SavedModelBundle exporter(...) build: mvn install [ERROR] tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[129,55] cannot find symbol symbol: variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor location: class org.tensorflow.proto.framework.StructProtos

[aday00]

This is follow-up to #97 (comment) and #97 (comment). I'm trying to build a tensorflow-java with SavedModelBundle exporter(...) but encounter a build error.

Specifically, I git checkout the save_model branch of https://github.com/karllessard/tensorflow-java and mvn install, but encounter a build error. I would raise an issue in https://github.com/karllessard/tensorflow-java, but https://github.com/karllessard/tensorflow-java/issues redirects to https://github.com/karllessard/tensorflow-java/pulls, so I cannot raise an issue there.

Would a colleague kindly advise how to fix the following build error:

root@478b80e86d3b:tensorflow-java-karl# mvn install -e
...
[INFO] --- maven-compiler-plugin:3.8.0:compile (default-compile) @ tensorflow-core-api ---
[INFO] Changes detected - recompiling the module!
[INFO] Compiling 1678 source files to /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/target/classes
[INFO] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Tensor.java: Some input files use unchecked or unsafe operations.
[INFO] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Tensor.java: Recompile with -Xlint:unchecked for details.
[INFO] -------------------------------------------------------------
[ERROR] COMPILATION ERROR : 
[INFO] -------------------------------------------------------------
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[129,55] cannot find symbol
  symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
  location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[135,55] cannot find symbol
  symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_fieldAccessorTable
  location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[479,57] cannot find symbol
  symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
  location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[485,57] cannot find symbol
  symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_fieldAccessorTable
  location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[536,57] cannot find symbol
  symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
  location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[82,65] cannot find symbol
  symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
  location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[88,65] cannot find symbol
  symbol:   variable internal_static_tensorflow_SaveableObject_fieldAccessorTable
  location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[290,67] cannot find symbol
  symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
  location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[296,67] cannot find symbol
  symbol:   variable internal_static_tensorflow_SaveableObject_fieldAccessorTable
  location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[329,67] cannot find symbol
  symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
  location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[INFO] 10 errors 
[INFO] -------------------------------------------------------------
[INFO] ------------------------------------------------------------------------
[INFO] Reactor Summary for TensorFlow Java Parent 0.1.0-SNAPSHOT:
[INFO] 
[INFO] TensorFlow Java Parent ............................. SUCCESS [  2.152 s]
[INFO] TensorFlow Tools Library ........................... SUCCESS [02:29 min]
[INFO] TensorFlow Core Parent ............................. SUCCESS [  0.136 s]
[INFO] TensorFlow Core Annotation Processor ............... SUCCESS [  4.371 s]
[INFO] TensorFlow Core API Library ........................ FAILURE [08:09 min]
[INFO] TensorFlow Core API Library Platform ............... SKIPPED
[INFO] TensorFlow Framework Library ....................... SKIPPED
[INFO] ------------------------------------------------------------------------
[INFO] BUILD FAILURE
[INFO] ------------------------------------------------------------------------
[INFO] Total time:  10:47 min
[INFO] Finished at: 2020-08-14T17:41:27Z
[INFO] ------------------------------------------------------------------------
[ERROR] Failed to execute goal org.apache.maven.plugins:maven-compiler-plugin:3.8.0:compile (default-compile) on project tensorflow-core-api: Compilation failure: Compilation failure: 
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[129,55] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[135,55] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_fieldAccessorTable
[ERROR]   location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[479,57] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[485,57] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_fieldAccessorTable
[ERROR]   location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[536,57] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[82,65] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[88,65] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_SaveableObject_fieldAccessorTable
[ERROR]   location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[290,67] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[296,67] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_SaveableObject_fieldAccessorTable
[ERROR]   location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[329,67] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] -> [Help 1]

Rather than debug https://github.com/karllessard/tensorflow-java, I would be happy if there were a save_model branch in https://github.com/tensorflow/java. However, I'm just trying to build something to test model save/export and load -- it does not matter to me how this is achieved.

[karllessard]

Hi @aday00 ,

I'm not sure why you are getting this error, it looks like your generated protos are out-of-sync somehow, maybe a mvn clean before can do it? (be aware that will retrigger a full Bazel build though)

But the save_model branch in my personal repo was just something very experimental showing what could be done and I was not really expecting supporting it for real usage. We just synced up on this topic during the SIG session and now we have a clear plan on how we want to design this before merging it to tensorflow/java. This should happen in the next month or so.

[aday00]

Thank you kindly @karllessard !

It appears the SIG document is https://docs.google.com/document/d/1RqSIKKFLE8kMSjUuZWsVS4JNCRHIrCYiuWcFZRNk2oU/edit# and mentions this:

Saved models and function graphs
  Shajan changes have been pushed to this branch
  Interesting thread about this topic
  We should build an API that is ready to:
    Serve session-centric models (estimators and TF1.x)
    Serve function-centric models (TF2.x)
    Create function graphs, referenced as an attribute of an op (e.g. `tf.dataset.map`)
    Create function graphs with signature (like `tf.function` does)
    Should we save session-centric models from Java? C API only supports this “officially” for now
    Confirmation: lets model the API with functions while still loading/saving session-centric graphs

It appears @Shajan's changes are discussed in #89 and the commit d845856

I don't see a unit test, so I'm not sure how to use this to save a model. Any tips would be greatly appreciated!

