# Portuguese NLP

Transformer model based on Distilbert architecture, specializing in Brazilian Portuguese sentiment analysis

A named entity recognition model for the Portuguese legal domain, compliant with Brazil's LGPD data protection law, specializing in identifying sensitive information in legal texts.

A Portuguese large language model fine-tuned based on Mistral-7B-Instruct-v0.2, focusing on improving Portuguese text generation capabilities

A hate speech classifier fine-tuned on the BERT-large Portuguese model, achieving 78.14% accuracy on the hate_speech_portuguese dataset

Portuguese zero-shot classification model based on DeBERTa-v3 architecture, suitable for performing text classification tasks without fine-tuning

A Portuguese POS tagger fine-tuned from BERTimbau model, trained on MacMorpho corpus with an F1 score of 0.9826.

Stanza is a precise and efficient multilingual text analysis toolkit that supports Portuguese text processing, syntactic parsing, and entity recognition.

This is a pre-trained language model based on the RoBERTa architecture, specifically optimized for Brazilian Portuguese.

A CPU-optimized Portuguese processing pipeline, including tokenization, POS tagging, dependency parsing, named entity recognition, and more.

A sentiment analysis model fine-tuned on social media brand datasets based on BERTweet

A Portuguese archive retrieval assistant model fine-tuned based on bert-base-multilingual-cased, mainly used for token classification tasks

spaCy's CPU-optimized Portuguese processing pipeline, including tokenization, POS tagging, dependency parsing, named entity recognition, etc.

This is a distilled Portuguese BERT model derived from BERTimbau, retaining 99% of the original model's accuracy.

A fine-tuned version of the small Portuguese GPT-2 model, specifically optimized for the TCU Acordaos dataset