-
Notifications
You must be signed in to change notification settings - Fork 237
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
add poseval #195
add poseval #195
Conversation
The documentation is not available anymore as the PR was closed or merged. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Awesome possum! 🦝
metrics/poseval/README.md
Outdated
|
||
## Limitations and bias | ||
|
||
In contrast to seqeval the poseval metric treats each token independantly and computes the classification report overal all concatenated sequences. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
maybe a link to Seqeval here?
also, a few minor typos: "In contrast to seqeval, the poseval metric treats each token independently and computes the classification report over all concatenated sequences."
metrics/poseval/poseval.py
Outdated
@@ -0,0 +1,113 @@ | |||
# Copyright 2020 The HuggingFace Evaluate Authors. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
A bit unrelated, but should we just massively update all of our code from 2020 to 2022? 😂
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Once merged to main:
You should merge the main branch into this PR branch to incorporate the fix:
git checkout poseval
git fetch origin main
git merge origin/main
git push -u origin poseval
* add poseval * fix style * add suggestion from review
This PR adds the Poseval metric. Since Seqeval does not work well with data that is not in IOB format (see here) and since POS usually assigns one label per word with poseval each token is evaluated independently (as suggested here).
cc @fxmarty @lewtun since that's useful for the
evalautor
andEval on the hub
.