类class io.agentscope.core.training.strategy 2.x

SelectionDecision

public class SelectionDecision

源码Source io/agentscope/core/training/strategy/SelectionDecision.java:29–113

Selection 决策，对智能体行为进行评测、追踪与可观测性分析。位于 io.agentscope.core.training.strategy 包。核心成员包括 accept、reject、shouldTrain、getReason、getLabels、getMetadata。

Selection Decision - Selection decision result

Encapsulates the decision of whether to train, along with related metadata (labels, reason, etc.)

方法摘要Method Summary

方法Method	说明Description
`public static` `SelectionDecision accept(String reason)` static	设置该参数。 Create a "should train" decision
`public static` `SelectionDecision accept(String reason, String... labels)` static	执行 `accept` 方法。 Create a "should train" decision with labels
`public static` `SelectionDecision accept(String reason, List<String> labels, Map<String, Object> metadata)` static	执行 `accept` 方法。 Create a "should train" decision with labels and metadata
`public static` `SelectionDecision reject(String reason)` static	设置拒绝。 Create a "should not train" decision
`public boolean shouldTrain()`	返回 `shouldTrain` 结果。 Returns the should train.
`public String getReason()`	获取推理。 Returns the reason.
`public List<String> getLabels()`	获取 `getLabels`。 Returns the labels.
`public Map<String, Object> getMetadata()`	获取元数据。 Returns the metadata.
`public String toString()`	返回 `toString` 结果。 Returns the to string.

方法详情Method Detail

accept

public static SelectionDecision accept(String reason)

static

设置该参数。

Create a "should train" decision

参数Parameters

名称Name	类型Type	说明Description
`reason`	String	原因（类型 `String`）。The reason.

返回Returns SelectionDecision

返回SelectionDecision。Returns selection decision.

源码Source io/agentscope/core/training/strategy/SelectionDecision.java:56–58

accept

public static SelectionDecision accept(String reason, String... labels)

static

执行 accept 方法。

Create a "should train" decision with labels

参数Parameters

名称Name	类型Type	说明Description
`reason`	String	原因（类型 `String`）。The reason.
`labels` varargs	String...	labels（类型 `String...`）。The labels.

返回Returns SelectionDecision

返回SelectionDecision。Returns selection decision.

源码Source io/agentscope/core/training/strategy/SelectionDecision.java:63–65

accept

public static SelectionDecision accept(String reason, List<String> labels, Map<String, Object> metadata)

static

执行 accept 方法。

Create a "should train" decision with labels and metadata

参数Parameters

名称Name	类型Type	说明Description
`reason`	String	原因（类型 `String`）。The reason.
`labels`	List<String>	labels（类型 `List`）。The labels.
`metadata`	Map<String, Object>	元数据（类型 `Map`）。The metadata.

返回Returns SelectionDecision

返回SelectionDecision。Returns selection decision.

源码Source io/agentscope/core/training/strategy/SelectionDecision.java:70–73

reject

public static SelectionDecision reject(String reason)

static

设置拒绝。

Create a "should not train" decision

参数Parameters

名称Name	类型Type	说明Description
`reason`	String	原因（类型 `String`）。The reason.

返回Returns SelectionDecision

返回SelectionDecision。Returns selection decision.

源码Source io/agentscope/core/training/strategy/SelectionDecision.java:78–80

shouldTrain

public boolean shouldTrain()

返回 shouldTrain 结果。

Returns the should train.

返回Returns boolean

返回布尔值，表示判断结果。Returns a boolean result.

源码Source io/agentscope/core/training/strategy/SelectionDecision.java:83–85

getReason

public String getReason()

获取推理。

Returns the reason.

返回Returns String

返回字符串结果。Returns a string.

源码Source io/agentscope/core/training/strategy/SelectionDecision.java:87–89

getLabels

public List<String> getLabels()

获取 getLabels。

Returns the labels.

返回Returns List<String>

返回列出。Returns list.

源码Source io/agentscope/core/training/strategy/SelectionDecision.java:91–93

getMetadata

public Map<String, Object> getMetadata()

获取元数据。

Returns the metadata.

返回Returns Map<String, Object>

返回映射。Returns map.

源码Source io/agentscope/core/training/strategy/SelectionDecision.java:95–97

toString

public String toString()

返回 toString 结果。

Returns the to string.

返回Returns String

返回字符串结果。Returns a string.

源码Source io/agentscope/core/training/strategy/SelectionDecision.java:99–112