Model Card for Model ID

Fine-tuned BERT trained to identify political party from speech in the UK Parliament. Trained with Labour/Conservative speeches using the hansard dataset 2000-2020

Model Details

  • License: MIT
  • Finetuned from model : BERT

Model Sources

Uses

Model can be used to predict whether an item of speech from the UK Parliament was said by a member of the Labour or Conservative Party.

Out-of-Scope Use

Model has only been exposed to UK data and speeches from Labour or Conservative. May not work as intended for other parties or geographies.

Training Data

biglam/hansard_speech

Preprocessing

[bert-cased]

Evaluation

Evaluated on held-out test data. Achieved test accuracy >80% and test loss <0.4 (binary cross-entropy).

Citation

BibTeX:

@misc{baly2020detect, title={We Can Detect Your Bias: Predicting the Political Ideology of News Articles}, author={Ramy Baly and Giovanni Da San Martino and James Glass and Preslav Nakov}, year={2020}, eprint={2010.05338}, archivePrefix={arXiv}, primaryClass={cs.CL} }

Downloads last month
6
Safetensors
Model size
108M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.