Skip to content

fix: restore standalone GPU trainer launch #3362

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

Open
LeonEricsson wants to merge 3 commits into
base: main
Choose a base branch
from
Open

fix: restore standalone GPU trainer launch #3362

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
39a9917
fix(trainer): restore standalone single-GPU launch
LeonEricsson Aug 23, 2026
769067d
chore(trainer): narrow standalone launch fix
LeonEricsson Aug 23, 2026
1e71a52
nit
LeonEricsson Aug 23, 2026
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
13 changes: 12 additions & 1 deletion src/prime_rl/trainer/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -189,7 +189,18 @@ def setup_torch_distributed(timeout: timedelta = DEFAULT_TIMEOUT, enable_gloo: b
# module default so every subsequently created PG inherits it.
dist.distributed_c10d.default_pg_timeout = timeout

dist.init_process_group(backend=backend, timeout=timeout, device_id=device_id)
if "RANK" in os.environ:
dist.init_process_group(backend=backend, timeout=timeout, device_id=device_id)
else:
get_logger().info("Using standalone single-process distributed mode")
dist.init_process_group(
backend=backend,
store=dist.HashStore(),
rank=0,
world_size=1,
timeout=timeout,
device_id=device_id,
)
get_logger().debug(f"Initialized torch distributed in {format_time(time.perf_counter() - t0)}")


Expand Down