From 4bcc508473c30e9bfd9c91df359896d05f645925 Mon Sep 17 00:00:00 2001 From: Lucas Armand Date: Tue, 25 Nov 2025 16:54:17 -0800 Subject: [PATCH] reduce vllm benchmark runs to 2 --- workers/openai/worker.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/workers/openai/worker.py b/workers/openai/worker.py index 105b8df..6cf17f0 100644 --- a/workers/openai/worker.py +++ b/workers/openai/worker.py @@ -57,7 +57,8 @@ worker_config = WorkerConfig( max_queue_time=60.0, benchmark_config=BenchmarkConfig( generator=completions_benchmark_generator, - concurrency=100 + concurrency=100, + runs=2 ) ), HandlerConfig(