From ee8f36d0b3dc36ad1751c2193b30feb2ba11e0da Mon Sep 17 00:00:00 2001 From: Woosuk Kwon Date: Thu, 16 Jul 2026 20:17:51 -0700 Subject: [PATCH] [Warmup] Show CuTeDSL compilation progress (#48881) Signed-off-by: Woosuk Kwon Co-authored-by: Bugen Zhao Co-authored-by: Giancarlo Delfin <32987265+TheEpicDolphin@users.noreply.github.com> Co-authored-by: Isotr0py Co-authored-by: Isotr0py Co-authored-by: Jee Jee Li Co-authored-by: Roger Wang Co-authored-by: Yifan Qiao Co-authored-by: Claude Fable 5 Co-authored-by: OpenAI Codex --- vllm/model_executor/warmup/cutedsl_warmup.py | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/vllm/model_executor/warmup/cutedsl_warmup.py b/vllm/model_executor/warmup/cutedsl_warmup.py index 5978e91a6e..991eab22b8 100644 --- a/vllm/model_executor/warmup/cutedsl_warmup.py +++ b/vllm/model_executor/warmup/cutedsl_warmup.py @@ -10,7 +10,9 @@ from collections.abc import Callable, Hashable, Iterable from dataclasses import dataclass import torch +from tqdm import tqdm +from vllm.distributed import is_global_first_rank from vllm.logger import init_logger from vllm.platforms import current_platform from vllm.tracing import instrument @@ -76,6 +78,8 @@ def _compile_cutedsl_warmup_units( compile_units: Iterable[CuTeDSLCompileUnit], ) -> int: compiled = 0 + if is_global_first_rank(): + compile_units = tqdm(compile_units, desc="Compiling CuTeDSL kernels") with torch.inference_mode(): for unit in compile_units: unit.compile()