ML Community Day is November 9! Join us for updates from TensorFlow, JAX, and more Learn more

d4rl_mujoco_ant

  • Description:

D4RL is an open-source benchmark for offline reinforcement learning. It provides standardized environments and datasets for training and benchmarking algorithms.

@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_ant/v0-expert (default config)

  • Download size: 131.34 MiB

  • Dataset size: 464.94 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 1,288
  • Features:
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v0-medium

  • Download size: 131.39 MiB

  • Dataset size: 464.78 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 1,122
  • Features:
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v0-medium-expert

  • Download size: 262.73 MiB

  • Dataset size: 929.71 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 2,410
  • Features:
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v0-mixed

  • Download size: 104.63 MiB

  • Dataset size: 464.93 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 1,320
  • Features:
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v0-random

  • Download size: 139.50 MiB

  • Dataset size: 464.97 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 1,377
  • Features:
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v1-expert

  • Download size: 220.72 MiB

  • Dataset size: 968.07 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 1,033
  • Features:
FeaturesDict({
    'algorithm': tf.string,
    'iteration': tf.int32,
    'policy': FeaturesDict({
        'fc0': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=tf.float32),
            'weight': Tensor(shape=(256, 111), dtype=tf.float32),
        }),
        'fc1': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=tf.float32),
            'weight': Tensor(shape=(256, 256), dtype=tf.float32),
        }),
        'last_fc': FeaturesDict({
            'bias': Tensor(shape=(8,), dtype=tf.float32),
            'weight': Tensor(shape=(8, 256), dtype=tf.float32),
        }),
        'last_fc_log_std': FeaturesDict({
            'bias': Tensor(shape=(8,), dtype=tf.float32),
            'weight': Tensor(shape=(8, 256), dtype=tf.float32),
        }),
        'nonlinearity': tf.string,
        'output_distribution': tf.string,
    }),
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'infos': FeaturesDict({
            'action_log_probs': tf.float32,
            'qpos': Tensor(shape=(15,), dtype=tf.float32),
            'qvel': Tensor(shape=(14,), dtype=tf.float32),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v1-medium

  • Download size: 222.39 MiB

  • Dataset size: 1023.23 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 1,179
  • Features:
FeaturesDict({
    'algorithm': tf.string,
    'iteration': tf.int32,
    'policy': FeaturesDict({
        'fc0': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=tf.float32),
            'weight': Tensor(shape=(256, 111), dtype=tf.float32),
        }),
        'fc1': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=tf.float32),
            'weight': Tensor(shape=(256, 256), dtype=tf.float32),
        }),
        'last_fc': FeaturesDict({
            'bias': Tensor(shape=(8,), dtype=tf.float32),
            'weight': Tensor(shape=(8, 256), dtype=tf.float32),
        }),
        'last_fc_log_std': FeaturesDict({
            'bias': Tensor(shape=(8,), dtype=tf.float32),
            'weight': Tensor(shape=(8, 256), dtype=tf.float32),
        }),
        'nonlinearity': tf.string,
        'output_distribution': tf.string,
    }),
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'infos': FeaturesDict({
            'action_log_probs': tf.float32,
            'qpos': Tensor(shape=(15,), dtype=tf.float32),
            'qvel': Tensor(shape=(14,), dtype=tf.float32),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v1-medium-expert

  • Download size: 442.25 MiB

  • Dataset size: 1.13 GiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 2,211
  • Features:
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'infos': FeaturesDict({
            'action_log_probs': tf.float32,
            'qpos': Tensor(shape=(15,), dtype=tf.float32),
            'qvel': Tensor(shape=(14,), dtype=tf.float32),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v1-medium-replay

  • Download size: 132.05 MiB

  • Dataset size: 175.09 MiB

  • Auto-cached (documentation): Only when shuffle_files=False (train)

  • Splits:

Split Examples
'train' 485
  • Features:
FeaturesDict({
    'algorithm': tf.string,
    'iteration': tf.int32,
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float64),
        'discount': tf.float64,
        'infos': FeaturesDict({
            'action_log_probs': tf.float64,
            'qpos': Tensor(shape=(15,), dtype=tf.float64),
            'qvel': Tensor(shape=(14,), dtype=tf.float64),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float64),
        'reward': tf.float64,
    }),
})

d4rl_mujoco_ant/v1-full-replay

  • Download size: 437.57 MiB

  • Dataset size: 579.52 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 1,319
  • Features:
FeaturesDict({
    'algorithm': tf.string,
    'iteration': tf.int32,
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float64),
        'discount': tf.float64,
        'infos': FeaturesDict({
            'action_log_probs': tf.float64,
            'qpos': Tensor(shape=(15,), dtype=tf.float64),
            'qvel': Tensor(shape=(14,), dtype=tf.float64),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float64),
        'reward': tf.float64,
    }),
})

d4rl_mujoco_ant/v1-random

  • Download size: 225.18 MiB

  • Dataset size: 583.49 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 5,741
  • Features:
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'infos': FeaturesDict({
            'action_log_probs': tf.float32,
            'qpos': Tensor(shape=(15,), dtype=tf.float32),
            'qvel': Tensor(shape=(14,), dtype=tf.float32),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v2-expert

