Skip to content

Latest commit

 

History

History
75 lines (57 loc) · 5.53 KB

CONTRIBUTING.md

File metadata and controls

75 lines (57 loc) · 5.53 KB
Raw

Contributing to ASKCOSv2

We want to make contributing to ASKCOSv2 as easy and transparent as possible.

Contributing to an existing repo

  1. Fork the repo and create your branch from dev, or main if dev doesn't exist.
  2. If you've added code that should be tested, add tests (which we gently enforce).
  3. Ensure the test suite passes.
  4. For askcos2_core and askcos2_site only, make sure your code lints.
  5. Submit the merge request into the forked branch (dev or main).

Contributing a new module

Step 1: wrap your prediction code as a microservice

  • Option 1 (basic): using Flask/FastAPI

We will use the site_selectivity module as a reference. Assuming you have some existing code that can run a prediction task, for example, SiteSelectivityModel().predict(). The first step is wrapping this function as an API endpoint, with the framework of your choice (we recommend FastAPI). See the relevant code section in site_selectivity_server.py as an example. This turns your prediction code into a service that can be run with

python your_server.py

and can be queried once started.

  • Option 2 (advanced): using Torchserve

Serving with Torchserve requires a few more steps, but makes it trivially easy to add new models later on. We will use the augmented_transformer module as a reference. Assuming you have some existing code that can run a prediction task, for example, ServerModel().run(). The first step is wrapping this function in your_handler.py under YourHandler.inference(). See the relevant code section in at_handler.py as an example.

Once the handler has been implemented, you need to package your code, including the handler, with other required files (e.g., model checkpoints, data files) into a servable model archive using torch-model-archiver, similar to how it was done for augmented Transformer. Note that while we typically perform this archiving step in Docker, this is not required and you can also do it in a conda environment. If successful, you should be able to obtain a model archive ending with .mar, which can then be served by

torchserve \
  --start \
  --foreground \
  --ncs \
  --model-store=${FOLDER_CONTAINING_MARS} \
  --models \
  ${MODEL_NAME}=${MODEL_NAME}.mar

and can be queried once started.

Step 2: containerized your microservice in Docker

Containerizing this service in Docker is then straightforward if you already have some conda environment. Simply create a Dockerfile similar to the sample Dockerfiles for site_selectivity, or for augmented_transformer to install the dependencies. Then build the Docker image with

export ASKCOS_REGISTRY=registry.gitlab.com/mlpds_mit/askcosv2
docker build -f your_Dockerfile -t ${ASKCOS_REGISTRY}/your_module:1.0-cpu .

after which the containerized microservice should be runnable with a single command, e.g.,

docker run --rm -p 9601:9601 -t ${ASKCOS_REGISTRY}/your_module:1.0-cpu

Or in the case of lengthier command, organize it into a start script similar to scripts/serve_cpu_in_docker.sh for augmented_transformer, which can then be run by

sh scripts/serve_cpu_in_docker.sh

Step 3 (optional): prepare deployment.yaml

TODO

Step 4: contact us

Once Steps 1-3 have been completed, contact the ASKCOS team ([email protected]) to initiate the integration of your service into ASKCOSv2. What follows would mostly be bookkeeping, such as adding wrapper that calls your service at the API gateway (in askcos2_core). Since we may be shifting data (to our cloud storage), reassigning ports, and adding additional supports (e.g., singularity and async task management), we would prefer to handle these hassles for you.

Contributing a new model for an existing module, or contributing new data (e.g., reactions, buyables).

Please contact the ASKCOS team directly ([email protected]).

Issues

We use GitLab issues to track public bugs. Please ensure your description is clear and has sufficient instructions to be able to reproduce the issue.

License

By contributing to ASKCOSv2, you agree that your contributions will be licensed under the LICENSE file in the root directory of the repo, if there is one.