GitLab

Previously #6086 added ROCm support but after numerous rebases it lost
critical changes. This PR restores the ROCm build.

There are many source file changes but most were automated using the
following:

```bash
for f in `grep -rl '#ifdef USE_CUDA'`
do
    sed -i 's@#ifdef USE_CUDA@#if defined(USE_CUDA) || defined(USE_ROCM)@g' $f
done

for f in `grep -rl '#endif  // USE_CUDA'`
do
    sed -i 's@#endif  // USE_CUDA@#endif  // USE_CUDA || USE_ROCM@g' $f
done
```

* [ci] [c++] use 'pre-commit' to run 'cpplint', upgrade to 'cpplint' 2.0.2

* remove bashisms

* one more pipefail use

* another pipefail

* [ROCm] add support for ROCm/HIP

- CMakeLists.txt ROCm updates, also replace glob with explicit file list
- initial warpSize interop changes
- helpers/hipify.sh script added
- .gitignore to ignore generated hip source files

* more rocm updates

- disable compiler warnings
- move PercentileDevice __device__ template function into header
- bug fixes for __host__ __define__ and __HIP__ preprocessor symbols

* more bug fixes

* warp 32 vs 64 updates

* lint fixes

* missing device_index variable

* accidental inclusion of hip headers

* copyright notice compliance

* Update CMakeLists.txt
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* fix lint issue

* clean up

* Update CMakeLists.txt
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* Update CMakeLists.txt
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* clean up CMakeLists.txt

use WARPSIZE

* use WARPSIZE

* fix share buffer size

---------
Co-authored-by: shiyu1994 <shiyu_k1994@qq.com>
Co-authored-by: James Lamb <jaylamb20@gmail.com>
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>
Co-authored-by: Yu Shi <yushi2@microsoft.com>

* add quantized training (first stage)

* add histogram construction functions for integer gradients

* add stochastic rounding

* update docs

* fix compilation errors by adding template instantiations

* update files for compilation

* fix compilation of gpu version

* initialize gradient discretizer before share states

* add a test case for quantized training

* add quantized training for data distributed training

* Delete origin.pred

* Delete ifelse.pred

* Delete LightGBM_model.txt

* remove useless changes

* fix lint error

* remove debug loggings

* fix mismatch of vector and allocator types

* remove changes in main.cpp

* fix bugs with uninitialized gradient discretizer

* initialize ordered gradients in gradient discretizer

* disable quantized training with gpu and cuda

fix msvc compilation errors and warnings

* fix bug in data parallel tree learner

* make quantized training test deterministic

* make quantized training in test case more accurate

* refactor test_quantized_training

* fix leaf splits initialization with quantized training

* check distributed quantized training result

* add cuda gradient discretizer

* add quantized training for CUDA version in tree learner

* remove cuda computability 6.1 and 6.2

* fix parts of gpu quantized training errors and warnings

* fix build-python.sh to install locally built version

* fix memory access bugs

* fix lint errors

* mark cuda quantized training on cuda with categorical features as unsupported

* rename cuda_utils.h to cuda_utils.hu

* enable quantized training with cuda

* fix cuda quantized training with sparse row data

* allow using global memory buffer in histogram construction with cuda quantized training

* recover build-python.sh

enlarge allowed package size to 100M

* add l1 metric for cuda_exp

* add huber/fair metric for cuda_exp

* add poisson/mape/gamma/gamma_deviance/tweedie  metrics for cuda_exp

* fix cpplint error

* fix return  error

Co-authored-by: James Lamb <jaylamb20@gmail.com>

* [ci] speed up if-else, swig, and lint conda setup

* add 'source activate'

* python constraint

* start removing cuda v1

* comment out CI

* remove more references

* revert some unnecessaary changes

* revert a few more mistakes

* revert another change that ignored params

* sigh

* remove CUDATreeLearner

* fix tests, docs

* fix quoting in setup.py

* restore all CI

* Apply suggestions from code review
Co-authored-by: shiyu1994 <shiyu_k1994@qq.com>

