Question

在使用DL4J的Scala项目中，该项目旨在导入Keras模型，我的SBT包含以下Sonatype快照对DL4J和ND4J的依赖关系。由于DL4J Keras Import: 2.2.x/2.3.x compatibility问题，我使用的是Sonatype快照，而不是Maven。

resolvers += Resolver.sonatypeRepo("snapshots")

libraryDependencies ++= Seq(
  // https://mvnrepository.com/artifact/org.deeplearning4j/deeplearning4j-core
  "org.deeplearning4j" % "deeplearning4j-core" % "1.0.0-SNAPSHOT",
  // https://mvnrepository.com/artifact/org.deeplearning4j/deeplearning4j-nn
  "org.deeplearning4j" % "deeplearning4j-nn" % "1.0.0-SNAPSHOT",
  // https://mvnrepository.com/artifact/org.deeplearning4j/deeplearning4j-modelimport
  "org.deeplearning4j" % "deeplearning4j-modelimport" % "1.0.0-SNAPSHOT",
  // https://mvnrepository.com/artifact/org.nd4j/nd4j-cuda-9.0
  "org.nd4j" % "nd4j-cuda-9.0" % "1.0.0-SNAPSHOT",
  // https://mvnrepository.com/artifact/org.nd4j/nd4j-cuda-9.0-platform
  "org.nd4j" % "nd4j-cuda-9.0-platform" % "1.0.0-SNAPSHOT",
  // https://mvnrepository.com/artifact/org.nd4j/nd4j-api
  "org.nd4j" % "nd4j-api" % "1.0.0-SNAPSHOT"
)

然后转到下面的代码段：

import org.deeplearning4j.nn.multilayer.MultiLayerNetwork
import org.deeplearning4j.nn.modelimport.keras.KerasModelImport

val model = KerasModelImport.importKerasSequentialModelAndWeights(
  """{Valid path to a working Keras model saved as an H5 file}""",
  true
)

我得到

Class org.nd4j.evaluation.IEvaluation not found - continuing with a stub.
val model = KerasModelImport.importKerasSequentialModelAndWeights(

错误。即使我在DL4J GitHub存储库中看到import org.nd4j.evaluation.IEvaluation;使用了many times，我也不知道要导入什么才能使其可用。

在使用Sonatype快照的deeplearning4j Scala项目中找不到ND4J IEvaluation

0 个答案: