odsbahia-ptbr

odsbahia/odsbahia-ptbr

<!-- Provide a quick summary of what the model is/does. -->

This model (ODSBahia-PTBR) serves to classify texts in relation to the 17 United Nations Sustainable Development Goals (SDGs), but it also covers the three new SDGs introduced in the document “2030 Agenda Guide: integrating SDGs, Education and Society”, prepared in 2020 in partnership between the University of Brasília (UnB) and the State University of São Paulo (UNESP,. "Racial Equality (SDG 18)", "Art, Culture and Communication (SDG 19)", and "Rights of Original Peoples and Traditional Communities (SDG 20)

Source:https://gtagenda2030.org.br/ Source:https://raizesds.com.br/pt/novos-ods/

Model Description

<!-- Provide a longer summary of what this model is. -->

This text classification model was developed by tuning the neuralmind/bert-base-portuguese-cased pre-trained model. The training data for this fine-tuned model was collected specifically for this task. This model was made as part of academic research at the Federal University of Southern Bahia in Brazil. The goal was to create a text classification model against the 20 SDGs based on Transformers that anyone could use. The main model details are highlighted below:

• Model type: Text classification
• Language(s) (NLP): Portuguese
• License: mit
• Finetuned from model [optional]: neuralmind/bert-base-portuguese-cased

Direct Use

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

This is a fine-tuned model and therefore does not require additional training.

How to Get Started with the Model

Use the code below to get started with the model.

from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("odsbahia/odsbahia-ptbr")
model = AutoModelForSequenceClassification.from_pretrained("odsbahia/odsbahia-ptbr")

Training Data

<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->

Training data includes texts from a wide variety of sources and academic research fields. Therefore, this adjusted model is not intended for a specific sector.

Training Hyperparameters

• SEED = 42
• MAX_LEN = 128
• BATCH_SIZE = 16
• NUM_EPOCHS = 34
• LEARNING_RATE = 9.46e-05
• WEIGHT_DECAY_RATE = 1.21e-05

Evaluation

Metrics

<!-- These are the evaluation metrics being used, ideally with a description of why. -->

• Accuracy = 0.98 sobre a base de dados (OSDG-CD)
• Precision = 0.89
• Recall = 0.70
• F1-Score = 0.79

Citation

Santos , Êmeris S., & Moraes , L. E. (2024). ODSBAHIA-PTBR: A Natural Language Processing Model to Support Sustainable Development Goals. Revista De Gestão Social E Ambiental, 18(12), e010230. https://doi.org/10.24857/rgsa.v18n12-039

<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->

Model Card Contact

[email protected]