vectorizer-v1-S-multilingual Open-source Multilingual Vectorizer - Free to Generate Vectors for Information Retrieval

Vectorizer V1 S Multilingual

Developed by sinequa

A multilingual vectorizer developed by Sinequa that generates embedding vectors for input paragraphs or queries, used for similarity calculation and information retrieval.

Text Embedding

Transformers

Supports Multiple Languages#Multilingual vectorization #Efficient retrieval #Low-latency inference

Downloads 322

Release Time : 7/10/2023

Model Overview

This model is a multilingual feature extractor based on the BERT-Small architecture, primarily used for sentence similarity calculation and information retrieval tasks, supporting four languages: English, French, German, and Spanish.

Model Features

Multilingual support

Supports text processing in four languages: English, French, German, and Spanish

Efficient inference

Demonstrates efficient inference speed across different GPUs, requiring only 5 milliseconds for batch processing of 32 samples under FP16 quantization

Case insensitivity

Insensitive to text case and accents, improving retrieval robustness

In-batch negative training

Uses query-passage pairs and in-batch negative samples for training to optimize vector representation

Model Capabilities

Text vectorization

Multilingual text processing

Semantic similarity calculation

Information retrieval

Use Cases

Information retrieval

Document retrieval system

Build a document retrieval system based on semantic similarity

Achieved an average Recall@100 of 0.448 on the BEIR benchmark

Multilingual Q&A system

Backend for a multilingual Q&A system

Achieved a French Recall@100 of 0.583 on the MIRACL benchmark

🚀 `vectorizer-v1-S-multilingual`

This model, developed by Sinequa, is a vectorizer that generates an embedding vector for a given passage or query. Passage vectors are stored in the vector index, and the query vector is used to find relevant passages in the index during query time.

🚀 Quick Start

This model is designed to generate embedding vectors for passages and queries. You can use it in your applications to perform tasks like sentence similarity and feature extraction.

✨ Features

Multilingual Support: The model supports English, French, German, and Spanish.
Efficient Inference: It offers different quantization types and batch sizes for efficient inference on various GPUs.

📦 Installation

The README does not provide installation steps, so this section is skipped.

💻 Usage Examples

The README does not provide code examples, so this section is skipped.

📚 Documentation

Supported Languages

The model was trained and tested in the following languages:

English
French
German
Spanish

Scores

Metric	Value
Relevance (Recall@100)	0.448

Note that the relevance score is computed as an average over 14 retrieval datasets (see details below).

Inference Times

GPU	Quantization type	Batch size 1	Batch size 32
NVIDIA A10	FP16	1 ms	5 ms
NVIDIA A10	FP32	3 ms	14 ms
NVIDIA T4	FP16	1 ms	12 ms
NVIDIA T4	FP32	2 ms	52 ms
NVIDIA L4	FP16	1 ms	5 ms
NVIDIA L4	FP32	2 ms	18 ms

GPU Memory Usage

Quantization type	Memory
FP16	300 MiB
FP32	600 MiB

Note that GPU memory usage only includes how much GPU memory the actual model consumes on an NVIDIA T4 GPU with a batch size of 32. It does not include the fix amount of memory that is consumed by the ONNX Runtime upon initialization which can be around 0.5 to 1 GiB depending on the used GPU.

Requirements

Minimal Sinequa version: 11.10.0
Minimal Sinequa version for using FP16 models and GPUs with CUDA compute capability of 8.9+ (like NVIDIA L4): 11.11.0
Cuda compute capability: above 5.0 (above 6.0 for FP16 use)

Model Details

Overview

Number of parameters: 39 million
Base language model: Homegrown Sinequa BERT - Small (Paper) pretrained in the four supported languages
Insensitive to casing and accents
Training procedure: Query - passage pairs using in - batch negatives

Training Data

Natural Questions (Paper, Official Page)
- Original English dataset
- Translated datasets for the other three supported languages

Evaluation Metrics

To determine the relevance score, we averaged the results that we obtained when evaluating on the datasets of the BEIR benchmark. Note that all these datasets are in English.

Dataset	Recall@100
Average	0.448

Arguana	0.835
CLIMATE - FEVER	0.350
DBPedia Entity	0.287
FEVER	0.645
FiQA - 2018	0.305
HotpotQA	0.396
MS MARCO	0.533
NFCorpus	0.162
NQ	0.701
Quora	0.947
SCIDOCS	0.194
SciFact	0.580
TREC - COVID	0.051
Webis - Touche - 2020	0.289

We evaluated the model on the datasets of the MIRACL benchmark to test its multilingual capacities. Note that not all training languages are part of the benchmark, so we only report the metrics for the existing languages.

Language	Recall@100
French	0.583
German	0.524
Spanish	0.483

🔧 Technical Details

The README does not provide in - depth technical details, so this section is skipped.

📄 License

The README does not provide license information, so this section is skipped.

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