Modeling Workbench - Machine Learning and Simulation Platform

Overview

The Modeling Workbench is one of the main components of Science Central™.

Primary Purpose

“Build, train, and deploy machine learning and simulation models”

Platform Description

“Platform provides a centralized environment for data scientists to build, train, and deploy machine learning models”

Tagline

“Explore your ideas”

Architecture

The Modeling Workbench consists of three main components:

  1. Data Science Environment

  2. TensorFlow Environment

  3. Tahoma OnDemand Computing Infrastructure

Component 1: Data Science Environment

Overview

Tagline: “Drive Insights With Multi-language Tools”

Tooltip Description: “A JupyterHub environment fully equipped with a suite of libraries designed for powerful data processing and analysis, supporting Python, R, and Julia.”

Directly Observed Features

JupyterLab Interface

When accessed, the Data Science environment opens a JupyterLab interface with:

  • File browser on the left sidebar

  • Main workspace with notebook tabs

  • Menu bar with File, Edit, View, Run, Kernel, Tabs, Settings, Help

  • Toolbar with common actions (save, add cell, cut, copy, paste, run, restart)

  • Status bar showing kernel information

Interface Elements

  • File Browser: Shows /home/jovyan directory

  • Notebook Environment: Interactive notebook with Python 3 (ipykernel)

  • Cell Types: Code, Markdown, Raw options available

  • Kernel Status: “Python 3 (ipykernel) | Connecting”

  • Tabs: “File Browser”, “Running Terminals and Kernels”, “Table of Contents”, “Extension Manager”

Observed Functionality

  • Code cells can be executed

  • File upload capability through file browser

  • New file creation options (notebook, folder, etc.)

  • Terminal access available

  • Save and checkpoint functionality

Component 2: TensorFlow Environment

Overview

Tagline: “Advance Your Deep Learning Journey”

Tooltip Description: “A JupyterHub environment designed with all the necessary libraries for conducting deep learning tasks using TensorFlow.”

Information

When clicking on the TensorFlow option, it opened a new browser tab but the specific interface details were not fully explored during the session.

Component 3: Tahoma OnDemand

Overview

Tagline: “Compute Infrastructure”

Tooltip Description: “Workspace to manage files, jobs, run GUI applications, and connect to an interactive shell.”

Directly Observed Features

Open OnDemand Dashboard

When accessed, Tahoma OnDemand opens to an Open OnDemand interface with:

  • URL: https://tahoma-ondemand.emsl.pnnl.gov/pun/sys/dashboard

  • Version: OnDemand version 3.1.14

  • Navigation Menu: Apps, Files, Jobs, Clusters, Interactive Apps, My Interactive Sessions, Help

Available Applications

The dashboard shows “Pinned Apps” with three main applications:

  1. Cluster Desktop

    • Description: “System Installed App”

    • Provides desktop environment access

  2. Jupyter Notebook/Lab

    • Description: “System Installed App”

    • Jupyter logo displayed

    • Alternative notebook access

  3. RStudio Server

    • Description: “System Installed App”

    • RStudio logo displayed

    • R statistical computing environment

Interface Description

  • Platform Description: “OnDemand provides an integrated, single access point for all of your HPC resources”

  • Footer: “Powered by Open OnDemand” with version information

Getting Started

Access Instructions

  1. Navigate to Science Central main page (https://sc.emsl.pnnl.gov/)

  2. Select “Modeling Workbench” from the main modules

  3. Choose from the three available environments:

    • Click “Data Science” for the JupyterHub environment

    • Click “TensorFlow” for deep learning environment

    • Click “Tahoma OnDemand” for HPC access

Environment Selection

Based on the tooltip descriptions:

  • Data Science: For multi-language data processing and analysis (Python, R, Julia)

  • TensorFlow: For deep learning tasks using TensorFlow

  • Tahoma OnDemand: For file management, job submission, GUI applications, and shell access

Integration with Science Central

The Modeling Workbench is one of the main components accessible through the Science Central navigation, alongside:

  • All Modules (main dashboard)

  • Proposal Management

  • LIMS

  • Data Portal

  • Insight Engine

Support

For technical support or questions about the Modeling Workbench, contact sc.support@pnnl.gov