Adaptive RAG¶

In [ ]:

var userHomeDir = System.getProperty("user.home");
var localRespoUrl = "file://" + userHomeDir + "/.m2/repository/";
var langchain4jVersion = "1.9.1";
var langchain4jbeta = "1.9.1-beta17";
var langgraph4jVersion = "1.7-SNAPSHOT";

var userHomeDir = System.getProperty("user.home"); var localRespoUrl = "file://" + userHomeDir + "/.m2/repository/"; var langchain4jVersion = "1.9.1"; var langchain4jbeta = "1.9.1-beta17"; var langgraph4jVersion = "1.7-SNAPSHOT";

In [2]:

%dependency /add-repo local \{localRespoUrl} release|never snapshot|always
// %dependency /list-repos
%dependency /add org.slf4j:slf4j-jdk14:2.0.9
%dependency /add org.bsc.langgraph4j:langgraph4j-core:\{langgraph4jVersion}
%dependency /add org.bsc.langgraph4j:langgraph4j-langchain4j:\{langgraph4jVersion}
%dependency /add dev.langchain4j:langchain4j:\{langchain4jVersion}
%dependency /add dev.langchain4j:langchain4j-open-ai:\{langchain4jVersion}

%dependency /resolve

%dependency /add-repo local \{localRespoUrl} release|never snapshot|always // %dependency /list-repos %dependency /add org.slf4j:slf4j-jdk14:2.0.9 %dependency /add org.bsc.langgraph4j:langgraph4j-core:\{langgraph4jVersion} %dependency /add org.bsc.langgraph4j:langgraph4j-langchain4j:\{langgraph4jVersion} %dependency /add dev.langchain4j:langchain4j:\{langchain4jVersion} %dependency /add dev.langchain4j:langchain4j-open-ai:\{langchain4jVersion} %dependency /resolve

Repository local url: file:///Users/bsorrentino/.m2/repository/ added.
Adding dependency org.slf4j:slf4j-jdk14:2.0.9
Adding dependency org.bsc.langgraph4j:langgraph4j-core:1.7-SNAPSHOT
Adding dependency org.bsc.langgraph4j:langgraph4j-langchain4j:1.7-SNAPSHOT
Adding dependency dev.langchain4j:langchain4j:1.9.1
Adding dependency dev.langchain4j:langchain4j-open-ai:1.9.1
Solving dependencies
Resolved artifacts count: 16
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/org/slf4j/slf4j-jdk14/2.0.9/slf4j-jdk14-2.0.9.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/org/slf4j/slf4j-api/2.0.9/slf4j-api-2.0.9.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/org/bsc/langgraph4j/langgraph4j-core/1.7-SNAPSHOT/langgraph4j-core-1.7-SNAPSHOT.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/org/bsc/async/async-generator/4.0.0-beta2/async-generator-4.0.0-beta2.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/org/bsc/langgraph4j/langgraph4j-langchain4j/1.7-SNAPSHOT/langgraph4j-langchain4j-1.7-SNAPSHOT.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/dev/langchain4j/langchain4j/1.9.1/langchain4j-1.9.1.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/dev/langchain4j/langchain4j-core/1.9.1/langchain4j-core-1.9.1.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/org/jspecify/jspecify/1.0.0/jspecify-1.0.0.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/com/fasterxml/jackson/core/jackson-annotations/2.20/jackson-annotations-2.20.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/com/fasterxml/jackson/core/jackson-core/2.20.1/jackson-core-2.20.1.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/com/fasterxml/jackson/core/jackson-databind/2.20.1/jackson-databind-2.20.1.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/org/apache/opennlp/opennlp-tools/2.5.4/opennlp-tools-2.5.4.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/dev/langchain4j/langchain4j-open-ai/1.9.1/langchain4j-open-ai-1.9.1.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/dev/langchain4j/langchain4j-http-client/1.9.1/langchain4j-http-client-1.9.1.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/dev/langchain4j/langchain4j-http-client-jdk/1.9.1/langchain4j-http-client-jdk-1.9.1.jar
Add to classpath: /Users/bsorrentino/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/com/knuddels/jtokkit/1.1.0/jtokkit-1.1.0.jar

Initialize Logger

In [3]:

try( var file = new java.io.FileInputStream("./logging.properties")) {
    java.util.logging.LogManager.getLogManager().readConfiguration( file );
}

var log = org.slf4j.LoggerFactory.getLogger("AdaptiveRag");

try( var file = new java.io.FileInputStream("./logging.properties")) { java.util.logging.LogManager.getLogManager().readConfiguration( file ); } var log = org.slf4j.LoggerFactory.getLogger("AdaptiveRag");

Test Issue #32¶

Issue concerns a problem on AdaptiveRag implementation referred to AnswerGrader task

In [4]:

import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.input.Prompt;
import dev.langchain4j.model.input.structured.StructuredPrompt;
import dev.langchain4j.model.input.structured.StructuredPromptProcessor;
import dev.langchain4j.model.openai.OpenAiChatModel;
import dev.langchain4j.model.output.structured.Description;
import dev.langchain4j.service.AiServices;
import dev.langchain4j.service.SystemMessage;
import java.time.Duration;
import java.util.function.Function;


public class AnswerGrader implements Function<AnswerGrader.Arguments,AnswerGrader.Score> {

    static final String MODELS[] =  { "gpt-3.5-turbo-0125", "gpt-4o-mini" };