In the mean time, I'm happy to continue with the save_model branch and SavedModelBundle exporter(...). Thanks for this pre-alpha feature, and all the time you've invested in the stewardship of this project! Hopefully I can contribute a bit by testing.

[karllessard]

Yes, the portion for saving the model is not in this branch yet but work will start soon and might differ a bit from what you see in my save_model branch right now because we will opt for a more "function-centric" API.

Keeping you posted! and please feel free to share any feedbacks on the draft version in my repo as they can also apply to the next version.

[aday00]

Thanks @karllessard ! Ran mvn clean as suggested:

root@478b80e86d3b:tensorflow-java-karl# mvn clean
[INFO] Scanning for projects...
[INFO] ------------------------------------------------------------------------
[INFO] Reactor Build Order:
[INFO] 
[INFO] TensorFlow Java Parent                                             [pom]
[INFO] TensorFlow Tools Library                                           [jar]
[INFO] TensorFlow Core Parent                                             [pom]
[INFO] TensorFlow Core Annotation Processor                               [jar]
[INFO] TensorFlow Core API Library                                        [jar]
[INFO] TensorFlow Core API Library Platform                               [jar]
[INFO] TensorFlow Framework Library                                       [jar]
[INFO] 
[INFO] -------------------< org.tensorflow:tensorflow-java >-------------------
[INFO] Building TensorFlow Java Parent 0.1.0-SNAPSHOT                     [1/7]
[INFO] --------------------------------[ pom ]---------------------------------
...
[INFO] --- maven-clean-plugin:2.5:clean (default-clean) @ tensorflow-framework ---
[INFO] Deleting /tmp/docker-share/tensorflow-java-karl/tensorflow-framework/target
[INFO] ------------------------------------------------------------------------
[INFO] Reactor Summary for TensorFlow Java Parent 0.1.0-SNAPSHOT:
[INFO] 
[INFO] TensorFlow Java Parent ............................. SUCCESS [  5.399 s]
[INFO] TensorFlow Tools Library ........................... SUCCESS [  0.227 s]
[INFO] TensorFlow Core Parent ............................. SUCCESS [  0.044 s]
[INFO] TensorFlow Core Annotation Processor ............... SUCCESS [  0.337 s]
[INFO] TensorFlow Core API Library ........................ SUCCESS [ 33.345 s]
[INFO] TensorFlow Core API Library Platform ............... SUCCESS [  0.313 s]
[INFO] TensorFlow Framework Library ....................... SUCCESS [  0.717 s]
[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------
[INFO] Total time:  41.954 s
[INFO] Finished at: 2020-08-14T19:25:24Z
[INFO] ------------------------------------------------------------------------

Clean worked, good. Next, mvn install, but same error:

root@478b80e86d3b:tensorflow-java-karl# mvn install -e
[INFO] Error stacktraces are turned on.
[INFO] Scanning for projects...
...
[INFO] --- maven-compiler-plugin:3.8.0:compile (default-compile) @ tensorflow-core-api ---
[INFO] Changes detected - recompiling the module!
[INFO] Compiling 1678 source files to /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/target/classes
[INFO] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Tensor.java: Some input files use unchecked or unsafe operations.
[INFO] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Tensor.java: Recompile with -Xlint:unchecked for details.
[INFO] -------------------------------------------------------------
[ERROR] COMPILATION ERROR : 
[INFO] -------------------------------------------------------------
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[129,55] cannot find symbol
  symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
  location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[135,55] cannot find symbol
  symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_fieldAccessorTable
  location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[479,57] cannot find symbol
  symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
  location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[485,57] cannot find symbol
  symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_fieldAccessorTable
  location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[536,57] cannot find symbol
  symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
  location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[82,65] cannot find symbol
  symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
  location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[88,65] cannot find symbol
  symbol:   variable internal_static_tensorflow_SaveableObject_fieldAccessorTable
  location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[290,67] cannot find symbol
  symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
  location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[296,67] cannot find symbol
  symbol:   variable internal_static_tensorflow_SaveableObject_fieldAccessorTable
  location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[329,67] cannot find symbol
  symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
  location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[INFO] 10 errors 
[INFO] -------------------------------------------------------------
[INFO] ------------------------------------------------------------------------
[INFO] Reactor Summary for TensorFlow Java Parent 0.1.0-SNAPSHOT:
[INFO] 
[INFO] TensorFlow Java Parent ............................. SUCCESS [  1.561 s]
[INFO] TensorFlow Tools Library ........................... SUCCESS [02:59 min]
[INFO] TensorFlow Core Parent ............................. SUCCESS [  0.132 s]
[INFO] TensorFlow Core Annotation Processor ............... SUCCESS [  4.513 s]
[INFO] TensorFlow Core API Library ........................ FAILURE [  15:00 h]
[INFO] TensorFlow Core API Library Platform ............... SKIPPED
[INFO] TensorFlow Framework Library ....................... SKIPPED
[INFO] ------------------------------------------------------------------------
[INFO] BUILD FAILURE
[INFO] ------------------------------------------------------------------------
[INFO] Total time:  15:03 h
[INFO] Finished at: 2020-08-15T10:30:05Z
[INFO] ------------------------------------------------------------------------
[ERROR] Failed to execute goal org.apache.maven.plugins:maven-compiler-plugin:3.8.0:compile (default-compile) on project tensorflow-core-api: Compilation failure: Compilation failure: 
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[129,55] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[135,55] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_fieldAccessorTable
[ERROR]   location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[479,57] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[485,57] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_fieldAccessorTable
[ERROR]   location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/BoundedTensorSpecProto.java:[536,57] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_BoundedTensorSpecProto_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.StructProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[82,65] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[88,65] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_SaveableObject_fieldAccessorTable
[ERROR]   location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[290,67] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[296,67] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_SaveableObject_fieldAccessorTable
[ERROR]   location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] /tmp/docker-share/tensorflow-java-karl/tensorflow-core/tensorflow-core-api/src/gen/java/org/tensorflow/proto/framework/SaveableObject.java:[329,67] cannot find symbol
[ERROR]   symbol:   variable internal_static_tensorflow_SaveableObject_descriptor
[ERROR]   location: class org.tensorflow.proto.framework.SavedObjectGraphProtos
[ERROR] -> [Help 1]

