Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks

03/09/2017
by   Chelsea Finn, et al.
0

We propose an algorithm for meta-learning that is model-agnostic, in the sense that it is compatible with any model trained with gradient descent and applicable to a variety of different learning problems, including classification, regression, and reinforcement learning. The goal of meta-learning is to train a model on a variety of learning tasks, such that it can solve new learning tasks using only a small number of training samples. In our approach, the parameters of the model are explicitly trained such that a small number of gradient steps with a small amount of training data from a new task will produce good generalization performance on that task. In effect, our method trains the model to be easy to fine-tune. We demonstrate that this approach leads to state-of-the-art performance on two few-shot image classification benchmarks, produces good results on few-shot regression, and accelerates fine-tuning for policy gradient reinforcement learning with neural network policies.

READ FULL TEXT

page 6

page 8

05/17/2019

Alpha MAML: Adaptive Model-Agnostic Meta-Learning

Model-agnostic meta-learning (MAML) is a meta-learning technique to trai...
10/31/2019

Hierarchical Expert Networks for Meta-Learning

The goal of meta-learning is to train a model on a variety of learning t...
03/15/2021

Robust MAML: Prioritization task buffer with adaptive learning process for model-agnostic meta-learning

Model agnostic meta-learning (MAML) is a popular state-of-the-art meta-l...
02/09/2019

Meta-Curvature

We propose to learn curvature information for better generalization and ...
10/27/2020

Why Does MAML Outperform ERM? An Optimization Perspective

Model-Agnostic Meta-Learning (MAML) has demonstrated widespread success ...
10/30/2019

Multimodal Model-Agnostic Meta-Learning via Task-Aware Modulation

Model-agnostic meta-learners aim to acquire meta-learned parameters from...
12/10/2020

Performance-Weighed Policy Sampling for Meta-Reinforcement Learning

This paper discusses an Enhanced Model-Agnostic Meta-Learning (E-MAML) a...

Code Repositories

pytorch-maml

PyTorch implementation of MAML: https://arxiv.org/abs/1703.03400


view repo

Model-Agnostic-Meta-Learning

None


view repo