From 0bcd2219ea550c4f34bb67702bf555fd16a06057 Mon Sep 17 00:00:00 2001
From: Lucas Armand <lucas@vast.ai>
Date: Wed, 3 Dec 2025 12:38:52 -0800
Subject: [PATCH] Increase model wait time for vLLM

---
 workers/openai/server.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/workers/openai/server.py b/workers/openai/server.py
index 8dc962f..63f21f9 100644
--- a/workers/openai/server.py
+++ b/workers/openai/server.py
@@ -35,6 +35,7 @@ backend = Backend(
     model_server_url=os.environ["MODEL_SERVER_URL"],
     model_log_file=os.environ["MODEL_LOG"],
     allow_parallel_requests=True,
+    max_wait_time=600.0,
     benchmark_handler=CompletionsHandler(benchmark_runs=3, benchmark_words=256),
     log_actions=[
         *[(LogAction.ModelLoaded, info_msg) for info_msg in MODEL_SERVER_START_LOG_MSG],