Package com.hw.langchain.agents.agent

Class AgentExecutor

java.lang.Object

com.hw.langchain.chains.base.Chain

com.hw.langchain.agents.agent.AgentExecutor

public class AgentExecutor
extends Chain

Consists of an agent using tools.

Author:

HamaWhite

Field Summary

Fields inherited from class com.hw.langchain.chains.base.Chain

memory

Constructor Summary

Constructors

Constructor	Description
AgentExecutor()

Method Summary

Modifier and Type	Method	Description
String	chainType()
static AgentExecutor	fromAgentAndTools(BaseSingleActionAgent agent, List<BaseTool> tools)	Create from agent and tools.
AgentFinish	getToolReturn(org.apache.commons.lang3.tuple.Pair<AgentAction,String> nextStepOutput)	Check if the tool is a returning tool.
protected Map<String,String>	innerCall(Map<String,Object> inputs)	Run text through and get agent response.
List<String>	inputKeys()	Return the input keys.
List<String>	outputKeys()	Return the singular output key.
Map<String,String>	processOutput(AgentFinish output, List<org.apache.commons.lang3.tuple.Pair<AgentAction,String>> intermediateSteps)
Object	takeNextStep(Map<String,BaseTool> nameToToolMap, Map<String,Object> inputs, List<org.apache.commons.lang3.tuple.Pair<AgentAction,String>> intermediateSteps)	Take a single step in the thought-action-observation loop.

Methods inherited from class com.hw.langchain.chains.base.Chain

asyncCall, asyncCall, asyncInnerCall, asyncRun, asyncRun, call, call, prepInputs, run, run

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

AgentExecutor

public AgentExecutor()

Method Details

fromAgentAndTools

public static AgentExecutor fromAgentAndTools(BaseSingleActionAgent agent,
 List<BaseTool> tools)

Create from agent and tools.

chainType

public String chainType()

Specified by:

chainType in class Chain

inputKeys

public List<String> inputKeys()

Return the input keys.

Specified by:

inputKeys in class Chain

Returns:

the list of input keys

outputKeys

public List<String> outputKeys()

Return the singular output key.

Specified by:

outputKeys in class Chain

Returns:

the list of output keys

processOutput

public Map<String,String> processOutput(AgentFinish output,
 List<org.apache.commons.lang3.tuple.Pair<AgentAction,String>> intermediateSteps)

takeNextStep

public Object takeNextStep(Map<String,BaseTool> nameToToolMap,
 Map<String,Object> inputs,
 List<org.apache.commons.lang3.tuple.Pair<AgentAction,String>> intermediateSteps)

Take a single step in the thought-action-observation loop. Override this to take control of how the agent makes and acts on choices.

Returns:

AgentFinish or List<Pair<AgentAction, String>>

innerCall

protected Map<String,String> innerCall(Map<String,Object> inputs)

Run text through and get agent response.

Specified by:

innerCall in class Chain

Parameters:

inputs - the inputs to be processed by the chain

Returns:

a map containing the output generated by the chain

getToolReturn

public AgentFinish getToolReturn(org.apache.commons.lang3.tuple.Pair<AgentAction,String> nextStepOutput)

Check if the tool is a returning tool.