Methods › General › Distributed Methods › IMPALA
IMPALA
Introduced by Lasse Espeholt et al. in IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
archive 2025-07-28 Description, source and code snippet are the archive's method entry.
IMPALA, or the Importance Weighted Actor Learner Architecture, is an off-policy actor-critic framework that decouples acting from learning and learns from experience trajectories using V-trace. Unlike the popular A3C-based agents, in which workers communicate gradients with respect to the parameters of the policy to a central parameter server, IMPALA actors communicate trajectories of experience (sequences of states, actions, and rewards) to a centralized learner. Since the learner in IMPALA has access to full trajectories of experience we use a GPU to perform updates on mini-batches of trajectories while aggressively parallelising all time independent operations.
This type of decoupled architecture can achieve very high throughput. However, because the policy used to generate a trajectory can lag behind the policy on the learner by several updates at the time of gradient calculation, learning becomes off-policy. The V-trace off-policy actor-critic algorithm is used to correct for this harmful discrepancy.
In Syntology repository known, no verified implementation (declared_code; repositories: 5 listed of 26 in the graph, strongest link: official). This answer is about arXiv:1802.01561, the paper the archive credits with this method.
Origin paper of this answer, arXiv:1802.01561 (its page on this site · its page on Syntology): Syntology's run record for its harvested code, 16 ran (of which 6 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 1 violated, 8 with no contract checked; 6 where Syntology's instrument failed) · 18 unverified (of 34 harvested samples) · 3 pointer-only (licence). The instrument figure counts failures of Syntology's instrument, not of the code. Ran means executed on a synthesized input, not that the code is correct.
Syntology's have check for this name (free, it answers coverage only), read 2026-10-05; ask it again live. A level is the served verification level of that one answer; nothing here is summed across methods.
Papers archive 2025-07-28
archive papers tagged: 16 · with a code link: 10 · where Syntology ran a sample: 4 (4 with a run with no instrument failure, 0 where every run was a failure of Syntology's instrument) Syntology
Show: all tagged papersonly where code ran (4 of 16 tagged: 4 with a run with no instrument failure, 0 where every run was a failure of Syntology's instrument)
Page 1 of 1: papers 1 to 16 of 16, newest first by the archive's date (ties by slug), in archive order.
Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code, as “N ran (of which C constructed an object rather than computing a result; K with no instrument failure: H honoured, V violated, P with no contract checked; I where Syntology's instrument failed) · U unverified”; the instrument figure counts failures of Syntology's instrument, not of the code. It is per sample and not a correctness claim. When the archive marks a repository official for the paper, the line starts with that repository's state (the archive's flag, not a verdict on who wrote the code; “community repositories only” when every sample that ran came from a community repository, “official: no sample here; runs from other or unrecorded repositories” when some came from a repository the paper names or has in its text, or from none recorded); hover it for the repositories the samples that ran came from.
-
World Model Agents with Change-Based Intrinsic Motivation 26 Mar 2025 · 1 repository · arXiv:2503.21047
-
Cleanba: A Reproducible and Efficient Distributed Reinforcement Learning Platform 29 Sep 2023 · 1 repository · arXiv:2310.00036Syntology official (archive's flag): 4 ran · 4 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 0 where Syntology's instrument failed) · 4 unverified (of 8 harvested samples) · 8 pointer-only (licence)
-
Exploring the Promise and Limits of Real-Time Recurrent Learning 30 May 2023 · 1 repository · arXiv:2305.19044
-
DoMo-AC: Doubly Multi-step Off-policy Actor-Critic Algorithm 29 May 2023 · 0 repositories · arXiv:2305.18501
-
Sharing Lifelong Reinforcement Learning Knowledge via Modulating Masks 18 May 2023 · 2 repositories · arXiv:2305.10997
-
Lifelong Reinforcement Learning with Modulating Masks 21 Dec 2022 · 4 repositories · arXiv:2212.11110
-
EnvPool: A Highly Parallel Reinforcement Learning Environment Execution Engine 21 Jun 2022 · 3 repositories · arXiv:2206.10558Syntology official (archive's flag): 3 ran · 5 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified (of 5 harvested samples) · 1 pointer-only (licence)
-
Semantic Exploration from Language Abstractions and Pretrained Representations 8 Apr 2022 · 0 repositories · arXiv:2204.05080
-
A Distributed Deep Reinforcement Learning Technique for Application Placement in Edge and Fog Computing Environments 24 Oct 2021 · 0 repositories · arXiv:2110.12415
-
MACRPO: Multi-Agent Cooperative Recurrent Policy Optimization 2 Sep 2021 · 1 repository · arXiv:2109.00882
-
PFPN: Continuous Control of Physically Simulated Characters using Particle Filtering Policy Network 16 Mar 2020 · 1 repository · arXiv:2003.06959
-
A Self-Tuning Actor-Critic Algorithm 28 Feb 2020 · 0 repositories · arXiv:2002.12928
-
IMPACT: Importance Weighted Asynchronous Architectures with Clipped Target Networks 30 Nov 2019 · 0 repositories · arXiv:1912.00167
-
TorchBeast: A PyTorch Platform for Distributed RL 8 Oct 2019 · 3 repositories · arXiv:1910.03552Syntology official (archive's flag): 3 ran · 6 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 4 where Syntology's instrument failed) · 3 unverified (of 9 harvested samples) · 3 pointer-only (licence)
-
Towards Combining On-Off-Policy Methods for Real-World Applications 24 Apr 2019 · 0 repositories · arXiv:1904.10642
-
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures 5 Feb 2018 · 24 repositories · arXiv:1802.01561Syntology community repositories only · 16 ran (of which 6 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 1 violated, 8 with no contract checked; 6 where Syntology's instrument failed) · 18 unverified (of 34 harvested samples) · 3 pointer-only (licence)
Tasks archive 2025-07-28
15 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.
Usage over time archive 2025-07-28
Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).
Categories archive 2025-07-28
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections