metadata

license: mit
language:
  - multilingual
tags:
  - zero-shot-classification
  - text-classification
  - pytorch
metrics:
  - accuracy
  - f1-score
extra_gated_prompt: >-
  Our models are intended for academic use only. If you are not affiliated with
  an academic institution, please provide a rationale for using our models.

  If you use our models for your work or research, please cite this paper:
  Sebők, M., Máté, Á., Ring, O., Kovács, V., & Lehoczki, R. (2024). Leveraging
  Open Large Language Models for Multilingual Policy Topic Classification: The
  Babel Machine Approach. Social Science Computer Review, 0(0).
  https://doi.org/10.1177/08944393241259434
extra_gated_fields:
  Name: text
  Country: country
  Institution: text
  E-mail: text
  Use case: text

xlm-roberta-large-english-media-cap-v3

Model description

An xlm-roberta-large model finetuned on multilingual training data containing texts of the media domain labelled with major topic codes from the Comparative Agendas Project.

How to use the model

from transformers import AutoTokenizer, pipeline

tokenizer = AutoTokenizer.from_pretrained("xlm-roberta-large")
pipe = pipeline(
    model="poltextlab/xlm-roberta-large-english-media-cap-v3",
    task="text-classification",
    tokenizer=tokenizer,
    use_fast=False,
    token="<your_hf_read_only_token>"
)

text = "We will place an immediate 6-month halt on the finance driven closure of beds and wards, and set up an independent audit of needs and facilities."
pipe(text)

Gated access

Due to the gated access, you must pass the token parameter when loading the model. In earlier versions of the Transformers package, you may need to use the use_auth_token parameter instead.

Model performance

The model was evaluated on a test set of 37345 examples (10% of the available data).
Model accuracy is 0.78.

label	precision	recall	f1-score	support
0	0.81	0.73	0.77	1705
1	0.76	0.58	0.66	1075
2	0.88	0.81	0.84	2167
3	0.8	0.73	0.77	473
4	0.67	0.64	0.66	801
5	0.83	0.89	0.86	2001
6	0.77	0.85	0.81	1274
7	0.84	0.83	0.84	1064
8	0.68	0.63	0.65	203
9	0.8	0.87	0.83	1796
10	0.79	0.75	0.77	3377
11	0.81	0.66	0.73	548
12	0.72	0.69	0.7	1007
13	0.81	0.72	0.76	2796
14	0.75	0.77	0.76	3138
15	0.78	0.67	0.72	759
16	0.79	0.67	0.72	515
17	0.75	0.84	0.8	4618
18	0.79	0.86	0.82	6475
19	0.7	0.55	0.61	559
20	0.67	0.69	0.68	981
21	0	0	0	13
macro avg	0.74	0.7	0.72	37345
weighted avg	0.78	0.78	0.78	37345

Inference platform

This model is used by the CAP Babel Machine, an open-source and free natural language processing tool, designed to simplify and speed up projects for comparative research.

Cooperation

Model performance can be significantly improved by extending our training sets. We appreciate every submission of CAP-coded corpora (of any domain and language) at poltextlab{at}poltextlab{dot}com or by using the CAP Babel Machine.

Debugging and issues

This architecture uses the sentencepiece tokenizer. In order to run the model before transformers==4.27 you need to install it manually.

If you encounter a RuntimeError when loading the model using the from_pretrained() method, adding ignore_mismatched_sizes=True should solve the issue.