Use multi core to run TaskNode::ToProto (#4820)
* Serialize proto in binary rather than text * move del ops out from loop * refine * Skip GenCollectiveBoxingPlan if no CollectiveBoxingTaskNode * multi core to proto * copy pointers explicitly * make toproto const method * reorder * larger tol * Update test_layers_conv1d.py * fix deadlock * remove ForeignCallBack in Operator::ToOpAttribute Co-authored-by:oneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by:
liujuncheng <liujuncheng1022@gmail.com> Co-authored-by:
clackhan <han_binbin@163.com>
Showing
- oneflow/core/graph/compute_task_node.cpp 1 addition, 1 deletiononeflow/core/graph/compute_task_node.cpp
- oneflow/core/graph/compute_task_node.h 1 addition, 1 deletiononeflow/core/graph/compute_task_node.h
- oneflow/core/graph/task_node.cpp 1 addition, 1 deletiononeflow/core/graph/task_node.cpp
- oneflow/core/graph/task_node.h 1 addition, 1 deletiononeflow/core/graph/task_node.h
- oneflow/core/job/compiler.cpp 25 additions, 9 deletionsoneflow/core/job/compiler.cpp
- oneflow/core/operator/operator.cpp 14 additions, 4 deletionsoneflow/core/operator/operator.cpp
Please register or sign in to comment