add support for model outputs without names

src/BUILD

@@ -1931,6 +1931,8 @@ cc_test(
         "test/dummy_tflite/1/dummy.tflite",
         "test/dummyUppercase/1/dummy.xml",
         "test/dummyUppercase/1/dummy.bin",
+        "test/no_name_output/1/model.xml",
+        "test/no_name_output/1/model.bin",    
         "test/increment_1x3x4x5/1/increment_1x3x4x5.xml",
         "test/increment_1x3x4x5/1/increment_1x3x4x5.bin",
         "test/mediapipe/config_mediapipe_openai_chat_completions_mock.json",

src/modelinstance.cpp

@@ -23,6 +23,7 @@
 #include <set>
 #include <string>
 #include <thread>
+#include <unordered_set>
 #include <utility>
 
 // TODO windows
@@ -305,7 +306,7 @@ static Status applyLayoutConfiguration(const ModelConfig& config, std::shared_pt
     }
 
     OV_LOGGER("ov::Model: {}, model->outputs()", reinterpret_cast<void*>(model.get()));
-    for (const ov::Output<ov::Node>& output : model->outputs()) {
+    for (ov::Output<ov::Node>& output : model->outputs()) {
         try {
             OV_LOGGER("ov::Output<ov::Node> output: {}, output.get_any_name()", reinterpret_cast<const void*>(&output));
             std::string name = output.get_any_name();
@@ -413,6 +414,29 @@ Status ModelInstance::loadTensors(const ModelConfig& config, bool needsToApplyLa
         SPDLOG_LOGGER_ERROR(modelmanager_logger, "Error during configuration validation against model");
         return status;
     }
+    // add output names if not present in the model
+    size_t outputIndex = 0;
+    for (ov::Output<ov::Node>& output : this->model->outputs()) {
+        try {
+            OV_LOGGER("ov::Output<ov::Node> output: {}, output.get_any_name()", reinterpret_cast<const void*>(&output));
+            if (output.get_names().size() == 0) {
+                std::unordered_set<std::string> dummy_name{"out_" + std::to_string(outputIndex)};
+                output.add_names(dummy_name);
+            }
+        } catch (const ov::Exception& e) {
+            SPDLOG_LOGGER_ERROR(modelmanager_logger, "Failed to set the missing name in output for model:{}; version:{}; Error:{}",
+                getName(),
+                getVersion(),
+                e.what());
+            return StatusCode::UNKNOWN_ERROR;
+        } catch (...) {
+            SPDLOG_LOGGER_ERROR(modelmanager_logger, "Failed to set the missing name in output for model:{}; version:{};",
+                getName(),
+                getVersion());
+            return StatusCode::UNKNOWN_ERROR;
+        }
+        outputIndex++;
+    }
     if (needsToApplyLayoutConfiguration) {
         status = applyLayoutConfiguration(config, this->model, getName(), getVersion());
         if (!status.ok()) {

src/test/modelinstance_test.cpp

@@ -112,6 +112,18 @@ TEST_F(TestUnloadModel, CanUnloadModelNotHoldingModelInstanceAtPredictPath) {
     EXPECT_TRUE(modelInstance.canUnloadInstance());
 }
 
+TEST_F(TestUnloadModel, NoNameOutput) {
+    ovms::ModelInstance modelInstance("UNUSED_NAME", UNUSED_MODEL_VERSION, *ieCore);
+    ASSERT_EQ(modelInstance.loadModel(NO_NAME_MODEL_CONFIG), ovms::StatusCode::OK);
+    ASSERT_EQ(ovms::ModelVersionState::AVAILABLE, modelInstance.getStatus().getState());
+    EXPECT_EQ(modelInstance.getInputsInfo().count("input1"), 1);
+    EXPECT_EQ(modelInstance.getInputsInfo().count("input2"), 1);
+    EXPECT_EQ(modelInstance.getOutputsInfo().count("out_0"), 1);
+    EXPECT_EQ(modelInstance.getOutputsInfo().count("out_1"), 1);
+    modelInstance.retireModel();
+    EXPECT_EQ(ovms::ModelVersionState::END, modelInstance.getStatus().getState());
+}
+
 TEST_F(TestUnloadModel, UnloadWaitsUntilMetadataResponseIsBuilt) {
     static std::thread thread;
     static std::shared_ptr<ovms::ModelInstance> instance;

