Source repo: ciml-summer-institute-2024 | Branch:
main| Last synced: 2026-04-24 10:27:17.425 UTC
Preparing for the CIML Summer Institute
SESSION: 0_preparation
Instructions:
You will get the most out of the CIML Summer Institute if you are prepared prior to the event. By brushing up on your knowledge of Linux and installing all necessary software on your laptop before the event, you’ll be able to focus your attention on the skills and topics that are most relevant to machine learning.
This section contains a set of detailed start-up instructions, beginning with setting up your account and connecting to Expanse, to configuring your laptop for running visualization software. Please read the documents and exercises carefully, and complete all necessary steps before event. Feel free to ask questions or if you have any problems with the start-up tasks
Contents
- Expanse User Guide
- HPC Systems Accounts
- Large Language Model (LLM) Accounts
- Basic Skills
- Computer Requirements
- Github, Slack and Zoom
Expanse User Guide
Please read the Expanse user guide and familiarize yourself with the hardware, file systems, batch job submission, compilers and modules. The guide can be found here:
- Expanse User Guide
- Expanse Landing Page: https://expanse.sdsc.edu
if you have any difficulties getting set up, please contact Institute staff at consult@sdsc.edu.
HPC Systems Accounts:
You will need a training account to access SDSC’s Expanse HPC system for the CIML Institute. We will be creating this account for you.
An email will be sent from consult@sdsc.edu with the subject line "CIML 2024 Summer Institute: Account Set-Up" that will include your training account username and password and instructions. We will send this email in time for "Prep Day," where we will be covering how to manage your training account, along with other useful information.
As a reminder, we are asking all attendees to use the training accounts that we are providing, even if you already have your own account. This will minimize problems related to custom configurations (e.g. environment variables, choice of shell, etc.) that may cause some of the hands-on examples to break.
This training account may be used:
- Via SSH directly to login.expanse.sdsc.edu
- Via a special page on the Expanse Portal: https://portal.expanse.sdsc.edu/training
-
- For this institute, please do not use portal.expanse.sdsc.edu.
Large Language Model Accounts:
Each participant will need to sign-up for a free account for the following, which will be used during the LLM sessions on day 3:
- ChatGPT: https://chat.openai.com/auth/login
- Gemini: https://gemini.google.com/
Basic Skills
There are several basic skills needed in order to access and run jobs on HPC systems. There are two GitHub repositories with several mini-tutorials that you can access:
- Basic HPC Linux Skills: https://github.com/sdsc-hpc-training-org/basic_skills
- HPC Security: https://github.com/sdsc-hpc-training-org/hpc-security
- Connecting to Expanse: https://github.com/sdsc-hpc-training-org/hpc-security/blob/master/connecting-to-hpc-systems/connect-to-expanse.md
- Basic_Linux_Skills on Expanse: https://github.com/sdsc-hpc-training-org/basic_skills/tree/master/basic_linux_skills_expanse
- Using Interactive Compute Nodes on Expanse
- How to Run Notebooks on Expanse: We use a secure notebook launching tool called
Galyleo:- Tutorial: https://github.com/mkandes/galyleo
- Video on Running Jupyter Notebooks on Expanse
Computer Requirements
- You will need to bring your own laptop for the summer institute. This will be used both for running software locally and connecting to SDSC’s supercomputers.
- WiFi will be available throughout the SDSC building
- Remember your charger. There will be extension cords throughout the room for charging devices
GitHub, Slack, and Zoom
Github:
- Training material will be located on the CIML Summer Institute GitHub repo.
- One of the hands-on sessions will require a GitHub account.
- If you do not already have one, you can create a free personal GitHub account here.
- For basic GitHub usage on SDSC systems, see here: https://github.com/sdsc-hpc-training-org/basic_skills/tree/master/using_github
Slack:
- We will also be using Slack as our main platform for announcements, where participants can communicate and ask for help. Download (Windows, MacOS, or Linux) and get started. Zoom chat will be disabled during the preparation day. Make sure to turn on your Slack notifications to receive alerts.
- See "Preparation Information | CIML Summer Institute 2024" email received for the link to join the Slack workspace for this institute.
Zoom:
- We will be using Zoom for preparation day on Tuesday, June 18, 2024, from 9am-11am (Pacific Time).
- You will need to install the latest Zoom client, which is available for Windows, MacOS and Linux. Once installed, you can test your microphone and camera interface with Zoom here. You can find more information on Zoom system requirements, including bandwidth requirements here.
Connection details are sent as a calendar invite to all CIML participants. You should have received an invite for the preparation day on Tuesday, June 18, 2024, from 9am - 11am (Pacific Time). If you did not receive this an invite, please contact cwong@sdsc.edu.
While Prep Day is a virtual event, the main CIML Summer Institute sessions (Tuesday, June 25 - Thursday, June 27) will be held in person and will not available remotely.