sagorsarker/bangla-bert-sentiment

The model is a pretrained model for bengali Sentiment Analysis using bangla-bert-base model . The model was trained with two combined datasets . It evaluated with 2178 sentences and F1 score of 0.3289 . The evaluation result is based on sentiment from the given sentence . The data was analysed using a simple binary classification script with total of 3 epochs in google colab gpu. The results are based on the data from two datasets. The model has been trained with simple transformers binary classification scripts. It has been used to calculate sentiment from given sentence. The data set includes a total of 2,000 sentences. The number of sentences that were evaluated. The,