d4rl_mujoco_hopper

  • Descrizione :

D4RL è un benchmark open source per l'apprendimento per rinforzo offline. Fornisce ambienti e set di dati standardizzati per algoritmi di addestramento e benchmarking.

I set di dati seguono il formato RLDS per rappresentare passaggi ed episodi.

@misc{fu2020d4rl,
    title={D4RL: Datasets for Deep Data-Driven Reinforcement Learning},
    author={Justin Fu and Aviral Kumar and Ofir Nachum and George Tucker and Sergey Levine},
    year={2020},
    eprint={2004.07219},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

d4rl_mujoco_hopper/v0-expert (configurazione predefinita)

  • Dimensione del download : 51.56 MiB

  • Dimensione del set di dati: 64.10 MiB

  • Auto-cache ( documentazione ): Sì

  • Divisioni :

Diviso Esempi
'train' 1.029
  • Struttura delle caratteristiche :
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v0-medium

  • Dimensione del download : 51.74 MiB

  • Dimensione del set di dati: 64.68 MiB

  • Auto-cache ( documentazione ): Sì

  • Divisioni :

Diviso Esempi
'train' 3.064
  • Struttura delle caratteristiche :
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v0-medium-expert

  • Dimensione del download : 62.01 MiB

  • Dimensione del set di dati: 77.25 MiB

  • Auto-cache ( documentazione ): Sì

  • Divisioni :

Diviso Esempi
'train' 2.277
  • Struttura delle caratteristiche :
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v0-misto

  • Dimensioni del download : 10.48 MiB

  • Dimensione del set di dati: 13.15 MiB

  • Auto-cache ( documentazione ): Sì

  • Divisioni :

Diviso Esempi
'train' 1.250
  • Struttura delle caratteristiche :
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v0-random

  • Dimensione del download : 51.83 MiB

  • Dimensione del set di dati: 66.06 MiB

  • Auto-cache ( documentazione ): Sì

  • Divisioni :

Diviso Esempi
'train' 8.793
  • Struttura delle caratteristiche :
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v1-expert

  • Dimensione del download : 93.19 MiB

  • Dimensione del set di dati: 608.03 MiB

  • Cache automatica ( documentazione ): No

  • Divisioni :

Diviso Esempi
'train' 1.836
  • Struttura delle caratteristiche :
FeaturesDict({
    'algorithm': string,
    'iteration': int32,
    'policy': FeaturesDict({
        'fc0': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=float32),
            'weight': Tensor(shape=(256, 11), dtype=float32),
        }),
        'fc1': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=float32),
            'weight': Tensor(shape=(256, 256), dtype=float32),
        }),
        'last_fc': FeaturesDict({
            'bias': Tensor(shape=(3,), dtype=float32),
            'weight': Tensor(shape=(3, 256), dtype=float32),
        }),
        'last_fc_log_std': FeaturesDict({
            'bias': Tensor(shape=(3,), dtype=float32),
            'weight': Tensor(shape=(3, 256), dtype=float32),
        }),
        'nonlinearity': string,
        'output_distribution': string,
    }),
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'infos': FeaturesDict({
            'action_log_probs': float32,
            'qpos': Tensor(shape=(6,), dtype=float32),
            'qvel': Tensor(shape=(6,), dtype=float32),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
algoritmo Tensore corda
iterazione Tensore int32
politica CaratteristicheDict
politica/fc0 CaratteristicheDict
politica/fc0/bias Tensore (256,) galleggiante32
politica/fc0/peso Tensore (256, 11) galleggiante32
politica/fc1 CaratteristicheDict
politica/fc1/bias Tensore (256,) galleggiante32
politica/fc1/peso Tensore (256, 256) galleggiante32
politica/ultimo_fc CaratteristicheDict
policy/last_fc/bias Tensore (3,) galleggiante32
policy/last_fc/peso Tensore (3, 256) galleggiante32
policy/last_fc_log_std CaratteristicheDict
politica/last_fc_log_std/bias Tensore (3,) galleggiante32
politica/last_fc_log_std/peso Tensore (3, 256) galleggiante32
politica/non linearità Tensore corda
politica/output_distribution Tensore corda
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante32
passi/informazioni/qpos Tensore (6,) galleggiante32
passaggi/informazioni/qvel Tensore (6,) galleggiante32
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v1-medium

  • Dimensione del download : 92.03 MiB

  • Dimensione del set di dati: 1.78 GiB

  • Cache automatica ( documentazione ): No

  • Divisioni :

