Skip to content

P
Paddle

PaddlePaddle / Paddle
大约 1 年 前同步成功

通知 2299
Star 20931
Fork 5422
P
Paddle
未验证 提交 1d3b0134 编写于 作者: H Huihuang Zheng 提交者： GitHub
浏览文件
操作

Error Message Enhancement (#23483) 

This PR enhances error messages of several API/OPs:

ParallelExecutor (python && C++)
Executor (python && C++)
StaticRNN (python)
IfElse (python)
cond (python)
split_lod_tensor (python && C++)
上级 6f7077db
隐藏空白更改
内联 并排
Showing with 267 addition and 115 deletion
paddle/fluid/framework/executor.cc
浏览文件 @ 1d3b0134
...@@ -21,6 +21,7 @@ limitations under the License. */ ...@@ -21,6 +21,7 @@ limitations under the License. */
#include "google/protobuf/io/zero_copy_stream_impl.h" #include "google/protobuf/io/zero_copy_stream_impl.h"
#include "google/protobuf/message.h" #include "google/protobuf/message.h"
#include "google/protobuf/text_format.h" #include "google/protobuf/text_format.h"
#include "paddle/fluid/framework/data_type.h"
#include "paddle/fluid/framework/feed_fetch_method.h" #include "paddle/fluid/framework/feed_fetch_method.h"
#include "paddle/fluid/framework/lod_rank_table.h" #include "paddle/fluid/framework/lod_rank_table.h"
#include "paddle/fluid/framework/lod_tensor_array.h" #include "paddle/fluid/framework/lod_tensor_array.h"
...@@ -145,8 +146,10 @@ std::shared_ptr<TrainerBase> Executor::InitForDataset( ...@@ -145,8 +146,10 @@ std::shared_ptr<TrainerBase> Executor::InitForDataset(
VLOG(3) << "Start to RunFromDataset in executor"; VLOG(3) << "Start to RunFromDataset in executor";
TrainerDesc trainer_desc; TrainerDesc trainer_desc;
bool success = trainer_desc.ParseFromString(trainer_desc_str); bool success = trainer_desc.ParseFromString(trainer_desc_str);
PADDLE_ENFORCE_EQ(success, true, "Fail to parse TrainerDesc from string:\n%s", PADDLE_ENFORCE_EQ(success, true,
trainer_desc_str.c_str()); platform::errors::PreconditionNotMet(
"Fail to parse TrainerDesc from string:\n%s",
trainer_desc_str.c_str()));
VLOG(3) << "Going to create trainer, trainer class is " VLOG(3) << "Going to create trainer, trainer class is "
<< trainer_desc.class_name(); << trainer_desc.class_name();
std::shared_ptr<TrainerBase> trainer; std::shared_ptr<TrainerBase> trainer;
...@@ -165,8 +168,9 @@ std::shared_ptr<TrainerBase> Executor::InitForDataset( ...@@ -165,8 +168,9 @@ std::shared_ptr<TrainerBase> Executor::InitForDataset(
} }
void Executor::RunFromDataset(std::shared_ptr<TrainerBase> trainer) { void Executor::RunFromDataset(std::shared_ptr<TrainerBase> trainer) {
PADDLE_ENFORCE_NE(trainer, nullptr, PADDLE_ENFORCE_NOT_NULL(
"Trainer is nullptr, invoke InitForDataset first"); trainer, platform::errors::InvalidArgument(
"Trainer is nullptr, invoke InitForDataset first"));
// training and finalize training // training and finalize training
VLOG(3) << "Trainer starts to run"; VLOG(3) << "Trainer starts to run";
trainer->Run(); trainer->Run();
...@@ -203,29 +207,41 @@ static bool has_feed_operators( ...@@ -203,29 +207,41 @@ static bool has_feed_operators(
if (op->Type() == kFeedOpType) { if (op->Type() == kFeedOpType) {
feed_count++; feed_count++;
// The input variable's name of feed_op should be feed_holder_name. // The input variable's name of feed_op should be feed_holder_name.
PADDLE_ENFORCE_EQ(op->Input("X")[0], feed_holder_name, PADDLE_ENFORCE_EQ(
"Input to feed op should be '%s'", feed_holder_name); op->Input("X")[0], feed_holder_name,
platform::errors::PreconditionNotMet(
"Input to feed op should be '%s', but received '%s'.",
feed_holder_name, op->Input("X")[0]));
std::string feed_target_name = op->Output("Out")[0]; std::string feed_target_name = op->Output("Out")[0];
PADDLE_ENFORCE( PADDLE_ENFORCE_NE(feed_targets.find(feed_target_name), feed_targets.end(),
feed_targets.find(feed_target_name) != feed_targets.end(), platform::errors::PreconditionNotMet(
"Feed operator output name '%s' cannot be found in 'feed_targets'", "Feed operator output name '%s' cannot be found in "
feed_target_name); "'feed_targets'",
feed_target_name));
} }
} }
if (feed_count > 0) { if (feed_count > 0) {
PADDLE_ENFORCE_EQ( PADDLE_ENFORCE_EQ(
feed_count, feed_targets.size(), feed_count, feed_targets.size(),
"The number of feed operators should match 'feed_targets'"); platform::errors::PreconditionNotMet(
"The number of feed operators should match 'feed_targets', but "
"received feed_count: %zu, required feed_targets.size(): %zu.",
feed_count, feed_targets.size()));
if (!feed_holder_name.empty()) { if (!feed_holder_name.empty()) {
// When feed operator are present, so should be feed_holder. // When feed operator are present, so should be feed_holder.
