feat: add onnx local reranking model

README.md

@@ -408,6 +408,7 @@ dat-parent/
 │   ├── dat-reranker-ms-marco-minilm-l6-v2-q/    # 【本地内置Reranking模型】
 │   ├── dat-reranker-ms-marco-tinybert-l2-v2/    # 【本地内置Reranking模型】
 │   ├── dat-reranker-ms-marco-tinybert-l2-v2-q/  # 【本地内置Reranking模型】
+│   ├── dat-reranker-onnx-local/                 # 【本地调用Reranking模型】
 │   └── dat-reranker-jina/
 ├── 💾 dat-storers/       # 向量存储后端
 │   ├── dat-storer-duckdb/    # 【本地内置向量存储】

dat-cli/pom.xml

@@ -713,6 +713,17 @@
                         </exclusion>
                     </exclusions>
                 </dependency>
+                <dependency>
+                    <groupId>ai.dat</groupId>
+                    <artifactId>dat-reranker-onnx-local</artifactId>
+                    <version>${project.version}</version>
+                    <exclusions>
+                        <exclusion>
+                            <groupId>com.microsoft.onnxruntime</groupId>
+                            <artifactId>onnxruntime</artifactId>
+                        </exclusion>
+                    </exclusions>
+                </dependency>
             </dependencies>
         </profile>
 

dat-rerankers/dat-reranker-onnx-local/pom.xml

@@ -0,0 +1,27 @@
+<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
+    <modelVersion>4.0.0</modelVersion>
+
+    <parent>
+        <groupId>ai.dat</groupId>
+        <artifactId>dat-rerankers</artifactId>
+        <version>${revision}</version>
+    </parent>
+
+    <artifactId>dat-reranker-onnx-local</artifactId>
+    <name>DAT : Rerankers : In-process(ONNX) local</name>
+
+    <dependencies>
+        <dependency>
+            <groupId>ai.dat</groupId>
+            <artifactId>dat-core</artifactId>
+            <version>${project.version}</version>
+        </dependency>
+        <dependency>
+            <groupId>dev.langchain4j</groupId>
+            <artifactId>langchain4j-onnx-scoring</artifactId>
+            <version>${langchain4j-beta.version}</version>
+        </dependency>
+    </dependencies>
+
+</project>

dat-rerankers/dat-reranker-onnx-local/src/main/java/ai/dat/reranker/onnx/OnnxScoringModelFactory.java

@@ -0,0 +1,67 @@
+package ai.dat.reranker.onnx;
+
+import ai.dat.core.configuration.ConfigOption;
+import ai.dat.core.configuration.ConfigOptions;
+import ai.dat.core.configuration.ReadableConfig;
+import ai.dat.core.factories.ScoringModelFactory;
+import ai.dat.core.utils.FactoryUtil;
+import dev.langchain4j.model.scoring.ScoringModel;
+import dev.langchain4j.model.scoring.onnx.OnnxScoringModel;
+
+import java.util.Collections;
+import java.util.LinkedHashSet;
+import java.util.List;
+import java.util.Set;
+
+/**
+ * @Author JunjieM
+ * @Date 2025/10/21
+ */
+public class OnnxScoringModelFactory implements ScoringModelFactory {
+
+    public static final String IDENTIFIER = "onnx";
+
+    public static final ConfigOption<String> MODEL_FILE_PATH =
+            ConfigOptions.key("model-file-path")
+                    .stringType()
+                    .noDefaultValue()
+                    .withDescription("""
+                            ONNX model file path, for example: /home/dat/model.onnx
+
+                            Local scoring (reranking) models, powered by [ONNX runtime](https://onnxruntime.ai/docs/get-started/with-java.html), running in the same Java process.
+                            Many models (e.g., from [Hugging Face](https://huggingface.co)) can be used, as long as they are in the ONNX format.
+                            Information on how to convert models into ONNX format can be found [here](https://huggingface.co/docs/optimum-onnx/onnx/usage_guides/export_a_model).
+                            Many models already converted to ONNX format are available [here](https://huggingface.co/Xenova).
+                            """);
+
+    public static final ConfigOption<String> TOKENIZER_FILE_PATH =
+            ConfigOptions.key("tokenizer-file-path")
+                    .stringType()
+                    .noDefaultValue()
+                    .withDescription("Tokenizer file path, for example: /home/dat/tokenizer.json");
+
+    @Override
+    public String factoryIdentifier() {
+        return IDENTIFIER;
+    }
+
+    @Override
+    public Set<ConfigOption<?>> requiredOptions() {
+        return new LinkedHashSet<>(List.of(MODEL_FILE_PATH, TOKENIZER_FILE_PATH));
+    }
+
+    @Override
+    public Set<ConfigOption<?>> optionalOptions() {
+        return Collections.emptySet();
+    }
+
+    @Override
+    public ScoringModel create(ReadableConfig config) {
+        FactoryUtil.validateFactoryOptions(this, config);
+
+        String modelFilePath = config.get(MODEL_FILE_PATH);
+        String tokenizerFilePath = config.get(TOKENIZER_FILE_PATH);
+
+        return new OnnxScoringModel(modelFilePath, tokenizerFilePath);
+    }
+}

dat-rerankers/dat-reranker-onnx-local/src/main/resources/META-INF/services/ai.dat.core.factories.ScoringModelFactory

@@ -0,0 +1 @@
+ai.dat.reranker.onnx.OnnxScoringModelFactory

dat-rerankers/pom.xml

@@ -18,6 +18,7 @@
         <module>dat-reranker-ms-marco-minilm-l6-v2-q</module>
         <module>dat-reranker-ms-marco-tinybert-l2-v2</module>
         <module>dat-reranker-ms-marco-tinybert-l2-v2-q</module>
+        <module>dat-reranker-onnx-local</module>
         <module>dat-reranker-jina</module>
     </modules>