Skip to content
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

low test accuracy in Sentiment classification with LEAF's Sent140 tutorial? #34

Open
ghaccount opened this issue Feb 6, 2022 · 0 comments

Comments

@ghaccount
Copy link
Contributor

❓ Questions and Help

Until we move the questions to another medium, feel free to use this as your question:

Question

I tried this tutorial https://github.com/facebookresearch/FLSim/blob/main/tutorials/sent140_tutorial.ipynb
And accuracy is less that random guess (50%)!

Any suggestions or approaches to improve accuracy for this tutorial?

from tutorial:
Running (epoch = 1, round = 1, global round = 1) for Test
(epoch = 1, round = 1, global round = 1), Loss/Test: 0.8683878255035598
(epoch = 1, round = 1, global round = 1), Accuracy/Test: 49.61439588688946
{'Accuracy': 49.61439588688946}

@JohnlNguyen JohnlNguyen reopened this Feb 10, 2022
JohnlNguyen added a commit to JohnlNguyen/FLSim that referenced this issue Feb 10, 2022
Summary:
## What?
Fix tutorial to word embedding to resolve the poor accuracy problem

## Why?
facebookresearch#34

Differential Revision: D34149392

fbshipit-source-id: 62f0713b8eb898c4a3175940be6f82ecee2a946b
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants