package tensor
- Alphabetic
- Public
- All
Type Members
-
final
case class
TensorProto(dtype: DataType = ..., tensorShape: Option[TensorShapeProto] = _root_.scala.None, versionNumber: Int = 0, tensorContent: ByteString = ..., halfVal: Seq[Int] = _root_.scala.Seq.empty, floatVal: Seq[Float] = _root_.scala.Seq.empty, doubleVal: Seq[Double] = _root_.scala.Seq.empty, intVal: Seq[Int] = _root_.scala.Seq.empty, stringVal: Seq[ByteString] = _root_.scala.Seq.empty, scomplexVal: Seq[Float] = _root_.scala.Seq.empty, int64Val: Seq[Long] = _root_.scala.Seq.empty, boolVal: Seq[Boolean] = _root_.scala.Seq.empty, dcomplexVal: Seq[Double] = _root_.scala.Seq.empty, resourceHandleVal: Seq[ResourceHandleProto] = _root_.scala.Seq.empty, variantVal: Seq[VariantTensorDataProto] = _root_.scala.Seq.empty, uint32Val: Seq[Int] = _root_.scala.Seq.empty, uint64Val: Seq[Long] = _root_.scala.Seq.empty, unknownFields: UnknownFieldSet = ...) extends GeneratedMessage with Updatable[TensorProto] with Product with Serializable
Protocol buffer representing a tensor.
Protocol buffer representing a tensor.
- tensorShape
Shape of the tensor. TODO(touts): sort out the 0-rank issues.
- versionNumber
Version number. In version 0, if the "repeated xxx" representations contain only one element, that element is repeated to fill the shape. This makes it easy to represent a constant Tensor with a single value.
- tensorContent
Serialized raw tensor content from either Tensor::AsProtoTensorContent or memcpy in tensorflow::grpc::EncodeTensorToByteBuffer. This representation can be used for all tensor types. The purpose of this representation is to reduce serialization overhead during RPC call by avoiding serialization of many repeated small items.
- halfVal
DT_HALF, DT_BFLOAT16. Note that since protobuf has no int16 type, we'll have some pointless zero padding for each value here.
- floatVal
DT_FLOAT.
- doubleVal
DT_DOUBLE.
- intVal
DT_INT32, DT_INT16, DT_INT8, DT_UINT8.
- stringVal
DT_STRING
- scomplexVal
DT_COMPLEX64. scomplex_val(2*i) and scomplex_val(2*i+1) are real and imaginary parts of i-th single precision complex.
- int64Val
DT_INT64
- boolVal
DT_BOOL
- dcomplexVal
DT_COMPLEX128. dcomplex_val(2*i) and dcomplex_val(2*i+1) are real and imaginary parts of i-th double precision complex.
- resourceHandleVal
DT_RESOURCE
- variantVal
DT_VARIANT
- uint32Val
DT_UINT32
- uint64Val
DT_UINT64
- Annotations
- @SerialVersionUID()
-
final
case class
VariantTensorDataProto(typeName: String = "", metadata: ByteString = ..., tensors: Seq[TensorProto] = _root_.scala.Seq.empty, unknownFields: UnknownFieldSet = ...) extends GeneratedMessage with Updatable[VariantTensorDataProto] with Product with Serializable
Protocol buffer representing the serialization format of DT_VARIANT tensors.
Protocol buffer representing the serialization format of DT_VARIANT tensors.
- typeName
Name of the type of objects being serialized.
- metadata
Portions of the object that are not Tensors.
- tensors
Tensors contained within objects being serialized.
- Annotations
- @SerialVersionUID()
Value Members
- object TensorProto extends GeneratedMessageCompanion[TensorProto]
- object TensorProtoCompanion extends GeneratedFileObject
- object VariantTensorDataProto extends GeneratedMessageCompanion[VariantTensorDataProto]