README.md

metadata

license: apache-2.0

BERTić-Tweet

BERTić-Tweet is a version of the BERTić-Tweet-Base*, fine-tuned for sentiment analysis using 5,610 annotated Serbian COVID-19 vaccination-related tweets. Specifically, it is tailored for five-class sentiment analysis to capture finer sentiment nuances in the social media domain using the following scale: very negative, negative, neutral, positive, and very positive.

*BERTić-Tweet-Base is based on the BERTić model, which has been additionally pretrained for the social media domain and is licensed under the Apache 2.0 license.

How to Use

To use the model, you can load it with the following code:

from transformers import AutoTokenizer, ElectraForSequenceClassification

model_name = "DarijaM/BERTic-Tweet"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = ElectraForSequenceClassification.from_pretrained(model_name)