mtrl
latest
Getting Started
MTRL
Algorithms
Supported Algorithms
Environments
Supported Environments
Tutorials
Overview
Running Code
Baselines
API:
mtrl package
mtrl
»
Index
Edit on GitHub
Index
A
|
B
|
C
|
D
|
E
|
F
|
G
|
H
|
I
|
L
|
M
|
N
|
O
|
P
|
Q
|
R
|
S
|
T
|
U
|
V
|
W
|
Z
A
act() (mtrl.agent.sac.Agent method)
action (mtrl.replay_buffer.ReplayBufferSample attribute)
Actor (class in mtrl.agent.components.actor)
add() (mtrl.replay_buffer.ReplayBuffer method)
Agent (class in mtrl.agent.abstract)
(class in mtrl.agent.deepmdp)
(class in mtrl.agent.distral)
(class in mtrl.agent.grad_manipulation)
(class in mtrl.agent.gradnorm)
(class in mtrl.agent.hipbmdp)
(class in mtrl.agent.pcgrad)
(class in mtrl.agent.sac)
(class in mtrl.agent.sac_ae)
(class in mtrl.agent.wrapper)
apply_vector_grad_to_parameters() (in module mtrl.agent.pcgrad)
AttentionBasedExperts (class in mtrl.agent.components.moe_layer)
AverageMeter (class in mtrl.logger)
B
BaseActor (class in mtrl.agent.components.actor)
buffer_index (mtrl.replay_buffer.ReplayBufferSample attribute)
build_dmcontrol_vec_env() (in module mtrl.env.builder)
build_envs() (mtrl.experiment.experiment.Experiment method)
(mtrl.experiment.metaworld.Experiment method)
(mtrl.experiment.multitask.Experiment method)
build_metaworld_vec_env() (in module mtrl.env.builder)
build_mlp() (in module mtrl.agent.utils)
build_mlp_as_module_list() (in module mtrl.agent.utils)
build_model() (mtrl.agent.components.critic.QFunction method)
C
check_if_should_use_multi_head_policy() (in module mtrl.agent.components.actor)
check_if_should_use_task_encoder() (in module mtrl.agent.components.actor)
Checkpointable (class in mtrl.utils.checkpointable)
chunks() (in module mtrl.utils.utils)
clear() (in module mtrl.experiment.utils)
close_envs() (mtrl.experiment.experiment.Experiment method)
ClusterOfExperts (class in mtrl.agent.components.moe_layer)
collect_trajectory() (mtrl.experiment.metaworld.Experiment method)
(mtrl.experiment.multitask.Experiment method)
complete_init() (mtrl.agent.abstract.Agent method)
(mtrl.agent.distral.Agent method)
(mtrl.agent.distral.DistilledAgent method)
(mtrl.agent.sac.Agent method)
(mtrl.agent.wrapper.Agent method)
Component (class in mtrl.agent.components.base)
compute_grad (mtrl.agent.ds.task_info.TaskInfo attribute)
copy_conv_weights_from() (mtrl.agent.components.encoder.Encoder method)
(mtrl.agent.components.encoder.FeedForwardEncoder method)
(mtrl.agent.components.encoder.MixtureofExpertsEncoder method)
(mtrl.agent.components.encoder.PixelEncoder method)
create_env_id_to_index_map() (mtrl.experiment.metaworld.Experiment method)
create_eval_modes_to_env_ids() (mtrl.experiment.metaworld.Experiment method)
(mtrl.experiment.multitask.Experiment method)
create_multitask_obs() (mtrl.env.vec_env.MetaWorldVecEnv method)
Critic (class in mtrl.agent.components.critic)
CurrentMeter (class in mtrl.logger)
D
delete_from_filesystem() (mtrl.replay_buffer.ReplayBuffer method)
DeterministicTransitionModel (class in mtrl.agent.components.transition_model)
dict_to_config() (in module mtrl.utils.config)
DistilledAgent (class in mtrl.agent.distral)
dump() (mtrl.logger.Logger method)
(mtrl.logger.MetersGroup method)
