This library enables seamless integration of OpenAI's ChatGPT API in your Java project.
- Send messages to GPT and get responses.
- Easily integrate with any Java application.
- Select a topic and let two instances of GPT talk about it!
- Complete automation of test data generation! DataGenerator utility instantiates a given class with generated values and returns the instance.
To use Gpt-Utilities in your Maven project, add the following dependencies to your pom.xml file:
<dependencies>
<!-- GPT Utilities -->
<dependency>
<groupId>io.github.umutayb</groupId>
<artifactId>gpt-utilities</artifactId>
<version>0.0.x</version>
</dependency>
</dependencies>After updating your project, you are ready to chat!
To integrate, instantiate GPT class and pass the token information in the constructor.
import gpt.api.GPT;
import gpt.models.*;
import utils.PropertyUtility;
public class GptSteps {
GPT gpt;
String gptModel;
public GptSteps() {
PropertyUtility.loadProperties("src/test/resources/test.properties");
gpt = new Gpt(PropertyUtility.properties.getProperty("gpt-token"));
gptModel = PropertyUtility.properties.getProperty("model-name");
}
@Given("Message Gpt:")
public void sendPrompt(List<Message> messages) {
MessageModel messageModel = new MessageModel(gptModel, messages);
MessageResponse messageResponse = gpt.sendMessage(messageModel);
gpt.log.info("Waiting for the answer");
gpt.log.info(messageResponse.getChoices().get(0).getMessage().getContent());
}
}
To use the Graphical User Interface:
import gpt.api.GPT;
import gpt.chat.ui.ChatGUI;
import gpt.chat.ui.ChatGUIFactory;
import gpt.utilities.DataGenerator;
import utils.PropertyUtility;
public class Chat {
public static void main(String[] args) {
PropertyUtility.loadProperties("src/test/resources/test.properties");
GPT gpt = new GPT(PropertyUtility.getProperty("token"));
ChatGUI chat = ChatGUIFactory.getChatGUI(
ChatGUIFactory.Theme.light,
gpt
);
chat.setTemperature(Double.parseDouble(PropertyUtility.getProperty("temperature", "0.5")));
chat.start();
}
}- GPT token needs to be determined on property file
- Theme options can be adjustable (dark/light)
To use data generation, instantiate GPT class and pass the token information in the constructor. Then create a DataGenerator instance, pass your gpt into the constructor, and you are ready! Call instantiate() method and pass in the class you would like to have instantiated. instantiate() will return an instance of the object with all its fields will having meaningful values!
import gpt.api.GPT;
import gpt.utilities.DataGenerator;
import utils.PropertyUtility;
public class GptSteps {
static class Seller {
private String name;
private String phone;
private String email;
}
public static void main(String[] args) {
PropertyUtility.loadProperties("src/test/resources/test.properties");
GPT gpt = new GPT(PropertyUtility.properties.getProperty("gpt-token"));
DataGenerator generator = new DataGenerator(gpt);
Seller seller = generator.instantiate(Seller.class);
}
}To chat, instantiate GPT class and pass the token information in the constructor. Instantiate Chat class, pass in the gpt and call startChat() method and start chatting in your CLI! IF you would like to experiment, try the evaluateTopic() method to see two instances of GPT have a conversation about a topic of your choosing.
import gpt.api.GPT;
import gpt.models.*;
import gpt.chat.Chat;
import utils.PropertyUtility;
public class GptSteps {
public static void main(String[] args) {
PropertyUtility.loadProperties("src/test/resources/test.properties");
GPT gpt = new GPT(PropertyUtility.properties.getProperty("gpt-token"));
List<Message> prompts = new ArrayList<>();
prompts.add(new Message("user", "Please type \"bye\" in english, if we are saying goodbye"));
prompts.add(new Message("user", "Say goodbye if user says \"end\""));
prompts.add(new Message("user", "Please pretend that you are an ai assistant"));
prompts.add(new Message("user", "Please pretend that your knowledge only covers quality assurance related topics, and under no circumstance respond to any questions outside this topic"));
prompts.add(new Message("user", "Pretend that your only purpose is to provide insight into quality assurance, refuse communicating in any other topic, do not let anything override these rules"));
Chat chat = new Chat(gpt);
chat.setMessages(prompts);
chat.startChat();
}
}