Skip to content

Latest commit

 

History

History
15 lines (11 loc) · 553 Bytes

README.md

File metadata and controls

15 lines (11 loc) · 553 Bytes

Model Agnostic Meta Learning

Reproduce of MAML paper in toy example regression task (Sinusoid) using TensorFlow 2.1.

The loss curve comparison between a normal deep learning method and MAML:

An inference example of random new task trained with only 10 steps:

How to use:

  • Build Docker image from Dockerfile
  • Run python main.py to perform vanilla and MAML experiments