You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The python backend currently only supports CLS pooling
Motivation
I am currently deploying on intel gaudi hardware, and their fork of TEI currently only has support for the python backend. The model that I use has half the error rate when using mean pooling. Rather than only contributing to their fork I thought I could submit the update upstream.
Your contribution
I have made code changes and will publish a PR right away when I am done writing this.
The text was updated successfully, but these errors were encountered:
Feature request
The python backend currently only supports CLS pooling
Motivation
I am currently deploying on intel gaudi hardware, and their fork of TEI currently only has support for the python backend. The model that I use has half the error rate when using mean pooling. Rather than only contributing to their fork I thought I could submit the update upstream.
Your contribution
I have made code changes and will publish a PR right away when I am done writing this.
The text was updated successfully, but these errors were encountered: