atb/infer_op_params.h
ActivationType
CommMode
FfnParam
FfnQuantParam
QuantType
ActivationOperation
AllGatherOperation
AllReduceOperation
AsStridedOperation
BroadcastOperation
ConcatOperation
CumsumOperation
ElewiseOperation
FillOperation
GatherOperation
IndexAddOperation
KVCacheOperation
LayerNormOperation
LinearOperation
LinearParallelOperation
LinearSparseOperation
MultinomialOperation
NonzeroOperation
OnehotOperation
PadOperation
PagedAttentionOperation
ReduceOperation
RepeatOperation
ReshapeAndCacheOperation
RmsNormOperation
RopeOperation
SelfAttentionOperation
SetValueOperation
SliceOperation
SoftmaxOperation
SortOperation
SplitOperation
TopkToppSamplingOperation
TransdataOperation
TransposeOperation
UnpadOperation
WhereOperation
父主题:
API参考