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

Refine InferShape for recurrent_network_op #3124

Merged
merged 8 commits into from
Aug 3, 2017
Merged

Refine InferShape for recurrent_network_op #3124

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
6cfb9a3
Refine InferShape for recurrent_network_op.
qingqing01 Jul 31, 2017
b89d15a
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into…
qingqing01 Jul 31, 2017
8925295
follow comments.
qingqing01 Aug 1, 2017
79e89ef
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into…
qingqing01 Aug 1, 2017
b94584c
Rename recurrent_network_op recurrent_op.
qingqing01 Aug 1, 2017
18aa0dc
resolve conflicts.
qingqing01 Aug 1, 2017
99b835c
resolve conficts.
qingqing01 Aug 2, 2017
7c49a4f
resovle conflicts.
qingqing01 Aug 2, 2017
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
147 changes: 60 additions & 87 deletions paddle/operators/recurrent_network_op.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,7 +29,8 @@ namespace rnn {

void SegmentInputs(std::vector<std::shared_ptr<Scope>>& step_scopes,
const std::vector<Link>& inlinks,
const size_t seq_len) {
const size_t seq_len,
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

const size_t&

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

const size_t is ok.

bool infer_shape) {
PADDLE_ENFORCE(!inlinks.empty(), "no in links are provided.");
for (size_t i = 0; i < inlinks.size(); ++i) {
Tensor* input =
Expand All @@ -42,28 +43,33 @@ void SegmentInputs(std::vector<std::shared_ptr<Scope>>& step_scopes,
Tensor* step_input = step_scopes[j]
->CreateVariable(inlinks[i].internal)
->GetMutable<Tensor>();
*step_input = input->Slice<float>(j, j + 1);
if (!infer_shape) {
*step_input = input->Slice<float>(j, j + 1);
}
step_input->Resize(step_dims);
}
}
}

void ConcatOutputs(std::vector<std::shared_ptr<Scope>>& step_scopes,
const std::vector<Link>& outlinks,
const size_t seq_len) {
const size_t seq_len,
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

const size_t&

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

const size_t is ok.

bool infer_shape) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

is_infer or infer_mode?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

infer_shape_mode

for (size_t i = 0; i < outlinks.size(); i++) {
Tensor* output =
step_scopes[0]->GetVariable(outlinks[i].external)->GetMutable<Tensor>();

// TODO(qingiqng) remove following code after adding
// InferShape in RecurrentGradientOp
DDim step_dims = step_scopes[0]
->GetVariable(outlinks[i].internal)
->GetMutable<Tensor>()
->dims();
std::vector<int> dims_vec = vectorize(step_dims);
dims_vec.insert(dims_vec.begin(), seq_len);
output->mutable_data<float>(make_ddim(dims_vec), platform::CPUPlace());
if (infer_shape) {
DDim step_dims = step_scopes[0]
->GetVariable(outlinks[i].internal)
->GetMutable<Tensor>()
->dims();
std::vector<int> dims_vec = vectorize(step_dims);
dims_vec.insert(dims_vec.begin(), seq_len);
output->Resize(make_ddim(dims_vec));
} else {
output->mutable_data<float>(platform::CPUPlace());
}

for (size_t j = 0; j < seq_len; j++) {
Tensor* step_output = step_scopes[j]
Expand All @@ -79,8 +85,9 @@ void ConcatOutputs(std::vector<std::shared_ptr<Scope>>& step_scopes,

void LinkMemories(std::vector<std::shared_ptr<Scope>>& scopes,
const std::vector<rnn::MemoryAttr>& memories,
size_t step_id,
int offset) {
const size_t step_id,
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

const size_t&

const int offset,
bool infer_shape) {
PADDLE_ENFORCE(step_id < scopes.size(),
"step [%d] is out of range of step scopes' size [%d]",
step_id,
Expand All @@ -97,18 +104,14 @@ void LinkMemories(std::vector<std::shared_ptr<Scope>>& scopes,
std::shared_ptr<Scope> scope = scopes[step_id];
std::shared_ptr<Scope> linked_scope = scopes[step_id + offset];
for (auto& attr : memories) {
auto mem = scope->CreateVariable(attr.pre_var)->GetMutable<Tensor>();
auto mem = scope->GetVariable(attr.pre_var)->GetMutable<Tensor>();
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

need an enforce on scope->GetVariable(xxx) != nullptr ?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done.