auto var = block.FindVar(feed_holder_name); auto var = block.FindVar(feed_holder_name);
PADDLE_ENFORCE_NOT_NULL(var, "Block should already have a '%s' variable", PADDLE_ENFORCE_NOT_NULL(
feed_holder_name); var,
PADDLE_ENFORCE_EQ(var->GetType(), proto::VarType::FEED_MINIBATCH, platform::errors::PreconditionNotMet(
"'%s' variable should be 'FEED_MINIBATCH' type", "Block should already have a '%s' variable", feed_holder_name));
feed_holder_name); PADDLE_ENFORCE_EQ(
var->GetType(), proto::VarType::FEED_MINIBATCH,
platform::errors::PreconditionNotMet(
"'%s' variable should be 'FEED_MINIBATCH' type, but received "
"'%s'.",
feed_holder_name, DataTypeToString(var->GetType())));
} }
} }
...@@ -247,29 +263,41 @@ static bool has_fetch_operators( ...@@ -247,29 +263,41 @@ static bool has_fetch_operators(
if (op->Type() == kFetchOpType) { if (op->Type() == kFetchOpType) {
fetch_count++; fetch_count++;
// The output variable's name of fetch_op should be fetch_holder_name. // The output variable's name of fetch_op should be fetch_holder_name.
PADDLE_ENFORCE_EQ(op->Output("Out")[0], fetch_holder_name, PADDLE_ENFORCE_EQ(
"Output of fetch op should be '%s'", fetch_holder_name); op->Output("Out")[0], fetch_holder_name,
platform::errors::PreconditionNotMet(
"Output of fetch op should be '%s', but received '%s'.",
fetch_holder_name, op->Output("Out")[0]));
std::string fetch_target_name = op->Input("X")[0]; std::string fetch_target_name = op->Input("X")[0];
PADDLE_ENFORCE( PADDLE_ENFORCE_NE(fetch_targets.find(fetch_target_name),
fetch_targets.find(fetch_target_name) != fetch_targets.end(), fetch_targets.end(),
"Fetch operator input name '%s' cannot be found in 'fetch_targets'", platform::errors::NotFound(
fetch_target_name); "Fetch operator input name '%s' cannot be found in "
"'fetch_targets'.",
fetch_target_name));
} }
} }
if (fetch_count > 0) { if (fetch_count > 0) {
PADDLE_ENFORCE_EQ( PADDLE_ENFORCE_EQ(
fetch_count, fetch_targets.size(), fetch_count, fetch_targets.size(),
"The number of fetch operators should match 'fetch_targets'"); platform::errors::PreconditionNotMet(
"The number of fetch operators should match 'fetch_targets', but "
"received fetch_count: %zu, required fetch_targets.size(): %zu.",
fetch_count, fetch_targets.size()));
if (!fetch_holder_name.empty()) { if (!fetch_holder_name.empty()) {
// When fetch operator are present, so should be fetch_holder. // When fetch operator are present, so should be fetch_holder.
auto var = block.FindVar(fetch_holder_name); auto var = block.FindVar(fetch_holder_name);
PADDLE_ENFORCE_NOT_NULL(var, "Block should already have a '%s' variable", PADDLE_ENFORCE_NOT_NULL(
fetch_holder_name); var,
PADDLE_ENFORCE_EQ(var->GetType(), proto::VarType::FETCH_LIST, platform::errors::PreconditionNotMet(
"'%s' variable should be 'FETCH_LIST' type", "Block should already have a '%s' variable.", fetch_holder_name));
fetch_holder_name); PADDLE_ENFORCE_EQ(
var->GetType(), proto::VarType::FETCH_LIST,
platform::errors::PreconditionNotMet(
"'%s' variable should be 'FETCH_LIST' type, but received '%s'.",
fetch_holder_name, DataTypeToString(var->GetType())));
} }
} }
...@@ -354,7 +382,11 @@ std::unique_ptr<ExecutorPrepareContext> Executor::Prepare( ...@@ -354,7 +382,11 @@ std::unique_ptr<ExecutorPrepareContext> Executor::Prepare(
const std::vector<std::string>& skip_ref_cnt_vars, bool force_disable_gc) { const std::vector<std::string>& skip_ref_cnt_vars, bool force_disable_gc) {
std::unique_ptr<ExecutorPrepareContext> ctx( std::unique_ptr<ExecutorPrepareContext> ctx(
new ExecutorPrepareContext(program, block_id)); new ExecutorPrepareContext(program, block_id));
PADDLE_ENFORCE_LT(static_cast<size_t>(block_id), program.Size()); PADDLE_ENFORCE_LT(static_cast<size_t>(block_id), program.Size(),
platform::errors::InvalidArgument(
"Input block id = %d, but it should be less than "
"program.size() which is %d",
static_cast<size_t>(block_id), program.Size()));
auto& block = program.Block(block_id); auto& block = program.Block(block_id);
for (auto& op_desc : block.AllOps()) { for (auto& op_desc : block.AllOps()) {
ctx->ops_.push_back(OpRegistry::CreateOp(*op_desc)); ctx->ops_.push_back(OpRegistry::CreateOp(*op_desc));
...@@ -367,14 +399,20 @@ std::vector<std::shared_ptr<ExecutorPrepareContext>> Executor::Prepare( ...@@ -367,14 +399,20 @@ std::vector<std::shared_ptr<ExecutorPrepareContext>> Executor::Prepare(
const ProgramDesc& program, const std::vector<int>& block_ids, const ProgramDesc& program, const std::vector<int>& block_ids,
const std::vector<std::vector<std::string>>& skip_ref_cnt_vars, const std::vector<std::vector<std::string>>& skip_ref_cnt_vars,
bool force_disable_gc) { bool force_disable_gc) {
PADDLE_ENFORCE( PADDLE_ENFORCE_EQ(
skip_ref_cnt_vars.empty() || skip_ref_cnt_vars.size() == block_ids.size(), skip_ref_cnt_vars.empty() || skip_ref_cnt_vars.size() == block_ids.size(),
"skip_ref_cnt_vars should be either empty or equals to block number %d", true,
