You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
 
Megvii Engine Team 998f71a8c9 perf(mge/imperative): do not busy wait in imperative's queue 4 years ago
..
ops feat(imperative): add inplace add_update option in optimizer 4 years ago
proxy_graph perf(mge/imperative): add mini graph to partially replace proxy graph 4 years ago
backward_graph_opt.cpp perf(mge): add memory optimization for backward graph 4 years ago
blob_manager_impl.cpp feat(interpreter): add command buffer for inplace 4 years ago
blob_manager_impl.h refactor(mge): cpp license 4 years ago
dnn_op_helper.h feat(imperative): add inplace add_update option in optimizer 4 years ago
event_pool.cpp refactor(mge/profiler): refactor profiler 4 years ago
event_pool.h refactor(mge/profiler): refactor profiler 4 years ago
function_hook.h feat(interpreter): add command buffer for inplace 4 years ago
interpreter_impl.cpp feat(interpreter): add command buffer for inplace 4 years ago
interpreter_impl.h perf(mge/imperative): do not busy wait in imperative's queue 4 years ago
mgb_cg_impl.h ci(mge): fix Windows CI 4 years ago
op_def.cpp feat(interpreter): add command buffer for inplace 4 years ago
op_trait.cpp refactor(mge/imperative): implement new tensor system 4 years ago
op_trait.h feat(interpreter): add command buffer for inplace 4 years ago
opr_utility.cpp fix(mge/trace): fix sublinear in trace 4 years ago
physical_tensor.cpp perf(mge/imperative): do not busy wait in imperative's queue 4 years ago
profiler.cpp feat(interpreter): add command buffer for inplace 4 years ago
proxy_graph.cpp feat(imperative): auto generated opdef header and python binding 4 years ago
proxy_graph.h perf(mge/imperative): misc optimizations 4 years ago
proxy_graph_detail.cpp feat(interpreter): add command buffer for inplace 4 years ago
tensor_sanity_check.cpp feat(imperative): add tensor sanity check 4 years ago

MegEngine 安装包中集成了使用 GPU 运行代码所需的 CUDA 环境,不用区分 CPU 和 GPU 版。 如果想要运行 GPU 程序,请确保机器本身配有 GPU 硬件设备并安装好驱动。 如果你想体验在云端 GPU 算力平台进行深度学习开发的感觉,欢迎访问 MegStudio 平台

C++ Cuda Python C SVG other