Skip to content

Pusher

Lerax port of Gymnasium's Pusher environment. A 7-DoF arm on a fixed base must push a free cylinder lying on the table to a fixed goal position. The cylinder is re-spawned each episode to a random position at least 0.17 from the goal.

Observation space

23-dim float vector: concat(qpos[:7], qvel[:7], tips_arm_xipos(3), object_xipos(3), goal_xipos(3)). Unbounded Box.

Action space

Box(low, high) from the model's actuator_ctrlrange — 7 continuous joint torques.

Reward

reward_dist + reward_ctrl + reward_near where

  • reward_near = -||object - tips_arm|| * reward_near_weight (default 0.5)
  • reward_dist = -||object - goal|| * reward_dist_weight (default 1.0)
  • reward_ctrl = -sum(action ** 2) * reward_control_weight (default 0.1)

Termination

Never terminates. No built-in truncation.

lerax.env.mujoco.Pusher

Bases: AbstractMujocoEnv[Float[Array, '...'], Float[Array, '...']]

MJX-based pusher environment matching Gymnasium's Pusher-v5.

unwrapped property 

unwrapped: Self

Return the unwrapped environment

action_mask 

action_mask(
    state: MujocoEnvState, *, key: Key[Array, ""]
) -> None

transition 

transition(
    state: MujocoEnvState,
    action: ActType,
    *,
    key: Key[Array, ""],
) -> MujocoEnvState

truncate 

truncate(state: MujocoEnvState) -> Bool[Array, '']

state_info 

state_info(state: MujocoEnvState) -> dict

default_renderer 

default_renderer() -> MujocoRenderer

render 

render(state: MujocoEnvState, renderer: AbstractRenderer)

render_states 

render_states(
    states: Sequence[StateType],
    renderer: AbstractRenderer | Literal["auto"] = "auto",
    dt: float = 0.0,
)

Render a sequence of frames from multiple states.

Parameters:

Name Type Description Default
states Sequence[StateType]

A sequence of environment states to render.

 required
renderer AbstractRenderer | Literal['auto']

The renderer to use for rendering. If "auto", uses the default renderer.

 'auto'
dt float

The time delay between rendering each frame, in seconds.

 0.0

render_stacked 

render_stacked(
    states: StateType,
    renderer: AbstractRenderer | Literal["auto"] = "auto",
    dt: float = 0.0,
)

Render multiple frames from stacked states.

Stacked states are typically batched states stored in a pytree structure.

Parameters:

Name Type Description Default
states StateType

A pytree of stacked environment states to render.

 required
renderer AbstractRenderer | Literal['auto']

The renderer to use for rendering. If "auto", uses the default renderer.

 'auto'
dt float

The time delay between rendering each frame, in seconds.

 0.0

reset 

reset(
    *, key: Key[Array, ""]
) -> tuple[StateType, ObsType, dict]

Wrap the functional logic into a Gym API reset method.

Parameters:

Name Type Description Default
key Key[Array, '']

A JAX PRNG key for any stochasticity in the reset.

 required

Returns:

Type Description
tuple[StateType, ObsType, dict]

A tuple of the initial state, initial observation, and additional info.

step 

step(
    state: StateType,
    action: ActType,
    *,
    key: Key[Array, ""],
) -> tuple[
    StateType,
    ObsType,
    Float[Array, ""],
    Bool[Array, ""],
    Bool[Array, ""],
    dict,
]

Wrap the functional logic into a Gym API step method.

Parameters:

Name Type Description Default
state StateType

The current environment state.

 required
action ActType

The action to take.

 required
key Key[Array, '']

A JAX PRNG key for any stochasticity in the step.

 required

Returns:

Type Description
tuple[StateType, ObsType, Float[Array, ''], Bool[Array, ''], Bool[Array, ''], dict]

A tuple of the next state, observation, reward, terminal flag, truncate flag, and additional info.

model_from_path staticmethod 

model_from_path(xml_file: str | Path) -> mujoco.MjModel