Class LiteRtContextOverflowException

Namespace

LiteRtLmSharp

Assembly

LiteRtLmSharp.dll

Raised instead of letting a send overflow the conversation's KV cache. The KV cache is sized by MaxNumTokens and the native runtime does not police it: a send that grows past the limit writes beyond the allocated cache, corrupting native memory — the typical symptom is a deferred process crash (0xC0000005 / 0xC0000374) on a LATER call, with nothing pointing back at the overflowing send. The binding throws this instead, before native state is damaged.

public sealed class LiteRtContextOverflowException : LiteRtException, ISerializable

Inheritance

object

Exception

LiteRtException

LiteRtContextOverflowException

Implements

ISerializable

Inherited Members

Exception.GetBaseException()

Exception.GetType()

Exception.ToString()

Exception.Data

Exception.HelpLink

Exception.HResult

Exception.InnerException

Exception.Message

Exception.Source

Exception.StackTrace

Exception.TargetSite

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.ReferenceEquals(object, object)

Remarks

Thrown by the send methods of LiteRtConversation when the engine was loaded with an explicit MaxNumTokens and either (a) the conversation is already at the limit (TokenCount ≥ MaxNumTokens), or (b) the message's measured prefill cost leaves no room to decode a reply. Without an explicit MaxNumTokens the limit is internal to the engine and the guard is off.

A conversation that hit the limit cannot continue: dispose it and start a fresh one — with a trimmed History if the thread must go on — or reload the engine with a larger MaxNumTokens. In the Extensions.AI stateful mode the client evicts the live conversation automatically when this throws; resuming its ConversationId then fails with ArgumentException like any evicted id.

Constructors

LiteRtContextOverflowException(string, int, int)

Creates the exception. tokenCount is the conversation's KV-cache size when the send was rejected; maxNumTokens is the engine's configured context limit.

public LiteRtContextOverflowException(string message, int tokenCount, int maxNumTokens)

Parameters

message string

tokenCount int

maxNumTokens int

Properties

MaxNumTokens

The engine's configured context limit (MaxNumTokens).

public int MaxNumTokens { get; }

Property Value

int

TokenCount

The conversation's KV-cache token count when the send was rejected (TokenCount).

public int TokenCount { get; }

Property Value

int