block_ids.size()); platform::errors::InvalidArgument("skip_ref_cnt_vars should be either "
"empty or equals to block number %d",
block_ids.size()));
std::vector<std::shared_ptr<ExecutorPrepareContext>> result; std::vector<std::shared_ptr<ExecutorPrepareContext>> result;
size_t idx = 0; size_t idx = 0;
for (auto& bid : block_ids) { for (auto& bid : block_ids) {
PADDLE_ENFORCE_LT(static_cast<size_t>(bid), program.Size()); PADDLE_ENFORCE_LT(static_cast<size_t>(bid), program.Size(),
platform::errors::InvalidArgument(
"Input block id = %zu, but it should be less than "
"program.size() which is %zu",
static_cast<size_t>(bid), program.Size()));
auto* ctx = new ExecutorPrepareContext(program, bid); auto* ctx = new ExecutorPrepareContext(program, bid);
auto& block = program.Block(bid); auto& block = program.Block(bid);
for (auto& op_desc : block.AllOps()) { for (auto& op_desc : block.AllOps()) {
...@@ -397,7 +435,8 @@ void Executor::RunPartialPreparedContext(ExecutorPrepareContext* ctx, ...@@ -397,7 +435,8 @@ void Executor::RunPartialPreparedContext(ExecutorPrepareContext* ctx,
bool create_local_scope, bool create_local_scope,
bool create_vars, bool keep_kids) { bool create_vars, bool keep_kids) {
platform::RecordBlock b(kProgramId); platform::RecordBlock b(kProgramId);
PADDLE_ENFORCE_NOT_NULL(scope); PADDLE_ENFORCE_NOT_NULL(
scope, platform::errors::InvalidArgument("Scope shouldn't be null"));
Scope* local_scope = scope; Scope* local_scope = scope;
if (create_vars) { if (create_vars) {
if (create_local_scope) { if (create_local_scope) {
...@@ -470,12 +509,14 @@ void Executor::RunPreparedContext( ...@@ -470,12 +509,14 @@ void Executor::RunPreparedContext(
const std::string& fetch_holder_name) { const std::string& fetch_holder_name) {
auto& global_block = ctx->prog_.Block(ctx->block_id_); auto& global_block = ctx->prog_.Block(ctx->block_id_);
PADDLE_ENFORCE( PADDLE_ENFORCE_EQ(
has_feed_operators(global_block, *feed_targets, feed_holder_name), has_feed_operators(global_block, *feed_targets, feed_holder_name), true,
"Program in ExecutorPrepareContext should has feed_ops."); platform::errors::PreconditionNotMet(
PADDLE_ENFORCE( "Program in ExecutorPrepareContext should has feed_ops."));
PADDLE_ENFORCE_EQ(
has_fetch_operators(global_block, *fetch_targets, fetch_holder_name), has_fetch_operators(global_block, *fetch_targets, fetch_holder_name),
"Program in the prepared context should has fetch_ops."); true, platform::errors::PreconditionNotMet(
"Program in the prepared context should has fetch_ops."));
// map the data of feed_targets to feed_holder // map the data of feed_targets to feed_holder
for (auto* op : global_block.AllOps()) { for (auto* op : global_block.AllOps()) {
......
paddle/fluid/framework/parallel_executor.cc
浏览文件 @ 1d3b0134
...@@ -155,7 +155,9 @@ class ParallelExecutorPrivate { ...@@ -155,7 +155,9 @@ class ParallelExecutorPrivate {
VLOG(10) << "find nccl_id_var:" << var_name << ", nccl_id:" << nccl_id; VLOG(10) << "find nccl_id_var:" << var_name << ", nccl_id:" << nccl_id;
} else { } else {
nccl_id = new ncclUniqueId(); nccl_id = new ncclUniqueId();
PADDLE_ENFORCE(platform::dynload::ncclGetUniqueId(nccl_id)); PADDLE_ENFORCE_EQ(
platform::dynload::ncclGetUniqueId(nccl_id), ncclSuccess,
platform::errors::PreconditionNotMet("Get NCCL unique ID failed."));
VLOG(10) << "can't find nccl_id_var:" << var_name VLOG(10) << "can't find nccl_id_var:" << var_name
<< ", nccl_id:" << nccl_id; << ", nccl_id:" << nccl_id;
} }
...@@ -178,7 +180,9 @@ class ParallelExecutorPrivate { ...@@ -178,7 +180,9 @@ class ParallelExecutorPrivate {
for (int i = 0; i < static_cast<int>(bst.nccl_comm_num_); i++) { for (int i = 0; i < static_cast<int>(bst.nccl_comm_num_); i++) {
std::string var_name = platform::GetFlatNCCLVarName(i); std::string var_name = platform::GetFlatNCCLVarName(i);
auto nccl_id_var = scope->FindVar(var_name); auto nccl_id_var = scope->FindVar(var_name);
PADDLE_ENFORCE(nccl_id_var, "can't find %s nccl_id_var", var_name); PADDLE_ENFORCE_NOT_NULL(
nccl_id_var,
platform::errors::NotFound("Can't find nccl_id_var '%s'.", var_name));
auto nccl_id = nccl_id_var->GetMutable<ncclUniqueId>(); auto nccl_id = nccl_id_var->GetMutable<ncclUniqueId>();
flat_nccl_ids.push_back(nccl_id); flat_nccl_ids.push_back(nccl_id);
} }
...@@ -191,7 +195,9 @@ class ParallelExecutorPrivate { ...@@ -191,7 +195,9 @@ class ParallelExecutorPrivate {
for (int i = 0; i < static_cast<int>(bst.nccl_comm_num_); i++) { for (int i = 0; i < static_cast<int>(bst.nccl_comm_num_); i++) {
std::string var_name = platform::GetHierarchicalInterNCCLVarName(i); std::string var_name = platform::GetHierarchicalInterNCCLVarName(i);
auto nccl_id_var = scope->FindVar(var_name); auto nccl_id_var = scope->FindVar(var_name);
PADDLE_ENFORCE(nccl_id_var, "can't find %s nccl_id_var", var_name); PADDLE_ENFORCE_NOT_NULL(nccl_id_var,
platform::errors::NotFound(
