multilingual-e5-small-optimized Open Source Model - Improved Multilingual Processing Performance While Maintaining High Quality

Multilingual E5 Small Optimized

Developed by elastic

This is the quantized version of multilingual-e5-small, optimized for inference performance through layer-wise quantization while retaining most of the original model's quality.

Text Embedding

PyTorch

Supports Multiple LanguagesOpen Source License:MIT #Multilingual Embedding #Quantization Optimization #Semantic Similarity

Downloads 201

Release Time : 11/16/2023

Model Overview

This model is a text embedding model generated through weakly supervised contrastive pre-training, supporting sentence similarity calculations in multiple languages.

Model Features

Multilingual Support

Supports sentence similarity calculations in over 100 languages.

Quantization Optimization

Optimizes the model through layer-wise quantization techniques to improve inference speed.

Performance Boost

Compared to the original model, inference speed is improved by 20-50%.

Model Capabilities

Multilingual sentence embedding

Sentence similarity calculation

Cross-language text matching

Use Cases

Information Retrieval

Multilingual Document Retrieval

Find the most relevant documents in a multilingual document collection based on a query.

Performs well on the MIRACL dataset.

Semantic Search

Cross-Language Search

Search for documents in one language using a query in another language.

🚀 Multilingual-e5-small-optimized

A quantized version of multilingual-e5-small for sentence similarity tasks.

This is a quantized version of multilingual-e5-small. The quantization was carried out layer by layer under the same conditions as our ELSERv2 model, as detailed here.

The model is based on the research paper Text Embeddings by Weakly-Supervised Contrastive Pre-training by Liang Wang, Nan Yang, Xiaolong Huang, Binxing Jiao, Linjun Yang, Daxin Jiang, Rangan Majumder, Furu Wei, arXiv 2022.

📚 Documentation

Model Information

Property	Details
Model Type	Quantized version of multilingual-e5-small for sentence similarity
Training Data	Not specified
Supported Languages	Multilingual, including af, am, ar, as, az, be, bg, bn, br, bs, ca, cs, cy, da, de, el, en, eo, es, et, eu, fa, fi, fr, fy, ga, gd, gl, gu, ha, he, hi, hr, hu, hy, id, is, it, ja, jv, ka, kk, km, kn, ko, ku, ky, la, lo, lt, lv, mg, mk, ml, mn, mr, ms, my, ne, nl, no, om, or, pa, pl, ps, pt, ro, ru, sa, sd, si, sk, sl, so, sq, sr, su, sv, sw, ta, te, th, tl, tr, ug, uk, ur, uz, vi, xh, yi, zh
License	MIT

Benchmarks

We conducted several small benchmarks to evaluate both the quality changes and inference latency compared to the baseline original model.

Quality

By measuring NDCG@10 using the dev split of the MIRACL datasets for selected languages, we found that the quantized model generally shows marginal quality changes.

	de	yo	ru	ar	es	th
multilingual-e5-small	0.75862	0.56193	0.80309	0.82778	0.81672	0.85072
multilingual-e5-small-optimized	0.75992	0.48934	0.79668	0.82017	0.8135	0.84316

To test the English out-of-domain performance, we used the test split of various datasets in the BEIR evaluation. Measuring NDCG@10, we observed a larger change in SCIFACT, but marginal changes in the other evaluated datasets.

	FIQA	SCIFACT	nfcorpus
multilingual-e5-small	0.33126	0.677	0.31004
multilingual-e5-small-optimized	0.31734	0.65484	0.30126

Performance

Using a PyTorch model traced for Linux and Intel CPUs, we performed performance benchmarking with various input lengths. Overall, the optimized model shows an average performance improvement of 50 - 20%.

input length (characters)	multilingual-e5-small	multilingual-e5-small-optimized	speedup
0 - 50	0.0181	0.00826	54.36%
50 - 100	0.0275	0.0164	40.36%
100 - 150	0.0366	0.0237	35.25%
150 - 200	0.0435	0.0301	30.80%
200 - 250	0.0514	0.0379	26.26%
250 - 300	0.0569	0.043	24.43%
300 - 350	0.0663	0.0513	22.62%
350 - 400	0.0737	0.0576	21.85%

Disclaimer

This e5 model, as defined, hosted, integrated, and used in conjunction with our other Elastic Software, is covered by our standard warranty.

📄 License

This project is licensed under the MIT license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご