You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

Leo Chen 3d015f1cf5 Set expected place in child thread for dataloader to avoid costing cuda memory on other card (#30338) ... * set expected place in child thread for dataloader * set device id when set tensor from numpy * revert tensor_py change * add compile guard * fix ci * fix bug		4 years ago
..
distributed	Fix/distributed proto (#29981)	4 years ago
framework	Recompute Offload (#30233)	4 years ago
imperative	Support unused parameters in dynamic graph distributed (#30224)	4 years ago
inference	Fix/distributed proto (#29981)	4 years ago
memory	optimize memcpy perf for kunlun (#30291)	4 years ago
operators	Recompute Offload (#30233)	4 years ago
platform	optimize memcpy perf for kunlun (#30291)	4 years ago
pybind	Set expected place in child thread for dataloader to avoid costing cuda memory on other card (#30338)	4 years ago
string	fix header file paths of gflags, commit 4, test=develop (#30274)	4 years ago
train	Add the strategy of skipping cc/cu test compilation and execution in CI (#29499)	4 years ago
.clang-format	…
API.spec	modify pipeline optimizer to only support the mode of sync pipeline training (#25065)	5 years ago
CMakeLists.txt	add ps table (#29463)	4 years ago