# Model-Agnostic Meta-Learning (MAML) - MAML attempts to answer the question: How to find an initialization for the meta-learner that is not only useful for adapting to various problems, but also can be adapted quickly (in a small number of steps) and efficiently (using only a few examples)? -  - MAML optimizes for a set of parameters such that when a gradient step is taken with respect to a particular task i, the parameters are close the optimal parameters for task i. - Doesn't make any assumptions on the form of the model. - No additional parameters introduced for meta-learning, and uses [[Stochastic Gradient Descent]]. ## Advantages of MAML - Substantially outperform a number of existing approaches on popular few-shot image classification benchmarks, Omniglot and MiniImageNet, including existing approaches that were much more complex or domain specific. - When MAML combined with [[Policy Gradient]] methods for [[Reinforcement Learning]]. MAML discovered a policy which let a simulated robot adapt its locomotion direction and speed in a single gradient update. ## First-order MAML - MAML is trained by backpropagating the loss through the within-episode gradient descent procedure. This normally requires computing second-order gradients, which can be expensive to obtain (both in terms of time and memory). For this reason, an approximation is often used whereby gradients of the within-episode descent steps are ignored. This approximation is called first-order MAML. ## ProtoMAML - Combines the complementary strengths of [[Prototypical Networks]] and MAML. - By allowing gradients to flow through the Prototypical Network-equivalent linear layer initialization, it significantly helps the optimization of this model and outperforms vanilla fo-MAML by a large margin. ## MAML++ https://arxiv.org/abs/1810.09502 ## REPTILE https://github.com/dragen1860/Reptile-Pytorch ## LEOPARD https://arxiv.org/pdf/1911.03863.pdf ## iMAML https://arxiv.org/pdf/1909.04630.pdf --- ## References 1. Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks. C. Finn, P. Abbeel, S. Levine. In ICML, 2017. ([pdf](https://arxiv.org/pdf/1703.03400.pdf), [code](https://github.com/cbfinn/maml)) 2. Learning to Learn, Chelsea Finnm Jul 2017 https://bair.berkeley.edu/blog/2017/07/18/learning-to-learn/ 3. Notes on iMAML https://www.inference.vc/notes-on-imaml-meta-learning-without-differentiating-through/ 4. How to train your MAML https://arxiv.org/abs/1810.09502 5. [Meta-Dataset: A Dataset of Datasets for Learning to Learn from Few Examples](https://openreview.net/forum?id=rkgAGAVKPr), Triantafillou et al. ICLR2020 6. MAML high level overview https://www.youtube.com/watch?v=ItPEBdD6VMk