Task Augmentation Strategies
Daniel discusses the efficiency of training on multiple tasks simultaneously, while Tim introduces the concept of transductive active fine tuning. Jan emphasizes the importance of data augmentation, particularly using symmetry techniques to enhance training despite limited examples. The conversation reveals innovative methods to manipulate data, aiming to improve model performance under constraints.In this clip
From this podcast
Machine Learning Street Talk (MLST)
Daniel Franzen & Jan Disselhoff - ARC Prize 2024 winners
Related Questions
How are Large Language Models (LLMs) fine-tuned post-training?
How are Large Language Models (LLMs) fine-tuned post-training in the episode Llama 3: Scaling open LLMs to AGI and the clip Fine Tuning Insights?
I have a question about the episode Navigating Machine Learning Careers: Insights from Meta to Consulting // Ilya Reznik // #286 and the clip Fine Tuning Insights. Can you explain the differences between fine-tuning and training models: LLMs vs custom model applications and when to use each?