32 Commits

Author SHA1 Message Date
Jiaming Yuan
441ffc017a
Copy data from Ellpack to GHist. (#8215) 2022-09-06 23:05:49 +08:00
Jiaming Yuan
16bca5d4a1
Support CPU input for device QuantileDMatrix. (#8136)
- Copy `GHistIndexMatrix` to `Ellpack` when needed.
2022-08-11 21:21:26 +08:00
Jiaming Yuan
142a208a90
Fix compiler warnings. (#8022)
- Remove/fix unused parameters
- Remove deprecated code in rabit.
- Update dmlc-core.
2022-06-22 21:29:10 +08:00
Jiaming Yuan
c311a8c1d8
Enable compiling with system cub. (#7232)
- Tested with all CUDA 11.x.
- Workaround cub scan by using discard iterator in AUC.
- Limit the size of Argsort when compiled with CUDA cub.
2021-09-17 14:28:18 +08:00
Robert Maynard
1a75f43304
Allow compilation with nvcc 11.4 (#7131)
* Use type aliases for discard iterators

* update to include host_vector as thrust 1.12 doesn't bring it in as a side-effect

* cub::DispatchRadixSort requires signed offset types
2021-07-27 20:05:33 +08:00
Jiaming Yuan
bd1f3a38f0
Rewrite sparse dmatrix using callbacks. (#7092)
- Reduce dependency on dmlc parsers and provide an interface for users to load data by themselves.
- Remove use of threaded iterator and IO queue.
- Remove `page_size`.
- Make sure the number of pages in memory is bounded.
- Make sure the cache can not be violated.
- Provide an interface for internal algorithms to process data asynchronously.
2021-07-16 12:33:31 +08:00
Jiaming Yuan
1c8fdf2218
Remove use of device_idx in dh::LaunchN. (#7063)
It's an unused parameter, removing it can make the CI log more readable.
2021-06-29 11:37:26 +08:00
Jiaming Yuan
86715e4cd4
Support categorical data for dask functional interface and DQM. (#7043)
* Support categorical data for dask functional interface and DQM.

* Implement categorical data support for GPU GK-merge.
* Add support for dask functional interface.
* Add support for DQM.

* Get newer cupy.
2021-06-18 13:06:52 +08:00
Andrew Ziem
3e7e426b36
Fix spelling in documents (#6948)
* Update roxygen2 doc.

Co-authored-by: fis <jm.yuan@outlook.com>
2021-05-11 20:44:36 +08:00
Jiaming Yuan
a7083d3c13
Fix dart inplace prediction with GPU input. (#6777)
* Fix dart inplace predict with data on GPU, which might trigger a fatal check
for device access right.
* Avoid copying data whenever possible.
2021-03-25 12:00:32 +08:00
Rory Mitchell
29745c6df2
Fix inclusive scan for large sizes (#6234) 2020-11-03 17:01:43 +13:00
Jiaming Yuan
bed7ae4083
Loop over thrust::reduce. (#6229)
* Check input chunk size of dqdm.
* Add doc for current limitation.
2020-10-14 10:40:56 +13:00
Jiaming Yuan
14afdb4d92
Support categorical data in ellpack. (#6140) 2020-09-24 19:28:57 +08:00
Jiaming Yuan
dd445af56e
Cleanup on device sketch. (#5874)
* Remove old functions.

* Merge weighted and un-weighted into a common interface.
2020-07-14 10:15:54 +08:00
Jiaming Yuan
1a0801238e
Implement iterative DMatrix. (#5837) 2020-07-03 11:44:52 +08:00
Jiaming Yuan
cacff9232a
Remove column major specialization. (#5755)
Co-authored-by: Hyunsu Cho <chohyu01@cs.washington.edu>
2020-06-05 16:19:14 +08:00
Jiaming Yuan
e533908922
Expose device sketching in header. (#5747) 2020-06-02 13:02:53 +08:00
Jiaming Yuan
eaf2a00b5c
Enhance nvtx support. (#5636) 2020-05-06 22:54:24 +08:00
Jiaming Yuan
67d267f9da
Move device dmatrix construction code into ellpack. (#5623) 2020-05-06 19:43:59 +08:00
Rory Mitchell
e268fb0093
Use thrust functions instead of custom functions (#5544) 2020-04-16 21:41:16 +12:00
Rory Mitchell
13b10a6370
Device dmatrix (#5420) 2020-03-28 14:42:21 +13:00
Rory Mitchell
b745b7acce
Fix memory usage of device sketching (#5407) 2020-03-14 13:43:24 +13:00
Rory Mitchell
3ad4333b0e
Partial rewrite EllpackPage (#5352) 2020-03-11 10:15:53 +13:00
Rory Mitchell
a38e7bd19c
Sketching from adapters (#5365)
* Sketching from adapters

* Add weights test
2020-03-07 21:07:58 +13:00
Jiaming Yuan
655cf17b60
Predict on Ellpack. (#5327)
* Unify GPU prediction node.
* Add `PageExists`.
* Dispatch prediction on input data for GPU Predictor.
2020-02-23 06:27:03 +08:00
Rong Ou
e4b74c4d22
Gradient based sampling for GPU Hist (#5093)
* Implement gradient based sampling for GPU Hist tree method.
* Add samplers and handle compacted page in GPU Hist.
2020-02-04 10:31:27 +08:00
Rong Ou
0afcc55d98 Support multiple batches in gpu_hist (#5014)
* Initial external memory training support for GPU Hist tree method.
2019-11-16 14:50:20 +08:00
Jiaming Yuan
7663de956c
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.
2019-11-06 16:13:13 +08:00
Jiaming Yuan
6ec7e300bd
Fix external memory race in colmaker. (#4980)
* Move `GetColDensity` out of omp parallel block.
2019-10-25 04:11:13 -04:00
Rong Ou
5b1715d97c Write ELLPACK pages to disk (#4879)
* add ellpack source
* add batch param
* extract function to parse cache info
* construct ellpack info separately
* push batch to ellpack page
* write ellpack page.
* make sparse page source reusable
2019-10-22 23:44:32 -04:00
Rong Ou
562bb0ae31 remove device shards (#4867) 2019-09-25 13:15:46 +08:00
Rong Ou
125bcec62e Move ellpack page construction into DMatrix (#4833) 2019-09-16 23:50:55 -04:00