Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix nn graph variable bind #6895

Merged
merged 3 commits into from
Nov 30, 2021
Merged

Fix nn graph variable bind #6895

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
23f7420
fix(AutoParallel): nn.Graph support auto_parallel change sbp
wyg1997 Nov 26, 2021
bd82f98
fix(AutoParallel): use tensor.set_data interface and add print sbp info
wyg1997 Nov 30, 2021
5c694c7
add comment
wyg1997 Nov 30, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
72 changes: 49 additions & 23 deletions oneflow/core/framework/nn_graph.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -97,7 +97,7 @@ const std::vector<std::string>& NNGraph::outputs_tensor_meta_str() const {
return outputs_tensor_meta_str_;
}

int64_t NNGraph::variable_op_size() const { return variable_op_name2eager_blob_.size(); }
int64_t NNGraph::variable_op_size() const { return variable_op_names_.size(); }

Maybe<void> NNGraph::RegisterInputOpNamesAndTensors(
const std::vector<std::string>& input_op_names,
Expand Down Expand Up @@ -142,21 +142,12 @@ Maybe<void> NNGraph::RegisterVariableOpNamesAndTensors(
const std::vector<std::shared_ptr<one::Tensor>>& variable_tensors) {
JUST(vm::CurrentRankSync());
CHECK_EQ_OR_RETURN(variable_op_names.size(), variable_tensors.size());
CHECK_OR_RETURN(variable_op_name2eager_blob_.empty());
for (int32_t i = 0; i < variable_op_names.size(); ++i) {
const std::shared_ptr<one::Tensor>& var = variable_tensors.at(i);
CHECK_OR_RETURN(var->is_eager());
Blob* var_blob = nullptr;
if (var->is_consistent()) {
// NOTE(chengcheng): var_blob maybe nullptr when consistent tensor has no cur_rank_phy_tensor.
const std::shared_ptr<one::MirroredTensor> local_var = JUST(var->cur_rank_phy_tensor());
var_blob = JUST(local_var->eager_blob_object())->mut_blob();
} else {
var_blob = JUST(var->eager_blob_object())->mut_blob();
}
const std::string& var_name = variable_op_names.at(i);
CHECK_OR_RETURN(!var_name.empty());
CHECK_OR_RETURN(variable_op_name2eager_blob_.emplace(var_name, var_blob).second);
CHECK_OR_RETURN(variable_op_name2tensor_.emplace(var_name, var).second);
CHECK_OR_RETURN(variable_op_names_.insert(var_name).second);
}
return Maybe<void>::Ok();
Expand All @@ -170,15 +161,8 @@ Maybe<void> NNGraph::RegisterFreeEagerTensorsToVariableOpNames() {
const std::string& var_name = pair.first;
const std::shared_ptr<one::Tensor>& var = pair.second;
CHECK_OR_RETURN(var->is_eager());
Blob* var_blob = nullptr;
if (var->is_consistent()) {
const std::shared_ptr<one::MirroredTensor> local_var = JUST(var->cur_rank_phy_tensor());
var_blob = JUST(local_var->eager_blob_object())->mut_blob();
} else {
var_blob = JUST(var->eager_blob_object())->mut_blob();
}
CHECK_OR_RETURN(!var_name.empty());
CHECK_OR_RETURN(variable_op_name2eager_blob_.emplace(var_name, var_blob).second);
CHECK_OR_RETURN(variable_op_name2tensor_.emplace(var_name, var).second);
CHECK_OR_RETURN(variable_op_names_.insert(var_name).second);
}
return Maybe<void>::Ok();
Expand Down Expand Up @@ -218,10 +202,7 @@ Maybe<void> NNGraph::CreateAndRegisterNewVariableOpInJobPass() {
}
std::shared_ptr<one::Tensor> tensor = JUST(one::functional::ConsistentConstant(
blob_desc.shape(), value, Symbol<DType>(dtype), placement, *sbp_tuple));
JUST(vm::CurrentRankSync());
const std::shared_ptr<one::MirroredTensor> local_var = JUST(tensor->cur_rank_phy_tensor());
Blob* var_blob = JUST(local_var->eager_blob_object())->mut_blob();
CHECK_OR_RETURN(variable_op_name2eager_blob_.emplace(var_name, var_blob).second);
CHECK_OR_RETURN(variable_op_name2tensor_.emplace(var_name, tensor).second);
CHECK_OR_RETURN(variable_op_names_.insert(var_name).second);

