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

Fix quantile tests running on multi-gpus #8775

Merged
merged 8 commits into from
Feb 13, 2023
Merged

Fix quantile tests running on multi-gpus #8775

Changes from 1 commit
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
f67b624
Fix quantile tests running on multi-gpus
rongou Feb 9, 2023
960f09e
Run some gtests with multiple GPUs
hcho3 Feb 10, 2023
757c304
fix mgpu test naming
rongou Feb 10, 2023
d5a4b09
Merge remote-tracking branch 'upstream/master' into fix-mgpu-quantile…
rongou Feb 10, 2023
05739ab
Instruct NCCL to print extra logs
hcho3 Feb 10, 2023
7127abd
Merge remote-tracking branch 'upstream/master' into fix-mgpu-quantile…
hcho3 Feb 12, 2023
4e9e6c5
Allocate extra space in /dev/shm to enable NCCL
hcho3 Feb 12, 2023
2669770
use gtest_skip to skip mgpu tests
rongou Feb 13, 2023
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
39 changes: 21 additions & 18 deletions tests/cpp/common/test_quantile.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -349,22 +349,24 @@ void TestAllReduceBasic(int32_t n_gpus) {

constexpr size_t kRows = 1000, kCols = 100;
RunWithSeedsAndBins(kRows, [=](int32_t seed, size_t n_bins, MetaInfo const& info) {
auto const device = collective::GetRank();

// Set up single node version;
HostDeviceVector<FeatureType> ft;
SketchContainer sketch_on_single_node(ft, n_bins, kCols, kRows, 0);
HostDeviceVector<FeatureType> ft({}, device);
SketchContainer sketch_on_single_node(ft, n_bins, kCols, kRows, device);

size_t intermediate_num_cuts = std::min(
kRows * world, static_cast<size_t>(n_bins * WQSketch::kFactor));
std::vector<SketchContainer> containers;
for (auto rank = 0; rank < world; ++rank) {
HostDeviceVector<float> storage;
HostDeviceVector<float> storage({}, device);
std::string interface_str = RandomDataGenerator{kRows, kCols, 0}
.Device(0)
.Device(device)
.Seed(rank + seed)
.GenerateArrayInterface(&storage);
data::CupyAdapter adapter(interface_str);
HostDeviceVector<FeatureType> ft;
containers.emplace_back(ft, n_bins, kCols, kRows, 0);
HostDeviceVector<FeatureType> ft({}, device);
containers.emplace_back(ft, n_bins, kCols, kRows, device);
AdapterDeviceSketch(adapter.Value(), n_bins, info,
std::numeric_limits<float>::quiet_NaN(),
&containers.back());
Expand All @@ -375,16 +377,16 @@ void TestAllReduceBasic(int32_t n_gpus) {
sketch_on_single_node.FixError();
}
sketch_on_single_node.Unique();
TestQuantileElemRank(0, sketch_on_single_node.Data(),
TestQuantileElemRank(device, sketch_on_single_node.Data(),
sketch_on_single_node.ColumnsPtr(), true);

// Set up distributed version. We rely on using rank as seed to generate
// the exact same copy of data.
auto rank = collective::GetRank();
SketchContainer sketch_distributed(ft, n_bins, kCols, kRows, 0);
HostDeviceVector<float> storage;
SketchContainer sketch_distributed(ft, n_bins, kCols, kRows, device);
HostDeviceVector<float> storage({}, device);
std::string interface_str = RandomDataGenerator{kRows, kCols, 0}
.Device(0)
.Device(device)
.Seed(rank + seed)
.GenerateArrayInterface(&storage);
data::CupyAdapter adapter(interface_str);
Expand All @@ -399,7 +401,7 @@ void TestAllReduceBasic(int32_t n_gpus) {
ASSERT_EQ(sketch_distributed.Data().size(),
sketch_on_single_node.Data().size());

TestQuantileElemRank(0, sketch_distributed.Data(),
TestQuantileElemRank(device, sketch_distributed.Data(),
sketch_distributed.ColumnsPtr(), true);

std::vector<SketchEntry> single_node_data(
Expand Down Expand Up @@ -437,12 +439,13 @@ void TestSameOnAllWorkers(int32_t n_gpus) {
constexpr size_t kRows = 1000, kCols = 100;
RunWithSeedsAndBins(kRows, [=](int32_t seed, size_t n_bins,
MetaInfo const &info) {
auto rank = collective::GetRank();
HostDeviceVector<FeatureType> ft;
SketchContainer sketch_distributed(ft, n_bins, kCols, kRows, 0);
HostDeviceVector<float> storage;
auto const rank = collective::GetRank();
auto const device = rank;
HostDeviceVector<FeatureType> ft({}, device);
SketchContainer sketch_distributed(ft, n_bins, kCols, kRows, device);
HostDeviceVector<float> storage({}, device);
std::string interface_str = RandomDataGenerator{kRows, kCols, 0}
.Device(0)
.Device(device)
.Seed(rank + seed)
.GenerateArrayInterface(&storage);
data::CupyAdapter adapter(interface_str);
Expand All @@ -451,7 +454,7 @@ void TestSameOnAllWorkers(int32_t n_gpus) {
&sketch_distributed);
sketch_distributed.AllReduce();
sketch_distributed.Unique();
TestQuantileElemRank(0, sketch_distributed.Data(), sketch_distributed.ColumnsPtr(), true);
TestQuantileElemRank(device, sketch_distributed.Data(), sketch_distributed.ColumnsPtr(), true);

// Test for all workers having the same sketch.
size_t n_data = sketch_distributed.Data().size();
Expand All @@ -468,7 +471,7 @@ void TestSameOnAllWorkers(int32_t n_gpus) {
thrust::copy(thrust::device, local_data.data(),
local_data.data() + local_data.size(),
all_workers.begin() + local_data.size() * rank);
collective::DeviceCommunicator* communicator = collective::Communicator::GetDevice(0);
collective::DeviceCommunicator* communicator = collective::Communicator::GetDevice(device);

communicator->AllReduceSum(all_workers.data().get(), all_workers.size());
communicator->Synchronize();
Expand Down