kerasltisubmission

Build Status PyPI version License Documentation Status Test Coverage

This python package allows you to submit your trained keras models to a kerasltiprovider (see here) that proxies any LTI conforming learning platform for grading.

$ pip install kerasltisubmission
import kerasltisubmission

# Looking for the provider?
# See https://github.com/into-ai/kerasltiprovider
provider = LTIProvider(
    input_api_endpoint="http://localhost:8080",
    submission_api_endpoint="http://localhost:8080/submit",
    user_token="7dd7367c-40c2-43cb-a052-bb04e1d0a858",
)

# Submit your keras model
submission = Submission(assignment_id=12, model=model)
provider.submit(submission)

For a complete example, see example.py. Also see the official documentation as well as the documentation of the kerasltiprovider.

Development

For detailed instructions see CONTRIBUTING.

Tests

You can run tests with

$ invoke test
$ invoke test --min-coverage=90     # Fail when code coverage is below 90%
$ invoke type-check                 # Run mypy type checks

Linting and formatting

Lint and format the code with

$ invoke format
$ invoke lint

All of this happens when you run invoke pre-commit.

Note

This project is still in the alpha stage and should not be considered production ready.