Run training with empty DMatrix. (#4990)

This makes GPU Hist robust in distributed environment as some workers might not
be associated with any data in either training or evaluation.

* Disable rabit mock test for now: See #5012 .

* Disable dask-cudf test at prediction for now: See #5003

* Launch dask job for all workers despite they might not have any data.
* Check 0 rows in elementwise evaluation metrics.

   Using AUC and AUC-PR still throws an error.  See #4663 for a robust fix.

* Add tests for edge cases.
* Add `LaunchKernel` wrapper handling zero sized grid.
* Move some parts of allreducer into a cu file.
* Don't validate feature names when the booster is empty.

* Sync number of columns in DMatrix.

  As num_feature is required to be the same across all workers in data split
  mode.

* Filtering in dask interface now by default syncs all booster that's not
empty, instead of using rank 0.

* Fix Jenkins' GPU tests.

* Install dask-cuda from source in Jenkins' test.

  Now all tests are actually running.

* Restore GPU Hist tree synchronization test.

* Check UUID of running devices.

  The check is only performed on CUDA version >= 10.x, as 9.x doesn't have UUID field.

* Fix CMake policy and project variables.

  Use xgboost_SOURCE_DIR uniformly, add policy for CMake >= 3.13.

* Fix copying data to CPU

* Fix race condition in cpu predictor.

* Fix duplicated DMatrix construction.

* Don't download extra nccl in CI script.
This commit is contained in:
Jiaming Yuan
2019-11-06 16:13:13 +08:00
committed by GitHub
parent 807a244517
commit 7663de956c
44 changed files with 603 additions and 272 deletions

View File

@@ -99,13 +99,13 @@ EllpackInfo::EllpackInfo(int device,
bool is_dense,
size_t row_stride,
const common::HistogramCuts& hmat,
dh::BulkAllocator& ba)
dh::BulkAllocator* ba)
: is_dense(is_dense), row_stride(row_stride), n_bins(hmat.Ptrs().back()) {
ba.Allocate(device,
&feature_segments, hmat.Ptrs().size(),
&gidx_fvalue_map, hmat.Values().size(),
&min_fvalue, hmat.MinValues().size());
ba->Allocate(device,
&feature_segments, hmat.Ptrs().size(),
&gidx_fvalue_map, hmat.Values().size(),
&min_fvalue, hmat.MinValues().size());
dh::CopyVectorToDeviceSpan(gidx_fvalue_map, hmat.Values());
dh::CopyVectorToDeviceSpan(min_fvalue, hmat.MinValues());
dh::CopyVectorToDeviceSpan(feature_segments, hmat.Ptrs());
@@ -116,7 +116,7 @@ void EllpackPageImpl::InitInfo(int device,
bool is_dense,
size_t row_stride,
const common::HistogramCuts& hmat) {
matrix.info = EllpackInfo(device, is_dense, row_stride, hmat, ba_);
matrix.info = EllpackInfo(device, is_dense, row_stride, hmat, &ba_);
}
// Initialize the buffer to stored compressed features.
@@ -189,7 +189,8 @@ void EllpackPageImpl::CreateHistIndices(int device,
const dim3 grid3(common::DivRoundUp(batch_nrows, block3.x),
common::DivRoundUp(row_stride, block3.y),
1);
CompressBinEllpackKernel<<<grid3, block3>>>(
dh::LaunchKernel {grid3, block3} (
CompressBinEllpackKernel,
common::CompressedBufferWriter(num_symbols),
gidx_buffer.data(),
row_ptrs.data().get(),