Staying In Touch
Thank you for participating! Now that the course is done, we’d love your feedback on how it went and what we can improve for future runs.
Please fill out the course survey: darc.stanford.edu/class-survey
Also, note that the DARC team runs the Yens and supports GSB researchers year-round. You are not expected to remember everything from this week — but you are expected to know where to ask.
Slack — #gsb-yen-users
Join the #gsb-yen-users channel on Stanford Slack. It’s where Yen users and the DARC team:
- Answer questions about the cluster, Slurm, storage, and software
- Share tips and scripts that didn’t make it into any tutorial
- Announce workshops, maintenance windows, and new hardware
- Collect feedback about what to improve
Join here: #gsb-yen-users
If the link does not open automatically, open the Slack app, search for #gsb-yen-users in Channels, and join from there.
For questions that need a direct answer from the team, or anything you’d rather not post in a channel:
Response time is typically one business day.
What to Do When You’re Stuck
| Situation | Where to go |
|---|---|
| “My Slurm job keeps failing” | #gsb-yen-users — someone has seen it |
| “Is this dataset ok to send to an LLM?” | Email DARC or ask your IRB coordinator |
| “I want to run something much bigger” | Email DARC — we can advise on allocations |
| “Is there a workshop on X?” | Watch #gsb-yen-users for announcements |
| “My code works on my laptop but not the Yens” | #gsb-yen-users — include your error output |
Keep Exploring
Everything you ran this week is in your fork. Future projects can start from the same patterns:
- More data, same pipeline: swap the input list in your Slurm array script
- Different model: change
base_urlandmodel— the rest is identical - New dataset type: adapt your Pydantic schema, rerun the pipeline
- Need a GPU: add
--gres=gpu:1(and the GPU partition) to your Slurm script — see How to Run LLMs on the Yens
The leaderboard stays up, and any optional exercises you didn’t finish are still there.