search

intel-analytics / analytics-zoo

Distributed Tensorflow, Keras and PyTorch on Apache Spark/Flink & Ray
https://analytics-zoo.readthedocs.io/
Apache License 2.0
16 stars 3 forks source link

Dllib several failed UTs #77

Closed Le-Zheng closed 3 years ago

Le-Zheng commented 3 years ago

Serval UTs are skipped temporarily on Jenkins nightly build

Check the error and mark it unignored. e.g replcaing ignore should "ImageChannelNormalize with std not 1 work properly" to `"ImageChannelNormalize with std not 1" should "work properly". Then run " mvn test -Dsuites=com.intel.analytics.bigdl.dllib.feature.FeatureSpec" on local to check the error.

  1. com.intel.analytics.bigdl.dllib.feature.FeatureSpec.ImageChannelNormalize with std not 1 should work properly also failed on local 561000 was not equal to 234900

  2. com.intel.analytics.bigdl.dllib.utils.EngineSpec should be right FAILED 6 was not equal to 4 (EngineSpec.scala:139) readConf should skip blockTransferService if bigdl.network.nio is set to false FAILED

  3. com.intel.analytics.bigdl.dllib.utils.EngineSpec.readConf should skip blockTransferService if bigdl.network.nio is set to false 5 was not equal to 3

  4. dllib.nn.keras.nn com.intel.analytics.bigdl.keras.nn.KerasStyleSpec.save and reload model should work correctly requirement failed: unrecognized type Seq[com.intel.analytics.bigdl.dllib.nn.Graph.ModuleNode[T]]

  5. dllib.nn.keras.nn com.intel.analytics.bigdl.keras.nn.KerasStyleSpec.save and reload sequential should work correctly com.intel.analytics.bigdl.dllib.utils.LayerException was thrown.

dding3 commented 3 years ago

The first failed ut can pass on my local env.

The left 4 have been fixed in https://github.com/intel-analytics/analytics-zoo/pull/4858