Skip to content

Create, save, load and run models using a new functional API #112

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
Sep 17, 2020
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
@@ -0,0 +1,291 @@
/*
* Copyright 2020 The TensorFlow Authors. All rights reserved.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.tensorflow;

import java.io.IOException;
import java.util.List;
import java.util.ListIterator;
import java.util.HashMap;
import java.util.Map;
import java.util.function.Function;
import org.tensorflow.op.Ops;
import org.tensorflow.proto.framework.SignatureDef;
import org.tensorflow.proto.framework.TensorInfo;

/**
* A graph that can be invoked as a single function, with an input and output signature.
*
* <p>A function can also invoke a
* <a href="https://www.tensorflow.org/api_docs/python/tf/function">tf.function</a>
* defined in a {@link SavedModelBundle}.
*
* <pre>{@code
* ConcreteFunction myFunction = savedModelBundle.function("myFunctionSignatureName");
* Map<String, Tensor<?>> outputTensorMap = myFunction.call(inputTensorMap);
* }</pre>
*/
public class ConcreteFunction implements AutoCloseable {

/**
* Creates a function by building a new graph.
*
* <p/>The {@code functionBuilder} must initialize the function graph from the provided
* {@link Ops} instance and return a valid signature that will be used to feed the input tensors
* and fetch the output tensors on execution.
*
* <p/>The function will be the owner of the new graph and its resulting session. Therefore,
* the function must be enclosed properly with a try-with-resources block to guarantee that
* all native resources will be freed once the function is discarded. For example:
*
* <pre>{@code
* public class MyModel {
*
* public static Signature addTwo(Ops tf) {
* Placeholder<TFloat32> input = tf.placeholder(TFloat32.DTYPE);
* Add<TFloat32> output = tf.math.add(input, tf.constant(2.0f));
* return Signature.builder("addTwo").input("x", input).output("y", output).build();
* }
*
* public static void main(String args[]) {
* try (ConcreteFunction function = ConcreteFunction.create(MyModel::addTwo);
* Tensor<TFloat32> x = TFloat32.scalarOf(2.0f)) {
* assertEquals(4.0f, function.call(x).expect(TFloat32.DTYPE).data().getFloat());
* }
* }
* }
* }</pre>
*
* @param functionBuilder function builder
* @return the new function
*/
public static ConcreteFunction create(Function<Ops, Signature> functionBuilder) {
Graph graph = new Graph();
try {
Ops tf = Ops.create(graph);
Signature signature = functionBuilder.apply(tf);
return new ConcreteFunction(signature, graph, new Session(graph), Ownership.GRAPH_AND_SESSION);
} catch (Exception e) {
graph.close();
throw e;
}
}

/**
* Create a function from a signature and an existing graph.
*
* <p/>The function will keep the ownership of the session used to run the graph but not
* the graph itself, meaning that the lifetime of the latter can extend beyond the scope
* of the function. For example:
*
* <pre>{@code
* try (Graph g = new Graph()) {
* Placeholder<TFloat32> input = tf.placeholder(TFloat32.DTYPE);
* Add<TFloat32> output = tf.math.add(input, tf.constant(2.0f));
* Signature signature = Signature.builder().input("x", input).output("y", output).build();
*
* try (ConcreteFunction f = ConcreteFunction.create(signature, g);
* Tensor<TFloat32> x = TFloat32.scalarOf(2.0f)) {
* assertEquals(4.0f, function.call(x).expect(TFloat32.DTYPE).data().getFloat());
* }
* // Graph g is still valid at this point
* }
* }</pre>
*
* @param signature signature of the function to create
* @param graph a valid and initialized graph
* @return a new function
*/
public static ConcreteFunction create(Signature signature, Graph graph) {
return new ConcreteFunction(signature, graph, new Session(graph), Ownership.SESSION_ONLY);
}

/**
* Create a function from a signature and a valid graph session.
*
* <p/>The function will not own the session nor its graph, meaning that their lifetime
* can extend beyond the scope of the function. Therefore the function does not need to be
* closed after its usage. For example:
*
* <pre>{@code
* try (Graph g = new Graph()) {
* Placeholder<TFloat32> input = tf.placeholder(TFloat32.DTYPE);
* Add<TFloat32> output = tf.math.add(input, tf.constant(2.0f));
* Signature signature = Signature.builder().input("x", input).output("y", output).build();
*
* try (Session s = new Session(g)) {
* // Auto-closing the function just as an example but this is not required since it has
* // no effect
* try (ConcreteFunction f = ConcreteFunction.create(signature, s);
* Tensor<TFloat32> t = TFloat32.scalarOf(2.0f)) {
* assertEquals(4.0f, function.call(x).expect(TFloat32.DTYPE).data().getFloat());
* }
* // Session s is still valid at this point
* }
* // Graph g is still valid at this point
* }
* }</pre>
*
* @param signature signature of the function to create
* @param graph a valid session to an initialized graph
* @return a new function
*/
public static ConcreteFunction create(Signature signature, Session session) {
return new ConcreteFunction(signature, session.graph(), session, Ownership.NONE);
}

/**
* Returns the signature of this function
*/
public Signature signature() {
return signature;
}

/**
* Invokes a function.
*
* <p>Caller is responsible for closing all Tensors.
*
* @param tensor input tensor
* @return output tensor
*/
public Map<String, Tensor<?>> call(Map<String, Tensor<?>> arguments)
throws IllegalArgumentException {

final SignatureDef signatureDef = signature.asSignatureDef();
final Session.Runner runner = session.runner();

signatureDef.getInputsMap().forEach((argName, t) -> {
Tensor<?> tensor = arguments.get(argName);
if (tensor == null) {
throw new IllegalArgumentException(String.format("Missing argument [%s]", argName));
}
runner.feed(t.getName(), tensor);
});

Map<String, TensorInfo> outputToNode = signatureDef.getOutputsMap();
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Are we guaranteed that the iteration order of the signature's outputs map is consistent?

Copy link
Contributor

@Shajan Shajan Sep 13, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Good observation, the ordering of signature's output map need not be consistent.

Here is why:

We depend on the ordering of output returned in the call List<Tensor<?>> resultTensors = runner.run(). Which depends on the order we pass in at runner.fetch(t.getName()), that happens to be the iteration order of signatureDef.getOutputsMap().

In runHelper, See creation of outputOpIndices based on order of outputs.add(output), which gets passed into Session.run(..)

Session.run uses outputOpIndices as the order of output, which is implemented in resolveOutputs and TF_Output

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

So it will be consistent within an execution of call but could change across different VM executions if the signatureDef.getOutputsMap() is backed by a regular HashMap rather than a LinkedHashMap?

I ask because we have a tendency to iterate the value set of maps in demo code, and I don't want users to think those will be consistent if they aren't. Usually it's because there is only a single output, but if we're not maintaining ordering then people might be confused if they try to apply what we do in the demos/tests to larger problems.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The order we fetch the tensors does not really matter, as we don't return the outputs as a list but as a map, forcing the user to retrieve them by name instead of by ordinal position, like Session.Run does. And since this is not demo code neither, I think it is safe enough to leave it as is.

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is this really correct? One line does outputToNode.values().forEach(t -> runner.fetch(t.getName())); and the other for (String nodeName: outputToNode.keySet()) { - is there any guarantee of consistent ordering between values() and keySet()!?

The code very much depends on this and there's no test to confirm (ConcreteFunctionTest.java only uses the simpler call()-method).

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Protobuf uses LinkedHashMap internally, so yeah it's fine. Wouldn't be a bad idea to convert it to one manually though.

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe use .values() in both to raise fewer red flags? Though relying on a Map producing the same values() order is perhaps also questionable?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Not if it's a LinkedHashMap, that's the point of the class, it provides stable iteration over a Map with minimal overhead.

outputToNode.values().forEach(t -> runner.fetch(t.getName()));

List<Tensor<?>> resultTensors = runner.run();
try {
ListIterator<Tensor<?>> resultTensorIter = resultTensors.listIterator();
Map<String, Tensor<?>> returnMap = new HashMap<String, Tensor<?>>();

// Use the output names as present in the signature definition
for (String nodeName: outputToNode.keySet()) {
returnMap.put(nodeName, resultTensorIter.next());
}
return returnMap;

} catch (Exception e) {
// Release tensors before throwing exception
for (Tensor<?> t : resultTensors) {
t.close();
}
throw e;
}
}

/**
* Invokes a function with a single input and output.
*
* <p>Caller is responsible for closing all Tensors.
*
* @param tensor input tensor
* @return output tensor
* @throws IllegalArgumentException if there are multiple input or output parameters defined
* in the function
*/
public Tensor<?> call(Tensor<?> tensor) throws IllegalArgumentException {
final SignatureDef signatureDef = signature.asSignatureDef();

if (signatureDef.getInputsCount() != 1) {
throw new IllegalArgumentException(
String.format("Function [%s] requires multiple inputs", signatureDef.getMethodName()));
}
String inputNodeName = signatureDef.getInputsMap().values().iterator().next().getName();

if (signatureDef.getOutputsCount() != 1) {
throw new IllegalArgumentException(
String.format("Function [%s] has multiple outputs", signatureDef.getMethodName()));
}
String outputNodeName = signatureDef.getOutputsMap().values().iterator().next().getName();

return session.runner().feed(inputNodeName, tensor).fetch(outputNodeName).run().get(0);
}

/**
* Export this function as a saved model.
*
* <p>This method is convenient shortcut equivalent to
* {@code SavedModel.exporter(exportDir).withFunction(this).export()}
*
* @throws IOException if saved model or variable state cannot be written on disk
*/
public void save(String exportDir) throws IOException {
SavedModelBundle.exporter(exportDir).withFunction(this).export();
}

/**
* Returns the session used to execute the graph when calling this function
*
* <p>In general, a user does not need to handle directly the session of a function and rely
* on {@link #call(Map)} to execute the graph instead. But in some cases, direct access to
* the session might be necessary, as it allows more running options.
*
* @return the function session
*/
public Session session() {
return session;
}

/**
* Returns the graph of this function
*/
public Graph graph() {
return graph;
}

@Override
public void close() {
if (ownership != Ownership.NONE) {
session.close();
if (ownership == Ownership.GRAPH_AND_SESSION) {
graph.close();
}
}
}

@Override
public String toString() {
return signature.toString();
}

private enum Ownership {
GRAPH_AND_SESSION, SESSION_ONLY, NONE;
}

private final Graph graph;
private final Session session;
private final Signature signature;
private final Ownership ownership;

ConcreteFunction(Signature signature, Graph graph, Session session, Ownership ownership) {
this.graph = graph;
this.session = session;
this.signature = signature;
this.ownership = ownership;
}
}
Loading