python3.7

google-cloud-sdk

poetry

Clone the project.

In credentials/ you need two files: ideal-pancake-docker-id.json ideal-pancake-google-application-credentials.json

These are created by creating service accounts on cloud.google.com. Get in touch with us if you don't have the access you should have. You will need to rename the files that are generated appropriately.

poetry install

. ./credentials/env.sh

poetry run ./scripts/generate_features.sh [environment] [dataset] --download_files [bool] Environment has the possible values remote or local. Which decides if a job should be run locally or dispatched as a job to gcp dataset has the possible values ["emip", "jetris", "emip-images", "mooc-images", "cscw", "fractions"] download_files decides if the dataset should be downloaded or local files should be used

If the feature extraction is run remotely, the features will be saved to a bucket which the classifier can fetch

poetry run ./scripts/train_model.sh [environment] [in_study_dataset] [out_of_study_dataset]

in_study_dataset and out_of_study has the possible values ["emip", "jetris", "emip-images", "mooc-images", "cscw", "fractions"] This script fetches the premade features from gcs and trains on them

poetry run ./scripts/generate_features.sh remote emip --download_files True poetry run ./scripts/train_model.sh local emip cscw