// maybe share variable is better?
auto linked_mem = linked_scope->GetVariable(attr.var)->GetMutable<Tensor>();
mem->ShareDataWith<float>(*linked_mem);

// TODO(qingqing) remove following code
// the memory of current step should be allocated in step net
auto m = scope->CreateVariable(attr.var)->GetMutable<Tensor>();
// for unit test, as addOp and mulOp are null currently, if not
// mutable_data, mem.data() in output will be error. We will
// remove this line after merge the correct addOp and mulOp.
m->mutable_data<float>(mem->dims(), platform::CPUPlace());
if (infer_shape) {
mem->Resize(linked_mem->dims());
} else {
mem->ShareDataWith<float>(*linked_mem);
}
}
}

Expand Down Expand Up @@ -176,61 +179,43 @@ void RecurrentAlgorithm::InferShape(const std::shared_ptr<Scope>& scope) const {
->GetMutable<Tensor>()
->dims()[0];
CreateScopes(scope);
auto step_scopes = GetStepScopes(scope);

// SegmentInputs is called in InferShape. The input must hold memory in
// SegmentInputs. But the other op only set dimension for the output in
// InferShape. That's a problem. Wether the RNN op needs InferShape or not?
// Wether the following functions (SegmentInputs, InitMemories, ...) need
// to rewrite for RNN op?
rnn::SegmentInputs(step_scopes, arg_->inlinks, seq_len_);
auto step_scopes = GetStepScopes(scope);
rnn::SegmentInputs(step_scopes, arg_->inlinks, seq_len_, true);

InitMemories(step_scopes[0]);
InitMemories(step_scopes[0], true);

PADDLE_ENFORCE(scope->HasVariable(arg_->step_net),
"stepnet [%s] is not in scope.",
arg_->step_net);
Variable* net = scope->GetVariable(arg_->step_net);
PADDLE_ENFORCE(net != nullptr, "failed to get step net");
// If the InferShape is called in OperatorBase's run function,
// the rnn op only needs to do InferShape for the first time step
for (size_t i = 0; i < seq_len_; i++) {
if (i > 0) {
rnn::LinkMemories(step_scopes, arg_->memories, i, -1);
rnn::LinkMemories(step_scopes, arg_->memories, i, -1, true);
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

true /* infer_mode */

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done.

}
net->GetMutable<NetOp>()->InferShape(step_scopes[i]);
}

auto outlinks = arg_->outlinks;
for (size_t i = 0; i < outlinks.size(); i++) {
DDim step_dims = step_scopes[0]
->GetVariable(outlinks[i].internal)
->GetMutable<Tensor>()
->dims();
std::vector<int> dims_vec = vectorize(step_dims);
// now only support fixed length
dims_vec.insert(dims_vec.begin(), seq_len_);
Tensor* output =
step_scopes[0]->GetVariable(outlinks[i].external)->GetMutable<Tensor>();
output->Resize(make_ddim(dims_vec));
}
rnn::ConcatOutputs(step_scopes, arg_->outlinks, seq_len_, true);
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

same as top

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done.

}

void RecurrentAlgorithm::Run(const std::shared_ptr<Scope>& scope,
const platform::DeviceContext& dev_ctx) const {
auto step_scopes = GetStepScopes(scope);

rnn::SegmentInputs(step_scopes, arg_->inlinks, seq_len_, false);

InitMemories(step_scopes[0], false);

Variable* net = scope->GetVariable(arg_->step_net);
for (size_t step_id = 0; step_id < seq_len_; step_id++) {
// the link memory is done in InferShape
// maybe remove following code after testing
if (step_id > 0) {
rnn::LinkMemories(step_scopes, arg_->memories, step_id, -1);
rnn::LinkMemories(step_scopes, arg_->memories, step_id, -1, false);
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

same as the top

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done.

