@@ -2178,11 +2178,15 @@ class GenerateResponseState<const Functions extends ChatModelFunctions | undefin
21782178 ! this . disengageInitiallyEngagedFunctionMode . hasInProgressStops
21792179 )
21802180 break ;
2181+
2182+ const stopRes = this . handleAbortTrigger ( "model" ) ?? this . handleMaxTokensTrigger ( "model" ) ;
2183+ if ( stopRes != null )
2184+ return stopRes ;
21812185 }
21822186
2183- const abortRes = this . handleAbortTrigger ( "model" ) ;
2184- if ( abortRes != null )
2185- return abortRes ;
2187+ const stopRes = this . handleAbortTrigger ( "model" ) ?? this . handleMaxTokensTrigger ( "model" ) ;
2188+ if ( stopRes != null )
2189+ return stopRes ;
21862190
21872191 if ( this . disengageInitiallyEngagedFunctionMode . hasTriggeredStops ) {
21882192 const lastTokensForDetokenizer = this . streamRegulator . getLastQueuedChunkTokens ( ) ;
@@ -2290,11 +2294,15 @@ class GenerateResponseState<const Functions extends ChatModelFunctions | undefin
22902294
22912295 if ( functionNameGenerationDoneDetector . hasTriggeredStops )
22922296 break ;
2297+
2298+ const stopRes = this . handleAbortTrigger ( "model" ) ?? this . handleMaxTokensTrigger ( "model" ) ;
2299+ if ( stopRes != null )
2300+ return stopRes ;
22932301 }
22942302
2295- const abortRes = this . handleAbortTrigger ( "model" ) ;
2296- if ( abortRes != null )
2297- return abortRes ;
2303+ const stopRes = this . handleAbortTrigger ( "model" ) ?? this . handleMaxTokensTrigger ( "model" ) ;
2304+ if ( stopRes != null )
2305+ return stopRes ;
22982306
22992307 const functionCallNameText = this . llamaChat . model . detokenize ( this . currentFunctionCallCurrentPartTokens ) ;
23002308 const functionName = functionNameGrammar . parseFunctionName ( functionCallNameText ) ;
@@ -2375,11 +2383,15 @@ class GenerateResponseState<const Functions extends ChatModelFunctions | undefin
23752383
23762384 if ( functionParamsGenerationDoneDetector . hasTriggeredStops )
23772385 break ;
2386+
2387+ const stopRes = this . handleAbortTrigger ( "model" ) ?? this . handleMaxTokensTrigger ( "model" ) ;
2388+ if ( stopRes != null )
2389+ return stopRes ;
23782390 }
23792391
2380- const abortRes = this . handleAbortTrigger ( "model" ) ;
2381- if ( abortRes != null )
2382- return abortRes ;
2392+ const stopRes = this . handleAbortTrigger ( "model" ) ?? this . handleMaxTokensTrigger ( "model" ) ;
2393+ if ( stopRes != null )
2394+ return stopRes ;
23832395
23842396 const functionCallParamsText =
23852397 this . llamaChat . model . detokenize ( this . currentFunctionCallCurrentPartTokens , false , lastPartTokens ) ;
@@ -2452,11 +2464,15 @@ class GenerateResponseState<const Functions extends ChatModelFunctions | undefin
24522464
24532465 if ( sectionSuffixDetector . hasTriggeredStops || ! sectionSuffixDetector . hasInProgressStops )
24542466 break ;
2467+
2468+ const stopRes = this . handleAbortTrigger ( "model" ) ?? this . handleMaxTokensTrigger ( "model" ) ;
2469+ if ( stopRes != null )
2470+ return stopRes ;
24552471 }
24562472
2457- const abortRes = this . handleAbortTrigger ( "model" ) ;
2458- if ( abortRes != null )
2459- return abortRes ;
2473+ const stopRes = this . handleAbortTrigger ( "model" ) ?? this . handleMaxTokensTrigger ( "model" ) ;
2474+ if ( stopRes != null )
2475+ return stopRes ;
24602476
24612477 if ( sectionSuffixDetector . hasTriggeredStops ) {
24622478 this . functionEvaluationMode = false ;
0 commit comments