skainet-lang-core/sk.ainet.lang.tensor.ops

Package-level declarations

Types

class AddOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

Basic math operations for graph-based execution

abstract class BaseOperation(val name: String, val type: String, val parameters: Map<String, Any> = emptyMap()) : Operation

Abstract base class for operations that provides common functionality

Conv2dOperation

class Conv2dOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

Convolutional operations

DifferentiableTensorOps

interface DifferentiableTensorOps<T : DType, V>

Contract for providing adjoint (backward) rules for differentiable operations. This interface is generated based on @Diff annotations in TensorOps.

DivideOperation

class DivideOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

FlattenOperation

class FlattenOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

GenericOperation

class GenericOperation(val name: String, val parameters: Map<String, Any> = emptyMap(), val type: String = "generic") : BaseOperation

A generic operation that can be used when a specialized operation class is not available.

InputOperation

class InputOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

Input tensor operation for graph representation

KspTensorOps

class KspTensorOps(base: TensorOps, sink: OpSink, session: TraceSession = TraceSession()) : TensorOps

Generated tracing wrapper for TensorOps. This class delegates all method calls to the base implementation while emitting OpTrace events to capture the computation graph.

MatmulOperation

class MatmulOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

Linear algebra operations

MaxPool2dOperation

class MaxPool2dOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

Pooling operations

MixedPrecisionTensorOps

interface MixedPrecisionTensorOps<V> : TensorOps

MultiplyOperation

class MultiplyOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

Operation

interface Operation

Interface for operations that can be represented as graph nodes. Each operation defines how to execute a computation and its metadata.

QuantizedMatmul

object QuantizedMatmul

Optimized matrix multiplication for quantized weight formats (Q8_0, Q4_K).

ReluOperation

class ReluOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

Activation functions

ReshapeOperation

class ReshapeOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

Shape operations

SigmoidOperation

class SigmoidOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

SoftmaxOperation

class SoftmaxOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

SqueezeOperation

class SqueezeOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

Additional shape operations

SubtractOperation

class SubtractOperation<T : DType, V>(parameters: Map<String, Any> = emptyMap()) : BaseOperation

TensorOps

interface TensorOps

TensorSpec

data class TensorSpec(val name: String, val shape: List<Int>?, val dtype: String, val requiresGrad: Boolean = false, val metadata: Map<String, Any> = emptyMap())

Specification for a tensor in the graph, containing shape and type information

TernaryMatmul

object TernaryMatmul

Optimized matrix multiplication for BitNet-style ternary weights.

TransposeOperation