src/test/no_name_output/1/model.xml

@@ -0,0 +1,86 @@
+<?xml version="1.0"?>
+<net name="no_name_output" version="11">
+	<layers>
+		<layer id="1" name="input1" type="Parameter" version="opset1">
+			<data shape="1,10" element_type="i8" />
+			<output>
+				<port id="0" precision="I8" names="input1">
+					<dim>1</dim>
+					<dim>10</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="0" name="input2" type="Parameter" version="opset1">
+			<data shape="1,10" element_type="i8" />
+			<output>
+				<port id="0" precision="I8" names="input2">
+					<dim>1</dim>
+					<dim>10</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="2" name="ADD" type="Add" version="opset1">
+			<data auto_broadcast="numpy" />
+			<input>
+				<port id="0" precision="I8">
+					<dim>1</dim>
+					<dim>10</dim>
+				</port>
+				<port id="1" precision="I8">
+					<dim>1</dim>
+					<dim>10</dim>
+				</port>
+			</input>
+			<output>
+				<port id="2" precision="I8">
+					<dim>1</dim>
+					<dim>10</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="4" name="MULTIPLY" type="Multiply" version="opset1">
+			<data auto_broadcast="numpy" />
+			<input>
+				<port id="0" precision="I8">
+					<dim>1</dim>
+					<dim>10</dim>
+				</port>
+				<port id="1" precision="I8">
+					<dim>1</dim>
+					<dim>10</dim>
+				</port>
+			</input>
+			<output>
+				<port id="2" precision="I8">
+					<dim>1</dim>
+					<dim>10</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="5" name="Result_5" type="Result" version="opset1">
+			<input>
+				<port id="0" precision="I8">
+					<dim>1</dim>
+					<dim>10</dim>
+				</port>
+			</input>
+		</layer>
+		<layer id="3" name="Result_6" type="Result" version="opset1">
+			<input>
+				<port id="0" precision="I8">
+					<dim>1</dim>
+					<dim>10</dim>
+				</port>
+			</input>
+		</layer>
+	</layers>
+	<edges>
+		<edge from-layer="0" from-port="0" to-layer="2" to-port="0" />
+		<edge from-layer="0" from-port="0" to-layer="4" to-port="0" />
+		<edge from-layer="1" from-port="0" to-layer="2" to-port="1" />
+		<edge from-layer="1" from-port="0" to-layer="4" to-port="1" />
+		<edge from-layer="2" from-port="2" to-layer="3" to-port="0" />
+		<edge from-layer="4" from-port="2" to-layer="5" to-port="0" />
+	</edges>
+	<rt_info />
+</net>

src/test/test_utils.hpp

@@ -79,6 +79,7 @@ const std::string passthrough_string_model_location = getGenericFullPathForSrcTe
 const std::string dummy_saved_model_location = getGenericFullPathForSrcTest(std::filesystem::current_path().u8string() + "/src/test/dummy_saved_model", false);
 const std::string dummy_tflite_location = getGenericFullPathForSrcTest(std::filesystem::current_path().u8string() + "/src/test/dummy_tflite", false);
 const std::string scalar_model_location = getGenericFullPathForSrcTest(std::filesystem::current_path().u8string() + "/src/test/scalar", false);
+const std::string no_name_output_model_location = std::filesystem::current_path().u8string() + "/src/test/no_name_output";
 
 const ovms::ModelConfig DUMMY_MODEL_CONFIG{
     "dummy",
@@ -215,6 +216,21 @@ const ovms::ModelConfig SCALAR_MODEL_CONFIG{
     scalar_model_location,  // local path
 };
 
+const ovms::ModelConfig NO_NAME_MODEL_CONFIG{
+    "no_name_output",
+    no_name_output_model_location,  // base path
+    "CPU",                          // target device
+    "1",                            // batchsize
+    1,                              // NIREQ
+    false,                          // is stateful
+    true,                           // idle sequence cleanup enabled
+    false,                          // low latency transformation enabled
+    500,                            // stateful sequence max number
+    "",                             // cache directory
+    1,                              // model_version unused since version are read from path
+    no_name_output_model_location,  // local path
+};
+
 constexpr const char* DUMMY_MODEL_INPUT_NAME = "b";
 constexpr const char* DUMMY_MODEL_OUTPUT_NAME = "a";
 constexpr const int DUMMY_MODEL_INPUT_SIZE = 10;

tests/models/no_name_output.py

@@ -0,0 +1,55 @@
+#
+# Copyright (c) 2025 Intel Corporation
+#
+# 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.
+#
+
+import openvino.runtime as ov
+import numpy as np
+import os
+
+batch_dim = []
+shape = [1, 10]
+dtype = np.int8
+model_name = "no_name_output"
+model_version_dir = model_name
+print(batch_dim + shape)
+in0 = ov.opset1.parameter(shape=batch_dim + shape, dtype=dtype, name="input1")
+in1 = ov.opset1.parameter(shape=batch_dim + shape, dtype=dtype, name="input2")
+op0 = ov.opset1.multiply(in1, in0, name="MULTIPLY")
+op1 = ov.opset1.add(in1, in0, name="ADD")
+
+model = ov.Model([op0, op1], [in0, in1], model_name)
+
+for idx, inp in enumerate(model.inputs):  
+    print(f"Input {idx}: {inp.get_names()} {inp.get_shape()} {inp.get_index()}")
+print(model.outputs)
+for idx, out in enumerate(model.outputs):  
+    print(f"Output {idx}: {out.get_names()} {out.get_shape()} {out.get_index()} ")
+    assert len(out.get_names()) == 0, "number of output names should be 0"
+
+try:
+    os.makedirs(model_version_dir)
+except OSError as ex:
+    pass  # ignore existing dir
+
+ov.serialize(model, model_version_dir + "/model.xml", model_version_dir + "/model.bin")
+
+ov_model = ov.Core().read_model(model_version_dir + "/model.xml")
+compiled_model = ov.Core().compile_model(model, "CPU")
+
+input_data = np.ones((1, 10),dtype=np.int8)*10
+results = compiled_model({"input1": input_data, "input2": input_data})
+assert np.all(results[0] == 100), "for inputs np.ones((1, 10), the expected output is 100 in every element: 10*10"
+assert np.all(results[1] == 20), "for inputs np.ones((1, 10), the expected output is 20 in every element: 10+10"
+