E
EMBEDDING (mtrl.agent.components.hipbmdp_theta.ThetaSamplingStrategy attribute)
encode() (mtrl.agent.components.actor.Actor method)
(mtrl.agent.components.actor.BaseActor method)
(mtrl.agent.components.critic.Critic method)
Encoder (class in mtrl.agent.components.encoder)
encoding (mtrl.agent.ds.task_info.TaskInfo attribute)
EnsembleOfExperts (class in mtrl.agent.components.moe_layer)
env_index (mtrl.agent.ds.task_info.TaskInfo attribute)
(mtrl.agent.grad_manipulation.EnvMetadata attribute)
env_index_count (mtrl.agent.grad_manipulation.EnvMetadata attribute)
env_obs (mtrl.agent.ds.mt_obs.MTObs attribute)
(mtrl.replay_buffer.ReplayBufferSample attribute)
EnvMetadata (class in mtrl.agent.grad_manipulation)
eval_mode (class in mtrl.agent.utils)
evaluate_vec_env_of_tasks() (mtrl.experiment.dmcontrol.Experiment method)
(mtrl.experiment.metaworld.Experiment method)
Experiment (class in mtrl.experiment.dmcontrol)
(class in mtrl.experiment.experiment)
(class in mtrl.experiment.metaworld)
(class in mtrl.experiment.multitask)
extra_repr() (mtrl.agent.components.moe_layer.Linear method)
F
FeedForward (class in mtrl.agent.components.moe_layer)
FeedForwardEncoder (class in mtrl.agent.components.encoder)
FiLM (class in mtrl.agent.components.encoder)
flatten_list() (in module mtrl.utils.utils)
forward() (mtrl.agent.components.actor.Actor method)
(mtrl.agent.components.actor.BaseActor method)
(mtrl.agent.components.critic.Critic method)
(mtrl.agent.components.critic.QFunction method)
(mtrl.agent.components.decoder.PixelDecoder method)
(mtrl.agent.components.encoder.Encoder method)
(mtrl.agent.components.encoder.FeedForwardEncoder method)
(mtrl.agent.components.encoder.FiLM method)
(mtrl.agent.components.encoder.IdentityEncoder method)
(mtrl.agent.components.encoder.MixtureofExpertsEncoder method)
(mtrl.agent.components.encoder.PixelEncoder method)
(mtrl.agent.components.hipbmdp_theta.ThetaModel method)
(mtrl.agent.components.moe_layer.AttentionBasedExperts method)
(mtrl.agent.components.moe_layer.FeedForward method)
(mtrl.agent.components.moe_layer.Linear method)
(mtrl.agent.components.moe_layer.MixtureOfExperts method)
(mtrl.agent.components.reward_decoder.RewardDecoder method)
(mtrl.agent.components.soft_modularization.RoutingNetwork method)
(mtrl.agent.components.soft_modularization.SoftModularizedMLP method)
(mtrl.agent.components.task_encoder.TaskEncoder method)
(mtrl.agent.components.transition_model.DeterministicTransitionModel method)
(mtrl.agent.components.transition_model.ProbabilisticTransitionModel method)
(mtrl.agent.components.transition_model.TransitionModel method)
forward_conv() (mtrl.agent.components.encoder.PixelEncoder method)
G
gaussian_kld() (in module mtrl.agent.distral)
get_action_when_evaluating_vec_env_of_tasks() (mtrl.experiment.dmcontrol.Experiment method)
get_alpha() (mtrl.agent.sac.Agent method)
get_component_name_list_for_checkpointing() (mtrl.agent.abstract.Agent method)
(mtrl.agent.wrapper.Agent method)
get_current_commit_id() (in module mtrl.utils.utils)
get_dirs_to_delete_from_experiment() (in module mtrl.experiment.utils)
get_env_metadata() (in module mtrl.experiment.experiment)
get_env_params_from_config() (in module mtrl.utils.config)
