TextWorldExpress: Simulating Text Games at One Million Steps Per Second
This work addresses a bottleneck for researchers in AI and language understanding by significantly accelerating simulation throughput for text game benchmarks.
The authors tackled the speed limitation of text-based games, which capped at 300 steps per second, by developing TextWorldExpress, a high-performance simulator that achieves over one million steps per second, enabling billion-step experiments in about a day.
Text-based games offer a challenging test bed to evaluate virtual agents at language understanding, multi-step problem-solving, and common-sense reasoning. However, speed is a major limitation of current text-based games, capping at 300 steps per second, mainly due to the use of legacy tooling. In this work we present TextWorldExpress, a high-performance simulator that includes implementations of three common text game benchmarks that increases simulation throughput by approximately three orders of magnitude, reaching over one million steps per second on common desktop hardware. This significantly reduces experiment runtime, enabling billion-step-scale experiments in about one day.