fix: incorrect default max_tokens set - legacy issue

extensions/inference-cortex-extension/bin/version.txt

@@ -1 +1 @@
-1.0.9-rc1
+1.0.9-rc2

web/containers/ModelDropdown/index.tsx

@@ -123,8 +123,8 @@
 
   const handleChangeStateOpen = useCallback(
     (state: boolean) => {
       setOpen(state)
       setModelDropdownState(state)
     },
     [setModelDropdownState]
   )
@@ -134,7 +134,7 @@
       configuredModels
         .concat(
           downloadedModels.filter(
             (e) => !configuredModels.some((x) => x.id === e.id)
           )
         )
         .filter((e) =>
@@ -147,22 +147,22 @@
               'local'
             )
           }
           return true
         })
         .sort((a, b) => a.name.localeCompare(b.name))
         .sort((a, b) => {
           const aInDownloadedModels = downloadedModels.some(
             (item) => item.id === a.id
           )
           const bInDownloadedModels = downloadedModels.some(
             (item) => item.id === b.id
           )
           if (aInDownloadedModels && !bInDownloadedModels) {
             return -1
           } else if (!aInDownloadedModels && bInDownloadedModels) {
             return 1
           } else {
             return 0
           }
         }),
     [configuredModels, searchText, searchFilter, downloadedModels, engineList]
@@ -184,8 +184,8 @@
     setShowEngineListModel((prev) => [
       ...prev,
       ...engineList
         .filter((x) => (x.engine?.api_key?.length ?? 0) > 0)
         .map((e) => e.name),
     ])
   }, [setShowEngineListModel, engineList])
 
@@ -205,8 +205,8 @@
 
   const isLocalEngine = useCallback(
     (engine?: string) => {
       if (!engine) return false
       return engineList.some((t) => t.name === engine && t.type === 'local')
     },
     [engineList]
   )
@@ -248,10 +248,10 @@
           ctx_len: model?.settings.ctx_len ? defaultContextLength : undefined,
           max_tokens: defaultContextLength
             ? Math.min(
-                model?.parameters.token_limit ?? 8192,
+                model?.parameters.max_tokens ?? 8192,
                 defaultContextLength
               )
-            : model?.parameters.token_limit,
+            : model?.parameters.max_tokens,
         }
 
         const modelParams = {