tvalentyn commented on PR #17470: URL: https://github.com/apache/beam/pull/17470#issuecomment-1148364215
I took a closer look today. It seems that this PR combines 3 different efforts: 1) Support for batched keyed examples. The key-batched values are passed to pytorch model via `kwargs` in model() call. 2) Support for non-batchable `kwargs` params. 3) Add an interpreted parameter `prediction_params`. There is collusion of `kwargs` meaning in 1 and 2. We could explain in the PR description that these are independent efforts (if my understanding is correct). While I can squash all the changes into one commit before merge, for those looking up the commit history later, it may be easier to follow if efforts 1-3 were separate commits in the repo. Do you want to squash commits accordingly? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
