---
base_model: INSAIT-Institute/BgGPT-7B-Instruct-v0.2
library_name: peft
license: apache-2.0
language:
- bg
tags:
- propaganda
---

# Model Card for identrics/BG_propaganda_detector




## Model Description

- **Developed by:** [`Identrics`](https://identrics.ai/)
- **Language:** Bulgarian
- **License:** apache-2.0
- **Finetuned from model:** [`INSAIT-Institute/BgGPT-7B-Instruct-v0.2`](https://huggingface.co/INSAIT-Institute/BgGPT-7B-Instruct-v0.2)
- **Context window :** 8192 tokens

## Model Description

This model consists of a fine-tuned version of BgGPT-7B-Instruct-v0.2 for a propaganda detection task. It is effectively a multilabel classifier, determining wether a given propaganda text in Bulgarian contains or not 5 predefined propaganda types.
This model was created by [`Identrics`](https://identrics.ai/), in the scope of the WASPer project.


## Propaganda taxonomy

The propaganda techniques we want to identify are classified in 5 categories:

1. Self-Identification Techniques:
These techniques exploit the audience's feelings of association (or desire to be associated) with a larger group. They suggest that the audience should feel united, motivated, or threatened by the same factors that unite, motivate, or threaten that group.


2. Defamation Techniques:
These techniques represent direct or indirect attacks against an entity's reputation and worth.

3. Legitimisation Techniques:
These techniques attempt to prove and legitimise the propagandist's statements by using arguments that cannot be falsified because they are based on moral values or personal experiences.

4. Logical Fallacies:
These techniques appeal to the audience's reason and masquerade as objective and factual arguments, but in reality, they exploit distractions and flawed logic.

5. Rhetorical Devices:
These techniques seek to influence the audience and control the conversation by using linguistic methods.




## Uses

To be used as a multilabel classifier to identify if the Bulgarian sample text contains one or more of the five propaganda techniques mentioned above.

### Example

First install direct dependencies:
```
pip install transformers torch accelerate
```

Then the model can be downloaded and used for inference:
```py
from transformers import AutoModelForSequenceClassification, AutoTokenizer

model = AutoModelForSequenceClassification.from_pretrained("identrics/BG_propaganda_classifier", num_labels=5)
tokenizer = AutoTokenizer.from_pretrained("identrics/BG_propaganda_classifier")

tokens = tokenizer("Газа евтин, американското ядрено гориво евтино, пълно с фотоволтаици а пък тока с 30% нагоре. Защо ?", return_tensors="pt")
output = model(**tokens)
print(output.logits)
```


## Training Details


During the training stage, our objective is to train the multi-label classifier on different types of propaganda using a dataset that includes both real and artificially generated samples. In the case of Bulgarian, there are 206 organic examples and 206 synthetic examples.
The data is carefully classified by domain experts based on our predetermined taxonomy, which covers five primary classifications. Certain examples are classified under just one category, others have several different groups, highlighting the complex structure of propaganda, where multiple techniques can be found inside a single text.



The model was then tested on a smaller evaluation dataset, achieving an F1 score of 

## Citation

If you find our work useful, please consider citing WASPer:

```
@article{...2024wasper,
  title={WASPer: Propaganda Detection in Bulgarian and English}, 
  author={....},
  journal={arXiv preprint arXiv:...},
  year={2024}
}
```