Unleashing Potential: Harnessing Foundation Models (Large Language Models) in Business and Research
Leveraging Large Language Models (LLMs) has marked a significant milestone in recent months, notably with the introduction of ChatGPT in early 2023. These models have demonstrated remarkable potential in addressing straightforward queries and tasks. However, to fully exploit their capabilities in handling routine inquiries, adept prompt engineering is essential. Furthermore, the adaptability of LLMs to novel tasks and domains is pivotal. It is crucial to recognize that each company or research field possesses unique requirements, necessitating tailored adaptations of LLMs. The specificity of these needs often hinges on domain-specific data, demanding meticulous consideration. How can these models be tailored to classify your data effectively? What strategies can be employed when dealing with a limited dataset? Complex scenarios, such as querying vast repositories of textual files stored in directories, underscore the challenges. These files encompass diverse modalities, formats, and structures, ranging from structured to entirely unstructured content.
In this research presentation, we will delve into an exploration of LLM capabilities and pinpoint the areas where they encounter limitations. Subsequently, we will elucidate various techniques for fine-tuning these models, especially in scenarios where data availability is constrained. By addressing these challenges, we aim to provide valuable insights into harnessing the full potential of LLMs, ensuring their optimal performance in diverse and data-intensive applications.
Date and Time
Location
Hosts
Registration
- Date: 24 Oct 2023
- Time: 05:00 PM to 06:00 PM
- All times are (GMT-08:00) US/Pacific
-
Add Event to Calendar
- UBC Okanagan
- Kelowna
- Kelowna, British Columbia
- Canada V1V 1V7
- Building: EME
- Room Number: EME 112
- Starts 11 October 2023 07:21 PM
- Ends 24 October 2023 05:00 PM
- All times are (GMT-08:00) US/Pacific
- No Admission Charge
Speakers
Fatemeh Hendijani Fard, PhD
Creating Accurate Predictions in the Stock Market
Biography:
-
Dr. Fatemeh Hendijani Fard is an Assistant Professor at The University of British Columbia, Okanagan, Canada, where she leads the Data Science and Software Engineering lab. Her research interests lie at the intersection of Natural Language Processing and Software Engineering, with a focus on code representation learning and transfer learning for low-resource languages, as well as mining software repositories. She collaborates closely with industry partners and has contributed her expertise as a program committee member and reviewer for esteemed journals and conferences, including IEEE Transactions on Software Engineering, ACM International Conference on the Foundations of Software Engineering, and IEEE/ACM International Conference on Automated Software Engineering. Dr. Fard is an IEEE Senior member, and she actively gives back to the community by mentoring females interested in Artificial Intelligence.
Email: