Mateusz P. Nowak
Mateusz P. Nowak
local_merge() function with calls to dpl::inplace_merge() is called to merge vector chunks coming from particular nodes
Joint work with @haichangsi
the dist_sort() function is too long to be clear, split it in reasonable way
A selection of OneDNN benchmarks added to the benchmark suite
The fix allows to use of _v1 and _v2 to select the l0 adapter and the use of arch to select the GPU, eg.: `llvm-lit --param "sycl_devices=level_zero_v2:arch-intel_gpu_mtl_u"` Necessary to run...