | ||
| auto& params_grads = result.Get<ParamsAndGrads>(kParamsAndGrads); | ||
| for (auto& node : topo_nodes) { | ||
| RecordParamsAndGrads(node, ¶ms_grads); | ||
| } | ||
|
|
||
| // Record parameters and gradients | ||
| std::unordered_map<std::string, ir::Node*> vars; | ||
| for (ir::Node* node : result.Nodes()) { | ||
| if (node->IsVar() && node->Var()) { | ||
| // Note: The graph may have the same name node. For example, parameter | ||
| // is the input of operator and it also is the output of optimizer; | ||
| vars.emplace(node->Var()->Name(), node); | ||
| } | ||
| } | ||
|
|
||
| // Note: Sort the parameters and gradient variables according | ||
| // to parameters' name to make variables' name correspond correctly. | ||
| auto& group_params_grads = | ||
| result.Get<GroupGradsAndParams>(kGroupGradsAndParams); | ||
| SortParamsAndGrads(vars, ¶ms_grads); | ||
| SetGroupGradsAndParams(vars, params_grads, &group_params_grads); | ||
|
|
||
| // Set Gradients as Persistable to prevent this var becoming reusable. | ||
| auto dtype = static_cast<proto::VarType::Type>(0); | ||
|
Contributor
Author
There was a problem hiding this comment. 0 is |
||
| for (auto& p_g : params_grads) { | ||
| // Get gradient var | ||
| auto iter = vars.find(p_g.second); | ||
| PADDLE_ENFORCE(iter != vars.end(), "%s is not found.", p_g.second); | ||
| iter->second->Var()->SetPersistable(true); | ||
|
|
||
| PADDLE_ENFORCE(IsSupportedVarType(iter->second->Var()->GetType())); | ||
|
|
||
| // Get Dtype | ||
| auto ele_dtype = iter->second->Var()->GetDataType(); | ||
| if (dtype == static_cast<proto::VarType::Type>(0)) { | ||
| dtype = ele_dtype; | ||
| PADDLE_ENFORCE_NE(ele_dtype, static_cast<proto::VarType::Type>(0)); | ||
| } | ||
| PADDLE_ENFORCE_EQ(ele_dtype, dtype); | ||
| } | ||
|
|
||
| std::vector<std::string> grads_name; | ||
| std::vector<std::string> params_name; | ||
| grads_name.reserve(params_grads.size()); | ||
| params_name.reserve(params_grads.size()); | ||
| for (auto& p_g : params_grads) { | ||
| params_name.emplace_back(p_g.first); | ||
| grads_name.emplace_back(p_g.second); | ||
| } | ||
|
|
||
| // Create the fused variable name. | ||
| const std::string prefix(kFusedVarNamePrefix); | ||
|
Contributor
Author
There was a problem hiding this comment. I think when we analyze the var's name, we can recognize the fused_var by the prefix. And it also prevents the name conflict with the existing names. |
||
| auto fused_var_name = prefix + "_GRAD"; | ||
| if (!result.Has(kFusedVars)) { | ||
| result.Set(kFusedVars, new FusedVars); | ||
| } | ||
| result.Get<FusedVars>(kFusedVars).emplace_back(fused_var_name); | ||
|
|
||
| // Insert alloc_continuous_space_for_grad to RunOnlyOnceProgram, | ||
| // which is executed before running the model with ParallelExecutor. | ||
| if (!result.Has(kRunOnlyOnceProgram)) { | ||
| result.Set(kRunOnlyOnceProgram, new RunOnlyOnceProgram); | ||
| } | ||
| result.Get<RunOnlyOnceProgram>(kRunOnlyOnceProgram).emplace_back(); | ||
| auto& program_desc = | ||
| result.Get<RunOnlyOnceProgram>(kRunOnlyOnceProgram).back(); | ||
|
Contributor
There was a problem hiding this comment. just run the operations in this pass. not need to save if for later? |
||
| auto* global_block = program_desc.MutableBlock(0); | ||
|
|
||
| AppendAllocSpaceForVarsOp(params_name, grads_name, fused_var_name, | ||
| global_block); | ||
|
|
||
| return std::move(graph); | ||
| } | ||
|
|
||
| template <typename AttrType> | ||
| void ResetAttribute(const std::string& attr_name, ir::Graph* graph) const { | ||
| if (graph->Has(attr_name)) { | ||
| VLOG(10) << attr_name << " is reset."; | ||
| graph->Erase(attr_name); | ||
|
Contributor
There was a problem hiding this comment. throw error? It's the job of build_strategy to do this.
