chore: target specific GPU

.gitignore

@@ -1,3 +1,4 @@
 ovh.conf
 .env
-test.js
+monitoring.sh
+notes.md

docker-compose.yaml

@@ -26,6 +26,7 @@ services:
             - driver: nvidia
               count: 1
               capabilities: [gpu]
+              device_ids: ['0']
     volumes:
       - "./.ollama-service-1:/root/.ollama"
     networks:
@@ -58,6 +59,7 @@ services:
             - driver: nvidia
               count: 1
               capabilities: [gpu]
+              device_ids: ['1']
     volumes:
       - "./.ollama-service-2:/root/.ollama"
     networks:

test.js

@@ -0,0 +1,37 @@
+import http from "k6/http";
+import { check } from "k6";
+
+export const options = { vus: 4, iterations: 40 };
+
+const url = `https://${__ENV.HOST_IP}.nip.io/api/generate`;
+
+const prompts = [
+  "Explain how a car engine works",
+  "Write a short story about a magical forest",
+  "What are the main principles of machine learning?",
+  "Describe the process of photosynthesis",
+  "Write a recipe for chocolate cake",
+  "Explain the theory of relativity",
+  "What are the benefits of meditation?",
+  "How does the internet work?",
+];
+
+const params = {
+  headers: {
+    "Content-Type": "application/json",
+    Authorization: `Basic ${__ENV.HOST_CREDENTIALS}`
+  },
+  timeout: '300s',
+};
+
+export default function () {
+  const randomIndex = Math.floor(Math.random() * prompts.length);
+  const prompt = prompts[randomIndex];
+
+  const payload = JSON.stringify({ prompt, model: "llama3.1" });
+  const response = http.post(url, payload, params);
+
+  check(response, {
+    "status is 200": (r) => r.status === 200,
+  });
+}