Merge branch 'paper-tests'

Dockerfile

@@ -34,4 +34,4 @@ RUN pip install jupyter notebook
 RUN pip install omegaconf scipy shapely
 RUN pip install invertedai
 RUN pip install torchdrivesim
-RUN pip install "torchdriveenv[baselines] @ git+https://github.com/inverted-ai/torchdriveenv.git"
+RUN pip install torchdriveenv[baselines]

README.md

@@ -6,34 +6,27 @@ The basic installation of torchdriveenv uses an OpenCV renderer, which is slower
 
 To install the “torchdriveenv” with opencv rendering:
 ```
-python3 -m venv $PHTHON_VIRTUAL_ENV_PATH
-source .venv/bin/activate
-pip install "torchdriveenv[baselines] @ git+https://github.com/inverted-ai/torchdriveenv.git"
+pip install torchdriveenv
 ```
 
 To run examples:
 Set the `$IAI_API_KEY` and `$WANDB_API_KEY`
 ```
-git clone git@github.com:inverted-ai/torchdriveenv.git
-cd torchdriveenv
+pip install torchdriveenv[baselines]
 cd examples
-source $PHTHON_VIRTUAL_ENV_PATH/bin/activate
 python rl_training.py
 ```
 
 ## Pytorch3d rendering
 
 To install the “torchdriveenv” with Pytorch3d rendering:
 ```
-git clone git@github.com:inverted-ai/torchdriveenv.git
-cd torchdriveenv
 docker build --target torchdriveenv-first-release -t torchdriveenv-first-release:latest .
 ```
 
 To run examples:
 Set the `$IAI_API_KEY` and `$WANDB_API_KEY`
 ```
-cd torchdriveenv
 cd examples
 docker compose up rl-training
 ```

pyproject.toml

@@ -22,12 +22,12 @@ dependencies = [
     "scipy",
     "imageio",
     "torch>=1.10.1",
-    "invertedai",
+    "invertedai>=0.0.16",
     "omegaconf",
     "opencv-python",
     "gymnasium",
     "lanelet2",
-    "torchdrivesim",
+    "torchdrivesim>=0.2.1",
 ]
 dynamic = ["version",]
 

torchdriveenv/gym_env.py

@@ -180,7 +180,7 @@ def build_simulator(cfg: EnvConfig, map_cfg, device, ego_state, scenario=None, c
     with torch.no_grad():
         traffic_light_controller = map_cfg.traffic_light_controller
         initial_light_state_name = traffic_light_controller.current_state_with_name
-        traffic_light_ids = [stopline.actor_id for stopline in map_cfg.stoplines if stopline.agent_type == 'traffic-light']
+        traffic_light_ids = [stopline.actor_id for stopline in map_cfg.stoplines if stopline.agent_type == 'traffic_light']
         driving_surface_mesh = map_cfg.road_mesh