Skip to content

Tutorial: Inference with a pre-trained model#

In this tutorial we will run inference on a pre-trained model for the PointNav task in the RoboTHOR environment. In this task the agent is tasked with going to a specific location within a realistic 3D environment.

For information on how to train a PointNav Model see this tutorial

We will need to install the RoboTHOR environment and download the RoboTHOR Pointnav dataset before we get started.

For this tutorial we will download the weights of a model trained on the debug dataset. This can be done with a handy script in the pretrained_model_ckpts directory:

bash pretrained_model_ckpts/ robothor-pointnav-rgb-resnet

This will download the weights for an RGB model that has been trained on the PointNav task in RoboTHOR to pretrained_model_ckpts/robothor-pointnav-rgb-resnet

Next we need to run the inference, using the PointNav experiment config from the tutorial on making a PointNav experiment. We can do this with the following command:


Where <PATH_TO_OUTPUT> is the location where the results of the test will be dumped, <PATH_TO_CHECKPOINT> is the location of the downloaded model weights, <BASE_DIRECTORY_OF_YOUR_EXPERIMENT> is a path to the directory where our experiment definition is stored, and <TIMESTAMP> is the unique timestamp associated with when the model was trained.

For our current setup the following command would work:

python \
    -o pretrained_model_ckpts/robothor-pointnav-rgb-resnet/ \
    -c pretrained_model_ckpts/robothor-pointnav-rgb-resnet/checkpoints/PointNavRobothorRGBPPO/2020-08-31_10-20-29/ \
    -b projects/tutorials pointnav_robothor_rgb_ddppo \
    -t 2020-08-31_10-20-29

For testing on all saved checkpoints we just need to omit <PATH_TO_CHECKPOINT>:

python \
    -o pretrained_model_ckpts/robothor-pointnav-rgb-resnet/ \
    -b projects/tutorials pointnav_robothor_rgb_ddppo \
    -t 2020-08-31_10-20-29


We also show examples of visualizations that can be extracted from the "valid" and "test" modes. Currently, visualization is still undergoing design changes and does not support multi-agent tasks, but the available functionality is sufficient for pointnav in RoboThor.

Following up on the example above, we can make a specialized pontnav ExperimentConfig where we instantiate the base visualization class, VizSuite, defined in utils.viz_utils, when in test mode.

Each visualization type can be thought of as a plugin to the base VizSuite. For example, all episode_ids passed to VizSuite will be processed with each of the instantiated visualization types (possibly with the exception of the AgentViewViz). In the example below we show how to instantiate different visualization types from 4 different data sources.

The data sources available to VizSuite are:

  • Task output (e.g. 2D trajectories)
  • Vector task (e.g. egocentric views)
  • Rollout storage (e.g. recurrent memory, taken action logprobs...)
  • ActorCriticOutput (e.g. action probabilities)

The visualization types included below are:

  • TrajectoryViz: Generic 2D trajectory view.
  • AgentViewViz: RGB egocentric view.
  • ActorViz: Action probabilities from ActorCriticOutput[CategoricalDistr].
  • TensorViz1D: Evolution of a point from RolloutStorage over time.
  • TensorViz2D: Evolution of a vector from RolloutStorage over time.
  • ThorViz: Specialized 2D trajectory view for RoboThor.

Note that we need to explicitly set the episode_ids that we wish to visualize. For AgentViewViz we have the option of using a different (typically shorter) list of episodes or enforce the ones used for the rest of visualizations.

class PointNavRoboThorRGBPPOVizExperimentConfig(

    viz_ep_ids = [
    viz_video_ids = [["FloorPlan_Train1_1_7"], ["FloorPlan_Train1_1_11"]]

    viz: Optional[VizSuite] = None

    def get_viz(self, mode):
        if self.viz is not None:
            return self.viz

        self.viz = VizSuite(
            # Basic 2D trajectory visualizer (task output source):
                path_to_target_location=("task_info", "target",),
            # Egocentric view visualizer (vector task source):
                max_video_length=100, episode_ids=self.viz_video_ids
            # Default action probability visualizer (actor critic output source):
            action_probs=ActorViz(figsize=(3.25, 10), fontsize=18),
            # Default taken action logprob visualizer (rollout storage source):
            # Same episode mask visualizer (rollout storage source):
            # Default recurrent memory visualizer (rollout storage source):
            # Specialized 2D trajectory visualizer (task output source):
                figsize=(16, 8),
                viz_rows_cols=(448, 448),
                scenes=("FloorPlan_Train{}_{}", 1, 1, 1, 1),

        return self.viz

    def machine_params(self, mode="train", **kwargs):
        res = super().machine_params(mode, **kwargs)
        res["visualizer"] = None
        if mode == "test":
            res["visualizer"] = self.get_viz(mode)

        return res

Running test on the same downloaded models, but using the visualization-enabled ExperimentConfig with

python \
    -o pretrained_model_ckpts/robothor-pointnav-rgb-resnet/ \
    -c pretrained_model_ckpts/robothor-pointnav-rgb-resnet/checkpoints/PointNavRobothorRGBPPO/2020-08-31_10-20-29/ \
    -b projects/tutorials pointnav_robothor_rgb_ddppo_viz \
    -t 2020-08-31_10-20-29

generates different types of visualization and logs them in tensorboard. If everything is properly setup and tensorboard includes the robothor-pointnav-rgb-resnet folder, under the IMAGES tab, we should see something similar to

Visualization example