"Can't find nccl_id_var '%s'.", var_name));
auto inter_nccl_id = nccl_id_var->GetMutable<ncclUniqueId>(); auto inter_nccl_id = nccl_id_var->GetMutable<ncclUniqueId>();
inter_nccl_ids.push_back(inter_nccl_id); inter_nccl_ids.push_back(inter_nccl_id);
} }
...@@ -200,7 +206,9 @@ class ParallelExecutorPrivate { ...@@ -200,7 +206,9 @@ class ParallelExecutorPrivate {
for (int i = 0; i < static_cast<int>(bst.nccl_comm_num_); i++) { for (int i = 0; i < static_cast<int>(bst.nccl_comm_num_); i++) {
std::string var_name = platform::GetHierarchicalExterNCCLVarName(i); std::string var_name = platform::GetHierarchicalExterNCCLVarName(i);
auto nccl_id_var = scope->FindVar(var_name); auto nccl_id_var = scope->FindVar(var_name);
PADDLE_ENFORCE(nccl_id_var, "can't find %s nccl_id_var", var_name); PADDLE_ENFORCE_NOT_NULL(nccl_id_var,
platform::errors::NotFound(
"Can't find nccl_id_var '%s'.", var_name));
auto nccl_id = nccl_id_var->GetMutable<ncclUniqueId>(); auto nccl_id = nccl_id_var->GetMutable<ncclUniqueId>();
exter_nccl_ids.push_back(nccl_id); exter_nccl_ids.push_back(nccl_id);
} }
...@@ -216,8 +224,9 @@ class ParallelExecutorPrivate { ...@@ -216,8 +224,9 @@ class ParallelExecutorPrivate {
const std::string var_name = "NCCLCommunicator"; const std::string var_name = "NCCLCommunicator";
auto var = scope->FindVar(var_name); auto var = scope->FindVar(var_name);
if (var != nullptr) { if (var != nullptr) {
PADDLE_ENFORCE(var->IsInitialized(), PADDLE_ENFORCE_EQ(var->IsInitialized(), true,
"if %s exists, it must be initialized", var_name); platform::errors::PreconditionNotMet(
"if %s exists, it must be initialized", var_name));
VLOG(1) << "find " << var_name VLOG(1) << "find " << var_name
<< " in scope, so use it and does not recreate!"; << " in scope, so use it and does not recreate!";
nccl_ctxs_ = var->GetMutable<platform::NCCLCommunicator>(); nccl_ctxs_ = var->GetMutable<platform::NCCLCommunicator>();
...@@ -225,15 +234,21 @@ class ParallelExecutorPrivate { ...@@ -225,15 +234,21 @@ class ParallelExecutorPrivate {
} }
if (bst->use_hierarchical_allreduce_) { if (bst->use_hierarchical_allreduce_) {
PADDLE_ENFORCE(bst->num_trainers_ > 1, "num_trainers:%llu < 1", PADDLE_ENFORCE_GT(
bst->num_trainers_); bst->num_trainers_, 1,
PADDLE_ENFORCE(bst->hierarchical_allreduce_inter_nranks_ > 1, platform::errors::PreconditionNotMet(
"inter_nranks:%d < 1", "The num_trainers should be greater than 1, but received %llu.",
bst->hierarchical_allreduce_inter_nranks_); bst->num_trainers_));
PADDLE_ENFORCE( PADDLE_ENFORCE_GT(
(bst->num_trainers_ % bst->hierarchical_allreduce_inter_nranks_ == 0), bst->hierarchical_allreduce_inter_nranks_, 1,
"num_trainers:%llu mod inter_nranks:%d != 0", bst->num_trainers_, platform::errors::PreconditionNotMet(
bst->hierarchical_allreduce_inter_nranks_); "The inter_nranks should be greater than 1, but received %d.",
bst->hierarchical_allreduce_inter_nranks_));
PADDLE_ENFORCE_EQ(
bst->num_trainers_ % bst->hierarchical_allreduce_inter_nranks_, 0,
platform::errors::PreconditionNotMet(
"num_trainers:%llu mod inter_nranks:%d != 0", bst->num_trainers_,
bst->hierarchical_allreduce_inter_nranks_));
bst->hierarchical_allreduce_exter_nranks_ = bst->hierarchical_allreduce_exter_nranks_ =
bst->num_trainers_ / bst->hierarchical_allreduce_inter_nranks_; bst->num_trainers_ / bst->hierarchical_allreduce_inter_nranks_;
...@@ -369,7 +384,8 @@ ir::Graph *ParallelExecutorPrivate::ApplyMemoryOptimizePass(ir::Graph *graph) { ...@@ -369,7 +384,8 @@ ir::Graph *ParallelExecutorPrivate::ApplyMemoryOptimizePass(ir::Graph *graph) {
max_memory_size)); max_memory_size));
VLOG(10) << "Created GarbageCollector at " << place; VLOG(10) << "Created GarbageCollector at " << place;
} else { } else {
PADDLE_THROW("Unsupported place for garbage collection"); PADDLE_THROW(platform::errors::PreconditionNotMet(
"Unsupported place for garbage collection"));
} }
#ifdef PADDLE_WITH_CUDA #ifdef PADDLE_WITH_CUDA
} }
...@@ -455,7 +471,9 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places, ...@@ -455,7 +471,9 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places,
} }
#if defined(PADDLE_WITH_CUDA) && defined(_WIN32) #if defined(PADDLE_WITH_CUDA) && defined(_WIN32)
if (member_->use_cuda_) { if (member_->use_cuda_) {
PADDLE_ENFORCE(places.size() == 1, "Windows can support Single GPU only."); PADDLE_ENFORCE_EQ(
places.size(), 1,
platform::errors::Unavailable("Windows can support Single GPU only."));
} }
#endif #endif
...@@ -484,7 +502,11 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places, ...@@ -484,7 +502,11 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places,
} }
} else { } else {
member_->own_local_scope_ = false; member_->own_local_scope_ = false;
PADDLE_ENFORCE_EQ(member_->places_.size(), local_scopes.size()); PADDLE_ENFORCE_EQ(member_->places_.size(), local_scopes.size(),
platform::errors::PreconditionNotMet(
"member_->places_.size() = %d is not equal to "
"local_scopes.size() = %d",
