Data Visualization and Analytics Training Series: Jupyter Workflow at OLCF
The OLCF is proud to offer the first event of the Data Visualization and Analytics training series: Jupyter Workflow at OLCF.
Jupyter notebooks offer a unique environment for users to analyze and process scientific data. The interactive nature of the web interface allows one to debug, visualize, and share code instantly, without having to handle everything on the command-line. Because Jupyter notebooks can contain both computer code and rich text elements (paragraphs, equations, figures, widgets, links), this allows you to create human-readable documents containing analytics, results, and descriptions.
OLCF offers a JupyterHub, which allows users to run CPU and GPU JupyterLab sessions with access to the Summit and Andes filesystems. The OLCF Jupyter environments also offer standard analysis and Machine Learning libraries (PyTorch, TensorFlow, Pandas, NumPy) pre-installed.
The training event will provide an overview of how to access Jupyter at OLCF and how to utilize the interface in an analytics setting. This includes hands-on examples of how to analyze data generated on OLCF systems, along with a Deep Learning workflow demo run with Jupyter.
Due to hardware limitations and the nature of accessing the OLCF JupyterHub, only 30 OLCF users attached to Moderate Security Enclave projects (able to access systems like Andes and Summit) can participate in the hands-on portion of the training, but others are welcome to sign-up and attend the virtual presentations. Although participation is limited during the event, the hands-on examples will be available after the event for Moderate users to access and work with on their own.
Warning: The datasets listed in the slides and recording have been moved from “/gpfs/alpine/world-shared/stf218/analysis_viz_training/07142022” to “/gpfs/alpine/world-shared/stf007/msandov1/jupyter_training”
- Video: Recording
- Slides: Overview of OLCF JupyterHub
- Slides: Dataset Workflows on Jupyter
- Slides: Deep Learning Workflow with Jupyter
- Training Series Website: https://github.com/olcf/dva-training-series
See agenda below:
|1:00 pm – 1:30 pm EDT||Overview of OLCF JupyterHub||Ryan Prout|
|1:30 pm – 2:00 pm EDT||Dataset Workflows on Jupyter||Benjamin Hernandez|
|2:00 pm – 2:15 pm EDT||Hands-on Examples||Benjamin Hernandez|
|2:15 pm – 2:45 pm EDT||Deep Learning Workflow with Jupyter||Junqi Yin|
|2:45 pm – 3:00 pm EDT||Extra Hands-on Examples||Benjamin Hernandez|
Registration has closed.
Joining information will be sent to you in a calendar event before the event.