Embark on a transformative generative AI journey with InstructLab

As artificial intelligence continues to be top of mind for enterprises, kick starting the generative AI journey has become of fundamental importance thanks to cutting-edge innovations, such as ChatGPT.

To get the ball rolling and open the doors to gen AI, Red Hat Inc. rolled out a cost-effective, open-source and community-based project called InstructLab that enhances the alignment of large language models, according to Máirín Duffy (pictured, left), software engineering manager at Red Hat. 

“The InstructLab project, basically, we’re using a typical upstream open-source process to enable anybody in the community to contribute knowledge and skills to get built into the model,” Duffy said. “We have the Granite-based models that you can take with InstructLab and they’re Apache-licensed, so when you’re playing with them and you’re building on top of them, we basically take on a regular cadence the contributions from the community, and then we make an InstructLab-tuned model with the community contributions.”

Duffy and Joe Fernandes (right), vice president and general manager of the gen AI foundation model platforms at Red Hat, spoke with theCUBE Research’s Rob Strechay and Paul Gillin at Red Hat Summit, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed why the generative AI journey is important in the present enterprise world and how Red Hat enhances it through InstructLab and Red Hat Enterprise Linux AI. (* Disclosure below.)

Red Hat’s Máirín Duffy and Joe Fernandes talk with theCUBE about the importance of the generative AI journey.

How RHEL AI accelerates the generative AI journey

As gen AI continues to be a stepping stone toward groundbreaking change, RHEL AI was devised to propel innovation in a trustworthy and transparent manner. As a result, RHEL AI uses an open-source approach and this helps eliminate entry barriers in the generative AI journey, according to Fernandes. 

“RHEL AI is our new foundation model platform for bringing these technologies together … the open Granite models, the InstructLab model alignment tools,” he noted. “It’s going to run on Linux. We package it as a bootable RHEL image using this new RHEL image mode technology that was also announced here at [Red Hat] Summit, we integrate acceleration for popular hardware, because Linux needs to run on Nvidia hardware or AMD hardware or Intel hardware.”

As the generative AI journey gains steam, InstructLab is accelerating this narrative, having been initiated by Red Hat and IBM Research. As a result, the RedHat-IBM partnership helps in the democratization of gen AI by deploying the prowess of an open-source community, according to Fernandes.

“This really highlights our partnership with IBM Research,” he stated. “IBM announced the Granite models a year ago at IBM Think, they’re part of the watsonx portfolio, but now, with this new open model strategy, they’ve begun open sourcing some of those models, so the Granite language model that we’re using in the InstructLab community, the lab methodology that drives the InstructLab project, also came out of IBM Research.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE Research’s coverage of Red Hat Summit:

(* Disclosure: Red Hat Inc. sponsored this segment of theCUBE. Neither Red Hat  nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE