Using and specializing large language models like ChatGPT
Discover our immersive training on chatbots based on large language models (LLM), intended for enthusiasts or evolving professionals wishing to understand and master this revolutionary technology. You will learn the basics of artificial intelligence, the principles of transformative models, and you will dive into practical cases to develop and specialize your own chatbots. In addition, we will address the essential ethical issues related to their use. This training is ideal for anyone interested in the future of AI in professional or non-professional fields.
Chatbots based on large language models (LLM) built using transformative neural networks such as ChatGPT are playing an increasing role in society, thanks to their good ability to generate text, images, sounds and human-like videos. They are/can/will be used in various areas such as: customer service, personal assistants, education, healthcare, accessibility, content creation, entertainment and much more.
However, while LLM-based chatbots offer many benefits, there are also potential risks and challenges. These include the risk of generating inappropriate or harmful content, difficulty understanding context, difficulty providing accurate information, and concerns about data privacy and security. Ethical considerations are crucial as these technologies continue to develop and become increasingly integrated into society.
The main parts of the training are:
- Introduction to AI and Deep Learning
- Large language models (LLM) and self-attention networks (Transformers)
- Understanding the ethical implications of AI
- Practical session on the use of LLMs
- Approaches for the creation of ChatBot based on generative AI
- Specialization of generative AI tools for a specific domain
Content
First Part
Introduction to AI and Deep Learning
Large language models (LLM) and self-attention networks (Transformers)
○ Understanding LLMs and Transformers
○ Applications and use cases
○ Challenges and limitations
Understanding the ethical implications of AI
○ Discuss potential risks and challenges (privacy, security, inappropriate content)
○ Importance of ethical considerations in AI development
Practical session on the use of LLMs
○ Practical experimentation with various LLMs
○ Practical exploration of different techniques and application areas
Second Part
Approaches for creating ChatBot programs
○ Introduction to current techniques
○ Building a Basic ChatBot using existing Transformers
Specialization of Transformers models
○ Understand how to specialize Transformers models
○ Focus on specific tasks and areas
○ Processing specific document types (text, PDF, audio, videos)
Practical session on model specialization
○ Hands-on practice on adapting and optimizing Transformers models
○ Evaluation of model efficiency and accuracy
○ Filtering Transformers Model Results
Learning Outcomes
At the end of this training, participants will be able to:
Understand the artificial intelligence landscape, with a focus on the role and importance of deep learning.
Understand what large language models (LLMs) built using self-attention networks are and understand their capabilities and applications.
Use various LLMs in different application areas through hands-on experience and hands-on exploration.
Understand the social and ethical dimensions of ChatBots, including potential risks and challenges such as inappropriate content generation, privacy concerns, and data security concerns.
Effectively use current technologies to create ChatBot programs that utilize self-attention networks.
Specialize ChatBots for specific tasks or domains, focusing particularly on processing specific documents in text or PDF format.
Develop an understanding of how to adapt and optimize Transformer models to ensure their efficiency and accuracy.
Critically evaluate the results generated by Transform models and make informed decisions on their usefulness for specific tasks or applications.
Training Method
Alternating conceptual presentation and practice sessions.
Organised By
Digital Learning Hub Luxembourg
Certification
Participation OnlyPrerequisites
The practical exercises are designed to be suitable for people who:
either have no knowledge of Python
either have basic or more advanced knowledge of Python
Depending on the level in Python, different practical exercises will be offered.
Planning and location
09:00 - 17:00
09:00 - 17:00
ESCO Skills
Your trainer(s) for this course

Nicolas GUELFI
