Merge b20ea70089 into 115e6b2a71

sample.config.toml

@@ -19,7 +19,7 @@ API_KEY = ""
 API_URL = "" # Ollama API URL - http://host.docker.internal:11434
 
 [MODELS.LMSTUDIO]
-API_URL = "" # LM STUDIO API URL - http://host.docker.internal:1234
+API_URL = "" # LM STUDIO API URL - http://host.docker.internal:1234/v1
 
 [MODELS.CUSTOM_OPENAI]
 API_KEY = ""

src/lib/providers/lmstudio.ts

@@ -4,10 +4,6 @@ import { getKeepAlive, getLMStudioApiEndpoint } from '../../config';
 import logger from '../../utils/logger';
 import axios from 'axios';
 
-const ensureV1Endpoint = (endpoint: string): string => {
-  return endpoint.endsWith('/v1') ? endpoint : `${endpoint}/v1`;
-};
-
 interface LMStudioModel {
   id: string;
   // add other properties if LM Studio API provides them
@@ -18,22 +14,6 @@ interface ChatModelConfig {
   model: ChatOpenAI;
 }
 
-const checkLMStudioAvailability = async (endpoint: string): Promise<boolean> => {
-  const v1Endpoint = ensureV1Endpoint(endpoint);
-  try {
-    await axios.get(`${v1Endpoint}/models`, {
-      timeout: 1000, // 1 second timeout
-      headers: {
-        'Content-Type': 'application/json',
-      },
-    });
-    return true;
-  } catch (err) {
-    logger.debug(`LM Studio server not available at ${endpoint}`);
-    return false;
-  }
-};
-
 export const loadLMStudioChatModels = async (): Promise<Record<string, ChatModelConfig>> => {
   const lmStudioEndpoint = getLMStudioApiEndpoint();
 
@@ -42,16 +22,8 @@ export const loadLMStudioChatModels = async (): Promise<Record<string, ChatModel
     return {};
   }
 
-  // Check if server is available before attempting to load models
-  const isAvailable = await checkLMStudioAvailability(lmStudioEndpoint);
-  if (!isAvailable) {
-    return {};
-  }
-
   try {
-    const v1Endpoint = ensureV1Endpoint(lmStudioEndpoint);
+    const response = await axios.get<{ data: LMStudioModel[] }>(`${lmStudioEndpoint}/models`, {
-    const response = await axios.get<{ data: LMStudioModel[] }>(`${v1Endpoint}/models`, {
-      timeout: 5000, // 5 second timeout for model loading
       headers: {
         'Content-Type': 'application/json',
       },
@@ -65,7 +37,7 @@ export const loadLMStudioChatModels = async (): Promise<Record<string, ChatModel
         model: new ChatOpenAI({
           openAIApiKey: 'lm-studio',
           configuration: {
-            baseURL: ensureV1Endpoint(lmStudioEndpoint),
+            baseURL: lmStudioEndpoint,
           },
           modelName: model.id,
           temperature: 0.7,
@@ -86,16 +58,8 @@ export const loadLMStudioEmbeddingsModels = async () => {
 
   if (!lmStudioEndpoint) return {};
 
-  // Check if server is available before attempting to load models
-  const isAvailable = await checkLMStudioAvailability(lmStudioEndpoint);
-  if (!isAvailable) {
-    return {};
-  }
-
   try {
-    const v1Endpoint = ensureV1Endpoint(lmStudioEndpoint);
+    const response = await axios.get(`${lmStudioEndpoint}/models`, {
-    const response = await axios.get(`${v1Endpoint}/models`, {
-      timeout: 5000, // 5 second timeout for model loading
       headers: {
         'Content-Type': 'application/json',
       },
@@ -109,7 +73,7 @@ export const loadLMStudioEmbeddingsModels = async () => {
         model: new OpenAIEmbeddings({
           openAIApiKey: 'lm-studio', // Dummy key required by LangChain
           configuration: {
-            baseURL: ensureV1Endpoint(lmStudioEndpoint),
+            baseURL: lmStudioEndpoint,
           },
           modelName: model.id,
         }),
@@ -122,4 +86,4 @@ export const loadLMStudioEmbeddingsModels = async () => {
     logger.error(`Error loading LM Studio embeddings model: ${err}`);
     return {};
   }
-};
+};