This repository has been archived by the owner on Sep 13, 2023. It is now read-only.

Inferencing types at saving is not feasible sometimes #691

Open

aguschin opened this issue Jul 11, 2023 · 0 comments

Labels

feedback serialization

Contributor

aguschin commented Jul 11, 2023

The type inference is nice. But it requires running the model on the machine packaging the model. This is a problem if I want to package a model requiring a GPU, on a machine which does not have a GPU (e.g. in CI/CD), or with a GPU too small to run the model with the parameters we want to have in production.

aguschin added serialization feedback labels

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.