File size: 4,576 Bytes
3067ef1 5a14c4a 3067ef1 7aa114f 3067ef1 ec69480 3067ef1 5a14c4a 3067ef1 5a14c4a 3067ef1 5a14c4a 3067ef1 5a14c4a 3067ef1 5a14c4a 3067ef1 5a14c4a 3067ef1 5a14c4a 3067ef1 5a14c4a 3067ef1 5a14c4a 3067ef1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 |
---
---
language:
- da
tags:
- text-classification
- pytorch
metrics:
- accuracy
- f1-score
---
# xlm-roberta-large-danish-parlspeech-cap-v3
## Model description
An `xlm-roberta-large` model fine-tuned on danish training data containing parliamentary speeches (oral questions, interpellations, bill debates, other plenary speeches, urgent questions) labeled with [major topic codes](https://www.comparativeagendas.net/pages/master-codebook) from the [Comparative Agendas Project](https://www.comparativeagendas.net/).
## How to use the model
This snippet prints the three most probable labels and their corresponding softmax scores:
```python
import torch
from transformers import AutoModelForSequenceClassification, AutoTokenizer
model = AutoModelForSequenceClassification.from_pretrained("poltextlab/xlm-roberta-large-danish-parlspeech-cap-v3")
tokenizer = AutoTokenizer.from_pretrained("xlm-roberta-large")
sentence = "This is an example."
inputs = tokenizer(sentence,
return_tensors="pt",
max_length=512,
padding="do_not_pad",
truncation=True
)
logits = model(**inputs).logits
probs = torch.softmax(logits, dim=1).tolist()[0]
probs = {model.config.id2label[index]: round(probability, 2) for index, probability in enumerate(probs)}
top3_probs = dict(sorted(probs.items(), key=lambda item: item[1], reverse=True)[:3])
print(top3_probs)
```
## Model performance
The model was evaluated on a test set of 44159 examples.<br>
Model accuracy is **0.94**.
| label | precision | recall | f1-score | support |
|:-------------|------------:|---------:|-----------:|----------:|
| 0 | 0.91 | 0.92 | 0.92 | 2310 |
| 1 | 0.9 | 0.9 | 0.9 | 1285 |
| 2 | 0.98 | 0.96 | 0.97 | 3400 |
| 3 | 0.95 | 0.95 | 0.95 | 1972 |
| 4 | 0.92 | 0.93 | 0.93 | 2679 |
| 5 | 0.96 | 0.96 | 0.96 | 2778 |
| 6 | 0.94 | 0.94 | 0.94 | 2458 |
| 7 | 0.96 | 0.94 | 0.95 | 1173 |
| 8 | 0.95 | 0.96 | 0.96 | 1948 |
| 9 | 0.95 | 0.97 | 0.96 | 3276 |
| 10 | 0.94 | 0.95 | 0.94 | 3224 |
| 11 | 0.92 | 0.93 | 0.93 | 2270 |
| 12 | 0.94 | 0.93 | 0.93 | 1510 |
| 13 | 0.89 | 0.89 | 0.89 | 1759 |
| 14 | 0.96 | 0.95 | 0.95 | 1941 |
| 15 | 0.95 | 0.93 | 0.94 | 1343 |
| 16 | 0.89 | 0.9 | 0.9 | 402 |
| 17 | 0.95 | 0.94 | 0.95 | 3337 |
| 18 | 0.92 | 0.92 | 0.92 | 3484 |
| 19 | 0.95 | 0.95 | 0.95 | 834 |
| 20 | 0.93 | 0.91 | 0.92 | 776 |
| macro avg | 0.94 | 0.93 | 0.94 | 44159 |
| weighted avg | 0.94 | 0.94 | 0.94 | 44159 |
### Fine-tuning procedure
This model was fine-tuned with the following key hyperparameters:
- **Number of Training Epochs**: 10
- **Batch Size**: 8
- **Learning Rate**: 5e-06
- **Early Stopping**: enabled with a patience of 2 epochs
## Inference platform
This model is used by the [CAP Babel Machine](https://babel.poltextlab.com), 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](https://babel.poltextlab.com).
## Reference
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
## Debugging and issues
This architecture uses the `sentencepiece` tokenizer. In order to use 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.
|