List of all items

Structs

• buffer::replay::ContinuousReplayBatch
• buffer::replay::ContinuousReplayBuffer
• buffer::replay::ContinuousReplayBurnTensors
• buffer::replay::PrioritizedBatch
• buffer::replay::PrioritizedBurnTensors
• buffer::replay::PrioritizedReplayBuffer
• buffer::replay::ReplayBatch
• buffer::replay::ReplayBuffer
• buffer::replay::ReplayBurnTensors
• buffer::replay::SumTree
• buffer::rollout::Minibatch
• buffer::rollout::MinibatchIterator
• buffer::rollout::RecurrentMinibatchIterator
• buffer::rollout::RecurrentRolloutBatch
• buffer::rollout::RecurrentRolloutBuffer
• buffer::rollout::RolloutBatch
• buffer::rollout::RolloutBuffer
• buffer::rollout::RolloutBurnTensors
• env::SpaceInfo
• env::StepInfo
• env::StepResult
• env::games::cartpole::CartPole
• env::games::cartpole::CartPoleState
• env::games::continuous_lqr::ContinuousLqr
• env::games::continuous_lqr::ContinuousLqrState
• env::games::flickering_cartpole::FlickeringCartPole
• env::games::flickering_cartpole::FlickeringCartPoleState
• env::games::grid_world::GridWorld
• env::games::grid_world::GridWorldState
• env::games::masked_cartpole::MaskedCartPole
• env::games::matching_pennies::MatchingPennies
• env::games::mountain_car_continuous::MountainCarContinuous
• env::games::mountain_car_continuous::MountainCarState
• env::games::n_player_matching_pennies::NPlayerMatchingPennies
• env::games::pendulum::PendulumState
• env::games::pendulum::PendulumSwingUp
• env::games::pong::Pong
• env::games::pong::PongState
• env::games::signaling::SignalingGame
• env::games::simple_bandit::SimpleBandit
• env::games::simple_bandit::SimpleBanditState
• env::games::snake::Food
• env::games::snake::GameState
• env::games::snake::Position
• env::games::snake::Snake
• env::games::snake::SnakeEnv
• env::games::t_maze::TMaze
• env::games::t_maze::TMazeState
• env::pool::EnvPool
• env::pool::PoolStepResult
• inference::ExportedModel
• inference::LayerWeights
• inference::snake::SnakeCNNInference
• multi_agent::comms::AgentMessage
• multi_agent::environment::MultiAgentResult
• multi_agent::joint::JointMultiAgentTrainer
• multi_agent::joint::JointRollout
• multi_agent::joint::JointStats
• multi_agent::joint::JointStepResult
• multi_agent::joint::JointTrainerConfig
• multi_agent::messages::Experience
• multi_agent::messages::PolicyUpdate
• multi_agent::messages::TrainingStats
• multi_agent::nfsp::NfspConfig
• multi_agent::nfsp::NfspIterationStats
• multi_agent::nfsp::NfspStats
• multi_agent::nfsp::NfspTrainer
• multi_agent::nfsp::ReservoirBuffer
• multi_agent::psro::AlphaRankMetaSolver
• multi_agent::psro::FictitiousPlayMetaSolver
• multi_agent::psro::PayoffCache
• multi_agent::psro::PsroConfig
• multi_agent::psro::PsroIterationStats
• multi_agent::psro::PsroStats
• multi_agent::psro::PsroTrainer
• multi_agent::psro::ReplicatorDynamicsMetaSolver
• multi_agent::psro::UniformMetaSolver
• policy::atari_cnn::NatureDqnBurnPolicy
• policy::atari_cnn::NatureDqnBurnPolicyRecord
• policy::atari_cnn::NatureDqnBurnPolicyRecordItem
• policy::atari_cnn::NatureDqnConfig
• policy::atari_cnn::NatureDqnQNetwork
• policy::atari_cnn::NatureDqnQNetworkRecord
• policy::atari_cnn::NatureDqnQNetworkRecordItem
• policy::continuous_q::ContinuousQNetwork
• policy::continuous_q::ContinuousQNetworkConfig
• policy::continuous_q::ContinuousQNetworkRecord
• policy::continuous_q::ContinuousQNetworkRecordItem
• policy::inference::InferenceModel
• policy::inference::SnakeCNNInference
• policy::inference::TrainingMetadata
• policy::lstm::LstmBurnConfig
• policy::lstm::LstmBurnPolicy
• policy::lstm::LstmBurnPolicyRecord
• policy::lstm::LstmBurnPolicyRecordItem
• policy::mlp::MlpBurnConfig
• policy::mlp::MlpBurnPolicy
• policy::mlp::MlpBurnPolicyRecord
• policy::mlp::MlpBurnPolicyRecordItem
• policy::multi_discrete_mlp::MultiDiscreteMlpBurnPolicy
• policy::multi_discrete_mlp::MultiDiscreteMlpBurnPolicyRecord
• policy::multi_discrete_mlp::MultiDiscreteMlpBurnPolicyRecordItem
• policy::q_network::QNetworkBurn
• policy::q_network::QNetworkBurnConfig
• policy::q_network::QNetworkBurnRecord
• policy::q_network::QNetworkBurnRecordItem
• policy::sac_actor::SacActor
• policy::sac_actor::SacActorConfig
• policy::sac_actor::SacActorRecord
• policy::sac_actor::SacActorRecordItem
• policy::snake_cnn::SnakeCnnBurnPolicy
• policy::snake_cnn::SnakeCnnBurnPolicyRecord
• policy::snake_cnn::SnakeCnnBurnPolicyRecordItem
• policy::universal_inference::ModelMetadata
• policy::universal_inference::OutputSpec
• policy::universal_inference::UniversalModel
• train::a2c::config::A2cConfig
• train::a2c::trainer::A2cStats
• train::a2c::trainer::A2cTrainer
• train::bc::config::BcConfig
• train::bc::dataset::Demonstrations
• train::bc::trainer::BcEpochStats
• train::bc::trainer::BcTrainer
• train::dqn::config::DQNConfig
• train::dqn::trainer::DQNStepStatsBurn
• train::dqn::trainer::DQNTrainerBurn
• train::optimizer::BurnOptimizer
• train::ppo::actor_learner::ActorChannels
• train::ppo::actor_learner::ActorHandle
• train::ppo::actor_learner::ActorStats
• train::ppo::actor_learner::ActorThrottle
• train::ppo::actor_learner::AsyncActorLearnerConfig
• train::ppo::actor_learner::LearnerReport
• train::ppo::config::PPOConfig
• train::ppo::recurrent_trainer::RecurrentPPOTrainer
• train::ppo::stats::AggregatedStats
• train::ppo::stats::TrainingStats
• train::ppo::trainer::PPOTrainerBurn
• train::sac::LogAlpha
• train::sac::SacConfig
• train::sac::SacStepStats
• train::sac::SacTrainer
• utils::normalize::RunningMeanStd

