saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-07-26 19:52:38 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Training fix

Browse source
This commit is contained in:
Ryan Hill 2021-05-10 15:09:35 -07:00
parent 7bdf68ffa2
commit 6076ef965b
2 changed files with 10 additions and 10 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

17
orttraining/orttraining/models/gpt2/main.cc
View file

@ -359,6 +359,7 @@ void setup_training_params(GPT2Parameters& params) {
0,
true,
0,
nullptr,
nullptr};
params.providers.emplace(kCudaExecutionProvider, CreateExecutionProviderFactory_Cuda(&info));
@ -439,17 +440,17 @@ static Status RunTraining(const GPT2Parameters& params, const Environment& env)
auto rank_in_data_parallel_group = MPIContext::GetInstance().GetWorldRank() / params.horizontal_parallel_size;
auto training_data_loader = std::make_unique<DataLoader>(params.input_name_map,
params.train_data_dir,
max_num_files_preload,
rank_in_data_parallel_group,
params.data_parallel_size);
params.train_data_dir,
max_num_files_preload,
rank_in_data_parallel_group,
params.data_parallel_size);
std::unique_ptr<DataLoader> test_data_loader;
// Evaluation is only done in device #0
if (MPIContext::GetInstance().GetWorldRank() == 0) {
test_data_loader = std::make_unique<DataLoader>(params.input_name_map,
params.test_data_dir,
max_num_files_preload);
params.test_data_dir,
max_num_files_preload);
}
if (!params.perf_output_dir.empty()) {
@ -463,8 +464,8 @@ static Status RunTraining(const GPT2Parameters& params, const Environment& env)
// only test and save trained model on device #0
if (MPIContext::GetInstance().GetWorldRank() == 0) {
test_data_loader = std::make_unique<DataLoader>(params.input_name_map,
params.test_data_dir,
max_num_files_preload);
params.test_data_dir,
max_num_files_preload);
ORT_RETURN_IF_ERROR(runner->EndTraining(test_data_loader.get()));
}

3
tools/ci_build/github/azure-pipelines/win-ci-pipeline.yml
View file

@ -6,8 +6,7 @@ parameters:
jobs:
- job: 'build'
pool: 'RyanWinGPU'
# pool: 'Win-CPU-2019'
pool: 'Win-CPU-2019'
strategy:
matrix:
x64_debug: