You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Because PRs from repo forks don't have access to GitHub secrets (see GHA docs), PRs from forked repos won't be able to pass the CI currently. This is because the pr-check CI job currently includes a step of updating the cached Coda data for tests (i.e. npm run refresh-test-data) which involves making requests to Coda, which requires the Coda authorization token. This was experienced here.
This isse was discussed on Discord here.
Proposed solution options are:
Remove redownloading from GH PR check action and make a separate GH action for updating cached test data (only for manual run. This could be similar to this previous GH action that was used to commit new versions of data for semantic search
Host the test cache externally (so as to remove the cache remove being directly in the repo). This is what is currently done to fetch semantic search data (fetch('https://storage.googleapis.com/stampy-nlp-resources/stampy-encodings.json') in tfWorker.js)
The text was updated successfully, but these errors were encountered:
Because PRs from repo forks don't have access to GitHub secrets (see GHA docs), PRs from forked repos won't be able to pass the CI currently. This is because the
pr-check
CI job currently includes a step of updating the cached Coda data for tests (i.e.npm run refresh-test-data
) which involves making requests to Coda, which requires the Coda authorization token. This was experienced here.This isse was discussed on Discord here.
Proposed solution options are:
fetch('https://storage.googleapis.com/stampy-nlp-resources/stampy-encodings.json')
in tfWorker.js)The text was updated successfully, but these errors were encountered: