24 #ifndef TVM_TOPI_X86_INJECTIVE_H_
25 #define TVM_TOPI_X86_INJECTIVE_H_
49 if (axis.size() == 4) {
53 sch[out].parallel(fused);
54 }
else if (!axis.empty()) {
55 sch[out].parallel(axis[0]);
Managed reference class to TargetNode.
Definition: target.h:200
Array, container representing a contiguous sequence of ObjectRefs.
Definition: array.h:289
void push_back(const T &item)
push a new item to the back of the list
Definition: array.h:457
A Compute op that compute a tensor on certain domain.
Definition: operation.h:226
Managed reference to FuseNode.
Definition: schedule.h:826
Global schedule container For operations and all the operations they depend on. The schedule per Oper...
Definition: schedule.h:326
Tensor structure representing a possible input, or intermediate computation result.
Definition: tensor.h:102
Generic function that can be specialzied on a per target basis.
Tensor expression language DSL.
Definition: extracted_task.h:33
Schedule create_schedule(Array< Operation > ops)
Create a schedule for array of ops(and their dependencies).
Definition: schedule.h:702
void AutoInlineInjective(Schedule sch)
To automatically inline operations with injective writes (i.e. writes without reduction or sequential...
Schedule schedule_injective(const Target &target, const Array< Tensor > &outs)
Create an x86 schedule for the given injective ops.
Definition: injective.h:68
Schedule schedule_injective_from_existing(Schedule sch, const Tensor &out)
Updates an existing schedule for the given injective ops.
Definition: injective.h:47
runtime implementation for LibTorch/TorchScript.
Definition: analyzer.h:36
Operation node can generate one or multiple Tensors.