// NOTE(chengcheng): just for tensor lifetime hold by session context in graph lifetime valid.
Expand Down Expand Up @@ -292,11 +273,56 @@ Maybe<void> NNGraph::CompileAndInitRuntime() {
PlanUtil::PopulateOpAttribute(&plan_, plan_.job_id2op_attribute_ref_table());

NewRuntimeBuffers();

JUST(DumpToVariable2BlobMap());
runtime_.reset(new Runtime(plan_, variable_op_name2eager_blob_));
runtime_inited_ = true;
return Maybe<void>::Ok();
}

Maybe<void> NNGraph::DumpToVariable2BlobMap() {
CHECK_OR_RETURN(variable_op_name2eager_blob_.empty());
JUST(vm::CurrentRankSync());
for (const std::string& var_name : variable_op_names_) {
auto iter = variable_op_name2tensor_.find(var_name);
CHECK_OR_RETURN(iter != variable_op_name2tensor_.end()) << var_name << " not found.";
std::shared_ptr<one::Tensor> tensor = iter->second;
Blob* var_blob = nullptr;
// Check sbp whether same or not and allow auto_parallel to change sbp.
if (tensor->is_consistent()) {
cfg::NdSbpSignature nd_sbp_signature = cfg::NdSbpSignature(
job_.job_parallel_view_conf().op_name2nd_sbp_signature_conf().at(var_name));
cfg::NdSbp job_nd_sbp = nd_sbp_signature.bn_in_op2nd_sbp().at("out");
if (*JUST(tensor->nd_sbp()) == job_nd_sbp) {
const std::shared_ptr<one::MirroredTensor> local_var = JUST(tensor->cur_rank_phy_tensor());
var_blob = JUST(local_var->eager_blob_object())->mut_blob();
} else {
CHECK_OR_RETURN(job_.job_conf().enable_auto_parallel())
<< "Only change sbp when enable auto_parallel.";
LOG(INFO) << "Variable with name `" << var_name << "` change sbp from "
<< NdSbpParallelToString(*JUST(tensor->nd_sbp())) << " to "
<< NdSbpParallelToString(job_nd_sbp);
std::vector<Symbol<cfg::SbpParallel>> job_sbp_parallels;
for (int i = 0; i < job_nd_sbp.sbp_parallel_size(); ++i) {
job_sbp_parallels.emplace_back(job_nd_sbp.sbp_parallel(i));
}
const auto& new_tensor = JUST(one::functional::ToConsistent(
tensor, JUST(tensor->parallel_desc()), job_sbp_parallels, {}));
JUST(vm::CurrentRankSync());
// Use tensor.set_data inferface and make new TensorImpl instead of the old one.
variable_op_name2tensor_.at(var_name)->set_data(new_tensor);
const std::shared_ptr<one::MirroredTensor> local_var =
JUST(new_tensor->cur_rank_phy_tensor());
var_blob = JUST(local_var->eager_blob_object())->mut_blob();
}
} else {
var_blob = JUST(tensor->eager_blob_object())->mut_blob();
}
CHECK_OR_RETURN(variable_op_name2eager_blob_.emplace(var_name, var_blob).second);
}
return Maybe<void>::Ok();
}

void NNGraph::NewRuntimeBuffers() {
auto* buffer_mgr = Global<BufferMgr<std::shared_ptr<JobInstance>>>::Get();
// NOTE(chengcheng):
Expand Down
2 changes: 2 additions & 0 deletions oneflow/core/framework/nn_graph.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -57,6 +57,7 @@ class NNGraph final : public NNGraphIf {
private:
Maybe<void> RegisterFreeEagerTensorsToVariableOpNames();
Maybe<void> CreateAndRegisterNewVariableOpInJobPass();
Maybe<void> DumpToVariable2BlobMap();

void NewRuntimeBuffers();
void CloseRuntimeBuffers();
Expand All @@ -68,6 +69,7 @@ class NNGraph final : public NNGraphIf {
std::vector<bool> output_tensors_valid_;
std::vector<std::string> inputs_tensor_meta_str_;
std::vector<std::string> outputs_tensor_meta_str_;
HashMap<std::string, std::shared_ptr<one::Tensor>> variable_op_name2tensor_;
HashMap<std::string, Blob*> variable_op_name2eager_blob_;
HashSet<std::string> variable_op_names_;
Job job_;
Expand Down