rlhfblender.data_collection package

Submodules

rlhfblender.data_collection.babyai_connector module

rlhfblender.data_collection.demo_session module

rlhfblender.data_collection.environment_handler module

rlhfblender.data_collection.episode_recorder module

rlhfblender.data_collection.feedback_model module

rlhfblender.data_collection.feedback_model_handler module

rlhfblender.data_collection.feedback_translator module

rlhfblender.data_collection.framework_selector module

rlhfblender.data_collection.imitation_connector module

rlhfblender.data_collection.metrics_processor module

rlhfblender.data_collection.metrics_processor.process_metrics(benchmark_results)[source]

Compute additional metrics on a per model/per benchmark basis :param benchmark_results: (RecordedEpisodes) Container of benchmark results :return metrics: (dict) Metrics

Parameters:: benchmark_results (RecordedEpisodesContainer)
Return type:: dict

rlhfblender.data_collection.sampler module

rlhfblender.data_collection.sb_zoo_connector module

Module contents

class rlhfblender.data_collection.RecordedEpisodesContainer(obs: numpy.ndarray, rewards: numpy.ndarray, dones: numpy.ndarray, actions: numpy.ndarray, infos: numpy.ndarray, renders: numpy.ndarray, features: numpy.ndarray, probs: numpy.ndarray, env_states: numpy.ndarray, episode_rewards: numpy.ndarray, episode_lengths: numpy.ndarray, additional_metrics: numpy.ndarray)[source]

Bases: object

Parameters:

obs (ndarray)
rewards (ndarray)
dones (ndarray)
actions (ndarray)
infos (ndarray)
renders (ndarray)
features (ndarray)
probs (ndarray)
env_states (ndarray)
episode_rewards (ndarray)
episode_lengths (ndarray)
additional_metrics (ndarray)

actions: ndarray

additional_metrics: ndarray

dones: ndarray

env_states: ndarray

episode_lengths: ndarray

episode_rewards: ndarray

features: ndarray

infos: ndarray

obs: ndarray

probs: ndarray

renders: ndarray

rewards: ndarray