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

[trainer] [deepspeed] refactor deepspeed setup devices #9880

Merged
merged 1 commit into from
Jan 29, 2021
Merged

[trainer] [deepspeed] refactor deepspeed setup devices #9880

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
30 changes: 15 additions & 15 deletions src/transformers/training_args.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -522,6 +522,20 @@ def _setup_devices(self) -> "torch.device":
self.local_rank = dist.get_local_rank()
device = torch.device("cuda", self.local_rank)
self._n_gpu = 1
elif self.deepspeed:
# deepspeed performs its own DDP internally, and requires the program to be started with:
# deepspeed ./program.py
# rather than:
# python -m torch.distributed.launch --nproc_per_node=2 ./program.py
from .integrations import is_deepspeed_available

if not is_deepspeed_available():
raise ImportError("--deepspeed requires deepspeed: `pip install deepspeed`.")
import deepspeed

deepspeed.init_distributed()
device = torch.device("cuda", self.local_rank)
self._n_gpu = 1
elif self.local_rank == -1:
# if n_gpu is > 1 we'll use nn.DataParallel.
# If you only want to use a specific subset of GPUs use `CUDA_VISIBLE_DEVICES=0`
Expand All @@ -536,21 +550,7 @@ def _setup_devices(self) -> "torch.device":
else:
# Here, we'll use torch.distributed.
# Initializes the distributed backend which will take care of synchronizing nodes/GPUs
#
# deepspeed performs its own DDP internally, and requires the program to be started with:
# deepspeed ./program.py
# rather than:
# python -m torch.distributed.launch --nproc_per_node=2 ./program.py
if self.deepspeed:
from .integrations import is_deepspeed_available

if not is_deepspeed_available():
raise ImportError("--deepspeed requires deepspeed: `pip install deepspeed`.")
import deepspeed

deepspeed.init_distributed()
else:
torch.distributed.init_process_group(backend="nccl")
torch.distributed.init_process_group(backend="nccl")
device = torch.device("cuda", self.local_rank)
self._n_gpu = 1

Expand Down