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

Jump to bottom

[IPU] merge recent changes #42078

Merged
merged 3 commits into from
May 5, 2022
Merged

[IPU] merge recent changes #42078

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
96ea994
merge recent changes
gglin001 Apr 21, 2022
73e5a2c
fix setting pipline
gglin001 Apr 21, 2022
496f7a2
Merge branch 'develop' into ipu_commit/misc_changes
gglin001 May 4, 2022
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
192 changes: 69 additions & 123 deletions paddle/fluid/platform/device/ipu/ipu_compiler.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -185,12 +185,9 @@ void Compiler::RegisterOpFunc() {
auto debug_context = BuildDebugContext(op_desc); \
auto aiGraphcoreOpset = builder_->aiGraphcoreOpset1(); \
auto aiOnnxOpset = builder_->aiOnnxOpset11(); \
PushNameScope(op_desc); \
NameScopeHelper ns_helper(op_desc, builder_.get()); \
auto output_ids = OnnxImpl(inputs Args, debug_context); \
PopNameScope(op_desc); \
SetIpuIndexStage(output_ids, op_desc); \
SetAMPAttributes(output_ids, op_desc); \
SetSerializeAttributes(output_ids, op_desc); \
PostLower(output_ids, op_desc); \
InsertTensors(output_names, output_ids); \
}}, // NOLINT
#include "paddle/fluid/platform/device/ipu/supported_ops_autogen.h"
Expand Down Expand Up @@ -273,10 +270,9 @@ void Compiler::LowerConstants(const Scope* scope) {
popart::TensorInfo tensor_info(PdDataType2PopartType(tensor->dtype()),
shape);
const_data.reset(new popart::ConstVoidData(tensor->data(), tensor_info));
PushNameScope(op_desc);
NameScopeHelper ns_helper(op_desc, builder_.get());
popart::TensorId result = builder_->aiOnnxOpset11().constant(*const_data);
PopNameScope(op_desc);
SetIpuIndexStage(result, op_desc);
PostLower(result, op_desc);
resources_->tensors.emplace(tensor_name, result);
}
}
Expand All @@ -285,42 +281,42 @@ void Compiler::LowerConstants(const Scope* scope) {

void Compiler::LowerWeights(const Scope* scope) {
VLOG(10) << "enter Compiler::LowerWeights";
// at this step, the graph doesn't contains optimizer related states
// At this step, the graph doesn't contains optimizer related states
for (auto id : graph_helper_->sorted_vars_id) {
auto* node = graph_helper_->nodes_id_map[id];
if (node->IsVar() && !node->IsCtrlVar() && node->Var()) {
if (node->Var()->Persistable() && node->inputs.empty()) {
auto var_name = node->Var()->Name();
if (resources_->tensors.count(var_name) != 0) {
VLOG(10) << "found existed one, skip lowering Weight: " << var_name;
continue;
}
if (var_name.rfind("learning_rate", 0) == 0) {
VLOG(10) << "skip learning_rate_var: " << var_name;
continue;
}
VLOG(10) << "lowering weight: " << var_name;

