← Back to OpenAI updates ← Terug naar OpenAI-updates
OpenAI ARTICLE ARTIKEL 8 March 2018 8 maart 2018

On first-order meta-learning algorithms On first-order meta-learning algorithms

Read paper(opens in a new window) Read paper(opens in a new window)

Article details Artikelgegevens
AI maker AI-maker OpenAI Type Type Article Artikel Published Gepubliceerd 8 March 2018 8 maart 2018 Updates Updates Videos Video's View original article Bekijk origineel artikel

Abstract

This paper considers meta-learning problems, where there is a distribution of tasks, and we would like to obtain an agent that performs well (i.e., learns quickly) when presented with a previously unseen task sampled from this distribution. We analyze a family of algorithms for learning a parameter initialization that can be fine-tuned quickly on a new task, using only first-order derivatives for the meta-learning updates. This family includes and generalizes first-order MAML, an approximation to MAML obtained by ignoring second-order derivatives. It also includes Reptile, a new algorithm that we introduce here, which works by repeatedly sampling a task, training on it, and moving the initialization towards the trained weights on that task. We expand on the results from Finn et al. showing that first-order meta-learning algorithms perform well on some well-established benchmarks for few-shot classification, and we provide theoretical analysis aimed at understanding why these algorithms work.

This paper considers meta-learning problems, where there is a distribution of tasks, and we would like to obtain an agent that performs well (i.e., learns quickly) when presented with a previously unseen task sampled from this distribution. We analyze a family of algorithms for learning a parameter initialization that can be fine-tuned quickly on a new task, using only first-order derivatives for the meta-learning updates. This family includes and generalizes first-order MAML, an approximation to MAML obtained by ignoring second-order derivatives. It also includes Reptile, a new algorithm that we introduce here, which works by repeatedly sampling a task, training on it, and moving the initialization towards the trained weights on that task. We expand on the results from Finn et al. showing that first-order meta-learning algorithms perform well on some well-established benchmarks for few-shot classification, and we provide theoretical analysis aimed at understanding why these algorithms work.

Authors

Alex Nichol, Joshua Achiam, John Schulman

Related articles

View all

Scaling laws for reward model overoptimization Publication Oct 19, 2022

Learning to play Minecraft with Video PreTraining Conclusion Jun 23, 2022

Dota 2 with large scale deep reinforcement learning Publication Dec 13, 2019

Dota 2 with large scale deep reinforcement learning Publication Dec 13, 2019

More from OpenAI Meer van OpenAI

All updates Alle updates

Gemini komt eraan