'으으으으' 방언

운영

mhlo.abs (mhlo::AbsOp)

복근 작동

통사론:

operation ::= `mhlo.abs` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

operand 텐서에 대해 요소별 절대 연산을 수행하고 result 텐서를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#abs

예:

%result = mhlo.abs %operand : tensor<3xi32>

특성: AlwaysSpeculatableImplTrait , Elementwise , SameOperandsAndResultShape

인터페이스: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

피연산자:

결과:

mhlo.add (mhlo::AddOp)

작업 추가

통사론:

operation ::= `mhlo.add` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

두 텐서 lhs 및 rhs 의 요소별 추가를 수행하고 result 텐서를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#add

예:

%result = mhlo.add %lhs, %rhs : tensor<2x2xi32>

특성: AlwaysSpeculatableImplTrait , Commutative , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

인터페이스: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

피연산자:

결과:

mhlo.add_dependency (mhlo::AddDependencyOp)

AddDependency 작업

통사론:

operation ::= `mhlo.add_dependency` operands attr-dict `:` functional-type(operands, results)

이 작업은 XLA 컴파일러 전용이므로 아직 사양이 없습니다.

비공식적으로 이 작업은 두 개의 피연산자(데이터 피연산자와 토큰)로 이루어집니다. 연산의 출력은 데이터 피연산자입니다. AfterAll과 함께 사용하면 부작용이 없는 작업(토큰 값을 생성하지 않는 작업)을 주문할 수 있습니다.

예:

%1 = mhlo.add_dependency %arg0, %0 : (tensor<3x4xf32>, !mhlo.token) -> tensor<3x4xf32>

특성: AlwaysSpeculatableImplTrait

인터페이스: ConditionallySpeculatable , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

피연산자:

결과:

mhlo.after_all (mhlo::AfterAllOp)

AfterAll 작업

통사론:

operation ::= `mhlo.after_all` $inputs attr-dict
              `:` custom<VariadicSameOperandsAndResultType>(ref($inputs), type($inputs), type($result))

result 에 의존하는 작업보다 먼저 inputs 생성하는 작업이 실행되도록 합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#after_all

예:

%result = mhlo.after_all %input0, %input1 : !mhlo.token

특성: AlwaysSpeculatableImplTrait

인터페이스: ConditionallySpeculatable , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

피연산자:

결과:

mhlo.all_gather (mhlo::AllGatherOp)

AllGather 작업

프로세스 그리드의 각 프로세스 그룹 내에서 all_gather_dim 따라 각 프로세스의 피연산자 텐서 값을 연결하고 결과 텐서를 생성합니다. computation operands 의 각 피연산자에 대해 개별적으로 적용되어 피연산자당 하나의 결과를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#all_gather

예:

%result = "mhlo.all_gather"(%operand) {
  all_gather_dim = 1 : i64,
  replica_groups = dense<[[0, 1]]> : tensor<1x2xi64>
  // channel_id = 0
  channel_handle = #mhlo.channel_handle<handle = 0, type = 0>,
  // use_global_device_ids = false
} : (tensor<2x2xf32>) -> tensor<2x4xf32>

특성: SameOperandsAndResultElementType

속성:

피연산자:

결과:

mhlo.all_reduce (mhlo::AllReduceOp)

AllReduce 작업

프로세스 그리드의 각 프로세스 그룹 내에서 각 프로세스의 피연산자 텐서 값에 축소 함수 computation 적용하고 결과 텐서를 생성합니다. computation operands 의 각 피연산자에 대해 개별적으로 적용되어 피연산자당 하나의 결과를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#all_reduce

예:

%result = "mhlo.all_reduce"(%operand) ({
  ^bb0(%arg0: tensor<f32>, %arg1: tensor<f32>):
    %0 = mhlo.add %arg1, %arg2 : tensor<f32>
    mhlo.return %0 : tensor<f32>
}) {
  replica_groups = dense<[[0, 1]]> : tensor<1x2xi64>
  // channel_id = 0
  channel_handle = #mhlo.channel_handle<handle = 0, type = 0>
  // use_global_device_ids = false
} : (tensor<4xf32>) -> tensor<4xf32>

특성: InferTensorType , SingleBlockImplicitTerminator<ReturnOp> , SingleBlock

인터페이스: InferShapedTypeOpInterface , InferTypeOpInterface

속성:

피연산자:

결과:

mhlo.all_to_all (mhlo::AllToAllOp)

AllToAll 작업

프로세스 그리드의 각 프로세스 그룹 내에서 split_dimension 따라 operand 텐서의 값을 여러 부분으로 분할하고, 분할된 부분을 프로세스 간에 분산시키고, concat_dimension 따라 분산된 부분을 연결하여 result 텐서를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#all_to_all

예:

%result = "mhlo.all_to_all"(%operand) {
  split_dimension = 1 : i64,
  concat_dimension = 0 : i64,
  split_count = 2 : i64,
  replica_groups = dense<[[0, 1]]> : tensor<1x2xi64>
} : (tensor<2x4xf32>) -> tensor<4x2xf32>

특성: AlwaysSpeculatableImplTrait , InferTensorType , SameOperandsElementType , SameOperandsShape , SameVariadicOperandSize

인터페이스: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

속성:

피연산자:

결과:

mhlo.and (mhlo::AndOp)

그리고 작동

통사론:

operation ::= `mhlo.and` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

두 텐서 lhs 및 rhs 의 요소별 AND를 수행하고 result 텐서를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#and

예:

%result = mhlo.and %lhs, %rhs : tensor<2x2xi32>

특성: AlwaysSpeculatableImplTrait , Commutative , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

인터페이스: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

피연산자:

결과:

mhlo.async_done (mhlo::AsyncDoneOp)

AsyncDone 작업

이 작업은 XLA 컴파일러 전용이므로 아직 사양이 없습니다.

비공식적으로 이 작업은 비동기 계산이 끝날 때까지 차단됩니다. 비동기 계산의 최종 결과를 반환합니다.

자세한 내용은 AsyncStart 설명서를 참조하세요.

인터페이스: InferTypeOpInterface

속성:

피연산자:

결과:

mhlo.async_start (mhlo::AsyncStartOp)

AsyncStart 작업

이 작업은 XLA 컴파일러 전용이므로 아직 사양이 없습니다.

비공식적으로 이 작업은 비동기 계산을 시작합니다.

이는 비동기 대기(예: DMA)와 스레드 내 계산을 모두 포함하는 함수가 있을 때 사용됩니다. 예를 들어 함수는 계산, DMA, 다른 계산, 두 번째 DMA 및 최종 계산으로 구성될 수 있습니다. 이는 async_start, async_update 및 async_done으로 표시됩니다. async_start는 스레드에서 첫 번째 계산을 수행한 다음 DMA를 시작합니다. async_update는 DMA가 아직 완료되지 않은 경우 완료될 때까지 기다린 다음 함수에서 두 번째 계산을 실행하고 두 번째 DMA를 시작합니다. 마지막으로 async_done은 이 마지막 DMA를 기다린 다음 스레드에서 실행해야 하는 마지막 계산을 실행하고 해당 최종 계산의 결과를 반환합니다.

operands called_computation 계산에 직접 전달됩니다._계산은 비동기적으로 실행될 함수입니다. execution_thread 는 실행될 스레드의 이름입니다. 메인 스레드를 "main"이라고 합니다. 모든 스레드에는 이름이 있습니다.

이는 비동기 작업 간에 필요한 모든 상태를 반환합니다. 버퍼 할당 후 반환 값은 입력, 결과 및 비동기 작업에 필요하거나 편집된 스크래치패드를 보관하는 데 필요한 공간을 나타냅니다.

속성:

피연산자:

결과:

mhlo.async_update (mhlo::AsyncUpdateOp)

AsyncUpdate 작업

이 작업은 XLA 컴파일러 전용이므로 아직 사양이 없습니다.

비공식적으로 이 작업은 동기화 장벽이 나타날 때까지 비동기 계산을 차단합니다. 이는 작업을 수행한 후 bundle 반환합니다.

자세한 내용은 AsyncStart 설명서를 참조하세요.

인터페이스: InferTypeOpInterface

속성:

피연산자:

결과:

mhlo.atan2 (mhlo::Atan2Op)

Atan2 작업

통사론:

operation ::= `mhlo.atan2` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

lhs 및 rhs 텐서에 대해 요소별 atan2 작업을 수행하고 result 텐서를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#atan2

예:

%result = mhlo.atan2 %lhs, %rhs : tensor<3xf32>

특성: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

인터페이스: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

피연산자:

결과:

mhlo.batch_norm_grad (mhlo::BatchNormGradOp)

BatchNormGrad 작업

grad_output 에서 역전파되는 BatchNormTrainingOp의 여러 입력에 대한 기울기를 계산하고 grad_operand , grad_scale 및 grad_offset 텐서를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#batch_norm_grad

예:

%grad_operand, %grad_scale, %grad_offset =
"mhlo.batch_norm_grad"(%operand, %scale, %mean, %variance, %grad_output) {
  epsilon = 0.0 : f32,
  feature_index = 2 : i64
} : (tensor<2x2x2xf32>, tensor<2xf32>, tensor<2xf32>, tensor<2xf32>,
    tensor<2x2x2xf32>) -> (tensor<2x2x2xf32>, tensor<2xf32>, tensor<2xf32>)

특성: AlwaysSpeculatableImplTrait , InferTensorType

인터페이스: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

속성:

피연산자:

결과:

mhlo.batch_norm_inference (mhlo::BatchNormInferenceOp)

BatchNormInference 작업

feature_index 차원을 제외한 모든 차원에서 operand 텐서를 정규화하고 result 텐서를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#batch_norm_inference

예:

%result = "mhlo.batch_norm_inference"(%operand, %scale, %offset, %mean, %variance) {
  epsilon = 0.0 : f32,
  feature_index = 2 : i64
} : (tensor<2x2x2xf32>, tensor<2xf32>, tensor<2xf32>, tensor<2xf32>, tensor<2xf32>) -> tensor<2x2x2xf32>

특성: AlwaysSpeculatableImplTrait , InferTensorType

인터페이스: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

속성:

피연산자:

결과:

mhlo.batch_norm_training (mhlo::BatchNormTrainingOp)

BatchNorm훈련 작업

배치 및 공간 차원 전체에서 평균과 분산을 계산하고 feature_index 차원의 각 기능에 대해 operand 텐서를 정규화하고 output , batch_mean 및 batch_var 텐서를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#batch_norm_training

예:

%output, %batch_mean, %batch_var = "mhlo.batch_norm_training"(%operand, %scale, %offset) {
  epsilon = 0.0 : f32,
  feature_index = 2 : i64
} : (tensor<2x2x2xf32>, tensor<2xf32>, tensor<2xf32>) -> (tensor<2x2x2xf32>, tensor<2xf32>, tensor<2xf32>)

특성: AlwaysSpeculatableImplTrait , InferTensorType

인터페이스: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

속성:

피연산자:

결과:

mhlo.bitcast (mhlo::BitcastOp)

비트캐스트 작업

통사론:

operation ::= `mhlo.bitcast` operands attr-dict `:` functional-type(operands, results)

이 작업은 XLA 컴파일러 전용이므로 아직 사양이 없습니다.