  • Download size: 317.69 MiB

  • Dataset size: 968.83 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 1,035
  • Features:
FeaturesDict({
    'algorithm': tf.string,
    'iteration': tf.int32,
    'policy': FeaturesDict({
        'fc0': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=tf.float32),
            'weight': Tensor(shape=(256, 111), dtype=tf.float32),
        }),
        'fc1': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=tf.float32),
            'weight': Tensor(shape=(256, 256), dtype=tf.float32),
        }),
        'last_fc': FeaturesDict({
            'bias': Tensor(shape=(8,), dtype=tf.float32),
            'weight': Tensor(shape=(8, 256), dtype=tf.float32),
        }),
        'last_fc_log_std': FeaturesDict({
            'bias': Tensor(shape=(8,), dtype=tf.float32),
            'weight': Tensor(shape=(8, 256), dtype=tf.float32),
        }),
        'nonlinearity': tf.string,
        'output_distribution': tf.string,
    }),
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'infos': FeaturesDict({
            'action_log_probs': tf.float32,
            'qpos': Tensor(shape=(15,), dtype=tf.float32),
            'qvel': Tensor(shape=(14,), dtype=tf.float32),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v2-full-replay

  • Download size: 437.57 MiB

  • Dataset size: 579.52 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 1,319
  • Features:
FeaturesDict({
    'algorithm': tf.string,
    'iteration': tf.int32,
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float64),
        'discount': tf.float64,
        'infos': FeaturesDict({
            'action_log_probs': tf.float64,
            'qpos': Tensor(shape=(15,), dtype=tf.float64),
            'qvel': Tensor(shape=(14,), dtype=tf.float64),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float64),
        'reward': tf.float64,
    }),
})

d4rl_mujoco_ant/v2-medium

  • Download size: 320.11 MiB

  • Dataset size: 1.01 GiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 1,203
  • Features:
FeaturesDict({
    'algorithm': tf.string,
    'iteration': tf.int32,
    'policy': FeaturesDict({
        'fc0': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=tf.float32),
            'weight': Tensor(shape=(256, 111), dtype=tf.float32),
        }),
        'fc1': FeaturesDict({
            'bias': Tensor(shape=(256,), dtype=tf.float32),
            'weight': Tensor(shape=(256, 256), dtype=tf.float32),
        }),
        'last_fc': FeaturesDict({
            'bias': Tensor(shape=(8,), dtype=tf.float32),
            'weight': Tensor(shape=(8, 256), dtype=tf.float32),
        }),
        'last_fc_log_std': FeaturesDict({
            'bias': Tensor(shape=(8,), dtype=tf.float32),
            'weight': Tensor(shape=(8, 256), dtype=tf.float32),
        }),
        'nonlinearity': tf.string,
        'output_distribution': tf.string,
    }),
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'infos': FeaturesDict({
            'action_log_probs': tf.float32,
            'qpos': Tensor(shape=(15,), dtype=tf.float32),
            'qvel': Tensor(shape=(14,), dtype=tf.float32),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v2-medium-expert

  • Download size: 637.02 MiB

  • Dataset size: 1.13 GiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 2,237
  • Features:
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'infos': FeaturesDict({
            'action_log_probs': tf.float32,
            'qpos': Tensor(shape=(15,), dtype=tf.float32),
            'qvel': Tensor(shape=(14,), dtype=tf.float32),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})

d4rl_mujoco_ant/v2-medium-replay

  • Download size: 132.05 MiB

  • Dataset size: 175.09 MiB

  • Auto-cached (documentation): Only when shuffle_files=False (train)

  • Splits:

Split Examples
'train' 485
  • Features:
FeaturesDict({
    'algorithm': tf.string,
    'iteration': tf.int32,
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float64),
        'discount': tf.float64,
        'infos': FeaturesDict({
            'action_log_probs': tf.float64,
            'qpos': Tensor(shape=(15,), dtype=tf.float64),
            'qvel': Tensor(shape=(14,), dtype=tf.float64),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float64),
        'reward': tf.float64,
    }),
})

d4rl_mujoco_ant/v2-random

  • Download size: 326.67 MiB

  • Dataset size: 583.57 MiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'train' 5,822
  • Features:
FeaturesDict({
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=tf.float32),
        'discount': tf.float32,
        'infos': FeaturesDict({
            'action_log_probs': tf.float32,
            'qpos': Tensor(shape=(15,), dtype=tf.float32),
            'qvel': Tensor(shape=(14,), dtype=tf.float32),
        }),
        'is_first': tf.bool,
        'is_last': tf.bool,
        'is_terminal': tf.bool,
        'observation': Tensor(shape=(111,), dtype=tf.float32),
        'reward': tf.float32,
    }),
})