From some SIGJVM email https://groups.google.com/a/tensorflow.org/d/msg/jvm/gGKO-hVS4Pc/LF4rLJOdAQAJ , it seems cherry-picking the save_model patch/changes and applying to the mainline tensorflow-java might work? Thanks also for #101 !

[aday00]

The save_model commit is karllessard@bdb0420

That patch won't apply cleanly to mainline tensorflow-java, so I'll try to make the change by hand.

[aday00]

Applying the saved_model-related changes by hand, to merge into the mainline tensorflow-java, the build seems to work, and I will test this soon:

root@11ade3e6890f:tensorflow-java-savemodel# mvn install -e
...
[INFO] --- maven-install-plugin:2.4:install (default-install) @ tensorflow-framework ---
[INFO] Installing /tmp/docker-share/tensorflow-java-savemodel/tensorflow-framework/target/tensorflow-framework-0.2.0-SNAPSHOT.jar to /root/.m2/repository/org/tensorflow/tensorflow-framework/0.2.0-SNAPSHOT/tensorflow-framework-0.2.0-SNAPSHOT.jar
[INFO] Installing /tmp/docker-share/tensorflow-java-savemodel/tensorflow-framework/pom.xml to /root/.m2/repository/org/tensorflow/tensorflow-framework/0.2.0-SNAPSHOT/tensorflow-framework-0.2.0-SNAPSHOT.pom
[INFO] ------------------------------------------------------------------------
[INFO] Reactor Summary for TensorFlow Java Parent 0.2.0-SNAPSHOT:
[INFO]
[INFO] TensorFlow Java Parent ............................. SUCCESS [  1.450 s]
[INFO] TensorFlow NdArray Library ......................... SUCCESS [02:26 min]
[INFO] TensorFlow Core Parent ............................. SUCCESS [  0.055 s]
[INFO] TensorFlow Core Annotation Processor ............... SUCCESS [  3.259 s]
[INFO] TensorFlow Core API Library ........................ SUCCESS [07:41 min]
[INFO] TensorFlow Core API Library Platform ............... SUCCESS [  0.873 s]
[INFO] TensorFlow Framework Library ....................... SUCCESS [ 38.209 s]
[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------
[INFO] Total time:  10:53 min
[INFO] Finished at: 2020-08-16T23:37:45Z
[INFO] ------------------------------------------------------------------------

My changes are almost identical to @karllessard's code karllessard@bdb0420. I sometimes made changes that did not alter the function of the code, but made the code more similar to Karl's branch, which has the save_model code. If we want a branch and PR from my work, I am happy to clean this up further. Prior email suggested that others (e.g. Alexey Zinoviev) may want to test model-saving code like this too https://groups.google.com/a/tensorflow.org/d/msg/jvm/gGKO-hVS4Pc/D0z5yu-gAQAJ.

My changes are mostly with changes related to NdArray namespacing. The diffs are below, starting with Ops.java:

tensorflow-java-savemodel# diff -u tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java ../tensorflow-java/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java

--- tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java	2020-08-16 19:30:40.402099828 -0400
+++ ../tensorflow-java/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java	2020-08-09 19:39:33.350722050 -0400
@@ -312,10 +312,10 @@
 
   public final ImageOps image;
 
-  public final DataOps data;
-
   public final ShapeOps shape;
 
+  public final DataOps data;
+
   public final IoOps io;
 
   public final DtypesOps dtypes;
@@ -338,10 +338,10 @@
 
   public final SignalOps signal;
 
-  public final QuantizationOps quantization;
-
   public final TrainOps train;
 
+  public final QuantizationOps quantization;
+
   private final Scope scope;
 
   private Ops(Scope scope) {
@@ -349,8 +349,8 @@
     nn = new NnOps(scope);
     summary = new SummaryOps(scope);
     image = new ImageOps(scope);
-    data = new DataOps(scope);
     shape = new ShapeOps(scope);
+    data = new DataOps(scope);
     io = new IoOps(scope);
     dtypes = new DtypesOps(scope);
     xla = new XlaOps(scope);
@@ -362,8 +362,8 @@
     math = new MathOps(scope);
     audio = new AudioOps(scope);
     signal = new SignalOps(scope);
-    quantization = new QuantizationOps(scope);
     train = new TrainOps(scope);
+    quantization = new QuantizationOps(scope);
   }
 
   /**

Graph.java:

tensorflow-java-savemodel# diff -u {../tensorflow-java/,./}tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Graph.java

--- ../tensorflow-java/tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Graph.java	2020-07-29 18:34:42.260476643 -0400
+++ ./tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Graph.java	2020-08-16 19:19:15.860833163 -0400
@@ -44,7 +44,16 @@
 import org.tensorflow.internal.c_api.TF_Status;
 import org.tensorflow.internal.c_api.TF_WhileParams;
 import org.tensorflow.op.Op;
+import org.tensorflow.op.Ops;
+import org.tensorflow.op.core.Constant;
+import org.tensorflow.op.core.NoOp;
+import org.tensorflow.op.core.Placeholder;
+import org.tensorflow.op.train.Restore;
+import org.tensorflow.op.train.Save;
 import org.tensorflow.proto.framework.GraphDef;
+import org.tensorflow.proto.util.SaverDef;
+import org.tensorflow.ndarray.StdArrays;
+import org.tensorflow.types.TString;
 
 
 /**
@@ -67,6 +76,11 @@
     this.nativeHandle = nativeHandle;
   }
 
+  Graph(TF_Graph nativeHandle, SaverDef saverDef) {
+    this(nativeHandle);
+    this.saverDef = saverDef;
+  }
+
   /**
    * Release resources associated with the Graph.
    *
@@ -287,6 +301,17 @@
     return addGradients(null, new Output<?>[] {y}, x, null);
   }
 
+  public SaverDef saverDef() {
+    if (saverDef == null) {
+      synchronized (this) {
+        if (saverDef == null) {
+          saverDef = addVariableSaver(this);
+        }
+      }
+    }
+    return saverDef;
+  }
+
   /**
    * Used to instantiate an abstract class which overrides the buildSubgraph method to build a
    * conditional or body subgraph for a while loop. After Java 8, this can alternatively be used to
@@ -405,6 +430,7 @@
   private final Object nativeHandleLock = new Object();
   private TF_Graph nativeHandle;
   private int refcount = 0;
+  private SaverDef saverDef;
 
   private final List<Op> initializers = new ArrayList<>();
 
@@ -726,6 +752,53 @@
     }
   }
 
+  private static SaverDef addVariableSaver(Graph graph) {
+    Ops tf = Ops.create(graph).withSubScope("save");
+
+    List<String> varNames = new ArrayList<>();
+    List<Operand<?>> varOutputs = new ArrayList<>();
+    List<DataType<?>> varTypes = new ArrayList<>();
+
+    for (Iterator<Operation> iter = graph.operations(); iter.hasNext();) {
+      Operation op = iter.next();
+      if (op.type().equals("VariableV2")) {
+        varNames.add(op.name());
+        varOutputs.add(op.output(0));
+        varTypes.add(op.output(0).dataType());
+      }
+    }
+
+    // FIXME Need an easier way to initialize an NdArray from a list
+    String[] tmp = new String[varNames.size()];
+    Constant<TString> varNamesTensor = tf.constant(StdArrays.ndCopyOf(varNames.toArray(tmp)));
+    Operand<TString> varSlices = tf.zerosLike(varNamesTensor);
+
+    Placeholder<TString> saveFilename = tf.placeholder(TString.DTYPE);
+    Save saveVariables = tf.train.save(
+        saveFilename,
+        varNamesTensor,
+        varSlices,
+        varOutputs
+    );
+    Restore restoreVariables = tf.train.restore(
+        saveFilename,
+        varNamesTensor,
+        varSlices,
+        varTypes
+    );
+    List<Op> restoreOps = new ArrayList<>(varOutputs.size());
+    for (int i = 0; i < varOutputs.size(); ++i) {
+      restoreOps.add(tf.assign(varOutputs.get(i), (Operand) restoreVariables.tensors().get(i)));
+    }
+    NoOp restoreAll = tf.withControlDependencies(restoreOps).noOp();
+
+    return SaverDef.newBuilder()
+        .setFilenameTensorName(saveFilename.op().name())
+        .setSaveTensorName(saveVariables.op().name())
+        .setRestoreOpName(restoreAll.op().name())
+        .build();
+  }
+
   static {
     TensorFlow.init();
   }

SavedModelBundle:

tensorflow-java-savemodel# diff -u {../tensorflow-java/,./}tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/SavedModelBundle.java

--- ../tensorflow-java/tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/SavedModelBundle.java	2020-07-29 18:34:42.260476643 -0400
+++ ./tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/SavedModelBundle.java	2020-08-16 19:18:29.544101268 -0400
@@ -20,6 +20,15 @@
 import static org.tensorflow.internal.c_api.global.tensorflow.TF_SetConfig;
 
 import com.google.protobuf.InvalidProtocolBufferException;
+import java.io.FileOutputStream;
+import java.io.IOException;
+import java.io.OutputStream;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.List;
+import java.util.Map;
 import org.bytedeco.javacpp.BytePointer;
 import org.bytedeco.javacpp.PointerPointer;
 import org.bytedeco.javacpp.PointerScope;
@@ -30,8 +39,16 @@
 import org.tensorflow.internal.c_api.TF_SessionOptions;
 import org.tensorflow.internal.c_api.TF_Status;
 import org.tensorflow.proto.framework.ConfigProto;
+import org.tensorflow.proto.framework.DataType;
 import org.tensorflow.proto.framework.MetaGraphDef;
+import org.tensorflow.proto.framework.MetaGraphDef.MetaInfoDef;
 import org.tensorflow.proto.framework.RunOptions;
+import org.tensorflow.proto.framework.SavedModel;
+import org.tensorflow.proto.framework.SignatureDef;
+import org.tensorflow.proto.framework.TensorInfo;
+import org.tensorflow.proto.framework.TensorShapeProto;
+import org.tensorflow.proto.framework.TensorShapeProto.Dim;
+import org.tensorflow.ndarray.Shape;
 
