Skip to content
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

Fix error caused by all_reduce call in domino #6880

Open
wants to merge 7 commits into
base: master
Choose a base branch
from
Open

Fix error caused by all_reduce call in domino #6880

Changes from 2 commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
b0ba18e
initialize communication backend
hwchen2017 Dec 16, 2024
2d3d688
Merge branch 'master' into hongwei/fix_domino_allreduce
hwchen2017 Dec 16, 2024
2833db7
add initialization check
hwchen2017 Dec 17, 2024
cc31ef5
Merge branch 'master' into hongwei/fix_domino_allreduce
tjruwase Dec 18, 2024
68698c7
Merge branch 'master' into hongwei/fix_domino_allreduce
hwchen2017 Dec 18, 2024
0ac1bb5
Merge branch 'master' into hongwei/fix_domino_allreduce
loadams Dec 19, 2024
327170c
Merge branch 'master' into hongwei/fix_domino_allreduce
tjruwase Dec 20, 2024
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
8 changes: 5 additions & 3 deletions deepspeed/runtime/domino/transformer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,13 +6,13 @@
import torch
import torch.nn.functional as F
from torch.nn.parameter import Parameter
import deepspeed
from deepspeed import comm as dist
import deepspeed.comm
from deepspeed.comm.comm import init_distributed
from deepspeed.accelerator import get_accelerator


def is_rank_0():
if dist.get_rank() == 0:
if deepspeed.comm.get_rank() == 0:
return True


Expand Down Expand Up @@ -249,6 +249,8 @@ def __init__(self,
output_bias=None):
super(DominoTransformerLayer, self).__init__()

init_distributed()
hwchen2017 marked this conversation as resolved.
Show resolved Hide resolved

self.llama_model = config.llama_model
self.layer_number = layer_number
self.layer_type = layer_type
Expand Down
Loading