Getting started
This page is designed to help you get started with Aloserving, a tool designed to configure Aloception on your robot.
Install
Make sure you follow the right installation (checkout the Installations section) steps from the documentation. We suggest you going with a Docker install for development. Then, run your docker based on the provided documentation.
Download configurations
Models configurations are necessary to be able to run Aloserving. Download the configurations below:
Models |
Value |
---|---|
bev_serving_baseline |
Candy-StereoP-DiNL-480x640-D-0.0.1-A-0.1.0-T-0.0.2 |
lux_bev_serving_baseline |
Candy-Lux-DiNL-480x640-D-0.0.1-A-0.0.2-T-0.0.3 |
Setup your project
This script will download your sample and setup your Aloserving configuration file.
python -m aloserving.tools.setup
You can also setup the script yourself into aloserving/aloserving_config.yaml
Here is an example:
data_folder: /data
weights_folder: /weights/
bev_serving_baseline: Candy-StereoP-DiNL-480x640-D-0.0.1-A-0.1.0-T-0.0.2
lux_bev_serving_baseline : Candy-Lux-DiNL-480x640-D-0.0.1-A-0.0.2-T-0.0.3
Note: If you’re using a Docker, the path is the volumes (docker -v) used to run the container.
If you just want to download the sample and edit your config file manually you can run:
python -m aloserving.tools.setup --sample_only
or you can also download the samples yourself and unzip the file into your data_folder
.
Run your first demo
Once the setup if finished you should be able to run Aloserving with the provided samples (specify your path to the downloaded fs_sample):
python -m aloserving.pipeline.pipeline_from_args --sensors fs_dataset --fs_base_folder PATH_TO/fs_sample/
You’re now ready to learn about the main Aloserving concepts or to run your first application!