member_->places_.size(), local_scopes.size()));
for (size_t i = 0; i < member_->places_.size(); ++i) { for (size_t i = 0; i < member_->places_.size(); ++i) {
member_->local_scopes_.emplace_back(&local_scopes[i]->NewScope()); member_->local_scopes_.emplace_back(&local_scopes[i]->NewScope());
} }
...@@ -492,8 +514,9 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places, ...@@ -492,8 +514,9 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places,
std::vector<ir::Graph *> graphs; std::vector<ir::Graph *> graphs;
if (member_->build_strategy_.async_mode_) { if (member_->build_strategy_.async_mode_) {
PADDLE_ENFORCE(!member_->use_cuda_, PADDLE_ENFORCE_EQ(member_->use_cuda_, false,
"gpu mode does not support async_mode_ now!"); platform::errors::Unavailable(
"gpu mode does not support async_mode_ now!"));
graphs.push_back(graph); graphs.push_back(graph);
for (size_t i = 1; i < places.size(); ++i) { for (size_t i = 1; i < places.size(); ++i) {
auto *tmp_graph = new ir::Graph(graph->OriginProgram()); auto *tmp_graph = new ir::Graph(graph->OriginProgram());
...@@ -622,8 +645,12 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places, ...@@ -622,8 +645,12 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places,
scope_map.emplace(scope, &local_exec_scope); scope_map.emplace(scope, &local_exec_scope);
} }
PADDLE_ENFORCE_EQ(member_->local_scopes_.size(), PADDLE_ENFORCE_EQ(
member_->local_exec_scopes_.size()); member_->local_scopes_.size(), member_->local_exec_scopes_.size(),
platform::errors::PreconditionNotMet(
"member_->local_scopes_.size() = %d is not equal to "
"member_->local_exec_scopes_.size() = %d",
member_->local_scopes_.size(), member_->local_exec_scopes_.size()));
std::vector<ir::Graph *> final_graphs; std::vector<ir::Graph *> final_graphs;
...@@ -655,8 +682,8 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places, ...@@ -655,8 +682,8 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places,
} }
} }
#else #else
PADDLE_THROW( PADDLE_THROW(platform::errors::PreconditionNotMet(
"Paddle should be compiled with CUDA for ParallelGraph Execution."); "Paddle should be compiled with CUDA for ParallelGraph Execution."));
#endif #endif
} else { } else {
bool has_drop_last_read_op = details::HasDropLastReadOp(*graph); bool has_drop_last_read_op = details::HasDropLastReadOp(*graph);
...@@ -746,7 +773,10 @@ void ParallelExecutor::BCastParamsToDevices( ...@@ -746,7 +773,10 @@ void ParallelExecutor::BCastParamsToDevices(
} }
PADDLE_ENFORCE_EQ(member_->places_.size(), buffers.size(), PADDLE_ENFORCE_EQ(member_->places_.size(), buffers.size(),
"variables' buffer size to bcast NOT equal to places"); platform::errors::PreconditionNotMet(
"variables' buffer size to bcast is %d, which is "
"NOT equal to places size %d",
buffers.size(), member_->places_.size()));
{ {
auto *nccl_ctxs = member_->nccl_ctxs_->DefaultFlatCtx(); auto *nccl_ctxs = member_->nccl_ctxs_->DefaultFlatCtx();
platform::NCCLGroupGuard guard; platform::NCCLGroupGuard guard;
...@@ -889,15 +919,17 @@ void ParallelExecutor::FeedAndSplitTensorIntoLocalScopes( ...@@ -889,15 +919,17 @@ void ParallelExecutor::FeedAndSplitTensorIntoLocalScopes(
"You should set the environment variable CPU_NUM in the system " "You should set the environment variable CPU_NUM in the system "
"to determine the number of devices you need."; "to determine the number of devices you need.";
} }
PADDLE_THROW(error_info); PADDLE_THROW(platform::errors::PreconditionNotMet(error_info));
} else if (is_persistable) { } else if (is_persistable) {
if (lod_tensors.size() == 1) { if (lod_tensors.size() == 1) {
lod_tensors.reserve(num_places); lod_tensors.reserve(num_places);
auto &tensor = lod_tensors.front(); auto &tensor = lod_tensors.front();
PADDLE_ENFORCE_EQ(tensor.dims(), pair.second.dims(), PADDLE_ENFORCE_EQ(
"The dim doesn't match."); tensor.dims(), pair.second.dims(),
PADDLE_ENFORCE_EQ(tensor.place(), member_->places_.at(0), platform::errors::PreconditionNotMet("The dim doesn't match."));
"The place doesn't match."); PADDLE_ENFORCE_EQ(
tensor.place(), member_->places_.at(0),
platform::errors::PreconditionNotMet("The place doesn't match."));
for (size_t i = 1; i < num_places; ++i) { for (size_t i = 1; i < num_places; ++i) {
lod_tensors.emplace_back(); lod_tensors.emplace_back();
auto &tmp = lod_tensors.back(); auto &tmp = lod_tensors.back();
...@@ -914,7 +946,7 @@ void ParallelExecutor::FeedAndSplitTensorIntoLocalScopes( ...@@ -914,7 +946,7 @@ void ParallelExecutor::FeedAndSplitTensorIntoLocalScopes(
"value, you should feed %d samples.", "value, you should feed %d samples.",
lod_tensors.size(), pair.first, num_places, lod_tensors.size(), pair.first, num_places,
(is_cpu_place ? "CPU" : "GPU"), pair.first, num_places, num_places); (is_cpu_place ? "CPU" : "GPU"), pair.first, num_places, num_places);
PADDLE_THROW(error_info); PADDLE_THROW(platform::errors::PreconditionNotMet(error_info));
} }
} }
......