    /**
     * Binary score to assess answer addresses question.
     */
    public static class Score {

        @Description("Answer addresses the question, 'yes' or 'no'")
        public String binaryScore;

        @Override
        public String toString() {
            return "Score: " + binaryScore;
        }
    }

    @StructuredPrompt("""
User question: 

{{question}}

LLM generation: 

{{generation}}
""")
    record Arguments(String question, String generation) {
    }

    interface Service {

        @SystemMessage("""
You are a grader assessing whether an answer addresses and/or resolves a question. 

Give a binary score 'yes' or 'no'. Yes, means that the answer resolves the question otherwise return 'no'
        """)
        Score invoke(String userMessage);
    }

    String openApiKey;

    @Override
    public Score apply(Arguments args) {
        var chatLanguageModel = OpenAiChatModel.builder()
                .apiKey( System.getenv("OPENAI_API_KEY")  )
                .modelName( MODELS[1] )
                .timeout(Duration.ofMinutes(2))
                .logRequests(true)
                .logResponses(true)
                .maxRetries(2)
                .temperature(0.0)
                .maxTokens(2000)
                .build();


        Service service = AiServices.create(Service.class, chatLanguageModel);

        Prompt prompt = StructuredPromptProcessor.toPrompt(args);

        log.trace( "prompt: {}", prompt.text() );
        
        return service.invoke(prompt.text());
    }

}

import dev.langchain4j.model.chat.ChatModel; import dev.langchain4j.model.input.Prompt; import dev.langchain4j.model.input.structured.StructuredPrompt; import dev.langchain4j.model.input.structured.StructuredPromptProcessor; import dev.langchain4j.model.openai.OpenAiChatModel; import dev.langchain4j.model.output.structured.Description; import dev.langchain4j.service.AiServices; import dev.langchain4j.service.SystemMessage; import java.time.Duration; import java.util.function.Function; public class AnswerGrader implements Function { static final String MODELS[] = { "gpt-3.5-turbo-0125", "gpt-4o-mini" }; /** * Binary score to assess answer addresses question. */ public static class Score { @Description("Answer addresses the question, 'yes' or 'no'") public String binaryScore; @Override public String toString() { return "Score: " + binaryScore; } } @StructuredPrompt(""" User question: {{question}} LLM generation: {{generation}} """) record Arguments(String question, String generation) { } interface Service { @SystemMessage(""" You are a grader assessing whether an answer addresses and/or resolves a question. Give a binary score 'yes' or 'no'. Yes, means that the answer resolves the question otherwise return 'no' """) Score invoke(String userMessage); } String openApiKey; @Override public Score apply(Arguments args) { var chatLanguageModel = OpenAiChatModel.builder() .apiKey( System.getenv("OPENAI_API_KEY") ) .modelName( MODELS[1] ) .timeout(Duration.ofMinutes(2)) .logRequests(true) .logResponses(true) .maxRetries(2) .temperature(0.0) .maxTokens(2000) .build(); Service service = AiServices.create(Service.class, chatLanguageModel); Prompt prompt = StructuredPromptProcessor.toPrompt(args); log.trace( "prompt: {}", prompt.text() ); return service.invoke(prompt.text()); } }

In [5]:

var grader = new AnswerGrader();

var args = new AnswerGrader.Arguments( "What are the four operations ? ", "LLM means Large Language Model" );
grader.apply( args );

var grader = new AnswerGrader(); var args = new AnswerGrader.Arguments( "What are the four operations ? ", "LLM means Large Language Model" ); grader.apply( args );

prompt: User question:

What are the four operations ? 

LLM generation:

LLM means Large Language Model
 

Out[5]:

Score: no

In [6]:

var args = new AnswerGrader.Arguments( "What are the four operations", "There are four basic operations: addition, subtraction, multiplication, and division." );   
grader.apply( args );

var args = new AnswerGrader.Arguments( "What are the four operations", "There are four basic operations: addition, subtraction, multiplication, and division." ); grader.apply( args );

prompt: User question:

What are the four operations

LLM generation:

There are four basic operations: addition, subtraction, multiplication, and division.
 

Out[6]:

Score: yes

In [7]:

var args = new AnswerGrader.Arguments( "What player at the Bears expected to draft first in the 2024 NFL draft?", "The Bears selected USC quarterback Caleb Williams with the No. 1 pick in the 2024 NFL Draft." );   
grader.apply( args );

var args = new AnswerGrader.Arguments( "What player at the Bears expected to draft first in the 2024 NFL draft?", "The Bears selected USC quarterback Caleb Williams with the No. 1 pick in the 2024 NFL Draft." ); grader.apply( args );

prompt: User question:

What player at the Bears expected to draft first in the 2024 NFL draft?

LLM generation:

The Bears selected USC quarterback Caleb Williams with the No. 1 pick in the 2024 NFL Draft.
 

Out[7]:

Score: yes