 /**
  * SavedModelBundle represents a model loaded from storage.
@@ -94,6 +111,78 @@
     private RunOptions runOptions = null;
   }
 
+  public static final class Exporter {
+
+    public Exporter withTags(String... tags) {
+      this.tags.addAll(Arrays.asList(tags));
+      return this;
+    }
+
+    public Exporter withSignature(Map<String, Operand<?>> inputs,  Map<String, Operand<?>> outputs) {
+      return withSignature("serving_default", "tensorflow/serving/predict", inputs, outputs);
+    }
+
+    public Exporter withSignature(String signatureName, String methodName, Map<String, Operand<?>> inputs,  Map<String, Operand<?>> outputs) {
+      SignatureDef.Builder signatureDefBuilder = SignatureDef.newBuilder();
+      for (Map.Entry<String, Operand<?>> inputEntry : inputs.entrySet()) {
+        signatureDefBuilder.putInputs(inputEntry.getKey(), toTensorInfo(inputEntry.getValue().asOutput()));
+      }
+      for (Map.Entry<String, Operand<?>> outputEntry : outputs.entrySet()) {
+        signatureDefBuilder.putOutputs(outputEntry.getKey(), toTensorInfo(outputEntry.getValue().asOutput()));
+      }
+      signatureDefBuilder.setMethodName(methodName);
+      metaGraphDefBuilder.putSignatureDef(signatureName, signatureDefBuilder.build());
+      return this;
+    }
+
+    public void export(Session session) throws IOException {
+      Graph graph = session.graph();
+      if (tags.isEmpty()) {
+        tags.add("serve");
+      }
+      // Important: it is imperative to retrieve the graphDef after the saverDef, as the former might add new ops. FIXME Better way for handling this?
+      MetaGraphDef metaGraphDef = metaGraphDefBuilder
+          .setSaverDef(graph.saverDef())
+          .setGraphDef(graph.toGraphDef())
+          .setMetaInfoDef(MetaInfoDef.newBuilder().addAllTags(tags))
+          .build();
+
+      // Make sure saved model directories exist
+      Path variableDir = Paths.get(exportDir, "variables");
+      variableDir.toFile().mkdirs();
+
+      // Save variable state, this must be done before we retrieve the `SaverDef` from the graph
+      session.save(variableDir.resolve("variables").toString());
+
+      // Save graph
+      SavedModel savedModelDef = SavedModel.newBuilder().addMetaGraphs(metaGraphDef).build();
+      try (OutputStream file = new FileOutputStream(Paths.get(exportDir, "saved_model.pb").toString())) {
+        savedModelDef.writeTo(file);
+      }
+    }
+
+    Exporter(String exportDir) {
+      this.exportDir = exportDir;
+    }
+
+    private final String exportDir;
+    private final MetaGraphDef.Builder metaGraphDefBuilder = MetaGraphDef.newBuilder();
+    private final List<String> tags = new ArrayList<>();
+
+    private static TensorInfo toTensorInfo(Output<?> operand) {
+      Shape shape = operand.shape();
+      TensorShapeProto.Builder tensorShapeBuilder = TensorShapeProto.newBuilder();
+      for (int i = 0; i < shape.numDimensions(); ++i) {
+        tensorShapeBuilder.addDim(Dim.newBuilder().setSize(shape.size(i)));
+      }
+      return TensorInfo.newBuilder()
+          .setDtype(DataType.forNumber(operand.dataType().nativeCode()))
+          .setTensorShape(tensorShapeBuilder)
+          .setName(operand.op().name() + ":" + operand.index())
+          .build();
+    }
+  }
+
   /**
    * Load a saved model from an export directory. The model that is being loaded should be created
    * using the <a href="https://www.tensorflow.org/api_docs/python/tf/saved_model">Saved Model
@@ -125,6 +214,10 @@
     return new Loader(exportDir);
   }
 
+  public static Exporter exporter(String exportDir) {
+    return new Exporter(exportDir);
+  }
+
   /**
    * Returns the <a
    * href="https://www.tensorflow.org/code/tensorflow/core/protobuf/meta_graph.proto">MetaGraphDef
@@ -176,7 +269,7 @@
    */
   private static SavedModelBundle fromHandle(
       TF_Graph graphHandle, TF_Session sessionHandle, MetaGraphDef metaGraphDef) {
-    Graph graph = new Graph(graphHandle);
+    Graph graph = new Graph(graphHandle, metaGraphDef.getSaverDef());
     Session session = new Session(graph, sessionHandle);
     return new SavedModelBundle(graph, session, metaGraphDef);
   }

Session:

tensorflow-java-savemodel# diff -u {../tensorflow-java/,./}tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Session.java

--- ../tensorflow-java/tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Session.java	2020-07-29 18:34:42.260476643 -0400
+++ ./tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Session.java	2020-08-15 16:30:35.751805310 -0400
@@ -36,6 +36,8 @@
 
 import java.util.ArrayList;
 import java.util.List;
+import org.tensorflow.proto.util.SaverDef;
+import org.tensorflow.types.TString;
 
 import static org.tensorflow.Graph.resolveOutputs;
 import static org.tensorflow.internal.c_api.global.tensorflow.*;
@@ -444,6 +446,14 @@
     runner().addTarget(op.op()).run();
   }
 