paddle/fluid/operators/split_lod_tensor_op.cc
浏览文件 @ 1d3b0134
...@@ -58,7 +58,8 @@ class SplitLoDTensorOp : public framework::OperatorBase { ...@@ -58,7 +58,8 @@ class SplitLoDTensorOp : public framework::OperatorBase {
framework::TensorCopy(mask, platform::CPUPlace(), dev_ctx, framework::TensorCopy(mask, platform::CPUPlace(), dev_ctx,
cpu_mask.get()); cpu_mask.get());
#else #else
PADDLE_THROW("Not supported GPU, Please compile WITH_GPU option"); PADDLE_THROW(paddle::platform::errors::Fatal(
"Not support GPU, Please compile WITH_GPU option"));
#endif #endif
} }
auto *mask_data = cpu_mask->data<bool>(); auto *mask_data = cpu_mask->data<bool>();
...@@ -146,34 +147,34 @@ class SplitLoDTensorOpProtoMaker : public framework::OpProtoAndCheckerMaker { ...@@ -146,34 +147,34 @@ class SplitLoDTensorOpProtoMaker : public framework::OpProtoAndCheckerMaker {
class SplitLoDTensorInferShape : public framework::InferShapeBase { class SplitLoDTensorInferShape : public framework::InferShapeBase {
public: public:
void operator()(framework::InferShapeContext *context) const override { void operator()(framework::InferShapeContext *context) const override {
PADDLE_ENFORCE(context->HasInput("X"), OP_INOUT_CHECK(context->HasInput("X"), "Input", "X", "SplitLoDTensor");
"SplitLoDTensorOp must have input X."); OP_INOUT_CHECK(context->HasInput("Mask"), "Input", "Mask",
PADDLE_ENFORCE(context->HasInput("Mask"), "SplitLoDTensor");
"SplitLoDTensorOp must have input Mask."); OP_INOUT_CHECK(context->HasOutput("OutTrue"), "Output", "OutTrue",
PADDLE_ENFORCE(context->HasOutput("OutTrue"), "SplitLoDTensor");
"SplitLoDTensorOp must have output OutTrue."); OP_INOUT_CHECK(context->HasOutput("OutFalse"), "Output", "OutFalse",
PADDLE_ENFORCE(context->HasOutput("OutFalse"), "SplitLoDTensor");
"SplitLoDTensorOp must have output OutFalse.");
auto mask_dim = context->GetInputDim("Mask"); auto mask_dim = context->GetInputDim("Mask");
PADDLE_ENFORCE_EQ(mask_dim.size(), 2, PADDLE_ENFORCE_EQ(
"If you are using IfElse OP:" mask_dim.size(), 2,
"\n\nie = fluid.layers.IfElse(cond=cond)\nwith " platform::errors::InvalidArgument(
"ie.true_block():\n out_1 = ie.input(x)\n\n" "If you are using IfElse OP:"
"Please ensure that the cond should be a 2-D tensor and " "\n\nie = fluid.layers.IfElse(cond=cond)\nwith "
"the second dim size of cond should be 1. " "ie.true_block():\n out_1 = ie.input(x)\n\n"
"But now the cond's shape is [", "Please ensure that the cond should be a 2-D tensor and "
*mask_dim.Get(), "].\n"); "the second dim size of cond should be 1. "
if (context->IsRuntime()) { "But now the cond's shape is [",
PADDLE_ENFORCE_EQ(mask_dim[1], 1, *mask_dim.Get(), "].\n"));
"If you are using IfElse OP:" PADDLE_ENFORCE_EQ(mask_dim[1], 1,
"\n\nie = fluid.layers.IfElse(cond=cond)\nwith " platform::errors::InvalidArgument(
"ie.true_block():\n out_1 = ie.input(x)\n\n" "If you are using IfElse OP:"
"Please ensure that the cond should be a 2-D tensor " "\n\nie = fluid.layers.IfElse(cond=cond)\nwith "
"and the second dim size of cond should be 1. " "ie.true_block():\n out_1 = ie.input(x)\n\n"
"But now the cond's shape is [", "Please ensure that the cond should be a 2-D tensor "
*mask_dim.Get(), "].\n"); "and the second dim size of cond should be 1. "
} "But now the cond's shape is [",
*mask_dim.Get(), "].\n"));
context->SetOutputDim("OutTrue", context->GetInputDim("X")); context->SetOutputDim("OutTrue", context->GetInputDim("X"));
context->SetOutputDim("OutFalse", context->GetInputDim("X")); context->SetOutputDim("OutFalse", context->GetInputDim("X"));
......
python/paddle/fluid/layers/control_flow.py
浏览文件 @ 1d3b0134
...@@ -110,9 +110,9 @@ def split_lod_tensor(input, mask, level=0): ...@@ -110,9 +110,9 @@ def split_lod_tensor(input, mask, level=0):
data into two parts. data into two parts.
Args: Args:
input(tuple|list|None): The input tensor that contains complete input(Variable|tuple|list|None): The input tensor that contains complete
lod information needed to construct the output. lod information needed to construct the output.
mask(list): A bool column vector which masks the input. mask(Variable|list): A bool column vector which masks the input.
level(int): The specific lod level to split. level(int): The specific lod level to split.
