frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-21 13:44:15 +08:00
Code Issues Packages Projects Releases Wiki Activity

Nccl update to 2.25.1 for cuda 12.4-12.8 (#146073)

Browse Source 
Should resolve: https://github.com/pytorch/pytorch/issues/144768
We use one common nccl version for cuda builds 12.4-12.8 : ``NCCL_VERSION=v2.25.1-1``
For CUDA 11.8 we use legacy ``NCCL_VERSION=v2.21.1-1``
We use pinned version of NCCL rather then submodule.
Move nccl location from ``third_party/nccl/nccl`` to ``third_party/nccl``

Pull Request resolved: https://github.com/pytorch/pytorch/pull/146073
Approved by: https://github.com/Skylion007, https://github.com/malfet, https://github.com/kwen2501, https://github.com/fduwjj
This commit is contained in:
atalman
2025-02-14 15:29:59 +00:00
committed by PyTorch MergeBot
parent cefd9805de
commit 06f4a5c0e5
19 changed files with 139 additions and 111 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
setup.py
View File

@ -401,6 +401,28 @@ def get_submodule_folders():
]
def read_nccl_pin() -> str:
nccl_file = "nccl-cu12.txt"
if os.getenv("DESIRED_CUDA", "").startswith("11") or os.getenv(
"CUDA_VERSION", ""
).startswith("11"):
nccl_file = "nccl-cu11.txt"
nccl_pin_path = os.path.join(cwd, ".ci", "docker", "ci_commit_pins", nccl_file)
with open(nccl_pin_path) as f:
return f.read().strip()
def checkout_nccl():
release_tag = read_nccl_pin()
report(f"-- Checkout nccl release tag: {release_tag}")
nccl_basedir = os.path.join(third_party_path, "nccl")
subprocess.check_call(
["git", "clone", "https://github.com/NVIDIA/nccl.git", "nccl"],
cwd=third_party_path,
)
subprocess.check_call(["git", "checkout", release_tag], cwd=nccl_basedir)
def check_submodules():
def check_for_files(folder, files):
if not any(os.path.exists(os.path.join(folder, f)) for f in files):
@ -419,16 +441,16 @@ def check_submodules():
# If none of the submodule folders exists, try to initialize them
if all(not_exists_or_empty(folder) for folder in folders):
try:
print(" --- Trying to initialize submodules")
report(" --- Trying to initialize submodules")
start = time.time()
subprocess.check_call(
["git", "submodule", "update", "--init", "--recursive"], cwd=cwd
)
end = time.time()
print(f" --- Submodule initialization took {end - start:.2f} sec")
report(f" --- Submodule initialization took {end - start:.2f} sec")
except Exception:
print(" --- Submodule initalization failed")
print("Please run:\n\tgit submodule update --init --recursive")
report(" --- Submodule initalization failed")
report("Please run:\n\tgit submodule update --init --recursive")
sys.exit(1)
for folder in folders:
check_for_files(
@ -484,7 +506,7 @@ def mirror_files_into_torchgen():
# all the work we need to do _before_ setup runs
def build_deps():
report("-- Building version " + version)
checkout_nccl()
check_submodules()
check_pydep("yaml", "pyyaml")
build_python = not BUILD_LIBTORCH_WHL