Diviso Esempi
'train' 6.328
  • Struttura delle caratteristiche :
FeaturesDict({
    'algorithm': string,
    'iteration': int32,
    'policy': FeaturesDict({
        'fc0': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=float32),
            'weight': Tensor(shape=(256, 11), dtype=float32),
        }),
        'fc1': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=float32),
            'weight': Tensor(shape=(256, 256), dtype=float32),
        }),
        'last_fc': FeaturesDict({
            'bias': Tensor(shape=(3,), dtype=float32),
            'weight': Tensor(shape=(3, 256), dtype=float32),
        }),
        'last_fc_log_std': FeaturesDict({
            'bias': Tensor(shape=(3,), dtype=float32),
            'weight': Tensor(shape=(3, 256), dtype=float32),
        }),
        'nonlinearity': string,
        'output_distribution': string,
    }),
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'infos': FeaturesDict({
            'action_log_probs': float32,
            'qpos': Tensor(shape=(6,), dtype=float32),
            'qvel': Tensor(shape=(6,), dtype=float32),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
algoritmo Tensore corda
iterazione Tensore int32
politica CaratteristicheDict
politica/fc0 CaratteristicheDict
politica/fc0/bias Tensore (256,) galleggiante32
politica/fc0/peso Tensore (256, 11) galleggiante32
politica/fc1 CaratteristicheDict
politica/fc1/bias Tensore (256,) galleggiante32
politica/fc1/peso Tensore (256, 256) galleggiante32
politica/ultimo_fc CaratteristicheDict
policy/last_fc/bias Tensore (3,) galleggiante32
policy/last_fc/peso Tensore (3, 256) galleggiante32
policy/last_fc_log_std CaratteristicheDict
politica/last_fc_log_std/bias Tensore (3,) galleggiante32
politica/last_fc_log_std/peso Tensore (3, 256) galleggiante32
politica/non linearità Tensore corda
politica/output_distribution Tensore corda
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante32
passi/informazioni/qpos Tensore (6,) galleggiante32
passaggi/informazioni/qvel Tensore (6,) galleggiante32
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v1-medium-expert

  • Dimensione del download : 184.59 MiB

  • Dimensione del set di dati: 230.24 MiB

  • Cache automatica ( documentazione ): solo quando shuffle_files=False (treno)

  • Divisioni :

Diviso Esempi
'train' 8.163
  • Struttura delle caratteristiche :
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'infos': FeaturesDict({
            'action_log_probs': float32,
            'qpos': Tensor(shape=(6,), dtype=float32),
            'qvel': Tensor(shape=(6,), dtype=float32),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante32
passi/informazioni/qpos Tensore (6,) galleggiante32
passaggi/informazioni/qvel Tensore (6,) galleggiante32
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v1-medium-replay

  • Dimensione del download : 55.65 MiB

  • Dimensione del set di dati: 34.78 MiB

  • Auto-cache ( documentazione ): Sì

  • Divisioni :

Diviso Esempi
'train' 1.151
  • Struttura delle caratteristiche :
FeaturesDict({
    'algorithm': string,
    'iteration': int32,
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float64),
        'discount': float64,
        'infos': FeaturesDict({
            'action_log_probs': float64,
            'qpos': Tensor(shape=(6,), dtype=float64),
            'qvel': Tensor(shape=(6,), dtype=float64),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float64),
        'reward': float64,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
algoritmo Tensore corda
iterazione Tensore int32
passi Set di dati
passi/azione Tensore (3,) galleggiante64
passi/sconto Tensore galleggiante64
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante64
passi/informazioni/qpos Tensore (6,) galleggiante64
passaggi/informazioni/qvel Tensore (6,) galleggiante64
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante64
passi/ricompensa Tensore galleggiante64

d4rl_mujoco_hopper/v1-full-replay

  • Dimensione del download : 183.32 MiB

  • Dimensione del set di dati: 114.78 MiB

  • Auto-cache ( documentazione ): Sì

  • Divisioni :

Diviso Esempi
'train' 2.907
  • Struttura delle caratteristiche :
FeaturesDict({
    'algorithm': string,
    'iteration': int32,
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float64),
        'discount': float64,
        'infos': FeaturesDict({
            'action_log_probs': float64,
            'qpos': Tensor(shape=(6,), dtype=float64),
            'qvel': Tensor(shape=(6,), dtype=float64),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float64),
        'reward': float64,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
algoritmo Tensore corda
iterazione Tensore int32
passi Set di dati
passi/azione Tensore (3,) galleggiante64
passi/sconto Tensore galleggiante64
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante64
passi/informazioni/qpos Tensore (6,) galleggiante64
passaggi/informazioni/qvel Tensore (6,) galleggiante64
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante64
passi/ricompensa Tensore galleggiante64

d4rl_mujoco_hopper/v1-random

  • Dimensione del download : 91.11 MiB

  • Dimensione del set di dati: 130.73 MiB

  • Cache automatica ( documentazione ): solo quando shuffle_files=False (treno)

  • Divisioni :

Diviso Esempi
'train' 45.265
  • Struttura delle caratteristiche :
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'infos': FeaturesDict({
            'action_log_probs': float32,
            'qpos': Tensor(shape=(6,), dtype=float32),
            'qvel': Tensor(shape=(6,), dtype=float32),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante32
passi/informazioni/qpos Tensore (6,) galleggiante32
passaggi/informazioni/qvel Tensore (6,) galleggiante32
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v2-expert

  • Dimensione del download : 145.37 MiB

  • Dimensione del set di dati: 390.40 MiB

  • Cache automatica ( documentazione ): No

  • Divisioni :

Diviso Esempi
'train' 1.028
  • Struttura delle caratteristiche :
FeaturesDict({
    'algorithm': string,
    'iteration': int32,
    'policy': FeaturesDict({
        'fc0': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=float32),
            'weight': Tensor(shape=(256, 11), dtype=float32),
        }),
        'fc1': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=float32),
            'weight': Tensor(shape=(256, 256), dtype=float32),
        }),
        'last_fc': FeaturesDict({
            'bias': Tensor(shape=(3,), dtype=float32),
            'weight': Tensor(shape=(3, 256), dtype=float32),
        }),
        'last_fc_log_std': FeaturesDict({
            'bias': Tensor(shape=(3,), dtype=float32),
            'weight': Tensor(shape=(3, 256), dtype=float32),
        }),
        'nonlinearity': string,
        'output_distribution': string,
    }),
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'infos': FeaturesDict({
            'action_log_probs': float64,
            'qpos': Tensor(shape=(6,), dtype=float64),
            'qvel': Tensor(shape=(6,), dtype=float64),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
algoritmo Tensore corda
iterazione Tensore int32
politica CaratteristicheDict
politica/fc0 CaratteristicheDict
politica/fc0/bias Tensore (256,) galleggiante32
politica/fc0/peso Tensore (256, 11) galleggiante32
politica/fc1 CaratteristicheDict
politica/fc1/bias Tensore (256,) galleggiante32
politica/fc1/peso Tensore (256, 256) galleggiante32
politica/ultimo_fc CaratteristicheDict
policy/last_fc/bias Tensore (3,) galleggiante32
policy/last_fc/peso Tensore (3, 256) galleggiante32
policy/last_fc_log_std CaratteristicheDict
politica/last_fc_log_std/bias Tensore (3,) galleggiante32
politica/last_fc_log_std/peso Tensore (3, 256) galleggiante32
politica/non linearità Tensore corda
politica/output_distribution Tensore corda
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante64
passi/informazioni/qpos Tensore (6,) galleggiante64
passaggi/informazioni/qvel Tensore (6,) galleggiante64
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v2-full-replay

  • Dimensione del download : 179.29 MiB

  • Dimensione del set di dati: 115.04 MiB

  • Auto-cache ( documentazione ): Sì

  • Divisioni :

Diviso Esempi
'train' 3.515
  • Struttura delle caratteristiche :
FeaturesDict({
    'algorithm': string,
    'iteration': int32,
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'infos': FeaturesDict({
            'action_log_probs': float64,
            'qpos': Tensor(shape=(6,), dtype=float64),
            'qvel': Tensor(shape=(6,), dtype=float64),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
algoritmo Tensore corda
iterazione Tensore int32
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante64
passi/informazioni/qpos Tensore (6,) galleggiante64
passaggi/informazioni/qvel Tensore (6,) galleggiante64
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v2-medium

  • Dimensione del download : 145.68 MiB

  • Dimensione del set di dati: 702.57 MiB

  • Cache automatica ( documentazione ): No

  • Divisioni :

Diviso Esempi
'train' 2.187
  • Struttura delle caratteristiche :
FeaturesDict({
    'algorithm': string,
    'iteration': int32,
    'policy': FeaturesDict({
        'fc0': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=float32),
            'weight': Tensor(shape=(256, 11), dtype=float32),
        }),
        'fc1': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=float32),
            'weight': Tensor(shape=(256, 256), dtype=float32),
        }),
        'last_fc': FeaturesDict({
            'bias': Tensor(shape=(3,), dtype=float32),
            'weight': Tensor(shape=(3, 256), dtype=float32),
        }),
        'last_fc_log_std': FeaturesDict({
            'bias': Tensor(shape=(3,), dtype=float32),
            'weight': Tensor(shape=(3, 256), dtype=float32),
        }),
        'nonlinearity': string,
        'output_distribution': string,
    }),
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'infos': FeaturesDict({
            'action_log_probs': float64,
            'qpos': Tensor(shape=(6,), dtype=float64),
            'qvel': Tensor(shape=(6,), dtype=float64),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
algoritmo Tensore corda
iterazione Tensore int32
politica CaratteristicheDict
politica/fc0 CaratteristicheDict
politica/fc0/bias Tensore (256,) galleggiante32
politica/fc0/peso Tensore (256, 11) galleggiante32
politica/fc1 CaratteristicheDict
politica/fc1/bias Tensore (256,) galleggiante32
politica/fc1/peso Tensore (256, 256) galleggiante32
politica/ultimo_fc CaratteristicheDict
policy/last_fc/bias Tensore (3,) galleggiante32
policy/last_fc/peso Tensore (3, 256) galleggiante32
policy/last_fc_log_std CaratteristicheDict
politica/last_fc_log_std/bias Tensore (3,) galleggiante32
politica/last_fc_log_std/peso Tensore (3, 256) galleggiante32
politica/non linearità Tensore corda
politica/output_distribution Tensore corda
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante64
passi/informazioni/qpos Tensore (6,) galleggiante64
passaggi/informazioni/qvel Tensore (6,) galleggiante64
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v2-medium-expert

  • Dimensione del download : 290.43 MiB

  • Dimensione del set di dati: 228.28 MiB

  • Cache automatica ( documentazione ): solo quando shuffle_files=False (treno)

  • Divisioni :

Diviso Esempi
'train' 3.214
  • Struttura delle caratteristiche :
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'infos': FeaturesDict({
            'action_log_probs': float64,
            'qpos': Tensor(shape=(6,), dtype=float64),
            'qvel': Tensor(shape=(6,), dtype=float64),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante64
passi/informazioni/qpos Tensore (6,) galleggiante64
passaggi/informazioni/qvel Tensore (6,) galleggiante64
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v2-medium-replay

  • Dimensione del download : 72.34 MiB

  • Dimensione del set di dati: 46.51 MiB

  • Auto-cache ( documentazione ): Sì

  • Divisioni :

Diviso Esempi
'train' 2.041
  • Struttura delle caratteristiche :
FeaturesDict({
    'algorithm': string,
    'iteration': int32,
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'infos': FeaturesDict({
            'action_log_probs': float64,
            'qpos': Tensor(shape=(6,), dtype=float64),
            'qvel': Tensor(shape=(6,), dtype=float64),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
algoritmo Tensore corda
iterazione Tensore int32
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante64
passi/informazioni/qpos Tensore (6,) galleggiante64
passaggi/informazioni/qvel Tensore (6,) galleggiante64
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32

d4rl_mujoco_hopper/v2-random

  • Dimensione del download : 145.46 MiB

  • Dimensione del set di dati: 130.72 MiB

  • Cache automatica ( documentazione ): solo quando shuffle_files=False (treno)

  • Divisioni :

Diviso Esempi
'train' 45.240
  • Struttura delle caratteristiche :
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(3,), dtype=float32),
        'discount': float32,
        'infos': FeaturesDict({
            'action_log_probs': float64,
            'qpos': Tensor(shape=(6,), dtype=float64),
            'qvel': Tensor(shape=(6,), dtype=float64),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': Tensor(shape=(11,), dtype=float32),
        'reward': float32,
    }),
})
  • Documentazione delle funzionalità :
Caratteristica Classe Forma Tipo D Descrizione
CaratteristicheDict
passi Set di dati
passi/azione Tensore (3,) galleggiante32
passi/sconto Tensore galleggiante32
passaggi/informazioni CaratteristicheDict
steps/infos/action_log_probs Tensore galleggiante64
passi/informazioni/qpos Tensore (6,) galleggiante64
passaggi/informazioni/qvel Tensore (6,) galleggiante64
passi/è_primo Tensore bool
passi/è_ultimo Tensore bool
passi/è_terminale Tensore bool
passi/osservazione Tensore (11,) galleggiante32
passi/ricompensa Tensore galleggiante32