Returns: Returns:
...@@ -135,6 +135,10 @@ def split_lod_tensor(input, mask, level=0): ...@@ -135,6 +135,10 @@ def split_lod_tensor(input, mask, level=0):
input=x, mask=y, level=level) input=x, mask=y, level=level)
""" """
check_type(input, 'input', (Variable, list, tuple, type(None)),
'fluid.layers.split_lod_tensor')
check_type(mask, 'mask', (Variable, list), 'fluid.layers.split_lod_tensor')
check_type(level, 'level', int, 'fluid.layers.split_lod_tensor')
helper = LayerHelper('split_lod_tensor', **locals()) helper = LayerHelper('split_lod_tensor', **locals())
out_true = helper.create_variable_for_type_inference(dtype=input.dtype) out_true = helper.create_variable_for_type_inference(dtype=input.dtype)
out_false = helper.create_variable_for_type_inference(dtype=input.dtype) out_false = helper.create_variable_for_type_inference(dtype=input.dtype)
...@@ -407,6 +411,7 @@ class StaticRNN(object): ...@@ -407,6 +411,7 @@ class StaticRNN(object):
AFTER_RNN_BLOCK = 2 AFTER_RNN_BLOCK = 2
def __init__(self, name=None): def __init__(self, name=None):
check_type(name, "name", (str, type(None)), "fluid.layers.StaticRNN")
self.helper = LayerHelper("static_rnn", name=name) self.helper = LayerHelper("static_rnn", name=name)
self.memories = {} # memory map, from pre_mem.name --> MemoryLink self.memories = {} # memory map, from pre_mem.name --> MemoryLink
self.inputs = [] # input variable list in current block self.inputs = [] # input variable list in current block
...@@ -511,6 +516,12 @@ class StaticRNN(object): ...@@ -511,6 +516,12 @@ class StaticRNN(object):
""" """
self._assert_in_rnn_block_('memory') self._assert_in_rnn_block_('memory')
check_type(init, "init", (Variable, type(None)),
"fluid.layers.StaticRNN.memory")
check_type(shape, "shape", (list, tuple, type(None)),
"fluid.layers.StaticRNN.memory")
check_type(batch_ref, "batch_ref", (Variable, type(None)),
"fluid.layers.StaticRNN.memory")
if init is None: if init is None:
if shape is None or batch_ref is None: if shape is None or batch_ref is None:
raise ValueError( raise ValueError(
...@@ -587,8 +598,7 @@ class StaticRNN(object): ...@@ -587,8 +598,7 @@ class StaticRNN(object):
""" """
self._assert_in_rnn_block_('step_input') self._assert_in_rnn_block_('step_input')
if not isinstance(x, Variable): check_type(x, "x", Variable, "fluid.layers.StaticRNN.step_input")
raise TypeError("step input takes a Variable")
if self.seq_len is None: if self.seq_len is None:
self.seq_len = x.shape[0] self.seq_len = x.shape[0]
elif x.shape[0] != -1 and self.seq_len != x.shape[0]: elif x.shape[0] != -1 and self.seq_len != x.shape[0]:
...@@ -641,8 +651,7 @@ class StaticRNN(object): ...@@ -641,8 +651,7 @@ class StaticRNN(object):
""" """
self._assert_in_rnn_block_('step_output') self._assert_in_rnn_block_('step_output')
if not isinstance(o, Variable): check_type(o, "o", Variable, "fluid.layers.StaticRNN.step_output")
raise TypeError("step output takes a Variable")
tmp_o = self.helper.create_variable_for_type_inference(dtype=o.dtype) tmp_o = self.helper.create_variable_for_type_inference(dtype=o.dtype)
self.helper.append_op( self.helper.append_op(
...@@ -715,8 +724,8 @@ class StaticRNN(object): ...@@ -715,8 +724,8 @@ class StaticRNN(object):
None None
""" """
if not isinstance(mem, Variable) or not isinstance(var, Variable): check_type(mem, "mem", Variable, "fluid.layers.StaticRNN.update_memory")
raise TypeError("update memory should take variables") check_type(var, "var", Variable, "fluid.layers.StaticRNN.update_memory")
self.memories[mem.name].mem = var self.memories[mem.name].mem = var
def _parent_block(self): def _parent_block(self):
...@@ -2134,6 +2143,8 @@ def cond(pred, true_fn=None, false_fn=None, name=None): ...@@ -2134,6 +2143,8 @@ def cond(pred, true_fn=None, false_fn=None, name=None):
return false_fn() return false_fn()
return None return None
check_variable_and_dtype(pred, "pred", ['bool'], "fluid.layers.cond")
check_type(name, "name", (str, type(None)), "fluid.layers.cond")
helper = LayerHelper('cond', **locals()) helper = LayerHelper('cond', **locals())
true_output = None true_output = None
false_output = None false_output = None
...@@ -2545,8 +2556,8 @@ class IfElse(object): ...@@ -2545,8 +2556,8 @@ class IfElse(object):
IN_IF_ELSE_FALSE_BLOCKS = 2 IN_IF_ELSE_FALSE_BLOCKS = 2
def __init__(self, cond, name=None): def __init__(self, cond, name=None):
if not isinstance(cond, Variable): check_type(cond, "cond", Variable, "fluid.layers.IfElse")
raise TypeError("cond must be a Variable") check_type(name, "name", (str, type(None)), "fluid.layers.IfElse")
self.helper = LayerHelper('ifelse', name=name) self.helper = LayerHelper('ifelse', name=name)
self.cond = cond self.cond = cond
self.input_table = {} self.input_table = {}
...@@ -2605,8 +2616,8 @@ class IfElse(object): ...@@ -2605,8 +2616,8 @@ class IfElse(object):
self.IN_IF_ELSE_TRUE_BLOCKS else 0] self.IN_IF_ELSE_TRUE_BLOCKS else 0]
parent_block = self._parent_block() parent_block = self._parent_block()
for each_out in outs: for each_out in outs:
if not isinstance(each_out, Variable): check_type(each_out, "each output", Variable,
raise TypeError("Each output should be a variable") "fluid.layers.IfElse.output")
# create outside tensor # create outside tensor
outside_out = parent_block.create_var( outside_out = parent_block.create_var(
name=unique_name.generate_with_ignorable_key("_".join( name=unique_name.generate_with_ignorable_key("_".join(
......
