A discussion of AI researcher Rich Sutton's critique of LLMs and his vision for AI inspired by animal learning, contrasting with current approaches.
Introducing TabICL, a state-of-the-art table foundation model that uses in-context learning and improved architecture for fast, scalable tabular data prediction.
Explores recent research on instruction finetuning for LLMs, including cost-effective data generation methods and an overview of new models like Gemma 2.
Introduces CARTE, a foundation model for tabular data, explaining its architecture, pretraining on knowledge graphs, and results.
A 1-hour video presentation covering the full development cycle of Large Language Models, from architecture and pretraining to finetuning and evaluation.
A tutorial on pre-training a BERT model from scratch using Hugging Face Transformers and Habana Gaudi accelerators on AWS.
