Malayalam-Doc-Topic-BERT

Malayalam-Doc-Topic-BERT model is an IndicSBERT(l3cube-pune/malayalam-sentence-bert-nli) model fine-tuned on Malayalam documents from the L3Cube-IndicNews Corpus [dataset link]https://github.com/l3cube-pune/indic-nlp.
This dataset consists of sub-datasets like LDC (Long Document Classification), LPC (Long Paragraph Classification), and SHC (Short Headlines Classification), each having different document lengths.
This model is trained on a combination of all three variants and works well across different document sizes.

More details on the dataset, models, and baseline results can be found in our [paper]https://arxiv.org/abs/2401.02254

Citing:

@article{mirashi2024l3cube,
  title={L3Cube-IndicNews: News-based Short Text and Long Document Classification Datasets in Indic Languages},
  author={Mirashi, Aishwarya and Sonavane, Srushti and Lingayat, Purva and Padhiyar, Tejas and Joshi, Raviraj},
  journal={arXiv preprint arXiv:2401.02254},
  year={2024}
}

Other document topic models for different Indic languages are listed below:
Hindi-Doc-Topic-BERT
Marathi-Doc-Topic-BERT
Bengali-Doc-Topic-BERT
Telugu-Doc-Topic-BERT
Tamil-Doc-Topic-BERT
Gujarati-Doc-Topic-BERT
Kannada-Doc-Topic-BERT
Odia-Doc-Topic-BERT
Malayalam-Doc-Topic-BERT
Punjabi-Doc-Topic-BERT
English-Doc-Topic-BERT

Downloads last month
42
Safetensors
Model size
238M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.