python/paddle/fluid/parallel_executor.py
浏览文件 @ 1d3b0134
...@@ -17,6 +17,7 @@ from . import core ...@@ -17,6 +17,7 @@ from . import core
from . import framework from . import framework
from . import executor from . import executor
from . import compiler from . import compiler
from .data_feeder import check_type
import sys import sys
__all__ = ['ParallelExecutor'] __all__ = ['ParallelExecutor']
...@@ -357,14 +358,14 @@ class ParallelExecutor(object): ...@@ -357,14 +358,14 @@ class ParallelExecutor(object):
parallel_exe.drop_local_exe_scopes() parallel_exe.drop_local_exe_scopes()
""" """
assert isinstance( check_type(self._compiled_program._executor,
self._compiled_program._executor, "the Executor of compiled program", core.ParallelExecutor,
core.ParallelExecutor), "The Executor should be ParallelExecutor." "ParallelExecutor.drop_local_exe_scopes")
self._compiled_program._executor.drop_local_exe_scopes() self._compiled_program._executor.drop_local_exe_scopes()
# This API is used to check whether DropLocalExeScopes can work. # This API is used to check whether DropLocalExeScopes can work.
def _need_create_local_exe_scopes(self): def _need_create_local_exe_scopes(self):
assert isinstance( check_type(self._compiled_program._executor,
self._compiled_program._executor, "the Executor of compiled program", core.ParallelExecutor,
core.ParallelExecutor), "The Executor should be ParallelExecutor." "ParallelExecutor._need_create_local_exe_scopes")
return self._compiled_program._executor._need_create_local_exe_scopes() return self._compiled_program._executor._need_create_local_exe_scopes()
python/paddle/fluid/tests/test_if_else_op.py
浏览文件 @ 1d3b0134
...@@ -221,5 +221,27 @@ class TestIfElseFalseBranch(TestIfElse): ...@@ -221,5 +221,27 @@ class TestIfElseFalseBranch(TestIfElse):
self.data = np.random.rand(25, 1).astype(np.float32) self.data = np.random.rand(25, 1).astype(np.float32)
class TestIfElseError(unittest.TestCase):
def test_input_type_error(self):
main_program = Program()
startup_program = Program()
with program_guard(main_program, startup_program):
src = layers.data(name='data', shape=[1], dtype='float32')
const_value = layers.fill_constant(
[1], dtype='float32', value=123.0)
ifcond = layers.less_than(x=src, y=const_value)
with self.assertRaises(TypeError):
ie = layers.IfElse(set())
with self.assertRaises(TypeError):
ie = layers.IfElse(ifcond, set())
with self.assertRaises(TypeError):
ie = layers.IfElse(ifcond)
with ie.true_block():
true_target = ie.input(src)
true_target = fluid.layers.exp(true_target)
ie.output([])
if __name__ == '__main__': if __name__ == '__main__':
unittest.main() unittest.main()
python/paddle/fluid/tests/unittests/test_cond.py
浏览文件 @ 1d3b0134
...@@ -539,5 +539,28 @@ class TestCondBackward(unittest.TestCase): ...@@ -539,5 +539,28 @@ class TestCondBackward(unittest.TestCase):
use_parallel_exe) use_parallel_exe)
class TestCondWithError(unittest.TestCase):
def test_input_type_error(self):
main_program = framework.Program()
startup_program = framework.Program()
with framework.program_guard(main_program, startup_program):
pred = fluid.data(name='y', shape=[1], dtype='bool')
def func():
return pred
with self.assertRaises(TypeError):
layers.cond(None, func, func)
with self.assertRaises(TypeError):
layers.cond(pred, func, set())
with self.assertRaises(TypeError):
layers.cond(pred, set(), func)
with self.assertRaises(TypeError):
layers.cond(pred, func, func, set())
if __name__ == '__main__': if __name__ == '__main__':
unittest.main() unittest.main()
python/paddle/fluid/tests/unittests/test_split_and_merge_lod_tensor_op.py
浏览文件 @ 1d3b0134
...@@ -221,5 +221,26 @@ class TestCPUSplitMergeLoDTensorGrad(unittest.TestCase): ...@@ -221,5 +221,26 @@ class TestCPUSplitMergeLoDTensorGrad(unittest.TestCase):
self.assertAlmostEqual(1.0, g_out_sum, delta=0.1) self.assertAlmostEqual(1.0, g_out_sum, delta=0.1)
class TestSplitLodTensorWithError(unittest.TestCase):
def test_error(self):
main_program = Program()
startup_program = Program()
with program_guard(main_program, startup_program):
x = layers.data(
name='x', shape=[1], dtype='float32', stop_gradient=False)
y = layers.data(
name='y', shape=[1], dtype='bool', stop_gradient=False)
level = 0
with self.assertRaises(TypeError):
split_lod_tensor(input=set(), mask=y, level=level)
with self.assertRaises(TypeError):
split_lod_tensor(input=x, mask=set(), level=level)
with self.assertRaises(TypeError):
split_lod_tensor(input=x, mask=set(), level=None)
if __name__ == '__main__': if __name__ == '__main__':
unittest.main() unittest.main()
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或