prerna.engine.impl.model.responses.IModelEngineResponseHandler Maven / Gradle / Ivy
package prerna.engine.impl.model.responses;
import java.util.List;
import java.util.Map;
/**
* The {@code IModelEngineResponseHandler} interface should be implemented for any {@code IModelEngine} that
* makes inference calls using REST directly. It defines how responses from a REST call should be
* built or consolidated so that a standard model engine response can be created.
*
*
* This interface defines the structure for handling responses from a model engine,
* which potentially involves streaming data.
*/
public interface IModelEngineResponseHandler {
/**
* This method is intended to append or add a stream (or partial response) to the handler.
*
* It takes an instance of IModelEngineResponseStreamHandler as its parameter,
* which represents a part of the overall response from the model engine.
*
* @param partial
*/
void appendStream(IModelEngineResponseStreamHandler partial);
/**
* This method is used to created the full response after it has been
* assimilated from partial responses
*/
void setResponse(Object response);
/**
* This method returns a list of partial responses.
* Each item in the list is an instance of {@code IModelEngineResponseStreamHandler},
* representing a segment or part of the full response from the model engine.
*/
List getPartialResponses();
/**
* This method returns the Class type of the stream handler. It ensures that the returned
* class is a subclass of IModelEngineResponseStreamHandler.
*
* This is used to instantiate a {@code IModelEngineResponseStreamHandler} when processing the request response.
*/
Class getStreamHandlerClass();
/**
* This method returns a map which should contain the following keys:
*
response
* numberOfTokensInPrompt
* numberOfTokensInResponse
*
*
* Ideally the values for these keys come from {@code getResponse}, {@code getPromptTokens},
* and {@code getResponseTokens} respectively.
*/
Map getModelEngineResponse();
/**
* This method retrieves the complete response from the model engine.
* The return type is generic (Object), since the response could be anything.
*/
Object getResponse();
/**
* This method returns the number of tokens in the prompt that was sent to the {@code IModelEngine} class.
* Tokens are typically units of text, like words or sentences, depending on the model's tokenization scheme.
*/
Integer getPromptTokens();
/**
* This method returns the number of tokens in the text generation response from the {@code IModelEngine} class.
* Tokens are typically units of text, like words or sentences, depending on the model's tokenization scheme.
*/
Integer getResponseTokens();
}