NetHack is known as the frontier of reinforcement learning research where learning-based methods still need to catch up to rule-based solutions.

In this work, we conduct a comprehensive study of NetHack and examine various learning mechanisms to enhance the performance of neural models.

Neural policy learning methods have achieved remarkable results in various control problems, ranging from Atari games to simulated locomotion.

Progress in Reinforcement Learning (RL) algorithms goes hand-in-hand with the development of challenging environments that test the limits of current methods. While existing RL environments are either sufficiently complex or based on fast simulation, they are rarely both. Here, we present the NetHack Learning Environment (NLE), a scalable, procedurally generated, stochastic, rich, and challenging environment for RL research based on the popular single-player terminal-based roguelike game, NetHack. We argue that NetHack is sufficiently complex to drive long-term research on problems such as exploration, planning, skill acquisition, and language-conditioned RL, while dramatically reducing the computational resources required to gather a large amount of experience. We compare NLE and its task suite to existing alternatives, and discuss why it is an ideal medium for testing the robustness and systematic generalization of RL agents. We demonstrate empirical success for early stages of the game using a distributed Deep RL baseline and Random Network Distillation exploration, alongside qualitative analysis of various agents trained in the environment.

Data is provided under the NetHack General Public License - A GPL style license that is used to covered the NetHack Game since 1989. Torch.IterableDataset interface, and handle metadata. The buffer's contents are then the Terminal Escape Sequences are printed in dark grey. The metadata consists of: 1. gameid (int) - A unique id for the game, created by the local database. The 13 roles are: Arc - Archaelogist Bar - Barbarian Cav - Cave(wo)man Hea - Healer Kni - Knight Mon - Monk Pri - Priest(ess) Ran - Ranger Rog - Rogue Sam - Samurai Tou - Tourist Val - V alkyrie Wiz - Wizard 14. race (str) - The Race of the player.

In addition, current offline RL benchmarks test agents on the exact same environment where the data was collected.

We support different padding strategies and alphabet sizes, but by default we choose an alphabet size of 96, where the last character is used for padding.

We respond to their concerns in detail below. We hope this addresses your key concern. We thank you for your supportive comments, and are somewhat surprised by the low score in light of them. As per your suggestion, we will release our full research code reproducing the paper's results as an additional supplement We would like to thank you for your thorough and supportive review, and the suggestions contained therein. We will update the paper to include a screenshot of Medusa's island in place of Figure 1, and update the link to the referenced video to include the exact time we refer to.