#73 - LLMs, RAG, Graph Neural Networks and Open Source with Maxime Labonne
If your into llms, open source, GNN, Gen AI topics this episode is a MUST watch. I had an amazing time with Maxime Labonne , thanks again for sharing ??
??? Who is Maxime Labonne?
Maxime Labonne is a Staff Machine Learning Scientist at Liquid AI . Maxime received his PhD from the Polytechnic Institute of Paris and has been working with machine learning since 2019. He has applied his expertise in various contexts, including R&D, industry, finance, and academia. He is also an AI/ML Google Developer Expert. Maxime is widely known for creating popular LLMs on Hugging Face, like AlphaMonarch-7B, Beyonder-4x7B, Phixtral, and NeuralBeagle14. He also released LLM tools, such as LLM AutoEval, LazyMergekit, LazyAxolotl, and AutoGGUF. Maxime is the author of the technical book "Hands-On Graph Neural Networks using Python," published with Packt. He has written technical articles on his blog and Towards Data Science and created the popular LLM course on GitHub, which has over 27,000 stars.
?? In this episode...
... we discuss Maxime's expertise in large language models (LLMs) and graph neural networks, as well as his journey from cybersecurity to machine learning. We also talk about insights on LLMS misconceptions, evaluation optimization techniques, and creating augmented datasets for fine-tuning. Maxime also talked about graph neural networks and non-transformer architectures and provided advice for leading language model research and mentoring language model research beginners. Finally, we discuss Maxime's passion for open source and knowledge sharing.
Most valuable lessons
1. Maxime have a very impressive career :)
2. It's possible to switch careers (from cybersecurity to AI in that case)
3. Open source tools and communities are amazing places for growth
4. Continuous learning and upskilling are essential for career growth in a fast-evolving field like AI
领英推荐
5. Following one's interests is key for career satisfaction and progression
6. Past experiences can be leveraged to excel in a different field
7. To learn AI, start small, use tutorials, share your work, consider filling any tooling gaps, and enjoy the learning process
?? Listen to this episode now!
??? Podcast ?? https://smartlink.ausha.co/let-s-talk-ai/
?? Youtube ?? https://www.youtube.com/@lets-talk-ai
Keep learning, keep creating, keep building, and let's have a positive impact!
Warm regards,
Thomas
AI & ML Innovator | Transforming Data into Revenue | Expert in Building Scalable ML Solutions | Ex-Microsoft
8 个月It sounds like a fascinating discussion! Maxime's journey from cybersecurity to AI is inspiringshows how diverse backgrounds can thrive in tech. Open source contributions and communities indeed foster incredible learning and collaboration. Switching careers into AI is challenging but rewarding, highlighting the importance of continuous learning. His insights on misconceptions about LLMs and optimizing evaluations are valuable for anyone exploring AI. It's encouraging to see experts like Maxime sharing knowledge and guiding newcomers. Overall, it reinforces the idea that passion and persistence drive career success in dynamic fields like AI.
LinkedIn 4X Top Voice | AI Speaker, Mentor & Trainer | Top 1%@Topmate.io | AI Developer & Researcher | GDGOnCampus CoOrganizer | Ex-Android Co Lead@ GDSC | ABC, WTM, GDG, IIC, GCI | Freelancer
8 个月Great work, Thomas Bustos! The course structure is well curated and the content is clear and on point. One of the best #AI podcasts, I have come across.
?? AI Geek | Book Author | Speaker | Mentor ??
8 个月Excited to listen to that one !
Co-Founder @Lyah | "Let's Talk AI" Podcast Host | Databricks Champion
8 个月??? Podcast ?? https://smartlink.ausha.co/let-s-talk-ai/ ?? Youtube ?? https://www.youtube.com/@lets-talk-ai