maxtoken

@@ -13,4 +13,6 @@ OPENSOURCE_MODELS="openchat_3.5,zephyr-7B-beta"
 OPENSOURCE_ENDPOINTS={"openchat_3.5": "http://192.168.22.180:5000/v1", "zephyr-7B-beta": "http://192.168.22.180:5000/v1"}
 OPENSOURCE_ENDPOINTS={"openchat_3.5": "http://192.168.22.180:5000/v1", "zephyr-7B-beta": "http://192.168.22.180:5000/v1"}
 OPENSOURCE_API_KEY={"openchat_3.5": "Q8QZG0HhS62piZSmvSR7CP3E53fY8J99Ad2wmJLdiW05cpjnnBAH5xlQlClYdqeh", "zephyr-7B-beta": "Q8QZG0HhS62piZSmvSR7CP3E53fY8J99Ad2wmJLdiW05cpjnnBAH5xlQlClYdqeh"}
 OPENSOURCE_API_KEY={"openchat_3.5": "Q8QZG0HhS62piZSmvSR7CP3E53fY8J99Ad2wmJLdiW05cpjnnBAH5xlQlClYdqeh", "zephyr-7B-beta": "Q8QZG0HhS62piZSmvSR7CP3E53fY8J99Ad2wmJLdiW05cpjnnBAH5xlQlClYdqeh"}
 
 
-REACT_APP_START_SERVER_URL= https://dev.start.ai-pro.org                                    
\ No newline at end of file
+REACT_APP_START_SERVER_URL= https://dev.start.ai-pro.org                                    
+
+MAX_TOKENS_chatbot_plus=100
\ No newline at end of file

@@ -69,6 +69,7 @@ app.use(cors())
 app.use(require('morgan')('dev'))
 app.use(require('morgan')('dev'))
 app.use(rateLimiter)
 app.use(rateLimiter)
 
 
+const max_tokens = process.env.MAX_TOKENS_chatbot_plus ? parseInt(process.env.MAX_TOKENS_chatbot_plus) : 512;
 // Routing
 // Routing
 
 
 // Primary Open AI Route
 // Primary Open AI Route
@@ -119,7 +120,7 @@ app.post('/api', async (req, res) => {
     const response = await openai.createCompletion({
     const response = await openai.createCompletion({
       model: `${currentModel}`,// "text-davinci-003",
       model: `${currentModel}`,// "text-davinci-003",
       prompt: query_prompt,
       prompt: query_prompt,
-      max_tokens: 3000,
+      max_tokens: max_tokens,
       temperature,
       temperature,
     });
     });
     let input = response.data.choices[0].text;
     let input = response.data.choices[0].text;
@@ -152,7 +153,7 @@ app.post('/api', async (req, res) => {
         message: "The output for your prompt is too long for us to process. Please reduce your prompt and try again.",
         message: "The output for your prompt is too long for us to process. Please reduce your prompt and try again.",
       })
       })
     } else {
     } else {
-      console.log(e.response);
+      // console.log(e.response);
     }
     }
   } finally {
   } finally {
     // console.log('We do cleanup here');
     // console.log('We do cleanup here');
@@ -181,7 +182,7 @@ async function runGPTTurbo(req, res) {
     const response = await openai.createChatCompletion({
     const response = await openai.createChatCompletion({
       model: `${currentModel}`,
       model: `${currentModel}`,
       messages: JSON.parse(message),
       messages: JSON.parse(message),
-      max_tokens: 3000,
+      max_tokens: max_tokens,
       temperature
       temperature
     });
     });
     input = response.data.choices[0].message.content
     input = response.data.choices[0].message.content
@@ -190,7 +191,7 @@ async function runGPTTurbo(req, res) {
     if (error_msg.indexOf('maximum context length') >= 0) {
     if (error_msg.indexOf('maximum context length') >= 0) {
       input = "The output for your prompt is too long for us to process. Please reduce your prompt and try again.";
       input = "The output for your prompt is too long for us to process. Please reduce your prompt and try again.";
     } else {
     } else {
-      console.log(e.response);
+      // console.log(e.response);
     }
     }
   } finally {
   } finally {
 
 
@@ -266,7 +267,7 @@ async function runOpensource(req, res) {
     if (error_msg.indexOf('maximum context length') >= 0) {
     if (error_msg.indexOf('maximum context length') >= 0) {
       input = "The output for your prompt is too long for us to process. Please reduce your prompt and try again.";
       input = "The output for your prompt is too long for us to process. Please reduce your prompt and try again.";
     } else {
     } else {
-      console.log(e.response);
+      // console.log(e.response);
     }
     }
   } finally {
   } finally {