Andrew Szot
Andrew Szot
In `imagination-augmented-agent.py` the imagination core does not correctly one hot encode the actions for the environment model as shown in the lines below: ``` if self.full_rollout: state = state.unsqueeze(0).repeat(self.num_actions, 1,...
## Motivation and Context Depends on [this PR](https://github.com/facebookresearch/habitat-lab/pull/883). Added: * Multi-agent support. To test multi-agent support in the interactive play script run: `python examples/interactive_play.py --cfg configs/tasks/rearrange/ma_rearrange.yam`. ## How Has This...
## Motivation and Context Task Changes: * Added challenge tasks: `rearrange_easy` and `rearrange`. * Refactored PDDL for complete support of [PDDL 1.2](https://planning.wiki/ref/pddl). The old code was difficult to maintain and...
# Setup Instructions * Download Fremont dataset, rename the folder to `fremont`, and place under `data/` in the base Habitat Lab directory. * Folders like `data/fremont/configs/scenes/` should exist. Generate the...
Evaluate policies trained in Gala in Habitat 2.0. Install with the most recent version of Habitat Sim (get the nightly conda install). Example of how to run: ``` python habitat_baselines/run.py...
## Motivation and Context Support for more state-based tasks in Habitat. Added following new environment IDs: * `HabitatGymCloseCabState-v0` * `HabitatGymOpenCabState-v0` * `HabitatGymPickState-v0` * `HabitatGymReachState-v0` Until the new data utility is...
# Setup Instructions * Download Fremont dataset, rename the folder to `fremont`, and place under `data/` in the base Habitat Lab directory. * Folders like `data/fremont/configs/scenes/` should exist. Generate the...
## Motivation and Context Fix the resume state not properly working on resuming a job from pre-emption. ## How Has This Been Tested Locally. ## Types of changes - **\[Bug...
## Motivation and Context For tracking SIRo project diff with main. ## How Has This Been Tested ## Types of changes - **\[Experiment\]** A pull request that add new features...