auto var = scope->FindVar(var_name);
if (var) {
auto tensor = var->Get<framework::LoDTensor>();
auto dtype = PdDataType2PopartType(tensor.dtype());
auto shape = std::vector<int64_t>();
for (size_t i = 0; i < tensor.dims().size(); ++i) {
shape.push_back(tensor.dims().at(i));
}
popart::TensorInfo tensor_info(dtype, shape);
popart::ConstVoidData const_data{tensor.data(), tensor_info};
if (!node->outputs.empty()) {
auto op_node = node->outputs[0];
PushNameScope(op_node->Op());
popart::TensorId result =
builder_->addInitializedInputTensor(const_data, var_name);
PopNameScope(op_node->Op());
resources_->tensors.emplace(var_name, result);
resources_->weights.push_back(var_name);
}
}
// Weights are var node and Persistable
if (node->IsVar() && !node->IsCtrlVar() && node->Var() &&
node->Var()->Persistable()) {
// Weights are Parameter in training mode
if (ipu_strategy_->is_training && !node->Var()->IsParameter()) {
continue;
}
auto var_name = node->Var()->Name();
// Some op has same input and output tensor, like batchnorm
if (resources_->tensors.count(var_name) != 0) {
VLOG(10) << "found existed one, skip lowering Weight: " << var_name;
continue;
}
VLOG(10) << "lowering weight: " << var_name;
auto var = scope->FindVar(var_name);
PADDLE_ENFORCE_NOT_NULL(
var, platform::errors::NotFound("Tensor %s is not found in the scope",
var_name));
auto tensor = var->Get<framework::LoDTensor>();
auto dtype = PdDataType2PopartType(tensor.dtype());
auto shape = std::vector<int64_t>();
for (size_t i = 0; i < tensor.dims().size(); ++i) {
shape.push_back(tensor.dims().at(i));
}
popart::TensorInfo tensor_info(dtype, shape);
popart::ConstVoidData const_data{tensor.data(), tensor_info};
if (!node->outputs.empty()) {
auto op_node = node->outputs[0];
NameScopeHelper ns_helper(op_node->Op(), builder_.get());
popart::TensorId result =
builder_->addInitializedInputTensor(const_data, var_name);
resources_->tensors.emplace(var_name, result);
resources_->weights.push_back(var_name);
}
}
}
Expand All @@ -341,10 +337,9 @@ void Compiler::LowerBody() {
} else if (op_type == "popart_checkpointoutput") {
auto inputs = GetOpInputs(op_desc);
auto outputs = GetOpOutputs(op_desc);
PushNameScope(op_desc);
NameScopeHelper ns_helper(op_desc, builder_.get());
auto output_ids = builder_->checkpointOutput(inputs);
PopNameScope(op_desc);
SetIpuIndexStage(output_ids, op_desc);
PostLower(output_ids, op_desc);
InsertTensors(outputs, output_ids);
} else if (op_type == "popart_custom_op") {
auto inputs = GetOpInputs(op_desc);
Expand All @@ -359,12 +354,11 @@ void Compiler::LowerBody() {
BOOST_GET_CONST(std::string, op_desc->GetAttr("__op_type"));
VLOG(10) << "Build graph from custom op: " << __op_type;
auto it = custom_ops_.find(__op_type);
PushNameScope(op_desc);
NameScopeHelper ns_helper(op_desc, builder_.get());
auto output_ids =
builder_->customOp(it->second.popart_op, it->second.popart_op.version,
inputs, outputs.size(), attributes, debug_context);
PopNameScope(op_desc);
SetIpuIndexStage(output_ids, op_desc);
PostLower(output_ids, op_desc);
InsertTensors(outputs, output_ids);
} else if (op_type == "popart_printtensor") {
auto inputs = GetOpInputs(op_desc);
Expand All @@ -373,11 +367,10 @@ void Compiler::LowerBody() {
auto print_gradient =
BOOST_GET_CONST(int64_t, op_desc->GetAttr("print_gradient"));
auto title = BOOST_GET_CONST(std::string, op_desc->GetAttr("title"));
PushNameScope(op_desc);
NameScopeHelper ns_helper(op_desc, builder_.get());
auto output_ids = builder_->aiGraphcoreOpset1().printtensor(
inputs, print_gradient, debug_context, title);
PopNameScope(op_desc);
SetIpuIndexStage(output_ids, op_desc);
PostLower(output_ids, op_desc);
InsertTensors(outputs, output_ids);
} else {
auto itr = name_function_.find(op_type);
Expand Down Expand Up @@ -625,12 +618,13 @@ void Compiler::InsertTensors(const std::vector<std::string>& output_names,
resources_->tensors.emplace(output_names[0], tensor_id);
}

void Compiler::SetIpuIndexStage(const std::vector<std::string>& tensor_ids,
const OpDesc* op_desc) {
VLOG(10) << "enter Compiler::SetIpuIndexStage";
void Compiler::PostLower(const std::vector<std::string>& tensor_ids,
const OpDesc* op_desc) {
// Set pipline
// Due to the limitation of popart, if an op has multiple outputs,
// pipline settings needs to be set at the same time
auto tensor_ids_set =
std::set<std::string>(tensor_ids.begin(), tensor_ids.end());

