mxnet::OpContext Struct Reference

All the possible information needed by Operator.Forward and Backward This is the superset of RunContext. We use this data structure to bookkeep everything needed by Forward and Backward. More...

#include <op_attr_types.h>

Collaboration diagram for mxnet::OpContext:

Public Member Functions
template<typename xpu >
mshadow::Stream< xpu > *	get_stream () const
	get mshadow stream from Context More...

Public Attributes
bool	need_grad
	whether there is a backward phase to compute gradients. More...
bool	is_train
	whether it is training phase More...
RunContext	run_ctx
	RunContext related resources. More...
engine::CallbackOnComplete	async_on_complete
	the callback when operation completes, used by asynchronize ops More...
std::vector< Resource >	requested
	Resources requested by the operator. More...

Detailed Description

All the possible information needed by Operator.Forward and Backward This is the superset of RunContext. We use this data structure to bookkeep everything needed by Forward and Backward.

See also

Resource

Member Function Documentation

template<typename xpu >

mshadow::Stream<xpu>* mxnet::OpContext::get_stream ( ) const

inline

get mshadow stream from Context

Returns

the mshadow stream

Template Parameters

xpu	the device type of the stream

Member Data Documentation

engine::CallbackOnComplete mxnet::OpContext::async_on_complete

the callback when operation completes, used by asynchronize ops

bool mxnet::OpContext::is_train

whether it is training phase

bool mxnet::OpContext::need_grad

whether there is a backward phase to compute gradients.

std::vector<Resource> mxnet::OpContext::requested

Resources requested by the operator.

RunContext mxnet::OpContext::run_ctx

RunContext related resources.

---

The documentation for this struct was generated from the following file:

• include/mxnet/op_attr_types.h