Beto Sentiment Analysis Es

🚀 beto_sentiment_analysis_es Model

A finetuned model for sentiment analysis in Spanish, leveraging the BETO base model and trained on Spanish movie reviews.

🚀 Quick Start

This model is designed for sentiment analysis of Spanish text, especially movie reviews. It's fine - tuned based on the BETO model and trained using Amazon SageMaker and Hugging Face's Deep Learning container.

✨ Features

• Based on BETO: Utilizes the BETO model, a BERT - base model pre - trained on a Spanish corpus, trained with the Whole Word Masking technique.
• Multilingual Dataset: Trained on a balanced dataset of about 50,000 Spanish movie reviews, with reviews and labels available in both English and Spanish.
• High Performance: Achieves high scores in accuracy, F1 score, precision, and recall on the test dataset.

📦 Installation

No specific installation steps are provided in the original document.

💻 Usage Examples

Basic Usage

import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification

tokenizer = AutoTokenizer.from_pretrained("edumunozsala/beto_sentiment_analysis_es")
model = AutoModelForSequenceClassification.from_pretrained("edumunozsala/beto_sentiment_analysis_es")

text ="Se trata de una película interesante, con un solido argumento y un gran interpretación de su actor principal"

input_ids = torch.tensor(tokenizer.encode(text)).unsqueeze(0)
outputs = model(input_ids)
output = outputs.logits.argmax(1)

📚 Documentation

Model Information

Model Index

• Name: beto_sentiment_analysis_es
• Results:
  • Task:
    • Name: Sentiment Analysis
    • Type: sentiment - analysis
  • Dataset:
    • Name: "IMDb Reviews in Spanish"
    • Type: IMDbreviews_es
  • Metrics:
    • Name: Accuracy, Type: accuracy, Value: 0.9101333333333333
    • Name: F1 Score, Type: f1, Value: 0.9088450094671354
    • Name: Precision, Type: precision, Value: 0.9105691056910569
    • Name: Recall, Type: recall, Value: 0.9071274298056156

BETO Citation

Spanish Pre - Trained BERT Model and Evaluation Data

@inproceedings{CaneteCFP2020,
  title={Spanish Pre-Trained BERT Model and Evaluation Data},
  author={Cañete, José and Chaperon, Gabriel and Fuentes, Rodrigo and Ho, Jou - Hui and Kang, Hojin and Pérez, Jorge},
  booktitle={PML4DC at ICLR 2020},
  year={2020}
}

Dataset

The dataset consists of about 50,000 Spanish movie reviews. It is balanced and includes reviews and labels in both English and Spanish.

• Train dataset: 42,500
• Validation dataset: 3,750
• Test dataset: 3,750

Intended Uses & Limitations

This model is intended for sentiment analysis of Spanish text, especially movie reviews, but can also be applied to other types of reviews.

Hyperparameters

{
    "epochs": "4",
    "train_batch_size": "32",    
    "eval_batch_size": "8",
    "fp16": "true",
    "learning_rate": "3e-05",
    "model_name": "\"dccuchile/bert-base-spanish-wwm-uncased\"",
    "sagemaker_container_log_level": "20",
    "sagemaker_program": "\"train.py\""
}

Evaluation Results

• Accuracy: 0.9101333333333333
• F1 Score: 0.9088450094671354
• Precision: 0.9105691056910569
• Recall: 0.9071274298056156

📄 License

This model is released under the Apache 2.0 license.

Created by Eduardo Muñoz/@edumunozsala