tokenizer and batcher are sane and work

i wonder how much pain it's gonna cause me to change
something in the code
This commit is contained in:
2019-11-30 22:16:42 -08:00
parent 04c35ed9b6
commit 101248965c
6 changed files with 321 additions and 156 deletions

View File

@@ -12,7 +12,7 @@ numpy_header = include_directories(run_command(
).stdout().strip())
executable(
'fedavg_mpi', 'main.c', 'cythoned/library.c',
'fedavg_mpi', 'main.c', 'cythoned/bridge.c',
dependencies: [mpi, python],
include_directories: numpy_header,
link_args: '-Wl,-w'