Nigel
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

!6996 In debugger, Add output filter for BatchNorm to avoid loading dirty output values 

Merge pull request !6996 from lichen_101010/batch_norm
This commit is contained in:
mindspore-ci-bot 2020-10-10 10:49:49 +08:00 committed by Gitee
parent fa4d2607c0 ea8c8361d6
commit f90529a4a8
1 changed files with 25 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
mindspore/ccsrc/runtime/device/gpu/gpu_kernel_runtime.cc
View File

@ -77,6 +77,24 @@ bool GPUKernelRuntime::Init() {
}
namespace {
std::vector<int> CheckRealOutput(const std::string &node_name, const size_t &output_size) {
// define a vector containing real output number
std::vector<int> real_outputs;
// P.FusedBatchNorm is used for training; P.BatchNorm is used for inference
// can add the filter list for more operators here....
if (node_name == "FusedBatchNorm") {
MS_LOG(INFO) << "loading node named FusedBatchNorm.";
real_outputs.insert(real_outputs.end(), {0, 3, 4});
} else {
// by default, TensorLoader will load all outputs
for (size_t j = 0; j < output_size; ++j) {
real_outputs.push_back(j);
}
}
return real_outputs;
}
void LoadKernelData(Debugger *debugger, const CNodePtr &kernel,
const std::vector<mindspore::kernel::AddressPtr> &kernel_inputs,
const std::vector<mindspore::kernel::AddressPtr> &kernel_workspaces,
@ -125,7 +143,13 @@ void LoadKernelData(Debugger *debugger, const CNodePtr &kernel,
// get outputs
auto output_size = AnfAlgo::GetOutputTensorNum(kernel);
for (size_t j = 0; j < output_size; ++j) {
auto node_name = AnfAlgo::GetCNodeName(kernel);
std::vector<int> real_outputs;
real_outputs = CheckRealOutput(node_name, output_size);
for (std::vector<int>::iterator it = real_outputs.begin(); it != real_outputs.end(); ++it) {
auto j = *it;
auto addr = kernel_outputs[j];
auto type = AnfAlgo::GetOutputInferDataType(kernel, j);
auto format = kOpFormat_DEFAULT;