Version 0.2
A lot has changed since release 0.1! Here is a rough outline:
-
C++17 and CUDA 11
-
MSVC support
-
improved API using C++17 CTAD
-
improved integration with Alpaka
-
dump mapping visualizations
-
add experimental Trace and Split meta mappings
-
lots of refactoring and code improvements
-
greatly updated documentation
-
turn some examples into proper unit tests
-
add more unit tests
-
CI support with unit tests, address sanitizer, amalgamated llama.hpp, doxygen etc.
-
replace png++ by stb_image
-
added .clang-format file