get_last_shared_layers() (mtrl.agent.abstract.Agent method)
(mtrl.agent.components.actor.Actor method)
(mtrl.agent.components.base.Component method)
(mtrl.agent.components.critic.Critic method)
(mtrl.agent.components.critic.QFunction method)
(mtrl.agent.components.decoder.PixelDecoder method)
(mtrl.agent.components.reward_decoder.RewardDecoder method)
(mtrl.agent.components.transition_model.DeterministicTransitionModel method)
(mtrl.agent.components.transition_model.ProbabilisticTransitionModel method)
(mtrl.agent.sac.Agent method)
(mtrl.agent.wrapper.Agent method)
get_mask() (mtrl.agent.components.moe_layer.MaskCache method)
get_optimizer_name_list_for_checkpointing() (mtrl.agent.abstract.Agent method)
(mtrl.agent.wrapper.Agent method)
get_parameters() (mtrl.agent.sac.Agent method)
get_shuffled_task_indices() (mtrl.agent.pcgrad.Agent method)
get_task_encoding() (mtrl.agent.hipbmdp.Agent method)
(mtrl.agent.sac.Agent method)
get_task_info() (mtrl.agent.sac.Agent method)
H
has_uncommitted_changes() (in module mtrl.utils.utils)
I
IdentityEncoder (class in mtrl.agent.components.encoder)
ids() (mtrl.env.vec_env.MetaWorldVecEnv property)
(mtrl.env.vec_env.VecEnv property)
init() (mtrl.utils.video.VideoRecorder method)
is_empty() (mtrl.replay_buffer.ReplayBuffer method)
is_integer() (in module mtrl.utils.utils)
L
Linear (class in mtrl.agent.components.moe_layer)
load() (mtrl.agent.abstract.Agent method)
(mtrl.agent.distral.Agent method)
(mtrl.agent.distral.DistilledAgent method)
(mtrl.agent.distral.TaskAgent method)
(mtrl.agent.wrapper.Agent method)
(mtrl.experiment.experiment.Experiment method)
(mtrl.replay_buffer.ReplayBuffer method)
(mtrl.utils.checkpointable.Checkpointable method)
load_latest_step() (mtrl.agent.abstract.Agent method)
(mtrl.agent.distral.Agent method)
(mtrl.agent.distral.DistilledAgent method)
(mtrl.agent.distral.TaskAgent method)
load_metadata() (mtrl.agent.abstract.Agent method)
log() (mtrl.logger.Logger method)
(mtrl.logger.MetersGroup method)
Logger (class in mtrl.logger)
M
make_config_immutable() (in module mtrl.utils.config)
make_config_mutable() (in module mtrl.utils.config)
make_decoder() (in module mtrl.agent.components.decoder)
make_dir() (in module mtrl.utils.utils)
make_encoder() (in module mtrl.agent.components.encoder)
make_model() (mtrl.agent.components.actor.Actor method)
make_transition_model() (in module mtrl.agent.components.transition_model)
mask_cache (mtrl.agent.components.moe_layer.OneToOneExperts attribute)
MaskCache (class in mtrl.agent.components.moe_layer)
MEAN (mtrl.agent.components.hipbmdp_theta.ThetaSamplingStrategy attribute)
MEAN_TRAIN (mtrl.agent.components.hipbmdp_theta.ThetaSamplingStrategy attribute)
MetaWorldVecEnv (class in mtrl.env.vec_env)
Meter (class in mtrl.logger)
MetersGroup (class in mtrl.logger)
MixtureOfExperts (class in mtrl.agent.components.moe_layer)
MixtureofExpertsEncoder (class in mtrl.agent.components.encoder)
mode() (mtrl.env.vec_env.MetaWorldVecEnv property)
(mtrl.env.vec_env.VecEnv property)
module
mtrl
mtrl.agent
mtrl.agent.abstract
mtrl.agent.components
mtrl.agent.components.actor
mtrl.agent.components.base
mtrl.agent.components.critic
