aphp / edsnlp

Modular, fast NLP framework, compatible with Pytorch and spaCy, offering tailored support for French clinical notes.
https://aphp.github.io/edsnlp/
BSD 3-Clause "New" or "Revised" License
111 stars 29 forks source link

Feature request: map_batches #273

Closed percevalw closed 5 months ago

percevalw commented 5 months ago

Feature type

Add a .map_batches method to lazy collections to allow filtering documents. ⚠️ We need to ensure that .map_batches is not called between two torch components that could share cache entries, otherwise the cache keys would be different between before and after .map_batches

percevalw commented 5 months ago

Added in #275