}
net->GetMutable<NetOp>()->Run(step_scopes[step_id], dev_ctx);
}

rnn::ConcatOutputs(step_scopes, arg_->outlinks, seq_len_);
rnn::ConcatOutputs(step_scopes, arg_->outlinks, seq_len_, false);
}

void RecurrentAlgorithm::CreateScopes(std::shared_ptr<Scope> scope) const {
Expand All @@ -246,6 +231,7 @@ void RecurrentAlgorithm::CreateScopes(std::shared_ptr<Scope> scope) const {
// Now all variables in scope must be created outside of op.
auto net_op = scope->GetVariable(arg_->step_net)->GetMutable<NetOp>();
for (auto& input : net_op->inputs_) {
// the weight are located in parent scope
step_scope->CreateVariable(input);
}
for (auto& output : net_op->outputs_) {
Expand All @@ -257,7 +243,8 @@ void RecurrentAlgorithm::CreateScopes(std::shared_ptr<Scope> scope) const {
}
}

void RecurrentAlgorithm::InitMemories(std::shared_ptr<Scope> step_scope) const {
void RecurrentAlgorithm::InitMemories(std::shared_ptr<Scope> step_scope,
bool infer_shape) const {
for (auto& attr : arg_->memories) {
Tensor* pre_mem =
step_scope->CreateVariable(attr.pre_var)->GetMutable<Tensor>();
Expand All @@ -267,14 +254,11 @@ void RecurrentAlgorithm::InitMemories(std::shared_ptr<Scope> step_scope) const {
attr.boot_var);
Tensor* boot_mem =
step_scope->GetVariable(attr.boot_var)->GetMutable<Tensor>();
pre_mem->ShareDataWith<float>(*boot_mem);

// TODO(qingqing) remove following code
// the memory of current step should be allocated in step net
// here for unit test
auto cur_step_mem =
step_scope->CreateVariable(attr.var)->GetMutable<Tensor>();
cur_step_mem->mutable_data<float>(boot_mem->dims(), platform::CPUPlace());
if (infer_shape) {
pre_mem->Resize(boot_mem->dims());
} else {
pre_mem->ShareDataWith<float>(*boot_mem);
}
}
}

Expand Down Expand Up @@ -336,35 +320,37 @@ void RecurrentGradientAlgorithm::Run(
const std::shared_ptr<Scope>& scope,
const platform::DeviceContext& dev_ctx) const {
auto step_scopes = GetStepScopes(scope);
rnn::SegmentInputs(step_scopes, arg_->inlinks, seq_len_);
rnn::SegmentInputs(step_scopes, arg_->inlinks, seq_len_, false);
PADDLE_ENFORCE(scope->HasVariable(arg_->step_net),
"step net is not in scope.");
Variable* net = scope->GetVariable(arg_->step_net);
PADDLE_ENFORCE(net != nullptr, "failed to get step net");
for (int step_id = seq_len_ - 1; step_id >= 0; --step_id) {
if (static_cast<size_t>(step_id) != seq_len_ - 1) {
rnn::LinkMemories(step_scopes, arg_->memories, step_id, 1);
rnn::LinkMemories(step_scopes, arg_->memories, step_id, 1, false);
}
net->GetMutable<NetOp>()->Run(step_scopes[step_id], dev_ctx);
}
LinkBootMemoryGradients(step_scopes[0]);
rnn::ConcatOutputs(step_scopes, arg_->outlinks, seq_len_);
LinkBootMemoryGradients(step_scopes[0], false);
rnn::ConcatOutputs(step_scopes, arg_->outlinks, seq_len_, false);
}

void RecurrentGradientAlgorithm::LinkBootMemoryGradients(
std::shared_ptr<Scope> step_scope) const {
std::shared_ptr<Scope> step_scope, bool infer_shape) const {
for (auto& attr : arg_->memories) {
Tensor* mem_grad =
step_scope->CreateVariable(attr.var)->GetMutable<Tensor>();
PADDLE_ENFORCE(mem_grad != nullptr,
"boot_tensor should be retrieved before");
PADDLE_ENFORCE(step_scope->HasVariable(attr.boot_var),
"memory [%s]'s boot variable [%s] not exists",
attr.var,
attr.boot_var);
Tensor* boot_mem_grad =
step_scope->CreateVariable(attr.boot_var)->GetMutable<Tensor>();
boot_mem_grad->ShareDataWith<float>(*mem_grad);
if (infer_shape) {
boot_mem_grad->Resize(mem_grad->dims());
} else {
boot_mem_grad->ShareDataWith<float>(*mem_grad);
}
}
}

Expand All @@ -374,7 +360,7 @@ void RecurrentGradientAlgorithm::InferShape(
->GetMutable<Tensor>()
->dims()[0];
auto step_scopes = GetStepScopes(scope);
rnn::SegmentInputs(step_scopes, arg_->inlinks, seq_len_);
rnn::SegmentInputs(step_scopes, arg_->inlinks, seq_len_, true);

PADDLE_ENFORCE(scope->HasVariable(arg_->step_net),
"step net is not in scope.");
Expand All @@ -383,25 +369,12 @@ void RecurrentGradientAlgorithm::InferShape(

for (int step_id = seq_len_ - 1; step_id >= 0; --step_id) {
if (static_cast<size_t>(step_id) != seq_len_ - 1) {
rnn::LinkMemories(step_scopes, arg_->memories, step_id, 1);
rnn::LinkMemories(step_scopes, arg_->memories, step_id, 1, true);
}
net->GetMutable<NetOp>()->InferShape(step_scopes[step_id]);
}

auto outlinks = arg_->outlinks;
for (size_t i = 0; i < outlinks.size(); i++) {
DDim step_dims = step_scopes[0]
->GetVariable(outlinks[i].internal)
->GetMutable<Tensor>()
->dims();
std::vector<int> dims_vec = vectorize(step_dims);
// now only support fixed length
dims_vec.insert(dims_vec.begin(), seq_len_);
Tensor* output =
step_scopes[0]->GetVariable(outlinks[i].external)->GetMutable<Tensor>();
output->Resize(make_ddim(dims_vec));
}
LinkBootMemoryGradients(step_scopes[0]);
rnn::ConcatOutputs(step_scopes, arg_->outlinks, seq_len_, true);
LinkBootMemoryGradients(step_scopes[0], true);
}

void RecurrentGradientOp::Init() {
Expand Down
16 changes: 10 additions & 6 deletions paddle/operators/recurrent_network_op.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,19 +72,22 @@ struct ArgumentName {
*/
void SegmentInputs(std::vector<std::shared_ptr<Scope>>& step_scopes,
const std::vector<Link>& inlinks,
const size_t seq_len);
const size_t seq_len,
bool infer_shape);

/**
* Process outputs of step nets and merge to variables.
*/
void ConcatOutputs(std::vector<std::shared_ptr<Scope>>& step_scopes,
const std::vector<Link>& outlinks,
const size_t seq_len);
const size_t seq_len,
bool infer_shape);

void LinkMemories(std::vector<std::shared_ptr<Scope>>& step_scopes,
const std::vector<MemoryAttr>& memories,
size_t step_id,
int offset);
const size_t step_id,
const int offset,
bool infer_shape);

void InitArgument(const ArgumentName& name, Argument* arg);

Expand Down Expand Up @@ -125,7 +128,7 @@ class RecurrentAlgorithm {
->GetMutable<std::vector<std::shared_ptr<Scope>>>();
}

void InitMemories(std::shared_ptr<Scope> step_scopes) const;
void InitMemories(std::shared_ptr<Scope> step_scopes, bool infer_shape) const;

private:
std::unique_ptr<rnn::Argument> arg_;
Expand All @@ -149,7 +152,8 @@ class RecurrentGradientAlgorithm {
void Run(const std::shared_ptr<Scope>& scope,
const platform::DeviceContext& dev_ctx) const;

void LinkBootMemoryGradients(std::shared_ptr<Scope> step_scopes) const;
void LinkBootMemoryGradients(std::shared_ptr<Scope> step_scopes,
bool infer_shape) const;

/**
* InferShape must be called before Run.
Expand Down
Loading