Fix 500 error on Warm Models Endpoint

Since we don't really use warm param the same and only support HF Inference in this playgound, we'll need to update to use this endpoint.

src/routes/+page.server.ts

@@ -4,7 +4,7 @@ import type { PageServerLoad } from "./$types";
 import { env } from "$env/dynamic/private";
 
 export const load: PageServerLoad = async ({ fetch }) => {
-	const apiUrl = "https://huggingface.co/api/models?pipeline_tag=text-generation&inference=warm&filter=conversational";
+	const apiUrl = "https://huggingface.co/api/models?pipeline_tag=text-generation&inference_provider=hf-inference&filter=conversational";
 	const HF_TOKEN = env.HF_TOKEN;
 
 	const res = await fetch(apiUrl, {