Imitation Learning Datasets: A Toolkit For Creating Datasets, Training Agents and Benchmarking

TL;DR

This paper introduces a comprehensive toolkit for creating, sharing, and benchmarking imitation learning datasets, addressing data scarcity and evaluation inconsistency issues in the field.

Contribution

It provides a toolkit with curated expert policies, ready-to-use datasets, and shared implementations to streamline dataset creation and evaluation in imitation learning.

Findings

Facilitates faster dataset creation with multithreaded expert policy support.

Provides standardized datasets and measurement tools for consistent evaluation.

Includes shared implementations of common imitation learning techniques.

Abstract

Imitation learning field requires expert data to train agents in a task. Most often, this learning approach suffers from the absence of available data, which results in techniques being tested on its dataset. Creating datasets is a cumbersome process requiring researchers to train expert agents from scratch, record their interactions and test each benchmark method with newly created data. Moreover, creating new datasets for each new technique results in a lack of consistency in the evaluation process since each dataset can drastically vary in state and action distribution. In response, this work aims to address these issues by creating Imitation Learning Datasets, a toolkit that allows for: (i) curated expert policies with multithreaded support for faster dataset creation; (ii) readily available datasets and techniques with precise measurements; and (iii) sharing implementations of common imitation learning techniques. Demonstration link: https://nathangavenski.github.io/#/il-datasets-video