From cd27baaffd3b642518c32ffc1679ac3853aac4ee Mon Sep 17 00:00:00 2001 From: Kangyan-Zhou Date: Fri, 1 May 2026 22:50:38 -0700 Subject: [PATCH] [ci][cu13] Bump torch_memory_saver to 0.0.9.post1; restore manual tests (#23182) Co-authored-by: Claude Opus 4.7 (1M context) --- python/pyproject.toml | 2 +- .../model_loading/test_utils_update_weights.py | 5 ----- .../rl/test_multi_instance_release_memory_occupation.py | 5 ----- 3 files changed, 1 insertion(+), 11 deletions(-) rename test/{manual => registered}/model_loading/test_utils_update_weights.py (95%) rename test/{manual => registered}/rl/test_multi_instance_release_memory_occupation.py (98%) diff --git a/python/pyproject.toml b/python/pyproject.toml index a906ebf9d..b54ff67cc 100755 --- a/python/pyproject.toml +++ b/python/pyproject.toml @@ -63,7 +63,7 @@ dependencies = [ "soundfile==0.13.1", "tiktoken", "timm==1.0.16", - "torch_memory_saver==0.0.9", + "torch_memory_saver>=0.0.9.post1", "torch==2.9.1", "torchao==0.17.0", "torchaudio==2.9.1", diff --git a/test/manual/model_loading/test_utils_update_weights.py b/test/registered/model_loading/test_utils_update_weights.py similarity index 95% rename from test/manual/model_loading/test_utils_update_weights.py rename to test/registered/model_loading/test_utils_update_weights.py index 18096ff7d..f79b6306c 100644 --- a/test/manual/model_loading/test_utils_update_weights.py +++ b/test/registered/model_loading/test_utils_update_weights.py @@ -14,11 +14,6 @@ from sglang.test.test_utils import DEFAULT_SMALL_MODEL_NAME_FOR_TEST register_cuda_ci(est_time=32, suite="stage-b-test-1-gpu-large") -""" -# TODO: torch_memory_saver wheel is built against libcudart.so.12, fails to LD_PRELOAD in Cu13 venv. Ref: https://github.com/sgl-project/sglang/actions/runs/24604424372/job/71968573867 -# Should move back to registered test after it's fixed -""" - class AsyncEngine(Engine): def __init__(self, **kwargs): diff --git a/test/manual/rl/test_multi_instance_release_memory_occupation.py b/test/registered/rl/test_multi_instance_release_memory_occupation.py similarity index 98% rename from test/manual/rl/test_multi_instance_release_memory_occupation.py rename to test/registered/rl/test_multi_instance_release_memory_occupation.py index 7cb550390..5484b55b8 100644 --- a/test/manual/rl/test_multi_instance_release_memory_occupation.py +++ b/test/registered/rl/test_multi_instance_release_memory_occupation.py @@ -1,8 +1,3 @@ -""" -# TODO: Fails on cu13 venv migration. Ref: https://github.com/sgl-project/sglang/actions/runs/24616960626/job/71980705675?pr=23119 -# Should move back to registered test after it's fixed -""" - import gc import multiprocessing import os