Skip to content

[NFC] polish colossalai/communication/collective.py #1262

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
binmakeswell merged 1 commit into from
Jul 12, 2022
Merged

[NFC] polish colossalai/communication/collective.py #1262

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 2 additions & 9 deletions colossalai/communication/collective.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,10 +10,7 @@
from colossalai.core import global_context as gpc


def all_gather(tensor: Tensor,
dim: int,
parallel_mode: ParallelMode,
async_op: bool = False) -> Tensor:
def all_gather(tensor: Tensor, dim: int, parallel_mode: ParallelMode, async_op: bool = False) -> Tensor:
r"""Gathers all tensors from the parallel group and concatenates them in a
specific dimension.

Expand Down Expand Up @@ -163,11 +160,7 @@ def broadcast(tensor: Tensor, src: int, parallel_mode: ParallelMode, async_op: b
return out


def reduce(tensor: Tensor,
dst: int,
parallel_mode: ParallelMode,
op: ReduceOp = ReduceOp.SUM,
async_op: bool = False):
def reduce(tensor: Tensor, dst: int, parallel_mode: ParallelMode, op: ReduceOp = ReduceOp.SUM, async_op: bool = False):
r"""Reduce tensors across whole parallel group. Only the process with
rank ``dst`` is going to receive the final result.

Expand Down