Dopamine: A Research Framework for Deep Reinforcement Learning

Code

• github.com/google/dopamine
  OfficialIn papertf★ 10,856
• github.com/floringogianu/atari-agents
  pytorch★ 99
• github.com/atavakol/action-hypergraph-networks
  tf★ 23
• github.com/google-research/google-research/tree/master/bisimulation_aaai2020
  tf★ 0
• github.com/ACampero/dopamine
  tf★ 0
• github.com/zhixuan-lin/ensemble-rl-discrete
  jax★ 0
• github.com/V0LsTeR/dopamine_prioritized_buffer
  tf★ 0
• github.com/AlexChaloner/RL-project
  tf★ 0
• github.com/GoingMyWay/dopamine_reward_decomposition
  tf★ 0
• github.com/V0LsTeR/DQN_heap
  tf★ 0

Abstract

Dopamine is a research framework for fast prototyping of reinforcement learning algorithms.

Tasks

Reproductions