Edit model card

MahaSent-PT

MahaSent-PT is a MahaBERT(l3cube-pune/marathi-bert-v2) model fine-tuned on L3Cube-MahaSent-PT Corpus, a political-domain Marathi sentiment analysis dataset.
This dataset was originally introduced in L3CubeMahaSent and is now part of extended corpus L3Cube-MahaSent-MD, a multi-domain Marathi sentiment analysis dataset.
The MahaSent-MD dataset contains domains like movie reviews, generic tweets, subtitles, and political tweets. This model is trained specifically on the political tweets domain.

The recommended multi-domain version of this model covering all domains is shared here: marathi-sentiment-md
[dataset link] (https://github.com/l3cube-pune/MarathiNLP)

More details on the dataset, models, and baseline results can be found in our [paper] (https://arxiv.org/abs/2306.13888)
Citing:

@article{pingle2023l3cube,
  title={L3Cube-MahaSent-MD: A Multi-domain Marathi Sentiment Analysis Dataset and Transformer Models},
  author={Pingle, Aabha and Vyawahare, Aditya and Joshi, Isha and Tangsali, Rahul and Joshi, Raviraj},
  journal={arXiv preprint arXiv:2306.13888},
  year={2023}
}
@inproceedings{kulkarni2021l3cubemahasent,
  title={L3CubeMahaSent: A Marathi Tweet-based Sentiment Analysis Dataset},
  author={Kulkarni, Atharva and Mandhane, Meet and Likhitkar, Manali and Kshirsagar, Gayatri and Joshi, Raviraj},
  booktitle={Proceedings of the Eleventh Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis},
  pages={213--220},
  year={2021}
}

Other Marathi Sentiment models from the MahaSent family are shared here:

MahaSent-MD (multi domain)
MahaSent-GT (generic tweets)
MahaSent-MR (movie reviews)
MahaSent-PT (political tweets)
MahaSent-ST (TV subtitles)
MahaSent v1 (political tweets)

Downloads last month
18
Safetensors
Model size
238M params
Tensor type
I64
·
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.