multilingual-e5-base-mar-16384

This model is a 64.53% smaller version of intfloat/multilingual-e5-base optimized for Marathi language via vocabulary size reduction using the trimming method.
This trimmed model should perform similarly to the original model with only 16,384 tokens and a much smaller memory footprint. However, it may not perform well for other languages as tokens not commonly used in the selected languages were removed from the vocabulary.

Model Statistics

Metric	Original	Trimmed	Reduction
Vocabulary size	250,037 tokens	16,384 tokens	93.44%
Model size	278,043,648 params	98,625,024 params	64.53%

Mining Dataset Statistics

• Number of texts used for mining: 200,000 texts
• Dataset: lbourdois/fineweb-2-trimming

Usage

from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("alphaedge-ai/multilingual-e5-base-mar-16384")
# Run inference with queries and documents
query = "My query in Marathi"
documents = [
    "Chunk in Marathi",
    "Chunk in Marathi",
    "Chunk in Marathi",
]
query_embeddings = model.encode_query(query)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# Compute similarities to determine a ranking
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)

Citations

Multilingual E5

@article{wang2024multilingual,
  title={Multilingual E5 Text Embeddings: A Technical Report},
  author={Wang, Liang and Yang, Nan and Huang, Xiaolong and Yang, Linjun and Majumder, Rangan and Wei, Furu},
  journal={arXiv preprint arXiv:2402.05672},
  year={2024}
}

Trimming blog post

@misc{hf_blogpost_trimming,
      title={Introduction to Trimming}, 
      author={Loïck BOURDOIS and Tom AARSEN and Bram VANROY and Christopher AKIKI and Woojun JUNG and Manuel ROMERO and Prithiv SAKTHI},
      year={2026},
      url={https://huggingface.co/blog/lbourdois/introduction-to-trimming}, 
}