비공식적으로 이 작업은 요소의 물리적 배열이 변경되지 않는 방식으로 입력의 모양을 변경합니다.

이 작업에는 "요소의 물리적 배열"을 이해하기 위한 레이아웃 정보가 필요하며 MHLO의 레이아웃 지원은 현재 진행 중인 작업입니다.

예:

%0 = mhlo.bitcast %arg0 : (tensor<3x4xf32>) -> tensor<3x4x1xf32>

특성: AlwaysSpeculatableImplTrait

인터페이스: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

피연산자:

결과:

mhlo.bitcast_convert (mhlo::BitcastConvertOp)

BitcastConvert 작업

통사론:

operation ::= `mhlo.bitcast_convert` operands attr-dict `:` functional-type(operands, results)

operand 텐서에 대해 비트캐스트 연산을 수행하고 result 텐서의 유형을 사용하여 전체 operand 텐서의 비트가 재해석되는 result 텐서를 생성합니다.

참조: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#bitcast_convert

예:

%result = mhlo.bitcast_convert %operand : (tensor<2xf32>) -> tensor<2x4xi8>

특성: AlwaysSpeculatableImplTrait

인터페이스: ConditionallySpeculatable , InferShapedTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과: MemoryEffects::Effect{}

피연산자:

결과:

mhlo.broadcast (mhlo :: broadcastop)

방송 작업

이 작업은 Scablehlo에서 벗어나기 때문에 사양에 포함되지 않습니다 : https://github.com/openxla/stablehlo/issues/3

비공식적 으로이 작업은 XLA의 방송과 동일한 작업을 수행합니다 : https://www.tensorflow.org/xla/operation_semantics#broadcast

예:

%result = mhlo.broadcast %operand, sizes = [1, 2] : (tensor<3xi32>) -> tensor<1x2x3xi32>

특성 : AlwaysSpeculatableImplTrait , InferTensorType , SameOperandsAndResultElementType

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

속성:

피연산자 :

결과:

mhlo.broadcast_in_dim (mhlo :: broadcastindimop)

방송 운영

operand 텐서의 데이터를 복제하여 입력 텐서의 치수 및/또는 순위를 확장하고 result 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#broadcast_in_dim 을 참조하십시오

예:

%result = mhlo.broadcast_in_dim %operand, dims = [2, 1] : (tensor<1x3xi32>) -> tensor<2x3x2xi32>

특성 : AlwaysSpeculatableImplTrait , HLO_CompatibleOperandsAndResultElementType

인터페이스 : ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

속성:

피연산자 :

결과:

mhlo.case (mhlo :: caseop)

케이스 작동

index 값에 따라 branches 에서 정확히 하나의 function 실행하여 출력을 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#case 를 참조하십시오

예:

%result0, %result1 = "mhlo.case"(%index) ({
  mhlo.return %result_branch0, %result_branch0 : tensor<2xi64>, tensor<2xi64>
}, {
  mhlo.return %result_branch1, %result_branch1 : tensor<2xi64>, tensor<2xi64>
}) : (tensor<i32>) -> (tensor<2xi64>, tensor<2xi64>)

TRACITS : RecursiveMemoryEffects , SingleBlockImplicitTerminator<ReturnOp> , SingleBlock

인터페이스 : InferTypeOpInterface

피연산자 :

결과:

mhlo.cbrt (mhlo :: cbrtop)

CBRT 운영

통사론:

operation ::= `mhlo.cbrt` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

operand 텐서에서 요소 별 입방 루트 작동을 수행하고 result 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#cbrt 를 참조하십시오

예:

%result = mhlo.cbrt %operand : tensor<4xf32>

특성 : AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

피연산자 :

결과:

mhlo.ceil (mhlo :: Ceilop)

천장 작동

통사론:

operation ::= `mhlo.ceil` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

operand 텐서의 원소 현저한 천장을 수행하고 result 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#ceil 을 참조하십시오

예:

%result = mhlo.ceil %operand : tensor<5xf32>

특성 : AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

피연산자 :

결과:

mhlo.cholesky (mhlo :: Choleskyop)

Cholesky Operation

매트릭스 배치의 Cholesky 분해를 계산합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#cholesky 를 참조하십시오

예:

%result = mhlo.cholesky %a, lower = true : tensor<3x3xf32>

특성 : AlwaysSpeculatableImplTrait , InferTensorType , SameOperandsAndResultElementType

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

속성:

피연산자 :

결과:

mhlo.clamp (mhlo :: 클램프)

클램프 작동

통사론:

operation ::= `mhlo.clamp` $min `,` $operand `,` $max attr-dict
              `:` custom<SameOperandsAndResultType>(type($min), type($operand), type($max), type($result))

operand 텐서의 모든 요소를 ​​최소와 최대 값 사이에서 클램핑하고 result 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#clamp 를 참조하십시오

예:

%result = mhlo.clamp %min, %operand, %max : tensor<3xi32>

특성 : AlwaysSpeculatableImplTrait , HLO_BroadcastingElementwise , InferTensorType , SameOperandsAndResultElementType

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

피연산자 :

결과:

mhlo.collective_broadcast (mhlo :: CollectiveBroadcastop)

집단 브로드 캐스트 운영

프로세스 그리드의 각 프로세스 그룹 내에서 operand 텐서의 값을 소스 프로세스에서 대상 프로세스로 보내고 result 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#collective_broadcast 를 참조하십시오

예:

%result = "mhlo.collective_broadcast"(%operand) {
  replica_groups = dense<[[0, 1]]> : tensor<1x2xi64>,
  channel_handle = #mhlo.channel_handle<handle = 0, type = 0>
} : (tensor<1x2xi64>) -> tensor<1x2xi64>

특성 : CompatibleOperandsAndResultType

인터페이스 : InferShapedTypeOpInterface , InferTypeOpInterface

속성:

피연산자 :

결과:

mhlo.collective_permute (mhlo :: CollectivePermuteop)

집단 관리 운영

프로세스 그리드의 각 프로세스 그룹 내에서, operand 텐서의 값을 소스 프로세스에서 대상 프로세스로 보내고 result 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#collective_permute 를 참조하십시오

예:

%result = "mhlo.collective_permute"(%operand) {
  source_target_pairs = dense<[[0, 1], [1, 2]]> : tensor<2x2xi64>,
  // channel_id = 0
  channel_handle = #mhlo.channel_handle<handle = 0, type = 0>
} : (tensor<4x2xf32>) -> tensor<4x2xf32>

특성 : AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

속성:

피연산자 :

결과:

mhlo.compare (mhlo :: compraleop)

작업을 비교하십시오

통사론:

operation ::= `mhlo.compare` $comparison_direction `,` $lhs `,` $rhs (`,` $compare_type^)?
              attr-dict `:` functional-type(operands, results)

comparison_direction 및 compare_type 에 따라 lhs 및 rhs 텐서의 요소 별 비교를 수행하고 result 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#compare 를 참조하십시오

예:

%result = mhlo.compare LT, %lhs, %rhs, FLOAT : (tensor<2xf32>, tensor<2xf32>) -> tensor<2xi1>

특성 : AlwaysSpeculatableImplTrait , Elementwise , InferTensorType , SameOperandsAndResultShape , SameOperandsElementType

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

속성:

피연산자 :

결과:

mhlo.complex (mhlo :: complexop)

복잡한 작동

통사론:

operation ::= `mhlo.complex` operands attr-dict
              `:` custom<ComplexOpType>(type($lhs), type($rhs), type($result))

lhs 및 rhs 한 쌍의 실제 및 가상 값 쌍에서 복잡한 값으로 요소 별 변환을 수행하고 result 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#complex 를 참조하십시오

예:

%result = mhlo.complex %lhs, %rhs : tensor<2xcomplex<f32>>

특성 : AlwaysSpeculatableImplTrait , Elementwise , SameOperandsAndResultShape , SameOperandsElementType

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

피연산자 :

결과:

mhlo.composite (mhlo :: compositeop)

복합 작업

통사론:

operation ::= `mhlo.composite` $name $inputs attr-dict `:` functional-type(operands, results)

다른 안정화 작업의 작업 (구성)을 캡슐화하고 inputs 및 composite_attributes 취하고 results 생성합니다. OP의 의미론은 decomposition 속성에 의해 구현됩니다. composite OP는 프로그램 의미를 변경하지 않고 분해로 대체 할 수 있습니다. 분해를 인라인으로 만드는 것이 동일한 OP 시맨틱을 제공하지 않는 경우 custom_call 사용하는 것이 좋습니다.

version 필드 (기본값 0에서 0 )는 복합 시맨틱이 변경 될 때를 나타냅니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#composite 를 참조하십시오

예:

%results = mhlo.composite "my.op" %arg0, %arg1 {
  decomposition = @my_op,
  composite_attributes = { my_attribute = "my_value" },
  version = 1 : i32
} : (tensor<f32>, tensor<f32>) -> tensor<f32>

인터페이스 : SymbolUserOpInterface

속성:

피연산자 :

결과:

mhlo.compute_reshape_shape (mhlo :: computershapheshapeop)

컴퓨터 샤핑 운영

통사론:

operation ::= `mhlo.compute_reshape_shape` operands attr-dict `:` functional-type(operands, results)

이 작업은 진행중인 작업이므로 아직 사양에 포함되어 있지 않습니다 : https://github.com/openxla/stablehlo/issues/8

비공식적 으로이 작업은 DynamicReshapeop의 오페라에서 num_elements 수의 요소 수와 TF의 reshape에 제공되는 dynamic_shape 모양의 요소 수에 대한 output_shape를 계산합니다 : https://www.tensorflow.org/api_docs/python/tf/reshape

예를 들어, num_elements = 12 및 dynamic_shape = [2, -1] 의 경우 result 는 [2, 6] 가 될 것입니다. 피연산자가 유효하지 않은 경우 (예 : 치수가 요소 수를 고르게 나누지 않거나 차원에 여러 값이있는 경우) 정의되지 않은 동작으로 이어집니다.

