File size: 4,503 Bytes
2930bce 4e299c6 2930bce 4e299c6 2930bce 4e299c6 2930bce 4e299c6 2930bce 4e299c6 2930bce 4e299c6 2930bce 4e299c6 2930bce 4e299c6 2930bce 4e299c6 2930bce 7e91909 2930bce 7e91909 2930bce 7e91909 2930bce 7e91909 2930bce 4e299c6 2930bce | 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 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 | ---
license: cc-by-sa-4.0
language:
- te
tags:
- dependency-parsing
- combo
- universal-dependencies
datasets:
- universal_dependencies
model-name: Combo Nlp Xlm Roberta Base Telugu Mtg Ud2.17
pipeline_tag: token-classification
---
# COMBO-NLP Model for Telugu
## Model Description
This is a Telugu-language model based on [combo-nlp](https://pypi.org/project/combo-nlp/), an open-source natural language preprocessing system. It performs:
- sentence segmentation (via [combo-seg](https://pypi.org/project/combo-seg/))
- tokenisation (via [combo-seg](https://pypi.org/project/combo-seg/))
- part-of-speech tagging
- morphological analysis
- lemmatisation
- dependency parsing
The Telugu model uses ``FacebookAI/xlm-roberta-base`` as its base encoder and is trained on [UD_Telugu-MTG](https://github.com/UniversalDependencies/UD_Telugu-MTG) (UD v2.17).
## Evaluation
Evaluation was performed on the UD_Telugu-MTG test split using the standard [CoNLL 2018 eval script](https://universaldependencies.org/conll18/conll18_ud_eval.py).
Two evaluation rows are reported:
- **Full-text (F1)**: raw text is segmented by [combo-seg](https://pypi.org/project/combo-seg/), then parsed and compared against gold — measures end-to-end pipeline performance including segmentation quality.
- **Aligned accuracy**: accuracy on correctly segmented (aligned) tokens — measures parsing quality on tokens that were correctly identified by the segmenter.
### Morphosyntactic Tagging
| Metric | Tokens | Sentences | Words | UPOS | XPOS | UFeats | AllTags | Lemmas |
| ------ | ------ | --------- | ----- | ---- | ---- | ------ | ------- | ------ |
| Full-text (F1) | 99.79 | 97.58 | 99.79 | 94.80 | 94.80 | 98.96 | 94.53 | 99.79 |
| Aligned accuracy | 0.00 | 0.00 | 0.00 | 95.00 | 95.00 | 99.17 | 94.72 | 100.00 |
### Dependency Parsing
| Metric | UAS | LAS | CLAS | MLAS | BLEX |
| ------ | --- | --- | ---- | ---- | ---- |
| Full-text (F1) | 90.64 | 83.16 | 79.43 | 76.19 | 79.43 |
| Aligned accuracy | 90.83 | 83.33 | 79.58 | 76.34 | 79.58 |
## Usage
Install the library from PyPI (assuming you have a virtual environment created):
```bash
pip install combo-nlp
```
The [combo-seg](https://pypi.org/project/combo-seg/) segmenter (used to split and
tokenise raw text) is installed automatically as a dependency of `combo-nlp`, so
no extra install step is needed.
```python
from combo import COMBO
# Load a pre-trained model with the corresponding combo-seg segmenter
nlp = COMBO("Telugu")
# Parse raw text (handles sentence splitting + tokenization)
result = nlp("వేగవంతమైన గోధుమ రంగు నక్క సోమరి కుక్క మీదుగా దూకుతుంది.")
# Inspect results
for sentence in result:
for token in sentence:
print(f"{token.form:<15} {token.lemma:<15} {token.upos:<8} head={token.head} {token.deprel}")
```
Refer to the combo-nlp documentation for installation and usage instructions:
- [https://pypi.org/project/combo-nlp/](https://pypi.org/project/combo-nlp/)
- [https://pypi.org/project/combo-seg/](https://pypi.org/project/combo-seg/)
## License
The training data license: cc-by-sa-4.0 is derived from the Universal Dependencies treebank. For the full license terms of each treebank, please refer to the corresponding `LICENSE.txt` file in the treebank repository:
- [UD_Telugu-MTG LICENSE.txt](https://github.com/UniversalDependencies/UD_Telugu-MTG/blob/master/LICENSE.txt)
## Citation
If you use this model, please cite:
Ulewicz, M., Jabłońska, M., Klimaszewski, M., Przybyła, P., Pszenny, Ł., Rybak, P., Wiącek, M., & Wróblewska, A. (2026). *COMBO-NLP Models Trained on UD v2.17*. Zenodo. https://doi.org/10.5281/zenodo.19650523
```bibtex
@software{combo_nlp_2026,
author = {Ulewicz, Michał and Jabłońska, Maja and Klimaszewski, Mateusz and Przybyła, Piotr and Pszenny, Łukasz and Rybak, Piotr and Wiącek, Martyna and Wróblewska, Alina},
title = {{COMBO-NLP} Models Trained on {UD} v2.17},
year = {2026},
publisher = {Zenodo},
doi = {10.5281/zenodo.19650523},
url = {https://doi.org/10.5281/zenodo.19650523}
}
```
## Resources
- combo-nlp: [https://pypi.org/project/combo-nlp/](https://pypi.org/project/combo-nlp/)
- combo-seg: [https://pypi.org/project/combo-seg/](https://pypi.org/project/combo-seg/)
- UD_Telugu-MTG: [https://github.com/UniversalDependencies/UD_Telugu-MTG](https://github.com/UniversalDependencies/UD_Telugu-MTG)
|