Member Corner
The GECF held the first workshop on ‘Artificial Intelligence in the Oil and Gas Industry’, with the participation of distinguished Professor Mérouane Debbah, who delivered a presentation titled “The Superpower of Open Source Large Language Models.”
Professor Debbah, currently the Chief Researcher in Technology Innovation Institute Abu Dhabi, United Arab Emirates, introduced large language models (LLM) as a type of machine learning model that uses natural language as input and output. He noted that the reach of language can refer to text, images, videos, and other forms of communication.
The presentation shed light on three important parameters in training large language models: the size of the model (number of neurons), the quality of the training data, and the training duration, emphasising the significance of these parameters in achieving good performance. Prof Debbah expanded the horizon of applications, including in the creation of Chabot, content creation (writing articles, generating code), text summarisation, language translation, data analysis, highlighting the potential impact of LLMs in different industries, including oil and gas.
Professor Debbah additionally promoted the recent development of the LLM "Falcon", a 40-billion-parameter model, trained in more than one trillion tokens designed to be released as open-source.
He invited proposals for use cases and applications of large language models and maintained that exceptional ideas will receive dedicated computing power as an investment.
Please find Prof Debbah’s presentation below.