bertin-roberta-base-zeroshot-esnli Open-source Model - Free for Spanish Text Classification Tasks

Bertin Roberta Base Zeroshot Esnli

Developed by somosnlp-hackathon-2022

This model is a zero-shot classifier trained on bertin-roberta-base-spanish, suitable for Spanish text classification tasks.

Text Classification

Transformers

Supports Multiple Languages#Spanish zero-shot classification #Natural language inference #Multi-label probability prediction

Downloads 54

Release Time : 4/4/2022

Model Overview

The model adopts a cross-encoder architecture, trained for natural language inference tasks, and can be used for zero-shot classification. Given a sentence and a set of candidate labels, the model outputs the probability of the sentence belonging to each label.

Model Features

Zero-shot classification capability

Can classify new categories without domain-specific training data

Spanish language support

Optimized specifically for Spanish text

Natural language inference architecture

Based on cross-encoder architecture, learning to predict 'contradiction', 'entailment', and 'neutral' relationships

Model Capabilities

Spanish text classification

Zero-shot learning

Natural language inference

Use Cases

Text classification

News categorization

Automatically categorize news articles into predefined topic categories

Social media content analysis

Analyze topic tendencies in Spanish social media posts

🚀 A zero-shot classifier based on bertin-roberta-base-spanish

This model, built on bertin-roberta-base-spanish, uses Cross encoder for NLI tasks and can perform zero-shot classifications via Hugging Face's pipeline.

🚀 Quick Start

This model can be used with Hugging Face's Zero - shot pipeline for zero - shot classifications. Given a sentence and a set of labels/topics, it will output the probability of the sentence belonging to each topic.

💻 Usage Examples

Basic Usage

The simplest way to use the model is through the huggingface transformers pipeline tool. Initialize the pipeline by specifying the task as "zero - shot - classification" and select "hackathon - pln - es/bertin - roberta - base - zeroshot - esnli" as the model.

from transformers import pipeline
classifier = pipeline("zero-shot-classification", 
                       model="hackathon-pln-es/bertin-roberta-base-zeroshot-esnli")

classifier(
    "El autor se perfila, a los 50 años de su muerte, como uno de los grandes de su siglo",
    candidate_labels=["cultura", "sociedad", "economia", "salud", "deportes"],
    hypothesis_template="Esta oración es sobre {}."
)

⚠️ Important Note

The hypothesis_template parameter is important and should be in Spanish. In the widget on the right, this parameter is set to its default value: "This example is {}.", so different results are expected.

🔧 Technical Details

Training

We used sentence - transformers to train the model.

Dataset

We used a collection of Natural Language Inference datasets as training data:

ESXNLI, only the Spanish part
SNLI, automatically translated
MultiNLI, automatically translated

The entire dataset used is available [here](https://huggingface.co/datasets/hackathon - pln - es/nli - es).

📚 Documentation

Authors

Property	Details
Pipeline Tag	zero - shot - classification
Tags	zero - shot - classification, nli
Language	es
Datasets	hackathon - pln - es/nli - es

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