예:

%result = mhlo.compute_reshape_shape %num_elements, %dynamic_shape
       : (index, tensor<2xi32>) -> tensor<2xi32>

특성 : AlwaysSpeculatableImplTrait

인터페이스 : ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

피연산자 :

결과:

mhlo.concatenate (mhlo :: concatenateop)

연결 작업

주어진 인수와 동일한 순서로 dimension 차원을 따라 inputs 에서 다양한 수의 텐서를 연결하고 result 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#concatenate 를 참조하십시오

예:

%result = mhlo.concatenate %input0, %input1, dim = 0 : (tensor<3x2xi64>, tensor<1x2xi64>) -> tensor<4x2xi64>

특성 : AlwaysSpeculatableImplTrait , SameOperandsAndResultElementType

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

속성:

피연산자 :

결과:

mhlo.constant (mhlo :: constantop)

일정한 작동

일정한 value 에서 output 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#constant 를 참조하십시오

예:

%output = mhlo.constant dense<[[0.0, 1.0], [2.0, 3.0]]> : tensor<2x2xf32>

특성 : AlwaysSpeculatableImplTrait , ConstantLike

인터페이스 : ConditionallySpeculatable , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

속성:

결과:

mhlo.convert (mhlo :: convertop)

운영을 변환하십시오

통사론:

operation ::= `mhlo.convert` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

operand 텐서에서 한 요소 유형에서 다른 요소 유형으로 요소가있는 변환을 수행하고 result 텐서를 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#convert 를 참조하십시오

예:

%result = mhlo.convert %operand : (tensor<3xi32>) -> tensor<3xcomplex<f32>>

특성 : AlwaysSpeculatableImplTrait , Elementwise , SameOperandsAndResultShape

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

피연산자 :

결과:

mhlo.convolution (mhlo :: Convolutionop)

컨볼 루션 작업

통사론:

operation ::= `mhlo.convolution` `(`operands`)`
              `dim_numbers` `=` custom<ConvolutionDimensions>($dimension_numbers) `,`
              `window` `=` `{` custom<WindowAttributes>($window_strides, $padding,
              $lhs_dilation, $rhs_dilation,
              $window_reversal) `}`
              attr-dict `:` functional-type(operands, results)

lhs 의 창과 rhs 조각 사이에 도트 제품을 계산하고 result 생성합니다.

https://github.com/openxla/stablehlo/blob/main/docs/spec.md#convolution 을 참조하십시오

예:

%result = "mhlo.convolution"(%lhs, %rhs) {
  window_strides = dense<4> : tensor<2xi64>,
  padding = dense<0> : tensor<2x2xi64>,
  lhs_dilation = dense<2> : tensor<2xi64>,
  rhs_dilation = dense<1> : tensor<2xi64>,
  window_reversal = dense<false> : tensor<2xi1>,
  dimension_numbers = #mhlo.conv<[b, 0, 1, f]x[0, 1, i, o]->[b, 0, 1, f]>,
  feature_group_count = 1 : i64,
  batch_group_count = 1 : i64,
  precision_config = [#stablehlo<precision DEFAULT>, #stablehlo<precision DEFAULT>]
} : (tensor<1x4x4x1xi32>, tensor<3x3x1x1xi32>) -> tensor<1x2x2x1xi32>

특성 : AlwaysSpeculatableImplTrait

인터페이스 : ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

속성:

피연산자 :

결과:

mhlo.copy (mhlo :: copyop)

복사 작업

통사론:

operation ::= `mhlo.copy` operands attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

이 작업은 XLA 컴파일러에 비공개이므로 아직 사양이 없습니다.

비공식적 으로이 작업은 operand 사본입니다. 작업에 부착 된 메타 데이터에 따라 NO-OP와 상당히 다르게 작동 할 수 있습니다.

예:

%0 = mhlo.copy %arg0 : tensor<f32>

특성 : AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise

인터페이스 : ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

효과 : MemoryEffects::Effect{}

속성:

피연산자 :

결과:

mhlo.cosine (mhlo::CosineOp)

Cosine operation

통사론:

operation ::= `mhlo.cosine` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise cosine operation on operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#cosine

예:

%result = mhlo.cosine %operand : tensor<2xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.count_leading_zeros (mhlo::ClzOp)

Clz operation

통사론:

operation ::= `mhlo.count_leading_zeros` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise count of the number of leading zero bits in the operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#count_leading_zeros

예:

%result = mhlo.count_leading_zeros %operand : tensor<2x2xi8>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.create_token (mhlo::CreateTokenOp)

CreateToken operation

통사론:

operation ::= `mhlo.create_token` attr-dict `:` type(results)

This operation is on its way out of StableHLO, so it is not included in the specification: https://github.com/openxla/stablehlo/issues/3

Informally, this operation does the same thing as AfterAllOp with 0 inputs: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#after_all

예:

%output = mhlo.create_token : !mhlo.token

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

결과:

mhlo.cross-replica-sum (mhlo::CrossReplicaSumOp)

CrossReplicaSum operation

This operation is on its way out of StableHLO, so it is not included in the specification: https://github.com/openxla/stablehlo/issues/3

Informally, this operation does the same thing as AllReduceOp with channel_id = 0 , use_global_device_ids = false and computation implementing addition: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#all_reduce

예:

%result = "mhlo.cross-replica-sum"(%operand) {
  replica_groups = dense<[[0, 1]]> : tensor<1x2xi64>
} : (tensor<4xf32>) -> tensor<4xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.cstr_reshapable (mhlo::CstrReshapableOp)

CstrReshapable operation

통사론:

operation ::= `mhlo.cstr_reshapable` operands attr-dict `:` functional-type(operands, results)

This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/8

Informally, this operation creates a witness on the constraint that ComputeReshapeShape would succeed with the provided operands.

예:

%result = mhlo.cstr_reshapable %num_elements, %dynamic_shape
       : (index, tensor<3xi32>) -> !shape.witness

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.custom_call (mhlo::CustomCallOp)

CustomCall operation

통사론:

operation ::= `mhlo.custom_call` custom<CustomCallTarget>($call_target_name) `(` $inputs `)`
              attr-dict `:` functional-type(operands, results)

Encapsulates an implementation-defined operation call_target_name that takes inputs and called_computations and produces results .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#custom_call

예:

%results = "mhlo.custom_call"(%input0) {
  call_target_name = "foo",
  has_side_effect = false,
  backend_config = "bar",
  api_version = 1 : i32,
  called_computations = [@foo]
} : (tensor<f32>) -> tensor<f32>

A custom call invokes code external to XLA. The `inputs` are passed to the
external code, and the external code is expected to produce a result of the
given type. The exact mechanism is backend-specific. For example, in the CPU
backend, a call instruction is emitted which targets a symbol with the name
`call_target_name`.

If XLA runtime is enabled for a backend, then custom calls use the runtime
custom call calling convention to call into the external functions. This
calling convention defines an ABI for encoding arguments, attributes and
results.

Depending on the API version there are two ways to pass extra bits of static
information to the external function:

1. For `API_VERSION_TYPED_FFI` custom calls `backend_config` must be a
   dictionary attribute, that will be encoded according to the custom call
   calling convention and passed to the external function as the attributes
   argument. External code is expected to use declarative bindings (see
   `xla/runtime/custom_call.h`) to decode them at run time. These custom
   calls are only supported if XLA uses XLA runtime.

2. For previous API versions it is the user responsibility to encode extra
   bits of static information as a string `backend_config` attribute, and
   decode it at run time.

Interfaces: MemoryEffectOpInterface

속성:

Operands:

결과:

mhlo.divide (mhlo::DivOp)

Div operation

통사론:

operation ::= `mhlo.divide` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise division of dividend lhs and divisor rhs tensors and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#divide

예:

%result = mhlo.divide %lhs, %rhs : tensor<4xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.domain (mhlo::DomainOp)

Domain operation

This operation is private to the XLA compiler, so it is does not yet have a specification.

Informally, these operations are used to group instructions with the same DomainMetadata property. ShardingMetadata is the main use case today to group instructions on the same device. Domain instructions provide two major benefits:

• Prevent unintentionally optimizing instructions across domains.
• Automatically assign the metadata of the instructions created in the domain. Without domain instructions, each HLO optimization pass would have to check and propagate the metadata, which would be easy to miss and also adds complexity to the compiler. Since domain instructions connect two different domains, each domain instruction is associated with two DomainMetadata -- one on the operand side and one on the user side of the domain.

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.dot (mhlo::DotOp)

Dot operation

This operation is on its way out of StableHLO, so it is not included in the specification: https://github.com/openxla/stablehlo/issues/3

Informally, this operation does the same thing as XLA's Dot: https://www.tensorflow.org/xla/operation_semantics#dot

예:

%0 = mhlo.dot %arg0, %arg1 : (tensor<1x2xi32>, tensor<2x1xi32>) -> tensor<1x1xi32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.dot_general (mhlo::DotGeneralOp)

DotGeneral operation

Computes dot products between slices of lhs and slices of rhs and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#dot_general

예:

%result = "mhlo.dot_general"(%lhs, %rhs) {
  dot_dimension_numbers = #mhlo.dot<
    lhs_batching_dimensions = [0],
    rhs_batching_dimensions = [0],
    lhs_contracting_dimensions = [2],
    rhs_contracting_dimensions = [1]
  >,
  precision_config = [#stablehlo<precision DEFAULT>, #stablehlo<precision DEFAULT>]
} : (tensor<2x2x2xi32>, tensor<2x2x2xi32>) -> tensor<2x2x2xi32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.dynamic_broadcast_in_dim (mhlo::DynamicBroadcastInDimOp)

DynamicBroadcastInDim operation

This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/8

Informally, this operation does the same thing as BroadcastInDimOp except that the result shape is specified dynamically via output_dimensions : https://github.com/openxla/stablehlo/blob/main/docs/spec.md#broadcast_in_dim

It also accepts optional attributes to express static knowledge about the expanding behavior of dimensions. If not specified, all dimensions are assumed to be possibly expanding. The sets of dimensions that are known to be expanding and the set of dimensions that are known to be non-expanding must be disjoint and they must be a subset of the operand's dimensions.