+  public void save(String prefix) {
+    SaverDef saverDef = graph.saverDef();
+    runner()
+        .addTarget(saverDef.getSaveTensorName())
+        .feed(saverDef.getFilenameTensorName(), TString.scalarOf(prefix))
+        .run();
+  }
+
   /**
    * Output tensors and metadata obtained when executing a session.
    *
@@ -463,6 +473,10 @@
     public RunMetadata metadata;
   }
 
+  Graph graph() {
+    return graph;
+  }
+
   private final Graph graph;
   private final Graph.Reference graphRef;

SavedModelBundleTest exercises the exporter(...) method to save a model:

tensorflow-java-savemodel# diff -u {../tensorflow-java/,./}tensorflow-core/tensorflow-core-api/src/test/java/org/tensorflow/SavedModelBundleTest.java

--- ../tensorflow-java/tensorflow-core/tensorflow-core-api/src/test/java/org/tensorflow/SavedModelBundleTest.java	2020-07-29 18:34:42.276476950 -0400
+++ ./tensorflow-core/tensorflow-core-api/src/test/java/org/tensorflow/SavedModelBundleTest.java	2020-08-16 19:26:24.250842779 -0400
@@ -15,21 +15,39 @@
 
 package org.tensorflow;
 
+import static org.junit.jupiter.api.Assertions.assertEquals;
 import static org.junit.jupiter.api.Assertions.assertNotNull;
 import static org.junit.jupiter.api.Assertions.assertTrue;
 import static org.junit.jupiter.api.Assertions.fail;
 
+import java.io.IOException;
 import java.net.URISyntaxException;
+import java.nio.file.Files;
+import java.nio.file.Path;
 import java.nio.file.Paths;
+import java.util.Collections;
 import org.junit.jupiter.api.Test;
 import org.tensorflow.exceptions.TensorFlowException;
+import org.tensorflow.op.Ops;
+import org.tensorflow.op.core.Init;
+import org.tensorflow.op.core.Placeholder;
+import org.tensorflow.op.core.ReduceSum;
+import org.tensorflow.op.core.Variable;
 import org.tensorflow.proto.framework.ConfigProto;
 import org.tensorflow.proto.framework.RunOptions;
+import org.tensorflow.proto.framework.SignatureDef;
+import org.tensorflow.proto.framework.TensorInfo;
+import org.tensorflow.ndarray.Shape;
+import org.tensorflow.ndarray.FloatNdArray;
+import org.tensorflow.ndarray.StdArrays;
+import org.tensorflow.types.TFloat32;
 
 /** Unit tests for {@link org.tensorflow.SavedModelBundle}. */
 public class SavedModelBundleTest {
 
+  private static final float EPSILON = 1e-7f;
   private static final String SAVED_MODEL_PATH;
+
   static {
     try {
       SAVED_MODEL_PATH = Paths.get(SavedModelBundleTest.class.getResource("/saved_model").toURI()).toString();
@@ -72,6 +90,73 @@
     }
   }
 
+  @Test
+  public void save() throws IOException {
+    Path testFolder = Files.createTempDirectory("tf-saved-model-export-test");
+    float reducedSum;
+    FloatNdArray xValue = StdArrays.ndCopyOf(new float[][] { { 0, 1, 2 }, { 3, 4, 5 } });
+    Shape xyShape = Shape.of(2, 3L);
+    try (Graph g = new Graph()) {
+      Ops tf = Ops.create(g);
+      Placeholder<TFloat32> x = tf.placeholder(TFloat32.DTYPE, Placeholder.shape(xyShape));
+      Variable<TFloat32> y = tf.variable(tf.random.randomUniform(tf.constant(xyShape), TFloat32.DTYPE));
+      ReduceSum<TFloat32> z = tf.reduceSum(tf.math.add(x, y), tf.array(0, 1));
+      Init init = tf.init();
+
+      try (Session s = new Session(g)) {
+        s.run(init);
+        try (Tensor<TFloat32> xTensor = TFloat32.tensorOf(xValue);
+            Tensor<TFloat32> zTensor = s.runner()
+                .feed(x, xTensor)
+                .fetch(z)
+                .run()
+                .get(0).expect(TFloat32.DTYPE)) {
+          reducedSum = zTensor.data().getFloat();
+        }
+        SavedModelBundle.exporter(testFolder.toString())
+            .withTags("test")
+            .withSignature(Collections.singletonMap("input", x), Collections.singletonMap("reducedSum", z))
+            .export(s);
+      }
+    }
+    assertTrue(Files.exists(testFolder.resolve(Paths.get("variables", "variables.index"))));
+    assertTrue(Files.exists(testFolder.resolve(Paths.get("variables", "variables.data-00000-of-00001"))));
+    assertTrue(Files.exists(testFolder.resolve("saved_model.pb")));
+
+    // Reload the model just saved and validate its data
+    try (SavedModelBundle savedModel = SavedModelBundle.load(testFolder.toString(), "test")) {
+      assertNotNull(savedModel.metaGraphDef());
+      assertNotNull(savedModel.metaGraphDef().getSaverDef());
+      assertEquals(1, savedModel.metaGraphDef().getSignatureDefCount());
+
+      SignatureDef signature = savedModel.metaGraphDef().getSignatureDefMap().get("serving_default");
+      assertNotNull(signature);
+      assertEquals(1, signature.getInputsCount());
+      assertEquals(1, signature.getOutputsCount());
+
+      TensorInfo inputInfo = signature.getInputsMap().get("input");
+      assertNotNull(inputInfo);
+      assertEquals(xyShape.numDimensions(), inputInfo.getTensorShape().getDimCount());
+      for (int i = 0; i < xyShape.numDimensions(); ++i) {
+        assertEquals(xyShape.size(i), inputInfo.getTensorShape().getDim(i).getSize());
+      }
+
+      TensorInfo outputInfo = signature.getOutputsMap().get("reducedSum");
+      assertNotNull(outputInfo);
+      assertEquals(0, outputInfo.getTensorShape().getDimCount());
+
+      // Run the saved model just loaded and make sure it returns the same result as before
+      try (Tensor<TFloat32> xTensor = TFloat32.tensorOf(xValue);
+          Tensor<TFloat32> zTensor = savedModel.session().runner()
+              .feed(inputInfo.getName(), xTensor)
+              .fetch(outputInfo.getName())
+              .run()
+              .get(0).expect(TFloat32.DTYPE)) {
+        assertEquals(reducedSum, zTensor.data().getFloat(), EPSILON);
+      }
+    }
+  }
+
   private static RunOptions sillyRunOptions() {
     return RunOptions.newBuilder()
         .setTraceLevel(RunOptions.TraceLevel.FULL_TRACE)

