Dev blog opened!
The Autonomous Learning Library (ALL) is an object-oriented reinforcement learning library for Pytorch. I’ve been working on it for the past 8 months or so as part of my research as a graduate student at UMass Amherst. While the project remains immature, I believe the ALL exceeds existing libraries in certain design considerations.
As work on the ALL has progressed, I have run many small tests and experiments. The results to date are disorganized and of little interest to anyone. However, as the project begins to mature, I believe records of these experiments are of increasing value. While the results are not “publishable” in the traditional sense, they shed light on many small questions of practical interest. I have therefore opened this blog in order to record my notes.