* Apply suggestions from code review

* completely remove cuda_exp, update docs

---------
Co-authored-by: shiyu1994 <shiyu_k1994@qq.com>

* add rmse metric for new cuda version

* add Init for CUDAMetricInterface

* fix lint errors

* fix rmse and add l2 metric for new cuda version

* use CUDAL2Metric

* explicit template instantiation

* write result only with the first thread

* pre allocate buffer for output converting

* fix l2 regression with cuda metric evaluation

* weighting loss in cuda metric evaluation

* mark CUDATree::AsConstantTree as override

* add rmse metric for new cuda version

* add Init for CUDAMetricInterface

* fix lint errors

* initial work for boosting and evaluation with CUDA

* fix compatibility with CPU code

* fix creating objective without USE_CUDA_EXP

* fix static analysis errors

* fix static analysis errors

* Correct spelling

Most changes were in comments, and there were a few changes to literals for log output.

There were no changes to variable names, function names, IDs, or functionality.

* Clarify a phrase in a comment
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Clarify a phrase in a comment
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Clarify a phrase in a comment
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Correct spelling

Most are code comments, but one case is a literal in a logging message.

There are a few grammar fixes too.
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Add check to prevent out of index lookup in the position discount table. Add debug logging to report number of queries found in the data.

* Change debug logging location so that we can print the data file name as well.

* Revert "Change debug logging location so that we can print the data file name as well."

This reverts commit 3981b34bd6e0530f89c4733e78e6b6603bf50d48.

* Add data file name to debug logging.

* Move log line to a place where it is output even when query IDs are read from a separate file.

* Also add the out-of-range check to rank metrics.

* Perform check after number of queries is initialized.

* Update

* fix warnings

* Apply suggestions from code review

* Update feature_group.h

* Update feature_group.h

* Update src/treelearner/serial_tree_learner.cpp

* Update multiclass_metric.hpp

* Implement average precision score

* Fix lint errors

* Change name to average_precision

* Add to R-package list of metrics

* Empty commit to trigger CI jobs

* Change name to average_precision

* Update auc_mu metric to use data weights if provided

* Calculate class sizes and total weights in Init so we only do it once

* Fix lint error

* Empty commit to trigger CI jobs

This reverts commit 656d2676.

* don't cache `num_thread`, to avoid change outside

* rename

* update document

* Update docs/Parameters.rst

* Update include/LightGBM/config.h

* Apply suggestions from code review
Co-Authored-By: Nikita Titov <nekit94-08@mail.ru>

* Apply suggestions from code review
Co-Authored-By: Nikita Titov <nekit94-08@mail.ru>
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* some refines

* more omp refactoring

* format define

* fix merge bug

* some fixes

* fix some warnings

* Apply suggestions from code review

* Apply suggestions from code review

* remove dup codes

* Fix bug where small values of max_bin cause crash.

* Revert "Fix bug where small values of max_bin cause crash."

This reverts commit fe5c8e2547057c1fa5750bcddd359dd7708fab4b.

* Add auc-mu multiclass metric.

* Fix bug where scores are equal.

* Merge.

* Change name to auc_mu everywhere (instead of auc-mu).

* Fix comparison between signed and unsigned int.

* Change name to AUC-mu in docs and output messages.

* Improve test.

* Use prefix increment.

* Update R package.

* Fix style issues.

* Tidy up test code.

* Read all lines first then process.

* Allow passing AUC-mu weights directly as a list in parameters.

* Remove unused code, improve example and docs.

* fix many cpp lint errors

* indent

* fix bug

* fix more

* fix gpu

* more fixes

* fix metric alias

* fix format

* updated docs

* simplify alias in objective function

* move the alias parsing to config.cpp

* updated docs

* fix multi-class aliases

* updated regression aliases in docs

* fixed trailing space

* Implement top-k multiclass error metric. Add new parameter top_k_threshold.

* Add test for multiclass metrics

* Make test less sensitive to avoid floating-point issues.

* Change tabs to spaces.

* Fix problem with test in Python 2. Refactor to use np.testing. Decrease number of training rounds so loss is larger and easier to compare.

* Move multiclass tests into test_engine.py

* Change parameter name from top_k_threshold to multi_error_top_k.

* Fix top-k error metric to handle case where scores are equal. Update tests and docs.

* Change name of top-k metric to multi_error@k.

* Change tabs to spaces.

* Fix formatting.

* Fix minor issues in docs.

* added all necessary includes - fixed build/include_what_you_use error

* fixed the order of includes (build/include_order)