# PaReS-sentimenTw-political-PL

This model is a fine-tuned version of [dkleczek/bert-base-polish-cased-v1](https://huggingface.co/dkleczek/bert-base-polish-cased-v1) to predict 3-categorical sentiment.
Fine-tuned on 1k sample of manually annotated Twitter data.


Mapping (id2label):
mapping = {
    0:'negative', 
    1:'neutral', 
    2:'positive'
}



## Intended uses & limitations

Sentiment detection in Polish data (fine-tuned on tweets from political domain).


## Training and evaluation data

Trained for 3 epochs, mini-batch size of 8.
Training results: loss: 0.1358926964368792


## Evaluation procedure


It achieves the following results on the test set (10%):

  Num examples = 100
  Batch size = 8
  Accuracy = 0.950
  F1-macro = 0.944

              precision    recall  f1-score   support

           0      0.960     0.980     0.970        49
           1      0.958     0.885     0.920        26
           2      0.923     0.960     0.941        25

    accuracy                          0.950       100
   macro avg      0.947     0.941     0.944       100
weighted avg      0.950     0.950     0.950       100