Revert "add timeout logic"

This reverts commit bff38cb.

Author: cupofpython

server.js

@@ -3,10 +3,6 @@ const cors = require("cors");
 const axios = require("axios");
 const app = express();
 
-// Stream timeout settings
-const STREAM_TIMEOUT = 60000; // 60 seconds max for streaming responses
-const HTTP_TIMEOUT = 30000;   // 30 seconds for HTTP requests
-
 app.use(cors());
 app.use(express.json());
 
@@ -31,17 +27,12 @@ app.post("/execute", async (req, res) => {
 app.post("/stream", handleStreamRequest);
 app.get("/stream", handleStreamRequest);
 
-// Handler function for stream requests with performance optimizations
+// Handler function for stream requests
 async function handleStreamRequest(req, res) {
     // Get parameters from either query (GET) or body (POST)
     const model = req.method === 'GET' ? req.query.model : req.body.model;
     const prompt = req.method === 'GET' ? req.query.prompt : req.body.prompt;
-    
-    // Get optional performance parameters
-    const temperature = parseFloat((req.method === 'GET' ? req.query.temperature : req.body.temperature) || "0.7");
-    const maxTokens = parseInt((req.method === 'GET' ? req.query.max_tokens : req.body.max_tokens) || "2048");
-    const topP = parseFloat((req.method === 'GET' ? req.query.top_p : req.body.top_p) || "0.9");
-    
+  
     if (!model || !prompt) {
       return res.status(400).json({ error: "Model name and prompt are required" });
     }
@@ -50,82 +41,31 @@ async function handleStreamRequest(req, res) {
     res.setHeader('Content-Type', 'text/event-stream');
     res.setHeader('Cache-Control', 'no-cache');
     res.setHeader('Connection', 'keep-alive');
-    
-    // Setup the timeout for the stream
-    let streamTimeout = setTimeout(() => {
-        console.warn("Stream timeout reached");
-        res.write(`data: ${JSON.stringify({ error: "Stream timeout", done: true })}\n\n`);
-        res.end();
-    }, STREAM_TIMEOUT);
 
     try {
         const host = ("REACT_APP_MODEL_SERVICE" in process.env) ? process.env.REACT_APP_MODEL_SERVICE : "model-published";
-        const startTime = Date.now();
 
-        // Make a streaming request to Ollama with optimized parameters
+        // Make a streaming request to Ollama
         const response = await axios({
             method: 'post',
             url: `http://${host}:11434/api/generate`,
             data: {
                 model: model,
                 prompt: prompt,
-                stream: true,
-                options: {
-                    temperature: temperature,
-                    top_p: topP,
-                    max_tokens: maxTokens,
-                    num_ctx: 2048,        // Context window size
-                    num_gpu: 1,           // Use 1 GPU if available
-                    num_thread: 4         // Use 4 CPU threads
-                }
+                stream: true
             },
-            responseType: 'stream',
-            timeout: HTTP_TIMEOUT  // HTTP request timeout
+            responseType: 'stream'
         });
 
-        console.log(`Stream started for model ${model} at ${new Date().toISOString()}`);
-        
-        // Forward the stream to the client with performance monitoring
-        let tokensGenerated = 0;
-        let firstTokenTime = null;
-
+        // Forward the stream to the client
         response.data.on('data', (chunk) => {
             try {
-                // Clear the timeout on each received chunk
-                clearTimeout(streamTimeout);
-                // Reset timeout
-                streamTimeout = setTimeout(() => {
-                    console.warn("Stream timeout between chunks");
-                    res.write(`data: ${JSON.stringify({ error: "Stream timeout between responses", done: true })}\n\n`);
-                    res.end();
-                }, STREAM_TIMEOUT);
-
                 const data = JSON.parse(chunk.toString());
-                
-                // Track first token time
-                if (data.response && tokensGenerated === 0) {
-                    firstTokenTime = Date.now();
-                    console.log(`First token after ${firstTokenTime - startTime}ms`);
-                }
-                
-                // Count tokens
-                if (data.response) {
-                    tokensGenerated += 1;
-                }
-                
                 // Send each chunk as an SSE event
                 res.write(`data: ${JSON.stringify(data)}\n\n`);
 
-                // If this is the final response, end the connection and log performance
+                // If this is the final response, end the connection
                 if (data.done) {
-                    const endTime = Date.now();
-                    const totalTime = endTime - startTime;
-                    const tokensPerSecond = tokensGenerated / (totalTime / 1000);
-                    
-                    console.log(`Stream completed: ${tokensGenerated} tokens in ${totalTime}ms (${tokensPerSecond.toFixed(2)} tokens/sec)`);
-                    
-                    // Clean up the timeout
-                    clearTimeout(streamTimeout);
                     res.end();
                 }
             } catch (err) {
@@ -138,13 +78,11 @@ async function handleStreamRequest(req, res) {
         response.data.on('error', (err) => {
             console.error("Stream error:", err);
             res.write(`data: ${JSON.stringify({ error: "Stream error" })}\n\n`);
-            clearTimeout(streamTimeout);
             res.end();
         });
     } catch (err) {
         console.error("Streaming error: ", err);
         res.write(`data: ${JSON.stringify({ error: "Server error", message: err.message })}\n\n`);
-        clearTimeout(streamTimeout);
         res.end();
     }
 }
@@ -154,17 +92,7 @@ function getResponse(model, prompt) {
     return axios.post(`http://${host}:11434/api/generate`, {
         model: model,
         prompt: prompt,
-        stream: false,
-        options: {
-            temperature: 0.7,
-            top_p: 0.9,
-            max_tokens: 2048,
-            num_ctx: 2048,       // Context window size
-            num_gpu: 1,          // Use 1 GPU if available
-            num_thread: 4        // Use 4 CPU threads
-        }
-    }, {
-        timeout: HTTP_TIMEOUT
+        stream: false
     })
     .then(res => {
         console.log("Output received");