Artificial intelligence
LG unveils Korea’s 1st open source AI to take on Meta, Google
LG says Exaone 3.0 performs better in some benchmark areas such as coding than Meta’s Llama 3.1, Google’s Gemma2
By Aug 07, 2024 (Gmt+09:00)
3
Min read
Most Read
LG Chem to sell water filter business to Glenwood PE for $692 million


KT&G eyes overseas M&A after rejecting activist fund's offer


Kyobo Life poised to buy Japan’s SBI Group-owned savings bank


StockX in merger talks with Naver’s online reseller Kream


Meritz backs half of ex-manager’s $210 mn hedge fund



LG Group, South Korea’s fourth-biggest conglomerate, on Wednesday introduced the country’s first open-sourced artificial intelligence large language model (AI LLM) to compete with the technology of Big Tech players like Meta Platforms Inc. and Google.
LG AI Research, an institute of the group’s holding company LG Corp., unveiled Exaone 3.0, an improved version of its predecessor showcased in July last year. The group revealed the first version of Exaone in December 2021.
“Its performance is as strong as other open-sourced AI models such as Meta’s Llama and Google’s Gemma,” said an LG group official. “We succeeded in improving performance and reducing its size to three-hundredths that of its initial model by focusing on lightweight, optimization technology.”
Exaone 3.0 has 7.8 billion parameters for learning on nearly 8 trillion tokens. By comparison, Google’s Gemma 2 has 9 billion parameters on 8 trillion tokens.
LG’s AI model ranked No. 1 in 13 benchmark areas, including real-world use cases, coding and math, demonstrating its global competitiveness compared with other global open-source AI of a similar size such as Llama 3.1 and Gemma 2, according to the Korean group.
OPEN SOURCE AI
LG Group decided to open-source the lightweight version of Exaone 3.0 to dominate the AI ecosystem by allowing industries to actively use the generative AI, industry sources in Seoul said.
Global Big Tech firms such as Microsoft Corp., Meta and Google rushed to develop lightweight AI models, which gained popularity thanks to better performance despite having fewer parameters than other LLMs.
“LG does not need to depend on Nvidia’s AI semiconductor as it succeeded in developing a lightweight model,” said an electronics industry source in Seoul.
“The more AI engineers and students use the open-sourced Exaone 3.0, the wider LG’s AI-related businesses will be.”
GLOBAL PARTNERSHIPS
The group plans to roll out the latest AI LLM to products and services of its affiliates such as world-leading home appliance maker LG Electronics Inc. and the conglomerate’s mobile carrier LG Uplus Corp.
Those units are expected to optimize Exaone 3.0 to their products and services to speed up innovation.
LG Group is currently in talks on forging global partnerships, sources said.
“We aim to vitalize the AI research ecosystem and improve national AI competitiveness by unveiling the country’s first open-sourced homegrown AI model,” said LG AI Research President Bae Kyunghoon.
CHATEXAONE
The group also unveiled ChatExaone, an enterprise AI agent based on the latest AI LLM and launched an open beta service for affiliates’ staff.
ChatExaone provides functions that enhance work productivity, including real-time web-based Q&A, document and image-based Q&A, coding and database management, LG Group said.
It will allow group employees to leverage AI in their jobs such as for searching, summarizing, translating, analyzing data, coding and more, the conglomerate added.
ChatExaone adopts retrieval-augmented generation (RAG) technology, which utilizes real-time web search results to understand the context of the user's instructions and provide answers that reflect the latest information, LG Group said.
Write to Chae-Yeon Kim at why29@hankyung.com
Jongwoo Cheon edited this article.
More to Read
-
Artificial intelligenceLG unveils upgraded hyperscale AI for future businesses
Jul 19, 2023 (Gmt+09:00)
2 Min read -
Artificial intelligenceLG AI Research: Foundation of group’s long-term future
Feb 27, 2022 (Gmt+09:00)
3 Min read -
Artificial intelligenceLG unveils multimodal ‘supergiant AI’ with unprecedented parameters
Dec 14, 2021 (Gmt+09:00)
2 Min read
Comment 0
LOG IN