mtrl.agent.components.decoder
mtrl.agent.components.encoder
mtrl.agent.components.hipbmdp_theta
mtrl.agent.components.moe_layer
mtrl.agent.components.reward_decoder
mtrl.agent.components.soft_modularization
mtrl.agent.components.task_encoder
mtrl.agent.components.transition_model
mtrl.agent.deepmdp
mtrl.agent.distral
mtrl.agent.ds
mtrl.agent.ds.mt_obs
mtrl.agent.ds.task_info
mtrl.agent.grad_manipulation
mtrl.agent.gradnorm
mtrl.agent.hipbmdp
mtrl.agent.pcgrad
mtrl.agent.sac
mtrl.agent.sac_ae
mtrl.agent.utils
mtrl.agent.wrapper
mtrl.app
mtrl.app.run
mtrl.env
mtrl.env.builder
mtrl.env.types
mtrl.env.vec_env
mtrl.experiment
mtrl.experiment.dmcontrol
mtrl.experiment.experiment
mtrl.experiment.metaworld
mtrl.experiment.multitask
mtrl.experiment.utils
mtrl.logger
mtrl.replay_buffer
mtrl.utils
mtrl.utils.checkpointable
mtrl.utils.config
mtrl.utils.types
mtrl.utils.utils
mtrl.utils.video
MTObs (class in mtrl.agent.ds.mt_obs)
mtrl
module
mtrl.agent
module
mtrl.agent.abstract
module
mtrl.agent.components
module
mtrl.agent.components.actor
module
mtrl.agent.components.base
module
mtrl.agent.components.critic
module
mtrl.agent.components.decoder
module
mtrl.agent.components.encoder
module
mtrl.agent.components.hipbmdp_theta
module
mtrl.agent.components.moe_layer
module
mtrl.agent.components.reward_decoder
module
mtrl.agent.components.soft_modularization
module
mtrl.agent.components.task_encoder
module
mtrl.agent.components.transition_model
module
mtrl.agent.deepmdp
module
mtrl.agent.distral
module
mtrl.agent.ds
module
mtrl.agent.ds.mt_obs
module
mtrl.agent.ds.task_info
module
mtrl.agent.grad_manipulation
module
mtrl.agent.gradnorm
module
mtrl.agent.hipbmdp
module
mtrl.agent.pcgrad
module
mtrl.agent.sac
module
mtrl.agent.sac_ae
module
mtrl.agent.utils
module
mtrl.agent.wrapper
module
mtrl.app
module
mtrl.app.run
module
mtrl.env
module
mtrl.env.builder
module
mtrl.env.types
module
mtrl.env.vec_env
module
mtrl.experiment
module
mtrl.experiment.dmcontrol
module
mtrl.experiment.experiment
module
mtrl.experiment.metaworld
module
mtrl.experiment.multitask
module
mtrl.experiment.utils
module
mtrl.logger
module
mtrl.replay_buffer
module
mtrl.utils
module
mtrl.utils.checkpointable
module
mtrl.utils.config
module
mtrl.utils.types
module
mtrl.utils.utils
module
mtrl.utils.video
module
N
next_env_obs (mtrl.replay_buffer.ReplayBufferSample attribute)
not_done (mtrl.replay_buffer.ReplayBufferSample attribute)
np_float32() (in module mtrl.logger)
np_int64() (in module mtrl.logger)
O
OneToOneExperts (class in mtrl.agent.components.moe_layer)
P
patch_agent() (mtrl.agent.distral.TaskAgent method)
periodic_save() (mtrl.experiment.experiment.Experiment method)
PixelDecoder (class in mtrl.agent.components.decoder)
PixelEncoder (class in mtrl.agent.components.encoder)
prepare_and_run() (in module mtrl.experiment.utils)
prepare_config() (in module mtrl.experiment.experiment)
preprocess_obs() (in module mtrl.agent.utils)
pretty_print() (in module mtrl.utils.config)
ProbabilisticTransitionModel (class in mtrl.agent.components.transition_model)
process_config() (in module mtrl.utils.config)
Q
QFunction (class in mtrl.agent.components.critic)
R
read_config_from_file() (in module mtrl.utils.config)
