Best-practices for pre-trained models.

TEAMuP-dev / pyharp

Companion repository which facilitates the creation of Gradio endpoints which are accessible from within Digital Audio Workstations (DAWs) through HARP.

BSD 3-Clause "New" or "Revised" License

23 stars 3 forks source link

Best-practices for pre-trained models. #6

Closed cwitkowitz closed 12 months ago

cwitkowitz commented 1 year ago

It might be nice to mention best-practices for utilizing pre-trained models in cases where users plan to leverage models/algorithms that are not easily accessible through python packages.

Store within repository using Git Large File Storage
Where to initialize models/algorithms (i.e., outside of process_fn)

cwitkowitz commented 1 year ago

Added in PR https://github.com/audacitorch/pyharp/pull/10 ...

oreillyp commented 12 months ago

Closed after merge