Xlm Roberta Large Pooled Cap Media Minor

🚀 xlm-roberta-large-pooled-cap-media-minor

An xlm-roberta-large model finetuned for multilingual text classification with specific topic codes.

🚀 Quick Start

To use this model, you can follow the code example below. It demonstrates how to load the model and perform text classification.

from transformers import AutoTokenizer, pipeline

tokenizer = AutoTokenizer.from_pretrained("xlm-roberta-large")
pipe = pipeline(
    model="poltextlab/xlm-roberta-large-pooled-cap-media-minor",
    task="text-classification",
    tokenizer=tokenizer,
    use_fast=False,
    truncation=True,
    max_length=512,
    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)

✨ Features

• Multilingual Support: Finetuned on multilingual (English, Danish) training data.
• Specific Topic Codes: Labelled with minor topic codes from the Comparative Agendas Project and additional 7 media codes.

💻 Usage Examples

Basic Usage

The following code shows the basic way to use the model for text classification:

from transformers import AutoTokenizer, pipeline

tokenizer = AutoTokenizer.from_pretrained("xlm-roberta-large")
pipe = pipeline(
    model="poltextlab/xlm-roberta-large-pooled-cap-media-minor",
    task="text-classification",
    tokenizer=tokenizer,
    use_fast=False,
    truncation=True,
    max_length=512,
    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

⚠️ Important Note

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.

📚 Documentation

Model description

An xlm-roberta-large model finetuned on multilingual (English, Danish) training data labelled with minor topic codes from the Comparative Agendas Project. Furthermore, we also added used the following 7 media codes:

• State and Local Government Administration (24)
• Weather and Natural Disaster (26)
• Fires(27)
• Sports and Recreation (29)
• Death Notices (30)
• Churches and Religion (31)
• Other, Miscellaneous and Human Interest (99)

Model performance

The model was evaluated on a test set of 91 331 examples.

• Weighted Average F1-score: 0.68

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

💡 Usage Tip

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.

📄 License

This model is released under the MIT license.