Tamil Sentence Bert Nli

🚀 TamilSBERT

TamilSBERT is a sentence similarity model. It is a TamilBERT model (l3cube - pune/tamil - bert) trained on the NLI dataset. This model is released as a part of project MahaNLP. A multilingual version supporting major Indic languages and cross - lingual capabilities is also available.

📋 Metadata

🎛️ Widget Examples

Example 1

• Source Sentence: "மக்கள் குழு பாடுகிறது"
• Comparison Sentences:
  • "சிலர் பாடுகிறார்கள்"
  • "ஒரு இளைஞன் பியானோ பாடுகிறான்"
  • "மனிதன் ஒரு கடிதம் எழுதுகிறான்"

Example 2

• Source Sentence: "நாய் பொம்மையை குரைக்கிறது"
• Comparison Sentences:
  • "ஒரு நாய் ஒரு பொம்மையில் குரைக்கிறது"
  • "ஒரு பூனை பால் குடிக்கிறது"
  • "ஒரு நாய் ஒரு பந்தைத் துரத்துகிறது"

Example 3

• Source Sentence: "நான் முதல் முறையாக விமானத்தில் அமர்ந்தேன்"
• Comparison Sentences:
  • "அது எனது முதல் விமானப் பயணம் "
  • "முதல் முறையாக ரயிலில் அமர்ந்தேன்"
  • "புதிய இடங்களுக்கு பயணம் செய்வது எனக்கு மிகவும் பிடிக்கும்"

🚀 Quick Start

This model can be used in two ways, with or without the sentence - transformers library.

📦 Installation

If you want to use the sentence - transformers library, you need to install it first:

pip install -U sentence-transformers

💻 Usage Examples

Basic Usage with Sentence - Transformers

from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]

model = SentenceTransformer('{MODEL_NAME}')
embeddings = model.encode(sentences)
print(embeddings)

Advanced Usage without Sentence - Transformers

from transformers import AutoTokenizer, AutoModel
import torch


#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)


# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']

# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('{MODEL_NAME}')
model = AutoModel.from_pretrained('{MODEL_NAME}')

# Tokenize sentences
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')

# Compute token embeddings
with torch.no_grad():
    model_output = model(**encoded_input)

# Perform pooling. In this case, mean pooling.
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])

print("Sentence embeddings:")
print(sentence_embeddings)

📚 Additional Information

This model is part of the project MahaNLP: https://github.com/l3cube - pune/MarathiNLP.

• A multilingual version of this model supporting major Indic languages and cross - lingual capabilities is available at indic - sentence - bert - nli .
• A better sentence similarity model (fine - tuned version of this model) is available at: https://huggingface.co/l3cube - pune/tamil - sentence - similarity - sbert.

📄 References

More details on the dataset, models, and baseline results can be found in our [paper] (https://arxiv.org/abs/2304.11434).

@article{deode2023l3cube,
  title={L3Cube-IndicSBERT: A simple approach for learning cross-lingual sentence representations using multilingual BERT},
  author={Deode, Samruddhi and Gadre, Janhavi and Kajale, Aditi and Joshi, Ananya and Joshi, Raviraj},
  journal={arXiv preprint arXiv:2304.11434},
  year={2023}
}

@article{joshi2022l3cubemahasbert,
  title={L3Cube-MahaSBERT and HindSBERT: Sentence BERT Models and Benchmarking BERT Sentence Representations for Hindi and Marathi},
  author={Joshi, Ananya and Kajale, Aditi and Gadre, Janhavi and Deode, Samruddhi and Joshi, Raviraj},
  journal={arXiv preprint arXiv:2211.11187},
  year={2022}
}

🔗 Related Models

• Monolingual Indic sentence BERT models:

  • Marathi SBERT
  • Hindi SBERT
  • Kannada SBERT
  • Telugu SBERT
  • Malayalam SBERT
  • Tamil SBERT
  • Gujarati SBERT
  • Oriya SBERT
  • Bengali SBERT
  • Punjabi SBERT
  • Indic SBERT (multilingual)

• Monolingual similarity models:

  • Marathi Similarity
  • Hindi Similarity
  • Kannada Similarity
  • Telugu Similarity
  • Malayalam Similarity
  • Tamil Similarity
  • Gujarati Similarity
  • Oriya Similarity
  • Bengali Similarity
  • Punjabi Similarity
  • Indic Similarity (multilingual)

📄 License

This model is released under the cc - by - 4.0 license.