---
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://gitlab.clarin-pl.eu/syntactic-tools/combo-nlp), an open-source natural language preprocessing system. It performs:

- sentence segmentation (via [LAMBO](https://gitlab.clarin-pl.eu/syntactic-tools/lambo))
- tokenisation (via [LAMBO](https://gitlab.clarin-pl.eu/syntactic-tools/lambo))
- 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 [LAMBO](https://gitlab.clarin-pl.eu/syntactic-tools/lambo), 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 | 96.60 | 99.79 | 94.94 | 94.80 | 98.96 | 94.66 | 99.79 |
| Aligned accuracy | 0.00 | 0.00 | 0.00 | 95.14 | 95.00 | 99.17 | 94.86 | 100.00 |

### Dependency Parsing

| Metric | UAS | LAS | CLAS | MLAS | BLEX |
| ------ | --- | --- | ---- | ---- | ---- |
| Full-text (F1) | 91.61 | 84.13 | 80.76 | 77.33 | 80.76 |
| Aligned accuracy | 91.81 | 84.31 | 80.92 | 77.48 | 80.92 |


## Usage

Install the library from PyPI (assuming you have a virtual environment created):

```bash
pip install combo-nlp
```

Install the Lambo segmenter - only needed when passing raw text strings to COMBO:

```bash
pip install --index-url https://pypi.clarin-pl.eu/ lambo
```

```python
from combo import COMBO

# Load a pre-trained model with corresponding Lambo 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://gitlab.clarin-pl.eu/syntactic-tools/combo-nlp](https://gitlab.clarin-pl.eu/syntactic-tools/combo-nlp)
- [https://gitlab.clarin-pl.eu/syntactic-tools/lambo](https://gitlab.clarin-pl.eu/syntactic-tools/lambo)

## Citation

<!--
```bibtex
@misc{combo_nlp_2026,
  title = {COMBO-NLP Models Trained on UD v2.17},
  author = {Ulewicz, Michał and Jabłońska, Maja and Klimaszewski, Mateusz and Przybyła, Piotr and Pszenny, Łukasz and Rybak, Piotr and Wiącek, Martyna and Wróblewska, Alina},
  year = {2026},
  howpublished = {\url{https://huggingface.co/collections/clarin-pl/combo-ud-217-models}}
}
```
-->

## Resources

- COMBO-NLP: [https://gitlab.clarin-pl.eu/syntactic-tools/combo-nlp](https://gitlab.clarin-pl.eu/syntactic-tools/combo-nlp)
- LAMBO: [https://gitlab.clarin-pl.eu/syntactic-tools/lambo](https://gitlab.clarin-pl.eu/syntactic-tools/lambo)
- UD_Telugu-MTG: [https://github.com/UniversalDependencies/UD_Telugu-MTG](https://github.com/UniversalDependencies/UD_Telugu-MTG)