We do highly appreciate your effort in helping us test our releases! Here we provide a step-by-step protocol on how to proceed with the test process.

In the testing process, please report the bugs, suggestions or any other experience issues to GitHub issues. If you do not have one, you can also use the alternative Google Spreadsheet.

Pre-requisites

You must have Docker installed. Please check Get Docker for how to install Docker on your machine.

You can either use the edge build of AID from Installation or get a internal build from the source code directly.

Testing Steps

Find Models

We have an experimental new Model Hub, please try it out!

Build Models

Probably you will choose ocr model, after installation as described on the Model Hub, you need to start to build it by the following command:

aid build aidmodels/ocr/ocrSolver

Then the system will let you know the id of the result, and you will need to bind it to a local port:

aid create {YOUR_ID_HERE} {YOUR_PORT_HERE}

For example:

aid create ef6782b0 8081

Then AID will show the new id of the service. Please note that it will be different from the id before. You can start the service by:

aid start {YOUR_ID_HERE}

Some of the solver names are listed below:

• aidmodels/ocr/ocrSolver
• aidmodels/detectron/detection
• eth-library-lab/speech-analysis/speechSolver

Testing Predicion with Curl

curl -X POST -F [email protected] http://127.0.0.1:{YOUR_PORT_HERE}/infer

Remove Installed Repositories

rm -rf ~/.autoai/aid

The above command will uninstall the repositories on your machine, but will not remove the docker images. If you want to remove the docker images, you can use the following command:

docker ps -a # Found all containers

docker stop {container_id} # Stop the container

docker rm {container_id} # Remove the container

docker images # Found all images

docker rmi {image_id} # Remove the image

By removing the contents, you can reset your system to the initial state, and make the environment clean enough for next test/other tasks.