Enums

• env::SpaceType
• env::games::snake::Cell
• env::games::snake::Direction
• env::games::t_maze::Cue
• multi_agent::comms::Delivery
• multi_agent::messages::ControlMessage
• multi_agent::messages::PolicyBroadcast
• policy::inference::InferenceActivation
• policy::mlp::BurnActivation
• policy::universal_inference::Activation
• policy::universal_inference::InputSpec
• policy::universal_inference::Layer
• policy::universal_inference::Tensor

Traits

• env::Environment
• multi_agent::comms::CommunicatingEnvironment
• multi_agent::environment::MultiAgentEnvironment
• multi_agent::joint::JointEnv
• multi_agent::joint::JointPolicy
• multi_agent::psro::MetaSolver
• train::optimizer::BackendOptimizer

Functions

• buffer::replay::sample
• buffer::replay::sample_continuous
• buffer::rollout::compute_advantages
• buffer::rollout::compute_advantages_multi_agent
• buffer::rollout::compute_advantages_partial
• buffer::rollout::compute_mc_returns
• buffer::rollout::compute_nstep_returns
• buffer::rollout::compute_vtrace_advantages
• buffer::rollout::generate_minibatch_indices
• buffer::rollout::normalize_advantages
• buffer::rollout::sample_minibatch
• buffer::rollout::shuffle_indices
• buffer::rollout::train_val_split
• env::games::pong::mirror_observation
• env::games::snake::make_snake_env
• multi_agent::comms::place_message
• multi_agent::comms::split_action
• policy::seeded_init::seeded_kaiming_uniform
• policy::seeded_init::seeded_orthogonal
• train::a2c::loss::compute_a2c_policy_loss
• train::a2c::loss::compute_a2c_value_loss
• train::bc::loss::compute_bc_loss
• train::dqn::loss::compute_dqn_loss
• train::dqn::loss::compute_dqn_loss_double
• train::dqn::loss::compute_loss
• train::dqn::loss::compute_td_target
• train::dqn::loss::compute_td_target_double
• train::dqn::loss::gather_action_q
• train::dqn::loss::huber_per_sample
• train::grad_clip::clip_grads_by_global_norm
• train::grad_clip::global_grad_norm
• train::optimizer::wrap_burn
• train::ppo::actor_learner::actor_thread
• train::ppo::actor_learner::learner_loop
• train::ppo::actor_learner::load_policy_from_broadcast
• train::ppo::actor_learner::serialize_policy
• train::ppo::actor_learner::spawn_actor
• train::ppo::loss::compute_entropy_loss
• train::ppo::loss::compute_policy_loss
• train::ppo::loss::compute_value_loss
• train::ppo::loss::generate_minibatch_indices
• train::ppo::loss::generate_minibatch_indices_with_rng
• train::ppo::loss::scalar_f64
• utils::cuda::default_burn_device

Type Aliases

• env::games::snake::SnakeEnvSingle
• multi_agent::messages::AgentId
• train::optimizer::OptimResult

Constants

• VERSION
• env::games::flickering_cartpole::DEFAULT_BURST_LEN
• env::games::flickering_cartpole::DEFAULT_FLICKER_PROBABILITY
• env::games::grid_world::DEFAULT_MAX_STEPS
• env::games::grid_world::DEFAULT_SEED
• env::games::grid_world::GOAL_REWARD
• env::games::grid_world::GRID_COLS
• env::games::grid_world::GRID_ROWS
• env::games::grid_world::HOLE_REWARD
• env::games::grid_world::NUM_ACTIONS
• env::games::grid_world::STEP_PENALTY
• env::games::signaling::LISTENER
• env::games::signaling::SPEAKER
• env::games::t_maze::DEFAULT_CORRIDOR_LENGTH
• env::games::t_maze::DEFAULT_SEED
• env::games::t_maze::NUM_ACTIONS
• env::games::t_maze::OBS_DIM
• env::games::t_maze::REWARD_CORRECT
• env::games::t_maze::REWARD_WRONG
• policy::sac_actor::LOG_STD_MAX
• policy::sac_actor::LOG_STD_MIN