Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Run training with empty DMatrix. (#4990)

Browse Source 
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
Download Patch File Download Diff File Expand all files Collapse all files

10
src/gbm/gbtree.h
View File

@@ -246,6 +246,14 @@ class GBTree : public GradientBooster {
std::unique_ptr<Predictor> const& GetPredictor(HostDeviceVector<float> const* out_pred = nullptr,
DMatrix* f_dmat = nullptr) const {
CHECK(configured_);
auto on_device = f_dmat && (*(f_dmat->GetBatches<SparsePage>().begin())).data.DeviceCanRead();
#if defined(XGBOOST_USE_CUDA)
// Use GPU Predictor if data is already on device.
if (!specified_predictor_ && on_device) {
CHECK(gpu_predictor_);
return gpu_predictor_;
}
#endif // defined(XGBOOST_USE_CUDA)
// GPU_Hist by default has prediction cache calculated from quantile values, so GPU
// Predictor is not used for training dataset. But when XGBoost performs continue
// training with an existing model, the prediction cache is not availbale and number
@@ -256,7 +264,7 @@ class GBTree : public GradientBooster {
(model_.param.num_trees != 0) &&
// FIXME(trivialfis): Implement a better method for testing whether data is on
// device after DMatrix refactoring is done.
(f_dmat && !((*(f_dmat->GetBatches<SparsePage>().begin())).data.DeviceCanRead()))) {
!on_device) {
return cpu_predictor_;
}
if (tparam_.predictor == "cpu_predictor") {