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.dynamic_conv (mhlo::DynamicConvOp)

DynamicConv operation

This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/8

Informally, this operation does the same thing as ConvolutionOp except that padding is specified dynamically via d_padding : https://github.com/openxla/stablehlo/blob/main/docs/spec.md#convolution

예:

%result = "mhlo.dynamic_conv"(%lhs, %rhs, %d_padding) {
  window_strides = dense<4> : tensor<2xi64>,
  lhs_dilation = dense<2> : tensor<2xi64>,
  rhs_dilation = dense<1> : tensor<2xi64>,
  window_reversal = dense<false> : tensor<2xi1>,
  dimension_numbers = #mhlo.conv<[b, 0, 1, f]x[0, 1, i, o]->[b, 0, 1, f]>,
  feature_group_count = 1 : i64,
  batch_group_count = 1 : i64,
  precision_config = [#stablehlo<precision DEFAULT>, #stablehlo<precision DEFAULT>]
} : (tensor<1x4x4x1xi32>, tensor<3x3x1x1xi32>, tensor<2x2xi64>) -> tensor<1x2x2x1xi32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.dynamic_gather (mhlo::DynamicGatherOp)

DynamicGather operation

This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/8

Informally, this operation does the same thing as GatherOp except that slice_sizes are specified dynamically: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#gather

예:

%result = "mhlo.dynamic_gather"(%operand, %start_indices, %slice_sizes) {
  dimension_numbers = #mhlo.gather<
    offset_dims = [2, 3],
    collapsed_slice_dims = [0],
    start_index_map = [0, 2],
    index_vector_dim = 2>,
  indices_are_sorted = false
} : (tensor<3x4x2xi32>, tensor<2x3x2xi64>, tensor<3xi64>) -> tensor<2x3x2x2xi32>

Traits: AlwaysSpeculatableImplTrait , InferTensorType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.dynamic_iota (mhlo::DynamicIotaOp)

DynamicIota operation

This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/8

Informally, this operation does the same thing as IotaOp except that the result shape is specified dynamically via output_shape : https://github.com/openxla/stablehlo/blob/main/docs/spec.md#iota

예:

%0 = mhlo.dynamic_iota %arg0, dim = 0 : (tensor<1xindex>) -> tensor<4xi32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.dynamic_pad (mhlo::DynamicPadOp)

DynamicPad operation

통사론:

operation ::= `mhlo.dynamic_pad` operands attr-dict `:` functional-type(operands, results)

Dynamically Pads the operand , with amount of padding added at low-end/high-end/interior is passed through input tensors.

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.dynamic_reshape (mhlo::DynamicReshapeOp)

DynamicReshape operation

통사론:

operation ::= `mhlo.dynamic_reshape` operands attr-dict `:` functional-type(operands, results)

This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/8

Informally, this operation does the same thing as ReshapeOp except that the result shape is specified dynamically via output_shape : https://github.com/openxla/stablehlo/blob/main/docs/spec.md#reshape

예:

%0 = mhlo.dynamic_reshape %arg0, %shape : (tensor<?xf32>, tensor<2xindex>) -> tensor<?x?xf32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.dynamic_slice (mhlo::DynamicSliceOp)

DynamicSlice operation

Extracts a slice from the operand using dynamically-computed starting indices and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#dynamic_slice

예:

%result = mhlo.dynamic_slice %operand, %start_indices0, %start_indices1, sizes = [2, 2]
  : (tensor<4x4xi32>, tensor<i64>, tensor<i64>) -> tensor<2x2xi32>

Traits: AlwaysSpeculatableImplTrait , InferTensorType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.dynamic_update_slice (mhlo::DynamicUpdateSliceOp)

DynamicUpdateSlice operation

통사론:

operation ::= `mhlo.dynamic_update_slice` operands attr-dict `:` functional-type(operands, results)

Produces a result tensor which is equal to the operand tensor except that the slice starting at start_indices is updated with the values in update .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#dynamic_update_slice

예:

%result = mhlo.dynamic_update_slice %operand, %update, %start_indices0, %start_indices1
  : (tensor<4x4xi32>, tensor<2x2xi32>, tensor<i64>, tensor<i64>) -> tensor<4x4xi32>

Traits: AlwaysSpeculatableImplTrait , InferTensorType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.einsum (mhlo::EinsumOp)

Einsum operation

This operation is on its way out of StableHLO, so it is not included in the specification: https://github.com/openxla/stablehlo/issues/3

Informally, this operation does the same thing as TF's einsum: https://www.tensorflow.org/api_docs/python/tf/einsum

예:

%result = "mhlo.einsum"(%lhs, %rhs) {
  einsum_config = "ab,bc->ac"
} : (tensor<4x16xf32>, tensor<16x4xf32>) -> tensor<4x4xf32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.erf (mhlo::ErfOp)

Erf operation

통사론:

operation ::= `mhlo.erf` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise erf operation on operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#erf

예:

%result = mhlo.erf %operand : tensor<2x2xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.exponential (mhlo::ExpOp)

Exp operation

통사론:

operation ::= `mhlo.exponential` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise exponential operation on operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#exponential

예:

%result = mhlo.exponential %operand : tensor<2x2xf64>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.exponential_minus_one (mhlo::Expm1Op)

Expm1 operation

통사론:

operation ::= `mhlo.exponential_minus_one` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise exponential minus one operation on operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#exponential_minus_one

예:

%result = mhlo.exponential_minus_one %operand : tensor<2xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.fft (mhlo::FftOp)

Fft operation

Performs the forward and inverse Fourier transforms for real and complex inputs/outputs.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#fft

예:

%result = mhlo.fft %operand, type = FFT, length = [4] : (tensor<4xcomplex<f32>>) -> tensor<4xcomplex<f32>>

Traits: AlwaysSpeculatableImplTrait , InferTensorType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.floor (mhlo::FloorOp)

Floor operation

통사론:

operation ::= `mhlo.floor` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise floor of operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#floor

예:

%result = mhlo.floor %operand : tensor<2xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.fusion (mhlo::FusionOp)

Fusion operation

This operation is private to the XLA compiler, so it is does not yet have a specification.

Informally, this operation consists of a group of basic ops (represented as a region attached to it). It serves as a hint to the backend that it is beneficial to emit the contained ops into a single loop nest or kernel.

속성:

Operands:

결과:

mhlo.gather (mhlo::GatherOp)

Gather operation

Gathers slices from operand tensor from offsets specified in start_indices and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#gather

예:

%result = "mhlo.gather"(%operand, %start_indices) {
  dimension_numbers = #mhlo.gather<
    offset_dims = [2, 3],
    collapsed_slice_dims = [0],
    start_index_map = [0, 2],
    index_vector_dim = 2>,
  slice_sizes = dense<[0, 2, 2]> : tensor<3xi64>,
  indices_are_sorted = false
} : (tensor<3x4x2xi32>, tensor<2x3x2xi64>) -> tensor<2x3x2x2xi32>

Traits: AlwaysSpeculatableImplTrait , InferTensorType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.get_dimension_size (mhlo::GetDimensionSizeOp)

GetDimensionSize operation

Produces the size of the given dimension of the operand .

See https://github.com/openxla/stablehlo/blob/main/docs/spec.md#get_dimension_size

예:

%result = mhlo.get_dimension_size %operand, dim = 1 : (tensor<2x3xf32>) -> tensor<i32>

Traits: AlwaysSpeculatableImplTrait , InferTensorType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.get_tuple_element (mhlo::GetTupleElementOp)

GetTupleElement operation

통사론:

operation ::= `mhlo.get_tuple_element` $operand `[` $index `]` attr-dict `:` functional-type(operands, results)

Extracts element at index position of the operand tuple and produces a result .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#get_tuple_element

예:

%result = mhlo.get_tuple_element %operand[0] : (tuple<tensor<2xf32>, tuple<tensor<i32>>>) -> tensor<2xf32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.if (mhlo::IfOp)

If operation

Produces the output from executing exactly one branch from true_branch or false_branch depending on the value of pred .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#if

Example: %result = "mhlo.if"(%pred) ({ "mhlo.return"(%result_true_branch) : (tensor ) -> () }, { "mhlo.return"(%result_false_branch) : (tensor ) -> () }) : (tensor ) -> tensor

Traits: RecursiveMemoryEffects , SingleBlockImplicitTerminator<ReturnOp> , SingleBlock

Interfaces: InferTypeOpInterface

Operands:

결과:

mhlo.imag (mhlo::ImagOp)

Imag operation

통사론:

operation ::= `mhlo.imag` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Extracts the imaginary part, element-wise, from the operand and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#imag

예:

%result = mhlo.imag %operand : (tensor<2xcomplex<f32>>) -> tensor<2xf32>

Traits: AlwaysSpeculatableImplTrait , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.infeed (mhlo::InfeedOp)

Infeed operation

Reads data from the infeed and produces results .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#infeed

예:

%results:2 = "mhlo.infeed"(%token) {
  infeed_config = ""
} : (!mhlo.token) -> (tensor<3x3x3xi32>, !mhlo.token)

속성:

Operands:

결과:

mhlo.iota (mhlo::IotaOp)

Iota operation

Fills an output tensor with values in increasing order starting from zero along the iota_dimension dimension.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#iota

예:

%output = mhlo.iota dim = 0 : tensor<4x5xi32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

결과:

mhlo.is_finite (mhlo::IsFiniteOp)

IsFinite operation

통사론:

operation ::= `mhlo.is_finite` $x attr-dict `:` functional-type(operands, results)

Performs element-wise check whether the value in x is finite (ie is neither +Inf, -Inf, nor NaN) and produces a y tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#is_finite

예:

%y = mhlo.is_finite %x : (tensor<7xf32>) -> tensor<7xi1>

Traits: AlwaysSpeculatableImplTrait , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.log (mhlo::LogOp)

Log operation

통사론:

operation ::= `mhlo.log` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise logarithm operation on operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#log

예:

%result = mhlo.log %operand : tensor<2x2xf64>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.log_plus_one (mhlo::Log1pOp)

Log1p operation

통사론:

operation ::= `mhlo.log_plus_one` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise logarithm plus one operation on operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#log_plus_one

예:

%result = mhlo.log_plus_one %operand : tensor<6xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.logistic (mhlo::LogisticOp)

Logistic operation

통사론:

operation ::= `mhlo.logistic` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise logistic operation on operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#logistic

예:

%result = mhlo.logistic %operand : tensor<2x2xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.map (mhlo::MapOp)