Contributor
Author
There was a problem hiding this comment. I think if this pass is run many times, the result is should be the same. |
||
| } | ||
| graph->Set(attr_name, new AttrType); | ||
| } | ||
|
|
||
| void SortParamsAndGrads( | ||
| const std::unordered_map<std::string, ir::Node*>& var_nodes, | ||
| ParamsAndGrads* params_grads) const { | ||
| // TODO(zcd): The sort should be removed. | ||
|
|
||
| // std::sort(params_grads->begin(), params_grads->end(), | ||
| // [](const std::pair<std::string, std::string>& a, | ||
| // const std::pair<std::string, std::string>& b) -> bool { | ||
| // return a.first < b.first; | ||
| // }); | ||
| } | ||
|
|
||
| void SetGroupGradsAndParams( | ||
| const std::unordered_map<std::string, ir::Node*>& var_nodes, | ||
| const ParamsAndGrads& params_grads, | ||
| GroupGradsAndParams* group_params_grads) const { | ||
| // group_size | ||
| const size_t group_size = 3; | ||
|
Contributor
Author
There was a problem hiding this comment. This is just an experimental. |
||
| size_t groups = (params_grads.size() + group_size - 1) / group_size; | ||
| group_params_grads->reserve(groups); | ||
|
|
||
| size_t j = 0; | ||
| for (size_t i = 0; i < groups; ++i) { | ||
| group_params_grads->emplace_back(); | ||
| auto& group_p_g = group_params_grads->back(); | ||
| group_p_g.reserve(group_size); | ||
| VLOG(10) << "Group:" << i; | ||
| std::stringstream out; | ||
| while (j < params_grads.size()) { | ||
| group_p_g.emplace_back( | ||
| std::make_pair(params_grads.at(j).second /*grad*/, | ||
| params_grads.at(j).first /*param*/)); | ||
| out << params_grads.at(j).second << "[" << params_grads.at(j).first | ||
| << "] "; | ||
| ++j; | ||
| if (j % group_size == 0) break; | ||
| } | ||
| VLOG(10) << out.str(); | ||
| } | ||
| } | ||
|
|
||
| private: | ||
| bool IsSupportedVarType(const proto::VarType::Type& type) const { | ||
| // Current only support LOD_TENSOR. | ||
| return type == proto::VarType::LOD_TENSOR; | ||
| } | ||
|
|
||
| void AppendAllocSpaceForVarsOp(const std::vector<std::string>& params_name, | ||
| const std::vector<std::string>& grads_name, | ||
| const std::string& fused_var_name, | ||
| BlockDesc* global_block) const { | ||
| auto op_desc = global_block->AppendOp(); | ||
| op_desc->SetType("alloc_continuous_space"); | ||
| op_desc->SetInput("Input", params_name); | ||
| op_desc->SetOutput("Output", grads_name); | ||
| op_desc->SetOutput("FusedOutput", {fused_var_name}); | ||
| } | ||
|
|
||
| void RecordParamsAndGrads(ir::Node* node, | ||
| ParamsAndGrads* params_grads) const { | ||
| try { | ||
| bool is_bk_op = | ||
| static_cast<bool>(boost::get<int>(node->Op()->GetAttr( | ||
| OpProtoAndCheckerMaker::OpRoleAttrName())) & | ||
| static_cast<int>(OpRole::kBackward)); | ||
| if (!is_bk_op) return; | ||
|
|
||
| // Currently, we assume that once gradient is generated, it can be | ||
| // broadcast, and each gradient is only broadcast once. | ||
| auto backward_vars = | ||
| boost::get<std::vector<std::string>>(node->Op()->GetNullableAttr( | ||
| OpProtoAndCheckerMaker::OpRoleVarAttrName())); | ||
| PADDLE_ENFORCE_EQ(backward_vars.size() % 2, static_cast<size_t>(0)); | ||
|
|
||
| for (size_t i = 0; i < backward_vars.size(); i += 2) { | ||
| VLOG(10) << "Trainable parameter: " << backward_vars[i] | ||
| << ", gradient: " << backward_vars[i + 1]; | ||
|
|
||
| params_grads->emplace_back(std::make_pair( | ||
| backward_vars[i] /*param*/, backward_vars[i + 1] /*grad*/)); | ||
| } | ||
| } catch (boost::bad_get e) { | ||
| } | ||
| } | ||
| }; | ||
|
|
||
| } // namespace details | ||
| } // namespace framework | ||
| } // namespace paddle | ||
|
|
||
| REGISTER_PASS(alloc_continuous_space_for_grad_pass, | ||
| paddle::framework::details::AllocContinuousSpaceForGradPass); | ||
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can you put your pass after memory optimize so you don't need to make them persistable?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Maybe, but if I make them as persistable, I don't need care about the position of memory_opt_pass.