reduce num workers to 1

paul-grundmann · Aug 7, 2024 · 3254489 · 3254489
1 parent f66846a
commit 3254489
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/vllm/model_executor/guided_decoding/outlines_decoding.py b/vllm/model_executor/guided_decoding/outlines_decoding.py
@@ -71,7 +71,7 @@ async def get_outlines_guided_decoding_logits_processor(
 
     if global_thread_pool is None:
         global_thread_pool = concurrent.futures.ThreadPoolExecutor(
-            max_workers=2)
+            max_workers=1)
     loop = asyncio.get_running_loop()
 
     return await loop.run_in_executor(global_thread_pool,