DeepMind Lab#

TorchWM supports DeepMind Lab through DMLabEnv, a lightweight adapter around the native deepmind_lab Python module. It converts Lab RGB observations into the channel-first image dictionary used by Dreamer and other pixel-based world-model code.

Installation#

DeepMind Lab includes native game-engine components and is distributed separately from TorchWM. Install or build a package that exposes the deepmind_lab Python module before using this adapter. One option is the external dmlab-gym build helper:

pip install dmlab-gym
dmlab-gym build

If you already have a working deepmind_lab installation, no additional runtime package is required by the adapter.

Dreamer configuration#

from world_models.configs.dreamer_config import DreamerConfig
from world_models.models.dreamer import DreamerAgent

cfg = DreamerConfig()
cfg.env_backend = "dmlab"
cfg.env = "rooms_collect_good_objects_train"
cfg.image_size = (64, 64)
cfg.dmlab_action_repeat = 4

agent = DreamerAgent(cfg)
agent.train()

DMLabEnv asks Lab for RGB_INTERLEAVED observations and returns them as obs["image"] with shape (3, H, W) and dtype uint8. Extra observation names can be requested with cfg.dmlab_observations; they are copied into the returned observation dictionary under their native DeepMind Lab names.

Direct usage#

from world_models.envs import make_dmlab_env

env = make_dmlab_env("rooms_collect_good_objects_train", seed=0, size=(64, 64))
obs = env.reset()
action = env.action_space.sample()
obs, reward, done, info = env.step(action)

The default action space is a normalized one-hot Box[-1, 1] over a compact set of navigation actions. You can pass action_set= to DMLabEnv or set cfg.dmlab_action_set to use a custom list of native seven-element Lab actions.

Backend options#

Option	Default	Description
`cfg.dmlab_action_repeat`	`4`	Native Lab frame repeat passed to `env.step(..., num_steps=...)`.
`cfg.dmlab_action_set`	`None`	Optional custom 2D array of native Lab actions.
`cfg.dmlab_observations`	`None`	Additional native observation names. `RGB_INTERLEAVED` is always included.
`cfg.dmlab_config`	`None`	Extra Lab config values. Width and height are derived from `cfg.image_size`.
`cfg.dmlab_renderer`	`"hardware"`	Renderer argument forwarded to `deepmind_lab.Lab`.

TorchWM’s shared Dreamer wrapper stack still applies cfg.action_repeat outside the DMLab adapter. If you only want native Lab frame repeat, leave cfg.action_repeat = 1.