XGBoost on OSX out-of-the-box

Hello everyone!

Can you advise if XGBoost should work on OSX if installed out of the box using just “pip install xgboost” command?

I’m using MBP15 with 2 GPUs, Python is 3.7.2. XGBoost comes from pip has version 0.9.0.

When I try to run it I got this error: “XGBoost version not compiled with GPU support”

More details:

/usr/local/lib/python3.7/site-packages/xgboost/core.py in _check_call(ret)
174 “”"
175 if ret != 0:
–> 176 raise XGBoostError(py_str(_LIB.XGBGetLastError()))
177
178

XGBoostError: [14:56:12] src/learner.cc:180: XGBoost version not compiled with GPU support.
Stack trace:
[bt] (0) 1 libxgboost.dylib 0x0000000125ad6579 dmlc::LogMessageFatal::~LogMessageFatal() + 57
[bt] (1) 2 libxgboost.dylib 0x0000000125ad9e70 xgboost::LearnerImpl::ConfigureUpdaters() + 2016
[bt] (2) 3 libxgboost.dylib 0x0000000125ad1971 xgboost::LearnerImpl::Configure(std::__1::vector<std::__1::pair<std::__1::basic_string<char, std::__1::char_traits, std::__1::allocator >, std::__1::basic_string<char, std::__1::char_traits, std::__1::allocator > >, std::__1::allocator<std::__1::pair<std::__1::basic_string<char, std::__1::char_traits, std::__1::allocator >, std::__1::basic_string<char, std::__1::char_traits, std::__1::allocator > > > > const&) + 3633
[bt] (3) 4 libxgboost.dylib 0x0000000125af12eb XGBoosterUpdateOneIter + 139
[bt] (4) 5 _ctypes.cpython-37m-darwin.so 0x000000010b80636f ffi_call_unix64 + 79
[bt] (5) 6 ??? 0x00007ffee7170a00 0x0 + 140732775467520

We do not support GPU training out-of-box on OSX. You may try manually compiling XGBoost from the source. In general, however, we do not provide any support or guarantee for GPUs on Mac OSX.