search

JulesBelveze / bert-squeeze

🛠️ Tools for Transformers compression using PyTorch Lightning ⚡
https://julesbelveze.github.io/bert-squeeze/
78 stars 10 forks source link

DistilAssistant data kwargs #39

Closed JulesBelveze closed 1 year ago

JulesBelveze commented 1 year ago

There are currently some caveats and unexpected behaviours when trying to pass data keywords arguments to the DistilAssistant.

One would expect the arguments to be applied to both the student and teacher modules.