Map operation

Applies a map function computation to inputs along the dimensions and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#map

예:

%result = "mhlo.map"(%input0, %input1) ({
  ^bb0(%arg0: tensor<i32>, %arg1: tensor<i32>):
    %0 = mhlo.multiply %arg0, %arg1 : tensor<i32>
    mhlo.return %0 : tensor<i32>
}) {
  dimensions = dense<[0, 1]> : tensor<2xi64>
} : (tensor<2x2xi32>, tensor<2x2xi32>) -> tensor<2x2xi32>

Traits: InferTensorType , RecursiveMemoryEffects , SameOperandsAndResultShape , SingleBlockImplicitTerminator<ReturnOp> , SingleBlock

Interfaces: InferShapedTypeOpInterface , InferTypeOpInterface

속성:

Operands:

결과:

mhlo.maximum (mhlo::MaxOp)

Max operation

통사론:

operation ::= `mhlo.maximum` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise max operation on tensors lhs and rhs and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#maximum

예:

%result = mhlo.maximum %lhs, %rhs : tensor<4xf32>

Traits: AlwaysSpeculatableImplTrait , Commutative , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.minimum (mhlo::MinOp)

Min operation

통사론:

operation ::= `mhlo.minimum` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise min operation on tensors lhs and rhs and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#minimum

예:

%result = mhlo.minimum %lhs, %rhs : tensor<4xf32>

Traits: AlwaysSpeculatableImplTrait , Commutative , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.multiply (mhlo::MulOp)

Mul operation

통사론:

operation ::= `mhlo.multiply` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise product of two tensors lhs and rhs and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#multiply

예:

%result = mhlo.multiply %lhs, %rhs : tensor<2xi32>

Traits: AlwaysSpeculatableImplTrait , Commutative , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.negate (mhlo::NegOp)

Neg operation

통사론:

operation ::= `mhlo.negate` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise negation of operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#negate

예:

%result = mhlo.negate %operand : tensor<2x3xi32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.not (mhlo::NotOp)

Not operation

통사론:

operation ::= `mhlo.not` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise NOT of tensor operand of type integer and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#not

예:

%result = mhlo.not %operand : tensor<5x3x1xi1>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.optimization_barrier (mhlo::OptimizationBarrierOp)

OptimizationBarrier operation

통사론:

operation ::= `mhlo.optimization_barrier` attr-dict ($operand^ `:` custom<PairwiseOpType>(type($operand), type($result))):(`(` `)`)?

Ensures that the operations that produce the operand are executed before any operations that depend on the result and prevents compiler transformations from moving operations across the barrier. Other than that, the operation is an identity, ie result = operand .

See https://github.com/openxla/stablehlo/blob/main/docs/spec.md#optimization_barrier

예:

%result0, %result1 = mhlo.optimization_barrier %operand0, %operand1 : tensor<f32>, tensor<f32>

Traits: AlwaysSpeculatableImplTrait , HLO_PairwiseSameOperandAndResultType

Interfaces: ConditionallySpeculatable , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.or (mhlo::OrOp)

Or operation

통사론:

operation ::= `mhlo.or` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise OR of two tensors lhs and rhs and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#or

예:

%result = mhlo.or %lhs, %rhs : tensor<2xi1>

Traits: AlwaysSpeculatableImplTrait , Commutative , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.outfeed (mhlo::OutfeedOp)

Outfeed operation

Writes inputs to the outfeed and produces a result token.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#outfeed

예:

%result = "mhlo.outfeed"(%input0, %token) {
  outfeed_config = ""
} : (tensor<3x3x3xi32>, !mhlo.token) -> !mhlo.token

Interfaces: InferTypeOpInterface

속성:

Operands:

결과:

mhlo.pad (mhlo::PadOp)

Pad operation

Expands operand by padding around the tensor as well as between the elements of the tensor with the given padding_value .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#pad

예:

%0 = mhlo.pad %arg0, %arg1, low = [0, 1], high = [2, 1], interior = [1, 2]
  : (tensor<2x3xi32>, tensor<i32>) -> tensor<5x9xi32>

Traits: AlwaysSpeculatableImplTrait , SameOperandsAndResultElementType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.partition_id (mhlo::PartitionIdOp)

PartitionId operation

통사론:

operation ::= `mhlo.partition_id` attr-dict `:` type(results)

Produces partition_id of the current process.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#partition_id

예:

%result = mhlo.partition_id : tensor<ui32>

Interfaces: InferTypeOpInterface

결과:

mhlo.popcnt (mhlo::PopulationCountOp)

PopulationCount operation

통사론:

operation ::= `mhlo.popcnt` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise count of the number of bits set in the operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#popcnt

예:

%result = mhlo.popcnt %operand : tensor<4xi8>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.power (mhlo::PowOp)

Pow operation

통사론:

operation ::= `mhlo.power` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise exponentiation of lhs tensor by rhs tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#power

예:

%result = mhlo.power %lhs, %rhs : tensor<6xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.real (mhlo::RealOp)

Real operation

통사론:

operation ::= `mhlo.real` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Extracts the real part, element-wise, from the operand and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#real

예:

%result = mhlo.real %operand : (tensor<2xcomplex<f32>>) -> tensor<2xf32>

Traits: AlwaysSpeculatableImplTrait , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.real_dynamic_slice (mhlo::RealDynamicSliceOp)

RealDynamicSlice operation

통사론:

operation ::= `mhlo.real_dynamic_slice` operands attr-dict `:` functional-type(operands, results)

This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/8

Informally, this operation does the same thing as SliceOp except that start_indices , limit_indices and strides are specified dynamically: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#slice

예:

%result = mhlo.real_dynamic_slice %operand,
            %start_indices, %limit_indices, %strides
       : (tensor<256x?xf32>, tensor<2xindex>, tensor<2xindex>, tensor<2xindex>) -> tensor<256x?xf32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.recv (mhlo::RecvOp)

Recv operation

Receives data from a channel with channel_id and produces results .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#recv

예:

%results:2 = "mhlo.recv"(%token) {
  // channel_id = 5 : i64,
  // channel_type = #stablehlo<channel_type HOST_TO_DEVICE>,
  channel_handle = #mhlo.channel_handle<handle = 5, type = 3>,
  is_host_transfer = true
} : (!mhlo.token) -> (tensor<3x4xi32>, !mhlo.token)

속성:

Operands:

결과:

mhlo.reduce (mhlo::ReduceOp)

Reduce operation

Applies a reduction function body to inputs and init_values along the dimensions and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#reduce

예:

%result = "mhlo.reduce"(%input, %init_value) ({
  ^bb0(%arg0: tensor<i32>, %arg1: tensor<i32>):
    %0 = "mhlo.add"(%arg0, %arg1) : (tensor<i32>, tensor<i32>) -> tensor<i32>
    "mhlo.return"(%0) : (tensor<i32>) -> ()
}) {
  dimensions = dense<1> : tensor<1xi64>
} : (tensor<1x6xi32>, tensor<i32>) -> tensor<1xi32>

Traits: InferTensorType , RecursiveMemoryEffects , SameVariadicOperandSize , SingleBlockImplicitTerminator<ReturnOp> , SingleBlock

Interfaces: InferShapedTypeOpInterface , InferTypeOpInterface

속성:

Operands:

결과:

mhlo.reduce_precision (mhlo::ReducePrecisionOp)

ReducePrecision operation

통사론:

operation ::= `mhlo.reduce_precision` $operand `,` `format` `=` custom<ExponentMantissa>($exponent_bits, $mantissa_bits)
              attr-dict `:` custom<SameOperandsAndResultType>(type($operand), type($output))

Performs element-wise conversion of operand to another floating-point type that uses exponent_bits and mantissa_bits and back to the original floating-point type and produces an output tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#reduce_precision

예:

%output = mhlo.reduce_precision %operand, format = e5m2 : tensor<6xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.reduce_scatter (mhlo::ReduceScatterOp)

ReduceScatter operation

Within each process group in the process grid, performs reduction, using computations , over the values of the operand tensor from each process, splits the reduction result along scatter_dimension into parts, and scatters the split parts between the processes to produce the result .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#reduce_scatter

예:

%result = "mhlo.reduce_scatter"(%operand) ({
  ^bb0(%arg0: tensor<f32>, %arg1: tensor<f32>):
  %0 = mhlo.add %arg0, %arg1 : tensor<f32>
  mhlo.return %0 : tensor<f32>
}) {
  scatter_dimension = 1 : i64,
  replica_groups = dense<[[0, 1]]> : tensor<1x2xi64>,
  // channel_id = 0
  channel_handle = #mhlo.channel_handle<handle = 0, type = 0>
  // use_global_device_ids = false
} : (tensor<2x4xf32>) -> tensor<2x2xf32>

속성:

Operands:

결과:

mhlo.reduce_window (mhlo::ReduceWindowOp)

ReduceWindow operation

Applies a reduction function body to windows of inputs and init_values and produces results .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#reduce_window

예:

%result = "mhlo.reduce_window"(%input, %init_value) ({
  ^bb0(%arg0: tensor<i32>, %arg1: tensor<i32>):
    %0 = mhlo.add %arg0, %arg1 : tensor<i32>
    mhlo.return %0 : tensor<i32>
}) {
  window_dimensions = dense<[2, 1]> : tensor<2xi64>,
  window_strides = dense<[4, 1]> : tensor<2xi64>,
  base_dilations = dense<[2, 1]> : tensor<2xi64>,
  window_dilations = dense<[3, 1]> : tensor<2xi64>,
  padding = dense<[[2, 1], [0, 0]]> : tensor<2x2xi64>
} : (tensor<3x2xi32>, tensor<i32>) -> tensor<2x2xi32>

Traits: InferTensorType , RecursiveMemoryEffects , SameVariadicOperandSize , SingleBlockImplicitTerminator<ReturnOp> , SingleBlock

Interfaces: InferShapedTypeOpInterface , InferTypeOpInterface

속성:

Operands:

결과:

mhlo.remainder (mhlo::RemOp)

Rem operation

통사론:

operation ::= `mhlo.remainder` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise remainder of dividend lhs and divisor rhs tensors and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#remainder

예:

%result = mhlo.remainder %lhs, %rhs : tensor<4xi64>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.replica_id (mhlo::ReplicaIdOp)

ReplicaId operation

통사론:

operation ::= `mhlo.replica_id` attr-dict `:` type(results)

Produces replica_id of the current process.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#replica_id

예:

%result = mhlo.replica_id : tensor<ui32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

결과:

mhlo.reshape (mhlo::ReshapeOp)

Reshape operation

통사론:

operation ::= `mhlo.reshape` operands attr-dict `:` functional-type(operands, results)

Performs reshape of operand tensor to a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#reshape

예:

%result = mhlo.reshape %operand : (tensor<2xf32>) -> tensor<1x2xf32>

Traits: AlwaysSpeculatableImplTrait , HLO_CompatibleOperandsAndResultElementType

Interfaces: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.return (mhlo::ReturnOp)

_This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/425

Informally, this operation serves as a terminator for regions defined by
the StableHLO ops. Non-StableHLO ops, e.g. `func.func`, have their own
terminators, e.g. `func.return`.

Example:

    ```mlir
    %result = "mhlo.reduce"(%input, %init_value) ({
      ^bb0(%arg0: tensor<i32>, %arg1: tensor<i32>):
        %0 = "mhlo.add"(%arg0, %arg1) : (tensor<i32>, tensor<i32>) -> tensor<i32>
        "mhlo.return"(%0) : (tensor<i32>) -> ()
    }) {
      dimensions = dense<1> : tensor<1xi64>
    } : (tensor<1x6xi32>, tensor<i32>) -> tensor<1xi32>
    ```_


Syntax:

```

operation ::= mhlo.return $results attr-dict ( : type($results)^)?

Traits: `AlwaysSpeculatableImplTrait`, `Terminator`

Interfaces: `ConditionallySpeculatable`, `NoMemoryEffect (MemoryEffectOpInterface)`

Effects: `MemoryEffects::Effect{}`

#### Operands:

| Operand | Description |
| :-----: | ----------- |
| `results` | variadic of ranked tensor of f8E4M3B11FNUZ type or f8E4M3FN type or f8E4M3FNUZ type or f8E5M2 type or f8E5M2FNUZ type or 16-bit float or 32-bit float or 64-bit float or bfloat16 type or pred (AKA boolean or 1-bit integer) or 4/8/16/32/64-bit signless integer or 4/8/16/32/64-bit unsigned integer or complex type with 32-bit float or 64-bit float elements or 4/8/16/32-bit uniform quantized signed integer or 4/8/16/32-bit uniform quantized unsigned integer or 4/8/16/32-bit uniform quantized per axis signed integer or 4/8/16/32-bit uniform quantized per axis unsigned integer values or token or nested tuple with any combination of ranked tensor of f8E4M3B11FNUZ type or f8E4M3FN type or f8E4M3FNUZ type or f8E5M2 type or f8E5M2FNUZ type or 16-bit float or 32-bit float or 64-bit float or bfloat16 type or pred (AKA boolean or 1-bit integer) or 4/8/16/32/64-bit signless integer or 4/8/16/32/64-bit unsigned integer or complex type with 32-bit float or 64-bit float elements or 4/8/16/32-bit uniform quantized signed integer or 4/8/16/32-bit uniform quantized unsigned integer values or ranked tensor of 4/8/16/32-bit uniform quantized per axis signed integer or 4/8/16/32-bit uniform quantized per axis unsigned integer values or token values


### `mhlo.reverse` (mhlo::ReverseOp)

_Reverse operation_

Reverses the order of elements in the `operand` along the specified
`dimensions` and produces a `result` tensor.

See:
<a href="https://github.com/openxla/stablehlo/blob/main/docs/spec.md#reverse">https://github.com/openxla/stablehlo/blob/main/docs/spec.md#reverse</a>

Example:
```mlir
%result = mhlo.reverse %operand, dims = [1] : tensor<3x2xi32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.rng (mhlo::RngOp)

Rng operation

Generates random numbers using the rng_distribution algorithm and produces a result tensor of a given shape shape .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#rng

예:

%result = mhlo.rng %a, %b, %shape, distribution = NORMAL : (tensor<i32>, tensor<i32>, tensor<2xi64>) -> tensor<3x3xi32>

Traits: InferTensorType

Interfaces: InferShapedTypeOpInterface , InferTypeOpInterface

속성:

Operands:

결과:

mhlo.rng_bit_generator (mhlo::RngBitGeneratorOp)

RngBitGenerator operation

Returns an output filled with uniform random data and an updated output state output_state given an initial state initial_state using the pseudorandom number generator algorithm rng_algorithm .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#rng_bit_generator

예:

%output_state, %output = mhlo.rng_bit_generator %initial_state, algorithm = THREE_FRY : (tensor<2xui64>) -> (tensor<2xui64>, tensor<2x2xui64>)

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.round_nearest_afz (mhlo::RoundOp)

Round operation

통사론:

operation ::= `mhlo.round_nearest_afz` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise rounding towards the nearest integer, breaking ties away from zero, on the operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#round_nearest_afz

예:

%result = mhlo.round_nearest_afz %operand : tensor<5xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.round_nearest_even (mhlo::RoundNearestEvenOp)

RoundNearestEven operation

통사론:

operation ::= `mhlo.round_nearest_even` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise rounding towards the nearest integer, breaking ties towards the even integer, on the operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#round_nearest_even

예:

%result = mhlo.round_nearest_even %operand : tensor<5xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.rsqrt (mhlo::RsqrtOp)

Rsqrt operation

통사론:

operation ::= `mhlo.rsqrt` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise reciprocal square root operation on operand tensor and produces a result tensor, implementing the rSqrt operation from the IEEE-754 specification.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#rsqrt

예:

%result = mhlo.rsqrt %operand : tensor<2x2xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.scatter (mhlo::ScatterOp)

Scatter operation

Produces results tensors which are equal to inputs tensors except that several slices specified by scatter_indices are updated with the values updates using update_computation .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#scatter

예:

%result = "mhlo.scatter"(%input, %scatter_indices, %update) ({
  ^bb0(%arg0: tensor<i32>, %arg1: tensor<i32>):
    %0 = mhlo.add %arg0, %arg1 : tensor<i32>
    mhlo.return %0 : tensor<i32>
}) {
  scatter_dimension_numbers = #mhlo.scatter<
    update_window_dims = [2,3],
    inserted_window_dims = [0],
    scatter_dims_to_operand_dims = [1, 0],
    index_vector_dim = 2>,
  indices_are_sorted = false,
  unique_indices = false
} : (tensor<3x4x2xi32>, tensor<2x3x2xi64>, tensor<2x3x2x2xi32>) -> tensor<3x4x2xi32>

Traits: RecursiveMemoryEffects , SameVariadicOperandSize

Interfaces: InferTypeOpInterface

속성:

Operands:

결과:

mhlo.select (mhlo::SelectOp)

Select operation

통사론:

operation ::= `mhlo.select` operands attr-dict `:`
              custom<SelectOpType>(type($pred), type($on_true), type($on_false), type($result))

Produces a result tensor where each element is selected from on_true or on_false tensor based on the value of the corresponding element of pred .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#select

예:

%result = mhlo.select %pred, %on_true, %on_false : tensor<2x2xi1>, tensor<2x2xi32>

Traits: AlwaysSpeculatableImplTrait , HLO_BroadcastingElementwise , InferTensorType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.select_and_scatter (mhlo::SelectAndScatterOp)

SelectAndScatter operation

Scatters the values from the source tensor using scatter based on the outcome of reduce_window of the input tensor using select and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#select_and_scatter

예:

%result = "mhlo.select_and_scatter"(%operand, %source, %init_value) ({
  ^bb0(%arg0: tensor<i32>, %arg1: tensor<i32>):
    %0 = "mhlo.compare"(%arg0, %arg1) {
      comparison_direction = #stablehlo<comparison_direction GE>
    } : (tensor<i32>, tensor<i32>) -> tensor<i1>
    "mhlo.return"(%0) : (tensor<i1>) -> ()
}, {
  ^bb0(%arg0: tensor<i32>, %arg1: tensor<i32>):
    %0 = "mhlo.add"(%arg0, %arg1) : (tensor<i32>, tensor<i32>) -> tensor<i32>
    "mhlo.return"(%0) : (tensor<i32>) -> ()
}) {
  window_dimensions = dense<[3, 1]> : tensor<2xi64>,
  window_strides = dense<[2, 1]> : tensor<2xi64>,
  padding = dense<[[0, 1], [0, 0]]> : tensor<2x2xi64>
} : (tensor<4x2xi32>, tensor<2x2xi32>, tensor<i32>) -> tensor<4x2xi32>

Traits: RecursiveMemoryEffects

Interfaces: InferTypeOpInterface

속성:

Operands:

결과:

mhlo.send (mhlo::SendOp)

Send operation

Sends inputs to a channel channel_id and produces a result token.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#send

예:

%result = "mhlo.send"(%operand, %token) {
  // channel_id = 5 : i64,
  // channel_type = #stablehlo<channel_type DEVICE_TO_HOST>,
  channel_handle = #mhlo.channel_handle<handle = 5, type = 2>,
  is_host_transfer = true
} : (tensor<3x4xi32>, !mhlo.token) -> !mhlo.token

Interfaces: InferTypeOpInterface

속성:

Operands:

결과:

mhlo.set_dimension_size (mhlo::SetDimensionSizeOp)

SetDimensionSize operation

This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/8

Informally, this operation does the same thing as XLA's SetDimensionSize: https://www.tensorflow.org/xla/operation_semantics#setdimensionsize

예:

%0 = mhlo.set_dimension_size %arg0, %arg1, dim = 1 : (tensor<4x2xf32>, tensor<i32>) -> tensor<4x2xf32>

Traits: AlwaysSpeculatableImplTrait , InferTensorType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.shift_left (mhlo::ShiftLeftOp)

ShiftLeft operation

통사론:

operation ::= `mhlo.shift_left` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise left-shift operation on the lhs tensor by rhs number of bits and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#shift_left

예:

%result = mhlo.shift_left %lhs, %rhs : tensor<6xi8>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.shift_right_arithmetic (mhlo::ShiftRightArithmeticOp)

ShiftRightArithmetic operation

통사론:

operation ::= `mhlo.shift_right_arithmetic` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise arithmetic right-shift operation on the lhs tensor by rhs number of bits and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#shift_right_arithmetic

예:

%result = mhlo.shift_right_arithmetic %lhs, %rhs : tensor<6xi8>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.shift_right_logical (mhlo::ShiftRightLogicalOp)

ShiftRightLogical operation

통사론:

operation ::= `mhlo.shift_right_logical` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise logical right-shift operation on the lhs tensor by rhs number of bits and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#shift_right_logical