record() (mtrl.utils.video.VideoRecorder method)
register_once() (in module mtrl.env)
reparameterize() (mtrl.agent.components.encoder.PixelEncoder method)
ReplayBuffer (class in mtrl.replay_buffer)
ReplayBufferSample (class in mtrl.replay_buffer)
reset() (mtrl.env.vec_env.MetaWorldVecEnv method)
(mtrl.env.vec_env.VecEnv method)
(mtrl.replay_buffer.ReplayBuffer method)
reward (mtrl.replay_buffer.ReplayBufferSample attribute)
RewardDecoder (class in mtrl.agent.components.reward_decoder)
RoutingNetwork (class in mtrl.agent.components.soft_modularization)
run() (in module mtrl.app.run)
(mtrl.experiment.experiment.Experiment method)
(mtrl.experiment.multitask.Experiment method)
S
sample() (mtrl.replay_buffer.ReplayBuffer method)
sample_action() (mtrl.agent.abstract.Agent method)
(mtrl.agent.distral.Agent method)
(mtrl.agent.distral.DistilledAgent method)
(mtrl.agent.sac.Agent method)
(mtrl.agent.wrapper.Agent method)
sample_an_index() (mtrl.replay_buffer.ReplayBuffer method)
sample_prediction() (mtrl.agent.components.transition_model.DeterministicTransitionModel method)
(mtrl.agent.components.transition_model.ProbabilisticTransitionModel method)
(mtrl.agent.components.transition_model.TransitionModel method)
save() (mtrl.agent.abstract.Agent method)
(mtrl.agent.distral.Agent method)
(mtrl.agent.distral.DistilledAgent method)
(mtrl.agent.distral.TaskAgent method)
(mtrl.agent.wrapper.Agent method)
(mtrl.experiment.experiment.Experiment method)
(mtrl.replay_buffer.ReplayBuffer method)
(mtrl.utils.checkpointable.Checkpointable method)
(mtrl.utils.video.VideoRecorder method)
save_components() (mtrl.agent.abstract.Agent method)
(mtrl.agent.wrapper.Agent method)
save_components_or_optimizers() (mtrl.agent.abstract.Agent method)
save_metadata() (mtrl.agent.abstract.Agent method)
save_optimizers() (mtrl.agent.abstract.Agent method)
(mtrl.agent.wrapper.Agent method)
select_action() (mtrl.agent.abstract.Agent method)
(mtrl.agent.distral.Agent method)
(mtrl.agent.distral.DistilledAgent method)
(mtrl.agent.sac.Agent method)
(mtrl.agent.wrapper.Agent method)
serialize_log() (in module mtrl.logger)
set_seed() (in module mtrl.utils.utils)
set_seed_everywhere() (in module mtrl.agent.utils)
set_struct() (in module mtrl.utils.config)
soft_update_params() (in module mtrl.agent.utils)
SoftModularizedMLP (class in mtrl.agent.components.soft_modularization)
split_on_caps() (in module mtrl.utils.utils)
startup_logs() (mtrl.experiment.experiment.Experiment method)
step() (mtrl.env.vec_env.MetaWorldVecEnv method)
(mtrl.env.vec_env.VecEnv method)
T
task_info (mtrl.agent.ds.mt_obs.MTObs attribute)
task_obs (mtrl.agent.ds.mt_obs.MTObs attribute)
(mtrl.replay_buffer.ReplayBufferSample attribute)
TaskAgent (class in mtrl.agent.distral)
TaskEncoder (class in mtrl.agent.components.task_encoder)
TaskInfo (class in mtrl.agent.ds.task_info)
ThetaModel (class in mtrl.agent.components.hipbmdp_theta)
ThetaSamplingStrategy (class in mtrl.agent.components.hipbmdp_theta)
tie_weights() (in module mtrl.agent.components.encoder)
to_dict() (in module mtrl.utils.config)
train() (mtrl.agent.abstract.Agent method)
(mtrl.agent.distral.Agent method)
(mtrl.agent.distral.DistilledAgent method)
(mtrl.agent.sac.Agent method)
(mtrl.agent.wrapper.Agent method)
training (mtrl.agent.components.actor.Actor attribute)
(mtrl.agent.components.actor.BaseActor attribute)
(mtrl.agent.components.base.Component attribute)
(mtrl.agent.components.critic.Critic attribute)
(mtrl.agent.components.critic.QFunction attribute)
(mtrl.agent.components.decoder.PixelDecoder attribute)
(mtrl.agent.components.encoder.Encoder attribute)
(mtrl.agent.components.encoder.FeedForwardEncoder attribute)
(mtrl.agent.components.encoder.FiLM attribute)
(mtrl.agent.components.encoder.IdentityEncoder attribute)
(mtrl.agent.components.encoder.MixtureofExpertsEncoder attribute)
(mtrl.agent.components.encoder.PixelEncoder attribute)
(mtrl.agent.components.hipbmdp_theta.ThetaModel attribute)
(mtrl.agent.components.moe_layer.AttentionBasedExperts attribute)
(mtrl.agent.components.moe_layer.ClusterOfExperts attribute)
(mtrl.agent.components.moe_layer.EnsembleOfExperts attribute)
(mtrl.agent.components.moe_layer.FeedForward attribute)
(mtrl.agent.components.moe_layer.Linear attribute)
(mtrl.agent.components.moe_layer.MixtureOfExperts attribute)
(mtrl.agent.components.moe_layer.OneToOneExperts attribute)
(mtrl.agent.components.reward_decoder.RewardDecoder attribute)
(mtrl.agent.components.soft_modularization.RoutingNetwork attribute)
(mtrl.agent.components.soft_modularization.SoftModularizedMLP attribute)
(mtrl.agent.components.task_encoder.TaskEncoder attribute)
(mtrl.agent.components.transition_model.DeterministicTransitionModel attribute)
(mtrl.agent.components.transition_model.ProbabilisticTransitionModel attribute)
(mtrl.agent.components.transition_model.TransitionModel attribute)
TransitionModel (class in mtrl.agent.components.transition_model)
U
unique_env_index (mtrl.agent.grad_manipulation.EnvMetadata attribute)
unset_struct() (in module mtrl.utils.config)
update() (mtrl.agent.abstract.Agent method)
(mtrl.agent.distral.Agent method)
(mtrl.agent.distral.DistilledAgent method)
(mtrl.agent.grad_manipulation.Agent method)
(mtrl.agent.sac.Agent method)
(mtrl.agent.wrapper.Agent method)
(mtrl.logger.AverageMeter method)
(mtrl.logger.CurrentMeter method)
(mtrl.logger.Meter method)
update_actor_and_alpha() (mtrl.agent.distral.TaskAgent method)
(mtrl.agent.sac.Agent method)
update_critic() (mtrl.agent.sac.Agent method)
update_decoder() (mtrl.agent.deepmdp.Agent method)
(mtrl.agent.sac.Agent method)
(mtrl.agent.sac_ae.Agent method)
update_task_encoder() (mtrl.agent.hipbmdp.Agent method)
(mtrl.agent.sac.Agent method)
update_transition_reward_model() (mtrl.agent.deepmdp.Agent method)
(mtrl.agent.sac.Agent method)
V
value() (mtrl.logger.AverageMeter method)
(mtrl.logger.CurrentMeter method)
(mtrl.logger.Meter method)
VecEnv (class in mtrl.env.vec_env)
VideoRecorder (class in mtrl.utils.video)
W
weight_init() (in module mtrl.agent.utils)
weight_init_conv() (in module mtrl.agent.utils)
weight_init_linear() (in module mtrl.agent.utils)
weight_init_moe_layer() (in module mtrl.agent.utils)
Z
ZERO (mtrl.agent.components.hipbmdp_theta.ThetaSamplingStrategy attribute)
Read the Docs
v: latest
Versions
latest
Downloads
pdf
On Read the Docs
Project Home
Builds