Tuning GPT-3 on a Single GPU via Zero-Shot Hyperparameter Transfer
Discover how to tune GPT-3 hyperparameters efficiently on a single GPU using µTransfer, a technique based on the maximal update parametrization that enables optimization of large language models with minimal resources.
Tensor Programs - Tuning Large Neural Networks via Zero-Shot Hyperparameter Transfer
Explore μTransfer: a technique for tuning large neural networks via zero-shot hyperparameter transfer, making optimal hyperparameters stable with respect to width scaling.
DALL-E - Zero-Shot Text-to-Image Generation - Paper Explained
Explore OpenAI's DALL-E: a groundbreaking AI model that generates images from text descriptions, combining concepts in plausible ways and showcasing emergent image-to-image capabilities.
Language Models as Zero-Shot Planners - Extracting Actionable Knowledge for Embodied Agents
Exploring how large language models can translate high-level tasks into actionable steps for virtual environments, with techniques to improve executability and potential real-world applications.
Fast Zero Shot Object Detection with OpenAI CLIP
Learn zero-shot object detection and localization using OpenAI CLIP, a multi-modal deep learning model. Implement practical techniques for efficient, domain-flexible computer vision tasks without fine-tuning.