예:

%result = mhlo.shift_right_logical %lhs, %rhs : tensor<6xi8>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.sign (mhlo::SignOp)

Sign operation

통사론:

operation ::= `mhlo.sign` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Returns the sign of the operand element-wise and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#sign

예:

%result = mhlo.sign %operand : tensor<7xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.sine (mhlo::SineOp)

Sine operation

통사론:

operation ::= `mhlo.sine` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise sine operation on operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#sine

예:

%result = mhlo.sine %operand : tensor<2xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.slice (mhlo::SliceOp)

Slice operation

Extracts a slice from the operand using statically-computed starting indices and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#slice

예:

%result = "mhlo.slice" (%operand) {
  start_indices = dense<[1, 2]> : tensor<2xi64>,
  limit_indices = dense<[3, 4]> : tensor<2xi64>,
  strides = dense<1> : tensor<2xi64>
} : (tensor<3x4xi64>) -> tensor<2x2xi64>

Traits: AlwaysSpeculatableImplTrait , SameOperandsAndResultElementType

Interfaces: ConditionallySpeculatable , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.sort (mhlo::SortOp)

Sort operation

Sorts a variadic number of tensors in inputs together, according to a custom comparator , along the given dimension and produces a variadic number of tensors as results .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#sort

예:

%result0, %result1 = "mhlo.sort"(%input0, %input1) ({
  ^bb0(%arg0: tensor<i32>, %arg1: tensor<i32>, %arg2: tensor<i32>, %arg3: tensor<i32>):
    %predicate = "mhlo.compare"(%arg0, %arg1) {
      comparison_direction = #stablehlo<comparison_direction GT>
      } : (tensor<i32>, tensor<i32>) -> tensor<i1>
    "mhlo.return"(%predicate) : (tensor<i1>) -> ()
}) {
  dimension = 0 : i64,
  is_stable = true
} : (tensor<2x3xi32>, tensor<2x3xi32>) -> (tensor<2x3xi32>, tensor<2x3xi32>)

Traits: InferTensorType , RecursiveMemoryEffects , SameOperandsAndResultShape

Interfaces: InferShapedTypeOpInterface , InferTypeOpInterface

속성:

Operands:

결과:

mhlo.sparse_dot (mhlo::SparseDotOp)

Sparse dot operation

Similar to dot_general operation, with one or both of the operands being sparse. An additional argument provides sparsity meta information. Disclaimer: this op is experimental / a work in progress.

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.sqrt (mhlo::SqrtOp)

Sqrt operation

통사론:

operation ::= `mhlo.sqrt` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise square root operation on operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#sqrt

예:

%result = mhlo.sqrt %operand : tensor<2x2xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.stochastic_convert (mhlo::StochasticConvertOp)

StochasticConvert operation

This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/295

Informally, this operation performs element-wise conversion of values from a bigger type to a smaller one with stochastic rounding using the random number passed in.

Traits: AlwaysSpeculatableImplTrait , Elementwise

Interfaces: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.subtract (mhlo::SubtractOp)

Subtract operation

통사론:

operation ::= `mhlo.subtract` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise subtraction of two tensors lhs and rhs and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#subtract

예:

%result = mhlo.subtract %lhs, %rhs : tensor<2xi32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.tan (mhlo::TanOp)

Tan operation

통사론:

operation ::= `mhlo.tan` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

This operation is a work in progress, so it is not yet included in the specification: https://github.com/openxla/stablehlo/issues/954

Informally, this operation returns Tan(operand) element-wise.

예:

%0 = mhlo.tan %arg0 : tensor<2xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.tanh (mhlo::TanhOp)

Tanh operation

통사론:

operation ::= `mhlo.tanh` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise hyperbolic tangent operation on operand tensor and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#tanh

예:

%result = mhlo.tanh %operand : tensor<2xf32>

Traits: AlwaysSpeculatableImplTrait , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.topk (mhlo::TopKOp)

TopK operation

통사론:

operation ::= `mhlo.topk` `(`$operand `,` `k` `=` $k (`,` `largest` `=` $largest^)? `)` attr-dict `:`
              type($operand) `->` `(`type($values)`,` type($indices)`)`

Returns top k values and their indices, along the last dimension of the operand if largest=true or the bottom k values if largest=false .

See: https://www.tensorflow.org/xla/operation_semantics#top-k

예:

%values, %indices = mhlo.topk(%operand, k=5, largest=true)
  : tensor<100xf32> -> (tensor<5xf32>, tensor<5xi32>)

Traits: InferTensorType , RecursiveMemoryEffects

Interfaces: InferShapedTypeOpInterface , InferTypeOpInterface

속성:

Operands:

결과:

mhlo.torch_index_select (mhlo::TorchIndexSelectOp)

TorchIndexSelect operation

This operation is on its way out of StableHLO, so it is not included in the specification: https://github.com/openxla/stablehlo/issues/3

Informally, this operation does the same thing as PyTorch's index_select, augmented with support for batch dimensions: https://pytorch.org/docs/stable/generated/torch.index_select.html

The batch_dims attribute specifies the number of major batch dimensions (0 or more) that act like a multidimensional loop over both the operand and the index.

예:

%result = "mhlo.torch_index_select"(%operand, %index) {
  dim = 2 : i64,
  batch_dims = 1 : i64
} : (tensor<8x128x3072x64xf32>, tensor<8x16x1024xi32>) -> tensor<8x128x16x1024x64xf32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.trace (mhlo::TraceOp)

Trace operation

통사론:

operation ::= `mhlo.trace` $operand `,` $tag attr-dict `:` type($operand)

This operation is on its way out of StableHLO, so it is not included in the specification: https://github.com/openxla/stablehlo/issues/604

It is not used by JAX, PyTorch or TensorFlow, so it looks like we should've classified it as "Private to XLA" and not included it in StableHLO in the first place. With that in mind, its semantics will not be documented here.

예:

mhlo.trace %arg0, "In test code." : tensor<5x1x5xi32>

속성:

Operands:

mhlo.transpose (mhlo::TransposeOp)

Transpose operation

Permutes the dimensions of operand tensor using permutation and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#transpose

예:

%0 = mhlo.transpose %arg0, dims = [2, 1, 0] : (tensor<1x2x3xi32>) -> tensor<3x2x1xi32>

Traits: AlwaysSpeculatableImplTrait , HLO_CompatibleOperandsAndResultElementType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.triangular_solve (mhlo::TriangularSolveOp)

TriangularSolve operation

Solves batches of systems of linear equations with lower or upper triangular coefficient matrices.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#triangular_solve

예:

%result = "mhlo.triangular_solve"(%a, %b) {
  left_side = true,
  lower = true,
  unit_diagonal = false,
  transpose_a = #stablehlo<transpose NO_TRANSPOSE>
} : (tensor<3x3xf32>, tensor<3x3xf32>) -> tensor<3x3xf32>

Traits: AlwaysSpeculatableImplTrait , InferTensorType , SameOperandsAndResultElementType

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.tuple (mhlo::TupleOp)

Tuple operation

통사론:

operation ::= `mhlo.tuple` $val attr-dict `:` custom<TupleOpType>(type($val), type($result))

Produces a result tuple from values val .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#tuple

예:

%result = mhlo.tuple %val0, %val1 : tuple<tensor<2xf32>, tuple<tensor<i32>>>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.unary_einsum (mhlo::UnaryEinsumOp)

UnaryEinsum operation

This operation is on its way out of StableHLO, so it is not included in the specification: https://github.com/openxla/stablehlo/issues/3

Informally, this operation does the same thing as TF's einsum: https://www.tensorflow.org/api_docs/python/tf/einsum

예:

%result = "mhlo.unary_einsum"(%operand) {
  einsum_config = "ab->a"
} : (tensor<4x16xf32>) -> tensor<4xf32>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

속성:

Operands:

결과:

mhlo.uniform_dequantize (mhlo::UniformDequantizeOp)

UniformDequantize operation

통사론:

operation ::= `mhlo.uniform_dequantize` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise conversion of quantized tensor operand to a floating-point tensor result according to the quantization parameters defined by the operand type.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#uniform_dequantize

예:

%result = mhlo.uniform_dequantize %operand : (tensor<16x16x!quant.uniform<i8:f32, 34.0:16>>) -> tensor<16x16xf32>

Traits: AlwaysSpeculatableImplTrait , Elementwise , InferTensorType , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.uniform_quantize (mhlo::UniformQuantizeOp)

UniformQuantize operation

통사론:

operation ::= `mhlo.uniform_quantize` $operand attr-dict
              `:` custom<SameOperandsAndResultType>(type($operand), type($result))

Performs element-wise conversion of floating-point tensor or quantized tensor operand to a quantized tensor result according to the quantization parameters defined by the result type.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#uniform_quantize

예:

%result = mhlo.uniform_quantize %operand : (tensor<16x16xf32>) -> tensor<16x16x!quant.uniform<ui8:f32, 34.0:16>>

Traits: AlwaysSpeculatableImplTrait , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

mhlo.while (mhlo::WhileOp)

While operation

Produces the output from executing body function 0 or more times while the cond function outputs true .

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#while

예:

%results0, %results1 = "mhlo.while"(%operand0, %operand1) ({
  ^bb0(%arg0: tensor<i32>, %arg1: tensor<i32>):
    %0 = "mhlo.compare"(%arg0, %arg1) {
      comparison_direction = #stablehlo<comparison_direction LT>
    } : (tensor<i32>, tensor<i32>) -> tensor<i1>
    "mhlo.return"(%0) : (tensor<i1>) -> ()
}, {
  ^bb0(%arg0: tensor<i32>, %arg1: tensor<i32>):
    %0 = "mhlo.add"(%arg0, %constant0) : (tensor<i32>, tensor<i32>) -> tensor<i32>
    "mhlo.return"(%0, %arg1) : (tensor<i32>, tensor<i32>) -> ()
}) : (tensor<i32>, tensor<i32>) -> (tensor<i32>, tensor<i32>)

Traits: RecursiveMemoryEffects , SingleBlockImplicitTerminator<ReturnOp> , SingleBlock

Interfaces: InferTypeOpInterface , OpAsmOpInterface

Operands:

결과:

mhlo.xla.rng_get_and_update_state (mhlo::XlaRngGetAndUpdateStateOp)

XlaRngGetAndUpdateState operation

통사론:

operation ::= `mhlo.xla.rng_get_and_update_state` attr-dict

This operation is private to the XLA compiler, so it is does not yet have a specification.

Informally, this operation represents the change of the global random number generator state for rng instructions. The global state is incremented by delta and the old state is returned.

The output is currently defined for a single output type. If this changes in the future to support multiple types, lowering to use of a global memref must ensure that a single memref is still used and updated appropriately.

Interfaces: InferTypeOpInterface

속성:

결과:

mhlo.xor (mhlo::XorOp)

Xor operation

통사론:

operation ::= `mhlo.xor` $lhs `,` $rhs attr-dict
              `:` custom<SameOperandsAndResultType>(type($lhs), type($rhs), type($result))

Performs element-wise XOR of two tensors lhs and rhs and produces a result tensor.

See: https://github.com/openxla/stablehlo/blob/main/docs/spec.md#xor

예:

%result = mhlo.xor %lhs, %rhs : tensor<2xi32>

Traits: AlwaysSpeculatableImplTrait , Commutative , CompatibleOperandsAndResultType , Elementwise , SameOperandsAndResultShape

Interfaces: ConditionallySpeculatable , InferShapedTypeOpInterface , InferTypeOpInterface , NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

결과:

속성

ArgResultAliasAttr

Attribute that models the alias relationship of entry function argument

This attribute captures the alias relationship of an MHLO main function argument to one of the results, denoted by resultIndex . The argTupleIndices and resultTupleIndices are used to index into nested tuples in operand and result respectively. If isMustAlias is true then the operand-result pair must alias.

This is meant to be used as an attribute on a function argument in MHLO. For example, in the following code it expresses that %arg1 may alias 0-th result.

func @main(%arg0: tensor<2xf32>, %arg1: tensor<3xf32> {mhlo.result_alias =
    mhlo.result_alias<result_index = [2], ...>}
  ) -> tensor<2xf32>, tensor<3xf32> {
  // function body ...
}

매개변수:

ChannelHandleAttr

two 64-bit integers 'handle' and 'type'

통사론:

#mhlo.channel_handle<
  int64_t,   # handle
  int64_t   # type
>

매개변수:

ComparisonDirectionAttr

Which comparison operation to perform.

통사론:

#mhlo.comparison_direction<
  ::mlir::mhlo::ComparisonDirection   # value
>

Enum cases:

• EQ ( EQ )
• NE ( NE )
• GE ( GE )
• GT ( GT )
• LE ( LE )
• LT ( LT ) #### Parameters:

ComparisonTypeAttr

Which comparison type to use.

통사론:

#mhlo.comparison_type<
  ::mlir::mhlo::ComparisonType   # value
>

Enum cases:

• NOTYPE ( NOTYPE )
• FLOAT ( FLOAT )
• TOTALORDER ( TOTALORDER )
• SIGNED ( SIGNED )
• UNSIGNED ( UNSIGNED ) #### Parameters:

ConvDimensionNumbersAttr

Structure of dimension information for conv op

매개변수:

CrossProgramPrefetchAttr

Argument that is prefetched from another program

통사론:

#mhlo.cross_program_prefetch<
  int64_t,   # parameter
  ::llvm::ArrayRef<int64_t>,   # indices
  std::optional<int64_t>   # offset
>

This attribute captures an argument that is prefetched from another program. For a given CrossProgramPrefetchAttr , parameter tells us which argument of the main function of the module is prefetched, and indices is a shape index telling us what subshape of that argument is prefetched.

A shape has a subshape iff it is a tuple. In that case, the subshape of the tuple by indices is the shape achieved after indexing by each element of indices in turn. For example, the [1,0] subshape of tuple<tuple<token, token>, tuple<tensor<i32>, token>> is tensor<i32> .

An empty value for indices means the whole shape is prefetched.

예를 들어,

module attributes { mhlo.cross_program_prefetch = [ #mhlo.cross_program_prefetch< parameter = 0, indices = [0]> ]} {
  func.func @copy(%arg0 : tuple<tensor<2x3xi32>, tensor<i32>>) -> tuple<tensor<2x3xi32>, tensor<i32>> {
    %0 = "mhlo.copy"(%arg0) {is_cross_program_prefetch}
    return %0 : tuple<tensor<2x3xi32>, tensor<i32>>
  }
  func.func @main(%arg0 : tuple<tensor<2x3xi32>, tensor<i32>>) -> tuple<tensor<2x3xi32>, tensor<i32>> {
    %1 = "mhlo.async_start"(%arg0) {called_computation=@copy}
    %2 = "mhlo.async_done"(%1) {called_computation=@copy}
    return %2 : tuple<tensor<2x3xi32>, tensor<i32>>
  }
}

The parameter = 0 tells us that the async copy of the 0 th parameter is a cross_program_prefetch , while the index of [0] tells us that the 0 th element of the tuple is prefetched while the other element of the tuple is not.

매개변수:

CustomCallScheduleAttr

Specifies the desired schedule for the custom-call.

통사론:

#mhlo.custom_call_schedule<
  ::mlir::mhlo::CustomCallSchedule   # value
>

Enum cases:

• NONE ( NONE )
• LATEST ( LATEST )
• EARLIEST ( EARLIEST ) #### Parameters:

DequantizeModeAttr

Dequantization mode. Only MIN_COMBINED is supported.

통사론:

#mhlo.dequantize_mode<
  ::mlir::mhlo::DequantizeMode   # value
>

Enum cases:

• MIN_COMBINED ( MIN_COMBINED ) #### Parameters:

DomainKindAttr

Kind of domain metatdata attached to an HLO domain.

통사론:

#mhlo.kind<
  ::mlir::mhlo::DomainKind   # value
>

Enum cases:

• sharding ( sharding ) #### Parameters:

DotDimensionNumbersAttr

Attribute that models the dimension information for dot.

매개변수:

FftTypeAttr

XLA fast fourier transform type.

통사론:

#mhlo.fft_type<
  ::mlir::mhlo::FftType   # value
>

Enum cases:

• FFT ( FFT )
• IFFT ( IFFT )
• RFFT ( RFFT )
• IRFFT ( IRFFT ) #### Parameters:

FusionKindAttr

fusion kind

통사론:

#mhlo.fusion_kind<
  ::mlir::mhlo::FusionKind   # value
>

Enum cases:

• kLoop ( kLoop )
• kInput ( kInput )
• kOutput ( kOutput )
• kCustom ( kCustom ) #### Parameters:

GatherDimensionNumbersAttr

Attribute that models the dimension information for gather

매개변수:

OutputOperandAliasAttr

Attribute that models the alias relationship of output and operand of a CustomCall op

통사론:

#mhlo.output_operand_alias<
  ::llvm::ArrayRef<int64_t>,   # outputTupleIndices
  int64_t,   # operandIndex
  ::llvm::ArrayRef<int64_t>   # operandTupleIndices
>

This attribute captures the alias relationship of the output to one of the operands for a CustomCall op, denoted by operand_index . The output_tuple_indices and operand_tuple_indices are used to index into output and operand types. These indices lists are empty if the corresponding types are not tuple types, and can be arbitrarily long in case of arbitrarily nested tuple types.

See https://www.tensorflow.org/xla/aliasing

Example when used as array with in mhlo.custom-call:

%0 = "mhlo.custom_call"(%arg0, %arg1) {
  // other attributes
  output_operand_alias = [
    #mhlo.output_operand_alias<output_tuple_indices = [0],
                               operand_index = 0,
                               operand_tuple_indices = [1]>
  ]
} : (tuple<tensor<1x1xf32>, tensor<2x3xf32>>, tensor<5x5xf32>) -> tuple<tensor<2x3xf32>>

The output and the 0th operand are both tuples. The aliasing shows the
relationship between the 0th element in output tuple with the 1st element in
the 0th operand. And both of them are of the same type: tensor<2x3xf32>.

매개변수:

PrecisionAttr

XLA precision for an operand. Has backend specific meaning.

통사론:

#mhlo.precision<
  ::mlir::mhlo::Precision   # value
>

Enum cases:

• DEFAULT ( DEFAULT )
• HIGH ( HIGH )
• HIGHEST ( HIGHEST )
• PACKED_NIBBLE ( PACKED_NIBBLE ) #### Parameters:

RngAlgorithmAttr

XLA PRNG algorithm to be used.

통사론:

#mhlo.rng_algorithm<
  ::mlir::mhlo::RngAlgorithm   # value
>

Enum cases:

• DEFAULT ( DEFAULT )
• THREE_FRY ( THREE_FRY )
• PHILOX ( PHILOX ) #### Parameters:

RngDistributionAttr

XLA PRNG distribution to be used.

통사론:

#mhlo.rng_distribution<
  ::mlir::mhlo::RngDistribution   # value
>

Enum cases:

• UNIFORM ( UNIFORM )
• NORMAL ( NORMAL ) #### Parameters:

ScatterDimensionNumbersAttr

Attribute that models the dimension information for scatter

매개변수:

SparsityDescriptorAttr

Describes structured (N:M) sparsity configuration

통사론:

#mhlo.sparsity<
  int64_t,   # dimension
  int64_t,   # n
  int64_t   # m
>

This attribute is defined for a sparse dot operation with a structured sparse input tensor. With (N=2,M=4), every 4 consecutive logical elements have exactly 2 non-zero physical elements in the input tensor.

$dimension defines the index of the contracting dimension that is sparse (it has to be the most minor dimension). The additional metadata operand in the sparse dot operation defines which logical elements are zeroed out.

매개변수:

TransposeAttr

Transpose options

통사론:

#mhlo.transpose<
  ::mlir::mhlo::Transpose   # value
>

Enum cases:

• TRANSPOSE_INVALID ( TRANSPOSE_INVALID )
• NO_TRANSPOSE ( NO_TRANSPOSE )
• TRANSPOSE ( TRANSPOSE )
• ADJOINT ( ADJOINT ) #### Parameters:

TypeExtensionsAttr

Attribute that extends tensor type with MHLO type properties.

통사론:

#mhlo.type_extensions<
  ::llvm::ArrayRef<int64_t>   # bounds
>

This attribute is used to extend MLIR tensor type with MHLO tensor specific properties. These properties aren't modeled in the MLIR type. This attribute is set in the encoding field of the tensor type.

See HLO_BoundedAttrInterface for documentation for bounds .

매개변수:

유형

AsyncBundleType

Opaque collection of other types

통사론:

!mhlo.async_bundle<
  ::llvm::ArrayRef<Type>   # types
>

매개변수: