Skip to content

Use different ports for different GPU IDs #214

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
merged 2 commits into from
Sep 23, 2025
Merged

Use different ports for different GPU IDs #214

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
ace6559
different ports for different GPU IDs
allenwang28 Sep 22, 2025
3a93b04
Update src/forge/controller/provisioner.py
allenwang28 Sep 23, 2025
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
7 changes: 5 additions & 2 deletions src/forge/controller/provisioner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -177,15 +177,18 @@ async def get_proc_mesh(
# We can't currently do this because HostMesh only supports single
# proc_mesh creation at the moment. This will be possible once
# we have "proper HostMesh support".
def bootstrap(gpu_ids: int):
def bootstrap(gpu_ids: list[str]):
# This works for single host, needed for vLLM currently.
import os

os.environ["CUDA_VISIBLE_DEVICES"] = ",".join(gpu_ids)
os.environ["MASTER_ADDR"] = socket.gethostname()
# Multiple actors trying to call _get_port doesn't work
# os.environ["MASTER_PORT"] = _get_port()
os.environ["MASTER_PORT"] = "12345"

# Setting the last digit to the first GPU id allows us to i.e.
# create multiple vLLM instances on the same local host.
os.environ["MASTER_PORT"] = f"1234{gpu_ids[0]}"
os.environ["HYPERACTOR_MESSAGE_DELIVERY_TIMEOUT_SECS"] = "600"
os.environ["HYPERACTOR_CODE_MAX_FRAME_LENGTH"] = "1073741824"

Expand Down
Loading