Skip to content

pruning old nightly docker containers, pass env vars #39

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
clayrosenthal merged 1 commit into from
Feb 4, 2026
Merged

pruning old nightly docker containers, pass env vars #39

Show file tree
Hide file tree
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
14 changes: 11 additions & 3 deletions .github/workflows/nightly-training.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,7 +37,7 @@ jobs:
extra_args: --command.setup_terms.motion_command.params.motion_config.motion_file=s3://far-holosoma-assets/data/motions/g1_29dof/whole_body_tracking/motion_dance_v3.npz --algo.config.buffer_size=384

name: Train ${{ matrix.experiment }} on ${{ matrix.multigpu == 'True' && 'multi-gpu' || 'single-gpu' }} using ${{ matrix.simulator }}
timeout-minutes: 720 # 12 hours
timeout-minutes: 1000 # ~16 hours

steps:
- name: Checkout code
Expand All @@ -48,13 +48,21 @@ jobs:
env:
WANDB_API_KEY: ${{ secrets.WANDB_PUB_API_KEY }}
shell: bash
timeout-minutes: 750 # 12.5 hours
timeout-minutes: 1000 # ~16 hours

run: |
# kill all old docker containers
Copy link
Contributor

@samuelgundry samuelgundry Feb 4, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

we'll eventually want multi runs on the same machine (for availability), so just noting we'll need to revisit this later

OLD_CONTAINERS=$(docker ps -a -q)
if [[ -n "$OLD_CONTAINERS" ]]; then
docker stop "$OLD_CONTAINERS" || true
docker rm --force "$OLD_CONTAINERS" || true
fi
eval "$(aws configure export-credentials --format env)"
docker run --gpus all --runtime=nvidia --shm-size=12g \
docker run --rm --gpus all --runtime=nvidia --shm-size=12g \
-v "$GITHUB_WORKSPACE:/workspace/holosoma" \
--env AWS_ACCESS_KEY_ID --env AWS_SECRET_ACCESS_KEY \
--env AWS_SESSION_TOKEN --env WANDB_API_KEY \
--env GITHUB_RUN_ID \
982423663241.dkr.ecr.us-west-2.amazonaws.com/holosoma:latest \
bash -c """
source scripts/source_${{ matrix.simulator }}_setup.sh
Expand Down
3 changes: 3 additions & 0 deletions src/holosoma/holosoma/config_types/experiment.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -78,6 +78,9 @@ class TrainingConfig:
name: str = "run"
"""Run name for logging. `logger.name` takes precedence if set."""

tags: tuple[str, ...] = ()
"""Optional tags to attach to the run for logging."""

# Evaluation settings
max_eval_steps: int | None = None
"""Maximum number of evaluation steps (None for unlimited)."""
Expand Down
2 changes: 1 addition & 1 deletion tests/nightly/nightly.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -100,7 +100,7 @@ def main():
config.logger,
project=f"nightly-{sanitized_exp}{multigpu_suffix}",
name=f"nightly-{sanitized_exp}{multigpu_suffix}-{now_timestamp()}",
tags=run_tags,
tags=tuple(run_tags),
),
)

Expand Down