if (op_desc->HasAttr(sIpuIndexAttr)) {
auto ipu_index = BOOST_GET_CONST(int, op_desc->GetAttr(sIpuIndexAttr));
builder_->virtualGraph(tensor_ids_set, ipu_index);
Expand All @@ -639,51 +633,43 @@ void Compiler::SetIpuIndexStage(const std::vector<std::string>& tensor_ids,
if (op_desc->HasAttr(sIpuStageAttr)) {
auto ipu_stage = BOOST_GET_CONST(int, op_desc->GetAttr(sIpuStageAttr));
builder_->pipelineStage(tensor_ids_set, ipu_stage);
VLOG(10) << "set " << sIpuStageAttr << "= " << ipu_stage
VLOG(10) << "set " << sIpuStageAttr << " = " << ipu_stage
<< " for op: " << op_desc->Type();
}
}
VLOG(10) << "leave Compiler::SetIpuIndexStage";

for (auto& tensor_id : tensor_ids) {
PostLower(tensor_id, op_desc, true);
}
}

void Compiler::SetIpuIndexStage(const std::string& tensor_id,
const OpDesc* op_desc) {
VLOG(10) << "enter Compiler::SetIpuIndexStage";
void Compiler::PostLower(const std::string& tensor_id, const OpDesc* op_desc) {
PostLower(tensor_id, op_desc, false);
}

if (op_desc->HasAttr(sIpuIndexAttr)) {
void Compiler::PostLower(const std::string& tensor_id, const OpDesc* op_desc,
bool skip_pipline) {
// Set pipline
if (!skip_pipline && op_desc->HasAttr(sIpuIndexAttr)) {
auto ipu_index = BOOST_GET_CONST(int, op_desc->GetAttr(sIpuIndexAttr));
builder_->virtualGraph(tensor_id, ipu_index);
VLOG(10) << "set " << sIpuIndexAttr << " = " << ipu_index
<< " for op: " << op_desc->Type();
if (op_desc->HasAttr(sIpuStageAttr)) {
auto ipu_stage = BOOST_GET_CONST(int, op_desc->GetAttr(sIpuStageAttr));
builder_->pipelineStage(tensor_id, ipu_stage);
VLOG(10) << "set " << sIpuStageAttr << "= " << ipu_stage
VLOG(10) << "set " << sIpuStageAttr << " = " << ipu_stage
<< " for op: " << op_desc->Type();
}
}
VLOG(10) << "leave Compiler::SetIpuIndexStage";
}

void Compiler::SetAMPAttributes(const std::vector<std::string>& tensor_ids,
const OpDesc* op_desc) {
if (op_desc->Type() == "popart_matmul") {
for (const auto& tensor_id : tensor_ids) {
SetAMPAttributes(tensor_id, op_desc);
}
}
}

void Compiler::SetAMPAttributes(const std::string& tensor_id,
const OpDesc* op_desc) {
VLOG(10) << "enter Compiler::SetAMPAttributes";
// Set amp
if (op_desc->Type() == "popart_matmul") {
if (set_amp_for_all_) {
auto amp = ipu_strategy_->available_memory_proportion;
if (amp < 0.0f || amp > 1.0) {
PADDLE_THROW(platform::errors::InvalidArgument(
"AvailableMemoryProportion %f is invalid, which should be set 0 <= "
"amp <= 1",
"AvailableMemoryProportion %f is invalid, which should be in "
"range [0.0, 1.0]",
amp));
}
if (amp > 0.0f) {
Expand All @@ -694,8 +680,8 @@ void Compiler::SetAMPAttributes(const std::string& tensor_id,
auto amp = BOOST_GET_CONST(float, op_desc->GetAttr(sAvailMemAttribute));
if (amp < 0.0f || amp > 1.0) {
PADDLE_THROW(platform::errors::InvalidArgument(
"AvailableMemoryProportion %f is invalid, which should be set 0 "
"<= amp <= 1",
"AvailableMemoryProportion %f is invalid, which should be in "
"range [0.0, 1.0]",
amp));
}
if (amp > 0.0f) {
Expand All @@ -705,17 +691,7 @@ void Compiler::SetAMPAttributes(const std::string& tensor_id,
}
}
}
}
VLOG(10) << "leave Compiler::SetAMPAttributes";
}

void Compiler::SetSerializeAttributes(
const std::vector<std::string>& tensor_ids, const OpDesc* op_desc) {
VLOG(10) << "enter Compiler::SetSerializeAttributes";
auto tensor_ids_set =
std::set<std::string>(tensor_ids.begin(), tensor_ids.end());