If anything looks incorrect, would a colleague kindly let me know? I plan to test soon. Thanks again!

[aday00]

Preliminary tests of model save/load pass. Test output shows orig_reduced_sum=17.154881 loaded_reduced_sum=17.154881, so the loaded model gives the same result as the original saved model. Good:

root@11ade3e6890f:tensorflow-java-savemodel# java -jar my.jar tf --verbose
DEBUG 1597658225600: Started ndarray TensorflowJavaTest.
DEBUG 1597658226181: matrix3d rank 3
DEBUG 1597658226186: Finished ndarray TensorflowJavaTest.
DEBUG 1597658226187: Started graph TensorflowJavaTest.
DEBUG 1597658230142: fetch_test fetched.data.getInt(0)=3 is 3? true
DEBUG 1597658230143: fetch_test fetched.data.getInt(1)=4 is 4? true
DEBUG 1597658230357: feed_test fetched.data=8,6,4,2
DEBUG 1597658230365: Started save_test saving to dir=tf-saved-model-export-test
DEBUG 1597658231576: Finished save_test
DEBUG 1597658231584: Started load_test
2020-08-17 09:57:11.595866: I external/org_tensorflow/tensorflow/cc/saved_model/reader.cc:31] Reading SavedModel from: /tmp/tf-saved-model-export-test8180187955358911902
2020-08-17 09:57:11.596308: I external/org_tensorflow/tensorflow/cc/saved_model/reader.cc:54] Reading meta graph with tags { test }
2020-08-17 09:57:11.596360: I external/org_tensorflow/tensorflow/cc/saved_model/loader.cc:295] Reading SavedModel debug info (if present) from: /tmp/tf-saved-model-export-test8180187955358911902
2020-08-17 09:57:11.597165: I external/org_tensorflow/tensorflow/cc/saved_model/loader.cc:234] Restoring SavedModel bundle.
2020-08-17 09:57:11.614106: I external/org_tensorflow/tensorflow/cc/saved_model/loader.cc:364] SavedModel load for tags { test }; Status: success: OK. Took 22272 microseconds.
DEBUG 1597658231632: load_test  xy_shape dimension_count=2  input_info dimension_count=2
DEBUG 1597658231640: load_test  xy_shape dimension_sizes=(2,3)  input_info dimension_sizes=(2,3)
DEBUG 1597658231643: load_test output_info dimension=0
DEBUG 1597658231659: load_test orig_reduced_sum=17.154881 loaded_reduced_sum=17.154881
DEBUG 1597658231660: Finished load_test
DEBUG 1597658231661: Finished graph TensorflowJavaTest.

The main run() method for testing is now:

    def run(): Int = {
      var ret = 0
      debug("Started ndarray TensorflowJavaTest.")
      // run simple data buffers test as in https://github.com/tensorflow/java/tree/master/ndarray
      val matrix3d = org.tensorflow.ndarray.NdArrays.ofInts( org.tensorflow.ndarray.Shape.of(2, 3, 2) )
      debug("matrix3d rank " + matrix3d.rank)
      debug("Finished ndarray TensorflowJavaTest.")

      debug("Started graph TensorflowJavaTest.")
      fetch_and_feed_tests()
      save_and_load_tests()


      debug("Finished graph TensorflowJavaTest.")
      return ret
    }

save_and_load_tests is:

    private def save_and_load_tests(): Unit = {
      val graph = new org.tensorflow.Graph() // defined in https://github.com/tensorflow/java/blob/master/tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/Graph.java
      val session = new org.tensorflow.Session(graph)
      val tf = org.tensorflow.op.Ops.create(graph)

      val (test_folder, x_value, xy_shape, z_tensor, reduced_sum) = save_test(graph, session, tf)
      val loaded_model = load_test(session, test_folder, x_value, xy_shape, z_tensor, reduced_sum)
    }

Model save is tested in save_test, which relies on @karllessard's SavedModelBundle exporter(...) method, thank you!

