Over the past five years, the Center for Research Computing has been instrumental in the software development, data collection and management, and data analysis aspects of the AEGIS program funded through UNITAID. The University of Notre Dame and Unitaid have announced that an innovative vector control tool for malaria called a spatial repellent showed a significant impact on reducing malaria infections and is highlighted as a featured project in an ND news release here.

NSWC Crane looks to expand its AI efforts through annual RAITE event

Naval Surface Warfare Center, Crane Division (NSWC Crane) hosted the second annual Robust Artificial Intelligence Test Event (RAITE) at Muscatatuck Training Center on October 21-25. This event allowed members from academia, industry, and government to collaborate on an unclassified level and test AI efforts in real-world scenarios on a cyber-physical range. Read more.

In the previous monthly newsletter, we announced that the CRC is running a locally hosted instance of the Open WebUI platform for interacting with open weight large language models (LLMs). Over 45 users have already signed up on the platform! :tada:. With the latest OpenWebUI version (v0.5.4), you will have access to new models, such as the latest Llama 3.3 70B. The Llama 3.3 70B is designed to support multiple languages and offers enhanced capabilities for fine-tune, coding, reasoning, tool use and general knowledge tasks. Since last month, we have also enabled Open WebUI Pipelines, offering a range of options for retrieval augmented generation (RAG), model building and agent-like actions using function calling pipelines. The CRC welcomes interesting research use cases that could leverage these features. All CRC Users are welcome to sign up at https://openwebui.crc.nd.edu/ by using your Notre Dame Google Login

Version 0.5 of Open WebUI introduces several updates. The latest version, released on January 5, 2025, includes the ability to clone shared chats, making collaboration more efficient. Various translation improvements have been made to provide a smoother global user experience. Bug fixes include better handling of image-only messages and improved exception feedback. These updates aim to enhance user experience and streamline workflows for all CRC hosted Open WebUI AI platform users. 

Managing Software Dependencies in Conda Environments

Managing software dependencies effectively is crucial for ensuring reproducibility and avoiding conflicts, especially when working with complex scientific applications. Conda environments provide a powerful way to handle these challenges. Here's how you can manage software dependencies in Conda environments to keep your workflows smooth and efficient.

1) Create a New Conda Environment

To start working with a specific set of dependencies, create a new Conda environment. This keeps your project isolated and prevents interference with other environments or system-wide packages.

conda create --name myenv python=3.10

Here, myenv is the name of your environment, and python=3.10 specifies the Python version. You can also add other packages in the creation command.

2) Activate and Deactivate Environments

Once an environment is created, activate it using:

conda activate myenv

This will switch to the environment and ensure that any subsequent package installs or executions happen within this environment. When done, deactivate it:

conda deactivate

3) Install Packages and Dependencies

To install packages in the active environment, use:

conda install <package_name>

For example, to install numpy:

conda install numpy

Conda will automatically handle the dependencies for you, ensuring that all required packages are installed at compatible versions.

4) Specifying Package Versions

If you need a specific version of a package, you can specify it like so:

conda install numpy=1.21

This ensures that you are working with the exact version required for your project, avoiding issues where newer or older versions could break your code.

5) Managing Package Conflicts

Sometimes, Conda might encounter conflicts between installed packages. To resolve these, you can: Use the --update-deps flag when installing a new package to automatically attempt to resolve dependencies:

conda install <package_name> --update-deps

If conflicts persist, try installing packages individually to narrow down which one causes the issue.

6) List Installed Packages

To view the list of installed packages in your environment:

conda list

This will display all the packages installed in the current environment along with their versions.

7) Export and Share Environments

To share your environment with others or recreate it on another machine, you can export your environment's configuration to a YAML file:

conda env export > environment.yml

This file will contain a list of all installed packages and their versions. To create an environment from this YAML file:

conda env create -f environment.yml

This ensures that the same set of dependencies and versions is installed, making your environment reproducible.

8) Updating Dependencies

To keep your environment up to date, run:

conda update --all

This will attempt to update all packages in your environment to their latest compatible versions.

9) Removing Packages

To remove a package from your environment:

conda remove <package_name>

10)  Cleaning Up Unused Packages and Caches

To free up space, you can clean up unused packages and caches from your Conda installation:

conda clean --all

This will remove unnecessary files, caches, and old package versions that are no longer in use.By following these practices, you'll be able to manage software dependencies effectively in Conda environments, ensuring smooth and reproducible scientific computing workflows.

Chemical & Biomolecular Engineering
529,990 CPU hours

Chemical & Biomolecular Engineering
423,945 CPU hours

Civil & Environmental Engineering & Earth Sciences
325,044 CPU hours

Chemical & Biomolecular Engineering
256,788 CPU hours

Chemical & Biomolecular Engineering
218,436 CPU hours

Aerospace & Mechanical Engineering
201,740 CPU hours

Chemical & Biomolecular Engineering
198,107 CPU hours

Civil & Environmental Engineering & Earth Sciences
182,678 CPU hours

Biological Sciences
165,641 CPU hours

Chemical & Biomolecular Engineering
113,503 CPU hours

Chemical & Biomolecular Engineering
69,623 GPU hours

Chemical & Biomolecular Engineering
62,021 GPU hours

Chemical & Biomolecular Engineering
19,765 GPU hours

Computer Science & Engineering
10,889 GPU hours

Computer Science & Engineering
6,805 GPU hours

Applied and Computational Mathematics and Statistics
3,969 GPU hours

Computer Science & Engineering
3,247 GPU hours

Chemistry & Biochemistry
2,729 GPU hours

Chemistry & Biochemistry
2,550 GPU hours

Computer Science & Engineering
2,302 GPU hours

Every Wednesday and Thursday
2:00 – 3:00 p.m.
812 Flanner Hall (map)

The CRC offers multiple training opportunities for both new and existing users. We periodically provide short courses and other learning opportunities, which are advertised on our website and through email lists. In-person office hours are held every Wednesday and Thursday from 2:00-3:30 p.m. in Flanner Hall, room 812, on a first-come, first-served basis. You can also arrange a Zoom meeting at your convenience by emailing CRCsupport@nd.edu with your availability. We recommend bringing a laptop to in-person sessions.

For more information, please contact CRCsupport@nd.edu.