if (op_desc->Type() == "popart_matmul") {
// Set serialize matmul
if (op_desc->HasAttr(sMatmulSerializeFactor)) {
auto factor =
BOOST_GET_CONST(int, op_desc->GetAttr(sMatmulSerializeFactor));
Expand All @@ -724,16 +700,9 @@ void Compiler::SetSerializeAttributes(
mode = BOOST_GET_CONST(std::string,
op_desc->GetAttr(sMatmulSerializeMode));
}
builder_->setSerializeMatMul(tensor_ids_set, mode, (int64_t)factor, true);
builder_->setSerializeMatMul({tensor_id}, mode, factor, true);
}
}
VLOG(10) << "leave Compiler::SetSerializeAttributes";
}

void Compiler::SetSerializeAttributes(const std::string& tensor_id,
const OpDesc* op_desc) {
std::vector<std::string> tensor_ids = {tensor_id};
SetSerializeAttributes(tensor_ids, op_desc);
}

void Compiler::SetCustomOps(
Expand Down Expand Up @@ -793,29 +762,6 @@ popart::DebugContext Compiler::BuildDebugContext(const OpDesc* op) {
return popart::DebugContext(op_identify_id);
}

void Compiler::PushNameScope(const OpDesc* op) {
auto op_namescope = BOOST_GET_CONST(std::string, op->GetAttr(sOpNamescope));
if (op_namescope == "/") {
return;
}
if (!op_namescope.empty()) {
op_namescope.pop_back();
}
if (!op_namescope.empty()) {
op_namescope.erase(op_namescope.begin());
}
VLOG(10) << "name_scope is: " << op_namescope;
builder_->pushNameScope(op_namescope);
}

void Compiler::PopNameScope(const OpDesc* op) {
auto op_namescope = BOOST_GET_CONST(std::string, op->GetAttr(sOpNamescope));
if (op_namescope == "/") {
return;
}
builder_->popNameScope();
}

} // namespace ipu
} // namespace platform
} // namespace paddle
41 changes: 28 additions & 13 deletions paddle/fluid/platform/device/ipu/ipu_compiler.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -70,7 +70,7 @@ struct CompilerResources {
std::unique_ptr<popart::Optimizer> optimizer;
};

// helper for lowering graph
// Helper for lowering graph
struct GraphHelper {
explicit GraphHelper(const Graph *);

Expand All @@ -81,6 +81,30 @@ struct GraphHelper {
std::vector<int> sorted_vars_id;
};

// Helper for adding namescope info
struct NameScopeHelper {
NameScopeHelper(const OpDesc *op, popart::Builder *builder)
: builder_(builder) {
auto op_namescope = BOOST_GET_CONST(std::string, op->GetAttr(sOpNamescope));
if (op_namescope.empty() || op_namescope == "/") {
return;
}
op_namescope.pop_back();
op_namescope.erase(op_namescope.begin());
builder->pushNameScope(op_namescope);
pushed_ = true;
}

~NameScopeHelper() {
if (pushed_) {
builder_->popNameScope();
}
}

bool pushed_ = false;
popart::Builder *builder_;
};

class Compiler {
public:
Compiler();
Expand Down Expand Up @@ -119,18 +143,9 @@ class Compiler {
const std::vector<std::string> &tensor_ids);
void InsertTensors(const std::vector<std::string> &output_names,
const std::string &tensor_id);
void SetIpuIndexStage(const std::vector<std::string> &tensor_ids,
const OpDesc *op_desc);
void SetIpuIndexStage(const std::string &tensor_id, const OpDesc *op_desc);
void SetAMPAttributes(const std::vector<std::string> &tensor_ids,
const OpDesc *op_desc);
void SetAMPAttributes(const std::string &tensor_id, const OpDesc *op_desc);
void SetSerializeAttributes(const std::vector<std::string> &tensor_ids,
const OpDesc *op_desc);
void SetSerializeAttributes(const std::string &tensor_id,
const OpDesc *op_desc);
void PushNameScope(const OpDesc *op);
void PopNameScope(const OpDesc *op);
void PostLower(const std::vector<std::string> &, const OpDesc *);
void PostLower(const std::string &, const OpDesc *);
void PostLower(const std::string &, const OpDesc *, bool);

private:
std::unique_ptr<popart::Builder> builder_;
Expand Down
Loading