    private def save_test(graph: org.tensorflow.Graph, session: org.tensorflow.Session, tf: org.tensorflow.op.Ops,
                          temp_dirname: String = "tf-saved-model-export-test"):
                          (Path, org.tensorflow.ndarray.FloatNdArray, org.tensorflow.ndarray.Shape,
                           org.tensorflow.Tensor[org.tensorflow.types.TFloat32], Float) = {
      // adapted from save() in https://github.com/karllessard/tensorflow-java/blob/bdb0420fd72264457a6d47e5e6c6f7a2e56b4271/tensorflow-core/tensorflow-core-api/src/test/java/org/tensorflow/SavedModelBundleTest.java#L116
      // use exporter() to save model.
      debug("Started save_test saving to dir=%s".format(temp_dirname))
      val test_folder: Path = Files.createTempDirectory(temp_dirname)

      //val x_value = org.tensorflow.tools.ndarray.StdArrays.ndCopyOf(new float[][] { { 0, 1, 2 }, { 3, 4, 5 } }) // FloatNdArray
      val x_value: org.tensorflow.ndarray.FloatNdArray = org.tensorflow.ndarray.StdArrays.ndCopyOf(Array( Array(0.0f,1.0f,2.0f),Array(3.0f,4.0f,5.0f) ))
      val xy_shape = org.tensorflow.ndarray.Shape.of(2, 3l)

      val x = tf.placeholder(org.tensorflow.types.TFloat32.DTYPE, org.tensorflow.op.core.Placeholder.shape(xy_shape))
      val y = tf.variable(tf.random.randomUniform(tf.constant(xy_shape), org.tensorflow.types.TFloat32.DTYPE))
      val z = tf.reduceSum(tf.math.add(x, y), tf.array(0, 1))
      val init = tf.init

      session.run(init)
      val x_tensor = org.tensorflow.types.TFloat32.tensorOf(x_value)
      val z_tensor = session.runner
                            .feed(x, x_tensor)
                            .fetch(z)
                            .run
                            .get(0).expect(org.tensorflow.types.TFloat32.DTYPE)
      val reduced_sum: Float = z_tensor.data.getFloat()
      org.tensorflow.SavedModelBundle.exporter(test_folder.toString)
                                     .withTags("test")
                                     .withSignature(Collections.singletonMap("input", x), Collections.singletonMap("reducedSum", z))
                                     .export(session)
      debug("Finished save_test")
      (test_folder, x_value, xy_shape, z_tensor, reduced_sum)
    }

Model load is tested in load_test, which is heavily based on Karl's helpful unit tests:

    private def load_test(session: org.tensorflow.Session,
                          test_folder: Path, x_value: org.tensorflow.ndarray.FloatNdArray,
                          xy_shape: org.tensorflow.ndarray.Shape,
                          orig_z_tensor: org.tensorflow.Tensor[org.tensorflow.types.TFloat32],
                          orig_reduced_sum: Float):
                          org.tensorflow.SavedModelBundle = {
      debug("Started load_test")
      // like save_test, adapted from https://github.com/karllessard/tensorflow-java/blob/bdb0420fd72264457a6d47e5e6c6f7a2e56b4271/tensorflow-core/tensorflow-core-api/src/test/java/org/tensorflow/SavedModelBundleTest.java#L116
      val saved_model: org.tensorflow.SavedModelBundle = org.tensorflow.SavedModelBundle.load(test_folder.toString, "test")
      val signature = saved_model.metaGraphDef.getSignatureDefMap.get("serving_default") // SignatureDef
      val input_info = signature.getInputsMap.get("input") // TensorInfo

      debug("load_test  xy_shape dimension_count=%d  input_info dimension_count=%d".format(xy_shape.numDimensions, input_info.getTensorShape.getDimCount)) // dimension counts should match

      debug("load_test  xy_shape dimension_sizes=(%s)  input_info dimension_sizes=(%s)".format(
            (0 until xy_shape.numDimensions).map( xy_shape.size(_) ).mkString(","),
            (0 until input_info.getTensorShape.getDimCount).map( input_info.getTensorShape.getDim(_).getSize ).mkString(","))) // dimension sizes should match

      val output_info = signature.getOutputsMap.get("reducedSum") // TensorInfo
      debug("load_test output_info dimension=%d".format(output_info.getTensorShape.getDimCount))

      val x_tensor = org.tensorflow.types.TFloat32.tensorOf(x_value)
      val z_tensor = saved_model.session.runner
                                        .feed(input_info.getName, x_tensor)
                                        .fetch(output_info.getName)
                                        .run
                                        .get(0).expect(org.tensorflow.types.TFloat32.DTYPE)

      debug("load_test orig_reduced_sum=%f loaded_reduced_sum=%f".format(orig_reduced_sum, z_tensor.data.getFloat()))

      debug("Finished load_test")
      saved_model
    }

The result is a saved model in /tmp. Would any colleagues kindly give their opinion -- does this look reasonable?

root@11ade3e6890f:tensorflow-java-savemodel# ls -hal /tmp/tf-saved-model-export-test8180187955358911902
total 16K
drwx------ 3 root root 4.0K Aug 17 09:57 .
drwxrwxrwt 1 root root 4.0K Aug 17 09:57 ..
-rw-r--r-- 1 root root 1.3K Aug 17 09:57 saved_model.pb
drwxr-xr-x 2 root root 4.0K Aug 17 09:57 variables
root@11ade3e6890f:tensorflow-java-savemodel# ls -hal /tmp/tf-saved-model-export-test8180187955358911902/variables    
total 16K
drwxr-xr-x 2 root root 4.0K Aug 17 09:57 .
drwx------ 3 root root 4.0K Aug 17 09:57 ..
-rw-r--r-- 1 root root   24 Aug 17 09:57 variables.data-00000-of-00001
-rw-r--r-- 1 root root  132 Aug 17 09:57 variables.index