DeepMind

Repositories

• open_spiel

  OpenSpiel is a collection of environments and algorithms for research in general reinforcement learning and search/planning in games.

  python swift games reinforcement-learning cpp multiagent
  C++ Apache-2.0 481 2,422 19 15 Updated Dec 3, 2020

• jax
  0 1 0 0 Updated Dec 3, 2020

• deepmind-research

  This repository contains implementations and illustrative code to accompany DeepMind publications
  Jupyter Notebook Apache-2.0 755 3,964 23 18 Updated Dec 3, 2020

• reverb

  Reverb is an efficient and easy-to-use data storage and transport system designed for machine learning research
  C++ Apache-2.0 24 288 8 (1 issue needs help) 1 Updated Dec 3, 2020

• chex
  Python Apache-2.0 5 130 0 0 Updated Dec 3, 2020

• kapitan

  Generic templated configuration management for Kubernetes, Terraform and other things

  python docker kubernetes yaml devops aws-kms jinja2
  Python 135 1,296 54 (5 issues need help) 22 Updated Dec 3, 2020

• surface-distance

  Library to compute surface distance based performance metrics for segmentation tasks.
  Python Apache-2.0 50 187 1 3 Updated Dec 3, 2020

• acme

  A library of reinforcement learning components and agents

  research reinforcement-learning agents
  Python Apache-2.0 171 1,787 33 3 Updated Dec 2, 2020

• dm-haiku

  JAX-based neural network library

  machine-learning deep-neural-networks deep-learning neural-networks jax
  Python Apache-2.0 56 812 14 3 Updated Dec 2, 2020

• tensor_annotations
  Python Apache-2.0 0 3 0 0 Updated Dec 2, 2020

• jaxline
  Python Apache-2.0 0 9 0 0 Updated Dec 1, 2020

• dm_control

  DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.

  machine-learning reinforcement-learning deep-learning artificial-intelligence neural-networks physics-simulation mujoco
  Python Apache-2.0 372 2,106 15 1 Updated Nov 30, 2020

• sonnet

  TensorFlow-based neural network library

  machine-learning deep-learning tensorflow artificial-intelligence neural-networks
  Python Apache-2.0 1,250 8,633 14 3 Updated Nov 30, 2020

• rlax
  Python Apache-2.0 33 491 2 1 Updated Nov 27, 2020

• bsuite

  bsuite is a collection of carefully-designed experiments that investigate core capabilities of a reinforcement learning (RL) agent
  Python Apache-2.0 145 1,104 3 6 Updated Nov 26, 2020

• ferminet

  An implementation of the Fermionic Neural Network for ab-initio electronic structure calculations
  Python Apache-2.0 36 323 0 0 Updated Nov 24, 2020

• jraph
  Python Apache-2.0 3 178 0 0 Updated Nov 23, 2020

• dm_hard_eight
  Python Apache-2.0 1 66 1 0 Updated Nov 19, 2020

• lab

  A customisable 3D platform for agent-based AI research

  machine-learning deep-learning artificial-intelligence neural-networks
  C GPL-2.0 1,289 6,332 32 2 Updated Nov 18, 2020

• lab2d

  A customisable 2D platform for agent-based AI research

  machine-learning deep-learning artificial-intelligence
  C++ Apache-2.0 26 284 1 0 Updated Nov 18, 2020

• optax

  Optax is a gradient processing and optimization library for JAX.
  Python Apache-2.0 11 193 6 (2 issues need help) 2 Updated Nov 17, 2020

• educational
  Jupyter Notebook Apache-2.0 4 21 0 0 Updated Nov 17, 2020

• mc_gradients
  Jupyter Notebook Apache-2.0 22 92 0 1 Updated Nov 13, 2020

• trfl

  TensorFlow Reinforcement Learning
  Python Apache-2.0 372 3,046 4 2 Updated Nov 6, 2020

• jax_verify

  Neural network verification in JAX
  Python Apache-2.0 4 45 0 1 Updated Nov 5, 2020

• dm_env_rpc

  A networking protocol for agent-environment communication
  Python Apache-2.0 5 39 0 0 Updated Nov 2, 2020

• dm_env

  A Python interface for reinforcement learning environments

  python api machine-learning reinforcement-learning deep-learning interface
  Python Apache-2.0 22 106 0 0 Updated Oct 30, 2020

• hanabi-learning-environment

  hanabi_learning_environment is a research platform for Hanabi experiments.
  Python Apache-2.0 105 512 11 4 Updated Oct 27, 2020

• dqn_zoo

  DQN Zoo is a collection of reference implementations of reinforcement learning agents developed at DeepMind based on the Deep Q-Network (DQN) agent.
  Python Apache-2.0 17 160 0 0 Updated Oct 21, 2020

• ai-safety-gridworlds

  This is a suite of reinforcement learning environments illustrating various safety properties of intelligent agents.
  Python Apache-2.0 98 473 0 0 Updated Oct 13, 2020